Introduction

Phytohormones influence distinct biological and developmental processes in plant ranging from seed germination to fruit ripening, and final seed dispersal (Goossens et al. 2016). Plant hormones such as jasmonate (JA), including jasmonic acid, methyl jasmonate (MeJA), and oxylipin derivatives control various developmental processes, including root formation (Wasternack and Hause 2013), cell cycle (Pauwels et al. 2008), resistance to pathogen invasion (Farmer et al. 2003), mechanical injuries (Devoto et al. 2005), leaf senescence (Weidhase et al. 1987), seed fertility and development (Xie et al. 1998; Qiang et al. 2014), and sex determination (Acosta et al. 2009). JAs are synthesised from linolenic acid, a precursor in JA biosynthesis. The beta-oxidation of linolenic acid occurres in six steps, each catalysed by various enzymes such as allene oxide synthase (AOS), 12‐oxo‐phytodienoate reductase 3 (OPR3), and lipoxygenase (LOX) (Creelman and Mullet 1997; Wasternack and Hause 2013). JA signalling is mediated by a F-box component E3 ubiquitin‐ ligase SKP1‐Cullin‐F‐box complex (SCFCOI1), CORONATINE INSENSITIVE 1 (COI1) (Yan et al. 2009), which ultimately lead to degradation of JAZ protein by 26S proteasome (Devoto et al. 2002; Xu et al. 2002; Chico et al. 2008).

JA responses have been well characterised through mutants analysis in Arabidopsis. For instance, Arabidopsis triple mutants (fad3-2, fad7-2, coi1) (Feys et al. 1994; McConn and Browse 1996) show reduced sensitivity to JA. The Arabidopsis JA-insensitive mutant coi1 show pleiotropic phenotypes, including leaf yellowing (Castillo and Leon 2008; Reinbothe et al. 2009), susceptibility to pathogen invasion (Sanders et al. 2000). Moreover, Arabidopsis COI1 involved apical dominance (Kim et al. 2013), root growth inhibition (Adams and Turner 2010), male sterility (Feys et al. 1994), and in inositol polyphosphates in wound signalling (Mosblech et al. 2011). Tomato and Arabidopsis have a single copy of the COI1 gene each, but other plant species have more than one homologue, such as rice with three homologues (OsCOI1a, OsCOIb, OsCOI2) (Lee et al. 2015). The tomato SlCOI1 share more than 68% sequence identity with Arabidopsis counterpart but show less than 57% sequence similarity with rice SOCI1 (Li et al. 2004). To date, the physiological role of COI1 in tomato is remained to be determined. In Solanum nigrum, COI1 involved in jasmonate metabolism and systematic protection against insect attacks (VanDoorn et al. 2011). In tomato, COI1 mediate signal perception JA is pivotal for steroidal glycoalkaloids (SGA) via a JA-responsive ETHYLENE RESPONSE FACTOR (ERF) transcription factor (JRE4) (Abdelkareem et al. 2017). However, in rice, OsCOI1a regulates a basic helix–loop–helix (bHLH) transcription factor 148 expression by SCFCOI1 complex, leading to enhance tolerance to drought stress (Seo et al. 2011), but OsCOI1b resulted in leaf yellowing (Lee et al. 2015).

Tomato is a typical model plant for studying genetics and developmental physiology because of its small genome and short life cycle (Meissner et al. 2002). However, little is known about the molecular characteristics of COI proteins in tomato. Here, a comprehensive bioinformatic analysis of the COI gene family in the whole genome of tomato was performed. A total of nine putative genes encoding COI proteins were identified in tomato. The results of this study display the bioinformatic characteristics of COI family proteins in tomato (SlCOIs), phylogeny, subcellular localization, and tissue-specific expression profile. We also investigated abiotic stress-induced expression analysis. Our data provide a handy reference for the functional analysis of the COI family in tomato.

Materials and methods

Tomato COI genes discovery and physicochemical characteristics

The Arabidopsis COI protein sequences were retrieved from TAIR database (https://www.arabidopsis.org/) (Reiser et al. 2017) and used as a query against the Solanaceae Genomics Network (SGN, https://www.solgenomics.net/) (Fernandez-Pozo et al. 2015) for tomato with default parameters. Moreover, the COI Pfam domain pattern was retrieved from the Pfam database (Finn et al. 2008). The COI proteins of tomato (SlCOIs) were predicted using a hidden Markov model (HMM) profile retrieved from the Pfam database (Finn et al. 2008). The S. lycopersicum COI protein sequences were searched using the HMMSEARCH program (Finn et al. 2011). All redundant COI sequences were excluded and validated whether candidate members of genes processes the COI domain in SMART (Schultz et al. 1998) and NCBI CDD (Marchler-Bauer et al. 2017). The characteristics of COI genes, including molecular weight (MW, kDa), the grand average of hydropathy (GRAVY), and theoretical isoelectric point (pI), were calculated using the sequence manipulation suite (SMS, http://www.bioinformatics.org/sms2/index.html) (Stothard 2000). All the genes were named in their chronological order on the chromosomes.

Subcellular location prediction of selected SlCOIs

WoLFPSORT program (https://wolfpsort.hgc.jp/) (Horton et al. 2007) was used to predict the in silico subcellular location of SlCOI proteins. To validate in silico predicted subcellular localization of selected SlCOIs, the full-length nucleotide sequences of SlCOI1, SlCOI2, SlCOI3 and SlCOI8 were cloned in frame with green fluorescent protein (GFP) under control of 35S in pMS4::GFP vector. Empty vector (control) and all SlCOI–GFP constructs were transformed into agrobacterium leaves as described by Li (2011). The green fluorescence was observed through the laser scanning confocal microscope (Leica TCS SP8, Germany) after infected for 3 days.

Phylogenetic analysis and gene duplication of SlCOI family members

To investigate the phylogenetic relationship of tomato COI protein with other plant species, the protein sequences of Physcomitrella patens (Pp), Arabidopsis thaliana (At), Populus trichocarpa (Pt), Sorghum bicolor (Sb), Oryza sativa (Os), Zea mays (Zm), Brachypodium distachyon (Bd), Selaginella moellendorffii (Sm), Hordeum vulgare L. (Hv), and Triticum urartu (Tu) were acquired from phytozome (https://phytozome.jgi.doe.gov/pz/portal.html) (Goodstein et al. 2012). An unrooted neighbour-joining (NJ) (Saitou and Nei 1987) dendrogram was constructed in MEGAX (Kumar et al. 2018) with bootstrap set at 1000 replicates. MCScanX program (https://github.com/wyp1125/MCScanX) was used to predict SlCOI gene duplication events in the tomato genome. The nonsynonymous (Ka), synonymous (Ks) nucleotide substitution rates, and the Ka/Ks ratios were predicted using k-estimator (http://en.bio-soft.net/format/KEstimator.html) (Comeron 1999). The divergence time (T, millions year ago (mya)) was calculated as follows: T = Ks/2y (y = 1.5 × 10−8) (Koch et al. 2000).

Gene structure analysis, conserved motif scan, and cis-regulatory motif prediction

The number and distribution of exon–intron in SlCOIs were visualised by submitting nucleotide coding sequences and corresponding genomic sequences to Gene Structure Display Server 2.0 (http://gsds.gao-lab.org/Gsds_about.php) (Hu et al. 2015). MEME suite (https://meme-suite.org/meme/tools/meme) (Bailey and Elkan 1994) was used to scan conserved motifs in COI proteins with default parameters. A 1-kb long 5′UTR nucleotide sequence from the start codon was extracted for each SlCOI gene from SNG and submitted in the PlantCARE database (http://bioinformatics.psb.ugent.be/webtools/plantcare/html/) (Lescot et al. 2002) for cis-regulatory elements prediction.

Plant material, abiotic stress and phytohormone treatment

The tomato plant cultivar Micro-Tom was grown in a greenhouse under control conditions: 14 h light / 12 h dark photoperiod, at 25°C / 20°C day/night temperature with relative humidity between 70% and 80%, and photon density of about 120 µmol photons m−2s−1. Six-week-old tomato seedling were used for tissue/organ-specific expression and abiotic stress treatments. For tissue/organ-specific expression, various plant parts including, root, shoot, leaves, flower bud, fully opened flower, and fruit at various developmental stages were harvested. The fruit development stages include 1 cm, 2 cm, 3 cm, mature green (MG), breaker (B), and 10-days breaker (B+10) fruits. All the samples were collected in triplicate and mixed thoroughly.

