Structural insights into the Middle East respiratory syndrome coronavirus 4a protein and its dsRNA binding mechanism
Maria Batool, Masaud Shah, Mahesh Chandra Patra, Dhanusha Yesudhas, and Sangdun Choi* Department of Molecular Science and Technology, Ajou University, Suwon, 16499, South Korea
*Corresponding Author Sangdun Choi, PhD Department of Molecular Science and Technology Ajou University, Suwon, 16499, Korea Phone: +82-31-219-2600 Fax: +82-31-219-1615 E-mail: [email protected]
Supplementary Table S1 | Comparative validation of p4a 3D models. This table summarizes the validation scores of p4a models obtained from different servers/tools. Ramachandran plot values were obtained from RAMPAGE server, ERRAT quality factor values were obtained from NIH server, Z-scores were obtained from ProSA-web, and QMEAN scores were obtained from QMEAN server. The model obtained from I-TASSER showed the best results among all.
No. of residues
Ramachandran Plot values (%)
ERRAT Quality Factor
Favored region: 90 Allowed region: 9 Outliers: 1
Favored region: 87.5 Allowed region: 6.9 Outliers: 5.6
Favored region: 83.8 Allowed region: 13.2 Outliers: 2.9
Favored region: 78.3 Allowed region: 15.9 Outliers: 5.8
Favored region: 86.4 Allowed region: 6.8 Outliers: 6.8
*Z-score: Z-score indicates the overall model quality using C-alpha atoms ** QMEAN: Qualitative Model Energy Analysis (Models of low quality are expected to have strongly negative QMEAN Zscores)
Supplementary Table S2 | Alanine scanning mutagenesis. The amino acid residues present at the p4a-dsRNA interacting interface were selected and mutated into alanine using protein design package distributed in MOE. The relative binding affinities and stabilities (dAffinities and dStabilities) of the mutant to wild-type residues were calculated using LowModeMD ensemble. The LowModeMD Search method generates conformations using ~1 ps run of molecular dynamics (MD) at constant temperature (300 K) followed by an all-atom energy minimization. The more positive dAffinity value represents the importance of wild-type residue at that particular position. Mutations with negative value indicates the relatively less importance of the wild-type residues. The top five mutations include the two experimentally reported mutations, K67A and K63A. Besides the already reported residues, other amino acid residues suggested in the table below (K27A, W45A, and N8A) might also be crucial in p4a-dsRNA interaction. The individual role of these tope five residues has been elaborated in the main text.
Mutation K27A W45A N8A K67A K63A T38A Y3A Q12A S38A G37A
dAffinity 10.109 5.611 3.359 0.578 0.225 -0.716 -1.785 -1.994 -3.168 -5.048
dStability 0.147 5.605 0.523 -0.202 -0.709 0.195 1.912 0.733 -0.043 -2.583
The unit for dAffinity and dStability is kcal/mol
Supplem mentary Fiigure S1 | C Complex off MERS-CooV p4a and d dsRNA. T The bindingg interface of p4a-ddsRNA com mplex obtainned from HE EX docking server. Thee hotspot ressidues N8, K K27, W45, K63, annd K67 are hhighlighted.. These residdues are loccated in heliix α1, helix α2 and loopp 2 of p4a and playy crucial rolles in p4a-dssRNA interaaction.
mentary Fiigure S2 | The root m mean squarre deviation n (RMSD) of MERS--CoV p4a Supplem variantts’ backbon ne atoms du uring three independeent MD sim mulations. Three T MD siimulations were peerformed foor each system using different d inittial velocities. Black, rred, and greeen colors represennt RMSDs ffrom first, second, s and third MD ttrajectory, rrespectively. The RMSD D plots of each sett of p4a vaariants convverged towaard the end of simulatiion, except K63A that exhibited relativelly higher baackbone devviations after 100 ns.
mentary Fiigure S3 | The T Radiuss of gyratioon (Rg) ploots of MER RS-CoV p4aa variants Supplem during three indeependent M MD simulattions. Blackk, red, and green colorrs representt Rg from Rg of all ME ERS-CoV p44a-dsRNA ccomplexes first, seccond, and thhird MD trajjectory, resppectively. R oscillateed around 200 Å during MD simulattion.
mentary Fiigure S4 | R Residue coontact map of p4a prootein. The contact mapp of wildSupplem type andd mutant reesidues highhlights the m movements at the residdue level. Thhe upper haalf of each correlatiion plot reepresents thhe contact map of w wild-type prrotein, wheereas the loower half represennts the mutaant complexx. Comparattively, K27A A mutant exxhibited a diistinctive coontact map; this suggests that K K27 could bee a crucial fo for p4a-dsRN NA complexx stability.
Supplem mentary Fiigure S5 | Hydrogen H b bond (h-bond) analysees of p4a-d dsRNA com mplex as a function n of time. ((a) Variationn in h-bond distances beetween wildd-type p4a’ss hotspot ressidues and dsRNA nucleotidess. (b) Changge in the tottal number of h-bonds between p44a and dsRN NA during MD sim mulation. A decrease inn the numbeer of h-bondds between pp4aW45A muutant and dssRNA was observed. The correesponding ccolor codes hhave been pprovided at the t bottom of o each figuure. 8
Supplem mentary Fiigure S6 | Principal ccomponent analysis (P PCA) of p44a variants.. 2D plots 9
represent the projection of motions of p4a and its variants by plotting the first three eigenvectors obtained from the last 100 ns trajectories. Color scale tracks the movement of eigenvectors during the trajectory from blue to red. The percentage of principal motions represented by their corresponding eigenvectors are shown in the lower half of each panel. ~80% of the motions correspond to the first 5-6 eigenvectors in wild type as well as mutant p4a.
LO OOP 2
Supplem mentary Fiigure S7 | P Porcupine p plots of K227A and W45A W complexes. Porcuupine plot represennts amplitudde and direcction of the most dominnant motionns in p4a andd bound dsR RNA. The directionn and magnnitude of moovements off loop 2 andd dsRNA caan be clearlyy seen from m the angle and lenggths of the spikes, resppectively. (aa) In K27A,, dsRNA mooves in oppposite directtion of the loop 2. (b) In W455A, loop 2 eexhibits a m more prominnent movem ment in the oopposing diirection of α2 tend to move towaard dsRNA.. This sugggests that dsRNA.. However, helix α1 aand helix α dsRNA does not move in conccert with thee fluctuatingg loop 2.