We have begun to investigate the use of oligonucleotide microarray Comparative Genome Hybridization (array CGH) as an adjunct to our current PCR-based method of screening for deletion mutants in C. elegans. We designed a pilot microarray chip, manufactured by NimbleGen Systems, Inc., that is composed of a tiled set of probes to nearly 90 percent of the exons on chromosomes X and II. Adjacent 50-mer probes in this set overlap by as much as 29 bp, and were selected using a set of criteria to minimize non-specific hybridization, the tendency to form secondary structure, and the range of predicted Tm values.Initial experiments demonstrated that homozygous deletions are easily detected by this method, with a very good signal-to-noise ratio. We successfully detected a 50-kb deletion affecting nine genes on the X chromosome in the strain VC100, as well as a deletion of just 1047 bp affecting the gene
ceh-39. The chip allows us to detect deletion breakpoints at the resolution of specific exons. A further experiment comparing hermaphrodite (XX) and male (XO) DNA samples demonstrated the ability to detect single copy number differences for probes across the entire X chromosome. We have also reliably detected a 1202-bp deletion in
dab-1 in selected heterozygotes and from wash-sampled balanced heterozygous populations. We are currently planning blind screens for novel deletions in progeny derived from mutagenized populations.Array CGH promises a number of advantages over our current screening strategy, including no constraint on the maximum detectable deletion size, the ability to screen for deletions in many thousands of genes in a single experiment, and a large reduction in the labour involved in the process of sibling selection. At current probe density three chips are required for genome-wide screens, but custom chips can be easily designed to encompass any subset of desired genomic regions. We are currently selecting probes to the other four chromosomes and generating chips to include those probe sets. The resolution afforded by this approach will also allow a better estimate of the frequency of induced deletions under different mutagenesis protocols.