Abstract
TNF-related apoptosis-inducing ligand (TRAIL), a new member of TNF family, induces apoptotic cell death of various tumor cells. We recently showed that TRAIL mediates perforin- and Fas ligand (FasL)-independent cytotoxic activity of human CD4+ T cell clones. In the present study, we investigated the expression and function of TRAIL on murine lymphocytes by using newly generated anti-murine TRAIL mAbs. Although freshly isolated T, B, or NK cells did not express a detectable level of TRAIL on their surface, a remarkable level of TRAIL expression was induced preferentially on CD3− NK1.1+ NK cells after stimulation with IL-2 or IL-15. In contrast, TRAIL expression was not induced by IL-18, whereas it efficiently potentiated lymphokine-activated killer activity of NK cells. In addition to perforin inactivation and neutralization of FasL by anti-FasL mAb, neutralization of TRAIL by anti-TRAIL mAb was needed for the complete inhibition of IL-2- or IL-15-activated NK cell cytotoxicity against mouse fibrosarcoma L929 target cells, which were susceptible to both FasL and TRAIL. These results indicated preferential expression of TRAIL on IL-2- or IL-15-activated NK cells and its potential involvement in lymphokine-activated killer activity.
Tumor necrosis factor TNF-related apoptosis-inducing ligand (TRAIL)3/APO-2L is a 40-kDa type II transmembrane protein belonging to the TNF family and structurally and functionally homologous to Fas ligand (FasL), a most well-characterized death factor (1, 2, 3). TRAIL preferentially induces apoptotic cell death in a variety of transformed cells but not in normal cells (1, 2, 4). Although TRAIL mRNA has been found in various tissues and cells, its expression at protein levels and its physiological roles remain largely unknown (1). Thus far, at least four receptors, TRAIL-R1/DR4, TRAIL-R2/DR5/TRICK2/killer, TRAIL-R3/TRID/DcR1/LIT, and TRAIL-R4/TRUNDD/DcR2, have been identified that bind to TRAIL with similar affinities (5, 6, 7, 8, 9). Two of these receptors, TRAIL-R1 and TRAIL-R2, contain a cytoplasmic death domain homologous to that in Fas and TNFR-I. Binding of trimeric TRAIL to TRAIL-R1 or TRAIL-R2 causes aggregation of the death domain and recruitment of caspase-8 or -10 via Fas-associated death domain or a Fas-associated death domain-like adopter molecule, which subsequently leads to activation of the caspase cascade (5, 6, 7, 9). In contrast to these proapoptotic TRAIL-Rs, TRAIL-R3 exists as a glycosylphosphatidylinositol-anchored cell surface protein and has been shown to act as a decoy receptor (6, 7, 9). Because TRAIL-R3 mRNA was preferentially found in normal cells but not in transformed cells, it is supposed that TRAIL-R3 might be responsible for the cellular resistance of normal cells to TRAIL-mediated apoptosis (6, 7, 9). TRAIL-R4 contains a truncated cytoplasmic death domain that cannot mediate apoptosis but might activate NF-κB, which may protect the cells from death domain-containing TRAIL-R-mediated apoptosis (8, 9). Although mRNAs for these TRAIL-Rs have been detected in a wide variety of tissues and cells, physiological roles of these multiple TRAIL-Rs remain largely unknown.
CTL rapidly kill target cells in vitro via two major effector pathways, the perforin-mediated and FasL-mediated pathways (10, 11, 12). Perforin is stored in the cytoplasmic granules of CTL and secreted into intercellular junction between a CTL and a target cell after granule exocytosis, which makes pores in the target plasma membrane and leads to osmotic lysis and influx of granule-derived granzyme B. Granzyme B also activates the caspase cascade which leads to apoptotic cell death of target cells (13, 14). In the FasL-mediated pathway, the intercellular interaction between FasL on a CTL and Fas on a target cell induces Fas-mediated apoptosis of target cells via Fas-associated death domain-mediated activation of the caspase cascade. In contrast to the perforin-mediated pathway that can potentially lyse any target cell, FasL-mediated pathway is dependent on the susceptibility of target cells to FasL which is determined primarily by the expression of Fas on the target cells and secondarily by intracellular antiapoptotic molecules such as FLICE-like inhibitory protein (15). In addition to these two effector pathways, we recently demonstrated that TRAIL was constitutively expressed on human CD4+ T cell clones and was involved in their cytotoxicity against TRAIL-sensitive target cells (16). Thomas et al. also reported that TRAIL was partly responsible for mediating the cytotoxicity of human CD4+ T cell clones against melanoma cells (17). These findings supplemented TRAIL as an additional component of T cell cytotoxicity. However, the expression and function of TRAIL on the other types of effector cells such as NK cells remain largely unknown.
