Under conditions of osmotic stress such as drought and salinity, many plants accumulate compatible organic solutes such as glycine betaine (Rhodes and Hanson, 1993). The primary metabolite choline is a precursor for glycine betaine synthesis in addition to being a component of phospholipids. In spinach leaves, choline synthesis involves three sequential N-methylations of phosphoethanolamine (PEA) in order to generate phosphocholine (PCho) via the pathway PEA -+ phosphomethylethanolamine (PMEA) -+ phosphodimethylethanolamine (PDEA) -+ PCho. The S-adenosyl-L-methionine (SAM) dependent N-methyltransferase phosphomethylethanolamine N-methyltransferase (PMEAMeT) can catalyze two of the three sequential steps: PMEA -+ PDEA -+ PCho (Dhadialla, 1999). This thesis describes a seven-step strategy for PMEAMeT purification from spinach leaves and provides evidence for the existence oftwo distinct enzymes with apparently overlapping capacities to use both PMEA and PCho as substrates. A seven step purification strategy was used in this project which included the initial four-step strategy used by Dhadialla ( 1999) to partially purify PMEAMeT approximately 70-fold. The seven steps included precipitation of soluble leaf protein from spinach leaves by 1.8-2.6M (NH^4)^2SO^4 fractionation followed by open column chromatography on DEAE Sepharose CL-6B, Phenyl Sepharose CL-4B, Macro-Prep® High Q, and Sephacryl S-100, then high performance chromatography on Mono QHR 5/5 and Protein Pak SW-300. The highest fold purifieation achieved for PMEAMeT was 7,507-fold and yielded a specific activity of 3,243 nmol min^-1 mg^-1 protein. SDS-PAGE analysis of this sample and silver staining of the polyacrylamide gel revealed that approximately 15 polypeptide bands are present in this sample and included two polypeptides with estimated molecular masses of 30 and 50 kDa. Anion exchange chromatography on a Mono Q matrix followed by photoaffinity cross-linking with aliquots of individual fractions shown to have high PMEAMeT activity, shows that the 30 and 50 kDa photoaffinity cross-linked species can be partially resolved from each other by this matrix. This observation is best explained if the two polypeptides are not subunits of the same methyltransferase enzyme. Evidence that both the 30 and 50 kDa [^3H]SAM-binding polypeptides contribute to PMEAMeT activity was provided by showing that the inclusion of either PMEA or PDEA in the photoaffinity cross-linking assay prevented the binding of [3H]:;AM to either polypeptide; a result consistent with PMEA and PDEA serving as substrates for the enzyme(s) associated with both polypeptides. In addition, the presence of PEA in the photoaffinity cross-linking assay did not prevent the binding of [^3H]SAM to either polypeptide showing that in these species cross-linking is not prevented by phosphobases that are not suitable substrates for PMEAMeT. This report describes for the first time the existence of two enzymes in spinach leaves that possess PMEAMeT activity. The role of these enzyme(s) in spinach might involve the maintenance of low PMEA and PDEA pool sizes. Low pools for these metabolites may be required to prevent the incorporation of PMEA and PDEA into the polar head groups of phospholipids in place of PCho. Whether the substitution of PMEA or PDEA for PCho is deleterious to phospholipid structure or function in plant membranes is unknown. / Thesis / Master of Science (MSc)

<p>In plants, S-adenosyl-L-methionine-dependent phospho-base <em>N</em>-methyl transferases catalyze the three sequential methylations of phosphoethanolamine to phosphocholine, the precursor for choline and the major membrane phospholipid phosphatidylcholine. The enzyme phosphoethanolamine <em>N</em>-methyltransferase (PEAMT) catalyzes the first and committing step in choline synthesis, a step for which no known by-pass has been found. In <em>Arabidopsis thaliana</em> there are two loci annotated as encoding PEAMT and a putative PEAMT, At3g18000 (<em>NMT1</em>) and<em> </em>At1g73600 (<em>NMT3</em>), respectively. A related gene product that catalyzes the last two methylations is encoded by locus At1g48600 (<em>NMT2</em>). The objective of this study was to investigate the role of <em>NMT3 </em>in <em>Arabidopsis</em>. Three SALK lines carrying independent T-DNA insertions in At1g73600 were used: SALK_062703, SALK_016929c and SALK_120703c.