Understanding the factors regulating host–microbiome interactions using Caenorhabditis elegans

How microbial associations and microbe-derived metabolites modify animal behaviours is an emerging area of study in host–microbiome interactions [32]. Can the gut microbiota influence what we eat and manipulate how we behave? The answer is, yes! Gut microbiota-derived metabolites and neurohormones can manipulate eating behaviours by impacting satiety [33], and can even modify the sense of taste and food cravings through different mechanisms [32]. In addition, human gut microbiota-derived metabolites such as short-chain fatty-acids (SCFAs) can directly induce the secretion of gut-derived hormones that indirectly modulate brain functions via the vagal nerve in the gut-brain axis, resulting in an impact on behaviour [32,34]. Interestingly, a recent study demonstrated the therapeutic potential of butyrate (an abundant SCFA) and human butyrogenic bacteria in ameliorating the toxic aggregation of metastable proteins induced by human enteric pathogens, using the C. elegans model [35].

In the wild, C. elegans is exposed to a variety of microbes, from pathogens to commensals. Given that C. elegans are bacterivores, many microbes also serve as food sources. It is likely that C. elegans uses an innate sense of smell to quickly profile its habitat and make immediate decisions, including foraging, avoidance, eating or egg-laying [36]. In addition to these innate behaviours, worms are also capable of long-term learned behaviours elicited by changes in their internal state [37,38]. As a result, while ingestion of nutritive bacteria leads to learned attraction for exploitation of a bacterial food source, ingestion of pathogens and toxins leads to learned aversion for avoiding stress and damage ([39–41]; figure 1a).

Microbial-derived odours are expected to impact seeking and ingestion of a food source by C. elegans. Interestingly, several bacteria from the natural habitat of C. elegans were found to release odours that include previously known attractants and repellents, studied for decades in C. elegans neurobiology and behaviour [42–44]. For example, Lactobacillus paracasei were isolated from rotten citrus fruits that also harboured wild C. elegans. This species of bacteria was found to naturally produce the chemical diacetyl, a known attractant of C. elegans [45]. This attractant is a volatile byproduct of citric acid metabolism by L. paracasei, and while this bacterium does not appear to colonize C. elegans, it possibly attracts the nematodes to rotting citrus fruits that are rich in other microbes [16].

In yet another interesting mechanism, a known C. elegans commensal bacterium, Providencia alcalifaciens (strain JUb39) [16], uses sensory attraction and sensory override mechanisms to manipulate C. elegans feeding behaviours. Providencia alcalifaciens releases isoamyl alcohol, another well-studied C. elegans attractant [44], that possibly makes this bacterial species a preferred food choice over the standard E. coli diet [46]. Then, on colonizing the C. elegans gut, Pr. alcalifaciens uses host-derived tyrosine to produce the neurotransmitter tyramine that is then used by C. elegans to synthesize octopamine, thus subverting the need for the host to produce tyramine. Thus, bacterial tyramine uses C. elegans octopamine signalling to essentially manipulate C. elegans into selecting a food source that is beneficial for both of them [46,47].

In an example of beneficial selection by the host, C. elegans were colonized on being attracted to a beneficial Pantoea strain of bacteria, that hindered the pathogenic colonization by Pseudomonas aeruginosa [29]. These observations imply a possible habitat-driven coevolution between this beneficial Pantoea sp. and C. elegans, as these are bacteria found in the same natural habitat with nematodes, and host colonization by beneficial commensals probably block infection by pathogens.

How does the microbiome shape C. elegans behaviour during food scarcity? Food deprivation is known to elicit behavioural changes from enhanced risk-taking during food search [48], to alternate developmental decisions, such as dauer formation. During unfavourable conditions including food scarcity, instead of further development into reproductive adults, C. elegans larva develop into a metabolically quiescent state called the dauer that is capable of surviving for prolonged periods [49]. Consistently, dauers are very prevalent in the natural habitat of C. elegans, and are probably the first to enter and the last to exit a new food source in a given habitat [18,21]. Dauers exhibit a phoretic behaviour called nictation, that enables their dispersal to geographically isolated habitats by hitchhiking on snails and isopods [18,21]. Interestingly, trimethylamine, a human gut flora metabolite was shown to mediate food-sensing behaviour in C. elegans and regulate dauer formation [50]. However, it remains unknown if microbial odours or metabolites derived from the C. elegans microbiota impact their dauer entry and exit decisions, or if dauer-associated microbes define the microbial communities in a new habitat. This is a fascinating area for future research in C. elegans, as it is likely that many host–microbiome co-evolutionary processes have occurred with the dauer state. Given that the dauer stage probably acts as a transport mechanism for some bacteria, dauers could possibly help in the survival of resident microbes between discontinuous food sources [51] and across geological time scales through suspended metabolism (cryptobiosis) [52].

