Improvement of functional quality is a complex task because the accumulation of any antioxidant compound is the result of complex metabolic processes. Consequently, the increase in the available information on antioxidant biosynthetic pathways (enzymes involved and regulation mechanisms) and the identification of mutant genotypes with beneficial pathway alterations for the antioxidant accumulation is essential to obtain higher precision and better results in the development of breeding programs. Different breeding strategies can be followed with this purpose, and they all depend on the existence of variability for the accumulation of bioactive compounds in the cultivated species or wild relatives. Additionally, a different approach can be followed via genetic engineering. Advances in the improvement of carotenoid and polyphenol composition of tomato are reviewed herein.
In tomato, breeding for fruit colour was one of the first fruit quality objectives demanded by markets. As this feature is conferred by carotenoid pigments (mainly lycopene and β-carotene) with important antioxidant properties, the improvement of fruit colour has indirectly led to the improvement of tomato nutritional and functional value. This situation justifies that breeding for improved carotenoid content is far more advanced than any other bioactive compounds. The high colour variability present in genus Solanum
allowed the first studies of diverse genetic variants (natural mutants) for this attribute (Table 2
). Studies on tomato natural mutants for carotenoid accumulation and on tomato transgenic plants with different carotenoid biosynthetic pathway alterations has allowed a detailed knowledge of the complete tomato carotenoid biosynthesis pathway (Figure 1
) and its regulation strategies.
In common red tomatoes, from the start of the maturing process (breaker stage) a high increment in synthesis of the enzymes PSY [101
], PDS [102
] and CRTISO [103
] occurs, which results in a high increment in the all-trans-lycopene synthesis (marked with thickness arrows in Figure 1
). At the same time, a strong repression of the synthesis of lycopene cyclases (enzymes involved in the formation of 6C cyclic end groups) occurs [104
]. Usually, in common red tomatoes, only lycopene-β-cyclases, LCY-B [102
] and the chromoplast specific CYC-B [105
] enzymes are present. The drastic diminution of the lycopene-β-cyclases entails that only a low amount of β-carotene synthesis at the expense of lycopene can be achieved (Figure 1
). Consequently, the prominent carotenoid in mature fruits will be all-trans
-lycopene, which confers to the ripe fruits its typical red colour. Several natural mutant genes with altered steps in this biosynthetic pathway have been identified (Table 2
). As a result, altered fruit carotenoid profile and fruit colours are obtained [103
]. In green tissues of the plant, the carotenoid biosynthesis pathway does not stop with lycopene accumulation, but it continues with the xanthophyll biosynthesis pathway (Figure 1
) and neoxanthin would be the last product synthesized, deriving in the abscisic acid (ABA) synthesis pathway [24
Regulation of the carotenoid biosynthesis process can be done at three different levels. The first level consists in the regulation of the initial amount of the precursor of the synthesis of carotenoids (pre-pathway substrate regulation, Figure 1
), the isopentenyl diphosphate (IPP), which determines the total amount of carotenoids that can be synthesised. IPP may arise at some developmental stages partly from the citoplasmic mevalonic (MVA) pathway [108
], but it is mainly synthesised through the methylerythritol-4-phosphate (MEP) pathway, apparently bound to the plastidic compartment [109
]. The first enzyme of MEP pathway, 1-deoxy-D-xylulose-5-phosphate synthase (DXS), has been proved to catalyse the first regulatory step in carotenoid biosynthesis [110
] and, consequently, highly influences the final amount of carotenoids. This pre-pathway substrate regulation also can be altered by a reduction of ABA synthesis which has as consequence an increase of plastid division enabling higher biosynthesis and accumulation of carotenoids [111
]. This was observed in high pigment-3 (hp-3) mutants. These mutants coding a defective zeaxanthin epoxidase (ZE) enzyme (Xanthophyll biosynthesis pathway, Figure 1
) which arrest ABA synthesis and, as explained above, increased carotenoid accumulation.
