Biophysical Chemistry 289 (2022) 106859

Contents lists available at ScienceDirect

Biophysical Chemistry
journal homepage: www.elsevier.com/locate/biophyschem

Understanding the conformational changes in the influenza B M2 ion

channel at various protonation states
Unmesh D. Chowdhury , B.L. Bhargava *
School of Chemical Sciences, National Institute of Science Education & Research – Bhubaneswar, an OCC of Homi Bhabha National Institute, P.O.Jatni, Khurda, Odisha
752050, India

A R T I C L E I N F O A B S T R A C T

Keywords: The characterization of influenza (A/B M2) ion channels is very important as they are potential binding sites for
Ion channels the drugs. We report the all-atom molecular dynamics study of the influenza B M2 ion channel in the presence of
Molecular dynamics explicit solvent and lipid bilayers using the high resolution solid-state NMR structures. The importance of the
Influenza B M2
various protonation states of histidine in the activation of the ion channel is discussed. The conformational
changes at the closed and the open structures clearly show that the increase in tilt angle is necessary for the
activation of the ion channel. Additionally, the free energy surfaces of the eight systems show the importance of
the protonation state of the histidine residues in the activation of the influenza B M2 ion channel. The pro­
tonation of the histidine residues increases the tilt angle and the intra-helix distance which is evident from the
superimposition of the structures corresponding to the maxima and the minima in the free energy landscape. The
findings imply differences in the singly protonated and double protonated conformational states of BM2 ion
channel and provide insights to help further studies of these ion channels as the drug targets for the influenza
virus.

1. Introduction channel [6], and NaK cation selective channel [7]. The viral ion chan­
nels are a class of short membrane proteins comprising of 50–120 amino
The ion channels are membrane proteins that regulate the flow of acids [8]. They play a pivotal role in the viral replication cycle. One of
ions across the cell membrane via the water filled pores. They are hy­ the primary reasons behind the study of these viral ion-channels is that
drophilic pathways for the ions and water across the low dielectric hy­ they can be potential drug targets for the regulation of the viruses, as
drophobic lipid bilayers [1]. The ion channels are distinguishable from these ion-channels play a significant role in the viral replication [9–11].
the aqueous pores on the account of their channel selectivity and the The most important transport channels in the influenza virus A or B
gating activity. They can regulate the permeability of the permeant virus matrix 2 are the AM2 and the BM2 channels [12,13]. The transport
molecules through the selectivity filter. The ion channels are gated, of the protons and water across these channels is essential for the life
allowing the opening of the channel via some external stimulus such as cycle of the virus. The influenza virus ion channels are a class of homo-
the voltage gradient across the membrane, binding of ligand, change in tetrameric, alpha-helical proton channels that are regulated by the pH of
pH or mechanical stress [2–4]. The external stimulus causes a confor­ the medium. In both B and A M2 ion channels there is a conserved
mational change in the protein structure which results in the opening of HxxxW motif which has a similar function. The His-19 residue is
the channel gate, regulating the ion-flux across the lipid membrane. responsible for the acid activation and the proton conduction. Given the
Biophysical characterization of a series of ion channels has revealed a prevalence of foreign viruses and pathogens in human life, the virus
wealth of information related to the conformational states of the acti­ pathology and the corresponding drug designing is a pertinent research.
vated and non-activated ion channels like EmRE [5], heat activated H+ The influenza virus B is the causative pathogen for nearly half of the

Abbreviation: QM/MM, quantum mechanics/molecular mechanics; POPC, 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine; POPG, 1-Palmitoyl-2-oleoyl-sn-

glycero-3-phosphoglycerol; CHOL, cholesterol; TM region, transmembrane region; H-bonding, hydrogen bonding; His, histidine; APL, area per lipid; MD, molecular
dynamics; RMSD, root mean square deviation; RMSF, root mean square fluctuation; MSD, mean square displacement; ss-NMR, solid state NMR; VDW, Van Der Waals;
PCA, principal component analysis..
* Corresponding author.
E-mail address: [email protected] (B.L. Bhargava).

https://doi.org/10.1016/j.bpc.2022.106859
Received 25 May 2022; Received in revised form 6 July 2022; Accepted 19 July 2022
Available online 23 July 2022
0301-4622/© 2022 Elsevier B.V. All rights reserved.
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

influenza diseases which almost exclusively infects humans [14,15]. Table 1

Literature reports show that A M2 inhibitors such as amantadine (AMT) Details of the simulated systems with different histidine protonation states.
and rimantadine (RMT) are unable to inhibit BM2 proton conductance Systems Total number of atoms Box length (nm × nm × nm)
[16,17]. After a lot of debate regarding the protonation states of BM2 0
closed(His-19 ) 39,576 6.05, 6.05, 10.45
channel, the best resolved structures are being reported [18]. Using closed (His-191+) 39,528 6.03, 6.03, 10.50
these structures, we have studied the structural and dynamical proper­ open (His-193+) 40,228 6.04, 6.04, 10.67
ties of BM2 channel at eight different protonation states by the aid of all- open (His-194+) 40,224 6.08, 6.08, 10.54
atom molecular dynamics (MD). MD simulations are useful to unearth closed(His-190/His-271+) 39,397 5.99, 5.99, 10.63
closed(His-191+/His- 39,771 6.05, 6.05, 10.52
important molecular mechanisms between the lipid bilayers and the 271+)
membrane proteins [19–23]. The earlier MD simulations have explored open(His-193+/His-273+) 43,748 6.44, 6.44, 10.23
the role of BM2 in the drug resistance (amantadine) using potential of open(His-194+/His-274+) 43,766 6.09, 6.09, 11.45
mean force (PMF) calculations [24]. There have been some attempts to
find the mechanism of proton shuttling via AM2 channel using QM/MM
[25,26]. MD simulations have also enabled us to understand the role of
lipid composition in the mechanism of activation and functioning of the
BM2 channel [27]. Using the constant pH molecular dynamics (cpHMD)
simulations, the protonation states of the histidine residues that are
responsible for the channel activation have been studied [28]. In this
study, we have looked at the effect of protonation of His-19 alone or
with His-27 on the activation of ion channel. Though the earlier studies
[29] suggest that the His-27 tetrad should become protonated prior to
the His-19 tetrad, we aim to model the BM2 opening through mono-
protonated His-19 tetrad compared to the doubly protonated His-19/
His-27 tetrads. Classical modeling used in our studies is not suitable to
capture the transport mechanism of the hydronium ion [26] through the
BM2 pore.

