Abstract
O-linked β-N-acetylglucosamine (O-GlcNAc) and O-fucose are two sugar-based post-translational modifications whose mechanistic role in plant signalling and transcriptional regulation is still largely unknown. Here we investigated how two O-glycosyltransferase enzymes of Arabidopsis thaliana, SPINDLY (SPY) and SECRET AGENT (SEC), promote the activity of the basic helix–loop–helix transcription factor SPATULA (SPT) during morphogenesis of the plant female reproductive organ apex, the style. SPY and SEC modify amino-terminal residues of SPT in vivo and in vitro by attaching O-fucose and O-GlcNAc, respectively. This post-translational regulation does not impact SPT homo- and heterodimerization events, although it enhances the affinity of SPT for the kinase PINOID gene locus and its transcriptional repression. Our findings offer a mechanistic example of the effect of O-GlcNAc and O-fucose on the activity of a plant transcription factor and reveal previously unrecognized roles for SEC and SPY in orchestrating style elongation and shape.
Similar content being viewed by others
Main
The gynoecium, a sophisticated organ within the flower, ensures fertilization and seed production in angiosperms. Depending on plant species and/or developmental window, gynoecium shape may greatly vary to complement its function1,2. Of notable importance is the morphological diversity observed at the apical-distal end of the gynoecium, encompassing the style and stigma. This region serves as the landing site for pollen, where germination occurs, and subsequently facilitates the pollen’s journey to fertilize the ovules2. Both plant fitness and seed production are thus intricately linked to the correct development of the style and stigma. This developmental process is highly dynamic, and it is underpinned by a complex molecular orchestration that operates in an intertwined manner3,4.
Among the transcription factors (TFs) presiding over gynoecium development, the activity of SPATULA (SPT)—a key regulator of medial tissue identity and style morphogenesis5—is pivotal in orchestrating auxin accumulation as well as coordinating the medio-lateral and adaxial–abaxial polarity axis6,7,8 and repressing cytokinin (CK)-mediated cell-proliferation input at the gynoecium apex9. SPT forms homo- and heterodimers with specific TF partners to modulate several aspects of style development4,9,10,11,12. Accordingly, SPT loss-of-function mutants fail to form a fused, radially symmetric style at the gynoecium apex5 (producing a so-called split style), and this phenotype is exacerbated by mutations in other basic helix–loop–helix (bHLH) TFs such as INDEHISCENT (IND)13 and HECATE 1, 2 and 3 (HEC1,2,3)8,11 as well as members of the NGATHA protein family4.
Because of its stable spatiotemporal expression in the apical-medial tissues during gynoecium development5,8, we hypothesized that the dynamic activity of SPT required for style patterning may be regulated at the protein level. Interestingly, a proteomic investigation of Arabidopsis flower proteins showed that SPT is post-translationally modified by O-linked β-N-acetylglucosamine (O-GlcNAc)13.
O-GlcNAc is a post-translational modification (PTM) associated with key cellular processes and stress responses in both animals and plants14,15,16,17,18. In both kingdoms, O-GlcNAc is attached to a plethora of proteins encompassing various cellular functions, among which TFs are highly abundant19,20,21. O-GlcNAc decoration of target proteins leads to the modulation of their stability, cellular localization, protein–protein interactions, transcriptional activity and other characteristics20,22. Mechanistic studies in the animal field have shown that O-GlcNAc modifications can impact TF activity by changing, for example, their DNA-affinity binding, their ability to bind transcriptional co-activators/co-repressors and/or their domain function20,22,23,24,25.
Despite its importance across kingdoms and biological scales, the translation of O-GlcNAc modification into specific protein functions remains predominantly uncharted territory. This knowledge gap is particularly pronounced in the plant kingdom, where the mechanistic understanding of how sugar-based PTMs influence the activity of TFs is notably lacking. The modification of SPT by O-GlcNAc might thus provide a rapid, adaptable mechanism to intricately regulate SPT activity during gynoecium development.
In Arabidopsis, the enzyme SECRET AGENT (SEC) catalyses the addition of O-GlcNAc from UDP–GlcNAc to Ser and/or Thr residues of acceptor-substrate proteins18. A functionally related enzyme encoded by the gene locus SPINDLY (SPY)26 functions as an O-fucosyltransferase (POFUT), attaching monofucose (O-fucose) to target substrates26,27. Notably, while single sec and spy mutant alleles are perfectly viable in Arabidopsis, sec spy double mutants are embryo lethal (similar to ogt mutants in animals28,29), highlighting the synergistic and fundamental importance of both enzymes for plant development30.
Examples of mechanistic outcomes resulting from the SEC-mediated O-glycosylation of their acceptor targets include alteration in protein–protein interactions of the RNA-binding protein TaGRP2 during wheat vernalization31 and of the gibberellin repressor RGA in Arabidopsis18. Given that RGA does not directly bind to DNA, it has been proposed that O-GlcNAc modification of this transcriptional repressor could elicit a conformational change in its protein structure. This structural alteration, in turn, would negatively impact the binding of RGA to its protein interactors, thereby facilitating the progression of gene expression18,26. In contrast, the modification of RGA by O-fucose would trigger an open conformation that retains the binding of RGA to its protein interactors, leading to repression of the downstream signalling pathways, opposite to the effect of O-GlcNAc18,26.
Our investigation allows a pioneering understanding of the functional implications of O-GlcNAc modifications for plant TF activity. By means of genetic, molecular, biochemical and proteomic experiments, we herein demonstrate a role for SEC and SPY in style development via post-translational regulation of SPT activity. We demonstrate that SPT directly interacts with SEC and SPY, which modify the amino terminus of SPT by O-GlcNAc and O-fucose, respectively, both in vivo and in vitro. Moreover, via a genetic complementation assay, we provide evidence that specific modified residues located in two N-terminal peptides are essential for SPT function in vivo, accounting for style morphogenesis. Furthermore, we show that both enzymes enhance the transcriptional activity and DNA-binding affinity of SPT to the PINOID (PID) gene locus32, but do not impact SPT nuclear localization, protein stability or dimerization events with itself, IND or HEC1. Accordingly, the genetic epistasis analysis between SEC/SPY and SPT, coupled with the style morphological defects displayed by the inducible SEC knockdown in a spy mutant background (SEC RNAi spy-3), corroborates a model in which SPT activity is primed by SEC and SPY transferases to fine-tune its downstream control of the hormonal balance and support style development.
Our findings provide a mechanistic insight into the role of O-GlcNAc and O-fucose PTMs in the activity of a plant TF, characterize a post-embryonic target modified synergistically by both SEC and SPY during style development, and reinforce the importance of the upstream regulation mediated by these sugar-based PTMs on key TFs important for organ morphogenesis in multicellular organisms.
Results
SPT is modified by O-GlcNAc and O-fucose via SEC and SPY
To understand whether a post-translation mechanism based on O-glycosylation could preside over SPT-mediated control of style development, we performed higher-energy collisional dissociation (HCD) and electron transfer higher-energy collision dissociation (EThcD) fragmentation mass spectrometry (MS/MS) analyses on nuclear extracts from fully rescued spt-12/SPT::SPT–sYFP complementation line inflorescences (Extended Data Fig. 1a,b). The analyses identified several Ser and Thr residues at the N terminus of SPT that were modified by two O-linked sugar moieties, O-fucose and O-GlcNAc. In vivo, the HCD spectra indicated that both sugar groups can be localized on Ser23 and Ser24 (recovered in peptide_1, LISSSSSSSVYDTR), while Ser25 was found modified by O-fucose (Fig. 1a–c and Extended Data Fig. 1c–e). EThcD analysis of the same peptide showed a wider distribution of modifications; it indicated that the glycans could be attached to Ser23, Ser24 and Ser25 and also showed that Ser26 and Ser27 can be modified (Supplementary Table 1).
The HCD spectra corresponding to the second peptide of interest, initiating with Ser60 (peptide_2: SSPLPSYYSPATTTTTASLIGVHGSGDPHADNSR), did not provide sufficient information for the precise localization of the glycan modifications. However, the observed precursor mass strongly suggests the presence of a modification consistent with O-GlcNAc (Fig. 1a). Furthermore, all EThcD spectra obtained for peptide_2 consistently indicated that one of the Thr71-to-Thr74 residues was the site of the modification (Supplementary Table 2). These findings align with a prior investigation performed on O-GlcNAc-modified Arabidopsis flower proteins21.
The quantification of both modifications on the two N-terminal peptides of SPT was then carried out by normalizing the abundances of each modified peptide version to the total sum of the abundances of all versions (modified and unmodified) of the same peptide. The levels of O-fucose modification were higher than those of O-GlcNAc on peptide_1, while on peptide_2, O-GlcNAcylation was conspicuously elevated and O-fucosylation was barely detected (Fig. 1d).
To assess whether modifications by O-GlcNAc and O-fucose on SPT were dependent on SEC and SPY activity, respectively, we crossed the spt-12/SPT::SPT–sYFP complementation line with a single loss-of-function mutant of SEC (sec-5)33 and a catalytically redundant mutant of SPY (spy-3)34 (Extended Data Fig. 2a). Using spt-12;sec-5/SPT::SPT–sYFP and spt-12;spy-3/SPT::SPT–sYFP inflorescences, we performed a similar HCD MS/MS analysis and compared the percentage of modifications in relation to the spt-12/SPT::SPT–sYFP segregating controls. Since O-fucosylation was hardly detected on peptide_2, we focused this analysis on peptide_1. Our data showed that O-fucose modification was completely abolished on peptide_1 in the spy-3 background, while O-GlcNAc was present at comparable levels (Fig. 1e). In contrast, in the spt-12;sec-5/SPT::SPT–sYFP line, we did not observe any change in the percentage of detectable O-fucose, while O-GlcNAc was no longer detected in this background (Fig. 1f). This analysis demonstrates that SPY and SEC modify SPT by O-fucose and O-GlcNAc, respectively, targeting both shared and specific residues. Our analysis highlights that the enzymes do not compensate for each other’s activity in their respective mutant backgrounds, since we did not observe changes in transcript levels of SPY and SEC in sec-5 and spy-3 inflorescences, respectively (Extended Data Fig. 2b), or an increase in their enzymatic activity as assessed by PTMs recovered on SPT (Fig. 1e,f).
To confirm the direct PTM of SPT by O-GlcNAc and O-fucose through SEC and SPY, we performed in vitro enzymatic assays. In these assays, we tested the ability of recombinant SEC (5TPR–SEC)26 and SPY (3TPR–SPY)26 to directly modify the full-length SPT protein (6xHis–SPT) (Extended Data Fig. 3a) in the presence and absence of their specific donor substrate (UDP–GlcNAc and GDP–fucose). Collectively, we found that the HCD spectra obtained in the in vitro experiments recapitulated the in vivo findings, showing that O-GlcNAc and O-fucose can modify Ser23 to Ser25 of SPT (Extended Data Fig. 3b–d).
Furthermore, no differences were observed for SPT transcript levels from spy-3 and sec-5 inflorescences compared to the wild type (Extended Data Fig. 2c), and no reduction in SPT stability (as measured by the intensity of the SPT–YFP signal) or alterations in its subcellular localization were observed in the single spy-3 and sec-5 mutants (Extended Data Fig. 2d,e). Altogether, our findings corroborate the idea that SEC and SPY work upstream of SPT, acting at the post-translational level.
