Insight into Buffalo (Bubalus bubalis) RIG1 and MDA5 Receptors: A Comparative Study on dsRNA Recognition and In-Vitro Antiviral Response

RIG1 and MDA5 have emerged as important intracellular innate pattern recognition receptors that recognize viral RNA and mediate cellular signals controlling Type I interferon (IFN-I) response. Buffalo RIG1 and MDA5 genes were investigated to understand the mechanism of receptor induced antiviral response. Sequence analysis revealed that RIG1 and MDA5 maintain a domain arrangement that is common in mammals. Critical binding site residues of the receptors are evolutionary conserved among mammals. Molecular dynamics simulations suggested that RIG1 and MDA5 follow a similar, if not identical, dsRNA binding pattern that has been previously reported in human. Moreover, binding free energy calculation revealed that MDA5 had a greater affinity towards dsRNA compared to RIG1. Constitutive expressions of RLR genes were ubiquitous in different tissues without being specific to immune organs. Poly I:C stimulation induced elevated expressions of IFN-β and IFN-stimulated genes (ISGs) through interferon regulatory factors (IRFs) mediated pathway in buffalo foetal fibroblast cells. The present study provides crucial insights into the structure and function of RIG1 and MDA5 receptors in buffalo.


Supporting Methods
: PCR cycling parameters used for the amplification of RLR genes.

Preparation of culture media
Complete Dulbecco Modified Eagle's medium (DMEM) was prepared as per manufacturer's guidelines, and supplemented with 25mM HEPES, 10ng/ml EGF and L-Glutamine (2 mM   respectively. The tertiary structures were predicted using advance modeling protocol of MODELLER 9.11 [4]. The resultant models were ranked based on Discrete Optimized Potential Energy (DOPE) score and the models with lowest DOPE scores were selected for further study.

Principal component analysis
Principal component analysis (PCA) was performed by constructing a covariance matrix containing dominant low-frequency, large scale motions of main chain atoms of the amino acid residues along the MD trajectory. The matrix was then diagonalized to give rise eigenvectors and their corresponding eigenvalues. Normally, the first eigenvector represents the largest contribution to the global fluctuation of the system followed by the second eigenvector, and so on.

Binding free energy calculation
The binding free energy (ΔG bind ) was calculated by GMXAPBS tool, which uses Molecular Mechanics /Poisson Boltzmann Surface Area (MM/PBSA) method as follows: Where, G complex , G protein and G ligand are the free energies of complex, protein and ligand, respectively.
The brackets indicate that the average of snapshots was taken from a single MD trajectory.
The free energy of each state was calculated as follows: <G >= <E MM > + <G PB > + <G SA > -T<S MM > T < S MM > is the entropic contribution of the solute, which was not evaluated in this study due to high computational cost and often this can produce unpredictable results.
G PB and G SA are the contributions from polar and nonpolar terms of the free energy of solvent continuum, the former is calculated via Poisson-Bolztmann equation and the latter is calculated as follows: Where, γ is the surface tension proportionality constant, β is the free energy of nonpolar solvation for a point solute and SASA is the solvent accessible surface area. E MM is the molecular mechanical energy, calculated as the sum of different contributions as follows: E vdW , E ele are the van der Waals (LJ) and the electrostatic energies, respectively. E int is the internal energy including bond, angle and torsional angle energies. It is worth noting that in the case of single-trajectory experiments, the variation of E int ( E int ) equals zero in calculating the binding free energy according to eqn.1, since the internal energies of the complex and the separated parts (protein and ligand) are calculated from the same trajectory [8]. Figure S1: Agarose gel electrophoresis of amplified products of different fragments of buffalo RIG1

Supporting Results
and MDA5 genes.  qualities of modeled proteins were found to be highly comparable to those of templates, indicating quality of the models were reasonably good to carry out further studies. For a good quality model the G-factor should be above the cutoff value of -0.5 [9]. e % of residues complementing sequence to structure agreement. For a good quality model the Verify 3D score should be greater than 80% [10].
f Errat score provides accuracy of the nonbonded atoms, and a good quality model must have a score greater than the acceptable value of 50% [6] g ProSA gives Z-score of a given model which should fall within range of protein structures of similar size available in PDB [11].
h ProQ analysis indicates that for a given model LGscore and MaxSub scores must be above 4.0 and 0.4, respectively [12].