NLRP6 inflammasome orchestrates the colonic host-microbial interface by regulating goblet cell mucus secretion

Abstract

Mucus production by goblet cells of the large intestine serves as a crucial antimicrobial protective mechanism at the interface between the eukaryotic and prokaryotic cells of the mammalian intestinal ecosystem. However, the regulatory pathways involved in goblet cell-induced mucus secretion remain largely unknown. Here, we demonstrate that the NLRP6 inflammasome, a recently described regulator of colonic microbiota composition and biogeographical distribution, is a critical orchestrator of goblet cell mucin granule exocytosis. NLRP6 deficiency leads to defective autophagy in goblet cells and abrogated mucus secretion into the large intestinal lumen. Consequently, NLRP6 inflammasome-deficient mice are unable to clear enteric pathogens from the mucosal surface, rendering them highly susceptible to persistent infection. This study identifies an innate immune regulatory pathway governing goblet cell mucus secretion, linking nonhematopoietic inflammasome signaling to autophagy and highlighting the goblet cell as a critical innate immune player in the control of intestinal host-microbial mutualism. PAPERCLIP:

Figure 1. NLRP6 protects from enhanced enteric infection

WT and Nlrp6^−/− mice were infected with 10⁹ CFU of bioluminescent C. rodentium and analyzed on day 15 p.i., unless otherwise stated. (A) In vivo whole body bioluminescence imaging of WT and Nlrp6^−/− mice on day 9 p.i. show increased bacterial growth in Nlrp6^−/− mice. (B) Both luminal (fecal matter) and adherent (extensively washed colons) bacterial colonization is enhanced in Nlrp6^−/− mice. Results are pooled from two separate experiments, n=12–14 per group. Significance determined using the Mann-Whitney U-test. (**p<= 0.0033; ****p < 0.0001). (C) H&E stained distal colon sections from WT and Nlrp6^−/− mice show an increase in inflammation and crypt ulceration throughout the mucosa of Nlrp6^−/− mice. Magnification = 5×, 10×; scale bar = 200 μm. (D) Histopathology scores from distal colon tissues of Nlrp6^−/− and WT mice. Each bar represents one individual mouse and shows scores for damage to the submucosa, mucosa, surface epithelium and lumen, n= 9 per group. (****p< 0.0001) (E, F) Secretion of pro-inflammatory cytokines in the colon (E) and spleen (F) is unchanged between WT and Nlrp6^−/− mice. Results are pooled from two separate infections of WT and Nlrp6^−/− mice, n=13 and 14, respectively. (G) C. rodentium-specific colonic IgA and systemic IgG titers. Results are pooled from two separate experiments, n=9–13 per group. (H–J) Quantitative RT-PCR showing expression of IL-22 (H), Reg3β (I) and Reg3γ (J) relative to gapdh in the distal colon of WT and Nlrp6^−/− mice over the course of C. rodentium infection, n= 4–9. See also Fig. S1.

Figure 2. Inflammasome signaling is required for clearance of C. rodentium infection

WT, Asc^−/− and Caspase-1/11^−/− mice were infected with 10⁹ CFU of bioluminescent C. rodentium and analyzed on day 9 post infection. (A, B, F, G) Representative images (A, F) and time course quantification (B, G) of in vivo whole body bioluminescence imaging shows elevated bacterial growth in the intestine of Asc^−/− (A, B) and Caspase-1/11^−/− mice (F, G). (C, H) Ex vivo imaging of extensively washed colonic explants shows enhanced bacterial attachment to colons of Asc^−/− (C) and Caspase-1/11^−/− (H) mice. (D, E, I) Bacterial plating demonstrates a higher colonic and systemic colonization of Asc^−/− (D, E) and Caspase-1/11^−/− (I) mice.

