Specificity of Pyridinium Inhibitors of the Ubiquinone Reduction Sites in Mitochondrial Complex I*

Dual binding sites for pyridinium-type inhibitors in bovine heart mitochondrial complex I have been proposed (Gluck, M. R., Krueger, M. J., Ramsay, R. R., Sablin, S. O., Singer, T. P., and Nicklas, W. J. (1994) J. Biol. Chem. 269, 3167–3174). The marked biphasic nature of the dose-response curve for inhibition of the enzyme by MP-6(N-methyl-4-[2-(p-tert-butylbenzyl)propyl]pyridinium) makes this compound the first selective inhibitor of the two sites (Miyoshi, H., Inoue, M., Okamoto, S., Ohshima, M., Sakamoto, K., and Iwamura, H. (1997) J. Biol. Chem.272, 16176–16183). Modifications of the structure of MP-6 show that a tert-butyl group on the benzene ring, a methyl group attached to the pyridine nitrogen atom, para-substitution pattern in the pyridine ring, and the presence of a branched structure in the spacer moiety are important for the selective inhibition. On the basis of the structural specificity, we synthesized a selective inhibitor, MP-24 (N-methyl-4-[2-methyl-2-(p-tert-butylbenzyl)propyl]pyridinium), which elicits greater selectivity. Characterization of the inhibitory behavior of MP-24 provided further strong evidence for the dual binding sites model.

Mitochondrial NADH-ubiquinone oxidoreductase (complex I) 1 is a large enzyme that catalyzes the oxidation of NADH by ubiquinone coupled to proton translocation across the inner membrane (1,2). There are a variety of inhibitors of mitochondrial complex I and with the exception of a few inhibitors which inhibit electron input into the enzyme (3,4), all inhibitors act at or close to the ubiquinone reduction site (5). Among the inhibitors, positively charged neurotoxic N-methyl-4-phenylpyridinium (MPP ϩ ) and its alkyl analogues exhibit unique inhibitory behavior with bovine heart mitochondrial complex I (6). A series of studies of the inhibition mechanism of MPP ϩ analogues by Singer and colleagues (6 -10) have suggested that MPP ϩ analogues are bound at two sites in the enzyme, one accessible to relatively hydrophilic inhibitors (termed the "hydrophilic site") and one shielded by a hydrophobic barrier on the enzyme (the "hydrophobic site"), and that occupation of both sites is required for complete inhibition. This concept may be helpful in elucidating the terminal electron transfer step in complex I and seems to be consistent with the existence of two EPR-detectable species of complex I-associated ubisemiquinones (11,12). Some experimental results with ordinary complex I inhibitors (13)(14)(15)(16) can be explained by assuming the existence of more than one inhibitor (or ubiquinone) binding site.
In the previous study (17), we synthesized a series of MPP ϩ analogues which are much more potent than the original MPP ϩ and demonstrated that the presence of hydrophobic counteranion tetraphenylboron (TPB Ϫ ) potentiates the inhibition by MPP ϩ analogues differently depending upon the molar ratio of TPB Ϫ to the inhibitors. In the presence of a catalytic amount of TPB Ϫ , the inhibitory potency of MPP ϩ analogues was markedly enhanced, and the extent of inhibition was almost complete. The presence of an excess amount of TPB Ϫ partially reactivated the enzyme activity, and the inhibition was partly saturated (ϳ50%). This complicated inhibitory behavior could be explained by the dual binding sites model mentioned above (6), which supposes quite different hydrophobic natures of the two sites and/or their environments.
If there are indeed two distinct binding sites of MPP ϩ analogues in bovine complex I, there should be specific inhibitors which act selectively at one of the two proposed binding sites since it is unlikely that the structural properties of the two sites are completely identical. We have synthesized such a selective inhibitor, MP-6 (N-methyl-4-[2-(p-tert-butylbenzyl)propyl]pyridinium, Fig. 1) (17). In the absence of TPB Ϫ , this inhibitor showed approximately 50% inhibition at 5 M in NADH-Q 1 oxidoreductase assay, but the inhibition reached a plateau at this level over a wide range of concentrations. Weak inhibition was again observed above ϳ80 M, and maximum inhibition (Ͼ90%) was obtained only when the concentration of the inhibitor was increased to ϳ250 M. Such a marked biphasic nature of the does-response curve has not been reported previously for usual complex I inhibitors. On the other hand, almost complete inhibition was readily obtained at low concentrations of MP-6 (Ͻ10 M) in the presence of 2 M TPB Ϫ . The site that is readily blocked by low concentrations of MP-6 without TPB Ϫ would be the hydrophilic binding site. Thus, MP-6 is a fairly selective inhibitor of one of the two proposed binding sites and could be a useful probe to examine the mechanism of the terminal electron transfer step of complex I. This inhibitor is, however, not completely selective in the strict sense because it elicits complete inhibition at high concentrations without TPB Ϫ , as mentioned above. Therefore, highly selective pyridinium-type inhibitors superior to MP-6 are required.
