A framework for testing the impact of co-infections on host gut microbiomes

Abstract

Parasitic infections disturb gut microbial communities beyond their natural range of variation, possibly leading to dysbiosis. Yet it remains underappreciated that most infections are accompanied by one or more co-infections and their collective impact is largely unexplored. Here we developed a framework illustrating changes to the host gut microbiome following single infections, and build on it by describing the neutral, synergistic or antagonistic impacts on microbial α- and ß-diversity expected from co-infections. We tested the framework on microbiome data from a non-human primate population co-infected with helminths and Adenovirus, and matched patterns reported in published studies to the introduced framework. In this case study, α-diversity of co-infected Malagasy mouse lemurs (Microcebus griseorufus) did not differ in comparison with that of singly infected or uninfected individuals, even though community composition captured with ß-diversity metrices changed significantly. Explicitly, we record stochastic changes in dispersion, a sign of dysbiosis, following the Anna-Karenina principle rather than deterministic shifts in the microbial gut community. From the literature review and our case study, neutral and synergistic impacts emerged as common outcomes from co-infections, wherein both shifts and dispersion of microbial communities following co-infections were often more severe than after a single infection alone, but microbial α-diversity was not universally altered. Important functions of the microbiome may also suffer from such heavily altered, though no less species-rich microbial community. Lastly, we pose the hypothesis that the reshuffling of host-associated microbial communities due to the impact of various, often coinciding parasitic infections may become a source of novel or zoonotic diseases.

Keywords: Co-infections; Disease ecology; Dysbiosis; Gut microbiome; Helminths; Non-human primate; One health; Parasites; Virus; Wildlife health.

Fig. 1

The impact of single infections on α- and ß-diversity of the host’s microbiome with examples. A Single infections can have a directional effect on microbial species diversity. Equine gut microbial α-diversity, for instance, decreased following helminth infection [17]. B Single infections may result in deterministic changes to the microbial community composition (i.e., ß-diversity), which are characterized by a shift of the centroid (= black dot; e.g., analysed by Permutational Multivariate Analysis of Variance, Permanova). In this case, the dispersion stays similar (e.g. analysed by Permutational Analysis of Multivariate Dispersions, Permdisp). For example, the gut microbial composition shifted in Adenovirus-infected mouse lemurs [11]. C Alternatively, single infections may lead to a changed dispersion, which can be visualized as distance to centroid (spread = black arrow). An example are chimpanzees infected by the simian immunodeficiency virus, which had a more dispersed gut microbiome [21]. D Single infections can also lead to both stochastic and deterministic effects. Three-spined stickleback (Gasterosteus aculeatus), for instance, infected with the cestode Schistocephalus solidus had a more dispersed and shifted gut microbial community [22]. * = significant differences (i.e., p-value < 0.05); ns = non-significant differences (i.e., p-value > 0.05)

Fig. 2

A framework to assess the impact of co-infections on α- and ß-diversity of the host’s microbiome. The top banner provides an overview of different α-diversity metrics, and the patterns created by a shift in centroid (i.e., a deterministic effect with centroid = black dot) or when plotted as distance from centroid (i.e., a stochastic effect with spread = black arrow), both describing ß-diversity. Based on the impact a single infection (yellow) has on uninfected hosts (blue), the effect of a co-infection (red) can be classified as either neutral, synergistic or antagonistic. Animal symbols are in reference to the focal organism of studies featured in Table 1 and the number represents the frequency a similar result was found. * = significant differences (i.e., p-value < 0.05); ns = non-significant differences (i.e., p-value > 0.05)

Fig. 3

Differences in gut microbial α- and ß-diversity in uninfected, single-infected and co-infected mouse lemurs (M. griseorufus). A α-diversity measured by Faith’s phylogenetic diversity, Chao1, Shannon diversity, Simpson (left to right) and B ß-diversity measured by weighted and C unweighted UniFrac distances and illustrated by non-metric multi-dimensional (NMDS) ordination plots. Displayed are uninfected (blue squares), single-infected (helminth⁺: yellow triangle, AdV⁺: orange circles) and co-infected (red diamonds) mouse lemurs

Fig. 4

Pathways to novel, potentially pathogenic bacterial disease agents. Aside from direct changes to the host microbiome caused by habitat disturbances (blue arrows), parasites directly impact host microbiome (red solid arrow) and indirectly via manipulation of host health (red dashed arrow). Since anthropogenically disturbed habitats facilitate the transmission and persistence of parasites, direct and indirect parasite-mediated changes to a host’s microbiome may become more frequent, and a dysbiotic gut may become a source of harmful bacteria. Adapted from [30]