Phytohormone inhibitor treatments phenocopy brassinosteroid–gibberellin dwarf mutant interactions in maize

Summary Phytohormone biosynthesis produces metabolites with profound effects on plant growth and development. Modulation of hormone levels during developmental events, in response to the environment, by genetic polymorphism, or by chemical application, can reveal the plant processes most responsive to a phytohormone. Applications of chemical inhibitors and subsequent measurements of specific phytohormones can determine whether, and which, phytohormone is affected by a molecule. In many cases, the sensitivity of biochemical testing has determined multiple pathways affected by a single inhibitor. Genetic studies are not subject to this problem, and a wealth of data about the morphological impacts of hormone biosynthetic inhibition have accumulated through the study of enzyme mutants. In this work, we sought to assess the specificity of three triazole inhibitors of cytochrome P450s by determining their abilities to recapitulate the phenotypes of single and double mutants affected in the production of brassinosteroid (BR) and gibberellin (GA) biosynthesis. The GA biosynthetic inhibitors uniconazole (UCZ) and paclobutrazol (PAC) were applied to the BR biosynthetic mutant nana plant2 (na2), and all double‐mutant phenotypes were recovered in the UCZ treatment. PAC was unable to suppress the retention of pistils in the tassels of na2 mutant plants. The BR biosynthetic inhibitor propiconazole (PCZ) suppressed tiller outgrowth in the GA biosynthetic mutant dwarf5 (d5). All treatments were additive with genetic mutants for effects on plant height. Due to additional measurements performed here but not in previous studies of the double mutants, we detected new interactions between GA and BR biosynthesis affecting the days to tassel emergence and tassel branching. These experiments, a refinement of our previous model, and a discussion of the extension of this type of work are presented.


| INTRODUCTION
Phytohormones control plant growth and development at very low concentrations. The entire life cycle of plants is influenced by the availability and amount of these metabolites. The number of identified classes of phytohormones currently includes auxins, cytokinins (CK), abscisic acid (ABA), jasmonic acid, salicylic acid, strigolactones (SL), brassinosteroids (BR), and gibberellins (GA) (Gomez-Roldan et al., 2008;Kende & Zeevaart, 1997;Santner, Calderon-Villalobos, & Estelle, 2009). The biosynthesis, transport, signaling, and responses to these phytohormones have been intensely studied in many species, including Zea mays (maize). Both biochemical and genetic studies identified genes encoding phytohormone biosynthetic steps and mutants blocking hormone biosynthesis in many genetic model systems including Pisum sativum (pea), Solanum lycopersicum (tomato), Arabidopsis thaliana (Arabidopsis), and maize (Bishop, Harrison, & Jones, 1996;Hartwig et al., 2011;Klahre et al., 1998;Lester, Ross, Davies, & Reid, 1997;Mendel, 1866). Despite these studies, the molecular identities of many steps are as yet unidentified. Furthermore, the roles of phytohormones in processes that do not occur in established genetic models are less well explored. Biochemical inhibitors, particularly the triazoles, have elucidated physiological effects of these phytohormones in plant systems that lack the expansive genetic resources of species with vibrant genetics research communities like Arabidopsis and maize. The utility of inhibitor treatments for determining the roles of phytohormones in plant processes is determined by the specificity of the biochemical inhibitors. A test of this specificity is available if we determine the concordance between phenotypes induced by reducing hormone biosynthesis via genetic ablation of biosynthetic enzymes and the phenotypes observed after chemical inhibition.
Mutants in BR and GA were additive for plant height, indicating independent effects on this trait. BR mutants were epistatic to GA mutants for tiller outgrowth, but GA mutants were epistatic to BR mutants for POPIT (Best, Hartwig et al., 2016).  Figure 1. PGR specificity within plants is known to be poor with wide impacts across P450s (Figure 2). It is currently unknown how wide these impacts actually are (Rademacher, 2000).
Propiconazole was first identified as a fungal growth inhibitor that blocked the C-14 demethylation of lanosterol ( Figure 1a) (Weete, 1987;Weete & Wise, 1987). PCZ is a specific and potent inhibitor of BR biosynthesis, as evidenced by the reversal of growth inhibition by the co-application of BR (Sekimata et al., 2002). PCZ binds the CYP90D1 of Arabidopsis (Figure 2a), but it is unknown whether PCZ also binds other P450s in BR biosynthesis such as CYP90B1/DWF4, CYP90C1/ROT3, or CYP90A1/CPD (Oh, Matsumoto, Hoshi, & Yoshizawa, 2016). The related BR biosynthesis inhibitor, brassinazole, effectively binds to CYP90B1, and the effectiveness of these two inhibitors may stem from an ability to simultaneously inhibit multiple CYP90/CYP85 BR biosynthetic enzymes.
