Portal de Periódicos da CAPES

A Functional Domain of the α1 Subunit of Soluble Guanylyl Cyclase Is Necessary for Activation of the Enzyme by Nitric Oxide and YC-1 but Is Not Involved in Heme Binding

2003; Elsevier BV; Volume: 278; Issue: 14 Linguagem: Inglês

10.1074/jbc.m212740200

1083-351X

Markus Koglin, Sönke Behrends,

Aldose Reductase and Taurine

Soluble guanylyl cyclase is a heterodimeric enzyme consisting of an α1 and a ॆ1subunit and is an important target for endogenous nitric oxide and the guanylyl cyclase modulator YC-1. The activation of the enzyme by both substances is dependent on the presence of a prosthetic heme group. It has been unclear whether this prosthetic heme group is sandwiched between the α1 and ॆ1 subunits or whether it exclusively binds to the ॆ1 subunit. Here we analyze progressive amino-terminal deletion mutants of the human α1 subunit after co-expression with the human ॆ1 subunit in the baculovirus/Sf9 system. Spectral, biochemical, and pharmacological analysis shows that the first 259 amino acids of the α1 subunit can be deleted without loss of sensitivity to nitric oxide (NO) or YC-1 or loss of heme binding of the respective enzyme complex with the ॆ1subunit. This is in contrast to previous data indicating that NO sensitivity and a functional heme binding site requires full-length amino termini of bovine α1 and ॆ1 subunits. Further deletion of the first 364 amino acids of the α1subunit leads to an enzyme complex with preserved heme binding but loss of sensitivity to NO or YC-1 despite induction of the typical spectral shift by NO binding to the prosthetic heme group. We conclude that 1) the amino-terminal part of the α1 subunit is not involved in heme binding and 2) amino acids 259–364 of the α1 subunit represent an important functional domain for the transduction of the NO activation signal and likely represent the target for NO-sensitizing substances like YC-1. Soluble guanylyl cyclase is a heterodimeric enzyme consisting of an α1 and a ॆ1subunit and is an important target for endogenous nitric oxide and the guanylyl cyclase modulator YC-1. The activation of the enzyme by both substances is dependent on the presence of a prosthetic heme group. It has been unclear whether this prosthetic heme group is sandwiched between the α1 and ॆ1 subunits or whether it exclusively binds to the ॆ1 subunit. Here we analyze progressive amino-terminal deletion mutants of the human α1 subunit after co-expression with the human ॆ1 subunit in the baculovirus/Sf9 system. Spectral, biochemical, and pharmacological analysis shows that the first 259 amino acids of the α1 subunit can be deleted without loss of sensitivity to nitric oxide (NO) or YC-1 or loss of heme binding of the respective enzyme complex with the ॆ1subunit. This is in contrast to previous data indicating that NO sensitivity and a functional heme binding site requires full-length amino termini of bovine α1 and ॆ1 subunits. Further deletion of the first 364 amino acids of the α1subunit leads to an enzyme complex with preserved heme binding but loss of sensitivity to NO or YC-1 despite induction of the typical spectral shift by NO binding to the prosthetic heme group. We conclude that 1) the amino-terminal part of the α1 subunit is not involved in heme binding and 2) amino acids 259–364 of the α1 subunit represent an important functional domain for the transduction of the NO activation signal and likely represent the target for NO-sensitizing substances like YC-1. soluble guanylyl cyclase 2,2-diethyl-1-nitroso-oxyhydrazine triethanolamine Soluble guanylyl cyclase (sGC)1 is an important target for endogenous nitric oxide (NO), NO-releasing drugs like glyceryl trinitrate, and novel substances like YC-1 or BAY 41-2272 that sensitize the enzyme for activation by NO (1Lucas K.A. Pitari G.M. Kazerounian S. Ruiz-Stewart I. Park J. Schulz S. Chepenik K.P. Waldman S.A. Pharmacol. Rev. 2000; 52: 375-414Google Scholar, 2Stasch J.-P. Becker E.M. Alonso-Alija C. Apeler H. Dembowsky K. Feurer A. Gerzer R. Minuth T. Perzborn E. Pleiss U. Schroeder H. Schroeder W. Stahl E. Steinke W. Straub A. Schramm M. Nature. 