Identification of Residues Crucially Involved in the Binding of the Heme Moiety of Soluble Guanylate Cyclase*

Soluble guanylate cyclase (sGC), a heterodimeric hemeprotein, is the only receptor for the biological messenger nitric oxide (NO) identified to date and is intimately involved in various signal transduction pathways. By using the recently discovered NO- and heme-independent sGC activator BAY 58-2667 and a novel cGMP reporter cell, we could distinguish between heme-containing and heme-free sGC in an intact cellular system. Using these novel tools, we identified the invariant amino acids tyrosine 135 and arginine 139 of the β1-subunit as crucially important for both the binding of the heme moiety and the activation of sGC by BAY 58-2667. The heme is displaced by BAY 58-2667 due to a competition between the carboxylic groups of this compound and the heme propionic acids for the identified residues tyrosine 135 and arginine 139. This displacement results in the release of the axial heme ligand histidine 105 and to the observed activation of sGC. Based on these findings we postulate a signal transmission triad composed of histidine 105, tyrosine 135, and arginine 139 responsible for the enzyme activation by this compound and probably also for transducing changes in heme status and porphyrin geometry upon NO binding into alterations of sGC catalytic activity.

tation, neurotransmission, and platelet aggregation (1)(2)(3). The pathogenesis of various diseases, especially those of the cardiovascular system, has been linked to inappropriate activation of sGC (4 -6). The concept of NO-dependent sGC activation as a mechanism underlying antianginal action has been validated by the successful clinical use of NO-releasing drugs for more than a century (7,8).
YC-1, an indazole derivative, was the first NO-independent activator of sGC discovered (9 -12). Recently, the YC-1-related substance BAY 41-2272 has been identified as a novel, more specific, and more potent compound that stimulates sGC in a concentration-dependent manner and shows a strong synergism when combined with NO (13). sGC activation through BAY 41-2272 requires the presence of the heme moiety, although a direct interaction with the heme-iron is not likely as suggested by spectroscopic studies (13). In contrast to BAY 41-2272, another recently described non-NO-based sGC activator BAY 58-2667 stimulates not only the native sGC but, even more potently, the heme-deficient or oxidized form of the enzyme (14,15) suggesting a novel mechanism of activation. The existence of these different types of non-NO-based sGC-activating compounds suggest that several unique allosteric regulatory sites are present on the enzyme and may open up new therapeutic avenues for cardiovascular diseases (6).
The degree of sGC activation by these two different types of synthetic sGC activators as well as NO is dependent on the presence or oxidation state of the heme moiety. The involvement of this prosthetic group and the heme-binding domain in nearly all mechanisms of sGC activation (and inhibition) has been difficult to probe due to the lack of compounds capable of activating the heme-free enzyme. The only known function, ascribed to this domain for many years, was binding of the heme moiety to the ␤-subunit of sGC via the histidine 105 and its alteration upon NO binding (16 -18).
As early as 1984, however, Ignarro and coworkers (19) could show the importance of the two propionic acid groups of the heme moiety for its binding to the enzyme and postulated: "These propionic acid groups, which are ionized at pH 7.4, may form tight ion pairs with positively charged groups in guanylate cyclase and thereby contribute to the binding of porphyrins." Using the indazole BAY 41-2272, the aminodicarboxylic acid BAY 58-2667 and a novel cGMP reporter cell we were able to identify the two residues tyrosine 135 and arginine 139 of the ␤ 1 -subunit of sGC as the postulated counterparts of the propionic acid groups of the heme moiety. Alteration of these amino acids resulted not only in the loss of the heme binding capacity of the enzyme but also in a reduction of the BAY 58-2667-induced sGC activation. Based on these findings we suggest a model of sGC activation with two high affinity binding sites for BAY 58-2667: one that is saturable at nanomolar concentrations and shows no direct interference with the heme moiety as described earlier (14), and a second one that exhibits a direct competition between BAY 58-2667 and the prosthetic heme for the tyrosine 135 and the arginine 139. This competition results in the displacement of the heme moiety and as a consequence to the observed activation of sGC probably due to the release of the axial heme ligand histidine 105. Based on these findings we postulate a signal transmission triad composed of histidine 105, tyrosine 135, and arginine 139 responsible for the enzyme activation by BAY 58-2667 and probably also for transducing changes in heme status and porphyrin geometry upon NO binding into alterations of sGC catalytic activity.
