Silencing of the phytoene desaturase (PDS) gene affects the expression of fruit-ripening genes in tomatoes

Background Past research has shown that virus-induced phytoene desaturase (PDS) gene silencing via agroinjection in the attached and detached fruit of tomato plants results in a pale-yellow fruit phenotype. Although the PDS gene is often used as a marker for gene silencing in tomatoes, little is known about the role of PDS in fruit ripening. In this study, we investigated whether the pepper PDS gene silenced endogenous PDS genes in the fruit of two tomato cultivars, Dotaerang Plus and Legend Summer. Results We found that the pepper PDS gene successfully silenced endogenous PDS in tomato fruit at a silencing frequency of 100% for both cultivars. A pale-yellow silenced area was observed over virtually the entire surface of individual fruit due to the transcriptional reduction in phytoene desaturase (PDS), zeta-carotene (ZDS), prolycopene isomerase (CrtlSO), and beta-carotene hydroxylase (CrtR-b2), which are the carotenoid biosynthesis genes responsible for the red coloration in tomatoes. PDS silencing also affected the expression levels of the fruit-ripening genes Tomato AGAMOUS-LIKE1 (TAGL1), RIPENING INHIBITOR (RIN), pectin esterase gene (PE), lipoxygenase (LOX), FRUITFULL1/FRUITFUL2 (FUL1/FUL2), and the ethylene biosynthesis and response genes 1-aminocyclopropane-1-carboxylate oxidase 1 and 3 (ACO1 and ACO3) and ethylene-responsive genes (E4 and E8). Conclusion These results suggest that PDS is a positive regulator of ripening in tomato fruit, which must be considered when using it as a marker for virus-induced gene silencing (VIGS) experiments in order to avoid fruit-ripening side effects.


Background
Virus-induced gene silencing (VIGS) is a widely used reverse genetics tool for the high-throughput analysis of the biological functions of target genes in plants due to its ability to rapidly degrade the mRNA of the target gene, its simplicity in use, and its rapid results. Application of VIGS was first achieved with the silencing of the phytoene desaturase (PDS) gene in tobacco leaves, resulting in visible leaf photo-bleaching [1]. Since then, VIGS has been used to investigate the function of genes involved in a number of research areas, including abiotic and biotic defense mechanisms, coloration (i.e., anthocyanin and carotenoid biosynthesis), and plant growth [2][3][4], and more than 30 plant species have been subject to this form of investigation [5].
With the aim of optimizing the VIGS approach, PDS is widely used as a common marker gene in many plant species due to its ease of detection. The use of VIGS in tomatoes was first achieved by silencing PDS in leaves [6], followed by many more successful PDS-silencing experiments on tomato fruit [7][8][9]. The increase in carotenoid content (in the order of a 10-to 14-fold increase) during the ripening of tomatoes has been reported [10].

