ABSTRACT
The role of ethylene in plant development is mostly inferred from its exogenous application. The usage of the mutants affecting ethylene biosynthesis proffers a better alternative to decipher its role. In tomato, 1-aminocyclopropane carboxylic acid synthase2 (ACS2) is a key enzyme regulating ripening-specific ethylene biosynthesis. We characterized two contrasting acs2 mutants; acs2-1 overproduces ethylene, has higher ACS activity, and increased protein levels, while acs2-2 is an ethylene under-producer, displays lower ACS activity, and protein levels than wild type. Consistent with high/low ethylene emission, the mutants show opposite phenotypes, physiological responses, and metabolomic profiles than the wild type. The acs2-1 showed early seed germination, faster leaf senescence, and accelerated fruit ripening. Conversely, acs2-2 had delayed seed germination, slower leaf senescence, and prolonged fruit ripening. The phytohormone profiles of mutants were mostly opposite in the leaves and fruits. The faster/slower senescence of acs2-1/acs2-2 leaves correlated with the endogenous ethylene/zeatin ratio. The genetic analysis showed that the metabolite profiles of respective mutants co-segregated with the homozygous mutant progeny. Our results uncover that besides ripening, ACS2 participates in vegetative and reproductive development of tomato. The distinct influence of ethylene on phytohormone profiles indicates intertwining of ethylene action with other phytohormones in regulating plant development.
INTRODUCTION
Ethylene is a simple gaseous molecule that also acts as a natural plant hormone. It participates in a multitude of development processes such as organ senescence, biotic and abiotic stresses, and fruit ripening (Abeles et al., 1992). The ethylene-mediated stimulation of the ripening process is restricted to climacteric fruits, where a surge in respiration marks the onset of ripening. The respiratory surge is preceded by increased ethylene biosynthesis that triggers ripening (Grierson, 2013). Antagonistically, mutations, or chemical treatments that block ethylene biosynthesis/perception, abolish or delay ripening of climacteric fruits (Brady, 1987; Martínez-Romero et al., 2007). For studies on climacteric fruit ripening, tomato (Solanum lycopersicum) has emerged as a preferred model system due to the availability of several monogenic mutants affecting the ripening process and ease of transgenic manipulations (Barry, 2014).
Studies carried out on tomato mutants defective in fruit ripening established a hierarchical genetic regulation of the ripening process. One of the extensively investigated mutants is ripening-inhibitor (rin) that lacks almost all ripening associated processes like the accumulation of lycopene, softening of fruits, and a climacteric burst of ethylene (Tigchelaar et al., 1978). The RIN gene encodes a MADs-type transcriptional factor, which positively regulates the onset of ripening process (Vrebalov et al., 2002). RIN directly binds to promoters of a large number of ripening-related genes, including those involved in ethylene biosynthesis (Fujisawa et al., 2012; Qin et al., 2012). The mutated rin gene encodes an in-frame fused RIN and Macrocalyx protein which strongly represses the expression of target genes (Ito et al., 2017). Consequently, the rin mutant does not ripen on exposure to exogenous ethylene, though it retains some ethylene-induced responses independent of RIN (Lincoln and Fischer, 1988).
The paramount role of ethylene in regulating tomato ripening is highlighted by the loss of ripening in tomato Nr mutant. The Nr mutant encodes a truncated ethylene receptor ETR3, resultantly compromised in ethylene perception (Wilkinson et al., 1995). The ripening can be restored in transgenic Nr fruits by antisense inhibition of mutated ETR3 gene suggesting receptor inhibition model of ethylene action (Hackett et al., 2000). Likewise, overexpression of EIL1 (EIN3-like transcription factor) restores ripening in Nr (Chen et al., 2004), indicating the operation of normal ethylene signal transduction in the mutant. The mutations in ETR1, ETR4, and ETR5 genes had a differential effect on tomato ripening, probably by affecting their ethylene sensitivity (Okabe et al., 2011; Mubarok et al., 2019).
The elegant studies initially conducted on wounded apple fruit tissues uncovered the ethylene biosynthesis pathway in higher plants named as Yang cycle (Adams and Yang, 1979). The ethylene is derived from C-3,4 of amino acid methionine, which is first converted to S-adenosyl-L-methionine (SAM) by SAM synthase. SAM, a common precursor of many biosynthetic pathways, is converted to 1-aminocyclopropane carboxylic acid (ACC) by ACC synthase (ACS). The conversion of ACC to ethylene is catalyzed by ACC oxidase (ACO) in the presence of oxygen. The formation of ACC by ACS constitutes the first committed step in ethylene biosynthesis.
ACS is considered as the rate-limiting enzyme in the ethylene biosynthesis pathway. In tomato, ACS is encoded by a multigene family consisting of at least nine genes (ACS1A, ACS1B, and ACS2-8) and five putative genes (Liu et al., 2015). During tomato fruit development and ripening, ACS genes are differentially regulated by endogenous mechanisms and ethylene-mediated autocatalytic induction. During tomato fruit development, expressions of ACS genes considerably differ in system-I and system-II responses (Klee and Giovannoni, 2011). During system-I that is confined to the fruit expansion phase, ACS genes are moderately expressed. During system-II, which marks ripening induction, ACS2, ACS4, and ACS6 are highly expressed (Barry et al., 1996, 2000; Van de Poel et al., 2012). The strong association between ripening and increased ACS2 expression was indicated by the total suppression of tomato fruit ripening by antisense inhibition of the ACS2 gene (Oeller et al., 1991). The high expression of ACS2 during ripening likely results from the binding of transcription factors like RIN (Ito et al., 2008) and TAGL1 (Itkin et al., 2009) to the promoter of ACS2 gene. Genome-wide analysis of DNA methylation demonstrated that tomato ripening is associated with the demethylation of several promoters, including ACS2 and ethylene receptors ETR3 and ETR4 (Zhong et al., 2013). Based on the absence of binding of TAGL to ACS4 promoter, Itkin et al. (2009) suggested ACS2 is the primary ethylene biosynthesis gene, regulating tomato ripening. The kinetic analysis of ethylene biosynthesis, ACC accumulation, ACS activity, and ACS genes expression indicated that during tomato ripening ACS4 and ACS6 are less important, whereas ACS2 seems to be the main biosynthetic enzyme as its expression is ethylene dependent and mainly occurs during system 2 (Van de Poel et al., 2014).
While the ACS2 gene along with other ethylene biosynthesis genes, seemingly contributes to the climacteric rise of ethylene during tomato ripening, information about its role in other developmental processes of tomato is limited. The paucity of information about its role stems from the near absence of tomato mutants compromised in the ethylene biosynthesis pathway. Here we describe isolation and characterization of two novel ACS2 mutants of tomato having diametrically opposite effects on ethylene emission. The acs2-1 mutant has high ethylene emission, and acs2-2 shows reduced ethylene emission. We show that stimulation/reduction of ethylene emission affects several developmental processes right from seed germination to fruit ripening. We also show that variation in ethylene emission affects hormonal and metabolome profiles in an opposite manner.