The microbiome is known to play a major role in the development and maintenance of host immunity. In parallel, host immunity has been identified as a common and critical factor in shaping the microbiome [53]. Especially, the composition of the first human microbiota is critical for the development and priming of the neonatal immune system. Vaginal births are known to result in better vertical transmission of microbiomes and greater gut microbiota diversity compared to cesarean delivery [54], which subsequently impact childhood pathologies and immune-mediated diseases [55]. Likewise, in the absence of gut microbiota, germ-free mice have morphological and functional defects in the development of adaptive immunity and regulation of immune cells [56].

The natural C. elegans microbiota are also known to contribute to host immunity, either directly or indirectly to counteract pathogens ([27,57]; figure 1b). Similarly, host immunity also impacts the microbial proliferation and composition [13]. Together, this can be conceptualized as the collective immunity of the host and its microbiome, which for C. elegans includes multiple innate immunity pathways combined with acquired immunity from the associated microbes that together dictate the carrying capacity and composition of its microbiome.

The microbiome can contribute to host immunity by directly counteracting pathogens. This occurs via inter-species or inter-kingdom interactions between microbes that shape the microbiome through cooperation or competition, sometimes through microbial derived products with anti-microbial effects [26]. For example, Pseudomonas lurida, strain MYb11, provides broad-range protection to C. elegans against the fungal pathogen Drechmeria coniospora [26], Gram-negative Ps. aeruginosa, and Gram-positive Bacillus thuringiensis bacteria [58]. Pseudomonas lurida was found to directly antagonize Bacillus thuringiensis growth by generating massetolide E, a cyclic lipopeptide surfactant of the viscosin group, which has a bacteriostatic effect on the pathogen. As such, worms colonized by Ps. lurida exhibited significant improvement in epithelial barrier function (leaky gut) induced by Bacillus thuringiensis ([58]; figure 1b). Pseudomonas lurida benefits its host in multiple ways, however gut colonization results in a small reduction in C. elegans lifespan, suggesting that there are some trade-offs to the host when colonized by these protective bacteria [59].

The microbiome can also contribute to host immunity by indirectly counteracting pathogens, through modulation of host immunity and stress pathways that regulate infection [27,30,60]. Pseudomonas fluorescens, strain MYb115, protects C. elegans without reducing the pathogen load of Bacillus thuringiensis. Therefore, it is hypothesized that Ps. fluorescens helps C. elegans cope with infection, possibly by enhancing damage repair in the gut epithelium [58]. This idea is supported by the fact that Ps. fluorescens decreases the epithelial barrier dysfunction induced by Bacillus thuringiensis infection [58], potentially by altering host lipid metabolism, and cytoskeleton reorganization through intermediate filaments [61], which are known to protect against microbial insults [62]. The impact of C. elegans microbiome on cytoskeleton reorganization and epithelial-barrier function requires further investigation.

The capacity of some microbiome bacteria to activate host immunity pathways, however, may be a result of their cryptic virulence, as they are capable of becoming pathogenic when host immunity is not activated. This was recently found to be the case for a large majority of the C. elegans CeMbio bacteria that induce the helix-loop-helix-30/transcription factor EB immune pathway, which were pathogenic when this pathway was knocked out, suggesting that these bacteria have characteristics of pathobionts (commensals with pathogenic potential) [63]. Consistent with this idea, the state of host immunity can also impact the composition of the microbiota. For example, the known C. elegans commensal bacteria, Enterobacter cloacae, strain CEent1, is known to protect from gram-positive Enterococcus faecalis infection [57]. However, Enterobacter cloacae itself becomes a pathogen in immune-compromised C. elegans strains with loss-of-function mutations in the transforming growth factor β/bone morphogenetic protein pathway, a conserved development and immunity pathway [13,64,65]. This rise in pathogenicity is evidenced by a selective Enterobacter spp. bloom that shortens host lifespan with enhanced susceptibility towards Enterococcus faecalis (figure 1b).

It is therefore evident from these studies in C. elegans, that both host immunity and microbe-derived immunity dictate microbiome composition, and a decline in host immunity plays a role in the shifting commensal-like microbes to pathogens (figure 1b). However, it is also possible that some microbes can actively switch from commensal-like to pathogenic traits through changes in their gene expression to produce toxins or other harmful metabolites [66], an aspect of the C. elegans microbiome that needs investigation.

