Motility Inhibition and Apoptosis Are Induced by Metastasis-Suppressing Gene Product CD82 and Its Analogue CD9, with Concurrent Glycosylation1

[CANCER RESEARCH 59, 2335–2339, May 15, 1999] Advances in Brief

Motility Inhibition and Apoptosis Are Induced by Metastasis-suppressing Gene Product CD82 and Its Analogue CD9, with Concurrent Glycosylation1

Masaya Ono,2 Kazuko Handa, Donald A. Withers, and Sen-itiroh Hakomori3 Pacific Northwest Research Institute, Seattle, Washington 98122, and Departments of Pathobiology and Microbiology, University of Washington, Seattle, Washington 98195

Abstract ITS4 medium (Collaborative Biomedical Products, Bedford, MA; Ref. 8) by changing the medium on day 1 to serum-free ITS with or without Gal (20 ␮M) Metastasis-suppressing gene product CD82 and its analogue CD9 are and/or GalNAc (200 ␮M). The effect of glycosylation on cellular function was considered to suppress the malignancy of various human cancers, al- determined on day 3 for cell migration and on day 10–11 for apoptosis. though the rationale for this effect is unknown. The present study ad- Preparation of cDNA Encoding CD9 and CD82, and Its Transfection dresses phenotypic changes in Chinese hamster ovary mutant cell line ldlD into ldlD-14 Cells. cDNAs for CD9 and CD82 were generated by reverse deficient in UDP-Glc 4-epimerase and expressing CD82 or CD9 by cDNA transcription-PCR using human placental RNA (Clontech, Palo Alto, CA) and transfection. Only CD82- or CD9-expressing cells grown in Gal-supple- Superscript II reverse transcriptase (Life Technologies, Inc., Gaithersburg, mented medium showed reduced motility and massive cell death, which MD). PCR amplification with Elongase (Life Technologies, Inc.) was per- are characteristic of apoptosis, after a latent period. Under this condition, formed with 5Ј primer AGAGGAATTCCATGCCGGTCAAAGGAGG and 3Ј endogenous GM3 synthesis was observed as a common factor, and N- primer GAGGGGATCCTTCCTGCTCAGGGATGTAAG for CD9 and with Ј Ј glycosylation occurred at a high level in CD82 and to a lesser extent in 5 primer GCGGGAATTCGATGGGCTCAGCCTGTATCAAAG and 3 CD9. Thus, the malignancy-suppressing effect of CD82 or CD9 is based primer GGCGGCTCGAGTCAGTACTTGGGGACCTTGC for CD82. The re- partially on cell motility inhibition and apoptosis induction promoted by sulting PCR products were cloned into pCDNA3 (Invitrogen, Carlsbad, CA) and sequenced to confirm identity by comparison with published sequences concurrent GM3 synthesis and N-glycosylation. (1, 3). pCDNA3 containing CD9 and CD82 was transfected into ldlD-14 cells Introduction using LipofectAMINE (Life Technologies, Inc.), transfectants were isolated by growing them in G418 (Life Technologies, Inc.)-containing medium, and Expression of the tetraspan membrane proteins CD82 (1, 2) and CD9 clones were selected by expression of the respective antigens probed by the (3) is down-regulated in various human cancers and is inversely corre- mAb, as described previously (9). lated with the patient survival rate (2, 4, 5). CD82 and CD9 suppress Haptotactic and Phagokinetic Motility Assay. Transwell assay using a tumor cell malignancy, although the mechanism for this phenomenon is 6.5-mm transwell assembly (8-␮m pore size; Costar, Cambridge, MA; Refs. 10 unknown. The cell biological significance of these proteins should be and 11) and the gold sol phagokinetic assay (12) were performed as described explored in relation not only to their expression per se, but also to the previously. Some modifications of the procedures are explained in the legend