Alteration of synonymous codon usage bias accompanies polyploidization in wheat

Abstract

The diploidization of polyploid genomes is accompanied by genomic variation, including synonymous nucleotide substitutions that may lead to synonymous codon usage bias (SCUB). SCUB can mirror the evolutionary specialization of plants, but its effect on the formation of polyploidies is not well documented. We explored this issue here with hexaploid wheat and its progenitors. Synonymous codons (SCs) ending in either cytosine (NNC) or guanidine (NNG) were more frequent than those ending in either adenosine (NNA) or thymine (NNT), and the preference for NNC/G codons followed the increase in genome ploidy. The ratios between NNC/G and NNA/T codons gradually decreased in genes with more introns, and the difference in these ratios between wheat and its progenitors diminished with increasing ploidy. SCUB frequencies were heterogeneous among exons, and the bias preferred to NNA/T in more internal exons, especially for genes with more exons; while the preference did not appear to associate with ploidy. The SCUB alteration of the progenitors was different during the formation of hexaploid wheat, so that SCUB was the homogeneous among A, B and D subgenomes. DNA methylation-mediated conversion from cytosine to thymine weakened following the increase of genome ploidy, coinciding with the stronger bias for NNC/G SCs in the genome as a function of ploidy, suggesting that SCUB contribute to the epigenetic variation in hexaploid wheat. The patterns in SCUB mirrored the formation of hexaploid wheat, which provides new insight into genome shock-induced genetic variation during polyploidization. SCs representing non-neutral synonymous mutations can be used for genetic dissection and improvement of agricultural traits of wheat and other polyploidies.

Keywords: DNA methylation; epigenetic variation; nucleotide substitution; polyploidy; synonymous codon usage bias; wheat.

FIGURE 1

SCUB is heterogeneous between hexaploid wheat and its progenitors. (A) Ratio between the numbers of C/G-ending SCs and of A/T-ending SCs for 18 amino acids (Met and Trp not included). (B) Frequency of NNA, NNT, NNC and NNG codons. NNA, NNT, NNC and NNG: SCs with A, T, C and G as the final base, respectively; N denotes any base. The frequency was calculated as the ratio between the number of all SCs ending with A, T, C or G and the total number of SCs. (C) Frequency of NNA/T and NNC/G codons. NNA/T and NNC/G: SCs with A and T or C and G as the final base, respectively; N denotes any base. The frequency was calculated as the ratio between the number of all SCs ending with A and T or C and G and the total number of SCs. Statistical comparison was conducted by Chi square (χ²) test; the difference between two species was calculated with Chi square partitioning; different lowercase letters represent significantly different values (p < 0.05).

FIGURE 2

Influence of the number of introns on SCUB. Frequencies of A-ending SCs (NNAs) (A), T-ending SCs (NNTs) (B), C-ending SCs (NNCs) (C), G-ending SCs (NNGs) (D), or A/T- and C/G-ending SCs (NNA/Ts and NNC/Gs) (E) in genes with up to nine introns. (F) Ratios between A/T-ending SCs and C/G-ending SCs (NNA/Ts and NNC/Gs) in genes with up to nine introns. N denotes any base. The difference between hexaploid wheat and its progenitors was calculated by Chi square (χ²) test, and the results are presented in Supplementary Table S5.

FIGURE 3

Heterogeneity of SCUB as a function of exon position in genes. Frequencies of A/T-ending SCs (NNA/Ts) (A) and C/G-ending SCs (NNC/Gs) (B) as a function of exon position in genes with one to nine introns. (C) Ratios between A/T-ending SCs and C/G-ending SCs (NNA/Ts and NNC/Gs) as a function of exon position in genes with one to nine introns. N denotes any base. The difference between hexaploid wheat and its progenitors was calculated by Chi square (χ²) test, and the results are presented in Supplementary Table S6.

FIGURE 4

Association between SCUB and DNA methylation-driven conversion of cytosines to thymines.(A) SCUB frequencies of NNA and NNG indicating the effect of the second nucleotide position of codons on the conversion of C to T at the third position on the antisense strand. (B) SCUB frequencies of NT|N and NC|N indicating the effect of the first nucleotide position of the next codon on the conversion of C to T at the third position of the previous codon on the sense strand. (C) Ratios between NNA and NNG codon frequencies. (D) Ratios between NT|N and NC|N triplets. NNA and NNG: SCs with A and G as the final bases and any base at the second position; N denotes any base. NT|N and NC|N: SCs with C and T as the final base of the previous codon and any base at the first position of the next codon. The difference between hexaploid wheat and its progenitors was calculated by Chi square (χ2) test, and the results are presented in Supplementary Table S7.

FIGURE 5

Ratios between A-ending SCs and G-ending SCs specifying various amino acids. The statistical comparison was conducted by Chi square (χ²) test, and the results are presented in Supplementary Table S8. The difference between the ratios for Ala, Pro, Ser, Thr and those for Arg, Glu, Gly, Leu, Lys, Val in a species was calculated with a two-sample Student’s t-test (p < 0.05).

FIGURE 6

Association between DNA methylation and SCUB heterogeneity as a function of exon number and position. (A) Ratios between NCA and NCG codons in genes with up to nine introns. (B) Ratios between NT|G and NC|G triplets in genes with up to nine introns. (C) Ratios between NCA and NCG codons as a function of exon position in genes with one to nine introns. (D) Ratios between NT|N and NC|N triplets as a function of exon position in genes with one to nine introns. The difference was calculated by Chi square (χ²) test, and the results are presented in Supplementary Table S9.