These authors contributed equally to this work.
Cross-presentation of antigens from apoptotic tumor cells by liver sinusoidal endothelial cells leads to tumor-specific CD8+ T cell tolerance
Article first published online: 13 OCT 2006
Copyright © 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim
European Journal of Immunology
Volume 36, Issue 11, pages 2960–2970, November 2006
How to Cite
Berg, M., Wingender, G., Djandji, D., Hegenbarth, S., Momburg, F., Hämmerling, G., Limmer, A. and Knolle, P. (2006), Cross-presentation of antigens from apoptotic tumor cells by liver sinusoidal endothelial cells leads to tumor-specific CD8+ T cell tolerance. Eur. J. Immunol., 36: 2960–2970. doi: 10.1002/eji.200636033
- Issue published online: 30 OCT 2006
- Article first published online: 13 OCT 2006
- Manuscript Accepted: 8 SEP 2006
- Manuscript Revised: 28 JUL 2006
- Manuscript Received: 9 MAR 2006
- DFG. Grant Numbers: SFB704, Kn437/9
- Flow Cytometry Core Facility, Institute of Molecular Medicine and Experimental Immunology, University of Bonn. Grant Number: HBFG-109–517
- Adaptive immunity;
- Liver sinusoidal endothelial cells;
Development of tumor-specific T cell tolerance contributes to the failure of the immune system to eliminate tumor cells. Here we report that hematogenous dissemination of tumor cells followed by their elimination and local removal of apoptotic tumor cells in the liver leads to subsequent development of T cell tolerance towards antigens associated with apoptotic tumor cells. We provide evidence that liver sinusoidal endothelial cells (LSEC) remove apoptotic cell fragments generated by induction of tumor cell apoptosis through hepatic NK1.1+ cells. Antigen associated with apoptotic cell material is processed and cross-presented by LSEC to CD8+ T cells, leading to induction of CD8+ T cell tolerance. Adoptive transfer of LSEC isolated from mice challenged previously with tumor cells promotes development of CD8+ T cell tolerance towards tumor-associated antigen in vivo. Our results indicate that hematogenous dissemination of tumor cells, followed by hepatic tumor cell elimination and local cross-presentation of apoptotic tumor cells by LSEC and subsequent CD8+ T cell tolerance induction, represents a novel mechanism operative in tumor immune escape.
liver sinusoidal endothelial cells
The liver is often a site of tumor metastasis once tumor cells reach the circulation. In particular for tumors arising in the gastrointestinal tract, the liver represents the first vascular bed that allows tumor seeding or promotes tumor elimination. The molecular mechanisms involved in tumor cell adhesion within the liver have been partly elucidated and are critical in the initiation of metastasis 1. There is mutual interaction between liver sinusoidal cells and tumor cells, which involves the mannose receptor, release of IL-1β and IL-18 as well as endothelial cell up-regulation of adhesion molecules such as VCAM-1 2, which all favor the development of hepatic metastasis.
At the same time, hepatic sinusoidal cell populations contribute to tumor defense. Kupffer cells as the hepatic macrophage population have the capacity to kill tumor cells efficiently through phagocytosis 3 and induction of apoptosis via release of oxygen metabolites and TNF-α 4. Consequently, elimination of Kupffer cells from the liver results in increased formation of hepatic metastasis 5. NK1.1+ cells constitute another important hepatic cell population involved in local tumor defense, which was first identified by Wisse in 1976 6. NK1.1+ cells in the liver form a heterogeneous population consisting of NK cells and NKT cells situated within and patrolling the liver sinusoid 7. These cells efficiently eliminate metastasizing tumor cells through induction of apoptosis through TRAIL, CD95L or the perforin/granzyme B pathway 8–11. The hepatic overall capacity to eliminate tumor cells seems to be linked to the physiological role of the liver to eliminate gut-derived material from portal venous blood, because the hepatic activity of tumor cell killing is reduced in germ-free mice 12. Furthermore, the liver is known to remove activated T cells from the circulation. Bone marrow-derived as well as organ-resident hepatic cell populations attract circulating activated T cells employing CD54/CD106 and induce T cell apoptosis 13–15. This similarity in retention and elimination of tumor cells and activated T cells suggests that not only may similar molecular mechanisms be employed, but also that the liver plays a role in the regulation of tumor-specific immune response 16.
