Substrate Specificity and Aspects of Deamination Catalyzed by Rabbit Muscle 5’-Adenylic Acid Aminohydrolase*

of 24

Rabbit, muscle 5'-adenylic acid aminohydrolase was prepared as described by Smiley et at. (1). Unless otherwise indicated, activity was determined in 0.15 M KCl, 0.05 M Trk-cacodylate, pH 6.3, and 50 pM Tris--4MP (Assay 1) or 0.10 M (CHp)4NC1, 0.05 M Tris-cacodylate, pH 6.3, and 50 pM Tris-AMP (Assay 2). Specific activities in micromoles per min per mg were calculated from the change in absorbance per unit time per mg of protein by the following relationship: micromoles per min per mg = AA per min per mg per F where F, the change in molar absorbance, is equal to 8.86 X lo3 M-' cm-l at 265 nm, 0.30 X lo3 M-I cm-' at 285 nm, and 0.12 X lo3 M-' cm-l at 290 nm.
N1-Methyl-AMP was synthesized from AMP at pH 4.5 by the method of Griffin and Reese (7). The purified product exhibited only one spot in Solvent A (see Legend, Table III), and the RF value of 0.75 agrees well with the Rp of 0.76 previously reported (7).
N6-Methyl-AMP was synthesized from N1-methyl AMP by the procedure of Brookes and Lawley (8). The absorption spectrum at pH 6.7 was identical within experimental error to that previously reported for N6-methyl adenosine in water, Xmax = 265 nm and Xmin = 229 nm (9). Paper chromatography of N6-methyl-AMP in solvent Systems A and in 1% ammonium 1314 5'-Adenylic Acid Aminohydrolase Vol. 246, No. 5 Deamination of AMP analogues was examined by observing the absorption spectrum of each before and after the addition of 10 to 40 pg of enzyme to both sample (Assay 1 containing approximately 60 PM analogue) and reference cells of a Beckman  were not hydrolyzed at a rate greater than 0.05 optical density per min per mg were not considered substrates.
Methylamine from Ne-methyl-AMP deamination was detected with an F and M 402 gas chromatograph equipped with a flame detector.
The column, 3 mm x 6 feet, packed with Chromosorb 103, 100 to 120 mesh (Anspec Company, Ann Arbor, Michigan), was conditioned overnight at 250" with Nz gas flow. The samples and methylamine standard (1.0 pg per ~1) were run at 105" with Nz as the carrier gas.

