/ . Biochem. 85, 887-899 (1979)

Isolation and Properties of a Glycohydrolase Specific for Nicotinamide Mononucleotide from Azotobacter vinelandii Tadayoshi IMAI Chemical Laboratory, College of General Education, Nagoya University, Furo-cho, Cnikusa-ku, Nagoya, Aichi 464 Received for publication, September 4, 1978

A glycohydrolase that catalyzes the irreversible conversion of NMN to nicotinamide and ribose 5-phosphate has been partially purified from a sonic extract of Azotobacter vinelandii. The enzyme is highly specific for NMN. NAD, NADP, nicotinic acid-adenine dinucleotide, nicotinamide riboside, and a-NMN are not significantly hydrolyzed by this enzyme, nor Ido they compete with NMN. The enzyme also exhibits an absolute dependence on guanylic acid derivatives with the following order of relative effectiveness: GTP, guanosine 5'-tetraphosphatodGTP, GDP, 2'-GMP, 3'-GMP>GMP, dGMP. A heat-resistant, nondialyzable factor which could replace the GTP requirement was found in the sonic extract. The KA for GTP and the Km for NMN in the presence of GTP at 1 mM were calculated to be 0.025 mM and 4.5 mM, respectively. GMP, dGMP, and dCMP were found to be effective inhibitors of the enzyme when 1 mM GTP was also present. The kinetic data suggest that the binding site for these mononucleotides is distinct from the active site or the GTP binding site. The ability of this enzyme to cleave NMN is suggestive of a metabolic role of the enzyme in selective conversion of NMN to nicotinamide, which, in turn, would be re-utilized by the cell as a precursor of NAD via nicotinic acid.

There have been a number of reports in the literature on the synthesis, breakdown, and resynthesis of NAD (for reviews, see references 1 and 2). An interesting aspect of pyridine nucleotide metabolism is its cyclic nature. Although the Abbreviations: NR, nicotinamide riboside; NaMN, nicotinic acid mononucleotide; NaAD, nicotinic acidadenine dinucleotide. The symbol a- represents the configuration of the glycosidic bond between the pyridine ring nitrogen and the ribose moiety. Unless otherwise indicated, the pyridine nucleotides and nucleosides have ^-configuration. Vol. 85, No. 4, 1979

887

pathway to NAD from nicotinamide via nicotinic acid has been believed to be the main route for NAD biosynthesis in most bacteria, Ohtu et al. (5) have found that Lactobacillus fructosus appears to form NAD from nicotinamide via NMN without being converted to nicotinic acid. Dietrich et al. (4) have isolated NMN pyrophosphorylase [EC 2.4.2.12] from mammalian tissues and thus supported the hypothesis that this type of enzyme could participate in biosynthesis of NAD (5, 6). In contrast, the author (7) has presented evidence that NMN is in part converted to NaMN in Azotobacter vinelandii. The NaMN thus formed could be re-utilized for NAD biosynthesis via

888

T. IMAI

NaAD far more economically than the nicotinamide turnover cycle proposed by Gholson (8, 9), in regard to the number of enzymatic steps and the energy required. During the course of studies on the NMN amidohydrolase reaction, an unexpected finding was made that not only NaMN but also free nicotinamide was formed from NMN on incubation with a sonic extract of A. vinelandii. This result suggested the existence of an alternate pathway of NMN metabolism, which involves the initial cleavage of NMN at the iV-ribosidic linkage, catalyzed perhaps by a glycolytic enzyme (7). This report is concerned with the isolation and some enzymatic properties of a glycohydrolase capable of catalyzing the hydrolysis of NMN to nicotinamide and ribose 5-phosphate. Andreoli et al. (10) have reported the occurrence of a membrane-bound NMN glycohydrolase in Escherichia coli but its properties have not been described in detail. EXPERIMENTAL PROCEDURES Materials—The following commercial materials were used: NAD, NADP, a-NAD, purineand pyrimidine nucleotides and their corresponding nucleosides and bases from Sigma; Sephadex G-200 and Sepharose 6B from Pharmacia; DEAEand ECTEOLA-cellulose from Brown; molecular weight markers (ferritin, M.W. 450,000; catalase, M.W. 240,000; aldolase, M.W. 158,000; bovine serum albumin, M.W. 67,000; hen egg albumin, M.W. 45,000; chymotrypsinogen A, M.W. 25,000) from Boehringer Mannheim; baker's yeast from Oriental Yeast Company. The remainder of the reagents were analytical-grade commercial products purchased from Sigma, Nakarai Chemical Company, or Wako Chemical Company. All other pyridinium compounds including NMN, NR, NaAD, NaMN, NaR, and their corresponding a-isomers used in this experiment were prepared as described in the previous paper (7). Calcium phosphate gel was prepared by the method of Sarker and Sumner (11). Paper Chromatography and Paper Electrophoresis—The following paper chromatographic solvent systems were employed generally in descending chromatography on Toyo No. 51A filter paper: solvent A, 1-butanol: 0.1 M ammonia (6:1); B,

