Design, Synthesis, and Biological Evaluation of Novel Cyclic Adenosine-Inosine Monophosphate (cAIMP) Analogs That Activate Stimulator of Interferon Genes (STING).

We describe novel STING-activating cyclic dinucleotides whose constituent nucleosides are adenosine and inosine and that vary by ribose substitution, internucleotide linkage position, and phosphate modification. In mammalian cells in vitro, some of these cAIMP analogs induce greater STING-dependent IRF and NF-κB pathway signaling than do the reference agonists for murine (DMXAA) or human (2',3'-cGAMP) STING. In human blood ex vivo, they induce type I interferons (IFNs) and proinflammatory cytokines: for the former, 3',3'-cAIMP (9; EC50 of 6.4 µM) and analogs 52-56 (EC50 of 0.4-4.7 µM), which contain one or two 2'-fluoro-2'-deoxyriboses and/or bis-phosphorothioate linkages, are more potent than 2',3'-cGAMP (EC50 of 19.6 µM). Interestingly, 9 induces type I IFNs more strongly than do its linkage isomers 2',3'-cAIMP (10), 3',2'-cAIMP (23), and 2',2'-cAIMP (27). Lastly, some of the cAIMP analogs are more resistant than 2',3'-cGAMP to enzymatic cleavage in vitro. We hope to exploit our findings to develop STING-targeted immunotherapies.

Structural determinants for N1/N7 cyclization of nicotinamide hypoxanthine 5'-dinucleotide (NHD+) derivatives by ADP-ribosyl cyclase from aplysia californica: Ca2+-mobilizing activity of 8-substituted cyclic inosine 5'-diphosphoribose analogues in T-lymphocytes.

A series of nicotinamide hypoxanthine 5'-dinucleotide (NHD+) analogues modified at C-8 (2-5) and 7-deaza-NHD+ were synthesized, and cyclization in the presence of Aplysia ADP-ribosyl cyclase was studied. All 8-substituted NHD+ analogues were converted into their N1-cyclic forms by the enzyme, while in contrast, 7-deaza-NHD+ 17 was hydrolyzed into 7-deazainosine 5'-diphosphoribose (7-deaza-IDPR) 25. Correlations are made showing that the conformation of the NHD+ substrate is the key to successful cyclization. The pharmacological activities of these novel cIDPR derivatives were evaluated in both permeabilized and intact Jurkat T-lymphocytes. The results show that in permeabilized cells both 8-iodo 1g and 8-N3-N1-cIDPR 1d have an activity comparable to that of cADPR, while 8-iodo 1g and 8-phenyl-N1-cIDPR 1c have a small but significant effect in intact cells and can therefore be regarded as membrane-permeant; thus, cIDPR derivatives are emerging as important novel biological tools to study cADPR-mediated Ca2+ release in T-cells.

Nucleotide modulation of f-channel activity in rabbit sino-atrial node myocytes.

f-channel nucleotide modulation was investigated in sino-atrial (SA) node cells isolated from rabbit hearts, using an inside-out macropatch configuration. Saturating doses (30 microM) of phosphorothioate derivatives of cAMP, Sp-cAMPS and Rp-cAMPS, were tested on the cAMP-induced shift of I(f) activation. Responses were not altered when Sp-cAMPS was combined with cAMP. When Rp-cAMPS was superfused with subsaturating cAMP concentrations (1-10 microM), it inhibited cAMP-induced I(f) activation shift. cGMP and cIMP reversibly shifted the I(f) conductance-voltage curve to more positive values; however they had a lesser specificity than that of cAMP. The efficacy ranking for I(f) activation by cyclic nucleotides was: cAMP > cGMP > cIMP. Non cyclic nucleotides (ATP, ADP and AMP) failed to change I(f) activation, indicating that the cyclic nature of nucleotides seems to be essential to f-channel modulation. Similarities between f-channels and cyclic nucleotide-gated (CNG) channels are discussed.

Nitric oxide depresses GABAA receptor function via coactivation of cGMP-dependent kinase and phosphodiesterase.

