There has been an increasing demand in the research reagent, diagnostic reagent, chemical process and pharmaceutical synthesis industries for protein-based catalysts possessing novel capabilities. At present, this need is largely addressed using enzymes purified from a variety of cultivated microorganisms, generally members of the domains Bacteria and Archaea. However, because less than l percent of naturally occurring microorganisms have been grown in culture, alternative techniques must be developed to exploit the full breadth of diversity for potentially valuable new products. We have developed a novel recombinant approach to generate and screen DNA libraries constructed from uncultured, mixed microbial populations. Because the microorganisms differ vastly in abundance in natural populations, a normalization strategy has been implemented. We now propose to implement phylogeny-based sorting of specific cell types from the environmental samples prior to DNA isolation. We also plan to modify existing library construction protocols to allow small DNA samples to be made into representative libraries. We predict that the resulting libraries will allow us to efficiently isolate genes encoding novel biological catalysts, other useful protein products and metabolic pathways. PROPOSED COMMERCIAL APPLICATION: NOT AVAILABLE