An Arabidopsis Retention and Splicing complex regulates root and embryo development through pre-mRNA splicing

Abstract

Pre-mRNA splicing is an important step in the posttranscriptional processing of transcripts and a key regulator of development. The heterotrimeric retention and splicing (RES) complex plays vital roles in the growth and development of yeast, zebrafish, and humans by mediating pre-mRNA splicing of multiple genes. However, whether the RES complex is conserved in plants and what specific functions it has remain unknown. In this study, we identified Arabidopsis (Arabidopsis thaliana) BUD13 (AtBUD13), GROWTH, DEVELOPMENT AND SPLICING 1 (GDS1), and DAWDLE (DDL) as the counterparts of the yeast RES complex subunits Bud site selection protein 13 (Bud13), U2 snRNP component Snu17 (Snu17), and Pre-mRNA leakage protein 1, respectively. Moreover, we showed that RES is an ancient complex evolutionarily conserved in eukaryotes. GDS1 directly interacts with both AtBUD13 and DDL in nuclear speckles. The BUD13 domain of AtBUD13 and the RNA recognition motif domain of GDS1 are necessary and sufficient for AtBUD13–GDS1 interaction. Mutants of AtBUD13, GDS1, and DDL failed to properly splice multiple genes involved in cell proliferation and showed defects in early embryogenesis and root development. In addition, we found that GDS1 and DDL interact, respectively, with the U2 small nuclear ribonucleoproteins auxiliary factor AtU2AF65B and the NineTeen Complex-related splicing factor SKIP, which are essential for early steps of spliceosome assembly and recognition of splice sites. Altogether, our work reveals that the Arabidopsis RES complex is important for root and early embryo development by modulating pre-mRNA splicing.

Introduction

Pre-mRNA splicing is a critical step in eukaryotic mRNA processing and gene expression regulation. RNA splicing is catalyzed by the spliceosome, a large RNA–protein complex composed of U1, U2, U4, U5, and U6 small nuclear ribonucleoproteins (snRNPs) and numerous additional non-snRNP protein factors (Shi, 2017). These non-snRNPs are important in spliceosome assembly and remodeling, definition of exons and introns, recognition and selection of polypyrimidine tracts (PPTs), branch point and 5′- and 3′-splice sites in introns, lariat formation and removal, and retention of unspliced RNAs in the nucleus (Will and Luhrmann, 2011; Nimeth et al., 2020). Non-snRNP-regulated splicing and alternative splicing are essential for plant growth, development, and response to the environmental stimuli (Wang et al., 2012; Sasaki et al., 2015; Lee et al., 2017; Laloum et al., 2018; Szakonyi and Duque, 2018; Li et al., 2019; Xiong et al., 2019a, 2019b).

The Retention and Splicing (RES) complex, a non-snRNP regulatory complex first identified in yeast (Saccharomyces cerevisiae), is critical for the splicing of numerous transcripts and nuclear retention of unspliced pre-mRNAs in the nucleus (Dziembowski et al., 2004). The RES complex consists of three subunits, namely U2 snRNP-associated protein 17 (Snu17)/IST3, bud site selection protein 13 (Bud13)/CWC26, and pre-mRNA leakage protein 1 (Pml1; Dziembowski et al., 2004; Yan et al., 2016). Both Bud13 and Snu17 bind directly to the pre-mRNA and are involved in splicing, whereas Pml1 is mainly linked to the retention of unspliced pre-mRNA (Gottschalk et al., 2001; Fernandez et al., 2018). In addition to splicing and nuclear retention of pre-mRNA, the RES complex is also required for the assembly and activation of the spliceosome (Gottschalk et al., 2001; Bao et al., 2017). Inactivation of individual RES components leads to reduced growth, increased thermo-sensitivity, and alteration in budding pattern in yeast (S. cerevisiae) (Zhou et al., 2013; Zhou and Johansson, 2017). This complex has also been identified in human and zebrafish (Danio rerio; Bessonov et al., 2010; Fernandez et al., 2018). In human, the three comprising subunits are RNA-binding motif protein X-linked 2 (RBMX2), hBUD13/MGC13125, and Smad nuclear-interacting protein 1 (SNIP1), which are the counterparts of yeast Snu17, Bud13, and Pml1, respectively (Zhang et al., 2018). Mutations in zebrafish RES complex genes lead to strong developmental defects and widespread cell death in the central nervous system during early embryogenesis (Fernandez et al., 2018).

Cooperation among the RES complex subunits increases the stability and the binding affinity of the complex to pre-mRNAs. Snu17 and its animal counterpart RBMX2 bind to the 3′-end sequence of the introns and act as a central platform, which independently binds the other two subunits, Bud13/hBUD13 and Pml1/SNIP1 (Yan et al., 2016; Zhang et al., 2018); no direct interaction between Bud13 and Pml1 has been detected (Trowitzsch et al., 2008, 2009; Brooks et al., 2009). The binding of Bud13 or Pml1 to Snu17 stabilizes the RNA recognition motif (RRM) fold of Snup17p, leading to rigidification of the Snu17 structure, and increasing the RNA-binding affinity of Snu17 during splicing (Wysoczanski et al., 2014, 2015; Wysoczanski and Zweckstetter, 2016). In addition, the assembly of RES complex is highly plastic due to Bud13 and Pml1 promoting the binding of each other to Snu17 (Wysoczanski et al., 2014, 2015; Wysoczanski and Zweckstetter, 2016).

During pre-mRNA splicing, the RES complex associates with U2 snRNP components and mediates the efficient transformation of the precatalytic B spliceosome into an activated B^act complex (Bessonov et al., 2008; Schneider et al., 2015; Bao et al., 2017). The mRNA splicing efficiency is greatly decreased in yeast strains lacking Snu17 or Bud13, whereas deletion of Pml1 leads to milder splicing defects (Gottschalk et al., 2001; Dziembowski et al., 2004; Bao et al., 2017). The RES complex is not essential for yeast viability although mutant strains for the three subunits show retarded growth under high temperature (Zhou et al., 2013; Zhou and Johansson, 2017). In contrast, this trimeric complex in zebrafish is preferentially expressed in the central nervous system during early embryo development and is essential for embryogenesis (Fernandez et al., 2018). Additionally, the homologs of Bud13 in mouse (Mus musculus) macrophages and humans are involved in withstanding virus infection and lipid metabolism, respectively (Jiang et al., 2001; Lin et al., 2016; Frankiw et al., 2019).