Nitric oxide (NO) is thought to play an essential role in neuronal processing, but the downstream mechanisms of its action remain unclear. We report here that NO analogs reduce GABA-gated currents in cultured retinal amacrine cells via two distinct, but convergent, cGMP-dependent pathways. Either extracellular application of the NO-mimetic S-nitroso-N-acetyl-penicillamine (SNAP) or intracellular perfusion with cGMP depressed GABA currents. This depression was partially blocked by a pseudosubstrate peptide inhibitor of cGMP-dependent protein kinase (PKG), suggesting both PKG-dependent and independent actions of cGMP. cAMP-dependent protein kinase (PKA) is known to enhance retinal GABA responses. 8-Bromoinosine 3',5'-cyclic monophosphate (8Br-cIMP), which activates a type of cGMP-stimulated phosphodiesterase that hydrolyzes cAMP, also significantly reduced GABA currents. 1-Methyl-3-isobutylxanthine (IBMX), a nonspecific phosphodiesterase (PDE) inhibitor, blocked both the action of 8Br-cIMP and the portion of SNAP-induced depression that was not blocked by PKG inhibition. Our results suggest that NO depresses retinal GABAA receptor function by simultaneously upregulating PKG and downregulating PKA.

The cyclic GMP-mediated calcium release pathway in sea urchin eggs is not required for the rise in calcium during fertilization.

The mechanisms required for cGMP-induced Ca2+ release in the sea urchin egg were investigated using both egg homogenates and intact eggs. The postulated pathway of cGMP-dependent protein kinase (PKG) activation of ADP-ribosyl cyclase for production of cADPR to activate the ryanodine receptor Ca2+ channel was tested with a variety of activators (cGMP analogs and cIMP) and inhibitors (Rp-8-pCPT-cGMPS, 3-aminopyridine NAD, nicotinamide, and spermine). Our observations are consistent with Ca2+ release by cGMP in the egg being dependent on an isoform of PKG that is distinct from the mammalian enzyme. PKG activity in the sea urchin egg was activated by cIMP, but was insensitive to cGMP analogs, which are potent activators of mammalian isoenzymes. Surprisingly, it appears the activation of the cGMP-dependent Ca2+ release pathway was unnecessary during fertilization. Inhibitors of either PKG or ADP-ribosyl cyclase activities did not prevent the transient rise in intracellular Ca2+ activity in heparin-loaded eggs during fertilization. These results suggest the synthesis of cADPR during fertilization is not necessary for regulating the Ca2+ event.

Molecular mechanism for ligand discrimination of cyclic nucleotide-gated channels.

Cyclic nucleotide-gated ion channels of retinal photoreceptors and olfactory neurons are differentially activated by ligands that vary only in their purine ring structure. The nucleotide selectivity of the bovine rod cyclic nucleotide-gated channel (cGMP > cIMP >> cAMP) was significantly altered by neutralization of a single aspartic acid residue in the binding domain (cGMP > or = cAMP > cIMP). Substitution by a nonpolar residue at this position inverted agonist selectivity (cAMP >> cIMP > or = cGMP). These effects resulted from an alteration in the relative ability of the agonists to promote the allosteric conformational change associated with channel activation, not from a modification in their initial binding affinity. We propose a general mechanism for guanine nucleotide discrimination, in common with that observed in high affinity GTP-binding proteins, involving the formation of a pair of hydrogen bonds between the aspartic acid side chain and N1 and N2 of the guanine ring.

Holoenzymes of cAMP-dependent protein kinase containing the neural form of type I regulatory subunit have an increased sensitivity to cyclic nucleotides.

Specific isoforms of the cAMP-dependent protein kinase are preferentially expressed within discrete neuronal regions in mouse brain (Cadd and McKnight (1989) Neuron 3, 71-79) suggesting that these subunits might have different functional properties. We have used recombinant techniques to express and purify the type I regulatory subunits, RI alpha and RI beta, the catalytic subunits C alpha and C beta, and then reconstituted holoenzymes with the various combinations of R and C subunits. The ability of the subunits to form inactive holoenzymes and then to be activated in the presence of cyclic nucleotides was examined. Holoenzymes containing C beta had essentially the same activation properties exhibited by C alpha holoenzymes. However, the presence of the neural form of RI, RI beta, led to formation of a holoenzyme which was activated at a 3-7-fold lower concentration of cyclic nucleotides compared to holoenzymes containing RI alpha. Expression of the RI beta protein in discrete regions of the central nervous system may provide a mechanism for increasing the sensitivity of the kinase to what would otherwise be subthreshold levels of stimulation. Two mutant forms of RI beta were constructed that converted the RI beta sequence to that of RI alpha at position 98 (RI beta Ala) or positions 98 and 99 (RI beta Ala/Ile). These sequences form part of a pseudosubstrate site thought to interact with the C subunit. Wild type and mutant R subunits were combined in vitro with purified bovine C subunits and half maximal activation constants (Ka) were determined with cyclic nucleotides. Holoenzymes containing RI beta Ala and RI beta Ala/Ile gave Ka values which were higher than wild type RI beta, with the double mutant shifting toward the Ka value of RI alpha holoenzymes by about 30%. These results suggest that amino acid differences in the pseudosubstrate site may account for some, but not all, of the increased sensitivity to cyclic nucleotides exhibited by RI beta.

