Brassinosteroid homeostasis is critical for the functionality of the Medicago truncatula pulvinus

Abstract

Many plant species open their leaves during the daytime and close them at night as if sleeping. This leaf movement is known as nyctinasty, a unique and intriguing phenomenon that been of great interest to scientists for centuries. Nyctinastic leaf movement occurs widely in leguminous plants, and is generated by a specialized motor organ, the pulvinus. Although a key determinant of pulvinus development, PETIOLULE-LIKE PULVINUS (PLP), has been identified, the molecular genetic basis for pulvinus function is largely unknown. Here, through an analysis of knockout mutants in barrelclover (Medicago truncatula), we showed that neither altering brassinosteroid (BR) content nor blocking BR signal perception affected pulvinus determination. However, BR homeostasis did influence nyctinastic leaf movement. BR activity in the pulvinus is regulated by a BR-inactivating gene PHYB ACTIVATION TAGGED SUPPRESSOR1 (BAS1), which is directly activated by PLP. A comparative analysis between M. truncatula and the non-pulvinus forming species Arabidopsis and tomato (Solanum lycopersicum) revealed that PLP may act as a factor that associates with unknown regulators in pulvinus determination in M. truncatula. Apart from exposing the involvement of BR in the functionality of the pulvinus, these results have provided insights into whether gene functions among species are general or specialized.

Nyctinasty is triggered by the pulvinus, a motor organ located at the base of the leaf and brassinosteroids is involved in functionality of pulvinus for leaf movement.

Introduction

Nyctinasty, a term used to describe the closure of leaves during the night, is common in Fabaceae, Maranthaceae, and Oxalidaceae species (Daphne and Osborne, 1998). It is assumed to be regulated by biological clock with a cycle of about 24 h (Moshelion et al., 2002). In this case, plants open their leaves during the day and close up at night, which has been described in detail by Charles Darwin after investigating more than 300 plant species (Darwin and Darwin, 1880).

The movement is triggered by the pulvinus, an organ located at the base of the leaf (or leaflet in the case of compound leaves; Satter et al., 1990). Pulvini are composed of a single vascular bundle surrounded by flexible, thin-walled parenchymatous cells, referred to as motor cells, which form an extensor layer and a flexor layer (Moran, 2007, 2015): flexor cells form on the adaxial surface of the pulvinus and extensor cells on the abaxial surface. Their cell volume is altered by a redistribution of ions (Moshelion et al., 2002; Oikawa et al., 2018; Ueda et al., 2019). During leaf movement, a number of molecular species are known to underlie nyctinasty (Ueda and Nakamura, 2007; Nakamura et al., 2011). As for the developmental determination of pulvinus, efforts have been made to understand the genetic basis for the motor organ. Mutants in which the pulvinus develops abnormally have been identified in Pisum sativum (apulvinic/apu), Lotus japonicus (sleepless/slp), and Medicago truncatula (petiolule-like pulvinus/plp; elongated petiolule1/elp1; Harvey, 1979; Kawaguchi, 2003; Chen et al., 2012; Zhou et al., 2012). In these mutants, the pulvinus is replaced by a petiolule-like structure, the motor cells being replaced by elongated petiole-like epidermal cells. APU, SLP1, and PLP/ELP1 are all LATERAL ORGAN BOUNDARIES domain (LBD) transcription factors, which are represented in the non-pulvinus forming species Arabidopsis thaliana by the LATERAL ORGAN BOUNDARIES (LOB; Shuai et al., 2002).

Brassinosteroids (BRs) are a group of phytohormones involved in the regulation of many plant developmental processes (Clouse and Sasse, 1998; Tong and Chu, 2018; Planas-Riverola et al., 2019; Nolan et al., 2020). BR-deficient or -insensitive mutants exhibit these defective phenotypes: dwarfism (Clouse et al., 1996; Szekeres et al., 1996), fused organ boundaries (Bell et al., 2012; Gendron et al., 2012), curled dark green leaves (Li et al., 1996), reduced apical dominance (Gonzalez-Garcia et al., 2011), and defective lamina joints (Sun et al., 2015). A large number of the enzymes involved in BR synthesis are encoded by members of the cytochrome P450 (CYP) gene family, such as DWARF4 (DWF4), CONSTITUTIVE PHOTOMORPHOGENESIS AND DWARFISM (CPD), and so on (Choe et al., 1998; Ohnishi et al., 2006; Ohnishi et al., 2012). The level of endogenous BR is regulated both by modulating the transcriptional activity of genes encoding proteins involved in the biosynthesis and by catabolism of BR. One of the BR inactivating genes has been identified in Arabidopsis, namely PHYB ACTIVATION TAGGED SUPPRESSOR1 (BAS1; Neff et al., 1999; Turk et al., 2003). BAS1 inactivates castasterone (CS) and brassinolide (BL; Turk et al., 2003), a form of BR recognized as highly active (Fujioka and Yokota, 2003). Moreover, genes involved in BR perception and subsequent signal transduction have been well investigated (Jaillais and Vert, 2016). The initiation of the signaling cascade is the perception of a BR molecule by a plasma membrane localized leucine-rich repeat (LRR) receptor-like kinase, BRASSINOSTERIOID INSENSITIVE 1 (BRI1), and many other signaling components such as the homologs BRI1-LIKE1 (BRL1), BRL3, and the coreceptor BRI1-ASSOCIATED KINASE1 (BAK1; Li and Chory, 1997; Li et al., 2002; Kinoshita et al., 2005).

BRs stimulate stem growth at low concentrations and inhibit it at high concentrations (Roddick et al., 1993; Clouse et al., 1996). Thus, BR homeostasis in specific tissues is critical for the development of plant organs. In Arabidopsis, LOB forms a feedback loop to orchestrate local BR accumulation during organ boundary formation via directly activating BAS1 transcription (Bell et al., 2012). PLP, as the putative ortholog of LOB in M. truncatula, has been shown to be required for the determination of the pulvinus, which is a distinct developmental event compared with Arabidopsis. If the pulvinus is considered to be the boundary between the leaflet and the petiole, then LOB and PLP define different boundaries (Zhou et al., 2012). These findings raise questions about whether the BR activity mediated by the LOB/BAS1 module is involved in pulvinus development and what is the mechanistic basis of the development of this organ. Here, the intention was to use M. truncatula knockout mutants of genes related to either BR metabolism or its signal perception to determine the dependence of pulvinus determination and functionality on BR homeostasis.

Results

Ectopic expression of PLP represses the elongation growth of plant organs

In M. truncatula, the adult leaves are in trifoliate form consisting of a terminal leaflet along with two lateral ones (Figure 1, A). The pulvinus is a cylindrical structure located at the base of each leaflet, which is responsible for leaflet movement (Figure 1, B–D). Scanning electron microscopy (SEM) images showed that the motor cells in the wild-type pulvinus had a highly convoluted surface (Figure 1, E), a morphology which differed markedly from that of petiole epidermal cells (Figure 1, F). In the plp-1 mutant, the pulvinus was replaced by a petiolule-like structure (Figure 1, G and H). The epidermal cells in this structure were smooth and elongated, resembling those in the petiole epidermis (Figure 1, I). These observations suggest that PLP plays a key role in the determination of pulvinus. To further investigate the role of PLP in pulvinus development, a GFP–PLP fusion driven by the cauliflower mosaic virus 35S promoter (35S:GFP-PLP) was introduced into wild-type plants (Supplemental Figure S1, A). The leaves developed by transgenic plants harboring the 35S:GFP-PLP construct were downward-curled (Figure 1, J) and the petioles and rachises were shorter than those formed by wild-type plants (Figure 1, K and L). The stature of transgenic plants was shorter than that of wild-type plants as a result of a reduction in internode length (Figure 1, M andSupplemental Figure S1, B and C). The obvious phenotypic changes in 35S:GFP-PLP plants indicate that PLP mainly functions in repressing the elongation of multiple plant organs.

