Introduction

The key enzyme for transcription of protein-encoding genes in eukaryotes is RNA polymerase II (RNAPII). The histone acetyltransferase (HAT) Elongator copurifies with RNAPII during transcriptional elongation, presumably by rendering the DNA more accessible for the passage of the polymerase1,2. Except for its important role in transcriptional elongation through the histone and α-tubulin acetylation3,4,5, there are a variety of mechanisms believed to be involved, including cytoskeleton organization, exocytosis and tRNA modification6,7,8,9. Elongator is composed of two subcomplexes: the core subcomplex with ELP1-3 and the accessory subcomplex with ELP4-610. ELP1 contains a nuclear localization sequence which is essential for Elongator function11,12, while ELP2 is dispensable for the integrity of the holo-Elongator complex because the complex lacking ELP2 even retains the ability to acetylate histones in vitro13. ELP3 is the catalytic subunit of the Elongator complex and encodes a histone acetyltransferase3. ELP3 also contains an Fe4S4 cluster which is characterized as S-adenosylmethionine (SAM) binding domain and is likely to carry out demethylation of histone residues14,15. In addition, Elp4-6 subcomplex forms a hexameric ring-shaped structure both in vitro and in vivo that is important for histone H3 and tRNA binding16,17.

The subunit structures of Elongator are highly conserved among yeast, human and plants. However, their functions are diverse in eukaryotic organisms1,18. For example, mutations in yeast Elongator subunits lead to resistance to the zymocin γ-toxin subunit and sensitivity to salt, caffeine and temperature19. Mutations in the IKBKAP (human ELP1) result in abnormal neuron development and cause human severe hereditary familial dysautonomia disease20. Presently limited genetic studies performed in Arabidopsis have emerged indicating that Elongator functions in plant growth and development and in responses to biotic and abiotic stress. At the macroscopic level, Elongator mutants in Arabidopsis display a pleiotropic phenotype, including reduced primary root growth, narrow and elongated leaves, aberrant inflorescence architecture, reduced germination frequency, delayed seedling growth and hypersensitivity to abscisic acid (ABA)21,22. Disruption of each Elongator subunit (ELP1–ELP6) results in similar phenotypes and double or triple mutants resemble single mutants in Arabidopsis, suggesting that plant Elongator subunits work as a complete complex and share common functions21,23. For example, all the four Elongator mutants (abo1/elo2/elp1, elp2, elo3/elp3 and elo1/elp4) have narrow leaves, reduced root growth, ABA hypersensitivity and an increased accumulation of anthocyanins23. Nelissen, et al.18 showed that Elongator likely controls plant development through regulating auxin-responsive genes, as a large number of auxin-responsive genes are differentially expressed in elp plants. Abnormal development may also derive from the upregulated jasmonic acid (JA) and ethylene (ET) signaling pathways observed in elongata (elo) mutants. Besides their critical functions in plant growth and development, Elongator has also been implicated in the response to biotic and abiotic stress. Disruption of both core and accessory subcomplexes results in the hyperinhibition by ABA of germination and seedling growth. Besides, the abo1/elo2/elp1, elp2, elo3/elp3 and elo1/elp4 mutants are all more resistant to oxidative stress than the wild-type23, suggesting that Elongator plays crucial roles in regulating plant responses to ABA, oxidative stress resistance and anthocyanin biosynthesis in Arabidopsis. Consistently, abo1/elo2/elp1 mutant exhibits a drought-resistant phenotype24. In summary, Arabidopsis Elongator appears to be essential for maintaining hormonal balance, as disruption of Elongator shifts hormonal signaling, resulting in pleiotropic growth and defense phenotypes.

Although some functional studies of Elongator have been performed in Arabidopsis, the biological, biochemical and molecular function of Elongator still need further elucidation and little is known about its roles in other plant species. Tomato (Solanum lycopersicum) is economically one of the most important crop plants. Unlike Arabidopsis, the tomato plant yields fleshy fruits and serves as another excellent model plant. In our present study, a tomato Elongator subunit 2-like gene, here designated as SlELP2L, was isolated and its expression profiles under various phytohormones and the tissue-specific expression were investigated. Silencing SlELP2L in tomato leads to reduced leaf growth, accelerated senescence of leaves and sepals and dark-green fruits with increased chlorophyll content and reduced carotenoid accumulation and down-regulation of ethylene- and ripening-associated genes, while the expression of DNA methyltransferase genes was up-regulated. Besides, the transgenic plants are hypersensitive to ABA during seedling growth. These results suggest that SlELP2L plays an important role in tomato growth and development, as well as in ABA signaling. Our results also suggest that SlELP2L in tomato plays a distinct role compared with in Arabidopsis, indicating that Elongator has functional diversity in different species.

