Immunosensitization of Prostate Carcinoma Cell Lines for Lymphocytes (CTL, TIL, LAK)-Mediated Apoptosis via the Fas–Fas-Ligand Pathway of Cytotoxicity

Immunosensitization of Prostate Carcinoma Cell Lines for Lymphocytes (CTL, TIL, LAK)-Mediated Apoptosis via the Fas–Fas-Ligand Pathway of Cytotoxicity

CELLULAR IMMUNOLOGY ARTICLE NO. 180, 70–83 (1997) CI971169 Immunosensitization of Prostate Carcinoma Cell Lines for Lymphocytes (CTL, TIL, LAK)-Med...

781KB Sizes 0 Downloads 22 Views

CELLULAR IMMUNOLOGY ARTICLE NO.

180, 70–83 (1997)

CI971169

Immunosensitization of Prostate Carcinoma Cell Lines for Lymphocytes (CTL, TIL, LAK)-Mediated Apoptosis via the Fas–Fas-Ligand Pathway of Cytotoxicity Patrick Frost,* Chuen Pei Ng,* Arie Belldegrun,† and Benjamin Bonavida* *Department of Microbiology and Immunology, and †Department of Urology, UCLA School of Medicine and Jonsson Comprehensive Cancer Center, University of California, Los Angeles, California 90095-1747 Received December 23, 1996; accepted June 25, 1997

lator EGTA/MgCl2 . Altogether, these findings show that drug-resistant, Fas/-expressing PC-3 and DU145 prostate tumor cells can be sensitized by CDDP and VP-16 to killing by Fas-L-bearing CTL, TIL, and LAK cells. Sensitization of tumor cells by drugs may augment the efficacy of immunotherapy in the eradication of tumor cells that are resistant to Fas-L-mediated killing. q 1997 Academic Press

Several reports suggest that immunotherapy mediated by cytotoxic lymphocytes is beneficial in the destruction of drug-resistant tumor cells. Cytotoxic T lymphocytes kill target cells by two main mechanisms, namely by the perforin pathway and by the Fas-ligand (Fas-L) pathway. The role of the Fas-L pathway in tumor cell killing is not clear because many Fas/-expressing tumor cells are resistant to the Fas-L agonist cytotoxic anti-Fas antibody. The human prostate tumor cell lines (PC-3, DU145, and LnCAP) express Fas on the cell surface but are resistant to killing by antiFas antibody. This study examined the sensitivity of prostate tumor cells to Fas-L-mediated cytotoxicity and sensitization of the tumor cells by drugs to Fas-Lmediated killing. All three prostate tumor cell lines are resistant to Fas-L killing as determined by the use of the murine CTL hybridoma PMMI that kills only through the Fas-L pathway. However, the addition of subtoxic concentrations of CDDP or VP-16 significantly sensitized the PC-3 and DU145, but not LnCAP, tumor cells to Fas-L killing and apoptosis by PMMI. The sensitization of tumor cells by drugs was inhibited by neutralizing anti-Fas antibody. These findings demonstrate that immunoresistant Fas/-expressing DU145 and PC-3 prostate tumor cells can be sensitized by drugs to Fas-L killing. We then examined the role of Fas-L killing by TIL and LAK cells. All three prostate tumor cell lines were sensitive to killing by TIL and LAK and cell killing was primarily mediated through the Ca2/-dependent perforin pathway because it was blocked by the addition of EGTA/MgCl2 . Sensitization by CDDP or VP-16 did not significantly augment killing of untreated tumor cells by TIL or LAK cells. However, in the presence of EGTA/MgCl2 , the addition of CDDP or VP-16 significantly augmented killing of PC3 and DU145, but not LnCAP, by TIL and LAK, and killing was blocked by neutralizing anti-Fas antibody. These findings demonstrate that both TIL and LAK exhibit a Fas-L-mediated killing pathway that is revealed once the perforin pathway is blocked by the Ca2/ che-

CI 1169

/

6c27$$$$61

One of the primary problems facing the treatment of cancer is the development of drug-resistant tumors (1, 2). For example, prostate cancer is the major form of cancer found in men in the United States (3). Prostate cancer is typically treated by hormonal ablation and with chemotherapeutic drugs (4). The initial effectiveness of these treatments is high; however, relapses of hormonally and drug-resistant tumors frequently occur (4). Furthermore, additional chemotherapy is rarely successful. Therefore, development of new and effective treatments for prostate cancer is needed. The development of tumor cell resistance to conventional therapy has led to the exploration of the immune system in eradicating resistant tumor cells. Various genetic approaches have been devised in the effort to elicit a specific antitumor immune response (5, 6). Several of the gene manipulation techniques have been shown to be successful in experimental model systems and a few are currently being explored clinically (7, 8). While many efforts have been directed at the induction of the effector phase of the cytotoxic response, little has been done to consider the susceptibility of the target cells to cytotoxic cells. It is reasonable to assume that many tumor cell types or subpopulations are resistant to cytotoxic lymphocytes and/or acquire resistance following therapy. Thus, the development of immune resistance by tumor cells may mimic the problems encountered with the acquisition of drug resistance. 70

0008-8749/97 $25.00 Copyright q 1997 by Academic Press All rights of reproduction in any form reserved.

