MiR1918 enhances tomato sensitivity to Phytophthora infestans infection

Late blight of tomato is caused by the oomycete pathogen Phytophthora infestans. In our previous work, we identified and characterized a miR1918 in P. infestans (pi-miR1918), and showed that its sequence is similar to the sequence of tomato miR1918 (sly-miR1918). In this study, we used Arabidopsis thaliana pre-miR159a as a backbone to synthesize pi-miR1918 via PCR and mutagenesis. The artificial pi-miR1918 was used to investigate the role of miR1918 in tomato-P. infestans interaction. Trangenic tomato plants that overexpressed the artificial pi-miR1918 displayed more serious disease symptoms than wild-type tomato plants after infection with P. infestans, as shown by increased number of necrotic cells, lesion sizes and number of sporangia per leaf. The target genes of pi-miR1918 and sly-miR1918 were also predicted for tomato and P. infestans, respectively. qPCR analysis of these targets also performed during tomato-P. infestans interaction. The expression of target gene, RING finger were negatively correlated with miR1918 in the all Lines of transgenic tomato plants. In addition, we used the 5′ RACE to determine the cleavage site of miR1918 to RING finger. These results suggested that miR1918 might be involved in the silencing of target genes, thereby enhancing the susceptibility of tomato to P. infestans infection.

Scientific RepoRts | 6:35858 | DOI: 10.1038/srep35858 Small RNAs (sRNAs) have been reported to move horizontally between plant and pathogen 12 . A total of 37 tomato miRNA and 38 tomato miRNA* sequences have been predicted, and most of the miRNA and miRNA* have high complementarity with the open reading frames (ORFs) of the genomic RNAs of CMV-Fny and CMV-Q, respectively, suggesting that these miRNAs may repress the translation or induce the cleavage of target genes 13 . In addition to plant miRNAs that can be transferred into viruses, several pathogens have also been shown to use RNA interference to suppress the expression of host genes 14,15 . For instance, genome-wide identification has predicted that vsiRNAs from Grapevine fleck virus (GFkV) and Grapevine rupestris stem pitting associated virus (GRSPaV) may target plant transcripts 16 . Three Botrytis cinerea sRNAs (Bc-SiR3.1, Bc-siR3.2 and Bc-siR5) have been shown to target different sets of A. thaliana genes involved in immunity against B. cinera 10,17 . Bc-siR3.1 targets the oxidative stress-related gene peroxiredoxin, whereas Bc-siR3.2 targets the mitogen-activated protein kinase 1 and protein kinase 2 gene (MPK1 and MPK2), while Bc-siR5 targets the cell wall-associated kinase gene.
Artificial miRNA (amiRNA) has emerged as a new gene regulation strategy, designed to suppress pathogen genes. AmiRNA was engineered by replacing the mature miRNA/miRNA* sequence with the complementary sequence taken from the host miRNA precursors [18][19][20] . At present, the precursors of Arabidopsis miR319, miR159, miR168 and miR171 are being used as the backbones to construct amiRNAs. For instance, two amiR-NAs have been constructed from the precursor of Arabidopsis miR159; one targets the overlapping sequence between CMV2a and 2b, and the other targets the untranslated region (UTR) of cucumber mosaic virus (CMV). Transgenic tomato plants overexpressing these two amiRNAs develop good resistance against CMV 21 . In addition, transient expression of vsiRNA-4A as amiRNAs in Nicotiana benthamiana can cause leaf twisting and stunting 22 . Overall, amiRNA provides a new way to study the function of sRNAs.
