Skip to main content
NCBI home page
Plant Physiology logoLink to Plant Physiology
. 2020 Sep 10;184(3):1424–1437. doi: 10.1104/pp.20.00536

OsHOX1 and OsHOX28 Redundantly Shape Rice Tiller Angle by Reducing HSFA2D Expression and Auxin Content1

Yong Hu a,b, Shuangle Li a, Xiaowei Fan a, Song Song a, Xin Zhou a, Xiaoyu Weng a, Jinghua Xiao a, Xianghua Li a, Lizhong Xiong a, Aiqing You b, Yongzhong Xing a,2,3
PMCID: PMC7608169  PMID: 32913047

Two homeodomain-leucine zipper genes shape tiller angle of rice by influencing lateral asymmetric distribution of the plant hormone auxin.

Abstract

Tiller angle largely determines plant architecture, which in turn substantially influences crop production by affecting planting density. A recent study revealed that HEAT STRESS TRANSCRIPTION FACTOR2D (HSFA2D) acts upstream of LAZY1 (LA1) to regulate tiller angle establishment in rice (Oryza sativa). However, the mechanisms underlying transcriptional regulation of HSFA2D remain unknown. In this study, two class II homeodomain-Leu zipper genes, OsHOX1 and OsHOX28, were identified as positive regulators of tiller angle by affecting shoot gravitropism. OsHOX1 and OsHOX28 showed strong transcriptional suppressive activity in rice protoplasts and formed intricate self- and mutual-transcriptional negative feedback loops. Moreover, OsHOX1 and OsHOX28 bound to the pseudopalindromic sequence CAAT(C/G)ATTG within the promoter of HSFA2D, thus suppressing its expression. In contrast to HSFA2D and LA1, OsHOX1 and OsHOX28 attenuated lateral auxin transport, thus repressing the expression of WUSCHEL-RELATED HOMEOBOX 6 (WOX6) and WOX11 in the lower side of the shoot base of plants subjected to gravistimulation. Genetic analysis further confirmed that OsHOX1 and OsHOX28 act upstream of HSFA2D. Additionally, both OsHOX1 and OsHOX28 inhibit the expression of multiple OsYUCCA genes and decrease auxin biosynthesis. Taken together, these results demonstrated that OsHOX1 and OsHOX28 regulate the local distribution of auxin, and thus tiller angle establishment, through suppression of the HSFA2D-LA1 pathway and reduction of endogenous auxin content. Our finding increases the knowledge concerning fine tuning of tiller angles to optimize plant architecture in rice.

Tillers are specified branches that function as the basic elements of the rice (Oryza sativa) plant body. Tillers initiate from the axial buds of leaves on their mother stems (Wang and Li, 2005) and grow at a certain angle (tiller angle) relative to the main culm (Xu et al., 1998). Tiller angle is a component of plant architecture that is crucial for rice yield. Loose tillers not only reduce harvest efficiency but also decrease planting density and yield production, while tillers that are too compact increase the risk of disease infection (Wang et al., 2018). In an agricultural setting, ideal rice plant architecture is characterized by a low number of productive tillers, an increased number of grains per panicle, and thick and sturdy stems. Hence, a comprehensive understanding of the molecular mechanisms underlying tiller angle is important for breeding rice varieties with ideal plant architecture.

Gravitropism, a conserved phenomenon in angiosperms, is divided into four sequential steps: gravity perception, signal transduction, asymmetrical distribution of auxin, and organ responses (Wu et al., 2016). Gravitropism contributes greatly to shaping plant architecture by altering the growth direction of both above- and belowground parts of the plant body (Chen et al., 1999). Recent studies have revealed close relationships between tiller angle establishment and shoot gravitropism. A large-scale transcriptomic analysis of rice shoots in response to gravistimulation revealed a core regulatory pathway involved in the regulation of tiller angle, which mainly includes two key genes: HEAT STRESS TRANSCRIPTION FACTOR 2D (HSFA2D) and LAZY1 (LA1; Zhang et al., 2018). The expression of HSFA2D is induced upon gravitropic treatment and positively regulates the transcription of LA1. LA1 encodes a protein containing a conserved GφL(A/T) IGT motif (Dardick et al., 2013) and promotes gravitropic responses and compact tiller growth (Li et al., 2007; Yoshihara and Iino, 2007). Loss of function of LA1 disrupts the asymmetric auxin distribution in the shoot base by inhibiting lateral auxin transport (LAT; Li et al., 2007). WUSCHEL-RELATED HOMEOBOX6 (WOX6) and WOX11 are two pivotal regulators in the step of organ responses of gravitropism (Zhang et al., 2018). Multiple auxin response elements have been identified in the promoters of WOX6 and WOX11 (Ponomarenko and Ponomarenko, 2015). The expression of WOX6 and WOX11 is induced by auxin. Failure to establish asymmetric auxin distribution in la1 and hsfa2d mutants altered the auxin-mediated asymmetric expression of WOX6 and WOX11 in the upper and lower sides of the shoot base of plants subjected to gravitropic treatment (Zhang et al., 2018). The asymmetric auxin gradient between the upper and lower sides of shoots leads to differential cell expansion and therefore shoot bending in the opposite direction of gravity (Chen et al., 1999; Wu et al., 2016).

Brevis Radix Like4 of rice (OsBRXL4) acts as a negative regulator of tiller angle by interacting with LA1 and affecting its nuclear localization, which is essential for its function. OsBRXL4 was demonstrated to regulate shoot gravitropism by affecting LAT and polar auxin transport (PAT; Li et al., 2019). Similar studies in Arabidopsis (Arabidopsis thaliana) have shown that the interaction between the CCL domain of AtLAZY1 and the Brevis Radix (BRX) domain of RCC1-like domain (RLD) proteins is important for the recruitment of RLDs from the cytoplasm to the plasma membrane (Furutani et al., 2020). Mutations of genes involved in the strigolactone (SL) biosynthetic or signaling pathway suppress the loose tiller phenotype of la1. Further analysis revealed that SLs inhibit auxin biosynthesis and attenuate rice shoot gravitropism by decreasing local indole-3-acetic acid (IAA) content (Sang et al., 2014). The most recent evidence, from a study on Tiller Angle Control (TAC4), revealed that reduction of auxin content caused by decreasing the expression of several OsYUCCA (OsYUC) genes in the tac4 mutant is responsible for its spreading tiller phenotype (Li et al., 2020).

In addition, plenty of other key regulators of tiller angle have been identified in the last two decades, including TAC1 (Yu et al., 2007), PLANT ARCHITECTURE AND YIELD1 (PAY1; Zhao et al., 2015), α-1,3-fucosyltransferase1 (FucT1; Harmoko et al., 2016), PROSTRATE GROWTH1 (PROG1; Jin et al., 2008), PROG7 (Hu et al., 2018), TAC3, DWARF 2 (D2; Dong et al., 2016), TILLER INCLINED GROWTH1 (TIG1; Zhang et al., 2019), Loose Plant Architecture1 (LPA1; Wu et al., 2013), RICE PLANT ARCHITECTURE DOMESTICATION (RPAD; Wu et al., 2018), and two NAM/ATAF1/ATAF2/CUC2 (NAC) members—OsNAC2 (Mao et al., 2007) and OsNAC106 (Sakuraba et al., 2015). Although these findings provide valuable information for understanding the molecular mechanisms underlying tiller angle regulation, more information is needed to construct a comprehensive regulatory network.

Homeodomain-Leu zipper (HD-ZIP) proteins are unique to the plant kingdom. These transcription factors contain a homeodomain and a Leu zipper domain. HD-ZIP members can be classified into four subgroups (I–IV; Brandt et al., 2014); members of the different subgroups have distinct binding motifs, with HD-ZIP II members specifically binding to the pseudopalindromic sequence CAAT(C/G)ATTG (Elhiti and Stasolla, 2009). Multiple Arabidopsis HD-ZIP II members contain an LxLxL type of Ethylene-responsive element binding factor-associated Amphiphilic Repression (EAR) motif, which implies that these proteins may function as negative regulators (Turchi et al., 2015). The Leu zipper domain enables dimerization of HD-ZIP proteins (Harris et al., 2011), which increases the DNA-binding efficiency compared with the monomeric form (Palena et al., 1999). Studies in Arabidopsis have shown that HD-ZIP II transcription factors are involved in the shade avoidance response (Turchi et al., 2015). Expression of HD-ZIP II members was induced under shaded conditions in a phytochrome-dependent manner. Overexpression of these genes resulted in similar phenotypes in nonshaded or open conditions, including increased hypocotyl length, decreased numbers of branches and decreased leaf size, which are similar to the phenotypic characteristics of wild-type plants in response to shade (Turchi et al., 2015). However, compared with those in Arabidopsis, the functions of HD-ZIP II members in rice have been less studied. OsHOX1 is involved in perivascular cell fate commitment, which was mainly revealed by cytological analysis and comparison of OsHOX1 expression patterns in vascular tissues between wild-type and OsHOX1 overexpression plants (Scarpella et al., 2000). Small Grain and Dwarf2 (SGD2/OsHOX3), another HD-ZIP II member, regulates plant height and seed size by upregulating GA biosynthesis genes (Chen et al., 2019). However, the function of HD-ZIP II members in regulating plant architecture remains to be elucidated.

