Human catalase: looking for complete identity

Editor-in-ChiefZihe Rao Nankai University, China

Deputy Editors-in-ChiefLe Kang Beijing Institutes of Life Science, Chinese Academy of Sciences, ChinaGeorge Fu Gao Beijing Institutes of Life Science, Chinese Academy of Sciences, ChinaRui-Ming Xu Institute of Biophysics, Chinese Academy of Sciences, ChinaSteven Y Cheng Nanjing Medical University, ChinaBob Sim University of Oxford, UK

Board MembersZhiqiang An University of Texas Health Science Center at Houston, USATed Baker University of Auckland, New ZealandJoan Bennett Rutgers University, USALeszek Borysiewicz Medical Research Council, UKHsiao Chang Chan Chinese University of Hong Kong, ChinaTse Wen Chang Academia Sinica, China Gursharan Singh Chhatwal Helmholtz Centre for Infection Research, Germany

Peter Colman Walter and Eliza Hall Institute, AustraliaZixin Deng Shanghai Jiao Tong University, ChinaHaian Fu Emory University, USAWei Gu Columbia University, USAJiahuai Han Xiamen University, ChinaShigang He Institute of Biophysics, Chinese Academy of Sciences, ChinaXi He Harvard University, USAZhigang He Harvard University, USARolf Hilgenfeld University of Lübeck, GermanyGeoff Howlett The University of Melbourne, Australia

Li Huang Institute of Microbiology, Chinese Academy of Sciences, ChinaNeil Isaacs University of Glasgow, UKAikichi Iwamoto University of Tokyo, JapanLi Jin Fudan University, ChinaAndrzej Joachimiak Argonne NationalLaboratory, USARobert Kaptein Utrecht University, The NetherlandsYoshihiro Kawaoka University of Wisconsin-Madison, USADaniel Klionsky University of Michigan, USAAdrian R. Krainer Cold Spring Harbor, USAChia-Wei Li Tsing Hua University, ChinaJinhua Lu National University of Singapore, SingaporeKeping Ma Institute of Botany, Chinese Academy of Sciences, ChinaBrian Matthews University of Oregon, USAAndrew McMichael University of Oxford, UKAnming Meng Institute of Zoology, Chinese Academy of Sciences, ChinaKunio Miki Kyoto University, JapanYi Rao Peking University, China

Beijing Institutes of Life ScienceChinese Academy of Sciences

Biophysical Society of China

Managing Editor Xiaoxue ZhangBeijing Institutes of Life ScienceChinese Academy of Sciences1 Beichen West Road, ChaoyangBeijing 100101, ChinaTel: +86-10-64862453Fax: +86-10-64880586Email: [email protected]

Higher Education Press4 Huixindongjie, ChaoyangBeijing 100029, ChinaTel: +86-10-58551834Fax: +86-10-58556034Email: [email protected]

Sponsored by

Ariel Ruiz i Altaba University of Geneva, SwitzerlandYigong Shi Tsinghua University, ChinaHong-Bing Shu Wuhan University, ChinaTimothy Springer Harvard University, USADavid Stuart University of Oxford, UKPeiqing SunThe Scripps Research Institute, USAJoel Sussman Weizmann Institute of Science, IsraelHong Tang Institute of Biophysics, Chinese Academy of Sciences, ChinaZhigang Tian University of Science and Technology of China, ChinaHans Vliegenthart Utrecht University, The NetherlandsJia-huai Wang Harvard University, USAPeng Wang Nankai University, China Xiaodong Wang University of Texas Southwestern Medical Center, USADick Wettenhall The University of Melbourne, AustraliaLuet Wong University of Oxford, UKChung-I Wu Beijing Institute of Genomics, Chinese Academy of Sciences, China

Jane Wu Northwestern University, USATao Xu Institute of Biophysics, Chinese Academy of Sciences, ChinaYongbiao Xue Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, ChinaGen Yamada Kumamoto University, JapanTadashi Yamamoto University of Tokyo, JapanLi Yu Tsinghua University, China Kwok-yung Yuen The University of Hong Kong, China Chenyu Zhang Nanjing University, China Kan Zhang Institute of Psychology, Chinese Academy of Sciences, ChinaHong ZhangNational Institute of Biological Sciences, ChinaMingjie Zhang The Hong Kong University of Science & Technology, ChinaYi Zhang University of North Carolina at Chapel Hill,USAYixian Zheng Carnegie Institution of Washington, USAWeimin Zhong Yale University, USAYang Zhong Fudan University, ChinaYi Zhong Cold Spring Harbor, USA

Editorial Board

Protein & Cell

NEWS AND VIEWSChemotaxis: new role for Ras revealedJianshe Yan, Dale Hereld, Tian Jin

PERSPECTIVE Phosphorylation of Rictor at Thr1135 impairs the Rictor/Cullin-1 complex to ubiquitinate SGK1Daming Gao, Lixin Wan, Wenyi Wei

RECOLLECTION

Commitment and dedication of a Chinese plant physiologistTingyun Kuang, Ming Li, Le Kang

REVIEWS Human catalase: looking for complete identityMadhur M. Goyal, Anjan Basak

Wnt pathway antagonists and angiogenesisBin Zhang, Jian-xing Ma

The late stage of autophagy: cellular events and molecular regulationJingjing Tong, Xianghua Yan, Li Yu

microRNAs: tiny RNA molecules, huge driving forces to move the cellShenglin Huang, Xianghuo He

COMMUNICATION A transcription assay for EWS oncoproteins in Xenopus oocytesKing Pan Ng, Felix Cheung, Kevin A.W. Lee

879

881

886

888

898

907

916

927

916

888

881

935

Protein & Cell

935

944

956

COVERMutations in the gene encoding gap junction protein connexin-31 (Cx31) are associated with hearing impairment (HI). Here, we show that expression of Cx31 in the mouse inner ear is developmentally regulated with a high level in adult inner hair cells and spiral ganglion neurons that are critical for the hearing process. In transfected cells, wild type Cx31 protein (Cx31wt) forms functional gap junction at cell-cell-contacts. In contrast, two HI-associated Cx31 mutants, Cx31R180X and Cx31E183K resided primarily in the ER and Golgi-like intracellular punctate structures, respectively, and failed to mediate lucifer yellow transfer. Expression of Cx31 mutants but not Cx31wt leads to upregulation of and increased association with the ER chaperone BiP indicating ER stress induction. Together, the HI-associated Cx31 mutants are impaired in trafficking, promote ER stress, and hence lose the ability to assemble functional gap junctions. The study reveals a potential pathological mechanism of HI-associated Cx31 mutations.

RESEARCH ARTICLES Trafficking abnormality and ER stress underlie functional deficiency of hearing impairment-associated connexin-31 mutantsKun Xia, Hong Ma, Hui Xiong, Qian Pan, Liangqun Huang, Danling Wang, Zhuohua Zhang

Interaction of Hsp40 with influenza virus M2 protein: implications for PKR signaling pathwayZhenhong Guan, Di Liu, Shuofu Mi, Jie Zhang, Qinong Ye, Ming Wang, George F. Gao, Jinghua Yan

DEXH-Box protein DHX30 is required for optimal function of the zinc-finger antiviral proteinPeiying Ye, Shufeng Liu, Yiping Zhu, Guifang Chen, Guangxia Gao

935

NEWS AND VIEWS

Chemotaxis: new role for Ras revealedJianshe Yan1✉, Dale Hereld2, Tian Jin1

1 Chemotaxis Signal Section, Laboratory of Immunogenetics, National Institute of Allergy and Infectious Diseases, NationalInstitutes of Health, Rockville, MD 20852, USA

2 Division of Metabolism and Health Effects, National Institute on Alcohol Abuse and Alcoholism, NIH, 5635 Fishers Lane,Rockville, MD 20852, USA

✉ Correspondence: [email protected]

A recent study of chemotaxis revealed a new role for theproto-oncogene Ras in the social ameba Dictyosteliumdiscoideum. Chemotaxis, the directional movement of cellstoward chemokines and other chemoattractants, plays criticalroles in diverse physiological processes, such as mobilizationof immune cells to fight invading microorganisms, targeting ofmetastatic cancer cells to specific tissues, and guidance ofsperm cells to ova during fertilization. This work, published inthe July 26 issue of The Journal of Cell Biology, wasconducted in Dr. Devreotes’ lab at John Hopkins Universityand Dr. Parent’s lab at National Cancer Institute. Thisresearch team demonstrated that RasC functions as anupstream regulator of TORC2 and thereby governs theeffects of TORC2-PKB signaling on the cytoskeleton andcell migration.

Many of the core components of the underlying chemotaxissignaling network have been elucidated in D. discoideum.Chemoattractants are sensed by G-protein-coupled receptors(GPCRs), which leads to the activation heterotrimeric G-proteins, small Ras-like G-proteins, and phosphoinsositide 3-kinase (PI3K), resulting in the generation of phosphatidylino-sitol-(3,4,5)-trisphosphate (PIP3). This phospholipid, in turn,prompts the membrane translocation of proteins containingpleckstrin homology (PH) domains, such as cytosolicregulator of adenylyl cyclase (CRAC) and protein kinase B(PKB), which regulate the cytoskeleton rearrangementsduring chemotaxis. Importantly, many of these componentsand their activation are highly localized to the leading edge ofcells undergoing chemotaxis, assuring that cytoskeletalchanges needed for directional movement are spatiallyrestricted (reviewed in Jin et al., 2009).

Although it has been well established that the PIP3 pathwayplays an important role in the regulation of chemotaxis,additional pathways that act in parallel with the PIP3 pathwayhave recently been revealed. For instance, phospholipase A2(PLA2) was reported to mediate chemotaxis in parallel with

PIP3 pathway (Chen et al., 2007). In addition, a PIP3-independent pathway in which PKB is activated by TORC2(target of rapamycin complex 2) was found to regulatechemotaxis (Lee et al., 2005; Kamimura et al., 2008).However, the mechanism by which TORC2 is regulated inchemotaxis was poorly understood prior to the publication ofthe recent report by Cai et al. (2010).

D. discoideum possesses two PKB homologs, namelyPKBA, which contains a PH domain and is dynamicallyrecruited to the plasma membrane by PIP3, and PKBR1,which is tethered to the plasma membrane via N-terminalmyristoylation. In their previous work, the authors discoveredthat both PKBA and PKBR1 are activated by TORC2-mediated phosphorylation of their hydrophobic motifs (HMs)(Kamimura et al., 2008) and phosphoinositide-dependentkinase (PDK)-mediated phosphorylation of their activationloops (ALs) (Kamimura and Devreotes, 2010). In the presentstudy, Cai and colleagues investigated whether Ras familyproteins activate TORC2 and, if so, what are the effects of thisRas-TORC2 pathway on chemotactic responsiveness.

In order to examine whether Ras proteins are required forTORC2-mediated activation of PKB, the scientists firstdetermined the PKB activity in different Ras knock-out cells.They found that phosphorylation of the HM of PKBR1, the ALsof both PKBR1 and PKBA, and many PKB substrates weresignificantly reduced in rasC− but not rasG− cells relative towild-type cells. To further explore rasC's role in PKBactivation, the authors examined the consequences ofexpressing activated RasC (RasCQ62L) and found that itdramatically prolonged the phosphorylation kinetics of thePKBR1 and multiple PKB substrates, suggesting that RasC isindeed involved in regulating the PKB pathway. In addition,RasCQ62L expression also prolonged actin polymerizationand impaired chemotaxis in wild-type cells, effects that weresuppressed in cells lacking Pianissimo (piaA−), an essentialcomponent of TORC2. Taken together these findings

© Higher Education Press and Springer-Verlag Berlin Heidelberg 2010 879

Protein Cell 2010, 1(10): 879–880DOI 10.1007/s13238-010-0119-6

Protein & Cell

represent the first genetic evidence that RasC functionsupstream of TORC2 to regulate PKBR1 and chemotaxis.

The finding that RasC is required for the activation of thePKB pathway in vivo encouraged the scientists to furtherelucidate the function of RasC in TORC2-PKB pathwayactivation in vitro. They demonstrated that PKB phosphoryla-tion can be reconstituted by mixing immunopurified TORC2with membranes containing RasCQ62L (but not those contain-ing inactive wild-type RasC). In addition, the authors foundthat TORC2 specifically co-immunoprecipitates with activatedRasC but not inactive RasC, suggesting that RasC andTORC2 physically interact in a regulated fashion, furthersupporting the conclusion that activated RasC stimulatesTORC2 activity.

In summary, this study by Cai and colleagues reveals anovel RasC-TORC2-PKB signaling pathway with importantroles in chemotaxis and suggests that RasC interacts directlywith TORC2. Presumably RasC interacts with the Ras-binding domain of Rip3, a component of D. discoideumTORC2 and an ortholog of mammalian TORC2 subunitmSin1. These findings are complemented by a recent reportfrom Charest et al. (2010) demonstrating the existence of anovel RasGEF-containing complex that translocates to theplasma membrane upon chemoattractant stimulation, pro-motes activation of RasC, and is the target of PKB-mediatednegative feedback that apparently terminates RasC activa-tion. Together, these studies provide novel mechanisticinsights into the activation of TORC2 and its regulation inchemotaxis. The findings that RasC activation dictates thekinetics of downstream signaling events and the demonstra-tion of PKB-mediate feedback regulation of RasC suggestthat RasC is a key regulatory node in the chemotaxis signalingnetwork and may be critical for the exquisite temporal andspatial regulation of the cytoskeleton that occurs in chemotaxis.

TORC2 is also known to mediate cell proliferation and

survival and thus is being aggressively pursued as a target fornovel cancer therapeutics. However, prior to this study, littlewas known about how TORC2 is activated. Therefore, itseems likely that the regulatory information about the RasCand TORC2 discovered here inD. discoideumwill lead to newtherapeutic strategies for treatment of inflammatory disordersand possibly also cancer.

REFERENCES

Cai, H., Das, S., Kamimura, Y., Long, Y., Parent, C.A., and Devreotes,P.N. (2010). Ras-mediated activation of the TORC2-PKB pathwayis critical for chemotaxis. J Cell Biol 190, 233–245.

Charest, P.G., Shen, Z., Lakoduk, A., Sasaki, A.T., Briggs, S.P., andFirtel, R.A. (2010). A Ras signaling complex controls the RasC-TORC2 pathway and directed cell migration. Dev Cell 18,

737–749.

Chen, L., Iijima, M., Tang, M., Landree, M.A., Huang, Y.E., Xiong, Y.,Iglesias, P.A., and Devreotes, P.N. (2007). PLA2 and PI3K/PTEN

pathways act in parallel to mediate chemotaxis. Dev Cell 12,603–614.

Jin T, Xu X, Fang J, Isik N, Yan J, Brzostowski JA, Hereld D. (2009).

How human leukocytes track down and destroy pathogens:lessons learned from the model organism Dictyostelium discoi-deum. Immunol Res 43, 118–27.

Kamimura, Y., Xiong, Y., Iglesias, P.A., Hoeller, O., Bolourani, P., andDevreotes, P.N. (2008). PIP3-independent activation of TorC2 andPKB at the cell's leading edge mediates chemotaxis. Curr Biol 18,

1034–1043.

Kamimura, Y., and Devreotes, P.N. (2010). Phosphoinositide-depen-

dent protein kinase (PDK) activity regulates phosphatidylinositol3,4,5-trisphosphate-dependent and -independent protein kinase Bactivation and chemotaxis. J Biol Chem 285, 7938–7946.

Lee, S., Comer, F.I., Sasaki, A., McLeod, I.X., Duong, Y., Okumura,K., Yates, J.R., Parent, C.A., and Firtel, R.A. (2005). TOR complex2 integrates cell movement during chemotaxis and signal relay in

Dictyostelium. Mol Biol Cell 16, 4572–4583.

880 © Higher Education Press and Springer-Verlag Berlin Heidelberg 2010

Jianshe Yan et al.Protein & Cell

PERSPECTIVE

Phosphorylation of Rictor at Thr1135 impairsthe Rictor/Cullin-1 complex to ubiquitinateSGK1Daming Gao*, Lixin Wan*, Wenyi Wei✉

Department of Pathology, Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA 02215, USA✉ Correspondence: [email protected]

ABSTRACT

The Rictor/mTOR complex plays a pivotal role in a varietyof cellular functions including cellular metabolism, cellproliferation and survival by phosphorylating Akt atSer473 to fully activate the Akt kinase. However, itsupstream regulatory pathways as well as whether it hasadditional function(s) remain largely unknown. Werecently reported that Rictor contains a novel ubiquitinE3 ligase activity by forming a novel complex withCullin-1, but not with other Cullin family members.Furthermore, we identified SGK1 as its downstreamtarget. Interestingly, Rictor, but not Raptor or mTOR,promotes SGK1 ubiquitination. As a result, SGK1expression is elevated in Rictor–/– MEFs. We furtherdefined that as a feedback mechanism, Rictor can bephosphorylated by multiple AGC family kinases includ-ing Akt, S6K and SGK1. Phosphorylation of Rictor at theThr1135 site did not affect its kinase activity towardsphosphorylating its conventional substrates includingAkt and SGK1. On the other hand, it disrupted theinteraction between Rictor and Cullin-1. Consequently,T1135E Rictor was defective in promoting SGK1 ubiqui-tination and destruction. This finding further expands ourknowledge of Rictor’s function. Furthermore, our workalso illustrates that Rictor E3 ligase activity could begoverned by specific signaling kinase cascades, and thatmisregulation of this process might contribute to SGKoverexpression which is frequently observed in varioustypes of cancers.

KEYWORDS mTORC2, Rictor, kinase, E3 ubiqutinligase, Cullin-1, SGK1

RICTOR-CONTAINING MTORC2 COMPLEX IS A

SER/THR PROTEIN KINASE

Mammalian target of rapamycin (mTOR), belongs to theserine/threonine kinase of the phosphatidylinositol kinase-related family, and is conserved in a broad range of eukaryoticorganisms (Sabatini, 2006; Wullschleger et al., 2006). It wasrecently established that the mTOR kinase plays a critical rolein a wide spectra of cellular homeostasis, cellular growth andsurvival pathways. This is achieved mainly by acting as asensor mechanism to receive upstream inputs from multiplegrowth-promoting signals and then transducing them to thedownstream effector signaling pathways (Reiling and Saba-tini, 2006; Guertin and Sabatini, 2007). By associating withdifferent subset of components, the mTOR kinase couldassemble into at least two different complexes that are termedas mTOR complex 1 (mTORC1) and mTOR complex 2(mTORC2) (Reiling and Sabatini, 2006; Guertin and Sabatini,2007). The mTORC1 complex contains the subunits mTOR,Raptor, PRAS40 and mLST8/GβL, and was reported tophosphorylate a faculty of downstream targets including S6Kand the 4E-BP protein. Phosphorylation by mTORC1 leads toactivation of S6K and inhibition of 4E-BP, thereby augmentingthe protein translation process (Yang and Guan, 2007).mTORC1 function has also been implicated in regulating cellgrowth and autophagy. On the other hand, the mTORC2complex is composed of mTOR, Rictor, mLST8/GβL,PRTOR/PRR5 and Sin1 (Fig. 1A) (Jacinto et al., 2006; Shiotaet al., 2006). Unlike mTORC1, the mTORC2 complex isrelatively insensitve to short-term rapamycin treatment,although more recent work demonstrated that prolongedrapamycin treatment also dissociated the mTORC2 complex

*These authors contributed equally to this work.


Protein Cell 2010, 1(10): 881–885DOI 10.1007/s13238-010-0123-x

Protein & Cell

(Sarbassov et al., 2006). The assembled mTORC2 complexphosphorylates the hydrophobic motif of Akt (Ser473)(Sarbassov et al., 2005) and SGK1 (Ser422) (Garcia-Martinez and Alessi, 2008), leading to the full activation ofthese two kinases. Since aberrant activation of Akt is ahallmark of many types of cancers (Manning and Cantley,2007), hyperactivation of mTORC2 activity has beenimplicated in cancer progression (Guertin and Sabatini,2007).

Previous work clearly demonstrated that mTORC1 activityis tightly controlled in vivo. In response to growth factors,activation of the PI3K kinase pathway leads to subsequentphosphorylation of the TSC2 and PRAS40 proteins by Akt,thus leading to activation of the mTORC1 kinase (Manningand Cantley, 2007). Recent work demonstrated that theactivity of mTORC1 can also be stimulated by the RagGTPase in response to nutrient stimulation (Sancak et al.,2008). In addition to the positive regulation, mTORC1 is alsosubject to layers of negative regulation. For example,phosphorylation of Raptor by AMP-activated protein kinase(AMPK) in response to low energy state represses mTORC1activity (Gwinn et al., 2008). Recent studies also revealed thatRaptor could form a complex with Cullin-4 and that theproteasome pathway might be critical for its kinase activity(Ghosh et al., 2008). This work indicated that Raptor might bemulti-functional, which contains activities other than its kinaseactivity. Compared to the accumulated knowledge ofmTORC1 regulation, very little is known so far regarding theregulatory mechanism for activation of mTORC2, althoughit is a critical upstream pathway governing Akt1 activity.Unlike Raptor, which is shown to be a WD40-repeat contain-ing protein, Rictor contains no characterized functionaldomain. However, Rictor was reported to associate withthe Cullin 4 complex (Ghosh et al., 2008) as well. Never-theless, the additional function(s) of Rictor remains largelyunknown.

RICTOR FORMS A NOVEL E3 LIGASE COMPLEX

WITH CULLIN-1 AND RBX1, AND

PHOSPHORYLATION OF RICTOR AT T1135

IMPAIRS ITS E3 LIGASE ACTIVITY

We obtained experimental evidence for a novel function of theRictor protein, which is promoting the ubiquitination anddestruction of SGK1 (Gao et al., 2010). Thus far the majorresearch focus of Rictor has been in identifying its down-stream phosphorylation targets, and how this affects theirfunctions. Although Rictor has been indicated to be asso-ciated with Cullin-4 (Ghosh et al., 2008), its role as an E3ligase has not been formally presented. We found that byspecific association with Cullin 1 and Rbx1, Rictor can form afunctional E3 ubiquitin ligase complex (Fig. 1B) (Gao et al.,2010). We further defined that this activity of Rictor does not

involve mTOR and other mTORC2 components such as Sin-1and GβL. This indicates that the kinase activity of Rictor andthe E3 ligase activity of Rictor could possibly be separated.Consistent with this notion, we showed that phosphorylationof Rictor at T1135 minimally affects the kinase activity ofRictor, but severely diminished its E3 ligase activity (Fig. 2). Inan effort to further define the minimal motif present in Rictorcritical for its kinase or E3 ligase activity, we found that both N-terminus and C-terminus of Rictor is required for properinteraction with mTOR (Huang et al., 2009; Gao et al., 2010).This indicates that the formation of the multi-componentmTORC2 complex requires an intact Rictor protein. On theother hand, although deletion of neither the N-terminusnor the C-terminus of Rictor affects its interaction with Cullin1, the resulting Rictor/Cullin 1 complex is defective inpromoting SGK1 ubiquitination (Gao et al., 2010). It ishighly possible that it requires the presence of both the N-terminus and C-terminus to maintain the Cullin/Rictor com-plex in an active conformation. However, more studies areneeded to fully understand the structural and functionaldifference and relationship between the kinase and E3 ligaseactivity of Rictor, and to elucidate the other possiblecomponents of the Rictor/Cullin-1 complex (Fig. 1A and1B).

It seems likely that by associating with mTOR, mLST8/GβL, PRTOR/PRR5 and Sin1, Rictor can form a kinasecomplex. On the other hand, Rictor can interact with Cullin-1,Rbx1 and other unknown factors to form an E3 ligasecomplex. Although we found that mTOR is not involved in itsE3 ligase activity, it still remains unclear whether these twocomplexes share some common scaffolding proteins andwhat the crosstalk is between these two complexes. Recentstudies revealed that many other kinases possess E3 ligaseactivity as well. For example, MEKK1 acts as an E3 ligase topromote ERK1 destruction (Lu et al., 2002), while the DYRK2protein can form an E3 ligase to promote katanin p60ubiquitination (Maddika and Chen, 2009). Since manyubiquitination processes require a prior phosphorylationevent for efficient substrate recognition, it has been shownthat by coupling both the kinase and the E3 ligase activity inadjacency provides for more efficient destruction. Forexample, the Skp2 E3 ligase is in tight association with theCyclin A/Cdk2 complex and this is thought to promote thedestruction of its substrate, p27, which is required to bephosphorylated by Cyclin A/Cdk2 (Carrano et al., 1999).However, we found that the phosphorylation of SGK1 byRictor at S422 did not serve to promote its subsequentubiquitination by Rictor (data not shown). It is possible thatthere are unknown Rictor ubiquitination substrate(s) thatrequire prior phosphorylation by Rictor. It is thus critical todefine Rictor ubiquitin substrates other than SGK1, which willhelp further reveal the in vivo physiological role of Rictor forcell growth and development.


Daming Gao et al.Protein & Cell

CROSSTALK BETWEEN SGK1 AND RICTOR SETS

UP A POSITIVE FEEDBACK LOOP ALLOWING

SGK1 ACCUMULATION AFTER MITOGENIC

STIMULATION

Furthermore, we identified that a variety of AGC family ofkinases including SGK1, Akt and S6K could potentially affect

the novel E3 ligase activity of Rictor by disrupting the Rictor/Cullin-1 complex through phosphorylation of Rictor at theT1135 site. On one hand, this provides another piece ofevidence about the complex regulatory mechanisms thatcontrol the mTOR pathway. On the other hand, it furtherindicates that besides activating Akt and SGK1 throughcomplexing with mTOR, Rictor also regulates cell signalingpathways by controlling the stability of key signaling

Figure 1. By association with a distinct subset of co-factors, Rictor can acquire either kinase or E3 ubiquitin ligaseactivity. (A) In association with mTOR, Sin1 and GβL, Rictor could form the mTORC2 complex, which is a Ser/Thr protein kinasecomplex that could phosphorylate Akt at Ser473 and SGK at Ser422, leading to their full activation. (B) Rictor could form a novel

complex with Cullin-1 and Rbx1 to promote SGK1 ubiquitination. It remains unclear whether Skp1, F-box or other unidentifiedcomponents are also involved in SGK1 ubiquitination.

Figure 2. Phosphorylation of Rictor at Thr1135 minimally affects the mTORC2 kinase activity, but severely impairs the E3ligase activity of the Rictor/Cullin-1 complex. In low AGC kinase activity conditions (such as serum starvation condition),

mTORC2 activity is low. As a result, neither Akt nor SGK is in its activated state, and the majority of Rictor is not phosphorylated atThr1135. Non-phosphorylated Rictor forms a complex with Cullin-1 and Rbx-1 to promote SGK1 ubiquitination and degradation. Inhigh AGC kinase activity conditions, mTORC2 complex activity is high, leading to phosphorylation of Akt at Ser473 and SGK1 atSer422. This modification fully activates both Akt and SGK1, both of which are capable of phosphorylating Rictor at Thr1135.

Phosphorylation of Rictor at Thr1135 dissociates Rictor from Cullin-1, thus leading to inactivation of the Rictor/Cullin-1 E3 ligasecomplex and stabilization of the SGK1 protein.


Rictor/Cullin 1 complex ubiquitinates SGK1 Protein & Cell

components like SGK1 (Gao et al., 2010). SGK1 activity hasbeen demonstrated previously to be activated by the PI3Kkinase pathway (Park et al., 1999; Loffing et al., 2006).Although SGK1 has been shown to be degraded by other E3ligases including Nedd4-2 and CHIP, it remains largelyunknown how the destruction of SGK1 by Nedd4-2 or CHIPis regulated and whether this process is related to PI3Kkinase activity. Our results offer a novel point of view for theregulation of SGK1 in the setting of the PI3K kinase signalingpathway. SGK1 is very unstable in normal conditions, andupon serum or growth factor induction, acute activation of thePI3K kinase pathway, which subsequently activates the Aktand S6K kinases, might also contribute to the induction ofSGK1 stability by disrupting the Rictor/Cullin-1 complex.Furthermore, since SGK1 could also negatively regulateRictor E3 ligase activity without affecting its kinase activity,this results in a positive feedback loop to boost the SGK1activity shortly post-stimulation. However, more studiesshould be performed to investigate whether the ability ofNedd4-2 and CHIP to degrade SGK1 is also affected by thePI3K/Akt pathway.

MISREGULATION OF RICTOR/CULLIN-1-

MEDIATED SGK1 DESTRUCTION MIGHT

CONTRIBUTE TO FREQUENT SGK1

OVEREXPRESSION IN CANCERS

Our results further implicate that aberrant activation of thePI3K/Akt pathway, which is a hallmark of many types ofcancers, might contribute to impair Rictor/Cullin-1-mediatedubiquitination of SGK1, thus contributing to the frequentelevation of SGK1 in many cancers. There are many reportedways to activate the PI3K/Akt pathway, which include loss-of-function mutations in the PTEN tumor suppressor, as well asgain-of-function mutations in upstream regulators such as thereceptors HER2, EGF-R and Ras (Majumder and Sellers,2005). In addition, there is also a high frequency ofconstitutively active PI3K mutations identified in severalhuman cancers (Samuels et al., 2004). We found thatdepletion of endogenous PTEN resulted in elevated phos-phorylation of Rictor at Thr1135, presumably through activa-tion of either the Akt or S6K kinase. Subsequently, we foundelevated SGK1 expression (Gao et al., 2010). SGK1 over-expression has been reported in multiple cancers includingbreast cancer. Interestingly, although SGK1 has beensuggested to have redundant functions with Akt, simulta-neous overexpression of both Akt1 and SGK1 has beenreported in many breast cancers (Sahoo et al., 2005). Ourwork suggested that since SGK1, but not Akt, is subject to theubiquitination by the Rictor/Cullin-1 complex, elevated Aktactivity might block the function of Rictor/Cullin-1, leading toaccumulation of SGK1. Thus, our work implicates a novelregulatory pathway for SGK1, which might provide a

mechanistic explanation for the frequent overexpression ofSGK1 in multiple cancers.

Collectively, our results provide novel insight into howRictor governs SGK1 abundance and activity by promoting itsubiquitination and subsequent destruction. We further definea feedback mechanism that could terminate the E3 ligaseactivity of the Rictor/Cullin-1 complex through phosphoryla-tion of Rictor at the T1135 site to dissociate Cullin-1 and Rictorinteraction. On one hand, our finding provides novelfunctional insight for Rictor, a key regulator of cell metabolismand cell growth. On the other hand, we also offer a novelmechanistic insight into SGK1 stability controlled by thePTEN/PI3K/Akt signaling pathway, which are hotspot muta-tion targets in cancers, and how misregulation of this processwill possibly contribute to SGK1 overexpression in cancers.

ACKNOWLEDGEMENTS

We thank Alan Lau, Hiroyuki Inuzuka and Pengda Liu for criticalreading of the manuscript, and other members of the Wei laboratoryfor useful discussions. Wenyi Wei is a Kimmel Scholar, V Scholar and

Karin Grunebaum Cancer Research Foundation Fellow. This workwas supported in part by the DOD Prostate New Investigator award toW.W., and NIH grant GM089763 to W.W.

REFERENCES

Carrano, A.C., Eytan, E., Hershko, A., and Pagano, M. (1999). SKP2is required for ubiquitin-mediated degradation of the CDK inhibitorp27. Nat Cell Biol 1, 193–199.

Gao, D., Wan, L., Inuzuka, H., Berg, A.H., Tseng, A., Zhai, B., Shaik,S., Bennett, E., Tron, A.E., Gasser, J.A., et al. (2010). Rictor formsa complex with Cullin-1 to promote SGK1 ubiquitination and

destruction. Mol Cell 39, 797–808.

Garcia-Martinez, J.M., and Alessi, D.R. (2008). mTOR complex 2

(mTORC2) controls hydrophobic motif phosphorylation andactivation of serum- and glucocorticoid-induced protein kinase 1(SGK1). Biochem J 416, 375–385.

Ghosh, P., Wu, M., Zhang, H., and Sun, H. (2008). mTORC1 signalingrequires proteasomal function and the involvement of CUL4-DDB1ubiquitin E3 ligase. Cell Cycle 7, 373–381.

Guertin, D.A., and Sabatini, D.M. (2007). Defining the role of mTOR incancer. Cancer Cell 12, 9–22.

Gwinn, D.M., Shackelford, D.B., Egan, D.F., Mihaylova, M.M., Mery,A., Vasquez, D.S., Turk, B.E., and Shaw, R.J. (2008). AMPKphosphorylation of raptor mediates a metabolic checkpoint. MolCell 30, 214–226.

Huang, J., Wu, S., Wu, C.L., and Manning, B.D. (2009). Signalingevents downstream of mammalian target of rapamycin complex 2

are attenuated in cells and tumors deficient for the tuberoussclerosis complex tumor suppressors. Cancer Res 69, 6107–6114.

Jacinto, E., Facchinetti, V., Liu, D., Soto, N., Wei, S., Jung, S.Y.,Huang, Q., Qin, J., and Su, B. (2006). SIN1/MIP1 maintains rictor-mTOR complex integrity and regulates Akt phosphorylation and

substrate specificity. Cell 127, 125–137.

Loffing, J., Flores, S.Y., and Staub, O. (2006). Sgk kinases and their


Daming Gao et al.Protein & Cell

role in epithelial transport. Annu Rev Physiol 68, 461–490.

Lu, Z., Xu, S., Joazeiro, C., Cobb, M.H., and Hunter, T. (2002). The

PHD domain of MEKK1 acts as an E3 ubiquitin ligase andmediates ubiquitination and degradation of ERK1/2. Mol Cell 9,945–956.

Maddika, S., and Chen, J. (2009). Protein kinase DYRK2 is a scaffoldthat facilitates assembly of an E3 ligase. Nat Cell Biol 11, 409–419.

Majumder, P.K., and Sellers, W.R. (2005). Akt-regulated pathways in

prostate cancer. Oncogene 24, 7465–7474.

Manning, B.D., and Cantley, L.C. (2007). AKT/PKB signaling:

navigating downstream. Cell 129, 1261–1274.

Park, J., Leong, M.L., Buse, P., Maiyar, A.C., Firestone, G.L., andHemmings, B.A. (1999). Serum and glucocorticoid-inducible

kinase (SGK) is a target of the PI 3-kinase-stimulated signalingpathway. EMBO J 18, 3024–3033.

Reiling, J.H., and Sabatini, D.M. (2006). Stress and mTORture

signaling. Oncogene 25, 6373–6383.

Sabatini, D.M. (2006). mTOR and cancer: insights into a complex

relationship. Nat Rev Cancer 6, 729–734.

Sahoo, S., Brickley, D.R., Kocherginsky, M., and Conzen, S.D.(2005). Coordinate expression of the PI3-kinase downstream

effectors serum and glucocorticoid-induced kinase (SGK-1) andAkt-1 in human breast cancer. Eur J Cancer 41, 2754–2759.

Samuels, Y., Wang, Z., Bardelli, A., Silliman, N., Ptak, J., Szabo, S.,

Yan, H., Gazdar, A., Powell, S.M., Riggins, G.J., et al. (2004). Highfrequency of mutations of the PIK3CA gene in human cancers.Science 304, 554.

Sancak, Y., Peterson, T.R., Shaul, Y.D., Lindquist, R.A., Thoreen, C.C., Bar-Peled, L., and Sabatini, D.M. (2008). The Rag GTPasesbind raptor and mediate amino acid signaling to mTORC1. Science

320, 1496–1501.

Sarbassov, D.D., Ali, S.M., Sengupta, S., Sheen, J.H., Hsu, P.P.,Bagley, A.F., Markhard, A.L., and Sabatini, D.M. (2006). Prolonged

rapamycin treatment inhibits mTORC2 assembly and Akt/PKB.Mol Cell 22, 159–168.

Sarbassov, D.D., Guertin, D.A., Ali, S.M., and Sabatini, D.M. (2005).

Phosphorylation and regulation of Akt/PKB by the rictor-mTORcomplex. Science 307, 1098–1101.

Shiota, C., Woo, J.T., Lindner, J., Shelton, K.D., and Magnuson, M.A.(2006). Multiallelic disruption of the rictor gene in mice reveals thatmTOR complex 2 is essential for fetal growth and viability. Dev Cell11, 583–589.

Wullschleger, S., Loewith, R., and Hall, M.N. (2006). TOR signaling ingrowth and metabolism. Cell 124, 471–484.

Yang, Q., and Guan, K.L. (2007). Expanding mTOR signaling. CellRes 17, 666–681.


Rictor/Cullin 1 complex ubiquitinates SGK1 Protein & Cell

RECOLLECTION

Commitment and dedication of a Chinese plantphysiologistTingyun Kuang1, Ming Li2, Le Kang2

1 Institute of Botany, Chinese Academy of Sciences, Beijing 100093, China2 Beijing Institutes of Life Science, Chinese Academy of Sciences, Beijing 100101, China

Professor Peisong Tang (Pei-sung Tang), born in 1903, is oneof the co-founders of the modern Chinese plant physiology.You might be familiar with some of his academic achieve-ments: he is the first one to discover respiratory enzymes(cytochorome oxidase) in plants; he proved that multiplerespiratory metabolic pathways and electron transfers exist inthe rice; he is the first one to experimentally prove theexistence of carbonic anhydrase in plants; together with Prof.Zhuxi Wang, he used thermodynamical method to illustratethe mechanism of intracellular water movement, which laterhas been referred to as the “Tang-Wang theory of cellularwater potential”. You probably also know that he is the firstChinese biologist who published in Nature and Science in1940s and 1950s. But there is something about him that youwill not learn from his over 200 publications: his commitment,as a Chinese, to his country and his people, and hisdedication, as a scientist, in pursuing science and truths.

Peisong Tang received his early education at TsinghuaCollege from 1917 to 1925, financially supported by the“Boxer Indemnity Scholarship”, which came through heavytaxation on the Chinese people. It was why Peisong Tangalways felt indebted to his country and his people, and madea commitment of returning home after studying abroad foryears. It was also at Tsinghua College where he first showedhis interests in biology. After the biology teacher describedhow the starch stored in the endosperm of a seed convertedinto a seedling, he raised a question that the teacher could notanswer: ‘How exactly do unorganized materials like starchbecome a seedling, something so well-organized andperfectly-structured?’ Little did he then know that the questionbecame the starting point of his scientific life.

After graduation from Tsinghua College, Peisong Tangwent to University of Minnesota and graduated with a B.A.degree. His curriculums at the university covered a broadspectrum of courses, which provided a solid knowledge basisfor his future research. Among the courses, he was

particularly inspired by physical chemistry lectures, whichstrengthened his interests in the energetics of living matter.After that, he went to John Hopkins to study with BurtonLivingtston, but the thesis project that Livingston chose forhim did not quite interest him. Unlike some other Chinesestudents who were humble and always did what they weretold to do, Peisong Tang started to perform some experimentsto study seed respiration and photosynthesis, meanwhile alsoworking on his thesis project. It was in the Livingston labwhere his concept of multiple pathways in respiratorymetabolism started to form.

After getting his PhD degree at John Hopkins, and before

Peisong Tang. (From Selected Works of Tang Pei-Sung,

China Esperanto Press, Beijing, 1993. (《汤佩松论文选集》,中国世界语出版社, 北京, 1993.) )


Protein Cell 2010, 1(10): 886–887DOI 10.1007/s13238-010-0122-y

Protein & Cell

he went to Harvard to work with W. J. Crozier, Peisong Tangspent a summer at the famous Marine Biological Laboratoryat Woods Hole in 1930, and returned for another summer in1931. It was at Woods Hole where he made quite a fewfriends such as Ralph S. Lillie, Ralph Gerald and W. J. V.Osterhout, whose influences made him decide to choosebioenergetics of cellular and plant respiration and of photo-synthesis as his life work.

During his three years at Harvard, he encountered thesame problem as he did at John Hopkins—Crozier wantedhim to work on the temperature characteristics of seedrespiration during germination, a project that did not interesthim much. Working harder than ever, Peisong Tang managedto publish quite a few papers on the temperature character-istic project; meanwhile, he also set up experiments of hisown, and for the first time discovered cytochrome oxidase inplants in 1932. In 1933, with his commitment to his country, hedeclined an offer for a position at a New York university, andaccepted an invitation from Wuhan University to return toChina.

Peisong Tang was quite productive during the first fewyears at Wuhan University. With the $2000 from theuniversity, he managed to set up a small lab for cell andplant physiology, and later published a series of seven paperson the kinetics of cell respiration. Then the eight-year Anti-Japanese War began in 1937. The research on cellrespiration had to stop, but Peisong Tang never stopped.With his knowledge and experience, he did everything hecould to serve the nation at war: first, he converted his lab intoa factory making active carbon for gas masks; after Wuhanwas occupied, he traveled to the city of Guiyang and helpedestablishing a medical school there. Later on, like mostChinese scientists, Peisong Tang moved to Kunming. Withthe relative safety and stability in Kunming, he built a small labof plant physiology, which also served as an assembly placefor young physiologists, many of whom later became thebackbone of Chinese physiology.

After the founding of the new China, Peisong Tang finallywas able to continue his work on plant respiratory metabolismand photosynthesis. No need to repeatedly count hisachievements, but it is important to point out that in hisacademic life, he was always creative and ahead of his time.One good example is the “Tang-Wang theory of cellular waterpotential” that he and Zhuxi Wang developed in 1940s. Theforerunning concept did not get much attention until 1960swhen P. J. Kramer and his colleges published similar theory.Another example is the study by Peisong Tang and Hsiang-YuWu on the adaptive formation of nitrate reductase in rice

seedling, the result of which was published in Nature in 1957.It is not only the first report to show that the nitrate reductaseis inducible, but also the first proved existence of inducibleenzymes in plants.

Even after he retired, the far-seeing vision of Peisong Tangstill played an important role in Chinese plant physiology. Inthe late 1980s, he suggested that inter-disciplinary collabora-tion was much needed to study the membrane proteins ofphotosynthesis, so he invited plant physiologist TingyunKuang, biophysicists Dongcai Liang and Wenrui Changover to his house and the four of them had a small meetingto discuss the possibility of collaboration. Over a decade later,after Peisong Tang passed away, Tingyun Kuang’s group andWenrui Chang’s group jointly solved the structure of the lightharvesting complex II from spinach, which was the firstmembrane protein structure from China. The result waspublished in Nature as a cover-story paper in 2004.

To celebrate his 80th birthday in1983, Annual Review ofPlant Physiology invited Peisong Tang to write an articleabout his academic life, which he described as “the devioustrail of a roaming plant physiologist” in the title of the article.However, with his commitment to his country and his people,and his dedication in pursuing science and truths, he alwaysknew where he was going and never stopped along the way.

Peisong Tang at his make-shift greenhouse in Kunming(From Selected Works of Tang Pei-Sung, China Esperanto

Press, Beijing, 1993. (《汤佩松论文选集》, 中国世界语出版

社, 北京, 1993.) )


Commitment and dedication of a Chinese plant physiologist Protein & Cell

REVIEW

Human catalase: looking for complete identity

Madhur M. Goyal✉, Anjan Basak

Department of Biochemistry, J. N. Medical College, Datta Meghe Insatitute of Medical Sciences (Deemed University), Wardha442004, India✉ Correspondence: [email protected] September 1, 2010 Accepted September 19, 2010

ABSTRACT

Catalases are well studied enzymes that play criticalroles in protecting cells against the toxic effects ofhydrogen peroxide. The ubiquity of the enzyme and theavailability of substrates made heme catalases the focusof many biochemical and molecular biology studies over100 years. In human, this has been implicated in variousphysiological and pathological conditions. Advancementin proteomics revealed many of novel and previouslyunknown features of this mysterious enzyme, but somefunctional aspects are yet to be explained. Along withdiscussion on future research area, this mini-reviewcompile the information available on the structure,function and mechanism of action of human catalase.

KEYWORDS human catalase, structure and function,mechanism of action, futuristic research area

INTRODUCTION

The ability of aerobic respiration using electron transportchain and other haem-containing proteins increased theefficiency of energy production in eukaryotic system. Amongmany other advantages of this evolutionary breakthrough,generation of reactive oxygen species (ROS) is also veryuseful phenomenon (Goth et al., 2004; Oktyabrsky andSmirnova, 2007), which if left unchecked would seriouslyaffect an organism’s viability. These ROS include hydrogenperoxide, superoxide anion radicals, singlet oxygen, hydroxylradicals and nitric oxide. More specifically, the role ofhydrogen peroxide (H2O2) has been implicated in deathinduction (Sancho et al., 2003).

To combat the destructive effects of ROS and makeoxidative cellular metabolism possible, aerobic organismsdeveloped protective antioxidant enzymes such as catalase,superoxide dismutase, peroxiredoxin and glutathione

peroxidase, among which, catalase is a well-known crucialenzyme to scavenge H2O2. This haem-containing protein isthe most efficient enzyme and can decompose millions ofhydrogen peroxide (~107 M/Sec) molecules every second(Young and Woodside, 2001) into molecular oxygen andwater without the production of free radicals (Fig. 1). Evenunder anaerobic conditions, catalase is considered neces-sary to certain parasitic microorganisms for protection againstH2O2 produced by host organisms (Rocha et al., 1996).

Figure 1. Two-stage mechanism of catalase action. Thereaction cycle of the catalase begins with the high spin ferric

(FeIII) state, which reacts with peroxide molecule to formcompound I intermediate, a porphyrin π-cation radical contain-ing FeIV. One of the protons of the hydrogen peroxide molecule

is being removed from one end of the molecule and placed atthe other end. The proton is transferred via a histidine residuein the active site. This action polarizes and breaks the O-Obond in hydrogen peroxide. In the next step, a second

hydrogen peroxide molecule is used as a reductant toregenerate the enzyme, producing water and oxygen. Oxida-tion of an electron donor (here second H2O2) returns

compound I, a highly-oxidising Fe(IV) species, to the nativeresting state Fe(III) (Andersson et al., 1991).


Protein Cell 2010, 1(10): 888–897DOI 10.1007/s13238-010-0113-z

Protein & Cell

Catalases include three types: typical catalases or mono-functional such as mammal type catalases, bifunctionalcatalase-peroxidases, and pseudo catalase. The mammaltype catalases are commonly isolated from animals, plants,fungi and bacteria, and their molecular features are similar:they are composed of four equal-size subunits containing2.5–4 protohemes-IX per tetramer, with a molecular mass of225–270 kDa. Typically, the monofunctional catalases displayminor peroxidase activity, and the target molecules are limitedto small organic substrates. Based on the subunit size, thisgroup can be further divided into catalases with small (55–69 kDa) and large (75–84 kDa) subunits. In addition to thesize, difference also exists in heme prosthetic group, withheme b present in small-subunit enzymes (e.g., bovine livercatalase (BLC)) and heme d present in large-subunitenzymes (e.g., E. coli HPII). Generally, the monofunctionalcatalases are active as tetramers, but dimers, hexamers andeven an unusual heterotrimer structure (from Pseudomonasaeruginosa) were also found (Peter et al., 2000). Catalasesexhibit a broad optimum pH range of 5–10; these areglycoproteins that are resistant to treatment with organicsolvents and are inhibited by 3-amino-1,2,4-triazole (Clai-borne et al., 1979; Nadler et al., 1986; Kim et al., 1994; Brown-Peterson et al., 1995; Terzenbach et al., 1998).

The second group—catalase-peroxidases, have beenisolated from bacteria and fungi, resemble plant and fungalperoxidases in sequence but have larger subunits (~80 kDa).They display several distinguished properties in comparisonto typical catalases: they are reduced by dithionite, they arenot glycoproteins, their activity is pH-dependent, and they aremore sensitive to heat, organic solvents and H2O2, but theyare insensitive to 3-amino-1,2,4-triazole (Nadler et al., 1986;Yumoto et al., 1990; Hochman and Goldberg, 1991; Brown-Peterson and Salin, 1993; Maricinkeviciene et al., 1995;Fraaije et al., 1996). These haem-containing catalases arebifunctional, acting as both catalase and peroxidase, and canuse a variety of organic substances as hydrogen donor.Because the catalase-peroxidases contain two similar,fused domains—one active and the other inactive, thegenes encoding these enzymes appear to evolve througha duplication event. They bear no resemblance to themonofunctional catalases in sequence, but the catalase-peroxidases do contain heme b and are active as dimers ortetramers. The third group, non-heme catalases, containsonly three characterized and sequenced enzymes fromdifferent bacterial species. Activity is derived from amanganese-rich reaction center rather than a heme group,and thus, they are called “pseudo-catalases” or non-haemmanganese-containing catalases (Peter et al., 2000).

In mammalian tissues, catalase activitiy is highest in liverand erythrocytes, relatively high in kidney and adipose tissue,intermediate in lung and pancreas, and very low in heart andbrain (Deisseroth and Dounce, 1970; Schonbaum andChance, 1976; Aebi and Wyss, 1978; Kang et al., 1996). In

human, catalase is absent in vascular smooth muscle cellsand endothelial cells (Shingu et al., 1985); however, it is foundfree in the cytosol of mature erythrocytes. Its activity has alsobeen shown in human milk that is about 10 times higher thanthe level determined in cow’s milk (Friel et al., 2002). In recentreviews, the biochemical history of catalase has beentabulated (Kirkman and Gaetani, 2006; Zamocky et al., 2008).

PHYSIOLOGICAL AND PATHOLOGICAL

SIGNIFICANCE

The enzyme catalase has a predominant role in controllingthe concentration of H2O2 (Gaetani et al., 1996; Mueller et al.,1997) and other cytotoxic oxygen derivatives (Renato et al.,1982). The very rigid and stable structure of tetramericcatalases makes them more resistant to pH, thermaldenaturation and proteolysis than most other enzymes.Their stability and resistance to proteolysis is an evolutionaryadvantage, especially since they are produced during thestationary phase of cell growth when levels of proteasesremain high and rate of protein turnover is rapid.

Catalase protects hemoglobin by removing over half of thehydrogen peroxide generated in normal human erythrocytes,which are exposed to substantial oxygen concentrations(Gaetani et al., 1989). It has been implicated as an importantfactor in inflammation (Halliwell and Gutteridge, 1984),mutagenesis (Vuillame, 1987), prevention of apoptosis(Sandstrom et al., 1993; Islam et al., 1997; Yabuki et al.,1999), and stimulation of a wide spectrum of tumors(Miyamoto et al., 1996). Shih Ho et al. (2004) suggestedthat the role of catalase in antioxidant defense is dependenton the type of tissue and the model of oxidant-mediated tissueinjury.