Tomato (Solanum lycopersicum) is a major crop plant that is widely cultured in the world 23 . However, field-grown tomatoes are susceptible to disease such as late blight (LB), which is one of the most devastating diseases of field-grown tomatoes, and it can cause severe economic losses for tomato growers [24][25][26] . Late blight is caused by the pathogen Phytophthora infestans. Our previous work have focused on the use of high-throughput sequencing to identify tomato miRNAs related to resistance against P. infestans as well as on the functional analysis of P. infestans-induced miRNAs in tomato, including miR169, miR398, miR482, miR6024 miR6026, miR6027 [27][28][29] . MiR1918 is expressed in all tomato tissues, and it was initially thought to be involved in the regulation of fruit development and maturation 30 . Recently, a number of P. infestans miRNAs (pi-miRNAs) have been predicted using P. infestans EST data and all the known miRNAs from miRBase, and pi-miR1918 is among these predicted pi-miRNAs, and its level was found to change dramatically in P. infestans-infected Solanum pimpinellifolium leaves 31 . In addition, a recent silico study has shown that tomato miR1918 might bind to the AC1 gene coding the replication-related protein in Tomato leaf curl virus (ToLCV). This suggests that it might also target the AC1 gene and inhibit ToLCV replication 32 .
The present study aimed to understand the roles of miR1918 in tomato -P. infestans interaction by analyzing the expressions of miR1918 and its target genes in tomato and P. infestans during the tomato-P. infestans interaction. The result suggested that miR1918 enhanced the susceptibility of the tomato to P. infestans infection.

Results
Tomato and P. infestans transcripts targeted by miR1918. To investigate the potential role of miR1918 during tomato-P. infestans interaction, we first predicted the target genes of sly-and pi-miR1918 through searching the tomato and P. infestans databases, respectively. Using psRNATarget and TargetAlign, which are bioinformatics tools for identifying mRNAs targeted by miRNA, we obtained a number of target genes for miR1918. 56 target genes of sly-miR1918 were predicted from tomato cDNA library by psRNATatget (maximum expectation ≤ 4) (Table S1). Three target genes which were maximum expectation ≤ 2.5, Os12g0283800-like gene (sly-TG1, Solyc10g045620.1.1), RING finger (sly-TG2, Solyc01g095820.2.1) and MATE efflux gene (sly-TG3, Solyc10g054110.1.1) were selected for next analysis. Little is known about the function of sly-TG1, but the functions of sly-TG2 and sly-TG3 have already been demonstrated. Sly-TG2 is generally associated with ubiquitination implicated in plant defense against biotic and abiotic challenges 33,34 and sly-TG3 is an essential component of SA and ABA-dependent signaling for disease and abiotic stress resistance 35,36 . As for pi-miR1918, three target genes were predicted from the data of P. infestans mRNAs and 3′ UTR sequences by TargetAlign. These were secreted RxLR effector peptide protein (pi-TG1, PITG_19655), trafficking protein particle complex subunit (pi-TG2, PITG_14806) and phosphoethanolamine N-methyltransferase (pi-TG3, PITG_15903) (Fig. 1a,c).
Quantitative real time RT-PCR (qRT-PCR) was used to examine the expression of sly-miR1918 and pi-miR1918 as well as their target genes in tomato and P. infestans, respectively after infection with P. infestans. The expression of sly-miR1918 was down-regulated except at 72 hpi (Fig. 1b). Its target gene sly-TG1 was up-regulated during the first 3 hpi, whereas sly-TG2 was up-regulated in the first 24 hpi. The expression of sly-TG3 actually decreased at all time points following infection. As for pi-miR1918, its expression during the first 3 hpi was not significantly changed, but was decreased slightly after 3 hpi. The expression of pi-TG3 increased at all time points following infection (Fig. 1d). In addition, the expression of secreted pi-TG1 was negatively correlated with that of pi-miR1918 from during the first 3 hpi and from 24 to 48 hpi, whereas the expression of secreted pi-TG2 was negatively correlated with that of pi-miR1918 from 24 to 72 hpi.
