MD simulations can be used as a reply to many types of issues. Here we used MD simulations to determine the mechanism of transport Fatty acid (pathway and rate) CD36 receptor. Figure 1b illustrates an overview result of the applications of Molecular Dynamics Simulations, the timescale and the molecular motions accessible with the current simulation methods.
Structural Studies of Mutational Effects on Fatty Acid Transport CD36 (K164A)
The Three-dimensional (3D) Structure of CD36 Mutant (K164A)
The tridimensional structures of virtual mutant K164A ectodomain of CD36 predicted by using Swiss Model server to construct the homology model based on the crystal structure CD36-CIDRa (PDB ID: 5LGD). The quality of the homology model was evaluated by Ramachandran plot, the overall quality factor ERRAT2 plot and PROVE produced by the SAVS server v6.0 (https://saves.mbi.ucla.edu/). The stereo chemical properties of the modeled demonstrates that a majority of its residues are in favored region (92.2%) and allowed region (7.8%) of the Ramachandran plot, that indicates reasonable quality of this homology model. Moreover, the ERRAT2 score was 95.918 and the total quality index (Z-score) was 0.352. We observed that the currently mutant modeled structure have structural qualities that are acceptable24 Fig. 1c.
Global Molecular Dynamics of the Simulations Analysis
Because protein flexibility represents a significant function in CD36 transport and signaling, we conducted MD simulations of 100 ns on wild type in absence of PLM, in addition to the Wild type and mutant type in the presence of PLM for further analysis. Also to study the link of Lys164 residues with PLM transport and binding, we first examined the interaction of CD36 wild type crystallized ecto-domain structure with (PDB Id: 5LGD) and models that were generated for the mutant CD36 in the presence of PLM as well as alone CD36 wild type using MD simulation. After 100 ps the structure of CD36w with PLM, revealed that PLM is located above Lys164 and near the upper helix. The room in the midst of group of carboxylic group of PLM and the Lys164 is 1.2 Å whereas the mutant residue Ala164 in helix I is 5.5 Å. Clearly, this shows that the mutant residue Ala164 is far from the binding site of CD36 (Fig. 1b) and it won’t instantly link with the PLM. Indeed, the mutation instantly impact the binding of PLM and hence, the K164A mutation has control of the overall conformation particularly, helix I, where the mutant amino acid K164A is existing. This is a pattern of the transportation channel of CD36 receptor25.
We calculated and evaluated the three simulation systems: the structure stability of the proteins in the time of the MD and RMSD/backbone, which were considered respective to the first conformation, as we had observed in the RMSD representation Fig. 2a, The plots of RMSD reflected equilibrium of the simulations that were signal of the global protein constancy. The peak RMSDs were at 0.25 nm to 0.55 nm for the three simulation systems. The plots were depicted in three sets. The first graph corresponded to CD36 wild-type only in absence of PLM, other showing the same for CD36 WT and MT in presence of PLM were illustrated separately with black, blue and red bars correspondingly. It is evident that the time taken for both the CD36-WT and CD36-MT in presence of PLM to reach convergence was more than their corresponding CD36r WT in absence PLM. The fluctuation size had a small difference in the medium RMSD value, which had a huge constancy from 38 ns. It was a dynamic constant comportment of the CD36r WT. The wild-type in presence of PLM attained early stability due to convergence from 43 ns whereas the mutated pair reached equilibrium much later.
