Introduction

Summary Introduction Procedures Results Discussion References

The role of the biogenic amine histamine is not clearly understood, but its release from various tissues of the human body is frequently associated with the inflammatory state. Histamine is also involved in gastric secretory activity and, thus, plays a role in gastric ulcer formation. The effects of histamine are brought about through the activation of the histamine H₁-, H₂-, and H₃-receptors. These receptors are distinguishable on the basis of their differing sensitivities to agonists and antagonists (1-4). Some tissues have predominantly one type of receptor, whereas others contain a mixture of the receptors. Gastric acid secretion is mediated almost exclusively through H₂-receptor activation (2). In the last 20 years, H₂-receptor antagonists, such as cimetidine (Tagamet), famotidine (Pepcid), and ranitidine (Zantac), have been widely used clinically in the treatment of peptic ulcers.

The metabolism of biogenic amines usually proceeds through aldehyde intermediates. Aldehyde dehydrogenase (EC 1.2.1.3) catalyzes the NAD⁺-linked dehydrogenation of aldehydes to acids and has a broad substrate specificity. Naturally occurring substrates include aldehyde metabolites of histamine, putrescine, and dopamine (5). Three isozymes of aldehyde dehydrogenase, the cytoplasmic E1 and E3 isozymes and the mitochondrial E2 isozyme, have been purified from human liver (6, 7), and their cDNAs have been cloned (8, 9). The genes coding for these isozymes have been chromosome-localized: the ALDH1 of E1 on chromosome 9; the ALDH2 gene of E2 on chromosome 12, and the ALDH9 gene of E3 on chromosome 1 (10, 11). Although all three isozymes exhibit broad substrate specificity, the activity with aminoaldehydes at low concentrations is confined to the E3 isozyme. The E3 isozyme catalyzes the conversion of -aminobutyraldehyde to -aminobutyric acid (K_m = 5-14 µM) (7, 12) and aldehyde metabolites of spermidine and spermine to corresponding carboxylic acids (12). More recently, it was also identified as a betaine aldehyde dehydrogenase (13). It appears that E3 isozyme may play a role in intermediary metabolism of putrescine, polyamines, and choline.

Study of this enzyme in more complex systems is hindered by the fact that up to the present time no inhibitors have been identified. Here, we report the potent inhibition of the E3 isozyme by the histamine H₂-receptor antagonists. The structures of the H₂-receptor antagonists used during this investigation are shown in Fig. 1. All are based on the chemical structure of histamine and consist of a basic substituted 5-membered ring (imidazole, thiazole, or furan) with a 4-atom side chain at position 4 of the ring. Each side chain bears a polar group, such as a thiourea, guanidine, or cyanoguanidine.

View larger version (13K): [in this window] [in a new window]   Fig. 1.   Chemical structures and nomenclature of the histamine H2-receptor antagonists used during this investigation. Each structure consists of a substituted 5-membered ring, with a flexible 4-atom side chain that bears a polar group at position 4 of the ring.

	Experimental Procedures

Summary Introduction Procedures Results Discussion References

Reagents. Aminodecyl agarose, aminoguanidine, cimetidine, 4-(dimethylamino)cinnamaldehyde, famotidine, furan, glycolaldehyde, histamine, histidine, nicotinamide hypoxanthine dinucleotide (NHD⁺), ranitidine hydrochloride, and thiazole were obtained from Sigma Chemical (St. Louis, MO). Guanidine hydrochloride was from United States Biochemicals (Cleveland, OH). Burimamide (SKF 91923), cimetidine guanidine (SKF 92408), and metiamide (SKF 92058) were a gift from SmithKline Beecham Pharmaceuticals (Philadelphia, PA). Cyanoguanidine and imidazole were obtained from Aldrich (Milwaukee, WI); dimaprit (S-[3-(N,N-dimethylamino)propyl]-isothiourea) dihydrochloride and cimetidine were from Research Biochemicals International (Natick, MA); and NAD⁺ (grade 1) was from Boehringer Mannheim (Indianapolis, IN). Tiotidine maleate was a gift from Dr. T. O. Yellin, SmithKline Beecham Pharmaceuticals, King of Prussia, PA. Tiotidine was also obtained from Tocris Cookson (St. Louis, MO). Stock solutions of the compounds were made up in water, 10 mM HCl, or 0.05 M sodium phosphate buffer, pH 7.4. Tiotidine was also dissolved in dimethyl sulfoxide and famotidine in N,N-dimethylformamide, both solvents having no effect at 1% v/v concentrations on E3 isozyme activity.

