Recent Advances in Studying the Regulation of Fruit Ripening in Tomato

1. Introduction

The fruits of angiosperms are included in the staple diet of humans and livestock. The transition of plants from the vegetative growth phase to the reproductive stage is the main switch in their life cycle. Fruit ripening is initiated and regulated by the combined action of various genetic factors in response to endo- and exogenous stimuli. The functions of ripening regulators are often conserved among different angiosperms.

In addition to the nutritional importance of tomatoes, this crop is a convenient object for studying the mechanisms of fruit ripening regulation. This is explained by its simple diploid genetics, small genome size, short life cycle, ease of transient and stable transformation, pronounced ripening phenotypes, an abundance of bioinformatics data, and well-characterized mutants.

Modern widespread varieties of tomatoes were obtained through domestication and subsequent selection. The selection of seemingly desirable traits carried out without an understanding of the nature of gene relationships has contributed to the reduction of genetic diversity in tomatoes. Consumer preferences and cultivation convenience have also contributed to this. On the other hand, the introgression of elite alleles into a cultivated variety helps create a hybrid genome with the allele of interest but also introduces undesirable genetic backgrounds in the form of linked genes from the allele donor. With the help of backcrossing, breeders can level out the manifestation of unwanted traits, but this is time-consuming and not consistently effective.

Genetic engineering methods provide significant potential for studying the genetic factors regulating fruit ripening. Thus, targeted genome editing technology using the CRISPR/Cas9 system allows researchers to create knockout alleles and make precise changes to the gene sequence. This, in turn, helps to study genes’ functions, relationships, and regulation. Despite the 10-year history of using this powerful technology, the number of publications using it, in which the tomato is the object, is growing steadily yearly (Figure 1). Thus, its research potential still needs to be exhausted.

An analysis of publications where the CRISPR/Cas9 system has been used in tomatoes over the past six years has made it possible to identify topics of most interest to the scientific community (Figure 2). It turned out that more than a quarter of the total number of works is devoted to the study of genes involved in the processes of fruit ripening (27%). Many of these genes encode transcription factors and transcriptional coregulators, microRNAs, or proteins involved in the epigenetic control of gene expression. In many cases, these regulators’ molecular mechanisms of action have yet to be studied, which is the reason for the growing interest in research in this area. Also, a considerable proportion of publications are devoted to the study of the regulation of the processes of flowering and fruit development (18%). The consistently current topic of stress (abiotic and biotic) occupies a third of the total number of publications. The remaining publications cover fields devoted to other physiological processes (15%) and plant architecture and morphology (8%).

Among the most prevalent genetic engineering methods used to study the regulation of ripening processes, the use of the CRISPR/Cas9 system is expected to increase (Figure 3a). At the same time, approaches that have already become classical, such as RNA interference gene silencing, and gene over- and heterologous expression, have not lost their relevance - the number of publications using them has remained consistently high over the past six years (Figure 3, b, c). Interestingly, there are a growing number of studies using gene overexpression to study ripening. All this suggests that, despite the large amount of accumulated data, the regulation of the ripening process still needs to be fully understood.

Ripening, as the final stage of development of tomato fruits, manifests itself in the form of bright pigmentation, increased aroma and taste, and softening of the pulp. These properties make the fruits attractive to animals, which act as seed dispersal vectors. Before the onset of ripening, these physiological changes are suppressed, but once the fruit enters the ripening phase, they occur in a highly synchronized manner with dramatic differences in gene expression patterns. Such radical changes are regulated by environmental factors such as temperature and light and by internal regulators, including hormones, transcription factors, and epigenetic modifications. Due to the commercial and biological importance of tomato fruit ripening, the mechanisms regulating this process are being actively studied.

2. Transcription factors regulating ripening

MADS-box genes are among the most widely represented and diverse transcription factors; consequently, they mediate various biological processes. Among them, regulating fruit ripening is one of the most prominent roles of MADS-box (MCM1, AGAMOUS, DEFICIENS, and SRF) genes. The transcription factor RIN (RIPENING INHIBITOR) has long been considered a major ripening regulator. RIN encodes a SEPALATA class MADS-box transcription factor. MADS-box family transcription factors typically function as multimers, and the MADS-box proteins TAGL1 (TAG-like) and two FRUITFULL (FUL) homologs (TDR4/FUL1 (tapetum degeneration retardation) and MBP7/FUL2 (MADS-box protein)), are coregulators with RIN and ripening regulators with overlapping functions [1]. Silencing of TAGL1 resulted in decreased levels of amino acids in fruit: aspartic acid, L-tyrosine, L-glutamine, L-phenylalanine, L-valine, L-leucine, isoleucine, and 5-caffeoylquinic acid [2]. TAGL1 was also found to regulate the synthesis of the glycoalkaloid α-tomatine negatively. As discussed in [3], TDR4/FUL1 and MBP7/FUL2 do not regulate ethylene biosynthesis but influence fruit ripening in an ethylene-independent manner. RIN often binds to demethylated sites in the promoter regions of ripening-related genes. RIN is induced early in ripening and stimulates ethylene-dependent and ethylene-independent pathways that promote ripening. Mediators in this process are response factors to ethylene (ERF, ethylene-responsive factor) and auxin (ARF, auxin response factor). ERF and ARF control their respective hormonal signaling pathways, regulating gene expression and hormonal signaling.

Several recent studies have clarified the function of RIN. Thus, [4] found that although RIN function is required for full ripening, RIN is not required for the initial ripening induction. The authors suggest that RIN acts redundantly (i.e., there are RIN homologs) or RIN-independent ripening induction occurs due to other transcription factors. In the second case, the authors concluded that a RIN-independent activator can induce the transcription of ripening-related genes even in RIN-deficient plants. Still, a mutant (defective) RIN protein can inhibit its activity. The chimeric transcription factor RIN-MC exhibits a negative role in ripening, promoting the mutant rin phenotype [5]. Other authors think that low ethylene concentrations initiate the ripening of mature green fruits, activate RIN expression, and lead to other changes, including a transition to a burst of autocatalytic ethylene synthesis [6]. Combined with the ethylene biosynthesis gene ACS2 (1-aminocyclopropane-1-carboxylate synthase), RIN has been shown to regulate the heat shock genes HSP17.7 [7] negatively. Therefore, RIN, ethylene, and other factors are necessary to complete the complete fruit ripening program. RIN is not only an activator of ripening but also a repressor of excessive softening [8]. It was found that when controlling the fruit ripening process, RIN binds to 6 lncRNAs [9].

