An Accessory Role for B Cells in the IL-12-Induced Activation of Resting Mouse NK Cells¹

Interleukin 12 is a prototypic proinflammatory cytokine produced by dendritic cells (DCs),³ monocytes/macrophages, and activated B cells in the early phase of an immune response to infection (1). The importance of IL-12 in immunity has been demonstrated by the profound susceptibility of IL-12-deficient humans and mice to various pathogens (2, 3, 4). IL-12 is a potent inducer of IFN-γ production, which plays a pivotal role in the initiation of inflammation and the establishment of protective Th1 responses (5, 6). IL-12 has also been used as a therapeutic reagent against various microbial diseases (7, 8) and more recently has proven to have potent antitumor/antimetastasis effects when administered to experimental animals and patients (9, 10, 11, 12). The anticancer effects of IL-12 are primarily mediated by IFN-γ. T cells, NK cells, and NKT cells, are thought to be the primary producers of IFN-γ in response to IL-12 (13, 14). Yet, resting human and mouse NK cells and T cells express none to low levels of the IL-12R and only express appreciable levels of IL-12R after activation (15, 16). Thus, the mechanisms by which resting NK cells are stimulated with IL-12 remain unclear.

IL-18 is another proinflammatory cytokine, initially identified as an IFN-γ-inducing cytokine (17, 18). It is produced as an inactive precursor, which must be cleaved to become biologically active, and is secreted by a variety of cells including epithelial cells, DCs, and macrophages (19). Like IL-12, IL-18 is a strong inducer of IFN-γ production from NK cells and T cells and is known to be involved in autoimmune and inflammatory diseases (20). Although IL-18 alone is not capable of driving IFN-γ production, it has strong synergistic effects with IL-12 and the combination of these two cytokines induces large amounts of IFN-γ production (21). Moreover, Fantuzzi et al. (22) have reported that both the neutralization of IL-18 and caspase 1 deficiency, which mediates the cleavage of the IL-18 precursor to active IL-18, significantly reduce the production of IFN-γ in mice receiving IL-12. These results suggest that IL-18 plays a role in the IL-12-induced production of IFN-γ. Recently, it has been reported that stimulation of ex vivo NK cells with IL-12 requires in vivo priming with IL-18 (23).

To elucidate the mechanisms by which resting NK cells are stimulated by IL-12, we stimulated spleen cells from normal mice with a suboptimal dose of IL-12 alone. After characterizing the cells producing IFN-γ, we tested the effects of IL-12 on various purified cell populations. We also examined mutant mice and conducted cell depletion experiments to determine the requirements for IL-12-induced IFN-γ production. Our results show that B cells are unexpectedly required for IL-12-induced IFN-γ production by NK cells.

Wild-type (WT) and TCRβ/TCRδ double knockout (KO) (TCRβδKO) C57BL/6 (B6) mice were purchased from The Jackson Laboratory and bred pathogen free in our animal facility. IL-18KO and B cell-deficient (BKO) (Igh-6^tm1Cgn) B6 mice were also purchased from The Jackson Laboratory and housed in our animal facility. Male or female 8- to 12-wk-old pathogen-free mice were used in this study. All animal use was approved by the animal care committee of the University of British Columbia and animals were maintained and euthanized under humane conditions in accordance with the guidelines of the Canadian Council on Animal Care.

Anti-CD16/CD32 FcRγ (III/II) (2.4G2; American Type Culture Collection) was purified from hybridoma supernatant. PE-, FITC-, allophycocyanin-, or PerCP-Cy5.5- conjugated mAbs to NK1.1, CD11c, CD3ε, CD19, B220, CD11b, F4/80, and IFN-γ and matching isotype controls were purchased from BD Biosciences. Mouse rIL-12 was purchased from StemCell Technologies. Mouse rIL-18 was purchased from Biovision. RPMI 1640 medium (StemCell Technologies) supplemented with 10% FBS (Life Technologies), penicillin, streptomycin (StemCell Technologies), and 5 × 10⁻⁵ M 2-ME (Sigma-Aldrich) was used for cell culture.

Freshly isolated spleens were ground through a 70-μm nylon sieve to prepare a single-cell suspension and washed with PBS. RBC were lysed with ammonium chloride solution and cells were washed twice. For bulk cultures, cells were dispensed at a density of 1 × 10⁶ cells/ml in 96-well plates and cultured with IL-12 (1 ng/ml) with or without IL-18 (10 ng/ml) for 48 h. To prepare cells for sorting, splenocytes were treated with 2.4G2 mAb to block Fc receptors, washed, and then stained with appropriate Abs for 30 min. To exclude dead cells, propidium iodide was added to a final concentration of 5 μg/ml.