2-butanol: 80% formic acid: water (200 : 49 : 81); C, butyric acid: 0.5 M ammonia (5 : 3); D, 95% ethanol: 1 M ammonium acetate, pH 7 (15:6). Paper electrophoresis was carried out by the method described previously (7) using 0.05 M acetic acid-ammonium acetate, pH 5. The pyridine nucleotides and bases were generally located on the paper by viewing under ultraviolet light. To locate N^substituted nicotinamide compounds, guide strips were cut and exposed to an atmosphere of ammonia-methyl ethyl ketone, 1 : 1 (v/v) (12). The compounds could then be located as fluorescent spots under ultraviolet light. The compounds were eluted from the remainder of the chromatogram or the electrophoretic strip with water. Procedures for the analytical paper chromatography of reaction products of NMN glycohydrolase were essentially the same as that described above except that the guide strips were stained with the silver nitrate reagent (13) for reducing sugar, the ammonium molybdate reagent (14) for phosphoester, or the periodate-benzidine reagent (15) for vicinal diol. Analytical Procedures—The following compounds were determined by the indicated methods: phosphate, by the method of Fiske and SubbaRow (16); ribose 5-phosphate, by the method of Schneider (17); nicotinamide, by a modification (18) of the method of Kodicek (19). Bacteria—The growth and harvesting of Azotobacter vinelandii strain O, Escherichia coli BH, and E. coli W3110 cells were described previously (7). Preparation of Cell-free Extract—All procedures were carried out at 0-4°C. Fifty grams of packed A. vinelandii cells was suspended in 100 ml of 0.025 M Tris-HCl, pH 7.5, containing 1 mM reduced glutathione and treated at 4.7±0.3 ampere for 30min in a Tomy sonicator, model UR-150P (20kHz). The cell debris was removed by centrifugation at 17,000 xg for 30min, and the resulting supernatant ("sonic extract") was used as an enzyme source. In the range from 2 to 60 min of sonication time, the specific activity of NMN glycohydrolase in the sonic extract did not change greatly, but the conditions described above were found to be best in regard to the total activity recovered. Cell-free extracts of E. coli, yeast, and rabbit /. Biochem.

NMN GLYCOHYDROLASE FROM A. vinelandii heart and homogenates of rabbit liver and of pancreas were prepared by the methods described previously (7). Assay Procedures for NMN Glycohydrolase— NMN glycohydrolase activity was determined spectrophotometrically by following the rate of NMN hydrolysis using a modification of the cyanide-addition method (20). The reaction mixture for routine assay contained 0.4 /imol of NMN, 7/imol of Tris-HCl, pH 8.7, 0.1 [tmo\ of GTP (see below), 0.02 (imo\ of reduced glutathione, and enzyme solution in a total volume of 0.1 ml. The mixture was incubated at 39°C for 50 min: the reaction was started by addition of the enzyme and terminated by heating for 90s at 100°C. After centrifugation, 50 fi\ of each reaction mixture was removed and mixed with 1 ml of 2 M KCN. The decrease of absorbance at 307 nm was measured and the amount of NMN hydrolyzed was calculated by using a value of 4.6 x 10s for molecular absorption coefficient (21). NMN amidohydrolase, if any, did not interfere this assay. In the case of purified enzyme preparations free of NMN amidohydrolase, NMN glycohydrolase activity was measured spectrophotometrically at 327 nm (e = 6.2 x 10*) (method 1).

889

The following enzyme activities were assayed by the indicated methods: NMN amidohydrolase, by the method described previously (7); NAD glycohydrolase [EC 3.2.2.5], by the method of Kaplan et al. (20); nicotinamide deamidase, by the method described by Petrack et al. (23); phosphodiesterase and phosphatase, by the method of Bjork (24). RESULTS

Partial Purification of NMN Glycohydrolase from Azotobacter vinelandii—Unless otherwise stated, all steps were performed at 0-4°C. In a typical preparation, 50 g of packed cells of A. vinelandii was used. All centrifugations were at 17,000 x g for 30 min. Step 1. Streptomycin precipitation: Five percent streptomycin sulfate in 0.025 M Tris-HCl, pH 7.5, containing 1 mM reduced glutathione (buffer A) was added slowly with stirring to the sonic extract (134 ml) of A. vinelandii to give a final concentration of 0.75%. The mixture was allowed to stand for 30 min and the supernatant liquid was collected by centrifugation (143 ml). Step 2. Ammonium sulfate fractionation: After the protein concentration of the supernatant (140 In some experiments, in which an iV^-sub- ml) had been adjusted to 20 mg per ml, solid ammostituted pyridinium compound was added to the nium sulfate was added slowly with stirring to the reaction mixture as an inhibitor, enzyme activity diluted supernatant to give a final concentration of was assayed by measuring nicotinamide formation. 33% saturation. Stirring was continued for an After the reaction had been terminated by addition additional 1 h, the mixture was centrifuged and of 30 fi[ of 1 % acetic acid, the reaction mixture was the precipitate discarded. The resulting superspotted on Toyo No. 51A filter paper and chro- natant solution was brought slowly to 50% saturamatographed with solvent A. The UV-absorbing tion with solid ammonium sulfate. The mixture spot corresponding to nicotinamide on the chro- was allowed to stand for 1 h and centrifuged. The matogram was cut out and eluted with water. precipitate was dissolved in buffer A and dialyzed Nicotinamide in the eluate was determined by the overnight against 4 liters of the same buffer (24 ml). method described above (method 2). For the Step 3. Sephadex G-200 gel filtration: The recontrol run, the enzyme was replaced with water sulting solution (23 ml) was applied to a Sephadex or a boiled enzyme preparation. The two methods G-200 column (2.7x121 cm), which had been described above gave nearly identical results. equilibrated with buffer A. The column was One unit of enzyme activity was defined as the eluted with the same buffer at a flow rate of 12 ml amount of enzyme required to hydrolyze 1 ^mol per h and 5.1 ml fractions were collected. A of NMN in 1 min at 39°C. Specific activity was typical elution profile is shown in Fig. 1. Enzyme expressed as units per mg of protein. Protein was activity was separated into two fractions: one determined according to the method of Lowry having a maximum at fraction 60 (Fraction A) and et al. (22), with crystalline bovine serum albumin the other at fraction 80 (Fraction B). as a standard. The reason for the occurrence of two fractions Measurement of Activities of the Enzyme with apparently identical substrate specificity is not Responsible for Pyridine Nucleotide Metabolism— known at present. The major peak, Fraction A, Vol. 85, No. 4, 1979

