Brave New World of Postgenomics

The Royal Society Discussion Meeting on utilizing the genome sequence of parasitic protozoa was held at the Royal Society in London, March 21-22, 2001.这篇文章汇集的大会的主要观点。

Six years ago Barry Bloom made the prediction that "Sequencing bacterial and parasite pathogens . . . could buy the sequence of every virulence determinant, every protein antigen and every drug target . . . for all time" [1]. Now that this prophecy is rapidly becoming reality, it is important to examine how best to exploit the massive amount of information from the parasite genome-sequencing projects.

In his introductory remarks, K. Vickerman (University of Glasgow, UK) warned the audience, which comprised more than 250 scientists, that we are in danger of being overwhelmed by huge tidal waves of data. Today, the average individual is exposed to as much information in a single day as someone living in the Renaissance would receive in a lifetime. No wonder science is becoming increasingly compartmentalized, with scientists struggling to keep abreast of their own specialized literature and with no time to reflect on the broader issues. When considering whole parasite genomes, comprising thousands of genes, the current challenge is to assemble, catalogue and analyze this information in a robust and useful manner. Bioinformatics is key to this effort, and it became evident during the conference that more tools are needed to tackle information retrieval and analysis if we are to understand parasite biology fully and to relate structure to function at the molecular level. One current problem with bioinformatics can be summarized as "rubbish in, rubbish out" and all scientists need to ensure that information deposited in electronic databases is kept up-to-date and correctly annotated. Only then can the ultimate goal of developing drugs, vaccines and diagnostics to control these devastating diseases be achieved. Welcome to the brave new world of post-genomics!

Thirteen speakers were invited to present their views on how best to address this challenge. The result was an interesting blend of informative and sometimes thought-provoking presentations. Some speakers chose to illustrate how particular techniques could be applied to exploit genomic information in order to address specific biological problems [e.g. kRNA editing (K. Stuart, University of Washington, USA) or erythrocyte invasion (A. Cowman, Walter & Eliza Hall Institute of Medical Research, Melbourne, Australia)]. Others reviewed various strategies that could be applied globally to genome analysis, including: database mining in order to predict metabolic functions (D. Roos, University of Pennsylvania, USA; A. Fairlamb, University of Dundee, UK; S. Oliver, University of Manchester, UK); the use of DNA microarrays to examine stage-specific gene expression (S. Beverley, University of Washington, USA; J. Blackwell, University of Cambridge, UK; D. Carucci, Naval Medical Research Center, USA); and the application of forward and reverse genetic techniques to elucidate the specific functional roles of gene products in areas such as invasion, virulence or drug resistance (A. Tait, University of Glasgow, UK; A. Waters, Leiden University, Netherlands; E. Ullu, Yale University, USA; D. Sibley, University of Washington, USA; and S. Beverley). Many of these lectures touched on the strengths and weaknesses of the available tools that are used to study global gene expression and gene function (e.g. DNA microarrays, RNA interference and proteomics). Not surprisingly, different techniques emerged as having greater utility in one parasite compared with another.

Two honorary members of the parasite community illustrated how functional genomics are being applied to yeast (S. Oliver) and bacterial pathogens (R. Moxon, University of Oxford, UK). Oliver introduced us to the "infoberg" (sequence data being the tip of the iceberg) and the world of "omes," in which the complete set of RNA molecules, proteins and metabolites within a cell (including macromolecules such as lipids and carbohydrates) are termed transcriptome, proteome and metabolome, respectively (figure 1). Unlike the genome, which is fixed in a cell, these vary depending on the environmental context and developmental stage of the organism. Many elegant techniques were described to illustrate the way in which the yeast community is systematically dissecting the Saccharomyces cerevisiae genome. Sadly, many of these techniques would not be applicable to malaria or trypanosomatid genomes. Moxon gave a fascinating account of functional genomics applied to the cell surface lipopolysaccharides (LPS) of Haemophilus influenzae and Neisseria meningitidis. In addition to having key roles in virulence and immune evasion, LPS also has potential as a vaccine candidate, and the systematic knockout of certain biosynthetic enzymes illustrated what might be termed "rational vaccine design."

So how far has Bloom's prediction been realized regarding drug and vaccine targets? Although extensive database analyses will undoubtedly identify putative functions by homology, the more interesting genes that encode the truly novel drug and vaccine candidates will probably remain hidden in the open reading frames of unknown function (ORFANS) that make up more than 50% of each parasite genome. Many of these will undoubtedly be unique to a particular species. For example, the existence of unique thiols, such as trypanothione [2] and ovothiol [3] in trypanosomatids and mycothiol [3] in Mycobacterium tuberculosis, could not have been predicted by database mining alone. Neither can bioinformatics accurately predict the precise structural details of the various complex lipophosphoglycans that adorn the surfaces of many of these organisms. Thus, fundamental research will continue to underpin our efforts to understand the functions of each particular genome. But, in case this sounds too negative, there have been some significant successes arising from parasite genomics. For example, the "thiol-specific antioxidant proteins" identified in EST (expressed-sequence tagged) databases of trypanosomes and leishmania are now known to belong structurally and mechanistically to the peroxiredoxins, a class of peroxidases with specificity for tryparedoxin (a homologue of thioredoxin) [4,5,6]. Similarly, mining of malaria databases has revealed the existence of hitherto unsuspected biochemical pathways, which are similar to those involved in bacterial metabolism. Examples of these pathways are the DOXP pathway [7] and type II or "dissociative" fatty acid biosynthesis [8], which have resulted in promising drug leads, fosmidomycin [7] (an antibiotic) and triclosan [9,10] (a component of toothpastes and mouthwashes), respectively.

It is humbling to think that humans might be only 5-10 times more complex genetically than the parasites within us. No wonder they are so difficult to control!