PerspectivesBiochemistry

Allosteric Regulation of Nucleotidyl Cyclases: An Emerging Pharmacological Target

See allHide authors and affiliations

Science Signaling  04 Sep 2012:
Vol. 5, Issue 240, pp. pe37
DOI: 10.1126/scisignal.2003466

Abstract

Adenylyl cyclases (ACs) and guanylyl cyclases (GCs) produce the second messengers cyclic adenosine monophosphate and cyclic guanosine monophosphate, respectively. ACs and GCs are differentially regulated by intercellular signaling molecules and are implicated in various disease states, including cardiovascular diseases, aging, pain, and neuropsychiatric ailments. Hence, ACs and GCs constitute interesting drug targets. Because the catalytic sites of these enzymes are highly conserved, it is difficult to achieve isoform specificity. However, studies have provided evidence for the notion that regulatory allosteric domains in the vicinity of the catalytic site provide new opportunities for pharmacological intervention. Here, we summarize the current status of such research and discuss future directions in this exciting field.

Nucleotidyl cyclases (NCs) catalyze the conversion of 5′-nucleoside triphosphates (NTPs) into the corresponding 3′,5′-cyclic nucleoside monophosphates (cNMPs), which act as second messenger molecules (14). NCs are divided into adenylyl cyclases (ACs) and guanylyl cyclases (GCs). ACs and GCs exist as either soluble (s) or particulate (p, also known as membranous) isoforms (Table 1). ACs catalyze the formation of cyclic adenosine monophosphate (cAMP), whereas GCs catalyze the formation of cyclic guanosine monophosphate (cGMP). ACs and GCs constitute interesting pharmacological targets because these enzymes are involved in various important diseases, including cardiovascular and metabolic diseases, aging, pain, and neuropsychiatric ailments (49). Depending on the specific disease, either activators or inhibitors of NCs may be therapeutically useful. Presently, the nitric oxide (NO)–releasing drugs, which activate sGC, have the highest clinical relevance. “Activators” and “stimulators” of sGC represent two distinct classes of regulatory ligands, and they act through hitherto unidentified allosteric sites (8). These compounds have emerged as promising drug classes for the treatment of systemic and pulmonary hypertension (8). pGCs are involved in cardiovascular function (7), bone growth (9), meiosis (10), and lipolysis (11). Synthetic natriuretic peptides [such as atrial natriuretic peptide (ANP)], which activate pGC-A, are currently under investigation for the treatment of heart failure (7). However, natriuretic peptides have to be injected intravenously and are rapidly eliminated by proteases and clearance receptors. Moreover, potent competitive inhibitors of mAC and sGC have been developed, but the high conservation of the catalytic sites of NCs renders it very difficult to develop isoform-specific NC inhibitors (4). Membranous ACs (mACs) possess an allosteric regulatory site in a position opposite to that of the catalytic site, which has structural similarities to the catalytic site and binds the plant diterpene forskolin (12). Forskolin activates all mAC isoforms except for AC9, but it has been difficult to develop mAC isoform–specific forskolin derivatives (4, 13). Hence, overall, research activity in the field of allosteric NC regulators has been low, in marked contrast to that in the field of heterotrimeric guanine nucleotide–binding protein (G protein)–coupled receptors (GPCRs), which has witnessed an explosion of research on allosteric regulators (14, 15).

