Activation of soluble guanylate cyclase by nitrovasodilators is inhibited by oxidized low-density lipoprotein

In the presence of oxidized low-density lipoprotein the stimulatory effects of nitric oxide, sodium nitroprusside and S-nitrosoglutathione on soluble guanylate cyclase partially purified from bovine platelets were diminished in a concentration-dependent manner with IC50 values around 100 micrograms total cholesterol/ml. This inhibitory effect was potentiated about 10-fold when the enzyme was pre-incubated with the lipoprotein for 10 minutes at 37 degrees C which indicates a direct interaction of the lipoprotein with the guanylate cyclase. As oxidized low-density lipoprotein is present in the wall of atherosclerotic arteries, we suggest that the impaired response of atherosclerotic blood vessels to vasodilators may be due to a diminished activation of smooth muscle guanylate cyclase.     

Oxidized low-density lipoprotein antagonizes the activation of purified soluble guanylate cyclase by endothelium-derived relaxing factor but does not interfere with its biosynthesis.

Oxidized low-density lipoprotein (LDLox) is a molecule with strong atherogenic properties. In a concentration dependent fashion, LDLox antagonized the activation of purified soluble guanylate cyclase by endothelium-derived relaxing factor (EDRF), which was produced in vitro by incubation of a partially purified EDRF-forming enzyme in the presence of L-arginine, Ca2+ and NADPH. The inhibitory effect of LDLox was potentiated by preincubation of the soluble guanylate cyclase with LDLox, but not when the EDRF-forming enzyme was pretreated with LDLox. As LDLox did not diminish the calmodulin-dependent conversion of L-arginine into L-citrulline by the EDRF-forming enzyme it would appear that EDRF-biosynthesis was not affected by LDLox. It is suggested that the impaired relaxant response of atherosclerotic blood vessels to endothelium-dependent vasodilators was not due to a reduced formation of EDRF but due to a diminished responsiveness of soluble guanylate cyclase.     

High-density lipoprotein inhibits the oxidative modification of low-density lipoprotein

Oxidatively modified low-density lipoprotein (LDL), generated as a result of incubation of LDL with specific cells (e.g., endothelial cells, EC) or redox metals like copper, has been suggested to be an atherogenic form of LDL. Epidemiological evidence suggests that higher concentrations of plasma high-density lipoprotein (HDL) are protective against the disease. The effect of HDL on the generation of the oxidatively modified LDL is described in the current study. Incubation of HDL with endothelial cells, or with copper, produced much lower amounts of thiobarbituric acid-reactive products (TBARS) as compared to incubations that contained LDL at equal protein concentrations. Such incubations also did not result in an enhanced degradation of the incubated HDL by macrophages in contrast to similarly incubated LDL. On the other hand, inclusion of HDL in the incubations that contained labeled LDL had a profound inhibitory effect on the subsequent degradation of the incubated LDL by the macrophages while having no effect on the generation of TBARS or the formation of conjugated dienes. This inhibition was not due to the modification of HDL as suggested by the following findings. (A) There was no enhanced macrophage degradation of the HDL incubated with EC or copper alone, together with LDL, despite an increased generation of TBARS. (B) HDL with the lysine groups blocked (acetyl HDL, malondialdehyde (MDA) HDL) was still able to prevent the modification of LDL and (C) acetyl HDL and MDA-HDL competed poorly for the degradation of oxidatively modified LDL. It is suggested that HDL may play a protective role in atherogenesis by preventing the generation of an oxidatively modified LDL. The mechanism of action of HDL may involve exchange of lipid peroxidation products between the lipoproteins.     

