Elevated μs-ms timescale backbone dynamics in the transition state analog form of arginine kinase

Abstract

Arginine kinase catalyzes reversible phosphoryl transfer between arginine and ATP. Crystal structures of arginine kinase in an open, substrate-free form and closed, transition state analog (TSA) complex indicate that the enzyme undergoes substantial domain and loop rearrangements required for substrate binding, catalysis, and product release. Nuclear magnetic resonance (NMR) has shown that substrate-free arginine kinase is rigid on the ps-ns timescale (average S²=0.84±0.08) yet quite dynamic on the µs-ms timescale (35 residues with R_ex, 12%), and that movements of the N-terminal domain and the loop comprising residues I182-G209 are rate-limiting on catalysis. Here, NMR of the TSA-bound enzyme shows similar rigidity on the ps-ns timescale (average S²=0.91±0.05) and substantially increased μs-ms timescale dynamics (77 residues; 22%). Many of the residues displaying μs-ms dynamics in NMR Carr-Purcell-Meiboom-Gill (CPMG) ¹⁵N backbone relaxation dispersion experiments of the TSA complex are also dynamic in substrate-free enzyme. However, the presence of additional dynamic residues in the TSA-bound form suggests that dynamics extend through much of the C-terminal domain, which indicates that in the closed form, a larger fraction of the protein takes part in conformational transitions to the excited state(s). Conformational exchange rate constants (k_ex) of the TSA complex are all approximately 2500s^-1, higher than any observed in the substrate-free enzyme (800-1900s^-1). Elevated μs-ms timescale protein dynamics in the TSA-bound enzyme is more consistent with recently postulated catalytic networks involving multiple interconnected states at each step of the reaction, rather than a classical single stabilized transition state.

Keywords: Arginine kinase; Dynamics; NMR; Relaxation dispersion; Transition state analog.

Figure 1

Structures and NMR spectra of the transition state analog (TSA) complex of arginine kinase. (A) The crystal structure of the TSA complex is overlaid on the substrate-free form, shown with transparent hues. Substrates are shown in stick model. The protein model is colored by dynamic domain, with blue, red, green, orange, and cyan corresponding to domains 1-5 (Niu et al., 2011). Residues which cannot be assigned to a dynamic domain are colored grey. (B) Chemical structure of the transition state analog bound to arginine kinase: an unreactive nitrate is held with non-covalent enzyme interactions between ADP and arginine in the active site, closely mimicking calculated structures with a γ-phosphoryl group bridging between nucleotide and guanidinium with the partial covalent bonding of the transition state (Zhou et al., 1998). The underlying enzyme structure is shown with high transparency. A stereo version is available as supplemental figure 4. (C)2D [¹⁵N,¹H]-TROSY of arginine kinase bound to the transition state analog. (D) Boxed-out area of panel C shown as an overlay of 2D [¹⁵N,¹H]-TROSY of arginine kinase in the substrate-free form (blue) on that of the TSA form (red). An overlay of the full spectra is shown in supplemental figure 5.

Figure 2

Generalized order parameters obtained for transition state analog bound arginine kinase. Secondary structure is noted above each plot with cylinders and arrows representing α-helices and β-strands, respectively.

Figure 3

Arginine kinase TSA complex ¹⁵N spin relaxation dispersion curves. For selected residues with μs-ms timescale dynamics , R₂^eff is plotted as a function of ν_CPMG for ν₀(¹⁵N) = 60.78 MHz (red squares) and ν₀(¹⁵N) = 81.04 MHz (blue circles). Lines represent two-state collective fits of each cluster of residues with R_ex: D71 (NTD motion), H185 (loop L8), F211 (dynamic domain 2), I230 (dynamic domain 4), and D293 (dynamic domain 5).

Figure 4

Conformational dynamics of transition state analog bound arginine kinase. The crystal structure of the closed (TSA) form of arginine kinase, with substrates shown in stick model, is colored by dynamic domain: domain 1 (blue), domain 2 (red), domain 3 (green), domain 4 (orange), and domain 5 (cyan) and shown as stereoscopic pairs. In grey are residues that are not assigned to a dynamic domain. Mapped onto the model are residues with R_ex measured by ¹⁵N spin relaxation dispersion, shown as spheres. They are located primarily at the NTD-CTD interface (approximated as a dashed line) and throughout the CTD. Shown are views looking down into the active site (top), with loop L8 and dynamic domains 2 and 3 in front (middle), and with dynamic domains 4 and 5 in front (bottom).