RESULTS
Isolation of acs2 mutants
To identify ACS2 gene mutants, we screened genomic DNA from 9,144 EMS-mutagenized M2 plants by TILLING. In total, nine alleles were identified, of which two named acs2-1 (M82-M3-112) and acs2-2 (M82-M2-162A) were characterized in detail (Supplemental Table S1). Sequencing of the full-length gene, including promoter, revealed two exonic and three intronic mutations in acs2-1, and one intronic and two promoter-localized mutations in acs2-2 (Figure 1A, Supplemental Table S2). Homozygous mutant lines were characterized in M6 generation and backcrossed twice to parent M82 plants. Genetic segregation analysis showed that acs2-1 and acs2-2 were inherited as a monogenic Mendelian trait. Segregation of mutation in F2 progeny was monitored by CEL-I endonuclease assay (Mohan et al., 2016) [acs2-1 BC1F2-120-plants, 29 (acs2-1/acs2-1), 61 (ACS2/acs2-1), and 30 (ACS2/ACS2), ratio 1:2:1, χ2 (0.224) P = 0.62]; [acs2-2 BC1F2-total 108-plants, 25 (acs2-2/acs2-2), 56 (ACS2/acs2-2), and 27 (ACS2/ACS2), ratio 1:2:1, χ2 (0.101) P = 0.95] (Supplemental Table S3). For the majority of the fruit ripening experiments, we compared respective mutants with parental wild type (WT), and BC1F2 progeny.
In silico analysis predicted over-expression/reduced-expression for acs2 mutants
Splice site analysis by NetGene2-server (http://www.cbs.dtu.dk/services/NetGene2/) predicted that A398G (K100=) mutation located in acs2-1 at 5′ splice site terminating the second exon leads to more efficient mRNA splicing that in turn may enhance acs2-1 transcript level (Supplemental Figure S1). Computational protein modeling and protein stability analysis indicated that V352E change affects bonding and folding pattern, improving the stability of ACS2-1 protein (Figure 1B, Supplemental Table S4–5). Taken together, enhanced transcript level and stable ACS2-1 protein may lead to an over-expression hypermorphic mutation.
It is reported that C(A/T)8G sites also have an affinity for RIN binding (Fujisawa et al., 2013). In acs2-2, T-106A mutation is located in C(A/T)9G location, therefore it is unlikely that it may have disrupted RIN binding site. However, T-106A mutation disrupts a SOC1 binding site and gains an AZF binding site (Kodaira et al., 2011; tomato homolog Solyc04g077980.1), a C2H2-type transcription factor negatively regulating ABA-mediated responses (Supplemental Table S6, Supplemental dataset S1). T-382A mutation is located at a methylated CpG site that shows progressive methylation reduction during ripening (http://ted.bti.cornell.edu/cgi-bin/epigenome/home.cgi) (Supplemental Table S7, Supplemental dataset S2). It is plausible that the above mutations reduce the affinity of regulatory transcription factors, thus compromising acs2-2 gene expression resulting in a reduced-expression phenotype.
acs2-1 shows faster seed germination
In acs2-1, the onset of seed germination was 12h earlier than WT, whereas it was delayed by 12h in acs2-2. The attainment of full germination followed the above pattern, with acs2-1 attaining 50% germination at 35h, WT at 55h, and acs2-2 at 77h (Figure 1C). Ethylene emission from respective mutant seedlings also showed a similar pattern. Three-day old acs2-1 seedlings showed higher emission (280%), and acs2-2 lower emission (40%) than WT (Figure 1D). Mutant seedlings showed a similar pattern in ethylene-mediated growth inhibition with acs2-1 being more sensitive and acs2-2 less sensitive than WT (Figure 1E-F).
acs2 mutants have contrasting phenotypes
During vegetative growth, acs2-1 displayed faster growth, elongated internodes, and higher ethylene emission from detached leaves than WT (Figure 2A,C). In contrast, acs2-2 exhibited slightly slower growth and lower ethylene emission than WT (Figure 2B,C). Akin to ethylene emission, detached mutant leaves showed faster (acs2-1) and slower (acs2-2) loss of coloration than WT (Figure 2D). In mutant leaves, chlorophylls and carotenoids levels similarly differed from WT (Supplemental Figure S2). Hormonal profiling of acs2-1 leaves showed up-regulation of jasmonic acid (JA), ABA, methyl jasmonate (MeJA), and down-regulation of IAA, IBA, salicylic acid (SA), and zeatin than WT. In contrast, the hormonal profile of acs2-2 was closer to WT, barring lower levels of SA, and higher levels of zeatin (Figure 2E,F). PCA of primary metabolites of mutants was distinctly different from WT (Figure 2G, Supplemental dataset S3). In general, more metabolites were downregulated in acs2-1 than acs2-2 (Figure 2H). Key TCA cycle metabolites, such as citrate, isocitrate, and malate, showed opposite levels indicating a differential effect of acs2-1 and acs2-2 on the TCA cycle. In acs2-2 levels of several amino acids were high, whereas, in acs2-1 barring threonine, other amino acids showed lower abundance than WT.
acs2-1 shows accelerated fruit ripening
Both acs2 mutants also affected the reproductive phase. acs2-1 had more flowers with a higher fruit set per truss, whereas acs2-2 had fewer flowers and a lower fruit set per truss than WT (Figure 3A, Supplemental Table S8). These results indicate that acs2-1 accelerates the overall reproductive phase and enhances fruit yield, and acs2-2 has the opposite phenotype. Ripened fruits of acs2-1 were smaller, and acs2-2 were bigger than WT (Figure 3B). acs2-1 fruits showed accelerated development reaching mature green (MG) stage by 31 days post-anthesis (dpa) (Figure 3B,C). Post-MG stage, acs2-1 showed accelerated ripening attaining red ripe (RR) stage by 38 dpa. RR stage of acs2-1 fruits was severely short (8-days) with the onset of fruit senescence (FS) at 46 dpa. Though WT and acs2-2 fruits reached the MG stage at the same time, the duration to attain the RR stage was significantly longer in acs2-2 (Figure 3C,D). Post-RR stage, the onset of FS was also significantly delayed in acs2-2 (40-days) than WT (25-days). While ripening acs2-1 fruits emitted more ethylene, acs2-2 fruits had lower ethylene emission than WT. Respiratory CO2 emission from acs2-1, acs2-2, and WT fruits followed a pattern similar to ethylene, indicating a linkage between these two processes (Figure 3E-F). Analysis of homozygous backcrossed progeny of acs2-1 and acs2-2 showed ethylene emission and ripening pattern, nearly similar to the respective mutant (Figure 3 C-E). acs2-1 fruits were less firm at TUR and RR, while acs2-2 fruits were more firm than WT. acs2 mutations also had a minor effect on °Brix and pH value of fruits (Supplemental Figure S3). The analysis of BC4F2 progeny of acs2-1 and acs2-2 in Arka Vikas (AV) and Pusa Early Dwarf (PED) cultivars also showed ethylene emission and ripening pattern, similar to respective mutant (Supplemental Figure S4).