RIN is reported to directly activate the expression of a novel gene, E6-2, involved in tomato fruit ripening [10]. Silencing of E6-2 leads to a delay in fruit ripening suppression of CNR (colorless non-ripening), PG (polygalacturonase), and ERF4 (ethylene-responsive factor) was observed, a decrease in the accumulation of carotenoids and lycopene (due to the suppression of PSY1, PDS and ZDS (phytoene synthase, phytoene desaturase, and zeta-carotene desaturase, respectively)), ethylene (decreased expression of the biosynthetic genes ACS2, ACO1 (1-aminocyclopropane-1-carboxylic acid oxidase), ACO3 and ethylene-sensitive E4, E8), increasing the content of pectin, cellulose, starch and soluble sugar (suppression of cell wall metabolism genes TBG4 (tomato beta-galactosidase), PL (pectate lyase), EXP1 (expansin), and XTH5 (xyloglucan endotransglucosylase/hydrolase)). The broad phenotypic pattern of RIN silencing is an attractive marker for testing molecular editing tools [11,12,13].

Genes with the NAC domain (NAM, ATAF1/2, and CUC2 (apical meristem, ARABIDOPSIS TRANSCRIPTION ACTIVATOR FACTOR, and cup-shaped cotyledon, respectively)) are considered to be another transcription factors that regulate tomato fruit ripening. It has been shown [14] that inhibition of NOR-like1 reduces ethylene production, delayed softening and loss of chlorophyll, and reduced lycopene accumulation. Activation of ethylene synthesis genes by NOR (non-ripening) and NOR-like genes is discussed further in [15]. Knockout of NAC-NOR suppressed fruit ripening (inhibition of ethylene synthesis, reduction of carotenoid accumulation, and fruit softening), and the opposite effect was observed with its overexpression [16]. Replacement of thymine with adenine in the ALC gene (alcobaca, NOR mutation) using homologous recombination contributed to an increase in the shelf life of tomato fruits [17]. Expression of peach NAC1 in tomatoes has been shown to enhance ripening in a delayed ripening (NOR) mutant and restore the synthesis of volatile esters [18]. Overexpression of NAC6 resulted in increased levels of endogenous abscisic acid, which affected the transcription of ripening genes [19]. Transfer of the kumquat NAC22 gene to tomato increased the expression of most carotenoid biosynthesis genes, accelerated the transformation of plastids into chromoplasts, and promoted color changes [20]. NAM1 (no apical meristem), another factor with an NAC domain responsible for the regulation of ethylene biosynthesis, also controls tomato ripening, as confirmed by delayed ripening in CRISPR/Cas9 mutants and accelerated ripening for lines overexpressing NAM1 [21]. Repression of NAM gene domains is carried out by miR164a [22,23,24]. In addition, the HWS (HAWAIIAN SKIRT) gene, encoding an F-box protein, regulates the number of floral organs by modulating the transcription levels of the miR164, CUC1 and CUC2 (cup-shaped cotyledon) genes. HWS is also involved in petals’ cell proliferation and mitotic growth [25].

It was previously shown that a representative of genes with the NAC domain NAP2 (Arabidopsis NAC domain-containing protein) activates the aging gene SAG113 (senescence-associated gene), protein phosphatase), chlorophyll degradation genes SGR1 (stay-green), PAO (polyamine oxidase), and NAP2, and also directly controls the expression of genes essential for the biosynthesis of abscisic acid NCED1 (9-cis-epoxycarotenoid dioxygenase), ABCG40 (Arabidopsis thaliana ATP-binding cassette), and CYP707A2 [26]. These interactions suggest the influence of NAP2 on leaf senescence and yield in tomato. Another new gene, named by the authors HEBE in honor of the Greek goddess of youth, has similar functions [27].

Numerous studies have shown that the CNR gene is the most important regulator of tomato fruit ripening. However, recent research [28] has called this assumption into question. CNR knockout lines exhibited only a ripening-arrested phenotype, while NOR knockout (non-ripening) lines exhibited a partial non-ripening phenotype similar to RIN mutants. Both knockouts differed from the strong, non-ripening phenotypes of their natural mutants. It became apparent that the expression of characteristic ripening genes, such as ACS2, ACO1, PSY, PG, and EXP, is not entirely suppressed in CRISPR/Cas9 lines compared to natural mutants. As the authors concluded, differences in the expression of the genes in question are explained by different degrees of methylation, and they also concluded that the regulatory network of transcription factor genes is redundant. Regulation of NOR may also be associated with something else: sulfoxidation of the NOR transcription factor with the help of MSRs (methionine sulfoxide reductase) proteins modulates the ripening process by reducing the DNA-binding ability of NOR [29].

It is known that the fruits of the tomato epimutant cnr fail to ripen and remain colorless. The SPL (SPOROCYTELESS) gene family consists of a group of genes encoding SBP (SQUAMOSA promoter binding proteins)-box transcription factors, and their protein products bind to the promoter of the floral meristem identity gene SQUAMOSA. Evidence shows that SPL-CNR interacts with SnRK1 (SNF1-related protein kinase) [30]. Suppression of SnRK1 by virus-induced gene silencing (VIGS) inhibits fruit ripening and leads to decreased expression of a wide range of ripening-related genes. This suggests that SnRK1 transcription and subsequent post-translational SPL-CNR-SnRK1 interaction are biologically crucial for tomato fruit ripening. The authors suggest that the involvement of SnRK1 in fruit ripening may be due to the physical interaction of proteins between the SnRK1 gene product and SPL-CNR and subsequent phosphorylation of SPL-CNR due to the kinase activity of SnRK1 [30].

The role of some MBP transcription factors in the ripening process has recently been studied. Thus, suppression of the MBP8 factor shortened the fruit ripening time, suggesting an increase in the activity of ethylene synthesis genes [31]. Meanwhile, carotenoids accumulated to higher levels, and the expression of PSY1, PDS, and ZDS was enhanced in MBP8 RNAi-silenced fruits. The activity of cell wall genes also changed, manifested in the softening of fruits. Silencing MBP15 in [32] delayed tomato ripening, and gibberellin, carotenoid, and ethylene biosynthesis genes were repressed. MBP15 was found to interact with RIN [32].