890

T. IMAI

is the primary concern of this paper. Fraction B will be described elsewhere. Fraction A was pooled and concentrated to 9 ml on a membrane filter in a SSrtorius ultrafiltration apparatus. Step 4. Calcium phosphate gel adsorption: The concentrated Fraction A was mixed with calcium phosphate gel (50 mg dry weight per 1 ml of enzyme preparation) which had been equilibrated with buffer A, stirred for 15 min, and centrifuged. The gel pellet was washed twice successively with 16 ml each of 0.04 M and 0.08 M phosphate (Na) buffer, pH 7.5, containing 1 mM reduced glutathione. The bulk of the enzyme activity was found in the first 0.08 M phosphate washing. The two 0.08 M phosphate fractions were combined, concentrated on a membrane filter and dialyzed against 1 liter of buffer A (5 ml).

Fraction A

10

0.4

0,2

fc 50 100 FRACTION DUMBER Fig. 1. Sephadex G-200 gel filtration of the ammonium sulfate fraction. T h e concentration of protein is expressed as the absorbance of the effluent at 280 nm. Details are given in t h e text. Fractions from 54 to 67 were pooled.

A summary of the purification scheme is shown in Table I. The final specific activity was 0.32 unit per mg protein, which represents a 48fold purification. The recovery was 20 %. Throughout the purification procedures described above, neither interconversion between Fraction A enzyme and Fraction B enzyme nor degradation of these to smaller molecular weight species occurred, as judged by gel filtration experiments. However, if the enzyme was purified by different methods using DEAE- or ECTEOLAcellulose, the resulting preparations behaved as smaller molecular weight proteins. The final preparation was free of NAD glycohydrolase, NMN amidohydrolase and activities degrading nucleoside triphosphates or diphosphates, but was still contaminated with a trace of AMP nucleosidase (less than 5 % of the activity of NMN glycohydrolase). Although more than 90% of the AMP nucleosidase activity present in Fraction A could be removed by calcium phosphate gel treatment, it was not possible to obtain an NMN glycohydrolase preparation completely free of the nucleosidase. Since this remaining nucleosidase activity did not interfere with the assay of NMN glycohydrolase, no further attempt was made to remove this minor contaminant. In the following experiments, the calcium phosphate fraction was used unless otherwise specified. Identification of Reaction Products—To identify the reaction products, the reaction was performed with 0.1 unit of purified enzyme on a scale 20 times greater than in the assay runs, and allowed to proceed for 3 h. The solution was then subjected to paper chromatography with solvent A. Only a single UV-absorbing band was detected as a reaction product and the material was recovered by elution with water. When aliquots of this

TABLE I. Purification of NMN glycohydrolase from A. vinelandii. Fraction Sonic extract Streptomycin supernatant Ammonium sulfate Sephadex G-200 Fraction A Calcium phosphate

Total activity Total protein (units) (mg) 39.9 50.9 38.9 20.7 7.6

6,140 3,660 1,330 210 24

Specific activity (units/mg protein)

Recovery (%)

0.0065 0.0139 0.0293 0.0987 0.317

100 128 98 52 19

/. Biochem.

NMN GLYCOHYDROLASE FROM A. vinelandii material were separately subjected to paper chromatography with solvents A, B, C, and D, the compound behaved in the same way as an authentic sample of nicotinamide. Its UV absorption spectra at pH 2 and 12 were also in agreement with those of nicotinamide. On the other hand, when aliquots of the reaction mixture were separately subjected to paper chromatography with solvents C and D and to paper electrophoresis at pH 5, staining with silver nitrate, periodate-benzidine, or perchloridatemolybdate indicated the formation of a reducing phosphoester which behaved in the same way as an authentic sample of ribose 5-phosphate. Thus, the results indicate that the enzyme hydrolyzes N M N to give rise to nicotinamide and ribose 5phosphate. Stoichiometry of the Reaction—The stoichiometry of N M N glycohydrolase reaction was TABLE II. Stoichiometry of reaction. The reaction mixture contained, in a final volume of 1 ml: NMN 4 /itnol, Tris-HCl, pH 8.7, 70 pmol, reduced glutathione 0.2 //mol, and GTP 1 pmol. The reaction was initiated by the addition of 0.017 unit of the most highly purified enzyme and the reaction mixture was allowed to stand at 39°C. At each indicated time, an aliquot (50 p\) was removed from the reaction mixture, heated at 100°C for 90 s and centrifuged. The supernatant solution was spotted on Toyo No. 51A paper and chromatographed with solvent A. Nicotinamide and NMN spots were eluted separately with water and determined by a modification (75) of the method of Kodicek (79) and by the cyanide-addition method (21), respectively. Another aliquot (50 /*I) treated as above was spotted on Toyo No. 51A paper and chromatographed with solvent C. The ribose 5-phosphate spot, which was located as described in the text, was eluted with water and analyzed for phosphate and reducing sugar by the method of Fiske and SubbaRow (16) and by the method of Schneider (77), respectively. A control sample which did not contain the enzyme was run concurrently. Incubation time (h)

Decrease of NMN (/imol) 0.043 0.074 0.110

Vol. 85, No. 4, 1979

Increase of Nicotinamide (ftmot) 0.048 0.076 0.115

Ribose 5-phosphate (/imo\)