Addition of exogenous catalase usually attenuated apop-tosis induction as well as the toxicity of antitumor drugs. Itshows both, direct antiapoptotic action of catalase as anantioxidant and indirect proapoptotic action as a suppressorof protective proteins (HSP70 and HSP27) in specialized cells(Sancho et al., 2003). Catalase protects pancreatic β-cellsfrom damage by H2O2 (Tiedge et al., 1997; Tiedge et al.,1998). It was suggested that deficiency of catalase andoxidant damage contribute to the development of diabetes(Góth, 2000; Góth and Eaton, 2000). Low catalase activitieshave been reported in patients with schizophrenia andatherosclerosis (Góth and Vitai, 1996). It also plays animportant role in sperm survival within female tract (Lapointeet al., 1998). Loss of catalase leads to the human geneticdisease known as acatalasemia, or Takahara's disease(Ogata, 1991). The structure of catalase gene and diseasesassociated with mutation and polymorphism has also beendescribed (Goth et al., 2004). Catalase was shown to beeffective in inhibiting the degeneration of neurons (Busciglioand Yankner, 1995; Mann et al., 1997). A growth promotingfactor derived from human erythrocytes with a wide target-cell


Human catalase Protein & Cell

spectrum was also identified as catalase (Takeuchi et al.,1995). In brain, the reaction of ethanol with catalase is animportant source of acetaldehyde (Mason et al., 1997;Zimatkin et al., 1998), which is implicated in the neurologicaleffects of alcohol in humans (Hunt, 1996). Treatment of ratswith the catalase inhibitor, 3-amino-1,2,4-triazole (3AT),decreases voluntary ethanol consumption (Aragon andAmit, 1992). There are also several studies on catalasesfrom agents that cause human disease in relation toprotection against the oxidative bursts of macrophages(Archibald et al., 1986; Bishai et al., 1994).

Reduced catalase activity in Xeroderma pigmentosumcells could be directly related to impaired DNA repair (Quillietet al., 1997) and elevation of catalase activity to an optimumlevel provides protection against doxorubicin-induced cardiacinjury (Kang et al., 1996). It has been proposed that catalase,at acid pH and in the presence of iodide or low concentrationsof hydrogen peroxide (H2O2), can exert a bactericidal effectsimilar to that of neutrophil myeloperoxidase. Erythrocytecatalase protects heterologous somatic cells against chal-lenge by high levels of exogenous H2O2, e.g., in areas ofinflammation (Agar et al., 1986).

STRUCTURE

The ubiquity of the enzyme and the availability of thesubstrates (H2O2 and alkyl peroxides) have made hemecatalases the focus of many biochemical and molecularbiology studies (Kani et al., 2004). The crystal structures ofeight heme-containing monofunctional catalases have beensolved, including those from animal (BLC) and humanerythrocytes catalase (HEC) (Fita et al., 1986; Vainshtein etal., 1986; Murshudov et al., 1992; Bravo et al., 1995; Gouet etal., 1995; Mate et al., 1999; Putnam et al., 2000; Ko et al.,2000; Carpena et al., 2003), revealing a highly conserved β-barrel core structure in all enzymes. Because of their largemolecular size, catalases have been exploited in a number ofcrystal-growth experiments (Sato et al., 1993; Malkin et al.,1995). The crystal structure of human erythrocyte catalase(HEC) has been determined (Putnam et al., 2000). It waspurified and crystallized in three different forms: orthorhom-bic, hexagonal and tetragonal (Ko et al., 2000).

HEC is a tetrameric protein of 244 kDa containing 1997amino acid residues in four identical subunits (named A, B, C,D) of 59.7 kDa, four heme groups, 393 water molecules andfour NADPH molecules (Bonaventura et al., 1972; Kirkmanand Gaetani, 1984; Fita and Rossmann, 1985). Thepolypeptide chain of each subunit has residues 4–502. Anadditional residue, Glu503, was included as an alanine insubunit B. Each subunit can be conceptually divided into fourdomains: β-barrel, N-terminal threading arm, wrapping loopand C-terminal helices (Fig. 2). The extensive hydrophobiccore of each subunit is generated by an eight-stranded antiparallel β-barrel (β1–8), surrounded by a number of α-helices

(Fita and Rossmann, 1985). The N-terminal threading arm(residues 5–70) intricately connects two subunits by hookingthrough a long wrapping loop (residues 380–438) aroundanother subunit. Finally, a helical domain at one face of the βbarrel is composed of four C-terminal helices (α16, α17, α18,and α19) and four helices derived from residues between β4and β5 (α4, α5, α6, and α7). Like BLC, the HEC bindsNADPH.

Two arm-exchanged dimers, which are related by the Q-axis of the P, Q, R molecular axes as defined for BLC (Fita etal., 1986), assemble to form the 222-symmetric tetramer (Koet al., 2000) that is roughly square with overall dimensions100 Å × 100 Å × 70 Å. Tetramerization forces the N-terminalthreading arms from the arm-exchanged dimer to cover theheme active site for the other pair of dimers (related by the R-axis). This human and other catalase structures both suggestthat tetramerization is essential for function and that tetramerassembly may proceed through arm-exchanged dimers asheme can still be loaded. The heme packs perpendicularlyagainst the β2, β3 and β4 and is held between a looppreceding the first β strand, an extended loop (residues332–336), two α-helices from the subunit (α4 and α12), andα2 from a subunit in the other arm-exchanged dimer pair.Tetramerization is important to ensure that the active site issequestered and that the enzyme is competent to completethe reaction cycle rather than allow generation of hydroxylradicals from exposed heme.

The 1.5 Å structure of the peroxiacetic acid (PAA) treatedcatalase reveals that human catalase is extensively hydrated.Throughout the protein, water fills in packing defects betweenthe four domains of the subunit, and between subunits withinthe tetramer. There were 393 bound waters in the currentHEC model (Matthews, 1968). Only the hydrophobic β-barreland the immediate vicinity of the active site are substantiallydevoid of these structural water molecules.

The tetrameric enzyme contains a central cavity andseveral channels that reach the active-site heme groups(Fita and Rossmann, 1985; Gouet et al., 1996; Sevinc et al.,1999; Diaz et al., 2004); access to these deeply buried hemeis via narrow channels restrict access to small molecules,explaining, at least in part, the weak peroxidatic activityamong catalases (Peter et al., 2000). Spectral analysis ofmammalian catalases confirmed the presence of a high spinpentacoordinate heme as active site. A water molecule on thedistal side and one tyrosine residue at proximal side presentas ligand with the heme (Sharma et al., 1989).

Human catalase binds NADPH at a cleft between thehelical domain and the β-barrel on the surface of the mole-cule. The 19 amino-acid residues of BLC involved in contactwith NADPH are also conserved in HEC. The acidic sidechain of Asp213 forms hydrogen bonds with the phosphate ofthe NADPH molecule. Ko et al. showed that the HEC can befully active with NADPH removed (Ko et al., 2000). Phlorizine,a bacterial product, have tendency to bind with NADPH

Protein & Cell


Madhur M. Goyal and Anjan Basak

binding site. This property is used to separate catalase byaffinity column chromatography (Kitlar et al., 1994).

MECHANISM OF ACTION

Catalase is intensively studied because it is easy to isolatefrom tissues like liver and blood, which facilitates itspurification to provide sufficient protein for detailed biochem-ical studies. The heme chromophore provided a convenienttool for the workers who intent to study the reactionmechanism, eventually lead to the characterization of twodistinct stages in the reaction pathway. Even with thetechnology advancement in proteomics, a clear understand-ing of how catalase maintains high selectivity for small-uncharged polar substrate hydrogen peroxide while at thesame time exhibiting turnover rates in excess of 106/sremained elusive (Nicholls et al., 2001; Chelikani et al.,2003). Its turnover number is close to the theoretical rate atwhich the reactants can diffuse together.

Putnam et al. (2000) explained the selectivity andproposed mechanism of action for human catalase, whichincludes substrate selection, reactions of compound-I leadingto the resting state enzyme or the tyrosine radical, the roles ofthe 3AT and cyanide inhibitors; it was termed as molecularruler mechanism for peroxide selection. Initial concentrationof hydrogen peroxide at the active site takes advantage of themolecular ruler generated by the narrow, hydrophobic tunnelthat promotes occupancy of one peroxide molecule and

blocks the passage of large molecules. According to thismechanism amino acids like Val, Phe, Tyr create hydrophobicinteraction a ~25 Å deep channel from the surface to theburied active site. 2–3 Å width allows only water, H2O2 andfew other small molecules to reach the heam molecule. At thetime only four molecules of water can be present in thischannel. Among them one molecule gets attached at thebeginning and another one at the end of the hydrophobicchannel by H-bonds. The space between these watermolecules in hydrophobic channel is too long to be bridgedby the rest of two water molecules and too short toaccommodate an additional fifth water molecule. This gapcan be filled with one H2O2 replacing one H2O among fourpresent in hydrophobic channel. The low dipole moment ofH2O2 is more suitable than water to enter in hydrophobicenvironment of channel. In this way H2O2 is selected andconcentrated at the active site through use of a narrowhydrophobic channel with two fixed water sites next to theactive site (His75 and Asn148) and at the other end of thehydrophobic channel (Asp128 and Gln168).

Haem-containing catalases break down hydrogen perox-ide by a two-stage mechanism in which hydrogen peroxidealternately oxidises and reduces the haem iron at the activesite (Fig. 1). Both the resting state and compound I of catalaseare neutral. At low H2O2 concentrations and in the presenceof one-electron donors, compound I may undergo a one-electron reduction toward the so-called compound II inter-mediate, which transforms back to the resting state by

Figure 2. Structure of human erythrocyte catalase. (A and B) Wrapping loop (a), C-terminal helices (b), β-barrel (c) and N-terminalthreading arm (d) in arm exchanged dimers. Heme molecules are shown in red color and NADPH molecules are in pink (not shown intetramer). Dimer ‘A’ after 90° rotation around axis-Q appears as ‘B’. Two dimers exchange their wrapping loops to form active tetramer ‘C’.

Structures were obtained with the assistance of RasMOL software based on published data (Putnam et al., 2000). Axis R is perpendicularto axis P and Q towards readers’ side.



another one-electron reduction step. Formation of compoundII and III has been described elsewhere (Kirkman et al., 1999;Rovira, 2005). The haem group in catalase is vital to thereaction, because Fe(III) bound at the center of porphyrin ringcan be oxidised to the very oxidised and less common Fe(IV),or ferryl species. This is enhanced by the presence of anearby tyrosine residue, which is a ligand to the iron in thehaem group. The tyrosine is in the ionized phenolate form(O−) and it has lost its proton due to the electron-withdrawingpower of the haem ring and of a nearby arginine residue.

Collectively, presence of amino acid residues at one planeof heam group where H2O2 molecules reached from hydro-phobic channel (His75 and Asn148), presence at oppositeplane (Tyr370), and the volume and shape of the hydrophobicregion play very important role in the heterolytic cleavage ofH2O2 in water and O2. They ensure optimum substrateaccess and electrostatic effects during catalysis (Chelikani etal., 2003). Good geometry for both iron coordination andhydrogen bond formation would require stretching of theperoxide bond, furthering the complex toward the cleavagetransition state. The roles of His75 and Asn148 in theformation and polarization of this peroxide complex is thestep disrupted by 3-amino-1,2,4-triazole (3AT), and preven-tion of iron ligation by peroxide is most likely the mechanismof action of cyanide. Less synthesis of heme brings adecrease in catalytic activity (Muppala et al., 2000).

High turnover number of catalase suggests the presence ofboth inlet and outlet routes to prevent interference betweenincoming H2O2 and exhausting O2. Amara et al. (2001)proposed that the major channel is used for both substrateentry and product exit. The exit of product molecules (H2O orO2) while substrate (H2O2) is entering, is facilitated bypresence of cavities located in the near vicinity of the activesite. In these cavities, product molecules can stay and wait forsubstrate molecule to be passed.

ROLE OF NADPH

Catalase is a major NADPH binding protein withinhuman erythrocytes (Kirkman et al., 1984, 1986). X-raycrystallographic studies revealed that each HEC subunitcontains (NADPH)b (bound NADPH on HEC surface) in anunusual configuration (Fita and Rossmann, 1985).The bind-ing sites have the relative affinities NADPH >NADH>NADP+>NAD+. This reduced dinucleotide is not essentialfor activity of catalase. The function of the bound NADPH isnot fully understood (Kirkman et al., 1984), but threehypothesis were proposed.

The first possibility is that the NADPH decreases thesusceptibility of catalase that is inactivated by low concentra-tions of its toxic substrate, H2O2. When reaction of compoundI with suitable reductants, is frustrated, catalase appears tooxidize bound NADPH as the preferred reductant (Kirkmanet al., 1999). When NADPH is unavailable, the one-electron

reduction of the porphyrin π-cation by an electron fromTyr370 appears to be the next best alternative. Thus, Tyr andNADPH serve as alternative reduction pathways for catalasetrapped in compound I at low peroxide concentrations or withNADPH oxidation being more rapid (Olson and Bruice 1995;Ivancich et al., 1996; Hoffschir et al., 1998; Putnam et al.,2000).

According to next possibility, the (NADPH)b of bovine livercatalase is a remnant of a system in which (NADPH)b wasonce a necessary intermediate in the prevention of compoundII formation by NADPH (present in solution at near vicinity),but the system has evolved so that NADPH bypasses(NADPH)b and provides its reducing equivalents directly forcompound II prevention (Gaetani et al., 2005).

Lastly, the purified samples of human and bovine catalasewere found to bind and release NADPH, suggesting thatcatalase may also function as a regulatory protein, releasingNADP+ when the cell is under peroxidative stress. Thisrelease would augment the removal of H2O2 by theglutathione reductase-glutathione peroxidase mechanisms(Kirkman et al., 1984) to operate more efficiently when the cellis under peroxidative stress. The initial step of that mechan-ism is catalyzed by glucose-6-phosphate dehydrogenase andis rate limiting in human erythrocytes. Catalase represents areservoir of 11–12 µM NADPH. Although this represents onlyone-third of the total NADPH of the erythrocyte, theseconcentrations exceed the 0–5 µM concentation of unboundNADP+ in normal erythrocytes (Kirkman et al., 1980, 1982;Gaetani et al., 1983).

The finding of (NADPH)b was followed by the finding thatadded NADPH largely prevents or reverses the formation ofcompound II by bovine liver or human catalase in thepresence of H2O2 (Jouve et al., 1986; Kirkman et al., 1987).The protective action of NADPH occurs at concentrations aslow as 2 µM (Kirkman et al., 1987). A subsequent studyrevealed that the action of NADPH is more one of preventionthan of reversal (Kirkman et al., 1999). Several groups ofauthors have proposed that the action of (NADPH)b orNADPH is via electron tunneling (Almarsson et al., 1993;Bicout et al., 1995; Gouet et al., 1995; Olson et al., 1995). Thetunnel to the haem group is large enough to accommodateH2O2, but much too small for NADPH. The human erythro-cyte, however, has most of its NADP in the form of NADPH,whereas its NAD is nearly all in the form of NAD+ (Canepa etal., 1991). Therefore, NADPH is known to be effective, and tobe oxidized, in preventing the inactivation of catalaseexposed to H2O2, but the function of (NADPH)b remainsunclear.

INHIBITION

Purdue and Paul (1996) identified a novel peroxisomaltargeting sequence (PTS) at the extreme COOH terminal ofhuman catalase that is necessary to targeting to peroxisomes

Protein & Cell



in human fibroblasts. Deletion of these residues or alterationof the penultimate asparagine to aspartate abolished locali-zation of human catalase to yeast peroxisomes. Catalaseassembly can also occur outside of the peroxisome.

Loss of catalase activity by a variety of chemicals includingH2O2 has been reported (Margoliash et al., 1960; Vetrano etal., 2005; Gibbons et al., 2006). Oxidation by H2O2 does notdirectly affect the active site domain of catalase but bringsabout conformational changes (necessary for catalysis) byoxidation of amino acid residues. Binding of cyanide clearlyblocks heme access to other potential iron ligands (Putnam etal., 2000). 3-amino-1,2,4-triazole (3AT) does not react directly(Jackson et al., 1985) but the near vicinity of active site.

Pyocyanin decreases cellular catalase activity via bothtranscriptional regulation and direct inactivation of theenzyme (O’Malley et al., 2003). Exogenous nitric oxide alsoinhibits catalase activity but the effects are reversible in theabsence of these agents (Sigfrid et al., 2003).

FUTURE RESEARCH AREAS

In this mini-review, we summarized the information availableon human catalase. When started working on humancatalase, we realized that there is a need of compiledinformation to understand its structure, function and mechan-ism of action so that further experiment can be designed. The

Figure 3. Fluorescent spectra of catalase (BLC) in Na-phosphate buffer (50mM, pH 7.4) with various concentration of urea(1–8M). (A) Tryptophan spectra, λex 295. (B) Tyrosine and tryptophan combine spectra, λex 280. Y-axis denotes flourescent intencityunits. In both spectra, fluoroscence peak shifted towards 350 nm with increasing molarity of urea.



information on gray areas is also important. This enzymeplays a significant role in normal function of body ambiance,and thus, better understanding can contribute a lot in field ofmedicine.

Few studies emphasized that the quantity of this enzyme inanimal cells is a balance between the rate of synthesis andthe rate of degradation (Ganschow and Schimke, 1969). It isthought to be synthesized from single gene and built up fromonly one type of subunit; exist in heterogenous form withrespect to their conformations and association status inbiologic system. Prakesh et al. (2002) observed an enzyma-tically active, folded dimmer of native BLC, which is notreported previously. Safo et al. (2001) reported tetragonalcrystal of HEC in which 20 residue N-terminal segmentscorresponding to the first exon of the human catalase genewas omitted. N-terminal segment is essential for dimerizationhence for catalytic activity. A new oxidase activity has recentlybeen characterized in catalase (Vetrano et al., 2005). Fromprotein unfolding study in our laboratory, we recently foundthat bovine liver catalase (of Sigma) produce hydroxylradicals in presence of 1–8M urea (unpublished data). Thishydroxyl radical generation by catalase has also been shownin physiologic conditions. The existences of active dimmers ormonomers of catalase tetramer thought to be responsible forthis activity.

There are very few articles available on the synthesis ofcatalase in mammalian systems. The events occur during thesynthesis of catalase by which heme is loaded in tetramer andthe entry into peroxisomes is not clearly elucidated. Thecomputer simulation proved that SNP (single nucleotidepolymorphism) in catalase gene may translate in the form ofpresence of different amino acid in wild type polypeptidechain. This can cause decrease or complete loss of enzymeactivity (Wood et al., 2008). In the face of few availablereports, detail screening is required to identify these varia-tions. In a recent review some gray areas related to catalasestructure, function and mechanism of action have beendiscussed (Kirkman and Gaetan, 2007). It shows that thepresence of NADPH with enzyme is not completely under-stood. The conversion of DNA damaging solar radiation intoless energetic oxidant species ROS by catalase is novel andpreviously unrecognized activity (Heck et al., 2003). Inerythrocytes, GPx is capable to remove endogenouslyproduced H2O2 thus the role of catalase is thought to be toremove exogenous H2O2 (Johnson et al., 2010). Aserythrocytes circulate in the entire body system, the role ofcatalase in regulation of systemic radox status (H2O2 flux)and comparative load bearing capacity against stress amongother antioxidant enzymes like glutathion peroxidase andthioredoxin is yet to be explored.

As Kirkman and Gaetani (2007) titled the mammaliancatalase ‘a venerable enzyme with new mysteries’, theenzyme coming out with new unexplained properties corro-borates with the title and presents a scope for furtherresearch.

REFERENCES

Aebi, H.E., and Wyss, S.R. (1978). In the metabolic basis of inherited

disease. Stanbury, J.B., Wyngaarden, J.B. and Fredrickson, D.S.,eds. (McGraw-Hill, New York), pp. 1792–1807.

Agar, N.S., Sadrzadeh, S.M.H., Hallaway, P.E., and Eaton, J.W.(1986). Erythrocyte catalase. A somatic oxidant defense? J ClinInvest 77, 319–321.

Almarsson, O., Sinha, A., Gopinath, E., and Bruice, T.C. (1993).Mechanism of one-electron oxidation of NADPH and function ofNADPH bound to catalase. J Am Chem Soc 115, 7093–7102.

Amara, P., Andreoletti, P., Jouve, H.M., and Field, M.J. (2001). Liganddiffusion in the catalase from Proteus mirabilis: a moleculardynamics study. Protein Sci 10, 1927–1935.

Andersson, L.A., and Dawson, L.A. (1991). EXAFS spectroscopy ofhemecontaining oxygenases and peroxidases. Structure andBonding 64, 1–40.

Aragon, C.M., and Amit, Z. (1992). The effect of 3-amino-1,2,4-triazole on voluntary ethanol consumption: evidence for brain

catalase involvement in the mechanism of action. Neuropharma-cology 31, 709–712.

Archibald, F.S., and Duong, M.N. (1986). Superoxide dismutase and

oxygen toxicity defenses in the genus Neisseria. Infect Immun 51,631–641.

Bicout, D.J., Field, M.J., Gouet, P., and Jouve, H.M. (1995).

Simulations of electron transfer in the NADPH-bound catalasefrom Proteus mirabilis PR. Biochim Biophys Acta 1252, 172–176.

Bishai, W.R., Smith, H.O., and Barcak, G.J. (1994). A peroxide/ascorbate-inducible catalase from Haemophilus influenzae ishomologous to the Escherichia coli katE gene product. J Bacteriol176, 2914–2921.

Bishai, W.R., Howard, N.S., Winkelstein, J.A., and Smith, H.O.(1994). Characterization and virulence analysis of catalase

mutants of Haemophilus influenzae. Infect Immun 62, 4855–4860.

Bonaventura, J., Schroeder, W.A., and Fang, S. (1972). Humanerythrocyte catalase: an improved method of isolation and a

reevaluation of reported properties. Arch Biochem Biophys 150,606–617.

Bravo, J., Verdaguer, N., Tormo, J., Betzel, C., Switala, J., Loewen, P.

C., and Fita, I. (1995). Crystal structure of catalase HPII fromEscherichia coli. Structure 3, 491–502.

Brown-Peterson, N.J., and Salin, M.L. (1993). Purification of acatalase-peroxidase from Halobacterium halobium: characteriza-tion of some unique properties of the halophilic enzyme. J Bacteriol175, 4197–4202.

Brown-Peterson, N.J., and Salin, M.L. (1995). Purification andcharacterization of a mesohalic catalase from the halophilic

bacterium Halobacterium halobium. J Bacteriol 177, 378–384.

Busciglio, J., and Yankner, B.A. (1995). Apoptosis and increasedgeneration of reactive oxygen species in Down’s syndrome

neurons in vitro. Nature 378, 776–779.

Canepa, L., Ferraris, A.M., Miglino, M., and Gaetani, G.F. (1991).Bound and unbound pyridine dinucleotides in normal and glucose-

6-phosphate dehydrogenase-deficient erythrocytes. Biochim Bio-phys Acta 1074, 101–104.

Carpena, X., Soriano, M., Klotz, M.G., Duckworth, H.W., Donald, L.J.,Melik-Adamyan, W., Fita, I., and Loewen, P.C. (2003). Structure ofthe Clade 1 catalase, CatF of Pseudomonas syringae, at 1.8 A

Protein & Cell



resolution. Proteins 50, 423–436.

Chelikani, P., Carpena, X., Fita, I., and Loewen, P.C. (2003). An

electrical potential in the access channel of catalases enhancescatalysis. J Biol Chem 278, 31290–31296.

Claiborne, A., Malinowski, D.P., and Fridovich, I. (1979). Purification

and characterization of hydroperoxidase II of Escherichia coli B. JBiol Chem 254, 11664–11668.

Deisseroth, A., and Dounce, A.L. (1970). Catalase: Physical and

chemical properties, mechanism of catalysis, and physiologicalrole. Physiol Rev 50, 319–375.

Díaz, A., Horjales, E., Rudiño-Piñera, E., Arreola, R., and Hansberg,W. (2004). Unusual Cys-Tyr covalent bond in a large catalase. JMol Biol 342, 971–985.

Fita, I., and Rossmann, M.G. (1985). The active center of catalase. JMol Biol 185, 21–37.

Fita, I., and Rossmann, M.G. (1985). The NADPH binding site on beef

liver catalase. Proc Natl Acad Sci U S A 82, 1604–1608.

Fita, I., Silva, A.M., Murthy, M.R.N., and Rossmann, M.G. (1986). The

refined structure of beef liver catalase at 2.5 Å resolution. ActaCrystallogr 42, 497–515.

Fraaije, M.W., Roubroeks, H.P., Hagen, W.R., and Van Berkel,

W.J.H. (1996). Purification and characterization of an intracellularcatalase-peroxidase from Penicillium simplicissimum. Eur JBiochem 235, 192–198.

Friel, J.K., Martin, S.M., Langdon, M., Herzberg, G.R., and Buettner,G.R. (2002). Milk from mothers of both premature and full-terminfants provides better antioxidant protection than does infant

formula. Pediatr Res 51, 612–618.

Gaetani, F., and Kirkman, H.N. (1983). [abstr.] Am J Hum Genet 35,43.

Gaetani, G.F., Galiano, S., Canepa, L., Ferraris, A.M., and Kirkman,H.N. (1989). Catalase and glutathione peroxidase are equally

active in detoxification of hydrogen peroxide in human erythro-cytes. Blood 73, 334–339.

Gaetani, G.F., Ferraris, A.M., Rolfo, M., Mangerini, R., Arena, S., and

Kirkman, H.N. (1996). Predominant role of catalase in the disposalof hydrogen peroxide within human erythrocytes. Blood 87,1595–1599.

Gaetani, G.F., Ferraris, A.M., Sanna, P., and Kirkman, H.N. (2005). Anovel NADPH:(bound) NADP+ reductase and NADH:(bound)NADP+ transhydrogenase function in bovine liver catalase.

Biochem J 385, 763–768.

Ganschow, R.E., and Schimke, R.T. (1969). Independent geneticcontrol of the catalytic activity and the rate of degradation of

catalase in mice. J Biol Chem 244, 4649–4658.

Gibbons, N.C.J., Wood, J.M., Rokos, H., and Schallreuter, K.U.

(2006). Computer simulation of native epidermal enzyme struc-tures in the presence and absence of hydrogen peroxide (H2O2):potential and pitfalls. J Invest Dermatol 126, 2576–2582.

Góth, L. (2000). Lipid and carbohydrate metabolism in acatalasemia.Clin Chem 46, 564–566. < /jrn >

Góth, L., and Eaton, J.W. (2000). Hereditary catalase deficiencies

and increased risk of diabetes. Lancet 356, 1820–1821.

Góth, L., Rass, P., and Páy, A. (2004). Catalase enzyme mutationsand their association with diseases. Mol Diagn 8, 141–149.

Góth, L., and Vitai, M. (1996). Hypocatalasemia in hospital patients.Clin Chem 42, 341–342.

Gouet, P., Jouve, H.M., and Dideberg, O. (1995). Crystal structure of

Proteus mirabilis PR catalase with and without bound NADPH. JMol Biol 249, 933–954.

Gouet, P., Jouve, H.M., Williams, P.A., Andersson, I., Andreoletti, P.,Nussaume, L., and Hajdu, J. (1996). Ferryl intermediates ofcatalase captured by time-resolved Weissenberg crystallographyand UV-VIS spectroscopy. Nat Struct Biol 3, 951–956.

Halliwell, B., and Gutteridge, J.M.C. (1984). Oxygen toxicity, oxygenradicals, transition metals and disease. Biochem J 219, 1–14.

Heck, D.E., Vetrano, A.M., Mariano, T.M., and Laskin, J.D. (2003).UVB light stimulates production of reactive oxygen species:unexpected role for catalase. J Biol Chem 278, 22432–22436.

Ho, Y.S., Xiong, Y., Ma, W., Spector, A., and Ho, D.S. (2004). Micelacking catalase develop normally but show differential sensitivityto oxidant tissue injury. J Biol Chem 279, 32804–32812.

Hochman, A., and Shemesh, A. (1987). Purification and character-ization of a catalase-peroxidase from the photosynthetic bacterium

Rhodopseudomonas capsulata. J Biol Chem 262, 6871–6876.

Hoffschir, F., Daya-Grosjean, L., Petit, P.X., Nocentini, S., Dutrillaux,B., Sarasin, A., and Vuillaume, M. (1998). Low catalase activity in

xeroderma pigmentosum fibroblasts and SV40-transformedhuman cell lines is directly related to decreased intracellular levelsof the cofactor, NADPH. Free Radic Biol Med 24, 809–816.

Hunt, W.A. (1996). Role of acetaldehyde in the actions of ethanol onthe brain—a review. Alcohol 13, 147–151.

Islam, K.N., Kayanoki, Y., Kaneto, H., Suzuki, K., Asahi, M., Fujii, J.,and Taniguchi, N. (1997). TGF-beta1 triggers oxidative modifica-tions and enhances apoptosis in HIT cells through accumulation ofreactive oxygen species by suppression of catalase and

glutathione peroxidase. Free Radic Biol Med 22, 1007–1017.

Ivancich, A., Jouve, H.M., and Gaillard, J. (1996). EPR evidence for a

tyrosyl radical intermediate in bovine liver catalase. J Am ChemSoc 118, 12852–12853.

Williams, R.N., Delamere, N.A., and Paterson, C.A. (1985). Inactiva-

tion of catalase with 3-amino-1,2,4-triazole: an indirect irreversiblemechanism. Biochem Pharmacol 34, 3386–3389.

Johnson, R.M., Ho, Y.S., Yu, D.Y., Kuypers, F.A., Ravindranath, Y.,

and Goyette, G.W. (2010). The effects of disruption of genes forperoxiredoxin-2, glutathione peroxidase-1, and catalase on ery-throcyte oxidative metabolism. Free Radic Biol Med 48, 519–525.

Jouve, H.M., Pelmont, J., and Gaillard, J. (1986). Interaction betweenpyridine adenine dinucleotides and bovine liver catalase: achromatographic and spectral study. Arch Biochem Biophys 248,

71–79.

Kang, Y.J., Chen, Y., and Epstein, P.N. (1996). Suppression of

doxorubicin cardiotoxicity by overexpression of catalase in theheart of transgenic mice. J Biol Chem 271, 12610–12616.

kani, P.C., Fita, I., Loewen, P.C. (2004). Diversity of structures and

properties among catalases. Cell Mol Life Sci 61, 192–208.

Kim, H., Lee, J.S., Hah, Y.C., and Roe, J.H. (1994). Characterizationof the major catalase from Streptomyces coelicolor ATCC 10147.

Microbiology 140, 3391–3397.

Kirkman, H.N. (1982). [abstr.] Fed. Proc. Fed. Am. Soc. Exp. Biol. 41,

1398.

Kirkman, H.N., and Gaetani, G.F. (1984). Catalase: a tetramericenzyme with four tightly bound molecules of NADPH. Proc Natl

Acad Sci U S A 81, 4343–4347.

Kirkman, H.N., and Gaetani, G.F. (2007). Mammalian catalase: avenerable enzyme with new mysteries. Trends Biochem Sci 32,



44–50.

Kirkman, H.N., Gaetani, G.F., and Clemons, E.H. (1986). NADP-

binding proteins causing reduced availability and sigmoid releaseof NADP+ in human erythrocytes. J Biol Chem 261, 4039–4045.

Kirkman, H.N., Galiano, S., and Gaetani, G.F. (1987). The function of

catalase-bound NADPH. J Biol Chem 262, 660–666.

Kirkman, H.N., Rolfo, M., Ferraris, A.M., and Gaetani, G.F. (1999).Mechanisms of protection of catalase by NADPH. Kinetics and

stoichiometry. J Biol Chem 274, 13908–13914.

Kirkman, H.N., Wilson, W.G., and Clemons, E.H. (1980). Regulation

of glucose-6-phosphate dehydrogenase I. Intact red cells. J LabClin Med 95, 877–887.

Kitlar, T., Döring, F., Diedrich, D.F., Frank, R., Wallmeier, H., Kinne, R.

K., and Deutscher, J. (1994). Interaction of phlorizin, a potentinhibitor of the Na+/D-glucose cotransporter, with the NADPH-binding site of mammalian catalases. Protein Sci 3, 696–700.

Ko, T.P., Safo, M.K., Musayev, F.N., Di Salvo, M.L., Wang, C., Wu, S.H., and Abraham, D.J. (2000). Structure of human erythrocytecatalase. Acta Crystallogr D Biol Crystallogr 56, 241–245.

Lapointe, S., Sullivan, R., and Sirard, M.A. (1998). Binding of a bovineoviductal fluid catalase to mammalian spermatozoa. Biol Reprod58, 747–753.

Malkin, A.J., Kuznetsov YuG, Land, T.A., DeYoreo, J.J., andMcPherson, A. (1995). Mechanisms of growth for protein and

virus crystals. Nat Struct Biol 2, 956–959.

Mann, H., McCoy, M.T., Subramaniam, J., Van Remmen, H., andCadet, J.L. (1997). Overexpression of superoxide dismutase and

catalase in immortalized neural cells: toxic effects of hydrogenperoxide. Brain Res 770, 163–168.

Margoliash, E., Novogrodsky, A., and Schejter, A. (1960). Irreversible

reaction of 3-amino-1:2:4-triazole and related inhibitors with theprotein of catalase. Biochem J 74, 339–348.

Marcinkeviciene, J.A., Magliozzo, R.S., and Blanchard, J.S. (1995).Purification and characterization of the Mycobacterium smegmatiscatalase-peroxidase involved in isoniazid activation. J Biol Chem270, 22290–22295.

Hamby-Mason, R., Chen, J.J., Schenker, S., Perez, A., andHenderson, G.I. (1997). Catalase mediates acetaldehyde forma-

tion from ethanol in fetal and neonatal rat brain. Alcohol Clin ExpRes 21, 1063–1072.

Maté, M.J., Zamocky, M., Nykyri, L.M., Herzog, C., Alzari, P.M.,

Betzel, C., Koller, F., and Fita, I. (1999). Structure of catalase-Afrom Saccharomyces cerevisiae. J Mol Biol 286, 135–149.

Matthews, B.W. (1968). Solvent content of protein crystals. J Mol Biol

33, 491–497.

Miyamoto, T., Hayashi, M., Takeuchi, A., Okamoto, T., Kawashima,

S., Takii, T., Hayashi, H., and Onozaki, K. (1996). Identification of anovel growth-promoting factor with a wide target cell spectrum fromvarious tumor cells as catalase. J Biochem 120, 725–730.

Mueller, S., Riedel, H.D., and Stremmel, W. (1997). Direct evidencefor catalase as the predominant H2O2 -removing enzyme in humanerythrocytes. Blood 90, 4973–4978.

Muppala, V.K., Lin, C.S., and Lee, Y.H. (2000). The role of HNF-1alpha in controlling hepatic catalase activity. Mol Pharmacol 57,93–100.

Murshudov, G.N., Melik-Adamyan, W.R., Grebenko, A.I., Barynin, V.V., Vagin, A.A., Vainshtein, B.K., Dauter, Z., and Wilson, K.S.(1992). Three-dimensional structure of catalase from Micrococcus

lysodeikticus at 1.5 A resolution. FEBS Lett 312, 127–131.

Nadler, V., Goldberg, I., and Hochman, A. (1986). Comparative study

of bacterial catalase. Biochim Biophys Acta 882, 234–241.

Nicholls, P., Fita, I., and Loewen, P.C. (2001). Enzymology andstructure of catalases. Adv Inorg Chem 51, 51–106.

O’Malley, Y.Q., Reszka, K.J., Rasmussen, G.T., Abdalla, M.Y.,Denning, G.M., and Britigan, B.E. (2003). The Pseudomonassecretory product pyocyanin inhibits catalase activity in human

lung epithelial cells. Am J Physiol Lung Cell Mol Physiol 285,L1077–L1086.

Ogata, M. (1991). Acatalasemia. Hum Genet 86, 331–340.

Oktyabrsky, O.N., and Smirnova G.V. (2007). Redox regulations of

cellular functions. Biochemistry (Moscow) 72, 2, 132–145.

Olson, L.P., and Bruice, T.C. (1995). Electron tunneling and ab initio

calculations related to the one-electron oxidation of NAD(P)Hbound to catalase. Biochemistry 34, 7335–7347.

Peter, C.L., Martin, G.K., and Daniel, J.H. (2000). Catalase–an “old”

enzyme that continues to surprise us. ASM News 66, 76–82.

Prakash, K., Prajapati, S., Ahmad, A., Jain, S.K., and Bhakuni, V.(2002). Unique oligomeric intermediates of bovine liver catalase.

Protein Sci 11, 46–57.

Purdue, P.E., and Lazarow, P.B. (1996). Targeting of human catalaseto peroxisomes is dependent upon a novel COOH-terminal

peroxisomal targeting sequence. J Cell Biol 134, 849–862.

Putnam, C.D., Arvai, A.S., Bourne, Y., and Tainer, J.A. (2000). Active

and inhibited human catalase structures: ligand and NADPHbinding and catalytic mechanism. J Mol Biol 296, 295–309.

Quilliet, X., Chevallier-Lagente, O., Zeng, L., Calvayrac, R., Mezzina,

M., Sarasin, A., and Vuillaume, M. (1997). Retroviral-mediatedcorrection of DNA repair defect in xeroderma pigmentosum cells isassociated with recovery of catalase activity. Mutat Res 385,

235–242.

Iozzo, R.V., MacDonald, G.H., and Wight, T.N. (1982). Immunoelec-tron microscopic localization of catalase in human eosinophilic

leukocytes. J Histochem Cytochem 30, 697–701.

Rocha, E.R., Selby, T., Coleman, J.P., and Smith, C.J. (1996).Oxidative stress response in an anaerobe, Bacteroides fragilis: a

role for catalase in protection against hydrogen peroxide. JBacteriol 178, 6895–6903.

Rovira, C. (2005). Structure, protonation state and dynamics ofcatalase compound II. Chemphyschem 6, 1820–1826.

Safo, M.K., Musayev, F.N., Wu, S.H., Abraham, D.J., and Ko, T.P.

(2001). Structure of tetragonal crystals of human erythrocytecatalase. Acta Crystallogr D Biol Crystallogr 57, 1–7.

Sancho, P., Troyano, A., Fernández, C., De Blas, E., and Aller, P.

(2003). Differential effects of catalase on apoptosis induction inhuman promonocytic cells. Relationships with heat-shock proteinexpression. Mol Pharmacol 63, 581–589.

Sandstrom, P.A., and Buttke, T.M. (1993). Autocrine production ofextracellular catalase prevents apoptosis of the human CEM T-cellline in serum-free medium. Proc Natl Acad Sci U S A 90,

4708–4712.

Sato, A., Furuno, T., Toyoshima, C., and Sasabe, H. (1993). Two-

dimensional crystallization of catalase on a monolayer film ofpoly(1-benzyl-L-histidine) spread at the air/water interface. Bio-chim Biophys Acta 1162, 54–60.

Schonbaum, G.R., and Chance, B. (1976). In the enzymes, Boyer, P.D. ed. (Academic, New York), pp.368–408.

Protein & Cell



Sevinc, M.S., Maté, M.J., Switala, J., Fita, I., and Loewen, P.C. (1999).

Role of the lateral channel in catalase HPII of Escherichia coli.Protein Sci 8, 490–498.

Sharma, K.D., Andersson, L.A., Loehr, T.M., Terner, J., and Goff,

H.M. (1989). Comparative spectral analysis of mammalian,fungal, and bacterial catalases. Resonance Raman evidence foriron-tyrosinate coordination. J Biol Chem 264, 12772–12779.

Shingu, M., Yoshioka, K., Nobunaga, M., and Yoshida, K. (1985).Human vascular smooth muscle cells and endothelial cells lackcatalase activity and are susceptible to hydrogen peroxide.

Inflammation 9, 309–320.

Sigfrid, L.A., Cunningham, J.M., Beeharry, N., Lortz, S., Tiedge, M.,Lenzen, S., Carlsson, C., and Green, I.C. (2003). Cytokines and

nitric oxide inhibit the enzyme activity of catalase but not its proteinor mRNA expression in insulin-producing cells. J Mol Endocrinol31, 509–518.

Takeuchi, A., Miyamoto, T., Yamaji, K., Masuho, Y., Hayashi, M.,Hayashi, H., and Onozaki, K. (1995). A human erythrocyte-derivedgrowth-promoting factor with a wide target cell spectrum: identifi-

cation as catalase. Cancer Res 55, 1586–1589.

Terzenbach, D.P., and Blaut, M. (1998). Purification and character-

ization of a catalase from the nonsulfur phototrophic bacteriumRhodobacter sphaeroides ATH 2.4.1 and its role in the oxidativestress response. Arch Microbiol 169, 503–508.

Tiedge, M., Lortz, S., Drinkgern, J., and Lenzen, S. (1997). Relationbetween antioxidant enzyme gene expression and antioxidativedefense status of insulin-producing cells. Diabetes 46, 1733–1742.

Tiedge, M., Lortz, S., Munday, R., and Lenzen, S. (1998).Complementary action of antioxidant enzymes in the protectionof bioengineered insulin-producing RINm5F cells against the

toxicity of reactive oxygen species. Diabetes 47, 1578–1585.

Vainshtein, B.K., Melik-Adamyan, W.R., Barynin, V.V., Vagin, A.A.,

Grebenko, A.I., Borisov, V.V., Bartels, K.S., Fita, I., and Rossmann,M.G. (1986). Three-dimensional structure of catalase fromPenicillium vitale at 2.0 A resolution. J Mol Biol 188, 49–61.

Vetrano, A.M., Heck, D.E., Mariano, T.M., Mishin, V., Laskin, D.L., andLaskin, J.D. (2005). Characterization of the oxidase activity inmammalian catalase. J Biol Chem 280, 35372–35381.

Vuillaume, M. (1987). Reduced oxygen species, mutation, inductionand cancer initiation. Mutat Res 186, 43–72.

Wood, J.M., Gibbons, N.C.J., Chavan, B., and Schallreuter, K.U.(2008). Computer simulation of heterogeneous single nucleotidepolymorphisms in the catalase gene indicates structural changesin the enzyme active site, NADPH-binding and tetramerization

domains: a genetic predisposition for an altered catalase inpatients with vitiligo? Exp Dermatol 17, 366–371.

Yabuki, M., Kariya, S., Ishisaka, R., Yasuda, T., Yoshioka, T., Horton,A.A., and Utsumi, K. (1999). Resistance to nitric oxide-mediatedapoptosis in HL-60 variant cells is associated with increasedactivities of Cu,Zn-superoxide dismutase and catalase. Free Radic

Biol Med 26, 325–332.

Young, I.S., and Woodside, J.V. (2001). Antioxidants in health and

disease. J Clin Pathol 54, 176–186.

Yumoto, I., Fukumori, Y., and Yamanaka, T. (1990). Purification andcharacterization of catalase from a facultative alkalophilic Bacillus.

J Biochem 108, 583–587.

Zimatkin, S.M., Liopo, A.V., and Deitrich, R.A. (1998). Distribution andkinetics of ethanol metabolism in rat brain. Alcohol Clin Exp Res

22, 1623–1627.

Zamocky, M., Furtmüller, P.G., and Obinger, C. (2008). Evolution of

catalases from bacteria to humans. Antioxid Redox Signal 10,1527–1548.



REVIEW

Wnt pathway antagonists and angiogenesis

Bin Zhang, Jian-xing Ma✉

Department of Physiology, Department of Medicine, The University of Oklahoma Health Sciences Center, Oklahoma City, Ok73104, USA✉ Correspondence: [email protected] September 19, 2010 Accepted September 27, 2010

ABSTRACT

Dysregulation of the Wnt pathway has been extensivelystudied in multiple diseases, including some angiogenicdisorders. Wnt signaling activation is a major stimulatorin pathological angiogenesis and thus, Wnt antagonistsare believed to have therapeutic potential for neovasculardisorders. Actually, some Wnt antagonists have beenidentified directly from the anti-angiogenic factor family.This review summarizes the recent progress towardunderstanding of the roles of Wnt pathway antagonists inangiogenic regulation and their mechanism of action,and exploring their therapeutic potential.

KEYWORDS sFRPs, Dkk, WIF-1, endostatin, SERPI-NA3K, curcumin

INTRODUCTION

The Wnt signaling pathway

‘WNT’ is originated from the Wg (wingless) and Int genes(Rijsewijk et al., 1987). Recessive mutations in the Wg genesaffect wing development in Drosophila (Sharma and Chopra,1976). In the tumor induced by mouse mammary tumor virus,the overexpressed INT genes were identified near virusintegration sites (Nusse et al., 1984). Subsequent workrevealed that the Int-1 gene and the Wg gene werehomologous and thus, named ‘Wnt’.

Wnts are a family of 19 secreted, cysteine-rich glycopro-teins (He et al., 2004; van Amerongen et al., 2008). Wntligands activate several signaling pathways involved in theregulation of development at different stages (Nusse, 2005;van Amerongen et al., 2008). The Wnt signaling cascade hastwo distinct branches: the canonical Wnt pathway and thenon-canonical pathway (Zerlin et al., 2008). The Wnt

signaling pathway is involved in multiple physiologic andpathological processes and is well studied in embryogenesisand carcinogenesis (Clevers, 2006). Recent evidence sug-gests that the Wnt pathway also plays an important role inangiogenesis (Masckauchán and Kitajewski, 2006).

Canonical Wnt signaling is initiated by Wnt ligands. In theabsence of Wnt ligands, the transcription factor β-catenin, adown-stream effector of the canonical Wnt pathway, isphosphorylated by a protein complex containing GSK-3 inthe cytosol and constantly degraded to prevent its accumula-tion (Li et al., 2002; Polakis, 2002). Upon binding of certainWnt ligands to the Frizzled (Fz) receptor and the co-receptor,low-density lipoprotein receptor-related protein 5 or 6(LRP5/6), the Fz receptor and LRP5/6 dimerize, forming areceptor/co-receptor complex (He et al., 2004). As a result,downstream signaling, including phosphorylation of LRP5/6and stabilization of β-catenin, is initiated (Orsulic and Peifer,1996; Dale, 1998). β-catenin is subsequently translocatedinto the nucleus, associates with T cell factor (TCF) for DNAbinding and regulates expression of target genes (Behrenset al., 1996; He et al., 2004). Some target genes of thecanonical Wnt pathway such as cyclinD1 and c-myc areimplicated in cellular proliferation (He et al., 1998; Tetsu andMcCormick, 1999). Recent evidence indicates that the Wnt/β-catenin pathway also plays important roles in pathologicalprocesses, such as inflammation and angiogenesis, and alsoregulates inflammatory and angiogenic factors such as Cox-2and VEGF (Howe et al., 1999; Easwaran et al., 2003; Ojesina,2004; Tachikawa et al., 2004).

The non-canonical Wnt pathway comprises several distinctbranches and is associated with some activities independentof Wnt/β-catenin signaling. The Wnt/PCP pathway is knownto regulate orientation of cellular structure in both vertebrates(Guo et al., 2004) and invertebrates (Mlodzik, 2002). The PCPpathway is regulated by the Fz receptor, Dishevelled (Dvl),the small GTPases RhoA and Rac1, and c-jun N-terminal



Protein & Cell

kinase (JNK) (Fanto and McNeill, 2004; Povelones et al.,2005). The relevance of PCP to vascularization is obscure.The Wnt/calcium pathway is another distinct branch of thenon-canonical Wnt pathway (Kühl et al., 2000). Upon bindingof certain Wnt ligands (e.g., Wnt5a and Fz2) with the Fzreceptors, G proteins are activated, leading to the release ofintracellular calcium (Slusarski et al., 1997; Kühl et al., 2000).Furthermore, some Ca2+-sensitive enzymes such as PKC(Sheldahl et al., 1999) and Ca2+-calmodulin kinase II (Kühl etal., 2000) are activated to regulate the downstream Wnt/calcium signaling. Although calcium signaling is known tomediate angiogenesis, the role of the Wnt/calcium pathway inneovascularization has not been established (Kohn et al.,1995).

There is complex crosstalk between the canonical Wntpathway and the non-canonical pathway. Evidence suggeststhat non-canonical Wnt signaling antagonizes Wnt/β-cateninsignaling (Ishitani et al., 2003; Topol et al., 2003). On the otherhand, Rac1 and JNK activation in the PCP pathway controlsnuclear translocation of β-catenin during canonical Wntsignaling (Wu et al., 2008). The crosstalk between the Wntsignaling components and other pathways, such as the Notchpathway, the MAP kinase pathway and the NF-κB pathway,have been identified (Behrens, 2000; Nakamura et al., 2007;Umar et al., 2009). All of these angiogenesis-related path-ways form a complicated net interacting with the Wntsignaling pathway.

Wnt signaling pathway and angiogenesis

Wnt signaling is essential for endothelial cell proliferation andangiogenesis. The roles of Wnt signaling in some pathologi-cal conditions with abnormal neovascularization have beenrevealed recently (Logan and Nusse, 2004; Masckauchánand Kitajewski, 2006). The role of Wnt signaling in vesselformation has been established based on phenotypes fromdisruptions and mutations in the Wnt/Fz genes in animalmodels or humans. Disruption of theWnt-2 andWnt-7b genesresults in defects in the placental and pulmonary vasculaturesin mice (Monkley et al., 1996; Shu et al., 2002). Thephenotype of Fz-5 knockout mice is embryonic lethal due tothe disorganization of the capillary plexus in the placenta (Heet al., 1997). Mutations in the Fz-4 gene have been linked toFEVR, which is associated with defective retinal angiogen-esis (Robitaille et al., 2002; Xu et al., 2004). Fz-4 knockoutmice also show a retinal vascular defect (Robitaille et al.,2002). Taken together, these studies have established crucialroles of Wnt ligands and their receptors in angiogenesis.

The Wnt signaling pathway is also a direct mediator ofendothelial cell growth and survival (Wright et al., 1999).Several Fz receptors are found on the surface of primary ECs,and these cells respond to Wnt ligands stimulation (Cheng etal., 2003; Masckauchán et al., 2005; Zerlin et al., 2008). Bothof Wnt1 and Wnt5a which target the canonical and non-

canonical Wnt pathways can induce EC proliferation (Masck-auchán et al., 2005, 2006).

Further, Wnt signaling upregulates several pro-angiogenicfactors. For example, vascular endothelial growth factor(VEGF) gene promoter region contains seven TCF/LEFbinding sites (Easwaran et al., 2003). In colon cancer cells,VEGF is upregulated as a result of the ectopic activation ofWnt/β-catenin signaling (Zhang et al., 2001; Easwaran et al.,2003). The Wnt pathway is considered a potential therapeutictarget to attenuate VEGF overexpression in tumor angiogen-esis (Hu et al., 2009). Interleukin-8 is another target of Wnt/β-catenin signaling (Masckauchán et al., 2005). IL-8 inducesendothelial cell proliferation and survival. IL-8 also inducesoverexpression of MMP-2 and MMP-9, which both participatein the angiogenesis process (Li et al., 2003). In addition, otherangiogenic factors have been reported as Wnt target genes,such as uPAR, FGF18 and 20, and MMP3 and 7 (Brabletz etal., 1999; Mann et al., 1999; Prieve and Moon, 2003;Shimokawa et al., 2003, 2005). Therefore, the angiogeniceffect of Wnt/β-catenin may also be through inducing pro-angiogenic factors.