GRAS (gibberellic acid insensitive (GAI), repressor of GAI (RGA), and scarecrow (SCR)) proteins are plant-specific transcription factors that play critical roles in plant development and stress response. It turned out that they also take part in regulating fruit ripening. For example, silencing GRAS2 reduces tomato fruit weight, which has been attributed to insufficient levels of gibberellic acid during initial ovary development [33]. Overexpression of GRAS4 accelerated fruit ripening (due to activation of expression in the promoter region of ethylene biosynthesis genes and repression of the negative regulator of ripening MADS1). It increased the total content of carotenoids [34]. GRAS24, in addition to flowering and ripening, is responsible for a variety of other agronomic traits, including plant height, leaf architecture, number of lateral branches, root length, and the observed pleiotropic effects in plants overexpressing GRAS24 are due to impaired modulation of gibberellin and auxin signaling [35].

Transcription factors of the WRKY superfamily exhibit up-regulation during fruit ripening. WRKY32 binds to W-box and similar motifs in the regulatory region of the YFT1 (yellow-fruited tomato) promoter and induces its expression [36]. YFT1 encodes the EIN2 protein, a major ethylene signal transduction component. Suppression of ethylene production resulted in delayed chromoplast development, decreased carotenoid accumulation, and a yellow fruit phenotype. Twelve WRKY genes were also shown to be ethylene-responsive (ER), 8 of which activated the promoters of color change-associated genes (PPH (pheophytinase), PAO (polyamine oxidases), PSY1, and PDS) [37]. In addition, protein interactions were found between WRKY17 and RIN/ERF2b/ERF7, WRKY33 and ERF7, WRKY54 and ERF2b, WRKY16 and WRKY1, which only confirms the complexity of the networks of ripening regulators [37].

3. Epigenetic modifications as regulators of ripening

DNA methylation has a critical role in a wide range of cellular functions. For example, a decrease in DNA methylation levels can be observed during fruit ripening, which is explained by DNA demethylases (DML) activation. In DML2 loss-of-function mutants generated by targeted editing, increased DNA methylation was found not only in genes induced during ripening but also in genes repressed during ripening [38]. However, a recent study found that the highly mobile protein positively regulates DML2 expression of 3-hydroxy-3-methylglutaryl-coenzyme A reductase (HMGA) [39]. In [40], expression of the mammalian demethylase TET3c (ten-eleven translocation) in tomatoes resulted in the activation of expression of the previously undescribed gene CEN1.1. The activation intensity of CEN1.1 expression correlated with increased hypomethylation in its promoter, suggesting that CEN1.1 expression is associated with DNA methylation of CHH promoter sites (H=A/C/T). Phenotypically, CEN1.1 has emerged as a repressor of flowering in tomatoes, leading to the development of leaves on inflorescences. Paradoxically, this led to an increase in the number of fruits but to a longer time for their ripening. Thus, this study provides an exciting approach to identifying methylation-associated genes.

In plants, cytosine methylation plays a crucial role in suppressing the movement of transposable elements. Methylation is maintained by DNA methyltransferases MET1, and CMT3 (chromomethylase), as well as additional proteins (for example, DDM1 (decreased DNA methylation)) involved in maintaining heterochromatic structure. The methylase encoded by MET1 is a key DNA methylase responsible for maintaining CG methylation in plants. Loss-of-function mutants of the MET1 gene had pleiotropic developmental phenotypes manifested as small curled leaves, defective flowers, and parthenocarpic fruits [41]. Also, the knockout of MET1 resulted in changes in the expression profiles of RIN target genes, such as ACC2 (acetyl-CoA carboxylase 2). In another study, suppression of MET1 by VIGS in a hypermethylated epimutant CNR promoted vivipary development [42]. The authors explain this by a decrease in the concentration of abscisic acid and NCED transcripts involved in its biosynthesis. The NCED silencing had similar consequences. Methyltransferase DRM7 (domains rearranged methyltransferase) has been shown to influence chloroplast development by modulating starch accumulation and chlorophyll synthesis. It has an epi-effect on leaf senescence, affecting tomatoes’ vegetative growth [43]. The transient expression of arginine methyltransferase PRMT1.5 in tomatoes inhibited the accumulation of carotenoids and anthocyanins [44].

Short interfering RNAs (siRNAs) also mediate DNA methylation through RNA-directed DNA methylation (RdDM). It has been established that heterochromatic mobile elements in plants with DDM1 dysfunction are deprived of mCG and mCHG, which generally keep them inactive [45]. Methylation of CHH sites increased for some heterochromatic transposons and, conversely, decreased for those localized in euchromatin. Knockout of CMT4 chromomethylase caused severe morphological changes in tomato plants, accompanied by defects in leaves, pollen, and seeds [46].

Another mode of epigenetic regulation is post-translational modifications of histones. Histone acetylation is known to be associated with gene activation. In contrast, histone methylation can be associated with either activation or repression depending on the lysine residue and the number of methyl groups added. In [47], RNA-seq profiling showed a significant increase in the expression of methylases MET1 and CMT3 and a minor increase in the demethylase DML2 during the fruit set, which is associated with their role in maintaining post-replication DNA methylation during extensive cell division characteristic of early stages of development fetus However, their abundance was significantly lower than that for histone marks H3K9ac and H3K4me3, determined using chromatin immunoprecipitation sequencing. This implies that changes in the transcriptional profile underlying the fruit set are more closely related to histone modifications than methylation. Histone modification is based on the histone methyltransferase genes SDG27, SDG5, and SDG16 (set domain group). However, the authors could not create homozygous loss-of-function mutants of these genes, suggesting their exceptional biological importance. Mutants heterozygous for these genes exhibited parthenocarpic fruits. The function of other histone lysine methyltransferases SDG33 and SDG34 was revealed in [48]. They were found to regulate the expression of nitrogen-responsive genes and physiological changes in an organ-specific manner.

