SAUR15 Promotes Lateral and Adventitious Root Development via Activating H⁺-ATPases and Auxin Biosynthesis

An early auxin response gene regulates auxin-mediated lateral and adventitious root development by stimulating plasma membrane H⁺-ATPases and promoting auxin accumulation.

Abstract

SMALL AUXIN-UP RNAs (SAURs) comprise the largest family of early auxin response genes. Some SAURs have been reported to play important roles in plant growth and development, but their functional relationships with auxin signaling remain unestablished. Here, we report Arabidopsis (Arabidopsis thaliana) SAUR15 acts downstream of the auxin response factors ARF6,8 and ARF7,19 to regulate auxin signaling-mediated lateral root (LR) and adventitious root (AR) formation. The loss-of-function mutant saur15-1 exhibits fewer LRs and ARs. By contrast, plants overexpressing SAUR15 exhibit more LRs and ARs. We find that the SAUR15 promoter contains four tandem auxin-responsive elements, which are directly bound by ARF6 and ARF7 and are essential for SAUR15 expression. LR and AR impairment in arf6 and arf7 mutants is partially reduced by ectopic expression of SAUR15. Additionally, we demonstrate that the ARF6,7-upregulated SAUR15 promotes LR and AR development using two mechanisms. On the one hand, SAUR15 interacts with PP2C-D subfamily type 2C protein phosphatases to inhibit their activities, thereby stimulating plasma membrane H⁺-ATPases, which drives cell expansion and facilitates LR and AR formation. On the other hand, SAUR15 promotes auxin accumulation, potentially by inducing the expression of auxin biosynthesis genes. A resulting increase in free auxin concentration likely triggers LR and AR formation, forming a feedback loop. Our study provides insights and a better understanding of how SAURs function at the molecular level in regulating auxin-mediated LR and AR development.

Lateral roots (LRs) play vital roles in plant anchorage, as well as in water and nutrient uptake (Casimiro et al., 2003; Hochholdinger and Zimmermann, 2008; Nibau et al., 2008; Den Herder et al., 2010; Petricka et al., 2012). LR formation has been extensively studied, especially in the model plant Arabidopsis (Arabidopsis thaliana). In Arabidopsis plants, LR formation comprises four landmark stages, primarily regulated by auxin, and defined as priming, initiation, primordium development, and emergence (Péret et al., 2009). A lateral root primordium (LRP) initiates from founder cells that are formed from xylem pole pericycle cells primed in the basal meristem of the primary root in response to auxin (Casimiro et al., 2001; Dubrovsky et al., 2008). According to Malamy and Benfey (1997), LRP development can be divided into seven main stages: stages I to VII. At stages I and II, founder cells undergo several rounds of anticlinal and asymmetric divisions, generating a dome-shaped LRP in the parent root. Subsequently, LRP enlarges via constant and highly organized cell divisions accompanied by cell expansion. The developmentally complete LRP then emerges, a process primarily driven by cell expansion. These cell division- and expansion-controlled LRP development and emergence events also depend on auxin, accumulated via local auxin biosynthesis and shoot-derived auxin transport. Hence, the key auxin biosynthesis enzymes, such as TRP AMINOTRANSFERASE RELATEDs (TARs) and YUCCAs (YUCs), and transporters, such as PIN-FORMEDs (PINs) and ATP-BINDING CASSETTE B19 (ABCB19), were found play important roles in regulating LR development (Geldner et al., 2001; Bhalerao et al., 2002; Friml et al., 2002; Lewis et al., 2011; Sukumar et al., 2013; Marhavý et al., 2014; Taylor-Teeples et al., 2016).

In addition to the complex regulation of auxin biosynthesis and transport, auxin signaling also plays a major role in LR organogenesis. Auxin binds to TRANSPORT INHIBITOR RESPONSE1/AUXIN SIGNALING F-BOXs receptors to trigger the degradation of AUXIN/INDOLE3-ACETIC ACID (Aux/IAAs) repressors, enabling the derepression of relevant AUXIN RESPONSE FACTORs (ARFs), which then activate the downstream auxin response genes to promote LR formation (Kepinski and Leyser, 2005; Weijers et al., 2005). Control of these processes by auxin signaling cascades are performed through different auxin signaling modules (Lavenus et al., 2013). For example, the IAA28-ARF5, -ARF6, -ARF7, -ARF8, and -ARF19 modules are important in the LR founder cell priming stage (De Rybel et al., 2010); the IAA14/SOLITARY ROOT (SLR)-ARF7 and -ARF19 and IAA12/BODENLOS-ARF5 modules control LR initiation (Okushima et al., 2007; Lee et al., 2009; De Smet, 2010; Berckmans et al., 2011); the IAA3/SHORT HYPOCOTYL2 (SHY2)-ARF7 and IAA14/SLR-ARF7 and -ARF19 modules are critical for LR emergence (Goh et al., 2012; Péret et al., 2012; Kumpf et al., 2013); and the IAA14/SLR-ARF7 and -ARF19 module regulates auxin response both in the LRP and outer tissues during LR development (Péret et al., 2012; Kumpf et al., 2013). Together, the functionally redundant transcription factors ARF7 and ARF19 are required for almost every stage of LR development. Indeed, LR formation is abolished in arf7 arf19 double mutants. In addition, downstream of the IAA14/SLR-ARF7 and -ARF19 signaling module, not only IAA3/SHY2 but also LATERAL ORGAN BOUNDARIES-DOMAINs (LBDs), such as LBD16, LBD18, and LBD29 are found as direct targets of ARF7 and ARF19 and identified as important components in regulating LR development (Okushima et al., 2007; Lee et al., 2009; Goh et al., 2012; Porco et al., 2016).

In Arabidopsis, ARs formed on the hypocotyl of etiolated seedlings share key elements of the genetic and hormonal regulatory networks for LRs formed on the primary root (Boerjan et al., 1995; Delarue et al., 1998; Sukumar et al., 2013). Both ARF7 and ARF19 play important roles in AR formation (Wilmoth et al., 2005). Further, IAA28, IAA3/SHY, and IAA14/SLR also act as determinants for AR initiation, with no or few ARs formed on the hypocotyl of the de-etiolated iaa28-1, slr-1/iaa14, or shy2-2/iaa3 mutant seedlings (Leyser et al., 1996; Tian and Reed, 1999; Rogg et al., 2001; Fukaki et al., 2005). Downstream of IAA14/SLR, the functionally redundant transcription factors ARF6 and ARF8 are identified as key positive regulators in the AR development pathway (Gutierrez et al., 2012; Lakehal et al., 2019). Additionally, auxin-inducible Gretchen Hagen3-like (GH3) proteins GH3-3, GH3-5, and GH3-6 fulfill a role in the fine-tuning of AR initiation via demodulation of jasmonate homeostasis (Gutierrez et al., 2012). Together, in the auxin signaling cascade, ARFs and these two major classes of early auxin response genes, Aux/IAAs and GH3s, play important roles in LR and AR formation and development.