acs2 mutants differently affect ACS activity
In plants, ethylene synthesis is a two-step process, wherein SAM is converted to ACC by ACS. In the next step, ACO converts ACC to final end-product ethylene (Yang and Hoffman, 1984). We determined whether higher/lower ethylene emission from acs2 fruits correlated with in vivo levels of ACC. Consistent with higher ethylene emission in acs2-1, the ACC level was higher, whereas, in acs2-2, it was lower than WT (Figure 4A). In vivo, ACC is also conjugated to malonic acid to form malonyl-ACC (MACC) by ACC malonyltransferase (Hoffman et al., 1982). Akin to higher ACC, MACC level was also higher in acs2-1, whereas acs2-2 had lower MACC levels than WT (Figure 4B). To ascertain whether high/low ACC levels in mutants reflected in ACS enzyme activity, we in vitro assayed ACS activity. Congruent to ACC levels, ACS activity was 2-fold higher in acs2-1, and 0.5-fold in acs2-2 compared to WT (Figure 4C). Next, we examined whether high/low ACC levels also influenced the activity of the subsequent enzyme-ACO. Even ACO activity was significantly higher in acs2-1, whereas acs2-2 had lower ACO activity (Figure 4D).
acs2 mutants differ in ACS protein levels
To determine whether altered ACS activity reflected changes in ACS2 protein levels, we raised polyclonal antibodies against ACS2 using a synthetic peptide, and antibody specificity was confirmed against the peptide (Supplemental Figure S5). In ripening tomato pericarp, the level of ACS protein is estimated to be <0.0001% of total soluble protein (Bleecker et al., 1986). Due to its low abundance, we could not detect ACS2 protein on Western blot on the direct loading of supernatants from fruit homogenates. Therefore, we purified IgG fraction and immunoprecipitated ACS2 protein to enrich it before Western blot analysis (Supplemental Figure S6). Reduction in ACS activity in the supernatant after immunoprecipitation indicated the specificity of antibodies toward ACS2 protein. Western blot analysis revealed no discernible ACS2 protein band in MG and TUR fruits of acs2-1, acs2-2, and WT. Only in RR fruits, a single band of 55 kD was detected corresponding to the reported molecular size of tomato ACS2 protein in acs2-1 and WT (Rottmann et al., 1991).
A more prominent band in acs2-1 compared to WT indicated that it is enriched in ACS2 protein. In contrast, a very faint band in acs2-2 RR fruits indicated a highly reduced ACS2 protein level (Supplemental Figure S6). To ascertain the relative amount of ACS2 protein in acs2-1, acs2-2, and WT RR fruits, we subjected samples to serial dilution. Matching the intensity of bands of serially diluted proteins indicated that the level of ACS2 protein in WT was about 4-fold less than acs2-1, whereas ACS2 protein level in acs2-2 was about 4-fold lower than WT (Figure 4E-F). Differences in ACS2 protein levels also support that acs2-1 is an over-expression, and acs2-2 is a reduced-expression mutant.
It is believed that ACS2 protein stability is regulated by protein phosphorylation at a conserved Ser-460 residue (Kamiyoshihara et al., 2010). Therefore, we examined whether mutation affected the phosphorylation status of ACS2 protein using a phospho-ser antibody. Considering that ACS2 protein level widely varied in acs2-1, WT, and acs2-2, the gel was run with 20 µg WT protein, 5 µg acs2-1 protein, and 80 µg acs2-2 protein, to have an equal amount of ACS2 protein in each lane. At equal levels of ACS2 protein loading, acs2-1, acs2-2, and WT lanes showed an equal amount of phosphorylated ACS2 protein, indicating that mutations did not affect ACS2 phosphorylation (Figure 4G).
acs2 mutations alter phytohormones level in fruits
Analysis of mutant fruits revealed that acs2 mutations also significantly influenced other plant hormones. Levels of zeatin, ABA, JA (Barring MG), MeJA, and SA were higher in acs2-1 than WT. Contrastingly, the IAA level was lower at MG, and at later stages, it was similar to WT (Figure 5A). Notably, homozygous F2 acs2-1/acs2-1 plants showed similar upregulation of zeatin, ABA, JA, MeJA, and SA compared to F2-WT (ACS2/ACS2). Contrasting to acs2-1, and compared to WT, levels of ABA, and SA was lower in acs2-2 (Barring MG), and in homozygous F2 (acs2-2/acs2-2) plants. Levels of MeJA, JA (at RR), and zeatin (at MG, TUR) was also lower than WT in homozygous F2 (acs2-2/acs2-2) plants. IAA level in homozygous F2 (acs2-2/acs2-2) plants (In acs2-2 only at TUR) was higher than WT (Figure 5A). These results indicate the opposite influence of acs2-1 and acs2-2 on the levels of other hormones during ripening.
acs2 mutants show higher carotenoid accumulation in fruits
The onset, progression, and completion of tomato ripening are visually monitored by the change of color imparted by carotenoids. While a total of 11 carotenoids were detected in fruits; only phytoene, lycopene, lutein, and β-carotene were detected at all ripening stages. In homozygous F2 (acs2-1/acs2-1) RR fruits, total carotenoid level was 1.85-fold higher than WT (acs2-1, 1.48-fold). The above increase was mainly contributed by a near doubling of lycopene and phytoene in acs2-1. Strikingly, acs2-2 and homozygous F2 (acs2-2/acs2-2) RR fruits also had a higher total carotenoid level, mainly due to higher accumulation of lycopene. In acs2-2, levels of different carotenoids were distinctly different from both WT and acs2-1 (Figure 5B). The higher lycopene accumulation at the RR stage was also observed in BC4F2 plants of AV (acs2-1/acs2-1) and PED (acs2-2/acs2-2) substantiating the influence of respective acs2 mutants on lycopene levels (Supplemental Figure S4).
acs2 mutants oppositely affect metabolites level in fruits
Ripening of tomato is driven by extensive metabolic shifts making fruits palatable. Several of these changes are triggered by ethylene, as fruits of Nr mutant, which is deficient in ethylene perception, show different patterns of metabolic shifts than WT (Osorio et al., 2011). Consistent with the opposite nature of acs2 mutants, PCA revealed distinct differences between metabolites of acs2-2, acs2-1, and WT (Figure 6A, Supplemental dataset S4). Notably, PCA profiles of F2 fruits were distinctly closer to respective parents. PCA profile of acs2-1 and homozygous F2 (acs2-1/acs2-1) overlapped. Similarly, PCA profiles of WT and homozygous F2 (ACS2-1/ACS2-1) distinctly overlapped at MG and TUR and were in close vicinity at RR. Strikingly, heterozygote F2 (ACS2-1/acs2-1) occupied an intermediate position in PCA between WT and acs2-1 (Figure 6B). A similar pattern in PCA profiles was also observed for acs2-2, where homozygous F2 mutant (acs2-2/acs2-2) overlapped with the respective parent, and heterozygous F2 (ACS2-2/acs2-2) showed intermediate PCA profile (Figure 6C).