It has been demonstrated that histone demethylation leads to activating tomato fruit ripening genes [49]. Here, the JMJ6 (Jumonji C-terminal domain-containing demethylase) gene was found to encode a histone lysine demethylase that specifically demethylates H3K27. Its overexpression accelerates the ripening of tomato fruits, which is associated with increased expression of the RIN, ACS4, ACO1, PL, and TBG4 genes. As study [50] showed, JMJ4 mediates abscisic acid-induced leaf senescence in tomatoes.

By knocking out the HTA1 genes of histone H2A and subsequent production of double homozygous mutants, changes were identified in the expression patterns of many biological ripening processes, including cell redox homeostasis, mRNA splicing, cell cycle regulation, translation, etc. [51]. Moreover, for three genes of carotenoid biosynthesis, PSY1, PDS, and VDE, expression was high regardless of the fruit ripening stage [51]. Histone deacetylation has been associated with transcriptional repression. Histone deacetylases carry out this process. There is evidence that they can act as both positive and negative ripening regulators. Indeed, RNAi silencing of the HDT3 (histone deacetylase) gene led to the suppression of genes for ethylene synthesis (ACS2, ACS4, ACO1, ACO3), carotenoids (PSY1), cell wall metabolism (HEX (acetylhexosaminidase), MAN (mannosidase), TBG4, XTH5, and XYL (xylanase)), as well as general genes, associated with ripening (RIN, E4, E8, PG, Pti4, LOXB (lipoxygenase)) [52]. In contrast, in [53,54], silencing of the HDT1 gene led to opposite results for these same transcripts. In this regard, the molecular mechanisms of regulation of these genes remain to be studied.

Histone acetyltransferase GCN5 acetylates histone H3 lysine (H3K14ac) and affects the levels of H3K9ac and H3K27ac. Its suppression leads to the loss of shoot apical dominance and a decrease in the size of the plant apical meristem [55]. It has also been established that GCN5 can increase WUSCHEL transcript levels. The expression of WUSCHEL can also be regulated by chromatin remodeling factors, such as the histone deacetylase HDA19 [56]. Here, the deacylation mechanism was found to involve the inhibitor gene IMA (inhibitor of meristem activity) acting as an adapter protein to form a chromatin remodeling complex together with the zinc finger protein C2H2 KNU (KNUCKLES) and the transcriptional corepressor TOPLESS.

4. Hormonal control of ripening

Auxin regulation is involved in all plant processes, including cell elongation and division, the formation of the architecture of roots, leaves, inflorescences, the development of embryos and fruits, and responses to stress. The leading site of auxin biosynthesis is young leaves and their primordia. The amino acid tryptophan is the primary precursor of auxins. Knockout of any of the auxin synthesis genes is associated with lethal phenotypes, so attention is paid to genes providing auxin-mediated inactivation (GH3, GRETCHEN HAGEN), transport (PIN, ABCB (PIN-FORMED, ATP binding cassette subfamily B, respectively)), and signal transduction (ARF (auxin response factor), Aux/IAA).

By conjugating auxins to amino acids for storage or degradation, members of the GH3 family, encoding acyl acid amidosynthetases, are critical for maintaining auxin homeostasis. In tomatoes, GH3.15 has been shown to regulate lateral root development and response to gravitropism by modulating auxin homeostasis [57], GH3.8 controls plant height [58], GH3.4 negatively regulates mycorrhization [59,60], GH3.2 affects fruit ripening in the early stages [61].

PINs are one of the facilitators of intercellular auxin transport. VIGS PIN1 accelerates flower abscission by increasing the accumulation of auxin in the ovule and reducing the auxin content in the abscission zone [62], and its negative regulator is the transcription factor MBP9 [63].

ARFs are plant-specific transcription factors that directly bind to auxin response elements in the promoters of auxin-responsive genes. ARF5 has been shown to regulate fruit set and development [64], ARF10 is involved in the accumulation of chlorophyll and sugar during fruit ripening [65], ARF19 is involved in leaf development [66], several ARFs (ARF6A, ARF8A, ARF8B, and ARF24) interact with the transcriptional repressor IAA9 [67] and regulate leaf shape [68], ARF10A is essential for the growth of leaf blades and formation of floral organs [69].

Phytohormones gibberellins regulate various physiological processes of plants. They promote plant growth participate in stem elongation, expansion of leaf blades, pollen development, flowering, and seed germination. Gibberellins are synthesized in plastids and then modified in the endoplasmic reticulum and cytosol.

DELLA (GRAS gene encodes protein containing D-E-L-L-A amino acid sequences) proteins are nuclear-localized negative growth regulators. Gibberellins promote DELLA degradation by assembling the E3 ubiquitin ligase complex, followed by protein degradation. DELLA is encoded by the PROCERA gene, and its loss of function in the homozygous state results in dwarfism [70] and parthenocarpy [71]. The degradation of proteins, including DELLA, is controlled by a complex regulatory network involving connections between several signaling pathways [72]. DELLA proteolysis is mediated by the gibberellin-activated receptor GID. Knockout of their coding genes also results in a dwarf phenotype [73]. There is evidence of cross-signaling between the gibberellin and abscisic pathways [74], and the DELLA protein is an activator of abscisic acid transporters (AIT), regulating transpiration through stomatal closure [75]. Tomato PROCERA activity is assumed to be necessary to transition tomatoes to flowering. DELLA protein directly or indirectly promotes the expression of SFT (SINGLE FLOWER TRUSS) in leaves, as well as SBP and AP1/MC, together with microRNAs in the shoot apex [76].

Recently, factors mediating gibberellin-dependent regulation have also received attention. Thus, silencing of the GRAS15 transcription factor led to pleiotropic phenotypes, including reduced plant height, small leaf size with pointed edges, as well as an increased number of nodes, lateral shoots, and petiole length, which is explained by the suppression of gibberellin synthesis genes [77]. The helix-loop-helix transcription factor gene PRE2 is induced by gibberellin. Its silencing has been shown to cause reductions in fruit size, seed size, pericarp thickness, and placental size [78]. These changes are associated with decreased expression of xyloglucan endotransglucosylases XTH2 and XTH5. PREs regulate many processes - their overexpression in tomatoes leads to multiple morphological changes, including changes in leaf angle, internode length, leaf curl, and pigment composition [79,80].