Results

Cloning and tissue expression pattern analysis of SlELP2L gene

In order to explore the potential functions of Elongator in tomato, we isolated a putative ELP gene (Elongator complex protein 2-like) from wild-type tomato fruits based on a cDNA clone (GenBank accession no. XM_004240442), here we named it SlELP2L. Nucleotide sequence analysis revealed that SlELP2L was consistent with the cDNA clone and contained an open reading frame (ORF) of 2190 nucleotides and encoded 729 amino acid residues with an estimated molecular mass of 80.4 kDa. The search of conserved structure domains showed that the location of 531–671 of SlELP2L protein are WD-repeats region and its theoretical isoelectric point is 6.2. The sequence alignment conducted by DNAMAN 5.2.2 program showed that tomato SlELP2L shared 64.2% identity with the Arabidopsis ELP2 (AT1G49540) protein at the amino acid level.

To extend our understanding of the role of SlELP2L in tomato growth and development, we examined its expression patterns in a wide range of tomato organs, including vegetative tissues such as young, mature and senescent leaves, stems and roots and reproductive tissues such as sepals, flowers and fruits at different stages of development. Quantitative RT-PCR analysis showed that SlELP2L was constitutively expressed at high levels in most of organs we examined, but was expressed at relatively low levels in flowers (Fig. 1). Previous reports have demonstrated that a number of Elongators are involved in regulating the essential steps in Arabidopsis growth and development25. The observation of high and constant SlELP2L expression in tomato suggests that SlELP2L may have a specialized function in growth or reproductive development of tomato plant.

Figure 1
figure 1

Expression patterns of SlELP2L gene in different tissues and during fruit development in wild-type tomato.

Ro, roots; St, stems; Se, sepals; Fl, flowers; Yl, young leaves; Ml, mature leaves; Sl, senescent leaves; IMG, immature green; MG, mature green; B, breaker; B+4, 4 day after breaker stage; B+7, 7 day after breaker stage. The expression data of root were normalized to 1. Each value represents the mean ± SD of three replicates.

Full size image

Expression profiles of SlELP2L gene under exogenous hormone treatments

In plants, growth and form are determined by the temporal and spatial regulation of cell division and cell expansion in which plant hormones play a crucial role. To determine whether SlELP2L is involved in phytohormone response, WT tomato seedlings were treated with different hormones. Quantitative RT-PCR analysis showed that the transcripts of SlELP2L mRNA remained unchanged at the beginning, while increased slightly at 24 h after ABA treatment (Fig. 2A). Whereas the expression of SlELP2L was inhibited by gibberellin acid (GA3) treatment (Fig. 2B). In addition, Fig. 2C showed that SlELP2L transcripts were also induced by ACC (the ethylene precursor) treatment. So far, many cis-elements have been reported for their critical roles in determining the hormone-responsive expression profiles of plant genes. Statistical analysis showed that there are multiple hormone-related cis-elements enriched in the promoter of SlELP2L gene, including ABA, GA and ethylene (Table S1). The existence of these cis-elements indicates that they likely have potential functions, but this is not always the case (For example, although there are GA-responsive cis-elements in SlELP2L promoter region, it is not induced by GA). The results suggest that SlELP2L may play a role in ABA and ACC response and signal transduction.

Figure 2
figure 2

Quantitative RT-PCR analysis of the expression of SlELP2L gene in leaves under ABA (A), GA3 (B) and ACC (C) treatments

The relative expression levels were normalized to 1 in unstressed plant (0 h). Each value represents the mean ± SD of three replicates. The asterisks indicate statistically significant differences between the unstressed plants and treated plants (P < 0.05).

Full size image

Silencing SlELP2L inhibits leaf growth and accelerates leaf senescence

To explore the physiological role of SlELP2L in greater depth, we obtained 11 independent SlELP2L-silenced lines by RNAi. To verify the repression of SlELP2L in the six transgenic lines which exhibited distinguishable alterations from WT tomato, total RNA was isolated from young leaves in WT and transgenic plants. Quantitative RT-PCR analysis showed that the accumulation of SlELP2L transcripts in lines 1 and 8 was significantly reduced to roughly 6–8% of control levels, while only about 70–85% of transcript accumulation was repressed in the other lines (Fig. 3A). Thus lines 1 and 8 were selected for further characterization. To verify specific repression of SlELP2L, the expression of SlELP1L (XM_004239729), SlELP3L (XM_004236311), SlELP4L (XM_004250120) and SlELP6L (XM_004251673) was also monitored. Fig. S1 showed that the expression of these four genes was generally higher or remained unchanged in SlELP2L-RNAi lines except SlELP1L expression was slightly lower at breaker stage in transgenic fruits than that of WT, suggesting the presence of a regulatory relationship between these genes. The results also indicate that SlELP2L mRNA was specifically targeted by the SlELP2L RNAi transgene RNA.

Figure 3
figure 3

SlELP2L gene expression analysis and growth phenotype of SlELP2L transgenic plants.