Enzymes. E1, E2, and E3 isozymes were purified from human liver as previously described (6, 7) and stored in 30% glycerol, at 4°, under nitrogen. Before use, the enzymes were extensively dialyzed against 30 mM sodium phosphate buffer, pH 7.0, 1 mM EDTA, to remove glycerol and -mercaptoethanol. The protein concentration was determined as described by Goa (14), using bovine serum albumin as a standard. Protein was also assayed spectrophotometrically at 280 nm, using an extinction coefficient of 1.0 (1 mg/ml)¹ cm¹ (7). E3 enzyme stability was checked using an assay mixture of 0.1 M sodium phosphate buffer, pH 7.4, containing 1 mM EDTA, 500 µM NAD⁺ and 0.1 M -aminobutyraldehyde as substrate (7). When partial loss of activity occurred, experimental results were adjusted to the maximal activity of 1.6 µmol/min/mg (7).

Kinetic studies. All assays were carried out in 0.05 M sodium phosphate buffer, pH 7.4, containing 1 mM EDTA, glycolaldehyde as substrate and either NAD⁺ or NHD⁺ as coenzyme, at 25°. NADH (or NHDH) formation was monitored at 340 nm using a Gilford spectrophotometer. An extinction coefficient of 6.22 mM¹ cm¹ for NADH (and NHDH) was used for the calculation of reaction rates. It was also used in the spectrophotometric determination of the concentration of glycolaldehyde stock solutions, as previously described (12). Inhibition experiments were performed in two ways: (i) for competition versus substrate, glycolaldehyde concentrations were varied at different fixed test-inhibitor concentrations, at a single NAD⁺ concentration (500 µM); (ii) for competition versus coenzyme, NHD⁺ concentrations were varied at different fixed test-inhibitor concentrations, at a single nonsaturating glycolaldehyde concentration. Reactions were initiated by the addition of enzyme. Reaction rates were determined by tangents to initial velocities. Each point represents duplicate determinations (which did not differ from each other by more than 5%) of the reaction rates. Kinetic data were obtained using the SlideWrite Plus Program, according to the method of Lineweaver and Burk (15), employing the linear least squares regression fit of reciprocals of the reaction rates versus reciprocals of substrate concentration. Regression coefficients were between 0.993 and 0.999 for all data.

	Results

Summary Introduction Procedures Results Discussion References

Effect of H₂-receptor antagonists on the E1, E2, and E3 isozymes of human aldehyde dehydrogenase. The effect of compounds I-VI on the catalytic activity of the three isozymes of liver aldehyde dehydrogenase (E1, E2, E3) is shown in Table 1. All the compounds (1 mM) had the greatest effect on the E3 isozyme; cimetidine and tiotidine completely abolished E3 isozyme catalytic activity. The compounds had less effect on the other two aldehyde dehydrogenase isozymes. However, tiotidine abolished more than 50% of the catalytic activity of both the E1 and E2 isozymes and burimamide had a similar effect on the E2 isozyme. It is interesting to note that metiamide, cimetidine guanidine, and especially famotidine produced a slight but reproducible activation of the E1 isozyme. The effect of ranitidine (compound VII) could not be determined because of its high absorbance in the range of NADH absorbance.

Inhibition studies of the E3 isozyme with glycolaldehyde as the varied substrate. NAD⁺ was used as the fixed substrate at a saturating concentration (500 µM) with glycolaldehyde (K_m = 221 µM, see footnote to Table 2). Thus, the K_i values shown in Table 2 represent dissociation constants of H₂-receptor antagonists from E3·NAD·inhibitor ternary complex. This concentration of NAD also approximates that in mammalian liver. Compounds I-VI (Fig. 1) inhibited the E3 isozyme in a competitive manner versus aldehyde substrate (as shown for cimetidine, Fig. 2A), allowing K_i values to be obtained from the slope replots (Fig. 2A, inset) which were all linear. K_i values shown in Table 2 are mean values from triplicate determinations. Although K_i values for cimetidine and tiotidine were in the low micromolar range, those for burimamide, metiamide, and cimetidine guanidine were larger by about 2 orders of magnitude, and that for famotidine was larger, by almost 4 orders of magnitude.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Lineweaver-Burk plots of cimetidine and cimetidine guanidine inhibition of E3 isozyme activity. The reaction velocity is expressed as micromoles of NADH or NHDH/min/mg of enzyme protein. A, Cimetidine inhibition of E3 isozyme activity at varied glycolaldehyde concentrations and fixed cimetidine concentrations of 0 µM (), 1.25 µM (), 2.5 µM (+), 5 µM (), and 10 µM (). Inset, slope replot of data represented in A. B, Cimetidine guanidine inhibition of E3 isozyme activity at varied NHD+ concentrations and fixed cimetidine guanidine concentrations of 0 µM (), 75 µM (), 150 µM (+), 300 µM (), and 600 µM (). Inset, slope and intercept replots of data represented in B; , slopes; , intercepts.

