Fig 1.
The ddc #16 EMS population shows overexpression of ATCOPIA28::GFP and upregulation of endogenous TEs.
(A) Schematic representation of the ATCOPIA28::GFP transgene. The 5’ long terminal repeat (LTR) promoter region of an ATCOPIA28 LTR-retrotransposon (AT3TE51900) is used to control the expression of GFP. The construct carries a Nuclear Localization Signal (NLS) to target the GFP in the nucleus. (B) WT and drm1 drm2 cmt3 (ddc) triple mutant plants carrying the ATCOPIA28::GFP transgene showed no and weak GFP fluorescence under UV light, respectively. By comparison, the ddc #16 EMS mutant showed strong GFP fluorescence. Insets show plants under white light. (C) Western blot using anti-GFP antibody confirmed ATCOPIA28::GFP overexpression in ddc #16. Coomassie staining of the large Rubisco subunit (rbcL) is used as a loading control. KDa: kilodalton. (D) Relative expression analyses of ATCOPIA28::GFP (GFP) and three endogenous TEs in ddc and ddc #16 assayed by Real-Time quantitative PCR (RT-qPCR). RT-qPCR analyses were normalized using the housekeeping RHIP1 gene, and transcript levels in the mutants are represented relative to WT. Error bars indicate standard deviation based on three independent biological replicates. Screening of EMS mutant populations was done on MS plates to allow for visualization of GFP-positive individuals under UV light.
Fig 2.
MAIN, DRM2 and CMT3 act synergistically to repress TEs.
(A) Representative pictures showing the developmental phenotype of 3-week-old ddc, main-3 and ddc main-3 mutants in comparison to WT plant. (B) Number of upregulated TEs in ddc, main-3 and ddc main-3, and classified by TE superfamily. (C) Chromosomal distributions of misregulated loci in ddc, main-3 and ddc main-3 over WT. Chromosome arms are depicted in light grey, pericentromeric regions in dark grey as defined in [50]. Upregulated genes and TEs are represented in blue and red, respectively; downregulated genes are represented in green. (D) Fraction of upregulated TEs in ddc, main-3 and ddc main-3 located in chromosome arms or in pericentromeric regions as defined in [50]. Asterisks indicate statistically significant enrichments of TEs in pericentromeric regions in comparison to the genomic distribution of all A. thaliana TEs (Chi-Square test, *: p-value≤ 0.05, **: p-value≤ 0.01 n.s: not significant). (E) Heatmap showing upregulated TEs in ddc, main-3 and ddc main-3 mutants in comparison to WT plants. (F-G) Relative expression analyses of ATCOPIA28 (F) and several endogenous TEs (G) in ddc, main-3, ddc main-3, cmt3 main-3 and drm1 drm2 (dd) main-3 assayed by RT-qPCR. RT-qPCR analyses were normalized using the housekeeping RHIP1 gene, and transcript levels in the mutants are represented relative to WT. Error bars indicate standard deviation based on three independent biological replicates. RNA-seq threshold: log2≥2, or log2≤-2; p-adj< 0.01.
Fig 3.
MAIN and MAIL1 are required for the proper expression of similar genes, and for TE silencing.
(A-B) Number of misregulated genes (A) and upregulated TEs (B) in main-2, mail1-1 and main-3 mutants in comparison to WT Col plants. TEs are classified by superfamily. (C) Heatmap showing misregulated loci in main-2, mail1-1 and main-3 in comparison to Col and WT controls, respectively. Asterisks represents loci that are commonly misregulated in the three mutant backgrounds. (D) Venn diagrams analyses representing the overlaps between misregulated loci in main-2, mail1-1 and main-3. Fisher's exact test statistically confirmed the significance of Venn diagram overlaps (p-value <2.2.10e-16). (E) Fraction of misregulated loci in main-2 and mail1-1 located in chromosome arms or in pericentromeric regions as defined in [50]. Asterisks indicate statistically significant enrichments of downregulated genes and upregulated genes and TEs in chromosome arms and pericentromeric regions, respectively, in comparison to the genomic distributions of all A. thaliana genes and TEs (Chi-Square test, *: p-value≤ 0.05, **: p-value≤ 0.01, n.s: not significant). Percentages of genes targeted by DNA methylation and H3K9me2 were calculated based on enrichment in heterochromatin states 8 and 9 as defined in [51]. RNA-seq threshold: log2≥2, or log2≤-2; p-adj< 0.01.
Fig 4.
The main-2 mutation has a slight effect on non-CG DNA methylation levels.
(A) Genome-wide DNA methylation levels along the five Arabidopsis chromosomes in main-2 versus WT Col plants. Chromosome arms are depicted in light grey, pericentromeric regions in dark grey as defined in [50]. Mb: megabase. (B-H) Boxplot analyses in two main-2 and WT Col biological replicates showing the DNA methylation levels of all pericentromeric TEs (B) and genes (C), all chromosome arms TEs (D) and genes (E), TEs that are upregulated in main-2 (F), and genes that are upregulated (G) and downregulated (H) in main-2. p-values were calculated using a Wilcoxon test. ***: p-value < 2.10e-16.
Fig 5.
MAIN, MAIL1 and PP7L physically interact together.