Cytokinins promote the development of shoots, provide stress resistance, and delay aging. These isopentenyladenine derivatives are formed mainly in the roots and transported to the aerial parts. The main rate-limiting enzyme for cytokinin synthesis is isopentenyltransferase (IPT). It has been shown that overexpression of the IPT gene leads to significant phenotypic changes and slower leaf senescence only under the control of a root-specific promoter [81].

Cytokinin catabolism is carried out by cytokinin oxidases (CKX). Overexpression of CKX2 in tomato fruit decreased cytokinin levels [82]. It is also shown here that endogenous cytokinins regulate the division of pericarp cells, which subsequently determines the size of the fetus. Ethylene is the simplest unsaturated hydrocarbon with the formula C2H2. In plants, it acts as a global regulator of developmental processes. The ethylene biosynthetic pathway includes three steps: S-adenosylmethionine synthetase (SAMS) modifies methionine to form S-adenosylmethionine (SAM), SAM is converted to 1-aminocyclopropane-1-carboxylic acid (ACC) by synthase (ACS), and in the last step ACC converts ACC oxidase (ACO) with the formation of ethylene.

In studies on tomatoes, ethylene is considered a participant in signaling cascades, including during the ripening process. It has been established that its synthesis during ripening is presumably regulated by FER receptor kinases (FERONIA). FERL6 and FERL1 were found to interact physically with the SAMS promoter [83]. Expression of FER genes in tomatoes showed negative regulation of ethylene accumulation at the initial stages of fruit development and, as a consequence, delayed fruit ripening.

The promoter of the transcription factor EIN3 (ethylene insensitive) has been shown to contain several motifs associated with hormones influencing fruit development and ripening [84]. Overexpression of EIN3 in tomatoes resulted in the activation of the expression of ethylene biosynthesis genes ACO1, ACS1, and SAMS1, which promoted early fruit ripening. Accordingly, EIN3 silencing showed the opposite effects. An EIN3-like gene causes premature onset of ovule senescence [85].

Ethylene is bound by a family of ETR (ethylene receptor) proteins located in the membrane of the endoplasmic reticulum. ETRs have functional redundancy. ETR3-mediated signaling inhibits pollen tube growth without sufficient ethylene [86]. ETR3 promotes the activation of cell wall remodeling genes and Ca2+ transporters — overexpression of ETR7 results in earlier flowering, short plants, and small fruits [87]. Targeted base substitution in ETR1/2 causes a delay in ripening and ensures prolonged storage of fruits [88,89].

Ethylene response factors (ERFs) are signaling components involved in ethylene-dependent developmental processes. They can perform both positive and negative regulation of target genes. Their number is large, as is the specificity of the reactions of tomato genes to ethylene: regulation of fruit ripening processes [90,91,92,93], control of aging [94], participation in the activation of protective reactions [95,96,97], growth [98,99], accumulation of chlorophyll and formation of chloroplasts [100], and regulation of other signaling pathways [101].

Brassinosteroids, which include various polyhydroxylated steroidal phytohormones, influence many critical agronomic traits related to growth, photosynthesis, morphology, and yield. The synthesis of brassinosteroids occurs along three pathways, in which campesterol is the initial substrate. Crosstalk between brassinosteroids and redox signals suggests a direct involvement of the former in the plant response to stress [102,103]. However, recent studies also reveal a connection between the brassinosteroid and ethylene pathways. Recently, it was demonstrated that overexpression of one of the genes for the brassinosteroid synthesis enzymes DWARF (DWF) in tomatoes promotes fruit softening, lycopene synthesis, and ethylene production, while gene knockout inhibits them [104]. It was concluded that APETALA2a (AP2a) promotes ethylene signaling to regulate brassinosteroid signaling. Also, tomatoes with overexpression and silencing of the cytochrome P450 monooxygenase CYP90B3 gene showed a correlation of the content of bioactive brassinosteroids with the processes of tomato fruit ripening, including softening, the content of soluble sugars and aromatic volatiles [105].

Research into brassinosteroid-dependent pathways is ongoing. The specific receptor for brassinosteroids is BRI1 (BRASSINOSTEROID INSENSITIVE1). Upon binding of brassinosteroids to its extracellular domain, dimerization of BRI1 and the coreceptor BAK1 (BRI 1-associated receptor kinase 1) occurs. The signal is then transmitted through a phosphorylation cascade involving BSK1 (BR-signaling kinase 1), CDG1 (constitutive differential growth), BSU1 (BRI1 SUPPRESSOR), and BIN2 (BRASSINOSTEROID INSENSITIVE2). Subsequently, BIN2 is inactivated, and two transcription factors, BZR1 (brassinazole resistant) and BES1 (BRI1-extra microsporocytes-suppressor 1), are dephosphorylated by protein phosphatase PP2A. BZR1, BES1, as well as other nuclear factors (for example, BIM1) are regulators of brassinosteroid-dependent genes.

Overexpression of BRI1 [106] in tomatoes improved carotenoid accumulation by increasing the expression of DXS (1-deoxy-D-xylulose 5-phosphate synthase), GGPS (geranylgeranyl pyrophosphate synthase), and PSY1. In addition, brassinosteroids induced the expression of genes involved in its ethylene biosynthesis (ACO1 and ACS2). Similar results were achieved by modifying threonine-1050 BRI1, resulting in plants with high levels of BRI1 autophosphorylation [107]. A recent study on BRI1 showed that the receptor also positively regulates a tomato’s tolerance to cold stress [108]. Brassinosteroids are capable of inducing early flowering. This is supported by the interaction of the suppressor of BIN2 signaling with the early flowering locus FRIGIDA [109]. There is evidence that canonical signaling pathways initiated by BRI1 are involved in xylem differentiation and wood formation in tomatoes through activation of the BZR1/2 transcription factors [110]. The BZR1 homolog has been shown to interact with BIM1 to act as a negative regulator of pericarp cell expansion [111]. According to available information, BZR1 is also a trans-activator of the promoter of the SUN gene (encodes Sad1/Unc-84 (SUN)-domain proteins), responsible for elongation of tomato fruits, and BZR1-knockout tomato phenotypes show redundancy of its homologs [112]. In addition, brassinosteroids promote bud growth in tomatoes through direct transcriptional regulation of BRANCHED1 (BRC1) via the signaling component BZR1 [113].