Although the conserved RES complex has been identified in yeast and animals, whether there is a functional RES complex and its roles in plant development remain unknown. In a previous study, we identified and characterized the Arabidopsis (Arabidopsis thaliana) counterpart of yeast Bud13, named AtBUD13. AtBUD13 is essential for early embryo development via mediating the splicing of smaller introns in a subset of genes (Xiong et al., 2019a). However, whether AtBUD13 acts as a part of plant RES complex has not been determined. The Arabidopsis forkhead-associated (FHA) domain-containing protein DAWDLE (DDL) has been identified as a homolog of human SNIP1. DDL is involved in miRNA processing and root development (Yu et al., 2008). The GROWTH, DEVELOPMENT AND SPLICING 1 (GDS1) is an important co-activator or helper protein in the early stage of the pre-mRNA splicing process and required for Arabidopsis growth and development (Kim et al., 2016). In this study, we show that Arabidopsis GDS1 and DDL are the functional counterparts of Snu17 and Plm1p, respectively, and form a RES complex with AtBUD13. GDS1 interacts directly with both AtBUD13 and DDL. The decreased expression of GDS1 and DDL resulted in defective pre-mRNA splicing comparable to those reported for atbud13 null mutants, reduced number of root meristematic cells, and abnormal cell division in early embryos, which consequently inhibited the roots growth and embryo development. We present evidence that there is a functional RES complex in plants with a critical role in the regulation of embryogenesis and root growth.

Results

Identification of the Arabidopsis homologs of Snu17 and Pml1

We have previously identified AtBUD13 as a homolog of Bud13, one of the three subunits of the yeast RES complex (Xiong et al., 2019a). To determine whether plants have a functional RES complex, we sought to identify the other RES subunits and investigate their functions. A high-throughput binary interactome based on a yeast two-hybrid (Y2H) assay showed that AtBUD13 interacts with GDS1 (encoded by At3G47120), a co-activator or helper protein in pre-mRNA splicing process (Arabidopsis Interactome Mapping Consortium, 2011; Kim et al., 2016). We performed Blast searches and found that GDS1 was the Arabidopsis protein with highest similarity to both yeast Snu17 and human RBMX2 (Supplemental Figure S1; Supplemental Table S1). Similar to Snu17 and RBMX2, GDS1 harbors a predicted RRM domain in its N-terminal region (Figure 1A). Although the overall amino acid sequence identity between GDS1 and Snu17 (19%) and RBMX2 (33%) is low, the RRM domain of GDS1 is 62% and 65% identical to those of Snup17p and RBMX2, respectively (Supplemental Figure S2A). As their yeast and animal counterparts, GDS1 contains the RRM-typical consensus sequence motifs RNP1 (77KGFAFLAY84) and RNP2 (38VYVGGI43) (Figure 1B; Supplemental Figure S2A; Gottschalk et al., 2001; Kielkopf et al., 2004; Tripsianes et al., 2014). Similar to Snu17 and RBMX22, GDS1 is a member of the RRM family of proteins with a long C-terminal part (Figure 1A; Supplemental Figure S2B; Gottschalk et al., 2001). In addition, two zinc-finger domains are predicted after the RRM domain of GDS1 but not in Snu17 and RBMX2 (Figure 1A). Additional comparative sequence analysis revealed that only GDS1 proteins from the land plants contain between one and three predicted zinc-finger domains (Supplemental Figure S2B).

To determine whether GDS1 is a functional homolog of yeast Snu17, GDS1 was expressed in yeast snu17Δ mutant to complement its thermosensitive growth phenotype. The entire open reading frame (ORF) of GDS1 was expressed under the control of proGAL1 promoter in the yeast snu17Δ mutant strain MJY548 (parental strain UMY2219). Consistent with a previous report (Zhou et al., 2013), growth retardation of snu17Δ cells was exacerbated at 37°C compared to that of wild-type. The expression of GDS1 was able to partially restore normal growth of the yeast snu17Δ strain at 37°C (Figure 1C), indicating that GDS1 is the functional counterpart of yeast Snu17.

Although the FHA domain protein DDL has been identified as an Arabidopsis homolog of human SNIP1 and is involved in small RNA biogenesis (Yu et al., 2008; Feng et al., 2016; Zhang et al., 2018), it is unclear whether DDL is involved in pre-mRNA splicing acting as part of a RES complex together with AtBUD13 and GDS1. Sequence analysis showed that, similar to yeast Pml1 and human SNIP1, the C-terminus of DDL contains an FHA domain that has been reported to bind phosphothreonine (Morris et al., 2006; Figure 1D). DDL shares 16% and 35% amino acid identity with Pml1 and SNIP1, respectively (Supplemental Figure S3). To test if DDL is able to functionally replace Pml1 in yeast, we expressed the full-length ORF of DDL under the control of the proGAL1 promoter in a yeast PML1 deletion mutant strain (pml1Δ). The pml1Δ cells grow well at 25°C but their growth is restricted at 37°C (Zhou et al., 2013). We found that the expression of DDL restored normal growth of pml1Δ at 37°C (Figure 1E), indicating that DDL is a functional homolog of yeast Pml1.

Interactions among AtBUD13, GDS1, and DDL

To determine whether AtBUD13, GDS1, and DDL interact, we first performed a GAL4-based Y2H assay. As shown in Figure 2A, the yeast cells expressing GDS1 with either AtBUD13 or DDL, but not the cells expressing both AtBUD13 and DDL, were able to grow on medium selected for interaction.

We further tested whether the three proteins could be part of a plant RES complex using bimolecular fluorescence complementation (BiFC) assays. We fused GDS1, AtBUD13, and DDL with the N- and/or C-terminal fragments of the enhanced yellow fluorescence protein (EYFP) (nEYFP and cEYFP, respectively) and transiently co-expressed them in pairs in Nicotiana benthamiana leaves and Arabidopsis leaf mesophyll protoplasts. Reconstituted YFP signal was detected in nuclei when either GDS1-nEYFP and AtBUD13-cEYFP, or GDS1-nEYFP and DDL-cEYFP were expressed together (Figure 2, B and C; Supplemental Figure S4). Moreover, the resulting EYFP signal in both cases colocalized with U1-70K-mCherry, a spliceosome marker localized to nuclear speckles (Wang et al., 2012). Consistent with the Y2H results, the co-expression of AtBUD13-nEYFP and DDL-cEYFP did not produce EYFP signal (Figure 2B; Supplemental Figure S4).