Another class of early auxin response genes is SAURs, which is the largest family of auxin-induced genes, with 79 members in the Arabidopsis genome (Hagen and Guilfoyle, 2002). These SAURs have been implicated in an increasing number of biological processes, including hypocotyl elongation, cotyledon opening, apical hook formation, and root growth (Spartz et al., 2012; Kong et al., 2013; Stamm and Kumar, 2013; Sun et al., 2016; Favero et al., 2017; Dong et al., 2019). The development of all these structures involves cell expansion, which is well explained by SAUR-PP2C-D-PM H⁺-ATPase signaling module-mediated cell expansion (Spartz et al., 2014; Sun et al., 2016; Favero et al., 2017; He et al., 2018). In brief, upon upstream signals provided by auxin, brassinosteroids (BRs), or light, expression of specific SAURs is induced, generating accumulated SAUR proteins that interact with and inhibit members of the PP2C-D clade of type 2C phosphatases at the plasma membrane (PM) to activate PM H⁺-ATPases by the phosphorylation of penultimate residue Thr, which then promotes cell expansion important for organ growth and development (Takahashi et al., 2012; Spartz et al., 2014; Ren et al., 2018; Uchida et al., 2018; Minami et al., 2019; Wong et al., 2019). LR and AR development are also dependent on cell expansion. However, no molecular evidence has yet been identified to explain whether some SAUR family members share the same action mechanism to regulate LR and AR growth and development.

In this study, we show that SAUR15, acting downstream of ARF6,8 and ARF7,19 in an auxin signaling cascade, modulates LR and AR formation via activating PM H⁺-ATPases and regulating auxin biosynthesis.

RESULTS

SAUR15 Positively Regulates Auxin-Mediated LR and AR Formation

The SAUR15 gene has been widely used as a marker for auxin-inducible gene expression (Gil et al., 1994; Roig-Villanova et al., 2007). To characterize the function of SAUR15 in plant growth and development, Arabidopsis loss- and gain-of-function mutant lines were used to assess changes in phenotype and physiology compared to wild-type plants. Stabilized SAUR15-overexpression (OE; Ren and Gray, 2015) lines (pSAUR15::SAUR15-FLAG and 35S::SAUR15-FLAG) were generated by overexpressing SAUR15 under a native promoter or a Cauliflower mosaic virus 35S promoter in Arabidopsis wild-type Col-0 plants. Phylogenetic analysis of SAUR family genes of Arabidopsis showed that SAUR15 is a single branch in the evolutionary tree (Supplemental Fig. S1). We speculated that SAUR15 is less likely to be functionally redundant with other SAURs. Thus, a SAUR15 transfer DNA (T-DNA) insertion line, named saur15-1 (GK_931C06) was used for loss-of-function studies (Supplemental Fig. S2A). Genotyping assays were used to confirm the presence of a T-DNA insert in saur15-1 that disrupts the coding region (Supplemental Fig. S2B). In addition, reverse transcription-quantitative PCR (RT-qPCR) was used to confirm that SAUR15 transcripts were obviously increased in SAUR15-OE lines and barely detectable in saur15-1 mutant seedlings compared with Col-0 (Supplemental Fig. S2C). When germinated and grown on one-half strength Murashige and Skoog (MS) solid medium, SAUR15-OE dark-grown, etiolated seedlings showed various auxin-related phenotypic differences from Col-0, including increased apical hook angles, open cotyledons (Supplemental Fig. S2, D and E), and longer hypocotyls (Supplemental Fig. S2, F and G) with ARs (Supplemental Fig. S2H). Light-grown SAUR15-OE seedlings had longer hypocotyls and bigger leaves compared to Col-0 (Supplemental Fig. S2, I and L). Meanwhile, dark-grown, etiolated saur15-1 seedlings showed reduced apical hook angles and hypocotyl lengths (Supplemental Fig. S2, D–G), and light-grown seedlings showed shorter hypocotyls and smaller leaves compared to Col-0 (Supplemental Fig. S2, I and L). Detailed analyses showed that the number of LRP and emerged LRs was higher in SAUR15-OE seedlings and lower in saur15-1 mutants compared to Col-0 (Fig. 1, A and B). Similarly, when 4-d-old dark-grown seedlings were transferred into light for 7 d, the number of light-induced AR primordia (ARP) and emerged ARs was higher in SAUR15-OE plants and lower in saur15-1 compared to Col-0 (Fig. 1, C and D). All of the above phenotypic defects in saur15-1 mutants were complemented by expression of SAUR15-FLAG driven by a native promoter pSAUR15 (Supplemental Fig. S2, J–L). These collective data indicate that, in addition to regulating apical hook angle, hypocotyl length, and leaf size, which have been previously documented for other SAURs (Spartz et al., 2012; Kong et al., 2013; Stamm and Kumar, 2013; Sun et al., 2016; Favero et al., 2017; Dong et al., 2019), SAUR15 promotes LR and AR formation.

Figure 1.

SAUR15 participates in LR and AR formation. A, LR phenotypes of wild-type (Col-0), T-DNA insertion mutant (saur15-1), and overexpression lines (pSAUR15::SAUR15-FLAG and 35S::SAUR15-FLAG). Plants were grown vertically for 7 d after germination on one-half strength MS medium. Bar = 1 cm. B, Statistical analysis of LRP and emerged LRs of plants treated as described in the legend of A. C, AR phenotypes of Col-0, saur15-1, and overexpression lines (pSAUR15::SAUR15-FLAG and 35S::SAUR15-FLAG). Four-day-dark-grown plants were transferred to the light and grown vertically for another 7 d on one-half strength MS medium. Bar = 1 cm. D, Statistical analysis of ARP and emerged ARs of plants treated as described in the legend of C. For A to D, experiments were repeated three times with similar results. E, GUS expression in LRP or emerged LRs of pSAUR15::GUS transgenic plants. Seedlings were grown vertically for 7 d after germination on one-fifth strength MS medium. All samples were treated with 5-bromo-chloro-3-indoxyl-beta-d-glucuronide cyclohexylammonium salt for 10 h. Bars = 100 μm. F, GUS expression in ARP or emerged ARs of pSAUR15::GUS transgenic plants. Four-day-dark-grown seedlings were transferred to the light and grown vertically for another 7 d on one-fifth strength MS medium. Bars = 100 μm. For B and D, asterisks indicate significant difference from Col-0 control (n > 10 plants per column; independent-samples t test; *P < 0.05, **P < 0.01).

LRP development can be divided into seven stages (Malamy and Benfey, 1997). ARP development is thought to share these steps (Bellini et al., 2014; Verstraeten et al., 2014). To further assess the function of SAUR15, we generated pSAUR15::GUS transgenic plants harboring a 1.6-kb SAUR15 promoter::GUS fusion construct. Staining of light-grown transgenic seedlings revealed GUS signals in leaves and hypocotyls (Supplemental Fig. S2M), but also in developing LRP at all stages and mature LRs (Fig. 1E). Similar strong GUS expression was observed at all stages of AR development (Fig. 1F). Treatment with the auxin IAA strongly enhanced these GUS signals at all developmental stages (Supplemental Fig. S2N). These GUS expression patterns are consistent with a role for SAUR15 in AR and LR formation during auxin signaling.

SAUR15 Acts Downstream of ARF6 and ARF7 in Response to Auxin

To further test if SAUR15 mediates auxin signaling important for LR and AR formation, 4-d-old light- or dark-grown Col-0, saur15-1, and SAUR15-OE seedlings were transferred to growth media with or without IAA for 4 or 6 d. The LR- and AR-defective phenotypes of saur15-1 were rescued to wild-type levels on media containing 100 nm of IAA (Fig. 2, A–F). However, SAUR15-OE seedlings developed more LRs and ARs than Col-0 on media with or without IAA (Fig. 2, A–F; Supplemental Fig. S3, A–F). These data demonstrate that SAUR15-regulated LR and AR formation is auxin dependent.