Despite this liver-resident cytolytic activity, local regulation of antigen-specific immune responses in the liver favors development of T cell tolerance. Different hepatic cell populations contribute to this tolerogenic function. It was demonstrated that hepatocellular MHC class I-restricted antigen presentation to CD8+ T cells leads to development of CD8+ T cell tolerance 17–19. Furthermore, the liver comprises a population of dendritic cells (DC), which is influenced in its immune function by the local microenvironment and induces T cell tolerance rather than immunity 20. Finally, liver sinusoidal endothelial cells (LSEC) serve as organ-resident antigen-presenting cells (APC), which have most efficient scavenger activity and share many similarities with immature DC 21. LSEC present exogenous antigens not only on MHC class II molecules to CD4+ T cells but also have the capacity to cross-present these antigens on MHC class I molecules to CD8+ T cells. Contact of naïve T cells with antigen-presenting LSEC leads to induction of CD4+ and CD8+ T cell tolerance characterized by loss of cytokine expression and failure to develop cytotoxic activity 22, 23.
Here we provide evidence that initial elimination of blood-borne tumor cells influences subsequent tumor-specific CD8+ T cell responses. LSEC participate in the clearance of apoptotic tumor cells from the blood, and process and cross-present tumor-associated antigens to naïve CD8+ T cells. The cross-presentation of apoptotic tumor cells by LSEC results in CD8+ T cell tolerance specific for tumor cell-associated antigens.
Initial elimination of blood-borne tumor cells modifies subsequent immune responses
We asked the question whether blood-borne tumor cell dissemination modulates subsequent development of tumor-associated antigen-specific immune responses. A single intravenous application of the NK-sensitive tumor cell line LB27.4 11 into H-2b mice consistently led to fast eradication of injected tumor cells that was dependent on the presence of NK1.1+ cells (Fig. 1A). When LB27.4 tumor cells were loaded with OVA (LB27/OVA), we observed, at later time points, alteration of OVA-specific immune responses. Subcutaneous implantation of RMA tumor cells stably transfected with OVA (RMA.OVA) into transgenic mice (OT-1), bearing an H-2Kb-restricted transgenic T cell receptor recognizing OVA-derived SIINFEKL, led to tumor cell eradication as a result of strong OVA-specific CD8+ T cell responses 23 and (Fig. 1B). However, if OT-1 mice were i.v. challenged 2 weeks before with LB27.4/OVA, we detected an increased rate of tumor growth after subcutaneous application of RMA-OVA tumor cells (Fig. 1B). This observation suggests that initial encounter with tumor cells modifies subsequent immune responses towards tumor-associated antigens. Eradication of RMA.OVA tumor cells requires CD8+ effector T cells 23; however, LB27.4 fails to cross-present OVA and prime specific naïve CD8+ OT-1 T cells, despite efficient MHC class II-restricted presentation to OVA-specific CD4+ T cells (Fig. 1C). Therefore we assumed that LB27.4 tumor cells were not responsible themselves for the observed alleviation of ovalbumin-specific CD8+ T cell responses in OT-1 mice. It seems rather likely that cellular fragments or apoptotic bodies of killed tumor cells were taken up by APC, and that interaction of T cells with APC cross-presenting tumor-associated antigens may lead to modulation of CD8+ T cell function.
Uptake of apoptotic tumor cells in the liver by LSEC
Given the fast elimination of blood-borne tumor cells through NK1.1+ cells, we next investigated, which APC population ingested tumor cell material in spleen or liver. We injected fluorochrome-labeled tumor cells i.v., and analyzed uptake of apoptotic tumor cell fragments into APC 18 h later. Uptake of apoptotic cells into sinusoidal lining cells of the liver was observed by confocal laser scanning microscopy (data not shown). The strong scavenger activity, as characterized by uptake of fluorochrome-labeled OVA, of hepatic cells taking up apoptotic cell material suggested that scavenger LSEC were involved, similar to the observation by Dini et al. 24. However, due to high hepatic background fluorescence we only detected larger apoptotic cell fragments showing strong fluorescence. We validated uptake of tumor cell material by LSEC by injecting tumor cells labeled simultaneously with two different fluorochromes. The majority of LSEC isolated from mice injected with tumor cells had taken up apoptotic tumor cell material and showed co-localization of both fluorochromes in endosomal vesicles (data not shown). These results indicate that uptake of tumor cell fragments in vivo by LSEC is an efficient process.