Our previous study revealed that OsHOX1 is one of the possible downstream genes of the crucial pleiotropic gene Grain number, plant height and heading date7 (Ghd7; Weng et al., 2014). Further investigation revealed that OsHOX1 functions in regulating tiller angle rather than heading date and plant height. In this study, we showed that two HD-ZIP transcription factors, OsHOX1 and OsHOX28, act redundantly in rice tiller angle establishment. Through regulation of the HSFA2D-LA1 pathway and endogenous auxin content, OsHOX1 and OsHOX28 affect the local distribution of auxin within the shoot base. Our results not only enrich the knowledge concerning the HSFA2D-LA1 regulatory pathway of plant architecture but also provide new gene resources for future genetic improvements to rice plant architecture.

RESULTS

OsHOX1 and OsHOX28 Regulate Shoot Gravitropism and Horizontal Tiller Angle in Rice

Tiller angle was originally defined as the angle between the side tiller and the main tiller (Xu et al., 1998), which is difficult to accurately measure. Here, the horizontal tiller angle (Fig. 1A, α) between the tiller and horizon was used to quantify tiller architecture. Lines overexpressing OsHOX1 (OsHOX1-OX; Supplemental Fig. S1A) and its homolog OsHOX28 (OsHOX28-OX; Supplemental Figs. S1B and S2) presented increased horizontal tiller angles under standard planting density (P < 0.05; Supplemental Fig. S3, A, B, and M, 16.7 × 26.7 cm) and low planting density (P < 0.05; Fig. 1, A–C and F, 33.4 × 53.4 cm) conditions. We obtained Cas9-free transgenic lines with either OsHOX1 (OsHOX1-CR) or OsHOX28 (OsHOX28-CR) or both (OsHOX1/28-CR) edited by the CRISPR/Cas9 strategy (Supplemental Fig. S1, C–E). Premature stop codons in OsHOX28 were induced in both OsHOX28-CR and OsHOX1/28-CR lines. Frameshift mutations in OsHOX1 were induced in both OsHOX1-CR and OsHOX1/28-CR lines, except for OsHOX1/28-CR line 1, which contained an amino acid deletion (Supplemental Fig. S1F). Expression analysis showed that CRISPR-induced mutations did not affect the expression of OsHOX1 and OsHOX28 (Supplemental Fig. S1, G and H). Compared with that of wild-type plants, the horizontal tiller angle of OsHOX1/28-CR lines slightly decreased only under low-density growing conditions (P < 0.05; Fig. 1, D and F), but no significant changes were detected in OsHOX1-CR and OsHOX28-CR lines under either growing condition (Supplemental Fig. S3, C–N).

Figure 1.

Figure 1.

Open in a new tab

Regulation of tiller angle and shoot gravitropism by OsHOX1 and OsHOX28. A to D, Phenotypes of wild-type (WT; cvZH11) plants, transgenic plants overexpressing OsHOX1 (OsHOX1-OX) or OsHOX28 (OsHOX28-OX), and the double-edited line of OsHOX1 and OsHOX28 (OsHOX1/28-CR, generated by CRISPR/Cas9 strategy) under low-density growing conditions (33.4 × 53.4 cm) in the natural field at the maturity stage. α represents the horizontal tiller angle, i.e. the angle between the tiller and horizontal level. Scale bars = 25 cm. E, Shoot curvature phenotype of seedlings of the wild type, OsHOX1-OX, OsHOX28-OX, OsHOX1/28-CR, and OsHOX1-CR and OsHOX28-CR single mutants after 36 h of gravitropic treatment (turning the plate by 90°). Fourteen-day-old seedlings were used for analysis. Scale bar = 2 cm. F, Horizontal tiller angles of wild-type, OsHOX1-OX, OsHOX28-OX, and OsHOX1/28-CR plants under low-density growing conditions. Data are shown as the means ± sd (n = 10). G, Comparison of shoot curvature among the plant materials described in E. Data are shown as the means ± sd (n = 18). Different lowercase letters indicate significant difference by Duncan’s test (P < 0.05).

To investigate the involvement of OsHOX1 and OsHOX28 in controlling shoot gravitropism, we examined the shoot gravitropic response of seedlings of OsHOX1-OX and OsHOX28-OX transgenic lines. The responses of both lines decreased in response to gravitropism treatment implemented by turning the plates 90° (Fig. 1, E and G). In contrast, seedlings of OsHOX1/28-CR displayed a subtle but significantly enhanced gravitropic response, although no significant difference was detected between the single edited lines (OsHOX1-CR and OsHOX28-CR) and the wild type (Fig. 1, E and G). These results demonstrated that OsHOX1 and OsHOX28 are redundantly involved in the regulation of both shoot gravitropism and tiller angle in rice.

OsHOX1 and OsHOX28 Act as Transcriptional Suppressors and Form Homo- and Heterodimers

There are 12 HD-ZIP II members in rice and 10 homologs in Arabidopsis. Phylogenetic analysis revealed that OsHOX1 and OsHOX28, together with two other rice HD-ZIP II members, OsHOX17 and OsHOX2, were closely grouped together with HAT1, HAT2, HAT3, ATHB-2, and ATHB-4 in a subbranch of the tree (Supplemental Fig. S2). Nine of 12 rice HD-ZIP II members, including OsHOX1 and OsHOX28, have the EAR transcriptional repression motif (Fig. 2A). A luciferase (LUC) assay was applied to analyze the transcriptional activity of these two proteins in rice protoplasts. The strong activation domain from viral protein16 (VP16) of the herpes simplex virus (Triezenberg et al., 1988) was fused to either the N or C terminus of OsHOX1 and OsHOX28 (Fig. 2B). Regardless of the fusion position, the chimeric proteins of OsHOX1 and OsHOX28 dramatically inhibited the induction activity of VP16 (Fig. 2, B and C), which demonstrated that OsHOX1 and OsHOX28 function as transcriptional suppressors. OsHOX1 and OsHOX28 could interact with each other and themselves in yeast two-hybrid analysis (Fig. 2D), which indicates that OsHOX1 and OsHOX28 are able to form homo- and heterodimers. These results were further verified by the LUC complementation imaging assay (Fig. 2E).

Figure 2.

Figure 2.

Open in a new tab

Transcriptional suppression activities of OsHOX1 and OsHOX28 proteins and interactions with each other and with themselves. A, Alignment of N-terminal fragments containing an EAR motif (indicated by the solid line) of HD-ZIP II proteins in rice and Arabidopsis. B, Constructs used in transcriptional regulation assays in rice protoplasts. 5× GAL4, Multimerized GAL4 binding site; TATA, TATA box; Nos, nopaline synthase terminator; GAL4DBD, DNA-binding site of the GAL4 protein; Ω, translation enhancer sequence; VP16, strong activation domain from VP16 of the herpes simplex virus. C, Effects of different effectors on the reporters. The relative luciferase (LUC) activity is represented by the ratio of signal values of firefly LUC to that of Renilla LUC (internal control). Data are shown as the means ± sd of three independent transformants. Different lowercase letters indicate significant difference by Duncan’s test (P < 0.05). D, Yeast two-hybrid assay of OsHOX1 and OsHOX28. Yeast cells transformed with the indicated plasmids were grown on control synthetic defined (SD)/-Trp-Leu medium and selective SD/-Trp-Leu-His-Ade medium. E, LUC complementation imaging assays of OsHOX1 and OsHOX28. OsHOX1 and OsHOX28 were fused with either the N or C terminus of LUC (nLUC or cLUC, respectively) and then infiltrated into Nicotiana benthamiana leaves. nLUC and cLUC are empty vectors.

