Published: November 29, 2011 r2011 American Chemical Society 886 dx.doi.org/10.1021/jp208767s | J. Phys. Chem. B 2012, 116, 886–894 ARTICLE pubs.acs.org/JPCB SRLS Analysis of 15 N Spin Relaxation from E. coli Ribonuclease HI: The Tensorial Perspective Eva Meirovitch,* ,† Yury E. Shapiro, † Mirco Zerbetto, ‡ and Antonino Polimeno ‡ † The Mina and Everard Goodman Faculty of Life Sciences, Bar-Ilan University, Ramat-Gan 52900, Israel ‡ Department of Chemistry, University of Padua, 35131 Padua, Italy 1. INTRODUCTION Ribonuclease HI (RNase H, EC 3.1.26.4) is an endonuclease that hydrolyzes the RNA strand in RNADNA hybrid mole- cules. 14 Escherichia coli RNase H participates in DNA replica- tion by defining the origin of replication in ColEl plasmids; 57 inhibiting replication from genomic sites other than oriC; 8,9 and removing RNA oligonucleotides from Okazaki fragments during lagging strand synthesis. 10 Retroviral reverse transcriptase (RT) contains a C-terminal RNase H domain that has been shown to be involved in at least three processes essential for reverse transcription. 11 Escherichia coli RNase consists of a single polypeptide chain comprising 155 amino acid residues. The three-dimensional structure of this enzyme was determined by X-ray crystallog- raphy 1214 and NMR. 15 RNase H is a α/β protein with five α-helices denoted α A to α E (residues 43 to 58, 71 to 80, 81 to 88, 100 to 112, and 127 to 142), and a five-stranded β-sheet, com- prising the strands denoted β 1 to β 5 (residues 4 to 13, 18 to 27, 32 to 42, 64 to 69, and 115 to 120). 12 The polypeptide chain segment containing residues 90 to 99 has been termed the handle region, 12 while the overlapping region from residues 81 to 99 has been termed the basic protrusion. 14 The backbone structure of RNase H is illustrated in Figure 1. 17 The secondary structure of the protein in solution agrees with its crystallographic counterpart. 18 Results obtained from model building 12,14,19 and NMR spectroscopy 19,20 are consistent with the presumed locations of the catalytic and binding sites. The loops between β 1 and β 2 , α C and α D , and β 5 and α E have been suggested to participate in substrate binding. RNase H has been studied extensively with NMR spectros- copy. 11,15,17,2127 Of particular interest to the present study are model-free (MF) analyses of autocorrelated relaxation param- eters, 15 N T 1 , T 2 and 15 N{ 1 H} NOE, 11,17,26,27 and cross- correlated relaxation parameters, η xy and η z , 25 associated with backbone amide bonds. MF is a simple method valid in the limit where the local motion of the NH bond and the global motion of the protein may be considered statistically independent (decoupled), and the tensorial properties are very simple. MF features analytical spectral densities (that underlie the expressions for the relaxation parameters) based on mathematical considerations. 2830 References 11, 17, and 25 describe careful MF analyses of good experimental data obtained from an RNase molecule considered spherically symmetric. Nevertheless, several signi ficant inconsisten- cies emerged. For example, the temperature-dependence of the generalized MF order parameter, S, was used to calculate a char- acteristic temperature, T*, for the motion of the backbone NH Received: September 11, 2011 Revised: November 29, 2011 ABSTRACT: 15 NH relaxation parameters from ribonuclease HI (RNase H), acquired in previous work at magnetic fields of 14.1 and 18.8 T, and at 300 K, are analyzed with the mode-coupling slowly relaxing local structure (SRLS) approach. In accordance with standard theoretical treatments of restricted motions, SRLS approaches NH bond dynamics from a tensorial perspective. As shown previously, a physically adequate description of this phenomenon has to account for the asymmetry of the local spatial restrictions. So far, we used rhombic local ordering tensors; this is straightforward but computationally demanding. Here, we propose substantiating the asymmetry of the local spatial restrictions in terms of tilted axial local ordering (S) and local diffusion (D 2 ) tensors. Although less straightforward, this description provides physically sound structural and dynamic information and is efficient computationally. We find that the local order parameter, S 0 2 , is on average 0.89 (0.84, and may be as small as 0.6) for the secondary structure elements (loops). The main local ordering axis deviates from the C i1 α C i α axis by less than 6°. At 300 K, D 2,^ is virtually the same as the global diffusion rate, D 1 = 1.8  10 7 s 1 . The correlation time 1/6D 2, ) ranges from 3125 (208344) ps for the secondary structure elements (loops) and is on average 125 ps for the C-terminal segment. The main local diffusion axis deviates from the NH bond by less than 2° (10°) for the secondary structure elements (loops). An effective data-fitting protocol, which leads in most cases to unambiguous results with limited uncertainty, has been devised. A physically sound and computationally effective methodology for analyzing 15 N relaxation in proteins, that provides a new picture of NH bond structural dynamics in proteins, has been set forth.