Figure 3. NLRP6 is expressed in goblet cells

(A) Analysis of NLRP6 expression in sorted colonic epithelial and hematopoietic (CD45⁺) cells. The purity of the sorted populations was analyzed by RT-qPCR using vil1 and ptprc as markers for epithelial and hematopoietic cells, respectively. NLRP6 expression closely mirrored that of colonic epithelial cells.(B–D) In situ hybridization with an NLRP6-specific probe, visible as black dots, with an H&E counter stain. The theca (housing all mucin-containing granules) within goblet cells is not stained with H&E and identified as un-stained circles allowing localization of goblet cells within the epithelium (outlined with black circles). (B) Representative localization of NLRP6 in a WT distal colon section, showing that staining is concentrated in the apical region of the epithelium. Magnifications demonstrate an enrichment of NLRP6 mRNA in proximity to goblet cells, seen as increased probe-binding to areas surrounding the theca of goblet cells. (C) As in (B), but in Asc^−/− mice. (D) No nonspecific probe binding is seen in Nlrp6^−/− distal colon sections.

Figure 4. NLRP6 inflammasome activity is required for goblet cell function and protection from C. rodentium invasiveness

(A) AB/PAS stained distal colon sections of WT, Nlrp6^−/−, Asc^−/−, and Caspase-1/11^−/− mice showing the inner mucin layer (“i”) and goblet cells (asterisks). Magnification = 400×; scale bar = 50 μm. (B) Quantification of inner mucus layer thickness in the distal colon. The inner mucus layer is absent in Nlrp6^−/− and Asc^−/− mice and significantly thinner in Caspase1/11^−/− mice, n= 8, 4, and 5 mice, respectively (***p = <0.0001). (C) Quantification of goblet cell number in the distal colon. Nlrp6^−/− (***p = 0.0001), Asc^−/− (***p = 0.0001) and Caspase1/11^−/− (***p = 0.0007) mice exhibit goblet cell hyperplasia, n= 8, 4, and 5 mice, respectively. (D) Representative transmission electron microscopy images taken from colonic sections of WT and Nlrp6^−/− mice, n= 4 mice per group. (E) Representative epifluorescence staining for mucus using the lectin UEA-1 (green) with DAPI (blue) as a counter stain. The inner mucin layer is absent in Nlrp6^−/− mice. i = inner mucin layer. Original magnification = 200×; scale bar = 50 μm. (F) Representative immunostaining for the C. rodentium effector Tir (green) and the mucus specific protein Muc2 (red) in colon, with DAPI (blue) as a counter stain, in WT and Nlrp6^−/− mice at 7 days p.i. The inner mucus layer is visible in WT mice and is lacking in the Nlrp6^−/− mice. Magnification = 200X; scale bar = 50 μm, i = inner mucus layer. (G) In Nlrp6^−/− mice, C. rodentium (green) appears to be more invasive, as shown by deeper penetration into the crypts, which often co-localizes with muc2 (red). See also Fig. S2 & Fig. S3.

Figure 5. NLRP6 inflammasome is required for mucus granule exocytosis

(A) Representative AB/PAS stained colon sections showing the inner mucus layer (i) in WT mice. Nlrp6^−/− mice show the presence of mucus granules-like structures within the lumen (inset “a”). Scale bar = 50 μm. (B) AB/PAS stained Nlrp6^−/− distal colon section showing accumulation of mucus granule-like structures in the lumen (arrowhead) and an increased number of large PAS⁺ goblet cells (asterisks). Scale bar = 50 μm. (C) Representative immunostaining for the goblet cell specific protein, Clca3 (green), with DAPI (blue) as a counter stain in distal colon sections. Arrowheads show diffuse staining of Clca3 in the WT lumen and punctate staining in the Nlrp6^−/− lumen. Representative transmission electron microscopy images (insets a and b) show intact mucus secretion by a goblet cell in WT and dysfunctional mucus granule exocytosis and the presence of granule-like structures in Nlrp6^−/− distal colon tissue. (D) Transmission electron microscopy image of the Nlrp6^−/− distal colon showing protrusion of mucus granules into the lumen without mucus secretion and intact mucus granules saturating the intestinal lumen, n= 4 mice. (E) Representative scanning electron microscopy images of the distal colon of WT and Nlrp6^−/− mice, n= 2 mice per group. Each experiment was repeated 3 times. A smooth intestinal epithelium is seen in WT mice. A large number of goblet cells with mucus granules protruding into the lumen (arrowheads) are seen in Nlrp6^−/− mice. (F) Enlarged scanning electron microscopy image of four goblet cells with protruding mucus granules into the Nlrp6^−/− intestinal lumen. See also Fig. S4.

