Gut Microbiota Composition Associated with Clostridioides difficile Colonization and Infection

Clostridioides difficile is an anaerobic Gram-positive and spore-forming bacterium. The majority of C. difficile strains produce two toxins, A and B, associated with the development of acute diarrhea and/or colitis. In this review, two situations are distinguished: C. difficile infection (CDI) and asymptomatic colonization (AC). The main objective of this review is to explore the available data related to the link between the gut microbiota and the development of CDI. The secondary aim is to provide more information on why some people colonized with toxigenic C. difficile develop an infection while others show no signs of disease. Several factors, such as the use of antibiotics and proton pump inhibitors, hospitalization, and age, predispose individuals to C. difficile colonization and/or C. difficile infection. The gut microbiota of people with AC showed decreased abundances of Prevotella, Alistipes, Bacteroides, Bifidobacterium, Dorea, Coprococcus, and Roseburia. The gut microbiota of people suffering from CDI showed reductions in the abundances of Lachnospiraceae, Ruminococcaceae, Blautia spp., Prevotella spp., Dialister spp., Bifidobacterium spp., Roseburia spp., Anaerostipes spp., Faecalibacterium spp. and Coprococcus spp., in comparison with healthy people. Furthermore, increases in the abundances of Enterococcaceae and Enterococcus were associated with C. difficile infection.


Introduction
Clostridioides difficile is an anaerobic, Gram-positive, and spore-forming bacterium recognized as the leading cause of health care-associated diarrhea. In the United States of America (USA), in 2017, the Centers for Disease Control and Prevention (CDC) considered C. difficile infection (CDI) to be a major health threat (with 223,900 national cases) among hospitalized patients, eventually leading to 12,800 deaths [1]. In France, in 2016, CDI incidence in acute care was estimated to be 3.6 cases per 10,000 patient days [2]. In 2016, a total of 7711 CDI cases were reported to the ECDC (European Centre for Disease Prevention and Control) in Europe (20 EU countries), of which 74.7% were associated with health care settings [3].
In this review, two situations are distinguished in order to understand the trigger symptoms: CDI and AC of C. difficile. CDI is defined by the presence of diarrhea and at least one other following criterion: the carriage of a toxigenic strain of C. difficile, the presence of toxins in the stool and/or a colonoscopy result showing pseudomembranous colitis. AC by C. difficile can be defined as the absence of diarrhea with at least one other criterion: carrying a strain of C. difficile and/or the presence of toxins in the stool [4]. Crobach et al., defined AC as the presence of C. difficile but without symptoms of CDI.
The main risks factors are PPI use, antibiotics use, corticoid use, hospital stay and age.
The main risks factors are PPI use, antibiotics use, corticoid use, hospital stay and age.
PPI use is a risk factor for rCDI (recurrent CDI) [5]. Their use increased stomach pH from 1.2 to 5, raising the possibility of developing CDI or even carrying the bacterium asymptomatically compared with subjects without this treatment [6]. The increase in stomach pH to a value of 5 during digestion does not influence the resistance of the spores of C. difficile, as they are able to survive at a normal stomach pH [6][7][8]. The vegetative cells of C. difficile can survive in the gastric content only if the pH is equal to or greater than 5 [7].
The use of antibiotics is a well-known risk factor for C. difficile asymptomatic colonization or infection due to modification of the gut microbiota. Most of these cases are associated with the use of four antibiotics: clindamycin, cephalosporins, carbapenem and fluoroquinolone [9][10][11]. Other antibiotics, including macrolides, sulfonamide, trimethoprim and penicillin, are less associated with CDI [9][10][11]. Antibiotic exposure increases the possibility of C. difficile colonization by 3.7-fold [12] and of developing CDI by 3.55-fold [9]. A previous study showed that among antibiotic-associated diarrhea (AAD) cases, the incidence of CDI was 1.14-1.89% [13], and the frequency of toxigenic C. difficile carriage was 18.1-19.0% [13].
The use of corticosteroids increases the risk of C. difficile colonization in adults admitted to the hospital [14] and immunosuppressive therapy is a risk factor for complicated CDI [15]. Immunocompromised patients have a higher risk of developing rCDI during hospitalization [16].
A recent hospitalization or a recent intensive care unit stay increases the risk of developing CDI by 2.2 and 6.5, respectively [11]. Hospitalization in the previous 6 months increases the risk of colonization by 2.18 [14]. Previous studies showed that in an ambulatory group (n = 43), in patients with short hospital stays (n = 48) and in patients with long hospital stays (n = 102), the percentages of C. difficile carriage were 9.5%, 8% and 13%, respectively [17]. CDC hospitalization rates are significantly higher among those 65 years of age and older (by 4-fold) and those over 85 years of age (by 10-fold) compared to those under 65 years [18,19].
