Both sense and antisense strands of the LTR of the Schistosoma mansoni Pao-like retrotransposon Sinbad drive luciferase expression

Mol Genet Genomics

ORIGINAL PAPER

Both sense and antisense strands of the LTR of the Schistosoma mansoni Pao-like retrotransposon Sinbad drive luciferase expression

Claudia S. Copeland · Victoria H. Mann · Paul J. Brindley

Received: 20 August 2006 / Accepted: 4 October 2006© Springer-Verlag 2006

Abstract Long terminal repeat (LTR) retrotranspo-sons, mobile genetic elements comprising substantialproportions of many eukaryotic genomes, are sonamed for the presence of LTRs, direct repeats about250–600 bp in length Xanking the open reading framesthat encode the retrotransposon enzymes and struc-tural proteins. LTRs include promotor functions aswell as other roles in retrotransposition. LTR retro-transposons, including the Gypsy-like Boudicca andthe Pao/BEL-like Sinbad elements, comprise a sub-stantial proportion of the genome of the human bloodXuke, Schistosoma mansoni. In order to deduce thecapability of speciWc copies of Boudicca and SinbadLTRs to function as promotors, these LTRs wereinvestigated analytically and experimentally. Sequenceanalysis revealed the presence of TATA boxes, canoni-cal polyadenylation signals, and direct inverted repeatswithin the LTRs of both the Boudicca and Sinbad ret-rotransposons. Inserted in the reporter plasmid pGL3,the LTR of Sinbad drove WreXy luciferase activity inHeLa cells in its forward and inverted orientation. Incontrast, the LTR of Boudicca did not drive luciferase

activity in HeLa cells. The ability of the Sinbad LTR totranscribe in both its forward and inverted orientationrepresents one of few documented examples of bidirec-tional promotor function.

Keywords Schistosome · Boudicca · Promotor · Promoter · Bidirectional · Inverse

Introduction

Eukaryotic genomes generally contain substantialamounts of repetitive sequences, many of which repre-sent mobile genetic elements (e.g., Lander et al. 2001;Holt et al. 2002). These mobile elements have playedfundamental roles in host genome evolution (Charles-worth et al. 1994; Deininger and Batzer 2002).Although less is known about the genome of the platy-helminth parasite Schistosoma than more extensivelystudied genomes such as human and mouse, recentWndings suggest that up to half of this genome may becomprised of repetitive sequences, and much of thisrepetitive complement will consist of mobile geneticelements (see Brindley 2005). Within this populationof mobile elements, 15% or more of the schistosomegenome may be composed of long terminal repeat(LTR) retrotransposons (Copeland et al. 2005b).

Long terminal repeat retrotransposons (and retrovi-ruses) are mobile genetic elements that reproducethrough an RNA intermediate. Typically 5–10 kb inlength, their general structure consists of two openreading frames (ORFs) Xanked by long direct terminalrepeats of »200–600 bp in length. These LTRs play apivotal role in retrotransposition and transcription ini-tiation. The 5� LTR encodes the promotor, which

Communicated by M.-A. Grandbastien.

C. S. Copeland · V. H. Mann · P. J. BrindleyDepartment of Tropical Medicine, and Interdisciplinary Program in Molecular and Cellular Biology, Tulane University Health Sciences Center, 1430 Tulane Avenue, New Orleans, LA 70112, USA

C. S. Copeland (&)Center for Medical, Agricultural, and Veterinary Entomology, USDA Agricultural Research Service, 1700 SW 23rd Drive, Gainesville, FL 32608, USAe-mail: [email protected]

1 3

Mol Genet Genomics

directs transcription by host cell RNA polymerase II.The LTRs’ roles in retrotransposition include the nine-step process of reverse transcription (Gilboa et al.1979) and the integration of the proviral form of theretrotransposon or retrovirus into the host cell chro-mosomes (Sharon et al. 1994; Hindmarsh and Leis1999). LTRs are constituted of three main parts, theU3 unique region, the repeated (R) region, and the U5unique region (CoYn et al. 1997) (Fig. 1a). The U3region includes both promotor and enhancer elements,and is only found in the 3� LTR in the RNA genome ofthe retrotransposon, since transcription begins down-stream of the U3 in the 5� LTR. Transcription begins at(and deWnes) the 5� end of the R region, which isrepeated at both ends of the transcript. The U5 fromthe 3� LTR is replaced with a poly-A tail, so that in thetranscript, the U5 is only found in the 5� LTR (CoYnet al. 1997). The U5 region includes a GT-rich regionthat encodes control elements for 3� end processing.The R region is integral to the transfer of DNA synthe-sis between the 5� and 3� ends of the mRNA templateduring reverse transcription. The poly-A tail begins atthe end of the R region of the 3� LTR, and by doing sodeWnes the 3� boundary of the R region. The polyaden-ylation signal is located either in the 3� end of the U3region or in the R region. (Polyadenylation signalsfound outside the LTRs of retroviruses such as HTLV-1 are associated with reverse ORFs (Cavanagh et al.2006). LTR retrotransposons and simpler retrovirusescontain only forward ORFs.) Located at the distal mar-gin of many LTR retrotransposons are small repeats ofsequence TCN-NCA, termed the DinucleotideInverted Repeats (DIR) (Bowen and McDonald 1999).These repeats are the byproducts of the loss of twonucleotides (nt) in the initial nucleophilic attack whichbegins the process of integration of the retrotranspo-son into the genome of the host (Hindmarsh and Leis1999). Longer (at least »8–12 bp for retroviruses),often perfectly inverted terminal sequences (terminalinverted repeats, or TIRs) are required for recognitionby integrase during the process of integration into hostgenomic DNA (Katzman et al. 1989).

