Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

NO, a short-lived bioregulatory molecule, participates in a variety of biological processes including neurotransmission, antitumorigenesis and regulation of vascular tone (Knowles and Moncada, 1994). It is synthesized from L-arginine by members of the NOSs in the presence of various cofactors as well as CaM. The activity of the constitutive forms, nNOS (Bredt et al., 1991; Bredt and Snyder, 1990; Schmidt and Murad, 1991a) and eNOS (Janssens et al., 1992; Marsden et al., 1992), is Ca⁺⁺/CaM dependent. There is also an iNOS (Stuehr et al., 1991a; Xie et al., 1992) which is a Ca⁺⁺-insensitive enzyme containing CaM as a tightly bound subunit. We have examined previously the structure and function analysis of nNOS through identification of the essential amino acids on its CaM-binding, application of a potent CaM antagonists, HF-2035, as a nNOS inhibitor, and finding of nNOS isoform specific phosphorylation (Watanabe et al., 1996, 1997; Win et al., 1996).

Isoform-selective inhibitors are useful tools, not only for elucidation of the physiological functions of nNOS, but also the therapeutic implications. A number of NOS inhibitors have been identified and used in pharmacological studies to investigate the biological significance of NO. Those reported to date include several N^G-derivatives of L-arginine (Hibbs et al., 1987; Moore et al., 1990), CaM antagonists (Bredt et al., 1990; Win et al., 1996), and antagonists acting at the flavin adenine dinucleotide binding site (Stuehr et al., 1991b). Recently, PIN (protein inhibitor of nNOS) was isolated as an endogenous inhibitor, and shown to specifically interact with nNOS (Jaffrey and Snyder, 1996). A range of heterocyclic compounds (e.g., 7-NI) (Mayer et al., 1994) and TRIM (Handy et al., 1996)) show relative selectivity towards nNOS in vivo. We have reported H-series heterocyclic compounds to be Ser/Thr kinase inhibitors. These include 1-(5-isoquinolinylsulfonyl)-2-methylpiperazine (H-7), N-[2-(methylamino)ethyl]-5-isoquinolinesulfonamide (H-8) (Hidaka et al., 1984), N-[2(p-bromocinamylamino)ethyl]-5-isoquinolinesulfonamide (H-89) (Chijiwa et al., 1990), and 1-(5-isoquinolinylsulfonyl)-homopiperazine (HA-1077) (Asano et al., 1989; Hidaka et al., 1984). Therefore, it is of interest whether some of these H-series heterocyclic compounds could exhibit inhibitory potency against nNOS enzyme activity.

The present report describes a newly developed heterocyclic compound, 1-(5-isoquinolinylsulfonyl)-7-methylhomopiperazine (HMN-1180), specific for nNOS but not for eNOS and iNOS, and related inhibitory mechanisms. Because HMN-1180 is a derivative of HA-1077, we also determined structure-activity relationships of such derivatives regarding inhibition of nNOS activity. The human neuroblastoma cell line SK-N-MC, was also applied as a model for assessing the effects of the drug on nNOS in situ.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. The cDNAs for rat brain nNOS, rat liver iNOS and rat endothelial NOS were gifts from Dr. Solomon H. Snyder (Johnes Hopkins, Baltimore, MD), Dr. Hiroyasu Esumi (National Cancer Center Research Institute, Tokyo, Japan) and Dr. Toshio Hayashi (Nagoya University School of Medicine, Nagoya, Japan), respectively. -Nicotinamide dinucleotide phosphate (NADPH) was purchased from Wako (Osaka, Japan). (6R)-5,6,7,8-Tetrahydro-L-biopterin (BH₄) was obtained from Research Biochemicals International (Natick, MA) and L-arginine from Sigma Chemical (St. Louis, MO). AG 50W cation exchange resin was a product of Bio-Rad (Hercules, CA). All other materials and reagents were of the highest quality available from commercial suppliers.

