Genome of the week - Deinococcus radiodurans Highly resistant to DNA damage –Most radiation resistant organism known Multiple genetic elements –2 chromosomes, 2 plasmids –Why call one a chromosome vs. plasmid?

Why sequence D. radiodurans? Learn how this bacterium is so resistant to DNA damage –This bacterium has nearly all known mechanisms for repairing DNA damage. –Redundancy of some DNA damage repair mechanisms. Use this organism in bioremediation. –Sites contaminated with high levels of radioactivity –DOE (Department of Energy) sequences many microbial genomes - JGI

Data normalization Why do we need to normalize microarray data? –Correct for experimental errors Northern blot example Microbial microarrays –Assume the expression of most genes don’t change –We know every gene - sum the intensity in both channels and make the equal. –Many other ways of normalizing data - not one standard way. Area of active research.

cy3 cy5 Log of Intensities Data Distribution Before and After Normalization Number of clones cy3 cy5

Experimental design Very important - often overlooked. Bacteria are easier to work with than more complex systems. Two types we will discuss in broad terms: –Direct comparison –Reference design –Also loop design (ANOVA)

Yang and Speed, 2002

Direct comparison Directly comparing all samples against each other. Best choice - lowest amount of variation in the experiment. Not the best design –Many samples are to be compared. –RNA is not easy to obtain (often not a problem for microbial systems. –If microarrays are limiting.

Reference design (indirect) Compare all samples to a common reference. –Usually a pool of all samples of RNA or genomic DNA Useful in comparing many samples. Drawbacks: –1/2 of the measurements are not biologically relevant –Each gene is expressed as a ratio/ratio. Variation in the ratios will be higher.

More complicated situations Multifactorial designs

Examples of applications Gene expression –Defining a regulon - targets of a transcription factor. –Functional annotation Identifying regions of DNA bound by a DNA binding protein Genome content Disease diagnosis

Characterization of the stationary phase sigma factor regulon (  H ) in Bacillus subtilis

What is a sigma factor? Directs RNA polymerase to promoter sequences Bacteria use many sigma factors to turn on regulatory networks at different times. –Sporulation –Stress responses –Virulence Wosten, 1998

Alternative sigma factors in B. subtilis sporulation Kroos and Yu, 2000

The stationary phase sigma factor:  H  most active at the transition from exponential growth to stationary phase  mutants are blocked at stage 0 of sporulation Many known sigH promoters previously identified –Array validation

Experimental approach Compare expression profiles of wt and ∆sigH mutant at times when sigH is active. Artificially induce the expression of sigH during exponential growth. –When Sigma-H is normally not active. –Might miss genes that depend additional factors other than Sigma-H. Identify potential promoters using computer searches.

∆sigHwild-type

Hour -1Hour 0Hour +1 wild type (Cy5) vs. sigH mutant (Cy3) citG sacT

Data from a microarray are expressed as ratios Cy3/Cy5 or Cy5/Cy3 Measuring differences in two samples, not absolute expression levels Ratios are often log2 transformed before analysis

Genes whose transcription is influenced by  H 433 genes were altered when comparing wt vs. ∆sigH. 160 genes were altered when sigH overexpressed. Which genes are directly regulated by Sigma-H?

Identifying sigH promoters Two bioinformatics approaches –Hidden Markov Model database HMMER 2.2 (hmm.wustl.edu) –Pattern searches (SubtiList) Identify 100s of potential promoters

Correlate potential sigH promoters with genes identified with microarray data. Genes positively regulated by Sigma-H in a microarray experiment that have a putative promoter within 500bp of the gene.

Directly controlled sigH genes 26 new sigH promoters controlling 54 genes Genes involved in key processes associated with the transition to stationary phase –generation of new food sources (ie. proteases) –transport of nutrients –cell wall metabolism –cyctochrome biogenesis Correctly identified nearly all known sigH promoters Complete sigH regulon: –49 promoters controlling 87 genes.

Identification of DNA regions bound by proteins. Iyer et al Nature, 409: