Plant material

Tomato plants (S. lycopersicum) cv. Rutgers were grown under greenhouse conditions at Zamora, Hermosillo, Sonora, México (20° 16’ 56.8” N, 110° 53’ 08.5” W). Inside the greenhouse, temperatures ranged from 25 to 32 °C. Tomato plants were watered daily, with weekly applications of a Hoagland solution enriched with 300 mM H₃PO₄ and 15 mM K₂SO₄. Self pollination was carried out by gently moving unpollinated flowers using a small brush. Immediately thereafter, a cardboard tag including the date was placed in the peduncle tissue. Tomato fruits were collected at 5, 10, 30 and 40 days after anthesis (DAA). Afterwards, tomato fruits were harvested at the mature green (MG), turning (TUR) and red ripe (RR) stages based on the United States Standards for Grades of Fresh Tomatoes (CFR51.1855-51.1877). In all samples, the fruit mesocarp was removed, frozen with liquid nitrogen and kept at -80 °C for further analysis.

Analysis of gene sequences

Amino acid sequences of the isoenzymes Solyc04g076630, Solyc04g076660 and Solyc11g011300 were obtained from the SolGenomics Network (SGN). Sequences were analyzed for the presence of domains using domain finder from NCBI (ncbi.nlm.nih.gov/Structure/cdd/wrpsb) and aligned with COBALT (Constraint-based Multiple Alignment Tool) from NCBI and Clustal-Omega from EMBL-EBI. The Expasy box shade online tool was used to represent the alignment. The reference sequence was rhamnogalacturonan lyase (GenBank id QFG75912.1).

Ethylene quantification

Ethylene was quantified based on the methodology published by ^{Ojeda-Contreras et al. (2008)}. For the assay, three biological samples composed of three tomatoe fruits were analyzed. Tomatoes were incubated for 1 h at 25 °C in a 1 L sealed glass container. After that time, fruit ethylene production was measured by collecting a 1-mL sample from the headspace with a syringe and injecting it into a chromatograph (Varian 3400 cx gas, Agilent Technologies, Santa Clara, California, USA) equipped with a HayeSep N column that was 2-m long with an internal diameter of 3.17 mm (Supelco Analytical, Inc., Bellefonte, Pennsylvania, USA) equipped with a thermal conductivity detector for CO₂ quantification and a flame ionization detector for ethylene quantification. The chromatography conditions were as follows: injection temperature of 100 °C, thermal conductivity detector temperature of 170 °C, and flame ionization detector temperature of 120 °C. Nitrogen was used as the carrier gas with a flow rate of 25 mL min^-1.

RNA Isolation and first strand DNA synthesis

The pericarp of fruits at different DAA was isolated, chopped into small pieces, frozen in liquid nitrogen, pulverized with a frozen mortar and pestle, and then stored at -80 °C until further analysis. Three tomato fruits harvested from individual plants were considered a single biological sample, and two biological samples were used for the analyses. RNA was isolated with the hot borate method ^{(Wilkins and Smart, 1994)}. RNA was quantified in a ultra-low volume spectrophotometer (Nanodrop 2000, Thermo Scientific, Waltham, Massachusetts, USA), and then RNA was treated with DNAse RQ1 (Promega) according to the manufacturer’s instructions. RNA integrity was corroborated by gel electrophoresis on a 1.2 % agarose gel. First strand DNA was synthesized from 1 (g of RNA, and then the retrotranscription reaction was performed by SuperScript II (Invitrogen) following the manufacturer instructions.

Quantification of transcripts

Primers were designed with QuantPrime ^{(Arvidsson et al., 2008)} using the tomato cDNA database (ITAG release 2.4). Primers were tested in a dynamic range assay. In the retrotranscription reaction, a mix of specific reverse primer pools was added for reference and RGL genes at a concentration of 400 pM for each reverse-specific primer ^{(Nolan et al., 2006)}. Gene for ubiquitin (UBI), a catalytic subunit of protein phosphatase 2A (PP2Acs) ^{(Løvdal and Lillo, 2009)} and a TIP41-like family protein ^{(Expósito-Rodríguez et al., 2008)} were used to normalize the gene expression results (Table 1).

To quantify the RGL isoenzyme and reference gene expression levels, quantitative real-time polymerase chain reaction (RT-qPCR) was performed in a total reaction volume of 20 (L. cDNA was measured by a Nanodrop 2000 after elimination of de RNA with RNAase H. Further, 20 ng of cDNA were used as a template for each reaction; then, 250 nM forward and reverse primers were used for the experimental reaction. PCR reagents and 10 (L of 2X HotStart-IT SYBR Green (Affymetrix) were added to the mix. Both biological samples were measured with three technical replicates in an Real-time PCR System (StepOne, Applied Biosystems, Waltham, Massachusetts, USA) under the following conditions: 2 min at 95 °C for denaturing and 40 cycles of 15 seconds at 95 °C and 30 seconds at 60 °C. The results were calculated with the 2^-(CT methodology ^{(Schmittgen and Livak, 2008)}.