Our data thus show that specific residues of SPT can host both O-GlcNAc and O-fucosyl moieties, which are attached by SEC and SPY in vivo and in vitro. Moreover, our data suggest that SEC and SPY may work redundantly but do not compensate for one another in the inflorescences of Arabidopsis.
SPT interacts in vivo and in vitro with SEC and SPY
To support a role for SEC and SPY in controlling SPT at the post-translational level, we tested whether these enzymes could directly interact with SPT. To this end, we employed co-immunoprecipitation (Co-IP) and split-luciferase complementation assays in Nicotiana benthamiana leaves (Fig. 2a,b) alongside yeast two-hybrid (Y2H) experiments (Fig. 2c and Extended Data Fig. 4), which all demonstrated direct interactions between full-length SPT–SPY and SPT–SEC proteins. Moreover, Y2H experiments also showed that both SPY and SEC bind to SPT via their N-terminal tetratricopeptide repeats (TPRs)26, TPRs 1–11 of SPY and TPRs 1–13 of SEC, while the carboxy-terminal catalytic domain showed no growth of the yeast cells on selective media (Extended Data Fig. 4).
Most of the residues modified on SPT were positioned at peptides flanking an amphipathic helix (AH) at the N terminus of SPT (Fig. 1a), a domain that supports the transcriptional activation activity of SPT35. Thus, to dissect the domains responsible for mediating the enzyme interactions with SPT, we performed Y2H experiments using the full-length sequences of SPY and SEC in combination with truncated versions of SPT. Our results showed the SPT N-terminal domain (spanning residues 1–180) containing the AH and acid (Ac) domains35, as well as a fragment containing only the AH domain (residues 1–135), were both sufficient to produce positive interactions with full-length SEC and SPY proteins (Fig. 2c and Extended Data Fig. 4) but did not interact with HEC1—a known interactor of SPT (Extended Data Fig. 4). In contrast, the SPT C-terminal truncation (residues 136–373), including both the Ac and bHLH domains, did not interact with SPY but could interact with SEC as well as HEC1 (Fig. 2c and Extended Data Fig. 4). These results are in line with a working model in which the binding of SPT with SPY occurs through the N-terminal domain of SPT containing the modified residues flanking the AH domain, while SEC interacts with the full-length SPT protein. These findings also provide insights into the signalling modulation at the N terminus of SPT, a crucial molecular site for both transcriptional regulation and carpel patterning35.
SEC and SPY synergistically control style morphogenesis
If the activity of Arabidopsis SEC and SPY controls style development at the gynoecium apex, defects in style specification or elongation or a break in radial symmetry (similar to spt) should be observed in mutants for sec and spy. Despite decades of genetic analysis on single mutants for these two enzymes, no defects have ever been reported30,36,37. Accordingly, our scanning electron microscopy (SEM) analysis of loss-of-function spy-4 (ref. 37) and sec-5 single mutant gynoecia, as well as the catalytically redundant mutants spy-3 and sec-2 (ref. 36), did not show any significant defects in style formation (Figs. 3a and 4a). The correct style development observed in sec and spy mutants is in line with normal SPT expression within the style of these single mutants (Extended Data Fig. 2c,d). This may suggest a redundant role for both enzymes in controlling the morphogenesis of the gynoecium apex.
Varying effects on carpel development were unveiled through haplo-deficiency genetic analysis of the SEC and SPY functions—that is, four-carpel ovaries and pin-like flowers36. To overcome the embryo-lethal effect elicited by the elimination of both enzymatic activities30,36 and reveal the synergistic post-embryonic roles of the two enzymes (encompassing style development), we produced a dexamethasone (DEX)-inducible SEC RNAi construct to lower SEC mRNA levels in the spy-3 background (SEC RNAi spy-3) (Extended Data Fig. 5a). Two specific target guides were designed (named a1SEC and a2SEC) for SEC, ensuring low off-target scores (Extended Data Fig. 5b,c). Only the Arabidopsis transgenic lines generated using the a2SEC guide showed reduced levels of SEC transcripts, as determined by quantitative real-time PCR (qRT-PCR) experiments conducted from DEX-sprayed inflorescences (Fig. 3b). Three independent transgenic lines recovered using the a2SEC guide were subsequently phenotypically analysed and compared to Col-0, spy-3 and SEC RNAi spy-3 lines that did not exhibit SEC downregulation, as negative controls (Fig. 3a,b). We observed defects in style morphogenesis after DEX treatment of transgenic lines with reduced levels of SEC expression—that is, a strong reduction in style length (Fig. 3a,c), a lack of wax/crenulations of stylar cells (used as a differentiation marker for style) (Fig. 3d), a reduction of cell elongation below the apical stigmatic tissue (Fig. 3d) and a consistent reduction in fruit length (Fig. 3e). The short-style phenotype and compromised cell elongation observed in the DEX-induced SEC RNAi spy-3 is remarkably different from the split-style phenotype of spt (Fig. 4a). It is consistent with a lack of apical-basal anisotropic growth displayed by cells at the developing apical gynoecium, where the style forms38,39. These results thus reveal a previously unrecognized role for SEC and SPY in style development and suggest that SEC and SPY control early developmental processes of style morphogenesis, presumably via a redundant and synergistic control of style- and fruit-specific targets. Consistent with a positive role for SEC and SPY in post-embryonic plant organogenesis, we observed a reduction of root length in SEC RNAi spy-3 (a2SEC) seedlings grown on DEX media for seven days compared with mock-treated samples and DEX-treated Col-0 and spy-3 parental line (Fig. 3f). Moreover, we observed arrest of growth after 13 days of growth on DEX media of the SEC RNAi spy-3 genetic background (Fig. 3g). Interestingly, while knocking down SEC activity did not affect the germination rate differently from the spy-3 control (Extended Data Fig. 5d–f), the reduction of root growth and lethality observed at the seedling stage confirm the fundamental roles of both enzymes in post-embryonic plant development.
Next, to investigate the role of SEC and SPY in style radialization via regulation of SPT activity, we analysed the gynoecia of spt sec (spt-12 sec-5 and spt-12 sec-2) and spt spy (spt-12 spy-4 and spt-12 spy-3) double mutants (Fig. 4a). Our genetic analysis showed that spt sec double mutants displayed a similar phenotype to the segregating spt-12 control, while spt spy double mutant combinations increased the frequency and severity of the spt split-style phenotype by strongly augmenting the percentage of bilateral versus radial styles observed (Fig. 4b), as well as the depth of the medial cleft (measured as the ratio between the cleft length and the style length) (Fig. 4c). These data indicate that spt is epistatic to the loss of both SEC and SPY enzymatic activities and suggest that O-fucosylation of other key regulators and cellular processes of style development may be regulated in parallel by SPY (for example, apical-basal anisotropic growth)38,39.
To further examine this possibility, sec and spy mutant gynoecia were treated with CK. No defects were displayed by sec mutants, but for the spy mutants we observed extensive proliferation of the medial-apical region that connects the stigma and replum (spy-3) as well as ectopic, unbalanced growth of the lateral shoulders (spy-4) (Extended Data Fig. 6), a phenotype associated with CK hypersensitivity that has not been previously observed. These data confirm that SPY plays multiple roles during gynoecium development. They also corroborate the idea that SPY and SPT work in similar pathways, as they both repress the CK proliferating output at the gynoecium apex, which is relevant as it provides a genetic framework for an antagonistic SPY–CK interaction.
Specific Ser/Thr residues of SPT promote style radialization
To test the effect of O-glycosyl decorations on SPT function during radial symmetry establishment at the gynoecium apex, we carried out a detailed mutant complementation assay by producing a series of loss-of-O-glycosylation mutant variants of SPT (expressed under the SPT native 5-kb promoter40) and analysed their ability to complement the spt split-style phenotype. To this end, specific Ser and Thr residues targeted by SEC and SPY in vivo and in vitro on SPT peptide_1 (Fig. 1a and Extended Data Figs. 1 and 3) and peptide_2 (ref. 21) were mutated to Ala to mimic the loss of modification41. We predicted that the lack of style complementation would indicate a positive functional role for those residues and the associated PTMs in sustaining SPT function. We produced and analysed the following mutant versions of SPT: Ser23-to-Ala; Ser23, Ser24 and Ser25-to-Ala (hereafter Ser23–25-to-Ala, on peptide_1); Thr71, Thr72, Thr73 and Thr74-to-Ala (hereafter Thr71–74-to-Ala21, on peptide_2); Ser23, Ser24, Ser25, Thr71, Thr72, Thr73 and Thr74-to-Ala (hereafter Ser+Thr-to-Ala, on peptide_1 and peptide_2); and Ser60 and Ser61-to-Ala (Ser60,61-to-Ala, as a negative control, on peptide_2) (Fig. 5a). First, to exclude the possibility that changes in the amino acid sequence of SPT would lead to alterations in transcript levels and protein stability or localization, we fused the SPT wild-type and mutant sequences with yellow fluorescent protein (YFP) and carried out qRT-PCR experiments (Extended Data Fig. 7a,b) and confocal microscopy analysis (Fig. 5b). None of the lines examined showed a significant reduction in SPT mRNA levels (Extended Data Fig. 7a,b). Furthermore, confocal microscopy analysis showed that a clear YFP signal was present at the apex of gynoecia—where style and stigmatic tissue form—expressing the wild-type SPT complementation line as well as the Ser23-to-Ala (peptide_1) and Ser60,61-to-Ala (peptide_2) point mutations (Fig. 5b). Accordingly, SEM analysis showed that the spt phenotype was complemented by Ser23-to-Ala and Ser60,61-to-Ala mutations to a similar level as the wild-type sequence, equal to 100% of radial styles (Fig. 5b). In contrast, despite clear SPT nuclear expression at the gynoecium apex of Ser23–25-to-Ala (peptide_1), Thr71–74-to-Ala (peptide_2) and Ser+Thr-to-Ala (both peptides), these point-mutation lines displayed a high percentage of bilateral styles, resembling the spt mutant. Specifically, 40.9%, 37.7% and 48.4% of the styles observed in the aforementioned transgenic lines were unfused—that is, they had the split-style phenotype (Fig. 5b).
These results thus indicate a functional role for specific SPT residues in style radial symmetry development in Arabidopsis via a mechanism distinct from transcriptional expression, protein stability and cellular localization.
O-glycosyl PTMs on SPT promote its transcriptional activity
Since SPT binds DNA (and thus regulates transcription) as a dimer4, we hypothesized that SEC and SPY may impact its ability to interact with protein partners and/or specific promoters. To understand how PTMs of SPT may play a role in orchestrating style morphogenesis, we tested whether SEC and SPY could enhance the formation of the SPT–SPT homodimers and/or SPT–IND and SPT–HEC1 heterodimers essential for correct style morphogenesis.
To test whether SEC and SPY could impact SPT dimerization events, we performed fluorescence resonance energy transfer (FRET)–fluorescence-lifetime imaging microscopy (FLIM) quantitative assays in tobacco leaves to assess whether the co-expression of Agrobacterium harbouring either SEC–HA (35S::SEC–3xHA) or SPY–HA (35S::SPY–3xHA) would lead to a further reduction in the lifetime of the FRET donor (GFP) compared with the formation of the homodimer SPT–GFP;SPT–RFP (35S::SPT::GFP;35S::SPT–RFP) and the heterodimers SPT–GFP;IND–RFP (35S::SPT::GFP;35S::IND–RFP) and SPT–GFP;HEC1–RFP (35S::SPT::GFP;35S::HEC1–RFP) (Fig. 6a).