Cotreatment of Arabidopsis with epi-brassinolide reversed the phenotypic effects of both brassinazole and PCZ, suggesting that the F I G U R E 1 Structures of chemicals used in this study. Chemical structure of (a) propiconazole, (b) paclobutrazol, and (c) uniconazole F I G U R E 2 All three plant growth regulators are known to affect more than one pathway. (a) Brassinosteroid, (b) gibberellin, (c) cytokinin, and (d) ABA metabolic pathways. Metabolites are indicated in black text and enzymes by black arrows with respective names in gray text. Functionally characterized enzymes in maize are indicated in parentheses. Previously identified enzymes inhibited by PCZ (yellow), PAC (purple), and UCZ (green) are indicated by colored ovals effects of these triazoles were predominantly due to BR inhibition (Asami, Min, & Nagata, 2000;Asami & Yoshida, 1999;Sekimata et al., 2002). PCZ is also used as a fungicide in agriculture due to inhibition of ergosterol biosynthesis. It is possible that PCZ also affects structural sterol biosynthesis in plants. To date, no experiments have been conducted on PCZ and enzymes in other metabolic pathways to determine effects on other phytohormones or sterol biosynthesis.
Paclobutrazol was also first identified as a fungicide (Figure 1b) (Sugavanam, 1984). In fungi, it inhibits C-14 demethylation, similar to PCZ. In barley and celery cell cultures, it was shown to inhibit the 14a-demethylase CYP51 required for the conversion of obtusifoliol to D 8 -14 -sterol ( Figure 2a) (Burden, Clark, & Holloway, 1987;Burden, Cooke, & Carter, 1989;Haughan, Lenton, & Goad, 1988). Supplementation of cell cultures with stigmasterol at high concentrations, or a mix of cholesterol and stigmasterol, reversed the growth retardation that resulted from PAC treatment of celery cell cultures.
Ratios of stigmasterol to sitosterol, a 22-desaturation carried out by CYP710A, were also altered in PAC-treated cells suggesting that this P450 may also be a target. The BR precursor campesterol was also reduced by PAC treatment, and it may be that PAC achieves growth retardation by affecting structural sterols as well as BR levels (Haughan et al., 1988). PAC is also known to reduce GA levels by inhibiting the ent-kaurene oxidase/CYP701 in plants ( Figure 2b) (Rademacher, 2000;. Hedden & Graebe Hedden & Graebe, 1985) showed the percent conversion of the CYP701 targets, ent-kaurene, ent-kaurenol, and ent-kaurenal, into later intermediates was greatly inhibited by 1 lM (2RS, 3RS)-PAC in a cell-free system from Cucurbita maxima endosperm. The different enantiomers of PAC have demonstrated differential effects on plants as compared to fungi.
The 2R, 3R diastereoisomer of PAC affected sterol biosynthesis in fungi but had no effect on plant height at 170.2 lM, while the 2S, 3S diastereoisomer dramatically reduced plant height at 34.0 lM (Sugavanam, 1984). A number of other downstream effects have been documented in PAC-treated plants including reduced accumulation of 1-aminocyclopropane-1-carboxylic acid (ACC) and ethylene production in water-stressed apple seedlings; however, the mode of action is unknown (Wang & Steffens, 1985).