2001; 410: 212-215Google Scholar). The enzyme has been purified from lung as a heterodimeric, heme-containing enzyme consisting of an α1 and a ॆ1 subunit. After cloning of the α1 and ॆ1 cDNAs, two other subunit cDNAs have been cloned by homology screening: the ॆ2 subunit from rat kidney and the α2subunit from human fetal brain (3Yuen P.S.T. Potter L.R. Garbers D.L. Biochemistry. 1990; 29: 10872-10878Google Scholar, 4Harteneck C. Wedel B. Koesling D. Malkewitz J. Böhme E. Schultz G. FEBS Lett. 1991; 292: 217-222Google Scholar). Co-expression of the α1/ॆ1 and α2/ॆ1cDNAs yielded NO-sensitive enzymes in expression systems (4Harteneck C. Wedel B. Koesling D. Malkewitz J. Böhme E. Schultz G. FEBS Lett. 1991; 292: 217-222Google Scholar), and the α2/ॆ1 heterodimeric enzyme has been demonstrated on the protein level in human placenta by co-precipitation experiments (5Russwurm M. Behrends S. Harteneck C. Koesling D. Biochem. J. 1998; 335: 125-130Google Scholar). We recently isolated a ॆ2 cDNA variant from rat kidney that shows NO-sensitive enzyme activity after expression in Sf9 or HEK-293 cells in the absence of a second subunit, most likely as ॆ2/ॆ2 homodimer (6Koglin M. Vehse K. Budaeus L. Scholz H. Behrends S. J. Biol. Chem. 2001; 276: 30737-30743Google Scholar). Most studies regarding the activation mechanism by binding of NO to the prosthetic heme group have concentrated on the α1/ॆ1 heterodimeric enzyme. Since the first purification of this enzyme isoform it has been assumed that the enzyme contains one prosthetic heme group per heterodimer (7Gerzer R. Böhme E. Hofmann F. Schultz G. FEBS Lett. 1981; 132: 71-74Google Scholar). Before the cDNA sequences of the two subunits were identified, it was proposed that one subunit was regulatory and bound the heme and that the other subunit was catalytic (8Kamisaki Y. Saheki S. Nakane M. Palmieri J.A. Kuno T. Chang B.Y. Waldman S.A. Murad F. J. Biol. Chem. 1986; 261: 7236-7241Google Scholar). However, analysis of the cDNA sequences revealed that the two subunits show a high degree of homology both in their amino-terminal and their carboxyl-terminal halves (9Koesling D. Harteneck C. Humbert P. Bosserhoff A. Frank R. Schultz G. Böhme E. FEBS Lett. 1990; 266: 128-132Google Scholar, 10Nakane M. Arai K. Saheki S. Kuno T. Buechler W. Murad F. J. Biol. Chem. 1990; 265: 16841-16845Google Scholar). While the carboxyl-terminal parts were assigned as being responsible for catalysis based on homology to the related adenylyl cyclases, it seemed plausible that both homologous amino-terminal regions of the α1 and ॆ1 subunits participate in binding of the prosthetic heme. This hypothesis was strengthened by findings using amino-terminal deletion mutants of the bovine α1and ॆ1 subunits showing that NO sensitivity and a functional heme binding site of sGC requires full-length amino termini of both subunits (11Wedel B. Harteneck C. Foerster J. Friebe A. Schultz G. Koesling D. J. Biol. Chem. 1995; 270: 24871-24875Google Scholar, 12Foerster J. Harteneck C. Malkewitz J. Schultz G. Koesling D. Eur. J. Biochem. 1996; 240: 380-386Google Scholar). In the present study, we used deletion mutagenesis to identify functional regions that are responsible for NO-heme and YC-1-mediated activation of sGC. To our surprise we found that the deletion of the first 259 amino acids of the human α1 subunit leads to an enzyme with strong sensitivity toward the heme-dependent activators NO and YC-1. Deletion of 364 amino acids of the α1 subunit leads to an enzyme complex that is insensitive to the heme-dependent activators NO and YC-1 but shows preserved heme binding with the typical shift in the spectral analysis by NO binding to the prosthetic heme group. This indicates that the amino-terminal part of the α1 subunit is not involved in heme binding and that amino acids 259–364 of the α1subunit represent an important functional domain for the transduction of the NO activation signal and likely represent the target for NO-sensitizing substances like YC-1. 3-(5′-Hydroxymethyl-2′-furyl)-1-benzylindazole (YC-1) was from Alexis Biochemicals (Lausen, Switzerland). 