research advances

Scaling up: single organisms and metagenomics

SBKB [doi:10.1038/fa_sbkb.2010.40]

Structural genomics and systems biology combine forces to uncover the central metabolic network of Thermotoga maritima.

Before the turn of the millennium, the prospect of determining the structure — and, hence, gaining insight into the function — of every protein produced by a bacterium was inconceivable. Development of the high-throughput structure-determination technologies enabled structural biology to achieve genome-wide scale.

The Protein Structure Initiative (PSI) aims to use a structural genomics approach to increase knowledge of three-dimensional (3D) protein structure. Such increased understanding can provide insights into how individual components of biological networks interact as a whole; indeed, it is through this approach that the PSI has greatly contributed to completely reconstructing, in 3D, the central metabolic network of the hyperthermophilic marine bacterium Thermotoga maritima.

With its relatively small genome (1,877 genes), T. maritima was deemed an ideal target for structural genomics analysis by PSI JCSG. By joining forces with PSI JCMM, the Department of Bioengineering, University of California at San Diego and the Sanford-Burnham Medical Research Institute at La Jolla, the central metabolic network of this bacterium was reconstructed using structural genomics and systems biology techniques 1 .

The investigators characterized a set of 478 proteins that was able to carry out the intracellular metabolic reactions required to form a self-sustaining, metabolic network that enables T. maritima to metabolize diverse carbohydrates and generate by-products such as hydrogen. The structures of 120 of the proteins were determined experimentally; the remaining 358 were predicted and modeled. The structural analysis results also supported the functional assignment of some of the individual gene products: some experimental structures already contained relevant metabolites, whereas others were determined in complexes with known ligands. Analysis of the conservation of folds as a function of network topology revealed that network expansion was probably driven by the recruitment of proteins that carry out similar reactions but are present in different pathways (the so-called 'patchwork' hypothesis of metabolic evolution). Convergent or parallel evolution of proteins carrying out similar reaction mechanisms also seems to have been important in the evolution of the network.

Notably, the investigators also found that there were 714 different domains, but only 182 distinct folds, within the 478 proteins. This number of folds, far fewer than the ∼300 that would be expected in an equivalent random set of proteins, most probably arises because 'core-essential' proteins carry out functions that are strongly linked to the presence of certain folds — their deletion or replacement would render the network non-viable. By contrast, expansion of the network by 'non-essential' proteins can be achieved by the addition of relatively few folds to the repertoire.

This collaborative approach is a far cry from the analysis of individual proteins outside of their full, systems-level, biological context by traditional structural biology methods, and provides an exciting example of the power of the application of high-throughput structural genomics by the PSI.

This T. maritima study is an example of the more traditional approach in microbiology of focusing on a single species. But the exciting new developments in DNA sequencing technology have opened yet another area of research — metagenomics — which involves the study of the genes of collections of microorganisms that are extracted directly from environmental samples, and offers a means of understanding how these microbes interact and prosper within their complex, multispecies communities. The structural genomics technologies can easily be scaled to tackle such problems. The PSI centers, in collaboration with sequencing groups studying marine (J. Craig Venter Institute) and human gut (Washington University, St Louis) environments have selected large numbers of targets that represent these two milieux, focusing primarily on novel protein families found only within organisms that live in these environments. Furthermore, almost 100 structures have been now been determined by the PSI from the dominant gut bacterium Bacteroides thetaiotaomicron, as well as exciting new structures from recently sequenced gut bacteria, such as Eubacteria rectale, that have never before been studied experimentally.

Katrin Legg

  1. Y. Zhang, I. Thiele, D. Weekes, Z. Li, L. Jaroszewski et al. Three-dimensional structural view of the central metabolic network of Thermotoga maritima.

    Science 325, 1544-1549 (2009). doi:10.1126/science.1174671


Explore proteins and this website