Cytotoxic Sensitivity to Tumor NecrosisFactor-a in PC3 and LNCaP Prostatic CancerCells Is Regulated by Extracellular Levels of

SGP-2 (Clusterin)

Sharon M. Sintich, Joseph Steinberg, James M. Kozlowski, Chung Lee,Shaina Pruden, Sakina Sayeed, and Julia A. Sensibar*

Department of Urology, Northwestern University Medical School, Chicago, Illinois

BACKGROUND. SGP-2 is a ubiquitous secreted glycoprotein that prevents cellular apopto-sis. This study was carried out to determine the extracellular action of SGP-2 in a model oftumor necrosis factor-a (TNF)-induced cytotoxicity using two human prostatic cancer lines,LNCaP and PC3. These two lines were selected because LNCaP cells are highly sensitive tothe cytotoxic effect of TNF, while PC3 cells are resistant to TNF at 24 hr.METHODS. Cells were cultured in the presence or absence of TNF (10 ng/ml). LNCaP cellswere treated with varying concentrations of exogenous SGP-2, while PC3 cells were treatedwith antisera to SGP-2 with and without exogenous SGP-2. Following a 24-hr treatment,cultures were assessed by counting of cell number and by the trypan blue exclusion assay.RESULTS. Western blot analysis of conditioned media revealed that PC3 secreted moreSGP-2 than did LNCaP. The sensitivity to TNF in LNCaP cells was reduced by the additionof exogenous SGP-2. PC3 cells became sensitive to TNF when SGP-2 antibody was added tothe culture. The effect of SGP-2 antibody on PC3 cells was reversed by the addition ofexogenous SGP-2 to the culture.CONCLUSIONS. These results suggest that SGP-2 can act as an extracellular mediator ofanti-TNF-induced cytotoxicity. Prostate 39:87–93, 1999. © 1999 Wiley-Liss, Inc.

KEY WORDS: cytotoxicity; SGP-2; clusterin; tumor necrosis factor-a; prostate cancer

INTRODUCTION

This report describes the biological action of sul-fated glycoprotein-2 (SGP-2, also known as clusterin,TRPM-2, apolipoprotein J, and complement lysis in-hibitor) in two established human prostate cancer celllines, PC3 and LNCaP. SGP-2 is a ubiquitous heterodi-meric sulfated glycoprotein first isolated from ramrete testes, and it was so named for its ability to clusterSertoli cells [1]. This protein has been found in a widerange of organisms and is highly conserved [2,3]. Ithas a multitude of proposed functions, including rolesin tissue remodeling, reproduction, lipid transport, in-hibition of complement-mediated cell lysis, and apop-tosis [4–6]. Results of our previous studies demon-strated a protective role of SGP-2 in apoptosis in pros-tatic cancer cells [7]. Since this discovery, it has

become apparent that the cytoprotective effect ofSGP-2 is also implicated in a large variety of patho-logical conditions. These conditions include prostaticregression, kidney injury, myocardial infarction, andthe negative selection of thymocytes [8–11]. To date,the exact mode of action of SGP-2 in target cells re-mains unknown. In the present study, we investigatedthe extracellular action of SGP-2. We continue to usetumor necrosis factor-a (TNF) as the cytotoxic inducerand human prostatic cancer cells as the target. LNCaP

Grant sponsor: NIH; Grant numbers: CA 69851, CA 09560.*Correspondence to: Julia A. Sensibar, Ph.D., Department of Urol-ogy and Lurie Cancer Center, Northwestern University MedicalSchool, 710 N. Fairbanks Court, Room 8360, Chicago, IL 60611. E-mail: jas846@nwu.eduReceived 8 December 1997; Accepted 21 October 1998

The Prostate 39:87–93 (1999)

© 1999 Wiley-Liss, Inc.

is a slow-growing cancer line and is exquisitely sensi-tive to the cytotoxic effects of TNF [7,12]. PC3 is anaggressive line that is more resistant to TNF [12] Wedemonstrate that sensitivity to TNF in these cells is, atleast in part, regulated by extracellular levels of SGP-2.We propose that SGP-2 may be an extracellular me-diator of anti-TNF-induced cytotoxicity.