Figure 2.

SAUR15-mediated LR and AR development is regulated by ARF6,8 and ARF7,19 in response to auxin. A, LR phenotypes of Col-0, saur15-1, pSAUR15::SAUR15-FLAG, and 35S::SAUR15-FLAG with or without IAA (mock) treatment. Four-day-old plants grown on one-half strength MS medium were transferred to one-half strength MS medium with or without 100 nm IAA and grown vertically for 4 d. Bars = 1 cm. B and C, Statistical analysis of LRP (B) and emerged LRs (C). Plants were treated as described in the legend of A. D, AR phenotypes of Col-0, saur15-1, pSAUR15::SAUR15-FLAG, and 35S::SAUR15-FLAG with or without IAA (mock) treatment. Four-day-dark-grown plants were transferred to one half-strength MS medium with or without 100 nm IAA and grown vertically for 6 d under light. Bars = 1 cm. E and F, Statistical analysis of ARP (E) and emerged ARs (F). Plants were treated as described in the legend of D. For B, C, E, and F, asterisks indicate significant difference from Col-0 control (n > 10 plants per column. Independent-samples t test; *P < 0.05, **P < 0.01). G and H, Effects of IAA on the expression of SAUR15 in the roots (G) and hypocotyls (H) of Col-0, arf6, arf6 arf8, 35S::ARF6-3HA, arf7, arf7 arf19, and 35S::ARF7-3HA plants. For G, 7-d-old light-grown seedlings were transferred to the medium with or without 1 μm IAA (mock), and roots were harvested for expression analysis after being treated for 1 h. For H, 4-d-dark-grown plants were transferred to the light for another 7 d and then treated with or without 1 μm IAA (mock). Hypocotyls were harvested for expression analysis after being treated for 1 h. The transcripts were analyzed by RT-qPCR, and the expression levels of SAUR15 were normalized to those of ACTIN2 and compared with those in Col-0 under control condition (mock) in G and H. n = 3 per column. Independent-samples t test. All experiments were repeated three times with similar results.

SAUR15 is a well-known auxin response gene under the regulation of ARFs. A previous study showed that ARF6 and ARF8 directly regulate SAUR15 during hypocotyl elongation (Oh et al., 2012). Given that ARF6,8 and ARF7,19 positively regulate LR and AR formation (Okushima et al., 2007; De Rybel et al., 2010; Gutierrez et al., 2012; Supplemental Fig. S3, L and M, and S4, B–D), we investigated a role for these ARFs as upstream regulators of SAUR15 in auxin-induced LR and AR formation. To test this role, the expression of SAUR15 was monitored in ARF mutant and overexpressing lines: arf6, arf6 arf8, 35S::ARF6-3HA, and arf7, arf7 arf19, 35S::ARF7-3HA. The arf6, arf6 arf8, arf7, and arf7 arf19 mutants were T-DNA insertion lines obtained from the Arabidopsis Biological Resource Center. The ARF6 and ARF7 overexpression lines were generated by overexpressing 35S::ARF6-3HA and 35S::ARF7-3HA constructs in which ARF6 and ARF7 tagged with hemagglutinin (HA) were expressed under the control of a Cauliflower mosaic virus 35S promoter in Arabidopsis Col-0 wild-type plants. RT-qPCR results indicated that gene transcript levels were significantly lower in each corresponding mutant compared with Col-0 (Supplemental Fig. S3, G–J). Western blotting results using an anti-HA antibody showed that ARF6 and ARF7 were overexpressed in 35S::ARF6-3HA and 35S::ARF7-3HA plants (Supplemental Fig. S3K). Col-0, arf6, arf6 arf8, 35S::ARF6-3HA, arf7, arf7 arf19, and 35S::ARF7-3HA seedlings were treated with or without IAA for 1 h. RT-qPCR results showed that under normal conditions, the expression of SAUR15 in arf6, arf6 arf8, arf7, and arf7 arf19 mutants was reduced, and in 35S::ARF6-3HA and 35S::ARF7-3HA plants was elevated, compared to Col-0 (Fig. 2, G and H). Under IAA treatment, the expression of SAUR15 was induced in Col-0, ARF6, and ARF7 overexpression seedlings and further upregulated in 35S::ARF6-3HA and 35S::ARF7-3HA seedlings, compared to Col-0 (Fig. 2, G and H). This inducement by IAA was lower in arf6 and arf7 single mutants, and there was no substantial change in expression of SAUR15 in the roots of arf6 arf8 or the hypocotyls of arf7 arf19 treated with or without IAA (Fig. 2, G and H). To further substantiate the above results, we additionally analyzed the expression of pSAUR15::GUS in arf6 and arf7 mutant backgrounds. These lines were obtained by crossing pSAUR15::GUS plants with either arf6 or arf7 mutants. Staining of these pSAUR15::GUS seedlings showed that GUS signals in LRP and ARP were diminished in arf6 and arf7 mutants (Supplemental Fig. S3N). Upon auxin treatment, GUS signals were enhanced in Col-0, but to a lesser degree in arf6 and arf7 mutants (Supplemental Fig. S3N). These collective data indicate that SAUR15 acts downstream of ARF6 and ARF7 in auxin-mediated LR and AR formation.

SAUR15 Is the Direct Target of ARF6 and ARF7, and Overexpression of SAUR15 Partially Suppresses the LR and AR Phenotypes of arf6 and arf7 Mutants

To determine if SAUR15 is a direct target of ARF6 and ARF7, we first searched for potential ARF binding sites in the SAUR15 promoter region (1.6 kb to 1 bp upstream of the translation start codon ATG). Four classical auxin-responsive elements are clustered in this region (Fig. 3A). Direct association of ARF6 and ARF7 to genomic DNA containing these sites was assayed by chromatin immunoprecipitation (ChIP). Seedling material from 35S::ARF6-3HA and 35S::ARF7-3HA transgenic plants was probed using an anti-HA antibody followed by qPCR with primers flanking the putative ARF6 and ARF7 binding site in the SAUR15 promoter (Fig. 3A). The ChIP-qPCR analyses showed that ARF6-3HA and ARF7-3HA binding is enriched at auxin-responsive elements in the SAUR15 promoter (Fig. 3B). Further, IAA significantly increased interaction of ARF6-3HA and ARF7-3HA with the SAUR15 promoter (Fig. 3C). These results demonstrate that ARF6 and ARF7 bind to the promoter of SAUR15.

Figure 3.

ARF6 and ARF7 directly bind to the promoter of SAUR15 to regulate its expression. A, Schematic diagram showing the ARF binding sites (vertical lines) in the promoter of SAUR15. B, ChIP-qPCR assays showed that ARF6 and ARF7 can directly bind to the ARF binding sites in the promoter of SAUR15. C, ChIP-qPCR assay showed that auxin treatment (1 µm IAA for 2 h) increases the capacity of ARF6 and ARF7 to bind to the promoter of SAUR15. For B and C, primers used in qPCR are listed in Supplemental Table S1. Data shown are average of three independent biological replicates. n = 3, Asterisks indicate significant difference based on independent-samples t test (*P < 0.05). D and G, Overexpression of SAUR15 can partially suppress the LR (top) and AR (bottom) defect phenotypes of arf6 (D) and arf7 (G) mutant seedlings. Bars = 1 cm. E and F, Statistical analysis of the emerged LRs (E) and ARs (F) of Col-0, arf6, and 35S::SAUR15-FLAG in arf6 seedlings. H and I, Statistical analysis of the emerged LRs (H) and ARs (I) of Col-0, arf7, and 35S::SAUR15-FLAG in arf7 seedlings. For D to I, seedlings were treated as described in the legend of Figure 1, A and C. Asterisks indicate significant difference from arf6 (E and F) and arf7 (H and I) by independent-samples t test (*P < 0.05 and **P < 0.01; n > 10 plants per column). Experiments were repeated three times with similar results.

