VSports app下载 - Composition and functional role of the mucus layers in the intestine

Epithelial cells and their glycocalyx

Before we look at the enterocytes and their glycocalyx, we should consider its dimensions in relation to the normal mucus of colon. As illustrated in Fig.  2a, the distal colon is protected by a two-layered mucus that reaches up to a millimeter from the epithelial cells [1, 9]. The inner of these layers is impermeable to bacteria and is 50–100 μm thick in mice and rat and probably even thicker in humans. The two mucus layers are formed by the gel-forming MUC2 mucin, with MUC2 being their main component and forming their skeleton. Below the two layers is the glycocalyx of the enterocytes, which has been estimated to reach at least 1 μm from the apical cell membrane. Although formally not proven, the transmembrane mucins MUC3, MUC12, and MUC17 are the major components of the intestinal glycocalyx. In addition, the very short MUC13 mucin is also found in the intestinal epithelium and is contributing to its protection [10]. The PTS domains of MUC3, MUC12 and MUC17 are 4,000–5,000 amino acids long (see http://www. medkem. gu. se/mucinbiology/databases) and when the abundant Ser and Thr have been O-glycosylated they form long and extended rods that are best described as bottle brushes VSports app下载. Assuming the same length per amino acid as determined for the mucin domains of MUC2 [11], the mucin domains of MUC3, MUC12, and MUC17 are about 0. 8 μm long. These mucins are densely packed and will in this way provide a barrier with similar properties as those of the mucus layers on the luminal side. The glycocalyx has not been studied in great detail, but it is assumed that they are quite similar on the enterocytes in the small and large intestine. .

MUC3, MUC12, and MUC17 belong to the family of SEA-(sea urchin sperm protein/enterokinase/agrin) domain-containing TM mucins and are located at the 7q22 locus [3]. The single SEA domain in these mucins is located extracellularly, close to the cell membrane (Fig. 2b), and have a sequence similar to that of the SEA domain in the MUC1 mucin [12]. The structure of this domain has been determined by us and shows a tightly packed domain with four anti-parallel β-pleated sheets, two in the outer domain and two in the membrane-anchored domain, with an intervening cleavage site (Fig. 2b). Interestingly, this cleavage is necessary for proper folding of the domain in the endoplasmic reticulum (ER) and this autoproteolysis, which occurs between the amino acids Gly and Ser, is in fact triggered by the folding energy [12]. To have such a break-point in the polypeptide chain, close to the membrane, might be a way to protect the cell from physical stress by allowing the two parts of these mucins to separate, while maintaining an intact epithelial cell layer.

The function of the TM mucins is most likely primarily cell protection, but they can probably also act as sensors for the luminal milieu, as suggested by the structure of the SEA domain. There are a few hydrophobic patches in the SEA domain, where other proteins might bind [12]. Moreover, the MUC3, MUC12, and MUC17 mucins all have a cytoplasmic C-terminal PDZ-binding motif that we have shown to bind several of the cytoplasmic PDZ proteins [13, 14]. These proteins are scaffolds for organizing and sorting transmembrane proteins, especially in polarized cells [15]. Many of the membrane proteins with C-terminal PDZ-binding motifs are ion channels, suggesting a connection between the TM mucins and ion transport. In fact, MUC17 binds PDZK1 and MUC3 binds GOPC, both being PDZ proteins known to interact with the CFTR ion channel [13, 14].

The gel-forming mucin MUC2 forms the skeleton of the intestinal mucus

As already pointed out, the gel-forming mucin MUC2 is the major component of the mucus layers of the intestine, which are produced by the goblet cells. The mucus is built around a primary MUC2 translation product of about 5,200 amino acids (this is not absolute, as the gene has not yet been fully sequenced and also because of genetic variation in the number of tandem repeats). The protein is translocated into the ER lumen where it folds and forms disulfide-bonded dimers due to the cysteine-knot in its far C-terminal end. The folding of MUC2 is a challenge for the cell due to the abundant Cys residues and there is a need for special chaperone proteins like AGR2 for this process [16]. The ER stress response system probably needs to be at least partly activated to increase the number of chaperones, as lack of the master regulator of the ER stress response, XBP1, causes accumulation of misfolded mucins and also inflammation [17].