AID

INTRODUCTION

09-08-97 09:09:24

cias

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

Therefore, it is important that new means are considered to overcome the immunoresistance of tumor cells. Cytotoxic T lymphocytes have been shown to kill target cells in a short time by two main mechanisms, namely the perforin and the Fas-ligand (Fas-L) pathways (9, 10). How these two pathways are activated and whether certain targets stimulate preferentially one and/or the other pathway are not clear. Several tumor cells express the Fas-R antigen on their cell surface but not all are sensitive to killing by anti-Fas antibody and presumably by Fas-L on lymphocytes (11– 13). Thus, it may be conceivable that certain Fas/ tumor cells that are resistant to Fas-L killing may also be resistant to CTL-mediated immunotherapy. Further, if such tumor cells either are resistant to the perforin pathway or preferentially trigger only the Fas-L pathway, such tumors will not be affected by any modalities that generate antitumor cytotoxic lymphocytes. Therefore, one must find means to sensitize such tumor cells in order to render them sensitive to killing by the FasL-mediated cytotoxic pathway. Recent studies from our laboratory have demonstrated that certain drugs can sensitize ovarian tumor cell lines to anti-Fas antibody-mediated cytotoxicity and apoptosis (12). In this study, we investigated the sensitization of three human prostate carcinoma cell lines to killing by the Fas-L pathway. This study was designed to investigate the following questions: (1) whether prostate carcinoma cell lines can be sensitized by chemotherapeutic drugs to Fas-L-mediated cytotoxicity by CTL; (2) whether TIL utilize the Fas-L pathway of cytotoxicity and whether prostate carcinoma cell lines can be sensitized to Fas-L killing by TIL; and (3) whether LAK cells exhibit a Fas-L component of cytotoxicity that is also regulated by sensitized prostate tumor cells.

71

centrifugation over Ficoll and washed three times in sterile PBS. Cells were incubated in the presence of 1000 U/ml of rhIL-2 for at least 7 days prior to use. Human TIL cells were supplied by Dr. Belldegrun and prepared as previously described (16). TIL were cultured from renal cell cancer tumors. Briefly, tumors were obtained from the operating room, minced, and digested overnight in RPMI 1640 (Cellgro, Washington, DC) with 0.01% hyaluronidase type V, 0.002% DNase type II, 0.1% collagenase type IV (Sigma Chemical Co., St. Louis, MO) 2 mM L-glutamine (Gibco, Grand Island, NY), and 50 mg/ml gentamycin. Singlecell suspensions were centrifuged over single-step Ficoll – Hypaque density gradients (LSM, Durham, NC). TIL and tumor cells were retrieved from the gradient interfaces, washed, counted, and cultured in 75-ml tissue culture flasks (Costar, Cambridge, MA) at a density of 0.5 1 106 cells/ml in RPMI 1640 medium plus 10% heat-inactivated human AB serum (Irvine Scientific, Santa Ana, CA), 50 IU/ml penicillin, 50 mg/ml streptomycin (JHR Biosciences, Lenexa, KS), 2 mM L-glutamine, and 400 U/ml interleukin-2 (Hoffman – LaRoche, Nutley, NJ). Cultures were maintained at 377C and 5% CO2 . All cells were cultured in 10% heat-inactivated fetal calf serum (Atlanta Biologicals) added to RPMI 1640 medium (Gibco) with 1% pyruvate (Gibco), 1% nonessential amino acids (Gibco), and 1% Fungi-bact solution (Irvine Scientific), which contains 10,000 U/ml penicillin G, 10 mg/ml streptomycin, and 25 mg/ml Fungizone. All cell lines were grown in a humidified atmosphere at 377C in 5% CO2 . Reagents

The PMMI murine CTL hybridoma is derived from BALB/PEL specific for the H-2b thymoma EL-4 (14). DT140 and LF0 cells are derived from the murine lymphoma cell line L1210. DT140 is transfected with a Fas overexpression construct, and LF0 is transfected with a Fas antisense construct (15). The human hormone-independent prostatic carcinoma cell lines, DU145 and PC-3, and hormone-dependent prostatic carcinoma cell line, LnCAP, were obtained from Dr. Belldegrun, and were maintained in culture as adherent cells. Human rhIL-20, activated LAK, and TIL cells were prepared from normal human donors. For LAK cells, whole blood was collected into 35-ml syringes with 1.5 ml sterile heparin. PBMCs were isolated by density

Actinomycin D, cisplatin (CDDP), etoposide (VP-16), ionomycin, and PMA were purchased from Sigma Chemical Co. Na251CrO4 was purchased from Amersham. The cytotoxic anti-Fas monoclonal antibody (IgM, clone CH-11), the neutralizing monoclonal antibody (IgG1 , clone ZB4), and the surface-staining monoclonal antibody (IgG1 , clone UB2) were purchased from Kamiya Biomedical Co. (Thousand Oaks, CA). The anti-CD95 Ig antibody clone B-G27, was obtained from Diaclone Research (Besanc¸on, France). The hamster anti-mouse Fas antibody (IgG, Jo2) and normal hamster IgG were purchased from PharMingen (San Diego, CA). Normal mouse IgM and mouse IgG1 were purchased from Amac (Westbrook, ME). Phycoprobe PEconjugated goat anti-mouse IgG (Biomeda Co., CA) and PE-conjugated anti-hamsters IgG (PharMingen) were used as a secondary antibody. Stock solutions of reagents were routinely prepared in PBS, medium, DMSO, or EtOH as appropriate.

AID

cias

MATERIALS AND METHODS Cell Lines

CI 1169

/

6c27$$$$62

09-08-97 09:09:24

72

FROST ET AL.