Next, we tested whether GDS1 directly interacts with both AtBUD13 and DDL by protein pull-down assays. Plasmid pGEX4T-1-GDS1-GST was transformed together with pET30a-AtBUD13-His, pET30a-DDL-His, or pET30a (negative control) in Escherichiacoli cells as described by Brooks et al. (2009). The co-expressed proteins were purified using nickel–nitrilotriacetic acid (Ni–NTA) beads and analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and immunoblotting with monoclonal anti-GST and anti-His. The blotting result confirmed that GDS1 directly bind to AtBUD13 and DDL (Figure 2D). Finally, we further confirmed protein interactions by co-immunoprecipitation (Co-IP) in Arabidopsis. The expression cassettes 35S::HA-GDS1 and either Super::AtBUD13-Myc or Super::DDL-Myc were co-transformed into Arabidopsis protoplasts. Expression of empty 35S::HA and Super::Myc vectors were used as negative controls. Using anti-HA antibodies, we were able to co-immunoprecipitate HA-GDS1 with either AtBUD13-Myc or DDL-Myc (Figure 2E; Supplemental Figures S5 and S6). Taken together, our results from Y2H, BiFC, pull-down and Co-IP assays are consistent with AtBUD13, GDS1, and DDL being part of a complex, with GDS1 directly interacting with AtBUD13 and DDL.

We next investigate whether AtBUD13 can interact with other RRM-containing proteins besides GDS1. We selected AtU2AF65B for our analysis since it is the second most similar Arabidopsis protein to human RBMX2 after GDS1 (Supplemental Figure S1; Supplemental Table S1). However, we detected no interaction between AtBUD13 and AtU2AF65B in a BiFC assay (Supplemental Figure S4), suggesting that GDS1 is the functional homolog of RBMX2 in Arabidopsis.

To explore the occurrence of the RES complex in eukaryotes, we performed a BLAST analysis to identify putative orthologs of the three subunits in genomes across a diverse group of eukaryotes. As shown in Supplemental Figure S7, the RES subunits are present nearly in all sampled eukaryotes, suggesting that the RES complex is ancient and widespread across eukaryotes. However, during evolution, the complex appears to entirely or partially disappear in various lineages, such as in Excavata, Rhizaria, and Amorphea.

Domains and key amino acid residues involved in protein–protein interactions within the RES complex

We next asked which domains of GDS1, AtBUD13, and DDL are responsible for their interactions. We used truncated protein fragments of AtBUD13 and GDS1 to perform Y2H assay and found that the BUD13 domain of AtBUD13 and the RRM motif of GDS1 were necessary and sufficient for the interaction between the two proteins (Figure 3A). We also expressed the BUD13 domain of AtBUD13 and the RRM motif of GDS1 in a BiFC assay and confirmed that they are sufficient to reconstitute the EYFP signal in nuclear speckles (Figure 3B; Supplemental Figure S8).

Previous studies using isothermal titration calorimetry analysis identified several conserved residues in the UHM ligand motif (ULM) of BUD13 domain of yeast Bud13, F₂₂₄, R₂₃₁, W₂₃₂, V₂₃₅, and R₂₃₇, with a putative role in mediating interaction with Snu17 (Trowitzsch et al., 2008; Wysoczanski et al., 2014). Sequence alignment analysis showed that the corresponding residues in AtBUD13 are Y₅₁₈, H₅₂₅, W₅₂₆, V₅₂₉, and R₅₃₁ within the BUD13 domain (Figure 3C). To determine whether these five residues are necessary for AtBUD13–GDS1 interaction, we performed alanine substitutions in each of the five positions to generate the AtBUD13 Y_518A, H_525A, W_526A, V_529A, and R_531A mutant proteins. In both Y2H and BiFC assays, we found that only the W_526A mutation abolished the interaction of AtBUD13 with GDS1 (Figure 3, D–E; Supplemental Figure S9).

To map the critical regions in DDL for interaction with GDS1, we generated N- and the C-terminal DDL fragments (nDDL and cDDL, respectively) and tested their ability to interact with GDS1 by a BiFC assay. The nDDL was able to interact with GDS1 in nuclear speckles, whereas the cDDL peptide, which harbors the complete FHA domain, was not (Supplemental Figure S10). This result indicates that the N-terminal region of DDL is required for the DDL–GDS1 protein interaction.

GDS1 and DDL are broadly expressed in Arabidopsis

Based on publicly available gene expression data, AtBUD13, GDS1, and DDL show similar expression patterns during Arabidopsis development (Supplemental Figure S11). Based on our own RT-qPCR analysis, GDS1 and DDL were expressed in all tested organs, with preferential expression in flowers, seedlings, siliques, and seeds (Figure 4, A and B). To better analyze the expression patterns of GDS1 and DDL, we generated transgenic plants expressing the promoter regions of GDS1 (1,945 bp) and DDL (1,300 bp) fused to the β-glucuronidase (GUS) reporter. Both GDS1 and DDL were ubiquitously expressed in tissues of seedlings, roots, leaves, flowers, ovules, and seeds (Figure 4, C and D; Supplemental Figure S12).

AtBUD13 is highly expressed in embryos, consistent with its crucial role in embryo development (Xiong et al., 2019a). Likewise, by analyzing the GUS reporter lines, we detected strong GDS1 and DDL expression during embryo development, including preglobular, globular, heart, and mature embryo stages (Figure 4, E and F). Together with the previous reports by Morris et al. (2006) and Kim et al. (2016), these results show that, similar to AtBUD13 (Xiong et al., 2019a), GDS1 and DDL are ubiquitously expressed in Arabidopsis.

Arabidopsis RES complex functions in root and embryo development by controlling cell proliferation

We previously found that AtBUD13 is essential for the proper cell division in zygotes and early embryos and for root growth (Xiong et al., 2019a). Here, we investigated the AtBUD13 amiRNA plants and found a significant reduction in the number of the root meristematic cells in the amiRNA lines (Supplemental Figure S13), indicating that AtBUD13 affects root growth by controlling cell proliferation in the root apical meristem.

