Previous studies with CTL lines and CTL hybridomas have suggested that functional CD95 (APO-1/Fas)-ligand (CD95L) expression on effector CTLs is a consequence of specific CTL-target recognition and TCR triggering of newly transcribed CD95L. Such a control on the expression of CD95L could provide a double safeguard for killing only cognate target cells. Here the regulation of CD95L expression and function was tested in in vivo primed, alloreactive peritoneal exudate CTL (PEL) from perforin-deficient (P0) mice. CD95L-based, PEL-mediated cytotoxicity was blocked by brefeldin A, an inhibitor of intracellular protein transport, but not by the protein synthesis inhibitor emetine, the immunosuppressive drug cyclosporin A, or the DNA transcription inhibitor actinomycin D. CD95L mRNA transcripts in freshly isolated PEL were shown by RT-PCR; CD95L surface expression was evident by staining with Fas-Fc as well as CD95L Abs. Undiminished CD95L expression and cytocidal activity were found in PEL incubated for 48 h in culture, without adding Ag, mitogen, or cytokines. PEL expressed functional CD95L, yet exhibited target cell-specific killing, except when encountering high CD95-expressing cells. The results indicate that PEL use CD95L probably expressed in the Golgi and/or on the cell surface and do not require newly transcribed CD95L upon target cell conjugation. Hence the TCR-triggered recruitment of preformed CD95L, rather than its biosynthesis, controls CD95L-based specific lysis induced by CTL.

Cytotoxic T lymphocytes constitute a primary immune surveillance system that can recognize and destroy foreign cells, or autologous cells expressing foreign or mutated self proteins. Clearly, at least two distinct cytocidal mechanisms are used by CTL in destroying such cells (1, 2). In the degranulation pathway, the secreted lytic protein perforin and a family of CTL-specific enzymes are thought to be responsible for target cell destruction (3). A perforin-independent killing pathway (4) is now recognized because perforin-deficient CTL-hybridomas (5), CTL clones (6), and mice (7, 8, 9) still possess a high degree of CTL-mediated lytic activity, most probably through the binding of CD95 (APO-1/Fas) on the target cell surface membrane by CD95 ligand (CD95L)3 expressed on the surface of CTL (2, 10, 11). It is widely believed that cross-linking of CD95 triggers a cascade of intracellular protein-protein interactions and proteolytic activities, culminating in apoptosis of the target cell (12, 13, 14).

CD95L is expressed after primary T cell activation, a process inhibited by the immunosuppressive agent cyclosporin A (CsA) (15, 16, 17). Little is known, however, about the regulation of CD95L expression and function in effector CTL (18). In CTL lines, transient CD95L expression is induced upon TCR engagement (15, 19), by CD3 Abs (20), by the polyclonal stimulator Con A (21), or by PMA and ionomycin (PI) (15, 17, 21). Induction of CD95L expression in CTL is calcium dependent and sensitive to macromolecule synthesis inhibitors; however, CD95L function in triggering CD95-based apoptosis is not (22). Based on studies conducted mainly with CTL lines and T cell hybridomas, transcriptional regulation of CD95L expression and function in CTL action against cognate target cells has been proposed (2). With that model, TCR-based recognition of the Ag presented by the MHC at the target cell surface transduces within the effector cell a transcriptional signal(s) that triggers CD95L gene expression. Enhanced CD95L mRNA expression is assumed to be due to increased transcription, although an increase in message stability has not been ruled out. Swift but transient expression of the CD95L protein then allows CD95 engagement at the target cell surface, signaling its demise. Obviously, failure to recognize the target would not signal CD95L expression nor its implementation. It has also been proposed that functional CD95L expression involves translocating previously made CD95L from storage compartments to the cell surface, or transforming CD95L from an inactive to a functional form (18). Here, we have taken a closer look at the regulation of CD95L expression and function on effector CTL, using PEL, a mouse model system of in vivo primed CTL (23).

C57BL/6 (H-2b) T cell leukemia EL4 and DBA/2 (H-2d) mastocytoma P815 were carried as ascites in syngeneic mice or maintained for short periods in culture. Leukemia L1210 of DBA/2 (H-2d) and BW of AKR (H-2k) were cultured in vitro. LF+ is an L1210 variant transfected with mouse CD95 overexpression construct (kindly provided by Dr. Pierre Golstein, Centre d’Immunologie, Marseille-Luminy, France) (10). LF is another L1210 subline that expresses little CD95 Ag because of transfection with a CD95 antisense construct (24). All cells were cultured in RHFM (RPMI 1640 containing heat-inactivated FCS (5%), sodium pyruvate (1 mM), HEPES (10 mM), penicillin (100 U/ml), streptomycin (100 μg/ml), and β-mercaptoethanol (5 × 10−5 M)). The CTL line AB.1 (H-2d anti-H-2b) has been previously described (25) and was maintained in vitro by periodic stimulation with irradiated C57BL/6 spleen cells and a minimal level of T cell growth factors (supernatant of Con A-stimulated rat splenocytes) required to support growth. To activate the AB.1 cytolytic function, cells were incubated for 2 h with PMA (Sigma; final concentration of 25 ng/ml) and ionomycin (Sigma, St. Louis, MO; final concentration of 0.5 μg/ml). The CTL hybridoma d11s (kindly provided by Dr. Pierre Golstein) exerts CD95-CD95L-based cytotoxicity when activated with PI (10).

Perforin-knockout (P0, H-2b) mice have been previously described (9). Two- to four-month-old C57BL/6, BALB/c, and P0 mice were supplied by the Animal Breeding Center of the Weizmann Institute.

PEL were generated, prepared, and purified as previously described (26). Briefly, P0, C57BL/6, and BALB/c mice were injected i.p. with allogeneic tumor cells LF+ (H-2d) or EL4 (H-2b) (25 × 106/mouse). Eight to eleven days after a primary alloimmunization, or 4 to 5 days after a secondary stimulation (given 6–12 wk after priming), the mice were sacrificed, and their peritoneal cavities were rinsed with PBS supplemented with 5% heat-inactivated newborn calf serum (PBS-NCS). The resulting crude peritoneal exudate cells were centrifuged, resuspended in medium, and incubated on nylon wool columns at 37°C to deplete adherent cells such as B cells and macrophages. After 60 min, the nonadherent cells were eluted by rinsing the columns with cold PBS-NCS. The eluted cells (PEL) contained >95% T cells, 80 to 90% of which were CD8+, about half of which formed specific conjugates. PEL blasts (PEB) were derived from PEL upon incubation in recombinant human IL-2 (500 U/ml), as previously described (5).