Three other key elements associated with the LTRare the primer binding site (PBS), polypurine tract(PPT), and encapsidation sequence, psi (�). The PBS,located immediately downstream of the 5� LTR, is thebinding site for the tRNA primer for negative strandDNA synthesis. The PPT is located at the other end ofthe retrotransposon, just upstream of the 3� LTR, andis involved in the second (plus strand) DNA strandsynthesis during reverse transcription. Downstream ofthe PBS and overlapping gag is the encapsidationsequence (�), also known as the packaging signal.

Retrotransposon LTRs and associated regulatory ele-ments can be considered just as critical as the codingsequences in that they can facilitate transposition ofthe retroelement, even when the coding sequence istoo degraded to encode functional proteins. Transposi-tion of degenerate retrotransposons incapacitated byinternal mutation(s) can be accomplished through theactivity of functional proteins encoded by non-mutatedcopies, a process known as retrotransposition in trans(Curcio and GarWnkel 1994).

The inXuence of LTRs extends beyond transcriptionand transposition of the retrotransposons and retrovi-ruses themselves. They have also been co-opted bytheir host genomes as promotors of native genes (Med-strand et al. 2001; Dunn et al. 2003). In some of thesecases, the LTRs serve as alternative promotors that areutilized in addition to the native promotors, and someof these initiate transcription in a tissue speciWc man-ner. For example, the human endothelin B receptorutilizes an LTR derived promotor in addition to itsnative promotor, but the LTR driven transcriptsappear only in placental tissue (Medstrand et al. 2001).Since expression of LTR retrotransposons has beenshown to increase in tissues related to reproduction(Dupressoir and Heidmann 1996; Lecher et al. 1997),this pattern of transcription by LTRs acting as promo-tors of native genes is a logical adaptation of functionby the host genome.

Schistosomiasis is considered the most important ofthe human helminthiases, and ranks second only tomalaria among all tropical diseases, in terms of mor-bidity and mortality (Chitsulo et al. 2004). Up to 15%of the schistosome genome may be composed of LTRretrotransposons (Copeland et al. 2005b). Understand-ing structural and functional aspects of these elementswill enhance the general understanding of the schisto-some genome. It is hoped that this will guide improve-ments in the control of schistosomiasis, including thedevelopment of vaccines and new antiparasite medica-tions. In addition to their natural roles as the promo-tors of LTR retrotransposons, LTRs are of interest fortheir potential in applications related to transgenesis(i.e., Tanaka et al. 1998). Because of their natural tar-get speciWcity, LTRs of native schistosome retrotrans-posons are of potential use as components oftransgenesis vectors to facilitate integration into theschistosome genome. Transduction of Schistosomamansoni using retroviral constructs has recently beendemonstrated (Kines et al. 2006). Stable integration oftransgenes into the genome would in turn enable theinvestigation of schistosome gene function and the devel-opment of transgenic schistosomes. The promotor func-tions of LTRs could also be exploited in transgenesis

1 3

Mol Genet Genomics

systems if the LTRs have been demonstrated to becapable of initiating transcription.

Boudicca and Sinbad are discrete LTR retrotranspo-sons recently characterized from the genome of S.mansoni (Copeland et al. 2003, 2005a). The 386 bplong LTRs of the copy of Sinbad characterized byCopeland et al. (2005a) were identical in sequence,implying recent insertion and competence to carry outtheir roles in transcription initiation and integration(Bowen and McDonald 1999). The LTRs of the copyof Boudicca characterized in Copeland et al. (2003)were 88.4% identical, with the 327 bp long 5� LTR rep-resenting an apparently more intact version, as com-pared with the consensus of several fragments of theLTRs of other copies of Boudicca (Copeland et al.2003). Here we investigated the structure of the LTRs

of Boudicca and Sinbad and their ability to drive WreXyluciferase expression in mammalian cells. We clonedboth forward and reverse versions of each LTR intothe luciferase expression vector pGL3, transformedHeLa cells with the constructs, and investigated thecapacity and performance of the LTRs of Boudiccaand Sinbad in driving luciferase expression in thesecells.

Materials and methods

Construction of LTR-luciferase expression vectors

The 5� LTRs of Boudicca and Sinbad were ampliWedfrom clones of a BAC library representing an eightfold

Fig. 1 Generalized structure of an LTR (a) and sequence alignments of 5� and 3� LTRs of (b) Sinbad and (c) Boudic-ca. Dinucleotide inverted repeats (DIRs) are marked with solid underline (red in online version). Seven nucleo-tide terminal inverted repeats in the Sinbad LTR are identi-Wed with double underline (orange in online version). Possible transcription initia-tion sites (TATA and CAAT sequences) are indicated with round dotted underline (green in online version). Polyadeny-lation signals (AATAAA or AATATA) are marked with square dotted underline (blue in online version). A GT rich region in Sinbad’s LTR is marked with dashed under-line (purple in online version)

1 3

Mol Genet Genomics

coverage of the S. mansoni genome in DH10B Escheri-chia coli cells (Le Paslier et al. 2000). BAC clone 53-J-5and 33-N-3 have been shown previously to contain cop-ies of Boudicca (Copeland et al. 2003) and Sinbad(Copeland et al. 2005a), respectively. Primers weredesigned from the S. mansoni genomic DNA justupstream of the 5� LTR of each retrotransposon (fwd)and the untranslated region just downstream of the 5�

