Leucine Aminopeptidase (Bovine Lens) EFFECT OF pH ON THE RELATIVE BINDING OF ZNZ+ AND MG*+ TO AND ON ACTIVATION OF THE ENZYME*

‘K., ,n,Mg with increasing the activation leucine aminopeptidase is to the the free Zn*+

leucine aminopeptidase by Mg *+ is attributed to the lowering of the free Zn*+ concentration relative to that of free Mg2+ caused by the formation of ZnOH+ and Zn(OH), complexes with increasing OHconcentration. When corrections are made for the binding of Zn*+ by OH-ions, the pH-independent ratio of association constants ( 'Kzn: 'K& = 'K&,,,) for the relative binding of Znz+ and Mgz+ at site 1 of leucine aminopeptidase is 29,800. From the effect of pH on the relative binding constant, a value (@*) for the product of the two stepwise association constants for the formation of Zn(OH), from Znz+ and OH-(Znz+ + OH-= ZnOH+; ZnOH+ + OH-k Zn(OH) ) ., was estimated to be 4.42 x lOlo M-*at 37". Values of K, at pH 7.5 and 30" with L-leucine p-nitroanilide as substrate in the presence of 0.01 M NaHCO, are 4.13 and 2.01 mM for the zinc-zinc and magnesium-zinc enzymes, respectively. Values for V,,,,, are 0.2 and 2.49 rmol/min/mg, respectively.
The discovery that Mgz+ activates leucine aminopeptidase (EC 3.4. 1.1) was made in 1936 by Johnson et al. (2). Since then, a number of investigations have shown that Mn*+ is also an activator (3), that the activation is time-dependent (4, 5), and that activation may be attributed to the binding of the metal ion to the protein (6, 7). The latter conclusion, based on the observation that specific activity increased in the presence of increasing free Mgz+ or MnZ+ concentrations (6, 7), was criticized by Malmstrijm and Rosenberg (8) because direct measurements of bound metals were not made and the possibility that Mgz+ or Mnz+ might compete with another bound metal was not considered. These criticisms became important when it was found that preparations of leucine aminopeptidase contained Zn 2+ (9, 10). The criteria of Vallee (11) for the establishment of leucine aminopeptidase as a zinc metalloenzyme were demonstrated in this laboratory (12, 13) by showing that a stoichiometric amount of ZnZ+ could be removed and reinstated with a concomitant loss and restoration of full enzymic activity. Furthermore, the latter reports indicated that 2 g atoms of Znz+ were bound per 54,000 g of enzyme subunit (12 g atoms/324,000 g of oligomer) in agree-* This work was supported in part by Grants EY 00813 and AM GO608 from the National Institutes of Health. A preliminary communication of these results has appeared (1).
$ Recipient of partial support as a Predoctoral Trainee of the United States Public Health Service (TI-GM-31).
The research was abstracted from a thesis submitted in November 1974 as partial fulfillment of the requirements for a Ph.D. degree. ment with some (14) but not all earlier reports (9, 10) and that activation of the zinc enzyme results in the replacement of one of the Zn*+ atoms per 54,000 daltons by competing Mg*+ or MnZ+. Although Mgz+ and MnZ+ activations have generally been performed at pH values around 8.0, Melbye and Carpenter (15,16) were the first to give a detailed report of pH dependence for MgZ+ activation. They found that the optimum activation pH was actually 9.5 and not 8.0. Melbye (17) proposed that the pH dependence of activation was a result of the deprotonation, with increasing pH, of a group or groups on the enzyme which had a.high affinity for Mg*+. In their report of the determination of the relative binding constant of the enzyme for Znz+ as compared with MgZ+ and MnZ+ at pH 9.5, Carpenter and Vahl (13) offered another explanation for the pH dependence based on the fact that OH-will donate its free oxygen electrons and complex metals. They reasoned that OH-ion has a higher affinity for Zn *+ than for Mg*+ and would therefore bring about a decrease in free Zn*+ ion that would allow Mgz+ to be a more effective competitor for the enzyme ligands, whatever their state of protonation. Vallee and Wacker (18) have stated that both the deprotonation of enzyme ligands and the importance of OH-ion as a ligand are significant factors to be accounted for in any experiments measuring binding of metals to proteins. Coleman and Vallee (19) presented an equation which can be used to correct for these factors if one knows the stability constants of the metal hydroxide complex (or metal- were stored with an added drop of toluene in plastic vials at 4" for periods of up to 3 months with complete stability. Enzyme concentrations were determined by measuring the absorbance at 280 nm using an Et,"m 0 f 10 as reported by Carpenter and Vahl (13), corresponding to a molar extinction coefficient of 5.4 x 10' M~I cm-' for a 54,000-dalton subunit. Specific activities are in micromoles/ min/mg and are calculated from the difference in molar extinction coefficients of 9,900 M-' cm-l between p-nitroaniline and leucine p-nitroanilide at 405 nm at pH 7.5 (24). The value of 9,900 M-' cm-' is unchanged between pH 5.0 and 10.5 (25). Michaelis-Menten parameters are reported for similar conditions with the exception that the enzyme concentration was 20 rg/ml for the zinc-zinc leucine aminopeptidase, the assay volume was 2.5 ml, the substrate concentration was varied between 1.5 and 2.5 mM, and KC1 was omitted.
