THE EFFECT OF METHYL JASMONATE ON Zea mays TASSEL DEVELOPMENT

Methyl jasmonate (MeJA) is a lipid-derived plant hormone that mediates diverse biological phenomena. Application of MeJA onto rice spikelet could exhibit abnormal floral organ development. Although jasmonic acid (JA) has been proved to be involved in maize tassel sex determination process, the roles of JA and its precursor MeJA in maize tassel development still remain obscure. In this study, we found that tassel development was decelerated by application of 2 mM MeJA. Exogenous MeJA also influenced the number of palea and stamens of tassel spikelets. Exogenous MeJA increased the expression level of some key regulator genes, which may responsible for the phenotypic change in MeJA-treated tassel, and may mediate the crosstalk between MeJA and other hormones.


INTRODUCTION
As one of the most important agricultural crops in the world, maize (Zea mays L. ssp. mays) is an excellent model plant for grass developmental biology research. Maize forms male and female inflorescences on a terminal tassel and on lateral ears, respectively (Supplementary figure 1A, TANAKA et al., 2013). At an early developmental stage, tassel and ear both initiate bisexual spikelets. Each spikelet contains two florets. Each floret consists of a lemma, a palea, two lodicules, three stamens, and a pistil, and is subtended by two glumes (Supplementary figure 1B). In a so-called sex determination process at a later developmental stage, the pistil in a tassel spikelet and the stamen in an ear spikelet will cease to develop and the plant will ultimately form unisexual spikelets (Supplementary figure 1C). Thus, to form a mature tassel spikelet, a delicate coordination between the establishment of floral organ identity and pistil abortion is required. Over the past three decades, molecular analyses have led to detailed insights into the genetic regulatory network for Arabidopsis flower morphogenesis (Ó'MAOILÉIDIGH et al., 2014). A classic ABCDE model explains how a series of genes, named class A, B, C, D and E genes, delicately specify the floral organ`s identity independently or coordinately (YANOFSKY et al., 1990;COEN;HONMA;GOTO, 2001 ). Among these transcription factor genes, the class C gene AGAMOUS (AG) is one of the most important regulators. Besides specifying carpel identity, the AG gene affects stamen identity in combination with two other class B genes APETALA3 (AP3) and PISTILLATA (PI) (YANOFSKY et al., 1990;COEN;PELAZ et al., 2000). Furthermore, AG also acts as a meristem gene by suppressing the class A gene APETALA1 (AP1) in the center of the flower (GUSTAFSON-BROWN et al., 1994) and is itself suppressed by the class B gene APETALA2 (AP2) in the sepal and petal domains (DREWS et al., 1991;YANT et al., 2010). Besides their complicated interactions with each other, some regulator genes, such as AG and AP1 are reported to mediate in hormone pathways to control floral organ development (Gómez- Mena et al., 2005;ITO et al., 2007;KAUFMANN et al., 2010).
Interestingly, in grass floral morphogenesis the interaction of JA and transcription factors seems not to be limited to stamen development (KIM et al., 2009;CAI et al., 2014). The floral meristem gene OsMADS1, working together with the rice EG1 gene, orthologous to DGL, controls floral meristem determinacy and impacts floral organ identity and number via a JA biosynthetic pathway (LI et al., 2009;CAI et al., 2014). Thus, Cai et al (2014) imply that the role of JA in reproduction has diversified during flowering plant evolution. Because of the different floral features between monocots and dicots and the comparatively few studies in monocotyledonous species (BORTIRI;HAKE, 2007;TANAKA et al., 2013), more work is needed to explain the roles of JA in grass flower morphogenesis.
JA has been proved to mainly affect the pistil abortion process in maize tassel spikelets (ACOSTA et al., 2009;YAN et al., 2012). The TASSELSEED1 (TS1) gene encodes a JA synthesis enzyme (ACOSTA et al., 2009). Maize ts1 mutants exhibit tasselseed phenotypes, in which pistils of tassel spikelets are not aborted as they normally would be. Compared to the well-documented interaction of JA and transcription factors in Arabidopsis floral organ morphogenesis, their relationship in maize still remains obscure, partly because the functions of most floral regulator genes in maize still await verification (CIAFFI et al., 2011;LI et al., 2014) To gain some insights into the relationship between JAs and some transcription factors in maize tassel development, we applied exogenous MeJA to the maize tassel at an early stage. By phenotypic observation and expression analysis, we found that application of MeJA altered the number of palea and stamens. We also identified some key transcription factors that can be influenced by MeJA. Our results provide clues for further understanding these genes' functions in tassel spikelet morphogenesis.