This preparation
of AMP-aminohydrolase catalyzed the deamination of several of the 24 analogues of AMP examined (Table I). Whether or not AMP-aminohydrolase was solely responsible for the observed deaminations was examined by three approaches.
First, the products of adenosine, ADP, and adenosine phosphoramidate deamination chromatographed with the same RF values as inosine, IDP, and inosine phosphoramidate when examined in three solvent systems (Table II). Second, the ratio of specific activities for AMP and ADP deamination was constant within experimental error throughout a single activity peak eluted from a cellulose phosphate column during enzyme preparation (Fig. 1). Third, the first order rate constant for inactivation of the enzyme as measured by the loss of AMP, ADP, and adenosine deaminating activities were essentially identical under two different conditions of heat inactivation (Table III).
Examination of Table I shows that in addition to the 6-amino group, the 6-methylamino, 6-ethylamino, and 6-chloro substituent (10) of the purine ribonucleotide were susceptible to hydrolysis.
AMP, however, was the preferred substrate, having On the other hand, in the absence of K+ (Assay 2) 0.165 mM 6-mercaptopurine 5'-ribonucleotide activated AMP deamination 32-fold. With respect to the 5' substituent, a free 5'-hydroxyl (adenosine) or second phosphoryl group (ADP) reduced the maximum rate of deamination to 1% or less of that observed for AMP at pH 6.3 to 6.5; only adenosine showed a significant change in K,, being 50 times greater than the K, for AMP while the K, for ADP at pH 6.3 was only increased by a factor of 2. At the pH optimum 5.2 for ADP, the K, was increased 20 times.
Adenosine phosphoramidate, which was deaminated at 73% of the rate of AMP, had a K, over 2 orders of magnitude greater than that of AMP. Interestingly, substitution of a methylene group for the oxygen of the phosphodiester linkage of ADP did not prevent deamination although it did prevent its role as an activator, i.e. Q!, ,&methylene-ADP could not substitute for ADP as an activator even at 1 x 10e4 M at which ADP activation is essentially maximal (11). ATP with three phosphoryl groups was not a substrate.
Substituting the sugar, psicose, for ribose completely eliminated substrate activity. , The purinc portion of the molecule with substituents at positions other than position 6 n'ere generally not substrates. Nl-Methyl AMP was neither a substrate nor an effective inhibitor at pH 6.3 (less than 15y0 inhibition was observed at 0.18 mix N'mcthgl AMP, Assay 1 with 32 ELM iz%IP). GMP was not deaminated although GDP and GTP inhibited the I(+-and ADP-activated systems (3). Inhibition of the calf brain and rat muscle enzyme by GTP have previously been reported (12, 13). Tubercidin 5'-monophosphate ( Fig. 21) and toyocamycin 5'-monophosphate (Fig. 211) were ncithcr substrates nor effcctire inhibitors Tvhile formycin 5'.monophosphate (Fig. 2111) was deaminai,cd, indicating that position 7 of the purine ring may be important for binding. 3-fl-n-Ribofuranosgladenine 5'.phosphntc ~vas not denminated, but rather was an effective inhibitor.
The replot of the slopes from the Lineweaver-Burk plot which intersected at the same V,n,, (data not shown) versus 3~p~n-J,ibofl;llanosyl:Ldeniae Zphosphate conccn-t&ion according to the iiicthod of Clcland (14) confirm linear competitive inhibition by 3-P-o-J,ibofui,anosyladcnille 5'.phosphatc with Kc = 32 pnr (Fig. 3). Setlow and Lowcnstein (15) reported an apparent Ki = 60 pnr for the competitive inhibition by 3-P-o-ribofuranosyladcnine 5'.phosphate of the calf brain enzyme. Lee (2) demonstrated that the products of AMP deamination were I;\II' and NH3. However, examination of the adenine moiety in light of the ring opening Dimroth rearrangement suggests that either N1 or N" of the purinc ring might be lost during the desminntion.
This suggestion could be tested since for the deamination of N%ncthyl Xl\IP 10~5 of N' would have resulted in Nl-methyl-IMP Tvhile loss of N6 would yield IMP. The single purine product of NQnethyl AMP deamination, observed by paper chromatography (Table II), exhibited spectral characteristics corresponding to ILII and not N1-methyl-IMP (Table IV).
Quantitative analysis of the purine product shorved greater than 98y0 conversion of NQnethyl-AMP to IMP (Table IV).
The amine product of the reaction had the same retention time (1.37 min) as commercial methylamine (I.35 min) upon gas chromatogrnphic analysis.
Comparat'ive measurements of the elutcd peak areas of the standard, methylamine, and the unknown shovxd greater than 957, recove'y of the amine product as methylaminc; liberation of ammonia was not examined. UISCUSSJON The broad substrate specificity observed for this preparation as noted in Table I might be interpreted in terms of cont~nminating enzymes such as (a) a second purine aminohydrolnse or (b) an enzyme such as a kinase or plrosphat~ase catalyzing t,he conversion of substrate under study to an alternate compound susceptible to attack by the AMP arninohydrolase. FIOKRVCJ-, neither of these interpretations are convincing in light of the heat innctivat'ion data or co&ant ratio of specific a.ct,ivities for AhIP and ADP dcamination throughout the nctivit,y peak obtained from the cellulose phosphate column during enzyme preparation.
Furthermore, prior conversion to an alternate compound was not observed for adenosine, A4DP, and adcnosine phosphor~amidate dcarnination since the corresponding inosine derivative was the only observed product in each case (Table II).
The different pH optima for adenosine and ilDP deamination (Table I), which may reflect the total charge of the substrate, have previously been observed mith a nonspecific adenine nucleotide aminohydrolnse from Porphyra crispatn, a red marine alga (16). As with the nonspecific algal enzyme, the pH optimum was lowest for ADP, intermediate for AMP, and highest for adcnosine.
Cornparidon of the maximum rates of deamination, tile R, values and the pH optima for adenosine, ~hMP, ADP, adenosine monosulfate, and adenosinc phosphoramidnte (Table I) and also the lack of substrate activity with AT1 would seem to emphasize the critical natme of the total negative charge of the substrate, or more specifically, at the position 5'. Wolfenden,Sharpless,and Bllan (17) have observed with the nonspecific Taka-diastase adenosine deaminase that nucleotides bearing monosubstituted phosphates appeared to bind only in the form in which the phosphate residue bears a single negative charge. With the muscle enzyme, a, more rigorous analysis of variations in f<, and V,,, values for these substrates as a function of p1-E is necessary before the nature of the influence of the negative charge on deamination can be evaluated. Steric factors, of course, cannot be entirely disregarded.
In summary, alterations in the purine, ribose, and phosphate moieties of AMP affect catalysis: analogues with substituents at the l-, 2-, 7-, and 3'-position of AMP were not substrates for the muscle enzyme whereas AMP with substitution or substituents at positions 6, 5', and occasionally 2' was deaminated? * The suggestion by a reviewer that the effective inhibitor, 3-pn-ribofuranosyladenine 5'-phosphate, be considered a transition state analogue (21) is worthy of consideration. tion of a Schiff base. The formation of a purinyl enzyme intermediate or a nucleophilic attack with formation of a tetrahedral ntermediate previously suggested for the calf and Taka-diastase adenosine aminohydrolase (23,29) cannot be excluded with the present data.3 It should be noted, however, that the adenosine aminohydrolase contains no reported cofactors, whereas the AMP aminohydrolase is a zinc metallo enzyme (31). This essential cation would presumably be involved in the catalytic event and as such assist in a nucleophilic attack by Hz0 at the position 6. AI.
Acknowledgments-We are grateful to Mr. Joseph Abbate for assisting in the gas chromatographic analysis.
3 Inhibition studies utilizing stable transition state analogues recentlv reuorted bv Evans and Wolfenden (30) lend further SUDport to"a nucleophiiic substitution reaction by'water in reactions catalyzed by the calf duodenum and Aspergillus oryzae adenosine