Uptake of tumor cell material was quantified by analyzing LSEC as well as hepatic and splenic CD11c+ DC through flow cytometric analysis. Uptake of tumor cell material in vivo was mainly observed into LSEC and hepatic CD11c+ DC, whereas no significant uptake was observed into splenic CD11c+ DC (Fig. 2A). Finally, we investigated uptake of apoptotic tumor cell fragments into LSEC in vitro, which confirmed the observations in vivo that LSEC are able to ingest apoptotic cell material (Fig. 2B). To further elucidate the cooperation between hepatic NK1.1+ cells and LSEC in induction of tumor cell apoptosis as well as elimination of apoptotic tumor cell material, respectively, we performed in vitro co-culture experiments. Freshly isolated hepatic lymphocytes containing NK1.1+ cells were co-cultured with LSEC and CFSE-labeled tumor cells. After 48 h of co-culture LSEC were analyzed for CFSE fluorescence by flow cytometry. If apoptosis of tumor cells was induced by UV irradiation before co-culture with LSEC, significant uptake of apoptotic material into LSEC was observed (Fig. 2C). Importantly, if LSEC were incubated with living tumor cells, we observed more efficient uptake of tumor cell material in the presence of NK1.1+ cells (Fig. 2C). These results indicate that tumor cell killing by NK1.1+ cells strongly enhanced the subsequent uptake of tumor cell fragments by LSEC. Residual uptake of tumor cell material by LSEC in the absence of NK1.1+ cells may be due to spontaneous tumor cell apoptosis or active apoptosis induction through LSEC via release of NO 25 or surface expression of CD95L (data not shown). Collectively, these results support the notion that local tumor cell elimination in the liver, mainly through NK1.1+ cells, is followed by local removal of apoptotic material through LSEC in addition to the well-known removal by phagocytosing Kupffer cells 5.
Cross-presentation of tumor-associated antigens by LSEC
Having identified the efficient uptake of apoptotic tumor cell fragments into LSEC, we next addressed the question whether antigens associated with apoptotic tumor cells were cross-presented on MHC class I molecules to CD8+ T cells. Fig. 3 demonstrates that H2Kb LSEC cross-presented OVA from apoptotic OVA-transfected H2Kd P815 tumor cells to B3Z, which released IL-2 in response to recognition of Kb-SIINFEKL. The lack of IL-2 release from B3Z cells after contact with apoptotic tumor cells alone (data not shown) or H2Kb fibroblasts incubated with apoptotic tumor cells demonstrates that expression of H2Kb-SIINFEKL by LSEC involved active antigen processing and was not related to peptide release from apoptotic tumor cells (Fig. 3).
Induction of CD8+ T cell tolerance towards tumor-associated antigens by LSEC in vitro
Induction of CD8+ T cell tolerance by LSEC does not involve clonal T cell elimination but is an active process with initial stimulation and proliferation of naïve cells before T cells finally develop a tolerant state 23 and (Diehl et al., manuscript in preparation). We investigated the ability of LSEC to prime naïve CD8+ T cells towards antigens associated with apoptotic tumor cells. LSEC or CD11c+ splenic DC were incubated in vitro with apoptotic tumor cells (LB27.4/OVA) followed by addition of naïve CFSE-labeled CD8+ T cells from OT-1 mice. After 5 days, strong proliferation was observed in OVA-specific CD8+ T cells primed by either CD11c+ DC or LSEC (Fig. 4A), demonstrating that cross-presenting tumor cell-associated antigen is an efficient process in LSEC. This result further suggests that low-level expression of costimulatory molecules (CD80/CD86) by LSEC is sufficient to drive naïve CD8+ T cell priming, even in conditions where antigen concentration is low.
To evaluate the functional consequences of cross-presentation of tumor-derived antigens through LSEC, we analyzed the cytotoxic activity and cytokine production of CD8+ T cells primed by cross-presenting LSEC. After stimulation by LSEC cross-presenting antigens derived from apoptotic LB27.4/OVA tumor cells in vitro, OT-1 CD8+ T cells failed to show significant specific cytotoxicity as compared to CD8+ T cells primed by OVA-loaded splenic DC used as positive control (Fig. 4B). Furthermore, OT-1 CD8+ T cells primed by LSEC cross-presenting tumor-associated OVA or soluble OVA failed to express IFN-γ after restimulation with anti-CD3ϵ Ab, whereas CD8+ T cells primed by DC showed strong IFN-γ expression, as shown by intracellular cytokine staining (Fig. 4C). These data indicate that cross-presentation of tumor-associated antigens, similar to cross-presentation of soluble exogenous antigens by LSEC to naïve CD8+ T cells, leads to induction of T cell tolerance in vitro.