Measurement of Bound Met&-Approximately 1 ml of sample was applied to a 24-ml column (1 x 30 cm) of Bio-Gel P-6 equilibrated in O:i M N-ethylmorpholine HCl at pH 7.5 and eluted with a-flow rate of 2 mUmin.
Fractions of approximately 1 ml were collected in polypropylene vials until 30 ml had been eluted. The enzyme concentration of each fraction was measured before determining its zinc and magnesium concentrations.
The values for bound metals in g atoms per 54,000 g of leucine aminopeptidase were calculated from the average of the bound zinc and magnesium in the several fractions comprising the enzyme peak.
Incubation with Magnesium Chloride at Various pH Values-Solutions of leucine aminopeptidase at 0.9 "g/ml were incubated for 10 hours at 37' in 0.155 M trimethvlamine .HCl at DH 9.14 with 1 M KC1 and various MgCl, concentrations between 1.6" and 46.5 mM. Activity assays were performed at pH 7.5 at 0, 3, and 6 hours. At 10 hours a 200-~1 aliquot was removed from each sample and set aside for total metal determination.
The and variok concentrations of-l,lO-phenanthroline between 0.005 and 1 mM. Aliquots were removed for assay of specific activity in 2.5 ml of 2 mM leucine p-nitroanilide assay solution not containing bicarbonate at 0, 1, 2, 4, 6. 7, and 10 hours. After 10 hours the samples were eluted through the P-6 columns and the eluates were measured for bound metals. In another experiment, zinc-zinc enzyme (1.0 "g/ml) was incubated at 37' in 0.2 M N-ethvlmorpholine.HC!I at pH 7.5 with 1 M KCI, 0.05 mM l,lO-phenanthroline, and various MgCl, concentrations between 0.04 and 4 mu. Aliquots were removed for specific activity assays in 2.5 ml of a 2 rn~ leucine p-nitroanilide assay solution not containing bicarbonate at 0, 1, 2, 4, 6, and 8 hours. After 8 hours the samples were eluted through the P-6 columns and the eluates were measured for bound metals.
Treatment of Data-Michaelis-Menten parameters were calculated using the computer program HYPERB (26, 27). The K, values were corrected by multiplication by 0.455, the fraction of the substrate calculated to be in the deprotonated state at pH 7.5 using a measured value of 7.58 for the pK, of L-leucine p-nitroanilide. This correction accounts for the fact that leucine aminopeptidase binds only substrates with deprotonated a-amino groups (28, 29). All other constants were calculated with a Control Data Corp. CDC-6400 computer which was programmed to find values of parameters that would make the best least squares fit to the data to a given theoretical equation (30). The previous work of Carpenter and Vahl (13) as well as the nresent investigation indicated that there are two metal binding sites per 54,000 subunit.
The total g atoms of metal bound to each enzyme subunit, designated as r, is composed of the g atoms of metal per subunit bound at site 1 (referred to as the "activation" site), designated as 'r, plus the g atoms of metal per subunit bound at site 2 (referred to as the "structural" site), designated as 9, so that r = lr + =r = 2.