The percentage of AC evolves as a function of age. It is high in the first months of life and decreases until adult age, and then it increases with advancing age. The percentage of individuals with AC over time is shown in Figure 1 [4,[20][21][22][23][24][25][26][27]. Patients aged > 65 years old have a 10-fold higher risk of developing CDI than patients in the other age groups [28,29]. Box plot illustrating the mean relative proportion of C. difficile asymptomatic carriers in function of category of age. These data were found in articles that studied the prevalence of AC [4,[20][21][22][23][24][25][26][27].
There are three lines of defense against pathogens: the epithelial barrier, innate immunity, and adaptive immunity. The first step is intestinal colonization by C. difficile. Once the bacterium can produce toxins, these toxins transgress the epithelial barrier through the activation of Rho glycosylation, which causes disruption of tight junctions. Secondly, pathogen-associated molecular patterns (PAMPs) are recognized by pattern recognition receptors (PRRs). This interaction induces a rapid innate immunity response. Finally, adaptive immunity provides a highly specific immune response against C. difficile [30].
The intestinal mucus is composed of two types of mucins: MUC1 (cell-surface) and MUC2 (secreted forms). Secreted MUC2 is found mostly in the feces of healthy people, while people who suffer from CDI have an imbalanced mucus composition; their stool mucus is composed mainly of MUC1, with significantly decreased MUC2 levels [31]. They also present an increase in terminal galactose residues (a known receptor for C. difficile toxin A in mice, hamster, rabbits and pigs) and N-acetyl glucosamine (GlcNAc) [31] and a decrease in N-acetyl galactosamine (GalNAc).
Essential to C. difficile spore germination is the presence of primary bile acids (PBAs). PBAs is produced by the liver, is discharged into the small intestine and helps with fatty digestion [20]. PBAs stimulates germination of C. difficile spores [32], and secondary bile acids (SBAs) inhibit germination [32].
Several studies have tried to understand the C. difficile pathogenesis in order of reduce the risks of development of the disease and find new therapeutic strategy. The animal experimentations using hamster have allowed to test the transplantation fecal efficiency, the use of non-toxigenic strain of C. difficile and use of monoclonal antibodies against toxin A and toxin B [33][34][35]. The piglet model of CDI is representative of the key characteristics of human CDI and helped to understand the virulence and new treatment [36]. Then, C. difficile studies use different methods in vitro in order to limit using animal experimentation: feces cultures [37], the continuous culture model [38,39], the triple-stage chemostat human gut model [40,41], and the Tim-2 model [42].
The main objective of this review is to explore the available data about the link between the gut microbiota and the development of CDI. The secondary aim is to provide more information on why some people colonized with toxigenic C. difficile develop CDI and others show no signs of disease.

Composition of the Normal Human Gut Microbiota
The composition of the gut microbiota is influenced by diet, age, the use of antibiotics, etc. [43]. In a normal gut, the dominant phyla are Firmicutes, Bacteroidetes, Actinobacteria, Proteobacteria, Fusobacteria, and Verrucomicrobia. Firmicutes and Bacteroidetes represent 90% of the gut microbiota [44,45]. Gut Bacteroidetes is mainly composed of two genera, Bacteroides and Prevotella [44,45]. The main genera in the Firmicutes phylum are Lactobacillus, Bacillus, Clostridium, Enterococcus, Faecalibacterium, Roseburia, and Ruminococcus. The Actinobacteria phylum is represented mainly by the Bifidobacterium genus [44,45]. The Metahit consortium classified human fecal metagenomic samples from three continents into three groups called "enterotypes": Bacteroides (enterotype I), Prevotella (enterotype II) and Ruminococcus (enterotype III) [45,46]. The enterotype concept is very controversial in the scientific community [46]. Gorvitovskaia et al., chose to validate the first two enterotypes (Prevotella and Bacteroides) [47], while Cheng et al., have shown that enterotypes are not constant over time and are influenced by age and diet [43].
The gut microbiota of infants is rich in Bifidobacterium spp. and the gut microbiota of elderly individuals has decreased proportions of Ruminococcaceae, Bifidobacterium, Lactobacillus and Faecalibacterium and increased proportions of Proteobacteria, Bacteroidetes and Clostridium spp. [44,48,49].

