Distinct organ-specific and temporal expression profiles of auxin-related genes during mango fruitlet drop

Plant Physiology and Biochemistry 115 (2017) 439e448

Contents lists available at ScienceDirect

Plant Physiology and Biochemistry journal homepage: www.elsevier.com/locate/plaphy

Research article

Distinct organ-specific and temporal expression profiles of auxinrelated genes during mango fruitlet drop Youlia Denisov a, 1, Shani Glick a, b, 1, Tali Zviran a, Mazal Ish-Shalom a, Adolfo Levin c, Adi Faigenboim a, Yuval Cohen a, Vered Irihimovitch a, * a b c

Institute of Plant Sciences, The Volcani Center, Agricultural Research Organization, Bet-Dagan 50250, Israel Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat Gan, Israel Migal e Galilee Technology Center, P.O. Box 831, Kiryat Shemona 11016, Israel

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 December 2016 Received in revised form 26 March 2017 Accepted 23 April 2017 Available online 24 April 2017

In mango, fruitlet abscission initiates with a decrease in polar auxin transport through the abscission zone (AZ), triggered by ethylene. To explore the molecular components affecting this process, we initially conducted experiments with developing fruitlet explants in which fruitlet drop was induced by ethephon, and monitored the expression patterns of distinct indole-3-acetic acid (IAA)-related genes, comparing control vs. ethephon-treated pericarp and AZ profiles. Over the examined time period (48 h), the accumulation of MiPIN1 and MiLAX2 IAA-efflux and influx genes decreased in both control and treated tissues. Nevertheless, ethephon-treated tissues displayed significantly lower levels of these transcripts within 18e24 h. An opposite pattern was observed for MiLAX3, which overall exhibited upregulation in treated fruitlet tissues. Ethephon treatment also induced an early and pronounced down-regulation of five out of six IAA-responsive genes, and a substantial reduction in the accumulation of two IAA-synthesis related transcripts, contrasting with significant up-regulation of Gretchen Hagen3 transcript (MiGH3.1) encoding an IAAeamino synthetase. Furthermore, for both control and treated AZ, the decrease in IAA-carrier transcripts was associated with a decrease in IAA content and an increase in IAAeAsp:IAA ratio, suggesting that fruitlet drop is accompanied by formation of this non-hydrolyzed IAA eamino acid conjugate. Despite these similarities, ethephon-treated AZ displayed a sharper decrease in IAA content and higher IAAeAsp:IAA ratio within 18 h. Lastly, the response of IAA-related genes to exogenous IAA treatment was also examined. Our results are discussed, highlighting the roles that distinct IAA-related genes might assume during mango fruitlet drop. © 2017 Elsevier Masson SAS. All rights reserved.

Keywords: Abscission Auxineamino acid conjugate Auxin-related genes Fruitlet Mango

1. Introduction In plants, abscission of reproductive structures has developed to facilitate the shedding of excess, damaged, or no longer necessary organs. In particular, fruitlet abscission occurring during fruit development is characterized by activation of a pre-differentiated abscission zone (AZ), located between the pedicel and fruitlet (Bonghi and Ramina, 2000; Sawicki et al., 2015). The main hormones controlling AZ formation are ethylene, acting as an inducer, and auxin (indole-3-acetic acid, IAA), acting as a suppressor. As such, diffusion of ethylene from fruit tissues to the AZ is suggested

* Corresponding author. E-mail address: [email protected] (V. Irihimovitch). 1 Equally contributed. http://dx.doi.org/10.1016/j.plaphy.2017.04.021 0981-9428/© 2017 Elsevier Masson SAS. All rights reserved.

to play a role in initiating abscission events by accelerating the activities of cell-wall-degrading enzymes at the AZ (Bonghi and Ramina, 2000; Sawicki et al., 2015). On the other hand, basipetal IAA flux though the AZ is thought to inhibit abscission by reducing the sensitivity of the AZ area to ethylene (Estornell et al., 2013; Sawicki et al., 2015). Auxin plays an important role in controlling many aspects of plant development (Vanneste and Firml, 2009). In particular, the balance of IAA synthesis, conjugation, degradation and transport is tightly regulated, leading to establishment of IAA homeostasis (Woodward and Bartel, 2005; Zhao, 2010). In principle, de-novo IAA synthesis, resulting from tryptophan (Trp)-dependent or independent pathways, is followed by its conjugation to amino acids, sugars and methylesters. Up to date, different pathways have been proposed to be involved in the production of IAA in plants. However, it has only recently been established that a simple two-step pathway,

440

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

which includes the sequential action of two groups of enzymes TAA1/TARs (TRYPTOPHAN AMINOTRANSFERASE of ARABIDOPSIS1/ TRYPTOPHAN AMINOTRANSFERASE RELATED) and YUCCAs (flavin monooxygenases), acts as the main route contributing to the synthesis of free IAA (see reviewed in (Korasick et al., 2013; LudwigMüller and 2011; Zhao, 2010)). Following biosynthesis and conjugation, release of IAA from its conjugates is achieved by hydrolytic cleavage (Korasick et al., 2013; Ludwig-Müller, 2011). Different IAA conjugates have been identified. Of these, IAAealanine (Ala), IAAeleucine (Leu) and IAAephenylalanine (Phe), among others, have been shown to hydrolyze back to free IAA. In contrast, IAAeaspartate (Asp) and IAAeglutamate (Glu) are thought to serve as precursors for IAA degradation, as there is no evidence indicating that they can be hydrolyzed back to IAA (Ludwig-Müller, 2011). The transport of free IAA from sites of synthesis to target cells is further facilitated by influx and efflux carriers. Specifically, IAA efflux is achieved by active transport mediated by members of the PINFORMED family of efflux carriers (PINs) (Petrasek and Friml, 2009), whereas IAA import is facilitated by AUXIN RESISTANT 1/LIKE AUXIN RESISTANT (AUX1/LAX) family members, (Peret et al., 2012). After reaching the target tissues, IAA perception and signaling seems to depend on two types of receptors: ABP1 (AUXIN-BINDING PROTEIN 1) and TIR1/AFB (TRANSPORT INHIBITOR RESISTANCE 1/ AUXIN SIGNALING F-BOX-TYPE) (Chapman and Estelle, 2009). While the role of ABP1 in mediating auxin responses is still controversial (Gao et al., 2015), the signaling pathways involving TIR1/AFBs have been well characterized (Chapman and Estelle, 2009; Weijers and Wagner, 2016). Briefly, three families of early auxin-responsive genes, including Aux/IAAs encoding nuclear transcriptional repressors, Gretchen Hagen3s (GH3s) encoding IAAconjugating enzymes, and small auxin-upregulated proteins (SAURs) encoding short-lived proteins that likely function in cell expansion, all contain one or more binding motifs to auxin response factor (ARF) transcription factors in their promoter regions. Whereas at low IAA concentrations, a heterodimer of an ARF and an Aux/IAA represses transcription, at higher IAA concentration, IAA binds to TIR1/AFB, triggering the degradation of Aux/IAAs, releasing ARF transcription factors and leading to activation of the early IAA-response genes (Chapman and Estelle, 2009). It should be noted, however, that ARFs may also function as transcription factors that mediate repression of gene expression, depending on the amino acid composition of their middle region (Weijers and Wagner, 2016). Over the last decade, knowledge has been gained on the functions of distinct IAA-related genes during reproductive organ abscission (Estornell et al., 2013; Sawicki et al., 2015). Moreover, recent molecular studies performed during flower abscission in tomato (Meir et al., 2010) and during fruitlet drop in apple (Botton et al., 2011; Dal Cin et al., 2009), citrus (Cheng et al., 2015), litchi (Li et al., 2015) and grapevine (Kühn et al., 2016) have shown that acquisition of ethylene sensitivity in the AZ is associated with, among other factors, disruption of IAA homeostasis and altered expression of distinct IAA-related genes. Interestingly, some of these studies reported that similar but not identical changes in expression patterns of IAA-related genes occur in both reproductive organs that are about to abscise and in their AZ (Kuang et al., 2012; Meir et al., 2010, 2015). Mango (Mangifera indica L.) is a very important tropical fruit crop. Despite satisfactory flowering and fruit set, mango production does not meet its potential due to intense natural fruitlet drop, leading to losses in revenue (Singh et al., 2005). Abscission events take place in mango during three distinct fruit-development stages: the first 2 months after fruit set, the mid-season (when fruit are 60e75 days old), and just before fruit maturity. Nevertheless, the highest abscission rates are observed during the early