References to a specific metal ion is made by the use of subscripts so that rzn = 'rzn + lrzn (2) means that the total g atoms of zinc bound per subunit (rZ,) is equal to that bound at site 1, 'rznr plus that bound at site 2, *rZn. Similarly, the total g atoms of magnesium bound per enzyme subunit (r & is equal to that bound at site 1, *r M8, plus that bound at site 2, *r Mg: r Mu = 'r Mg + + Mg. ( However, under the conditions of exchange used by Carpenter and Vahl (13) as well as those used in this report, Mg*+ competes with Zn*+ for only one of the sites (the activation site) on the enzyme. Thus, [n addition, under these conditions, it was assumed that the two binding sites were saturated with metal and that when the measured sum of the g atoms of Zna+ and Mg2+ was less than 2.0, that some of the less tightly bound Mg2+ had been removed from the enzyme during the separation of free and bound metals by the column procedure. This assumption was based on the consideration that the sum of the bound metals was found to be 2.0 when the enzyme was incubated with no magnesium and was present only as the zinc-zinc enzyme, but generally was found to be progressively less than 2.0 as the amount of bound magnesium increased.
As a consequence of this treatment, The concentration of free magnesium in each sample was calculated by the following formula, where K" is the corrected constant, (L) is the molar concentration of the competing anion, and /3" is the product of n stepwise constants for the nth complex.
The competing anion in the buffer solutions used here was OH-Data from Sillen and Martell (33) indicates that while the OHconcentration is not high enough for it to be an effective nragnesiumbinding agent in the pH range of these experiments, it is important in the calculation of the true free zinc concentration. According to Fulton and Swinehart (34). the primary species of zinc in solution between pH R.0 and 9.25 at 25' are Zn*'. ZnOH+. and Zn(OH),.
In terms of a simple treatment as stepwise ahsociatlons. the formation of these species can be written as ZnOH' t OH-c:::? Zn(OH),; K, The sum of these equations is Since K,K, is the product of two stepwise association constants, n in Equation 14 is 2 and K,K, becomes pz. Assuming that the concentrations of these species are still much greater than that of Zn(OH); at 37" in this pH range, Equation 14 as it applies to the present work can be written 'K:,, = 'K,,,(l + Pz[OH-1') (15) where 'K& is the corrected association constant for the binding of Zn2+ at site 1.
The value of 'K,,, Mg as a function of pH is related to the corrected constant, 'KY,, ML 7 'K",,l'K$,, 0 where 'K,,, is taken as equal to IKI,,, as follows:  by passage of the enzyme solution through the P-6 column. As a consequence of these observations and the assumption that the two metal sites on each enzyme subunit were completely saturated before the enzyme solution was placed on the P-6 column, the g atoms of Mg*+ bound per subunit (rMg) was calculated as being equal to 2 minus the measured value for bound zinc (2 -rzn).
Measurement of Bound Metals-An elution diagram (Fig. 1) for the separation of free and bound metals shows that a 24-ml column of Bio-Gel P-6 completely separates 0.05 M MgCl, from a l-ml sample of the enzyme. Ultraviolet absorbance measurements at 280 nm on the eluates indicate that Fractions 6 to 10 contain the enzyme. Associated with the enzyme peak are peaks in the atomic absorption measurements of Zna+ and Mga+, corresponding to the enzyme-bound forms of these metals. The free Mg*+ peak is very large, extending from Fractions 12 to 31. No free zinc peak was observed, probably because the , column dilutes the small amount of free zinc present to a level below the detection limit of atomic absorption using the Boling burner. Table I shows the results obtained from measuring the Znz+ and Mg *+ bound per subunit for the series of solutions incubated at pH 9.14. The sum of the Znz+ plus Mg*+ bound per subunit decreases from 2.0 for enzyme incubated in the absence of added Mg2+ to 1.82 for enzyme incubated in 46.5 mM MgCI,. In general, the more Mg*+ bound to the enzyme, the greater is the difference between 2.0 and the sum of the Mg*+ plus Zn *+ bound per enzyme subunit. These results indicate that enzyme-bound zinc is completely stable to the column procedure used to remove unbound metal ions but that a portion of the enzyme-bound magnesium is removed were measured on each l-ml fraction. tiduse- Fig.  2 shows the change in specific activity with incubation time at 37" for various pH values and MgCl, concentrations. Equilibrium was reached, as judged by activity measurements, at 10 hours at which time the g atoms of zinc bound per subunit were determined by the column method. The relation between specific activity and bound Mg'+ or Zn*+ is shown in Fig. 3. Extrapolation of the line defined by the least squares fit of the data to the intercept at 1 g atom per subunit of either Mg*+ or Zn*+ yields a specific activity for the l-magnesium/l-zinc enzyme of 2.35 + 0.03 ~mol/minlmg in the standard assay using leucine p-nitroanilide at pH 7.5,

Determination of Ratio of Association Constants (*KZnIMN)
as Function of pH- Fig.   4 illustrates the g atoms of zinc bound per subunit ( lrzn) at site 1 as a function of [Znl+],/ [Mg*+j, at various pH values plotted in the reciprocal form used for graphical calculation of relative binding constants (36). From the data shown in Fig. 4, two major observations can be extracted. First, the ordinate intercept of the plot gives a value for the reciprocal of the number of binding sites involved. In this case, it is clear that Mg*+ is competing with Zn*+ for only one site on each enzyme subunit in the pH range studied. Second, the value of the slope, where only one class of site is involved, is equal to the number of binding sites divided by the relative binding constant.