Splenocytes were cultured at a density of 4 × 10⁶/ml in 14-ml round-bottom polypropylene tubes and treated with IL-12 (1 ng/ml) with or without IL-18 (10 ng/ml) for 12–48 h at 37°C. Brefeldin A was added during the final 6 h of culture. Cells were stained with appropriate mAbs before being fixed and permeabilized with a Cytofix/Cytoperm Plus Kit (BD Biosciences) as per the manufacturer’s instructions. Permeabilized cells were then stained with allophycocyanin-conjugated anti-IFN-γ mAb. A FACSCalibur (BD Biosciences) was used for acquisition and FlowJo software (BD Biosciences) was used for analysis.

B6 splenocytes were stained for the depletion of various cell populations as follows: CD11c⁺NK1.1⁻ for DCs, CD11b⁺F4/80⁺CD11c⁻NK1.1⁻ for macrophages, and CD19⁺ for B cells. A FACSVantage SE (BD Biosciences) was used to sort the DC, macrophage, or B cell populations and the remaining population that was considered “depleted” of the population of interest. Bulk splenocytes stained with mAbs but not sorted were used as a control. For add-back experiments, cell populations of interest and depleted cells were mixed at the same ratio as that in the original bulk spleen population. For IL-12 stimulation, 2.5 × 10⁵ cells were cultured with 1 ng/ml IL-12 in 200 μl of medium in 96-well round-bottom plates for 4 days.

WT splenocytes were stained with PE-conjugated anti-CD11c and CD11c⁺ cells (DCs) were isolated using a StemCell Technologies EasySep Murine PE-positive selection kit. DCs (6 × 10³) were cultured alone or mixed with 3 × 10⁴ NK or NK1.1⁺T cells in 96-well round-bottom tissue culture plates (BD Falcon) in a total volume of 200 μl. IL-12 (1 ng/ml) was added at the beginning of culture and supernatants were harvested after 48 h.

WT and IL-18KO B6 mouse splenocytes were stained with mAbs and NK cells (NK1.1⁺CD3⁻) or B cells (CD19⁺) and purified by two rounds of cell sorting to a purity of 95–98%. Cells were cultured in 96-well round-bottom plates at varying ratios of B:NK cells with the NK cell number fixed at 3 × 10⁴ cells in a final volume of 200 μl. Cultures were stimulated with IL-12 (1 ng/ml) for 48 h. For Transwell cultures, purified B and NK cells were cultured at 4:1 or 10:1 ratios in a final volume of 600 μl in 24-well plates with cell-impermeable (0.4- μm pore size) Transwell inserts (BD Falcon). NK cells (9 × 10⁴ cells) in the inserts and B cells in bottom wells were stimulated with IL-12 (1 ng/ml) for 4 days, unless otherwise indicated.

IL-12 (1 μg) in 200 μl of PBS containing 0.1% endotoxin-free BSA (Sigma-Aldrich) was administered i.p. into each mouse daily for 3 consecutive days. Control mice received three injections of PBS-BSA. One day after the final injection, blood was collected via cardiac puncture for serum cytokine analysis.

Cytokines in culture supernatants or sera were measured by ELISA for IFN-γ (eBioscience or BD Opteia, BD Biosciences) and IL-18 (MBL) according to the manufacturers’ protocols. Values below 15–25 pg/ml cytokine were not detected.

All statistical analyses of experimental mean values were performed using the Student paired t test. Values of p < 0.05 were considered significant.

