- Email: [email protected]
Analysis of granzyme B and caspases pathway in xenogeneic cytotoxicity
Analysis of Granzyme B and Caspases Pathway in Xenogeneic Cytotoxicity H. Nakajima, I. Fujiwara, N. Mizuta, K. Sakaguchi, and H. Yamagishi
A
LTHOUGH there is increased evidence that natural killer (NK) cells play important roles in early phases of xenograft rejection via antibody-dependent cell-mediated cytotoxicity (ADCC), the molecular mechanism of how NK cells can give damage on xenogeneic target cells is still unknown. We have previously reported that perforin/granzymes-cell mediated cytotoxicity (P/G-CMC) plays a predominant role in human–anti-pig ADCC and induces apoptosis in target cells,1,2 however, its downstream biochemical events are not investigated. According to the observation using caspase inhibitors, caspase family is thought to be a common mediator of apoptotic cell death. Sequential activation of caspase 8 and caspase 3 during Fas-mediated apoptosis has been shown to be a critical event for apoptosis.3 Among many granzymes contained in lymphocyte granules, on the other hand, granzyme B (GB) has been shown to be the most critical protease to causes apoptotic cell death when perforin is present.4 GB has a substrate specificity and inhibitor profile similar to that of the caspase family, that also cleave substrates after aspartyl residues.5,6 In addition, GB has been shown in vitro to cleave and activate various members of the caspase subfamily: caspase-3 (CPP32), caspase-7 (Mch-3), caspase-10 (Mch-4), and caspase-8 (FLICE/Mch5), but not the ICE-related subfamily (caspase-1), leading directly to cleavage of nuclear target proteins such as caspase-activated-DNAase (CAD), poly adenosine diphosphate ([adp]-ribose)polymerase, and lamin. Those observation suggest that an interaction between GB and caspases plays a prominent role in the pathway of NK cell–mediated killing. Although contribution of these caspases in allogeneic cytotoxicity has been reported recently,7 much less is known about xenogeneic cytotoxicity. In this study, we used in vitro xenogeneic ADCC model and investigated the role of GB and caspases in xenogeneic cytotoxicity. MATERIALS AND METHODS Cytotoxicity Assays Human peripheral blood lymphocyte cells (PBL) were isolated by Ficoll-Paque gradient centrifugation and purified from adherent cells by passage over nylon wool column. These PBL in 5% heat-inactivated human serum were used as ADCC effector cells. PK15 (pig kidney cell line) were obtained from American Type Cell Culture Collection (Rockville, MD) and used as target cells. To detect membrane damage as cell lysis or DNA damage as a marker 0041-1345/00/$–see front matter PII S0041-1345(00)01044-7 932
for apoptosis in cytotoxicity assay, PK15 were labeled with 51Cr or 3 H-thymidine and tested for their release in 4-hour assay, as described previously.1 In both cytotoxicity assays, results are given as mean ⫾ SE of triplicate wells. All the results are representative of at least three independent experiments.
GB and Caspases Inhibition To inhibit GB activity, effector PBL were preincubated in 40 mol/L of zAAD-cmk(benzyloxycarbonyl-Ala-Ala-Asp-chlorometyl ketone)8 at 37°C for 30 minutes, and used in cytotoxicity assays as described above. In terms of caspases inhibition, zYVAD-fmk (z-Tyr-Val-Ala-Asp-fluoromethyl ketone, caspase-1 inhibitor), zDEVD-fmk (z-Asp-Glu-Val-Asp-fmk, caspase-3 inhibitor), zIETD fmk (z-Ile-Glu-Thr-Asp-fmk, caspase-8 inhibitor), and zFA-fmk (z-Phe-Ala-fmk, negative control) (Kamiya Biochemical Co., WA) were used. These compounds not only lead to irreversible enzyme inhibition but also are esterified to promote membrane permeability.9,10 Target PK15 were preincubated in 200 mol/L of each peptide-fmk at 37°C for 1 hour and the cytotoxicity assays were carried out in 20 mol/L of each of these inhibitors.