For salinity, drought and phytohormone stresses, 6-week-old tomato seedlings were treated with 200 mM NaCl, cold (4°C), 4% PEG and 0.01 mM ABA, GA3, SA, MeJA, and IAA following Waseem et al. (2018). The control plants were treated with fresh water. Seedlings were harvest at 0h, 3h, 6h, 12h and 24h after treatment. Three independent biological replicates were collected, and six seedlings were used for each treatment. All the samples were immediately frozen in liquid nitrogen and stored at −80°C until further analysis.

cDNA preparation and reverse transcription quantitative real-time polymerase chain reaction (RT-qPCR) analysis

Total RNA was extracted from collected samples using TRIZOL reagent following the manufacture's protocol. The RNA was quantified using nanodrop (Thermo USA), and quality was determined through 2% (w/v) stained agarose gel electrophoresis. The first complementary DNA strand was synthesised using Prime Script RT reagent Kit with gDNA Eraser (Takara, Japan). SYBR-Premix Ex Taq-II (TliRNaseH Plus) on CFX96 touch real-time PCR detection system (Bio-Rad, USA) was used to conduct RT-qPCR. The relative expression was normalized using the housekeeping gene of SlUBQ (Solyc01g056940). We first calculated the amplification efficiency of all the primer pairs analysed in this study (figure 1 in electronic supplementary material at http://www.ias.ac.in/jgenet/). Finally, the relative expression was calculated following 2−ΔΔCt method (Livak and Schmittgen 2001).

Statistical analysis

A Student's t-test was used to determine the statistical significance of transcript levels between treatments and their corresponding controls. Differential expression data were statistically significant and were marked by a P-value of *P < 0.05, ** P < 0.01, and *** P < 0.001.

Results

Characteristics of tomato COI gene family

In this study, we used Arabidopsis amino acid sequences (AtCOIs) for BLASTP and HMM searches to screen COI members in the genome of tomato. Further, the presence of conserved COI characteristic F-box domain was confirmed by NCBI CDD-Search and SMART. These nine SlCOI genes were subsequently named as SlCOI1 to SlCOI9 according to their genomic locus. Detailed physicochemical analysis of putative SlCOI proteins revealed that proteins peptide length and predicted MW ranged widely from 421 (aa)/74.12 kDa (SlCOI6) to 675 (aa)/73.78 kDa (SlCOI9). The pI ranged from 4.9 (SlCOI6) to 7.76 (SlCOI1). It was found that some COI proteins include SlCOI2, SlCOI3, SlCOI4, SlCOI6 and SlCOI8 was hydrophilic due to negative GRAVY values, while others were hydrophobic. In silico subcellular localization prediction revealed that SlCOIs were localized in three organelles: nucleus, cytoplasm, and chloroplast. SlCOI1, SlCOI2, SlCOI3, SlCOI6 and SlCOI8 were in the nucleus; SlCOI4 and SlCOI9 were found in the cytoplasm; SlCOI5 and SlCOI7 were in the chloroplast (table 1). The tomato SlCOI proteins were distributed in seven chromosomes. SlCOI1, SlCOI2, SlCOI3, SlCOI4, SlCOI8 and SlCOI9 were scattered on chromosomes 1, 2, 4, 5, 9 and 10, respectively. Three SlCOI: SlCOI5, SlCOI6 and SlCOI7 was distributed on chromosome 6 (figure 1). Only a single pair of SlCOI genes (SlCOI5-SlCOI8) displayed segmental duplication while tandem duplications were absent (figure 1). Based on a substitution rate, the duplication events for (SlCOI5-SlCOI8) segmental duplication was estimated to have occurred ~27.8 mya (table 1 in electronic supplementary material at http://www.ias.ac.in/jgenet/).

Table 1 Physiochemical feature of SlCOI genes in tomato genome.
Full size table
Figure 1
figure 1

Chromosomal location and synteny plot of nine COI genes in tomato chromosomes. A segmental duplication pair is marked in red.