Figure 1.

The phenotype of pulvinius in wild-type and phenotypic consequences of altering the expression of PLP. A and B, Fully expanded leaves (A) and close-up view of pulvinus (B) in 30-d-old wild type (WT). Arrows point to the pulvinus in (B). TL, terminal leaflet; LL, lateral leaflet; Ra, rachis; Pe, petiole; and Pu, pulvinus. C and D, The close-up view of pulvinus of terminal leaflet (C) and pulvinus of lateral leaflet (D). E and F, SEM images of pulvinus epidermal cells (E) and petiole epidermal cells (F). G and H, Fully expanded leaves (G) and close-up view of petiolule-like pulvinus (H, arrowed in [G]) in 30-d-old plp-1 mutant. Pet, petiolule-like pulvinus. I, SEM image of the petiolule-like pulvinus epidermal cells (boxed region in [H]). J, A fully expanded leaf of a transgenic plant harboring 35S: GFP-PLP. K–M, The length of the rachis (K), the petiole (L), the internode (M) of 60-d-old WT and 35S: GFP-PLP transgenic plants. Values shown in the form mean ± sd (n = 10). *, **, and *** means differ significantly (P < 0.05, < 0.01, < 0.001, t test). Bar = 5 mm in (A), (G), and (J), 1 mm in (B), 0.5 mm in (C), (D), and (H), and 10 µm in (E), (F), and (I).

PLP suppresses BR activity by activating MtBAS1

The short stature and curled leaf phenotype in PLP over-expressors resembled that of BR deficient or insensitive mutants, implying that PLP may play a role in activating BAS1 ortholog in M. truncatula. To verify this hypothesis, molecular interaction between LOB/PLP and the putative BAS1 orthologs was performed. First, a BLAST search identified Medtr4g113650 as the closest M. truncatula ortholog of BAS1 (Supplemental Figure S2), so this gene was hereafter referred to as MtBAS1. Second, PLP exhibited trans-activation activity, based on the activation of luciferase activity in protoplasts (Supplemental Figure S3). Third, the transient transformation of Arabidopsis mesophyll protoplasts with a plasmid containing the LUC reporter gene driven by either the BAS1 or MtBAS1 promoter and REN driven by the CaMV35S promoter, together with an effector plasmid harboring either PLP or LOB, indicated that both PLP and LOB are able to interact with both the BAS1 and the MtBAS1 promoters (Figure 2, A and B).

Figure 2.

Functional analysis of PLP and LOB in modulating accumulation of BR. A and B, A transient expression assay in A. thaliana protoplasts shows that both PLP and LOB activate BAS1 (A) and MtBAS1 (B). LUC activity was normalized by REN activity. Values shown in the form mean ± sd (n = 4). * and ** means differ significantly (P < 0.05, < 0.01, t test). C, Fully expanded leaves of WT and transgenics harboring either 35S: GFP-LOB or 35S: GFP-PLP. D and E, The abundance of MtBAS1 transcript in the leaves of 40-d-old transgenic plants harboring 35S: GFP-LOB (D) and 35S: GFP-PLP (E). Values shown in the form mean ± sd (n = 3). ** means differ significantly (P < 0.01, t test). F and G, Transcript abundance of MtDWF4, MtCPD1, and MtCPD2 in the leaflets of 40-d-old transgenic plants harboring 35S: GFP-LOB (F) and 35S: GFP-PLP (G) Values shown in the form mean ± sd (n = 3). ** and *** means differ significantly (P < 0.01, <0.001, t test). H, 30-d-old A. thaliana WT and transgenic plants harboring 35S: GFP-LOB or 35S: GFP-PLP. I, Leaves of the plants shown in (H). J, A model indicating the conserved roles of the PLP and LOB in negatively modulating accumulation of BR via the activation of BAS1 orthologs. Bar = 1 cm in (H), 5 mm in (C) and (I).

To further confirm this conserved regulatory mechanism between PLP and LOB in vivo, a 35S:GFP-LOB construct was introduced into M. truncatula (Supplemental Figure S4, A). Ectopic expression of LOB in M. truncatula resulted in a phenotype similar to that of transgenic plants harboring the transgene 35S:GFP-PLP (Figure 2, C andSupplemental Figure S4, B). A reverse transcription quantitative PCR (RT-qPCR) assay showed that the abundance of MtBAS1 transcripts was notably higher in both 35S:GFP-PLP and 35S:GFP-LOB transgenic plants than in wild-type plants (Figure 2, D and E). Previous studies reported that DWF4 and CPD are involved in BR biosynthesis and are used as the BR markers (Goda et al., 2004; Bai et al., 2007; Zhang et al., 2009). Thus, the orthologs of these genes were identified in M. truncatula. Two putative orthologs of CPD, MtCPD1, and MtCPD2 were obtained by a BLAST search, and both of them were expressed highly in leaves (Supplemental Figure S5, A and B). Moreover, two genes with a close relationship to DWF4 were identified. However, only Medtr5g020020 was expressed in leaves, thus, named MtDWF4 (Supplemental Figure S5, C and D). The expression level of these genes was measured and the results showed that MtCPD1 transcription was enhanced in the 35S:GFP-LOB plants, as were those of MtDWF4 and MtCPD1 in the 35S:GFP-PLP plants (Figure 2, F and G). The increased transcriptional level of these markers suggested a possible negative feedback regulation in M. truncatula due to the low endogenous BR level. To further investigate the conserved function of LOB and PLP, 35S:GFP-PLP and 35S:GFP-LOB were separately introduced into Arabidopsis (Supplemental Figure 6, A and B). The phenotype of Arabidopsis plants constitutively expressing either of the two transgenes was reminiscent of that generated in M. truncatula, namely a reduction in stature and the curling of its leaves (Figure 2, H and I); a further effect was the up-regulation of BAS1 (Supplemental Figure S6, C). Taken together, these data demonstrate that PLP, like LOB, suppresses the BR activity by activating MtBAS1, suggesting conserved roles of LOB and PLP across species (Figure 2, J).

BRs homeostasis affects the shape and function of the pulvinus

The leaflets of M. truncatula display nyctinastic movement with circadian rhythms (Zhou et al., 2012). The two lateral leaflets of the wild-type leaf start to move vertically at ZT13 (ZT: Zeitgeber time). The angle formed between the two lateral leaflets reaches ∼160° at ZT14, reaching complete closure by ZT22. The leaflets of 35S:GFP-PLP transgenics, however, failed to close (Figure 3, A and B), indicating that high expression level of PLP affected the normal function of the pulvinus in leaf movement. To investigate whether the development of pulvinus is defective due to the ectopic PLP expression, SEM analysis was performed. The results showed that the epidermal cells of pulvinus in both the wild-type and the transgenic plants formed a highly convoluted surface, which is characteristic of motor cells (Figure 3, C). However, the wild-type and transgenic pulvini differed from one another in that the cells formed in the latter were somewhat irregularly shaped (Figure 3, C) and the entire organ was shorter (Figure 3, D). These findings indicate that the proper expression level of PLP is critical for the shape and function of pulvinus. To further evaluate the effects of BR on pulvinus, the BR biosynthetic inhibitor (propiconazole, PPZ) was applied to wild-type plants (Figure 3, E and F). The effect of treating wild-type plants with PPZ was that PPZ reduced the size of the leaves and induced them to curl, mimicking the phenotype of the 35S: GFP-PLP transgenic plants (Figure 3, F). The shape of the motor cells formed by PPZ-treated plants was irregular (Figure 3, F), resulting in a reduction in the length of the pulvinus (Figure 3, E) and a failure of leaf movement (Supplemental Figure S7). These observations indicate that BR homeostasis exerts a strong influence over the shape and function of the pulvinus.