We then investigated whether GDS1 and DDL1 also play roles in embryo and root development. Because the transcript-null mutant of GDS1 is severely dwarf (Kim et al., 2016), we obtained two independent GDS1 transcript-knockdown mutants, gds1-2 (SAIL_56_E10) and gds1-3 (SK33362), in which the T-DNA fragments were inserted in the 5′ UTR region (Supplemental Figure S14, A–C). Although the overall growth and development of gds1-2 and gds1-3 plants seemed normal, the roots of these gds1 seedlings were shorter than those of wild-type (Figure 5, A and B; Supplemental Figure S14D and E). As observed for the AtBUD13 amiRNA lines, the number of cells in the root apical meristems of gds1-2 and gds1-3 seedlings were significantly decreased compared to wild-type (Figure 5, C and D). In addition, we detected aborted seeds within the gds1-2 and gds1-3 developing siliques (Figure 5E) and the number of seeds in mature gds1-2 silique (56.2 ± 2.3 seeds per silique) was significantly reduced compared to wild-type (60 ± 2.1 seeds per silique; Figure 5, F and G). To confirm that the observed phenotypic defects were due to altered GDS1 expression, we expressed a construct consisting of 3,378 bp including the GDS1 promoter region (1,945-bp upstream of the start codon) and the GDS1 coding region fused to GFP (GDS1g-GFP) in gds1-2 plants. The root length, cell number of root apical meristems and seed yield in independent gds1-2 GDS1g-GFP transgenic lines were restored to wild-type levels (Figure 5, A–G), indicating that reduced expression of GDS1 is responsible for defective root development and seed abortion in the gds1-2 mutant.

To test the underlying cause for seed abortion, we analyzed embryo development in gds1-2 developing siliques. We found that some gds1-2 embryos showed delayed or arrested development at octant, early-globular, and heart-shaped embryo stages. In addition, we detected abnormal cell proliferation in both suspensor and embryo proper at early stages of embryogenesis (Figure 5H). At 6-day after pollination (DAP), about 3.2% of the developing gds1-2 seeds contain arrested/aborted embryos, consistent with early seed abortion. These results support the notion that both GDS1 and AtBUD13 control cell proliferation and are required for root and embryo development.

DDL has been reported to participate in plant development and fertility, based on the observation of the transcript-null mutants (ddl-1 and ddl-2) with seriously defective roots and reduced fertility, and the transcript-knockdown lines (ddl-6 and ddl-7) with smaller plants (Morris et al., 2006; Feng et al., 2016). We found that ddl-6 and ddl-7 developed shorter roots than wild-type seedlings (Supplemental Figure S15, A and B), consistent with a previous report on the ddl-1 and ddl-2 mutants (Morris et al., 2006). We then tested whether the decreased root length and fertility in these ddl mutants is connected to cell proliferation. The number of cells in root apical meristems of these four mutants was significantly decreased (Supplemental Figure S15, C and D). The two transcript-null ddl-1 and ddl-2 mutants showed irregular and disorganized cell arrangement patterns in the root apical meristems of seedlings (Supplemental Figure S15D) and 86% (±4.6) seed abortion rate in developing siliques (Supplemental Figure S15E). We detected delayed or arrested embryos at globular, octant, and even earlier stages (3–4 DAP; Supplemental Figure S15F). By 6 DAP, a large proportion (85.5%) of developing seeds contain aborted embryos. Taken together, these results showed that, just like AtBUD13 and GDS1, DDL is required for root and embryo development by controlling cell division.

In order to confirm the interactions of GDS1 with AtBUD13 and DDL in vivo in Arabidopsis, we generated double mutants for AtBUD13 and GDS1, and for GDS1 and DDL. The gds1 atbud13 and gds1 ddl double mutants did not show additional or more pronounced developmental defects compared to the corresponding single mutants (Supplemental Figures S16 and S17). These genetic results were also consistent with the presence of a plant RES complex formed by GDS1, AtBUD13, and DDL.

To further support that AtBUD13, GDS1, and DDL work together in cell division and proliferation, we determined the expression levels of several key genes controlling cell division including WUSCHEL-RELATED HOMEBOX 2 (WOX2), AtMERISTEM LAYER 1 (ATML1), PROTODERMAL FACTOR2 (PDF2) and SCARECROW (SCR) in the atbud13^−/+, gds1, and ddl mutant siliques according to our previous RNA-Seq analysis (Xiong et al., 2019a). Our RT-qPCR result showed that the transcripts of ATML1, WOX2, PDF2, and SCR were all significantly decreased in these mutants compared to the wild-type Arabidopsis (Supplemental Figure S18).

Arabidopsis RES complex is involved in pre-mRNA splicing of a subset of genes modulating embryo and root development

To investigate whether GDS1 and DDL act together with AtBUD13 to mediate pre-mRNA splicing of the same group of genes, we performed RT-PCR using cDNAs from the gds1 and ddl mutants and tested intron retention of genes that are known to be misspliced in atbud13 mutant plants (Xiong et al., 2019a; Figure 6A). We found similar abnormal splicing patterns in atbud13, gds1, and ddl mutants compared to the wild-type controls and complemented lines (atbud13-1/+ AtBUD13g-GFP and gds1-2 GDS1g-GFP) (Figure 6B). These results further support that AtBUD13, GDS1, and DDL are part of the same complex and act together in pre-mRNA splicing.

Our previous RNA-seq analysis showed that mutation in AtBUD13 lead to the defective splicing of 52 genes connected to early embryo development, causing embryo lethality of null atbud13 mutants (Xiong et al., 2019a). We, therefore, analyzed the transcript splicing patterns of ATML1 and ARGONATE10/ZWILLE (AGO10/ZLL), two genes necessary for apical meristem development in early embryos (Iida et al., 2019; Han et al., 2020). In young developing siliques (1–3 DAP) of atbud13-1/+, gds1-2, and ddl-1 mutants, we detected increased levels of intron retention in intron 3 of ATML1 and intron 6 of AGO10/ZLL transcripts compared to wild-type (Figure 6, C and D). The splicing efficiencies of ATML1 intron 3 and AGO10/ZLL intron 6 were quantified by RT-qPCR. The result showed that the splicing efficiencies of these two introns were all significantly decreased in atbud13-1/+, gds1-2, and ddl-1 mutants (Figure 6, E and F), consistent with the increased retention of introns.

We also previously showed aberrant splicing of 36 genes involved in root development in the atbud13 transcript-null mutant (Xiong et al., 2019a; Supplemental Table S2). Among these genes, we selected AGO1 and SHORT ROOT IN SALT MEDIUM 1 (RSA1)/EMBRYO DEFECTIVE 1579 (EMB1579; Trolet et al., 2019; Ré et al., 2020; Zhang et al., 2020) for further analysis by RT-qPCR. We found that the splicing efficiencies of AGO1 intron 8 and RSA1/EMB1579 intron 2 were all significantly decreased in the seedlings of atbud13 (AtBUD13-amiRNA), gds1, and ddl mutants (Figure 7A).