Transcriptional Regulation of OsHOX1 and OsHOX28 by Negative Feedback Regulation

Complicated negative feedback regulatory networks within the HD-ZIP family are essential for the functions of HD-ZIP members in the eudicot plant Arabidopsis (Ohgishi et al., 2001; Sawa et al., 2002). However, it is not clear whether feedback regulation occurs in the monocot plant rice. In this study, the coding DNA sequences of OsHOX1 and OsHOX28 were used for generation of corresponding overexpression lines (Fig. 3A). We analyzed the expression of OsHOX1 and OsHOX28 in OsHOX1-OX and OsHOX28-OX plants. Primers designed to bind within coding DNA sequences and untranslated regions (UTRs) were used to measure the total and endogenous expression levels, respectively, of target genes (Fig. 3B). In the OsHOX1-OX plants, the high expression of OsHOX1 greatly suppressed its endogenous expression and the expression of its homolog OsHOX28 (Fig. 3C). Similarly, the endogenous expression of OsHOX28 and OsHOX1 was greatly suppressed by ectopic expression of exogenous OsHOX28 (Fig. 3D). The putative HD-ZIP II binding site CAAT[G/C]ATTG was identified ∼5 kb upstream of OsHOX1 (Fig. 3B), suggesting the possibility of direct negative feedback regulation of OsHOX1 by its own protein and its homolog OsHOX28. A yeast one-hybrid assay revealed that OsHOX1 and OsHOX28 indeed bound to fragment 1p2, which is upstream of OsHOX1 and contains the binding site, but not the mutated fragment 1p2m (Fig. 3E). Furthermore, electrophoretic mobility shift assays (EMSAs) confirmed the interactions between fragment 1p2 and both OsHOX proteins (Fig. 3F). OsHOX1 was found to be highly enriched at 1p2 site upstream of OsHOX1 in the chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) analysis (Fig. 3G). Interestingly, enrichments of OsHOX1 at p3, p4, and p5 within 2 kb upstream of OsHOX28 were also detected (Fig. 3G), which is consistent with the transcriptional regulation effect of OsHOX1 on OsHOX28 (Fig. 3C). These results suggested that expression of both OsHOX1 and OsHOX28 was fine-tuned by negative feedback regulation.

Figure 3.

Figure 3.

Open in a new tab

Transcriptional expression of OsHOX1 and OsHOX28 by negative feedback regulation. A, Schematic diagram of constructs for overexpression of OsHOX1 and OsHOX28, with numbered black boxes representing the exons of the indicated gene. B, Gene structure of OsHOX1 and OsHOX28. The empty boxes on the left and right sides represent the 5′ and 3′ UTRs, respectively; the labeled black boxes represent exons; lines between the black boxes represent introns; and lines before the 5′ UTR and after the 3′ UTR are sequences up- and downstream of genes. qHOX1 and qHOX1-3′UTR represent the fragments amplified with primers for RT-qPCR to measure the total and endogenous expression, respectively, of OsHOX1 in the overexpression lines. qHOX28 and qHOX28-5′UTR are similar designs for OsHOX28. The red characters indicate the HD-ZIP II binding cis-element CAAT[G/C]ATTG and its mutated type. C and D, Analysis of total and endogenous expression of OsHOX1 and OsHOX28 in OsHOX1-OX (C) and OsHOX28-OX lines (D). Fourteen-day-old seedlings were used for the analysis. Data are shown as the means ± sd of three biological replicates. Asterisks indicate significant difference by Student’s t test (**P < 0.01). E, Yeast one-hybrid assay showing the interaction of OsHOX1 and OsHOX28 with fragment 1p2 containing the cis-element 1p2 but not the mutated 1p2m within the promoter of OsHOX1. F, EMSA showing direct interactions between GST-tagged proteins (OsHOX1:GST and OsHOX28:GST) and 1p2 probes. The 100- and 400-fold excess unlabeled probes were used for competitions. G, Enrichments of OsHOX1 on the promoter of OsHOX1 and OsHOX28. ChIP assays were performed with 25-d-old wild-type (WT) and OsHOX1-FLAG transgenic line 3 plants. The precipitated fragments were analyzed by qPCR using the primer sets indicated by the short solid lines in B. Data represent the means ± sd of three biological replicates. Asterisks indicate significant difference by Student’s t test (*P < 0.05 and **P < 0.01).

OsHOX1 and OsHOX28 Act Upstream of HSFA2D and LA1

We hypothesized that OsHOX1 and OsHOX28 might act upstream of the HSFA2D-LA1 pathway in the regulation of tiller angle, because the phenotypes of the overexpression lines (OsHOX1-OX and OsHOX28-OX) were similar to the phenotype of the la1 mutant. To test this hypothesis, we compared the expression of HSFA2D and LA1 in seedlings of the overexpression lines and wild-type plants. Interestingly, ectopic expression of OsHOX1 and OsHOX28 significantly suppressed the expression of both HSFA2D and LA1 (Fig. 4A). Consistent with these results, the OsHOX1/28-CR lines displayed the opposite effects (Fig. 4A). However, no significant difference in the expression of either OsHOX1 or OsHOX28 was observed between the seedlings of the LA1 CRISPR edited lines (LA1-CR) and the wild type (Supplemental Fig. S4, A–C). These results clearly demonstrated that OsHOX1 and OsHOX28 act upstream of HSFA2D and LA1.

Figure 4.

Figure 4.

Open in a new tab

Regulation of HSFA2D by OsHOX1 and OsHOX28 in vitro and in vivo. A, Expression analysis of HSFA2D and LAZY1 (LA1) in 14-d-old seedlings of OsHOX1-OX, OsHOX28-OX, and OsHOX1/28-CR transgenic lines. Data are shown as the means ± se of three replicates. Asterisks indicate significant difference by Student’s t test (*P < 0.05 and **P < 0.01)). B, Gene structure of HSFA2D. Empty boxes on the left and right sides represent the 5′ and 3′ UTRs, respectively; black boxes represent exons; lines between black boxes represent introns; and lines before the 5′ and after the 3′ UTRs are sequences up- and downstream of genes. Fragments p1 and p2 indicate two CAAT[G/C]ATTG HD-ZIP II binding cis-elements (highlighted in red) within the promoter of HSFA2D. C, Yeast one-hybrid assay showing that OsHOX1 and OsHOX28 interacted with fragment p12, which contains two copies of the CAAT[G/C]ATTG cis-element at −1,522 and −1,440 bp upstream of HSFA2D. D, Reporter constructs used in the transient LUC assay. The reporter gene LUC was driven by the CaMV 35S promoter. The fragment −1,736 to −1,227 bp upstream of the promoter of HSFA2D and its mutated derivatives were inserted in the region flanked by the CaMV 35S promoter and TATA box. Wild type (WT) represents the inserted fragment with the original CAAT[G/C]ATTG cis-element (red characters) at both the p1 and p2 sites, while mp1 and mp2 represent the inserted fragments with mutation in the cis-element to poly-A (blue lowercase characters) at the p1 and p2 sites, respectively. mp12 represents a mutation in the cis-element at both the p1 and p2 sites. E, Effector constructs used in the transient LUC assay. The full-length ORFs of OsHOX1 and OsHOX28 were driven by the CaMV 35S promoter, which generated the effectors of each gene. A construct with no gene inserted (None) was used as a negative control. F, Relative LUC activity of combinations of the different effectors and reporters described above. Relative LUC activity is represented by the ratio of signal values of firefly LUC to that of Renilla LUC (internal control). Data are shown as the means ± sd of three independent transformants. Different lowercase letters indicate significant differences among the different reporters (wild type, mp1, mp2, and mp12) under the control of the same effector. Different uppercase and lowercase letters indicate significant differences among the different effectors (None, OsHOX1, and OsHOX28) with the same reporter by Duncan’s test (P < 0.05). G, OsHOX1 directly associates with the promoter of HSFA2D. ChIP assays were performed with 25-d-old wild-type and OsHOX1-FLAG transgenic line 3 plants. The precipitated fragments were analyzed by qPCR using the primer sets indicated by the short solid lines with the primer names shown below. Data represent the means ± sd of three biological replicates. Asterisks indicate significant difference by Student’s t test (**P < 0.01). H, OsHOX1 and OsHOX28 proteins interact with the p1 and p2 regions of the HSFA2D promoter individually. Probes (53 bp; Supplemental Table S1) containing the cis-element at the p1 or p2 site were used for these assays. The 25- and 100-fold excess unlabeled probes were used for competitions.

OsHOX1 and OsHOX28 Suppress the Expression of HSFA2D by Binding to Its Promoter

Two linked HD-ZIP II binding cis-elements (p1 and p2; Fig. 4B were identified 2 kb upstream of HSFA2D, which suggested the possibility of potential direct interactions between OsHOX1/OsHOX28 and the promoter of HSFA2D. A yeast one-hybrid assay demonstrated that both OsHOX1 and OsHOX28 interacted with fragment p12, which contains both p1 and p2 (Fig. 4, B and C). To further confirm these regulatory patterns, transient expression assays were performed in rice protoplasts. A fragment containing both p1 and p2 cis-elements (−1,736 to −1,227 kb upstream of the promoter of HSFA2D) and its derivatives with mutation(s) in p1 and/or p2 were inserted between the cauliflower mosaic virus (CaMV) 35S promoter and a LUC reporter gene, constituting reporter constructs (Fig. 4D). The coding sequences of OsHOX1 and OsHOX28 were driven by the CaMV 35S promoter and used as effectors (Fig. 4E). Compared with the negative control (effector None), the OsHOX1 and OsHOX28 effectors exhibited significant suppressive effects on the wild-type reporter. However, mutations within both cis-elements p1 and p2 (mp12) totally abolished these suppressive effects, because the relative LUC activities were restored to the levels of the negative control. Under control of either the OsHOX1 or OsHOX28 effector, intermediate relative LUC activities were detected in reporters with mutation in either p1 (mp1) or p2 (mp2), compared with that of the wild type and mp12 (Fig. 4F). Altogether, these results suggested that suppression of HSFA2D by OsHOX1 or OsHOX28 requires an intact cis-element of either p1 or p2. The EMSAs further revealed that OsHOX1 and OsHOX28 directly interact with both p1 and p2 individually (Fig. 4G). Consistent with these results, ChIP-PCR analysis further verified the binding of OsHOX1 to p12 cis-elements in vivo (Fig. 4H). Taken together, these results clearly support that OsHOX1 and OsHOX28 bound to the binding sites CAAT[G/C]ATTG within the promoter of HSFA2D.

