- Email: [email protected]
granzyme B expression
BBRC Biochemical and Biophysical Research Communications 337 (2005) 1319–1323 www.elsevier.com/locate/ybbrc
Hyperthermia suppresses the cytotoxicity of NK cells via down-regulation of perforin/granzyme B expression Tomoaki Koga a, Hideki Harada b, Tea Seow Shi a, Seiji Okada b, Mary Ann Suico a, Tsuyoshi Shuto a, Hirofumi Kai a,* a
Department of Molecular Medicine, Faculty of Medical and Pharmaceutical Sciences, Kumamoto University, 5-1 Oe-honmachi, Kumamoto 862-0973, Japan b Division of Hematopoiesis, Center for AIDS Research, Kumamoto University, 2-2-1 Honjo, Kumamoto 860-0811, Japan Received 26 September 2005 Available online 10 October 2005
Abstract Hyperthermia, which is used as an adjunctive therapy for cancer, is known to modulate the activity of natural killer (NK) cells in vitro, but its effect in vivo is unclear. In the present study, we used a whole body hyperthermia (WBH) device heated by infrared rays to evaluate the effect of WBH on mice models. We demonstrate here that wild type C57BL/6J mice exposed to 42 C for 60 min had reduced NK cell cytolytic activity against YAC-1 target cells as determined by cytolytic assay. This result was confirmed using Rag-2 knockout mice, which possess functional NK but not cytolytic T or NK-T cells. Moreover, WBH decreased the mRNA expression of perforin and granzyme B in spleens of mice. But the expression of TNF cytokines (Fas ligand and TRAIL) was unchanged. These data suggest that the suppression of NK cell activity induced by WBH could be mediated through the perforin/granzyme pathway. 2005 Elsevier Inc. All rights reserved. Keywords: Whole body hyperthermia; NK cells; Perforin; Granzyme B; FasL/TRAIL; Infrared box; Rag-2 knockout mouse
Hyperthermia has been applied as an adjunctive therapy with various established cancer treatments such as radiotherapy and chemotherapy, and favorable results have been obtained from several clinical phase III studies [1,2]. However, the biological mechanisms underlying the clinical effects are still poorly understood. It is hypothesized that hyperthermia could have both direct effects on tumor cells and indirect effects on lymphocytes, such as natural killer (NK) cells, which may be involved in the control of tumor growth [3,4]. NK cells are highly sensitive to heat [5–10] and it is well known that they have a crucial role in innate immunity as they are capable of effectively responding to virus-infected and tumor cells [11–13]. Several in vivo and in vitro studies have confirmed the importance of NK cell-mediated cytotoxicity in tumor elimination [14–18]. NK cells exhibit their cytotoxic activity via two major pathways: one way is by the exocytosis of cytolytic granule *
Corresponding author. Fax: +81 96 371 4405. E-mail address: [email protected] (H. Kai).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.09.184
complex perforin/granzyme, which, when released, binds to target cell and cause cell lysis; another is through the deathreceptor-mediated apoptosis by exhibiting various ligands including FAS ligand (FasL) and TNF-related apoptosis inducing ligand (TRAIL) [12,19]. However, the molecular mechanisms of hyperthermia on NK cells are still under extensive investigations. Data obtained from various in vivo and in vitro studies of hyperthermia effect on NK cell-mediated cell lysis are rather controversial and inconsistent [20–23]. Effect of hyperthermia on different targets in the cell and the diverse possible regulations of various genes suggests that a better understanding of this type of temperature-dependent mechanism will clearly open up new clinical implications for cancer therapy. In the present study, we analyzed the in vivo effect of whole body hyperthermia (WBH) on NK cell cytotoxicity by exposing to 42 C heat shock for 1 h wild type mice and Rag-2 knockout (KO) mice, which have functional NK cells but do not have cytolytic T cells and NK T cells.
1320
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
In vivo hyperthermia was carried out using a relatively novel WBH device heated using infrared rays (we termed, IR box) and which mimics WBH in clinical setting. Here we show that WBH reduced the cytolytic activity of mouse NK cells. Moreover, the expressions of perforin and granzyme B, but not that of FasL and TRAIL, were down-regulated, suggesting that the hyperthermic inhibition of NK cell cytotoxicity may be mediated through the perforin/ granzyme B pathway. Materials and methods Cells. YAC-1, target cell of murine natural killer cells, was obtained from RIKEN Cell Bank (Tsukuba, Japan). The cells were grown in RPMI-1640 medium supplemented with 10% FBS, 50 IU/ml penicillin, 50 lg/ml streptomycin, and 2 mM L-glutamine, and maintained at 37 C in 5% CO2. Animals. Six to ten-week-old C57BL/6J wild type (WT) mice were obtained from Charles River (Charles River Japan). Rag-2 KO (C57BL/6J strain) mice were provided by Dr. T. Taniguchi (University of Tokyo, Japan). Mice used in this study were housed individually in specific pathogen-free condition at the Center for Animal Resources and Development, Kumamoto University. The animals were fed with sterilized water and chow ad libitum. All experiments were performed according to the protocols approved by the Animal Welfare Committee of Kumamoto University. Antibodies. Anti-Hsp72 polyclonal antibody and anti-Hsc70 antibody were obtained from Stressgen Biotechnologies (Victoria, BC, Canada), anti-perforin polyclonal antibody was from Santa Cruz Biotechnology (Santa Cruz, CA), and horseradish peroxidase (HRP)-conjugated Affinipure goat anti-rabbit IgG and goat anti-rat IgG antibodies were from Jackson Immuno Research (West Grove, PA). PE-conjugated anti-NK1.1 and FITC-conjugated anti-CD3 monoclonal antibodies were purchased from BD Pharmingen (San Diego, CA). Whole body hyperthermia. To perform WBH, the mice were placed in a well-ventilated, thermally controlled box, generating homogeneous heat from infrared rays (IR box, generously provided by Blast, Tokyo, Japan). Mice were anesthetized with pentobarbital (80 mg/kg) before exposing to hyperthermia for 60 min. Rectal temperatures of mice were monitored every 2 min and were maintained between 40.5 and 42.5 C. The control mice were anesthetized with half the dosage of pentobarbital used in sample mice (40 mg/kg). The control group was maintained at a normothermic condition of ambient room temperature (24 C) for 60 min. Each group of animals was allowed to recover at room temperature for 5 h before being sacrificed to obtain the organs. Western blotting. Five hours after WBH, proteins from whole tissues were extracted by sonication in ice-cold glycerol buffer [10% glycerol, 5 mM EDTA, 10 mM Tris/HCl (pH 7.4), 200 mM NaCl, 1/100 (v/v) protease inhibitor cocktail (Sigma, St Louis, MO), and 1 mM PMSF]. The protein concentration was measured using bicinchoninic acid assay (BioRad). Fifty micrograms of samples was electrophoresed through 7.5% SDS–PAGE gels before transferring to polyvinylidene difluoride membrane (Millipore, MA). After blocking the membrane with 5% (w/v) skim milk in 0.1% PBS-Tween for 1 h at room temperature, membranes were