To begin, we confirmed the expression of SEC–HA and SPY–HA recombinant proteins by western blotting (Extended Data Fig. 8a). We then tested whether SEC and SPY could modify SPT in tobacco leaves, by means of HCD MS/MS experiments (Extended Data Fig. 8b,c). In line with our in vivo and in vitro findings (Fig. 1 and Extended Data Figs. 1c–e and 3b–d), co-infiltration of SEC or SPY with SPT resulted in an elevated percentage of O-GlcNAc or O-fucose modifications, respectively, detected on peptide_1 (Extended Data Fig. 8b) as well as a higher ratio of modified peptides compared with the SPT control (Extended Data Fig. 8c).
Next, we confirmed a direct interaction in the nuclei co-expressing (;) SPT–GFP;SPT–RFP as well as SPT–GFP;IND–RFP and SPT–GFP;HEC1–RFP (without co-expression of the enzymes) with FRET–FLIM assays. The GFP fluorescence lifetime was significantly shorter in the presence of SPT–GFP;SPT–RFP, SPT–GFP;IND–RFP and SPT–GFP;HEC1–RFP interactions than in the negative control: SPT–GFP;RFP–NLS (which targets the RFP protein in the nucleus) (Fig. 6a). These changes in GFP fluorescence lifetime are consistent with FRET efficiencies of 5.7%, 9.6% and 7.4% for SPT;SPT, SPT;IND and SPT;HEC1, respectively (Fig. 6a). The positive FRET control, SPT–GFP–RFP (in which the two fluorescent tags are both cloned in cis to SPT) showed an even stronger decrease in GFP fluorescence lifetime, which translated into a FRET efficiency of 18.8% (Fig. 6a). However, the presence of either SEC–HA or SPY–HA had no significant effect on the FRET efficiencies observed in the presence of interactions between SPT–GFP and its RFP-tagged partners (Fig. 6a). In principle, this would suggest that O-GlcNAc and O-fucose do not promote the formation of SPT-containing dimers. However, the enzymes may modulate dimer formation via other mechanisms—for example, by conformational rearrangement.
We then investigated whether SEC and SPY would impact SPT-mediated gene expression. The best-characterized downstream target of SPT (and IND) during style development is the PID kinase6,9,42,43, which is downregulated to support the accumulation of the auxin morphogenic signal within specific medial-apical cells6. We thus tested whether the downregulation of PID expression by SPT–IND was exacerbated by SEC and SPY by performing a transient transactivation assay in N. benthamiana leaves using the same constructs used for the FRET–FLIM experiments and a 1-kb promoter fragment of PID, containing the cis-elements directly bound by both IND9,10,42,43 and SPT in vivo (Extended Data Fig. 9a).
The PID promoter was fused to the GUS gene as a reporter. To begin, we confirmed that the expression of SEC–HA or SPY–HA did not affect GUS expression (Extended Data Fig. 9b), and the transcriptional read-out of SPT/IND was determined by qRT-PCR experiments using different ODs for co-infiltration (Fig. 6b and Extended Data Fig. 9c). We first tested for equal expression of the co-infiltrated recombinant proteins by western blots and then compared the GUS transcription levels (Fig. 6b). The expression of SPT–GFP and IND–RFP either alone (homodimers) or co-expressed (heterodimer) significantly diminished pPID::GUS expression levels (Fig. 6b and Extended Data Fig. 9c), while the expression of either SEC–HA or SPY–HA had no significant effect on transcription alone (Extended Data Fig. 9b). However, co-expression of both SEC–HA and SPY–HA together with SPT–GFP and IND–RFP led to a significant further reduction of the background levels of PID expression (Fig. 6b). Furthermore, co-expression of the catalytic SPY (spy-3–HA) and SEC (sec-5–HA) mutant enzymes, together with SPT–GFP and IND–RFP recombinant proteins, led to a reversion of the GUS transcriptional levels to the basal repression trigged by the SPT–GFP/IND–RFP heterodimer (Fig. 6b). Altogether, these results demonstrate that both SEC and SPY enhance SPT/IND-mediated repression of PID transcription.
To corroborate the idea that O-glycosyl PTMs of SPT directly influence its transcriptional activity rather than dimerization events, we tested whether the affinity of SPT for the PID promoter was enhanced by SEC and SPY. To this end, we employed electrophoretic mobility shift assay (EMSA) experiments performed using the full-length 6xHis–SPT recombinant protein (Extended Data Fig. 3a), which we showed was able to bind a 171-bp fragment of the PID promoter encompassing the cis-elements (G-box, CACGTG) recognized in vivo by SPT (Extended Data Fig. 9a,d). Moreover, the electrophoresis mobility shift signal was strongly abolished when the wild-type G-box sequence of the PID promoter was mutated (TGATGA)44 (Extended Data Fig. 9d), in agreement with previous data19. Notably, when SPT was incubated with the PID promoter fragment in the presence of recombinant 5TPR–SEC–HA and 3TPR–SPY–HA proteins and their respective donor substrates, we observed a stronger mobility shift occurring at lower concentrations of SPT recombinant protein (Fig. 6c).
Altogether, our findings show that SEC and SPY can modulate SPT-mediated control of PID expression by increasing its DNA-binding affinity and/or transcriptional activity rather than its ability to form homo- or heterodimers.
Discussion
This work provides the genetic and molecular foundation for elucidating the role of Arabidopsis thaliana SEC18 and SPY26 O-glycosyl enzymes in style development5,6. More broadly, our data provide a molecular characterization of the functional roles of O-GlcNAc and O-fucose PTMs in the regulation of a plant bHLH TF—that is, SPT, which plays pivotal roles in tissue identity5, body-axis formation8 and bilateral-to-radial symmetry transition6 during gynoecium development. As comparable biological processes and conserved molecular players (that is, O-GlcNAc14 and bHLH TFs45) are present in both the plant and animal kingdoms, this study sheds light on a fundamental mechanism that could be employed more generally during the orchestration of organ morphogenesis and symmetry establishment in higher eukaryotes.
Ever since its initial discovery in the 1980s by Torres and Hart46, the significance of O-GlcNAc as a signalling molecule in processes such as nutritional sensing, stress response and animal development has been long established. Despite its importance in several cellular processes and the extensive range of targets modified by O-GlcNAc19, the impact of this PTM on protein functions has been comprehensively elucidated for only a limited number of targets. Mechanistic examples that might share similarity to the effect of O-GlcNAc on SPT include three well-characterized TFs from the animal kingdom: PDX-1 (ref. 25), Oct4 (ref. 22) and C/EBPβ24.
In plants, it has been previously proposed that SEC and SPY control the activity of other acceptor substrates by promoting protein stability47, controlling cellular localization48, enhancing or disrupting protein–protein interactions26,49 and fine-tuning gene expression48. Among these, the best-characterized molecular example is the DELLA protein RGA, which is antagonistically regulated by SEC and SPY during Arabidopsis post-embryonic development18,26.
Here we showed that the OGT enzyme SEC and the POFUT enzyme SPY can both physically interact with SPT in vivo and in vitro (Fig. 2) and modify several Ser/Thr residues at the N terminus of SPT (Fig. 1 and Extended Data Figs. 1 and 3) flanking its AH domain (Fig. 2c and Extended Data Fig. 4). Mutations of the SPT sequence at Ser23 to Ser25 and Thr71 to Thr74 (to Ala) demonstrated a functional role for those residues in style morphology, thus linking specific O-GlcNAc and O-fucose modifications to organ function and development.
The AH domain of SPT is quite different from the N-terminal sequence of other plant bHLH TFs35, reinforcing the idea that it hosts specific residues for signal transduction mediated by O-GlcNAc and O-fucose. Notably, the AH domain has been found to be either highly conserved across some SPT orthologues (for example, tomato) or missing from other plant species completely (for example, in monocot rice)40. This supports a scenario in which the evolution of the N terminus of SPT and its regulation by sugar decorations might be the site of style/stigma shape diversification in monocots and dicots. It has long been known that the AH domain of SPT supports the essential role of the Ac domain in carpel function35, but the molecular basis of its modulation is yet to be determined.
O-glycosylation of SPT did not lower its levels, either at the transcriptional or at the protein level (Extended Data Fig. 2c–e), even though O-fucosylation of SPT by SPY might, in addition, fine-tune SPT protein stability, as we observed a slight increase of SPT protein levels at the gynoecium apex in the spy-3 single mutant background (Extended Data Fig. 2d,e). Although the N terminus of SPT is predicted to be unstructured by AlphaFold50 (AF-Q9FUA4-F1) and the residues modified by PTMs are located within a disordered region of SPT, we excluded a possible effect of the mutations on protein folding or instability as the C-terminal YFP, cloned in cis to the wild-type and mutant versions of SPT, was visible, suggesting that the transgene protein precursors are correctly folded and expressed (Fig. 5b). Since the N-terminal region, including both the AH and Ac domains, is important for SPT transcriptional activation35, the attachment of PTMs flanking the AH domain supports two mechanistic scenarios for the function of the AH domain as a modulator of the Ac domain activity: (1) a direct effect via a conformational change or (2) an indirect effect via interaction with other transcriptional activators and/or co-repressors.
While SEC and SPY did not enhance SPT’s ability to form protein–protein interactions with itself and its known dimerization partners IND and HEC1 in tobacco leaves (Fig. 6a), we observed an augmented transcriptional repression of its known downstream target PID6,9 (Fig. 6b). Accordingly, the effect of glycosylation on SPT increased the affinity of SPT to a DNA fragment of the PID locus (Fig. 6c). Altogether, this supports a working model where SEC and SPY, and the PTMs they transfer onto SPT, can directly modify the affinity of SPT for its DNA targets and its transcriptional activity (Fig. 6d). To delve deeper into this scenario, in vivo transcriptomic studies that employ cellular resolution across various gynoecium tissue types will be essential to provide insights into the spatiotemporal contribution of PTMs to SPT activity.
O-GlcNAc has been linked to the cold-temperature response in both animals51 and plants31. Interestingly, in addition to being a master regulator of style development, SPT can transduce environmental cues into developmental programmes by integrating external signals such as cold temperature52,53 and light quality54 during seed germination and style development. Even though SPT was the first regulatory gene described to control the cold response during seed germination52, the molecular mechanism underpinning this process is still unelucidated, and post-translational control of SPT was hypothesized to underpin its mechanistic activity53. This raises the intriguing possibility that SPT might be regulated by O-GlcNAc at the post-transcriptional level to integrate environmental cues into organ development.
A recent O-fucosylomics55 study performed in young Arabidopsis seedlings identified an N-terminal peptide of SPT modified by O-fucose, coinciding with our peptide_1 (Fig. 1), whose modification was lost in the spy-4 mutant background. This strengthens the importance of sugar-based PTM of SPT in other developmental contexts—for example, roots and cotyledons. Such a regulatory mechanism would fit well with the view of SPT as integrator of abiotic signals and genetic factors, similar to that proposed for the DELLA protein RGA56. Conversely, even though 262 O-GlcNAc-modified flower proteins have been identified21, a gap in knowledge remains regarding the SPY targets and the impact of O-fucose on key regulators of gynoecium development. This includes the CK signalling pathway, as implied by the heightened responsiveness of spy alleles to CK treatments (Extended Data Fig. 6). The severe phenotype of the spt spy double mutant is thus in line with a scenario in which other key regulators of style development and hormonal regulators are O-fucosylated by SPY in parallel to SPT, leading to the augmented phenotype observed in the spt spy mutant (Fig. 4).