Comparison of metabolite profiles of acs2-1 and acs2-2 (Figure 7) and their backcrossed progeny with WT highlighted a strong linkage between the mutated gene and metabolome. The metabolome of acs2-1 and F2 (acs2-1/acs2-1) was closer to each other than WT. Similarly, the metabolome of WT and F2 (ACS2-1/ACS2-1) showed only mild differences. The same was observed for metabolite profiles of acs2-2 and F2 (acs2-2/acs2-2) fruits (Figure 7, Supplemental dataset S4). Metabolome of ACS2-1/acs2-1 and ACS2-2/acs2-2 was intermediate. Observed profiles of metabolome were also consistent with PCA profiles of mutated plants and their backcrossed progeny.
A comparison of metabolite profiles of acs2-1 and acs2-2 highlighted distinct differences between the two mutants. The effect of acs2-1 on the relative up-/down-regulation of most metabolites was milder than acs2-2. In acs2-2, the majority of amino acids derived from the glycolysis pathway were upregulated, whereas acs2-1 only mildly affected the aminome except for 10-fold higher methionine in RR fruits. Commensurate with high ethylene emission and methionine, ACC levels were also high in acs2-1 fruits at the RR stage. In acs2-2, 2-oxoproline was upregulated at all ripening stages. Similar to aminome, most organic acids were downregulated in acs2-1 fruits. Only a few organic acids were upregulated in acs2-1, such as citrate, α-ketoglutaric acid, isocitric acid (TUR, RR), and succinate (TUR). Similarly, only a few organic acids were upregulated in acs2-2 [Lactate; Oxalate (MG, TUR); Phosphoric acid; Maleate (MG)]. Compared to acs2-2, fewer sugars were higher in acs2-1 than WT. In acs2-2, sucrose and glucose-6-phosphate were higher than WT (Supplemental figure 7).
acs2-1 mutation elevates ETR3/4 expression
System II ethylene synthesis in ripening tomato fruits is associated with higher transcript levels of ACS2, ACS4, ACO1, ACO2, and ACO3 (Liu et al., 2015). During ripening, transcript levels of ACS2, ACS4, ACO1, and ACO2 were upregulated in acs2-1, in homozygote-F2 (acs2-1/acs2-1) and heterozygote-F2 (ACS2-1/acs2-1) at TUR and RR stages than parental WT and WT-F2 (ACS2-1/ACS2-1) (Figure 8A). Contrastingly, in acs2-2, in homozygote-F2 (acs2-2/acs2-2), and in heterozygote-F2 (ACS2-2/acs2-2) transcript levels of ACS2 (barring RR in acs2-2), ACS4 (barring TUR in acs2-2), ACO1, ACO2 and ACO4 (barring ACS2-2/acs2-2), were lower during ripening WT. Both acs2-1 and acs2-2 and their respective homozygous F2 progeny had no consistent influence on transcript levels of ACO3. Out of six ETR genes in tomato, acs2-1 specifically upregulated ETR3, ETR4 at the TUR stage (Figure 8A). In contrast, acs2-2 and its homozygote-F2 (acs2-2/acs2-2), downregulated ETR1, ETR2, and ETR4 (barring acs2-2 at BR) at all stages of ripening. For ETR5, only homozygote-F2 (acs2-2/acs2-2) showed downregulation at BR and RR stages. For ETR6, no consistent influence of acs2-1 or acs2-2 and their homozygote was observed.
We next examined whether higher carotenoids level in acs2-1 and acs2-2 was also associated with increased expression of carotenoid biosynthesis genes. Influence of acs2-1, its homozygote-F2 (acs2-1/acs2-1), and heterozygote-F2 (ACS2-1/acs2-1) on the expression of carotenoid biosynthesis genes at TUR and RR stages was mainly confined to three genes of pathway-phytoene synthase1 (PSY1), phytoene desaturase (PDS), and carotenoid isomerase (CRTISO). Higher transcript levels of PSY1, PDS, and CRTISO in heterozygote-F2 (ACS2-1/acs2-1) fruits indicate a semi-dominant influence of acs2-1 on these genes. acs2-1 homozygote also upregulated chromoplast-specific lycopene-β-cyclase (CYCB). Besides the above genes, acs2-1 upregulated NCED1 and ZEP (only at TUR) expression. Interestingly, the expression of PSY1 followed an opposing pattern in acs2-2 with lower transcript levels. Few other genes in acs2-2 were downregulated at specific stages, viz. ZDS (at TUR in homozygote), CRITSO (homo- and hetero-zygote at TUR, homozygote at RR), and CYCB (homo- and hetero-zygote at RR) (Figure 8B).
Discussion
Ethylene plays several roles throughout the life cycle of plants, regulating a plethora of developmental responses. Among these, ethylene regulation of the ripening of climacteric fruits has drawn the most attention. In this study, a comparison of two contrasting acs2 mutants uncovers that besides regulating fruit ripening; ethylene produced from ACS2 also modulates several aspects of tomato phenotype right from seed germination.
Ethylene promotes seed germination
It is established that in higher plants, seed germination is antagonistically regulated by ABA and GA (Shu et al., 2016). Our results indicate that endogenous ethylene participates in tomato seed germination. Faster germination of acs2-1 is likely related to higher ethylene emission. Slower germination of acs2-2 associated with reduced ethylene emission conforms to the above assumption. Consistent with this, peak emission of ethylene from acs2-1 seeds coincides with the completion of germination (Lashbrook et al., 1998). Considering that transgenic tomato seeds overexpressing ERF2 show faster germination (Pirrello et al., 2006), it is likely that higher ethylene emission from acs2-1 amplifies ethylene signaling.
