To investigate the role of CD16, NKG2D and NCR1 in NK cell activation we generated ‘triple knockout’ (TKO) mice, i.e. animals with genetic deficiency for all three receptors. TKO mice were healthy and did not show changes in the frequency of B cells, CD4⁺ and CD8⁺ T cells compared to WT controls ( Supplementary Figure 1A ). In addition, the frequency and numbers of NK cells in the spleen, bone marrow (BM), liver, kidney and blood were comparable to those in WT controls ( Figure 1A ).

Next, we wanted to investigate whether the lack of these three activating receptors influences NK cell development. In the bone marrow we only observed minor changes in maturation, with a decrease in stage 1 (CD122⁺NK1.1^-NCR1⁺CD11b^-c-kit^-) and 2 (CD122⁺NK1.1⁺NCR1^-CD11b^-c-Kit) cells ( Figure 1B ). In contrast, when looking at CD27 and CD11b (15), TKO mice had NK cells with a more mature phenotype, both in the spleen and in the bone marrow ( Figure 1C ). To determine whether the change in NK cell maturation depended on an individual receptor, we analyzed the phenotype of NK cells in the spleens of animals lacking only a single gene. The maturation status of NK cells in Klrk1 ^-/- and Ncr1 ^gfp/gfp mice was not altered compared to WT mice, whereas animals deficient for FcγRIIIa (Fcgr3a ^-/-) displayed a reduced frequency of late-stage NK cells ( Supplementary Figure 1B ). In the bone marrow, lower percentages of CD11b^-CD27⁺ and a concomitant increase of CD11b⁺CD27⁺ cells were observed ( Figure 1C ). This implies that early maturation is the same but that signals in the periphery (and lesser extent BM) drive late maturation.

To investigate whether other markers of late maturation are affected by triple deficiency of non-paired receptors, we investigated expression of CD43 and KLRG1 (16, 17). We observed a significant increase in both markers on NK cells in the spleen and BM of TKO mice ( Figure 1D ). The difference in KLRG1 expression was not observed in any of the single knockout mice ( Supplementary Figure 1C ).

Taken together, these data indicate that the lack of CD16, NKG2D and NCR1 has only a minor, but synergistic impact on NK cell development. Their deficiency results in a more mature phenotype of NK cells in the periphery, whereas it does not affect NK cell numbers and frequency.

We previously observed that the lack of one NK cell receptor may impact the expression of others. We therefore analyzed the expression profile of other activating and inhibitory NK cell receptors. We found that several Ly49 receptors were downregulated, including Ly49D and Ly49A ( Figure 2A ). To investigate whether the differential control of Ly49 receptors depended on the synergistic loss of three non-paired receptors, we investigated their expression on animals lacking only a single gene. A similar downregulation of Ly49A was observed in Klrk1 ^-/- and Ncr1 ^gfp/gfp mice, but not in Fcgr3a ^-/- mice ( Figure 2B ). No differences were seen in Ly49D expression on NK cells of Klrk1 ^-/- and Fcgr3a ^-/- mice compared to WT mice, whereas there was significant upregulation of this receptor in Ncr1 ^gfp/gfp mice ( Figure 2C ). The expression levels of Ly49H were similar across the genotypes ( Figure 2D ).

Interestingly, a higher expression of DNAM-1 was observed in TKO mice ( Figure 2A ). In contrast, TIGIT, an inhibitory receptor that competes with DNAM-1 for binding to PVR (18), was found downregulated in TKO mice ( Figure 2A ). Upregulation of DNAM-1 was also seen in Klrk1 ^-/- and Ncr1 ^gfp/gfp mice ( Figure 2E ) but was not associated with a concomitant decrease in TIGIT ( Figure 2F ). Deficiency of CD16 did not impact expression of DNAM-1 or TIGIT. Thus, whereas loss of individual activating receptors impacts the DNAM-1/TIGIT balance, deficiency of all three has the strongest effect.

CD16 is known to impact functional properties of macrophages and dendritic cells (DC) (11, 19). To determine whether the observed differences in receptor expression were cell intrinsic, we generated mixed bone marrow chimeras in which TKO and WT bone marrow cells were mixed in a 1:1 ratio and transferred to irradiated mice with different congenic markers. Differences in receptor expression observed in mixed bone marrow chimeras corresponded with the ones seen in the TKO mice ( Figure 2G ).