MATERIALS AND METHODS

Cell Cultures and Conditions

LNCaP and PC3 cells were purchased from Ameri-can Type Culture Collection (Rockville, MD). All cellswere maintained in RPMI-1640 containing 10% fetalbovine serum and penicillin (100 units/ml)/streptomycin (100 mg/ml). Cells were plated in serum-containing media in 24-well plates for 24 hr prior totreatment. Subsequently, cultures were changed tomedia containing RPMI-1640 supplemented with ITS+(5 mg/ml insulin, 5 mg/ml transferrin, 5 ng/ml sele-nous acid, 1.25 mg/ml bovine serum albumin (BSA),and 5 mg/ml linoleic acid; Collaborative Research,Inc., Bedford, MA) supplemented with or without 10ng/ml TNF (specific activity, 6.27 × 107 units/mg; Col-laborative Research, Inc.). This dose of TNF was se-lected based on our previous finding that it yielded amaximal level of cytotoxic effect in LNCaP cells [7,12].At the time of TNF treatment, LNCaP cells weretreated with varying concentrations of exogenousSGP-2 (Quidel, San Diego, CA), while PC3 cells weretreated with sheep anti-rat SGP-2 (Quidel) with andwithout exogenous SGP-2. A 24-hr culture period wasused to assess the cytotoxic effect of TNF on targetcells, as described [7].

Cell Counting

For determination of cell number, cells in 24-wellplates were detached by 0.5 ml trypsin-EDTA solution(Hazelton Biologics, Lenexna, KS). Detached cellswere transferred to a counting vessel containing 9.5 mlisotonic solution (Isoton II, Coulter Corp., Hialeah, FL)and counted using a Coulter counter (Coulter Corp.).

Trypan Blue Dye Exclusion Assay

To determine the cytotoxic effect of TNF, trypanblue dye exclusion assay was employed at the end ofeach culture period. Cells in 24-well plates weretreated with TNF in addition to SGP-2 protein (or an-tibody to SGP-2) for 24 hr, as described above. Theywere detached by 0.5 ml trypsin-EDTA solution (Ha-zelton Biologics) and pelleted by centrifugation.Trypan blue dye (Sigma Chemical Co., St. Louis, MO)

was added to each pellet, and the number of cellswhich absorbed the dye was counted using a hemo-cytometer. A minimum of 200 cells was counted. Thepercentage of viable cells in each culture was calcu-lated according to the following equation:

% viable cells =Total cells − Viable cells

Total cells

This value was used as an end point of cytotoxicity.

Preparation of Conditioned Media

Conditioned media were prepared by growing cellsto 80% confluence in RPMI-1640 medium containing10% fetal bovine serum. Cultures were then switchedto serum-free media for 24 hr. The resulting condi-tioned medium was harvested and centrifuged to re-move floating cells. The harvested medium was con-centrated with Centriprep centrifugal concentrators(Amicon, Inc., Beverly, MA) with a molecular cutoff of3.0 kDa according to the manufacturer’s instructions.The protein content of the concentrated conditionedmedium was determined by the method of Bradford[13].

