Deletion of IL-4 Receptor Alpha on Dendritic Cells Renders BALB/c Mice Hypersusceptible to Leishmania major Infection

In BALB/c mice, susceptibility to infection with the intracellular parasite Leishmania major is driven largely by the development of T helper 2 (Th2) responses and the production of interleukin (IL)-4 and IL-13, which share a common receptor subunit, the IL-4 receptor alpha chain (IL-4Rα). While IL-4 is the main inducer of Th2 responses, paradoxically, it has been shown that exogenously administered IL-4 can promote dendritic cell (DC) IL-12 production and enhance Th1 development if given early during infection. To further investigate the relevance of biological quantities of IL-4 acting on DCs during in vivo infection, DC specific IL-4Rα deficient (CD11ccreIL-4Rα-/lox) BALB/c mice were generated by gene targeting and site-specific recombination using the cre/loxP system under control of the cd11c locus. DNA, protein, and functional characterization showed abrogated IL-4Rα expression on dendritic cells and alveolar macrophages in CD11ccreIL-4Rα-/lox mice. Following infection with L. major, CD11ccreIL-4Rα-/lox mice became hypersusceptible to disease, presenting earlier and increased footpad swelling, necrosis and parasite burdens, upregulated Th2 cytokine responses and increased type 2 antibody production as well as impaired classical activation of macrophages. Hypersusceptibility in CD11ccreIL-4Rα-/lox mice was accompanied by a striking increase in parasite burdens in peripheral organs such as the spleen, liver, and even the brain. DCs showed increased parasite loads in CD11ccreIL-4Rα-/lox mice and reduced iNOS production. IL-4Rα-deficient DCs produced reduced IL-12 but increased IL-10 due to impaired DC instruction, with increased mRNA expression of IL-23p19 and activin A, cytokines previously implicated in promoting Th2 responses. Together, these data demonstrate that abrogation of IL-4Rα signaling on DCs is severely detrimental to the host, leading to rapid disease progression, and increased survival of parasites in infected DCs due to reduced killing effector functions.


Introduction
Leishmania spp. are protozoan parasites that are transmitted by Phlebotomus spp. sandflies and can cause several forms of disease in humans, ranging from localized cutaneous lesions to visceral Leishmaniasis, where parasites invade internal organs such as the spleen and liver. The incidence of disease is approximately 1.5 million per annum for cutaneous Leishmaniasis, and 500 000 per annum for visceral Leishmaniasis, which is usually fatal if left untreated [1]. Currently there is no vaccine. To identify correlates of immune protection, which may aid in vaccine design and therapeutic strategies, experimental models of cutaneous Leishmaniasis have been established in which disease is induced by infecting mice subcutaneously with L. major. Susceptible BALB/c mice show progressive lesion development with dissemination of parasites to visceral organs, while resistant C57BL/6 mice are able to control infection and heal lesions [2][3][4]. Lack of healing in BALB/c mice is associated with a T helper (Th) 2 response characterized by secretion of interleukin (IL)-4, IL-5, IL-9 and IL-13 [3,[5][6][7][8], high anti-Leishmania antibody titres [8,9] and alternative activation of macrophages [9,10]. In contrast, resistant C57BL/6 mice develop protective Th1 responses with production of IL-12 and IFN-c, associated with classical activation of macrophages and killing of parasites by effector nitric oxide production [9,[11][12][13][14]. IL-4 and IL-13, both of which signal through a common receptor chain, the IL-4 receptor alpha (IL-4Ra) are known to be important susceptibility factors in L. major infection [3,6,8,15,16]. Both BALB/c and C57BL/6 mice secrete IL-4 early after infection however, production of IL-4 is sustained in susceptible BALB/c mice and transient in resistant C57BL/6 mice [17,18]. It appears that resistant mouse strains redirect the early Th2 response in an IL-12-dependent mechanism, while in susceptible mice the Th2 response persists and dominates the disease outcome by suppressing effector mechanisms needed for parasite killing [3].
While IL-4 is the primary inducer of Th2 responses [19], paradoxically it has also been shown that IL-4 promotes IL-12 production by bone marrow-derived dendritic cells (BMDCs) stimulated with CpG or LPS [20][21][22][23]. Furthermore, administration of 1 mg of recombinant IL-4 at 0 and 8 hours after infection with L. major led to increased IL-12 mRNA expression by dendritic cells (DCs) in vivo, promoted Th1 responses and rendered mice resistant to infection [21]. It has also been shown that global abrogation of IL-4Ra renders mice resistant to L. major only in the acute phase of infection, with mice continuing to develop necrotic footpad lesions during the chronic phase [15]. However, specific abrogation of IL-4Ra on CD4 + T cells does lead to resistance, indicating a protective role for IL-4Ra signalling on non-CD4 + T cells [24].
A candidate for this protective role may therefore be DCs. These sentinels of the immune system are specialized antigenpresenting cells, proficient at uptake of antigen, migration to the lymph nodes (LN) and activation of lymphocytes. Consequently, they play a critical role in the initiation and differentiation of the adaptive immune response [25,26]. To investigate the role of IL-4Ra signaling on DCs in resistance to Leishmania, CD11c cre IL-4Ra -/lox mice, deficient in IL-4Ra signaling on DCs, were generated and infected with L. major LV39 and IL81 strains. CD11c cre IL-4Ra -/lox mice were hypersusceptible to both strains of L. major, with increased footpad swelling and necrosis and substantially increased parasite burdens in peripheral organs, including the brain. Hypersusceptibility in CD11c cre IL-4Ra -/lox mice was associated with an upregulation of Th2 responses, impairment in iNOS production by macrophages and inflammatory DCs and increased parasite loads in LN and spleen DCs. Therefore, it is clear that IL-4Ra signaling has important effects on DC phenotype during cutaneous L. major infection, and is necessary to avoid rapid disease progression in the host. This study therefore expands our knowledge on the role of dendritic cells during cutaneous Leishmaniasis and on the effects of IL-4Ra signaling on dendritic cells.
To assess functional impairment of DCs in CD11c cre IL-4Ra -/lox mice, we generated bone marrow-derived dendritic cells and stimulated them with LPS in the presence or absence of IL-4 or IL-13. IL-4 is known to enhance DC production of IL-12 in an IL-4Ra dependent manner, so called ''IL-4 DC instruction'' [21][22][23]. As expected, BMDCs derived from IL-4Ra -/lox mice and BALB/c wildtype controls had significantly increased IL-12 production after the addition of IL-4 ( Figure 1E). In contrast, LPS/IL-4 stimulated BMDCs derived from CD11c cre IL-4Ra -/lox mice or from global IL-4Ra -/mice showed similar levels of IL-12 to those stimulated with LPS alone, with IL-4 having no effect. This demonstrates functional impairment of IL-4Ra signaling on DCs from CD11c cre IL-4Ra -/lox mice. In fact, after the addition of LPS alone, BMDCs with a functional IL-4Ra already showed a trend towards increased IL-12p40 levels, suggesting that endogenous levels of IL-4 found in the culture could influence these BMDCs. IL-13 did not increase levels of IL-12, confirming previous DC stimulation studies [22]. As previously reported [29], IL-4 and IL-13 had no significant effect on BMDC maturation, as shown by similar expression of MHCII, CD86, CD80, CD83 and CD40 (data not shown). Total yield of BMDCs per precursor cell seeded was similar in CD11c cre IL-4Ra -/lox mice and littermate controls and survival after maturation was not significantly different (data not shown).

