Reaction versus subsite stereospecificity of peptidylglycine alpha-monooxygenase and peptidylamidoglycolate lyase, the two enzymes involved in peptide amidation.

Carboxyl-terminal amidation, a required post-translational modification for the bioactivation of many neuropeptides, entails sequential enzymatic action by peptidylglycine alpha-monooxygenase (PAM, EC 1.14.17.3) and peptidylamidoglycolate lyase (PGL, EC 4.3.2.5). The monooxygenase, PAM, first catalyzes conversion of a glycine-extended pro-peptide to the corresponding alpha-hydroxyglycine derivative, and the lyase, PGL, then catalyzes breakdown of this alpha-hydroxyglycine derivative to the amidated peptide plus glyoxylate. We have previously established that PAM and PGL exhibit tandem reaction stereospecificities, with PAM producing, and PGL being reactive toward, only alpha-hydroxyglycine derivatives of absolute configuration (S). We now demonstrate that PAM and PGL exhibit dramatically different subsite stereospecificities toward the residue at the penultimate position (the P2 residue) in both substrates and inhibitors. Incubation of Ac-L-Phe-Gly, Ac-L-Phe-L-Phe-Gly, or (S)-O-Ac-mandelyl-Gly with PAM results in complete conversion of these substrates to the corresponding alpha-hydroxylated products, whereas the corresponding X-D-Phe-Gly compounds undergo conversions of < 1%. The KI of Ac-D-Phe-Gly is at least 700-fold higher than that of Ac-L-Phe-Gly, and the same pattern holds for other substrate stereoisomers. This S2 subsite stereospecificity of PAM also holds for competitive inhibitors; thus, the KI of 45 microM for Ac-L-Phe-OCH2CO2H increases to 2,247 microM for the -D-Phe- enantiomer. In contrast, incubation of PGL with Ac-L-Phe-alpha-hydroxy-Gly, Ac-D-Phe-alpha-hydroxy-Gly, (S)-O-Ac-mandelyl-alpha-hydroxy-Gly, or (R)-O-Ac-mandelyl-alpha-hydroxy-Gly results in facile enzymatic conversion of each of these compounds to their corresponding amide products. The simultaneous expression of high reaction stereospecificity and low S2 subsite stereospecificity in the course of PGL catalysis was illustrated by a series of experiments in which enzymatic conversion of the diastereomers of Ac-L-Phe-alpha-hydroxy-Gly and Ac-D-Phe-alpha-hydroxy-Gly was monitored directly by HPLC. Kinetic parameters were determined for both substrates and potent competitive inhibitors of PGL, and the results confirm that, in sharp contrast to PAM, the configuration of the chiral moiety at the P2 position has virtually no effect on binding or catalysis. These results illustrate a case where catalytic domains, which must function sequentially (and with tandem reaction stereochemistry) in a given metabolic process, nevertheless exhibit sharply contrasting subsite stereospecificities toward the binding of substrates and inhibitors.

Elucidation of the stereochemistry of carboxyl-terminal amidation is a critical issue for detailed mechanistic studies on the enzymology of this process and for the rational design of pseudosubstrates and inhibitors targeted at PAM and PGL. In this regard, we recently demonstrated that PAM and PGL exhibit tandem reaction stereospecificities in carrying out the two requisite steps in carboxyl-terminal amidation (29). Thus, PAM produces only ␣-hydroxyglycine derivatives of absolute configuration (S), and PGL is reactive only toward (S)-␣-hydroxyglycines (Scheme 1). While these results elucidate the reaction stereospecificity of both PAM and PGL toward the COOH-terminal (S)-␣-hydroxyglycine moiety, the quite distinct question of whether PAM and PGL differ with respect to subsite stereospecificity remains unresolved.