Immunoprecipitation and Western Blot Analysis

Protein A Sepharose CL4B beads (Pharmacia,Uppsala, Sweden) were reconstituted in phosphate-buffered saline (PBS) for 2 hr at 4°C prior to incubationwith the antibody. The sepharose beads were coatedwith 20 mg/ml of sheep anti-rat SGP-2 antibody(Quidel) by incubation at 4°C overnight. Subse-quently, the beads were washed with PBS and mixedwith 417 ng/ml of concentrated conditioned media at4°C. The beads were then washed, reconstituted withLaemmli sample buffer, and boiled for 5 min [14].Equal volumes of each sample were loaded on a 10%polyacrylamide gel. Following electrophoresis, pro-teins in the gel were transferred to a 0.45-mM nitrocel-lulose membrane using the XCell-II Mini-Cell System(Novex, San Diego, CA) according to the manufactur-er’s instructions. The membrane was incubated with ablocking agent (5% Carnation nonfat dry milk in PBS)for 3 hr followed by overnight incubation with 20 mg/ml sheep anti-rat SGP-2 antibody (Quidel). The mem-brane was washed and incubated with horseradishperoxide-conjugated donkey anti-sheep IgG (SigmaChemical Co.) for 2 hr. The immunoreactive bandswere visualized with 3,38,5,58-tetramethylbenzidinemembrane peroxidase substrate (Kirkegaard & PerryLaboratories, Gaithersburg, MD).

88 Sintich et al.

Statistical Analysis

Numerical values were expressed as the mean ±standard error of the mean. Student’s t-test was car-ried out to determine differences among means. P <0.05 was considered statistically significant.

RESULTS

Differential Effects of TNF on PC3 and LNCaPCells

A 24-hr culture period was used in the presentstudy as a rapid and accurate method to assess thecytotoxic effect of TNF on prostatic cancer cells [7]. Asindicated in a previous report, during this period, adecrease in cell number correlates well with an in-crease in cell death as assessed by the release of radio-activity from cells prelabeled with 3H-thymidine, withthe extent of DNA fragmentation, and by the trypanblue dye exclusion test [7]. In the present study, resultsof both counting of cell number and percent of viablecells were reported for each treatment. Figure 1a indi-cates that treatment of LNCaP cells with 10 ng/mlTNF resulted in a 30% decrease in cell number. Treat-ment of PC3 cells with TNF under the same conditionshad no significant effect on cell number when com-pared to untreated controls. The trypan blue exclusionassay was used to verify that the decrease in cell num-ber was due to a reduction in cellular viability or anincrease in cytotoxicity. Figure 1b demonstrates that

the viability of LNCaP cells decreased by 40% withTNF treatment, while this treatment had no significanteffect on the viability of PC3 cells.

Levels of SGP-2 in Conditioned Media of LNCaPand PC3 Cells

The amount of SGP-2 present in samples of condi-tioned media was determined by immunoprecipita-tion and Western blot analysis. Immunoprecipitationwas utilized to specifically concentrate the SGP-2 pro-tein in the conditioned medium of PC3 and LNCaPcultures. Care was taken so that a constant amount ofthe antibody-coated sepharose beads was incubatedwith the same amount of protein (417 ng/ml). Figure2 reveals that PC3 cells (lane 1) produced more SGP-2than LNCaP cells (lane 2) during a 24-hr period, sincethe immunoreactive bands in lane 1 were more in-tensely stained than those in lane 2. The presence ofthree bands was expected, as human prostate cancercells express a precursor form and both the a and bsubunits of SGP-2, with molecular weights compa-rable to previously reported values [7].

Effect of Exogenous SGP-2 and TNF on LNCaPCells

LNCaP cells are exquisitely sensitive to the cyto-toxic effect of TNF. However, addition of exogenousSGP-2 will protect LNCaP cells from the effect of TNF.

Fig. 1. Effect of TNF on LNCaP and PC3 cells in culture. PC3 (j) and LNCaP (h) cells were cultured in the absence (−) or presenceof (+) TNF (10 ng/ml) for 24 hr. Each value is the average of quadruple replicates. Error bars denote standard error of the mean. a: Totalcell counts. There was a significant decrease (P < 0.05) in LNCaP cell number with TNF treatment. There was no significant differencebetween TNF-treated and control cultures for PC3 cells. b: Cell viability was assessed by trypan blue exclusion. LNCaP viability wassignificantly decreased (P < 0.05) with TNF treatment, while PC3 cell viability was not significantly effected by TNF treatment.

