Background
Toll-like receptors (TLRs) are the main family of pattern recognition receptors (PRRs) through which immune and non-immune cells sense pathogen-associated molecular patterns (PAMPs) [
1,
2]. Several TLRs are implicated in the recognition of fungal pathogens such as
Candida albicans [
3,
4]. The interaction between TLRs and yeasts during candidiasis stimulates immune cells to generate inflammatory and immunomodulatory mediators that shape the host immune response. Unlike TLR4, TLR2 recognizes both blastoconidia and hyphal forms of
C. albicans [
5]. TLR2 forms heterodimers with either TLR1 or TLR6 which have been implicated in ligand discrimination [
6]. TLR2 senses phospholipomannans, which are expressed in the cell wall of
C. albicans [
7]. In addition, TLR2 in combination with galectin-3 also senses β-mannosides [
8].
TLRs are expressed not only in myeloid cells and leukocytes, but also in the intestinal epithelium, which contributes to mucosal homeostasis by preventing the penetration of commensal microbiota into the intestine [
9,
10]. In an animal model of colitis, TLR2
−/− mice developed more severe colonic inflammation than wild-type mice [
11]. Moreover, mutations in TLRs, including the
TLR2 gene, have been associated with predisposition to and maintenance of inflammatory bowel disease (IBD) [
12‐
14]. Interestingly, in patients with ulcerative colitis, Pierik et al. [
15] observed an association between
TLR1 and
TLR2 gene polymorphisms and pancolitis, and a negative relationship between
TLR6 polymorphisms and pancolitis, suggesting that TLR2 and its co-receptors TLR1 and TLR6 are involved in the initial immune response to pathogens in the development of IBD.
The aim of this study was to determine the impact of TLR1, TLR2, and TLR6 deficiency on inflammatory parameters associated with C. albicans colonization and acute colitis induced by DSS by comparing wild-type, TLR1−/−, TLR2−/−, and TLR6−/− mice. We also assessed intestinal permeability, serological response, and colonic expression levels of pro-inflammatory and anti-inflammatory cytokines in control and TLR-deficient mice. Finally, we explored the effects of TLR deficiency on neutrophil-mediated C. albicans phagocytosis/death.
Discussion
IBD is characterized by microbial dysbiosis related to the abundance of a range of pathogenic microbial species, in particular,
C. albicans [
16‐
18].
C. albicans colonization is more frequent and more severe in patients with Crohn’s disease than in control subjects [
19]. The abundance of yeast in the gut can cause inflammation by activating cells through PRRs, including TLRs [
20,
21]. TLRs are highly expressed in mucosal immune and epithelial cells and their triggering stimulates cytokine release and microbicidal activity [
22]. In the present study, we investigated the role of TLR1, TLR2, and TLR6 in intestinal inflammation and
C. albicans colonization. We previously developed a
C. albicans colonization model, which is promoted by intestinal inflammation induced by DSS [
23]. This
C. albicans colonization is not maintained without the presence of intestinal inflammation; otherwise
C. albicans is eliminated immediately in mice. In this DSS model, we employed a low concentration of DSS (1.5%) to promote
C. albicans colonization without inducing mouse mortality or severe colitis. In the present study, we compared DSS mice and mice treated with DSS and colonized with
C. albicans. Initially, we performed experiments without DSS by administering
C. albicans only to mice used as controls (TLR1
−/−, TLR2
−/−, TLR6
−/−, and wild-type mice). No
C. albicans colonization was observed in any of these mouse strains a few days later. Interestingly, we observed that TLR1
−/− and TLR2
−/− mice were more susceptible to DSS-induced colitis than TLR6
−/− and wild-type mice. Rakoff-Nahoum et al. [
10] demonstrated that mice deficient in TLR2 are highly susceptible to DSS-induced colitis, suggesting that TLRs contribute to prevent an aberrant immune response against the commensal flora. TLR2 forms heterodimers with either TLR1 or TLR6 to recognize different configurations of lipoproteins/lipopeptides [
6]. It has been reported that TLR2/TLR6 recognizes yeast-derived zymosan, which is mainly composed of mannans and β-glucans [
24]. Jouault et al. [
7,
8] showed that TLR2 binds to phospholipomannans of
C. albicans in the presence of galectin-3, and β-mannosides of
C. albicans, to induce pro-inflammatory responses by macrophages. TLR2 can also collaborate with dectin-1 to bind fungal β-glucans [
25,
26].
