Viral infections can strongly stimulate both NK cell and allospecific CD8 T cell responses, and these same effector cells can lyse allogeneic cell lines in vitro. However, the impact of viral infections on the effector systems mediating rejection of allogeneic tissues in vivo has not been fully explored. Using in vivo cytotoxicity assays, we evaluated the effector systems mediating the rejection of CFSE-labeled allogeneic splenocytes after an infection of C57BL/6 (B6) mice with lymphocytic choriomeningitis virus. Naive B6 mice predominantly used a NK cell-effector mechanism to reject allogeneic splenocytes because they rejected BALB/C (H2d) splenocytes but not CBA (H2k) splenocytes, and the rejection was prevented by immunodepletion of NK1.1+ or Ly49D+ NK cells. This rapid and efficient in vivo cytotoxicity assay recapitulated the specificity of NK cell-mediated rejection seen in longer duration in vivo assays. However, as early as 1 day after infection with lymphocytic choriomeningitis virus, a CD8 T cell-dependent mechanism participated in the rejection process and a broader range of tissue haplotypes (e.g., H2k) was susceptible. The CD8 T cell-mediated in vivo rejection process was vigorous at a time postinfection (day 3) when NK cell effector functions are peaking, indicating that the effector systems used in vivo differed from those observed with in vitro assays measuring the killing of allogeneic cells. This rapid generation of allospecific CTL activity during a viral infection preceded the peak of viral epitope-specific T cell responses, as detected by in vivo or in vitro cytotoxicity assays.
Viral infections precipitate the rejection of transplanted MHC-mismatched tissue grafts in humans and in mouse models (1, 2, 3, 4). Although this virus-induced rejection may be a consequence of altered immunoregulation or altered maintenance of tolerance, it may also be attributed to the direct activation of allospecific effector mechanisms (5, 6, 7, 8, 9, 10). NK cells and allospecific T cells are two effector systems that recognize and reject allogeneic tissue (11, 12). NK cells do so through a combination of negatively and positively signaling surface receptors that can bind to class I MHC molecules (13, 14). T cell recognition of alloantigens is mediated by TCR that either directly recognize class I or class II MHC molecules or indirectly recognize alloantigens that are cross-presented by syngeneic APC (15, 16). The frequencies of alloreactive T cells specific for any given allogeneic haplotype can be quite high in naive mice (17, 18, 19). Notably, during the course of a viral infection, both of these effector systems can be profoundly activated (9, 10, 20).
The host response to a viral infection comes in two waves, an early innate cytokine-driven response that is associated with the activation of NK cells and a later adaptive or acquired immune response that is characterized by the expansion of virus-specific T and B cells (20, 21, 22). During the innate response, NK cells become cytolytically activated by virus-induced type 1 IFN and are induced to proliferate by IL-15 and to synthesize IFN-γ by IL-12 (23, 24). These events occur early during infection, with NK cell activity peaking by day 3 postinfection in most mouse models (25, 26, 27, 28). NK cell numbers can continue to increase moderately thereafter, but their activation state is at a lower level. Substantial work has shown that activated NK cells isolated 3 days after a virus infection can lyse a wide variety of syngeneic and allogeneic cell lines in vitro (10, 25, 29). Viral infections or poly(I:C), a synthetic IFN inducer, can greatly augment NK cell-dependent rejection of implanted tumors in murine hosts (30, 31, 32). During the virus-induced acquired immune response, allospecific CD8+ CTL are generated, and these cross-react with virus-derived peptide epitopes presented in the context of self-class I MHC (2, 9, 33, 34, 35, 36). In C57BL/6 (B6) mice infected with lymphocytic choriomeningitis virus (LCMV),3 allospecific CTL are undetected before day 5 and peak ∼8 days postinfection, with kinetics that mirror the generation of virus-specific CTL activity (9, 10, 33). A portion of these allospecific CD8 T cells generated by infection are maintained into memory, creating a reservoir of memory allospecific T cells in a host that has never been exposed to alloantigens (9, 33, 37).