Seedlings of acs2-1 grown in the air do not manifest constitutive triple response alike Arabidopsis ETO mutants (Guzman and Ecker, 1990) or tomato Epi mutant (Barry et al., 2001). Since ethylene biosynthesis/action inhibitors do not revert Epi phenotype, Epi probably affects an ethylene-signaling component (Fujino et al., 1989; Barry et al., 2001). Considering that Arabidopsis eto2 (ACS5) and eto3 (ACS9) seedlings produce 20-90 fold higher ethylene than the wild-type, 3-4 fold higher ethylene emission by acs2-1 seedlings may not be sufficient to elicit a constitutive triple response (Kieber et al., 1993; Chae et al., 2003). Nevertheless, higher ethylene emission from acs2-1 seems to elicit shorter hypocotyls in etiolated seedlings than WT. Further shortening of hypocotyls in ethylene-treated seedlings likely reflects combined action of external and endogenous ethylene produced by seedlings, as hypocotyls shortening followed acs2-1>WT>acs2-2.
Ethylene affects both vegetative and reproductive phenotypes
Contrasting phenotypes of acs2-1 and acs2-2 seem to result from high/low ethylene emission from respective mutants. Despite higher ethylene emission, acs2-1 phenotype had no resemblance to 35S::ACS2 transgenic tomato displaying epinasty, reduced growth, and higher ethylene emission (Lee et al., 1997). Likewise, the acs2-1 phenotype was also distinct from the tomato Epi mutant that displays an erect growth habit, curled leaves, thickened stems, and petioles. Notably, the Epi effect is restricted to vegetative development (Barry et al., 2001), whereas acs2-1 affects both vegetative and reproductive phenotypes.
Contrasting phenotypes of acs2-1 and acs2-2, while emanate from high/low ethylene emission, modulation of hormones and metabolites also underlie phenotypic differences. Hormonal profiles of acs2-1 and acs2-2 leaves indicated crosstalk between ethylene and other hormones. In conformity with reduced ethylene emission delays tomato leaf senescence (John et al., 1995), detached leaves of acs2-1 showed faster senescence, while acs2-2 showed slower senescence. In tomato plants grown under high salinity, onset and progression of leaf senescence correlated with endogenous zeatin/ACC ratio (Ghanem et al., 2008). Likewise, opposite zeatin and ethylene levels in acs2-2 and acs2-1 leaves imply that the ethylene/zeatin ratio is a key determinant regulating leaf senescence. Similarly, high ABA/ethylene level in acs2-1 leaves agrees with a parallel increase in ABA and ethylene in salt-stressed tomato plants (Ghanem et al., 2008).
Generally, plants undergoing pathogen or insect damage show a concerted action of ethylene with SA and JA to activate defense responses (Yang et al., 2015). In tomato plants infected with Alternaria, ethylene and JA acted synergistically to promote susceptibility. Conversely, SA promoted resistance to Alternaria and antagonized ethylene signaling (Jia et al., 2013). Therefore, it is conceivable that higher ethylene emission in acs2-1 affects levels of defense hormones such as JA and SA. The upregulation of JA, MeJA, and downregulation SA in acs2-1 leaves, conforms with this assumption.
The relationship between auxin and ethylene is complex and has both synergistic and antagonistic interactions (Muday et al., 2012). Auxin and ethylene co-act to facilitate tomato root tip penetration in soil (Santisree et al., 2011). Conversely, ACC-treatment reduces free auxin in tomato roots indicating antagonistic action of ethylene (Negi et al., 2010). Reduction in IAA and IBA levels in acs2-1 points toward an antagonistic action of ethylene on the IAA/IBA level. Since barring zeatin and SA, other hormones are not affected in acs2-2 leaves; it seems that hormonal modulation in acs2-1 leaves reflects a response analogous to stress, where on surpassing a threshold, ethylene modulates levels of stress-related hormones.
Distinct PCA profiles of acs2 mutants are consistent with their contrasting responses. The cross-comparison showed that most metabolites were downregulated in acs2-1, while in acs2-2, several were upregulated than WT. Ostensibly, altered ethylene emission from acs2-1 and acs2-2 leaves affects metabolites profiles oppositely. Higher citrate, isocitrate levels, and lower malate levels in acs2-1 leaves indicate an upsurge in respiratory metabolism, a hallmark of stress. Purportedly, citrate accumulation is construed as a response mechanism to alleviate stress (Gupta et al., 2012). Differences in ethylene emission also influence the reproductive phenotype, as both mutants display opposite flower numbers and fruit set than WT. Taken together, it is implicit that distinct phytohormone and metabolite profiles of acs2 mutants underlie differences in their phenotypes.
acs2 mutants show contrasting ripening progression
It is believed that ACS2 contributes to system-II ethylene emission, whereas its role in system-I ethylene emission is minimal (Barry et al., 2000). Nonetheless, faster progression to the MG stage in acs2-1 implies a role for ACS2 in system-I ethylene emission. Consistent with this, ethylene-treated Micro-Tom fruits, too showed faster attainment of the MG stage (Kevany et al., 2007). Our results support the view that the transition from MG to RR stage is linked with ethylene-induced ripening. Post-MG stage, faster progression to RR stage, and the onset of senescence in acs2-1 are consistent with the pivotal role of ACS2 in the progression of ripening. The above assumption is also corroborated with slower progression to the RR stage and delayed senescence in acs2-2 fruits. Shorter MG to RR transition period in F2-acs2-1/acs2-1 and a far-longer period in F2-acs2-2/acs2-2 progeny of M82, AV and PED, further corroborate close linkage with ACS2.
Ethylene also influences the expression of ETR3/4 genes
The higher transcript level of ETR3 and ETR4 in acs2-1 fruits at the TUR stage indicates the likelihood of stimulation by ethylene (Lashbrook et al., 1998; Kevany et al., 2007). Lowered transcript levels of ETR1, 2, 4, 5 in acs2-2 indicates that their expression is also related to ethylene levels (Okabe et al., 2011; Mubarok et al., 2019). Higher ACS2, ACS4, ACO1, and ACO2 transcript levels in ripening acs2-1 fruits conform with the notion that ethylene in an autocatalytic fashion enhances its synthesis. Consistent with this, reduced ethylene emission from acs2-2 fruits lowers ACS2, ACS4, ACO1, ACO2, and ACO4 transcript levels. Taken together, it appears that higher ethylene emission from acs2-1 fruits promotes higher transcript levels of ripening-specific ethylene biosynthesis genes and ethylene receptors, and it is opposite in acs2-2.