A second level of regulation consists in the hormonal growth regulation of the fruit ripening process (light independent regulation, Figure 1
) which also plays a major role in the control of the described carotenoid synthesis pathway. During the ripening process, ethylene has a strong positive control of the increment in the mRNA levels for the lycopene-producing enzymes phytoene synthase (PSY) and phytoene desaturase (PDS), at the same time, the mRNA levels of the genes for the lycopene β- and ε-cyclases diminish and completely disappear [104
]. This explains that mutations affecting ethylene synthesis or perception, such as long life mutations rin
..., not only delay the normal ripening process also result in altered carotenoid content [112
Finally, there is a third level of regulation in which fruit localized phytochromes also regulate the extent of carotenoid accumulation [113
]. Phytochrome response to light presence enables carotenoid synthesis that stops in darkness, thus conditioning day-cyclical biosynthesis periods (light-dependent regulation box, Figure 1
). Several natural mutants present alterations in the normal phytochrome regulation with enhanced global carotenoid content (Table 2
): the mutation high pigment-1
), and the allelic hp-1w
results in the plant acting as perceiving continuously the light [114
]. The mutation high pigment-2
), and the allelic hp-2j
], affects the photomorphogenesis regulatory gene TDET1, and also affects the light signal-transduction machinery [116
]. Another mutant, Intense pigment (Ip
) is implicated in a promotion of phytochrome signal amplification [117
All the breeding strategies targeted to improve carotenoid content in tomato try to achieve a gene combination enabling a higher accumulation of one or more carotenoids in a desired genotype with good agronomic performance. Two main strategies can be used to do it: one is the use of germplasm with a high potential to accumulate one or more antioxidants as a gene donor in order to transfer them by conventional breeding programs and the other the use of advanced biotechnology to transfer foreign genes (genetic engineering) to allow a beneficial biosynthesis alteration of the desired antioxidant compound.
Both strategies can be complementary, and in fact genetic engineering can be used to avoid the negative side effects of certain genes changing their promotor (reviewed by Cebolla-Cornejo et al.
]). Nevertheless, the level of public scepticism in certain regions (e.g., Europe) hinders the commercialization of transgenic varieties [3
] and commercial approaches are right now based in conventional breeding programs.
In summary, the first approach involved the exploitation of natural diversity present in the genus (mainly genotypes with natural gene mutations identified in many studies and described before) as their use as donor parents in conventional breeding programs following different hybridization strategies and selection generations. Following this approach, some of the more successful fresh tomato cultivars developed carrying the Beta, B, gene with its modifier Beta-modifier, moB (Figure 1
) have been Caro Red [119
] and Caro Rich [120
] and the processing tomato cultivar Caro beta [121
]. These cultivars show orange fruits with β-carotene contents up to 5 mg 100 g−1
fw (roughly up to 10-fold the normal content in standard red cultivars). Unfortunately, these orange fruited cultivars with high β-carotene content have not been commercially successful, as consumers seem to prefer red tomato fruits.
Other approaches focused in the increase of lycopene content have obtained improved cultivars using mutants old gold
) and old gold crimson
) which inhibits the synthesis of β-carotene by cyclisation of lycopene resulting in higher lycopene contents (up to 30% of normal content) and lower β-carotene accumulations [122
]. Tomato cultivars of the crimson type have been successfully commercialized due to their intense red pigmentation even when the fruits are not completely ripe. One of the problems of the use of mutants affecting single steps of the biosynthesis pathway is that the increase in the level of one carotenoid is obtained at the expense of another, thus little effect can be expected for total carotenoid content. Thus, mutants affecting the regulation of the pathway would have a more dramatic effect. In fact, better results have been obtained with cultivars carrying both crimson and high pigment genes (hp-1
alleles), as increments in the lycopene content up to 3- to 4-fold of common cultivars has been obtained [124
]. Although, some deleterious effects on seed germination, plant vigour and yield have been associated with the high pigment mutants. The best expectations where deposited in the use of Intense pigment
) mutant gene. This gene allows carotenoid accumulation similar to those of hp
genes, but with lower deleterious effects and it is dominant (commercial hybrid development more interesting than with recessive genes as hp
Regarding the second breeding strategy, the high advances in the knowledge of the carotenoid biosynthesis (metabolic pathway, precursors and regulation mechanisms) allowed the use of this information to obtain several experimental transgenic tomato lines with modified genes controlling some biosynthetic steeps. Most of this works try to emulate the carotenoid biosynthetic performance of some of the natural mutants identified, and some of them also try to avoid undesirable side-effects found in these mutants.