reacted with 1/1000 dilution of anti-Hsp72, anti-Hsc70 or 1/200 dilution of anti-perforin antibodies for 1 h at room temperature. The secondary antibody used for Hsp72 and perforin was HRP-conjugated goat antirabbit IgG. Anti-rat antibody was used for Hsc70. Visualization of the immunoblot was performed by SuperSignal West Pico chemiluminescent substrate (PIERCE, Rockford, IL). Cytotoxicity assay. The cytotoxic activity of the NK cells was measured by flow cytometry-based cytotoxicity assay with CFSE-labeled Yac-1 cells and propidium iodide (PI), as described by Marcusson-Stahl et al. [24]. Briefly, target Yac-1 cells were stained with 2 lM of fluorescence labeling reagent, 5-(6)-carboxy-fluorescein succinimidyl ester (CFSE, Molecular Probes, OR), as described in manufacturerÕs instruction. Fifty microliters aliquots of CFSE-labeled Yac-1 cells (1 · 105 cell/ml) were placed in 5-ml round-bottomed tubes, and 400 ll of twofold serial diluted mononuclear cells from spleen was added into the tubes as effector cells. Effector or target cells alone were used as negative controls. After 4 hincubation at 37 C, 50 ll of 20 lg/ml PI was added and cells were incubated for an additional 15 min. Target cells in FSC (forward scatter) vs. SSC (side scatter) dot plots were gated, and CFSE and PI were measured by LSR II flow cytometer (Becton Dickinson, San Jose, CA). Cytotoxic activity was calculated as follows: A/(A + B) · 100 · C (%), where A is the percentage of PI+EGFP+ cells; B is the percentage of PIGFP+ cells at each E/T ratio; C is the percentage of spontaneous PI+ cells without effector cells (A/(A + B) · 100 (%) at E/T ratio = 0). Flow cytometry analysis. The spleens were cut into small fragments and suspended in lysis buffer (0.155 M ammonium chloride, 0.1 M disodium EDTA, and 0.01 M potassium bicarbonate). Single cell suspensions were prepared in staining medium (PBS with 3% FCS and 0.1% sodium azide) and were stained with the antibodies described above. After 20 min of incubation on ice, cells were washed two times and resuspended in staining medium supplemented with 1 lg/mL PI. Stained cells were analyzed on a LSR II and analyzed with FlowJo software (Tree Star, San Carlos, CA). Semi-quantitative reverse transcription-polymerase chain reaction (RTPCR). Total RNA was extracted from Rag-2 KO mice spleen tissue using Isogen reagent (Nippongene, Japan). RT-PCR was performed with a Takara RNA PCR kit ver.3 (Takara, Japan) according to the manufacturerÕs instructions. RT reaction was done at 42 C for 30 min, 99 C for 5 min, and 5 C for 5 min. PCR conditions were: 1 cycle of initial denaturing at 98 C for 2 min; followed by denaturing at 98 C for 30 s; annealing for 30 s at 67, 62, 60, and 57 C for FasL, perforin, granzyme B, and TRAIL primers, respectively; extension at 72 C (30 s) and final extension for 1 cycle at 72 C (5 min). The primers used for amplications are displayed in Table 1.
Results and discussion Whole body hyperthermia was effectively carried out using the IR box Previous investigations suggested that hyperthermia suppressed the cytotoxic activity of NK cells in vitro [9,20]. Due to the scarcity of in vivo report, we attempted to investigate the effect of WBH on NK cells using mouse model. WBH treatment was carried out in a well-ventilated,
Table 1 List or primers Primers
Upstream
Downstream
Hsp72 Granzyme B Perforin TRAIL FasL b-Actin
5 0 -GCTGACCAAGATGAAGGAGATC-3 0 5 0 -AGATCATCGGGGGACATG-3 0 5 0 -GAACCCTAGGCCAGAGGCAAAC-3 0 5 0 -TCACCAACGAGATGAAGCAGC-3 0 5 0 -CAGCTCTTCCACCTGCAGAAGG-3 0 5 0 -TAAAACGCAGCTCAGTAACAGTCGG-3 0
5 0 -GAGTCGATCTCCAGGCTGGC-3 0 5 0 -TTACACACAAGAGGGCCTCCAG-3 0 5-CCTGGTTGGTGACCTTTGAATCC-3 0 5 0 -CTCACCTTGTCCTTTGAGACC-3 0 5 0 -AGATTCCTCAAAATTGATCAGAGAGAG-3 0 5 0 -TGCAATCCTGTGGCATCCATGAAAC-3 0
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
thermally insulated box heated using infrared rays (IR box), which was designed to mimic the clinical hyperthermia device. We first determined whether the IR box could effectively sustain an increased body temperature of C57BL/6J wild type mice. Fig. 1A shows that mouse body temperature was increased within 10 min after starting the procedure and maintained within the duration of WBH treatment. The elevated body temperature correlated well with the induction of Hsp72, an inducible heat shock protein that has protective functions during heat stress [25]. The increased expression of Hsp72 was detected by immunoblotting of protein lysates from various tissues of WBHtreated mice (Fig. 1B). The expression of the housekeeping gene Hsc70, used as a control, was not changed by WBH treatment. These data indicated that the IR box could work effectively as a WBH device for mice. Whole body hyperthermia suppressed the cytolytic activity of NK cells To evaluate the effect of WBH on the activity of NK cells in vivo, we analyzed the cytolytic activity of NK cells
Fig. 1. Whole body hyperthermia was effectively carried out using the IR box. (A) The rectal temperature of C57BL/6J mice was monitored every 2 min throughout the 1 h WBH treatment in IR box and was maintained between 40.5 and 42.5 C. (B) Western blotting was performed with 50 lg protein lysates from various tissues of heat-shocked or non-heat-shocked C57BL/6J mice. After blotting, membranes were reacted with anti-Hsp72 or Hsc70 antibodies.