LTR (rev). The 5� LTR of Boudicca was ampliWedfrom BAC 53-J-5 and the 5� LTR of Sinbad was ampli-Wed from BAC 33-N-3, using as the templates BACplasmid DNA isolated from transformed E. coli usingthe Phaseprep kit from Sigma (St. Louis, MO USA).DNA was allowed to denature at 95°C, followed by 30denaturation-annealing-extension cycles of 95°C(1 min), 56°C (1 min), and 72°C (1 min). Primersincluded restriction sites for Kpn 1 and Hind III toallow cloning of forward and reverse LTRs. Four setsof primers were designed; two with the Kpn 1 and HindIII sites positioned such that the LTR would be clonedinto an expression vector in the forward orientation,and two with restriction sites positioned such that theLTR would be cloned in the inverted orientation(Fig. 2a) Primer sequences are shown in Table 1.The PCR products were isolated using the Wizardsystem (Promega, Madison, WI, USA), after which

gel-puriWed PCR products and the promotorless lucif-erase expression vector pGL3-Basic (Promega) weredigested with Kpn I and Hind III. The ends of the line-arized pGL3 vector were dephosphorylated by incu-bating with Antarctic Phosphatase (New EnglandBioLabs, Beverly, MA, USA) for 30 min at 37°C fol-lowed by heat inactivation for 5 min at 65°C. Thedigested LTRs were ligated into the dephosphorylated,linearized pGL3 plasmid after which the ligation prod-ucts were employed to transform electrocompetent E.coli cells (EP-Max 10B Competent Cells, Bio-Rad,Hercules, CA, USA) by electroporation (Bio-RadGenePulser Xcell). Plasmids were puriWed from trans-formed cells and the sequence of the inserts deter-mined (Davis Sequencing, Davis, CA, USA) toconWrm identity and orientation of the retrotransposonLTR inserts.

Sequence analysis and alignment of LTR sequences

Nucleotide sequence data were analyzed using SeqMan(DNAstar Inc., Madison, WI, USA) and VecScreen(NCBI) software. The LTR sequences have beenassigned GenBank accession numbers; AY965072,AY965073, AY965074, and AY965075. Sequence align-ments were accomplished with ClustalW (Thompson

Fig. 2 a Luciferase expression vector constructs. Promotors(LTRs) were cloned into a multiple cloning site 5� of the lucifer-ase coding sequence. The pGL3 Basic construct does not have apromotor for the luciferase gene. b Mean luciferase activity, mea-sured in relative light units per mg total protein (RLU/mg pro-tein). From left to right, constructs measured are: pGL3-Basic,

the promotorless luciferase expression vector (negative control),pGL3 with the Boudicca LTR as promotor, pGL3 with an invert-ed Boudicca LTR as promotor, pGL3 with the Sinbad LTR aspromotor, pGL3 with an inverted Sinbad LTR as promotor,pGL3 with the S. mansoni glutathione-S-transferase promotor,GST28-A (positive control)

B

pGL3-basicBoudiccaBoudicca, inv.SinbadSinbad, inv.GST28-A

A

Lu

c+

Promotor

Lu

c+

Promotor

Lu

c+

Amp

F1 ori

Amp

Amp

pGL3-Basic(no promotor)

pGL3-Control (SV40)pGL3-Sinbad LTRpGL3-Boudicca LTRpGL3-GST-28A

pGL3-Sinbad LTR,Inverse

pGL3-Boudicca LTR,Inverse

F1 oriF1 ori

1 3

Mol Genet Genomics

et al. 1994) and MacVector software (Accelrys, SanDiego, CA, USA).

Transfection of HeLa cells

The HeLa cell line was purchased from ATCC(Manassas, VA, USA) and maintained as recom-mended by the vendor. HeLa cells were transfectedwith pGL3 plasmids using the TransFectin lipidreagent system (Bio-Rad). About 5 £ 105 cells wereseeded onto 60 mm diameter plates (triplicates ofexperimental groups, along with negative and positivecontrols) and cultured overnight in Dulbecco’s Modi-Wed Eagle’s Medium (DMEM) supplemented with 1%penicillin/Streptomycin and 10% fetal calf serum. Onemicrogram plasmid DNA of each pGL3 construct wasdiluted in 50 �l serum free DMEM medium and 15 �lTransFectin lipid reagent (Bio-Rad) was added to 35 �lserum free DMEM. The diluted DNA and diluted lipidreagent were then combined and incubated at roomtemperature (RT) for 20 min. The overnight culturemedium was removed from the cells and replaced with400 �l serum free medium and the diluted DNA-Trans-Fectin complexes (100 �l). Seven treatment groupswere included: Boudicca LTR; Boudicca LTR ininverse orientation; Sinbad LTR; Sinbad LTR ininverse orientation; S. mansoni GST28-A promoter;SV40-pGL3 control; and pGL3 Basic (Fig. 2a) [the S.mansoni GST28-A-pGL3 construct included a 1.3 kbfragment of the promoter region, exons 1 and 2, andintron 1 of the gene encoding glutathione-S-transferase28 kDa (GST28) of S. mansoni. This construct, kindlyprovided by Dr. Edward Pearce, University of Pennsyl-vania, was included here as a positive control promoterof an endogenous schistosome gene known to be capa-ble of driving luciferase expression in HeLa cells (P. J.Brindley et al., unpublished data)]. The cells were incu-bated at 37°C for 4 h in the presence of the transfectionreagents, with gentle rocking. Subsequently, 2.5 ml ofDMEM with 10% serum was added to each plate, after

which they were incubated for 24 h at 37°C under 5%CO2 in air. The assay was replicated four times.

Luciferase assay

Luciferase activity was determined using the Lucifer-ase Assay System from Promega. Media were removedfrom transfected cells, after which 400 �l of LuciferaseCell Culture Lysis Reagent (CCLR, Promega: 25 mMTris–phosphate (pH 7.8), 2 mM DTT, 2 mM 1,2-diami-nocyclohexane-N,N,N’,N’-tetraacetic acid, 10% glyc-erol, 1% Triton X-100) was added to each plate to lysethe HeLa cells. The lysate was removed and half of itwas used for the luciferase assay (the other half wasstored at ¡80°C for subsequent protein measurement).One hundred microliters of Luciferase Assay Reagent(Promega) (luciferin substrate in ATP/Mg2+ buVer)was added to 100 �l cell lysate immediately beforemeasuring the luciferase activity. Luciferase activitywas measured using a luminometer (Sirius, BertholdDetection Systems, Bad Wildbad, Germany), withunits of output (raw readings) in Relative Light Units(RLU). Two 100 �l samples from each plate (above)were analyzed, and the RLU results from these tworeadings were averaged to yield the raw RLU readingsfor each plate.