We report here a series of experiments, which demonstrate that PAM and PGL indeed exhibit dramatically different subsite stereospecificities toward the residue at the penultimate position (the P 2 residue) 2 of their respective substrates. In addition, we introduce new competitive inhibitors for PAM and PGL, and we demonstrate that the distinctive difference in S 2 subsite 2 stereospecificities for substrates is also reflected in the binding of competitive inhibitors targeted at either enzyme, respectively. Finally, we illustrate the simultaneous expression of high reaction stereospecificity and low S 2 subsite stereospecificity in the course of PGL catalysis by a series of experiments in which enzymatic conversion of all four diastereomers of Ac-Phe-␣-hydroxy-Gly was monitored directly by HPLC. Thus, these results provide a clear example of a case where catalytic domains, which must function sequentially (and with tandem reaction stereochemistry) in a given metabolic process, nevertheless exhibit contrasting subsite stereospecificities toward the binding of both substrates and inhibitors.

Enzyme Purification and Assays
PAM and PGL were isolated from bovine pituitaries as described previously (9). PGL assays were performed in 100 mM sodium 2-[Nmorpholino]ethanesulfonate, pH 6.5, containing 0.2 mg/ml bovine serum albumin, at 37°C. PGL inhibition was established using TNP-D-Tyr-Val-␣-hydroxy-Gly as the PGL substrate. PAM assays were performed as described previously (11) with the addition of 0.2 mg/ml bovine serum albumin in assay mixtures. PAM inhibition by O-glycolate ester derivatives was established using the continuous assay developed in our laboratory (31). Typically, PAM was incubated with the O-glycolate ester derivatives, in the presence of Ac-D-Tyr-Val-Gly (PAM substrate). Values for V max and K m were obtained from the inverse plots by using a least-squares fit program; standard errors of 10% or less were calculated for the kinetic parameters listed in the tables.

Ac-L-Phe-L-Phe-Gly and Ac-L-Phe-D-Phe-Gly
These compounds were synthesized on a Rainin PS3 automated solid phase peptide synthesizer using the standard Fmoc protocol. All procedures were done according to the instructions in the manual. The time for the coupling step was 50 min, and acetylation was carried out after the synthesis of the tripeptide-resin. Each of the tripeptide products exhibited a single peak on a C8 RP-HPLC with a gradient from 15% CH 3 CN, 85% H 2 O, 0.2% phosphoric acid to 75% CH 3 CN, 25% H 2 O, 0.2% phosphoric acid over 60 min.

Ac-D-Phe-Gly
Method I-A mixture of N-(tert-butoxycarbonyl)-D-phenylalanine (2.6 g), glycine ethyl ester hydrochloride (1.4 g), triethylamine (1.0 g), 1-hydroxybenzotriazole hydrate (1.5 g), and dimethylformamide (15 ml) were cooled in an ice bath, and 1,3-dicyclohexylcarbodiimide (2.3 g) was added. The mixture was stirred overnight at 4°C, and the dimethylformamide was evaporated under vacuum. To the solid, 30 ml of ethyl acetate was added and the remaining solid was removed by filtration. The ethyl acetate layer was washed sequentially with 10% citric acid, saturated sodium bicarbonate water, and saturated sodium chloride water (3 ϫ 30 ml each), and dried over MgSO 4 . Evaporation of the ethyl acetate afforded a white solid product, Boc-D-Phe-Gly-OEt (2.4 g). Cleavage of the Boc group by trifluoroacetic acid (5 times molar excess) gave D-Phe-Gly-OEt. Ac-D-Phe-Gly-OEt (0.8 g) was obtained by acetylation of D-Phe-Gly-OEt (1.0 g) in a mixture of 10 ml of pyridine and 4 ml of acetic anhydride for 30 min at room temperature.
The crude Ac-D-Phe-Gly-OEt (0.5 g, 1.7 mmol) was dissolved in a mixture of methanol and water (20 ml, 1:1), and 3.8 ml of 0.5 M NaOH (1.9 mmol) was then added. The mixture was stirred at room temperature, and the hydrolysis of Ac-D-Phe-Gly-OEt was monitored on a C8 PR-HPLC column with an isocratic solution of H 2 O/CH 3 CN (80/20, 0.2% phosphoric acid). Nearly complete hydrolysis was achieved in about 2 h, after which the pH of the solution was adjusted to neutrality with dilute HCl. The mixture was extracted with methylene chloride (2 ϫ 30 ml) to remove any unreacted Ac-D-Phe-Gly-OEt. The pH of the solution pH was then adjusted to about 3.0 and extracted with ethyl acetate (2 ϫ 30 ml). The ethyl acetate layer was dried over MgSO 4 and evaporated to dryness to give Ac-D-Phe-Gly (0.3 g). 1  Method II-This compound was synthesized using the solid-phase method, with the procedure being analogous to the synthesis of Ac-L-Phe-L-Phe-Gly.
