Search

Cassava is one of the most important food security crops in tropical countries, and a competitive resource for the starch, food, feed and ethanol industries. However, genomics research in this crop is much less developed compared to other economically important crops such as rice or maize. The International Center for Tropical Agriculture (CIAT) maintains the largest cassava germplasm collection in the world. Unfortunately, the genetic potential of this diversity for breeding programs remains underexploited due to the difficulties in phenotypic screening and lack of deep genomic information about the different accessions. A chromosome-level assembly of the cassava reference genome was released this year and only a handful of studies have been made, mainly to find quantitative trait loci (QTL) on breeding populations with limited variability. This work presents the results of pooled targeted resequencing of more than 1500 cassava accessions from the CIAT germplasm collection to obtain a dataset of more than 2000 variants within genes related to starch functional properties and herbicide tolerance. Results of twelve bioinformatic pipelines for variant detection in pooled samples were compared to ensure the quality of the variant calling process. Predictions of functional impact were performed using two separate methods to prioritize interesting variation for genotyping and cultivar selection. Targeted resequencing, either by pooled samples or by similar approaches such as Ecotilling or capture, emerges as a cost effective alternative to whole genome sequencing to identify interesting alleles of genes related to relevant traits within large germplasm collections. Keywords: Cassava, Pooled targeted resequencing, Herbicide tolerance, Starch biosynthesis, SNP detection

Improvements to massively parallel sequencing have allowed the routine recovery of natural and induced sequence variants. A broad range of biological disciplines have benefited from this, ranging from plant breeding to cancer research. The need for high sequence coverage to accurately recover single nucleotide variants and small insertions and deletions limits the applicability of whole genome approaches. This is especially true in organisms with a large genome size or for applications requiring the screening of thousands of individuals, such as the reverse-genetic technique known as TILLING, where whole genome approaches become cost prohibitive. Using PCR to target and sequence chosen genomic regions provides an attractive alternative as the vast reduction in interrogated bases means that sample size can be dramatically increased through amplicon multiplexing and multidimensional sample pooling while maintaining suitable coverage for recovery of small mutations. Direct sequencing of PCR products is limited, however, due to limitations in read lengths of many next generation sequencers. In the present study we show the optimization and use of ultrasonication for the simultaneous fragmentation of multiplexed PCR amplicons for TILLING highly pooled samples. Thirty-two PCR products were produced from genomic DNA pools representing 265 pooled barley mutant lines. Mutant lines were produced with the chemical mutagen ethyl methanesulfonate. Samples were subjected to 2×300PE illumina sequencing. Evaluation of read coverage and base quality across amplicons suggests this approach is suitable for high-throughput TILLING and other applications employing highly pooled complex sampling schemes. Induced mutations previously identified in a traditional TILLING screen were recovered in this dataset further supporting the efficacy of the approach.

Nucleotide variation, whether induced or natural, is responsible for a vast majority of heritable phenotypic variation. Evaluation of genomic DNA sequence is therefore fundamental for both functional genomics and marker-assisted breeding. The starting point for this is the extraction of high-quality DNA of a suitable quantity for planned downstream applications. A myriad of kits and techniques exists, but not all are suitable for laboratories with limiting infrastructure due to high costs and reliance on toxic chemicals. We describe here a protocol for extraction of high-quality genomic DNA from leaves of a variety of plant species that uses self-prepared buffers and reagents and obviates the need for toxic organic chemicals such as chloroform. We also provide a protocol for using free image analysis software to quantify isolated DNA.

Single-strand-specific nucleases are used to cleave mismatches in otherwise double-stranded DNA. Assays typically involve PCR amplification followed by a denaturing and annealing step to generate heteroduplexed DNA molecules from PCR products containing nucleotide polymorphisms. This is followed by digestion with nucleases that cleave at the site of the mismatch. The molecular weights of the cleavage products indicate the approximate position of nucleotide polymorphisms in PCR amplicons. Cleaved DNA products are observed by one of several readout platforms such as native gel electrophoresis, denaturing gel electrophoresis, capillary electrophoresis, or denaturing high-performance liquid chromatography (DHPLC). This approach is highly suitable for accurate discovery of natural and induced single-nucleotide polymorphisms (SNPs) and also small insertions and deletions (indels). The use of self-extracted single-strand-specific nucleases, typically prepared from crude extracts of celery (Apium graveolens), is common for reverse-genetics (e.g. Targeting Induced Local Lesions IN Genomes (TILLING)) and studies of natural nucleotide polymorphisms (e.g. Ecotilling). While protocols have been published describing the preparation of single-strand-specific nuclease from plant tissues, many rely on toxic chemicals, dialysis and large-volume preparatory centrifuges and also require steps to be performed at 4 °C. We provide here a streamlined room temperature extraction protocol that uses standard microcentrifuges and eliminates toxic chemicals and traditional dialysis.