(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Type II taste cells participate in mucosal immune surveillance [1]

['Yumei Qin', 'School Of Food Science', 'Bioengineering', 'Zhejiang Gongshang University', 'Hangzhou', 'Peoples Republic Of China', 'Monell Chemical Senses Center', 'Philadelphia', 'Pennsylvania', 'United States Of America']

Date: 2023-01

The oral microbiome is second only to its intestinal counterpart in diversity and abundance, but its effects on taste cells remains largely unexplored. Using single-cell RNASeq, we found that mouse taste cells, in particular, sweet and umami receptor cells that express taste 1 receptor member 3 (Tas1r3), have a gene expression signature reminiscent of Microfold (M) cells, a central player in immune surveillance in the mucosa-associated lymphoid tissue (MALT) such as those in the Peyer’s patch and tonsils. Administration of tumor necrosis factor ligand superfamily member 11 (TNFSF11; also known as RANKL), a growth factor required for differentiation of M cells, dramatically increased M cell proliferation and marker gene expression in the taste papillae and in cultured taste organoids from wild-type (WT) mice. Taste papillae and organoids from knockout mice lacking Spib (Spib KO ), a RANKL-regulated transcription factor required for M cell development and regeneration on the other hand, failed to respond to RANKL. Taste papillae from Spib KO mice also showed reduced expression of NF-κB signaling pathway components and proinflammatory cytokines and attracted fewer immune cells. However, lipopolysaccharide-induced expression of cytokines was strongly up-regulated in Spib KO mice compared to their WT counterparts. Like M cells, taste cells from WT but not Spib KO mice readily took up fluorescently labeled microbeads, a proxy for microbial transcytosis. The proportion of taste cell subtypes are unaltered in Spib KO mice; however, they displayed increased attraction to sweet and umami taste stimuli. We propose that taste cells are involved in immune surveillance and may tune their taste responses to microbial signaling and infection.

The most important component of adaptive immunity at the mucosae is the mucosa-associated lymphoid tissue (MALT) [ 22 , 23 ]. MALTs are immune inductive sites that sample the luminal microbes and generate an appropriate mucosal immune response. MALT consists of lymphoid follicles containing B and T cells, which are overlaid by an epithelial layer called follicle-associated epithelium (FAE) [ 24 ]. Specialized epithelial cells in the FAE called Microfold cells (M cells) transcytoses luminal microbes and pass them on to antigen-presenting cells (APCs) such as dendritic cells and macrophages housed in their basal pocket [ 25 , 26 ]. Antigen processing and presentation by APCs stimulate the B and T lymphocytes in the underlying lymphoid follicles that mount an appropriate immune response [ 24 , 26 , 27 ]. These effector cells then migrate to other parts of the mucosae and systemically in blood. Thus, M cells play a central role in mucosal immunity. M cells express several receptors for microbes such as glycoprotein 2 (GP2), peptidoglycan recognition protein 1 (PGLYRP1), and the poliovirus receptor (PVR) that bind to and internalize luminal microbes [ 28 – 30 ]. They have a well-developed microvesicular system, but poorly developed lysosomes enabling rapid transport of their microbial cargo mostly intact across the epithelium [ 26 ]. M cell differentiation is induced by tumor necrosis factor ligand superfamily member 11 (TNFSF11; also called RANKL) secreted by connective tissue cells underlying the MALT epithelium [ 31 ]. RANKL binds to the receptor tumor necrosis factor receptor superfamily member 11A (TNFRSF11A), which activates the noncanonical nuclear factor kappa B (NF-κB) signaling pathway to turn on the expression of early M cell marker genes [ 31 – 34 ]. The most prominent among them is Spib, a transcription factor that orchestrates the later stages of M cell differentiation [ 35 , 36 ]. Using single-cell RNASeq (scRNASeq) of GFP-labeled mouse taste cells, we found that taste cells including sweet and umami receptor cells that express taste 1 receptor member 3 (Tas1r3) (Tas1r3+ cells) express several M cell marker genes [ 37 ]. Consistent with this gene expression profile, RANKL administration led to marked up-regulation of M cell marker genes in the taste papillae and cultured taste organoids from wild-type (WT) but not Spib knockout (Spib KO ) mice. Spib KO mice also showed reduced expression of NF-κB signaling pathway components and proinflammatory cytokine gene expression in their taste papillae and attracted fewer immune cells to the papillae. However, lipopolysaccharide (LPS)-induced expression of cytokines was highly up-regulated in Spib KO mice compared to their WT littermates. Using a fluorescently labeled microbead uptake assay, we show that taste cells from WT but not Spib KO mice are capable of transcytosis of luminal microparticles. Spib ablation did not affect the proportion of taste cell subtypes in the papillae but caused increased attraction to sweet and umami taste stimuli. Our results indicate that taste cells possess M cell–like properties and might modulate taste signaling in response to microbial colonization and infection.

