Background
The mammalian colon is covered by a constantly renewing, double mucus layer that forms the first line of defence in the colon. The main component of both layers is the highly glycosylated Mucin 2 (MUC2) which is produced, stored and secreted by goblet cells [
1]. Increased mucus secretion in the crypts is a response to stress stimuli such as bacterial invasion [
2,
3]. The elevated secretion facilitates the elimination of the pathogens by quickly flushing them away. Once secreted, MUC2 unfolds and builds a polymeric and viscous network due to numerously hydrated
O-glycosylations connected to repeating domains rich in the amino acids proline, threonine, and serine (PTS-domains) [
4]. The MUC2 glycosylation is a post-translational modification that takes place as a stepwise process in the Golgi apparatus [
5]. Firstly, the Tn-antigen is created by
N-acetylgalactosamine binding to serine or threonine of the protein backbone, then the galactosaminyltransferase C1GALT1 and
N-acetylglucosaminyltransferases B3GNT6 and GCNT1-4 catalyse the elongation of the
O-linked oligosaccharide chains [
5,
6]. Various
O-glycans can be bound by specific glycosyltransferases, forming long carbohydrate chains and different core structures with different functions. Mucin-type
O-glycan chains can be expanded or branched by sialic acids on terminal, non-reducing ends expressed both on cell-surfaces and secreted glycoproteins. The transfer of sialic acid residues from a sugar nucleotide donor, mainly
N-acetylneuraminic acid, to an acceptor substrate is catalysed by sialyltransferases such as ST3 beta-galactoside alpha-2,3-sialyltransferase 1 (ST3GAL1) and 2 (ST3GAL2) [
7]. The negatively charged sialic acid improves water binding of MUC2 and prevents degradation of the protein backbone [
8,
9].
Apart from host
O-glycans, pathogenic bacteria also have characteristic oligosaccharides on their surface that serve host-pathogen interactions and are due to co-evolution with their hosts [
2,
10]. Here,
O-glycans on mucins can act as ligands for bacterial adhesins and specific connections between bacterial and MUC2 glycans can enhance adherence and invasion of intestinal pathogens. Some gastrointestinal pathogens, such as
Campylobacter jejuni, are able to bind host mucins mediated by glycan-glycan interactions and thereby penetrate the otherwise sterile inner mucus layer invading and infecting the underlaying epithelium [
2]. In humans,
C. jejuni is the leading cause of foodborne bacterial gastroenteritis (campylobacteriosis) worldwide (
http://www.who.int). In pigs and poultry though, the zoonotic bacteria can colonise the intestinal tract asymptomatically without causing disease [
11]. Even though
C. jejuni infections in humans are highly prevalent, only little is known about the pathomechanisms, and further research is needed to understand the molecular and regulatory backgrounds with special emphasis on host glycans and their modifications.
Previous work has already assessed that microRNAs (miRNAs) play an important role in the mammalian immune response after bacterial or viral infections [
10,
12‐
14]. miRNAs are short, non-coding RNAs that are involved in regulating gene expression and silencing [
15]. After binding their target sites, they initiate degradation or translational repression of the mRNA [
5]. While one miRNA can mediate the activity of different genes, a single gene could also be regulated by several miRNAs, resulting in complex regulatory networks that control cellular processes such as differentiation, proliferation or immune response [
15]. For instance, it has been shown that
Mycobacterium tuberculosis induces the overexpression of miR-125b to block tumour necrosis factor (TNF) biosynthesis for a suppressed immune response [
16]. Also, decreased expression of miR-125a-5p was associated with the presence of
Helicobacter pylori in human gastritis [
17]. Our recent study in a mouse model of human campylobacteriosis based on secondary abiotic IL10
−/− mice [
18] has shown that the mammalian conserved miRNAs miR-615-3p and miR-125a-5p are dysregulated upon
C. jejuni infection [
10]. The study revealed that the identified miRNAs are mutually involved in regulation of several glycosyl- and sialyltransferases participating in mucin-type
O-glycosylation in the murine colon. In the mentioned study, we mainly focused on the interaction between miR-615-3p and
St3gal2 and were able to demonstrate their molecular interaction in an infection-dependent context [
19]. However, advanced knowledge about the role of these two miRNAs in gastrointestinal bacterial infections is still scarce.
