BAF60/SWP73 subunits define subclasses of SWI/SNF chromatin remodelling complexes in Arabidopsis

Abstract

Evolutionarily conserved switch-defective/sucrose nonfermentable (SWI/SNF) ATP-dependent chromatin remodelling complexes (CRCs) alter nucleosome positioning and chromatin states, affecting gene expression to regulate important processes such as proper development and hormonal signalling pathways. We employed transcript profiling, chromatin immunoprecipitation (ChIP), mass spectrometry, yeast two-hybrid and bimolecular fluorescence complementation protein-protein interaction studies, along with hormone and metabolite profiling and phenotype assessments, to distinguish the SWP73A and SWP73B subunit functions in Arabidopsis. We identified a novel subclass of SWI/SNF CRCs defined by the presence of the SWP73A subunit. Therefore, we propose a refined classification of SWI/SNF CRCs in Arabidopsis, introducing BRM-associated SWI/SNF (BAS)-A (containing SWP73A) and BAS-B (containing SWP73B) subclasses. The SWP73A- and SWP73B-carrying SWI/SNF CRCs exhibit differential properties, demonstrated by distinct chromatin binding patterns and divergent effects on hormone biosynthesis and metabolism. We additionally found that SWP73A plays a specific role in the regulation of auxin signalling, root development, metabolism and germination that cannot be fully compensated by SWP73B. We recognised that some atypical subclasses of SWI/SNF CRCs may be likely formed in mutant lines with inactivated SWP73 subunits. Our study reveals that the duplication of the SWP73 subunit genes contributes to unique and shared functions of SWI/SNF CRC subclasses in the regulation of various processes in Arabidopsis.

Keywords: Arabidopsis; SWI/SNF; SWP73/BAF60; chromatin remodelling complexes; hormonal signalling; metabolome.

Fig. 1

SWP73 subunits of switch‐defective/sucrose nonfermentable (SWI/SNF) chromatin remodelling complexes (CRCs) differentially influence gene expression in Arabidopsis. (a) The phenotypes of 3‐wk‐old wild‐type (WT) plants and T‐DNA insertional lines used in RNA‐Seq analysis. (b) Differentially expressed genes (DESeq2, FC > 1.5, adjusted P‐value < 0.05) show increasing numbers among swp73a, swp73b and swp73a/SWP73A; swp73b/swp73b, indicative of unequal functional redundancy of SWP73A and SWP73B. (c) swp73a, swp73b mutants, and swp73a/SWP73A; swp73b/swp73b sesquimutant exhibit specific and partially overlapping transcriptomic changes among upregulated genes. (d) swp73a, swp73b mutants, and swp73a/SWP73A; swp73b/swp73b sesquimutant exhibit specific and partially overlapping transcriptomic changes among downregulated genes.

Fig. 2

Arabidopsis SWP73A and SWP73B exhibit differential genome‐wide distribution and are present in various subclasses of the switch‐defective/sucrose nonfermentable (SWI/SNF) chromatin remodelling complexes (CRCs) but may cooperate on some target genes. (a) SWP73A and SWP73B bind a set of unique and common targets (reanalysis based on Jégu et al., ; Huang et al., 2021). Asterisks denote statistical significance P < 8.848e‐61 determined by the hypergeometric test. (b) Complementary chromatin occupancy profile of SWP73A and SWP73B generated based on all protein‐coding genes from Araport11 and targets shared for SWP73 indicate differential genome‐wide distribution of SWP73 subunits. (c) Graphical representation of the SWP73A‐specific SWI/SNF subclass: BRM‐associated SWI/SNF (BAS) lacking SWP73A impacts plant growth under standard, long‐day conditions. (d) Graphical representation of the SWP73B‐specific SWI/SNF subclasses: BAS, MINUSCULE‐associated SWI/SNF (MAS), and SYD‐associated SWI/SNF (SAS) lacking SWP73B impacts plant growth under standard, long‐day conditions.