Synthetic Ac-D-Phe-Gly has a specific rotation value of Ϫ11.8°in methanol.

Ac-L-Phe-Gly
This compound was synthesized by the same procedures used for the synthesis of Ac-D-Phe-Gly.
Synthetic Ac-L-Phe-Gly has a specific rotation value of 11.9°in methanol.

(R)-O-Ac-mandelyl-Gly and (S)-O-Ac-mandelyl-Gly
Both (R)-and (S)-O-Ac-mandelyl-Gly tert-butyl esters were synthesized by the synthetic procedure used for the synthesis of Boc-D-Phe-Gly ethyl ester. Cleavage of the tert-butyl group by trifluoroacetic acid (5 ϫ molar excess) gave the desired products.

Ac-D-Phe-amide, (S)-O-Ac-mandelic amide, and (R)-O-Ac-mandelic amide
These compounds were synthesized according to a published procedure (32). The amide products exhibit the following specific rotation in methanol

(R)-2-Phenylbutyl-␣-hydroxy-Gly
(R)-2-Phenylbutyric acid (1.0 g, 6.09 mmol) was converted to its acid chloride by reaction with cyanuric chloride, according to a published procedure (33). The acid chloride was then dissolved in anhydrous ether (20 ml) and treated with anhydrous ammonia gas to form the corresponding amide. The insoluble ammonium chloride salt was filtered off. The filtrate was washed twice with aqueous sodium bicarbonate solution, dried over MgSO 4 , concentrated under reduced pressure, and recrystallized from ether/hexane. 1 This compound (0.20 g) was synthesized as described for (R)-2-Phenylbutyl-␣-hydroxy-Gly, starting from (S)-2-phenylbutyric acid (1.0 g). To a solution of bromoacetic acid (4.5 g, 32 mmol) in ethyl acetate (500 ml) were added phenacyl bromide (33 g, 0.17 mol) and triethyl amine (4.5 ml, 32 mmol) (35). The reaction mixture was stirred overnight at room temperature and filtered to remove the insoluble triethylamine hydrobromide salt. The filtrate was washed with saturated NaHCO 3 solution until no more bromoacetic acid showed on TLC (eluting solvent: EtOAc), then washed with water twice. The organic layer was dried over NaSO 4 and concentrated under reduced pressure. Excess phenacyl bromide was removed by distillation under vacuum and the residue recrystallized from EtOAc/hexane, to give pure 2-bromophenacyl acetate (4.74 g). Deprotection of the terminal carboxylic acid was achieved by stirring the ester with zinc dust in glacial acetic acid (20 ml) at room temperature, for 3 h (37). The zinc dust was filtered off. The filtrate was concentrated under reduced pressure and dried under vacuum overnight to remove all traces of acetic acid. The residue was then dissolved in water (50 ml) and washed twice with ethyl acetate (50 ml). The aqueous layer was then concentrated, dried under vacuum over P 2 O 5 , and the residue recrystallized from ethanol/hexane, giving N-Ac-Gly acetic acid ester (0.56 g). This compound was synthesized as described for Ac-Gly-OCH 2 CO 2 H, except for the final purification step, starting from Ac-L-Phe (1.0 g). After evaporation of acetic acid, the residue was dissolved in water, washed once with EtOAc/hexane 2:8, then with hexane. The aqueous layer was lyophilized to give Ac-L-Phe-OCH 2

RESULTS AND DISCUSSION
Subsite Stereospecificity of PAM toward Peptide Substrates-While in early work using crude preparations of "amidating enzyme," Bradbury et al. (38,39) observed that a tripeptide possessing a D-Ala residue at the penultimate position did not undergo amidation; it is unknown whether this reflects the stereospecificity of PAM, of PGL, or of both enzymes. We therefore began by determining the subsite stereospecificity of PAM. Glycine-extended peptides possessing residues of either the L or D configuration at their respective penultimate (P 2 ) positions were synthesized and tested as substrates for purified PAM ( Table I). Incubation of Ac-L-Phe-Gly, Ac-L-Phe-L-Phe-Gly, (S)-O-Ac-mandelyl-Gly, or TNP-D-Tyr-L-Val-Gly (the latter being our standard PAM substrate; Ref. 40) with PAM resulted in progressive conversion of all of these compounds to their respective ␣-hydroxylated products. The time course of the PAMcatalyzed conversion of Ac-L-Phe-Gly to Ac-L-Phe-␣-hydroxy-Gly is illustrated in Fig. 1. For all four compounds, nearly 100% conversion to the ␣-hydroxylated product was observed after a reaction time of 2 h or less.