Factors Influencing the Healthy Gut Microbiota
The use of proton pump inhibitors (PPIs) influences the pH of the stomach and therefore the gut microbiota. The use of PPIs decreased gut microbial diversity and the abundances of Clostridiales and Ruminococcaceae [50,51] and increased the abundances of the Enterococcaceae and Staphylococcaceae families [50,52] and Veillonella parvula and Streptococcus mutans [53]. These modifications of the gut microbiota are strongly correlated with CDI development.
In human feces, clindamycin, cephalosporins and fluoroquinolone treatments impact the microbiota, resulting in an increase in Enterococcus abundance [54]. These three categories of antibiotics induce a reduction in the abundances of Streptococcus spp., Anaerococcus spp., Peptoniphilus spp., Porphyromonas spp. and Prevotella spp., and increase the abundance of Sphingomonas spp. [54]. Antibiotics also seem to decrease the proportions of Ruminococcaceae, Lachnospiraceae and Bifidobacterium and to increase the proportions of Lactobacilliaceae and Streptococcaceae [49]. Werkhoven and collaborators (2021) showed that the carriage of toxigenic C. difficile increased the incidence of developing CDI 10-fold after antibiotic treatment. Specifically, the use of carbapenem increased the incidence of CDI 5-fold and increased the abundance of Enterococcus 5-fold [13].
The age-modified gut microbiota. The gut microbiota of elderly individuals has decreased proportions of Ruminococcaceae, Bifidobacterium, Lactobacillus and Faecalibacterium and increased proportions of Proteobacteria, Bacteroidetes and Clostridium spp. [44,48,49]. Two interesting recent studies have addressed the gut microbiota composition in hospitalized elderly patients with CDI, showing lower microbial diversity, lower proportions of gut commensals with putative functions and a reduction in butyrate-producing bacteria in CDI samples [55,56].
The state of dysbiosis can be defined as a decrease in the obligate anaerobic bacteria and an increase in the relative abundance of facultative anaerobic bacteria, such as Enterobacteriaceae [57,58], a decrease in microbial diversity and a decrease in anti-inflammatory species such Faecalibacterium prausnitzii [59]. In Inflammatory Bowel Disease (IBD), a decrease in microbial diversity, a decrease in F. prausnitzii and an increase in Streptococcus and Escherichia/Shigella are observed [59,60].

Composition of Microbiota among Patients with AC
As previously described, AC by C. difficile can be defined as the absence of diarrhea with at least one other criterion: carrying a strain of C. difficile and/or the presence of toxins in the stool [4]. In the literature, few studies differentiate between AC by C. difficile and CDI. The composition of the microbiota of patients with AC was similar to that of the control group and included the phyla Bacteroidetes (40.95%), Firmicutes (36.23%), and Proteobacteria (15.73%) [61].
Within the phylum Bacteroidetes, decreases in two families (Bacteroidaceae and Prevotellaceae) were observed [17]. Zhang et al., showed decreases in the AC of Prevotella spp., Alistipes spp., Bacteroides spp. and an increase in the AC of Parabacteroides spp. [61].
Within the phylum Proteobacteria, several studies showed an increase in AC by Enterococcus spp. and Klebsiella spp. [61]. In the phylum Actinobacteria, Zhang et al., showed a decrease in AC by Bifidobacterium spp. [61]. Within the phylum Verrucomicrobia, a decrease in AC by Akkermansia spp. was observed [63].
Some studies showed no differences in microbial diversity between an AC group and a healthy group (HG) who presented a negative stool test for C. difficile [63]; however, some studies showed that the microbial richness (Chao index) decreased in patients with AC compared with patients in the HG [61].

Microbiota Composition of Adults Suffering from CDI
Many studies have described the gut microbiota composition of patients with CDI. In Table 1, the modifications of gut microbiota with increases (red) and decreases (green) in the abundances of various phyla and genera when a patient is suffering from CDI are represented. These data are from the original research publications.
The abundance of the phylum Actinobacteria seems to decrease with adult CDI [61,62,64]. The main genus responsible for the decrease is Bifidobacterium spp.
Within the phylum Verrucomicrobia, some studies showed that an increase in the abundance of A. muciniphila is associated with CDI [62,68,71,77], while others reported that an increase in the abundance of Akkermansia protected against CDI [56,63,69].
Hernandez et al., (2019) classified CDI into two groups according to prognosis. Cluster A showed high abundances of Enterococcaceae and Enterococcus and decreases in the abundances of Bacteroidaceae and Lachnospiraceae. This cluster was associated with more severe diarrhea, more aggressive therapy and a poor prognosis. Cluster B had high abundances of Bacteroidaceae and Lachnospiraceae. This cluster was associated with less severe diarrhea, less aggressive therapies and a good prognosis [67].
