Brassinosteroid signaling integrates multiple pathways to release apical dominance in tomato

Significance

For almost a century, auxin had been well-known as the master regulator of apical dominance. Recently, however, sugars were shown to be the initial regulator of apical dominance, while strigolactones (SLs) and cytokinins (CKs) act downstream of auxin to control bud outgrowth. However, the interactions of the different pathways have remained outstanding questions. Here, we report that brassinosteroids (BRs) are essential for the release of apical dominance in tomato. CK signaling relays information from auxin, SL, and sugars to promote the production of BRs, which activate the BZR1 transcription factor to suppress the expression of BRANCHED1, an inhibitor of bud outgrowth. These findings demonstrate that hormonal and metabolic pathways impinge on a common BR signaling for controlling shoot branching.

Abstract

The control of apical dominance involves auxin, strigolactones (SLs), cytokinins (CKs), and sugars, but the mechanistic controls of this regulatory network are not fully understood. Here, we show that brassinosteroid (BR) promotes bud outgrowth in tomato through the direct transcriptional regulation of BRANCHED1 (BRC1) by the BR signaling component BRASSINAZOLE-RESISTANT1 (BZR1). Attenuated responses to the removal of the apical bud, the inhibition of auxin, SLs or gibberellin synthesis, or treatment with CK and sucrose, were observed in bud outgrowth and the levels of BRC1 transcripts in the BR-deficient or bzr1 mutants. Furthermore, the accumulation of BR and the dephosphorylated form of BZR1 were increased by apical bud removal, inhibition of auxin, and SLs synthesis or treatment with CK and sucrose. These responses were decreased in the DELLA-deficient mutant. In addition, CK accumulation was inhibited by auxin and SLs, and decreased in the DELLA-deficient mutant, but it was increased in response to sucrose treatment. CK promoted BR synthesis in axillary buds through the action of the type-B response regulator, RR10. Our results demonstrate that BR signaling integrates multiple pathways that control shoot branching. Local BR signaling in axillary buds is therefore a potential target for shaping plant architecture.

Shoot architecture has significant impacts on light capture, photosynthesis, and resource allocation. Moreover, it is important in ensuring reproductive success and crop productivity (1–3). Optimizing shoot architecture is thought to be a promising approach for increasing crop yield to meet the challenges of an increasing global population and climate change (2). Shoot branching is mainly determined by the formation of axillary buds in the leaf axils, together with subsequent outgrowth or dormancy. The production of axillary buds is controlled genetically (4), while the outgrowth of axillary buds is controlled by a range of factors and displays a high level of plasticity in response to sugar availability, hormonal signals, and changing environment conditions (5, 6). Multiple pathways converge on a common transcription factor named BRANCHED1 (BRC1) to control bud outgrowth in different plant species (7). BRC1 is expressed specifically in axillary buds and acts as a central regulator that inhibits bud outgrowth (8, 9).

The phytohormone regulatory network plays a principal role in regulating the outgrowth of axillary buds. The growth of buds is inhibited by auxin and strigolactones (SLs) and is promoted by cytokinins (CKs). To date, several hormone-based models have been used to explain the control of bud outgrowth. The canalization model proposes that auxin synthesized in the young leaves of shoot apex is transported basipetally through the main stem in a polar manner to inhibit shoot branching indirectly by competing for access to the main polar auxin transport stream (10, 11). The second messenger model proposes that CKs and SLs act downstream of auxin to control shoot branching. In this model, local CK synthesis in the nodal stem is sufficient to promote axillary bud outgrowth, whereas auxin in the polar transport stream suppress CK synthesis by inhibiting the expression of Isopentenyltransferase (12). SLs are synthesized mainly in roots and transported acropetally to inhibit shoot branching (13–15). SLs synthesis is mediated by the sequential action of Carotenoid Cleavage Dioxygenase 7 (CCD7), Carotenoid Cleavage Dioxygenase 8 (CCD8) and the cytochrome P450 named More Axillary Growth1 (MAX1) (16). Notably, auxin signaling regulates the expression of SL synthesis genes (17, 18).

Recently, sugars were shown to be major players in the regulation of bud outgrowth. Sugar levels in the dormant buds increase when the buds start to grow before any changes in the auxin content are observed in the adjacent stem (19). However, internode elongation may inhibit shoot branching by diverting sugars away for the growing buds (20). Hence, genes regulating stem elongation, especially those related to gibberellin (GA) signaling, influence shoot branching (21, 22). Little information is available, however, concerning how hormone signaling and metabolic cues are linked to the regulation of BRC1.

Brassinosteroids (BRs) are a group of steroid hormones that play critical roles in plant growth and development (23). BR binding to the receptor Brassinosteroid Insensitive1 (BRI1) triggers a phosphorylation cascade that results in the inactivation of the negative regulator Brassinosteroid Insensitive2 (BIN2), leading to dephosphorylation and activation of the transcription factors BRASSINAZOLE RESISTANT1 (BZR1) and BRI1 EMS SUPPRESSOR1 (BES1) (24). Dephosphorylated BZR1 (dBZR1) can regulate the expression of a variety of genes involved in plant growth and development as well as stress responses (25–27). BR also promotes shoot branching. For example, the tissue-specific expression of the BR synthesis genes CYP724B and CYP90B increases tiller number in rice (28), whereas BR synthesis and signaling mutants have lower tiller numbers than the wild-type (WT) rice (29, 30). To date, relatively few studies have considered the mechanisms whereby BR regulates shoot branching. Recent studies have demonstrated that BES1 or BZR1 forms a complex with D53-like proteins, that are regulators of SL signaling, to inhibit the expression of BRC1 in Arabidopsis or FC1 in rice (31, 32). However, the precise role of BR synthesis and signaling in the control of shoot branching remains to be defined, and little is known about the interactions between BR and other hormone and sugar pathways in the shoot branching regulatory network.