Figure 3.

The over-expression of PLP compromises the development of the pulvinus and inhibits nyctinastic leaf movement. A, A closed WT leaf and an open 35S: GFP-PLP transgenic leaf imaged at ZT22. B, The angle subtended between two lateral leaflets of 30-d-old WT and 35S:GFP-PLP transgenic plants. Values shown in the form mean ± sd (n = 10). Black bar indicates the dark condition and white bar indicates the light condition. C, SEM images of pulvinus epidermal cells in WT and 35S: GFP-PLP transgenic plants. D, The length of the WT pulvinus and its equivalent in a 35S: GFP-PLP transgenic plant. Values shown in the form mean ± sd (n = 10). * means differ significantly (P < 0.05, t test). E, The length of the WT pulvinus as affected by exposure to PPZ. Values shown in the form mean ± sd (n = 10). * means differ significantly (P < 0.05, t test). F, Adult WT leaf (left-hand panel), close-up view (center panel), and SEM image (right-hand panel) of the pulvinus of 20-d-old plants treated with 20 µM PPZ or DMSO as control. G, Measurement of endogenous BR level in pulvinus of WT and 35S:GFP-PLP, and in petiolule-like structure of plp-1 mutant. Values shown in the form mean ± sd (n = 3). *** means differ significantly (P < 0.001, t test). Bar = 5 mm in (A) and the left-hand panels of (F), 0.5 mm in the center panels of (F), 20 µm in (C), and the right-hand panels of (F).

To further evaluate the BR homeostasis, endogenous BR levels were measured in pulvinus of wild type and 35S:GFP-PLP, and in petiolule-like structure of plp-1 mutant (Figure 3, G). The results showed that 24-epibrassinolide (24-epiBL), and the precursors of BL, CS, and 6-deoxo-CS (6-deoxo-castasterone) were detected in these plant tissues, but typhasterol (TY) and BL were not. The levels of 24-epiBL, CS, and 6-deoxo-CS were significantly increased in plp-1 mutant, suggesting that PLP plays key role in repression of BR. However, BR level was not decreased in pulvinus of 35S:GFP-PLP plant as expected. To investigate the possible causes, the transcriptional level of PLP was measured (Supplemental Figure S6, D). The data showed that expression level of PLP in pulvinus of wild type was up-regulated about 400-fold compared with that in petiolule-like structure of plp-1 mutant. Meanwhile, PLP in pulvinus of 35S:GFP-PLP plant was up-regulated only six-fold compared with that in pulvinus of wild type. Additionally, expression level of MtBAS1 in pulvinus of wild-type and 35S:GFP-PLP plant was similar (Supplemental Figure S6, E). These data suggest that endogenous expression of PLP is very high in pulvinus of wild type. Thus, the exogenous expression of PLP in 35S:GFP-PLP plant essentially did not increase the repression of PLP on BR activity. However, ectopic expression of PLP indeed affects the BR compounds, as evidenced by the increased level of 24-epiBL, probably due to the feedback regulation of the BR biosynthetic pathway.

Isolation and characterization of the BR biosynthesis mutant mtdwarf4

Given that PLP represses BR activity and the BR level is low in pulvinus, it was intriguing to explore whether a low BR activity is able to rescue the petiolule-like structure of plp-1 mutant. To answer this question, BR biosynthesis-related mutants were isolated. The target chosen for mutagenesis in M. truncatula was the homolog of DWF4 in M. truncatula, MtDWF4, which was identified by a BLAST search (Figure 4, A andSupplemental Figure S5, C and D). MtDWF4 was highly expressed in flower, mature leaf, and pulvinus (Supplemental Figure S8, A). The spatial localization of MtDWF4 during leaf development was further examined by in situ hybridization analysis. The mRNA expression of MtDWF4 was detected in the pulvinus in terminal leaflet primordia at the P2 and P3 stages, and in lateral leaflet primordia at the P5 stage (Supplemental Figure S8, B). To determine the extent of functional equivalence between MtDWF4 and DWF4, CDS of MtDWF4 driven by the CaMV35S promoter was introduced into the Arabidopsis dwf4-102 mutant (Figure 4, B). The transgenic plants showed that MtDWF4 fully complemented the mutant phenotype of dwf4-102 (Figure 4, C). These data indicate that MtDWF4 plays a similar role as DWF4 in BR biosynthesis in M. truncatula.

Figure 4.

Molecular cloning and characterization of MtDWF4. A, Phylogenetic analysis of MtDWF4 with DWF4 genes from other species. The bootstrap values (%) presented at the branches were calculated from 1,000 replicates. The scale bar indicates the sequence divergence is 0.1 per unit bar, which represents 10% substitutions per nucleotide position. B and C, The phenotype of the dwf4-102 mutant (B), a dwf4-102 plant harboring 35S: MtDWF4 (C). D, The MtDWF4 sequence comprises eight exons (blocks) and seven introns (lines). The ATG start and TGA stop codons are indicated. E, A RT-qPCR assay of MtDWF4 transcription in WT and mtdwf4 mutant plants. Values shown in the form mean ± sd (n = 3). *, **, and *** means differ significantly (P < 0.05, < 0.01, < 0.001, t test). F and G, The appearance of the adult leaf of WT (F) and the mtdwf4-1 mutant (G). H, The phenotype of 70-d-old WT and mtdwf4-1 mutant plants. I–K, The length of the rachis (I), the petiole (J), and the internode (K) of 70-d-old WT and mtdwf4-1 mutant. Values shown in the form mean ± sd (n = 10). *, **, and ***: means differ significantly (P < 0.05, < 0.01, < 0.001, t test). Bar = 10 cm in (H); 5 mm in (C), (F), and (G); and 1 mm in (B).

Through both forward and reverse genetic screening of the retrotransposon Tnt1-tagged mutant population (Cheng et al., 2014), five independent mtdwf4 mutant lines were identified (Figure 4, D). mtdwf4-1, -2, -4, and -5 each experienced the insertion of a single tobacco Tnt1 retrotransposon, while mtdwf4-3 harbored an insertion of the native M. truncatula retrotransposon Mere1 (Rakocevic et al., 2009). In four of the mutants (mtdwf4-1, -2, -3, and -5), the insertion occurred within an exon (a different exon in each case), while the mtdwf4-4 insertion interrupted the final intron. A RT-qPCR assay measuring the abundance of MtDWF4 transcript in wild type and each of the mtdwf4 mutants was shown in Figure 4, E: very little transcript was produced by either the mtdwf4-1, -4, or -5 mutants, while the mtdwf4-2 and -3 mutants experienced only a modest reduction in transcript abundance. Sequence comparison indicated that the Tnt1 insertion in mtdwf4-2 caused the loss of the sixth and seventh exons of the transcript product of MtDWF4, resulting in a shift of the MtDWF4 open-reading frame. Additionally, the Mere1 insertion in mtdwf4-3 caused a 45-bp deletion in the seventh exon of MtDWF4, resulting in a truncated protein product. Thus, the expression level of MtDWF4 was slightly decreased in mtdwf4-2 and mtdwf4-3. All five mutants shared a similar phenotype, comprising downward-curled leaves (Figure 4, F and G), reduced stature (Figure 4, H), and shortened petioles, rachises, and internodes (Figure 4, I–K), reminiscent of the phenotypes observed in the 35S:GFP-PLP transgenic plants (Figure 1, J–M). In addition, transcript levels of MtCPD1 and MtCPD2 were up-regulated in mtdwf4 mutants due to negative feedback regulation (Supplemental Figure S8, C), suggesting a low BR level in MtDWF4 knockout mutants. Furthermore, the relationship between MtDWF4 and PLP was investigated at the transcriptional level. RT-qPCR data showed that expression level of PLP in pulvinus of mtdwf4-1 was similar to that of wild type, suggesting that transcription of PLP is not under the regulation of the MtDWF4-involved BR biosynthetic pathway (Supplemental Figure S8, D). Additionally, the expression level of MtDWF4 was decreased in petiolule-like structure of plp-1 mutant, compared with that in pulvinus of wild type, indicating that BR biosynthesis pathway is affected in plp-1 (Supplemental Figure S8, E).

