First, within Workpackage 1, analyses of DNaseI and RNA sequencing during de-etiolation of C3 Arabidopsis thaliana were combined with functional analysis of candidate regulators in planta. By combining this approach with in silico analysis, we identified the first transcription factor network (MYC and MYB transcription factors) to control expression in bundle sheath cells. We showed both partners are necessary and sufficient for expression in this cell type. Using A. thaliana, we also used this approach to define additional elements in cis, positioned in various gene features necessary for expression of other genes in bundle sheath cells. Moreover, we realised that incorporation of data from the distantly related C3 species Oryza sativa (rice) allowed us to identify genes likely to be expressed in bundle sheath cells of all plants derived from their last common ancestor. So, we developed a conserved blueprint associated with this cell type.
Second, in Workpackage 2, analogous sequencing analysis of C4 Gynandropsis gynandra has also been combined with functional analysis to test candidate regulators in cis and trans. This allowed us identify duons within C4 genes, and to provide functional evidence for a critical role of these duons in patterning gene expression to the bundle sheath. We also identified elements in promoter regions and untranslated regions of C4 genes that restrict expression to bundle sheath cells. In collaboration with international partners we developed a chromosome level genome assembly for G. gynandra, comprehensive data for transcripts and therefore gene models, as well as genome-wide data on transcription factor binding in vivo. Much of this was instrumental in our findings from Workpackage 3.
Third, in Workpackage 3, which aimed to combine insights from the first two Workpackages we made progress in two main areas. First, this analysis showed that the amplitude in the response of C4 genes to light increases in C4 compared with C3 leaves. Thus, C4 genes are already part of photosynthesis gene regulatory networks in the ancestral C3 state, but their response to these networks is amplified in the C4 state. We used functional testing to show that this increased responsiveness is associated with cis-elements known to be bound by transcription factors that respond to light. Different C4 genes have acquired distinct cis-elements, but they are all bound by light responsive transcription factors. Second, we have now identified multiple examples of components of gene regulatory networks that operate in the ancestral C3 state, and have been co-opted by additional genes in the C4 state to restrict their expression to specific cells. The MYC-MYB module that patterns gene expression to bundle sheath cells of A. thaliana (WP1) is also found in C4 genes of G. gynandra. Moreover, we also identified the sequences bound by these transcription factors in the GLDP gene from C3-C4 intermediate species. A switch from ubiquitous GLDP gene expression in leaves of C3 plants to bundle sheath expression is considered one of the early events in the evolution of photosynthesis, but the molecular events underpinning this event were not clear. We were able to therefore provide the first mechanistic insight into how this key evolutionary event took place. In addition, we have also discovered a short sequence of DNA and its cognate transcription factor family, which causes a domain switch from vascular cells of C3 A. thaliana to mesophyll cells of C4 leaves. This is the first example of a molecular mechanism underpinning a domain shift during the evolution of C4 photosynthesis.