CLASSICAL, ENDOGENOUS WNT ANTAGONISTS

Secreted frizzled related proteins (sFRPs) family

The sFRPs family consists of a group of Wnt binding proteinscontaining a frizzled-type cysteine-rich domain (CRD) (Ratt-ner et al., 1997; Jones and Jomary, 2002; Kawano and Kypta,2003). The CRD domain which contains ten cysteine residuesand other conserved residues is located in the N terminus ofthe protein (Rattner et al., 1997). The C terminus of sFRPsalso contains a netrin (NTR) domain, containing six cysteineresidues and conserved segments of hydrophobic residues(Bányai and Patthy, 1999). In mammals, there are five sFRPs,designated as sFRP1–5, with the predicted molecular weightof 33–40 kDa (around 300 amino acids) (Jones and Jomary,2002). Since their CRD domains share sequence homology(30%–50%) with CRD in the Wnt receptors (Fz), it is predictedthat sFRPs can regulate the Wnt pathway, and the function ofNTR domain remains unclear (Melkonyan et al., 1997;Rattner et al., 1997; Bányai and Patthy, 1999; Jones andJomary, 2002; Kawano and Kypta, 2003).

Based on the sequence homology, sFRPs are divided intotwo subgroups. One subgroup comprises sFRP1, sFRP2 andsFRP5. Another subgroup contains sFRP3 and sFRP4(Jones and Jomary, 2002). In the sFRP family, the firstidentified member is sFRP3, which was originally named Frzb(Hoang et al., 1996). In the study of Xenopus embryosdevelopment, sFRP3/Frzb was reported to interact with Xwnt-8 and block the Xwnt-8 signaling (Leyns et al., 1997; Wang etal., 1997a). Furthermore, in mammalian cell cultures, sFRP3/Frzb was found to bind to Wnt-1 and inhibit the β-cateninaccumulation induced by Wnt-1 (Lin et al., 1997; Wang et al.,


Wnt pathway antagonists and angiogenesis Protein & Cell

1997b). Since the discovery of sFRP3, other members havebeen identified, and some of them are identical to otherexisting genes. Therefore, each sFRP has several alternativenames. For example, sFRP1 is also named Frp-1, SARP2and FrzA; sFRP2 has the alternative names as SDF-5 andSARP1; sFRP3 was originally identified as Frzb, Frzb-1, Fritzand Frezzled; sFRP4 has been known as DDC-4, frpAP,frpHE and Frzb-2; sFRP5 has other names like SARP3,hFRP-1b and Frzb-1b (Jones and Jomary, 2002; Kawano andKypta, 2003). In addition to sFRP3, sFRP1 and sFRP2 werealso reported to antagonizeWnt activity (Finch et al., 1997; Xuet al., 1998; Ladher et al., 2000; Uren et al., 2000). Aninteresting finding from sFRP1 is that it is not always anantagonist of Wnt signaling. At low concentrations, sFRP1 actas an agonist in the tissue culture experiments, suggestingthat there may be different binding sites for it (Uren et al.,2000).

The roles of sFRPs in tumor growth have been well studied(Shi et al., 2007), relating to their Wnt modulating activities.Since angiogenesis is also known to be regulated by Wntsignaling, the function of sFRPs in angiogenesis is predict-able. A few studies have reported the anti-angiogenicactivities of sFRPs. sFRP4 downregulates endothelial cellmigration and proliferation, disrupts the stability of endothelialrings and inhibits the development of sprouts and pseudopo-dia. The inhibitory effects on endothelial cells are from theantagonizing effects on both the canonical Wnt pathway andthe non-canonical Wnt/planar cell polarity pathway (Muley etal., 2010). In in vitro experiments, the sFRP1/FrzA mRNAlevel was high in the growing endothelial cells and wasdecreased in the confluent cells (Duplàa et al., 1999). Therate of endothelial cell proliferation was also inhibited by FrzA(Duplàa et al., 1999). However, in vivo studies of sFRP1/FrzAin angiogenesis showed conflict results. sFRP1/FrzA wasfound to inhibit angiogenesis in a hepatocellular carcinomamodel, while inducing angiogenesis in a chick chorioallantoicmembrane model (Dufourcq et al., 2002; Hu et al., 2009). Thedisparity may be ascribed to its concentration-dependentWnt-regulation activities.

Wnt inhibitory factor 1 (WIF-1)

WIF-1 is another Wnt inhibitor which binds to Wnt ligands,similar to sFRPs. In 1999, it was first identified to inhibit thesomitogenesis in Xenopus embryo development, and it wasfound to interact with Drosophila Wingless and XenopusWnt8 (Hsieh et al., 1999). WIF-1 does not share the CRDsequence with Fz or sFRPs, however, it has a highlyconserved N-terminal domain named WIF domain (WD),containing 150 amino acids. WIF-1 also contains a hydro-philic domain at C terminus with 45 amino acids, and fiveepidermal growth factor (EGF)-like repeats. The downregu-lated expression levels of WIF-1 were correlated withprostate, breast, lung and bladder cancers, suggesting an

important role of WIF-1 in tumorigenesis (Wissmann et al.,2003).

Treatment in hepatocellular carcinoma with WIF-1 resultedin reduced microvessel density as well as the inhibition oftumor growth (Hu et al., 2009). The Wnt-regulated pro-angiogenic factors, such as stromal cell-derived factor-1(SDF-1) and VEGF, were also downregulated at the sametime. Further experiments demonstrated that WIF-1 signifi-cantly inhibited migration and tube formation in culturedhuman microvascular endothelial cells. In mouse endothelialprogenitor cells (EPC), WIF-1 not only inhibited cell migrationand tube formation, but also obstructed EPC differentiationand even induced EPC apoptosis (Hu et al., 2009). Therefore,the anti-angiogenic effect of WIF-1 is derived from its Wnt-antagonizing function and may play roles in the anti-tumoractivity.

Dickkopf (Dkk) family

The human Dkk family is composed of four members, Dkk-1,2, 3 and 4, and a unique Dkk-3-related protein which is namedSoggy (Sgy). They are secreted proteins containing ~250amino acids, except for Dkk-3 which has 350 amino acids.There are two conserved cysteine-rich domains (Cys-1 andCys-2) in Dkks, containing 10 cysteine residues in eachposition (Krupnik et al., 1999). However, Dkk-3 related Sgylacks cysteine-rich domains (Krupnik et al., 1999). In theXenopus embryo development assay, Dkk-1 and Dkk-4 inhibitthe Wnt-induced secondary axis formation, while Dkk-2,Dkk-3 and Sgy lack this activity (Krupnik et al., 1999; Mao andNiehrs, 2003).

In the Dkk family, Dkk-1 is the most studied member. It wasfirst identified to be a secreted inducer of Spemann’sorganizer in Xenopus embryo in 1998 (Glinka et al., 1998).Dkk-1 is essential for head formation and also a potent Wntsignaling antagonist. Dkk-1 inhibited the effect of Xwnt-8 inXenopus embryo, but did not block Dishevelled (Dvl/Ssh)-induced secondary axis formation, suggesting a potentialupstream target in the Wnt signaling pathway (Glinka et al.,1998). The binding partner of Dkk-1 was not identified until2001. Two groups established the receptor for Dkk-1 almostat the same time, which are LRP5/6, the co-receptors of Wntproteins (Mao et al., 2001; Semënov et al., 2001). The bindingof Dkk-1 to LRP5/6 results in the blockage of Fz-LRPassociation and subsequently, specific inhibition of the Wnt/β-catenin pathway (Semënov et al., 2001). In addition toLRP5/6, Kremen1 (Krm1) and Kremen2 (Krm2) are also thehigh-affinity receptors for Dkk-1 and can cooperate with Dkk-1to inhibit Wnt signaling (Mao et al., 2002). Krm2, Dkk-1 andLRP6 have been reported to form a ternary complex andinduce rapid endocytosis and removal of the Wnt receptorLRP6 from the plasma membrane (Mao et al., 2002). In Dkk-1, the C-terminal fragment containing the Cys-2 domain bindsto LRP6 and Krm and further inhibits the Wnt signaling

Protein & Cell


Bin Zhang and Jian-xing Ma

activation as shown by the Xenopus embryo secondary axisinduction and the LEF-1 luciferase reporter assays (Brott andSokol, 2002; Li et al., 2002; Mao and Niehrs, 2003). Itsuggests that the C-terminal domain of Dkks is moreimportant for Wnt/LRP6 regulation. However, whether Dkk-1can induce LRP6 internalization is still controversial. In 2008,a group reported that Dkk-1 and Wnt-3a induced internaliza-tion of LRP6 through distinct pathways (Yamamoto et al.,2008). At the same time, another group reported that Dkk-1blocked Wnt signaling but did not promote LRP6 internaliza-tion and degradation (Semënov et al., 2008). Therefore, themechanism for the antagonizing effect of Dkk-1 on LRP6 andWnt/β-Catenin pathway remains unclear.

Dkk-4 has displayed the Wnt antagonist activity similar toDkk-1 (Krupnik et al., 1999). However, the binding betweenDkk-4 and LRP6 has not been demonstrated. Dkk-3 has noinhibitory effect onWnt signaling and does not bind to LRPs orKrm1/2 (Krupnik et al., 1999; Mao et al., 2001, 2002; Mao andNiehrs, 2003). Dkk-2 binds to LRP6 with a lower affinity andthe binding Kd value is approximately 2 folds of that for Dkk-1(0.73 nM vs 0.34 nM) (Mao et al., 2001). This may be one ofthe reasons explaining why Dkk-2 shows conflicting functionsin the Xwnt8-induced axis duplication by different groups. Twogroups reported that Dkk-2 is a poor inhibitor of Wnt signaling(Krupnik et al., 1999; Wu et al., 2000), while another groupreported the Wnt antagonizing activity of Dkk-2 (Brott andSokol, 2002). This disparity may be ascribed to differentexpression levels of Dkk-2 in their experiments. The Wntantagonist activity of Dkk-2 is from C-terminal domaininvolving Cys-2, same as Dkk-1 (Brott and Sokol, 2002; Liet al., 2002; Mao and Niehrs, 2003). However, the N-terminaldomains in Dkk-1 and Dkk-2 have different functions. The N-terminal fragment of Dkk-2 synergies with LRP6 to induceWnt signaling activation (Brott and Sokol, 2002), but the N-terminal domain of Dkk-1 has no such function. Together withother evidence, Dkk-2 is suggested to play the role as anagonist in low-Wnt/high-LRP6 condition, and acts as anantagonist in environment with high-Wnt levels (Wu et al.,2000; Mao et al., 2001; Brott and Sokol, 2002; Li et al., 2002;Mao et al., 2002; Mao and Niehrs, 2003).

As a Wnt-mediated event, angiogenesis is predicted to beregulated by Wnt-antagonizing Dkks, especially Dkk-1, whichis the most studied and potent family member. Theameliorative effect of Dkk-1 has been found in the ocularpathological neovascularization with ectopically higher acti-vation level of Wnt signaling in the eye (Chen et al., 2007). Inthe eyecups of the very low-density lipoprotein receptor(VLDLR) gene knockout (Vldlr–/–) mice, expression levels ofWnt co-receptor LRP5/6 were significantly upregulated,concomitant with accumulation of β-catenin, suggesting anover activation of the Wnt pathway. Knockdown of VLDLR incultured endothelial cells by siRNA also upregulated LRP5/6expression and stabilized β-catenin. Consistent with therole of Wnt signaling in angiogenesis, subretinal

neovascularization develops in Vldlr–/– mice. This phenotypepresents similarly to the neovasculazation in patients with wetAMD. This subretinal neovascularization correlates withoverexpression of VEGF in Vldlr–/– mice. In this abnormalWnt-induced angiogenic model, Dkk-1 effectively decreasedVEGF and β-catenin levels in the RPE of Vldlr–/– mice and incultured cells with VLDLR knockdown (Chen et al., 2007).Recently, another study found that the Wnt pathway is overactivated in the retinas of the OIR mouse model, correlatingwith retinal neovascularization. Further, Dkk-1 also effectivelyattenuated retinal neovascularization as well as Wnt signaling(Chen et al., 2009). These mouse studies provide newevidence to support the anti-angiogenic role of Dkk-1 in ocularangiogenesis.

ANTI-ANGIOGENIC FACTORS AS WNT

ANTAGONISTS

Some Wnts act as pro-angiogenic factors and some Wntantagonists play anti-angiogenic roles. Furthermore, severalendogenous anti-angiogenic factors have also been reportedto inhibit Wnt signaling activity, such as endostatin andSERPINA3K, possible new mechanisms for anti-angiogenicfunctions of these factors.

Endostatin

Endostatin is a C-terminal fragment of collagen XVIII and hasbeen identified to be an anti-angiogenic factor (O'Reilly et al.,1997; Hanai et al., 2002). It is a 20 kDa peptide which caninhibit endothelial cell proliferation and tumor growth (O'Reillyet al., 1997). Endostatin is involved in many aspects ofembryo development. In a Xenopus embryo developmentalstudy, endostatin overexpression blocked the Xenopus axisduplication induced by β-catenin, indicating an inhibitoryeffect of endostatin on the Wnt signaling pathway (Hanai etal., 2002). Endostatin inhibited β-catenin-induced TCF/LEFtranscription activity, but failed to inhibit the effect ofTCF-VP16 (a constitutive downstream activator, independentof β-catenin) (Hanai et al., 2002). At the same time,suppression of endothelial cell migration and inhibition ofthe cell cycle by endostatin were reversed by TCF-VP16(Hanai et al., 2002). It suggested that the potential target forendostatin was at the upstream of TCF, but not earlier than β-catenin. Actually, endostatin can degrade both normal and“stabilized” (no N-terminal phosphorylation sites) forms of β-catenin, revealing a GSK3-independent β-catenin-degrada-tion signaling pathway (Hanai et al., 2002). This mechanismof endostatin is distinct from other secreted Wnt-antagoniststhat act at the Wnt ligand/receptor level. This novel findingalso suggests that the anti-angiogenic activity of endostatinmay act through inhibition of TCF-dependent, canonical Wnt-mediated transcription.



SERPINA3K

SERPINA3K was first identified as a specific inhibitor of tissuekallikrein and thus named kallikrein binding protein (Chao etal., 1986, 1990). Amino acid sequence analysis classifiedSERPINA3K into the serine proteinase inhibitor (serpin)family (Gettins, 2002). Tissue kallikrein is a serine proteinaseand releases bioactive kinins from kininogens (Clements,1989; Murray et al., 1990). SERPINA3K specifically binds totissue kallikrein, forming a covalent complex and inhibitsproteolytic activities of tissue kallikrein (Chao et al., 1990; Maet al., 1995). Later studies suggest that SERPINA3K hasfunctions in addition to the inhibition of tissue kallikrein.SERPINA3K has been found to inhibit angiogenesis and toreduce vascular permeability (Miao et al., 2002; Gao et al.,2003). These anti-angiogenic effects of SERPINA3K havebeen shown to be independent of its interactions with thekallikrein-kinin system (Gao et al., 2003; Zhang et al., 2008).

As an angiogenic inhibitor, SERPINA3K was found todownregulate several pro-angiogenic cytokines such asCTGF and VEGF, which are also target genes of the Wntpathway (Zhang et al., 2009, 2010a, b). In experimentaldiabetic models, SERPINA3K blocked Wnt signaling activa-tion in the retinas with microvascular complications and incells treated with high glucose, suggesting that SERPINA3Kmay regulate angiogenesis through interacting with the Wntpathway (Zhang et al., 2010a, b). Further, SERPINA3K wasidentified as a Wnt antagonist since it blocked the Wnt ligand-induced Wnt pathway activation, blocking phosphorylation ofLRP6, cytosolic β-catenin stabilization, TCF/LEF-mediatedtranscription activity and Xenopus axis duplication. Co-precipitation and ligand binding assay showed thatSERPINA3K binds to LRP6 with a Kd of 10 nM. Under thesame conditions, SERPINA3K did not bind to the Fz receptoror low-density lipoprotein receptor. The interaction betweenSERPINA3K and the extracellular domain of LRP6 blockedFz/LRP6 (receptor/co-receptor) dimerization induced by aWnt ligand. Together with the pro-angiogenic role of Wntsignaling, this study suggests that the antagonizing activity ofSERPINA3K to LRP6 is responsible, at least in part, for itsanti-angiogenic activities.

THE WNT PATHWAY OFFERS MULTIPLE TARGETS

FOR THERAPEUTIC COMPOUNDS

Since Wnt signaling displays potent angiogenic activities,components of the Wnt pathway have been consideredpromising drug targets for the treatment of neovasculardisorders. In addition to the natural Wnt antagonist proteins,more and more small compounds have been developed toinhibit the Wnt pathway through various pathway targets (Reyand Ellies). One of the advantages of the drugs blocking theWnt/β-catenin pathway is their potential to target multiplepotential factors. The agonists and antagonists of the Wnt

receptor/co-receptor can be used to regulate Wnt signalingfrom the extracellular side. Small molecular compounds canbe designed to target the intracellular components of the Wntpathway, such as β-catenin, some proteases and axin1. Thephosphorylation sites on LRP6, DVL, GSK3 and the relativekinase activities all represent potential drug targets. Drugscan also be designed to regulate nuclear translocation andnuclear protein binding activity of β-catenin. Recently, JNKand AKTwere found to phosphorylate β-catenin and enhanceβ-catenin nuclear translocation (Wu et al., 2008). Therefore,the inhibitors targeting JNK and AKTcan potentially inhibit thecanonical Wnt pathway.

Curcumin

Curcumin is a small compound isolated from the spiceturmeric, and it is a member in the ginger family. Curcuminwas shown to inhibit cancer (Conney et al., 1991; Huang etal., 1994), and was later reported to function as an angiogenicinhibitor (Arbiser et al., 1998). Since this report, the anti-angiogenic effect of curcumin and its potential mechanism ofaction have been studied extensively in different models andorgans (Adams et al., 2004; Furness et al., 2005; Bhandarkarand Arbiser, 2007; Kunnumakkara et al., 2008; Varinska et al.,2010). Several signaling transduction pathways have beenfound to be regulated by curcumin, including protein kinase C,NF-kappaB and AP-1 (Bhandarkar and Arbiser, 2007).Recently, curcumin has been identified to be an inhibitor ofthe Wnt/β-catenin pathway. It can induce β-catenin degrada-tion via increasing caspase activity, and further plays anti-angiogenic roles (Park et al., 2005; Leow et al., 2009).Another group reported that the natural derivatives ofcurcumin also inhibited the Wnt/β-catenin pathway, but it isthrough downregulation of p300, one of the transcriptional co-activators (Ryu et al., 2008). These studies suggest that theWnt-antagonizing activity of curcumin can contribute to itsanti-cancer and anti-angiogenesis effects.

SUMMARY

Most of the classical Wnt antagonists were first identified fromXenopus embryo, and their functions were only known indevelopment. Recently, more andmoreWnt antagonists havebeen reported from other protein families, such as the anti-angiogenic factor family, directly implying the Wnt pathway inpathogenesis of angiogenic disorders. At the same time,classical Wnt antagonists have also been found to inhibitangiogenesis, suggesting the potential application of the Wntantagonists in anti-angiogenic therapies. Small moleculecompounds have the advantages in the high-throughputscreening and drug delivery. There are multiple drug targets inthe Wnt pathway for compound design, which may representa major effort in the future drug development to treatangiogenic disorders.

Protein & Cell



REFERENCES

Adams, B.K., Ferstl, E.M., Davis, M.C., Herold, M., Kurtkaya, S.,

Camalier, R.F., Hollingshead, M.G., Kaur, G., Sausville, E.A.,Rickles, F.R., et al. (2004). Synthesis and biological evaluation ofnovel curcumin analogs as anti-cancer and anti-angiogenesisagents. Bioorg Med Chem 12, 3871–3883.

Arbiser, J.L., Klauber, N., Rohan, R., van Leeuwen, R., Huang, M.T.,Fisher, C., Flynn, E., and Byers, H.R. (1998). Curcumin is an in vivo

inhibitor of angiogenesis. Mol Med 4, 376–383.

Bányai, L., and Patthy, L. (1999). The NTR module: domains ofnetrins, secreted frizzled related proteins, and type I procollagen C-

proteinase enhancer protein are homologous with tissue inhibitorsof metalloproteases. Protein Sci 8, 1636–1642.

Behrens, J. (2000). Cross-regulation of the Wnt signalling pathway: a

role of MAP kinases. J Cell Sci 113, 911–919.

Behrens, J., von Kries, J.P., Kühl, M., Bruhn, L., Wedlich, D.,

Grosschedl, R., and Birchmeier, W. (1996). Functional interactionof beta-catenin with the transcription factor LEF-1. Nature 382,638–642.

Bhandarkar, S.S., and Arbiser, J.L. (2007). Curcumin as an inhibitor ofangiogenesis. Adv Exp Med Biol 595, 185–195.

Brabletz, T., Jung, A., Dag, S., Hlubek, F., and Kirchner, T. (1999).

beta-catenin regulates the expression of the matrixmetalloproteinase-7 in human colorectal cancer. Am J Pathol 155,1033–1038.

Brott, B.K., and Sokol, S.Y. (2002). Regulation of Wnt/LRP signalingby distinct domains of Dickkopf proteins. Mol Cell Biol 22,6100–6110.

Chamorro, M.N., Schwartz, D.R., Vonica, A., Brivanlou, A.H., Cho, K.R., and Varmus, H.E. (2005). FGF-20 and DKK1 are transcriptional

targets of beta-catenin and FGF-20 is implicated in cancer anddevelopment. EMBO J 24, 73–84.

Chao, J., Chai, K.X., Chen, L.M., Xiong, W., Chao, S., Woodley-Miller,

C., Wang, L.X., Lu, H.S., and Chao, L. (1990). Tissue kallikrein-binding protein is a serpin. I. Purification, characterization, anddistribution in normotensive and spontaneously hypertensive rats.

J Biol Chem 265, 16394–16401.

Chao, J., Tillman, D.M., Wang, M.Y., Margolius, H.S., and Chao, L.(1986). Identification of a new tissue-kallikrein-binding protein.

Biochem J 239, 325–331.

Chen, Y., Hu, Y., Lu, K., Flannery, J.G., and Ma, J.X. (2007). Very lowdensity lipoprotein receptor, a negative regulator of the wnt

signaling pathway and choroidal neovascularization. J Biol Chem282, 34420–34428.

Chen, Y., Hu, Y., Zhou, T., Zhou, K.K., Mott, R., Wu, M., Boulton, M.,Lyons, T.J., Gao, G., and Ma, J.X. (2009). Activation of the Wntpathway plays a pathogenic role in diabetic retinopathy in humansand animal models. Am J Pathol 175, 2676–2685.

Cheng, C.W., Smith, S.K., and Charnock-Jones, D.S. (2003). Wnt-1signaling inhibits human umbilical vein endothelial cell proliferation

and alters cell morphology. Exp Cell Res 291, 415–425.

Clements, J.A. (1989). The glandular kallikrein family of enzymes:tissue-specific expression and hormonal regulation. Endocr Rev

10, 393–419.

Clevers, H. (2006). Wnt/beta-catenin signaling in development anddisease. Cell 127, 469–480.

Conney, A.H., Lysz, T., Ferraro, T., Abidi, T.F., Manchand, P.S.,

Laskin, J.D., and Huang, M.T. (1991). Inhibitory effect of curcuminand some related dietary compounds on tumor promotion andarachidonic acid metabolism in mouse skin. Adv Enzyme Regul31, 385–396.

Dale, T.C. (1998). Signal transduction by the Wnt family of ligands.Biochem J 329, 209–223.

Dufourcq, P., Couffinhal, T., Ezan, J., Barandon, L., Moreau, C.,Daret, D., and Duplàa, C. (2002). FrzA, a secreted frizzled relatedprotein, induced angiogenic response. Circulation 106,

3097–3103.

Duplàa, C., Jaspard, B., Moreau, C., and D’Amore, P.A. (1999).Identification and cloning of a secreted protein related to the

cysteine-rich domain of frizzled. Evidence for a role in endothelialcell growth control. Circ Res 84, 1433–1445.

Easwaran, V., Lee, S.H., Inge, L., Guo, L., Goldbeck, C., Garrett, E.,Wiesmann, M., Garcia, P.D., Fuller, J.H., Chan, V., et al. (2003).beta-Catenin regulates vascular endothelial growth factor expres-sion in colon cancer. Cancer Res 63, 3145–3153.

Fanto, M., and McNeill, H. (2004). Planar polarity from flies tovertebrates. J Cell Sci 117, 527–533.

Finch, P.W., He, X., Kelley, M.J., Uren, A., Schaudies, R.P., Popescu,N.C., Rudikoff, S., Aaronson, S.A., Varmus, H.E., and Rubin, J.S.(1997). Purification and molecular cloning of a secreted, Frizzled-

related antagonist of Wnt action. Proc Natl Acad Sci U S A 94,6770–6775.

Furness, M.S., Robinson, T.P., Ehlers, T., Hubbard, R.B. 4th, Arbiser,

J.L., Goldsmith, D.J., and Bowen, J.P. (2005). Antiangiogenicagents: studies on fumagillin and curcumin analogs. Curr PharmDes 11, 357–373.

Gao, G., Shao, C., Zhang, S.X., Dudley, A., Fant, J., and Ma, J.X.(2003). Kallikrein-binding protein inhibits retinal neovascularizationand decreases vascular leakage. Diabetologia 46, 689–698.

Gettins, P.G. (2002). Serpin structure, mechanism, and function.Chem Rev 102, 4751–4804.

Glinka, A., Wu, W., Delius, H., Monaghan, A.P., Blumenstock, C., andNiehrs, C. (1998). Dickkopf-1 is a member of a new family ofsecreted proteins and functions in head induction. Nature 391,

357–362.

Guo, N., Hawkins, C., and Nathans, J. (2004). Frizzled6 controls hairpatterning in mice. Proc Natl Acad Sci U S A 101, 9277–9281.

Hanai, J., Gloy, J., Karumanchi, S.A., Kale, S., Tang, J., Hu, G., Chan,B., Ramchandran, R., Jha, V., Sukhatme, V.P., et al. (2002).

Endostatin is a potential inhibitor of Wnt signaling. J Cell Biol 158,529–539.

He, T.C., Sparks, A.B., Rago, C., Hermeking, H., Zawel, L., da Costa,

L.T., Morin, P.J., Vogelstein, B., and Kinzler, K.W. (1998).Identification of c-MYC as a target of the APC pathway. Science281, 1509–1512.

He, X., Saint-Jeannet, J.P., Wang, Y., Nathans, J., Dawid, I., andVarmus, H. (1997). A member of the Frizzled protein familymediating axis induction by Wnt-5A. Science 275, 1652–1654.

He, X., Semenov, M., Tamai, K., and Zeng, X. (2004). LDL receptor-related proteins 5 and 6 in Wnt/beta-catenin signaling: arrows pointthe way. Development131, 1663–1677.

Hoang, B., Moos, M. Jr, Vukicevic, S., and Luyten, F.P. (1996).Primary structure and tissue distribution of FRZB, a novel protein

related to Drosophila frizzled, suggest a role in skeletal morpho-



genesis. J Biol Chem 271, 26131–26137.

Howe, L.R., Subbaramaiah, K., Chung, W.J., Dannenberg, A.J., and

Brown, A.M. (1999). Transcriptional activation of cyclooxygenase-2 in Wnt-1-transformed mouse mammary epithelial cells. CancerRes 59, 1572–1577.

Hsieh, J.C., Kodjabachian, L., Rebbert, M.L., Rattner, A., Smallwood,P.M., Samos, C.H., Nusse, R., Dawid, I.B., and Nathans, J. (1999).A new secreted protein that binds to Wnt proteins and inhibits their

activities. Nature 398, 431–436.

Hu, J., Dong, A., Fernandez-Ruiz, V., Shan, J., Kawa, M., Martínez-Ansó, E., Prieto, J., and Qian, C. (2009). Blockade of Wnt signaling

inhibits angiogenesis and tumor growth in hepatocellular carci-noma. Cancer Res 69, 6951–6959.

Huang, M.T., Lou, Y.R., Ma, W., Newmark, H.L., Reuhl, K.R., and

Conney, A.H. (1994). Inhibitory effects of dietary curcumin onforestomach, duodenal, and colon carcinogenesis in mice. CancerRes 54, 5841–5847.

Ishitani, T., Kishida, S., Hyodo-Miura, J., Ueno, N., Yasuda, J.,Waterman, M., Shibuya, H., Moon, R.T., Ninomiya-Tsuji, J., andMatsumoto, K. (2003). The TAK1-NLK mitogen-activated protein

kinase cascade functions in the Wnt-5a/Ca(2+) pathway toantagonize Wnt/beta-catenin signaling. Mol Cell Biol 23, 131–139.

Jones, S.E., and Jomary, C. (2002). Secreted Frizzled-relatedproteins: searching for relationships and patterns. Bioessays 24,811–820.

Kawano, Y., and Kypta, R. (2003). Secreted antagonists of the Wntsignalling pathway. J Cell Sci 116, 2627–2634.

Kohn, E.C., Alessandro, R., Spoonster, J., Wersto, R.P., and Liotta, L.

A. (1995). Angiogenesis: role of calcium-mediated signal transduc-tion. Proc Natl Acad Sci U S A 92, 1307–1311.

Krupnik, V.E., Sharp, J.D., Jiang, C., Robison, K., Chickering, T.W.,Amaravadi, L., Brown, D.E., Guyot, D., Mays, G., Leiby, K., et al.(1999). Functional and structural diversity of the human Dickkopfgene family. Gene 238, 301–313.

Kühl, M., Sheldahl, L.C., Malbon, C.C., and Moon, R.T. (2000).Ca(2+)/calmodulin-dependent protein kinase II is stimulated by

Wnt and Frizzled homologs and promotes ventral cell fates inXenopus. J Biol Chem 275, 12701–12711.

Kühl, M., Sheldahl, L.C., Park, M., Miller, J.R., and Moon, R.T. (2000).

The Wnt/Ca2+ pathway: a new vertebrate Wnt signaling pathwaytakes shape. Trends Genet 16, 279–283.

Kunnumakkara, A.B., Anand, P., and Aggarwal, B.B. (2008).

Curcumin inhibits proliferation, invasion, angiogenesis and metas-tasis of different cancers through interaction with multiple cellsignaling proteins. Cancer Lett 269, 199–225.

Ladher, R.K., Church, V.L., Allen, S., Robson, L., Abdelfattah, A.,Brown, N.A., Hattersley, G., Rosen, V., Luyten, F.P., Dale, L., et al.(2000). Cloning and expression of the Wnt antagonists Sfrp-2 and

Frzb during chick development. Dev Biol 218, 183–198.

Leow, P.C., Tian, Q., Ong, Z.Y., Yang, Z., and Ee, P.L. (2009).

Antitumor activity of natural compounds, curcumin andPKF118–310, as Wnt/beta-catenin antagonists against humanosteosarcoma cells. Invest New Drugs. In press.

Leyns, L., Bouwmeester, T., Kim, S.H., Piccolo, S., and De Robertis,E.M. (1997). Frzb-1 is a secreted antagonist of Wnt signalingexpressed in the Spemann organizer. Cell 88, 747–756.

Li, A., Dubey, S., Varney, M.L., Dave, B.J., and Singh, R.K. (2003). IL-8 directly enhanced endothelial cell survival, proliferation, and

matrix metalloproteinases production and regulated angiogenesis.J Immunol 170, 3369–3376.

Li, L., Mao, J., Sun, L., Liu, W., and Wu, D. (2002). Second cysteine-rich domain of Dickkopf-2 activates canonical Wnt signalingpathway via LRP-6 independently of dishevelled. J Biol Chem277, 5977–5981.

Lin, K., Wang, S., Julius, M.A., Kitajewski, J., Moos, M. Jr, and Luyten,F.P. (1997). The cysteine-rich frizzled domain of Frzb-1 is required

and sufficient for modulation of Wnt signaling. Proc Natl Acad SciU S A 94, 11196–11200.

Logan, C.Y., and Nusse, R. (2004). The Wnt signaling pathway in

development and disease. Annu Rev Cell Dev Biol 20, 781–810.

Ma, J.X., Yang, Z., Chao, J., and Chao, L. (1995). Intramusculardelivery of rat kallikrein-binding protein gene reverses hypotension

in transgenic mice expressing human tissue kallikrein. J Biol Chem270, 451–455.

Mann, B., Gelos, M., Siedow, A., Hanski, M.L., Gratchev, A., Ilyas, M.,Bodmer, W.F., Moyer, M.P., Riecken, E.O., Buhr, H.J., et al. (1999).Target genes of beta-catenin-T cell-factor/lymphoid-enhancer-factor signaling in human colorectal carcinomas. Proc Natl Acad

Sci U S A 96, 1603–1608.

Mao, B., and Niehrs, C. (2003). Kremen2 modulates Dickkopf2

activity during Wnt/LRP6 signaling. Gene 302, 179–183.

Mao, B., Wu, W., Davidson, G., Marhold, J., Li, M., Mechler, B.M.,Delius, H., Hoppe, D., Stannek, P., Walter, C., et al. (2002). Kremen

proteins are Dickkopf receptors that regulate Wnt/beta-cateninsignalling. Nature 417, 664–667.

Mao, B., Wu, W., Li, Y., Hoppe, D., Stannek, P., Glinka, A., and Niehrs,

C. (2001). LDL-receptor-related protein 6 is a receptor for Dickkopfproteins. Nature 411, 321–325.

Masckauchán, T.N., Agalliu, D., Vorontchikhina, M., Ahn, A.,Parmalee, N.L., Li, C.M., Khoo, A., Tycko, B., Brown, A.M., andKitajewski, J. (2006). Wnt5a signaling induces proliferation andsurvival of endothelial cells in vitro and expression of MMP-1 and

Tie-2. Mol Biol Cell 17, 5163–5172.

Masckauchán, T.N., and Kitajewski, J. (2006). Wnt/Frizzled signaling

in the vasculature: new angiogenic factors in sight. Physiology(Bethesda) 21, 181–188.

Masckauchán, T.N., Shawber, C.J., Funahashi, Y., Li, C.M., and

Kitajewski, J. (2005). Wnt/beta-catenin signaling induces prolifera-tion, survival and interleukin-8 in human endothelial cells.Angiogenesis 8, 43–51.

Melkonyan, H.S., Chang, W.C., Shapiro, J.P., Mahadevappa, M.,Fitzpatrick, P.A., Kiefer, M.C., Tomei, L.D., and Umansky, S.R.(1997). SARPs: a family of secreted apoptosis-related proteins.

Proc Natl Acad Sci U S A 94, 13636–13641.

Miao, R.Q., Agata, J., Chao, L., and Chao, J. (2002). Kallistatin is anew inhibitor of angiogenesis and tumor growth. Blood 100,

3245–3252.

Mlodzik, M. (2002). Planar cell polarization: do the samemechanisms

regulate Drosophila tissue polarity and vertebrate gastrulation?Trends Genet 18, 564–571.

Monkley, S.J., Delaney, S.J., Pennisi, D.J., Christiansen, J.H., and

Wainwright, B.J. (1996). Targeted disruption of the Wnt2 generesults in placentation defects. Development 122, 3343–3353.

Muley, A., Majumder, S., Kolluru, G.K., Parkinson, S., Viola, H., Hool,

L., Arfuso, F., Ganss, R., Dharmarajan, A., and Chatterjee, S.(2010). Secreted frizzled-related protein 4: an angiogenesis

Protein & Cell



inhibitor. Am J Pathol 176, 1505–1516.

Murray, S.R., Chao, J., Lin, F.K., and Chao, L. (1990). Kallikrein

multigene families and the regulation of their expression. JCardiovasc Pharmacol 15, S7–S16.

Nakamura, T., Tsuchiya, K., and Watanabe, M. (2007). Crosstalk

between Wnt and Notch signaling in intestinal epithelial cell fatedecision. J Gastroenterol 42, 705–710.

Nusse, R. (2005). Wnt signaling in disease and in development. Cell

Res 15, 28–32.

Nusse, R., van Ooyen, A., Cox, D., Fung, Y.K., and Varmus, H.

(1984). Mode of proviral activation of a putative mammaryoncogene (int-1) on mouse chromosome 15. Nature 307, 131–136.

O’Reilly, M.S., Boehm, T., Shing, Y., Fukai, N., Vasios, G., Lane, W.S.,

Flynn, E., Birkhead, J.R., Olsen, B.R., and Folkman, J. (1997).Endostatin: an endogenous inhibitor of angiogenesis and tumorgrowth. Cell 88, 277–285.

Ojesina, A.I. (2004). The role of beta-catenin in regulating angiogen-esis in Wilms tumor. J Pediatr Surg 39, 1446–1447.

Orsulic, S., and Peifer, M. (1996). Cell-cell signalling: Wingless landsat last. Curr Biol 6, 1363–1367.

Park, C.H., Hahm, E.R., Park, S., Kim, H.K., and Yang, C.H. (2005).

The inhibitory mechanism of curcumin and its derivative againstbeta-catenin/Tcf signaling. FEBS Lett 579, 2965–2971.

Polakis, P. (2002). Casein kinase 1: a Wnt’er of disconnect. Curr Biol

12, R499–R501.

Povelones, M., Howes, R., Fish, M., and Nusse, R. (2005). Genetic

evidence that Drosophila frizzled controls planar cell polarity andArmadillo signaling by a common mechanism. Genetics 171,1643–1654.

Prieve, M.G., and Moon, R.T. (2003). Stromelysin-1 and mesothelinare differentially regulated by Wnt-5a and Wnt-1 in C57mg mousemammary epithelial cells. BMC Dev Biol 3, 2.

Rattner, A., Hsieh, J.C., Smallwood, P.M., Gilbert, D.J., Copeland, N.G., Jenkins, N.A., and Nathans, J. (1997). A family of secretedproteins contains homology to the cysteine-rich ligand-binding

domain of frizzled receptors. Proc Natl Acad Sci U S A 94,2859–2863.

Rey, J.P., and Ellies, D.L. (2010). Wnt modulators in the biotech

pipeline. Dev Dyn 239, 102–114.

Rijsewijk, F., Schuermann, M., Wagenaar, E., Parren, P., Weigel, D.,and Nusse, R. (1987). The Drosophila homolog of the mouse

mammary oncogene int-1 is identical to the segment polarity genewingless. Cell 50, 649–657.

Robitaille, J., MacDonald, M.L., Kaykas, A., Sheldahl, L.C., Zeisler, J.,Dubé, M.P., Zhang, L.H., Singaraja, R.R., Guernsey, D.L., Zheng,B., et al. (2002). Mutant frizzled-4 disrupts retinal angiogenesis in

familial exudative vitreoretinopathy. Nat Genet 32, 326–330.

Ryu, M.J., Cho, M., Song, J.Y., Yun, Y.S., Choi, I.W., Kim, D.E., Park,B.S., and Oh, S. (2008). Natural derivatives of curcumin attenuate

the Wnt/beta-catenin pathway through down-regulation of thetranscriptional coactivator p300. Biochem Biophys Res Commun377, 1304–1308.

Semënov, M.V., Tamai, K., Brott, B.K., Kühl, M., Sokol, S., and He, X.(2001). Head inducer Dickkopf-1 is a ligand for Wnt coreceptorLRP6. Curr Biol 11, 951–961.

Semënov, M.V., Zhang, X., and He, X. (2008). DKK1 antagonizesWntsignaling without promotion of LRP6 internalization and degrada-tion. J Biol Chem 283, 21427–21432.

Sharma, R.P., and Chopra, V.L. (1976). Effect of the Wingless (wg1)

mutation on wing and haltere development in Drosophila melano-gaster. Dev Biol 48, 461–465.

Sheldahl, L.C., Park, M., Malbon, C.C., and Moon, R.T. (1999).

Protein kinase C is differentially stimulated by Wnt and Frizzledhomologs in a G-protein-dependent manner. Curr Biol 9, 695–698.

Shi, Y., He, B., You, L., and Jablons, D.M. (2007). Roles of secreted

frizzled-related proteins in cancer. Acta Pharmacol Sin 28,1499–1504.

Shimokawa, T., Furukawa, Y., Sakai, M., Li, M., Miwa, N., Lin, Y.M.,and Nakamura, Y. (2003). Involvement of the FGF18 gene incolorectal carcinogenesis, as a novel downstream target of thebeta-catenin/T-cell factor complex. Cancer Res 63, 6116–6120.

Shu, W., Jiang, Y.Q., Lu, M.M., and Morrisey, E.E. (2002). Wnt7bregulates mesenchymal proliferation and vascular development in

the lung. Development 129, 4831–4842.

Slusarski, D.C., Corces, V.G., and Moon, R.T. (1997). Interaction ofWnt and a Frizzled homologue triggers G-protein-linked phospha-

tidylinositol signalling. Nature 390, 410–413.

Tachikawa, K., Schröder, O., Frey, G., Briggs, S.P., and Sera, T.(2004). Regulation of the endogenous VEGF-A gene by exogen-

ous designed regulatory proteins. Proc Natl Acad Sci U S A 101,15225–15230.

Tetsu, O., and McCormick, F. (1999). Beta-catenin regulatesexpression of cyclin D1 in colon carcinoma cells. Nature 398,422–426.

Topol, L., Jiang, X., Choi, H., Garrett-Beal, L., Carolan, P.J., and Yang,Y. (2003). Wnt-5a inhibits the canonical Wnt pathway by promotingGSK-3-independent beta-catenin degradation. J Cell Biol 162,

899–908.

Umar, S., Sarkar, S., Wang, Y., and Singh, P. (2009). Functionalcross-talk between beta-catenin and NFkappaB signaling path-

ways in colonic crypts of mice in response to progastrin. J BiolChem 284, 22274–22284.

Uren, A., Reichsman, F., Anest, V., Taylor, W.G., Muraiso, K., Bottaro,

D.P., Cumberledge, S., and Rubin, J.S. (2000). Secreted frizzled-related protein-1 binds directly to Wingless and is a biphasicmodulator of Wnt signaling. J Biol Chem 275, 4374–4382.

van Amerongen, R., Mikels, A., and Nusse, R. (2008). Alternative wntsignaling is initiated by distinct receptors. Sci Signal 1, re9.

Varinska, L., Mirossay, L., Mojzisova, G., and Mojzis, J. (2010).Antiangogenic effect of selected phytochemicals. Pharmazie 65,57–63.

Wang, S., Krinks, M., Lin, K., Luyten, F.P., and Moos, M. Jr. (1997a).Frzb, a secreted protein expressed in the Spemann organizer,binds and inhibits Wnt-8. Cell 88, 757–766.

Wang, S., Krinks, M., and Moos, M. Jr. (1997b). Frzb-1, an antagonistof Wnt-1 and Wnt-8, does not block signaling by Wnts -3A, -5A, or-11. Biochem Biophys Res Commun 236, 502–504.

Wissmann, C., Wild, P.J., Kaiser, S., Roepcke, S., Stoehr, R.,Woenckhaus, M., Kristiansen, G., Hsieh, J.C., Hofstaedter, F.,Hartmann, A., et al. (2003). WIF1, a component of the Wnt

pathway, is down-regulated in prostate, breast, lung, and bladdercancer. J Pathol 201, 204–212.

Wright, M., Aikawa, M., Szeto, W., and Papkoff, J. (1999). Identifica-tion of a Wnt-responsive signal transduction pathway in primaryendothelial cells. Biochem Biophys Res Commun 263, 384–388.

Wu, W., Glinka, A., Delius, H., and Niehrs, C. (2000). Mutual



antagonism between dickkopf1 and dickkopf2 regulates Wnt/beta-catenin signalling. Curr Biol 10, 1611–1614.

Wu, X., Tu, X., Joeng, K.S., Hilton, M.J., Williams, D.A., and Long, F.(2008). Rac1 activation controls nuclear localization of beta-catenin during canonical Wnt signaling. Cell 133, 340–353.

Xu, Q., D’Amore, P.A., and Sokol, S.Y. (1998). Functional andbiochemical interactions of Wnts with FrzA, a secreted Wntantagonist. Development 125, 4767–4776.

Xu, Q., Wang, Y., Dabdoub, A., Smallwood, P.M., Williams, J., Woods,C., Kelley, M.W., Jiang, L., Tasman, W., Zhang, K., et al. (2004).Vascular development in the retina and inner ear: control by Norrin

and Frizzled-4, a high-affinity ligand-receptor pair. Cell 116,883–895.

Yamamoto, H., Sakane, H., Yamamoto, H., Michiue, T., and Kikuchi,

A. (2008). Wnt3a and Dkk1 regulate distinct internalizationpathways of LRP6 to tune the activation of beta-catenin signaling.Dev Cell 15, 37–48.

Zerlin, M., Julius, M.A., and Kitajewski, J. (2008). Wnt/Frizzled

signaling in angiogenesis. Angiogenesis 11, 63–69.

Zhang, B., Abreu, J.G., Zhou, K., Chen, Y., Hu, Y., Zhou, T., He, X.,

and Ma, J.X. (2010a). Blocking the Wnt pathway, a unifyingmechanism for an angiogenic inhibitor in the serine proteinaseinhibitor family. Proc Natl Acad Sci U S A 107, 6900–6905.

Zhang, B., Hu, Y., and Ma, J.X. (2009). Anti-inflammatory andantioxidant effects of SERPINA3K in the retina. Invest OphthalmolVis Sci 50, 3943–3952.

Zhang, B., Ma, J.X., and Cookson, M.R. (2008). SERPINA3Kprevents oxidative stress induced necrotic cell death by inhibitingcalcium overload. PLoS ONE 3, e4077.

Zhang, B., Zhou, K.K., and Ma, J.X. (2010b). Inhibition of connectivetissue growth factor overexpression in diabetic retinopathy bySERPINA3K via blocking the WNT/beta-catenin pathway. Dia-

betes 59, 1809–1816.

Zhang, X., Gaspard, J.P., and Chung, D.C. (2001). Regulation of

vascular endothelial growth factor by the Wnt and K-ras pathwaysin colonic neoplasia. Cancer Res 61, 6050–6054.

Protein & Cell



REVIEW

The late stage of autophagy: cellular eventsand molecular regulation

Jingjing Tong1,2, Xianghua Yan2, Li Yu1✉

1 State Key Laboratory of Biomembrane and Membrane Biotechnology, School of Life Science, Tsinghua University, Beijing100084, China

2 College of Animal Sciences and Technology, Huazhong Agricultural University, Wuhan 430070, China✉ Correspondence: [email protected] October 6, 2010 Accepted October 18, 2010

ABSTRACT

Autophagy is an intracellular degradation system thatdelivers cytoplasmic contents to the lysosome fordegradation. It is a “self-eating” process and plays a“house-cleaner” role in cells. The complex processconsists of several sequential steps—induction, autop-hagosome formation, fusion of lysosome and autopha-gosome, degradation, efflux transportation ofdegradation products, and autophagic lysosome refor-mation. In this review, the cellular and molecular regula-tions of late stage of autophagy, including cellular eventsafter fusion step, are summarized.

KEYWORDS autophagy, autophagosome, lysosome,fusion, degradation

INTRODUCTION

The balance between synthesis and degradation of cellularcomponents is important for cellular homeostasis. In eukar-yote cell, two powerful hydrolytic mechanisms, the protea-some system and the lysosome system, are responsible fordegradation. Autophagy is a lysosome based degradationpathway. In contrast to the ubiquitin-proteasome system,which only degrades ubiquitinated proteins, autophagy canengulf and degrade large portion of cytoplasm in a highlyregulated manner. There are three types of autophagy,macroautophagy, microautophagy, and chaperone-mediatedautophagy (Klionsky, 2007). In this review we will focus onmacroautophagy, hereafter referred to as autophagy. Basedon the recent studies, the whole process of autophagyproceeds through the following sequential steps: 1. induction

of autophagy, 2. formation of autophagosome precursor, 3.formation of autophagosome, 4. fusion between autophago-some and lysosome, 5. degradation of the contents, 6.release of the degradation products to the cytoplasm, 7.reformation of lysosome.

In eukaryotic cells, basal level autophagy occurs in mostcell types in nutrient rich condition for its housekeepingfunctions (Mizushima, 2007). Dramatic upregulation ofautophagy occurs in the presence of various stresses, suchas starvation. mTOR (mammalian target of rapamycin) is oneof the key regulators of cell growth. When nutrients arelimited, mTOR is inactivated, which in turn induces autop-hagy. Autophagy is initiated by the formation or elongation ofisolated membrane (IM). In yeast, the pre-autophagosomalstructure (PAS) serves as the initial site of autophagy-related(Atg) protein recruitment. PAS is not observed in mammaliancells, but recent research found a structure named omego-some on endoplasmic reticulum. These structures provide amembrane platform for accumulation of autophagosomalproteins, expansion of autophagosomal membranes, andeventually formation of autophagosome (Yorimitsu andKlionsky, 2005; Mizushima, 2007; Xie and Klionsky, 2007;Nakatogawa et al., 2009; Hamasaki and Yoshimori, 2010).The origin of autophagosome membranes still remainscontroversial, but in mammalian cells, the endoplasmicreticulum seems to be the major source of autophagosomemembrane (Axe et al., 2008). Autophagosome formationis regulated by a complicated molecular machinery. Amongthem, two ubiquitin-like conjugation systems, Atg12-Atg5conjugation system and Atg8-PE conjugation system, parti-cipate in the expansion of autophagosome. After autophago-some formation, autophagosome will fuse with endosomeand lysosome, to form a hybrid organelle named


Protein Cell 2010, 1(10): 907–915DOI 10.1007/s13238-010-0121-z

Protein & Cell

autolysosome. Inside autolysosome, cellular contentengulfed by autophagosome is degraded. The degradedproducts such as amino acids and monosaccharide arereleased from autolysosome through lysosomal efflux trans-porter. Finally, lysosome components in autolysosome arerecycled through an evolutionary conserved mechanismnamed autophagic lysosome reformation, to form the newfunctional lysosome.

So far, the formation of autophagosome has beensubjected to extensive study and many excellent reviewshave been dedicated to this topic (Longatti and Tooze, 2009;Ravikumar et al., 2009). The goal of this review is to providean overview of recent advances in our understanding ofcellular events after autophagosome formation. For the sakeof clarity, this review divides the continuous cellular eventsfollowing autolysosome formation into three parts: fusionbetween autophagosome and lysosome, degradation inautolysosome and efflux of degradation products, and finally,autophagic lysosome reformation.

