SAXS experiments were performed at the BL23A SAXS beamline at the TLS of NSRRC, using a monochromatic X-ray beam (λ = 0.828 Å), with an integrated HPLC system of an Agilent-Bio SEC-3 300 Å column (Agilent Technologies, Inc. Santa Clara, CA). Protein samples (44 μM MERS-CoV N and MERS-CoV N:P3 complex prepared by incubating the 44 μM native protein with 440 μM P3) were prepared in a buffer consisting of 50 mM Tris-HCl (pH 8.5) and 150 mM NaCl on ice for 1 h. Then, a 100 μL aliquot was injected into the column at a flow rate of 0.02 mL min–1. After passing through the column, the sample solution was directed into a quartz capillary (2 mm dia.) for subsequent buffer and sample SAXS measurement at 288 K. The sample-to-detector distance of 2.5 m used covered a scattering vector q range of 0.01–0.20 Å–1. Here, q is defined as q = (4π/λ) sin θ, with the scattering angle 2θ. Thirty-six frames were collected for each sample elution with an X-ray frame exposure time of 30 s. Frames of good data overlapping (namely, of low radiation damage effects) were merged for improved data statistics and analyzed to determine initial Rg using PRIMUS (version 3.1). The P(r) distance distribution and Dmax were calculated from the experimental scattering curve using GNOM (version 4.1). An ensemble optimization method (EOM) analysis was performed through the EMBL Hamburg web interface.50 Modeling of the rigid body crystal structure was calculated and generated using CRYSOL (ATSAS Program Suite v. 2.8.2).51 The crystal structures of MERS-CoV NTD (PDB ID: 4UD1)27 and MERS-CoV NTD:P3 (solved in this study) and the CTD domain of MERS-CoV N protein (PDB ID: 6G13)23 were used as rigid bodies in EOM analysis. With the EOM analysis, 1000 models were generated in the beginning as a structural pool. Selected from the SAXS profiles of the structural pool was an ensemble of models that could fit the experimental scattering curve with their linear combination. Tetrameric MERS-CoV NP conformations and 16-mer MERS-CoV:P3 conformations were selected because their ensemble generated curves fit best to the experimental SAXS results.