Table 1

Overview of the structure, function, and pharmacological modulation of mAC, sGC, and pGC. mACs possess two membranous domains (M1 and M2) and two cytosolic domains (C1 and C2). Physiologically, mACs are activated by Gs-proteins after activation of GPCRs (not shown here). sGC consists of an α subunit and a β subunit and is activated by NO. NO binds to heme (shown as a chemical structure), resulting in a conformational change in the protein and, hence, enzyme activation. pGC-A is a dimeric protein that binds to ANP through its extracellular domain. Through a propagated conformational change, catalysis is activated. Despite substantial structural and regulatory differences, a common theme of these NCs is that they all possess an allosteric regulatory site near the catalytic site. mACs use ATP as substrate, whereas sGC has a much broader substrate specificity. pGC uses GTP and ITP as substrates. All NCs can be inhibited by various nucleotides. To this end, mACs and sGC have been studied in much greater detail than pGC. 2,4,6-trinitrophenyl (TNP)– and 2′(3′)-N-methylanthraniloyl (MANT)–substituted NTPs are very potent competitive inhibitors of mAC and sGC. P-site inhibitors inhibit mACs noncompetitively. The best-characterized allosteric mAC activators are the diterpenes, which are derived from forskolin. Some diterpenes can also act as allosteric mAC inhibitors. sGC possesses two allosteric regulatory sites; one targeted by nonnucleotides, such as YC1 and BAY 41-2272, and the other targeted by nucleotides. Note that some nucleotides, such as GTP and ITP, stimulate sGC, whereas other nucleotides, such as ATP and XTP, inhibit sGC. Only allosteric nucleotide activators of pGC have been identified thus far, but it is likely that inhibitors will be found. On the basis of results of gene knockout and overexpression studies as well as (patho)physiological studies, numerous clinical applications of allosteric NC modulators can be envisaged. So far, the development of nonnucleotide allosteric ligands for sGC is clinically most advanced. Systematic and rigorous analysis of the catalytic and allosteric sites of NCs is certainly warranted and bears great potential for future drug development.

CREDIT: ADAPTED BY Y. HAMMOND/SCIENCE SIGNALING

However, advances in the NC field are about to change this situation. It has been known for a long time that adenosine 5′-triphosphate (ATP) increases the activity of pGCs through an unknown mechanism (16). Robinson and Potter (17), in an elegant series of biochemical and mutagenesis studies, provide compelling evidence for the notion that pGCs possesses a regulatory allosteric nucleotide-binding site close to the catalytic site. In the presence of ATP, atrial natriuretic peptides dramatically increase the maximum initial velocity (Vmax) and decrease the Michaelis constant (Km) of pGC-A, resulting in enhanced cGMP production with high amplification of the signal. Intriguingly, initial studies indicate that there are differences in the structure-activity relationships (SARs) of nucleotides that bind to the catalytic and regulatory sites of pGC-A. Specifically, both GTP and ATP bind to the allosteric site, whereas 2′-dNTPs do not (17). In marked contrast, 3′-dATP is an effective activator of pGC, suggesting that the position of the 2′-ribosyl hydroxyl group is critical for interaction with the protein. This is an indication of the crucial importance of a precisely located hydrogen bond between the nucleotide and the protein for allosteric activation. By analogy, allosteric activation of mAC by forskolin is also regulated by a defined hydrogen bond between the ligand and the protein (12), which points to a highly conserved allosteric regulatory mechanism among NCs (18). Notably, 2′-dGTP is an effective substrate for pGC (19) and sGC (20). By analogy, 2′-dATP is a substrate of AC (21). Adenosine diphosphate (ADP) binds to the allosteric 5′-site of pGC-A, too, but adenosine does not (17). Thus, these adenosine 5′-monophosphate (AMP) initial studies with a few nucleotides already indicate that it is, in principle, possible to discriminate between the catalytic and the allosteric sites. Based on the literature, it is clear that allosteric regulatory sites are present in most, if not all, pGCs (3).

Allosteric regulation by NTPs is not restricted to pGCs but is also found for sGC. Previous studies showed that ATP inhibits the catalytic activity of sGC through an allosteric mechanism (22). Moreover, a second allosteric binding site in sGC, distinct from the ATP-binding site, has been postulated. This site binds to YC-1 [3-(5′-hydroxymethyl-2′-furyl)-1-benzylindazole], a NO-independent activator of sGC (22). We have recently confirmed the ATP-binding site and, moreover, found guanosine 5′-triphosphate (GTP) and inosine 5′ triphosphate (ITP) to be allosteric enhancers of the catalytic activity of sGC (23). Strikingly, allosteric regulation of sGC by NTPs is dependent on the cation present. Specifically, allosteric enhancement of catalysis by GTP and ITP is observed only in the presence of Mn2+ but not in the presence of Mg2+, whereas allosteric inhibition by ATP is independent of the specific cation present. Xanthosine 5′-triphosphate (XTP) is an allosteric sGC inhibitor only in the presence of Mn2+. These data also imply that binding of a specific cation to sGC constitutes a powerful regulatory mechanism. Unfortunately, to this end, we do not know which cation (Mn2+, Mg2+, or either one in alteration) is present in NCs in vivo. Generally, it is assumed that Mg2+ is physiologically relevant because it is much more abundant than Mn2+, but formal proof is lacking (23). Robinson and Potter (17) did not study the effects of Mn2+ versus Mg2+ on allosteric regulation of pGC-A by NTPs but conducted their studies exclusively in the presence of Mg2+. Hence, future analysis of pGCs in the presence of Mn2+ may yield surprises. Conversely, it is not yet known how 2’-dNTPs and 3′-dNTPs interact with the catalytic and allosteric sites in sGC.