2. Methodology and simulation details

The high resolution influenza B M2 coordinates were taken from the

published structures by Hong and coworkers [18]. The solid-state Fig. 1. The snapshots of the initial configurations of the BM2 ion channel
structures are resolved at 1.4–1.5 Å. The PDB ID of the closed state is corresponding to the pdb id – 6pvr and 6pvt in the (a) closed and (b) open
6pvr and that of the open state is 6pvt. The simulations of the closed conformation along with the lipid bilayer and water. The N and C-terminus of
state structures were performed taking the overall charge as (0) and the transmembrane (TM) region are labelled in the Figure.
(+1) for the His-19 residues, and that of the open state structures with
the His-19 protonation states of (+3) and (+4) [29]. His-19 (0) corre­ protein molecules was monitored through the C-α RMSD given in the
sponds to a state where none of the His-19 residues of the channel are supplementary information. Leap frog algorithm with a time step of 2 fs
protonated, His-19 (+1) corresponds to a state where one of the four His- was used to carry out the simulations. The particle mesh Ewald (PME)
19 residues is protonated and so on. In the closed state at most one His- method with a coulomb cutoff of 1.2 nm was used to compute the long
19 residue is protonated, whereas in the open state at least three of the range electrostatic interactions [40]. The van der Waals interaction was
four His-19 residues are protonated. Along with these four systems, four modeled using the Lennard-Jones potential with a cut-off distance of 1.0
additional systems were simulated in which the His-27 residues were nm to 1.2 nm with the force based switching function [41]. The tem­
also protonated in the pH range of the protonation of the His-19 resi­ perature was maintained at 310 K using the Nosé-Hoover thermostat
dues. Overall, we have considered these eight models in our analysis with a coupling constant of 1 ps. The pressure was maintained using the
(four mono-protonated and four doubly-protonated) to compare across Parrinello-Rahman barostat with a coupling constant of 5 ps and
the different pH states of His-19 and His-27 residues of the ion-channel. compressibility of 4.5 × 10− 5 bar− 1. The bonds involving the hydrogen
The proteins were inserted into the lipid bilayers with a composition atoms were constrained using the LINCS algorithm [42]. The images
of POPC: POPG: cholesterol at the molar ratios of 3: 1: 1 using the were rendered using VMD software [43]. The analysis scripts have been
membrane builder in CHARMM-GUI [30,31]. The choice of the lipid written in MDAnalysis python library [44] along with the gmx modules.
bilayers is made in accordance with the earlier studies [27,28,32–34]. Pore radii of the ion channels were calculated using HOLE [45]. Normal
The orientation of the channel axis is kept perpendicular to the lipid mode analysis was carried out using ProDy [46]. Further details of the
plane. The resulting number of lipids was 61 in the outer leaflet and 59 simulations are given in Table 1.
in the inner leaflet. A water bed of 15 Å was kept on each side of the The snapshots of the starting configurations of the BM2 ion channel
bilayer. The neutrality of the systems was maintained by adding chloride in closed and open conformations are shown in Fig. 1.
counter-ions. KCl concentration of 0.15 M was used to mimic the
physiological conditions. TIP3P was used to model water molecules 3. Results and discussions
[35].
All the simulations were carried out using the GROMACS simulation
engine (version 5.1.4) [36]. CHARMM36 was used to model the protein
and the lipid molecules [37–39]. The initial minimization was carried
out using the steepest descent method to negate the bad contacts. 3.1.1. Structural characteristics
Stepwise system equilibration was carried out at constant NVT condi­ The BM2 ion channel is a tetrameric alpha helix with 33 residues in
tions with restraints on the biomolecules as implemented in the the transmembrane region. The inside of the ion channel is lined by six
CHARMM-GUI. Production runs were carried out in constant NPT residues as shown in Fig. 2(a). The tetramer has a well defined hydro­
ensemble without any restraints at 310 K for 1.2 μs for each of the eight phobic core that has F5 and W23 residues at the N- and the C-terminal
systems, giving a total simulation time of 9.6 μs. The equilibration of the

2
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Fig. 2. (a) The structure of the influenza BM2 ion-channel in closed conformation showing the pore lining residues. (b) The root mean square fluctuations (RMSF) of
the C-α atoms in the eight systems with different protonation states of His-19 residue. (c) and (d) are the view of the ion channel from the N-terminus and the C-
terminus respectively, in the closed structure (PDB id – 6pvr), (e) and (f) are the views from the N-terminus and the C-terminus respectively, in the open structure
(PDB id – 6pvt). The α-helices are given in new cartoon representation and the L8 and H27 residues are shown explicitly in blue. (For interpretation of the references
to colour in this figure legend, the reader is referred to the web version of this article.)

sides of the channel respectively. The residues at the top of the channel for the transport of the water molecules. Below the lining of the serine
among the pore lining residues are the hydrophobic L8 valve followed residues are the histidine tetrads responsible for the opening of the
by the serine at resid-12 and resid-16. The L8 residue opens outward channel. The histidine box or the histidine tetrads open up with the
giving way for the change in the tilt angle of the helix bundle, eventually protonation hence opening the pore. The opening of the His-tetrads
leading to the activation of the ion channel [18]. These hydrophilic make way for the tryptophan motion. The H19 and W23 are impor­
residues that constitute the core of the α-helical tetramer, are necessary tant in the proton selectivity and gating mechanism [47,48]. The

3
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Table 2 residues are accommodated as seen in Fig. 2(c) -(d). The open channel
The tilt angle of the systems studied. The standard deviations of has a more uniform pore radius across both the N-terminus and the C-
the tilt angle are given in the parenthesis. terminus as seen in Fig. 2(e)–(f). The large W23 residues play an
Systems Tilt Angle important role in the water dynamics in their vicinity.
closed(His-190) 15∘ (1∘)
In our study, we have illustrated the importance of the different
closed (His-191+) 16∘ (2∘) charges on the His-residues. The convergence of the simulation was
open (His-193+) 18∘ (1∘) checked using the structural fluctuations determined from the Root
open (His-194+) 21∘ (1∘) Mean Square Displacement (RMSD) of the C-α atoms of the protein that
closed(His-190/His-271+) 16∘ (4∘)
is given as Supplementary Information (Fig. S1). To check the fluctua­
closed(His-191+/His-271+) 18∘ (1∘)
open(His-193+/His-273+) 19∘ (3∘) tions of the residues comprising the ion channel, the root mean square
open(His-194+/His-274+) 22∘ (2∘) fluctuations (RMSF) of the pore residues were computed. The increasing
protonation of the Histidine residues leads to the increasing fluctuations
in the N-terminal region for the (+3) state in the open structure and
orientation of the helices within the tetramer is such that the proposed decreases the fluctuations in residues 19–21 at the middle of the α-helix
pore-lining residues (L8, S12, S16, H19, W23 and H27) are all facing the bundle and towards the C-terminal region in the open structures. The C-
center of the tetrameric bundle. The bundle exhibits an approximate 4- terminal region shows higher conformational mobility with respect to
fold symmetry. The helices are packed together more tightly at the N- the N-terminus. There is also an increase in the structural fluctuations of
terminal region than at the C-terminal region where the large W23 the residues between the H19 and W23 across the structures. These

Fig. 3. (a) – (d) are the intra-helical distance profiles of the closed and the open structures. (a) and (b) are for the closed structures at the neutral and the (+1)
protonation respectively, and (c) and (d) are for the open structures at (+3) and (+4) protonation respectively. A and B are the closed and the open channel
respectively, viewed from the top of the N-termini.

