Gene Expression of Flavanone 3-Hydroxylase (F3H), Anthocyanidin Synthase (ANS), and p-Coumaroyl Ester 3-Hydroxylase (C3H) in Tzimbalo Fruit

The current investigation emphasizes the expression of candidate genes for future fruit quality improvement. This study aims to describe morphological variation on Solanum caripense Dunal (tzimbalo) ecotypes; identify gene expression of F3H and ANS and analyze gene expression of C3H. This study employed Ecuadorian (BIO) and Peruvian (IBT) as samples of the study. Morphological descriptors for Solanum muricatum Aiton were used in this study. RNA was isolated for identification of F3H and ANS transcripts in BIO-Ltg1 and BIO-Cyb1 through reverse transcription followed by semiquantitative PCR (RT-PCR). C3H relative expression was analyzed in IBT-Lib1 for zero, five and fourteen days under the influence of controlled conditions (10 ± 2 °C; 16 h day/8 h night) through reverse transcription followed by quantitative PCR (RT-qPCR). The cophenetic correlation (0.88) of conglomerate analysis (CA) pointed out good similarity for Ecuadorian ecotypes and two subgroups for Peruvian ecotypes. The first three principal components (PC) explained qualitatively 71.39% and quantitatively 81.34% of total variation; Fr-Flavour, Se-Diameter, Fl-CorollaColour, Frstripes, Fr-Length, Fr-PlacentLength, and Fr-PlacentBreadth were characters that contributed more to the variability. The expression of F3H was identified in BIO-Ltg1. The expression of ANS was similar (BIO-Ltg1→48.20 ng·μL; BIO-Cyb1→36.19 ng·μL). The mean fold change value in C3H expression was 3.32, 4.52, and 6.24 for zero, five, and fourteen days; C3H transcripts level was significantly different and increased 2.92 units after fourteen days. These results demonstrate the expression of F3H and ANS in BIOLtg1 and BIO-Cyb1, differential expression of C3H in IBT-Lib1, and focus the nutritional value of tzimbalo fruit. Keywords— Reverse transcription; wild relatives; fruit quality; improvement; commercial potential. Manuscript received 9 Oct. 2020; revised 12 Jan. 2021; accepted 7 Feb. 2021. Date of publication 30 Apr. 2021. IJASEIT is licensed under a Creative Commons Attribution-Share Alike 4.0 International License.