0.046 0.066 0.091

891 followed during a 4 h incubation period. The results are shown in Table IL The decrease in the amount of N M N was matched by the increase in the amount of nicotinamide as well as the increase in the amount of ribose 5-phosphate. The conversion of N M N to nicotinamide and ribose 5-phosphate appears to be irreversible, for all attempts to demonstrate the formation of N M N from these products have been unsuccessful. Substrate Specificity—As shown in Table HI, no hydrolysis was noted when N A D , N A D P , N R , N a A D , N a R , a - N A D , a - N M N , a-NR, a - N a A D , a - N a M N or a - N a R was incubated with the enzyme preparation. N a M N was hydrolyzed slowly (at about 8 % of the rate of N M N hydrolysis) to give rise to nicotinic acid and ribose 5-phosphate. Properties of the Enzyme and the Enzyme Reaction—1) Stability of the enzyme: At all stages in the purification procedures the enzyme remained active for at least 1 week at 4°C in 0.025 M Tris-HCl, p H 7.5, containing 1 mM reduced glutathione, and could be stored at either 0 or — 16°C for a month without significant loss of activity. In 0.025 M Tris-HCl, p H 9, containing 1 mM reduced TABLE m . Specificity for substrate. Test compounds were added at a concentration of 2 mM. Standard assay conditions were used with 2 x 10"3 unit of the most highly purified enzyme for each measurement. Activities were determined by measurement of the hydrolysis of each pyridine compound using the cyanide-addition method (27). Compound NMN NaMN NAD NADP NR NaAD NaR a-NMN a-NAD a-NR a-NaAD a-NaMN o-NaR

Relative activity (%) 100 8 2 1 2 0 4 0 0 0 0 0 0

892

T. IMAI

glutathione, however, about one-half of the enzyme activity was lost in 5 days at 0°C. 2) Molecular weight: Gel filtration with Sephadex G-200 was used according to the procedure of Andrews (25) to estimate the molecular weight of NMN glycohydrolase. The elution profile of the enzyme relative to standard compounds indicated that its apparent molecular weight is approximately 213,000 (Fig. 2). A similar experiment with Sepharose 6B gave a slightly higher value, 240,000. 3) Effects of enzyme concentration and incubation time: Under the standard assay conditions, the rate of reaction was proportional to enzyme concentration (1-12 //g protein) and, when 6 ^g of enzyme was used, remained constant with time of incubation for at least 100 min. 4) Effect ofpH: The maximum activity was observed at pH 8.5-9.0 in Tris-HCl buffer (Fig. 3).

4.5 £ 1

In contrast, when the enzyme was assayed in glycine-NaOH buffer, the enzyme was less active in this pH range. The enzyme was also shown to be quite sensitive to buffer concentration. With Tris-HCl, pH 8.7, the enzyme activity was maximal at a buffer concentration of 0.07 M. In 0.04 M, 0.03 M, 0.02 M, or 0.01 M of the same buffer, the activity was 93, 81, 63, or 21 %, respectively, of the activity in 0.07 M Tris-HCl, pH 8.7. 5) Effect of temperature: As Fig. 4A indicates, the rate of NMN hydrolysis was maximal at 39°C. The activation energy calculated from the negative slope —E/2.303R (E is the activation energy, and R is the gas constant) of a standard Arrhenius plot, i.e., the logarithm of reaction velocity plotted against the reciprocal of absolute temperature, has a value of 11 kcal per mol (Fig. 4B). 6) Effects of cations, onions, and -SH inhibitors: The enzyme did not require the addition of any metal ions for full activity, and several metals were inhibitory when assayed under the standard assay conditions. Cd t+ , Cu1+, and Zn2+, at 1 mM, caused considerable inhibition, ranging from 40% to 90%. Although the enzyme did not seem to require the addition of any metal for activity, 1 mM EDTA inhibited NMN hydrolysis by 30%. The reason for this effect of EDTA is not known.

1.5 Ve / Vo

Fig. 2. Molecular weight estimation of NMN glycohydrolase by gel filtration on Sephadex G-200. A column (1.2x77.2 cm) of Sephadex G-200 equilibrated with 0.025 M Tris-HCl, pH 7.5, containing 1 mM reduced glutathione, was prepared. The column was run at 4°C at a downward flow rate of 5 ml/h and 2 ml fractions were collected. The following standard proteins were used: (1) ferritin, M.W. 450,000; (2) catalase, M.W. 240,000; (3) aldolase, M.W. 158,000; (4) bovine serum albumin, M.W. 67,000; (5) hen's egg albumin, M.W. 45,000. The void volume ( Vo) was determined with blue dextran. The elution position of A. vinelandii NMN glycohydrolase is marked by the filled circle ( • ) . The estimated molecular weight for the enzyme was 213,000.

10

Fig. 3. Effect of pH on NMN glycohydrolase activity. The enzyme reaction was performed under standard conditions using 0.07 M Tris-HCl (O), glycine-NaOH (A), or phosphate buffer (D) at various pH's. /. Biochem.