Author Summary
Leishmaniasis is a parasitic infection caused by protozoan parasites of Leishmania species and is transmitted by the sandfly. Disease in humans ranges from localized cutaneous lesions to disseminated visceral Leishmaniasis. Mouse models of Leishmania major infection have demonstrated that a ''healing'' response in C57BL/6 mice requires the secretion of protective T helper (Th) 1 cytokines, including IFN-c, which mediates parasite killing by inducing nitric oxide production. Conversely, ''non-healer'' BALB/c mice are unable to control infection and develop a Th2 immune response characterized by the production of IL-4 and IL-13 cytokines. Although IL-4 is the main inducer of Th2 responses, it has been shown that IL-4 can instruct dendritic cell (DC)-derived IL-12 production and Th1 development if administered during DC activation. To further investigate the role of DCs, a DC specific IL-4Radeficient mouse model was established. L. major studies demonstrated hypersusceptibility to infection and strikingly increased parasite loads in peripheral organs of mice lacking IL-4Ra on DCs. Moreover, increased parasite burdens were observed in host cells, including DCs, which showed reduced killing effector functions. In summary, this study demonstrates that IL-4Ra-mediated instruction of DCs occurs in vivo and is necessary to avoid rapid progression of disease in the host.
during acute infection with both L. major strains (Figure 2A and 2D), which correlated with low parasite numbers in infected footpads ( Figure 2B and 2E) and draining lymph nodes ( Figure 2C and 2F). Susceptible WT BALB/c and IL-4Ra -/lox littermate control mice developed progressive footpad swelling after infection with both strains (Figure 2A and 2D), with increased parasite burdens in the infected footpads ( Figure 2B and 2E) and draining LN ( Figure 2C and 2F). Hemizygous (IL-4Ra -/lox mice) had slightly reduced footpad swelling compared to BALB/c mice in IL81 infection. The greater virulence of IL81 is reflected in more rapid disease progression, with footpad swelling and parasite burden reaching similar levels by 4 weeks to those obtained with LV39 in 8 weeks. Of importance, CD11c cre IL-4Ra -/lox mice were hypersusceptible to acute L. major infection compared to heterozygous littermate controls and BALB/c mice, showing considerably worsened disease progression when infected with either strain (Figure 2A and 2D), with earlier and dramatically larger footpad lesions, and development of early necrosis (Figure 2A and 2D). Increased disease progression was accompanied by significantly higher parasite numbers in the footpads ( Figure 2B and 2E) and LN ( Figure 2C and 2F) of infected animals. In addition, infection with a 10-fold lower dose of L. major LV39 also resulted in a hypersusceptible phenotype in CD11c cre IL-4Ra -/lox mice (Supplementary Figure S1 A-C). Histopathological analysis of CD11c cre IL-4Ra -/lox footpads at week 4 after infection with the virulent IL81 revealed severe destruction of epidermis, connective tissue and bone as a result of footpad necrosis, accompanied by increased inflammatory infiltrates and a high load of extracellular L. major amastigotes ( Figure 2G). In contrast, infected footpads of IL-4Ra -/lox revealed moderate dermal inflammatory infiltrates with mostly intact epidermis, connective tissue and bone. Together, these data reveal that IL-4Ra signaling on DCs play an important role in host protection against acute L. major infection.
A shift towards Th2 responses in CD11c cre IL-4Ra -/lox BALB/c mice Th1/Th2-type responses were investigated in CD11c cre IL-4Ra -/lox mice and controls during acute cutaneous leishmaniasis (IL81). Antigen-specific restimulation of CD4 + T cells sorted from the LN of infected mice and co-cultured with fixed antigenpresenting cells and soluble Leishmania antigen (SLA) revealed a significantly reduced IFN-c response in CD11c cre IL-4Ra -/lox mice in comparison to the resistant C57BL/6 or IL-4Ra -/strains as well as to the susceptible IL-4Ra -/lox littermate controls ( Figure 3A). Conversely, the levels of IL-4, IL-13 and IL-10 were significantly higher in CD11c cre IL-4Ra -/lox mice compared to IL-4Ra -/lox , IL-4Ra -/and C57BL/6 mice ( Figure 3B, 3C and 3D). The observed shift in cytokine responses was confirmed in LN cells, stimulated with anti-CD3 or SLA (data not shown) and systemically in the quality of Leishmania-specific antibody immune responses. Sera of week 4 infected mice revealed a predominant type 1 antibody response in IL-4Ra -/mice, as shown by elevated levels of Leishmania-specific IgG2a ( Figure 3E). In contrast, CD11c cre IL-4Ra -/lox mice displayed a predominant type 2 antibody response shown by marked production of IgG1 and total IgE, which was significantly higher than that observed in littermate IL-4Ra -/lox mice ( Figure 3F and 3G). A shift towards Th2-type responses also occurred in CD11c cre IL-4Ra -/lox mice in a 10-fold lower dose L. major LV39 infection (Supplementary Figure S1 D-H).
As IFN-c-induced nitric oxide synthase (iNOS) production by classically activated macrophages (caMphs) is a key control mechanism in L. major infection [14], the activation state of macrophages was determined in the infected footpad at week 4 after infection. Inflammatory macrophages (CD11b + MHCII + CD11c 2 ) from CD11c cre IL-4Ra -/lox mice had significantly reduced iNOS expression compared to those of littermate IL-4Ra -/lox control mice ( Figure 3H). Conversely, expression of arginase 1, a marker of alternatively activated macrophages (aaMphs), was higher in macrophages of CD11c cre IL-4Ra -/lox mice ( Figure 3I). This altered phenotype was confirmed in iNOS and arginase activity assays performed on total footpad cells stimulated with LPS ( Figure 3J and 3K). Together, these results demonstrate a shift towards Type 2 responses and reduced macrophage effector functions in CD11c cre IL-4Ra -/lox mice.
CD11c cre IL-4Ra -/lox mice have increased parasite loads in peripheral organs In L. major LV39 infection, parasites were present only in footpads and the draining lymph nodes at week 3, whereas by week 8 parasites had disseminated to the spleen and liver in both CD11c cre IL-4Ra -/lox mice and littermate controls ( Figure 4A and 4B). Parasite burdens were much higher in the organs of infected CD11c cre IL-4Ra -/lox mice, compared to littermate control mice. Moreover, in some CD11c cre IL-4Ra -/lox mice, but not in control mice, L. major parasites had disseminated as far as the brain by week 8 after infection ( Figure 4B). Similar disease progression was observed after infection with L. major IL81 ( Figure 4C), where CD11c cre IL-4Ra -/lox mice already displayed noticeable splenomegaly at 4 weeks post infection (data not shown), and had strikingly increased parasite burdens in all organs analyzed, including the brain ( Figure 4C). Histological analysis confirmed the increased presence of disseminated parasites in the spleen and liver of CD11c cre IL-4Ra -/lox mice (IL81, week 4), as shown by the high load of extracellular L. major amastigotes (spleen) and the prevalence of inflammatory foci and leishmanial bodies in mononuclear cells (liver) ( Figure 4D). The presence of parasites in brains of perfused CD11c cre IL-4Ra -/lox mice (IL81, week 