A universal tree of protein fold architecture, showing the three most ancient structures that rise at its base. Ancient folds share a common architecture of sheets (in yellow) and helices (in purple) that form either barrels or are interleaved and are highly symmetrical. Example proteins (from left to right) showing these structures include the nitrogenase iron protein from Azotobacter vinelandii, an enzyme important for nitrogen fixation, the xynalase from Penicillium simplicissimum, and an alcohol dehydrogenase from humans.

GENOMES AND EVOLUTION
The evolution of molecular architecture and phylogenomics

The principal focus of research for the next several years will be in the crossroads of genomics and evolution. One important challenge in the now incipient post-genomic era relates to the ‘mapping’ of genotype, phenotype, function and fitness to each other. We are particularly interested in the origins of molecular diversification, transcript networks, biological processes that are linked to co-evolutionary phenomena (such as plant pathogenesis and symbiosis), and the study of levels and patterns of genome-wide mutation. Our research program will focus inquiry on the molecular evolution of macromolecular structure and on phylogenomics.

In the past few years, genome, proteome and transcriptome research resulted in rapid acquisition of nucleic acid and protein sequences. While acquired information has been largely analyzed at the polymer sequence level, there is continuing recognition that higher order structure is fundamental to establish structure-function relationships in biological macromolecules. This has led for example to structural genomic initiatives (e.g., the creation of a complete inventory of protein folds from structural information). Similarly, advances in crystallography have provided unusual mechanistic views of complex macromolecular ensembles, such as the RNA polymerase complex, the nucleosome, and the ribosome. The increasing acquisition of structural information therefore promises to unravel details on the function, interaction and evolution of nucleic acid and protein molecules. However, this necessitates the development of tools for comparative analysis that focus on high-order macromolecular structure.

The idea that biological entities can be related through history of common descent constitutes a general and powerful organizing principle in biology and the basis for phylogenetic analysis of molecules and organisms. Phylogenies can be traced at different levels, from nucleic acid sequences, genes, and molecules to features in individuals, populations, lineages, and species. Since most functional constraints on evolutionary divergence of molecules operate at the level of tertiary structure, three-dimensional structures are generally more evolutionarily conserved than sequences. We have therefore chosen to reconstruct phylogenetic history directly from the structure of proteins and nucleic acids. Using cladistic analysis, we have compared RNA structures at a wide range of phylogenetic levels, from the subspecies analysis of a fungal tree pathogen (Caetano-Anolles et al. 2001) to the universal tree of life (Caetano-Anolles 2002). In these studies, structural attributes were treated as ordered multi-state cladistic characters, and these characters were polarized by a state transformation sequence (grounded in statistical mechanic principles) that assumes that molecules are optimized by a process that increases molecular order. This phylogenetic approach has been extended to the study of a wide variety of macromolecules, and can be used to unravel evolutionary processes and uncover functional relationships in transcript RNA and protein molecules.

Current studies: (1) compare systematically the structure of proteins and nucleic acids at different evolutionary levels, (2) establish which are the ‘contextual’ constraints imposed by the function and inherent properties of these molecules, and (3) delimit a structural morphospace for phylogenomic analysis. Characters that describe how folded, branched, plastic, modular and stable are macromolecules, are used to infer models of molecular change and explore the origin and diversification of life, the existence of lateral gene transfer, and the role of mRNA structure in transcript networks.

CROSSTALK DURING THE INTERACTION OF PLANTS AND MICROORGANISMS
Plant responses to symbionts and pathogens

Legumes and bacteria (rhizobia) interact to establish a unique association that results in the formation of a highly specialized organ, the nitrogen-fixing nodule. This symbiosis is central to agronomical improvement because it provides a biological alternative to the use of nitrogen fertilizer in agriculture. To satisfy the expectations of a growing human population which will surely double in the next 40 years and currently demands 23 million tons of nitrogen per year, crop yield must increase without fertilizer-mediated pollution (leaching of toxic nitrates and ozone depletion). Therefore, biological nitrogen fixation offers the natural framework of a sustainable agricultural system and the possibilities of improvement of legumes through breeding and genetic engineering.

Studies focus on the genes and molecular mechanisms that are responsible for switching on or off the cascade of developmental changes that occur when legume roots are infected by rhizobia, the symbiotic bacteria (Caetano-Anolles 1997). Infection is initiated by rhizobial lipo-oligosaccharide signals. It results in cortical cell re-differentiation that leads to the formation of a new meristem which then develops into a nodule. Although the bacteria trigger this developmental sequence, it is the host which controls most facets of it. We have identified two very important genes, nnr and nts, that appear to be involved in these central aspects of host control. These genes are involved in signal transduction and control the efficiency with which nodule meistems progress into mature nodules.

Other aspects of the signal interplay between microbes and plants are being investigated. Bacterial N-acyl homoserine lactone (AHL) signals coordinate the behavior of individual microbial cells. The successful infection of eukaryotic hosts by bacteria seems particularly dependent on such AHL-mediated “quorum sensing” regulation. The model legume Medicago truncatula detects nanomolar to micromolar levels of bacterial AHLs and responds to changes in the accumulation of hundreds proteins, by the activation of gusA reporter fusions in a gene and tissue-specific manner, and by altered secretion of compounds that mimic AHLs (Mathesius et al. 2003). Eukaryotes respond extensively and specifically to bacterial-quorum signals and this behavior could play important roles in the beneficial or pathogenic outcomes of the interaction between plants and microbes. We are currently exploring links between these responses and the developmental trigger of symbiotic and pathogenic responses.

SPONTANEOUS MUTATION AND GENETIC DIVERSITY
DNA markers for plant and microbe identification and the study of spontaneous mutation in evolution

Molecular biotechnology provides tools for the evaluation and management of resources and change. Nucleic acid markers are important for such endeavor because they can be made to measure constitution, diversity and evolution of genetic material, serving as suitable indicators of wealth, health and future of life. A number of techniques for genome analysis have been applied to the characterization of plant and microbes. These involve both nucleic acid amplification and hybridization. Some of these tools are currently being applied to study plant and microbial diversity and to determine the role of spontaneous mutation in evolution.