NMN GLYCOHYDROLASE FROM A. vinelandii

oy o b o.i

< B )

- 2.6

-

20

30

40

50

TEHPERATURE (°C)

3.2

2.7

3.3

TABLE IV. Requirements of NMN glycohydrolase. The complete incubation mixture contained 0.4 ftmol of NMN, 0.1 fimol of GTP, 7 fimol of Tris-HCl, pH 8.7, 0.02 fimol of reduced glutathione, 3 x 10"3 unit of the enzyme and, where indicated, 20 fi\ of boiled extract, in a total volume of 0.1 ml. The mixture was maintained at 39°C for 50 min. After reaction, the amounts of NMN hydrolyzed were measured by the cyanide method, as described in the text. System

NMN hydrolyzed

( 1/T )xl03 (

Fig. 4. The velocity of NMN hydrolysis as a function of temperature. (A) Conditions were the same as for the standard assay, except that incubation was carried out at the indicated temperatures. 3 x 10"1 unit of the most highly purified enzyme was used. (B) Effect of temperature on the kinetic parameters of NMN glycohydrolase. The logarithms of the initial velocities of the reaction are plotted against the reciprocal temperature in °K. The plot was linear between 25-38°C. The slope corresponds to —E/2.303R, which indicates that the activation energy is 10.6 kcal/mol. Anions such as phosphate, pyrophosphate, and sulfate had no effect on the enzyme. Further addition of reduced glutathione to give a final concentration of 1.2 mM had no effect on the enzyme activity. Nevertheless, the enzyme was sensitive to reagents which attack primarily sulfhydryl groups; p-chloromercuribenzoate (50% inhibition at 3 ^ M ) was considerably more potent than JV-ethylmaleimide (50% inhibition at 80 fiM) in this regard. Iodoacetamide and iodoacetic acid had no significant effect at 1 mM. 7) Effects of various structural analogs: Effects of the reaction products and various other compounds metabolically related to pyridine nucleotide on NMN glycohydrolase were tested at 2mM. Thus, NAD, NADP, NaAD, NaMN, nicotinic acid, quinolinic acid, 5-phosphoribosyl 1-pyrophosphate, nicotinamide, and ribose 5phosphate had little or no inhibitory effect on the enzyme. In the presence of both nicotinamide and ribose 5-phosphate at 4 mM, 28 % inhibition was observed. 8) Effect of crude boiled extract: In the course of attempted purification it became evident that the purified enzyme lost over 90% of its activity after Sephadex G-200 treatment and that the missing activity could be totally restored by the Vol. 85, No. 4, 1979

893

Complete Complete Complete Complete Complete Complete

0.150 minus GTP 0.003 minus GTP plus boiled extract 0. 113 plus boiled extract 0.155 minus enzyme 0.000 minus enzyme plus boiled extract 0. 000

addition of a boiled sonic extract (referred to as "boiled extract") of the same cells. GTP (1 mM) could substitute for this boiled extract. The requirements for these materials of the enzyme are summarized in Table IV. As shown in Fig. 5, the enzyme activity depended on the concentration of the boiled extract and increased to 40-fold at maximum. The rate of reaction was proportional to the concentration of the boiled extract in the range of 1-5 (i\ per incubation mixture, and a saturating level was reached with 20 (t\ per incubation mixture. The boiled extract of A. vinelandii cells was obtained by heating the sonic extract for 5 min at 100°C followed by centrifugation and dialysis against 1 mM Tris-HCl, pH 7.5 (protein concentration, 1.5 mg/ml; density units at 260 ran, 50/ml). The nature of this cellular component is unclear at present and remains to be elucidated. 9) Requirement for guanylic acid derivatives: Table V shows the effects of various purine and pyrimidine nucleotides on the enzyme activity. Of all the compounds tested, guanylic acid derivatives were found to be the potent activating agents, with the following order of relative effectiveness; GTP, guanosine 5'-tetraphosphate>dGTP, GDP, 2'-GMP, 3'-GMP>GMP, dGMP. ATP is a less effective activator than GTP. CTP, UTP, TTP,

894

T. I M A I

TABLE V. Effects of various nucleotides on NMN glycohydrolase. Experimental conditions were as in Table IV, except that the enzyme activity was assayed in the absence of GTP and boiled extract (Experiment 1), in the presence of GTP at 1 mM (Experiment 2), or in the presence of boiled extract (20 ii\ per 0.1 ml) (Experiment 3). The indicated nucleotides or nucleoside were added to give a final concentration of 1 mM. The results are given as relative activity standardized in terms of the enzyme activity in the presence of 1 mM GTP as an effector (Experiment 1). Relative activity (°Yo) Compound 0

20 10 BOILED EXTRACT , pL/REACTION MIXTURE Fig. 5. Dependence of N M N glycohydrolase activity on boiled extract. Conditions were the same as for the standard assay, except that the boiled extract was added with n o G T P addition. 3 x 1 0 " * unit of the most highly purified enzyme was used. After 50 min at 39°C, the enzyme activities were assayed by the method described in the text.

None Guanosine GMP dGMP 2'-GMP 3'-GMP 3',5'-cGMP GDP GTP dGTP Guanosine 5'-tetraphosphate ATP dCMP