RT–PCR /

Poly(A) RNA from TIL cells or LAK cells was purified using oligo(dt)25-bound Dynabeads according to the instruction manual and subjected to cDNA synthesis. PCR reaction was performed from 5 1 104 cells with equivalent cDNA under the following conditions: 1 cycle at 947C for 3 min, 607C for 2 min, and 727C for 3 min; 28 or 38 cycles at 947C for 1 min, 607C for 2 min, and 727C for 3 min; 1 cycle at 947C for 1 min, 607C for 2 min, and 727C for 7 min. Ten microliters of PCR reaction was subjected for analysis. The PCR was performed using the following upstream (U) and downstream (D) primers: Fas-ligand (U) 5*-CCTGACTCACCAGCTGCCATGC-3*; Fas-ligand (D) 5*-CTCTTAGAGCTTATATAGCCG-3*; b-actin (U) 5*-ATCTGGCACCACACCTTCTACAATGAGCTGCG-3*; b-actin (D) 5*-CGTCATACTCCTTGCTTGCTGATCCACATCTGC-3*. CTL-Mediated Cytotoxicity PMMI cells were activated in the presence of 10 ng/ ml PMA and 2 mg/ml ionomycin and incubated for 3 hr at 377C in 5% CO2 (15). At the end of the incubation period, the cells were washed once in PBS and resuspended at a final concentration of 1 1 106 cells/ml and used immediately in the cytotoxicity assay. TIL and LAK cells were washed three times in PBS and resuspended at a final concentration of 1 1 106 cells/ml and used immediately in the cytotoxicity assay. DT140 and LF0 nonadherent cells were grown overnight in the presence or absence of drug for 18 hr at 377C in 5% CO2 and then collected and washed once in PBS. The cells were incubated in 100 mCi of Na251CrO4 for 1 hr at 377C in 5% CO2 and then washed three times in medium, and 104 cells were added to V-bottom 96well culture plates and used immediately in the cytotoxic assay. DU145, LnCAP, and PC-3 adherent cells were trypsinized for 5 min, collected, and washed once in PBS. The cells were incubated in 100 mCi of Na251CrO4 for 1 hr at 377C in 5% CO2 and then washed three times in medium, and 104 cells were added to flat-bottom 96well culture plates in the presence or absence of drug. The plates were allowed to incubate for 18 hr at 377C in 5% CO2 . Two hours prior to addition of effector cells, the supernatant was removed and 100 ml fresh medium plus or minus 1 mg/ml of neutralizing antibody (ZB4, MBL, MA, or B-G27, Diaclone Research) was added to the cells and incubated at 377C in 5% CO2 . At the time of the experiment, 100 ml of effector cells in the presence or absence of 3 mM EGTA/2 mM MgCl2 was added at the indicated E:T ratio. The plates were centrifuged and incubated for 5–7 hr at 377C in 5% CO2 . Following the incubation, 100 ml of supernatant was harvested from each well and counted in a Beckman gamma-4000 gamma counter. Total 51Cr release was determined by

AID

CI 1169

/

6c27$$$$62

09-08-97 09:09:24

lysing target cells with 50 ml of 10% SDS buffer and collecting 150 ml for count. The percentage-specific 51Cr release was determined as follows: % Å

51

Cr release

experimental release 0 spontaneous release 1 100. total release 0 spontaneous release

All values are presented as the mean { SD of triplicate samples. XTT Cytotoxicity Assay Cytotoxicity was assessed using the XTT assay kit (Boehringer-Mannheim), which measures the metabolic activity of viable cells. Prostate tumor cells were plated in a 96-well plate in 50 ml RPMI 1640, 10% FBS medium at 1 1 104 cells/well. Twenty-five microliters of various concentrations of chemotherapeutic drugs (ADR, VP-16, and CDDP) and 25 ml of anti-Fas IgM antibody clone CH11 (Kamiya, Seattle, WA) were then added to the tumor cells, giving a final volume of 100 ml/well. Positive control cells were left untreated and substituted with 50 ml of medium. For negative control, 100 ml of the medium alone without cells was used. The cells were incubated for 24 hr at 377C for maximal killing. The next day, XTT-labeling mixture was prepared by mixing XTT-labeling reagent and electroncoupling in 50:1 ratio, and 50 ml of the mixture was added to each well. After 2 hr of incubation in XTTlabeling mixture at 377C, absorbance readings were taken at 490 nm using a multiwell spectrophotometer (Titertek Multiscan MCC/340). The percentage cytotoxicity was calculated using the background-corrected readings as follows: % cytotoxicity Å [1 0 (OD of experimental well/OD of positive control well)] 1 100. Acrydine Orange/Ethidium Bromide Staining Apoptotic cell death was determined by the dye-exclusion method. DU145 cells were incubated in the presence or absence of 10 mg/ml CDDP or vehicle for 18 hr at 377C and 5% CO2 . Following the incubation, the cells were harvested, washed two times in medium / 10% FBS, and resuspended at 1 1 106 cells/ml. One hundred microliters of cell suspension was added to a 15-ml centrifuge tube in the presence or absence of 1 ml of 1 1 106 activated PMMI cells. The cells were then centrifuged at 1000 rpm/5 min and allowed to incubate for 5 hr at 377C and 5% CO2 . At the end of incubation, the supernatant was removed and 100 ml of fresh medium was added. To visualize apoptosis, 25 ml of cell suspension was collected and 1 ml of acrydine orange/ ethidium bromide was added. Cells were observed using a Zeiss fluorescent microscope.

cias

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

Flow Cytometric Analysis The expression of the Fas and Fas-L antigen on cells was determined by flow cytometry. Staining was performed in U-bottom, 96-well culture plates. Dilutions and washings were performed with PBS containing 0.1% sodium azide. Before they were stained, the cells (5 1 105) were washed once and resuspended in 100 ml of PBS. The human tumor cells DU145 and PC-3 were incubated for 1 hr with 1 mg/ml of mouse anti-human Fas (IgG1 clone UB2) or 1 mg/ml mouse IgG1 . The human tumor cells, LnCAP, were incubated for 1 hr with 1 mg/ml of mouse anti-human Fas (clone CH11) or 1 mg/ml mouse IgM. For Fas-L expression, TIL cells were incubated for 30 min with 1 mg/ml of hamster antihuman Fas-L (clone 4H9) or hamster IgG control. The cells were washed three times in PBS and resuspended in 100 ml PBS containing goat anti-mouse PE-conjugated IgG or IgM antibody or goat anti-hamster PEconjugated IgG antibody for an additional hour. The cells were then washed three times in PBS and fixed in 2% paraformaldehyde solution. The murine lymphoma cells, DT140 and LF0, were incubated for 1 hr with 1 mg/ml PE-conjugated hamster anti-mouse Fas (clone Jo2, PharMingen). The cells were then washed three times in PBS and fixed in 2% paraformaldehyde solution. Flow cytometry was performed using an Epic C flow cytometer.