Uniconazole differs from PAC by one desaturation leading to a double bond ( Figure 1c) and has been shown to inhibit GA biosynthesis by affecting the same enzymes ( Figure 2b) (Izumi, Kamiya, Sakurai, Oshio, & Takahashi, 1985;Izumi, Yamaguchi, Wada, Oshio, & Takahashi, 1984;Rademacher, 2000). However, the height of a GA 3 -insensitive DELLA mutant in Helianthus annuus (sunflower) was further reduced by UCZ treatment (Best, Wang et al., 2016). UCZ has also been shown to affect structural sterol levels in Oryza sativa (rice) and pea ( Figure 2a) (Khan et al., 2009;Wagatsuma et al., 2015). In addition, UCZ treatment reduced castasterone accumulation in pea and blocked BR-induced tracheary element differentiation in Zinnia elegans L. mesophyll cells (Iwasaki & Shibaoka, 1991;Yokota et al., 1991). Whether these phenotypes result from BR inhibition, the dependence of BR-induced tracheary element differentiation on GA, or some other impact is unknown. Recently, it has been shown that UCZ affects ABA catabolism by inhibiting CYP707A in Arabidopsis and tobacco cell cultures (Figure 2d) (Kitahata et al., 2005;Mizutani & Todoroki, 2006;Saito et al., 2006). UCZ also affects cytokinin biosynthesis; however, there are conflicting reports depending on the species tested. In Arabidopsis, it inhibited transzeatin accumulation (Figure 2c), but UCZ treatment in rice and Glycine max showed higher levels of trans-zeatin (Izumi et al., 1988;Sasaki et al., 2013;Zhang et al., 2007). Izumi et al. (1988) showed that ethylene levels were higher but that ABA was unaffected in UCZ-treated rice. Of the three PGRs used in this study, UCZ has been the most studied. The conflicting reports on hormonal levels in different species highlight the necessity of studying the specificity of PGRs within the species being tested due to potential differences in uptake, transport, or binding affinity by potential target P450s (Rademacher, 2000).
We sought to test the model for BR-GA interaction in maize derived from our previous genetic experiments (Best, Hartwig et al., 2016). Given the multiple interactions between the mutants, it is less likely that the same effects, and direction of interaction, will be generated by non-specific growth retardation or general toxicity of a chemical treatment. In addition, using the genetics experiments as orthogonal data confirming the utility of these inhibitors' specificity in maize would permit stronger interpretation of inhibitor studies across the grasses, including species without the genetic resources of maize. Orthogonal data, if confirmatory, will also bolster our confidence in the interpretation of previous genetic interactions between BR and GA discovered in maize (Best, Hartwig et al., 2016).
In this study, we treated the na2-1 BR and d5 GA biosynthetic mutants with each of the three triazole inhibitors described above and measured phenotypes known to result from BR and GA reduction. All three inhibitors were additive with the two mutants for plant height. UCZ had the strongest impact on plant height. Consistent with the expectation for reduced BR biosynthesis, PCZ enhanced the POPIT phenotype of BR mutants but was unable to induce POPIT in wild-type siblings or d5 mutants at the concentra-       (Table 1). Organs and internodes were shorter in PCZ-treated plants than in mocktreated plants for both na2-1 and wild-type siblings, except internode 4 for na2-1 (Table 1 and Table 2). PCZ treatment had a dramatic effect on na2-1 mutant plant height indicating either bioactive BR accumulation in the mutant or non-specific inhibition of other growth-promoting metabolites by PCZ application (Table 1). Primary tassel branch number was 10-fold lower in PCZ-treated na2-1 than in treated wild-type siblings, whereas mutants and wild types were indistinguishable in the mock-treated samples (Table 1). Tassel length was also non-additively affected by PCZ application to na2-1 mutants, causing a decrease by 40% in wild type and 80% in na2-1 mutants. Just as tassels emerged later in na2-1 mutants than in wild types, PCZ treatment extended the number of days to tassel emergence in wild types. PCZ treatment of the mutants further enhanced this delay (Table 1). Yet, in neither comparisons between na2-1 and wild type nor in PCZ treatments did the number of nodes per plant change. Taken together, PCZ reproduced all but the POPIT phenotype in wild-type treated plants and enhanced all loss-of-BR phenotypes in the na2-1 mutant, including increased POPIT penetrance.
Organs and overall plant height were reduced by PAC treatment in both na2-1 and wild-type siblings, except internode 13 and 14 for na2-1 (Table 1 and Table 2). Treatment of wild-type siblings with F I G U R E 3 Effects of plant growth regulators and genotype on internode length. Mean internode lengths in cm and SD of (a) mock-treated, (b) PCZ-treated, (c) PAC-treated, and (d) UCZ-treated na2-1 and wild-type siblings. Asterisks indicate statistical difference by Student's t test at p < .05 between na2-1 and wild-type siblings at the respective internode PAC resulted in a significant increase in tiller outgrowth similar to the GA biosynthetic mutant d5. The mock-treated na2-1 mutants did not tiller, and na2-1 suppressed the tiller outgrowth affected by PAC treatment. PAC did not suppress POPIT in na2-1 mutants. Similar to the phenotypes of d5 mutants (Table 3), PAC treatment of wild-type siblings caused the lower leaf angle to become more upright. Contrastingly, the upper leaf angles of wild-type siblings went the opposite direction and were less upright following PAC treatment. As was the case for tillering, na2-1 suppressed the effects of PAC on leaf angle and no discernable difference between mock and PAC treatments was observed. Similar to PCZ, primary tassel branch numbers of na2-1 plants were reduced by PAC treatments. The days to tassel emergence was increased by PAC treatment of both na2-1 and wild types. Surprisingly, PAC treatment reduced the number of nodes below the top ear and the total number of nodes per plant in na2-1 but not wild-type siblings. Thus, node number decreased as days to tassel emergence increased. For all phenotypes except suppression of POPIT and induction of anther-ear, PAC recapitulated the phenotypes observed in d5 mutants as well as the genetic interactions of d5 and na2.