Stimulation of glucose transport in rat cardiac myocytes by guanosine 3',5'-monophosphate.

Glucose transport in isolated rat cardiomyocytes is stimulated by insulin, catecholamines, and anoxia approximately 2- to 3-fold over basal rates. The molecular mechanisms controlling these responses are unknown. In our search for possible cellular mediators of glucose transport stimulation, we examined the effects of a number of nucleotides on 3-O-methylglucose transport in heart cells. The nucleotides and/or permeable analogs (monosuccinyl, 8-bromo, and dibutyryl derivatives) included cUMP, cIMP, cCMP, cAMP, and cGMP at concentrations ranging from 10 nM to 1 mM. Of all the nucleotides tested only cGMP analogs induced a significant stimulation of transport at concentrations as low as 100 nM. This effect was observed in both the 8-bromo- and dibutyryl derivatives and with 1 mM cGMP itself. The effect was concentration dependent for both analogs and produced a maximal response equivalent to that of 100 nM insulin. This insulinomimetic effect of cGMP was examined in more detail in order to evaluate its role as a potential mediator of this response. Agents that are known to stimulate guanylate cyclase in the heart produced a clear stimulation of transport when added to cardiomyocytes. These include insulin, aminophylline, histamine, beta-estradiol, and biotin-nitrophenyl ester. Methylene blue, an inhibitor of guanylate cyclase, blocked the insulin response when added to cells before insulin, but was ineffective when added after insulin. In addition, agents that raise intracellular cGMP levels by inhibiting cyclic nucleotide phosphodiesterases were also examined for effects on glucose transport. Out of several phosphodiesterase inhibitors tested, only Zaprinast (which selectively increases cGMP in heart) stimulated transport in a concentration-dependent manner to within 80% of the maximal insulin effect. These results are consistent with the notion that cGMP may be involved in glucose transport stimulation.

Activation of type I cyclic AMP-dependent protein kinases is impaired by a point mutation in cyclic AMP binding sites.

The type I regulatory (R-I) subunit of cyclic AMP-dependent protein kinase (A-kinase) was expressed in E. coli, and a single amino acid substitution in cyclic AMP binding sites A or B was introduced by site-directed mutagenesis. The cyclic AMP binding activity and cyclic AMP-stimulated phosphotransferase activity of the holoenzymes formed by wild-type or mutant R-Is and the purified bovine catalytic subunit of A-kinase were then examined. The wild-type holoenzyme was activated by low concentrations of cyclic AMP, a finding in accord with its high-affinity binding to cyclic AMP. In contrast, although the two mutant holoenzymes showed high-affinity cyclic AMP binding at their non-mutated sites, both holoenzymes were resistant to activation by cyclic AMP. Thus, binding of cyclic AMP to the non-mutated cyclic AMP binding site is not sufficient to dissociate the catalytic subunit from the mutant R-Is upon cyclic AMP binding. These results suggest that both A and B cyclic AMP binding sites are required for efficient coupling between cyclic AMP binding and activation of the enzyme.

Inhibition of growth of Rhizobium japonicum by cyclic GMP.

Exogenous cyclic guanosine-3',5'-monophosphate (cGMP) inhibited the growth of Rhizobium japonicum at less than 100 microM. Other nucleotides, including cyclic AMP, cyclic IMP, and cyclic CMP, had no inhibitory effect even at higher concentrations nor was the inhibition by cGMP reversed by cyclic AMP. The inhibitory effect was independent of the carbon and nitrogen source(s) used. cGMP did not inhibit the growth of any other species of bacterium tested, including several fast-growing Rhizobium species. The kinetics of growth inhibition are multiphasic, with no apparent effect for several hours after addition, followed by a period of total inhibition. Subsequently, growth resumed at a slower rate. Resumption of growth was not due to destruction of the nucleotide. Studies of the intracellular cGMP concentration did not reveal significant changes in cells grown under aerobic or microaerobic conditions. No effect of cGMP on the derepression of respiratory nitrate reductase was observed.

Inhibition of utilization of hypoxanthine and guanine in cells treated with the carbocyclic analog of adenosine. Phosphates of carbocyclic nucleoside analogs as inhibitors of hypoxanthine (guanine) phosphoribosyltransferase.