effects of their N-glycosylation and surrounding endogenous gangliosides of Fig. 2. Determination of GM3 and Other GSLs. Packed cells (106 cells har- in membrane, because these factors profoundly affect membrane receptor vested for one run) were extracted in 2 ml of chloroform-methanol (2:1 by function in general (for a review, see Ref. 6). For this purpose, Chinese volume) by sonication for 5 min and centrifuged, and the pellet was extracted hamster ovary mutant cell line ldlD-14 deficient in UDP-Glc 4-epimerase two more times with the same solvent. The GSL fraction from the combined and expressing CD82 or CD9 by cDNA transfection was utilized. Phe- extract was separated using an acetylation procedure (13). notypic changes of the transfectants, in terms of cell motility inhibition Determination of CD9 and CD82 with Glycosylation Status. Cells were and enhancement of apoptosis, were clearly observed in Gal-supple- solubilized in lysis buffer, subjected to SDS-PAGE, electrophoretically trans- mented medium. These changes reflect not only the effect of CD82 or ferred to Immobilon-P (Millipore, Bedford, MA), and subjected to Western blotting with a chemiluminescence system (Pierce, Rockford, IL; Ref. 11). CD9 expression per se, but also the effect of endogenous GM3 synthesis Protein content was determined with a BCA kit (Pierce). Antihuman CD9 as a common factor and a high level of N-glycosylation in CD82 and a mouse mAb M-L13 was from PharMingen (San Diego, CA). Antihuman CD82 lower level of N-glycosylation in CD9. These findings provide a rationale mouse mAb B-L2 was from Serotec (Raleigh, NC). for the suppression of malignancy by CD82 and CD9 with proper Determination of Apoptosis. Morphological changes characteristic of glycosylation. apoptosis (14), i.e. nuclear compaction, cytoplasmic condensation, and disintegration into dense particles, were examined. DNA fragmentation was determined by Materials and Methods terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (15) using the in situ cell death detection kit (Boehringer Mannheim, Indianapolis, IN). ldlD-14 Cell Culture, Antibodies, Reagents, and Flow Cytometry. Cell Translocation of phosphatidylserine associated with apoptosis (16) was deter- line ldlD-14, a Chinese hamster ovary cell mutant deficient in UDP-Glc mined by the annexin-V-fluos staining kit (Boehringer Mannheim). 4-epimerase (7), was maintained in Ham’s F-12 medium supplemented with 5% fetal bovine serum. The glycosylation pattern was altered in serum-free Results and Discussion CD82- and CD9-expressing ldlD-14 Cells. ldlD-14 clones ex- Received 2/25/99; accepted 4/5/99. pressing CD82 (ldlD/CD82; clones 515 and 1209) and clones The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with expressing CD9 (ldlD/CD9; clones 24, 28, and 47) were estab- 18 U.S.C. Section 1734 solely to indicate this fact. lished (Fig. 1A). SDS-PAGE and Western blotting indicate that: (a) 1 Supported by National Cancer Institute Outstanding Investigator Grant CA42505 (to S. H.). 2 Fellowship awardee from the Department of Surgery, University of Tokyo. 4 The abbreviations used are: ITS, insulin-transferrin-selenium; Gal, galactose; 3 To whom requests for reprints should be addressed, at Pacific Northwest Research GalNAc, N-acetylgalactosamine; GM3, NeuAc␣2-3Gal␤1-4Glc␤Cer; GSL, glycosphin- Institute, 720 Broadway, Seattle, WA 98122. Phone: (206) 726-1222; Fax: (206) 726- golipid; mAb, monoclonal antibody; SDS-PAGE, sodium dodecyl sulfate polyacrylamide 1212; E-mail: [email protected]. gel electrophoresis. 2335