Notably, the complex network of TFs orchestrating style development includes several layers of cross- and self-regulation that act at both the transcriptional and translational levels15. Although our FRET–FLIM experiments in tobacco leaves indicate that SEC and SPY may not regulate SPT–SPT, SPT–IND and SPT–HEC1 dimer formation (Fig. 6a), it is still possible that specific TF interactions might be mediated by O-GlcNAc and O-fucose in a cell-/tissue-type manner and/or at a specific developmental stage. Also, additional upstream regulation can be hypothesized, as three NGA family members have been identified as targets of O-GlcNAc modification in Arabidopsis inflorescences21.
Our genetic analysis and the phenotypes unveiled by the inducible spy-3 SEC RNAi mutant confirm fundamental roles for these two enzymes in post-embryonic growth, including seedling and root growth as well as fruit and style development (Fig. 3). The synergistic role of SEC and SPY in Arabidopsis reproductive development has been previously observed by haplo-insufficiency genetic analysis, which revealed that heterozygosis of SEC/sec-1 in a spy-4 ga1 double mutant background increased the number of gynoecia with four carpels36, a phenotype previously associated with the spy-2 allele57. In contrast, heterozygosis of SPY/spy-4 in a sec-1 ga1 double mutant background produced pin-like structures32,36, which are reminiscent of the pid mutant32. This is in line with our work showing that SEC and SPY have multiple, overlapping functions in gynoecium development, including tissue specification and style elongation, presumably via the regulation of a plethora of targets, including SPT. It also suggests that the ratio of O-GlcNAc and O-fucose on target proteins might trigger very specific developmental outputs in vivo. It is possible to speculate that, during style development, different arrays of PTMs on SPT (and other targets), triggered by both genetic and environmental cues, might determine the selectivity of DNA targets and/or protein partners bound by SPT, to specify medial tissue identity5, coordinate the medio-lateral7 and adaxial–abaxial8 body axes and control the antagonistic auxin/CK balance, to establish radial symmetry6,11,58.
Taking the data as a whole, we propose a working model (Fig. 6d) in which O-glycosylation of SPT by O-GlcNAc and O-fucose is mediated by SEC and SPY, respectively, to regulate style development by supporting cell elongation and the SPT-mediated balance of the auxin/CK crosstalk at the gynoecium apex.
Methods
Plant materials and growth conditions
The following loss-of-function mutant lines in the ecotype Columbia (Col-0) background were used in this study: sec-2 (ref. 30), sec-5 (ref. 33), spy-3 (ref. 34), ind-2 (ref. 13) and spt-12 (ref. 59). The spy-4 mutant34, which was originally in the Wassilewskija background, was back-crossed to Col-0 three consecutive times, and homozygous segregating seeds were used for this study. The SPT::SPT–sYFP transgenic line was described previously41. The SPT::SPT–sYFP line in the Col-0 background was crossed with spt-12 to obtain the complementation line used in this study. Furthermore, the spt-12/SPT::SPT–sYFP line was crossed with sec-5 and spy-3 mutants. The plants were grown at 22 °C in long-day conditions (16 h light / 8 h dark) in controlled-environment rooms, unless otherwise specified.
Immunoprecipitation from Arabidopsis inflorescences
SPT–sYFP was immune-precipitated from young inflorescences of spt-12/SPT::SPT–sYFP (segregated control), spt-12;sec-5/SPT::SPT–sYFP sec-5 and spt-12;spy-3/SPT::SPT–sYFP plants. Young buds (close sepals) were collected from the inflorescences and immediately frozen in liquid N2 after picking; 5 g were used in each of the three biological replicates for each genotype. Each sample was ground in liquid N2 and extracted using 10 ml of buffer A as previously described48. After centrifugation at 16,000 g for 30 min, the supernatant was incubated with 30 μl of GFP-Trap magnetic beads (ChromTek). After 1 h of rotation at 4 °C, the beads were sedimented with a magnetic rack (GE Healthcare) for 1 min and washed with buffer A three times. Proteins on the beads were eluted with 1× loading buffer (Merck) and then separated by 10% SDS–PAGE and used for MS analysis.
MS and data processing
Gel slices were prepared according to standard procedures adapted from Shevchenko et al.60. Briefly, the slices were washed with 50 mM TEAB buffer, pH 8 (Sigma), incubated with 10 mM DTT for 30 min at 65 °C and then incubated with 30 mM iodoacetamide at room temperature (both in 50 mM TEAB). After washing and dehydration with acetonitrile, the gels were soaked with 50 mM TEAB containing 10 ng µl−1 Sequencing Grade Trypsin (Promega) and incubated at 40 °C for 8 h. The extracted peptide solution was dried down, and the peptides were dissolved in 0.1% TFA/3% acetonitrile. Aliquots were analysed by nanoLC–MS/MS on an Orbitrap Eclipse Tribrid mass spectrometer coupled to an UltiMate 3000 RSLCnano LC system (Thermo Fisher Scientific). The instruments were controlled by the Orbitrap Eclipse Tune Application v.3.4, Thermo Scientific Xcalibur v.4.4.16.14 and the Thermo Scientific SII for Xcalibur v.1.6.0.60983 software. The samples were loaded and trapped using a pre-column with 0.1% TFA at 15 µl min−1 for 4 min. The trap column was then switched in-line with the analytical column (nanoEase M/Z column, HSS C18 T3, 100 Å, 1.8 µm; Waters) for separation using the following gradient of solvents A (water, 0.1% formic acid) and B (80% acetonitrile, 0.1% formic acid) at a flow rate of 0.2 µl min−1: 0–3 min, 3% B; 3–10 min, increase B to 7% (curve 4); 10–70 min, linearly increase B to 37%; 70–90 min, linearly increase B to 55%, followed by a ramp to 99% B and re-equilibration to 3% B. Data were acquired with the following mass spectrometer settings in positive ion mode: MS1/OT, resolution 120 K, profile mode, mass range m/z 300–1,800, AGC 4e5, fill time 50 ms; MS2/IT, data-dependent analysis with the following parameters: 1.5 s cycle time in IT turbo mode, centroid mode, isolation window 1 Da, charge states 2–5, threshold 1e4, HCD CE = 33, AGC target 1e4, max. inject time auto, dynamic exclusion 1 count, 15 s exclusion, exclusion mass window ±10 ppm.
Alternatively (for the tobacco sample analysis), the following gradient of solvents A (water, 0.1% formic acid) and B (80% acetonitrile, 0.1% formic acid) was used at a flow rate of 0.2 µl min−1: 0–3 min, 3% B; 3–10 min, increase B to 7% (curve 4); 10–100 min, linearly increase B to 37%; 100–148 min, linearly increase B to 50%, followed by a ramp to 99% B and re-equilibration to 3% B. The mass spectrometer settings for those samples were: MS1/OT, resolution 120 K, profile mode, mass range m/z 300–1,800, AGC 4e5, fill time 50 ms; MS2/IT, data-dependent analysis with the following parameters: FAIMS device set to three compensation voltages (−35V, −50V and −65V) for 1 s each; MS2/IT: for each CV, data-dependent analysis with the following parameters: IT turbo mode, centroid mode, isolation window 1 Da, charge states 2–5, threshold 1e4, HCD CE = 30, AGC target 1e4, max. inject time dynamic, dynamic exclusion 1 count, 15 s exclusion, exclusion mass window ±10 ppm.
LC–MS analysis using EThcD fragmentation was performed on an Orbitrap Fusion Tribrid mass spectrometer (Thermo Fisher Scientific) using similar basic parameters as described above. Specific parameters for EThcD fragmentation included: targeted inclusion of the peptides of interest with ±25 ppm tolerance, ETD active with calibrated charge dependent parameters, EThcD true with hcd25, detector type: ion trap rapid.
The raw data from all acquisitions were processed in Proteome Discoverer v.2.4 or v.3.0 (Thermo Scientific); all mentioned tools of the following workflow are nodes of the Proteome Discoverer software. Spectra were recalibrated, and identification was performed using an in-house Mascot Server v.2.8.0 to v.2.8.2 (Matrixscience) with the TAIR10_pep_20101214 A. thaliana protein sequence database (arabidopsis.org, 35,386 entries) or the Nicotiana tabacum database61 to which the SPT–YFP fusion sequence was added. The MaxQuant contaminants database (maxquant.org, 245 entries) was included in the search. The parameters were enzyme trypsin, 2 missed cleavages, 6 ppm precursor tolerance, 0.6 Da fragment tolerance, carbamidomethylation (C) as fixed modification and oxidation (M), deamidation (N/Q), acetylation (protein N terminus), dHex (NST, +146.058 Da) and HexNAc (NST, +203.079 Da) as variable modifications. Evaluation was performed using Percolator with an FDR target of 0.01.
For processing the EThcD data, Proteome Discoverer v.3.0 with Mascot Server v.2.8.0 set to EThcD was used with a custom database containing the SPT sequence and common contaminants. The parameters were enzyme trypsin, 1 missed cleavage, 10 ppm precursor tolerance, 0.6 Da fragment tolerance, carbamidomethylation (C) as fixed modification and dHex (NST, +146.058 Da) and HexNAc (NST, +203.079 Da) as variable modifications. Evaluation was performed using Fixed Value PSM Validator with a maximum delta Cn of 0.05.
For all data, the Minora Feature Detector was used for peak detection and quantification with a minimum trace length of 7 and S/N 3. Peptide abundances were determined as peak areas. After normalization to the total peptide amount, the quantification was based on the top three unique peptides per protein. Missing values were imputed by low abundance resampling. For hypothesis testing, a background-based t-test was applied. The results were exported to Microsoft Excel. The percentage of the peptide modified with O-GlcNAc or O-fucose was calculated on the basis of the normalized abundances of the modified peptide compared with the sum of the abundances of all versions of the peptide.
For overviews and spectrum presentation, the MS raw files were converted to .mgf files using msconvert (Proteowizard v.3, https://proteowizard.sourceforge.io/index.html); these files were used for a database search via Mascot Server v.2.8.0 to v.2.8.2 using the abovementioned databases and Mascot search parameters. The Mascot search results were then imported into Scaffold v.4.11.0 (www.proteomesoftware.com).
DNA constructs
To produce the recombinant 6xHis–SPT protein, the full-length SPT coding sequence was cloned into the pRSFDuet vector by using the BamHI and PstI restriction sites. To produce recombinant 5TPR–SEC and 3TPR–SPY proteins, we used the following strategy. First, to generate a protein expression vector with the 10xHis–MBP tag, we amplified a 10xHis–MBP coding fragment by PCR with the primer pair 10xHis–MBP-F and 10xHis–MBP-R from the pMAL–c2X vector (Addgene) and cloned it into the pTrcHis vector (Addgene) using the XhoI and HindIII restriction enzymes, obtaining a pTrc10xHis–MBP vector. Second, code-optimized full-length SEC and SPY coding sequences were synthesized by Sangon Biotech (Shanghai). The coding sequences of 5TPR–SEC and 3TPR–SPY were amplified using the gene-specific primers pair 5TPR–SEC–opt_F/5TPR–SEC–opt_R and 3TPR–SPY–opt_F/3TPR–SPY–opt_R and cloned into the pTrc10xHis–MBP vector using the SacI and SalI restriction enzymes. All constructs were verified by sequencing and transformed into the Escherichia coli Rosetta (DE3) strain by heat shock.