Experiment Experiment 1Experiment 3 1 1.8 0 23 22 90 85 3 94 100 91

100 99 18 18 114 114 105 95 105 100

75 74 27 23 86 85 86 84 103 96

and their corresponding nucleoside mono- and diphosphates failed to support the catalytic activity. 114 96 110 Figure 6 shows the results of plotting concen31 50 99 trations of these effective nucleotides with respect 0 41 27 to enzyme activity. The concentration of each nucleotide that permits the enzyme to function at half-maximal velocity, i.e. KA value, was determined by means of the usual double-reciprocal tives, GMP and dGMP are unique in their effects plot (I/velocity versus 1/nucleotide concentration). on the enzyme activation (Fig. 6): this suggests The extrapolated intercept on the 1/nucleotide that these mononucleotides have a bimodal funcconcentration axis was used as a measure of — 1/ tion. 10) Specificity and kinetics of activation of apparent A"A for the effector. The values obtained are as follows: GTP, 0.025 HIM; dGTP, 0.080 mM; NMN glycohydrolase by GTP: Figure 7 shows guanosine 5'-tetraphosphate, 0.020 mM; 3'-GMP, the results of plotting substrate concentration with 0.21 mM; 2'-GMP, 0.030 mM; GMP, 0.23 mM; respect to enzyme activity in the presence and absence of GTP. In both cases, a hyperbolic dGMP, 0.44 mM; and ATP, 0.38 mM. It should be noted that GTP and guanosine relationship of these variables is noted and a 5'-tetraphosphate operate as the most potent straight line is obtained when these data are plotted activators and that the apparent KA values for in terms of 1/v with respect to 1/[S]. The data these nucleotides are very low and virtually identi- indicate that the apparent Km value for NMN is cal. Apparently, GTP and guanosine 5'-tetra- 2 mM in the absence of GTP and that GTP increases phosphate act as a "prime effector"1 for NMN the Vmtx value, concomitant with an alteration in glycohydrolase. Among the guanylic acid deriva- the apparent Km for NMN (4.5 mM). When the Km value for NMN was determined 1 in the presence of various concentrations of GTP, The terminology was suggested by Beck (26). /. Biochem.

NMN GLYCOHYDROLASE FROM A. vlnelandii

895

C 0 N C. 2

3

1

NUCLEOT1DE (J1)

Fig. 6. Comparison of the effects of increasing concentrations of several guanylic acid derivatives and ATP on NMN hydrolysis. Conditions were the same as for the standard assay, except that GTP was absent and the concentrations of nucleotides were varied. Other details were as in the legend to Fig. 5. G tetraP, guanosine 5'-tetraphosphate. 0.5 mM dGTP, or 0.5 mM guanosine 5'-tetraphosphate, the value obtained was 4.5 mM (Table VI). Since the Km value for NMN appeared to be independent of the concentration of GTP in the reaction mixture and of the structure of the guanylic acid derivatives that serve as effectors, and since a typical double-reciprocal plot was obtained for GTP, it is suggested that the enzyme contains an independent site for binding GTP. The kinetic data for NMN glycohydrolase action upon its substrate can be expressed by the empirical Hill equation. If the initial rate of the enzyme activity as a function of substrate concentration is plotted as shown in Fig. 8A, one obtains a straight line with a slope of unity. The Hill equation can also be applied to the kinetics of activation of the enzyme by GTP. A plot of the enzyme activity as a function of increasing concentration of GTP also gives a straight line with a slope of unity (Fig. 8B). These data indicate that there is no cooperative binding of either NMN or GTP to the enzyme molecule. Vol. 85, No. 4, 1979

O F NMN

Fig. 7. Dependence of NMN glycohydrolase activity on NMN concentration in the presence or absence of GTP. Enzyme assays were carried out under the standard conditions, except that the NMN concentration was varied in the absence (O) or presence of 1 mM GTP ( • ) . 5x10-' unit of the most highly purified enzyme was used. The incubation was performed at 39°C for 90min in the experiment without GTP and for 20 min in the experiment with the effector. The initial velocities are expressed as nmol of NMN hydrolyzed per min. Inset shows a double-reciprocal plot of the reaction rate versus NMN concentration.

TABLE VI. Km of NMN glycohydrolase for NMN in the presence of various effectors. Conditions for incubation were as described in the legend to Fig. 7, except for changes in the concentration of GTP or its replacement by other guanylic acid derivatives as indicated. (mM)

Km for NMN (rrw)

None GTP

— 0.20

2.0 4.5

GTP

0.50

4.5

GTP

1.0

4.5

Guanosine 5'-tetraphosphate

0.50

4.4

dGTP

0.52

4.0

Effector

Concentration

896

T. IMAI I ( A

>

)

.03/

1

J

3 -

0 "

/

•n-1.07

o -

-1

2

I >

lo» [SMO

•

log [GTP1

Fig. 8. Order with respect to N M N (A) and G T P (B) of the reaction catalyzed by N M N glycohydrolase. The initial velocities, substrate and G T P concentrations, and assay conditions for the data in (A) and (B) were those shown in Figs. 7 and 6, respectively.

A

—*-

A

1

2 1

/ C GTP 3

Fig. 10. Kinetics of the inhibition of N M N glycohydrolase by G M P and dCMP with respect to GTP. Standard assays were carried out, except that the G T P concentration was varied in the presence of 1 mM G M P ( • ) , 0.50 mM G M P ( T ) , or 0.50 mM dCMP ( A ) . (O), Control. Other details and data treatment were as in the legend to Fig. 9. The data are shown as a doublereciprocal plot of velocity versus GTP concentration.

i / CNHH: Fig. 9. Kinetics of the inhibition of N M N glycohydrolase by G M P , dGMP, and dCMP with respect to N M N . Standard assays were carried out, except that the N M N concentration was varied in the presence of l m M G M P ( • ) , 0.50 mM G M P ( T ) , 1 mM d G M P ( • ) , or 1 mM d C M P ( A ) . (O), Control. Incubations were performed with 1.5x10"* unit of the most highly purified enzyme. The initial velocities are expressed as nmolof N M N h y d r o l y z e d p e r m i n . The data are shown as Lineweaver-Burk plots.