Fig. 1. Expression of CD9 and CD82 (A), SDS- PAGE profiles of CD9 and CD82 (B), and patterns of GM3 and other GSLs in ldlD transfectants (C). A, flow cytometric pattern on EPICS XL (Coulter Corp.) showing CD9 expression (a) and CD82 expression (b). i, parental ldlD. ii, ldlD/CD9 clone 24 (clones 28 and 47 show similar reactivity). iii, ldlD/ CD82 clone 515 (clone 1209 shows similar reactivity). B, SDS-PAGE profiles of CD9 in ldlD/CD9 (a) and CD82 in ldlD/CD82 (b). Lane 1, cells grown in ITS medium. Lane 2, ITS with Gal. Lane 3, ITS with GalNAc. Lane 4, ITS with Gal plus GalNAc. Num- bers on the left indicate molecular mass ϫ 10Ϫ3. C, high performance thin-layer chromatography pattern of GSLs in parental ldlD (a), ldlD/CD9 (b), and ldlD/CD82 (c). Lanes 1–4, same as those in B. Lanes x, y, and z, reference LacCer, GM1, and GM3 re- vealed by orcinol-sulfuric acid.

the molecular mass (23 kDa) of CD9 in ldlD/CD9 cells was Apoptosis Was Induced Only in CD9 and CD82-expressing unchanged, regardless of the addition of Gal (Fig. 1 B, a); (b) Cells, with Concurrent GM3 Synthesis and Complete N-Glycosy- CD82 had a much higher molecular mass and a broader band lation. Massive cell death (apoptosis) occurred after 11 days in ldlD/ (ϳ45–65 kDa) in ldlD/CD82 cells grown in the presence of Gal or CD9 (Fig. 3A, first column) or ldlD/CD82 (second column) culture in Gal plus GalNAc (Fig. 1B, b); and (c) endogenous GM3 synthesis ITS added with Gal (Fig. 3A, row b) or with Gal plus GalNAc (row occurred for both CD9- and CD82-expressing cells grown in the c). Cell death did not occur in ITS medium without Gal addition (row presence of Gal or Gal plus GalNAc (Fig. 1C). a) or for parental ldlD14 cells, regardless of Gal addition (third Only CD9- or CD82-expressing Cells with Concurrent GM3 column). The effect of CD9 and CD82 with concurrent glycosylation Synthesis and Complete N-Glycosylation Display Inhibition of (Gal addition) on the induction of apoptosis was confirmed by flow Haptotactic and Phagokinetic Motility. Haptotactic motility of cytometry after nick end labeling, i.e. 3Ј hydroxyl end detected on day ldlD/CD9 clones 24 and 28 (Fig. 2A) and ldlD/CD82 clones 515 and 11 in cells cultured on a Matrigel-coated plate. Clear labeling was 1209 (Fig. 2B) as well as the phagokinetic motility of ldlD/CD9 observed in ldlD/CD9 and ldlD/CD82 in Gal-supplemented medium clones 24 and 47 (Fig. 2D) and ldlD/CD82 clones 515 and 1209 (Fig. (Fig. 3B, row b) and in Gal plus GalNAc-supplemented medium (row 2E) were all reduced in cells grown in the presence of Gal or Gal plus c). Labeling was minimal in ITS medium without Gal addition (row GalNAc, but not those grown in the presence of GalNAc alone. a) or for parental ldlD14 cells, regardless of Gal addition (third Parental ldlD cells, which do not express CD9 or CD82, showed no column). change of motility on the addition of Gal or GalNAc (Figs. 2 C Changes in the number of live cells/mm2 are shown in Fig. 4. The and F). number of ldlD/CD9 and ldlD/CD82 cells, but not of parental ldlD14 2336