Host–microbiota assemblies are expected to be governed by the availability and distribution of nutrients in a given habitat. For microbiome bacteria, it is expected that the bulk of nutrition comes from the host. However, microbes reciprocally allow their host to use resources that they otherwise cannot digest or metabolize on their own, including essential vitamins. Consistent with this idea, the lack of a metabolically active gut microbiome causes caloric restriction in germ-free mice [67], and axenically grown C. elegans [68], showing a host's dependence on their microbiome for optimal nutrition. Similarly, our diet can dramatically impact our microbiome, as a Mediterranean-style diet rich in fruits and fibres, and low in fat, can have a positive impact on the gut microbial diversity and metabolism [69–71]. By contrast, a Western-style diet low in fibre and high in fat, can result in reduced levels of beneficial bacteria with low diversity [72], resulting in obesity and altered behaviours [73].

Caenorhabditis elegans are bacterivores and thus have a more complicated relationship with their microbiota, which can also serve as a nutrition source. As such, it is not so straightforward to separate the non-dietary commensal- or pathogen-like impact of microbes from their potential dietary (or nutritive) roles. We have only begun to understand the impact of microbiome derived metabolites on C. elegans life-history traits. For example, of the several known essential nutrients commonly required by both humans and C. elegans [74], vitamin B12 is an example of a microbiota-derived essential cofactor whose deficiency leads to slow growth (figure 1c). Among the known members of the C. elegans microbiome, several Pseudomonas and Ochrobactrum species have genes in the biosynthetic pathway of vitamin B12 [75]. Without B12, C. elegans exhibit stunted development, loss of fertility, and reduced lifespan [76], as the cofactor is required in the methionine/S-adenosyl methionine cycle for converting homocysteine to methionine [77]. As another example, C. elegans are cholesterol auxotrophs as they lack the first three enzymes in the canonical cholesterol biosynthesis pathway. However, it was recently found that these animals are capable of converting dietary sterols obtained from fungi and plants into cholesterol [78], making it possible that microbiome bacteria or fungi can provide sterols for cholesterol production in C. elegans, an area that remains to be investigated (figure 1c). Besides providing for nutrients and essential vitamins, microbiota can also impact host metabolism. For an example, a known C. elegans commensal, Ps. lurida (MYb11) was found to increase vitellogenin protein (yolk protein) production in young C. elegans adults, believed to hasten their reproductive timing [61].

In addition, it is predicted that C. elegans microbiome bacteria use their other unique metabolic competencies to colonize and flourish in the gut. Along these lines, a recent study identified two key metabolic traits in known C. elegans commensals, namely, the ability to ferment pyruvate to acetoin which probably influences gut colonization, and the ability to degrade hydroxyproline which probably correlates with C. elegans population growth [75]. The fermentation of pyruvate to acetoin releases diacetyl, whose buttery odour is known to attract C. elegans and promote feeding behaviour [79], which would result in enhanced bacterial uptake and colonization. Similarly, hydroxyproline degradation is known to produce reactive oxygen species, which could induce C. elegans growth and reproduction, or alternatively bacteria degrading hydroxyproline can proliferate by using the amino acids as a carbon source [75].

Finally, an interesting aspect of host–microbiome associations is the ability of hosts to provide niche-specific food to microbiota species that are uniquely adapted to forage on them. In mammals, the mucus layer, in addition to its role as a protective physical barrier for the epithelia, is also known to provide nutrition to its microbes. Mucus is made of heavily glycosylated mucin proteins which can become a foraging avenue for glycolytic microbes that metabolize the complex glycosylated moieties to release sugars for their consumption [80]. The production, diversification and maintenance of a mucus layer is akin to a metabolic collaboration between the host and their microbiota. Consistent with this idea, germ-free mice have an underdeveloped mucus layer [81], and glycan metabolism is known to shape the human gut microbiota [80]. Several prominent human gut commensals include mucin-degrading bacteria, such as Akkermansia muciniphila, that can utilize mucin as their sole source of carbon and nitrogen (reviewed in [82]). Similarly, Bacteroides thetaiotaomicron exhibit adaptive mucus foraging to utilize glycan sugars in the absence of dietary polysaccharides [81], thereby contributing to the metabolic homeostasis in the gut. However, in C. elegans, we do not yet know whether its microbiota can induce glycan and mucin production, nor whether there are mucolytic and glycolytic foragers in its microbiota (figure 1c). Overall, for successful colonization, microbes likely integrate their metabolism with host metabolism to provide an optimal growth condition for both.