One approach used has been the modification of isoprenoid precursor’s pathway (pre-pathway substrate regulation) intervening in the both mevalonate (MVA; [109
]) and methylerythritol-4-phosphate (MEP; [126
]) pathways (Figure 1
) trying to increase the total amount of carotenoids increasing the levels of the precursor. Only transformation with gene encoding 1-deoxy-D-xylulose-5-phosphate synthase (DXS) from Escherichia coli
to increase MEP pathway showed interesting results (2.2 times higher β-carotene content).
The most explored approach has been placed in the development of transgenic lines with altered expression of the most important enzymes in the carotenoid synthetic pathway with the objective to alter the carotenoid profile and to increase the content of certain of interesting carotenoids. The focus was put on the enzymes phytoene synthase [127
], phytoene desaturase [129
] and mainly on lycopene cyclases [105
] due to their role in the regulation of the whole pathway and its role in the partition of the main carotenoids lycopene and β-carotene. Several strategies have been used: constitutive or selective expression of foreign genes from several species and overexpression or repression of target genes by several mechanisms, but results obtained were contradictory. To avoid these problems, in recent years, a more refined strategy targeted to obtain a better expression of transgenes using more selective and specific promoters have been explored. In this sense, the characterization of the tomato PDS promoter [135
], as well as S. habrochaites
lycopene β-cyclase (CYC-B) promoter [136
] bring information to enhance the use of new transgenes. Other efforts were targeted to use promoters which allows a selective expression of transgenes in fruit tissues and adequate developmental stages as occurs with fruit specific promoters such as ethylene response genes E8 and E4m polygalacturonase and lipoxygenase, mainly acting in the late-ripening stage or the LA22CD07 and LesAffx.6852.1.Sl at in green and red-ripening fruits [137
Finally, the third approach used is based on the modification of some of the regulatory mechanisms of the carotenoid biosynthesis pathway. In this sense, some transgenic experimental lines tried to emulate the performance of high pigments mutants hp-1
] and hp-2
]. In this last case, the strategy used involved the suppression of photomorphogenesis regulatory gene TDET1 in fruits (using fruit specific promoters combined with a RNA interference approach based in inverted repeat constructs) to reduce negative collateral vegetative effects, and gave interesting results. Other works were focused in the development of transgenic experimental lines simulating performance of hp-3
mutants. In this sense, the down-regulation of ABA synthesis during ripening using an RNAi construct of the SINCED1
gene driven by the fruit specific promoter E8 [139
], caused a reduction in ABA concentration, an increase of ethylene production and resulted in an increment in lycopene and β-carotene contents.
The polyphenol content of tomatoes has gained importance during the last decade. Consequently, the achievements of breeding efforts still lag behind those obtained for carotenoids. Nevertheless, quite a lot of information is available regarding polyphenol biosynthesis in this crop. In fact, the polyphenol metabolic pathway has also been ascertained. Several transcription factors related with the regulation of polyphenol biosynthesis have been identified, but a lot of information regarding the spatial accumulation of polyphenols in the fruit and how it can be reverted is still required.