Fig. 2. Haptotactic and phagokinetic motility of ldlD/CD9 and ldlD/CD82 cells. Motility was determined by transwell assay (A–C) and by phagokinetic gold sol assay (D–F). In the transwell assay (10, 11), the lower surface of the filter was coated by placing 10 ␮l of solution (containing 1.5 ␮g of Matrigel; Collaborative Biomedical Products) in distilled water, and the filter was dried at room temperature. Cell suspension (100 ␮l; containing 5 ϫ 104 cells) in Ham’s F-12 medium with 0.25% BSA was placed in the upper compartment, and 0.6 ml of Ham’s F-12 medium with 0.25% BSA was added in the lower compartment. Cells that migrated during the 24-h period were counted in five areas after hematoxylin staining. For the phagokinetic gold sol assay (12), cells were incubated and grown for 24 h on gold sol-coated plates. Cells were then fixed for 30 min in 3.5% formaldehyde solution, and the migrated area (in ␮m2) was measured using the Scion Image 1.62a computer program (Scion Corporation, Frederick, MD). A and B, transwell motility of ldlD/CD9 and ldlD/CD82, respectively, in ITS medium (a), ITS ϩ Gal (b), ITS ϩ GalNAc (c), or ITS ϩ Gal ϩ GalNAc (d). Ⅺ and u represent ldlD/CD9 clones 24 and 28, respectively, in A and Ⅺ and ■ represent IdlD/CD82 clones 515 and 1209 in B. Statistical significance of the differences between the groups are indicated by P values. NS, not significant. C, transwell motility of parental ldlD in ITS with or without the addition of Gal and/or GalNAc. D and E, phagokinetic motility of ldlD/CD9 and ldlD/CD82, respectively, in different media as described in A and B. ᮀ and ■ represent clones 24 and 47, respectively, in D and clones 515 and 1209 in E. Statistical significance is indicated as described in A and B. F, phagokinetic motility of parental ldlD in ITS with or without the addition of Gal and/or GalNAc. cells decreased significantly upon culture in Gal or Gal plus GalNAc- CD82 and CD9 inhibits haptotactic and phagokinetic cell motility supplemented medium on Matrigel-coated plates (Fig. 4A) or non- and induces cell death with features typical of apoptosis, and that coated plates (Fig. 4B). this process is promoted by concurrent GM3 synthesis and com- Dead cells in the above experiments showed a typical apoptotic plete N-glycosylation. Because N-glycosylation in CD9 is minimal, morphology (nuclear compaction, cytoplasmic condensation, and dis- endogenous GM3 synthesis has the major effect on CD9. On the integration into dense particles). other hand, there is a high level of N-glycosylation in CD82, as The Malignancy-suppressing Effect of CD82 and CD9 Is Due to indicated by the great increase in molecular mass and heterogene- Motility Inhibition and Apoptosis and Is Promoted by Concurrent ity in ldlD/CD82 cells grown in Gal-supplemented medium. Inhi- GM3 Synthesis and Complete N-Glycosylation. CD82 (KAI-1) bition of cell motility and induction of apoptosis in ldlD/CD82 was originally assigned as a metastasis-suppressing gene product, cells are ascribable to enhanced N-glycosylation. The effect of and its expression was shown to be down-regulated in a few types endogenous GM3 synthesis on CD82 must be similar to the effect of human cancer (1, 2, 5). Its analogue, CD9, is also down- on CD9, i.e. endogenous GM3 is the common factor for both CD9- regulated in metastatic tumors (3, 4). Expression of both CD82 and and CD82-dependent inhibition of cell motility and induction of CD9 was correlated with the patient survival rate (2, 4, 5). How- apoptosis. ever, the cell biological mechanism for these effects of CD82 and Expression of CD82 or CD9 per se is an essential, but not suffi- CD9 is unknown. cient, factor to induce cell motility inhibition and apoptosis. Comple- Results of the present study indicate that expression of both tion of N-glycosylation and endogenous GM3 synthesis, as observed 2337

