Cancer stem cell-like SP cells have a high adhesion ability to the peritoneum in gastric carcinoma

doi: 10.1111/j.1349-7006.2009.01211.x Cancer Sci | August 2009 | vol. 100 | no. 8 | 1397–1402© 2009 Japanese Cancer Association

Blackwell Publishing Asia

Cancer stem cell-like SP cells have a high adhesion ability to the peritoneum in gastric carcinomaTakafumi Nishii,1 Masakazu Yashiro,1,2,3 Osamu Shinto,1 Tetsuji Sawada,1 Masaichi Ohira1 and Kosei Hirakawa1

1Department of Surgical Oncology, 2Oncology Institute of Geriatrics and Medical Science, Osaka City University Graduate School of Medicine, Abeno-ku, Osaka, Japan

(Received December 23, 2008/Revised March 30, 2009/Accepted April 19, 2009/Online publication May 31, 2009)

Cancer stem cells (CSCs) are considered to be responsible for cancermetastasis, but the evidence to conclusively prove this hypothesisremains uncertain. The side population (SP), as evaluated by a flowcytometric analysis using Hoechst 33342, has been known as CSC-rich population. The aim of this study was to clarify thecharacterization of the SP cells in peritoneal metastasis of gastriccarcinoma. Gastric cancer cell lines OCUM-2M, OCUM-2D, andOCUM-2MD3 (a daughter cell line with high potential for peritonealmetastasis) were used. We isolated SP cells from OCUM-2M andOCUM-2D using flow cytometry. Serial sorting was performed threetimes to enrich SP cells, and they were designated as OCUM-2M/SPand OCUM-2D/SP cells. Flow cytometric analysis showed 0.46%,0.29%, 5.24%, 6.49%, and 11.3% of the SP cells to be found inOCUM-2M, OCUM-2D, OCUM-2MD3, OCUM-2M/SP, and OCUM-2D/SP cells, respectively. The intraperitoneal inoculation of SP cells andOCUM-2MD3 cells produced peritoneal metastasis, but parent cellsdid not. The adhesion ability of SP and OCUM-2MD3 cells wassignificantly high in comparison to that of parent cells. Theexpression level of adhesion molecules a2-, a5-, b3-, and b5-integrin,and CD44, was high in SP cells compared to parent cells. Theexpression of stemness markers, Oct3/4 and Sox2, increased in theSP-cell-injected tumors. These findings suggested that CSC-like SPcells expressing a2-, a5-, b3-, and b5-integrin, and CD44, may playan important role for peritoneal metastasis in gastric carcinoma.Oct3/4 and Sox2 may be associated with CSC in gastric cancer.(Cancer Sci 2009; 100: 1397–1402)

Recently, the concept of cancer stem cells (CSCs) has beenreported in the solid tumors of a wide variety of organs,

including the breast,(1) brain,(2) liver,(3) and gastrointestinaltumor.(4,5) CSCs are defined as a unique subpopulation in thetumors that possess the ability to initiate tumor growth andsustain self-renewal,(6) which explains the tumorigenicity andthe maintenance of bulk. Since gastric carcinoma shows ahistologic heterogeneity,(7) multipotent CSCs may explain thehistologic heterogeneity found in gastric tumors.(8) Recently,some human primary cancers and cell lines have been shown topossess side population (SP) cells that have been described as aCSC-rich population.(9) SP cells may provide a useful tool andmay be a readily accessible source for stem cell studies in boththe normal and cancerous settings.

CSCs have been proposed to play an important role in themetastasis of cancer cells,(10) although the evidence supportingthis hypothesis remains ambiguous. Gastric cancer frequentlydevelops into peritoneal metastasis, resulting in a poor prognosis.Although CSCs may be responsible for the peritoneal metastasisof gastric cancer, no study of peritoneal metastasis of CSCs hasbeen reported in gastric carcinoma. Peritoneal metastasis is theresult of a multistep phenomenon, including the detachment ofcancer cells from the primary tumor, freeing of the cells fromthe peritoneal cavity, and subsequent attachment to the perito-neum.(11) We previously reported that adhesion molecules, such

as integrins and CD44, were associated with a high potential forperitoneal metastasis.(11–13) In addition, Oct3/4 and Sox2 weresuggested to be stemness factors that render the reprogrammingcapability of adult cells into germ-line-competent-inducedpluripotent stem cells.(14–16) The aim of this study was to clarifythe characterization of peritoneal metastasis of the CSC-like SPcells in gastric carcinoma.