Plant Materials and MeJA Treatment
Plants of the maize cultivar B73 were grown under natural conditions in an experimental field. For the MeJA treatment experiments, we treated tassels of six-leaf stage plants with a solution containing 0.2 mM MeJA (Sigma-Aldrich, MeJA diluted in 1% alcohol solution), 2 mM MeJA, or with 1% alcohol solution as a control. Using plastic pipettes, 2 mL solutions were directly applied onto the tassel inflorescences every 2~3 days until the tassel spikelets had reached maturity.

Phenotypic Characterization and Microscopy
Inflorescence meristems of control plants (treated with 1% alcohol solution) and MeJA-treated plants (<2 nm long) were fixed in 2% glutaraldehyde in phosphate buffer overnight at 4°C, then dehydrated in an ethanol series and subjected to critical point drying. Mounted samples were sputter coated with gold and viewed with an FEI 1474 The effect of meja on zea… TAN, J. et al.

QUANTA600
(Holland) scanning electron microscope. To perform phenotypic analysis and stereoscopic microscope observation, mature tassel spikelets were dissected and observed under a SteReo Lumar V12 microscope (Zeiss, Germany).

Gene Expression Analysis
For promoter analysis, 2500bp of upstream sequences of some regulator genes were analyzed with PlantCARE database. For quantitative realtime PCR experiments, mature tassels and ears were collected from at least four independent plants. Total RNA extraction and cDNA preparation were performed according to the previous report (ZHENG et al., 2010). Primers were designed with Beacon Designer 7 software and listed in Supplementary Table 1. A master mix of qPCR reaction components was prepared according to the manufacturer's protocol for SYBR Green RealMasterMix (Promega, USA). Thermocycling and fluorescence detection were performed using a Roche LightCycle480 Real-Time PCR machine (Roche, Switzerland). Each experiment was repeated three times. Relative fold changes in gene expression were calculated using the comparative 2 -∆∆Ct method with GAPDH (glyceraldehyde-3phosphate dehydrogenase) as the endogenous control.

MeJA Treatment Influences the Development of Tassel
To investigate the effects of MeJA on the development of the male inflorescence, we observed the early developmental stage of wild-type and MeJA-treated tassels by scanning electron microscopy. At an early developmental stage, wildtype tassel had a main inflorescence with the spikelet pair starting to initiate and branches differentiating at the base of the main inflorescence ( Figure 1A). At this stage, the 2 mM MeJA-treated tassel apparently developed slowly and seemed to just start its branch differentiation ( Figure 1B). When the wild-type spikelets were approaching a typical bisexual stage, containing a lower floret (red arrow) and an upper floret as mentioned above ( Figure 1C), the MeJA-treated branches lagged far behind those of wild-type and were just at the glume primordia differentiation stage ( Figure 1D, star). Based on these observations, we concluded that the application of exogenous MeJA decelerates the developmental pace of the tassel. Since 0.2 mM MeJA treatment had a similar effect on the development of tassel as well, though to a lesser extent(data not shown), we used 2 mM MeJAtreated samples for further research in this study.

MeJA Treatment Influences the Development of Tassel Floral Organs
To further examine the effect of exogenous MeJA on the development of tassel spikelets, we dissected wild-type and MeJA-treated spikelets and observed their phenotypic features. A mature wildtype tassel spikelet contains two florets subtended by two glumes. Each floret consists of a lemma, a palea, two lodicules and three stamens (Figure 2A).
We found some mature spikelets in MeJA-treated tassels only contained one floret and three stamens ( Figure 2B). The size of the spikelet also tended to be smaller than that of wild-type. By analysis of at least 100 spikelets, we found that approximately 8% of the MeJA-treated spikelets changed the number of palea and stamens (Table 1). Taken together, our results demonstrated that the application of MeJA could affect floral organ number of tassel spikelets.

