In Bolgheri, regardless the developmental stage there were many clusters with a very high abundance level in Cabernet Sauvignon . In Montalcino and even more in Riccione we also observed differences between the expressions of clusters in the two cultivars, with ripened and green berries showing an almost opposite profile in terms of number of clusters more expressed in Cabernet Sauvignon or Sangiovese. When the berries were green, in Montalcino Cabernet Sauvignon shows the highest number of up-regulated clusters, while in Riccione, Sangiovese has the highest number of up-regulated clusters. The opposite behavior was noticed in ripened berries, with Sangiovese having the highest number of up-regulated clusters in Montalcino and Cabernet Sauvignon in Riccione . Notably, we observed a small percentage of regulated clusters exhibiting at least a 10-fold higher abundance of small RNA in Cabernet Sauvignon or Sangiovese when compared to each other . An examination of those clusters showed that a substantial difference could exist between the cultivars, depending on the vineyard and the developmental stage. For example, in Riccione, a cluster matching a locus encoding a BURP domain-containing protein showed a fold change of 390 when comparing green berries of Sangiovese with Cabernet Sauvignon. The small RNAs mapping in this region were mainly 21-nt and produced from both strands . Similarly, hydroponic channel the majority of the highly differentially expressed clusters showed a similar profile: strong bias toward 21-nt sRNAs and a low strand bias.
These findings suggest that these small RNAs might be the product of RDR polymerase activity rather than degradation products of mRNAs.We applied a pipeline adapted from Jeong et al. and Zhai et al. to identify annotated vvi-miRNAs, their variants, novel species-specific candidates and, when possible, the complementary 3p or 5p sequences. Starting from 25,437,525 distinct sequences from all the 48 libraries, the first filter of the pipeline removed sequences matching t/r/sn/snoRNAs as well as those that did not meet the threshold of 30 TP4M in at least one library or, conversely, that mapped in more than 20 loci of the grapevine genome . Only sequences 18–26-nt in length were retained. Overall, 27,332 sequences, including 56 known vvi-miRNAs, passed through this first filter and were subsequently analyzed by a modified version of miREAP as described by Jeong et al. . miREAP identified 1819 miRNA precursors producing 1108 unique miRNA candidates, including 47 known vvi-miRNA. Next, the sequences were submitted to the third filter to evaluate the single-strand and abundance bias retrieving only one or two most abundant miRNA sequence for each precursor previously identified. A total of 150 unique miRNA corresponding to 209 precursors were identified as candidate miRNAs. Among these 209 candidate precursors, 61 belonged to 31 known vvi-miRNA that passed allthe filters and 148 were identified as putatively novel miRNA candidates. To certify that they were novel candidates rather than variants of known vvi-miRNAs we compared their sequences and coordinates with the miRNAs registered in miRBase . In order to reduce false positives and the selection of siRNA-like miRNAs, we considered only 20, 21, and 22 nt candidates whose stemloop structures were manually evaluated .
Eventually, 26 miRNAs homologous to other plant species were identified with high confidence. Twenty-two were new members of nine known V. vinifera families, whereas the other four belong to three families not yet described in grapevine . For 16 homologs we were able to retrieve also the complementary sequence. Finally, excluding these 26 miRNAs and other si-RNA like miRNAs, we identified 7 completely novel bona fide miRNAs. Apart from the 61 known vvi-miRNAs identified by the pipeline, we searched the dataset for others known vvi-miRNAs eliminated throughout the pipeline, looking for isomiRs that were actually more abundant than the annotated sequences. Their complementary 3p or 5p sequence was also retrieved when possible. Hence 89 known vvi-miRNAs were identified in at least one of our libraries . Among the known vvi-miRNAs identified, 24 had an isomiR more abundant than the annotated sequence and 4 have the complementary sequence as the most abundant sequence mapping to their precursor. We found 16 vvi-miRNA isomiRs that were either longer or shorter than the annotated sequence, 7 vvi-miRNAs that mapped in the precursor in a position shifted with respect to the annotated ones and one miRNA that contains a nucleotide gap when compared to the annotated sequence . An extreme case of shifted position was found in vvi-miRNA169c, where the annotated sequence had only 5 TP4M when summing its individual abundance in the 48 libraries. Another sequence, shifted 16 bp as compared to its annotated position on the precursor had an abundance sum of 1921 TP4M, and was retained together with the annotated sequence, and named vvi-miRNA169c.1. For 36 of the 48 V. vinifera miRNA families deposited in miRBase we found at least one member. An in silico prediction of miRNA targets was performed for the 191 mature miRNAs here identified. Using the miRferno tool , and considering only targets predicted with high stringency, 1192 targets were predicted for 143 miRNAs, including six completely novel vvi-miRNA candidates .