Studying the structure and regulation of soluble guanylyl cyclase

Soluble guanylyl cyclase acts as the receptor for the signaling molecule nitric oxide. The enzyme consists of two different subunits. Each subunit shows the cyclase catalytic domain, which is also conserved in the membrane-bound guanylyl cyclases and the adenylyl cyclases. The N-terminal regions of the subunits are responsible for binding of the prosthetic heme group of the enzyme, which is required for the stimulatory effect of nitric oxide (NO). The five-coordinated ferrous heme displays a histidine as the axial ligand; activation of soluble guanylyl cyclase by NO is initiated by binding of NO to the heme iron and proceeds via breaking of the histidine-to-iron bond. Recently, a novel pharmacological and possibly physiological principle of guanylyl cyclase sensitization was demonstrated. The substance YC-1 has been shown to activate the enzyme independent of NO, to potentiate the effect of submaximally effective NO concentrations, and to turn carbon monoxide into an effective activator of soluble guanylyl cyclase.     

Characterization of oxidized guanosine 5′-triphosphate as a viable inhibitor of soluble guanylyl cyclase

The guanine base is prone to oxidation by free radicals regardless of the cellular moiety it is bound to. However, under conditions of oxidative stress, 8-oxoguanosine triphosphate (oxo⁸GTP) formation has been shown to occur without oxidation of the guanine base in DNA. In vitro studies have suggested that oxo⁸GTP could impact G-protein signaling and RNA synthesis. Whether increased levels of oxo⁸GTP translate into cellular malfunction is unknown. Data presented herein show that oxo⁸GTP is formed in cell-free preparations as well as in PC12 cells after exposure to physiologically relevant oxidative conditions generated with 10 μM copper sulfate and 1 mM l-ascorbic acid (Cu/Asc). We also determined that oxo⁸GTP has biological activity as a potent inhibitor of nitric oxide-stimulated soluble guanylyl cyclase (sGC). The increase in oxo⁸GTP formation in purified GTP and PC12 cells exposed to Cu/Asc caused a significant reduction in the product of sGC activity, cGMP. This oxidation of GTP was attenuated by the addition of reduced glutathione under these same Cu/Asc conditions, thus preventing the decrease in sGC activity. This suggests that oxo⁸GTP is produced by free radicals in vivo and could have significant impact on cell functions regulated by sGC activity such as synaptic plasticity in the central nervous system.     

Nucleotidyl cyclase activity of soluble guanylyl cyclase in intact cells

Soluble guanylyl cyclase (sGC) is activated by nitric oxide (NO) and generates the second messenger cyclic GMP (cGMP). Recently, purified sGC α₁β₁ has been shown to additionally generate the cyclic pyrimidine nucleotides cCMP and cUMP. However, since cyclic pyrimidine nucleotide formation occurred only the presence of Mn²⁺ but not Mg²⁺, the physiological relevance of these in vitro findings remained unclear. Therefore, we studied cyclic nucleotide formation in intact cells. We observed NO-dependent cCMP- and cUMP formation in intact HEK293 cells overexpressing sGC α₁β₁ and in RFL-6 rat fibroblasts endogenously expressing sGC, using HPLC–tandem mass spectrometry. The identity of cCMP and cUMP was unambiguously confirmed by HPLC–time-of-flight mass spectrometry. Our data indicate that cCMP and cUMP play second messenger roles and that Mn²⁺ is a physiological sGC cofactor.     

The effect of peroxynitrite on the catalytic activity of soluble guanylyl cyclase.

