|Home | About | Journals | Submit | Contact Us | Français|
Nucleotide-binding domain leucine-rich repeat containing receptors (NLRs) are cytosolic receptors that initiate immune responses to sterile and infectious insults to the host. Studies have demonstrated that Nlrp3 is critical for the control of Candida albicans infections and in the generation of anti-fungal Th17 responses. Here we show that the NLR family member Nlrp10 also plays a unique role in the control of disseminated C. albicans infection in vivo. Nlrp10-deficient mice had increased susceptibility to disseminated candidiasis as indicated by decreased survival and increased fungal burdens. In contrast to Nlrp3, Nlrp10-deficiency did not affect innate proinflammatory cytokine production from macrophages and dendritic cells challenged with C. albicans. However, Nlrp10-deficient mice displayed a profound defect in Candida-specific Th1 and Th17 responses. These results demonstrate a novel role for Nlrp10 in the generation of adaptive immune responses to fungal infection.
Members of the NLRP subfamily contain a central nucleotide-binding domain (NACHT), an N-terminal pyrin domain (PYD) and C-terminal leucine rich repeats (LRR) thought to function in ligand sensing (1). Recently we, and others, have shown that the NLR family member Nlrp3 plays an important role in host defense against C. albicans through triggering the assembly and activation of the Nlrp3 inflammasome (2-4). Nlrc4 also functions within the mucosal stroma to control oral C. albicans infections (5). However, other than Nlrp3 and Nlrc4, the role of NLR family members in fungal pathogenesis remains unknown. Of interest, one NLR family member, Nlrp10, lacks the C-terminal LRR domain and has therefore been hypothesized to function as a negative regulator of inflammasome activation (6, 7).
In this study we demonstrate that Nlrp10, unlike Nlrp6, Nlrp12 and Nlrc4, is required for control of a disseminated C. albicans infection in vivo. We also show that in contrast to Nlrp3, the absence of Nlrp10 in macrophages (Mϕ) and dendritic cells (DC) does not affect inflammasome activation in response to C. albicans or other inflammasome activators. A recent study has also demonstrated that Nlrp10-deficient DCs have defective migration (8); here we demonstrate that despite normal inflammasome activation, Nlrp10-deficient mice display a profound defect in the generation of Candida-specific Th1 and Th17 responses. Thus our results implicate Nlrp10 as a novel NLR involved in the generation of anti-fungal adaptive immune responses against C. albicans through a mechanism that is independent of the Nlrp3 inflammasome and the production of IL-1β.
The generation of Nlrp10−/−, Nlrp6−/−, Nlrp12−/−, Nlrc4−/−, and ASC−/− mice has been described previously (8-12). Sex and age matched C57BL/6 (NCI) mice were used as controls. All protocols used in this study were approved by the Institutional Animal Care and Use Committee at the University of Iowa. Bone marrow chimeras were generated as described (8). Reconstitution was greater than 97% in Nlrp10−/− mice reconstituted with WT bone marrow and 82% in WT mice reconstituted with Nlrp10−/− bone marrow.
The C. albicans clinical isolate FC20 was used in this study (2). Culture conditions for C. albicans yeast and hyphae have been previously described (2). Mice were infected i.v. with 5 × 105 CFU of C. albicans and survival assessed; mice found in a moribund state for more than 4 h were considered terminal and euthanized. Kidneys were harvested at the indicated time post-infection (p.i.) and dilutions of homogenized organs were plated and counted to determine CFUs. Serum blood urea nitrogen (BUN) and creatinine levels were quantified at the Animal Fluid Analysis Core at the University of Iowa. To assess renal cytokine levels kidneys were homogenized, resuspended in lysis buffer (50 mM Tris·HCl, 5 mM EDTA, 150 mM NaCl, 1% Triton X-100, and a protease inhibitor cocktail (Roche)) and cytokine levels measured by ELISA.
Mice were infected i.v. with a sublethal dose (5 × 104 CFU) of C. albicans. 14 d p.i. spleens were collected and splenocytes cultured in the presence or absence of 1 × 107 ml−1 heat-killed C. albicans for 72 h. Supernatants were collected and IL-17 and IFNγ levels assessed by ELISA (eBiosciences).