Statistical analysis

Expression data were analyzed under a completely randomized design with a statistical significance of P ≤ 0.05. One-way analysis of variance (ANOVA) was used, and when ANOVA identified statistically significant results, the Tukey-Kramer test was used to identify significant difference between the means. All statistical analyses were performed by NCSS 7. For the Pearson correlation coefficient, an R script was designed to correlate the gene expression data with ethylene production during fruit ripening.

In silico analysis of amino acid sequences

The functional domains between the three RGL isozymes analyzed were compared, including a characterized rhamnogalacturonan lyase (QFG75912.1) from beach strawberry (Fragaria chiloensis) as a reference sequence. Figure 1 shows the catalytic domain, active site, and cofactor binding domain motifs in the three tomato RGL genes and in the reference rhamnogalacturonan gene from strawberry. It was recorded a 54.79, 52.73 and 57.01 of percentage identity with the reference sequence in Solyc11g011300, Solyc04g076630 and Solyc04g07660 genes, respectively. Furthermore, approximately 54 % identity was recorded among the tomato RGL sequences. The low identity is most likely because the Solyc04g07660 and Solyc04g076630 genes are located on chromosome 4, and the Solyc11g011300 gene is on chromosome 11. The catalytic domain RGL_4N (accession cl15675) is responsible for cleaving the glycosidic bond (Figure 1, underlying continuous black line) between the rhamnose and galacturonic acid residues with the configuration L-rhamnopyranosyl-(1->4)-alpha-D-galactopyranosyluronide by a β-elimination mechanism ^{(McDonough, et al., 2004)}. In this domain, the residues Lys-Tyr-Leu and His are conserved in the three aligned RGL genes (Figure 1, highlighted in black). In agreement with this result, ^{Uluisik and Seymour (2020)} reported that β-elimination requires basic residues such as Arg, Lys, or His to separate the α-proton at the C5 atom of the substrate galactose residue. The middle RGL4_M (accession cd10316) and carbohydrate binding module (CMB, accession pfam14683) are shown with underlying gray and red lines, respectively, in Figure 1. Furthermore, these modules are present in Solyc11g011300, Solyc04g076630, Solyc04g076660 and the reference (QFG75912.1) rhamnogalacturonan gene. Additionally, the conserved calcium binding site within the CBM domain ^{(Uluisik and Seymour, 2020)} is highlighted in light green.

Figure 1 Amino acid sequence alignment of the three RGL genes. The continuous black line shows the rhamnogalacturonan lyase domain RGL_4N (accession cl15675). The amino acid located at the active site is highlighted in gray. The dotted black line underlines the catalytic site, and the amino acid playing a role in the catalysis of RG-l is highlighted in black. The continuous green line underlines the carbohydrate binding domain RGL4_M (accession cd10316). The continuous red line underlines the carbohydrate binding module (accession pfam14683). The amino acids that bind calcium ions are highlighted in light green. 

The main function of RGL4_M and CMB is to interact with the spatial arrangement of the RG-I polymer ramification ^{(Silva et al., 2016)}. These three domains were also found within the amino acid sequence of an RGL enzyme isolated from two species of strawberry ^{(Méndez-Yañez et al., 2020}; ^{Molina-Hidalgo et al., 2013)}. In the three tomato RGL sequences, the three domains were found to play a role in cleaving and interacting with the RG-I pectin polymer of the cell wall. Furthermore, the presence of the conserved amino acid residues Lys, Tyr, Leu and His in the catalytic domain was observed. Based on the above results, Solyc04g076660, Solyc04g076630 and Solyc11g011300 show the presence of domains indicating that they are rhamnogalacturonan lyase enzymes, although more experimental evidence is needed to support this statement.

Gene expression and ethylene production

With the aim of elucidating the physiological function of RGL on the PCW during fruit development and ripening, the expression patterns of three genes encoding different RGL isoenzymes were obtained. Quantification of the Solyc04g076660 transcript (Figure 2A) showed low levels of gene accumulation from 5 to 40 DAA. Furthermore, during fruit ripening, higher quantities of Solyc04g076660 transcripts at the MG and RR stages were observed, with no statistically significant differences (P > 0.05). Figure 2B shows the expression changes of the Solyc04g076630 gene. Very low expression was recorded during fruit development, except at 10 DAA, in which there was an eightfold increase in expression relative to that at 5 DAA (P ≤ 0.05) and a more than fourfold increase relative to that at the MG stage of fruit ripening (P ≤ 0.05).

