The gut microbiota of wild rodents: Challenges and opportunities

Abstract

The gut microbiota can have important, wide-ranging effects on its host. To date, laboratory animals, particularly mice, have been the major study system for microbiota research. It is now becoming increasingly clear that laboratory animals often poorly model aspects of the biology of wild animals, and this concern extends to the study of the gut microbiota. Here, the relatively few studies of the microbiota of wild rodents are reviewed, including a critical assessment of how the gut microbiota differs between laboratory and wild rodents. Finally, the many potential advantages and opportunities of wild-animal systems for research into the gut microbiota are considered.

Introduction

It is now well established that the gut microbiota can have profoundly important, systemic effects on an animal’s biology. Studies with laboratory animals have been central to discovering many of the wider effects of the gut microbiota. But, equally, it is now becoming apparent that artefacts of laboratory animal studies may be bringing unhelpful complexity to the study of the gut microbiota,^e.g.1 and so may have sometimes frustrated progress in this area.

The fundamental rationale for the use of laboratory animals is that they are models, but ‘model’ can mean different things. A laboratory animal can be a general model for a group of related species, for example, laboratory mice being a general model for other mammals or vertebrates: this is model, sense 1. A laboratory animal can also be a model for a specific second species, for example, laboratory mice being a model for humans: model, sense 2. A third type of model is where laboratory animals are models of wild animals: model, sense 3. These three, non-mutually exclusive types of model are therefore making both cross-species comparisons and comparisons between animals leading different lifestyles (i.e. wild v. laboratory). There are obviously many ways in which laboratory animals differ from wild animals: laboratory animals are usually genetically homogenous, well-nourished and free from infection and disease. This contrasts with wild animals where members of a population will be genetically heterogeneous, as well as having different nutritional, reproductive and infection states. Many aspects of an animal’s biology are impacted by its environment and it is reasonable to anticipate that the gut microbiota will be similarly affected in this way. Indeed, studies of laboratory animals have shown very significant differences in the microbiota of the same mouse strain kept in different research facilities, itself testament to the effect of environment on an animal’s microbiota. ²

Here, I will review what is actually known about the gut microbiota of wild rodents, how the microbiota of wild and laboratory animals compares, and ask what important opportunities are provided by studying the gut microbiota in wild-rodent systems.

Gut microbiota of wild mice, Mus spp.

The laboratory mouse (Mus musculus domesticus) is perhaps the pre-eminent laboratory model, but there has only been limited study of wild Mus spp.^3,4 Analysis of the microbiota of M. m. domesticus from eight different European sites showed that the patterns of microbiota diversity was most strongly affected by the geographical origins of the mice, more so than their genetic relationship. ⁵ Comparison of the caecal mucosa-associated and caecal luminal bacterial communities, found that these were more similar within individual mice, than mucosa-associated or lumen-associated microbiota were among different mice, overall pointing to the substantial inter-individual differences in wild-mouse microbiota composition. ⁵

A second study of wild M. m. domesticus also found substantial inter-individual differences in the gut microbiota, with individuals’ microbiota largely clustering according to the three different locations from which they were obtained. ⁶ This work also found that the caecal and rectal microbiota of individuals were on average more similar within an individual, than were caecal or rectal samples among individuals, ⁶ broadly consistent with previous work. ⁵ Despite a relatively small sample size, there were significant correlations between microbiota diversity and animals’ age, body mass, body mass index, as well as virus and macroparasite infection status. ⁵ Analysis of wild M musculus from New York City found consistent bacterial taxa across the city and over a six-month period, which was broadly similar to those described from wild mice elsewhere.^6,7 There was evidence of bacteria potentially pathogenic in people among these New York City mouse populations. ⁷

Analysis of the spatial organisation of bacteria at 10 sites along the gut of wild M. musculus showed that measures of microbial diversity changed along the gut, with the highest diversity in the caecum, colon, rectum and faeces, where anaerobic species were more abundant too. ⁸ Consequently, predicted microbiota gene function also differed between the upper- and lower-intestinal sites. Consistent with other studies in Mus,^5,6 there were significant differences among individual mice in their gut microbiota, with this effect more pronounced for the more anterior gut sites. ⁸ Analogous observations in wild woodrats, Neotoma spp., also described variation in the number of live bacterial cells along the host gut, with the taxonomic variety of these also differing longitudinally and being discordant between neighbouring gut sections. ⁹

In summary, studies of Mus spp. have concluded that there are significant inter-individual differences in the gut microbiota, which raises important questions of the causes and consequences of this for wild animals.