The RMSF of C-alpha atoms was used to infer with the residue specific flexibility in order to study the fluctuation of every single residue. In other words, during MD simulation three RMSF plots for CD36 wild-type in absence of PLM, the CD36 wild-type and CD36m mutant in presence of PLM were generated Fig. 2,a. For the bleu curve, a trimodal distribution at the regions close to the individual portal was observed for the complex CD36-PLM wild type compared to their wild-type only black curve. The elevation of RMSF at the upper helix domain and coil was supposed to be produced by the amino acid lys 164 impacts of first assays to enter from these portals, which surround the bottom of the cavity. Interestingly, CD36 mutant red curve, exhibited elevated flexibility at N-terminal region (from 50 to 160) and C-terminal at 400 end did not displayed a lot of displacement at the binding region Lys164 .The difference in fluctuation patterns CD36 could be attributed to their subtle importance of Lys164 residues and structural features. The flexibility of the upper part of the helix/loop is important for the organization of the unlocking and locking of the path. The representation of the RMSF showed that the conformational order occurred in the mutant. The utmost modification has taken place in the loop and upper helix region. Clearly, the access of PLM from this portal requires protein conformational change in upper helix from Lys164, which might explain why it was the most preferred route in our multiple independent runs. To conclude the RMSF plot confirmed clearly that in the helix regions of the upper loop the fluctuations of the residues were more articulated. Since the radius of gyration (Rg) is a very powerful system in analyzing the configurational compactness, the form, and the coil of the global WT, WT-PLM and MT-PLM conformation at several period; we used it to executed Rg examination in order to look for the structural variations and dynamic stability of the wild type and mutate (K164A) structures in the presence of PLM. The average Rg values of WT, WT-PLM and MT-PLM from the trajectory were calculated as 2.2, 2.21 and 2.18 nm respectively see Fig. 2,a. Throughout the simulation, the Wild type CD36 and complex wild type CD36-PLM are marked with black and blue lines respectively, exhibiting almost a similar pattern in terms of the Rg values on the contrary the complex Mutant Type CD36m-PLM which was marked with red line; This suggested that there was a change in the global conformation and coil of the preceptor after the mutation. The K164A mutation shows a significant decrease in Rg, implying a strong compactness that will influence the closure of the fatty acid transport cavity.
Moreover, in order to search for the hydrophobic center behavior of the wild type CD36 (WT) and mutant type (MT) structures in presence and absence of PLM, we had investigated the solvent surface accessibility area (SASA). The analysis of SASA displayed the residues of the complex MT-PLM structures CD36 (K634A) which had the same SASA estimation to the complex WT-PLM CD36 protein, and they both conserved the openability at 44 ns Fig. 1a and upper to WT in the absence of PLM. PLM made this amino acid more accessible to the solvent. An important decrease of the SASA mean was verified in the complex WT CD36-PLM. The CD36 K164A mutation which has brought about larger changes in the SASA value because it had a great surface area exposed to the solvent, this could cause an exposure of hydrophobic amino acids and subsequently the deployment of the receptor.
A calculation of hydrogen bond (HB) in CD36 WT, WT-PLM and MT-PLM mutants was given to interpret the stability and flexibility of the receptor. Figure 2,a shows the quantity of intra molecular HB in the WT and mutant CD36 receptor in presence and absence of PLM throughout the simulation. The total of HB decreased in complex CD36 WT-PLM and MT-PLM mutants as compared to the WT at 40 ns. WT CD36 has 340 HB, whereas complex WT-PLM and MT-PLM have 320 HBs, respectively. This diminution in HB was considered a loss of conformation and compactness of the CD36 receptor in presence of PLM. Overall, a significant diminution MT-PLM after a simulation time of 40 ns disrupts the structure, dynamics, and the stability, which leads to a deregulated and dysfunctional CD36-PLM complex. Throughout the period of simulation, the energetic value of the complex WT-PLM is an indication of strong basis of the fact that the molecule has a stable structure as well as relatively higher amount of total interaction compared to complex MT-PLM. Destabilization in folding and thermodynamic stability may affect the total energy of the biomolecules. The results demonstrated a very slight rise in the total energy value in mutant structure this ruled out the notion of stability losses induced by mutation.
The secondary structural propensities become significant in exploring the structural properties of the protein. However, there were significant structural variations measured at each simulation time. These outcomes were consistent with the secondary structure evaluations with the DSSP utilizing the Fig. 2,b. Therefore, this gave us a clear evidence that the secondary and tertiary constructing (a-helix and b-sheets) of the CD36 WT-PLM and MT-PLM structures were not stable during the simulation. The mutation K164A induced change in the A-helix content but it delayed at 70 ns on the contrary in WT-PLM at 40 ns. Our MD analysis has clearly showed that K164 change conformational, dynamics, and flexibility might be a possible cause of transport of fatty acid.
Principal component analysis PCA,
We use the Principal component analysis (PCA) in order to identify the motions and the direction of the flexibilities in the residues. The principal components analysis was applied to the backbone atoms for CD36 WT, WT-PLM and MT-PLM such analysis showed that the first account 25%, 56% and 79% of variances of the total variance observed in the simulation data respectively Fig. 3a.
