Pyridine nucleotide interaction with rat liver dihydropteridine reductase.

The interactions of a homogeneous preparation of rat liver dihydropteridine reductase with NADH, NADPH, NAD+, NADP+, and the 1-N6-ethenoadenine derivative of NAD+ have been investigated by fluorescence titration, circular dichroism, equilibrium dialysis, Sephadex G-25 chromatography, and polyacrylamide gel electrophoresis. The procedures indicate that the dimeric enzyme has a definite preference for NADH, but binds only 1 mol of this nucleotide per mol of enzyme. The binary complex of enzyme with NADH is only partially stable to exhaustive dialysis and gel electrophoresis, where it shows greater mobility (0.26) than the free enzyme (0.21); however, the complex can be isolated by Sephadex G-25 chromatography, and characterized with respect to its absorbance spectrum. No ternary complexes are observed when samples of reductase, preincubated with excess NADH, and either the reaction product, 2-amino-4-hydroxy-6,7-dimethyl-5,6,7,8-tetrahydropteridine, or the inhibitor, methotrexate, are subjected to polyacrylamide gel electrophoresis.

The interactions of a homogeneous preparation of rat liver dihydropteridine reductase with NADH, NADPH, NAD+, NADP+, and the l-A@-ethenoadenine derivative of NAD+ have been investigated by fluorescence titration, circular dichroism, equilibrium dialysis, Sephadex G-25 chromatography, and polyacrylamide gel electrophoresis. The procedures indicate that the dimeric enzyme has a definite preference for NADH, but binds only 1 mol of this nucleotide per mol of enzyme. The binary complex of enzyme with NADH is only partially stable to exhaustive dialysis and gel electrophoresis, where it shows greater mobility (0.26) than the free enzyme (0.21); however, the complex can be isolated by Sephadex G-25 chromatography, and characterized with respect to its absorbance spectrum. No ternary complexes are observed when samples of reductase, preincubated with excess NADH, and either the reaction product, 2-amino-4-hydroxy-6,7-dimethyl-5, 6,7,8tetrahydropteridine, or the inhibitor, methotrexate, are subjected to polyacrylamide gel electrophoresis.
Reduced pteridines are substrates in several mixed function oxygenase reactions (l-3). Of particular interest are the reduction-hydroxylation sequences necessary for the conversion of phenylalanine to tyrosine, and the subsequent conversion of the latter compound to dihydroxyphenylalanine.
The reductase, first detected in sheep liver extracts (6), has since been purified from several mammalian tissues (7)(8)(9), and from a species of Pseudomonas (10). The enzymes from all of these sources have similar molecular weights (42,000 to 52,-000), are dimeric, and generally demonstrate a marked preference for NADH as a cofactor, although Nakanishi et al. (11) have indicated that an NADPH-requiring pteridine reductase is present in bovine liver. More recently, however, Hasegawa (9) has also purified an NADH-dependent reductase from the same source, which contains 2 tightly bound molecules of nucleotide. [el = $.$ (2) 8, observed ellipticity in degrees; M,, molecular weight (51,000 for rat liver diiydropteridine reductase); I, path iength in cm; and C, enzyme concentration in g/ml. The CD spectra of the enzyme were corrected for the contribution of the nucleotide at each concentration used in the titration; it was assumed that the spectrum of the nucleotide did not alter upon interaction with the enzyme.