Here, we show that the transcription factor BZR1 mediates BR to promote axillary bud outgrowth in tomato through the direct suppression of BRC1. Bud activation in response to a depletion of auxin/SLs levels or by treatment with CKs/sucrose is dependent on an increase in BR signaling. Conversely, enhanced GA signaling results in suppression of shoot branching, together with a decrease in BR and BZR1 levels. Auxin, SL, sucrose, and GA were found to regulate CK accumulation, which promoted the expression of BR biosynthesis gene through the type-B response regulator (RR), RR10. These findings demonstrate that multiple hormone and sugar signals interact to regulate shoot branching through a common BR–BZR1–BRC1 signaling cascade.

Results

BR Biosynthesis and Signaling Promote Shoot Branching in Tomato.

We first determined whether BR regulates bud outgrowth in tomato. The dwf mutant that is defective in the BR biosynthesis gene DWARF (DWF) had significantly fewer lateral buds than the WT, while overexpression of DWARF (OE-DWF) promoted bud outgrowth (Fig. 1 A and B). Mutation of the BR signaling gene BZR1 (bzr1) suppressed bud outgrowth, while overexpression of BZR1 (OE-BZR1) significantly promoted bud outgrowth. Virus-induced gene silencing (VIGS) of homologs of BIN2, which is a negative regulator of BR signaling, revealed that BIN2.2 is a suppressor of bud outgrowth in tomato (SI Appendix, Fig. S1). The BIN2.2 loss-of-function mutant showed significant increases in total lateral bud length. Together with these bud outgrowth phenotypes, the levels of transcripts encoding BRC1, which is a central regulator suppressing shoot branching, were increased in the dwf and bzr1 mutant buds and decreased in the buds of the bin2.2 mutant as well as those of the OE-DWF and OE-BZR1 plants (Fig. 1C). In agreement with these findings, histochemical analysis of β-glucuronidase (GUS) expression driven by BRC1 promoter (P_BRC1) showed that the signal was greater in the dwf mutant than the WT (Fig. 1D). In situ hybridization studies showed that the levels of BRC1 messenger RNA in the meristem and leaf primordia of the axillary buds were higher in the dwf mutant than the WT, whereas BRC1 transcripts were barely detectable in the OE-DWF plants (Fig. 1E). Taken together, these results indicated that the growth of lateral buds is promoted by increasing the expression of BR biosynthesis and signaling genes in tomato.

Fig. 1.

BR signaling regulates bud outgrowth in tomato. (A and B) Bud outgrowth phenotypes of BR biosynthesis (dwf) and signaling (bzr1, bin2.2) mutants and transgenic plants overexpressing BR biosynthesis gene DWF and signaling gene BZR1. (C) qPCR analysis of relative transcript of BRC1 in axillary buds. (D) Histochemical analysis of GUS expression driven by P_BRC1. (E) In situ hybridization of messenger RNA of BRC1 in axillary buds. (F) IAA content in nodal stems and roots. (G) Relative content of orobanchol, didehydro-orobanchol, and solanacol in roots. (H) CKs content in nodal stems of dwf mutant, WT, and OE-DWF plants. Data are presented as the means of three replicates ± SD. Different letters indicate significant differences (P < 0.05) according to Tukey’s test.

BR-Induced Bud Outgrowth Is Not Attributable to Changes in the Levels of Auxin, SLs, and CKs.

Given the roles of auxin, SLs, and CKs in bud growth, we next examined whether BR promotes bud outgrowth by altering the levels of auxin, SLs, and CKs in tomato plants. The dwf mutant showed lower levels of indole-3-acetic acid (IAA) in nodal stems and roots than the WT, while OE-DWF plants showed increased IAA levels in roots without differences in the IAA levels in the stems (Fig. 1F). Consistent with the observations of root IAA levels, the expression of the auxin-responsive reporter gene DR5::GUS was decreased in the root tips of the dwf mutant and increased in the OE-DWF roots (SI Appendix, Fig. S2A). Compared to the WT, the expression of CCD7, CCD8, and MAX1 was suppressed in the roots of the dwf mutant and increased in those of the OE-DWF plants (SI Appendix, Fig. S2B). In addition, the content of solanacol and didehydro-orabanchol was decreased in the dwf roots but increased in those of the OE-DWF plants (Fig. 1G). There were no significant differences in the orabancol contents that accompanied the changes in the BR levels in different lines. In contrast to SLs, the levels of Isopentenyltransferase2 (IPT2) transcripts in the nodal stems were inversely correlated with BR levels (SI Appendix, Fig. S2B). While OE-DWF plants did not show any changes in the CK content of the nodal stems, the accumulation of the bioactive CKs isopentenyladenosine (iP) and trans-zeatin (tZ) and of the ribosides isopentenyladenosine riboside (iPR), trans-zeatin riboside (tZR), and dihydro-zeatin riboside (DHZR) was significantly increased in the dwf mutant (Fig. 1H). As bud outgrowth is promoted by CKs and suppressed by auxin and SLs, it is unlikely that BR promotes bud outgrowth through impacts on IAA, SL, and CK homeostasis in tomato plants.

BR Signaling Directly Regulates the Expression of BRC1.

Since the dwf and bzr1 mutants showed increased BRC1 expression in the lateral buds, we next determined whether BR signaling releases apical dominance by regulating the expression of BRC1. The brc1 mutant generated by CRISPR/Cas9 showed a significantly increased total lateral bud length. However, knocking out BRC1 in the OE-DWF plants did not further increase total lateral bud length (Fig. 2 A and B), suggesting that BRC1 may be involved in BR-induced bud outgrowth. Silencing of BRC1 released buds from dormancy in the dwf and bzr1 mutants (Fig. 2 C and D) and resulted in a similar total lateral bud length in the bzr1 mutant (but not the dwf mutant) to the WT. Furthermore, silencing of DWF or BZR1 inhibited bud outgrowth in the WT but did not affect the total lateral bud length of the brc1 mutant (SI Appendix, Fig. S3). These results indicate that BRC1 acts downstream of BR signaling to regulate bud outgrowth.

Fig. 2.