Materials and Methods

Cell culture. The human gastric cancer cell lines OCUM-2M,(17)

OCUM-2D,(18) and OCUM-2MD3(19) were derived from a diffusetype of gastric carcinoma. OCUM-2MD3 cells, a daughter cellline with high potential for peritoneal metastasis, were establishedfrom OCUM-2M cells, a parental cell line with poor metastaticpotential.(19) These cells were cultured in medium which consistedof Dulbecco’s modified Eagle medium (DMEM; Nikken Bio.,Kyoto, Japan) with the addition of 10% fetal bovine serum(FBS; Life Technologies, Grand Island, NY, USA), 100 IU/mLpenicillin (ICN Biomedicals, Costa Mesa, CA, USA), 100 μg/mLstreptomycin (ICN Biomedicals), and 0.5 mM sodium pyruvate(Cambrex, Walkersville, MD, USA), and they were incubated at37°C in an atmosphere containing 5% CO2.

Side population analysis using flow cytometry. The cells weresuspended at 1 × 106 cells/mL in PBS supplemented with 4%FBS. These cells were then incubated at 37°C for 60 min with5 μg/mL Hoechst 33342 (Sigma Chemicals, St Louis, MO,USA), either alone or in the presence of 500 μmol/L verapamil(Sigma). After incubation, 1 μg/mL propidium iodide was addedand then filtered through a 40-μm cell strainer (Becton Dickinson,San Diego, CA, USA) to obtain single-suspension cells.Analyses and sorting were performed using BD LSR (B.D.) orFACS Vantage SE (Becton Dickinson). Hoechst 33342 wasexcited with the UV laser at 350 nm and fluorescence emissionwas measured with 405/BP30 (Hoechst blue) and 570/BP20(Hoechst red) optical filters.

Animal models. Female BALB/c nude mice (Nihon CLEA,Tokyo, Japan) were used for the in vivo study. They were housedin specific pathogen-free conditions and fed standard chowpellets and water ad libitum. All experiments were performedaccording to the standard guidelines for animal experimentsof Osaka City University Medical School. We examined thetumorigenicity in each cell line by intraperitoneal or subcutaneousinoculation. OCUM-2M, OCUM-2D, OCUM-2M/SP, OCUM-2D/SP, and OCUM-2MD3 cells in various numbers wereinjected into the abdominal cavity or subcutaneously into theback area of the mice, and monitored for tumor development.

3To whom correspondence should be addressed. E-mail: [email protected]

1398 doi: 10.1111/j.1349-7006.2009.01211.x© 2009 Japanese Cancer Association

Adhesion assay. The adhesion ability of cancer cells toperitoneal components was examined using the adhesion assay.The binding of cancer cells to peritoneal components, fibronectin(Mallinckrodt Specialty Chemicals, St. Louis, MO, USA),matrigel (Collaborative Research, Bedford, MA, USA), andlaminin (Mallinckrodt Specialty Chemicals), were investigatedas follows. Briefly, a 96-well microtiter plate was coated withfibronectin (4 μg/well), matrigel (2 μg/well), or laminin (4 μg/well). Gastric cancer cells and SP cells (4 × 105 cells) wereseeded onto these components in 96-well microtiter plate(Falcon, Franklin Lakes, NJ). They were allowed to adhere toeach well for 30 min at 37°C and then were gently washed twicein PBS. The adhesion cancer cells were quantified by the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-tetrazolium bromide (MTT;Sigma) colorimetric assay. The percentage of cells adhering wascalculated as follows: % bindings = (OD of treated surface –only ECM component)/OD of total surface × 100. Total surfacemeans total cancer cells, 4 × 105 cells, seeded on the microtiterplates with matrigel, fibronectin, and laminin.