A standard 51Cr release assay was used. Target cells were incubated with Na51Cr2O4 (1 h at 37°) and washed twice with PBS-NCS before use. Lytic assays were conducted in U-shaped, 96-well microtiter plates with 3 × 104 labeled target cells per well, and effector cells at the indicated ratios. The plates were centrifuged to promote conjugate formation and incubated at 37° for 4 to 5 h and then recentrifuged. One hundred microliters of supernatant from each well was harvested, and its radioactivity was determined in a gamma counter. The percentage of cytotoxicity was calculated as follows: % cytotoxicity = [(experimental release − spontaneous release)/(total release − spontaneous release)] × 100.

Total RNA was isolated from various CTL and control cells by TRI REAGENT (Molecular Research Center, Cincinnati, OH). Titan One Tube RT-PCR System (Boehringer Mannheim, Mannheim, Germany) was used to analyze these RNAs for CD95L expression compared with GAPDH expression. In this system, reverse transcription and PCR are performed in a single step. Each 50-μl reaction mixture contained 2 μg RNA, 15 pmol downstream primer (5′-CTT GGG CTC CTC CAG GGT CAG T-3′), 15 pmol upstream primer (5′-TCT CCT CCA TTA GCA CCA GAT CC-3′), nucleoside 5′-triphosphate NTP (0.2 mM), DTT (5 mM), MgCl2 (1.5 mM), RNase inhibitor (10 U), 5× RT-PCR buffer (10 μl), and 1 μl enzyme mixture containing avian myeloblastosis virus RT and Expand High Fidelity (Boehringer Mannheim). Each sample was mixed, briefly centrifuged, overlaid with 30 μl mineral oil, and placed in the thermocycler (Programmable Thermal Controller, MJ Research, Watertown, MA), equilibrated at 50°C, for 30 min and set for thermocycling as follows: 1 min denaturation at 94°C, 35 cycles of 1 min denaturation at 94°C, 1 min annealing at 55°C, and 1 min elongation at 72°C, and a last elongation of 7 min at 72°C. The samples were then resolved on a 1% agarose gel and observed with ethidium bromide staining and UV light.

Fas-Fc staining.

Analysis of surface CD95L expression by Fas-Fc staining has been previously described (27, 28). Briefly, cells (2.5 × 105/tube) were washed in cold buffer (consisting of PBS, 1% BSA, and 0.02% sodium azide), centrifuged, resuspended (30 μl/tube), and incubated (30 min on ice) with murine Fas-Fc (Immunex, Seattle, WA; Lot No. 8) (27, 28), 20 μg/ml. After two washes in the above buffer, the cells were resuspended in a 30-μl buffer containing 5 μg/ml biotinylated goat anti-human IgG Fc (Jackson ImmunoResearch Laboratories) and incubated on ice for 30 min. After two washes, the cells were resuspended in a 30-μl buffer containing 5 μg/ml phycoerythrin (PE)-Avidin (Jackson ImmunoResearch, West Grove, PA) and incubated in the dark for 30 min on ice. Finally, the cells were washed as above and resuspended in 0.5 ml PBS containing 0.02% sodium azide and analyzed by FACScan.

CD95L Ab staining.

We used the following CD95L Abs: a) mAb to CD95L (peptide, aa 196–220) (FITC), clone A11-2, (Alexis, Laufelfingen, Switzerland; catalogue No. 804-009F100), b) M anti-MFasL, clone Kay-10, C57BL/6 gld anti-mFasL-transfected L5178Y T lymphoma (PharMingen; catalogue No. 09931D), and c) Rb anti-MCD95L (peptide, aa 261–277) (Ab-1), (Oncogene Research Products, Calbiochem, La Jolla, CA; catalogue No. PC78). The staining procedure by these CD95L Abs followed that described for Fas-Fc except for the necessary variations. a) Staining by the Alexis Ab was in 30 μl, of 8 μg/ml). b) In staining by the PharMingen Ab, cells were first incubated in 30 μl (40 μg/ml) of anti-mouse CD95L, washed, and secondarily incubated in 30 μl (5 μg/ml) of FITC-goat anti-mAb (Zymed, San Francisco, CA). c) In staining by the Oncogene Ab, cells were first incubated in 30 μl (8 μg/ml) polyclonal Rb anti-m CD95L, washed, and secondarily incubated in 30 μl (20 μg/ml) FITC-goat anti-Rb Ab (Jackson ImmunoResearch).

Unlike perforin-mediated lysis, CD95-based cytotoxicity induced by CTL can occur in Ca2+-free medium (10, 16, 21, 29) and is inhibited by CD95 Abs (30) and by Fas-Fc (31). Using perforin-deficient (P0) mice (H-2b) and L1210 cells (H-2d) transfected with CD95 or CD95-antisense (LF+ and LF, respectively), we have found that P0 anti-LF+ (or anti-LF) (b anti-d) PEL-mediated lysis of LF+ cells was calcium independent and inhibitable (to about 15% of the control) by either a CD95 Ab or Fas-Fc. The P0 PEL showed poor cytocidal activity toward either LF(H-2d) or third party (H-2b) EL4 cells (Fig. 1). These results confirmed that the cytocidal activity of P0 anti-LF+ PEL is indeed CD95 based and Ag specific.

FIGURE 1.

Fas-based lymphocytotoxicity induced by perforin-deficient CTL, P0 anti-LF+ PEL (b anti-d), 4 days after the secondary alloimmunization, were preincubated with soluble Fas-Fc or Fas(Jo2) Ab, both at 5 μg/ml at room temperature for 30 min, before adding 51Cr-labeled targets (3 × 104 cells/well). The E:T ratio was 5 to 1. The inhibitors were present throughout the lytic reaction (5 h, 37°). EGTA, Mg2+EGTA (4 mM Mg2+/2 mM EGTA); SR, spontaneous release; LF+, 8.9%; LF+ in the presence of EGTA, 11.1%; LF, 10.2%; EL4, 18.9%.