In the present study, we describe a virus-induced rejection of implanted primary tissue grafts and evaluate the effector systems mediating this rejection during the progression of a viral infection. Previous work evaluating the rejection of allogeneic skin grafts or the engraftment of allogeneic bone marrow revealed rejection systems that evolved over a time period of several days to weeks (38). To succinctly pinpoint the mechanisms of rejection of primary tissue allogeneic grafts in discrete units of time after a viral infection, we used an in vivo cytotoxicity assay to monitor the rejection of CSFE-labeled primary splenocytes (39, 40). Our results show that NK1.1+ and Ly49D+ NK cells in naive mice reject H2d-splenocytes but not H2k-splenocytes. However, within 3 days after infection with LCMV, the rejection of allogeneic tissues is mediated by a combination of NK cells and, surprisingly, CD8 T cells. Both H2d- and H2k-splenocytes were susceptible to this virus-induced CD8 T cell-mediated rejection. This rapid activation of a CD8 T cell effector mechanism for allogeneic graft rejection occurred at the very early stages of the viral infection, before the detection of virus- or allospecific CD8 T cell responses ex vivo. Our findings suggest that mechanisms mediating the rejection of allogeneic tissues can be dramatically influenced by a viral infection and raise concerns for transplant recipients that contract viral infections.
Materials and Methods
Male B6 (H-2b), BALB/CJ (H-2d), and CBA/J (H-2k) mice were purchased from The Jackson Laboratory at 4–5 wk of age and used between 6 and 12 wk of age. B6.129S2-Tcrbtm1Mom (αβ TCR-KO) mice were bred and maintained under specific pathogen-free conditions within the Department of Animal Medicine at the University of Massachusetts Medical School. All experiments were done in compliance with institutional guidelines as approved by the Institutional Animal Care and Use Committee of the University of Massachusetts Medical School.
In vivo cytotoxicity assay
The in vivo cytotoxicity assay was performed as described previously (39, 40). Briefly, spleens were harvested from the indicated mouse strains, and single-cell suspensions were prepared. Splenocytes were washed with HBSS without Phenol Red (Invitrogen Life Technologies), and each population was incubated with various concentrations of CFSE (Molecular Probes) ranging from 2 to 0.016 μM for 15 min at 37°C. Splenocytes were washed with HBSS to remove excess CFSE, the populations being combined at equal ratios, and 3 × 107 total cells were adoptively transferred i.v. into recipient mice. Spleens from recipient mice were harvested 20 h later, a time when rejection of allogeneic splenocytes was easily detectable. Survival of each population was assessed by flow cytometry using a Becton Dickinson LSR2 (BD Biosciences). For visualization of greater than three populations, each group of splenocytes was also labeled with 2 μM of Cell Trace Far Red DDAO-SE (Molecular Probes) to allow discrimination from the recipient cells. To determine peptide-specific killing, syngeneic B6 splenocytes were pulsed with 1 μM of the indicated peptides or with medium alone for 30 min at a concentration of 2 × 107 cells/ml at 37°C/5% CO2 before fluorescent labeling. Specific lysis was calculated by comparing the relative survival of each target population to the survival in NK cell-depleted (i.e., treated with mAb to NK1.1; see equation below) naive mice, using the following equation as described previously (41): 100 − (((percentage of target population in experimental/percentage of syngeneic population in experimental)/(percentage of target population in NK1.1-depleted naive/% syngeneic population in NK1.1-depleted naive)) × 100). To determine specific lysis, the survival of allogeneic splenocytes in NK1.1-depleted naive mice was used as a baseline because optimal survival of these populations was found under these conditions.
In vivo depletions
Depletion of immune cell subsets from recipient mice was done as follows: NK cells were depleted 3 days before adoptive transfer using a dose of anti-NK1.1 Ab (PK136) titrated to eliminate LCMV-induced NK cell but not CTL activity (42). CD4 cells were depleted with 0.5 mg/mouse anti-CD4 Ab (GK1.5) 3 days before adoptive transfer, and CD8 cells were depleted with two 2.5-mg doses of anti-CD8 Ab (2.43) on days 5 and 3 before adoptive transfer. Depletion of Ly49 subsets was performed as previously described (43) using mAb to Ly49D (44) and Ly49G2 (45). Depletions of all subsets were verified by FACS analysis.