Since the number of sites is equal to unity, the slope is the reciprocal of the relative binding constant and is seen to vary with pH. The value of 'KznlMg, the ratio of the apparent association constants of the enzyme for Zn'+ to that for Mg", decreases as the pH is increased, which indicates that it is easier for Mg*+ to replace Zn*+ at the activation site at high pH than at low pH. This result is consistent with the reports on the effect of pH on the activation of leucine aminopeptidase by either Mg*+ or Mn*+ (16,37).
It is proposed that this effect of pH on the Mgz+ activation of the enzyme can be attributed on one hand to a competition between Znz+ and Mg*+ for one site on the enzyme, and on the other hand, by a competition between the enzyme and OH-ion for the free Zn I+ in solution. At high pH values the OH-effectively competes with the enzyme for zinc and lowers the free Zn*+ concentration of the solution enough so that the less tightly bound Mg*+ (which forms much less stable hydroxides than zinc) can effectively compete with lowered Znz+ for one site on the enzyme.
Correction of the Relative Binding Constant (LKZn,Mg) for Formation of Zn(OH),-The measurements of 'rZn and [Zn*+],/[Mg"], at each pH were fitted by the least squares method to Equation 19 and yielded a value of 29,800 for the corrected relative binding constant (lK&.,J and 4.42 x lOlo M-* for Be the product of the stepwike association constants of Zna+ and OH-. Putting these values into Equation 18 yields the sigmoid curue in Fig. 5 representing the relation between the apparent relative binding constant (lKZnlMg) and pH. The points along this curve are the individual values of '&n,w at each pH that were derived in the preceding section. The point of inflection in this curve is at pH 8.33 and represents the pH at which the molar concentration of free zinc is equal to that of Zn(OH),.
The upper curue in Fig. 5 represents the pH-independent value of 'K&,, of 29,800. The points along the line are individual values of this constant that were calculated from Equation 18 using the values of the apparent relative binding constants (lKZnNg ) derived in the preceding section (plotted along the lower curve) and the fitted value of 8,. The comparison of these two curves demonstrates clearly that correction of the data for the formation of Zn(OH), yields a relative binding constant that is independent of pH. In other words, the phenomenon of pH-dependent Mg'+ activation of leucine aminopeptidase is a result of the lowered concentration of free zinc caused by the formation of zinc hydroxides. Activation of Leucine Aminopeptidase at pH 7.5 with Magnesium in Presence of l,lO-Phenanthroline-In order to show that other Zn2+-specific complexing agents will produce the same effect on Mg*+ activation of leucine aminopeptidase at pH 7.5 that OH-ion produces at higher pH, the enzyme was incubated with various concentrations of both MgCl, and l,lO-phenanthroline in 1 M KC1 and 0.2 M N-ethylmorpholine.HCl at pH 7.5. In the first series of experiments shown in Fig. 6A, the MgCl, concentration was held constant at 10 mM, while l,lO-phenanthroline was varied from 0.005 to 1 mM. The graph shows that up to a concentration of 0.05 mtu, increasing concentrations of the complexing agent enhanced activation, while further increases resulted in the eventual destruction of activity. Analysis of the samples for bound metals (Fig. 6A, inset) shows that Mga+ bound per subunit increases in correspondence with rising activity, while those samples which had lost activity at high l,lO-phenanthroline concentration had nearly all of their bound metals removed. In addition, the latter samples appeared to have nearly equal amounts of bound Mg2+ and Zn?+, leading to the speculation that Mgz+ will not bind to site 1 if there is no metal in site 2.
The results of the second series of experiments (Fig. 6B) where the MgCl, concentration was varied from 0.04 to 4 mM in the presence of 0.05 mM l,lO-phenanthroline shows corresponding increases in both specific activity and Mg"+ bound per subunit with increasing MgCl, concentration.
The highest activity attained, 1.6 Fmol/min/mg (using the standard leucine p-nitroanilide assay without added NaHCO,), was with the 4 mM sample.