The phenylpropanoid biosynthetic pathway is the first step in the polyphenol biosynthesis and it uses the amino acid phenylalanine from the shikimate pathway as initial substrate (Figure 2
). This biosynthetic pathway is common for two of the main classes of tomato polyphenols: hydroxycinnamic acids and flavonoids. The first step involves the conversion of phenylalanine into trans
-cinnamic acid using the enzyme phenylalanine ammonia lyase (PAL). The enzyme cinnamate 4-hydroxylase (C4H) catalyses the conversion of the resulting product into p
-coumaric acid. At this point, the flavonoid biosynthetic pathway continues with the conversion of p
-coumaric acid into 4-coumaroyl-CoA
as a result of the action of 4-coumarate-CoA ligase (4CL) [141
Meanwhile the hydroxycinnamic acid pathway continues with the transformation of p
-coumaric acid into caffeic acid catalysed by the p-coumarate 3-hydroxylase (C3H) [142
]. Chlorogenic acid, one of the main polyphenols present in tomato [26
], is formed from caffeoyl-CoA, which is transesterificated with quinic acid by hydroxycinnamoyl-Coenzyme A:quinate hydroxycinnamoyl transferase, HQT [143
]. Caffeoyl-CoA would be obtained from 4-coumaroyl-CoA in three steps involving the successive activities of cinnamoyl CoA shikimate/quinate transferase (HCT), p
-coumaroyl ester 3-hydroxylase (C3H) and HCT [144
]. Finally, the third main hydroxycinnamic acid in tomato, ferulic acid, would be obtained from caffeic acid with the enzyme caffeic acid O
-methyltransferase (COMT) [142
On the other hand, the core flavonoid biosynthetic pathway starts with the conversion of the resulting 4-coumaroyl-CoA
from the phenylpropanoid pathway into the yellow-coloured naringenin chalcone [141
]. This key reaction is performed by the enzyme chalcone synthase (CHS) which begins with the condensation of one molecule of 4-coumaroyl-CoA
with three molecules of malonyl-CoA
. In most plants, including tomato, chalcones are not the end-product of the pathway. The enzyme chalcone isomerase (CHI) isomerizes naringenin chalcone into the flavanone naringenin. Finally, the core flavonoid intermediates pathway finishes with the formation of the dihydroflavonol dihydrokaempferol as a result of the action of the flavanone 3-hydroxylase (F3H). From this central intermediate, the flavonoid biosynthetic pathway diverges into several side branches, each resulting in a different class of flavonoid [141
One of the most important classes of flavonoids in tomato are flavonols [26
], which are synthesised from dihydrokaempferol. The three more important in tomato are kaempferol, quercetin, and myricetin. The three are formed from the corresponding dihydroflavonol, and have dihydrokaempferol as a starting point.
Flavonol synthase (FLS) catalyses the direct conversion of dihydrokaempferol into kaempferol. On the other hand, with dihydrokaempferol as substrate, the enzymes flavonoid 3’-hydroxylase (F3’H) and FLS would produce respectively dihydroquercetin and quercetin [147
]. Quercetin is especially important in tomato, as it is the base for the formation of rutin. This quercetin glycoside is quite abundant in tomato [32
], and it is obtained from quercetin by the action of the enzymes glucosyltransferase (GTF) and rhamnosyltransferase (RTF) [146
The third flavonol, myricetin, would be derived in from dihydroquercetin in two steps. The first one catalysed by flavonoid 3’5’-hydroxylase, F3’5’H [145
], would produce dihydromyricentin which would be converted into myricetin by FLS. Moreover, myricetin could also be directly obtained from dihydrokaempferol by means of the action of F3’5’H and FLS [145
Anthocyanins are not naturally accumulated in the tomato fruit, but as it will be shown, interspecific crosses can restore this pathway in the fruit. Their synthesis would start from the three commented dihydroflavonols. Dihydroflavonol 4-reductase (DFR) would catalyse the conversion of dihydroflavonols into flavan-3,4-diols, which would be then transformed into anthocyanidins by anthocyanidin synthase (ANS). The final step would require the addition of sugars to form anthocyanins, which are anthocyanidin glycosides [141
]. The main anthocyanins present in tomato would be derived from three anthocyanidins: delphidin, that would be synthesised following this scheme from dihydromyricetin, cyaniding from dihydroquercetin and pelargonidin from dihydrokaempferol. Other anthocyanidins would be derived from these. For example, delphinidin can be methylated on its 3’ hydroxyl group to form petunidin or on both its 3’ and 5’ hydroxyl groups to form malvidin [150
As in the case of carotenoids, several mutants have been identified regarding polyphenol accumulation (Table 3
), and those more used in breeding programs are related to the regulation of the pathway. It seems that CHI plays a central role in the rate determining step in the production of flavonols [151
]. In fact, its over-expression increases dramatically the levels of flavonols at the expense of naringenin chalcone [33
]. Ballester et al.
