Accepted Manuscript

Activation of human insulin by vitamin E: A molecular dynamics simulation study

Hossein Soleymani, Mohammad Ghorbani, Abdollah Allahverdi, SeyedehsamanehShojaeilangari, Hossein Naderi-manesh

PII: S1093-3263(19)30129-9

DOI: https://doi.org/10.1016/j.jmgm.2019.06.006

Reference: JMG 7389

To appear in: Journal of Molecular Graphics and Modelling

Received Date: 3 March 2019

Revised Date: 2 June 2019

Accepted Date: 3 June 2019

Please cite this article as: H. Soleymani, M. Ghorbani, A. Allahverdi, S. Shojaeilangari, H. Naderi-manesh, Activation of human insulin by vitamin E: A molecular dynamics simulation study, Journal ofMolecular Graphics and Modelling (2019), doi: https://doi.org/10.1016/j.jmgm.2019.06.006.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service toour customers we are providing this early version of the manuscript. The manuscript will undergocopyediting, typesetting, and review of the resulting proof before it is published in its final form. Pleasenote that during the production process errors may be discovered which could affect the content, and alllegal disclaimers that apply to the journal pertain.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

1

Activation of human insulin by vitamin E: A molecular dynamics

simulation study

Hossein Soleymani1, Mohammad Ghorbani1, Abdollah Allahverdi1, Seyedehsamaneh Shojaeilangari2, and

Hossein Naderi-manesh1, 3*

1 Biophysics Department, Faculty of Biological Sciences, Tarbiat Modares University, 14115-154, Tehran, Iran

2 School of Cognitive Science, Institute for Research in Fundamental Sciences (IPM), 193955746, Tehran, Iran

3 School of Biological Science, Institute for Research in Fundamental Sciences (IPM), 19395-5746, Tehran, Iran

EMAIL:

Hossein Soleymani: hsoleymani@modares.ac.ir

Mohammad Ghorbani: mohammad.ghorbani.janatabad@gmail.com

Abdollah Allahverdi: a-allahverdi@modares.ac.ir

Seyedehsamane Shojaeilangari: s.shojaei@ipm.ir

Corresponding Author: Hossein Naderi-manesh: naderman@modares.ac.ir

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

2

Abstract:

Lack of perfect insulin signaling can lead to the insulin resistance, which is the hallmark of

diabetes mellitus. Activation of insulin and its binding to the receptor for signaling process

initiates via B-chain C-terminal hinge conformational change through an open structure to

“wide-open” conformation. Observational studies and basic scientific evidence suggest that

vitamin D and E directly and/or indirectly prevent diabetes through improving glucose

secretion and tolerance, activating calcium dependent endopeptidases and thus improving

insulin exocytosis, antioxidant effect and reducing insulin resistance. On the contrary, clinical

trials have yielded inconsistent results about the efficacy of vitamin D supplementations for

the control of glucose hemostasis. In this work, best binding modes of vitamin D3 and E on

insulin obtained from AutoDock Vina were selected for Molecular Dynamic, MD, study. The

binding energy obtained from Molecular Mechanics- Poisson Boltzman Surface Area, MM-

PBSA method, revealed that Vitamins D3 and E have good affinity to bind to the insulin and

vitamin E has higher binding energy (-46 kj/mol) by engaging more residues in binding site.

Distance and angle calculation results illustrated that vitamin E changes the B-chain

conformation and it causes the formation of wide-open/active form of insulin. Vitamin E

increases the ValB12-TyrB26 distance to ~15 Å and changes the hinge angle to ~65°.

Consequently, essential hydrophobic residues for binding to insulin receptor exposed to

surface in the presence of vitamin E. However, our data illustrated that vitamin D3 cannot

change B-chain conformation. Thus our MD simulations propose a model for insulin

activation through vitamin E interaction for therapeutic approaches.

Keywords: Insulin, Hinge opening, Diabetes mellitus, Molecular dynamics simulation, MM-

PBSA

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

3

Abbreviations:

COM, Center Of Mass; LINCS, Linear Constraint; MD, Molecular Dynamics; MM-PBSA,

Molecular Mechanics- Poisson Boltzman Surface Area; PDB, Protein Data Bank; SPC,

Simple Point Charge; PME, Particle Mesh Ewald; Rg, Gyration Radius; RMSD, Root Mean

Square Deviation; SPC water, Simple Point-charge Water; VMD, Visual Molecular

Dynamics;

Introduction:

Observational studies and basic scientific evidence suggest that vitamin D and E directly

and/or indirectly prevent diabetes through improving glucose secretion and tolerance,

activating calcium dependent endopeptidases and thus improving insulin exocytosis,

antioxidant effect and reducing insulin resistance [1-3]. It has been reported that type 2

diabetes mellitus patients are associated with vitamin D3 deficiency [4]. Recently, it has been

demonstrated that vitamin D supplementations are inversely affiliated with insulin resistance,

which may have beneficial effects on controlling several complications of childhood obesity

[5]. Vitamin D and E supplementations could diminish insulin resistance in patients with non-

insulin dependent diabetes mellitus with substantial improvements in serum fasting plasma

glucose (FPG) and insulin [6, 7]. New clinical trial on 2,423 participants has been established

that vitamin D supplementations decrease the risk of diabetes [8]. F. Cadario et. al in 2017

reported that co-administration of omega 3 fatty acid and high dose vitamin D could assist to

halt the progression of type 1 diabetes [9]. Low dairy intake of vitamin E supplementations

during the second pregnancy trimester can dwindle the risk of hyperglycemia and insulin

resistance [10].