Treatment with UCZ reproduced all of the effects of a loss of GA biosynthesis and the genetic interactions between d5 and na2.
The effect of UCZ on organ length and plant height of na2-1 and wild-type siblings was similar to what was observed in PAC and PCZ treatments (Table 1 and Table 2). Concordant with a strong inhibition of GA biosynthesis, UCZ caused ears of both na2-1 and wildtype siblings to exhibit florets with persistent anthers. Similar to d5 mutants and PAC treatment, UCZ-treated wild types profusely tillered. Consistent with the genetic BR-GA interactions, na2-1 strongly suppressed UCZ-induced tillering. UCZ treatment reduced the frequency of POPIT in na2-1 by over twofold, but this difference was not statistically significant from mock-treated na2-1 mutants.
UCZ treatments had a synergistic effect on leaf width in combination with na2-1 and caused na2-1 leaves to become narrower. UCZ resulted in a dramatic decrease in primary tassel branch number, similar to the other inhibitors. UCZ treatment delayed the number of days before tassel emergence in both mutants and wild-type siblings.
UCZ treatment nominally, but not statistically significantly, reduced both the number of nodes before the top ear and the total number of nodes per plant in na2-1, but not wild-type siblings. Thus, UCZ treatments of na2-1 delayed tassel emergence without the additional production of nodes. Taken together, UCZ treatments reproduced all of the phenotypes present in GA biosynthetic mutants.

| Effect of hormone biosynthetic inhibitors on d5 and wild-type siblings
The d5 mutants were shorter, had shorter organs, exhibited persistent anthers in the ear florets, had delayed tassel emergence, reduced primary tassel branch numbers, and tillered profusely (Table 3 and   anther-ear and the reduction in tassel length was less pronounced in PAC-treated wild types than in mock-treated d5 mutants. PAC further reduced the height and organ lengths of d5 plants (Table 3 and   Table 4). Similar to the effects of PAC on primary tassel branch number in na2-1 mutants, PAC reduced primary tassel branch number to zero in d5. The d5 mutants had greater leaf widths than wildtype controls. PAC treatment increased leaf widths of wild types, but PAC-treated d5 mutants had narrower leaves than mock-treated mutants. The d5 mutants had the same number of nodes as wild type, and unlike PAC-treated na2-1, PAC did not affect node number of d5 mutants. PAC treatment delayed tassel emergence in wildtype siblings and further delayed tassel emergence in d5 mutants as compared to mock treatments. Thus, a delay in tassel emergence without an increased number of nodes was observed for d5 treated with PAC.
As was observed in applications of UCZ to na2-1 and wild types, application of UCZ to d5 and wild-type siblings recapitulated all lossof-GA phenotypes (Figure 6a-b,d). Application of UCZ decreased plant height and the lengths of all organs and internodes for both d5 and wild-type siblings (Table 3 and Table 4). Similar to the effects of UCZ on na2-1, UCZ narrowed both the upper and lower leaves of The BR inhibitor PCZ reduced overall plant height, internode lengths, and the lengths of all measured organs for both d5 and wild-type siblings (Table 3 and Table 4). Primary tassel branching decreased dramatically in both PCZ-treated d5 mutants and their wild-type siblings. Just as in the wild-type controls in the na2-1 experiment (Table 1), PCZ application to d5 and wild-type siblings did not reproduce the na2-1 effect on POPIT (Table 3). PCZ treatment also failed to suppress persistence of anthers in the ear florets of d5 mutants. The d5 mutants exhibited increased tillering and PCZ treatment suppressed tiller outgrowth induced by the loss of d5 (Table 3). PCZ treatment increased the angle of the upper leaf in d5 mutants and the angle of the lower leaf in wild types. Days to tassel emergence were greater in PCZ-treated d5 and wild-type siblings consistent with the tassel emergence delay in PAC treatment of na2-1 plants. Thus, PCZ treatment of d5 mutants and wild-type siblings delayed tassel emergence, but no discernable difference was observed in the total number of nodes.