Reducing BR compromises leaf movement, but does not rescue the defective pulvinus in the plp mutant

The lateral leaflets of the mtdwf4-1 mutant exhibited little nyctinastic movement: the angle subtended between them remained ∼170° throughout the day and night (Figure 5, A and B), which is similar to the 35S:GFP-PLP transgenic plants (Figure 3, B). A close-up view showed that the mutant’s pulvini were shorter than those formed in wild-type plants (Figure 5, C and G). Examination of pulvinus cross-sections failed to reveal any marked difference between the wild type and the mutant, both of which were composed of a central vascular bundle surrounded by parenchymatous cells (Figure 5, D and H), which is distinct from that in the petiole of wild type (Figure 5, K). However, SEM analysis revealed that the surface of the motor cells in the mutant was irregular and the cell boundaries were blurry (Figure 5, E, F, I, and J). Even so, they were still recognizable as motor cells, rather than epidermal cells of the wild-type petiole (Figure 5, L).

Figure 5.

Nyctinastic leaf movement is disrupted in the mtdwf4 mutant. A, Leaf movement in WT and mtdwf4-1 mutant plants at ZT14, ZT16, ZT17, and ZT18. B, The angle subtended between two lateral leaflets of 30-d-old WT and mtdwf4-1 mutant plants. Values shown in the form mean ± sd (n = 10). Black bar indicates the dark condition and white bar indicates the light condition. C–F, A close-up view (C), cross-section (D), SEM image (E), and a close view of SEM image (F) of the pulvinus present in WT. The sectioning region in (D) is shown in (C) by white line. The image in (E) is same with that in Figure 1, E. G–J, A close-up view (G), cross-section (H), SEM image (I), and a close view of SEM image (J) of the pulvinus present in WT. The sectioning region in (H) is shown in (G) by white line. K and L, A cross-section (K) and an SEM image (L) of petiole in WT. M–P, A adult leaf of plp-1 mutant (M) and a close-up view (N), cross-section (O), and SEM image (P) of the petiolule-like structure present in plp-1 mutant. The sectioning region in (O) is shown in (N) by white line. Q–T, A adult leaf of mtdwf4-1 plp-1 double mutant (Q) and a close-up view (R), cross-section (S), and SEM image (T) of the petiolule-like structure present in mtdwf4-1 plp-1 double mutant. The sectioning region in (S) is shown in (R) by white line. Bar = 1 cm in (A), (M), and (Q); 0.5 mm in (C), (G), (N), and (R); 0.1 mm in (D), (H), (K), (O), and (S); and 20 µm in (E), (F), (I), (J), (L), (P), and (T).

To test whether a reduction in BR is able to rescue the defective pulvinus seen in the plp-1 mutant, a mtdwf4-1 plp-1 double mutant was generated. Knockout of MtDWF4 in plp-1 mutant did not recover the petiolule-like structure, although its length was slightly reduced compared with that formed by the plp-1 mutant (Figure 5, M, N, Q, and R). Cross sections of the petiolule-like structure in plp-1 and mtdwf4-1 plp-1 showed a similar pattern of vascular bundle arrangement (Figure 5, O and S), which is also similar to that in the wild-type petiole (Figure 5, K). Moreover, SEM showed that the epidermal cells of the petiolule-like structure in mtdwf4-1 plp-1 were shorter than those in plp-1, but lost the features of motor cells (Figure 5, P and T). To further confirm this observation, plp-1 mutant was treated with PPZ to reduce the BR level (Supplemental Figure S9). When the plp-1 mutant was treated with PPZ, the leaves exhibited downward curling, but still formed a petiolule-like structure at the base of leaflets, demonstrating that PPZ application was unable to rescue the defective pulvinus in plp-1 mutant (Supplemental Figure S9). These observations suggest that the determination of the pulvinus is not influenced by BR level, but proper BR level is critical for the function of pulvinus.

Blocking BR signal perception does not affect the determination of the pulvinus

Knockout of MtDWF4 does not rescue the petiolule-like structure in the plp-1 mutant. One possibility is that other BR biosynthesis genes function redundantly, leading to the relatively weak BR deficiency phenotype in the mtdwf4-1 mutant. To test this hypothesis, the BR receptor that functions as the first step of BR signal perception was identified in M. truncatula. By searching the genome sequence database of M. truncatula, Medtr3g095100 was identified as MtBRI1, since it shows the greatest similarity with BRI1 (Figure 6, A and B and Supplemental Figure S10, A). Expression analysis of MtBRI1 indicated that the highest transcriptional level of MtBRI1 was detected in pulvinus (Supplemental Figure S10, B). Moreover, in situ hybridization analysis showed that the expression of MtBRI1 was detected in the pulvinus in terminal leaflet primordia at the P6 stage, instead of P3 and P4 stage, indicating its possible roles in development of pulvinus (Supplemental Figure S10, C). By a reverse genetic screening in a mutant population, one mtbri1 allele was isolated. The mutation was caused by the Tnt1 insertion at 476-bp downstream of the MtBRI1 start codon (Figure 6, A and C), which is identical to mtbri1-3 in a previous report (Cheng et al., 2017). The mutant was extremely dwarfed (Figure 6, D andSupplemental Figure S10, D); its leaves were thickened, curled, and dark green, and the size was greatly reduced; its petioles and rachises failed to elongate (Figure 6, E and F).

Figure 6.

The isolation and characterization of the mtbri1 mutant. A, The structure of the intronless gene MtBRI1. The ATG start and TGA stop codons are shown. B, Phylogenetic analysis of MtBRI1 with BRI1 genes from other species. The bootstrap values (%) presented at the branches were calculated from 1,000 replicates. The scale bar indicates the sequence divergence is 0.1 per unit bar, which represents 10% substitutions per nucleotide position. C, A RT-qPCR assay of MtBRI1 transcription in WT and mtbri1-3 mutant plants. Values shown in the form mean ± sd (n = 3). *** means differ significantly (P < 0.001, t test). D, The appearance of a 60-d-old mtbri1-3 mutant plant. E and F, The adaxial side (E) and the abaxial side (F) of the adult leaf of a 40-d-old mtbri1-3 mutant plant. G–J, A close-up view (G), cross-section (H), SEM image (I), and a close view of SEM image (J) of the pulvinus present in mtbri1-3 mutant. The sectioning region in (H) is shown in (G) by white line. K–N, A close-up view (K), cross-section (L), SEM image (M), and a close view of SEM image (N) of the pulvinus present in mtbri1-3 plp-1 double mutant. The sectioning region in (L) is shown in (K) by white line. Bar = 5 mm in (D), (E), and (F); 0.5 mm in (G) and (K); 0.1 mm in (H) and (L); and 50 µm in (I), (J), (M), and (N).