FIGURE 1.

Fas-based lymphocytotoxicity induced by perforin-deficient CTL, P0 anti-LF+ PEL (b anti-d), 4 days after the secondary alloimmunization, were preincubated with soluble Fas-Fc or Fas(Jo2) Ab, both at 5 μg/ml at room temperature for 30 min, before adding 51Cr-labeled targets (3 × 104 cells/well). The E:T ratio was 5 to 1. The inhibitors were present throughout the lytic reaction (5 h, 37°). EGTA, Mg2+EGTA (4 mM Mg2+/2 mM EGTA); SR, spontaneous release; LF+, 8.9%; LF+ in the presence of EGTA, 11.1%; LF, 10.2%; EL4, 18.9%.

Close modal

Because CD95-based, Ag-independent cytotoxicity has been observed with reactivated CTL lines and hybridoma cells, but not with nonreactivated effectors (10, 11), constitutive, sustained expression of CD95L on in vivo primed CTL has been overlooked. In vivo primed PEL are highly specific killers (23). If PEL expressed CD95L and did not require additional TCR-based stimulation, they might kill nonspecifically high, but not low, CD95-expressing cells. We have tested this hypothesis with BALB/c anti-EL4 PEL (d anti-b), shown to exhibit potent, specific cytocidal activity (Table I), using LF+ and LF target cells (both are d). In the absence of Ca2+ (4 mM Mg2+/2 mM EGTA), a condition in which perforin was neither secreted nor lytic, d anti-b PEL nonspecifically killed high, but not low, CD95-expressing cells, LF+ and LF, respectively (Table I). The nonspecific lytic activity of PEL against the high Fas-expressing, noncognate target LF+ was considerably higher than against the low-expressing noncognate P815 or L1210. Yet PEL preferentially killed EL4, even though EL4 expressed less Fas than LF+ (Table I). Finally, we have confirmed (data not shown) that cognate PEL-target interaction facilitated the moderate killing of high CD95-expressing noncognate targets presented as bystanders, as reported before (32, 33).

Table I.

Cytocidal activity of BALB/c anti-EL4 PELa

TargetH-2 TypeCD95 Expression (% cells)% 51Cr Released at E:T of
10:15:12.5:1SR
Specificity       
EL4 34.7 74.3 69.0 44.4 11.8 
EL4 + EGTAb 34.7 49.2 42.9 27.9 24.6 
P815 30.2 2.0 0.0 0.0 10.4 
L1210 22.2 0.0 0.0 0.0 9.2 
CD95 dependence       
LF++ EGTA 70.0 21.1 14.5 9.0 10.6 
LF + EGTA 9.4 3.6 3.4 4.6 10.5 
TargetH-2 TypeCD95 Expression (% cells)% 51Cr Released at E:T of
10:15:12.5:1SR
Specificity       
EL4 34.7 74.3 69.0 44.4 11.8 
EL4 + EGTAb 34.7 49.2 42.9 27.9 24.6 
P815 30.2 2.0 0.0 0.0 10.4 
L1210 22.2 0.0 0.0 0.0 9.2 
CD95 dependence       
LF++ EGTA 70.0 21.1 14.5 9.0 10.6 
LF + EGTA 9.4 3.6 3.4 4.6 10.5 
a

BALB/c anti-EL4 (d anti-b) PEL, 4 days after a secondary alloimmunization were mixed with the 51Cr-labeled targets (3 × 104 cells/well); lysis was measured after 5 h at 37°C.

b

EGTA, Ca2+-free medium (4 mM MgCl2/2 mM EGTA), blocking perforin secretion and function.

To determine whether the CD95-based cytocidal activity exhibited by PEL correlated with the active expression of CD95L, CD95L mRNA transcripts in PEL were tested by PCR. Figure 2 a shows that the nonactivated hybridoma d11s and CTL line AB.1, P0 PEL-blasts (PEB), BALB/c anti-EL4 PEL, P0 anti-LF+ PEL, but not control LF+ or EL4, expressed CD95L mRNA. The CD95L PCR was performed with the same RNA preparations used for GAPDH control PCR.

FIGURE 2.

CD95L mRNA transcripts and cell surface expression of CD95L. a, RT-PCR. Lanes: 1, d11s (not PI activated); 2, P0 PEB (derived from P0 anti-LF+ PEL); 3, PEL (BALB/c anti-EL4, 9 days after primary alloimmunization); 4, P0 PEL (P0 anti-LF+, 4 days after secondary alloimmunization); 5, AB.1 (not PI activated); 6, LF+; 7, EL4. b, FACS analysis of CD95L expression on PEL and CTL lines stained by Fas-Fc. c, CD95 staining of PEL detected by CD95L Abs from Alexis (M), PharMingen (N), and Oncogene (O) (see Materials and Methods).

FIGURE 2.

CD95L mRNA transcripts and cell surface expression of CD95L. a, RT-PCR. Lanes: 1, d11s (not PI activated); 2, P0 PEB (derived from P0 anti-LF+ PEL); 3, PEL (BALB/c anti-EL4, 9 days after primary alloimmunization); 4, P0 PEL (P0 anti-LF+, 4 days after secondary alloimmunization); 5, AB.1 (not PI activated); 6, LF+; 7, EL4. b, FACS analysis of CD95L expression on PEL and CTL lines stained by Fas-Fc. c, CD95 staining of PEL detected by CD95L Abs from Alexis (M), PharMingen (N), and Oncogene (O) (see Materials and Methods).