| DISCUSSION
We report here the effects of PCZ, PAC, and UCZ treatment on the phenotypes of na2-1, d5, and wild-type siblings at maturity. PCZ treatment of wild-type plants was able to recapitulate the phenotype of na2-1, except for the POPIT phenotype (Table 1, Table 2, and Figure 4b-c). However, a previous study used PCZ at twice the concentration (500 lM) that we used in this study and they did observe POPIT in PCZ-treated wild-type plants (Hartwig et al., 2011). Offtarget and non-specific effects of inhibitors are a greater concern as concentrations increase, and we sought to minimize these with lower concentrations. The observation that PCZ treatment even at a lower concentration decreased plant height and increased POPIT in na2-1 mutants (  (Hong et al., 2005). Targeted metabolic analysis of all possible BR intermediates in the maize na1, na2, and brd1 mutants is required to determine whether this is the case.  Data are presented as means with SD. Lowercase letters indicate connecting letter report as determined by ANOVA with post hoc analysis using the Holm-Sidak algorithm with p < .05.
Brassinosteroid and GA interact to control plant development, and the concordance or discordance of PCZ, PAC, and UCZ with the phenotypes presented in the mutants of BR and GA biosynthesis is shown in Figure 7. This summarizes the hypothesis test central to this study, that these inhibitors achieved their phenotypes via the specific inhibition of BR or GA biosynthesis. UCZ treatment to wildtype plants was able to phenocopy d5 mock-treated plants, and UCZ-treated na2-1 plants displayed all the interactions observed following the loss of both GA and BR biosynthesis in na2-1/d5 double mutants (Best, Hartwig et al., 2016) (Table 1, Table 3, and Figure 7).
UCZ reduced plant height, promoted tiller outgrowth, induced anther persistence in the ear, and suppressed POPIT induced by the loss of BR biosynthesis in the na2-1 mutant (Table 1, Table 3, and Figure 7).
These results are consistent with UCZ affecting maize growth via specific inhibition of GA biosynthesis at the concentrations employed. However, the suppression of tiller outgrowth induced by a loss of na2 was the least effective of all treatment and genetic combinations observed (  (Saito et al., 2006). This reaction contributes to the formation of phaseic acid which has been demonstrated to have additional biological activities of its own (Figure 2d) (Hill et al., 1992;Sharkey & Raschke, 1980). The perfect agreement of the developmental phenotypes caused by UCZ and the GA mutants was somewhat surprising in light of these additional modes of action ( Figure 2a-d). It is possible that the weak suppression of tiller formation by na2-1 in UCZ treatments may result from inhibition of other P450s. It is formally possible, although we have no evidence, that UCZ might affect P450s in strigolactone biosynthesis or that ABA might enhance tillering in maize. UCZ also delayed tassel emergence in wild-type, d5, and na2-1 plants (Tables 1 and 3). No effect on node number was observed in wild-type and d5 plants treated with UCZ as compared to controls (Table 3). However, UCZ-treated na2-1 displayed fewer nodes per plant than UCZ-and mock-treated wild types (Table 1 and Figure 8). The simplest interpretation of our results is that UCZ affects all of the traits we measured via inhibition of GA, for which there is ample evidence (Rademacher, 2000). Further research, particularly work comprehensively evaluating the binding and catalytic inhibition of recombinant P450s by UCZ, as was done for ABA catabolism and UCZ, is needed to test these hypotheses and establish the specificity of these triazole inhibitors.