Additionally, nyctinastic leaf movement was completely abolished in mtbri1-3, even though a visible pulvinus was present (Figure 6, G), indicating the function of pulvinus is severely affected. The cross section and SEM showed that the mutant’s pulvini were coiled, including a single central vascular bundle (Figure 6, H) and its motor cells were irregularly shaped (Figure 6, I and J). These observations indicate that the pulvinus still determined in mtbri1-3, but the shape is abnormal. To further elucidate potential genetic interaction between MtBRI1 and PLP, a double mutant was generated. The leaves of the mtbri1-3 plp-1 double mutant were small and downward-curled (Supplemental Figure S10, E), and the pulvinus was replaced by a petiolule-like structure (Figure 6, K–N), essentially similar to that formed by the plp-1 mutant (Figure 5, M–P), indicating that blocking the BR signaling pathway does not rescue the defects in plp-1 pulvinus. These observations confirmed that a low level of BR is insufficient for determining the formation of pulvinus, but is required for pulvinus function.

Increasing endogenous BR activity affects the functionality of pulvinus, but not its development

PLP promotes the expression of MtBAS1 to repress the activities of BR in pulvinus. Therefore, by treating wild-type plants with bioactive BL, we further investigated whether increasing BR level in pulvinus is involved in its determination. When wild-type plants were supplied with BL, there was no visible phenotypic effect on the leaves, but the motor cells became elongated and the length of the pulvinus was increased (Figure 7, A and B). The leaf movement of BL-treated plants was abnormal; the lateral leaflets could not completely close between ZT18 and ZT22, compared control plants (Supplemental Figure S7). This observation indicates that BL affects the shape and function of pulvinus, but not its determination.

Figure 7.

The isolation and characterization of the mtbas1 mutant. A, Adult leaves, a close-up view and an SEM image of the pulvinus of 20-d-old WT plants treated with either 1 µM BL or DMSO. The pulvinus epidermal cells are shown in red in the BL-treated plants and in green in the DMSO-treated plants. B, The effect of BL treatment on the length of the pulvinus formed by WT plants. Values shown in the form mean ± sd (n = 10). **: means differ significantly (P < 0.01, t test). C, The MtBAS1 sequence comprises five exons (blocks) and four introns (line). The ATG start and TGA stop codons are shown. D, Phylogenetic analysis of MtBAS1 with BAS1 genes from other species. The bootstrap values (%) presented at the branches were calculated from 1,000 replicates. The scale bar indicates the sequence divergence is 0.1 per unit bar, which represents 10% substitutions per nucleotide position. E, A RT-qPCR assay of MtBAS1 transcription in WT and mtbas1 mutant plants. Values shown in the form mean ± sd (n = 3). ***: means differ significantly (P < 0.001, t test). F, The angle subtended between two lateral leaflets of 30-d-old WT and mtbas1-1 mutant plants. Values shown in the form mean ± sd (n = 10). G, The pulvinus formed by WT and mtbas1-1 mutant plants, and a cross-section and an SEM image of the mtbas1-1 pulvinus. The epidermal cells of pulvinus are shown in red. The sectioning region is shown by white line. H, The length of the pulvinus formed by WT and mtbas1-1 mutant plants. Values shown in the form mean ± sd (n = 10). **: means differ significantly (P < 0.01, t test). Bar = 10 mm in adult leaves in (A); 1 mm in close-up views in (A) and (G); 0.1 mm in cross-sections in (G); and 20 µm in SEM images of pulvinus in (A) and (G).

To test if increasing endogenous BR activity can affect the pulvinus, MtBAS1 was further characterized. First, the expression level of MtBAS1 was examined in different plant organs. The data showed that MtBAS1 was highly expressed in pulvinus (Supplemental Figure S11, A). RNA in situ hybridization revealed that the expression of MtBAS1 was detected in the pulvinus in terminal leaflet primordia at the P5, P7, and P9 stages (Supplemental Figure S11, B). Moreover, loss-of-function of PLP resulted in decreased expression of MtBAS1 in the petiolule-like structure of plp-1 mutant (Supplemental Figure S11, C). These findings are consistent with the positive regulation of MtBAS1 by PLP in pulvinus (Figure 2, J). Second, a reverse genetic screen succeeded in identifying two independent Tnt1 retrotransposon-induced mtbas1 mutants: the insertion sites were in the fourth exon (mtbas1-1) and fifth exon (mtbas1-2; Figure 7, C and D), and in both cases, the transcription of MtBAS1 was diminished (Figure 7, E). In addition, the transcript level of BR marker gene MtCPD2 was down-regulated in mtbas1-1 (Supplemental Figure S11, D), implying increased BR levels in the mutant. Third, leaf movement of mtbas1 mutant was investigated. Between ZT14 and ZT0.5, the angle subtended between the lateral leaflets was higher in mtbas1-1 plants than in wild-type ones (Figure 7, F), showing that the mutant’s nyctinastic leaf movement was compromised. The morphology of the mtbas1-1 pulvini, like those of wild type, comprised a single central vascular bundle and their epidermis surface was similarly convoluted (Figure 7, G), indicating that the pulvinus had developed. However, the length of the mutant’s motor cells and pulvinus was greater than in the wild type (Figure 7, G and H), which is consistent with the phenotypes observed in BL-treated plants (Figure 7, A and B). These findings indicate that increasing the level of BR activity by knocking out MtBAS1 cannot affect the determination of pulvinus; however, it does affect the functionality of motor organs.

PLP functions in a species-specific manner in pulvinus development

While Arabidopsis lacks a pulvinus, it does produce LOB, a protein which is very similar to PLP. Therefore, it was of interest to test whether LOB is able to determine the development of pulvinus in legumes. To address the question of whether differences in the expression patterns between LOB and PLP underlie the formation of a boundary in Arabidopsis or pulvinus in M. truncatula, promoter swap experiments were performed between two species. The GUS gene was fused to both the 2,201-bp LOB promoter (ProLOB:GUS) and the 2,411-bp PLP promoter (ProPLP:GUS), respectively, and each of the constructs was transformed into both Arabidopsis and M. truncatula (Figure 8, A–D). In Arabidopsis, both transgenes were expressed at the boundary of the cauline leaf axils (Figure 8, A and B), while in M. truncatula, both were strongly expressed in the pulvinus (Figure 8, C and D). The similar expression patterns between LOB and PLP indicate the comparable cis regulatory mechanisms between them. Meanwhile, these results imply that species-specific activities of trans regulatory factors are required for the correct expression pattern of LOB and PLP between Arabidopsis and M. truncatula. To investigate whether LOB is able to determine pulvinus formation, a series of transgenic experiments were performed. The plp-1 mutant was independently transformed with the two transgenes; those harboring ProPLP:PLP, but not those harboring ProLOB:PLP, restored the phenotype of plp-1 mutant (Figure 8, E–H), implying that the LOB promoter lacks a regulatory element(s) needed to promote a sufficient level of PLP expression in M. truncatula. This hypothesis was strengthened by the observation that in both Arabidopsis and M. truncatula, the ProPLP:GUS transgene generated a higher level of GUS activity than the ProLOB:GUS one (Figure 8, A–D). In addition, neither a ProLOB:LOB nor a ProPLP:LOB transgene was able to rescue the defective pulvinus in the plp-1 mutant (Figure 8, I and J). Moreover, ProLOB:PLP was introduced in the lob::DsE mutant which showed the fusion between branch and the subtending cauline leaf. The defects of lob::DsE mutant were completely rescued by ProLOB:PLP (Supplemental Figure S12), indicating that PLP and LOB play conserved roles in organ boundaries to limit growth within the boundary domain.