Construction of artificial P. infestans miR1918 vector. As previous studies have shown that miR159 (ath-miR159) is expressed at high levels in A. thaliana, the hairpin structure of ath-pre-miR159a was chosen as the backbone for the design of amiRNA. The design of artificial P. infestans miR1918 (pi-amiR1918) is shown in Fig. 2a. Oligonucleotide-directed mutagenesis was used to replace ath-miR159a and ath-miR159a* with pi-miR1918 and pi-miR1918*, respectively. The last round of PCR was performed to amplify the sequence of pi-amiR1918. The products from these three rounds of PCR were detected using agarose gel electrophoresis. The lengths of the amplified fragments were consistent with the expected sizes (Fig. 2b). In addition, sequence alignment between ath-pre-miR159a and pi-amiR1918 showed that both sequences were conserved except for two regions of 21-nt, marked as region 1 and region 2 in Fig. 2c. Pi-miR1918 was derived from ath-miR159a* by a mutation within region 1 or from ath-miR159a by a mutation within region 2. The secondary structures of ath-pre-miR159a and pi-amiR1918 obtained with Mfold 3.5 were consistent each other (Fig. S1). Overall, the data suggested that pi-amiR1918 was successfully constructed by using ath-pre-miR159a as the backbone.
To further investigate the role of miR1918, a plasmid for the overexpression of pi-amiR1918 was constructed on the basis of the pBI121 vector. The DNA containing the pi-amiR1918 sequence was cloned into Xbal I-Sac I cut pBI121, placing it under the control of the Cauliflower mosaic virus (CaMV) 35S promoter (Fig. 2d). The pBI121-pi-amiR1918 construct obtained was sequenced to confirm the presence of pi-amiR1918 and the result of multiple sequence alignment between pBI121-pi-amiR1918 and pi-amiR1918 indicated that pBI121-pi-amiR1918 was successfully constructed (Fig. S2).
Overexpression of amiR1918 by agrobacterium infiltration. The pBI121-pi-amiR1918 plasmid was mobilized into Agrobacterium tumefaciens strain GV3101 by the freeze-thaw method. As shown in Fig.  S3, pBI121-pi-amiR1918 was successfully mobilized into GV3101 by PCR using the primers amiR1918*-F and amiR1918-R (Table S1). Introduction of A. tumefaciens harbouring pBI121-pi-ami1918 into tomato leaves resulted in significant up-regulation of miR1918 (sly-miR1918 and amiR1918) expression during the first seven days (Fig. 3a). The expression of miR1918 was also negatively correlated with the expression of the target genes. Three days after the transient overexpression of amiR1918, the leaves were infected with P. infestans. Resistance analysis showed that the extent of the disease was more severe and the lesions were significantly larger in P. infestans-infected leaves that transiently overexpressed amiR1918 (OE1918) than in P. infestans-infected leaves that did not overexpress amiR1918 (EV) (Fig. 3b,c). These results suggested that overexpression of amiR1918 may have weakened the resistance of tomato against P. infestans.

Overexpression of amiR1918 in tomato enhances sensitivity to P. infestans infection. To inves-
tigate whether miR1918 is involved in tomato-P. infestans interaction, pBI121-pi-amiR1918 was introduced into tomato. After callus induction and shoot regeneration, three representative positive transgenic lines were confirmed by selection on medium containing kanamycin. Four-week-old tomato seedlings (WT and transgenic lines) were transferred to pots filled with perlite/vermiculite (1:3, v/v) and grown for 2 weeks before they were subjected to qRT-PCR analysis to estimate the levels of miR1918 expression in the leaf tissues (Fig. 4). Compared
to the wild types, all the transgenic lines expressed a higher level of amiR1918, with the levels of amiR1918 in Line 1, Line 2 and Line 3 being approximately 1.5 fold, 2.4 fold, and 1.4 fold the level of amiR1918 in WT, respectively.