The first outstanding featured motions during the simulation were analyzed through secondary structures which includes an opening-closing motion of transport tunnel. Our observation focused on the movements of the C-coil, A-helix. We found, that the flexible motions of theses secondary structures in the distinct active states are quite dissimilar from CD36 wild type (WT) in absence of PLM and complexes CD36 WT-PLM and MT-PLM Fig. 3a,b,c. The top principal components for model PC1 for CD36 WT in absence PLM, WT-PLM and MT-PLM are illustrated in Fig. 3,b. In the top two eigenvectors, the most prominent motions observed are propeller loops of CD36 MT-PLM which are the most flexible in moving in the opposite direction. Particularly, in holo structure the propeller loops attached to the upper, moves leftwards or rightwards for CD36WT and it moves in the opposite direction upwards to downwards for the CD36 MT in presence of PLM. In contrast, the helix and coil at termination K164 has an expanding or contracting movements in the WT in the presence of PLM, which is the opposite in the WT only in absence of PLM. Inclusively, the mutation results in a change in the movement patterns of these secondary elements. We have visualized this movement of the top two eigenvectors Fig. 3,c: they show the orientation and the volume of selected eigenvectors by each of the backbone atoms. Although, there might be differences between the simulations with respect to the movement sampled, it is evident that this may suggest that the site K164 is able to settle the paradigm by reducing the movement of the loops to a reduce module.
Covariance Analysis and Motions
For a preferable comprehension of the association through the movement of the upper helix, coil and Lys164 residue, as well as the movements of the adjoining part, covariance analyses were executed during each of the separate simulation on the three groups CD36 WT, WT-PLM and MT-PLM. They were illustrated utilizing 3D system plots. As shown in Fig. 3,d, positive correlations are mapped in light blue green color while negative correlations are mapped in pink color. The deeper color indicates more correlations (or anti-correlations).
We detected a cluster with correlated motions in the middle the secondary structures and the adjoining zone in the active situation of the related groups. The correlated motions of these parts grow extremely in the CD36MT-PLM state Fig. 3,e correlation and anticorrelation resented in red and blue color line. However, the mutation K164A increased the anticorrelated motions between the upper helix and C-terminal and N terminal (tunnel). In addition, strongly correlated motions were also observed within the beta sheets regions and helix C-terminal and N-terminal regions, which signal that the conjunction of these parts retired and proposed that the flexibility of the helix coil is in charge for the movements of the adjoining pathways. Besides, the movements of these two secondary structures are correlated with the movement of upper helix that is essential for the structural stability of CD36 structure for PLM transport.
In the next parts, we revealed cavities in the several conformations during the trajectory for the six MD. A more functional study has indicate the elaborated organizational connection in the midst the motions of the upper helix, and coils, along with the movements of the adjoining conduits. The channels in holo conformations of CD36 crystal structure.
Transport Channel Analysis
The channels in holo conformations of CD36 crystal structure with two co-crystal of PLM14 were identified and analyzed using CAVER (Fig. 4,a) in order to determine the associated motion between upper helix and the transport channel. To examine mutation effect K164A and active area of transport channel bottleneck amino acids were geted by study the Channel Extraction and Visualization (CHEXVIS) during MD simulations this observed the conformation constituents Table 1.
CHEXVIS was used to study Scenes from 100 ns MD for CD36WT, CD36WT-PLM and CD36MT-PLM; Fig. 4,b represents tunnel width, length, and bottleneck width after 50 ns and 100 ns. The width indicats the extreme flexible domain for fatty acid, the bottleneck can elucidate the hardness of absorption of fatty acid from receptor hole. A reduce bottleneck showed in CD36MT-PLM complex signals the problem of the receptor in absorption and transporting the fatty acid. From the above results, it can be concluded that K164A influence helix conformation and the ability of the site closing and binding and the transport of PLM. The transport channel of the fatty acid in the apo and holo conformations were identified based on the PCA analysis using CHEXVIS for CD36WT, CD36WT-PLM and CD36MT-PLM Table 1. The width channel of holo structure has enlarged, that demonstrate the unlocking of the channel, as indicate in Fig. 4,c,IV,V,I (The bottle-neck 4.059, 4.243,3.592 and Length 59.978,55.923,62.427 for CD36 WT, WT-PLM and MT-PLM channel) In contrast, the channel in the apo conformation showed a decline in the width of channel especially at bottle-neck level of, as shown in Fig. 4,c,I,II,II).The bottle-neck 3.746, 4.264,3.754 and Length 48.392,59.430,75.344 for WT, WT-PLM and MT-PLM channel respectively. In addition, only the width of the channel for CD36 MT-PLM of holo form was very small in bottle-neck Fig. 4,c,VI remains in the inactive state correspond to an inaccessible pathway for substrate. It also it indicate that the mutation influences the conduit unlocking-locking by modifying the movement of the upper helix and loops. However, an issue appears related the way the mutation influences the movement of the helix and loops.