LSEC induce CD8 T cell tolerance to tumor-associated antigens in vivo
Having demonstrated the uptake of apoptotic cell material by LSEC in vivo and the ability of LSEC to tolerize CD8+ T cells in vitro towards tumor-associated antigens, we addressed the question whether LSEC were involved in modulation of CD8+ T cell reactivity towards tumor-associated antigens in vivo. To study the functional relevance of antigen presentation by LSEC in vivo we adoptively transferred LSEC, which leads to preferential homing and orthotopic implantation of these cells into the liver as previously described 23, 26. LSEC were isolated from H-2Kb mice challenged with H-2Kd P815.OVA tumor cells, and were adoptively transferred into RAG2–/– mice, which were previously reconstituted with CFSE-labeled spleen cells from OT-1 mice. It is known that transfer of naïve T cells into lymphopenic RAG2–/– mice leads to IL-7-driven homeostatic T cell proliferation 27, which was also observed in our experiments (Fig. 5A). Adoptive transfer of LSEC from mice challenged with P815.OVA induced strong proliferation of OT-1 CD8+ T cells within 10 days, indicating that LSEC cross-presented tumor-associated OVA efficiently to CD8+ T cells in vivo (Fig. 5A). In comparison, adoptive transfer of splenic CD11c+ DC obtained from the same P815.OVA-challenged animals elicited only weak CD8+ T cell proliferation (Fig. 5A), which may be explained by low uptake of apoptotic cell material by splenic DC in vivo after i.v. challenge with tumor cells (Fig. 2A). Homeostatic proliferation of T cells in lymphopenic RAG2–/– mice is known to lead to a pseudo-memory phenotype even in the absence of antigen-specific stimulation 27. We also observed this in our experiments, where significant in vivo cytotoxicity and ex vivo cytokine production was determined in OVA-specific CD8+ T cells in PBS-treated mice as well as in mice that received DC by adoptive transfer (Fig. 5B). In contrast, CD8+ T cells primed in vivo by adoptively transferred LSEC displayed a strong reduction of in vivo cytotoxicity and released little if any cytokines upon antigen-specific re-stimulation ex vivo (Fig. 5B). This implies that LSEC induced CD8+ T cell tolerance even in experimental conditions favoring development of T cell effector function. Collectively, these experiments demonstrate that LSEC induced immune tolerance in CD8+ T cells towards tumor-associated antigens in vivo.
As the high precursor frequency of antigen-specific T cells in TCR-transgenic animals may lead to overestimation of the relevance of T-cell mediated immune responses, we repeated the experiments in C57BL/6 mice expressing a normal TCR repertoire. To improve uptake of apoptotic cell fragments by DC in the spleen we injected tumor cells that had already undergone apoptosis in vitro. We adoptively transferred either splenic CD11c+ DC, hepatic CD11c+ DC or LSEC from P815.OVA-challenged H-2Kb mice into littermates and analyzed OVA-specific CD8+ T cell responses 10 days later. Determination of OVA-specific cytotoxicity in vivo revealed that adoptive transfer of either splenic or hepatic CD11c+ DC induced T cell immunity (Fig. 5C). Specific cytotoxicity towards tumor-associated OVA in vivo was low but reflects the small precursor frequency of CD8+ T cells with specificity for OVA. However, after adoptive transfer of LSEC, we did not detect significant levels of specific cytotoxicity in vivo, although similar numbers of OVA-specific CD8+ T cells were present (data not shown). This result is compatible with the induction of CD8+ T cell tolerance by LSEC. To further validate induction of T cell tolerance, we isolated CD8+ T cells from the animals and re-stimulated equal numbers of CD8+ T cells in an antigen-specific fashion in vitro. Clearly, no expression of IFN-γ or IL-2 was observed in CD8+ T cells obtained from animals adoptively transferred with LSEC, whereas CD8+ T cells primed by adoptively transferred splenic or hepatic CD11c+ DC expressed large amounts of these cytokines. Taken together, our data demonstrate that LSEC can cross-present tumor-associated antigens after uptake of apoptotic bodies and induce antigen-specific CD8+ T cell tolerance in vitro and in vivo.
