Under the Microscope Functional genomics of Streptomyces coelicolor

Following the decision, in the late 1990s, The PCR-generated microarrays have to sequence the genome of Streptomyces Prof Colin P Smith been tested with commercially important coelicolor A3(2), the Biotechnology and streptomycetes such as S. clavuligerus School of Biomedical and Biological Sciences Research Council Molecular Sciences and found to readily detect their (BBSRC) of the UK committed substantial University of Surrey respective orthologous genes. This Guildford, Surrey GU2 7XH, UK funding for a coordinated functional opens the way for exploiting the arrays Tel: (44) 1483 68 6937 genomics programme aimed at building a Fax: (44) 1483 30 0374 more broadly in commercial and detailed understanding of this model E-mail: [email protected] taxonomic research programmes. actinomycete species. For current information on the S. coelicolor genome A great advantage of having the in-house DNA microarrays and associated training capacity to produce spotted arrays is that and genome-related tools, see: http://jic- they can be customised. For example, bioinfo.bbsrc.ac.uk/S.coelicolor/. (http://www.surrey.ac.uk/sbms/Fgenomic) (e-mail, [email protected]). We gene probes representing antibiotic gene clusters from other species can be The programme involves a number of UK have focussed largely on producing PCR- designed and spotted alongside the ‘core’ centres, covering the following post- based DNA microarrays, although we S. coelicolor genes, allowing the parallel genomic technologies: have recently completed successful side- measurement of species-specific gene • DNA microarray development, by-side trials with long oligonucleotide expression. provision and training: Colin Smith arrays; a complete ‘long oligo’ set has (University of Surrey, formerly at been produced in collaboration with For generating PCR products, we UMIST, Manchester). MWG Biotech, and are now being used to designed an automated primer selection programme. Similarity searches on each • Proteomics: Keith Chater (John Innes produce arrays at UniS (Figure 1). Centre, Norwich).

• Systematic gene disruption: Keith Chater and Tobias Kieser (JIC) and Paul Dyson (University of Swansea).

• Bioinformatics: Andy Brass (University of Manchester); Douglas Kell (UMIST, formerly at University of Aberystwyth); Chater (JIC); Smith (Surrey).

These technologies are now well established in the respective laboratories and are starting to provide new biological insights. Whole genome DNA microarrays have been produced for gene expression profiling and ‘genomotyping’, more than 1,000 protein spots have been identified, more than 1,000 genes have been knocked out by ‘PCR targeting’ and in vitro transposition technologies, and genome, transcriptome and proteome databases are actively under development.

Figure 1. A Streptomyces coelicolor whole genome DNA microarray produced by We have established a BBSRC-supported the UniS Functional Genomics Laboratory. The cDNA is labelled with Cy3 resource centre at the University of (green) and the genomic DNA is labelled with Cy5 (red). Each spot is Surrey (UniS) for providing Streptomyces approx 150 microns in diameter.

16 MICROBIOLOGY AUSTRALIA • MAY 2004 Under the Microscope candidate probe predicted potential We are now exploiting DNA microarrays metabolite formation. Several transiently cross-hybridisations and allowed to study global patterns of gene induced novel genes have already been selection of ‘unique’ PCR products; these expression in Streptomyces and are identified that are likely to play roles in were then generated by a two-stage particularly focussing on investigating the regulation of development and procedure, firstly using gene-specific patterns of gene expression in time- antibiotic production. primers with universal tags and, course experiments to investigate In parallel, the proteomics effort has subsequently, using universal primers, changes that correlate with identified (by 2-D PAGE and subsequent one of them being 5’-amidated. Arrays developmental and metabolic transitions. MALDI-MS) more than 1,000 protein comprising the majority of open reading Our analysis has concentrated mainly on spots (e.g. see http://qbab.aber.ac.uk/s_ frames have been produced and ‘surface-grown’ cultures and has revealed coeli/referencegel/ for information), and a protocols for RNA isolation, cDNA dramatic changes in gene expression at significant number of unusual post- synthesis and hybridisation have been the ‘decision’ phase prior to the onset of translational modifications have been optimised. aerial mycelium and secondary revealed.

Streptomyces viewed from the inside: the application of proteomics to a model streptomycete

Streptomyces coelicolor A3(2) has Dr Andy Hesketh analytical possibilities that emerge directly from the sequence, and also become the model system for this genus Prof Keith Chater of antibiotic-producing bacteria, thanks to various functional genomics approaches. Department of Molecular Microbiology the life-time commitment of Sir David John Innes Centre Norwich Research Park, Colney, Norwich One particularly exciting prospect is the Hopwood to its genetic analysis 1. The NR4 7UH, UK high throughput application of MALDI- determination of its complete genome Tel: (44) 1603 450000 Fax: (44) 1603 450778 ToF mass spectrometry to the analysis of 2 sequence has made it an even more E-mail: [email protected] the overall protein content of cell or [email protected] valuable model, opening up many culture extracts. In such a proteomics approach, proteins separated by multiple fractionations, most typically by pI pI isoelectric properties and size in classical SCO1947 SCO1947 mwt SCO7511 SCO7511 2D gel electrophoresis, are digested with a protease (typically trypsin) before mass spectrometric analysis. Usually, the resulting mass fingerprint reveals about half of the fragments predicted for any 'bldA (early stationary) Parent (early stationary) one protein, with a mass accuracy sufficient to give unambiguous identification of the cognate gene. Figure 1. Comparison of 2D gel separations reveals that the abundance of gene products from SCO7511, one of three genes for glyceraldehyde-3- phosphate dehydrogenase, is increased by mutation in bldA, while Proteomic analysis of S. ceolicolor synthesis from another, SCO1947, is not significantly affected. The third extends the range of global gene homologue was not detected. expression analysis beyond the

MICROBIOLOGY AUSTRALIA • MAY 2004 17