Full size image

Phylogeny of SlCOI gene family

To insight into the phylogenetic relationship of tomato SlCOI proteins with COIs from different plant species, an unrooted NJ tree revealed that all COI proteins were clustered into eight distinct clades. The tomato COI genes were clustered in five clades with COI proteins from P. patens (Pp), A. thaliana (At), P. trichocarpa (Pt), S. bicolor (Sb), O. sativa (Os), Zea mays (Zm), B. distachyon (Bd), S. moellendorffii (Sm), H. vulgare L. (Hv), and T. urartu (Tu) indicating high sequence similarity with COI proteins of these plant species (figure 2). SlCOI8, SlCOI5, and SlCOI6 were clustered in group II, SlCOI1 in III, SlCOI3 in V, SlCOI4 in VI, SlCOI7, SlCOI1 and SlCOI9 in group VII. Similarly, we found that COI proteins from different species were also distributed in different groups. For example, Arabidopsis COI proteins clustered in groups I, II and V but only maize in group VI. Moreover, rice COI proteins grouped in I, II, IV, V, and VI while P. patens in group III and VI only (figure 2), suggesting that COI proteins exhibited differences in evolution among different species.

Figure 2
figure 2

Phylogeny of COI genes. A NJ phylogenetic tree of COI proteins from Physcomitrella patens (Pp), Arabidopsis thaliana (At), Populus trichocarpa (Pt), Sorghum bicolor (Sb), Oryza sativa (Os), Zea mays (Zm), Brachypodium distachyon (Bd), Selaginella moellendorffii (Sm), Hordeum vulgare L. (Hv), and Triticum urartu (Tu) including tomato (Sl, S. lycopercium) was divided into eight groups (I to VIII). The bootstrap was set at 1000 replicates.

Full size image

Gene structure, conserved motifs, and cis-regulatory elements in SlCOIs

To understand the evolution of the SlCOI gene family, the numbers and distribution of intron/exon in putative tomato COI genes were analysed using corresponding coding and genomic sequences. It was revealed that the exon number in all members of the COI gene family ranged from three to eight. Five exon and four introns were present in SlCOI5 and SlCOI6, while eight exon and seven introns in SlCOI1, SlCOI7 and SlCOI9. We found that the position and number of exons and introns in genes belonging to the same class or subclass were similar. This finding supports the phylogenetic relationship of COI family genes (figure 2 and figure 2 in electronic supplementary material and that exon gain and loss have occurred in the SlCOI gene family. The protein architecture of COI proteins from different plant species, including tomato in the MEME program revealed that most of the COIs possess 10 motifs in groups III, V and VI. COI proteins in group I have 10 motifs; II and IV containes eight motifs.

Moreover, five and six motifs were predicted in groups VIII and VII, respectively (figure 3). In plants, cis-regulatory elements play an essential role in regulating gene expression under normal and response to stress stimuli. To study the response of the SLCOI gene to various signal factors, we searched 1 kb sequences upstream of the start codon of the SlCOI family for stress response elements. Figure 3 in electronic supplementary material shows hormone-responsive elements associated with abscisic acid (ABRE), ethylene (ERE), auxin (TGA-element), gibberellins (TATC-box), salicylic acid (TCA-element), and jasmonate (CGTCA/TGACG-motif) were identified in the promoter region of SlCOIs. Further, stress-responsive regulatory elements, defence/stress responses (TC-rich repeats), low-temperature responses (LTR), associated with drought inducibility (MBS, MYB), and wound responsive elements (WUN-motif) were identified (figure 3 in electronic supplementary material. These results suggest that the COI family may play a crucial role in the growth and development of tomato and various hormones and stress.

Figure 3
figure 3

MEME predicted conserved motifs in COI proteins from Physcomitrella patens (Pp), Arabidopsis thaliana (At), Populus trichocarpa (Pt), Sorghum bicolor (Sb), Oryza sativa (Os), Zea mays (Zm), Brachypodium distachyon (Bd), Selaginella moellendorffii (Sm), Hordeum vulgare L. (Hv), and Triticum urartu (Tu) including tomato (Sl; S. lycopercium). Thus, a total of 10 motifs (1 to 10) were identified in COI proteins.