Here, we investigated the
C. jejuni infection dependent interaction between miR-125a-5p and the sialyltransferase ST3GAL1
, a conserved target in both human and mouse. We identified target sites of miR-125a-5p in the 3’ untranslated region (UTR) of
St3gal1 and validated them in vitro. By applying the above-mentioned in vivo mouse model of human campylobacteriosis [
18], we demonstrated that
St3gal1 and
B4galt1 possess anti-correlated gene expression compared to miR-125a-5p during
C. jejuni infection. This was supported by increased ST3GAL1 protein levels in colon samples of infected mice.
Discussion
Some pathogenic bacteria such as the zoonotic C. jejuni can infiltrate and penetrate the protective mucus layer in the colon and infect the mucosa. Even though C. jejuni is such a relevant human pathogen, the mechanisms underlying the infection have not been entirely unravelled yet and questions regarding colonisation and infection as well as the role of the host intestinal immune response remain unanswered to date. One major aspect within these processes forms the mucus. However, the role of the highly glycosylated MUC2, the main component of both colonic mucus layers, is not fully understood. In this study, we aimed to gain a more detailed insight into the regulatory mechanisms of mucin-type O-glycosylation in C. jejuni infections using a mouse model of human campylobacteriosis.
The importance of well-functioning MUC2 layers as a first line of defence against pathogens in the colon was revealed in previous studies [
22‐
24]. As a response to
C. jejuni infection, we have shown here that enzymes involved in the mucin-type
O-glycosylation exhibit altered intestinal expression, which may lead to differential MUC2 glycosylation and consequently affect the protective properties of the glycoprotein. This was also reported upon infections with other pathogenic species. For example,
H. pylori can modulate
O-glycosylation, leading to changes in mucosal glycans and enhanced mucin sialylation due to dysregulations of sialyltransferases [
25‐
28]. In the first line of
H. pylori infection, the adhesion of the pathogen is mediated by antigen binding adhesin (BabA) binding at H-type 1 and Lewis b structures present on mucin (28]. The increase of sialylated mucin structures during the infection process promotes a closer membrane attachment of
H. pylori mediated by the interaction of the sialic acid binding adhesin (SabA) with sialylated mucin structures, facilitating the infection [
28]. This might also be the mechanism how
C. jejuni penetrates the inner mucus layer and enables infection of the epithelium.
In the colon, ST3GAL1 primarily adds sialic acid to the core 1
O-glycans, a dominant core structure in mucin glycosylation, while ST3GAL2 transfers sialic acid to the terminal galactose residues found in glycoconjugates. B4GALT1, in contrast, catalyses the transfer of galactose to non-reducing ends of the sugar chain [
29‐
31].
C. jejuni dependent dysregulation of miR-125a-5p and the previously reported miR-615-3p as well as the targeted glycosyltransferases in the applied mouse model suggest that a miRNA-dependent modification of mucin-type
O-glycosylation takes place in the colon upon infection. It is still to be determined whether this is caused by
C. jejuni directly or triggered by host immune responses. On the one hand, it was shown that modification of mucins and the increased mucus secretion is controlled by the epithelial cells responding to signals from the innate and adaptive immune system [
26]. Pathogens have developed numerous mechanisms to penetrate the sterile inner mucus layer, thus, an altered mucin structure could serve as a host response preventing infection. This was seen in
H. pylori infections and could also be assumed in infections with
C. jejuni [
32]. However, the altered
O-glycosylation could also be pathogen-derived and may result in diminished mucin barrier function facilitating the infection. These alterations could affect colonisation and infection by the pathogen.