BR signaling is involved in bud outgrowth through transcriptional regulation of BRC1 by BZR1. (A and B) Bud outgrowth phenotypes of WT, brc1 mutant, OE-DWF plants, and OE-DWF plants in the brc1 background (OE-DWF/brc1). (C and D) Effects of silencing of BRC1 on the bud outgrowth phenotypes of dwf and bzr1 mutants. (E) Y1H analysis of BZR1 binding to the P_BRC1. (F) ChIP-qPCR analysis of BZR1 binding to the P_BRC1. P1 to P5 represented the DNA fragments containing the E-box in the P_BRC1. The asterisk indicated statistical difference between WT and OE-BZR1 plants (Student’s t test, P < 0.05). (G) Dual-luciferase assay for the regulatory effect of BZR1 on the expression of BRC1. WT and mutated P_BRC1 were used for the assay. The ratio of LUC/REN of the EV plus promoter was set as one. Data are presented as the means of three replicates ± SD. Different letters indicate significant differences (P < 0.05) according to Tukey’s test.

Next, we examined whether BZR1 directly regulates the expression of BRC1. In the yeast one-hybrid (Y1H) assay (Fig. 2E), the yeast cells containing the P_BRC1-bait vector and the pGADT7-BZR1 vector grew on the SD/Leu media containing aureobasidin A (AbA). In contrast, transformants without BZR1 failed to grow on this media. There are five fragments (P1 to P5) containing E-box that are the potential binding sites for BZR1 in P_BRC1 (Fig. 2F). Chromatin immunoprecipitation (ChIP)-qPCR analysis indicated that BZR1 bound to the P4 fragment of the P_BRC1. Dual-luciferase assay showed that the activity of P_BRC1 was inhibited by 67% as a result of BZR1 binding (Fig. 2G), whereas a mutation of the E-box in the P4 fragment abolished the regulatory effects of BZR1 on BRC1 without changing the basal promoter activity. Collectively, these findings demonstrated that BZR1 binds to the P_BRC1 at the E-box motif and then suppresses BRC1 expression. This down-regulation of BRC1 ultimately promotes bud outgrowth.

BR Signaling Is Required for Release of Apical Dominance in Buds.

To study the role of BR signaling in apical dominance, we analyzed the response of lateral buds to decapitation, a traditional means to release apical dominance. Decapitation led to a rapid increase in the levels of transcripts of DET2 and DWF that encode BR synthesis enzymes. This increase was observed within 3 to 6 h of decapitation, while the levels of CYP90B1 and CPD transcripts were either decreased or unaltered, respectively (Fig. 3A). Decapitation also resulted in the suppression of the expression of BR inactivation gene CYP734A8 and the negative regulator signaling genes BIN2.2 and BIN2.3 in the buds. In addition, the levels of brassinolide (BL), the active end product of BR synthesis, were significantly increased in buds and nodes after decapitation (Fig. 3B). Moreover, an increase in the accumulation of the BZR1 protein, particularly the active form dBZR1, was observed in buds after decapitation (Fig. 3C). Treatment of buds with 24-epibrassinolide resulted in a similar increase in dBZR1 levels (SI Appendix, Fig. S4). Together, these findings suggest that the release from apical dominance is associated with enhanced BR signaling in buds.

Fig. 3.

BR signaling is required for the release from apical dominance. (A) Changes in transcript of BR biosynthesis and signaling genes in axillary buds in response to decapitation. (B) Changes in BL in axillary buds and nodal stems in response to decapitation. (C) Changes in relative accumulation of dBZR1, the active form involved in gene regulation, in axillary buds in response to decapitation. Signal intensity of dBZR1 protein was analyzed with ImageJ. The dBZR1 levels of control plants were set as one. (D and E) Effects of decapitation on the total lateral bud length of mutants in BR biosynthesis (dwf) and signaling (bzr1) or WT plants silenced for BR receptor gene BRI1. (F and G) qPCR analysis of the relative transcript of BRC1 in axillary buds. Data are presented as the means of three replicates ± SD. Different letters indicate significant differences (P < 0.05) according to Tukey’s test.

After decapitation, the total length of lateral buds was increased in the WT and in the empty vector (EV) based on Tobacco Rattle Virus (TRV) plasmid (Fig. 3 D and E and SI Appendix, Fig. S5). However, the suppression of lateral bud growth was not completely relieved by decapitation in the dwf and bzr1 mutants or the TRV–BRI1 plants. The release from apical dominance in the WT and TRV plants was associated with the suppression of BRC1 expression (Fig. 3 F and G). However, the levels of BRC1 transcripts remained higher in the buds of the dwf, bzr1, or TRV–BRI1 plants than the WT or pTRV plants after decapitation. These results indicate that BR plays a crucial role in apical dominance.

BR Signaling Is Involved in Hormone- and Sucrose-Regulated Shoot Branching in Tomato.

Since auxin, SLs, CKs, and sugars are involved in apical dominance, we next studied whether BR mediates hormone and sugar signals for controlling the bud growth. Silencing of FLOOZY (FZY), which is an auxin synthesis gene, and cosilencing of CCD7 and CCD8 (CCD7/8) genes that are required for SL biosynthesis resulted in a significant increase in total length of lateral buds in the WT. Similarly, the application of 6-benzylaminopurine (6-BA, a synthetic CK) to buds at a concentration of 50 μM or the application of sucrose to leaves at a concentration of 20 mM promoted lateral bud growth in the WT (Fig. 4 A and B and SI Appendix, Figs. S6 and S7). In contrast to these observations, only minor to moderate increases in the total length of lateral buds were observed in the dwf and bzr1 mutants after silencing of FZY or CCD7/8 or the application of 6-BA or sucrose. While the expression of BRC1 was suppressed in the WT plants after silencing FZY or CCD7/8, or the application of 6-BA or sucrose, high levels of BRC1 transcripts were observed in the dwf and bzr1 mutants, regardless of the loss of FZY and CCD7/8 functions or the application of 6-BA and sucrose (Fig. 4 D and E). Decreasing auxin and SLs levels via VIGS or the application of 6-BA or sucrose resulted in enhanced accumulation of DET2 and DWF transcripts (SI Appendix, Figs. S6 and S7), followed by increases in BL levels and the accumulation of the dBZR1 protein in the buds (Fig. 4 G, H, and J).