Cell culture. SK-N-MC human neuroblastoma cells were kindly provided by Jun Yoshida (Nagoya University School of Medicine) and cultured in minimum essential medium (Nissui, Tokyo, Japan) supplemented with 10% heat-inactivated fetal bovine serum, 2 mM glutamine, 50 units/ml penicillin G, 50 µg/ml streptomycin and 0.1 mM MEM nonessential amino acids (GIBCO, Grand island, NY).

Construction of plasmids. cDNAs encoding rat brain nNOS, rat liver iNOS and bovine aortic eNOS were introduced into pVL1393 (InVitrogen, Carlsbad, CA) or pFastBac1 vectors (GIBCO) resulting in pVLnNOS, pVLiNOS and pFastBaceNOS, respectively. An N-terminal deletional nNOS mutant (nNOS1-227) construct (pFastBacnNOS1-227) was made as follows. After digesting the cDNA for nNOS with Bpu1102I (Takara) and filling-in, a 2.8 kilobase pair nNOS fragment was obtained by digestion with NotI (Takara) and subcloning into the pFastBac1 plasmid.

Purification of expressed NOS proteins. Expression of proteins in Sf9 cells and their purification on 2',5'-ADP-Sepharose (Sigma) were performed as described previously (Watanabe, 1996). Protein was estimated by the method of Bradford, using bovine serum albmin (Fraction V) as the standard.

NOS activity assay. The activity of NOS was determined by measuring conversion of L-[³H]arginine (Amersham, Arlington Heights, WI) to L-[³H]citrulline as described previously (Bredt and Snyder, 1990). Unless otherwise indicated, the standard reaction mixture contained 50 mM Tris-HCl (pH 7.5), 1 mM CaCl₂, 100 µM NADPH, 100 µM BH₄ and 100 nM CaM.

In vitro dimerization of NOS. Purified (4 µg aliquots) NOS isoforms were incubated for 10 min at 37°C in 40 µl of 50 mM triethanolamine buffer (pH 7.0), 10 µM BH₄ and L-arginine in the presence or absence of HMN-1180. Incubation was terminated by the addition of 5 µl of Laemmli buffer containing 125 mM Tris-HCl buffer (pH 6.8), 4% (w/v) SDS, 10% (v/v) 2-mercaptoethanol, 20% glycerol and 0.02% (w/v) bromophenol blue. Samples were separated by 6% SDS-PAGE under with a constant 30 mA current at 4°C. NOS's were blotted to PVDF membrane (Millipore, Bedford, MA), then detected by ECL Western blotting detection reagents (Amersham, Arlington Heights, IL) using NOS specific antibodies (Transduction Laboratories, Lexington, KY).

HMN-1180-coupled affinity chromatography. For assaying the binding of NOS's HMN-1180 was coupled to Hitrap NHS-activated gel using the manufacturer's protocol (Pharmacia Biotech, Uppsala, Sweden); 10 µmol of HMN-1180 was coupled to 1 ml of the resin.

Nitrite analysis. Then, 4 × 10⁵ human neuroblastoma SK-N-MC cells were resuspended in 500 µl of Krebs' buffer and stirred at 1000 rpm with a platelet aggregometer. The cells were preincubated for 10 min with 100 µM HMN-1180 or 300 µM L-NAME (Sigma) and then stimulated by addition of 3 mM glutamate. Samples were collected 0 min, 30 min and 60 min after stimulation. The intact cells were centrifuged at 13,000 × g for 5 min and nitrite (NO₂) levels were measured using an automated NO detector-HPLC system ENO-10 (Eicom, Kyoto, Japan) as described recently (Yamada and Nabeshima, 1997).

Protein kinase assays. The activity of cAMP-dependent protein kinase (PKA) was measured by incorporation of [-³²P]ATP (Amersham) into CREB peptide as described previously (Hashimoto and Soderling, 1987). Ca⁺⁺/CaM-dependent protein kinase II (CaMKII) activity was measured as described previously (Tokumitsu et al., 1990).

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of HMN-1180 on NOSs enzyme activities. As illustrated in figure 1, measurement of conversion of L-[³H]arginine to L-[³H]citrulline showed inhibition of nNOS enzyme activity by HMN-1180. However no significant effects up to the concentration of 100 µM were exerted on eNOS and iNOS (eNOS, 0%; iNOS, 40% inhibition). Thus, HMN-1180 found to have a relatively selective inhibitory action against nNOS.