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

4

Supernumerary evidence has strongly proposed that vitamin D plays a significant role in

amending the risk of diabetes. However, the findings of type 2 diabetes are derived

particularly from observational and basic studies, which may be missed by a variety of

parameters. Only a few and often underpowered clinical trials have been designed to test the

effectiveness of vitamin D and E administration in diabetes [11]. Many ongoing randomized

trials are trying to test the hypothesis that vitamin D and E lower type 2 diabetes risk [12].

Insulin is a peptide produced by beta cells of islets of pancreas that plays its role as one of the

main anabolic hormones in the body. It regulates the metabolism of proteins, fats, and

carbohydrates by promulgating the absorption of glucose from the blood into skeletal muscle,

fat and liver cells [13]. Investigations in the last century illustrated that insulin served a key

role in the advancement of structural biology, peptide chemistry, pharmacology and cell

signaling [14, 15].

The functional and biologically active form of insulin is a monomer comprising 51 amino

acids containing two chains, an A-chain of 21 and a B-chain of 30 residues, which are linked

together by two inter-chain disulfide bridges between CysA7-CysB7 and CysA20-CysB19. The

A-chain has an intra-chain disulfide bridge between CysA6-CysA11 as well. The A-chain has

an N-terminal helix from GlyA1 to ThrA8, continued with an antiparallel C-terminal helix from

SerA12 to CysA20. The B-chain also has two N- and C-terminal strands surrounding a central

helix GlyB8 to CysB19. This conformation is known as T conformation (Fig. 1a and b). At

micromolar concentration, insulin dimerizes through a β sheet structure of both C-terminal

portions of the B-chain, and in the higher concentrations in the presence of zinc ions

associates into the beautiful symmetrical structures of hexamer [16, 17].

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

5

Fig. 1. Structural details of the insulin molecule and vitamins D3 and E. (a) 3D structure of human insulin

(PDB ID: 2G4M). A-chain (yellow); B-chain N-terminal (residues PheB1 to GlyB8) (red), B-chain α-helix

(residues SerB9 to CysB19) (black), B-chain C-terminal (residues GlyB20 to ThrB30) (blue). The colored spheres

demonstrate the Cα atoms of the first and last residues of the A and B-chains. (b) Side chains of the insulin

hydrophobic core: GlyB8, LeuB11, ValB12, LeuB15, PheB24, TyrB26 (shown in purple), IleA2 and ValA3 (shown in

green).

The mechanism of monomeric insulin conformational changes and its engagements with the

receptor has long been still enigmatic [18]. Attentions have been recently focused on B-chain

C-terminus portion that is critical for self-assembly in β cells and receptor attachment in

target tissues. In particular, the C-terminal residues of the B-chain (nearly PheB24-ThrB30)

egresses from the helical core to divulge buried nonpolar surfaces (the detachment model)

[19]. Therefore, the C-terminal β-strand (PheB24-ThrB27) and β-turn (GlyB20-GlyB23) rotate

away from the alpha-helical core of insulin to expose its conserved aromatic motif

(PheB24,PheB25,TyrB26) to insulin receptor binding site [20]. Most recent findings showed that

opening of B-chain C-terminus is inherently stochastic and then transition through an open

structure to “wide-open” conformation. The wide-open conformation is critical for receptor

binding, however, this conformation occurs very seldom. B-chain C-terminus opens as a

zipper mechanism; also PheB24 acts as a hinge that is retained in the prevailing closed/inactive

state by neighboring TyrB26 hydrophobic effect. The TyrB26 is the essential residue for

initiation of B-chain C-terminal opening and activation of insulin [21].

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

6

Engagement of insulin to its receptor in the primary hormone binding site (α-subunit domains

L1 and α-CT) occurs due to hinge-like rotation at insulin’s GlyB20-GlyB23 β-turn, coupling

reorientation of PheB24 to a 60o rotation of the PheB25-ThrB30 away from the protein

hydrophobic core. Inauguration of this hinge exposes conserved nonpolar side chains (PheB24,

PheB25, ValB12, IleA2 and ValA3) to engage the receptor [22]. Restraining the hinge

reconstitution by nonstandard mutagenesis preserves native conformation but blocks receptor

binding [20, 23].

Different strategies have been reported to treat diabetes, such as immunomodulation [24],

metabolic/bariatric surgery [25, 26], islet transplantation [27], amino acid substitution for

insulin stability improvement [28] and ligand binding [29, 30]. In the last decade, vitamins

have scintillated more attention to diabetes treatment [31]. Vitamin D directly and/or

indirectly prevents diabetes through improving glucose secretion and tolerance, activating

calcium dependent endopeptidases and thus improving insulin exocytosis, antioxidant effect

and reducing insulin resistance [2, 32].

Phenolic components widely used for maintaining insulin preparations can have a profound

result on the insulin’s conformation in solution [33]. Phenol group stabilizes more helix

structures in the insulin hexamer [34]. This conformational change can affect the hormone’s

stability and function [35]. Vitamin E as a phenolic compound reduces oxidative stress and

protein glycosylation as well as improves insulin sensitivity and insulin action after oral

administration [7, 10]. Recently, it was found that cholecalciferol and α-tocopherol -the

active forms of vitamin D and E in blood stream in the vicinity of insulin, strongly bind to

hydrophobic patch of human insulin and change insulin’s structure and improve its stability

[36].

In the present work, we investigated the interactions between α-tocopherol and

cholecalciferol with human insulin monomer individually using molecular dynamic

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

7

simulation approach. The aim was to study the effect of vitamin binding on insulin B-chain

conformational transition, which play an important role in the insulin/insulin receptor

attachment as a critical component for insulin function.

