On the conservation of the slow conformational dynamics within the amino acid kinase family: NAGK the paradigm

Abstract

N-acetyl-L-glutamate kinase (NAGK) is the structural paradigm for examining the catalytic mechanisms and dynamics of amino acid kinase family members. Given that the slow conformational dynamics of the NAGK (at the microseconds time scale or slower) may be rate-limiting, it is of importance to assess the mechanisms of the most cooperative modes of motion intrinsically accessible to this enzyme. Here, we present the results from normal mode analysis using an elastic network model representation, which shows that the conformational mechanisms for substrate binding by NAGK strongly correlate with the intrinsic dynamics of the enzyme in the unbound form. We further analyzed the potential mechanisms of allosteric signalling within NAGK using a Markov model for network communication. Comparative analysis of the dynamics of family members strongly suggests that the low-frequency modes of motion and the associated intramolecular couplings that establish signal transduction are highly conserved among family members, in support of the paradigm sequence-->structure-->dynamics-->function.

Figure 1. Structure of the closed form of NAGK.

The NAGK dimer is complexed with the ATP analogue AMPPNP and the amino acid NAG (PDB code 1GS5) . The substrates ATP and NAG (ball and sticks) bind to the C- (red) and N-domains (blue) respectively. N-terminal domains form the interface between the two monomers. Inset shows a closer view of the NAG lid and both ligands.

Figure 2. Structure and fluctuation dynamics of the open conformation of NAGK.

(A) Color-coded ribbon diagram of NAGK where regions involved in substrate binding are indicated by arrows. All secondary structure elements are highlighted with different colors. (B) Comparison of experimentally observed (blue) and computationally predicted (green) B-factors. The theoretically predicted B-factors are rescaled based on the experimentally observed B-factors averaged over all residues. Experimental data refer to the PDB structure 2WXB (to be published). The range of residues of N and C domains is highlighted. The different parts of the protein have been numbered as follows: (1) β1+αA; (2) β2+αB; (3) β3–β4, the NAG lid; (4) αC+β5; (5) β6–β7; (6) αD+β8; (7) β9–β10;(8) αE; (9) β11+β12–β13+β14; (10) αF+αF–αG; (11) αG+β15; (12) αH+β16. The color code of the numbered parts of the protein is the same in both subunits, and indicated along the upper abscissa.

Figure 3. Mobilities of residues.

Results are presented for the softest three GNM modes. On the left, the ribbon diagrams of NAGK, color-coded according to the mobilities of residues in the respective modes are displayed. The most mobile residues are colored blue, and the most rigid ones, red. The green dots on the diagrams indicate the position of the hinge sites. Note that in mode 2 the hinge sites closely neighbour the residue D162, which is a key catalytic residue.

Figure 4. Comparison with experimental conformational changes.

Cumulative overlaps CO(m) between ANM modes and the experimentally observed deformation between the open and closed forms of NAGK are plotted for subsets of m modes, in the range 1≤m≤50 (see equation (7) in Methods ). We note that the first 3 ANM modes (among 3N-6 = 1542 modes) accessible to the open form (red) yield an overlap of 0.75 with the experimentally observed reconfiguration from open to closed state of the enzyme. In the case of the closed form, the first 3 modes yield an overlap of 0.61. In either case, a small subset of modes intrinsically accessible to the structure attain a cumulative overlap of >0.80, pointing to the pre-disposition of the structure to undergo its functional changes in conformation between the open and closed forms.

Figure 5. Movement along the slowest ANM mode.

Motion of active site residues between open and closed conformers along the 1^st ANM mode accessible to the open form. The position of these residues in different conformations is shown: open conformation (yellow), intermediate positions (green and blue) and closed conformation (atom-colored). (A) Color-coded ribbon diagram for motions along the 1^st mode (generated with the ANM web server and Pymol [64]). (B) Movement of catalytic residues with respect to the ATP analogue. (C) Movement of ATP binding residues with respect to the nucleotide. (D) Movement of NAG binding residues with respect to NAG.

Figure 6. Communication properties of NAGK.

(A) Mean hitting time profile for the open and closed (with and without ligands) forms of NAGK. Vertical lines indicate the positions of catalytic (solid line) and ligand-binding residues (dotted line). Note that catalytic residue tend to occupy minima positions, indicative of their efficient communication properties. (B) Difference map between the contribution to hitting times from cross-correlations () (equation (8) in Methods ) of the open and ligand-bound closed forms. Dashed lines set the boundaries of N- and C- domains and also enclose those pairs of domains that undergo the largest changes in the contribution from cross-correlations upon ligand binding. (C) Mean path lengths for linking different parts of the protein: N- and C-domains (blue), N-domain and catalytic site (red), and C-domain and catalytic site (green). (D) Standard deviation in the mean paths displayed in panel (C).

Figure 7. Communication pathways between catalytic site and ligand-binding residues in EcNAGK.

The communication pathways are represented by the network of residues (each atom is shown as a dot) along the shortest paths of communication between the following residue pairs: R66-L209 (yellow), R66-D162 (orange) and L209-D162 (green). These three cases are representative of the communication between a NAG-binding residue (R66) on the N-domain, a catalytic site residue (D162) and an AMPPNP binding residue on the C-domain (L209). These residues at the endpoints of the pathways are colored by atom name. N- and C- domains are colored blue and red, respectively. These pathways are shown for the three states considered: (A) Open state. (B) Closed state. The pathways R66-L209 and L209-D162 have three residues in common (colored in light green). (C) Closed state with ligands AMPPNP and NAG. The ligands are shown with gray sticks for a better visualization and participate in all three pathways considered in the figure. Note that the ligands directly establish the communication between the pairs of residues considered.

Figure 8. Conservation of the lowest frequency modes of motion among the AAK family members.

Correlation cosines between the ten slowest modes of EcNAGK and those of three other members of the AAK family. (A) Dimeric PfCK. (B) EcNAGK-like dimer of TmNAGK. (C) Monomer of PfUMPK. Ribbon diagrams represent the structure of the corresponding dimers of the AAK family that are compared with EcNAGK. The C- and N-domains are colored green and yellow. In panel B, the N-terminal helix (red) is a key element for the hexamerization of TmNAGK. In panel C, one of the monomers is shadowed to highlight that it has been considered as the coupling environment of the monomer that is being compared with a monomer of EcNAGK.