In the study [114], the authors focused on BES1, a key transcription factor in the brassinosteroid signaling pathway. BES1 was found to bind to the promoter of the fruit-softening inhibitor PMEU1 (pectin methylesterase). Knockdown or knockout of BES1 in tomatoes resulted in increased shelf life without negatively affecting the appearance and nutritional composition of the fruit.

Altered regulation of brassinosteroids may influence cell elongation and division, leading to altered fetal morphology. For example, a premature stop codon at the GLOBE locus containing a brassinosteroid hydroxylase sequence resulted in a spherical phenotype of tomato fruit, which had a flattened shape in the wild type [115]. Since GLOBE and FW3.2 (KLUH) were found to be members of the same cytochrome P450 family, the authors hypothesized that both may act similarly in regulating fruit size and shape.

During oxidative stress following pesticide application, plants use glutathione to clear excess reactive oxygen species. Brassinosteroids induce pesticide metabolism by activating GRX (glutaredoxin) gene expression through transcription factors [116,117].

Abscisic acid is considered an antagonist of brassinosteroids during fruit ripening. The abscisic acid signaling pathway consists of the family of receptor proteins PYR (PYRABACTIN RESISTANCE), PYL (PYR1-like), RCAR (regulatory components of ABA receptors), protein phosphatases PP2C, and SnRK2 kinases. There is evidence of positive regulation of abscisic acid biosynthesis by brassinosteroid signaling. BZR1 (brassinazole resistant) was found to mediate brassinosteroid signaling by promoting abscisic acid biosynthesis through direct transcriptional regulation of NCED1 [118]. Here, we showed that BIN2 negatively regulated BZR1 protein accumulation and cold tolerance by suppressing abscisic acid biosynthesis.

Suppression of PP2C3 expression in tomatoes accelerated the onset of fruit ripening and affected their glossiness by changing the external structure of the epidermis [119]. In transgenic plants, an increase in the expression of SnRK2, PYL receptors, various cutin synthesis and transfer genes, and CYP (cytochrome P) genes was observed. The role of PP2C as a negative regulator in abscisic acid signaling was further supported in [120], where alteration of PP2C5 expression affected fruit quality traits, including pericarp thickness, shape, seed number, and soluble solid content. In addition, PP2C1 silencing increased accumulation of endogenous abscisic acid and accelerated ethylene release in transgenic tomatoes compared to wild-type fruit [121]. PP2C1-RNAi lines had abnormal flowers, and pedicel abscission was impaired.

Abscisic acid homeostasis is regulated by its conjugation with glucose using uridine diphosphate glucosyltransferases (UGT). It was shown that RNAi silencing of the UGT75C1 gene significantly increases the level of expression of the CYP707A2 hydrolase gene while not affecting the expression of the key gene for abscisic acid biosynthesis NCED1 [122]. Suppression of UGT75C1 expression significantly accelerated fruit ripening by increasing abscisic acid levels and promoting early ethylene release.

The PYL9 protein has been identified as a positive regulator of abscisic acid signaling [123]. Depending on abscisic acid concentration, PYL9 can inhibit the protein phosphatase PP2C. In tomatoes overexpressing PYL9, fruit ripening was significantly accelerated due to the early release of ethylene.

The abscisic acid-induced oxidase gene DAO2 (dioxygenase for auxin oxidation) inhibited hypocotyl elongation in tomatoes, exhibiting an antagonistic role to auxins [124].

SnRK phosphorylation is mediated by the protein kinase MAPK11, thereby regulating abscisic acid biosynthesis and signaling [125].

In addition, a new transcriptional repressor of abscisic acid biosynthesis, EAD1 (ERF-associated amphiphilic repression (EAR) motif-containing ABA down-regulated), was recently discovered [126]. Although the authors have not studied the molecular mechanism of repression, its implementation is possible either through the recruitment of histone deacetylases with subsequent formation of a complex with co-suppressors or through direct or indirect binding to transcription factors.

Hydrogen sulfide counteracts the effects of ethylene during ripening. The assimilation of sulfates in chloroplasts can produce endogenous hydrogen sulfide, and the main enzymes in this process are sulfite reductases. Cytosolic hydrogen sulfide can also be generated from cysteine by cysteine desulfhydrase 1 (DES1/LCD1). Loss-of-function mutations in LCD1 in tomatoes [127] increase the expression of genes for ethylene synthesis (ACO1, ACO3, and ACS2), carotenoids (PSY1, PDS, and ZDS), and cell wall metabolism (CEL2, EXP, XTH5, PG, and TBG4). Knockout of the tomato D-cysteine desulfhydrase (DCD) gene results in increased expression of ripening-related genes, including NYC1, PAO, SGR1, PDS, PSY1, ACO1, ACS2, E4, CEL2, and EXP [128].

An attempt to understand hydrogen sulfide-mediated regulation of ripening is made in [129]. The authors suggest that the ubiquitin-protein ligase BRG3 undergoes persulfidation at two cysteine residues, leading to a decrease in ubiquitinating activity and its interaction with the repressor transcription factor WRKY71. This leads to increased binding of WRKY71 to the promoter of cyanoalanine synthase CAS1, which inhibits its transcription and, thus, prolongs fruit ripening.

There is also confirmation that hydrogen sulfide is a regulator of aging, which is noticeable in changes in the expression of chlorophyll degradation genes (NYC1, PAO, PPH, SGR1) and the aging-associated gene SAG [130].

5. Abiotic ripening factors

Fruit ripening is also regulated by signaling systems activated in response to abiotic stimuli. It has been reported that changes in light sensitivity and light-sensitive signaling in tomatoes can significantly change fruit development and quality characteristics. In this context, phytochromes act as molecular switches in response to light. Following light activation, phytochromes deactivate photomorphogenic response repressor proteins (e.g., COP1, CUL4, DDB1, DET1, and PIF).

Using RNAi silencing of the phytochrome genes PHYA, PHYB1, and PHYB2, it was shown that PHYA positively affects the differentiation and division of tomato plastids through changes in the expression of both light-dependent genes and cytokinin-dependent genes [131]. Regulators of carotenoid biosynthesis (GGPS, PSY1, and PDS) were also affected, resulting in decreased carotenoid biosynthesis during fruit ripening.