73

In order to ascertain whether sensitization of Fas/ tumor cell killing to anti-Fas antibody can also take place through the Fas-L pathway, we initially used a well-defined murine system. Both Fas-negative (LF0) cells and Fas/-transfectant (DT140) cells (Fig. 1), derived from L1210, were used as target cells. Further, a murine CTL hybridoma, PMMI, has been shown to kill only by the Fas-L pathway and is not species specific (15). Figure 2 demonstrates that DT140, but not LF0, target cells are killed by activated PMMI and the extent of killing is a function of the E:T ratio used. Further, treatment of DT140 with CDDP augmented the cytotoxic activity and the augmentation was dependent on the concentration of CDDP used. In contrast, the Fas-negative LF0 cell line was not sensitized by CDDP to killing by PMMI. The parental cell line L1210, which expresses an intermediate level of Fas between DT140 and LF0, was also sensitized by CDDP to killing by PMMI (data not shown). These findings demonstrate that CDDP can sensitize Fas / target cells to killing by Fas-L-bearing CTL. Sensitization of Human Prostate Carcinoma Cell lines to Fas-L-Mediated Killing by PMMI

We first examined if human prostate cell lines can be sensitized to killing by anti-Fas antibody. All three prostate lines, PC-3, DU145, and LnCAP, express the Fas receptor, albeit with different degrees (Fig. 1). Initially, PC-3 and DU145 cells were resistant to both CDDP and anti-Fas antibody. However, pretreatment with CDDP, followed by anti-Fas antibody, resulted in a significant augmentation of cytotoxicity (Table 1). LnCAP tumor cells, which express low levels of Fas receptor, were resistant to killing by anti-Fas antibody and were not sensitized by drug treatment (data not shown). Similar findings of sensitization by CDDP were observed with ADR and VP-16 (data not shown). These findings demonstrate that CDDP can sensitize PC-3 and DU145 to anti-Fas antibody-mediated cytotoxicity.

Based on the above findings with murine targets, we examined whether human prostate tumor cell lines can also be sensitized to killing via the Fas-L pathway. All three lines, PC-3, DU145, and LnCAP, express the Fas receptor, albeit to different degrees (Fig. 1). We examined the effect of CDDP and VP-16 on PC-3 and DU145 tumor cell killing by PMMI. Figure 3 demonstrates that PC-3 and DU145 are resistant to killing by PMMI; however, treatment with CDDP or VP-16 significantly potentiated killing and the extent of the killing was a function of both the drug concentration and the E:T ratio used. In order to confirm that PMMI-mediated killing of sensitized target cells was through apoptosis, we stained DU145 cells in the presence or absence of activated PMMI cells with acrydine orange/ethidium bromide. Acrydine orange stains living cells green with a green nucleus, while ethidium bromide is uptaken by dead cells (apoptotic and necrotic) and stains the nuclei orange (17). In apoptotic cells, the cytoplasm stains green and the condensed nuclei stains bright orange. In necrotic cells, the chromatin does not condense and the nuclei and cytoplasm stain orange. Thus, it is possible to differentiate between living, apoptotic, and necrotic cells based on the staining pattern. DU145 control (Fig. 4A) and CDDP-treated DU145 (Fig. 4B) cells do not exhibit either apoptotic or necrotic morphology. Furthermore, control DU145 cells are not killed by activated PMMI (Fig. 4C). In contrast, about 10–15% of CDDP-treated DU145 cells incubated with activated

AID

cias

Statistical Analysis All assays were set up in triplicates, and the results were expressed as the mean { SD. Statistical analysis was determined by the Student t test. RESULTS Chemosensitization of Fas/ Murine Tumor Cells to Killing by Anti-Fas Antibody and by Fas-LExpressing Cytotoxic Lymphocytes

CI 1169

/

6c27$$$$62

09-08-97 09:09:24

74

FROST ET AL.

FIG. 1. Flow cytometry analysis of target cells for expression of Fas following CDDP treatment. Human PC-3 and DU145 were stained with mouse anti-human UB4 (IgG) antibody and then goat anti-mouse IgG–PE-conjugated secondary antibody. LnCAP cells were stained with mouse anti-human CH-11 (IgM) antibody and then goat anti-mouse IgM–PE secondary antibody. Murine DT140 and LF0 cells were stained with hamster anti-mouse Fas Jo2-PE (IgG) antibody. Untreated cells and cells treated with different concentrations of CDDP for 18 hr were analyzed. Isotype control (A, F, K, P, U), anti-Fas alone (B, G, L, Q, U), 0.1 mg/ml CDDP (M), 1 mg/ml CDDP (C, H, V, N, R, W), 3 mg/ml CDDP (S, X), 5 mg/ml CDDP (D, I), 10 mg/ml CDDP (E, J, O, T, Y).

PMMI CTLs exhibit the characteristic apoptotic morphology as determined by bright orange areas of condensed chromatin (17). The sensitization of tumor cells by drugs to Fas-mediated killing was corroborated by the demonstration that neutralizing anti-Fas antibody significantly inhibited the cytotoxic activity (Fig. 5). The hormonal-dependent LnCAP tumor cell line was

TABLE 1

shown to express a low level of Fas but was resistant to killing by anti-Fas antibody even if pretreated with drugs. We expected that this tumor cell line will not be sensitized to PMMI killing. Indeed, LnCAP was not killed even in the presence of high concentrations of CDDP (data not shown). These findings demonstrate that PMMI can kill Fas/ human PC-3 and DU145 cells following treatment with the sensitizing drugs. However, the LnCAP cells were resistant to sensitization.

Sensitization of DU145 and PC-3 to Killing by Anti-Fas Antibody

Cell line

CDDP (mg/ml)

Anti-Fas IgM (CH11) (ng/ml)

DU145

0 1 1 5 5 10 10 0 1 1 5 5 10 10

100 0 100 0 100 0 100 100 0 100 0 100 0 100

PC-3

% Cytotoxicity { SDa 8.0 0.0 22.5 11.0 72.2 29.7 73.2 17.7 20.4 9.6 15.6 81.7 13.2 80.9

{ { { { { { { { { { { { { {

4.6 14.6 5.5 12.2 6.0 7.3 1.4 7.3 14.0 3.0 15.7 2.2 5.7 2.2

a

Cytotoxicity was determined by the XTT assay as described under Materials and Methods.

