Douglas Bretzing and Professor David Busath, Physiology and Developmental Biology
The ability of influenza A virus to unpack its genome, replicate, and infect its host is contingent upon acidification of the viral interior. This is achieved by proton conductance via M2, an integral membrane protein that forms proton channels in the viral lipid envelope of influenza A virus (1). The drugs amantadine and rimantadine have previously been successful in blocking proton flux into the virus, however viral resistance to these drugs has recently reached over 90% (2). The goal of this project is to provide direct evidence as to the actual site of interaction between rimantadine and M2, thereby aiding in the development of a replacement anti-influenza drug.
The first step of this project was to use site directed mutagenesis (with a QuickChange Kit) to mutate the residues of interest to cysteine. Ian Nelson generously performed the site directed mutagenesis for two mutants, S31C and G34C, and subsequently transfected E. Coli with each plasmid. He is currently working on performing the DNA sequencing to confirm each sequence. I received training for M2 growth, harvest, and purification from November 2010 through January 2011. I then began attempting these procedures on my own. The first problem I ran into was that I was getting significantly lower yield of purified M2 than we expected to produce. I was able to more than double the initial yield of purified M2 by increasing the incubation period when growing E.Coli, decreasing the incubation period of M2 with nickel column resin (to achieve more selective binding of M2 to the resin), increasing the volume of nickel resin (to avoid saturation of the resin), and improving my nickel column elution techniques.
Following M2 purification, I ran the purified M2 sample on a Tris-Hepes-SDS gel for approximately 40 minutes at 150 V. The results of the finished gel were unexpected. The plasmid we used to transfect the E. Coli included the DNA corresponding to the mutated M2 (22-62) linked through a TEV protease recognition site to maltose binding protein (MBP). Each M2 (22-62) monomer is approximately 5kDa and each TEV protease cleavable MBP is approximately 44.1kDa. Because the M2 monomers are unable to form tetramers before TEV cleavage, we expected a single band corresponding to uncleaved M2 at approximately 49 kDa. Instead, we observed significant bands at 10kDa, 15kDa, 25kDa, and 55kDa. The 10kDa, 15kDa, and 25kDa bands could correspond to TEV-cleaved M2 dimers, trimers, and tetramers, respectively. The 55kDa band could possibly correspond to uncleaved M2 monomers. However, this explanation suggests that the MBP-fusion protein is spontaneously cleaving (without being exposed to TEV protease), and therefore requires further investigation. It will also be necessary to run the purified M2 sample on a gel under reducing conditions to see if disulfide linkages are forming between the cysteines of each monomer. In an effort to better understand how to interpret SDS gels, I am currently studying the effects of running gels with multiple purified proteins of known molecular weights under many different conditions.
In order to further investigate the possibility of the MBP-fusion protein spontaneously cleaving, I incubated my next M2 sample with TEV protease for about 22 hours at 4˚ C while M2 was still bound to 0.5ml nickel resin. Theoretically, the MBP-fusion protein should be cleaved while bound to the resin, releasing the cleaved M2. The protease has an HQ tag that binds to nickel resin with high affinity. Consequently, the column runoff should contain only the cleaved, purified M2. After collecting the column runoff, I ran the sample on a SDS gel. The only significant band was located at 48kDa, corresponding to the molecular weight of the protease. We determined that the resin probably became saturated, resulting in the presence of protease in the runoff. We attributed the minimal M2 in the runoff to the low temperature (4˚ C) at which the reaction took place, although we also considered the possibility that a fraction of the protease bound to the resin before catalyzing cleavage of the MBP-fusion protein.
With my next M2 sample, I eluted the M2 from the nickel column before incubating the sample with protease. I performed the incubation at room temperature (approximately 25˚ C) for 20 hours. Following MBP-fusion protein cleavage, I incubated the sample with 1 ml of nickel resin, collected the column runoff, and then eluted the column. I then ran uncleaved M2, nickel column runoff (cleaved M2), and nickel column elution (MBP and TEV protease) on a SDS gel. To our surprise, the column runoff and column elution lanes appeared identical to each other on the gel, with bands corresponding to each component. It is possible that the nickel resin, although increased from 0.5 ml in the previous trial to 1 ml in this trial, was still saturated, thereby accounting for the presence of protease and MBP in the runoff. The presence of cleaved M2 in the column elution fraction is a bit harder to explain, but it is possible that its presence is a result of my not washing the column prior to elution. With my next purified sample, I will increase the nickel resin volume to at least 2 ml and will wash the column prior to elution of the MBP and TEV protease. I will then run each fraction on a SDS gel as before, but I expect the column runoff to contain only cleaved M2 and the column elution to contain only MBP and TEV protease.
After having grown and harvested M2 S31C multiple times, I combined each fraction for the purpose of increasing M2 concentration. I then used ultrafiltrate centrifugation to further concentrate the combined fractions and to move the M2 from suspension in buffer to suspension in methanol. My first attempt resulted in precipitation of the M2. Subsequent attempts with different M2 samples were successful as I decreased the speed of centrifugation (from 4000 x g to 3200 x g) and added methanol before the sample could become too concentrated and precipitate. These M2 samples, now suspended in methanol, are ready to be reconstituted into liposomes and tested for functionality, as described in Sharma et al. (3). After the M2 mutants prove to be functional in liposomes, I plan to induce in situ amide bond formation, using native chemical ligation, between the rimantadine-derived drug generously being developed by Dr. Steven Castle and the M2 cysteine residue of interest. Finally, I will calculate M2 proton conductance to determine the percent blockage of the rimantadine-derived drug and will then confirm formation of the rimantadine-M2 amide bond using mass spectrometry.
Scholarly Sources
- Schnell, J. R., J. J. Chou. 2008. Structure and mechanism of the M2 proton channel of influenza A virus. Nature. 451:591-595.
- Stouffer, A., R. Acharya, D. Salom, A. S. Levine, L. Di Costanzo, C. S. Soto, V. Tereshko, V. Nanda, S. Stayrook, W. F. DeGrado. 2008. Structural basis for the function and inhibition of an influenza virus proton channel. Nature. 451:596-599.
- Sharma, M., M. Yi, H. Dong, H. Qin, E. Peterson, D. D. Busath, H. X. Zhou, T. A. Cross. 2010. Insight into the mechanism of the influenza A proton channel from a structure in a lipid bilayer. Science. 330:509-512.