Expression Analysis of Some Regulator Genes
To get a clue about genes that may be related to the floral phenotypic changes triggered by exogenous MeJA, we selected some key regulator genes related to floral development (Table 2). Next, we analyzed the promoter sequences of these selected genes, paying attention to modules that are related to hormones. Among these eight genes, ZAP1, SI1, ZAG1, ZMM2, and TS1 had MeJA responsive elements, which may imply they react directly to MeJA application. All of the eight genes, except ZMM2, have GA responsive elements. Other hormone-responsive elements also were found in the promoter regions of these genes. The correlation of these hormone elements and the function of these genes needs further experimental test.
Since some selected genes still await functional verification (CIAFFI et al., 2011;LI et al., 2014; Table 2), we tested their expression patterns in tassel and ear by quantitative real-time PCR. Among eight genes examined in our study, class B gene SI1 was expressed slightly more in tassel than in ear, and putative class A gene ZAP1, class C gene ZAG1, meristem genes ZAG3 and IDS1, were expressed mainly in ear. The other three genes, putative class C genes ZMM2 and ZMM23, and sex determination gene TS1 displayed no biased expression pattern between tassels and ears ( Figure  3A, 3C). The expression level of all examined genes, except ZMM2, were increased by application of MeJA ( Figure 3B, 3D). Taking these genes` promoter analysis into consideration, we speculated that ZAP1, SI1, ZAG1 and TS1 may directly respond to exogenous MeJA, while other genes that showed increased expression level in MeJA-treated tassels may respond to MeJA via other hormones. 1476 The effect of meja on zea… TAN, J. et al.

DISCUSSION
In this study, we showed that exogenous application of MeJA to maize impacts its tassel development. The effects included a reduced developmental pace of spikelets and alterations in floral organ numbers. Exogenous MeJA also increased the expression level of key developmental floral organ identity genes, meristem genes and tassel sex-determination genes, suggesting that MeJA has a role in maize floral organ development, at least in the development of stamen and palea.

Class C gene function may divergent in maize inflorescence development
Maize contains three genes that are homologous to AG1, ZAG1 and ZMM2/ZMM23 (ZAHN et al., 2006). ZAG1 is mainly expressed in carpel, and ZMM2 is mainly expressed in anthers (SCHMIDT et al., 1993;MENA et al. 1996). Though mutants of ZMM2/ZMM23 have not yet been identified, it is thought that ZMM2/ZMM23 probably specify the organ identity of stamens and ZAG1 specifies floral meristem determination (SCHMIDT et al., 1993;THEISSEN et al., 1995;MENA et al., 1996;LI et al., 2014). Consistent with this hypothesis, our results also showed that ZAG1 is expressed mainly in the ear, and ZMM2 and ZMM23 are expressed slightly higher in the tassel than in ears ( Figure 3A). A noticeable expression difference also existed between ZMM2 and ZMM23. ZMM23, containing putative GA responsive elements rather than MeJA responsive elements in its promotor, showed a MeJA responsive expression pattern, while ZMM2 did not. Our results may provide another clue about their divergent role in maize inflorescence development mediated by different hormone pathways.

JA and regulator genes in maize tassel development
JAs have been proved to be involved in stamen development and usually result in male sterility. In our study, application of 2 mM MeJA did not result in male sterility (data not shown). However, alterations in the number of stamens and palea in MeJA-treated tassels were phenotypically similar to those in MeJA-treated rice plants, which showed an alteration in spikelet organ numbers (KIM et al., 2009). Our study proved that exogenous MeJA has a similar influence in maize floral organ development, at least in stamen and palea development.
indeterminate spikelet1 (ids1) is an APETALA2 (AP2)-like gene which is required for the timely conversion of the spikelet meristem into the floral meristem. IDS1 has also been reported to be a target gene of the TASSELSEED4 gene. Mutation in the IDS1 gene will result in a tasselseed phenotype which is similar to the JA-deficient mutant ts1 (CHUCK et al., 2007(CHUCK et al., , 2008. In our study, the expression level of IDS1 and TS1 both increased in MeJA-treated tassels. It will be worthwhile to study whether there exists an interaction between the ts1 pathway and the ts4 pathway during the tassel sex determination process. By qPCR, we found 7 regulator genes displayed MeJA-responsive expression patterns, although their roles in MeJA-treated tassel undoubtedly need more detailed study. Among these genes, ZAP1, ZAG1 ZAG3 and IDS1 were mainly expressed in the ear. Considering that all of these four genes contain GA responsive elements in their promoters and GA is reported to interact with JA in Arabidopsis stamen development (PENG, 2009), it will be interesting to explore whether the crosstalk between JA and GA exists in maize tassel development.