11) Inhibition by GMP, dGMP, and dCMP in the presence of GTP: Since the maximum velocity obtainable with GMP as an activator is considerably lower than that observed with GTP, GMP can be regarded as an inhibitor of NMN hydrolysis, provided that GTP is also present. Indeed, marked inhibitions of the enzyme by GMP, dGMP, and dCMP were observed in the presence of a saturating concentration of GTP or the boiled extract (Experiments 2 and 3 in Table V). When the inhibition of the enzyme at a constant GMP concentration was studied with various substrate concentrations, a noncompetitive inhibition of the enzyme with respect to the substrate was observed. This is illustrated in Fig. 9, from which the average Ki value can be calculated to be 0.23 mM. The Ki value obtained is essentially equal to the KA value for GMP as an effector for the enzyme (see above). The data in Fig. 9 also indicate that dGMP and dCMP are noncompetitive /. Biochem.

NMN GLYCOHYDROLASE FROM A. vinelandii inhibitors with respect to the substrate with Kt values of 0.37 HIM and 0.70 mM, respectively. When the inhibition of the enzyme at a con^ stant GMP concentration was studied with various GTP concentrations, no mutual competition was observed (Fig. 10). The data show that the degree of enzyme activity in the presence of GMP in a range of 0.5-1 mM is dependent not on GTP concentration but on GMP concentration, suggesting that GMP may function to desensitize the GTP effect on the enzyme. In the case of dCMP, which exhibits no activation of the enzyme (Table V), the inhibition mechanism appears to be identical with that of GMP inhibition (Fig. 10). In order to confirm these complex effects of GMP on the enzyme, further studies were done. Figure 11 shows the results of plotting the concentration of GMP with respect to enzyme activity in the absence or presence of various concentrations of GTP. The data show that at lower concentrations of GMP the enzyme activity is dependent on GTP concentration but at higher concentrations

897

above 1 mM, the enzyme activity is strictly dependent on GMP concentration regardless of the GTP concentration. This result reinforces the view of the inhibition mechanism by GMP described above. All these results indicate that GMP does not show any competition with NMN or with GTP. Therefore, the enzyme must have at least one locus for GMP binding, which is topologically distinct from the active site or GTP binding site. NMN Glycohydrolase and Related Enzymes in Sonic Extract of A. vinelandii—The activities of some enzymes responsible for pyridine nucleotide metabolism in A. vinelandii were measured as a function of cell growth (Fig. 12). In contrast to NMN amidohydrolase, which had been shown to have a fairly constant activity (specific activity, 3 x 10~* (7)), NMN glycohydrolase activity varied with the growth phase of the cells. Nevertheless, this enzyme appeared to be at least 10 times more active than NMN amidohydrolase throughout the culture period. Both NAD glycohydrolase and

100

0 OF G M P

10 20 CULTIVATION TIME (hr)

Fig. 11. Effect of GMP concentration on NMN glyco- Fig. 12. NMN glycohydrolase, NAD glycohydrolase, hydrolase activity at various concentrations of GTP. nicotinamide deamidase, and phosphodiesterase activiStandard assays were carried out, except that GMP was ties as functions of cultivation time. Specific activities added to the reaction mixture to give the indicated con- of each enzyme were determined by the methods decentration in the absence (O) or presence of 1 mM GTP . scribed in the text at various times during culture. D, ( • ) or 0.1 mM GTP ( A ) . Other details were as in the Cell growth expressed as increase in absorbance at legend to Fig. 9. The data are given as relative activity 600nm; • , NMN glycohydrolase; O, NAD glycostandardized in terms of the enzyme activity in the hydrolase; • , nicotinamide deamidase; A, phosphodiesterase. presence of 1 mM GTP alone. Vol. 85, No. 4, 1979

T. IMAI

898

phosphodiesterase activities also varied with culture time, while the nicotinamide deamidase activity varied less significantly. Distribution in Other Organisms—The specific activities of NMN glycohydrolase in crude extracts were: A. vinelandii (6.5±10)xl0-'; E. coli BH, 2.3x10-'; E. CO//W3110, 1.7x10"'; and bakers' yeast, 0.4xlO" 3 . Those in homogenates were: rabbit liver, 1.3x10"*; rabbit pancreas, 10x10"'; and rabbit heart, 1.7 x 10~3. The data indicate that NMN glycohydrolase is distributed more widely than NMN amidohydrolase (7). In all the extracts and homogenates examined, the specific activities of NMN glycohydrolase were higher than those of NMN amidohydrolase.

Fig. 13. Proposed pathways of pyridine nucleotide metabolism in A. vinelandii.