Measurement of total protein

Total protein per sample was measured using the BCAProtein Assay Reagent Kit (Pierce, Rockford, IL,USA). A standard curve was prepared by dilutingknown concentrations of bovine serum albumin (BSA)in CCLR. Two replicates of each standard and experi-mental sample were incubated in a cupric sulfate basedreagent provided in the BCA kit (0.08% cupric sulfate,0.1 M sodium hydroxide, and undisclosed proportionsof sodium carbonate, sodium bicarbonate, bicinchoni-nic acid and sodium tartrate) at 37°C for 30 min, andthen were cooled at RT. Absorbances at wavelength

Table 1 Primers used for PCR ampliWcation of forward and inverse LTRs of Boudicca and Sinbad

Retrotransposon LTR orientation

Primer Sequence

Boudicca Forward fwd 5�-CGGGGTACCTGTAGCTGTAAATAGTTCCCC-3�Boudicca Forward rev 5�-CCCAAGCTTTGAGGATACGACTAGTATTCTA-3�Boudicca Inverse fwd 5�-CGGGGTACCTGAGGATACGACTAGTATTCTA-3�Boudicca Inverse rev 5�-CCCAAGCTTTGTAGCTGTAAATAGTTCCCC-3�Sinbad Forward fwd 5�-CGGGGTACCAAAGACGCAAAGAGATGGTT-3�Sinbad Forward rev 5�-CCCAAGCTTATCCAGTGTGAACACAGAACC-3�Sinbad Inverse fwd 5�-CGGGGTACCATCCAGTGTGAACACAGAACC-3�Sinbad Inverse rev 5�-CCCAAGCTTAAAGACGCAAAGAGATGGTT-3�

1 3

Mol Genet Genomics

562 nm (OD592) were determined with a spectropho-tometer (Bio-Rad, SmartSpec 3000) within 10 min ofeach other. The average OD592 reading for the blankstandard (CCLR only) was subtracted from the aver-age reading for each sample to yield the individualabsorbance readings. The standard curve was used tocalculate total protein for each cell lysate sample. Indi-vidual luciferase RLU values for each sample werethen divided by total calculated protein for each sam-ple to yield RLU/mg protein for each sample. RLU/mgresults for the three samples of each experimentalgroup were determined along with standard error ofmean values.

Results

Structural analysis of retrotransposon LTRs

A BLAST search of the non-redundant nt databaserevealed a highly conserved region in the Sinbad LTRbetween nt positions »200 and 230. As this region con-tains a potential TATA box (Fig. 1b), this probablyrepresents the transcription start site, in the vicinity ofthe U3/R boundary (CoYn et al. 1997). This indicatedthat the lengths of the U3 and the R + U5 region of theSinbad LTR were each »200 bp. A BLAST searchwith the Boudicca LTR also revealed a conservedregion, between nt positions »170 and 200. This wasone of the predicted positions for a potential TATAbox for the LTR of Boudicca (Fig. 1c), and so this rep-resents the probable location for the transcription startsite for Boudicca. Further, this indicated that thelength of Boudicca’s U3 and R + U5 regions are»150 bp each in length. In addition to the polymerasebinding sites, canonical polyadenylation signals ofsequence AATAAA (CoYn et al. 1997) and the vari-ant AATATA were apparent in the LTRs of both ret-rotransposons.

The LTRs of both Boudicca and Sinbad exhibitedmany of the hallmark features of LTRs of retroviruses.The polyadenylation signal for the Boudicca LTR islocated about 10–40 bp downstream of the putativePBS (at nts 208–213) placing it either in the R region,as is seen in HIV-1 and other retroviruses, or at the 3�

end of the U3, as is seen in HTLV-1 and other retrovi-ruses (CoYn et al. 1997) (Fig. 1a, c). Sinbad’s poly-adenylation signal is located at 265–270 bp, about 40–70 bp downstream of the putative PBS, probably in theR region. In addition, the Sinbad LTR also includes aGT-rich region 12 bp downstream of the forward poly-adenylation signal (TTTGGAGTTTTGGTT, nts 283–297 of the LTR). The LTR of Boudicca does not

apparently include a similar GT-rich region (Fig. 1b, c).The LTRs of both retrotransposons include DIRs, butonly Sinbad’s LTR includes more extended invertedrepeats. The locations of these regulatory signals andthe segment lengths they imply are consistent with thestructures of the LTRs of many retroviruses (CoYnet al. 1997), but the Sinbad LTR includes more recog-nizable signals similar to those of retroviral LTRs.

Luciferase assay

Luciferase activity and total protein was measured forseven triplicate sets of transfected cells, in each of fourreplicates of the assay. These included triplicates foreach experimental condition, Boudicca LTR forward,Boudicca LTR inverted, Sinbad LTR forward, and Sin-bad LTR inverted, and for three controls (Fig. 2a). Allconstructs were sequenced prior to their use in theassay and were identical to the expected sequences, inGenBank accession numbers AY965072, AY965073,AY965074, and AY965075. The pGL3 Control SV40driven construct produced strong levels of luciferaseactivity (about 100 £ that of pGL3-Basic, as expectedbased on data from Promega) (not shown) while theGST28-A driven construct produced moderatelystrong activity (Fig. 2b). The pGL3-Basic constructproduced weak, though detectable, activity. In general,both the forward and inverted Boudicca LTR con-structs drove only weak or negligible luciferase activ-ity, with levels similar to those from the promotorless,pGL3 basic, control. In contrast, the Sinbad LTR for-ward construct drove luciferase activity at levels com-parable to those seen with the S. mansoni GST28-Aconstruct (Fig. 2b). In addition, surprisingly, the Sin-bad LTR inverted construct drove luciferase activity ata similar level to the Sinbad LTR forward construct(Fig. 2b). These Wndings showing luciferase activitywith both forward and inverse orientation versions ofthe Sinbad LTR construct were observed consistently,in each of the four replicates of the experiment (notshown).