SGP-2 and Prevention of TNF-Induced Cytotoxicity 89

Figure 3 indicates that both LNCaP cell number (Fig.3a) and cell viability (Fig. 3b) decreased followingTNF treatment. This cytotoxic effect of TNF was di-minished by treatment with increasing doses of exog-enous SGP-2. This cytoprotective effect was dose-dependent, with high concentrations of SGP-2 (15 mg/ml) completely abolishing TNF-induced cytotoxicityas assessed by percent viability and total cell number.

Effect of Anti-SGP-2 Antibody on PC3 Responseto TNF

Under the present culture conditions, PC3 cellswere resistant to the cytotoxic effect of TNF. However,neutralization of SGP-2 with specific antiserum in PC3cultures will sensitize these cells to the cytotoxic ef-fects of TNF. Figure 4 shows that the addition of an-tiserum to SGP-2 caused a decrease in PC3 cell num-ber (Fig. 4a) and cell viability (Fig. 4b) in the presenceof 10 ng/ml of TNF. This sensitization to TNF-inducedcytotoxicity occurred in a dose-dependent fashion,with high concentrations of anti-SGP-2 (50 mg/ml)causing a 76% decrease in cell number and 77% de-crease in cell viability. This effect was specific to SGP-2antiserum, as treatment of PC3 cells with TNF in thepresence of a nonimmune IgG fraction was unable tosensitize the cells to TNF. The effect of SGP-2 antise-rum could be negated with the addition of exogenousSGP-2. In Figure 5, addition of exogenous SGP-2 coun-teracted the decrease in cell number caused by 10 mg/ml anti-SGP-2. Again, this effect was specific to SGP-2,

because the addition of nonspecific protein (sheepIgG) did not desensitize PC3 cells to the cytotoxic ef-fect of TNF.

DISCUSSION

Results of the present study demonstrated that thesensitivity to TNF-induced cytotoxicity in two pros-tatic cancer cell lines, PC3 and LNCaP, can be regu-lated by extracellular levels of SGP-2. Under the pres-ent culture conditions, LNCaP cells were sensitive tothe cytotoxic effect of TNF, while PC3 cells were re-sistant to it. The present results indicated that the sen-sitivity to TNF could be regulated by the level ofSGP-2 present in the culture medium. LNCaP cellsdeveloped resistance to TNF when the concentrationof SGP-2 in the culture was increased. Likewise, PC3cells developed sensitivity to TNF when the effectivelevel of SGP-2 in the culture was reduced by the use ofan anti-SGP-2 antibody. Since this antibody neutral-ized the effect of SGP-2, it most likely did not enter theintracellular space. Therefore, the present results haveadvanced our understanding that the protective actionof SGP-2 against cellular cytotoxicity is at least in partmediated through an extracellular mechanism.

The cytotoxic effect of TNF on cancer cells is well-known. Our previous studies demonstrated that TNFshowed variable degrees of cytotoxicity to differentlines of prostatic cancer cells but not to benign pros-tatic epithelial or stromal cells. LNCaP cells were mostsensitive to the cytotoxic effects of TNF, while PC3and DU145 cells were minimally sensitive [12]. PC3cells have demonstrated variable responses to TNF,depending on the time points sampled [15]. Conse-quently, in the present study we conducted all experi-ments after 24-hr treatment, since at this time pointPC3 cells are resistant to the cytotoxic effect of TNF. Itis now clear that the amount of SGP-2 produced byPC3 and LNCaP cells can partially determine the de-gree of sensitivity to the cytotoxic effects of TNF. In thepresent study, PC3 cells secreted a significantly higherlevel of SGP-2 than that produced by LNCaP cells and,therefore, PC3 cells are less sensitive to TNF-inducedcytotoxicity than LNCaP cells. Furthermore, additionof exogenous SGP-2 was able to reverse the sensitivityto TNF in LNCaP cells. Reduction of a protective levelof SGP-2 produced by PC3 cells by a specific antibodywas able to restore TNF sensitivity in PC3 cells. Theseresults are consistent with those of our previous stud-ies, in which an overexpression of SGP-2 in LNCaPcells protected them from TNF-induced apoptosis [7].These findings are also consistent with another studyin which increased levels of SGP-2 staining intensitycorrelated with increasing tumor grade in archivalprostate cancer specimens [16]. Taken together, an im-