Soluble guanylyl cyclase (sGC) is a key enzyme of the *NO/cGMP pathway. Many cardiovascular disorders are associated with reduced *NO-mediated effects, while vascular superoxide (O(2)*(-)) production is increased. Both radicals rapidly react to peroxynitrite. We investigated whether peroxynitrite affects the activity and protein expression of sGC in intact vascular preparations. Catalytic sGC activity and expression of the sGC-beta(1) subunit was measured by conversion of radiolabeled GTP and western blot, respectively, using cytosolic extracts from rat aorta that had been incubated for 4 h with *NO/O(2)*(-) systems (devoid of free *NO) generating either 0.13 microM or 7.5 microM peroxynitrite/min. Incubation of rat aorta with 0.13 microM peroxynitrite/min had no effect. In striking contrast, incubation with 7.5 microM peroxynitrite/min resulted in a shift of the concentration-response curve obtained with a *NO donor (p =.0004) and a reduction of maximal specific activity from 3579 +/- 495 to 2422 +/- 265 pmol cGMP/mg/min (p =.036). The expression of the sGC-beta(1) subunit was unchanged. Exposure of aorta to the O(2)*(-) component had no effect, while exposure to the *NO-component reduced sGC expression to 58.8 +/- 7% (p <.001) and maximal sGC activity from 4041 +/- 992 to 1429 +/- 491 pmol cGMP/mg/min (p =.031). These data suggest that continuous generation of extracellular peroxynitrite might interfere with the *NO/cGMP signaling in vascular cells.     

Amphipathic alpha-helix mediates the heterodimerization of soluble guanylyl cyclase

Soluble guanylyl cyclase (soluble GC) is an enzyme consisting of alpha and beta subunits and catalyzes the conversion of GTP to cGMP. The formation of the heterodimer is essential for the activity of soluble GC. Each subunit of soluble GC has been shown to comprize three functionally different parts: a C-terminal catalytic domain, a central dimerization domain, and an N-terminal regulatory domain. The central dimerization domain of the beta(1) subunit, which contains an N-terminal binding site (NBS) and a C-terminal binding site (CBS), has been postulated to be responsible for the formation of alpha/ beta heterodimer. In this study, we analyzed heterodimerization by the pull-down assay using the affinity between a histidine tag and Ni(2+) Sepharose after co-expression of various N- and C-terminally truncated FLAG-tagged mutants of the alpha(1) subunit and the histidine-tagged wild type of the beta(1) subunit in the vaculovirus/Sf9 system, and demonstrated that the CBS-like sequence of the alpha(1) subunit is critical for the formation of the heterodimer with the beta(1) subunit and the NBS-like sequence of the alpha(1) subunit is essential for the formation of the enzymatically active heterodimer, although this particular sequence was not involved in heterodimerization. The analysis of the secondary structure of the alpha(1) subunit predicted the existence of an amphipathic alpha-helix in residues 431-464. Experiments with site-directed alpha(1) subunit mutant proteins demonstrated that the amphipathicity of the alpha-helix is important for the formation of the heterodimer, and Leu(463) in the alpha-helix region plays a critical role in the formation of a properly arranged active center in the dimer.     

Nitric oxide-dependent allosteric inhibitory role of a second nucleotide binding site in soluble guanylyl cyclase

The mechanism of desensitization of the nitric oxide (NO) receptor (alpha1.beta1 isoform of soluble guanylyl cyclase, sGC) is not known. Models of the structure of alpha1.beta1, based on the x-ray crystal structure of adenylyl cyclase (AC) suggest the existence of a nucleotide-like binding site, in addition to the putative catalytic site. We have previously reported that mutating residues that coordinate Mg(2+)GTP (substrate) binding in alpha1.beta1 into those present in AC fully reverts GC activity to AC activity. The wild-type form of alpha1.beta1 (GC-wt) and the mutant form (AC-mut, alpha1R592Q.beta1E473K,C541D) were purified, and their sensitivities to various nucleotides were assessed. In using the AC-mut as well as other mutants that coordinate purine binding, we were able to distinguish allosteric inhibitory effects of guanine nucleotides from competitively inhibitory effects on catalytic activity. Here we report that several nucleotide analogs drastically alter sGC and AC-mut activity by acting at a second nucleotide site, likely pseudosymmetric to the catalytic site. In particular, Mg(2+)GTP gamma S and Mg(2+)ATP gamma S inhibited cyclase activity through a mixed, non-competitive mechanism that was only observable under NO stimulation and not under basal conditions. The non-competitive pattern of inhibition was not present in mutants carrying the substitution beta1D477A, the pseudosymmetric equivalent to alpha1D529 (located in the substrate-binding site and involved in substrate binding and catalysis), or with the double mutations alpha1E525K,C594D, the pseudosymmetric equivalent to beta1E473K,C541D. Taken together these data suggest that occupation of the second site by nucleotides may underlie part of the mechanism of desensitization of sGC.     

