The cyanobacterium Anabaena, which utilizes both photosynthesis and nitrogen fixation, provides an excellent model for studying the differentiation in cells. It is an ideal organism because of its simple genetic makeup. In the course of investigating the origins of differentiation within the multicellular filaments of the cyanobacterium, we have isolated the genes dmnA, dmnB, and dmnC, which encode methyltransferases (Matveyev, Young, Meng, Elhai, submitted).

Studies of the amino acid sequences of over 50 methylases have led researchers to identify three general groups: 5mC-methyltransferases (cytosine), N6mA-methyltransferases (adenine), and N4mC-methyltransferases (cytosine). All three groups share a common 3-D structure. In addition, N6mA and N4mC methyltransferases appear to be closely related in their sequences. Enzymes methylating the N6 amine of adenine have been further subdivided into three groups, a, b, and g, based on sequence similarities.

This research proposal focuses on DmnA, a DNA methyltransferase that recognizes the sequence GATC. It is highly unusual due to the fact that its amino acid sequence shows no similarity to any of the methylase groups. I am interested to determine whether or not DmnA has a similar 3-D structure to methyltransferases in any of the three groups. Because the gene coding for DmnA has only recently been isolated (Matveyev, Young, Meng, Elhai, submitted), few studies have been directed towards this enzyme.

Current studies are discovering new facts about this mysterious protein. Mutation of its gene leads to cell death. Therefore, it is concluded that DmnA serves a pivotal role in the life processes of the cyanobacterium. The protein, DmnA, methylates the adenine in the GATC sequence in the N6 position. In E. coli, the Dam protein, differing in sequence from that of DmnA, also methylates the adenine in the GATC sequence in order to allow the cell to identify newly synthesized DNA. The protein sequence of DmnA has been found to have remote similarities in sequence the following known methyltransferases: M.MunI; MT-A70, found in humans; Mouse orf, found in mice; Spo8, found in Saccharomyces; AT-T12H20 and AT-T5L19, both found in the plant Arabidopsis. Together they may constitute a new, fourth group of methyltransferases.

While DmnA has a noticeably different primary structure, or amino acid sequence, than the other noted methylases, the methylating function is similar. I propose that research be directed towards DmnA in order to determine the tertiary structure of the protein. Knowledge of the 3-D structure of DmnA will allow us to look for similarities among the structure of DmnA and other known methyltransferases. Conclusions may then be drawn to relate the folding of the protein into an active site with the methylating function. Furthermore, information gathered from this research would allow us to propose universal 3-D structures that would allow proteins to successfully methylate nucleotide bases.

Two methods prevail when determining the tertiary structure of a protein: X-ray crystallography and 2-dimensional nuclear magnetic resonance (NMR). The former is a technique which requires specially trained scientists to analyze the collected data. Crystals of the protein structure are subjected to X-rays which are diffracted through the crystalline structure. The resulting diffraction pattern can be analyzed to determine the number of atoms, the bonding pattern, and how the atoms are arranged. The later method, NMR, is the more easily accessible procedure which I propose for determining the structure of DmnA. To prepare for NMR, the molecule being tested is first dissolved in a deuterium solvent. The protein is usually contained in a small glass tube. In H (hydrogen) NMR, the nuclei of molecules in question are first subjected to a magnetic field. As a result, the internal spins of hydrogen nuclei are directed in one direction by the magnetic field. Radio waves are then used to bombard the molecule, and a spectra is produced. This spectra serves as a fingerprint to identify key structural components in molecules. It is used to learn more about the number of hydrogen atoms, the distances between hydrogen atoms, and the torsional angles between bonds. Certain functional groups exhibit characteristic NMR spectra which may then be used as references to the spectra obtained from NMRs of unknown molecules.

NMR is the ultimate procedure to determine the protein structure of DmnA and will be performed in collaboration with Suzanne O'Handley (Chemistry, University of Richmond) and Neel Scarsdale (Biochemistry and Biophysics, Virginia Commonwealth University). However, given the allotted time, I propose that the 10 weeks of summer research be used to first prepare the protein for future NMR analysis. Protein preparation includes isolating the desired protein, DmnA, from E.coli, in which its gene has been cloned. Only a pure protein sample can yield NMR spectra that can be properly interpreted. Many time-consuming steps are involved in this process. First, the already isolated dmnA gene must be subcloned into a specialized plasmid so that the protein is overproduced. An enzyme assay is next needed to quantify the amount of protein present. The protein desired must then be purified from vector protein by using such methods as affinity column chromatography. This purified protein can now be dissolved in solvent for NMR spectroscopy for further analysis. I will also concurrently study the biochemical properties of the enzyme.

Results from NMR testing will provide us with more information regarding the structure of the protein, DmnA. There are three main plausible results. In the first, the NMR might show us that while the protein differs in its amino acid sequence, it has the same 3-D structure as other known methylases. This would suggest that the known structure of DmnA methyltransferases may be achieved by different amino acid sequences, thereby allowing the enzyme to perform methylating functions identical to similarly structured methylases. In the second case, NMR might show us that an entirely different and new structure exists for DmnA. This possibility would prompt us to find a correlation between this unknown structure and its methylating capacities. Finally, it may well be that the protein is difficult to purify for NMR spectroscopy. Should this happen, we would be prompted to find a new means of purification for further structure testing.

I believe this project has great potential. Close examination of the Anabaena protein, DmnA, will provide more knowledge regarding its methylating function.

References:
Billeci, Todd. Home page.
<http://picasso.ucsf.edu/~billeci/>

Hornak, Joseph. "The Basics of NMR." © 1997-1999.
<http://www.cis.rit.edu/htbooks/nmr/inside.htm>

Matveyev AV, Young KT, Elhai J (submitted). DNA methyltransferases of the cyanobacterium Anabaena PCC 7120.

Elhai, J. (1998) NSF grant proposal. DNA Modification and Regulation over Patterned Heterocyst Differentiation.