Real-time reverse transcription-PCR. mRNA expression of CD44,the integrin family, Sox2, Oct3/4, and CD133 was examined asfollows. Total cellular RNA was extracted from these cells withTrizol (Life Technologies) according to the manufacturer’sprotocol. cDNAs were synthesized from 1 μg of the RNA witha Moloney murine leukemia virus-reverse transcription kit (LifeTechnologies) using random hexamers. To determine foldchanges in each gene, real-time reverse transcription-PCR wasdone on the ABI Prism 7000 (Applied Biosystems, Foster City,CA, USA) using a commercially available gene expressionassay for α1-integrin, α2-integrin, α3-integrin, α5-integrin,β1-integrin, β3-integrin, β5-integrin, CD44, Oct3/4, Sox2, andCD133 (Hs00235030, Hs00158148, Hs00233722, Hs00233732,Hs01127543, Hs00173978, Hs00609896, Hs00174139,Hs00999632, Hs00415716, and Hs01009250, respectively).PCR was carried out at 95°C for 15 s and 60°C for 60 s for 40cycles. As an internal standard to normalize mRNA levels fordifferences in sample concentration and loading, amplificationof glyceraldehyde-3-phosphate dehydrogenase (GAPDH) wasused. The threshold cycle values were used to calculate therelative expression ratios between control and treated cells usingthe formula described by Pfaffl.(20)

Immunohistochemical techniques. Hematoxylin–eosin stainingand the immunohistochemical determination of Sox2, Oct3/4,and CD133 were examined using subcutaneous tumors. Briefly,paraffin-embedded sections were cut at 4 μm. Followingdeparaffinization and heating in Target Retrieval Solution(Dako, Carpinteria, CA, USA), the sections were incubated witha 1:50 dilution of a Sox2 antibody (MAB2018; R&D Systems,Minneapolis, MN, USA), with a 1:300 dilution of an Oct3/4

antibody (sc-5279; Santa Cruz Biotechnology, CA, USA), witha CD133 antibody (Miltenyi Biotec, Auburn, CA, USA), andwere treated with a secondary antibody. The mean percentage ofpositive tumor cells for the expression of Oct3/4, Sox2, andCD133 was determined at 400-fold magnification. Stainingintensity was marked by ‘+++’, ‘++’, ‘+’, and ‘–’, followed bythe percentage of positive tissue samples. Further, cases withmore than 50% positively stained tumor cells were defined as‘+++’, with more than 20% and less than 50% were defined as‘++’, with few cells were defined as ‘+’, and with no positivecells were defined as ‘–’.

Statistical analysis. Comparisons among the data sets weremade by Student’s t-test or Fisher’s exact test. Differences wereconsidered to be statistically significant when the P-value was0.05 or less.

Results

Side population (SP) fraction in gastric cancer cell lines. OCUM-2M, OCUM-2D, and OCUM-2MD3 cells were labeled withHoechst 33342 and analyzed by flow cytometry. Figure 1a is arepresentative picture of flow cytometric analysis. The SP cells,which disappear in the presence of verapamil (bottom panel),are outlined and shown as a percentage of the total cellpopulation. OCUM-2M and OCUM-2D cells sorted as SP cellswere cultured for 7–10 days and re-sorted by flow cytometry,recovered for an additional 7–10 days, and then re-analyzed byflow cytometry. Two cell lines were therefore re-sorted threetimes, and then designated as OCUM-2M/SP and OCUM-2D/SP cells. Each successive sort demonstrated the enrichment ofSP cells. The average SP percentages of OCUM-2M, OCUM-2M/SP, OCUM-2D, OCUM-2D/SP, and OCUM-2MD3 were0.46%, 6.49%, 0.29%, 11.3%, and 5.24%, respectively (Fig. 1b).The percentage of SP cells in OCUM-2M/SP and OCUM-2MD3 was significantly higher compared with that in OCUM-2M, and the percentage of SP cells in OCUM-2D/SP wassignificantly higher compared with that in OCUM-2D.