To ascertain that the ARF binding site in the SAUR15 promoter mediates ARF6 and ARF7 as a transactivation function, we used a transient expression assay in Nicotiana benthamiana leaves. Wild-type and mutant SAUR15 promoters, in which the auxin-responsive elements were deleted, were fused to luciferase (LUC) to generate pSAUR15-WT::LUC and pSAUR15-d::LUC, respectively. Coinfiltration of 35S::ARF6-3HA with pSAUR15-WT::LUC and 35S::ARF7-3HA with pSAUR15-WT::LUC led to obvious induction of luminescence intensity (Supplemental Fig. S4A). However, the activation effect of ARF6 and ARF7 on pSAUR15-d::LUC was completely abolished (Supplemental Fig. S4A). These data suggest that the ARF6 and ARF7 binding motifs are critical for induction of SAUR15 gene expression.

Since ARF6 and ARF7 acting on SAUR15 positively regulate auxin-mediated LR and AR development, we further tested whether ectopic expression of SAUR15 rescues the LR and AR defects of arf6 and arf7 mutants. We generated transgenic plants overexpressing SAUR15 under the control of 35S promoter in arf6, arf7, and arf7 arf19 mutant backgrounds (35S::SAUR15-FLAG in arf6, 35S::SAUR15-FLAG in arf7, and 35S::SAUR15-FLAG in arf7 arf19). The LR and AR phenotypic defects of arf6 were almost completely rescued (Fig. 3, D–F) and partially rescued in arf7 and arf7 arf19 (Fig. 3, G–I; Supplemental Fig. S4, B–D). These results demonstrated that SAUR15 acts genetically downstream of ARF6 and ARF7 in auxin-mediated LR and AR development.

PM-Localized SAUR15 Inhibits the Activities of PP2C-D Phosphatases

It was previously reported the PM-localized SAUR19 protein can interact with and inhibit PP2C-D phosphatases to promote cell expansion (Spartz et al., 2014). The light-responsive SAURs are also reported to inhibit PP2C-D activities (Sun et al., 2016). These observations suggest that SAUR15 interacts with PP2C-Ds on the PM to exert its functions. A series of experiments was performed to test this hypothesis. First, a transient expression assay showed that the SAUR15-GFP fusion proteins are localized on the PM of epidermis cells in N. benthamiana leaves (Fig. 4A). We then tested for physical interactions between SAUR15 and PP2C-Ds to indicate complex formation. Yeast two-hybrid analyses using a mating-based split ubiquitin system (mbSUS) showed that SAUR15 protein interacts strongly with PP2C-D1, PP2C-D2, PP2C-D5, and PP2C-D6 and weakly with PP2C-D8 (Fig. 4B). Bimolecular fluorescence complementation experiments further demonstrated that SAUR15 associates with PP2C-D1, PP2C-D2, PP2C-D5, and PP2C-D6 on the PM (Fig. 4C). The collective data demonstrated that SAUR15 interacts with PP2C-D1, PP2C-D2, PP2C-D5, and PP2C-D6 both in vitro and in vivo, suggesting that SAUR15 may possess the same function as SAUR19 to inhibit the activities of PP2C-D phosphatases.

Figure 4.

SAUR15 proteins interact with PP2C-D phosphatases to inhibit their enzyme activities. A, Subcellular location of SAUR15 proteins. Bars = 10 μm. B, Interaction assay between SAUR15 and five D-clade PP2Cs via yeast two-hybrid mbSUS system. SAUR15 and AHA2 proteins were fused to the C-terminal part of ubiquitin, respectively; PP2C-D1, PP2C-D2, PP2C-D5, PP2C-D6, and PP2C-D8 were fused to the N-terminal part of ubiquitin, respectively. Yeast was grown on a synthetic complete medium containing Ade and His (SC + Ade + His) for diploid cells selection or on a synthetic dextrose minimal medium (SD) with 150 or 400 μm Met for interaction detection. C, Bimolecular fluorescence complementation assay for SAUR15 and four D-clade PP2Cs. SAUR15 proteins were fused to the N-terminal part of YFP; PP2C-D1, PP2C-D2, PP2C-D5, and PP2C-D6 were fused to the C-terminal part of YFP. The indicated split YFP constructs were transiently coexpressed in N. benthamiana leaves, and fluorescent images of epidermal cells were obtained by confocal microscopy. Bars = 10 μm. D and E, Phosphatase activity assays of PP2C-D2 (D) and PP2C-D5 (E). For D, p-nitrophenylphosphate (pNPP) phosphatase assays containing 0.2 μm MBP-PP2C-D2 ± 1 μm MBP or MBP-SAUR15. For E, pNPP phosphatase assays containing 0.2 μm MBP-PP2C-D5 ± 1 μm MBP or MBP-SAUR15. Absorbance at 405 nm was recorded every minute up to 8 min.

We next tested whether SAUR15 interaction regulates PP2C-D activities. First, maltose binding protein (MBP) control or MBP-tagged PP2C-D2, PP2C-D5, and SAUR15 were expressed in Escherichia coli cells. Addition of a chromogenic pNPP substrate was used to test the activities of PP2C-D2 and PP2C-D5 in vitro. The recombinant PP2C-D2 and PP2C-D5 exhibited strong phosphatase activities as reported previously (Spartz et al., 2014). This activity was repressed by addition of MBP-SAUR15, but not MBP alone (Fig. 4, D and E), demonstrating SAUR15 inhibits the activity of PP2C-D phosphatases.

PP2C-D Phosphatases Negatively Regulate LR and AR Formation

SAUR15 binding-mediated repression of PP2C-D activity suggests that PP2C-Ds and SAUR15 function in the same genetic pathway to control LR and AR development. To test this idea, LRs and ARs were measured in seedlings of PP2C-D T-DNA insertion mutants (pp2c-d2, pp2c-d5 single and pp2c-d2 pp2c-d5 double mutants) and PP2C-D-overexpressing lines (35S::PP2C-D2-3HA line 1 and line 2) and compared to Col-0 and saur15-1 controls (Fig. 5, A–F; Supplemental Fig. S5A). Similar to SAUR15-OE, the number of LRs and ARs was increased in the pp2c-d2, pp2c-d5, and pp2c-d2 pp2c-d5 seedlings compared to wild type (Fig. 5, A–C). The hook angles of SAUR15-OE and pp2c-d2 pp2c-d5 seedlings were also increased (Supplemental Fig. S5, B and C). By contrast, the number of LRs and ARs was decreased in PP2C-D2-overexpressing lines compared to wild type (Fig. 5, D–F). Further genetic testing showed that PP2C-D2 loss-of-function partially rescues the LR and AR defect of saur15-1 mutants (Fig. 5, G–I), whereas PP2C-D2 overexpression partially corrects the LR and AR phenotypes of SAUR15-OE plants (Fig. 5, J–L). These collective data show that PP2C-D2 and PP2C-D5 negatively regulate LR and AR formation and genetically interact with SAUR15 to regulate LR and AR formation. These results align with Spartz et al. (2014), who previously reported that SAUR19 positively regulates cell elongation via inhibition of PP2C-D activities. Further, SAUR15-OE plants displayed all of the same phenotypes as SAUR19-OE plants exhibited, which are similar to the constitutively active PM H⁺-ATPase mutant ost2 (OPEN STOMATA2; Merlot et al., 2007), suggesting that SAUR15 proteins might play a role in PM H⁺-ATPase activation similar to SAUR19 via interaction with and inhibition of PP2C-Ds in regulating auxin-mediated LR and AR development.