1321
obtained from the spleen of WBH-treated or -untreated C57BL/6J mice against YAC-1 cells, a prototypical NK cell-sensitive target. Cytolytic activity of NK cells in these mice was significantly decreased after hyperthermia (Fig. 2A). To confirm this observation, we repeated the cytotoxicity assay using Rag-2 KO mice. We obtained similar results using this mouse model (Fig. 2A). Since Rag-2 KO mice do not contain cytolytic T lymphocytes and NK T cells (Fig. 2B), the only possible immune component that is functional in these mice is the NK cell population. Consistently, these mice displayed higher cytolytic activity compared with wild type mice (Fig. 2A) due to its higher percentage population of NK cells (Fig. 2B). Taken together, these data suggest that WBH (42 C, 60 min) decreased the cytolytic activity of NK cells in vivo. Our results are consistent with some in vitro studies which showed a reduction of lymphocyte cytotoxic activity at 42 C [9,23,26]. The down-regulation of NK cell activity in our study could not be attributed to the decrease in cell viability since there were no significant PI+ splenocytes after WBH treatment (unpublished data). This is in agreement with previous
Fig. 2. Whole body hyperthermia suppressed the cytolytic activity of NK cells. (A) Cytotoxic activity of splenic NK cells of C57BL/6J wild type or Rag-2 KO mice, treated or untreated with WBH, was tested against YAC1 target cells. The results are expressed as percent specific lysis at different effector (E)/target (T) cell ratios. Data shown are representative of three independent experiments. (B) Flow cytometry analyses were performed with peripheral blood mononuclear cells (PBMCs) from peripheral blood of C57BL/6J mice and Rag-2 KO mice. PBMCs were stained with antiCD3 and NK1.1 antibodies.
1322
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
findings that the decrease of NK activity effected by hyperthermia is not related to cell death [9,27]. It was previously reported, however, that WBH performed at lower temperatures (39–40 C) and longer duration resulted in the enhancement of the activity of NK cells [22]. This inconsistency could be explained by the diversity of experimental parameters deemed optimal for clinical advantage. But this seemingly contradictory report underscores the complexity and sensitivity of NK cellsÕ response to its thermal environment. Whole body hyperthermia down-regulates the expression of cytotoxic granules in Rag-2 KO mice spleen NK cells are known to mediate cell death via several possible pathways [12]. To clarify the mechanisms underlying the suppression of NK cell cytolytic activity after hyperthermia, we screened several groups of molecules including the membrane-anchored cytotoxic molecules that bind to the death receptor of target cells (TRAIL, FasL) and secretory cytotoxic granules (perforin, granzyme). The mRNA expression levels were analyzed using semiquantitative RT-PCR. Interestingly, the expression of perforin and granzyme B in Rag-2 KO mice spleen tissues was decreased after WBH (Fig. 3A), but there was no detectable change observed in the mRNA level of mouse FasL and TRAIL after hyperthermia (Fig. 3B). The expression of Hsp72 was induced after WBH, indicating that heat shock was properly carried out (Fig. 3C). We quantified the mRNA expressions of perforin and granzyme B (Fig. 3D) as well as FasL and TRAIL (Fig. 3E) based on the band intensities of the transcripts using Image Gauge software. We observed a significant inhibition of the perforin and granzyme B expression in the spleens of Rag-2 KO mice after WBH (Fig. 3D). It is pertinent to note here that the average decrease of splenic perforin and granzyme in WBH-treated Rag-2 KO mice was 35% from the nontreated controls. This is highly consistent with the result of the cytolytic assay of NK cell activity of Rag-2 KO mice, which showed 35–40% decrease after WBH (Fig. 2A). The WBH experiment was reiterated using ICR nude mice (Crj:CD-1 ICR-nu/nu SPF/VAF; KBT Oriental, Saga, Japan) and in vitro heat shock was performed using human NK cells obtained from healthy volunteers. In both systems, the perforin and granzyme mRNA expressions were suppressed after hyperthermia (unpublished data). Previous studies of perforin-deficient mice established the indispensable role of perforin in NK cell-mediated tumor cell killing [29–32]. Thus, we checked the protein expression of perforin by immunoblotting. Consistent with the RT-PCR data, the protein level of perforin was decreased in mice subjected to WBH (Fig. 3F). Collectively, these results suggest that WBH inhibited NK cell cytolytic activity possibly through the down-regulation of perforin and granzyme B, not through the down-regulation of death receptor ligands. Granzymes are necessary for triggering apoptosis in target cells, but they depend on being appro-
Fig. 3. Whole body hyperthermia down-regulated the expression of perforin and granzyme B in NK cells. (A–C) Five hours after whole body hyperthermia (42 C, 60 min), total RNA was extracted from spleen tissue of Rag-2 KO mice untreated or treated with WBH. Semi-quantitative RT-PCR was performed using 0.5 lg RNA for each gene indicated. (D) Quantitative analyses of perforin (36 cycles) and granzyme B (40 cycles) gene expressions were performed using Image Gauge software (Ver. 3.45, Fuji Film, Japan). Relative mRNA expression levels were normalized to the level of b-actin (20 cycles) that served as an internal control for the amount of RNA used in each reaction. Values are means ± SE from three independent experiments. *p < 0.005 as compared to their respective controls computed using StudentÕs t test. (E) Quantitative analyses of FasL and TRAIL at 36 cycles were performed using Image Gauge. Relative mRNA expression levels were normalized to the level of the internal control b-actin at 20 cycles. (F) Western blotting was performed using protein lysates extracted from whole spleen of Rag-2 KO mice untreated or treated with WBH. After blotting, membranes were reacted using antiperforin and Hsp72 antibodies. Ponceau S staining was performed on membranes to determine equal loading of protein samples.
