Ecological drift during colonization drives within- and between-host heterogeneity in animal symbiont populations

Abstract

Host-associated microbiomes vary greatly in composition both within and between host individuals, providing the raw material for natural selection to act on host-microbe associations. Nonetheless, the drivers of compositional heterogeneity in host-associated microbiomes have only rarely been examined. To understand how this heterogeneity arises, we utilize the squash bug,Anasa tristis, and its bacterial symbionts in the genusCaballeronia. We artificially modulate symbiont bottleneck size and strain diversity during colonization to demonstrate the significance of ecological drift, which causes stochastic fluctuations in community composition. Consistent with predictions from the neutral theory of biodiversity, ecological drift alone can account for heterogeneity in symbiont community composition between hosts, even when two strains are nearly genetically identical. When acting on competing, unrelated strains, ecological drift can maintain symbiont genetic diversity among different hosts by stochastically determining the dominant strain within each host. Finally, ecological drift mediates heterogeneity in isogenic symbiont populations even within a single host, along a consistent gradient running the anterior-posterior axis of the symbiotic organ. Our results demonstrate that symbiont population structure across scales does not necessarily require host-mediated selection, but emerges as a result of ecological drift acting on both isogenic and unrelated competitors. Our findings illuminate the processes that might affect symbiont transmission, coinfection, and population structure in nature, which can drive the evolution of host-microbe symbiosis and microbe-microbe interactions within host-associated microbiomes.

A strong priority effect in the assembly of a specialized insect-microbe symbiosis

“…If priority effects play a role in community assembly in this symbiosis, the early-arriving colonist (referred to here as the resident strain) should dominate symbiont communities regardless of which strain or fluorophore is used. Conversely, if community assembly is driven primarily by competition between colonizing strains, then the order of colonization should not be significant, and GA-OX1 should tend to exclude SQ4a, as observed when both strains colonize the host simultaneously (35). When bugs were inoculated first with GA-OX1 GFP, and then with SQ4a RFP, all bugs were colonized exclusively with the first colonist, GA-OX1 GFP (Figure 1B, top); SQ4a RFP was not recovered from any of these nymphs (Figure S1A, left).…”

“…When bugs were inoculated first with GA-OX1 GFP, and then with SQ4a RFP, all bugs were colonized exclusively with the first colonist, GA-OX1 GFP (Figure 1B, top); SQ4a RFP was not recovered from any of these nymphs (Figure S1A, left). Conversely, when bugs were inoculated first with SQ4a RFP and then with GA-OX1 GFP, all bugs were colonized exclusively with the first colonist, SQ4a RFP (Figure 1B, bottom; Figure S1A, right), despite its documented competitive disadvantage during simultaneous pairwise competition with GA-OX1 (35). Thus, symbiont composition between these two treatments significantly differed according to colonization order but not according to strain identity (Figure S1A, W = 0, p-value = 9.98×10 -13 ).…”

“…concitans SQ4a (Figure 1A), each expressing a green or red fluorescent protein marker (GFP or RFP, respectively) that represent different clades within the bacterial genus Caballeronia. In previous experiments in which bugs were co-inoculated with large doses of both strains, GA-OX1 tended strongly to outcompete SQ4a (35). If priority effects play a role in community assembly in this symbiosis, the early-arriving colonist (referred to here as the resident strain) should dominate symbiont communities regardless of which strain or fluorophore is used.…”

“…Nymphs were killed with 70% denatured ethanol, and intact nymphs were imaged on an Olympus SZX16 stereomicroscope with an Olympus XM10 monochrome camera and Olympus cellSens Standard software ver. 1.13, as described in (35). Nymphs were immersed in a shallow volume of PBS in 6 cm plastic Petri dishes, and images were taken in darkfield (30 ms exposure 11.4 dB gain), brightfield (autoexposure, 11.4 dB gain), a GFP channel (autoexposure, 18 dB gain), and an RFP channel (autoexposure, 11.4 dB gain).…”

“…Although this symbiosis is highly specific, many genetically distinct Caballeronia symbiont isolates are capable of colonization in A. tristis , with no apparent difference in the fitness benefits that they confer to the host (33, 34). There is also striking heterogeneity between individual squash bugs in the strain-level taxonomic composition of their symbiont communities (34), raising the question of how community structure and strain diversity emerges and is maintained (35).…”