Paclobutrazol treatment was unable to phenocopy all of the effects of GA-deficient mutants in maize (Figure 7). PAC treatment of wild type and na2-1 did not promote anther retention in the ear florets nor did it suppress POPIT in na2-1 plants (Table 1). This contrasts with both UCZ treatments and our previous genetic results (Best, Hartwig et al., 2016). The concentration of PAC used was the same as UCZ, but it may be that PAC is a weaker inhibitor as has been demonstrated in sunflower, Impatiens walleriana, Salvia splendens, Tagetes erecta L., and Petunia hybrid (Barrett & Nell, 1992;Whipker & Dasoju, 1998). If PAC is effectively inhibiting GA biosynthesis, this would contradict maize double-mutant analyses where GA biosynthetic mutants were epistatic to BR biosynthetic mutants for the expression of POPIT (Best, Hartwig et al., 2016). More likely the discordance between PAC and genetic ablation resulted from incomplete inhibition of GA biosynthesis by our repeated root drenches with 60 lM PAC. The effects of treatment on height and stimulation of tiller outgrowth in wild-type plants, and suppression of outgrowth by na2-1, matched the effects of reduced GA biosynthesis (Tables 1 and 3). The inability of PAC treatment to induce anther-ear in any genotype, except for d5, or suppress POPIT in na2-1 may be due to the concentration used, unspecific inhibition of other pathways besides GA biosynthesis (Figure 2), or combination of the two. The observation that PAC growth arrest can be reversed by application of sterols raises the possibility that PAC achieves some height reduction by affecting both BR and GA levels (Haughan et al., 1988). If BR biosynthesis is inhibited by PAC treatment, this could increase the penetrance of POPIT in the na2-1 mutant, as was observed in PCZ treatment, and require greater reduction in GA to prevent floral organ persistence. However, the lack of anthers in ear florets in PAC treatments, which were observed in all maize GA mutants, suggested that PAC might simply be a weaker GA biosynthesis inhibitor.
F I G U R E 7 Summary of inhibitor tests of the genetic model of BR-GA interaction during maize growth. Check marks indicate concordance of inhibitor treatment and combined inhibitor and mutant with respective biosynthetic single-and double-mutant analysis presented in Best, Hartwig et al. (2016). PCZ is indicated in yellow, UCZ in blue, and PAC in red. Dashed lines indicate trait measured; lines with arrows indicate positive regulation; lines with blocked arrows indicate negative regulation; and lines with diamond heads indicate synergistic positive regulation. GA mutants were epistatic to BR mutants for the presence of pistils in tassel (POPIT), and both PCZ and UCZ treatments were concordant with this interaction, while PAC was not. GA inhibits anther-ear (AE), and both UCZ and PAC phenocopied the mutants. BR mutants were epistatic to GA mutants for tiller outgrowth (T), and both PCZ and UCZ were concordant for this interaction, while PAC was not. GA and BR were additive for plant height (PH), and all three inhibitors were concordant for this phenotype BEST ET AL.

| 13
The combined effects of loss of na2 and any of the inhibitor treatments on plant height resulted in extreme dwarfism. The ability of PCZ to further inhibit the growth of na2-1 mutants, and to enhance the POPIT phenotype of na2-1 plants, is consistent with residual BR activity in na2-1. The presence of BR activity in mutants of the rice ortholog of na2-1 (Hong et al., 2005) is consistent with pathway non-linearity and accumulation of BR intermediates with limited biological activity in single mutants. The height reduction in na2-1 treated with PCZ (7.1 cm; Table 1) was indistinguishable from na2-1 treated with UCZ (6.5 cm) and both of these were significantly shorter than na2-1 treated with PAC (9.7 cm). In previous studies of Arabidopsis seedlings, GA biosynthesis was under the control of BR signaling where BR signaling promoted GA biosynthesis (Tong et al., 2014;Unterholzner et al., 2015). Our previous study of the interactions between the na2 and d5 mutants on floral organ retention in maize suggested that the opposite of this might be true in maize tassel florets where GA synthesis was downstream of a loss of BR (Figure 7). The finding that PAC was less effective than UCZ at suppressing both plant height and POPIT in the na2-1 BR-deficient background suggested that a complete inhibition of GA biosynthesis is required to recapitulate the genetic interactions with inhibitor treatments. The failure of UCZ and PAC to affect elongation beyond the effect of PCZ suggests that there is little to no compensation for the loss of BR by increased GA biosynthesis in maize stems. Measurements of BR and GA levels and of the expression of signaling outputs such as DELLA-dependent and BZR1/BES1-dependent gene expression in the mutants and inhibitor treatments are required to clarify the relationships between these hormones and resolve the apparent differences between tissues within maize and the difference between maize tassels and Arabidopsis seedlings. The rice GRAS-domain transcription factor DWARF AND LOW-TILLERING (DLT) affects both BR and GA (Li et al., 2010;Tong et al., 2009Tong et al., , 2012, resulting in a suppression of tiller formation much like the na2 mutants of maize. We expect, based on the phenotypes in rice and Arabidopsis, that double mutants between na2-1 and loss-offunction alleles at the maize DLT ortholog (AC234164.1_FG004) will exhibit increased POPIT due to an increase in GA biosynthesis and that double mutants between d5 and loss-of-function alleles at the maize DLT ortholog will suppress tiller outgrowth affected by loss of GA biosynthesis in d5.