The liver is considered to promote immune tolerance rather than immunity. Here we provide evidence that LSEC take up apoptotic cell fragments of tumor cells killed locally by NK1.1+ hepatic cells and cross-present tumor cell-associated antigens to CD8+ T cells. The consequence of cross-presentation of tumor cell-associated antigens through LSEC to naïve CD8+ T cells was induction of T cell tolerance characterized by loss of cytokine expression and lack of specific cytotoxicity. Our experiments suggest that hematogenous metastasis of tumor cells and tumor cell elimination in the liver followed by cross-presentation of tumor-associated antigens to CD8 T cells through LSEC contributes to development of tumor-specific T cell tolerance.
Tumor growth is considered to result either from failure of immune surveillance, i.e., outgrowth of mutated tumor cells escaping the immune response 28, 29, or from tumor-specific T cell tolerance induced by the tumor itself 30. It was initially believed that development and growth of tumor cells in non-lymphatic tissue prevents immune recognition of these tumors 31. Meanwhile, it became clear that immune responses can be generated outside secondary lymphatic tissue 23, 32 and that tumors are not simply ignored by the immune system even if they grow at extra-lymphatic sites 33. Cross-presentation of tumor-associated antigens by DC rather than direct antigen presentation by tumor cells is operative in recognition of tumor cells by the immune system and induction of immunity or tolerance 34–36. However, processing of tumor antigens critically depends on DC subtype and on the mode of antigen delivery 37, 38. Another important determinant of tumor-specific T cell responses is the functional state of the antigen-presenting DC. Activated, mature DC promote tumor immunity 34, whereas immature DC rather favor development of T cell tolerance towards tumor-associated antigens 39. Moreover, uptake of apoptotic cell material itself leads to a hypo-reactive state of DC 40. Furthermore, the function of DC presenting tumor-associated antigen and the quality of ensuing immune responses has to be seen in context of the local microenvironment. Increased recruitment of DC into the tumor, local activation of DC in the tumor and removal of CD4+ regulatory T cells from the tumor promote strong induction of tumor-specific immunity 41–43.
While local factors determine tumor-specific immunity, formation of tumor metastasis in the liver is thought to be a consequence of immune escape. The liver appears to be involved in this process: first, tumor cells circulating in the blood stream are either trapped in the hepatic sinusoidal circulation due to the small vessel diameter or are arrested through mutual interaction with hepatic endothelial cells 2, 44; second, Kupffer cells, the large population of hepatic NK cells and liver endothelial cells itself have potent anti-tumor activity leading to killing of tumor cells trapped in the liver 5, 25, 45; third, apoptotic tumor cell material is eliminated locally in the liver by APC (Fig. 2) 24. We have reported that LSEC are a unique population of organ-resident APC capable of cross-presenting exogenous antigens to CD8+ T cells and promoting development of tolerance in naïve CD8+ T cells 23. We therefore wondered whether uptake of cell-associated antigens by LSEC would also result in cross-presentation and induction of CD8+ T cell tolerance. Indeed, LSEC are capable of cross-presenting antigens associated with apoptotic tumor cells to CD8+ T cells (Fig. 3 and 4) similar to DC 46. In contrast to DC, LSEC do not require maturation for efficient cross-presentation. LSEC were as efficient as DC in cross-presenting of cell-associated antigens to naïve CD8+ T cells, as determined by T cell proliferation (Fig. 4).