Figure 5.

PP2C-D2 and PP2C-D5 negatively regulate LR and AR formation. A, The pp2c-d2, pp2c-d5 single and pp2c-d2 pp2c-d5 double mutants exhibit increased LR and AR phenotypes. B and C, LR (B) and AR (C) statistical analysis of seedlings treated as described in the legend of A. D, LR and AR phenotypes of Col-0 and saur15-1, as well as two independent PP2C-D2 overexpression lines. E and F, LR and AR statistical analysis of seedlings treated as described in the legend of D. G, LR and AR phenotypes of Col-0, saur15-1, pp2c-d2, and pp2c-d2 saur15-1 double mutants. H and I, Statistical analysis of the LRs and ARs of seedlings treated as described in the legend of G. J, PP2C-D2 overexpression can suppress the LR and AR phenotypes of SAUR15 overexpression plants. K and L, LR and AR statistical analysis of seedlings treated as described in the legend of J. Asterisks indicate significant difference from Col-0 control by independent-samples t test (*P < 0.05 and **P < 0.01; n > 10 plants per column). Bars = 1 cm. All experiments were repeated three times with similar results.

SAUR15 Positively Regulates PM H⁺-ATPase Activity to Promote LR and AR Formation

Physiological, genetic, and biochemical experiments were performed to examine if SAUR15 regulates the activation of PM H⁺-ATPase. First, like SAUR19-OE plants and ost2 mutants and consistent with increased proton pump activity, overexpression of SAUR15 resulted in increased media acidification, which facilitates LR and AR formation and outgrowth, whereas saur15-1 resulted in decreased media acidification (Fig. 6, A and B). We then assayed the apoplastic pH of rosette leaves. Consistent with the media acidification results, apoplastic pH was significantly increased in saur15-1 mutants, whereas apoplastic pH was diminished in SAUR15-OE lines (Fig. 6C), also suggesting that SAUR15 activates PM H⁺-ATPases.

Figure 6.

SAUR15 positively regulates PM H⁺-ATPase activity, facilitating LR and AR formation. A and B, Medium acidification assays around LRs and ARs. Twelve-day-old light-grown seedlings (A) and 4-d-dark-grown + 8-d-light-grown seedlings (B) were transferred to plates containing the pH indicator dye bromocresol purple. Color changes were recorded after 24 h. Bars = 1 cm. C, Leaf apoplastic pH. Emission wavelength of8-hydroxypyrene-1,3,6-trisulfonic acid trisodium was collected at both 510 and 530 nm. At least 50 leaves were used for each replicate. D, Relative vanadate-sensitive ATP hydrolytic activity and PM H⁺-ATPase level in PM fractions prepared from 10-d-old seedlings. For C and D, apoplastic pH and ATP hydrolytic activity shown are average of three independent biological replicates. E, Phosphorylation level of H⁺-ATPase in PM fractions prepared from 3-week-old seedlings. Bottom shows the equal loading of the PM H⁺-ATPase for each sample. In the top, double amount of each PM H⁺-ATPase sample was loaded to detect the phosphorylation level. F, LR and AR phenotypes of Col-0 and aha1 mutant. G and H, LR and AR statistical analysis of seedlings treated as described in the legend of F. I, Overexpression of AHA1 can partially suppress the LR and AR defects of saur15-1. J and K, Statistical analysis of LRs and ARs of seedlings treated as described in the legend of I. For C, D, G, H, J, and K, asterisks indicate significant difference from Col-0 control based on independent-samples t test ( *P < 0.05 and **P < 0.01; n > 10 plants per column). Bars = 1 cm (F and I).

To directly assess the effect of SAUR15 proteins on PM H⁺-ATPase activity, we measured the vanadate-sensitive ATP hydrolytic activity present in PM fractions prepared from Col-0, saur15-1, and SAUR15-OE seedlings. Compared with the equal PM of Col-0, membranes prepared from SAUR15-OE lines displayed a 23% to 38% increase in ATPase activity, while that in saur15-1 PM was decreased by about 42% (Fig. 6D). However, the H⁺-ATPase protein level in saur15-1 is much more than that of Col-0, whereas the protein levels in two independent SAUR15-OE lines are less compared to Col-0 (Fig. 6D). These data demonstrate that PM H⁺-ATPases are constitutively active in SAUR15-overexpression seedlings and the accumulation of PM H⁺-ATPases is regulated by SAUR15. To confirm whether the PM H⁺-ATPases are activated by phosphorylation of the penultimate residue Thr, the phosphorylation level of H⁺-ATPases in PM fractions prepared from Col-0, saur15-1, and SAUR15-OE seedlings was estimated by using an anti-pThr947-AHA2 antibody (Abmart). Our results showed that when the PM H⁺-ATPase protein levels are equal, its phosphorylation level in SAUR15-OE is stronger than that of Col-0 (Fig. 6E). The phosphorylation level of H⁺-ATPase in saur15-1, on the other hand, is much lower compared to that in Col-0 (Fig. 6E). These results are consistent with the findings from previous reports, indicating that SAUR proteins activate PM H⁺-ATPases by phosphorylating the penultimate residue Thr (Takahashi et al., 2012; Uchida et al., 2018; Minami et al., 2019).

Genetic experiments further showed that, like saur15-1 seedlings, PM H⁺-ATPase isoform mutant aha1 seedlings have a previously undescribed decrease in LRP/LR and ARP/AR numbers compared to wild-type seedlings (Fig. 6, F–H; Supplemental Fig. S6, A and B). AHA1 overexpression partially suppressed the LR and AR defects of saur15-1 (Fig. 6, I–K), whereas overexpression of both AHA1 and AHA2 corrected saur15-1 mutant defects in full (Supplemental Fig. S6H). These results demonstrated that PM H⁺-ATPases act genetically downstream of SAUR15 to regulate LR and AR formation.

Lastly, we showed that SAUR15-OE lines are sensitive to toxic cations Li⁺ (LiCl; Supplemental Fig. S6, C and D; Haruta et al., 2010) and insensitive to PM H⁺-ATPase stimulant fusicoccin (FC; Supplemental Fig. S6, E–G; Singh and Roberts, 2004), consistent with ost2 and SAUR19-OE previous reports (Spartz et al., 2014). Inversely, saur15-1 seedlings were sensitive to LiCl and were hypersensitive to FC (Supplemental Fig. S6, C–G). Unlike wild-type and saur15-1 seedlings, the SAUR15 overexpression lines exhibited no additional LR formation in response to FC (Supplemental Fig. S6, C–G). These characterizations were not described in previous studies on SAUR19 overexpression lines and ost2 plants.

Taken together, these results suggest SAUR15 positively regulates PM H⁺-ATPase activity to promote LR and AR formation.