In contrast, as shown in Fig. 1, when Ac-D-Phe-Gly is incubated with a large quantity of PAM, product formation is barely detectable. Examination of HPLC traces at high sensitivity revealed appearance of a very small amount, i.e. at most, 0.5% conversion, of product with retention time corresponding to that of authentic Ac-L-Phe-␣-hydroxy-Gly within the first few minutes of incubation, with no further conversion being detectable even after 20 h. Similarly, incubation of PAM with either Ac-L-Phe-D-Phe-Gly or (R)-O-Ac-mandelyl-Gly results in conversions of less than 0.2% and 0.9%, respectively. Control experiments confirmed that PAM remains fully active after incubation with these substrates.
In order to demonstrate directly that this S 2 subsite stereospecificity affects substrate binding, apparent K I values Ac-L-Phe-Gly 7.9 5.6 Ac-L-Phe-L-Phe-Gly 1. were determined for both Ac-L-Phe-Gly and its non-reactive -D-Pheenantiomer using our standard assay substrate, TNP-D-Tyr-L-Val-Gly. As is evident from the data in Table I, there is a dramatic 700-fold difference between the apparent K I values for the two Ac-Phe-Gly enantiomers. Moreover, the true K I of Ac-L-Phe-Gly is actually even lower than the value of 2 M listed in the table, since some reaction of Ac-L-Phe-Gly itself occurs during the initial rate measurements. Although reaction of Ac-Phe-Gly is transparent to our assay, it causes the measured K I to be higher than the true value. The K I of 580 M for Ac-L-Phe-D-Phe-Gly is 480-fold higher than the K m for its L-Phe-L-Phe-diastereomer, and the same pattern is true for (R)and (S)-O-Ac-mandelyl-Gly. Thus, it is clearly evident from these results that substrate reactivity toward PAM is virtually abolished by the presence of a D-amino acid residue, or of the stereotopically equivalent (R)-mandelyl residue, at the P 2 position.
Subsite Stereospecificity of PAM toward Inhibitors-To determine whether the S 2 subsite stereospecificity observed with PAM substrates is also reflected in the potency of competitive inhibitors, O-glycolate ester derivatives of N-acetylated amino acids of the L or D configuration were synthesized. Initial experiments with Ac-L-Phe-OCH 2 CO 2 H confirmed that this ester is a potent inhibitor, and not a substrate, for PAM. Kinetic experiments were then carried out with the esters listed in Table II. As is evident from the K I values in the table, inhibitor potency is markedly reduced when the configuration of the amino acid residue at the P 2 position is inverted. Thus, the K I of 45 M for Ac-L-Phe-OCH 2 CO 2 H increases to 2,247 M for the -D-Pheenantiomer; the corresponding values for Ac-L-Leu-OCH 2 CO 2 H and its -D-Leu-enantiomer are 60 M and 2,115 M, respectively. Indeed, Ac-L-Phe-OCH 2 CO 2 H and Ac-L-Leu-OCH 2 CO 2 H are the most potent PAM competitive inhibitors yet reported; inhibition is purely competitive, as evidenced from Dixon plots. Taken together, these results demonstrate the high S 2 subsite stereospecificity of PAM toward both substrates and inhibitors.