Full size image

Subcellular localization of SlCOI proteins

Prediction analysis indicated that SlCOIs exhibit various patterns of subcellular localization (table 1). Subcellular localization implied the working position of a protein and is essential for functional gene characterization. To further determined the subcellular localization of SlCOI proteins, GFP fused with selected SlCOI proteins was transiently expressed in tobacco leaf. Consistent with the in silico predicted localization using bioinformatics approaches, as shown in figure 4; four COI proteins: SlCOI1, SlCOI2, SlCOI3 and SlCOI8, localized in the nucleus.

Figure 4
figure 4

Subcellular localization analysis of SlCOIs proteins. Tobacco leaves transiently expressed SlCOI–GFP fusion proteins that observed through the laser scanning confocal microscope. Scale bars represent 50 μm.

Full size image

Expression analysis of COI protein family members in different parts of the tomato plant

To elucidate the role of COI protein in plant development, the expression profile analysis in various plant parts, including root, leaf, flower at the bud and fully opened stage and fruits at various developmental stages was investigated (figure 5). It was observed that some genes expressed at relatively high levels in root include SlCOI1, SlCOI3, SlCOI4, SlCOI7 and SlCOI9, indicating that these genes might play a developmental or physiological role in root tissues. Further, some genes had higher transcript abundance at a specific developmental stage concerning fruit development, such as SlCOI expressed specifically in 3 cm fruit, SlCOI4 in fruit at breaker stage, SlCOI6 at 1 cm and SlCOI8 in 3 cm fruit. Additionally, it was observed that the expression of SlCOI1, SlCOI2, SlCOI3 and SlCOI8 was decreased with the fruit development, and the onset of fruit ripening indicated that these genes might involve in tomato fruit development and ripening (figure 5). Further, we found that promoter sequences of these genes contained ethylene-responsive elements (figure 3 in electronic supplementary material and may involve in ethylene-dependent tomato fruit ripening.

Figure 5
figure 5

Tissue-specific expression analysis of tomato Pht1 genes. The RNA was prepared from different tissues including root, FB, flower bud; F, flower; leaves: 1cm_F, 2cm_F, 3cm_F; _F, fruit, MG_F, mature green; B_F, breaker fruit; B+10_F, 10 days breaker fruit. The relative expression levels of each of the tomato COI genes were indicated as a percentage of the constitutive SlUBQ expression activity. Each bar was the mean of five biological replications with standard error.

Full size image

The inducible expression analysis of SlCOIs under salt, drought, and cold

To study the potential roles of SlCOI in abiotic stress responses, we performed expression analysis of SlCOIs under various abiotic stress conditions using quantitative RT-qPCR. The RT-qPCR data from these stresses at different time intervals (0h, 3h, 6h, 23h, and 24h) is shown in figure 6, demonstrating that SlCOI genes significantly responded to different abiotic stress treatments. For salt stress, the expression abundance of some genes show temporal expression while suppressed for others (figure 6a). For instance, SlCOI1, SlCOI3, SlCOI5, SlCOI6 and SlCOI9 were upregulated at all points. SlCOI4, SlCOI8 and SlCOI7 had similar expression patterns, with the highest expression at 3h and 6h points. SlCOI2 was upregulated at 6h (figure 6a). SlCOI1, SlCOI3, SlCOI6 and SlCOI8 were induced under drought stress at all point intervals except for SlCOI1 and SlCOI8 downregulated at 24h. However, expressions of SlCOI4, SlCOI5, SlCOI7, and SlCOI9 was upregulated at 3h but showed a gradual decrease in expression until 24h. SlCOI5 showed strong expression at 6h but downregulated in 12 and 24h (figure 6b). SlCOI2 was upregulated at all time interval, SlCOI1, SlCOI5 and SlCOI7 were upregulated at 12 h intervals, but at 6h SlCOI4 was peak under cold conditions. SlCOI1 peaked at 12h; SlCOI3, SlCOI8 and SlCOI9 had peaked expression at 24h (figure 6c). It was observed that SlCOI6 expressions were peaked at all point against all stresses (salt, drought and cold); SlCOI showed similar expression under salt and drought, which were downregulated at 3h and 24h intervals; opposite trends were observed under drought stress. Similarly, SlCOI9 showed ascending expression under salinity (figure 6a), but descending trends were observed under drought (figure 6b). The extensive involvement of SlCOI genes in response to abiotic stresses implied their potential essential functions.