As for the repressor proteins mentioned above, their effect on ripening is also being studied. Thus, according to [132], overexpression of COP1 (CONSTITUTIVE PHOTOMORPHOGENIC) from Solanum melongena in tomatoes caused a delay in fruit ripening by 3-6 weeks. These transgenic plants showed decreased ethylene production due to suppressing the expression of the central genes of its biosynthesis ACO1, ACO3, and ACS2. The carotenoid biosynthesis genes PSY1, PDS, and ZDS were also downregulated. In [133], using the DDB1, DET1, and CYC-B genes as an example, the multiplex Target-AID (activation-induced cytidine deaminase) technique was developed. As a result, the authors obtained two lines of triple mutants, in which each gene had two-point substitutions, which showed a higher accumulation of carotenoids and lycopene compared to the wild type. In tomatoes, PIF-dependent light signaling has been reported to regulate fruit development and influence nutritional value and ripening time. Transient overexpression of PIF3 in tomato fruit resulted in decreased GGDR mRNA levels, which was inversely related to PIF3 transcript levels [134]. These data indicate that PIF3 mediates PHY-dependent regulation of tocopherol biosynthesis through transcriptional inhibition of geranylgeranyl diphosphate reductase GGDR expression in tomato fruit. Evidence shows that PIF4 can regulate hypocotyl elongation, plant growth, flowering, and leaf senescence in response to light and temperature [135]. The authors support this statement by obtaining tomatoes with RNAi-mediated knockdown of PIF4, which showed increased carotenoid content, accelerated fruit ripening time, and delayed leaf senescence. A small number of flowers and a decrease in vegetative mass were observed in such plants. Knockout of PIF3 using CRISPR/Cas9 led to the arrest of phase I of pollen mitosis, which was reflected in its non-viability [136]. Glutamate synthase (GLT1) and cell wall invertase (CWIN9), involved in auxin and sugar homeostasis, respectively, have also been shown to colocalize with PIF3 in anthers and are directly regulated by PIF3. Knockout lines of GLT1 and CWIN9 (cell wall invertase) showed a similar phenotype. VIGS-mediated silencing of the light-signaling transcription factors HY5 and PIF3 led to changes in glycoalkaloid levels in tomato leaves compared to wild type, suggesting their involvement in the regulation of target genes of glycoalkaloid metabolism [137].

While the most abundant antioxidant in tomato fruit is the lipophilic carotenoid lycopene, levels of water-soluble flavonoids (including anthocyanins) are suboptimal. Plants accumulate anthocyanins in response to various stress events such as low temperature, drought, UV radiation, intense light, and nutrient deficiency, acting as an antioxidant and photoprotective agent. The bZip transcription factor HY5 is believed to be a significant regulator of anthocyanin accumulation in plants in response to light [138,139]. However, research [140] has cast doubt on the accuracy of this statement. By creating HY5 knockout mutants, the authors demonstrated a reduced anthocyanin content, which suggests the presence of additional pathways for their synthesis independent of HY5. Indeed, eight candidate anthocyanin transcription factors have been identified.

A recent study has uncovered the function of the little-studied PHY-F. It turned out that PHY-F is a low-flux radiation sensor [141]. It forms dimers with PHYA and/or PHYB, with which it makes additive contributions to various processes of photomorphogenesis.

In addition to the phytochromes of red and far-red light receptors, there are also cryptochromes of blue light receptors – CRY1 and CRY2. Tomato lines overexpressing CRY1a showed significant accumulation of anthocyanins through the regulation of genes encoding key enzymes of anthocyanin biosynthesis (e.g., AN2, DFR (dihydroflavonol 4-reductase)) [142]. The same study showed that blue light consistently induced overexpressing tagged HY5 protein accumulation in tomatoes. In addition, it was shown that under the influence of blue radiation, repression of COP1 (CONSTITUTIVE PHOTOMORPHOGENIC) transcription was observed, which was confirmed by the creation of lines with RNAi-COP1. Ultimately, the silencing of HY5 and two anthocyanin biosynthesis genes (CHS1 (chalcone synthase) and DFR) in CRY1a lines was accompanied by a decrease in anthocyanin accumulation. Moreover, CRY1a was found to be critical for regulating starch accumulation in chloroplasts by inducing starch degradation through the transcription factor HY5 [143]. Induction of transcription of genes associated with starch degradation under the influence of blue radiation in CRY1a- or HY5-overexpressing plants was also confirmed.

It is known that in tomato, the R2R3-MYB group of factors regulating anthocyanin biosynthesis is represented by AN (ANANTHA) genes. Currently, their biological function is being actively clarified. For example, by generating loss-of-function mutants of AN2, the authors identified it as a positive regulator of anthocyanin biosynthesis in tomato vegetative tissues [144]. In addition to reduced anthocyanin content, the mutants had a dwarf phenotype. Overexpression of AN2 resulted in changes in multiple fruit qualities [145]. Thus, increased production of ethylene and increased content of anthocyanins, phenols, and flavonoids were observed. The content of aromatic volatiles such as aldehydes, phenylpropanoid derivatives, and terpene volatiles was also increased in these fruits. Thus, AN2 was shown here to regulate the transcription of genes in several metabolic pathways. Additionally, it was found that loss-of-function mutations in the AN2 ortholog in wild tomato impair anthocyanin synthesis [146]. Overexpression of ANT1 in tomatoes enriched the anthocyanins in leaves, contributing to more intense light absorption in the blue and red spectrum [147]. However, introducing knockout mutations into the AN2-like gene rather than ANT1 (ANTHOCYANIN) essentially eliminates the accumulation of anthocyanins [149,150]. It was found that AN2-like activated the expression of DFR; however, when AN1 was knocked out, anthocyanin pigmentation in the fruits was also eliminated. The AN2-like antagonist is the R3-MYB protein MYBATV. Meanwhile, a similar conclusion regarding MYBATV was made earlier [151]. It can be summarized that HY5 activates AN2-like, promotes the expression of AN1 and MYBATV, and MYBATV protein competes with AN2-like for binding to AN1 and thereby negatively regulates anthocyanin biosynthesis. Moreover, in [152], overexpression of AN2-like was found to increase jasmonic acid accumulation, activate the defense signaling pathway against Botrytis cinerea, and also increase fruit shelf life by inhibiting the expression of genes associated with the modification cell wall.