stages of fruit development, followed by gradually less, albeit still intense, fruitlet drop rates as the fruit reach maturity (Nunez-Elisea and Davenport, 1986; Singh et al., 2005). Interestingly, mango studies have shown that natural fruitlet drop correlates with enhanced endogenous ethylene production in both seed and pericarp tissues, however, pedicels containing the AZ produced no detectable ethylene prior to or at the moment of abscission (NunezElisea and Davenport, 1986). To gain preliminary insight into the molecular mechanism regulating mango fruitlet drop, we previously cloned a mango gene encoding an ERS1-type ethylene receptor and its expression was monitored during fruitlet drop and fruit ripening (Ish-Shalom et al., 2011). Our data highlighted MiERS1's function in regulating fruitlet abscission (Ish-Shalom et al., 2011), a conclusion that was also recently corroborated in a study by Hagemann et al. (2015). Notably, in the latter study, it was also determined that mango fruitlet drop is associated with a reduction in polar auxin transport capacity through the fruitlet pedicel, although the biological events controlling this process were not elucidated (Hagemann et al., 2015). Here, to determine whether mango fruitlet drop is accompanied by specific deactivation or activation of distinct IAA-related genes, experiments were first conducted with developing fruitlet explants in which fruitlet abscission was induced by the ethylene-releasing compound ethephon. The spatial and temporal expression patterns of selected IAA-related genes were investigated by comparing their expression profiles in control vs. ethephon-treated fruitlet tissues. The study was next expanded to quantify free IAA and IAA conjugates in control and ethephon-treated AZ tissues and to explore the response of IAA-related genes to exogenous treatment with synthetic IAA. Our results highlight the potential roles of distinct IAA-related genes in decreasing IAA levels and increasing the IAAeAsp:IAA ratio in the AZ, thus affecting mango fruitlet drop. 2. Materials and methods 2.1. Plant material, induction of fruitlet abscission by ethephon, and auxin treatment Mango (Mangifera indica L.) explants were collected in the spring from commercially bearing 'Kent' trees grown in Ramot, in northeast Israel, on 12 May 2014, and on 07 May, 2016. The explants, each bearing 1 to 2 fruitlets per panicle (approx. 20 cm long), were kept in water and brought to the laboratory within 2.5 h of collection. In the laboratory, the explants were divided into six (May, 2014) or nine (May, 2016) experimental units, each composed of 110e120 explants. The basal end of each explant was placed in a 50-mL tube containing water and kept at 25
C. On May, 2014, fruitlet abscission was induced in three experimental units using ethephon (Ethrel, 1.4 g L1, Agan Chemicals, Ashdod, Israel), as previously described (Ish-Shalom et al., 2011). The remaining untreated experimental units served as controls. On May 2016, fruitlet abscission was induced in three experimental units using ethephon, three experimental units were sprayed with 2,4dichlorophenoxyacetic acid (2,4-D) (Hadranol™, 0.2 g L1, Machteshim, Beer Sheva, Israel), and the remaining experimental units served as controls. A non-ionic surfactant, Triton X-100, was included in all sprays at 0.025%. In both occasions, samples from the AZ and fruitlet pericarp of treated and control experimental units were collected at different time points, frozen in liquid nitrogen and kept at 80
C until further analysis. 2.2. RNA isolation and cDNA synthesis Plant tissues were ground in liquid nitrogen using an IKA-A11 analytical grinding mill (IKA®-Werke GmbH & Co. KG, Staufen,