RESULTS
Although there is considerable evidence that perforin pore formation allows the access of granzymes to the target cells cytoplasm where these molecules induce DNA fragmentation by activating caspases in allogeneic cytotoxicity,4,7 less is known about xenogeneic cytotoxicity. Since we have previously shown that P/G-CMC is an only effector mechanism in this xenogeneic ADCC model,1,2 we have tried to inhibit GB by preloading zAAD-cmk to effector PBL. As shown in Figure 1, membrane damage was partially rescued, however, DNA fragmentation was greatly inhibited to the almost basal levels. These results indicate that GB is not only a key factor for inducing nuclear damage, but also contributes to lysis of the target cells in xenogeneic ADCC. Next, to examine which caspases were involved in NK-cell– induced xenogeneic ADCC, we designed the experiments in From the Second Department of Surgery (H.N., I.F., N.M., K.S.); Ayabe Municipal Hospital (I.F.); Department of Digestive Surgery (H.Y.), Kyoto Prefectural University of Medicine, Kyoto, Japan. Address reprint requests to Hiroo Nakajima, MD, PhD, Second Department of Surgery, Kyoto Prefectural University of Medicine, 465 Kawaramachi Hirokoji, Kamigyo, Kyoto 602-0841, Japan. © 2000 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010 Transplantation Proceedings, 32, 932–934 (2000)
XENOGENEIC CYTOTOXICITY
933
Fig 1. Effect of granzyme B and caspases inhibition on xenogeneic ADCC. 51Cr- or 3H-thymidine-labeled PK15 were incubated with PBL for 4 hours in 5% heat-inactivated human serum at ET ratio ⫽ 50 and 51Cr (upper panels) and 3H-DNA (lower panels) release were compared. Effector PBL were preincubated with zAAD-cmk to inhibit granzyme B (A1, 2). Target PK15 were preincubated with zYVAD-fmk (B1, 4), zDEVD-fmk (B2, 5), and zIETD-fmk (B3, 6), and cytotoxicity assays were performed.
which target PK15 were loaded with tetrapeptide inhibitors known to inhibit caspases. zYVAD-fmk, a selective inhibitor of caspase-1, inhibited neither membrane nor DNA damage (Fig 1). In contrast, zDEVD-fmk and zIETD-fmk, a selective inhibitor of caspase-3 and caspase-8, respectively, partially inhibited DNA fragmentation (Fig 1), but not the membrane damage (Fig 1). These results were consistent with previous observation that GB did activate caspase-3 and -8, but did not caspase 1,6 and that caspases inhibition did not affect the lytic activity, whereas DNA damage is reduced by the inhibition of CPP32-related caspase subfamily.7 DISCUSSION
Concerning the effector mechanism of xenogeneic ADCC, although we have recently found that it can cause apoptotic cell death in PK15 via P/G-CMC but not Fas-mediated cytotoxicity,1,2 its precise signal transduction was not elucidated so far. In this study, we focused on the main effector mechanism of P/G-CMC, ie, GB6 and caspases,5 and clarified their roles in xenogeneic ADCC. Our results that
DNA fragmentation was greatly reduced but cell lysis was partially reduced by GB inhibitor indicate that human GB plays a similar critical role in inducing xenogeneic target cell death to that in allogeneic cell death.4,7 In addition, incomplete inhibition of DNA damage is consistent with the previous report that CTL from GB-deficient mice were unable to induce rapid DNA fragmentation on allogeneic target cells, but this defect was rescued by long time incubation.11 Probably, the residual nucleolytic activities would be caused by other granzymes, and the retained cytolytic activities would be due to perforin itself or perforin plus other granzyme activities. To further analyze the signal transduction from GB, caspase inhibition was performed. As compared with the partial nucleolysis inhibition by peptide caspase inhibitors, no effect on cell lysis was observed by these inhibitors. The results on cell lysis are consistent, but those on nucleolysis are incompatible with the recent findings that broad spectrum caspase inhibitor, ie, zVAD-fmk, completely inhibited the CTL granuleinduced targets nuclear damage, but not their lysis in
934
