ARCHIVES OF RIOCHEMISTRY AND BIOPHYSICS 134, 597-602 (1969)

Leucine Aminopeptidase: A Zinc Metalloenzyme

S. RALPH HIMMELHOCH

Laboratory of Biochemistry, National Cancer Institute, National Institutes of Health, Department of

Health, Education, and Welfare, Bethesda, Maryland 20014
Received July 7, 1969

The leucine aminopeptidase of the supernatant fraction of porcine kidney is in-

stantaneously inhibited by orthophenanthroline, bipyridyl, cupferron, sodium di-
ethyldithiocarbamide, sodium sulfide, and sodium cyanide. All of these inhibitions
are reversed by addition of group IIb metal ions in appropriate quantity to the
reaction mixture. During purification of the enzyme under conditions which minimize
the addition of extraneous metal, the zinc-to-protein ratio in the fractions rises to a
final value between 4 and 6 g-atoms per 369,096 g of protein, as estimated by three
independent analytical methods, while the content of all other metals falls to negligi-
ble levels. Removal of the zinc from the purified enzyme or its replacement by cad-
mium results in loss of enzyme activity in proportion to the loss of zinc. Replacement
of the cadmium in cadmium leucine aminopeptidase by zinc restores native levels of
activity. Manganese can also partially replace the zinc of the purified enzyme yielding
a derivative more active toward bot)h leucine amide and leucine p-nitroanilide than
the native enzyme. These data suggest that leucine aminopeptidase is a zinc metallo-
enzyme.

Two studies of the role of metals in cataly- only failed to eliminate contamination from
sis by leucine aminopeptidase have been the reagents used, but actually involved pre-
made. In the first, Smith and his collaborators cipitation with zinc as a purification step.
(1) investigated the effects of added metals Thus, their finding of zinc in the final prod-
on the activity of porcine renal leucine uct can only be considered a provisional
aminopeptidase and found that manganese indication of the native metal content of
and magnesium activated the enzyme leucine aminopeptidase, particularly since
whereas “heavy metals” inhibited it Al- they also found that the zinc present in their
though the authors felt that their evidence final product, although firmly bound, was
favored the hypothesis that leucine amino- easily exchanged for manganese, clearly
peptidase is a magnesium or manganese demonstrating that the metal content of the
metalloenzyme, they did not measure the enzyme can be influenced by the metal
metal content of the native or the activated content of the solution from which it is
protein nor did they study inhibitors other isolated.
than EDTA and citrate ions, reagents that We have, therefore, undertaken a study
complex both the alkaline earth metals and of the role of metals in catalysis by native
members of the transition series. Thus, their leucine aminopeptidase from porcine kidney,
data are pertinent only to the phenomenon utilizing methods which satisfy the require-
of manganese activation, not to the nature ments outlined by Vallee (3) ; namely: 1.
of the native prosthetic group of the enzyme. isolation of the enzyme by methods which
The second study (2) measured the metal exclude contaminating metals; 2. measure-
content of the leucine aminopeptidase from ment of all metals present in the preparation
bovine lens. However, the enzyme prepara- to show that only one metal is present in
tions used were made by a method which not stoichiometrically significant quantity in the
597
 59s HIMMELHOCH