16s RNA sequencing and metagenomics studies have shown that the oral cavity including the tongue dorsum is colonized by a diverse array of microbial species [ 8 – 11 ]. Unlike cells in the non-taste (NT) lingual epithelium, taste cells are continually exposed to the oral microbiota through their microvilli that project to the lingual surface through taste pores. However, their effects on taste signaling and taste cell regeneration have not received the deserved level of attention. The taste papillae are patrolled by a diverse population of immune cells, mostly dendritic cells including the Langerhans cell subtype, T cells, and macrophages [ 12 , 13 ]. Most oral mucosa-resident microbes are harmless or beneficial, and the host develops immune tolerance toward them [ 14 ]. However, oral dysbiosis can cause taste loss or distortion, commonly seen in patients with conditions such as influenza, oral thrush (candidiasis), HIV, bacterial infection, and COVID-19 [ 15 – 21 ]. Identifying the pathways underlying the finely tuned crosstalk between taste cells, the oral microbiome and epithelial immune cells will help uncover how microbiota influence taste cells in health and disease.

In mammals such as the mouse, most taste buds are found in three types of taste papillae on the dorsal surface of the tongue [ 1 , 2 ]. Among them, the fungiform papillae (FFP) located on the anterior tongue each house a single taste bud, and the foliate (FOP) and circumvallate (CVP) papillae located laterally and medially on the posterior tongue host a few hundred taste buds each [ 1 , 2 ]. The taste buds in CVP and FOP are arranged around trenches that extend down into the tongue. Each taste bud contains approximately 50 to 100 mature receptor cells classified as type I, type II, type III, and type IV cells based on morphology and marker gene expression [ 1 – 3 ]. These cells are further classified into functional subtypes that respond to basic taste qualities of sweet, bitter, umami (subtypes of type II cells), sour, and salty (mostly subtypes of type III cells) [ 1 – 3 ]. Taste cells have a half-life ranging from 8 to 24 days and are continually regenerated from a population of stem cells located at the base of taste papillae [ 4 – 7 ].

Given its role in regeneration and function of M cells, we asked if ablation of Spib cause defects in taste cell regeneration and/or taste preference. Immunostaining with antibodies against the taste marker proteins T1R3, TRPM5, GNAT3, and CAR4 showed that the proportion of type II and III cells are unaltered in the CVP of Spib KO mice compared to WT ( S8A–S8I Fig ). These results were confirmed by qPCR analysis of the corresponding mRNAs ( S8J Fig ). However, in brief access taste tests, Spib KO mice displayed increased attraction to the prototypical sweet and umami taste stimuli sucrose and monopotassium glutamate compared to WT littermates ( Fig 4A and 4C ). The attraction to sucralose appeared to be higher as well in Spib KO mice, although not statistically significant ( Fig 4B ). On the other hand, the responses to denatonium benzoate (bitter), sodium chloride (salty), and citric acid (sour) that are mediated by taste cell types other than Tas1r3+ cells were unchanged between the two strains ( Fig 4D–4F ).

( A - F ) Indirect immunofluorescence confocal microscopy of CVP sections from WT and Spib KO mice stained with antibodies against mouse immune cell markers CD45 ( A , D ), CD11B ( B , E ), and CD3 ( C , F ). Compared to WT mice, Spib KO mice had fewer immune cells in the CVP. ( G - H ) Uptake of 200-nm diameter fluorescent beads in taste cells from CVP of WT ( G ) and Spib KO mice ( H ) were observed using confocal microscopy. Taste cells are visualized by staining with a pan-taste cell marker KCNQ1 in panels A - H . ( I ) Uptake of beads was quantitated by image analysis and normalized so that the average uptake in WT mice was 1.0. (The data underlying the graphs can be found in Data I in S6_Data.) Compared to WT mice, Spib KO mice took up fewer beads. ***p < .001. Scale bars, 50 μm.