FUSION BETWEEN AUTOPHAGOSOMES AND

LYSOSOMES

After autophagosome formation, autophagosomes will fusewith lysosomes or vacuoles to form autolysosomes. Thefusion process can be viewed as three steps: movementtoward the target compartment, tethering/docking with theacceptor membrane, and the eventual fusion of the lipid bi-layers. Fusion events between autophagosomes and lyso-somes have been studied extensively. Studies indicate thatthe fusion between autophagosome and lysosomes inmammalian cells and yeast is a multi-step process regulatedthrough complex molecular machinery.

ESCRT

The endosomal sorting complex required for transport(ESCRT) originally identified for their roles in sortingubiquitinylated membrane proteins into mutivescular bodieshas been found to play an important role in autophagosome-lysosome fusion (Rothman and Wieland, 1996; Rusten et al.,2007; Raiborg and Stenmark, 2009). Under the condition ofESCRT-III inactivation, such as inhibiting the expression ofESCRT subunits through RNA interference or geneticmutation, autophagosome accumulated (Lee et al., 2007).Hrs belongs to ESCRT-0 complex and is a master regulator inendosomal protein sorting (Bache et al., 2002; Lloyd et al.,2002; Kanazawa et al., 2003). It contains a FYVE domain,and is localized to the autophagosomes, and its depletionsignificantly decreases the number of autolysosomes in cells(Tamai et al., 2007). These results indicate ESCRT-0 andESCRT-III complex may be crucial in the autophagosome-lysosome fusion. Also, there are data suggesting that ESCRThas a role in autophagosome formation (Rusten and

Stenmark, 2009). Further study on this mechanism will beneeded to sort out the precise role of ESCRT in autophagy.

Microtubules

Autophagosomes appear to form randomly inside the mam-malian cells. However, in most of cell types, lysosomes areclustered around the microtubule-organizing center (MTOC;located near the nucleus) (Jahreiss et al., 2008), thus fusionevent requires lysosome or autophagosome to move closer.Recent literatures show that autophagosomes move alongmicrotubules toward lysosomes in a dynein-dependentmanner (Ravikumar et al., 2005). In perinuclear region,lysosomes and autophagosomes tether, dock and fuse witheach other. But the mechanism for this directed movement isnot well understood yet. Treating cells with microtubuledisrupting reagents such as nocodazole and vinblastinecauses delay in autophagosome-lysosome fusion (Aplin etal., 1992; Seglen et al., 1996). Conversely, microtubulestabilization by taxol increases the fusion between autophagicvacuoles and lysosomed. In addition, Fass et al. (2006), usingtime-lapse video microscopy, revealed only mature autopha-gosome but not isolation membrane can associate withmicrotubules and move along microtubule tracks. Recentresearch demonstrated autophagosomes move bi-direction-ally along microtubules. However, what drives autophago-somes moving along microtubules? Dynein, which movestoward the minus end of microtubules, is a motor protein (Gillet al., 1991; Schroer and Sheetz, 1991), and its partnerdynactin, which is essential for motor activity, can interact withdynein to form a large complex. Modulating dynein by variousmethods, including using dynein ATPase adenosine deami-nase inhibitor, RNAi or microinjecting anti-dynein intermediatechain antibodies, causes impairment of autophagosomestrafficking and decreases the fusion of these structures,suggesting dynein may be the motor protein that drivesautophagosome along microtubules. Interestingly, Kimura etal. (2008) reported the average velocity of rapid autophago-some movement (5m/sec), which is consistent with thevelocity of dynein motor (4m/sec) (Lakadamyali et al.,2003). So far, it is unknown how dynein interacts withautophagosome. One possibility is that dynein is directly orindirectly associated with LC3. Kimura et al. (2008) demon-strated GFP-LC3 and dynein partially co-localize in autopha-gosome, and anti-LC3 antibody microinjection inhibitsautophagosome movement. Structural study has shown thatthe N-terminal extension in LC3 consists of two-alpha-helices(Sugawara et al., 2004; Kouno et al., 2005) and it binds tomicrotubules (Mann and Hammarback, 1994; Kouno et al.,2005). Microinjection of anti-N-terminal LC3 antibodyabolishes autophagosome movement (Kimura et al., 2008).These data suggest the essential role of LC3 in autophago-some movement is possibly by promoting autophagosomeconnecting to microtubules. In yeast, treating cells with


Jingjing Tong et al.Protein & Cell

nocodazole does not affect autophagy, demonstrating thatmicrotubules are not required for autophagy in yeast.

A novel protein FYCO1 (FYVE and coiled-coil [CC] domaincontaining 1) have been identified by Johansen grouprecently (Pankiv et al., 2010). It can associate with LC3 andRab7 to form an adaptor complex, which promotes autopha-gic vesicles transportation along microtubule plus end.FYCO1 contains an N-terminal RUN domain, 850-aa-longCC region, FYVE domain, LIR (LC3-interacting region) motif,and globular domain. RUN domain is an α-helical protein-protein interaction domain shown to bind to the smallGTPases of Rab and Rap families (Callebaut et al., 2001;Recacha et al., 2009); CC region is responsible for thedimerization of FYCO1 and has many protein binding regions,C-terminal part of the CC is responsible for the co-localizationof FYCO1 with Rab7, FYVE is a PI3P binding domain whichmay take part in the membrane targeting. They have alsodemonstrated that the CC domain in FYCO1 which is requiredfor FYCO1 self-interacting is essential for membrane recruit-ment of FYCO1. LIR is another important domain in FYCO1. Itis necessary and sufficient for mediating FYCO1 binding toLC3, and deletion of LIR resulted in loss of co-localizationbetween FYCO1 and LC3. Endogenous FYCO1 is localizedto punctuated structures concentrated in the juxtanuclearregion, but FYCO1 redistributes to other parts of thecytoplasm during starvation. Depletion of FYCO1 leads tothe accumulation of perinuclear clustering of autophago-somes, these autophagosomes co-localize with Rab7 but notthe lysotracker. All these data suggested that FYCO1 binds toLC3 and Rab7 to form a complex which can promoteautophagosome from perinuclear to MT plus end and regulatebidirectional transport of autophagosomes along the MT track(Pankiv et al., 2010).

RAB, HOPS and SNARE

So far, the evidence suggests that fusion between autopha-gosome and lysosome depends on the canonical fusionmachinery, the RAB-SNARE (soluble N-ethylmalemide-sen-sitive factor attachment protein receptor) system, and the setof involved molecules are thought to be almost identical tothose involved in vacuole-vacuole homotypic fusion (Cai etal., 2007). SNAREs are membrane-anchored proteins. Theycan adjoin their transmembrane domain through a conservedcoiled-coil domain, which is required for regulated fusionbetween lipid bilayers (Ungermann and Langosch, 2005;Jahn and Scheller, 2006; Cai et al., 2007; Langosch et al.,2007). Based on their localization, SNAREs are classified asvesicle (v)-SNAREs (Ykt6/Nyv1) and target membrane(t)-SNAREs (Vam3/Vam7/Vti1) (Rothman, 1994; Darsow et al.,1997; Nichols et al., 1997; Sato et al., 1998; Ungermann andWickner, 1998; Ungermann et al., 1998, 1999). The fusion-competent assembly of SNAREs localize on opposingmembranes, termed the trans-SNARE complex (Weber

et al., 1998). In mammalian cells, knocking down (t)-SNAREVti1b delays the maturation of autophagosomes (Parlati et al.,2000; Atlashkin et al., 2003). In yeast, it has been demon-strated that Vam3 and Vti1 are needed for the fusion betweenautophagosomes and vacuoles (Wang and Klionsky, 2003;Klionsky, 2005).

HOPS (homotypic fusion and protein sorting), is aconserved protein complex consisting of four C-Vps proteins(Vps11, 16, 18, 33), and Vps41 and Vps39 (Price et al., 2000;Seals et al., 2000; Wurmser et al., 2000). HOPS interacts witht-SNARE vam3 and the vacuolar Rab GTPase Ypt7 throughVps33 (Sato et al., 2000; Dulubova et al., 2001; Stroupe et al.,2006). During fusion, the larger size of HOPS complex helps itreach over relatively longer distances compared to SNAREcomplex, thus, mediating the first contact between vacuoles(Cai et al., 2007). Recent research demonstrates HOPScomplex prevents the disassembly of trans-SNARE com-plexes by Sec17/Sec18 during membrane fusion (Mima et al.,2008; Xu et al., 2010). Many data indicate HOPS complexplays a role in the early stages of docking at the vesiclesurface and takes part in the vesicle fusion with SNAREs andRab (Sato et al., 2000; Seals et al., 2000; Wurmser et al.,2000). Early in 1997, Stephanie E. Rieder group havedemonstrated that Vps18 takes part in the delivery ofautophagosome to the vacuole in yeast. The mutant ofVps18 shows numbers of autophagic bodies within thecytoplasm but not the vacuoles (Rieder and Emr, 1997). InDrosophila, the Vps18 homolog Deep orange (Dor) haspreviously been shown to mediate fusion of autophagosomeswith lysosomes. It has been demonstrated class C Vpsproteins are components of a heter-oligomeric proteincomplex that achieves their function (Rieder and Emr.,1997). Vps16 binds to Dor and Vps33 in Drosophila, andVps16 knockdown causes loss of Dor and accumulation ofautophagosomes (Pulipparacharuvil et al., 2005). Vam3p, asyntaxin homolog, is required for the autophagosome-lysosome fusion. The mutant of Vam3 accumulated multipleautophagosomes in the cytoplasm but no detectable accu-mulation of autophagic bodies in the vacuole (Darsow et al.,1997). It is likely Vam3 and Vps33 do in fact function togetherto direct the docking and fusion of transport intermediates withthe vacuole.

It is well established that Rab proteins which peripherallyassociate with membranes via a geranylgeranyl lipid tail playan important role in tethering/docking of vesicles to theirtarget compartment during vesicles fusion. The vesicleassociated GTP-bound form of Rab proteins is thought tobe active and interact with its effector proteins. The GDP-bound form of Rab proteins is inactive and disassociates fromvesicles (Somsel Rodman andWandinger-Ness, 2000). Eachtransport step requires activated Rab proteins binding tosoluble factors which act as effector proteins (Marino andHeidi., 2001). HOPS complex seems to be the effector thatmediates Rab7-dependent tethering (Price et al., 2000b). In


The late stage of autophagy: cellular events and molecular regulation Protein & Cell

yeast, Vps39 has been shown to confer GTPases exchangefactor (GEF) activity to Ypt7p (the yeast Rab7 ortholog). Inmammalian cells, hVps39 regulates the recruitment/activa-tion of Rab7 onto the Rab5-labeled early endosomes. Rabeffectors are not randomly distributed on the organellemembrane but are clustered in distinct functional domains(Novick and Zerial., 1997). The interaction between Rabeffector and SNARE provides a new understanding how Rabproteins directly regulate SNARE function during fusion(Wurmser et al., 2000). Rab GTPases and their effectorsprovide the complementary specificity to SNARE complexesduring membrane tethering and fusion. Based on the recentresearch, Rab7 GTPases is targeted to the autophagosomemembrane and required for autophagosome maturation(Gutierrez et al., 2004; Jäger et al., 2004). Liang et al.(2008a) demonstrate that UVRAG bind to class C-Vpscomplex and this interaction promotes activation of Rab7GTPase, which takes part in the mature of autophagosome(Liang et al., 2008b). Overexpression of a Rab7 dominantnegative mutant impairs fusion between autophagosomesand the late endosome/lysosome (Gutierrez et al., 2004;Jäger et al., 2004). In addition, Rab24 shows a perinuclearreticular localization and partially overlapps with ER, cis-Golgi, and the ER-Golgi intermediates compartment undernormal condition. But its distribution changes dramaticallyunder starvation. Co-localization between Rab24 and LC3has been detected under starvation condition (Munafó andColombo, 2002). Yoshimori group (2005) demonstrated thatRab24 may be involved in transportation of autophagosomemembrane compartment to lysosome (Egami et al., 2005).Biochemical studies in hepatocytes have indicated thatautophagosomes can fuse with endosomes to form amphi-some before they fuse with lysosome. Such phenomenon canalso be found in pancreatic cells, fibroblasts and HeLa cells.Rab11 is required for MVBs formation and its normal function(Satoh et al., 2005). Overexpression of wild-type Rab11 andits active mutant Rab11Q70L generates large MVBs markedby Rab11 and a remarkable colocalization with LC3. Mutantform of Rab11 hampered the interaction between MVBs andautophagosomes. And this process does not require Rab7which plays a role in fusion between autophagosome andlysosome and amphisome and lysosome (Fader et al., 2008).Rab22 associates with early endosome and late endosomebut not lysosome, and can affect the morphology andphysiology of the endocytic pathway; interestingly, its activemutant can associate to lysosome and autophagosome(Mesa et al., 2001). Further research demonstrated it mayplay a role in autophagic process, but it will need more clearevidence.

Others

The peripheral membrane protein complex of Mon1-Ccz1 hasbeen discussed as an additional factor involved in docking at

the vacuole (Wang et al., 2002). Ccz1, initially discovered as aprotein functionally linked to Ypt7 in yeast, is conserved inhumans and C.elegans (Kucharczyk et al., 2001). Mon1 andCcz1 are membrane-associated proteins, and Mon1-Ccz1complex has been proposed to contain longin domains, whichalso have been detected in several other trafficking proteinssuch as SNAREs (Kucharczyk et al., 2000; Wang et al.,2002). In addition, either Mon1 or Ccz1 requires the class Csubunits of the HOPS complex to associate with membrane.Wang et al. (2003) have demonstrated that vacuole fusion isstrongly impaired when either protein is missing or when theproteins are inhibited by specific antibodies (Wang et al.,2003). It is possible that the Mon1-Ccz1 is a cofactor of theHOPS tethering complex in the fusion between autophago-some and lysosome (Wang and Klionsky, 2003; Klionskyet al., 2005).

LAMPs (lysosomal-associated membrane proteins) areheavily glycosylated lysosomal transmembrane proteins(Eskelinen et al., 2003). Lysosome membrane proteinLamp2 has been recently demonstrated a role in autophago-some maturation (Huynh et al., 2007; Saftig et al., 2008).Starvation-induced degradation of long-lived proteins isimpaired in the lamp2-deficient hepatocytes (Tanaka et al.,2000). Interestingly, unlike Lamp2 single-deficient hepato-cytes, the degradation of long-lived proteins is not affected inLamp1/Lamp2 double deficient fibroblasts (Eskelinen et al.,2004), the difference is possibly due to the differentautophagy rates in different cell types (Eskelinen, 2005).More interestingly, the delivery of Rab7 to autophagicvacuoles is impaired in the lamp double-deficient fibroblasts(Eskelinen et al., 2004). So there is a hypothesis that the roleof Lamp2 in autophagosome maturation is to regulate thedistribution or targeting of Rab7.

UVRAG (UV irradiation resistance-associated gene) is aBeclin1-binding autophagic tumor suppressor, and it hasbeen reported to bind Beclin1 to take part in the formation ofautophagosomes (Liang et al., 2006, 2007; Takahashi et al.,2007, 2008). Interestingly, Liang et al. (2008a) reportedUVRAG can also interact with class C-VPS complex throughVps16 and has a role in the fusion between autophagosomeand lysosome. And the role of UVRAG-Class C-Vps complexin autophagosome maturation is different from the role ofUVRAG-Beclin1-mediated autophagosome formation.UVRAG mutant that was defective in binding to Beclin1 butcan still interact with C-Vps was able to regulate assembly ofLC3+ and LAMP1+structure. However, a mutant that cannotbind to C-Vps but still can bind to Beclin1 showed markedlyattenuated ability to promote autophagosome maturation(Liang et al., 2008a). And this is due to UVRAG-Class C-Vpscomplex can activate Rab7 GTPase activity but not theUVRAG mutant, which cannot bind to Class C-Vps complex(Liang et al., 2008b; Peplowska et al., 2008).

AAA ATPases (AAA adenosine triphosphatases) was firstdefined as a subset of P-loop ATPases in the early 1990s,



based on the homology within the ATP binding domain. Theyare conserved from prokaryotes to humans, and they use theenergy by ATP hydrolysis to remodel their target substrates.They have many functions, such as vesicle transport,organelle assembly, and protein unfolding. Disassembly ofprotein complex by AAA proteins plays an important role inseveral cell biologic function including the fusion in autopha-gic pathway (White and Lauring, 2007; Mehrpour et al., 2010).NSF is one of the AAA proteins and it binds to SNAREcomplex and utilizes ATP hydrolysis to disassemble them inorder to regenerate free SNAREs, thus allowing fusionproceed as we described above (Block et al., 1998;Ungermann et al., 1998). SKD1 (a mammalian homolog ofyeast VPS4), another AAA protein, is critical for thedisassembly of the ESCRT-III complex once cargo selectionis completed. Overexpression of a SKD1 mutant whichcannot hydrolyse ATP induced a defection in autophago-somes maturation (Shirahama et al., 1997; Nara et al., 2002;Rusten et al., 2007).

DRAM (damage-regulated autophagy modulator), a trans-membrane protein present in the lysosome, has also beendemonstrate a role in the later stage of autophagy inmammalian (Crighton et al., 2006).

DEGRADATION AND LYSOSOME EFFLUX

After autophagosome-lysosome fusion, the outer membraneof autophagosome is incorporated into lysosome/vacuole.The step of degradation may contain two steps: breakdownthe autophagosome membrane to deliver its contents to thelysosome lumen, and the degradation of these contents bythe various enzymes inside lysosome/vacuole.

Autophagosome is double membrane vesicle. After fusionwith lysosome or vacuole, autophagosome membrane will besubject to lysosome degradation (Nakamura et al., 1997;Takeshige et al., 1992). Aut5p (a yeast homolog ofmammalian Atg15), contains one to three potential trans-membrane domains and a lipase active-site motif. It localizeson ER and is transported into vacuolar/lysosome via anautophagy-independent route (Odorizzi et al., 1998, 2000).As a lipase, it can disintegrate the membrane of autophago-some and release the contents into vacuole for degradation inAut5p–/– cells accumulated autophagosomes (Epple et al.,2001, 2003). The maturation of proaminopeptidase, which isdependent on delivery to the vacuole by autophagy, isimpaired in Aut5p–/– cells.

The breakdown process of autophagic content depends onproteases and acidification of vacuole lumen in yeast(Takeshige et al., 1992; Nakamura et al., 1997). ProteinaseA belongs to aspartic proteinase superfamily, which areproteolytic enzymogen. It is initially synthesized as an inactiveprecursor (zymogen), which transits to the endoplasmicreticulum. Proteinase A activation can occur either throughproteinase B catalyzed cleavage which is a vacuolar serine

protease or via autoactivation. Proteinase A is required toprocess inactive precursors of various enzymes (Parr et al,2007). Experiments on the effects of gene disruption andinhibitors of proteases (proteinase A and B) showed theaccumulation of autophagic bodies in wild-type cells and thatthese autophagic bodies disappeared rapidly from thevacuoles once the proteases activity were restored (Mechlerand Wolf, 1981; Takeshige et al., 1992; Klionsky, 2005).

The degradation products, such as amino acids andmonosaccharide, will be transported out of lysosome/vacuolethrough a group of lysosomal transmembrane proteins namedlysosomal efflux transporters (Lloyd, 1996). Atg22 was firstidentified as Aut4 as an integral membrane protein located onmembrane of lysosome/vacuole (Teter et al., 2001). It wasfirst identified as a protein required for autophagic degrada-tion (Suriapranata et al., 2000), however, the follow up workby Klionsky’s group demonstrated that Atg22 is not essentialfor autophagic cargo degradation. Instead, they found Atg22is a tyrosine and leucine efflux transporter on vacuolarmembrane. The Atg22 mutant that they examined isauxotrophic for leucine which could be rescued by additionof leucine and the mutant shows inhibition of protein synthesisunder autophagy-inducing condition.

AUTOPHAGIC LYSOSOME REFORMATION

One of the key roles of autophagy is to recycle nutrientsthrough degradation of unessential cellular content. As wehave mentioned above, autophagosome formation is trig-gered by mTOR inactivation shortly after starvation. Interest-ingly, Yu et al. (2010) recently reported the amino acidsreleasing form autolysosome will reactivate mTOR. mTORreactivation will trigger the disassembly of autolysosome,resulting in a tubular structure which mainly containslysosomal membrane components extended from autolyso-some. These tubular structures are highly dynamic, under-going constantly fission of budding process by which smallLamp1 positive vesicles named “proto-lysosome” are formed.Proto-lysosomes are not acidic and do not have degradationcapacity, but over a period of 2–3 h, proto-lysosomes matureinto functional lysosome through acquiring lysosomal lumenproteins by a M6PR-dependent mechanism. This process isnamed as autophagic lysosome reformation (ALR). ALR isdirectly regulated by mTOR activities. When mTOR isinhibited, ALR is blocked, resulting in enlarged long-lastingautolysosome. So far, the molecular machinery regulatingALR is largely unknown. Rab7 is required to disassociatefrom autolysosome before ALR, and forced staying of Rab7on autolysosome membrane by overexpressing constitutiveactive form of Rab7 blocks ALR (Yu et al., 2010).

CONCLUSION

So far, later stage of autophagy have received relatively less



attention comparing to the earlier stage of autophagy,however, recently work had demonstrated that the fusion,degradation and lysosome reformation are regulated by a setof elaborated and complicated molecular machinery, anddefecting in these later stage autophagy events has seriousphysiologic consequence. At this point, better understandingthe cellular events and of molecular regulation of later stageautophagy is clearly needed, as many intriguing questionsstill remain unsolved, for example, how lysosome recognizesand selectively fuses with “mature” autophagosome but notisolation membrane? What is the molecular machinery toregulate the formation of reformation tubules? How compo-nents from lysosome are separated from autophagosomecomponents during autophagic lysosome reformation? Weare confident that with more researchers from differentbackground attracted into this emerging research direction,these questions can be solved soon.

ABBREVIATIONS

AAA ATPases, AAA adenosine triphosphatases; ALR, autophagic

lysosome reformation; Atg, autophagy-related; Dor, Deep orange;DRAM, damage-regulated autophagy modulator; ESCRT, endosomalsorting complex required for transport; FYCO1, FYVE and coiled-coil

[CC] domain containing 1; GEF, GTPases exchange factor; HOPS,homotypic fusion and protein sorting; IM, isolatedmembrane; LAMPs,lysosomal-associated membrane proteins; LIR, LC3-interacting

region; MTOC, microtubule-organizing center; SNARE, soluble N-ethylmalemide-sensitive factor attachment protein receptor; mTOR,mammalian target of rapamycin; PAS, pre-autophagosomal structure;t-SNAREs, target membrane SNAREs; UVRAG, UV irradiation

resistance-associated gene; v-SNAREs, vesicle SNAREs

REFERENCES

Aplin, A., Jasionowski, T., Tuttle, D.L., Lenk, S.E., and Dunn, W.A. Jr.(1992). Cytoskeletal elements are required for the formation andmaturation of autophagic vacuoles. J Cell Physiol 152, 458–466.

Atlashkin, V., Kreykenbohm, V., Eskelinen, E.-L., Wenzel, D.,Fayyazi, A., and Fischer von Mollard, G. (2003). Deletion of theSNARE vti1b in mice results in the loss of a single SNARE partner,

syntaxin 8. Mol Cell Biol 23, 5198–5207.

Axe, E.L., Walker, S.A., Manifava, M., Chandra, P., Roderick, H.L.,

Habermann, A., Griffiths, G., and Ktistakis, N.T. (2008). Autopha-gosome formation from membrane compartments enriched inphosphatidylinositol 3-phosphate and dynamically connected tothe endoplasmic reticulum. J Cell Biol 182, 685–701.

Bache, K.G., Raiborg, C., Mehlum, A., Madshus, I.H., and Stenmark,H. (2002). Phosphorylation of Hrs downstream of the epidermal

growth factor receptor. Eur J Biochem 269, 3881–3887.

Block, M.R., Glick, B.S., Wilcox, C.A., Wieland, F.T., and Rothman, J.E. (1988). Purification of an N-ethylmaleimide-sensitive protein

catalyzing vesicular transport. Proc Natl Acad Sci U S A 85,7852–7856.

Cai, H., Reinisch, K., and Ferro-Novick, S. (2007). Coats, tethers,

Rabs, and SNAREs work together to mediate the intracellulardestination of a transport vesicle. Dev Cell 12, 671–682.

Callebaut, I., de Gunzburg, J., Goud, B., and Mornon, J.P. (2001).

RUN domains: a new family of domains involved in Ras-likeGTPase signaling. Trends Biochem Sci 26, 79–83.

Cao, X., and Barlowe, C. (2000). Asymmetric requirements for a Rab

GTPase and SNARE proteins in fusion of COPII vesicles withacceptor membranes. J Cell Biol 149, 55–66.

Crighton, D., Wilkinson, S., O’Prey, J., Syed, N., Smith, P., Harrison,

P.R., Gasco, M., Garrone, O., Crook, T., and Ryan, K.M. (2006).DRAM, a p53-induced modulator of autophagy, is critical forapoptosis. Cell 126, 121–134.

Darsow, T., Rieder, S.E., and Emr, S.D. (1997). A multispecificitysyntaxin homologue, Vam3p, essential for autophagic andbiosynthetic protein transport to the vacuole. J Cell Biol 138,

517–529.

Dulubova, I., Yamaguchi, T., Wang, Y., Südhof, T.C., and Rizo, J.

(2001). Vam3p structure reveals conserved and divergent proper-ties of syntaxins. Nat Struct Biol 8, 258–264.

Egami, Y., Kiryu-Seo, S., Yoshimori, T., and Kiyama, H. (2005).

Induced expressions of Rab24 GTPase and LC3 in nerve-injuredmotor neurons. Biochem Biophys Res Commun 337, 1206–1213.

Epple, U.D., Suriapranata, I., Eskelinen, E.-L., and Thumm, M.

(2001). Aut5/Cvt17p, a putative lipase essential for disintegrationof autophagic bodies inside the vacuole. J Bacteriol 183,5942–5955.

Epple, U.D., Eskelinen, E.L., and Thumm, M. (2003). Intravacuolarmembrane lysis in Saccharomyces cerevisiae. Does vacuolartargeting of Cvt17/Aut5p affect its function? J Biol Chem 278,

7810–7821.

Eskelinen, E.L. (2005). Maturation of autophagic vacuoles in

mammalian cells. Autophagy 1, 1–10.

Eskelinen, E.L., Tanaka, Y., and Saftig, P. (2003). At the acidic edge:emerging functions for lysosomal membrane proteins. Trends Cell

Biol 13, 137–145.

Eskelinen, E.L., Schmidt, C.K., Neu, S., Willenborg, M., Fuertes, G.,Salvador, N., Tanaka, Y., Lüllmann-Rauch, R., Hartmann, D.,

Heeren, J., et al. (2004). Disturbed cholesterol traffic but normalproteolytic function in LAMP-1/LAMP-2 double-deficient fibro-blasts. Mol Biol Cell 15, 3132–3145.

Fader, C.M., Sánchez, D., Furlán, M., and Colombo, M.I. (2008).Induction of autophagy promotes fusion of multivesicular bodieswith autophagic vacuoles in k562 cells. Traffic 9, 230–250.

Fass, E., Shvets, E., Degani, I., Hirschberg, K., and Elazar, Z. (2006).Microtubules support production of starvation-induced autophago-

somes but not their targeting and fusion with lysosomes. J BiolChem 281, 36303–36316.

Gasch, A.P., Spellman, P.T., Kao, C.M., Carmel-Harel, O., Eisen, M.

B., Storz, G., Botstein, D., and Brown, P.O. (2000). Genomicexpression programs in the response of yeast cells to environ-mental changes. Mol Biol Cell 11, 4241–4257.

Gill, S.R., Schroer, T.A., Szilak, I., Steuer, E.R., Sheetz, M.P., andCleveland, D.W. (1991). Dynactin, a conserved, ubiquitouslyexpressed component of an activator of vesicle motility mediated

by cytoplasmic dynein. J Cell Biol 115, 1639–1650.

Gutierrez, M.G., Munafó, D.B., Berón, W., and Colombo, M.I. (2004).Rab7 is required for the normal progression of the autophagic

pathway in mammalian cells. J Cell Sci 117, 2687–2697.

Hamasaki, M., and Yoshimori, T. (2010). Where do they come from?

Insight into autophagosome formation. FEBS Lett 584, 1296–1301.



Hayakawa, A., Hayes, S.J., Lawe, D.C., Sudharshan, E., Tuft, R.,

Fogarty, K., Lambright, D., and Corvera, S. (2004). Structural basisfor endosomal targeting by FYVE domains. J Biol Chem 279,5958–5966.

Huynh, K.K., Eskelinen, E.L., Scott, C.C., Malevanets, A., Saftig, P.,and Grinstein, S. (2007). LAMP proteins are required for fusion oflysosomes with phagosomes. EMBO J 26, 313–324.

Itoh, T., Fujita, N., Kanno, E., Yamamoto, A., Yoshimori, T., andFukuda, M. (2008). Golgi-resident small GTPase Rab33B interactswith Atg16L and modulates autophagosome formation. Mol Biol

Cell 19, 2916–2925.

Jäger, S., Bucci, C., Tanida, I., Ueno, T., Kominami, E., Saftig, P., andEskelinen, E.L. (2004). Role for Rab7 in maturation of late

autophagic vacuoles. J Cell Sci 117, 4837–4848.

Jahn, R., and Scheller, R.H. (2006). SNAREs—engines for mem-

brane fusion. Nat Rev Mol Cell Biol 7, 631–643.

Jahreiss, L., Menzies, F.M., and Rubinsztein, D.C. (2008). Theitinerary of autophagosomes: from peripheral formation to kiss-

and-run fusion with lysosomes. Traffic 9, 574–587.

Kanazawa, C., Morita, E., Yamada, M., Ishii, N., Miura, S., Asao, H.,Yoshimori, T., and Sugamura, K. (2003). Effects of deficiencies of

STAMs and Hrs, mammalian class E Vps proteins, on receptordownregulation. Biochem Biophys Res Commun 309, 848–856.

Kimura, S., Noda, T., and Yoshimori, T. (2008). Dynein-dependentmovement of autophagosomes mediates efficient encounters withlysosomes. Cell Struct Funct 33, 109–122.

Klionsky, D.J. (2005). The molecular machinery of autophagy:unanswered questions. J Cell Sci 118, 7–18.

Klionsky, D.J. (2007). Autophagy: from phenomenology to molecular

understanding in less than a decade. Nat Rev Mol Cell Biol 8,931–937.

Kouno, T., Mizuguchi, M., Tanida, I., Ueno, T., Kanematsu, T., Mori, Y.,

Shinoda, H., Hirata, M., Kominami, E., and Kawano, K. (2005).Solution structure of microtubule-associated protein light chain 3and identification of its functional subdomains. J Biol Chem 280,

24610–24617.

Kucharczyk, R., Dupre, S., Avaro, S., Haguenauer-Tsapis, R.,

Słonimski, P.P., and Rytka, J. (2000). The novel protein Ccz1prequired for vacuolar assembly in Saccharomyces cerevisiaefunctions in the same transport pathway as Ypt7p. J Cell Sci 113,4301–4311.

Kucharczyk, R., Kierzek, A.M., Slonimski, P.P., and Rytka, J. (2001).The Ccz1 protein interacts with Ypt7 GTPase during fusion of

multiple transport intermediates with the vacuole in S. cerevisiae. JCell Sci 114, 3137–3145.

Kutateladze, T.G. (2006). Phosphatidylinositol 3-phosphate recogni-

tion and membrane docking by the FYVE domain. BiochimBiophys Acta 1761, 868–877.

Lakadamyali, M., Rust, M.J., Babcock, H.P., and Zhuang, X. (2003).

Visualizing infection of individual influenza viruses. Proc Natl AcadSci U S A 100, 9280–9285.

Langosch, D., Hofmann, M., and Ungermann, C. (2007). The role of

transmembrane domains in membrane fusion. Cell Mol Life Sci 64,850–864.

Lee, J.A., Beigneux, A., Ahmad, S.T., Young, S.G., and Gao, F.B.(2007). ESCRT-III dysfunction causes autophagosome accumula-tion and neurodegeneration. Curr Biol 17, 1561–1567.

Liang, C., Feng, P., Ku, B., Dotan, I., Canaani, D., Oh, B.H., and Jung,

J.U. (2006). Autophagic and tumour suppressor activity of a novelBeclin1-binding protein UVRAG. Nat Cell Biol 8, 688–699.

Liang, C., Feng, P., Ku, B., Oh, B.H., Jung, J.U., Oh, B., and Jung, J.(2007). UVRAG: a new player in autophagy and tumor cell growth.Autophagy 3, 69–71.

Liang, C., Lee, J.S., Inn, K.S., Gack, M.U., Li, Q., Roberts, E.A.,Vergne, I., Deretic, V., Feng, P., Akazawa, C., et al. (2008a).Beclin1-binding UVRAG targets the class C Vps complex to

coordinate autophagosome maturation and endocytic trafficking.Nat Cell Biol 10, 776–787.

Liang, C., Sir, D., Lee, S., Ou, J.H., and Jung, J.U. (2008b). Beyond

autophagy: the role of UVRAG in membrane trafficking. Autophagy4, 817–820.

Lindmo, K., Simonsen, A., Brech, A., Finley, K., Rusten, T.E., and

Stenmark, H. (2006). A dual function for Deep orange inprogrammed autophagy in the Drosophila melanogaster fat body.Exp Cell Res 312, 2018–2027.

Lloyd, J.B. (1996). Metabolite efflux and influx across the lysosomemembrane. Subcell Biochem 27, 361–386.

Lloyd, T.E., Atkinson, R., Wu, M.N., Zhou, Y., Pennetta, G., andBellen, H.J. (2002). Hrs regulates endosome membrane invagina-tion and tyrosine kinase receptor signaling in Drosophila. Cell 108,

261–269.

Longatti, A., and Tooze, S.A. (2009). Vesicular trafficking andautophagosome formation. Cell Death Differ 16, 956–965.

Lupas, A., Van Dyke, M., and Stock, J. (1991). Predicting coiled coilsfrom protein sequences. Science 252, 1162–1164.

Mann, S.S., and Hammarback, J.A. (1994). Molecular characteriza-tion of light chain 3. A microtubule binding subunit of MAP1A andMAP1B. J Biol Chem 269, 11492–11497.

Marchler-Bauer, A., Anderson, J.B., Chitsaz, F., Derbyshire, M.K.,DeWeese-Scott, C., Fong, J.H., Geer, L.Y., Geer, R.C., Gonzales,N.R., Gwadz, M., et al. (2009). CDD: specific functional annotation

with the Conserved Domain Database. Nucleic Acids Res 37,D205–D210.

Marino, Z., and Heidi, M. (2001). Rab proteins as membrane

prganizers. Natl Rev 2, 107–118.

Mechler, B., and Wolf, D.H. (1981). Analysis of proteinase A function

in yeast. Eur J Biochem 121, 47–52.

Mehrpour, M., Esclatine, A., Beau, I., and Codogno, P. (2010).Overview of macroautophagy regulation in mammalian cells. Cell

Res 20, 748–762.

Mesa, R., Salomón, C., Roggero, M., Stahl, P.D., and Mayorga, L.S.(2001). Rab22a affects the morphology and function of the

endocytic pathway. J Cell Sci 114, 4041–4049.

Mima, J., Hickey, C.M., Xu, H., Jun, Y., and Wickner, W. (2008).

Reconstituted membrane fusion requires regulatory lipids,SNAREs and synergistic SNARE chaperones. EMBO J 27,2031–2042.

Mizushima, N. (2007). Autophagy: process and function. Genes Dev21, 2861–2873.

Munafó, D.B., and Colombo, M.I. (2002). Induction of autophagy

causes dramatic changes in the subcellular distribution of GFP-Rab24. Traffic 3, 472–482.

Nakamura, N., Matsuura, A., Wada, Y., and Ohsumi, Y. (1997).

Acidification of vacuoles is required for autophagic degradation inthe yeast, Saccharomyces cerevisiae. J Biochem 121, 338–344.

Nakatogawa, H., Suzuki, K., Kamada, Y., and Ohsumi, Y. (2009).



Dynamics and diversity in autophagy mechanisms: lessons fromyeast. Nat Rev Mol Cell Biol 10, 458–467.

Nara, A., Mizushima, N., Yamamoto, A., Kabeya, Y., Ohsumi, Y., andYoshimori, T. (2002). SKD1 AAA ATPase-dependent endosomaltransport is involved in autolysosome formation. Cell Struct Funct27, 29–37.

Nichols, B.J., Ungermann, C., Pelham, H.R.B., Wickner, W.T., andHaas, A. (1997). Homotypic vacuolar fusion mediated by t- and v-

SNAREs. Nature 387, 199–202.

Novick, P., and Zerial, M. (1997). The diversity of Rab proteins invesicle transport. Curr Opin Cell Biol 9, 496–504.

Odorizzi, G., Babst, M., and Emr, S.D. (1998). Fab1p PtdIns(3)P 5-kinase function essential for protein sorting in the multivesicularbody. Cell 95, 847–858.

Odorizzi, G., Babst, M., and Emr, S.D. (2000). Phosphoinositidesignaling and the regulation of membrane trafficking in yeast.

Trends Biochem Sci 25, 229–235.

Olkkonen, V.M., Dupree, P., Killisch, I., Lütcke, A., Zerial, M., andSimons, K. (1993). Molecular cloning and subcellular localization

of three GTP-binding proteins of the rab subfamily. J Cell Sci 106,1249–1261.

Pankiv, S., Alemu, E.A., Brech, A., Bruun, J.A., Lamark, T., Overvatn,

A., Bjørkøy, G., and Johansen, T. (2010). FYCO1 is a Rab7 effectorthat binds to LC3 and PI3P to mediate microtubule plus end-directed vesicle transport. J Cell Biol 188, 253–269.

Parlati, F., McNew, J.A., Fukuda, R., Miller, R., Söllner, T.H., andRothman, J.E. (2000). Topological restriction of SNARE-depen-dent membrane fusion. Nature 407, 194–198.

Parr, C.L., Keates, R.A., Bryksa, B.C., Ogawa, M., and Yada, R.Y.(2007). The structure and function of Saccharomyces cerevisiae

proteinase A. Yeast 24, 467–480.

Peplowska, K., Cabrera, M., and Ungermann, C. (2008). UVRAGreveals its second nature. Nat Cell Biol 10, 759–761.

Price, A., Seals, D., Wickner, W., and Ungermann, C. (2000a). Thedocking stage of yeast vacuole fusion requires the transfer ofproteins from a cis-SNARE complex to a Rab/Ypt protein. J Cell

Biol 148, 1231–1238.

Price, A., Wickner, W., and Ungermann, C. (2000b). Proteins needed

for vesicle budding from the Golgi complex are also required for thedocking step of homotypic vacuole fusion. J Cell Biol 148,1223–1229.

Pulipparacharuvil, S., Akbar, M.A., Ray, S., Sevrioukov, E.A., Haber-man, A.S., Rohrer, J., and Krämer, H. (2005). Drosophila Vps16Ais required for trafficking to lysosomes and biogenesis of pigment

granules. J Cell Sci 118, 3663–3673.

Raiborg, C., and Stenmark, H. (2009). The ESCRT machinery inendosomal sorting of ubiquitylated membrane proteins. Nature

458, 445–452.

Ravikumar, B., Acevedo-Arozena, A., Imarisio, S., Berger, Z., Vacher,C., O’Kane, C.J., Brown, S.D., and Rubinsztein, D.C. (2005).

Dynein mutations impair autophagic clearance of aggregate-proneproteins. Nat Genet 37, 771–776.

Ravikumar, B., Futter, M., Jahreiss, L., Korolchuk, V.I., Lichtenberg ,M., Luo, S., Massey, D.C., Menzies, F.M., Narayanan, U., Renna,M., et al. (2009). Mammalian macroautophagy at a glance. J CellBiol 122, 1707–1711.

Recacha, R., Boulet, A., Jollivet, F., Monier, S., Houdusse, A., Goud,B., and Khan, A.R. (2009). Structural basis for recruitment of Rab6-

interacting protein 1 to Golgi via a RUN domain. Structure 17,21–30.

Rieder, S.E., and Emr, S.D. (1997). A novel RING finger proteincomplex essential for a late step in protein transport to the yeastvacuole. Mol Biol Cell 8, 2307–2327.

Rothman, J.E. (1994). Mechanisms of intracellular protein transport.Nature 372, 55–63.

Rothman, J.E., and Wieland, F.T. (1996). Protein sorting by transport

vesicles. Science 272, 227–234.

Rusten, T.E., and Stenmark, H. (2009). How do ESCRT proteins

control autophagy? J Cell Sci 122, 2179–2183.

Rusten, T.E., Vaccari, T., Lindmo, K., Rodahl, L.M., Nezis, I.P., Sem-Jacobsen, C., Wendler, F., Vincent, J.P., Brech, A., Bilder, D., et al.

(2007). ESCRTs and Fab1 regulate distinct steps of autophagy.Curr Biol 17, 1817–1825.

Saftig, P., Beertsen, W., and Eskelinen, E.L. (2008). LAMP-2: a

control step for phagosome and autophagosome maturation.Autophagy 4, 510–512.

Sato, T.K., Darsow, T., and Emr, S.D. (1998). Vam7p, a SNAP-25-likemolecule, and Vam3p, a syntaxin homolog, function together inyeast vacuolar protein trafficking. Mol Cell Biol 18, 5308–5319.

Sato, T.K., Rehling, P., Peterson, M.R., and Emr, S.D. (2000). Class CVps protein complex regulates vacuolar SNARE pairing and isrequired for vesicle docking/fusion. Mol Cell 6, 661–671.

Satoh, A.K., O’Tousa, J.E., Ozaki, K., and Ready, D.F. (2005). Rab11mediates post-Golgi trafficking of rhodopsin to the photosensitiveapical membrane of Drosophila photoreceptors. Development 132,

1487–1497.

Schroer, T.A., and Sheetz, M.P. (1991). Two activators of microtubule-based vesicle transport. J Cell Biol 115, 1309–1318.

Seals, D.F., Eitzen, G., Margolis, N., Wickner, W.T., and Price, A.(2000). A Ypt/Rab effector complex containing the Sec1 homolog

Vps33p is required for homotypic vacuole fusion. Proc Natl AcadSci U S A 97, 9402–9407.

Seglen, P.O., Berg, T.O., Blankson, H., Fengsrud, M., Holen, I., and

Strømhaug, P.E. (1996). Structural aspects of autophagy. Adv ExpMed Biol 389, 103–111.

Shirahama, K., Noda, T., and Ohsumi, Y. (1997). Mutational analysis

of Csc1/Vps4p: involvement of endosome in regulation ofautophagy in yeast. Cell Struct Funct 22, 501–509.

Söllner, T., Bennett, M.K., Whiteheart, S.W., Scheller, R.H., and

Rothman, J.E. (1993). A protein assembly-disassembly pathway invitro that may correspond to sequential steps of synaptic vesicledocking, activation, and fusion. Cell 75, 409–418.

Somsel Rodman, J., and Wandinger-Ness, A. (2000). Rab GTPasescoordinate endocytosis. J Cell Sci 113, 183–192.

Stroupe, C., Collins, K.M., Fratti, R.A., and Wickner, W. (2006).Purification of active HOPS complex reveals its affinities forphosphoinositides and the SNARE Vam7p. EMBO J 25,

1579–1589.

Sugawara, K., Suzuki, N.N., Fujioka, Y., Mizushima, N., Ohsumi, Y.,and Inagaki, F. (2004). The crystal structure of microtubule-

associated protein light chain 3, a mammalian homologue ofSaccharomyces cerevisiae Atg8. Genes Cells 9, 611–618.

Suriapranata, I., Epple, U.D., Bernreuther, D., Bredschneider, M.,

Sovarasteanu, K., and Thumm, M. (2000). The breakdown ofautophagic vesicles inside the vacuole depends on Aut4p. J CellSci 113, 4025–4033.



Takahashi, Y., Coppola, D., Matsushita, N., Cualing, H.D., Sun, M.,

Sato, Y., Liang, C., Jung, J.U., Cheng, J.Q., Mulé, J.J., et al. (2007).Bif-1 interacts with Beclin 1 through UVRAG and regulatesautophagy and tumorigenesis. Nat Cell Biol 9, 1142–1151.

Takahashi, Y., Meyerkord, C.L., and Wang, H.G. (2008). BARgainingmembranes for autophagosome formation: Regulation of autop-hagy and tumorigenesis by Bif-1/Endophilin B1. Autophagy 4,

121–124.

Takeshige, K., Baba, M., Tsuboi, S., Noda, T., and Ohsumi, Y. (1992).Autophagy in yeast demonstrated with proteinase-deficient

mutants and conditions for its induction. J Cell Biol 119, 301–311.

Tamai, K., Tanaka, N., Nara, A., Yamamoto, A., Nakagawa, I.,Yoshimori, T., Ueno, Y., Shimosegawa, T., and Sugamura, K.

(2007). Role of Hrs in maturation of autophagosomes inmammalian cells. Biochem Biophys Res Commun 360, 721–727.

Tanaka, Y., Guhde, G., Suter, A., Eskelinen, E.L., Hartmann, D.,Lüllmann-Rauch, R., Janssen, P.M., Blanz, J., von Figura, K., andSaftig, P. (2000). Accumulation of autophagic vacuoles andcardiomyopathy in LAMP-2-deficient mice. Nature 406, 902–906.

Teter, S.A., Eggerton, K.P., Scott, S.V., Kim, J., Fischer, A.M., andKlionsky, D.J. (2001). Degradation of lipid vesicles in the yeast

vacuole requires function of Cvt17, a putative lipase. J Biol Chem276, 2083–2087.

Ungermann, C., and Langosch, D. (2005). Functions of SNAREs in

intracellular membrane fusion and lipid bilayer mixing. J Cell Sci118, 3819–3828.

Ungermann, C., Nichols, B.J., Pelham, H.R.B., and Wickner, W.

(1998). A vacuolar v-t-SNARE complex, the predominant form invivo and on isolated vacuoles, is disassembled and activated fordocking and fusion. J Cell Biol 140, 61–69.

Ungermann, C., von Mollard, G.F., Jensen, O.N., Margolis, N.,Stevens, T.H., andWickner, W. (1999). Three v-SNAREs and two t-SNAREs, present in a pentameric cis-SNARE complex on isolated

vacuoles, are essential for homotypic fusion. J Cell Biol 145,1435–1442.

Ungermann, C., and Wickner, W. (1998). Vam7p, a vacuolar SNAP-

25 homolog, is required for SNARE complex integrity and vacuoledocking and fusion. EMBO J 17, 3269–3276.

Wang, C.-W., and Klionsky, D.J. (2003). The molecular mechanism ofautophagy. Mol Med 9, 65–76.

Wang, C.-W., Stromhaug, P.E., Kauffman, E.J., Weisman, L.S., and

Klionsky, D.J. (2003). Yeast homotypic vacuole fusion requires theCcz1-Mon1 complex during the tethering/docking stage. J Cell Biol163, 973–985.

Wang, C.-W., Stromhaug, P.E., Shima, J., and Klionsky, D.J. (2002).The Ccz1-Mon1 protein complex is required for the late step ofmultiple vacuole delivery pathways. J Biol Chem 277,

47917–47927.

Weber, T., Zemelman, B.V., McNew, J.A., Westermann, B., Gmachl,M., Parlati, F., Söllner, T.H., and Rothman, J.E. (1998). SNAR-

Epins: minimal machinery for membrane fusion. Cell 92, 759–772.

White, S.R., and Lauring, B. (2007). AAA + ATPases: achieving

diversity of function with conserved machinery. Traffic 8,1657–1667.

Wurmser, A.E., Sato, T.K., and Emr S.D. (2000).New component of

the vacuolar class C–Vps complex couples nucleotide exchangeon the Ypt7 GTPase to SNARE-dependent docking and fusion. JCell Biol 151, 551–62.

Xie, Z., and Klionsky, D.J. (2007). Autophagosome formation: coremachinery and adaptations. Nat Cell Biol 9, 1102–1109.

Xu, H., Jun, Y., Thompson, J., Yates, J., and Wickner, W. (2010).HOPS prevents the disassembly of trans-SNARE complexes bySec17p/Sec1p during membrane fusion. J EMBO 29, 1948–1960.

Yang, Z., Huang, J., Geng, J., Nair, U., and Klionsky, D.J. (2006).Atg22 recycles amino acids to link the degradative and recyclingfunctions of autophagy. Mol Biol Cell 17, 5094–5104.

Yorimitsu, T., and Klionsky, D.J. (2005). Autophagy: molecularmachinery for self-eating. Cell Death Differ 12, 1542–1552.

Yu, L., McPhee, C.K., Zheng, L., Mardones, G.A., Rong, Y., Peng, J.,Mi, N., Zhao, Y., Liu, Z., Wan, F., et al. (2010). Termination ofautophagy and reformation of lysosomes regulated by mTOR.Nature 17, 942–946.



REVIEW

microRNAs: tiny RNA molecules, huge drivingforces to move the cell

Shenglin Huang, Xianghuo He✉

State Key Laboratory of Oncogenes and Related Genes, Shanghai Cancer Institute, Shanghai Jiao Tong University School ofMedicine, Shanghai 200032, China✉ Correspondence: [email protected] September 20, 2010 Accepted October 9, 2010

ABSTRACT

Cell migration or movement is a highly dynamic cellularprocess, requiring precise regulation that is essential fora variety of biological processes. microRNAs (miRNAs)are a class of tiny non-coding RNA molecules thatfunction as critical post-transcriptional regulators ofgene expression. Emerging evidence demonstrates thatmiRNAs play important roles in cell migration anddirectly contribute to extracellular matrix (ECM) remodel-ing, cell adhesion, and cell signalling that controls cellmigration by targeting a large number of protein-codinggenes. Accordingly, the dysregulation of these miRNAshas been linked to several migration-related diseases. Inthis review, we summarize and highlight the recentadvances concerning the roles and validated targets ofmiRNAs in the control of cell movement.

