Using high throughput cutting edge methods like RNA-Seq, DAP-Seq, ChIP-Seq, and IP-MS, we have identified several novel targets and TAC protein partners at the genome and proteome levels. These extended the very reduced repertoire of direct targets, so far including only SP6A itself and GERMINs. These two genes had been shown to be activated on co-expression of the SP6A and the potato FDL1 bZIP factor. Furthermore, we found that multiple cytokinin pathway genes were directly activated by the TAC, showing that changes in the auxin-CK balance presumably have a pivotal role in inducing the switch of vascular cambium to a storage fate. Interestingly, SP6A had been shown to bind in addition to FDL1 (the potato homolog of FD) other group III bZIP factors, as ABL1 and AREBs. Same as FDL1, these bZIPs include a SAP motif at the C-terminal end that is required for 14.3.3s interaction, and may be implicated in a drought scape tuberization response. Through ChIP-Seq experiments with Hd3a-GFP potato lines (express the rice FT homolog which promotes tuberization), we here showed that the TAC binds the promoter of AGL8/FUL (this gene displays in all RNA-seq studies a co-regulated expression with SP6A), and also APETALA 1 (AP1), CAULIFLOWER (CAL), and SEPALLATA 4 (SEP4), consistently with these MADS-box transcription factors acting as direct TAC targets. Notably, GO term analyses showed that TAC targets were enriched in flowering- as well as cytokinin (CK)-related genes, indicating that CK signaling has a pivotal role in tuber initiation.
Furthermore, we performed DAP-Seq studies on potato genomic DNA by using MBP fusions of the FDL1a, FDL1c, and ABL1 factors expressed in E.coli. As expected, comparison of the FDL1a and ABL1 DAP-Seq peaks identified several binding sites in common with the Hd3a ChIP-Seq, but a significant number of peaks were observed to be unique to Hd3a or the FDL1/ABL1 factors. Peaks unique to Hd3a were found to be enriched in flowering and CK/ JA-related genes, rather than in the salt stress and ABA signaling GO terms as for FDL1a or ABL1. Genes differentially expressed in SP6Aox lines are actually enriched in JA pathway genes, indicating that Hd3a/SP6A activate these targets via complex formation with other regulators distinct to the bZIP factors. This is a fully novel finding that will be further investigated by the host group.
Proteomic studies of chromatin samples of the Hd3a-GFP lines immunopurified using anti-GFP magnetic beads led to the identification of several novel transcription factors that bound the Hd3a FT protein and thus might form alternative transcriptionally active complex with this tuberigen signal. These included a zinc finger factor, TFs of the indeterminate (IDD) family, FRIGIDA-like, GRAS and TCP families. Surprisingly, FDL1 or ABL1 were not included in this protein dataset and therefore parallel studies were carried out with chromatin samples from stolons, to assess whether we were able to detect such bZIPs in these samples and which are currently being analyzed by MS/MS. We cloned the potato IDD, TCP, FRIGIDA-like and GRAS proteins in Y2H vectors to carry out specific protein-protein interaction studies in yeast cells to further analyze their interaction ability with the SP6A, SP5G, BRC1b, FDL1 and ABL1 baits. Our initial Y2H experiments showed very encouraging results which indicated that IDDs, TCPs, and GRAS proteins directly bind various of these tuberization regulators. We are currently verifying these results by using complementary methods like Co-IP.