2. Methods:

2.1. Molecular docking

In the MD simulations, we used a crystal structure of human insulin monomer with a

resolution of 1.35 Å (PDB ID: 5ENA [37]). Vitamins structures were obtained from the

Brookhaven Protein Data Bank (PDB IDs: 3C6G [38] for vitamin D3 and 3CXI [39] for

vitamin E). Molecular docking was done using AutoDock Vina with standard parameters

[40]. Both vitamins were docked to the human insulin monomer in 26 × 26 × 34 Å grid box.

Best binding structures for each vitamin were selected using the best score of AutoDock Vina

results. By analyzing the binding energy and RMSD, certain insulin-vitamin complexes were

selected for simulation studies.

2.2. MD simulations

MD simulations were performed using the Gromacs 4.6.7 package using Gromos96 53a6

force field [41, 42]. The ligands topologies were generated by the PRODRG server [43]. The

structure of insulin monomer was solvated in SPC water molecules. Na+ and Cl- ions were

placed in the solvated system to neutralize it and to set the NaCl concentration to 0.1 M. The

system was then exposed to a primary minimization of water molecules followed by

equilibration. Steepest descent energy minimization was performed down to a maximum

gradient of 1000 KJmol-1nm-1. The system was simulated in two steps, 1000 ps of position

restrained dynamics using LINCS algorithm followed by full 100 ns standard MD [44]. The

system was coupled to a Berenson’s thermostat method with the reference temperature of 300

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

8

k, further, standard pressure of 1.0 bar with a coupling constant of 1000 ps for pressure. It is

noteworthy, during this time, positions of all bonds were constrained. PME electrostatics

were applied using the Lennard-Jones cutoff of 1.4 nm and a Coulomb cutoff of 0.9 nm [45].

MD simulations were done during 100 ns for all minimized structures with time step of 2 fs.

Equilibrium of the system was monitored from the RMSD plot. Images were created using

PyMOL molecular graphic system [46], VMD [47] and the program LigPlot+ v 1.0 [48],

which generates schematic 2-D representations of protein ligand complexes from the PDB

file input [49, 50].

2.3. MM-PBSA analysis

The binding affinity of vitamins could be investigated on the basis of binding free energy.

Gibbs free energy was calculated by the procedure MM-PBSA [51, 52]. Interaction energy

for insulin with vitamin D3 and E were calculated. MM-PBSA methods were implemented

for 90 to 100 ns. For each simulated system, 1000 snapshots were taken from the last 10 ns of

the trajectory at the intervals of 10 ps. In this study, different components of interaction

energy which contribute to vitamin binding were estimated. This includes van der Waals

interactions, electrostatic interactions, polar and non-polar solvation energy. The MM-PBSA

analysis also offers a unique contribution to the binding affinity and per residue distribution

that provides significant amino acid in all cases. Energy decomposition per residues and MM-

PBSA revealed the relative importance of amino acid involved in binding and different

components of binding energy.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

9

3. Result and Discussion:

3.1. Molecular docking

The blind-docking results indicated that insulin monomer has only one binding site for

vitamins. The grid dimensions for accurate-dock were adjusted according to the binding site.

AutoDock Vina scored twenty interaction modes for each complex based on the affinity and

distance from the best mode. Best docking conformations of vitamin-insulin complexes,

generated by Autodock, were taken as initial structure for MD simulation. Table 1 shows the

binding energies of the chosen complexes. Vitamin D3 and E bind to hydrophobic groove of

insulin with the lowest binding energy of -29.7 and -36.5 kj/mol, respectively.

Interaction modes and the orientation of vitamins are also illustrated in Fig. 2. In Fig. 2, it is

shown both vitamins D3 and E are bound to hydrophobic region between two α-helix

structures in A and B-chain interface. After binding of vitamins, the B-chain C-terminal

hinge is free to couple with its receptor by formation of wide-open conformation. In other

words, vitamins do not inhibit insulin interaction with its receptor and insulin activity is

maintained. Fig. 3 shows the hydrophobic interactions between insulin and vitamins created

by LigPlot+ program in 2D representation [53]. LeuA13, TyrA14, GluA17, PheB1, GlnB4, AlaB14,

LeuB17, ValB18 and ArgB22 residues participate in hydrophobic interaction.

Table 1. Binding energy. Obtained binding energies of the molecular docking and MM-PBSA method for

vitamin D3 and vitamin E.

Insulin-Vitamin D3 Insulin-Vitamin E

Docking Binding Energy (kj/mol) -29.7 -26.8

MD Binding Energy (kj/mol) -36.5 -46

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

10

Fig. 2. The orientations of the D3 (shown in blue) (a) and vitamins E (shown in green) (b) bound to the A-

chain; Hydrophobic surface schematics of human insulin monomer interacted with vitamin D3 (c) and vitamin E

(d). Insulin is shown in hydrophobic representation.

Fig. 3. 2D plot of protein (insulin)-ligand (vitamin D3 and E) interaction (generated by LigPlot+ obtained

from docking results). Ligplot of insulin with vitamin D3 (a). Ligplot of insulin with vitamin E (b). Red

semicircle shows hydrophobic interactions.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

11

3.2. Conformational stability of insulin

MD simulation is a reliable method to find out the structural movement of molecules and

atoms as a function of time. To explore the conformational stability of insulin and effective

conformation samplings during the MD simulations, the root mean square deviation (RMSD)

was computed based on the starting structure. The RMSD method averaging all the particles

at a time. This action is based-on the comparison of the structure after the simulation with the

first structure and the smaller its numerical value, the simulated system is more stable [54].