Inhibition studies of the E3 isozyme with NHD⁺ as the varied substrate

Studies with NHD⁺ were conducted to determine points of inhibitor binding to the E3 isozyme. The K_m for NAD⁺ for the E3 isozyme is low (4 µM), which made variation of coenzyme concentration difficult even at the highest sensitivity of our instruments. Therefore, the NAD⁺ analog NHD⁺ (K_m = 203 µM, see footnote to Table 3) was used in the kinetic studies where coenzyme was varied. Compounds I-V (Fig. 1) were shown to inhibit the E3 isozyme in a noncompetitive manner, producing both slope and intercept effects (see Fig. 2B, which shows cimetidine guanidine), versus NHD⁺. K_i slope and K_i intercept values (Table 3) were obtained from secondary plots (as shown in Fig. 2B, inset). K_i intercept values are modified by the fixed concentration of glycolaldehyde used in the experiment. True K_i values were, therefore, calculated using the following equation: K_i = K_i intercept/(1 + [concentration of glycolaldehyde]/K_m glycolaldehyde).

Structurally related compounds as inhibitors of the E3 isozyme. Imidazole, thiazole, and cyanoguanidine (Table 4) at 1 mM concentration had no effect on E3 isozyme activity. Dimaprit (S-[3-(N,N-dimethylamino)propyl]-isothiourea), an H₂-receptor agonist, had only slight inhibitory effect. Inhibition of E3 isozyme by imidazole and thiazole was not observed at concentrations below 10 mM. Cyanoguanidine did not exhibit any inhibitory properties up to a concentration of 100 mM.

Comparison of inhibition constants for the E3 isozyme with those for the histamine H₂-receptor. Comparison of the K_i values for the E3 isozyme with the in vitro dissociation constants, K_B, of compounds I-V for the histamine H₂-receptor (4, 16-18) showed that the K_i value of cimetidine for the E3 isozyme was indistinguishable from that for the H₂-receptor (Table 2). The K_i values of cimetidine guanidine were also similar, whereas those of tiotidine, burimamide, and metiamide for the E3 isozyme were 2 orders of magnitude higher than for the H₂-receptor. Interestingly, famotidine, which was a poor inhibitor of the E3 isozyme, had an extremely low (17 nM) K_i value for the H₂-receptor.

	Discussion

Summary Introduction Procedures Results Discussion References

The histamine H₂ receptor antagonists (compounds I-VI, Fig. 1) inhibited or activated all three isozymes of human aldehyde dehydrogenase to varying degrees (Table 1). With all the compounds, inhibition was strongest of the E3 isozyme. Cimetidine and tiotidine proved to be potent inhibitors with K_i values of ~ 1 µM (Table 2). Although still in the micromolar range, the K_i values of burimamide, metiamide, and cimetidine guanidine were 1-2 orders of magnitude larger than those of cimetidine and tiotidine. Thus, H₂-receptor antagonists are the first ever reported potent and selective inhibitors of the E3 isozyme. Only famotidine was a poor inhibitor with a K_i value in the millimolar range.