Fig. 2. SGP-2 secretion by LNCaP and PC3 cells in culture. Cellswere cultured in serum-free media for 24 hr, and the collectedconditioned medium was concentrated as described in Materialsand Methods. SGP-2 was immunoprecipitated and subjected toSDS-PAGE, followed by Western blot analysis. Lane 1, PC3-conditioned medium; lane 2, LNCaP-conditioned medium. Upperarrow denotes the precursor form; lower arrow indicates the aand b subunits of human SGP-2.

90 Sintich et al.

portant biological role of SGP-2 in prostatic cancer ap-optosis and progression begins to emerge. Our datasuggest that cancer cells that express a large amount ofSGP-2 may have a greater ability to resist the cytotoxiceffect of TNF, thus displaying a more aggressive phe-notype.

Several other proteins have also been implicated inthe protection of cells from TNF-induced cytotoxicityin various model systems. These include the overex-pression of manganese superoxide dismutase, heatshock protein 70, A20 zinc finger protein, bcl-2, bcl-x,and functional p53 [17–20]. It is clear that cells may

Fig. 3. Effect of exogenous SGP-2 on TNF-treated LNCaP cells. Cells were cultured in the presence of TNF with varying doses of SGP-2for 24 hr. Cell number was measured by cell counting (a), while cell viability was assessed by trypan blue exclusion (b). Each value is theaverage of quadruple replicates. Error bars denote standard error of the mean. Values with asterisks are significantly greater than thosetreated with TNF alone (P < 0.05).

Fig. 4. Effect of SGP-2 antibody on TNF-treated PC3 cells. a: Cells were cultured in the presence of TNF (10 ng/ml) with varying dosesof SGP-2 antibody (aSGP-2, 10–50 µg/ml, h), or IgG (10–50 µg/ml, j) for 24 hr. Total cell number was determined by cell counting. b:Cell viability was determined by trypan blue exclusion in the presence of TNF (10 ng/ml, +) with increasing concentrations of SGP-2antibody (10–50 µg/ml, h) or IgG (10–50 µg/ml, j). Each value is the average of quadruple replicates. Error bars denote standard errorof the mean. In the presence of SGP-2 antibody, TNF treatment resulted in a significant decrease in cell count and cell viability. This effectof the antibody was specific, as the same dose of IgG had no effect on cell count or cell viability. Values with asterisks are significantly lowerthan those treated with TNF alone (P < 0.05).

SGP-2 and Prevention of TNF-Induced Cytotoxicity 91

employ a multitude of strategies to overcome the cy-totoxic effect of TNF. The role of SGP-2 in these alter-nate pathways remains to be determined.

SGP-2 is a sulfated glycoprotein which is usuallysecreted into the extracellular environment. Currently,the mechanism of SGP-2 action is unknown. In thetestis, SGP-2 is mainly secreted by Sertoli cells and byepididymal principal cells [5,21]. A layer of SGP-2 hasbeen seen to coat the surface of sperm in the repro-ductive tract [5]. It has been hypothesized that itsfunction is mediated through the stabilization of cellmembranes [6,22]. The present findings demonstratethat SGP-2 extracellularly protects cells from the cyto-toxic effects of TNF. The following are some specula-tions as to its possible sites of action in protection fromcellular apoptosis. First, SGP-2 has been characterizedas a “sticky” protein: its protective effect may be me-diated through direct interaction with the TNF mol-ecules, rendering the cytokine unavailable to targetcells. Another possibility is that SGP-2 may act at thelevel of TNF receptors. If SGP-2 competes with TNFfor its 55-kDa receptor [23], it could protect target cellsfrom the cytotoxic effect of TNF. A third possibility isthat SGP-2 may interact with cell surface moleculessuch as sphingomyelin and interfere with TNF signaltransduction [24]. These speculations will be the topicsof our future investigations.

