CLE (CLV3/ESR-like) genes share a putative N-terminal signal peptide and a conserved C-terminal 14-amino-acid CLE motif, from which, recent studies suggest, a new family of 12-amino-acid plant peptide hormones is derived. Using computational approaches, for the first time, the 3D structures of 12-amino-acid CLV3 peptide and CLV1 and CLV2 receptors, and their interacting models were predicted in this study. The putative interacting partners for some CLE peptides were also inferred using statistic machine learning methods. The results from over-expression of some CLE genes suggest that CLE proteins function in a tissue-specific manner. This leads to the question of which region(s) of the CLE sequence determine its functional tissue specificities in planta. Using domain swapping and deletion my results suggest that the CLE motif likely determines much of the functional tissue-specificity of the proteins. However, the results also support a view that sequences outside the CLE motif contribute to CLE function, and show that the CLE signal peptides are required for the entrance of CLE proteins into the secretory pathway and for their functionalities in vivo. But what role, if any, does the variable domain play? Due to probable proteolytic separation, use of a gene-tag system such as GFP may only indicate the sub-cellular localizations of CLE proproteins. Nevertheless, the cleaved tags may indicate where the processing of CLE peptides takes place. Using GFP-tagged and domain swapping, my results show that neither the CLE motif nor the CLE signal peptide controls the localization of the GFP signals of the CLE proteins. Rather, the results show that the variable region determines the localization, implying that the variable region influences where the CLE-GFP is processed. These results further imply that the CLE motif itself may determine its functional tissue-specificity by dictating the direct interaction of each CLE peptide with its optimal receptor(s), whereas the receptor(s) may be available in a tissue-specific manner. On the other hand, the sequences outside the CLE motif may influence CLE function by affecting the processing of CLE peptides, resulting in a change in the availability of CLE peptides in specific tissues and/or cells.

Subsequently, my studies on the CLE family focused on CLE14 and CLE20, because over-expression of CLE14/CLE20 caused strong short-root phenotypes. Using CLE promoter and GFP translational fusions, CLE14 and CLE20 were found to be expressed in highly specific domains. Both CLE14 and CLE20 peptides inhibit irreversibly root growth by reducing cell division rates in the RAM. Intriguingly, it was found that exogenous application of cytokinin, but not auxin was able to partially rescue the short-root phenotype induced by over-expression of CLE14/CLE20 in planta, but there was no rescue by in vitro application of the synthesized CLE peptide, suggesting a requirement for a condition provided only by the living plants. The results imply that cytokinin may influence the CLE functions by affecting the processing of CLE peptides in vivo, resulting in a change in the availability and/or abundance of CLE peptides.

Gain-of-function analysis provides important insights into understanding the function of different CLE sequence regions. However, the gain-of-function analysis cannot truly reveal the in vivo function of the CLE peptides, as the activity of a given CLE member depends on its availability, which is tightly restricted in a spatial and temporal manner. To explore the in vivo functional nature of CLE14 and CLE20, efforts were made to isolate loss-of-function mutants for CLE14/CLE20. Unfortunately, this effort by screening the T-DNA insertion lines, and using the RNAi silencing approach did not result in the identification of any loss-of-function mutants for CLE14 and CLE20. This result may be explained by the fact that CLE genes have a small coding region with a low expression level, and thus may be less likely to be the targets for T-DNA insertion or for the RNAi silencing mechanism.

To date, except for CLV1, CLV2, and the recently identified CORYNE receptor kinase, no other signal receptor has been associated with any other CLE peptide ligands. Furthermore, except for the signal perception mediated by CLV1-3 complex at the cell surface, other cellular components involved in CLAVATA-CLE signaling transduction pathways remain largely unknown. Thus identification of novel corresponding receptors for the distinct CLE peptide ligands, and/or other interacting partners or downstream components will significantly advance our understanding of CLE signal transduction and stem cell homeostasis in plants. To identify such components I screened for the suppressors of the short-root phenotype caused by over-expressing CLE14/CLE20 using EMS mutagenesis and attempted to isolate the corresponding gene(s) using the Illumina sequencing approach. Seven such suppressor lines showing resistance to the application of the CLE peptides were isolated, and 35 candidate loci were obtained from Illumina sequencing of one suppressor (M-59). Further identification and confirmation of the mutation loci are underway.

In addition, I identified in Arabidopsis a CLE-like putative peptide family, which shares an overall similar domain structure with the CLE family, e.g., a N-terminal putative signal peptide, a variable region in the middle, and a conserved 13-amino-acid C-terminal CLE-like motif. Exogenous application of a synthetic 12-/13-amino-acid peptide derived from the CLE-like motif of this family caused wavy roots. Similar phenotypes were observed when the full-length sequences of two CLE-like genes are over-expressed individually in the Arabidopsis plants. The finding of a CLE-like family in Arabidopsis opens a new avenue for studies of cell signaling, root growth and development.

Finally, I isolated a loss-of-function phenotype for a membrane-associated thioredoxin (Trx h9) characterized by dwarf plants with shortened root meristems and small yellowish leaves. Trx h9 was found to be able to move from cell to cell. Mutagenesis analysis demonstrated that glycine at position 2 (Gly2) was required for its membrane binding, possibly via myristoylation. Both Gly2 and a cysteine in position 4 (Cys4) were needed for the movement, the latter seemingly for protein structure and palmitoylation. These results extend the known boundaries of thioredoxin and suggest a role in cell-to-cell communication in plant growth and specifically in root meristem development.

The results from this study provide new evidences and insights towards understanding the molecular basis of the cell signaling in plant meristem activity and development in Arabidopsis.