Discussion

This study illustrated not only the ability of Sinbad’sLTR to create active transcripts, but also showed thatit is capable of functioning bidirectionally. The Boud-icca LTR, in contrast, did not appear to be functionalin HeLa cells, though it may be capable of transcrip-tion in diVerent systems (although luciferase activitydriven by the Boudicca LTR was signiWcantly greaterin the inverse versus forward orientation, the level of

1 3

Mol Genet Genomics

luciferase activity was low compared to that of the pos-itive control).

Transcription of LTR retrotransposons can be tissuespeciWc (Awasaki et al. 1996; Dupressoir and Heid-mann 1996; Lecher et al. 1997), and genomic Xankingsequence can have a major impact on retrotransposonpromotor activity (Lavie et al. 2006). Boudicca tran-scription could therefore be aVected by the absence ofthe natural cellular and genomic context of copies inte-grated in the schistosome genome. Since numeroustranscripts of both Boudicca and Sinbad have beendetected in both larval and adult stages of S. mansoni(Copeland et al. 2003, 2004, 2005a), it is likely thatLTRs of both retrotransposons are active within schis-tosome tissues. The Boudicca transcripts seen in theseprevious studies, however, may have been driven byLTRs of diVerent copies of the Boudicca retrotranspo-son. The structural analysis shown in Fig. 1 suggestedthat the copy of Boudicca examined here may notinclude fully functional LTRs. Alternatively, the fail-ure of the Boudicca LTR to drive luciferase activity inHeLa could indicate that either the LTR may not beactive in heterologous (mammalian) cells and/or that itcannot drive transcription from the WreXy luciferasegene. Interestingly, the LTR of the same copy of Boud-icca examined here can drive green Xuorescent protein(GFP) expression in adult schistosomes transfected byparticle bombardment (B. H. Kalinna, personal com-munication), a Wnding that suggests that the absence ofluciferase activity in HeLa cells could reXect tissuespeciWcity, perhaps schistosome or molluscan tissuespeciWcity, rather than simple lack of promotor func-tion [Boudicca-like retrotransposons may be present inthe genomes of some snail hosts of schistosomes, butdo not appear to be present in the human genome(Copeland et al. 2006)]. Further experiments usingwhole schistosomes or cell lines such as Bge, a linederived from embryonic Biomphalaria glabrata snailcells (Yoshino et al. 1998), may help to clarify whetherthe Boudicca LTR is capable of functioning in non-human tissues. Alternatively, expression may be con-struct-speciWc, with this LTR capable of driving GFPexpression but not luciferase expression. Despite thesuccess with GFP expression, the LTRs of the 53-J-5copy of Boudicca are only 88.4% identical. Althoughmulti-copy analysis of other Boudicca LTRs indicatesthat the 3� LTR is degraded and the 5� LTR (investi-gated here) is similar to the consensus (Copeland et al.2003), the more intact 5� LTR nonetheless does notrepresent the promotor of a recently integrated retro-transposon. In addition, the lack of terminal invertedrepeats beyond the DIRs and the lack of a recogniz-able GC-rich area downstream of the polyadenylation

signal may be due to mutations in the Boudicca 5� LTRaccumulated over time. The LTRs of other, morerecently integrated copies of Boudicca may be found tobe permissive enough to drive robust levels of expres-sion in mammalian and other non-schistosome cells.

In contrast to Boudicca, the Sinbad promotor exhib-ited levels of expression approaching those of the schis-tosome GST28-A promotor. The LTRs of this copy ofSinbad were 100% identical, suggesting recent inser-tion and lack of mutation. The present Wndings indicatethat the LTR of the Sinbad retrotransposon is func-tional as a promotor. Interestingly, retrotransposonsclosely related to Sinbad have been found in verte-brates (Wsh) as well as platyhelminths (Copeland et al.2005a), whereas retrotransposons closely related toBoudicca have only been found in platyhelminths (Baeet al. 2001; Bae and Kong 2003) and insects (Abe et al.2000; Tubio et al. 2005). This may reXect less stringenthost speciWcity in terms of transcription factors andother host speciWc elements on the part of Sinbad.

Whereas LTR driven expression of luciferase wasnot surprising, expression of luciferase driven by theinverted Sinbad LTR (antisense) was unanticipated.Analysis of the LTR showed that it was not palin-dromic and sequencing of the construct conWrmed thatthe insert represented a single LTR in the inverted ori-entation. While apparently uncommon, other promo-tors have been shown to function both in the forwardand inverse orientation, including the LTR of the IAPmurine endogenous retroviruses (Maksakova et al.2006). Some promotors, such as the Drosophila H2A-H2B spacer (Crayton et al. 2004), and the mammalianACACA and TADA2L shared gene promotor (Tra-vers et al. 2005), function bidirectionally in nature, ini-tiating transcription for coding sequences on eitherside. In addition, 3� LTRs have been shown to produceantisense transcripts encoding reverse ORFs in retrovi-ruses (Cavanagh et al. 2006). Bidirectional promotorshave also been produced synthetically in order toexpress two transgenes in a coordinated manner(Amendola et al. 2005). Promotors that function natu-rally only in the forward direction have also beenshown to exhibit some transcription in the reversedirection in vitro (e.g., Johnson and Friedmann 1990).

