Infection Leishmania Major Th1 Response and Control Cutaneous

Mice Lacking NK Cells Develop an Efﬁcient Th1 Response and Control Cutaneous Leishmania major Infection1

Abhay R. Satoskar,2* Luisa M. Stamm,* Xingmin Zhang,† Anjali A. Satoskar,‡ Mitsuhiro Okano,* Cox Terhorst,† John R. David,* and Baoping Wang†

NK cells are believed to play a critical role in the development of immunity against Leishmania major. We recently found that transplantation of wild-type bone marrow cells into neonatal tg⑀ 26 mice, which are deficient in T and NK cells, resulted in normal T cell development, but no or poor NK cell development. Using this novel model we analyzed the role of NK cells in the development of Th1 response and control of cutaneous L. major infection. Mice selectively lacking NK cells (NK؊T؉) developed an efficient Th1-like response, produced significant amounts of IL-12 and IFN-␥, and controlled cutaneous L. major infection. Ad- ministration of neutralizing IL-12 Abs to NK؊T؉ mice during L. major infection resulted in exacerbation of the disease. These results demonstrate that NK cells are not critical for development of protective immunity against L. major. Furthermore, they indicate that IL-12 can induce development of Th1 response independent of NK cells in NK؊T؉ mice following L.major infection. The Journal of Immunology, 1999, 162: 6747–6754.

he leishmaniases comprising a number of diseases caused involved in host defense against this parasite (11). Furthermore, a by the intracellular protozoan parasite Leishmania have a recent study indicated that NK cells are involved in protection and T wide spectrum of clinical manifestations (1). While sus- healing of cutaneous leishmaniasis in humans (12). Therefore, we ceptible BALB/c mice develop large nonhealing lesions following examined the development of Th1 response and growth of L. ma- L. major infection, most other mouse strains, including C3H, jor in mice specifically lacking NK cells. Our results show that NK CBA/J, and C57BL/6, are resistant and develop small lesions that cells are not essential for the development of Th1 response and heal spontaneously. It is widely accepted that protective immunity immunity to L. major infection in these mice. against cutaneous L. major infection is associated with the development of a Th1-like response and the production of cytokines Materials and Methods such as IL-12, IL-2, and IFN-␥ (2–4), whereas susceptibility to L. Mice major is associated with the development of a Th2-like response Tg⑀26 mice were maintained through sib breeding in the animal facility of and the production of cytokines such as IL-4 and IL-10 (5). the Beth Israel Deaconess Medical Center (Boston, MA) (13, 14). Al- ⑀ ϫ NK cells are a subpopulation of bone marrow-derived large, though the tg 26 transgenic founder was a (C57BL/6 CBA/J)F2, all Ϫ ϩ granular lymphocytes that lack T cell- and B cell-specific subset tg⑀26 mice used in this study were H-2k. Mice lacking NK cells (NK T ) ␥␦⑀ were generated by transplanting fetal liver or bone marrow cells from the markers (TCR-, CD4-, CD8-, CD3 -, and Ig-), but express some ϫ ⑀ (C57BL/6 CBA/J)F1 mice into neonatal tg 26 mice, as described re- specific markers, such as NK 1.1 and ASGM1 (6). NK cells have cently.3 Age- and sex-matched wild-type CBA/J (H-2k), C57BL/6 (H-2b), been shown to play a critical role in innate immunity against a and BALB/c (H-2d) mice were purchased from The Jackson Laboratory variety of viruses, bacteria, fungi, and parasites (7). The protective (Bar Harbor, ME). Two types of immunocompetent mice with functionally role of NK cells has been attributed to their ability to secrete im- competent NK and T cells were used as wild-type controls in this study. ϫ b/k ␥ One type was (C57BL/6 CBA/J)F1 (H-2 ) generated through breeding munoregulatory cytokines, such as IFN- (8), lyse host cells in- of C57BL/6 ϫ CBA/J. The other type was tg⑀26 mice reconstituted with fected with the intracellular pathogens, and directly inhibit growth ϫ b/k (C57BL/6 CBA/J)F1 (H-2 ) bone marrow or fetal liver cells at 2–3 wk of microorganisms (9, 10). of age, instead of neonatally as in the generation of NKϪTϩ mice. These ϩ ϩ Previous studies have demonstrated that depletion of NK cells immunocompetent mice were termed as NK T (tg⑀26Y) mice. Five ϩ ϩ ⑀ using anti-asialo GM1 antiserum significantly reduces early IFN-␥ weeks after the transplantation, NK T (tg 26Y) mice develop functionally competent NK and T cells, and their levels are comparable to those production in resistant C3H/HeN mice and renders them suscep- ϫ ϩ ϩ observed in wild-type (C57BL/6 CBA/J)F1 mice. Furthermore, NK T tible to cutaneous L. major infection, suggesting that NK cells are (tg⑀26Y) and NKϪTϩ mice have similar levels of CD4ϩ and CD8ϩ T cells. Of note, all the NKϩTϩ,NKϩTϩ(tg⑀26Y), and NKϪTϩ mice used in this study were analyzed by flow cytometry of PBL before the infection, and *Department of Immunology and Infectious Diseases, Harvard School of Public † the lymph node and spleen cells upon sacrificing animals to confirm the Health, Boston MA 02115; and Department of Medicine, Division of Immunology, lack or presence of NK cells and the presence of T cells. and ‡Department of Pathology, Beth Israel Deaconess Medical Center, Boston, MA 02115 Parasites and infection protocols Received for publication December 1, 1998. Accepted for publication March 17, 1999. L. major. LV39 was maintained by serial passage of amastigotes inocu- lated s.c. into the shaven rumps of BALB/c mice. Amastigotes isolated 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 from the lesions of infected mice were grown to stationary-phase promas- with 18 U.S.C. Section 1734 solely to indicate this fact. tigotes as described previously (15). Mice were injected in the hind footpad with 2 ϫ 106 L. major stationary-phase promastigotes. Disease progression 1 This work was supported by National Institutes of Health Grants A122532-13 (to J.R.D.), AI17651 (to C.T.), and HD35562-01(to B.W.). B.W. is the recipient of a Basil O’Connor Starter Scholar Research Award. 3 B. Wang, K. Nguyen, X. Zhang, A. Nichogiannopoulou, S. J. Simpson, J. Guimond, 2 Address correspondence and reprint requests to Dr. Abhay R. Satoskar, Department B. A. Croy, J.-C. Gutierrez Ramos, G. A. Hollander, C. A. Biron, and C. Terhorst. of Immunology and Infectious Diseases, Harvard School of Public Health, 665 Hun- 1999. Distinct homing of engrafted hematopoietic stem cells in neonatal mice differ- tington Avenue, Boston, MA 02115. E-mail address: [email protected] entially affects T lymphocyte and NK cell development. Submitted for publication.