The backbone RMSD of A and B-chain and Rg (gyration radius) of insulin-vitamin

complexes in comparison to the apo-insulin are plotted in Fig. 4 and average values are

shown in Table 2. The entire conformation of the apo-insulin structure was observed stable

about 0.12 and 0.3 nm for A and B-chain, respectively, during the simulation. The RMSD

was generally stable, however, some interesting variations occurred in B-chain insulin, which

is related to B-chain C-terminal fluctuation. In the insulin-Vitamin E simulation, the average

RMSD of B-chain was calculated about 0.5 nm (Fig. 4b). As expected, the RMSD of insulin-

Vitamin E shows persistent variation, probably because of phenol effect on the insulin

conformational changes and hinge opening [55].

The radius of gyration (Rg) was measured to examine the dynamic stability and compactness

of protein in the presence of vitamins. A time evolution plot of Rg (backbone) is shown in

Fig. 4c. Several consecutive and continuous fluctuation is seen in Rg trajectory from

beginning of simulation, which is due to major structural changes of protein. This drift in Rg

is maintained up to 50 ns. Thereafter, a constant fluctuation is seen up to 75 ns,a gain in Rg is

observed up 90 ns. The vitamin E changes the gyration radius of insulin due to its influence

on the B-chain hinge conformational transition.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

12

Fig. 4. RMSD and Gyration radius. RMSD of backbone atoms of insulin-vitamin complexes in respect to apo-

insulin for A-chain (a) and B-chain (b). Rg of whole protein with respect to 100 ns molecular dynamics

simulations (c).

Table 2. RMSD profile of protein backbone atoms and Rg profile of whole protein calculated over the course of

100 ns molecular dynamic simulations.

3.3. MM-PBSA analysis

To estimate the strength of interaction between vitamins and insulin and difference in the

binding energy between vitamins, MM-PBSA method was applied for 98 to 100 ns. The

obtained values indicated that the contribution of van der Waals interaction in the total

interaction energy was much larger than the electrostatic energy (Table 3). The positive value

of polar solvation energy indicated a little contribution to vitamin binding with insulin [56].

In order to recognize the hotspot residues of insulin ligand binding, the contribution of every

residue was explored. Per residue binding energy decomposition for insulin with both

vitamins suggested that IleA10, LeuA13, LeuA16, PheB1, GlnB4, LeuB6, AlaB14, LeuB17, ValB18

were the major residues interacting with vitamins (Fig. 5). Per residue energy contributions

Apo-Insulin Insulin-Vitamin D3 Insulin-Vitamin E

A-chain RMSD (nm) 0.12 0.14 0.13

B-chain RMSD (nm) 0.3 0.27 0.5

Rg (nm) 1.03 1.02 1.05

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

13

illustrated that the favorable binding free energy originates predominantly from the nonpolar

terms.

Table 3. Binding energy values (kj/mol) and individual component energy calculated with MM-PBSA method

for insulin with vitamin D3 and vitamin E.

Van der Waal

energy

Electrostatic

energy

Polar solvation

energy SASA energy Binding energy

Insulin-Vitamin D3 -40.413 +/- 5 -0.258 +/- 0.18 9.453 +/- 2.12 -5.233 +/- 0.65 -36.451 +/- 4.14

Insulin-Vitamin E -54.087 +/- 7.1 -0.079 +/- 0.085 14.281 +/- 2.67 -6.099 +/- 0.799 -45.984 +/- 5.41

Fig. 5. Energy decomposition graph for individual amino acid residues majorly contributing to binding

energy. IleA10, LeuA13, LeuA16, PheB1, GlnB4, LeuB6, AlaB14, LeuB17, ValB18 were the major residues interacting

with vitamin E.

3.4. Opening of B-chain C-terminal hinge

There are several strategies for investigation of B-chain C-terminal conformational changes

to study how vitamins can affect insulin structure and function. Distance computation

between B-chain C-terminal and α-helix residues and also calculating angle changes of B-

chain used for hinge conformational transition analysis. We selected three pair residues from

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

14

the B-chain C-terminal and B-chain α-helix for calculating distances of pair residues and

studying conformational changes of B-chain. GlyB8, ValB12 and LeuB15 were selected from α-

helix portion and distances were calculated from 3 C-terminal residues (PheB24 ,TyrB26

,ProB28) for all trajectories. Fig. 6a shows the distances between the Cα atoms of the three

pairs from the MD simulation. These distances provide the most useful indexes for opening

of the B-chain C-terminus portion and hence must pay more attention to it.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

15

Fig. 6. Illustration of the residue pairs for (a) distance calculation and (b) Angle computation. The spheres

represent the Cα atoms of the B-chain α-helix and B-chain C-terminal residues, respectively. (c and e) Up and

lateral view of Insulin-Vitamin E complex, and (d and f) Up and lateral view of Insulin-Vitamin D3 complex

after simulation.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

16

The MD simulation result for the vitamin apo-insulin shows that it mainly remains in the

closed conformation and the B-chain C-terminus portion makes only sporadic excursions

away from the B-chain α-helix (Fig. 7). Fig. 7 shows the stability of hinge in apo-insulin.

Distance calculation in apo-insulin hinge shows distance in all indexes remain under 8.5 Å.

Based on previous experimental and computational results, calculated distance in apo-insulin

B-chain hinge is about 10 Å, so B-chain hinge is in the closed conformation. These findings

were validated by comparing with the recently reported MD simulations by A. Papaioannou,

et al. [21] and corresponding crystal structure 2G4M. The average distances between the Cα

atoms of B-chain C-terminal and B-chain α-helix approbate with the experimental values

within 10 Å (Fig. 7). As displayed in Fig. 1b, the side chains of PheB24, PheB24 and TyrB26

participate in the formation of hydrophobic core, which play the main role of maintaining

insulin in the closed state.