NK cells play important roles in immune surveillance against transformed cells and virus-infected cells. In response to infection by certain viruses and pathogens, NK cells produce proinflammatory and antiviral cytokines such as TNF-α and IFN-γ (18, 19). Moreover, NK cells directly kill transformed cells and virus-infected cells in a MHC-unrestricted and Ag-independent manner. Previous studies with perforin- or FasL-deficient mice showed that these effector molecules constitute predominant pathways of NK cell cytotoxicity (20, 21). It has been also reported that the FasL- and perforin-mediated NK cell cytotoxicity can be augmented by some cytokines including IL-2, IL-15, and IL-18 (22, 23, 24, 25).
In the present study, by using newly generated anti-murine TRAIL mAbs, we examined the expression and function of TRAIL on murine lymphocytes. Cell surface TRAIL expression was not detectable on freshly isolated splenic lymphocytes as assessed by flow cytometry. However, a remarkable level of TRAIL was induced preferentially on CD3− NK1.1+ NK cells after stimulation with IL-2 or IL-15, which was involved in their cytotoxicity against TRAIL-sensitive target cells along with perforin and FasL. In contrast to IL-2 and IL-15, IL-18 did not induce TRAIL expression but augmented NK cell cytotoxicity in a perforin- and FasL-dependent manner. These results indicated that IL-2 and IL-15 are potent inducers of TRAIL-mediated NK cell cytotoxicity against tumor target cells. The physiological and clinical relevancies of this finding are discussed.
Materials and Methods
Animals
Six-wk-old male C57BL/6 (B6) mice and F344/DuCrj rats were purchased from SLC (Shizuoka, Japan).
Cell lines
Mouse B lymphoma 2PK-3, myeloma P3U1 (P3 × 63Ag8U.1), mastocytoma P815 (H-2d), T lymphoma YAC-1 (H-2a), melanoma B16 (H-2b), macrophage RAW 264, and hamster fibroblast BHK21 were obtained from the American Type Culture Collection (Manassas, VA) and cultured in RPMI 1640 containing 10% FCS, 100 μg/ml streptomycin and penicillin, and 2 mM glutamine (culture medium). A mouse fibrosarcoma L929 (H-2k) was obtained from Japanese Cancer Research Resource Bank (Tokyo, Japan) and maintained in the culture medium. 2PK-3 transfectants, mTRAIL/2PK-3 and mFasL/2PK-3, that stably express mTRAIL and mFasL, respectively, were also maintained in the culture medium. For preparation of IL-2, IL-15, or IL-18 blasts, B6 splenocytes were cultured at 2 × 106 cells/ml in the culture medium supplemented with IL-2 (500 U/ml), IL-15 (150 ng/ml), or IL-18 (100 ng/ml) for 6 days.
Reagents
Mouse IL-15 was purchased from PharMingen (San Diego, CA). Human IL-2 was provided by Shionogi (Kamakura, Japan). Mouse IL-18 was prepared as described before (26). An anti-mFasL mAb (MFL2) and soluble DR5-Ig fusion protein were prepared as described previously (16, 27). An anti-murine TNF-α mAb (MP-6 XT22) and mouse TNF-α were purchased from PharMingen.