View larger version (24K): [in this window] [in a new window]   Fig. 1.   Effects of HMN-1180 on NOS activity. Purified recombinant nNOS (), eNOS () and iNOS () (100 nM of each protein) were assayed with 30 µl of 10 µM L-[3H] arginine, 0.1 µM CaM, 1 mM CaCl2, 10 µM BH4, 100 µM NADPH and various concentrations of HMN-1180 for 10 min at 30°C. The data were normalized to give percentages of the control values, defined as the enzyme activities of NOSs without HMN-1180. The data are the mean ± S.E.M. from two different experiments.

Kinetic analysis of inhibition of nNOS by HMN-1180. To elucidate mechanisms involved in the inhibition of nNOS enzyme activity, HMN-1180 was tested for its ability to compete with Ca⁺⁺/CaM or L-arginine binding to the enzyme. Inhibition patterns of HMN-1180 against nNOS enzyme activity were analyzed by double reciprocal plots (fig. 2). HMN-1180 competitively inhibited the enzyme activity with L-arginine with a K_i value of 5.4 µM (fig. 2, inset).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Kinetic analysis of nNOS inhibition by HMN-1180. Recombinant nNOS (100 nM) was assayed in either in the presence (10 µM, (); 30 µM, ()) or the absence () of HMN-1180 under the same conditions described in figure 1 with 5 to 50 µM L-[3H] arginine. The apparent Km/Vmax is plotted as a function of the concentration of HMN-1180. Each point represnt represent the mean ± S.E.M. from different experiments.

Inhibition of nNOS dimerization by HMN-1180. Previous characterization of NOS indicated that the native protein is a homodimer (Hevel et al., 1991; Schmidt et al., 1991b), and dimerization has been shown to be necessary for catalytic activity of the enzyme (Tzeng et al., 1995). We therefore explored the possibility that HMN-1180 affects nNOS dimerization. It is known that the nNOS homodimer is stabilized by tetrahydrobiopterin and L-arginine during low-temperature SDS-PAGE (Klatt et al., 1995). In this assay, dimerized nNOS was diminished after incubation with a 30 µM concentration of HMN-1180. The dimerized nNOS increased as the concentration of L-arginine in the assay mixture was increased (fig. 3A). However, incubation with HMN-1180 resulted in minimal loss of dimerized eNOS and iNOS (fig. 3, B and C).

View larger version (38K): [in this window] [in a new window]   Fig. 3.   Inhibition of nNOS dimerization by HMN-1180. Four µg aliquots of purified recombinant nNOS (A), eNOS (B) or iNOS (C) were analyzed by low-temperature SDS-PAGE followed by Western blotting after incubation for 10 min at 37°C with increasing concentrations of L-arginine, as indicated, in the presence (+) or absence () of 30 µM HMN-1180. A boiled sample (CNT) is included to show the expected mobility of the NOS monomer. The illustrated membrane is representative of four experiments.

HMN-1180-coupled Sepharose affinity chromatography. To demonstrate that HMN-1180 binds directly to nNOS, a sample solution containing 7 µg of the enzyme was applied to HMN-1180-coupled Sepharose column (fig. 4A), and the through fractions, washed with 1 M NaCl and boiled in SDS-sample buffer, were analyzed by SDS-PAGE. As shown in figure 4, nNOS appeared in the boiled fraction (fig. 4B; lane 3), but eNOS and iNOS did not (lanes 6 and 9).

View larger version (37K): [in this window] [in a new window]   Fig. 4.   HMN-1180-coupled Sepharose affinity chromatography. Purified recombinant nNOS, eNOS and iNOS (7 µg each) each in 50 mM Tris-HCl (pH 7.5) (200 µl) were applied to HMN-1180-coupled Sepharose (100 µl) which had been equilibrated with the same buffer. The columns were washed with 1 ml of buffer then 500 µl of the buffer containing 1 M NaCl. The retained proteins were boiled for 2 min with Laemmli's buffer (50 µl). Then, 40 µl of each fraction was applied to SDS-6% polyacrylamide gels for electrophoresis. Panel A; applied sample, panel B; lanes 1-3, nNOS; lanes 4-6, eNOS; lanes 7-9, iNOS; lanes 1, 4 and 7, void fractions; lanes 2, 5 and 8, washed fraction with 1.0 M NaCl; lanes 3, 6 and 9, gel boiled fraction. The illustrated membrane is representative of three experiments.