Figure 2 Changes in gene expression of the different rhamnogalacturonan lyase isoenzymes studied during different stages of fruit development and ripening. Panel A shows the relative expression of Solyc04g076660, panel B shows the relative expression of Solyc04g076630 and panel C shows the relative expression of Solyc11g011300 along with the changes in ethylene production and images of tomato fruits at different DAA. DAA: days after anthesis, MG: mature green, TUR: turning. RR: red ripe. Asterisks on top of the bars indicate statistically significant differences (P ≤ 0.05). Vertical lines on the bars indicate the standard deviation. 

The accumulation of transcripts from the Solyc04g076630 gene in fruits at 10 DAA can be associated with cell wall remodeling during fruit development. In the two phases of fruit growth (fruit division and fruit expansion), the pectin polymer RG-l and other polymers constantly changed, as reviewed by ^{Wang et al. (2018)}. This study showed increase in the Solyc04g076630 transcript level during fruit development when cell division ended at 10 DAA (Figure 2B). As reported by ^{Goulao and Oliveira (2008)}, modification of the plant cell wall in the early stages of fruit development can be carried out by the coordinated activity of different isozyme families, as observed in the wild tomato Ailsa Craig, pear and apple. Furthermore, evidence reported by ^{Catalá et al. (2000)} and ^{Wu et al. (1996)} suggests that auxin induces the expression of enzymes that modify cell wall polymers, such as xyloglucan endo-transglycosidase (LeEXT1), expansin (LeExp2) and endo 1-4-(-glucanase during the early stage of fruit development; at this stage, the tomato fruit diameter is between 0.5 and 3 cm and the fruit is growing mainly by cell division. In fact, the production of auxins by fruits in the cell division phase (between 5 and 10 DAA) has been reported ^{(Mapelli et al., 1978)}. Additionally, the SlARF gene family, whose activation is regulated by auxins, contains the regulatory box TGTCTC in the promoter region, which has been shown to induce the expression of the gene SlARF5 ^{(Liu et al., 2018)}. In this study, it is suggested that the expression of Solyc04g076630 at 10 DAA may be due to the presence of the regulatory box TGTCTC contained in the RGL promoter. Considering that the transcriptomic analyses showed that the Solyc04g076630 and Solyc04g076660 genes are expressed in the pericarp of fruit development at 4 DAA ^{(Pattison et al., 2015)}, it is feasible to suggest that these genes play a role in tomato fruit growth phenomena by cell division ^{(Azzi et al., 2015)}, although more experimental evidence is needed to further support this statement. Based on results of this study, it is difficult to explain the expression of Solyc04g076630 in MG and Solyc04g076660 in MG and RR, although they may play a role during fruit ripening along with the gene Solyc11g011300. Studies during the last 20 years, had shown the coordinated expression of several gene families acting over the PCW, such as peroxidases ^{(Andrews et al., 2000)}, polygalacturonase, pectate lyase ^{(Crookes and Grierson, 1983)} and pectin methyl esterase ^{(Giovane et al., 1993)}, have been reported; the combined action of these gene families results in complex changes during fruit ripening phenomena ^{(Rivera-Domínguez et al., 2008)}.

The expression of an RGL gene during ripening was reported in two species of strawberry, Fragaria × annannasa ^{(Molina-Hidalgo et al., 2013)} and Fragaria chiloensis L. ^{(Méndez-Yañez et al., 2020)} . Additionally, the role of the RGL gene in fruit strawberry (Fragaria × annannasa) during softening was demonstrated by linkage association. Sequence comparison between tomato Solyc11g011300 and the strawberry rhamnogalacturonan lyase gene (FaRGLyase1) resulted in 65 % identity ^{(Berumen-Varela et al., 2018)}. In this experiment, high expression of the Solyc11g011300 gene was found during fruit ripening (Figure 2C), similar to the expression of the FaRGLyase1 gene ^{(Molina-Hidalgo et al., 2013)}. Based on the above mentioned findings, it is possible to suggest that the Solyc11g011300 gene from tomato plays a role in fruit softening, although more experimental evidence is needed to support this statement.

Figure 2C also shows the changes in ethylene production. In the MG stage, ethylene production of 0.13 (L kg^-1 h^-1 was recorded, but ethylene was increased 10-fold (1.31 (L kg^-1 h^-1) in the RR stage relative to the MG stage. It was found that Solyc11g011300 gene expression and ethylene production showed a similar trend during fruit ripening, with a positive correlation of 0.9, which is rather high. The accumulation of Solyc11g011300 transcripts and ethylene production begin at the same time, maintaining a large positive correlation during fruit ripening. In addition, the ethylene responsive element ERELE4 was found in the Solyc11g011300 promoter ^{(Berumen-Varela et al., 2017)}. In climacteric fruits, ethylene is responsible for the activation of genes that induce color changes, alter starch-sugar metabolism, soften fruits and synthesize volatiles related to aroma, among other changes. In this regard, the Solyc11g011300 gene is probably one of the genes that are activated by ethylene to carry out the changes in texture and firmness observed in fruits during ripening, although more experimental evidence is needed to support this statement.