Molecular basis for GTP recognition by light-activated guanylate cyclase RhGC

Abstract

Cyclic guanosine 3',5'-monophosphate (cGMP) is an intracellular signalling molecule involved in many sensory and developmental processes. Synthesis of cGMP from GTP is catalysed by guanylate cyclase (GC) in a reaction analogous to cAMP formation by adenylate cyclase (AC). Although detailed structural information is available on the catalytic region of nucleotidyl cyclases (NCs) in various states, these atomic models do not provide a sufficient explanation for the substrate selectivity between GC and AC family members. Detailed structural information on the GC domain in its active conformation is largely missing, and no crystal structure of a GTP-bound wild-type GC domain has been published to date. Here, we describe the crystal structure of the catalytic domain of rhodopsin-GC (RhGC) from Catenaria anguillulae in complex with GTP at 1.7 Å resolution. Our study reveals the organization of a eukaryotic GC domain in its active conformation. We observe that the binding mode of the substrate GTP is similar to that of AC-ATP interaction, although surprisingly not all of the interactions predicted to be responsible for base recognition are present. The structure provides insights into potential mechanisms of substrate discrimination and activity regulation that may be common to all class III purine NCs. DATABASE: Structural data are available in Protein Data Bank database under the accession number 6SIR. ENZYMES: EC4.6.1.2.

Keywords: cGMP; cyclic GMP; guanylate cyclase; guanylyl cyclase; retinylidene photoreceptor.

Figure 1

Characterization of the crystallized construct. (A) Schematic domain organization of the full‐length RhGC (top) and its potential dimeric assembly (bottom). M: membrane. From the N terminus: a‐N: N‐terminal autoinhibitory region. α0: predicted additional transmembrane helix. RhI: type I rhodopsin domain. SH: signalling helix. GC: GC domain. On the bottom panel, one of the monomers is colour‐coded, whereas the second one is shown in grey. (B) SEC‐MALS analysis of the CaGC construct. The chromatogram displays normalized Rayleigh ratio (black) and molar mass calculated by MALS (20 kDa, green). (C) Activity of the CaGC construct in solution in the presence of different divalent cations measured as concentration of generated pyrophosphate. Activity was measured in triplicates. Depicted values represent the mean ± SD.

Figure 2

CaGC·GTP·Ca²⁺ structure overview. (A) Cartoon representation of the CaGC (light green, PDB 6SIR) and CaAC monomer (grey, PDB 5OYH 11). Secondary structure elements are depicted and labelled β1–β8 for the strands and α1–α5 for the helices. N and C terminus are labelled N and C, respectively. (B) Comparison of the tongue regions in CaGC and CaAC. Top: conformation of the β7/β8 tongues differs between CaGC monomers (light and dark green) whereas it is same for all CaAC monomers (grey). Grey and dark green conformations are almost perfectly overlapping. Bottom: conformation of the β4/β5 tongues differs between CaAC monomers within each dimer (both grey), whereas it is distal for all CaAC monomers (light green). (C) Top: comparison of the CaGC dimer (light and dark green) and its ligand‐free form modelled based on the inactive sGC dimer (PDB 6JT1 21). View from the ventral side. Structures were superimposed via the light green monomer shown as single surface. Calcium ions are depicted as green spheres. GTP molecules are shown as sticks. Bottom: CaGC·GTP·Ca²⁺ dimer along its twofold axis viewed from the ventral (left) and dorsal side (middle) represented as solvent accessible surface. Right bottom panel shows cross section of the dimer with the view on one of the substrate binding pockets. (D) Comparison of CaGC·GTP·Ca²⁺ dimer and its model based on the sGC·GMPCPP·2×Mg²⁺ assembly (PDB 6JT2 21). Structures were superimposed via the light green monomers shown as single surface. Second subunit in CaGC·GTP·Ca²⁺ (dark green cartoon) is in almost identical relative orientation as in sGC·GMPCPP·2×Mg²⁺ (grey cartoon). (E) 2F _o–F _c electron density map carved within 2 Å distance around GTP and calcium, displayed at 1.5 σ level. For clarity, second monomer was omitted. Molecular graphics and analyses were performed with the ucsf chimera package 52.

Figure 3

Active site of CaGC·GTP·Ca²⁺ complex. (A) Active site in CaGC·GTP·Ca²⁺ structure. Chains A and B are in light and dark green, respectively. The calcium ion is shown as a green sphere. The GTP, residues crucial for substrate binding and turnover are shown as sticks. Water molecules are shown as red spheres. The metal coordination and selected hydrogen bonds between protein, triphosphate and ribose are shown as black dashed lines. (B) Close‐up view of the protein–guanine base interaction in the CaGC·GTP·Ca²⁺ structure. (C) From left to right: close‐up view of the protein–base interaction in CaGC·ATPαS·Ca²⁺ (PDB 5OYH 11, sGC·GMPCPP·2×Mg²⁺ (PDB 6JT2 21), Ma1120_CHD·ATP·Ca²⁺ (PDB 5D15 10 and Ma1120_CHD ^{( KDA → EGY )} ·GTP·Ca²⁺ (PDB 5D0G 10) structures. Introduced mutations are marked in red. In all panels, maximum acceptor–donor distance is 3.5 Å and colour coding as in A. Molecular graphics and analyses were performed with the ucsf chimera package 52.

Figure 4

GTP–CaGC interaction in one of the active sites. Residues involved in hydrophobic contacts are shown as a curved comb. Ionic and hydrogen bond interactions (maximal donor–acceptor distance = 3.5 Å) are shown as dark green dotted lines. Calcium and waters are represented by light green and light blue spheres, respectively. Figure generated using ligplot+ 53.

Figure 5

Alignment of sequences of cyclases: Catenaria anguillulae CaGC (PDB 6SIR, this work), Blastocladiella emersonii BeGC (PBD 6AOB 20), Mycobacterium avium Ma1120 (PDB 5D15 10), Synechocystis sp. Cya2 (PDB 2W01 16), Chlamydomonas reinhardtii GC (PDB 3ET6 15). Amino acids that coordinate to the catalytical ion (●), phosphate (▼), ribose (■) or base (*) are marked. Positions marked with red (*) are mutated in CaAC. Positions marked with red (*) and with red (■) are mutated in Ma1120_CHD ^{( KDA → EGY )} mutant. Alignment was created using clustal omega 54 and jalview 55.

Figure 6

Hypothetical model of the CaGC active site in the presence of two catalytic Mg²⁺ cations. Model was based on the Ma1120_CHD·ATP·Ca²⁺ crystal structure (PDB 5D15) 10. Magnesium ions are shown as blue spheres. Selected hydrogen bonds between the protein and GTP are shown as black dashed lines. Colour coding is as in Fig. 3. Molecular graphics and analyses were performed with the ucsf chimera package 52.