priately delivered by perforin [19]. Since perforin and granzymes function cooperatively [28], the attenuation of one or the other would impair cytolysis through the granulemediated pathway. Our present results reinforce the importance of perforin/ granzyme-mediated exocytosis as the main pathway used by NK cells in tumor surveillance and elimination [29– 32]. The actual mechanism of how hyperthermia decreases the expression of perforin and granzyme B remains to be clarified but these current findings may lead us to reconsider the current modalities of hyperthermia for cancer treatment. Further understanding of the upstream pathways
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
that inhibit perforin/granzyme may be able to rescue the hyperthermic inactivation of NK cells and could help elucidate the molecular pathway of NK cells after WBH for better therapeutic development. Acknowledgments We thank Dr. T. Taniguchi (University of Tokyo, Japan) for the Rag-2 KO mice (C57BL/6J strain). This work was supported by grants from the Ministry of Education, Science, Sports and Culture of Japan and the Takeda Science Foundation. References [1] B. Hildebrandt, P. Wust, O. Ahlers, A. Dieing, G. Sreenivasa, T. Kerner, R. Felix, H. Riess, The cellular and molecular basis of hyperthermia, Crit. Rev. Oncol. Hematol. 43 (2002) 33–56. [2] P. Wust, B. Hildebrandt, G. Sreenivasa, B. Rau, J. Gellermann, H. Riess, R. Felix, P.M. Schlag, Hyperthermia in combined treatment of cancer, Lancet Oncol. 3 (2002) 487–497. [3] T.L. Whiteside, N.L. Vujanovic, R.B. Herberman, Natural killer cells and tumor therapy, Curr. Top. Microbiol. Immunol. 230 (1998) 221– 244. [4] C. Christophi, A. Winkworth, V. Muralihdaran, P. Evans, The treatment of malignancy by hyperthermia, Surg. Oncol. 7 (1998) 83– 90. [5] K.S. Zanker, J. Lange, Whole body hyperthermia and natural killer cell activity, Lancet 1 (1982) 1079–1080. [6] M. Onsrud, Effect of hyperthermia on human natural killer cells, Recent Results Cancer Res. 109 (1988) 50–56. [7] B. Kubista, K. Trieb, H. Blahovec, R. Kotz, M. Micksche, Hyperthermia increases the susceptibility of chondro- and osteosarcoma cells to natural killer cell-mediated lysis, Anticancer Res. 22 (2002) 789–792. [8] J.F. Downing, M.W. Taylor, The effect of in vivo hyperthermia on selected lymphokines in man, Lymphokine Res. 6 (1987) 103–109. [9] M.P. Fuggetta, E. Alvino, M. Tricarico, S. DÕAtri, R. Pepponi, S.P. Prete, E. Bonmassar, In vitro effect of hyperthermia on natural cellmediated cytotoxicity, Anticancer Res. 20 (2000) 1667–1672. [10] R.N. Shen, N.B. Hornback, H. Shidnia, L. Lu, H.E. Broxmeyer, Z. Brahmi, Effect of whole-body hyperthermia and cyclophosphamide on natural killer cell activity in murine erythroleukemia, Cancer Res. 48 (1988) 4561–4563. [11] M.J. Smyth, Y. Hayakawa, K. Takeda, H. Yagita, New aspects of natural-killer-cell surveillance and therapy of cancer, Nat. Rev. Cancer 2 (2002) 850–861. [12] M.J. Smyth, E. Cretney, J.M. Kelly, J.A. Westwood, S.E. Street, H. Yagita, K. Takeda, S.L. van Dommelen, M.A. Degli-Esposti, Y. Hayakawa, Activation of NK cell cytotoxicity, Mol. Immunol. 42 (2005) 501–510. [13] W.M. Yokoyama, S. Kim, A.R. French, The dynamic life of natural killer cells, Annu. Rev. Immunol. 22 (2004) 405–429. [14] K. Karre, H.G. Ljunggren, G. Piontek, R. Kiessling, Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. 1986, J. Immunol. 174 (2005) 6566–6569.
1323
[15] T. Barlozzari, C.W. Reynolds, R.B. Herberman, In vivo role of natural killer cells: involvement of large granular lymphocytes in the clearance of tumor cells in anti-asialo GM1-treated rats, J. Immunol. 131 (1983) 1024–1027. [16] E. Gorelik, B. Rosen, D. Copeland, B. Weatherly, R.B. Herberman, Evaluation of role of natural killer cells in radiation-induced leukemogenesis in mice, J. Natl. Cancer Inst. 72 (1984) 1397–1403. [17] C. Riccardi, A. Santoni, T. Barlozzari, P. Puccetti, R.B. Herberman, In vivo natural reactivity of mice against tumor cells, Int. J. Cancer 25 (1980) 475–486. [18] H.G. Ljunggren, K. Karre, Host resistance directed selectively against H-2-deficient lymphoma variants. Analysis of the mechanism, J. Exp. Med. 162 (1985) 1745–1759. [19] J.A. Trapani, M.J. Smyth, Functional significance of the perforin/ granzyme cell death pathway, Nat. Rev. Immunol. 2 (2002) 735– 747. [20] T. Kalland, I. Dahlquist, Effects of in vitro hyperthermia on human natural killer cells, Cancer Res. 43 (1983) 1842–1846. [21] H. Yang, W. Lauzon, I. Lemaire, Effects of hyperthermia on natural killer cells: inhibition of lytic function and microtubule organization, Int. J. Hyperthermia 8 (1992) 87–97. [22] R. Burd, T.S. Dziedzic, Y. Xu, M.A. Caligiuri, J.R. Subjeck, E.A. Repasky, Tumor cell apoptosis, lymphocyte recruitment and tumor vascular changes are induced by low temperature, long duration (fever-like) whole body hyperthermia, J. Cell Physiol. 177 (1998) 137– 147. [23] R.N. Shen, L. Lu, P. Young, H. Shidnia, N.B. Hornback, H.E. Broxmeyer, Influence of elevated temperature on natural killer cell activity, lymphokine-activated killer cell activity and lectin-dependent cytotoxicity of human umbilical cord blood and adult blood cells, Int. J. Radiat. Oncol. Biol. Phys. 29 (1994) 821–826. [24] M. Marcusson-Stahl, K. Cederbrant, A flow-cytometric NK-cytotoxicity assay adapted for use in rat repeated dose toxicity studies, Toxicology 193 (2003) 269–279. [25] D.D. Mosser, A.W. Caron, L. Bourget, C. Denis-Larose, B. Massie, Role of the human heat shock protein hsp70 in protection against stress-induced apoptosis, Mol. Cell Biol. 17 (1997) 5317–5327. [26] A. Izumi, S. Koga, M. Maeta, Effects of in vitro hyperthermia on murine and human lymphocytes, Cancer 51 (1983) 2061–2065. [27] A. Yoshioka, Y. Miyachi, S. Imamura, Immunological effects of in vitro hyperthermia, J. Clin. Lab. Immunol. 29 (1989) 95–97. [28] S.S. Metkar, B. Wang, M. Aguilar-Santelises, S.M. Raja, L. UhlinHansen, E. Podack, J.A. Trapani, C.J. Froelich, Cytotoxic cell granule-mediated apoptosis: perforin delivers granzyme B-serglycin complexes into target cells without plasma membrane pore formation, Immunity 16 (2002) 417–428. [29] D. Kagi, B. Ledermann, K. Burki, P. Seiler, B. Odermatt, K.J. Olsen, E.R. Podack, R.M. Zinkernagel, H. Hengartner, Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforindeficient mice, Nature 369 (1994) 31–37. [30] M.E. van den Broek, D. Kagi, F. Ossendorp, R. Toes, S. Vamvakas, W.K. Lutz, C.J. Melief, R.M. Zinkernagel, H. Hengartner, Decreased tumor surveillance in perforin-deficient mice, J. Exp. Med. 184 (1996) 1781–1790. [31] M.F. van den Broek, D. Kagi, R.M. Zinkernagel, H. Hengartner, Perforin dependence of natural killer cell-mediated tumor control in vivo, Eur. J. Immunol. 25 (1995) 3514–3516. [32] M.F. van den Broek, H. Hengartner, The role of perforin in infections and tumour surveillance, Exp. Physiol. 85 (2000) 681–685.