Interaction of HMN-1180 with the N-terminal domain of nNOS. The N-terminal domain of nNOS is unique to the neuronal isoform and contains 220-230 amino acids that are not present in iNOS and eNOS. We constructed a deletion mutant, nNOS1-227, lacking the first 227 amino acids. However, the NOS enzyme activity itself and the effect of HMN-1180 on the nNOS1-227 mutant were essentially indistinguishable from these with the wild-type enzyme (fig. 5), indicating that the N-terminal domain of nNOS is not involved in the drug-enzyme interaction.

View larger version (25K): [in this window] [in a new window]   Fig. 5.   The role of the extended N-terminus of nNOS in the selective inhibition by HMN-1180 the enzyme activity. Purified recombinant wild-type nNOS or nNOS1-227 (100 nM of each protein) were assayed under the same conditions as described in figure 2. The data were normalized to give percentages of the control values, defined as the enzyme activities of NOSs without HMN-1180. Purified wild-type nNOS (2 µg) and nNOS1-227 (1 µg) were separated on 6% SDS-PAGE and analyzed by Coomassie Blue staining (inset). The data are the mean ± S.E.M. of two experiments.

Structure-activity relationships of HMN-1180 and its derivatives as inhibitors of NOS. To assess the activity-structure relationships of HMN-1180 regarding inhibition of NOSs activity, PKA and CaMKII, we examined the effects of the position of the methyl group in the homopiperazine molecule of various HMN-1180 derivatives (table 1). Among these agents, HMN-1180 (7-methyl) was the most potent inhibitor of nNOS. The derivative with no methyl group in the homopiperazine molecule (HA-1077) was the weakest. So HA-1077 was the most potent inhibitor of PKA, and HMN-1182 (6-methyl) was to CaMKII in these derivatives, that indicated the methyl group in the homopiperazine molecule may be responsible for the affinity of this compound for nNOS.

Effects of HMN-1180 on nNOS-dependent NO production. The effects of HMN-1180 on nitrite production in the stirred human neuroblastoma cell line SK-N-MC were also analyzed. In the supernatants of stirred unstimulated SK-N-MC cells, we found a basal nitrite (NO₂) level (~6.6 nmol/ml), which was presumably due to iNOS because expression was detected by RT-PCR using human iNOS-specific primers (data not shown). When the cells were incubated with glutamate (3 mM) for 0.5 and 1 hr, a time-dependent increase of nitrite synthesis was observed (fig. 6). This phenomenon was inhibited by preincubation with L-NAME (500 µM), thus demonstrating the glutamate-increased nitrite generation to be associated with the L-arginine-NO pathway (fig. 6). Expression of nNOS was also analyzed in SK-N-MC cells by Western blotting using antihuman nNOS antibodies. The expected 162-kDa immunoreactive band was observed (fig. 6, inset). When the cells were preincubated with HMN-1180 (100 µM) for 10 min followed by incubation with glutamate (3 mM), the glutamate-induced nitrite production was significantly inhibited in two separate experiments (fig. 6). These results suggested that HMN-1180 can cross the cell membrane and block nNOS enzyme activity in SK-N-MC cells.