Among the 19 cationic inhibitors synthesized previously (17), only MP-6 exhibited such a unique inhibitory behavior, suggesting that some specific structural feature(s) of this compound would be responsible for its behavior. Elucidation of such structural features is essential to develop further selective inhibitors. In the present study, we systematically modified the structure of the lead compound MP-6 ( Fig. 1) and investigated the effects of structural modifications on its inhibitory behavior. On the basis of the structural information, we synthesized a selective inhibitor superior to MP-6. Characterization of the inhibitory behavior of the new inhibitor provides further evidence for the dual binding sites model.

EXPERIMENTAL PROCEDURES
Materials-Q 1 was a generous gift from Eisai Co. (Tokyo, Japan). Piericidin A was generously provided by Dr. Shigeo Yoshida (RIKEN, Japan). Other chemicals were commercial products of analytical grade.
Synthesis-All synthetic compounds were characterized by 1 H NMR spectroscopy (Bruker ARX-300) and elemental analyses for carbon, hydrogen, and nitrogen, within an error of Ϯ 0.3%. N-Methylation (or N-ethylation) of the compounds was carried out by the previously reported method (17).
Synthesis of MP-6 and Its Derivatives-The previous synthetic method for MP-6 (17) was not suitable for preparation of a variety of derivatives because of low reaction yield and limits of structural variations of starting materials. We therefore synthesized MP-6 by a new method as follows (Method 1 in Fig. 2).
To a mixture of tert-butylbenzene (16 g, 0.12 mol) and n-propionyl chloride (9.3 g, 0.1 mol), was added aluminum chloride (14.7 g, 0.11 mol) slowly at 0°C. After stirring at room temperature for 3 h, the mixture was poured into ice water and extracted with ether, washed with 1 N HCl, 1 N NaOH, and brine, dried over MgSO 4 , and evaporated. The crude product was purified by silica gel column chromatography (ethyl acetate/hexane, 1:20) to give 4-propionyl-tert-butylbenzene in a 47% yield.
To a solution of lithium diisopropylamide in THF, was added 4-propionyl-tert-butylbenzene (7.7 g, 40.5 mmol) slowly at Ϫ78°C and the reaction mixture was stirred for 1 h. After adding 4-pyridinecarboxaldehyde (4.4 g, 41 mmol) dropwise, the mixture was stirred for 1 h. The reaction was quenched with saturated aqueous NH 4 Cl, ether was added, and the organic phase was washed with brine and dried over MgSO 4 . The solvent was removed under reduced pressure and the residue was purified by silica gel column chromatography (ethyl acetate/hexane, 1:1) to give 4-[1-hydroxy-2-(p-tert-butylbenzoyl)propyl]pyridine in a 90% yield.
The mixture was evaporated and then the residue was extracted with ether, washed with brine, dried over MgSO 4 , and concentrated. The crude product was purified by silica gel column chromatography (ethyl acetate/hexane, 3:7) to give 4-[2-(p-tert-butylbenzoyl)propenyl]pyridine in a 93% yield.
To MP-11 and MP-12 were prepared by the same method as used for MP-6, except that p-n-or p-sec-butylbenzene was used, respectively, in place of p-tert-butylbenzene in step a. MP-11: 1   Synthesis of MP-24 and Its Derivatives-MP-24 was synthesized by Method 2 as shown in Fig. 2. To a suspension of NaH (2, 2 g, 91.0 mmol) in 50 ml of THF, 4-acetylpyridine (5.0 g, 41.3 mmol) was added slowly at 0°C under N 2 and the mixture was stirred for 30 min, after which methyliodide (12.8 g, 90.0 mmol) was added and stirred at room temperature overnight. The reaction mixture was extracted with ether and washed with brine. The organic phase was dried over MgSO 4 and evaporated. The crude product was purified by silica gel column chromatography (ethyl acetate/hexane, 3:7) to give 4-(2-methylpropionyl)pyridine in a 56% yield.