Hemoglobin mediated nitrite activation of soluble guanylyl cyclase

Nitrite has long been known to be vasoactive when present at large concentrations but it was thought to be inactive under physiological conditions. Surprisingly, we have recently shown that supraphysiological and near physiological concentrations of nitrite cause vasodilation in the human circulation. These effects appeared to result from reduction of nitrite by deoxygenated hemoglobin. Thus, nitrite was proposed to play a role in hypoxic vasodilation. We now discuss these results in the context of nitrite reacting with hemoglobin and effecting vasodilation and present new data modeling the nitric oxide (NO) export from the red blood cell and measurements of soluble guanylate cyclase (sGC) activation. We conclude that NO generated within the interior of the red blood cell is not likely to be effectively exported directly as nitric oxide. Thus, an intermediate species must be formed by the nitrite/deoxyhemoglobin reaction that escapes the red cell and effects vasodilation.     

Differential effects of cGMP produced by soluble and particulate guanylyl cyclase on mouse ventricular myocytes

Particulate guanylyl cyclase (pGC) and soluble guanylyl cyclase (sGC) are cGMP-generation systems distributed in different intracellular locations. Our aim was to test the hypothesis that the functional effects of cGMP produced by pGC and sGC on contraction and Ca2+ transients would differ in ventricular myocytes. We measured myocyte shortening from adult mice using a video edge-detector and investigated the functional changes after stimulating pGC with C-type natriuretic peptide (CNP; 10(-8) M and 10(-7) M) or sGC with S-nitroso-N-acetyl-penicillamine (SNAP; nitric oxide donor; 10(-6) M and 10(-5) M). Significant concentration-dependent decreases in percentage shortening (PCS), maximal rate of shortening (RSmax), and relaxation (RRmax) were produced by CNP. To a similar degree, SNAP concentration-dependently reduced PCS, RSmax, and RRmax. The addition of Rp-8-[(4-chlorophenyl)thio]-cGMPS triethylamine (cGMP-dependent protein kinase inhibitor; 5 x 10(-6) M) or erythro-9-(2-hydroxy-3-nonyl) adenine (cGMP-stimulated cAMP phosphodiesterase inhibitor; 10(-5) M) reduced the responses induced by CNP or SNAP, suggesting that their actions were through cGMP-mediated pathways. While SNAP significantly increased intracellular cGMP concentration by 57%, CNP had little effect on cGMP production. We also found that CNP markedly decreased the amplitude of Ca2+ transients while SNAP had little effect, suggesting the cGMP generated by sGC may decrease myofilament Ca2+ sensitivity. The small amount of cGMP generated by pGC had a major effect in reducing Ca2+ level. This study suggested the existence of compartmentalization for cGMP in ventricular myocytes.     

Regulation of nitric oxide-responsive recombinant soluble guanylyl cyclase by calcium.

