Transposable elements contribute to activation of maize genes in response to abiotic stress

Abstract

Transposable elements (TEs) account for a large portion of the genome in many eukaryotic species. Despite their reputation as "junk" DNA or genomic parasites deleterious for the host, TEs have complex interactions with host genes and the potential to contribute to regulatory variation in gene expression. It has been hypothesized that TEs and genes they insert near may be transcriptionally activated in response to stress conditions. The maize genome, with many different types of TEs interspersed with genes, provides an ideal system to study the genome-wide influence of TEs on gene regulation. To analyze the magnitude of the TE effect on gene expression response to environmental changes, we profiled gene and TE transcript levels in maize seedlings exposed to a number of abiotic stresses. Many genes exhibit up- or down-regulation in response to these stress conditions. The analysis of TE families inserted within upstream regions of up-regulated genes revealed that between four and nine different TE families are associated with up-regulated gene expression in each of these stress conditions, affecting up to 20% of the genes up-regulated in response to abiotic stress, and as many as 33% of genes that are only expressed in response to stress. Expression of many of these same TE families also responds to the same stress conditions. The analysis of the stress-induced transcripts and proximity of the transposon to the gene suggests that these TEs may provide local enhancer activities that stimulate stress-responsive gene expression. Our data on allelic variation for insertions of several of these TEs show strong correlation between the presence of TE insertions and stress-responsive up-regulation of gene expression. Our findings suggest that TEs provide an important source of allelic regulatory variation in gene response to abiotic stress in maize.

Figure 1. Cold stress effects plant growth and gene expression.

(A) Exposure of maize seedlings to cold stress resulted in leaf lesions visible after two days of recovery. A B73 leaf not exposed to cold stress is shown on the left and cold-stressed B73 leaf is shown on the right. (B) Seedlings subjected to cold stress showed decreased growth as measured on the 7^th day of recovery (p-value <0.05; 20 plants were measured for each condition; standard error is shown with vertical lines). Similar decreases in growth and fitness were detected for three other stress conditions. (C) Abiotic stress exposure results in up- or down-regulation for numerous maize genes in each genotype. The log₂(stress/control) values for all differentially expressed FGS genes were used to perform hierarchical clustering of the gene expression values. The genotypes (B73 - B, Mo17 - M, and Oh43 - O) and stress treatments are indicated below each column.

Figure 2. Several TE families are associated with stress-induced up-regulation of gene expression.

(A) and (B) Fold enrichment for down-regulated (A) and up-regulated (B) genes for 283 TE families with the number of expressed WGS genes over 10 is shown as a heat map for four abiotic stress conditions. (C) Fold-enrichment values for each of the 20 TE families associated with gene up-regulation in response to abiotic stress are shown as a heat map. (D) Comparison of distributions of log₂ (stress/control) values between all genes and genes located near certain TE families. The distribution of all genes is shown using a violin plot while the expression changes for individual genes are shown using colored dots. Genes located near ipiki elements are shown on the left and genes located near etug elements are shown on the right with the colors indicating the different environmental stresses. (E) The relative proportion of WGS genes turned on or up-regulated following stress that are associated with the TE families (from C) is indicated for each stress condition in B73. Total number of up-regulated genes is shown for each stress. The expected proportion of genes with insertions of TEs from the enriched families for all expressed genes is less than 1% for all stresses.

Figure 3. Stress-induced up-regulation of gene expression correlates with the variation in TE presence.

(A) Proportion of genes up-regulated in B73 that are also up-regulated in Mo17 and Oh43 is shown for all TE families under the stress condition with highest enrichment for the TE family. (B) The relative expression levels in stress compared to control treatments (log₂ ratio) is shown for B73, Mo17, and Oh43 for each of the 10 expressed genes that are polymorphic for insertions of TEs. The presence/absence of the TE for each genotype-inbred combination is shown by ‘+’ and ‘-‘ symbols. The genes are as follows: 1-GRMZM2G102447; 2-GRMZM2G108057; 3-GRMZM2G071206; 4-GRMZM2G108149; 5-GRMZM2G400718; 6-GRMZM2G347899; 7-GRMZM2G517127; 8-GRMZM2G378770; 9-GRMZM2G177923; 10- GRMZM2G504524. All genes with TE insertion polymorphism are listed in S8 Table.

Figure 4. Validations of correlation between stress-induced up-regulation of gene expression and presence of TEs.

The presence/absence of insertions of ZM00346 elements in the promoter of GRMZM2G108149 (A), GRMZM2G071206 (B), GRMZM2G400718 (C), GRMZM2G102447 (D), and GRMZM2G108057 (E) was assessed by PCR and genotypes were divided according to whether this insertion is present or not (displayed in alphabetical order). The changes in gene expression are shown as log₂(stress/control) values determined using qRT-PCR for each genotype. Vertical brackets correspond to standard error based on three technical replicates of qRT-PCR experiments. The functional annotations for these genes are as follows: GRMZM2G108149- conserved protein involved in amino acid metabolism; GRMZM2G071206- conserved expressed protein involved in nitrogen metabolism; GRMZM2G400718 - unknown conserved protein; GRMZM2G102447 – GCIP interacting protein involved in regulating cell cycle; GRMZM2G108057 – cation transporting ATPase.