KEYWORDS microRNA, cell migration, metastasis

INTRODUCTION

Cell migration or movement is essential for a variety ofbiologic processes, such as embryonic morphogenesis,wound healing, immune response, and cancer metastasis(Lauffenburger and Horwitz, 1996). It is a highly dynamicphenomenon that requires the precise regulation andintegration of multiple signaling pathways (Ridley et al.,2003; Friedl andWolf, 2010). The initial response of a cell to amigration-promoting agent involves polarization and theformation of a protrusion in the direction of migration (Parentand Devreotes, 1999). Next, the adhesion receptors bind tothe extracellular matrix (ECM) or adjacent cells, forming linksto the actin cytoskeleton. These adhesions serve as tractionpoints for migration but also stabilize the protrusion via

structural connections to actin filaments. Finally, the adhe-sions disassemble at the cell rear, which allows the cell todetach and contract, thus pulling the cell forward. In the pastdecades, great progress has been made in understanding thecomplexities and subtleties of the molecular mechanisms ofcell migration. These works mainly focused on investigatingthe proteins that regulate cell migration. Recently, a growingnumber of reports have described a new class of small non-coding RNA molecules termed microRNAs (miRNAs) that areinvolved in cell migration.

miRNAs are evolutionarily conserved non-coding smallRNA molecules that function as critical post-transcriptionalregulators of gene expression (Bartel, 2004, 2009). Theywere first discovered in Caenorhabditis elegans (Lee et al.,1993). In human cells, there are about 1000 miRNAs thatcollectively regulate the expression of more than 30% ofprotein-coding genes (Bentwich et al., 2005; Friedman et al.,2009). miRNAs are initially transcribed as long primarytranscripts (pri-miRNAs) by RNA polymerase II in the nucleus,which are subsequently cleaved by Drosha into the stem loopstructured precursor miRNAs (pre-miRNAs) (Lee et al.,2004). The pre-miRNAs are then exported to the cytoplasm,where they are further processed by the RNase III enzymeDicer into mature miRNAs (Murchison and Hannon, 2004;Gregory et al., 2006; Ji, 2008). In the cytoplasm, maturemiRNA molecules associate with the RNA-induced silencingcomplex (RISC) and regulate gene expression primarilythrough binding to the 3′ un-translated regions (UTRs) oftarget mRNAs (mRNAs), resulting in mRNA degradation orthe blockade of mRNA translation (Rana, 2007).

Since miRNAs were discovered, they have been shown toplay fundamental roles in a variety of physiologic andpathological processes (He and Hannon, 2004; Esquela-Kerscher and Slack, 2006). A growing body of evidence hasdocumented that miRNAs are involved in the control of cell



Protein & Cell

movement and directly contribute to ECM remodeling, celladhesion, and cell signaling during cell migration. This reviewsummarizes and highlights the recent advances in under-standing miRNAs and their validated targets underlying cellmovement.

miRNAs AND EXTRACELLULAR MATRIX

REMODELING

The extracellular matrix (ECM) is a complex structural entitythat is composed of three major classes of biomolecules:structural proteins (e.g., collagens and elastin), specializedproteins (e.g., fibrillin, fibronectin, and laminin), and proteo-glycans. The ECM serves as the structural and molecularscaffold for cell adhesion and migration. An increasingnumber of miRNAs are being identified as upstreamregulators of the ECM-related genes that thereby regulateECM remodeling and influence the mode and efficiency of cellmigration (Fig. 1). Two miRNAs, let-7g and miR-29c, werefound to directly target the expression of collagen, which isthe major protein comprising the ECM. There are at least 30different collagen genes dispersed throughout the humangenome. These genes generate proteins that combine in avariety of ways to create over 20 different types of collagenfibrils. Let-7g was reported to be present at significantly lowerlevels in metastatic hepatocellular carcinomas (HCCs) andmay inhibit HCC cell migration by targeting type I collagen a2(COL1A2) (Ji et al., 2010). miR-29c, which is remarkablydownregulated in nasopharyngeal carcinomas (NPCs), cantarget multiple collagens (collagen 1A1, 1A2, 3A1, 4A1, 4A2,

15A1) and laminin 1 (Sengupta et al., 2008). In addition, miR-335 can target the progenitor cell transcription factor SOX4and the extracellular matrix component tenascin C tosuppress breast cancer cell migration and metastasis(Tavazoie et al., 2008). The upregulation of miR-128 inhibitsReelin and DCX expression and reduces neuroblastoma cellmotility and invasiveness (Evangelisti et al., 2009). Reelin, ahigh-molecular-weight secreted glycoprotein, is thought to actas a guide for migratory neurons via interactions with two cellsurface receptors, the very low density lipoprotein receptor(VLDLR) and the apolipoprotein E receptor 2 (ApoER2); itthen triggers a tyrosine kinase signaling cascade. miR-143,induced by myocardin, attenuates ECM versican proteinexpression and inhibits smooth muscle cell (SMC) migration(Wang et al., 2010b). Versican is a chondroitin sulfateproteoglycan within the ECM that is produced by syntheticSMCs and promotes SMC migration and proliferation. miR-143 can also promote HCC cell invasion and metastasis byrepressing the expression of fibronectin type III domaincontaining 3B (FNDC3B) (Zhang et al., 2009). Fibronectinand the fibronectin type III domain containing 3A (FNDC3A)are two targets that are repressed by miR-17, which candecrease cell adhesion, migration and proliferation (Shan etal., 2009). Transgenic mice overexpressing miR-17 showedoverall growth retardation, smaller organs and greatlyreduced hematopoietic cell lineages.

Some miRNAs can also regulate the expression of ECMmodulators, such as matrix metalloproteinases (MMPs) andtissue inhibitors of metalloproteinases (TIMPs). miR-146binhibits cell migration and the invasion of glioblastoma cells by

Figure 1. miRNAs and extracellular matrix (ECM) remodeling. miRNAs are identified as upstream regulators of ECM-relatedgenes. FNDC3A: fibronectin type-III domain containing 3A; FNDC3B: fibronectin type III domain containing 3B; MMP: matrix

metalloproteinases; TIMP: tissue inhibitors of metalloproteinases; VLDLR: very low density lipoprotein receptor; ApoER2:apolipoprotein E receptor 2.


The role of miRNAs in cell migration Protein & Cell

reducing the expression of MMP16 (Xia et al., 2009). MMP16was found to possess proteolytic activity against ECMcomponents, such as type III collagen. MMP16 was alsoidentified as a functional target of miR-31, which cansuppress breast tumor cell metastasis (Valastyan et al.,2009). On the other hand, miR-21 contributes to gliomamalignancy by downregulating RECK and TIMP3 matrixmetalloproteinase inhibitors, which leads to the activation ofMMPs, thus promoting the invasiveness of cancer cells(Gabriely et al., 2008). miR-221 and miR-222 also directlyregulate the expression of the protein phosphatase 2Asubunit B (PPP2R2A) and TIMP3 tumor suppressors(Garofalo et al., 2009), leading to the activation of the AKTpathway and metallopeptidases to promote HCC cell invasionand metastasis. TIMP3 is also a functional target of miR-181bthat is induced by TGF-β and enhances MMP2 and MMP9activity by modulating TIMP3 levels and promoting migrationand invasion of HCC cells (Wang et al., 2010a).

miRNAs AND CELL ADHESION

Cell adhesion is mediated by adherent junction proteins,including cadherins, integrins, and other cell adhesionmolecules. These proteins can directly or indirectly connectto actin and/or intermediate filament cytoskeleton and therebyprovide mechanically robust but dynamic coupling. Recentreports have shown that some miRNAs can modulate theseadherent junction genes. It was recently reported that miR-9can directly target CDH1, the E-cadherin-encoding mRNA,leading to increased cell motility and invasiveness in breastcancer cells (Ma et al., 2010). In both morphogenesis andcancer models, the loss of E-cadherin results in weakenedcell junctions followed by cell detachment and the onset of asingle-cell mode of migration. This is termed the epithelial-mesenchymal transition (EMT). EMT describes the molecularreprogramming and phenotypic changes characterizing theconversion of polarized immotile epithelial cells to motilemesenchymal cells. This process allows the remodeling oftissues during embryonic development and is implicated inthe promotion of tumor invasion and metastasis. miR-9 cansuppress the expression of E-cadherin to promote carcinomacell motility and invasiveness and to activate β-cateninsignaling. The latter contributes to an elevated expressionof VEGFA, leading to the induction of tumor-associatedangiogenesis. In addition, the loss of expression of miR-200family members has been shown to play a critical role in therepression of E-cadherin by ZEB1 and ZEB2 during EMT,enhancing cancer cell migration and invasion (Korpal et al.,2008).

Integrins are a major family of cell-cell and cell-ECMadhesion proteins. Integrins contribute to cell-cell cohesionindirectly through intercellular ECM components, such as thebinding of α5β1 integrin to intercellular deposits of fibronectinand the binding of α6β1 integrin to intercellular laminin. In

addition to this adhesive function, integrins are also involvedin intracellular signaling and the regulation of cytoskeletalformation and play important roles in promoting cell migration.Each integrin consists of non-covalently linked α and βsubunits. Integrin β1 (ITGB1) was recently shown to bedirectly regulated by miR-183 (Li et al., 2010). The regulationof ITGB1 expression by miR-183 provides a new mechanismfor the anti-metastatic role of miR-183 and suggests that thismiRNA could influence the development and function ofneurosensory organs and contribute to functional alterationsassociated with cellular senescence in human diploidfibroblasts and human trabecular meshwork cells. In additionto ITGB1, miR-183 was also found to target KIF2A, a kinesinessential for both bipolar spindle assembly and chromosomemovement (Li et al., 2010).

miRNAs can also directly regulate other cell adhesionmolecules, such as CD117 and CD44. CD44, the receptor forhyaluronic acid (HA), was identified as a functional target ofmiR-373 and miR-520c, which can stimulate breast cancercell migration and invasion (Huang et al., 2008). CD44mediates cell-cell and cell-matrix interactions through itsaffinity for HA and plays an important role in cell migration,tumor growth, and progression. CD44 is also directly targetedby miR-328, which regulates zonation morphogenesis (Wanget al., 2008a). CD117 (c-kit) is the receptor for the cytokinestem cell factor (SCF) and plays a key role in endothelialprogenitor cell migration and homing. miR-221, a specificmiRNA identified in human umbilical vein endothelial cells(HUVECs), affects the expression of c-kit and participates inthe regulation of angiogenesis (Poliseno et al., 2006). Underhyperglycemic conditions, miR-221 is induced in HUVECs,which consequently triggers inhibition of c-kit and impairmentof HUVECmigration (Li et al., 2009c). Downregulation of miR-221 can attenuate high-glucose-induced suppression of c-kitand migration in HUVECs.

miRNAs AND CELL SIGNALING OF CELL

MIGRATION

miRNAs and HGF/c-Met signaling

The hepatocyte growth factor (HGF)/c-Met signaling cascadeis considered to be involved in embryonic organ develop-ment, adult organ regeneration, wound healing, and tumormetastasis. HGF interacts with the proto-oncogenic c-Metreceptor tyrosine kinase and regulates cell growth, cellmotility, and morphogenesis. Recently, c-Met has beenshown to be directly regulated by miR-1, miR-206, miR-34a,miR-23b, and miR-199a-3p (Fig. 2). miR-1 and miR-206,highly expressed in skeletal muscles, can suppress c-Metexpression and inhibit cell proliferation and migration ofrhabdomyosarcoma (Yan et al., 2009). miR-34a decreased c-Met-induced phosphorylation of extracellular signal-regulatedkinases 1 and 2 (ERK1/2) and inhibited HCC cells migration

Protein & Cell


Shenglin Huang and Xianghuo He

and invasion (Li et al., 2009a). miR-23b decreased theproliferation and migration abilities of HCC cells by inhibitingc-Met and urokinase-type plasminogen activator (uPA) (Salviet al., 2009). The latter is a critical functional downstreamtarget of HGF/c-Met signaling. miR-199a-3p reduced HCCcells invasive capability by targeting c-Met and mTOR(Fornari et al., 2010). In addition, miR-101, a miRNA that isrepressed in HCC, downregulated the expression of the fosoncogene, thereby reducing HGF-induced cell invasion andmigration (Li et al., 2009b).

miRNAs and epidermal growth factor receptor (EGFR)signaling

The epidermal growth factor receptor (EGFR) belongs to theERBB family of receptor tyrosine kinases, which consists offour members: EGFR (ErbB1, HER1), ErbB2 (HER2), ErbB3(HER3), and ErbB4 (HER4). ErbB2 is a unique member of theErbB family because it does not bind any of the known ligandswith high affinity, but it is the preferred heterodimeric partnerfor other EGFRs. These receptors couple binding of extra-cellular growth factor ligands to intracellular signaling path-ways and regulate diverse responses, including proliferation,differentiation, and cell motility. miRNAs have been shown todirectly regulate the expression of these receptors (Fig. 3).The regulation of miR-146a by breast cancer metastasissuppressor 1 (BRMS1) can suppress breast cancer cell

migration and metastasis by targeting the expression ofEGFR (Lin et al., 2008). The overexpression of miR-125a ormiR-125b reduced ERBB2 and ERBB3 at both the transcriptand protein level in these cells, leading to reduced ERK1/2and AKT signaling (Scott et al., 2007). Functionally, miR-125a- or miR-125b-overexpressing SKBR3 cells displayedmarkedly reduced cell migration and invasion capacities.Intriguingly, miR-125 was shown to be significantly down-regulated upon EGF stimulation (Wang et al., 2009). Thus,the EGF-miR-125- ERBB2/3 axis may be a positive feedbackloop for EGFR signaling.

miRNAs and PI3K signaling

Phosphatidylinositol 3-kinase (PI3K) signaling plays animportant role in the regulation of cell migration andparticularly in controlling the polarity of migrating cells. ThePI3K signaling cascade can be stimulated by EGF. Theactivation of this pathway increases the activity of the AKTkinase, which can phosphorylate mTOR (mammalian targetof rapamycin). The PI3K/AKT pathway is controlled by thetumor suppressor lipid phosphatase PTEN. Recently, PTENwas identified as a direct target of miR-21, miR-221, and miR-222. Aberrant expression of miR-21 cannot only contribute toHCC growth but can also mediate HCC cell invasion bydirectly targeting PTEN (Meng et al., 2007). miR-21 can alterfocal adhesion kinase (FAK) phosphorylation and the

Figure 2. miRNAs and hepatocyte growth factor (HGF)/c-Met signaling.HGF interacts with the proto-oncogenic c-Met receptortyrosine kinase and regulates cell growth, cell motility, and morphogenesis. The c-Met mRNA level has been shown to be directlyregulated by miR-1, miR-206, miR-34a, miR-23b, and miR-199a-3p. PAK, p21-activated kinase; uPA: urokinase-type plasminogen

activator; PI3K: phosphatidylinositol 3-kinase; FAK: focal adhesion kinase.



expression of the matrix metalloproteases MMP2 and MMP9,both downstream mediators of PTEN involved in cellmigration and invasion. PTEN was also found to be the directtarget of miR-221 and miR-222, which induce TRAILresistance and enhance HCC cell migration. Recently,mTOR was identified as a target of miR-199a-3p, which canblock the G1-S transition and reduce HCC cell invasion(Fornari et al., 2010). In addition, the DNA damage-inducibletranscript 4 (DDIT4), a modulator of the mTOR pathway, wasalso found to be a target of miR-221 and miR-222 (Pineau etal., 2010).

miRNAs and Rho GTPases signaling

The Rho GTPases signaling cascade plays a central role inregulating cell adhesion, migration, and cytoskeletal reorga-nization. The Rho family of GTPases is a family of smallsignaling G proteins and is a subfamily of the Ras superfamily.In mammals, Rho GTPases contain about 20 members,which are largely divided into the Cdc42, Rac1, and Rho(RhoA, RhoB, and RhoC) subfamilies. The activity of RhoGTPases is tightly controlled by several families ofregulators, including guanine nucleotide exchange factors(GEFs), GTPase-activating proteins (GAPs), and Rho GDP-dissociation inhibitor (RhoGDI). Rho GTPases carry outdistinct functions by activating various downstream effectors,such as Rho-associated kinases (ROCK) and p21-activated

kinases (PAK). A growing body of evidence shows thatmiRNAs can affect cell migration by regulating expression ofthe Rho GTPase members, their effectors and their regulators(Fig. 4). For example, miR-31 can directly target Rho A, whichis involved in the inhibition of several steps of breast cancercell metastasis, including local invasion, extravasation, initialsurvival at a distant site, and metastatic colonization(Valastyan et al., 2009). miR-10b is highly expressed inmetastatic breast cancer cells and positively regulates cellmigration and invasion. miR-10b proceeds to inhibit transla-tion of the mRNA encoding homeobox D10, resulting inincreased RhoC expression (Ma et al., 2007). RhoC wasidentified as a functional target of miR-138, which suppressestongue squamous cell carcinomas (TSCC) cell migration andinvasion. Furthermore, miR-138 can directly regulate theexpression of the Rho-associated kinase ROCK2, a down-stream signaling molecule of RhoC. By concurrently targetingRhoC and ROCK2, miR-138 leads to the reorganization of thestress fibers and the subsequent cell morphology change to around bleb-like shape as well as the suppression of cellmigration and invasion (Jiang et al., 2010). In addition,ROCK1 was substantially suppressed by miR-146a, whichwas decreased in hormone-refractory prostate carcinomas(HRPCs) and reduced cell invasion and metastasis tohuman bone marrow endothelial cell monolayers (Lin et al.,2008). Ezrin and stathmin are the effectors of ROCK and arealso found to be regulated by miRNAs. Ezrin is a target of

Protein & Cell

Figure 3. miRNAs and EGFR signaling. The epidermal growth factor receptors (EGFRs) belong to the ERBB family of receptor

tyrosine kinases. HER2 (ErbB2) is a unique member of the ErbB family because it does not bind any of the known ligands with highaffinity, but it is the preferred heterodimeric partner for other EGFRs. These receptors couple the binding of extracellular growthfactor ligands to intracellular signaling pathways and regulate diverse responses, including proliferation, differentiation, and cell

motility. MiRNAs have been shown to directly regulate the expression of these receptors. PTEN: phosphatase and tensin homolog;mTOR: mammalian target of rapamycin.



miR-183, which has been reversely correlated with themetastatic potential of lung cancer cells and inhibits migrationand invasion in lung cancer cells (Wang et al., 2008b). miR-9was shown to promote proliferation but to suppress themigration of human neural progenitor cells (hNPCs) bydirectly downregulating the expression of stathmin (Delaloyet al., 2010), which increases microtubule instability. miR-7introduction inhibits the motility, invasiveness, anchorage-independent growth, and tumorigenic potential of highlyinvasive breast cancer cells by directly targeting p21-activated kinase 1 (PAK1) expression (Reddy et al., 2008).PAK1 is a critical effector that links Cdc42 and Rac tocytoskeletal reorganization and nuclear signaling. miRNAscan also modulate regulators of Rho GTPases. miR-151, afrequently amplified miRNA on chromosome 8q24, increasesHCC cell migration and invasion by directly targetingRhoGDIA (Ding et al., 2010). RhoGDIA can prevent nucleo-tide exchange and membrane association of Rho GTPasesand thus block their activation. Moreover, miR-151 canfunction synergistically with the host gene FAK to enhanceHCC cell motility and spreading.

miRNAs AND OTHER REGULATORS OF CELL

MIGRATION

miRNAs can also target other migration-related genesincluding chemokine (IL8), cell surface proteins (ADAM10,ADAM17, Ephrin-A3, and LRP1), adapter proteins (Crk andGNAI2), and transcription factors (Pax3, Pax7, MTA1, c-Myb,

FOXO3, and Mitf-M). miR-17/20 directly represses chemo-kine IL8 expression and inhibits cellular invasion and tumormetastasis of breast cancer (Yu et al., 2010). miR-122 wasshown to inhibit HCC cell metastasis by directly targetingthe expression of ADAM10 and ADAM17 (Bai et al., 2009;Tsai et al., 2009), which are cell surface proteins with a uniquestructure, possessing both potential adhesion and proteasedomains. miRNA-210 modulates the endothelial cellresponse to hypoxia and inhibits the receptor tyrosinekinase ligand ephrin-A3 (Fasanaro et al., 2008). miRNA-205inhibits tumor cell migration by downregulating the expressionof the LDL receptor-related protein 1 (LRP1) (Song and Bu,2009). miRNA-126 inhibits invasion in non-small cell lungcarcinoma cell lines by the repression of adapter protein Crkexpression (Crawford et al., 2008). miR-30d can enhanceintrahepatic and distal pulmonary metastasis of HCC cells byrepressing the expression of Galphai2 (GNAI2) (Yao et al.,2010). The transcription factors related to cell migration werealso identified as functional targets of miRNAs. miR-27bregulates Pax3 protein levels, and this downregulationensures rapid and robust entry into the myogenic differentia-tion program and inhibits progenitour cell migration (Cristet al., 2009). miR-196 was shown as an essential regulatourof tail regeneration by targeting the expression of BMP4and Pax7 (Sehm et al., 2009). miR-661 inhibited themotility, invasiveness, and tumorigenicity of invasive breastcancer cells by downregulating the expression of metastatictumor antigen 1 (MTA1) (Reddy et al., 2009). miR-150effectively reduced c-Myb expression and enhanced cell

Figure 4. MiRNAs and Rho GTPases. The Rho GTPases signaling cascade plays a central role in regulating cell adhesion,migration, and the cytoskeleton. MiRNAs can affect cell migration by regulating the expression of the Rho GTPases members as

well as their effectors and regulators. GEFs: guanine nucleotide exchange factors; RhoGDIA: Rho GDP-dissociation inhibitor;ROCK: Rho-associated kinases; PAK: p21-activated kinases; HOXD10: homeobox D10.



migration in HMEC-1 cells (Zhang et al., 2010). AberrantmiR-182 expression promotes melanoma metastasis byrepressing FOXO3 and microphthalmia-associated tran-scription factor-M (Mitf-M) (Segura et al., 2009). miR-10bpromotes cell migration and invasion by downregulating theexpression of KLF4 in human esophageal cancer cells (Tianet al., 2010).

CONCLUSION

Current research shows that many miRNAs have significantroles in the driving forces of cell movement by directlytargeting a large number of critical migration-related genes.These miRNAs and their target mRNAs construct a complexinteracting network controlling cell movement. As shown inTable 1, some miRNAs have been shown to target multiplegenes involved in cell migration. For example, miR-17/20 cantarget IL-8, CXCL1, CK8, fibronectin, and FNDC3A, and miR-221 can target PTEN, TIMP3, DDIT4, and c-kit. As these

target genes are all involved in cell migration, individualmiRNAs thus have robust roles in cell migration byconcurrently regulating their corresponding target genes.Conversely, a single gene may be modulated by multiplemiRNAs. As shown in Table 1, c-Met, PTEN, MMP16, andTIMP3 have been found to be regulated by at least twomiRNAs. Note that some miRNAs (miR-9 and miR-221) havedual roles in their effect on cell migration. For example, miR-9can increase breast cancer cell motility and invasiveness bytargeting CDH1, the E-cadherin-encoding mRNA, whereasmiR-9 is shown to suppress the migration of human neuralprogenitor cells (hNPCs) by downregulating the expression ofstathmin. This heterogeneity may result from the divergentcell types and subsequent different target genes. In conclu-sion, the discovery of miRNAs and their correspondingtargets is a new field that allows for the investigation of theunderlying molecular mechanisms of cell movement and forthe development of therapies for the treatment of migration-related disorders.

Protein & Cell

Table 1 The miRNAs and the validated targets in the driving forces of cell migration

miRNA regulation targets migration function references

Let-7 – COL1A2 – inhibits HCC cell migration Ji et al., 2010

miR-1/206 – c-Met –inhibits rhabdomyosarcoma

developmentYan et al., 2009

miR-10b twist HOXD10, Tiam1, Rac1, KLF4 +

promotes cell migration and inva-siveness in breast cancer, glioma,

and squamous cell carcinomacells

Ma et al., 2007;

Tian et al., 2010

miR-101 – FOS – Inhibits HCC cell invasion Li et al., 2009b

miR-122 – ADAM10, ADAM17 –inhibits HCC cell migration and

metastasisBai et al., 2009;Tsai et al., 2009

miR-125 EGFR regulated ErbB2/3 –

inhibits cell migration and invasionin breast cancer and lung cancer

cells

Scott et al., 2007;

Wang et al., 2009

miR-126 – Crk –inhibits lung cancer cell migration

and invasionCrawford et al., 2008

miR-128 – Reelin, DCX –inhibits neuroblastoma cell motility

and invasivenessEvangelisti et al., 2009

miR-138 – RhoC, ROCK2 –

inhibits tongue squamous cell

carcinoma cell migration andinvasion

Jiang et al., 2010

miR-143 NF-κB, myocardin Versican, FNDC3B –inhibits smooth muscle cells

migrationWang et al., 2010b;Zhang et al., 2009

miR-146 BRMS1 EGFR, MMP16, ROCK1 –inhibits cell migration and invasionin breast cancer and glioma cells

Lin et al., 2008;Xia et al., 2009

miR-150 – C-myb +promotes endothelial cell migra-

tionZhang et al., 2010

miR-151 amplification RhoGDIA +promotes HCC cell migration and

metastasisDing et al., 2010

miR-17/20 c-mycIL-8, CXCL1, CK8, fibronectin,

FNDC3A–

inhibits breast cancer cell migra-

tion and metastasis

Shan et al., 2009;

Yu et al., 2010



(Continued)

miRNA regulation targets migration function references

miR-181b TGF-β TIMP3 +promotes HCC cell migration and

invasionWang et al., 2010a

miR-182 amplification FOXO3, Mitf-M +promotes melanoma cell migra-

tion and metastasisSegura et al., 2009

miR-183 Integrin 1, kinesin 2, ezrin –inhibits cell migration and invasion

in HeLa and lung cancer cells

Wang et al., 2008b;

Li et al., 2010

miR-196 BMP4, Pax7 –inhibits tail regeneration of sala-

mandersSehm et al., 2009

miR-199a-3p mTOR, c-Met – inhibits HCC cell invasion Fornari et al., 2010

miR-200 ZEB1, ZEB2 –

counteracts epithelial-to-

mesenchymal transition andreduces cell migration/invasion

Korpal et al., 2008

miR-205 LRP1 – inhibits cell migration and invasion Song and Bu, 2009

miR-21 TIMP3, RECK, PTEN +promotes cell migration and inva-sion in breast cancer, glima, and

HCC cells

Meng et al., 2007;

Gabriely et al., 2008

miR-210 HIF-1 Ephrin-A3 +promotes endothelial cell migra-

tion in response to hypoxiaFasanaro et al., 2008

miR-221 c-jun PTEN, TIMP3, DDIT4, c-kit ±

promotes migration and metasta-sis in NSCLC and HCC cells.

Inhibits vascular smooth musclecells migration

Garofalo et al., 2009;Li et al., 2009c;

Pineau et al., 2010;Poliseno et al., 2006

miR-23b – uPA, c-Met –inhibits HCC cell migration and

invasionSalvi et al., 2009

miR-27b – Pax3 –

regulates myogenic differentiationand inhibits muscle stem cell

migration

Crist et al., 2009

miR-29c – Collagens, laminin 1, –inhibits nasopharyngeal carcino-

mas cell migration and metastasisSengupta et al., 2008

miR-30d – Gnail2 +promotes HCC cell migration and

metastasisYao et al., 2010

miR-31 – MMP16, RhoA –inhibits cell migration and metas-

tasis in breast cancer cellsValastyan et al., 2009

miR-328 – CD44 +regulates zonation morphogen-

esisWang et al., 2008a

miR-335 – SOX4, tenascin C –inhibits breast cancer cell migra-

tion and metastasisTavazoie et al., 2008

miR-34a – c-Met –inhibits HCC cell migration and

invasionLi et al., 2009a

miR-373/520 – CD44 +promotes breast cancer cellmigration and metastasis

Huang et al., 2008

miR-661 c/EBPA MTA1 –inhibits breast cancer cell migra-

tion and invasionReddy et al., 2009

miR-7 HoxD10 PAK1 –inhibits breast cancer cell migra-

tion and invasionReddy et al., 2008

miR-9 MYC/MYCN E-cadherin, stathmin ±

promotes breast cancer cell

migration and metastasis. Inhibitshuman neural progenitor cells

migration

Delaloy et al., 2010;Ma et al., 2010



ACKNOWLEDGEMENTS

This work was partially supported by the National Natural ScienceFoundation of China (Grant No. 81071637), the Science andTechnology Commission of Shanghai Municipality (Grant No.

10JC1414200), and the Doctoral Program of Higher Education ofChina (Grant No. 200802480076). We apologize to those colleagueswho have contributed to this exciting field but whose work could notbe cited because of space limitations.

REFERENCES

Bai, S., Nasser, M.W., Wang, B., Hsu, S.H., Datta, J., Kutay, H.,Yadav, A., Nuovo, G., Kumar, P., and Ghoshal, K. (2009).

MicroRNA-122 inhibits tumorigenic properties of hepatocellularcarcinoma cells and sensitizes these cells to sorafenib. J BiolChem 284, 32015–32027.

Bartel, D.P. (2004). MicroRNAs: genomics, biogenesis, mechanism,and function. Cell 116, 281–297.

Bartel, D.P. (2009). MicroRNAs: target recognition and regulatoryfunctions. Cell 136, 215–233.

Bentwich, I., Avniel, A., Karov, Y., Aharonov, R., Gilad, S., Barad, O.,

Barzilai, A., Einat, P., Einav, U., Meiri, E., et al. (2005). Identificationof hundreds of conserved and nonconserved human microRNAs.Nat Genet 37, 766–770.

Crawford, M., Brawner, E., Batte, K., Yu, L., Hunter, M.G., Otterson,G.A., Nuovo, G., Marsh, C.B., and Nana-Sinkam, S.P. (2008).MicroRNA-126 inhibits invasion in non-small cell lung carcinoma

cell lines. Biochem Biophys Res Commun 373, 607–612.

Crist, C.G., Montarras, D., Pallafacchina, G., Rocancourt, D.,

Cumano, A., Conway, S.J., and Buckingham, M. (2009). Musclestem cell behavior is modified by microRNA-27 regulation of Pax3expression. Proc Natl Acad Sci U S A 106, 13383–13387.

Delaloy, C., Liu, L., Lee, J.A., Su, H., Shen, F., Yang, G.Y., Young, W.L., Ivey, K.N., and Gao, F.B. (2010). MicroRNA-9 coordinatesproliferation and migration of human embryonic stem cell-derived

neural progenitors. Cell Stem Cell 6, 323–335.

Ding, J., Huang, S., Wu, S., Zhao, Y., Liang, L., Yan, M., Ge, C., Yao,J., Chen, T., Wan, D., et al. (2010). Gain of miR-151 on

chromosome 8q24.3 facilitates tumour cell migration and spread-ing through downregulating RhoGDIA. Nat Cell Biol 12, 390–399.

Esquela-Kerscher, A., and Slack, F.J. (2006). Oncomirs - microRNAs

with a role in cancer. Nat Rev Cancer 6, 259–269.

Evangelisti, C., Florian, M.C., Massimi, I., Dominici, C., Giannini, G.,

Galardi, S., Buè, M.C., Massalini, S., McDowell, H.P., Messi, E., etal. (2009). miR-128 up-regulation inhibits Reelin and DCXexpression and reduces neuroblastoma cell motility and invasive-ness. FASEB J 23, 4276–4287.

Fasanaro, P., D’Alessandra, Y., Di Stefano, V., Melchionna, R.,Romani, S., Pompilio, G., Capogrossi, M.C., and Martelli, F.

(2008). MicroRNA-210 modulates endothelial cell response tohypoxia and inhibits the receptor tyrosine kinase ligand Ephrin-A3.J Biol Chem 283, 15878–15883.

Fornari, F., Milazzo, M., Chieco, P., Negrini, M., Calin, G.A., Grazi, G.L., Pollutri, D., Croce, C.M., Bolondi, L., and Gramantieri, L. (2010).miR-199a-3p regulates mTOR and c-Met to influence the

doxorubicin sensitivity of human hepatocarcinoma cells. CancerRes 70, 5184–5193.

Friedl, P., and Wolf, K. (2010). Plasticity of cell migration: a multiscale

tuning model. J Cell Biol 188, 11–19.

Friedman, R.C., Farh, K.K., Burge, C.B., and Bartel, D.P. (2009).Most mammalian mRNAs are conserved targets of microRNAs.

Genome Res 19, 92–105.

Gabriely, G., Wurdinger, T., Kesari, S., Esau, C.C., Burchard, J.,Linsley, P.S., and Krichevsky, A.M. (2008). MicroRNA 21 promotes

glioma invasion by targeting matrix metalloproteinase regulators.Mol Cell Biol 28, 5369–5380.

Garofalo, M., Di Leva, G., Romano, G., Nuovo, G., Suh, S.S.,Ngankeu, A., Taccioli, C., Pichiorri, F., Alder, H., Secchiero, P., et al.(2009). miR-221&222 regulate TRAIL resistance and enhancetumorigenicity through PTEN and TIMP3 downregulation. Cancer

Cell 16, 498–509.

Gregory, R.I., Chendrimada, T.P., and Shiekhattar, R. (2006).

MicroRNA biogenesis: isolation and characterization of themicroprocessor complex. Methods Mol Biol 342, 33–47.

He, L., and Hannon, G.J. (2004). MicroRNAs: small RNAs with a big

role in gene regulation. Nat Rev Genet 5, 522–531.

Huang, Q., Gumireddy, K., Schrier, M., le Sage, C., Nagel, R., Nair, S.,Egan, D.A., Li, A., Huang, G., Klein-Szanto, A.J., et al. (2008). The

microRNAs miR-373 and miR-520c promote tumour invasion andmetastasis. Nat Cell Biol 10, 202–210.

Ji, J., Zhao, L., Budhu, A., Forgues, M., Jia, H.L., Qin, L.X., Ye, Q.H.,Yu, J., Shi, X., Tang, Z.Y., et al. (2010). Let-7g targets collagen typeI alpha2 and inhibits cell migration in hepatocellular carcinoma. JHepatol 52, 690–697.

Ji, X. (2008). The mechanism of RNase III action: how dicer dices.Curr Top Microbiol Immunol 320, 99–116.

Jiang, L., Liu, X., Kolokythas, A., Yu, J., Wang, A., Heidbreder, C.E.,Shi, F., and Zhou, X. (2010). Downregulation of the Rho GTPasesignaling pathway is involved in the microRNA-138-mediated

inhibition of cell migration and invasion in tongue squamous cellcarcinoma. Int J Cancer 127, 505–512.

Korpal, M., Lee, E.S., Hu, G., and Kang, Y. (2008). The miR-200

family inhibits epithelial-mesenchymal transition and cancer cellmigration by direct targeting of E-cadherin transcriptional repres-sors ZEB1 and ZEB2. J Biol Chem 283, 14910–14914.

Lauffenburger, D.A., and Horwitz, A.F. (1996). Cell migration: aphysically integrated molecular process. Cell 84, 359–369.

Lee, R.C., Feinbaum, R.L., and Ambros, V. (1993). The C. elegansheterochronic gene lin-4 encodes small RNAs with antisensecomplementarity to lin-14. Cell 75, 843–854.

Lee, Y., Kim, M., Han, J., Yeom, K.H., Lee, S., Baek, S.H., and Kim, V.N. (2004). MicroRNA genes are transcribed by RNA polymerase II.EMBO J 23, 4051–4060.

Li, G., Luna, C., Qiu, J., Epstein, D.L., and Gonzalez, P. (2010).Targeting of integrin beta1 and kinesin 2alpha by microRNA 183. JBiol Chem 285, 5461–5471.

Li, N., Fu, H., Tie, Y., Hu, Z., Kong, W., Wu, Y., and Zheng, X. (2009a).miR-34a inhibits migration and invasion by down-regulation of c-Met expression in human hepatocellular carcinoma cells. Cancer

Lett 275, 44–53.

Li, S., Fu, H., Wang, Y., Tie, Y., Xing, R., Zhu, J., Sun, Z., Wei, L., and

Zheng, X. (2009b). MicroRNA-101 regulates expression of the v-fos FBJ murine osteosarcoma viral oncogene homolog (FOS)oncogene in human hepatocellular carcinoma. Hepatology 49,

1194–1202.

Protein & Cell



Li, Y., Song, Y.H., Li, F., Yang, T., Lu, Y.W., and Geng, Y.J. (2009c).

MicroRNA-221 regulates high glucose-induced endothelial dys-function. Biochem Biophys Res Commun 381, 81–83.

Lin, S.L., Chiang, A., Chang, D., and Ying, S.Y. (2008). Loss of mir-

146a function in hormone-refractory prostate cancer. RNA 14,417–424.

Ma, L., Teruya-Feldstein, J., and Weinberg, R.A. (2007). Tumour

invasion and metastasis initiated by microRNA-10b in breastcancer. Nature 449, 682–688.

Ma, L., Young, J., Prabhala, H., Pan, E., Mestdagh, P., Muth, D.,Teruya-Feldstein, J., Reinhardt, F., Onder, T.T., Valastyan, S., et al.(2010). miR-9, a MYC/MYCN-activated microRNA, regulates E-cadherin and cancer metastasis. Nat Cell Biol 12, 247–256.

Meng, F., Henson, R., Wehbe-Janek, H., Ghoshal, K., Jacob, S.T.,and Patel, T. (2007). MicroRNA-21 regulates expression of the

PTEN tumor suppressor gene in human hepatocellular cancer.Gastroenterology 133, 647–658.

Murchison, E.P., and Hannon, G.J. (2004). miRNAs on the move:

miRNA biogenesis and the RNAi machinery. Curr Opin Cell Biol 16,223–229.

Parent, C.A., and Devreotes, P.N. (1999). A cell’s sense of direction.

Science 284, 765–770.

Pineau, P., Volinia, S., McJunkin, K., Marchio, A., Battiston, C., Terris,

B., Mazzaferro, V., Lowe, S.W., Croce, C.M., and Dejean, A.(2010). miR-221 overexpression contributes to liver tumorigenesis.Proc Natl Acad Sci U S A 107, 264–269.

Poliseno, L., Tuccoli, A., Mariani, L., Evangelista, M., Citti, L., Woods,K., Mercatanti, A., Hammond, S., and Rainaldi, G. (2006).MicroRNAs modulate the angiogenic properties of HUVECs.

Blood 108, 3068–3071.

Rana, T.M. (2007). Illuminating the silence: understanding thestructure and function of small RNAs. Nat Rev Mol Cell Biol 8,

23–36.

Reddy, S.D., Ohshiro, K., Rayala, S.K., and Kumar, R. (2008).MicroRNA-7, a homeobox D10 target, inhibits p21-activated kinase

1 and regulates its functions. Cancer Res 68, 8195–8200.

Reddy, S.D., Pakala, S.B., Ohshiro, K., Rayala, S.K., and Kumar, R.

(2009). MicroRNA-661, a c/EBPalpha target, inhibits metastatictumor antigen 1 and regulates its functions. Cancer Res 69,5639–5642.

Ridley, A.J., Schwartz, M.A., Burridge, K., Firtel, R.A., Ginsberg, M.H.,Borisy, G., Parsons, J.T., and Horwitz, A.R. (2003). Cell migration:integrating signals from front to back. Science 302, 1704–1709.

Salvi, A., Sabelli, C., Moncini, S., Venturin, M., Arici, B., Riva, P.,Portolani, N., Giulini, S.M., De Petro, G., and Barlati, S. (2009).MicroRNA-23b mediates urokinase and c-met downmodulation

and a decreased migration of human hepatocellular carcinomacells. FEBS J 276, 2966–2982.

Scott, G.K., Goga, A., Bhaumik, D., Berger, C.E., Sullivan, C.S., and

Benz, C.C. (2007). Coordinate suppression of ERBB2 and ERBB3by enforced expression of micro-RNA miR-125a or miR-125b. JBiol Chem 282, 1479–1486.

Segura, M.F., Hanniford, D., Menendez, S., Reavie, L., Zou, X.,Alvarez-Diaz, S., Zakrzewski, J., Blochin, E., Rose, A., Bogunovic,D., et al. (2009). Aberrant miR-182 expression promotes mela-

noma metastasis by repressing FOXO3 and microphthalmia-associated transcription factor. Proc Natl Acad Sci U S A 106,1814–1819.

Sehm, T., Sachse, C., Frenzel, C., and Echeverri, K. (2009). miR-196

is an essential early-stage regulator of tail regeneration, upstreamof key spinal cord patterning events. Dev Biol 334, 468–480.

Sengupta, S., den Boon, J.A., Chen, I.H., Newton, M.A., Stanhope, S.

A., Cheng, Y.J., Chen, C.J., Hildesheim, A., Sugden, B., andAhlquist, P. (2008). MicroRNA 29c is down-regulated in nasophar-yngeal carcinomas, up-regulating mRNAs encoding extracellular

matrix proteins. Proc Natl Acad Sci U S A 105, 5874–5878.

Shan, S.W., Lee, D.Y., Deng, Z., Shatseva, T., Jeyapalan, Z., Du, W.W., Zhang, Y., Xuan, J.W., Yee, S.P., Siragam, V., et al. (2009).

MicroRNA miR-17 retards tissue growth and represses fibronectinexpression. Nat Cell Biol 11, 1031–1038.

Song, H., and Bu, G. (2009). MicroRNA-205 inhibits tumor cell

migration through down-regulating the expression of the LDLreceptor-related protein 1. Biochem Biophys Res Commun 388,400–405.

Tavazoie, S.F., Alarcón, C., Oskarsson, T., Padua, D., Wang, Q., Bos,P.D., Gerald, W.L., and Massagué, J. (2008). Endogenous humanmicroRNAs that suppress breast cancer metastasis. Nature 451,

147–152.

Tian, Y., Luo, A., Cai, Y., Su, Q., Ding, F., Chen, H., and Liu, Z. (2010).

MicroRNA-10b promotes migration and invasion through KLF4 inhuman esophageal cancer cell lines. J Biol Chem 285,7986–7994.

Tsai, W.C., Hsu, P.W., Lai, T.C., Chau, G.Y., Lin, C.W., Chen, C.M.,Lin, C.D., Liao, Y.L., Wang, J.L., Chau, Y.P., et al. (2009).MicroRNA-122, a tumor suppressor microRNA that regulates

intrahepatic metastasis of hepatocellular carcinoma. Hepatology49, 1571–1582.

Valastyan, S., Reinhardt, F., Benaich, N., Calogrias, D., Szász, A.M.,

Wang, Z.C., Brock, J.E., Richardson, A.L., and Weinberg, R.A.(2009). A pleiotropically acting microRNA, miR-31, inhibits breastcancer metastasis. Cell 137, 1032–1046.

Wang, B., Hsu, S.H., Majumder, S., Kutay, H., Huang, W., Jacob, S.T.,and Ghoshal, K. (2010a). TGFbeta-mediated upregulation ofhepatic miR-181b promotes hepatocarcinogenesis by targeting

TIMP3. Oncogene 29, 1787–1797.

Wang, C.H., Lee, D.Y., Deng, Z., Jeyapalan, Z., Lee, S.C., Kahai, S.,Lu, W.Y., Zhang, Y., Yang, B.B., and Bähler, J. (2008a). MicroRNA

miR-328 regulates zonation morphogenesis by targeting CD44expression. PLoS ONE 3, e2420.

Wang, G., Mao, W., and Zheng, S. (2008b). MicroRNA-183 regulatesEzrin expression in lung cancer cells. FEBS Lett 582, 3663–3668.

Wang, G., Mao, W., Zheng, S., and Ye, J. (2009). Epidermal growth

factor receptor-regulated miR-125a-5p—a metastatic inhibitor oflung cancer. FEBS J 276, 5571–5578.

Wang, X., Hu, G., and Zhou, J. (2010b). Repression of versican

expression by microRNA-143. J Biol Chem 285, 23241–23250.

Xia, H., Qi, Y., Ng, S.S., Chen, X., Li, D., Chen, S., Ge, R., Jiang, S.,

Li, G., Chen, Y., et al. (2009). microRNA-146b inhibits glioma cellmigration and invasion by targeting MMPs. Brain Res 1269,158–165.

Yan, D., Dong, X.E., Chen, X., Wang, L., Lu, C., Wang, J., Qu, J., andTu, L. (2009). MicroRNA-1/206 targets c-Met and inhibitsrhabdomyosarcoma development. J Biol Chem 284,

29596–29604.

Yao, J., Liang, L., Huang, S., Ding, J., Tan, N., Zhao, Y., Yan, M., Ge,C., Zhang, Z., Chen, T., et al. (2010). MicroRNA-30d promotes



tumor invasion and metastasis by targeting Galphai2 in hepato-cellular carcinoma. Hepatology 51, 846–856.

Yu, Z., Willmarth, N.E., Zhou, J., Katiyar, S., Wang, M., Liu, Y., McCue,P.A., Quong, A.A., Lisanti, M.P., and Pestell, R.G. (2010).microRNA 17/20 inhibits cellular invasion and tumor metastasisin breast cancer by heterotypic signaling. Proc Natl Acad Sci U S A

107, 8231–8236.

Zhang, X., Liu, S., Hu, T., Liu, S., He, Y., and Sun, S. (2009). Up-

regulated microRNA-143 transcribed by nuclear factor kappa Benhances hepatocarcinoma metastasis by repressing fibronectinexpression. Hepatology 50, 490–499.

Zhang, Y., Liu, D., Chen, X., Li, J., Li, L., Bian, Z., Sun, F., Lu, J., Yin,Y., Cai, X., et al. (2010). Secreted monocytic miR-150 enhancestargeted endothelial cell migration. Mol Cell 39, 133–144.

Protein & Cell



COMMUNICATION

A transcription assay for EWS oncoproteins inXenopus oocytes

King Pan Ng, Felix Cheung, Kevin A.W. Lee✉

Department of Biology, The Hong Kong University of Science and Technology, Clear Water Bay, Kowloon Hong Kong, China✉ Correspondence: [email protected] September 6, 2010 Accepted September 18, 2010

ABSTRACT

Aberrant chromosomal fusion of the Ewing's sarcomaoncogene (EWS) to several different cellular partnersproduces the Ewing's family of oncoproteins (EWS-fusion-proteins, EFPs) and associated tumors (EFTs).EFPs are potent transcriptional activators, dependent onthe N-terminal region of EWS (the EWS-activation-domain, EAD) and this function is thought to be centralto EFT oncogenesis and maintenance. Thus EFPs arepromising therapeutic targets, but detailed molecularstudies will be pivotal for exploring this potential. Suchstudies have so far largely been restricted to intactmammalian cells while recent evidence has indicatedthat a mammalian cell-free transcription system may notsupport bona fide EAD function. Therefore, the lack ofmanipulatable assays for the EAD presents a significantbarrier to progress. Using Xenopus laevis oocytes wedescribe a plasmid-based micro-injection assay thatsupports efficient, bona fide EAD transcriptional activityand hence provides a new vehicle for molecular dissec-tion of the EAD.

KEYWORDS EWS/ATF1, Ewing's sarcoma, microinjec-tion, Xenopus oocytes, transcription, EWS-activation domain

INTRODUCTION

Chromosomal translocations involving the Ewing's sarcomaoncogene (EWS) or the related genes TAF15 and TLS/FUS(the TET family, Law et al., 2006) give rise to the Ewing'sfamily of oncoproteins (EWS-fusion-proteins, EFPs) and theirassociated tumors (EFTs; Kim and Pelletier, 1999; Arvandand Denny, 2001; Janknecht, 2005). EFPs include EWS/FLI1(EFT: Ewings sarcoma), EWS/WT1 (desmoplastic smallround cell tumor), EWS/ATF1 (clear cell sarcoma), EWS/

TEC (chondrosarcoma), EWS/ZSG (small round cell tumor),TLS/ERG (myeloid leukemia) and TLS/CHOP (liposarcoma).EFPs are gene-specific transcriptional activators dependenton the N-terminal region of EWS (the EWS-activation-domain, EAD) and a C-terminal DNA binding domainprovided by the particular fusion partner (Kim and Pelletier,1999). Transcriptional de-regulation by EFPs is most likelycentral to EFT oncogenesis but other effects of EFPs,including transcriptional repression or perturbation of pre-mRNA splicing are also likely to be important (Arvand andDenny, 2001).

EFPs may be promising therapeutic targets (Kovar et al.,1999) because they are absolutely tumor specific and theirfunction is quite distinct from the parental TET proteins(Arvand and Denny, 2001). In addition at least in some cases(Prieur et al., 2004; Davis et al., 2006) EFPs appear to play arole in tumor cell survival, indicating the therapeutic potentialof EFP inhibitors. Detailed biochemical analysis of EFPs willbe essential for exploration of the above possibility and theEAD is of particular interest because it is common to the entireEFT family.

Transient introduction of exogenous EFPs into mammaliantissue culture cells has uncovered some essential structure/function relations for the EAD (Feng and Lee, 2001; Ng et al.,2007). Notably, systematic mutational analysis of intact EADshowed that multiple dispersed Tyr residues are crucial forEAD function in mammalian cells (Ng et al., 2007). Ofrelevance to the current study, the availability of informativeEAD mutants (Fig. 1C) provides an essential tool forevaluating novel assays for the EAD. Despite the aboveprogress several other factors hindered biochemical char-acterization of the EAD. First, the EAD is complex (spanning~250 residues) and harbors dispersed functional elements(Pan et al., 1998). Second, the highly biased composition of theEAD (resulting from the presence of a degenerate hexapeptiderepeat [DHR] with consensus SYGQQS, Fig. 1B) confers the


Protein Cell 2010, 1(10): 927–934DOI 10.1007/s13238-010-0114-y

Protein & Cell

properties of an intrinsically disordered protein (IDP) region (Nget al., 2007) and thus precludes classical structural analysis.Third, the EAD most likely interacts with a complex array ofproteins (Rual et al., 2005) and identification of critical EADpartners has not been achieved.

One further barrier to biochemical studies of the EAD is thatavailable functional assays in mammalian cells and yeast(Zhou and Lee, 2001) are not easy to manipulate. Crucially ithas also been shown that mammalian cell extracts do notsupport bona fide EAD function (Ng et al., 2009). Develop-ment of more tractable experimental systems to study theEAD is therefore of significance. Xenopus leavis oocytes offersome advantages as a heterologous host system formammalian transcription factors. Exogenous mammalian

promoters can be activated by the endogenous oocytetranscription machinery (Gurdon and Wickens, 1983; GurdonandWakefied, 1986). In addition, PolII promoters are active atlow template levels in oocytes (Gurdon and Wakefied, 1986)and are efficiently assembled into chromatin thus favoringsupport of regulated events (Gurdon and Wickens, 1983).Finally, the large size of oocytes (relative to mammalian cells)readily allows for microinjection of macromolecules (RNA,proteins/antibodies and peptides) and thus greatly facilitatestesting of potential inhibitors. Here we describe the use ofXenopus oocytes and a plasmid-based micro-injection assaythat supports efficient, bona fide EAD transciptional activityand thus provides a new vehicle for molecular dissection ofthe EAD.