Our findings show that the surface expression of individual receptors is fine-tuned by the presence of others. This suggests that a compensating mechanism exists to maintain the activation equilibrium in NK cells even in the absence of three major activating receptors.

Our findings indicate that the lack of activating receptors is compensated by modifying the expression of others. To investigate whether this compensatory effect also operates at a functional level, we performed a series of in vitro stimulations to test the responsiveness of TKO NK cells to activating signals. IL-15 is an essential cytokine for the survival and proliferation of NK cells (20). Therefore, splenic NK cells of WT and TKO mice were cultured in the presence of IL-15 and after 4 days their proliferation was measured. There was no significant difference in the proliferation capacity of TKO and WT NK cells ( Figure 3A ). The expression levels of CD132 and CD122, subunits of the IL-15 receptor, were also comparable to those in WT controls ( Figure 3B ). These data indicate that lack of non-paired activating receptors does not influence the proliferation capacity of NK cells in response to IL-15.

We next investigated the responsiveness of other activating receptors in NK cells of TKO mice. When stimulated through NK1.1, TKO NK cells showed a significantly higher production of IFNγ compared to the WT NK cells ( Figures 3C, D ). This effect was also seen in CD16 and NKG2D deficient mice, but not in animals lacking NCR1 ( Supplementary Figure 2A ). Differences in responsiveness were not observed following Ly49H, IL-12 or IL-12 plus IL-18 stimulation, as production of IFNγ was comparable to those of WT NK cells ( Figures 3C, D ). To determine whether the observed effects were indeed NK cell intrinsic, we performed NK1.1, Ly49H and IL-12 stimulation of NK cells isolated from mixed bone marrow chimeras ( Figure 3E ). The NK1.1 hyper-responsiveness was also observed in the chimeric system, whereas the sensitivity of other receptors was not affected by receptor triple-deficiency. Stimulation of TKO NK cells with the receptor-bypassing stimulus, phorbol-12-myristate-13-acetate (PMA) plus ionomycin did not result in differences in IFNγ production, indicating that modulation of NK cell responsiveness operates at the receptor level and not in the downstream signaling machinery, while differences were observed in CD16 and NKG2D deficient mice ( Figures 3C, D , Supplementary Figure 2B ).

Thus, NK cells of mice lacking three non-paired activating receptors have altered sensitivity to other activating NK cell receptors in a cell-intrinsic fashion.

Since we observed alterations in receptor expression, we wanted to investigate the ability of TKO NK cells to respond to putative hematopoietic (RMA-S) and non-hematopoietic (B16, YAC-1) tumor targets in vitro. Splenic NK cells were co-cultured with B16, RMA-S and YAC-1 cells, and their IFNγ production was measured. B16 is a mouse melanoma that expresses NCR1 and DNAM-1 ligands (21–24). RMA-S is a mouse lymphoma cell line that is MHC-I deficient and therefore susceptible to “missing self”-mediated killing (25), while YAC-1 is a mouse lymphoma cell line that expresses NKG2D ligands (26). Only in the case of YAC-1 co-cultivation did we observe a significant decrease in IFNγ production, whereas co-cultivation of TKO cells with RMA-S or B16 cells did not result in a changed production of IFNγ ( Figure 4A ).

We next assessed the cytolytic capacity of TKO NK cells following in vitro co-cultivation of these cells with tumor targets. Splenic TKO NK cells killed YAC-1, B16 and RMA-S target cells with an efficiency equal to that of WT NK cells ( Figure 4B ). Thus, the loss of CD16, NKG2D and NCR1 does not cause any defect in the ability of NK cells to kill hematopoietic and non-hematopoietic tumor targets in vitro.