View larger version (48K): [in this window] [in a new window]   Fig. 6.   Effects of HMN-1180 on nNOS-dependent NO production. Nitrite levels were measured using the Griess reaction in supernatants of stirred SK-N-MC cells treated with glutamate (3 mM) for 0.5 and 1 hr. When the cells were preincubated with L-NAME (500 µM) or HMN-1180 (100 µM) for 10 min followed by incubation with glutamate (3 mM), nitrite levels were normalized by the control of NaNO2 (10 µM). nNOS protein was detected by Western blot using antihuman nNOS antibody (inset). Molecular weights are given on the left. The data shown is mean ± S.E.M representative of two experiments.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

This study characterized HMN-1180 and a related group of isoquinolinesulfonamides as specific inhibitors of nNOS. The observation that varying the concentration of L-arginine incubated with recombinant nNOS affected the NOS inhibitory activity of HMN-1180 suggests that this compound competes with L-arginine for a substrate binding site on the synthase. We also demonstrated that nNOS binds to a HMN-1180-coupled Sepharose 4B column, while eNOS and iNOS do not (fig. 4). Kinetic (Lineweaver-Burk) analysis revealed a Km for L-arginine of 4.4 µM (similar to the published values of 2.0 ± 0.4 µM, (Richards and Marletta, 1994)) and a K_i value of 5.4 µM for HMN-1180 (fig. 2).

To date, well over are hundred NOS inhibitors have been reported in the literature. Many amino acid derivatives inhibit NOS enzyme activity. In general, none of these compounds reported to date shows significant selectivity for nNOS either in vitro or in vivo (Hibbs, 1987; Moore, 1990). The interest in HMN-1180 is due to its being structurally distinct from known amino acid derivatives. Heterocyclic compounds that have been shown to inhibit nNOS, include 7-NI and TRIM carrying net electronegative charges and thought to interact with the heme prosthetic group of NOS. Isoquinolinesulfonamide protein kinase inhibitors of the H series are among the most widely used inhibitors of Ser/Thr kinases (Chijiwa et al., 1990; Hidaka et al., 1984; Inagaki et al., 1984). Because isoquinolinesulfonamide compounds are heterocyclic, we focused on their nNOS inhibitory abilities. Preliminary attempts were also made to examine the structure-activity relationships of HMN-1180 and a related group of isoquinolinesulfonamides. The inhibitory potency of these derivatives seems to be dependent on the position of the methyl group in the homopiperazine molecule (table 1). HMN-1180 possesses a lone pair of electrons on the nitrogen same nitrogen at the 2-position of the isoquinoline and the 4-position of the homopiperazine. How the nNOS inhibitory activity of the derivatives targets the nNOS isoform (but not eNOS or iNOS), is related to an electrochemical charge effect remain and whether this to be determined.

Basically, isoquinolinesulfonamide derivatives act in competition for ATP but not the substrate. In fact, kinetic analysis indicated that the inhibition of PKA, and CaMKII by HMN-1180 was competitive with respect to ATP (data not shown). HA-1077 exhibited no significant effects up to a concentration of 1 mM on nNOS activity (table 1). Crystal structures of PKA bound to H7, H8, or H-89 have been reported and the mode of inhibitory action and factors governing selectivity have been identified (Engh et al., 1996). We demonstrated direct binding of HMN-1180 to nNOS, however, analyses of structural-enzyme inhibitory relationships need to be performed for elucidating the underlying mechanisms. We previously reported that the nucleotide binding consensus sequence G-Q-G-A-G-S (residues 166-171 in nNOS) might be important for regulation of autophosphorylation of nNOS because it is not contained in iNOS and eNOS (Watanabe et al., 1996). Because the region of the glycine flap in the kinases could contribute to the selectivity of isoquinolinesulfonamide inhibitors, the role of the extended N-terminal region of nNOS in the selective inhibition by HMN-1180 was analyzed. However, the NOS enzyme activity itself and the effects of HMN-1180 on the nNOS1-227 mutant were essentially indistinguishable from those for the wild-type enzyme (fig. 5).

Although it has been confirmed that PKA, PKC and CaMKII phosphorylate nNOS, the physiological significance remains uncertain as there is either no detectable effect of phosphorylation on enzyme activity (PKA) or the effect on activity is controversial (PKC, CaMKII) (Bredt et al., 1992; Brune and Lapetina, 1991; Nakane et al., 1991). Although HMN-1180 could inhibit the nNOS enzyme activity in vivo, an involvement of PKC and CaMKII-induced phosphorylation could not be excluded. However, the present study suggests that the nNOS specific inhibitor, HMN-1180, provides a valuable tool for elucidating NO-mediated cellular and signal transduction.