4) was also confirmed by confocal microscopy ( Figure 4E). Parasites were not visible in the brains of littermate controls (data not shown). These results demonstrate a drastic increase in numbers of disseminated parasites in peripheral organs of infected CD11c cre IL-4Ra -/lox mice. Although it has been reported that dissemination could occur within hours after high-dose parasite inoculation [30], infection with GFP + L. major IL81 and analysis by flow cytometry demonstrated that GFP + parasites was not detectable in the spleen at 1 or 3 days post infection, whereas at BALB/c mice were intercrossed with CD11c cre expressing and IL-4Ra lox/lox mice to generate CD11c cre IL-4Ra -/lox BALB/c mice. (B) Genotyping of CD11c cre IL-4Ra -/lox mice. The deleted IL-4Ra PCR is 471 base pairs, loxP is 450 base pairs (floxed) or 356 base pairs (wildtype) and CD11c cre specific is 517 base pairs. (C) IL-4Ra surface expression was analyzed by flow cytometry from naïve IL-4Ra -/lox (solid line), IL-4Ra -/-(dashed line) and CD11c cre IL-4Ra -/lox (grey tinted) mice. DCs were CD11c + MHCII + (SiglecF 2 in lungs), alveolar macrophages were CD11c + SiglecF + , peritoneal macrophages were F480 + CD11b + , B cells were CD19 + CD3 2 and T cells were CD3 + CD19 2 . GM = geometric mean. (D) Genomic DNA was extracted from spleen DCs and IL-4Ra exon 8 (deleted in IL-4Ra deficient cells) was determined by RT-PCR and normalized to exon 5 (present in all cells). (E) Bone marrow-derived DCs were stimulated with LPS in the presence of absence of IL-4 or IL-13 and IL-12p40 was measured in the supernatants 48   In order to determine if dendritic cells could harbor L. major parasites, GFP-expressing L. major parasites (IL81) were used to track infected cell populations in different organs by flow cytometry at different time-points (day 3, day 7 and week 4) after infection. Parasite replication occurred in GFP + cell populations that were sorted and plated out for limiting dilution assays, indicating that GFP positivity was a good marker for viable parasites associated with cells (Supplementary Figure S3). At day 3 after GFP-L. major IL81 infection, plasmacytoid DCs (pDCs), macrophages and neutrophils had infiltrated the infected footpad ( Figure 5A). By 4 weeks post infection, numbers of infiltrating cells had increased substantially, with conventional DCs (cDCs) also now present in high numbers ( Figure 5B). The number of infiltrating cells was significantly higher in CD11c cre IL-4Ra -/lox mice compared to IL-4Ra -/lox mice ( Figure 5B). At the early time point in FP, macrophages were infected with GFP + Leishmania, with similar numbers in CD11c cre IL-4Ra -/lox mice and littermate controls ( Figure 5C). This was in contrast to the draining lymph node, where conventional and plasmacytoid DCs were infected, with higher numbers of DCs infected in CD11c cre IL-4Ra -/lox mice compared to controls ( Figure 5D). Similar results were obtained at day 7 post-infection (data not shown). At week 4 post infection, the footpad harbored a pool of infected cells, namely macrophages, cDCs and neutrophils ( Figure 5E), while in the draining lymph node, the cDCs were still infected compared to the other cell types ( Figure 5F). Again the number of infected DCs was significantly higher in CD11c cre IL-4Ra -/lox mice ( Figure 5E and 5F) compared to littermate controls. However, overall numbers of DCs infiltrating the LN at week 4 after L.
major IL81 infection were similar in both CD11c cre IL-4Ra -/lox mice and littermate control mice (data not shown), suggesting that differences in parasite killing and not DC migration were responsible for the increased number of infected DCs in CD11c cre IL-4Ra -/lox mice. Total cells were isolated from footpads at week 4 after infection, surface-stained for CD11b high MHCII high CD11c 2 macrophages followed by intracellular staining for iNOS (H) and arginase 1 (I). GM = geometric means. (J-K) Production of NO and arginase 1 in total footpad cells. Total cells were isolated from footpads at week 4 after infection and stimulated with 10 ng/ml LPS for 72 h. Production of NO was determined in cell supernatants (J) and cell lysates were assayed for arginase 1 production (K). A representative of two individual experiments is shown with mean values 6SEM. Statistical analysis was performed defining differences to IL-4Ra -/lox mice as significant (*, p#0.05, **, p#0.01; ***, p#0.001). doi:10.1371/journal.ppat.1003699.g003 Infected DCs were also found in the spleen, with significantly increased numbers of infected cells in CD11c cre IL-4Ra -/lox mice compared to controls ( Figure 6A). However, overall numbers of DCs infiltrating the spleen were also increased to a similar degree in both CD11c cre IL-4Ra -/lox mice and littermate controls at week 4 (data not shown), again suggesting that differences in parasite . Impairment of IL-4Ra signaling in vivo results in increased L. major parasite loads in peripheral organs. CD11c cre IL-4Ra -/lox and littermate mice were infected subcutaneously with 2610 6 stationary phase metacyclic L. major (L. m LV39) promastigotes into the hind footpad. Parasite load was determined by limiting dilution assay (LDA) of single-cell suspensions from homogenized footpad, lymph node, spleen, liver and brain at week 3 (A) and week 8 (B) after infection. Similarly, organs were harvested from mice infected with GFP-expressing L. major (L. m IL81) at week 4 after infection for limiting dilution assay (C). At the same time point, histopathology was analysed using formalin-fixed spleen and liver (D) stained with H&E and Giemsa, respectively (original magnification 6100; asterisks indicate inflammatory foci and insets, arrows indicate amastigote parasites 6800). Frozen brain sections (E) were stained with Hoechst nuclear stain (blue) and visualized by confocal microscopy for the presence of GFP-L. major amastigote parasites (original magnification 6400). A representative of two individual experiments is shown with mean values 6SEM. Statistical analysis was performed defining differences to IL-4Ra -/lox mice (*, p#0.05, **, p#0.01, ***, p#0.001). doi:10.1371/journal.ppat.1003699.g004 IL-4Ra Deficient DCs in Cutaneous Leishmaniasis PLOS Pathogens | www.plospathogens.org killing and not DC migration were responsible for the increased parasite loads in CD11c cre IL-4Ra -/lox mice. Although it is well known that iNOS-mediated NO production in classically-activated macrophages drives intracellular killing of L. major parasites, a recent study has now implicated a population of iNOS +producing inflammatory DCs in controlling Leishmania infection [31]. We therefore examined iNOS production by DCs in CD11c cre IL-4Ra -/lox and littermate control mice using intracellular FACS. In hypersusceptible CD11c cre IL-4Ra -/lox mice, a significantly reduced percentage of CD11c high MHCII high DCs produced iNOS compared to DCs from IL-4Ra -/lox littermate control mice ( Figure 6B). This was confirmed at the level of intracellular NO expression, which was also reduced in DCs from CD11c cre IL-4Ra -/lox mice ( Figure 6C). Together, these data demonstrate that DCs from CD11c cre IL-4Ra -/lox mice have reduced NO killing effector functions, further explaining the increased parasite burdens in the DCs of these mice.