Our findings indicate that NKG2D, NCR1 and CD16 are not required for the activation of NK cells in vitro. However, we hypothesized that their role may be indispensable for their functionality in vivo. Therefore, RMA-S cells were injected in WT or TKO mice and tumor progression or survival of animals was followed. We did not observe any differences in tumor growth after s.c. injection of RMA-S cells ( Figure 5A ), nor was survival of animals affected in TKO mice ( Figure 5B ). Next, we examined the ability of these cells to control B16 melanoma cells. Mice were injected i.v. with B16 cells and survival was followed. Compared with WT controls, TKO mice showed significantly reduced survival rates ( Figure 5C ). Observed differences were due to NCR1 deficiency, as mice lacking only this receptor showed comparable survival to TKO animals ( Supplementary Figure 3 ). This goes in line with previously published data on Ncr1-deficient animals (14, 23).

To elucidate why TKO NK cells mediate reduced control of B16 tumor growth, tumor -associated NK cells (TANKs) from tumors of equal size were analyzed and compared with those in the spleen and blood of WT and TKO mice. Although there was no difference in the percentage of infiltrating NK cells ( Figure 5D ), phenotypic alternations were observed. TANKs from TKO mice showed reduced IFNγ production ( Figure 5E ), a cytokine which is known to play a pivotal role in in vivo control of B16 melanoma (14). Moreover, reduced IFNγ production was observed in TKO NK cells obtained from the spleen and blood ( Figure 5E ). TKO NK cells also showed significantly reduced levels of Perforin ( Figure 5F ), while Granzyme B production was only slightly reduced ( Figure 5G ). The observed reduction in IFNγ and Perforin production corresponds with the increase in expression of exhaustion markers such as PD-1 and Ly6C in TKO NK cells ( Figures 5H, I ). Notably, the expression level of CD69 and KLRG1 were significantly reduced in the spleen, blood and tumor of TKO animals, indicating that these cells are less activated compared to those in WT controls ( Figures 5J, K ).

Altogether, these data indicate that TKO NK cells are suppressed in the B16 tumor micro-environment compared to WT NK cells. This is most likely the result of NCR1 deficiency, resulting in reduced survival in the B16 tumor model.

NK cells are very important for the early control of viral infection (27). To investigate whether receptor triple-deficiency compromises anti-viral responsiveness of NK cells, we infected TKO mice with murine cytomegalovirus (MCMV). At 3.5 days post infection viral titers (PFU) were determined in several organs. Viral titers in the spleen and lungs of TKO mice were significantly higher than those in the control group ( Figure 6A ), whereas the viral titers in the liver were not affected ( Figure 6A ). The increase in lung viral titers was not observed in mice lacking only a single activating receptor ( Supplementary Figure 4A ). NK cell depletion of mice before the infection negated differences between WT and TKO mice, indicating that differences in viral titers are the result of impaired functionality of these cells ( Figure 6B , Supplementary Figure 4B ). Nevertheless, NK depletion increased titers both in WT and TKO mice, indicating that deficiency of the three receptors does not abrogate NK cell activity completely.

We next investigated why NK cells of TKO mice have a reduced ability to control MCMV infection. We did not observe changes in the number of NK cells in the spleen after infection ( Figure 6C ). In accordance with this notion, proliferation of cells quantified by Ki-67 staining was comparable between WT and TKO animals ( Figure 6D ). We therefore analyzed the functional capacity of these cells. As in the B16 model, we observed that NK cells of TKO mice have reduced induction of CD69 compared to WT animals ( Figure 6E ). The percentage of KLRG1⁺ NK cells was increased in both the spleen and the liver of TKO mice ( Figure 6F ), as well as TIM-3, TIGIT and LAG-3 ( Supplementary Figure 4D ), suggesting a more exhausted phenotype of these cells. Increased KLRG1 levels were also observed in splenic NK cells from Klrk1 ^-/- and Ncr1 ^gfp/gfp mice, whereas lower induction of CD69 was only seen in the latter genotype ( Supplementary Figure 4C ). Interestingly, whereas Perforin expression was not affected, NK cells of TKO mice expressed significantly higher levels of Granzyme B ( Figures 6G, H ). This upregulation was present both in the spleen and the liver. The same effect was also observed in the liver and spleen of Ncr1 ^gfp/gfp mice, indicating that this phenotype comes from the loss of the NCR1 receptor ( Supplementary Figure 4C ). During viral infection cytokines play an important role in the initial NK cell activation, particularly type I interferons. Indeed, stimulation with IFN-β resulted in an increase of granzyme B production of WT NK cells, which was further increased in TKO cells ( Figure 6I ). Notably, expression of the degranulation marker CD107a was lower on the surface of TKO NK cells, suggesting that Granzyme B accumulates in these cells as a result of a reduced secretion rate ( Figure 6J ). The maturation status of NK cells in TKO mice was predominantly unaltered, with the exception of CD27⁺CD11b⁺ population which was slightly decreased in TKO mice ( Figure 6K ).

