Investigation of the structure and dynamics of the M2 transmembrane peptide of the influenza A virus by solid-state nuclear magnetic resonance
Solid-state nuclear magnetic resonance (SSNMR) is an important tool for the study of structure, function and dynamics of many chemical and biological systems. By using this technique we are able to study molecules that are insoluble or would not be in their native state when dissolved in solution. This is of particular importance in the study of membrane proteins, whose structures have been difficult to analyze through the use of traditional x-ray crystallography or solution NMR techniques. In this thesis, we have studied the transmembrane segment of the M2 proton channel of the influenza A virus. The M2 proton channel is essential to virus replication, as it acidifies the viral interior allowing for the uncoating and release of viral RNA. The M2 transmembrane peptide (M2TMP) is a 25-residue α-helical transmembrane segment that spontaneously forms tetramers in the lipid bilayer. We are interested in the structure and dynamics of M2TMP, and the interaction of the peptide with an anti-influenza drug, amantadine.
We determine the orientation of M2TMP relative to the bilayer normal in model phosphocholine lipid bilayers such as DLPC and POPC, with and without the amantadine drug. This is achieved through the measurement of 15 N-1H dipolar coupling and 15N chemical shift anistropy, which are directly related to tensor orientation for a uniaxially rotationally diffusing molecule. From this information we show that the orientation of M2TMP is dependent on lipid bilayer thickness. We also observe that the addition of amantadine affects the orientation of the peptide. We have measured the orientation of both singly and multiply 15N labeled M2TMP.
In order to observe the effects of drug binding on M2TMP secondary structure, we measure the 13C and 15N chemical shifts for 11 sites along the channel, including several proposed binding sites. Chemical shift is known to be very sensitive to local torsion angle perturbations. The largest drug-induced chemical shift change is observed at Serine 31, one of the proposed binding sites. The chemical shifts are used to simulate backbone torsion angles, and an amantadine-bound M2TMP structure is deposited in the Protein Data Bank with accession code 2KAD. We also establish a relationship between 13C chemical shift of sidechain methyl carbons and sidechain torsion angles. Here we show that sidechain rotamer can be predicted from 13C methyl chemical shift and direct measurement of Val χ 1 correlates with the predicted rotamer.
Finally, we have measured the 13C T2 and 1H T1ρ relaxation rates and 13C linewidths of M2TMP in DLPC bilayers with and without amantadine. Substantial 13C line narrowing occurs when amantadine is bound to the peptide, and the line-narrowing mechanism is attributed to faster peptide motion and a more homogeneous peptide conformation. We observe an increase in 13C T2 relaxation time for the amantadine-bound peptide relative to the apo state, with the most significant increases occurring near Serine 31. From 1H T1ρ measurements, we are able to calculate the relative activation energies and motional correlation times for uniaxial rotational diffusion of the apo- and amantadine bound peptides. An increase in Ea and a reduction in τ C was observed for the amantadine-bound state. We propose that the addition of amantadine to the sample results in a more tightly packed and homogeneous tetramer bundle that is able to diffuse faster in the lipid bilayer.