Control of cell division in the early embryo and root apical meristem is critical for embryo and root development. Our previous RNA-seq data showed that mutations in AtBUD13 resulted in abnormal splicing of 20 genes with critical roles in cell cycle and proliferation (Supplemental Table S3; Xiong et al., 2019a), such as cyclins, cyclin-like proteins, and their partners (e.g. MITOTIC-LIKE CYCLIN 3B FROM ARABIDOPSIS [CYC3B], Cyclin/Brf1-like TBP-binding protein [BRF1], CYCLIN D7, CYCLIN T [CYCT] 1;2, CYCT1;3 and CYCT 1;5), and components of anaphase-promoting complex (Cui et al., 2007; Rojas et al., 2009; Guo et al., 2016). We found that the splicing efficiencies of BRF1 intron 11 was markedly reduced in the atbud13, gds1, and ddl mutant seedlings whereas the splicing of CYCT1;5 intron 5 was impaired in atbud13 and ddl seedlings (Figure 7B). These results further support a role of AtBUD13, GDS1, and DDL in cell proliferation through pre-mRNA splicing of genes involved in cell cycle regulation. Taken together, all these findings reveal that AtBUD13, GDS1, and DDL act as a complex to modulate pre-mRNA splicing of genes required for cell proliferation and plant development, especially, in the early embryo and root apical meristems.

Arabidopsis RES complex binding with splicing factors SKIP and AtU2AF65B

To explore further the function and regulation of RES complex in plant pre-mRNA splicing, we searched for proteins interacting with AtBUD13, GDS1, and DDL. Structural studies of yeast spliceosome showed that the formation of RES complex is regulated and stabilized by the NineTeen complex (NTC)-related component Prp45, a protein required for the early stages of spliceosome assembly (Yan et al., 2016; Hálová et al., 2017). Residues 193–350 of Prp45 wrap around Pml1 (DDL homolog) and Bud13 (Yan et al., 2016), but direct interaction between Prp45 and any of the RES complex components has not been reported. Arabidopsis SNW/Ski-interacting protein domain protein (SKIP) is a homolog of Prp45 that regulates flowering time and environmental fitness by modulating recognition or cleavage of 3′- and 5′-splice sites (Wang et al., 2012; Cui et al., 2017; Li et al., 2019). To investigate whether the RES complex subunits interact with SKIP, we performed Y2H and BiFC assays. We found that DDL, but not AtBUD13 or GDS1, interacts with SKIP (Figure 8, A and B; Supplemental Figure S19A). Furthermore, the reconstituted EYFP signal indicative of a positive interaction/proximity between DDL and SKIP in our BiFC assay was colocalized with U1-70K in nuclear speckles (Figure 8B; Supplemental Figure S19A). These results suggest that RES complex functions might be facilitated by the physical contact between DDL and SKIP.

In yeast, Bud13 binds Snu17 (GDS1 homolog), facilitating the formation of an extended β-sheet in Snu17 that is involved in specific recognition of the 3ʹ-end intron sequences (Yan et al., 2016; Zhang et al., 2018). U2AF65, the large subunit of U2 snRNP auxiliary factor (U2AF), recognizes the 3′ splice sites and recruits U2 snRNP to introns during the assembly of the spliceosomal complex A (Guth et al., 1999; Chusainow et al., 2005). AtU2AF65B, a homolog of mammalian U2AF65 and yeast Mud2, is required for transcript splicing of a subset of genes in Arabidopsis (Jang et al., 2014; Park et al., 2017, 2019; Xiong et al., 2019b). We hypothesized that the Arabidopsis RES complex might interact with AtU2AF65B through GDS1. As we expected, both Y2H and BiFC assays showed that GDS1 interacts with AtU2AF65B in nuclear speckles (Figure 8, C and D; Supplemental Figure S19B), whereas no interaction was detected between either AtBUD13 or DDL and AtU2AF65B. Collectively, these results indicate that Arabidopsis RES complex might function in the early stages of spliceosome assembly and recognition of the 3ʹ-end intron sequences by interaction with SKIP and U2 auxiliary splicing factor.

Discussion

In this study, we have identified AtBUD13, GDS1, and DDL as components of the RES complex in plants, with essential functions in plant development. The RES complex identified in yeast, zebrafish, and humans plays roles in pre-mRNA splicing by facilitating formation and activation of the spliceosome (Deckert et al., 2006; Bessonov et al., 2008; Wysoczanski et al., 2014; Bao et al., 2017; Fernandez et al., 2018). In vivo splicing assay revealed that the yeast RES complex enhances splicing, especially of those introns that contain poor consensus sequence at the splice site (Dziembowski et al., 2004; Scherrer and Spingola, 2006). Plants provide an excellent model system for studying function of the RES complex in pre-mRNA splicing since plant genes contain far more introns than those in yeast. Together with our previous analysis of AtBUD13 function (Xiong et al., 2019a), our findings suggest that the Arabidopsis RES complex is critical for embryogenesis and root development by mediating pre-mRNA splicing of a subset of genes with roles in cell cycle and proliferation, and root apical meristem establishment and early embryo development (Supplemental Figure S20).

Within the yeast RES complex, Snu17 acts as a central component that organizes and binds the other two subunits (Trowitzsch et al., 2008; Brooks et al., 2009; Tripsianes et al., 2014). Interestingly, Snu17 contains the RRM-typical consensus sequence motifs RNP1 and RNP2 that mediate RNA interactions in canonical RRM domains but lacks the Arg-X-Phe motif characteristic of the U2AF homology motif (UHM) that mediates protein–protein interactions (Tripsianes et al., 2014). We found that GDS1, the homolog of yeast Snu17, directly binds AtBUD13 and DDL. GDS1 contains a typical RRM at the N-terminal with highly conserved RNP1 and RNP2 motifs, two Zinc-finger domains, and conserved cysteine and histidine residues required to target RNAs for pre-mRNA splicing (Kim et al., 2016). We also found that a conserved tryptophan residue (W526) in the BUD13 domain of AtBUD13 is required for the interaction with the RRM domain of GDS1. The corresponding W232 residue in yeast Bud13 is important for RRM binding and stability of the Bud13-Snu17 dimer; although it is unknown whether a single W232A substitution abolish interaction of Bud13 with Snu17 (Brooks et al., 2009; Tripsianes et al., 2014; Wysoczański et al., 2015). Collectively, our results demonstrate that the GDS1 RRM domain and the conserved tryptophan residue (W526) in the BUD13 domain of AtBUD13 are essential for the AtBUD13–GDS1 interaction.

