Stamp


The crystal structure of the catalytic domain of a eukaryotic guanylate cyclase


Jonathan A. Winger, Emily R. Derbyshire, Meindert H. Lamers, Michael A. Marletta and John Kuriyan


BMC Structural Biology 2008, 8:42 doi:10.1186/1472-6807-8-42 (local copy)

Abstract / Figures from the paper / Coordinates / Supplementary Information


Abstract:

Background

Soluble guanylate cyclases generate cyclic GMP when bound to nitric oxide, thereby linking nitric oxide levels to the control of processes such as vascular homeostasis and neurotransmission. The guanylate cyclase catalytic module, for which no structure has been determined at present, is a class III nucleotide cyclase domain that is also found in mammalian membrane-bound guanylate and adenylate cyclases.

Results

We have determined the crystal structure of the catalytic domain of a soluble guanylate cyclase from the green algae Chlamydomonas reinhardtii at 2.55 Å resolution, and show that it is a dimeric molecule.

Conclusion

Comparison of the structure of the guanylate cyclase domain with the known structures of adenylate cyclases confirms the close similarity in architecture between these two enzymes, as expected from their sequence similarity. The comparison also suggests that the crystallized guanylate cyclase is in an inactive conformation, and the structure provides indications as to how activation might occur. We demonstrate that the two active sites in the dimer exhibit positive cooperativity, with a Hill coefficient of ~1.5. Positive cooperativity has also been observed in the homodimeric mammalian membrane-bound guanylate cyclases. The structure described here provides a reliable model for functional analysis of mammalian guanylate cyclases, which are closely related in sequence.

Illustrations from the paper.

Click on the small image to get a bigger one.


Figure 1. Structural alignment of selected guanylate and adenylate cyclase catalytic domains. Secondary structure annotation and numbering correspond to the guanylate cyclase homolog CYG12 from C. reinhardtii. Sequences are grouped as follows: A, atypical soluble guanylate cyclases; B, membrane-bound guanylate cyclases; C, NO-sensing soluble guanylate cyclases; D, putative bacterial guanylate cyclases; E, mammalian and bacterial adenylate cyclases. Functional residues are indicated by symbols: metal binding (*); ribose binding (Δ); guanine/adenine binding (◆); triphosphate binding (#). Accession numbers are as follows: Chlamydomonas reinhardtii CYG12 (GenBank XP_001700847), Caenorhabditis elegans GCY35 (GenBank O02298), Rattus norvegicus sGCβ2 (GenBank BAB68564), Drosophila melanogaster GYC-88E (GenBank Q8INF0), Homo sapiens RetGC1 (GenBank Q02846), R. norvegicus GCA (GenBank P18910), C. elegans GCY7 (GenBank AAQ62451), Strongylocentrotus purpuratus mGC (GenBank P16065), R. norvegicus sGCβ1 (GenBank BAC55087), zias latipes sGCβ1 (GenBank BAA76691), duca sextaCβ1 (GenBank AAC61264), melanogasterβ1 (GenBank), norvegicus sGCα1 (GenBank AAB17953), O. latipes sGCα1 (GenBank BAA76690), M. sexta sGCα1 (GenBank AAC61263), D. melanogaster sGCα1 (GenBank AAF56917), Anabaena sp. PCC7120 all1118 (GenBank NP_485161), Nostoc punctiforme PCC73102 NpR1313 (GenBank YP_001864972), N. punctiforme PCC73102 NpR0352 (GenBank ACC79135), Trichodesmium erythraeum IMS101 Tery_4585 (GenBank ABG53561), T. erythraeum IMS101 Tery_3412 (GenBank ABG52512), Synechocystis sp. PCC6803 sll0646 (GenBank BAA16969), Canis familiaris ACV_C1 (GenBank 1CJU_A), R. norvegicus ACII_C2 (GenBank 1CJU_B), Mycobacterium tuberculosis Rv1264 (GenBank 1Y11_A), M. tuberculosis Rv1900c (GenBank 1YBU_C), Spirulina platensis CyaC (GenBank 1WC1_C). Initial alignments were carried out using the program ClustalX [67]. Sequences were adjusted manually with comparison to results from a structural homology search using the DALI server [45]. Figure 1 was prepared using the program ESPRIPT [68]. Regions containing residues of > 70% equivalence (red letters) are boxed with a thin blue line, and absolutely conserved residues are highlighted in red.



Figure 2. Structural features of the guanylate cyclase domain. A) Structural representation of a guanylate cyclase domain monomer. Elements of secondary structure are labeled according to the nomenclature depicted in Figure 1. B) The guanylate cyclase catalytic domain. Monomer A is colored green, and monomer B is multi-colored, ranging from blue at the N-terminus to red at the C-terminus. Two of eight phosphate ions are shown and are depicted as stick figures: phosphorus, orange; oxygen, red.