For transient co-expression in tobacco, 35S::SPT–3xFLAG, 35S::SEC–3xHA and 35S::SPY–3xHA were constructed as follows: the full-length coding sequences of SPT, SEC and SPY were amplified using gene-specific primers (listed in Supplementary Table 3), and the PCR products were digested with SifI (NEB) and cloned into the empty pCambia1305–35S::3xFLAG or pCambia1305–35S::3xHA vector (gifted by Y. Ding), which was pre-digested with DraIII (NEB). All constructs were verified by sequencing and introduced into the Agrobacterium tumefaciens GV3101 strain for infiltration in N. benthamiana leaves.
For the split-luciferase complementation assay, the coding sequences of SPT, SEC and SPY were amplified by PCR using gene-specific primers (listed in Supplementary Table 3) and inserted into the pCambia1305–35S::nLuc and pCambia1305–35S::cLuc vectors (gifted by Y. Ding), respectively, using DraIII restriction sites, thus producing the 35S::SPT–nLuc, 35S::SEC–cLuc and 35S::SPY–cLuc constructs.
For the Y2H assay, full-length and domain fragments of coding sequences from SPT, SEC and SPY were cloned into pDONR221 (Invitrogen) using primers in Supplementary Table 3, and then recombined into either the pGDAT7 or pGBKT7 vector (Clontech) following the manufacturer’s instructions.
The DEX-inducible amiRNA constructs for SEC were generated as follows: specific amiRNA primers were generated using the online tool ‘Web MicroRNA Designer’ (WMD3) (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi)62. According to WMD3 guidelines, the amiRNAs were designed with uridine at position 1; in addition, one mismatch to the target gene was introduced in the 3′ part of the amiRNAs to reduce the likelihood that the amiRNA would act as a primer for RNA-dependent RNA polymerases to trigger secondary RNAi. Two specific regions, ‘ACGTGCAACCTTCTACACACC’ (named a1SEC) and ‘CTGGGCTCCCTGTAAAACGTG’ (named a2SEC), located in the 15th and 18th exons of the SEC gene, respectively, were chosen (Extended Data Fig. 5). Using pRS300 (ref. 62) as a PCR template, the fragments were amplified, and subsequently each fragment was inserted downstream of a DEX-inducible promoter to generate two pGAL6::amiRNA plasmids. The resultant plasmids were recombined with the 35S::GVG::Nos cassette (pICSL11041, Synbio TSL) and the Basta selection marker cassette (pICH41308, Synbio TSL), using the standard Golden Gate cloning method, to produce the binary vectors p35S::GVG–amiSEC–Basta–pICSL4723 (p35S::ami1SEC–GR and p35S::ami2SEC–GR). The vectors were transformed into the spy-3 mutant background.
To generate the transgenic SPT–YFP point-mutation lines analysed in this study, we amplified a 5-kb SPT promoter fragment from Col-0 genomic DNA using the primer pair pSPT-F and pSPT-R (see Supplementary Table 3 for the sequences). The SPT promoter was cloned into the pCambia1305 vector by using the EcoRI and BspEI restriction enzymes, thus producing the pSPT–pCambia1305 plasmid. The SPT wild-type genomic coding sequence fused in frame to the sYFP fluorescent tag (hence named SPT–YFP) was then amplified as a whole segment (2.5 kb) from genomic DNA extracted from original SPT::SPT–sYFP48 seedlings using the primer pair 2.5k_(BspEI)_F and 2.5k_(KpnI)_R and cloned into the pSPT–pCambia1305 plasmid, thus obtaining the pSPT::SPT–YFP–pCambia1305 construct.
To generate constructs harbouring specific point mutations of SPT, we used the pSPT::SPT–YFP–pCambia1305 plasmid as a template and used mutagenesis primers (listed in Supplementary Table 3) introducing specific mutations in the SPT coding sequence. All constructs were verified by sequencing and introduced into the A. tumefaciens GV3101 strain by heat shock, then transformed into the spt-12 background by floral dipping (note, since spt-12 produces few seeds, each construct was transformed in 30 mutant plants). The transgenic seeds were selected on Murashige and Skoog plates supplied with 15 mg l−1 Basta, and at least eight positive, independent T1 lines were screened using a confocal microscope for the presence of nuclear YFP signals in roots.
For the FRET–FLIM assay, the coding sequences of SPT and IND were first inserted into pCambia1305 using the DraIII sites; then, the coding sequences of EGFP and RFP were inserted in-frame using the XbaI and PstI restriction enzymes.
For the transactivation assay, we first cloned the GUS coding sequence into the pCambia1305 vector using the HindIII and BstEII sites, producing the pCambia1305–GUS construct. The 1-kb promoter region of PID was then cloned into the pCambia1305–GUS vector using the PstI and HindIII restriction sites. All constructs were verified by sequencing and introduced into the A. tumefaciens GV3101 strain for transformation in N. benthamiana leaves.
Recombinant protein expression
Cells were grown in LB media at 37 °C until the OD600 value reached 0.5; the culture was then cooled down before induction. The expression of 6xHis–SPT was induced by 0.4 mM IPTG at 25 °C for 4 h. The expression of 10xHis–MBP–5TPR–SEC and 10xHis–MBP–3TPR–SPY was induced by 0.4 mM IPTG at 16 °C for 16 h. Recombinant proteins were purified with nickel sepharose according to the manufacturer’s instructions (QIAGEN). The purified recombinant protein was dialysed overnight against dialysis buffer (20 mM Tris-HCl, pH 8.0, 50 mM NaCl, 1 mM DTT, 5% glycerol) at 4 °C. All proteins were aliquoted, flash-frozen in liquid N2 and stored at −80 °C.
In vitro O-glycosylation assay
The direct modification of SPT protein by SEC and SPY was tested by an in vitro glycosylation assay as previously described in Zentella et al.26. Briefly, to test O-GlcNAclation, a 50 ml reaction was carried out by mixing 10 mg of 6xHis–SPT, 5 mg of 5TPR–SEC, 20 mM Tris-HCl (pH 7.2), 12.5 mM MgCl2 and 200 μM UDP–GlcNAc (Merck). To test O-fucosylation, the 50 ml reaction contained 10 mg of 6xHis–SPT, 5 mg of 3TPR–SPY, 50 mM Tris-HCl (pH 8.2), 50 mM NaCl, 5 mM MgCl2 and 200 μM GDP–fucose (Merk). After incubation for 2 h at 25 °C, the protein samples were separated by 10% SDS–PAGE, and the band containing recombinant 6xHis–SPT (43 kDa) was excised and treated for MS analysis as described above.
Co-IP assay in tobacco leaves
A. tumefaciens GV3101 strains harbouring the 35S::SPT–3xFLAG, 35S::SEC–3xHA and 35S::SPY–3xHA constructs were transiently expressed either alone or co-infiltrated in four-week-old leaves of N. benthamiana. To enhance gene expression, an Agrobacterium strain harbouring P19 was always co-infiltrated. After 48 h from infiltration, the leaves were harvested, and the total protein was extracted using the protein extraction buffer (25 mM Tris-HCl, pH 7.4, 1 mM EDTA, 150 mM NaCl, 10% glycerol, 0.15% NP-40, 1 mM NaF, 10 mM DTT, 2% PVPP, 1 mM PMSF and 1× protein inhibitor). Total protein extracts were immunoprecipitated with anti-HA magnetic beads (Thermo Scientific); input and IP samples were analysed by immunoblotting using anti-HA (Sigma) and anti-FLAG antibodies (Sigma) separately.
Split-luciferase complementation assay in tobacco leaves
The split-luciferase complementation assay was performed as described previously55. A. tumefaciens GV3101 strains harbouring the 35S::SEC–cLuc and 35S::SPY–cLuc constructs were co-infiltrated with either 35S::SPT–nLuc or a 35S::Ø–nLuc empty vector, while 35S::SPT–nLuc was also co-infiltrated with a 35S::Ø–cLuc empty vector, into four-week-old leaves of N. benthamiana. To enhance gene expression, an Agrobacterium strain harbouring P19 was always co-infiltrated. After 48 h infection at room temperature, 0.4 mM d-luciferin (ThermoFisher) was infiltrated into the leaves, and LUC activity was measured using the NightOWL system equipped with a cooled CCD imaging apparatus (Berthold Technologies) and analysed with the Indigo software (v.2.0.5.0).
Y2H assay
For the Y2H experiments, coding DNA sequences were cloned into pDONR207 and recombined into pGDAT7 or pGBKT7 vectors (Clontech). Plasmids were transformed into the AH109 yeast strain by the lithium acetate method. Co-transformed strains were selected on SD/−Leu/−Trp (Merk) at 28 °C for three to four days. Transformed yeast were serially diluted (100, 10−1, 10−2 and 10−3) and dotted on SD/−Ade/−His/−Leu/−Trp and SD/−Ade/−His/−Leu/−Trp/ 2.5 mM 3-amino-1,2,4-triazole (3-AT, Merk) to examine protein interactions. Growth was observed after five days of incubation at 28 °C.
Gynoecium treatments
For the DEX treatments, 35S::amiSEC–GR inflorescences were sprayed with either 10 µM DEX (Sigma) or mock (DMSO, Sigma) three times every five days over two weeks, and gynoecia were fixed after five days from the last treatment. For the CK (6-benzylaminoadenine, BA) treatments, the inflorescences of Col-0, sec-2, sec-5, spy-3 and spy-4 were sprayed with either 50 µM or 100 µM BA (Merk) or mock (NaOH (200 mM)) three times every five days over two weeks, and gynoecia were fixed after five days from the last treatment. All spray treatments used a 0.015% final concentration of Silwet L-77.
SEM
Inflorescences were fixed overnight in FAA solution (3.7% formaldehyde, 5% glacial acetic acid and 50% ethanol). After complete dehydration through an ethanol series from 50% to 100%, the inflorescences were critical point dried using the Leica EM CPD300. Gynoecia were hand-dissected using a stereomicroscope (Leica S9D), coated with gold particles and examined with an FEI Nova NanoSEM 450 emission scanning electron microscope equipped with xT microscope control software (v.6.3.4), using an acceleration voltage of 3 kV. The experiments were conducted in biological triplicates, and gynoecia were taken from distinct inflorescences and plants each time. The total number of gynoecia analysed for each experiment is indicated in the figure legends. Data analysis was carried out using Microsoft Excel (v.2311).
Confocal microscopy
Confocal microscopy analysis was performed on a Zeiss LSM 880 confocal microscope with a ×40 water-immersion lens, using ZEN-Black-LSM880 software (v.2.3) (Zeiss). YFP signals were excited by an argon-514 nm, 10 mW solid laser with emission at 550–570 nm. To view the gynoecia, floral buds were dissected using a stereomicroscope (Leica S9D); then, the gynoecia were mounted along their longitudinal axis in water. To quantify the YFP fluorescence intensity, the z-series images of epidermal cells on the surface layer (L1) at the style region were collected with the z-step set at 1 μm. Maximum intensity projection of the z-series of the nucleus was used to quantify the YFP intensity using the software ImageJ (v.1.53).