2,2-Diethyl-1-nitroso-oxyhydrazine (DEA/NO) and all other chemicals, in the highest grade of purity, were obtained from Sigma. Products for Sf9 cell culture were from Invitrogen. Cloning of the α1 subunit (a kind gift of Dr. Georges Guellaën, Créteil; Ref. 13Giuili G. Scholl U. Bulle F. Guellaen G. FEBS Lett. 1992; 304: 83-88Google Scholar) and the ॆ1 subunit has been described previously (6Koglin M. Vehse K. Budaeus L. Scholz H. Behrends S. J. Biol. Chem. 2001; 276: 30737-30743Google Scholar, 14Koglin M. Behrends S. Biochim. Biophys. Acta. 2000; 1494: 286-289Google Scholar). Cloning of diverse α1 deletion mutants was carried out by digestion with different restriction endonucleases. For construction of the α1 ΔN364 mutant, aBsmFI/HindIII fragment of full-length α1 was cloned using StuI/HindIII into the pFASTBAC vector. Before cloning into pFASTBAC theBsmFI 5′-ends were filled in with Taq polymerase (Invitrogen). For construction of the α1ΔN259 mutant, a single nucleotide exchange was done with the QuikChangeTM kit (Stratagene, La Jolla, CA) using the following primer pair: P 216, 5′-GCG AGT TTG TGA ATC AGC CCT ACTAGT TGT ACT CCG-3′; and P 217, 5′-CGG AGT ACA ACT AGT AGG GCT GAT TCA CAA ACT CGC-3′. The modified nucleotide is underlined. A SpeI/HindIII fragment of the mutated α1 full-length clone was then ligated using SpeI/HindIII into the pFASTBAC vector. Recombinant baculoviruses of respective subunits were generated according to the BAC-TO-BACTM System (Invitrogen). Sf9 cells were cultured in Sf-900 II serum-free medium supplemented with 17 penicillin/streptomycin and 107 fetal calf serum. Spinner cultures were grown to a cell density of 3.0 × 106 cells/ml and then diluted to 1.2 × 106 cells/ml for infection. 30 ml of cell solution were infected with multiplicities of infection of 1. After 74 h cells were harvested and collected by centrifugation (1500 × g for 10 min at 4 °C). All following steps were performed at 4 °C or on ice. The cell pellet was resuspended in 4 ml of homogenization buffer containing 50 mm TEA/HCl, pH 7.6, 0.2 ॖm benzamidine, 1 mm EDTA, pH 8.0, and freshly dissolved dithiothreitol with a final concentration of 10 mm. The cells were passed through a Sterican® needle (0.45 × 25 mm, B. Braun, Melsungen, Germany) several times for lysis. To remove complete cells and nuclei the solution was centrifuged for 2 min at 800 × g. Cytosolic fractions were obtained by a second centrifugation step for 30 min at 40,000 ×g. All purification steps were performed at 4 °C. The cell pellet from 1800 ml of cell solution infected with the respective subunits was homogenized with a cell disruption bomb (Parr, Moline, IL) at 60 bars for 1 h in 180 ml of 50 mm TEA/HCl, pH 8.0 containing 10 mmdithiothreitol, 1 mm benzamidine, 10 ॖg/ml phenylmethylsulfonyl fluoride, and 900 ॖl of protease inhibitor mixture (Sigma). The homogenate was centrifuged at 40,000 ×g for 30 min, and 180 ml of supernatant were collected. All chromatographic steps were performed on a FPLC system (Amersham Biosciences). The protease inhibitor benzamidine (1 mm), dithiothreitol (10 mm), and phenylmethylsulfonyl fluoride (10 ॖg/ml) were used in all chromatographic steps. The supernatant was immediately applied to a Q-Sepharose column (20-ml volume) at 2 ml/min. Ion exchange buffer A contained 50 mm TEA/HCl, pH 8.0. Ion exchange buffer B contained 5 mm potassium phosphate (pH 7.2). Ion exchange buffer C was prepared by adding 1m NaCl to buffer B. The column was washed at 3 ml/min with buffer A, buffer B, and 87 buffer C untilA280 was stable. A linear gradient from 87 C to 307 C for 828 ml was used to elute sGC. The sGC-containing fractions were pooled by determining sGC activity at basal or NO-stimulated conditions after each column. The pooled fractions (104 ml) were diluted with 104 ml of 5 mm potassium phosphate (pH 7.2) and applied immediately to a ceramic hydroxyapatite column (Bio-Rad, 5 ml volume) at 1.5 ml/min. Hydroxyapatite buffer A contained 5 mm potassium phosphate (pH 7.2), and hydroxyapatite buffer B contained 400 mm potassium phosphate (pH 6.6). The column was then washed with 107 B until the A280 was stable. The enzyme was eluted with a linear gradient running from 107 B to 507 B for 180 ml. The sGC-containing fractions (44 ml) were again pooled and applied immediately to a blue Sepharose column (Amersham Biosciences, 5 ml volume) at 1.5 ml/min. Blue Sepharose buffer A contained 5 mm potassium phosphate (pH 7.2), and blue Sepharose buffer B was prepared by adding 1 m NaCl to buffer A. The column was then washed with 107 buffer B until theA280 was stable. The enzyme was eluted with a linear gradient