Enteric defensins are essential regulators of intestinal microbial ecology

Abstract

Antimicrobial peptides are important effectors of innate immunity throughout the plant and animal kingdoms. In the mammalian small intestine, Paneth cell alpha-defensins are antimicrobial peptides that contribute to host defense against enteric pathogens. To determine if alpha-defensins also govern intestinal microbial ecology, we analyzed the intestinal microbiota of mice expressing a human alpha-defensin gene (DEFA5) and in mice lacking an enzyme required for the processing of mouse alpha-defensins. In these complementary models, we detected significant alpha-defensin-dependent changes in microbiota composition, but not in total bacterial numbers. Furthermore, DEFA5-expressing mice had striking losses of segmented filamentous bacteria and fewer interleukin 17 (IL-17)-producing lamina propria T cells. Our data ascribe a new homeostatic role to alpha-defensins in regulating the makeup of the commensal microbiota.

Figure 1. Expression of Paneth cell effector genes in Mmp7 ^−/−, DEFA5 tg (+/+) and wild-type (WT) mice

a,b. The absolute mRNA copy numbers of transcripts encoding murine Paneth cell lysozyme (M and P isoforms Lyz1, Lyz2), secretory phospholipase A2 (Pla2g2), α-defensins (cryptdins) 1 (Defcr1) and 4 (Defcr4), cryptdin related sequence (Defa-rs1c), and DEFA5 were determined by quantitative real-time PCR using RNA isolated from the distal small intestine of (a) DEFA5 tg (+/+) (n=3) and wild-type (n=3) FVB mice or (b) Mmp7 ^−/− (n=4) and wild-type (n=4) B6 mice. The absolute mRNA copy numbers were determined using the same primers and standards for all mouse strains, except that the cryptdin 4 used strain specific primers (Supplementary Table 3). Mean values and standard deviations are presented. No statistically significant differences were identified.

Figure 2

Community comparisons and phylogenetic analysis of bacterial composition of the distal small intestine. Genomic DNA was isolated from the distal small intestines of Mmp7 ^−/− (a) and DEFA5 tg (+/+) (b) mice and littermate controls (WT). 16S rRNA sequences were obtained by PCR amplification, subcloning, and sequencing. Using R-package software, the sequence data assigned by the RDP classifier were used for the comparison of microbial communities and the generation of heat maps. For each animal, the number of sequences in each phylum was normalized by the total number of sequences obtained for that animal, producing a percentage. The percentage was used as the matrix for generating the heat map. The color key values represent the percentage (0–100%) of sequences in each phylum.

Figure 3

Quantitative analysis of intestinal bacterial groups. a,b. Bacterial composition of the distal small intestines of Mmp7 ^−/− and DEFA5 tg (+/+) mice was determined by analysis of subclone sequences (a) and qPCR (b). Stacked graphs show relative percentages of dominant bacterial groups in the distal small intestine of offspring from Mmp7^+/− and DEFA5 tg (+/−) breeding pairs with total bacterial sequences (a) or total bacterial copies as determined by amplification with universal bacterial primers (b) used as the denominator for subclone and qPCR analyses, respectively. c,d. Log number of total copies of specific bacterial 16S rDNA in the distal small intestine of each mouse was measured by pPCR. In offspring of DEFA5 tg (+/−) breeding pairs, the presence of SFB was not detectable (ND) by the reliable limits of detection by this assay. * P < 0.05.

Figure 4

FISH of adherent bacteria in mouse distal small intestine. a. Sections (3 micron) of distal small intestine from vendor-obtained wild-type (WT) mice or offspring from DEFA5 tg (+/−) breeding pairs were hybridized with a mixture of oligonucleotide probes recognizing SFB (6Fam-SFB) and total bacteria (TR-Bact338), and visualized by fluorescence microscopy. Where indicated, mice were analyzed 14 days post-gavage with feces containing SFB. All mice were on a pure FVB background. b. Sections of distal small intestine (3 microns) from offspring from Mmp7^+/− breeding pairs were hybridized and visualized as in (a). All mice in (b) were on a pure B6 background. Arrows point to SFB bacteria. Arrowheads point to non-SFB bacteria. Non-specific autofluorescence staining is noted in the mucus when viewing with FITC filters (SFB panels), and does not correlate with any specific bacteria (visualized using Texas Red filters-All Bacteria panels).

Figure 5

Influence of maternal exposure on SFB colonization in mouse distal small intestine. a,b. Sections of distal small intestine (3 microns) from 5-week-old offspring of indicated breeding pairs were analyzed by FISH as in Fig. 4 to detect epithelial associated SFB. Arrows point to SFB. Arrowheads point to non-SFB bacteria. Non-specific auto-fluorescence may be noted in the mucus in the FITC channel (SFB panels), but this does not correspond to any bacteria noted in the Texas Red channel (All Bacteria panels). c. Log numbers and percentages of SFB 16S rDNA in the distal small intestine were analyzed by qPCR. Only DEFA5 tg (+/−) offspring from indicated breeding pairs were analyzed (n=9–11 mice per litter). *** P=0.0002.

Figure 6

Quantitative comparison of total bacteria by intestinal segment. Log number of total copies of bacterial 16S rDNA were determined for distal small intestine (DSI), cecum, and large intestine (LI) of each mouse by qPCR, using universal bacterial primers, and quantified using a standard curve generated from Ruminococcus productus bacterial genomic DNA as a reference strain. Offspring from DEFA5 tg (+/−) (a) and Mmp7^+/− (b) breeding pairs were analyzed.

Figure 7

T_H17 profile of distal small intestinal LPL. (a) Representative flow cytometry analysis of IL-17A and IFN-γ expression in lamina propria CD4⁺ T cells from indicated mice after stimulation with PMA and ionomycin. (b) The percent and the number of CD4⁺ IL-17A⁺ T cells in LPL from distal small intestine of the indicated groups of mice. Each dot represents an individual mouse. ** P<0.001.