Open Access
Plant Methods *Correspondence: kblim@knu.ac.kr; ckkim@knu.ac.kr † Aung Htay Naing and Swum Yi Kyu contributed equally to this work 1 Department of Horticultural Science, Kyungpook National University, Daegu, South Korea Full list of author information is available at the end of the article The PDS gene encoding the phytoene desaturase enzyme is involved in carotenoid biosynthesis pathway [10] and its silencing in the fruit of tomatoes reduced carotenoid accumulation and led to yellow-colored fruit [8,9]. In addition, transcriptional control of PDS in carotenoid biosynthesis in tomato flowers and fruits has also been reported [11][12][13]. However, the molecular mechanism how the PDS silencing affects carotenoid biosynthesis in tomato fruits has not been well explored. Su et al. [14] reported that carotenoid accumulation during the ripening of tomatoes is controlled by the transcript levels of beta-carotene hydroxylase (CrtR-b2) gene. In addition, Fantini et al. [13] also claimed that reduction of carotenoid contents in tomato fruits is associated with down-regulation of the carotenoid biosynthesis genes zeta-carotene desaturase (ZDS) and prolycopene isomerase (CrtISO). Hence, we investigated whether PDS silencing influences the regulation of the above mentioned genes involved in carotenoid biosynthesis pathway.
Previously, Kim et al. [15] reported that the silencing of the pepper PDS gene, which has high homology to the PDS gene in tomato and petunia plants, in pepper leaves and fruit led to distinct photo-bleaching. Recently, Naing et al. [16] reported that the pepper PDS gene significantly suppressed endogenous PDS gene in petunia and reduced total chlorophyll content. Hence, pepper PDS was exploited to silence the endogenous PDS gene in tomato fruit. Based on our preliminary work, the silencing of the endogenous PDS gene in detached tomatoes led to fruit with a yellow color that continued until the fruit had softened. Vrebalov et al. [17] observed that the reduction of the RIPENING INHIBITOR (RIN) gene results in the failure of fruit to ripen. In addition, Vrebalov et al. [18] and Itkin et al. [19] also claimed that Tomato AGAMOUS-LIKE1 (TAGL1) is involved in the regulation of fruit ripening and its suppression results in yellow-orange fruits, decreased carotenoids and delayed ripening. Moreover, the involvement of pectin esterase gene (PE), lipoxygenase (LOX), and the two homologous genes FRUITFULL1 (FUL1) and FRUITFUL2 (FUL2) in the ripening of tomatoes has been identified [17,[19][20][21]. In the present study, we were interested in investigating whether the expression patterns of the fruit-ripening related genes would change in PDS-silenced tomato fruits.
Because tomatoes are climacteric fruit, their ripening is associated with the expression of ethylene biosynthesis genes such as 1-aminocyclopropane-1-carboxylate oxidase (ACO) [20,[22][23][24]. Alexander and Grierson [25] and Hu et al. [26] reported that the suppression of SlACO1 delayed fruit ripening and ethylene biosynthesis, while the transcript levels of SlACO1 and SlACO3 are markedly high when the ripening of tomatoes is triggered [25,27]. Similarly, Lincoln et al. [28] and Zhang et al. [20] observed that expression of E4 is positively associated with ethylene biosynthesis, while Kneissl and Deikman [29] also reported that E8 is a fruit ripening-specific gene that is activated during the fruit ripening process, and it is widely used as a fruit-specific promoter in transgenic tomatoes [30,31].
In summary, in the present study, PDS expression was silenced in the fruit of two commercial tomato cultivars using VIGS to elucidate how PDS silencing affects the fruit phenotype and the expression of the genes mentioned above that are involved in the ripening of tomatoes.

Plant materials
Seeds of two tomato cultivars, Dotaerang Plus (Takii Korea Co., Ltd.) and Legend Summer (Haesung Seed Plus Co., Ltd.), were sown in plug trays filled with the soil-less mixture BM7 (Berger Co., Quebec, Canada) in a growth chamber at 23 °C, with a 16 h photoperiod (400 µmol m −2 s −1 ) and 70% relative humidity (RH). After 2 weeks, healthy germinated plants of uniform size were transplanted into pots filled with BM7 and moved to a greenhouse at 25-27 °C (daytime) and 16-18 °C (night), with a 16-h photoperiod and 70% RH for plant growth and fruiting. When the fruit had reached the mature green stage, healthy fruits of uniform size were picked at 8 AM from each cultivar and used for VIGS testing.

Preparation of Agrobacterium suspension
The use of the tobacco rattle virus (TRV1 and TRV2) in VIGS experiments has been reported for tomatoes [6]. Agrobacterium tumefaciens strain GV3101 harboring the tobacco rattle virus vectors pTRV1, pTRV2 (without PDS), and pTRV2-PDS was provided by Prof. Je Min Lee (Kyungpook National University, Daegu), while the PDS gene inserted into the vector pTRV2-PDS was cloned from pepper. Bacterial samples each harboring one of the three vectors were separately cultured in LB broth containing 25 mg L −1 kanamycin and rifampicin in a shaking incubator at 300 rpm and 28 °C. When the optical density of the bacteria was OD 600 0.6, the cells were centrifuged and the obtained pellets suspended in an inoculation buffer of 10 mM MgCl 2 , 10 mM MES (pH 5.6), and 200 µM acetosyringone, to obtain a final OD 600 of 1.0 for each culture. The cells were then placed in a rotary shaker at 300 rpm and 28 °C for 6 h.