We observed that
C. albicans colonization enhances the susceptibility of TLR1
−/− and TLR2
−/− mice, but not TLR6
−/− mice, to intestinal inflammation. TLR6
−/− mice were more resistant to DSS-induced colitis as reflected by lower clinical and histologic scores for inflammation and mortality. These data corroborate findings from a recent study which showed that mice deficient in TLR6
−/− are protected from intestinal inflammation and have a low number of Th17 cells [
27]. This evidence emphasizes the role played by TLR6 in the induction of an inflammatory response to
C. albicans sensing. Netea et al. [
28] showed that, in contrast to TLR1, TLR6 is involved in the recognition of
C. albicans and modulation of the Th1/Th2 cytokine balance. Moreover, TLR2
−/− mice had significantly impaired survival to
Candida infection, indicating that TLR2 confers protection against primary disseminated candidiasis [
3].
All TLRs except TLR3 are expressed by neutrophils, which represent the primary first line of defense against
C. albicans infection [
1,
29]. In the present study, neutrophils from TLR1, TLR2, and TLR6 deficient mice were not affected in terms of
C. albicans phagocytosis. On the other hand, adhesion and migration of neutrophils from TLR1
−/− and TLR2
−/− mice to
C. albicans was impaired. In line with these findings, neutrophils and macrophages from TLR2
−/− mice internalized and killed
C. albicans as efficiently as wild-type cells [
4]. Moreover, Weindl et al. [
30] showed neutrophil-dependent TLR4-mediated protective mechanisms against
C. albicans infection at epithelial surfaces, a phenomenon that was independent of neutrophil migration to
C. albicans or epithelial cells.
Analysis of colon cytokine expression revealed that TLR1 deficiency strongly up-regulated TNF, IL-1β, and IL17A, whereas TLR6 deletion down-regulated TNF, IL-1β, and IL17A colonic expression when compared to controls. These data suggest that TLR1 plays a role in dampening the cytokine response and preventing excessive immune-mediated tissue damage, whereas TLR6 is involved in exacerbating the inflammatory response associated with tissue damage upon recognition of
C. albicans and DSS-induced colitis [
31,
32]. Notably,
C. albicans colonization increased TLR2 expression in the colons of DSS-treated wild-type mice, whereas TLR2 expression was reduced in galectin-3 deficient mice leading to high pro-inflammatory cytokine expression and aggravated intestinal inflammation [
20].
IL-10 is essential for host defense against
C. albicans infection and can limit the potential tissue damage caused by inflammation [
33]. In the present study, we showed that in contrast to the state of TLR6
−/− mice, DSS-induced colitis induced the production of pro- and anti-inflammatory cytokines, including IL-10, in TLR1
−/− mice. In line with this study, the absence of TLR6 dramatically increased survival and decreased IL-10 production, whereas the absence of TLR1 led to decreased survival and higher IL-10 levels in mice infected with
Yersinia pestis, suggesting that TLR6 is a distinct TLR receptor driving regulatory IL-10 responses [
34]. Netea et al. [
35] showed that TLR activation, in particular TLR2, can suppress the immune defense against