Figure 8.

Differences of the expression pattern and protein function between PLP and LOB. A–D, The expression of ProLOB: GUS (A, C) and ProPLP: GUS (B, D) in A. thaliana (A, B) and M. truncatula (C, D). E and F, Adult leaves of WT (E) and the plp-1 mutant (F). The image in (F) is same with that in Figure 1, G. G–J, Adult leaves of a plp-1 transgenic harboring the transgenes ProLOB: PLP (G), ProPLP: PLP (H), ProLOB: LOB (I), and ProPLP: LOB (J). K–N, Models of the LOB and PLP homodimers. The modeled surface of LOB (M) and PLP (N) protein structures from (K) and (L) shows their surface charge distribution. The surface is colored on the basis of electrostatic potential: positively charged surfaces appear in blue and negatively charged in red. N, N-terminal region; C, C-terminal region. Bracketed regions indicate the C-terminus. O and P, Adult leaves of WT tomato (O), and tomato harboring the 35S: PLP transgene (P). Q, A close-up view of the leaflets shown in the boxed region of (P); the arrow indicates the petiolule. Bar = 1 cm in (O)–(Q), 5 mm in (C)–(J), 2 mm in (A) and (B), and 0.5 mm in the close-up views shown in (C) and (D).

To investigate the functional difference between LOB and PLP, the homodimeric structure of LOB and PLP was predicted based on the crystal structure of LBD protein TtRa2LD from wheat (Chen et al., 2019), since homodimerization of the LOB domain is crucial for DNA binding and recognition of potential target sites. Both LOB and PLP mainly consist of a zinc finger, a GAS motif, a highly conserved DPVYG motif, and a leucine zipper-like motif (LZLM; Supplemental Figure S13, A). The output modeling results showed that the overall structure of LOB was similar to that of PLP (Figure 8, K and L). Moreover, the LBD family proteins possess a variable C-terminal region (Supplemental Figure S13, B). Surface charge potential analysis showed that the C-terminus of LOB presents a wider negative charge property compared with that of PLP (Figure 8, M and N). The LZLM motif in LBD family proteins is critical for protein dimerization (Shuai et al., 2002). Therefore, LOB and PLP probably interact with different partners in a species-specific manner. To investigate whether PLP is sufficient for pulvinus formation in non-leguminous species, 35S:PLP was introduced into tomato (Solanum lycopersicum). Similar to M. truncatula, the leaves in tomato are compound, and consist of leaflets and petiolules (Figure 8, O). The transgenic plants exhibited the BR-deficient phenotype (Figure 8, P andSupplemental Figure S14), which is similar to the 35S:GFP-PLP plants in Arabidopsis and M. truncatula (Figures 1, J, and 2, H and I). However, a pulvinus was not induced in 35S:GFP-PLP plants in tomato (Figure 8, Q). These results suggest that one or more yet unknown species-specific regulators cooperatively function with PLP to form a pulvinus.

Discussion

BR homeostasis is critical for the functionality of the pulvinus but not for its determination

BR homeostasis is essential for normal growth and development in land plants (He et al., 2005; Tanaka et al., 2005). BR-deficient mutants suffer characteristically from reduced stature, leaf curling, male sterility, and failure to adjust to darkness (Clouse et al., 1996). Treatment with a bioactive BR such as BL induces swelling of hypocotyl and roots (Wang et al., 2002; Zhao and Li, 2012). Here, the consequences of knocking out both the M. truncatula BR synthesis gene MtDWF4 and the BR biosignaling gene MtBRI1 were explored. Both mutants displayed a typical BR-deficient phenotype. When the BR inactivation gene MtBAS1 was induced by ectopically expressing either LOB or PLP, up-regulation of the BR biosynthesis genes MtDWF4 and MtCPD was observed. The inference is that endogenous BR levels in M. truncatula are regulated by a balance between biosynthesis and inactivation.

Cell elongation is regulated by a number of hormonal and environmental signals, the former including BR, a hormone synthesized in the same tissue in which it functions (Bishop et al., 1996). To control cell elongation therefore, plants need to monitor the BR level at the single cell level (Symons and Reid, 2004). As for the species-specific pulvinus, its outer cells are termed the “motor cells,” which are responsible for nyctinastic leaf movement through cellular volume changes (Cortizo and Laufs, 2012). Several lines of experimental evidence were provided here to demonstrate that the shape and function of the pulvinus are BR-dependent processes: first, both knocking out MtBAS1 and treating wild-type plants with BL lengthened the pulvinus, as a result of the elongation of the epidermal cells; second, knocking out MtDWF4 or MtBRI1 and treating wild-type plants with PPZ shortened the pulvinus. Nyctinastic leaf movement in each of the mtdwf4, mtbri1, and mtbas1 mutants was compromised. These results raise several questions concerning the regulation of BR in both shape determination and function of the pulvinus. How do plants sense the changes of BR concentrations in the pulvinus to control its shape? Does the defective leaf movement result from abnormal motor cells? Is BR directly involved in pulvinus function? Leaf movement in Samanea saman is known to be induced by the asymmetrical swelling/shrinking of the motor cells, a process facilitated by alterations in osmotic pressure, brought about by a change in the flow of ions through both anion and potassium ion channels (Oikawa et al., 2018). Endogenous BR is also involved in the regulation of potassium channel activity in guard cells (Inoue et al., 2017). Since the osmotic volume changes of motor cells are analogous to those of stomatal guard cells (Moran, 2007; Ueda and Nakamura, 2007), it is reasonable to suggest that the effects of BR on motor cell activity similarly reflect its regulation of anion and/or potassium ion channels.

While the normal pulvinus contains a single vascular bundle surrounded by the motor cells (Satter et al., 1990), the petiole features three vascular bundles. The pulvinus’ epidermal cells are highly convoluted, unlike those of the petiole. With respect to the pulvinus in M. truncatula, the PLP-MtBAS1 module governs the repression of BR activity. However, its determination is not dependent on its BR content, since the knocking out of either MtDWF4 or MtBRI1 in a plp mutant background failed to restore a pulvinus. According to SEM and light microscopic observations, the pulvinus in both mtdwf4 and mtbri1 mutants retains the characteristics of motor cells. In addition, knocking out MtBAS1 (unlike knocking out PLP) did not compromise the pulvinus. Taken together, these findings demonstrate that BR homeostasis is essential for the function, but not for the determination of the pulvinus (Figure 9).

Figure 9.

A proposed model for the functional comparison between M. truncatula PLP and Arabidopsis LOB. PLP and LOB are expressed in pulvinus and lateral organ boundary, respectively. The cis elements between LOB and PLP promoters are similar, while the species-specific activities of trans regulatory factor in Arabidopsis and M. truncatula determinate their correct expression patterns. Furthermore, BR activity is repressed in the pulvinus and lateral organ boundary by the conserved PLP-MtBAS1 and LOB-BAS1 regulatory module cross species. Low BR level is required for boundary formation in Arabidopsis. Meanwhile, BR homeostasis is critical for the functionality of pulvinus in nyctinastic leaf movement, but not for its determination. PLP may act as a factor that associates with unknown regulators to determine pulvinus formation in a species specific manner.