When 10⁶ cells/ml bulk splenocytes from naive B6 mice were stimulated with 1 ng/ml IL-12 for 24 h, ∼1.5 ng/ml IFN-γ was detected in culture supernatants, and higher concentrations of IL-12 (10 ng/m) did not induce more IFN-γ (Fig. 1,A). We therefore used 1 ng/ml IL-12 for all additional experiments. To identify early responders to IL-12, bulk spleen cells were stimulated for a total of 12 h, stained for surface markers, and analyzed for intracellular IFN-γ (Fig. 1,B). The IFN-γ response was limited to the NK1.1⁺ compartment in spleen. Approximately 10% of NK (NK1.1⁺CD3⁻) and 7% of NKT (NK1.1⁺CD3⁺) cells were positive for intracellular IFN-γ, while no T cells (NK1.1⁻CD3⁺) or other spleen cells (NK1.1⁻CD3⁻) were IFN-γ⁺ (Fig. 1,B). Similar results were obtained with longer periods (24, 48, and 72 h) of IL-12 stimulation (data not shown). IL-12 and IL-18, which are known to synergize and induce high levels of IFN-γ secretion from NK and T cells (24), were also used to stimulate resting splenocytes. The combination of IL-12 and IL-18 induced a much higher percentage of NK cells to become IFN-γ⁺ than stimulation with IL-12 alone (Fig. 1,B); the effect on NK1.1⁺ T cells was only marginally higher than that with IL-12 alone. The percentage of IFN-γ-producing cells among IL-12-stimulated splenocytes was consistently ∼10%. To confirm that NK1.1⁺ cells (NK cells and NK1.1⁺ T cells) were indeed the source of the IFN-γ detected in bulk spleen cultures stimulated with IL-12, NK1.1⁺ cells were depleted from bulk splenocyte populations. Depletion of all NK1.1⁺ cells from bulk spleen cells resulted in an almost complete loss of IFN-γ secretion (Fig. 1 C). Adding NK1.1⁺ cells back to the depleted population restored the production of IFN-γ to a level similar to that from undepleted cells (p > 0.1). These results show that stimulation of bulk spleen cells with IL-12 induces IFN-γ production by NK and NK1.1⁺ T cells but not T cells or other spleen cells. NK1.1⁺ cells comprise ∼4% of splenocytes and only 7–10% of them produce IFN-γ upon stimulation with IL-12. Thus, ∼4 × 10³/ml NK1.1⁺ cells among 10⁶/ml bulk splenocyte cultured with IL-12 produced almost 1500 pg/ml (∼0.4 pg/cell) IFN-γ.

Although the above results showed that NK1.1⁺ cells were the source of IL-12-induced IFN-γ, purified NK cells (3 × 10⁵/ml) stimulated with IL-12 produced only very small amounts (60–150 pg/ml) of IFN-γ (Fig. 2,A, left). Purified NK1.1⁺ T cells stimulated with IL-12 alone produced a significantly higher level (500–1000 pg/ml) of IFN-γ (Fig. 2,A). Both populations produced large amounts of IFN-γ (10–40 ng/ml) when both IL-12 and IL-18 were added (Fig. 2,A, right). These results suggested that IFN-γ produced by IL-12-stimulated bulk splenocytes might derive mostly from NK1.1⁺ T cells while NK cells might secrete very little IFN-γ. Alternatively, the stimulation of NK cells with IL-12 is indirect and requires other cells. Resting NK cells do not express the IL-12R to appreciable levels (15). Whereas NKT cells express the IL-12R constitutively (16). Therefore, IL-12 might primarily stimulate NKT cells and induce production of cytokines, which in turn stimulates NK cells. To test these possibilities, spleen cells from TCRβδ double KO mice, which lack all T cells including NK1.1⁺ T cells, were tested. Bulk spleen cells from TCRβδ double KO mice stimulated with IL-12 produced the same amount of IFN-γ as WT spleen cells (p > 0.1, n = 9; Fig. 2 B), indicating that NK cells can be stimulated with IL-12 to produce IFN-γ in the absence of NKT cells.

The above results suggested that the stimulation of NK cells with IL-12 requires the presence of other cells. Because the importance of cellular interaction with DCs for the stimulation of NK cells and NKT cells has been demonstrated (25, 26, 27), we first investigated whether DCs are required. Some NK cells express CD11c (28, 29) and CD11c⁺ NK cells are preferentially stimulated by IL-12 (supplemental Fig. 1⁴), we identified DCs by CD11c⁺NK1.1⁻ and divided bulk splenocytes into DCs and DC-depleted populations. DC-depleted spleen cells cultured for 4 days in the presence of IL-12 produced a slightly lower level of IFN-γ than nondepleted controls (Fig. 3,A). Purified DCs stimulated with IL-12 did not produce a detectable level of IFN-γ and adding back DCs to the depleted cells did not significantly increase (p > 0.1, n = 9) the level of IFN-γ production. Moreover, coculturing highly purified NK cells and DCs in the presence of IL-12 for 2 days did not induce the production of a significant amount of IFN-γ, whereas DCs enhanced the production of IFN-γ by NK1.1⁺ T cells by 2-fold (Fig. 3,B). Thus, DCs promote NK1.1⁺ T cell stimulation, but they do not seem to play a significant role in the stimulation of NK cells with IL-12. Like DCs, macrophages constitutively express the IL-12R (30). Therefore, we tested the effects of depletion of those cells on IL-12-induced IFN-γ production by bulk splenocytes. Because some NK cells also express CD11b and CD11c, macrophages were identified by CD11b⁺F4/80⁺CD11c⁻NK1.1⁻. The depletion of macrophages had no effect (p > 0.1, n = 9) on the amount of IFN-γ being secreted in stimulated cultures (Fig. 3 C).