] suggested that upon ripening an increase in the expression CHS expression and a coordinated decrease in the expression of CHI would result in the accumulation of naringenin chalcone and in a limitation of flavonol contents. Previously, Willits et al.
] suggested that the lack of expression of CHI in the peel of the fruit, probably caused by a mutation in a fruit specific promoter would explain the high levels of naringenin chalcone in this tissue. The authors also assumed that cultivated tomato would have lost the expression of CHI in the peel in an early step of the domestication process.
Naringenin chalcone is one of the prominent polyphenols in tomato. Its accumulation in the peel gives a yellow colour that in combination with red flesh results in an external red colour of the fruit. Several tomato landraces are characterized by an external pink colour, resulting from the lack of accumulation of naringenin chalcone and a consequent transparent peel. The evaluation of introgression lines of Solanum chmielewskii
in tomato enabled the identification of the gene responsible for this mutation (yellow
), initially described in 1925 [154
]. This gene encodes SlMYB12, a transcription factor involved in the regulation of the phenylpropanoid and/or flavonoid pathway [152
The expression of the Arabidopsis
form AtMYB12 in tomato induced primary and secondary metabolism, binding directly to the promoters of different genes encoding enzymes of the primary metabolism such as 3-deoxy-D-arabino-heptulosonate 7-phosphate synthase, DAHPS, and plastidial enolase, ENO [155
]. DAHPS is a key determinant if the flow of the shikimate pathway. A first step needed for the accumulation of phenylalanine, a substrate necessary for the flavonoid pathway. Additionally, it would also bind to promoters of genes related enzymes participating in the flavonoid pathway (PAL5A, PAL5C, PAL5D, CHS1 and F3H). As result a dramatic increase in flavonol and hydroxycinnamates is observed, reaching up to 10% of fruit dry weight.
Pandey et al.
] also expressed constitutively AtMYB15, obtaining similar results. In this case, 305 unigenes were upregulated in the fruit tissue and 419 downregulated. Specifically, several enzymes of the flavonoid pathway were upregulated, especially CHS (300-fold enhanced expression in the fruit) and FLS (300-fold enhanced expression in the fruit). Additionally, genes involved in ethylene biosynthesis and signalling, ABA, auxin and Ga signalling were also modulated. Primary metabolism was also altered, and a differential regulation of genes involved in aromatic amino acid biosynthesis and carbohydrate metabolism was observed. Probably this modulation may be related to the elevated demand of C-source to support the enhanced biosynthesis of polyphenols.
Recent transcriptome analysis has identified additional transcription factors involved in the regulation of the pathway [157
]. At least 20 transcription factors would correlate with expression of genes participating in the flavonoid biosynthesis pathway. As expected, SlMYB12 is included in this set. Other examples include LIM, which is highly correlated with the expression of genes involved in both the biosynthesis of ascorbic acid and flavonoids, and other MYB and bHLH genes.
The expression of endogenous genes encoding key enzymes of the flavonoid pathway (PAL, CHS, CHI, F3H and FLS) in tomato pericarp and collumela tissues has been found to be under detection limits [146
]. This lack of expression would explain the low amounts of flavonoids found in these tissues. The expression of these genes can be achieved. For example, an overall increase in flavonol accumulation in the whole fruit was achieved with the expression Lc and C1 transcription factors from maize [151
]. In this case, a clear over-expression of genes encoding CHS and F3H was observed, suggesting that the expression of these two genes would be necessary to improve flavonoid accumulation in tomato flesh. Other studies confirmed that the expression of the gene encoding FLS would also be required alter all [34
]. In fact, the joint over-expression of genes encoding CHS and FLS would be required to pull the carbon flux towards the accumulation of flavonols [146
Regarding the light regulation of the pathway, Giuntini et al.