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

Germany). Total RNA was extracted from 2 g of ground frozen plant material using the phenoleSDS method, as previously described (Ish-Shalom et al., 2011). RNA integrity was evaluated on a 1.2% formaldehydeeagarose gel. Following confirmation of RNA integrity, 5 mg of total RNA, pretreated with 1 unit of RQ1 DNase, was used as the template in the synthesis of first-strand cDNA with an anchored oligo-dT primer and SuperScript® III Reverse Transcriptase (Thermo Scientific, MA, USA) according to the manufacturer's instructions. The reaction products were then used for geneexpression analyses. 2.3. Selection of genes for expression analyses Distinct expressed sequence tags (ESTs), representing either full-length or partial sequences putatively encoding different IAArelated genes, were retrieved from a 'Keitt' mango transcriptome (Sherman et al., 2015). For each selected gene, specific primers were designed using the web-based Primique tool (http://cgi-www. daimi.au.dk), which allows automatic design of specific primers for a single sequence within a gene family (Table S1). Prior to the expression analyses, specificity of each pair of primers was validated by PCR using cDNA synthesized from collected tissues (a mixture of samples collected at different time points). The obtained PCR products were sequenced at Hy-labs Laboratories (Rehovot, Israel), ligated into the pGEM-T Easy vector and used as a template to create the corresponding standard curves for RT-PCR analysis (see below). 2.4. Quantitative (q) real-time-PCR RNA isolated from pericarp and AZ tissues of the control and ethephon or auxin-treated experimental units, taken at various time intervals, were used for analyses. Transcript accumulation of genes of interest was estimated by qRT-PCR, using Power SYBR® Green PCR Master Mix (Applied Biosystems, Warrington, UK). Reactions were carried out using 3 mL of cDNA products (1:10 dilution), 6 mL SYBR Green PCR Master mix and 200 nM primers from the relevant primer pair, in a final volume of 12 mL. Analyses were conducted in a Rotor GENE 6000 instrument (Corbett Life Science, Sydney, Australia). cDNA samples were analyzed in triplicate with each reaction being subjected to melting point analysis to confirm the presence of single amplified products. To estimate the efficiently of the reactions, a decimal dilution series of pGEM-T Easy vector containing cDNA-amplified PCR fragments of the relevant genes, were generated, and standard curves for each gene were established using pairs of specific primers. The obtained data were analyzed with manually set threshold of 0.02 and target Ct range of 15e35. Relative gene expression was calculated using the DDCt method (Livak and Schmittgen, 2001), where averaged crossing points were normalized to the endogenous reference gene MiEF1a (mango_rep_c22857) (detector normalization), and to geneexpression level at time 0 (sample normalization). The equations used were: DCt ¼ target RNA Ct value - reference RNA Ct value; DDCt ¼ selected time point DCt - time point zero DCt (Livak and Schmittgen, 2001). Following 2DDCt calculation, values were log2 transformed and mean values and standard errors were calculated. 2.5. Quantification of IAA and IAA conjugates Samples of AZ tissue, collected at the initial of the experimental set up on May 2014, together with ethephon-treated AZ tissue collected at 4, 18, 24 and 48 h after treatment, and control AZ tissue collected 24 and 48 h after treatment, were lyophilized for 24 h and subjected to hormonal analysis. IAA and IAAeconjugate analyses were performed at the Plant Biotechnology Institute of the National

441

Research Council of Canada (http://www.pbi.nrc.gc.ca/ENGLISH/ technology-platforms/plant-hormone-profiling.htm). For extraction and purification, an aliquot (100 mL) containing deuterated internal IAA and IAAeconjugate standards, each at a concentration of 0.2 pg mL1, was added to freeze-dried homogenized AZ samples (approximately 50 mg). Isopropanol:water:glacial acetic acid (80:19:1, v/v; 3 mL) was then added, and the samples were agitated in the dark for 24 h at 4
C. Samples were then centrifuged - and the supernatant was isolated and dried on a Büchi Syncore Polyvap (Büchi, Flawil, Switzerland). Samples were reconstituted in 100 mL acidified methanol, adjusted to 1 mL with acidified water, and then partitioned against 2 mL hexane. After 30 min, the aqueous layer was isolated and dried as above. Dry samples were reconstituted in 800 mL acidified methanol and adjusted to 1 mL with acidified water. The reconstituted samples were passed through equilibrated Sep-PaK C18 cartridges (Waters, Mississauga, ON, Canada), and eluted with acetonitrile:water:glacial acetic acid (30:69:1, v/v). The eluate was then dried on a centrivap concentrator (Labconco Corporation, Kansas City, MO, USA). The samples were next subjected to UPLC/ESIeMS/MS (ultra-performance liquid chromatography electrospray tandem mass spectrometry) analysis and quantified as described by Lulsdorf et al. (2013) 2.6. Statistical analyses Statistical analyses were performed using JMP software, version 10 (SAS Institute, Cary, NC, USA). The analyses of the real-time qPCR data were performed by least significant difference (LSD) test, according to pairwise comparison by Student's t-test, with P  0.05. At each time point, log2-fold changes in gene expression values of treated ethephon samples, or treated auxin samples, were compared to their respective controls. Statistical analyses of IAA, IAAeconjugates and the calculated IAAeAsp:IAA ratios were performed using one-way analysis of variance (ANOVA) by TukeyeKramer multiple comparison test, with P  0.05. 3. Results 3.1. Monitoring the effect of ethephon treatment on fruitlet abscission rate As mentioned above, in mango, fruit drop occurs throughout the developmental period (Singh et al., 2005). In particular, field records under our conditions indicate that drop of persisting 'Kent' fruitlets, monitored from the end of April through the end of July, reaches its highest level (approximately 70% of total fruitlets), by the end of May (See Fig. S1). With this observation in mind, to explore molecular components affecting mango fruitlet abscission, experiments were conducted with 'Kent' fruitlet explants that were collected at the onset of May. In these experiments, fruitlet abscission was artificially induced using ethephon (Ethrel) and fruitlet abscission levels were recorded. The abscission records obtained indicated that ethephon treatment induced intense fruitlet abscission, which initiated within 24 h of treatment, and reached 94% 72 h post-treatment (Fig. 1). Abscission of the entire ethephon-treated fruitlet population was observed at 120 h posttreatment. In contrast, under control conditions, fruitlet drop reached only 20% at 120 h after treatment. From this time point, the rate of fruitlet drop also increased under control conditions but only reached 40% by the end of the observation period (168 h). 3.2. Sequence analysis of distinct auxin-related mango genes selected for expression analyses For expression analyses, fourteen ESTs putatively encoding

442

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

acids, and group III contains proteins only found in Arabidopsis, Brassica and Gossypium (Staswick et al., 2005; Terol et al., 2006). Here, using phylogenetic analysis, we found that MiGH3.1, MiGH3.3 and MiGH3.6 were all clustered with the distinct group II GH3 proteins (Fig. S2). Four ESTs including an EST sequence encoding a full-length TAA1/TAR-type protein, and three ESTs putatively encoding distinct YUCCA genes, were also chosen for our expression analyses. In all cases, typical anticipated domains characterizing each protein group's function were found in their predicted translated sequences (see Table S2). 3.3. Expression profiles of auxin-related genes in control vs. ethephon-treated fruitlet tissues

Fig. 1. Mango fruitlet abscission. (A) Changes in control and ethephon-treatment abscission rates. Fruitlet abscission was induced with ethephon using three experimental units of explants bearing fruitlets, each composed of 110e120 explants. Three untreated experimental units served as controls. The experiment was conducted in May 2014. The number of intact fruitlets in each experimental unit was documented at several time points and abscission rates were calculated. Each value represents the mean of three biological replicates with the standard error (SE) indicated by vertical bars. (B) Mango 'Kent' fruitlets before (left) and after (right) abscission. S e seed, P e pericarp, AZ e abscission zone. The dotted lines indicate the pedicel AZ and P tissues used for sampling.