Next, we compared various aspects of mucosal immunity in CVP of WT and Spib KO mice. Multiple stimuli, including RANKL, other TNF family ligands, and pathogen-associated molecular patterns such as LPS signal through the NF-κB pathway [ 39 ]. We asked if Spib KO mice show defects in expression of the components of the NF-κB signaling pathway and its target genes. qPCR analysis showed that the expression of the adapter protein Myd88, regulator proteins Tollip, Irak3, and Irak4, and the transcription factors Nfkb1, Nfkb2, Rela, Rel, and Irf6 are down-regulated in Spib KO mice ( S6A and S6B Fig ). In agreement with these findings, expression of several proinflammatory cytokines regulated by NF-κB, namely, Il1b, Il6, Mcp1, and Tnf is down-regulated, while that of the anti-inflammatory cytokines Il10 and Il12 and Ifnγ did not change in Spib KO mice ( S6C Fig ). LPS is known to cause robust proinflammatory cytokine expression in CVP, and we tested if this is recapitulated in Spib KO mice [ 20 ]. In contrast to baseline expression levels noted above, LPS administration triggered exaggerated cytokine gene expression in CVP of Spib KO mice compared to WT littermates ( S7 Fig ). Next, we asked if the reduced baseline cytokine expression in Spib KO mice affected the recruitment of immune cells to the CVP. Immunostaining with antibodies against CD11B and CD45, which label a broad spectrum of immune cells or the T cell marker CD3, showed that Spib KO mice had fewer immune cells patrolling the CVP ( Fig 3A–3F ). Finally, we tested if taste cells are capable of microbial transcytosis and if this is impaired in Spib KO mice. The uptake of fluorescently labeled microbeads, a proxy for transcytosis is evident in KCNQ1-expressing taste cells in CVP from WT, but not Spib KO mice ( Fig 3G–3I ).

In the PP and intestinal villi, RANKL treatment induces expression of various M cell marker genes in a stereotypical temporal order [ 35 ]. To determine if this is conserved in taste cells, we turned to taste organoids derived from Spib KO :Lgr5-EGFP knockin or control (Lgr5-EGFP) strains. Like in PP and intestinal organoids, expression of the early M cell markers Marcksl1 and Ccl9 mRNAs and the corresponding proteins started increasing by day 1 and reached their peaks at day 2; that of Spib increased gradually and reached the peak by day 3; and that of the late marker Gp2 increased more slowly and reached its peak at day 4 after RANKL administration in organoids derived from control mice ( S5A–S5P Fig ). Taste organoids derived from Spib KO mice, on the other hand, showed a dramatically impaired induction of Gp2, Ccl9, and Marcksl1 mRNAs upon RANKL administration ( S5M–S5P Fig ).

We next asked if RANKL can trigger Spib-dependent increase in expression of M cell marker genes in taste papillae like it does in other MALT [ 31 , 35 , 36 ]. The basal expression levels of M cell marker proteins GP2, CCL9, MARCKSL1, and SPIB are comparatively low in the WT animals, and the mRNAs encoding Gp2 and Ccl9 are lower in CVP of Spib KO mice ( S4A , S4D , S4I and S4K Fig ). Administration of RANKL led to a dramatic increase in the proportion of GP2, CCL9, MARCKSL1, and SPIB expressing cells and the levels of corresponding mRNAs in the CVP of WT but not Spib KO mice ( S4E and S4L Fig ). At the same time, qPCR analysis showed that the expression of Gp2, Ccl9, and Marcksl1 mRNAs were not up-regulated by RANKL treatment in the NT epithelium from both strains ( S4M Fig ).

Double label immunohistochemistry also showed that two other M cell markers, GP2 and CCL9, appear to be coexpressed with the type II cell marker TRPM5, but not with the Type III cell marker CAR4 in the CVP, although the coexpression was not quantified ( S2 Fig ). Of note, the strong GP2 staining at the apex of taste buds, presumably in taste cell microvilli, is similar to the pattern observed in M cells in other MALT for this receptor protein ( S2C and S2F Fig ) [ 28 ]. CVP from Pou2f3 knockout mice that lack all type II taste cells including Tas1r3+ cells do not express CCL9, GP2, and SPIB ( S3 Fig ) [ 38 ]. Collectively, the results above conclusively show that taste cells, in particular, the type II subtype, express several M cell marker genes.

To identify the taste cell types that express M cell marker genes, we turned to RNAscope Hiplex Fluorescent Assay and double label immunohistochemistry. RNAscope assay was done with probe sets against the M cell marker genes Spib, Gp2, and Tnfrsf11a and taste cell marker genes Tas1r3, Gnat3, Trpm5 (marker for all type II cells), and Ddc (marker for all type III cells) in the CVP. In agreement with the scRNASeq data, strong colocalization was observed between Spib and Tas1r3 with 95% of Spib+ cells coexpressing Tas1r3 ( Fig 1A and S2 Table ). About 27% of Spib+ cells coexpressed Gnat3, 93% coexpressed Trpm5, and about 18% coexpressed Ddc ( Fig 1B–1D and S2 Table ). Notably, the expression level of Spib in Ddc+ and Gnat3+ cells (indicated by the number of fluorescent spots per cell) is much weaker than in Tas1r3+ and Trpm5+ cells ( Fig 1A–1D ). Gp2 expression is distributed among all cell types tested, although it tended to be more strongly associated with type II taste cells ( Fig 1E–1H and S2 Table ). Tnfrsf11a expression is evenly distributed across all cell types ( Fig 1I–1L and S2 Table ).