ACS activity of mutants is correlated with ACS2 protein levels
The in vitro assays corroborated that higher ethylene emission was causally related to ACS activity. In ripening tomato fruits, endogenous ACC content closely correlates with ethylene emission (Hoffman and Yang, 1980). Higher ethylene emission from acs2-1 fruits also correlates with higher ACC as well as MACC levels. Also, acs2-1 fruits had higher ACS and ACO enzyme activities. Considering that acs2-1 fruits accumulate 4-fold higher ACS2 protein, it may have contributed to higher ACS activity. Conversely, lower ethylene emission from acs2-2 was associated with lower ACC and MACC levels, reduced ACS, and ACO enzyme activities and lower amounts of ACS2 protein.
Substantial differences in immuno-detectable ACS2 protein in acs2-1 and acs2-2 fruits indicated that the above mutations likely affect transcription or stability of the protein. The likelihood of enhanced stability of ACS2 protein in acs2-1 by post-translational modification seems to be remote. Typical of type-I ACS, tomato ACS2 protein has three conserved serine phosphorylation sites in C-terminal (Kamiyoshihara et al., 2010). However, there was no difference in phosphorylation magnitude in ACS2 protein in mutants and WT. In Arabidopsis eto2 (ACS5) and eto3 (ACS9) mutants encoding Type-II ACS, mutations close to C-terminal stabilizes ACS protein by making it resistant to E3 ligase dependent proteolysis (Chae and Kieber, 2005). However, Type-I ACS proteins are not subjected to ubiquitin-mediated proteolysis. It remains possible that α-helix change by V352E confers protein stability or stimulates ACS2 catalytic activity. The above possibility is supported by in silico analysis showing higher stability for ACS2-1 protein than WT, which may have contributed to increased ACS2 levels in acs2-1.
Notwithstanding predicted higher ACS2-1 protein stability, increased ACS2 transcript level in acs2-1 seems to also contribute to the higher amount of ACS2 protein. The abundance of mRNA is governed by a combination of transcription, splicing, and turnover. A398G mutation in acs2-1 is located at the 5′ splice site of second-exon and third-intron junction. The presence of G at splice junction is predicted to confer more efficient mRNA splicing than WT (Hebsgaard et al., 1996), which in turn may boost ACS2 transcript levels. A large-scale analysis revealed that synonymous alleles bearing guanine or cytosine at the third position of a codon increased mRNA half-life (Duan et al., 2013). Considering A398G (K=) is a synonymous mutation, it may have enhanced half-life of acs2-1 mRNA. During the climacteric phase of tomato ripening, ethylene upregulates ACS2 transcripts in a positive feedback fashion (Alba et al., 2005). It is plausible that higher ethylene emission from acs2-1 fruits in an autocatalytic fashion boosted ACS2 transcript levels, which in turn led to a higher level of ACS2 protein. Available evidence points that enhanced transcript levels of acs2-1 and a likely increase in protein stability seem to contribute to the increased ACS2 protein level.
Considering acs2-2 harbors two mutations in the promoter, the lower level of ACS2 protein seems to be a consequence of reduced ACS2 transcript level in acs2-2 fruits. These mutations most likely affect transcript levels of ACS2 by perturbing interactions between transcription factors and their binding sites. Among transcription factors interacting with mutated sites, AZF is a known repressor of ABA-mediated gene expression (Kodaira et al., 2011). It is plausible that the gain of the AZF binding site due to T-106A mutation in the acs2-2 promoter may reduce its expression. Besides, mutation at T-382A may hinder demethylation during ripening, thus affecting acs2-2 transcript levels. Anyhow, contrasting transcript, ACS2 protein levels, ACS enzyme activity, and ethylene emission support the notion that acs2-1 is a hypomorphic enhanced-expression, and acs2-2 is a reduced-expression mutation.
acs2 mutants also influence other phytohormones in fruits
While ethylene is considered a master regulator of tomato ripening, it may crosstalk and influence other hormones during ripening. Consistent with the above, hormone profiling revealed alteration in other phytohormones’ levels in acs2 mutants. acs2-1 displayed higher levels of zeatin, ABA, JA, MeJA, and SA. Downregulation of ABA and SA in acs2-2 conforms to its opposite action. Higher auxin levels in acs2-2/acs2-2 (MG, TUR) fruits are consistent with reported auxin/ethylene antagonism during the early phase of tomato ripening (Su et al., 2015). It is reported that ABA acts as a positive regulator of ethylene biosynthesis at the onset of ripening in tomato (Zhang et al., 2009). Our results indicate that ethylene acts as a positive regulator of ABA in ripening fruits, as ABA levels are higher in acs2-1 and lower in acs2-2. Increase in the ABA level is probably mediated by the upregulation of NCED1 expression in acs2-1, a gene involved in ABA biosynthesis. Considering that both JA and MeJa levels are higher in acs2-1 and MeJa levels are lower in acs2-2/acs-2-2, there seems to be a positive correlation between ethylene and JA levels in tomato fruits. In conformity with this, jasmonate-deficient tomato mutants show lower ethylene emission during fruit ripening (Liu et al., 2012). Interaction of ethylene with phytohormones also seems to be organ-specific, as evidenced by lower zeatin levels in acs2-1 leaf, whereas fruits have higher zeatin levels. Whilst ethylene modulates hormonal level in ripening fruits, the interrelationship among these hormones is complex, and likely involves the intersection of various regulatory pathways (Breitel et al., 2016).
ACS2 mutants show high lycopene level in fruits
The onset of fruit-specific carotenogenesis is closely linked with ethylene biosynthesis as antisense suppression of ACS2 (Theologis et al., 1993) or defect in ethylene perception in Nr mutant (Lanahan et al., 1994) inhibits characteristic red coloration of tomato fruits. Higher carotenoids level in acs2-1 is consistent with the linkage between ethylene emission and carotenoid accumulation. Analogous to influence on the hormonal level, acs2-1 in a semi-dominant fashion boosted carotenoid levels in heterozygous acs2-1 plants. Ethylene seems to stimulate carotenoid levels in acs2-1 by upregulating transcripts of PSY1, PDS, and CRTISO, the key genes regulating phytoene and lycopene biosynthesis. Surprisingly, though most carotenoid intermediates and key genes transcripts such a PSY1 and CRTISO were downregulated in acs2-2, yet it accumulated a higher level of lycopene. In tomato fruits, carotenoid biosynthesis is geared towards lycopene accumulation, which is sequestered and stored in plastoglobules and lycopene crystals in chromoplasts (Kilambi et al., 2013; Nogueira et al., 2013). It is likely that prolonged transition period from MG to RR stage in acs2-2 facilitated continued synthesis and sequestration of lycopene, in turn leading to higher lycopene levels despite having reduced transcript levels of carotenogenic genes. This view is also substantiated by higher lycopene accumulation in the introgressed progeny of acs2-1, acs2-2 in AV, and PED cultivars.