Agroinjection into detached tomatoes
Green mature tomatoes were harvested at 8 AM from each cultivar and immediately divided into three groups to be separately injected with a buffer solution alone or one of the bacterial suspensions, pTRV1 + pTRV2 (TRV1/2) or pTRV1 + pTRV2-PDS (TRV1/2-PDS; Fig. 1). The suspensions were injected through the carpopodium with a 1-mL syringe at room temperature. Once the suspension had infiltrated the tissue, it appeared in the sepals. There were 20 tomatoes per treatment and three replications of each treatment. The injected fruits were placed in a growth chamber at 23 °C, with a 16-h photoperiod (400 µmol m −2 s −1 ) and 70% relative humidity.

Evaluation of fruit phenotype
Two weeks after injection, the presence of pale yellow and red coloring of the pericarp was evaluated for both cultivars. To determine the silencing frequency, the following calculation was used:

Measurement of gene expression levels using quantitative real time-PCR
Two weeks after injection, RNA was extracted from the pericarp of the tomatoes injected with the different inoculums; for those injected with TRV1/2-PDS, RNA was extracted from visibly silenced areas. Total RNA extraction was performed using an RNAqueous kit (Ambion Inc., Austin, TX, USA). Reverse transcription was conducted using 1 µg of total RNA and an oligo dT20 primer with a ReverTra Ace-α kit (Toyobo Co. Ltd., Osaka, Japan). Transcript levels of the endogenous tomato PDS gene, the carotenoid biosynthesis genes (ZDS, CrtR-b2, and CrtlSO), the fruit-ripening-associated genes (FUL1, FUL2, PE, LOX, TAGL1, and RIN), and the ethylene biosynthesis and response genes (ACO1, ACO3, E4, and E8) Silencing frequency (%) = number of tomatoes with visible silencing/ total number of tomatoes injected × 100 were measured using a StepOnePlus Real-Time PCR system (Thermo Fisher Scientific Inc., Waltham, MA, USA). Expression levels were normalized to the actin gene to minimize variation in the cDNA template. The primers, accession numbers, and PCR conditions for the examined genes are listed in Additional file 1: Table S1.

Statistical analysis
For qRT-PCR analysis, total RNA was extracted from three different biological samples, and this analysis was repeated three times for each biological sample. The data were analyzed using SPSS v.11.09 (IBM Corporation, Armonk, NY, USA), and the values of the relative gene expression levels described in the study were means of the three biological replicates. The least significant difference test (LSDT) was used to identify differences between the means. A P-value of less than 0.05 was considered statistically significant.

PDS-silencing fruit phenotype
Mature green tomatoes harvested from the two cultivars (Dotaerang Plus and Legend Summer) were immediately injected with the buffer, TRV1/2, or TRV1/2-PDS. Two weeks after injection, silenced phenotypes associated with PDS were observed in the TRV1/2-PDS tomatoes, but not in those injected with the buffer or TRV1/2. A distinct pale-yellow coloration was observed in fruits of both cultivars injected with TRV1/2-PDS ( Fig. 2A, B). In this study, all tomatoes injected with TRV1/2-PDS exhibited a silencing efficiency of 100%, and the silencing patterns of the fruit within the same cultivar were also similar based on visual observations (data not shown). The silenced areas of cv. Legend Summer covered Fig. 1 Schematic of injections to the carpopodium of green mature tomato fruits using a 1 mL syringe virtually the entire fruit, while this was not observed in cv. Dotaerang Plus ( Fig. 2A, B), indicating that there is variation in the effectiveness of silencing depending on the cultivars used. Tomatoes injected with the buffer solution or TRV1/2, which were used as controls, did not exhibit any silencing symptoms and developed the customary red color, although those tomatoes injected with the buffer solution were a deeper red compared to those injected with TRV1/2 for both cultivars ( Fig. 2A, B).