Tumorigenicity of SP cells in vivo. Tumorigenicity was examinedby the intraperitoneal inoculation or subcutaneous inoculation ofcancer cells (Table 1). The intraperitoneal inoculation of 5 × 105,1 × 106, and 5 × 106 OCUM-2M/SP cells resulted in peritonealmetastasis in one (20%), two (40%), and five (100%) of fivetreated mice, respectively. In OCUM-2MD3 cells, the intraper-itoneal inoculation of 5 × 105, 1 × 106, and 5 × 106 cells resultedperitoneal metastasis in none of five mice, two (25%) of eightmice, and 11 (100%) of 11 mice, respectively. In OCUM-2D/SPcells, the intraperitoneal inoculation of 1 × 106 and 5 × 106 cellsresulted peritoneal metastasis in none of four mice and three(75%) of four mice, respectively. In contrast, peritoneal metastasis

Table 1. Tumorigenicity of cancer cells

Cell line

Number of cancer cells

Intraperitoneal inoculation Subcutaneous inoculation

5 × 105 1 × 106 5 × 106 2 × 107 5 × 107 1 × 104 5 × 104 1 × 105 1 × 106 5 × 106

OCUM-2M 0/6† 0/14 0/16 0/10 0/4 0/4 0/6 0/9 2/10 3/14OCUM-2M/SP 1/5 2/5 5/5** ND‡ ND 0/5 4/5* ND ND NDOCUM-2MD3 0/5 2/8 11/11** 4/4 4/4 ND ND ND ND NDOCUM-2D ND 0/10 0/10 ND ND ND ND 0/8 3/8 10/10OCUM-2D/SP ND 0/4 3/4** ND ND ND ND 3/11 8/8** ND

The tumor growth was determined at 8 weeks after inoculation.†Number of mice bearing a tumor/total number of mice.‡ND, not determined.*P < 0.05; versus OCUM-2M or OCUM-2D.**P < 0.01; versus OCUM-2M or OCUM-2D.

Nishii et al. Cancer Sci | August 2009 | vol. 100 | no. 8 | 1399© 2009 Japanese Cancer Association

after intraperitoneal inoculation of 5 × 105, 1 × 106, 5 × 106,2 × 107, and 5 × 107 of OCUM-2M cells and 1 × 106 and 5 × 106

cells of OCUM-2D cells was not found in any mice.Tumorigenicity of OCUM-2M/SP and OCUM-2MD3 cells byintraperitoneal inoculation was significantly greater (P < 0.01,Fisher’s exact test) than that of OCUM-2M cells. Tumorigenicityof OCUM-2D/SP cells by intraperitoneal inoculation was alsosignificantly greater than that of OCUM-2D cells. Thesubcutaneous injection of only 5 × 104 OCUM-2M/SP cellsresulted in xenografted tumor formation in four (80%) of fivemice, while 1 × 105, 1 × 106, and 5 × 106 cells of OCUM-2Mresulted in xenografted tumor formation in none of nine mice,two (20%) of 10 mice, and three (25%) of 14 mice, respectively.In OCUM-2D/SP cells, subcutaneous injection of 1 × 105 and1 × 106 cells resulted xenografted tumor formation in three(27%) of 11 mice and eight (100%) of eight mice, respectively.Tumorigenicity by subcutaneous injection of OCUM-2M/SPand OCUM-2D/SP cells was also significantly greater (P < 0.05,Fisher’s exact test) than that of OCUM-2M and OCUM-2D cells(Table 1).

The adhesion ability of SP cells to peritoneal components.Figure 2a shows cancer cells binding to fibronectin, matrigel,and laminin. The number of adhesion cells of OCUM-2MD3and OCUM-2M/SP was greater compared to that of OCUM-2Mand the number of adhesion cells of OCUM-2D/SP was alsogreater compared to that of OCUM-2D. The adhesion ability tofibronectin, matrigel, and laminin of OCUM-2M/SP cells(20.0%, 26.3%, 19.3%) and OCUM-2MD3 cells (25.5%, 35.7%,31.6%) was significantly increased compared with that ofOCUM-2M cells (1.2%, 4.0%, 11.7%). Taken together, theadhesion ability of OCUM-2D/SP cells (69.0%, 16.7%, 27.1%)was significantly increased compared with that of OCUM-2Dcells (24.6%, 8.3%, 10.5%) (Fig. 2b).