The three transgenic and WT tomato plants were infected with P. infestans, and seven days after the infection, the physical appearance of the plants was assessed. The leaves of the transgenic plants showed necrotic lesions, whereas the leaves of WT plants exhibited fewer symptoms of the disease (Fig. 5a). Trypan blue staining also revealed increased sensitivity to the pathogen in the case of the transgenic plants, as shown by the presence of increased number of necrotic cells compared to WT (Fig. 5b). In addition, the transgenic plants also had larger lesion size and more sporangia per leaf than the WT plants (Fig. 5c,d). These results suggested that overexpression of amiR1918 enhanced the susceptibility of the plants to P. infestans infection.
The identification of target genes by 5′ RACE. The expressions of the target genes also varied between transgenic lines and WT. In the case of sly-TG1, its level in Line 1 was approximately 1.5 fold that of WT, but did not really increase in Line 2. As for sly-TG2, its level in Line 1 was approximately 3.3 fold that of WT. In the case of sly-TG3, its levels in Line 1 and Line 3 were about 3 fold that of WT. Intriguingly, the expression level of amiR1918 was the highest in Line 2. Also in Line 2, the expressions of sly-TG1 and -TG2 were down-regulated, whereas the expression of sly-TG3 was up-regulated. The expression pattern analysis showed there was no negative correlation between the expression of amiR1918 and the expression of sly-TG1 and sly-TG3 (Fig. 1b). In addition, we also used another reference control gene, α -tubulin to perform the qRT-PCR analysis in tomato. The results indicated there was negative correlation between expression of miR1918 and the expression of sly-TG2 and only sly-TG2 expression was down-regulated in all transgenic lines (Fig. S4).
The expression of miR1918 and its target genes (sly-TG1, sly-TG2 and sly-TG3) in WT and three transgenic lines were also detected by qRT-PCR seven days after infection with P. infestans. The expression of miR1918 was significantly up-regulated in Line 2 and Line 3 (p < 0.05), but not in Line 1. The expressions of sly-TG1 and sly-TG2 were down-regulated in all three transgenic lines, while the expression of sly-TG3 was significantly down-regulated in Line 2 and Line 3 (Fig. 5e). However, the most significant changes in the expression levels of miR1918 and its target genes were detected in Line 2. To identify the target genes of miR1918 in tomato, we used the 5′ RACE to determine the cleavage site. Total RNA was extracted from overexpression miR1918 tomato (Line 2); 5′ RNA adapter was ligated using T4 RNA ligase; the ligated mRNA was reversed using Oligo dT. After PCR and Sanger sequence, we only found that 3 out of 10 cloned sequence had the same cleavage site of sly-TG2 between 10 th nucleotide and 11 th nucleotide (Fig. S5). This suggested that only sly-TG2 was the target genes of miR1918 in tomato.

Discussion
As a novel type of regulatory factor, plant miRNAs play an important role in plant-pathogen interaction 37 . For instance, our previous work has shown that the miR396 family miRNAs target growth-regulating factor (GRF) genes. Tobacco plants overexpressing Sp-miR396a-5p show increased susceptibility to P. nicotianae infection 38 . Although plant miRNAs have been extensively studied, few miRNAs have been discovered in microbes. Computational analysis has identified 78 mRNAs as potential target transcripts of 14 milRNAs from M. anisopliae 39 A number of miRNAs-like RNAs have been identified in the plant fungal pathogen Sclerotinia sclerotiorum by high-throughput sequencing 11 . In our previous study, some miRNAs in P. infestans, including miR1918, were predicted from the P. infestans EST data and all the known miRNAs from miRBase 27 . P. infestans and tomato are two unrelated organisms, but both contain the conserved sequence of miR1918, and therefore studying the function of miR1918 would be of great significance.