] showed that UV-B radiation differentially affects the expression of genes of the flavonoid pathway and results in altered content of flavonoids and hydroxycinnamic acids. The conventional cultivar Esperanza showed higher levels of naringenin chalcone, quercetin and rutin and of sinapic, caffeic, ferulic and p
-coumaric acids in the flesh of UV-B shielded fruits at the red ripe stage, while in the high lycopene cultivar DRW5981 limited effects were observed at this stage, especially for flavonoid content. In the conventional cultivar Esperanza, the expression of genes encoding CHS and CHI expression was higher and that corresponding to F3H and F3’H was reduced in UV-B shielded fruits. In both cultivars, carotenoid accumulation followed the same response of flavonoid accumulation. In the high lycopene cultivar DRW5981 several genes of the flavonoid pathway were upregulated in the fruit flesh with the exception of CHI
. As a result a dramatic increase in naringenin chalcone and a modest increase in quercetin were observed.
As commented in the carotenoid section, the high pigment
) tomato mutants also affect the light regulation of flavonoid synthesis. This overproduction of bioactive compounds is also associated with increased plastid biogenesis and therefore plastid-accumulating metabolites would be expected (reviewed by Azari et al.
The increase in the accumulation of flavonoids in conventional breeding programs, excluding genetic engineering has offered limited results. In fact, the analysis of different tomato germplasm for flavonol accumulation offered limited levels of variation (up to 10-fold). Quercetin in hydrolysed peel extracts varied between 6.3 and 64.9 μg g−1
fw, but the highest levels only represented a 2.5-fold increase compared to a standard commercial variety [34
]. Similar levels of variation have been found by other authors, with the highest levels been found in the smaller cherry tomato fruits originating from sunny climates [35
]. The limited variation found in the cultivated tomato suggested that it would be difficult to improve flavonol accumulation via conventional breeding.
Nevertheless, the use of the primary gene pool has enabled the identification and use of wild species from the Solanum
as sources of variation. Following this approach, tomato flavonoid content has been increased using the wild tomato species Solanum pennellii
Correl, in order to restore the flavonoid pathway in fruit flesh. With this objective, germplasm expressing chalcone isomerase in the flesh was selected and used in the development of hybrids with the cultivated species with higher levels of quercetin diglycoside [153
Something similar happened with the improvement of anthocyanins in tomato. In fact, in the cultivated species anthocyanins are not produced in the fruit, but as a result of interspecific crossed in breeding programs, several mutants with anthocyanin accumulation in the fruit have been identified. The three mutants that can lead to increased anthocyanin content in the peel of the fruit are Anthocyanin fruit (Aft) Aubergine (abg) and atroviolacea (atv).
The Anthocyanin fruit
, mutant was identified in crosses with Solanum chilense
Dunal. This gene is located in chromosome 10 and its presence in tomato leads to increased levels of delphinidin, malvidin and petunidin, as well as higher levels of the flavonols quercetin, 3.6-fold, and kaempferol, 2.7-fold [37
Sapir et al.