Figure 6
figure 6

Expression profile of tomato COI genes under (a) 200 mM salt (b) drought (4% PEG), (c) cold (4°C), 0.01 mM, and (d) ABA. Six-week-old tomato seedling was selected for treatments. The fold changes in gene expression were obtained according to the 2−ΔΔCt method. Data are presented as mean and SD. P values obtained from pairwise comparisons using student t-test are shown (*P < 0.05, **P < 0.01, ***P < 0.001).

Full size image

Phytohormone induced expression profiling of tomato COI genes

The responsiveness of SlCOI genes to different hormones such as ABA, IAA, GA3, SA, and MeJA was investigated by RT-qPCR. For ABA stress, SlCOI1, SlCOI2 and SlCOI3 upregulated at 12h and 24h intervals; SlCOI4 and SlCOI5 upregulated at 24h. SlCOI8 peaked at 3h, 6h and 12h, while SlCOI7 had peaked expressions at 6h and 24h. SlCOI6 was upregulated at 6h but SlCOI9 peaked at 24h (figure 6d). SlCOI1, SlCOI2 and SlCOI3 showed peaks at 24h under SA stress. SlCOI7 decreased with time while SlCOI6 opposite trends till 12h then downregulated at 24h. SlCOI5 upregulated at 3h; SlCOI4 expressions was peaked at 6h; SlCOI8 and SlCOI9 upregulated at 3h, and 24h were upregulated (figure 7a). For GA treatment, SlCOI2, SlCOI6 and SlCOI8 were induced with strong expression at 12h and 24h; SlCOI7 upregulated at 12h, but SlCOI4 and SlCOI5 downregulated. SlCOI1, SlCOI6, showed temporal expression in all points (figure 7b). SlCOI2 and SlCOI8 were upregulated at 24 and 3h points, respectively, against MeJA. SlCOI1, SlCOI5 and SlCOI7 were upregulated at all point intervals; SlCOI4, SlCOI8, and SlCOI9 showed opposite trend (figure 7c). For IAA treatment, most of the genes respond at early point of time. SlCOI1, SlCOI4, SlCOI7 and SlCOI9 were upregulated at all point intervals except at 24h. SlCOI8 was upregulated at 3h and 12h, suppressed at 24h; SlCOI6 upregulated at 3h, SlCOI9 upregulated at 3h but downregulated at later time intervals (figure 7d).

Figure 7
figure 7

Expression profile of tomato COI genes under (a) SA, (b) GA (c) MeJA, and (d) IAA. Six-week-old tomato seedling was selected for treatments. Relative fold changes in gene expression were obtained according to the 2−ΔΔCt method. Data are presented as mean and SD. In addition, P values obtained from pairwise comparisons using student t-test are shown (*P < 0.05, **P < 0.01, ***P < 0.001).

Full size image

Notably, we found that the expressions of SlCOI1 increased after treated with all the mentioned hormones used in this study, while the expressions of SlCOI7 showed a similar pattern between MeJA and IAA treatments but opposite patterns with SA. Expression of SlCOI9 and SlCOI8 showed opposite trends between MeJA and GA also between ABA and MeJA treatments, suggested that these hormones may play an antagonistic effect on expressions SlCOIs. Taken together, the variational expression of SlCOI genes under different plant hormone treatment implied that this gene family involved multiple hormonal signals in a complicated way. The detailed role of this gene family in the crosstalk of plant hormones was thus worth studying and may provide us with new insight into the field.

Discussion

The oxylipin (jasmonate and their derivatives) plays a pivotal role in various aspects of plant growth and development and tolerance against abiotic and biotic stresses as including, pigmentation (Zhang et al. 2009), root growth and development (Cheng et al. 2011), fertility (Turner et al. 2002), fruit growth and development (Kondo et al. 2000), herbivory (Abuqamar et al. 2008), seeds maturation and trichrome formation (Li et al. 2004), and leaf yellowing (Reinbothe et al. 2009). An F-Box protein, COI1, is required for JA mediated responses. COI interacts to form SCF(COI1) E3 ubiquitin ligase complex and trigger jasmonate ZIM-domain (JAZ) proteins for degradation via 26S proteasome (Yan et al. 2009).