Poly(I:C) and viruses
Poly(I:C) (Sigma-Aldrich) was resuspended in PBS to a concentration of 2 mg/ml, and 0.2 mg was delivered i.p. into recipient mice 48 h before the adoptive transfer of target cells. LCMV, strain Armstrong, and Pichinde virus strain AN3739 (PV) stocks were prepared in baby hamster kidney cells (BHK21) as described previously (46). For the generation of acute virus-specific T cell responses, mice were infected ip with 5 × 104 PFU of LCMV.
Synthetic peptides listed were generated by BioSource International and were purified with reverse phase-HPLC to a 90% purity. Final products were analyzed by mass spectroscopy. Peptides used included the following: LCMV-NP396-404 (FQPQNGQFI) and LCMV-GP33-41 (KAVYNFATC).
Two-tailed t tests were performed where indicated using GraphPad InStat.
NK cell-mediated rejection of allogeneic splenocytes in naive mice
NK cells in naive B6 mice have been shown previously to mediate the rejection of H2d-allogeneic bone marrow (47). To directly monitor the lysis of implanted allogeneic splenocytes, we used an in vivo cytotoxicity assay as described in Materials and Methods. Splenocytes from BALB/C (H2d), B6 (H2b), or CBA (H2k) mice were labeled with distinct CFSE concentrations that allowed each population to be distinguished by fluorescent intensity. Twenty hours posttransfer, spleens were recovered from recipient B6 mice and analyzed for the survival of each allogeneic population relative to the reference syngeneic splenocytes (H2b). The percent-specific lysis of each allogeneic population was determined using the equation described in the Materials and Methods. Naive B6 mice efficiently rejected BALB/C (H2d) splenocytes by 20 h posttransfer, but treatment of recipient mice with mAb to either NK1.1 or Ly49D abrogated the rejection (Fig. 1, A and B). Ly49D-positive NK cells recognize Dd and have been shown in prior studies to mediate the rejection of H2d bone marrow in B6 mice (44, 48). In the present study, we show that Ly49D-positive NK cells can also reject mature leukocyte populations. Treatment of recipient mice with mAb to the negatively signaling Ly49G2, which also recognizes H2Dd (49, 50), did not diminish the rejection of H2d splenocytes in B6 recipient mice and may have slightly enhanced rejection, although this was not statistically significant (p = 0.08, Ly49G2 depletion vs naive). In contrast to the rejection of H2d splenocytes, no rejection of CBA (H2k) splenocytes was detected at 20 h posttransfer (Fig. 1, A and B). This is consistent with previous studies showing the resistance of H2k bone marrow to NK cell-mediated rejection in B6 mice (51). Treatment with NK1.1-, Ly49D-, and Ly49G2-specific Abs was effective in depleting the targeted populations (Fig. 1 C and data not shown). These results indicate that the in vivo cytotoxicity assay is an effective method for evaluating NK cell-mediated rejection of allogeneic splenocytes.
Splenocyte rejection under conditions of NK cell activation
NK cells can be cytolytically activated and induced to proliferate in vivo by viral infections and other IFN inducers, such as poly(I:C) (20, 52, 53). Numerous studies have shown that these highly active, B6-derived NK cells can lyse a wide variety of allogeneic targets in vitro, including H2k-expressing fibroblasts, many H2d-expressing cell lines, and even syngeneic targets (10, 29). To examine the ability of activated NK cell populations to reject implanted allogeneic tissues, we evaluated the rejection of H2d and H2k splenocytes after treatment of recipient B6 mice with poly(I:C). Forty-eight hours following poly(I:C) treatment, B6 mice were inoculated with allogeneic (both H2d and H2k) and syngeneic (H2b) splenocytes, and the survival of each allogeneic population was determined 20 h later in the spleen. The H2d population was again efficiently rejected in both naive and poly(I:C)-treated mice, with a slight increase in killing (p = 0.04 for naive vs poly(I:C) treated) after treatment with poly(I:C) (Fig. 2, A and B). Interestingly, the killing of H2d splenocytes in poly(I:C)-treated recipient B6 mice was now only partially sensitive to treatment with mAb to either NK1.1 or Ly49D, suggesting that the rejection was also mediated by a second effector population. NK cells were effectively depleted in poly(I:C)-treated mice as shown in Fig. 2,C. Treatment with poly(I:C) did not enhance the rejection of H2k splenocytes (Fig. 2, A and B), suggesting that the mechanism for NK cell-mediated lysis of allogeneic target cells in vivo may be different from the mechanism of NK cell lysis of continuous cell lines that occurs in vitro.