Fig. 3. CD9- and CD82-dependent apoptosis promoted by glycosylation. A, morphology change of ldlD cells at day 11 on a Matrigel-coated plate. Three cell lines (as indicated at the bottom) were grown in ITS medium alone (a), ITS ϩ Gal (b), and ITS ϩ Gal ϩ GalNAc (c). The number of adherent cells is dramatically decreased in Gal-containing medium (b and c) for CD9- and CD82-expressing cells but is minimally changed for parental ldlD14 cells. Dead cells are characterized by round shape, nuclear compaction, cytoplasmic condensation, and disintegration into dense particles. B, flow cytometric patterns showing nick end labeling of ldlD/CD9, ldlD/CD82, and parental ldlD cells. Cells were cultured on Matrigel-coated plates in ITS medium (top row), ITS ϩ Gal (second row), or ITS ϩ Gal ϩ GalNAc (third row) and stained on day 11. Note that the high fluorescence peak representing apoptotic cells is clearly observed only in CD9- or CD82-transfected cells in the presence of Gal. This trend is much weaker in parental cells without CD9 or CD82. typically in ldlD cells grown in Gal-supplemented medium, are re- about the correlation between glycosylation and apoptosis, although quired to manifest motility inhibition as well as CD82- or CD9- Lex (17) and Ley (18) expressed in colorectal carcinoma are associ- dependent induction of apoptosis, although the latter process requires ated with apoptosis. Glycosylation that defines malignancy depends at least a 10-day latent period. What type of molecular mechanism largely on which molecular species is glycosylated and is susceptible takes place during the latent period is yet unidentified. Little is known to gangliosides. Tetraspan membrane proteins that suppress tumor 2338