Figure 6. NLRP6 is required for autophagosome formation in the intestinal epithelium

(A) Representative immunofluorescence image of WT (LC3, top panel) and NLRP6-deficient (LC3:Nlrp6^−/− , bottom panel) intestinal epithelium shows abrogated autophagy in the absence of NLRP6. Goblet cells are stained with the mucus specific protein Muc2 (red), epithelial cell nuclei are indicated with DAPI (blue). Formation of autophagosomes is visualized utilizing the LC3-GFP endogenously expressed protein (green). Scale = 70 μm. (B) Magnification of intestinal epithelial cells showing WT goblet cells (Muc2 positive; red) active in the formation of autophagosomes, seen as punctate staining with the LC3-GFP endogenous protein co-localizing with Muc2 positive cells. (C) Quantitation of autophagosome formation through enumeration of LC3 puncta per 100 epithelial cells, n= 5 mice per group (***p < 0.0001). (D) Immunoblot analysis of total LC3-GFP, and p62 proteins in isolated intestinal epithelial cells from WT LC3-GFP transgenic mice and NLRP6-deficient GFP-LC3 transgenic mice. (E) LC3-GFP band intensities from (D) were quantified and normalized to actin band intensity, n= 5 mice per group (**p = 0.0067). (F) Immunoblot analysis of total endogenous LC3-I/II and p62 proteins in isolated intestinal epithelial cells of WT, Nlrp6^−/−, Asc^−/− and Caspase-1/11^−/− mice. LC3-I and LC3-II denote the nonlipidated (cytosolic) and lipidated (autophagosome membrane) forms of LC3, respectively. (G) Accumulation of LC3-I in isolated epithelial cells from Nlrp6^−/− (**p = 0.0015), Asc^−/− (**p = 0.0013) and Caspase-1/11^−/− (**p = 0.0025) mice, as shown by the fraction of LC3-I band density out of total LC3 band density. Data represent n= 6 (WT, Nlrp6^−/−, Asc^−/−) or n= 4 (Caspase-1/11^−/−) mice. (H) Increased abundance of p62 in Nlrp6^−/− (*p = 0.0349), Asc^−/− (ns, p = 0.2115) and Caspase-1/11^−/− (*p = 0.0284) mice, as shown by quantification of p62 band intensity. Data represent n= 6 (WT, Nlrp6^−/−, Asc^−/−) or n= 4 (Caspase-1/11^−/−) mice. (I) Mitochondria were scored and enumerated in WT and Nlrp6^−/− intestinal epithelial cells as healthy (black), unhealthy (red) and dense inclusion body containing (blue), n= 25 or 28 epithelial cells, respectively. Mitochondrial dysfunction was characterized in Nlrp6^−/− mice as a decrease in total healthy mitochondria (***p < 0.0001) and an accumulation of unhealthy (***p < 0.0001) and dense inclusion body-containing (***p = 0.0002) mitochondria. Representative transmission electron microscopy images are shown (magnification = 11500×) and healthy (black asterisk), unhealthy (red asterisk), and dense inclusion body containing (blue asterisk) mitochondria are depicted within WT and Nlrp6^−/− intestinal epithelial cells.

Figure 7. Autophagy is required for goblet cell function and mucus secretion in the intestine

(A) Representative AB/PAS stained colon sections showing the inner mucus layer (i) in WT mice. Atg5 heterozygous mice show reduced production of the inner mucus layer and goblet cell hyperplasia (asterisk). Scale bar = 50 μm. (B) Quantification of inner mucus layer thickness in the distal colon. The inner mucus layer is significantly thinner in the Atg5^+/− distal colon, n= 3 mice (***p = <0.0001). (C) Quantification of goblet cell number in the distal colon. Atg5^+/− mice exhbit goblet cell hyperplasia, n= 3 mice (**p = 0.0030). (D) Transmission electron microscopy image of Atg5^+/− showing reduced mucus secretion. Theca of WT mice fuse with surface of epithelium resulting in mucus granule shedding and release of contained mucins. Fusion and granule release is stalled in Atg5^+/− mice.