Generation of a cGMP Reporter Cell-To characterize sGC mutants in an intracellular environment, a cGMP reporter cell was constructed based on a method reported earlier (22). Briefly, a CHO cell line expressing cytosolic aequorin was stably transfected with a plasmid coding for the cGMP-gated ion channel CNG2 under a zeocin resistance. Thereafter, zeocin-resistant clones were characterized for channel expression, and active clones were subcloned by the limited dilution technique. Selected clones were cultured in Dulbecco's modified Eagle's medium/F-12 with L-glutamine (Invitrogen, Carlsbad, CA), 1 mM sodium pyruvate, and 0.075% sodium bicarbonate, supplemented with 50 units/ml penicillin, 50 g/ml streptomycin, 2.5 g/ml amphotericin B, and 10% (v/v) inactivated fetal calf serum.
Transient Transfection-cGMP readout cells were seeded on 96-well microtiter plates at a density of 20,000 cells per well and were cultured for 2 days at 37°C and 5% CO 2 to ensure confluent growth. Afterward, cells were cotransfected applying a transfection mixture containing 75 ng of ␣ 1 -and 75 ng of ␤ 1 -plasmid, 0.5 l of Plus® reagent and 1 l of LipofectAMINE® (Invitrogen) in 100 l of Opti-MEM® serum-free medium (Invitrogen) according to the manufacturer's protocol. After 3 h at 37°C the transfection medium was replaced with serum-containing medium, and cells were incubated for 48 h at 37°C to ensure optimal expression of the WT-sGC and mutant-sGC.
cGMP Readout-Cells were cultured and transiently transfected with WT-sGC or mutant-sGC as described above. 48 h after transfection medium was removed, and the transfected cells were incubated with calcium-free buffer (130 mM NaCl, 5 mM KCl, 20 mM HEPES, 1 mM MgCl 2 , 4.8 mM NaHCO 3 , pH 7.4) containing 0.83 g/ml coelenterazine for 3 h at 37°C. For the determination of the sGC activation profile, cells were incubated with various concentrations of sGC activators in a volume of 50 l at 37°C for 15 min. cGMP readout was initiated by application of buffer containing 10 mM CaCl 2 , and emitted light was measured as relative light units (RLU) in a light-tight box using a charge-coupled device camera.
PPIX Reconstitution-To investigate whether the performed mutations influenced the heme binding capacity of the enzyme, WT-, ␤Y135F-, and R139L-sGC were transiently expressed in CHO cells as described above. After 48 h, cells were harvested and centrifuged at 100,000 ϫ g. The sGC-containing supernatant was applied to the sGC activity assay. Reconstitution was performed with increasing concentrations of PPIX in the presence of 0.5% Tween 20 and 1 M BAY 41-2272 to amplify the PPIX-induced sGC activation (23, 24).
sGC Purification and Activity Assay-We expressed sGC by using a baculovirus/Sf9 expression system and measured enzyme activity as described earlier (25). Briefly, a bioreactor was filled with 2.5 liters of medium (SF900II, 10% fetal calf serum) and inoculated with SF9 cultures. The culture was incubated at 28°C until a final density of 2 ϫ 10 6 cells/ml. After infection with virus stocks containing both subunits of sGC, cells were incubated for further 88 h at 28°C and pelleted by centrifugation. The pellet was harvested by sonication on ice, and sGC was purified as described (25).
Spectroscopic Studies-UV-visible spectra were recorded from 300 to 600 nm on a Beckman DU 640 spectrophotometer as reported earlier (15). 1 mM and 10 mM stock solutions of BAY 58-2667 were prepared in Me 2 SO and added to the samples to obtain a final concentration of 10 and 100 M. ODQ was prepared as a 100 M stock solution in MilliQ H 2 O and diluted to a final concentration of 10 M. Samples containing 15 g of sGC were incubated at 37°C for 10 min in the absence or presence of BAY 58-2667 (10 and 100 M) alone or combined with ODQ. Before the UV-visible spectra were recorded, the enzyme was separated by ion exchange chromatography from BAY 58-2667, ODQ, or free heme that might interfere with the measurement.
Separation of sGC from Detergent-Separation of sGC from detergent and unbound heme was performed as previously described (26). Briefly, sGC-containing samples were loaded onto ion exchange columns and washed once to remove any traces of detergent. Bound sGC was eluted with 60 l of elution buffer (300 mM NaCl), and UV-visible spectra were recorded as described above.