mRNA expression of adhesion molecules in SP cells. The expressionof α2-, α3-, α5-, β1-, β3-, and β5-integrin, CD44, Oct3/4, andSox2 mRNA was significantly increased 11.1-, 4.9-, 29-, 2.7-,10.8-, 7.6-, 5.8-, 1.4-, and 37.6-fold in OCUM-2M/SP, and 13.9-,7.7-, 5.2-, 1.9-, 1.1-, 6.8-, 5.4-, 3.7-, and 1.9-fold in OCUM-2MD3 cells, compared with that in OCUM-2M cells. Incontrast, the expression of α1-integrin and CD133 mRNA inOCUM-2M/SP and OCUM-2MD3 cells was significantlydecreased compared with that of OCUM-2M cells (Fig. 3a). InOCUM-2D/SP cells, the expression of α2-, α5-, β3-, and β5-integrin, CD44, Oct3/4, and Sox2 mRNA was significantlyincreased 2.3-, 1.6-, 2.3-, 3.6-, 9.8-, 3.4-, and 10.4-fold incomparison with that of OCUM-2D cells, while the expressionof α1-, α3-, and β1-integrin, and CD133, mRNA wassignificantly decreased (Fig. 3b).

Stem cell marker expression in tumors by SP cells. The evaluationof Oct3/4 staining of xenografted tumor was 3+ positive byOCUM-2M/SP, OCUM-2D/SP, and OCUM-2MD3 cells, whilethat by OCUM-2M and OCUM-2D cells was negative. Theevaluation of Sox2 expression of xenografted tumor by OCUM-2M/SP and OCUM-2D/SP cells was 3+ positive, while those byOCUM-2M, OCUM-2D, and OCUM-2MD3 were ‘+’, negative,and negative, respectively. The evaluation of CD133 staining ofxenografted tumor by OCUM-2M/SP and OCUM-2D/SP cellswas negative, while those by parent cells were 2+ (Fig. 4).

Discussion

The rate of SP cells in OCUM-2M/SP (6.49%) and OCUM-2D/SP (11.4%) significantly increased 14.1 and 38.6 times, incomparison to that in parent OCUM-2M (0.46%) and OCUM-2D (0.29%), respectively. The serial sorting and reanalysis ofOCUM-2M/SP and OCUM-2D/SP cells demonstrated an

Fig. 1. Side population (SP) cells in gastric cancercell lines. (a) Typical examples of flow cytometricanalysis. The frequency of SP cells in OCUM-2M,OCUM-2M/SP, OCUM-2D, CUM-2D/SP, and OCUM-2MD3 cells was 1.4%, 6.4%, 0.4%, 14.7%, and6.0%, respectively. (b) The percentages of SP cellsin OCUM-2M/SP cells (6.49%) and OCUM-2MD3cells (5.24%) were significantly increasedcompared to OCUM-2M cells, and thepercentages of SP cells in OCUM-2D/SP cells(11.3%) were significantly increased compared toOCUM-2D cells. Results presented are the meansof three independent experiments, and barsindicate the SD. *, P < 0.01. **, P < 0.001.

1400 doi: 10.1111/j.1349-7006.2009.01211.x© 2009 Japanese Cancer Association

enrichment of the SP cells (from 22.3% to 6.49% and from27.5% to 10.0%) and the presence of non-SP cells. Thesefindings suggest that gastric cancer cell line OCUM-2M andOCUM-2D cells contain SP cells that are able to self-renew andproduce non-SP cells. In this study, the peritoneal inoculationof both OCUM-2M/SP and OCUM-2D/SP cells resulted inperitoneal adhesion, while the intraperitoneal inoculation ofgreater amounts of OCUM-2M and OCUM-2D cells resulted inan absence of peritoneal metastasis in mice. Although OCUM-2M and OCUM-2D cells have tumorigenicity of subcutaneoustumors as a result of the inoculation of 5 × 106 parent cells, theirparent cells do not have tumorigenicity to the peritoneum. Thesefindings suggest that cancer cells in SP fractions possess highpotential for peritoneal metastasis. OCUM-2M/SP cells may beuseful for the analysis of the mechanisms responsible forperitoneal metastasis of CSCs.