The K_i values (Table 2) were used to identify the inhibiting moieties of the histamine H₂-receptor antagonists. The ring moieties of the structures were considered first. Metiamide and burimamide, have similar structures. The imidazole ring of metiamide, however, has a methyl substitution at position 5. Metiamide also contains a sulfur in the side chain. It is, therefore, not clear whether the slightly lower K_i value for metiamide for the E3 isozyme is due to the ring substitution or to the presence of sulfur in the side chain. Cimetidine and tiotidine, too, have similar structures, the only difference being that the methyl-imidazole ring of cimetidine is replaced by a guanidinothiazole in tiotidine. However, their K_i values for the E3 isozyme were identical (Table 2), which suggests that replacement of the methylimidazole by a guanidinothiazole ring did not affect inhibition. Comparison of the side chains of the compounds revealed that the removal of the cyano group of the cyanoguanidino side chain of cimetidine resulted in a loss of inhibitory properties. This is seen in the 68-fold increase in the K_i value for cimetidine guanidine. An even higher loss of inhibitory properties (110-fold increase in the K_i value) occurred with the replacement of the cyanoguanidino group (cimetidine) by a thiourea group (metiamide). Replacement of the methyl cyanoguanidino group of tiotidine by a sulfonylamido group (famotidine) increased the K_i value by 3 orders of magnitude. From this, it appears that the side-chain polar groups strongly influence inhibition, with the methyl cyanoguanidino group having the strongest influence. When cyanoguanidine was tested (Table 4) for inhibition of the E3 isozyme; however, it proved to be a poor inhibitor. It thus appears that spatial configuration and a 5-membered heterocyclic ring is important; cyanoguanidine is a potent inhibitor only as a part of the side chain of cimetidine or tiotidine. Although the side chain of cimetidine guanidine has structural resemblance to a known substrate of E3 isozyme, -guanidinobutyraldehyde (8), substrates resembling the methyl cyanoguanidine-containing side chain of cimetidine and tiotidine have not yet been identified.

The K_i values of the H₂-receptor antagonists for the E3 isozyme were also compared with their K_i values for the histamine H₂-receptor (Table 2). Although the K_i values of cimetidine were identical, the K_i values of the other H₂-receptor antagonists for the E3 isozyme were larger than those for the H₂-receptor. The most notable difference was with famotidine, whose K_i value for the E3 isozyme was 5 orders of magnitude larger than that for the H₂-receptor. The recognition of these compounds by the E3 isozyme must, therefore, occur in a manner different from that by the histamine H₂ receptor. From data in Table 2 it appears that the side-chain polar groups influence the inhibition of the E3 isozyme by the H₂-receptor antagonists more strongly than the ring moieties. In contrast, the ring moieties of these compounds are more important for H₂-receptor recognition, and act cooperatively with the side-chain polar groups to confer receptor binding (4, 16, 19). This is demonstrated by the fact that replacement of the methylimidazole ring of cimetidine by a thiazole ring (tiotidine and famotidine) or a furan ring (ranitidine, H₂-receptor, K_i = 0.063 µM) resulted in more potent antagonists of the H₂-receptor (Table 2) (4, 19).

All six compounds inhibited the E3 isozyme in a noncompetitive manner versus varied NHD⁺, producing both slope and intercept effects (Fig. 2B). Only an intercept effect versus the varied coenzyme would be expected if the H₂-receptor antagonists bound solely to the E3·coenzyme binary complex (20). The slope effect with varied NHD⁺ shows that the H₂ receptor antagonists can also bind before the coenzyme in the reaction sequence, binding to the free enzyme. This binding is prevented when coenzyme is saturating (Table 2). Cimetidine was found to specifically elute the E3 isozyme from an affinity column (to be published elsewhere as a part of improved purification procedure), confirming that cimetidine can bind to the free enzyme. Because of the use of two different coenzymes, the ternary complexes shown in Tables 2 and 3 are different, the one represented by the K_i values in Table 2 is from the E3·NAD⁺·I complex and that represented by K_i intercept values in Table 3 is from the E3·NHD⁺·I complex. Despite these differences, the K_i slope values in Table 2 and values calculated from the K_i intercept in Table 3 are similar and possibly identical within the experimental error of the procedure employed. The dissociation constants for the E·I binary complex (represented by the K_i slope values, Table 3), except for burimamide, were also similar to those of E3·coenzyme·I ternary complexes (K_i values in Table 2 and calculated K_i values in Table 3), differing by only a factor of 2 at the most. For the E·I binary complex, the dissociation constant (Table 3, K_i slope) for burimamide was 10-fold lower than that for the E3·NAD⁺·I ternary complex (Tables 2 and 3). This suggests that, with the exception of burimamide, the inhibitors and the coenzyme bind independently to the enzyme.

H₂-receptor antagonists are the first ever described potent and selective inhibitors of the E3 isozyme. Their immediate use is envisaged in the purification of the E3 isozyme. They could also be valuable in further studies of the E3 isozyme both in vitro and in intact animals. The low K_i values of the E3 isozyme with cimetidine and tiotidine (Table 2) suggest that it may be inhibited in vivo during treatment for stomach ulcer, where concentrations of cimetidine are known to rise to ~ 200 µM (21). Such inhibition could result in altered metabolism of polyamines, -aminobutyric acid and betaine which are important in cell growth, differentiation, neurotransmission, and osmoregulation.