Haplotype-resolved genomes of geminivirus-resistant and geminivirus-susceptible African cassava cultivars

Abstract

Background Cassava is an important food crop in tropical and sub-tropical regions worldwide. In Africa, cassava production is widely affected by cassava mosaic disease (CMD), which is caused by the African cassava mosaic geminivirus that is transmitted by whiteflies. Cassava breeders often use a single locus, CMD2, for introducing CMD resistance into susceptible cultivars. The CMD2 locus has been genetically mapped to a 10-Mbp region, but its organization and genes as well as their functions are unknown. Results We report haplotype-resolved de novo assemblies and annotations of the genomes for the African cassava cultivar TME (tropical Manihot esculenta), which is the origin of CMD2, and the CMD-susceptible cultivar 60444. The assemblies provide phased haplotype information for over 80% of the genomes. Haplotype comparison identified novel features previously hidden in collapsed and fragmented cassava genomes, including thousands of allelic variants, inter-haplotype diversity in coding regions, and patterns of diversification through allele-specific expression. Reconstruction of the CMD2 locus revealed a highly complex region with nearly identical gene sets but limited microsynteny between the two cultivars. Conclusions The genome maps of the CMD2 locus in both 60444 and TME3, together with the newly annotated genes, will help the identification of the causal genetic basis of CMD2 resistance to geminiviruses. Our de novo cassava genome assemblies will also facilitate genetic mapping approaches to narrow the large CMD2 region to a few candidate genes for better informed strategies to develop robust geminivirus resistance in susceptible cassava cultivars.

The Cassava Source–Sink project: opportunities and challenges for crop improvement by metabolic engineering

“…This suggests a role for these enzymes in determining storage root yield, which would be consistent with genetic studies in sweet potato that link sucrose synthase to tuber yield (Gemenet et al ., 2020) and the fact that overexpressing the enzyme in potato increased tuber yield (Baroja‐Fernandez et al ., 2009). Today, the availability of chromosome scale cassava reference genomes (Prochnik et al ., 2012; Kuon et al ., 2019) and application of next generation sequencing methods (Tecle et al ., 2014; Rabbi et al ., 2014a; Rabbi et al ., 2014b; Bredeson et al ., 2016) makes GWAS a more comprehensive tool for assessing the genetic and phenotypic variation of the cassava gene pool (Zhao et al ., 2011). Extending GWAS approaches to a range of additional metabolic and physiological traits to reveal their interactions would therefore be a highly useful strategy with respect to the identification of additional targets for the genetic improvement of cassava.…”

“…We used cassava genome information (Bredeson et al ., 2016; Kuon et al ., 2019) to assemble the necessary metabolic toolset for building cassava GSM models that simulate metabolism in source (leaf) and sink (storage root) tissues. These models help us to identify likely metabolic bottlenecks that can be engineered to increase leaf productivity, to facilitate metabolite transport, and to increase storage root growth and starch production in storage roots.…”

Comprehensive genotyping of Brazilian Cassava (Manihot esculentaCrantz) Germplasm Bank: insights into diversification and domestication

“…Genetic gain from breeding in cassava has made little progress over the last century compared to other crops (Ceballos et al 2004). Many challenges confronting cassava breeding (limiting breeding efficiency) include its heterozygous genome, long breeding cycles, clonal propagation, and non-recovery of recurrent genome after single trait introgression (Ceballos et al 2016;Kuon et al 2019). While cassava is mostly clonally propagated, it is an outcrosser with plants still capable of sexual reproduction, and the conscious or unconscious inclusion of seedlings into clonally propagated stock is continually generating new genotypes within a population, thereby increasing allelic variation (McKey et al 2010).…”

Genomic structural variation in ‘Nebbiolo’ grapevines at the individual, clonal and cultivar levels

“…At the same time, long-range genomic platforms have also contributed to improve the de novo genome assembly process, yielding highly contiguous assemblies up to chromosome-scale level. In particular, high-quality diploid genome assemblies for highly heterozygous crops have been obtained, such as: Brassica rapa, Brassica oleracea, Manihot esculenta 12 , as well as several V. vinifera L. cultivars 3,4,[13][14][15][16] . These assemblies offer a smoother starting-point to identify novel features, previously hidden in collapsed and fragmented genome assemblies 4,12 .…”

“…In particular, high-quality diploid genome assemblies for highly heterozygous crops have been obtained, such as: Brassica rapa, Brassica oleracea, Manihot esculenta 12 , as well as several V. vinifera L. cultivars 3,4,[13][14][15][16] . These assemblies offer a smoother starting-point to identify novel features, previously hidden in collapsed and fragmented genome assemblies 4,12 . Integration of multiple platforms is a common practice for obtaining high-quality genome assemblies 2,3,12,14,17 .…”