Peritoneal metastasis develops through a sequential process,including the detachment of cancer cells from the serosa, freeingof the cells from the peritoneal cavity, and adhesion to theperitoneum. The adhesion ability of SP cells to fibronectin,matrigel, and laminin was significantly high compared to that ofthe parent cells. The mRNA expression level of adhesion mole-cules of α2-, α5-, β3-, and β5-integrin, and CD44, in OCUM-2M/SP and OCUM-2D/SP cells was significantly high comparedto their parent cells. α3β1-integrin binds to laminin andfibronectin, α5β1-integrin binds to fibronectin, and CD44 bindsto collagen type IV, and αvβ3-integrin and αvβ5-integrin bindto vitronectin and fibronectin. αvβ3-integrin and αvβ5-integrinare also involved in cancerous biological activity such as cell

adhesion and migration to several extracellular matrix compo-nents. We previously established OCUM-2MD3 as a cell linewith high potential for peritoneal metastasis from parentOCUM-2M cells of a poor metastatic potential by in vivo cloningmethods;(19) thus suggesting that α3β1-integrin, α2β1-integrin,and CD44(11–13) were associated with a high potential for peritonealmetastasis in OCUM-2MD3 cells. The SP fraction rate inOCUM-2MD3 (5.24%) was also significantly high compared tothat in OCUM-2M. These findings suggest that adhesionmolecules of α2-, α5-, α3-, and β5-integrin, and CD44, may beassociated with the high potential for peritoneal metastasis ofSP cells.

Recently, several investigators have reported that certainintegrins may be stem cell markers in certain tissues, includingthe colonic epithelium,(21) epidermal keratinocytes,(22) and prostateepithelium.(23) In addition, Oct3/4 and Sox2 have been suggestedto be factors that render the reprogramming capability of adultcells into germ-line-competent-induced pluripotent stem cells.(14–16)

On the other hand, no distinct markers for CSCs exist thus far.Recently, the expression of CD44(24) and integrin,(1,25,26) Oct4,(27–29)

and Sox2 has been shown in some CSC-like cells, suggestingthat its expression may be implicated in self-renewal and tumori-genesis via activating its downstream target genes. In our study,the expression level of Oct3/4, Sox2, and CD44 was signifi-cantly high in SP cells, in comparison to parent cells, thus sug-gesting that CD44, Oct3/4, and Sox2 may be associated withCSC in gastric cancer. In contrast, the mRNA and proteinexpression of CD133 were decreased in both OCUM-2M/SPand OCUM-2D/SP cells, in comparison with their parent cells.

Fig. 2. Adhesion ability of side population (SP)cells. (a) Photograph of adhesive cells. In thephotograph of OCUM-2M and OCUM-2D, fewcells were observed that adhered to thefibronectin, matrigel, and laminin. In thephotograph of OCUM-2M/SP and OCUM-2MD3,adherent cells were increased compared withOCUM-2M. In the photograph of OCUM-2D/SP,adherent cells were increased compared withOCUM-2D cells. (b) The percentage of adherentgastric cancer cells to ECM. The percentage ofOCUM-2M/SP cells adherent to the various ECMcomponents (fibronectin, matrigel, and laminin)was 20%, 26.3%, and 19.3%, respectively, andsignificantly greater (P < 0.01, fibronectin,matrigel; and P < 0.05, laminin) than those ofOCUM-2M (1.2%, 4%, 11.7%). The percentage ofOCUM-2MD3 cells adherent to fibronectin was25.5% significantly greater (P = 0.024), tomatrigel 35.7% significantly greater (P = 0.036),and to laminin 31.6% significantly greater(P < 0.01) than that of OCUM-2M/SP cells. Thepercentage of OCUM-2D/SP cells adherent tofibronectin, matrigel, and laminin (69.0%, 16.7%,27.1%) was significantly greater (P < 0.01) thanthose of OCUM-2D (24.6%, 8.3%, 10.5%),respectively. Results presented are the mean offour independent experiments, and bars indicatethe SD. *, P < 0.05. **, P < 0.01.

Nishii et al. Cancer Sci | August 2009 | vol. 100 | no. 8 | 1401© 2009 Japanese Cancer Association

Fig. 3. mRNA expression of adhesion molecule and stemness genes. (a) In OCUM-2M/SP and OCUM-2MD3 cells, the expression of α2-, α3-, α5-, β1-,β3-, and β5-integrin, CD44, Oct3/4, and Sox2 mRNA was significantly increased compared with OCUM-2M cells. α1-integrin and CD133 mRNA wasnot amplified in OCUM-2M/SP and OCUM-2MD3 cells. (b) The expression of α2-, α 5-, β3-, and β5-integrin, CD44, Oct3/4, and Sox2 mRNA in OCUM-2D/SP cells was significantly increased compared with OCUM-2D cells. α1-, α3-, and β1-Integrin and CD133 mRNA expression was not increased inOCUM-2D/SP cells. Quantitative PCR reactions were done in triplicate. *, P < 0.05. **, P < 0.01.

