|Home | About | Journals | Submit | Contact Us | Français|
Natural Killer (NK) cells respond rapidly during viral infection. The development, function, and survival of NK cells are thought to be dependent on Interleukin (IL)-15. In mice lacking IL-15, NK cells are found in severely decreased numbers. Surprisingly, following infection of IL-15- and IL-15Rα-deficient mice with mouse cytomegalovirus (MCMV), we measured a robust proliferation of Ly49H-bearing NK cells in lymphoid and non-lymphoid organs, capable of cytokine secretion and cytolytic function. Remarkably, even in Rag2−/− × Il2rg−/− mice, a widely used model of NK cell deficiency, we detected a significant number of NK cells one week after MCMV infection. In these mice, we measured a greater than 300-fold expansion of NK cells, which was dependent on recognition of the m157 viral glycoprotein ligand and IL-12. Together, these findings demonstrate a previously unrecognized independence of NK cells on IL-15 or other common-γ signaling cytokines during their response against viral infection.
IL-15 and its receptor (IL-15Rα) are important in the homeostasis of NK cells and memory CD8+ T cells (1–10). IL-15 bound to the receptor IL-15Rα on the surface of dendritic cells is “trans-presented” to IL-15–responsive cells bearing the shared IL-2 and IL-15 receptor common-β chain (CD122) (11–17). During infection, dendritic cells respond to inflammatory cytokines, leading to the production of IL-15 and IL-15Rα (13–15, 18–20). Expression of IL-15 and IL-15Rα on activated myeloid cells has thus been thought to contribute to NK cell responses against pathogens. Although mice deficient in IL-15 or the IL-15 receptor severely lack peripheral NK cells, a small population of NK cells (<0.1%) is detectable in the spleen (3, 5). We sought to determine whether these NK cells that arise in the absence of IL-15 signals can mount effector responses against viral infection.
C57BL/6 (B6) and Rag2−/− × Il2rg−/− B6 mice were purchased from the National Cancer Institute and Taconic, respectively. Il15−/−, Il15ra−/−, Il15−/− × Il15ra−/−, and Rag1−/− × Il2rb−/− B6 mice were bred at UCSF. Experiments were done according to the UCSF Institutional Animal Care and Use Committee guidelines. 5×104 PFU of a salivary gland stock of MCMV (Smith strain) or MCMV-Δm157 was injected intraperitoneally (21). 750 μg of neutralizing anti-IL-12 p70 (clone C17.8) was injected intraperitoneally 24 hours prior to infection.
Cells were stained with antibodies against NK1.1, CD3, Ly49H, Ly49D, KLRG1, NKp46, NKG2D, CD27, and DX5 (CD49b) (eBioscience or BD Pharmingen). Flow cytometry was performed using a LSRII with CELLQuest software (Becton Dickinson). Splenocytes were enriched for NK cells by using a NK cell isolation kit (Miltenyi Biotec), followed by AutoMACS magnetic bead separation. NK cells were incubated in tissue culture plates treated with N-(1-(2,3-dioleoyloxyl)propyl)-N,N,N-trimethylammonium methylsulphate (Sigma) and coated with anti-NK1.1 or anti-Ly49H or PBS for 5 h at 37 °C in the presence of Golgiplug (BD Pharmingen), followed by staining for LAMP-1 and intracellular IFN-γ (BD Pharmingen) (22). NK cells were used as effector cells in a 4 hr 51Cr release assay (23) against Ba/F3 and m157-transfected Ba/F3 cells (22).
The spleens of Il15ra−/− mice contain <0.1% CD3−, NK1.1+ NK cells, compared with 2–5% in wild-type (WT) B6 mice (5). The absolute number of NK cells is decreased and the percentage of NK cells bearing the Ly49H receptor is lower in Il15ra−/− (~10%) compared to WT mice (~50%) (Fig. 1A). During the NK cell response against MCMV in WT mice, the Ly49H+ NK cells preferentially proliferate during the first several days of infection (21, 24, 25), a response specific for the m157 gene product of MCMV (22, 26). When we infected WT and Il15ra−/− mice with MCMV, both mice showed an increase in Ly49H+ NK cell numbers and comprised >80% of total NK cells at day 7 post-infection (PI) (Fig. 1A). A similar expansion was not observed in the Ly49D+ Ly49H− NK cell subset (Fig. 1A). With precursor numbers of ~2×104 total Ly49H+ NK cells in the spleen, the absolute number of Ly49H+ NK cells in Il15ra−/− mice at day 7 PI expanded approximately 72-fold to become comparable to the numbers found in uninfected WT B6 mice (greater than 106) (Fig. 1B). NK cells from MCMV-infected Il15ra−/− mice expressed comparable levels of activating receptors (NK1.1, NKp46, Ly49H, and NKG2D) and activation markers (KLRG1 and CD27) as WT mice (Fig. 1C). When NK cells at day 7 PI were stimulated ex vivo with anti-NK1.1 or -Ly49H, these cells upregulated LAMP-1 and produced IFN-γ (Fig. 1D), demonstrating that NK cells do not require IL-15 signals to mediate effector functions during MCMV infection.
Il15−/− mice are also deficient in NK cells (3). On day 7 PI, we observed robust expansion of Ly49H+ NK cells in the spleen of MCMV-infected Il15−/− mice (Fig. 2A). Expression of KLRG1, a NK cell activation marker (27), was comparable in WT and Il15−/− mice (Supplemental Fig. 1). With <104 total Ly49H+ NK cells in the spleen prior to infection, the absolute number of Ly49H+ NK cells in Il15−/− mice at day 7 PI was >105, representing a 50-fold increase in absolute numbers (Fig 2A). We tested the ability of Ly49H+ NK cells from Il15−/− mice to kill m157-bearing target cells. Ly49H+ NK cells isolated at day 7 PI from MCMV-infected Il15−/− mice were able to efficiently lyse m157-bearing target cells (Fig. 2B).