running from 107 B to 1007 B. The sGC-containing fractions (18 ml) were again pooled and concentrated in centrifugal devices with a 50-kDa cut-off (Millipore, Bedford, MA) to 1.5 ml. The enzyme was then loaded on a Superdex 200 column (Amersham Biosciences, 60 × 2.6 cm) and eluted overnight with 50 mm TEA/HCl, pH 8.0 containing 250 mm NaCl at 0.15 ml/min. Fractions with the highest sGC activity were pooled and concentrated as described above to a final volume of ∼200 ॖl. For spectroscopic measurements 100 ॖl of purified enzyme were used. Purified enzyme was diluted with 50 mm TEA/HCl, pH 8.0 containing 250 mm NaCl and stored with 107 (v/v) glycerol at −80 °C. Protein concentrations were determined by the method of Bradford using bovine plasma gamma globulin (Protein Assay Standard I, Bio-Rad) as standard. sGC activity of Sf9 cytosol (approximately 40 ॖg of protein per assay tube) or purified protein (50 ng of protein per assay tube) was determined by incubation for 10 min at 37 °C in the presence of 1 mm cGMP, 0.5 mm [32P]GTP (about 0.2 ॖCi), 3 mm MgCl2, 50 mm TEA/HCl, pH 7.4, 0.25 g/liter creatine kinase, 5 mm creatine phosphate, and 1 mm 3-isobutyl-1-methylxanthine in a total volume of 0.1 ml as described by Schultz and Böhme (15Schultz G. Böhme E. Bergmeyer H.U. Bergmeyer J. Grassl M. 3rd Ed. Methods of Enzymatic Analysis. 4. Verlag Chemie, Weinheim, Germany1984: 379-389Google Scholar). Reactions were started by the addition of protein and incubation at 37 °C. All experiments were stopped by ZnCO3 precipitation, and purification of the enzyme-formed cGMP was performed as described previously (15Schultz G. Böhme E. Bergmeyer H.U. Bergmeyer J. Grassl M. 3rd Ed. Methods of Enzymatic Analysis. 4. Verlag Chemie, Weinheim, Germany1984: 379-389Google Scholar). Basal enzyme activity measurements were performed in the absence of NO or YC-1. NO-stimulated measurements were performed in the presence of the NO donor DEA/NO, and NO/YC-1-stimulated enzyme activity measurements were performed in the presence of both DEA/NO and YC-1 in variable concentrations. YC-1 was dissolved in 257 (v/v) Me2SO so that the final Me2SO concentration in the enzyme assay did not exceed 2.57 (v/v). At this concentration no effects of Me2SO on enzyme activity were observed. DEA/NO was dissolved in 10 mm NaOH, which also did not affect the enzyme activity. The α1-1200 antibody was raised against two peptides (EP012493, H2N-FTPRSREELPPNFP-COOH; and EP012494, H2N-CFQKKDVEDGNANFLGKASGID-COOH) of the carboxyl-terminal domain of the human α1 subunit, and the ॆ1-89 antibody was raised against the carboxyl-terminal peptide (EP990255, H2N-CSRKNTGTEETKQDDD-COOH). Antibodies were coupled by an additional cysteine to keyhole limpet hemocyanin. Rabbits were immunized on days 0, 14, 28, and 56 and were finally bled on day 80. Successful antigen response was estimated by enzyme-linked immunosorbent assay. For monitoring the purity of enzyme preparations and for the determination of apparent molecular masses of the purified enzyme, SDS-polyacrylamide gel electrophoresis was performed in 107 slab gels, and proteins were stained with Coomassie Blue G-250. For immunoblotting, protein fractions were subjected to 107 SDS-PAGE and then transferred electrophoretically to a nitrocellulose membrane. The membrane was reversibly stained with Ponceau S, and unspecific binding sites were saturated by immersing the membrane for 1 h in TBST buffer (10 mm Tris/HCl, pH 8.0, 150 mmNaCl, 0.057 Tween 20) containing 57 nonfat dry milk. The membranes were incubated for 1.5 h in TBST buffer containing α1-1200 and ॆ1-89 in a 1:1000 dilution and 0.57 dry milk. Negative control reactions were run in the presence of synthetic peptides used for immunization in different combinations (5 ॖg/ml). The membranes were washed three times for 10 min with TBST and subsequently incubated for 1 h with horseradish peroxidase-labeled anti-rabbit IgG antibodies (diluted 1:4000, Sigma). After three washes with TBST the membranes were processed with the ECL Western blotting detection system according to the recommendations of the manufacturer (Amersham Biosciences). All results were controlled for their statistical significance by one-way analysis of variance followed by a Newman-Keuls post test. A value of p < 0.05 was considered