RESULTS
Sequence Alignments-The N-terminal 200 amino acids of the ␤ 1 -subunit of sGC are sufficient to bind the prosthetic heme moiety with spectral characteristics comparable to those of the native sGC (27,28). Based on this information, we initiated a BLAST search with this putative heme-binding domain. Various sGC ␣and ␤-subunits from different species as well as predicted proteins of unknown function were found. A subsequent alignment of sequences known to bind heme identified invariant or conserved amino acids (Fig. 1). From these residues, the positively charged and polar amino acids ␤H 105 , ␤Y 135 , and ␤R 139 were chosen for site-directed mutagenesis as indicated by the asterisk.
Screening sGC Mutants-For screening sGC mutants we constructed a novel cGMP reporter cell line based on a CHO cell stably transfected with a cGMP-dependent cation channel (CNG2) and cytosolic aequorin. After maximum activation of this reporter cell in a 96-well microtiter plate, a signal of about 2,500,000 RLUs in each well could be measured. To ensure that the measured sGC activity was within the linear range of the readout system the experimental conditions were chosen to obtain a maximum signal between 500,000 and 1,000,000 RLUs after maximal sGC activation (see Supplemental Material to Fig. 2). The different mutants of the sGC ␤ 1 -subunit were constructed according to the results of the alignment (  Fig. 2). In addition, a concentration-dependent stimulation of 15.9-fold could be achieved by the NO-independent but heme-dependent sGCstimulator BAY 41-2272 ( Fig. 2A and Table I). In the presence of 10 nM DEA/NO, a concentration that exhibits only negligible effects on sGC activation (2.2-fold; Fig. 3A), the maximal activation of BAY 41-2272 was increased 60% ( Fig. 2A and Table I).
The NO-and heme-independent sGC activator BAY 58-2667 showed a concentration-dependent activation of the transiently transfected enzyme of 7.4-fold that was potentiated up to 25.6-fold in the presence of the sGC inhibitor ODQ ( Fig. 2A and Table I).
␤H105F-The mutant ␤H105F was generated as a control that causes the expression of heme-free sGC. Indeed, the mutant ␤H105F exhibited the activation pattern expected of heme-free sGC: no measurable activation of the enzyme in the presence of DEA/NO (Fig. 3, A and B), a negligible activation (2.1-fold) after incubation with BAY 41-2272, not further elevated when combined with NO ( Fig. 2B and Table I), and activation by heme-independent BAY 58-2667 (7.3-fold), which was not enhanced by the presence of ODQ ( Fig. 2B and Table I).
␤Y135F and ␤Y135A-Neither mutants ␤Y135F or ␤Y135A were activated by DEA/NO (Fig. 3, A and B). Both mutations abolished nearly completely the enzyme activation by BAY 41-2272 alone or when combined with NO (Fig. 2, C and D, and Table I). In contrast, both mutants were responsive to BAY 58-2667 with similar EC 50 values reaching a maximal activation of 34-fold and 6.6-fold, respectively (Table I). This activation was not further potentiated by additional ODQ (Fig. 2, C and D, and Table I).
␤R139L and ␤R139A-These mutations caused the loss of the responsiveness toward NO and BAY 41-2272 (Figs. 2E, 2F, 3A, 3B, and Table I). In contrast, both mutants could be activated by BAY 58-2667 with a minimal effective concentration of about 1 M. ␤R139L achieved a maximal activation of 21.9fold in the absence and 21.1-fold in the presence of ODQ. The exchange ␤R139A was not saturable within the tested concentration range of BAY 58-2667 (Fig. 2F). After incubation with 10 M BAY 58-2667, an activation of 16.8-fold was determined increasing to 18.7-fold in the presence of ODQ (Fig. 2F). Based on the extrapolated concentration-response curves, EC 50 values of about 7 M were assumed.
␤Y135F plus R139L and ␤Y135A plus ␤R139A-The construct with both Y135F and R139L resulted in an enzyme that was insensitive to NO, BAY 41-2272, and the combination of both (Figs. 2G, 3A, 3B, and Table I). Incubation with 10 M BAY 58-2667 led to an activation of 13.2-fold increasing to 16.0-fold upon addition of ODQ (Fig. 2G). A saturable activation could not be reached within the tested concentration range. According to the calculated concentration-response curves, EC 50 values of about 8 M and a maximal stimulation factor of 25 were assumed (Table I). The exchange of both residues with alanine abolished any sGC activation by NO, BAY 41-2272, and BAY 58-2667 within the tested concentration range (Figs. 2H, 3A, and 3B).