RNA extraction and qRT-PCR
Three independent biological repeats of total RNA extracted from each genotype were isolated from young inflorescences using an RNeasy Plant Mini Kit (Qiagen), including treatment with RNase-free DNase (Qiagen) following the manufacturer’s instructions. Four micrograms of extracted RNA sample was reverse transcribed using M-MLV Reverse Transcriptase (Promega) from each RNA sample. The qRT-PCR experiments were performed in quadruplicates using SYGREEN BLUE qPCR MIX (PCRBIO) with a Chromo4 Real-Time PCR Detection System (Bio-Rad). Target gene expression levels were quantified by the 2−ΔΔct method with UBIQUITIN10 as the internal control using Bio-Rad CFX manager software (v.3.1) and Microsoft Excel (v.2311). The gene-specific primers are listed in Supplementary Table 3.
Western blot analysis
Protein samples were separated by 10% SDS–PAGE and transferred to a nitrocellulose membrane (GE Healthcare). After being blocked in 1× PBST buffer containing 5% skimmed milk, the membrane was incubated with the selected primary antibody using a 1,000-fold dilution overnight at 4 °C, washed three times with 1× PBST (10 min each) and incubated with the selected secondary antibody conjugated with horseradish peroxidase using a 3,000-fold dilution for 1 h at room temperature. After three washes with 1× PBST (10 min each), the film was illuminated and photographed with ImageQuant800 (GE Healthcare). AMERSHAM ImageQuot 800 software (v.1.2.0) was used for western blotting analysis; GeneSys software (v.1.3.8.0) was used for gel imaging. If the primary antibodies were already conjugated with horseradish peroxidase, there was no need to incubate the membrane with secondary antibodies. The following antibodies were used: anti-GFP (GF28R, 1:1,000 dilution, Thermo Scientific), anti-RFP (ab34771, 1:2,500 dilution, Abcam), anti-FLAG (F3165, 1:3,000 dilution, Sigma), anti-HA (3F10, 1:3,000 dilution, Sigma), anti-mouse (sc-516102, 1:1,000 dilution, Santa Cruz) and anti-rabbit (ab205718, 1:5,000 dilution, Abcam).
Seed germination assay
Freshly harvested seeds were sterilized and sowed on 0.9% agar plates supplemented with 10 µM DEX or mock (DMSO). The plates were grown in long-day (16 h light / 8 h dark) conditions. Seed germination was photographed from day 0 for three consecutive days, and the rate of germination was calculated as the ratio of germinated seeds over the total number of seeds plated, for each genotype every 24 hours. Data analysis was carried out using Microsoft Excel (v.2311).
Root length measurement
Seeds were sterilized and stratified in a dark, cold room for 48 h before being sowed on Murashige and Skoog plates supplemented with 10 µM DEX or mock (DMSO). The seed plates were grown vertically (with a slight angle) in short-day (8 h light / 16 h dark) conditions. The seedlings were imaged at day 7 after germination, and the primary root lengths were measured using ImageJ (v.1.53).
FRET–FLIM assay
FLIM images were captured using a Leica Stellaris 8 FALCON confocal microscope equipped with a ×40 water-immersion objective (HC Plan 40x/NA 1.10). The samples were excited with the 488 nm output of a pulsed white-light laser working at a 20 MHz repetition rate (Pulsed SuperK Fianium FIB-12 PP white-light laser from NK Photonics). The full width at half maximum of the laser pulse was ~170 ps, as determined from the instrument response function that was obtained by imaging an Erythrosin B (Sigma-Aldrich, >95% purity) water solution saturated with potassium iodide (Sigma-Aldrich). The EGFP fluorescence signal was collected in the 503–530 nm range using a Leica HyD X detector. Four-week-old N. benthamiana leaves were co-infiltrated with GV3101 strains to express SPT–EGFP with RFP–NLS or in combination with IND–RFP, HEC1–RFP and SPT–RFP, with or without SEC–3xHA and SPY–3xHA recombinant proteins. To enhance gene expression, an Agrobacterium strain harbouring P19 was always co-infiltrated. After 48 h, the leaves were cut and imaged on the microscope. Image acquisition and analysis was carried out using LAS X software (v.4.2, Leica Microsystems). At least 50 nuclei per condition from three biological replicates were analysed. A region of interest was hand-drawn around the nucleus, and the arrival times of all photons within the region of interest were used to generate a fluorescence decay histogram. The fluorescence decay was fitted using a bi-exponential decay function, and the amplitude-weighted average fluorescence lifetime63 obtained from the fit was used to compute the energy transfer efficiency, E, in accordance with E = 1 − τDA/τD (ref. 64), where τDA and τD are, respectively, the average fluorescence lifetimes of EGFP in the presence and absence of the FRET acceptor, RFP. The latter fluorescence lifetime, τD, was obtained from the negative control, SPT–EGFP::RFP–NLS.
Chromatin immunoprecipitation–qPCR assay
The chromatin immunoprecipitation assay was performed using young inflorescences (close sepals) of the SPT–sYFP spt-12 complementation line as previously described65. Three biological repeats were performed, using 3 g of plant material for each repeat. IP was conducted using GFP-Trap beads (ChromTek). The enrichment of the PID promoter regions was quantified using qPCR normalized with the ACTIN2 gene with the appropriate primers listed in Supplementary Table 3.
Transient transactivation assay
Transactivation assays were performed on four-week-old N. benthamiana Agrobacterium-infiltrated leaves using pPID::GUS alone, pPID::GUS combined with 35S::SPT–GFP and 35S::IND–RFP or together with 35S::SEC–3xHA, 35S::sec-2–3xHA, 35S::SPY–3xHA or 35S::spy-3–3xHA. To enhance gene expression, an Agrobacterium strain harbouring P19 was always co-infiltrated. All transformed bacteria were infiltrated using OD = 0.5, except 35S::SPT–GFP and 35S::IND–RFP (Fig. 6b), which were infiltrated using OD = 0.1. After 48 h from infiltration, the leaves were harvested, and the expression levels of GUS transcript were quantified via qRT-PCR; the hygromycin gene in the pPID::GUS vector was used as the internal control. Protein expression levels were detected by western blot. The primers used are listed in Supplementary Table 3.
EMSA
The EMSA reaction was performed as follows: 150 ng of amplified and gel-extracted wild-type or mutated DNA nucleotides together with the indicated concentration of recombinant 6His–SPT protein were incubated in 1× EMSA buffer (250 mM Tris-HCl, pH 8.0, 500 mM NaCl, 25% glycerol, 10 mM DTT) on ice for 20 min. To test the effect of O-glycosylation on SPT, recombinant SPT was first incubated with 5TPR–SEC or 3TPR–SPY recombinant proteins in reaction buffer for 1 h at 25 °C with 200 µM UDP–GlcNAc or GDP–fucose, respectively (enzymatic assay). Then, the DNA nucleotides and EMSA buffer were added to the reaction. The reaction was analysed by electrophoresis on 5% native acrylamide gel in 1× TBE buffer at 150 V for 50 min. After electrophoresis, the gels were stained with EB for 20 min followed by imaging with the UV imager (G:BOX F3 gel doc system, SYNGENE). The primers used for amplification of the wild-type and mutated PID 171-nucleotide fragments are listed in Supplementary Table 3.
The company names and catalogue numbers of the commercial reagents used in this study are summarized in Supplementary Table 4.
Statistical analysis
Statistical analysis was performed as indicated in each figure legend, using GraphPad Prism v.9 (Dotmatics). The exact P values are provided in the ‘Statistical source data1’ file.
Reporting summary
Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.
Data availability
All data needed to evaluate the conclusions in this paper are present in the paper and/or its supplementary materials. All raw proteomic data are available in the PRIDE repository (https://www.ebi.ac.uk/pride/; accession numbers: PXD037917 (in vivo HCD MS/MS from A. thaliana); PXD043987 (in vivo EThcD MS/MS from A. thaliana); PXD043957 (in vivo HCD MS/MS from N. benthamiana); and PXD044034 (in vitro HCD MS/MS from enzymatic assay)). The following databases were used in this study: TAIR10_pep_20101214 Arabidopsis thaliana protein sequence database (arabidopsis.org); the N. tabacum database61, to which the SPT–YFP fusion sequence was added; and the MaxQuant contaminants database (maxquant.org). All new genetic material (high-order mutants and transgenic lines) and expression vectors will be made available to the scientific community upon request and with no limitation. Source data are provided with this paper.
References
Simonini, S. & Østergaard, L. Female reproductive organ formation: a multitasking endeavor. Curr. Top. Dev. Biol. 131, 337–371 (2019).
Jiang, Y. & Moubayidin, L. Floral symmetry: the geometry of plant reproduction. Emerg. Top. Life Sci. 6, 259–269 (2022).
Zúñiga-Mayo, V. M., Gómez-Felipe, A., Herrera-Ubaldo, H. & de Folter, S. Gynoecium development: networks in Arabidopsis and beyond. J. Exp. Bot. 70, 1447–1460 (2019).
Ballester, P. et al. A transcriptional complex of NGATHA and bHLH transcription factors directs stigma development in Arabidopsis. Plant Cell 33, 3645–3657 (2021).
Heisler, M. G., Atkinson, A., Bylstra, Y. H., Walsh, R. & Smyth, D. R. SPATULA, a gene that controls development of carpel margin tissues in Arabidopsis, encodes a bHLH protein. Development 128, 1089–1098 (2001).
Moubayidin, L. & Ostergaard, L. Dynamic control of auxin distribution imposes a bilateral-to-radial symmetry switch during gynoecium development. Curr. Biol. 24, 2743–2748 (2014).
Larsson, E., Roberts, C. J., Claes, A. R., Franks, R. G. & Sundberg, E. Polar auxin transport is essential for medial versus lateral tissue specification and vascular-mediated valve outgrowth in Arabidopsis gynoecia. Plant Physiol. 166, 1998–2012 (2014).
Carabelli, M. et al. Coordination of biradial-to-radial symmetry and tissue polarity by HD-ZIP II proteins. Nat. Commun. 12, 4321 (2021).
Girin, T. et al. INDEHISCENT and SPATULA interact to specify carpel and valve margin tissue and thus promote seed dispersal in Arabidopsis. Plant Cell 23, 3641–3653 (2011).
Gremski, K., Ditta, G. & Yanofsky, M. F. The HECATE genes regulate female reproductive tract development in Arabidopsis thaliana. Development 134, 3593–3601 (2007).
Schuster, C., Gaillochet, C. & Lohmann, J. U. Arabidopsis HECATE genes function in phytohormone control during gynoecium development. Development 142, 3343–3350 (2015).
Groszmann, M., Paicu, T., Alvarez, J. P., Swain, S. M. & Smyth, D. R. SPATULA and ALCATRAZ, are partially redundant, functionally diverging bHLH genes required for Arabidopsis gynoecium and fruit development. Plant J. 68, 816–829 (2011).
Liljegren, S. J. et al. Control of fruit patterning in Arabidopsis by INDEHISCENT. Cell 116, 843–853 (2004).
Gambetta, M. C., Oktaba, K. & Müller, J. Essential role of the glycosyltransferase sxc/Ogt in polycomb repression. Science 325, 93–96 (2009).
Webster, D. M. et al. O-GlcNAc modifications regulate cell survival and epiboly during zebrafish development. BMC Dev. Biol. 9, 28 (2009).
Gambetta, M. C. & Müller, J. O-GlcNAcylation prevents aggregation of the Polycomb group repressor polyhomeotic. Dev. Cell 31, 629–639 (2014).
Xu, S. et al. The protein modifications of O-GlcNAcylation and phosphorylation mediate vernalization response for flowering in winter wheat. Plant Physiol. 180, 1436–1449 (2019).