I. INTRODUCTION
The S. caripense is a phylogenetically complex plant [1], commonly named tzimbalo [2], mostly wild, and widely distributed in Ecuador and Peru [3], [4], also present in Colombia, Bolivia, Venezuela, Costa Rica and Panama [5]. This species is an herbaceous plant, native of the Andean region [6], growing on damp places of highlands [7], even until 3800 m.a.s.l. [3]. The tzimbalo plant is compact, and similar to the morphology of S. muricatum (pepino) this species produces vertical branches [8], is considered, with high likelihood, a close relative to the pepino, and its ancestor [9], [10], due to chromosome similarities and the possibility for harvesting of the obtained interspecific hybrids [11]; it belongs to the section Basarthrum, series Caripensia, and complex Caripense [1], [12], [13].
The fruit of S. caripense has many seeds, its high germination percentage [14], [15] lets the investigations to discard the presence of primary dormancy and physical lethargy [16]. In contrast to seeds of other wild species of the genus Solanum [17], [18], the fruit of S. caripense  contains significantly more sucrose, vitamin C [19], and minerals [20], compared to modern cultivated varieties of S. muricatum, and other species of the series Caripensia.
Despite the great potential of S. caripense for interspecific gene flow towards related commercial crops as Solanum tuberosum L., Solanum lycopersicon L. and S. muricatum, there are limited genomic studies of this species [21], [22]. Modern biotechnological tools help overcome production, commercialization, and export limitations, such as gene expression analysis [23] and genetic transformation [24].
Plant breeding with S. caripense and S. muricatum accessions is carried out through backcrosses to the Pepino. The estimation of sucrose and ascorbic acid concentrations [25], and others decided by candidate genes involved in fruit quality improvements such as anthocyanins and chlorogenic acid contents [22], leads to heritability studies.
Relative quantification measures the expression levels and their relative change. It determines the change in certain mRNA levels of a gene across multiple samples. It does not require a calibration curve or standards with known concentration, and the reference genes can be any transcript, as long as its sequence is known [26]. The measures aim to express relative quantities of the unit used is arbitrary, and its quantities can be compared across multiple RT-qPCR experiments [27], [28]. Relative quantification assumes an optimal doubling of the complementary DNA (cDNA) of interest during each performed qPCR cycle [29]; its model is derived from the exponential nature of the PCR; the amplification efficiency is close to one for the 2 ∆∆ method, the target gene quantity is normalized using reference reactions, and it is relative to calibrator reactions [30].
The anthocyanin biosynthetic pathway is already characterized and comprises a well-conserved mechanism in many plants [31]. It is an extension of the general pathway for flavonoid synthesis [32]. The primary anthocyanidin delphinidin shows violet/blue hues [33]. Single methylation of delphinidin results in petunidin and double methylation in malvidin; anthocyanins based on delphinidin are found in purple tissues of Solanum melongena L. and Capsicum spp. The expression of genes involved in the accumulation of anthocyanins covering specific tissues during certain stages of development can be stimulated by exposure to white light and low temperature [22], [34], [35]. Anthocyanins are phenolic compounds or secondary metabolites of the flavonoid subclass, soluble in water and important due to its antioxidant ability [36].
Phenolic compounds cause the high antioxidant activity in S. melongena [37], [38], and S. tuberosum [39], these are hydroxycinnamic acid (HCA) conjugates synthesized by phenylalanine conversion into cinnamic acid. Chlorogenic acid (CGA, 5-O-caffeoyl-quinic acid) is an HCA conjugate that reaches 70% and exceeds over 95% of phenolic content totality. Great diversity is observed in the total content of phenolic and CGA concentrations, caused by genetic and environmental factors. Molecular breeding for high CGA content, low polyphenol oxidase (PPO) activity, and consequently low degree of browning helps develop improved varieties for higher bioactive properties [37]. A candidate gene approach is promising for this purpose, given that genes involved in the biosynthetic pathway of CGA are well characterized [39]. In the Solanaceae, the abundance of CGA is strongly associated with different genes of its biosynthetic pathway [40], [41]. The development and storage stage influences gene expression and phenolic content [42]. Postharvest conditions pretend to prolong shelf life and increase the agronomic quality of vegetal products.
This work aims at contributing with biotechnological tools for plant breeding programs and emphasizes the expression of S. caripense genes, which belong to biosynthetic pathways of anthocyanins and chlorogenic acid; with a future propose of innovating the local production, improving the fruit quality, and converting this species into a novel alternative for consumption and derivative uses. The specific objectives were to 1) describe a morphological variation on S. caripense ecotypes; 2) identify gene expression of F3H and ANS associated with anthocyanins; 3) analyze gene expression of C3H associated with chlorogenic acid. Morphological description was performed using descriptors for S. muricatum; total RNA was isolated from S. caripense for the identification of F3H and ANS transcripts in BIO-Ltg1 and BIO-Cyb1 through RT-PCR; the relative expression of C3H was analyzed in IBT-Lib1 for zero, five, and fourteen days under the influence of controlled temperature (10 ± 2 °C) and photoperiod (16 h day/8 h night) through RT-qPCR. The studied genes belong to biosynthetic pathways that codify beneficial human enzymes, with industrial potential [43], due to their biological activities and antioxidant properties.

A. Plant Material
Regions, departments, and provinces related to the geographic distribution of S. caripense plants were taken as reference [4], [5]; individual plants were identified and in situ described on mostly wild S. caripense ecotypes (Table I).

B. Morphological Description by CA and PCA
Morphological descriptors for S. muricatum and wild related species were used (Table II) [45]; descriptors of plant (P), stem (St), leaf (L), inflorescence (I), flower (Fl), fruit (Fr) and seed (Se) were evaluated; mode and mean values were obtained for three observations per plant, differentiating between qualitative and quantitative variables, respectively [46], [47]; CA and PCA were performed.