PPIX Reconstitution-To explore the influence of the tyrosine 135 and the arginine 139 for the binding of the heme moiety, reconstitution studies with protoporphyrin IX (PPIX), which mimics the nitrosyl-heme complex, were performed with WT-, ␤Y135F-, and ␤R139L-sGC (Fig. 3C). Due to its activating effect PPIX represents a useful tool to investigate sGC reconstitution (19,24,29). Reconstitution was performed in the presence Tween 20 for both the removal of the native heme moiety and to facilitate the subsequent reconstitution with PPIX (30). Furthermore, BAY 41-2272 was added to the incubation buffer to amplify the sGC activation upon PPIX reconstitution (23, 24). WT-sGC could be reconstituted with increasing concentrations of PPIX leading to a maximal sGC activation of 79.6-fold with an EC 50 of 3.1 M in the presence of 1 M BAY 41-2272 (Fig. 3C). In contrast, 30 M PPIX was required to achieve even a slight increase (3.1-fold) in the activity of the ␤R139L mutant. The ␤Y135F exchange was not responsive even at the highest applied concentration of PPIX (Fig. 3C).
Spectroscopic Studies-To explore potential interactions between the heme moiety of sGC and BAY 58-2667, the enzyme was incubated with BAY 58-2667 in the absence and presence of ODQ and separated subsequently by ion exchange chromatography. The native ferrous sGC showed the Soret band at 431 nm that was shifted to 392 nm after oxidizing the heme by addition of ODQ (Fig. 4A). Incubation of the native sGC with 10 M BAY 58-2667 did not result in any shift of the Soret peak (Fig. 4B). Incubation of sGC with 10 M BAY 58-2667 in the presence of ODQ led to the removal of the prosthetic heme moiety as did incubation with 100 M BAY 58-2667 without additional ODQ (Fig. 4B).
sGC Activity Assay-Purified WT-sGC was incubated with increasing concentrations of BAY 58-2667 from 100 pM to 200 M in the absence and presence of 10 M ODQ. As shown in Fig.  4C, a biphasic activation of sGC was observed in the absence of ODQ. The first step displayed a concentration-dependent activation of the enzyme from 1 nM to 50 nM with an EC 50 of 3.6 nM followed by a phase of slight increase of sGC activation with increasing concentrations of BAY 58-2667 up to about 3 M. Thereafter, a second phase of sGC activation started from 3 M until the maximal solubility of this compound was reached (200 M). In the presence of ODQ a sigmoidal concentration response curve with an EC 50 of 9.6 nM was observed reaching a maximal specific activity of 19,050 nmol⅐min Ϫ1 ⅐mg Ϫ1 .

DISCUSSION
Here we report the identification of amino acids crucial for binding of the prosthetic heme moiety to sGC as well as for NO-independent sGC activation through BAY 58-2667. Based on these findings, we propose a model for the BAY 58-2667induced activation of sGC summarized in Fig. 5. This model might also be useful to understand the NO-driven activation of the enzyme via a signal transmission triad consisting of the histidine 105, tyrosine 135, and arginine 139 located within the ␤ 1 -subunit of sGC. These residues might be involved in the transduction of heme status and porphyrin geometry upon NO binding into alterations of sGC catalytic activity.
In the early 1980s, studies with different porphyrin derivatives suggested that the propionic acid groups of the porphyrin interact with basic residues of the enzyme (19). More than two decades later our knowledge of the heme binding domain and the intramolecular signal transduction has advanced little due to the lack of any crystal structure of the enzyme (31). Moreover, elucidating the function of this domain by mutagenesis studies failed due to the lack of compounds capable of activating the heme-free enzyme. Using a novel cGMP reporter cell line that obviated the need to purify sGC mutants together with the newly discovered heme-independent sGC activator BAY 58-2667 (14), the heme-dependent sGC stimulator BAY 41-2272, NO and the sGC inhibitor ODQ, it was possible for the first time to distinguish between heme-containing and hemefree sGC within their cytosolic environment. This cGMP reporter cell, stably transfected with a cGMP-gated cation channel and aequorin, transduces intracellular cGMP concentrations via Ca 2ϩ influx into bioluminescence that can be easily measured through a charge-coupled device camera.