allogeneic cytotoxicity.7 Several possible explanations can be made: (1) GB have a direct nucleolytic pathway independent from caspases.6 (2) GB can activate not only caspase-3 and -8, but also caspase-2, -6, -7, -9 and -10 to lead to nuclear damage. (3) Different signal transduction pathways would be operated in xenogeneic cytotoxicity. In any case, it is concluded that human GB is able to cause both apoptotic nuclear and cytolytic damage on xenogeneic pig target cells, and that the signals to nucleolysis are partially transduced through caspase-3 and -8. REFERENCES 1. Fujiwara I, Nakajima H, Yamagishi H, et al: Xenotransplantation 5:50, 1998 2. Fujiwara I, Nakajima H, Matsuda T, et al: Transplant Proc 30:2488, 1998
NAKAJIMA, FUJIWARA, MIZUTA ET AL 3. Smith DJ, McGuire MJ, Tocci MJ, et al: J Immunol 158:163, 1997 4. Nakajima H, Park HL, Henkart PA: J Exp Med 181:1037, 1995 5. Salvesen GS, Dixit VM: Cell 91:443, 1997 6. Talanian RV, Yang XH, Turbov J, et al: J Exp Med 186:1323, 1997 7. Sarin A, Williams MS, Alexander-Miller MA, et al: Immunity 6:209, 1997 8. Odake S, Kam CM, Narasimhan L, et al: Biochemistry 30:2217, 1991 9. Jacobson MD, Weil M, Raff MC: J Cell Biol 133:1041, 1996 10. Xiang J, Chao DT, Korsmeyer SJ, et al: Proc Natl Acad Sci USA 93:14559, 1996 11. Heusel JW, Wesselschmidt RL, Shresra S, et al: Cell 76:977, 1994
A
LTHOUGH there is increased evidence that natural killer (NK) cells play important roles in early phases of xenograft rejection via antibody-dependent cell-mediated cytotoxicity (ADCC), the molecular mechanism of how NK cells can give damage on xenogeneic target cells is still unknown. We have previously reported that perforin/granzymes-cell mediated cytotoxicity (P/G-CMC) plays a predominant role in human–anti-pig ADCC and induces apoptosis in target cells,1,2 however, its downstream biochemical events are not investigated. According to the observation using caspase inhibitors, caspase family is thought to be a common mediator of apoptotic cell death. Sequential activation of caspase 8 and caspase 3 during Fas-mediated apoptosis has been shown to be a critical event for apoptosis.3 Among many granzymes contained in lymphocyte granules, on the other hand, granzyme B (GB) has been shown to be the most critical protease to causes apoptotic cell death when perforin is present.4 GB has a substrate specificity and inhibitor profile similar to that of the caspase family, that also cleave substrates after aspartyl residues.5,6 In addition, GB has been shown in vitro to cleave and activate various members of the caspase subfamily: caspase-3 (CPP32), caspase-7 (Mch-3), caspase-10 (Mch-4), and caspase-8 (FLICE/Mch5), but not the ICE-related subfamily (caspase-1), leading directly to cleavage of nuclear target proteins such as caspase-activated-DNAase (CAD), poly adenosine diphosphate ([adp]-ribose)polymerase, and lamin. Those observation suggest that an interaction between GB and caspases plays a prominent role in the pathway of NK cell–mediated killing. Although contribution of these caspases in allogeneic cytotoxicity has been reported recently,7 much less is known about xenogeneic cytotoxicity. In this study, we used in vitro xenogeneic ADCC model and investigated the role of GB and caspases in xenogeneic cytotoxicity. MATERIALS AND METHODS Cytotoxicity Assays Human peripheral blood lymphocyte cells (PBL) were isolated by Ficoll-Paque gradient centrifugation and purified from adherent cells by passage over nylon wool column. These PBL in 5% heat-inactivated human serum were used as ADCC effector cells. PK15 (pig kidney cell line) were obtained from American Type Cell Culture Collection (Rockville, MD) and used as target cells. To detect membrane damage as cell lysis or DNA damage as a marker 0041-1345/00/$–see front matter PII S0041-1345(00)01044-7 932
for apoptosis in cytotoxicity assay, PK15 were labeled with 51Cr or 3 H-thymidine and tested for their release in 4-hour assay, as described previously.1 In both cytotoxicity assays, results are given as mean ⫾ SE of triplicate wells. All the results are representative of at least three independent experiments.