C. Gene Expression by RT-PCR and RT-qPCR
Based on S. melongena sequences for F3H, ANS [22], [48], C3H genes [37]; and 5.8S rRNA [49], primers were synthetized to be used in S. caripense (Table III). Total RNA was isolated from the fruit of S. caripense using the reagents innuPREP Plant RNA (Analytik Jena AG, Germany) and PureLink® ARN Mini Kit (Ambion, Life Technologies, USA); it was purified for the synthesis of the first strand of cDNA through reverse transcription using the reagents 5X All-In-One RT MasterMix (with AccuRT Genomic DNA Removal Kit) (Applied Biological Materials Inc., Canada). The microtubes for RT-PCR in a final volume of 10 µL contained the necessary components for amplification in thermocycler Mastercycler EP Gradient 96 well Thermal Cycler (Eppendorf, Germany). The microtubes for RT-qPCR in a final volume of 10 µL contained the necessary components for amplification in thermocycler QuantStudio® 3 (Applied Biosystems, USA) (Table IV). When the reaction finished, it was assessed for genespecific amplification fragments by melting curve. The CT values of samples were exported to an Excel® (Microsoft, USA) calculation sheet, and the relative expression was determined with the 2 ∆∆ method [30], [50]; where: ΔΔCT = (CT,C3H -CT,rRNA)time x -(CT,C3H -CT,rRNA)time 0 (1) The values of C3H expression were normalized using 5.8S rRNA, and the levels of expression were relative to day zero. The relative expression in each level corresponds to the mean with four biological replicates (± S.E., n = 4) [39], [48], [51], and three technical replicates, whose CT values were manually controlled for S.D. > 0.5 [52]. 2) Gene Expression: The sample for analysis of ANS expression in BIO-Ltg1 and BIO-Cyb1, consisted in 7 fruits per plant conserved at chilling temperature (10 ± 2 °C) with photoperiod (16 h day/8 h night) from fluorescent lights (1250 lx) for fourteen days; RNA was isolated from fine sheets of fruit skin. The data were disposed under CRD with two treatments, seven observations per treatment and a linear additive statistic model [56]. Statistic packages GelAnalizer 2010 for comparative analysis and Minitab 17 for data processing were used.
The sample for analysis of C3H expression in IBT-Lib1, consisted in 12 fruits conserved at chilling temperature (10 ± 2 °C) with photoperiod (16 h day/8 h night) from fluorescent lights (1250 lx) for fourteen days; RNA was isolated from thin sheets of fruit flesh. The data were disposed under CRD; first an assay of the experiment was performed with three treatments and four biological replicates, and then with three technical replicates for analysis with the mean values; the statistic model is the same as above. Statistic packages InfoStat 2018, Minitab 17 and RStudio 1.2.1335 were used.

A. Morphological Description
The phenogram corresponding to the CA for qualitative descriptors of S. caripense (Fig. 1), represented a cophenetic correlation coefficient ( xy = 0.88) higher than 0.8; this implies a good representation of the similarity matrix [57].  The mode of P-Size descriptor evaluated in S. caripense ecotypes could be similar for the tzimbalo accessions BIRM/S 1034, E-7, EC-40 and QL-013; and it is perceived a bit higher, and superior if were compared to the related wild species P-80, P-62, E-257 and E-34, or to the pepino cv. Sweet Long or cv. Puzol, respectively. Furthermore, rare radicular protuberances in node were observed in S. caripense, as greater steam pubescence density, and more compound leaves, similar to related wild species [58], [59]. Not all S. caripense plants presented fruit stripes, a descriptor of broad variability, important for agronomic purposes of these species; wild relatives are sources of variation for plant breeding and for studies about the process of domestication. The tzimbalo plants presented greater style exsertion, as high pollen production and many seeds per fruit, as related wild species, which contribute to cross pollination and germplasm dispersion; in contrast to modern cultivated varieties of S. muricatum [58], [59].
The descriptors Fr-Flavour, Se-Diameter, Fr-AddColour, Fl-CorollaColour, Fr-Stripes and others, represent sources of variation for the breeding of S. muricatum and related studies. The environment does not regulate the dominant type effects of qualitative expression in a monogenic or oligogenic mode, these are ideal for its high heritability [60].
The PCA approach for quantitative descriptors of S. caripense explained 91.88% of total variance until the PC4, with Eigen values higher than one (PC1 = 6.46, PC2 = 4.01, PC3 = 1.73, PC4 = 1.58); the PC1 and PC2 represented 43.05% and 26.75% of total variation, respectively (Fig. 3).  On the other hand, the PC1 (43.05%) of the PCA for quantitative descriptors of S. caripense was mostly correlated with Fr-Length, Fr-PlacentLength, Fl-SepalLength, and others. These descriptors have additive type effects, are regulated by the environment in a polygenic mode. Therefore, it is optimal to evaluate them by variance decomposition in genotype, environmental, and interactions effects. According to plant reproduction systems, breeding methods are related to the species of interest [60]; plant breeding with S. caripense accessions is carried out through backcrossing to S. muricatum [22].