In contrast to the results from the DC and macrophage depletions, the removal of B cells (CD19⁺) resulted in a striking loss of the IFN-γ production by IL-12-stimulated splenocytes, while adding B cells back to the B cell-depleted splenocytes significantly restored the production of IFN-γ (Fig. 4,A). To further test the role of B cells in IL-12-mediated stimulation, splenocytes from BKO mice were stimulated with IL-12 and the amount of IFN-γ produced was compared with WT. As expected from the B cell depletion experiment, IL-12-stimulated BKO splenocytes secreted significantly less IFN-γ than WT splenocytes (Fig. 4,B, left). BKO mouse spleens had slightly higher percentages of NK and NKT cells (data not shown), and BKO splenocytes produced larger amounts (p < 0.05, n = 9) of IFN-γ when stimulated with IL-12 plus IL-18 (Fig. 4,B, right). Therefore, the low level of IFN-γ production by IL-12-stimulated BKO splenocytes was unlikely due to defective NK or NKT cells in BKO spleen. To determine whether B cells are sufficient for the stimulation of NK cells with IL-12, we highly purified (95–98%) NK and B cells from TCRβδ KO mice and cocultured them with IL-12 at varying ratios of B:NK cells (2:1, 4:1, and 10:1), with the NK cell number fixed at 30,000/well (Fig. 4 C). The amounts of IFN-γ production increased in a dose-dependent manner as the ratios of B:NK cells increased. Purified B cells without NK cells did not produce any IFN-γ with IL-12 stimulation (data not shown). Additionally, we could detect low levels (37 ± 8 pg/ml, n = 3) of TNF-α, another inflammatory cytokine, in the supernatants of these IL-12-stimulated B-NK cocultures (data not shown). NK cells or B cells cultured alone did not secrete any detectable TNF-α upon IL-12 stimulation. We also tested whether B cells promote IL-12-induced NK cell cytotoxicity. Our data show that IL-12-stimulated NK cells are poorly cytotoxic (supplemental Fig. 2). There was not a statistically significant difference between the WT and BKO splenocytes with respect to killing ability after IL-12 stimulation (p = 0.21).

The above results showed that purified NK cells are not stimulated with IL-12 alone and require B cells, whereas purified NK cells can be stimulated with a combination of IL-12 and IL-18 in the absence of B cells. Human tonsillar B cells constitutively express IL-18 mRNA (31) and mouse B cells secrete IL-18 upon Ly49D cross-linking (32). Moreover, a recent study has shown that the in vivo priming of NK cells with IL-18 is important for the stimulation of NK cells with IL-12 (23). To clarify the role of IL-18 and B cells, we first tested bulk splenocytes from IL-18KO mice. IL-18KO splenocytes stimulated with IL-12 produced significantly lower levels of IFN-γ than WT splenocytes (Fig. 5,A, left). This reduced response to IL-12 was not because of a functional deficiency in the IL-18KO splenocytes since they produced high levels of IFN-γ when stimulated with a combination of IL-12 and IL-18 (Fig. 5,A, right). To determine whether B cells play a role in the effect of IL-18 deficiency on NK cell stimulation, NK cells and B cells were purified from WT and IL-18KO splenocytes and cocultured in various combinations in the presence of IL-12. WT B cells promoted both WT and IL-18KO NK cell stimulation with IL-12, whereas IL-18KO B cells did not have much effect on either WT or IL-18KO NK cells (Fig. 5 B).