Two novel candidates seem to be involved in the regulation of important secondary metabolites biosynthesis. Among the six targets predicted for grape-m1191, the TT12 gene is known to be involved in the vacuolar accumulation of proanthocyanidins in grapevine . For grape-m1355, 12 targets were predicted and all of them are involved in secondary metabolism pathways. Nine targets code a bifunctional dihydroflavonol 4-reductase that is responsible for the production of anthocyanins , catalyzing the first step in the conversion of dihydroflavonols to anthocyanins . Another targeted gene codes a phenylacetaldehyde reductase which, in tomato, was demonstrated to catalyze the last step in the synthesis of the aroma volatile 2-phenylethanol, important for the aroma and flavor . Still this same miRNA candidate was predicted to target with high confidence a cinnamoyl reductaselike protein that is part of polyphenol biosynthetic pathway . The grapem1355 candidate maps on chromosome 3, exactly on thefirst exon of its target , in a region where another two isoforms of the same gene are located . The last target of this miRNA candidate, codes a cinnamyl alcohol dehydrogenase known to be involved in the lignin biosynthesis . Other novel vvi-miRNA candidates seem to be involved in cell proliferation and in chloroplasts-related functions . Furthermore, for the new vvi-miRC482b candidate, besides the already known involvement of this miRNA family with disease resistance also predicted here, one predicted target encodes an anthocyanin 5-aromatic acyltransferase-like protein known to be involved in the biosynthesis of anthocyanin in different species . As for the conserved known vvi-miRNAs, most of the well-established examples of miR-targets, such as miR156-SPB, miR166-HD-ZIP, miR171-GRAS, miR172-AP2, confirmed in several plant species and already predicted in grapevine, were also predicted here.We studied miRNA profile of accumulation in the different samples. Using their normalized abundance , i.e., their relative cloning frequency, we set an empirical cut off value equal to at least 10 TP4M in both biological replicates to consider a miRNA as expressed in a given library. Also, a miRNA was considered specific when it was expressed in one or more libraries of a unique cultivar, hydroponic dutch buckets unique environment or unique developmental stage. According to our established cut off, 175 miRNAs were classified as expressed in at least one of our libraries . The libraries constructed from Sangiovese berries at bunch closure collected in Bolgheri showed only 24 expressed miRNAs . For all the other libraries, expressed miRNAs ranged from 76 to 148 . We found very few miRNAs specific to a given condition. The number of specific miRNAs for each cultivar, developmental stage and environment is reported in Figures 8A–C, respectively. Thirty-nine vvi-miRNAs were highly expressed in almost all libraries [21 ubiquitous plus 18 expressed in all libraries except in Bol_SG_bc ], whereas other miRNAs had different accumulation patterns. The normalized expression values of miRNAs were subjected to hierarchical clustering and represented in a heat map . To examine the relatedness among cultivars, environments and developmental stages, we generated a correlation dendrogram . The dendrogram shows, as already suggested by the heatmaps, that a fundamental dichotomy emerges between ripened and green berries. The most evident pattern of expression is observed when comparing different developmental stages, and confirm previous observation of miRNA modulation during fruit ripening . For example, some members of the vvi-miRNA156 family were highly expressed in all ripened berries, but weakly or not expressed in green berries. Differently, vvi-miR396a-3p and vvi-miR396b-3p showed the opposite profile. Similarly, vvi-miR172d, vvi-miR166b-5p, vvi-miR166f-5p, and vvi-miR396d-5p were highly expressed in green berries but weakly expressed in ripened berries and the members of the vvi-miR319 family showed a gradient of decreasing abundance from pea size to harvest.To gain statistical evidence of miRNA differential expression driven by the environment and/or genotype, we made pairwise comparisons, keeping constant the developmental stage, and evaluating the miRNA modulation among vineyards or between cultivars .