1402 doi: 10.1111/j.1349-7006.2009.01211.x© 2009 Japanese Cancer Association

Recently, it was reported that CD133 was not increased in sometypes of stem cell or CSC-like cells.(30) We therefore concludedthat CD133 might not be an appropriate marker of CSC in gastriccarcinoma.

In conclusion, CSC-like SP cells have a high adhesion abilityto the peritoneum which may be associated with α2-, α5-, α3-,and β5-integrin, and CD44. Oct3/4 and Sox2 may be associatedwith CSC in gastric cancer.

Acknowledgments

This study was partially funded by Grants-in-Aid for Scientific Research(Nos. 18591475 20591073 and 18390369) from the Ministry of Education,Science, Sports, Culture and Technology of Japan; a JSGE Grant-in-Aidfor Scientific Research; a Grant-in-Aid from the Kobayashi Foundationfor Innovative Cancer Chemotherapy; a Grant-in-Aid from the SagawaFoundation for Cancer Research; and a Grant-in-Aid from the OsakaMedical Research Foundation for Incurable Diseases.

References

1 Al-Hajj M, Wicha MS, Benito-Hernandez A, Morrison SJ, Clarke MF.Prospective identification of tumorigenic breast cancer cells. Proc Natl AcadSci USA 2003; 100: 3983–8.

2 Singh SK, Clarke ID, Terasaki M et al. Identification of a cancer stem cellin human brain tumors. Cancer Res 2003; 63: 5821–8.

3 Ma S, Chan KW, Hu L et al. Identification and characterization oftumorigenic liver cancer stem/progenitor cells. Gastroenterology 2007; 132:2542–56.

4 Haraguchi N, Inoue H, Tanaka F et al. Cancer stem cells in humangastrointestinal cancers. Hum Cell 2006; 19: 24–9.

5 O’Brien CA, Pollett A, Gallinger S, Dick JE. A human colon cancer cellcapable of initiating tumour growth in immunodeficient mice. Nature 2007;445: 106–10.

6 Clarke MF, Dick JE, Dirks PB et al. Cancer stem cells – perspectives oncurrent status and future directions: AACR Workshop on cancer stem cells.Cancer Res 2006; 66: 9339–44.

7 Tatematsu M, Tsukamoto T, Inada K. Stem cells and gastric cancer: role ofgastric and intestinal mixed intestinal metaplasia. Cancer Sci 2003; 94: 135–41.

8 Tu SM, Lin SH, Logothetis CJ. Stem-cell origin of metastasis andheterogeneity in solid tumours. Lancet Oncol 2002; 3: 508–13.

9 Goodell MA, Rosenzweig M, Kim H et al. Dye efflux studies suggest thathematopoietic stem cells expressing low or undetectable levels of CD34antigen exist in multiple species. Nat Med 1997; 3: 1337–45.

10 Glinsky GV, Berezovska O, Glinskii AB. Microarray analysis identifies adeath-from-cancer signature predicting therapy failure in patients withmultiple types of cancer. The J Clin Invest 2005; 115: 1503–21.

11 Koyama T, Yashiro M, Inoue T, Nishimura S, Hirakawa YSCK. TGF-beta1secreted by gastric fibroblasts up-regulates CD44H expression and stimulatesthe peritoneal metastatic ability of scirrhous gastric cancer cells. Int J Oncol2000; 16: 355–62.

12 Nishimura S, Chung YS, Yashiro M, Inoue T, Sowa M. Role of alpha 2 beta1- and alpha 3 beta 1-integrin in the peritoneal implantation of scirrhousgastric carcinoma. Br J Cancer 1996; 74: 1406–12.

13 Matsuoka T, Yashiro M, Nishimura S et al. Increased expression ofalpha2beta1-integrin in the peritoneal dissemination of human gastriccarcinoma. Int J Mol Med 2000; 5: 21–5.