AID

CI 1169

/

6c27$$$$63

09-08-97 09:09:24

FIG. 2. Fas-ligand-mediated cytotoxicity of murine PMMI cells versus CDDP-treated DT140 and LF0 cell lines. The percentage cytotoxicity was measured using the 51Cr release assay at indicated E:T ratios in 96-well plates. Target cells were treated overnight in the presence of CDDP or DMSO prior to assay. Spontaneous release of control cells treated with drugs alone was below 10%.

cias

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

75

FIG. 3. Fas-ligand-mediated cytotoxicity of murine PMMI cells versus CDDP or VP-16-treated human PC-3 (A and B) and DU145 (C and D) prostate cell lines. The percentage cytotoxicity was measure using 51Cr release assay at indicated E:T ratio in 96-well plates. Target cells were treated overnight in the presence of drug or DMSO prior to assay. Spontaneous release of control cells treated with drugs alone was less than 5%.

Sensitization of PC-3 and DU145 Prostate Carcinoma Cells to Killing by TIL Cells The above findings demonstrate that sensitized prostate carcinoma cell lines are killed by Fas-L-bearing murine CTL. These findings corroborated that killing by the Fas–Fas-L pathway is independent of MHC compatibility. Therefore, we considered that human TIL propagated in vitro might also kill through the Fas-L pathway in a non-MHC-restricted manner. TIL cells were derived from patients with renal carcinoma and propagated in vitro (16). PCR analysis demonstrated that TIL express the Fas-L message (Fig. 6A, Lane 2) and protein (Fig. 6B). TIL cells killed effectively DU145 tumor cells and killing was slightly augmented by treatment with CDDP (Fig. 7A) or VP-16 (Fig. 7C). To delineate whether the augmented killing that was achieved was due to Fas–Fas-L-mediated cytotoxicity, we treated the TIL with EGTA/MgCl2 , which blocks the Ca2/-dependent perforin pathway but not the FasL pathway (18). As shown in Figs. 7B and 7D treatment

with EGTA/MgCl2 significantly reduced the cytotoxic activity against untreated DU145 tumor cells. However, the addition of CDDP or VP-16 significantly augmented the cytotoxic activity and the extent of killing was a function of both drug concentrations and E:T ratios used. To further corroborate the role of Fas–FasL in killing by TIL, the cultures were treated with the neutralizing anti-Fas (ZB4) antibody. Clearly, the addition of ZB4 inhibited the sensitization of DU145 by CDDP and inhibited killing in the EGTA/MgCl2treated cultures (Fig. 8B). The failure to sensitize LnCAP to killing by PMMI was confirmed with TIL (data not shown). These findings demonstrate that sensitization of DU145 and PC-3 by drugs render these tumor cells more susceptible to killing by TIL. Furthermore, the findings with EGTA/MgCl2 demonstrate that, while untreated TIL reveal a minor component of killing by the Fas-L, the Fas-L pathway can be potentiated in drugtreated cells.

AID

cias

CI 1169

/

6c27$$$$63

09-08-97 09:09:24

76

FROST ET AL.

FIG. 4. Apoptotic killing of CDDP-treated DU145 cells by activated PMMI. Apoptosis was determined by acrydine orange/ethidium bromide dye exclusion. Target cells were treated overnight in the presence of 10 mg/ml CDDP or DMSO prior to assay. Target cells alone (A, B) or in the presence of activated PMMI (10:1) were incubated for 5 hr (C, D). Apoptotic cells were characterized by condensed chromatin stained bright orange. This figure is representative of five experiments.

AID

CI 1169

/

6c27$$1169

09-08-97 09:09:24

cias

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

FIG. 4.—Continued

AID

CI 1169

/

6c27$$1169

09-08-97 09:09:24

cias

77

78

FROST ET AL.

FIG. 5. Blocking of Fas-L-mediated cytotoxicity of murine PMMI cells versus CDDP-treated human DU145 (A) and PC-3 (B) prostate cell lines. The percentage cytotoxicity was measured using the 51Cr release assay at indicated E:T ratio in 96-well plates. Target cells were treated overnight in the presence of drug or DMSO prior to assay. Two hours prior to addition of effector cells, CDDP was washed from target cells and fresh medium containing 1 mg/ml of the neutralizing anti-Fas antibody, ZB4, or isotype control was added. Spontaneous release of control cells treated with drugs alone was less than 1% of untreated cells. The percentage cytotoxicity of drug / anti-Fas antibodytreated cells (j) was significantly less than that of drug-treated cells (l) alone at the 10:1 E:T ratio (P õ 0.05). This figure is representative of four experiments. Spontaneous release of control cells treated with drugs alone was less than 1% of untreated cells.

Sensitization of DU145 and PC-3 Tumor Cells to Killing by LAK Cells The above findings demonstrated that TIL activity can be potentiated by drug-sensitized tumor cells. We

examined whether LAK cells also exert the same effect. Unlike TIL, LAK are more easily available and easily propagated in culture and offer a good source of antitumor cytotoxic lymphocytes. Like TIL, LAK also expresses the Fas-L mRNA (Fig. 6, Lane 1). Human LAK cells effectively kill DU145 tumor cells (Fig. 9A), and in the presence of EGTA/MgCl2 , most of the killing is abrogated (Fig. 9B). In contrast, EGTA/MgCl2-treated LAK revealed a significant killing in the presence of CDDP (Fig. 9B) and killing was inhibited by neutralizing anti-Fas antibody (Fig. 9D). Like the above findings with PMMI and TIL, the LnCAP cells were not sensitized to killing by the Fas-L pathway of cytotoxicity mediated by LAK (Fig. 10B). These findings demonstrate that LAK cells can kill by the Fas–Fas-L pathway and sensitization of prostate tumor cells reveals the Fas-L pathway in LAK cells. DISCUSSION

FIG. 6. Fas-ligand expression in IL-2 activated human TIL and LAK cells. Fas-ligand expression was assayed using PCR (A) and flow cytometry (B). (A) M, molecular weight standards; Lane 1, LAK cells; Lane 2, TIL cells; Lane 3, b-actin. (B) Top, isotype control; bottom, anti-human Fas-L (4H9).