Arabidopsis RES complex is involved in pre-mRNA splicing via binding to distinct splicing factors. GDS1 is known to interact with splicing factors U1-70K, SR1, and SRm102 (Kim et al., 2016). Arabidopsis U1-70K is a U1 snRNP, and SR1 and SRm102 belong to SR (serine/arginine)-related proteins (Koncz et al., 2012). U1 snRNP is involved in the recognition of 5′-splice sites through base pairing of the U1 snRNA to complementary sequences of pre-mRNA (Cho et al., 2011). Binding of U1 snRNP with the 5′-splice site initiates the assembly of the spliceosome (Stark et al., 2001). Human SRm160, a homolog of SRm102, is involved in U2 snRNP binding for intron definition in pre-mRNAs (Wagner et al., 2003). Interestingly, in this study, we identified a GDS1-interacting protein, AtU2AF65B (Xiong et al., 2019b). AtU2AF65B, also an RRM-motif containing protein, is a homolog of yeast Mud2 and human U2AF65 which binds to the PPT, facilitates U2 snRNP recruitment, and stabilizes the interaction of U2 snRNP with the branchpoint during assembly of spliceosomal complex A (Guth et al., 1999; Chusainow et al., 2005). Additionally, a few other putative GDS1 interactors have been detected via Y2H screening (Kim et al., 2016). These different interacting partners of GDS1 might facilitate the participation of the Arabidopsis RES complex in the assembly and activation of the spliceosome as well as pre-mRNA splicing, consistent with the roles of the yeast RES complex (Dziembowski et al., 2004; Bao et al., 2017). Moreover, we also found that DDL interacts with SKIP. SKIP is a component of the NTC and modulates recognition of 3′- and 5′-splice sites in Arabidopsis (Wang et al., 2012). SKIP and AtU2AF65B play important roles in plant development and responses to environmental stimuli by controlling pre-mRNA splicing of multiple genes (Cui et al., 2017; Li et al., 2019; Park et al., 2019; Xiong et al., 2019a, 2019b). Collectively, our study together with previously published data show that the Arabidopsis RES complex has a variety of functions in pre-mRNA splicing with the GDS1 RRM domain binding pre-mRNAs as well as interacting with other splicing factors at early stages of spliceosome assembly.

We also identified multiple target genes of the RES complex in Arabidopsis. Our previous RNA-Seq analysis showed that AtBUD13 is required for the proper splicing of 52 genes including AGO10/ZLL and ATML1 involved in embryo development (Xiong et al., 2019a). In addition to these 52 genes, our RNA-Seq analysis together with RT-qPCR validation also revealed that AtBUD13, GDS1 and DDL participate in pre-mRNA splicing of 36 genes involved in root development and 20 genes required for cell cycle and proliferation. These genes include RETARDED ROOT GROWTH (RRG), ELONGATA 1 (ELO1), AGO1, CYC3B, BRF1, CYCLIN D7, CYCT1;2, CYCT1;3, CYCT1;5, APC11, and CDC27a. RRG encodes a mitochondria-localized protein required for cell proliferation in the root meristem (Zhou et al., 2011). ELO1 is a subunit of the elongator complex that promotes root cell proliferation (Qi et al., 2019). AGO1 is a critical RNA slicer in microRNAs and siRNAs processing required for root growth and development, and cell division. AGO1 binds to a set of 19-nucleotide, tRNA-derived fragments during the cell cycle progression, and therefore promotes cell division and root meristem activity (Trolet et al., 2019). Moreover, AGO1 enhances plant tolerance to UV-B by reducing DNA damage and programmed cell death in the root meristematic cells (Ré et al., 2020). CYC3B, BRF1, CYCLIN D7, CYCT1;2, CYCT1;3, and CYCT1;5 are cyclins, cyclin-like proteins, and their partners, which are core cell cycle proteins regulating cell division in roots (Ferreira et al., 1994; Cui et al., 2007; Zhang et al., 2019, 2020). APC11/ZGY1 activates the first division of the zygote by interacting and degrading cyclin B1, which is critical for cell cycle synchronization in Arabidopsis (Guo et al., 2016, 2018). CDC27a is a subunit of the anaphase-promoting complex which is an essential ubiquitin ligase that targets cell cycle proteins for proteasome-mediated degradation. Arabidopsis CDC27a regulates mitotic progression and cell differentiation during the entire life cycle (Pérez-Pérez et al., 2008). Overexpression of CDC27a is associated to apical meristem restructuration and altered cell cycle marker expression in Arabidopsis (Rojas et al., 2009). Collectively, all these findings support that AtBUD13, GDS1, and DDL in the same complex regulates plant development via modulating pre-mRNA splicing of genes required for cell proliferation, and early embryo and root development.

Consistent with the role of the RES complex in pre-mRNAs processing of genes involved in root and embryo development, and cell cycle regulation, the bud13, gds1, and ddl mutants showed lower number of cells in root apical meristems, impaired root growth, and abnormal patterns of cell division during early embryo development.

Conclusions

Our results revealed that AtBUD13, GDS1, and DDL form the RES complex and functions in pre-mRNA splicing in Arabidopsis. The RRM domain of GDS1 and the BUD13 domain of AtBUD13 are necessary and sufficient for AtBUD13–GDS1 interaction. Decreased expression of GDS1, AtBUD13, and DDL resulted in reduced root meristematic cell number and root growth, abnormal patterns of cell division, and defective embryonic development with aberrant splicing of multiple genes involved in regulation of cell cycle and development. The direct interactions of GDS1 with the splicing factor AtU2AF65, and DDL with SKIP suggest that the Arabidopsis RES complex might function in recognition of splice sites and spliceosome assembly at earliest steps of pre-mRNA processing. Taken together, this work highlights the importance of the plant RES complex in root and early embryo development by modulating pre-mRNA splicing.

Materials and methods

Plant materials and growth conditions

All plant materials used in this study were in the Columbia-0 (Col-0) ecotype and Wassilewskija (WS) backgrounds of Arabidopsis (A.thaliana). Seeds of gds1-2 (SAIL_56_E10) and gds1-3 (SK33362), ddl-6 (SALK_045025), and ddl-7 (SAIL_1281_F08) mutants were obtained from the SALK T-DNA mutant collection (Arabidopsis Biological Resource Center). The characters of atbud13-1, atbud13-2, ddl-1, and ddl-2 (WS), and ddl-6 and ddl-7 (Col-0) have been described previously (Walker et al., 2008; Feng et al., 2016; Xiong et al., 2019a). The genotypes were analyzed by PCR with gene-specific primers and T-DNA-specific primers (Supplemental Table S4). Seeds were sterilized and placed on 0.5× MS medium. Seedlings and plants were grown under long day (16-h/8-h light-dark [LD] photocycles) at 20°C–22°C.