In the Golgi apparatus, the MUC2 dimers are decorated with O-glycans to reach a molecular mass in the range of 5 MDa. The dimers are then sorted to the regulated secretory pathway where the mucins are densely packed in secretory vesicles. At this stage, the N-terminal ends of the dimers are coupled together by disulfide bonds into trimeric structures, Fig. 3a [18]. Once released from the cells, the MUC2 molecules are expanded and will form large net-like structures due to the C-terminal dimers and the N-terminal trimers (Fig. 3b). Such a model should have a built-in property to form flat sheets.

The disulfide-bonded di- and trimers are not the only interaction points for the MUC2 mucin. Several years ago Carlstedt et al. [19] discovered that MUC2 is also interconnected via non-reducible bonds. The chemical nature of this bond is still unknown, but it is formed during biosynthesis in the goblet cells and it also gives the MUC2 mucin its typical property of insolubility in chaotropic salts [20]. Moreover, all the human gel-forming mucins, except MUC6, have two or more CysD domains interrupting the PTS domains [3]. These domains are 110 amino acid long, with ten conserved Cys amino acids that form five intramolecular disulfide bonds [21]. The MUC2 mucin has two such domains and we could recently show that its second CysD domain forms very stable, non-covalent dimers [21]. It is very likely that other CysD domains have similar properties and that they are there to form non-covalent cross-links in the mature mucus gel for further stability (Fig. 3c). However, this would imply that the CysD domains do not interact during vesicular storage, before secretion of the mucin. The exact mechanisms for this remain to be elucidated, but may involve differences in Ca²⁺-concentrations and pH. Finally, we have discovered that there is a large protein called FCGBP (immunoglobulin Fc γ binding protein) that is covalently attached to the MUC2 mucin in the colonic mucus [22]. FCGBP has 13 von Willebrand D domains (vWD), 11 of which have a GDPH (Gly–Asp–Pro–His) sequence suggesting an autocatalytic cleavage, as found in the fourth vWD of the MUC2 mucin. This was shown to be cleaved at low pH between the Asp and Pro and by this generating a new C-terminus that forms an internal anhydride with the side-chain of the Asp [23]. The vWD domain is still held together via disulfide bonds across the cleavage site, bridging the cleaved parts of the protein. The newly formed anhydride is very unstable and will react with primary amines or hydroxyl groups (but only if not hydrolyzed in water). As FCGBP has up to 11 of these attachment sites, it can be predicted to cross-link components of the mucus to the MUC2 mucin. However, it is not known how and when the GDPH cleavage is triggered, nor if the attachment sites are random or if there is a specificity. A similar system connecting bikunin with heavy chain 3 of the pre-α-inhibitor protein has been shown to be specific for the C6 hydroxyl group of GalNAc in the bikunin chondroitin sulfate [24]. Taken together, this suggests that the MUC2 mucin can be further covalently cross-linked by endogenous properties of the FCGBP protein once the mucus gel has been secreted and formed.

The basic structure of the MUC2 mucin thus has a net-like appearance that can be further cross-linked by both non-covalent and covalent mechanisms. This suggests that the mucus gel, as organized by MUC2, is well structured and well stabilized.

The mucus of the small intestine

The small intestine is protected by mucus that has the MUC2 mucin as its main and structural component, just as in the large intestine. Despite this, the properties of the mucus in these two locations are different. As described by Atuma et al. [9], the small intestine has a mucus that is easy to aspirate and remove. We recently extended our studies of the small intestine, using explant systems. In the ileum, the mucus covers the villi (Fig. 4) and this mucus is easy to aspirate, removing the mucus all the way down to the crypt openings. Atuma et al. have, in contrast, in their experimental set-up, seen a thin layer of non-removable mucus in the small intestine. These differences may be caused by the technical difficulties in removing mucus in the small intestine due to the presence of villi. Our preliminary studies also suggest that this small intestinal mucus is permeable to bacteria. The observations that the ileal mucus has only one layer, is not attached to the epithelium, and is permeable to bacteria are all in contrast to the large intestine. The molecular reasons for why MUC2 is forming another type of mucus in the small as compared to the large intestine are not understood today.