RNA silencing is considered to be an innate immune response in plant and it is important that a pathogen evades this line of defense in order to establish successful infection. Small RNA silencing that targets endogenous mRNAs may have a distinct biogenesis and involves miRNAs or siRNAs. During plant-microbe interaction, plant miRNAs may regulate their own effector genes or transposable elements (TEs) 12 , and RNA silencing further affects the resistance of a plant to pathogen 29 . Besides, sRNAs have been reported to move horizontally between plant and pathogen. Plant miRNAs can be transferred to viruses, and in return pathogen-derived sRNAs can bind to host transcripts when the plant interacts with microbes, including virus and fungus 12,16,17,40 . Overall, a plant can recognize and develop various means to defend itself against the attacks by pathogens, and thus the pathogens need to have strategies that would facilitate their infection and create a suitable environment to grow and reproduce inside the host plant. A negative correlation was observed between the expression of pi-miR1918 and some of its targets in tomato infected with P. infestans (Fig. 1b). Moreover, sly-miR1918 was detected in tomato leaves inoculated with P. infestans and its expression level was down-regulated at all time points except at 72 h following the inoculation. Its target genes, sly-TG1and sly-TG2 were up-regulated following the inoculation, although at different time points (Fig. 1a). These results suggested that pi-and sly-miR1918 could silence their respective targets.
Using 5′ -RACE, sly-TG2 was identified as the target genes of miR1918 in tomato. Sly-TG2 (RING finger gene) belongs to a large family of zinc finger-coding genes that play important roles in the regulation of growth and development, hormone signaling, and responses to biotic and abiotic stresses in plants. RING finger proteins, which are generally associated with ubiquitination, are implicated in plant defense against biotic and abiotic factors 33,34 . For example, the expression of a RING finger protein-coding gene from the Chinese wild grapevine Vitis pseudoreticulata (designated VpRFP1) has been shown to increase rapidly in mildew-resistant V. pseudoreticulata plants upon inoculation with grapevine powdery mildew Uncinula necator 41 . Overexpression of VpRFP1 in Arabidopsis plants has been found to enhance resistance to Arabidopsis powdery mildew Golovinomyces cichoracearum and a virulent bacterial pathogen Pseudomonas syringae pv. tomato DC3000. Another V. pseudoreticulata RING finger protein, EIRP1, can activate plant defense responses by inducing proteolysis of the VpWRKY11 transcription factor 42 . Furthermore, the expression of a novel E3 ubiquitin ligase RING1 gene (CaRING1) from pepper (Capsicum annuum) was found to be induced upon infection with avirulent Xanthomonas campestris pv vesicatoria, and overexpression of CaRING1 in Arabidopsis can confer enhanced  resistance to infection by hemibiotrophic Pseudomonas syringae pv tomato and biotrophic Hyaloperonospora arabidopsidis 43 . On the basis of these results, genes coding for RING finger proteins could be key regulators of plant immunity. Our results showed that amiR1918 was negatively correlated with the expression of RING finger-coding genes and that there were more disease symptoms and larger lesion size in the transgenic tomato plants (Fig. 4). This suggested that amiR1918 might be involved in silencing genes coding for RING fingers, a process that would eventually result in the tomato having enhanced susceptibility to P. infestans infection.
The predicted targets of pi-miR1918 in P. infestans might play important roles in tomato-P. infestans interaction. For example, pi-TG3 (which codes for phospoethanolamine N-methyltransferase) is related to glycerophospholipid metabolism. It participates in the synthesis of stress-associated second messenger such as phosphatidic acid. Phosphatidic acid can regulate various signal pathways that might be involved in the susceptibility of tomato to P. infestans infection. The RxLR effector gene, pi-TG1 is a target gene of pi-miR1918 in P. infestans. RxLR effector protein mediates its own entry into host cells after being secreted from the pathogens 44 . One previous study has shown that an RxLR effector from P. infestans can prevent the re-location of two plant NAC transcription factors to promote disease progression 45 . Another P. infestans RxLR effector interacts with host MAPKKK to suppress plant immune signaling 44 . In addition, RxLR effectors also suppress RNA silencing in plants, leading to enhanced susceptibility to Phytophthora. Two RxLR-like effectors, PSR1 and RSR2 are found in the oomycete plant pathogen P. sojae 46 . PSR2 is a member of the conserved RxLR effector family that is widely distributed in Phytophthora spp and it may contribute to the virulence of the fungus by suppressing host RNA silencing processes in various plants 47 . On the other hand, PSR1 can target PINP1 (a previously unidentified component of RNA silencing that regulates distinct classes of small RNAs in plants) in order to promote infection 48 .