Although expression driven by the antisensesequence of the LTR of Sinbad may seem to suggest anatural function for this phenomenon, how this mightconfer an increase in Wtness on an LTR retrotranspo-son remains unclear (no obvious reverse orientationORFs extend from the 3� LTR of Sinbad; all reverseorientation ORFs are short, and are not located adja-cent to the 3� LTR). Although the co-evolution of amulti-copy mobile element with a Xanking host gene

1 3

Mol Genet Genomics

seems unlikely, L1 elements encode antisense promo-tors in their 5� UTRs that can drive the transcription ofadjacent genes (Speek 2001), and as mentioned abovethe LTR of IAP can drive ectopic gene expression inan antisense direction. Perhaps the LTR of Sinbad isderived from the bidirectional promotor of a host genepair, or from the LTR of a retrovirus encoding reversereading frames; LTR retrotransposons are thought tohave evolved through the acquisition of both genes andLTRs from exogenous sources (Xiong and Eickbush1990; McClure 1991; Malik et al. 2000).

Functionality of the antisense sequence of the SinbadLTR could be due to a relatively recent acquisition of abidirectional promotor, or could be rooted in some asyet unknown function for the retrotransposon. Func-tional studies clarifying the deWnitive polyadenylationsignals, as well as the deWnitive transcription initiationsites, would be useful in investigating these possibilities.In addition, investigation of the activity of the LTRs ofother Pao/BEL elements might clarify whether bidirec-tional LTRs are unique to Sinbad or are characteristicof Pao/BEL clade retrotransposons at large.

The ability of retrotransposon and retroviral promo-tors to function bidirectionally has important implica-tions for genome function and evolution. This abilityexpands the known roles of retrotransposons as inser-tional mutagens and generators of transcripts 3� of the3� LTR to now include the possibility of generation oftranscripts 5� of the 5� LTR. Promotors capable offunctioning bidirectionally should thus be capable ofaVecting upstream genes or regulatory regions in addi-tion to those downstream, enabling LTR retrotranspo-sons with these promotors to transcribe geneticmaterial on either side of their insertion sites. Theinvestigation of the promotor function of other LTRs,as well as analyses of insertion sites of genes or regula-tory elements upstream of bidirectionally functionalLTRs, can be expected to shed light on how bidirec-tional LTRs may inXuence genome function andgenome evolution.

Acknowledgments We thank Dr. Edward Pearce for the provi-sion of the GST28-A-pGL3 construct. We thank Tulane Univer-sity for Dissertation Fellowship support for CSC. PJB is arecipient of a Burroughs Welcome Fund scholar award in Molec-ular Parasitology.

References

Abe H, Ohbayashi F, Shimada T, Sugasaki T, Kawai S, Mita K,Oshiki T (2000) Molecular structure of a novel gypsy-Ty3-like retrotransposon (Kabuki) and nested retrotransposableelements on the W chromosome of the silkworm Bombyxmori. Mol Gen Genet 263:916–924

Amendola M, Venneri MA, BiY A, Vigna E, Naldini L (2005)Coordinate dual-gene transgenesis by lentiviral vectors car-rying synthetic bidirectional promoters. Nat Biotechnol23:108–116

Awasaki T, Juni N, Yoshida KM (1996) An eye imaginal disc-spe-ciWc transcriptional enhancer in the long terminal repeat ofthe tom retrotransposon is responsible for eye morphologymutations of Drosophila ananassae. Mol Gen Genet251:161–166

Bae YA, Kong Y (2003) Divergent long-terminal-repeat retro-transposon families in the genome of Paragonimus wester-mani. Korean J Parasitol 41:221–231

Bae YA, Moon SY, Kong Y, Cho SY, Rhyu MG (2001) CsRn1, anovel active retrotransposon in a parasitic trematode, Clonor-chis sinensis, discloses a new phylogenetic clade of Ty3/gypsy-like LTR retrotransposons. Mol Biol Evol 18:1474–1483

Bowen NJ, McDonald JF (1999) Genomic analysis of Caenor-habditis elegans reveals ancient families of retrovirus-likeelements. Genome Res 9:924–935

Brindley PJ (2005) The molecular biology of schistosomes.Trends Parasitol 21:533–536

Cavanagh MH, Landry S, Audet B, Arpin-Andre C, Hivin P,Pare ME, Thete J, Wattel E, Marriott SJ, Mesnard JM, Bar-beau B (2006) HTLV-I antisense transcripts initiating in the3�LTR are alternatively spliced and polyadenylated. Retro-virology 3:15

Charlesworth B, Sniegowski P, Stephan W (1994) The evolution-ary dynamics of repetitive DNA in eukaryotes. Nature371(6494):215–220

Chitsulo L, Loverde P, Engels D (2004) Schistosomiasis. Nat RevMicrobiol 2:12–13

CoYn JM, Hughes SH, Varmus HE (1997) Retroviruses. ColdSpring Harbor Laboratory Press, Plainview, NY

Copeland CS, Brindley PJ, Heyers O, Michael SF, Johnston DA,Williams DJ, Ivens A, Kalinna BH (2003) Boudicca, a retro-virus-like, LTR retrotransposon from the genome of the hu-man blood Xuke, Schistosoma mansoni. J Virol 77:6153–6166

Copeland CS, Heyers O, Kalinna BH, Bachmair A, Stadler PF,Hofacker IL, Brindley PJ (2004) Structural and evolutionaryanalysis of the transcribed sequence of Boudicca, a Schisto-soma mansoni retrotransposon. Gene 329:103–114

Copeland CS, Laha T, Brindley PJ (2005b) Schistosome Long ter-minal repeat retrotransposons. In: Brindley PJ (ed) Mobilegenetic elements in metazoan parasites. Eurekah Press/Lan-des Bioscience, Georgetown, TX, pp315–323

Copeland CS, Lewis FA, Brindley PJ (2006) IdentiWcation of theBoudicca and Sinbad retrotransposons in the genome of thehuman blood Xuke Schistosoma haematobium. Mem InstOsv Cruz 101(5):565–571

Copeland CS, Mann VH, Morales ME, Kalinna BH, Brindley PJ(2005a) The Sinbad retrotransposon from the genome of thehuman blood Xuke, Schistosoma mansoni, and the distribu-tion of related Pao-like elements. BMC Evol Biol 5:20