Restrained LAT and Asymmetric Expression of WOX6 and WOX11 by OsHOX1 and OsHOX28

The establishment of asymmetric distribution of auxin in the upper and lower sections of responding stems through LAT is a critical step in gravitropism responses. Loss of function of HSFA2D and LA1 largely abolishes LAT, leading to a deficiency in shoot gravitropism (Li et al., 2007; Zhang et al., 2018). To study whether OsHOX1 and OsHOX28 affect this process, we assessed the capability of LAT in the transgenic lines and wild-type plants. 3H-IAA was delivered to the apical end of coleoptile sections, which were then positioned horizontally for incubation for 2.5 h. Relative radioactivity was measured thereafter for both the upper and lower sides (Fig. 5A). Significantly less accumulation of 3H-IAA was detected in the lower side than in the upper side in the OsHOX1 and OsHOX28 overexpression lines compared to wild-type plants, while the OsHOX1/28-CR line presented greater accumulation in the lower side (Fig. 5B). To further confirm the effects of OsHOX1 and OsHOX28 on LAT, the expression levels of the auxin-responsive marker gene OsIAA20 were monitored before and after gravitropic stimulation. The ratio of expression in the lower side to that in the upper side (LS/US expression ratio) was used to evaluate the local auxin distribution. Before treatment (0 h), no difference of LS/US expression ratio was detected in overexpression lines, the double edited line, or the wild type (Fig. 5C). However, after 6 h of gravitropism treatment, the LS/US expression ratio of OsIAA20 in overexpression lines was markedly lower than in the wild type. In contrast, the OsHOX1/28-CR line showed a higher ratio than the wild type, although not statistically significant. Furthermore, LS/US expression ratios of WOX6 and WOX11, which function downstream of HSFA2D, were similar to that of OsIAA20 (Fig. 5, D and E). Collectively, these results demonstrated that OsHOX1 and OsHOX28 largely affected the asymmetric distribution of auxin by inhibiting LAT progress, thus restraining the differential expression of WOX6 and WOX11 in the upper and lower sides. In addition, we compared the auxin content among transgenic lines and the wild type. Significantly reduced and increased amounts of IAA were detected in overexpression lines and the double edited line, respectively (Fig. 5F). Consistently, expression levels of four members of the OsYUC gene family (OsYUC1, OsYUC5, OsYUC6, and OsYUC7) that are responsible for the biosynthesis of auxin were suppressed in overexpression lines of OsHOX1 and OsHOX28 but enhanced in the double edited line (Fig. 5G). Taken together, these findings suggest that OsHOX1 and OsHOX28 inhibit auxin biosynthesis and restrain LAT and asymmetric expression of WOX6 and WOX11.

Figure 5.

Figure 5.

Open in a new tab

Regulation of the asymmetric distribution of auxin and the asymmetric expression of OsIAA20, WOX6, and WOX11 by OsHOX1 and OsHOX28. A, Schematic diagram of splicing of the shoot base into upper and lower sides for expression analysis. B, Comparison of the lateral distribution of 3H-IAA (LAT) in dark-grown wild-type (WT), OsHOX1-OX line 1, OsHOX28-OX line 7, and OsHOX1/28-CR coleoptiles of seedlings subjected to gravitropic treatment. The ratio indicates the relative radioactivity of the lower side to that of the upper side of the coleoptiles in response to gravistimulation. Data are shown as the means ± sd of four independent replicates. C to E, Comparison of the LS/US expression ratio of OsIAA20 (C), WOX6 (D), and WOX11 (E) in seedlings of wild-type and OsHOX1-OX, OsHOX28-OX, and OsHOX1/28-CR transgenic plants subjected to gravitropic treatment. Data are shown as the means ± sd of three replicates. F, Comparison of IAA content in 10-d-old seedlings of wild-type, OsHOX1-OX, OsHOX28-OX, and OsHOX1/28-CR lines. G, The expression levels of OsYUC genes in 14-d-old seedlings of OsHOX1-OX, OsHOX28-OX, and OsHOX1/28-CR transgenic lines and the wild type. Data are shown as the means ± sd of three replicates. Different lowercase letters indicate significant differences among the indicated genes by Duncan’s test (P < 0.05).

OsHOX1 and OsHOX28 Regulate the Tiller Angle Mainly by Suppressing HSFA2D

To further confirm the relationship between OsHOX1/28 and HSFA2D, transgenic lines overexpressing both OsHOX1 and HSFA2D or both OsHOX28 and HSFA2D were generated (Supplemental Fig. S5). OsHOX1, OsHOX28, and HSFA2D were indeed overexpressed in their corresponding overexpression lines (Fig. 6, A and B). As expected, the expression of LA1, which acts downstream of HSFA2D, was significantly higher in the double-overexpression lines than in the wild-type and single-overexpression lines (OsHOX1-OX and OsHOX28-OX; Fig. 6, A and B). Moreover, internal expression of HSFA2D was also inhibited in the double-overexpression lines (Supplemental Fig. S5, C–E). The horizontal tiller angle of both double-overexpression lines was between that of the overexpression lines (OsHOX1-OX and OsHOX28-OX) and wild-type plants (Fig. 6, C–I). These results demonstrated that HSFA2D overexpression partially rescued the tiller angle of OsHOX1-OX and OsHOX28-OX plants.

Figure 6.

Figure 6.

Open in a new tab

Genetic and transcriptional expression analyses of OsHOX1, OsHOX28, and HSFA2D. A, Expression analysis of OsHOX1, HSFA2D, and LA1 in wild-type (WT), OsHOX1-OX, OsHOX1 and HSFA2D double-overexpression (OsHOX1/HSFA2D-OX) lines. Data are shown as the means ± sd of three independent replicates. B, Expression analysis of OsHOX28, HSFA2D, and LA1 in wild-type, OsHOX28-OX, OsHOX28 and OsHOX28/HSFA2D-OX lines. Data are shown as the means ± sd of three independent replicates. Different lowercase letters indicate significant differences among the indicated genes by Duncan’s test (P < 0.05). C to H, Phenotypes of the horizontal tiller angle of wild-type, OsHOX1-OX, OsHOX1/HSFA2D-OX, OsHOX28-OX, and OsHOX28/HSFA2D-OX lines under low-density growing conditions in the natural field. Scale bars = 25 cm. I, Horizontal tiller angles of transgenic lines described in B under the conditions described in Figure 1A. Data are shown as the means ± sd (n = 10). Different lowercase letters indicate significant differences by Duncan’s test (P < 0.05).

DISCUSSION

Regulation of Tiller Angle by OsHOX1 and OsHOX28 Mainly Occurs via the HSFA2D-LA1 Pathway

A recent study revealed a core pathway in which HSFA2D acts upstream of LA1 and connects the gravitropic response to the process of tiller angle establishment in rice (Zhang et al., 2018). In this study, we investigated the functions of two HD-ZIP II transcription factors, OsHOX1 and OsHOX28, in the regulation of rice tiller angle. Expression levels of HSFA2D and LA1 were repressed in OsHOX1 and OsHOX28 overexpression lines but enhanced in the OsHOX1/28-CR lines (Fig. 4A), which indicates that OsHOX1 and OsHOX28 act upstream of the HSFA2D-LA1 pathway. Our LUC assay revealed that mutation of p1 or p2 alone attenuated the repressive effects of both OsHOX1 and OsHOX28, although it was not statistically significant in the case of OsHOX1 (Fig. 4, D–F). Mutations in both p1 and p2 totally abolished the suppressive effects of both OsHOX1 and OsHOX28 (Fig. 4, D–F). These results, together with those of yeast one-hybrid assays (Fig. 4, B and C) and EMSAs (Fig. 4H), consistently demonstrated that both OsHOX1 and OsHOX28 directly targeted HSFA2D by binding to two adjacent CAAT[G/C]ATTG cis-elements within its promoter. Multiple shifted bands were obtained in EMSAs of OsHOX1 and OsHOX28 (Figs. 3F and 4G), consistent with previous studies of the HD-ZIP proteins REVOLUTA (Xie et al., 2014) and ATHB-52 (Miao et al., 2018). It was reported that both the monomer and dimer of HD-ZIP proteins are able to bind DNA although the monomeric form possesses a weaker affinity (Palena et al., 1999). Furthermore, OsHOX1 and OsHOX28 were able to form homodimers (Fig. 2, D and E). Thus, the greater shifts in EMAS are probably supershifted complexes containing dimer forms of OsHOX1 and OsHOX28 (Figs. 3F and 4G).