Infection with LCMV elicits an NK cell-independent rejection mechanism
To examine the rejection of allogeneic implants in recipients undergoing a viral infection, LCMV-infected B6 mice were inoculated with the indicated populations of H2d and H2k splenocytes on either 1 or 3 days postinfection, and survival of allogeneic populations was assessed 20 h later. NK cell-dependent rejection of H2d splenocytes was again evident in naive mice, with no rejection of H2k targets (Fig. 3, A and B). In mice infected 1 day previously with LCMV, NK cells accounted for a majority of the H2d rejection, but a low level of NK cell-independent rejection was also detectable. However, by 3 days postinfection, an alternative effector population was contributing significantly to the rejection of H2d splenocytes because depletion of NK cells diminished only minimally the rejection (Fig. 3, A and B). As shown in Fig. 3 and above (Figs. 1 and 2), naive mice or mice treated with poly(I:C) did not reject H2k splenocytes after 20 h in vivo. In contrast, recipient mice infected with LCMV demonstrated NK cell-independent rejection of H2k splenocytes as early as 1 day after infection, and this increased on day 3 (Fig. 3, A and B). These results demonstrate that a viral infection can dramatically influence the rejection of allogeneic tissues by altering the effector mechanisms and increasing the number of haplotypes that are susceptible to this rapid rejection.
LCMV infection stimulates CD8 T cell-mediated rejection of allogeneic splenocytes
Infection with LCMV generates a vigorous virus-specific T cell response that includes a cross-reactive allospecific T cell population (9, 33). To determine whether infection with LCMV stimulates a T cell-dependent rejection of allogeneic splenocytes, TCR-αβ-knockout mice were inoculated with the indicated populations of splenocytes on either 1 or 3 days postinfection, and survival of allogeneic populations was assessed 20 h later. In TCR-αβ-knockout mice, rejection of H2d splenocytes was dependent completely on NK cells in both naive and LCMV-infected mice (Fig. 4, A and B). Only minimal rejection of H2k splenocytes was detectable in knockout mice infected with LCMV, which is consistent with the results above showing that NK cells in B6 mice are not proficient in rejecting H2k-expressing target populations. These results suggest that infection of normal mice with LCMV promotes a T cell effector mechanism for the rejection of allogeneic targets cells during the early stages of infection.
To determine whether CD8 and/or CD4 T cells mediated the early NK cell-independent rejection of allogeneic populations induced by LCMV, B6 recipient mice were depleted of either T cell subset before inoculation with the indicated populations of splenocytes. Separate groups of B6 recipient mice infected 3 days previously with LCMV were treated with mAb to NK1.1 only or in combination with mAb to either CD4 or CD8. The effectiveness of the CD8 and CD4 depletions is shown in Fig. 5,A. The results shown in Fig. 5,B indicate that the NK-cell independent rejection of both H2d and H2k splenocytes was attributable to CD8 T cells because depletion of CD8 T cells but not depletion of CD4 T cells abrogated rejection when used in combination with anti-NK1.1 mAb. To determine whether NK cells and CD8 T cells were both functioning to reject allogeneic populations by 3 days after infection with LCMV, infected recipient mice were treated with a mAb specific for either NK1.1 or CD8 or with both mAbs. LCMV-infected recipient mice that were depleted of either NK cells or CD8 T cells still rejected H2d splenocytes, and this rejection was only completely abrogated after depletion of both effector populations (Fig. 5, B and C). Thus, after infection with LCMV, NK cells retain the ability to reject H2d splenocytes, but a CD8 T cell effector mechanism is also activated. In contrast, LCMV-induced rejection of H2k splenocytes was abrogated by treatment with mAb to CD8 only but not by depletion of NK cells, suggesting that CD8 T cells predominantly mediated H2k rejection after infection. We next showed that B6 mice infected 3 days previously with PV also displayed CD8 T cell-mediated rejection of both H2d and H2k splenocytes (Fig. 6). These results indicate that infection with viruses rapidly activate CD8 T cells to reject allogeneic splenocytes early during infection, although NK cell-mediated rejection is still apparent.