AID

CI 1169

/

6c27$$$$63

09-08-97 09:09:24

Cytotoxic lymphocytes involved in tumor cells lysis have been reported to exert killing by at least three distinct mechanisms. Two occur in a short time (less than 4 hr) and consist of the perforin/granzyme and Fas-L pathways. The third mechanism is delayed and consists of TNF-mediated killing (9, 10). The relative contribution by either of these three mechanisms in tumor cell killing may depend on the effector cells, the tumor cells, or both. This study investigated the role of the Fas-L-mediated killing by CTL, TIL, and LAK

cias

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

79

FIG. 7. TIL-mediated cytotoxicity of CDDP and VP-16-treated human DU145 prostate tumor cells. The percentage cytotoxicity was measured using the 51Cr release assay at indicated E:T ratio in 96-well plates in the absence (A, C) or presence (B, D) of 3 mM EGTA/4 mM MgCl2 . Target cells were treated overnight in the presence of drug or DMSO prior to assay. This figure is representative of four experiments. Spontaneous release of control cells treated with drugs alone was less than 5%.

against three human prostate carcinoma cell lines. All three lines express moderately the Fas-R antigen but are resistant to killing by anti-Fas antibody. The data in this study confirm the resistance of prostate tumor cell killing by the Fas-L-mediated pathway of CTL. Further evidence demonstrates that the chemotherapeutic drugs (CDDP, VP-16), used at subtoxic concentrations, can sensitize the hormonal-independent PC3 and DU145, but not the hormonal-dependent LnCAP, and drug-resistant human prostatic tumor cells lines to the Fas-L-mediated cytotoxic pathway. The drugmediated sensitization was blocked by neutralizing anti-Fas antibody, demonstrating that the drugs facilitate the death signal through the Fas-R. The sensitization was not due to Fas-R upregulation on the tumor cell surface. This study also demonstrates that IL-2activated human TIL and LAK cells express Fas-L and are capable of killing sensitized PC-3 and DU145 cells via the Fas–Fas-L-mediated cytotoxic pathway. Although sensitive to TIL and LAK-mediated killing by

the perforin pathway, treatment of LnCAP by CDDP and VP-16 did not render the tumor cells sensitive to Fas-L-mediated killing. These findings demonstrate that certain tumor cells that are Fas/ but resistant to Fas-L killing can be rendered sensitive by treatment with subtoxic concentrations of drugs like CDDP and VP-16. This drug-mediated sensitization process may be beneficial in treatment by immunotherapy, particularly in cases when the Fas-L pathway is the predominant pathway of killing by the cytotoxic lymphocytes. Fas is a member of the tumor necrosis factor (TNF)/ nerve growth factor super families (19–21). Fas is a transmembrane protein that has been shown to induce apoptotic cell death when reacted with anti-Fas antibodies, soluble Fas-L, and T and NK cells expressing Fas-L (21). Fas is expressed on immature thymocytes, activated T cells, and cells in the ovary, heart, and testes. Studies with lymphoproliferative mice (lpr and gld) have shown that Fas and Fas-L play a role in removing activated T cells following the immune re-

AID

cias

CI 1169

/

6c27$$$$63

09-08-97 09:09:24

80

FROST ET AL.

FIG. 8. Blocking of TIL-mediated cytotoxicity of CDDP-treated human DU145 prostate tumor cells. The percentage cytotoxicity was measured using the 51Cr release assay at indicated E:T ratio in 96-well plates in the absence (A) or presence (B) of 3 mM EGTA/4 mM MgCl2 . Target cells were treated overnight in the presence of drug or DMSO prior to assay. Two hours prior to addition of effector cells, CDDP was washed from target cells and fresh medium containing 1 mg/ml of the anti-Fas neutralizing antibody, ZB4, or isotype control was added. In B, the percentage cytotoxicity of drug / anti-Fas antibody-treated cells (j) in the presence of EGTA/MgCl2 was significantly less than that of drug-treated cells (l) alone at the 10:1 E:T ratio (P õ 0.05). This figure is representative of four experiments. Spontaneous release of controlled cells treated with drugs alone was less than 1%.

sponse (22). Fas-ligand is a type I integral membrane protein. It has been shown that soluble Fas-L and COS cells transfected with Fas-L are capable of killing FasR/ target cells (23). Fas–Fas-ligand-mediated cytotoxicity has been shown to be Ca2/-independent (18). To delineate the sensitivity of the prostate lines to CTL killing by the Fas-L pathway, we made use of the murine PMMI hybridoma cell line that kills by the FasL pathway (14). Suda and Nagata (21) have reported that Fas–Fas-L killing is not species specific and both murine and human effector or target cells can be used interchangeably. In this system, we show that the Fasexpressing prostate lines were all resistant to killing by PMMI. However, treatment with CDDP or VP-16 significantly sensitized PC-3 and DU145, but not LnCAP, to killing and apoptosis. Killing was Fas–FasL-mediated because it was blocked in the presence of neutralizing anti-Fas antibody. Both TIL and LAK cells have been used in immunotherapy against certain drug-refractory cancers (5). The mechanism by which these cells act in vivo is not clear and, furthermore, it is not clear why some tumors are responsive by such treatments and others are not. In vitro, both TIL and LAK kill tumor cells in nonMHC-restricted fashion. Both mediate killing in short term by the perforin/granzyme and Fas-L-mediated pathways and long term by the TNF-a pathway (24– 27). We examined the Fas-L component in both TIL and LAK by the use of the Ca2/ chelator EGTA/MgCl2 . Perforin-mediated cytotoxicity is strictly dependent on extracellular Ca2/, which is required for CTL-regulated granule exocytosis as well as the lytic activity of perforin (28). In contrast, Fas-based cytotoxicity has been