(A) Representative pictures of 3-week-old main-2 and mail1-1 mutants, and epitope-tagged complementing lines in comparison to WT Col plants. (B) Western blots using anti-FLAG and anti-MYC antibodies showing the accumulation of epitope-tagged PMD proteins at the expected sizes in the different complementing lines. Coomassie staining of the large Rubisco subunit (rbcL) is used as a loading control. KDa: kilodalton. (C-E) Relative expression analyses of upregulated TEs (C), upregulated genes (D) and downregulated genes (E) in the different complementing lines assayed by RT-qPCR. RT-qPCR analyses were normalized using the housekeeping RHIP1 gene, and transcript levels in the complementing lines and mutants are represented relative to WT Col. Error bars indicate standard deviation based on three independent biological replicates. (F) FLAG-tagged MAIN and MAIL1 proteins were immunoprecipitated and putative interacting proteins were identified by mass spectrometry. Numbers of identified spectra, peptides and the normalized spectral abundance factor (NSAFe5) are shown for two independent experiments, including three main-2 and two mail1-1 replicates. WT replicates are used as a negative control. Only proteins reproducibly enriched in all the FLAG-MAIN and FLAG-MAIL1 IP, and depleted in WT controls across multiple replicates are described in the table. (G) MAIL1-MYC was co-immunoprecipitated with MAIN-FLAG in F1 plants obtained by crossing MAIL1-MYC and MAIN-FLAG lines together. Parental MAIL1-MYC and MAIN-FLAG lines were used as negative controls. (H) The MAIN-MYC line was supertransformed with the PP7L-FLAG construct, and MAIN-MYC was co-immunoprecipitated with PP7L-FLAG. Plants expressing only MAIN-MYC or PP7L-FLAG were used as negative controls. (I) Same as H but using MAIL1-MYC plants supertransformed with the PP7L-FLAG construct. Epitope-tagged proteins were detected by Western blotting. Arrowheads indicates expected bands. Asterisks indicates non-specific hybridization. Co-exp: plants co-expressing PP7L-FLAG and MAIN-MYC (H) or PP7L-FLAG and MAIL1-MYC (I).
Fig 6.
main-2, mail1-1, pp7l-2 single and mail1-1 pp7l-2 double mutants display similar developmental and molecular phenotypes.
(A) Representative pictures of 3-week-old main-2, mail1-1, pp7l-2 single and mail1-1 pp7l-2 double mutants in comparison to WT Col plant. (B) Heatmap showing misregulated loci in main-2, mail1-1, pp7l-2 and mail1-1 pp7l-2 mutants in comparison to WT Col plants using the datasets of RNA-seq Exp1, Exp2 and Exp3 (S2 and S4 Tables). One asterisk defines the loci that are commonly misregulated in all mutant backgrounds. Two asterisks define the loci that are misregulated in the mail1-1 pp7l-2 double mutant. (C) Venn diagrams analyses representing the overlaps between misregulated loci in main-2, mail1-1, pp7l-2 and mail1-1 pp7l-2. Fisher's exact test statistically confirmed the significance of Venn diagram overlaps (p-value <2.2.10e-16). (D) Relative expression analyses of upregulated TEs, genes and downregulated genes in the different genotypes assayed by RT-qPCR. RT-qPCR analyses were normalized using the housekeeping RHIP1 gene, and transcript levels in the different mutants are represented relative to WT Col. Error bars indicate standard deviation based on three independent biological replicates. (E-G) Boxplots analyses showing average RPKM values of upregulated TEs (E), upregulated genes (F) and downregulated genes (G) in mail1-1 pp7l-2 in the indicated genotypes of RNA-seq Exp3. These analyses are based on the misregulated loci datasets defined by ** in panel B. P-values were calculated using a Wilcoxon test, and only significant p-values are shown. *: p-value< 1.10e-3; **: p-value < 3.10–6; ***: p-value< 2.10e-16.
Fig 7.
Constitutive heterochromatin appears unaltered in pp7l-2 mutant.
Proportion of nuclei showing condensed, partially decondensed (intermediate), or decondensed chromocenters in the pp7l-2 mutant in comparison to WT control (Col) based on H3K9me2 immunostaining of nuclei. Representative pictures of nuclei displaying condensed, partially decondensed or decondensed chromocenters. DAPI: DNA stained with 4′,6-diamidino-2-phenylindole.
Fig 8.
Evolutionary history of PMD-C and PP7 proteins in plants.
(A) An alignment of the PMD-C motifs from 30 representative Eudicot species was used to construct a phylogenetic tree. The two major clades (MAIL2/MAIL2-like and MAIL3) are indicated. The species codes are given in S11 Table, and corresponding protein sequences in S12 Table). In red are genes presenting a fusion between a PMD-C and a PP7 motif. Statistical supports of key nodes calculated with the approximate likelihood-ratio test are indicated. Scale bar indicates one substitution/site. The tree was rooted using the Amborella trichopoda PMD-C motif (Atr1PMDC). (B) Phylogenetic tree constructed using an alignment of the PP7 motif from the same species as in (A). The two major clades (PP7 and PP7L) are indicated. In red are genes presenting a fusion between a PP7 and a PMD-C motif. Statistical supports of key nodes calculated with the approximate likelihood-ratio test are indicated. Scale bar indicates one substitution/site. The tree was rooted using the A. thaliana PP5 motif (AtPP5).