to nicotinamide, which would, in turn, be subjected to deamidation by nicotinamide deamidase to yield nicotinic acid. The nicotinic acid thus produced could be re-utilized as a raw material for the synthesis of NAD via NaMN and NaAD. The same organism has been suggested to have DISCUSSION another pathway in which deamidation takes place The results described in this paper show that A. at the mononucleotide level (7). The NaMN vinelandii contains a glycohydrolase which catalyzes formed, then, could be utilized for NAD biosynthe irreversible hydrolysis of the iV-ribosidic linkage thesis via NaAD. These observations lead to a of NMN. Several nucleoside hydrolases capable proposal for a new pyridine nucleotide cycle (Fig. of hydrolyzing nicotinamide riboside are known 13) which is more complex than has been believed in bakers' yeast (27), Lactobacillus delbrueckii (28), so far (8, 9). and Pseudomonas fluorescens (29). It was also The TV-glycosidic bond hydrolysis (reaction X) reported previously (50) that NAD glycohydrolase and the deamidation (reaction VII) provide isolated from bull semen can hydrolyze (in addition branches. Although two separate enzymes could to NAD and NADP) NMN at a rate of approxi- participate in NMN metabolism, NMN glycomately 10% of that of NAD hydrolysis. The hydrolase is more active and is more likely to be NMN glycohydrolase described here is distinctly subject to control by various nucleotides and different from these enzymes in its high substrate cellular factors than is NMN amidohydrolase. In specificity and its absolute requirement for GTP. this respect, it is noteworthy that NMN glycoIt is remarkable that the enzyme shows such hydrolase is activated by various guanylic acid a high degree of specificity for NMN. With one effectors, especially by GTP and guanosine 5'exception, no pyridine nucleotides and nucleosides tetraphosphate. Furthermore, several lines of are cleaved. The exception is NaMN, but the rate evidence indicate that GMP, at concentrations of its hydrolysis is only about 8 % of the rate of above 0.5 mM, desensitizes the GTP effect on the enzyme (Figs. 10 and 11). The mechanism of NMN cleavage. In the previous study (7), the author showed regulation by these nucleotides remains to be that in A. vinelandii the enzymes in the "nicotinic elucidated, but these observations can be interacid pathway," NaMN pyrophosphorylase [EC preted as suggesting that the differences between 2.4.2.11], NaMN adenylyltransferase [EC 2.7.7.18], the sensitivities of NMN amidohydrolase and NAD synthetase [EC 6.3.5.1], NAD glycohydrolase, NMN glycohydrolase towards various nucleotides and nicotinamide deamidase, were present, whereas may be an important factor for regulating the two the enzymes in the "nicotinamide pathway," NMN catabolic routes of NMN. The importance of pyrophosphorylase and NMN adenylyltransferase, NMN in pyridine nucleotide metabolism is further were either absent or present in very low amounts illustrated by the occurrence in A. vinelandii of a (see Table VI in Ref. 7). In this organism, there- phosphodiesterase and a pyrophosphorylase, both fore, pyridine nucleotides would be cleaved at capable of catalyzing the formation of NMN from both mono- and dinucleotide levels to give rise NAD (7). / . Biochenu

NMN GLYCOHYDROLASE FROM A. vinelandii The author is grateful to Dr. S. Suzuki, Department of Chemistry, Faculty of Science, Nagoya University, for critical readings of the manuscript and useful suggestions. REFERENCES 1. Kaplan, N.O. (1960) in The Enzymes (Boyer, P.D., Lardy, H., & Myrback, K., eds.) Vol. 3, pp. 105-169, Academic Press, New York 2. Kaplan, N.O. (1968) /. Vitaminology 14, 103-113 3. Ohtu, E., Ichiyama, A., Nishizuka, Y., & Hayaishi, O. (1967) Biochem. Biophys. Res. Commun. 29, 635-641 4. Dietrich, L.S., Fuller, L., Yero, I.L., & Martinez, L. (1966) /. Blol. Chem. 241, 188-191 5. Kornberg, A. (1950) /. Biol. Chem. 182, 779-793 6. Kornberg, A. (1950) /. Biol. Chem. 182, 805-813 7. Imai, T. (1973) /. Biochem. 73, 139-153 8. Gholson, R.K. & Kori, J. (1964) / . Biol. Chem. 239, PC2399-2400 9. Gholson, R.K. (1968) /. Vitaminology 14, 114-122 10. Andreoli, A.J., Okita, T.W., Bloom, R., & Grover, T.A. (1972) Biochem. Biophys. Res. Commun. 49, 264-269 11. Sarker, N.K. & Sumner, J.B. (1951) Enzymologia 14, 280 12. Kodicek, E. & Reddi, K.K. (1951) Nature 168, 475-477 13. Trevelyan, W.E., Procter, D.P., & Harrison, J.S. (1950) Nature 166, 444-445

Vol. 85, No. 4, 1979

899 14. Suzuki, S. (1962) /. Biol. Chem. 237, 1393-1399 15. Gordon, H.T., Thornburg, W., & Werum, L.N. (1956) Anal. Chem. 28, 849-855 16. Fiske, C.H. & SubbaRow, Y. (1925) /. Biol. Chem. 66, 375-400 17. Schneider, W.C. (1957) in Methods in Enzymology (Colowick, S.P. & Kaplan, N.O., eds.) Vol. 3, pp. 680-684, Academic Press, New York 18. Suzuki, K., Nakano, H., & Suzuki, S. (1967) /. Biol. Chem. 242, 3319-3325 19. Kodicek, E. (1940) Biochem. J. 34, 712-723 20. Kaplan, N.O., Colowick, S.P., & Nason, A. (1951) / . Biol. Chem. 191, 473-483 21. Imai, T. & Suzuki, S. (1971) Seikagaku (in Japanese) 43, 346-362 22. Lowry, O.H., Rosebrough, N.J., Farr, A.L., & Randall, R.J. (1951) J. Biol. Chem. 193, 265-275 23. Petrack, B., Greengard, P., Craston, A., & Sheppy, F. (1965) /. Biol. Chem. 240, 1725-1730 24. Bjork, W. (1963) /. Biol. Chem. 238, 2487-2490 25. Andrews, P. (1965) Biochem. J. 96, 595-606 26. Beck, W.S. (1967) /. Biol. Chem. 242, 3148-3158 27. Heppel, L.A. & Hilmoe, R.J. (1952) /. Biol. Chem. 198, 683-694 28. Takagi, Y. & Horecker, B.L. (1956) /. Biol. Chem. 225, 77-86 29. Terada, M., Tatibana, M., & Hayaishi, O. (1967) J. Biol. Chem. 242, 5578-5585 30. Abdel-Latif, A.A. & Alivisatos, S.G.A. (1962) /. Biol. Chem. 237, 500-505