Fig. 4.

BR signaling is involved in hormone and sugar-regulated bud outgrowth. (A) Effects of silencing of auxin biosynthesis gene FZY or cosilencing of SLs biosynthesis genes CCD7 and CCD8 on the total lateral bud length of dwf and bzr1 mutants. (B) Effects of application of 50 μM 6-BA, a synthetic CK, to buds and foliar application of sucrose at 20 mM on the total lateral bud length of dwf and bzr1 mutants. (C) Total lateral bud length of WT and DELLA-deficient pro mutant. (D–F) qPCR analysis of the relative transcript of BRC1 in axillary buds. (G–I) Changes in BL in axillary buds. (J) Changes in relative accumulation of dBZR1 in axillary buds. The signal intensity of dBZR1 protein was analyzed with ImageJ. The dBZR1 levels of control or TRV plants were set as one. Data are presented as the means of three replicates ± SD. Different letters indicate significant differences (P < 0.05) according to Tukey’s test.

In contrast to the above observations, a significant decrease in total lateral bud length was observed in the procera (pro) mutant, which is defective in a DELLA protein, an inhibitor of GA signaling (Fig. 4C and SI Appendix, Fig. S8A). The levels of BRC1 transcripts were increased, whereas the expression of BR biosynthesis genes (DET2 and DWF) was decreased, together with a decrease in the BL contents in the buds of the pro mutant (Fig. 4 F and I and SI Appendix, Fig. S8B). In addition, silencing of PRO resulted in a decreased accumulation of dBZR1 (Fig. 4J). WT plants treated with the GA biosynthesis inhibitor (paclobutrazol [PAC]) showed an increase in total lateral bud length and decreased levels of BRC1 transcripts. In contrast, PAC-induced bud growth was not observed in the dwf mutant (SI Appendix, Fig. S8 C–E). These data suggest that BR signaling is involved in the hormone and sugar-regulated shoot branching in tomato.

CK Signaling Mediates Multiple Hormone and Sugar Signals to Promote BR Biosynthesis in Buds.

CKs have been extensively studied as promoters of shoot branching (11, 12). The decreases in the auxin and SLs levels of the FZY- and CCD7/8-silenced plants, as well as the sucrose treatment resulted in an increase in the levels of IPT2 transcripts that encode the CK synthesis enzyme and in the total CK content in nodal stems (Fig. 5A and SI Appendix, Fig. S9A). The levels of CKs (iP, iPR, tZ, and tZR) were increased by cosilencing of CCD7/8 or by sucrose treatment, while iP and tZR levels were increased by silencing FZY. In contrast, the levels of IPT2 transcripts and the total CK contents were decreased in the pro mutant. Of the measured CKs, iPR and tZR were decreased in the pro mutant (SI Appendix, Fig. S9 B–E).

Fig. 5.

CK signaling regulates BR biosynthesis in axillary buds. (A) Effects of silencing of FZY or cosilencing of CCD7/8, treatment of 6-BA (50 μM) or foliar application of sucrose at 20 mM, and enhanced GA signaling in pro mutant on total CK content in nodal stems. (B and C) Bud outgrowth phenotypes of rr10 mutant and plants OE-RR10, which encodes a type-B RR in CK signaling. (D) qPCR analysis of the relative transcript of BRC1 in axillary buds. (E) qPCR analysis of the relative transcript of BR biosynthesis gene DWF in axillary buds. (F) Changes in BL in axillary buds. (G) ChIP-qPCR analysis of RR10 binding to the DWF promoter. The asterisk indicated statistical difference between WT and OE-BZR1 plants (Student’s t test, P < 0.05). (H) Dual-luciferase assay for the regulatory effect of RR10 on the expression of DWF. The ratio of LUC/REN of the EV plus promoter was set as one. Data are presented as the means of three replicates ± SD. Different letters indicate significant differences (P < 0.05) according to Tukey’s test. (I) Y1H analysis of RR10 binding to the DWF promoter. (J) The model in which BR signaling integrates multiple hormone and sugar signals to promote bud outgrowth. Auxin, SL, and GA reduce, while CK increases BR accumulation in axillary buds. BR activates BZR1, which transcriptionally suppresses BRC1, resulting in bud outgrowth. CK promotes BR biosynthesis through transcriptional activation of DWF with the action of RR10. Arrows indicate activation, and blunt-ended lines indicate inhibition.

Type-B RRs function as primary transcription factors in the regulation of gene expression in response to CKs (33). VIGS of RR8, RR9, RR10, RR13, and RR21 revealed that only silencing of RR10 significantly inhibited bud outgrowth (SI Appendix, Fig. S10). Based on these observations, rr10 mutant was generated using CRISPR/Cas9. The rr10 mutant showed a decrease in total lateral bud length, whereas overexpression of RR10 (OE-RR10) significantly promoted bud outgrowth relative to the WT (Fig. 5 B and C). BRC1 expression was increased in the rr10 mutant but was suppressed in the OE-RR10 plants (Fig. 5D). In addition, the rr10 mutant was insensitive to 6-BA with regard to bud outgrowth. Moreover, the RR10 protein levels were increased by 6-BA (SI Appendix, Fig. S11). Intriguingly, the expression of the key BR biosynthesis gene, DWF, was increased in the OE-RR10 plants but suppressed in the rr10 mutant. In agreement with these findings, OE-RR10 resulted in an increase of BL contents in buds, whereas BL accumulation was decreased in the buds of rr10 mutant (Fig. 5 E and F). However, no consistent changes in the levels of DET2 transcripts were observed (SI Appendix, Fig. S12). Y1H assay and ChIP-qPCR analyses indicated that RR10 directly bound to the promoter of DWF gene. Dual-luciferase assay confirmed the direct regulation of DWF expression by RR10 (Fig. 5 G–I). Taken together, these results indicate that RR10-mediated CK response promotes BR biosynthesis through transcriptional regulation of the DWF gene, leading to bud outgrowth in tomato.