GlyB8-ProB28 and ValB12-TyrB26 distances provide the most useful indicators for opening and

wide-opening of the B-chain C-terminal, respectively. There is a good confirmation between

our results and the experiments, which further validates the MD simulations. Obtained results

show the LeuB15-PheB24 distance of apo-insulin does not change, in accordance with a hinge

behavior, while others increase to varying values. Average distances between the Cα atoms of

the pairs LeuB15- PheB24, ValB12-TyrB26 and ProB28-GlyB8 are approximately 7.5, 8.3 and 6.8 Å

respectively. It is obvious from all the MD simulations that there is no systematic dynamic

behavior or opening mechanism like a switch or periodic motion between two states. The

insulin activation casually occurs at the simulated times and appears to be an accidental event

(Fig. 7).

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

17

Fig 7. Stability of the hinge. Distances between the Cα atoms of the LeuB15-PheB24, ValB12-TyrB26 and GlyB8-

ProB28 pairs, which represent at the hinge, for the apo-insulin.

Time series of the distances between Cα atoms of GlyB8-ProB28 illustrated that open

conformation of B-chain C-terminal hinge is formed affected by vitamin E (Fig. 8a). As it is

obvious from the time series of the distances between Cα atoms of ValB12-TyrB26, increasing

of distance and opening of hinge occurred only in the presence of vitamin E and hence only

vitamin E can induce the wide-open conformation, not vitamin D3 (Fig. 8b). ValB12-TyrB26

average distance increased from 0.83 Å in apo-insulin to 1.34 Å in insulin-Vitamin E that

clearly shows the wide-open conformation.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

18

Fig. 8. Time series of the distances between the Cα atoms of (a) GlyB8-ProB28 , (b) ValB12-TyrB26 and (c)

LeuB15-PheB24. These distances are used as a criterion for the opening of the B-chain hinge and activation of

insulin.

Examination of the time series of the distances of insulin hinge in insulin-Vitamin D3

complex shows that the ValB12-PheB24 and GlyB8-ProB28 distances do not change, so vitamin

D3 could not activate insulin for binding to its receptor. Distances between Cα atoms of

LeuB15-PheB24 shows B-chain hinge most stable in all simulations, illustrated the importance

of β-turn in insulin structure.

How insulin changes conformation to engage the insulin receptor has long been unclear. The

recent structural experiment of insulin and its receptor demonstrated that B-chain protective

hinge of insulin opens to initiate receptor binding procedure [20]. It explained that, on

receptor binding, B-chain C-terminal fragment undergoes reorientation of PheB24 to a 60°

rotation of the PheB25-LysB29 strand away from the insulin core, to expose buried residues to

its receptor.

The angle changes were calculated for the confirmation of distance results and

characterization of the wide-open conformation. To analysis of angle changes in B-chain C-

terminal, GlyB8, GluB21 and LysB29 were selected for structure of open conformation (PheB25-

LysB29). Since wide-open conformation occurs in the B-chain hinge (PheB24-TyrB26), GlyB8,

GluB21 and TyrB26 were selected for the computation of wide-open conformation. In Fig 9(a

and b), representative angle variation throughout the trajectories for all simulations are

shown; four Cα atoms were selected for angle calculation. The positions of these atoms

provide a good representation of the movement of the C-terminal coil relative to the α-helix

and opening of the hinge: GluB21 is situated in the center of the hinge, which is between helix

and coil structure, GlyB08 is situated at the opposite side of helix in front of the coil structure,

LysB29 is situated at the flexible end of C-terminus coil, and TyrB26 is situated in the middle of

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

19

the C-terminus coil and instantly after β-turn, which is the key residue in formation of the

wide-open conformation. Accordingly, Cα atoms of these amino acids move as the hinge

opens.

Fig. 9. Variation of the B-chain hinge angle along the MD simulation trajectories. The angles are measured

between Cα atoms of GlyB8, GluB21 and (a) LysB29, (b) TyrB26. Hydrogen bond of B-chain hinge with A-chain in

(c) apo-insulin and (d) insulin-Vitamin E.

The maximum opening for the apo-insulin in the ValB12-TyrB26 is ~8 Å (Fig. 7), while

surprisingly vitamin E increases the ValB12-TyrB26 distances to ~15 Å (Fig. 8b). The average

opening for the apo-insulin is ~20 degrees (GlyB8, GluB21 and LysB29) (Fig. 9a) and ~30

degrees (GlyB8, GluB21 and TyrB26) (Fig. 9b), while in the presence of vitamin E, the hinge

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

20

opens to ~85 and ~65 degrees, respectively. Reorientation of PheB24 to a 60o rotation of the

PheB25-ThrB30 away from the protein hydrophobic core occurs in wide-open conformation of

insulin [20]. Obtained results reveal that vitamin E bounded to A and B-chain interface

induces conformational changes in the B-chain hinge and obviously opens it.

A contact map is an exclusively useful 2D representation of aminoacid-aminoacid

interactions in protein 3D structure [57]. Intramolecular contact maps were computed as

shown in Fig. 10. The contact maps clearly show the contact changes in the presence of

vitamins. An inspection of this contact maps reveal that distances of PheB25-ThrB30 from

GlnB4-ValB18 increases in the insulin-Vitamin E complex related to the apo-insulin. As can be

clearly seen from distance and angle variation and contact map results, the wide-opening of

the B-chain hinge occurs in the presence of vitamin E. Vitamin D3 does not affect on the

Activation of insulin by hinge opening.

Fig. 10. Contact map results. Mean smallest distance for insulin in the presence of vitamins. Distances of

PheB25-ThrB30 from GlnB4-ValB18 increases in the insulin-Vitamin E complex related to the apo-insulin.