Hyperthermia suppresses the cytotoxicity of NK cells via down-regulation of perforin/granzyme B expression Tomoaki Koga a, Hideki Harada b, Tea Seow Shi a, Seiji Okada b, Mary Ann Suico a, Tsuyoshi Shuto a, Hirofumi Kai a,* a
Department of Molecular Medicine, Faculty of Medical and Pharmaceutical Sciences, Kumamoto University, 5-1 Oe-honmachi, Kumamoto 862-0973, Japan b Division of Hematopoiesis, Center for AIDS Research, Kumamoto University, 2-2-1 Honjo, Kumamoto 860-0811, Japan Received 26 September 2005 Available online 10 October 2005
Abstract Hyperthermia, which is used as an adjunctive therapy for cancer, is known to modulate the activity of natural killer (NK) cells in vitro, but its effect in vivo is unclear. In the present study, we used a whole body hyperthermia (WBH) device heated by infrared rays to evaluate the effect of WBH on mice models. We demonstrate here that wild type C57BL/6J mice exposed to 42 C for 60 min had reduced NK cell cytolytic activity against YAC-1 target cells as determined by cytolytic assay. This result was confirmed using Rag-2 knockout mice, which possess functional NK but not cytolytic T or NK-T cells. Moreover, WBH decreased the mRNA expression of perforin and granzyme B in spleens of mice. But the expression of TNF cytokines (Fas ligand and TRAIL) was unchanged. These data suggest that the suppression of NK cell activity induced by WBH could be mediated through the perforin/granzyme pathway. 2005 Elsevier Inc. All rights reserved. Keywords: Whole body hyperthermia; NK cells; Perforin; Granzyme B; FasL/TRAIL; Infrared box; Rag-2 knockout mouse
Hyperthermia has been applied as an adjunctive therapy with various established cancer treatments such as radiotherapy and chemotherapy, and favorable results have been obtained from several clinical phase III studies [1,2]. However, the biological mechanisms underlying the clinical effects are still poorly understood. It is hypothesized that hyperthermia could have both direct effects on tumor cells and indirect effects on lymphocytes, such as natural killer (NK) cells, which may be involved in the control of tumor growth [3,4]. NK cells are highly sensitive to heat [5–10] and it is well known that they have a crucial role in innate immunity as they are capable of effectively responding to virus-infected and tumor cells [11–13]. Several in vivo and in vitro studies have confirmed the importance of NK cell-mediated cytotoxicity in tumor elimination [14–18]. NK cells exhibit their cytotoxic activity via two major pathways: one way is by the exocytosis of cytolytic granule *
Corresponding author. Fax: +81 96 371 4405. E-mail address: [email protected] (H. Kai).
0006-291X/$ - see front matter 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2005.09.184
complex perforin/granzyme, which, when released, binds to target cell and cause cell lysis; another is through the deathreceptor-mediated apoptosis by exhibiting various ligands including FAS ligand (FasL) and TNF-related apoptosis inducing ligand (TRAIL) [12,19]. However, the molecular mechanisms of hyperthermia on NK cells are still under extensive investigations. Data obtained from various in vivo and in vitro studies of hyperthermia effect on NK cell-mediated cell lysis are rather controversial and inconsistent [20–23]. Effect of hyperthermia on different targets in the cell and the diverse possible regulations of various genes suggests that a better understanding of this type of temperature-dependent mechanism will clearly open up new clinical implications for cancer therapy. In the present study, we analyzed the in vivo effect of whole body hyperthermia (WBH) on NK cell cytotoxicity by exposing to 42 C heat shock for 1 h wild type mice and Rag-2 knockout (KO) mice, which have functional NK cells but do not have cytolytic T cells and NK T cells.
1320
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
In vivo hyperthermia was carried out using a relatively novel WBH device heated using infrared rays (we termed, IR box) and which mimics WBH in clinical setting. Here we show that WBH reduced the cytolytic activity of mouse NK cells. Moreover, the expressions of perforin and granzyme B, but not that of FasL and TRAIL, were down-regulated, suggesting that the hyperthermic inhibition of NK cell cytotoxicity may be mediated through the perforin/ granzyme B pathway. Materials and methods Cells. YAC-1, target cell of murine natural killer cells, was obtained from RIKEN Cell Bank (Tsukuba, Japan). The cells were grown in RPMI-1640 medium supplemented with 10% FBS, 50 IU/ml penicillin, 50 lg/ml streptomycin, and 2 mM L-glutamine, and maintained at 37 C in 5% CO2. Animals. Six to ten-week-old C57BL/6J wild type (WT) mice were obtained from Charles River (Charles River Japan). Rag-2 KO (C57BL/6J strain) mice were provided by Dr. T. Taniguchi (University of Tokyo, Japan). Mice used in this study were housed individually in specific pathogen-free condition at the Center for Animal Resources and Development, Kumamoto University. The animals were fed with sterilized water and chow ad libitum. All experiments were performed according to the protocols approved by the Animal Welfare Committee of Kumamoto University. Antibodies. Anti-Hsp72 polyclonal antibody and anti-Hsc70 antibody were obtained from Stressgen Biotechnologies (Victoria, BC, Canada), anti-perforin polyclonal antibody was from Santa Cruz Biotechnology (Santa Cruz, CA), and horseradish peroxidase (HRP)-conjugated Affinipure goat anti-rabbit IgG and goat anti-rat IgG antibodies were from Jackson Immuno Research (West Grove, PA). PE-conjugated anti-NK1.1 and FITC-conjugated anti-CD3 monoclonal antibodies were purchased from BD Pharmingen (San Diego, CA). Whole body hyperthermia. To perform WBH, the mice were placed in a well-ventilated, thermally controlled box, generating homogeneous heat from infrared rays (IR box, generously provided by Blast, Tokyo, Japan). Mice were anesthetized with pentobarbital (80 mg/kg) before exposing to hyperthermia for 60 min. Rectal temperatures of mice were monitored every 2 min and were maintained between 40.5 and 42.5 C. The control mice were anesthetized with half the dosage of pentobarbital used in sample mice (40 mg/kg). The control group was maintained at a normothermic condition of ambient room temperature (24 C) for 60 min. Each group of animals was allowed to recover at room temperature for 5 h before being sacrificed to obtain the organs. Western blotting. Five hours after WBH, proteins from whole tissues were extracted by sonication in ice-cold glycerol buffer [10% glycerol, 5 mM EDTA, 10 mM Tris/HCl (pH 7.4), 200 mM NaCl, 1/100 (v/v) protease inhibitor cocktail (Sigma, St Louis, MO), and 1 mM PMSF]. The protein concentration was measured using bicinchoninic acid assay (BioRad). Fifty micrograms of samples was electrophoresed through 7.5% SDS–PAGE gels before transferring to polyvinylidene difluoride membrane (Millipore, MA). After blocking the membrane with 5% (w/v) skim milk in 0.1% PBS-Tween for 1 h at room temperature, membranes were