Close modal

Given the PCR data, we next investigated whether PEL express cell surface CD95L by testing Fas-Fc binding to PEL. Fas-Fc is a soluble fusion protein composed of mouse CD95 and the Fc portion of human IgG and is capable of binding CD95L (28). Although the specificity of Fas-Fc binding to membrane-bound CD95L is not absolute, its blocking of CD95L-based PEL action was robust and comparable to that of the CD95 Ab (Jo2) (Fig. 1). Using FACS analysis, we found that freshly isolated PEL of various origins, at peak lytic ability after primary or secondary immunization, expressed CD95L (Fas-Fc binding) without requiring stimulation by cognate target cells (Fig. 2,b, A and B). The small differences in ligand staining intensity among the various PEL populations tested (Fig. 2,b, A, C, and E) probably reflected experimental variations. The FACS experiments were conducted with different batches of cells, various settings of the cytometer, and newly prepared reagents. CD95L expression on PEL detected by Fas-Fc staining was validated by staining with three independently derived CD95L Abs (Fig. 2 c).

To exclude the possibility that stable CD95L expression on PEL was due to residual exposure of the PEL to antigenic stimulation in vivo, the PEL were harvested, depleted of adherent cells on nylon wool columns, and then incubated in vitro for 48 h, without deliberate antigenic or IL-2 stimulation, and CD95L expression and cytocidal activity were monitored. Figure 2,b, C, D, E, and F, shows that CD95L continued to be expressed on in vitro cultured PEL, which also maintained undiminished lytic activity (Table II). We found that the CTL line AB.1 (d anti-b) showed Fas-Fc staining even without activation (Fig. 2 b, I and J), although its (nonspecific, CD95-based) lytic activity was enhanced after PI stimulation, as previously reported (27). Interestingly, PI stimulation of P0 anti-LF+ PEL affected neither their cytocidal activity against cognate, high CD95-expressing LF+ nor against the low Fas-expressing target LF. With BALB/c anti-EL4 PEL, inhibition of EL4 lysis by PI was noted in a few experiments, whereas the lysis of noncognate LF+ was markedly enhanced and was not Ca2+ dependent (data not shown).

Table II.

The lytic activity of freshly isolated, cultured PELa

EffectorTarget% 51Cr Released by PEL
Freshly isolatedCultured for 48 h
C57BL/6 anti-LF+ PEL LF+ 67.2 66.3 
P0 anti-LF PEL LF+ 40.8 51.0 
EffectorTarget% 51Cr Released by PEL
Freshly isolatedCultured for 48 h
C57BL/6 anti-LF+ PEL LF+ 67.2 66.3 
P0 anti-LF PEL LF+ 40.8 51.0 
a

PEL cytocidal activity was determined on the day of their removal from the peritoneal cavity of alloimmunized mice and after 48 h in culture without deliberate stimulation. (E:T, 5:1, 5 h at 37°C). Spontaneous release was 8.6% and 15.8%, respectively.

The immunosuppressive agent CsA selectively inhibits those T cell activation pathways associated with an increase in intracellular Ca2+. A cytoplasmic, T cell-specific component of the transcription factor NF-AT, necessary for IL-2 gene transcription, must be dephosphorylated by the Ca2+/calmodulin-dependent protein phosphatase calcineurin for NF-AT to translocate into the nucleus and activate transcription. Calcineurin is inhibited by binding to a complex of CsA and cyclophilins (34, 35). Table III shows that, even with 100 nM of CsA, the lysis of LF+ mediated by P0 anti-LF+ PEL was only slightly inhibited. To ascertain the efficacy of the CsA used in blocking lymphocyte activation, we determined its effects on the lymphoproliferative response during a two-way mixed lymphocyte reaction (P0 vs BALB/c). As little as 5 nM CsA blocked the incorporation of [3H]thymidine by 95% (Table III), thus proving the efficacy of the inhibitor used. These results indicate that transcriptionally regulated (and CsA-sensitive) activation is not required in the course of CD95-based, PEL-induced lysis.

Table III.

Transcriptional and translational control of CD95-based cytotoxicity induced by P0 PELa

Blocking Agent% 51Cr Released at E:T[3H]Thymidine Uptake
10:15:12.5:1CsA (nM)cpm
CsA (nM)      
60.5 43.2 ND 3913 
25 63.7 38.3 ND 228 
50 50.0 39.9 ND 20 221 
100 56.2 36.7 ND 75 392 
Act D (μg/ml)      
60.8 45.9 32.3   
1.5 46.2 40.9 24.6   
3.0 51.4 41.4 25.7   
Emetine (μM)    Protein Synthesis (% inhibition)  
51.7 36.6 23.2  
0.25 51.2 (1.0) 30.8 (15.8) 18.2 (21.5) 52.9  
0.5 47.1 (8.9) 28.2 (22.9) 17.5 (24.5) 85.2  
2.5 43.8 (15.2) 29.6 (19.1) 12.4 (46.5) 93.7  
Blocking Agent% 51Cr Released at E:T[3H]Thymidine Uptake
10:15:12.5:1CsA (nM)cpm
CsA (nM)      
60.5 43.2 ND 3913 
25 63.7 38.3 ND 228 
50 50.0 39.9 ND 20 221 
100 56.2 36.7 ND 75 392 
Act D (μg/ml)      
60.8 45.9 32.3   
1.5 46.2 40.9 24.6   
3.0 51.4 41.4 25.7   
Emetine (μM)    Protein Synthesis (% inhibition)  
51.7 36.6 23.2  
0.25 51.2 (1.0) 30.8 (15.8) 18.2 (21.5) 52.9  
0.5 47.1 (8.9) 28.2 (22.9) 17.5 (24.5) 85.2  
2.5 43.8 (15.2) 29.6 (19.1) 12.4 (46.5) 93.7  
a

P0 anti-LF+ PEL, 4 days after secondary alloimmunization, were preincubated with CsA (30 min at room temperature), ActD (present during lytic reaction), or emetine (37°C for 1 h), at the indicated concentrations, prior to adding 51Cr-labeled LF+ target cells (3 × 104 cells/well). Lysis was measured after 5 h at 37°C. Spontaneous release was 10 to 12%. For [3H]thymidine incorporation, two-way mixed lymphocyte culture (P0 versus BALB/c splenocytes), 2.5 × 105 cells each, were incubated for 4 days at 37°C, with the indicated concentrations of CsA. [3H]thymidine, 1.5 μCi/tube, was added for the last 8 h. Incorporation into TCA (10%)-insoluble material was determined by a liquid scintillation beta counter. Numbers in parentheses (in the emetine results) indicate the percentage inhibition of lysis. To determine inhibition of protein synthesis, emetine-treated cells were used to measure [35S]methionine uptake (0.2 ml of PEL (1 × 106 cells)/6 μCi [35S]methionine, for 4 h).