Equilibrium studies under these conditions were thwarted by an apparently irreversible removal of site 2 zinc by l,lOphenanthroline which complicated the determination of equilibrium. However this approach unequivocally demonstrates that introduction of a complexing agent, which is more specific for zinc than magnesium, to compete with the enzyme for free Zn"+ in solution has the same effect on Mg2+ activation of leucine aminopeptidase as raising the pH. Determination The determinations were complicated by the fact that K, for both forms of the enzyme was higher than the solubility of leucine p-nitroanilide in aqueous solution at pH 7.5. Therefore it was difficult to obtain measurements over a wide enough concentration range to achieve high precision.
The value of K, does not change significantly between the zinc-zinc and magnesium-zinc forms, being 4.13 f 2.20 and 2.01 A 0.98 mM, respectively. The value of V,,, is increased from 0.20 * 0.09 to 2.49 i 0.83 rmollminlmg in going from the zinc-zinc to the magnesium-zinc enzyme.

Measurement
of Bound Metals-In previous reports from this laboratory (13) free divalent ions were separated from enzyme-bound metals by passage of the solution through a short column of sulfonated polystyrene resin (Beckman type 50A). The enzyme was in contact with the resin for less than 30 s, during which time there was complete removal of free divalent ion with virtually no effect on the bound Mg2+ and Zn2+. In the present experiments, incubation with divalent ions was performed for the most part in the presence of 1 M KCl. Since this quantity of salt in the eluate resulted in considerable interference with the measurement of Zn2+ and Mg2+ by atomic absorption spectrophotometry, it was necessary to use a technique that would largely remove these monovalent ions along with the unbound divalent metal ions. Several Sephadex and Bio-Gel resins were investigated in order to determine an effective separation of bound from unbound ions in a short exposure time. As shown in Fig. 1 the use of a Bio-Gel P-6 colymn (1 x 30 cm) gave a good separation of bound from unbound ions in a l-ml sample when operated at a flow rate of 2 ml/min. Under these conditions the enzyme is exposed to the gel resin for about 4 min. Even though bound zinc was completely retained by the enzyme under these conditions, the longer exposure time (4 min for the Bio-Gel P-6 uersus 30 s for the sulfonated polystyrene) resulted in some removal of bound magnesium.
Ratio of Association Constants ('K,,,,,)-The values obtained for the ratios of apparent association constants at various pH values are in good agreement with that obtained by Carpenter and Vahl (13) at pH 9.5. Using the values of A and 'KhMg reported here in Equation 18, the predicted apparent relative binding constant ( lKz,,& at pH 9.5 is 134, extremely close to the value of 152 reported by Carpenter and Vahl(13).
Problems associated with obtaining these ratios and their corrected values are numerous. From a practical point of view, the most critical problem is the measurement of pH. Equations 18 and 19 show that values of 1K&,,g are extremely sensitive to pH as this parameter occurs in an exponent. Thus, an error of 0.05 pH unit can affect 'K&,,, by as much as 20%.
Another critical factor is the measurement of metal. Although atomic absorption is an extremely sensitive method, and is perhaps also the most precise for quantitative purposes, it is still subject to at least 5% standard error unless impractical numbers of determinations are performed. Therefore, all measurements made by this technique are subject to at least this level of error. It was for this reason that, instead of determining the free zinc concentration from the difference between two atomic absorption measurements of similar magnitude (total minus bound zinc), a protein concentration measurement, which has much higher precision, was substituted for one of the atomic absorption measurements. This procedure was not used for the determination of free magnesium concentration.
In this case the atomic absorption measurements used to calculate the free Mg2+ concentration differed from one another by at least 1 order of magnitude. Consequently the error involved in the two measurements was insignificant as compared with their difference.
No determinations of the relative binding constants at pH values lower than 8.16 are reported. In order to obtain measurable amounts of Mg 2+ bound to the enzyme (>0.15 g atom per subunit) at these or lower pH values, MgCI, concentrations greater than 0.05 M must be employed. Such concentrations brought about a slow crystallization of the enzyme with resulting uncertainties in the obtainment of equilibrium and in the measurement of bound metals. Although it was not possible to determine the relative binding constant at pH values lower than 8.16, it was possible to demonstrate that the enzyme could be at least partially activated by Mg" at pH values as low as 7.5. Incubation of the zinc-zinc enzyme with 0.5 M MgCl, at pH 7.5 led to the formation of crystals which contained at least 0.63 g atom of Mg2+/subunit and possessed an enhanced specific activity (1.61 pmol/min/mg) commensurate with the magnesium content (30). These results indicate that magnesium can replace zinc at the activation site at pH values around neutrality if one can obtain a high enough ratio of magnesium to zinc ions to compete for the site. Attempts to drive this displacement to completion by performing the incubation in dialysis sacks which were placed in large volumes of zinc-free MgCl, solutions were only partially successful. Under such conditions zinc ions were also removed from the structural site (site 2), leading to inactivation and precipitation of a large portion of the enzyme (30).