reported to work in the absence of Ca2/ (18, 29, 30). However, Lowin et al. (31) have reported that depletion of Ca2/ in primary MLC abolished both perforin- and Fas-mediated killing. Ca2/ was needed for the TCR signal-induced expression of Fas-L, but not for interaction of Fas-L with Fas. In our studies, the Fas-L killing by PMMI was not mediated by TCR engagement and was induced by PMA/ionomycin prior to interaction with target cells; therefore, Ca2/ was not required. Thus, in our studies, EGTA/MgCl2 discriminates between the perforin pathway and the Fas-L pathway. IL-2-activated TIL and LAK cells express both perforin and Fas-L and are able to effectively kill untreated LnCAP, PC-3, and DU145 target cells. However, the addition of EGTA/MgCl2 completely abolished TIL- and LAK-mediated killing. In contrast, following drug treatment of PC-3 and DU145, but not LnCAP target cells, TIL and LAK develop a Ca2/-independent cytolytic component. Furthermore, this Ca2/-independent cytotoxicity is blocked by anti-Fas antibody. These findings demonstrate that both TIL and LAK express the Fas-L pathway that is not readily revealed since the prostate lines are resistant to Fas-L. The addition of drugs revealed a Fas-L pathway in EGTA/MgCl2treated cultures. The perforin pathway is potent and may override the overall cytotoxic activity and, thus, the Fas-L pathway component may be masked. Clearly, under circumstances in which tumor cells are resistant to perforin/granzyme and resistant to Fas-L, chemosensitization of tumor cells by drugs revealed a significant cytotoxic activity. The failure of sensitizing LnCAP by drugs is noteworthy due to the hormonal dependency of LnCAP. The

AID

cias

CI 1169

/

6c27$$$$63

09-08-97 09:09:24

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

81

FIG. 9. LAK-mediated cytotoxicity of CDDP-treated human DU145 prostate cell line. The percentage cytotoxicity was measured using the 51Cr release assay at indicated E:T ratio in 96-well plates in the absence (A) or presence (B) of 3 mM EGTA/4 mM MgCl2 . Target cells were treated overnight in the presence of drug or DMSO prior to assay. For blocking of Fas-mediated killing of DU145 tumor cells, 2 hr prior to addition of effector cells, CDDP was washed from target cells and fresh medium containing 1 mg/ml of the anti-Fas neutralizing antibody, B-G27, or isotype control was added in the absence (C) or presence (D) of 3 mMEGTA 4 mM MgCl2 . In D, the percentage cytotoxicity of drug / anti-Fas antibody-treated cells (j) in the presence of EGTA/MgCl2 was significantly less than that of drug-treated cells (l) alone at the 10:1 E:T ratio (P õ 0.05). This figure is representative of four experiments. Spontaneous release of control cells treated with drugs alone was less than 5%.

resistance to Fas killing is not due to mutation of the Fas-R because transfection with the wild-type Fas does not render LnCAP sensitive for killing by anti-Fas antibody (32). This suggests that the Fas–Fas-L cytotoxic pathway is blocked at the level of the receptor for PC3 and DU145 cells and downstream of the receptor for LnCAP cells, and supports our findings that PC-3 and DU145, but not LnCAP tumor cells, can be sensitized to Fas-L-mediated killing. This study has shown that sensitization to Fas-L can be obtained by subtoxic concentrations of drugs acting at different biological levels. The mechanisms by which drugs sensitize cells to the Fas–Fas-L cytotoxic pathway are yet to be determined. Studies of gene expression in PC-3 cells following treatment with CDDP or VP-16 suggested that sensitization was due to altered levels of antiapoptotic and/or oncogene expression (33).

Other studies have implicated a role for the lipid second messenger, ceramide, in drug-mediated apoptosis in Fas and TNF-mediated cytotoxicity (34). For example, blocking of ceramide synthase abrogated daunorubicine-induced apoptosis in P388 lymphocytic and U937 leukemia cell lines (35). Alternatively, it is possible that subtoxic concentrations of drugs may modify gene expression and either upregulate ‘‘death signal’’ proteins, such as FADD and ICE, or downregulate antiapoptotic genes, such as FAP-1 (36). The development of resistance to drugs and CTLs, LAK, or TIL by tumor cells had important implications in immunotherapy. The development of antitumor CTL activity and the use of TIL and LAK cells and gene therapy have met with limited success in the cure of tumor cells (25, 26). Furthermore, some tumors express Fas-R but are resistant to killing by the Fas-L. Thus,

AID

cias

CI 1169

/

6c27$$$$64

09-08-97 09:09:24

82

FROST ET AL.

FIG. 10. LAK-mediated cytotoxicity of CDDP-treated human LnCAP prostate cell line. The percentage cytotoxicity was measured using Cr release assay at indicated E:T ratio in 96-well plates in the absence (A) or presence (B) of 3 mM EGTA/4 mM MgCl2 . Target cells were treated overnight in the presence of drug or DMSO prior to assay. Spontaneous release of control cells treated with drugs alone was less than 15%.

51

it is possible that resistance to Fas-L may also be accompanied by resistance to killing by CTLs through the perforin pathway. Therefore, in such cases, overcoming resistance to Fas–Fas-L-mediated cytotoxicity may overcome resistance to CTL-mediated killing. We have shown that IL-2-activated TIL and LAK cells express Fas-L and can kill sensitized PC-3 and DU145 prostate via the Fas–Fas-L cytotoxic pathway. These findings support the hypothesis that tumor cells can be sensitized by subtoxic concentrations of chemotherapeutic drugs and that the Fas–Fas-L pathway may be an effective target for intervention in tumor cells’ resistance to cytotoxic lymphocytes in immunotherapy. ACKNOWLEDGMENTS We acknowledge Dr. William Clark for providing the PMMI hybridoma, Dr. Shunsuke Mori for the RT–PCR, and Dr. Anahid Jewett for flow cytometry. We also thank Dr. Randir for providing the TIL. This work was supported in part by the Concern Foundation, the Boiron Research Foundation, and the Carolan Foundation. We also thank Samantha Nguyen and Sonia Choi for preparing the manuscript.