Preparation of mTRAIL or mFasL transfectants
mTRAIL cDNA was prepared by RT-PCR amplification of total RNA from Con A-activated B6 splenocytes with an oligonucleotide primer corresponding to the first six codons as the 5′-primer and that corresponding to the last six codons as the 3′-primer, according to the published sequence (1). The 5′- and 3′-primers were tagged with a XhoI and a NotI site, respectively. After XhoI and NotI digestion, the PCR product of 850 bp was subcloned into pBluescript II SK(+) and the nucleotide sequence was confirmed by using an automated sequencer (Applied Biosystems, Foster City, CA) and a fluoresceinated dye terminator cycle sequencing method. The 850-bp cDNA was then transferred into the XhoI and NotI sites of the pMKITNeo expression vector (kindly provided by Dr. K. Maruyama, Tokyo Medical and Dental University, Tokyo, Japan). For generating mTRAIL transfectants, mTRAIL/pMKITNeo vector was transfected into 2PK-3 cells by electroporation (290 V, 960 μF) with a Gene Pulser (Bio-Rad, Hercules, CA). After selection with 1 mg/ml G418 and cloning by limiting dilution, a stable transfectant, designated mTRAIL/2PK-3, was selected by the staining with DR5-Ig. In a similar way, BHK21 cells stably expressing mTRAIL (mTRAIL/BHK) were prepared. Murine FasL-expressing 2PK-3 cells (mFasL/2PK-3) were generated as described before (27).
Flow cytometric analysis
2PK-3, mTRAIL/2PK-3, and mFasL/2PK-3 cells (1 × 106) were incubated with l μg of biotinylated mAb or DR5-Ig for 1 h at 4°C followed by PE-labeled avidin or PE-labeled anti-human IgG (PharMingen), respectively. After washing with PBS, the cells were analyzed on a FACScan (Becton Dickinson, San Jose, CA), and data were processed by using the CELLQuest program (Becton Dickinson). In some experiments, lymphocytes were stained with 1 μg biotinylated mAb followed by PE-labeled avidin, FITC-labeled anti-NK1.1 mAb (PharMingen), and peridinin chlorophyll protein-labeled anti-CD3 mAb (PharMingen).
Generation of anti-mTRAIL mAbs
A F344/DuCrj rat was immunized with mTRAIL/2PK-3 (2 × 107 cells) several times at 10-day intervals. Three days after final immunization, the splenocytes were fused with P3U1 mouse myeloma cells as described previously (28). After hypoxanthine-aminopterin-thymidine selection, the Abs that can inhibit cytotoxic activity of mTRAIL/2PK-3 against L929 were screened. Two hybridomas producing mAbs (N2B1 and N2B2) were identified by their strong inhibitory effects and cloned by limiting dilution. N2B1 and N2B2 (both rat IgG2a,κ) were purified from culture supernatants by protein G affinity chromatography.
Cytotoxic assay
A standard 51Cr release assay was performed as described previously (29). Briefly, 51Cr-labeled target cells (1 × 104) and effector cells were mixed in U-bottom wells of a 96-well microtiter plate at the indicated E:T ratios. After 8 h of incubation, cell-free supernatants were collected, and radioactivity was measured in a gamma counter. The percentage of specific 51Cr release was calculated as described before (30). In some experiments, the effector cells were pretreated with 20 nM concanamycin A (CMA) (Wako Pure Chemicals, Osaka, Japan) for 2 h to inactivate perforin (31). Anti-mFasL mAb (MFL2) and/or N2B2 were added to a final concentration of 10 μg/ml at the start of the cytotoxic assay.
In some experiments, IL-2- or IL-15 blasts were stained with PE-labeled anti-CD3 mAb (PharMingen) or PE-labeled anti-NK1.1 mAb (PharMingen). Then, CD3− (NK) and NK1.1− (T) populations, respectively, were isolated by sorting on a FACStar (Becton Dickinson) and used as effector cells. The purities were >95% CD3−NK1.1+ and >90% CD3+NK1.1−, respectively.
A [3H]TdR release assay was performed as described previously (16).
Results
Characterization of mTRAIL transfectants
Although we and others have recently showed that TRAIL is involved in cytotoxic activity of human CD4+ T cell clones against certain target cells, physiological functions of TRAIL remain largely unknown. Further studies in the murine system will be helpful for elucidating the physiological and pathological roles of TRAIL. To characterize the expression and function of TRAIL in the murine system, we first generated stable cDNA transfectants expressing mTRAIL. The mTRAIL cDNA was isolated by RT-PCR from Con A-activated B6 splenocytes and subsequently subcloned into expression vector pMKITNeo. Mouse B lymphoma 2PK-3 cells, which were totally resistant to recombinant TRAIL-induced cytotoxicity (our unpublished data), were transfected with mTRAIL/pMKITNeo. Cell surface expression of TRAIL was first verified by cell surface staining with DR5-Ig. As represented in Fig. 1⇓A, DR5-Ig specifically stained mTRAIL/2PK-3 but not 2PK-3. An anti-mFasL mAb (MFL2), one we previously generated (27), stained mFasL/2PK-3 but not mTRAIL/2PK-3 (not shown), indicating that MFL2 does not cross-react with mTRAIL.