C. albicans through the induction of IL-10 and regulatory T-cells.
In this study, we observed that TLR6
−/− mice eliminate
C. albicans from the digestive tract more rapidly than TLR1
−/− and TLR2
−/− mice. In particular, colonization in the stomach was dramatically higher in TLR1
−/− mice, supporting the notion of a protective role of TLR1 and TLR2 against
C. albicans colonization. Additionally, in agreement with our results,
TLR1 gene polymorphisms have been associated with an increased susceptibility to candidemia in patients [
21]. In addition, polymorphisms in the
TLR2 gene have been identified as a risk factor for candidemia [
36,
37]. Although TLR2 interacts with a large number of non-TLR molecules, in particular Galectin-3 and Dectin-1-mediated ERK/MAPK activation [
8,
25], allowing the recognition of fungal ligand varieties, in our study, we found that mice deficient in TLR1 had higher levels of colonization with
C. albicans in the stomach and colon than TLR2
−/− mice. These data show that TLR1 is more involved in
C. albicans elimination from the gut than TLR2, suggesting that the absence of TLR1 can have a great impact on the formation of the heterodimer TLR1/TLR2 when compared to the absence of TLR2. Van Duin et al. [
38] showed that alterations in baseline TLR1 surface expression were increased in elderly individuals, whereas TLR2 surface expression was unaffected, suggesting that the defect in TLR1 expression can contribute to the increased infection-related morbidity and mortality observed in elderly individuals.
Changes in commensal bacterial diversity can differentially modulate mucosal TLR responsiveness, leading to TLR-mediated hyper- or hypo-reactive immune responses [
10,
39,
40]. In the present study, intestinal inflammation alone was sufficient to promote the overgrowth of
E. coli. Additionally, both inflammation and
C. albicans colonization maintained
E. coli overgrowth in TLR1
−/−, TLR2
−/− and wild-type mice when compared to that in TLR6
−/− mice. These data are consistent with Lupp et al. [
41] finding which suggested that the commensal
E. coli increased during colitis and displayed a strong pro-inflammatory potential. Ey et al. [
14] showed that an unaltered microbiota was required for colitis exacerbation in TLR2/MDR1A double-knockout mice once protection from colitis was observed upon antibiotic treatment.
In the present study, intestinal inflammation alone was sufficient to promote the overgrowth of
E. coli. Additionally, both inflammation and
C. albicans colonization maintained
E. coli overgrowth in TLR1
−/−, TLR2
−/− and wild-type mice when compared to that in TLR6
−/− mice. These data are consistent with Lupp et al. [
41] finding which suggested that the commensal
E. coli increased during colitis and displayed a strong pro-inflammatory potential.
Experimental and clinical studies emphasize the role of the serologic response during the development of IBD [
20,
42]. Crohn’s disease patients who express high levels of serologic markers experience more aggressive disease, suggesting that the presence of Crohn’s disease marker antibodies reflect a specific mucosal immune-mediated response [
43]. We observed that mannanemia was elevated only in TLR1
−/− and TLR2
−/− mice. High mannanemia in TLR1
−/− and TLR2
−/− mice was correlated with increased intestinal permeability, which facilitates the passage of fungi into the bloodstream. Clinically, Pierik et al. [
15] showed a negative association between
TLR6 S249P and ulcerative colitis with proctitis. In our study, we found in a clinical cohort of 26 healthy subjects (between 25- and 65-years-old) an association between the
TLR6 rs5743810 homozygous wild-type genotype and ASCA (anti-
Saccharomyces cerevisiae antibody) level (
P = 0.0317). Homozygous healthy subjects (
TLR6 A/A wild-type) have significantly higher ASCA levels than heterozygous (
TLR6 A/G) and homozygous mutants (
TLR6 G/G) indicating a possible association of this
TLR6 rs5743810 polymorphism with IBD (Additional file
1). These clinical data corroborate findings from our experimental study, which showed that deletion of TLR6 decreased the circulating mannan level in mice. Taylor et al. [
44] showed that among African Americans, women who carried 1 or 2 of the
TLR6 rs5743810 alleles had decreased odds of endometritis and upper genital tract infection and there was a similar trend among white participants. We intend to widen our study to larger groups of healthy subjects and patients with CD.
Authors’ contributions
LC, HV, DLR, and SJ performed the experiments. LC, HV, LD, LR, TJ, DP, BS, TC, TR, and SJ analyzed data. LC, HV, LD, LR, TJ, DP, BS, TC, TR, and SJ interpreted results of experiments. DLR, LD, LR, TC, and TR contributed reagents/materials/analysis tools. SJ designed the experiments and drafted the manuscript. All authors read and approved the final manuscript.