(A). Relative expression profiles of SlELP2L between WT and SlELP2L-RNAi lines. The expression data of WT plants were normalized to 1. Each value represents the mean ± SD of three replicates. (B). Growth characteristics of WT and transgenic tomato plants. Seeds of WT and T1 generation SlELP2L-RNAi tomato were sowed for one week, then transplanted into pots, picture was taken for the representative plants after seven weeks. (C). Leaves obtained from the same node of WT and SlELP2L-RNAi plants. D and E. Comparisons of GA content (D) and IAA content (E) in the leaves of WT and transgenic plants. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines. (F). Growth characteristics of WT and transgenic tomato plants grown for three months in the greenhouse. G and H. Relative expression profiles of Rbcs-2 (G) and RCHY1 (H) in the leaves of different stages of senescence between WT and SlELP2L-RNAi plants. a, mature leaves (no visible senescence); b, early senescing leaves (<10% yellowing); c, senescent leaves (about 30% yellowing). The expression data of mature leaves were normalized to 1. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

Then the seeds of WT and the two T1 SlELP2L-silenced transgenic lines were germinated and cultivated concurrently in the greenhouse. The germination rate of transgenic tomato seeds was similar to that of WT tomato under normal conditions (data not shown). About seven weeks later, SlELP2L-silenced tomato plants showed a visible pleiotropic phenotype in leaves, such as reduced leaf rachis length and rapid senescence of tomato leaves (Fig. 3B). The compound leaves were collected from WT and SlELP2L-silenced tomato plants at the same nodes. Fig. 3C showed that the leaves from WT plants had longer leaf rachis length than that of SlELP2L-silenced tomato plants. This dwarf phenotype is usually observed in gibberellin disordered plants. GA has a crucial role in regulating internode length and leaf size and the latter is more sensitive to GA content modification in tomato26. Besides, indole-3-acetic acid (IAA), the main auxin in higher plants, also has profound effects on plant growth and development, including embryogenesis, seedling growth and flower development27. In order to explore the underlying reasons for the smaller shapes of transgenic plants, endogenous GA and IAA contents were determined in the leaves of WT and SlELP2L-silenced plants using an enzyme-linked immunosorbent assay (ELISA). The results showed that these two plant hormone contents were significantly decreased in transgenic plants (Fig. 3D, E).

Besides, about three months later, the lower leaves of WT plants remained green, while the leaves of SlELP2L-silenced tomato plants showed yellowing and senescence (Fig. 3F). To gain further information on the accelerated senescence phenotype observed in transgenic plants, the transcription of small subunit of ribulose-1, 5-biphosphate carboxylase-oxygenase (Rubisco) gene (Rbcs-2) was determined in tomato leaves at different stages of senescence. Rubisco initiates the pathways of photosynthetic carbon reduction and photorespiratory carbon oxidation, plays a decisive impact on net photosynthetic rate28. Fig. 3G showed that Rbcs-2 mRNA was highly expressed in fully expanded mature leaves of WT plants, then declined gradually during leaf senescence. Whereas Rbcs-2 expression in SlELP2L-silenced lines was markedly decreased at all three stages detected compared with that of WT, indicating that the photosynthetic rate may also be reduced in SlELP2L-silenced lines. In addition, Elongator complex has been reported to have a functional connection with the Urm1p, an ubiquitin-like protein (UBLs), suggesting a potential relationship between Elongator and protein ubiquitination process29. RCHY1 (ring finger and CHY zinc finger domain-containing 1), as a member of the RING finger family of E3 ubiquitin ligases can facilitate protein degradation via the ubiquitin-proteasome pathway30. Quantitative RT-PCR analysis showed that RCHY1 expression was increased significantly in the early senescence of leaves (less than 10% yellowing) of SlELP2L-silenced lines (Fig. 3H), indicating that the increase of protein ubiquitination may trigger a rapid senescence of SlELP2L-silenced transgenic tomato leaves.

Silencing SlELP2L produces dark-green fruits and accelerates sepal senescence

Visible alterations were displayed during the process of tomato fruit development and ripening, it is interesting to note that SlELP2L-silenced transgenic fruits showed darker green phenotype from IMG to breaker stage compared with WT (Fig. 4A–C). However, no significant difference was observed at B+4 and B+7 stages. In view of the color changes, the time from anthesis to breaker stage was measured, we found that there is no significant difference in the ripening time between WT and SlELP2L-silenced fruits (Table 1). Besides, another striking phenotype was observed in SlELP2L-RNAi fruits. The sepals of transgenic fruits began to lose their chlorophyll apparently and accelerated senescence was observed from B+4 stage, while WT sepals were still green (Fig. 4A, C). We further measured the time from breaker stage to sepal senescence (characterized as the first appearance of yellowing in sepals) and observed that the senescence time of SlELP2L-silenced sepals was about six days in advance compared with that of WT (Table 1), indicating that the silencing of SlELP2L in tomato leads to accelerated sepal senescence.

Table 1 Days from anthesis to breaker of fruits and days from breaker of fruits to sepals show senescence between wild-type and SlELP2L-RNAi fruits. Values represent the means ± SD in days for at least 15 fruits of each line
Full size table
Figure 4
figure 4

Phenotypes of fruits and sepals in WT and SlELP2L-RNAi lines.