IL-4Ra-deficient DCs have impaired DC instruction during infection in vivo
Previous studies using BMDCs found that IL-4-mediated instruction results in reduced IL-10 production that is responsible for increased IL-12p40 production by DCs upon stimulation with IL-4 plus CpG or LPS [21,23]. To test whether endogenous amounts of IL-4 could mediate DC instruction in vivo, CD11c cre IL-4Ra -/lox mice and controls were infected with L. major IL81. At 4 weeks post infection, total LN cells were restimulated with SLA and cytokines were measured in the supernatant. Lymph node cells from infected CD11c cre IL-4Ra -/lox mice produced significantly reduced IL-12p40 but increased IL-10 compared to littermate IL-4Ra -/lox mice ( Figure 7A). Moreover, intracellular cytokine staining revealed that DCs from CD11c cre IL-4Ra -/lox mice produced less IL-12p40 and more IL-10 than those from littermate IL-4Ra -/lox controls ( Figure 7B and Figure S4). Quantification of mRNA found decreased expression of the Th1-promoting cytokine genes for IL-12p40 ( Figure 7C) and IL-18 ( Figure 7D) in sorted LN DCs from CD11c cre IL-4Ra -/lox mice compared to controls. In contrast, there was a trend towards increased mRNA expression of IL-10 as well as significantly increased mRNA expression of IL-23 and activin A, cytokines which are involved in inducing Th2 responses by promoting Th17 and alternative activation of macrophages, respectively [32,33] ( Figure 7E-G). In addition, differences in IL-12p70 production were detected in vitro. L. major/IL-4 stimulated BMDCs derived from IL-4Ra -/lox mice showed increased IL-12p70 production, whereas IL-4 had no additive effect on IL-12p70 production in BMDCs from CD11c cre IL-4Ra -/lox mice ( Figure 7H). IL-13 did not increase IL-12p70 production, as previously shown [22].