4
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

residues (from H19 to W23) play an important role in the gating Table 3
mechanism of the channel opening. The overall mean fluctuations The average number of H-bonds observed between the pore lining residues and
decrease about 10% in going from singly protonated closed state His-190 water in the closed and the open structures across the protonation states for all
to the open His-194+ conformation. However, it is observed that in case the systems. The standard deviations are given in the parenthesis.
of doubly-protonated structures, it increases nearly 52% in going from His-19 closed-(0) closed-(+1) open-(+3) open-(+4)
closed to the open state. L8 0.27 (0.25) 0.15 (0.21) 3.19 (0.41) 2.75 (0.35)
S12 6.91 (0.09) 5.95 (0.86) 5.43 (1.00) 7.13 (0.82)
3.1.2. Helix packing S16 5.58 (0.08) 6.27 (0.85) 6.00 (0.95) 6.74 (0.80)
The role of the hydrophobic mismatch between the bilayers and the H19 8.11 (1.02) 6.51 (0.90) 6.31 (0.75) 4.74 (0.67)
W23 2.92 (0.52) 3.38 (0.76) 3.91 (0.72) 4.10 (0.70)
length of the transmembrane helix is well documented in earlier studies H27 9.95 (0.54) 11.41 (1.09) 11.09 (1.11) 12.04 (1.15)
on a series of ion channels [49–51]. It has been reported that the tilt
angle of the helix increases with the concomitant decrease of the bilayer
His-19/His-27
thickness [52,53]. Hong and coworkers have shown that the increase in L8 0.15 (0.24) 0.19 (0.21) 2.83 (0.49) 3.44 (0.39)
the tilt angle (angle between the principal axis of helix and the mem­ S12 5.83 (0.80) 6.28 (0.75) 7.14 (0.86) 7.15 (0.85)
brane normal) of the helix bundle and the intra-helical distance (dis­ S16 5.53 (0.76) 4.11 (0.68) 6.73 (0.81) 6.79 (0.83)
tance between the center of mass of the helices) leads to the activation of H19 7.55 (0.92) 7.59 (0.92) 6.17 (0.81) 6.02 (0.73)
W23 3.76 (0.77) 3.40 (0.74) 3.15 (0.69) 2.72 (0.75)
the BM2 ion channel [18]. We have traced the change in helix tilt in
H27 8.64 (0.84) 9.77 (1.00) 8.22 (0.93) 7.14 (0.98)
going from the closed structure to the open structure across the eight
protonation states of the His-tetrads which are given in Table 2. There is
an increase in the mean tilt angle of the helix bundle in the closed state this study, we have adapted the geometric criteria to classify the
with the neutral charge to the (+1) charge from a tilt angle of 15∘ to 16∘. hydrogen bonds [63]. In a strong hydrogen bond, the hydrogen atom
The tilt angle for the open structure also increases from 18∘ to 21∘ when and the acceptor are separated by a distance less than 2.2 Å, and the
the protonation state of the His-tetrads is changed from (+3) to (+4). For angle made by the donor, hydrogen atom, and the acceptor is within the
the double protonation states, the tilt angle for the (+3) and the (+4) range 130∘ – 180∘. The corresponding distance and angular range are
His-protonated states are 19∘ and 22∘ respectively. Previous studies 2.0–3.0 Å and 90∘ – 180∘, respectively, in a weak hydrogen bond. We
based on the coarse-grained modeling of the ion channel have reported have analyzed the H-bonding interactions between the water and the
the tilt angle of the closed structure to be 13 ± 3∘ [54]. The experi­ pore lining residues in the open and the closed conformations in the
mentally calculated values for the tilt angle is 14 ± 2∘ for the closed state eight states to understand the pore hydration. No strong hydrogen bonds
and 20 ± 2∘ for the open state [18]. Hence, one can conclude from both were observed in any of the eight systems. The average number of
the experiments and computational studies that the increase in the tilt hydrogen bonds along with the standard deviations is given in Table 3.
angle leads to the activation of the ion channel at acidic pH. The shape of the pore and the structure of the ion channel is pivotal
Another important characteristic of the BM2 activation is the change for the accessibility of the water molecules which in-turn increases the
in the intra-helical packing. The packing of the helix bundle is important average water – pore residue hydrogen bonding. Thus, in general the
for the stability, folding and the association of the membrane proteins activation of BM2 leads to the increase of the hydrogen bonding between
[55–57]. The internal packing of membrane proteins is in many ways the pore lining residues and the pore water molecules except for a few
different from that of the soluble proteins. The hydrophobic core of in­ cases towards the C-terminal region. The most notable change in the H-
tegral membrane proteins is most often formed by well-packed mem­ bond number happens in the L8 residues for both the singly protonated
brane spanning α-helices, like in the case of influenza B M2 channel. and the doubly protonated systems in agreement with the previous
Only in a few rare cases, there are other structures, like in case of porrins studies [64]. The water density of the ion channel with the His pro­
consisting of antiparallel bundle of β-strands [58], or a combination of tonations in the closed and the open structures are shown in the Sup­
both the helix and the β-strands like in case of acetylcholine receptor plementary Information (Fig. S5). The relative density of the waters is
[59,60]. The change in the center-of-mass intra-helical distance in case projected over the bilayer systems with the density being shown as a
of the open and the closed structures across all the protonation states is colorbar. The open channels show a broader distribution of the water
shown in Fig. 3. The center-of-mass distance of the helix-chains are molecules in both the singly and the doubly protonated systems.
calculated and the chains are being referred to as ch 1 for chain 1 and ch To further illustrate the pore hydration, the pore radius of the ion
2 for chain 2 in the Fig. 3. The distance between the helical chains for the channels is shown. At first, the molecular dynamics trajectories are
doubly protonated states are given in Supplementary Information clustered according to the GROMOS algorithm [65] to find the most
(Fig. S2). The intra-helical distance is also observed to increase with the predominant conformations across the eight trajectories. The RMSD
protonation of the His-19 in the closed state. The closed state structure at based clustering analysis was implemented with a cut-off of 0.2 nm. The
(+1) protonation shows less deviation in the intra-helix distance with most probable cluster conformations are shown in Fig. 4A – D. The
respect to the (0) protonation state. But, with the increasing protonation corresponding pore radii are shown in Fig. 4B and D. The increased
in histidine residues in the open state, there is an increase in the intra- protonation leads to an increase in the pore radius. Among the mono-
helix distance. The intra-helical distance though decrease in the (+3) protonated states, the (+3) and (+4) states show similar pore radii,
state as shown in the Fig. 3(c). For the doubly protonated states the whereas the doubly protonated states show pronounced differences
change in intra-helical distances are higher than the mono-protonated among the pore radii.
states as shown in the Supplementary Information (Fig. S2). The con­
tact maps showing the increasing intra-helical distances are shown in 3.2. Lipid properties
Supplementary Information (Figs. S3 and S4). Thus our MD simulations
support the claim that BM2 activation takes place through the increasing 3.2.1. Deuterium order parameter (SCD)
tilt angle and the increasing intra-helical distances. The opening and the closing of the BM2 ion channel leads to the
reorganization of the lipid bilayers. The fluidity of the lipid bilayers in
3.1.3. Hydrogen bonding the presence of the ion channels is characterized using the Deuterium
To elucidate the interactions of the protein residues with the solvent Order Parameter (SCD) for the lipid tails (sn-1 and sn-2 chains). The
water, we have characterized the hydrogen bonding formed between higher values of SCD correlates with higher rigidity as it denotes how
them. Hydrogen bonding plays an important role in determining the straight or kink the lipid chain is. The terminal atoms of the lipid tails are
nature of molecular association in membrane proteins [53,61,62]. In relatively mobile, have more kink and hence have less value of SCD

