Development and evaluation of functional gene arrays for detection of selected genes in the environment

To determine the potential of DNA array technology for assessing functional gene diversity and distribution, a prototype microarray was constructed wi th genes involved in nitrogen cycling: nitrite reductase (nirS and nirK) ge nes, ammonia mono-oxygenase (amoA) genes, and methane mono-oxygenase (PmoA) genes from pure cultures and those cloned from marine sediments. In experi ments using glass slide microarrays, genes possessing less than 80 to 85% s equence identity were differentiated under hybridization conditions of high stringency (65 degreesC). The detection limit for nirS genes was approxima tely I ng of pure genomic DNA and 25 ng of soil community DNA using our opt imized protocol. A linear quantitative relationship (r(2) = 0.89 to 0.94) w as observed between signal intensity and target DNA concentration over a ra nge of I to 100 ng for genomic DNA (or genomic DNA equivalent) from both pu re cultures and mixed communities. However, the quantitative capacity of mi croarrays for measuring the relative abundance of targeted genes in complex environmental samples is less clear due to divergent target sequences. Seq uence divergence and probe length affected hybridization signal intensity w ithin a certain range of sequence identity and size, respectively. This pro totype functional gene array did reveal differences in the apparent distrib ution of nir and amoA and pmoA gene families in sediment and soil samples. Our results indicate that glass-based microarray hybridization has potentia l as a tool for revealing functional gene composition in natural microbial communities; however, more work is needed to improve sensitivity and quanti tation and to understand the associated issue of specificity.