PLP functions in determining pulvinus formation in a species-specific manner

LBD proteins are key regulators of lateral organ development (Xu et al., 2016). LOB is induced by BR at boundary sites, and acts to repress BR activity by activating BAS1 (Bell et al., 2012). In M. truncatula, PLP is strongly expressed in the pulvinus, which is located in the boundary between the leaflets and the petiole. Functional homologs of PLP, in particular APU in pea and SLP in L. japonicus, similarly regulate pulvinus formation, implying functional conservation among legume species of the genetic network underlying this trait. Although the expression of PLP and LOB occurs in different types of boundaries, both proteins act to repress BR activity by activating BAS1 orthologs. While many UTR-associated cis regulatory elements are conserved across plant taxa (Vaughn et al., 2012), species-specific regulation has also been recognized (Srivastava et al., 2018). For example, the class I KNOX proteins regulate leaf shape in Arabidopsis (Piazza et al., 2010) but control the formation of leaflets in its close relative Cardamine hirsuta (Hay and Tsiantis, 2006). Since PLP and LOB both influence cell elongation and expansion by repressing BR activity, a series of promoter swap experiments was carried out. These showed that both ProLOB:GUS and ProPLP:GUS displayed the same expression pattern within either Arabidopsis or M. truncatula, but the expression pattern in each species was distinct (at the boundary of the cauline leaf axils in Arabidopsis or in the pulvinus in M. truncatula). The intensity of the GUS signal in ProPLP:GUS transgenic plants is higher than that in ProLOB:GUS transgenic plants, implying that variation in cis regulatory elements is responsible for controlling the expression level of transgenes. The proposition is therefore that the species-specific activity of trans regulatory factors is probably required to ensure the appropriate expression of LOB/PLP in the two species (Figure 9).

Conserved gene modules are thought to function in a context-dependent manner (Efroni et al., 2010; Hay and Tsiantis, 2010). Thus, any process regulated by BR and its regulation may be specific to certain cells in certain tissues of certain species (Tong and Chu, 2018). Although the PLP-MtBAS1/LOB-BAS1 modules appear to be well conserved between Arabidopsis and M. truncatula, PLP and LOB are involved in distinct developmental events. Attempts to rescue the plp-1 mutant using LOB and PLP promoter swap constructs showed that LOB and PLP are not functionally equivalent in pulvinus determination. However, ProLOB:PLP could rescue the defects in lob::DsE mutant. These observations imply that PLP and LOB play conserved roles in the formation of organ boundaries in Arabidopsis, but PLP may act as an associated factor in determining pulvinus formation in M. truncatula. Furthermore, the loss-of-function of PLP changed the pulvinus into a petiolule-like structure (this is also the case in other legume species). The implication is that a necessary prerequisite for the formation of a pulvinus is a capacity to form a petiolule. However, an attempt to induce the formation of a pulvinus in tomato (which forms petiolules) by constitutively expressing PLP produced a plant which exhibited symptoms of BR deficiency but failed to form a pulvinus. The conclusion is that one or more yet unknown species-specific regulators cooperatively function with PLP to form a pulvinus in M. truncatula.

LBD proteins are key integrators of development (Xu et al., 2016). A structural comparison between LOB and PLP reveals a clear difference in surface charge potential in the C-terminal region of these proteins. The C terminus of LBD, which includes the LZLM domain responsible for dimerization, is variable (Chen et al., 2019). An analysis of PLP’s upstream regulators, its protein partners, and downstream targets should reveal the molecular network underlying the PLP-dependent determination of the pulvinus in legumes, which will lead to better understanding of the genetic basis and evolution of the diversity observed in legumes and other species.

Materials and methods

Plant materials and growth conditions

Medicago truncatula ecotype R108 was used as the wild-type accession. All M. truncatula mutants were derived from of a Tnt1 retrotransposon-tagged mutant collection (Tadege et al., 2008). The M. truncatula plants were grown in the greenhouse with a 16-h photoperiod with 150 μmol/m²/s, a day/night temperature of 23°C/20°C and 70%–80% relative humidity. Arabidopsis (wild-type Col-0 and derived transgenics) plants were grown in a growth chamber with a constant temperature of 22°C and an 16-h photoperiod with a light intensity at 120 μmol/m²/s. Seeds of the Arabidopsis dwf4 allele dwf4-102 (Salk_020761) were obtained from TAIR (http://www.arabidopsis.org/). All mutants are listed in Supplemental Table S1.

Gene constructs

To generate constitutive expressors of LOB and PLP, the LOB and PLP coding sequences were amplified using the primer pairs, LOB-CDS-F/-R and PLP-CDS-F/-R, respectively. The amplicons were purified, then inserted into the pENTR/D-TOPO cloning vector (Invitrogen), and subsequently into the pEarleyGate 103 vector via attLX attR recombination reactions (Invitrogen). The ProPLP:LOB transgene comprised the 561-bp LOB open-reading frame driven by the 2,411-bp PLP promoter, while the ProLOB:LOB transgene comprised the same LOB sequence driven by the 2,201-bp LOB promoter and the ProLOB:PLP transgene comprised the 579-bp PLP open-reading frame driven by the LOB promoter. All three constructs were transferred into the pBGWFS7 vector (Karimi et al., 2002) using a Gateway LR reaction (Invitrogen). To generate the ProPLP:GUS and ProLOB:GUS transgenes, the 2,411-bp PLP promoter or the 2,201-bp LOB promoters were amplified and the amplicons were cloned into pBGWFS7. To generate the MtDWF4 overexpression construct, the CDS of MtDWF4 was amplified and cloned into pENTR/D-TOPO cloning vector (Invitrogen) using the primers MtDWF4-CDS-F and MtDWF4-CDS-R, then transferred into the pEarleyGate 100 vector (Earley et al., 2006) by attLXattR recombination reactions (Invitrogen). Primer sequences used in this study are listed in Supplemental Table S2.

Stable plant transformation

Transgene constructs were introduced into disarmed Agrobacterium tumefaciens EHA105 cells using the freezing/heat shock method. The protocol for transforming M. truncatula was as described previously (Cosson et al., 2006). Arabidopsis plants were transformed using the floral dip technique (Clough and Bent, 1998). All transgenic plants are listed in Supplemental Table S3.

BL and PPZ treatments

BL and PPZ were first dissolved in DMSO and then diluted in water to 1 and 20 μM, respectively. Seeds of wild type M. truncatula and the plp-1 mutant were surface-sterilized by immersion in 2% w/v (weight/volume) NaClO for 30 min, plated on solidified half strength MS medium, and held at 4°C for 7 d. Successfully germinated seeds were selected and transferred to a fresh plate in which the medium was supplemented with BL or PPZ; the seedlings were grown for 20 d in a growth chamber with a 16-h photoperiod and a constant temperature of 23°C. Pulvinus lengths were estimated from digital images using the ImageJ software (https://imagej.nih.gov/ij/).