To determine whether cell contact between B cells and NK cells is also required for the stimulation of purified NK cells with IL-12, they were cocultured in Transwells. When NK cells and B cells were separated by a cell-impermeable membrane, only a small amount of IFN-γ was produced upon IL-12 stimulation, regardless of the B:NK ratios. Coculturing NK and B cells in the same compartment, which facilitates cell-cell contact, resulted in the B cell dose-dependent production of significantly higher amounts (p < 0.005, n = 9) of IFN-γ (Fig. 5 C). Therefore, B cell-NK cell physical contact appears important for the activation of NK cells by IL-12.

The above results showed that B cells are necessary and also sufficient for the in vitro production of IFN-γ from NK cells induced with IL-12. To test whether B cells also play a role in an in vivo response to IL-12, we i.p. injected IL-12 into normal B6 and BKO mice and compared the levels of serum IFN-γ. Our preliminary studies showed that a daily injection of 1 μg/mouse for 3 consecutive days induced a significant level of IFN-γ in the serum of WT mice. Under these conditions, ∼1200 pg/ml IFN-γ was detected in the sera of WT mice, whereas a 4-fold lower level of IFN-γ was detected in the sera of BKO mice (Fig. 6,A). No IFN-γ was detected in the sera from vehicle (PBS plus 0.1% BSA)-injected WT or BKO mice. Since our in vitro results also suggested that endogenous IL-18 is important for IL-12-induced IFN-γ production, we also measured serum IL-18. Unlike IFN-γ, IL-18 was detected in the sera of vehicle-injected WT and BKO mice. Whereas the levels of serum IL-18 increased significantly (p < 0.05) following IL-12 injection in both WT and IL-18KO mice, they were significantly lower (p < 0.01, n = 4) in BKO mice than in WT mice (Fig. 6 B).

The data presented above have revealed a novel role for B cells in the stimulation of mouse NK cells with IL-12 in vitro and in vivo. IL-12 is considered to be a potent stimulator of NK cells and stimulation of bulk spleen cells with a suboptimal dose of IL-12 alone induces IFN-γ production by NK cells and NK1.1⁺ T cells. However, purified NK cells are not stimulated by IL-12 alone, whereas in the presence of B cells purified NK cells are stimulated by IL-12 and secrete significant amounts of IFN-γ. B cell depletion from normal spleen inhibits IL-12-induced IFN-γ production, and spleen cells from B cell-deficient mice produce much lower levels of IFN-γ than WT spleen cells. Moreover, purified B cells, but not DCs, support IL-12-induced IFN-γ production by purified NK cells. Therefore, B cells are necessary and also sufficient for NK cell stimulation with IL-12. Importantly, the critical role for B cells in IL-12-induced IFN-γ production is not restricted to in vitro cultures. Injection of IL-12 into B cell-deficient mice results in significantly lower levels of IFN-γ than those in WT mice. It has been proposed that this difference might be due to IFN-γ production by B cells (33). However, our current study has shown that B cells do not secrete detectable levels of IFN-γ; it should also be noted that B cell-deficient spleen cells stimulated by IL-12 do produce some IFN-γ, albeit significantly less than WT spleen cells do. This is most probably due to a B cell-independent stimulation of NK1.1⁺ T cells, as purified NK1.1⁺ T cells are stimulated by IL-12 alone and produce IFN-γ. Similarly, the reduced IFN-γ production in B cell-deficient mice injected with IL-12 may also be due to NK1.1⁺ T cell stimulation. Nevertheless, spleen cells from TCRβδ double KO mice, which lack all T cells including NK1.1⁺ T cell, produce no less IFN-γ than WT spleen cells, indicating that NK cells are an important primary source of IFN-γ upon IL-12 stimulation.