WT mice were infected i.v. with 5 × 104 CFU of C. albicans; 10 d p.i. mice were rechallenged i.v. with 1 × 106 heat-killed C. albicans. 5 d later splenic CD4+ T cells were isolated using MACS microbeads (Miltenyi Biotec). CD4+ T cells isolated from uninfected WT mice were used as naïve controls. 5 × 106 naïve or immune CD4+ T cells were transferred i.v. into Nlrp10−/− mice; 24 h following the adoptive transfer mice were infected i.v. with 5 × 105 CFU of C. albicans and survival monitored.
Bone marrow-derived Mϕ (BMMϕ) and bone marrow-derived DC (BMDC) (13, 14) were either left unprimed or primed with 50 ng/ml LPS (Invivogen) for 3-4 h and then infected with C. albicans (MOI 10:1), F. tularensis LVS (MOI 50:1) or P. aeruginosa PAK strain (MOI 10:1) for 6 h or as indicated. LPS-primed BMMϕ were challenged with 50 μg/cm2 silica (Min-U-Sil-5; Pennsylvania Glass Sand Corporation), 5 mM ATP (Sigma) or 20 μM nigericin (Sigma) for 6 h. For ATP and nigericin treated cells media was replaced with fresh media 30 min after stimulation. Antibody pairs for ELISA were from eBiosciences except for IL-1β (R&D Systems).
Phylogenetic analysis of the NLR family demonstrates that Nlrp6, Nlrp10 and Nlrp12 NACHT sequences are closely related to that of Nlrp3 (1). In addition, Nlrc4, which activates caspase-1 in response to cytosolic flagellin and bacterial type III secretion systems (1), has also been shown to play a role in controlling mucosal C. albicans infections (5). We hypothesized, that similar to Nlrp3, these receptors might contribute to the in vivo immune response against a systemic C. albicans infection. In order to assess this we tested the susceptibility of Nlrp6-, Nlrp10-, Nlrp12- and Nlrc4-deficient mice to a systemic infection with C. albicans. Nlrp6-, Nlrp12- and Nlrc4-deficient mice did not show increased susceptibility to i.v. infection with C. albicans compared to WT mice (Fig. 1A). Surprisingly, Nlrp10−/− mice were highly susceptible to C. albicans infection with 100% mortality by day 16 p.i. (Fig. 1B).
Sepsis is the main cause of death in hematogenously disseminated candidiasis; in this model, renal dysfunction strongly correlates with increased kidney fungal burdens as well as increased mortality (15). We therefore evaluated kidneys of WT and Nlrp10−/− mice 9 d p.i. with C. albicans. Histologic sections of kidneys revealed more severe early fibroplasia and parenchymal loss in kidneys of Nlrp10−/− mice compared to WT (Fig. 1C). Kidneys from Nlrp10−/− mice had significantly more collagen deposition as shown by Masson’s trichrome stain than WT, indicating greater damage (Fig. 1C and Suppl. Fig. 1A). Very few yeast were detected histologically in kidney sections from WT mice in contrast to Nlrp10−/− mice where C. albicans yeast and hyphae were readily observed in the renal cortex and medulla (Fig. 1C and Suppl. Fig. 1B). Surprisingly, despite increased C. albicans within the renal parenchyma of Nlrp10−/− mice there was no significant difference in the percent of parenchymal Mϕ and neutrophil staining between WT and Nlrp10−/− mice (Suppl. Fig. 1C) suggesting a possible functional defect in the inflammatory response observed in the absence of Nlrp10.
Consistent with the increased renal damage observed by histology, 9 d p.i. Nlrp10−/− mice had diminished renal function as reflected by significantly higher serum blood urea nitrogen (BUN) and creatinine levels (Fig. 1D). Increased damage in Nlrp10-deficient kidneys correlated with elevated IL-1α and IL-6 levels within the kidney at day 9 p.i., although IL-1β, IL-18, IL-12p40 and IL-23 levels were unaffected by Nlrp10-deficiency (Suppl. Fig. 1D). Nlrp10−/− mice also had significantly higher fungal burdens in the kidney at 9 d p.i. indicating a role for Nlrp10 in controlling the replication of C. albicans in vivo (Fig. 1E). Surprisingly, during the early stages of infection, examined at day 3 and 6, there was no difference between WT and Nlrp10−/− mice in renal function (Fig. 1D), fungal burdens (Fig. 1E) and kidney cytokines (data not shown). These data suggest that early innate mechanisms required to control C. albicans replication in vivo remain intact in Nlrp10-deficient mice.