</p> <p>Genomic DNA was used for PCR and sequence analysis of the products established the insertion of T-DNA in the protein coding region of At1g73600 for all three lines. Gene expression was analyzed by q-PCR. Primer design was particularly important for <em>NMT3 </em>transcript quantification by q-PCR. In SALK_062703 <em>nmt3 </em>mutants, the T-DNA is in exon 8 and in the SALK_120703c line it is in intron 6. In both cases, no <em>NMT3 </em>transcripts were detected using primers that annealed to sites 3’ to the position of the T-DNA in the gene. However, low levels of transcripts were detected using primers that annealed at positions 5’ to the site of T-DNA insertion. In the SALK_016929c line the position of the T-DNA insertion was in exon 2 and primers annealing near the site of the T-DNA insertion showed no <em>NMT3 </em>expression in this mutant but amplifying the mid portion of the gene showed WT levels of <em>NMT3 </em>transcripts. Thus all the mutants produce truncated <em>NMT3 </em>transcripts and by avoiding areas that overlap truncated transcript regions we could differentiate between <em>NMT3</em> knock-out or knock-down expression.</p> <p>Wild-type (<em>NMT3</em>) and <em>nmt3 </em>seedlings from the three lines grown on defined media plates showed no difference with respect to primary root length, number or density of lateral roots, and total root length. Exposing seedlings to salt (50 or 75 mM NaCl) led to reductions in root growth but again, roots of wild-type plants were indistinguishable from the mutant seedlings. One anomaly relates to the <em>nmt3</em> SALK_120703c<em> </em>line which showed two root phenotypes. On saline media most of the seedlings had longer roots that resembled the wild-type and other mutant lines and about a third had shortened roots. Whether the seedlings had long or short roots on salt, they all lacked <em>NMT3 </em>transcripts. This line is likely carrying another insertion that yields a salt-sensitive root phenotype. Mutant plants at four-weeks looked like wild-type plants and time of flowering was not reproducibly delayed or accelerated in mutant plants relative to wild-type.</p> <p>In wild-type seedlings the relative expression level of the three <em>NMT </em>genes is similar at day or night with transcript abundance ranked in the order <em>NMT3</em> > <em>NMT2 </em>> <em>NMT1. nmt3 </em>seedlings harvested midday showed no detectable <em>NMT3</em> expression but the abundance of <em>NMT1 </em>transcripts was 6.2-fold and 1.7-fold higher relative to wild-type in shoots and roots, respectively. At night, <em>NMT1 </em>expression in shoots of<em> nmt3 </em>seedlings decreased 4.8-fold relative to the level of <em>NMT1 </em>expression at midday while transcripts detected in roots increased slightly (1.3-fold). Using SALK_036291 <em>nmt1 </em>seedlings we found that <em>NMT3 </em>expression in shoots and roots was modestly up-regulated in the absence of <em>NMT1 </em>expression and the expression of <em>NMT3 </em>is lower at night than during the day. Also, regardless of the genotype or time of day, <em>NMT2 </em>expression remained constant even when <em>NMT1 </em>and <em>NMT3 </em>transcripts underwent several-fold changes in abundance. Interestingly, four-week old <em>nmt3 </em>plants of the SALK_062703 line showed that <em>NMT3 </em>expression is knocked-out in leaves but only knocked-down in roots.</p> <p><em> NMT3 </em>was the most highly expressed of the three <em>NMT </em>genes monitored by q-PCR. Nonetheless, three independent T-DNA insertion lines defective for <em>NMT3</em> expression were wild-type by appearance and as such, offer compelling evidence that NMT3 is not required by <em>Arabidopsis. </em>The increased expression of <em>NMT1 </em>in <em>nmt3 </em>plants and <em>NMT3</em> in <em>nmt1 </em>plants strongly suggests that plants compensate for the loss of one gene by up-regulating, to varying extents, the expression of the remaining <em>NMT </em>gene. If this is the case, a reasonable prediction made for a cross between <em>nmt1 </em>and <em>nmt3 </em>plants is that it would be lethal unless plants have yet another way to circumvent the loss of an essential enzyme for this committing metabolic bottleneck in choline synthesis.</p> / Master of Science (MSc)