Zentella, R. et al. O-GlcNAcylation of master growth repressor DELLA by SECRET AGENT modulates multiple signaling pathways in Arabidopsis. Genes Dev. 30, 164–176 (2016).
Wulff-Fuentes, E. et al. The human O-GlcNAcome database and meta-analysis. Sci. Data 8, 25 (2021).
Comer, F. I. & Hart, G. W. O-GlcNAc and the control of gene expression. Biochim. Biophys. Acta 1473, 161–171 (1999).
Xu, S.-L. et al. Proteomic analysis reveals O-GlcNAc modification on proteins with key regulatory functions in Arabidopsis. Proc. Natl Acad. Sci. USA 114, E1536–E1543 (2017).
Constable, S., Lim, J.-M., Vaidyanathan, K. & Wells, L. O-GlcNAc transferase regulates transcriptional activity of human Oct4. Glycobiology 27, 927–937 (2017).
Chen, Y.-X. et al. Alternative O-GlcNAcylation/O-phosphorylation of Ser16 induce different conformational disturbances to the N terminus of murine estrogen receptor beta. Chem. Biol. 13, 937–944 (2006).
Li, X. et al. O-linked N-acetylglucosamine modification on CCAAT enhancer-binding protein beta: role during adipocyte differentiation. J. Biol. Chem. 284, 19248–19254 (2009).
Gao, Y., Miyazaki, J.-I. & Hart, G. W. The transcription factor PDX-1 is post-translationally modified by O-linked N-acetylglucosamine and this modification is correlated with its DNA binding activity and insulin secretion in min6 β-cells. Arch. Biochem. Biophys. 415, 155–163 (2003).
Zentella, R. et al. The Arabidopsis O-fucosyltransferase SPINDLY activates nuclear growth repressor DELLA. Nat. Chem. Biol. 13, 479–485 (2017).
Sun, T.-P. Novel nucleocytoplasmic protein O-fucosylation by SPINDLY regulates diverse developmental processes in plants. Curr. Opin. Struct. Biol. 68, 113–121 (2021).
Shafi, R. et al. The O-GlcNAc transferase gene resides on the X chromosome and is essential for embryonic stem cell viability and mouse ontogeny. Proc. Natl Acad. Sci. USA 97, 5735–5739 (2000).
Sinclair, D. A. R. et al. Drosophila O-GlcNAc transferase (OGT) is encoded by the Polycomb group (PcG) gene, super sex combs (sxc). Proc. Natl Acad. Sci. USA 106, 13427–13432 (2009).
Hartweck, L. M., Scott, C. L. & Olszewski, N. E. Two O-linked N-acetylglucosamine transferase genes of Arabidopsis thaliana L. Heynh. have overlapping functions necessary for gamete and seed development. Genetics 161, 1279–1291 (2002).
Xiao, J. et al. O-GlcNAc-mediated interaction between VER2 and TaGRP2 elicits TaVRN1 mRNA accumulation during vernalization in winter wheat. Nat. Commun. 5, 4572 (2014).
Benjamins, R., Quint, A., Weijers, D., Hooykaas, P. & Offringa, R. The PINOID protein kinase regulates organ development in Arabidopsis by enhancing polar auxin transport. Development 128, 4057–4067 (2001).
Xing, L. et al. Arabidopsis O-GlcNAc transferase SEC activates histone methyltransferase ATX1 to regulate flowering. EMBO J. 37, e98115 (2018).
Jacobsen, S. E., Binkowski, K. A. & Olszewski, N. E. SPINDLY, a tetratricopeptide repeat protein involved in gibberellin signal transduction in Arabidopsis. Proc. Natl Acad. Sci. USA 93, 9292–9296 (1996).
Groszmann, M., Paicu, T. & Smyth, D. R. Functional domains of SPATULA, a bHLH transcription factor involved in carpel and fruit development in Arabidopsis. Plant J. 55, 40–52 (2008).
Hartweck, L. M., Genger, R. K., Grey, W. M. & Olszewski, N. E. SECRET AGENT and SPINDLY have overlapping roles in the development of Arabidopsis thaliana L. Heyn. J. Exp. Bot. 57, 865–875 (2006).
Silverstone, A. L. et al. Functional analysis of SPINDLY in gibberellin signaling in Arabidopsis. Plant Physiol. 143, 987–1000 (2007).
Eldridge, T. et al. Fruit shape diversity in the Brassicaceae is generated by varying patterns of anisotropy. Development 143, 3394–3406 (2016).
Gómez-Felipe, A. et al. Competing differentiation gradients coordinate fruit morphogenesis. Preprint at bioRxiv https://doi.org/10.1101/2023.01.19.524793 (2023).
Wendrich, J. R. et al. Framework for gradual progression of cell ontogeny in the Arabidopsis root meristem. Proc. Natl Acad. Sci. USA 114, E8922–E8929 (2017).
Moon, S. P., Javed, A., Hard, E. R. & Pratt, M. R. Methods for studying site-specific O-GlcNAc modifications: successes, limitations, and important future goals. JACS Au 2, 74–83 (2022).
Sorefan, K. et al. A regulated auxin minimum is required for seed dispersal in Arabidopsis. Nature 459, 583–586 (2009).
Simonini, S. et al. A noncanonical auxin-sensing mechanism is required for organ morphogenesis in Arabidopsis. Genes Dev. 30, 2286–2296 (2016).
Wild, M. et al. Tissue-specific regulation of gibberellin signaling fine-tunes Arabidopsis iron-deficiency responses. Dev. Cell 37, 190–200 (2016).
de Martin, X., Sodaei, R. & Santpere, G. Mechanisms of binding specificity among bHLH transcription factors. Int. J. Mol. Sci. 22, 9150 (2021).
Torres, C. R. & Hart, G. W. Topography and polypeptide distribution of terminal N-acetylglucosamine residues on the surfaces of intact lymphocytes: evidence for O-linked GlcNAc. J. Biol. Chem. 259, 3308–3317 (1984).
Steiner, E. et al. The putative O-linked N-acetylglucosamine transferase SPINDLY inhibits class I TCP proteolysis to promote sensitivity to cytokinin. Plant Physiol. 171, 1485–1494 (2016).
Liang, L. et al. O-fucosylation of CPN20 by SPINDLY derepresses abscisic acid signaling during seed germination and seedling development. Front. Plant Sci. 12, 724144 (2021).
Tarbet, H. J., Toleman, C. A. & Boyce, M. A sweet embrace: control of protein–protein interactions by O-linked β-N-acetylglucosamine. Biochemistry 57, 13–21 (2018).
Jumper, J. et al. Highly accurate protein structure prediction with AlphaFold. Nature 596, 583–589 (2021).
Radermacher, P. T. et al. O-GlcNAc reports ambient temperature and confers heat resistance on ectotherm development. Proc. Natl Acad. Sci. USA 111, 5592–5597 (2014).
Penfield, S. et al. Cold and light control seed germination through the bHLH transcription factor SPATULA. Curr. Biol. 15, 1998–2006 (2005).
Sidaway-Lee, K. et al. SPATULA links daytime temperature and plant growth rate. Curr. Biol. 20, 1493–1497 (2010).
Reymond, M. C. et al. A light-regulated genetic module was recruited to carpel development in Arabidopsis following a structural change to SPATULA. Plant Cell 24, 2812–2825 (2012).
Bi, Y. et al. SPINDLY mediates O-fucosylation of hundreds of proteins and sugar-dependent growth in Arabidopsis. Plant Cell 35, 1318–1333 (2023).
Josse, E.-M. et al. A DELLA in disguise: SPATULA restrains the growth of the developing Arabidopsis seedling. Plant Cell 23, 1337–1351 (2011).
Jacobsen, S. E. & Olszewski, N. E. Mutations at the SPINDLY locus of Arabidopsis alter gibberellin signal transduction. Plant Cell 5, 887–896 (1993).
Reyes-Olalde, J. I. et al. The bHLH transcription factor SPATULA enables cytokinin signaling, and both activate auxin biosynthesis and transport genes at the medial domain of the gynoecium. PLoS Genet. 13, e1006726 (2017).
Ichihashi, Y., Horiguchi, G., Gleissberg, S. & Tsukaya, H. The bHLH transcription factor SPATULA controls final leaf size in Arabidopsis thaliana. Plant Cell Physiol. 51, 252–261 (2010).
Shevchenko, A., Tomas, H., Havlis, J., Olsen, J. V. & Mann, M. In-gel digestion for mass spectrometric characterization of proteins and proteomes. Nat. Protoc. 1, 2856–2860 (2006).
Edwards, K. D. et al. A reference genome for Nicotiana tabacum enables map-based cloning of homeologous loci implicated in nitrogen utilization efficiency. BMC Genomics 18, 448 (2017).
Schwab, R., Ossowski, S., Riester, M., Warthmann, N. & Weigel, D. Highly specific gene silencing by artificial microRNAs in Arabidopsis. Plant Cell 18, 1121–1133 (2006).
Li, Y. et al. Investigations on average fluorescence lifetimes for visualizing multi-exponential decays. Front. Phys. 8, 576862 (2020).
Gadella, T. W. J. FRET and FLIM Techniques 560 (Elsevier Science, 2008).
Kuhn, A. & Østergaard, L. Chromatin immunoprecipitation (ChIP) to assess histone marks in auxin-treated Arabidopsis thaliana inflorescence tissue. Biol. Protoc. 10, e3832 (2020).
Acknowledgements
We thank N. E. Olszewski (University of Minnesota, USA) for providing the spy-3, spy-4 and sec-2 mutants; L. Xing (Institute of Botany, CAS, China) for the sec-5 mutant line; and D. Weijers (Wageningen University, the Netherlands) for the SPT translational fusion line SPT::SPT–sYFP. We thank J. Sklenar (Sainsbury Laboratory, Norwich, UK) for help with developing the ETD method and Y. Ding, S. Xu, Y. Dong, E. Tee and S. Mithoe (John Innes Centre) for help in setting up the Co-IP and split-luciferase complementation assays, the chromatin immunoprecipitation assay, the DEX inducible construct, the FRET–FLIM assay and tissue collection, respectively. We also thank L. Feng and T. Maxwell (John Innes Centre) for helping set up the EMSA assay and L. Wang (Institute of Botany, CAS, China) for useful suggestions for setting up the in vitro modification assay. We are grateful for the access to the John Innes Centre Bioimaging Facility including staff assistance and training, as well as the Horticultural Platform. We thank T.-p. Sun (Duke University, North Carolina, USA) for discussions and helpful comments on the manuscript. This work was supported by the Royal Society University Research Fellowship no. URF\R1\180091 (L.M.), the Royal Society Enhanced Research Expenses no. RF\ERE\210323 (L.M.), the Royal Society Research Fellows Enhancement Awards no. RGF\EA\181077 (L.M.), Biotechnology and Biological Sciences Research Council NRP Doctoral Training Programme Studentship no. BB/T008717/1 (L.M. and S.C.-F.), the British Society for Developmental Biology Gurdon summer scholarship (L.A.) and the Institute Strategic Programme grant (no. BB/X01102X/1) to the John Innes Centre from the Biotechnology and Biological Sciences Research Council.