The requirements for de novo RNA synthesis were tested using actinomycin D (ActD), a powerful inhibitor of DNA transcription. ActD at 0.5 μg/ml completely inhibited the proliferation of the tumor cell lines P815, L1210, EL4, and LF+ (not shown). With up to 3 μg/ml of ActD, only a slight inhibition of P0 anti-LF+ PEL activity was evident (Table III). Next the requirements for protein synthesis were tested by Emetine, a potent, nonreversible inhibitor of protein synthesis of mammalian cells. With 0.5 to 2.5 μM of Emetine, [35S]methionine and [35S]cysteine incorporation into PEL was reduced to 10% of the control with only 1 to 15% inhibition of the lytic CTL activity at 50% lysis (Table III). Apparently, new DNA transcription or protein synthesis is not needed by the CD95-CD95L pathway of PEL action, which supports our theory that PEL express CD95L before cognate binding of the respective target.

BFA blocks constitutive protein transport by disrupting the Golgi apparatus (36, 37). Since PEL-mediated cytotoxicity and most intracellular constitutive protein transport are not calcium dependent, we tested whether BFA would also inhibit the P0 PEL-mediated killing. Table IV shows that, with BFA at 2.5 and 10 μg/ml, cytolysis of LF+ decreased by 85%. BFA is known to have other intracellular targets, and its effects are concentration dependent and influenced by the cell type and their biologic state (36, 37). However, no toxic effect against either the effector or target cell was detected by eosine dye exclusion at the working concentrations and up to 15 μg/ml (data not shown); at the highest concentrations used (10 μg/ml), BFA had only a minimal effect on 51Cr-release from LF+ cells (4.2% at 10 μg/ml, Table IV). Furthermore, BFA only marginally (16%) affected the lysis of LF mediated by C57BL/6 anti-LF+ PEL (Table IV), excluding the possibility that BFA inhibition of P0 PEL-induced lysis was due to its effects on PEL viability or interference with the recognition and signaling of target apoptosis. Hence, it was concluded that BFA inhibited lymphocytotoxicity by blocking the constitutive transport pathway of CD95L membrane expression. PEL express functional CD95L via the constitutive pathway of intracellular protein transport rather than by regulated secretion. This was consistent with the reduced Fas-Fc staining of cell-surface CD95L, as a result of incubating P0 PEL and the AB.1 CTL line with BFA (Fig. 2 b, G, H, K, and L).

Table IV.

Inhibition of CD95-based cytotoxicity by BFAa

EffectorTargetBFA% 51Cr Release at E:T
5:12.5:1SR
  (μg/ml)    
P0 anti-LF+ LF+ 49.3 32.1 8.7 
  2.5 15.6 5.6  
  10.0 13.2 4.2  
C57BL/6 anti-LF+ LF 65.5 47.2 12.6 
  2.5 61.9 46.3  
  10.0 55.0 40.0  
EffectorTargetBFA% 51Cr Release at E:T
5:12.5:1SR
  (μg/ml)    
P0 anti-LF+ LF+ 49.3 32.1 8.7 
  2.5 15.6 5.6  
  10.0 13.2 4.2  
C57BL/6 anti-LF+ LF 65.5 47.2 12.6 
  2.5 61.9 46.3  
  10.0 55.0 40.0  
a

CTL (PEL, 5 days after secondary alloimmunization) was preincubated with BFA (at the indicated concentrations) for 30 min prior to adding 51Cr-labeled-target cells (3 × 104 cells/well); the drug was present throughout the entire lytic reaction (4.5 h at 37°C).

After activation by Ag, by mitogen, and particularly by PI, certain CTL lines and hybridomas were found to express CD95L and kill CD95-expressing targets specifically and even nonspecifically (10, 38). Similarly, secreted CD95L, CD95L-expressing tumor cell lines, as well as transfected nonlymphoid COS cells can kill CD95-expressing target cells (11). On the other hand, the hallmark of CTL-mediated killing is that it is highly specific and MHC restricted, making constitutive, sustained expression of functional CD95L on effector CTL unlikely. Target cell specificity for lysis induced by CD95L-expressing CTL has therefore been explained by the activation of the CD95L gene and the transient surface membrane expression of CD95L upon (but not before) specific effector CTL-target conjugation and TCR triggering (2). It has been argued that constitutive expression of CD95 on tissues would make constitutive expression of CD95L on effector CTL quite dangerous (39), further supporting the above control mechanism.

We have found that effector CTL such as PEL show sustained expression of functional CD95L without apparent Ag stimulation (Table II; Figure 2,b, CF) and in the presence of inhibitors of transcription and translation (Table III). Similar CD95L-expressing cytocidal CTL (CD8+CD45RA+CD27) not requiring in vitro stimulation have been recently isolated from human blood (40). Thus, the proposed transcriptional regulation of CD95L expression and cytolytic function (2) must apply to the initial activation steps of naive (precursor) and the reactivation of memory CTL, CTL lines and hybridomas, but not to the action of effector CTL in vivo such as PEL, where constitutively expressed CD95L can signal target cell apoptosis upon TCR-mediated cognate interaction. Hence, the control of specific CD95L-based CTL action is at the cognitive TCR-mediated level, not the gene expression of ligands.

CD95-based lymphocytotoxicity induced by PEL has been shown to be blocked by BFA (Table IV). Importantly, killing by perforin-expressing PEL was not inhibited (Table IV). Neither CD95 expression nor cell apoptosis induced by CD95 Abs was affected by BFA (41). This indicated that BFA blocking was due to interference with constitutive CD95L expression, as was confirmed by FACS staining (Fig. 2). On the other hand, PEL-induced lysis was only marginally inhibited by the protein synthesis inhibitor Emetine (Table III), but not at all by the immunosuppressive drug CsA, or by the DNA transcription inhibitor ActD (Table III). From RT-PCR for CD95L mRNA transcripts (Fig. 2,a) and Fas-Fc staining of cell surface CD95L (Fig. 2,b), effector CTL, such as PEL, apparently express CD95L on the surface, not requiring fresh TCR triggering of CD95L gene expression brought about by the cognate target cells. Cytocidal PEL expressed functional CD95L upon removal from the peritoneal cavity and continue to express it for at least 2 days under in vitro conditions (Table II; Figure 2 b, CF).