The availability of genome sequence technology opens new horizons in functional genomics and identifies new genes annotations (Pfeifer et al. 2014; Liu et al. 2017; Wang et al. 2017; Waseem et al. 2018). In this sense, the COI gene family has been identified and characterized in various plant species such as Arabidopsis (Xie et al. 1998), rice (Lee et al. 2015), Aquilaria sinensis (Liao et al. 2015), and tobacco (Ishiga et al. 2013). In this study, nine tomato COI genes were identified in the whole genome using four Arabidopsis COI sequences as the query. The bioinformatics analysis revealed that SlCOIs showed hydrophobic and hydrophilic characteristics and was in silico localized in the nucleus, cytoplasm, and chloroplast (table 1). These genes were physically located in seven chromosomes and shared molecular weight as low as 74.14 kDa and as high as 73.78 kDa. The phylogenetic analysis of tomato SlCOI proteins with other plant species revealed that all the SlCOI proteins were clustered into five clades (figure 2). It was also observed that SlCOIs with similar gene configuration (exons/introns) grouped in the same cluster (figure 2 in electronic supplementary material).

The expression analysis of SlCOIs in different tomato plant parts was investigated to examine COI functions (figure 5). It was observed that these proteins show significant expression levels in a specific tissue or plant organ. For example SlCOI1, SlCOI3, SlCOI4, SlCOI7 and SlCOI9 was specifically expressed in root tissues with high transcript abundance but SlCOI6 in 1 cm fruit; SlCOI8 in 3 cm fruit, and SlCOI4 in breaker stage fruit. These findings suggested that these genes might play a pivotal role in developing these plant organs. For further insight into the functions of SlCOIs, cis-regulatory elements (figure 3 in electronic supplementary material) and expression under salt, drought, cold and exogenous phytohormones such as ABA, IAA, GA, MeJA and SA were analysed (figure 2 in electronic supplementary material). The hormone associated cis-elements such as ABA, IAA, GA, MeJA and SA, and stress-responsive cis-regulatory sequences such as salinity, light, drought, and defence were found in promoters of tomato COI genes. These suggested that SlCOI genes involved in different regulatory mechanisms activated under abiotic and biotic stimuli.

The COI genes expression regulated by various plant hormones. For instance, Arabidopsis (AtCOI1) involved in JA mediated pollen development (McConn and Browse 1996) and insects defence (Mcconn et al. 1997). COI genes were also induced by some abiotic stresses. TaCOI1 was upregulated under salt and drought stresses, but TaCOI3, TaCOI6 and TaCOI8 were downregulated (Jian-fang et al. 2018). However, tomato COI genes were induced by salt, drought, and cold treatments (figure 6, a–c). SlCOI1, SlCOI3, SlCOI5, and SlCOI6 were upregulated by salt; SlCOI3, SlCOI6, and SlCOI8 induced by drought; SlCOI6, SlCOI7, SlCOI8, and SlCOI9 were peaked under cold stress. In this study, inducible expression analyses revealed that SlCOI expression could be induced by at least one hormone. In wheat, TaCOI2 and TaCOI6 were downregulated by treatment with SA, MeJA, PEG, and IAA (Jian-fang et al. 2018). Consistent with this, the expression of SlCOI genes was also at least induced under one exogenous phytohormone (figures 6, d and 7), such as SlCOI1, SlCOI2, SlCOI3, SlCOI4, SlCOI5, and SlCOI7 were peaked under ABA, GA, SA and MeJA; SlCOI6 was upregulated by GA; SlCOI7, SlCOI8, and SlCOI9 were upregulated by IAA; SlCOI5, SlCOI7 and SlCOI4 were downregulated by GA, SA and MeJA. These suggest that tomato COIs might be directly or indirectly associated with enhanced tolerance against these stresses and plant development. Further investigations will need the functional characterization of these genes in the future.

In conclusion, these results provide a comprehensive overview of the COI gene family in tomato model plant, including phylogeny, in silico subcellular location analysis, and expression profile analysis in various plant parts under normal conditions. Moreover, the role of SlCOIs in response to salinity, drought, cold and exogenous phytohormones such as ABA, IAA, GA, SA, and MeJA provides a more comprehensive understanding of these genes' functions. Further, functional characterisation of these genes in the future validates the conclusion presented in this study.