As an indispensable step in the gravitropism response, LAT promotes the asymmetric accumulation of auxin such that the auxin content in the lower side is greater than that in the upper side in the responding organ, which results in a high ratio of auxin in the lower side to the upper side, inducing an erect-growth phenotype. Diminished LAT in both the la1 mutant and the mutant of its upstream regulator hsfa2d attenuated the gravitropic response and thus increased the tiller angle of rice (Li et al., 2007; Zhang et al., 2018). Here, through estimating the transportation of isotope-labeled auxin in response to gravitropic treatments, we found that OsHOX1 and OsHOX28 negatively regulate LAT progress (Fig. 5, A and B). This result was further confirmed by the finding that the auxin response marker gene OsIAA20 exhibited a reduced LS/US expression ratio in overexpression lines (Fig. 5C). However, the OsHOX1/28-CR line exhibited a statistically comparable expression ratio of OsIAA20 to that of the wild type (Fig. 5C), which may be attributed to the weak tiller angle phenotypic change in the OsHOX1/28-CR line. Failure to establish asymmetrical distribution of auxin in the OsHOX1-OX and OsHOX28-OX lines reduced the extent of induction of WOX6 and WOX11 expression in the lower side of the shoot base (Fig. 5, D and E), the effects of which are similar to those of mutations in la1 or hsfa2d (Li et al., 2007; Zhang et al., 2018). These results further indicated the involvement of OsHOX1 and OsHOX28 in the HSFA2D-LA1 pathway. Notably, the tiller angle of both double-overexpression lines (OsHOX1/HSFA2D-OX and OsHOX28/HSFA2D-OX) was between that of the wild-type plants and the single-overexpression lines (OsHOX1-OX and OsHOX28-OX; Fig. 6, C–I), which suggested that OsHOX1 and OsHOX28 function partially through the HSFA2D-LA1 pathway.

Interestingly, our results demonstrated that OsHOX1 and OsHOX28 act as negative regulators of auxin biosynthesis and thus reduce the endogenous auxin content of rice seedlings (Fig. 5, F and G). In contrast, TAC4 enhances shoot gravitropism by increasing auxin content through upregulating the expression of OsYUC genes and affecting auxin distribution (Li et al., 2020). Moreover, a recent study revealed that increased auxin biosynthesis caused by mutations in genes regulating SL biosynthesis or signaling could increase auxin levels at the lower side of the shoot base subjected to gravistimulation and promote more compact growth of tillers. This SL-mediated regulation of auxin biosynthesis and gravitropism response was proposed to be independent of the LA1 pathway, since the la1 mutation does not affect auxin content in rice seedlings (Sang et al., 2014). Hence, we propose that apart from the HSFA2D-LA1 pathway, it is possible that reduction of auxin biosynthesis by OsHOX1 and OsHOX28 may also contribute to the regulation of tiller angle establishment (Fig. 7), which could explain the partial restoration of the horizontal tiller angle in the double overexpression lines (Fig. 6, C–I).

Figure 7.

Figure 7.

Open in a new tab

Working model of the mechanism by which OsHOX1 and OsHOX28 regulate tiller angle establishment. The expression of each of the OsHOX1 and OsHOX28 genes is tightly regulated by its own encoded protein and by the protein encoded by the other gene either directly (solid lines) or indirectly (dotted lines). An unknown factor(s) (green box with a question mark) is proposed to interact with OsHOX1 and guide it to the promoter of OsHOX28. Moreover, both OsHOX1 and OsHOX28 are regulated by other unknown regulators (triangles with question marks). The OsHOX1 and OsHOX28 proteins bind to the promoter of HSFA2D and suppress its expression and that of LA1. In addition, both OsHOX1 and OsHOX28 negatively regulate auxin biosynthesis and content. The repressed LA1 expression and the reduced auxin content together affect the local auxin distribution and thus the differential expression of WOX6 and WOX11 in the upper and lower sides of the shoot base, ultimately decreasing the horizontal tiller angle.

Conserved Self- and Mutual Regulation of HD-ZIP II Genes in Rice and Arabidopsis

Endogenous expression of OsHOX1 was greatly suppressed in both OsHOX1-OX and OsHOX28-OX plants (Fig. 3C). The results of our yeast one-hybrid assays and EMSAs further suggested that both OsHOX1 and OsHOX28 bound to the promoter of OsHOX1 (Fig. 3E). Likely, exogenous overexpression of OsHOX28 suppressed the endogenous expression of OsHOX28 and OsHOX1 (Fig. 3D). Although no potential CAAT[G/C]ATTG binding element was found within 5 kb upstream of OsHOX28 (Fig. 3B), ChIP-PCR analysis revealed high enrichment of OsHOX1 protein at the p3, p4, and p5 sites within 2 kb upstream of OsHOX28 (Fig. 3G), which suggested that regulation of OsHOX28 by OsHOX1 could be facilitated by another partner that could target to the promoter of OsHOX28. Therefore, intricate negative feedback loops between OsHOX1 and OsHOX28 and for each individual component are proposed to control their expression (Fig. 7).

A similar regulatory pattern was reported in Arabidopsis. Overexpression of HAT2 not only suppressed endogenous expression of HAT2 but also strongly repressed the expression of most HD-ZIP II members in Arabidopsis (Supplemental Fig. S2), including HAT1, HAT3, HAT9, HAT22, ATHB-2, and ATHB-4 (Sawa et al., 2002). ATHB-2 likely was able to regulate its own expression as well as that of HAT1, HAT2, HAT3, HAT9, and HAT22 (Ohgishi et al., 2001). Furthermore, overexpression of ATHB-4 suppressed the expression of ATHB-2, HAT1, HAT2, HAT3, and HAT22 (Sorin et al., 2009). Taken together, these findings suggested that HD-ZIP II members share a conserved self- and mutual regulatory pattern in rice and Arabidopsis. This fine-tuned regulatory mechanism may assist the establishment of homeostasis, which stabilizes the output of the signal to ensure the normal development of plant architecture.

OsHOX1 and OsHOX28 Function Redundantly in the Regulation of Tiller Angle Establishment

Gene redundancy among HD-ZIP II genes has been reported in Arabidopsis. Overexpression of individual HD-ZIP II genes such as HAT1 (Ciarbelli et al., 2008), HAT2 (Sawa et al., 2002), HAT3, ATHB-4 (Sorin et al., 2009), and ATHB-2 (Steindler et al., 1999) resulted in enhanced hypocotyl elongation, whereas no phenotypic differences were observed in single-mutant lines (Sorin et al., 2009). Here, we found that overexpression of OsHOX1 or OsHOX28 resulted in an enhanced tiller angle (Fig. 1, A–C; Supplemental Fig. S3, A and B), and CRISPR editing of either of these two genes resulted in no phenotypic changes under either condition (Supplemental Fig. S3, C–J). Heterodimer formation between OsHOX1 and OsHOX28 (Fig. 2, D and E) implies that they may cooperatively participate in the regulation of tiller architecture, which could be the reason for no phenotypic changes observed in the single-edited lines. However, the double-edited line presented a subtle but significant change in tiller angle under growing conditions with low planting density (Fig. 1, D and F). Since the OsHOX1/28-CR lines 1 and 2 showed undistinguishable horizontal tiller angles under low planting density conditions (Fig. 1F), we speculated that one amino acid deletion in OsHOX1 of OsHOX1/28-CR line 1 should have the same effect as the frameshift mutation in OsHOX1/28-CR line 2. Thus, OsHOX1 and OsHOX28 act redundantly in the regulation of tiller angle. The close relationship among OsHOX2, OsHOX17, OsHOX1, and OsHOX28 (Supplemental Fig. S2) implies that both OsHOX2 and OsHOX17 probably possess functions similar to those of OsHOX1 and OsHOX28 in the regulation of tiller architecture. Additional analyses with higher-order mutants may help to confirm this hypothesis. A previous report also revealed mutations in another HD-ZIP II gene, OsHOX3/SGD2, alone dramatically reduced plant height and seed size in rice, both of which were further explained by altered expression of multiple genes involved in GA biosynthesis (Chen et al., 2019). Functional divergence likely exists within the HD-ZIP II subfamily in rice, which needs to be further tested for other members. It will also be interesting to address in future studies why tiller number was reduced in both OsHOX1 and OsHOX28 overexpression lines (Supplemental Fig. S3O).

