Transcriptional networks of TCP transcription factors in Arabidopsis development

Leaves are a plant’s main organs of photosynthesis and hence the development of this organ is under strict control. The different phases of leaf development are under the control of both endogenous and exogenous influences. In this work we were interested in a particular class of transcription factors known to control leaf development: the TCP transcription factor family. Members of the TCP transcription factor family are found to be involved in transcriptional control of leaf development in different plant species, including tomato, snapdragon and the model plant Arabidopsis thaliana. The 24 TCP genes encoded in the Arabidopsis genome are divided into two classes, class I and class II TCPs. Based on the putative consensus binding sites of TCP proteins belonging to these classes, the theory has been postulated that class I and class II proteins antagonistically regulate common target genes. In general, not much is known about TCP functions, and almost all knowledge comes from analyses of class II TCP mutants, where divergent phenotypes have been characterized in single and multiple gene knockouts. Most of the time, single TCP knockouts have inconspicuous or no divergent phenotypes that can be explained by strong functional redundancy within the TCP transcription factor family. The most prominent example is the JAGGED AND WAVY (JAW) phenotype, in which a group of five TCP transcription factors that are under control of the microRNA miR319a a knocked down by overexpression of the microRNA. Different ways exist to circumvent problems with genetic redundancy. One way is to cross knockout lines for closely related homologues in order to knock-out complete functions. In the TCP family, where sequences are highly variable outside the so called TCP domain, highest sequence homology is not always a good predictor for functional redundancy. Instead, integration of expression and other functional data can help determining the level of functional redundancy between closely related genes. We could show that from the four TCPs that are closely related to TCP4 the transcription factor, TCP10 has the highest overlap in sequence, expression, and protein-protein interaction capacity. Further investigation shows a strong overlap in target genes as well, specifically covering genes that are involved in jasmonate (JA) synthesis and response, indicating common functions and verifying earlier studies on the importance of JA signaling in mediating class II TCP functioning in leaf development. Another way to circumvent genetic redundancy as a problem to analyze gene function, which is suitable for the analysis of transcription factors, is to identify the genes that are under direct transcriptional control of the transcription factor. We identified direct target genes of the class I TCP transcription factor TCP20, because no phenotypic alterations could be observed in tcp20 single knockout mutants. Among the target genes found there was a significant proportion of JA synthesis and response genes. Surprisingly, cell cycle genes, which were supposed to be under the control of TCP20, were not found in our study. Because LIPOXYGENASE2 (LOX2), which is also under the control of the class II TCP transcription factors TCP4 and TCP10, was found to be a target of TCP20 we were able to investigate and partially verify the previously theorized antagonistic control of target genes by class I and class II TCPs. Another group of genes that was found to be over-represented in the TCP20 target gene list were genes involved in iron homeostasis in both roots and leaves. The so called subgroup Ib basic helix-loop-helix transcription factors were analyzed for their role in leaf development. We could show that these transcription factors are involved in photomorphogenesis during the switch from proliferative to differentiated leaf cells, and that they are inhibited by TCP20 as a means to suppress cell differentiation during early leaf development. As TCPs were not found to control the cell cycle directly, we identified potential direct regulators in a large scale analysis of transcription factor binding to three selected cell cycle promoters. Transcription factors of different families were identified, some of which previously have not been associated to cell cycle control. Of these families, especially the MYB and NAC families of transcription factors stand out.

In sum, we found both class I and class II TCPs to be involved in hormonal control of leaf development, but also in cell wall control and iron homeostasis, increasing the number of cellular functions TCPs are probably involved in. More importantly we could not find any indications that core cell cycle genes are direct targets of these transcription factors, despite the growth effects discovered when knocking out several TCP genes. The fact that in these tcp mutants ultimately cell proliferation is affected leads to the assumption that TCP genes indirectly control the cell cycle via downstream targets. For this, especially the hormones under their control are strong candidates. We conclude that the broadly expressed members from the TCP transcription factor family are not the key regulators of growth, but act as co-factors or mediators in this biological process.