CONCLUSIONS

The present results demonstrate that the sensitivityof prostatic cancer cells to the cytotoxic effect of TNFcan be regulated by extracellular levels of SGP-2. Thisfinding has provided additional support for the beliefthat SGP-2 has a protective effect on target cells tocytotoxicity induced by TNF. Furthermore, the pre-sent observations suggest that the cellular membraneis a potential site of action for SGP-2. We propose thatthe mechanism of SGP-2 protection from TNF-induced cytotoxicity is extracellular.

ACKNOWLEDGMENTS

This work was supported in part by NIH grant CA69851 to J.A.S., by NIH grant CA 09560 to S.M.S, andby a gift to the Abraham “Terry” Rogovin UrologicalOncology Laboratory, Lurie Cancer Center, North-western University Medical School.

REFERENCES

1. Fritz IB, Burdzy C, Setchell B, Blaschuk O. Ram rete testis fluidcontains a protein (clusterin) which influences cell-cell interac-tions in vitro. Biol Reprod 1983;28:1173–1188.

2. Bettuzzi S, Hiipakka RA, Gilna P, Liao S. Identification of anandrogen-repressed mRNA in rat ventral prostate as coding forsulfated glycoprotein-2 by cDNA cloning and sequence analy-sis. Biochem J 1989;257:293–296.

3. Lee C, Sensibar J. Proteins of the rat prostate. II. Synthesis ofnew proteins in the ventral lobe during castration-induced re-gression. J Urol 1987;138:903–908.

4. Jenne DE, Tschopp J. Molecular structure and functional char-acterization of a human complement cytolysis inhibitor found inblood and seminal plasma: identity to sulfated glycoprotein-2, aconstituent of rat testis fluid. Proc Natl Acad Sci USA 1989;86:7123–7127.

5. Sylvester SR, Morales C, Oko R, Griswold M. Localization ofsulfated glycoprotein-2 (clusterin) on spermatozoa and in thereproductive tract of the male rat. Biol Reprod 1991;45:195–207.

6. Jordan-Stark TC, Witte DP, Aronow BJ, Harmony JAK. Apoli-poprotein J: a membrane policeman. Curr Opin Lipidol 1992;3:75–85.

7. Sensibar JA, Sutkowski DM, Raffo A, Buttyan R, Griswold MD,Sylvester SR, Kozlowski JM, Lee C. Prevention of cell deathinduced by tumor necrosis factor alpha in LNCaP cells by over-expression of sulfated glycoprotein-2 (clusterin). Cancer Res1995;55:2431–2437.

8. Buttyan R, Olsson CA, Pintar J, Chang C, Bandyck M, Ng P-Y,Sawczuk I. Induction of the TRPM-2 gene in cells undergoingprogrammed death. Mol Cell Biol 1989;9:3473–3481.

9. Duguid JR, Bohmont N, Liu N, Tourtellotte WW. Changes inbrain gene expression shared by scrapie and Alzheimer’s dis-ease. Proc Natl Acad Sci USA 1989;86:7260–7264.

10. Kyprianou N, English HF, Isaacs JT. Programmed cell deathduring regression of PC-82 human prostate cancer followingandrogen ablation. Cancer Res 1990;50:3748–3753.

11. Pasientti GM, Finch CE. Sulfated glycoprotein-2 (SGP-2) mRNAis expressed in rat striatal astrocytes following ibotenic acidlesions. Neurosci Lett 1991;130:1–4.