Figure 1. Structures of EWS/ATF1 (EFPs), EAD, and experimental activators and reporters. (A) Structure of EWS/ATF1

(EFPs). The Ewing's family of oncoproteins (EFPs) contains the N-terminal region of the Ewing's sarcoma oncogene (EWS) fused tovarious transcription factor partners (Kim and Pelletier, 1999). All EFPs contain at least EWS residues 1–264 (the EWS-activation-domain, EAD) and a sequence-specific DNA binding domain from distinct partners. EWS/ATF1 is a typical EFP and contains EWSresidues 1–325 fused to the C-terminal region of ATF1 (ΔATF1, residues 66–271), including the DNA binding (bZIP) domain (Zucman

et al., 1993). EWS/ATF1 is a potent constitutive activator of ATF-dependent promoters (Ng et al., 2007) dependent on the EAD andthe bZIP domain of ATF1. (B) Primary structure of the EAD. The EAD contains multiple degenerate hexapeptide repeats (DHRs,purple boxes) with consensus sequence SYGQQS. DHR degeneracy is indicated by the % occurrence of conserved residues

including the absolutely conserved Tyr residue (red). Seven additional Tyr residues (dark gray boxes) are also present. Spacesbetween DHRs are generally only a few residues except in two cases (white boxes) of 12 and 25 residues, respectively. Several SH2binding sites (YxxP, black circle) and SH3 binding sites (PxxP, open triangle) are indicated. (C) Structure of experimental activators

and reporters. EZA contains the intact EAD (EWS residues 1–245), ΔATF1 and the zta bZIP domain (Ribeiro et al., 1994), ZΔE isequivalent to EZA but lacks the EAD. N3Z contains residues 1–166 of the EAD but is otherwise equivalent to EZA. Several Tyrresidues within DHRs (purple boxes in N3Z) that are critical for transcriptional activation in mammalian cells (Ng et al., 2007) areshown. N3ZA and N3ZI are essentially the same as N3Z except that the Tyr residues highlighted (see materials and methods for

precise coordinates) are changed to Ala (N3ZA, blue) or Isoleucine (N3ZI, orange). eN3Z corresponds to N3Z with EGFP fused to theN terminus and likewise for eN3ZA and eN3ZI (not shown). Expression vectors pSVEZA (Li and Lee, 2000), pZΔE (Feng and Lee,2001), pN3Z, pN3ZA and pN3ZI (Ng et al., 2007) were described elsewhere. All expression vectors were derived from pSG424

(Sadowski and Ptashne, 1989) containing the SV40 early promoter and polyadenylation signals for expression in Xenopus oocytes.All proteins contained the KT3 epitope PPPEPET (MacArthur and Walter, 1984) at the C-terminus.


King Pan Ng et al.Protein & Cell

RESULTS AND DISCUSSION

To test EAD function in oocytes, we exploited EWS/ATF1(Brown et al., 1995) or close derivatives (Ng et al., 2007) thatserve as useful EFP models. EWS/ATF1 is a potent activatorin mammalian cells (Brown et al., 1995) and, crucially, wehave recently characterized informative EAD mutants (Fig.1C) in this context (Ng et al., 2007). Because oocytes containendogenous CREB/ATF-related activators that could impactEWS/ATF1 activity (either via binding to ATF sites in thereporter or via heterodimerization) we employed a subtledomain swap (Ribeiro et al., 1994) involving substitution ofthe ATF1 bZIP domain with that of the EBV zta protein (Fig.1C). Promoters containing multiple zta binding sites fused toCAT (Z7E4TCAT, referred to as Z7CAT) or Luciferase (Z7Luc)reporters give essentially background activity in oocytes (Fig.2) thereby allowing response to exogenous activator.

EZA (Fig. 2) contains essentially the intact EAD (EWSresidues 1–245) fused to the part of ATF1 present in EWS/ATF1 (ΔATF1) and the zta-bZIP domain. Initially we askedwhether EZA could activate the Z7CAT reporter in oocytes.An expression vector for EZA driven by the SV40 earlypromoter that is active in oocytes (Jones et al., 1983) andZ7CAT reporter plasmid were co-injected into stageVI oocytegerminal vesicles (GV) and CAT assays (using pooledextracts, each from three oocytes) performed 40 h post-injection. Compared with Z7CAT reporter alone, EZA sig-nificantly activated transcription (48-fold) suggesting that theEAD is functional in the oocytes (Fig. 2).

The CAT reporter assay is relatively insensitive/timeconsuming and so we employed Luciferase to improve theassay and test EAD-dependence. EZA stimulates transcrip-tion of the pZ7Luc reporter ~180-fold (Fig. 2 and Table 1Exp#1) and deletion of the EAD (ZΔE) reduced activation to~8-fold (Fig. 2 and Table 1 Exp#1). Western blotting ofextracts (pooled from seven oocytes) using KT3 antibody(Fig. 2) indicates that (consistent with results in mammaliancells) EZA is expressed at lower levels than ZΔE and thus theeffect of the EAD in oocytes may be underestimated. Toassess the absolute magnitude of trans-activation by EZA weinjected oocytes with pGL3 which contains the SV40 earlypromoter expressing Luciferase (Fig. 2) and found that EZAactivity was generally about 2–4 fold less than that of pGL3(see also Fig. 4 for eN3Z). Together the above results showthat the EAD can efficiently activate transcription in Xenopusoocytes.

Due to low expression of EZA or the relative sensitivity ofthe KT3 antibody, it was not possible to scrutinize individualoocytes or determine the correlation between EZA levels andactivity. We therefore sought to sensitize the assay andestablish data for individual oocytes. Previously we haveshown in mammalian cells that proteins containing the intactEAD (such as EZA) are expressed at lower levels than thosewith deletions of the C-terminal region of EAD (Pan et al.,

1998). Thus a protein containing EAD residues 1–166 (N3Z,Fig. 1C) is expressed at higher levels than EZA and remains apotent activator in mammalian Jeg3 cells (Fig. 3A). To confirmthat introduction of the zta-bZIP domain in N3Z does not affectEAD activity we employed a mutant protein N3ZA (Fig. 1C)that harbors several Tyr to Ala substitutions (Fig. 1C) knownto inactivate the EAD (Ng et al., 2007). N3ZA is transcription-ally inactive in Jeg3 cells as expected (Fig. 3A) but isexpressed at high levels and retains full DNA binding activity(Fig. 3A). Characterization of several EAD mutants (Ng et al.,2007) has established that the effect of mutations present inN3ZA (or N3ZI, Fig. 1C) reflects the crucial role of Tyrresidues in EAD activity and does not reflect a gross proteinmalfunction engendered by the significant mutational burden.Specifically a similar degree of Ala substitutions in non-tyrosine EAD residues (Gln to Ala or Ser/Thr to Ala) does not

Figure 2. EAD-mediated trans-activation in Xenopus

oocytes. (A) The germinal vesicle of unfertilised stage VIXenopus oocytes was injected with 2 ng of a Cat reporterplasmid (pZ7E4TCAT) containing seven zta binding siteseither alone (control) or in the presence of 2 ng of vector

(pSVEZA) expressing EZA (+ EZA) containing the intact EAD.Cat assays were performed 40 h post-injection and anautoradiogram of the CATassay is shown (c, chloramphenicol;

ac, acetylated chloramphenicol). Each sample in the Cat assayshows the activity from three injected oocytes pooled together.(B) Oocytes were injected with 2 ng of a Luciferase reporter

(pZ7Luc) containing seven zta binding sites and expressionplasmids (2 ng) for EZA and ZΔE or pGL3 (an SV40 promoter-Luciferase reporter). Luciferase reporter activity (kRLU/sec ±SEM) was determined for individual oocytes and the mean

values are shown (see Table 1 Exp#1) for uninjected oocytes(U), pZ7Luc alone, EZA, ZΔE and pGL3. Western blot usingKT3 antibody (left hand side) shows expression levels for EZA

and ZΔE. Extract equivalent to 7 oocytes (derived from a totalof 34 oocytes) was loaded on the gel.


A transcription assay for EWS oncoproteins in Xenopus oocytes Protein & Cell

inactivate the EAD (Ng et al., 2007). Thus, the Tyr to Ala or Ilemutations (Fig. 1C) enable verification of authentic EADactivity in other systems. We tested N3Z in Xenopus oocytesand detected high levels of trans-activation in individualoocytes (Fig. 3B and Table 1 Exp#2). The large variation forN3Z activity in different oocytes is consistently observed (seealso Fig. 4) and is discussed later. N3ZA gives onlybackground levels of activity in oocytes (Fig. 3B and Table 1Exp#2) and thus the above results indicate that the oocyteassay reflects authentic EAD function.

Although N3Z is expressed at higher levels than EZA, wewere still unable to readily detect N3Z expression in individualoocytes using KT3 antibody (data not shown). To overcomethe above obstacle we produced enhanced green fluorescentprotein (EGFP) derivatives (eN3Z, eN3ZA and eN3Z1, Fig.1C) and exploited a very sensitive EGFP antibody (JL8antibody, Clontech). eN3Z strongly activates transcription inmammalian Jeg3 cells as control (Fig. 3C) and thecorresponding EAD mutant proteins eN3ZA and eN3ZI (Fig.1C) are defective, as expected (0.1% and 0.4% of eN3Zactivity, respectively). Western blotting using anti-EGFP JL8antibody confirms expression eN3ZA and eN3ZI proteins(Fig. 3C) and N3ZA expression is also elevated relative toeN3Z (but less so than in the case of N3ZA versus N3Z, Fig.3A). The above results show that eN3Z-derivatives behaveappropriately in mammalian cells and can serve as reliabletools for studying the EAD in oocytes (or in other systems).eN3Z strongly activated the Z7Luc reporter in oocytes (Fig.3C and Table 1 Exps#3–5). In one side-by-side test eN3Z was~2-fold more active than N3Z (Fig. 3C and Table 1 Exp#3). Asexpected from their lack of activity in mammalian cells,eN3ZA and eN3ZI had no detectable activity in oocytes (Fig.3C).

Significantly, eN3Z protein (and eN3ZA/eN3ZI) can readilybe detected in single oocytes by Western blotting using theanti-EGFP JL8 antibody (Fig. 4) and this allowed correlationof activator levels and transcriptional activity in individualoocytes. Ten oocytes were injected with pGL3 (as a robustindicator of transcriptional competence in the oocyte popula-tion) or co-injected with the Z7Luc reporter and eN3Z, eN3ZIor eN3ZA. eN3ZI and eN3ZA are both expressed in oocytes(Fig. 4) and thus the lack of activity for eN3ZA and eN3ZI inoocytes is not explained by poor expression. Trans-activationby eN3Z is detectable in ~80% of injected oocytes but variesover a wide range for individual oocytes (Fig. 4). The aboveresult is typical and is reflected by a similar and relatively largeSEM (22%–36% of the mean values) in three differentexperiments (Table 1, Exps#3–5). The small number ofinactive oocytes that do not express eN3Z is probably dueto poor injection (Gurdon and Wakefield, 1986; Guille, 1999).For active oocytes there is not a particularly strong correlationbetween activator (eN3Z) levels and transcription activity.However, variability in results for the SV40 promoter (pGL3) issimilar to that for eN3Z with SEMs equal to 27% and 28% ofthe mean value in two experiments (Table 1, Exps#1 and 4).This variation is within expectation (Gurdon and Wickens,1983) since similar variation also occurs for simple tests (GL3in our experiments) requiring activation of injected PolIIpromoters by endogenous oocyte factors (Gurdon andWickens, 1983). Thus, variable EAD activity reflects generalexperimental/oocyte variation rather than the existence ofspecific problems related to EAD-mediated trans-activation.

CONCLUSION

The DNA-based oocyte micro-injection assay described

Table 1 Data summary

activator No. oocytes Luc. activity fold activation

Exp. #1 uninjected 4 0.27 ± 0.01 −

pZ7Luc alone 17 0.34 ± 0.02 −

GL3 11 102 ± 29 −

EZA 15 59 ± 18 174

ZΔE 26 2.8 ± 0.3 8

Exp. #2 N3Z 20 86 ± 19 254

N3ZA 20 background 1

Exp. #3 N3Z 6 56 ± 12 165

eN3Z 5 125 ± 41 368

Exp. #4 eN3Z 10 78 ± 17 229

GL3 10 214 ± 58 −

Exp. #5 eN3Z 10 41 ± 15 121

eN3ZA 10 background 1

eN3ZI 10 background 1

The table shows the activator protein in each experiment (activator), number of oocytes analyzed (No. oocytes), transcriptional activity (Luc. activity,

kRLU/sec ± SEM) and fold activation compared with the pZ7Luc reporter alone (fold activation).



herein supports efficient, bona fide EAD transcriptionalfunction. Even the lower range of N3Z activity is ~100-foldhigher than the Z7Luc reporter background thus providing abroad window for experimentation. Although there is systemicvariability, robust data can typically be obtained by averagingthe N3Z activity from as few as ten oocytes. This iscomparable with other transcription assays employinginjected oocytes (Gurdon and Wickens, 1983).

Considering the advantage of single cell experiments usingXenopus oocytes, the likely conservation of cellular pathwaysrelevant to the EAD (see below) and the indication that simplemammalian cell-free transcription systems do not supportEAD function (Ng et al., 2009), Xenopus oocytes may provideadvantages for biochemical dissection of the EAD. Mostnotably compared with mammalian cells, oocytes can bereadily microinjected (with RNA, proteins/antibodies andpeptides) using relatively simple equipment and this willgreatly facilitate inhibitor studies. The EAD is a potential targetfor drug design and the ability to test macromolecular

inhibitors under well defined conditions will enable effectiveevaluation of early lead compounds. With regard to develop-ment of a soluble transcription system for the EAD, it remainsto be seen whether oocyte extracts will faithfully support EADactivity or whether they will exhibit the same deficiency asmammalian cell extracts (Ng et al., 2009).

The EAD is not broadly conserved in the animal kingdombut is highly conserved between frogs, zebra fish (Azuma etal., 2007) and mammals, including multiple Tyr phosphoryla-tion sites and SH2/SH3 binding motifs (Fig. 1B). Frog oocytesare thus likely to harbor crucial endogenous factors that haveco-evolved with the EAD and this should be advantageousover alternative (non-mammalian) systems for EAD analysis,such as yeast (Zhou and Lee, 2001). For example Tyrphosphorylation can modulate EAD activity under particularcircumstances (Kim et al., 1999, 2000) and other phosphor-ylation events (Olsen and Hinrichs, 2001) and O-GlcNAcyla-tion (Bachmaier et al., 2009) have been shown to influenceEFP/EAD activity.

Figure 3. Analysis of N3Z and EGFP-N3Z proteins. (A) Transcriptional activity of N3Z and N3ZA in mammalian cells wasdetected by Cat assay using the Z7Cat reporter. Western blot analysis using KT3 antibody shows activator expression levels. DNAbinding activity was detected by gel mobility shift assay (right hand side). Extracts from transfected cells were incubated with 1ng of

labeled probe containing a single zta binding site. DNA-protein complexes were resolved on non-denaturing polyacrylamide gels anddetected by autoradiography. Positions of unbound DNA probe and DNA-protein complexes are indicated to the right. The presenceof excess competitor oligonucleotide (100 ng) containing a functional (wt) or mutated (mt) zta binding site is indicated. (B)

Transcriptional activity (kRLU/sec ± SEM) of N3Z and derivatives in oocytes was determined as described in Fig. 2 using the Z7Lucreporter. The figure shows the individual data for each injected oocyte (left hand plot) and the mean values with SEM (right hand side,see Table 1 Exp#2). (C) Transcriptional activity of EGFP-N3Z derivative (eN3Z) and mutants eN3ZA and eN3ZI was determined inmammalian cells by Cat assay using the Z7Cat reporter (left hand side). AWestern blot using αEGFP antibody JL8 (Clontech) shows

activator expression levels. The right hand graph shows the activity of eN3Z in oocytes using the Z7Luc reporter (mean values andSEM are shown, see Table 1 Exps#3–5). eN3ZA and eN3ZI both gave only background activity (b).



MATERIALS AND METHODS

Plasmids

All expression vectors were derived from pSG424 (Sadowski and

Ptashne, 1989) containing the SV40 early promoter and polyadenyla-tion signals that allow expression in both mammalian cells andXenopus oocytes. Expression vectors pSVEZA (Li and Lee, 2000)

and pZΔE (Feng and Lee, 2001) are previously described. pN3Z wasderived from pΔ167C (Pan et al., 1998) by replacing the ATF1 bZIPdomain with the zta bZIP domain using an Nde1 site engineered for

bZIP domain swaps (Ribeiro et al., 1994). pN3ZA and pN3ZI wereobtained by inserting HindIII/BglII ended synthetic DNA fragmentsobtained by total gene synthesis (TOP Gene Technologies, Montreal)directly into pZΔE digested withHindIII/BglII. pZ7E4CAT (Carey et al.,

1992) and pGL3-control vector (Promega) are described elsewhere.pZ7Luc was obtained by inserting a HindIII/KpnI fragment frompZ7E4TCAT into pVIPRSVluc vector (Masson et al., 1992). peN3Z

was prepared by joining a AgeI/SalI fragment from pEGFPC-3(Clontech) and a HindIII/SalI fragment from pN3Z, using anoligonucleotide with HindIII/AgeI overhangs. peN3ZA and peN3ZI

were obtained by replacing the wild type EAD sequence betweenHindIII/BglII in peN3Z with Hind3/Bgl2 ended synthetic DNAfragments harboring the desired mutations. All proteins contained

the KT3 monoclonal epitope PPPEPET MacArthur and Walter, 1984)at the C-terminal adjacent to the zta bZIP domain.

EAD mutants

The EAD mutations exploited in this study (Fig. 1C) have beenfunctionally characterized in mammalian cells in the context of the

natural EFP, EWS/ATF1 (Ng et al., 2007). The above proteins havemultiple Tyr to Ala (N3ZA and eN3ZA) or Tyr to Ile (eN3ZI) changesdispersed throughout the EAD (corresponding to Tyr residues 9, 18,

29, 36, 44, 52, 61, 70, 77, 86, 93, 112, 118, 127, 158 and 165 in thenormal EWS protein). Characterization of several other EAD mutants(Ng et al., 2007) have established that the effect of the Tyr to Ala/Ilemutations reflects the critical role of multiple Tyr residues in EAD

activity, as opposed to a gross protein malfunction engendered by themutational burden. Specifically, a similar degree of Ala substitutions innon-tyrosine EAD residues (Gln to Ala or Ser/Thr to Ala) do not

inactivate the EAD (Ng et al., 2007). Thus, the Tyr to Ala/Ile mutantsare valuable tools for verifying authentic EAD activity in othersystems.

Oocyte preparation

Anesthesia of frogs using 0.1% ethyl m-aminobenzoate, surgery and

oocyte maintenance prior to injection were broadly according toestablished procedure (Goldin, 1992). Lobes were washed exten-sively in calcium-free OR2medium (OR2MC, contains 82.5mMNaCl,2.5mM KCl, 1 mM MgCl2 and 5mM Hepes, adjusted to pH 7.5 using

NaOH), placed in a 50mL tube containing 20mL of collagenase(1mg/mL; Sigma C5138; 388 U/mg [collagen digestion activity] or0.9 U/mg FALGPA hydrolysis activity) in OR2MC and incubated at

room temperature with very gentle agitation on a rocking platform for20min. The collagenase solution was then removed, oocytes washedin calcium-free OR2 medium and then incubated in fresh collagenase

for a further 40min. Separated oocytes were further washed inOR2MC, stage VI oocytes were selected and kept at 18°C and at lowdensity (about 20 oocytes per 10 cm petri dish) in Modified Barthssolution (MBS).

Microinjection

Visually healthy oocytes were injected the day after surgery.

Preparation of needles and microinjection were performed byestablished procedures (Guille, 1999). Needles with external dia-meter of 12–15 μm were used for the microinjection and were

calibrated by filling with trypan blue and injecting into light mineral oilusing a Medical Systems Corporation PLI-100 micro-injector. Forexperiments ~10 nL of DNA sample in water (containing 2 ng of

activator plasmid and 2 ng of reporter plasmid) was injected into theoocyte germinal vesicle (GV) and this was commonly achieved usingan injection pressure of 10–15 psi for 60–l00ms. Use of 4 ng ofinjected plasmid is just sub-saturating (Gurdon and Wickens, 1983)

and should be optimal for transcriptional studies. Typically 10–20oocytes were injected for each test and following injection, individualoocytes were kept in isolation at 18°C in MBS. Success rate for GV

injection and subsequent oocyte viability was>80% as expected(Gurdon and Wakefield, 1986; Guille, 1999) and reporter activity wasassayed 40 h post-injection. Mean values ± the standard error (SEM)

were calculated and presented in the results summary Table 1.

Figure 4. Analysis of activator levels and EAD activity inindividual oocytes. For pGL3 control reporter (blue bars,

average activity 214 ± 58 kRLU/sec) and activation of Z7Luc byeN3Z (green bars, average 78 ± 17 kRLU/sec) the activitydetected in ten individual oocytes is shown. The correspondinglevel of eN3Z expression for each oocyte (detected byWestern

blot using αEFGP JL8 antibody) is shown below the activitygraph. Expression levels for eN3ZA and eN3ZI in individualoocytes (detected using αEFGP) is shown for the same

experiment as eN3Z. eN3ZA and eN3ZI yield only backgroundsignals for trans-activation (Fig. 3C).



Oocyte reporter assays

Injected oocytes were lysed by vigorous resuspension in 100 μL of

MBS using a micropipette yellow tip. Cell extract supernatant wasobtained by centrifugation at 12,000 rpm for 5min in an Eppendorfmicrofuge. Luciferase assays were performing by mixing 50 μL of

extract with 50 μL of Steady-Glo Luciferase substrate (Promega).

Mammalian reporter assays

Freshly passaged cells were transfected by calcium phosphate co-

precipitation method as previously described (Brown et al., 1995) with5 μg of plasmid expressing pN3Z/peN3Z (or EAD mutants), 5 μg ofpZ7E4TCAT as reporter and 15 μg of pGem3 as carrier. CAT assays

were performed at 40 h post-transfection as previously described(Brown et al., 1995).

DNA binding assays

Gel mobility shift assays using extracts from mammalian cells wereperformed as previously described (Ribeiro et al., 1994; Krajewski

and Lee, 1994). The labeled zta DNA probe contains a single ztabinding site with the TTGCTAA core motif and competitor oligonu-cleotides have either a wt zta binding site or a mutant site with thecore motif changed to GACACAC. 15 μL binding reactions contained

~1.5 ng of labeled DNA probe and 1 μg of poly(dIdC) and wereincubated for 20min at 30°C. Half of the reaction mixture wasresolved on 5% polyacrylamide gels run in 50% TBE. Gels were fixed

in 50% methanol and exposed at minus 80°C for ~ 2 h against KodakBiomax XAR film.

Western blotting

Western blotting was performed in PBS containing 3% dried milk.Tagged proteins were detected either using primary antibody KT3

(MacArthur and Walter, 1984) and alkaline phosphatase conjugatedanti-mouse secondary antibody (DAKO D0486) or EGFP antibody(mouse monoclonal JL8, Clontech) and anti-mouse HRP conjugatedsecondary antibody (Amersham NA931) and detection using an ECL

kit (Amersham NA931).

ACKNOWLEDGEMENTS

This work was supported by The Association for International CancerResearch (AICR) (grant 03-131 to K.A.W.L.). We also thank TOPGene Technologies Inc., Montreal, for efficient and economical genesynthesis.

ABBREVIATIONS

EWS, Ewing's sarcoma oncogene; EFP, EWS-fusion-protein; EAD,

EWS-activation-domain; EFT, EWS-family-tumor; IDP, intrinsicallydisordered protein; DHR, degenerate hexapeptide repeat; MBS,Modified Barths solution; GV, germinal vesicle; CAT, chloramphenicolacetyl transferase; SEM, standard error of the mean

REFERENCES

Arvand, A., and Denny, C.T. (2001). Biology of EWS/ETS fusions in

Ewing’s family tumors. Oncogene 20, 5747–5754.

Azuma, M., Embree, L.J., Sabaawy, H., and Hickstein, D.D. (2007).

Ewing sarcoma protein Ewsr1 maintains mitotic integrity andproneural cell survival in the zebra fish embryo. PLoS ONE 10,e979.

Bachmaier, R., Aryee, D.N.T., Jug, G., Kauer, M., Kreppel, M., Lee, K.A.W., and Kovar, H. (2009). O-GlcNAcylation is involved in thetranscriptional activity of EWS-FLI1 in Ewing’s sarcoma. Onco-

gene 28, 1280–1284.

Brown, A.D., Lopez-Terrada, D., Denny, C.T., and Lee, K.A.W. (1995).Promoters containing ATF-binding sites are de-regulated in

tumour-derived cell lines that express the EWS/ATF1 oncogene.Oncogene 10, 1749–1756.

Carey, M., Kolman, J., Katz, D.A., Gradoville, L., Barberis, L., and

Miller, G. (1992). Transcriptional synergy by the Epstein-Barr virustransactivator ZEBRA. J Virol 66, 4803–4813.

Davis, I.J., Kim, J.J., Ozsolak, F., Widlund, H.R., Rozenblatt-Rosen,O., Granter, S.R., Du, J., Fletcher, J.A., Denny, C.T., Lessnick, S.L.,et al. (2006). Oncogenic MITF dysregulation in clear cell sarcoma:defining the MiT family of human cancers. Cancer Cell 9, 473–484.

Feng, L., and Lee, K.A.W. (2001). A repetitive element containing acritical tyrosine residue is required for transcriptional activation by

the EWS/ATF1 oncogene. Oncogene 20, 4161–4168.

Goldin, A.L. (1992). Maintenance of Xenopus laevis and oocyteinjection. Methods Enzymol 207, 266–279.

Guille, M. (1999). Microinjection into Xenopus oocytes and embryos.Methods Mol Biol 127, 111–123.

Gurdon, J.B., and Wakefield, L. (1986). Microinjection of AmphibianOocytes and Eggs for the Analysis of Transcription. Microinjectionand Organelle Transplantation Techniques. Academic Press Inc.,

pp 269–299.

Gurdon, J.B., andWickens, M.P. (1983). The use of Xenopus oocytesfor the expression of cloned genes. Methods Enzymol 101,

370–386.

Janknecht, R. (2005). EWS-ETS oncoproteins: the linchpins of Ewingtumors. Gene 363, 1–14.

Jones, N.C., Richter, J.D., Weeks, D.L., and Smith, L.D. (1983).Regulation of adenovirus transcription by an E1a gene in

microinjected Xenopus laevis oocytes. Mol Cell Biol 3, 2131–2142.

Kim, J., Lee, J.M., Branton, P.E., and Pelletier, J. (1999). Modification

of EWS/WT1 functional properties by phosphorylation. Proc NatlAcad Sci U S A 96, 14300–14305.

Kim, J., Lee, J.M., Branton, P.E., and Pelletier, J. (2000). Modulation

of EWS/WT1 activity by the v-Src protein tyrosine kinase. FEBSLett 474, 121–128.

Kim, J., and Pelletier, J. (1999). Molecular genetics of chromosometranslocations involving EWS and related family members. PhysiolGenomics 1, 127–138.

Kovar, H., Aryee, D., and Zoubek, A. (1999). The Ewing family oftumors and the search for the Achilles’ heel. Curr Opin Oncol 11,275–284.

Krajewski, W., and Lee, K.A.W. (1994). A monomeric derivative of thecellular transcription factor CREB functions as a constitutiveactivator. Mol Cell Biol 14, 7204–7210.

Law,W.J., Cann, K.L., and Hicks, G.G. (2006). TLS, EWS and TAF15:a model for transcriptional integration of gene expression. BriefFunct Genomics Proteomics 5, 8–14.



Li, K.K.C., and Lee, K.A.W. (2000). Transcriptional activation by the

Ewing’s sarcoma (EWS) oncogene can be cis-repressed by theEWS RNA-binding domain. J Biol Chem 275, 23053–23058.

MacArthur, H., and Walter, G. (1984). Monoclonal antibodies specific

for the carboxy terminus of simian virus 40 large T antigen. J Virol52, 483–491.

Masson, N., Ellis, M., Goodbourn, S., and Lee, K.A.W. (1992). Cyclic

AMP response element-binding protein and the catalytic subunit ofprotein kinase A are present in F9 embryonal carcinoma cells butare unable to activate the somatostatin promoter. Mol Cell Biol 12,

1096–1106.

Ng, K.P., Li, K.K.C., and Lee, K.A.W. (2009). In vitro activity of theEWS oncogene transcriptional activation domain. Biochemistry 48,

2849–2857.

Ng, K.P., Potikyan, G., Savene, R.O., Denny, C.T., Uversky, V.N., and

Lee, K.A.W. (2007). Multiple aromatic side chains within adisordered structure are critical for transcription and transformingactivity of EWS family oncoproteins. Proc Natl Acad Sci U S A 104,479–484.

Olsen, R.J., and Hinrichs, S.H. (2001). Phosphorylation of the EWSIQ domain regulates transcriptional activity of the EWS/ATF1 and

EWS/FLI1 fusion proteins. Oncogene 20, 1756–1764.

Pan, S., Ming, K.Y., Dunn, T.A., Li, K.K.C., and Lee, K.A.W. (1998).

The EWS/ATF1 fusion protein contains a dispersed activationdomain that functions directly. Oncogene 16, 1625–1631.

Prieur, A., Tirode, F., Cohen, P., and Delattre, O. (2004). EWS/FLI-1silencing and gene profiling of Ewing cells reveal downstreamoncogenic pathways and a crucial role for repression of insulin-likegrowth factor binding protein 3. Mol Cell Biol 24, 7275–7283.

Ribeiro, A., Brown, A.D., and Lee, K.A.W. (1994). An in vivo assay formembers of the CREB family of transcription factors. J Biol Chem

269, 31124–31128.

Rual, J.F., Venkatesan, K., Hao, T., Hirozane-Kishikawa, T., Dricot, A.,Li, N., Berriz, G.F., Gibbons, F.D., Dreze, M., Ayivi-Guedehoussou,

N., et al. (2005). Towards a proteome-scale map of the humanprotein-protein interaction network. Nature 437, 1173–1178.

Sadowski, I., and Ptashne, M. (1989). A vector for expressing GAL4

(1–147) fusions in mammalian cells. Nucleic Acids Res 17, 7539.

Zhou, H., and Lee, K.A.W. (2001). An hsRPB4/7-dependent yeast

assay for trans-activation by the EWS oncogene. Oncogene 20,1519–1524.

Zucman, J., Delattre, O., Desmaze, C., Epstein, A.L., Stenman, G.,

Speleman, F., Fletchers, C.D.M., Aurias, A., and Thomas, G.(1993). EWS and ATF-1 gene fusion induced by t(12;22)translocation in malignant melanoma of soft parts. Nat Genet 4,

341–345.



RESEARCH ARTICLE

Trafficking abnormality and ER stress underliefunctional deficiency of hearing impairment-associated connexin-31 mutants

Kun Xia1*, Hong Ma2*, Hui Xiong2*, Qian Pan1, Liangqun Huang1, Danling Wang1, Zhuohua Zhang1,2✉

1 State Key Laboratory of Medical Genetics, Central South University, Changsha 410083, China2 Burnham Institute for Medical Research, La Jolla, CA 92037, USA✉ Correspondence: [email protected] September 26, 2010 Accepted October 9, 2010

ABSTRACT

Hearing impairment (HI) affects 1/1000 children and over2% of the aged population. We have previously reportedthat mutations in the gene encoding gap junction proteinconnexin-31 (Cx31) are associated with HI. The patholo-gical mechanism of the disease mutations remainsunknown. Here, we show that expression of Cx31 in themouse inner ear is developmentally regulated with a highlevel in adult inner hair cells and spiral ganglion neuronsthat are critical for the hearing process. In transfectedcells, wild type Cx31 protein (Cx31wt) forms functionalgap junction at cell-cell-contacts. In contrast, two HI-associated Cx31 mutants, Cx31R180X and Cx31E183Kresided primarily in the ER and Golgi-like intracellularpunctate structures, respectively, and failed to mediatelucifer yellow transfer. Expression of Cx31 mutants butnot Cx31wt leads to upregulation of and increasedassociation with the ER chaperone BiP indicating ERstress induction. Together, the HI-associated Cx31mutants are impaired in trafficking, promote ER stress,and hence lose the ability to assemble functional gapjunctions. The study reveals a potential pathologicalmechanism of HI-associated Cx31 mutations.

KEYWORDS gap junction, bip, inner ear, protein folding

INTRODUCTION

HI has dramatic effects on speech acquisition and literacywhen it presents in early childhood and seriously compro-mises the quality of life in individuals affected at late onset.

Over 100 genes are associated HI in human with more than30 being identified. Among these identified genes associatedwith HI, several encode gap junction proteins (Petit et al.,2001).

Gap junctions, formed by hemichannels of the same ordifferent connexins, mediate cell-cell communication by directexchange of intracellular small molecules (≤1 kDa) (Elfganget al., 1995). Gene disruption studies in mice demonstrateessential roles of connexins in the development of variousorgans and in the maintenance of cellular homeostasis(Simon and Goodenough, 1998; Cohen-Salmon et al.,2002). Mutations in connexin genes are linked to multiplehuman diseases, including hearing loss, neuropathy, skin andheart diseases (Goodenough et al., 1996; Bone et al., 1997;Denoyelle et al., 1997; Zelante et al., 1997; Simon andGoodenough, 1998; Grifa et al., 1999; Kelsell et al., 2001a, b;Petit et al., 2001; Watts and Chance, 2002).

Mutations in connexin-26 (Cx26), connexin-30 (Cx30),connexin 30.3 (Cx30.3) and connexin-31 (Cx31) are linked toboth non-syndromic and syndromic deafness (Denoyelleet al., 1997; Kelsell et al., 1997; Zelante et al., 1997; Xia etal., 1998; Grifa et al., 1999; Lopez-Bigas et al., 2001, 2002b).Electrophysiological studies of Xenopus oocytes expressingdifferent connexin mutants revealed that HI-associated Cx26mutants failed to generate intercellular conductance (Whiteet al., 1998; Bruzzone et al., 2001, 2003). Coexpression ofwild type and Cx26- or Cx30-disease mutants markedlyinhibited the intercellular conductance (White et al., 1998;Grifa et al., 1999; Bruzzone et al., 2001). These findingsindicate that the human-disease-linked mutations in connex-ins impair gap junctional function and may dominant-negatively overwhelm their wild-type counterparts.

*These authors contributed equally to the work.



Protein & Cell

Mutations in Cx31 are identified from patients with HI,erythrokeratodermia variabilis (EKV), and peripheral neuro-pathy (Richard et al., 1998; Xia et al., 1998; Lopez-Bigaset al., 2001). Interestingly, Cx31 mutants identified from HIpatients and skin-disease patients show exclusive diseasephenotypes (Richard et al., 1998; Xia et al., 1998). In addition,disruption of the Cx31 gene results only in transient placentaldysmorphogenesis that does not explain its involvement ineither HI or skin disease (Dahl et al., 1996; Reuss et al., 1996;Plum et al., 2001). Therefore, deciphering functional mechan-isms of Cx31 disease mutations likely yield importantinformation about the roles of Cx31 not only in normalphysiological processes but also in initiating pathogenesis ofmultiple diseases including HI and skin diseases.

In this study, we showed expression of the Cx31 protein inthe mouse cochlea, a peripheral organ of hearing. Two HI-associated Cx31 mutants, Cx31R180X (C-terminal deletionafter amino acid 180) and Cx31E183K (E to K substitution atamino acid 183), were neither assembled into gap junctionalplaques nor functional in the lucifer dye transfer assay. Wefurther demonstrated that the two HI mutant proteins wereimpaired in trafficking and induced ER stress.

RESULTS AND DISCUSSION

Expression of Cx31 in developing mouse inner ears

To investigate the involvement of Cx31 in hearing, we firstdetermined its expression in inner ear, a primary organ of thehearing process. Expression of Cx31 in inner ear tissuesdissected from postnatal day 0 (P0), day 3 (P3), day 6 (P6),day 9 (P9) and adult mice was examined with a Cx31 specificantibody. A strong Cx31 signal was detected in the spiralganglion neurons from P0 to adulthood (Fig. 1, left panel). Inthe organ of Corti, Cx31 is highly expressed in Hensen’s cellsfrom P0 to adult (Fig. 1, right panel). Cx31 was found only atthe apical surface of the outer hair cells beginning at P3 andreached the highest level at P9 (Fig. 1D, F, H, and J).Immunoreactivity in adult outer hair cells appeared to bepresent throughout the cuticular plate, and not just at theedges where contacts with Deiters’ cell processes occur. Ininner hair cells, Cx31 was first detected at P3 and reached thehighest level of expression in adulthood (Fig. 1D, F, H, and J).Preimmune serum yielded no specific staining of inner eartissue (not shown).

HI-associated Cx31 mutant proteins fail to assemble intogap junctional plaques

We next determined the gap junction formation by Cx31wtand two Cx31 mutants (E183K and R180X) identified fromindividuals with high-frequency HI (Xia et al., 1998). In orderto identify exogenously expressed Cx31 variants, a GFP or amyc-epitope tag was added to the C-termini of the proteins.

GFP- or myc-tagged Cx31 variants were expressed in HeLacells (Fig. 2A) as well as COS, HEK293, HT1080 and HaCatcells (not shown). Cx31wt and mutant Cx31 proteins weredetected as doublets by immunoblotting (Fig. 2A). Thedoublets are resistant to phosphotase treatment, likely dueto protein degradation (not shown).

Immunofluorescence revealed that both Cx31wtGFP andCx31wtmyc were assembled into patch-like structures at cell-cell contacts (Fig. 2B and 2C), indicating that Cx31wt wasassembled effectively into gap junctions. In contrast, no gapjunctional plaque-like structures were detected in cellsexpressing Cx31E183K (including E183KGFP andE183Kmyc) or Cx31R180X (R180XGFP and R180Xmyc).Similar observation was made in both transient transfectants(Fig. 2B) and in stably expressors of the Cx31 variants(Fig. 2C). Thus, HI-associated Cx31 mutations fail to assem-ble into microscopic detectable gap junctions.

To examine whether the Cx31 mutants can form amicroscopic undetectable but functional gap junction, weperformed lucifer yellow dye transfer assays. Clusters ofHeLa cells expressing GFP-tagged Cx31 variants wereidentified under fluorescence microscopy. One cell in eachcluster was injected with 4% lucifer yellow. In cells transfectedwith Cx31wt, lucifer yellow was transferred into adjoiningGFP-brightened cells but not to non-GFP-expressing cellswithin 3min after dye injection. In contrast, lucifer yellowtransfer was not detected in either Cx31E183K orCx31R180X transfectants even 20min after dye injection.Representative images of the lucifer yellow transfer assay inCx31 variant transfectants are shown in Fig. 2D. Consistentwith a previous report that the hearing loss and neuropathyassociated Cx31 D66del does not form channel activity (Diet al., 2002), the HI-associated Cx31 mutant proteins do notform functional gap junctions. These results suggest that gapjunctional activity of Cx31 likely play important roles in normalhearing.

Differential degradation and subcellular localization ofHI-associated Cx31 mutant protein

To elucidate the molecular mechanism for the inability ofCx31 mutant proteins to assemble into functional gapjunction, we examined the half-life of Cx31 variantsexpressed in HeLa cells (Fig. 3) and COS cells (not shown).The Cx31wt and Cx31E183K proteins exhibited a similar half-life of slightly more than 4h. The Cx31R180X protein wasdegraded faster than Cx31wt and Cx31E183K, with a half-lifeof about 1 h (Fig. 3B). Similar results were obtained in threeindependent experiments. Thus, rapid turn-over of the mutantprotein may constitute a mechanism for defects in gapjunction assembly seen in cells expressing Cx31R180X butnot Cx31E183K.

Immunofluorescence revealed Cx31wt at cell-cellcontacts and its co-localization with the ER protein calnexin


Kun Xia et al.Protein & Cell

(Fig. 4A–C) and the Golgi protein GM130 (Fig. 4J–L). Incontrast to Cx31wt, although Cx31E183K protein was alsodetected in Golgi and ER, it was not detected at cell-cellcontacts (Fig. 4D–F and 4N–O). In cells expressing lowerlevel of Cx31E183K, most of the protein was seen in the Golgi(Fig. 2B). Unlike either Cx31wt or Cx31E183K, Cx31R180Xwas found primarily in the ER (Fig. 4G–I and 4P–R). Thus, theCx31R180X protein is largely restricted to the early secretorypathway, while the Cx31E183K protein is transported to theGolgi and Cx31wt protein is further transported to the latesecretory pathway and plasma membrane. The results

suggest that HI-associated Cx31 mutant proteins aredefective in intracellular trafficking. The abnormal traffickingof Cx31mutant proteins likely accounts for the lack of ability ofCx31 mutant proteins to form a functional gap junction.

Correlation between formation of Triton X-100 insolubleCx31 variants and their intracellular localization

Previous studies suggest that the assembled connexin-43(Cx43) gap junction is Triton X-100 insoluble (Musil andGoodenough, 1991). Triton X-100 insolubility of Cx43 is

Figure 1. Expression of C31 in mouse inner ear. Inner ears were dissected from P0 (A, B), P3 (C, D), P6 (E, F), P9 (G, H) andadult (I, J) mice and immunostained with an anti-Cx31 antibody. Representative images of inner ear (left panel, 100 ×) and organ ofcorti (right panel, 400 ×) are shown. Arrows indicate localization of Cx31 at tips of adult outer hearing cells (J). Differential expression

of Cx31 by Reissner’smembrane (RM), organ of corti (OC), spiral ganglion (SG), Hensen’s cells (HC), outer hair cells (OH) and innerhair cells (IH) at different stages of inner ears is shown.


Cx31 mutants induce ER stress Protein & Cell

correlated more with hemichannel interlocking or cell-cellchannel formation than with channel clustering or largeconnexin plaque formation (Wang et al., 1995). We thereforeanalyzed whether Cx31 variants formed Triton X-100insoluble hemichannels in the cell as does Cx43. As shownin Fig. 5A, Cx31wt and E183K were detected in both solubleand insoluble fractions with an apparent enrichment in theinsoluble fractions. However, the Cx31R180X protein waslargely present in the soluble fraction with only a small amountpresent in the insoluble fraction. Higher molecular-weightaggregates were also found in cells expressing Cx31wt and

Cx31E183K but not in cells expressing Cx31R180X (notshown).

To examine whether the inability of Cx31R180X to form aTriton X-100 insoluble complex is due to a lack of anintermolecular interaction that requires the C-terminus,Cx31R180XGFP and Cx31R180Xmyc proteins were co-expressed in cells and analyzed by co-immunoprecipitation.The myc-tagged Cx31R180X was co-immunoprecipitatedwith GFP-tagged Cx31R180X and vice versa (Fig. 5B). Thus,the intermolecular interaction between Cx31R180X mole-cules remains intact. Yet, Cx31R180X is incapable of forming

Figure 2. Gap junctional plaque assembly and lucifer yellow transfer of Cx31 variants. HeLa cells were transfected withplasmids encoding myc- or GFP-tagged Cx31wt (WT), Cx31E183K (E183K), or Cx31R180X (R180X). (A) Expression of Cx31variants. Cell lysates of expressing Cx31 variants tagged with either the myc tag (Cx31myc) or GFP (Cx31GFP) were detected with

an anti-myc antibody (upper panel) or an anti-GFP antibody (lower panel). Lysate of cells transfected with empty plasmid served asnegative control (C). (B) Localization of Cx31 variants. Cells expressing Cx31myc variants (upper panel) and Cx31GFP variants(lower panel) are shown. Cx31wt (WT) was detected at gap junctional plaques (white arrow). Cx31E183K and Cx31R180X wereseen at Golgi and ER-like structures, respectively. (C) Localization of Cx31 in stable transfected Hela cells. Cells stably transfected

with Cx31GFP variants (lower panel) are shown. Cx31wt (WT) was detected at gap junctional plaques (white arrow). Intracellularlocalization of Cx31E183K and Cx31R180X was found. (D) Lucifer yellow transfer of cells expressing Cx31 variants. HeLa cellsexpressing either GFP-tagged Cx31 variants were examined. One cell in each GFP illuminated cell cluster (indicated by *) was

injected with lucifer yellow. Note that lucifer yellow transfer was visualized in cells expressing Cx31wt but not in cells expressingCx31E183K or Cx31R180X.



Triton X-100 insoluble complexes. The lack of ability ofCx31R180X to form Triton X-100 insoluble complexes maycause a failure in Cx31R180X semi-channel formation.Furthermore, the retention of the Cx31R180X protein in theER indicates that wild-type Cx31 semi-channel interlockingoccurs in the late secretory pathway.

Induction of BiP expression by HI-associated Cx31mutants

It has been well established that unfolded and misfoldedprotein will induce cellular stress and abnormal trafficking(Ellgaard et al., 1999). Mutations in plasma membrane

Figure 3. Stability of Cx31 variants. HeLa cells expressing Cx31 variants, including Cx31wt (WT), Cx31E183K, and

Cx31R180X, were pulse labeled with 35S-methionine and chased with unlabeled methionine for periods indicated on the top of thefigure. A representative result of the pulse-chase experiments (A) and quantitation of Cx31 variant degradation in three independentexperiments (B) are shown. Note that Cx31wt (empty squares) and Cx31E183K (empty circles) have similar degradation kinetics

with a half-life of approximately 4 h while Cx31R180X (empty triangles) has a half-life of around 1 h.

Figure 4. Cellular localization of Cx31 variants. HeLa cells expressing myc-tagged Cx31wt, Cx31E183K, or Cx31R180X were

immunostained with an anti-myc antibody (A, D, G, J, M, P). ER and Golgi were labeled with an anti-calnexin antibody (B, E, H) andan anti-GM130 antibody (K, N, Q), respectively. Colocalization of Cx31 variants and the ER protein calnexin (yellow, C, F, I) wasobserved in cells expressing all three variants. Colocalization of Cx31 variants with the Golgi protein GM130 was detected in cells

expressing Cx31wt (J, K, L) and Cx31E183K (M, N, O), but not in cells expressing Cx31R180X (P, Q, R). Gap junctional plaques areonly detected in cells expressing Cx31wt (arrows. A, C, J, L).



proteins lead to either rapid degradation, aggregation, orinhibition of intracellular trafficking of the mutant proteins (Kimand Arvan, 1998; Aridor and Balch, 1999). To explore whetherCx31 mutant proteins are abnormal folded and induce ERstress, we analyzed the expression of the ER chaperone BiPin cells expressing Cx31 variants. Cells expressingCx31R108X and Cx31E183K showed increased level ofBiP detected by immunofluorescence. However, little BiPinduction was evidenced in cells expressing Cx31wt(Fig. 6A). Immunoprecipitation demonstrated a co-precipita-tion of Cx31 mutant proteins, but not Cx31wt, with BiP(Fig. 6B). Thus, there is an increased interaction between theER chaperone BiP and Cx31 mutant proteins. The resultssuggest that HI-associated Cx31 mutants induce ER stress inthe cell.

DISCUSSION

We have shown in this study that HI-associated Cx31 mutantproteins are defective in functional gap junction formation andintracellular trafficking that are likely a consequence of themutant proteins promoted unfolded protein response (UPR).These cellular abnormalities induced by mutant proteins maycontribute to pathogenesis of Cx31-assocaited hearingimpairment.

Expression of endogenous Cx31 protein in inner hair cellsand spiral ganglion neurons strongly supports a function ofCx31 in hearing. It is well recognized that these cells playcritical roles in transmitting hearing information into the brain.The observation of the Cx31 protein in the cuticular plates ofouter hair cells, where the cells do not have cell-cell contacts,

Figure 5. Triton X-100 solubility of Cx31 variants. (A)Cells were transfected with either a control plasmid (C) or

plasmids encoding myc-tagged Cx31 variants (WT, R180X,E183K). Triton X-100 insoluble proteins (IS, upper panel) andsoluble proteins (S, lower panel) were fractionated and

immunodetected with an anti-myc antibody. (B) Intermolecularinteraction of Cx31R180X protein. Lysates from cells co-expressing Cx31R180XGFP and Cx31R180Xmyc were immu-noprecipitated with either an anti-myc antibody (upper panel)

or an anti-GFP antibody (lower panel) followed by immunoblot-ting with an anti-GFP antibody (upper panel) or an anti-mycantibody (lower panel), respectively. As controls (C), cell

lysates were precipitated with unrelated mouse IgG (upperpanel) or rabbit IgG (lower panel).

Figure 6. Induction of BiP expression by R180X andE183K. (A) Induction of BiP expression by Cx31 mutants, but

not Cx31wt. Hela cells expressing myc-tagged Cx31wt (WT),Cx31R180X and Cx31E183K were immunostained with ananti-myc antibody (upper panel) and an anti-BiP antibody

(lower panel), respectively. Note that cells expressingCx31R180X and Cx31E183K but not Cx31wt were co-labeledwith anti-BiP antibody. (B) Co-precipitation of BiP with Cx31

mutant proteins, but not with Cx31wt. Cell lysates made fromcells transfected with a control plasmid (C), and myc-taggedCx31wt (WT), Cx31R180X or Cx31E183K were immunopre-cipitated with an anti-BiP antibody followed by immunoblotted

with an anti-myc antibody (upper panel). Co-precipitation ofBiP and Cx31R180X and Cx31E183K, but not Cx31wt (WT) isnoted. Expression of BiP (middle panle) and Cx31 variants

(lower panel) are shown. All corresponding molecules areindicated at the right side of the figure.



suggest that Cx31 potentially function as unpaired channelsin hair cells (Goodenough and Paul, 2003). A previous studyfailed to detect Cx31 transcripts in adult inner hair and outerhair cells using in situ hybridization (Lopez-Bigas et al.,2002a), may have been due to mRNA instability, low rates oftranscription while protein half-life is relatively long, ormethodological differences between the two studies.

Connexins function as intercellular channels at the cell-cellcontact via gap junction formation (Goodenough et al., 1996).The observation that two hearing loss-associated Cx31mutants fail to be assembled into cell surface semi-channelssuggests a loss of function of the mutants. This is consistentwith previous reports that HI-associated Cx26 mutants do nothave gap junction activity in Xenopus oocytes (White, 2000).The defect of gap junction formation by Cx31 diseasemutantsis likely caused by misfolding of the mutant proteins. Thesemisfolded mutant proteins are either rapid degraded orimpaired in trafficking. C-terminal deletion mutant Cx31d179is largely located at ER while substitution mutant Cx31E183Kprotein is detected in Golgi. The results indicate that thesorting of Cx31 mutant proteins is affected by diseasemutations. Another critical step for cell to make a functionalgap junction is the oligomerization of connexins. Cx31d179does not form Triton X-100 hexamers while Cx31E183K andCx31wt do. Since intramolecular interaction of Cx31d179 isnot visibly disrupted by the mutation, one possible explana-tion is that Cx31 oligomerization occurs in the late secretorypathway. Cx31d179 protein is largely located in ER whileCx31E183K and Cx31wt proteins are transported to Golgiand late secretary pathway. Different connexin may beoligomerized in different organelles (Kumar et al., 1995;VanSlyke et al., 2000; Das Sarma et al., 2001, 2002). Cx32 islikely oligomerized in ER while Cx46 is in Golgi or latersecretary pathway (VanSlyke et al., 2000; Das Sarma et al.,2001, 2002). Our results suggest that Cx31 forms hexamersin Golgi or in the secretary pathway after Golgi.