(A). Representative fruits and sepals from WT and transgenic tomato. IMG, immature green; MG, mature green; B, breaker; B+4, 4 day after breaker stage; B+7, 7 day after breaker stage. (B). Color characterization of WT and SlELP2L-RNAi fruits on the plants. (C). Phenotypic characteristics of WT and SlELP2L transgenic fruits corresponding with Fig. 4A.

Full size image

The typical features of green-to-red color change in ripening tomato fruit is primarily because of chlorophyll degradation and carotenoid accumulation31. To determine whether or not the dark-green phenotype represented a change in total chlorophyll and carotenoid contents between WT and SlELP2L-silenced fruits, total chlorophyll and carotenoid were extracted from pericarps. The results showed that SlELP2L-silenced fruits had higher levels of chlorophyll content compared with the WT fruits at mature green (MG) stage as expected (Fig. 5A), which is responsible for the dark-green coloration of transgenic fruits. The total carotenoid content in B+4 stage fruits was also determined. Fig. 5B showed that SlELP2L-silenced fruits exhibited a slightly reduced carotenoid concentration, this slight reduction in transgenic fruits was not reflected apparently in the phenotype (Fig. 4A, C). SlELP2L thus may also affect the accumulation of pigments in fruits.

Figure 5
figure 5

Chlorophyll and carotenoid accumulation between WT and SlELP2L-RNAi fruits in pericarps.

(A). Pericarp chlorophyll content in mature green (MG) fruits. (B). Total carotenoid accumulation in red ripe fruits of B+4 (4 day after breaker stage). Biological replicates (2–3 fruits per fruit ripening stage) were performed in triplicate and the data are presented as mean ± SD. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

Ethylene- and ripening-related genes were down-regulated in SlELP2L-silenced fruits

The synthesis, perception and signal transduction of ethylene is essential for the coordination and completion of fruit ripening32 and ethylene also regulates the accumulation of carotenoid during ripening33. Currently, two systems of ethylene regulation have been proposed to operate in climacteric fruit. System-1 is responsible for producing the basal levels of ethylene detectable in all tissues. Whereas system-2 is autostimulatory during the climacteric fruit ripening and petal senescence and requires the induction of both of ACS (ACC synthase) and ACO (ACC oxidase) genes, which catalyze the rate limiting and final steps in ethylene biosynthesis, respectively34. The dark-green fruits and slightly reduced content of carotenoid in ripe fruit were observed in transgenic tomato (Fig. 4, 5). To further characterize the potential molecular mechanisms of SlELP2L in fruit ripening, the transcript levels of ethylene-related genes were examined from MG to B+7 stage between WT and transgenic tomato fruits. Fig. 6 showed that the expression of ACS2 was inhibited by about half mainly at breaker and B+4 stages, while ACS4 was down-regulated by about two thirds at almost all of the stages tested in SlELP2L-silenced lines. ACO1 was decreased by about two thirds mainly at B+4 stage and the most significant repression (about 75–85%) of ACO3 expression was observed at breaker and B+7 stages. Furthermore, the expression of three known ethylene-regulated ripening-related genes was also detected between WT and transgenic fruits. The expression of RIN, which encodes a MADS-box protein necessary for the onset and completion of fruit ripening35, was reduced by about one third in SlELP2L-silenced fruits mainly at B+4 and B+7 stages compared with that of WT fruits (Fig. 6). NOR, a NAC transcription factor gene necessary for fruit ripening36, its expression was inhibited by about half at B+4 stage. For E8, which responds specifically to ethylene37, its transcript was decreased by about half in SlELP2L-silenced fruits at almost all of the stages detected (Fig. 6). These results suggest that the silencing of SlELP2L may suppress the expression of these ethylene- and ripening-related genes which are normally induced during ripening.

Figure 6
figure 6

Relative expression profiles of ethylene-related and ripening-related genes in pericarps between WT and SlELP2L-RNAi fruits.

MG, mature green; B, breaker; B+4, 4 day after breaker stage; B+7, 7 day after breaker stage. The expression data of mature green (MG) fruit were normalized to 1. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

Given that fruit ripening is correlated with the occur of cell wall alterations38, thus the expression of four ripening-related cell wall metabolism genes, PG (polygalacturonase)39, PE1 (pectineaterase)40, CEL2 (endo-beta-1,4-glucanase)41 and XTH5 (Xyloglucan endotransglucosylase/hydrolase)42 was also detected between WT and transgenic tomato fruits from breaker to B+7 stage. Fig. 7 showed that the expression of PG was inhibited by about two thirds at B+4 stage, while PE1 expression was decreased by about half at breaker stage. For CEL2, its expression was down-regulated significantly at all of the stages tested in SlELP2L-silenced lines, while 50–70% inhibition of XTH5 expression was observed at B+4 and B+7 stages. Consistent with the down-regulated expression of ethylene- and ripening-related genes, these results further suggest that SlELP2L may affect the process of tomato fruit ripening.

Figure 7
figure 7

Relative expression profiles of cell wall metabolism genes in pericarps between WT and SlELP2L-RNAi fruits.