Upon CTL-target conjugate formation, we postulate that the constitutively expressed CD95L, either present at or drawn from the ER/Golgi complex to the effector CTL/target contact site under the influence of TCR, triggers signaling through CD95 receptors on the targets, ultimately leading to their demise (Fig. 3). This hypothesis is supported by a number of earlier observations, some of which were made with PEL. For example, the polarization of the cytoskeletal element (e.g., microtubules organizing center) and the Golgi apparatus of PEL to the CTL/target contact region, found in our laboratory (42, 43), is in agreement with the proposed model of ligand focusing, as is the sequential killing of two target cells bound to a single effector (44).

FIGURE 3.

A model for CD95L expression and function in CTL bound to cognate target cell. After CTL-target conjugate formation, CD95L, either present at or drawn from the ER/Golgi complex to the CTL/target contact site under the influence of TCR, triggers signaling through CD95 receptors on the targets, ultimately leading to their apoptosis. Interactions of molecules other than TCR-MHC and CD95-CD95L (e.g., LFA-1) may further modulate the outcome of the CTL/target cell encounter.

FIGURE 3.

A model for CD95L expression and function in CTL bound to cognate target cell. After CTL-target conjugate formation, CD95L, either present at or drawn from the ER/Golgi complex to the CTL/target contact site under the influence of TCR, triggers signaling through CD95 receptors on the targets, ultimately leading to their apoptosis. Interactions of molecules other than TCR-MHC and CD95-CD95L (e.g., LFA-1) may further modulate the outcome of the CTL/target cell encounter.

Close modal

The proposed model does not preclude bystander killing. That TCR-mediated cognate recognition enhances low level of Fas-based killing of bystander targets has already been demonstrated (32, 33) and confirmed by us with PEL (not shown). Importantly, bystander killing is a two-step event (32). In the first step of cognate recognition/killing, bystanders are spared. It is only in the second stage that the activated CTL, now expressing more adhesion molecules (notably LFA-1), bind to and exert low level killing against Fas-expressing bystanders. That the killing of bystanders does not occur simultaneously with the killing of the cognate target supports a focal distribution on the cell surface of newly recruited CD95L (Fig. 3).

Why then is in vivo CD95L expression on CTL not dangerous to CD95-expressing neighboring cells? In humans, regulating surface membrane CD95L expression by metalloproteinase has been thought to prevent the accumulation of the membrane-bound form of the ligand, thus avoiding bystander killing (45, 46). On the other hand, in studies on CD95-based CTL action, selected or artificially transfected high CD95-expressing target cells are used (e.g., LF+ in mouse, SKW in humans, etc.); and third party and bystander killing is frequently seen (32). Although BALB/c PEL do lyse high CD95-expressing LF+ cells nonspecifically, and even more so as bystanders, they still preferentially kill the cognate target cells (Table I), even though EL4 express less CD95 than LF+. This reflects the selective advantage of cognitive binding affected by the T cell receptor, which then facilitates effective interaction of CD95L with target cell CD95. In our experience, most tissues in vivo express less CD95 than the commonly used CD95-transfected LF+ cell line and would therefore probably not be lysed nonspecifically by CTL. CD95L-expressing CTL, such as PEL, would kill cognate target cells efficiently and yet be of little threat to noncognate bystanders, except perhaps in isolated diseases such as hepatitis and myocarditis, where Fas expression on the target organ is high.

The kinetics of the two pathways of lymphocytotoxicity (degranulation and CD95-CD95L) are quite different (27, 47). Whereas the Ca2+-dependent degranulation pathway leads to cell death shortly after the onset of target cell contact with the CTL, CD95-dependent cytotoxicity in Ca2+-free media has an extended lag period (27, 47, 48). The lag phase may reflect the time required to attract sufficient CD95L for surface membrane expression and its implementation to ensure effective killing, as well as for downstream signaling of apoptosis. Note also that a lag period of several hours is required for CD95-based apoptosis triggered by a powerful CD95 Ab such as Jo2.

The regulation of CD95L expression seems to vary widely (15, 20, 22, 24, 27, 47, 49). The characteristics of the different CTL lines and hybridomas probably determine how they express CD95L. Actually, IL2, Con A, PI, and TCR triggering can all lead to CD95L expression in T cells (10, 22, 49). Hence, multiple signaling pathways may exist. Naive and memory CTL do not express CD95L, whereas cytocidal effectors do, as clearly demonstrated by Hamann et al. (40). PEL, like naive and memory cells, are small- to medium-sized nondividing lymphocytes (7–9 microns in diameter); however, they express potent, specific cytocidal activity (26) and may be regarded as activated, cytocidal effectors such as those described by Hamann et al. In in vivo studies, we have proved that these cells originate from Ag-stimulated dividing lymphoblasts (50). The small cytocidal PEL, upon IL-2 treatment, undergo blast transformation into large, dividing cytocidal blasts (51), expressing CD95L. Late-stage PEL (e.g., 17–20 days and longer after priming) exhibit no cytocidal activity and do not express CD95L; an extended PI treatment is required for their reactivation. The data presented here suggest that, like the triggering of T cell cytokines, CD95L expression and function in CTL is intrinsically related to the differentiation and activation state of the cells and is not directly dependent on continuous Ag-mediated TCR signaling.

1

Financial support has been obtained from an anonymous source of the Weizmann Institute and from the Israel Science Foundation.

3

Abbreviations used in this paper: CD95L, CD95 ligand; CsA, cyclosporin A; PI, PMA and ionomycin; PEL, peritoneal exudate CTL; P0, perforin knockout; NCS, newborn calf serum; PEB; PEL blasts; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; BFA, brefeldin A; ER, endoplasmic reticulum; ActD, actinomycin D.

