The chromosome-scale genome and population genomics reveal the adaptative evolution of Populus pruinosa to desertification environment

Abstract

The Populus pruinosa is a relic plant that has managed to survive in extremely harsh desert environments. Owing to intensifying global warming and desertification, research into ecological adaptation and speciation of P. pruinosa has attracted considerable interest, but the lack of a chromosome-scale genome has limited adaptive evolution research. Here, a 521.09 Mb chromosome-level reference genome of P. pruinosa was reported. Genome evolution and comparative genomic analysis revealed that tandemly duplicated genes and expanded gene families in P. pruinosa contributed to adaptability to extreme desert environments (especially high salinity and drought). The long terminal repeat retrotransposons (LTR-RTs) inserted genes in the gene body region might drive the adaptive evolution of P. pruinosa and species differentiation in saline-alkali desert environments. We recovered genetic differentiation in the populations of the northern Tianshan Mountain and southern Tianshan Mountain through whole-genome resequencing of 156 P. pruinosa individuals from 25 populations in China. Further analyses revealed that precipitation drove the local adaptation of P. pruinosa populations via some genetic sites, such as MAG2-interacting protein 2 (MIP2) and SET domain protein 25 (SDG25). This study will provide broad implications for adaptative evolution and population studies by integrating internal genetic and external environmental factors in P. pruinosa.

Figure 1

Morphological characteristics and genome overview (v2.0) of the Populus pruinosa. (a), (b) P. pruinosa appearance in summer and winter. (c) Shoot and fruit. (d) Long-oval leaves. (e) Ovate leaves. (f) Broad-ovate leaves. (g) The Hi-C heatmap at 100-kb resolution of P. pruinosa genome assembly. Chr1-Chr19 represented the 19 chromosomes. (h) Circos plot of P. pruinosa genome assembly. (a) Assembled 19 chromosomes. (b–i) The distribution of the gene density, GC density, transposon density, tandem repeat density, SSR density, LTR density, Gypsy density, and Copia density, respectively, with densities calculated in 1-Mb windows. (j) Relationship between syntenic blocks, as indicated by lines.

Figure 2

Genome evolution of Populus pruinosa. (a) Evolutionary scenario of the five dicotyledons (P. pruinosa, Populus euphratica, Populus trichocarpa, Arabidopsis thaliana, and Vitis vinifera) from the ancestral eudicot karyotype (AEK) of seven protochromosomes. (b, c) Classification of gene duplicates origin in the genomes of P. pruinosa and P. euphratica. The origins of gene duplicates were classified into five types: WGD/segmental duplication, tandem duplication, proximal duplication, dispersed duplication, and singleton. (d, e) Biological preference of tandem duplicated genes in P. pruinosa and P. euphratica (biological process category). (f) Insertion time of LTR-RTs in five Populus species. (g) Ka/Ks distribution of orthologous genes between P. pruinosa and P. euphratica in LTR-RTs-inserted genes patterns (upstream 2 kb, gene body, and downstream 2 kb). (h) Gene expression changes of orthologous genes between P. pruinosa and P. euphratica. Up-2 kb LTR-RTs, Body LTR-RTs, and Down-2 kb LTR-RTs represented the LTR-RTs-inserted genes in upstream 2 kb of genes, LTR-RTs-inserted genes in the gene body, and LTR-RTs-inserted genes in downstream 2 kb of genes, respectively. Up-2 kb non-LTR-RTs, Body non-LTR-RTs, and Down-2 kb non-LTR-RTs represented the non-LTR-RTs-inserted genes in upstream 2 kb of genes, non-LTR-RTs-inserted genes in the gene body, and non-LTR-RTs-inserted genes in downstream 2 kb of genes, respectively. The asterisks denote significant differences identified through the Wilcox test (^**P < 0.01).

Figure 3

Comparative genomic analysis of Populus pruinosa. (a) Venn diagram of gene families across five Populus species. (b) Phylogenetic trees and gene family evolution in 14 species. Pink indicated expansions, green indicated contractions, and blue indicated positive selections. (c) Biological preference of unique gene families in P. pruinosa. (d) Biological preference of expanded gene families in P. pruinosa.

Figure 4

Population structure of Populus pruinosa. (a) Geographical distribution of 156 P. pruinosa accessions from 25 populations. Blue, and red tree tags on the Xinjiang, China map represented P. pruinosa populations of southern and northern Tianshan, respectively. (b) Population structure of 156 resequencing accessions. ‘NTM’, northern group of Tianshan Mountains. ‘STM’, southern group of Tianshan Mountains. (c) PCA analysis of P. pruinosa accessions. (d) Genetic diversity (Pi) and divergence between NTM and STM. (e) The assessment of effective population size (N_e).

Figure 5

Identification of candidate loci for local adaptation. (a) Spearman’s correlation (above the diagonal) and gradient forest analysis ranking (below the diagonal) of 19 environmental variables. The bold variables indicate the six environmental variables selected for RDA analysis. (b) PCA plot based on RDA axes 1 and 2. (c) and (d) represented Manhattan plots of variants associated with annual precipitation (BIO12) and coldest season precipitation (BIO19), respectively. Dashed horizontal lines represent significance thresholds (P = 0.05). (e), (g) Allele frequencies of candidate adaptive SNPs (e, Chr15:571990; g, Chr2:90871) associated with BIO12 (e) and BIO19 (g) across the 25 populations. (f), (h) Alleles of candidate adaptive SNPs alter protein sequences in exon regions MIP2 (f) and SDG25 (h).