Most importantly, LSEC cross-presenting tumor-associated antigens induced CD8+ T cell tolerance in vitro (Fig. 4). The functional significance of LSEC-mediated cross-presentation of cell-associated antigens was investigated by adoptively transferring LSEC from mice that received cell-associated antigen via the i.v. route into immunodeficient RAG2–/– mice following reconstitution with antigen-specific CD8+ T cells. In this model, we observed that LSEC cross-presenting cell-associated antigen induced CD8+ T cell tolerance as determined by low specific cytotoxicity in vivo and reduced T cell expression of IFN-γ and IL-2. It is of interest to note that CD8+ T cell tolerance developed despite homeostatic proliferation of CD8+ T cells in lymphopenic RAG2–/– mice, which is known to lead to activation and maturation of CD8+ T cells into effector cells 27. Moreover, induction of CD8+ T cell tolerance through LSEC was not only observed in TCR-transgenic models but also in animals with a normal TCR repertoire (Fig. 5C). Although uptake of apoptotic tumor cells through DC occurs efficiently in vitro, systemic distribution of tumor cells via the blood stream leads to trapping of tumor cells in the liver (data not shown) followed by local tumor cell killing and local removal of apoptotic cells through LSEC and liver DC in addition to Kupffer cells. In contrast, dissemination of apoptotic lymphocytes or cell fragments of already apoptotic tumor cells in the blood is followed by significant uptake through DC in the spleen 40 (and unpublished data). The localization of tumor cell fragments following induction of apoptosis has relevance for the subsequent immune response. Whereas cross-presentation of tumor cell associated antigens by LSEC leads to induction of CD8+ T cell tolerance, local induction of apoptosis in the primary tumor, e.g., in the skin, is associated with induction of strong CD8+ T cell immunity 47. As isolation and adoptive transfer of DC may lead to their activation, the induction of tumor-specific immunity through DC observed in our experiments may rely in part on such stimulation (Fig. 5C). In the absence of activation-inducing signals in vivo DC may also act as inducers of tumor-specific tolerance 48. Interestingly, in contrast to this functional plasticity of DC, LSEC are resistant to such a functional maturation through ligands of TLR3, TLR4, TLR7 or TLR9, despite constitutive expression of these receptors (Scholz et al., submitted). The sentinel role of LSEC is apparently restricted to expression of cytokines such as IL-6 and adhesion-promoting molecules, and does not extend to functional maturation 49, 50. Therefore, we assume that induction of CD8 T cell tolerance by LSEC cross-presenting tumor-derived antigens occurs despite signals derived from pattern-recognition receptors, while immunity triggered by DC rather depends on such activation.
In a physiological situation, induction of peripheral immune tolerance towards cell-associated antigens is thought to avoid induction of autoimmunity towards body cells undergoing apoptosis 40. In particular, uptake and removal of apoptotic leukocytes occurs efficiently in the spleen by DC, which does not promote DC maturation, and as a consequence leads to maintenance of peripheral immune tolerance towards autoantigens 40, 51. Elimination of apoptotic cells also occurs in the liver 13, 24, and here we report for the first time on the consequence of cross-presentation of cell-associated antigens by liver-derived APC. The preferential elimination of tumor cells circulating in the blood together with local removal of apoptotic tumor cell fragments through local cells, in particular tolerogenic LSEC, supports the notion that hematogenous metastasis of tumor cells may accidentally abuse the physiological mechanisms of peripheral tolerance towards autoantigens.
Collectively, our results suggest that hematogenous metastasis of tumor cells into the liver is not merely a consequence of already established tumor immune escape. Rather, initial NK cell-mediated killing of tumor cells and subsequent cross-presentation of tumor antigens by tolerogenic LSEC contribute to development of skewed tumor-specific CD8 T cell response.
Materials and methods
Mice and cell lines
Mice were maintained under SPF conditions at the animal facilities of the ZMBH (Heidelberg, Germany) and the Institute for Molecular Medicine in Bonn. Experiments were performed according to animal experimental ethics committee guidelines. C57BL/6 mice were purchased from Jackson Laboratories (Bar Harbor, ME, USA). OVA257–264-specific MHC class I (OT-1) 52 and the OVA 323–339-specific MHC class II (OT-2) 53 -restricted TCR transgenic animals, and RAG2–/–54 mice have been described previously.
The RMA tumor cell line transfected with OVA (RMA-OVA) and the C57 fibroblast cell line were described previously 23. The OVA257–264-specific CD8+ T cell hybridoma B3Z 55 was provided by N. Shastri. the B cell line LB27.4 is a B cell line generated by fusion of A20 B cell tumor with C57BL/10 splenocytes 56, the P815 cell line, transfected with a minigene encoding the OVA257–264 peptide, was provided by F. Momburg.
Reagents and antibodies
The peptide SIINFEKL (OVA257–264) 52 was obtained from Pineda (Berlin, Germany), OVA (grade II and VII) and saponin were obtained from Sigma (Deisenhofen, Germany). Antibodies for IL-2 (clone JES6–1A12 and JES6–5H4) or IFN-γ (R4–6A2 and XMG1.2) ELISA were obtained from BD Biosciences (Heidelberg, Germany) and the assays were performed according to the manufacturer's instruction. mAb conjugated to different fluorochroms or biotin, and streptavidin-conjugates (SA-PerCP-Cy5.5 and SA-PE-Cy5) were purchased from BD Biosciences. The ME-9F1 mAb was kindly provided by A. Hamann. AcLDL conjugated with AlexaFluor 488 and the celltracer dyes DDAO-SE (FarRed) and CFSE were obtained from Molecular Probes (Leiden, The Netherlands). NK1.1+ cells were depleted in vivo by i.p. injection of 0.5 mg/mouse αNK1.1Ab (PK136) 2 days before the experiments. KbSIINFEKL-Tetramers were kindly provided by T. Schumacher.