Characterization of mTRAIL transfectants. A, Cell surface staining with DR5-Ig. 2PK-3 and mTRAIL/2PK-3 cells were stained with DR5-Ig and PE-labeled anti-human IgG Ab (——). ·····, Background staining with control human IgG and PE-labeled anti-human IgG Ab. B, Cytotoxic activity of mTRAIL or mFasL transfectants against tumor cell lines. Cytotoxic activity of mTRAIL/2PK-3 (▪), mFasL/2PK-3 (▨), or 2PK-3 (□) was tested against the indicated target cells in an 8-h 51Cr release assay at an E:T ratio of 10. Data represent mean ± SD of triplicate samples. Similar results were obtained in two independent experiments.
Although various human tumor cell lines were sensitive to TRAIL-induced cytotoxicity in vitro (1, 2), little is known about the sensitivity of murine tumor cell lines to TRAIL-induced cytotoxicity. Then, we examined the susceptibility of some mouse tumor cell lines to TRAIL- and FasL-induced cytotoxicity by using mTRAIL/2PK-3 and mFasL/2PK-3 transfectants, respectively. As represented in Fig. 1⇑B, both mTRAIL/2PK-3 and mFasL/2PK-3 cells efficiently lysed L929 cells in the 8-h 51Cr release assay, whereas no significant cytotoxicity was observed with the parental 2PK-3 cells. Because L929 cells are highly sensitive to TNF-mediated cytotoxicity, we tested mTRAIL/2PK-3 or mFasL/2PK-3 cytotoxicity against L929 in the presence of anti-TNF mAb. However, no inhibition was observed even with 50 μg/ml anti-TNF mAb (not shown). In addition, neither soluble TNF activity in their supernatants nor membrane-bound TNF on these cells was detected (not shown). These results ruled out the possible involvement of TNF in the cytotoxicity of these transfectants against L929. A mastocytoma cell line P815 was sensitive to FasL but not to TRAIL. A melanoma cell line B16 and a T lymphoma cell line YAC-1 were resistant to both TRAIL and FasL. Similar levels of mTRAIL/2PK-3 or mFasL/2PK-3 cytotoxicity against these targets were also observed as estimated by [3H]TdR release assay (not shown), suggesting apoptotic death of these target cells.
Characterization of anti-mTRAIL mAbs
To further characterize the expression and function of mTRAIL, we generated two mAbs that specifically bind to mTRAIL and block its cytotoxic activity. Hybridomas were prepared from splenocytes from a F344/DuCrj rat immunized with the mTRAIL/2PK-3 cells. Two hybridomas, producing mAbs designated N2B1 and N2B2 (both rat IgG2aκ), were selected by their strong ability to block the mTRAIL/2PK-3 cytotoxicity against L929 cells. As represented in Fig. 2⇓A, both N2B1 and N2B2 bound to mTRAIL/2PK-3, but not to 2PK-3, cells as estimated by cell surface staining. These mAbs also reacted with mTRAIL/BHK, but not BHK, cells (not shown). Furthermore, as represented in Fig. 2⇓B, both N2B1 and N2B2 mAbs neutralized mTRAIL/2PK-3 cytotoxicity against L929 cells in a dose-dependent manner. Neither N2B1 nor N2B2 could inhibit the cytotoxic activities of recombinant mTNF or mFasL/2PK-3 against L929 cells (not shown). In addition, these mAbs could not stain mFasL/2PK-3 cells or LPS-stimulated RAW 264 cells expressing membrane TNF (not shown). These results indicated that these mAbs do not cross-react with mFasL or mTNF.