3.5. Secondary structure

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

21

Insulin forms a small helix-rich (44%) globular protein with 30% random coil, 9% β-strands

and 19% turns [58]. The Dictionary Secondary Structure of Proteins (DSSP) analysis

performed to identify the secondary structures of each insulin residue in all MD simulations

(Fig. 11). The DSSP analysis showed that the secondary structure elements are intact in the

apo-insulin (Fig. 11a). As observed in Fig. 11b, the insulin secondary structures began to lost

its helical contents mostly in 3-9 residues at the presence of vitamin D3, which is in

agreement with our previous experimental study [36]. Vitamin D3 destroys helical structure at

GlyA1-CysA8 residues, which is somewhat unstable in apo-insulin. The DSSP analysis also

revealed that vitamin E does not change the helical contents of human insulin during the MD

simulation. As it is shown in Fig. 11, β-bridges (isolated backbone hydrogen bonds) are

present in residues CysA11 and HisB5 of apo-insulin. In insulin-Vitamin D3 complex CysA11

and HisB5 β-bridges deform in the last 20 ns of simulation time. In insulin-vitamin E complex

CysA11 and HisB5 β-bridges fully deform and CysA20 and PheB24 β-bridges form, which are

stable during MD simulation. Vitamin E also destroys CysB19-ArgB22 turn structure by

disrupting hydrogen bonds between TyrA19-PheB25. It seems that TyrA19-PheB25 hydrogen

bond disruption may changethe turn structure and induces hinge conformational changes.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

22

Fig. 11. The time evolution of the secondary structure of human insulin monomer in (a) apo-insulin, (b)

insulin-Vitamin D3, and (c) insulin-Vitamin E. Per residue secondary structure density during the last 10 ns of

trajectory are shown in (d) apo-insulin, (e) insulin-Vitamin D3, and (f) insulin-Vitamin E.

Conclusion:

Vitamin D and E supplementations are conventionally used as a long-term oral intake for

diabetes treatment. In this work, for the first time, we investigated the interaction of vitamin

E and vitamin D3 with human insulin by computational insights. Our binding energy MM-

PBSA calculation results reveal that Vitamins D3 and E bind to the insulin, also vitamin E

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

23

engages to more residues and reaches high binding energy of -46 kj/mol. Our data is in

agreement with our previous experimental results [36]. Per residue energy contributions

illustrated that the favorable binding free energy originates predominantly from the nonpolar

terms. IleA10, LeuA13, LeuA16, PheB1, GlnB4, LeuB6, AlaB14, LeuB17, ValB18 were the major

residues interacting with vitamins. Distance and angle calculation results illustrated that

vitamin E changes the B-chain conformation and it causes the formation of wide-open/active

form of insulin. ValB12-TyrB26 distance provides a useful index for studying the

conformational changes and forming wide-open structure of B-chain hinge. Vitamin E

increases the ValB12-TyrB26 distance from ~8 Å in apo-insulin to ~15 Å. In addition, the hinge

angle rises to ~65° in the presence of vitamin E. Distance and angle histogram show that

wide-open conformation is most stable in insulin-Vitamin E complex. On the other hand,

vitamin D cannot change the hinge conformation, therefore, activation of insulin does not

occur in insulin-Vitamin D3 complex. The DSSP analysis revealed that insulin secondary

structure does not much change in both vitamin-insulin complexes. It seems that vitamin E

increases the helical contents of dimeric insulin, due to its phenolic structure, and it does not

lead to significant changes in the monomeric insulin helical contents.

As a result, our MD simulations propose a model for insulin activation through vitamin E

interaction for therapeutic approaches. In conclusion, these findings provide new insights on

the conformational changes of insulin B-chain hinge that is crucial to enable its binding to the

insulin receptor. Finally, we propose a new approach for activation of insulin without any

gene manipulation or mutagenesis for diabetes treatment.

Appendix A. Supplementary Data:

MD simulations repeated using the Gromacs 2019.2 package and supplementary appendix

has been written as per reviewer’s suggestion.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

24

Acknowledgement:

This research has been supported by Tarbiat Modares Research grant, IG-39708. We thank

Dr. Mehrdad Ravanshad for providing computational facility in Tarbiat Modares computer

center.

References: Uncategorized References

1. Riek, A.E., et al., Vitamin D3 supplementation decreases a unique circulating monocyte

cholesterol pool in patients with type 2 diabetes. The Journal of Steroid Biochemistry and

Molecular Biology, 2018. 177: p. 187-192.

2. Pham, T.-M., et al., The relationship of serum 25-hydroxyvitamin D and insulin resistance

among nondiabetic Canadians: a longitudinal analysis of participants of a preventive health

program. PLoS One, 2015. 10(10): p. e0141081.

3. Hochberg, I., et al., Interaction between the haptoglobin genotype and vitamin E on

cardiovascular disease in diabetes. Current Diabetes Reports, 2017. 17(6): p. 42.

4. Saha, P., et al., Patients with type 2 diabetes have reduced levels of plasma vitamin D. Asian

Journal of Medical Sciences| May-Jun, 2016. 7(3): p. 35.

5. Kelishadi, R., et al., Effects of vitamin D supplementation on insulin resistance and

cardiometabolic risk factors in children with metabolic syndrome: a triple-masked controlled

trial. Jornal de Pediatria, 2014. 90(1): p. 28-34.

6. Talaei, A., M. Mohamadi, and Z. Adgi, The effect of vitamin D on insulin resistance in patients

with type 2 diabetes. Diabetology & Metabolic Syndrome, 2013. 5(1): p. 8.