was monitored by measuring the increase in thickness of the infected foot- Results pad using a dial-gauge micrometer (Mitutoyo, Kanagawa, Japan) at weekly NKϪTϩ mice have functionally normal CD4ϩ and CD8ϩ T cells intervals up to 10 wk after infection and comparing this to the thickness of the contralateral uninfected footpad. but lack cytotoxic NK cells Quantitation of parasite loads Recently, we demonstrated that transplantation of wild-type bone marrow or fetal liver cells into neonatal tg⑀26 mice results in nor- Parasite burdens in the infected footpad were determined by homogenizing mal development of T cells, but poor NK cell development.3 All footpads of individual mice and carrying out limiting dilution analysis as described previously (15). The results were expressed as reciprocal log the neonatal transplanted tg⑀26 mice used in this study had a nor- ϩ ϩ parasite titers. mal number of CD4 and CD8 T cells as confirmed by flow ϩ Ϫ ␣␤Ϫ Ab ELISA cytometry. The NK1.1 CD3 TCR- , which represent 2–4% of wild-type splenocytes, however, were markedly diminished in Peripheral blood was collected at 2-wk intervals from tail snips of all the neonatal transplanted mice (Fig. 1, A and B). Peripheral lymph experimental animals infected with L. major. Blood was centrifuged at Ϫ ϩ ϫ nodes from NK T mice also showed markedly diminished levels 200 g, and serum was collected to determine titers of Th1-associated ϩ Ϫ Ϫ ϩ ϩ IgG2a and Th2-associated IgG1 Leishmania-specific Abs by ELISA as of NK1.1 CD3 TCR-␣␤ cells than those from NK T mice described before (15). (1–2% in NKϩTϩ mice and background levels (Ͻ0.3%) in Ϫ ϩ ϩ ␣␤ϩ Flow cytometric analysis NK T mice). In contrast, NK1.1 TCR- T cells, which represent 0.4–1% of wild-type splenocytes, were present in the neo- The lymph node cells, spleen cells, and PBL were analyzed by three-color 6 natal transplanted mice (Fig. 1B). The selective NK cell-deficient flow cytometry as described previously (15). Briefly, 0.5–1 ϫ 10 cells in Ϫ ϩ ϩ ϩ 50 ␮l were incubated with prestaining buffer (PBS, 4% BSA, 0.5% sodium mice were termed as NK T mice. Both CD4 and CD8 T cells Ϫ ϩ azide, 15% mixture of normal hamster, normal rat, and normal mouse sera, from NK T mice were functionally competent as assessed by anti-Fc receptor Ab) for 5 min. The cells were then stained with biotinyl- MLR and CTL assays, respectively (Fig. 1, C and D). NK cell lytic ated Ab for 30 min, washed once, followed by staining with a mixture of function, as measured by splenocyte cytotoxicity against NK cell- streptavidivin-RED670 (0.4 ␮l/sample; Life Technologies, Rockville, MD), PE- and FITC- conjugated Abs (0.5 ␮g/sample) for 30 min. The cells sensitive YAC-1 cells, however, was generally nondetectable or were washed twice, fixed in 1% formaldehyde, and analyzed with a FAC- Ͻ10% of wild-type control levels (Fig. 1E). When young tg⑀26 Scan using CellQuest software (Becton Dickinson, Mountain View, CA). mice were reconstituted with F1 bone marrow or fetal liver cells at All procedures were performed on ice until analysis. The following Abs ⑀ ⑀ ␣␤ 2–3 wk of age (tg 26Y), these mice had comparable wild-type (F1) were used: 145-2C11 (CD3- , H57-597 (TCR- ), RM4-5 (CD4), 53-5.8 Ϯ Ϯ Ϯ (CD8␣), 53-2.1 (Thy 1.2), RA3-6B2 (B220), PK136 (NK1.1), and DX5 T (Fig. 2A) and NK cell (2.5 0.63%, 1.3 0.05%, and 0.3 ϩ Ϫ ϩ ϩ ϩ ϩ (all purchased from PharMingen, San Diego, CA). 0.1% of NK 1.1 TCR-␣␤ cells in NK T ,NK T (tg⑀26Y), and NKϪTϩ mice, respectively) levels. Furthermore, NK cells Allogenic T cell proliferation and cytotoxicity assays from NKϩTϩ(tg⑀26Y) mice were functionally as competent as Allogenic T cell proliferation and cytotoxicity assays were performed by those from NKϩTϩ mice (Fig. 2B). Therefore, these mice were k kϫb stimulating responsive spleen cells (H-2 or H-2 ) with irradiated allo- ϩ ϩ d termed as NK T (tg⑀26Y). Since they have identical background genic BALB/c (H-2 ) spleen cells as described previously (14). NK cell Ϫ ϩ ϩ ϩ cytotoxicity assay was performed using yeast artificial chromosome as the NK T mice, the NK T (tg⑀26Y) mice were also used as (YAC)4 cells as described before (13). NK cell competent controls, in addition to the wild-type mice ϩ ϩ T cell proliferation and cytokine assays (NK T ). The draining popliteal lymph nodes were removed from L. major-infected mice at week 10 postinfection, and T cell proliferation assays were per- NKϪTϩ mice control cutaneous L. major infection formed as previously described (15). Briefly, 3 ϫ 105 lymph node cells Ϫ ϩ ϩ ϩ were added in triplicate to the wells of 96-well flat-bottom tissue culture Following infection with L. major,NKT ,NKT , and plates and stimulated with either 20 ␮g/ml L. major Ag (prepared from NKϩTϩ(tg⑀26Y) mice developed lesions, which resolved sponta- stationary-phase promastigotes by six cycles of freezing at Ϫ70°C and ␮ neously within 60–70 days (Figs. 2C and 3A). The course of L. thawing at 37°C) or 1 g/ml of Con A. Culture supernatants from these Ϫ ϩ ϩ ϩ assays were analyzed for production of IL-4 (reagents purchased from major infection was similar in NK T ,NKT , and ϩ ϩ Endogen, Cambridge, MA; detection limit, 5 pg/ml) and IFN-␥ (reagents NK T (tg⑀26Y) mice (Figs. 2C and 3A). In contrast, concomi- purchased from PharMingen; detection limit, 20 pg/ml) by capture ELISA tantly infected tg⑀26 (NKϪTϪ) and BALB/c mice developed large as described previously (15). nonhealing lesions and did not control the infection (Figs. 2C and Histopathology 3A). The lesion grew significantly faster in NKϪTϪ mice than in Ϫ ϩ ϩ ϩ Ϫ Ϫ BALB/c mice (Fig. 3A). Examination of the histopathology of the Infected foot pads from L. major-infected NK T ,NK T , and NK T Ϫ Ϫ mice were excised and fixed in decalcifying solution F (Stephens Lab, infected footpads from NK T and BALB/c mice revealed ulcer- Riverdale, NJ) for 7 days. The tissues were processed and embedded in ation and extensive s.c. tissue destruction with a diffuse inflam- ␮ paraffin, and 4- to 8- m sections were cut. The sections were hydrated and matory infiltrate consisting of heavily parasitized macrophages, stained by routine hematoxylin eosin staining. eosinophils, and neutrophils (Fig. 3C). On the other hand, infected Anti-IL-12 neutralizing Ab treatment foot pads from NKϪTϩ and NKϩTϩ mice displayed inflammatory Rat anti-mouse IL-12 (p40/p70) (clone: C17.8) neutralizing mAb was infiltrate comprised predominantly of lymphocytes and macro- kindly provided by Dr. T. Veldman (Genetics Institute, Cambridge, MA). phages with few parasites (Fig. 3, D and E). There were no sig- Ϫ ϩ NK T mice were treated by i.p. administration of 0.5 mg anti-IL-12 nificant differences in the parasite burdens in footpads of NKϩTϩ neutralizing Ab or control Ab 1 day before L. major infection and weekly and NKϪTϩ mice. The lesions from L. major-infected NKϪTϪ dose of 0.5 mg/mouse thereafter until 7 wk. and BALB/c mice, however, contained significantly more parasites Statistical analyses (at least 10 logs more) than NKϪTϩ and NKϩTϩ mice (Fig. 3B). Student’s unpaired t test was used to determine significance of values ob- These results indicate that, although NK cells may play a role in tained. Differences in Ab endpoint titers were determined using Mann- innate immunity to L. major as reported in previous studies using Whitney U prime test. SCID and RAG-2Ϫ/Ϫ mice, they suggest that NK cells are not essential for control of L. major infection when immunocompetent 4 Abbreviation used in this paper: YAC, yeast artificial chromosome. T cells are present. The Journal of Immunology 6749