distinct auxin-related genes were chosen from a ‘Keitt’ transcriptome (Sherman et al., 2015). Four ESTs encoding mango auxinefflux and influx carriers (named: MiPIN1, MiPIN3, MiLAX2 and MiLAX3), were selected for our study. As reported for other IAAefflux and influx proteins, hydropathic analysis confirmed the presence of transmembrane domains, indicating subcellular membrane localization of the selected mango IAA-carriers (see Table S2). It should be noted that in Arabidopsis, eight members of the PIN family, divided into “long” and “short” PIN subgroups, have been characterized, of which only “long” PINs are suggested to be involved in polar transport of IAA (Petrasek and Friml, 2009). In our case, based on sequence analysis, we concluded that both MiPIN1 and MiPIN3 are “long” PIN-type IAA-efflux proteins (see Table S2). Furthermore, six early IAA-responsive genes, three encoding distinct mango AUX/IAAs and three encoding different GH3 proteins, were also selected for expression studies. Based on best matches with different AUX/IAAs and GH3 genes from Arabidopsis, generic nomenclature was assigned to each coding sequence (see Table S2). Analysis of the predicted complete or partial MiAUX/IAA and MiGH3 sequences revealed that all selected sequences contain the typical domains that are required for their function (Tiwari et al., 2004; Staswick et al., 2005). It should be noted that based on sequences, GH3 proteins can be divided into three groups. Specifically, group I contains GH3-11-like (JAR1) proteins that have been shown to adenylate jasmonic acid, group II contains IAAeamido synthetases, shown to conjugate IAA to various amino

To understand the possible roles of the selected auxin-related genes during mango fruitlet drop, qRT-PCR analyses were carried out. Total RNA extracted from control and ethephon-treated fruitlet tissues, collected at different time points up to 48 h post-treatment, was analyzed using specific primers designed to amplify the different gene products. The expression patterns of genes encoding IAA carriers were examined first. Fig. 2 shows that in the pericarp, the expression of MiPIN1 and MiLAX2 was moderately downregulated within 24e48 h under control conditions; however, following ethephon treatment, their levels were already significantly lower within 18e24 h (Fig. 2, left panels). Furthermore, in the same tissue, MiPIN3 expression, which initially increased under both control and treatment conditions, was also slightly downregulated within 48 h of treatment. An exception of this ethephon-induced decrease pattern was observed for MiLAX3, which exhibited an overall up-regulation pattern following treatment. Furthermore, in the parallel AZ analysis, similar differences in MiPIN1, MiLAX2 and MiLAX3 mRNA accumulation were observed between control and treated tissues, while MiPIN3 expression profiles did not differ significantly between control and treatment conditions (Fig. 2, right panels). Decreasing accumulation of specific IAA-carrier transcripts during induction of fruitlet drop might hamper or decrease IAA flux from the developing fruitlet into the AZ, or in the AZ itself. Since this situation is expected to impact the accumulation of distinct early IAA responsive genes (Chapman and Estelle, 2009), we further examined the influence of ethephon treatment on the expression patterns of selected MiAUX/IAAs and MiGH3 genes. Under control conditions, in both the pericarp and AZ, the levels of the three examined MiAUX/IAA genes remained relatively stable or only moderately decreased within 48 h (Fig. 3). In contrast, in ethephontreated tissues, the levels of these genes decreased markedly within 18e24 h and remained significantly low until the end of the experimental time period (Fig. 3, left panels). With respect to MiGH3 expression profiles in the pericarp, two of the examined genes displayed a rather stable or decreased pattern under both control and treatment conditions (MiGH3.3 and MiGH3.6, respectively), although the levels of both were significantly lower within 4e24 h of treatment. More striking, however, was the expression pattern observed for MiGH3.1, which overall, exhibited enhanced transcript accumulation from 4 h of treatment. Interestingly, treated AZ also displayed up-regulation of the MiGH3.1 transcript, whereas under control conditions, MiGH3.1 levels first slightly decreased and then increased back to initial levels within 24 h. Lastly, the levels of the other two examined GH3s did not differ between control and treated AZ (see Fig. 3, right panels). Disruption of IAA homeostasis might be caused by various reasons, including an interference with normal de-novo IAA synthesis (Korasick et al., 2013; Ludwig-Müller, 2011; Zhao, 2010). Fig. 4 summarizes our expression analysis results, in which the

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

443

Fig. 2. Expression analysis of IAA-carrier genes. qRT-PCR was performed with cDNA prepared from total RNA extracted from control and ethephon-treated tissues collected at different time points. Gene expression was calculated using the DDCt method, with averaged crossing points normalized to the endogenous reference gene MiEF1a (detector normalization), and to gene-expression level at time 0 (sample normalization). Log2-transformed data for a given gene are expressed as fold change in expression relative to its expression at time 0. Data are means ± SE of three independent biological replicates. Asterisks denote a significant difference between ethephon-treated and control tissue samples at each sampling point (P < 0.05). P e pericarp tissue, AZ e abscission zone. Sampling intervals are presented as hours from initial setting.

profiles of MiTAA1/TARs and MiYUCCA IAA-synthesis related genes were monitored. Examination of their profiles revealed that in the pericarp, under control conditions, the four examined genes exhibited an initial increase in their accumulation levels, followed by decreased expression. Notably, the ethephon-treated pericarp tissues exhibited analogous expression patterns. Nevertheless, MiYUCCA2 levels decreased sharply within 4 h of treatment and remained low, while MiTAR2, MiYUCCA4 and MiYUCCA10 levels became significantly lower than in controls within 24e48 h. The parallel AZ results showed that under control conditions, MiTAR2 transcript initially increased in levels, then slightly decreased within 24 h, and subsequently increased again. By comparison, throughout the examination period, ethephon-treated AZ

displayed a decrease in MiTAR2 transcription, which become significant within 24 h. Furthermore, both under control and treatment conditions, the AZ displayed a decline in MiYUCCA10 transcription. However, in this case, the levels of MiYUCCA10 did not differ significantly between control and treated tissues. Finally, since the presence of MiYUCCA2 and MiYUCCA4 in the AZ could not be reliably detected in the qRT-PCR analysis, due to very low transcript abundance in both control and treated tissue, the relevant AZ data were omitted from this study. 3.4. Quantification of auxin and auxin conjugates To expand the results of

the expression

studies, the

444

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

Fig. 3. Expression analysis of AUX/IAA and GH3 genes. qRT-PCR was performed with cDNA prepared from total RNA extracted from control and ethephon-treated tissues collected at different time points. Gene expression was calculated using the DDCt method as described in the legend to Fig. 2. Data are means ± SE of three independent biological replicates. Asterisks denote a significant difference between ethephon-treated and control tissue samples at each sampling point (P < 0.05). P e pericarp tissue, AZ e abscission zone. Sampling intervals are presented as hours from initial setting.