To understand better the biological function of Nlrp10, we examined the tissue distribution of Nlrp10 in WT mice. Consistent with previous reports we found high expression of Nlrp10 mRNA in the heart (7); in addition Nlrp10 was highly expressed in the tongue, testis and spleen (Suppl. Fig. 1D). Within the hematopoietic compartment Nlrp10 was expressed in Mϕ, DCs, CD4+ T cells, CD19+ B cells and neutrophils, but minimally in CD8+ T cells (Suppl. Fig. 1E). Stimulation of Mϕ and DCs with live C. albicans in vitro resulted in a reduction of Nlrp10 expression in these cells (Suppl. Fig. 1F). In contrast, LPS and heat-killed C. albicans did not significantly alter Nlrp10 mRNA expression (Suppl. Fig. 1F).
Given that Nlrp10 is expressed in both hematopoietic cells as well as stromal cells we wanted to determine whether the increased susceptibility of Nlrp10−/− mice to disseminated candidiasis was the result of loss of Nlrp10 in the hematopoietic compartment. To do this we generated bone marrow chimeric mice in which Nlrp10 deficiency was restricted to either the hematopoietic or non-hematopoietic compartment. WT mice that received Nlrp10−/− bone marrow were susceptible to disseminated candidiasis and recapitulated the phenotype observed in Nlrp10-deficient mice (Fig. 2A). Conversely, Nlrp10−/− mice that received WT bone marrow did not have significantly increased mortality compared to WT mice that received WT bone marrow (Fig. 2A). These results suggest that the increased susceptibility of Nlrp10−/− mice to C. albicans infection was primarily due to a deficiency of Nlrp10 within the hematopoietic compartment. A recent study by Lautz et al. demonstrates that Nlrp10 contributes to proinflammatory cytokine release by epithelial cells and dermal fibroblasts in response to infection with Shigella flexneri (16) and may possibly explain the reduction, although not significant, in survival of Nlrp10−/− mice that received WT bone marrow.
Recent studies using in vitro overexpression of Nlrp10 as well as Nlrp10 transgenic mice suggested that Nlrp10 could inhibit the activation of Nlrp3 and Nlrc4 inflammasomes as well as suppress NF-κB activation (6, 7). We examined the ability of Nlrp10-deficient Mϕ to secrete IL-1β in response to specific inflammasome agonists. LPS-primed BMMϕ from WT and Nlrp10−/− mice secreted comparable levels of IL-1β when challenged with C. albicans yeast (Fig. 2B and C). Similarly, the Nlrp3 agonists silica and nigericin induced similar levels of IL-1β secretion from Nlrp10-deficient BMMϕ when compared to WT BMMϕ (Fig. 2D). In addition, Pseudomonas aeruginosa and Francisella tularensis LVS, activator of the Nlrc4 and AIM2 inflammasomes respectively (1), also induced comparable levels of IL-1β secretion from Nlrp10−/− and WT BMMϕ (Fig. 2C). As expected, IL-1β secretion in response to C. albicans, P. aeruginosa and F. tularensis LVS was dependent on the presence of the inflammasome adaptor molecule ASC (Fig. 2C). Similar to our findings with BMMϕ, BMDC from WT and Nlrp10−/− mice secreted comparable levels of IL-1β when challenged with C. albicans yeast (Suppl. Fig. 2A). In addition, both unprimed and LPS-primed WT and Nlrp10−/− BMMϕ failed to secrete IL-1β in response to C. albicans hyphae (Suppl. Fig. 2B). C. albicans hyphae were capable of inducing the secretion of IL-1β from LPS-primed WT BMDC, although this was again similar to levels of IL-1β secreted from LPS-primed Nlrp10−/− BMDC (Suppl. Fig. 2C). These data indicate that a deficiency of Nlrp10 in Mϕ or DCs does not affect the activation of Nlrp3, Nlrc4 and AIM2 inflammasomes. Consistent with a recent study by Eisenbarth et al. (8), in response to stimulation with the TLR4 agonist LPS the production of IL-6, TNFα and IL-12 p40 was unaffected by Nlrp10-deficiency in both BMMϕ and BMDC (Fig. 2E and Suppl. Fig. 2D) suggesting that Nlrp10 also does not suppress NF-κB activation in these cells.