FIGURE 1. NKϪTϩ mice lack NK1.1ϩCD3ϪTCR-␣␤Ϫ cells but have normal T cell development. A, CD4 and CD8 proﬁle of lymph node cells from a wild-type (NKϩTϩ) and an NKϪTϩ mouse. B, NK1.1 and TCR-␣␤ proﬁle of spleen cells, indi- cating that NKϪTϩ mice have markedly diminished NK1.1ϩ␣␤Ϫ cells, but have a relatively normal number of NK 1.1ϩ␣␤ϩ cells. C,NKϪTϩ mice have functionally normal CD4ϩ T cells as assessed by MLR assay. D, NKϪTϩ mice have functionally normal CD8ϩ T cells as assessed by CTL assay. E,NKϪTϩ cell mice lack NK cell cytotoxicity against YAC-1 cells. Data for C–E represents the analysis of the spleen cells obtained from a single NKϩTϩ and two NKϪTϩ mice.

NKϪTϩ mice develop efficient Th1-like response following L. NKϩTϩ and NKϪTϩ mice, respectively). NK 1.1ϩ cells, how- major infection ever, were markedly reduced in the lymph nodes from NKϪTϩ Ϯ Ϯ ϩ ϩ Ϫ ϩ NK cells have been shown to be a major source of IFN-␥, a cy- mice (2.1 0.31% and 0.4 0.06% in NK T and NK T mice, respectively). At this time, lymphocytes from NKϩTϩ, tokine critical for development of the Th1 lymphocyte subset of Ϫ ϩ ϩ NK T , and BALB/c mice displayed greater Ag-specific prolif- the CD4 T cell population in resistant mice following L. major Ϫ Ϫ infection (11). Therefore, we compared IL-12 and IFN-␥ produc- erative responses than those derived from NK T mice. At this Ϫ ϩ tion by Leishmania Ag-stimulated draining lymph node cells from time, LmAg-stimulated lymph node cells from both NK T and ϩ ϩ L. major-infected NKϪTϩ and NKϩTϩ mice. On day 60 postin- NK T mice produced IL-12 and IFN-␥, although IL-12 levels Ϫ ϩ fection, the draining lymph node cells from L. major-infected were significantly higher in NK T mice (Fig. 4, A and B; p Ͻ NKϩTϩ and NKϪTϩ mice contained a similar number of lym- 0.05). Ag-stimulated lymph node cells from concomitantly in- phocytes (1.02 Ϯ 0.13 ϫ 107 and 1.35 Ϯ 0.4 ϫ 107 in NKϩTϩ and fected BALB/c mice produced significant amounts of IL-4, but no Ϫ ϩ NKϪTϩ mice, respectively; p Ͻ 0.375). Furthermore, there were IL-4 was detected in culture supernatants from NK T and ϩ ϩ no significant differences in proportions of B220ϩ (43.7 Ϯ 5.4% NK T mice (Fig. 4C). Neither IL-4 nor IFN-␥ was detectable in and 41.6 Ϯ 1.9% in NKϩTϩ and NKϪTϩ mice, respectively), the lymph node cell culture supernatants from NKϪTϪ mice, CD4ϩ (28.6 Ϯ 4.2% and 32.9 Ϯ 4.7% in NKϩTϩ and NKϪTϩ which contained basal levels of IL-12. (Fig. 4, A–C). All groups mice, respectively), and CD8ϩ (10.1 Ϯ 1.6% and 11.1 Ϯ 1.1% in produced comparable levels of IL-10 (data not shown). Similarly, 6750 NK CELL-DEFICIENT MICE CONTROL CUTANEOUS LEISHMANIASIS