Discussion
Understanding mechanisms of immune control in cutaneous Leishmaniasis is critical for the design of effective therapeutics and vaccines. Although several studies have clearly established that IL-4 is a key cytokine in the development of non-healing disease in BALB/c mice [8,19,34,35], apparently contradictory evidence also suggests that IL-4 has the ability to instruct protective Th1 responses [21,[36][37][38][39][40][41]. The term ''instruction theory'' was coined when IL-4 was found to promote increased production of IL-12 by BMDCs [20][21][22]. IL-4, but not IL-13, enhances the production of IL-12 induced by pathogen products via signalling through the type 1 IL-4 receptor [21,22]. The mechanism behind instruction was found to be inhibition of IL-10 by IL-4, leading to higher levels of IL-12 and increased protective Th1 responses [23]. Several studies also indicate that IL-4 and IL-13 may play a role in promoting DC maturation [22,42]. However, most in vitro and in vivo studies on the effects of IL-4 and IL-13 on DCs have been conducted with exogenously administered IL-4 or IL-13, and thus the relevance of biological quantities of IL-4 signalling through IL-4Ra on DCs during disease in vivo has not been demonstrated. To address these issues, dendritic cell-specific (CD11c cre IL-4Ra -/lox ) BALB/c mice were generated using the cre/loxP recombinase system under control of the cd11c locus. These mice were found to have abrogated IL-4Ra expression on DCs and alveolar macrophages, with other cell types maintaining IL-4Ra expression and functioning.
Infection of CD11c cre IL-4Ra -/lox mice with L. major LV39 and IL81 revealed IL-4Ra signaling on DCs to be highly important in protection against cutaneous Leishmaniasis. Compared to IL-4Ra -/lox littermate controls, CD11c cre IL-4Ra -/lox mice showed dramatically worsened disease progression, with increased footpad swelling and necrosis, and significantly higher parasite burdens both locally and in visceral organs such as the spleen and liver. As expected, genetically resistant C57BL/6 mice effectively controlled infection, as did global IL-4Ra -/mice, which have been shown to be resistant during the acute phase of L. major infection, with disease progression in the chronic phase only [15,24]. Progressive disease during L. major infection in BALB/c mice has been attributed to the predominance of Th2 cytokines and type 2 antibody immune responses [8,9,11], with a previous study by our laboratory showing that CD4 + T cell specific IL-4Ra deficient mice were highly resistant to L. major infection [24]. Analysis of CD4 + T cell cytokine responses in CD11c cre IL-4Ra -/lox mice revealed a decrease in IFN-c accompanied by a marked increase in IL-4, IL-13 and IL-10, while increased secretion of IgG1 and IgE by B cells confirmed a shift towards a Th2-type immune phenotype. Aside from its role in instruction, IL-10 is known to be a susceptibility factor for L. major infection, being produced at higher levels in susceptible BALB/c mice and capable of suppressing Th1-mediated effector functions [3,43]. In humans, IL-10 is strongly associated with persistent infection [44].
IFN-c plays an important role in mediating protective immunity during L. major infection by classically-activating macrophages to induce nitric oxide synthase-mediated NO production for intracellular killing of parasites [9,14,45,46]. Latent Leishmaniasis is reactivated in chronically infected healthy C57BL/6 mice by inhibition of endogenous NOS-2, indicating that iNOS expression is crucial for the sustained control of L. major infection [9,31,47]. Induction of iNOS-mediated NO production is counter-regulated by IL-4/IL-13 and IL-4Ra, which promote the development of alternatively activated macrophages and arginase 1 production through depletion of L-arginine as a substrate for iNOS. Interestingly, IL-10 has also been shown to suppress intracellular killing of pathogens in macrophages by suppressing IFN-c responses [48][49][50] and can induce an alternatively activated macrophage type phenotype in the absence of IL-4 and IL-13 [51]. Parasites such as Leishmania can utilize polyamines generated by arginase 1 activity for their own growth, making alternatively activated macrophages a favorable environment for their survival [52][53][54]. Accumulating reports have demonstrated a role for alternative macrophage activation and arginase 1 expression in influencing susceptibility to L. major infection [7,9,55,56]. LysM cre IL-4Ra -/lox mice which lack IL-4/IL-13 induced alternative activation of macrophages were found to have increased resistance to infection [9], while neutralization of endogenous arginase 1 with N-hydroxy-nor-L-arginine leads to complete healing in BALB/c mice [55].
Macrophages from the footpads of CD11c cre IL-4Ra -/lox mice were found to have reduced iNOS expression and increased arginase 1 expression compared to those from littermate control IL-4Ra -/lox mice, demonstrating a shift in macrophage effector function most likely as a consequence of increased IL-4, IL-13 and IL-10. Recently it has been shown that DCs can also become alternatively activated by upregulating markers such as Ym-1 and RELM-a after administration of IL-4 [29]. In our study, the data suggest that IL-4Ra-independent alternative activation of DCs is also possible, as DCs from CD11c cre IL-4Ra -/lox had decreased iNOS expression, possibly a consequence of reduced IFN-c and/ or increased IL-10 and activin A, and had higher parasite loads than those from littermate controls. Previous studies have revealed that iNOS-producing DCs constitute a major Th1-regulated effector cell population and contribute to resistance to infection by L. major [31], L. monocytogenes [57] and Brucella spp. [58]. The reduced ability of both macrophages and DCs to initiate NOmediated killing of L. major in CD11c cre IL-4Ra -/lox mice is therefore likely to play a role in the uncontrolled parasite replication observed both in the footpad and at peripheral sites.