The analyses reveal that some miRNAs are differentially expressed between the two genotypes grown in the same environment, but also that a number of miRNAs are modulated by the environment. In particular the number of differentially expressed miRNAs is higher in ripened berries , while no miRNAs are differentially expressed at bunch closure stage . In details, 14 reads are differentially expressed at pea size stage, in at least one comparison, corresponding to 6 distinct miRNA families; 27 reads are modulated at 19 ◦Brix stage, corresponding to 12 miRNA families and 35 reads are differentially expressed in berries at harvest, corresponding to 12 miRNA families. It is worth noting that 4 of the 6 families modulated in the berries at pea size, are still present among the miRNAs differentially expressed in the berries sampled at 19 ◦Brix and at harvest , even though not always in the same comparisons. Some of the modulated miRNAs, both novel and known are intriguingly connected to berry development and secondary metabolism, even though most of the modulated families are still uncharacterized, or with targets not clearly involved in berry ripening and development, and deserve further studies to fully understand their biological roles.Using high throughput sequencing coupled with robust bio-informatics pipelines we analyzed small RNAs derived from the berries of Cabernet Sauvignon and Sangiovese, grown sideby-side in three vineyards, representative of different grapevine cultivation areas in Italy . The size distribution profiles of our libraries were in general consistent with previous reports in berry grapevine, where the 21-nt class was more abundant than the 24-nt class . Our analysis revealed dynamic features of the regulatory network mediated by miRNAs and other small RNAs, at the basis of genotype-environment interactions.Plants evolved a series of pathways that generate small RNAs of different sizes with dedicated functions . Although the various small RNA classes have been intensively studied, we are still far from understanding how many small RNA pathways exist, and how they are connected . Additionally, new classes of small non-coding RNAs continue to be discovered and many studies demonstrate a substantial redundancy and cross-talk between known small RNA pathways . Estimating the exact percentage of the plant genome covered by small RNA-generating loci still remains a challenge. By applying static cluster analysis, we investigated small RNA abundances across the genome, identifying 4408 small RNAs producing hotspots. We analyzed their expression in different cultivars, environments and developmental stages, highlighting that the majority of the considered small RNA producing regions was modulated in different conditions. This suggests a strong influence of small RNAs in the response to environment in grapevine berries. Only 462 small RNA-generating loci, corresponding to about 10% of the total, were expressed in all the analyzed libraries, possibly involved in essential biological pathways. Comparing the two cultivars, we observed, with few exceptions, that Cabernet Sauvignon berries have a higher number of expressed sRNA-generating loci than Sangiovese berries when collected in the same conditions . Considering the fact that small RNAs are implicated in the regulation of gene expression in several processes , the higher number of small RNAs expressed in Cabernet Sauvignon compared to Sangiovese berries may reflect a buffering effect of small RNAs influencing grapevine response to diverse growing environments. We believe that these characteristics may have contributed to the wide diffusion of Cabernet Sauvignon, allowing its wide cultivation in almost all wine producing countries. This is not the case for Sangiovese whose cultivation is more restricted. It is worth noting that Sangiovese is considered a very unsettled grapevine cultivar , showing a wide range of variability in response to year, clone and bunch exposure .