Topology and evolution of gene regulatory networks modulating aroma synthesis in plants

Plants are the source of a large diversity of compounds that aid in their interaction with the environment. These compounds vary among species and are the product of specialized (secondary) metabolic pathways derived from primary metabolism. Among these, the isoprenoid pathway is one of the major biosynthetic processes of specialized metabolism and it results in the production of aromas (e.g., terpenoids or apocarotenoids), pigments (i.e., carotenoids) or hormones among many other types of compounds. Terpene synthases (TPSs), responsible for the synthesis of sesqui-/mono-/diterpenes, are present in most plants and have been largely investigated for their function and substrate specificity. They contribute to the array of volatile organic compounds (VOCs) with several roles in animal attraction, herbivore repellence, antioxidant processes, and even antipathogenic properties. Despite the terpenoid sub-pathway (or branch) is quite conserved in the plant kingdom, end-products differ between species, and this may be the result of differences in both structural gene composition and gene expression regulation. In contrast to the extensive studies on their enzymatic activity, the transcriptional regulation of TPS genes has been less explored. Nevertheless, a few transcription factor families have been implied as associated to the regulation of terpenoid biosynthesis, including members of the R2R3-MYB family. Here, we focused in functionally characterizing the regulatory roles of subgroup 19 genes belonging to this family. Previous studies have shown that S19-MYBs are regulators of flower maturation, being VOCs emission a particularly conserved event in late developmental stages of flowers. In particular, we explored the capacity of S19-MYBs to bind and regulate terpenoid biosynthesis genes, using grapevine (Vitis vinifera), tomato (Solanum lycopersicum), and Cannabis sativa as study cases. We then evaluated the extent of conservation of S19-MYBs in plants to understand if TPS regulation by S19-MYBs was only present in Angiosperms. The analysis in grapevine begun by predicting in which tissues the only S19-MYB member, VviMYB24, was exerting its role. We analyzed public transcriptomic data generated from different tissues at different developmental stages. The expression profile showed that VviMYB24 was highly expressed in flowers, es expected, but also showed high expression in berry, especially at the late-ripening stages. The aggregated gene co-expression network (AggGCNs) tool, established in our research group, was used to inspect VviMYB24‘s most related processes in flower and fruit organs. The terpene synthase (TPS) term was in fact significantly enriched among VviMYB24 co-expressed genes. Its co-expression with a large fraction of the TPS family was further explored throughout berry development of cultivars (cv.) ‘Pinot Noir’ and ‘Cabernet Sauvignon’ in two consecutive seasons. Many TPSs, including TPS04, TPS10, TPS11, TPS26, TPS35, TPS76, and TPS81, appeared highly co-expressed with VviMYB24. We further explored VviMYB24 bound genes (i.e., potential direct target genes) by performing DNA affinity purification sequencing (DAP-seq). VviMYB24 binds to 22 TPSs in the cv. ‘Cabernet Sauvignon’ genome, and also to other terpenoid-related (e.g., belonging to the carotenoid pathway) and photosynthesis/light-related genes. The promoter activity of some of these bound genes, including TPS09, TPS13, TPS35, CRTISO2, and HYH was shown to be enhanced through the interaction of VviMYB24 with VviMYC2, as confirmed by dual-luciferase assays (DLA). To examine the expression of targets and quantifying terpene accumulation, we also took advantage of a rare ‘Bequignol Noir’ variegated berry, where VviMYB24 was more highly expressed in white skin sections. All DLA-validated targets were upregulated, in white skin sections, in correlation with higher amounts of terpenes in these sections. We thus demonstrated that VviMYB24 positively and directly regulates terpene synthases, promoting terpene accumulation in the grape berry. In tomato, two S19-MYB homologs were found, SlMYB21 and SlMYB24. While SlMYB21 had been previously associated with ovule development, SlMYB24 had not been characterized, and its exon-intron organization was actually wrongly annotated in the genome. These two S19-SlMYBs were functionally validated to test whether transcriptional regulation of terpenoid biosynthesis was conserved also in this species. Their expression was studied by exploring transcriptomic data of cv. ‘Heinz’ and ‘Micro-Tom’, both cultivars considered as reference tomato models. Two public platforms (Tomato eFP Browser and TomExpress) were used, and samples across different tissues and different developmental stages were inspected, showing high expression in flowers as expected. To determine the regulatory targets of S19-SlMYBs, DAP-seq assays was again conducted for SlMYB21 and SlMYB24 using genomic DNA of cv. ‘Micro-Tom’. A similar DNA-binding motif is found for both SlMYBs, which shared 82.3% of their bound genes. Among bound genes we found 11 and 15 TPS genes, out of 34 (functional TPSs), as being bound by SlMYB21 and SlMYB24, respectively. However, TPS expression profiles showed similarity of these two transcription factors (TFs) to just a few TPSs, such as TPS7 and TPS17. We further generated slmyb24 knock-out plants with CRISPR-Cas9 and collected flowers from two independent lines at two developmental stages. The transcriptomic analysis of WT and edited flowers showed some carotenoid- and pathogen defense-related genes as being down-regulated in slmyb24 mutants at day 8 after bud emergence. In fact, analysis of public RNA-seq data showed that SlMYB21 is induced by Pseudomonas syringae infection, in correlation to TPS7, TPS33, TPS46, and TPS51 upregulation. This SlMYB-TPS association may be stimulus-dependent and not much affected during flower development. Cannabis sativa floral trichomes largely accumulate VOCs mostly corresponding to mono- and sesquiterpenes. We examined whether CsMYB24, the Cannabis S19-MYB gene member, was related to terpene production in this species. Two CsMYB24 variants (CsMYB24.1 and CsMYB24.2) were identified by gene isolation from cv. ‘Purple Kush’ flowers, where they differ by the presence of the S19 motif (S19M). BLASTn search in different Cannabis public genomes revealed that CsMYB24 was duplicated in the cv. ‘Pineapple Banana Bubba Kush’, ‘Jamaican Lion DASH’, in addition to ‘Purple Kush’, while it was present in only one copy in cv. ‘Finola’, ‘CBDRx-18’, ‘Cannatonic’, ‘LA Confidential’, and ‘Chemdog91’. The binding capacity of the two CsMYB24 variants was examined using leaf and flower genomic DNA of cv. ‘Purple Kush’ via DAP-seq. The two CsMYB24 variants presented a highly similar DNA-binding consensus profile, although CsMYB24.1 specifically bound to around 1500 genes in flower gDNA. CsMYB24 expression was analyzed using public transcriptomic data, encompassing a large collection of flower structures and developmental stages together with other organs where TPS are not generally expressed. This analysis enables us to find that CsMYB24 was highly and almost specifically expressed in female flowers and trichomes of the cultivar ‘Cannbio-2’. To explore the correlation of CsMYB24 and TPS expression in Cannabis sativa, 69 TPSs were identified in the ‘Purple Kush’ genome, and public transcriptomic data was processed accordingly. CsMYB24 is highly co-expressed with several TPSs, three of them (CsPK_01G0002600, CsPK_02G0023200, and CsPK_06G0013550) being bound by CsMYB24. These results offer the possibility to study terpene regulation by CsMYB24 in more detail, especially regarding the two variants that differ in the presence of the activation domain (S19M). The study of S19-MYBs as regulators of flower development and VOCs emission has mostly been assessed in eudicots. In fact, S19 MYBs have only been characterized in three monocot species and present slightly different S19Ms. BLASTp search for S19 and S20 (the closest subgroup) genes showed that S19 members from monocots were not clustered with eudicot-S19 MYBs, while S20 members are all grouped together. A new group, named S19-like, was found in monocot-S19 MYBs. Interestingly, S19 MYBs were absent in basal angiosperms and early-diverging plants as genes with the highest similarity to S19 were actually clustered in S20. On the other hand, eudicots have one or two S19 members, with some few species, such as Arabidopsis, which have three. We further compared the consensus DNA-binding motif of S19-MYBs, taken from DAP-seq experiments, using the TF motif database JASPAR. S19 and S20 members share a highly similar DNA-binding sequence, identified as an AC-element. Furthermore, to examine the degree of conservation of S19-MYBs in terpene transcription regulation, we searched for S19-MYBs binding motifs in upstream regions of de novo identified terpene synthases throughout 18 plant species (early-diverging plants, angiosperm, monocots, and eudicots). Interestingly, the AC-element shows a similar distribution in the upstream regions of terpene synthase genes. Thus, the transcriptional regulation of R2R3-MYBs on terpene synthases through AC-element targeting appears to be conserved in plants, although in early-diverging plants this network is probably governed by S20 members.