reacted with 1/1000 dilution of anti-Hsp72, anti-Hsc70 or 1/200 dilution of anti-perforin antibodies for 1 h at room temperature. The secondary antibody used for Hsp72 and perforin was HRP-conjugated goat antirabbit IgG. Anti-rat antibody was used for Hsc70. Visualization of the immunoblot was performed by SuperSignal West Pico chemiluminescent substrate (PIERCE, Rockford, IL). Cytotoxicity assay. The cytotoxic activity of the NK cells was measured by flow cytometry-based cytotoxicity assay with CFSE-labeled Yac-1 cells and propidium iodide (PI), as described by Marcusson-Stahl et al. [24]. Briefly, target Yac-1 cells were stained with 2 lM of fluorescence labeling reagent, 5-(6)-carboxy-fluorescein succinimidyl ester (CFSE, Molecular Probes, OR), as described in manufacturerÕs instruction. Fifty microliters aliquots of CFSE-labeled Yac-1 cells (1 · 105 cell/ml) were placed in 5-ml round-bottomed tubes, and 400 ll of twofold serial diluted mononuclear cells from spleen was added into the tubes as effector cells. Effector or target cells alone were used as negative controls. After 4 hincubation at 37 C, 50 ll of 20 lg/ml PI was added and cells were incubated for an additional 15 min. Target cells in FSC (forward scatter) vs. SSC (side scatter) dot plots were gated, and CFSE and PI were measured by LSR II flow cytometer (Becton Dickinson, San Jose, CA). Cytotoxic activity was calculated as follows: A/(A + B) · 100 · C (%), where A is the percentage of PI+EGFP+ cells; B is the percentage of PIGFP+ cells at each E/T ratio; C is the percentage of spontaneous PI+ cells without effector cells (A/(A + B) · 100 (%) at E/T ratio = 0). Flow cytometry analysis. The spleens were cut into small fragments and suspended in lysis buffer (0.155 M ammonium chloride, 0.1 M disodium EDTA, and 0.01 M potassium bicarbonate). Single cell suspensions were prepared in staining medium (PBS with 3% FCS and 0.1% sodium azide) and were stained with the antibodies described above. After 20 min of incubation on ice, cells were washed two times and resuspended in staining medium supplemented with 1 lg/mL PI. Stained cells were analyzed on a LSR II and analyzed with FlowJo software (Tree Star, San Carlos, CA). Semi-quantitative reverse transcription-polymerase chain reaction (RTPCR). Total RNA was extracted from Rag-2 KO mice spleen tissue using Isogen reagent (Nippongene, Japan). RT-PCR was performed with a Takara RNA PCR kit ver.3 (Takara, Japan) according to the manufacturerÕs instructions. RT reaction was done at 42 C for 30 min, 99 C for 5 min, and 5 C for 5 min. PCR conditions were: 1 cycle of initial denaturing at 98 C for 2 min; followed by denaturing at 98 C for 30 s; annealing for 30 s at 67, 62, 60, and 57 C for FasL, perforin, granzyme B, and TRAIL primers, respectively; extension at 72 C (30 s) and final extension for 1 cycle at 72 C (5 min). The primers used for amplications are displayed in Table 1.
Results and discussion Whole body hyperthermia was effectively carried out using the IR box Previous investigations suggested that hyperthermia suppressed the cytotoxic activity of NK cells in vitro [9,20]. Due to the scarcity of in vivo report, we attempted to investigate the effect of WBH on NK cells using mouse model. WBH treatment was carried out in a well-ventilated,
Table 1 List or primers Primers
Upstream
Downstream
Hsp72 Granzyme B Perforin TRAIL FasL b-Actin
5 0 -GCTGACCAAGATGAAGGAGATC-3 0 5 0 -AGATCATCGGGGGACATG-3 0 5 0 -GAACCCTAGGCCAGAGGCAAAC-3 0 5 0 -TCACCAACGAGATGAAGCAGC-3 0 5 0 -CAGCTCTTCCACCTGCAGAAGG-3 0 5 0 -TAAAACGCAGCTCAGTAACAGTCGG-3 0
5 0 -GAGTCGATCTCCAGGCTGGC-3 0 5 0 -TTACACACAAGAGGGCCTCCAG-3 0 5-CCTGGTTGGTGACCTTTGAATCC-3 0 5 0 -CTCACCTTGTCCTTTGAGACC-3 0 5 0 -AGATTCCTCAAAATTGATCAGAGAGAG-3 0 5 0 -TGCAATCCTGTGGCATCCATGAAAC-3 0
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
thermally insulated box heated using infrared rays (IR box), which was designed to mimic the clinical hyperthermia device. We first determined whether the IR box could effectively sustain an increased body temperature of C57BL/6J wild type mice. Fig. 1A shows that mouse body temperature was increased within 10 min after starting the procedure and maintained within the duration of WBH treatment. The elevated body temperature correlated well with the induction of Hsp72, an inducible heat shock protein that has protective functions during heat stress [25]. The increased expression of Hsp72 was detected by immunoblotting of protein lysates from various tissues of WBHtreated mice (Fig. 1B). The expression of the housekeeping gene Hsc70, used as a control, was not changed by WBH treatment. These data indicated that the IR box could work effectively as a WBH device for mice. Whole body hyperthermia suppressed the cytolytic activity of NK cells To evaluate the effect of WBH on the activity of NK cells in vivo, we analyzed the cytolytic activity of NK cells
Fig. 1. Whole body hyperthermia was effectively carried out using the IR box. (A) The rectal temperature of C57BL/6J mice was monitored every 2 min throughout the 1 h WBH treatment in IR box and was maintained between 40.5 and 42.5 C. (B) Western blotting was performed with 50 lg protein lysates from various tissues of heat-shocked or non-heat-shocked C57BL/6J mice. After blotting, membranes were reacted with anti-Hsp72 or Hsc70 antibodies.