5
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Fig. 4. The most probable structures generated from the clustering of the trajectories and their corresponding pore profiles. (A) and (C) are the most probable
structures obtained from the clustering of the trajectories, (B) and (D) are the pore profiles generated from their structures. The phosphate groups are colored in blue,
the lipid tails are colored in green and the protein structures are colored in red. (For interpretation of the references to colour in this figure legend, the reader is
referred to the web version of this article.)

whereas the central atoms in the chains are relatively straighter and the decreasing ordering in the lipid chains near the middle carbon
have high SCD. The conformational changes in the lipid tails for the C–H atoms. This is evident from both the single and the double histidine
bonds are computed through protonations in both the sn1 and sn2 chains across the open and the
closed states. The closed BM2 ion channel leads to the high values of SCD
1( )
SCD = − 3cos2 θ − 1 (1) and thus higher ordering of the lipid tails in the middle part of the lipids.
2

where θ is the angle between the C – H bond and the bilayer normal. 3.2.2. Bilayer thickness
The SCD values of the acyl chains in all the systems are shown in The lipid packing of the bilayers has been studied using the bilayer
Fig. 5. The protonation of the histidine residues in the open BM2 leads to thickness. The bilayer thickness is calculated as the distance between the

6
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Fig. 6. The Q(x) for the four mono-protonated structures as a function of time.

1 ∑ 1
Q(X) = [ ( ) ] (2)
|S| (i,j)∈S1 + exp β rij (X) − λrij0

Fig. 5. The Deuterium order parameters (SCD) across all the systems. (a) and (b) where X is a conformation, rij(X) is the distance between the atoms i and j
are for the His-19 protonation states and the (c) and (d) are for the His-19/His- in conformation X, and S is the set of the heavy atoms (i, j) belonging to
27 protonation states in both the sn-1 and the sn-2 lipid tails. residues θi and θj such that ∣θi – θj∣ > 3 and the distance between the i and
j is less than 4.5 Å. r0ij is the distance between the i and j in the native
state. The smoothing parameter β is set to be 5 Å− 1 and the contact
Table 4 fluctuation is taken into account through the parameter λ = 1.8. The
Bilayer thickness of various systems studied. The standard de­ summation is carried out over all the ∣S∣ pairs. The native contacts are
viations are given in the parenthesis.
computed for all the systems from uncharged His-residue to (+4)-
System Bilayer thickness (nm) charged His tetrad. The Q(x) is shown for the entire trajectory in Fig. 6.
His-190 4.32 (0.02) Assuming that the Q(x) captures the reaction co-ordinate for the protein
His-191+ 4.30 (0.03) folding, the greater deviation of the Q(x) relates to larger deviation from
His-193+ 4.29 (0.08) the native structure. The closed and the open ssNMR structures of BM2
His-194+ 4.23 (0.02)
are taken as the reference for computing Q(x). Fig. 6 shows the Q(x) for
His-190/His-271+ 4.31 (0.08)
His-191+/His-271+ 4.33 (0.03) the mono-protonated structures. The (0) and (+1) states show larger
His-193+/His-273+ 4.29 (0.08) deviation from the average Q(x) value. The Q value of the neutral and
His-194+/His-274+ 4.29 (0.07) the (+1) states are 0.675 and 0.769 respectively at the point of
convergence beyond which the values do not change. In case of the
structures in the (+3) and the (+4) charges, the corresponding mean
phosphate headgroups in the two leaflets of the bilayer. The bilayer
values are 0.85 and 0.849 respectively. The changes in the Q value for
thickness calculated for our systems is given in Table 4 along with the
the closed structures are thus more significant with the increasing pro­
standard deviations.
tonation. The protonation induces greater changes in the native contacts
The thickness of the lipid bilayers is greater in the closed B M2
in the closed structure compared to the open structure. The Q(x) for the
channels than in the open ion channels. The linearity or the kink of the
doubly protonated states are shown in the Supplementary Information
lipid channels is responsible for increasing or decreasing the bilayer
(Fig. S6). The converged Q value of the closed structures are 0.783 and
thickness. It’s been known from the earlier studies that the thicker
0.786 respectively. The corresponding Q values of the open structures
membrane induces closer lipid packing [66,67]. We envisage that the
are 0.835 and 0.886 respectively. The open structure at the (+4) pro­
ordering of the lipid tails (as shown in Fig. 5) in the closed BM2 state
tonation thus has the Q value higher than the mono-protonated (+4)
induces greater lipid thickness in the closed state over the open state.
state. Thus overall the doubly protonated states (except for the His-
The greater ordering in the closed states mean higher linearity in the
193+/His-273+ system) are showing higher value of Q than the singly
lipid chains (with the lowering of the kink) which increases the bilayer
protonated states.
thickness. Hence, the opening of the BM2 channel leads to the change in
the lipid packing due to kink induced in the lipid tails.
3.2.4. Structure factor (S2)
The rotational correlation of the backbone NH bonds is important for
3.2.3. Native contacts
correlating the experimentally probed NMR relaxation data with MD
The native contacts (Q(x)) determine the folding mechanism of a
simulations [69]. The time scale of the internal motions can be probed
protein structure. Thus Q(x) traces the folding mechanism of a protein
using the MD data through suitable correlation functions and order
[68] and provides a suitable reaction coordinate for the protein folding
parameters. NMR measures the backbone dipolar relaxation of the two
through the atomistic and coarse-grained simulations. We have calcu­
nuclei (backbone NH dipole) that can be expressed as
lated the Q(x) as a reaction coordinate to track the change in the native
contacts responsible for the opening of the BM2 ion channel. The frac­ ′ ′
Ci (t) = 〈P2 (μi (t )⋅μi (t + t) ) 〉 (3)
tion of native contacts is defined as follows
where μi(t) is the dipole moment of the backbone NH bond and the P2(x)

7
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Fig. 7. S2 values for the pore lining residues across the eight systems.