Reactivity of anti-mTRAIL mAbs. A, Cell surface staining of mTRAIL transfectants. 2PK-3 and mTRAIL/2PK-3 cells were stained with biotinylated anti-mTRAIL mAbs, N2B1 and N2B2, followed by PE-labeled avidin (——). ·····, Background staining with biotinylated control IgG and PE-labeled avidin. B, Inhibition of mTRAIL cytotoxicity by anti-mTRAIL mAbs. Cytotoxic activity of mTRAIL/2PK-3 was tested against L929 cells in the absence (▴) or presence of serially diluted N2B1 (○) or N2B2 (•) in an 8-h 51Cr release assay at an E:T ratio of 10. Data represent mean ± SD of triplicate samples. Similar results were obtained in two independent experiments.
Expression of mTRAIL on IL-2- or IL-15-activated NK cells
We recently showed that some human CD4+ T cell clones constitutively expressed TRAIL on their surface as estimated by staining with anti-human TRAIL mAbs (16). However, it remains to be determined whether TRAIL can be expressed not only on particular T cell clones but also on primary lymphocytes. Especially in the murine system, expression of TRAIL at the protein level remains largely unknown. Thus, we first examined the expression of TRAIL on freshly isolated splenic lymphocytes by 3-color flow cytometric analysis with N2B2. Freshly isolated B6 splenocytes can be classified by their expression of CD3 and NK1.1 into four populations, CD3+ NK1.1− (T cells), CD3−NK1.1+ (NK cells), CD3−NK1.1− (mostly B cells), and CD3+NK1.1+ (NKT cells). No detectable level of TRAIL expression was found on all of these cells (Fig. 3⇓A). Likewise, neither freshly isolated thymocytes nor lymph node cells expressed TRAIL on their surface (not shown). We also examined the surface expression of FasL, which constitutes another pathway of T and NK cell cytotoxicity (10, 11, 12, 20). As represented in the bottom panels of Fig. 3⇓A, a marginal but significant level of FasL expression was observed on NK cells but not on the other cells, as reported by others (20).
Expression of mTRAIL and mFasL on resting or cytokine-activated lymphocytes. Freshly isolated B6 splenocytes (SPC) (A) or those stimulated with IL-2 (500 U/ml), IL-15 (150 ng/ml), or IL-18 (100 ng/ml) for 6 days (B) or with IL-2 (500 u/ml) for the indicated periods (C) were stained with biotinylated N2B2 or MFL2 followed by PE-labeled avidin, FITC-labeled anti-NK1.1 mAb, and peridinin chlorophyll protein-labeled anti-CD3 mAb. ——, staining with N2B2 or MFL2 on the cells gated in R1–R5. ·····, Background staining with biotinylated control IgG and PE-labeled avidin. In C, only the CD3−NK1.1+ cells were electronically gated. Similar results were obtained in three independent experiments
It has been reported that FasL can be up-regulated on T and NK cells after stimulation with some cytokines such as IL-2, IL-15, and IL-18 (22, 23, 24, 32). Thus, we next examined whether TRAIL expression can be induced by these cytokines. B6 splenocytes were stimulated with IL-2, IL-15, or IL-18 for 6 days, and then the expression of TRAIL was examined by three-color flow cytometric analysis (Fig. 3⇑B). When stimulated with IL-2, the resulting lymphoblasts were composed of two populations, CD3+NK1.1− (T) and CD3−NK1.1+ (NK) cells. A remarkable level of TRAIL expression was found on the NK cells but not on the T cells. IL-15, which shares β and γ subunits of the receptor complex with IL-2 (33, 34, 35), exhibited an effect similar to that of IL-2, which produced both T and NK populations and induced TRAIL expression preferentially on the latter. In contrast to these cytokines, IL-18, which stimulates a distinct receptor (IL-18R/IL-1Rrp) (36, 37), induced only lymphoblasts of the NK phenotype (Fig. 3⇑B). No detectable level of TRAIL expression was found on these IL-18 blasts. Taken together, these results indicated that IL-2 and IL-15 but not IL-18 are potent inducers of TRAIL expression on murine NK cells. We further examined the kinetics of TRAIL expression on NK cells after IL-2 stimulation. The TRAIL expression appeared on NK cells after 2 days after the IL-2 stimulation and reached peaks at 5 days (Fig. 3⇑C). A similar kinetics of TRAIL expression on NK cells was found after the IL-15 stimulation (not shown). We also examined whether IL-2, IL-15, and IL-18 can up-regulate the surface expression of FasL and found that all these cytokines substantially up-regulated the FasL expression on both T and NK cells (Fig. 3⇑B bottom). These results suggested a differential regulation of TRAIL and FasL expression by these cytokines.