In conclusion, an artificial pi-miR1918 was generated by PCR using the precursor of A. thaliana pre-miR159a as the backbone, and tomato plants that overexpressed this artificial pi-miR1918 displayed more serious disease symptoms than WT plants after infection with P. infestans. A few target genes of miR1918 were predicted for tomato and their expressions were found to be negatively correlated with the expression of miR1918. These results suggested that miR1918 might be involved in the silencing of target genes, thereby enhancing the susceptibility of tomato to P. infestans infection.

Collection of tomato and P. infestans specimens during their interaction. The tomato plant
S. lycopersicum Zaofen No.2 was chosen as the host plant. It was grown in a greenhouse under 16 h light at a temperature range of 22-28 °C. P. infestans strain P12103 used in the experiments was provided by Prof. Shan from Northwest A&F University of China. The specimen had been collected from Pingliang City, Zhuanglang County, Gansu Province, China, in 2009. The treated tomato plants (Ts, 5-6-leaf stage) were inoculated with a suspension of P. infestans spores (10 6 zoospores/mL; 5 individuals each) before being placed in a 100% relative humidity room or chamber in the dark to ensure spore germination at 20 ± 1 °C. The controls (Cs, 5 individuals) were treated with the same volume of sterile water and grown under the same conditions. Leaf samples were collected from each plant immediately after inoculation (0 hpi) and at 3, 12, 24, 36, 48 and 72 hpi, and the leaves from Ts and Cs plants were pooled. P. infestans P12103 was maintained on oat medium that had been covered with a layer of cellophane inside a 9-cm culture dish. The dish was kept at 20 °C in the dark. After 10 days, the cellophane, which was filled with P. infestans, was transferred to a new culture dish. Tomato leaves were used to cover the upper and lower sides of this cellophane as described previously 31 . P. infestans samples were collected from the cellophane at 0, 3, 12, 24, 36, 48 and 72 hpi. All samples were quickly frozen in liquid nitrogen and stored − 80 °C for storage until DNA and RNA isolation.
Construction of amiR1918 vector. AmiR1918 was generated by PCR, which involved amplification and mutagenesis using the precursor of ath-pre-miR159a as the backbone 21,49 . The sequence of ath-pre-miR159a was cloned by PCR using the primers, ath-pre-miR159-F and ath-pre-miR159-R (Table S1). The PCR fragment was cloned into pMD-19T vector (TaKaRa) according to the manufacturer's instruction.
To construct pi-amiR1918, four primers were designed: amiR1918-F, amiR1918-R, amiR1918*-F and amiR1918*-R (Table S1). Three rounds of PCR were performed to amplify pi-amiR1918. In the first round of PCR, pMD-19T-ath-pre-miR159a, amiR1918-F and amiR1918-R were used to amplify a fragment containing miR1918. In the second round of PCR, pMD-19T-ath-pre-miR159a, amiR1918 * -F and amiR1918 * -R were used to amplify a fragment containing miR1918 * . In the third round of PCR, the PCR products of the first and second rounds were used as the template along with amiR1918 * -F and amiR1918-R. The fragment from the third round of PCR was cloned to pMD-19T and its sequence was checked by DNA sequencing. Plasmid construction and transient overexpression in tomato. The DNA fragment obtained from the third round of PCR was digested with SacI and XbaI and then cloned into SacI-XbaI digested pBI121 to generate the recombinant plasmid pBI121-pi-ami1918, in which pi-amiR1918 was placed under the control of the strong constitutive CaMV35S promoter. This recombinant plasmid was transformed into A. tumefaciens strain GV3101 by freeze-thaw method.