is the second order Legendre polynomial. The angular brackets represent

the averaging over the dipole moments at different time origin. The
protein structures are initially refined to get rid of the center-of-mass
translational and rotational motion for computing the Ci(t). This is
because the total correlation function is assumed to be resolved through
the correlation motions for the overall motion Co(t) and the internal
motion Ci(t). The motions related to the overall Brownian tumbling and
internal fluctuations are mutually independent.
C(t) = Co (t)⋅Ci (t) (4) Fig. 8. Vibrational modes corresponding to the lowest frequency computed
across the C-α atoms for all the mono-protonated systems. The arrows attached
In accordance with the model free approach of Lippari-Szabo, it has to each backbone atom indicate the direction of the eigenvector and the length
been assumed that the internal and the overall motions are independent. of each arrow shows the magnitude of the corresponding eigenvalue. (a) and (b)
The internal correlation function Ci(t) is expressed by the following are the closed states at neutral and (+1) protonation of the His-tetrad, and (c)
relation and (d) are the open states at (+3) and (+4) protonation states of His-tetrad.
( )
Ci (t) = S2 + 1 − S2 e− t/τe (5)
based on the elastic network model calculations to calculate the col­
2 lective dynamics of the complex protein structures. According to the
where S is the order parameter and the τe is the effective internal self-
model, the i-th and j-th C-α atoms interact through a harmonic potential
correlation time of the NH bond vector. The structure factor S2 accounts
given by the following equation,
for the backbone fluctuations in the protein. The higher value of the S2
leads to greater ordering and less conformational fluctuations where a γ ( )2
Vij = Γij rij − rij0 (6)
value of 1 defines a rigid state and a value of 0 defines an unhindered 2
motion of the NH dipole. The pore lining residues in different proton­
ation states show different ordering owing to the differences in the where rij is the instantaneous distance, r0ij is the equilibrium distance
rotational fluctuations of the backbone NH dipole as shown in Fig. 7. The between the i-th and the j-th C-α atoms, γ is the force constant taken to
differences in the structure factors are more pronounced between the be 1 (in relative units), Γij is the matrix of contacts which becomes “1”
neutral and (+1) states in the closed conformation than between the when r0ij ≤ rc and “0” otherwise (where rc = 1.5 nm). The vibrational
(+3) and (+4) states in the open conformation. The previously reported modes are calculated by constructing the second derivative matrix for
structure factors computed using the ss-NMR techniques show that the the C-α modes of the protein. The Hessian (H) for the BM2 ion channel
histidine and tryptophan residues of the AM2 give rise to concerted having 132C-α atoms comprises of 396 × 396 elements with 132 × 132
dynamics owing to their proximity [70]. Due to the conserved HxxxW superelements (Hij). Hij carries the information of the anisotropy of a
domain of AM2 and the BM2, we envisage common features between particular ij pair.
these two. Among the single protonation states, the structure factor of γ ( T )
Hij = − ( )2 B ⋅B (7)
the backbone NH bonds corresponding to the pore lining residues show rij0
similar characteristics in case of the (+3) and (+4) structures. On the
contrary, the neutral and (+1) structures are more distinct, especially in where B = [xi − xj, yi − yj, zi − zj]. The decomposition of H gives 3 N – 6
case of the S8 residue. For the doubly protonated states the C-terminal eigenvalues and eigenvectors that give us the respective frequencies and
shows higher fluctuations. Overall the doubly protonated states are the individual modes.
more disordered compared to the singly protonated states. The disor­ The open channels and the closed channels show different collective
dered doubly protonated states pave the way for the increasing pore motions across various protonation states of the His-tetrads. The
hydration and increasing pore radius. changes in the histidine protonation state change the principle motion of
the whole ion channels. We consider the low frequency motions across
3.2.5. Free energy landscape the eight structures to visualize the highest mean collective motion
We have determined the distinct low-frequency and large-amplitude across all the channels. A large region of the overall fluctuations of the
modes of fluctuations across the eight protonation states of the BM2 protein can often be accounted for by a few low-frequency eigenvectors
channel through the anisotropic network model (ANM). ANM analysis is

8
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

Fig. 9. The free energy landscape of the ion channels across the first two principal components. (a) and (b) are for the closed state at the neutral and (+1) pro­
tonation states of the His-19 residue, respectively and (c) and (d) are for the open state at the protonation states of (+3) and (+4) respectively. Figs. (A-D) are the
superimposed structures corresponding to minima and the maxima of the free energy landscape with the colour code used for the individual α-helices. The con­
formations at the maxima are colored in teal whereas each α-helix in the conformation corresponding to the minima are colored differently. (For interpretation of the
references to colour in this figure legend, the reader is referred to the web version of this article.)