Gut microbiota of wild and laboratory animals compared

Comparison of the gut microbiota of wild M. musculus, laboratory M. m. domesticus and laboratory M. m. musculus showed strong differentiation of the microbiota of the wild and laboratory animals, while the microbiota of the two laboratory sub-species did not differ. ¹⁰ In a separate study, analysis of the microbiota of wild M. m. domesticus and M. m. musculus across their hybrid zone in central Europe and a comparison with laboratory-generated hybrids, showed that measures of the microbiota diversity differed between wild and laboratory animals, between sub-species in the laboratory (but not in the wild), between sub-species and hybrids, but not between male and female animals, nor due to macroparasite infection. ¹¹ Genetic mapping of the difference between the mouse sub-species’ microbial abundance and microbiota diversity showed that fewer than 20 loci could explain much of this. ¹¹

The differences observed in the microbiota of wild and laboratory Mus raise the question of the extent to which the microbiota of wild animals persist when they are brought into captivity? This has been explicitly studied in the desert woodrat Neotoma lepida, which found that during 6 months’ captivity there was a persistence of the wild microbiota, though the relative abundance of the community members changed. ¹² The reverse experiment (where laboratory M. m. domesticus were released into a near-wild environment) showed that the microbiota of rewilded mice rapidly changed away from those of laboratory maintained controls. ¹³

In wild M. m. domesticus there is evidence of two different microbiota ‘enterotypes’ – community of signature taxa. ¹² Tracking of wild animals from when they moved into the laboratory showed that the microbiota moved to just one of the two enterotypes, with this change in enterotype hypothesised to be driven by changes in the animals’ diet. ¹⁴

Direct comparison of the ileocaecal microbiota of wild and laboratory M. m. domesticus found that in mice from different locations the microbiota had a similar community structure, but one that differed to that of laboratory mice. ¹⁵ A series of elegant gut microbiota transplant experiments showed that the microbiota of laboratory mice could be altered by the transplant of wild-mouse microbiota, though this didn’t completely recreate the microbiota of wild mice. ¹⁵ The gut microbiota of mice had a profound effect on their immune responses following a virus infection, such that standard laboratory mice were killed by the infection, whereas laboratory mice that had received wild mouse microbiota were resistant to it. ¹⁵ These different infection outcomes were due to reduced inflammatory responses in laboratory mice in receipt of the wild mouse microbiota, thus showing the very significant effect that the gut microbiota can have on the host immune response. ¹⁵

In summary, these studies show that the microbiota of wild and laboratory animals differ, and also suggest that the gut microbiota changes as animals move from the wild to the laboratory, though the dynamics of this process needs to be explored further.

Gut microbiota of other wild rodents

The gut microbiota of other wild rodents has also been studied. A two-year, longitudinal study of wood mice, Apodemus sylvaticus, showed marked seasonal changes in the composition of the gut microbiota, hypothesised to be driven by seasonal changes in the animals’ diet. ¹⁶ Work with A. flavicollis sought relationships between the gut microbiota and macroparasite infection, finding no such effect on microbiota diversity, but significant effects on the composition on the microbiota and on the abundance of specific bacterial taxa. ¹⁷

Analysis of the gut microbiota of two species of Peromyscus spp. showed no significant differences in the species’ microbiota, though there was substantial inter-individual variation in the microbial communities. ¹⁸ The gut microbiota of wild naked mole-rats, Heterocephalus glaber, has also been described and found to be distinct from those of a range of other mammals. ¹⁹

Gut microbiota effects on an animals’ biology

Studies of laboratory animals have shown that the microbiota has multiple, important effects on animals’ development, immune responses, acquisition of nutrition from food, behaviour, and other physiological processes. Analogous effects are likely to occur in wild animals, such that multiple aspects of the lives of wild animals will be affected by their microbiota. ²⁰ To date, this has hardly been examined in wild animals, but obviously it is a priority to do so.