concentrations of free IAA and specific IAA conjugates were further analyzed in control and ethephon-treated AZ tissues, which served to monitor gene expression. Fig. 5A shows that during the experimental period, IAA content decreased as compared to the initial time point (time 0) in both control and ethephon-treated AZ samples. However, whereas under control conditions, IAA content only decreased by 50%, in ethephon-treated tissue, the amount of free IAA decreased sharply within 4e18 h, reaching a null value within 24 h. Moreover, Fig. 5A also shows that in both control and ethephon-treated samples, IAA conjugates were only represented by the non-hydrolyzed IAA amino conjugate IAAeAsp, as IAAeAla, IAAeGlu and IAAeLeu signals were below the limit of detection (not shown). Surprisingly however, whereas under control conditions, concomitant with the attenuated decrease in free IAA levels, IAAeAsp content increased significantly, in ethephon-treated tissue, the sharp decrease in IAA levels that was observed within 18 h was only accompanied by a slight, non-significant increase in IAAeAsp content. From that time point on, concurrent with the complete decay of IAA, IAAeAsp conjugates were not detected in the treated samples. On the other hand, a parallel examination of IAAeAsp:IAA ratios indicated that this parameter reached its maximum value (0.57) within 18 h of treatment. Under control conditions, 30 additional hours were required for this ratio to reach a value of 0.5 (Fig. 5B). 3.5. Testing the reproducibility of the obtained expression gene results and the response of selected genes to auxin treatment The above data indicated that the induction of mango fruitlet abscission is associated with changes in the accumulated levels of distinct IAA-related genes in both pericarp and AZ tissues. To provide additional evidence supporting this conclusion, and to test the effect of auxin treatments on the expression patterns of the

examined genes, we further analyzed the expression levels of the selected genes using plant material which was collected from an independent experiment, carried out in a subsequent season. In this experiment, 'Kent' fruitlet explants were treated either with ethephon or with 2,4-D synthetic auxin, while a remaining group of untreated explants served as controls (see Materials and methods). In this case, ethephon-mediated induction of fruitlet abscission also initiated within 24 h, reaching an 100% abscission rate within 96 h. By comparison, under control conditions, fruitlet drop reached only 20% at the end of the experimental period (168 h), while 2,4-D treatment completely inhibited fruitlet drop (not shown). Following treatments, the RNA extracted from control and treated tissues at time 0 and 24 h later was subjected to qRT-PCR analysis. In line with our results from the previous season, the levels of MiPIN1, MiLAX2, MiAUX/IAA4, MiAUX/IAA16 and MiGH3.3 transcripts in the ethephon-treated pericarp tissue dropped significantly, relative to control samples, whereas significant increments in MiLAX3 and MiGH3.1 transcript levels were observed in this tissue following ethephon spraying (see Fig. S3). In the same tissue, 2,4-D treatment induced an increment in MiPIN1 transcript levels, prevented the decrease in MiLAX2 mRNA levels, and had no effect on MiLAX3 expression. Moreover, and as expected, 2,4-D treatment also induced the accumulation of four examined early auxinresponding genes, including MiAUX/IAA4, MiAUX/IAA16, MiGH3.1 and MiGH3.3. Similar expression patterns were also obtained in the AZ. However, in this case, MiLAX3 levels did not significantly differ from control following ethephon treatment, whereas 2,4-D only slightly but not significantly prevented the drop in MiLAX2 levels, as compared to the control. Finally, with regard to the expression patterns of genes involved in IAA synthesis, the qRT-PCR data partially coincided with our data from the previous season, as a significant decrease in MiTAR2 and MiYUCCA10 accumulation, but not in MiYUCCA4 levels, was only

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

445

Fig. 4. Expression analysis of TAA1/TARs and YUCCA-related genes. qRT-PCR was performed with cDNA prepared from total RNA extracted from control and ethephon-treated tissues collected at different time points. Gene expression was calculated using the DDCt method as described in the legend to Fig. 2. Data are means ± SE of three independent biological replicates. Asterisks denote a significant difference between ethephon-treated and control tissue samples at each sampling point (P < 0.05). P e pericarp tissue, AZ e abscission zone. Sampling intervals are presented as hours from initial setting.

observed between control and ethephon-treated pericarp tissue. In addition, both in the pericarp and in the AZ, 2,4-D treatment did not affect the expression patterns of these genes, as compared to controls (see Fig. S3).

4. Discussion Previous physiological studies have demonstrated that natural mango fruitlet abscission involves the synthesis of ethylene in fruitlet tissues, which diffuses into the AZ to induce fruit drop (Nunez-Elisea and Davenport, 1986). The role of ethylene in inducing mango fruitlet abscission was further corroborated using ethephon treatment (Hagemann et al., 2015; Ish-Shalom et al., 2011; Singh et al., 2005). Here, results from our experiments showed that ethephon treatment induced high abscission rates, once again confirming the notion that ethephon hastens fruitlet abscission. Note, however, that under control conditions, abscission rates reached a value of 20e40% at the end of the experimental period, suggesting that a priori, the utilized fruitlet explants contained a proportion of fruitlets that were naturally close to abscission and/or that the setup of our experimental system induced a low level of abscission events. Examination of our expression analyses results revealed that one of the changes induced by ethephon was down-regulation of specific IAA-influx and efflux carriers in both pericarp and AZ. Specifically, ethephon-treated tissues exhibited significantly lower

MiPIN1 and MiLAX2 levels within18e24 h post-treatment, and lower MiPIN3 levels within 48 h. An exception to this declining pattern was observed for the MiLAX3 transcript, which increased in content in the pericarp and the AZ tissues following ethephon treatment. It is well established that PIN and AUX1/LAX family member proteins, contributing to the formation of local IAA maxima and minima, assume essential roles during plant growth and development, and also regulate shedding events (Pattison et al., 2014; Petrasek and Friml, 2009). For example, in tomato, where the expression of SlPIN1 and SlPIN2 was shown to peak in very young fruit immediately after anthesis, it was suggested that the role of SlPIN1 and SlPIN2 during fruit set is to enhance the transport of IAA to peduncles, suppressing abscission events (Nishio et al., 2010). In contrast, a decline in the abundance of specific IAAinflux and efflux carrier genes, probably obstructing IAA transport, has been suggested to be responsible for fruitlet abscission in apple (Botton et al., 2011), litchi (Li et al., 2015), grapevine berries (Kühn et al., 2016) and during abscission of mature melon fruit (Corbacho et al., 2013). In agreement with these latter reports, our qRT-PCR results, revealing a rapid decline in MiPIN1 and MiLAX2 mRNA levels following ethephon treatment, suggest a role for these IAA-carrier proteins in controlling mango fruitlet drop. Moreover, our data also imply that the minor decline in MiPIN1 and MiLAX2 levels observed under control conditions, in both examined tissues within 24e48 h, might be linked to the posteriorly observed lower rate of fruitlet drop. The question that remains unanswered,