9
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

with large eigenvalues [71]. The first eigenvectors (or the most promi­ hindered by the M2 inhibitors. The classical molecular dynamics studies
nent mode) of vibration for the four mono-protonated systems are reported here provide insights into the pore hydration in case of the BM2
shown in Fig. 8. The eigenvectors of the four doubly protonated systems channels with singly-protonated and doubly-protonated histidine resi­
are shown in Supplementary Information (Fig. S7). In the case of the dues in the open and closed conformations.
closed systems in the mono-protonated states, there is a clear preference Atomistic molecular dynamics simulations have been performed
for the alpha helices to undergo the tilting motion. The vibrational mode using the ss-NMR structures having a resolution of 1.4–1.5 Å in presence
for the helix-tilt is more predominant at the N-terminus of the helix of the lipid bilayers having the composition of POPC:POPG:CHOL at the
bundle across all the systems. In case of the (+3) and the (+4) proton­ molar ratios of 3:1:1. The histidine protonation in the influenza channel
ated states in the mono-protonated open structure, the mode of the tilt is taken to be neutral and (+1) in the closed state at the high pH and
motion of the bundle seems to be predominant. This happens via (+3) and (+4) in the open state at the low pH. The eight systems at the
increasing tilt angle of the helix bundle and concomitant change in the various histidine protonation are studied here. Our computational
inter-helical packing in the open and the closed states. For the doubly studies reveal the following details: (a) With the increase in the pro­
protonated states the vibrational modes corresponding to the four states tonation state of the His-tetrads from the neutral closed state to the (+4)
are different from the singly protonated structures. The common mode open state, the tilt angle of the ion channel and the intra-helical dis­
of vibration for the doubly protonated states is the tilting of the helices tances increase. (b) The hydrogen bonding of the pore waters and the
and the intra-helical breathing mode (that gives rise to the fluctuations pore lining residues show the most discrepancy near the L8 residues. (c)
in the intra-helical distance). The AM2 channels also show similar The greater ordering of the lipid tails in the closed state of influenza BM2
vibrational modes (which leads to the increasing the tilt angle) as results in greater bilayer thickness. (d) The free energy landscape of the
observed through the corresponding ensemble of the TM regions of M2 four systems across the first two principal components gives us the
obtained by X-ray crystallography, NMR, and ss-NMR [72]. important conformations across the sampling space and corroborates
To put forward the hypothesis of the BM2 activation through the that the increase in protonation leads to increasing differences in the
increasing tilt-angle, we have also projected the first two eigenvalues of intra-helical distances and the tilt angle. The overlap of the maxima and
the principal components (PCs) into the Free Energy Landscape. The the minima structures is greater in the (0) state in the closed structure
Boltzmann inversion of the PCA is used to get the free energy landscape pointing to the inactivity of the ion channel. The other structures show
according to the equation the activation and the opening of the ion channel due to the increase in
[ ] the intra-peptide distances and the tilt angles. Thus, histidine proton­
ΔG(R) = − kB T lnP(R1 , R2 ) − lnP(R)max (8)
ation is pivotal for the activation of the BM2 ion channel.
The findings from these simulations explain the structural and
where P(R1, R2) are the population of the principal components in the
dynamical changes taking place in the BM2 ion channel upon a change
phase space and P(R)max is the maximum of the distribution when the
in the protonation state of the His-19 tetrad and provide insights into the
minimum of the energy landscape is set to zero. The free energy basins
debated protonation state of the histidine tetrads.
are different for each of the four mono-protonated systems as shown in
Fig. 9(a–d). The maximum and the minimum structures borne out of the
free energy landscape are superimposed in Fig. 9(A–D). The free energy CRediT authorship contribution statement
basins for the doubly protonated systems are shown in the Supplemen­
tary Information (Fig. S8). The conformational states at the maxima and Unmesh D. Chowdhury: Methodology, Investigation, Software,
the minima of the energy landscape show some distinct characteristics. Data curation, Writing – original draft. B.L. Bhargava: Methodology,
It was found that in all the cases, except for the neutral histidine pro­ Writing – review & editing, Supervision, Project administration.
tonation in the closed state, there is a tendency for the tilt angle to in­
crease. The intra-helical distances are found to increase in the
conformations corresponding to the minima. It is evident from the su­ Declaration of Competing Interest
perimposition of the conformations, that the differences among the
structures corresponding to the maximum and the minimum on the The authors announce no known competing financial interests.
landscape are less prominent only in the neutral state. In case of the
closed conformation in (0) state, the structures corresponding to the Acknowledgments
minima and the maxima are similar which is evident from the super­
imposition. However, in the case of protonated states across all the other The authors gratefully acknowledge NISER, Bhubaneswar for the
structures, there is a clear mismatch among the helices. The mismatch is computational resources.
observed due to the change in the tilt angle, and increasing or decreasing
of intra-helical distances. Hence, we can say that the BM2 channel gets Appendix A. Supplementary data
activated through the increasing tilt angle. The protonation of the
histidine-residues lead to the activation of the ion channel due to an The supplementary information comprises the Root Mean Square
increase in tilt angle and intra-helix distance which is evident from the Deviation (RMSD) of the protein structures at various protonation state
superimposed structures at the (+3) and (+4) state. across the closed and the open states (Fig. S1), the intra-helical distances
for the doubly protonated systems (Fig. S2), contact maps for both the
4. Conclusions single and doubly protonation (Fig. S3 – S4), water density over the
bilayer plane (Fig. S5), fraction of native contacts for the doubly pro­
The AM2 and BM2 ion channels of the influenza virus are important tonated states (Fig. S6), normal mode analysis of the doubly protonated
for the viral life cycle and hence are the target for a number of influenza systems (Fig. S7), free energy landscapes for the doubly protonated
drugs. These viroporins are the critical drug binding sites for thera­ systems (Fig. S8).
peutics. Rimantadine and Amantadine are the M2 channel inhibitors.
Amantadine can inhibit the AM2 channels by blocking the water chan­
References
nel and hindering the proton transport [33,73]. Though AM2 and BM2
channels consist of conserved HxxxW motifs, the search for BM2 in­ [1] B. Hille, Ion Channels of Excitable Membranes, 2001.
hibitors is still in progress. The continuous water wire is important to [2] D. Purves, G. Augustine, D. Fitzpatrick, L. Katz, A. LaMantia, J. McNamara,
maintain the proton conductance across the BM2 channel which is S. Williams, Voltage-Gated Ion Channels, Neuroscience, Sinauer Associates,
Sunderland, 2001.