To identify genes involved in mucosal immunity expressed in taste cells, we analyzed scRNASeq data from Gnat3-EGFP-expressing (Gnat3-EGFP+, primarily bitter taste receptor expressing in CVP, type II), Tas1r3-EGFP+ (sweet and umami receptor expressing, type II), and Gad1-EGFP+ (type III) taste cells from CVP of respective EGFP transgenic mice. Our analysis revealed that all cell types, especially the Tas1r3+ cells selectively expressed several genes critical for M cell maturation and function ( S1 Table ) [ 37 ]. The expression of a subset of these genes at the mRNA and protein levels in taste cells was confirmed using molecular and histological techniques. Endpoint PCR and quantitative real-time PCR showed robust expression of M cell marker genes Gp2, Marcksl1, Ccl9, Anxa5, Sgne1, and Spib in the CVP, while they were either expressed at much lower levels or not at all in the NT lingual epithelium ( S1A and S1B Fig ). RNAscope hybridization with an Spib-specific probe set showed that it is expressed in the taste cells in CVP and FFP ( S1C and S1D Fig ). These results were confirmed at the protein level using indirect immunohistochemistry with an antibody against SPIB, which stained the nuclei of subpopulations of taste cells in CVP, FOP, and PP ( S1E and S1G Fig ). The specificity of the SPIB antibody and secondary antibody was confirmed by lack of staining in the PP and CVP of Spib KO mice and when the primary antibody was omitted while staining the CVP of WT mice ( S1H and S1J Fig ).

Discussion

Mucosae are important routes of microbial colonization, and animals have evolved a strong mucosal immune system comprised of both innate and adaptive components to counter infection [40–42]. The MALT is a type of secondary lymphoid tissue critical for mucosal adaptive immunity. The MALT in the gut (PP) and tonsils are the best studied, but MALT occurs in other mucosae such as the salivary glands, nasopharynx, conjunctiva, and tear ducts as well [24,43,44]. The mucosa in the tongue dorsum is heavily colonized by the oral microbiome, even more so than the buccal mucosa [9,45]. Most of the tongue surface is made up of keratinized stratified squamous epithelium that acts as an effective barrier to microbial invasion. However, the microvilli of taste cells project to the tongue surface through the taste pores in the taste buds and presumably represent an easier route for microbial invasion. The trenches in the CVP and FOP surrounded by taste buds are ideal sites for long-term microbial colonization, as they are largely shielded from salivary flushing. However, the effects of the oral microbiota on taste cells have not been studied in sufficient detail so far. The expression of Toll-like receptors (TLRs), interferon receptors, and their downstream signaling pathway components in taste cells has been documented [46–48]. Administration of LPS or double-stranded RNA that binds to TLRs and mimics bacterial and viral infection, respectively, activates TLR and interferon signaling pathways in taste cells and diminishes taste cell regeneration and taste transduction, likely by promoting secretion of proinflammatory cytokines [20,47,49,50]. Similarly, knocking out Tlr4, the primary receptor for LPS in mice, leads to diminished taste response to sugars, lipids, and umami [48]. Interestingly, sweet and bitter taste receptors bind microbial metabolites and mediate immune responses in several extra oral tissues, but whether they do so in taste papillae itself is not known [51,52].