References 1. Dong, J-T., Lamb, P. W., Rinker-Schaeffer, C. W., Vukanovic, J., Ichikawa, T., Isaacs, J. T., and Barrett, J. C. KAI1, a metastasis suppressor gene for prostate cancer on human chromosome 11p11.2. Science (Washington DC), 268: 884– 886, 1995. 2. Dong, J. T., Suzuki, H., Pin, S. S., Bova, S., Schalken, J. A., Isaacs, W. B., Barrett, J. C., and Isaacs, J. T. Down-regulation of the KAI1 metastasis suppressor gene during the progression of human prostatic cancer infrequently involves gene mutation or allelic loss. Cancer Res., 56: 4387–4390, 1996. 3. Miyake, M., Koyama, M., Seno, M., and Ikeyama, S. Identification of the motility- related protein (MRP-1), recognized by monoclonal antibody M31–15, which inhibits cell motility. J. Exp. Med., 174: 1347–1354, 1991. 4. Cajot, J-F., Sordat, I., Silvestre, T., and Sordat, B. Differential display cloning identifies motility-related protein (MRP1/CD9) as highly expressed in primary compared to metastatic human colon carcinoma cells. Cancer Res., 57: 2593– 2597, 1997. 5. Adachi, M., Taki, T., Ieki, Y., Huang, C., Higashiyama, M., and Miyake, M. Correlation of KAII/CD82 gene expression with good prognosis in patients with non-small cell lung cancer. Cancer Res., 56: 1751–1755, 1996. 6. Hakomori, S. Tumor malignancy defined by aberrant glycosylation and sphingo(gly- co)lipid metabolism. Cancer Res., 56: 5309–5318, 1996. 7. Kingsley, D. M., Kozarsky, K. F., Hobbie, L., and Krieger, M. Reversible defects in O-linked glycosylation and LDL receptor expression in a UDP-Gal/UDP-GalNAc 4-epimerase deficient mutant. Cell, 44: 749–759, 1986. 8. Krieger, M., Reddy, P., Kozarsky, K., Kingsley, D., Hobbie, L., and Penman, M. Analysis of the synthesis, intracellular sorting, and function of glycoproteins using a mammalian cell mutant with reversible glycosylation defects. Methods Cell Biol., 32: 57–84, 1989. 9. Handa, K., White, T., Ito, K., Fang, H., Wang, S., and Hakomori, S. P-selectin- dependent adhesion of human cancer cells: requirement for co-expression of a “PSGL-1-like” core protein and the glycosylation process for sialosyl-Lex or sialosyl- Lea. Int. J. Oncol., 6: 773–781, 1995. 10. Albini, A., Iwamoto, Y., Kleinman, H. K., Martin, G. R., Aaronson, S. A., Kozlowski, J. M., and McEwan, R. N. A rapid in vitro assay for quantitating the invasive potential Fig. 4. CD9- or CD82-dependent cell death of ldlD/CD9 or ldlD/CD82 in Gal-added of tumor cells. Cancer Res., 47: 3239–3245, 1987. ϫ 4 medium. Cells (5 10 ) were plated on 24-well dishes coated with or without Matrigel 11. Ichikawa, D., Handa, K., Withers, D. A., and Hakomori, S. Histo-blood group A/B in 5% fetal bovine serum with Ham’s F-12 medium. The medium was changed to ITS with versus H status of human carcinoma cells as correlated with haptotactic cell motility: or without Gal and GalNAc the next day (day 1). The adherent cell number was counted approach with A and B gene transfection. Cancer Res., 57: 3092–3096, 1997. as four random places for each dish using a videocapture system (0.2 mm2/screen; Scion 12. Albrecht-Buehler, G. The phagokinetic tracks of 3T3 cells. Cell, 11: 395–404, Corp.). Experiments were run in duplicate. Ordinate, cell number/mm2. Cell numbers and 1977. morphology showed the same pattern until day 9 for dishes coated with (A) or without Matrigel (B). Massive cell death occurred at day 11 for ldlD/CD9 and ldlD/CD82 cells 13. Saito, T., and Hakomori, S. Quantitative isolation of total glycosphingolipids from grow in in ITS ϩ Gal ( )orITSϩ Gal ϩ GalNC (f). Cell death was minimal for cells animal cells. J. Lipid Res., 12: 257–259, 1971. grown in ITS alone (e) and for parental cells not expressing CD9 or CD82, regardless of 14. Martin, S. J., Green, D. R., and Cotter, T. G. Dicing with death: dissecting the the addition of sugars (A, ldlD14). The degree of cell death was greater for cells on dishes components of the apoptosis machinery. Trends Biochem. Sci., 19: 26–30, 1994. without Matrigel (B). Statistical significance of the differences between the groups is 15. Gavrieli, Y., Sherman, Y., and Ben-Sasson, S. A. Identification of programmed cell indicated by P values. death in situ via specific labeling of nuclear DNA fragmentation. J. Cell Biol., 119: 493–501, 1992. 16. Koopman, G., Reutelingsperger, C. P., Kuijten, G. A., Keehnen, R. M., Pals, S. T., malignancy are undoubtedly target molecules of glycosylation, con- and van Oers, M. H. Annexin V for flow cytometric detection of phosphatidylserine trolling motility and apoptosis. expression on B cells undergoing apoptosis. Blood, 84: 1415–1420, 1994. 17. Akamatsu, S., Yazawa, S., Zenita, K., Matsumoto, H., Tachikawa, T., and ␣ Acknowledgments Kannagi, R. Elevation of an (1,3)fucosyltransferase activity correlated with apoptosis in the human colon adenocarcinoma cell line, HT-29. Glycoconj. J., 13: 1021–1029, 1996. We thank Dr. Monty Krieger for the donation of ldlD-14 cells, Wendy 18. Hiraishi, K., Suzuki, K., Hakomori, S., and Adachi, M. Ley antigen expression is Smith and Jon McBride for technical assistance, and Dr. Stephen Anderson for correlated with apoptosis (programmed cell death). Glycobiology, 3: 381–390, scientific editing and preparation of the manuscript. 1993.

2339

Downloaded from cancerres.aacrjournals.org on September 28, 2021. © 1999 American Association for Cancer Research. Motility Inhibition and Apoptosis Are Induced by Metastasis-suppressing Gene Product CD82 and Its Analogue CD9, with Concurrent Glycosylation

Masaya Ono, Kazuko Handa, Donald A. Withers, et al.

Cancer Res 1999;59:2335-2339.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/59/10/2335

Cited articles This article cites 18 articles, 11 of which you can access for free at: http://cancerres.aacrjournals.org/content/59/10/2335.full#ref-list-1

Citing articles This article has been cited by 31 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/59/10/2335.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/59/10/2335. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.