Generation of transgenic plants

A 1,945-bp and 1,300-bp predicted promoter sequence upstream from the start codon of GDS1 and DDL were amplified with specific primer (Supplemental Table S4), respectively. The PCR products were cloned into Gateway Entry vector (pCR8/GW/TOPO TA Cloning Kit, Invitrogen, Waltham, MA, USA), and recombined into the pMDC163 vector by LR reaction to generate proGDS1::GUS and proDDL::GUS constructs. The 3,402-bp genomic sequence of GDS1 including the 1,945-bp promoter was amplified and subcloned into the TOPO Gateway Entry vector to generate pENTR-proGDS1::GDS1. Then, it was recombined into the pMDC163 vector by LR reaction to create proGDS1::GDS1-GFP construct.

The proGDS1::GUS and proDDL::GUS constructs were introduced into Col-0 plants and proGDS1::GDS1g-GFP construct was transformed into the gds1-2 plants as described before (Clough and Bent, 1998). More than three T3 transgenic lines were used for analysis.

Analysis of cell number and embryo development

For scoring the number of root meristematic cells, seedlings were grown on 0.5 MS× medium for 5 days and stained with propidium iodide (PI). The PI fluorescence was examined by LSM880 laser scanning confocal microscope (Carl Zeiss, Oberkochen, Germany) with the excitation and emission wavelengths of 532 and 588 nm, respectively.

For the analysis of embryo development, siliques (1–7 DAP) from wild-type and mutant plants were incubated in Hoyer’s solution as described by Xiong et al. (2019a, 2019b), and imaged by differential interference contrast.

Complementation analysis in yeast

The CDSs of GDS1 and DDL were subcloned into the PYES2 vector with restriction sites for Kpn I/Xho I and Hind III/BamH I under the control of the yeast GAL1 promoter to generate GAL1:GDS1 and GAL1:DDL constructs, respectively. The reconstructs were transformed into snu17⊿ (MJY548) and pml1⊿ (MJY535) mutant yeast cells, which are thermosensitive compared to wild-type cells (UMY2219; Zhou et al., 2013). The transformed strains are grown at 25°C or 37°C on the appropriate selective media (containing 20 g L⁻¹ X-α-D-galactoside, 10 g L⁻¹ yeast extract, and 20 g L⁻¹ peptone) for 3 days.

Y2H assays

The CDSs with or without mutation and different truncated fragments of proteins were cloned into the TOPO vector and then recombined into either the GAL4 activation domain plasmid (pDEST22) or the GAL4 binding domain plasmid (pDEST32) using the LR reaction. Both plasmids for the detection of protein interaction were then cotransformed into yeast strain Gold. SKIP-AD and U1-70K-BD were used as the positive control (Wang et al., 2012). Protein interactions were determined by measuring the growth of transformants after 3-day growth on medium lacking Leu and Trp, or lacking Ade, His, Leu, and Trp (Clontech, Mountain View, CA, USA) containing 125 ng/mL amino-1,2,4-triazole and 32 ug/mL X-α-gal.

Pull-down assay

The pull-down assay was performed by coexpressing two proteins in E. coli cells (Brooks et al., 2009). GDS1 CDS was cloned into pGEX4T-1 to express GDS1-GST. CDSs of AtBUD13 and DDL were cloned into pET30a vector to express AtBUD13-His and DDL-His. The pGEX4T-1-GDS1 vector was co-transformed with either pET30a-AtBUD13, pET30a-DDL or pET30a empty vector (negative control) into E.coli BL21 (DE3) competent cells for protein co-expression. Co-expressed proteins in E.coli cells were purified using Ni–NTA agarose beads for 2 h at 4°C in 50-mM PBS buffer according to the manufacturer’s instructions (Beyotime, Jiangsu, China). Purified proteins were collected and subjected to immunoblotting analysis with monoclonal anti-GST and anti-His antibodies (1:1,000).

BiFC assay

We performed BiFC assay of protein–protein interactions among GDS1, AtBUD13, and DDL both in epidermal cells of N.benthamiana leaves and Arabidopsis protoplasts from leaf mesophyll. The full-length CDSs of GDS1/AtBUD13/DDL with or without mutations and truncated fragments were cloned into Gateway Entry vector, and then transferred into pSITE-nEYFP-C1 vector (ABRC CD3-1648) or pSITE-cEYFP-C1 vector (ABRC CD3-1649) by LR reaction.

BiFC assays in epidermal cells of N. benthamiana leaves were performed as previously described (Xiong et al., 2019b). One of recombined nEYFP constructs and one of recombined cEYFP constructs were co-transformed into GV3101 cells, and then were co-infiltrated into N. benthamiana leaves to transient expression proteins. U1-70K-mCherry located in nuclear speckles was used as indicator for distribution of splicing factor (Wang et al., 2012). The YFP and mCherry signals were examined by LSM880 laser scanning confocal microscope (LSCM; Carl Zeiss) with the excitation and emission wavelengths of 514 nm and 530–590 nm, and 561 nm and 580–620 nm (488 nm laser power: 2.4%; Master gain: 488–548 nm; Digital gain: 1.0) (561 nm laser power: 9.1%–11.1%; Master gain: 712–745 nm; Digital gain: 1.0), respectively. Pearson correlation coefficients (R) were calculated to estimate YFP and mCherry colocalization.

Interactions between GDS1 and AtBUD13/DDL were investigated in leaf mesophyll protoplasts of Arabidopsis (Jiang et al., 2019). Col-0 plants were grown under a LD (16-h/8-h) photoperiod at 20°C–22°C. Leaf mesophyll protoplasts were prepared from leaves of 3- to 4-week-old plants as described previously (Wang et al., 2009). Recombined nEYFP and cEYFP constructs were transiently transformed into leaf mesophyll protoplasts. Then these protoplasts were incubated in the dark at 22°C for 18 h before fluorescent examination using LSCM. A BP 530–590 nm filter was used to detect YFP fluorescence and an LP650 nm filter was used to detect chlorophyll fluorescence (488-nm laser power: 26.6%; Master gain: 788; Digital gain: 1.0) (561nm laser power: 20%; Master gain: 626; Digital gain: 1.0).