7. Fang, F., Z. Kang, and C. Wong, Vitamin E tocotrienols improve insulin sensitivity through

activating peroxisome proliferator-activated receptors. Molecular Nutrition & Food

Research, 2010. 54(3): p. 345-352.

8. LeBlanc, E.S., et al., Baseline Characteristics of the Vitamin D and Type 2 Diabetes (D2d)

Study: A Contemporary Prediabetes Cohort That Will Inform Diabetes Prevention Efforts.

Diabetes Care, 2018: p. dc180240.

9. Cadario, F., et al., Can type 1 diabetes progression be halted? Possible role of high dose

vitamin D and omega 3 fatty acids. European Review for Medical & Pharmacological

Sciences, 2017. 21(7): p. 1604-1609.

10. Ley, S., et al., Lower dietary vitamin E intake during the second trimester is associated with

insulin resistance and hyperglycemia later in pregnancy. European Journal of Clinical

Nutrition, 2013. 67(11): p. 1154.

11. Maddaloni, E., et al., Vitamin D and diabetes mellitus, in Vitamin D in Clinical Medicine. 2018,

Karger Publishers. p. 161-176.

12. Mitri, J. and A.G. Pittas, Vitamin D and diabetes. Endocrinology and Metabolism Clinics,

2014. 43(1): p. 205-232.

13. Brange, J. and L. Langkjœr, Insulin structure and stability, in Stability and Characterization of

Protein and Peptide Drugs. 1993, Springer. p. 315-350.

14. Mayer, J.P., F. Zhang, and R.D. DiMarchi, Insulin structure and function. Peptide Science,

2007. 88(5): p. 687-713.

15. Boucher, J., A. Kleinridders, and C.R. Kahn, Insulin receptor signaling in normal and insulin-

resistant states. Cold Spring Harbor Perspectives in Biology, 2014. 6(1): p. a009191.

16. Frankær, C.G., et al., Insulin fibrillation: The influence and coordination of Zn2+. Journal of

Structural Biology, 2017. 199(1): p. 27-38.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

25

17. De Meyts, P., Insulin and its receptor: structure, function and evolution. Bioessays, 2004.

26(12): p. 1351-1362.

18. Haeusler, R.A., T.E. McGraw, and D. Accili, Biochemical and cellular properties of insulin

receptor signalling. Nature Reviews Molecular Cell Biology, 2018. 19(1): p. 31.

19. De Meyts, P., Insulin/receptor binding: the last piece of the puzzle? Bioessays, 2015. 37(4): p.

389-397.

20. Menting, J.G., et al., Protective hinge in insulin opens to enable its receptor engagement.

Proceedings of the National Academy of Sciences, 2014. 111(33): p. E3395-E3404.

21. Papaioannou, A., S. Kuyucak, and Z. Kuncic, Molecular dynamics simulations of insulin:

elucidating the conformational changes that enable its binding. PloS One, 2015. 10(12): p.

e0144058.

22. Menting, J.G., et al., A minimized human insulin-receptor-binding motif revealed in a Conus

geographus venom insulin. Nature Structural & Molecular Biology, 2016. 23(10): p. 916-920.

23. Zˇáková, L., et al., Structural integrity of the B24 site in human insulin is important for

hormone functionality. Journal of Biological Chemistry, 2013: p. jbc. M112. 448050.

24. Abdi, R., et al., Immunomodulation by mesenchymal stem cells: a potential therapeutic

strategy for type 1 diabetes. Diabetes, 2008. 57(7): p. 1759-1767.

25. Mingrone, G., et al., Bariatric–metabolic surgery versus conventional medical treatment in

obese patients with type 2 diabetes: 5 year follow-up of an open-label, single-centre,

randomised controlled trial. The Lancet, 2015. 386(9997): p. 964-973.

26. Paoli, P., et al., The insulin-mimetic effect of Morin: a promising molecule in diabetes

treatment. Biochimica et Biophysica Acta (BBA)-General Subjects, 2013. 1830(4): p. 3102-

3111.

27. Bruni, A., et al., Islet cell transplantation for the treatment of type 1 diabetes: recent

advances and future challenges. Diabetes, Metabolic Syndrome and Obesity: Targets and

Therapy, 2014. 7: p. 211.

28. Hirsch, I.B., Insulin analogues. New England Journal of Medicine, 2005. 352(2): p. 174-183.

29. Ward, C.W. and M.C. Lawrence, Similar but different: ligand-induced activation of the insulin

and epidermal growth factor receptor families. Current Opinion in Structural Biology, 2012.

22(3): p. 360-366.

30. Xu, Q., et al., BPN, a marine-derived PTP1B inhibitor, activates insulin signaling and improves

insulin resistance in C2C12 myotubes. International Journal of Biological Macromolecules,

2018. 106: P. 379-386.

31. Rochette, L., et al., Diabetes, oxidative stress and therapeutic strategies. Biochimica et

Biophysica Acta (BBA)-General Subjects, 2014. 1840(9): p. 2709-2729.

32. Saha, P., et al., Patients with type 2 diabetes have reduced levels of plasma vitamin D3 and

are well correlated with the oxidative stress parameters. Asian Journal of Medical Sciences,

2016. 7(3): p. 35-40.

33. WOLLMER, A., et al., Phenol-promoted structural transformation of insulin in solution.

Biological Chemistry Hoppe-Seyler, 1987. 368(2): p. 903-912.

34. Derewenda, U., et al., Phenol stabilizes more helix in a new symmetrical zinc insulin hexamer.

Nature, 1989. 338(6216): p. 594.

35. Dunn, M.F., Zinc–ligand interactions modulate assembly and stability of the insulin hexamer–

a review. Biometals, 2005. 18(4): p. 295-303.