MATERIALS AND METHODS

Plant Growing Conditions

Rice (Oryza sativa) plants were grown in the experimental field of Huazhong Agricultural University in Wuhan (30.4°N, 114.2°E) under normal (16.7 × 26.7 cm) and low (33.4 × 53.4 cm) planting-density conditions. Horizontal tiller angle (Fig. 1A, α) was measured with a digital horizontal ruler, and the mean value of three replicates generated from measuring three different tillers was used for each individual plant.

Generation of Constructs and Genetic Transformation

The full-length open reading frames (ORFs) were PCR amplified using primers HOX1-UF and HOX1-UR for OsHOX1 and HOX28-UF and HOX28-UR for OsHOX28 and then inserted into pU1301 according to the Gibson Assembly method (Gibson et al., 2009). Similarly, the full-length ORF of OsHOX1 was amplified with primers HOX1-UFLAG-F and HOX1-UFLAG-R and then inserted into a pU1301-FLAG vector to generate a FLAG-tagged OsHOX1 overexpression construct (OsHOX1-FLAG).

To generate CRISPR constructs for OsHOX1 and OsHOX28 single mutants, the OsU3 promoter was used to drive the single-guide RNA (sgRNA) scaffold. Target sequences were designed with the online CRISPR-P tool (http://crispr.hzau.edu.cn/CRISPR2/) and then inserted into the sgRNA scaffold via overlap PCR extension. The OsU3-sgRNA fragment was further inserted into a pCXUN-CAS9 backbone at the KpnI site according to the Gibson Assembly method (Gibson et al., 2009). To generate a construct targeting two genes, the OsU6 promoter was used to drive sgRNA containing the target sequence of OsHOX28 (OsU6-sgRNA-OsHOX28), and the OsU3 promoter was used to drive sgRNA containing the target sequence of OsHOX1 (OsU3-sgRNA-OsHOX1). Thereafter, these two fragments were inserted into the pCXUN-CAS9 backbone at the KpnI and SacI sites sequentially according to the Gibson Assembly method. Information concerning the primers is listed in Supplemental Table S1.

To generate constructs overexpressing both OsHOX1 and HSFA2D, full-length ORFs of each gene were amplified with corresponding primers (HOX1-bamh1F and HOX1-bamh1R for OsHOX1, HSF-35SF and HSF-35SR for HSFA2D) and then inserted sequentially into a pCX vector backbone (Supplemental Fig. S5A) at the BamHI and SpeI sites. Similar steps were performed to generate a construct for the overexpression of both HSFA2D and OsHOX28 by using primers HOX28-bamh1F and HOX28-bamh1R for amplification of OsHOX28 (Supplemental Fig. S5B).

All of these constructs were transformed into the japonica variety cv Zhonghua 11 (ZH11, Oryza sativa ssp.) through Agrobacterium tumefaciens-mediated transformation.

Shoot Gravitropism Assays

Seeds were sterilized with 0.15% (w/v) mercury chloride and germinated on one-half strength Murashige and Skoog (MS) medium (pH 5.8) at 28°C for 48 h. The seeds with the same state were then transferred to square plates that contained one-half strength MS medium and the plates were maintained in a vertical position at 28°C. Four-day-old seedlings were subjected to gravitropic treatment by reorienting the square plates by 90°. The shoot curvature angles were measured with the ImageJ program (https://imagej.nih.gov/ij/) after 36 h of treatment.

Quantification of Gene Expression

Total RNA from tested tissues was extracted using TRIzol reagent (TransGen Biotech) according to the manufacturer’s instructions. Genomic DNA within the samples was removed using DNaseI (Invitrogen), and first-strand complementary DNA (cDNA) was synthesized using Moloney murine leukemia virus reverse transcriptase (Invitrogen) according to the manufacturer’s instructions. Reverse transcription quantitative PCR (RT-qPCR) was performed using gene-specific primers (Supplemental Table S1) in a reaction system containing FastStart Universal SYBR Green Master Mix (Roche). A Quant-Studio 6 Flex Real-Time PCR System (Applied Biosystems, Thermo Fisher Scientific) was used to perform the RT-qPCR according to the manufacturer’s instructions. The PCR amplifications were conducted in triplicate for each sample of three independent biological replicates, and the rice ubiquitin gene (Os03g0234200) was used for normalization.

Yeast One-Hybrid Assays

With respect to yeast one-hybrid assays, full-length ORFs of OsHOX1 and OsHOX28 were cloned into a pJG45AD vector using corresponding primers and then cotransformed along with a pLacZi2μ vector, which contains a LacZ reporter gene driven by the 1p2 fragment of the OsHOX1 promoter or the p12 fragment of the HSFA2D promoter, into an EGY48 yeast strain. The resulting transformants were subsequently grown on SD/-Ura-Trp plates at 30°C. Single clones were then transferred to new SD/-Ura-Trp plates that contained 5-bromo-4-chloro-3-indolyl-β-d-galactopyranoside for blue color development at 30°C. Information about the primers used to generate the constructs is shown in Supplemental Table S1.

EMSAs

EMSAs were performed using a LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific) according to the manufacturer’s instructions. By using corresponding primers (Supplemental Table S1), full-length ORFs of OsHOX1 and OsHOX28 were cloned into pGEX-4T-1 and pGEX-6p-1 vectors, respectively. Glutathione S-transferase (GST)-tagged OsHOX1 and OsHOX28 proteins were expressed in Escherichia coli strain BL21 (DE3; TransGen Biotech) and then subjected to purification with glutathione sepharose beads (GE Healthcare) according to the manufacturer’s instructions. Biotin-labeled and unlabeled single-stranded DNA oligonucleotides corresponding to the OsHOX1- and OsHOX28-binding regions within the promoters of OsHOX1 (Supplemental Table S1, 1p2) and HSFA2D (Supplemental Table S1, p1 and p2) were synthesized and renatured to produce biotin-labeled probes and unlabeled probes, respectively. The indicated proteins and labeled probes or unlabeled probes were incubated together in 20 μL reaction mixtures consisting of 10 mm Tris-HCl (pH 7.5), 50 mm KCl, 1 mm dithiothreitol, and 50 ng mL−1 poly(deoxyinosinic-deoxycytidylic) for 20 min at room temperature and then separated on a 6% native polyacrylamide gel.

Rice Protoplast Preparation and Transformation

Rice protoplasts were isolated from 14-d-old seedlings of cv ZH11 as described previously (Xie and Yang, 2013), with some modifications. The rice protoplasts were isolated by incubating rice sheath strips with digestion solution (10 mm MES [pH 5.7], 0.6 m mannitol, 1 mm CaCl2, 5 mm β-mercaptoethanol, 0.1% [w/v] bovine serum albumin, 0.3% [w/v] Cellulase RS [Yakult Honsha], and 0.75% [w/v] Macerozume R10 [Yakult Honsha]) together for 4 h with gentle shaking at 25°C. The protoplasts were then incubated in W5 solution (2 mm MES [pH 5.7], 154 mm NaCl, 5 mm KCl, and 125 mm CaCl2) for 30 min, collected via centrifugation at 100g for 8 min, and ultimately resuspended in MMG solution (4 mm MES [pH 5.7], 0.6 m mannitol, and 15 mm MgCl2). For transformation, 6 μg of plasmids, 100 μL of protoplasts, and 110 μL of polyethylene glycol-CaCl2 solution (0.6 m mannitol, 100 mm CaCl2, and 40% [w/v] polyethylene glycol 4000) were mixed together, after which the mixture was incubated at room temperature for 15 min in the dark. Two volumes of W5 solution were added to stop the transformation. The transformed protoplasts were collected by centrifugation and resuspended in WI solution (4 mm MES [pH 5.7], 0.6 m mannitol, and 4 mm KCl), after which they were cultured in 2-mL tubes for 12 h in the dark. The protoplasts were ultimately collected by centrifugation at 100g for 8 min.

Transcriptional Regulation Assay in Protoplasts

To analyze the transcriptional activity of OsHOX1 and OsHOX28, the fragment encoding the DNA-binding domain of the GAL4 protein (GAL4DBD; Ma and Ptashne, 1987) was driven by the CaMV 35S promoter; this construct was used as a negative effector control. With respect to the positive control, the strong activation domain from VP16 of the herpes simplex virus (Triezenberg et al., 1988) was fused to the GAL4DBD to generate a GAL4DBD-VP16 vector. OsHOX1 and OsHOX28 were then fused to either the C or N terminus of the VP16 domain in the GAL4DBD-VP16 vector to generate the remaining effectors. With respect to the reporter construct, the GAL4-binding site multimerized five times together with a minimal TATA box were placed upstream of the firefly LUC reporter gene. To validate the direct regulations of OsHOX1 and OsHOX28 on HSFA2D, the full-length ORFs of OsHOX1 and OsHOX28 driven by the CaMV 35S promoter were used as the effectors. The −1,736 to −1,227 bp upstream fragment from the promoter of HSFA2D and its mutated derivatives were inserted upstream of firefly LUC to generate the reporter constructs. As a reference, the Renilla LUC gene (Promega) under the control of the CaMV 35S promoter (35S: RLUC) was used. The effector and reporter constructs, together with the reference construct, were cotransformed into rice protoplast in a ratio of 5:5:1 (effector:reporter:reference). The transformed protoplasts were cultured for 12 h at 25°C in the dark, collected by centrifugation, and then quantified for LUC activity. The luciferase activity was measured using the Dual Luciferase Reporter Assay System (Promega) according to the manufacturer’s instructions. The relative LUC activity was indicated as the ratio of firefly LUC signal values to those of Renilla LUC for three independent transformants.

