Antimicrobial peptides and the enteric mucus layer act in concert to protect the intestinal mucosa

Abstract

The intestinal mucosa squares the circle by allowing efficient nutrient absorption while generating a firm barrier toward the enteric microbiota, enteropathogenic microorganisms and high luminal concentrations of potent immunostimulatory molecules. The mucus layer together with local antimicrobial and anti-inflammatory peptides significantly contribute to this ability. Here we summarize the recent progress made to better understand the critical importance of this dynamic, complex and highly structured anti-inflammatory and antimicrobial barrier.

Abbreviations

CRS

Cryptdin-related sequence

HD

Human defensin

IAP

Intestinal alkaline phosphatase

LBP

Lipopolysaccharide binding protein

RegIIIγ

Regenerating islet-derived protein 3 γ

SIgA

Secretory immunoglobulin A.

The enteric mucosa facilitates efficient nutrient and fluid uptake while preventing translocation of macromolecules and commensal and enteropathogenic bacteria from the densely colonized gut lumen. During the last decades, a detailed understanding of the role of the epithelium as well as mucosal myeloid and lymphoid immune cells in host-microbial homeostasis, inflammation and antimicrobial host defense was obtained. Yet, the critical role of the most upper protective layer of the enteric surface barrier, the mucus layer produced by specialized epithelial cells named goblet cells has remained largely undefined. This was in part due to the fact that formalin or cryo-fixation, the 2 most commonly used methods to preserve tissues for histological analysis, fail to preserve the delicate mucus structure. As a consequence, standard histological images incorrectly suggested a direct contact between the epithelium and the luminal microbiota. In addition, commonly used epithelial cell culture models failed to produce significant amounts of mucus, and the mucus matrix itself, composed of highly glycosylated and interlinked proteins, was not amenable to standard molecular biological analysis. Despite these technical obstacles, a number of studies during the past decade have revealed the synthesis and structural organization of the mucus matrix and its critical role during homeostasis and antimicrobial host defense.^1-3

Using the mucus preserving Carnoy fixative, Matsuo et al. and Swidsinski et al. demonstrated the existence of a bacteria-free zone separating the enteric epithelium from the microbiota.^4-5 Later, Johansson et al. showed the presence of bacteria within the outer, but not the inner, colonic mucus layer consistent with the finding that mucus represents an important nutrient source of enteric commensals.^6,7 The structural organization of the firmly adherent inner mucus layer was later shown to physically hinder bacterial penetration in the colon.^8,9 In contrast, the mucus in the small intestine only consists of one layer, probably to facilitate nutrient absorption.¹⁰ In this organ, the mucus layer completely covers the crypts and lower part of the small intestinal villi and may leave the upper villus area exposed to the intestinal lumen content. Although it was shown to be permeable to particles with the size of a bacterium, a bacteria-free exclusion zone overlaying the epithelium was also observed in the murine small intestine.^11,12 The mucus layer therefore seems to act as a physical barrier to keep the local microbiota at a distance from the epithelium.

Figure 1.

Antimicrobial and anti-inflammatory properties of the intestinal mucus barrier. In the small intestine, Paneth cells are the main producers of antimicrobial peptides. Antimicrobial peptides generate a gradient within the mucus layer from the cell surface to the lumen and from the bottom of the crypt (where the Paneth cells are located) to the villus tip. In the colon, the antimicrobial peptide gradient is reinforced by a bi-layered mucus matrix. The lower mucus layer represents a physical barrier for penetrating bacteria due to its dense net structure. Due to the anti-inflammatory and antimicrobial nature of antimicrobial peptides, the mucus layer generates a protective shield to prevent bacterial translocation and inappropriate immune stimulation of the mucosal tissue. It also restricts the action of antimicrobial peptides mostly to the epithelial surface. Antimicrobial peptides are subject to both constitutive and inducible expression. PAMP, pathogen-associated molecular pattern.

