Expression and promoter analysis of Glycine max nodule autoregulation receptor kinase gene

Legume-rhizobia symbioses contribute at least 20% of the biosphere's supply of reactive nitrogen. These unique associations rely on the exchange of specific molecular signals between nitrogen-fixing soil bacteria, collectively called rhizobia, and their host plants and, with few exceptions, result in the formation of root nodules, which provide an environment suitable for nitrogen fixation. However, nitrogen fixation is energetically expensive and nodule proliferation, in much the same manner as the proliferation of other meristems in plants, must be controlled in order to attain equilibrium between cell proliferation and differentiation. Nodule proliferation is controlled primarily and systemically by autoregulation of nodulation (AON). A gene central to this process was first isolated by map-based cloning from soybean (Glycine max) and was named G. max Nodule Autoregulation Receptor Kinase (GmNARK) in accordance with its biochemical and physiological functions. Expression patterns of GmNARK have been described by several investigators; however, these reports were based on either non-quantitative methods or a limited number of tissue types. More importantly, the expression domains of GmNARK were completely unknown. The study described in this thesis utilised techniques such as quantitative RT-PCR (QRTPCR), transcription start site mapping, promoter-reporter gene fusion, and promoter deletion, to analyse the expression levels and domains of GmNARK across a variety of tissues as well as identify the promoter elements that are responsible for the basal and tissue-specific expression of GmNARK. In addition, the promoter activity of GmNARK was also compared with that of Lotus japonicus HAR1, the GmNARK orthologue, in both homologous and heterologous transformation systems. Based on QRT-PCR, GmNARK was expressed to varying levels throughout the plant; the transcript was detected at high levels in mature leaves and roots but to a lesser extent in young leaves, shoot tips and nodules. The transcript level was not significantly affected by Bradyrhizobium japonicum during the first week following inoculation. Histochemical analysis of L. japonicus plants carrying either a 1.7 kb GmNARK promoter or 2.0 kb LjHAR1 promoter fused to a beta-glucuronidase reporter gene localised GUS activity to living cells within vascular bundles, especially phloem cells in leaves, stems, roots, and nodules. Phloem-specific expression was also detected in soybean hairy roots carrying these constructs. These results suggested that both cis- and trans-acting elements required for the transcriptional regulation of these orthologous genes are likely to be conserved. In contrast, 1.7 kb of the GmNARK promoter did not drive phloem-specific expression in Arabidopsis thaliana, indicating the absence of the trans-acting elements required for the tissue-specificity of GmNARK in this distantly related species. The comparison of 2.0 kb of promoter sequences of GmNARK, LjHAR1 and Medicago truncatula SUNN, another GmNARK orthologue, using bioinformatics and computational approaches indicated several highly conserved motifs including a putative negative regulatory region (NRR), which was previously reported to repress gene expression in non-phloem cell types. Deletion analysis of the GmNARK promoter, however, ruled out the possibility that this motif, found at -308 bp with respect to the translation start site, was truly functional and located the region controlling phloem-specific expression to DNA sequence between 908 bp and 1.7 kb upstream of the start codon. Two other candidate regions were identified by Multiple EM for Motif Elicitation (MEME). These regions, namely MEME3 and MEME4 showed strong sequence similarity to the corresponding regions of the LjHAR1 promoter. Interestingly, the MEME3 motif was also found in the MtSUNN promoter at a similar location to that of LjHAR1. Potential NRRs in the LjHAR1 and MtSUNN promoters were found in the MEME3 motifs, whereas only a variant form of a NRR in the GmNARK promoter was found in this region. Additionally, an identical semi-palindromic sequence was also observed in the MEME3 motifs of the three orthologous promoters. Based on these findings, the semi-palindromic sequence and the variant form of the NRR are proposed to be positive and negative regulatory elements for the phloem-specific expression of GmNARK, respectively. The computational approaches also identified two potential TATA elements in the GmNARK promoter. Rapid amplification of 5' cDNA ends and promoter deletion analysis have confirmed that they were functional. The two TATA elements in GmNARK promoter appeared to cooperatively direct transcription of GmNARK, but either was adequate for basal transcription. The finding that the expression of AON receptor-like kinase genes is phloem-specific has contributed to a better understanding of AON signalling pathways. The QRT-PCR study and the discovery of cis-acting regulatory regions have also provided crucial information on the transcriptional regulation of GmNARK as well as plant genes in general. Additionally, the promoters of GmNARK and LjHAR1 could potentially be used to drive phloem-specific expression in legume biotechnology research