Pool deconvolution approach for high-throughput gene mining from Bacillus thuringiensis

Novel genes from Bacillus thuringiensis (Bt) are required for effective deployment in agriculture, human health, and forestry. In an improvement over conventional PCR-based screening, next generation sequencing (NGS) has been used for identification of new genes of potential interest from Bt strains, but cost becomes a constraint when several isolates are to be sequenced. We demonstrate the potential of a DNA pooling strategy known as pool deconvolution to identify commercially important toxin genes from 36 native Bt isolates. This strategy is divided into three steps: (a) DNA pooling, (b) short read sequence assembly followed by gene mining, and (c) host isolate identification. With this approach, we have identified insecticidal protein (ip) genes including nine three-domain (3D) cry genes, three cyt-type genes, three mtx genes (mosquitocidal toxin), and one bin and vip-type gene each. Three cry-type and three cyt-type genes were cloned, out of which, two cry-type genes, ip11 and ip13, were named as cry4Ca2 and cry52Ca1, respectively by the Bacillus thuringiensis nomenclature committee. Our results show that the pool deconvolution approach is well suited for high-throughput gene mining in bacteria.