Author information
Authors and Affiliations
Contributions
L.M. conceptualized the project. Y.J. and L.M. designed the experiments. Y.J. performed most of the experimental work with help from S.C.-F., S.W.H.K., I.J., B.G., L.A., S.G.L., C.M., G.S. and L.M. All authors analysed the data. Y.J. prepared the figures. L.M. wrote the paper, and all authors commented on and edited the paper.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Peer review
Peer review information
Nature Plants thanks Stefan de Folter, Beatriz Gonçalves and the other, anonymous, reviewer(s) for their contribution to the peer review of this work.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Extended data
Extended Data Fig. 1 In vivo proteomic analysis of spt/SPT-YFP complementation line inflorescences.
a, Confocal image (split-channels) of a stage-12 gynoecium apex of the spt-12/SPT::SPT-sYFP complementation line (hereafter use SPT-YFP to represent the translated protein product), which fully restored the split-style phenotype of spt-12 mutant. Note, expression of nuclear YFP signal is shown in the style. Scale bars indicate 500 μm. b, Quantification of SPT expression levels by qRT-PCR in inflorescences of Col-0, spt-12 and spt-12/SPT::SPT-sYFP complementation line. Values shown are means ± SD from three biological repeats. P values indicate two-tailed t test results. c-e, Representative HCD spectra of SPT pep_1 indicating modifications of residue Ser24 by both O-fucose (c) and O-GlcNAc (d) (zoom-in spectrum, lower panel) and residue Ser25 modified by O-fucose (e) (zoom-in spectrum, lower panel).
Extended Data Fig. 2 Quantification of SPT, SPY, and SEC expression patterns.
a, Illustration of the SEC, SPY and SPT gene structure indicating the T-DNA (sec-5, sec-2, spy-4, spt-12) and point-mutation (spy-3) lines and qRT-PCR primers (arrows F and R) used for these experiments. Scale bars represent 100 bp. b, Quantification of SEC and SPY expression by qRT-PCR in mutant inflorescences of the transferase enzymes sec-5 and spy-3. Values shown are means ± SD from three biological repeats. P values indicate two-tailed t test results. c, Quantification of SPT expression by qRT-PCR in mutant inflorescences of the transferase enzymes sec-5 and spy-3. Values shown are means ± SD from three biological repeats. P values indicate two-tailed t test results. d, z-stack confocal images of stage-10 gynoecia of SPT complementation line in sec-5 and spy-3 backgrounds compared to their respective segregating control, spt-12/pSPT:SPT:sYFP. Scale bars represent 100 μm. e, Quantification of the YFP signal intensity in nuclei of gynoecia of the genotypes depicted in panel (d). The box plots indicate the median (line within the box), the lower and upper quartiles (box), and the largest and smallest data points (whiskers). All data points are plotted in the graph. For spt-12/pSPT:SPT:sYFP in sec-5 background, n = 11 for control (spt-12) and n = 8 for spt-12 sec-5, respectively; For spt-12/SPT::SPT-sYFP in spy-3 background, n = 10 for both control (spt-12) and spt-12 spy-3. The P values indicate results of two-tailed t test.
Extended Data Fig. 3 SPT is modified by O-GlcNAc and O-fucose via SEC and SPY in vitro.
a, SDS-PAGE gel showing purified SEC, SPY and SPT recombinant proteins from E. coli. The arrows indicate the band of the proteins at the expected size, that is, 10xHis-MBP-5TPR-SEC (SEC), 112 kDa; 10xHis-MBP-3TPR-SPY (SPY), 107 kDa; and 6xHis-SPT (SPT), 44 kDa. The panel shows representative results obtained from at least three repetitions. b, Schematic representation of SPT protein domains and amino acid sequence (pep_1 and pep_2 are underlined in black) showing a summary of the residues found modified in vitro by O-fucose (red residue) and both modifications (O-GlcNAc and O-fucose, yellow residues) on pep_1, from HCD-MS experiments. The spectra indicate that modification by O-GlcNAc can occur on pep_2 around residues T71-T74 (blue underline). c, d, Representative HCD spectra indicating that SPT pep_1 is modified by O-fucose (c) and O-GlcNAc (d) (zoom-in spectrum, lower panel) on residue Ser23.
Extended Data Fig. 4 SEC and SPY interact with SPT via their N-terminal domains.
Yeast-two-Hybrid experiments to determine SPT-SEC (top panel) and SPT-SPY (bottom panel) protein-protein interactions. Full-length (FL) proteins and functional domains (TPRs and Cat) structures are depicted on the left panel. HEC1 (top panel) was used as negative and positive control to test interactions with SPT-AH domain and the SPT-C (Ac plus bHLH domains combined), respectively. The assays were performed using SEC and SPY as the bait (BD) and SPT as prey (AD). Growth on selection media SD-LWHA supplemented with 2.5 mM 3-AT indicates strong interactors. The dilutions examined in the experiments encompass ratios of 1:1, 1:10, 1:100, and 1:1000.
Extended Data Fig. 5 DEX-inducible SEC RNAi cloning strategy and seed germination rate analysis.
a, Schematic illustration of the working principle of the construct for DEX inducible RNAi of SEC. b, Prediction of highly specific amiRNA sequences for SEC gene (AT3G04240) in WMD3 (http://wmd3.weigelworld.org/cgi-bin/webapp.cgi). The top-ranking two sequences (a1 and a2, in red) were selected for analysis. HE and HE* indicate the hybridization energy of the amiRNA to a perfectly matching complement (HE, column 2) and to the target site in the target gene (HE*, column 4); identifier of the target gene (column 3). c, Schematic illustration of the a1 and a2 complementary sequences on the SEC mRNA (top), highlighting the mismatched nucleotides on the amiRNA guides, in red. Bottom, BLAST results for the a1SEC (bottom left) and a2SEC (bottom right) guide specificity in the NCBI database. The BLAST results show very high specificity for both sequences in Arabidopsis. d, Representative images of Col, spy-3, sec-5 and SEC RNAi spy-3 (a2SEC line 3) seeds germinated on either mock or DEX plates. e,f, Statistical analysis of the germination rate of Col, spy-3, sec-5 and SEC RNAi spy-3 (a2SEC line 3) seeds germinated on either mock (e) or DEX (f) plates. More than 300 seeds from four biological repeats for each genotype were analysed. Values shown are means ± SD. Different letters at the same timepoint indicate significant differences (P < 0.05) determined by ordinary one-way ANOVA (Tukey’s multiple comparison test). The exact P values are provided in the Statistical Source Data1 file. The colours of the letters are consistent with the genotype in the graph.
Extended Data Fig. 6 spy mutant gynoecia are hypersensitive to cytokinin treatments.
Representative SEM images of stage-13 gynoecia of Col-0, sec-2, sec-5, spy-3 and spy-4 genetic backgrounds after pharmacological treatment with cytokinin (either 50 μM or 100 μM BA) and mock. Scale bars represent 500 μm. The panels show representative results from one of three independent experiments.
Extended Data Fig. 7 Molecular characterization of spt-12/SPT-YFP S/T-to-A point mutation lines.
a, Comparison of SPT transcript levels in young inflorescences of three independent spt-12/SPT-YFP complementation (wild-type sequence) and point mutation lines, by qRT-PCR experiments. The experiment was performed once on three independent transgenic lines per construct, with four technical repeats. Values shown are means ± SEM. b, Quantification of SPT transcript levels in selected transgenic lines of spt-12/SPT-YFP complementation and point mutation lines depicted in the graph. Values shown are means ± SD from three biological repeats. The P values indicated comparison with wild-type by two-tailed student’s t test.
Extended Data Fig. 8 SPT is modified by O-GlcNAc and O-fucose via SEC and SPY in tobacco leaves.
a, Immunodetection of SEC-HA (35S::SEC-3xHA) and SPY-HA (35S::SPY-3xHA) from Agro-infiltrated N. benthamiana leaves. The CBB (Coomassie Brilliant Blue) bands were used as sample loading control. The data shows representative results from one of three independent experiments. b, Percentage of O-GlcNAc-modified, O-fucose-modified or unmodified SPT peptide_1 detected in HCD MS/MS experiments from tobacco leaves infiltrated with GV3101 strains carrying either construct of SPT alone or with SEC and SPY. The graph shows representative data generated from one of the three independent biological repeats. c, Table depicting the abundance ratio of SPT peptide_1 modified by O-GlcNAc (HexNac) and O-fucose (dHex), resulting from co-expression with SEC and SPY, in comparison to the abundance of the same peptide observed in SPT-alone expression. The colour scale indicates the range of ratio values.
Extended Data Fig. 9 Molecular interaction between SPT and PINOID promoter in vivo and in vitro.
a, Top, schematic representation of the PID promoter fragment including G-boxes (red lines) and E-box (blue line), and the position of the fragments amplified across those cis-elements. Bottom, qRT-PCR quantification of ChIP experiments carried out from young inflorescences of spt-12/pSPT:SPT:sYFP complementation line. The enrichment of the SPT binding to the region (R2) containing the two G-boxes demonstrate SPT binds in vivo to the PID promoter. Values shown are means ± SD from three biological repeats for R1 and R2; four biological repeats for R3. The P values indicate results of two-tailed t test. b, qRT-PCR quantification of GUS expression in transactivation assay from tobacco leaves infiltrated with either the pPID:GUS construct alone or with SEC-HA (35S::SEC-3xHA) or SPY-HA (35S::SPY-3xHA). OD = 0.5 was used for both SEC and SPY in each experiment. Note, SEC and SPY have no effect on pPID:GUS expression. Values shown are means ± SD from three biological repeats. The P values indicate results of two-tailed t test. c, qRT-PCR experiments showing quantification of GUS expression in transactivation assays from tobacco leaves following infiltration with IND-RFP (IND) and SPT-GFP (SPT) alone or both (IND + SPT) compared to the pPID:GUS construct alone. OD = 0.5 was used for both SPT and IND in each experiment. Values shown are means ± SD from three biological repeats. The P values indicate results of two-tailed t test. d, Graphic representation of the PID promoter including the 171-bp fragment used in EMSA experiments. Recombinant SPT (6xHis-SPT) binds to wild-type sequences of the G-box (red lines) present in the PINOID promoter, in a concentration-dependent manner (left panel). Incubation with a mutated (Mut) G-boxes (from CACGTG to TGATGA) abolish this binding. Similar results were obtained from two independent experiments.
Supplementary information
Supplementary Table 1
EThcD MS proteomics analysis results for peptide_1 in SPT (LISSSSSSSVYDTR) in vivo.
Supplementary Table 2
EThcD MS proteomics analysis results for peptide_2 in SPT (SSPLPSYYSPATTTTTASLIGVHGSGDPHADNSR) in vivo.
Supplementary Table 3
Sequences and purposes for all primers used in this study.
Supplementary Table 4
Company names and catalogue numbers of commercial reagents used in this study.
Source data
Source Data Figs. 1 and 3–6 and Extended Data Figs. 1, 2, 5 and 7–9
Statistical source data for Figs. 1 and 3–6 and Extended Data Figs. 1, 2, 5 and 7–9.
Source Data Figs. 2 and 6 and Extended Data Fig. 8
Unprocessed western blots and/or gels for Figs. 2 and 6 and Extended Data Fig. 8.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Jiang, Y., Curran-French, S., Koh, S.W.H. et al. O-glycosylation of the transcription factor SPATULA promotes style development in Arabidopsis. Nat. Plants 10, 283–299 (2024). https://doi.org/10.1038/s41477-023-01617-4
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1038/s41477-023-01617-4
- Springer Nature Limited