FIGURE 2. NKϩTϩ(tg⑀26Y) mice develop Th1 response and control L. major infection as efficiently as NKϩTϩ mice. A, CD4 and CD8 profile of lymph node cells from a wild-type (NKϩTϩ) and NKϩTϩ(tge26Y) mice. B,NKϩTϩ(tge26Y) mice have normal NK cell cytotoxicity against YAC-1 cells similar to that observed in NKϩTϩ mice. C, Course of L. major infection in NKϩTϩ,NKϩTϩ(tg⑀26Y), NKϪTϩ, and NKϪTϪ mice. Disease progression was monitored by measuring increase in the thickness of infected footpad, as described before. D, In vitro LmAg-induced IFN-␥ production by the lymph node cells. Data expressed as mean Ϯ SE. Asterisk indicates statistically significant difference between two groups. ND, not detectable.

at an earlier time point on day 36 postinfection, both NKϪTϩ and NKϩTϩ and NKϩTϩ(tg⑀26Y) mice, respectively). Similar results NKϩTϩ mice produced IL-12 (0.114 Ϯ 0.4 ng/ml and 1.3 Ϯ 0.17 were observed on days 30 and 45 postinfection (data not shown). ng/ml in NKϩTϩ and NKϪTϩ mice, respectively) and IFN-␥ (1.068 Ϯ 0.6 ng/ml and 2.9 Ϯ 0.25 ng/ml in NKϩTϩ and NKϪTϩ IL-12 is critical for development of Th1 response and controls mice, respectively), but no IL-4. In additional experiments, L. major infection in NKϪTϩ mice NKϩTϩ(tg⑀26Y) mice were also examined. NKϩTϩ(tg⑀26Y) ϩ ϩ To determine whether IL-12 is critical for development of Th1 mice controlled L. major infection as efficiently as NK T mice response in NKϪTϩ mice during L. major infection, we treated L. (Fig. 2C). Furthermore, there was no significant difference in major-infected NKϪTϩ with i.p. injections of IL-12 neutralizing ␥ IFN- production by LmAg-stimulated lymph node cells from Ab or control Ab 1 day before infection and weekly thereafter for ϩ ϩ ϩ ϩ ⑀ NK T and NK T (tg 26Y) mice (Fig. 2D). 7 wk. Anti-IL-12 mAb-treated NKϪTϩ mice developed signifi- NKϪTϩ mice fail to produce Leishmania-specific IgG2a, despite cantly larger lesions than control animals following L. major infection (Fig. 6). At wk 8 postinfection, Ag-stimulated lymph node development of Th1 response Ϫ ϩ Ϫ ϩ ϩ ϩ Ϫ Ϫ cells from control NK T mice produced significantly higher Ab responses in L. major-infected NK T ,NK T ,NK T , and amounts of IFN-␥ (mean levels, 1.34 Ϯ 0.4 ng/ml) than those from BALB/c mice were analyzed by measuring titers of Leishmania- anti-IL-12-treated NKϪTϩ mice, which produced only basal levels specific Th1-dependent IgG2a and Th2-dependent IgG1 Abs on (Ͻ0.05 ng/ml; p Ͻ 0.05). days 30, 45, and 60 postinfection. On day 30 and thereafter, NKϩTϩ and BALB/c mice developed significant levels of Leish- mania-specific IgG1 and IgG2a, although BALB/c mice produced Discussion significantly more (Fig. 5A). On the other hand, L. major-infected The results presented here indicate that although NK1.1ϩCD3ϪTCR- NKϪTϩ mice displayed high titers of Leishmania-specific IgG1, ␣␤Ϫ (NK) cells play a role in innate immunity to L. major, they are but failed to produce any measurable quantities of Leishmania- not required for development of Th1-like response and control of specific IgG2a throughout the course of infection (Fig. 5B). Both L. major infection in resistant mice. Furthermore, they also dem- NKϩTϩ and NKϩTϩ(tg⑀26Y) mice produced significant titers of onstrate that in the absence of NK cells, IL-12 can directly induce LmAg-specific IgG2a (2666.2 Ϯ 533 and 16200 Ϯ 12049 in development of a Th1 response during L. major infection. The Journal of Immunology 6751

FIGURE 3. NKϪTϩ mice control cutaneous L. major infection. A, Course of L. major infection following infection with 2 ϫ 106 stationary-phase promastigotes in NKϪTϩ,NKϩTϩ, BALB/c, and tg⑀26 (NKϪTϪ) mice. Progress of lesion growth was monitored by measuring the increase in thickness of the infected footpad and comparing this to the thickness of the contralateral uninfected footpad. NKϪTϪ were sacrificed on day 30 after the infection, due to the development of large lesions. All other mice were monitored up to 60 days. B, Footpad parasite burdens in L. major-infected NKϩTϩ,NKϪTϩ, and NKϪTϪ mice. Data expressed as log parasite titer Ϯ SE. C–E, Hematoxylin-eosin stained skin lesions from L. major-infected NKϪTϪ,NKϩTϩ, and NKϪTϩ mice. Lesions from NKϪTϪ mice showed ulceration and extensive tissue destruction with inflammatory infiltrate comprising of parasitized macrophages, neutrophils, and eosinophils. C and D, Similarly stained skin from the inoculation sites of NKϩTϩ and NKϪTϩ mice displayed a more preserved skin structure with lymphocytes and some macrophages with few intracellular parasites (original magnification, ϫ40). Results are representative of three experiments with four to five animals per group. Data expressed as mean Ϯ SE.