Involvement of TRAIL in IL-2- or IL-15-activated NK cell cytotoxicity
It has been well known that IL-2-activated lymphocytes exhibit MHC-unrestricted, Ag-independent cytotoxicity against various tumor cells, including NK-resistant targets such as P815, i.e., so-called lymphokine-activated killer (LAK) activity (32, 38). We compared the LAK activity of the IL-2, IL-15, or IL-18 blasts against those of several target cells. The B6-derived IL-2 blasts spontaneously killed not only the NK-susceptible YAC-1 but also NK-resistant B16, P815, and L929 target cells in an MHC-unrestricted manner (Fig. 4⇓). A similar level of LAK activity against these target cells was observed with the IL-15 blasts. The IL-18 blasts exhibited much higher cytotoxic activities against all these target cells than those of the IL-2 and IL-15 blasts (Fig. 4⇓). Similar results were also obtained as estimated by [3H]TdR release assay (not shown).
LAK activity of IL-2-, IL-15-, or IL-18 blasts. Cytotoxic activity of B6 splenocytes that had been cultured with IL-2 (○), IL-15 (•), or IL-18 (▵) for 6 days was tested against YAC-1, B16, P815, or L929 target cells by an 8-h 51Cr release assay at the indicated E:T ratios. Data represent mean ± SD of triplicate samples. Similar results were obtained in three independent experiments.
It has been reported that IL-2, IL-15, and IL-18 can augment T and NK cell cytotoxicity by up-regulating FasL- and/or perforin-mediated cytotoxicity (22, 23, 24, 25, 32). Then, we examined the contribution of TRAIL, FasL, and perforin to LAK activity of the IL-2, IL-15, or IL-18 blasts. N2B2 and MFL2 were used to assess the contribution of TRAIL and FasL, respectively. CMA pretreatment that inactivates perforin (31) was used to assess the contribution of perforin. Cytotoxic activities of the IL-2 blasts and the IL-15 blasts against the FasL- and TRAIL-resistant target cells (YAC-1 and B16) were almost completely abrogated by CMA alone (Fig. 5⇓A), indicating a dominant contribution of perforin. Although cytotoxic activity of the IL-18 blasts against B16 was almost completely abrogated by the CMA treatment, that against YAC-1 was not completely inhibited even by the combination of CMA, MFL2, and N2B2. This suggests a possible existence of another cytotoxic pathway that does not depend on perforin, FasL, or TRAIL. When the FasL-sensitive but TRAIL-resistant P815 was used as the target, CMA mostly inhibited and the combination with MFL2 completely abrogated cytotoxic activities of the IL-2, IL-15, and IL-18 blasts (Fig. 5⇓A), indicating major and minor contribution of perforin and FasL, respectively. When the TRAIL- and FasL-sensitive L929 was used as the target, CMA alone only partially inhibited cytotoxic activities of the IL-2 blasts and the IL-15 blasts (Fig. 5⇓A). Further inhibition was observed by the combination with either MFL2 or N2B2, and almost complete inhibition was achieved by the combination with both. These results indicated that TRAIL, in addition to perforin and FasL, is involved in the cytotoxic activities of the IL-2 and IL-15 blasts against L929. In contrast, the IL-18 blasts appeared to exert perforin- and FasL-dependent but TRAIL-independent cytotoxicity, because the combination of CMA and MFL2 completely inhibited the cytotoxic activity of the IL-18 blasts against L929 in the absence of N2B2.
Contribution of perforin, FasL, and TRAIL to IL-2-, IL-15-, or IL-18-activated NK cell cytotoxicity. A, Cytotoxic activity of B6 splenocytes that had been stimulated with IL-2 (500 U/ml), IL-15 (150 ng/ml), or IL-18 (100 ng/ml) for 6 days was tested against YAC-1, B16, P815, or L929 target cells in the presence or absence of 20 nM CMA, 10 μg/ml each of N2B2, and/or MFL2 by an 8-h 51Cr release assay at an E:T ratio of 5. Data represent mean ± SD of triplicate samples. Similar results were obtained in three independent experiments. B, Restricted expression of functional TRAIL on NK cells. Cytotoxic activity of NK1.1+CD3− NK or NK1.1−CD3+ T population in IL-2- or IL-15 blasts was tested against L929 cells in the presence or absence of 20 nM CMA, 10 μg/ml each of N2B2 and/or MFL2 by an 8-h 51Cr release assay at an E:T ratio of 5. Data represent mean ± SD of triplicate samples. Similar results were obtained in three independent experiments.