Table 1
Cavity's profile as seen by the ChExVis program
Simulations | Rank | Pt 1 | Pt 2 | Length | Bottle-neck | Straightness | Score |
Wild CD36 (holo form) | 1 | 2 | 7 | 59.978 | 4.059 | 0.903 | 0.844 |
Wild CD36 - PLM (holo form) | 1 | 3 | 4 | 55.923 | 4.243 | 0.876 | 0.852 |
CD36m - PLM (holo form) | 1 | 1 | 9 | 62.427 | 3.592 | 0.821 | 0.795 |
Wild CD36 (apo form) | 1 | 3 | 10 | 48.392 | 3.746 | 0.858 | 0.826 |
Wild CD36 - PLM (apo form) | 1 | 4 | 6 | 59.430 | 4.264 | 0.895 | 0.835 |
CD36m - PA (apo form) | 1 | 5 | 7 | 75.344 | 3.754 | 0.769 | 0.868 |
Analysis of Properties of Helix
To determine the associated motion between the Lys164 residue and C-coil and the surrounding channels, a precedent research proposed that the upper helix plays a function in stability between the unlocking and locking structures of CD36. The K164A mutation based on the PCA analysis, both the coil and the helix of the wild-types and mutants switch in contrary orientations in apo and holo form Fig. 5,a. In order to provide the detailed view of the observed transitions, we have analyzed the MD trajectories for CD36 WT-PLM and MT-PLM complexes by calculating the average geometrical parameters of upper helix using Simulaid software. K164 residues, therefore, directly contribute to its global helix that was turned towards the right of Y axis Fig. 5,b,I, while the mutation A164 mobility turned towards the left of Y and Z axis Fig. 5,b,II. Lysine side chains, allowed the detection of their rotation in our simulations. As a result, it retains the helix rotation, local helix litl and turn helix per residue.
To study the mechanism of the rotation in our simulations of the mutation on the upper helix of CD36 proteins, we concentrated on the structure analysis of upper helix using Bendix analyses it analyzed static peptide's local angles along the length of the helix showed in Fig. 5c,d. The result of the analyses of static peptide's local angles along the length of the helix had 152–164 residue and the helical angles were measured with 6.5 curvature degree for WT-PLM Fig. 5,c in comparison with MT-PLM Fig. 5,d the length of the helix 152–165 residue. And the helical angles were measured with 8 curvature degree. On the whole, our investigation of upper helix proposes that the upper helix at the MT-PLM mutate situation of CD36 receptor tends to be wrapped, that influences the movement of the helix and loops of opening fatty acid transport site as revealed by the covariance analysis.
Portals PLM Transport and Pathway
In the X-ray structure of CD36- complex, the first PLM occupied the channel connecting the helical portal region and the inner cavity thus14, if the helical portal is the only choice to enter the cavity, does it exist an alternative portal that might allow the direct exchange between PLM? But there was no direct evidence showing the feasibility of such a process. In order to evaluate the feasibility of the structural dynamics underlying this hypothesis was directly applied to only PLM in ligated protein complex. A successful penetration of PLM from principal portal was observed Fig. 6.
Figure 6 presents snapshots at 100ps, 800ps, 15ns, 21ns, 23ns, 25ns, 100ns, 200ns presenting the pathway of PLM transport at the CD36 taken during a simulation, where one palmitic acid (PLM) were randomly located in the simulation box together with the crystal CD36 (PDB code : 5LGD) and records the RMSD calculated for the wholes system as it evolves with time. At about 100ps after the initiation of the simulation Fig. 6,a the PLM hydrophobic tail joined the Lys164 residue first one colored blue, due to hydrophobic and the electrostatic interaction to attach to the protein. 800ps later, the tight interaction between PLM and Lys164, stimulated the conformation change caused by partially penetrating from the portal region initially. Then, it tried to slide over a loop dissociated from the gap between a loop and a-helix Fig. 6,a (red circle) by the carboxylic part of a PLM molecule in the inner cavity of the protein. 15–150 ns later the PLM had accessed the inner channel and finished its way ted till the termination of the simulation. During the 200ns/time, the FA had approached the central inner cavity protein. Residues, which were commonly encountered by the Palmitic acid during expulsion, were important for the constitution of individual portals. On the starting protein structure of our MD simulations, the constituting residues, identifying these residues, that are highlighted Fig. 6,a are Arg63,Phe65,leu140,Ala143,Ala144,Ile148,Leu161,Ser162,Lys164,Leu187,Leu188,Leu189,Pro191,TYR192,Pro193,Val194,Thr195,Val198,Thr197,Gly199,Leu200,Phe201,Pro203,Tyr204,Thr207,Ser268,Ser269,Asp270,Ile271,Arg273,Ile330,Lys334,Glu335,Gly336,Arg337,Pro338,Ser339,Val339,Lys385, Leu387, Val389, Leu343, Ile341, Thr369, Leu391,Leu416, Glu418. Small openings at the portal regions were closed in the starting structure, however, certain parts of portal (e.g. Arg337, Thr195, Val194 at loop and Ala144, Ile148 at upper helical) showing the potentiality of creating an opening at this region. Since the PLM molecule bind in to Arg337 was close to this portal, so it would be the most favorable route. This portal was the most recorded one in multiple repetitions of MD simulations; as a result this portal is believed to be the only choice for Arg337. We therefore conducted MD simulations of 200 ns on wild type in presence of one PLM as shown in the RMSD profile Fig. 6,b. The plots of RMSD revealed to be stable during simulation. The result indicates that Lys164 residue which provides the structural basis of forming an opening at the region of the principal portal for the dissociation of palmitic acid molecules from this path.