In cell cultures treated with the carbocyclic analog of adenosine (C-Ado, (+/-)-aristeromycin), the utilization of hypoxanthine and guanine has been observed to be blocked. In an attempt to define the mechanism of this inhibition, we have reexamined the metabolism of C-Ado and its effects on the metabolism of guanine and hypoxanthine. In cultures of L1210 cells, C-Ado at a concentration of 25 microM inhibited the utilization of hypoxanthine and guanine for nucleotide synthesis by more than 90% but produced little or no inhibition of the utilization of these bases in cultures of L1210/MeMPR cells which lack adenosine kinase and cannot phosphorylate C-Ado. In cultures of mammalian cells (L1210, HEp-2, and colon-26 cells), C-Ado was converted to the triphosphate (as previously observed) and also to the triphosphate of the carbocyclic analog of guanosine. The presence of coformycin in the medium at a concentration sufficient to inhibit AMP deaminase almost completely prevented the formation of carbocyclic GTP; thus, the deamination of C-Ado monophosphate is essential for the formation of phosphates of carbocyclic guanosine. Since hypoxanthine (guanine) phosphoribosyltransferase is known to be subject to end product inhibition, it was considered likely that phosphates of carbocyclic guanosine or carbocyclic inosine, present in C-Ado-treated cells, were responsible for inhibition of utilization of hypoxanthine and guanine. The 5'-phosphates of the carbocyclic analogs of inosine and guanosine were synthesized and found to be effective inhibitors of the phosphoribosyltransferase. Carbocyclic GMP was a better inhibitor than carbocyclic IMP and was also superior to GMP and IMP; the concentration of C-GMP that produced a 50% inhibition of GMP formation was approximately 1 microM. It is probable that the presence of phosphates of carbocyclic guanosine accounts for the inhibition of utilization of hypoxanthine and guanine in C-Ado-treated cells.

A unique cyclic nucleotide-dependent protein kinase.

During the course of studying the soluble cyclic nucleotide-dependent protein kinases of a developing insect, three different enzymes were isolated. Two of these were found to be cAMP-dependent enzymes eluting from DEAE-cellulose in a manner identical with protein kinases I and II found in vertebrate muscle. The third enzyme appears to be unique. It has high affinity for either cAMP or cGMP (KA of 43 nM and 25 nM, respectively), the only cyclic nucleotide-dependent kinase described, to have this property. The enzyme has lower affinity for cIMP and cCMP (KA of 160 nM and 340 nM, respectively). Binding to cyclic nucleotide does not alter enzyme size. The KM for ATP is 86 microM, and among several types of histones tried, the slightly lysine-rich subgroup f2a was the best phosphate acceptor. Maximum activity was obtained with 1 mM Mg2+ while Mn2+ was completely ineffective. This new enzyme was purified to homogeneity on a cAMP affinity column as judged by two-dimensional electrophoresis. On the basis of molecular sieving and sodium dodecyl sulfate electrophoresis we have reached the preliminary conclusion that the native enzyme is a dimer of identical subunits with a molecular weight of 180,000. If the mammalian cAMP and cGMP enzymes are indeed homologous proteins, perhaps we have in this new kinase a species that represents a common ancestral protein.

Comparison of cyclic nucleotide specificity of guanosine 3',5'-monophosphate-dependent protein kinase and adenosine 3',5'-monophosphate-dependent protein kinase.

Guanosine 3',5'-monophosphate-dependent protein kinase (cyclic GMP-dependent protein kinase) and adenosine 3',5'-monophosphate-dependent protein kinase (cyclic AMP-dependent protein kinase) exhibited a high degree of cyclic nucleotide specificity when hormone-sensitive triacylglycerol lipase, phosphorylase kinase, and cardiac troponin were used as substrates. The concentration of cyclic GMP required to activate half-maximally cyclic dependent protein kinase was 1000- to 100-fold less than that of cyclic AMP with these substrates. The opposite was true with cyclic AMP-dependent protein kinase where 1000- to 100-fold less cyclic AMP than cyclic GMP was required for half-maximal enzyme activation. This contrasts with the lower degree of cyclic nucleotide specificity of cyclic GMP-dependent protein kinase of 25-fold when histone H2b was used as a substrate for phosphorylation. Cyclic IMP resembled cyclic AMP in effectiveness in stimulating cyclic GMP-dependent protein kinase but was intermediate between cyclic AMP and cyclic GMP in stimulating cyclic AMP-dependent protein kinase. The effect of cyclic IMP on cyclic GMP-dependent protein kinase was confirmed in studies of autophosphorylation of cyclic GMP-dependent protein kinase where both cyclic AMP and cyclic IMP enhanced autophosphorylation. The high degree of cyclic nucleotide specificity observed suggests that cyclic AMP and cyclic GMP activate only their specific kinase and that crossover to the opposite kinase is unlikely to occur at reported cellular concentrations of cyclic nucleotides.