ACS2 mutants show contrasting effects on metabolite profiles
Consistent with the pleiotropic effect of ethylene, levels of several metabolites in fruits were altered in acs2-1 and acs2-2. Different PCA profiles of acs2-1, acs2-2, and WT are consistent with the view that hormonal signaling perturbation strongly influences metabolite levels (Bastías et al., 2014). Though acs2-2 and Nr mutant are genetically distinct, a comparison of acs2-2 and Nr metabolite profiles (Osorio et al., 2011) showed a similar shift in levels of at least 13 metabolites from MG to RR stage including key TCA cycle constituents such as citric acid, isocitric acid, and malic acid. This entails that the above similarity in metabolites between Nr and acs2-2 likely emanates from reduced ethylene-mediated signal transduction in respective mutants.
Reduced CO2 emission from acs2-2 fruits seems to be linked with reduced ethylene emission (Tigchelaar et al., 1978). Higher levels of amino acids in acs2-2 may be linked with reduced respiration diverting glycolysis/TCA cycle intermediates to amino acid synthesis. Likewise, reduced ACS2 activity in acs2-2 likely leads to high 5-oxoproline, which is part of the reactions needed to close the methionine salvage cycle in plants (Ellens et al., 2015). Low levels of 2-ketoglutarate in acs2-2, where 5-oxoproline accumulates, and high level in acs2-1, where 5-oxoproline does not accumulate, indicates that methionine salvage cycle differently operates in acs2-1 and acs2-2, to prevent the accumulation of α-ketoglutaramate, a toxic molecule (Cooper, 2004).
Our study highlights that perturbation in the function of a single gene distinctly influences a wide range of metabolites. Importantly, metabolic shifts associated with acs2-2 and acs2-1 is largely retained in their backcrossed homozygous progeny. Likewise, in backcrossed homozygous WT, metabolic profiles revert to parental WT. Studies on the genetic control of metabolism have underscored QTLs as major determinants of metabolic shifts and profiles (Fernie and Tohge, 2017). Co-segregation of metabolite profiles with respective acs2 mutations conforms with the above view. Importantly, it also highlights that an increase/decrease in ethylene levels in mutants has a cascading effect on plant metabolism, which in turn, accelerates/decelerates the ripening process.
In conclusion, our results highlight that besides a pivotal role in fruit ripening, ACS2 participates in several facets of tomato development. Faster germination of acs2-1 implies the role of ACS2 in seed germination. Contrasting vegetative and reproductive phenotypes of mutants, including leaf/fruit senescence, indicate a role of ACS2, as a key ethylene biosynthesis enzyme, during tomato development. Faster transition to the MG stage in acs2-1 mutant suggests the role of ACS2 in early fruit development. Our study also entails that alleles of ACS2 can be potentially used for modulating on-vine tomato ripening.
Materials and Methods
Mutant isolation
Two EMS-mutagenized populations of tomato (Solanum lycopersicum) cultivars, M82 (Menda et al., 2004), and Arka Vikas were used for screening of mutants. Genomic DNA was isolated from 8-fold pooled plants (Sreelakshmi et al., 2010), and mutation detection was carried out on the LICOR-4300 DNA analyzer (TILL et al., 2006) (Supplemental Table 9). The presence of mutated gene copy and its zygosity in backcrossed plants AV, M82, and PED was monitored by the CEL-I endonuclease assay (Mohan et al., 2016).
Characterization of mutants
The seeds sterilized with 4% (v/v) NaOCl were sown on filter papers, and germination was visually monitored. For phenotype studies, after sowing on agar (0.8% w/v), seedlings were grown in light (100 µmol/m2/s) or darkness. For ethylene-mediated growth inhibition, seedlings were grown in darkness in airtight plastic boxes, and at the onset of the experiment, a known volume of ethylene was injected in the box.
Plants were grown in the greenhouse under natural photoperiod (12-14h day, 28±1°C; 10-12h night, 14-18°C). For senescence, pigments, hormonal levels, and ethylene emission studies, leaves were harvested from the seventh node of 45-day-old plants. The floral and inflorescence morphology were from the second and third truss. For senescence study, the leaves laid on moist filter papers in Petri plates were placed under white light (100 µmol/m2/s), or darkness. The fruit development was monitored on-vine from the day of post-anthesis through different ripening stages until the fruit skin wrinkled. The firmness of detached fruits, pH, and °Brix was measured as described in Gupta et al. (2014).
Estimation of ethylene and CO2 emission
The ethylene emission was monitored using a previously described procedure (Kilambi et al., 2013). The seedlings were transferred to an airtight glass vial on 0.8% (w/v) agar for 24h. For leaves, the detached leaves were placed on moistened filter paper and enclosed in airtight Petri plates for 4h. For ethylene emission, WT and mutant fruits were harvested from the second truss. The fruits were placed in an airtight container for 4h. At the end of the above-mentioned periods, one mL of headspace was withdrawn from the respective containers to estimate ethylene. For CO2 emission, the fruits were enclosed in a chamber containing the CO2 sensor (Vernier, OR, USA) for 10 min. After that, the CO2 emission was monitored using preinstalled Graphical AnalysisTM software.
Estimation of ACC levels, ACS and ACO activity
The ACC, MACC levels, and ACS, ACO activity, were estimated using closed vials and calculated using Bulens et al. (2011) protocol. For ACC estimation, the fruit tissues were homogenized in 5% (w/v) sulphosalicylic acid, followed by 10 mM HgCl2 addition and release of ethylene by addition of 4% (w/v) NaOCl and saturated NaOH (2:1, v/v). To determine ACS (1gm tissue) and ACO (500mg tissue) activity, fruit tissue was homogenized in liquid N2 and mixed with the extraction buffer by vortexing. The supernatant obtained after centrifugation (21,000g) was desalted on the Sephadex-G25 column. The eluted fraction was assayed for ACS activity in a reaction mixture containing S-adenosyl-L-methionine. The reaction was terminated by adding HgCl2, and ethylene was released by adding NaOCl/NaOH mixture. The ACO activity was estimated in a reaction mixture containing 1 mM ACC. For all of the above, at the end of the reaction, one mL headspace was withdrawn for ethylene estimation.
Western blotting
The polyclonal antibodies were raised in rabbits against ACS2 protein using a peptide “EHGENSPYFDGWKAYDSD” custom synthesized by PEPTIDE-2.0, USA. The Ig fraction was purified using DEAE-Sepharose column using standard protocols. The fruits were homogenized in Tricine buffer (200 mM, pH 8.5), 2 mM pyridoxal-5-phosphate, 10 mM DTT with 50 mg PVPP. The homogenate was centrifuged (21,000g, 4°C) and the supernatant was desalted on Sephadex G25 column. Given the very low abundance of ACS2 protein, the eluted sample was immunoprecipitated after protein estimation (Bradford, 1976). The supernatant was gently shaken with purified IgG fraction at 37°C for 1h, followed by 1h shaking with protein A-Sepharose at 4°C. The beads were recovered by centrifugation (12,000g, 4°C). The supernatants were analyzed for the reduction in ACS2 activity, and protein A-Sepharose beads were used for Western blotting.