To a solution of 4-[2-methyl-2-(p-tert-butylbenzyl)propionyl]pyridine (0.7 g, 2.3 mmol) in 10 ml of ethylene glycol, hydrazine monohydrate (0.4 g, 7.0 mmol) and KOH (0.4 g, 7.0 mmol) were added and stirred at 150°C for 5 h. The mixture was extracted with ether and washed with brine. The organic phase was dried over MgSO 4 and evaporated. The crude product was purified by silica gel column chromatography (ethyl acetate/hexane, 3: MP-26 and MP-27 were synthesized by the same method as used for MP-24, except that p-n-butylbenzyliodide or p-tert-amylbenzyliodide were used, respectively, in place of p-tert-butylbenzylbromide in step h. MP-26: 1  Methods-Bovine heart submitochondrial particles were prepared by the method of Matsuno-Yagi and Hatefi (19) and stored in a buffer containing 0.25 M sucrose and 10 mM Tris/HCl (pH 7.4) at Ϫ78°C. NADH-Q 1 oxidoreductase activity was measured as the rate of NADH oxidation with a Shimadzu UV-3000 at 340 nm (⑀ ϭ 6.2 mM Ϫ1 cm Ϫ1 ). The reaction medium contained 50 mM potassium phosphate (pH 7.4), 0.25 M sucrose, 1 mM MgCl 2 , 2 mM KCN, 0.4 M antimycin A, and the final mitochondrial protein concentration was 30 g/ml. The reaction was started by adding 50 M NADH after submitochondrial particles were incubated with inhibitor for 4 min at 30°C. Q 1 (50 M) was used as an electron acceptor since this substrate is the best electron acceptor, yielding high reaction rate with linear kinetics (20). Unless otherwise noted, the concentration of TPB Ϫ was set at 2 M throughout the present study since the inhibition by TPB Ϫ itself was not negligible above this concentration. The experimental conditions for investigation of the effects of incubation temperature or incubation time on the inhibition by MP-24 are described in the figure legends.
Ferricyanide and dichloroindophenol reduction by complex I was assayed at 420 -500 and 600 nm, respectively, under the same experimental conditions. When ferricyanide (1 mM) or dichloroindophenol (50 M) was used as an electron acceptor, the concentration of NADH was set at 150 or 50 M, respectively.

Effects of Structural Modification on Inhibitory Behavior of MP-6 -
The I 75 /I 25 ratio was used as an index of the biphasic nature of the titration curve (Table I), wherein the I 75 and I 25 were the molar concentrations to give 75 and 25% inhibition of the control NADH-Q 1 oxidoreductase activity without TPB Ϫ , respectively. The I 50 values, i.e. the molar concentrations which gave 50% inhibition with 2 M TPB Ϫ , are also listed in Table I to compare the inhibitory potencies when complete inhibition was achieved with the aid of TPB Ϫ .
The dose-response curve of MP-6 (without TPB Ϫ ) for the inhibition of NADH-Q 1 oxidoreductase activity is shown in Fig.  3 (open squares) as a reference. Transformation of a tert-butyl group attached to the benzene ring of MP-6 to an n-butyl group (MP-11, closed circles) or a sec-butyl group (MP-12, closed squares) resulted in reduction of the biphasic nature of the dose-response curve. In particular, saturation of the inhibition by the two derivatives was less clear than that by MP-6. Considering that the hydrophobicities of these three derivatives are almost identical, these findings indicate that a bulky and rigid tert-butyl group of MP-6 is important for the selective inhibition. In the presence of 2 M TPB Ϫ , these three derivatives exhibited similar inhibitory potencies, as shown in the inset in Fig. 3. Complete inhibition was readily attained at concentrations less than 10 M irrespective of different substituents. The presence of TPB Ϫ makes the selective inhibition by pyridinium-type inhibitors ambiguous probably by facilitating inhibitor passage through the hydrophobic barrier to the remaining (i.e. hydrophobic) site (6).
The dose-response curve of the N-ethyl derivative (MP-13) showed no marked biphasic nature, although its inhibitory potency was almost identical to that of MP-6. This result indicates that a methyl group attached to the pyridine nitrogen atom is critical for the selective inhibition. In addition, the observation that the dose-response curves of MP-14 and MP-15 were much less biphasic than that of MP-6 indicates that the para-substitution pattern at the pyridine ring is also important.