LCMV-induced rejection of allogeneic splenocytes occurs before in vivo killing of virus-specific targets
The virus-induced CD8 T cell-mediated rejection of allogeneic splenocytes occurred very early during infection. The LCMV-specific CD8 T cell response is normally first detectable at days 4 and 5 after infection, using conventional protocols such as MHC tetramer staining and intracellular cytokine assays (54). The early rejection of allogeneic implants occurred before the detection of virus-specific CD8 T cell responses in vitro, but the in vivo cytotoxicity assay may allow much earlier detection of virus-specific T cells. Therefore, we compared the rejection of LCMV peptide-coated syngeneic splenocytes to the rejection of allogeneic targets using the in vivo cytotoxicity assay. This comparison required the differentiation of five populations by FACS. Allogeneic splenocytes (both H2d and H2k), GP33- or NP396-pulsed syngeneic splenocytes, and a reference population of unpulsed-syngeneic splenocytes were labeled with the same concentration of DDAO-FarRed dye, which allows the five transferred populations to be easily distinguished from the host cells. Each DDAO-FarRed-labeled population was also labeled with a different concentration of CFSE to allow each group to be distinguished from each other. A representative histogram showing the distinct CFSE-staining intensities of the transferred populations (gating on the cells staining positive for DDAO-FarRed) is shown in Fig. 7,A. The results shown in Fig. 7,B again demonstrate the rapid activation of CD8 T cell-mediated rejection during the early stages of a LCMV infection. The killing of LCMV-specific targets is shown in Fig. 7 C. Cytotoxicity against NP396-peptide-pulsed splenocytes was not detectable until day 5, but low levels of cytotoxicity against GP33-coated targets were detected by day 3 postinfection and increased by day 5. These results indicate that the in vivo cytotoxicity assay can detect virus-specific T cell activity early during infection, but the pronounced LCMV-induced rejection of allogeneic populations occurs when virus-specific cytotoxicity is at very low levels.
A majority of our knowledge on allospecific immune responses and transplant rejection has been based on studies in immunologically naive animals, but these studies do not address the potential obstacles posed by the continuous exposure to pathogens that a normal patient would experience under “real world” conditions. Infections with viruses and bacteria can precipitate the rejection of MHC-mismatched transplants (3, 4, 55, 56), but the impact of infections on the effector systems mediating graft rejection has not been evaluated extensively. Viral infections can activate NK cells and T cells, which are both effective in rejecting allogeneic hemopoietic tissues (12, 38). In the present study, we show, using an in vivo cytotoxicity assay, that an acute viral infection will induce a surprisingly rapid activation of a CD8 T cell-dependent rejection mechanism.
NK cells that have been activated by IFN-αβ induced as a result of either exposure to poly(I:C) or a viral infection display enhanced cytotoxicity against allogeneic cell lines in vitro and have an increased capacity to reject tumor cells in vivo (10, 29, 30, 52, 53). Unexpectedly, mice treated with poly(I:C) or infected with LCMV for as little as 1 day also displayed NK cell-independent rejection of allogeneic splenocytes. Prior studies have demonstrated that poly(I:C) treatment prevented the engraftment of allogeneic bone marrow in irradiated mice that would have normally accepted the transplant and that the rejection was mediated by radiation-resistant allospecific CD8 T cells and not NK cells (51, 57). Although there may be experimental differences between the in vivo rejection of allogeneic splenocytes with the engraftment of allogeneic bone marrow, our results showing that infection with LCMV stimulated CD8 T cell-dependent rejection of allogeneic splenocytes would be consistent with results from bone marrow reconstitution studies using poly(I:C).