We and others showed that the intestinal mucus material harbors significant antibacterial activity whereas only low activity was detected in the luminal content.^1,13 Immunostaining and proteomic analysis of the intestinal mucus identified lysozyme as well as a number of antimicrobial peptides with well-known antibacterial function in the mucus layer (see Table 1).^1,2,13,14 Antimicrobial peptides possess a broad spectrum activity against most pathogenic but also commensal bacteria.² In the small intestine, Paneth cells situated at the crypt basis are the main producers of antimicrobial peptides, mainly α-defensins and cryptdin related sequence (CRS) peptides as well as lysozyme. Lack of proteolytic defensin maturation or transgene expression of human defensin (HD) 5 in mice alters the susceptibility to enteropathogenic infection as well as the microbiota composition.^15-17 Similarly reduced Paneth cell numbers were described in a number of animal models and might contribute to an enhanced susceptibility to inflammation or infection.^18-21 Upon secretion, antimicrobial peptides can reach a concentration of up to 100 mg/mL within the crypt lumen.²² The positively charged antimicrobial peptides appear not to diffuse into the gut lumen but to remain attached to the polyanionic sulfated and sialylated mucin glycoproteins.^1,2,13 The enrichment of antimicrobial peptides within the mucus layer facilitates the peptide concentrations required to efficiently kill bacteria and simultaneously reduces the pressure on luminal commensal bacteria to acquire resistance. The antimicrobial peptide-mucin interaction does not affect peptide activity and is reversible.^2,13 In accordance, cryptdin-coated bacteria were visualized within the mucus layer.² Also, the c-type lectin regenerating islet-derived protein 3 gamma (RegIIIγ) produced by both Paneth cells and small intestinal epithelial cells was localized to the mucus layer, reducing the number of epithelium associated bacteria in vivo.¹¹ The mucus layer thus represents a physicochemical shield that traps and kills bacteria coming too close to the body surface. As a functional homolog, the human α-defensin (HD)6, which possesses no detectable antimicrobial activity, was recently shown to act by trapping bacteria in a peptide-derived net structure.²³

Table 1.

Antimicrobial peptides and other related molecules identified in human and mouse intestinal mucus material

Antimicrobial molecules detected in intestinal mucus material	Reference
Human β-defensin (HBD) 1 and 3	13
Mouse α-defensin (cryptdin) 1, 2, 3, 4 and 6	1-2;14
CRS 1C (mouse)	14
LL-37 (human)	13
Lysozyme (human and mouse)	1;13-14
Angiogenin 4 (mouse)	14
SIgA (human)	28
LBP (mouse)	26
Histone H2A and H2B (human and mouse)	13-14
HMGN2 (human and mouse)	13-14
Thymosin-β (mouse)	1;14
Ubiquicidin (human and mouse)	1;13

Even when killed, bacteria can still release a magnitude of potent immunostimulatory molecules such as endotoxin, lipopeptides, nucleic acids or peptidoglycan fragments that are able to penetrate the mucus matrix. Translocation of these molecules to systemic body sites must be prevented to avoid an inappropriate inflammatory reaction. Recent work has highlighted the anti-inflammatory activity of antimicrobial peptides also within the mucus matrix.^2,24,25 The presence of antimicrobial peptides in bacteria/host cell co-cultures simultaneously caused bacterial killing and abrogated the cellular innate immune response.² With respect to the enteric mucus layer, this could allow immunologically “silent” killing and help to restrict the proinflammatory activity of microbiota-derived immunostimulatory molecules to the gut lumen. Consistent with this hypothesis, a 1000-fold gradient of the endotoxin concentration was found between the intestinal lumen and the epithelial cell layer.² Additional molecules with potent anti-inflammatory properties such as lipopolysaccharide binding peptide (LBP), intestinal alkaline phosphatase (IAP) or secretory IgA (sIgA) were identified within the mucus layer.^26-28 LBP, a potent modulator of lipopolysaccharide recognition, is highly enriched in enteric mucus.²⁶ IAP, expressed by intestinal epithelial cells, dephosphorylates the bioactive lipid A portion of the endotoxin molecule and renders it less stimulatory, significantly reducing the systemic toxicity of microbiota-derived endotoxin and promoting mucosal wound healing.^29,30 The finding that IAP could not be detected at the mucosa of mucin (Muc) 2-deficient mice suggests that IAP also needs to bind to the intestinal mucus layer after being secreted by the epithelial cells.²⁷ Finally, SIgA within the enteric mucus layer provides high affinity protection against enteric pathogens and toxins but additionally provides low affinity binding to commensal bacteria and thereby reinforces the antimicrobial and anti-inflammatory barrier.³¹