The precise mechanisms by which B cells help NK cell stimulation with IL-12 are still unclear. As reported by Chaix et al. (23), spleen cells from IL-18-deficient mice stimulated by IL-12 produce significantly less IFN-γ than WT spleen cells. This has been suggested to be due to an in vivo priming of NK cells by IL-18. In our current study, a very small amount of IFN-γ is produced by purified WT NK cells stimulated by IL-12 alone. Furthermore, NK cells from IL-18KO mice can be stimulated by IL-12 in the presence of WT B cells, whereas WT NK cells are not stimulated by IL-12 in the presence of IL-18KO B cells. These results suggest that the production of IL-18 by B cells, rather than the in vivo priming of NK cells by IL-18, is critical. However, we have been unable to detect IL-18 in the cultures of purified NK cells and B cells stimulated with IL-12. It should be noted that purified NK cells treated with IL-12 require as little as 10 pg/ml recombinant mouse IL-18 to produce the level (∼1500 pg/ml) of IFN-γ commonly seen in our cultures (data not shown). It is then possible that the level of IL-18 secreted by B cells in our cultures while sufficient for synergizing with IL-12 for IFN-γ is too low to be detected by ELISA. At the same time, our results with Transwell cultures indicate that physical contact between the B cells and NK cells is required. Therefore, IL-18 secretion may not be the only mechanism by which B cells support IL-12-mediated NK cell stimulation. It is possible that for IL-18 production IL-12-stimulated B cells require contact-dependent costimulatory signals from the NK cells to produce IL-18. To examine this possibility, we have tested spleen cells from CD40 KO and CD80/86 double KO mice. These cultures produced only slightly lower levels of IFN-γ than WT cells upon IL-12 stimulation (data not shown). An alternative possibility is that B cells may not only produce IL-18 but also present it to NK cells, perhaps via the IL-18R. The IL-18R consists of two chains, the α-chain that binds IL-18 and the β-chain that does not efficiently bind IL-18 but associates with the complex of IL-18/IL-18Rα and transduces signals (34). To test whether B cell-derived IL-18 is bound to IL-18Rα on B cells and presented to NK cells expressing IL-18Rβ, we have tested IL-18Rα KO mouse spleen cells and found that IL-18Rα deficiency does not have a significant effect on B cell-NK cell cooperation in IL-12 stimulation (data not shown). Therefore, it is unlikely that B cells trans-present IL-18 to NK cells. Recently, human DCs have been shown to make contact with NK cells and secrete IL-18 into the cleft at the contact site (35). Whether B cells also form similar contact with NK cells remains to be determined.

In 1991, Wyatt and Dawson (36) showed that human B cells could form conjugates with NK cells, which resulted in the release of increased levels of IFN-γ. Yuan and colleagues (37) and Yuan (38) also found that in vivo-activated splenic B cells induce IFN-γ production by IL-2-stimulated NK cells, and this was dependent upon direct physical contact between the B and NK cells. However, it should be noted that the combination of IL-2 and IL-12 or IL-18 strongly stimulates NK cells and induces IFN-γ production without B cells. In our current study, resting NK cells and B cells cooperate in NK cell stimulation induced by IL-12.

It is unexpected that B cells, rather than DCs, are critical for the IL-12-mediated stimulation of NK cells. It has been shown that DCs promote NK cell stimulation by trans-presenting IL-15 (39). IL-12 also stimulates IL-15 production by DCs (40). However, in our study with IL-12, DCs show no effect on IL-12-mediated NK cell stimulation. It should be noted that DCs used for most studies on NK-DC interaction are generated in vitro, whereas DCs in our current study are freshly isolated CD11c⁺ cells. DCs thus purified from normal spleen do not seem to be stimulated by IL-12 because they do not support NK cell stimulation with IL-12. Even stimulation with IL-12 plus IL-18, which strongly stimulates NK cells and NK1.1⁺ T cells, does not induce much IFN-γ production by DCs. Thus, although DCs are known to be the major sources of IL-12 and IL-18 (41, 42), they do not seem to produce IL-18 when stimulated with IL-12 alone. Monocytes and macrophages have also been shown to be important partners to NK cell IFN-γ production. In response to TLR ligands (43, 44, 45) or pathogens (46, 47), monocytes and macrophages can up-regulate NK cell stimulatory factors like NKG2D ligands, IL-15, IL-12, and/or IL-18 and activate NK cells. However, our depletion of monocytes/macrophages had no effect on the IL-12-induced IFN-γ production by NK cells, most likely because monocytes/macrophages are not stimulated with IL-12 alone.

IL-12 is often used to enhance immune responses against infections and cancer. The adjuvant effect of IL-12 is thought to be primarily mediated by IFN-γ. NK cells, NKT cells, and T cells have been reported to produce IFN-γ upon stimulation by IL-12 (48, 49, 50, 51), but our results have clearly shown that NK cells and NK1.1⁺ T cells are the only significant IFN-γ producers upon IL-12 stimulation. Furthermore, our finding that B cells are critical for IFN-γ production by NK cells in vitro as well as in vivo has important implications to the clinical use of IL-12.

We thank the BC Cancer Research Center Flow Core Facility for expertise in cell sorting and the BC Cancer Animal Resource Center for care of all of the mice used in the study.

The authors have no financial conflict of interest.