Subsite Stereospecificity of PGL toward Peptide Substrates-Incubation of synthetic Ac-L-Phe-␣-hydroxy-Gly, Ac-D-Phe-␣hydroxy-Gly, (S)-O-Ac-mandelyl-␣-hydroxy-Gly, or (R)-O-Acmandelyl-␣-hydroxy-Gly with PGL resulted in facile enzymatic conversion of each of these compounds to their corresponding amide products. The time courses of the PGL-catalyzed conversion of Ac-D-Phe-␣-hydroxy-Gly to Ac-D-Phe-NH 2 and of (R)-O-Ac-mandelyl-␣-hydroxy-Gly to (R)-O-Ac-mandelyl-NH 2 are shown in Fig. 2. It is evident from these time courses that enzymatic turnover terminates prior to complete conversion of the substrates to amide products by PGL. Therefore, prolonged incubation experiments with PGL were carried out to quantitate the maximal percentage conversion of Ac-L-Phe-␣-hydroxy-Gly, Ac-D-Phe-␣-hydroxy-Gly, (S)-O-Ac-mandelyl-␣-hydroxy-Gly, and (R)-O-Ac-mandelyl-␣-hydroxy-Gly to their corresponding amide products. As shown in Table III, maximal conversion in all cases is about 50%. To further confirm this result, a double addition experiment was carried out wherein Ac-L-Phe-␣-hydroxy-Gly and Ac-D-Phe-␣-hydroxy-Gly were each first incubated overnight with PGL then analyzed by HPLC to confirm 50% conversion, after which the enzyme was removed by ultrafiltration. Each filtrate was then again incubated with a large quantity of PGL, and monitored by HPLC. As expected, no additional enzymatic conversion of unreacted substrate to amide products occurred for either substrate.

FIG. 2. The time courses of PGL-catalyzed conversion of Ac-D-Phe-␣-hydroxy-Gly and (R)-O-Ac-mandelyl-␣-hydroxy-Gly to their amide products. Ac-D-Phe-␣-hydroxy-Gly (200 M) (E) and (R)
O-Ac-mandelyl-␣-hydroxy-Gly (200 M) ( ) were each incubated with 0.2 g/ml purified PGL at 37°C, and the formation of the amide products were measured using C8 RP-HPLC. hydroxy-Gly and (S)-O-Ac-mandelyl-␣-hydroxy-Gly, while Ac-D-Phe-␣-hydroxy-Gly and (R)-O-Ac-mandelyl-␣-hydroxy-Gly should have been unreactive toward PGL. Therefore, our finding that 50% conversion actually occurs with all these substrates (Table III) is a strong indication that, unlike PAM, PGL reacts with substrates possessing residues of either the (S) or (R) configuration at the P 2 position. In order to demonstrate directly this simultaneous expression of high reaction stereospecificity and low S 2 subsite stereospecificity in the course of PGL catalysis, the series of HPLC experiments illustrated in Fig. 3 was carried out. Panel A shows the HPLC elution profile of Ac-L-Phe-␣-hydroxy-Gly produced from Ac-L-Phe-Gly by PAM. Since PAM catalysis produces only ␣-hydroxyglycines of the (S) configuration, this peak represents the retention time of the (S,S) diastereomer of Ac-L-Phe-␣-hydroxy-Gly. Panels B and C show the elution profiles of synthetic Ac-L-Phe-␣-hydroxy-Gly and Ac-D-Phe-␣-hydroxy-Gly, respectively. It is evident from panels B and C that the two diastereomers of synthetic Ac-L-Phe-␣-hydroxy-Gly (with the configurations (S,R) and (S,S)) and the two diastereomers of synthetic Ac-D-Phe-␣-hydroxy-Gly (with the configurations (R,S) and (R,R)) are well separated on a C8 RP-HPLC column using a mobile phase of 2% CH 3 CN, 98% H 2 O, 0.2% H 3 PO 4 . Panels D and E show the elution profiles obtained after reaction of Ac-L-Phe-␣-hydroxy-Gly and Ac-D-Phe-␣-hydroxy-Gly, respectively, with PGL. It is quite clear from panel D that reaction of PGL with synthetic Ac-L-Phe-␣-hydroxy-Gly results in conversion of only one substrate diastereomer to Ac-L-Phe-NH 2 . Since the reactive diastereomer has retention time identical to that of Ac-L-Phe-␣-hydroxy-Gly (with (S,S) configuration) produced by PAM (Fig. 3, panel A), it is clear that this reactive diastereomer has the (S,S) configuration. Correspondingly, as shown in panel E, reaction of PGL with synthetic Ac-D-Phe-␣-hydroxy-Gly also results in conversion of only one diastereomer of this substrate to Ac-D-Phe-NH 2 . From a comparison with the previous panels, it is quite clear that the reactive diastereomer of this substrate has the (R,S) configuration. Taken together, these experiments unequivocally confirm the sharp contrast between the reaction and the subsite stereospecificities of PGL; while PGL is reactive only toward an ␣-hydroxyglycine moiety of the (S) configuration, this lyase reacts readily with substrate diastereomers of either the (S) or (R) configuration at the P 2 position.