A starting point for our work was the hypothesis of Ignarro and coworkers (19), who postulated positively charged or polar amino acids of sGC as counterparts interacting with the propionic acid groups of the heme moiety. The N-terminal 200 residues of the ␤ 1 -subunit form a domain capable of binding heme with spectral characteristics comparable to that of the native enzyme, suggesting that the postulated amino acids probably reside in this subunit (27,28). Consequently, we initiated a BLAST search and included in the subsequent multisequence alignment all sequences of proteins known to bind heme. Very recently this approach was supported by Iyer and coworkers (32) who postulated an ancient heme-NO binding domain in the N-terminal region of the ␤ 1 -subunit of sGC that could be found even in prokaryotic organisms (32). Two identified invariant amino acids in this alignment that fit the prediction of Ignarro and coworkers (␤Y 135 and ␤R 139 ) were exchanged by alanine, and the ability of these muteins to bind heme was evaluated by transfection into the cGMP readout cell.
As controls for the analysis we used the WT-sGC, which contains heme, and ␤H105F, which does not (16 -18, 30). The cGMP readout cell transiently transfected with WT-sGC displayed an activation profile similar to that of the isolated enzyme responding to DEA/NO and/or BAY 41-2272 (13,25,33). Additionally, incubation with BAY 58-2667 led to an increase in sGC activity that was potentiated in the presence of ODQ as described for the purified enzyme (14,15). Conversely, the cGMP readout cell transfected with the ␤H105F mutant behaved as expected for the heme-free apo-enzyme in that it responded to neither NO nor BAY 41-2272 (16 -18, 30). Finally, the lack of potentiation of the BAY 58-2667-induced activity by ODQ is further indication of the loss of the heme moiety, which is essential for the oxidizing effect of ODQ (14, 34 -36). Basal activity of this mutant was slightly higher compared with the WT-sGC resulting in lower activation factors in agreement with findings of Martin and coworkers (18).
Having established that the cellular detection system is a sensitive and reliable method to distinguish between hemecontaining and heme-free sGC in its cytosolic environment, we evaluated the various novel muteins. The replacements of the tyrosine 135 (␤Y135A and ␤Y135F) and arginine 139 (␤R139A and ␤R139L) resulted in enzymes that were not responsive to NO, BAY 41-2272, or the combination of both. BAY 58-2667mediated enzyme activation was observed, but was not further potentiated by ODQ. Therefore, it can be assumed that the removal of the hydroxyl-group of the tyrosine 135 or the positive charge of the arginine 139 led to the expression of hemefree sGC. Interestingly, substituting the conserved aromatic amino acids surrounding the tyrosine 135, such as histidine 134 and tyrosine 136 with alanine had no effect on heme binding to sGC (data not shown). This is in agreement with findings described by others (16,17). The importance of tyrosine 135 and arginine 139 for heme binding was also confirmed by PPIX reconstitution studies. PPIX shares the propionic acid moieties with the native heme and might be expected to share also the requirement for the two identified residues. In contrast to the heme-depleted WT-sGC, the mutants ␤Y135F and ␤R139L could not be reconstituted with PPIX even at micromolar concentrations. These observations are summarized in Fig. 5. The sGC prosthetic heme is embedded in a hydrophobic binding pocket between the histidine 105, tyrosine 135, and the arginine 139. Comparable interactions were described for various other heme binding proteins such as cytochrome P-450, cytochrome b, or FixL (37)(38)(39).