Expression of the PDS gene
To clarify whether the silencing of PDS was associated with the phenotypic variation, the transcript levels of PDS expressed in the pericarp of the injected fruit were determined using quantitative real time-PCR (qRT-PCR). For both cultivars, the transcript levels detected in tomatoes injected with TRV1/2-PDS (pale yellow) were significantly lower than those of the control groups (red), while the expression levels in tomatoes injected with TRV1/2 were slightly lower than those injected with the buffer (Fig. 2C, D). These results indicate that VIGS significantly inhibits PDS expression, which is consequently associated with the observed fruit phenotypes.

Expression of carotenoid biosynthesis genes
Because PDS encodes a key enzyme involved in the carotenoid biosynthesis pathway, its silencing may affect carotenoid accumulation. We investigated whether PDS silencing affected carotenoid accumulation in tomatoes by measuring the expression levels of the carotenoid biosynthesis genes (ZDS, CrtR-b2, and CrtlSO), which are involved in carotenoid biosynthesis, using qRT-PCR (Fig. 3A-F). Silencing of the PDS gene significantly reduced the expression levels of the biosynthesis genes compared to the control groups for both cultivars, which could lead to a reduction in the carotenoids in PDSsilenced tomatoes. This association is supported by the visual appearance of the experimental tomatoes, which appeared to have lower levels of carotenoids than the control fruit.

Expression of fruit-ripening genes
In this study, PDS-silenced tomatoes did not ripen, remaining pale yellow until the fruit had softened, while the control tomatoes underwent natural ripening, with softening occurring approximately 10 days earlier than in the PDS-silenced tomatoes. The expression levels of the known ripening genes (TAGL1, RIN, PE, LOX, FUL1, and FUL2) were analyzed for all tomatoes and found to be significantly down-regulated in the PDS-silenced fruits compared to the controls (Fig. 4A-L). These results suggest that PDS might influence the ripening of tomatoes by regulating the expression of these genes.

Expression of ethylene biosynthesis and response genes
Tomatoes are climacteric, meaning their ripening is associated with the transcriptional activation of ethylene biosynthesis and response genes. We further explored the impact of reduced PDS mRNA levels on ethylene biosynthesis during the ripening process by determining the expression levels of the ethylene biosynthesis genes (ACO1 and ACO3) and response genes (E4 and E8). Results indicated that these genes were significantly down-regulated in PDS-silenced fruits compared to the controls (Fig. 5A-H). These results suggest that the silencing of PDS might inhibit the transcription of these genes.