ChIP Assay

For the ChIP assay, 25-d-old wild-type and OsHOX1-FLAG transgenic lines were used for chromatin extraction and immunoprecipitation as described previously (Zhang et al., 2017), with some modification. Briefly, chromatin was extracted and fragmented with a Bioruptor Sonication Device (Diagenode) to 200 to 800 bp. The fragmented chromatin suspension was incubated with washed Anti-FLAG M2 Magnetic Beads (Sigma-Aldrich) at 4°C overnight. After extensive washing and de-crosslinking, the precipitated and input DNA samples were analyzed by qPCR. Primers for ChIP-qPCR are listed in Supplemental Table S1. Data normalized with input DNA are shown as the means ± sd from three biological repeats. Each biological replicate was tested with three technical repeats.

LAT Assay

The polar auxin transport assays were performed as previously described with some modification (Li et al., 2007). Five-day-old dark-grown seedlings were used for the assay. One-centimeter segments from the basal ends were excised and incubated in one-half strength MS (pH 5.8) liquid medium for 2 h with gentle shaking to remove the endogenous IAA. Coleoptile segments were then laid horizontally in 0.6-mL tubes with their apical ends in the 10-μL agar blocks that contained 0.5 mm 3H-labeled IAA. After 2.5 h of treatment in the dark, 5-mm segments at the end away from the apex were split into upper and lower halves. Each half was incubated in 1 mL scintillation liquid (27 mm polyphenol oxidase, 0.27 mm 1,4-bis(5-phenyloxazol-2-yl) benzene, 60% [v/v] dimethylbenzene, and 40% [v/v] ethanol) for 18 h. Radioactivity of incubated scintillation liquid in each tube was counted with a liquid scintillation counter (1450 MicroBeta TriLux, PerkinElmer).

Yeast Two-Hybrid Assays

Constructs for yeast two-hybrid analysis were generated using pGBKT7 and pGADT7, which express protein fusions to the GAL4 DNA-binding domain or the activation domain, respectively. Full-length cDNA of OsHOX1 and OsHOX28 was inserted into both pGBKT7 and pGADT7 at the same EcoRI sites by the Gibson Assembly method. According to the Matchmaker Gold Yeast Two-Hybrid System manual, this analysis was performed in strain AH109 carrying HIS3 and MEL1 reporters for reconstituted GAL4 activity.

LUC Complementation Imaging Assays

Full-length cDNA of OsHOX1 and OsHOX28 was inserted into both pCAMBIA1300-nLUC and pCAMBIA1300-cLUC (Chen et al., 2008) at the SalI and KpnI/SalI sites, respectively, by the Gibson Assembly method. Each of the nLUC- or cLUC-fused plasmids was transformed into A. tumefaciens strain EHA105 and then infiltrated into leaves of Nicotiana benthamiana. Two days after infiltration, the LUC activity was measured using an in vivo imaging system (Tanon) with potassium luciferin as the substrate.

Phylogenetic Analysis

The phylogenetic analysis was inferred by using the maximum-likelihood method and the Poisson correction model with MEGA X (Zuckerkandl and Pauling, 1965; Kumar et al., 2018). The bootstrap consensus tree inferred from 1,000 replicates is taken to represent the evolutionary history of the taxa analyzed. Branches corresponding to partitions reproduced in <50% of bootstrap replicates are collapsed. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (1,000 replicates) is shown next to each branch. Protein sequences of Arabidopsis and rice HD-ZIP II members were downloaded from websites http://www.arabidopsis.org/ and http://rice.plantbiology.msu.edu/index.shtml, respectively.

Statistical Analysis

Statistical analyses were performed with SPSS 20. For comparison of two groups, we used Students’s t test. For multiple comparisons, ANOVA was used followed by Duncan’s test, as mentioned in the figure legends.

Accession Numbers

Sequence data form this article can be found in the Rice Genome Annotation Project under the following accession numbers: LOC_Os10g41230 (OsHOX1) and LOC_Os06g04850 (OsHOX28).

Supplemental Data

The following supplemental materials are available.

Acknowledgments

We thank Jianbo Wang for his excellent work in the field.

Footnotes

1

This study was supported by the National Key Research and Development Program of China (grant no. 2016YFD0100403) and the National Natural Science Foundation of China (grant nos. 91935302, 31771751, and 31821005).