In susceptible BALB/c mice, L. major parasites can disseminate within 24 hours from the site of infection in the footpad to the popliteal lymph nodes, spleen, liver, lungs and bone marrow [30,59]. However, L. major parasites were not detected at early time points during IL81 infection (day 1 and day 3) but were detected at week 4, and were also detected at week 8 but not at week 3 during LV39 infection, suggesting that parasite dissemination may have occurred at a later stage of infection. Dissemination is inhibited by the administration of recombinant IL-12 and resistant mouse strains restrict the spread of the parasites [30]. While several susceptible mouse strains have been reported to show some increase in dissemination [60][61][62], disseminated parasite loads in CD11c cre IL-4Ra -/lox mice were unusually dramatic, with relatively higher parasite burdens in the spleens and footpads compared to other susceptible strains. Unexpectedly, parasites were even identified within the brain of some of the CD11c cre IL-4Ra -/lox mice. This suggests that the L. major parasites managed to cross the immunological blood-brain barrier, which has only rarely been reported for this cutaneous strain with very low levels of parasites detected [63]. However, dissemination of parasites to the central nervous system (CNS) has been frequently observed in visceral Leishmaniasis in both humans and dogs [64][65][66][67]. It has been suggested that parasites arrive in the CNS via infected leukocytes [65] and/or disruption to the blood brain barrier caused by inflammation [67]. Studying the mechanisms by which other pathogens, such as bacteria, invade the CNS may lend insights into Leishmania dissemination. Many intracellular organisms such as Mycobacterium tuberculosis, Listeria monocytogenes, Brucella spp. and Salmonella spp. appear to make use of the ''Trojan-horse'' mechanism, using phagocyte facilitated invasion for entry into the CNS [68]. After infection with intracellular pathogens, phagocytes undergo phenotypical changes, such as increased migratory activity and increased expression of adhesion molecules and proinflammatory cytokines, all of which could aid in dissemination and crossing of the blood-brain barrier [68,69]. Whether infected phagocytes are recruited to the CNS by specific or non-specific means is unknown [69]. In order to determine which cells were infected by L. major, mice were infected with GFP-IL81 parasites and cell populations containing parasites were identified by flow cytometry.
At day 3 and 7 after infection, macrophages harbored L. major in the footpad, while pDCs and cDCs were found to be infected in the lymph node. Similar to other reports, this indicates that DCs were responsible for transporting parasites to the lymph node [70]. At week 4, L. major parasites were still detected in macrophages in the footpads, as well as in DCs and neutrophils, but in the LN they were primarily found in DCs. The number of infected DCs in both footpad and LN was significantly higher in CD11c cre IL-4Ra -/lox mice. A previous study also reported that DCs were the primary infected cell population in the draining LN of L. major infected mice [70]. DCs were also infected with L. major parasites in the spleen, with CD11c cre IL-4Ra -/lox mice again showing a greater number of infected DCs. Numbers of DCs infiltrating the LN and spleen were equivalent in both CD11c cre IL-4Ra -/lox mice and littermate controls during infection. This suggests that the increased survival and/or growth of parasites in DCs, as a consequence of significantly reduced DC iNOS production, was responsible for the increase in infected cell numbers in CD11c cre IL-4Ra -/lox mice. Interestingly, a recent study found that infected DCs, which are monocyte-derived CD11b + inflammatory DCs expressing Ly6C, F480, Ly6G and iNOS, showed a unique ability to disseminate to peripheral sites in M. tuberculosis infection [71]. Furthermore, CD11b + Ly6C + cells were found to be the principal phagocytic cells harboring L. monocytogenes in circulation [69,72]. We hypothesize that dendritic cells may therefore play a role in disseminating L. major parasites to peripheral sites and that their killing effector responses could be important in controlling disease.
The reduced Th1 and increased Th2 responses in CD11c cre IL-4Ra -/lox mice suggests that instruction theory is relevant in vivo, and more importantly, that biological quantities of IL-4 acting through DCs can promote resistance to Leishmania infection. DCs from lymph nodes of CD11c cre IL-4Ra -/lox mice produced more IL-10 and less IL-12 than those from IL-4Ra -/lox mice. Quantification of mRNA expression also revealed interesting differences in DCs from CD11c cre IL-4Ra -/lox mice. Expression of the Th1-promoting genes for IL-12p40 and IL-18 was decreased compared to DCs from littermate control mice, while expression of the Th2-promoting genes for IL-23p19 and activin A were significantly increased. IL-23 production by DCs has been shown to promote Th17 [32], leading to increased neutrophils that enhance susceptibility to L. major by acting as Trojan horses [73]. Activin A is a pleiotropic cytokine belonging to the TGF-beta superfamily, and has previously been found to promote alternative activation of macrophages by inducing Arginase 1 and decreasing IFN-c-induced expression of iNOS [33]. The absence of IL-4Ra signalling on DCs therefore appears to have a more complex influence on the dendritic cells than just affecting IL-12 production during cutaneous Leishmaniasis in vivo.
Dendritic cell instruction may not be restricted to Leishmaniasis, since other disease models have also demonstrated a protective role for IL-4. Experimental infections with Candida albicans in IL-4 deficient mice led to impaired development of Th1 responses [38], and a Th1 promoting effect of IL-4 has also been observed in autoimmunity [36,40,74], tumor immunity [39,75,76] and contact sensitivity reactions [41,77]. There is also evidence to suggest that IL-4 may promote Th1 development in humans, since both human and mouse DCs produce increased levels of bioactive IL-12 after stimulation with IL-4 [20]. A similar effect was observed in human peripheral blood mononuclear cells (PBMCs) treated with IL-4 plus lipopolysaccharide or Staphylococcus aureus [78]. Incorporating exogenous IL-4 as an adjuvant for enhancing strong Th1 responses could therefore be utilised to boost vaccine efficiency against cutaneous Leishmaniasis. Accordingly, parallel studies examining the efficacy of IL-4 as an adjuvant during BMDC-mediated vaccination against L. major, found that IL-4 instruction of DCs was critical in eliciting protective immune responses [79]. The role of IL-4Ra signalling on DCs in eliciting immunity to other intracellular pathogens is therefore of interest to vaccination strategies, and an exciting avenue to be explored.

