Morphological mimicry among organisms has long been recognized as an adaptive strategy, but mimicry also occurs at the molecular level. One emerging example is microbial pathogens' use of structural mimics that engage host-cell receptors.
A century and a half ago, in the rainforests of the Amazon, Henry Walter Bates discovered that certain edible butterflies mimic the appearance of poisonous ones to avoid predation. Such adaptation to approximate existing forms, now referred to as Batesian mimicry, illustrates the astounding potential for evolutionary change driven by natural selection. But hidden from the celebrated naturalist's view were similar instances of mimicry occurring at the sub-microscopic level. A growing collection of investigations is now revealing the extraordinary extent to which pathogenic bacteria and viruses mimic host functions. Writing in Cell Host & Microbe, Drayman et al.1 provide new evidence for structural mimicry of host ligands for cellular receptors.
Several pathogenic microorganisms are known to produce proteins that mimic the form and functions of host proteins to exploit cellular machinery and counter immune defences. Indeed, molecular mimics have been identified that subvert all major cellular processes — including cell-cycle progression, membrane trafficking, cytoskeleton dynamics and signal transduction — to promote infections. Certain microbes seem to be especially adept at deploying molecular mimicry2,3. For instance, among the hundreds of genes carried by some large DNA viruses are several that encode proteins with definitive amino-acid sequence similarity to host proteins. These genes are often acquired from host genomes by a process known as horizontal gene transfer; once captured, selective pressure may lead to the accrual of changes to these genes that increase the fitness of the pathogen, thereby enhancing the advantage conferred by mimicry.
Many mimics, however, cannot be recognized by sequence identity owing to either extensive divergence from transferred host genes or convergent evolution2,4, whereby a similar structure or function arises in the absence of common ancestry. As such, the use of X-ray crystallography and related approaches to resolve high-resolution protein structures from pathogen mimics has been central to furthering our understanding of molecular mimicry4. In this spirit, Drayman et al. used existing structural information to screen for novel examples of pathogen mimicry, focusing on potential mimicry of ligands for host receptors.
The authors used established algorithms5 for scoring similarities between protein structures to identify surface proteins from viral and bacterial pathogens that share structural features with host-cell ligands. Reasoning that interactions of ligand mimics with host receptors might facilitate pathogen adhesion and access to host cells, the researchers then tested whether putative mimics could directly bind to their cognate receptors. Despite the small data set used for the comparisons, the approach identified several candidates of mimicry, highlighting the potential of melding structural and evolutionary biology not only for delineating mechanisms of pathogen mimicry, but also for identifying unknown mimics and their targets.
One example of mimicry proposed by the authors was a structural feature of the simian polyoma virus SV40 protein VP1, which resembles Gas6, a ligand of the receptor tyrosine kinase Axl (Fig. 1). The identification of Axl as a probable target of ligand mimicry hints at its importance in viral infections. Axl function has also been implicated in infections with Kaposi's sarcoma-associated herpesvirus and Ebola virus6,7, but its precise role is unclear. This case illustrates how, although informative, the use of structural approaches as a predictive tool for studying mimicry provides only a partial view of the underlying function of the identified mimics. Furthermore, because functional mimicry can evolve in the absence of any structural similarity, many cases of convergent mimicry cannot be uncovered by structural comparisons. To move forward from structure-based screens, we will need to decipher the molecular basis of mimic functions and their contributions to pathogenesis.
Drayman and colleagues' study exemplifies how increasingly powerful tools for exploring the biology of molecules and genomes are helping researchers to uncover additional examples of molecular mimicry, some as exquisite as the classically described cases of morphological mimicry. Molecular mimicry, much like its Batesian counterpart, has major consequences for the evolution of both hosts and pathogens. Mimics pose a daunting challenge to hosts — how does a host differentiate between its native ligands and pathogen mimics, while maintaining core functions?
Research is beginning to illuminate how the rapid accumulation of protein-coding mutations, a conspicuous sign of positive selection, alters host proteins at crucial binding interfaces with pathogen mimics. See-sawing adaptations between hosts and pathogens at these protein surfaces can lead to molecular 'arms races', characterized by intense bouts of positive selection over millions of years8. Studying such evolutionary patterns can provide important insight into the potential for hosts to combat the challenge of mimicry9. Thus, the thriving alliance between evolutionary and experimental approaches10 holds great promise for enhancing our knowledge of the evolution of pathogen mimicry. In this light, it is exciting to consider how Bates's observations of rainforest butterflies long ago might help to inform our understanding of infectious disease today.