References

  1. Brandt R, Cabedo M, Xie Y, Wenkel S(2014) Homeodomain leucine-zipper proteins and their role in synchronizing growth and development with the environment. J Integr Plant Biol 56: 518–526 [DOI] [PubMed] [Google Scholar]
  2. Chen H, Zou Y, Shang Y, Lin H, Wang Y, Cai R, Tang X, Zhou J-M(2008) Firefly luciferase complementation imaging assay for protein-protein interactions in plants. Plant Physiol 146: 368–376 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Chen R, Rosen E, Masson PH(1999) Gravitropism in higher plants. Plant Physiol 120: 343–350 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Chen W, Cheng Z, Liu L, Wang M, You X, Wang J, Zhang F, Zhou C, Zhang Z, Zhang H, et al. (2019) Small Grain and Dwarf 2, encoding an HD-Zip II family transcription factor, regulates plant development by modulating gibberellin biosynthesis in rice. Plant Sci 288: 110208. [DOI] [PubMed] [Google Scholar]
  5. Ciarbelli AR, Ciolfi A, Salvucci S, Ruzza V, Possenti M, Carabelli M, Fruscalzo A, Sessa G, Morelli G, Ruberti I(2008) The Arabidopsis homeodomain-leucine zipper II gene family: Diversity and redundancy. Plant Mol Biol 68: 465–478 [DOI] [PubMed] [Google Scholar]
  6. Dardick C, Callahan A, Horn R, Ruiz KB, Zhebentyayeva T, Hollender C, Whitaker M, Abbott A, Scorza R(2013) PpeTAC1 promotes the horizontal growth of branches in peach trees and is a member of a functionally conserved gene family found in diverse plants species. Plant J 75: 618–630 [DOI] [PubMed] [Google Scholar]
  7. Dong H, Zhao H, Xie W, Han Z, Li G, Yao W, Bai X, Hu Y, Guo Z, Lu K, et al. (2016) A novel tiller angle gene, TAC3, together with TAC1 and D2 largely determine the natural variation of tiller angle in rice cultivars. PLoS Genet 12: e1006412. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Elhiti M, Stasolla C(2009) Structure and function of homodomain-leucine zipper (HD-Zip) proteins. Plant Signal Behav 4: 86–88 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Furutani M, Hirano Y, Nishimura T, Nakamura M, Taniguchi M, Suzuki K, Oshida R, Kondo C, Sun S, Kato K, et al. (2020) Polar recruitment of RLD by LAZY1-like protein during gravity signaling in root branch angle control. Nat Commun 11: 76. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Gibson DG, Young L, Chuang RY, Venter JC, Hutchison CA III, Smith HO(2009) Enzymatic assembly of DNA molecules up to several hundred kilobases. Nat Methods 6: 343–345 [DOI] [PubMed] [Google Scholar]
  11. Harmoko R, Yoo JY, Ko KS, Ramasamy NK, Hwang BY, Lee EJ, Kim HS, Lee KJ, Oh DB, Kim DY, et al. (2016) N-glycan containing a core α1,3-fucose residue is required for basipetal auxin transport and gravitropic response in rice (Oryza sativa). New Phytol 212: 108–122 [DOI] [PubMed] [Google Scholar]
  12. Harris JC, Hrmova M, Lopato S, Langridge P(2011) Modulation of plant growth by HD-Zip class I and II transcription factors in response to environmental stimuli. New Phytol 190: 823–837 [DOI] [PubMed] [Google Scholar]
  13. Hu M, Lv S, Wu W, Fu Y, Liu F, Wang B, Li W, Gu P, Cai H, Sun C, et al. (2018) The domestication of plant architecture in African rice. Plant J 94: 661–669 [DOI] [PubMed] [Google Scholar]
  14. Jin J, Huang W, Gao J-P, Yang J, Shi M, Zhu M-Z, Luo D, Lin H-X(2008) Genetic control of rice plant architecture under domestication. Nat Genet 40: 1365–1369 [DOI] [PubMed] [Google Scholar]
  15. Kumar S, Stecher G, Li M, Knyaz C, Tamura K(2018) MEGA X: Molecular Evolutionary Genetics Analysis across computing platforms. Mol Biol Evol 35: 1547–1549 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Li H, Sun H, Jiang J, Sun X, Tan L, Sun C (2020) TAC4 controls tiller angle by regulating the endogenous auxin content and distribution in rice. Plant Biotechnol J doi:10.1111/pbi.13440 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Li P, Wang Y, Qian Q, Fu Z, Wang M, Zeng D, Li B, Wang X, Li J(2007) LAZY1 controls rice shoot gravitropism through regulating polar auxin transport. Cell Res 17: 402–410 [DOI] [PubMed] [Google Scholar]
  18. Li Z, Liang Y, Yuan Y, Wang L, Meng X, Xiong G, Zhou J, Cai Y, Han N, Hua L(2019) OsBRXL4 regulates shoot gravitropism and rice tiller angle through affecting LAZY1 nuclear localization. Molecular Plant 12: 1143–1156 [DOI] [PubMed] [Google Scholar]
  19. Ma J, Ptashne M(1987) Deletion analysis of GAL4 defines two transcriptional activating segments. Cell 48: 847–853 [DOI] [PubMed] [Google Scholar]
  20. Mao C, Ding W, Wu Y, Yu J, He X, Shou H, Wu P(2007) Overexpression of a NAC-domain protein promotes shoot branching in rice. New Phytol 176: 288–298 [DOI] [PubMed] [Google Scholar]
  21. Miao Z-Q, Zhao P-X, Mao J-L, Yu L-H, Yuan Y, Tang H, Liu Z-B, Xiang C-B(2018) HOMEOBOX PROTEIN52 mediates the crosstalk between ethylene and auxin signaling during primary root elongation by modulating auxin transport-related gene expression. Plant Cell 30: 2761–2778 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Ohgishi M, Oka A, Morelli G, Ruberti I, Aoyama T(2001) Negative autoregulation of the Arabidopsis homeobox gene ATHB-2. Plant J 25: 389–398 [DOI] [PubMed] [Google Scholar]
  23. Palena CM, Gonzalez DH, Chan RL(1999) A monomer-dimer equilibrium modulates the interaction of the sunflower homeodomain leucine-zipper protein Hahb-4 with DNA. Biochem J 341: 81–87 [PMC free article] [PubMed] [Google Scholar]
  24. Ponomarenko PM, Ponomarenko MP(2015) Sequence-based prediction of transcription upregulation by auxin in plants. J Bioinform Comput Biol 13: 1540009. [DOI] [PubMed] [Google Scholar]
  25. Sakuraba Y, Piao W, Lim J-H, Han S-H, Kim Y-S, An G, Paek N-C(2015) Rice ONAC106 inhibits leaf senescence and increases salt tolerance and tiller angle. Plant Cell Physiol 56: 2325–2339 [DOI] [PubMed] [Google Scholar]
  26. Sang D, Chen D, Liu G, Liang Y, Huang L, Meng X, Chu J, Sun X, Dong G, Yuan Y, et al. (2014) Strigolactones regulate rice tiller angle by attenuating shoot gravitropism through inhibiting auxin biosynthesis. Proc Natl Acad Sci USA 111: 11199–11204 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Sawa S, Ohgishi M, Goda H, Higuchi K, Shimada Y, Yoshida S, Koshiba T(2002) The HAT2 gene, a member of the HD-Zip gene family, isolated as an auxin inducible gene by DNA microarray screening, affects auxin response in Arabidopsis. Plant J 32: 1011–1022 [DOI] [PubMed] [Google Scholar]
  28. Scarpella E, Rueb S, Boot KJ, Hoge JH, Meijer AH(2000) A role for the rice homeobox gene Oshox1 in provascular cell fate commitment. Development 127: 3655–3669 [DOI] [PubMed] [Google Scholar]
  29. Sorin C, Salla-Martret M, Bou-Torrent J, Roig-Villanova I, Martínez-García JF(2009) ATHB4, a regulator of shade avoidance, modulates hormone response in Arabidopsis seedlings. Plant J 59: 266–277 [DOI] [PubMed] [Google Scholar]
  30. Steindler C, Matteucci A, Sessa G, Weimar T, Ohgishi M, Aoyama T, Morelli G, Ruberti I(1999) Shade avoidance responses are mediated by the ATHB-2 HD-zip protein, a negative regulator of gene expression. Development 126: 4235–4245 [DOI] [PubMed] [Google Scholar]
  31. Triezenberg SJ, Kingsbury RC, McKnight SL(1988) Functional dissection of VP16, the trans-activator of herpes simplex virus immediate early gene expression. Genes Dev 2: 718–729 [DOI] [PubMed] [Google Scholar]
  32. Turchi L, Baima S, Morelli G, Ruberti I(2015) Interplay of HD-Zip II and III transcription factors in auxin-regulated plant development. J Exp Bot 66: 5043–5053 [DOI] [PubMed] [Google Scholar]
  33. Wang B, Smith SM, Li J(2018) Genetic regulation of shoot architecture. Annu Rev Plant Biol 69: 437–468 [DOI] [PubMed] [Google Scholar]
  34. Wang Y, Li J(2005) The plant architecture of rice (Oryza sativa). Plant Mol Biol 59: 75–84 [DOI] [PubMed] [Google Scholar]
  35. Weng X, Wang L, Wang J, Hu Y, Du H, Xu C, Xing Y, Li X, Xiao J, Zhang Q(2014) Grain number, plant height, and heading date7 is a central regulator of growth, development, and stress response. Plant Physiol 164: 735–747 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Wu D, Huang LZ, Gao J, Wang YH(2016) The molecular mechanism of plant gravitropism. Yi Chuan 38: 589–602 [DOI] [PubMed] [Google Scholar]
  37. Wu X, Tang D, Li M, Wang K, Cheng Z(2013) Loose Plant Architecture1, an INDETERMINATE DOMAIN protein involved in shoot gravitropism, regulates plant architecture in rice. Plant Physiol 161: 317–329 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Wu Y, Zhao S, Li X, Zhang B, Jiang L, Tang Y, Zhao J, Ma X, Cai H, Sun C, et al. (2018) Deletions linked to PROG1 gene participate in plant architecture domestication in Asian and African rice. Nat Commun 9: 4157. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Xie K, Yang Y(2013) RNA-guided genome editing in plants using a CRISPR-Cas system. Mol Plant 6: 1975–1983 [DOI] [PubMed] [Google Scholar]
  40. Xie Y, Huhn K, Brandt R, Potschin M, Bieker S, Straub D, Doll J, Drechsler T, Zentgraf U, Wenkel S(2014) REVOLUTA and WRKY53 connect early and late leaf development in Arabidopsis. Development 141: 4772–4783 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Xu Y, McCouch SR, Shen Z(1998) Transgressive segregation of tiller angle in rice caused by complementary gene action. Crop Sci 38: 12–19 [Google Scholar]
  42. Yoshihara T, Iino M(2007) Identification of the gravitropism-related rice gene LAZY1 and elucidation of LAZY1-dependent and -independent gravity signaling pathways. Plant Cell Physiol 48: 678–688 [DOI] [PubMed] [Google Scholar]
  43. Yu B, Lin Z, Li H, Li X, Li J, Wang Y, Zhang X, Zhu Z, Zhai W, Wang X, et al. (2007) TAC1, a major quantitative trait locus controlling tiller angle in rice. Plant J 52: 891–898 [DOI] [PubMed] [Google Scholar]
  44. Zhang H, Zhao Y, Zhou DX(2017) Rice NAD+-dependent histone deacetylase OsSRT1 represses glycolysis and regulates the moonlighting function of GAPDH as a transcriptional activator of glycolytic genes. Nucleic Acids Res 45: 12241–12255 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Zhang N, Yu H, Yu H, Cai Y, Huang L, Xu C, Xiong G, Meng X, Wang J, Chen H, et al. (2018) A core regulatory pathway controlling rice tiller angle mediated by the LAZY1-dependent asymmetric distribution of auxin. Plant Cell 30: 1461–1475 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Zhang W, Tan L, Sun H, Zhao X, Liu F, Cai H, Fu Y, Sun X, Gu P, Zhu Z(2019) Natural variations at TIG1 encoding a TCP transcription factor contribute to plant architecture domestication in rice. Mol Plant 12: 1075–1089 [DOI] [PubMed] [Google Scholar]
  47. Zhao L, Tan L, Zhu Z, Xiao L, Xie D, Sun C(2015) PAY1 improves plant architecture and enhances grain yield in rice. Plant J 83: 528–536 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Zuckerkandl E, Pauling L(1965) Evolutionary divergence and convergence in proteins In Bryson V, and Vogel HJ, eds, Evolving Genes and Proteins. Academic Press, Cambridge, MA, pp 97–166 [Google Scholar]

Articles from Plant Physiology are provided here courtesy of Oxford University Press

RESOURCES