SAUR15 Positively Regulates Auxin Synthesis

Since overexpression of SAUR15-FLAG fusion proteins stimulates AR formation on the hypocotyl of dark-grown etiolated seedlings (Supplemental Fig. S2H) reminiscent of auxin overproduction mutants superroot1 (sur1) and sur2 (Boerjan et al., 1995; Delarue et al., 1998), we speculated that SAUR15 might regulate auxin synthesis. We first tested for auxin accumulation using DR5::GUS transgenic plants in which the GUS staining intensity is positively related to the native auxin distribution (Ulmasov et al., 1997). We constructed DR5::GUS transgenic plants in saur15-1 and SAUR15-OE backgrounds (DR5::GUS in saur15-1, DR5::GUS in pSAUR15::SAUR15-FLAG, and DR5::GUS in 35S::SAUR15-FLAG) by crossing DR5::GUS in Col-0 plants with the corresponding mutant and overexpression lines. GUS staining showed that expression of DR5::GUS is substantially elevated throughout SAUR15-OE seedlings, whereas expression is decreased in saur15-1 mutants compared to Col-0 seedlings (Fig. 7, A–C). The changes in auxin levels were confirmed by measuring free IAA contents in these plants (Fig. 7D). In addition, we measured the transcript levels of auxin biosynthesis genes in Col-0, saur15-1, and SAUR15-OE seedlings. RT-qPCR analyses showed that TAR4, YUC6, YUC7, ALDEHYDE OXIDASE1 (AAO1), CYTOCHROME P450 CYP71A13, and NITRILASE 1 (NIT1) were greatly induced in SAUR15-OE lines (Fig. 7E). The expression levels of these genes, however, were significantly reduced in saur15-1 mutants (Fig. 7E). These data suggested that SAUR15 positively regulates auxin synthesis.

Figure 7.

SAUR15 positively regulates auxin synthesis. A to C, GUS staining of the DR5::GUS transgenic plants: Col-0, saur15-1, and SAUR15 overexpression lines. Seven-day-light-grown (A) and 4-d-dark-grown (B) whole plants were performed for detecting. GUS staining intensity of detail organs of seedlings treated as described in legends of A, B, and Figure 1E are shown in C. Bars = 1 cm (A and B) and 100 μm (C). D, Free IAA levels in Col-0, saur15-1, and pSAUR15::SAUR15-FLAG seedlings. Seven-day-light-grown seedlings were used for analysis. E and F, RT-qPCR results showing transcript levels of auxin biosynthesis genes (E) and cell expansion related genes (F) in 14-d-old light-grown Col-0, saur15-1, and pSAUR15::SAUR15-FLAG. The expression levels of each gene are normalized to ACTIN2 and compared with that in Col-0 control. For D to F, the data shown are representative of three independent experiments. Asterisks indicate significant difference from Col-0 control based on independent-samples t test (* P < 0.05 and **P < 0.01; n = 3). G, A model for SAUR15 regulation LR and AR formation. Black words and arrows represent the regulatory components and events analyzed in present study, and gray words and arrows represent the regulatory factors and events reported by previously studies.

Given that accumulation of auxin can trigger auxin signaling, we also tested if SAUR15-promoted auxin production can active auxin signaling by analyzing expression levels of EXPANSIN14 (EXP14) and EXP17 genes (Lee et al., 2009, 2013), which encode cell-wall-loosening factors associated with auxin response and LR development. RT-qPCR showed that the transcript levels of EXP17 are significantly elevated in SAUR15-OE lines and decreased in saur15-1 mutants compared to Col-0 (Fig. 7F). These data demonstrate that SAUR15-mediated auxin synthesis might activate auxin signaling to regulate LR and AR in a positive feedback loop.

DISCUSSION

The SAUR15 gene has been used widely as a marker for auxin-inducible gene expression. However, its functional relationship with auxin signaling has been unclear. In this study, we demonstrate that SAUR15 acts downstream of ARF6,8 and ARF7,19 to positively regulate auxin-induced LR and AR formation. Based on combination of genetic, molecular, and biochemical data, we propose a working model of how SAUR15 promotes auxin-induced LR and AR formation (Fig. 7G). In response to auxin, ARF6, ARF7, ARF8, and ARF19 direct expression of SAUR15, which in turn acts in two ways to promote LR and AR formation. On the one hand, SAUR15 proteins inhibit PP2C-D subfamily phosphatases to stimulate PM H⁺-ATPase activity, which drives cell expansion required for LR and AR formation. On the other hand, SAUR15 promotes the synthesis of auxin. The resulting increase in free auxin level induces signal transduction, and this positive feedback loop may trigger LR and AR formation by promoting the accumulation of SAUR proteins and/or cell wall-loosening factors (Fig. 7G).

SAUR15 is an early auxin response gene under the transcriptional control of ARFs in the auxin signaling pathway. However, the identity of these ARFs was partly unknown. We demonstrate that SAUR15 is the direct target gene of ARF6 and ARF7 and functions in the same genetic pathway as ARF6,8 and ARF7,19 to promote LR and AR formation. In support, mutation of SAUR15 and overexpression of stabilized SAUR15 proteins phenocopies arf6, arf6 arf8, arf7, arf7 arf19 mutants, and ARF6-, ARF7-overexpressed plants, respectively (Fig. 1, A–D; Supplemental Fig. S3, L and M). SAUR15, ARF6,8, and ARF7,19 showed overlapped expression patterns in AR and LR (Fig. 1, E and F; Okushima et al., 2005; Gutierrez et al., 2009). Further, ARF6 and ARF7 preferentially bind to an auxin-responsive elements-containing region of the SAUR15 promoter to induce expression in response to auxin (Fig. 3, B and C). These data agree with prior studies suggesting a role for ARF6,8 and ARF7,19 in promoting LR and AR formation (Okushima et al., 2007; Gutierrez et al., 2012) and showing direct regulation of SAUR15 by ARF6 and ARF8 during hypocotyl elongation (Oh et al., 2012).

We further show that SAUR15 overexpression can partly rescue the LR and AR defect of arf7 and arf7 arf19 mutants (Fig. 3, D–I; Supplemental Fig. S4, B–D), indicating that SAUR15 acts downstream of ARF7 and ARF19 with other potential factors involved in LR and AR development. This finding is consistent with previous reports demonstrating that ARF7 and ARF19 act together with LBD members and downstream cell wall loosening EXP14 and EXP17 to positively regulate LR formation (Lee et al., 2009, 2013). LBDs are also involved in AR development (Lee et al., 2019). On the other hand, loss of SAUR15 does not completely block formation of LRs and ARs (Fig. 1, A–D), suggesting the involvement of other SAURs. A likely candidate is SAUR19 whose expression is induced by auxin and protein product localizes to the PM in promoting hypocotyl cell expansion (Spartz et al., 2012, 2014). Identifying these SAURs through construction of higher-order mutants is a challenge for the future.