The effects of divalent cations on allosteric regulation of NCs by NTPs is conceivable, given that cations form ion bond networks between the polyphosphate chain and the protein (24). However, the effect of cations on allosteric NC regulation goes well beyond NTPs. Particularly, the presence of Mg2+ rather than Mn2+ exhibited an unexpected effect on the SARs for diterpenes with respect to regulation of mACs 1, 2, and 5. Most impressively, in the presence of Mg2+, the bulky diterpene BODIPY (dipyrromethene boron difluoride)–forskolin exhibits a much larger inhibitory effect on AC2 activity than it does in the presence of Mn2+ (13). Because divalent cations do not form direct contacts between diterpenes and mAC (25), these data may point to the existence of another, as yet unidentified, regulatory cation-binding site or a propagated conformational change from the two cations bound to the nucleotide in the catalytic site of mAC (26). In addition, the existence of a specific regulatory cation-binding site in sGC and pGC cannot be excluded. Moreover, the presence of a regulatory nucleotide-binding domain in mACs cannot be completely dismissed. Rather, this issue has not yet been studied rigorously. A major reason for this lack of knowledge is the fact that the classic radioisotopic AC assay used in standard experiments precludes analysis of low-affinity regulatory sites because of the lack of sensitivity of the assay at higher nucleotide concentrations (radioisotope dilution) and overloading of the alumina columns separating cNMPs from NTPs (23). These technical problems are avoided by the application of high-performance liquid chromatography coupled to tandem mass spectrometry methods (23). To complicate matters even further, mACs contain an allosteric nucleotide-binding site within the catalytic site; that is, nucleoside 3′-phosphates rather than nucleoside 5′-phosphates act as allosteric (noncompetitive) inhibitors or so-called P-site inhibitors (24). To this end, P-site inhibitors have not revealed striking AC isoform specificity (4, 27), which is not surprising in view of the high conservation of the catalytic site (1, 25).

Considering that ATP is present at millimolar concentrations in cells (28), allosteric regulation of NCs by ATP seems likely to be of physiological relevance. It is also possible that, although ATP is an allosteric regulator of pGC and sGC (3, 22), another, as yet unidentified, molecule acts as physiological ligand. This notion is supported by the fact that the nucleotide-binding affinities of the allosteric sites of sGC and pGC are low—that is, in the millimolar range (17, 23). Consistent with this, the endogenous ligand for the allosteric site in mACs is not yet known (4). The inclusion of diterpenes was crucial for the successful crystallization of mACs (25). Conversely, the lack of inclusion of allosteric ligands may be the reason that holo-sGCs have evaded successful crystallization for decades (2).

In conclusion, NCs, by analogy to GPCRs (14, 15), can be considered as allosteric signal transduction machines that are regulated in a subtle manner by NTPs. The integration of allosteric sites and their ligands will also facilitate future crystallography studies urgently needed for sGC and pGCs. Once we know the crystal structures of these enzymes, including their catalytic and allosteric sites, it will also be much easier to understand their interaction and to target them pharmacologically. We anticipate that systematic medicinal chemistry studies on these sites will reveal NC-specific allosteric ligands, probably because the allosteric site is less conserved than the catalytic site. It will also be an exciting endeavor to identify the endogenous ligands for these allosteric sites. The development of fluorescent ligands will facilitate the analysis of dynamic interactions between catalytic and regulatory sites of NCs, as has already been shown for mAC (26). With the present study of Robinson and Potter (17) and other advances (23), the door has been opened into a new field of research on “difficult” drug targets, the NCs.

References and Notes

Funding: This work was supported by grants of the Deutsche Forschungsgemeinschaft (Se 529 | 5 2) to R.S.
View Abstract

Stay Connected to Science Signaling

Navigate This Article