Equilibrium Dialysis-Equilibrium dialysis measurements of ligand binding were obtained with the following system. Dialysis tubing (length, 10 cm; diameter, 1 cm) was sealed at one end and attached, at the other, to a glass tube inserted through a rubber stopper. The unit was sealed into a 2-liter Erlenmeyer flask containing stirred 0.05 M Tris-hydrochloride buffer (pH 7.0, determined at the start and finish of each experiment). Enzyme and r3H]NADH, or [3H]NAD', in identical buffer solutions (3.5 ml), were placed in turn inside the sac, and each system was allowed to equilibrate at 4'C. Aliquots (20 ~1) were extracted from the sac at intervals in both experiments and their radioactive content was measured in Aquasol (10 ml) using a Beckman LS-233 liquid scintillation counter. The concentration of ligand remaining inside the sac was then calculated and plotted as a function of time. The results were analyzed by the following relationship (18): [NADH] is the concentration of nucleotide present within the dialysis sac at any time (t), kl and kz are the rate constants of two parallel first order rate processes, and by delinition, A + B is the initial concentration of r3H]NADH. A semilogarithmic plot of [NADH] uersus t afforded kz as the slope of the slow phase of nucleotide release (Fig. 7), and B was obtained by extrapolating the slope to t = 0. A plot of ln([NADH] -Bemkd) versus t gave kr and A as slope and ordinate intercept, respectively. kr, the apparent initial rate constant in the presence of the reductase, was compared to the rate of diffusion (A) of the nucleotide from the dialysis sac when enzyme was excluded. In addition, from the values of A, B, A,, and AZ, a Hewlett-Packard calculator model MP-9810A was used to reconstruct a curve relating [NADH] and t, which was then checked for correlation with the experimental data.
Polyacrylamide Gel Electrophoresis-Polyacrylamide gel electrophoresis was performed without a tracking dye on 7.5%, w/v, gels at pH 8.3 according to the procedure of Omstein (19) and Davis (20). Mixtures of enzyme and ligand were incubated in 0.05 M Tris-hydrochloride, pH 7.0, for 30 mm at room temperature prior to application. Gels were stained overnight in 0.2% Aniline blue black and destained electrophoretically in 7% acetic acid.
Protein concentrations were determined by the method of Lowry (21). NADH and NAD* concentrations were calculated using the extinction coefficients reported by Gurr et al.

AND DISCUSSION
Fluorescence-Dihydropteridine reductase exhibits a fluorescence excitation maximum at 285 nm and an emission peak at 335 run. These fluorescence characteristics and the absorbance spectrum of the homogeneous enzyme are illustrated in Fig. 1. Titration of the enzyme with NADH, NADPH, NAD+, and NADP' quenched the enzyme fluorescence, but did not alter the wavelengths of the excitation or emission maxima. The fluorescence of the enzyme at 335 nm was quenched by approximately 70% upon addition of increasing quantities of the reduced pyridine nucleotides, although much lower concentrations of NADH were required than NADPH (Fig. 2).
Addition of similar concentrations of the oxidized compounds quenched the fluorescence of the enzyme by approximately 25% (cft Fig. 2).
The quenching of the enzyme fluorescence by increasing concentrations of NADH was examined in more detail by the experiment illustrated in Fig. 3.  Table I. Michaelis  constants for  NADH and NADPH, which have been reported previously (16), are included for comparison.
In a complementary series of experiments, the enzymenucleotide interaction was monitored by measuring the fluorescence enhancement at 454 nm exhibited by the reduced pyridine nucleotides upon binding to the enzyme. Maximum fluorescence enhancement (&fold) occurred with a 1:l molar ratio of enzyme and NADH as is shown in Fig. 4. Additional elevation of the nucleotide concentration did not lead to further enhancement in fluorescence. Identical results were observed when the experiment was carried out in either 0.05 M Tris-hydrochloride or in 0.1 M potassium phosphate buffer solutions at pH 7.0. These observations suggest a single binding site for the NADH molecule, although the existence of a second weaker binding site, or a site which causes little alteration in the fluorescence emissions of enzyme or nucleotide, cannot be excluded.
The fluorescence changes observed upon titration of the reductase with NADPH in a similar experiment gave an equivalence point of 7 mol of nucleotide/mol of enzyme and afforded only a 20% increase in fluorescence. These measurements suggest the occurrence of nonspecific binding and support the evidence from kinetic data (cf Table I) that NADPH is not the natural cofactor for this enzyme.