Calcium (Ca2+) and cyclic GMP (cGMP) subserve antagonistic functions that are reflected in their coordinated reciprocal regulation in physiological systems. However, molecular mechanisms by which Ca2+ regulates cGMP-dependent signaling remain incompletely defined. In this study, the inhibition of recombinant nitric oxide (NO)-stimulated soluble guanylyl cyclase (SGC) by Ca2+ was demonstrated. The alpha- and beta-subunits of recombinant rat SGC were heterologously coexpressed in HEK 293 cells which do not express NO synthase, whose Ca2+-stimulated activity can confound the effects of that cation on SGC. Ca2+ inhibited basal and NO-stimulated SGC in a concentration- and guanine nucleotide-dependent fashion. This cation inhibited SGC in crude cell extracts and immunopurified preparations. Ca2+ lowered both the Vmax and Km of SGC via an uncompetitive mechanism through direct interaction with the enzyme. In intact HEK 293 cells, increases in the intracellular Ca2+ concentration induced by ionomycin, a Ca2+ ionophore, and thapsigargin, which releases intracellular stores of that cation, inhibited NO-stimulated intracellular cGMP accumulation. Similarly, carbachol-induced elevation of the intracellular Ca2+ concentration inhibited NO-stimulated intracellular cGMP accumulation in HEK 293 cells. These data demonstrate that SGC behaves as a sensitive Ca2+ detector that may play a central role in coordinating the reciprocal regulation of Ca2+- and cGMP-dependent signaling mechanisms.     

The Dictyostelium homologue of mammalian soluble adenylyl cyclase encodes a guanylyl cyclase

A new Dictyostelium discoideum cyclase gene was identified that encodes a protein (sGC) with 35% similarity to mammalian soluble adenylyl cyclase (sAC). Gene disruption of sGC has no effect on adenylyl cyclase activity and results in a >10-fold reduction in guanylyl cyclase activity. The scg- null mutants show reduced chemotactic sensitivity and aggregate poorly under stringent conditions. With Mn(2+)/GTP as substrate, most of the sGC activity is soluble, but with the more physiological Mg(2+)/GTP the activity is detected in membranes and stimulated by GTPgammaS. Unexpectedly, orthologues of sGC and sAC are present in bacteria and vertebrates, but absent from Drosophila melanogaster, Caenorhabditis elegans, Arabidopsis thaliana and Saccharomyces cerevisiae.     

Sensitizing soluble guanylyl cyclase to become a highly CO-sensitive enzyme.

It took at least a decade to realize that the toxic gas NO is the physiological activator of soluble guanylyl cyclase (sGC), thereby acting as a signaling molecule in the nervous and cardiovascular systems. Despite its rather poor sGC-activating property, CO has also been implicated as a physiological stimulator of sGC in neurotransmission and vasorelaxation. Here, we establish YC-1 as a novel NO-independent sGC activator that potentiates both CO- and NO-induced sGC stimulation. As this potentiating effect is also observed with protoporphyrin IX which activates sGC independently of a gaseous ligand, we conclude that stabilization of the enzyme's active configuration is the underlying mechanism of YC-1's action. Moreover, the results obtained with YC-1 reveal that CO is capable of stimulating sGC to a degree similar to NO, and thus provide the molecular basis for CO functioning as a signaling molecule.     

Immunolocalization of soluble guanylyl cyclase subunits in rat kidney

Stimulation of soluble guanylyl cyclase (SGC) by nitric oxide (NO) results in the generation of cyclic guanosine monophosphate (cGMP). We recently described expression of abundant nitric oxide synthase, the enzyme by which NO is generated froml-arginine, in macula densa cells of rat kidney at the protein and mRNA level. In the present study we looked for possible targets of NO in the kidney. By light and electron microscopy, we applied polyclonal antisera against four subunits (₁,₂, ₁₂) of SGC in immunocytochemical studies of frozen sections of rat kidney. We demonstrate the presence of ₁-subunit in glomerular podocytes and of ₂-subunit in principal cells of the collecting duct. In both cell types a cytosolic localization was evident from ultrastructural analysis. Regarding the collecting duct, NO was shown by other authors to inhibit sodium reabsorption in cultured mouse cortical collecting duct principal cells. In podocytes NO may relax the contractile system of podocyte foot processes, the tone of which has been suggested to counteract the elastic distension of the capillary wall.     

Differential effects of soluble and particulate guanylyl cyclase on Ca(2+) sensitivity in airway smooth muscle.