Interestingly, the exchange of tyrosine 135 and arginine 139 not only caused the loss of heme binding by sGC but also reduced the potency of BAY 58-2667. In contrast, although the heme was lost from ␤H105F, the potency of BAY 58-2667 was undiminished. In our recent work (14) we postulated a model with two binding sites for BAY 58-2667 at the sGC: one high affinity binding site independent from the presence and oxida- tion state of the heme and a second one, which changes its affinity for BAY 58-2667 from low to high upon oxidation or removal of the heme moiety of sGC. For the putative hemeindependent binding site a K D value of 3.2 nM was determined and a saturation of binding was reached at concentrations of about 50 nM (14). These results are in agreement with the first step of the biphasic activity profile showing an EC 50 value of 3.6 nM and a saturable activation of the enzyme at about 50 nM indicating that this putative binding site for BAY 58-2667 may serve as an explanation for the activation of the native ferrous sGC at nanomolar concentrations. The second postulated binding site for BAY 58-2667 changes its affinity from low to high upon oxidation or removal of the heme moiety (14). Under the latter conditions there was also a doubling of maximal binding in the receptor binding study, an increase in sGC activation, and a shift in the photoaffinity labeling pattern of a derivative of BAY 58-2667 from the ␣ 1 -to the ␤ 1 -subunit (14). At the time these observations were first reported, they were difficult to interpret. The results reported in this work may complete the puzzle.  Because a space-filling electronic model of BAY 58-2667 closely resembles that of the heme moiety (Fig. 6), we hypothesize that the second binding site of BAY 58-2667 is in the heme pocket and that the carboxylic groups in BAY 58-2667 interact with tyrosine 135 and arginine 139 in place of the heme propionic groups. To test this hypothesis, sGC was incubated with BAY 58-2667 in the presence and absence of ODQ. Heme that became dissociated from sGC was removed by ion exchange chromatography before UV-visible spectra were recorded. The Soret peak remained unchanged after incubation with BAY 58-2667 (10 M), however, in the presence of additional ODQ a loss of the heme moiety was observed. Interestingly, a further increase of the concentration of BAY 58-2667 up to 100 M rendered the enzyme heme deficient even in its ferrous state. Oxidation of the heme by the sGC inhibitor ODQ is known to change the secondary structure of sGC (40,41) and weaken the binding of the heme moiety (5) as described for other heme proteins such as myoglobin (42)(43)(44). Therefore, the ferric heme could be displaced at lower concentrations of BAY 58-2667, whereas for the displacement of the tightly bound ferrous heme higher concentrations of BAY 58-2667 were needed.
These findings are consistent with the biphasic activation curve of BAY 58-2667 on the ferrous form that might be explained by a model of sGC activation with two high affinity binding sites for BAY 58-2667: one that is saturable at nanomolar concentrations and shows no direct interference with the heme moiety, and a second one that exhibits a direct competition between BAY 58-2667 and the prosthetic group for the tyrosine 135 and the arginine 139. Within this model the biphasic activation profile observed for BAY 58-2667 at the native ferrous sGC can be easily explained. At concentrations up to 50 nM BAY 58-2667 the putative heme-independent high affinity binding site may be saturated in agreement with the findings of the receptor binding studies reported earlier (14), whereas the second binding site remains occupied by the heme moiety of sGC. Total displacement of the heme finally happened at high micromolar concentrations in agreement with the spectroscopic studies. In contrast, in the presence of ODQ the weakly bound ferric heme was displaced immediately by BAY 58-2667 resulting in the observed strong one-step activation of the enzyme. However, until this model is confirmed by cocrystallization studies, it cannot be excluded that the activation of the native heme-containing enzyme contributes to a small percentage of heme-free sGC in the enzyme preparation. The competition between the heme and BAY 58-2667 together with the observed involvement of the tyrosine 135 and arginine 139 in the sGC activation by this compound implicate the possibility that the heme propionate interaction with these residues may be more than a simple rigid coordination. It might be that this interaction is part of intramolecular signal transmission cascades involved in translating changes of the porphyrin geometry and heme status into structural changes of the protein as observed for other heme proteins (38). This proposed signal transmission triad consisting of histidine 105, tyrosine 135, and arginine 139 may also explain why sGC is activated by PPIX or CO to a lesser extent than by NO, namely due to differences in porphyrin geometry. CONCLUSION Using the novel cGMP reporter cell and the recently described heme-independent sGC activator BAY 58-2667, we were able to distinguish between heme-free and heme-containing sGC in an intact cellular system. In addition to the established histidine 105, we identified two additional residues, tyrosine 135 and arginine 139, to be crucially important for heme binding to sGC. Moreover, the heme-dependent high affinity binding site of BAY 58-2667 was characterized, exhibiting a direct competition between this compound and the heme moiety of sGC. Based on these findings we postulate a signal transmission triad composed of histidine 105, tyrosine 135, and arginine 139 in the ␤ 1 -subunit of sGC, responsible for the BAY 58-2667-induced sGC activation and probably involved in transducing changes in the heme geometry upon NO binding into alterations of the catalytic rate of the enzyme. These results await confirmation via crystallization-and cocrystallization studies with sGC and BAY 58-2667. Meanwhile, the use of different sGC activators, such as BAY 58-2667 and BAY 41-2272, with sequence alignments and mutagenesis studies, seems to be a promising approach to elucidate mechanisms of sGC activation and intramolecular signaling.