1
Berke, G..
1995
. The CTL kiss of death.
Cell
81
:
9
2
Nagata, S., P. Golstein.
1995
. The Fas death factor.
Science
267
:
1449
3
Shiver, J. W., L. Su, P. A. Henkart.
1992
. Cytotoxicity with target DNA breakdown by rat basophilic leukemia cells expressing both cytolysin and granzyme A.
Cell
71
:
315
4
Berke, G..
1991
. The mechanism of lymphocyte-mediated killing: lymphocyte-triggered internal target disintegration.
Immunol. Today
12
:
396
5
Berke, G., D. Rosen, D. Ronen.
1993
. Mechanism of lymphocyte-mediated cytolysis: functional cytolytic T cells lacking perforin and granzymes.
Immunology
78
:
105
6
Kojima, H., N. Shinohara, S. Hanaoka, Y. Someya-Shirota, Y. Takagaki, H. Ohno, T. Saito, T. Katayama, H. Yagita, K. Okumura, Y. Shinkai, F. W. Alt, A. Matsuzawa, S. Yonehara, H. Takayama.
1994
. Two distinct pathways of specific killing revealed by perforin mutant cytotoxic T lymphocytes.
Immunity
1
:
357
7
Kagi, D., B. Ledermann, K. Burki, P. Seiler, B. Odermatt, K. J. Olsen, E. R. Podack, R. M. Zinkernagel, H. Hengartner.
1994
. Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforin-deficient mice.
Nature
369
:
31
8
Lowin, B., M. Hahne, C. Mattmann, J. Jschopp.
1994
. Cytolytic T-cell cytotoxicity is mediated through perforin and Fas lytic pathways.
Nature
370
:
650
9
Walsh, C. M., M. Matloubian, C. C. Liu, R. Ueda, C. G. Kurahara, J. L. Christensen, M. T. Huang, J. D. Young, R. Ahmed, W. R. Clark.
1994
. Immune function in mice lacking the perforin gene.
Proc. Natl. Acad. Sci. USA
91
:
10854
10
Rouvier, E., M. F. Luciani, P. Golstein.
1993
. Fas involvement in Ca++ independent T cell-mediated cytotoxicity.
J. Exp. Med.
177
:
195
11
Suda, T., T. Takahashi, P. Golstein, S. Nagata.
1993
. Molecular cloning and expression of the Fas ligand, a novel member of the tumor necrosis factor family.
Cell
75
:
1169
12
Nagata, S..
1997
. Apoptosis by death factor.
Cell
88
:
355
13
Itoh, N., S. Yonehara, A. Ishii, M. Yonehara, S. I. Mizushima, M. Sameshima, A. Hase, Y. Seto, S. Nagata.
1991
. The polypeptide encoded by the cDNA for human cell surface antigen Fas can mediate apoptosis.
Cell
66
:
233
14
Debatin, K. M., C. K. Goldmann, R. Bamford, T. A. Waldmann, P. H. Krammer.
1990
. Monoclonal-antibody-mediated apoptosis in adult T-cell leukemia.
Lancet
335
:
497
15
Anel, A., M. Buferne, C. Boyer, A. M. Schmitt-Verhulst, P. Golstein.
1994
. T cell receptor-induced Fas ligand expression in cytotoxic T lymphocyte clones is blocked by protein tyrosine kinase inhibitors and cyclosporin A.
Eur. J. Immunol.
24
:
2469
16
Dhein, J., H. Walczak, C. Baumler, K. M. Debatin, P. H. Krammer.
1995
. Autocrine T-cell suicide mediated by APO-1/(Fas/CD95).
Nature
373
:
438
17
Rogers, A. M., A. R. B. Thilenius, J. H. Russell.
1997
. Cyclosporine-insensitive partial signaling and multiple roles of Ca2+ in Fas ligand-induced lysis.
J. Immunol.
159
:
3140
18
Takayama, H., H. Kojima, N. Shinohara.
1995
. Cytotoxic T lymphocytes: the newly identified Fas (CD95)-mediated killing mechanism and a novel aspect of their biological functions.
Adv. Immunol.
60
:
289
19
Ramsdell, F., M. S. Seaman, R. E. Miller, T. W. Tough, M. R. Alderson, D. H. Lynch.
1994
. Gld/gld mice are unable to express a functional ligand for Fas.
Eur. J. Immunol.
24
:
928
20
Hanabuchi, S., M. Koyanagi, A. Kawasaki, N. Shinohara, A. Matsuzawa, Y. Nishimura, Y. Kobayashi, S. Yonehara, H. Yagita, K. Okumura.
1994
. Fas and its ligand in a general mechanism of T-cell-mediated cytotoxicity.
Proc. Natl. Acad. Sci. USA
91
:
4930
21
Vignaux, F., P. Golstein.
1994
. Fas-based lymphocyte-mediated cytotoxicity against syngeneic activated lymphocytes: a regulatory pathway.
Eur. J. Immunol.
24
:
923
22
Vignaux, F., E. Vivier, B. Malissen, V. Depraetere, S. Nagata, P. Golstein.
1995
. TCR/CD3 coupling to Fas-based cytotoxicity.
J. Exp. Med.
181
:
781
23
Berke, G..
1994
. The binding and lysis of target cells by cytotoxic lymphocytes: molecular and cellular aspects.
Annu. Rev. Immunol.
12
:
735
24
Walsh, C. M., A. A. Glass, V. Chiu, W. R. Clark.
1994
. The role of the Fas lytic pathway in a perforine-less CTL hybridoma.
J. Immunol.
153
:
2506
25
Ostergaard, H., K. Gorman, W. R. Clark.
1988
. Cloned CTL target cells fail to induce early activation events in effector CTLs.
Cell. Immunol.
114
:
188
26
Berke, G., K. A. Sullivan, D. B. Amos.
1972
. Rejection of ascites tumor allografts. I. Isolation, characterization and in vitro reactivity of peritoneal lymphoid effector cells from BALB/c mice immune to EL4 leukosis.
J. Exp. Med.
135
:
1334
27
Glass, A., C. M. Walsh, C. H. Lynch, W. R. Clark.
1996
. Regulation of the Fas lytic pathway in cloned CTL.