10
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

[3] A. Tasneem, L.M. Iyer, E. Jakobsson, L. Aravind, Identification of the prokaryotic amber, openmm, and charmm/openmm simulations using the charmm36 additive
ligand-gated ion channels and their implications for the mechanisms and origins of force field, J. Chem. Theory Comput. 12 (1) (2016) 405–413.
animal cys-loop ion channels, Genome Biol. 6 (1) (2005) R4. [32] J.K. Williams, D. Tietze, M. Lee, J. Wang, M. Hong, Solid-state nmr investigation of
[4] I. Kass, I.T. Arkin, How ph opens a h+ channel: the gating mechanism of influenza a the conformation, proton conduction, and hydration of the influenza b virus m2
m2, Structure 13 (12) (2005) 1789–1798. transmembrane proton channel, J. Am. Chem. Soc. 138 (26) (2016) 8143–8155.
[5] K. Henzler-Wildman, Analyzing conformational changes in the transport cycle of [33] Y. Zhang, Q.-C. Zheng, In silico analysis revealed a unique binding but ineffective
emre, Curr. Opin. Struct. Biol. 22 (1) (2012) 38–43. mode of amantadine to influenza virus b m2 channel, J. Phys. Chem. Lett. 12 (4)
[6] H. Chen, J. Deng, Q. Cui, B. Chanda, K. Henzler-Wildman, Mapping temperature- (2021) 1169–1174.
dependent conformational change in the voltage-sensing domain of an engineered [34] Y. Zhang, Q.-C. Zheng, What are the effects of the serine triad on proton conduction
heat-activated k+ channel, Proc. Natl. Acad. Sci. U. S. A. 118 (14) (2021), of an influenza b m2 channel? An investigation by molecular dynamics
e2017280118. simulations, Phys. Chem. Chem. Phys. 21 (17) (2019) 8820–8826.
[7] A. Lewis, V. Kurauskas, M. Tonelli, K. Henzler-Wildman, Ion-dependent structure, [35] W.L. Jorgensen, J. Chandrasekhar, J.D. Madura, R.W. Impey, M.L. Klein,
dynamics, and allosteric coupling in a non-selective cation channel, Nat. Commun. Comparison of simple potential functions for simulating liquid water, J. Chem.
12 (1) (2021) 6225. Phys. 79 (2) (1983) 926–935.
[8] W.B. Fischer, M.S. Sansom, Viral ion channels: structure and function, Biochim. [36] M.J. Abraham, T. Murtola, R. Schulz, S. Páll, J.C. Smith, B. Hess, E. Lindahl,
Biophys. Acta Biomembr. 1561 (1) (2002) 27–45. Gromacs: high performance molecular simulations through multi-level parallelism
[9] J.-Y. Min, K. Subbarao, Cellular targets for influenza drugs, Nat. Biotechnol. 28 (3) from laptops to supercomputers, SoftwareX 1 (2015) 19–25.
(2010) 239–240. [37] J. Huang, A.D. MacKerell Jr., Charmm36 all-atom additive protein force field:
[10] N. Hari Narayana Moorthy, V. Poongavanam, V. Pratheepa, Viral m2 ion channel validation based on comparison to nmr data, J. Comput. Chem. 34 (25) (2013)
protein: a promising target for anti-influenza drug discovery, Mini-Rev. Med. 2135–2145.
Chem. 14 (10) (2014) 819–830. [38] J.B. Klauda, R.M. Venable, J.A. Freites, J.W. O’Connor, D.J. Tobias, C. Mondragon-
[11] A.A. Ahmed, M. Abouzid, Arbidol targeting influenza virus a hemagglutinin; a Ramirez, I. Vorobyov, A.D. MacKerell Jr., R.W. Pastor, Update of the charmm all-
comparative study, Biophys. Chem. 277 (2021), 106663. atom additive force field for lipids: validation on six lipid types, J. Phys. Chem. B
[12] R. Acharya, V. Carnevale, G. Fiorin, B.G. Levine, A.L. Polishchuk, V. Balannik, 114 (23) (2010) 7830–7843.
I. Samish, R.A. Lamb, L.H. Pinto, W.F. DeGrado, M.L. Klein, Structure and [39] J.B. Lim, B. Rogaski, J.B. Klauda, Update of the cholesterol force field parameters
mechanism of proton transport through the transmembrane tetrameric m2 protein in charmm, J. Phys. Chem. B 116 (1) (2012) 203–210.
bundle of the influenza a virus, Proc. Natl. Acad. Sci. U. S. A. 107 (34) (2010) [40] T. Darden, D. York, L. Pedersen, Particle mesh ewald: an n-log (n) method for
15075–15080. ewald sums in large systems, J. Chem. Phys. 98 (12) (1993) 10089–10092.
[13] V. Balannik, V. Carnevale, G. Fiorin, B.G. Levine, R.A. Lamb, M.L. Klein, W. [41] D. Van Der Spoel, P.J. van Maaren, The origin of layer structure artifacts in
F. DeGrado, L.H. Pinto, Functional studies and modeling of pore-lining residue simulations of liquid water, J. Chem. Theory Comput. 2 (1) (2006) 1–11.
mutants of the influenza a virus m2 ion channel, Biochemistry 49 (4) (2010) [42] B. Hess, H. Bekker, H.J. Berendsen, J.G. Fraaije, Lincs: a linear constraint solver for
696–708. molecular simulations, J. Comput. Chem. 18 (12) (1997) 1463–1472.
[14] D. Owusu, J. Hand, M.W. Tenforde, L.R. Feldstein, J. DaSilva, J. Barnes, G. Lee, [43] W. Humphrey, A. Dalke, K. Schulten, Vmd: visual molecular dynamics, J. Mol.
J. Tran, T. Sokol, A.M. Fry, L. Brammer, M.A. Rolfes, Early season pediatric Graph. 14 (1) (1996) 33–38.
influenza b/victoria virus infections associated with a recently emerged virus [44] N. Michaud-Agrawal, E.J. Denning, T.B. Woolf, O. Beckstein, Mdanalysis: a toolkit
subclade — Louisiana, 2019, Morb. Mortal. Wkly Rep. 69 (2) (2020) 40. for the analysis of molecular dynamics simulations, J. Comput. Chem. 32 (10)
[15] M. Nyirenda, R. Omori, H.L. Tessmer, H. Arimura, K. Ito, Estimating the lineage (2011) 2319–2327.
dynamics of human influenza b viruses, PLoS One 11 (11) (2016), e0166107. [45] O.S. Smart, J.G. Neduvelil, X. Wang, B. Wallace, M.S. Sansom, Hole: a program for
[16] C. Wang, K. Takeuchi, L. Pinto, R.A. Lamb, Ion channel activity of influenza a virus the analysis of the pore dimensions of ion channel structural models, J. Mol. Graph.
m2 protein: characterization of the amantadine block, J. Virol. 67 (9) (1993) 14 (6) (1996) 354–360.
5585–5594. [46] A. Bakan, L.M. Meireles, I. Bahar, Prody: protein dynamics inferred from theory
[17] R.M. Pielak, J.R. Schnell, J.J. Chou, Mechanism of drug inhibition and drug and experiments, Bioinformatics 27 (11) (2011) 1575–1577.
resistance of influenza a m2 channel, Proc. Acad. Natl. Sci. U.S.A. 106 (18) (2009) [47] J. Wang, R.M. Pielak, M.A. McClintock, J.J. Chou, Solution structure and
7379–7384. functional analysis of the influenza b proton channel, Nat. Struct. Mol. Biol. 16 (12)
[18] V.S. Mandala, A.R. Loftis, A.A. Shcherbakov, B.L. Pentelute, M. Hong, Atomic (2009) 1267–1271.
structures of closed and open influenza b m2 proton channel reveal the conduction [48] R.M. Pielak, J.J. Chou, Influenza m2 proton channels, Biochim. Biophys. Acta 1808
mechanism, Nat. Struct. Mol. Biol. 27 (2) (2020) 160–167. (2) (2011) 522–529.
[19] S. Maurya, S. Vishweshwaraiah, G. Ayappa, R. Roy, Pore assembly of bacterial [49] J.A. Killian, Hydrophobic mismatch between proteins and lipids in membranes,
alpha pore-forming toxin (αpft), cytolysin a on lipid membranes, Biophys. J. 118 Biochim. Biophys. Acta Biomembr. 1376 (3) (1998) 401–416.
(3) (2020) 364a. [50] O. Bavi, M. Vossoughi, R. Naghdabadi, Y. Jamali, The combined effect of
[20] S. Panzuela, D. Tieleman, L. Mederos, E. Velasco, Molecular ordering in lipid hydrophobic mismatch and bilayer local bending on the regulation of
monolayers: an atomistic simulation, Langmuir 35 (42) (2019) 13782–13790. mechanosensitive ion channels, PLoS One 11 (3) (2016), e0150578.
[21] A. Ardalan, S. Sowlati-Hashjin, S.O. Uwumarenogie, M. Fish, J. Mitchell, [51] R. Baccouch, E. Rascol, K. Stoklosa, I.D. Alves, The role of the lipid environment in
M. Karttunen, M.D. Smith, M. Jelokhani-Niaraki, Functional oligomeric forms of the activity of g protein coupled receptors, Biophys. Chem. 106794 (2022).
uncoupling protein 2: strong evidence for asymmetry in protein and lipid bilayer [52] A. Ramamoorthy, S.K. Kandasamy, D.-K. Lee, S. Kidambi, R.G. Larson, Structure,
systems, J. Phys. Chem. B 125 (1) (2020) 169–183. topology, and tilt of cell-signaling peptides containing nuclear localization
[22] A. Martyna, B. Bahsoun, J.J. Madsen, F.S.J. Jackson, M.D. Badham, G.A. Voth, J. sequences in membrane bilayers determined by solid-state nmr and molecular
S. Rossman, Cholesterol alters the orientation and activity of the influenza virus m2 dynamics simulation studies, Biochemistry 46 (4) (2007) 965–975.
amphipathic helix in the membrane, J. Phys. Chem. B 124 (31) (2020) 6738–6747. [53] S.K. Kandasamy, D.-K. Lee, R.P. Nanga, J. Xu, J.S. Santos, R.G. Larson,
[23] I.I. Ponmalar, R. Cheerla, K.G. Ayappa, J.K. Basu, Correlated protein A. Ramamoorthy, Solid-state nmr and molecular dynamics simulations reveal the
conformational states and membrane dynamics during attack by pore-forming oligomeric ion-channels of tm2-Gabaa stabilized by intermolecular hydrogen
toxins, Proc. Natl. Acad. Sci. U. S. A. 116 (26) (2019) 12839–12844. bonding, Biochim. Biophys. Acta 1788 (3) (2009) 686–695.
[24] Y. Zhang, H. Shen, M. Zhang, G. Li, Exploring the proton conductance and drug [54] S.L. Rouse, T. Carpenter, P.J. Stansfeld, M.S. Sansom, Simulations of the bm2
resistance of bm2 channel through molecular dynamics simulations and free proton channel transmembrane domain from influenza virus b, Biochemistry 48
energy calculations at different ph conditions, J. Phys. Chem. B 117 (4) (2013) (42) (2009) 9949–9951.
982–988. [55] M. Heyden, J.A. Freites, M.B. Ulmschneider, S.H. White, D.J. Tobias, Assembly and
[25] L.C. Watkins, R. Liang, J.M. Swanson, W.F. DeGrado, G.A. Voth, Proton-induced stability of α-helical membrane proteins, Soft Matter 8 (30) (2012) 7742–7752.
conformational and hydration dynamics in the influenza a m2 channel, J. Am. [56] R.F. Walters, W.F. DeGrado, Helix-packing motifs in membrane proteins, Proc.
Chem. Soc. 141 (29) (2019) 11667–11676. Acad. Natl. Sci. U.S.A. 103 (37) (2006) 13658–13663.
[26] L.C. Watkins, W.F. DeGrado, G.A. Voth, Influenza a m2 inhibitor binding [57] M. Eilers, S.C. Shekar, T. Shieh, S.O. Smith, P.J. Fleming, Internal packing of
understood through mechanisms of excess proton stabilization and channel helical membrane proteins, Proc. Acad. Natl. Sci. U.S.A. 97 (11) (2000)
dynamics, J. Am. Chem. Soc. 142 (41) (2020) 17425–17433. 5796–5801.
[27] Y. Zhang, H.-X. Zhang, Q.-C. Zheng, In silico study of membrane lipid composition [58] S.W. Cowan, Bacterial porins: lessons from three high-resolution structures, Curr.
regulating conformation and hydration of influenza virus b m2 channel, J. Chem. Opin. Struct. Biol. 3 (4) (1993) 501–507.
Inf. Model. 60 (7) (2020) 3603–3615. [59] J.-L. Galzi, J.-P. Changeux, Neurotransmitter-gated ion channels as unconventional
[28] Y. Zhang, H. Zhang, Q. Zheng, A unique activation–promotion mechanism of the allosteric proteins, Curr. Opin. Struct. Biol. 4 (4) (1994) 554–565.
influenza b m2 proton channel uncovered by multiscale simulations, Phys. Chem. [60] N. Unwin, Nicotinic acetylcholine receptor at 9 å resolution, J. Mol. Biol. 229 (4)
Chem. Phys. 21 (6) (2019) 2984–2991. (1993) 1101–1124.
[29] B. Kwon, M. Roos, V.S. Mandala, A.A. Shcherbakov, M. Hong, Elucidating relayed [61] J. Medeiros-Silva, S. Jekhmane, M. Baldus, M. Weingarth, Hydrogen bond strength
proton transfer through a his–trp–his triad of a transmembrane proton channel by in membrane proteins probed by time-resolved 1h-detected solid-state nmr and md
solid-state nmr, J. Mol. Biol. 431 (14) (2019) 2554–2566. simulations, Solid State Nucl. Magn. Reson. 87 (2017) 80–85.
[30] S. Jo, T. Kim, V.G. Iyer, W. Im, Charmm-gui: a web-based graphical user interface [62] S.-Y. Sheu, E.W. Schlag, H.L. Selzle, D.-Y. Yang, Hydrogen bonds in membrane
for charmm, J. Comput. Chem. 29 (11) (2008) 1859–1865. proteins, J. Phys. Chem. B 113 (15) (2009) 5318–5326.
[31] J. Lee, X. Cheng, J.M. Swails, M.S. Yeom, P.K. Eastman, J.A. Lemkul, S. Wei, [63] G.R. Desiraju, T. Steiner, The Weak Hydrogen Bond: In Structural Chemistry and
J. Buckner, J.C. Jeong, Y. Qi, S. Jo, V.S. Pande, D.A. Case, C.L. Brooks, A. Biology vol. 9, International Union of Crystal, 2001.
D. MacKerell, J.B. Klauda, W. Im, Charmm-gui input generator for namd, gromacs,