Fig. 3

SWP73A and SWP73B differentially impact the Arabidopsis response to selected phytohormones and hormone biosynthesis. (a) Pinoid‐like structures occurring in swp73b flowers. (b) swp73b plants exhibit hypersensitivity to 2,4‐D demonstrated as enhanced growth inhibition. Letters correspond to statistical significance determined by the Kruskal–Wallis ANOVA with post hoc Dunn's test (P < 0.001). (c) swp73a mutation causes increased frequency of lateral root formation in response to indole‐3‐acetic acid (IAA) and 2,4‐D. Charts represent lateral root numbers in the wild‐type (WT) and swp73a counted from first to third day after transferring 6‐d‐old seedlings from a ½‐strength Murashige & Skoog medium (½MS), to ½MS, ½MS with 50 μM IAA or 1 mM 2,4‐D. Error bars denote the SD, while asterisks indicate statistical significance (P‐value < 0.05, Student’s t‐test) (d) SWP73A protein is present in the root meristem zone, lateral root (including root primordia) and root hair. (e) Indole‐3‐acetic acid level in the WT, swp73a and swp73b mutants. Error bars denote the SD, while asterisks denote statistically significant enrichment compared with WT plants (t‐test, the asterisk denotes P‐value < 0.05). (f) swp73b plants exhibit specific cell cycle alterations demonstrated by the increased cell number in leaves epidermis that are reversed by the gibberellin spraying (100 mM GA_(4 + 7)). Counted from at least 10 different leaves for each sample. Letters correspond to statistical significance determined by the Kruskal–Wallis ANOVA with Dunn's post hoc test. On the right, representative images of leaf surface collected from 3‐wk‐old WT, swp73a, swp73b treated or nontreated with gibberellic acid (GA) through the life cycle are given. (g) swp73b affects gibberellin metabolite levels. Error bars denote the SD, while asterisks denote statistically significant enrichment compared with WT plants (t‐test, the asterisk denotes P‐value < 0.05). (h) swp73b is hypersensitive to SA treatment. Growth ratio of swp73a and swp73b in response to 5 μM salicyclic acid (SA). Error bars denote the SD, while letters correspond to statistical significance determined by the Kruskal–Wallis ANOVA with post hoc Dunn’s test (P < 0.05). (i) SA amount in the WT, swp73a and swp73b mutants. Error bars denote the SD, while asterisks denote statistically significant enrichment compared with WT plants (t‐test; *, P‐value < 0.00005; **, P‐value < 0.000005).

Fig. 4

SWP73A and SWP73B play differential roles in salicylic acid (SA) biosynthesis in Arabidopsis. (a) Simplified scheme of SA biosynthesis pathway in Arabidopsis. Solid arrows indicate a single‐step enzymatic reaction leading to the formation of the subsequent metabolite. Dashed lines represent a multistep process not detailed in the scheme. Blunt‐ended arrows denote an inhibitory process. (b) Differential effect of swp73a and swp73b on the expression of ICS1, (c) ICS2 and (d) MES2 and MES9 genes. Data are shown as the mean ± SD (n = 3). Error bars denote the SD, while the asterisks show statistical significance P‐values by the student's t‐test (P < 0.05). (e) Scheme of the genes involved in SA biosynthesis with amplicon positions corresponding to (f–j). (f) SWP73B binds directly the ICS1, ICS2 and MES9 loci. (g) SWP73A directly binds MES9 gene body. Data are shown as the mean ± SD (n = 3). The asterisks represent P‐values by Student's t‐test (P < 0.05). Samples were collected at the end of the night. (h) SWI3B occupancy on the regulatory region of genes involved in SA biosynthesis is affected in the swp73b mutant. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). swp73b mutation causes alterations in histone modification presence on (i) ICS1 and (j) ICS2 loci. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). Samples for (f, h–j) were collected at midday. ICS, isochorismate synthase.

Fig. 5

swp73a and swp73b mutations lead to defects in gibberellin biosynthesis and catabolism. (a) Simplified scheme of gibberellin biosynthesis and catabolism pathway in Arabidopsis. Green colour represents bioactive form of gibberellin, while magenta indicates inactive form. (b) The swp73a and swp73b mutations differentially affect the expression of genes involved in gibberellin biosynthesis. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (c) Scheme of the genes involved in GA biosynthesis with amplicon positions corresponding to (d–f). (d) SWP73A and SWP73B bind to the regulatory region of genes involved in gibberellin biosynthesis. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (e) SWI3B binding to loci of genes involved in gibberellin biosynthesis is differentially affected in swp73a and swp73b plants. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (f) swp73a and swp73b lines exhibit differential effects on chromatin status on loci of genes involved in gibberellin biosynthesis. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test).