Kinetic parameters for the two pairs of diastereomeric PGL substrates are listed in Table IV. It is apparent that the V max values are virtually unaffected by whether the residue at the P 2 position has the (S) or (R) configuration, whereas small effects are apparent on the K m values. Both pairs of compounds are comparable in reactivity to N-benzoyl-␣-hydroxy-Gly.
Subsite Stereospecificity of PGL toward Inhibitors-A systematic specificity study revealed that the substrate analog, phenylacetyl-␣-OH-Gly, is an inhibitor of PGL with a K I value of 770 M; inhibition is purely competitive as evidenced by Dixon plots. Therefore, in order to determine the effect of stereochemical configuration at the P 2 position on the potency of PGL inhibitors, we synthesized (R)-2-phenylbutyl-␣-hy-droxy-Gly and (S)-2-phenylbutyl-␣-hydroxy-Gly, analogs of phenylacetyl-␣-OH-Gly each possessing a chiral center of opposite configuration at the penultimate position. As is the case for the PGL substrates discussed above, synthetic (R)-2-phenylbutyl-␣-hydroxy-Gly actually consists of an equimolar mixture of the (R,S) and (R,R) diastereomers; likewise, synthetic (S)-2-phenylbutyl-␣-hydroxy-Gly is an equimolar mixture of (S,S) and (S,R) diastereomers.
Both (R)-2-phenylbutyl-␣-hydroxy-Gly and (S)-2-phenylbutyl-␣-hydroxy-Gly were found to indeed be PGL inhibitors. We also observed that they both undergo very slow enzymatic FIG. 3. Simultaneous expression of high reaction stereospecificity and low S 2 subsite stereospecificity in PGL catalysis. The two diasteromers of synthetic Ac-L-Phe-␣-hydroxy-Gly and the two diasteromers of synthetic Ac-D-Phe-␣-hydroxy-Gly were separated on a C8 RP-HPLC with a mobile phase of 2% CH 3 CN, 98% H 2 O, 0.2% H 3 PO 4 . Panel A, HPLC chromatogram of Ac-L-Phe-␣-hydroxy-Gly produced by PAM. Since PAM only produces the peptidyl-␣-hydroxyGlycine derivatives with the S configuration at the ␣-hydroxyglycine moiety, the peak at 46.5 min has the configuration (S,S). The small peak at 51 min is Ac-L-Phe-NH 2 . Panel B, HPLC chromatogram of synthetic Ac-L-Phe-␣hydroxy-Gly. By comparison with panel A, it is evident that the two diastereomers with retention times of 42 and 46.5 min, respectively, have the configurations (S,R) and (S,S). Panel C, HPLC chromatogram of synthetic Ac-D-Phe-␣-hydroxy-Gly. By comparison with panel A, it is evident that the two diastereomers with retention times of 42 and 46.5 min, respectively, have the configurations (R,S) and (R,R). The small peak at 51 min is Ac-D-Phe-NH 2 . Panel D, HPLC chromatogram of the sample obtained after PGL-catalyzed conversion of Ac-L-Phe-␣-hydroxy-Gly to Ac-L-Phe-NH 2 . This panel demonstrates that only the diastereomer of the (S,S) configuration was converted to Ac-L-Phe-NH 2 . Panel E, HPLC chromatogram of the sample obtained after PGL-catalyzed conversion of Ac-D-Phe-␣-hydroxy-Gly to Ac-D-Phe-NH 2 . This panel demonstrates that only the diastereomer of the (R,S) configuration was converted to Ac-D-Phe-NH 2 .