To confirm that the altered response of TKO NK cells following MCMV infection are indeed cell intrinsic, mixed bone marrow (BM) chimeras were generated in which WT and TKO tissue was transplanted in wild type recipients. Chimeras were infected with MCMV and analyzed at 1.5- and 3.5-days post infection. The upregulation of Granzyme B and downregulation of CD107a expression were also present in the chimeric system ( Figures 7A, B ), whereas Perforin, Ki-67 and CD69 expression levels remained unchanged between groups ( Figures 7C–E ). Increased KLRG1 expression was observed in NK cells from TKO mice isolated from the liver, but not from the spleen ( Figure 7F ).

Our results demonstrate that lack of all the three activating receptors results in a compromised control of MCMV infection compared to WT mice causing higher viral titers in spleen and lungs, possibly as a result of impaired degranulation.

Mice were strictly age- and sex-matched within experiments and were held in SPF conditions and handled in accordance with institutional, national and/or EU guidelines. Mice used in experiments were between 6 and 12 weeks of age. Klrk1^−/− mice were produced in our laboratory by gene targeting the Klrk11 locus in the C57BL/6 genetic background (13). Wild-type C57BL/6 (strain 000664) and FcγRIIIa ^-/- (strain 003171) mice were from the Jackson Laboratory. Ncr1 ^gfp/gfp mice were kindly provided by O. Mandelboim (Hebrew University Hadassah Medical School). FcγRIIIa ^-/- Klrk1 ^-/- Ncr1^gfp/gfp were generated by interbreeding Klrk1^-/- , Ncr1^gfp/gfp and FcγRIIIa ^-/- mice. Permission for our experiments was given by Ethical Committee of the Faculty of medicine, University of Rijeka and Croatian Ministry of Agriculture, veterinary and Food Safety Directorate (UP/I -322-01/19-01/64,525-10/0543-19-4).

C57BL/6 Ly^5.1/5.1 mice were lethally irradiated with 9.5 Gy and 1 d later received 5–10 × 10⁶ bone marrow cells isolated from C57BL/6 Ly^5.1/5.2 and FcγRIIIa ^-/- Klrk1 ^-/- Ncr1^gfp/gfp Ly^5.2/5.2 mice. Recipients were analyzed 8 wk after transfer. Recipients were given antibiotics (Enroxil 10%; Krka) diluted in a 1:1000 ratio with drinking water for 8–10 d following irradiation.

B16 cell line was purchased from the American Type and Culture Collection (ATCC). Cells were cultured in complete DMEM, supplemented with 10mM HEPES (pH 7.2), 2mM L-glutamine, 10⁵ U/L Penicillin, 0.1 g/L Streptomycin, and 10% FCS. RMA-S and YAC-1 cells were cultured in complete RPMI 1640, supplemented with 10mM HEPES (pH 7.2), 2mM L-glutamine, 10⁵ U/L Penicillin, 0.1 g/L Streptomycin, and 10% FCS. Cells were maintained in incubator at 37°C, 5% CO₂.