Fig. 5. Effect of exogenous SGP-2 and antibody to SGP-2 onTNF-treated PC3 cells. Cells were cultured in the presence ofTNF (10 ng/ml) and antisera against SGP-2 (+, 10 µg/ml) andvarying doses of exogenous SGP-2 (µg/ml, h) or exogenous IgG(µg/ml, j) for 24 hr. Each value is the average of quadruple rep-licates. Error bars denote standard error of the mean. ExogenousSGP-2 at the highest dose (20 µg/ml) was able to reverse the effectof the antibody (10 µg/ml).

92 Sintich et al.

12. Sherwood E, Ford T, Lee C, Kozlowski J. Therapeutic efficacy ofrecombinant tumor necrosis factor a in an experimental modelof human prostatic carcinoma. J Biol Response Mod 1990;9:44–52.

13. Bradford MA. Rapid and sensitive method for the quantitationof microgram quantities of protein utilizing the principle of pro-tein dye-binding. Anal Biochem 1976;72:248–251.

14. Laemmli UK. Cleavage of the structure proteins during assem-bly of the head of bacteriophage T4. Nature 1970;227:680–685.

15. Rokhin OW, Bishop GA, Hostager BS, Waldschmidt TJ,Sidorenko SP, Pavloff N, Kiefer MC, Umansky SR, Glover RA,Cohen MB. Fas-mediated apoptosis in human prostatic carci-noma cell lines. Cancer Res 1997;57:1758–1768.

16. Steinberg J, Oyasu R, Lang S, Rademaker A, Lee C, KozlowskiJM, Sensibar JA. Intracellular levels of SGP-2 (clusterin) corre-late with tumor grade in prostate cancer. Clin Cancer Res 1997;3:1707–1711.

17. Wong GH, Elwell JH, Oberley LW, Goeddel DV. Manganoussuperoxide dismutase is essential for cellular resistance to cyto-toxicity of tumor necrosis factor. Cell 1989;58:923–931.

18. Opipari AW, Hu HM, Yabkowitz R, Dixit VM. The A20 zincfinger protein protects cells from tumor necrosis factor cytotox-icity. J Biol Chem 1992;267:12424–12427.

19. Jaattela M, Benedict M, Shayman JA, Dixit VM. Bcl-x and Bcl-2

inhibit TNF and Fas-induced apoptosis and activation of phos-pholipase A2 in breast carcinoma cells. Oncogene 1995;10:2297–2305.

20. Cai Z, Capoulade C, Moyret-Lalle C, Amor-Gueret M, FeunteunJ, Larsen AK, Bressac-de Paillerets B, Chouaib S. Resistance ofMCF7 human breast carcinoma cells to TNF-induced cell deathis associated with loss of p53 function. Oncogene 1997;15:2817–2826.

21. Sensibar JA, Qian Y, Griswold MD, Sylvester SR, Bardin CW,Cheng CY, Lee C. Localization and molecular heterogeneity ofsulfated glycoprotein-2 (clusterin) among ventral prostate,seminal vesicle, testis, and epididymis of rats. Biol Reprod 1994;49:233–242.

22. French LE, Sappino AP, Tschopp J, Schifferli JA. Distinct sites ofproduction and deposition of the putative cell death markerclusterin in human thymus. J Clin Invest 1992;90:1919–1925.

23. Yanaga F, Watson SP. Tumor necrosis factor a stimulates sphin-gomyelinase through the 55 KDa receptor in HL-60 cells. FEBSLett 1992;314:297–300.

24. Kolesnick RN, Haimovitz-Friedman A, Fuks Z. The sphingomy-elin signal transduction pathway mediates apoptosis for tumornecrosis factor, Fas, and ionizing radiation. Biochem Cell Biol1994;72:471–474.

SGP-2 and Prevention of TNF-Induced Cytotoxicity 93