Supervisors info:
Aντώνιος Κολοκούρης Αναπλ. Καθηγητής (Επιβλέπων), Ζωή Κούρνια Ερευνήτρια Δ΄, Εμμανουήλ Μικρός Καθηγητής
Summary:
Influenza A/M2 protein forms a tetramer and its transmembrane domain (TM)
comprises the pore of a proton channel activated by low pH in the viral
endosome. The pore activation triggers releasing of the the viral RNA into the
cell cytoplasma, which then leads to the viral proliferation and pathogenesis.
It has been shown that aminoadamantane derivatives (Aamt), such as amantadine
(Am) and rimantadine (Rim), have been reported to block the M2 membrane protein
of influenza A virus A/M2TM, but their use has been discontinued due to
reported resistance in humans. Understanding the mechanism of action of
aminoadamantane derivatives could help the development of novel potent
inhibitors which can overcome A/M2TM resistance.
Two different M2TM peptide sequences were synthesized through Solid Phase
Peptide Synthesis (SPPS), that is, Udorn (WT) and Weybridge (V28I, L38F), and
they were then characterized. Different samples including aminoadamantane
ligands and M2TM tetramers in dipalmytoylphosphatidylcholine (DPC) micelles at
pH 8 have been prepared and used in affinity constant measurements with
Isothermal Titration Calorimetry (ITC) for both sequences. The ITC measurements
confirm the literature known binding constants of amantadine and rimantadine to
A/M2 protein.
Molecular Dynamics (MD) simulations were performed for the complex M2TM-Am in
high pH for Udorn (experimental structure from protein data bank: 2ΚQT.pdb) and
Weybridge sequences (no available experimental structure). The obtained
structures after the end of MD simulations were used to Molecular Docking
calculations. The binding constants measured against A/M2TM tetramers at pH 8
through ITC were used to evaluate and optimize the calculations results. The
energy scores of ChemScore function for Udorn sequence had a correlation of R2
= 0.50 with experimental values. However, all the other scoring functions
failed to give a good correlation with the experimental data. Due to the the
empirical nature of these functions more sophisticated methods were used based
on statistical mechanics, such as Free Energy Perturbation.
Thus, Free Energy Perturbation calculations coupled with Molecular Dynamics
simulations (FEP/MD) were applied for the M2TM-Aamt complex embedded in a
1,2-dipalmytoyl-sn-glycero-3-phosphocholine (DPPC) lipid bilayer in order to
establish a reliable protocol able to predict the relevant free energy of
several Aamt derivatives inside the A/M2TM pore. FEP calculates the binding
affinity to the protein between two similar chemical compounds with an accuracy
of 1 kcal/mol. This method was used because it has been shown that the full
protein flexibility in a hydrated DPPC lipid membrane is crucial for FEP/MD to
be a successful methodology for predicting relative binding affinities of a
series of aminoadamantane ligands bound inside the hydrated pore of A/M2TM and
thus for a future optimization of new ligands. Nine and seven alchemical
transformations were calculated for the Udorn and Weybridge sequences, which
binding affinities were measured with ITC. The correlation between the
experimental data and the calculated values was R2 = 0.85 and 0.53
respectively. Thus, the results of the calculations are quite satisfactory
considering that FEP/MD drug optimization is usually applied in other protein
binding pockets, in which the binding site of the proton is better defined and
diffused, in contract to the fully hydrated M2TM which represents a more
challenging system.
Finally, the desolvation penalty should be considered for drug design purposes
in membrane proteins. The results indicate that a reduced desolvation penalty
can facilitate the entrance of the ligands into the binding site crevice
increasing binding affinity. For example, the additional CHCH3 attached to
adamantane from amantadine to rimantadine can reduce the desolvation cost
leading to the higher binding affinity of rimantadine relative to amantadine,
as it has been shown by their experimental binding affinity constants and
confirmed by the FEP calculations. However, increasing lipophilicity can also
reduce enthaplic bonding interactions inside the pore resulting in the lower
affinity of 2-alkyl-2-adamantanamines relative to 2-adamantanamine. A
compromise between the two effects should be aimed.
Keywords:
Free energy calculations, Perturbation, Virus, Docking calculations, Isothermal titration calorimetry