Determination of endogenous BRs levels

Three biological replicates of pulvinus in wild-type and 35S:GFP-PLP plants and the petiolule-like structure in plp-1 mutant were collected for BRs measurements. The quantification of endogenous BRs levels was performed based on the method reported previously with some simplifications in sample pretreatment (Xin et al., 2013). The harvested plant materials were first ground to a fine powder. Two hundred milligrams of the powder was extracted with of 90% aqueous methanol (MeOH) in an ultrasonic bath for 1 h. Simultaneously D₃-BL, D₃-CS, D₃-6-deoxo-CS, and D₃-TY were added to the extract as internal standards for BRs content measurement. After the MCX cartridge was activated and equilibrated with MeOH, water, and 40% MeOH in sequence, the crude extracts reconstructed in 40% MeOH were loaded onto the cartridge. The MCX cartridge was washed with 40% MeOH, and then BRs were eluted with MeOH. After being dried with N₂ stream, the eluent was redissolved with ACN to be derivatized with 2-methoxypyridine-5-boronic acid (MPyBA) prior to UPLC-MS/MS analysis. BRs analysis was performed on a quadrupole linear ion trap hybrid MS (QTRAP 6500, AB SCIEX) equipped with an electrospray ionization source coupled with a UPLC (Waters; Xin et al., 2020). As for BL, D₃-BL, CS, D₃-CS, 6-deoxo-CS, D₃-6-deoxo-CS, TY, and D₃-TY, the MRM transition 598.3 > 178.1, 601.3 > 178.1, 582.4 > 178.1, 585.4 > 178.1, 568.4 > 178.1, 571.4 > 178.1, 566.4 > 548.3, and 569.4 > 548.3 was used for quantification.

GUS staining and SEM analysis

GUS activity in transgenic leaf material was detected as described by Zhou et al. (2011). For SEM, tissue samples were fixed overnight in 3% v/v (volume/volume) glutaraldehyde in 25 mM phosphate buffer (pH 7.0), dehydrated with an ethanol series, then dried. SEM images were captured using a TM-3000 device (Hitachi) delivering an accelerating voltage of 15 kV.

In situ hybridization analysis

For RNA in situ hybridization, the probe fragments of 526-bp MtDWF4 CDS, 457-bp MtBRI1 CDS, and 543-bp MtBAS1 CDS were amplified using primers listed in Supplemental Table S2, then ligated into pGEM-T vector (Promega, USA). The sense and anti-sense probes were labeled with digoxigenin-11-UTP (Roche, Switzerland) and used for hybridization. In situ hybridization was performed with vegetative buds of 4-week-old wild-type plants as described previously (Zhou et al., 2011).

RNA extraction, RT-PCR, and RT-qPCR

At least three biological replicates of leaves or pulvini of 40-d-old wild-type and mutant plants were harvested for the purpose of assessing the transcription of key genes. The procedures used to isolate, purify and reverse-transcribe RNA, and then subject the cDNA to either reverse transcription PCR or RT-qPCR following the manufacturer’s instructions (Invitrogen). The relevant primer sequences are given in Supplemental Table S2. The RT-qPCR was executed on CFX Connect (Bio-Rad) using Roche SYBR-green fluorescence dye (FastStart Essential DNA Green Master Kit). The relative expression levels of genes were calculated using 2^−△△CT method with MtUBI gene as the internal control.

Protein structure

The structures of PLP and LOB were predicted from the crystalline structure of the wheat homolog TtRa2LD (Chen et al., 2019) using routines implemented in the Swiss-model database (Biasini et al., 2014). Structural illustrations were generated using the PyMOL v.2.3.2 software (https://pymol.org/).

Phylogenetic analysis

Alignment of multiple sequences was performed using the ClustalW software (https://www.ebi.ac.uk/Tools/msa/clustalw2/) with default parameters. Neighbor-joining phylogenetic trees were constructed using the routines implemented in the MEGA6 software suite (https://www.megasoftware.net/) imposing 1,000 bootstrap replicates.

Transient gene expression assays

Protoplasts from 21-d-old Arabidopsis seedlings were prepared as described by Lee et al. (2008). Isolated Arabidopsis mesophyll protoplasts were transfected with plasmid DNA using polyethylene glycol (PEG). LUC activity was measured using a Centro XS LB960 (Berthold).

Accession numbers

Sequence data from this article can be found in the National Center for Biotechnology Information/GenBank and Medicago truncatula Hapmap Project under the following accession numbers: LOB (At5G63090), PLP (Medtr3g077240), MtDWF4 (Medtr5g020020), MtBAS1 (Medtr4g113650), BAS1 (At2G26710), MtCPD1 (Medtr5g082520), MtCPD2 (Medtr3g070380), GmDWF4a (XP_003517233), GmDWF4b (XP_003538851), CcDWF4 (XP_020218337), VaDWF4 (XP_017426057), ApDWF4 (XP_027360511), CaDWF4 (XP_004512038), PsDWF4 (BAF56239), AtDWF4 (At3G50660), OsDWF4 (XP_015633105), ZmDWF4 (APQ46083), TpBRI1 (PNY09948), CaBRI1 (XP_004502878), CcBRI1 (XP_020232732), ApBRI1 (XP_027332115), PsBRI1 (BAC99050), GmBRI1a (NP_001345944), GmBRI1b (NP_001237411), AtBRI1 (At4G39400), OsBRI1 (XP_015621030), ZmBRI1a (NP_001309780), ZmBRI1b (AKG58825), SlBRI1 (XP_004237477), OsCYP734A2 (XP_015623670), OsCYP734A4 (XP_015641577), OsCYP734A5 (XP_015648047), OsCYP734A6 (XP_015631278), CaCYP734A1 (XP_004508536), PsCYP734A (BAF56240), ApCYP734A1 (XP_027368795), LaCYP734A1a (XP_019439804), LaCYP734A1b (XP_019452180), ZmCYP734A1 (ONM29579), ZmCYP734A8 (ACG45585), GmCYP734A1a (XP_003549781), GmCYP734A1b (XP_003524531), and GmCYP734A1c (XP_003528297).

Supplemental data

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

Supplemental Figure S1 . The identification and phenotype of 35S: GFP-PLP transgenic plant.

Supplemental Figure S2 . Phylogenetic analysis of BAS1 in M. truncatula and Arabidopsis.

Supplemental Figure S3 . Transactivation analysis of PLP in Arabidopsis protoplasts.

Supplemental Figure S4 . The phenotype of 35S:GFP-LOB transgenic plants.

Supplemental Figure S5 . Phylogenetic analyses and expression patterns of MtCPD and MtDWF4.

Supplemental Figure S6 . Expression level of BAS1 in wild type and transgenic plants of Arabidopsis.

Supplemental Figure S7 . Leaf movement in wild type under the treatment of PPZ and BL.

Supplemental Figure S8 . Characterization of MtDWF4 and its loss-of-function mutant.

Supplemental Figure S9 . The phenotype of adult leaf in plp-1 treated with PPZ.

Supplemental Figure S10 . Characterization of MtBRI1 and its loss-of-function mutant.

Supplemental Figure S11 . Characterization of MtBAS1 and its loss-of-function mutant.

Supplemental Figure S12 . PLP rescues lob::DsE mutant phenotype.

Supplemental Figure S13 . Alignment and motif organization of PLP and LOB proteins.

Supplemental Figure S14 . The phenotypes resulting from overexpressing PLP in tomato.

Supplemental Table S1. List of mutant alleles

Supplemental Table S2. Primers used in this study

Supplemental Table S3. List of the number of transgenic lines

 

Y.K. and C.Z. conceived the study, designed the experiments, and wrote the manuscript. Y.K., Z.M., H.W., Y.W., Y.Z., M.W., L.H., R.L., J.Z., and L.H. performed the research. K.S.M. and J.W. contributed new reagents/analytic tools. P.X. and J.C. measured the BRs level. Y.K., Z.M., M.B., X.Y., F.K., Y.K., J.W., and C.Z. analyzed the data, provided the critical discussion on the work, and edited the manuscript.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is: Chuanen Zhou (czhou@sdu.edu.cn).