The number of nodes was decreased in na2-1 treated with PAC and UCZ; however, d5 treated with PCZ showed no effect on total nodes (Figure 8a). The discordance between these may stem from the incomplete inhibition of BR biosynthesis by PCZ. If so, the loss of both BR and GA results in a synergistic reduction of total nodes per plant (Figure 8a). Days to tassel emergence was significantly longer for na2-1 and d5 (Tables 1 and 3) compared to their wild-type siblings, similar to our earlier work (Best, Hartwig et al., 2016). The tassels of wild-type siblings treated with PCZ, PAC, or UCZ emerged from the leaf whorl later than mock-treated wild-type controls. Every mutant-triazole combination further increased days to tassel emergence, and this difference was always statistically significant. The delay in tassel emergence in triazole-treated na2-1 mutants was associated with the production of fewer nodes per plant despite the , and all three inhibitors were concordant for these phenotypes 2-3 extra weeks of growth before tassel emergence. This difference in node number itself was significantly different than na2-1 mocktreated controls for only the PAC treatment, but the delay in flower emergence without additional node production suggests the plastochron depends on BR and GA in maize. The plastochron is the time it takes to initiate each lateral organ from the meristem (Erickson & Michelini, 1957;Meicenheimer, 2014). We quantified total number of nodes and time to tassel emergence, so we may overestimate the effect of these treatments on flowering time as decreased elongation should delay emergence of the tassel. However, the average increase in days to tassel emergence for na2-1 or d5 treated with any inhibitor was 19.0 days and 13.4 days, respectively (Tables 1 and 3). Furthermore, PAC and UCZ treatment of na2-1 mutants caused the top ear to reside on a lower node (Table 1); thus, the developmental stage of these plants was different from controls even when observed at the same number of nodes. This could result from BR and GA effects on developmental rate and flowering time or inhibition of another cytochrome P450 by these triazoles. Overexpression of the maize plastochron1 gene, which encodes cytochrome P450 CYP78A1, increased leaf number but decreased plastochron (Sun et al., 2017). Loss-of-function CYP78A1 mutants exhibited slower leaf elongation, but interpretation is limited by the presence of multiple CYP78A paralogs in the maize genome.
An inhibitor of all CYP78A paralogs would circumvent the need to isolate loss-of-function mutants in all paralogs. As yet, we do not know the phenotype of low CYP78A activity on maize plastochron or the identity of the bioactive metabolite affected by CYP78A enzymes in plants.
We measured traits in this study that were not previously assessed in na2-1/d5 double mutants. Thus, we cannot use these observations as a test of a genetic model. Rather, we hypothesize the BR-GA interactions, or lack thereof, based on this inhibitor study to be tested in future genetic experiments. No interaction between BR and GA for length of the upper and lower leaves was evident.
na2-1 and treatment with PAC or UCZ additively reduced both traits ( Figure 8b and c). A similar additive effect was observed for d5 treated with PCZ (Figure 8b and c). Leaf width did not consistently respond in mutants, treatments, and their combination (Figure 8b and c). The width of the upper leaf was not significantly different for na2-1 and d5 as compared to their respective wild types. Mutants of na2-1 treated with either GA inhibitor and d5 mutants treated with PCZ had narrower upper leaves than mock-treated mutants, but only the UCZ-treated na2-1 was significantly different from mock-treated mutants (Tables 1 and 3). Lower leaf width was not consistently affected by na2-1 or PCZ (Table 1). The width of the lower leaf of d5 mutants and PAC-and UCZ-treated wild-type siblings were significantly wider than mock-treated wild-type plants (Table 3). Taken together, this suggests that loss of GA widened the lower, but not upper, leaf and that loss of BR and GA synergistically resulted in narrower upper leaves (Figure 8c). Similar to the effect on leaf widths, there were no consistent effects of BR and GA on upper leaf angle (Tables 1 and 3). Lower leaves were more upright in both na2-1 and d5 than wild-type siblings. Treatment of d5 with PCZ and na2-1 with PAC or UCZ did not make the leaves more upright. Therefore, we propose that BR and GA act on the same pathway to increase lower leaf inclination (Figure 8c). Like plant height, tassel length was additively reduced by a loss of BR and GA (Figure 8d) in comparisons between mock-treated mutant controls and na2-1 treated with PAC and UCZ or d5 treated with PCZ (Tables 1 and 3).