GB and Caspases Inhibition To inhibit GB activity, effector PBL were preincubated in 40 mol/L of zAAD-cmk(benzyloxycarbonyl-Ala-Ala-Asp-chlorometyl ketone)8 at 37°C for 30 minutes, and used in cytotoxicity assays as described above. In terms of caspases inhibition, zYVAD-fmk (z-Tyr-Val-Ala-Asp-fluoromethyl ketone, caspase-1 inhibitor), zDEVD-fmk (z-Asp-Glu-Val-Asp-fmk, caspase-3 inhibitor), zIETD fmk (z-Ile-Glu-Thr-Asp-fmk, caspase-8 inhibitor), and zFA-fmk (z-Phe-Ala-fmk, negative control) (Kamiya Biochemical Co., WA) were used. These compounds not only lead to irreversible enzyme inhibition but also are esterified to promote membrane permeability.9,10 Target PK15 were preincubated in 200 mol/L of each peptide-fmk at 37°C for 1 hour and the cytotoxicity assays were carried out in 20 mol/L of each of these inhibitors.
RESULTS
Although there is considerable evidence that perforin pore formation allows the access of granzymes to the target cells cytoplasm where these molecules induce DNA fragmentation by activating caspases in allogeneic cytotoxicity,4,7 less is known about xenogeneic cytotoxicity. Since we have previously shown that P/G-CMC is an only effector mechanism in this xenogeneic ADCC model,1,2 we have tried to inhibit GB by preloading zAAD-cmk to effector PBL. As shown in Figure 1, membrane damage was partially rescued, however, DNA fragmentation was greatly inhibited to the almost basal levels. These results indicate that GB is not only a key factor for inducing nuclear damage, but also contributes to lysis of the target cells in xenogeneic ADCC. Next, to examine which caspases were involved in NK-cell– induced xenogeneic ADCC, we designed the experiments in From the Second Department of Surgery (H.N., I.F., N.M., K.S.); Ayabe Municipal Hospital (I.F.); Department of Digestive Surgery (H.Y.), Kyoto Prefectural University of Medicine, Kyoto, Japan. Address reprint requests to Hiroo Nakajima, MD, PhD, Second Department of Surgery, Kyoto Prefectural University of Medicine, 465 Kawaramachi Hirokoji, Kamigyo, Kyoto 602-0841, Japan. © 2000 by Elsevier Science Inc. 655 Avenue of the Americas, New York, NY 10010 Transplantation Proceedings, 32, 932–934 (2000)
XENOGENEIC CYTOTOXICITY
933
Fig 1. Effect of granzyme B and caspases inhibition on xenogeneic ADCC. 51Cr- or 3H-thymidine-labeled PK15 were incubated with PBL for 4 hours in 5% heat-inactivated human serum at ET ratio ⫽ 50 and 51Cr (upper panels) and 3H-DNA (lower panels) release were compared. Effector PBL were preincubated with zAAD-cmk to inhibit granzyme B (A1, 2). Target PK15 were preincubated with zYVAD-fmk (B1, 4), zDEVD-fmk (B2, 5), and zIETD-fmk (B3, 6), and cytotoxicity assays were performed.