Discussion

Shoot branching is generally considered to be controlled by the crosstalk between auxin, CKs, SLs, and sugars. Here, we provide evidence supporting that BR signaling is crucial to the control of shoot branching through the direct transcriptional suppression of BRC1. Auxin, CK, SL, and sugars impinge on a common BR–BZR1–BRC1 cascade that regulates bud outgrowth. Furthermore, our results demonstrate that RR10 mediates the CK response and regulates the expression of DWF to promote BR synthesis and bud outgrowth. Based on these findings, we propose a model in which BR signaling integrates hormonal and sugar signals to control shoot branching (Fig. 5J).

BR Signaling Is Essential for the Release of Apical Dominance by the Direct Suppression of BRC1 Expression.

The results presented in this study provide several lines of evidence showing that BR is directly involved in the regulation of bud outgrowth. Defects in BR synthesis (DWF) or in the positive regulator genes of BR signaling (BRI1 and BZR1) inhibited bud outgrowth. In contrast, OE-DWF or mutant in the negative regulator of BR signaling BIN2.2 promoted lateral bud outgrowth (Figs. 1 and 3). Decapitation induced a rapid increase in the levels of transcripts encoding BR biosynthesis enzymes, together with decreases in transcripts involved in BR inactivation or negative regulation of BR signaling in the axillary buds. These changes were followed by increased levels of BL and dBZR1. However, decapitation was less effective in the suppression of BRC1 expression and in the promotion of bud outgrowth in the dwf and bzr1 mutants or when the BR receptor gene BRI1 was silenced (Fig. 3). These findings suggest that BR signaling acts downstream of signals triggered by removal of shoot apex and that it is essential for the release of apical dominance.

Since no long-distance transport of BRs has been documented (34), BR signaling may act locally in buds to promote shoot branching. Indeed, BRC1, which is expressed in developing buds (8, 9), was found to be transcriptionally regulated by BR signaling. The results of a GUS reporter gene driven by the BRC1 promoter (P_BRC1::GUS) expression, in situ hybridization, and qPCR analysis demonstrate that BR suppresses the expression of BRC1 (Fig. 1). Molecular approaches provide further evidence that BZR1 is the key component of BR signaling that mediates the transcriptional regulation of BRC1 (Fig. 2). Importantly, we provide genetic evidence showing that brc1 mutant does not show enhanced shoot branching in the OE-DWF background, whereas silencing of BRC1 rescued the bud growth phenotypes in the dwf and bzr1 mutants, suggesting that BRC1 acts downstream of BR in the regulation of shoot branching. These results indicate that BR regulates shoot branching at least in part through BZR1-dependent regulation of BRC1 expression in lateral buds. Since the presence of BRC1 alone is not sufficient to prevent bud outgrowth in some cases (35), the incomplete recovery of bud growth achieved by the silencing of BRC1 in the dwf mutant suggests that BR may regulate bud growth through both BRC1-dependent and -independent routes. Since BR is a well-known regulator of cell division and elongation (36), these processes may be involved in BR-induced bud outgrowth.

BR Signaling Is Regulated by Hormone and Sugar Signals Involved in Apical Dominance.

Accumulating evidence has demonstrated the intensive crosstalk between auxin, SL, CK, and sugars in the control of apical dominance. In this regulatory network, auxin and SL act as repressors, while CK and sugars act as inducers of shoot branching (7, 11). In the present study, we found that the dwf mutant showed a decreased accumulation of IAA in the nodal stem, together with lower SL levels in the roots and an increased accumulation of CKs in the nodal stem (Fig. 1). These findings suggest that BR-dependent regulation of bud growth is not attributable to changes in auxin, SL, or CK levels. However, these hormones may regulate BR synthesis and signaling, which promote bud outgrowth.

The results presented here show that silencing of FZY or CCD7 and CCD8, which are involved in auxin and SL synthesis, stimulated bud growth, together with an increased abundance of DET2 and DWF transcripts and an increase in BL levels and BZR1 accumulation in the axillary buds. The observed increases in bud outgrowth were compromised in mutants that are defective in BR synthesis (dwf) or signaling (bzr1) (Fig. 4). These results indicate that BR signaling is essential for the auxin- and SL-dependent regulation of bud growth. The synthesis and/or signaling of BR and auxin show reciprocal regulation (37). Decreases in auxin resulting from decapitation resulted in a down regulation of the BR biosynthesis gene CYP90B1 and the catabolism gene CYP734A8 (Fig. 3). These findings are consistent with earlier findings that auxin increases the expression of CYP90B1 and CYP734A1 in Arabidopsis roots (38, 39). However, further studies are required to fully understand the mechanisms by which auxin signaling regulates BR accumulation in buds. We found that SLs negatively regulate the expression of BR biosynthesis genes as well as the accumulation of BL and dBZR1 (Fig. 4). These findings are consistent with an earlier observation that SLs activated MAX2, a subunit of an SCF E3 ligase, promotes the degradation of BZR1 protein in Arabidopsis (40). We conclude that BR mediates SL signaling in the regulation of shoot branching. Suppression of SL synthesis genes in the dwf mutant (Fig. 1) may be due to feedback regulation. Like auxin, GA has also been shown to be involved in the control of apical dominance (21, 41). The DELLA-deficient pro mutant showed reduced shoot branching with lower BR signaling in buds. Moreover, bud growth arising from the inhibition of GA synthesis by PAC was abolished in the dwf mutant (Fig. 4 and SI Appendix, Fig. S8). These results confirm the role of BR in apical dominance and indicate that GA inhibits BR signaling in axillary buds, possibly through increased auxin sensitivity (42).

Earlier studies have demonstrated that sugar availability is critical for regulating apical dominance (19, 43). We found that the application of sucrose to leaves promotes bud outgrowth, together with increased expression of BR synthesis genes and accumulation of BL and BZR1 (Fig. 4 and SI Appendix, Fig. S7). However, sucrose-driven bud growth was largely compromised in the dwf and bzr1 mutants, indicating that BR synthesis and signaling are required for sugar-induced shoot branching in tomato. Recently, mutations in the BR synthesis gene DET2 were found to compromise hypocotyl elongation in response to sugars (44). Moreover, the stability of BZR1 was shown to be regulated by the sugar-activated target of rapamycin protein kinase in Arabidopsis (45). Taken together, these results suggest that sugar acts as a signal in the regulation of BR biosynthesis and signaling in diverse developmental processes, including shoot branching.