Crayton ME, Ladd CE, Sommer M, Hampikian G, StrausbaughLD (2004) An organizational model of transcription factorbinding sites for a histone promoter in D. melanogaster. InSilico Biol 4:537–548

Curcio MJ, GarWnkel DJ (1994) Heterogeneous functional Ty1elements are abundant in the Saccharomyces cerevisiaegenome. Genetics 136:1245–1259

Deininger PL, Batzer MA (2002) Mammalian retroelements. Ge-nome Res 12(10):1455–1465

Dunn CA, Medstrand P, Mager DL (2003) An endogenous retro-viral long terminal repeat is the dominant promoter for hu-man beta1,3-galactosyltransferase 5 in the colon. Proc NatlAcad Sci USA 100:12841–12846

1 3

Mol Genet Genomics

Dupressoir A, Heidmann T (1996) Germ line-speciWc expressionof intracisternal A-particle retrotransposons in transgenicmice. Mol Cell Biol 16:4495–4503

Gilboa E, Mitra SW, GoV S, Baltimore D (1979) A detailed mod-el of reverse transcription and tests of crucial aspects. Cell18:93–100

Hindmarsh P, Leis J (1999) Retroviral DNA integration. Micro-biol Mol Biol Rev 63:836–843

Holt RA, Subramanian GM, Halpern A, Sutton GG, Charlab R,Nusskern DR, Wincker P, Clark AG, Ribeiro JM, Wides R,Salzberg SL, Loftus B, Yandell M, Majoros WH, Rusch DB,Lai Z, Kraft CL, Abril JF, Anthouard V, Arensburger P,Atkinson PW, Baden H, de Berardinis V, Baldwin D, BenesV, Biedler J, Blass C, Bolanos R, Boscus D, Barnstead M,Cai S, Center A, Chaturverdi K, Christophides GK, ChrystalMA, Clamp M, Cravchik A, Curwen V, Dana A, Delcher A,Dew I, Evans CA, Flanigan M, Grundschober-Freimoser A,Friedli L, Gu Z, Guan P, Guigo R, Hillenmeyer ME, HladunSL, Hogan JR, Hong YS, Hoover J, Jaillon O, Ke Z, KodiraC, Kokoza E, Koutsos A, Letunic I, Levitsky A, Liang Y, LinJJ, Lobo NF, Lopez JR, Malek JA, McIntosh TC, Meister S,Miller J, Mobarry C, Mongin E, Murphy SD, O’Brochta DA,Pfannkoch C, Qi R, Regier MA, Remington K, Shao H, Sha-rakhova MV, Sitter CD, Shetty J, Smith TJ, Strong R, Sun J,Thomasova D, Ton LQ, Topalis P, Tu Z, Unger MF, WalenzB, Wang A, Wang J, Wang M, Wang X, Woodford KJ,Wortman JR, Wu M, Yao A, Zdobnov EM, Zhang H,Zhao Q, Zhao S, Zhu SC, Zhimulev I, Coluzzi M, dellaTorre A, Roth CW, Louis C, Kalush F, Mural RJ, MyersEW, Adams MD, Smith HO, Broder S, Gardner MJ,Fraser CM, Birney E, Bork P, Brey PT, Venter JC, Weiss-enbach J, Kafatos FC, Collins FH, HoVman SL (2002) Thegenome sequence of the malaria mosquito Anophelesgambiae. Science 298(5591):129–149

Johnson P, Friedmann T (1990) Limited bidirectional activity oftwo housekeeping gene promoters: human HPRT and PGK.Gene 88:207–213

Katzman M, Katz RA, Skalka AM, Leis J (1989) The avian retro-viral integration protein cleaves the terminal sequences oflinear viral DNA at the in vivo sites of integration. J Virol63:5319–5327

Kines KJ, Mann VH, Morales ME, Shelby BD, Kalinna BH, Go-bert GN, Chirgwin SR, Brindley PJ (2006) Transduction ofSchistosoma mansoni by vesicular stomatitis virus glycopro-tein-pseudotyped Moloney murine leukemia retrovirus. ExpParasitol 112:209–220

Lander ES, Linton LM, Birren B, Nusbaum C, Zody MC, Bald-win J, Devon K, Dewar K, Doyle M, FitzHugh W, Funke R,Gage D, Harris K, Heaford A, Howland J, Kann L, Le-hoczky J, LeVine R, McEwan P, McKernan K, Meldrim J,Mesirov JP, Miranda C, Morris W, Naylor J, Raymond C,Rosetti M, Santos R, Sheridan A, Sougnez C, Stange–Thomann N, Stojanovic N, Subramanian A, Wyman D, Rog-ers J, Sulston J, Ainscough R, Beck S, Bentley D, Burton J,Clee C, Carter N, Coulson A, Deadman R, Deloukas P,Dunham A, Dunham I, Durbin R, French L, Grafham D,Gregory S, Hubbard T, Humphray S, Hunt A, Jones M,Lloyd C, McMurray A, Matthews L, Mercer S, Milne S,Mullikin JC, Mungall A, Plumb R, Ross M, Shownkeen R,Sims S, Waterston RH, Wilson RK, Hillier LW, McPhersonJD, Marra MA, Mardis ER, Fulton LA, Chinwalla AT, Pe-pin KH, Gish WR, Chissoe SL, Wendl MC, Delehaunty KD,Miner TL, Delehaunty A, Kramer JB, Cook LL, Fulton RS,Johnson DL, Minx PJ, Clifton SW, Hawkins T, BranscombE, Predki P, Richardson P, Wenning S, Slezak T, Doggett N,Cheng JF, Olsen A, Lucas S, Elkin C, Uberbacher E, Frazier