conversion to the corresponding phenylbutyramides (as confirmed by HPLC analysis). The substrate activity (V/K) of these compounds is ϳ10,000 times slower than that of our normal PGL assay substrate, TNP-D-Tyr-L-Val-␣-hydroxy-Gly; therefore, inhibition kinetics are readily performed in the usual manner. Both compounds are highly potent competitive inhibitors, and the K I values obtained are 280 M and 310 M for (R)-2-phenylbutyl-␣-OH-Gly and (S)-2-phenylbutyl-␣-OH-Gly, respectively (Table V). It is thus clear that, in sharp contrast to PAM, the configuration of the chiral moiety at the P 2 position has no effect on the potency of these PGL inhibitors.
Table V also lists the kinetic constants measured for the very slow PGL-catalyzed reaction of these two compounds. As expected, V max and K m are identical for (R)-and (S)-2-phenylbutyl-␣-OH-Gly. Moreover, the measured K m values are identical to the K I values obtained in the inhibition experiments against TNP-D-Tyr-L-Val-␣-hydroxy-Gly, thus confirming that the K m values represent true binding constants.
Taken together, the results reported here establish that despite their tandem reaction stereospecificities with respect to the COOH-terminal (S)-␣-hydroxyglycine moiety, PAM and PGL exhibit sharply contrasting S 2 subsite stereospecificities in the binding of both substrates and inhibitors. Physiologically, these two enzymatic steps must occur sequentially in order to convert a given glycine-extended pro-peptide to the corresponding mature amidated peptide (9,11). Thus, from a metabolic viewpoint, the stereospecificities of these catalytic domains would have been expected to correlate. From a mechanistic perspective, there is, of course, a vast difference between the catalytic site of a metallo-monooxygenase on the one hand, which entails participation of both active-site copper and an electron donor in catalysis, and that of a lyase on the other. Indeed, the contrasting stereospecificities we report here are likely reflections of corresponding differences in the active site topographies of PAM and PGL.
To our knowledge, this is the first demonstration of a case where catalytic domains that must function sequentially nevertheless exhibit contrasting binding stereospecificities. Most likely, the possibility that stereospecificities of such domains might differ has not been addressed since, of course, metabolites flow through the catalytic domains in sequence; a given domain would never "see" a metabolite of altered chirality such that it would have been precluded from binding at the preceding domain. Thus, inspection of the stereochemical configurations of metabolic intermediates does not reveal possible differences in the subsite stereospecificities of the domains; stereochemical studies using substrate analogs and inhibitors of deliberately altered chirality are required for this purpose. For example, in the case of chorismate mutase-prephenate dehydrogenase, a multifunctional protein catalyzing sequential reactions that has been the subject of much recent interest (41)(42)(43), the relative stereochemistry at the ring position para to the site of enzymatic reaction clearly affects reactivity of substrate analogs in the dehydrogenase reaction. However, a comparison of the effect of commensurate changes in subsite stereochemistry on binding and turnover for the individual catalytic domains has not been carried out, and, unfortunately, synthetic approaches to the synthesis of individual substrate analog stereoisomers that might be used in such investigations have been unsuccessful (41)(42)(43). N-Benzoyl-␣-hydroxy-Gly 0.46 42 Ac-L-Phe-␣-hydroxy-Gly 0.23 46 Ac-D-Phe-␣-hydroxy-Gly 0.50 43 (S)-O-Ac-mandelyl-␣-hydroxy-Gly 0.56 89 (R)-O-Ac-mandelyl-␣-hydroxy-Gly 2.6 99 (R)-2-Phenylbutyl-␣-hydroxy-Gly 0.28 0.30 0.28 (S)-2-Phenylbutyl-␣-hydroxy-Gly 0.31 0.33 0.38 Phenylacetyl-␣-hydroxy-Gly 0.77 TNP-D-Tyr-L-Val-␣-hydroxy-Gly 2.7 ϫ 10 Ϫ2 2.7 ϫ 10 2