to be statistically significant. To determine the function of the amino-terminal part of the α1 subunit a series of recombinants containing progressive deletions of the amino-terminal sequences of α1 were constructed and expressed in Sf9 cells together with the dimerizing subunit ॆ1. On Western immunoblots using an antibody directed against a carboxyl-terminal sequence of the α1 subunit, full-length α1and α1 deletion mutants (α1ΔN259 and α1 ΔN364) exhibited molecular masses corresponding to those predicted from their deduced amino acid sequences (Fig. 1). Expression levels of the full-length α1, α1 deletion mutants, and ॆ1 were very similar (see Fig. 1). Guanylyl cyclase activity was measured in the respective cytosols from Sf9 cells under basal conditions, activation with NO, and activation with the combination of NO and YC-1 (Fig.2). Guanylyl cyclase activity was similar under all experimental conditions for α1 and the α1 deletion mutant α1 ΔN259. The deletion mutant α1 ΔN364 showed a complete loss of sensitivity toward NO or YC-1 and a slight decrease of guanylyl cyclase activity under basal conditions (see Fig. 2).Figure 2Guanylyl cyclase activity in the respective cytosolic fractions of Sf9 infected cells. Guanylyl cyclase activity was measured under basal conditions (only 3 mmMg2+, black columns), in the presence of 100 ॖm DEA/NO (white columns), or in the presence of additional 100 ॖm YC-1 (gray columns). The columns represent means ± S.E. of at least four independent experiments.View Large Image Figure ViewerDownload (PPT) Concentration response experiments were done using the NO donor DEA/NO (Fig. 3). The EC50 values for DEA/NO showed no significant differences and were 421 ± 69 nm for full-length α1 and 585 ± 246 nm for α1 ΔN259. α1 ΔN364 was NO-insensitive. Concentration-response curves for YC-1 were performed both in the absence (Fig. 4A) and presence (Fig. 4B) of DEA/NO (100 ॖm). The EC50 values for YC-1 showed no significant differences for full-length α1 and α1 ΔN259and were 25 ± 7 and 19 ± 6 ॖm in the absence and 0.89 ± 0.05 and 1.94 ± 0.64 ॖm in the presence of DEA/NO, respectively. α1ΔN364 was YC-1-insensitive (see Fig. 4A).Figure 4Concentration-dependent effect of YC-1 on guanylyl cyclase activity in Sf9 cell cytosol after infection with the respective α1subunits in combination with ॆ1. The results show the effect of YC-1 in a range between 0.1 and 300 ॖm YC-1. Ashows dose-response curves for α1 (closed circles), α1 ΔN259 (open circles), and α1 ΔN364 (closed squares) for basal guanylyl cyclase activity, whereas Bshows DEA/NO (100 ॖm)-stimulated curves. Enzymatic activity of α1 ΔN364 was measured only for the highest YC-1 concentration at basal conditions. All points of YC-1-dependent curves represent means (±S.E.) of at least three independent experiments performed in duplicate.View Large Image Figure ViewerDownload (PPT) All mutant and wild type enzymes were purified to apparent homogeneity. Coomassie Blue-stained SDS-PAGE analyses are shown in Fig.5. Spectroscopic analysis of the purified wild type enzyme revealed absorption maxima at 432 nm in the absence and 400 nm in the presence of the NO donor DEA/NO (100 ॖm) (Fig. 6A). Analysis of the purified α1ΔN259/ॆ1 enzyme revealed an almost identical spectrum with very similar absorption maxima (431 nm and 399 nm, respectively; see Fig. 6B). Purified α1ΔN364/ॆ1 enzyme showed absorption maxima at 432 nm in the absence and 399 nm in the presence of the NO donor DEA/NO (Fig. 6C). Although these maxima were almost identical to the wild type enzyme, the ratio of the absorption at 432 nm to 280 nm was lower, indicating lower heme content (see Fig. 6C). During the purification of the wild type and the α1ΔN259/ॆ1 enzyme, fractions from each column were pooled by determining sGC activity at basal and NO-stimulated conditions. For the NO-insensitive α1ΔN364/ॆ1 enzyme fractions could only be tested for basal enzyme activity after each column. Thus we selected for heme-containing, NO-sensitive enzyme in the case of wild type and α1 ΔN259/ॆ1 enzyme. To control whether this effect accounts for the lower amount of heme in α1 ΔN364/ॆ1 enzyme, we purified α1 ΔN259/ॆ1 enzyme and assayed the fractions after each column only for basal enzyme activity. This resulted also in a significantly