The previously mentioned dihydroflavonol 4-reductase (DFR) is involved in the reduction of dihydroflavonols to leukoanthocyanidins during the synthesis of the pigments pelargonidin, cyanidin, and delphinidin. The DFR gene in the tomato genome is represented by a single copy, which prompted its use in developing a natural genome editing marker based on homologous recombination with restoration of the DFR function [153]. DFR expression is also regulated by BBX20, which binds to its promoter region to activate expression [154].

6. System of regulation of tomato fruit ripening process

As a result of our literature review, we have presented a putative model of ripening factor regulatory pathways (Figure 4). We recognize five components of the maturation regulation system: transcription factors, hormones, epigenetics, external stimuli, and ncRNAs.

A significant contribution to regulation is provided by the ethylene-dependent pathway involving important polycistronic regulators like ethylene-sensitive genes, ethylene response genes, and the MADS-RIN complex. The fact that there are many additional factors with which RIN directly interacts suggests the existence of various ad hoc regulatory complexes consisting of several units of transcription factors. Quite often, functional redundancy is observed for them. There are 3 possible relationships of transcription factors: redundancy, additivity, and dependency. Redundancy is manifested in the functional identity of transcription factors. Additivity is associated with the provision of function through the joint contribution of each element. Direct dependence involves activation or repression of the role of one factor only after interaction with another. Moreover, autocatalytic regulation of the participants of regulatory cascades is possible.

As practice shows, turning off the function of any of the factors does not interrupt the entire cascade of ripening gene regulation but only leads to a delay phenotype. This is explained by the mechanism of gene compensation, i.e., the existence of homologs of the nonfunctional factor and/or other regulatory pathways. Indeed, we show that there are other regulators of the ethylene pathway, such as factors of the brassinosteroid and abscisic pathways. Although SA/JA pathways are absent in the proposed scheme, we do not exclude the presence of SA/JA regulators as additional maturation factors. Nevertheless, the number of studies aimed at clarifying their role in these processes does not allow us to conclude that they contribute significantly to the regulation of ripening-related genes.

Activation or deactivation of genes involved in ripening regulation can be mediated epigenetically, as discussed previously with specific examples. RIN-mediated regulation also requires interaction with promoters of lncRNAs, which are regulators of other genes, including those associated with maturation. This provides the so-called ethylene-independent regulatory pathway of maturation genes. Because epigenetic regulation and lncRNA regulation are potentially applicable to each element of regulatory cascades, their display on the scheme is redundant.

7. Future Prospects and Challenges

Significant progress has recently been made in understanding signal transduction systems and processes. The discovery of the function of genes and their regulatory systems in tomato ripening process can allow breeding tomatoes with an increased number of fruits and improved nutritional properties.

Although today there are attempts to generalize the cross-relationship of cascades of hormonal signaling pathways [155,156,157], they are only sometimes consistent with each other. They are united by the general concept of the participation of gibberellin- and auxin-sensitive genes at the stages of flowering, fertilization, and early fruit development, and ethylene-sensitive genes at the direct ripening of tomato fruits. Genes of mediator factors ensure the regulation of these processes. Of course, the proposed concepts still need to be completed. Thus, it was shown that ABA and BS are involved in all stages of tomato reproductive processes. It was also demonstrated with RIN and CNR that established hypotheses are not always perfect. The expanding understanding of the influence of epigenetic changes, microRNAs, and external factors on signaling regulation is also worth mentioning.

All this reminds us of a puzzle: to start putting it together, you need to find the edge of the image; otherwise, it can take infinite time to compare individual elements. But signaling pathways do not have “edges”; they must be created manually. One of the modern approaches to solving problems in this area is the use of a wide range of genetic engineering and bioengineering methods, including, among other things, the collection and processing of bioinformatics data, new sequencing technologies, targeted genome editing using CRISPR technology [158,159,160,161,162,163,164,165], use of base and prime editing for precision gene correction [166,167,168], creation of unique and simple markers for detection of transgenes [169], application transcriptomics, metabolomics and other omics technologies.

However, even here, some difficulties arise. For example, orthologous genes often have different regulatory mechanisms and are unreliable predictors of expression in related species. In addition, not all studies are field-validated, which reduces the significance of the data obtained using transgenic plants. Ideal and, at the same time, simplified laboratory conditions cannot determine all the possible subtleties of gene expression and their regulation. Moreover, changes in the expression of regulatory genes do not always entail significant changes in the transcriptome due to the possible presence in the genome of paralogs of the genes being studied. For genes higher in the regulatory cascades, pleiotropic changes occur. It is worth mentioning that the reproduction processes are discrete and appear in various tissues with excellent synchronization, determined by the life cycle stage, so their identification and description are difficult.

Undoubtedly, the existing conclusions about the regulation of plant life cycle processes obtained on model objects such as the tomato, although extensive, still need to be completed. In the future, genetic approaches will continue to make essential contributions to identifying new candidate genes involved in tomato reproductive signaling cascades. This may open up a broader cluster of regulatory signaling networks involving currently unknown factors and stimuli.

As for the perspectives of the practical application of the study of the considered ripening-related genes, these are improving fruit quality due to increased nutrient content, accelerated ripening, prolonged shelf life, and much more.

8. Conclusions

This review examined recent advances in studying tomato ripening factors using various genetic approaches. Modern research into the factors of tomato fruit ripening continues to expand our understanding of the molecular basis and physiological mechanisms of these processes, which has significant applications for improving breeding methods and growing new varieties with enhanced phenotypic traits that meet the requirements of the modern agricultural industry and consumer demand.

The abundance of current scientific reports cited in this review article reflects the relative ease of studying tomato and potential biotechnological improvements and the breadth of approaches and methods. Despite significant advances, a large number of biochemical pathways, the involvement of hundreds of genes in the fruit ripening process, and fine regulation involving transcription factors and microRNAs, at present, it is premature to talk about a complete understanding of the described processes. The complex new data presented in the review and proposed model of gene regulation will allow us to understand the principle of tomato fruit ripening better and complement the overall knowledge picture.