Speciation trajectories in recombining bacterial species

Pekka Marttinen, William P. Hanage

Introduction

In eukaryotic biology, species are traditionally defined by reproductive isolation. However, bacteria present a significant challenge to this paradigm. While they reproduce clonally, they frequently engage in horizontal gene transfer and recombination, often across established species lines. This "promiscuity" raises a fundamental question: why do bacterial populations not merge into a single genetic continuum? This paper introduces the "Overlapping Habitats Model" to explore the speciation trajectories of recombining bacteria, identifying conditions under which populations either diverge into new species or remain as stable, related clusters.

The Overlapping Habitats Model

The authors address the limitations of previous models, such as the "Ecotype Model," which often fail to account for the nuances of habitat structure and partial competition. They developed two complementary tools to study these dynamics:

  1. Stochastic Simulation: A Wright-Fisher forward simulation tracking core genes, mutation, recombination, migration, and habitat overlap.
  2. Deterministic Approximation: A mathematical framework that predicts population structure evolution and identifies "stationary points" where genetic distance reaches equilibrium.

A central innovation in this study is the use of ecoSNPs—genetic variants present in all members of one cluster but absent in another. The genomic distribution of these variants serves as a diagnostic tool to differentiate between groups undergoing rapid divergence and those maintained in a stable equilibrium.

Satellite Species and Equilibrium

The model reveals that speciation is not always a binary outcome. When significant habitat overlap exists, the cohesive force of recombination can balance the diversifying force of mutation. This leads to the formation of satellite species: stable clusters that maintain a constant genetic distance from a main population indefinitely.

Interestingly, recombination acts as a double-edged sword. While it generally promotes cohesion, increasing sequence divergence reduces the probability of successful recombination. This can eventually trigger a "runaway" divergence once a certain threshold is passed. In stable equilibria, ecoSNPs tend to concentrate in specific genes that have "escaped" recombination, whereas in rapid clonal divergence, they accumulate across the entire genome.

Empirical Validation: S. pneumoniae and C. jejuni

The researchers applied their model to two distinct genomic datasets to validate their theoretical findings:

  • Streptococcus pneumoniae: Analysis of sequence cluster 12 (SC12), an atypical group, revealed an estimated 41% habitat overlap. The model suggests SC12 is a stable satellite species tethered to the main population by frequent recombination in shared environments.
  • Campylobacter jejuni: In contrast, a divergent minority population in C. jejuni showed a lower habitat overlap of approximately 24%. The model predicts this group has passed the tipping point for runaway divergence and is on a trajectory to become a fully separate species.

The study also found that these population structures are relatively resilient to varying migration rates, provided some movement between habitats occurs.

Conclusion and Future Outlook

The Overlapping Habitats Model provides a robust framework for understanding how bacterial diversity is maintained despite genetic exchange. However, the current "two-type" system is a simplification of the multi-niche complexity found in nature. Future research should explore "fragmented speciation," where different genomic regions speciate at different rates, and incorporate evolving habitat specificity to better reflect the dynamic nature of bacterial evolution.

Summary updated

Abstract

Author summary Species are conventionally defined as groups of individuals that breed with each other, but not with those of other species. However, this does not apply to bacteria because, even if they reproduce clonally, DNA may be donated between distinct species. Nevertheless, bacterial species do exist, and a fundamental question is how they are created. We present a mathematical model to describe bacterial speciation. The model predicts that two groups of ecologically different bacteria, assumed to live in partially overlapping habitats, may evolve into genetically distinguishable clusters, without being able to proceed to full separation. Analysis of a divergent Streprococcus pneumoniae subgroup shows that such ‘satellite species’ exist and can be distinguished from more rapidly diverging clusters, like the one we detect in Campylobacter jejuni.
BibTeX
@article{marttinen2017speciation,
  author = {Pekka Marttinen and William P. Hanage},
  title = {{Speciation trajectories in recombining bacterial species}},
  journal = {PLOS Computational Biology},
  year = {2017},
  doi = {10.1371/journal.pcbi.1005640},
}
Figure