Backbone Dynamics of the Ribonuclease Binase Active Site Area Using Multinuclear ( 15 N and 13 CO) NMR Relaxation and Computational Molecular Dynamics † Yuxi Pang, ‡,| Matthias Buck, §,⊥ and Erik R. P. Zuiderweg* ,‡ Department of Biological Chemistry, UniVersity of Michigan, Biophysics Research DiVision, 930 North UniVersity AVenue, Ann Arbor, Michigan 48109-1055 ReceiVed August 14, 2001; ReVised Manuscript ReceiVed December 13, 2001 ABSTRACT: The nano-pico second backbone dynamics of the ribonuclease binase, homologous to barnase, is investigated with 15 N, 13 C NMR relaxation at 11.74 and 18.78 T and with a 1.1 ns molecular dynamics simulation. The data are compared with the temperature factors reported for the X-ray structure of this enzyme. The molecular dynamics and X-ray data correspond well and predict motions in the loops 56- 61 and 99-104 that contain residues that specifically recognize substrate and are catalytic (His101), respectively. In contrast, the 15 N relaxation data indicate that these loops are mostly ordered at the nano- pico second time scale. Nano-pico second motions in the recognition loop 56-61 are evident from 13 CO- 13 CR cross relaxation data, but the mobility of the catalytic loop 99-104 is not detected by 13 CO cross relaxation either. From the results of this and previous work [Wang, L., Pang, Y., Holder, T., Brender, J. R., Kurochkin, A., and Zuiderweg, E. R. P. (2001) Proc. Natl. Acad. Sci. U.S.A., 98, 7684-7689], the following dynamical characterization of the active site area of binase emerges: a beta sheet, rigid at all probed time scales, supports the catalytic residue Glu 72. Both substrate-encapsulating loops are mobile on both fast and slow time scales, but the fast motions of the loop which contains the other catalytic residue, His 101, as predicted by B-factors and computational molecular dynamics is not detected by NMR relaxation. This work strongly argues for the use of several measures in the study of protein dynamics. The characterization of protein dynamics may help to correlate structural properties with biological activities (1-3). Enzyme active sites often contain protein loops, which are by nature dynamic. Part of such a site may need to sequester a substrate, or some times even isolate it from the solvent into the protein interior (4-10). In some cases, the sequestering loops contain catalytic residues (11-13). Here we focus on the study of the dynamical behavior of the active-site area of the ribonuclease binase, using nuclear magnetic resonance (NMR) 1 spectroscopy and simulated molecular dynamics (MD). NMR spectroscopy is a powerful tool to study protein dynamics at multiple atomic-sites covering a wide range of time-scales (14-17). Molecular dynamics simulations have recently become sufficiently efficient to probe similar time scales as the NMR relaxation and can thus be used to help interpret and complement the NMR methods (18-20). Binase is a guanyl-specific extracellular ribonuclease secreted by Bacillus intermedius, which catalyzes the endo- nucleotic cleavage of single-stranded RNA (21). It consists of a single polypeptide chain (109 amino acid residues) that is folded into a R+ compact structure as determined by X-ray crystallography and NMR spectroscopy (22-24). The tertiary structure of binase is virtually superimposable with that of the well-studied enzyme barnase, a homologue with 85% sequence identity (25, 26). As shown in Figure 1, three R helices are located in the N terminal part of binase, while five antiparallel  strands are in the C terminal part. Of the two catalytic residues, Glu 72 is located on the “bottom” of the active site cleft on strand 2, while His 101 is perched on the loop L5 connecting strands 4 and 5. Loop L2, formed by residues 56-67 connecting the first and second  sheet, is responsible for the binding of the guanine group of the substrate. The interactions between loop L2 and the substrate, in particular with residues Phe 55, Arg 58, and Glu 59, contribute to the majority of the enzyme’s specificity (21, 27). The crystal structure of the homologous enzyme barnase with the inhibitor dCGAG suggests that the substrate in this class of enzymes is also interacting with Lys 26 of helix R2, with Tyr102 on the loop connecting strands 4 and 5, and with Arg 86 on loop L3 (binase count) (28). † Supported by Grant MCB 9814431 of the National Science Foundation. * To whom correspondence should be addressed. Phone: (734) 936- 3850. Fax: (734) 764 3323. E-mail: zuiderwe@umich.edu. ‡ Biophysics Research Division, Department of Biological Chemistry and Department of Chemistry, The University of Michigan, 930 N University Avenue, Ann Arbor, MI 48109. § Department of Chemistry and Chemical Biology, Harvard Uni- versity, 12 Oxford Street, Cambridge, MA 02139. || Present address: Laboratory of Functional and Molecular Imaging, 10 Center Drive, 10/B1D118, NINDS/NIH, Bethesda, MD 20892. ⊥ Present address: Memorial Sloan-Kettering Cancer Center, 1275 York Avenue, New York, NY 10021. 1 Abbreviations: NMR, nuclear magnetic resonance; MD, molecular dynamics; CSA, chemical shift anisotropy; CPMG, Carr-Purcell- Meiboom-Gill; RMSD, root-mean-square displacement; RMSF, root- mean-square fluctuation; R1, longitudinal spin relaxation rate; R2, transverse spin relaxation rate; R1F, rotating frame spin relaxation rate; ηxy, transverse cross correlation rate; NOE, nuclear Overhauser effect. 2655 Biochemistry 2002, 41, 2655-2666 10.1021/bi011657f CCC: $22.00 © 2002 American Chemical Society Published on Web 01/30/2002