] proved that Aft
fruits not only showed high levels of anthocyanins in the skin and outer pericarp of the fruit, but also of the flavonols quercetin and kaempferol. They also showed that Aft
is encoded by a single locus on chromosome 10 fully associated with Anthocyacin1
was discovered in a T-DNA insertional mutagenesis program and was identified as a MYB transcription factor. Vegetative tissues of Slant1
showed intense purple colour and fruits displayed purple spotting on the epidermis and pericarp. The overexpression of Slant1
upregulated genes encoding enzymes of early and later steps of anthocyanidin biosynthesis as well as genes involved in the glycosylation and transport of anthocyanins into the vacuole [160
]. It seems that the original allele from S. chilense ScAnt1
would be more efficient in the production of anthocyanins than the tomato counterpart [161
paralog gene SlAn2
, similar to Petunia anthocyanin2
, is found in chromosome 10 and has also been related with the Aft
encodes an R2R3-MYB transcription factor and its overexpression in tomato results in increased anthocyanin accumulation in fruit peel [162
]. Additionally, overexpressing fruits display ripening related phenotypes, with enhanced ethylene levels, reduced carotenoid accumulation and faster fruit softening. In fact, the authors found a concomitant accumulation of Rin
) transcripts suggesting that the functions of SlAn2
may be related. It has been recently studied that between SlAnt1
only the latter acts as a positive regulator of anthocyanin synthesis in vegetative tissues under high light or low temperature conditions [163
) mutant was identified in a cross with Solanum lycopersicoides
]. This gene also relies in the chromosome 10 and it may be allelic to Aft
. The difficulties in the management of S. lycopersicoides
introgressions has hindered the development of allelic studies to confirm or discard this relation. In this sense, a paracentric inversion has been identified in the long arm of chromosome 10 of S. lycopersicoides
, resulting in the absence of recombination events in this segment [165
The recessive gene atroviolacea
), mapping in chromosome 7, was identified in a segregant population S. cheesmaniae
(L. Riley) Fosberg [166
]. It has a strongest effect on the accumulation of anthocyanins of vegetative tissues and has a limited effect on the fruit. The atv
mutant shows an exaggerated response in the red broad band light, suggesting specificity for the phytochrome phyB1 high irradiance response pathway [167
]. Thus, this mutation per se has no important effect on the accumulation of anthocyanins on the fruit. Nonetheless, it has been proved that in the double homozygous mutants Aft Aft atv atv
a synergistic effect of both genes arises on the transcription of specific genes of the anthocyanin pathway, resulting in higher anthocyanin contents than in the individual mutants [168
]. This effect also applies to the combination Abg
. In fact the best combinations to improve anthocyanin content in the fruit include Abg atv atv
and Aft Aft atv atv
in small fruits, with contents up to 415 and 116 mg 100g-1
fw respectively in the epidermis and subepidermis [36
Another mutant, anthocyanin free
), was identified in the fifties as a mutagenesis variant characterized by the lack of anthocyanin production in all plant tissues. Recently, Kang et al.
] showed that af
encodes SlCHI1, but complementation assays demonstrated that SlCHI1 not only complements flavonoid synthesis in the af
mutant, but also complements a defect in terpenoid production, suggesting a link between both pathways.
None of these mutants result in the accumulation of anthocyanins in the fruit flesh. Alternatively, tomatoes with purple flesh have been obtained via genetic engineering. In this sense, the transcription Del
have been involved in the activation of the flavonoid pathway in tomato [147
], as the expression of these factors from Antirrhinum
resulted in the accumulation of purple anthocyanins in both peel and flesh. It seems that these factors would stimulate the transcription of the genes encoding PAL, CHI and F3’5’H. This upregulation of the flavonoid pathway and the opening of the anthocyanin gate (F3’5’H) resulted in anthocyanin levels up to 3 mg g−1
fw. Another possible success of the combination of both transcription factors may be related with the upregulation of genes involved in the side chain modification of anthocyanins and genes associated with the transport and accumulation in vacuoles [170
]. Later works with the same genes obtained higher levels up 5.2 mg g−1
dry weight (dw), with petunidin-3-(trans
-coumaroyl)-rutinoside-5-glucoside and delphinidin-3-(trans
-coumaroyl)-rutinoside-5-glucoside representing an 86% of the total anthocyanins [171
The enzyme F3’5’H would be in fact a key enzyme in restoring the accumulation of anthocyanins in tomato. In fact, the absence of anthocyanins in LC/C1 fruits was attributable primarily to an insufficient expression of F3’5’H, in combination with a strong preference of the tomato dihydroflavonol reductase (DFR) to use dihydromyricetin as a substrate [151