B, breaker; B+4, 4 day after breaker stage; B+7, 7 day after breaker stage. The expression data of breaker fruits were normalized to 1. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

Expression analysis of tomato DNA methyltransferase genes

In eukaryotic nuclei, DNA is tightly wrapped around octamers of histones to form the chromatin and chromatin organisation is crucial for the control of gene expression and other nuclear processes. Changes in chromatin structure are determined by protein complexes involved in the ATP-dependent repositioning of nucleosomes, in histone posttranslational modifications and in the methylation of cytosine residues43,44. In Arabidopsis, ELP2 has been shown to be required for pathogen-induced dynamic DNA methylation changes in defense related genes. The elp2 mutation alters (increases or decreases) methylation levels of specific methylcytosines45. To investigate whether SlELP2L also modulates DNA methylation in tomato, the transcript levels of genes encoding DNA methyltransferases were examined from MG to B+7 stage between WT and transgenic lines. Consistent with the results reported previously45, the expression of SlDRM7, SlDRM8 and SlMET1 genes43 was higher in SlELP2L-silenced lines than in the WT (Fig. 8), indicating increased methylation levels. Among which, the expression of SlDRM7 and SlDRM8 in SlELP2L-silenced lines was about twice as WT. SlMET1 expression was only slightly up-regulated in transgenic lines, while there is no obvious change was detected in SlDRM5 expression. These results indicate that SlELP2L may regulate gene expression through DNA methylation.

Figure 8
figure 8

Relative expression profiles of DNA methyltransferase genes in pericarps between WT and SlELP2L-RNAi fruits.

MG, mature green; B, breaker; B+4, 4 day after breaker stage; B+7, 7 day after breaker stage. The expression data of mature green (MG) fruits were normalized to 1. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

SlELP2L-silenced tomato was hypersensitive to ABA in seedling growth

Previous reports showed that Elongator mutants are hypersensitive to ABA, suggesting that they are involved in ABA signaling23,24. We showed that the transcripts of SlELP2L mRNA were induced slightly by ABA treatment (Fig. 2A). To further investigate the effect of ABA on tomato seedling growth, WT and transgenic seeds of consistent germination were transferred to MS medium containing 0 and 10 μM ABA for 7 days. As shown in Fig. 9A, no discernible difference in the length of shoot and root between WT and SlELP2L-RNAi seedlings was observed when grown in the control medium. Addition of 10 μM ABA to the incubation medium, the seedling growth was inhibited in both WT and transgenic tomato, while stronger inhibition was observed in transgenic seedlings (Fig. 9A). Then the length of roots was measured. Fig. 9B showed that the root length of transgenic seedlings was reduced more significantly than that of WT under ABA treatment, indicating that the silencing of SlELP2L causes a obvious increase in ABA sensitivity during seedling growth. We further detected the content of endogenous ABA in the leaves of WT and SlELP2L-silenced plants using an ELISA. The results show that ABA content was increased in transgenic plants (Fig. 9C).

Figure 9
figure 9

Seedling growth of SlELP2L-RNAi lines was hypersensitive to ABA.

A and B. Representative growth performance (A) and root length (B) of WT and transgenic seedlings. WT and SlELP2L transgenic tomato seeds were surface-sterilized and germinated in moist and disinfectant filter papers for three days, then the seeds of consistent germination (n = 20 each) were transferred to MS medium containing 0 and 10 μM ABA. Seedlings were photographed seven days after transfer. (C). ABA content in the leaves of WT and SlELP2L transgenic plants. Each value represents the mean ± SD of three replicates. Asterisks indicate a significant difference (P < 0.05) between WT and transgenic lines.

Full size image

Discussion

The multi-subunit complex Elongator has been reported to be involved in diverse biological processes in yeast, human and Arabidopsis, including exocytosis, cell proliferation, cell differentiation, embryogenesis and response to biotic and abiotic stress4,8,19,21,25,46. The Arabidopsis Elongator mutants have narrow and elongated leaves, decreased primary root growth, disorganized shoot apical meristem, aberrant inflorescence and flower architecture, reduced germination frequency and delayed seedling growth21,22,25. Leaf growth is a complex developmental process controlled by genetic and environmental factors and is determined by a proliferation, expansion and maturation phase47. Leaves of Arabidopsis elo mutants have larger and fewer cells, suggesting that the cell division rate of these leaves is reduced, which may be responsible for the reduced plant growth21. Mutations in Elongator subunits also result in aberrant DNA replication and increased DNA damage in plant cell cycle progression and alter cell division affecting leaf polarity formation48. Here, we explored the function of tomato SlELP2L by RNAi-mediated gene silencing. Similar to the reports in Arabidopsis, the silencing of SlELP2L in tomato also causes pleiotropic phenotypes, such as delayed seedling growth, rapidly senescing leaves and sepals, dark-green fruits and hypersensitivity to ABA (Fig. 3, Fig. 4, Fig. 9), suggesting SlELP2L plays an important role in diverse developmental processes. It was reported that Elongator modulated crosstalks of several hormones like auxin, jasmonic acid, ethylene and ABA, appearing to be essential to maintain hormonal balance, as disruption of Elongator shifts hormonal signaling, resulting in pleiotropic growth and defense phenotypes25. In our present study, reduced GA and IAA contents were detected in SlELP2L-RNAi plants (Fig. 3). GA has a crucial role in regulating internode length and leaf size26 and IAA has profound effects on plant growth and development, including seedling growth27. Thus possible hormonal misregulation might contribute to the pleiotropic phenotypes of SlELP2L-silenced plants.