To test whether specific viral ligand (and not inflammation alone) is required to drive NK cell proliferation, we infected Il15−/− mice with MCMV or a mutant strain lacking m157 (MCMV-Δm157). Unlike MCMV-infected Il15−/− mice, which contained a large percentage and absolute number of Ly49H+ NK cells at day 7 PI (45.3-fold expansion), infection of Il15−/− mice with MCMV-Δm157 did not generate many NK cells (1.7-fold expansion) compared to uninfected controls (Fig. 2, C and D). The diminished proliferation of NK cells during infection with MCMV-Δm157 is not due to defective replication as this mutant virus is equally or more virulent than WT MCMV (28). Adoptive transfer of WT NK cells into Il15−/− recipient mice results in the rapid loss of the transferred NK cells (1–10); however, during infection with MCMV, we measured large numbers of transferred NK cells (CD45.1+) at day 7 PI in spleen and liver of the Il15−/− recipients (Supplemental Fig. 2, A and B). At later time points after MCMV infection (day 15 and 30 PI), transferred NK cells were difficult to recover (data not shown), suggesting that following the resolution of infection, NK cells again require IL-15 for survival. Expansion and survival of adoptively transferred WT NK cells were not observed in Il15−/− mice infected with MCMV-Δm157 (Supplemental Fig. 2, A and B). Altogether, these experiments demonstrate that both viral infection and m157 are required for robust NK cell proliferation in the setting of IL-15 deficiency.
Rag2−/− × Ilr2g−/− mice are currently the best model of NK cell deficiency. Without the common-γ chain (γC), NK cells cannot receive signals from any cytokine of the γC family, including IL-2, IL-4, IL-7, IL-9, IL-15, and IL-21. In naïve Rag2−/− × Il2rg−/− mice, NK cells were barely detectable (0.05% in spleen and 0.2% in liver) (Fig. 3A and Supplemental Fig. 3). When we infected WT and Rag2−/− × Il2rg−/− mice with MCMV and measured NK cell responses at day 7 PI, Rag2−/− × Il2rg−/− mice showed an increase in total NK cell numbers in spleen (comprising 1.6% of splenocytes) and liver (comprising 1% of hepatic lymphocytes) (Fig. 3A and Supplemental Fig. 3). In Rag2−/− × Il2rg−/− mice, as with the other models of IL-15 deficiency, only the NK cells expressing Ly49H (and not Ly49D+ Ly49H− NK cells) expanded vigorously, upregulating KLRG1 (Fig. 3A). With <1000 total Ly49H+ NK cells in the spleens of uninfected Rag2−/− × Il2rg−/− mice, the absolute number of Ly49H+ NK cells at day 7 PI became >105 (320-fold expansion) (Fig. 3B). Similar results were obtained analyzing Rag1−/− x Il2rb−/− mice (Supplemental Figure 4). Collectively, these data demonstrate that during MCMV infection NK cells do not require cytokines of the γC family for their activation and proliferation.
IL-12 is produced by dendritic cells and granulocytes in response to viral and bacterial infection and is required for the generation of Th1 cells, as well as inducing proliferation and IFN-γ in activated CD8+ T cells and NK cells (reviewed in Nat Rev Immunol. 2003 Feb;3(2):133–46). Additionally, IL-12 plays an important role in NK cell production of IFN-γ and NK cell blastogenesis during MCMV infection (29, 30), and NK cell proliferation in response to MCMV infection is somewhat impaired in Il12−/− mice (31, 32). To address whether IL-12 contributes to NK cell expansion in the setting of IL-15 deficiency, we injected Il15−/− × Il15ra−/− mice with a neutralizing anti-IL-12 antibody prior to infection. Uninfected Il15−/− × Il15ra−/− mice have very few peripheral Ly49H+ NK cells, but 7 days following infection, significant numbers and percentages of Ly49H+ NK cells were detected in the spleen (78%) and liver (91%) (Fig. 4A). However, absolute numbers of Ly49H+ NK cells at day 7 PI were ~30-fold less in anti-IL-12 treated mice compared to control mice (Fig. 4B). The overall expansion of Ly49H+ NK cells in Il15−/− × Il15ra−/− mice was ~70-fold, versus a 2-fold increase in anti-IL12 treated Il15−/− × Il15ra−/− mice (Fig. 4B). Thus, IL-12, not IL-15, contributes greatly to the overall NK cell response following MCMV infection in mice lacking the ability to produce or respond to IL-15. Future studies are required to determine whether the small number of NK cells that do proliferate during MCMV infection represent a unique IL-15-independent subset or new bone marrow emigrants that are rescued from death by IL-12 and inflammatory cytokine signaling. Moreover, although we have shown that IL-12 is involved in NK cell expansion in the absence of IL-15, other factors might also contribute to their proliferation and survival. In conclusion, our surprising findings contribute added insight into the cytokines (or lack thereof) that NK cells require during an immune response against viral infection.
We thank Drs. Jody Baron, Wayne Yokoyama, Anne Hill, and Ulrich Koszinowski for generously providing reagents.
1J.C.S. is an Irvington Postdoctoral Fellow of the Cancer Research Institute. This study was supported by NIH grant AI068129 and L.L.L. is an American Cancer Society Professor.
The authors have no financial conflicts of interest.