36. Soleymani, H., et al., Vitamin E induces regular structure and stability of human insulin, more

intense than vitamin D 3. International Journal of Biological Macromolecules, 2016. 93: p.

868-878.

37. Mandal, K., et al., Crystallization of Enantiomerically Pure Proteins from Quasi-Racemic

Mixtures: Structure Determination by X-Ray Diffraction of Isotope-Labeled Ester Insulin and

Human Insulin. ChemBioChem, 2016. 17(5): p. 421-425.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

26

38. Strushkevich, N., et al., Structural analysis of CYP2R1 in complex with vitamin D3. Journal of

Molecular Biology, 2008. 380(1): p. 95-106.

39. dos Santos, J.I., A.M. Soares, and M.R. Fontes, Comparative structural studies on Lys49-

phospholipases A2 from Bothrops genus reveal their myotoxic site. Journal of Structural

Biology, 2009. 167(2): p. 106-116.

40. Trott, O. and A.J. Olson, AutoDock Vina: improving the speed and accuracy of docking with a

new scoring function, efficient optimization, and multithreading. Journal of Computational

Chemistry, 2010. 31(2): p. 455-461.

41. Hess, B., et al., GROMACS 4: algorithms for highly efficient, load-balanced, and scalable

molecular simulation. Journal of Chemical Theory and Computation, 2008. 4(3): p. 435-447.

42. Oostenbrink, C., et al., A biomolecular force field based on the free enthalpy of hydration and

solvation: the GROMOS force-field parameter sets 53A5 and 53A6. Journal of Computational

Chemistry, 2004. 25(13): p. 1656-1676.

43. Schüttelkopf, A.W. and D.M. Van Aalten, PRODRG: a tool for high-throughput

crystallography of protein–ligand complexes. Acta Crystallographica Section D: Biological

Crystallography, 2004. 60(8): p. 1355-1363.

44. Hess, B., et al., LINCS: a linear constraint solver for molecular simulations. Journal of

Computational Chemistry, 1997. 18(12): p. 1463-1472.

45. Darden, T., D. York, and L. Pedersen, Particle mesh Ewald: An N⋅ log (N) method for Ewald

sums in large systems. The Journal of Chemical Physics, 1993. 98(12): p. 10089-10092.

46. DeLano, W.L., The PyMOL molecular graphics system. http://www. pymol. org, 2002.

47. Humphrey, W., A. Dalke, and K. Schulten, VMD: visual molecular dynamics. Journal of

Molecular Graphics, 1996. 14(1): p. 33-38.

48. Wallace, A.C., R.A. Laskowski, and J.M. Thornton, LIGPLOT: a program to generate schematic

diagrams of protein-ligand interactions. Protein Engineering, Design and Selection, 1995.

8(2): p. 127-134.

49. Laskowski, R.A. and M.B. Swindells, LigPlot+: multiple ligand–protein interaction diagrams

for drug discovery. 2011, ACS Publications.

50. Seeliger, D. and B.L. de Groot, Ligand docking and binding site analysis with PyMOL and

Autodock/Vina. Journal of Computer-Aided Molecular Design, 2010. 24(5): p. 417-422.

51. Kumari, R., et al., g_mmpbsa A GROMACS tool for high-throughput MM-PBSA calculations.

Journal of Chemical Information and Modeling, 2014. 54(7): p. 1951-1962.

52. Homeyer, N. and H. Gohlke, Extension of the free energy workflow FEW towards implicit

solvent/implicit membrane MM–PBSA calculations. Biochimica et Biophysica Acta (BBA)-

General Subjects, 2015. 1850(5): p. 972-982.

53. Gao, M. and J. Skolnick, A comprehensive survey of small-molecule binding pockets in

proteins. PLoS Computational Biology, 2013. 9(10): p. e1003302.

54. Aier, I., P.K. Varadwaj, and U. Raj, Structural insights into conformational stability of both

wild-type and mutant EZH2 receptor. Scientific Reports, 2016. 6: p. 34984.

55. Karavassili, F., et al., Structural studies of human insulin cocrystallized with phenol or

resorcinol via powder diffraction. Acta Crystallographica Section D: Biological

Crystallography, 2012. 68(12): p. 1632-1641.

56. Kumar, A., et al., Docking, molecular dynamics, binding energy-MM-PBSA studies of

naphthofuran derivatives to identify potential dual inhibitors against BACE-1 and GSK-3β.

Journal of Biomolecular Structure and Dynamics, 2018(just-accepted): p. 1-37.

57. Toon, A. and G. Williams, A dynamical approach to contact distance based protein structure

determination. Journal of Molecular Graphics and Modelling, 2012. 32: p. 75-81.

58. Hua, Q.-x. and M.A. Weiss, Mechanism of insulin fibrillation the structure of insulin under

amyloidogenic conditions resembles a protein-folding intermediate. Journal of Biological

Chemistry, 2004. 279(20): p. 21449-21460.

MANUSCRIP

T

ACCEPTED

ACCEPTED MANUSCRIPT

Highlights:

• Imperfection of insulin signaling can lead to insulin resistance, which is the hallmark

of diabetes mellitus. Vitamin E and D3 play a crucial role in insulin sensitivity

improvement and diabetes treatment.

• Activation of insulin and binding to its receptor for signaling process initiates via B-

chain C-terminal hinge conformational change.

• The wide-open structure of B-chain hinge is critical for receptor binding. Val12-

Tyr26 distances provide the most useful indicator for wide-open conformation

• Vitamin E effectively activates human insulin monomer via B-chain hinge opening.

Val12-Tyr26 distances of apo-insulin increases from ~8.3 Å to ~13 Å and hinge angle

increases from ~30° to ~65° in the presence of vitamin E.