RNAi repression of SlELP2L resulted in elevated mRNA accumulation of SlELP1L, SlELP3L, SlELP4L and SlELP6L (Fig. S1). Thus SlELP2L and the other tomato Elongator subunits expression appears a certain degree of functional compensation. This is consistent with that the disruption of each Elongator subunit (ELP1–ELP6) results in similar phenotypes in Arabidopsis, suggesting that plant Elongator subunits work as a complete complex and share common functions21,23. Although SlELP2L is highly homologous to Arabidopsis ELP2, distinct phenotypes were observed between Arabidopsis elp2 mutant and SlELP2L-silenced tomato plants, such as reduced leaf rachis length, rapid senescence of leaves and sepals and dark-green fruits (Fig. 3, Fig. 4), indicating the diverse functions of Elongator in different organisms. Similar alteration of fruit coloration phenotype was observed in the light-hypersensitive mutants dark green (dg) and high pigment 1 and 2 (hp1 and hp2) in tomato characterized by their exaggerated light responsiveness. dg is an alternative allele at the HP-2 locus, identified as the tomato DET1 gene which involves in the light signal transduction and morphogenesis49,50. DET1 was also reported to bind the amino-terminal tail of histone H2B51. Besides, Arabidopsis histone acetyltransferase TAF1/HAF2, GCN5 and histone deacetylase HD1/HDA19 are necessary for the light regulation of plant development and gene expression52. These evidences establish a potential connection between histone modification and light signal transduction in plants. Furthermore, Elongator plays a vital role in transcriptional elongation via histone and α-tubulin acetylation3,4,5. Therefore, the silencing of SlELP2L in tomato may lead to abnormal histone modification and thus affects the light signaling pathway, which further affects fruit development and coloration and results in dark-green fruits.

Elongator was first isolated together with RNAPII and the studies in yeast and humans suggest that one of Elongator's functions is to facilitate mRNA transcription elongation during transcription by RNA polymerase II2,23,53, thus the mutations in Elongator would decrease the transcript levels of their target mRNAs. In yeast and Arabidopsis, Elongator subunit deletions or mutations lead to both up- and down-regulated gene expression23,54. In this study, we found that repressed SlELP2L reduces the expression of Rbcs-2 (Fig. 3G), ethylene-related genes (ACS2, ACS4, ACO1 and ACO3) and ripening-related genes (E8, NOR and RIN) (Fig. 6). Besides, the expression of ripening-related cell wall metabolism genes, PG, PE1, CEL2 and XTH5 was also down-regulated (Fig. 7). We also found, however, that the expression of RCHY1 was upregulated in SlELP2L-RNAi plants (Fig. 3H). To date it has been well documented that ethylene plays a key regulatory role in fruit ripening55 and most ripening-deficient mutants fruit characterized are the consequence of defects in carotenoid biosynthesis56. In addition, reduced expression of SlELP2L also results in reduced carotenoid content (Fig. 5B). Because SlELP2L may be involved in the mRNA elongation process, SlELP2L may not directly participate in the regulation of tomato growth and development, but by regulating another gene (s) that plays a role in plant growth and development. Besides, the higher transcript levels of DNA methyltransferase genes, SlDRM7, SlDRM8 and SlMET1, were observed in SlELP2L-silenced lines than that of WT (Fig. 8), suggesting that SlELP2L may regulate gene expression through DNA methylation, thereby affecting the expression of related genes. Together, these results indicate that SlELP2L may modulate DNA methylation in tomato and the silencing of SlELP2L in tomato affects fruit ripening possible via regulating the expression of ethylene- and ripening-related genes. Besides, consistent with the role of Rbcs-2 and RCHY1, the SlELP2L-RNAi plants expressed reduced Rbcs-2 and improved RCHY1 may partially explain the phenotypes of accelerated senescence of leaves and sepals.