A sample from A. tumefaciens culture was centrifuged at 4,000 × g for 10 min and the supernatant was discarded. The pellet was resuspended in infiltration medium (10 mM MES, 10 mM MgCl2 and 20 μ M acetosyringone) (OD600 = 0.1). Agrobacterium tumefaciens was introduced into tomato leaves by infiltration. Agrobacterium tumefaciens with empty vector was used as a control. After 3 days, the infiltrated leaf regions were inoculated with 20 μ l of P. infestans (10 6 zoospores/mL) per infiltrated leaf region. These regions were collected for RNA extraction at the indicated times (0, 2 and 4 d). In addition, the sizes of the lesions were calculated from the image obtained by ImageJ 17 .
Generation of pi-amR1918 transgenic tomato lines and P. infestans resistance analysis. Tomato cotyledon were infected with A. tumefaciens strain GV3101 containing the modified plasmid pBI121-pi-ami1918 via Agrobacterium-mediated leaf disk method. Putative transgenic plants were selected on Murashige and Skoog (MS) agar medium containing 50 mg l −1 kanamycin.
Four-week-old tomato seedlings (WT and transgenic lines) were transferred to pots filled with perlite/vermiculite (1:3, v/v) and grown for 2 weeks. The expression levels of mR1918 and target genes were examined by qRT-PCR. The plants (WT and transgenic) were each infected with P. infestans by inoculating with a suspension of P. infestans spores. After 4 days, the leaves were collected for RNA extraction to examine the expression levels of pi-amiR1918 and target genes. After 7 days, the relative lesion size, sporangia per leaf, and detection of dead cells were determined by ImageJ, direct counting and trypan Blue staining 38 , respectively.

RNA isolation and reverse transcription.
Total RNA was extracted from tomato using RNAiso Plus (TaKaRa) and the corresponding cDNAs were synthesized with SYBR ® PrimeScriptTM miRNA RT-PCR Kit (TaKaRa) and PrimeScriptTM RT Master Mix (TaKaRa) according to the manufacturers' instructions.
qRT-PCR analysis. Quantitative real-time PCR (qRT-PCR) was used to analyze the expression levels of miR1918 and its target genes. The specific forward primers for each miRNA gene were designed according to a previously described method 50 , and qRT-PCR was performed with the SYBR ® PrimeScriptTM miRNA RT-PCR kit (TaKaRa) according the manufacturer's protocol using the ABI 7300 Fast Real-time PCR machine (Applied Biosystems, Foster City, CA, USA). The primers for the target genes were designed following the manufacturer's instructions as described in the SYBR ® Premix Ex TaqTM II kit (TaKaRa). The presence of miRNA-mediated cleavage of target genes at a site located between the Forward and Reverse primers was ensured. Actin andα -tubulin were used as reference control gene for the miRNA and target gene qRT-PCR analysis. The sequence information of all primers is provided in Table S2. Mixed samples were used in qRT-PCR analysis, and three biologically independent replicates were analyzed for each sample.
5′-Rapid amplification of cDNA ending (5′-RACE) analysis. 5′ -RACE was used to determine the cleavage site of miR1918 on target genes in tomato. The GeneRacer TM (Invitrogen) kit was used to perform the 5′ -RACE.according the manufacturer's protocol. Total RNA was extracted from overexpression miR1918 tomato (Line 2), 5′ RNA adapter was ligated using T4 RNA ligase and the ligated mRNA was reversed using Oligo dT. The sequence information was provided in Table S2.
Statistical analysis. All statistical analysis of data was performed with SPSS, and all data were expressed as the means ± SEs from three independent experiments. The Duncan's multiple range test (p < 0.05) was selected where was appropriate.