Deletion of a branched methyl group in the spacer moiety linking the pyridine and benzene rings (MP-17) and transformation of the methyl to hydroxymethyl (MP-16) results in loss of the biphasic nature of the dose-response curve. These findings demonstrate that the branched methyl group is also an essential structural factor. 2 Other derivatives (MP-18 to MP-23) in which the spacer moiety of MP-6 was modified showed no or poorly biphasic titration curves with the exception of MP-23, indicating that three carbon atoms is the best length as the spacer moiety. MP-23, in which the methyl group in MP-6 was replaced by an ethyl group, showed a more significant biphasic nature than that of MP-6 (closed circles in Fig. 4). These results strongly suggest that the branched structure in the spacer moiety is very important to maintain selective inhibition. This is probably because some specific conformation of the two aromatic rings is regulated by steric congestion arising from the branched structure, as discussed below.
MQ-20, -21, and -22 did not show biphasic dose-response curves, indicating that a quinoline ring cannot substitute for the pyridine ring. The enhancement of their inhibitory poten-  cies by TPB Ϫ was slight compared with the corresponding pyridinium compounds (MP-21, MP-22, and MP-6, respectively). This was probably because quinoliniums do not readily form ion pairs with TPB Ϫ due to steric hindrance.
Structure/Activity Relationship of Pyridinium-type Inhibitors-The above results clearly indicated that there are rigid structural constraints for the selective inhibition by MP-6. Namely, a tert-butyl group on the benzene ring, a methyl group attached to the pyridine nitrogen atom, para-substitution in the pyridine ring, and the presence of a branched structure in the spacer moiety are important structural factors for selectivity.
However, the structural factors needed to elicit inhibition per se, not selective, have yet to be defined. Based on the results of structure/activity studies of a wide variety of pyridinium-type inhibitors (6,10,17,21) including the present study, the following conclusions may be drawn: 1) the inhibitory potency increases as the hydrophobicity of the inhibitors increases; 2) a methyl group attached to the pyridine nitrogen atom is not essential because N-ethyl or even bulky N-benzyl derivatives retain the activity; 3) the substitution position as well as the steric shape of the substituents on the pyridinium ring are not restricted; 4) the N-methylpyridinium ring itself is not essential for the activity since other aromatic rings such as bulky N-methylquinolinium can functionally substitute for it. Thus, the physicochemical structural factors, except for hydrophobicity, of the pyridinium-type inhibitors required for inhibition have yet to be defined. It is, however, likely that only two factors, a positive charge (i.e. electrophilic property), which may interact with the proposed anionic residue(s) in the binding site (6), and marked hydrophobicity which facilitates the access of the cationic inhibitor to the binding site through the hydrophobic environment in the membrane, are required for the inhibitory action. We recently showed that the ubiquinone reduction sites of bovine complex I are sufficiently spacious to accommodate exogenous bulky ubiquinone (22). This structural property is extremely unusual compared with other ubiquinone-mediated enzymes such as bovine heart mitochondrial complexes II and III (23), glucose dehydrogenase (24) and terminal ubiquinol oxidases (25) in Escherichia coli. Therefore, the loose recognition by the enzyme of the structure of pyridinium-type inhibitors may reflect the large cavity-like structure of the ubiquinone reduction sites.
Synthetic Development of Novel Selective Inhibitor-On the basis of the above structure/activity relationship, we synthesized MP-24 which fills the structural requirements for selective inhibition and possesses geminal dimethyl groups in the space moiety. This inhibitor showed a much clearer biphasic dose-response curve (Fig. 4, closed squares) as compared with MP-6 and MP-23. The complete inhibition by MP-24 without TPB Ϫ could not be determined because of solubility limit above ϳ400 M, but was readily attained at less than 10 M in the presence of 2 M TPB Ϫ (open squares). The selectivity of MP-24 in terms of the I 75 /I 25 ratio was at least 3-fold clearer than that of MP-6. As discussed for MP-6 (17), the site that is readily blocked by low concentrations of MP-24 without TPB Ϫ would be the hydrophilic binding site. The residual enzyme activity in the presence of MP-24 alone was completely inhibited by piericidin A (data not shown). This compound did not entirely inhibit ferricyanide or dichloroindophenol reduction by complex I at concentrations up to ϳ400 M, consistent with an earlier report of the original MPP ϩ (7).