Recent studies have suggested that mRNA metabolism is closely related to ABA signal transduction. For example, an ABH1 (ABA-hypersensitive 1) mutation encodes an mRNA capbinding protein which, together with CBP20, forms a heterodimeric nuclear cap-binding complex. Mutations in either ABH1 or CBP20 render plants hypersensitive to ABA in seed germination and seedling growth57,58. Other important components, such as transcription factors and those in the ABA signaling pathway can also affect the sensitivity to ABA59,60. The plant hormone ABA is involved in regulating many aspects of plant growth and development, such as seed dormancy, seed germination, seedling growth and plays important roles in plant responses to environmental stresses. ABA also has a role in inducing leaf abscission, flower and leaf senescence23. Interestingly, we found that SlELP2L-RNAi plants had greater sensitivity to ABA than the WT plants during seedling growth (Fig. 9). Similarly, it has been shown that all the four Elongator mutants (abo1/elo2/elp1, elp2, elo3/elp3 and elo1/elp4) in Arabidopsis exhibit ABA hypersensitivity in seed germination and seedling growth, suggesting important roles for Elongator in ABA signaling23. Transcriptional elongation mediated by RNA polymerase II is a critical process in gene regulation and is highly regulated in eukaryotes by numerous factors in mRNA biogenesis and maturation61. Together, these results suggest that plant responses to hormone ABA can be regulated by different molecular mechanisms and at different stages during mRNA processing, including mRNA metabolism, transcriptional initiation, transcriptional elongation and protein modifications23,59,60,62.

In conclusion, this study concerned the morphological, physiological and molecular features of SlELP2L-RNAi transgenic tomato. Presently, the functional studies of Elongator in plants are mainly limited in the model plant Arabidopsis and the specific biological functions of most plant Elongator still remain to be elucidated. Clearly, understanding the role of SlELP2L will not only extend knowledge of the biological function of Elongator, but also provide insight into exploring further the significance of Elongator in regulating plant growth and development, as well as in responses to ABA. Besides, previous reports indicated that mutants of different subunits of Elongator complex exhibit similar phenotypes in yeast, humans and plants18,54,63, while whether other subunits of tomato Elongator complex have similar functions with SlELP2L still remain to be further elucidated.

Methods

Plant materials and growth conditions

The seeds of wild-type (WT) tomato (Solanum lycopersicum Mill. cv. Ailsa Craig) and SlELP2L-RNAi transgenic tomato (T1) were germinated and the seedlings were grown in a greenhouse under following conditions: 16/8 h day/night cycle, 25/18°C day/night temperature and 80% humidity. All the plants were watered daily. For organ-specific expression profiling of SlELP2L, the tissues of roots, stems, leaves, sepals, flowers and fruits of different period from tomato plant with ripe fruits were collected. Flowers were tagged at anthesis and fruit development recorded as days post-anthesis (DPA). Immature green (IMG) fruits were defined as 28 DPA. Mature green (MG) fruits were defined as 35 DPA and characterized as full fruit expansion and shiny with no obvious mature-related color change. Breaker (B) were defined as fruit emerging the first clearly orange. Subsequent ripening stages were defined in days post-breaker: B+4 (4 day after breaker) and B+7 (7 day after breaker). All plant materials for total RNA extracting were taken in the same time each day and frozen in liquid nitrogen and stored at −80°C until needed.

Plant hormone treatments

To verify the expression profiles of SlELP2L under phytohormone treatments, the WT tomato seeds were germinated and the seedlings were grown in greenhouses. Then three-week old WT tomato seedlings were treated with 100 μM ABA, 50 μM GA3 and 100 μM ACC, respectively64. Plants were enclosed in plastic immediately following spraying and left for 3, 6, 12 and 24 h. Individual seedlings were used in each treatment with three biological replicates. Unstressed seedlings were used as control. All the samples were collected at certain time points and frozen in liquid nitrogen for Quantitative RT-PCR analysis.

ABA sensitivity experiment was performed as follows: WT and transgenic tomato seeds were germinated according to the above procedure, then the seeds of consistent germination were selected and transferred to MS medium containing 0 and 10 μM ABA. Culture vessels containing the seeds were incubated in a growth chamber for 7 d, then seedlings were taken picture and root length was measured.

SlELP2L isolation and sequence analysis

The primers for cloning the coding region of SlELP2L were as follows: Forward, 5' TACACTGGCTGGTCATAAAG 3' and Reverse, 5' GACAAGGCTAAAGCAAACAT 3'. The PCR procedure was incubated at 94°C for 5 min followed by 35 cycles of 30 s at 94°C, 30 s at 56°C and 2 min at 72°C and a final extension at 72°C for 10 min. The amplified products were subcloned into pMD18-T vector (Takara, Dalian, China) and confirmed by sequencing. The translation of SlELP2L gene using the ExPASy translate tool, grand average of hydropathicity (GRAVY) and theoretical molecular weight (Mw) were calculated using the ExPASy ProtParam tool (http://www.expasy.org/). Sequence alignment was conducted using the DNAMAN 5.2.2 program. Conserved structure domains were annotated according to ScanProsite (http://prosite.expasy.org/scanprosite/). To analyze putative cis-elements in the promoter region of SlELP2L gene, promoter sequence (1000 bp regions upstream the 5′ end of the predicted ORF) of SlELP2L gene was extracted from SGN database and searched against the promoter database PLACE (http://www.dna.affrc.go.jp/PLACE/index.html)65.