In order to investigate the binding of oxidized nucleotides to the enzyme, a fluorescent derivative of NAD', E-NAD+, was synthesized. The quenching of enzyme fluorescence observed on titration with this material was less (-5%) than was observed with the unmodified oxidized nucleotide and the Kd was 20% higher (cf Table I). In addition, enhancement of the +NAD+ fluorescence following interaction with the reductase   was less than that observed between the enzyme and its natural cofactor, NADH. However, maximum fluorescence enhancement was again observed with a 1:l molar equivalence of enzyme and E-NAD+, as is shown in Fig. 5. Although the binding of the fluorescent nucleotide analog, E-NAD+, is somewhat less than that of NADH, it probably interacts with the enzyme at the same site as NADH, as the oxidized nucleotide is leaving the group after enzymatic reduction of the pteridine substrate. Occurrence of a maximal fluorescence increment at the one molar equivalence point provides additional evidence for the existence of a stable 1:l complex between the nucleotide and the dimeric reductase.
Polyacrylamide Gel Electrophoresis-Stable enzyme. substrate and enzyme. cofactor complexes of both dihydrofolate reductase (23) and thymidylate synthetase (24) have been identified previously by measuring their altered mobilities following polyacrylamide gel electrophoresis. Similarly, when the pteridine reductase in 0.05 M Tris-hydrochloride buffer, pH 7.0, was incubated for 30 min with an excess of NADH at 25°C prior to gel application, the formation of an enzyme*NADH complex was visible after electrophoresis as a second band of increased mobility, RF 0.26, (Gel B, Fig. 6) compared to the native enzyme, RF 0.21 (Gel A, Fig. 6). However, variations in the incubation conditions (time, 1 to 120 min; temperature, 4-37°C; up to a 50-fold excess concentration of the pyridine nucleotide) did not produce quantitative conversion to the nucleotide. enzyme complex. In addition, no bands suggesting the formation of the ternary complexes were observed when a 5-fold excess of either the reaction product, 2-amino-4-hydroxy-6,7-dimethyl-5,6,7,8-tetrahydropteridine, or the inhibitor, methotrexate, were added to the incubation mixture of enzyme and NADH. The limited stability of the enzyme. NADH complex was also demonstrated by preincubating the rat liver dihydropteridine reductase (40 pg) with a 1:l molar ratio of [3H]NADH at room temperature for 60 min prior to electrophoresis. The gel was then frozen and sliced into 2-mm sections, each of which was homogenized with 0.5 ml of 0.1 M Tris-hydrochloride buffer, pH 7.2. After centrifugation, the supernatant fractions were assayed for enzyme activity and radioactivity.
The main peak of enzymatic activity showed a mobility of 0.21 corresponding to free enzyme, and the majority (95%) of the radioactivity moved with the front. However, a small peak of mobility 0.26, which contained both C3H]NADH and enzyme activity, suggested the formation of an enzyme. cofactor complex.
Equilibrium for Kz, and 4.62 pM for B. The similarity of this latter value to that of the initial enzyme concentration suggested the existence of a stable 1:l complex between enzyme and nucleotide.
In order to search for evidence of a second nucleotide binding site, the early part of the efflux rate curve was examined in more detail. both with, and in the absence of enzyme, suggested that the early loss of nucleotide was independent of enzyme presence. Only when a 1:l equivalence of enzyme and nucleotide was approached did the rate alter significantly from that of a simple diffusion efflux. Such results suggest the lack of a second nucleotide binding site.
Control experiments established that under the conditions of equilibrium dialysis, free NADH was oxidized at a rate of approximately 7% in each 24-h period, as measured by the loss of absorption at 340 run. However, it was determined that the presumed product, NAD', did not bind to the enzyme, therefore measured radioactivity retained within the dialysis sac was an accurate representation of bound NADH. For example, in experiments where the initial concentrations of NADH were varied, such that total equilibration periods ranged from 10 to 160 h, no alteration in the binding portion of the curve profile (Fig. 7) was observed. Retention of enzymatic activity by the reductase. NADH complex was confiied by the procedure of Hasegawa (9), modified by the replacement of ferricytochrome c with 2,6-dichloroindophenol.
Circular Dichroism-CD measurements were made in 0.05 M T&-hydrochloride buffer, pH 7.0, between 245 and 450 run.