lower ratio of the absorption at 432 nm to 280 nm but absorption maxima very similar to those obtained before (432 nm in the absence and 399 nm in the presence of DEA/NO).Figure 6Spectroscopic analysis of purified guanylyl cyclase enzyme complexes. Spectroscopic analysis shows relative absorption values at basal (solid line) or NO-stimulated (100 ॖm DEA/NO, dotted line) conditions.A and B show the spectra of wild type enzyme α1/ॆ1 (A, 0.27 ॖg/ॖl protein) and the deletion mutant α1ΔN259/ॆ1 (B, 0.28 ॖg/ॖl protein). Both spectra were the result of a purification controlled with NO-stimulated guanylyl cyclase activity for pooling after each column (see 舠Experimental Procedures舡). C andD show spectra of the deletion mutants α1ΔN364/ॆ1 (C, 0.87 ॖg/ॖl protein) and α1 ΔN259/ॆ1(D, 0.33 ॖg/ॖl protein), respectively. The spectra were the results of a purification controlled with basal guanylyl cyclase activity for pooling after each column (see 舠Experimental Procedures舡).View Large Image Figure ViewerDownload (PPT) To compare the kinetic properties of the purified enzyme complexes, cGMP formation was determined in the presence of increasing GTP concentrations. A Lineweaver-Burk plot of the data revealed apparentKm values that showed no significant differences between full-length α1 (134 ± 19 ॖm), α1 ΔN259 (119 ± 13 ॖm), and α1 ΔN364 (163 ± 12 ॖm) (Fig. 7).Vmax values showed no significant differences and were 145 ± 16 nmol of cGMP/min × mg for full-length α1, 104 ± 15 nmol of cGMP/min × mg for α1 ΔN259, and 98 ± 15 nmol of cGMP/min × mg for α1 ΔN364. Guanylyl cyclase activity of the purified enzymes was measured under basal conditions and activation with NO to investigate the status of the enzymes (Fig. 8). Guanylyl cyclase activity was not significantly different under all experimental conditions for α1 and the α1 deletion mutant α1 ΔN259. In the presence of the NO donor DEA/NO (100 ॖm) enzyme activity was increased by 242-fold for α1 and 252-fold for α1ΔN259. Analysis of the purified deletion mutant α1 ΔN364 confirmed the complete loss of sensitivity toward NO and demonstrated a slight decrease of guanylyl cyclase activity under basal conditions (see Fig. 8). Concentration-response curves for YC-1 were performed both in the absence (Fig. 9A) and presence (Fig. 9B) of a submaximally active DEA/NO concentration (100 nm). The EC50 values for YC-1 showed significant differences between full-length α1 and α1 ΔN259 and were 51 ± 7 and 30 ± 5 ॖm (p < 0.05) in the absence and 2.62 ± 0.26 and 4.53 ± 0.71 ॖm(p < 0.05) in the presence of DEA/NO (100 nm), respectively. α1ΔN364 was YC-1-insensitive (see Fig. 9A).Vmax values showed significant differences only in the absence of DEA/NO and were 9717 ± 216 nmol of cGMP/min × mg for full-length α1 and 5745 ± 420 nmol of cGMP/min × mg for α1ΔN259 (p < 0.001) at basal conditions and 24,259 ± 3237 nmol of cGMP/min × mg for full-length α1 and 17,369 ± 3308 nmol of cGMP/min × mg for α1 ΔN259 in the presence of 100 nm DEA/NO. Previous studies have used amino-terminal deletion mutants of bovine NO-sensitive guanylyl cyclase subunits to map functional regions of this enzyme family (11Wedel B. Harteneck C. Foerster J. Friebe A. Schultz G. Koesling D. J. Biol. Chem. 1995; 270: 24871-24875Google Scholar). The deletion of only 131 amino-terminal amino acids of the α1 subunit and co-expression of this deletion mutant with the ॆ1 subunit in the baculovirus system led to a 10-fold reduction in enzyme activity in co-infected Sf9 cytosol versus the respective α1full-length construct and an almost complete loss of NO sensitivity (11Wedel B. Harteneck C. Foerster J. Friebe A. Schultz G. Koesling D. J. Biol. Chem. 1995; 270: 24871-24875Google Scholar). By contrast, in our study there was no significant decline in basal enzyme activity and stimulation by NO after deletion of 259 amino acids of the α1 subunit. It is conceivable that our approach of using endogenously occurring methionines rather than newly introduced methionines as start codons poses less risk of unwanted conformational changes resulting in lower enzyme activity. It is also possible that sequence differences between the bovine and human α1 subunits might explain the discrepancies. Subsequent to the study by Wedel and co-workers (11Wedel B. Harteneck C. Foerster J. Friebe A. Schultz G. Koesling D. J. Biol. Chem. 1995; 270: 24871-24875Google Scholar), the same group analyzed the purified α1 ΔN131/ॆ1 enzyme complex and showed a loss of affinity in binding of the prosthetic heme group, which binds NO (12Foerster J. Harteneck C. Malkewitz J. Schultz G. Koesling D. Eur. J. Biochem. 