Fig. 6

BRM‐associated SWI/SNF (BAS)‐A complex regulates germination speed through the repression of GA3OX1 and GA3OX2 expression in Arabidopsis. (a) Seed germination of the wild‐type (WT) and the swp73a mutant and the SWP73A::SWP73A‐GFP swp73a line, after 2 d of stratification. Germination was scored at given time point. Error bars show the SD of four biological replicates. *, P < 0.05. (b) Expression of GA3OX1 is affected in swp73a. (c) Expression of GA3OX2 is affected in swp73a. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student’s t‐test). (d) Scheme of the genes involved in gibberellic acid (GA) biosynthesis with amplicon positions corresponding to (e–g, i, j). SWP73A occupancy measured by chromatin immunoprecipitation (ChIP)‐qPCR alters in dry seeds (e), and imbibed seeds after 24 h (f) and 48 h (g). Values were normalised to WT, and TA3 was used as a control. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (h) Confocal microscope analysis of SWP73A‐GFP signal indicates the presence of SWP73A protein in dry seeds and imbibed seeds after 24 h, whereas after 48 h SWP73A was not detected. Bar, 50 μm. (i) swp73a affects H3K9me2 and H3K4me3 occupancy on GA3OX1 and (j) GA3OX2. In swp73a, H3K4me3 was indicated by a green line and rectangles, while H3K9me2 was denoted by a dark red line and circles. Values were normalised to WT and to TA3. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test).

Fig. 7

swp73a and swp73b mutations have a different effect on metabolome profiles in Arabidopsis. (a) Quantity of primary metabolites in the swp73a and swp73b mutants collected at night's end. (b) Quantity of primary metabolites in the swp73a and swp73b mutants collected at the end of the day. Relative expression of RAFFINOSE SYNTHASE 5 (c) and RAFFINOSE SYNTHASE 6 (d) in swp73a and swp73b. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (e) Scheme of the RS5, RS6 and IGMT1 genes with amplicon positions corresponding to (f, i). (f) SWP73A and SWP73B bind to the regulatory region of RS5 and RS6 (n = 3, P‐value < 0.05, Student's t‐test). (g) 4‐Methoxy‐3‐indolylmethylglucosinolate amount in swp73a and swp73b collected at the end of the day. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student’s t‐test). (h) Relative expression of INDOLE GLUCOSINOLATE O‐METHYLTRANSFERASE 1 in swp73a and swp73b. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test). (i) SWP73A binds to the IGMT1 locus, as shown by chromatin immunoprecipitation (ChIP)‐qPCR analysis. Error bars denote the SD, while asterisks indicate statistical significance (n = 3, P‐value < 0.05, Student's t‐test).

Fig. 8

Hypothetical working model summarising regulatory functions of Arabidopsis switch‐defective/sucrose nonfermentable (SWI/SNF) complex subclasses containing SWP73A and SWP73B. SWP73A can form a subclass of BRM‐associated SWI/SNF (BAS)‐A BRM‐containing SWI/SNF complex and binds to gene body regions. SWP73A controls genes involved in auxin biosynthesis and response. Its loss of function leads to decreased auxin levels and affected lateral root formation upon auxin treatment. It represses the expression of raffinose and glucosinolate biosynthesis genes and negatively modulates germination dynamics. However, SWP73B has broader functions than SWP73A. SWP73B forms the BAS‐B subclass of BRM‐SWI/SNF chromatin remodelling complexes (CRCs), and it is indispensable in the formation of MINUSCULE‐associated SWI/SNF (MAS)‐ and SYD‐associated SWI/SNF (SAS)‐SWI/SNF complexes. SWP73B‐containing SWI/SNF subclasses mainly occupy promoter and 5'UTR regions. swp73b mutation leads to decreased gibberellin levels and overaccumulation of salicylic acid. SWP73B affects metabolism control, including raffinose and glucosinolates. Created in BioRender (https://BioRender.com/z26m589).