Finally, binding of unfolded proteins to ER chaperones is acommon observation of the primary quality control mechan-ism that promotes folding and assembly of these proteins.The interaction between ER chaperones and unfoldedproteins is also sufficient to cause trafficking alteration ofthese proteins (Ellgaard et al., 1999; Ellgaard and Helenius,2001). Our observation suggests that the disease-associatedCx31 mutations may compromise the normal folding andinduces the UPR, leading to abnormal trafficking and rapiddegradation. Consistent with this notion, Cx31R180X exhibitsits primary localization in the ER with a shorter half-life thanCx31wt. It is possible that folding impairment for Cx31E183Kis less severe and enable it to escape ER quality control withthe help of increased ER chaperones compared toCx31R180X. Nevertheless, the Cx31E183K protein may bedefective in the anterograde transport from the Golgi to thecell surface leading to accumulation in the Golgi, thereforepreventing it from reaching the cell surface in sufficientquantity to form a functional gap junction.


Cell lines, antibodies, and constructs

Cell lines were purchased from ATCC and cultured under recom-mended conditions. Antibodies specific for green fluorescent protein(GFP), myc epitope (9E10.2), calnexin, GM-130 were purchased from

Clontech, ATCC, Stressgen and Bioscience, respectively. A poly-clonal antibody specific for Cx31was generated using a peptidelocated at amino acid 101–119 (ERRHRQKHGDQCAKLYDNAG)

(Abgent). Lucifer yellow was from Molecular Probes. Other reagentswere from Sigma. All constructs were made by PCR amplification andverified by sequencing. For amplification, genomic DNA from a

normal individual and from individuals harboring the correspondingmutations was used as the template (Xia et al., 1998). Primers(forward: 5 ′-CGGAATTCTGGGCGCCATGGACTGGAAGA-CACTCCA-3 ′ ; reverse: 5 ′-GCGTCGACTGGATGGGGGT-

CAGGTTGGG-3′) were used for Cx31wt and Cx31E183K. Primers(forward: 5 ′-CGGAATTCTGGGCGCCATGGACTGGAAGA-CACTCCA-3′; reverse: 5′-CCAAGCTTGGGGCAATGTAGCAGTC-

CACG-3′) were used for Cx31R180X. DNAs encoding Cx31variants were cloned into both pEGFP/N1 (Clontech) andpcDNA3.1(−)MycHisB (Invitrogen) to generate GFP-tagged or myc-

tagged Cx31 variants, respectively.

Transfection and immuno-assay

Transfection was performed with Lipofectamine 2000 reagent

(Invitrogen) according to the manufacturer’s instruction. Stableexpressors were selected using G418 followed by ring cloning.Immunofluorescent staining was performed essentially as previouslydescribed and analyzed under confocal microscopy (BioRad) (Zhang

et al., 1993).Immunoprecipitation and immunoblotting were done as described

(Zhang et al., 1998). Briefly, transfected cells were lysed either in

0.7% NP-40 buffer (10mM Hepes, pH7.5, 142.4mM KCl, 5mMMgCl2, 1mM EGTA, and 0.7% NP-40) or RIPA buffer (Zhang et al.,1998) on ice. Insoluble cellular debris was removed by centrifugation

at 14,000g at 4°C for 30min. Corresponding antibodies (3 µg) andprotein G beads (Roche, 25µL) were added to cell lysates andincubated at 4°C overnight on a nutator. Protein-antibody-beadcomplexes were washed with corresponding buffer. The proteins

were separated on 4%–20% Tris-glycine gels (Invitrogen), electro-transferred onto PVDF membranes (Millipore), immunoblotted withindicated antibodies and detected by ECL (Amersham).

Immunohistology of the inner ear: Postnatal day (P) 0, 3, 6, 9, andadult mice were fixed via transcardiac perfusion with 4% paraformal-dehyde in 0.1M phosphate buffer. Adult mouse heads including inner

ears were extirpated and decalcified in 8% EDTA in 4% paraformal-dehyde for 10 days, immersed overnight in 0.1M phosphate buffercontaining 30% sucrose at 4°C and frozen in OCT. Serial sections

(12mm) were generated on a cryostat and mounted on slides.Masked epitopes were retrieved using 10mM sodium citrate bufferheated to 95°C for 7 min followed by incubation at room temperaturefor 20 min, in 0.3% H2O2 for 5 min, and 1% Triton-100/PBS for 20min.

After blocking with 5% BSA/5% horse serum in PBS for 1 h, sectionswere incubated in primary antibody (1:250) at 4°C overnight, washed3 times with PBS, further incubated in the second antibody (1:500) for

3 h, then developed with DAB solution (Roche).



Metabolic labeling

Cells at 90% confluence were rinsed with methionine-deficient DMEM

containing 10% dialyzed fetal bovine serum (FBS) followed bylabeling with the same medium containing 35S-methionine (150µCi/mL) for 20min. The cells were then rinsed and chased with normal

culture medium containing 450µg/mL L-methionine. Cx31 variantswere immunoprecipitated and analyzed with a phosphorimager.

Triton X-100 solubility analysis of Cx31 variants

Cells were rinsed once with PBS and incubated on ice for 30 min inPBS containing 1% Triton X-100 and protease inhibitor cocktails(Roche). The cells were collected by scraping and further incubated

at 4°C for 1 h. The samples were then centrifuged at 100,000 g for30 min to separate soluble and insoluble fractions. Both fractionswere lysed in SDS sample buffer (0.5M TrisHCl, pH6.8, 20%

glycerol, 4% SDS). Equal amounts of protein from each sample wereanalyzed by immunoblotting.

Dye transfer assay

HeLa cells were transfected with GFP-tagged Cx31 variants. Luciferyellow (4%) was injected into one cell in a cluster of GFP-positive cellsusing a loose patch clamp whole cell recording technique. Dye

transfer was assessed 20 min after injection using fluorescencemicroscopy (Nikon). Finally, cells were fixed with 3.7% paraformalde-hyde. Images were captured using confocal microscopy.

ACKNOWLEDGEMENTS

This work was supported by grants from the Chinese NationalScience Foundation (ZZ, KX, DLW) and the NIH/NIDCD (ZZ).

ABBREVIATIONS

Cx26, connexin-26; Cx30, connexin-30; Cx30.3, connexin 30.3;

Cx31, connexin-31; Cx31wt, wild type Cx31 protein; FBS, fetalbovine serum; GFP, green fluorescent protein; HI, hearing impairment

REFERENCES

Aridor, M., and Balch, W.E. (1999). Integration of endoplasmicreticulum signaling in health and disease. Nat Med 5, 745–751.

Bone, L.J., Deschenes, S.M., Balice-Gordon, R.J., Fischbeck, K.H.,

and Scherer, S.S. (1997). Connexin32 and X-linked Charcot-Marie-Tooth disease. Neurobiol Dis 4, 221–230.

Bruzzone, R., Gomes, D., Denoyelle, E., Duval, N., Perea, J.,

Veronesi, V., Weil, D., Petit, C., Gabellec, M.M., D'Andrea, P., et al.(2001). Functional analysis of a dominant mutation of humanconnexin26 associated with nonsyndromic deafness. Cell Com-

mun Adhes 8, 425–431.

Bruzzone, R., Veronesi, V., Gomes, D., Bicego, M., Duval, N., Marlin,S., Petit, C., D'Andrea, P., and White, T.W. (2003). Loss-of-function

and residual channel activity of connexin26 mutations associatedwith non-syndromic deafness. FEBS Lett 533, 79–88.

Cohen-Salmon, M., Ott, T., Michel, V., Hardelin, J.P., Perfettini, I.,Eybalin, M., Wu, T., Marcus, D.C., Wangemann, P., Willecke,K., et al. (2002). Targeted ablation of connexin26 in the inner ear

epithelial gap junction network causes hearing impairment and celldeath. Curr Biol 12, 1106–1111.

Dahl, E., Manthey, D., Chen, Y., Schwarz, H.J., Chang, Y.S., Lalley, P.A., Nicholson, B.J., and Willecke, K. (1996). Molecular cloning andfunctional expression of mouse connexin-30,a gap junction genehighly expressed in adult brain and skin. J Biol Chem 271,

17903–17910.Das Sarma, J., Meyer, R.A., Wang, F., Abraham, V., Lo, C.W., and

Koval, M. (2001). Multimeric connexin interactions prior to thetrans-Golgi network. J Cell Sci 114, 4013–4024.

Das Sarma, J., Wang, F., and Koval, M. (2002). Targeted gap junction

protein constructs reveal connexin-specific differences in oligo-merization. J Biol Chem 277, 20911–20918.

Denoyelle, F., Weil, D., Maw, M.A., Wilcox, S.A., Lench, N.J., Allen-

Powell, D.R., Osborn, A.H., Dahl, H.H., Middleton, A., Houseman,M.J., et al. (1997). Prelingual deafness: high prevalence of a30delG mutation in the connexin 26 gene. Hum Mol Genet 6,

2173–2177.

Di, W.L., Monypenny, J., Common, J.E., Kennedy, C.T., Holland, K.A.,Leigh, I.M., Rugg, E.L., Zicha, D., and Kelsell, D.P. (2002).

Defective trafficking and cell death is characteristic of skindisease-associated connexin 31 mutations. Hum Mol Genet 11,2005–2014.

Elfgang, C., Eckert, R., Lichtenberg-Frate, H., Butterweck, A., Traub,O., Klein, R.A., Hulser, D.F., and Willecke, K. (1995). Specificpermeability and selective formation of gap junction channels in

connexin-transfected HeLa cells. J Cell Biol 129, 805–817.

Ellgaard, L., and Helenius, A. (2001). ER quality control: towards anunderstanding at the molecular level. Curr Opin Cell Biol 13,

431–437.

Ellgaard, L., Molinari, M., and Helenius, A. (1999). Setting the

standards: quality control in the secretory pathway. Science 286,1882–1888.

Goodenough, D.A., Goliger, J.A., and Paul, D.L. (1996). Connexins,

connexons, and intercellular communication. Annu Rev Biochem65, 475–502.

Goodenough, D.A., and Paul, D.L. (2003). Beyond the gap: functions

of unpaired connexon channels. Nat Rev Mol Cell Biol 4, 285–294.

Grifa, A., Wagner, C.A., D'Ambrosio, L., Melchionda, S., Bernardi, F.,Lopez-Bigas, N., Rabionet, R., Arbones, M., Monica, M.D., Estivill,

X., et al. (1999). Mutations in GJB6 cause nonsyndromicautosomal dominant deafness at DFNA3 locus. Nat Genet 23,16–18.

Kelsell, D.P., Di, W.L., and Houseman, M.J. (2001a). Connexinmutations in skin disease and hearing loss. Am J Hum Genet 68,

559–568.

Kelsell, D.P., Dunlop, J., and Hodgins, M.B. (2001b). Humandiseases: clues to cracking the connexin code? Trends Cell Biol

11, 2–6.

Kelsell, D.P., Dunlop, J., Stevens, H.P., Lench, N.J., Liang, J.N.,Parry, G., Mueller, R.F., and Leigh, I.M. (1997). Connexin 26

mutations in hereditary non-syndromic sensorineural deafness.Nature 387, 80–83.

Kim, P.S., and Arvan, P. (1998). Endocrinopathies in the family of

endoplasmic reticulum (ER) storage diseases: disorders of proteintrafficking and the role of ER molecular chaperones. Endocr Rev19, 173–202.

Kumar, N.M., Friend, D.S., and Gilula, N.B. (1995). Synthesis andassembly of human beta 1 gap junctions in BHK cells by DNA



transfection with the human beta 1 cDNA. J Cell Sci 108 (Pt 12),3725–3734.

Lopez-Bigas, N., Arbones, M.L., Estivill, X., and Simonneau, L.(2002a). Expression profiles of the connexin genes, Gjb1 andGjb3, in the developing mouse cochlea. Mech Dev 119 Suppl 1,S111–115.

Lopez-Bigas, N., Melchionda, S., Gasparini, P., Borragan, A.,Arbones, M.L., and Estivill, X. (2002b). A common frameshift

mutation and other variants in GJB4 (connexin 30.3): Analysis ofhearing impairment families. Hum Mutat 19, 458.

Lopez-Bigas, N., Olive, M., Rabionet, R., Ben-David, O., Martinez-

Matos, J.A., Bravo, O., Banchs, I., Volpini, V., Gasparini, P.,Avraham, K.B., et al. (2001). Connexin 31 (GJB3) is expressed inthe peripheral and auditory nerves and causes neuropathy and

hearing impairment. Hum Mol Genet 10, 947–952.

Musil, L.S., and Goodenough, D.A. (1991). Biochemical analysis ofconnexin43 intracellular transport, phosphorylation, and assembly

into gap junctional plaques. J Cell Biol 115, 1357–1374.

Petit, C., Levilliers, J., and Hardelin, J.P. (2001). Molecular genetics ofhearing loss. Annu Rev Genet 35, 589–646.

Plum, A., Winterhager, E., Pesch, J., Lautermann, J., Hallas, G.,Rosentreter, B., Traub, O., Herberhold, C., andWillecke, K. (2001).

Connexin31-deficiency in mice causes transient placental dys-morphogenesis but does not impair hearing and skin differentia-tion. Dev Biol 231, 334–347.

Reuss, B., Hellmann, P., Dahl, E., Traub, O., Butterweck, A.,Grummer, R., and Winterhager, E. (1996). Connexins and E-cadherin are differentially expressed during trophoblast invasion

and placenta differentiation in the rat. Dev Dyn 205, 172–182.

Richard, G., Smith, L.E., Bailey, R.A., Itin, P., Hohl, D., Epstein, E.H.,Jr., DiGiovanna, J.J., Compton, J.G., and Bale, S.J. (1998).

Mutations in the human connexin gene GJB3 cause erythroker-atodermia variabilis. Nat Genet 20, 366–369.

Simon, A.M., and Goodenough, D.A. (1998). Diverse functions of

vertebrate gap junctions. Trends Cell Biol 8, 477–483.

VanSlyke, J.K., Deschenes, S.M., and Musil, L.S. (2000). Intracellular

transport, assembly, and degradation of wild-type and disease-linked mutant gap junction proteins. Mol Biol Cell 11, 1933–1946.

Wang, Y., Mehta, P.P., and Rose, B. (1995). Inhibition of glycosylation

induces formation of open connexin-43 cell-to-cell channels andphosphorylation and triton X-100 insolubility of connexin-43. J BiolChem 270, 26581–26585.

Watts, G.D., and Chance, P.F. (2002). Molecular basis of hereditaryneuropathies. Adv Neurol 88, 133–146.

White, T.W. (2000). Functional analysis of human Cx26 mutationsassociated with deafness. Brain Res Brain Res Rev 32, 181–183.

White, T.W., Deans, M.R., Kelsell, D.P., and Paul, D.L. (1998).

Connexin mutations in deafness. Nature 394, 630–631.

Xia, J.H., Liu, C.Y., Tang, B.S., Pan, Q., Huang, L., Dai, H.P., Zhang,B.R., Xie, W., Hu, D.X., Zheng, D., et al. (1998). Mutations in the

gene encoding gap junction protein beta-3 associated withautosomal dominant hearing impairment. Nat Genet 20,370–373.

Zelante, L., Gasparini, P., Estivill, X., Melchionda, S., D'Agruma, L.,Govea, N., Mila, M., Monica, M.D., Lutfi, J., Shohat, M., et al.(1997). Connexin26 mutations associated with the most common

form of non-syndromic neurosensory autosomal recessive deaf-ness (DFNB1) in Mediterraneans. Hum Mol Genet 6, 1605–1609.

Zhang, Z., Hartmann, H., Do, V.M., Abramowski, D., Sturchler-Pierrat,C., Staufenbiel, M., Sommer, B., van de Wetering, M., Clevers, H.,Saftig, P., et al. (1998). Destabilization of beta-catenin by mutationsin presenilin-1 potentiates neuronal apoptosis. Nature 395,

698–702.

Zhang, Z., Morla, A.O., Vuori, K., Bauer, J.S., Juliano, R.L., and

Ruoslahti, E. (1993). The alpha v beta 1 integrin functions as afibronectin receptor but does not support fibronectin matrixassembly and cell migration on fibronectin. J Cell Biol 122,235–242.



RESEARCH ARTICLE

Interaction of Hsp40 with influenza virus M2protein: implications for PKR signalingpathway

Zhenhong Guan1,2, Di Liu2, Shuofu Mi2, Jie Zhang2, Qinong Ye3, Ming Wang1, George F. Gao1,2,4,Jinghua Yan2✉

1 College of Veterinary Medicine, China Agricultural University, Beijing 100094, China2 CAS Key Laboratory of Pathogenic Microbiology and Immunology, Institute of Microbiology, Chinese Academy of Sciences,Beijing 100101, China

3 Beijing Institute of Biotechnology, Beijing 100850, China4 Beijing Institutes of Life Science, Chinese Academy of Sciences, Beijing 100101, China✉ Correspondence: [email protected] September 25, 2010 Accepted October 9, 2010

ABSTRACT

Influenza virus contains three integral membrane pro-teins: haemagglutinin, neuraminidase, andmatrix protein(M1 and M2). Among them, M2 protein functions as an ionchannel, important for virus uncoating in endosomes ofvirus-infected cells and essential for virus replication. Inan effort to explore potential new functions of M2 in thevirus life cycle, we used yeast two-hybrid system tosearch for M2-associated cellular proteins. One of thepositive clones was identified as human Hsp40/Hdj1, aDnaJ/Hsp40 family protein. Here, we report that both BM2(M2 of influenza B virus) and A/M2 (M2 of influenza Avirus) interacted with Hsp40 in vitro and in vivo. Theregion of M2-Hsp40 interaction has been mapped to theCTD1 domain of Hsp40. Hsp40 has been reported to be aregulator of PKR signaling pathway by interacting withp58IPK that is a cellular inhibitor of PKR. PKR is a crucialcomponent of the host defense response against virusinfection. We therefore attempted to understand therelationship among M2, Hsp40 and p58IPK by furtherexperimentation. The results demonstrated that both A/M2 and BM2 are able to bind to p58IPKin vitro and in vivoand enhance PKR autophosphorylation probably viaforming a stable complex with Hsp40 and P58IPK, andconsequently induce cell death. These results suggestthat influenza virus M2 protein is involved in p58IPK-mediated PKR regulation during influenza virus infection,therefore affecting infected-cell life cycle and virusreplication.

KEYWORDS M2 protein of influenza virus, Hsp40,P58IPK, protein interaction, PKR signal pathway

INTRODUCTION

Influenza virus is an important human and zoonotic pathogen,prevalent throughout the world for centuries. The twopredominant types of influenza viruses that infect humansare influenza A and B viruses. Both influenza A and influenzaB viruses in the family Orthomyxoviridae have negative-stranded RNA genomes consisting of eight RNA segments.RNA segments 1–3 encode the three polymerase proteins,PB1, PB2, and PA; RNA segment 4 encodes haemagglutinin(HA); RNA segment 5 encodes nucleoprotein (NP); RNAsegment 6 encodes neuraminidase (NA); RNA segment 7encodes two matrix proteins: M1 and M2; RNA segment 8encodes two non-structural proteins NS1 and NS2 (nuclearexport protein/NEP).

The M2 protein of influenza A virus (A/M2) is translatedfrom a spliced mRNA, containing 97 amino acid residues inlength (Lamb et al., 1981). A/M2 has an ion channel activity topermit protons to enter virions and cause RNP-M1 dissocia-tion during uncoating of virions in endosomes (Sugrue andHay, 1991; Pinto et al., 1992; Takeda et al., 2002). A/M2channel activity was required for the activation of inflamma-somes by influenza and was sufficient to activate inflamma-somes in primed macrophages and dendritic cells (Ichinoheet al., 2010). A/M2 also functions to equilibrate the pHgradient between the lumen of the trans-Golgi network (TGN)and the cytoplasm to prevent HA from adopting a low-pH-


Protein Cell 2010, 1(10): 944–955DOI 10.1007/s13238-010-0115-x

Protein & Cell

induced conformation in the Golgi apparatus (Ciampor et al.,1992; Shimbo et al., 1996). Cytoplasmic tail of the A/M2 playsa vital role in infectious virus production by coordinating theefficient packaging of genome segments into influenza virusparticles (McCown and Pekosz, 2006). Unlike the A/M2protein, the M2 protein of influenza B virus (BM2) is translatedfrom a bicistronic mRNA derived from RNA segment 7(Briedis et al., 1982; Horvath et al., 1990), containing 109amino acid residues, with a small N-terminal ectodomain(1– 7 aa), a single transmembrane domain (8–27 aa) and a C-terminal cytoplasmic tail (28–109 aa) (Pinto and Lamb,2006). In addition to its obvious ion channel activity (Mouldet al., 2003), BM2 protein has recently been found to beessential for influenza B virus replication as shown by theexperimental data of reverse genetics technology (Hatta etal., 2004).

There is increasing evidence suggesting that the protein-protein interactions between virus and host play an importantrole in the life cycle and pathogenicity of viruses (Liu et al.,2009). But so far no interacting host partners of influenza virusM2 protein have been identified, except for our recent findingsof the host ATPase β1 unit (Mi et al., 2010). In this study, ayeast two-hybrid system was used to screen a human kidneycDNA library so as to identify host proteins that interact withBM2. One clone encoding heat shock protein Hsp40 (Hdj1/DnajB1/DjB1), a DnaJ/Hsp40 family protein, was identified.We further presented our experimental data to show that bothA/M2 and BM2 interacted with Hsp40/Hdj1 in vitro and in vivo.It has been reported that Hsp40 associates with P58IPK, acellular inhibitor of PKR, an interferon-induced double-stranded RNA activated serine-threonine kinase (Melville etal., 1997), which is a novel “CIHD” member of the host innatedefense response against pathogenic virus. Infection ofP58IPK knockout mice with influenza virus resulted inincreased lung pathology, immune cell apoptosis, PKRactivation, and mortality (Goodman et al., 2009). It has beenalso previously reported that Influenza virus infection pro-motes the disruption of the Hsp40-P58IPK complex and theactivation of P58IPK (Katze et al., 1988; Lee et al., 1994; Leeand Katze, 1994). The released P58IPK is then capable ofinteracting with PKR by binding to amino acids 244–296,which prevents dimerization and activation of PKR (Tan et al.,1998). Our results also show that M2 proteins could bind toP58IPK and be able to promote PKR autophosphorylation andactivation in vitro and in vivo. Therefore, it is probable that M2proteins form a stable complex with Hsp40-P58IPK and hinderthe disassociation of Hsp40-P58IPK. As M2 is synthesizedin the late stage of virus infection (Odagiri et al., 1999),together with the early-synthesized NS1 to inactivate the PKRactivity (Bergmann et al., 2000), influenza virus regulates/interferes cell functions to facilitate its replication through itsprotein products (NS1 and M2) interacting with PKR-relatedproteins.

RESULTS

Identification of a BM2-interacting protein in the yeasttwo-hybrid system

Since influenza virus is able to infect several tissues in humanbeings, such as lung, liver, brain etc, a prey library of humankidney cDNA was used in a yeast two-hybrid screen toidentify proteins that interact with BM2, with the portion of thecytoplasmic domain of BM2 (BM2C) as bait. One positiveclone containing cDNA with entire open reading frameencoding Hsp40/Hdj1 was isolated from a library of approxi-mately 1 ×106 independent clones. To confirm the observedinteraction between BM2 and Hsp40 in the yeast, we set up agrowth experiment on SD plate lacking adenine, tryptophan,leucine and histidine in the two hybrid system. In this assay,growth on SD medium is supported only when the two hybridproteins interact and induce transcription from the his reportergene. We found that two yeast clones co-expressing AD-Hsp40/BD-BM2 and AD-p53/BD-T antigen (positive control)constructs grew on this medium (Fig. 1A). We also carried outliquid β-galactosidase assays by co-transforming the recom-binant plasmids to yeast strain SFY526. As shown in Fig. 1B,co-transformation of AD-Hsp40 and BD-BM2C resulted in astrong induction of β-galactosidase activity, significantlyhigher than that of the respective control co-transformations.These results collectively indicate that BM2 and Hsp40interact with each other.

Interaction between BM2 and Hsp40 in vitro and in vivo

To verify and extend the binding data obtained in yeast two-hybrid assay, we performed GST pull-down experiments.Bacterially expressed GST-BM2C or GST bound to glu-tathione-Sepharose beads was allowed to react with in vitrotranslated [35S]methionine-labeled Hsp40 in reaction buffer.Consistent with the yeast two-hybrid results, Hsp40 specifi-cally bound to GST-BM2C, but not GST (Fig. 2A).

To further assess the interaction between BM2 and Hsp40in vivo, the plasmids that express Flag-Hsp40 or GFP-BM2were co-transfected into 293T cells. The cell lysates werethen immunoprecipitated with the anti-Flag M2-conjugatedagarose and subsequently immunobloted with anti-GFPantibody. Consistent with the GST pull-down and yeast two-hybrid results, BM2 specifically interacted with Hsp40 (Fig.2B). A reciprocal co-immunoprecipitation experiment alsoshowed the physical interaction between BM2 and Hsp40(Fig. 2C).

Mapping the BM2 binding domain of Hsp40

To determine the region of Hsp40 that is responsible for itsinteraction with BM2, a series of Hsp40 deletion mutants were


Interaction of Hsp40 and influenza virus M2 protein Protein & Cell

constructed to test for their ability to interact with BM2 in GSTpull-down assays (Fig. 3A). Full-length Hsp40 and its deletionmutants were translated, [35S]-methionine labeled in vitro,and then incubated with GST-BM2C or GSTalone. As shownin Fig. 3B, full-length Hsp40, Hsp40 (1–246 aa) including J

Protein & Cell

Figure 2. GST pull-down and co-immunoprecipitation

(IP) experiments. (A) In vitro translated 35S-labeled Hsp40was incubated with GST or GST-BM2C fusion proteinimmobilized onto glutathione-Sepharose beads. Binding was

viewed on SDS-PAGE by autoradiography. (B) 293Tcells werecotransfected with the expression vectors for Flag-taggedHsp40 and GFP-tagged BM2 or GFP as indicated. Lysatesfrom the transfected cells were immunoprecipitated using anti-

Flag M2-agarose, and the immunoprecipitates were probed(IB) with an anti-GFP antibody (Santa Cruz). Controls of proteininput and relevant IB antibodies (anti-GFP and anti-Flag) were

shown in two lower panels. (C) The reciprocal co-IP assay forBM2 (flag tagged) and Hsp40 (GFP tagged)as indicated .

Figure 1. Yeast two-hybrid analysis. (A) Yeast strain

SYF526 transformed with the indicated expression constructswere grown on an SD plate lacking adenine, tryptophan,leucine and histidine. Growth of the yeast on this plate is

indication of interaction of the two expressed proteins. BD,pGBKT7 vector expressing GAL4 DNA binding domain; BD-x(e.g., BD-BM2C), the in-frame cloned plasmid of the relevantgene in pGBKT7 (pGBKT7-BM2C). AD, pACT2 vector

expressing the transcription activation domain; AD-x (e.g.,AD-Hsp40), the in-frame cloned plasmid of the relevant gene inpACT2 (pACT2-Hsp40). BD-T antigen, BD-LaminC and AD-

p53 are control vectors supplied by Clontech. Interaction ofBD-T antigen with AD-p53 is used as a positive controlwhereas BD-LaminC and AD-p53 is used as a negative

control. (B) Quantification of β-galactosidase activity. Datashown are the means of three separate experiments and errorbars are standard deviations. Co-transfection pair plasmidcombinations are as follows: 1, pACT2-Hsp40 and pGBKT7-

BM2C; 2, pACT2-Hsp40 and pGBKT7; 3, pACT2-Hsp40 andpAS2-LaminC; 4, pACT2 and pGBKT7; 5, pACT2 andpGBKT7-BM2C; 6, pGBKT7-T antigen and pACT2-p53 (posi-

tive control); 7, pGBKT7-LaminC and pACT2-p53 (negativecontrol).


Zhenhong Guan et al.

domain, G/F domain and CTD1, Hsp40 (162–340 aa)including CTD1, CTD2 and DD, and Hsp40 (162–246 aa)only including CTD1 were able to interact with BM2, but notHsp40 (1–162 aa) including J domain and G/F domain. Asnegative control, the full-length and deletion mutants ofHsp40 did not bind to GST alone. These results indicatedthat Hsp40 interacted with BM2 through its CTD1.

Interaction between A/M2 and Hsp40 in vitro and in vivo

A/M2 protein of influenza A virus is structurally andbiochemically similar to BM2, so we examined the possibilityof A/M2 binding to Hsp40 in GST pull-down and co-IP assays.Purified GST and GST-Hsp40 immobilized on glutathione-Sepharose beads were used to pull down in vitro translatedand [35S]-methionine labeled A/M2. The results indicated thatHsp40 also specifically bound to A/M2 in vitro, but not GST(Fig. 4A). To examine if A/M2 can interact with Hsp40 in vivo,293Tcells were transfected with the vectors expressing Flag-Hsp40 or GFP-A/M2. The cell lysates were then immunopre-cipitated with the anti-Flag M2-agarose and subsequentlyimmunoblotted with anti-GFP antibody. As shown in Fig. 4B,like BM2, A/M2 also bound to Hsp40 physically in 293Tcells.

Figure 3. Definition of BM2 binding domain of Hsp40. (A) Structural representation of domain structures of wild-type Hsp40

and deletion constructs as indicated (ΔHsp40). The subdomains of Hsp40 are labeled as follows: JD, J-domain; G/F, glycine/phenylalanine rich domain; CTD1 and CTD2, carboxyl-terminal domain 1 and 2; DD, a predicted dimerization domain. Positions ofterminal amino acids are indicated. (B) GST pull-down assays were performed using 35S-labeled Hsp40 or ΔHsp40 and GST-BM2C

fusion protein. GST protein was used as a control.

Figure 4. GST pull-down (A) and co-immunoprecipitation(IP) (B) experiments between A/M2 and Hsp40. Experimentswere carried out as indicated in Fig. 2 except 35S-labeled A/M2

or the relevant A/M2 constructs were used.



Association of M2 protein with P58IPK and influence onthe binding of Hsp40 to P58IPK by M2 proteins

Since it has been reported that Hsp40 interacts withP58IPKand influenza virus infection functionally activates theP58IPK pathway by promoting the disassociation of Hsp40from P58IPK, we next investigated whether BM2 is a factorthat results in their disassociation. Purified soluble GST-BM2C protein incubated with GST-P58IPK immobilized ontoglutathione-Sepharose beads, and then in vitro translatedand [35S]-methionine labeled Hsp40 was added to theincubation buffer to detect the effect of BM2 on the interactionof P58IPK and Hsp40. Interestingly, BM2 does not block theassociation between P58IPK and Hsp40 as we expected(Fig. 5). Using GST pull-down and co-IP approaches, wefound that A/M2 and BM2 were able to bind to P58IPK, Invirus-infected 293T cells, overexpressed Flag-tagged P58IPK

was co-immunoprecipitated with M2 protein (Fig. 6). Theseresults suggest that M2 protein possibly forms a stablecomplex with Hsp40 and P58IPK and inhibits P58IPK activity.

Enhancement of PKR autophosphorylation in vitro and invivo by M2 proteins

P58IPK is a negative regulator of PKR, the repression of itsactivity may result in increase in autophosphorylation of PKRand subsequent enhancement of phosphorylation of the αsubunit of eukaryotic initiation factor 2 (eIF2α). Therefore, toexamine the functional consequences of the interactiondescribed above, we first performed an in vitro kinaseassay using purified Flag-PKR immobilized onto anti-FlagM2-conjugated agarose in the presence of GST-P58IPK aloneor both GST-BM2C and GST-P58IPK as described under

“Materials and Methods.” As shown in Fig. 7A, the autopho-sphorylation of PKR was blocked by P58IPK in vitro, and thenwas reversed by BM2 protein when its activator dsRNA wasadded.

To determine whether M2 protein has an effect on PKRautophosphorylation in mammalian cells, 293Tcells were co-transfected with expression vectors for Flag-P58IPK andFlag-A/M2 or Flag-BM2 constructs at 1:1 ratio and treatedwith IFN-α and poly(I:C) as described in “Materials andMethods.” As a control, cells were transfected with emptyvector alone. Cell extracts were then prepared and analyzedby SDS-PAGE and Western blotting with phospho-PKRantibody. As shown in Fig. 7B, overexpression of BM2 or A/M2 in 293T cells increased the autophosphorylation of PRKdue to the inhibition of P58IPK activity via forming a complexwith M2 protein.

Induction of death in HeLa cells with M2 proteins

To examine whether death was initiated in the cellstransfected with pCAGGS-AM2/BM2, PI staining and flowcytometry were used to analyze the percentages of death inthe total cell population. Dead cells have a weaker fluores-cence (M1 zone) (Fig. 8A). Ratio of M1 and M1 +M2represents the percentages of dead cells (Fig. 8B). Theresults showed that HeLa cells transfected with pCAGGS-AM2/BM2, exhibited cell death in a time-dependent manner(Fig. 8). The maximal cell death occurred at 72 h posttrans-fection, the dead cells increased to 13.04% and 15.78%respectively, in cells transfected with pCAGGS-AM2/BM2 Incontrast, only 5.75% dead cells were observed in cellstransfected with an empty vector.

DISCUSSION

Hsp40 family is involved in numerous cellular functions,including regulation of protein folding, translocation andassembly by cooperating with Hsp70 (Cheetham and Caplan,1998; Ohtsuka and Hata, 2000). However, it has beenrecently reported that Hsp40 interacts with HBV core proteinand inhibits viral replication (Sohn et al., 2006). Here wereport for the first time that BM2 interacts with Hsp40, whichhas been identified by yeast two-hybrid screening and furtherconfirmed by GST pull-down and immunoprecipitation experi-ments. Hsp40/Hdj1 belongs to the Type II Hsp40s, containinga J-domain, a G/F-rich domain, two conserved carboxyl-terminal domains (CTD1 and CTD2) and a predicteddimerization domain (Mohler et al., 2004). Our resultsindicated that BM2 interacted with Hsp40 through CTD1domain. Furthermore, the data that Hsp40 was pulled downand immunoprecipitated by A/M2 indicated that the interac-tion of M2 protein with Hsp40 is a common feature ofinfluenza A and B viruses.

PKR is a key component in the establishment of the

Protein & Cell

Figure 5. BM2 protein enhances binding of Hsp40 toP58IPK. GST-P58IPK immobilized onto glutathione-Sepharosebeads was incubated with increasing amounts of soluble GST-BM2C protein for 2 h at 4°C, and then Hsp40 labeled with [35S]-

methionine was added to incubation buffer for 3 h. Afterwashing, the protein complexes were dissociated from thebeads and subjected to SDS-PAGE followed by autoradio-

graphy. 35S-Hsp40 signals were increasing with more GST-BM2C added. Note, left lane is 35S-Hsp40 input.



interferon-mediated cellular antiviral and antiproliferativeresponses (Gale et al., 1996). By binding to dsRNA, PKRundergoes a conformational change and becomes autopho-sphorylated at multiple serine and threonine sites, and theactivated PKR then phosphorylates specific substrate eIF2αat Ser51, leading to an inhibition in protein synthesis and ablock in viral replication (Meurs et al., 1992; Srivastava et al.,1998). Many viruses have evolved elaborate mechanisms toevade the host defense, such as production of multifunctionalproteins binding to dsRNA or direct interaction with PKR(Gale and Katze, 1998). Influenza A virus has developed twostrategies to block the activation of PKR. First, it encodes anon-structural protein (NS1) that can bind to dsRNA toprevent PKR autophosphorylation (Lu et al., 1995); Secondly,the infection of influenza A virus activates P58IPK, a cellularinhibitor of PKR, which can prevent dimerization andactivation of PKR through directly binding to PKR. P58IPK

was originally characterized as an influenza virus-activatedprotein. Hsp40 was shown to normally bind to and negativelyregulate P58IPK. The disruption of the Hsp40-P58IPK complexwas found during influenza virus infection. However, what

causes the dissociation of Hsp40 and P58IPK remains to bedetermined. Therefore, we speculate that the interaction ofinfluenza virus M2 protein with Hsp40 possibly results in thedissociation of Hsp40 and P58IPK and activation of P58IPK.Unexpectedly, our results from GST pull-down assay showedthat purified BM2 protein could not block binding of Hsp40 toP58IPK. The further observation that both BM2 and A/M2 canbind to P58IPK suggests that M2 protein, Hsp40 and P58IPK

probably form a stable complex in virus-infected cells. Thefindings that the level of PKR autophosphorylation wasenhanced by both A/M2 and BM2 in vitro and in vivo indicatedthat P58IPK was arrested and inactivated as a result ofinteraction with Hsp40 and M2 protein.

Studies using the herpes simplex virus translocatingprotein VP22 to carry influenza virus proteins into cells haveshown that expression of M2 protein induces Hela cellapoptosis (Morris et al., 2002). It is reported that A/M2 ishighly toxic for mammalian cells, yeast and insect cells(Ilyinskii et al., 2007, 2008). Similar observations have beenmade in our experiments. Taken together, our resultsdemonstrate that, during influenza virus infection, M2 protein

Figure 6. GST pull-down and co-immunoprecipitation (IP) to show direct interaction of P58IPK andM2. (A) In vitro translatedBM2 or A/M2 was incubated with GST-P58IPK fusion protein immobilized onto glutathione-Sepharose beads. GST protein was usedas a control. (B) Co-IP of P58IPK and M2 protein. 293T cells growing in 6-cm-diameter dishes were transiently cotransfected with

plasmids coding for P58IPK and BM2 or A/M2, and at 48 h after transfection, cell extracts were collected, immunoprecipitation wasperformed with anti-Flag antibody. and analyzed by Western blotting with GFP or Flag antibody as shown. (C) Interaction of P58IPK

and M2 protein in virus-infected cells. 293T cells transiently expressed Flag-tagged P58IPK were infected by influenza A virus,immunoprecipitation was performed with anti-A/M2 antibody. Non-immunized mouse IgG was used as a control.



associating with host partners Hsp40 and P58IPK leads to theautophosphorylation of PKR, reduction of host proteinssynthesis, and finally, induction of cell apoptosis.

It has been thought that the induction of apoptosis is a hostdefense response, stopping the replication and spread ofvirus. However, the increasing evidence has shown thatapoptosis induction is beneficial for influenza virus replication.First of all, the expression of anti-apoptosis protein Bcl-2which inhibits influenza virus-induced apoptosis reducesvirus replication, spread and HA glycosylation (Olsen et al.,1996). Furthermore, the inhibition of caspase 3 activity whichis a member of the central component of the apoptoticmachinery strongly impairs influenza virus propagation(Wurzer et al., 2003). It seems that influenza virus hasacquired the capability to take advantage of the protectionmachinery of the host cells, thereby supporting viral replica-tion. The virus probably needs some mechanisms to keep thebalance between limitation of antiviral response and main-tenance of sufficient signaling strength to support virusgrowth. Such a balance may be controlled by proteinsencoded by influenza virus. There are several proteins ofinfluenza virus that has been reported to act as apoptosispromoters: NS1, PB1-F2, NA and M2. But the data that NS1acts as apoptosis inducer (Schultz-Cherry et al., 2001) waschallenged by the finding that recombinant influenza viruslacking NS1 still induced cell apoptosis. It is expressed duringthe early stage of infection and has been described as aninhibitor of PKR to promote viral protein synthesis. NAwas thefirst influenza virus protein shown to have a role in theinduction of apoptosis (Schultz-Cherry and Hinshaw, 1996;Morris et al., 1999). It can activate TGF-β at the cell surface byfacilitating cleavage of TGF-β into its active form. However,NA is not the sole contributor to apoptosis as UV-irradiatedvirus, which retains 100% NA activity, weakly inducedapoptosis. PB1-F2 is known to localize in the mitochondriaof the infected cell and to sensitize cells to death throughinteractions with two mitochondrial proteins, ANT3 andVDAC1 (Zamarin et al., 2005). These interactions promotethe permeabilization of the mitochondria, facilitate the releaseof mitochondrial products and trigger cell apoptosis. LikePB1-F2, M2 protein is expressed during the later stages of theinfection cycle. This correlates well with late requirement forTRAIL and caspase activity in the viral replication cycle. Theirpro-apoptotic effect most likely is not inhibitory to viralreplication.

Based on our results, we propose the following model forthe regulation of PKR pathway by influenza virus proteinsduring the infection (Fig. 9). PKR remains latent in unstimu-lated cells, and its activation requires binding of specificactivators. P58IPK is also inactive before influenza virusinfection because it is bound to Hsp40. Early in the infection,PKR is activated by dsRNA generated by viruses. Moreover,NS1 is expressed and blocks the dsRNA-mediated activationof PKR to fight against host defense and support normal viralreplication. P58IPK is also activated to block the dimerizationand activation of PKR due to the disruption of Hsp40-P58IPK

complex. During later stage of the infection, M2 is expressed

Protein & Cell

Figure 7. M2 (BM2 and A/M2 respectively) proteinenhances PKR phosphorylation. (A) Flag-PRK immobilizedonto anti-Flag M2-conjugated agarose was incubated withpurified soluble GSTcontrol (lane 1), GST-P58IPK (lane 2) and

the complex of GST-BM2C and GST-P58IPK (lane 3) in thepresence of [32P] ATP. Reaction mixtures were subjected toSDS-PAGE and visualized by autoradiography. (B) Over-

expression of M2 enhances PRK phosphorylation. 293T cellswere transfected with the indicated plasmids encoding for Flag-P58IPK, Flag-A/M2, or Flag-BM2 and treated with IFN-α and

poly(I:C) as described in “Materials and Methods.” Equalamounts of cell extracts were analyzed by SDS-PAGE andsubjected to immunoblotting with phospho-PKR, PKR and β-actin antibody. The top two panels show protein levels of

phosphorylated PKR and total PKR detected by phospho-PKR(P-PKR) and PKR antibodies (PKR). The third panel showsprotein level of β-actin acted as a loading control. The bottom

panel shows protein levels of P58IPK, A/M2 and BM2 detectedby Western blotting.



and associates with Hsp40 and P58IPK to prevent thedisruption of Hsp40-P58IPK complex. This leads to theactivation of PKR, and then may induce cell apoptosis andcontrol virus replication. Through this sophisticated way,influenza virus manipulates the host cells to favor itsreplication and release.

METHODS

Plasmid construction

To generate GSTor Flag epitope-tagged full-length protein ofHsp40, PCR was performed to amplify Hsp40 DNA fragmentfrom the original yeast two-hybrid library clone (pACT2-Hsp40) (see below). The PCR product was digested with

BamH I-Xho I and then inserted into pGEX6p-1 and pcDNA3-Flag (Clontech). Deletion mutants of Hsp40 were constructedby inserting PCR-generated fragments from the correspond-ing cDNAs into the pcDNA3-Flag vector. The mammalianexpression plasmids of wild-type BM2 and A/M2 fused withFlag epitope were generated by constructing to pCAGGS/MCS vector (kindly provided by Dr Y. Kawaoka, TheUniversity of Tokyo). The plasmid pcDNA3-Flag-PKR wasconstructed to use for purification of PKR protein byamplifying a full-length PKR cDNA from a human kidneycDNA library (Clontech) and then inserting to pcDNA3-Flagvector. For the expression of P58IPK in mammalian cells andE. coli, a P58IPK cDNA was obtained from Hela cells by RT-PCR and inserted to pcDNA3-Flag and pGEX6p-1 vector,respectively. The PCR fragment product (BM2C) of cytoplas-

Figure 8. Supravital PI assays identify the dead cells. HeLa cells were transfected with plasmid pCAGGS-AM2/BM2, stained

with PI (see Methods) and analyzed by flow cytometry. (A) Representative DNA histograms showing the proportions of deadhypodiploid nuclei detected by flow cytometry at 24–72h. (B) Proportions of dead cells by PI staining after transfected with pCAGGS,pCAGGS-AM2 or pCAGGS-BM2 at 24–72h. Results are expressed as means ± SD of results from three independent experiments.

Control, cells transfected with pCAGGS empty vector; AM2, cells transfected with pCAGGS-AM2; BM2, cells transfected withpCAGGS-BM2.



mic domain (amino acids 28–109) of the BM2 gene derivedfrom Influenza B virus strain (B/Yamagata/K542/2001)was inserted into pGBKT7 (Clontech) in frame with GAL4DNA binding domain (BD), resulting into plasmid pGBKT7-BM2C.

Yeast two-hybrid screening

For the initial screening, pGBKT7-BM2C was used as baitand pACT cDNA library (Clontech) from human kidney wasused as a source of prey genes. The bait pGBKT7-BM2Cplasmid and the pACT2 cDNA library were transformed intothe yeast strain AH109 by lithium acetate method (yeastprotocols handbook, Clontech). Transformants were platedonto SD medium lacking tryptophan, leucine and histidine butcontaining 1mM 3-aminotriazole. The candidate clones wererescued from the yeast cells and introduced to the yeast strainSFY526 to verify the interaction by detecting β-galactosidaseactivity. For quantitative β-galactosidase assays, colonieswere grown to mid-logarithmic phase in liquid selectionmedium before cells were harvested and lysed by theglass-bead method (yeast protocols handbook, Clontech).The procedures for library amplification, yeast cell transfor-mation, screening for growth in the absence of histidine, andmeasurement of β-galactosidase activity followed the Match-maker protocol (Clontech).

Cell lines and reagents

293T and Hela cells were cultured in high-glucose DMEM(Gibco) supplemented with 10% heat-inactivated fetal bovineserum (FBS), 2mM L-glutamine, and penicillin-streptomycin(100 units/mL; Invitrogen). Anti-Flag monoclonal M2 antibody,M2 anti-Flag-agarose, and poly(I:C) (synthetic dsRNA) werepurchased from Sigma. The polyclonal anti-PKR antibody (N-18; sc-6282) and phospho-specific anti-PKR antibody(against phosphorylated threonine 446) were purchasedfrom Santa Cruz Biotechnology.

Cell lysis and immunoblotting experiments

To analyze whether M2 protein affects PKR autophosphor-ylation in vivo, we performed the following assays. 293Tcellswere transfected with Flag-tagged A/M2, BM2 and/or P58IPK

constructs. Twenty-four hours later, the cells were treated withalpha interferon (IFN-α) at 1000U/mL for 24 h, and then poly(I:C) (100 μg/mL) was added directly to the cell culturemedium for 6 h to activate PKR. To harvest, the cells werewashed once with PBS buffer and proteins were extracted inice-cold lysis buffer containing 50mM Tris-HCl (pH8.0),150mM NaCl, 0.5% NP-40, 1mM dithiothreitol (DTT), 1mMphenylmethylsulfonyl fluoride (PMSF), CPI cocktail (Boehrin-ger Mannheim) as the source of protease inhibitors. The

Protein & Cell

Figure 9. Proposedmodel of modulation of PKR function by influenza virus. In the uninfected cells, both PKR and P58IPK arein inactivated state. At the start of virus infection, viral dsRNA activates PKR but in the early stage of influenza virus infection, theearly-synthesized protein NS1 binds to dsRNA thereof promoting the viral protein synthesis. In the late stage, late-synthesized viral

M2 activates PKR through its interaction with Hsp40 and P58IPK to inhibit host protein synthesis therefore promoting the viruspackaging and budding (for detail, please see text).



extracts were clarified by microcentrifugation at 13,000 rpmfor 10min. The protein concentration was determined bybicinchoninic acid (BCA) protein assay kit (Pierce Biotechnol-ogy), and 40μg of protein were fractionated by sodiumdodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The proteins were then electroblotted onto PVDFmembrane, which was incubated with each of the followingantibodies: phosphorylation site-specific antibody to PKR(Thr446), rabbit polyclonal antibodies to PKR and to β-actinas a control for protein loading. The secondary antibodieswere horseradish peroxidase-conjugated anti-rabbit or anti-goat IgG (Amersham Biosciences) used at a dilution of1:5000 in blocking solution.

Co-immunoprecipitation (Co-IP) assay

Immunoprecipitation assays were performed essentially asdescribed previously (Yan et al., 2003; Fan et al., 2006).Briefly, 293T cells were transiently transfected with theindicated plasmids using Lipofectamine 2000 reagents(Invitrogen). After 48 h of cultivation, the cells were washedand resuspended in 0.5mL lysis buffer (mentioned above).Equal amounts of cleared cell lysates were subjected toimmunoprecipitate with anti-Flag monoclonal antibody M2-conjugated agarose. The reactions were performed overnightat 4°C, and then the beads were centrifuged at 3000 rpm for2min and washed three times with lysis buffer. The antibody-protein complexes were then resolved by SDS-PAGE, andthe GFP or Flag-tagged proteins were identified by Westernblotting with an anti-GFP/Flag antibody probe using anenhanced chemiluminescence system. For the detection ofinteraction between P58IPK and A/M2 in the virus-infectedcells, 293T cells were transfected with pFlag- P58IPK plasmidfor 24 h, and then infected by influenza virus A/WSN/33. After24 h, co-immunoprecipitation was performed with anti-A/M2antibody (Abcam), non-immunized mouse IgG as a negativecontrol.

Preparations of GST fusion proteins and GST pull-downassay

The interactions between M2 and Hsp40 or P58IPK in vitrowere examined by GST pull-down assays. Transformants ofE. coli BL21 bearing plasmids encoding GSTor different GSTfusion proteins were grown to an optical density at 600 nm(OD600) of 0.6–0.7, and IPTG was added to 0.1mM to induceexpression of the GST proteins for 12 h at 16°C. Total proteinswere extracted by sonicating cells in PBS buffer containingNP-40 and protein inhibitors, followed by centrifugation at12,000 × g for 15min at 4°C. The supernatant were purified byglutathione-Sepharose 4B beads according to the manufac-turer’s recommended protocol (Amersham Pharmacia Bio-tech).

35S-labeled A/M2/BM2, Hsp40 and P58IPK proteins were

produced by in vitro transcription and translation usingplasmids pcDNA3/BM2, pcDNA3/Hsp40, pcDNA3/P58IPK,respectively. [35S]-Pro Mix (mixture of [35S]methionine;Amersham) and the TNT T7 coupled reticulocyte lysatesystem (Promega) were used as instructed by the manufac-turer. GST binding assays were conducted as follows: The35S-labeled proteins were incubated with 10 μg of GSTderivatives bound to glutathione-Sepharose beads in 0.5mLbinding buffer (50mM Tris-HCl (pH8.0), 150mM NaCl, 1mMEDTA, 0.5% NP-40, 1mM DTT, 1mM PMSF and proteaseinhibitors). The binding reaction was performed at 4°Covernight and the beads were subsequently washed fourtimes with the binding buffer. The beads were thenresuspended in 10μL of 2 ×SDS sample buffer, resolved bySDS-PAGE and followed by autoradiography.