NK cells have been demonstrated to be involved in the first line cells are not required for the control of cutaneous L. tropica in- of defense against viruses, bacteria, and parasites (7, 16). The im- fection, which supports our observations in the present study. portance of NK cells in early antibacterial immunity has been dem- These differences are probably due to the different experimental onstrated by a study showing that SCID mice that lack ␣␤ and ␥␦ models of NK cell deficiency used and the differences in the ex- T cells but have NK cells develop activated macrophages and par- perimental approaches. For example, although poly(I:C) activates tially control Listeria monocytogenes (17–19). Later, it was dem- NK cells and significantly reduces the parasite burdens in the early onstrated that spleen cells from naive SCID and nude mice pro- course of L. major infection in BALB/c mice (26), it also induces duced significant levels of IFN-␥ following incubation with heat- production of type I (IFN-␣/␤) IFN from NK cells, which has been killed L. monocytogenes (18). Furthermore, administration of shown to induce expression of nitric oxide synthase 2 (NOS2) in neutralizing anti-IFN-␥ mAb or NK cell depletion before infection vivo and regulate innate immunity to L. major (27). Conversely, abolished macrophage activation in SCID mice. Together, these depletion of NK cells prior to L. major infection using anti- data indicate that NK cell-derived IFN-␥ is involved in vivo mac- AsGM1 Ab significantly decreased early IFN-␥ production and rophage activation following L. monocytogenes infection (19). exacerbated the infection in resistant C3H/HeN, but had no effect Similarly, many studies have demonstrated that NK cells also play on the ultimate disease outcome (11, 26). a critical role in immunity against viruses such as murine CMV, Systemic depletion of cells using Ab treatment is efficient but Coxsackievirus B4, and influenza virus (20–22). This has been not absolute (28). Furthermore, in a recent study, NK cell depletion attributed to their cytolytic capacity and ability to produce type I using anti-NK1.1 as well as anti-AsGM-1 Abs failed to alter the IFNs that have antiviral activity (23). Th1/Th2 balance of Ag-driven cytokine synthesis (29). The ability Although some studies using NK cell-deficient beige mice had of NK-depleted C3H/HeN mice to heal L. major infection, how- suggested that NK cells were required for the control of visceral ever, could be attributed to the repopulation of NK cells in these leishmaniasis (24), others using SCID mice indicated that NK cell- mice following cessation of anti-AsGM-1 treatment (11, 26). Fi- derived IFN-␥ was unlikely to participate in the early regulation of nally, beige mice, which controlled L. tropica (30), exhibit normal visceral leishmaniasis caused by L. donovani (25). Similarly, some numbers of NK 1.1ϩCD3Ϫ cells and normal NK cell cytotoxicity studies in murine cutaneous leishmaniasis indicate that the NK against viruses, although they have very low cytotoxicity against cell-derived IFN-␥ plays an important role in early resistance and YAC-1 cells (31, 32). By our use of a novel model of specific development of a Th1 response following L. major infection (11, murine NK cell deficiency, these possibilities have been excluded. 26). Others, however, using beige mice have demonstrated that NK Unlike beige mice, the NKϪTϩ mice lack NK cell cytotoxicity 6752 NK CELL-DEFICIENT MICE CONTROL CUTANEOUS LEISHMANIASIS

FIGURE 5. L. major-speciﬁc IgG1 and IgG2a production in NKϩTϩ and NKϪTϩ mice on day 60 postinfection presented as reciprocal endpoint titers on a log scale. Similar results were observed in three independent experiments. NKϪTϪ mice were sacriﬁced on day 30 postinfection, due to the development of large lesions. ND, not detectable.

vivo (33). This mechanism, however, could not account for NK cell deficiency in the neonatal tg⑀26 model described in this study. First, in the neonatal tg⑀26 model, NK cell deficiency is independent of T cell development,3 and there was no overproduction of TNF-␣ in sera from L. major-infected NKϩTϩ and NKϪTϩ mice ϩ ϩ Ϫ ϩ FIGURE 4. NKϪTϩ mice develop a Th1-like response. In vitro LmAg- (97 Ϯ 28 pg/ml and 71 Ϯ 40 pg/ml in NK T and NK T mice, induced (20 ␮g/ml) IL-12 (A), IFN-␥ (B), and IL-4 (C) production by the respectively). Second, we demonstrated that the lack of cytotoxic ϩ ϩ Ϫ ϩ lymph node cells from NK T ,NK T , and BALB/c mice measured on NK cells in the neonatally transplanted tg⑀26 mice was due to the Ϫ Ϫ day 60 postinfection. NK T (tg⑀26) mice were sacrificed on day 30 failure of transplanted hematopoietic stem cells to home to bone postinfection for proliferation assays and cytokine production, due to the marrow, whereas the T cell development in the same mice was due Ϯ development of large lesions. Data expressed as mean SE. Asterisks to the migration of the hematopoietic stem cells during neonatal indicate statistically significant differences between NKϩTϩ and NKϪTϩ period to the thymus.3 mice. Similar results were observed in two independent experiments. ND, ␣␤Ϫ/Ϫ not detectable. We have previously demonstrated that TCR- mice lacking T cells on genetically resistant background (15) and RAG-2Ϫ/Ϫ mice that have innately high levels of IFN-␥ and IL-12 control against lymphocytic choriomeningitis virus and YAC-1 cells.3 early lesion growth of L. major but later succumb to the disease Therefore, these mice are truly deficient in NK cell lytic function, (our unpublished observations). Our findings in the present study and the number of NK1.1ϩCD3Ϫ cells are either absent or mark- that tg⑀26 mice (NKϪTϪ) are susceptible to L. major and develop edly diminished in these mice.3 We have reported previously that lesions significantly faster than similarly infected RAG2Ϫ/Ϫ or transplantation of wild-type bone marrow into adult tg⑀26 mice TCR-␣␤Ϫ/Ϫ mice (NKϩTϪ) suggest that NK cells may be in- results in aberrant NK cell development, which is caused by the volved in controlling early resistance to L. major. However, the generation of aberrant T cells that leads to high levels of TNF-␣ in ability of NKϪTϩ mice to control cutaneous L. major infection as The Journal of Immunology 6753