Several recent studies have focused on comparing the gut microbiota of people with CDI with people with negative C. difficile detection and symptomatology. A significantly decrease in the bacterial population diversity (Shannon index) [54,[64][65][66]70,74] and a significantly lower richness (Chao1 index) [64,66] has often been observed in patients with CDI.
People who developed a single CDI had higher levels of IgM anti-toxin A, toxin B and non-toxigenic antigens on Day 3 and significantly higher IgG anti-toxin A on Day 12 than people who developed recurrent CDI forms [30].
In Table 2, more details about these studies are provided.    CDI group (n = 31; 64.0 y). Three periods: pretreatment ATB, two days after treatment, seven days after treatment or discharge.
No control group The alpha diversity (Shannon index) was lower with CDI pretreatment in the recurrent group. * Veillonella dispar * predictor of recurrence. /

Fecal Microbiota Transplantation (FMT)
FMT is the transfer of the fecal microbiota containing bacteria from a healthy volunteer into a diseased patient [82,83]. In 1958, Dr Ben Eiseman described the "fecal enema" as a treatment for pseudomembranous enterocolitis [83]. FMT is administered in several ways: with capsules for ingestion, with an endoscope and with a nasoenteric tube [84]. The criteria for selecting a healthy donor have been described in several papers [84,85]. Briefly, the blood and the stool must be tested to ensure that no infectious diseases or pathogenic bacteria are present, and a series of criteria are checked, including recent antibiotic use, history of diarrhea, and history of immune disorders [84,85].
In the case of primary CDI, this treatment could be used before using antibiotics or in addition of antibiotics to avoid rCDI [86]. In the case of rCDI, FMT is the second line of treatment. FMT has been reported to be successful in 80-92% of patients with rCDI [84,[87][88][89] and with primary CDI [89]. This treatment is safe and effective in adults [87,88], in elderly [90][91][92] and in children [93].
Regarding the composition of the gut microbiota of patients with FMT, the alpha diversity (Shannon index) seems to be lowest pretreatment among patients with CDI, and the alpha diversity is restored after FMT [87,88,90,91]. The composition of the microbiota has an impact on the recurrence of CDI and the success of FMT. After FMT, the phylum Firmicutes increased significantly in rCDI (<65 y) and the phylum Proteobacteria decreased significantly in rCDI (>65 y) [90]. Lachnospiraceae, Ruminococcaceae and Bifidobacteriaceae increased significantly [90] and Enterobacteriaceae decreased significantly after FMT in rCDI [89,90]. Blautia, Ruminococcus, Coprococcus, Bifidobacterium [90] and Faecalibacterium [89] spp. increased significantly after FMT in rCDI [89,90]. These modifications of the gut microbiota after FMT strongly suggested that these bacterial populations are associated with healthy people (See Table 1) and will favor a good prognosis. Staley et al. (2018) showed that a follow-up analysis of 16S rDNA extracted from feces can be used to predict an eventual recurrence of CDI. After FMT, high proportions of Lactobacillales, Enterobacteriaceae, Enterococcus, Klebsiella, Streptococcus and Veillonella and reductions in Roseburia, Blautia, Lachnospiraceae, Ruminococcaceae, Anaerostipes, Coprococcus, Dorea and Clostridiales [88,91] will disadvantage the patient and promote rCDI. These bacterial population are associated with the gut microbiota of CDI cases (see in Table 1).
Before using the FMT, a preventive probiotic administration before the antibiotics use is effective [94]. The use of probiotics before and at the same time as antibiotics reduces the risk of CDI by >50% in hospitalized adults [94].

