Evolutionary dynamics following HGT
An unexpected finding from analysis of sequenced microbial genomes is the extent to which genetic material has been transferred across distant lineages. Horizontal gene transfer (HGT) has the potential to introduce large gene clusters encoding entire blocks of metabolism which may spark quantum leaps in organism function. Current research on HGT is dominated by comparative genomics approaches that have identified many genes putatatively transferred at distant times in the past (millions to billions of years ago).
Surprisingly little attention has been paid to understanding the immediate consequences of the incorporation of new functions through HGT. Although regions acquired via HGT may provide an initial fitness benefit, many aspects (expression, activity, regulation, etc) of the transferred region may be suboptimal, especially with increasing phylogenetic distance between the two organisms. Furthermore, the recipient’s physiology is likely suboptimally adapted to performing a novel metabolism encoded by the transferred DNA.
I suggest a model for the acquisition of new traits via HGT whereby rare transfer events provide the raw genetic material for subsequent optimization within the context of the recipient’s genome (see figure). In Curve 1 the transferred region provides substantial fitness gain, followed by only slight optimization. In curves 2 and 3, beneficial mutations after HGT result in significant optimization. The degree and type of epistasis (interaction between mutations) will affect the shape of the curve. Transitive beneficial mutations (additive and irrespective of order), or antagonistic interactions, will display a steep rise followed by diminishing returns (Curve 2). Alternatively, multiple mutations, individually of small benefit, may interact synergistically to result in significant fitness gains (Curve 3).
A desirable system for addressing the role of HGT in the evolution of novel capacities needs to somehow also be achievable within a reasonable (fundable) timeline. Thus, rather than comletely inventing a new capacity from scratch, our efforts are beginning with an engineered HGT system has been devised to provide an unrelated (different enzymes and intermediates), but functionally analogous formaldehyde oxidation pathway to a mutant background that lacks the endogenous formaldehyde oxidation pathway. This is rather like swapping out a gas engine for an electric one – you may still have a car that can turn and stop and move – but is unlikely to be set-up in an optimal way.
Prior work established that the tetrahydromethanopterin-dependent pathway of Methylobacterium could be functionally replaced with a glutathione-dependent pathway fromParacoccus denitrificans (ref 7 of publications).This was true despite the fact that I had ‘cheated’ by using a strong endogenous promoter to express the foreign glutathione pathway – thereby removing one potential hurdle (sufficient expression) that material acquired via HGT would likely face. The complemented strains grew ~4x more slowly than wild-type on methanol, however, leading to the hypothesis that significant optimization would occur subsequent to the transfer (such as in the figure above, curves 2 or 3).
Although this experiment is still in early stages, it is clear that significant improvement is possible very early. One population, for example, has evolved to be 80% more fit in less than 100 generations – the same degree of improvement experienced by E. coli grown in glucose for 35,000 generations! Among the current directions of this project are to use genetic techniques to distinguish whether fitness increases derive from the host genome, the HGT region, or both. The hypothesis that adaptation following HGT is a coevolutionary process predicts that there will come to be specificity as to the effect of combinations of given beneficial mutations in the HGT region and the rest of the genome. In particular, cases where the HGT region apparently contains beneficial mutation(s) will prompt sequencing of these defined regions in order to identify the nature of the mutations.