The three-dimensional structure of MP-24 obtained by x-ray crystallographic analysis is shown in Fig. 5. 3 This compound is bent at the middle of the molecule due to steric congestion arising from the methyl groups and the aromatic rings. Considering a significant role of the branched structure, the bent conformation must be one of the important structural factors for the selective inhibition.
We confirmed that transformation of para-substitution to ortho-substitution (MP-25) and a tert-butyl group of MP-24 to an n-butyl (MP-26) resulted in marked reduction of selectivity, as observed for pairs of MP-6 versus MP-14 and MP-11, respectively. This structural specificity indicated that selective inhibition is the result of some sort of specific interaction between the inhibitor molecule and the enzyme (or its environment), and exclude the possibility that the apparent saturation of inhibition (i.e. biphasic dose-response curve) was due to artifacts such as solubility limits of the inhibitor.
Transformation of the tert-butyl group of MP-24 to a tertamyl (MP-27) resulted in slight reduction of the selectivity, indicating that a tert-butyl is the best substituent for this moiety. In addition, MP-28 did not show selective inhibition and its inhibitory potency without TPB Ϫ was much poorer than that of MP-24. The inhibitory potency of this compound was not significantly enhanced in the presence of 2 M TPB Ϫ . This was probably because MP-28 does not readily form an ion pair with TPB Ϫ due to its bulky and polar branched structure.
Effects of TPB Ϫ on the Inhibition by MP-24 -Previously, we reported that in the presence of a small amount of TPB Ϫ , the inhibitory potency of some potent MPP ϩ analogues (such as MQ-17 and MQ-18 in Ref. 17) was markedly enhanced, and complete inhibition was attained at inhibitor concentrations of less than 10 M (17). In contrast, the presence of an excess amount of TPB Ϫ partially reactivated the enzyme activity, and the inhibition was saturated at incomplete level. On the basis of an earlier report (6), we interpreted this partial reversal of inhibition as due to dissociation of the inhibitor from the hydrophilic binding site as a result of an increase in ion-pair formation. However, the inhibitors used previously were not selective inhibitors like MP-24, and variation of TPB Ϫ concentrations was not sufficient; therefore, the above hypothesis was re-examined using MP-24 in the presence of various concentrations of TPB Ϫ . Fig. 6 shows the effects of TPB Ϫ on the extent of inhibition by MP-24 at two different concentrations (0.9 and 95 M). At these concentrations, MP-24 exhibited about 30 and 50% inhibition without TPB Ϫ , respectively. It is likely that the concentration of 0.9 M MP-24 was sufficiently low to solely inhibit the hydrophilic site. The extent of inhibition by 0.9 M MP-24 increased as the concentration of TPB Ϫ increased, and maximum inhibition was attained in the presence of approximately equimolar TPB Ϫ (closed circles). Further increases in TPB Ϫ gradually reactivated the enzyme activity. The inhibition by 95 M MP-24 was also potentiated by TPB Ϫ in a concentration-dependent manner, and complete inhibition was attained in the presence of a catalytic amount of TPB Ϫ (closed squares), whereas reactivation was no longer observed. This result supports the previous hypothesis that the presence of an excess amount of TPB Ϫ resulted in dissociation of the inhibitor from the hydrophilic binding site.
pH Dependence of the Inhibition by MP-24 -On the basis of pH dependence of the I 50 value of MPP ϩ (plus 10 M TPB Ϫ ), Gluck et al. (6) suggested the existence of two distinct ionizable amino acids at or near the MPP ϩ -binding sites. Their experimental conditions, however, were complicated since MPP ϩ binds simultaneously to both of the sites in the presence of 10 M TPB Ϫ and the inhibition by TPB Ϫ itself is unavoidable at this concentration as discussed in the literature (6). We could not reproduce their experimental data (Fig. 5 in Ref. 6), which showed that the I 50 value of MPP ϩ (plus 10 M TPB Ϫ ) varies by about 2 orders of magnitude within the pH range of 6 -9 with three inflection points.