J. Immunol.
156
:
3638
28
Lynch, D. H., M. L. Watson, M. R. Alderson, P. R. Baum, R. E. Miller, T. Tough, M. Gilbson, T. Davis-Smith, C. A. Smith, K. Hunter, D. Bhat, W. Din, R. G. Goodwin, M. F. Seldin.
1994
. The mouse Fas-ligand gene is mutated in gld mice and is part of a TNF family gene cluster.
Immunity
1
:
131
29
Esser, M. T., D. M. Haverstick, C. L. Fuller, C. A. Gullo, V. L. Braciale.
1998
. Ca2+ signaling modulates cytolytic T lymphocyte effector functions.
J. Exp. Med.
187
:
1057
30
Kagi, D., F. Vignaux, B. Ledermann, K. Burki, V. Depraetere, S. Nagata, H. Hengartner, P. Golstein.
1994
. Fas and perforin pathways as major mechanisms of T cell-mediated cytotoxicity.
Science
265
:
528
31
Suda, T., S. Nagata.
1994
. Purification and characterization of the Fas-ligand that induces apoptosis.
J. Exp. Med.
179
:
873
32
Kojima, H., K. Eshima, H. Takayama, M. V. Sitkovsky.
1997
. Leukocyte function-associated antigen-1-dependent lysis of Fas+ (CD95+ /Apo-1+) innocent bystanders by antigen-specific CD8+ CTL.
J. Immunol.
159
:
2728
33
Smyth, M. J..
1997
. Fas ligand-mediated bystander lysis of syngeneic cells in response to an allogeneic stimulus.
J. Immunol.
158
:
5765
34
Liu, J., J. D. Farmer, W. S. Lane, J. Friedman, I. Weissman, S. L. Schreiber.
1991
. Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP-FK506 complexes.
Cell
66
:
807
35
Sigal, N. H., F. J. Dumont.
1992
. Cyclosporin A, FK-506 and rapamycin: pharmacologic probes of lymphocyte signal transduction.
Annu. Rev. Immunol.
10
:
519
36
Klausner, R. D., J. G. Donaldson, J. Lippincott-Schwartz.
1992
. Brefeldin A: insights into the control of membrane traffic and organelle structure.
J. Cell Bio.
116
:
1071
37
Pelham, H. R. B..
1991
. Mutiple targets for BFA.
Cell
67
:
449
38
Clark, W. R., C. M. Walsh, A. A. Glass, F. Hayashi, M. Matloubian, R. Ahmed.
1995
. Molecular pathways of CTL-mediated cytotoxicity.
Immunol. Rev.
146
:
33
39
Ogasawara, J., R. Watanabe-fukunaga, M. Adachi, A. Matsuzawa, T. Kasugai, Y. Kitamura, N. Itoh, T. Suda, S. Nagata.
1993
. Lethal effect of the anti-Fas antibody in mice.
Nature
364
:
806
40
Hamman, D., P. A. Baars, M. H. G. Rep, B. Hooibrink, S. R. Kerkhof-garde, M. Klein, R. A. W. V. Lier.
1997
. Phenotypic and functional separation of memory and effector human CD8+ cells.
J. Exp. Med.
186
:
1407
41
Kataoka, T., N. Shinohara, H. Takayama, K. Takaku, S. Kondo, S. Yonehara, K. Nagai.
1996
. Concanamycin A, a powerful tool for characterization and estimation of contribution of perforin- and Fas-based lytic pathways in cell-mediated cytotoxicity.
J. Immunol.
156
:
3678
42
Zagury, D., J. Bernard, N. Thierness, M. Feldman, G. Berke.
1975
. Isolation and characterization of individual functionally reactive cytotoxic T lymphocytes: conjugation, killing and recycling at the single cell level.
Eur. J. Immunol.
5
:
818
43
Geiger, B., D. Rosen, G. Berke.
1982
. Spatial relationships of microtubule-organizing centers and the contact area of cytotoxic T lymphocytes and target cells.
J. Cell Biol.
95
:
137
44
Zagury, D., J. Bernard, P. Jeannesson, N. Thiernesse, J. C. Cerottini.
1979
. Studies on the mechanism of T cell-mediated lysis at the single effector cell level. I. Kinetic analysis of lethal hits and target cell lysis in multicellular conjugates.
J. Immunol.
123
:
1604
45
Mariani, S. M., B. Matiba, C. Baumler, P. H. Krammer.
1995
. Regulation of cell surface APO-1/Fas (CD95) ligand expression by metalloproteases.
Eur. J. Immunol.
25
:
2303
46
Kayagaki, N., A. Kawasaki, T. Ebata, H. Ohmoto, S. Ikeda, S. Inoue, K. Yoshino, K. Okumura, H. Yagita.
1995
. Metalloproteinase-mediated release of human Fas ligand.
J. Exp. Med.
182
:
1777
47
Lowin, B., C. Mattman, M. Hahne, J. Jschopp.
1996
. Comparison of Fas (Apo-1/CD95)- and perforin-mediated cytotoxicity in primary T lymphocytes.
Int. Immunol.
8
:
57
48
Tirosh, R., G. Berke.
1985
. T lymphocyte mediated lysis as an excitatory process of the target.
Cell. Immunol.
95
:
113
49
Suda, T., T. Okazaki, Y. Naito, T. Yokota, N. Arai, S. Ozaki, K. Nakao, S. Nagata.
1995
. Expression of the Fas ligand in cells of T cell lineage.
J. Immunol.
154
:
3806
50
Denizot, F., A. Wilson, F. Battye, G. Berke, K. Shortman.
1986
. Clonal expansion of T cells: a cytotoxic T-cell response in vivo that involves precursor cell proliferation.
Proc. Natl. Acad. Sci. USA
83
:
6089
51
Berke, G., D. Rosen.
1988
. Highly lytic in vivo primed cytolytic T lymphocytes devoid of lytic granules and BLT-esterase activity acquire these constituents in the presence of T cell growth factors upon blast transformation in vitro.
J. Immunol.
141
:
1429