11
U.D. Chowdhury and B.L. Bhargava Biophysical Chemistry 289 (2022) 106859

[64] M.D. Gelenter, V.S. Mandala, M.J. Niesen, D.A. Sharon, A.J. Dregni, A.P. Willard, [69] F. Kummerer, S. Orioli, D. Harding-Larsen, F. Hoffmann, Y. Gavrilov, K. Teilum,
M. Hong, Water orientation and dynamics in the closed and open influenza b virus K. Lindorff-Larsen, Fitting side-chain nmr relaxation data using molecular
m2 proton channels, Commun. Biol. 4 (1) (2021) 1–14. simulations, J. Chem. Theory Comput. 17 (8) (2021) 5262–5275.
[65] X. Daura, K. Gademann, B. Jaun, D. Seebach, W.F. Van Gunsteren, A.E. Mark, [70] J.K. Williams, Y. Zhang, K. Schmidt-Rohr, M. Hong, Ph-dependent conformation,
Peptide folding: when simulation meets experiment, Angew. Chem. 38 (1–2) dynamics, and aromatic interaction of the gating tryptophan residue of the
(1999) 236–240. influenza m2 proton channel from solid-state nmr, Biophys. J. 104 (8) (2013)
[66] A. Schumann-Gillett, M.L. O’Mara, The effects of oxidised phospholipids and 1698–1708.
cholesterol on the biophysical properties of popc bilayers, Biochim. Biophys. Acta [71] M.M. Teeter, D.A. Case, Harmonic and quasiharmonic descriptions of crambin,
Biomembr. 1861 (1) (2019) 210–219. J. Phys. Chem. 94 (21) (1990) 8091–8097.
[67] W.-C. Hung, M.-T. Lee, F.-Y. Chen, H.W. Huang, The condensing effect of [72] J. Wang, J.X. Qiu, C. Soto, W.F. DeGrado, Structural and dynamic mechanisms for
cholesterol in lipid bilayers, Biophys. J. 92 (11) (2007) 3960–3967. the function and inhibition of the m2 proton channel from influenza a virus, Curr.
[68] R.B. Best, G. Hummer, W.A. Eaton, Native contacts determine protein folding Opin. Struct. Biol. 21 (1) (2011) 68–80.
mechanisms in atomistic simulations, Proc. Natl. Acad. Sci. U. S. A. 110 (44) [73] S. Llabres, J. Juarez-Jimenez, M. Masetti, R. Leiva, S. Vazquez, S. Gazzarrini,
(2013) 17874–17879. A. Moroni, A. Cavalli, F.J. Luque, Mechanism of the pseudoirreversible binding of
amantadine to the m2 proton channel, J. Am. Chem. Soc. 138 (47) (2016)
15345–15358.

12