Auxin is a primary signaling molecule regulating LR formation at all stages of development (Dubrovsky et al., 2008, 2011; De Smet, 2012; Lavenus et al., 2013). Boerjan et al. (1995) reported that sur1 and sur2 auxin overproducing mutants develop excessive LRs, and under dark-grown conditions, etiolated seedlings form ARs on hypocotyls. Interestingly, we found that overexpression of stabilized SAUR15 proteins result in sur1- and sur2-like formation of ARs on the hypocotyl of dark-grown seedlings (Supplemental Fig. S2H). Collective changes in the staining intensity of DR5::GUS in saur15-1 and SAUR15 overexpression lines, and a corresponding decrease or increase of auxin levels compared to Col-0 (Fig. 7, A–D), indicate that SAUR15 promotes the synthesis of auxin in localized cells. SAUR15 is expressed throughout LR and AR development (Fig. 1, E and F), making it possible that SAUR15 continuously promotes local auxin synthesis to facilitate this process. In accordance, upstream transcription factors ARF7 and ARF19 are involved in almost every step of LR development (Lavenus et al., 2013). OsSAUR39 and OsSAUR45 are also seen to regulate auxin synthesis in rice (Oryza sativa; Kant et al., 2009; Xu et al., 2017). How SAUR15 proteins regulate auxin biosynthesis requires further study.

In addition to the LRs and ARs, SAUR15 is expressed in the cotyledons, hypocotyl, and primary root of the light-grown seedlings (Supplemental Fig. S2M). SAUR15 overexpression plants show the corresponding phenotypes such as larger leaves, elongated hypocotyls, and longer primary roots. These data suggest that SAUR15 plays a broader role in auxin-induced cell expansion of organs. These data are consistent with prior studies indicating that SAUR-PP2C-D-PM H⁺-ATPase signaling modules are also implicated in hypocotyl elongation, cotyledon opening, and apical hook formation processes, all of which involve cell expansion (Spartz et al., 2014; Favero et al., 2017; He et al., 2018; Sun et al., 2016; Dong et al., 2019). Involvement of SAUR-PP2C-D-PM H⁺-ATPase signaling modules in these organs’ developmental processes suggests that these distinct biological processes adopted the same regulatory mechanisms during evolution. However, SAUR15 expression is notably absent from the primary root of dark-grown seedlings (Supplemental Fig. S2M), possibly explaining why the etiolated seedling primary root is short and no LRs formed on the etiolated seedlings.

Lastly, we note that in addition to auxin, the expression of many SAUR genes are induced by other growth-promoting pathways, including BRs, gibberellins, and PHYTOCHROME INTERACTING FACTOR (PIF) transcription factors (Oh et al., 2012; Sun et al., 2016). BRs also play an important role in the cell expansion. For example, BR signaling factor mutants show reduced hypocotyl length and LR number (Cho et al., 2014). This intriguing link raises the possibility that SAUR may be involved in BR signaling to promote plant growth.

MATERIALS AND METHODS

Plant Materials and Growth Conditions

All plants used in this study are in the Arabidopsis (Arabidopsis thaliana) ecotype Columbia-0 (Col-0) background. The DR5::GUS transgenic plant line was kept in Jia Li’s lab (Zhou et al., 2018). The saur15-1 mutant (GK-931C06) seeds were ordered from GABI-Kat, and other T-DNA insertion alleles arf6 (CS24606), arf6 arf8 (CS24632), arf7 (CS24607), arf7 arf19 (CS24630), pp2c-d2 (WsDsLOX493G12), pp2c-d5 (GK_330E08), and aha1 (SALK_118350) were obtained from the Arabidopsis Biological Resource Center. Double mutants pp2c-d2 pp2c-d5 and pp2c-d2 saur15-1 were generated by crossing. PCR-based methods were used to genotype mutants. Primers are listed in Supplemental Table S1.

The coding sequences of SAUR15, ARF6, ARF7, PP2C-D1, PP2C-D2, PP2C-D5, PP2C-D6, PP2C-D8, AHA1, and AHA2 were PCR-amplified and introduced into a pDONR/zeo vector using Gateway technology (Invitrogen). The resulting entry clones were used to transfer target sequences into appropriate destination vectors for plant expression. The pSAUR15::SAUR15-FLAG expression construct was generated by cleaving the construct 35S::SAUR15-FLAG with HindIII and KpnI to replace the CaMV35S promoter sequence with the native promoter sequence of SAUR15. The pSAUR15::SAUR15-FLAG construct was used to complement the saur15-1 mutant. The other overexpression constructs, including 35S::SAUR15-FLAG, 35S::ARF6-3HA, 35S::ARF7-3HA, 35S::PP2C-D2-3HA, 35S::AHA1-FLAG, and 35S::AHA2-FLAG were used to create single gene transgenic plants or to create double transgenic plants. These gene expressions were tested by RT-qPCR. ARF6-3HA protein and ARF7-3HA protein were detected by western blot with an anti-HA antibody (Abmart, M20003). The GUS-fusion expression construct was generated by cloning the PCR-amplified promoter sequences from Col-0 plants into binary vector pBIB-BASTA-GWR-GUS (Gou et al., 2010; Wu et al., 2016) and subsequently introduced into Col-0. These transgenic plants in Col-0 were used to cross with other appropriate genotype plants to generate related transgenic plants, and homozygous plants were identified among F2 populations for further GUS staining analyses, apoplastic pH, or ATPase activity assays.

Plants cultivated in pots were grown under long-day conditions (16 h white light per day, ∼100 µmol m⁻² s⁻¹ light intensity; 22°C ± 2°C). For LR analyses, seedlings were grown on one-half strength MS medium under long-day conditions. Photographs were taken with a digital camera, and LRP/LRs were counted with an Olympus light microscope. For AR analyses, 4-d-old dark-grown seedlings were transferred to one-half strength MS medium and grown under long-day light conditions. Pictures were taken with a digital camera, and ARP/ARs were counted with an Olympus light microscope. Apical hook angles were measured from 5-d-old etiolated seedlings as described in Yue et al. (2016).

Root growth on one-half strength MS medium and medium acidification assays were performed as described previously (Spartz et al., 2014).

Statistical Analysis

Quantitative data on the phenotypes, such as measurements of root length, hypocotyl length, and number of LRs and ARs, were subjected to statistical analysis for every pairwise comparison, and significance was assessed by independent-samples t test.

GUS Staining Analyses

GUS activity was assayed with 2 g L⁻¹ 5-bromo-chloro-3-indoxyl-beta-d-glucuronide cyclohexylammonium salt as described by Wu et al. (2016). Images were obtained with a no. DFC 7000 T camera (Leica) affixed to number DM6 B upright fluorescence scope (Leica).

RNA Isolation and RT-qPCR

For RT-qPCR analysis, 7-d-old light-grown or 4-d dark-grown + 6-d light-grown seedlings of different genotypes were treated with or without 1 μm IAA for 1 h. Roots or hypocotyls were collected and immediately frozen in liquid nitrogen and stored at −80°C. Total RNA was isolated using an RNeasy plant mini kit (TIANGEN) and then subjected to first-strand cDNA synthesis according to the manufacturer’s protocol (TaKaRa Biotechnology). RT-qPCR was conducted with Power SYBR green master mix (TaKaRa Biotechnology) and finished on a StepOne real-time PCR thermocycler (Applied Biosystems). ACTIN2 was used as the internal control gene. Primer-BLAST (https://www.ncbi.nlm.nih.gov/tools/primer-blast) was used for primer design. A standard curve and efficiency test were performed for each set of primers (Supplemental Table S1).