1. 2. 3. 4.

Goldstein, L. J., Curr. Probl. Cancer 19, 65, 1995. Hichman, J. A., Eur. J. Cancer 32A, 921, 1996. Wingo, P. A., Tong, T., and Bolden, S., Cancer Stat. 45, 8, 1995. Isaacs, J. T., Lundom, P. I., Berges, R., Martilainen, P., Kyprianou, N., and English, H. F., J. Androl. 13, 457, 1992. 5. Rosenberg, S. A., Annu. Rev. Med. 47, 481, 1996. 6. Brinkmann, U., and Pastan, I., Biochim. Biophys. Acta 1198, 27, 1994. 7. Freinberg, B., Kurzurock, R., Talpaz, M., Blick, M., Saks, S., and Gutterman, J. U., J. Clin. Oncol. 6, 1328, 1992.

CI 1169

/

6c27$$$$64

9. Kagi, D., Vignaux, F., Ledermann, D., Burki, K., Depraetere, V., Nagata, S., Hengartner, H., and Golstein, P., Science 265, 528, 1994. 10. Lowin, B., Beermann, F., Schmidt, A., and Tschopp, J., Proc. Natl. Acad. Sci. USA 91, 11571, 1994. 11. Morimoto, H., Yonehara, S., and Bonavida, B., Cancer Res. 53, 2591, 1993. 12. Uslu, R., Jewett, A., and Bonavida, B., Gynecol. Oncol. 62, 282, 1996. 13. Friesen, C., Herr, I., Krammer, P. H., and Debatin, K., Nature Med. 2, 574, 1996. 14. Kaufmann, Y., and Berke, G., J. Immunol. 131, 50, 1983. 15. Walsh, C. M., Glass, A. A., Chiu, V., and Clark, W. R., J. Immunol. 153, 2514, 1994. 16. Steger, G. G., deKernion, J. B., Figlin, R., and Belldegrun, A., Br. J. Cancer 72, 101, 1995. 17. McGahon, A. J., Martin, S. J., Bissonnette, R. P., Mahboubi, A., Shi, Y., Mogil, R. J., Nishioka, W. K., and Green, D. G., in ‘‘Methods in Cell Biology,’’ (L. M. Schwartz and B. A. Osborne, Eds.), Vol. 46, p. 153, Academic Press, Inc., San Diego, 1995. 18. Rouvier, E., Luciani, M., and Golstein, P., J. Exp. Med. 177, 195, 1993. 19. Nagata, S., Adv. Immunol. 57, 129, 1994.

REFERENCES

AID

8. Schmidt-Wolf, G., and Schmidt-Wolf, I. G., Ann. Hematol. 69, 273, 1994.

09-08-97 09:09:24

20. Takahashi, T., Tanaka, M., Brannan, C. I., Jenkins, N. A., Copeland, N. G., Suda, T., and Nagata, S., Cell 76, 969, 1994. 21. Suda, T., and Nagata, S., J. Exp. Med. 17, 873, 1994. 22. Vignaux, P., and Golstein, P., Eur. J. Immunol. 24, 923, 1994. 23. Suda, T., Takahashi, T., Golstein, P., and Nagata, S., Cell 75, 1169, 1993. 24. Garner, R., Helgason, C. D., Atkinson, E. A., Pinkoski, M. J., Ostergaard, H. L., Sorensen, O., Fu, A., Lapchak, P. H., Rabinovitch, A., McElhaney, J. E., Berke, G., and Bleackley, R. C., J. Immunol. 153, 5413, 1994. 25.

cias

Facchetti, P., Prigione, I., Ghiotto, F., Tasso, P., Garaventa,

IMMUNOSENSITIZATION OF TUMOR CELLS TO Fas-L-MEDIATED CYTOTOXICITY

26. 27. 28. 29. 30.

AID

A., and Pistoia, V., Cancer Immunol. Immunother. 42, 170, 1996. Whiteside, T. L., and Parmiani, G., Cancer Immunol. Immunother. 39, 15, 1994. Lee, R. K., Spielman, J., Zhao, D. Y., Olsen, K. J., and Podack, E. R., J. Immunol. 157, 1919, 1996. Young, J. D., Clark, W. R., Liu, C. C., and Cohn, J. Exp. Med. 166, 1894, 1987. Vujanovic, N. L., Nagashima, S., Herberman, R. B., and Whiteside, T. L., J. Immunol. 157, 1117, 1996. Lowin, B., Hahne, M., Mattmann, C., and Tschopp, J., Nature 370, 650, 1994.

CI 1169

/

6c27$$$$64

09-08-97 09:09:24

83

31. Lowin, B., Mattman, C., Hahne, M., and Tschopp, J., Int. Immunol. 8, 57, 1996. 32. Takeuchi, T., Sasaki, Y., Ueki, T., Kaziwara, T., Moriyama, N., Kawabe, K., and Kakizoe, T., Int. J. Cancer 67, 709, 1996. 33. Sinha, B. K., Yamazaki, H., Eliot, H. M., Schneider, E., Borner, M. M., and O’Connor, P. M., Biochim. Biophys. Acta 1270, 12, 1995. 34. Bose, R., Verheij, M., Haimovitz-Friedman, A., Scotto, K., Fuks, Z., and Kolesnick, R., Cell 82, 405, 1995. 35. Kolesnick, R., and Fuks, Z., J. Exp. Med. 181, 1949, 1995. 36. Mori, S., Murakami-Mori, K., Jewett, A., Nakamura, S., and Bonavida, B., Cancer Res. 56, 1874, 1996.

cias