Conclusions
Many studies have characterized the gut microbiota composition of patients with CDI, but confusion is still present in the literature between CDI and AC. Few studies have differentiated AC by C. difficile and CDI. This review explores the available data related to the link between the gut microbiota and the development of C. difficile infection. The causes of the development of CDI are clearly multifactorial. An external cause (such as a medication) can disrupt the homoeostasis of the gut microbiota. PPI and antibiotic use decrease the richness of the gut microbiota [95]. This imbalance promotes the growth of some bacteria (for example, A. muciniphila), and these bacteria can degrade the mucus layer and allow the pathogenic bacteria and toxins access to the epithelium. Moreover, the abundances of some bacteria (Lachnospiraceae, Ruminococcaceae and Blautia) decrease, and these bacteria are involved in bile metabolism and can increase the primary bile acid concentration. Higher PBAs concentrations are favorable to C. difficile germination and multiplication. Some bacteria also have a positive correlation (Enterococcus, Enterobacteriaceae) or negative correlation (Blautia, Prevotella, Roseburia, Dorea, Collinsella, Coprococcus, Ruminococcus, Ruminococcaceae, Lachnospiraceae) with C. difficile colonization and/or CDI. The gut microbiota will promote the development of the CDI. Through all the studies, the CDI has a gut microbiota footprint with the decrease and the increase in some bacteria. In this review, a lot of bacteria are singled out for giving an advantage or a disadvantage when developing CDI. Some of these bacteria have an impact on gut health. Faecalibacterium prausnitzii is considered as a species of the healthy gut microbiota. This bacterium is reduced in gut dysbiosis, in IBD, in obesity, in diabetes, etc. [96]. Lachnospiraceae is protective against CDI [20]. C. scindens, member of Lachnospiraceae have a protective effect against CDI [20,32]. Amrane et al. (2018) showed that C. scindens is present in the feces when patient developing CDI [97]. Alistipes spp. indicated a controversial pathogenicity. On the one hand, the bacteria have protective effects against liver fibrosis, cancer immunotherapy and cardiovascular disease [98]. On the other hand, this genus is associated with colorectal cancer and mental disease [98].
In this review, CDI can be associated with an increase or a decrease in A. municiphila and AC is associated with a decrease A. municiphila. The presence of this bacteria is positive against obesity, diabetes, cardiometabolic disease and low-grade inflammation [99]. It is actually used to manage obesity [100]. In human intestinal organoids, C. difficile is capable of decreasing MUC2 production, but it is not responsible for altering host mucus oligosaccharide composition [31]. Furthermore, it has been reported that C. difficile is not capable of degrading mucin glycans, although coculture with mucin-degrading Akkermansia muciniphila, Bacteroides thetaiotaomicron and Ruminococcus torques allowed the pathogen to grow in media that lacked glucose but contained purified MUC2 [101]. When mucus is degraded by bacteria, oligosaccharides (GlcNAc, GalNAc, galactose, mannose and fucose) are salted out [101], and C. difficile is capable of using these oligosaccharides [31].
The Enterobacteriaceae family is associated with the dysbiosis state [58,59]. Enterococcus spp. is a controversial bacterium [102]. It is a commensal bacterium of intestinal flora, vagina, and mouth microbiota [102]. E. faecium and E. faecalis are used as probiotics [102,103] and Enterococcus spp. is used in meat and cheese [102,104] fermentation. Recently, it was shown that E. faecalis and E. faecium are potentially pathogenic bacteria due to their ability to adapt in new environment [102,105]. Additionally, a resistance to vancomycin has emerged in this genus [102,105]. Romyasamit et al. (2020) exhibited that six E. faecalis strains have a probiotic effect and anti-C. difficile activity [106]. Klebsiella pneumonae is present in the mucus layer with C. difficile [107].
The second objective of this review was to provide more information on why some people colonized with toxigenic C. difficile develop C. difficile infection and others show no signs of disease. The answer to this question is still unknown, but some facts will improve the understanding. The response of the adaptative immune system impacts the development of the disease. Patients exhibiting AC were shown to have higher antibody levels against C. difficile than people who developed CDI. It has been reported that sixty percent of the general population has had AC or has been infected with C. difficile, as determined by the observation of detectable seric IgG and IgA antibodies to toxins A and B [108]. IgG and IgA titers against toxins A and B are significantly higher in children positive for toxigenic strains than in people carrying non-toxigenic strains [109]. IgG antibody levels against toxin A are significantly higher within 3 days of colonization in AC patients than in those who develop CDI [30,108,110]. IgG levels against toxin B and non-toxin antigens seem to be higher among individuals who develop AC [30,108,110].
Some risks factors will predispose patients to developing CDI (antibiotics use, PPI use, age, etc.). The decrease in the diversity described in elderly gut microbiota [18], the decrease in some bacterial population (Ruminococcaceae, Bifidobacterium, Faecalibacterium) and the increase in some bacterial population (Proteobacteria, Bacteroidetes and Clostridium spp.) suggest why this population have a 10-fold higher risk of developing CDI.
Some treatments involving bacteria are commonly used and effective against CDI. FMT allows the recovery of patients with recurrent CDI with an increase in the abundances of some bacteria (Blautia, Collinsella, Anaerostipes, Coprococcus, Dorea and Roseburia) and a decrease in the abundances of others (Lactobacillales, Enterobacteriaceae; Enterococcus, Klebsiella, Streptococcus and Veillonella). More research with strict inclusion criteria is needed to measure AC and CDI gut microbiota analysis. The purpose of this work was to study the impact of in vivo control measures for the gut microbiota to decrease colonization and CDI or to improve recovery.