Phosphate transporter gene families in rye (Secale cereale L.) - genome-wide identification, characterization and sequence diversity assessment via DArTreseq

Abstract

Phosphorus is a macronutrient indispensable for plant growth and development. Plants utilize specialized transporters (PHT) to take up inorganic phosphorus and distribute it throughout the plant. The PHT transporters are divided into five families: PHT1 to PHT5. Each PHT family has a particular physiological and cellular function. Rye (Secale cereale L.) is a member of Triticeae, and an important source of variation for wheat breeding. It is considered to have the highest tolerance of nutrient deficiency, among Triticeae. To date, there is no report about genes involved in response to phosphorus deficiency in rye. The aim of this study was to: (i) identify and characterize putative members of different phosphate transporter families in rye, (ii) assess their sequence diversity in a collection of 94 diverse rye accessions via low-coverage resequencing (DArTreseq), and (iii) evaluate the expression of putative rye Pht genes under phosphate-deficient conditions. We identified 29 and 35 putative Pht transporter genes in the rye Lo7 and Weining reference genomes, respectively, representing all known Pht families. Phylogenetic analysis revealed a close relationship of rye PHT with previously characterized PHT proteins from other species. Quantitative RT PCR carried out on leaf and root samples of Lo7 plants grown in Pi-deficient and control condition demonstrated that ScPht1;6, ScPht2 and ScPht3;3 are Pi-deficiency responsive. Based on DArTreseq genotyping of 94 diverse rye accessions we identified 820 polymorphic sites within rye ScPht, including 12 variants identified by the SIFT algorithm as having a potentially deleterious effect, of which three are scored as high confidence. SNP density varied markedly between ScPht genes. This report is the first step toward elucidating the mechanisms of rye's response to Pi deficiency. Our findings point to multiple layers of adaptation to local environments, ranging from gene copy number variation to differences in level of polymorphism across Pht family members. DArTreseq genotyping permits for a quick and cost-effective assessment of polymorphism levels across genes/gene families and supports identification and prioritization of candidates for further studies. Collectively our findings provide the foundation for selecting most promising candidates for further functional characterization.

Keywords: Pht genes; Secale cereale L.; expression profiling; gene diversity; low-coverage resequencing; phosphate deficiency; phylogenetic relationships; rye.

Figure 1

Distribution of ScPht genes on the rye Lo7 and Weining chromosomes. (A) Overview of the seven rye Lo7 chromosomes and the ScPht genes location. (B) Overview of the seven Weining Lo7 chromosomes and the ScPht genes location.

Figure 2

Distribution of ScPht genes on the rye chromosomes. Lo7 and Weining chromosomes are shown in green and blue, respectively. Only chromosome segments containing Pht genes are shown, at varying magnification, to ensure sufficient resolution. The scale is in kMb. Green, blue and purple lines indicate putative orthologs within each Pht family. Red lines indicate putatively duplicated genes.

Figure 3

Phylogenetic relationships of PHT protein families from Secale cereale L. Lo7 and Weining, Oryza sativa, Sorghum bicolor, Glycine max and Arabidopsis thaliana. The protein tree was constructed using the maximum likelihood method. with the WAG+G4+F as the best-fitting substitution model. The ultrafast bootstrap and single branch test were inferred from 1000 replicates.

Figure 4

Phylogenetic relationships of PHT1 proteins from different monocot grass species. The phylogenetic protein tree contains PHT1 proteins from Secale cereale L., Triticum aestivum, Hordeum vulgare, Oryza sativa, Sorghum bicolor, Zea mays, Setaria viridis, Brachypodium distachyon, Glycine max and Arabidopsis thaliana. The protein tree was constructed using the maximum likelihood method with the JTT+R5 as the best-fitting substitution model. The ultrafast bootstrap and single branch test were inferred from 1000 replicates.

Figure 5

Location of P1BS cis-regulatory elements in the promoter region of ScPhts genes.

Figure 6

Relative expression levels of phosphate transporter genes assessed by qRT-PCR in leaf (A) and root (B) tissue under two treatment conditions: phosphate deficiency (Pi deficiency) and phosphorus sufficiency (control), at two-time points (14 days and 21 days). Each bar plot represents the mean 2^-ΔΔCt values obtained from three independent biological replicates, with error bars indicating the standard error of the mean (SEM). Statistical significance between control and treatment conditions was determined using the Kruskall test (*p < 0.05).

Figure 7

Genetic diversity of 94 rye accessions analyzed by DArTreseq. (A) PCoA plot based on 190430 genome-wide SNPs. (B) NJ tree based on 190430 genome-wide SNPs.

Figure 8

Diagram of the nonsynonymous deleterious SNP location within the ScPht members. (A) Location of deleterious SNP within the coding regions of ScPht genes. (B) Amino acid changes due to deleterious SNPs in the protein sequences of PHT transporters.