Quantitative RT-PCR analysis

Total RNA was isolated using Trizol (Invitrogen, Shanghai, China) according to the manufacturer's instructions and was treated with the Dnase-I (Promega, Beijing, China) for 30 min at 25°C. The first strand cDNA synthesis was performed using M-MLV reverse transcriptase (Promega, Beijing, China) with oligo (dT)18 primer. Quantitative RT-PCR was carried out using the CFX96™ Real-Time System (Bio-Rad, USA). All reactions were performed using the SsoFast™ EvaGreen® Supermix (TaKaRa, Dalian, China) according to the manufacturer's recommended protocol. Quantitative RT-PCR reactions were performed using a two-step method: 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. For each gene analysis, NTC (no template control) were also performed. Tomato CAC gene was selected as internal standard for organ-specific expression studies66 and tomato EFα1 gene was used as internal control under hormone treatments67. The analysis of relative gene expression levels were detected using the 2−ΔΔCT method68. All Quantitative RT-PCR experiments were taken in three biological replicates and each reaction was run in triplicate. Gene-specific primers used for quantitative RT-PCR are listed in Table S2 and a standard curve by serial dilution was analyzed for each specific gene using wild-type tomato cDNA.

RNAi vector construction and plant transformation

A 313 bp SlELP2L-specific DNA fragment was amplified from wild-type fruit cDNA using the following primers: Forward, 5' CGGTGGTACCAAGCTTGGAAATCCGTTGGTCGTT 3' with Kpn I & Hind III restriction sites and Reverse, 5' CCGCTCGAGTCTAGATTGAAAGGTGGAAGTGTC 3' with Xho I & Xba I restriction sites, respectively. The PCR products were divided into two groups: one group was digested with Hind III & Xba I, linked into the pHANNIBAL plasmid at Hind III & Xba I restriction sites in the sense orientation, then another group was digested with Kpn I & Xho I and linked into the above plasmid digested with the same restriction sites in the antisense orientation. After that, the double-stranded (ds) RNA expression unit containing the cauliflower mosaic virus (CaMV) 35S promoter and terminator, was digested with Sac I & Spe I and linked into the plant binary vector pBIN 19 to form SlELP2L-RNAi vector. Subsequently, this resulted vector was transformed into WT tomato cotyledon explants via Agrobacterium strain LBA440469. Positive transgenic lines were screened for kanamycin (50 mg L−1) resistance. Genomic DNA of transgenic tomato plants was extracted according to the manufacturer's protocol of Genomic DNA Extraction Kit (Invitrogen, Shanghai, China) and was used for detection of the presence of T-DNA by PCR with primers NPTII-F (5' GACAATCGGCTGCTCTGA 3') and NPTII-R (5' AACTCCAGCATGAGATCC 3'). T1 seeds from the selected transgenic lines were germinated on a medium with kanamycin (50 mg L−1) and positive lines were selected for further study.

Extraction and quantification of plant phytohormones

The total endogenous contents of ABA, IAA and GA in the leaves of wild-type and transgenic plants were determined using one-step double-antibody sandwich method enzyme-linked immunosorbent assay (ELISA). For each test, a 0.5 g sample was ground in anice-cooled mortar in 5 mL 80% (v/v) methanolextraction medium, incubated at 4°C for 4 h, then the extracts were centrifuged at 4,000 rpm for 15 min at 4°C and supernatant was analyzed by ELISA. All of the ELISA experiments were carried out by using the same procedure according to the manufacturer's instructions (http://www.wksubio.com/product/5794959.html). Contents for three kinds of hormones were all measured at 450 nm using a Bio-Rad680 microplate reader. The final results are means of three replicates. The correlation coefficients (R2) of ABA, IAA and GA standard curves were 0.992, 0.995 and 0.991, respectively. The concentration of each sample was calculated from the linear regression equations.

Fruit chlorophyll and carotenoid determination

Tomato chlorophyll and carotenoid were extracted from pericarp tissues using a modified protocol from Forth and Pyke70. A section of fruit pericarp was cut from an area above a locule in a 5 mm wide strip around the equator of fruits, weighed and ground into powder in a mortar with liquid nitrogen, then this powder was transferred into a 50 mL centrifuge tube and extracted with 10 mL of hexane:acetone (60:40, v/v). The extract was centrifuged at 4000 rpm for 5 min and the supernatant was carefully transferred to a fresh tube. The cell debris was repeatedly extracted with fresh solvent until colorless and the pooled supernatants volume was measured. The optical absorbance of the supernatants was immediately measured in a lambda 900 scanning spectrophotometer (PerkinElmer). The amount of pericarp chlorophyll and carotenoid contents was calculated by the following equation: total chlorophyll mg mL−1 = 8.02 (OD643) + 20.2 (OD647) and total carotenoid mg mL−1 = (OD450)/0.25. Individual tissue samples were taken from 3–4 fruits for each ripening stage in triplicate.

Measurement of plant leaf area

The digital image processing method was adopted to measure plant leaf area. The photos of plant leaf were taken, then Photoshop CS5 software was used to count pixels and determine the unit pixel area. The leaf area was calculated by the following equation: Leaf area = number of leaf pixels/the number of pixels per unit area contained.