In this region, the enzyme shows an overall negative ellipticity with an aromatic side chain Cotton effect at 260 to 310 rim. The major band is centered at 285 nm ([a = 54,700 deg. cm'. dmol-I). Titration of the enzyme with NADH leads to the spectral changes shown in Fig. 8: a band of positive ellipticity is generated at 255 nm, the negative band at 285 nm is sharpened, the shoulder at 265 nm disappears, and a positive extrinsic Cotton effect is generated between 340 and 350 nm (an absorbance maximum of NADH). All of the transitions reach a maximum at a 1:l molar ratio of enzyme to nucleotide, as is shown for the 255 nm band in the inset to Fig. 8.
When a similar experiment was performed with the oxidized nucleotide and the spectra were corrected for the CD contribution of NAD+, no changes could be observed.  on that used by Hasegawa (9) to isolate a 2:l complex between NADH and dihydropteridine reductase from bovine liver. The enzyme, which was preincubated with a 5-fold excess of NADH for 15 min at room temperature, was applied to a column (1.3 x 35 cm) of Sephadex G-25 previously equilibrated with 0.05 M Tris-hydrochloride buffer, pH 7.0. Elution was performed with the same buffer at a flow rate of 20 ml/h. Each fraction was monitored for protein concentration, determined by the Lowry method (21), enzymatic activity, and nucleotide absorbance at 340 nm, pH 8.6, and the profile shown in Fig. 9 was obtained. From these results, it could be calculated that a 1:l molar ratio of enzyme and NADH was present in the protein peak. Unbound NADH was eluted from the column in a second peak which contained no protein.
The spectral properties of the enzyme.NADH complex at pH 7.0 are illustrated in the inset to Fig. 9. The complex showed the same specific activity (62.5 pmol of NADH oxidized/min/mg) as did the enzyme alone prior to the addition of NADH. It was found to be stable at 4°C for greater than 7 days and no loss of activity occurred on storage at -15°C for 6 months.
The ability of the enzyme to form a stable complex with NAD+ was examined in a similar experiment. A 1:5 mixture of enzyme and C3H]NAD+ was incubated and chromatographed under conditions identical with those described above. No radioactivity was found associated with the peak containing enzymatic activity, indicating that any complex formed between the enzyme and oxidized nucleotide was unstable under these conditions.
The combined results of the experiments described in this report support the concept that an equimolar complex is formed between rat liver dihydropteridine reductase and 1 mol of its preferred pyridine nucleotide cofactor, NADH. The formation of a similar complex between the enzyme and the oxidized nucleotide, NAD', is suggested by fluorescence measurements with the analog e-NAD+; however, this less stable complex could not be detected by the other procedures utilized to measure the NADH . enzyme interaction.
Because previous reports have indicated that the reductase possesses a dimeric structure composed of two identical subunits (7,8,16), it was expected that two nucleotide binding sites would exist and that the reductase would exhibit an affinity for 2 molecules of NADH. The dihydropteridine reductase isolated by Hasegawa (9) from bovine liver does, in fact, show these properties. It is possible that a second binding site may also be available in the rat liver enzyme, but that it may be activated only in the presence of the quinonoid dihydropteridine substrate. The unstable nature of this substrate (25), and the rapid turnover rate of the enzymatic reaction (16), preclude use of such a pteridine to test this hypothesis directly by the types of experiment described in this report. Electrophoresis experiments designed to detect ternary complex formation between the enzymatic reaction product, 2amino-li-hydroxy8,7-dimethyl-5,6,7,8-tetrahydropteridine, or the inhibitor, methotrexate, and NADH plus enzyme were unsuccessful. No evidence was seen of new protein staining bands with altered mobility, which might suggest the formation of such complexes. This unusual property of an apparent single nucleotide binding site shown by the rat liver reductase is not unlike that of thymidylate synthetase from methotrexate-resistant Lactobacillus casei, a dimeric enzyme which also shows a preference for substrate attachment (2'-deoxyuridine 5'-phosphate) to only one of the two sites which are available for binding (26,27). In this instance, the second site was revealed by the use of substrate analogs which inhibit the enzymatic reaction and bind more strongly to the active sites. At present, no similar analogs are known for the pteridine reductase.