In humans,
Campylobacter causes symptoms such as severe diarrhoea and abdominal cramping [
11]. During infection,
Campylobacter trigger a range of immune responses and studies have shown that miRNAs are involved in regulating immune responses in several ways. The mechanisms known so far range from regulation of inflammation-dependent genes to control of apoptosis of infected host cells [
10,
13,
33]. Understanding miRNA function can provide new insights into the pathogenesis of microorganisms. Here, the role of miR-125a-5p in infections with bacterial pathogens is of particular interest. In
Mycobacterium avium infections, miR-125a-5p was overexpressed in human macrophages as part of the innate immune response to the bacterial infection [
34]. On the other hand, upregulated miR-125a in
Mycobacterium tuberculosis infections was associated with enhanced bacterial survival in human and murine macrophages [
35]. Furthermore,
H. pylori infections in humans were accompanied by decreased miR-125a-5p expression [
17] and downregulation of miR-125a in mesenteric lymph nodes of piglets after
Salmonella Typhimurium infection was identified, triggering diverse immune responses [
36]. Finally,
C. jejuni infections in mice have been recently found to be associated with reduced miR-125a-5p and enhanced miR-613-3p expression in the colon [
10]. Interestingly, all studies that have focussed on infections with intestinal pathogens revealed decreased expression of miR-125a-5p. In the present study, we focused on the interaction of miRNA-125a-5p with putative targets conserved in human and mice to emphasise its relevance in
C. jejuni infections. In silico analysis determined two overlapping targets of both dysregulated miRNAs involved in mucin-type
O-glycosylation being conserved in human and mice,
St3gal1 and
B4galt1. Here, miR-125a-5p was selected because of its known involvement in several bacterial infections, including
C. jejuni infections [
10,
17].
Decreased gene expression of St3gal1 and B4galt1 was observed after RNAi with miR-125a-5p mimics in our in vitro transfection experiments. Similar counter-regulated expression profiles of miR-125a-5p with St3gal1 and B4galt1 mRNAs were also observed in colonic tissue of the infected mice. Therefore, it is likely that expressions of both genes are negatively regulated by miR-125a-5p. In contrast, expression of St3gal2 was not significantly regulated after RNAi with miR-125a-5p mimics and therefore does not seem to be a target of miR-125a-5p. The fact that miR-615-3p tends to be upregulated after miR-125a-5p transfection suggests a synergistic mode of action of both miRNAs. However, the underlying regulatory mechanisms still need to be investigated. Overall, miRNAs can play an important role in regulating the expression of glycosyltransferases during infection, and understanding these interactions can not only provide new insights into the pathogenesis of infectious diseases but also help to develop novel therapeutics.
Apart from matching mRNA and protein levels of ST3GAL1 after intestinal infection of mice, we determined the localisation of ST3GAL1 in the infected tissue and furthermore, at the cellular level. Mucin-type
O-glycosylation takes place in the Golgi apparatus, and sialylated glycoproteins are stored in granules before being exported to the cell membrane or secreted [
1,
37]. IF staining showed cytoplasmic ST3GAL1 signal, suggesting concordant localisation of ST3GAL1 in the Golgi apparatus. As adhesins of various pathogens can bind to sialic acids on mucins, abundance of ST3GAL1 could lead to increased integration of sialic acids in MUC2 glycans enhancing pathogen binding [
26,
38]. The observation that the glycosyltransferases are dysregulated after interacting with the miRNAs during infection with
C. jejuni supports our hypothesis that a miRNA-dependent modification of the
O-glycosylation of MUC2 or other signalling proteins might take place here. Further research will focus on altered glycosylation pattern of mucin after
C. jejuni infection.
Conventional laboratory mice with a complex commensal gut microbiota are not susceptible to
C. jejuni infection, whereas secondary abiotic IL10
−/− mice generated upon antibiotic pre-treatment can not only be infected by the enteropathogens, but also display key features of human campylobacteriosis including acute enterocolitis within 6 days post-infection [
18]. Although the model simulates the course of the disease in humans well, there are limitations, as in any translational model. Thus, the results cannot be completely transferred to human conditions and differences between mice and humans need to be considered. There are clear species differences in the extent of intestinal colonisation by
C. jejuni. In contrast to humans and secondary abiotic mice, the enteropathogens colonise the intestinal tract of poultry and pigs asymptomatically without causing clear clinical signs of infection. Here, comparative studies of the interactions investigated may reveal possible differences between the animal species and humans to understand why the aforementioned animals are less susceptible to
C. jejuni. This would even have applied effects on food safety, because a better understanding of the factors that contribute to colonisation can be used to reduce the load of zoonotic agents in the food chain.
The two miRNAs studied here and previously, miR-125a-5p and miR-615-3p, could be involved in the colonisation and infection process of the gut by C. jejuni. Through their targets, they could be part of the mucosal and cellular host responses upon bacterial infection. Modified mucin-type O-glycosylation could be a direct mucosal response to hinder pathogenic colonisation and infection. On the contrary, C. jejuni could actively induce dysregulation of miRNAs facilitating host cell adhesion and invasion.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.