CK Signaling Directly Regulates BR Biosynthesis.

Auxin in the nodal stem inhibits CKs synthesis (12), while D53, a central regulator of SL signaling, was recently shown to regulate CK degradation (46). In addition, increased sugar availability positively regulates CK synthesis (43). Consistent with previous studies, the results presented here show that CKs are regulated by auxin, SLs, GA signaling, and sucrose in tomato (Fig. 5). Based on these results, it will be interesting to study the mechanism whereby CK mediates auxin, SL, and sugar signals to regulate BR synthesis. CKs regulate hormone crosstalk through the action of type-B RRs (47, 48). Loss of RR10 functions suppressed bud outgrowth in tomato, whereas OE-RR10 promoted bud outgrowth (Fig. 5). In addition, we present evidence showing that RR10 is essential for CK-induced bud growth (SI Appendix, Fig. S11). The results of a series of physiological and molecular studies demonstrate that RR10 directly promotes BR synthesis by the transcriptional activation of DWF gene in tomato (Fig. 5). In this way, RR10-mediated BR synthesis integrates hormone and sugar signals to regulate shoot branching. CKs are known to suppress the expression of BRC1, while the mechanisms involved in the transcriptional regulation of BRC1 by CK signaling remain unclear (49). Our results indicate that CK signaling may regulate the expression of BRC1 through BR synthesis. Notably, CKs play an important role in shoot branching, but they are not absolutely required for bud growth in response to decapitation (50). BR functions locally in buds and may be more closely coupled with the release of apical dominance. Other signals such as SLs and sugars may also regulate BR synthesis and signaling through CK-dependent and CK-independent pathways to control shoot branching.

In conclusion, the data presented here show that BR is required for the release of apical dominance. BR does not promote bud growth through decreasing apical dominance, but rather it acts locally in buds to suppress the expression of BRC1 through the action of BZR1. Traditional signals involved in apical dominance such as auxin, SLs, CKs, and sugars, together with GA regulate BR biosynthesis in buds. In addition, auxin, SL, sugar, and GA may regulate BR via CK, which regulates BR synthesis through RR10-mediated transcriptional activation of DWF gene (Fig. 5J). Based on these findings, together with the results of previous studies, we propose that BR signaling integrates multiple pathways in the regulation of shoot branching.

Materials and Methods

Plant Materials and Growth Conditions.

The dwf and pro mutants, together with the corresponding WT, Solanum lycopersicum cv. (cultivar) Condine Red and cv. Ailsa Craig were obtained from Tomato Genetics Resource Center (https://tgrc.ucdavis.edu). Condine Red was used as the WT for CRISPR/Cas9 mutants and transgenic plants. Transgenic plants overexpressing DWF and OE-BZR1, and CRISPR/Cas9 mutant of bzr1 were generated during previous studies (51, 52). Seeds were germinated in Petri dishes at 28 °C and the germinated seeds were sown in a mixture of peat and vermiculite (2:1, volume/volume [vol/vol]). When the plants had grown to the two-leaf stage, they were transferred to pots (height × diameter, 15 cm × 10 cm) containing the same substrate. The plants were grown in a controlled growth chamber with 12 h light (200 μmol m⁻² ⋅ s⁻¹) at 23 °C and 12 h dark at 20 °C. The relative humidity was kept at 70%. Plants were watered with Hoagland nutrient solution every 2 d.

VIGS.

For gene silencing, complementary DNA (cDNA) fragments of target genes were amplified using primers listed in SI Appendix, Table S1. Purified PCR products were cloned into TRV2 vector. After confirmation by sequencing, the plasmids were transformed into Agrobacterium tumefaciens strain GV3101. When the cotyledons fully expanded, the seedlings were infiltrated with a mixture of A. tumefaciens strain carrying a TRV2 derivative and the strain carrying the helper vector TRV1 as previously described (53). The infiltrated plants were kept in the aforementioned growth chamber. qPCR was performed to ensure silencing efficiency before experiments.

Cloning Procedures and Plant Transformation.

CRISPR/Cas9 system was used to generate bin2.2, brc1, and rr10 mutants as previously described (54). The single guide RNA (sgRNA) sequences were designed by the CRISPR-P program (crispr.hzau.edu.cn/CRISPR2/) as follows: BIN2.2: ACC​TCA​GCA​CCA​TAA‐TCC​GC; BRC1: TCT​TCT​CCT​TGT​ATG​CAA​TA; RR10: GGG​TCT​GTT​CTC​AGT​TCG​AC.

The synthesized sgRNA sequence was annealed and introduced into BbsI site of an AtU6–sgRNA–AtUBQ–Cas9 vector. The resulting plasmid was digested by HindIII and KpnI and then inserted into pCAMBIA1301 binary vector digested by the same restriction enzymes. After confirmation by sequencing, the vector containing sgRNA and Cas9 was transformed into A. tumefaciens strain GV3101. Plant transformation was performed as described previously (51). Primers used for genotyping were as follows: BIN2.2: TGA​GAA​CAC​GAG​AAA​AAT​AT and ACT​ACC​TTT​TCG​GAT​TCC​ATT; BRC1: ATCACTTTGGTCAATCCA and TCA​TCT​CCT​TTC​TTT​TCG; RR10: GTA​AGA​TAA‐ACC​CCC​CAA​AAA​G and CTT​CAT​AGT​GAC​AGT​TCT​TGA​GC.

An insertion of one base pair (bp) was found in the +38, +70, and +33 position in the coding sequence of BIN2.2, BRC1, and RR10, respectively. The homozygous lines were used in this study.