As shown in Fig. 3⇑B, the IL-2 or IL-15 blasts contain two populations of T and NK cells. To investigate the effector mechanisms for these cells separately, we isolated each population by flow cytometry. Then, the cytotoxic activities of these cells against L929 were tested in the presence or absence of CMA, MFL2, and/or N2B2. As shown in Fig. 5⇑B, IL-2-stimulated NK cells exhibited substantially higher cytotoxicity than T cells. The combination of CMA, N2B2, and MFL2 was required to completely inhibit the cytotoxic activity of IL-2-activated NK cells against L929. In contrast, N2B2 was not required for complete inhibition of the IL-2-activated T cell cytotoxicity. Similar results were obtained with the IL-15 blasts (Fig. 5⇑B). These results indicated that functional TRAIL was expressed on IL-2- or IL-15-activated NK cells but not on T cells.
Discussion
We recently showed that TRAIL is constitutively expressed on some human CD4+ T cell clones and is involved in their cytotoxicity against certain tumor target cells (16). In the present study, we investigated the expression and function of TRAIL in the murine system by using newly established neutralizing mAbs against mTRAIL. No detectable level of TRAIL expression was found on freshly isolated lymphocytes. However, a remarkable level of TRAIL expression was induced preferentially on CD3−NK1.1+ NK cells after IL-2 or IL-15 stimulation, which partly mediated LAK cytotoxicity against a TRAIL-sensitive target, L929. These results indicated the potential involvement of TRAIL in mediating NK cell cytotoxicity and that IL-2 and IL-15 are potent inducers of TRAIL expression on NK cells.
In the present study, we observed that IL-2 and IL-15 exhibited similar effects; i.e., both induced the surface expression of TRAIL preferentially on NK cells and induced LAK activities to a similar level ( Figs. 3–5⇑⇑⇑). These results are consistent with previous reports indicating that IL-2 and IL-15 share two receptor components (IL-2Rβ/IL-15Rβ and IL-2Rγ/IL-15Rγ chains) and exert mostly overlapped biological activities including the up-regulation of NK activity (33, 34, 35). It has been reported that exogenously administered or genetically expressed IL-2 or IL-15 exhibited antitumor effects in mouse models by activating NK cells (38, 39, 40). The TRAIL-mediated NK cell cytotoxicity might be partly responsible for the antitumor effects of IL-2 and IL-15 in vivo. It has also been shown that IL-15 plays an important role in the development and maturation of NK cells in vivo (41, 42). Indeed, IL-15Rα-chain knockout mice have a defect in the development of NK cells and almost completely lack mature NK cells (41). Thus, IL-15 might up-regulate NK activity in vivo not only by inducing the development of NK cells but also by augmenting the perforin-, FasL-, and TRAIL-dependent cytotoxicity.
As demonstrated in Fig. 3⇑, no detectable level of TRAIL expression was found on freshly isolated lymphocytes. Furthermore, in our ongoing studies with N2B2, we could not find a detectable level of mTRAIL expression on T and B cells, even when these cells were activated by various stimuli such as anti-CD3 mAb, PHA, CD40 ligand, LPS, and PMA + ionomycin (our unpublished data). Thus, the cell surface expression of mTRAIL presently appeared to be largely confined to activated NK cells in vitro. This is inconsistent with a previous report (43), which demonstrated the cell surface expression of TRAIL on freshly isolated murine B cells and T cell blasts by utilizing a rabbit polyclonal Ab (pAb) to a peptide sequence in the extracellular region of human TRAIL. However, they did not address functional activity of TRAIL on these cells. It has been noticed that some commonly used pAbs to FasL exhibited aberrant reactivities (44). In this respect, their staining might result from some nonspecific reactivity of their pAb.