14 Okita K, Ichisaka T, Yamanaka S. Generation of germline-competentinduced pluripotent stem cells. Nature 2007; 448: 313–7.

15 Yu J, Vodyanik MA, Smuga-Otto K et al. Induced pluripotent stem cell linesderived from human somatic cells. Science 2007; 318: 1917–20.

16 Park IH, Zhao R, West JA et al. Reprogramming of human somatic cells topluripotency with defined factors. Nature 2008; 451: 141–6.

17 Yashiro M, Chung YS, Sowa M. Role of orthotopic fibroblasts in thedevelopment of scirrhous gastric carcinoma. Jpn J Cancer Res 1994; 85:883–6.

18 Yashiro M, Chung YS, Nishimura S, Inoue T, Sowa M. Establishment of twonew scirrhous gastric cancer cell lines: analysis of factors associated withdisseminated metastasis. Br J Cancer 1995; 72: 1200–10.

19 Yashiro M, Chung YS, Nishimura S, Inoue T, Sowa M. Peritoneal metastaticmodel for human scirrhous gastric carcinoma in nude mice. Clin ExpMetastasis 1996; 14: 43–54.

20 Pfaffl MW. A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res 2001; 29: e45.

21 Fujimoto K, Beauchamp RD, Whitehead RH. Identification and isolation ofcandidate human colonic clonogenic cells based on cell surface integrinexpression. Gastroenterology 2002; 123: 1941–8.

22 Jones PH, Harper S, Watt FM. Stem cell patterning and fate in humanepidermis. Cell 1995; 80: 83–93.

23 Collins AT, Habib FK, Maitland NJ. Neal DE. Identification and isolation ofhuman prostate epithelial stem cells based on alpha(2)beta(1)-integrinexpression. J Cell Sci 2001; 114: 3865–72.

24 Patrawala L, Calhoun-Davis T, Schneider-Broussard R, Tang DG.Hierarchical organization of prostate cancer cells in xenograft tumors: theCD44+alpha2beta1+ cell population is enriched in tumor-initiating cells.Cancer Res 2007; 67: 6796–805.

25 Zhang X, Komaki R, Wang L, Fang B, Chang JY. Treatment ofradioresistant stem-like esophageal cancer cells by an apoptotic gene-armed,telomerase-specific oncolytic adenovirus. Clin Cancer Res 2008; 14: 2813–23.

26 Li C, Heidt DG, Dalerba P et al. Identification of pancreatic cancer stemcells. Cancer Res 2007; 67: 1030–7.

27 Ezeh UI, Turek PJ, Reijo RA, Clark AT. Human embryonic stem cell genesOCT4, NANOG, STELLAR, and GDF3 are expressed in both seminoma andbreast carcinoma. Cancer 2005; 104: 2255–65.

28 Chiou SH, Yu CC, Huang CY et al. Positive correlations of Oct –4 andNanog in oral cancer stem-like cells and high-grade oral squamous cellcarcinoma. Clin Cancer Res 2008; 14: 4085–95.

29 Chen Y, Shi L, Zhang L et al. The molecular mechanism governing theoncogenic potential of SOX2 in breast cancer. J Biol Chem 2008; 283:17969–78.

30 Bussolati B, Bruno S, Grange C, Ferrando U, Camussi G. Identification of atumor-initiating stem cell population in human renal carcinomas. Faseb J2008; 22: 3696–705.

Fig. 4. Expression of stem cell markers. Thenumber of Oct3/4-positive cancer cells wasincreased in tissue injected with OCUM-2M/SPand OCUM-2MD3 cells, compared to OCUM-2Mcells. The evaluation of Oct3/4 staining was ‘+++’in the tumor by OCUM-2D/SP cells, while in thetumor by OCUM-2D cells, positive cancer cellswere not found. The evaluation of Sox2 stainingby OCUM-2M/SP and OCUM-2D/SP was ‘+++’,and that by OCUM-2M was ‘+’. In the tumors byOCUM-2D and OCUM-2MD3 cells, Sox2 stainingwas negative. The evaluation of CD133 stainingin the tumors by OCUM-2M and OCUM-2D cellswas ‘++’, while in the tumors by SP and OCUM-2MD3 cells the evaluation was ‘–’. (Originalmagnification, ×400).