NK cells have been shown to induce Ab production from activated murine B cells (40) as well as resting human B cells (41). A recent study, however, has demonstrated that IL-12 enhances Ab responses and increase levels of IgG2a to T-independent polysaccharide vaccines in the absence of T and NK cells (42). Interest- ingly, despite the production of IL-12 and development of Th1-like response, L. major-infected NKϪTϩ mice failed to produce Th1- associated Leishmania-specific IgG2a Ab throughout the course of infection. Similarly, in our ongoing studies, NKϪTϩ produced significant amounts of IFN-␥ but displayed baseline levels of Ag- specific IgG2a following immunization with keyhole limpet he- mocyanin (KLH) or OVA in CFA, which induces Th1-biased response (our unpublished observations). Furthermore, there was no difference in levels of Th2-type cytokines IL-4 and IL-10 produced by Ag-stimulated spleen cells from these mice (our unpub- Ϫ ϩ FIGURE 6. Administration of IL-12 neutralizing Ab to NK T mice lished observations). This data has been reproduced in four inde- inhibits Th1 development and exacerbates cutaneous L. major infection. pendent studies conducted so far in our laboratory using KLH and Ϯ Data expressed as mean lesion size SE. OVA. These results suggest that NK cells may play a critical role in production of IgG2a against T-dependent Ags. efficiently as wild-type (NKϩTϩ) mice indicates that, in the pres- Recently, a population of CD4 T cells that express NK1.1 and ence of functional T cells, NK cells are not essential for the control TCR-␣␤ (NKT cells) has been shown to produce IL-4 (43) as well of L. major lesion growth. Furthermore, they also indicate that NK as IFN-␥ (44). Although early studies had indicated that NKT cells cell-derived IFN-␥ is not essential for the Th1 differentiation of may be the initial source of IL-4 that induces Th2 development CD4ϩ T cell following cutaneous L. major infection. (43), recent studies indicate that in a susceptible mouse, L. major Acquired protective immunity against cutaneous L. major in- induces rapid IL-4 production by CD4ϩ cells that are NK 1.1- fection in genetically resistant mice is dependent on the ability to negative (45). Recent studies have demonstrated that IL-12-stim- develop a CD4 Th1-type lymphocyte response and produce cyto- ulated NK1.1ϩ T cells produce high levels of IFN-␥ (46, 47), as kines such as IL-2 and IFN-␥ (2, 34). Many recent observations well as exhibit cytotoxicity against tumor cells (46). Furthermore, indicate that IL-12 plays a critical role in the development of a endogenous IL-12 has been shown to down-regulate IL-4-produc- Th1-like response in resistant mice following L. major infection ing NK 1.1ϩ T cells in liver and improve protective immunity (3, 4, 35). Thus, for example, IL-12 administered simultaneously at against listeriosis (48). NKϪTϩ mice used in the present study the time of vaccination with Leishmania Ag facilitated Th1 devel- have a normal number of NK1.1ϩCD4ϩ T cells. Additionally, ad- opment (35). Furthermore, in two independent studies, susceptible ministration of anti-IL-12 Ab L. major-infected NKϪTϩ mice in- BALB/c mice treated with IL-12 during L. major infection devel- hibited Th1 development and rendered them susceptible to infec- oped a significant Th1-like response and healed (3, 4). In contrast, tion. Therefore, one may speculate that NK 1.1ϩ T cell-derived administration of IL-12 neutralizing Ab to resistant C3H/HeN IFN-␥ initiates Th1 development in these mice. A recent study, mice rendered them susceptible to L. major (36). Similarly, genet- however, demonstrated that RAG-2/IFN-␥Ϫ/Ϫ (double mutant) ically resistant mice lacking the IL-12 gene failed to mount a Th1- mice reconstituted with the wild-type CD4ϩ NK 1.1Ϫ T cells de- like response and developed large nonhealing lesions following L. velop Th1 response and control cutaneous L. major infection (38). major infection (37). Taken together, these results indicate that neither NK cell-derived Neutralization of IL-12 in vivo has been shown to abrogate NK IFN-␥ nor NK 1.1ϩCD4ϩ T cells are critical for the development cell cytotoxicity and decrease early IFN-␥ production by lymph of protective Th1 response following L. major infection. node cells from L. major-infected C3H mice (36). This result sug- In conclusion, L. major-infected NKϪTϩ mice on genetically gested that the protective role of IL-12 was mediated by its ability resistant background develop an efficient Th1-like response as to activate NK cells to produce IFN-␥ critical for the subsequent measured by significant IFN-␥ production by the lymph node cells development of the Th1 subset of the CD4ϩ T cells (36). On the following stimulation with Leishmania Ag and control cutaneous other hand, a recent study demonstrated that IFN-␥ derived from L. major infection. Furthermore, administration of neutralizing anti- CD4ϩ T cells is sufficient to mediate Th1 development following IL-12 Abs to NKϪTϩ mice during L. major infection inhibits de- L. major infection (38). In the present study, we found that L. velopment of Th1-like response and enhances cutaneous lesion major-infected NKϪTϩ mice produced significantly higher growth. amounts of IL-12 and IFN-␥ than similarly infected NKϩTϩ mice and efficiently controlled the infection. Conversely, administration References Ϫ ϩ of anti-IL-12 neutralizing Abs to NK T mice during L. major 1. Peters, W., and R. Killick-Kendrick. 1987. The Leishmaniasis in Biology and infection significantly reduced IFN-␥ production and resulted in Medicine, Vols. 1 and 2. Academic Press, London, p. 941. 2. Heinzel, F. P., M. D., Sadick, S. S. Mutha, and R. M. Locksley. 1991. Production the development of large rapidly growing lesions. Possibly, the ␥ ϩ Ϫ ϩ of interferon , interleukin 2, interleukin 4, and interleukin 10 by CD4 T lym- increased IL-12 production in NK T mice is due to the devel- phocytes in vivo during healing and progressive murine leishmaniasis. Proc. opment of compensatory mechanisms or loss of NK cell-mediated Natl. Acad. Sci. USA 88:7011. 3. Heinzel, F. P., D. S. Schoenhaut, R. M. Reeko, L. E. Rosser, and M. K. Gately. negative feedback regulating IL-12 production by macrophages. In 1993. Recombinant interleukin 12 cures mice infected with Leishmania major. fact, recent studies have demonstrated that NK cell-derived type 1 J. Exp. Med. 179:447. IFN (IFN-␣/␤), which has been shown to regulate innate responses 4. Sypek, J. P., C. L. Chung, S. E. H. Mayor, J. M. Subramanyam, S. J. Goldman, D. S. Sieburth, S. F. Wolf, and R. G. Schaub. 1993. Resolution of cutaneous in L. major infection (27), can down-regulate IL-12 production leishmaniasis: interleukin 12 initiates a protective T helper type 1 immune re- (39). Nevertheless, taken together, these results suggest that in the sponse. J. Exp. Med. 177:1505. 5. Locksley, R. M., F. P. Heinzel, B. J. Holaday, S. S. Mutha, S. L. Reiner, and absence of NK cells, IL-12 can directly induce Th1 development M. D. Sadick. 1991. Induction of Th1 and Th2 CD4ϩ subsets during murine in vivo following L. major infection. Leishmania major infection. Res. Immunol. 142:28. 6754 NK CELL-DEFICIENT MICE CONTROL CUTANEOUS LEISHMANIASIS