Internalization and killing of Candida is an indispensable function of Mϕ in the control of candidal infections. Nlrp10-deficient BMMϕ and BMDC did not display any defect in their ability to phagocytose C. albicans compared to WT BMMϕ and BMDC (Suppl. Fig. 2E). Growth of C. albicans within Mϕ was also comparable between WT and Nlrp10−/− BMMϕ (Suppl. Fig. 2F). Similarly, we did not observe defects in the phogocytosis or growth of C. albicans within Nlrp10−/− thioglycollate-elicited peritoneal neutrophils (Suppl. Fig. 2E and F). Taken together these data suggest that phagocytosis, intracellular killing and the generation of proinflammatory cytokines by Mϕ and DCs remain intact in the absence of Nlrp10.
Adaptive immune responses play a crucial role in host defense against C. albicans. Th17 responses, in particular, are important for control of C. albicans infections through the recruitment of neutrophils to the infection site. As such, mice deficient in the cytokine receptor IL-17 receptor A (IL-17RA) have increased susceptibility to both disseminated and mucosal candidiasis (17, 18). Given our findings that Nlrp10-deficiency had little effect on Mϕ and DC production of pro-inflammatory cytokines we next examined if the generation of CD4+ T helper cell responses to C. albicans remained intact in Nlrp10−/− mice. WT and Nlrp10−/− mice were infected i.v. with a sublethal dose of C. albicans; 15 d p.i. Th1 and Th17 responses were evaluated by measuring IFNγ and IL-17 release, respectively, from splenocytes restimulated with heat-killed C. albicans for 72 h. As expected, WT mice displayed a mixed Th1 and Th17 response to infection with C. albicans as evidenced by the secretion of IFNγ and IL-17 by restimulated splenocytes (Fig. 3A). Surprisingly, Nlrp10−/− mice displayed a profound defect in the generation of both Candida-specific Th1 and Th17 responses (Fig. 3A), which indicate that Nlrp10 is required for driving appropriate adaptive immune responses to C. albicans in vivo.
To determine if the inability to generate appropriate adaptive immune responses was the cause of the increased mortality of Nlrp10−/− mice following C. albicans infection we adoptively transferred naïve and C. albicans-immune WT CD4+ T cells into Nlrp10−/− mice. Following CD4+ T cell adoptive transfer, Nlrp10−/− mice were infected i.v. with C. albicans. Nlrp10−/− mice that received CD4+ T cells from naïve WT mice succumbed to C. albicans infection at a similar rate to Nlrp10−/− control mice (Fig. 3B). However, Nlrp10−/− mice that received CD4+ T cells from WT mice that had previously been challenged with a sublethal dose of C. albicans and boosted with heat-killed C. albicans displayed significantly improved survival in response to a lethal C. albicans challenge compared to Nlrp10−/− control mice (Fig. 3B). Hence, taken together these data suggest that Nlrp10 is required for the generation of protective anti-fungal adaptive immune responses in vivo.
The defect in generation of specific T helper cell responses in Nlrp10−/− mice was not restricted to C. albicans; immunization of Nlrp10−/− mice with antigen in the presence of a number of adjuvants, including LPS, aluminum hydroxide and complete Freund’s adjuvant has also been shown to result in defective adaptive immune responses (8). In addition, Nlrp10-deficient DCs were shown to have an intrinsic defect in their ability to emigrate from a site of inflammation resulting in a lack of antigen transport to the draining lymph node and explaining the lack of priming of naïve CD4+ T cells in Nlrp10−/− mice (8). Hence Nlrp3 and Nlrp10 play distinct roles in shaping adaptive immune responses against fungal pathogens. Whereas Nlrp3 inflammasome-driven IL-1β production drives Th17 differentiation during a C. albicans infection (19), the role of Nlrp10 in the generation of specific T helper cell responses is likely to be at the level of appropriate DC migration and antigen presentation to naïve CD4+ T cells.
We thank Millennium Pharmaceuticals for providing knockout mice and Vickie Knepper-Adrian for assistance with animal husbandry.
††NIH grants R01 AI087630 (F.S.S.), K08 AI067736 (S.L.C.), K08 AI085038 (S.C.E) and an Edward Mallinckrodt, Jr. Foundation scholarship (F.S.S.) supported this work. R.A.F. is an Investigator of the Howard Hughes Medical Institute.