1996; 240: 380-386Google Scholar). Based on the findings of this study, it has been suggested that heme binding to sGC requires the presence of both subunits (α1 and ॆ1) in full-length and that both homologous amino-terminal regions of the α1 and ॆ1 subunits participate in binding of the prosthetic heme (12Foerster J. Harteneck C. Malkewitz J. Schultz G. Koesling D. Eur. J. Biochem. 1996; 240: 380-386Google Scholar). Expression of only the amino-terminal part of the ॆ1 subunit expressed in Escherichia coliindicated that the amino terminus of the ॆ1 subunit (ॆ1-(1–385)) is sufficient for heme binding with preserved binding of NO (16Zhao Y. Marletta M.A. Biochemistry. 1997; 36: 15959-15964Google Scholar). As pointed out by the authors of the latter study (16Zhao Y. Marletta M.A. Biochemistry. 1997; 36: 15959-15964Google Scholar), this result does not rule out an in vivo heme binding site involving residues contributed by both the α and ॆ subunits as suggested by Foerster et al. (12Foerster J. Harteneck C. Malkewitz J. Schultz G. Koesling D. Eur. J. Biochem. 1996; 240: 380-386Google Scholar). Given the degree of homology in the amino terminus of the subunits, it is possible that the ॆ1-(1–385) homodimer is an in vitro outcome of expression in the absence of the α1subunit and that the second ॆ1 subunit provides crucial amino acids for heme binding that would come from the α1subunit in vivo (16Zhao Y. Marletta M.A. Biochemistry. 1997; 36: 15959-15964Google Scholar). The results of the current study argue against an in vivo heme binding site involving amino-terminal residues contributed by both the α and ॆ subunits as suggested by Foerster et al. (12Foerster J. Harteneck C. Malkewitz J. Schultz G. Koesling D. Eur. J. Biochem. 1996; 240: 380-386Google Scholar) and a model where the heme in sGC is sandwiched between the amino-terminal parts of the two subunits (16Zhao Y. Marletta M.A. Biochemistry. 1997; 36: 15959-15964Google Scholar). The results of the current study rather indicate that the amino-terminal region of the α1 subunit is not involved in heme binding. In fact, we could fully reproduce the findings by Zhao and Marletta (16Zhao Y. Marletta M.A. Biochemistry. 1997; 36: 15959-15964Google Scholar) that the amino-terminal part of the ॆ1 subunit can be expressed as a soluble, nitric oxide-, and heme-binding protein in E. coli. 2S. Behrends, unpublished data. In contrast, the expression of longer constructs of the ॆ1 subunit including the catalytic domain or co-expression with the α1 subunit in E. coli led to insoluble protein in the form of inclusion bodies under different experimental conditions.2 These findings should encourage approaches to solve the structure of the heme binding domain of sGC by focusing on the expression of the amino-terminal part of the ॆ1 subunit. We find in the current study that deletion of 259 amino acids of the α1 subunit leaves the enzyme functionally intact but that deletion of the first 364 amino acids leads to an enzyme complex with preserved heme binding but loss of sensitivity to NO or YC-1. NO still binds to the enzyme variant lacking 364 amino acids since it induces the typical spectral shift of the prosthetic heme group. This indicates that amino acids 259- 364 of the α1 subunit are either directly important for the transduction of the activation signal by NO or that the deleted region is indirectly involved in the mediation of the NO effect, e.g. by the stabilization of another domain of the enzyme. The cysteine 238 and cysteine 243 region in the α1 subunit has been mapped as the likely binding site of the YC-1-related substance BAY 41-2272 using photoaffinity labeling (2Stasch J.-P. Becker E.M. Alonso-Alija C. Apeler H. Dembowsky K. Feurer A. Gerzer R. Minuth T. Perzborn E. Pleiss U. Schroeder H. Schroeder W. Stahl E. Steinke W. Straub A. Schramm M. Nature. 2001; 410: 212-215Google Scholar). We show in the current study that both cysteines including an additional 14 amino-terminal residues can be deleted without loss of YC-1 sensitivity. Given the data by Stasch and colleagues (2Stasch J.