Fatty Acid Transport Rates
Here we described the outcome of MD simulations, using high concentrations of palmitate molecules. In the X-ray structure of CD36 complex, two PLM crystallized as described elsewhere occupied the channel connecting the portal area14. Consequently, there were no hypothetical proposals of how it would absorb from the receptor and even the maximal rates of palmitic uptake. To determine the rate of the fatty acid transported via the CD36 receptor and it existence other intermediate portals were entry portals using the same MD parameters with those for one PLM. Figure 7,a presents snapshots at 100ps, 50,100, 150, and 200 ns taken during each simulation, which depicted the pathway, transport mode and rate of three PLM to the portal CD36 receptor. In this simulation, 200 ns long, a crystal human CD36 (PDB code 5LGD) was equilibrated with three palmitic acid (PLM) that were placed in the simulation box at random locations outside. Figure 7,b records the RMSD, RMSF, Rg calculated for the whole systems for three PLM outside and three PLM (two inside and one outside) as it evolves with time simulation. At the beginning we started with simulation (5 ns) Fig. 7,a, the first PLM approximated to Lys164 after, a second PLM molecule continued and the third molecule moved away of the protein range electrostatic attraction of one of the carboxylates to the positive charge of the Lys164 amine side chain. 100 ns later, the first PLM moved towards the central portal zone due to hydrophobic reaction and second continued and one single PLM into the interior hole of the receptor. The third PLM had stayed in a liberate structure until the stop of the simulation. So, it formed an dimmer in conflictn connected to the receptor. During this time, the liberate PLM had approached the receptor a shortly, but none of them led to the creation of a stable structure. To ascertain that the encounter between the protein and the two PLM was not accidental; we repeated the simulation using the toad crystal structure (PDB Id: 5LGD) and the pre-formed of two PLM inside and one outside molecules. We found out that the PLM outside was not absorbed till the simulation was terminated at 40 ns. As seen in Fig. 7 after the results of the dynamics of the system with three PLMs that the protein had transported two PLMs one by one to confirm these results we determined that the encounter between the protein and the two PLM was not accidental. Indeed, the simulation was repeated using the toad the same crystal structure and a pre-formed of two PLM inside and one outside the molecules. The PLM outside was dissociated till the simulation was terminated at 18 ns. Another system had to be carried out to put two PLMs inside and one outside after the simulation of the third PLM always remained outside and according to the RMSD, RMSF and radius of gyration (Rg) analysis it was always stable by contribution to the first simulation three PLM ext. The fluctuations of protein structure variation, in the time of the absorption from the three PLM, was observed in Fig. 7,b (The green and pink curves respectively).
The principal structure fluctuation, as expected, occurred at the proximity of the specific portal. For the green line, the augmentation of RMSF at the upper helix region and coil was thought to be affected by the amino acids impacts of the starting trys to exit from this area. The domain with the growing RMSF, immediately participated by the ligand entrance from the gateway, K164 helix I loop, surrounding the bottom of the cavity. The outside absorption of the three PLM from this gateway had imposed a structure conformational modification contrariwise of the system two PLM one inside and one outside. Clearly, the two PLM are the most preferred to CD36 structure stabilization. Our CD36 structure model therefore supported a model in which fatty acids move through the cavity, identifying just one main possible entrance and favorable positions through which translocating ligands moved as they went towards the membrane proximal exit and the plasma membrane.