6. Lanier, I., H. Spits, and J. H. Phillips. 1992. The developmental relationship 28. Fu, Y. X., C. E. Roark, K. Kelly, D. Drevets, P. Campbell, R. O’Brien, and W. between NK cells and T cells. Immunol. Today 13:392. Born 1994. Immune protection and control of inflammatory tissue necrosis by ␥␦ 7. Trinchieri, G. 1989. Biology of natural killer cells. Adv. Immunol. 47:187. T cells. J. Immunol. 144:713. 8. Dunn, P. L., and R. J. North. 1991. Early ␥ interferon production by natural killer 29. Wang, M., C. A. Ellison, J. G. Gartner, and K. T. HayGlass. 1998. Natural killer cells is important in defense against murine listeriosis. Infect. Immun. 59:1892. cell depletion fails to influence initial CD4 T cell development in vivo in exog- 9. Levitz, S. M., and M. P. Dupont. 1993. Phenotypic and functional characteriza- enous antigen-stimulated cytokine and antibody responses. J. Immunol. tion of human lymphocytes activated by interleukin-2 to directly inhibit growth 160:1098. of Cryptococcus neoformans in vitro. J. Clin. Invest. 91:1490. 30. Kirkpatrick, C. E., and J. P. Farrell. 1982. Leishmaniasis in Beige mice. Infect. 10. Murphy, J. W., M. R. Hidore, and S. C. Wong.1993. Direct interactions of human Immun. 38:1208. lymphocytes with the yeast-like organism, Cryptococcus neoformans. J. Clin. 31. McKinnon, K. P., A. H. Hale, and M. J. Ruebush. 1981. Elicitation of natural Invest. 91:1553. killer cells in beige mice by infection with vesicular stomatitis virus. Infect. 11. Scharton, T. M., and P. Scott. 1993. Natural killer cells are a source of interferon Immun. 32:204. ϩ ␥ that drives differentiation of CD4 T cell subsets and induces early resistance 32. Welsh, R. M. J., and R. W. Kiessling. 1980. Natural killer cell response to lym- to Leishmania major in mice. J. Exp. Med. 178:567. phocytic choriomeningitis virus in Beige mice. Scand. J. Immunol. 11:363. 12. Maasho, K., F. Sanchez, E. Schurr, A. Hailu, and H. Akuffo. 1998. Indications of 33. Wang, B., G. A. Hollander, A. Nichogiannopoulou, S. J. Simpson, J. S. Orange, the protective role natural killer cells in human cutaneous leishmaniasis in an area J. C. Gutierrez-Ramos, S. J. Burakoff, C. A. Biron, and C. Terhorst. 1996. Natural of endemicity. Infect. Immun. 66:2698. killer cell development is blocked in context of aberrant T lymphocyte ontogen. 13. Wang, B., C. Biron, J. She, K. Higgins, M. J. Sunshine, E. Lacy, N. Lonberg, and Int. Immunol. 8:939. C. Terhorst. 1994. A block in both early T lymphocyte and natural killer cell 34. Reed, S. G., and P. Scott. 1993. T-cell and cytokine responses in leishmaniasis. ⑀ development in transgenic mice with high-copy numbers of the human CD3 Curr. Opin. Immunol. 5:524. gene. Proc. Natl. Acad. Sci. USA 91: 9402. 35. Afonso, L. C. C., T. M. Scharton, L. Q. Vieira, M. Wysocka, G. Trinchieri, and 14. Hollander, G. A., B. Wang, A. Nichogiannopoulou, P. P. Platenburg, P. Scott. 1994. The adjuvant effect of interleukin-12 in a vaccine against Leish- W. van Ewijk, S. J. Burakoff, J. C. Gutierrez-Ramos, and C. Terhorst. 1995. mania major. Science 263:235. Developmental control point in induction of thymic cortex regulated by a sub- 36. Scharton-Kersten, T., L. C. C. Afonso, M. Wysocka, G. Trinchieri, and P. Scott. population of prothymocytes. Nature 373:350. ␥␦ 1995. IL-12 is required for natural killer cell activation and subsequent T helper 15. Satoskar, A, M. Okano, and J. R. David. 1997. T cells are not essential for 1 cell development in experimental leishmaniasis. J. Immunol. 154:5320. control of cutaneous Leishmania major infection in genetically resistant C57BL/6 37. Mattner, F., J. Magram, J. Ferrante, P. Launois, K. Di Padova, R. Behin, mice. J. Infect. Dis. 176:1649. M. K. Gately, J. A. Louis, and G. Alber. Genetically resistant mice lacking in- 16. Bancroft, G. J. 1993. The role of natural killer cells in innate resistance to interleukin-12 are susceptible to infection with Leishmania major and mount a fection. Curr. Opin. Immunol. 5:503. polarized Th2 cell response. Eur. J. Immunol. 26:1553. 17. Newborg, M. S., and R. J. North. 1980. On the mechanism of T cell independent 38. Wakil, A. E., Z. -E. Wang, J. C. Ryan, D. J. Fowell, and R. M. Locksley. 1998. anti-Listeria resistance in nude mice. J. Immunol. 