To generate plants expressing P_BRC1::GUS, a 1,369 bp P_BRC1 sequence was amplified by PCR using specific primers (SI Appendix, Table S2). The PCR product was ligated into the pBI121 vector, which harbors the GUS reporter gene. After confirmation by sequencing, the vector was transformed into A. tumefaciens strain GV3101, which was used for plant transformation. The RR10-OE plants were generated as described previously (51). The full-length coding sequence of RR10 was amplified with the primers as shown in SI Appendix, Table S2. T3 homozygous lines generated from T1 individuals carrying a single insertion of the transgene were used in this study. The P_BRC1::GUS transgene was introduced into the dwf mutant and the 35S::DWF transgene was introduced into the brc1 mutant by crossing.

Treatments and Measurement of Lateral Bud Length.

Treatments of plants were performed when plants have five fully expanded leaves. For decapitation, young leaves and shoot apex above the fifth node were removed by a sterilized scalpel. Nodes were numbered acropetally from the first true leaf. For 6-BA treatment, 22.5 mg 6-BA (Sigma-Aldrich, Merck Life Science Co.) was dissolved in 10 mL NaOH solution (0.1 M) and diluted with 40 mL distilled water to make a 2 mM 6-BA stock solution, then the stock solution was diluted 40 folds to make a 50 μM working solution. The 6-BA solution was applied in a volume of 10 μL directly to the axillary buds every two days. For sugar feeding, 20 mM sucrose solution was sprayed to fully expanded leaves of the whole plant. Each plant was treated with 15 mL sucrose solution twice a day until the end of the experiment. For the treatment of GA synthesis inhibitor PAC, PAC (Supelco, Merck Life Science Co.) was firstly dissolved in a minimum volume of ethanol to make a stock solution and then diluted with distilled water to make a 20 μM PAC working solution. The PAC solution was sprayed to fully expanded leaves of the whole plant in a volume of 15 mL. The treatment was applied twice a day until the end of the experiment. The length of lateral buds was measured 7 d after different treatments using a numeric caliper, perpendicular to the stem.

Histochemical GUS Analysis.

For GUS staining, transgenic plants bearing the P_BRC1::GUS were fixed in 50 mM phosphate buffer (pH 7.0) containing 1% (vol/vol) formaldehyde and 0.04% Triton-X 100 for 30 min. After washing twice with phosphate buffer, samples were incubated at 37 °C overnight with GUS staining solution (50 mM phosphate buffer, pH 7.0, 0.4 mg ⋅ mL⁻¹ 5-bromo-4-chloro-3-indolyl-β-D-glucuronic acid, 1 mM potassium ferricyanide, and 1 mM potassium ferrocyanide). Following staining, samples were washed with a graded ethanol series to extract chlorophyll. Expression of P_BRC1::GUS was observed using a Zeiss STEMI305 stereo microscope.

In Situ Hybridization.

Tomato axillary buds (about 2 mm) were fixed with 4% paraformaldehyde at 4 °C overnight, washed with phosphate buffer three times (5 min each), dehydrated through an ethanol series (30%, 50%, 70%, 80%, 90%, 95%, and 100%) for 30 min in each solution, and then treated with 100% ethanol for 1 h. After that, samples were treated through a graded series of xylene in ethanol (25%, 50%, 75%, and 100%), washed with 50% paraffin in xylene at 65 °C for 1 h, incubated at 65 °C with 100% paraffin overnight, embedded in paraffin in molds (MEIKO EC 360), and allowed to solidify at room temperature. Paraffin blocks were cut into 8 µm sections using a Manual Rotary Microtome (Thermo Scientific HM 325) and collected on adhesion microscope slides (MeVid). After dewaxing, in situ hybridization was performed using the RNAscope 2.5 HD Detection Kit (Advanced Cell Diagnostics [ACD]) following the manufacturer’s protocol. BRC1 probe and a negative control probe for an irrelevant bacterial gene dapB were provided by ACD. Slides were imaged on a Zeiss Axio Scope A1 microscope (Zeiss) with a Zeiss Axiocam 503 color camera and Zeiss ZEN imaging software.

Y1H Assay.

Y1H assay was performed using Matchmatch Gold Yeast One-Hybrid System (Clontech) according to the manufacturer’s instructions. BRC1 or DWF promoter was ligated into the pAbAi vector and the full-length coding region of BZR1 or RR10 was also amplified using specific primers (SI Appendix, Table S2) and fused to the pGADT7 vector. The linearized pAbAi constructs containing BRC1 or DWF promoter were transformed into Y1HGold yeast strain. pGADT7-BZR1, pGADT7-RR10, or an empty AD vector was transformed into the modified Y1HGold yeast strain. The transformed yeast cells were selected on SD/Leu− media supplemented with 50 ng ⋅ mL⁻¹ AbA.

ChIP-qPCR.

ChIP assay was performed using the EpiQuik Plant ChIP Kit (Epigentek). One gram of buds was fixed in 1% formaldehyde for 10 min under a vacuum and neutralized with glycine. After washing three times with cold sterilized water, the tissues were homogenized. Following isolation and sonication of chromatin, BZR1-HA (hemagglutinin) or RR10-HA protein was immuno-precipitated using anti-HA antibody. The enriched DNA was amplified by qPCR using specific primers (SI Appendix, Table S3). Enrichment was calculated by percentage of the immunoprecipitated DNA relative to the input. The goat anti-mouse immunoglobulin G (IgG) was used as the negative control. Relative enrichment was calculated by the ratio of the percentage of anti-HA–immunoprecipitated DNA to the percentage of IgG-immunoprecipitated DNA.

Dual-Luciferase Assay.

Dual-luciferase assay was used to confirm BZR1 binding to the P_BRC1 and RR10 binding to the DWF promoter according to a previous study (55). BRC1 and DWF promoters were amplified and ligated into the pGreen II 0800-LUC vector (SI Appendix, Table S2). A variant of P_BRC1, which contains a mutation in the BZR1-binding site, was generated using overlapping PCR, in which two amplified fragments with an intended mutation in the overlapping region are joined together by PCR (56). The primer pairs used to create mutations are as follows (substituted nucleotides are underlined): reverse: 5′-CGT​GAA​GTA​AAT​AAA​GTTT​TTTACT​TTG​TCA​TCC‐TGA​GCC​G-3′; forward: 5′-CGG​CTC​AGG​ATG​ACA​AAG​TAAA​AAACTT​TAT​TTA‐CTT​CAC​G-3′.