Unspecific and specific stimulating effects of 3',5'-cyclic nucleotides (cGMP, cAMP, cIMP, cCMP) on the in vitro biosynthesis of proteochondroitin-4,-6-sulfate and on other anabolic processes from calf rib cartilage.

With cartilage slices from calf ribs, cGMP as well as cAMP accelerate dose-dependently and specifically label rates of Ch-4-,-6-S protein; they slightly elevate rates of anaerobic glycolysis dose-independently and unspecifically, similar to their 5-monophosphate compounds. cAMP, but not cGMP, slightly stimulates labeling of total protein dose-dependently. Guanosine and adenosine (as well as adenine) accelerate more significantly all three anabolic processes in the order Ch-4-,-6-S protein formation greater than or equal to total protein labeling greater than anaerobic glycolysis. Acceleration of some of the processes rises further after adding theophylline or SQ 20.009, depending on the nucleoside used. diBu-cAMP (but not 8-Br-cAMP) stimulates the three processes more than cAMP; diBu-cGMP and 8-Br-cGMP alone increase the labeling rates of protein more than cGMP, cCMP and cIMP slightly accelerate at least one of the three processes dose-independently and unspecifically, similar to their 5-monophosphate compounds. cUMP was almost inactive. The results point to specific and unspecific effects of cGMP similar or different to those of cAMP.

Stimulation by cyclic nucleotides of prostaglandin E production in isolated graafian follicles.

Rat Graafian follicles isolated intact responded to 8-Br-cyclic AMP and 8-Br-cyclic GMP with increased prostaglandin E (PGE) production during a 6 h incubation. By contrast, 8-Br-cyclic IMP, 8-Br-5' AMP and 8-Br-5' GMP were inactive in this respect. The effect of 8-Br-cyclic AMP and 8-Br-cyclic GMP was noted only after a lag period of about 4 h. Choleragen, LH, and the phosphodiesterase inhibitor (3-isobutyl-l-methyl-xanthine; IBMX) also stimulated PGE production. Actinomycin D and cycloheximide given simultaneously with 8-Br-cyclic AMP or LH prevented the stimulatory effect of these agents. Concomitant addition of arachidonic acid did not overcome the effect of these inhibitors. Administration of hCG in vivo or incubation with LH in vitro did not elevate endogenous ovarian free arachidonate, while PGE production was enhanced. Dexamethasone prevented this stimulatory effect of hCG. Collectively, the results suggest that stimulation of ovarian PGE production by cyclic nucleotides and LH is dependent on de novo synthesis of one or more components of the PG synthetase system rather than on substrate availability. Cyclic nucleotides may mediate the stimulatory effect of gonadotropins on PGE production.

Substrate and effector specificity of a guanosine 3':5'-monophosphate phosphodiesterase from rat liver.

Guanosine 3':5'-monophosphate phosphodiesterases, which appear to be under allosteric control, have been partially purified from rat liver supernatant and particulate fractions. The preferred substrate for both phosphodiesterases was cGMP (Km values: cGMP less than cIMP less than cAMP). At subsaturating concentrations of substrate, the phosphodiesterases were stimulated by purine cyclic nucleotides. The order of effectiveness for activation of cyclic nucleotide hydrolysis was cGMP greater than cIMP greater than cAMP greater than cXMP. Using cAMP derivatives as activators of cIMP hydrolysis, modifications in the ribose, cyclic phosphate, and purine moieties were shown to alter the ability of the cyclic nucleotide to activate the supernatant enzyme. cGMP, at concentrations that stimulated cyclic nucleotide hydrolysis, enhanced chymotryptic inactivation of the supernatant phosphodiesterase. At similar concentrations, cAMP was not effective. It appears that on interaction with appropriate cyclic nucleotides, this phosphodiesterase undergoes conformational changes that are associated with increased catalytic activity and enhanced susceptibility to proteolytic attack. Divalent cation may not be required for the nucleotide-phosphodiesterase interaction and resultant change in conformation.