1321
obtained from the spleen of WBH-treated or -untreated C57BL/6J mice against YAC-1 cells, a prototypical NK cell-sensitive target. Cytolytic activity of NK cells in these mice was significantly decreased after hyperthermia (Fig. 2A). To confirm this observation, we repeated the cytotoxicity assay using Rag-2 KO mice. We obtained similar results using this mouse model (Fig. 2A). Since Rag-2 KO mice do not contain cytolytic T lymphocytes and NK T cells (Fig. 2B), the only possible immune component that is functional in these mice is the NK cell population. Consistently, these mice displayed higher cytolytic activity compared with wild type mice (Fig. 2A) due to its higher percentage population of NK cells (Fig. 2B). Taken together, these data suggest that WBH (42 C, 60 min) decreased the cytolytic activity of NK cells in vivo. Our results are consistent with some in vitro studies which showed a reduction of lymphocyte cytotoxic activity at 42 C [9,23,26]. The down-regulation of NK cell activity in our study could not be attributed to the decrease in cell viability since there were no significant PI+ splenocytes after WBH treatment (unpublished data). This is in agreement with previous
Fig. 2. Whole body hyperthermia suppressed the cytolytic activity of NK cells. (A) Cytotoxic activity of splenic NK cells of C57BL/6J wild type or Rag-2 KO mice, treated or untreated with WBH, was tested against YAC1 target cells. The results are expressed as percent specific lysis at different effector (E)/target (T) cell ratios. Data shown are representative of three independent experiments. (B) Flow cytometry analyses were performed with peripheral blood mononuclear cells (PBMCs) from peripheral blood of C57BL/6J mice and Rag-2 KO mice. PBMCs were stained with antiCD3 and NK1.1 antibodies.
1322
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
findings that the decrease of NK activity effected by hyperthermia is not related to cell death [9,27]. It was previously reported, however, that WBH performed at lower temperatures (39–40 C) and longer duration resulted in the enhancement of the activity of NK cells [22]. This inconsistency could be explained by the diversity of experimental parameters deemed optimal for clinical advantage. But this seemingly contradictory report underscores the complexity and sensitivity of NK cellsÕ response to its thermal environment. Whole body hyperthermia down-regulates the expression of cytotoxic granules in Rag-2 KO mice spleen NK cells are known to mediate cell death via several possible pathways [12]. To clarify the mechanisms underlying the suppression of NK cell cytolytic activity after hyperthermia, we screened several groups of molecules including the membrane-anchored cytotoxic molecules that bind to the death receptor of target cells (TRAIL, FasL) and secretory cytotoxic granules (perforin, granzyme). The mRNA expression levels were analyzed using semiquantitative RT-PCR. Interestingly, the expression of perforin and granzyme B in Rag-2 KO mice spleen tissues was decreased after WBH (Fig. 3A), but there was no detectable change observed in the mRNA level of mouse FasL and TRAIL after hyperthermia (Fig. 3B). The expression of Hsp72 was induced after WBH, indicating that heat shock was properly carried out (Fig. 3C). We quantified the mRNA expressions of perforin and granzyme B (Fig. 3D) as well as FasL and TRAIL (Fig. 3E) based on the band intensities of the transcripts using Image Gauge software. We observed a significant inhibition of the perforin and granzyme B expression in the spleens of Rag-2 KO mice after WBH (Fig. 3D). It is pertinent to note here that the average decrease of splenic perforin and granzyme in WBH-treated Rag-2 KO mice was 35% from the nontreated controls. This is highly consistent with the result of the cytolytic assay of NK cell activity of Rag-2 KO mice, which showed 35–40% decrease after WBH (Fig. 2A). The WBH experiment was reiterated using ICR nude mice (Crj:CD-1 ICR-nu/nu SPF/VAF; KBT Oriental, Saga, Japan) and in vitro heat shock was performed using human NK cells obtained from healthy volunteers. In both systems, the perforin and granzyme mRNA expressions were suppressed after hyperthermia (unpublished data). Previous studies of perforin-deficient mice established the indispensable role of perforin in NK cell-mediated tumor cell killing [29–32]. Thus, we checked the protein expression of perforin by immunoblotting. Consistent with the RT-PCR data, the protein level of perforin was decreased in mice subjected to WBH (Fig. 3F). Collectively, these results suggest that WBH inhibited NK cell cytolytic activity possibly through the down-regulation of perforin and granzyme B, not through the down-regulation of death receptor ligands. Granzymes are necessary for triggering apoptosis in target cells, but they depend on being appro-
Fig. 3. Whole body hyperthermia down-regulated the expression of perforin and granzyme B in NK cells. (A–C) Five hours after whole body hyperthermia (42 C, 60 min), total RNA was extracted from spleen tissue of Rag-2 KO mice untreated or treated with WBH. Semi-quantitative RT-PCR was performed using 0.5 lg RNA for each gene indicated. (D) Quantitative analyses of perforin (36 cycles) and granzyme B (40 cycles) gene expressions were performed using Image Gauge software (Ver. 3.45, Fuji Film, Japan). Relative mRNA expression levels were normalized to the level of b-actin (20 cycles) that served as an internal control for the amount of RNA used in each reaction. Values are means ± SE from three independent experiments. *p < 0.005 as compared to their respective controls computed using StudentÕs t test. (E) Quantitative analyses of FasL and TRAIL at 36 cycles were performed using Image Gauge. Relative mRNA expression levels were normalized to the level of the internal control b-actin at 20 cycles. (F) Western blotting was performed using protein lysates extracted from whole spleen of Rag-2 KO mice untreated or treated with WBH. After blotting, membranes were reacted using antiperforin and Hsp72 antibodies. Ponceau S staining was performed on membranes to determine equal loading of protein samples.