For NK killer assay B16/YAC-1/RMA-S cells were labeled with proliferation dye eFlour450 (eBioscience) and co-cultured with C57BL/6J or FcγRIIIa ^-/- Klrk1 ^-/- Ncr1^gfp/gfp splenocytes, respectively, at an effector to-target ratio adjusted to the number of NK cells. After 4h co-culture in a 5% CO₂ atmosphere at 37°C, specific lysis was determined in triplicate by flow cytometric analysis as measured by Fixable Viability Dye eFluor780 (eBioscience) incorporation. Spontaneous death of cells was determined in wells containing targets only (42). In co-cultivation assays for cytokine production analysis, splenocytes from WT or FcγRIIIa ^-/- Klrk1 ^-/- Ncr1^gfp/gfp mice were co-cultured with B16/YAC-1/RMA-S cells in a 1:1 ratio overnight at 5% CO₂ at 37°C in addition to IL-2 (100U/mL, R&D Systems), IL-12 (50 pg, R&D Systems) and addition of Brefeldin A (eBioscience) for the last 4h. IFNγ production by NK cells was analyzed by flow cytometry. For in vitro NK cell stimulations and analysis of cytokine production 5 x 10⁵ splenocytes were stimulated for 4h at 5% CO₂ atmosphere and 37°C in addition to IL-2 (100U, Preprotech) and Brefeldin A (eBioscience). For stimulation through specific activating receptor 10 µg/mL of antibody in PBS was pre-coated on an ELISA plate (αLy49H (3D10), αNK1.1 (PK136)). Cytokines were added in suspension: IL12 (10ng/mL, R&D Systems), IL18 (20ng/mL, R&D Systems), IFNβ (20ng/mL, R&D Systems) as well as phorbol-12-myristate-13-acetate (PMA) plus ionomycin. For the proliferation assay, NK cells were enriched from splenocytes using biotinylated DX5 antibodies, streptavidin-coated beads and magnetic cell sorting (Milteny) and then cultured in the presence of IL15 (50ng/mL, R&D Systems) for 4 days. In experiments of in vitro NK cell analysis we used minimally 5 mice per group and experiments were repeated two to three times.

Cells were pretreated with Fc block (clone 2.4G2, produced in-house). To-pro3 (life Technologies) or Fixable Viability Dye (eBioscience) was used to exclude dead cells. Cells were stained and analyzed in PBS containing 1% BSA and NaN3 with antibodies listed below. For intracellular staining, permeabilization and fixation of cells was done with the Fix/Perm kit (BD Biosciences). Cells were measured on a FACSVerse, FACSAria flow cytometer (BD Biosciences) or MACSQuant (Milteny), and data were analyzed using FlowJo v10 software (Tree Star, Ashland, OR). For flow cytometry, we used monoclonal antibodies to mouse CD3e (145-2C11), CD45.1(A20), CD45.2 (104), NK1.1 (PK136), CD49b (DX5), Ly49H (3D10), Ly49D (4e4), IFNγ (XMG1.2), TNFα (MP6-XT22), NKG2D (CD314) (CX5), CD122 (TM-b1), CD27 (O323), Nkp46 (CD335), DNAM-1 (CD226), TIGIT (MBSA43), Ly49I (YLI-90), Ly49A/D (12A8), CD43 (84-3C1), KLRG1 (2F1), Ly-6A/E (Sca1), Ki67 (SolA15), EOMES (Dan11mag), CD69 (FN50), CD49a (Integrin alpha 1)(TS2/7), CD107a (LAMP-1), CD200R (OX108), Perforin (eBioOMAK-D), Granzyme B (NGZB), (CD11b (M1/70), TIM-3 (CD336)(RMT3-23), LAG-3 (CD223)(3DS223H) and c-kit (CD117)(ACK2) from eBioscience. For hematopoietic stem cell staining, bone marrow cells were stained with Abs against CD4 (GK1.5), CD8 (53-6.7), B220 (RA3-6B2), Gr-1 (RB6-8C5), CD11b (M1/70), TER119 (Ter-119) were used to exclude Lin⁺ cells. Abs were purchased from eBioscience or BD Biosciences.

The tissue culture grown mCMV C3XR129 was produced in mouse embryonic fibroblasts according to standard protocol. The mice were either treated with αNK1.1 (PK136, 250µg/mouse) or PBS 24h before i.v. injection of 2 x 10⁵ PFU of C3XR129 MCMV. Mice were sacrificed at two time points: one and a half day, and three and a half days. Viral titers were determined in spleens, liver and lungs by standard virus plaque assay. In these experiments 4-6 mice per group were used and experiment was repeated two times.

To analyze statistical significance, we used Student’s t-test and ANOVA, with Bonferroni’s post-test correction for multiple comparisons. To assess survival rates, the Kaplan–Meier model was used followed by Log-rank (Mantel-Cox) test for pairwise group comparisons. Statistical significance is defined as: * p < 0.05; ** p < 0.01; *** p < 0.001.