which target PK15 were loaded with tetrapeptide inhibitors known to inhibit caspases. zYVAD-fmk, a selective inhibitor of caspase-1, inhibited neither membrane nor DNA damage (Fig 1). In contrast, zDEVD-fmk and zIETD-fmk, a selective inhibitor of caspase-3 and caspase-8, respectively, partially inhibited DNA fragmentation (Fig 1), but not the membrane damage (Fig 1). These results were consistent with previous observation that GB did activate caspase-3 and -8, but did not caspase 1,6 and that caspases inhibition did not affect the lytic activity, whereas DNA damage is reduced by the inhibition of CPP32-related caspase subfamily.7 DISCUSSION
Concerning the effector mechanism of xenogeneic ADCC, although we have recently found that it can cause apoptotic cell death in PK15 via P/G-CMC but not Fas-mediated cytotoxicity,1,2 its precise signal transduction was not elucidated so far. In this study, we focused on the main effector mechanism of P/G-CMC, ie, GB6 and caspases,5 and clarified their roles in xenogeneic ADCC. Our results that
DNA fragmentation was greatly reduced but cell lysis was partially reduced by GB inhibitor indicate that human GB plays a similar critical role in inducing xenogeneic target cell death to that in allogeneic cell death.4,7 In addition, incomplete inhibition of DNA damage is consistent with the previous report that CTL from GB-deficient mice were unable to induce rapid DNA fragmentation on allogeneic target cells, but this defect was rescued by long time incubation.11 Probably, the residual nucleolytic activities would be caused by other granzymes, and the retained cytolytic activities would be due to perforin itself or perforin plus other granzyme activities. To further analyze the signal transduction from GB, caspase inhibition was performed. As compared with the partial nucleolysis inhibition by peptide caspase inhibitors, no effect on cell lysis was observed by these inhibitors. The results on cell lysis are consistent, but those on nucleolysis are incompatible with the recent findings that broad spectrum caspase inhibitor, ie, zVAD-fmk, completely inhibited the CTL granuleinduced targets nuclear damage, but not their lysis in
934
allogeneic cytotoxicity.7 Several possible explanations can be made: (1) GB have a direct nucleolytic pathway independent from caspases.6 (2) GB can activate not only caspase-3 and -8, but also caspase-2, -6, -7, -9 and -10 to lead to nuclear damage. (3) Different signal transduction pathways would be operated in xenogeneic cytotoxicity. In any case, it is concluded that human GB is able to cause both apoptotic nuclear and cytolytic damage on xenogeneic pig target cells, and that the signals to nucleolysis are partially transduced through caspase-3 and -8. REFERENCES 1. Fujiwara I, Nakajima H, Yamagishi H, et al: Xenotransplantation 5:50, 1998 2. Fujiwara I, Nakajima H, Matsuda T, et al: Transplant Proc 30:2488, 1998
NAKAJIMA, FUJIWARA, MIZUTA ET AL 3. Smith DJ, McGuire MJ, Tocci MJ, et al: J Immunol 158:163, 1997 4. Nakajima H, Park HL, Henkart PA: J Exp Med 181:1037, 1995 5. Salvesen GS, Dixit VM: Cell 91:443, 1997 6. Talanian RV, Yang XH, Turbov J, et al: J Exp Med 186:1323, 1997 7. Sarin A, Williams MS, Alexander-Miller MA, et al: Immunity 6:209, 1997 8. Odake S, Kam CM, Narasimhan L, et al: Biochemistry 30:2217, 1991 9. Jacobson MD, Weil M, Raff MC: J Cell Biol 133:1041, 1996 10. Xiang J, Chao DT, Korsmeyer SJ, et al: Proc Natl Acad Sci USA 93:14559, 1996 11. Heusel JW, Wesselschmidt RL, Shresra S, et al: Cell 76:977, 1994