446

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

Fig. 5. Changes in IAA and IAAeAsp content in ethephon-treated and control mango abscission zone (AZ) tissues.(A) Levels of IAA and IAAeAsp were determined at time 0 (T0) and at the indicated time points in control (C) and ethephon-treated (E) AZ. Means ± SE of three independent biological replicates, for each condition and time point, are presented. Means with different letters are significantly different (P  0.05). (lower and upper cases for IAA and IAA-Asp, respectively). (B) IAAeAsp:IAA ratios. For each condition and time point, IAAeAsp:IAA ratios were calculated by dividing IAAeAsp content by its IAA content. Means ± SE of three independent biological replicates, are presented. Means with different letters are significantly different (P  0.05).

however, is what might be the function of MiLAX3's upregulation, observed in the pericarp and AZ tissues upon ethephon treatment? In this context, it is important to note that recent Arabidopsis studies have revealed that AtAUX1 and AtLAX3 IAA influx-encoding transcripts are upregulated near the site of floral organ shedding (Basu et al., 2013). Based on these observations, it was further suggested that while elevated concentrations of IAA help to prevent the shedding process, a low level of IAA might serve to induce abscission events (Basu et al., 2013). In the case studied here, the finding that MiLAX3 transcript was upregulated by ethephon in the fruitlet tissues fits, in principle, with the above hypothesis. However, whether the function of MiLAX3 in delivering IAA from the fruitlet to the AZ is equivalent to that of other carriers remains to be established. In this study, the expression patterns of distinct genes related to IAA synthesis, including four TAA1/TARs and YUCCA family members, were also investigated. In Arabidopsis, these two groups of proteins are encoded by small gene families comprised of five and eleven members, respectively (Korasick et al., 2013; Zhao, 2010). In particular, the involvement of TAA1, TAR1 and TAR2 in auxin biosynthesis has been established experimentally, whereas different sets of YUCCA proteins seem to play overlapping functions, contributing to de-novo auxin synthesis, depending on tissue and cell type (reviewed in (Korasick et al., 2013; Zhao, 2010)). Here, our pericarp qRT-PCR results showed that within 24 h of ethephon treatment, two of the four genes (MiTAR2 and MiYUCCA10), were significantly downregulated as compared to controls. Together, these results might suggest that mango fruitlet drop is also associated with a decrease in de-novo IAA synthesis in the fruitlet tissue.

An additional insight from our expression analyses was the observation that ethephon treatment induces a general downregulation of five out of six examined early IAA-responsive genes, contrasting with upregulation of MiGH3.1 transcript, which was observed in both examined tissues. Abscission events associated with a general decline in genes encoding early IAA-induced proteins have been previously documented (Estornell et al., 2013; Sawicki et al., 2015). The finding of a similar phenomenon during ethephon-induced mango fruitlet drop was therefore predicted. In contrast, and of particular interest, was the upregulation pattern of MiGH3.1 transcript, which was observed within 4 h in both treated fruitlet tissues, and within 24 h after its levels initially decreased in the control AZ. It is important to note that our understanding of the function of GH3 genes is just beginning to emerge. For example, consistent with their potential role in affecting IAA homeostasis by conjugating excess IAA to amino acids, the activity of GH3.1 has been recently shown to be associated with control of fruit ripening. Especially in grape berries, VvGH3.1 expression has been shown to increase at the beginning of ripening, in parallel with a decrease in free IAA content and an increase in IAAeAsp levels (Bottcher et al., 2010). Notably, a transient increment in litchi GH3.3 transcript levels was also recently observed in activated fruitlet AZ, suggesting that LcGH3.3 plays a role in controlling the abscission process (Kuang et al., 2012). Assuming that MiGH3.1 functions as a group II GH3 protein, conjugating free auxin to amino acid, the elevated expression pattern of MiGH3.1 observed in both tissues upon ethephon treatment could suggests a role for MiGH3.1 in establishing low free IAA concentrations during mango fruitlet drop. As already mentioned, by means of [3H]-IAA measurements, it was recently shown that mango fruitlet drop begins with a reduction in polar IAA transport capacity through the fruitlet pedicel (Hagemann et al., 2015). In agreement with this notion, our IAA-quantification analyses showed that the sharp decrease in accumulation of the IAA carriers MiPIN1 and MiLAX2 observed in the ethephon-treated tissues was accompanied by a rapid and drastic decline in IAA content in the AZ. Similarly, it was noted that the attenuated decline in MiPIN1 and MiLAX2 content observed in control tissues was also accompanied by a certain degree of decay of free IAA content in the control AZ. Moreover, quantification analysis of IAA conjugates in the AZ showed that only IAAeAsp conjugates, thought to serve as precursors for the IAA-degradation pathway (Korasick et al., 2013; Ludwig-Müller, 2011), were present in treated and control samples. Surprisingly, however, where one would expect to see a higher concentration of IAAeAsp in ethephon-treated tissues, as compared to controls, an opposite result was observed. A possible explanation for this apparently contradictory result is that enhanced formation of IAAeAsp, induced by ethephon, is followed by its rapid degradation. In support of this scenario, the IAAeAsp:IAA ratio was indeed highest within 18 h of treatment, whereas 6 h after that, IAAeAsp conjugates were no longer detected in this tissue. It should be mentioned that degradation of IAAeAsp conjugates can only be proven by the identification of their respective metabolites, as no genes involved in the oxidative metabolism of IAAeAsp have been identified (Ludwig-Müller, 2011). Possible rapid formation of IAAeAsp metabolites in the activated mango AZ thus remains to be established. In addition, our IAA conjugate data also raise questions regarding the specific function of MiGH3.1. Specially, the finding that treated AZ tissue exhibited an increase in MiGH3.1 content, combined with the observation that IAAeAsp:IAA levels were highest in this tissue, suggest a role for MiGH3.1 in conjugating IAA to Asp. The finding that in the control AZ tissue, MiGH3.1 content decreased within 4 h but then increased back to its original level, concurrent with an increase in IAAeAsp level, also fits with the postulated role of MiGH3.1. However, an unexpected and striking characteristic of