Figure 3. Comparison between guanylate and adenylate cyclase active sites. Monomer B of the guanylate cyclase catalytic domain was superimposed onto the C2 domain of mammalian adenylate cyclase (PDB ID: 1CJU) [49]. Residues and structural elements involved in catalysis and nucleotide recognition are shown. A) Comparison of guanylate and adenylate cyclase catalytic residues. B) Comparison of guanylate and adenylate cyclase base recognition residues. Side chains and structural elements from the guanylate cyclase and adenylate cyclase catalytic domains are colored green and grey, respectively. The nucleotide 2',3'-dideoxyadenosine triphosphate (ddATP) and the two Mg2+ ions are from 1CJU. Non-carbon atoms are colored as follows: phosphorus, orange; oxygen, red; nitrogen, blue; sulfur, yellow; arsenic, violet; magnesium, white.




Figure 4. Proposed guanylate cyclase activation mechanism. Comparison of helix α1 in the guanylate cyclase structure with helix α1 of the active mammalian adenylate cyclase structure indicates that the guanylate cyclase structure is in an inactive state. Monomer B of the guanylate cyclase structure was superimposed onto the C2 domain of the active adenylate cyclase structure (PDB ID: 1CJU) [49] and monomer A was superimposed onto the C1 domain of the same adenylate cyclase structure. The C2 domain is omitted for clarity. The C1 domain of the active adenylate cyclase structure is colored blue, and the guanylate cyclase structure is colored green. The nucleotide 2',3'-dideoxyadenosine triphosphate (ddATP) and Mg2+ ions from 1CJU are shown as a stick figure and spheres: phosphorus, orange; oxygen, red; nitrogen, blue; magnesium, white.




Figure 5. Communication between active sites. Plot of guanylate cyclase activity at increasing concentrations of substrate GTP. Guanylate cyclase (5 μg) was incubated for 2 min at 24 °C with the indicated concentrations of GTP in the presence of 4 mM MnCl2 and cGMP was measured. Data were fit to the equation (Vmax(S)n)/((S0.5)n+Sn)), where Vmax is the maximum activity, S is the concentration of GTP, S0.5 is the substrate concentration at which half-maximal velocity is reached, and n is the Hill coefficient. From the fit, Vmax = 2795 ± 117 nmoles cGMP/min/mg, S0.5 = 269 ± 26 μM, and n = 1.49 ± 0.16. A Hill coefficient greater than 1 indicates the presence of interacting active sites.




Figure 6. A potential binding site for a regulatory control element. A) Structure of the mammalian adenylate cyclase catalytic domain bound to the activator Gsα (PDB ID 1CJU) [49]. The switch II helix of Gsα binds in a groove on the C2 domain between the α1–α2 and α3-β4a loops, priming the catalytic domain for nucleotide binding. A surface representation of the adenylate cyclase catalytic domain is shown, and Gsα is shown as a ribbon cartoon. The C1 domain is colored blue, the C2 domain is colored orange, and Gsα is colored teal. B) Surface representation of the guanylate cyclase catalytic domain in the same orientation as the adenylate cyclase domain in A. A groove similar to that used by adenylate cyclase to bind to Gsα is located between the α1–α2 and α3-β4a loops, and may serve as a site for interaction of control elements with the guanylate cyclase catalytic domain. Monomer A is colored teal, and monomer B is colored green.




Coordinates


Coordinates in the Protein Data Bank: 3ET6

Supplementary Information



Additional file 1. Dimethylarsenic cysteine modifications. Five cysteine residues are modified through a reaction with the sodium cacodylate and dithiothreitol in the crystallization buffer, resulting in addition of dimethylarsenic to the cysteine thiols. Experimental anomalous difference density contoured at 5 σ is shown superimposed onto the refined guanylate cyclase structure. Sidechains are show as stick figures and colored as follows: carbon, green; sulfur, yellow; arsenic, violet.




Additional file 2. Activation mechanism of mammalian adenylate cyclase. Helix α1 in the C1 domain of mammalian adenylate cyclase undergoes a significant conformational change going from an inactive structure to an active structure. A hypothetical inactive structure of the mammalian adenylate cyclase catalytic domain was generated as previously described [35]: the C2 domain structure from 1AZS was superimposed on chain A from the structure of a C2 domain homodimer (PDB ID: 1AB8) [41], and the C1 domain structure from 1AZS was superimposed on chain B from 1AB8. The nucleotide- and Gsα-bound active C1/C2 dimer structure 1CJU [49] was then superimposed onto the C2 domain of the inactive model. The inactive model is colored white, and the active structure is colored blue. Gsα and the C2 domain of 1CJU are omitted for clarity. The nucleotide 2',3'-dideoxyadenosine triphosphate (ddATP) and Mg2+ ions from 1CJU are shown as a stick figure and spheres: phosphorus, orange; oxygen, red; nitrogen, blue; magnesium, white.