Materials and Methods
Generation and genotyping of CD11c cre IL-4Ra -/lox BALB/c mice CD11c cre mice [27] were crossed with IL-4Ra lox/lox BALB/c mice [28] and complete IL-4Ra -/-BALB/c mice [15] to generate hemizygous CD11c cre IL-4Ra -/lox mice. Mice were backcrossed to a BALB/c background for 9 generations to generate CD11c cre IL-4Ra -/lox BALB/c mice. Hemizygous littermate controls (IL-4Ra -/lox ) were used as controls in all experiments. Mice were genotyped as described previously [28]. All mice were housed in specific-pathogen free barrier conditions in individually ventilated cages. Experimental mice were age and sex matched and used between 8-12 weeks of age.

Ethics statement
This study was performed in strict accordance with the recommendations of the South African national guidelines and University of Cape Town of practice for laboratory animal procedures. All mouse experiments were performed according to protocols approved by the Animal Research Ethics Committee of the Health Sciences Faculty, University of Cape Town (Permit Number: 009/042). All efforts were made to minimize suffering of the animals.

Analysis of IL-4Ra deletion efficiency
Genomic DNA was isolated from spleen DCs (CD11c + MHCII + ) sorted using a FACS Vantage flow cytometer (BD Immunocytometry systems). Purity was determined by flow cytometry and checked by cytospin and staining with the Rapidiff Stain set (Clinical Diagnostics CC, Southdale, South Africa) and was at least 99%. A standard curve was prepared from serial 10fold DNA dilutions of cloned IL-4Ra exon 5 and exon 8 DNA and RT-PCR was performed using the following primers; exon 5: forward 59 AACCTGGGAAGTTGTG 39 and reverse 59 CA CAGTTCCATCTGGTAT 39, exon 8: forward 59 GTA-CAGCGCACATTGTTTTT 39 and reverse 59 CTCGGCGCA CTGACCCATCT 39.

ELISAs
Cytokines in cell supernatants were measured by sandwich ELISA as previously described [15]. For antibody ELISAs, blood was collected in serum separator tubes (BD Bioscience, San Diego, CA). Antigen-specific IgG1, IgG2a and IgG2b were quantified by ELISA, as previously described [15]. Detection limits were 5 ng/ ml for IgG1 and IgG2b and 0.1 ng/ml for IgG2a and IgG3. Total IgE was determined as described [15]. The detection limits was 8 ng/ml for total IgE.

Histology
Footpads, spleens and livers were fixed in 4% formaldehyde in phosphate buffered saline and embedded in wax. Tissue sections were stained with either haemotoxylin and eosin or Giemsa.

Immunohistochemistry
Following infection of mice with GFP-L. major IL81 parasites for 4 weeks, isolated brain tissue was immediately embedded in OCT (Tissue-Tek; Sakura, Zoeterwoude, Netherlands) medium. Prefixing of tissues was avoided to minimize background staining from the fixative. OCT-embedded brain tissue were cut into 10 mm frozen sections and mounted on 3-aminopropyltriethoxysilanecoated slides. Following acetone fixation of tissue, sections were stained with nuclear stain Hoechst. Coverslips were then mounted on sections using Mowiol 4-88 mounting medium (Calbiochem) with anti-fade (Sigma). Images were acquired and analyzed by Ziess LSM 510 confocal microscope (Jena, Germany).

Detection of viable parasite burden
Infected organ and tissue cell suspensions were cultured in Schneider's culture medium (Sigma). Prior to removal of mouse brain tissue for detection of parasite burden, animals were perfused with 20 ml sterile saline solution. Detection of viable parasite burden was estimated by two-fold limiting dilution assay as previously described [15].

Antigen-specific restimulation
CD4 + T cells were positively selected using anti-CD4 MACS beads (Miltenyi Biotec) according to the manufacturer's instructions (purity .95%). Thy1.2-labeled splenocytes were T cell depleted by complement-mediated lysis to enrich antigen presenting cells (APCs). APCs were fixed with mitomycin C (50 mg/ml, 20 min at 37uC) and washed extensively in complete IMDM. A total of 2610 5 purified CD4 + T cells and 1610 5 APCs were cultured with SLA (50 mg/ml). After 72 h incubation at 37uC, supernatants were collected and cytokine production analysed as previously described [28].