priately delivered by perforin [19]. Since perforin and granzymes function cooperatively [28], the attenuation of one or the other would impair cytolysis through the granulemediated pathway. Our present results reinforce the importance of perforin/ granzyme-mediated exocytosis as the main pathway used by NK cells in tumor surveillance and elimination [29– 32]. The actual mechanism of how hyperthermia decreases the expression of perforin and granzyme B remains to be clarified but these current findings may lead us to reconsider the current modalities of hyperthermia for cancer treatment. Further understanding of the upstream pathways
T. Koga et al. / Biochemical and Biophysical Research Communications 337 (2005) 1319–1323
that inhibit perforin/granzyme may be able to rescue the hyperthermic inactivation of NK cells and could help elucidate the molecular pathway of NK cells after WBH for better therapeutic development. Acknowledgments We thank Dr. T. Taniguchi (University of Tokyo, Japan) for the Rag-2 KO mice (C57BL/6J strain). This work was supported by grants from the Ministry of Education, Science, Sports and Culture of Japan and the Takeda Science Foundation. References [1] B. Hildebrandt, P. Wust, O. Ahlers, A. Dieing, G. Sreenivasa, T. Kerner, R. Felix, H. Riess, The cellular and molecular basis of hyperthermia, Crit. Rev. Oncol. Hematol. 43 (2002) 33–56. [2] P. Wust, B. Hildebrandt, G. Sreenivasa, B. Rau, J. Gellermann, H. Riess, R. Felix, P.M. Schlag, Hyperthermia in combined treatment of cancer, Lancet Oncol. 3 (2002) 487–497. [3] T.L. Whiteside, N.L. Vujanovic, R.B. Herberman, Natural killer cells and tumor therapy, Curr. Top. Microbiol. Immunol. 230 (1998) 221– 244. [4] C. Christophi, A. Winkworth, V. Muralihdaran, P. Evans, The treatment of malignancy by hyperthermia, Surg. Oncol. 7 (1998) 83– 90. [5] K.S. Zanker, J. Lange, Whole body hyperthermia and natural killer cell activity, Lancet 1 (1982) 1079–1080. [6] M. Onsrud, Effect of hyperthermia on human natural killer cells, Recent Results Cancer Res. 109 (1988) 50–56. [7] B. Kubista, K. Trieb, H. Blahovec, R. Kotz, M. Micksche, Hyperthermia increases the susceptibility of chondro- and osteosarcoma cells to natural killer cell-mediated lysis, Anticancer Res. 22 (2002) 789–792. [8] J.F. Downing, M.W. Taylor, The effect of in vivo hyperthermia on selected lymphokines in man, Lymphokine Res. 6 (1987) 103–109. [9] M.P. Fuggetta, E. Alvino, M. Tricarico, S. DÕAtri, R. Pepponi, S.P. Prete, E. Bonmassar, In vitro effect of hyperthermia on natural cellmediated cytotoxicity, Anticancer Res. 20 (2000) 1667–1672. [10] R.N. Shen, N.B. Hornback, H. Shidnia, L. Lu, H.E. Broxmeyer, Z. Brahmi, Effect of whole-body hyperthermia and cyclophosphamide on natural killer cell activity in murine erythroleukemia, Cancer Res. 48 (1988) 4561–4563. [11] M.J. Smyth, Y. Hayakawa, K. Takeda, H. Yagita, New aspects of natural-killer-cell surveillance and therapy of cancer, Nat. Rev. Cancer 2 (2002) 850–861. [12] M.J. Smyth, E. Cretney, J.M. Kelly, J.A. Westwood, S.E. Street, H. Yagita, K. Takeda, S.L. van Dommelen, M.A. Degli-Esposti, Y. Hayakawa, Activation of NK cell cytotoxicity, Mol. Immunol. 42 (2005) 501–510. [13] W.M. Yokoyama, S. Kim, A.R. French, The dynamic life of natural killer cells, Annu. Rev. Immunol. 22 (2004) 405–429. [14] K. Karre, H.G. Ljunggren, G. Piontek, R. Kiessling, Selective rejection of H-2-deficient lymphoma variants suggests alternative immune defence strategy. 1986, J. Immunol. 174 (2005) 6566–6569.
1323
[15] T. Barlozzari, C.W. Reynolds, R.B. Herberman, In vivo role of natural killer cells: involvement of large granular lymphocytes in the clearance of tumor cells in anti-asialo GM1-treated rats, J. Immunol. 131 (1983) 1024–1027. [16] E. Gorelik, B. Rosen, D. Copeland, B. Weatherly, R.B. Herberman, Evaluation of role of natural killer cells in radiation-induced leukemogenesis in mice, J. Natl. Cancer Inst. 72 (1984) 1397–1403. [17] C. Riccardi, A. Santoni, T. Barlozzari, P. Puccetti, R.B. Herberman, In vivo natural reactivity of mice against tumor cells, Int. J. Cancer 25 (1980) 475–486. [18] H.G. Ljunggren, K. Karre, Host resistance directed selectively against H-2-deficient lymphoma variants. Analysis of the mechanism, J. Exp. Med. 162 (1985) 1745–1759. [19] J.A. Trapani, M.J. Smyth, Functional significance of the perforin/ granzyme cell death pathway, Nat. Rev. Immunol. 2 (2002) 735– 747. [20] T. Kalland, I. Dahlquist, Effects of in vitro hyperthermia on human natural killer cells, Cancer Res. 43 (1983) 1842–1846. [21] H. Yang, W. Lauzon, I. Lemaire, Effects of hyperthermia on natural killer cells: inhibition of lytic function and microtubule organization, Int. J. Hyperthermia 8 (1992) 87–97. [22] R. Burd, T.S. Dziedzic, Y. Xu, M.A. Caligiuri, J.R. Subjeck, E.A. Repasky, Tumor cell apoptosis, lymphocyte recruitment and tumor vascular changes are induced by low temperature, long duration (fever-like) whole body hyperthermia, J. Cell Physiol. 177 (1998) 137– 147. [23] R.N. Shen, L. Lu, P. Young, H. Shidnia, N.B. Hornback, H.E. Broxmeyer, Influence of elevated temperature on natural killer cell activity, lymphokine-activated killer cell activity and lectin-dependent cytotoxicity of human umbilical cord blood and adult blood cells, Int. J. Radiat. Oncol. Biol. Phys. 29 (1994) 821–826. [24] M. Marcusson-Stahl, K. Cederbrant, A flow-cytometric NK-cytotoxicity assay adapted for use in rat repeated dose toxicity studies, Toxicology 193 (2003) 269–279. [25] D.D. Mosser, A.W. Caron, L. Bourget, C. Denis-Larose, B. Massie, Role of the human heat shock protein hsp70 in protection against stress-induced apoptosis, Mol. Cell Biol. 17 (1997) 5317–5327. [26] A. Izumi, S. Koga, M. Maeta, Effects of in vitro hyperthermia on murine and human lymphocytes, Cancer 51 (1983) 2061–2065. [27] A. Yoshioka, Y. Miyachi, S. Imamura, Immunological effects of in vitro hyperthermia, J. Clin. Lab. Immunol. 29 (1989) 95–97. [28] S.S. Metkar, B. Wang, M. Aguilar-Santelises, S.M. Raja, L. UhlinHansen, E. Podack, J.A. Trapani, C.J. Froelich, Cytotoxic cell granule-mediated apoptosis: perforin delivers granzyme B-serglycin complexes into target cells without plasma membrane pore formation, Immunity 16 (2002) 417–428. [29] D. Kagi, B. Ledermann, K. Burki, P. Seiler, B. Odermatt, K.J. Olsen, E.R. Podack, R.M. Zinkernagel, H. Hengartner, Cytotoxicity mediated by T cells and natural killer cells is greatly impaired in perforindeficient mice, Nature 369 (1994) 31–37. [30] M.E. van den Broek, D. Kagi, F. Ossendorp, R. Toes, S. Vamvakas, W.K. Lutz, C.J. Melief, R.M. Zinkernagel, H. Hengartner, Decreased tumor surveillance in perforin-deficient mice, J. Exp. Med. 184 (1996) 1781–1790. [31] M.F. van den Broek, D. Kagi, R.M. Zinkernagel, H. Hengartner, Perforin dependence of natural killer cell-mediated tumor control in vivo, Eur. J. Immunol. 25 (1995) 3514–3516. [32] M.F. van den Broek, H. Hengartner, The role of perforin in infections and tumour surveillance, Exp. Physiol. 85 (2000) 681–685.