M, Gibbs RA, Muzny DM, Scherer SE, Bouck JB, SodergrenEJ, Worley KC, Rives CM, Gorrell JH, Metzker ML, NaylorSL, Kucherlapati RS, Nelson DL, Weinstock GM, Sakaki Y,Fujiyama A, Hattori M, Yada T, Toyoda A, Itoh T, KawagoeC, Watanabe H, Totoki Y, Taylor T, Weissenbach J, HeiligR, Saurin W, Artiguenave F, Brottier P, Bruls T, Pelletier E,Robert C, Wincker P, Smith DR, Doucette-Stamm L, Ru-benWeld M, Weinstock K, Lee HM, Dubois J, Rosenthal A,Platzer M, Nyakatura G, Taudien S, Rump A, Yang H, Yu J,Wang J, Huang G, Gu J, Hood L, Rowen L, Madan A, QinS, Davis RW, Federspiel NA, Abola AP, Proctor MJ, MyersRM, Schmutz J, Dickson M, Grimwood J, Cox R, Olson MV,Kaul R, Raymond C, Shimizu N, Kawasaki K, Minoshima S,Evans GA, Athanasiou M, Schultz R, Roe BA, Chen F, PanH, Ramser J, Lehrach H, Reinhardt R, McCombie WR, dela Bastide M, Dedhia N, Blocker H, Hornischer K, NordsiekG, Agarwala R, Aravind L, Bailey JA, Bateman A, Batzog-lou S, Birney E, Bork P, Brown DG, Burge CB, Cerutti L,Chen HC, Church D, Clamp M, Copley RR, Doerks T, EddySR, Eichler EE, Furey TS, Galagan J, Gilbert JG, HarmonC, Hayashizaki Y, Haussler D, Hermjakob H, Hokamp K,Jang W, Johnson LS, Jones TA, Kasif S, Kaspryzk A, Ken-nedy S, Kent WJ, Kitts P, Koonin EV, Korf I, Kulp D, LancetD, Lowe TM, McLysaght A, Mikkelsen T, Moran JV, Mul-der N, Pollara VJ, Ponting CP, Schuler G, Schultz J, Slater G,Smit AF, Stupka E, Szustakowski J, Thierry-Mieg D, Thi-erry-Mieg J, Wagner L, Wallis J, Wheeler R, Williams A,Wolf YI, Wolfe KH, Yang SP, Yeh RF, Collins F, Guyer MS,Peterson J, Felsenfeld A, Wetterstrand KA, Patrinos A,Morgan MJ, de Jong P, Catanese JJ, Osoegawa K, ShizuyaH, Choi S, Chen YJ, International Human GenomeSequencing Consortium (2001) Initial sequencing and analy-sis of the human genome. Nature 409(6822):860–921

Lavie L, Maldener E, Brouha B, Meese EU, Mayer J (2006) Thehuman L1 promoter: variable transcription initiation sitesand a major impact of upstream Xanking sequence on pro-moter activity. Genome Res 14(11):2253–2260

Le Paslier MC, Pierce RJ, Merlin F, Hirai H, Wu W, WilliamsDL, Johnston D, LoVerde PT, Le Paslier D (2000) Construc-tion and characterization of a Schistosoma mansoni bacterialartiWcial chromosome library. Genomics 65:87–94

Lecher P, Bucheton A, Pelisson A (1997) Expression of the Dro-sophila retrovirus gypsy as ultrastructurally detectable parti-cles in the ovaries of Xies carrying a permissive Xamencoallele. J Gen Virol 78:2379–2388

Maksakova IA, Romanish MT, Gagnier L, Dunn CA, van deLagemaat LN, Mager DL (2006) Retroviral elements andtheir hosts: insertional mutagenesis in the mouse germ line.PLoS Genet 2:e2

Malik HS, HenikoV S, Eickbush TH (2000) Poised for contagion:evolutionary origins of the infectious abilities of invertebrateretroviruses. Genome Res 10:1307–1318

McClure MA (1991) Evolution of retroposons by acquisition ordeletion of retrovirus-like genes. Mol Biol Evol 8:835–856

Medstrand P, Landry JR, Mager DL (2001) Long terminal re-peats are used as alternative promoters for the endothelin Breceptor and apolipoprotein C-I genes in humans. J BiolChem 276:1896–1903

Sharon G, Burkett TJ, GarWnkel DJ (1994) EYcient homologousrecombination of Ty1 element cDNA when integration isblocked. Mol Cell Biol 14:6540–6551

Speek M (2001) Antisense promoter of human L1 retrotranspo-son drives transcription of adjacent cellular genes. Mol CellBiol 21:1973–1985

Tanaka AS, Tanaka M, Komuro K (1998) A highly eYcient meth-od for the site-speciWc integration of transfected plasmids

1 3

Mol Genet Genomics

into the genome of mammalian cells using puriWed retroviralintegrase. Gene 216:67–76

Thompson JD, Higgins DG, Gibson TJ (1994) CLUSTAL W:improving the sensitivity of progressive multiple sequencealignment through sequence weighting, position-speciWc gappenalties and weight matrix choice. Nucleic Acids Res22:4673–4680

Travers MT, Cambot M, Kennedy HT, Lenoir GM, Barber MC,Joulin V (2005) Asymmetric expression of transcripts de-rived from the shared promoter between the divergently ori-ented ACACA and TADA2L genes. Genomics 85:71–84

Tubio JMC, Naveira H, Costas J (2005) Structural and evolution-ary analyses of the Ty3/gypsy group of LTR retrotranspo-sons in the genome of Anopheles gambiae. Mol Biol Evol22:29–39

Xiong Y, Eickbush T (1990) Origin and evolution of retroele-ments based upon their reverse transcriptase sequences.EMBO J 9:3353–3362

Yoshino TP, Wu XJ, Liu HD (1998) Transfection and heat-induc-ible expression of molluscan promoter-luciferase reportergene constructs in the Biomphalaria glabrata embryonicsnail cell line. Am J Trop Med Hyg 59:414–420

1 3