Isolation of footpad and spleen cells
Muscle tissue was separated from infected footpads and digested in DMEM medium supplemented with Collagenase IV IL-4Ra Deficient DCs in Cutaneous Leishmaniasis (Sigman-Aldrich; 1 mg/ml) and DNase I (Sigma-Aldrich; 1 mg/ ml) at 37uC for 60 min. Following incubation, single cell suspensions were isolated by straining through 40 mM cellstrainers. Spleen cells were isolated by pressing through 70 mM cell-strainers, red blood cell lysis was performed and white blood cells were washed and resuspended in 10% DMEM (Gibco).

Cell sorting
Total lymph node or footpad cells were labeled with specific mAbs and populations isolated by cell sorting on a FACS Vantage machine. Macrophages from the footpad were gated as CD11b high MHCII high CD11c 2 cells and DCs, macrophages, neutrophils and B cells from the lymph node were gated as CD11c high MHCII high , CD11b high MHCII high CD11c 2 , GR-1 high -SSC high FSC high CD11c 2 and CD19 + CD3 2 CD11c 2 cells, respectively. Cells were .98% pure and used for further analysis.

Quantitative RT-PCR
Dendritic cells were stained with specific mAb and sorted from the LN of infected mice. Total RNA was extracted from dendritic cells using Tri reagent (Applied Biosystems, Carlsbad, Calif) and mini-elute columns (Qiagen) according to the manufacturer's protocol. cDNA was synthesized with Transcriptor First Strand cDNA synthesis kit (Roche), and real-time PCR was performed by

Nitric oxide synthase and arginase
Lymph node and footpad cells collected at week 4 after infection were restimulated with LPS (Sigma-Aldrich; 10 ng/ml). Supernatants were collected at 48 hours for quantification of nitric oxide [81] while arginase activity was measured in cell lysates [81]. Expression of intracellular iNOS and arginase was analyzed in CD11b high MHCII high CD11c 2 macrophages and CD11c high MH-CII high DCs by flow cytometry using rabbit anti-mouse iNOS (Abcam) with goat anti-rabbit PE (Abcam) and goat anti-mouse arginase (Santa Cruz Biotechnology) with donkey anti-goat PE (Abcam). Purified goat IgG and rabbit IgG were used as controls.

Statistics
Data is given as mean 6 SEM. Statistical analysis was performed using the unpaired Student's t test or 1-way Anova with Bonferroni's post test, defining differences to IL-4Ra -/lox mice as significant (*, p#0.05; **, p#0.01; ***, p#0.001) unless otherwise stated. (Prism software: http://www.prism-software. com). Figure S1 CD11c cre IL-4Ra -/lox mice are hypersusceptible to a 10-fold lower dose infection with L. major. Mice were infected with L. major LV39 (MRHO/SV/59/P) parasite strain. Footpad swelling was measured at weekly intervals in mice (5 per group) infected subcutaneously with a 10-fold lower dose of 2610 5 stationary phase metacyclic L. major promastigotes into the hind footpad (A). ''N'' indicates necrosis or ulceration/mouse. Parasite burden was determined by limiting dilution of single-cell suspensions from homogenized footpads (B) and lymph nodes (C) at week 8 after infection. (D-F) Antigen-specific IgG2a (D), IgG1 (E) and total IgE (F) antibody production was quantified in sera by ELISA at week 8 post infection. (G-H) Total cells from the draining lymph node were incubated for 72 hrs with medium, aCD3 or soluble Leishmania antigen (SLA). The production of IFNc (G) and IL-4 (H) was determined by ELISA. A representative of two individual experiments is shown with mean values 6SEM. Statistical analysis was performed defining differences to IL-4Ra -/lox mice as significant (*, p#0.05, **, p#0.01; ***, p#0.001).

Supporting Information
(TIF) Figure S2 Infiltration of GFP + -L. major parasites in immune cell populations in spleen during infection in CD11c cre IL-4Ra -/lox mice. Mice were infected subcutaneously with 2610 6 stationary phase metacyclic GFP-expressing L. major IL81 (MRHO/SV/59/P) strain into the hind footpad. At Day 0, Day 1, Day 3 and Week 4 after infection, total spleen cells were surface stained for dendritic cells (DCs-CD11c high MHCII high ), Macrophages (Mph-CD11b high MHCII high CD11c 2 ) and neutrophils (Neut-GR1 high CD11c 2 ). The percentage of infiltrating GFP + -infected cells were determined by flow cytometry. (TIF) Figure S3 Viability of GFP + -L. major in immune cell populations during acute L. major IL81 infection by limiting dilution assay. Experimental mice were infected subcutaneously with 2610 6 stationary phase metacyclic GFPexpressing L. major IL81 promastigotes into the hind footpad. At week 4 after infection, total lymph node cells were isolated and DCs (CD11c high MHCII high ), macrophages (CD11b high MHCII high CD11c 2 ), neutrophils (GR1 high CD11c 2 ) and B cells (CD19 + CD3 2 CD11c 2 ) were isolated by cell sorting on a FACS Vantage machine. Sorted cells were plated to determine viable parasite burden by limiting dilution assay in two-fold dilutions. (TIF) Figure S4 Intracellular IL-12 and IL-10 in lymph node DCs. Experimental mice were infected subcutaneously with 2610 6 stationary phase metacyclic L. major IL81 promastigotes into the hind footpad. Total lymph node cells were incubated with PMA/Ionomycin/Monensin for 4 h at 37uC, then surface-stained for CD11c high MHCII high DCs followed by intracellular FACS staining for IL-12 and IL-10. Dot plots of percent cytokine producing cells are shown. (TIF)