"V体育ios版" The mucus of the distal large intestine

As already pointed out, the colon has a two-layered mucus (Fig. 5a). The inner mucus layer is formed from the goblet cell secretion, remains attached to the epithelium and is not possible to remove experimentally by gentle aspiration. This inner layer has a stratified appearance as could be predicted from the previously discussed model of the MUC2 polymers, with flat sheets of MUC2 (Fig. 3). This inner layer has special physical properties as it is not soluble in for example guanidinium chloride [1]. This layer is then converted to the outer mucus layer by processes that are not yet understood, followed by a protease-dependent fourfold volume expansion [25]. The outer layer can be aspirated and this mucus layer is thus also called the loose mucus. Immunostained tissue sections, made with techniques that preserve the mucus, show a sharp border between the inner and outer layer, suggesting a very well controlled system. As germ-free animals also show both an inner firm layer and an outer loose layer, this conversion and expansion must happen through endogenous mechanisms and proteases [1]. Using the experimental explant tissue set-up, the inner layer is determined to be about 50 μm in mice, the same thickness as in live animals [1]. As also observed in live animals, the mucus layer grows over time, as shown in Fig. 5b for distal colon. After 45 min, the mucus has grown almost 100 μm (Fig. 5b). Thus, the colon has a sophisticated mucus protection system that is under control of the host, even though it is located at a large distance from the epithelial cell surface.

Localization of bacteria in the distal large intestine

With this elaborate mucus system in the colon, it was obvious to ask where the commensal bacteria were to be found. Surprisingly, the bacteria were found only in the outer loose mucus layer and were totally absent in the inner attached layer [1]. In fact, the commensal bacteria have the outer mucus layer as their habitat. This is illustrated in Fig. 6, where the commensal Lactobacilli are revealed by specific in situ hybridization in the outer mucus layer, with the inner mucus layer being totally devoid of bacteria [26]. Thus, the inner mucus layer is impermeable to bacteria. This organization is probably critical for a normal homeostasis and also effectively separates bacteria from cells of the immune system. This means that the immune system does not need to monitor bacterial friends from bacterial foes in the colon. This model has led to the questioning of several of the more popular models of how the immune system handles the commensal flora of the colon. Interestingly, lack of the MUC2 mucin allow the bacteria to come in direct contact with the epithelial cells, penetrate down into the normally sterile crypts, and even into the epithelial cells [1]. These mice develop a relatively severe inflammation with diarrhea, rectal prolapse, weight loss and, after some time, also colon cancer [1, 27, 28], a situation that is similar to the human disease ulcerative colitis.

Why is the outer mucus layer such a good habitat for the commensal microbiota? A first prerequisite is that the bacteria are able to penetrate this mucus. Secondly, the bacteria can use the high number of glycans on the mucins as attachment sites. Many bacteria are known to carry adhesins of lectin-type that can be used to anchor the bacteria. As the glycan repertoire varies considerably between species, it can be suggested that there are only certain bacteria that has the appropriate adhesins to adapt to each host. In fact, it is now known that there is a core microbiota in each species, supporting this model [29] and that the glycans on mucosal surfaces also varies dramatically [30, 31]. Further proof for such a model is our recent observation that the MUC2 O-glycans in the sigmoid colon are uniform in all humans, except in individuals with an active colon inflammation [32, 33]. This is in contrast to the mucin O-glycans on other mucosal surfaces of the body. Together, these observations suggest a co-evolution between host and bacteria for an optimal symbiotic relation. Last but not least, all the oligosaccharides present on the mucins provide an enormous food supply for the bacteria. It is well known that the commensal bacteria have an especially rich source of glycan-degrading enzymes [34]. These enzymes are typically of the exoglycosidase type that removes one sugar residue at a time, starting from the non-reducing end. This is probably important, as this will slow down the degradation that will finally reach the protein core of the mucin, and when this is degraded the whole mucin polymer network will be dissolved. Such a scenario is suggested by our recent studies, in collaboration with Dr. Lijun Xia, of mice lacking the Core 1 glycosyltransferase. This glycosylation deficiency will give rise to a MUC2 mucin with shorter O-glycans and this will allow a faster degradation. These animals develop a relatively severe colitis, just as the animals that are totally devoid of MUC2 [35]. The released sugars are then utilized as food for the bacteria, but they are also converted to short fatty acids that can diffuse through the mucus and reach the epithelium, which will in this way regain some of the energy used for the biosynthesis of the quickly turned-over mucus.