Another comment is that the treatment proposed in this paper ignores electrostatic interactions (38). Since the measurements involved determining the ratio of binding constants rather than individual constants, electrostatic effects, which should be approximately equal for Mgz+ and Zn*+, should cancel out in the ratio determination.
The question as to whether or not factors other than OHcould explain the pH activation phenomenon warrants further discussion. In order to avoid complications from ions which chelate metals and whose concentrations change with pH, such as ammonia, carbonate, or Tris, buffers were prepared from either N-ethylmorpholine or trimethylamine, both of which show little tendency to bind divalent ions (33). This left OHas the primary component of the system which could bind metals and whose concentration would change with pH. The possibility proposed by Melbye (17) that ionization of functional groups on the protein are involved seems unlikely. Functional groups are better chelators at high pH as it is the free electrons in the deprotonated form that participate in the metal-ligand association, or in other words, at high pH the H+ concentration is low enough to allow the metal ions to compete for association with the ligand anion. In addition, the functional groups present in proteins which associate with zinc and magnesium all bind zinc more strongly than magnesium (18). To propose that the ionization of a functional group on the protein at increased pH causes a stronger association with Mg2+ compared to Zn2+ goes against all previous reports in this field. In most likelihood, the reverse is true, namely, that increased pH causes an even stronger association with ZnZ+ compared to Mg'+.
The theoretical treatment used to calculate t'he effect of [OH-] on the concentration of free Zn*+ in solution is based on the work of Fulton and Swinehart (34), who determined that, at pH values around 9 the predominant species of zinc at 25" are Znz+, ZnOH+, and Zn(OH),. It is assumed that these same equilibria are predominant at 37". Fulton and Swinehart (34) determined the constants for the following two equilibria at 25" Zn(OH),(s) F;=? Zn(OH),; K,, Zn(OH),(s) +=--k Zn" + 20H-; Ksp and reported values of K,, equal to 4 x 10e6 M and K,, equal to 7 x lo-Is MS. Rearrangement and addition of these two equations gives ZnzT + 20H-,=A Zn(OH),; KsJKs, where K,,IK,, is equal to @, of Equation 15. The value of p, at 25" of 5.73 x 10" M-', .calculated from the Fulton and Swinehart constants, can be compared with the value reported here for /3, of 4.42 x 10" Mm', calculated from the binding data obtained at 37". Although these values differ by a factor of 10, the difference is not out of line with what might be expected from the effect of temperature on the stepwise stability constants of Zn2+ with other ligands such as ethylenediamine (39,40). In addition, Perrin (41)  If 0, equals 4.42 x 10'" M-~, then K,, the second stepwise association constant, must be equal to 1.84 x lo5 M-l. Thus, K, and K, are of equal magnitude, in agreement with the general observation that the n stepwise association constants between a metal with a high coordination number and univalent anions are usually of similar magnitude (42).
Michaelis-Menten Parameters-Since different assay conditions were used in this investigation than in previous work, new measurements were made of K, and V,,, for both the magnesium-zinc and zinc-zinc enzymes. These results agree with previous observations (13,23,43) in that K, is not changed significantly by substitution of Mga+ for Zn*+, while V m8x is increased approximately 12-fold. The actual values of K, for the zinc-zinc and magnesium-zinc enzymes (4.13 and 2.87 mM, respectively) are somewhat higher than those reported by Lasch et al. (23). However, this difference may be attributed to different assay conditions plus the limitations imposed by the low solubility of the substrate. Because of the low substrate solubility, K, values reported here as well as in Lasch et al. (23) are estimated from a small range of substrate concentrations, all of which are below K,, and consequently are subject to considerable error. Acknowledgment-Some of the preparations of leucine aminopeptidase used in these investigations were those of Shamshir Kang, Gary Glasser, or Allen Taylor. We are indebted to Sharon Glasser Fisher for technical assistance.