On the other hand, if the two components of total enzyme activity exhibiting different sensitivities to MP-24 without TPB Ϫ (Fig. 4) are indeed attributable to the two distinct binding sites, different pH dependences of the two components would be observed. Therefore, MP-24 might be a suitable inhibitor for the investigation of pH dependence to characterize the properties of the binding sites for pyridinium-type inhibi-  tors. Fig. 7 shows the dose-response curves of MP-24 without TPB Ϫ determined at various pH values. The scale of the horizontal axis in the figure differs between the left and right sides. It was clearly shown that inhibitor sensitivity of the high sensitivity component was not affected by the pH changes, whereas that of the low sensitivity component increased markedly when the pH increased over 8.0. This result indicated that MP-24 indeed interacts with the two distinct binding sites in complex I.
Origin of the Selective Inhibition by MP-24 -The marked biphasic nature of the dose-response curve of MP-24 disappeared in the presence of a catalytic amount of TPB Ϫ , and complete inhibition was readily attained at low concentrations of the inhibitor (Ͻ10 M) (Fig. 4). Since it is unlikely that MP-24 occupies the binding sites as an ion pair with bulky TPB Ϫ , this marked potentiation by TPB Ϫ suggested that the energetic barrier preventing the access of positively charged inhibitors to the hydrophobic site is reduced with the aid of TPB Ϫ (6, 10). We examined whether this energetic barrier can be overcome by prolongation of incubation time. Fig. 8 shows the dose-response curves of MP-24 determined without TPB Ϫ after various incubation periods at 30°C. It was clear that with longer incubation periods, the biphasic nature became less distinct, indicating that access of MP-24 to the hydrophobic site is promoted by prolongation of incubation time. However, in contrast to the observations in the presence of TPB Ϫ , about 15% activity remained even after incubation for 45 min, indicating that complete occupation of the hydrophobic binding site could not be achieved solely by prolongation of incubation time. In our previous study (17), the dose-response curve of MP-6 was not significantly affected by prolongation of incubation time (30 or 120 min). This discrepancy might have been because in the previous study, incubation was performed on ice and then the temperature of the reaction mixture was raised to 30°C 5 min before starting the enzyme reaction to prevent deactivation of the enzyme. We confirmed that the dose-response curve of MP-24 was not affected by prolongation of incubation time under the previous experimental conditions (data not shown).
The above results strongly suggest that the process of the inhibitor passage to the binding sites is responsible for the apparently selective inhibition. Since raising the temperature has the same effect as prolongation of the incubation period, we next examined the effects of reaction temperature on the inhibition by MP-24. Fig. 9 shows the dose-response curves of the inhibitor determined at various temperatures with 4 min incubation without TPB Ϫ (open symbols). Unexpectedly, the relative inhibition was saturated at ϳ35, ϳ20, and ϳ10% at 25, 20, and 15°C, respectively. We confirmed that this was not due to insolubility of the inhibitor, and the residual enzyme activity was completely inhibited by 0.1 M piericidin A (data not shown). In contrast to the observations at 30°C in which complete inhibition by MP-24 was readily attained at less than 10 M in the presence of 2 M TPB Ϫ (Fig. 4), the presence of TPB Ϫ no longer facilitated complete inhibition at 15 or 20°C (closed squares and circles, respectively). Complete inhibition by MP-24 with TPB Ϫ was achieved at around 200 M at 25°C (closed triangles). These results suggested that there is a significant energetic barrier preventing access of the inhibitor to the hydrophilic site as well as the hydrophobic site, although the latter is greater than the former.
On the other hand, the I 50 values with 2 M TPB Ϫ were almost identical between closely related derivatives such as MP-6, MP-11, and MP-12, and MP-24 and MP-26, although their selectivities in terms of the I 75 /I 25 ratio were entirely different (Table I). This excludes the possibility that a difference in the intrinsic inhibitor sensitivities between the two binding sites was responsible for the observed selective inhibition. This is probably because the inhibitor structure is not strictly recognized by the binding sites, as discussed above. Thus, it is likely that the structural specificity is closely related to the level of the energetic barrier preventing access of the inhibitor to the hydrophobic site. The molecular basis of how the structural specificity is concerned in ease of inhibitor passage is unclear because of the limited available information on three-dimensional structure of the enzyme. Nevertheless, as the sensitivity of the hydrophobic site to the inhibition by MP-24 was significantly enhanced as the pH increased, it is likely that a positively charged amino residue(s) with pK a of around neutral pH (probably histidine, cf. Ref. 6) interferes with access of the inhibitor to the hydrophobic site. Neutralization of the residue(s) by an increase in pH may reduce the energetic barrier of this kinetic process.