B. Gene Expression
The expression of F3H was identified in the skin of BIO-Ltg1 fruit with 2 µL of cDNA (100 ng·µL -1 ) per reaction and primers alignment at 52 °C (Fig. 4). The expression of F3H in BIO-Ltg1 seems to increase slightly after five days of postharvest conditions; after fourteen days well defined and intense F3H transcripts were observed.  [34] showed through RT-PCR that F3H expression increases gradually in controlled conditions. Nevertheless, considering an increase of F3H transcripts in BIO-Ltg1, it is mentioned that early expression of the structural gen F3H is positively correlated with the increase of anthocyanins content in S. tuberosum tuber. It is different from the fruit of S. melongena, S. lycopersicon and Capsicum spp. [32]. This suggests that reached F3H in its biosynthetic pathway, the enzymatic action can follow or redirect it. It takes another way apart of that for anthocyanins accumulation, such as flavonols (kaempferol) formation in the presence of flavonol synthases [61].
The expression of ANS was identified with 1 µL of cDNA (750 ng·µL -1 ) per reaction and primers alignment at 55 °C; the transcripts of ANS in BIO-Ltg1 (48.20 ng·µL -1 ; n=6) and BIO-Cyb1 (36.19 ng·µL -1 ; n=5) fruit were quantified by comparative analysis of bands intensity on the agarose gel. They were taken as reference molecular weight marker (WM) bands of known concentration. The ANOVA did not return significant differences for the expression of ANS; p-value = 0.206, D.F. = 1 according to Tukey test (p-value < 0.05).
In relation to the expression of ANS (145 bp) in BIO-Ltg1 and BIO-Cyb1 on agarose gel (Fig. 5), for short amplicons ˂ 150 bp, sometimes is observed very weak and fuzzy bands which migrate ahead of the major-specific bands. A Superstructured or single-stranded version of the specific transcripts in an equilibrium state should be considered specific [62]. The gel composition sometimes interferes with band definition. Therefore, polyacrylamide gels can provide a higher resolution. It could be aberrant reactions that influence the identification, quantification, and analysis. Nevertheless, RT-qPCR provides higher specificity. The plants BIO-Ltg1 and BIO-Cyb1 are phenotypically those that presented greater fruit surface covered by additional color and fruit stripes corresponding to the expression of anthocyanin pigments associated with genes (ANS) identified in S. caripense through RT-PCR. The expression of the structural gene ANS is positively correlated with the increase of anthocyanins concentration in S. tuberosum tuber, also this occurs in the fruit of S. lycopersicon, S. melongena and Capsicum spp. [32]. Moreover, ANS is required for the production of characteristic pigments of anthocyanins [52].
The expression of C3H in IBT-Lib1 fruit was identified through RT-PCR. It was induced by exposure to postharvest conditions and monitored through RT-qPCR. The term evaluated the expression of the reference gene 5. (2) The corresponding p-value = 0.1649 indicates that it was not significantly different during the days of controlled conditions. The log2-transformed values (Table V) represent asymmetric logarithmic scale [63], [64]. The ANOVA applied to the mean fold change of C3H expression returned significant differences, p-value = 0.0085 (Table VI). The mean fold change of C3H expression in IBT-Lib1 for day fourteen (6.24 ± 0.73) was significantly different from that calculated for day zero (3.32 ± 0.33) and similar for day five (4.52 ± 0.34) (Fig. 6); transcripts level of C3H expression increased in 2.92 units after fourteen days in postharvest conditions. Fig. 6 Gene expression of C3H (mean ± S.E., n = 4) in tzimbalo IBT-Lib1 fruit exposed to controlled temperature (10 ± 2 °C) and photoperiod ( It is mentioned that the CGA content in S. melongena fruit cv. Lucía increments after two weeks of storage at 10 °C [38]. Furthermore, C3H transcripts levels in Andean varieties of S. tuberosum, increments drastically after storage at 10 °C in darkness, preceded by exposure to drought stress during tuberization; the increase of C3H expression in cv. Huata Colorada coincides with the CGA content caused by drought [41]. Also, the expression level of C3H in S. tuberosum increments in 2.4 units by induction of CGA biosynthesis with sucrose 120 mM [39].
Previous investigations about pepino [65], eggplant [38], and potato [39], [41], support the obtained results with tzimbalo [15], [44], for the future development of improved varieties and the enhancement of the commercial potential of these species [10], [66]. Additionally, the concentrations for phenolic compounds of S. caripense fruit are greater than phenolic contents of melon and cucumber, and these are useful for the development of new varieties of S. muricatum, focused on the improvement of nutritional and bioactive values of the fruit [20].

IV. CONCLUSIONS
The morphological description of tzimbalo ecotypes indicates that Fr-Flavour, Se-Diameter, Fl-CorollaColour, Fr-Stripes, Fr-Length, Fr-PlacentLength and Fr-PlacentBreadth were characters that contribute more to the variability, and these are agronomical distinctive to be utilized in breeding programs. The expression of F3H and ANS identified through RT-PCR in BIO-Ltg1 and BIO-Cyb1, and the expression of C3H in IBT-Lib1 fruit, constitutes an analysis applied to the exploration of candidate genes, for subsequently transcript quantification in real time.
The expression levels of C3H in the flesh of IBT-Lib1 fruit influenced by postharvest conditions were significantly different; opening the possibility of selecting genotypes that demonstrate good performance in front of different crop conditions. The approach of candidate genes and their expression represents a promising tool for introducing tzimbalo into plant breeding programs, focused on the conservation and utilization of Andean resources (Fig. 7).