Maximal relaxation of airway smooth muscle (ASM) in response to atrial natriuretic peptide (ANP), which stimulates particulate guanylyl cyclase (pGC), is less than that produced by nitric oxide (NO) and other compounds that stimulate soluble guanylyl cyclase (sGC). We hypothesized that stimulation of pGC relaxes ASM only by decreasing intracellular Ca(2+) concentration ([Ca(2+)](i)), whereas stimulation of sGC decreases both [Ca(2+)](i) and the force developed for a given [Ca(2+)](i) (i.e., the Ca(2+) sensitivity) during muscarinic stimulation. We measured the relationship between force and [Ca(2+)](i) (using fura 2) under control conditions (using diltiazem to change [Ca(2+)](i)) and during exposure to ANP, diethylamine-NO (DEA-NO), sodium nitroprusside (SNP), and the Sp diastereoisomer of beta-phenyl-1,N(2)-etheno-8-bromoguanosine-3',5'-cyclic monophosphorothionate (Sp-8-Br-PET-cGMPS), a cell-permeant analog of cGMP. Addition of DEA-NO, SNP, or Sp-8-Br-PET-cGMPS decreased both [Ca(2+)](i) and force, causing a significant rightward shift of the force-[Ca(2+)](i) relationship. In contrast, with ANP exposure, the force-[Ca(2+)](i) relationship was identical to control, such that ANP produced relaxation solely by decreasing [Ca(2+)](i). Thus, during muscarinic stimulation, stimulation of pGC relaxes ASM exclusively by decreasing [Ca(2+)](i), whereas stimulation of sGC decreases both [Ca(2+)](i) and Ca(2+) sensitivity.     

Receptors for modified low-density lipoproteins on human endothelial cells: different recognition for acetylated low-density lipoprotein and oxidized low-density lipoprotein

We examined the uptake pathway of acetylated low-density lipoprotein and oxidatively modified LDL (oxidized LDL) in human umbilical vein endothelial cells in culture. Proteolytic degradation of 125I-labeled Ac-LDL or Ox-LDL in the confluent monolayer of human endothelial cells was time-dependent and showed saturation kinetics in the dose-response relationship, which suggests that their incorporation is receptor-mediated. Cross-competition studies between acetylated LDL and oxidized LDL showed that the degradation of 125I-labeled acetylated LDL was almost completely inhibited by excess amount of unlabeled acetylated LDL, while only partially inhibited by excess unlabeled oxidized LDL. On the other hand, the degradation of 125I-labeled oxidized LDL was equally inhibited by excess amount of either acetylated or oxidized LDL. Cross-competition results of the cell-association assay paralleled the results shown in the degradation assay. These data indicate that human endothelial cells do not have any additional receptors specific only for oxidized LDL. On the contrary, they may have additional receptors, as we previously indicated on mouse macrophages, which recognize acetylated LDL, but not oxidized LDL.     

Mapping of heme-binding domains in soluble guanylyl cyclase beta1 subunit.

Soluble guanylyl cyclase (sGC) is activated upon the interaction of NO with heme in the sGC beta1 subunit. To identify the domains contributing to heme-binding, we constructed a series of deletion mutants of the beta1 subunit, and evaluated their heme-binding capability. Deletion mutants consisting of residues 1-120 [beta1(1-120)] and 80-385 [beta1(80-385)] were the shortest mutants exhibiting heme binding among the C-terminal and N-terminal-truncated mutants, respectively. The region common to both beta1(1-120) and beta1(80-385), i.e., residues 80-120, is therefore essential for heme binding, although the residues 341-385 play an auxiliary role in heme binding. Two deletion mutants, beta1(80-195) and beta1(60-195), which include only the essential region, exhibited strong heme binding and spectral properties similar to those of the nitrosyl complex of native sGC. Thus, these heme-binding core proteins may serve as model proteins for future studies on the tertiary structure of the nitrosyl complex of sGC.