Molecular mechanisms for the restricted induction of TRAIL expression on NK cells by IL-2 or IL-15 remain to be determined. This could not be explained by a differential expression of IL-2R and IL-15R among lymphocytes, because FasL expression was inducible on both NK and T cells after IL-2 or IL-15 stimulation (Fig. 3⇑B, bottom). Further studies are needed to elucidate the mechanisms for restricted expression of TRAIL on NK cells.
IL-18 is a recently identified cytokine that strongly induces IFN-γ production by NK cells (26, 37) and augments T and NK cell cytotoxicity (24, 25, 26, 37). In the present study, we found that the stimulation of splenocytes with IL-18 selectively produced lymphoblasts of NK cell phenotype. This might be the result of the constitutive expression of IL-18R on NK cells but not on the other types of splenic lymphocytes (25, 45, 46). IL-18 augments NK cell cytotoxicity in a FasL- and perforin-dependent manner (24, 25). Consistently, we also found that IL-18-activated NK cells up-regulated the cell surface expression of FasL and that these cells killed Fas-sensitive targets such as P815 and L929 via FasL- and perforin-mediated pathways (Fig. 5⇑A). However, in contrast to IL-2 and IL-15, IL-18 did not induce the TRAIL expression (Fig. 3⇑). The apparently distinct regulation of TRAIL-, FasL-, and perforin-mediated cytotoxicity by these cytokines might be relevant to differential roles of these effector molecules in the immune system.
We observed that the IL-18-activated NK cells could lyse YAC-1 target cells even in the presence of CMA, anti-FasL, and anti-TRAIL mAbs (Fig. 5⇑A). It is has been known that membrane-bound or soluble form of TNF-α can participate in CTL or IL-2-induced LAK cytotoxicity (32, 47). However, the addition of a neutralizing anti-TNF-α mAb did not abolish the IL-18-induced LAK cytotoxicity against YAC-1 even in the presence of CMA + anti-FasL + anti-TRAIL (data not shown). These results suggest a possible existence of another effector mechanism for NK cytotoxicity, which can be specifically induced by IL-18. Further study is now under way to characterize this perforin-, FasL-, TRAIL-, and TNF-α–independent cytotoxic pathway.
It has been suggested that perforin and FasL constitute two predominant pathways of NK cell cytotoxicity (20, 21). In the present study, we demonstrated an additional TRAIL-mediated pathway in NK cell cytotoxicity. A very recent study by Zamai et al. also showed that human CD161+CD56− immature NK cell lines, which were established from CD34+Lin− progenitor cells in cord blood by using stem cell factor + IL-2, exhibited TRAIL-dependent but FasL- and perforin-independent cytotoxicity (48). Their results together with our present study supplement TRAIL as an additional component of NK cell cytotoxicity.
Although our present study demonstrated the potential involvement of TRAIL in NK cell cytotoxicity, its physiological roles remain to be determined. It has been well known that NK cells play important roles in immune surveillance against virus-infected or transformed cells (18, 19). In this respect, the TRAIL-mediated NK cytotoxicity may at least partly participate in the elimination of virus- or intracellular pathogen-infected cells. In fact, IL-15 production was inducible following infection with certain viruses such as herpesvirus (49). The neutralizing anti-mTRAIL mAb generated in the present study will be useful for further investigation of the physiological functions of TRAIL in the murine system.
Footnotes
-
↵1 This work was supported by grants from the Science and Technology Agency, the Ministry of Education, Science and Culture, and the Ministry of Health, Japan. N.K. is a Research Fellow of the Japan Society for the Promotion of Science.
-
↵2 Address correspondence and reprint requests to Dr. Hideo Yagita, Department of Immunology, Juntendo University School of Medicine, 2-1-1 Hongo, Bunkyo-ku, Tokyo 113-8421, Japan. E-mail address: hyagita{at}med.juntendo.ac.jp
-
↵3 Abbreviations used in this paper: TRAIL, TNF-related apoptosis-inducing ligand; mTRAIL, murine TRAIL; FasL, Fas ligand; mFasL, murine FasL; pAb, polyclonal Ab; CMA, concanamycin A; LAK, lymphokine-activated killer.
- Received March 16, 1999.
- Accepted June 9, 1999.
- Copyright © 1999 by The American Association of Immunologists
References
In this issue
