Castor bean organelle genome sequencing and worldwide genetic diversity analysis

Abstract

Castor bean is an important oil-producing plant in the Euphorbiaceae family. Its high-quality oil contains up to 90% of the unusual fatty acid ricinoleate, which has many industrial and medical applications. Castor bean seeds also contain ricin, a highly toxic Type 2 ribosome-inactivating protein, which has gained relevance in recent years due to biosafety concerns. In order to gain knowledge on global genetic diversity in castor bean and to ultimately help the development of breeding and forensic tools, we carried out an extensive chloroplast sequence diversity analysis. Taking advantage of the recently published genome sequence of castor bean, we assembled the chloroplast and mitochondrion genomes extracting selected reads from the available whole genome shotgun reads. Using the chloroplast reference genome we used the methylation filtration technique to readily obtain draft genome sequences of 7 geographically and genetically diverse castor bean accessions. These sequence data were used to identify single nucleotide polymorphism markers and phylogenetic analysis resulted in the identification of two major clades that were not apparent in previous population genetic studies using genetic markers derived from nuclear DNA. Two distinct sub-clades could be defined within each major clade and large-scale genotyping of castor bean populations worldwide confirmed previously observed low levels of genetic diversity and showed a broad geographic distribution of each sub-clade.

Figure 1. Map of the castor bean chloroplast.

The large (LSC, 89,652 bp) and small (SSC, 18,817 bp) single copy regions are separated by two inverted repeats (IRa and IRb, 27,347 bp each), which are colored in grey. The outer most whorl (4) contains genes transcribed clockwise, and the genes on whorl 3 are transcribed counter-clockwise. The inner-most whorl (1) shows the alignment of a nucleotide blast of R. communis against all cassava (M. esculenta) tRNAs and rRNAs from the plastid genome. Whorl 2 shows the alignment of a nucleotide to protein blast (BLASTX) of R. communis against all proteins of the cassava plastid genome (M. esculenta). * Split genes or genes with introns. ** Trans-spliced gene rps12.

Figure 2. Chloroplast read enrichment in methylation filtration libraries.

Percentage of chloroplast reads in each shotgun (blue) or MF (green) library. Available shotgun reads for cultivars from Ethiopia and El Salvador are shown to highlight the enrichment in chloroplast sequences in MF libraries.

Figure 3. Draft assemblies of the chloroplast genomes of 7 castor bean accessions.

Assembled contigs corresponding to the chloroplast genomes of the different accessions are shown in orange, mapped to the reference v. Hale chloroplast genome (x-axis). Individual reads are shown in blue, aligned to the same reference genome. Left y-axis corresponds to the percent identity of each read relative to the reference cv. Hale sequence. Right y-axis shows the percent identity of contig relative to the Hale sequence. Each panel corresponds to a different cultivar: A) US Virgin Islands, B) El Salvador, C) Puerto Rico, D) India, E) Ethiopia, F) Greece, and G) Mexico.

Figure 4. SNP identification and phylogeny obtained from genomic sequencing of 8 castor bean accessions.

A) 83 high quality SNPs identified in the chloroplast genomes of the 8 sequenced chloroplast accessions. The two major clades are shown in red and blue and SNPs that differentiate different sub-clades are shown in orange or green. B) Phylogeny of the 8 castor bean accessions based on the identified SNPs. Two major clades, “A” and “B”, are shown separated by 69 SNPs. Members of each sub-clade are indicated in each branch of the phylogeny.

Figure 5. Worldwide distribution of chloroplast genotypes of castor bean from a collection of 894 accessions.

Distribution of chloroplast genotypes corresponding to the sub-clades shown in Figure 4, based on the origin of each accession. The pie chart corresponding to Florida is expanded to reflect the larger number of samples (n = 272) that came from that state, relative to other parts of the world.