There are several functional and developmental similarities between taste and intestinal epithelia. For instance, intestinal epithelial cells and taste cells in the CVP and FOP are of endodermal origin and arise from Lgr5-expressing stem cells [4,53,54]. Taste receptors and other members of the taste transduction machinery are also expressed in the nutrient-sensing enteroendocrine cells and the microbe- and parasite-sensing tuft cells in the intestine [55–60]. Finally, both taste and intestinal epithelial cells are heavily exposed to the respective microbiota, and it is plausible that some taste cell types may share functional features of M cells. Indeed, we identified an M cell–like gene expression signature in taste cells using scRNASeq (Figs 1, 2, S1, and S2 and S1 Table). Strikingly, RANKL administration led to M cell proliferation in CVP and up-regulation of M cell marker genes in cultured taste organoids in the same temporal order observed in the PP and PP-derived organoids (S4 and S5 Figs). Finally, as in the PP, M cell marker gene expression in CVP and taste organoids requires Spib. SpibKO mice have low basal levels of M cell gene expression and fail to respond to RANKL (S4 and S5 Figs). In agreement with these observations, taste cells from WT but not SpibKO mice are able to transcytose luminal microbeads (Fig 3G–3I). Thus, all evidence indicates that taste cells have M cell–like properties. However, it is not clear if all taste cell types have properties of M cells. In the basal level, type II cells, in particular the Tas1r3+ population, strongly coexpress Spib and Gp2 at both mRNA and protein levels (Figs 1, 2, and S2). The expression level of Spib (as determined by the number of fluorescent spots per cell) in Ddc+ cells is much lower than in type II cells (Fig 2A–2H). However, it is unlikely that all the cells expressing M cell marker genes upon RANKL stimulation are type II cells or if only type II cells are capable of transcytosis. Some type III taste cells also express Tnfrsf11a (RANKL receptor), and it possible that they form a sizeable proportion of cells expressing M cell marker genes upon RANKL stimulation (Fig 1L and S1 and S2 Tables). Similarly, it is possible that type III taste cells are also capable of transcytosis, although a subtype-level transcytosis assay was not done due to technical difficulties. Nevertheless, the basal level expression of SPIB, GP2, and CCL9 is abrogated in Pou2f3KO animals, indicating a prominent role for type II taste cells in immune surveillance (S3 Fig). It is unlikely that type I or IV cells are capable of transcytosis since their microvilli do not project out of the taste pore [3]. In summary, our data indicate that type II taste cells, in particular, Tas1r3+ cells, likely form the bulk of M cell–like taste cells, although it is possible that a smaller subset of other taste cells, likely type III cells can also acquire M cell properties.

Despite the similarities outlined above, there are several key differences between taste cells and classical M cells. Taste cells do not have the basal pocket found in M cells that houses APCs. In this respect, they are analogous to villous M cells that spontaneously transdifferentiate from enterocytes in apices of intestinal villi [61,62]. Similarly, villous M cells and the taste papillae are not associated with underlying germinal centers [63]. Villous M cell–stimulated T and B cell maturation may occur in the intestinal lamina propria, but it is not known if this occurs in the lamina propria of taste papillae [27]. However, taste papillae from WT mice contain a diverse population of immune cells, and SpibKO mice have far fewer of them, strongly supporting a role for taste cells in immune surveillance (Fig 3A–3F). Of note, although not present in rodents, humans and most other mammals have lingual tonsils located near the CVP and FOP. They contain MALT, where taste cell-stimulated T and B cells may mature, although this might happen in the cervical or other lymph nodes as well. In fact, the mesenteric lymph nodes appear to be the primary sites of T and B cell maturation in intestinal mucosa [64].

What other roles might the M cell–like properties of taste cells serve? Taste cells secrete cytokines such as IL-10 and TNF alpha, which affect taste signaling [49,50,65]. M cells are known to regulate secretion of cytokines in the PP[66]. SpibKO mice express lower levels of proinflammatory cytokines and have fewer immune cells in the taste papillae while the expression of anti-inflammatory cytokines is unaltered (Figs 3A–3F and S6C). The expression of components of the NF-κB signaling pathway that regulate acute phase cytokine gene expression is lower in CVP of SpibKO mice, which might underlie this phenotype (S6A and S6B Fig). Of note, increased T cell recruitment to taste papillae was observed in IL-10 knockout mice, which also had smaller taste buds and fewer taste cells per bud [65]. The lower levels of immune cell recruitment in CVP of SpibKO mice likely reflects down-regulation of cytokines, especially chemokines (Figs 3A–3F and S6C). However, upon LPS stimulation, SpibKO mice show exaggerated inflammatory cytokine gene expression (S7 Fig). These data and the inability of taste cells in SpibKO mice to transcytose luminal particles indicate that mucosal immune responses are impaired in the CVP in the absence of antigen surveillance. Spib is required for the development and regeneration of M cells and plasmacytoid dendritic cells [35,36,67]. Strikingly, the proportion of taste cell subtypes in SpibKO mice is unaltered, suggesting that its role in taste cells is likely restricted to regulating expression of genes involved in mucosal immunity (S8 Fig). A direct role in regulating taste signaling itself remains an open question. The higher behavioral attraction to sweet and umami stimuli in SpibKO mice could arise from reduced expression of proinflammatory cytokines or changes in expression of downstream components of the taste signaling pathway (Fig 4). Our findings indicate that taste cell–mediated immune surveillance is a key aspect of oral mucosal immunity and that the dysregulation of this pathway may lead to microbial infection and taste loss.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3001647

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.