Protein kinase assay

Purified recombinant Flag-tagged PKR immunoprecipitatedfrom cell extracts were used for the in vitro kinase reactions inkinase reaction buffer containing 20mM HEPES, pH 7.4,1mM DTT, 5mM MgCl2, 20μM ATP, 5μCi of [γ-32P]ATP(6mCi/mmol). Whenever indicated, poly(I:C) was added to afinal concentration of 1μg/mL. The kinase reactions wereincubated for 30min at 30°C, stopped by the addition of 2 ×SDS-PAGE sample buffer, boiled for 5min, and analyzed bySDS-PAGE (10%). The degree of PKR phosphorylation wasvisualized by autoradiography.

Cell death assay

HeLa cells were cultured in 6-well plates and allowed to growto 75%–80% confluency, and then were transfected withpCAGGS-AM2/BM2 (3μg/well). Cells were collected at 24, 48and 72 h post-transfection, washed twice with PBS and fixedby 75% cooling ethanol overnight. The fixed cells were thenstained with 50μg/mL PI in the dark at room temperature for15min. A minimum of 1 ×105 cells for each group wasanalyzed by fluorescence activated cell sorting (FACS).

ACKNOWLEDGEMENTS

This work was supported by National Natural Sciences Foundation ofChina (NSFC) (Grant Nos. 30670091 and 30599434), National BasicResearch Program (Project 973) of China Ministry of Science and

Technology (Grant No. 2011CB504703), National Key TechnologiesR&D Program (Grant No. 2006BAD06A01). GFG is a leadingprincipal investigator of the NSFC Innovative Research Group(Grant No. 81021003).

REFERENCES

Bergmann, M., Garcia-Sastre, A., Carnero, E., Pehamberger, H.,Wolff, K., Palese, P., and Muster, T. (2000). Influenza virus NS1

protein counteracts PKR-mediated inhibition of replication. J Virol74, 6203–6206.



Briedis, D.J., Lamb, R.A., and Choppin, P.W. (1982). Sequence of

RNA segment 7 of the influenza B virus genome: partial amino acidhomology between the membrane proteins (M1) of influenza A andB viruses and conservation of a second open reading frame.Virology 116, 581–588.

Cheetham, M.E., and Caplan, A.J. (1998). Structure, function andevolution of DnaJ: conservation and adaptation of chaperone

function. Cell Stress Chaperones 3, 28–36.

Ciampor, F., Thompson, C.A., Grambas, S., and Hay, A.J. (1992).Regulation of pH by the M2 protein of influenza A viruses. Virus

Res 22, 247–258.

Fan, Z., Zhuo, Y., Tan, X., Zhou, Z., Yuan, J., Qiang, B., Yan, J., Peng,X., and Gao, G.F. (2006). SARS-CoV nucleocapsid protein binds to

hUbc9, a ubiquitin conjugating enzyme of the sumoylation system.J Med Virol 78, 1365–1373.

Gale, M. Jr, and Katze, M.G. (1998). Molecular mechanisms ofinterferon resistance mediated by viral-directed inhibition of PKR,the interferon-induced protein kinase. Pharmacol Ther 78, 29–46.

Gale, M. Jr, Tan, S.L., Wambach, M., and Katze, M.G. (1996).Interaction of the interferon-induced PKR protein kinase with

inhibitory proteins P58IPK and vaccinia virus K3L is mediated byunique domains: implications for kinase regulation. Mol Cell Biol16, 4172–4181.

Goodman, A.G., Fornek, J.L., Medigeshi, G.R., Perrone, L.A., Peng,X., Dyer, M.D., Proll, S.C., Knoblaugh, S.E., Carter, V.S., Korth, M.J., et al. (2009). P58(IPK): a novel “CIHD” member of the host

innate defense response against pathogenic virus infection. PLoSPathog 5, e1000438.

Hatta, M., Goto, H., and Kawaoka, Y. (2004). Influenza B virus

requires BM2 protein for replication. J Virol 78, 5576–5583.

Horvath, C.M., Williams, M.A., and Lamb, R.A. (1990). Eukaryoticcoupled translation of tandem cistrons: identification of the

influenza B virus BM2 polypeptide. EMBO J 9, 2639–2647.

Ichinohe, T., Pang, I.K., and Iwasaki, A. (2010). Influenza virus

activates inflammasomes via its intracellular M2 ion channel. NatImmunol 11, 404–410.

Ilyinskii, P.O., Gabai, V.L., Sunyaev, S.R., Thoidis, G., and Shneider,

A.M. (2007). Toxicity of influenza A virus matrix protein 2 formammalian cells is associated with its intrinsic proton-channelingactivity. Cell Cycle 6, 2043–2047.

Ilyinskii, P.O., Gambaryan, A.S., Meriin, A.B., Gabai, V., Kartashov,A., Thoidis, G., Shneider, A.M., and Blagosklonny, M. (2008).Inhibition of influenza M2-induced cell death alleviates its negative

contribution to vaccination efficiency. PLoS ONE 3, e1417.

Katze, M.G., Tomita, J., Black, T., Krug, R.M., Safer, B., andHovanessian, A. (1988). Influenza virus regulates protein synth-

esis during infection by repressing autophosphorylation andactivity of the cellular 68,000-Mr protein kinase. J Virol 62,3710–3717.

Lamb, R.A., Lai, C.J., and Choppin, P.W. (1981). Sequences ofmRNAs derived from genome RNA segment 7 of influenza virus:colinear and interrupted mRNAs code for overlapping proteins.

Proc Natl Acad Sci U S A 78, 4170–4174.

Lee, S.B., Green, S.R., Mathews, M.B., and Esteban, M. (1994).

Activation of the double-stranded RNA (dsRNA)-activated humanprotein kinase in vivo in the absence of its dsRNA binding domain.Proc Natl Acad Sci U S A 91, 10551–10555.

Lee, T.G., and Katze, M.G. (1994). Cellular inhibitors of the interferon-

induced, dsRNA-activated protein kinase. Prog Mol Subcell Biol14, 48–65.

Liu, D., Liu, X., Yan, J., Liu, W.J., and Gao, G.F. (2009). Interspecies

transmission and host restriction of avian H5N1 influenza virus. SciChina C Life Sci 52, 428–438.

Lu, Y., Wambach, M., Katze, M.G., and Krug, R.M. (1995). Binding of

the influenza virus NS1 protein to double-stranded RNA inhibits theactivation of the protein kinase that phosphorylates the elF-2translation initiation factor. Virology 214, 222–228.

McCown, M.F., and Pekosz, A. (2006). Distinct domains of theinfluenza a virus M2 protein cytoplasmic tail mediate binding to theM1 protein and facilitate infectious virus production. J Virol 80,

8178–8189.

Melville, M.W., Hansen,W.J., Freeman, B.C., Welch, W.J., and Katze,

M.G. (1997). The molecular chaperone hsp40 regulates the activityof P58IPK, the cellular inhibitor of PKR. Proc Natl Acad Sci U S A94, 97–102.

Meurs, E.F., Watanabe, Y., Kadereit, S., Barber, G.N., Katze, M.G.,Chong, K., Williams, B.R., and Hovanessian, A.G. (1992).Constitutive expression of human double-stranded RNA-activated

p68 kinase in murine cells mediates phosphorylation of eukaryoticinitiation factor 2 and partial resistance to encephalomyocarditisvirus growth. J Virol 66, 5805–5814.

Mi, S., Li, Y., Yan J., and GAO, G.F. (2010). Na+/K+-ATPase β1 subunitinteracts with M2 proteins of influenza A and B viruses and affectsthe virus replication. Sci China C Life Sci 53, 1–8.

Mohler, P.J., Hoffman, J.A., Davis, J.Q., Abdi, K.M., Kim, C.R., Jones,S.K., Davis, L.H., Roberts, K.F., and Bennett, V. (2004). Isoformspecificity among ankyrins. An amphipathic alpha-helix in the

divergent regulatory domain of ankyrin-b interacts with themolecular co-chaperone Hdj1/Hsp40. J Biol Chem 279,25798–25804.

Morris, S.J., Price, G.E., Barnett, J.M., Hiscox, S.A., Smith, H., andSweet, C. (1999). Role of neuraminidase in influenza virus-inducedapoptosis. J Gen Virol 80, 137–146.

Morris, S.J., Smith, H., and Sweet, C. (2002). Exploitation of theHerpes simplex virus translocating protein VP22 to carry influenzavirus proteins into cells for studies of apoptosis: direct confirmation

that neuraminidase induces apoptosis and indications that otherproteins may have a role. Arch Virol 147, 961–979.

Mould, J.A., Paterson, R.G., Takeda, M., Ohigashi, Y., Venkataraman,P., Lamb, R.A., and Pinto, L.H. (2003). Influenza B virus BM2protein has ion channel activity that conducts protons acrossmembranes. Dev Cell 5, 175–184.

Odagiri, T., Hong, J., and Ohara, Y. (1999). The BM2 protein ofinfluenza B virus is synthesized in the late phase of infection and

incorporated into virions as a subviral component. J Gen Virol 80,2573–2581.

Ohtsuka, K., and Hata, M. (2000). Molecular chaperone function of

mammalian Hsp70 and Hsp40—a review. Int J Hyperthermia 16,231–245.

Olsen, C.W., Kehren, J.C., Dybdahl-Sissoko, N.R., and Hinshaw, V.S.

(1996). bcl-2 alters influenza virus yield, spread, and hemagglu-tinin glycosylation. J Virol 70, 663–666.

Pinto, L.H., Holsinger, L.J., and Lamb, R.A. (1992). Influenza virus M2protein has ion channel activity. Cell 69, 517–528.

Pinto, L.H., and Lamb, R.A. (2006). The M2 proton channels of

Protein & Cell



influenza A and B viruses. J Biol Chem 281, 8997–9000.

Schultz-Cherry, S., Dybdahl-Sissoko, N., Neumann, G., Kawaoka, Y.,

and Hinshaw, V.S. (2001). Influenza virus ns1 protein inducesapoptosis in cultured cells. J Virol 75, 7875–7881.

Schultz-Cherry, S., and Hinshaw, V.S. (1996). Influenza virus

neuraminidase activates latent transforming growth factor beta. JVirol 70, 8624–8629.

Shimbo, K., Brassard, D.L., Lamb, R.A., and Pinto, L.H. (1996). Ion

selectivity and activation of the M2 ion channel of influenza virus.Biophys J 70, 1335–1346.

Sohn, S.Y., Kim, J.H., Baek, K.W., Ryu, W.S., and Ahn, B.Y. (2006).Turnover of hepatitis B virus X protein is facilitated by Hdj1, ahuman Hsp40/DnaJ protein. Biochem Biophys Res Commun 347,764–768.

Srivastava, S.P., Kumar, K.U., and Kaufman, R.J. (1998). Phosphor-ylation of eukaryotic translation initiation factor 2 mediates

apoptosis in response to activation of the double-stranded RNA-dependent protein kinase. J Biol Chem 273, 2416–2423.

Sugrue, R.J., and Hay, A.J. (1991). Structural characteristics of the

M2 protein of influenza A viruses: evidence that it forms atetrameric channel. Virology 180, 617–624.

Takeda, M., Pekosz, A., Shuck, K., Pinto, L.H., and Lamb, R.A.(2002). Influenza a virus M2 ion channel activity is essential forefficient replication in tissue culture. J Virol 76, 1391–1399.

Tan, S.L., Gale, M.J. Jr, and Katze, M.G. (1998). Double-strandedRNA-independent dimerization of interferon-induced proteinkinase PKR and inhibition of dimerization by the cellular P58IPK

inhibitor. Mol Cell Biol 18, 2431–2443.

Wurzer, W.J., Planz, O., Ehrhardt, C., Giner, M., Silberzahn, T.,Pleschka, S., and Ludwig, S. (2003). Caspase 3 activation is

essential for efficient influenza virus propagation. EMBO J 22,2717–2728.

Yan, J., Zhu, J., Zhong, H., Lu, Q., Huang, C., and Ye, Q. (2003).

BRCA1 interacts with FHL2 and enhances FHL2 transactivationfunction. FEBS Lett 553, 183–189.

Zamarin, D., García-Sastre, A., Xiao, X., Wang, R., and Palese, P.(2005). Influenza virus PB1-F2 protein induces cell death throughmitochondrial ANT3 and VDAC1. PLoS Pathog 1, e4.



RESEARCH ARTICLE

DEXH-Box protein DHX30 is required foroptimal function of the zinc-finger antiviralprotein

Peiying Ye1,2, Shufeng Liu1, Yiping Zhu1,2, Guifang Chen1, Guangxia Gao1✉

1 Center for Infection and Immunity, Institute of Biophysics, Chinese Academy of Sciences, Beijing 100101, China2 Graduate School of Chinese Academy of Sciences, Beijing 100039, China✉ Correspondence: [email protected] August 29, 2010 Accepted September 26, 2010

ABSTRACT

The zinc-finger antiviral protein (ZAP) is a host factor thatspecifically inhibits the replication of certain viruses byeliminating viral mRNAs in the cytoplasm. In previousstudies, we demonstrated that ZAP directly binds to theviral mRNAs and recruits the RNA exosome to degradethe target RNA. In this article, we provide evidence that aDEXH box RNA helicase, DHX30, is required for optimalantiviral activity of ZAP. Pull-down and co-immunopreci-pitation assays demonstrated that DHX30 and ZAPinteracted with each other via their N terminal domains.Downregulation of DHX30 with shRNAs reduced ZAP’santiviral activity. These data implicate that DHX30 is acellular factor involved in the antiviral function of ZAP.

KEYWORDS zinc-finger antiviral protein, RNA heli-case, DHX30

INTRODUCTION

The zinc-finger antiviral protein (ZAP) was originally cloned asa host factor that prevented cells from infection by Moloneymurine leukemia virus (MLV) (Gao et al., 2002). Over-expression of ZAP inhibits the replication of MLV, Ebolavirus and Marburg virus (Muller et al., 2007), and severalmembers of the Alpha virus genus (Bick et al., 2003). ThemRNA level of endogenous ZAP in mouse bone marrow-derived dendritic cells can be significantly up-regulated byinterferon treatment or Sindbis virus (SINV) infection (Zhanget al., 2007). These results suggest that ZAP may play animportant role in the host antiviral system in vivo.

ZAP specifically eliminates the cytoplasmic viral mRNA

(Gao et al., 2002). It directly binds to the target viral mRNAsthrough the zinc-finger motifs (Guo et al., 2004) and recruitsthe RNA degradation machineries to degrade the target viralmRNA (Guo et al., 2007; Zhu and Gao, 2008). mRNAdegradation is a highly organized and complex process(Guhaniyogi and Brewer, 2001; Garneau et al., 2007),involving many co-factors. We previously reported that theDEAD-box RNA helicase p72 interacted with ZAP and wasrequired for the optimal function of ZAP (Chen et al., 2008). Itis reasonable to speculate that more co-factors may beinvolved in the antiviral function of ZAP.

DHX30 is an RNA helicase belonging to the DExD/H familysince it bears the DEVH signature sequence in the Walker Bmotif (Zhou et al., 2007). There are three isoforms of DHX30reported on NCBI. Isoform 1 and 2 contain the DEVHsignature helicase domain and a putative type A dsRBD atthe N-terminal region. Isoform 3 (NM_138614.1) lacks thehelicase core domain and is believed to exist at very lowabundance because it is a nonsense-mediated mRNA decay(NMD) candidate. Annotation of AceView suggests thatDHX30 gene contains 39 different introns. Furthermore,transcription of DHX30 produces 22 different mRNAs byalternative splicing, which have good potential to betranslated into 18 different isoforms. Indeed, besides Isoform1, 2 and 3, a specific isoform of DHX30 was identified as amitochondrial nucleoid protein. The antibody raised againstthe N-terminal domain of DHX30 detects both nuclear andmitochondrial isoforms of DHX30 (Wang and Bogenhagen,2006). It has been reported that overexpression of DHX30enhanced HIV-1 gene expression by augmenting basaltranscription from the HIV-1 LTR promoter. However, knock-down of DHX30 did not distinctly affect HIV-1 LTR-dependenttranscription (Zhou et al., 2007). Moreover, overexpression of



Protein & Cell

DHX30 severely restrained the packaging of HIV-1 RNA, andthereby reduced virus infectivity (Zhou et al., 2007).

In the present study, we identified DHX30 as a ZAP-interacting protein involved in the antiviral function of ZAP.

RESULTS

Identification of the DHX30 RNA helicase as aZAP-interacting protein

To further study the mechanism for ZAP’s antiviral activity, weset out to identify cellular factors that associated with ZAP. His-myc-tagged rat ZAP (rZAP) was expressed in 293TRex-ZAPcells in a tetracycline-inducible manner. rZAP and its asso-ciated proteins were purified using Ni-NTA Agarose, resolvedon SDS/PAGE and visualized by Coomassie blue staining. Toprevent possible nonspecific RNA tethering, the lysates weretreated with RNase A. Compared with the precipitates of293TRex control cells, a specific band of 130 kD was detectedin the rZAP-expressing cells (Fig. 1A). This band was excisedand subjected to matrix-assisted laser desorption ionizationtime-of-flight (MALDI-TOF) mass spectrometry (MS) analysisto identify the protein. The result of database search identifiedthe DHX30 RNA helicase as a candidate.

To confirm the interaction between rZAP and DHX30,coimmunoprecipitation assays were performed. Flag-taggedrZAP and myc-tagged DHX30 were co-transfected intoHEK293T cells. Immunoprecipitation of DHX30 using theanti-myc antibody coprecipitated Flag-tagged rZAP in theabsence of RNase A (Fig. 1B). Treatment of the cell lysateswith RNase A reduced the interaction between rZAP andDHX30. The reduced interaction was likely due to the fact thatDHX30 precipitated easily after RNase A treatment, leavingrelatively low amount of DHX30 in the supernatant forimmunoprecipitation (Fig. 1B, lower panel). In addition, it isalso possible that both RNA-dependent and independentZAP-DHX30 interactions may exist and that RNase treatmentmay have removed the RNA-dependent interaction. TheRNA-independent interaction between ZAP and DHX30 wasfurther confirmed in later experiments (see below). In areverse experiment, immunoprecipitation of myc-taggedrZAP coprecipitated DHX30 in the absence or presence ofRNase A (Fig. 1C). The interaction between human ZAP(hZAP) and DHX30 was also investigated and as expected,hZAP also interacted with DHX30 (Fig. 1D).

Mapping the binding domains of ZAP and DHX30

The N-terminal domain of ZAP fused with the zeocinresistance gene (NZAP-Zeo) displayed the same antiviralactivity as the full-length ZAP (Gao et al., 2002). If DHX30 isinvolved in the function of ZAP, it would be expected tointeract with the N-terminal domain of ZAP. The N-terminaldomain of rZAP (rNZAP254), the C-terminal domains of rZAP

(rCZAP193 and rCZAP236) were individually co-expressedwith DHX30 and analyzed for their interactions with DHX30.As expected, immunoprecipitation of DHX30 coprecipitatedonly the N-terminal domain of rZAP, but not the C-terminaldomain (Fig. 2A). In the reverse experiment, only immuno-precipitation of the N-terminal domain of rZAP, but not the C-terminal domains of rZAP, coprecipitated DHX30 (Fig. 2B).These results established that DHX30 interacted with the N-terminal domain of rZAP.

To determine the domains of DHX30 responsible for itsinteraction with ZAP, DHX30 truncation mutants were con-structed and analyzed for their abilities to bind to ZAP (Fig. 3).Immunoprecipitation of N-terminal domain of DHX30(NDHX30) coprecipitated rZAP (Fig. 3A). The truncationmutant containing the N-terminal domain and the core domain(N + core) also interacted with rZAP, although the band of co-immunoprecipitated ZAP by N + core was weaker comparedwith the N-terminal domain alone. The weaker band likelyreflected the relatively low level of N + core (Fig. 3A, lowerpanel) since the intensities of the bands are roughlyproportional. In contrast, the core domain alone or the C-terminal domain of DHX30 failed to interact with rZAP(Fig. 3A). The interaction between the N-terminal domain ofDHX30 and rZAP was not affected by the treatment of the celllysate with RNase A (Fig. 3B). Pull-down assays wereemployed to further prove the interaction between the N-terminal domain of DHX30 and ZAP using bacteriallyexpressed NDHX30 fused with maltose binding protein(MBP). With or without RNase A treatment, MBP-NDHX30pulled down rZAP, hZAP and rNZAP (Fig. 3C). The directinteraction between NDHX30 and hNZAP was also confirmedby pull-down assays using purified bacterially expressedproteins (Fig. 3D).

Downregulation of DHX30 reduced ZAP’s activity

To investigate the role of DHX30 in ZAP’s antiviral function,three shRNAs directed against DHX30 were designed andvalidated. The plasmids expressing the shRNAs directedagainst DHX30 (shDHX30i) were first tested for their ability todown-regulate the expression of DHX30 by cotransfected into293A cells with the construct expressing the myc-taggedDHX30. Compared with the control shRNA (ctrl), the shRNAsdirected against DHX30 (Di-1, Di-2 and Di-3) significantlyreduced the expression level of myc-tagged DHX30 (Fig. 4A).The efficiency of these three shRNAs against endogenousDHX30 was confirmed by RT-PCR (Fig. 4B). To test the effectof DHX30 downregulation on the activity of ZAP, 293TRex-ZAP cells were cotransfected with pMLV-Luc reporter andshDHX30i, and assayed for inhibition of the reporter. TheshRNA directed against p72 (72i), which has been shown toreduce ZAP’s activity, was used as a positive control. All threeshRNAs against DHX30 reduced the activities of both rZAPand hZAP (Fig. 4C and 4D).


DHX30 is involved in the antiviral function of ZAP Protein & Cell

Protein & Cell

Figure 1. DHX30 interacted with ZAP. (A) Identification of DHX30 as a putative ZAP-interacting protein 293TRex or 293TRex-

ZAP cells were treated with tetracycline to induce ZAP expression. ZAP and associated proteins were purified by using Ni-NTAAgarose and resolved on 10% SDS-PAGE. The proteins were visualized by Coomassie blue staining. The band specific to the ZAP-expressing cells (indicated by the arrow) was excised and subjected to MALDI-TOF mass spectrometry analysis. Ctrl: 293TRex

control cells; ZAP: ZAP-expressing 293TRex-ZAP cells. (B) co-IP of rZAP with DHX30. The plasmids expressing Flag-tagged rZAPand myc-tagged DHX30 were transiently cotransfected into HEK293T cells. At 48 h posttransfection, the cells were lysed. The celllysates were immunoprecipitated with the anti-myc antibody in the presence (+) or absence (−) of RNase A andWestern blotted withthe anti-Flag and anti-myc antibodies. (C) co-IP of DHX30 with rZAP. The plasmids expressing Flag-tagged DHX30 and myc-tagged

rZAP were transiently cotransfected into HEK293T cells. At 48 h posttransfection, the cells were lysed. The cell lysates wereimmunoprecipitated with the anti-myc antibody in the presence (+) or absence (−) of RNase A and Western blotted with the anti-Flagand anti-myc antibodies. (D) co-IP of DHX30 with hZAP. The plasmids expressing Flag-tagged DHX30 and myc-tagged hZAP were

transiently cotransfected into HEK293T cells. The cell lysates were immunoprecipitated with the anti-Flag antibody in the presence(+) of RNase A and Western blotted with the anti-myc and anti-Flag antibodies.


Peiying Ye et al.

In an attempt to test whether DHX30 is also involved inother RNA degradation processes, DHX30i was analyzed forits effect on the degradation of AU-rich element (ARE)-containing mRNA by AUBP, which binds to ARE and recruitsexosome to degrade the ARE-containing RNA. DHX30i hadlittle effect on ARE-containing reporter (Fig. 4E), suggestingthat DHX30 RNA helicase is not involved in ARE mediatedmRNA decay.

DISCUSSION

ZAP specifically inhibits the replication of certain viruses byrecognizing the ZAP-responsive element (ZRE) present in the

target viral mRNAs and recruiting the cellular RNA degrada-tion machineries to degrade the target viral mRNAs. Amongthe ZREs so far identified, no obvious sequence similarity orcommon motifs has been observed. The only commonfeature of these ZREs is that they are at least 500 nucleotideslong. It is plausible that the ZRE-containing viral mRNAs formcomplex tertiary structures in vivo. Studies of reconstitutedRNA exosome revealed that the exosome could process onlyextended RNA substrate but not stem-loop structured RNAsubstrate (Liu et al., 2006). Thus, it is reasonable to speculatethat disruption of the secondary structures by RNA helicasesis necessary for efficient ZAP-mediated viral mRNA degrada-tion.

Figure 2. Mapping the binding domains of ZAP. (A) (B) co-IP of rZAP truncation mutants with DHX30. (A) The plasmidexpressing the indicated myc-tagged truncated rZAP protein was individually cotransfected into HEK293T cells with the plasmidexpressing Flag-tagged DHX30. The cell lysates were immunoprecipitated with the anti-Flag antibody in the presence of RNase Aand Western-blotted with the anti-myc and anti-Flag antibodies. Input, total cell lysate; Ctrl, pCMV-HA-Flag empty vector; EV,

pcDNA4 empty vector; D30, DHX30. Schematic representations of the rZAP proteins are shown, and their binding activities aresummarized. (B) The plasmid expressing the indicated myc-tagged truncated rZAP protein was individually cotransfected intoHEK293T cells with the plasmid expressing Flag-tagged DHX30. The cell lysates were immunoprecipitated with the anti-myc

antibody in the presence of RNase A and Western-blotted with the anti-Flag and anti-myc antibodies. Ctrl, pCMV-HA-Flag emptyvector; EV, pcDNA4 empty vector; D30, DHX30.



Protein & Cell

Figure 3. Mapping the binding domains of DHX30. (A) co-IP of rZAP with DHX30 truncation mutants. The plasmid expressingthe indicated Flag-tagged truncated DHX30 protein was individually cotransfected into HEK293Tcells with the plasmid expressingmyc-tagged rZAP. The cell lysates were immunoprecipitated with the anti-Flag antibody in the presence (+) or absence (−) of RNaseA and Western-blotted with the anti-myc and anti-Flag antibodies. Input, total cell lysate; Ctrl, pCMV-HA-Flag; EV, pcDNA4. Thepositions of the truncated DHX30 proteins are indicated by asterisks. Schematic representations of the DHX30 proteins are shown,


Peiying Ye et al.

Defined by the highly conserved sequence motifs requiredfor RNA binding and ATP hydrolysis, most RNA helicaseshave the intrinsic capability to unwind RNA duplexes(Jankowsky and Fairman, 2007). Furthermore, variable N-or C-terminal domain flanking the highly conserved helicasecore usually grants unique substrate specificity. It is con-ceivable that multiple RNA helicases may be needed tounwind complex RNA tertiary structure. We previouslyreported that the DEAD box RNA helicase p72 was requiredfor the optimal function of ZAP. Here, we observed thatDHX30 interacted with ZAP through its N-terminal domain,and downregulation of DHX30 reduced ZAP’s antiviralactivity. Whether and how these two RNA helicasescoordinate to contribute to ZAP’s activity awaits furtherinvestigation.

It is unclear yet how DHX30 is involved in ZAP-mediatedantiviral activity. The likely possibility is that ZAP binds to thetarget RNA and recruits DHX30 to help unwind the RNA tofacilitate efficient RNA degradation. However, other possibi-lities also exist. For example, ZAP may recruit DHX30 just asa co-factor to further recruit other RNA degradation machi-neries. Further investigation is needed to fully understandhow DHX30 is involved in the antiviral activity of ZAP.


Plasmids

The plasmids pcDNA4TO/myc-ZAP, pcDNA4TO/myc-NZAP,

pcDNA4TO/myc-CZAP-A, and pcDNA4TO/myc-CZAP-C, whichexpress myc-tagged full-length rZAP, rNZAP (amino acids 1–254),rCZAP-I (amino acids 193–776), and rCZAP-II (amino acids

236–776), respectively, have been described previously (Guo et al.,2007). pcDNA4TO-ZAP-Flag expresses Flag-tagged ZAP. Thecoding sequence of rZAP was amplified from pcDNA4TO/myc-ZAP

by using forward primer ZAP-5 bearing a KpnI site and reverse primerZAP-3 bearing a NotI site, and cloned into the expression vectorpcDNA3-Flag by using these two sites: ZAP-5: 5′-GGGGTAC-CATGGCAGATCCCGGGGTATGCTGTTTC-3′ ; ZAP3: 5′-ATAA-

GAATGCGGCCGCTCTGGACCTCTTCTCTTCTGC-3′.pcDNA4TO/myc-hZAP expresses myc-tagged human ZAP iso-

form 2. The coding sequence of human ZAP iso2 was amplified from

a human fetal liver cDNA library by using forward primer hZAP-upbearing an BamHI site and reverse primer hZAP-down bearing a NotIsite and cloned into the expression vector pcDNA4TO/myc-his by

using these two sites: hZAP-up: 5'-AATAGGATCCGCCAC-

CATGGCGGACCCGGAGGTGTGC-3' and hZAP-down: 5'-ATCT-GAGCGGCCGCGGTCTGGCCCTCTCTTCATCTGCT-3'.

The DHX30-deletion mutants were generated by cloning PCR-derived DHX30 fragments into pCMV-HA-Flag (Guo et al., 2007). AnEcoRI site was built in the forward primers, and a NotI site was built in

the reverse primers. The primers are listed with the restriction sitesitalicized:

DHX30-FP2, 5′-GGAATTCGCAGCTTCTAGGGACCTATTAAA-3′;DHX30-RP2,5′-ATATAGGCGGCCGCTCAGTCGTCAGCTGTCTT

GCG-3′;DHX30N-RP, 5′-ATATAGGCGGCCGCTCATGGGTCCACAGGTA

GCTG-3′;

DHX30-core-FP, 5′-GGAATTCCATCGGGACACCATCCTCA-3′;DHX30-core-RP, 5′-ATATAGGCGGCCGCTCATGTGATGGAAGT

CTCAGC-3′;

DHX30C-FP, 5′-GGAATTCATCAATGACATCGTGCATG-3′.The plasmid-expressing N-terminal portion of hZAP with GST

fused at the N-terminus in E. coli has been described previously (Lawet al., 2010). The plasmid expressing the N-terminal portion of DHX30

with MBP fused at the N-terminus in E.coli was generated by cloningPCR-derived DHX30 fragments into pMAL-c2x-linker vector. AnEcoRI site was built in the forward primers, MBP-DHX30-FP (5′-

GGAATTCGCAGCTTCTAGGGACCTATTAAA-3′) and a NotI site wasbuilt in the reverse primers DHX30N-RP.

pSUPER.puro-DHX30 RNAi expresses shRNA directed against

DHX30. Oligonucleotides DHX30-RNAi FP and DHX30-RNAi RPwere annealed and cloned into pSUPER.puro vector by using theBglII and HindIII sites to generate pSUPER.puro-DHX30 RNAi.

pSUPER.puro control RNAi and GFP RNAi were constructed byusing the same strategy. The sequences of the oligos are listedbelow:

DHX30 RNAi-1-FP: 5′-GATCCCCCAGCTGAATCCAGAGAG-

TATTCAAGAGATACTCTCTGGATTCAGCTGTTTTTA-3′;DHX30 RNAi-1-RP: 5′-AGCTTAAAAACAGCTGAATCCAGAGAG-

TATCTCTTGAATACTCTCTGGATTCAGCTGGGG-3′;

DHX30 RNAi-2-FP: 5′-GATCCCCGCTGTGGACAGTCCAAA-CATTCAAGAGATGTTTGGACTGTCCACAGCTTTTTA-3′;

DHX30 RNAi-2-RP:5′-AGCTTAAAAAGCTGTGGACAGTCCAAA-

CATCTCTTGAATGTTTGGACTGTCCACAGCGGG-3′;DHX30 RNAi-3-FP: 5’-GATCCCCGATGGATCAGAAGGCCA-

TATTCAAGAGATATGGCCTTCTGATCCATCTTTTTA-3';

and their binding activities are summarized. (B) co-IP of rZAP with N-terminal domain of DHX30. The plasmids expressing Flag-

tagged NDHX30 and myc-tagged rZAP were transiently cotransfected into HEK293Tcells. The cell lysates were immunoprecipitatedwith the anti-Flag antibody in the presence (+) or absence (−) of RNase A and Western blotted with the anti-myc and anti-Flagantibodies. (C) Pull-down of ZAP with MBP-N-DHX30. Bacterially expressed MBP or MBP-NDHX30 was immobilized onto Amylose

resin and incubated with the lysates of the cells expressing the indicated myc-tagged ZAP proteins in the presence of RNase A at 4°Cfor 2 h. The resins were washed and boiled in the sample loading buffer. The proteins were resolved by SDS-PAGE and detected byWestern blotting using the anti-myc antibody. Input, total cell lysate; EV, pcDNA4. The positions of MBP and MBP-NDHX30 are

indicated by arrows. (D) Pull-down of MBP-N-DHX30 with GST-N-hZAP. Bacterially expressed GSTor GST-N-hZAP was immobilizedonto glutathione-Sepharose 4B resin and incubated with purified MBP-NDHX30 in the presence of RNase A at 4°C for 2 h. The resinswere washed and boiled in the sample loading buffer. The proteins were resolved by SDS-PAGE and detected by Coomassie staining(upper panel) or Western blotting using the anti-ND30 antibody (lower panel). Input, purified fusion proteins; Ctrl, MBP; ND30, MBP-

NDHX30. The positions of the fusion proteins are indicated by asterisks. The band of the pull-downed protein is indicated by an arrow.



Protein & Cell

Figure 4. Downregulation of DHX30 reduced ZAP’s activity. (A) The effect of DHX30 shRNA on DHX30 expression. The

plasmids expressing myc-tagged DHX30 and GFP was cotransfected into HEK293A cells with the plasmid expressing controlshRNA (Ctrl1, Ctrl2) or the shRNA directed against DHX30 (Di-1, Di-2 or Di-3). The expression levels of the DHX30 proteins weremeasured by Western blotting using the anti-myc antibody. GFP-myc served as an internal control. (B) The effect of DHX30 shRNAon the mRNA level of endogenous DHX30. The plasmid expressing the indicated shRNA was transfected into HEK293 cells. The

cells were selected in puromycin to remove the untransfected cells. At 48 h after puromycin treatment, total RNA was isolated andthe RNA level of endogenous DHX30 was measured by Realtime PCR using GAPDH as an internal control. (C) (D) The effect ofDHX30 shRNA on ZAP’s activity. The pMLV-Luc reporter was cotransfected with the plasmid expressing the indicated shRNA into

(C) 293TRex-rZAP cells or (D) 293TRex-hZAP cells. At 4 h posttransfection, the cells were treated with puromycin to remove theuntransfected cells and tetracycline was added to induce ZAP expression. At 48 h posttransfection, the cells were lysed andluciferase activities were measured. Fold inhibition was calculated as the luciferase activity in the mock-treated cells divided by the

luciferase activity in the tetracycline-treated cells. Relative fold inhibition was calculated as percentage of the fold inhibition in thepresence of the Ctrl RNAi (average of Ctrl1 and Ctrl2) divided by the fold inhibition in the presence of the indicated RNAi. 72i servedas a positive control. The data are means +SE of four independent experiments. The asterisks (*) denote P<0.001. (E) The effect ofshDHX30 on AUBP-mediated mRNA degradation. The ARE-containing reporter, pGL3-ARE-Luc, or pGL3 empty vector was

cotransfected with the plasmid expressing the indicated shRNA into HEK293 cells. At 48 h posttransfection, the cells were lysed andluciferase activities were measured. Fold inhibition was calculated as the luciferase activity in the cells transfected with pGL3 dividedby the luciferase activity in the cells transfected with pGL3-ARE-Luc. Relative fold inhibition was calculated as percentage of the fold

inhibition in the presence of the Ctrl RNAi (average of Ctrl1 and Ctrl2) divided by the fold inhibition in the presence of the indicatedRNAi. The data are means + SE of three independent experiments.


Peiying Ye et al.

DHX30 RNAi-3-RP: 5′-AGCTTAAAAAGATGGATCAGAAGGCCA-TATCTCTTGAATATGGCCTTCTGATCCATCGGG-3′;

Control RNAi-FP: 5′-GATCCCCGAGCACTCTGAACTACCTGTT-

CAAGAGACAGGTAGTTCAGAGTGCTCTTTTTGGAAA-3′;Control RNAi-RP: 5′-AGCTTTTCCAAAAAGAGCACTCTGAAC-

TACCTGTCTCTTGAACAGGTAGTTCAGAGTGCTCGGG-3′.

Cell culture

All the cells were maintained in DMEM supplemented with 10% FBS.

Transfection was performed by using Lipofectamine 2000 (Invitrogen)following the manufacturer’s instructions. 293TRex and 293TRex-ZAP cell lines have been described previously (Guo et al., 2004). Toanalyze the effect of the RNAi on ZAP’s activity, 293TRex-ZAP cells

were transfected with the pMLV-Luc reporter, pRL-TK (for normalizingtransfection efficiency) and the effector-expressing plasmid. Immedi-ately after transfection, the cells were mock-treated or treated with

tetracycline to induce ZAP expression. Forty-eight hours aftertransfection, luciferase activities were measured and normalized bydividing the firefly luciferase activity with the Renilla luciferase activity.

Fold inhibition by ZAP was calculated as the normalized luciferaseactivity in the mock-treated cells divided by the normalized luciferaseactivity in the tetracycline-treated cells. To assay the effect of RNAi onthe activity of AUBP, 293A cells were cotransfected with pGL3 or a

type II ARE-containing reporter, pGL3-ARE-Luc (Guo et al., 2004),together with pRL-TK, and the shRNA-expressing plasmid. At 48hposttransfection, the luciferase activities were measured. Fold

inhibition by AUBP was calculated as the normalized luciferaseactivity in the pGL3-transfected cells divided by the normalizedluciferase activity in the pGL3-ARE-Luc-transfected cells. The

statistical significance of the data is analyzed with an SPSS program.

Identification of ZAP-interacting proteins

293TRex-ZAP cells were treated with tetracycline at a finalconcentration of 1 μg/mL to induce ZAP expression. Thirty-six hoursafter induction, the cells were lysed with lysis buffer B [30mM Hepes(pH 7.6), 100mM NaCl, 0.5% Nonidet P-40, and protease inhibitors

mixture]. The lysates were clarified by centrifugation at 13,000 rpmfor 10min at 4°C in a microcentrifuge (Sorvall Fresco), treated with50μg/mL RNase A for 15min at 37°C. ZAP and associated proteins

were purified by using Ni-NTA Agarose (QIAGEN) following themanufacturer’s instructions. The ZAP complex was eluted with100mM imidazole and resolved on 10% acrylamide gels for SDS-

PAGE. The proteins were visualized by Coomassie blue staining. Thespecific bands from the ZAP-expressing cells were excised from thegel and digested with trypsin. The resulting peptides were analyzed

by MALDI-TOF MS. The acquired MS data were analyzed with ahuman nucleotide/protein database by using a MASCOT databasesearch tool for peptide identification. The protein was identified whenmultiple peptides corresponding to the ORF of a protein were

identified.

RNA isolation, reverse transcription and realtime-PCR

Total RNA was isolated from HEK293A cells using the RNeasy kit(Qiagen) following the manufacturer's protocol and checked for itsquality and concentration. 2μg of total RNA was reverse transcribed

using the M-MLV reverse transcriptase. The cDNA was amplified by

using 1 μL of the RT reaction into a 20-μL PCR reaction containing 10pmol of the following primers: DHX30 realtime FP 5′-GCACAAGTC-GACCATTAACAGGGAG-3 ′ and DHX30 realt ime RP 5 ′-

ACTGTCGCTCAGTGAGATGGTGGCC-3′. Amplification was carriedout under standard conditions using 2×PCR Master Mix (TiangenBiotech, Beijing, China) on the Rotor-Gene 6000TM real-time PCR

instrument (Corbett Research).

Antibodies

The anti-myc mouse monoclonal antibody 9E10 (Santa CruzBiotechnology) was used to immunoprecipitate or detect proteinswith the myc epitope, and the anti-Flag mouse monoclonal antibodyM2 (Sigma-Aldrich) was used to immunoprecipitate or detect proteins

with the Flag epitope. Bacterially expressed N-terminal portion ofDHX30 was used to immunize rabbits to generate polyclonalantibodies against DHX30. The antibodies were affinity-purified by

using the cognate protein.

Coimmunoprecipitation

Cells were lysed in lysis buffer B [30mM Hepes (pH 7.6), 100 mMNaCl, 0.5% Nonidet P-40, and protease inhibitors mixture] mock-treated or treated with RNase A (5 mg/mL) on ice for more than one

hour, and the lysates were clarified by centrifugation at 4°C for 20minat 13,000 rpm. The supernatant was mixed with protein G plusagarose (Santa Cruz Biotechnology) and the antibody and incubatedat 4°C for 2 h. The resins were then washed three times with lysis

buffer B, and the bound proteins were detected by Western blotting.

ACKNOWLEDGEMENTS

We thank Jing Sun and Yihui Xu for technical support and Dr. FuquanYang for help with the MALDI-TOF analyses. This work was in partsupported by the grant to Guangxia Gao from National ScienceFoundation of China (Grant No. 81030030), and by the grant toGuifang

Chen from National Science Foundation (Grant No. 30800053).

ABBREVIATIONS

AMD, ARE mediated mRNA decay; Ctrl, control plasmid; Di-1,shRNA directed against DHX30i number 1; Di-2, shRNA directedagainst DHX30i number 2; Di-3, shRNA directed against DHX30inumber 3; dsRBD, double-stranded RNA-binding domain; EV, empty

vector; F.I., fold inhibition; MLV, murine leukemia virus; ZAP, zinc-finger antiviral protein; ZRE, ZAP responsive element

REFERENCES

Bick, M.J., Carroll, J.W., Gao, G., Goff, S.P., Rice, C.M., and

MacDonald, M.R. (2003). Expression of the zinc-finger antiviralprotein inhibits alphavirus replication. J Virol 77, 11555–11562.

Chen, G., Guo, X., Lv, F., Xu, Y., and Gao, G. (2008). p72 DEAD box

RNA helicase is required for optimal function of the zinc-fingerantiviral protein. Proc Natl Acad Sci U S A 105, 4352–4357.

Gao, G., Guo, X., and Goff, S.P. (2002). Inhibition of retroviral RNA

production by ZAP, a CCCH-type zinc finger protein. Science 297,1703–1706.

Garneau, N.L., Wilusz, J., Wilusz, C.J. (2007). The highways and



byways of mRNA decay. Nat Rev Mol Cell Biol 8, 113–126.

Guhaniyogi, J., Brewer, G. (2001). Regulation of mRNA stability in

mammalian cells. Gene 265, 11–23.

Guo, X., Carroll, J.W., Macdonald, M.R., Goff, S.P., and Gao, G.(2004). The zinc finger antiviral protein directly binds to specific

viral mRNAs through the CCCH zinc finger motifs. J Virol 78,12781–12787.

Guo, X., Ma, J., Sun, J., and Gao, G. (2007). The zinc-finger antiviral

protein recruits the RNA processing exosome to degrade the targetmRNA. Proc Natl Acad Sci U S A 104, 151–156.

Jankowsky, E., and Fairman, M.E. (2007). RNA helicases-one fold formany functions. Curr Opin Struct Biol 17, 316–324.

Law, L.M., Albin, O.R., Carroll, J.W., Jones, C.T., Rice, C.M., and

Macdonald, M.R. (2010). Identification of a dominant negativeinhibitor of human zinc finger antiviral protein reveals a functionalendogenous pool and critical homotypic interactions. J Virol 84,

4504–4512.

Liu, Q., Greimann, J.C., and Lima, C.D. (2006). Reconstitution,

activities, and structure of the eukaryotic RNA exosome. Cell 127,1223–1237.

Muller, S., Moller, P., Bick, M.J., Wurr, S., Becker, S., Gunther, S., andKummerer, B.M. (2007). Inhibition of filovirus replication by the zincfinger antiviral protein. J Virol 81, 2391–2400.

Wang, Y., and Bogenhagen, D.F. (2006). Human mitochondrial DNAnucleoids are linked to protein folding machinery and metabolicenzymes at the mitochondrial inner membrane. J Biol Chem 281,

25791–25802.

Zhang, Y., Burke, C.W., Ryman, K.D., and Klimstra, W.B. (2007).Identification and characterization of interferon-induced proteins

that inhibit alphavirus replication. J Virol 81, 11246–11255.

Zhou, Y., Ma, J., Bushan, R.B., Wu, J.Y., Pan, Q., Rong, L., and Liang,C. (2007). The packaging of human immunodeficiency virus type 1

RNA is restricted by overexpression of an RNA helicase DHX30.Virology 372, 97–106.

Zhu, Y., and Gao, G. (2008). ZAP-mediated mRNA degradation. RNABiol 5, 65–67.

Protein & Cell


Peiying Ye et al.

Copyright

Submission of a manuscript implies: that the workdescribed has not been published before (except in theform of an abstract or as part of a published lecture,review, or thesis); that it is not under consideration forpublication elsewhere; that its publication has beenapproved by all co-authors, if any, as well as – tacitly orexplicitly – by the responsible authorities at the institutionwhere the work was carried out. The author warrantsthat his/her contribution is original and that he/she hasfull power to make this grant. The author signs for andaccepts responsibility for releasing this material onbehalf of any and all co-authors. Transfer of copyrightto Higher Education Press and Springer becomeseffective if and when the article is accepted forpublication. After submission of the Copyright TransferStatement signed by the corresponding author, changesof authorship or in the order of the authors listed will notbe accepted by Higher Education Press and Springer.The copyright covers the exclusive right (for U.S.government employees: to the extent transferable) toreproduce and distribute the article, including reprints,translations, photographic reproductions, microform,electronic form (offline, online) or other reproductionsof similar nature.

An author may self-archive an author-created versionof his/her article on his/her own website and/or on his/her institutional repository. He/she may also deposit thisversion on his/her funder’s or funder’s designatedrepository at the funder’s request or as a result of alegal obligation, provided it is not made publicly availableuntil 12 months after official publication. He/she may notuse the publisher’s PDF version which is posted onwww.springerlink.com for the purpose of self-archivingor deposit. Furthermore, the author may only post his/her version provided acknowledgement is given to theoriginal source of publication and a link is inserted to the

published article on Springer’s website. The link must beaccompanied by the following text: “The originalpublication is available at springerlink.com.”

All articles published in this journal are protected bycopyright, which covers the exclusive rights to reproduceand distribute the article (e.g., as offprints), as well as alltranslation rights. No material published in this journalmay be reproduced photographically or stored onmicrofilm, in electronic data bases, video disks, etc.,without first obtaining written permission from thepublishers. The use of general descriptive names,trade names, trademarks, etc., in this publication, evenif not specifically identified, does not imply that thesenames are not protected by the relevant laws andregulations.

While the advice and information in this journal isbelieved to be true and accurate at the date of its goingto press, neither the authors, the editors, nor thepublishers can accept any legal responsibility for anyerrors or omissions that may be made. The publishermakes no warranty, express or implied, with respect tothe material contained herein.

Special regulations for photocopies in the USA:Photocopies may be made for personal or in-houseuse beyond the limitations stipulated under Section107 or 108 of U.S. Copyright Law, provided a fee ispaid. All fees should be paid to the CopyrightClearance Center, Inc., 222 Rosewood Drive, Danvers,MA 01923, USA, Tel: + 1-978-7508400, Fax: + 1-978-6468600, http://www.copyright.com, stating the ISSNof the journal, the volume, and the first and last pagenumbers of each article copied. The copyright owner’sconsent does not include copying for general distribu-tion, promotion, new works, or resale. In these cases,specific written permission must first be obtained fromthe publishers.

Protein & Cell Aims & ScopeProtein & Cell publishes original research articles, reviews, and commentaries concerning the latest developments in multidisciplinary areas in biology and biomedicine, with an emphasis on protein and cell research.

Our subject areas include, but are not limited to, biochemistry/biophysics, cell biology, developmental biology, genetics, immunology, microbiology,

molecular biology, neuroscience, oncology, protein science, structural biology and translational medicine. In addition, Protein & Cell addresses research highlights, news and views, and commentaries covering research policies and funding trends in China as well as in other countries, and seeks to provide a forum to foster academic exchange among researchers across different fields of the life sciences.

Supervised byMinistry of Education of the People’s Republic of China

Sponsored byBeijing Institutes of Life Science, Chinese Academy of Sciences Biophysical Society of China

Administered byBeijing Institutes of Life Science, Chinese Academy of SciencesHigher Education Press

Submission informationManuscripts should be submitted to: http://mc.manuscriptcentral.com/protein-cell [email protected]

Subscription information ISSN print edition: 1674-800X ISSN electronic edition: 1674-8018

Subscription rates:For information on subscription rates please contact: Customer [email protected] and South [email protected] North and South [email protected]

Orders and inquiries中国大陆地区高等教育出版社100029 北京市朝阳区惠新东街4号富盛大厦15层电话：+86-10-58556485http://journal.hep.com.cn

ChinaHigher Education Press4 Huixin Dongjie, Beijing 100029, ChinaTel: +86-10-58556485Fax: +86-10-58556034North and South AmericaSpringer New York, Inc.Journal FulfillmentP.O. Box 2485Secaucus, NJ 07096 USA

Tel: 1-800-SPRINGER or 1-201-348-4033Fax: 1-201-348-4505E-mail: [email protected]

Outside North and South AmericaSpringer Distribution Center Customer Service JournalsHaberstr. 769126 Heidelberg, GermanyTel: +49-6221-345-0, Fax: +49-6221-345-4229E-mail: [email protected]

Cancellations must be received by September 30 to take effect at the end of the same year.

Changes of address Allow for six weeks for all changes to become effective. All communications should include both old and new addresses (with postal codes) and should be accompanied by a mailing label from a recent issue. According to § 4 Sect. 3 of the German Postal Services Data Protection Regulations, if a subscriber’s address changes, the German Federal Post Office can inform the publisher of the new address even if the subscriber has not submitted a formal application for mail to be forwarded. Subscribers not in agreement with this procedure may send a written complaint to Customer Service Journals within 14 days of publication of this issue.Microform editions are available from: ProQuest. Further information available at http://www.il.proquest.com/uni

Electronic editionAn electronic version is available at springerlink.com

ProductionHigher Education Press4 Huixin Dongjie, Beijing 100029, China

Printed in People’s Republic of China

Jointly Published byHigher Education Press and Springer

Springer-Verlag is part of Springer Science + Business Media

Human catalase: looking for complete identity

Documents