Another finding of our experiments was an effect of GA biosynthesis on tassel branch numbers, particularly in the na2-1 background.
The d5 mutants had fewer primary tassel branches than wild-type siblings, but na2-1 mutant primary tassel branch numbers were unaffected in both the results of this study (Tables 1 and 3) and our previous work (Best, Hartwig et al., 2016). Treatment of wild-type plants with PAC or UCZ also reduced primary tassel branch numbers (Tables 1 and 3) consistent with GA promotion of tassel branching ( Figure 8d). All three triazoles severely reduced branching in na2-1 and d5 mutants (Tables 1 and 3). The wild-type siblings of d5 mutants, but not the wild-type siblings of na2-1, exhibited fewer primary tassel branches following PCZ treatment. That na2-1 mutants did not have fewer branches than wild type in mock treatments may result from these genetic backgrounds differentially responding to BR reduction.
As was the case for POPIT (Table 1), the treatment of na2-1 mutants with PCZ did reduce tassel branch numbers (Table 1). Thus, we propose that GA and BR additively promote tassel branching (Figure 8d), although a synergistic interaction or off-target effects of the inhibitors are possible. Residual BR biosynthetic activity in all na2 alleles is suggested by the stronger phenotype of brd1 mutants and na1/na2 double mutants (Dilkes and Best, data not shown) and it may be that the residual BR are sufficient to maintain normal tassel branching in na2-1 mutants (Makarevitch et al., 2012). Alternatively, PCZ might have effects on regulators of tassel branching encoded by P450s outside of BR biosynthesis. In our previous work, exogenous application of GA to na2-1, d5, and wild-type growing apices did not alter tassel branch numbers (Best, Hartwig et al., 2016). Additional work determining the physiological basis of triazole application on tassel branching may uncover as yet unknown impacts of GA on tassel architecture, or implicate triazoles in the inhibition of other branch-promoting signals.
We repeatedly treated plants with PGRs, and this continuous treatment was required to observe the dramatic decrease in plant height. This requirement suggests that maize can catabolize or inactivate these PGRs over the course of a few days. Reduction in the concentrations of bioactive PGRs should permit the biosynthesis of phytohormones to return and promote growth in responsive cells.
We ceased treatment at tassel emergence during the elongation of the uppermost internodes, and the lengths of these internodes were not as affected in our treated plants (Table 2 and Figure 3). We propose that PGR inactivation rather than phytohormone-independent growth underlies these differences in internode responses.
The ability to combine chemical and genetic inhibition of phytohormone biosynthesis to assess interactions and combinatorial effects is limited only by resource availability. In conclusion, PCZ treatment of wild-type siblings largely phenocopied na2-1 mutants and PAC or UCZ treatment predominantly phenocopied d5 mutants (Figure 7). PAC did not affect sexual organ persistence, unlike GA biosynthetic mutants or UCZ treatment. There was no strong evidence of non-specific growth inhibition by these three triazole inhibitors in maize. Unexpectedly, primary tassel branch number was reduced in na2-1 or d5 treated with PCZ, PAC, or UCZ.
Thus, either all three triazoles affect another pathway controlling tassel branching or BR and GA biosynthetic pathways interact to affect tassel branching. Days to tassel emergence was synergistically increased by PCZ, PAC, or UCZ treatment of both na2-1 and d5. The mechanism for this is unknown, and we did not measure flowering time or days to tassel emergence in na2-1/d5 double mutants in our previous study (Best, Hartwig et al., 2016). The similarity of the effects on plant height, tiller outgrowth, and floral organ persistence in the mutant analyses and inhibitor treatments was consistent with the genetic model proposed previously (Best, Hartwig et al., 2016).
The concordance of phenotypes increased the confidence of interpretation for UCZ and PCZ as GA and BR inhibitors, respectively. Using these inhibitors in the greenhouse via simple repeated pot drench will allow for rapid assessment of phytohormone function, even in the Maize genotypes na2-1, d5, and their respective wild types were described previously (Best, Hartwig et al., 2016). Both mutants were recovered from segregating families. As such, all comparisons are between mutants and wild-type siblings. as evidenced by silk outgrowth, the plant was scored as "yes." Ears were inspected for anthers and given a binary yes/no score similar to that for POPIT. Flowering time was estimated as DAG to the day the tassel rachis was visible within the leaf whorl. Statistical analysis of differences between genotypes and between treatments for con-