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

MiGH3.1 was its high accumulation, lasting up to 48 h of treatment, when IAAeAsp content had already declined. Accordingly, any conclusion about the precise role of MiGH3.1 should be carefully drawn. It is still possible that other MiGH3-encoding genes play a role in conjugating IAA to Asp. Lastly, in this study, the effects of a synthetic auxin treatment on both fruitlet drop and on the expression of auxin-related genes were also investigated. Reports on synthetic IAA treatments reducing abscission events and activating specific auxin-related genes exist in the literature (Abebie et al., 2007; Xie et al., 2015). Furthermore, it is also known that in some cases, improved mango fruitlet retention could be achieved utilizing exogenous IAA treatments (see review by (Singh et al., 2005)). Interestingly, here, our expression analyses results showed that within 24 h, concomitantly with the inhibition of fruitlet drop events by 2,4-D, this IAA treatment prevented the down-regulation of MiPIN1 and of MiLAX3 transcripts in the examined fruitlet tissues. Moreover, and as expected, our expression analyses also revealed elevated expression levels of four early IAA responsive genes examined, including MiGH3.3 and MiGH3.1, following 2,4-D treatment. Taken together, these results might suggest that 2,4-D application improves fruitlet retention by enhancing IAA transport into the AZ. Moreover, as it was recently shown that 2,4-D can be metabolized in plants to from reversible 2,4-D-Asp and 2,4-D-Glu conjugates (Eyer et al., 2016), it would be of interest to further test the effects of 2,4-D, also at the level of the orchard, in conjunction with IAA-related expression studies and quantification assays of 2,4-D and 2,4-D conjugates levels. To conclude, considerable progress, based on studies performed in distinct fruit tree species, has been made in understanding the biological events controlling fruitlet abscission. In contrast, such information was completely lacking in mango. Our data offer, for the first time, insight into the fruitlet drop machinery of mango, pinpointing specific IAA-related genes with potential roles in decreasing IAA free content through the AZ, thereby affecting fruitlet abscission. Considering that this study focused on IAArelated gene-expression profiles in the pericarp and AZ, an indepth exploration of processes related to the induction of mango fruitlet drop is warranted. A better understanding of the mechanisms regulating mango fruitlet drop may assist in the future development of reliable methods to mitigate this problem, and might provide potential candidate genes for further biotechnological applications. Funding This work was supported by the Chief Scientist of the Ministry of Agriculture, Israel [grant number 203-1013 to VI and YC]. Author contributions Conceived and designed the experiments: YD, YC, VI. Performed the experiments: YD, SG, TZ, MIS, AL. Analyzed the data: YD, SG, TZ, AF, VI, YC. Contributed reagents/materials/analysis tools: VI,YC. Contributed to the writing of the manuscript: YD, YC, VI. Acknowledgements We thank Mr. Reuven Dor for providing access to his 'Kent' mango orchard. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.plaphy.2017.04.021.

