The history of inquiry in the aphid–Buchnera system highlights how genomics can expand on classical approaches. In 1910, Pierantoni and Sule showed that structures within the aphid, now known as bacteriomes, housed microorganisms, the role of which was unknown. More than a half century later, a series of experiments in which Buchnera-harboring and Buchnera-free aphids were fed on artificial diets like those utilized by MacDonald et al. (2011) strongly suggested that these bacteria provided their hosts with essential amino acids not available in the aphid plant sap diet (reviewed in Shigenobu & Wilson (2011)). While sequencing the first whole genome of Buchnera provided the necessary verification that these bacteria indeed have the capacity to synthesize essential amino acids, it also revealed that a handful of amino acid metabolic pathways were missing important genes. Furthermore, the paucity of Buchnera-encoded transporters provided little insight into the metabolic integration of the symbiosis (Shigenobu et al. 2000). The metabolic mystery was solved last year following sequencing of the first whole genome sequence for an aphid. Remarkably, genome annotation revealed that pea aphids have the genetic capacity to serve as within pathway metabolic collaborators with their symbiotic partners (International Aphid Genomics Consortium 2010; Wilson et al. 2010). For example, in the biosynthetic pathways of five essential amino acids, aphids have some of the genes and do some of the work, while Buchnera have the other genes and do the remainder of the work (see Shigenobu & Wilson (2011) for a complete review). However, these host genome results, while insightful, still explained little about ecologically important variation determining the host ranges of different aphid lineages. Significantly, both MacDonald et al. (2011) and Vogel & Moran (2011) have taken this next step by exploring the genetics underlying the nutritional requirement of different aphid–Buchnera pairs.
While the aphid–Buchnera symbiosis is arguably the best-studied symbiotic system, it is not the only system in which researchers are addressing ecological questions with genomic and postgenomic tools. In fact, the scientific trajectory of the aphid–Buchnera system is being repeated across many symbiotic partnerships. While whole-genome sequencing focused first on smaller microbial genomes because of costs and time constraints, technological advances have facilitated a rapid expansion of available eukaryotic genomes. In fact, the recent expansion of sequenced whole genomes provides us with the first handful of host/symbiont genome pairs (Fig. 2). Amazingly, the genomes themselves are providing important insight into the mechanisms maintaining these intimate associations. For example, sequencing of Bifidobacterium longum revealed that these bacteria, acquired by human infants early in life, utilize complex metabolic machinery to process sugars provided in breast milk (Sela et al. 2008). The human genome affirms that humans are incapable of processing these sugars, suggesting selection on human hosts to provide nutrients for their symbionts. Simultaneous sequencing of the genomes of the body louse Pediculus humanus humanus and its obligate bacterial symbiont Candidatus Riesia pediculicola revealed that the bacteria possess a unique arrangement of genes that underlie synthesis of pantothenate, an essential vitamin not available in the louse diet, revealing selection on the symbionts to provide nutrients for their hosts (Kirkness et al. 2010). Insights gained via genomics are particularly important in systems like the lice–Riesia system because these microbial symbionts cannot be cultured or experimentally manipulated outside of their hosts.
The order of nucleotides within genomes tells us very little in isolation. While the presence of genes is suggestive of underlying process and function, thoughtful follow-up experimentation is essential to confirm genomic suspicions. MacDonald et al. (2011) highlight several of the critical next steps needed to understand host–symbiont interactions following initial genomic revelations. First, comparative genomics within species can reveal both evolutionary processes and ecologically important variation. Comparison of the 12 sequenced pea aphid Buchnera genomes (Shigenobu et al. 2000; Moran et al. 2009; MacDonald et al. 2011), for example, indicates that most mutations resulting in protein changes are likely selected against and rapidly removed from the population. However, a few do remain, and some of those contribute to host diet breadth (Vogel & Moran 2011). Comparative genomics across more symbioses will reveal additional examples of important host and symbiont variation affecting symbiotic function. Second, as complex interaction networks appear to underlie many symbiotic associations, tools to analyse these networks will facilitate building models of the genetic integration of symbioses. More specifically, as metabolism is central to many symbiotic associations, study of metabolic networks, as carried out by MacDonald et al. (2011) through flux balance analysis, can reveal how alteration of host parameters can affect symbiont functioning. Such models, of course, are only as good as their inputs and require both well-annotated genomes and physiological studies. With respect to all nutritional symbioses, the current holy grail remains the development of genetic tools for the generation of transgenic lineages containing knock-downs, knock-outs, or knock-ups of host and symbiont genes so as to facilitate hypothesis-driven dissection of host/symbiont pair function and integration. Future efforts developing such tools for ecologically interesting, nonmodel systems will undoubtedly revolutionize the study of symbiosis once again.
Even in its infancy, investigation of symbioses via paired genomics is powerful. With both host and symbiont perspectives in hand, we see that these interactions are much more intimate than previously understood, with integrated pathways and processes shared across the partners. The extent of this intimacy is yet to be fully explored. While it is clear that metabolic functioning is a result of coordinated effort between many animal and plant hosts and their symbionts, undoubtedly, other functions are similarly affected by symbioses. For example, in the aphid–Buchnera system, even when all essential amino acids are provided to aphids through diet, aphids without their symbiont partner cannot reproduce, suggesting that symbiosis is fundamentally necessary for development as well as metabolism. We further see that these interactions at the genome level have fundamental consequences for ecological interactions, including plant–insect interactions whose boundaries are defined by the metabolic functioning of host and symbiont. As more paired genomes become available, underlying dynamics at the genomic level will undoubtedly provide insight into the functioning of other ecological interactions as well.