The ability of NK cells to mediate rejection of allogeneic implants in naive mice is dependent on the haplotype of the engrafted tissues. Previous studies, using proliferation-based assays to monitor the engraftment of allogeneic bone marrow in irradiated hosts, have shown that irradiated B6 mice reject H2d bone marrow via a Ly49D-dependent mechanism (48) but do not readily reject H2k bone marrow (51). The in vivo cytotoxicity assay shown here recapitulates the specificity of NK cell functions shown with more cumbersome long-term assays in vivo and, surprisingly, shows that this can be done with mature splenocyte targets, long thought to be relatively resistant to NK cell-mediated lysis. A recent study has shown that splenocytes from β2m-knockout mice, and therefore not expressing class I MHC, were rejected by NK cells when transferred into wild-type, syngeneic recipients, also demonstrating that mature splenocytes are susceptible to NK cell killing (40). Using the in vivo cytotoxicity assay, we show that while naive B6 mice were unable to reject H2k-splenocytes within the narrow time frame of the assay, infection with LCMV rapidly stimulated a CD8 T cell-dependent rejection of H2k implants. In contrast, poly(I:C) did not stimulate rejection of H2k-splenocytes, and this suggests that viral-derived Ags may be necessary for the activation of H2k-specific T cells, which may be stimulated through cross-reactivity. The rapid CD8 T cell-dependent rejection of both H2d and H2k implants during a virus-specific immune response indicates that an infection can significantly accelerate the generation of allospecific T cell responses.
Viruses such as LCMV, PV, vaccinia virus, and murine-CMV in mice and EBV in humans have been shown to stimulate allospecific CD8 T cells (2, 10, 34, 58). The virus-specific CD8 T cells generated by infection can directly cross-react with alloantigens, and these cross-reactive T cells may mediate graft rejection (9, 33). In addition to cross-reactive mechanisms, the rich cytokine environment generated by viral infections might aid in the activation of allospecific T cells stimulated by third-party Ags on the implanted allogeneic cells over the course of the 20-h in vivo cytotoxicity assay. A low level (15–30% lysis) of CD8 T cell-mediated rejection of allogeneic splenocytes was detectable by 3 days after infection using a short-term, 4-h assay, suggesting that allospecific T cells are cytolytically active before their exposure to the implants. However, this result does not distinguish whether the higher level of rejection at 20 h is due to the allospecific T cells having more time to exert their effector function or to the further activation of the allospecific T cells by the implanted allogeneic splenocytes.
The virus-induced generation of allospecific T cells has been well documented previously (6, 59), but in the present study, we show that the in vivo efficiency of virus-induced allospecific T cells is far greater than expected and seems to dominate at a time when NK cell-mediated mechanisms should have been at their peak. NK cells from mice treated with poly(I:C) or infected with virus can lyse H2k-expressing cell lines in vitro but are unable to reject implanted H2k splenocytes. Of note is that 3 days postinfection with LCMV, virtually all killing of L-929 (H2k) targets in vitro is mediated by NK cells (10), whereas all in vivo killing of H2k splenocytes is mediated by CD8 T cells. Therefore, NK cell-mediated killing of splenocytes in vivo is very different from the killing of tumor cells lines in vitro, which may express additional NK cell ligands such as Rae-1 (60, 61, 62). The rapid virus-induced rejection of allogeneic tissues mediated by CD8 T cells shows that infections can dramatically influence the effector systems mediating the rejection of allogeneic tissues.
We thank Dr. Sung-Kwon Kim for helpful discussions and technical assistance.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
This work was supported by National Institutes of Health Research Grants AR-35506, CA-34461, and AI-46629 (to R.M.W.), a Charles A. King Trust Fellowship from the Medical Foundation (to M.A.B.), and by Institutional Diabetes Endocrinology Research Center Grant DK-52350.
Abbreviations used in this paper: LCMV, lymphocytic choriomeningitis virus; PV, Pichinde virus.