Co-IP assay

Co-IP assays of GDS1 with both AtBUD13 and DDL in Arabidopsis were performed as described by Liu et al. (2021). The ORFs of AtBUD13 and DDL were amplified and cloned into pSuper1300 vectors by homologous recombination to generate Super::AtBUD13-Myc and Super::DDL-Myc constructs, respectively. 35S::HA-GDS1 construct was generated using a similar method. Plasmids of 35S::HA-GDS1 and Super::AtBUD13-Myc/Super::DDL-Myc were co-transformed into Arabidopsis leaf protoplasts and expressed for 18–24 h. Empty 35S::HA and Super::Myc vectors were transformed as negative controls. The total proteins (input) extracted from the co-expressed Arabidopsis protoplasts were immunoprecipitated with anti-HA agarose (Sigma-Aldrich, St Louis, MO, USA; A2095). The Co-IP products were subjected to immunoblot analysis. Anti-HA and anti-Myc antibodies were used to detect AtBUD13-Myc/DDL-Myc and HA-GDS1, respectively.

GUS staining

Before GUS staining, samples were fixed using 90% acetone for 20 min and were immersed in the staining solution (1-mM 5-bromo-4-chloro 3-indolyl glucuronic acid, 100-mM sodium phosphate pH 7.0, 0.1-mM EDTA, 0.5-mM ferricyanide, 0.5-mM ferrocyanide, and 0.1% Triton X-100), and then were incubated in staining solution at 37°C after vacuum infiltration for 10 min on ice. Chlorophyll in stained specimens was cleared by 30 min incubation through a series increasing concentration of ethanol solutions. Cleared specimens were imaged using an Olympus BX51 microscope.

RT-PCR and RT-qPCR

Total RNA was extracted using the TRIzol isolation reagent (Qiagen, Hilden, Germany). After RNase-free DNase I treatment, 1 µg of total RNA was used for the first-strand cDNA synthesis (PrimeScript RT reagent Kit with gDNA Eraser; TaKaRa, Shiga, Japan). GAPDH and TUBULIN2 were used as quantitative controls for RT-qPCR and RT-PCR, respectively. All experiments were repeated 3 times. For analysis of splicing patterns of pre-mRNA, the primers for RT-PCR assay were designed in two adjacent exons to ensure specific amplification of spliced RNA and unspliced RNA (intron retention). To test the unspliced RNA, intron retention, the primers for RT-qPCR detection were designed in the introns and their 5′-exon according to the description by Xiong et al. (2019a, 2019b; Supplemental Table S4). Splicing efficiency for tested introns was calculated by normalizing the spliced RNA level to the unspliced RNA level (Marquardt et al., 2014; Xiong et al., 2019a, 2019b). Statistically significant differences were analyzed by two-tailed Student’s t test or with Tukey’s post-hoc test.

Accession numbers

Sequence data from this article can be found in the GenBank data libraries under the following accession numbers: Arabidopsis AtBUD13 (AT1G31870), GDS1 (AT3G412700), DDL (AT3G20550), AtU2AF65B (AT1G60900) and SKIP (AT1G77180); yeast Bud13 (NP_011341.3), Snu17 (P40565.1) and Pml1 (NP_013116.1); Human hBUD13 (NP_116114.1), RBMX2 (NP_057108.2), and SNIP1 (NP_078976.2).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. BLAST sequence analysis of Arabidopsis proteins with homology to human RBMX2 and yeast IST3/Snu17.

Supplemental Figure S2. Amino acid sequence analysis of GDS1 and its putative homologs.

Supplemental Figure S3. Alignment of amino acid sequences from Arabidopsis DDL, human SNIP1, and yeast Pml1.

Supplemental Figure S4. BiFC assay of AtBUD13, GDS1, and DDL in N. benthamiana leaves.

Supplemental Figure S5. Co-IP assay between AtBUD13 and GDS1 in Arabidopsis.

Supplemental Figure S6. Co-IP of GDS1 with DDL in Arabidopsis.

Supplemental Figure S7. Occurrence of RES complex subunits in eukaryotes.

Supplemental Figure S8. BiFC assay between the GDS1 RRM domain and the BUD13 domain of AtBUD13.

Supplemental Figure S9. Critical role of tryptophan 526 of AtBUD13 for AtBUD13–GDS1 interaction.

Supplemental Figure S10. Interactions between GDS1 and different regions of DDL.

Supplemental Figure S11. Expression patterns of AtBUD13, GDS1, and DDL in various tissues and organs during Arabidopsis development.

Supplemental Figure S12. Expression patterns of GDS1 and DDL.

Supplemental Figure S13. Quantification of root meristematic cells in amiRNA-AtBUD13 seedling roots.

Supplemental Figure S14. Characterization of mutants for GDS1.

Supplemental Figure S15. Defective development of the ddl mutants.

Supplemental Figure S16. Growth and seed abortion of gds1 atbud13 double mutant.

Supplemental Figure S17. Defective growth and development of the gds1 ddl double mutant.

Supplemental Figure S18. Decreased expression levels of ATML1, WOX2, SCR, and PDF2 in atbud13, gds1, and ddl mutants.

Supplemental Figure S19. Interactions of the RES complex subunits with SKIP and AtU2AF65B.

Supplemental Figure S20. A proposed working model of the RES complex functions in Arabidopsis development.

Supplemental Table S1. BLAST analysis with RBMX2 against Arabidopsis protein data.

Supplemental Table S2. Genes involved in root development with abnormal splicing detected by RNA-Seq analysis.

Supplemental Table S3. Genes involved in cell cycle and cell division with abnormal splicing detected by RNA-Seq analysis.

Supplemental Table S4. Primers used in this study.

X.L.W. and F.X. conceived the project. F.X., Z.Z., H.D.Q., J.J.R., and Y.Y.W. performed the research. X.L.W. and F.X. wrote the paper with contribution of M.S.O. All authors commented on the manuscript. X.L.W. agrees to serve as the author responsible for contact and ensures communication.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://dbpia.nl.go.kr/plphys/pages/general-instructions) is Xiu-Ling Wang ([email protected]).

Acknowledgments

We are grateful to the Arabidopsis Biological Resource Center for providing the T-DNA insertion lines gds1 (SAIL_56_E10 and SK33362) and ddl (Col-0) (SALK_045025 and SAIL_1281_F08). We thank Dr. Morris ER and Walker JC (University of Missouri) for the mutant seeds of ddl1-1 and ddl1-2, Dr. Marcus J.O. Johansson (Umea˚ University) for the yeast cells wild-type cells (UMY2219), snu17⊿ (MJY548) and pml1⊿ (MJY535), and Dr. Ligeng Ma (Capital Normal University) for the plasmid of pro35S::U1-70K-mCherry construct.

Funding

This work was supported by the National Natural Science Foundation of China (31970330 and 32170361).

Conflict of interest statement. The authors declare no conflicts of interest.

References

Author notes