Transcriptional Programs and Regulators Underlying Age-Dependent and Dark-Induced Senescence in Medicago truncatula

Abstract

In forage crops, age-dependent and stress-induced senescence reduces forage yield and quality. Therefore, delaying leaf senescence may be a way to improve forage yield and quality as well as plant resilience to stresses. Here, we used RNA-sequencing to determine the molecular bases of age-dependent and dark-induced leaf senescence in Medicago truncatula. We identified 6845 differentially expressed genes (DEGs) in M3 leaves associated with age-dependent leaf senescence. An even larger number (14219) of DEGs were associated with dark-induced senescence. Upregulated genes identified during age-dependent and dark-induced senescence were over-represented in oxidation-reduction processes and amino acid, carboxylic acid and chlorophyll catabolic processes. Dark-specific upregulated genes also over-represented autophagy, senescence and cell death. Mitochondrial functions were strongly inhibited by dark-treatment while these remained active during age-dependent senescence. Additionally, 391 DE transcription factors (TFs) belonging to various TF families were identified, including a core set of 74 TFs during age-dependent senescence while 759 DE TFs including a core set of 338 TFs were identified during dark-induced senescence. The heterologous expression of several senescence-induced TFs belonging to NAC, WKRY, bZIP, MYB and HD-zip TF families promoted senescence in tobacco leaves. This study revealed the dynamics of transcriptomic responses to age- and dark-induced senescence in M. truncatula and identified senescence-associated TFs that are attractive targets for future work to control senescence in forage legumes.

Keywords: forage legumes; heterologous expression; transcription factors; transcriptome analysis.

Figure 1

Physiological and molecular analysis of leaf growth and senescence. (A) Phenotypic analysis of leaf growth and senescence in M3 leaves at 10 d, 15 d, 21 d, 28 d, 35 d and 42 d after sowing. (B) Chlorophyll content (chla, chlb and total chlorophyll (chl-ab) in M3 leaf at 10 d, 15 d, 21 d, 28 d, 35 d and 42 d. Data represent mean values (±SD; n = 4), and were analyzed using one-way ANOVA LSD test (p < 0.05). Bars with different letters are statistically different from each other. (C) Expression analysis of photosynthesis-associated gene (PhAG, RBCS) at 21 d, 28 d, 35 d and 42 d. (D) Expression analysis of senescence-associated genes (SAGs, NYE-1 and PaO) at 21 d, 28 d, 35 d and 42 d. Data in (C) and (D) represent mean values (±SD; n = 4) and were analyzed using Student’s t-test (NS; not significant, ** p < 0.01, *** p < 0.001) against 21 d time point.

Figure 2

Transcriptomic analysis of age-dependent senescence. (A) Numbers of differentially expressed genes (DEGs) (Padj ≤ 0.05, Log2FC ± 1) in M3 leaves at day 28, 35 and 42 vs. day 21, calculated from the FPKM values from three biological replicates for each time point. (B) Venn diagram showing overlap among DEGs in M3 leaves at 28, 35 and 42 DAS vs. 21 DAS, using DiVenn (https://divenn.tch.harvard.edu; accessed on 30 March 2022). (C) STEM clustering of DEGs exhibiting similar expression patterns at 21, 28, 35 and 42 DAS. The numbers outside the boxes in (C) represent gene numbers. The numbers in the upper left corner inside the boxes in (C) indicate the profile number, and in the lower left corner, the p values.

Figure 3

Transcriptional regulation of age-dependent senescence. (A) Graphical representation of DE TF families at 28, 35, 42 DAS vs. 21 DAS. (B) DiVenn analysis showing overlap among DE TFs in M3 leaves at 28, 35, 42 DAS vs. 21 DAS.

Figure 4

Physiological and molecular analysis of dark-induced leaf senescence. (A) Phenotype of M3 trifoliate leaves from 28-day-old plants at 0 d, or after an additional 1 d, 2 d, 3 d or 4 d of dark treatment (dD—days of dark treatment). (B) Chlorophyll concentration (Chl-a, Chl-b and total chlorophyll, Chl-ab) in M3 trifoliate leaf at 0 dD, 1 dD, 2 dD, 3 dD and 4 dD. Data represent mean values (±SD; n = 3) and were analyzed using one-way ANOVA LSD test (p < 0.05) and bars with different letters in (B) are statistically not similar to each other. Expression analysis of (C) photosynthesis-associated genes (PhAG, RBCS) and (D) senescence-associated genes (SAGs, NYE-1 and PaO), at 0 dD, 1 dD, 2 dD and 3 dD. Data in (C) and (D) represent mean values (±SD; n = 3), and were analyzed using Student’s t-test (*** p < 0.001) against 0 dD time point.

Figure 5

Transcriptomic analysis of dark-induced senescence in M. truncatula. (A) Number of differentially expressed genes (DEGs) (Padj ≤ 0.05, Log2FC ± 1) at 1 dD, 2 dD, 3 dD vs. 0 dD (dD—day after dark treatment), calculated from the FPKM values from three biological replicates for each time point. (B) Venn diagram showing overlap among DEGs at 1 dD, 2 dD, 3 dD vs. 0 dD using DiVenn. (C) STEM clustering of DEGs exhibiting similar temporal expression patterns at 0 dD, 1 dD, 2 dD and 3 dD. The numbers outside the boxes in (C) represent the gene numbers. The numbers in the upper left corner inside the boxes in (C) indicate the profile number, and in the lower left corner, the p values.

Figure 6

Transcriptional regulation of dark-induced senescence. (A) Graphical representation of differentially expressed transcription factor families at 1 d, 2 d, 3 d vs. 0 d of dark treatment (dD). (B) Venn diagram analysis showing overlap among differentially expressed transcription factors in M3 leaves at 1 d, 2 d, 3 d vs. 0 d of dark treatment using DiVenn.

Figure 7

Comparison of age-dependent and dark-induced transcript profiles. (A) Venn diagram showing overlap among all the DE upregulated genes in age-dependent and dark-induced leaf senescence. (B) Venn diagram showing overlap among all the DE downregulated genes in age-dependent and dark-induced leaf senescence. (C) GO-term enrichment analysis of common upregulated genes during age-dependent and dark-induced leaf senescence. (D) GO-term enrichment analysis of common downregulated genes during age-dependent and dark-induced leaf senescence.

Figure 8

Comparison of M. truncatula and A. thaliana age-dependent and dark-induced transcriptomic profiles. (A) Comparison of M. truncatula age-dependent transcriptomic data with that of A. thaliana reported by Breeze et al., 2011. (B) Comparison of M. truncatula dark-induced transcriptomic data with that of A. thaliana reported by Breeze et al., 2011. (C) Comparison of M. truncatula age-dependent transcriptomic data with A. thaliana dark-induced reported by Law et al., 2018. (D) Comparison of M. truncatula dark-induced transcriptomic data with A. thaliana dark-induced transcriptomic data reported by Law et al., 2018.

Figure 9

Transient expression of putative senescence-associated transcription factors (SA-TFs). (A) Venn diagrams of senescence-induced TFs at different time points in age-dependent and dark-induced senescence. (B) Transient overexpression of ORE-1 and AtNAP (35S:TFs) in tobacco leaves through the agroinfiltration method. Expression of GFP under 35S promoter was used as a positive control for gene expression and negative control for senescence induction. (C) SA-TFs were cloned into expression vectors under 35S CaMV promoter and transiently expressed in tobacco leaves using the agroinfiltration method. Leaves were photographed six to seven days after infiltration.