447

References Abebie, A., Lers, A., Philosoph-Hadas, S., Goren, R., Riov, J., Meir, S., 2008. Differential effects of NAA and 2,4-D in reducing floret abscission in Cestrum (Cestrum elegans) cut flowers are associated with their differential activation of Aux/IAA homologous genes. Ann. Bot. (Lond) 101, 249e259. Basu, M.M., Gonzalez-Carranza, Z.H., Azam-Ali, S., Tang, S.Y., Shahid, A.A., Roberts, J.A., 2013. The manipulation of auxin in the abscission zone cells of Arabidopsis flowers reveals that indoleacetic acid signaling is a prerequisite for organ shedding. Plant Physiol. 162, 96e106. Bonghi, C., Ramina, A., 2000. Biochemical and molecular aspects of fruitlet abscission. Plant Growth Regul. 31, 35e42. Bottcher, C., Keyzers, R.A., Boss, P.K., Davies, C., 2010. Sequestration of auxin by the indole-3-acetic acid-amido synthetase GH3-1 in grape berry (Vitis vinifera L.) and the proposed role of auxin conjugation during ripening. J. Exp. Bot. 61, 3615e3625. Botton, A., Eccher, G., Forcato, C., Ferrarini, A., Begheldo, M., Zermiani, M., Moscatello, S., Battistelli, A., Velasco, R., Ruperti, B., Ramina, A., 2011. Signaling pathways mediating the induction of apple fruitlet abscission. Plant Physiol. 155, 185e208. Chapman, E.J., Estelle, M., 2009. Mechanism of auxin-regulated gene expression in plants. Annu. Rev. Genet. 43, 265e285. Cheng, C.Z., Zhang, L.Y., Yang, X.L., Zhong, G.Y., 2015. Profiling gene expression in citrus fruit calyx abscission zone (AZ-C) treated with ethylene. Mol. Genet. Genomics 290, 1991e2006. Corbacho, J., Romojaro, F., Pech, J.C., Latche, A., Gomez-Jimenez, M.C., 2013. Transcriptomic events involved in melon mature-fruit abscission comprise the sequential induction of cell-wall degrading genes coupled to a stimulation of endo and exocytosis. PLoS One 8 (3), e58363. http://dx.doi.org/10.1371/ journal.pone.0058363. Dal Cin, V., Velasco, R., Ramina, A., 2009. Dominance induction of fruitlet shedding in Malus x domestica (L. Borkh): molecular changes associated with polar auxin transport. BMC Plant Biol. 9, 139. http://dx.doi.org/10.1186/1471-2229-139. Estornell, L.H., Agusti, J., Merelo, P., Talon, M., Tadeo, F.R., 2013. Elucidating mechanisms underlying organ abscission. Plant Sci. 199, 48e60. , B., Oklestkova, J., Barbez, E., Kozubíkova , H., Pospísil, T., Eyer, L., Vain, T., Parízkova nek, M., Strnad, M., Robert, S., Novak, O., 2016. Wierzbicka, R., Kleine-Vehn, J., Fra 2,4-D and IAA amino acid conjugates show distinct metabolism in Arabidopsis. PLoS One 11 (7), e0159269. http://doi.org/10.1371/journal.pone.0159269. Gao, Y., Zhang, Y., Zhang, D., Dai, X., Estelle, M., Zhao, Y., 2015. Auxin binding protein 1 (ABP1) is not required for either auxin signaling or Arabidopsis development. Proc. Natl. Acad. Sci. U. S. A. 112, 2275e2280. Hagemann, M.H., Winterhagen, P., Hegele, M., Wunsche, J.N., 2015. Ethephon induced abscission in mango: physiological fruitlet responses. Front. Plant Sci. 6, 706. http://dx.doi.org/10.3389/fpls.2015. Ish-Shalom, M., Dahan, Y., Maayan, I., Irihimovitch, V., 2011. Cloning and molecular characterization of an ethylene receptor gene, MiERS1, expressed during mango fruitlet abscission and fruit ripening. Plant Physiol. Biochem. 49, 931e936. Korasick, D.A., Enders, T.A., Strader, L.C., 2013. Auxin biosynthesis and storage forms. J. Exp. Bot. 64, 2541e2555. Kuang, J.F., Wu, J.Y., Zhong, H.Y., Li, C.Q., Chen, J.Y., Lu, W.J., Li, J.G., 2012. Carbohydrate stress affecting fruitlet abscission and expression of genes related to auxin signal transduction pathway in litchi. Int. J. Mol. Sci. 13, 16084e16103. Kühn, N., Serrano, A., Abello, C., Arce, A., Espinoza, C., Gouthu, S., Deluc, L., ArceJohnson, P., 2016. Regulation of polar auxin transport in grapevine fruitlets (Vitis vinifera L.) and the proposed role of auxin homeostasis during fruit abscission. BMC Plant Biol. 16, 234e251. Li, C.Q., Wang, Y., Ying, P.Y., Ma, W.Q., Li, J.G., 2015. Genome-wide digital transcript analysis of putative fruitlet abscission related genes regulated by ethephon in litchi. Front. Plant Sci. 6, 502. http://dx.doi.org/10.3389/fpls.2015.00502. Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-time quantitative PCR and the 2(T)(-Delta Delta C) method. Methods 25, 402e408. Ludwig-Müller, J., 2011. Auxin conjugates: their role for plant development and in the evolution of land plants. J. Exp. Bot. 62, 1757e1773. Lulsdorf, M.M., Yuan, H.Y., Slater, S.M.H., Vandenberg, A., Han, X.M., Zaharia, L.I., Abrams, S.R., 2013. Endogenous hormone profiles during early seed development of C. arietinum and C. anatolicum. Plant Growth Regul. 71, 191e198. Meir, S., Philosoph-Hadas, S., Sundaresan, S., Selvaraj, K.S., Burd, S., Ophir, R., Kochanek, B., Reid, M.S., Jiang, C.Z., Lers, A., 2010. Microarray analysis of the abscission-related transcriptome in the tomato flower abscission zone in response to auxin depletion. Plant Physiol. 154, 1929e1956. Meir, S., Srivignesh, S., Riov, J., Ishangi, A., Philosoph-Hadas, S., 2015. Role of auxin depletion in abscission control. Stewart Postharvest Rev. 2 http://dx.doi.org/ 10.2212/spr.2015.2212.2212. Nishio, S., Moriguchi, R., Ikeda, H., Takahashi, H., Fujii, N., Guilfoyle, T.J., Kanahama, K., Kanayama, Y., 2010. Expression analysis of the auxin efflux carrier family in tomato fruit development. Planta 232, 755e764. Nunez-Elisea, R., Davenport, T.L., 1986. Abscission of Mango fruitlets as influenced by enhanced ethylene biosynthesis. Plant Physiol. 82, 991e994. Pattison, R.J., Csukasi, F., Catala, C., 2014. Mechanisms regulating auxin action during fruit development. Physiol. Plant. 151, 62e72. Peret, B., Swarup, K., Ferguson, A., Seth, M., Yang, Y.D., Dhondt, S., James, N., Casimiro, I., Perry, P., Syed, A., Yang, H.B., Reemmer, J., Venison, E., Howells, C.,

448

Y. Denisov et al. / Plant Physiology and Biochemistry 115 (2017) 439e448

Perez-Amador, M.A., Yun, J.G., Alonso, J., Beemster, G.T.S., Laplaze, L., Murphy, A., Bennett, M.J., Nielsen, E., Swarup, R., 2012. AUX/LAX genes encode a family of auxin influx transporters that perform distinct functions during Arabidopsis development. Plant Cell. 24, 2874e2885. Petrasek, J., Friml, J., 2009. Auxin transport routes in plant development. Development 136, 2675e2688. Sawicki, M., Ait Barka, E., Clement, C., Vaillant-Gaveau, N., Jacquard, C., 2015. Crosstalk between environmental stresses and plant metabolism during reproductive organ abscission. J. Exp. Bot. 66, 1707e1719. Sherman, A., Rubinstein, M., Eshed, R., Benita, M., Ish-Shalom, M., SharabiSchwager, M., Rozen, A., Saada, D., Cohen, Y., Ophir, R., 2015. Mango (Mangifera indica L.) germplasm diversity based on single nucleotide polymorphisms derived from the transcriptome. BMC Plant Biol. 15, 277. http://dx.doi.org/ 10.1186/s12870-015-0663-6. Singh, Z., Malik, A.U., Davenport, T.L., 2005. Fruit drop in mango. Hortic. Rev. 31, 111e153. Staswick, P.E., Serban, B., Rowe, M., Tiryaki, I., Maldonado, M.T., Maldonado, M.C., Suza, W., 2005. Characterization of an Arabidopsis enzyme family that

conjugates amino acids to indole-3-acetic acid. Plant Cell. 17, 616e627. Terol, J., Domingo, C., Talon, M., 2006. The GH3 family in plants: genome wide analysis in rice and evolutionary history based on EST analysis. Gene 371, 279e290. Tiwari, S.B., Hagen, G., Guilfoyle, T.J., 2004. Aux/IAA proteins contain a potent transcriptional repression domain. Plant Cell. 16, 533e543. Vanneste, S., Firml, J., 2009. Auxin: a trigger for change in plant development. Cell 136, 1005e1016. Weijers, D., Wagner, D., 2016. Transcriptional responses to the auxin hormone. In: Merchant, S.S. (Ed.), Annual Review of Plant Biology, vol. 67, pp. 539e574. Woodward, A.W., Bartel, B., 2005. Auxin: regulation, action, and interaction. Ann. Bot. 95, 707e735. Xie, R.J., Pang, S.P., Ma, Y.Y., Deng, L., He, S.L., Yi, S.L., Lv, Q., Zheng, Y.Q., 2015. The ARF, AUX/IAA and GH3 gene families in citrus: genome-wide identification and expression analysis during fruitlet drop from abscission zone A. Mol. Genet. Genomics 290, 2089e2105. Zhao, Y., 2010. Auxin biosynthesis and its role in plant development. Annu. Rev. Plant Biol. 61, 49e64.