Genome biology of the paleotetraploid perennial biomass crop Miscanthus

Abstract

Miscanthus is a perennial wild grass that is of global importance for paper production, roofing, horticultural plantings, and an emerging highly productive temperate biomass crop. We report a chromosome-scale assembly of the paleotetraploid M. sinensis genome, providing a resource for Miscanthus that links its chromosomes to the related diploid Sorghum and complex polyploid sugarcanes. The asymmetric distribution of transposons across the two homoeologous subgenomes proves Miscanthus paleo-allotetraploidy and identifies several balanced reciprocal homoeologous exchanges. Analysis of M. sinensis and M. sacchariflorus populations demonstrates extensive interspecific admixture and hybridization, and documents the origin of the highly productive triploid bioenergy crop M. × giganteus. Transcriptional profiling of leaves, stem, and rhizomes over growing seasons provides insight into rhizome development and nutrient recycling, processes critical for sustainable biomass accumulation in a perennial temperate grass. The Miscanthus genome expands the power of comparative genomics to understand traits of importance to Andropogoneae grasses.

Fig. 1. Allotetraploidy in miscanthus.

a Syntenic relationships between sorghum and M. sinensis subgenomes MsA and MsB. Distribution of subgenome-specific 13-mer sequences (blue for MsA, red for MsB) is shown for each M. sinensis chromosome (see text and Supplementary Note 7.1). b Clustering of counts of 13-mers that differentiate homeologous chromosomes enables the consistent partitioning of the genome into two subgenomes. Blue chromosome names correspond to the A subgenome, red chromosome names correspond to the B subgenome. c Timetree of Andropogoneae showing the timeline of allotetraploidy in the Miscanthus lineage, with divergence and hybridization times of the A and B progenitors estimated from sequence comparisons (Supplementary Note 8). Source Data underlying Fig. 1b are provided as a Source Data file.

Fig. 2. Miscanthus–Saccharum synteny.

Dotplot showing co-orthologs between Miscanthus sinensis and Saccharum spontaneum. All syntenic genes with c score ≥ 0.7 are shown using the mcscan ortholog algorithm. Note the 2:4 ratio of miscanthus:sugarcane chromosome segments. Source Data are provided as a Source Data file.

Fig. 3. Post-allotetraploidy reciprocal exchanges.

a Example of a chromosome pair without reciprocal exchange (chr01–chr02), and two chromosome pairs with distal reciprocal exchanges (chr05–chr06 and chr11–chr12). Red and blue dots represent occurrences of subgenome-specific 13-mers. Black bars identify A and B ancestry inferred from a Hidden Markov Model (Supplementary Note 7.2). b Relative expression of homeologous gene pairs. Across tissues and seasonal sampling times, there is a 3.8% median bias toward the expression of the B member of the pair. c Homeologous gene pairs within reciprocally exchanged regions show the expression bias of their ancestral location. Source Data underlying Fig. 3b, c are provided in the Source Data file.

Fig. 4. Seasonal gene expression changes in miscanthus.

a Shows asparagine in rhizome, stem, and leaves over the growing seasons normalized to total nitrogen in the sample. The error bars represent the standard deviation. b Principal component analysis of RNA-seq read counts normalized using the DESeq2 variance-stabilizing transformation method. PC1/2 distinguishes the three tissues from each other. c PC3/PC4 separates samples based on their nutrient mobilization status. The color scheme for the organs and dates matches b. d Heatmap across all tissues in the study comparing the expression of a subset of genes expressed in tissues that are actively remobilizing nutrients. Source Data are provided as a Source Data file.

Fig. 5. Miscanthus population structure and segmental ancestry.

a Population structure of 407 miscanthus accessions, including 57 M. × giganteus, 120 M. sacchariflorus (Msa), and M. sinensis (Msi) from China (75), Korea (15), and Japan (140). b Principal component analysis of 407 miscanthus accessions where “x” marks admixtures of Msa and Msi. Such hybrids are collectively referred to as M. × giganteus (Mxg), and can be diploid, triploid, or tetraploid. Separation of Japanese and mainland Asian populations is largely consistent with structure analysis in a. Whole-genome shotgun (WGS)-sequenced accessions are labeled. c Segmental ancestry of miscanthus accessions based on WGS sequencing. Each horizontal bar denotes one (imputed) haploid chromosome set; red and blue indicate Msa and Msi ancestry, respectively. The number of bars represents ploidy. Introgression of M. sacchariflorus into M. sinensis (MsiEF148, Undine, DH2, DH2P) is common among cultivated European types (Supplementary Fig. 10). Source Data underlying Fig. 5c are provided as a Source Data file.