The immunoprecipitated ACS2 protein was recovered by boiling in SDS-PAGE loading buffer and separated in 12% gel following Laemmli (1970) protocol. The gel was electroblotted on the PVDF membrane, followed by Western blotting (Towbin et al., 1979). The non-specific binding sites on the membrane were blocked by incubating with milk powder. The membrane was first incubated with ACS2-antibody (1/1500 dilution), followed by incubation with an anti-rabbit IgG goat antibody (1/80,000 dilution), coupled with alkaline phosphatase. The alkaline phosphatase activity was visualized using standard BCIP-NBT assay.
Profiling of phytohormones, carotenoids, and metabolites
The phytohormone levels were determined from leaves from the 7th node of 45-day-old plants and fruits using Orbitrap Exactive-plus LC-MS following the protocol described earlier (Pan et al., 2010; Bodanapu et al., 2016). Carotenoid profiling was carried out following the procedure of Gupta et al. (2015). Metabolite analysis by GC-MS was carried out by a method modified from Roessner et al. (2000) described in Bodanapu et al. (2016).
Real-time PCR
The RNA was isolated from fruit pericarp using TRI reagent (Sigma), and 2 µg DNase-treated RNA was reverse transcribed using a cDNA Synthesis Kit (Agilent), following the respective manufacturer’s protocol. RT-PCR was performed using iTaq Universal SYBR Green Supermix (BioRad) (Supplemental Table S10). The relative differences were determined by normalizing Ct values of each gene to the mean expression of β-Actin and Ubiquitin genes and calculated using the equation 2−∆Ct.
In silico characterization
The target protein structure and ACS2 sequence (PDB ID: 1IAX) (Huai et al., 2001) (https://www.rcsb.org/structure/1IAX), having 2.8Å resolution was retrieved from protein data bank (http://www.rcsb.org/pdb/). The 3-dimensional structure of ACS2-1 (V352E) mutant and WT proteins were visualized using PyMol (http://www.pymol.org). The stabilization of ACS2-1 versus WT protein was analyzed using the following software CUPSAT (http://cupsat.tu-bs.de/), MAESTROWeb (https://pbwww.che.sbg.ac.at/maestro/web), PoPMuSiCv3.1 (https://soft.dezyme.com/query/create/pop), STRUM (https://zhanglab.ccmb.med.umich.edu/STRUM/), and DynaMut (http://biosig.unimelb.edu.au/dynamut/). Promoter analysis was performed using PCbase (http://pcbase.itps.ncku.edu.tw/) and manually for ripening-specific transcription factors.
Statistical analysis
All results are expressed as mean±SE of three or more independent replicates. The StatisticalAnalysisOnMicrosoft-Excel (http://prime.psc.riken.jp-/MetabolomicsSoftware/StatisticalAnalysisOn-MicrosoftExcel/) was used to obtain significant differences between data points. Heat maps and 3D-PCA plots were generated using Morpheus (https://software.broadinstitute.org/morpheus/) and MetaboAnalyst-4.0 (https://www.metaboanalyst.ca/), respectively. The statistical significance was determined using Student’s t-test (* for P≤0.05, # for P≤0.01 and $ for P≤0.001).
Supplemental Data
Supplemental Figure S1. Splice site analysis of acs2-1 mutant by NetGene2 server.
Supplemental Figure S2. Chlorophyll, carotenoids, and xanthophylls levels in the leaf of WT, acs2-1, and acs2-2 mutant.
Supplemental Figure S3. Fruit firmness, total soluble solids (Brix) and pH of WT, acs2 mutants, and their BC1F2 progenies at different ripening stages
Supplemental Figure S4. On-vine ripening, ethylene emission, and lycopene levels in BC4F2 progeny (acs2-1/acs2-1), (acs2-2/acs2-2) in AV and PED cultivars.
Supplemental Figure S5. Raising and validation of antibodies specific for ACS2 peptides.
Supplemental Figure S6. Immunoprecipitation and Western blotting of ACS2 protein using purified IgG fraction.
Supplemental Figure S7. The metabolic shifts in acs2 mutant fruits during ripening in comparison to WT.
Supplemental Table S1. EMS-mutagenized tomato populations used for the isolation of ACS2 mutants.
Supplemental Table S2. ACS2 mutant lines identified and confirmed for mutation.
Supplemental Table S3. The genetic segregation of acs2-1 and acs2-2 mutants in BC1F2 generation.
Supplemental Table S4. Increases in ACS2-1 protein stability predicted by different software.
Supplemental Table S5. The alteration of bonding pattern in ACS2 protein in acs2-1 mutant compared to WT protein.
Supplemental Table S6. Disruption in transcription factor binding site in acs2-2 due to promoter mutation.
Supplemental Table S7. Changes in the methylation status of ACS2 promoter at −106 and −382 position during fruit ripening.
Supplemental Table S8. Comparisons of flower numbers and fruit set in WT and acs2 mutants.
Supplemental Table S9. List of primers used for screening for mutations in ACS2 by TILLING
Supplemental Table S10. List of genes and the primers used for qRT-PCR analysis.
Supplemental dataset S1. The transcription factor binding sites in the ACS2 and acs2-2 promoter predicted by PCbase.
Supplemental dataset S2. Changes in the methylation status of ACS2 promoter during fruit ripening.
Supplemental dataset S3. List of metabolites identified in leaves of WT and acs2 mutants.
Supplemental dataset S4. List of metabolites identified at different fruit ripening stages of WT and acs2 mutants.
DECLARATION OF INTERESTS
The authors declare no competing interests.
ACKNOWLEDGMENTS
We thank Dr. Dani Zamir for EMS-mutagenized M82 cultivar lines and Dr. Alok Sinha for Phospho-Ser antibody. S.G. was recipient of DST Young Scientist grant and K.S. was recipient of UGC-JRF. S.S. gratefully acknowledges the financial support from the DBT-RA Program in Biotechnology and Life Sciences.
Footnotes
One sentence summary: Genetic and functional analysis of ACS2 mutants reveal high/low ethylene emission oppositely modulates developmental processes and metabolite profiles in tomato.
Funding: This work was funded by Department of Biotechnology, India grants (BT/PR/5275/AGR/16/465/2004, BT/PR11671/PBD/16/828/2008, BT/PR/7002/PBD/16/1009/2012, and BT/COE/34/SP15209/2015) to RS and YS.
The version of manuscript is revised to update the title and an additional supplemental file, and deletion of some information, not related to the core theme of the paper.