-P. Becker E.M. Alonso-Alija C. Apeler H. Dembowsky K. Feurer A. Gerzer R. Minuth T. Perzborn E. Pleiss U. Schroeder H. Schroeder W. Stahl E. Steinke W. Straub A. Schramm M. Nature. 2001; 410: 212-215Google Scholar), we think that it is likely that the region adjacent to cysteine 238 and cysteine 243 (from amino acid 259 to 364) of the α1 subunit represents the binding site of YC-1. The possibility that YC-1 still binds to the α1 ΔN364 mutant enzyme complex and that the activation signal is not transduced within the enzyme directly or indirectly as discussed for NO is also a plausible explanation of our results. In the present study, sGC in Sf9 cytosol is activated 25-fold, while the purified enzyme is activated 80-fold. The -fold stimulation by YC-1 in the literature varies from 7-fold (17Martin E. Lee Y.C. Murad F. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12938-12942Google Scholar), 10-fold (18Mülsch A. Bauersachs J. Schäfer A. Stasch J.-P. Kast R. Busse R. Br. J. Pharmacol. 1997; 120: 681-689Google Scholar), and 14-fold (19Friebe A. Russwurm M. Mergia E. Koesling D. Biochemistry. 1999; 38: 15253-15257Google Scholar) up to close to 100-fold (20Hoenicka M. Becker E.M. Apeler H. Sirichoke T. Schroder H. Gerzer R. Stasch J.-P. J. Mol. Med. 1999; 77: 14-23Google Scholar). YC-1 activation of sGC is to a very large degree heme-dependent (17Martin E. Lee Y.C. Murad F. Proc. Natl. Acad. Sci. U. S. A. 2001; 98: 12938-12942Google Scholar, 21Friebe A. Koesling D. Mol. Pharmacol. 1998; 53: 123-127Google Scholar), and thus -fold stimulation of YC-1 is a function of the heme content of the enzyme similar to the heme-dependent activator NO. The result that stimulation factors by YC-1 are significantly higher for the purified enzymes than in Sf9 cytosol can be explained by a higher percentage of heme-containing versus heme-free enzyme that may also form by expression in Sf9 cells. Since we have pooled the fractions during our purification protocol according to the determination of sGC activity at basal and NO-stimulated conditions, we have purified selectively heme-containing enzyme. The purified α1 ΔN364 enzyme complex contained less heme than the wild type enzyme or the α1ΔN259 enzyme complex purified under regular conditions. Because of the lack of NO sensitivity, the α1ΔN364 enzyme complex had to be purified by controlling fractions after each column for basal guanylyl cyclase activity only. In the case of the wild type enzyme and the α1ΔN259 enzyme complex we have pooled the fractions during our purification protocol according to the determination of guanylyl cyclase activity at basal and NO-stimulated conditions and have thus purified selectively heme-containing enzyme. Attempts to establish a purification protocol with wild type enzyme that would be applicable to all three enzyme variants by pooling fractions after each column according to their absorbance at 430 nm in the spectrophotometer were not successful. Especially at the crucial initial stages of the purification protocol the measurement of the absorbance at 430 nm gave no reliable results with respect to sGC-containing fractions when compared with NO-activated guanylyl cyclase activity measurements. To find out whether the relatively low heme content of the α1 ΔN364 enzyme complex was due to the impossibility of pooling fractions according to NO-activated guanylyl cyclase activity, we purified the α1ΔN259 enzyme under the same conditions. This resulted in an enzyme preparation with a similarly reduced heme content. While this argues in favor of the hypothesis that the different purification procedure is responsible for the relatively low heme content, we cannot rule out the possibility that the α1 ΔN364mutant enzyme complex shows reduced heme binding affinity. In summary, we show that 1) the amino-terminal part of the α1 subunit is not involved in heme binding and 2) amino acids 259–364 of the α1 subunit represent an important functional domain for the transduction of the NO activation signal and likely represent the target for NO-sensitizing substances like YC-1. The expert technical assistance of Jutta Starbatty, Jenny Behrens, and Alexandra Zielinski is gratefully acknowledged.