High resolution RNA-seq profiling of genes encoding ribosomal proteins across different organs and developmental stages in Arabidopsis thaliana

Abstract

In Arabidopsis thaliana, each ribosomal protein (RP) is encoded by a small gene family consisting of two or more highly homologous paralogues, which results in ribosome heterogeneity. It is largely unknown that how genes from multiple member containing RP families are regulated at transcriptional level to accommodate the needs of different plant organs and developmental stages. In this study, we investigated the transcript accumulation profiles of RP genes and found that the expression levels of RP genes are varied dramatically in different organs and developmental stages. Although most RP genes are found to be ubiquitously transcribed, some are obviously transcribed with spatiotemporal specificity. The hierarchical clustering trees of transcript accumulation intensity of RP genes revealed that different organs and developmental stages have different population of RP gene transcripts. By interrogating of the expression fluctuation trend of RP genes, we found that in spite of the fact that most groups of paralogous RP genes are transcribed in concerted manners, some RPs gene have contrasting expression patterns. When transcripts of paralogous RP genes from the same family are considered together, the expression level of most RP genes are well-matched but some are obviously higher or lower, therefore we speculate that some superfluous RPs may act outside the ribosome and a portion of ribosomes may lack one or even more RP(s). Altogether, our analysis results suggested that functional divergence may exist among heterogeneous ribosomes that resulted from different combination of RP paralogues, and substoichiometry of several RP gene families may lead to another layer of heterogeneous ribosomes which also have divergent functions in plants.

Keywords: Arabidopsis thaliana; functional specialization; gene duplication; paralogue; ribosomal protein; ribosome heterogeneity; transcript profiling.

FIGURE 1

A boxplot figure showing the overall transcripts accumulation level of 240 RP genes and total genes across 79 different tissues and developmental stages in Arabidopsis thaliana. The average median value of RPKM (Reads Per Kilobase per Million mapped reads) of 240 RP genes and total genes were calculated from two biological replicates for each sample. The median value of log₂(RPKM) together with upper quartile and lower quartile were plotted against different samples. Blue, RP genes; yellow, total genes

FIGURE 2

A heatmap with hierarchical cluster trees of transcripts accumulation intensity of RP genes across different organs and developmental stages. The value of log₂(RPKM) of each RP gene was plotted against each examined sample. Yellow, relative high expression; black, relative low expression. The vertical hierarchical cluster tree shows the Euclidean distance of examined samples and the horizontal hierarchical cluster tree shows the Euclidean distance of different RP genes based on the log₂(RPKM) value

FIGURE 3

Diversified spatiotemporal transcripts accumulation pattern within RPS15 family. A heat map with hierarchical cluster trees of transcripts accumulation intensity of RPS15 family containing six members (RPS15A, RPS15B, RPS15C, RPS15D, RPS15E, and RPS15F) was made. Yellow, relative high expression; black, relative low expression. The vertical hierarchical cluster tree shows the Euclidean distance of examined samples whereas the horizontal hierarchical cluster tree shows the Euclidean distance of different RP genes based on the log₂(RPKM) value

FIGURE 4

Samples of which the number of RP gene families largely deviating from 1 (<−0.5 or >2) is more than 5. Examined tissues such as (a) F.AN; (b) S.M; (c) L.PET.sn; (d) F.AN.ad; (e) IN.sn; (f) SD.d, of which the number of RP gene families largely deviating from 1 (<−0.5 or >2) is more than 5. (g), proposed hypothesis of free RPs which may act outside ribosomes and heterogeneous ribosomes contributed by some substoichiometric RPs. Relative level was calculated with RPKM value of each RP gene divided by the median RPKM value of total RP genes in each examined tissue. Error bars indicate SD from two independent experiments; asterisk indicates significant difference (t‐test, p < .05)

FIGURE 5

Concerted expression patterns of paralogous RP genes. The expression fluctuation trend of RP genes was investigated from four types of tissues representing important growth and development stages (Parts of 1‐day seedling (hypocotyl, cotyledons, apical meristem with adjacent tissues); seed germination (SD.g1~SD.g3); meristem (M1~M10), and flower development (F1~19.)). RPL34 and RPS29 families were used as examples of RP gene paralogues with concerted expression pattern

FIGURE 6

Contrasting expression patterns of paralogous RP genes. Families such as RPLP0, RPL3, RPL7, RPS2, RPS9, RPS15a, and RPS16 were used as examples of RP gene paralogues with contrasting expression pattern

FIGURE 7

Consensus expression patterns of non‐paralogous RP genes in different developmental stages of meristem. The RPKM value of paralogous and non‐paralogous RP genes from different developmental stages of meristem was treated in accordance with Z‐score using R packages TCseq with the parameters“algo = "cm", k = 6”s