Protein Interaction Analyses

An mbSUS yeast (Saccharomyces cerevisiae) two-hybrid system (Obrdlik et al., 2004) was used to detect interactions between SAUR15 and PP2C-Ds according to the manual. The PCR-amplified CDS sequences of SAUR15 and PP2C-Ds were mixed with linearized pMetYCgate vector or pX-NubWTgate vector to transform yeast strain THY.AP4 or THY.AP5 through in vivo DNA recombination. Diploid yeast cells were selected by growing on SC + Ade + His media. Interactions were assessed by spotting serial dilutions of diploid yeast cells on SD media containing 0, 150, or 400 µm Met. AHA2-Cub with PP2C-D1-Nub was used as a positive control (Spartz et al., 2014), and pMetYCgate with pX-NubWTgate was used as a negative control (Obrdlik et al., 2004).

Bimolecular fluorescence complementation analyses were performed as previously described (Zhao et al., 2016). Agrobacterium tumefaciens strain GV3101 containing binary construct SAUR15-YN (SAUR15 fused to the N-terminal half of YFP) was respectively mixed with clones containing PP2C-D-YC (PP2C-D fused to the C-terminal half of YFP) and then infiltrated into young leaves of 3- to 4-week-old Nicotiana benthamiana. YFP fluorescence signals were detected using a confocal microscope (Hu et al., 2002) after 2 d. SAUR15-YN were used as negative control.

ChIP and Luciferase Assays

ChIP was performed as described previously (Ni et al., 2009). ChIP to detect ARF6 and ARF7 binding to the promoter of SAUR15 was performed with 35S::ARF6-3HA and 35S::ARF7-3HA transgenic seedlings or Col-0 as the negative control. Seedlings were grown on one-half strength MS medium for 10 d and then transferred to one-half strength MS liquid medium with or without 1 µm IAA for 2 h before harvesting. Two grams of seedlings and anti-HA tag agarose (M20012M, Abmart) were used for ChIP. Precipitated DNA was dissolved in 50 µL of double-distilled water and 2 µL was used for qPCR.

For luciferase assays, wild-type and mutant SAUR15 promoters were introduced into binary vector pGWB235 to create the constructs pSAUR15-WT::LUC and pSAUR15-d::LUC, respectively. 35S::ARF6-3HA and 35S::ARF7-3HA were used as effectors. Luciferase activity was imaged using a ChemiDoc MP imaging system (Bio-Rad). Experiments were repeated three times with similar results.

Apoplastic pH Measurement

Apoplastic pH was measured according to Cho et al. (2012). At least 50 rosette leaves per genotype were submerged in 50 mL water and subjected to 5 min vacuum cycles with rapid vacuum release for four times. The leaves were dried with filter paper and put into a 5-mL injector without plunger. The exhaust of the injector was linked with a 0.5-mL conical tube and placed into a 50-mL centrifuge tube. The whole mounting was centrifuged at 1000g for 10 min at 4°C. One hundred sixty microliters of apoplastic fluid was mixed with 40 μL 100 mg mL⁻¹ 8-hydroxypyrene-1,3,6-trisulfonic acid trisodium and fluorescence was collected at 510 and 530 nm using an excitation wavelength of 460 nm. The standard curve was made with Britton-Robinson universal buffer containing 0.05 m H₃BO₃, 0.05 m H₃PO₄, 0.05 m CH₃COOH, which was adjusted to pH ranging from 4.5 to 8.0 with NaOH.

Western Blot Analysis

Arabidopsis seedlings grown on one-half strength MS medium were harvested and total protein was extracted and separated on 8% (w/v) SDS polyacrylamide gel to detect the ARF6-3HA and ARF7-3HA protein expression by using an anti-HA antibody (M20003, Abmart). PM of 10-d-old or 3-week-old light-grown Arabidopsis seedlings was prepared according to Minami et al. (2017). The protein concentration was determined by the method of Bradford (1976). PM protein was separated on 7% or 8% (w/v) SDS polyacrylamide gel and PM H⁺-ATPase protein expression level was detected by anti-PM H⁺-ATPase antibodies (AS07260, Agresera). Phosphorylation level of the penultimate residue of PM H⁺-ATPase was estimated by using an anti-pThr947-AHA2 antibody, which was prepared by Abmart Biotechnology, following the methods of Takahashi et al. (2012). Three biological replicates were performed.

ATPase Activity Assays

The same PM samples used for detecting the PM H⁺-ATPase protein expression level was used for H⁺-ATPase activity assay. H⁺-ATPase activity of PM was analyzed by the method of Xu et al. (2012). For each assay, 0.5 mL of reaction solution contained 3 µg membrane protein, 0.02% (w/v) Brij 58, 30 mm 1,3-bis(tris(hydroxymethyl)methylamino)-propane/MES, pH 6.5, 5 mm MgSO₄, 50 mm KCl, and 4 mm TRIS-ATP. Reactions proceeded for 30 min at 30°C and stopped with 1 mL stopping solution containing 2% (v/v) concentrated H₂SO₄, 5% (w/v) SDS, 0.7% (w/v) sodium molybdate, and 50 μL of 10% (w/v) ascorbic acid. After 30 min color development of the phosphomolybdate complex, A₇₀₀ was measured with a spectrophotometer. PM H⁺-ATPase activity was calculated as the phosphorus liberated in excess of boiled-membrane controls. To assess the purification of the PM, the result was expressed as the activity difference with or without 0.1 mm sodium vanadate. Three biological replicates were performed.

Phosphatase Activities Assays

Coding sequences of PP2C-D2, PP2C-D5, and SAUR15 were PCR-amplified and cloned into the pMAL-cRI vector with a MBP tag (New England Biolabs) to generate MBP-tagged protein. Proteins expressed in Escherichia coli were purified with amylose resin (E8021S, New England Biolabs) according to the manufacturer’s instruction. For in vitro phosphatase activities assays, 0.2 μm MBP-PP2C-D2 or MBP-PP2C-D5 was preincubated with 1.0 μm MBP-SAUR15 protein or MBP protein or an equivalent amount of elution buffer for 10 min on ice. Protein mixtures were then added to assay buffer containing 75 mm Tris-HCl, pH 7.6, 10 mm MnCl₂, 100 mm NaCl, 0.5 mm EDTA, and 5 mm pNPP to obtain a final volume of 100 μL. Absorbance at 405 nm was recorded every minute up to 10 min on a Varioskan Flash (Thermo Scientific).

Measurement of Free IAA Levels

Seven-day light-grown seedlings were collected for measurement of IAA levels. The measurement was performed by Suzhou Comin Biotechnology.

Accession Numbers

Sequence data for genes used in this study can be found in The Arabidopsis Information Resource (http://www.arabidopsis.org/) under the following accession numbers: SAUR15 (AT4G38850), ARF6 (AT1G30330), ARF7 (AT5G20730), ARF8 (AT5G37020), ARF19 (AT1G19220), PP2C-D1 (AT5G02760), PP2C-D2 (AT3G17090), PP2C-D5 (AT4G38520), PP2C-D6 (AT3G51370), PP2C-D8 (AT4G33920), AHA1 (AT2G18960), AHA2 (AT4G30190), and ACTIN 2 (AT3G18780).

Supplemental Data

The following supplemental materials are available.

• Supplemental Figure S1. Phylogenetic tree of SAUR family genes from Arabidopsis.

• Supplemental Figure S2. Auxin-related phenotypes.

• Supplemental Figure S3. Expression of SAUR15 is regulated by auxin.

• Supplemental Figure S4. ARF6 and ARF7 directly bind to the promoter of SAUR15.

• Supplemental Figure S5. Auxin-related phenotype of pp2c-d mutants.

• Supplemental Figure S6. SAUR15 activates PM H⁺-ATPases.

• Supplemental Table S1. Primers used in current study.