The full-length coding region of BZR1 or RR10 was amplified and fused to the pGreen II 0029 62-SK vector (SI Appendix, Table S2). The above constructs were transformed into A. tumefaciens strain GV3101. To determine the activity of promoter as influenced by transcription factor, a mixture of A. tumefaciens strain carrying the pGreen II 0800-LUC vector or pGreen II 0029 62-SK vector in a 1:1 ratio was infiltrated into Nicotiana benthamiana leaves. Three days later, the activity of promoter was measured by the ratio of enzyme activity of firefly luciferase (LUC) and renilla luciferase (REN), which acted as an internal reference. The LUC/REN value in the absence of BZR1 or RR10 protein was set as one. Measurements were carried out using a Modulus Luminometer (Promega).

Measurement of Phytohormones.

For measurement of IAA, 0.1 g samples were collected and homogenized in 1 mL ethyl acetate, which was spiked with D₆-IAA (C/D/N Isotopes Inc.) as internal standards. The homogenates were shaken overnight at 4 °C and centrifuged at 13,000 rpm for 10 min. The supernatants were transferred to new tubes and the pellets were resuspended in 1 mL ethyl acetate. The suspensions were shaken for 1 h and then centrifuged at 13,000 rpm for 10 min. The supernatants were combined and dried under gaseous N₂ flush. Samples were dissolved in 0.5 mL 70% methanol (vol/vol) for analysis.

SLs measurement was performed according to the method of Ruiz-Lozano et al. (57) with modifications. Root samples (0.5 g) were ground in a mortar filled with liquid nitrogen and then extracted with 0.5 mL 40% acetone (vol/vol). The homogenates were vortexed for 2 min and centrifuged at 8,000 g for 5 min. The supernatants were discarded, and the pellets were extracted twice with 0.5 mL 50% acetone (vol/vol). The supernatants were filtered through membrane filters (0.22 μm) and then analyzed.

For analysis of CKs, samples (0.2 g) frozen in liquid N₂ were ground to fine powders and then ground in 1 mL ice-cold extraction solution (15:1:4 [vol/vol/vol] methanol: formic acid: water), which was spiked with [²H₆]N⁶-iP, [²H₆]N⁶-iPR, [²H₃]-dihydrozeatin, [²H₃]-DHZR, [²H₅]-tZ, and [²H₅]-tZR. The extracts were kept at −30 °C overnight and centrifuged at 10,000 g for 15 min. The supernatants were collected and flowed through a hydrophilic lipophilic balance column (Oasis) which was pretreated with formic acid (1 M). An aliquot of 0.3 mL extraction solution was flowed through the column. The liquid was collected and dried under gaseous N₂ flush. Samples were dissolved in 1 mL formic acid (1 M) and flowed through a mixed-mode cation (MCX) column (Oasis) which was pretreated with formic acid (1 M). The column was washed sequentially with 1 mL formic acid (1 M), 1 mL methanol, 1 mL ammonia solution (0.35 M), and 1 mL 60% methanol containing 0.35 M ammonia. The final liquid was dried under gaseous N₂ flush and dissolved in 200 μL 1% acetic acid. The mixture was vortexed and then analyzed.

For measurement of BL, 0.1g bud or stem samples were ground into powder with liquid N₂ and then ground in 1 mL ice-cold acetonitrile, which was spiked with [26-²H₃]BL (Olchemim) as internal standard. After being extracted overnight at 4 °C, the homogenate was centrifuged at 10,000 g for 10 min at 4 °C. The supernatant was transferred into a new tube which contained 0.3g MCX@BBII and 3 mL ddH₂O and was vortexed for 5 min. MCX@BBII was prepared according to Luo et al. (58). After centrifugation at 10,000 g for 1 min at 4 °C, the pellet was resuspended with 5 mL 90% acetone which contained 0.5% formic acid and vortexed for 1 min. The mixture was centrifugated at 10,000 g for 1 min at 4 °C. The supernatant was discarded, and 1.2 mL 90% acetone followed by 20 to 50 mg ammonium acetate was added to the tube. The mixture was vortexed for 1 min and centrifuged at 10,000 g for 3 min. The upper phase was dried under mild N₂ flow and was dissolved in 100 μL 45% acetonitrile for analysis.

Phytohormones except BL were analyzed by Agilent 1290 ultra-high performance liquid chromatography coupled to 6460 triple quadruple mass spectrometer. BL was analyzed by using ACQUITY ultra performance liquid chromatography I-Class coupled to Waters Xevo TQ-XS triple quadruple mass spectrometer.

Western Blot.

Protein extraction from bud samples and Western blot were performed as described previously (52). For Western blot, proteins were separated by SDS-PAGE using 10% (wt/vol) acrylamide gels and electrophoretically transferred to nitrocellulose membranes (Millipore). The BZR1-HA and RR10-HA proteins were detected with commercial antibodies raised against anti-HA monoclonal antibody (Thermo Fisher Scientific).

Gene Expression Analysis.

Total RNA extraction was performed using an RNAprep pure Plant Kit (TIANGEN) according to the operation manual. The first-strand cDNA was synthesized using ReverTraAce qPCR Reverse Transcription Kit with genome-DNA–removing enzyme (Toyobo). qPCR was performed on LightCycler480 detection system (Roche) using SYBR Super Mix (Takara, RR420A). Primers for target genes were shown in SI Appendix, Table S4. Tomato housekeeping gene Actin was used as internal reference. Relative expression of target genes was calculated according to Livak and Schmittgen (59).

Statistical Analysis.

The experimental design was a completely randomized design. Data were subjected to statistical analysis of variance using the SPSS package (SPSS 19.0). The differences between the means were separated by Tukey’s test at a level of P < 0.05.

Data Availability

All study data are included in the article and/or SI Appendix.