Epithelial signaling through the interleukin-1 and Toll-like receptor adapter molecule MyD88 as well as the MyD88 adaptor-like (Mal) was shown to promote epithelial integrity in models of enteric inflammation and infection.^2,32-35 The phenotype of MyD88^−/− mice might at least in part be the result of a dysfunctional mucus layer since mucus production, antimicrobial peptide and RegIIIγ expression, as well as luminal SIgA transport were all shown to require intact epithelial MyD88 signaling.^11,36 Accordingly, the thinner inner colonic mucus layer observed in germ-free mice was restored to the size of the one observed in conventional mice by administration of Toll-like receptor ligands.³⁷ Also stimulatory cytokines such as tumor necrosis factor (TNF) and interleukin (IL)-13 have been shown to induce mucin gene expression and interferon-γ receptor signaling was implicated in goblet cell secretion upon bacterial infection.³⁸ Of note, antimicrobial peptide production in the small intestine and in the colon is subject to both constitutive and regulated expression. Intestinal RegIIIγ expression is constitutive in conventional mice but strongly enhanced by epithelial innate immune stimulation as well as IL-22 receptor activation. The colonic epithelium additionally expresses constitutively produced and transcriptionally regulated β-defensins. In the small intestine, Paneth cell α-defensin and CRS peptide expression occurs mostly constitutively and only to a minor degree depends on transcriptional regulation (for a review, see ref 39). Paneth cell secretion into the crypt lumen, however, is induced by microbial, neurogenic and endogenous stimuli.^22,40

The critical role of the mucus layer in host microbial homeostasis and antimicrobial host defense was confirmed using mucin-deficient animals. Mice deficient in the main intestinal mucus constituent, muc2, suffer from increased epithelial cell turnover, crypt elongation and spontaneous intestinal inflammation.^41,42 Inflammatory changes are detected after weaning, i.e. with uptake of solid food and the establishment of a mature enteric microbiota consistent with the rise in mucin expression during this time period (own unpublished observations).⁴³ Muc1^−/− and Muc2^−/− mice also exhibit enhanced susceptibility to Campylobacter jejuni, Salmonella enterica or Citrobacter rodentium infection.^27,44,45 Similarly, mice deficient in muc13 or core 3 ß1,3-N-acetylglucosaminyl-transferase (C3GnT) with therefore a reduced mucin O-glycosylation exhibit an enhanced susceptibility to inflammation.^46,47 Also, aberrant mucin assembly-associated ER stress in goblet and Paneth cells was reported in 2 strains of muc2 mutant mice leading to spontaneous inflammation and increased mucosal permeability.⁴⁸ Altered mucin expression and reduced antimicrobial peptide expression has also been observed in patients with ileal Crohn's disease or ulcerative colitis.^49,50 Also, decreased goblet cell numbers and enhanced permeability of the inner mucus layer for bacteria or beads of a similar size was described in patients with ulcerative colitis.⁵¹ Thus, a reduced antimicrobial-mucus shield might cause and/or maintain inflammation in patients with inflammatory bowel syndrome.

Similar to the situation in vivo, oral administration of a synthetic anti-inflammatory and antibacterial peptide led to enrichment within the small intestinal mucus layer in mice. Although significant alterations of the microbiota composition were noted, this approach resulted only in a temporary improvement of the clinical course of enteric infection with Salmonella probably due to insufficient local peptide concentration.² Topic application of a peptide-enriched mucus-like material on inflamed mucosal tissue e.g., by endoscopy might allow sufficient local peptide concentration and facilitate an improved clinical outcome and warrants further investigations. On the other hand, the well-orchestrated regulation and compartmentalized production of the antimicrobial and anti-inflammatory mucus layer in vivo illustrates the complex requirements of this protective mechanism during host-microbial homeostasis (Fig. 1). Stimulation of endogenous antimicrobial peptide secretion and/or mucus production might therefore represent a more effective future therapeutic strategy in patients with mucosal barrier disruption.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Funding

AD KB and MWH received support from the German Ministry for Science and Education BMBF 01GU0825. In addition, MWH was supported by the individual grant Ho2236/8-1, the Collaborative Research Center SFB900 and the DFG priority program SPP 1656 from the German Research Foundation.