124:571. ϩ Interferon ␥ derived from CD4 T cells is sufficient to mediate T helper cell type 18. Bancroft, G. J., R. D. Schreiber, G. C. Bosma, M. J. Bosma, and E. R. Unanue. 1 development. J. Exp. Med. 188:1651. 1987. A T-cell independent mechanism of macrophage activation by interferon-␥. 39. Cousens, L. P., J. S. Orange, H. C. Su, and C. A. Biron. 1997. Interferon-␣/␤ J. Immunol. 139:1104. inhibition of interleukin 12 and interferon-␥ production in vitro and endogenously 19. Bancroft, G. J., K. C. F. Sheehan, R. D. Schreiber, and E. R. Unanue. 1989. during viral infection. Proc. Natl. Acad. Sci. USA 94:634. Tumor necrosis factor is involved in the T cell independent pathway of macrophage activation in SCID mice. J. Immunol. 143:127. 40. Snapper, C. M., H. Yamaguchi, M. A. Moorman, R. Sneed, D. Smoot, and 20. Orange, J. S., B. Wang, C. Terhorst, and C. A. Biron. 1995. Requirement for J. J. Mond. 1993. Natural killer cells induce activated murine B cells to secrete natural killer cell-produced interferon ␥ in defense against murine cytomegalo- Ig. J. Immunol. 151:5251. virus infection and enhancement of this defense by interleukin 12 administration. 41. Dixon Gray, J., and D. A. Horwitz. 1995. Activated human NK cells can stim- J. Exp. Med. 182:1045. ulate resting B cells to secrete immunoglobulin. J. Immunol. 154:5656. 21. Vella, C., and H. Festenstein. 1992. Coxsackievirus B4 infection of the mouse 42. Buchanan, R. M., B. P. Arulanandam, and D. W. Metzger. 1998. IL-12 enhances pancreas: the role of natural killer cells in the control of virus replication and antibody responses to T-independent polysaccharide vaccines in the absence of T resistance to infection. J. Gen. Virol. 73:1379. and NK cells. J. Immunol. 161:5525. 43. Yoshimoto, T., A. Bendelac, C. Watson, L. J. Hu, and W. E. Paul. 1995. Role of 22. Karupiah, G., A. J. Ramsay, I. A. Ramshaw, and R. V. Bladen. 1992. Recom- ϩ binant vaccine vector induced protection of athymic, nude mice from Influenza A NK1.1 T cells in a Th2 response and immunoglobulin E production. Science virus infection: analysis of protective mechanisms. Scand. J. Immunol. 36:99. 270:1845. 23. Phillips, J. H., T. Horri, A. Nagler, N. Bhat, H. Spits, and L. Lanier. 1992. 44. Arase, N., H. Arase, S. Y. Park, H. Ohno, C. Ra, and T. Saito. 1996. Association with far ␥ is essential for activation signal through NKR-P1 (CD161) in natural Ontogeny of human natural killer (NK) cells: fetal NK cells mediate cytolytic ϩ function and express cytoplasmic CD3⑀, ␦ proteins. J. Exp. Med. 175:1055. killer (NK) cells and NK1.1 T cells. J. Exp. Med. 186:1957. 24. Kirkpatrick, C. E., J. P. Farrell, J. F. Warner, and G. Denner. 1985. Participation 45. Launois, P., T. Ohteki, K. Swihart, H. R. MacDonald, and J. A. Louis. 1995. In of natural killer cells in the recovery of mice from visceral leishmaniasis. Cell. susceptible mice, Leishmania major induce very rapid interleukin-4 production ϩ Ϫ Immunol. 92:163. by CD4 T cells which are NK 1.1 . Eur. J. Immunol. 25:3298. 25. Kaye, P. M., and G. J. Bancroft. 1992. Leishmania donovani infection in SCID 46. Leite-De-Moraes, M. C., G. Moreau, A. Arnold, F. Machavoine, C. Garcia, mice: lack of tissue response and in vivo macrophage activation correlates with M. Papiernik, and M. Dy. 1998. Interleukin-4 producing NKT cells are biased failure to trigger natural killer cell-derived ␥ interferon production in vitro. Infect. towards IFN-␥ production by IL-12: influence of the microenvironment on the Immun. 60:4335. functional capacities of NKT cells. Eur. J. Immunol. 28:1507. 26. Laskay, T., M. Rollinghoff, and W. Solbach. 1993. Natural killer cells participate 47. Kawamura, T., K. Takeda, S. K. Mendiratta, H. Kawamura, L. Van Kaer, in the early defense against Leishmania major in mice. Eur. J. Immunol. 23:2237. H. Yagita, T. Abo, and K. Okumura. 1998. Critical role of NK1ϩ T cells in 27. Diefenbach, A., H. Schindler, N. Donhauser, E. Lorenz, T. Laskay, IL-12-induced immune responses in vivo. J. Immunol. 160:16. J. MacMicking, M. Rollinghoff, I. Gresser, and C. Bogdan. 1998. Type 1 inter- 48. Emoto, Y., M. Yomoto, and S. H. E. Kaufmann. 1997. Transient control of IL-4 feron (IFN ␣/␤) and type 2 nitric oxide synthase regulate the innate immune producing natural killer T cells in the livers of Listeria monocytogenes-infected response to a protozoan parasite. Immunity 8:77. mice by interleukin-12. Infect. Immun. 65:5003.