Anatomical and tissue-specific factors are expected to drive the community composition of a microbiome, and here we focus specifically on the animal gut. In mammals, the gut is compartmentalized along the length, into the stomach, small intestine (duodenum and ileum), and the large intestine, each having unique structure and function. The different microbial composition in these tissues is predominantly dictated by their niche-specific environments, such as pH, oxygen, nitrate and bile acid metabolites (reviewed in [83]). Microbial interactions in the gut mainly occur with the secreted mucus layer or the glycocalyx, a layer of heavily glycosylated proteins (including membrane-bound mucins) at the apical surface of the epithelia. These coats create a protective barrier between the underlying gut epithelium and microbes. Besides being the first-line of defence, the mucus layer can provide substrate for bacterial adhesion, with host mucins bound by bacterial fimbriae or pili [84]. Research suggests that there is a potential cross-communication between the host gut and the microbiota. In fact, altered mucous glycosylation can impact the microbiota composition and intestinal architecture [85]. In turn, microbes are known to induce host secretion of mucus. This idea is supported by the ability of the mucolytic human gut commensal, Bacteroides thetaiotaomicron, to induce mucin secretion by regulating the differentiation of goblet cells [86].

Caenorhabditis elegans has a simpler and less compartmentalized alimentary canal. It begins with the pharynx, containing a neuro-muscular grinder that is very efficient in masticating bacteria. The grinder empties into the worm gut, an epithelial tube made of 20 polarized epithelial cells [87,88]. In order to colonize C. elegans, microbes must first survive mastication in the grinder. Then, they must adapt to the lumenal pH and oxygen conditions in the gut, while avoiding getting flushed out by gastric motility and defecation. In the gut, some bacteria can directly adhere to the intestinal epithelial cells to create a niche, whereas others remain sessile by persisting and/or replicating in the intestinal lumen without direct adherence ([31]; figure 1d). The C. elegans intestinal lumen is suggested to be aerobic, as their gut microbiota includes several known obligate aerobes, but is also colonized by several fermenting bacteria [75]. Additionally, the pH of the intestinal lumen of C. elegans is weakly acidic and discontinuous [89]. The low intestinal pH of the gut is a possible selection force, also known to modulate inter-microbial interactions during community assembly [90]. In turn, microbes can also condition the C. elegans intestinal environment to maintain pH gradients that are critical for the function of pH-dependent intestinal transporters [68]. In fact, the intestinal pathogens Ps. aeruginosa and Enterococcus faecalis were shown to neutralize the intestinal pH of the lumen upon infection [91], suggesting that microbiome bacteria are likely to play a role in pH alterations upon colonization.

Bacterial adhesion to host mucus glycans is a potential host selection mechanism that permits their gut retention and colonization [92]. Currently, we still do not know if C. elegans contains a flushable mucus layer, similar to other mammals, but we have visualized a potential mucus-like layer on intestinal epithelial cells that may be thick glycocalyx containing mucin-like proteins [93]. While many gut commensals in C. elegans have not been shown to colonize by adherence [21], a few are emerging with obvious direct interactions with gut epithelial cells [14,31]. A possible role of C. elegans mucins in microbial adherence remains understudied and is an active area of investigation. It is possible that in C. elegans, the heavily glycosylated glycocalyx can contribute to the adherence of gut commensals, and the gut microbiota in turn induce mucin-like protein secretion in the glycocalyx, somewhat similar to exchanges seen in mammals. In fact, a recent study has shown a requirement for the C. elegans mucin-like gene, mul-1, in Ps. aeruginosa pathology [94], and mul-1 was previously shown to be upregulated during Ps. aeruginosa infection [95,96]. Super-resolution electron microscopy can be used to visualize the impact of C. elegans gut microbiota on the mucin-glycocalyx layer, and the integrity of underlying microvilli and brush border assembly [97,98]. In addition, in vivo molecular imaging of C. elegans mucin-type O-glycans can be implemented to understand how microbes drive the mucobiology of the C. elegans gut [99].

There are other known factors that impact the homeostasis of the gut microbiome, its mucus layer and the underlying epithelium. Age-related decline in the grinder efficiency and defecation frequency results in more live bacteria entering and proliferating in the gut, leading to constipation and death [17]. In fact, the standard C. elegans diet bacteria, OP50, are efficiently destroyed in the grinder of healthy, young animals, but can colonize the intestinal lumen in aged animals [100]. In addition, age-related depletion of the mucus-like layer and decline in gut barrier function is known to enhance intestinal leakiness and gut dysbiosis, which negatively impacts overall health and lifespan ([101,102]; figure 1d). Different members of the microbiota are known to impact the gut epithelium, with some pathogens disrupting it and a number of commensals fortifying it [103]. For example, several known commensal bacteria in C. elegans have been shown to improve the gut barrier function following pathogenic infections that would increase gut leakiness [58,61]. However, how C. elegans commensals modify gut mucin layer and the underlying epithelium is not well explored. In C. elegans, the impact of microbes on the gut barrier function can be readily accessed by implementing the Smurf assays, where a blue or florescent dye is fed to animals to measure the extent of gut leakiness, as the dye leaks from the damaged gut lumen into the body cavity [102,104].

This article has no additional data.