Discussion
The silencing of PDS using VIGS has been reported for a variety of plants, including tomato, pepper, and eggplant leaves, that belong to the Solanaceae family [6,32,33]. In these studies, PDS was used as a marker gene due to the photo-bleaching caused by its inactivation. However, only a few reports have described the silencing of PDS in the fruit of tomatoes [8,9], and they did not examine PDS gene silencing beyond photo-bleaching. Because PDS encodes an enzyme involved in the carotenoid biosynthesis pathway, it is interesting to investigate whether its silencing would inhibit the expression of carotenoid biosynthesis genes and those involved in fruit ripening and ethylene biosynthesis.
In this study, all tomato fruit of both cultivars (Legend Summer and Dotaerang Plus) injected with TRV1/2-PDS exhibited a pale yellow phenotype, representing a significant difference in appearance compared to non-PDS-silenced fruits ( Fig. 2A, B). Expectedly, the exploitation of the PDS gene to silence the endogenous PDS gene in the tomato cultivars resulted in similar phenotypes to those reported by Orzaez et al. [8], Wang et al. [34], and Romero et al. [9], who silenced PDS in attached and detached tomatoes (cvs. Micro Tom, Maxifort, Jinfen) using the tomato PDS gene. In the control fruits, those injected with a buffer exhibited a deeper red color than those injected with TRV1/2. It is likely that "TRVI/2" itself is able to degrade PDS mRNA to some extent because the transcript levels of PDS were slightly lower in the TRVI/2-treated tomatoes than in the buffer-treated fruit. The slight differences observed in the PDS-silenced areas of the two different cultivars could be due to genetic differences. A significant reduction in PDS transcript levels was observed in PDS-silenced tomatoes compared to those of the controls, as illustrated by the resultant fruit phenotypes. Orzaez et al. [8] and Romero et al. [9] reported similar findings in their studies. These results suggest that the pepper PDS gene was able to silence the endogenous PDS gene in the tomato cultivars.
Su et al. [14] observed that carotenoid accumulation during the ripening of tomatoes is linked to the transcript levels of CrtR-b2 genes. In addition, the association between the carotenoid accumulation and upregulation of PSY1, ZDS, CrtlSO had also been reported in tomato [11][12][13]. In this study, the transcript levels of the carotenoid biosynthesis genes ZDS, CrtR-b2, and CrtlSO were also significantly lower in the PDS-silenced fruits than in the controls; these lower transcript levels were likely to prevent the accumulation of enough carotenoids to produce a red color in the fruit, as indicated by the lack of red in the PDS-silenced tomatoes. The lower accumulation of carotenoids resulting from the silencing of PDS in the fruit of tomatoes has been reported previously, although the transcription levels of these biosynthesis genes were not specifically examined [8,9,13].
The downregulation of the fruit-ripening genes TAGL1, RIN, PE, LOX, FUL1, and FUL2 was also observed in the PDS-silenced tomatoes but not in the controls, suggesting that PDS silencing inhibited fruit ripening. Downregulation of these genes resulting in defective ripening with fruit softening has also been reported in previous research [18,19,[35][36][37]. In addition, Bemer et al. [36] and Shima et al. [37] also reported that the tomato FUL1 and FUL2 genes participate in fruit ripening via their interaction with RIN. FUL1/FUL2 and RIN genes code for enzymes upstream of the ethylene signaling pathway during ripening, and suppression of these genes results in a ripening-defective phenotype with lower ethylene production [18,[37][38][39]. Because the downregulation of the FUL1/FUL2 and RIN genes was observed in PDSsilenced tomatoes in this study, we also expected that the downregulation of the ethylene biosynthesis and response genes ACO1, ACO3, E4, and E8 would occur, and our analysis revealed that these genes were indeed downregulated in PDS-silenced fruit. Zhang et al. [20] also suggested that the transcription levels of PDS in tomato fruit were positively associated with those of other genes, such as TAGL1, FUL1/FUL2, RIN, ACO1, ACO3, E4, and E8, that are involved in fruit ripening. Taken together, we conclude that the silencing of the PDS gene led to the downregulation of ethylene biosynthesis and response genes and fruit-ripening genes, leading to a ripening-defective phenotype. In some VIGS studies of fruit, PDS was used as marker gene to characterize the functional role of target genes; for example, Orzaez et al. [40] identified the specific role of Delia (D) and Rosea (R) in anthocyanin accumulation in tomatoes with the combined construct TRV2:DR:PDS. However, it is relatively difficult to specifically characterize the function of a target gene due to potential side effects of PDS in terms of fruit pigmentation. Our results suggest that these effects should be taken into account when using PDS as a marker gene in the VIGS-based research of fruit.

Conclusion
We demonstrated that the pepper PDS gene effectively silenced transcription of the endogenous PDS gene in the fruit of two tomato cultivars, and this silencing affected the regulation of the ZDS, CrtlSO, and CrtR-b2 genes, which are involved in the carotenoid biosynthesis pathway. PDS-silencing also appeared to affect fruit ripening, acting as a positive regulator by modulating the fruit-ripening genes TAGL1, RIN, PE, LOX, FUL1, and FUL2 and their associated ethylene biosynthesis and response genes ACO1, ACO3, E4, and E8. These results suggest that PDS silencing not only affects the carotenoid pathway but also leads to the inhibition of other genes involved in the fruit-ripening process in tomatoes. We expect that our work will aid in the better understanding of the regulatory mechanisms of PDS in the fruit-ripening process.
Additional file 1: Table S1. Primers, accession numbers, and PCR conditions used for qRT-PCR analysis of the genes.