One important role of the microbiota can be to detoxify components of an animal’s diet. ²⁰ This has been studied in wild woodrats, where dietary toxins commonly present in their diet can increase the diversity of the microbiota (and so detoxify the toxins), whereas novel toxins depress the microbiota diversity. ²¹ The oxalate-rich diet of wild woodrats, Neotoma albigula, appears to be important in driving the presence of potentially oxalate-degrading bacteria in its gut,^22,23 consistent with other studies showing how an animal’s diet can finely-tune the composition of its microbiota. ²⁴ The important insight from these studies is showing that hosts have an evolved association with their microbiota and that it contributes to host fitness.^21,25,26 Studies in bank voles, Myodes glareolus, also show this. Lines selected for maintaining body mass when fed a high-fibre diet had an altered gut microbiota community structure (compared with control lines), with the caecal microbiota being comparatively more diverse. ²⁷

Outstanding questions about the gut microbiota

Despite the relatively few studies of wild rodents, the picture that is emerging is that in these populations there are significant inter-individual differences in the gut microbiota. This is, perhaps, not surprising, but it is important, especially since this is something that is not captured in studies of laboratory animals. Indeed, one of the underlying rationales of using laboratory animals is to minimise differences among individual animals. Key questions pertaining to these inter-individual differences in gut microbial diversity in wild animals are what causes this (for example, the relative importance of genetic and environmental effects and their interaction), what are the dynamics of this, and what are the consequences of this both for the constituent taxa of the microbiota but also for the host itself? As has been elegantly shown, the gut microbiota has a major effect on the host immune response. ¹⁵ By extension, differences among individual wild animals in their microbiota would be predicted to drive immune-heterogeneity among wild animals. ³ This will also feedback since the differing immune state of animals will itself affect components of the microbiota. ¹⁵ A major focus of laboratory animal studies is the immune system. That the gut microbiota can change an animals’ immune response to an infection (changing an animal from being fully susceptible to fully resistant to viral infection) ¹⁵ requires a major re-evaluation of the utility of immunological studies in model laboratory animals. If laboratory mice are being used as a broad-view model of mammals (model, sense 1, above) then these models would continue to have some validity. However, if they are being used as models of free-living animals, be those human or non-human (model, senses 2 and 3, above), then the models’ validity and utility must be severely compromised.

Gut microbiota research opportunities with wild animals

There are several major themes of gut microbiota research that have been pursued using laboratory animals, including the study of the immunological interface between the microbiota and its host, the processes that assemble and maintain the microbiota community composition, and the functional effect (particularly nutritional) of the microbiota. These have yet to be investigated in wild animals, and here I suggest how addressing these questions in wild, rather than laboratory, animals may greatly facilitate microbiota research.

Host-microbiota immunological interface

Work with laboratory animals has shown that the gut microbiota stimulates a host immune response, notably the generation of secretory immunoglobulin A (sIgA) antibodies by the gut mucosa. ²⁸ This response corrals the microbiota to the gut and can also alter its composition. ²⁹ However, the mechanisms by which sIgA and other components of the immune response affect the gut bacteria is far from fully elucidated. ³⁰

The gut microbiota is a mixture of many different bacterial taxa. Many, if not most, of these taxa are commensal; a small number are pathogens. But, beyond this simple dichotomy, whether or not a bacterial taxon is pathogenic also depends upon its context – which other bacterial taxa it occurs with, but also the physiological and immunological state of the host. The host sIgA response differs for different bacterial taxa, shown by some microbiota bacteria having IgA on their surface, but others not. ³¹ IgA-negative gut bacteria are, presumably, commensal, but many of the IgA-positive taxa are presumably commensal too, though among this fraction some will also be pathogens, or potential pathogens. Discordance between the microbiota of wild and laboratory animals (and fewer potential pathogens in laboratory animals) suggest that unpicking the complex interplay between commensal and pathogenic taxa is most appropriately studied in wild animals.

Also of relevance here is work that has studied the immune responses of wild animals, finding that they differ quantitatively in their immune responses compared with laboratory animals, having elevated humoral and cellular responses, but depressed cytokine responses. ⁴ These immunological differences are likely to be due to wild animals being exposed to more infections in the wild resulting in stimulation of the immune system, with down regulation of some immune components being required to avoid inducing immunopathology.^3,32

How the gut immune response shapes the gut microbiota community, and how the gut microbiota itself modifies the gut immune response, is not yet well understood. Given the extensive inter-individual heterogeneity in gut microbiota composition in wild animals, this raises the questions of (a) whether this is caused by immunoheterogeneity among wild animals, for example due to different exposure to other infectious agents, and/or (b) whether different gut microbiota have systemic immunological effects that affects animals’ susceptibility or resistance to other infectious agents?