Cell isolation procedures and adoptive transfer of cells
LSEC 22 and lymphocytes 57 from murine liver were isolated as described. LSEC were identified by uptake of i.v. injected AcLDL-AlexaFluor488 (12 µg/mouse). The orthotropic implantation of transferred LSEC (1 × 107 cells/mouse) has been reported previously. Splenic CD8+ or CD4+ T cells were purified using MACS (Miltenyi Biotech, Bergisch-Gladbach, Germany) according to the manufacturer's instruction; 5 × 106 CD8+ T cells were injected i.v. into recipient mice. DC were obtained after organs were digested with collagenase D (Roche diagnostics, Mannheim, Germany) and DNase I (Boehringer, Mannheim, Germany). DC were purified by positive selection with αCD11c-conjugated MACS beads (Miltenyi Biotech); 1 × 106 DC were injected i.v. into recipient mice. Apoptosis was induced in tumor cells by UV irradiation (5 min, 15 mJ) using a CL1000 cross-linker from Stratagene. Induction of apoptosis was verified by Annexin staining.
For staining of cell surface molecules, cells were suspended in PBS with 2% FCS and 0.02% NaN3 and stained with fluorochrome-conjugated Ab at (0.1–1 µg/106 cells) for 15 min in a total volume of 50 μL on ice. FcγR-blocking antibody anti-CD16/32 (2.4G2) and unconjugated rat IgG (Jackson ImmunoResearch, West Grove, USA) were added to prevent nonspecific binding. If biotin-conjugated Ab were used, cell bound Ab was detected with streptavidin conjugates (1:200) in a second incubation step. To preparations intended for intracellular staining GolgiPlug and GolgiStop (BD Bioscience, Germany and San Diego, USA) were added for the last 5 h of incubation. For staining of intracellular molecules, cells were fixed for 15 min at room temperature with 2% paraformaldehyde in PBS. Cells were washed and incubated for 20 min in PBS with 2% FCS and 0.5% saponin. Followed by additional 30-min incubation in saponin-containing buffer with fluorochrome-conjugated Ab and unconjugated rat IgG. Cells were washed and analyzed with FACSCalibur or LSR II (BD Bioscience) and data were processed with Flow Jo (Tree Star Inc., San Carlos, USA) software.
In vivo cytotoxicity assays were performed as described 58. In brief, spleen cells were either pulsed with SIINFEKL peptide (1 µg/mL, 45 min at 37°C) and labeled with a high concentration of CFSE (1 μM, 15 min at 37°C; CFSEhigh cells) or were mock treated and labeled with a low concentration of CFSE (0.1 μM; CFSElow cells). Cells were washed twice with PBS and equal numbers of cells from each population were injected i.v. (1 × 107 target cells). Animals were either killed after 4 h (in the case of mice bearing TCR transgenic T cells) or after 18 h (C57BL/6 mice) and presence of target cells in spleen was determined by flow cytometry (BD, CellQuest Pro, Germany). For the determination of the NK cell-mediated cytotoxicity in vivo semi-allogenic LB27.4 cells (CFSEhigh) and syngenic RMA cells (CFSElow) were utilized as targets cells and analyzed 5 h later.
For in vitro cytotoxicity assays RMA (CFSEhigh) and RMA-OVA (CFSElow) tumor cells were utilized as targets and incubated with CD8+ T cells for 4 h at a ratio of 1:1. To calculate specific lysis of the in vivo or in vitro cytotoxicity assays, the following formula were used: % specific cytotoxicity = 100 – [100 × (CFSEhigh / CFSElow)primed / (CFSEhigh / CFSElow)control].
Experiments were performed in triplicates and representative data from one of at least two independent experiments are shown. The statistical analysis was calculated using two-tailed student test or log rank-test.
The authors thank A. Hamann for providing ME9F1 antibody. We acknowledge the assistance of E. Endl and A. Dolf from the Flow Cytometry Core Facility at the Institute of Molecular Medicine and Experimental Immunology, University of Bonn. This work was supported by DFG grants to P.K..