More broadly still, beyond bacteria and viruses, the gut of wild animals is also home to protozoan and helminth parasites, each of which will likely affect the bacterial microbiota, both directly and indirectly. Indeed, it has been shown that one nematode parasite, Trichuris muris, requires a bacterial microbiota for the hatching of its eggs and to facilitate parasite establishment in a host,^33,34 showcasing the potential complex, multi-trophic interactions occurring within a gut ecosystem.

While in laboratory animals there has been detailed study of the immunology of infections of single parasite species (and sometimes multiple species), the rich, complex gut community present in wild animals is unlikely to be usefully replicated in laboratory animals. It seems inevitable that these multiple parasite species in the gut will have far-reaching effects on the gut bacterial microbiota, emphasising that these phenomena need to be studied in naturally infected animal populations to comprehensively understand the gut microbiota. The different infections that individual wild animals will have is potentially a challenge when seeking to unpick the relative effects of these other infections and of the microbiota on animals. It is important therefore, that such studies will need to be sufficiently robust to be able to account for these potential multifactorial effects.

Drivers of microbiota community composition

Understanding how the microbiota community is constructed and maintained is a major theme in microbiota research, in part overlapping with immunological considerations (above). Studies in laboratory animals have shown that the abundance of components of the microbiota can be heritable – a measure of the proportion of the phenotypic variance of the trait in question that can be explained genetically.^35,36 There is considerable interest in understanding the mechanistic basis of such heritability.

That there is significant inter-individual heterogeneity in wild rodents’ gut microbiota suggests that wild animals have more microbiota phenotypic variation to be explained, compared with laboratory animals. This may have the consequence that the actual heritability of these microbiota traits will be lower in wild animals, compared with estimates derived from studies of laboratory animals. This therefore suggests that to properly quantify the heritability of aspects of the gut microbiota requires that this is studied in wild animals.

The lives of wild animals also vary in time and space. Of particular note, wild animals’ diet varies particularly seasonally which (as discussed above) can result in the modification of the microbiota. Clearly, such variation in diet is not a feature of laboratory animal studies, though their diet can obviously be experimentally manipulated. Understanding the dynamic aspects of the biology of the microbiota (and perhaps temporally changing the heritability of aspects of the microbiota too) would appear to be best achieved by studying these phenomena in wild, rather than laboratory animals.

Measuring the microbiota

Sequence-based characterisation of the microbiota enumerates operational taxonomic units (OTU), from which the relative proportions of different taxa are calculated, so quantifying the microbiota diversity. Recent work in humans casts doubt on the validity of this approach. ³⁷ Specifically, a comparison of people shows that they differ very significantly in the number of bacteria they have in their gut microbiota. Accounting for this when describing the bacterial taxa present (so-called ‘quantitative microbiome profiling’ (QMP)) shows that individuals differ quantitatively in their bacterial microbiome diversity. ³⁷ Critically, this QMP approach gives very different answers than the standard proportional approach. Cleary these results are relevant to studies in laboratory and wild animals, though this now needs to be studied in these systems directly. This is also directly relevant to the immunological interface between the microbiota and the host, given that the host immune response may play a critical role in controlling both the number and the type of bacteria present in a host’s gut.

Summary

Fundamental to the approach of using laboratory animals to understand biological phenomena is allowing the variable under investigation to be manipulated while all other sources of variation are minimised. With this perspective, the genotypic and phenotypic heterogeneity within wild animal populations would seem an unhelpful situation. In fact, this heterogeneity is a powerful research opportunity that can be exploited to rigorously explore microbiota biology and the effects of it on its host. Critically, by studying heterogeneous wild systems, the results are directly relevant to real-world settings rather than being models which can clearly diverge from that which they aim to model.

Study of wild rodents’ gut microbiota is still in its infancy, but this field must rapidly mature to validate, or not, the utility of studies that use laboratory animals to seek an understanding of the important effects that the gut microbiota has in animal biology.