final product; 3. demonstration that removal TABLE I

of the metal results in proportional loss of ZINC CONTENT OF FRACTIONS OBTAINED DURING
activity, and replacement in quantitatively PURIFICATION OF LEUCINE AMINOPEPTIDASE
proportional restoration of activity; and 4.
Protein Zinc Zinc
Fractiona
determination by kinetic analysis of the Is) bg) h&g)
action of metal-complexing agents whether
the metal is at or near the active site of the Detergent-supernatant solu- 82 8ooo 98
enzyme. tion
The data obtained indicate that zinc is a Ion-exchange filtrate 13 1800 138
Ammonium sulfate precipi- 2.6 280 110
native prosthetic group of porcine leucine
tate
aminopeptidase.
Sephadex G-200 fractions 0.24 62 260
METHODS Fractions from ion-exchange 0.032 32 980
chromatography
Precautions were taken to exclude contamina-
tion of reagents and apparatus by metals (4). 5 The fraction designations are in accordance
Enzyme samples. Five samples of leucine amino- with Himmelhoch and Peterson (5). Protein
peptidase were prepared as described by Himmel- determined by the procedure of Lowry et al. (7)
hoch and Peterson (5), except that the final and zinc by atomic absorption spectrography.
products were dialyzed against metal-free 0.04 M
Tris-HCl buffer, pH 8.0, omitting magnesium.
The purified euzyme was stable for several months dex G-200 (5). The total zinc content of the
at 4” in the absence of magnesium. These enzyme fractions fell during purification presumably
preparations contained all of the chromatographic reflecting removal of other zinc-containing
variants of leucine aminopeptidase described speciesfrom the enzyme.
earlier (5).
The metal contents of five highly purified
Measurement of metals. Zinc, cadmium, mag-
nesium, manganese, iron, nickel, and copper were
preparations of leucine aminopeptidase are
measured by atomic absorption spectroscopy as shown in Table II. All of these contained
described previously (4). Total metal analysis stoichiometrically significant quantities of
was by AC-spark-emission spectrography on zinc, estimated both by atomic absorption
samples dry-ashed in platinum (4). Dithizone spectroscopy and b\- the dithizone method.
zinc determinations were by the method of Vallee The values ranged from 3.9 g-atoms per
and Gibson (6). Reference protein measurements mole (dithizone estimate of preparation V)
were by the method of Lowry et al. (7). to 6.2 g-atoms per mole (atomic absorption
Measurement of activity and chelate inhibition. estimate of preparation IV), assuming a
The standard assay employed 1.66 X 10T3 M L-
leucine-p-nitroanilide as substrat,e dissolved in
molecular weight of 300,000. A good correla-
0.1 M Tris-HCl buffer, pH 7.8. No magnesium or
tion was found between the zinc contents of
manganese was added to the reaction mixture. these preparations and their specific activi-
Measurements were carried out as previously ties (seefar right-hand column of Table II).
described (5), by following the increase in ab- No other metal was found in stoichiometri-
sorbance at 405 rnp at 25” in a Beckman DU spec- tally significant quantity in these prepara-
trophotometer. Variations in assay procedure are tions, except for a significant quantity of iron
described with each experiment in Results. in preparation V. In separate experiments
in which these preparations were analyzed
RESULTS
for magnesium after extensive dialysis, it
The zinc and protein contents of the frac- was found that their magnesium content
tions obtained during purification of leucine could be reduced to undetectable levels
aminopeptidase are shown in Table I. During without affecting the activity of the en-
purification, the ratio of zinc to protein in- zyme.
creased IO-fold to a value (for this prepara- Measurement of the total metal content of
tion) corresponding to 4.6 g-atoms of zinc Preparation V by AC spark-emission spec-
per 300,000 g of protein (the molecular troscopy indicated a zinc content of 700
weight of this leucine aminopeptidase as pg/g, an iron content of 180 pg/g and less
determined by chromatography on Sepha- than 0.1 g-atoms/mole of all other metals.
 ZINC IN LEUCINE AMINOPEPTIDASE 599

TABLE II
METAL CONTENTS OF PREPARATIONS OF LEUCINE AMINOPEPTIDASE

Copper Zinc/
Preparation C’b C’C
Wg)”

I 890 (865) 14 - - 100 8.9

II 870 (884) - 10 30 40 96 9.0
III 980 (1040) - - - 114 8.6
IV 1320 (1100) 12 - - - 158 8.3
11 1100 (830) 16 120 - - 99 11

a As det,ermined by atomic absorption spectrophotometry. Values in parentheses were those obtained

by dithizone analysis of the same samples.
b C’ is the first order rate constant/mg of enzyme nitrogen expressed in rnirl-1 (5).
c Obtained in arbitrary units by dividing the metal content of the preparation by its specific activity
toward leucine amide.

TABLE III When a group IIb metal, such as zinc, was

INHIBITION OF LEUCINE AMINOPEPTIDASE BY added to the reaction mixture with the com-
MET.4L-COMPLEXING AGENTS plexing agent, this inhibitory effect was
Agents vI/vcb largely abolished. This is shown in Fig. 1
where the effect of successive additions of
1,lO Phenanthroline .14 zinc chloride to a reaction mixture contain-
Cupferron .33 ing orthophenanthroline (OP) at a concen-
Sodium sulfide .GO tration sufficient to inhibit the enzyme to
Sodium diethyldithiocarbamide .62 50% of control activity is illustrated. The
Bipyridyl .67 enzyme activity rose with each addition of
Sodium cyanide .77 zinc to a maximum of 92% of the control
Ethylenediaminetetraacetic acid 1.00
value when the molar ratio of zinc to ortho-
Sodium fluoride 1.00
phenanthroline was closeto 1: 3, after which
a All inhibitors were present in a concentration the previously documented inhibition of
of 10-a molar in the standard assay medium. leucine aminopeptidase by ionic zinc was
b VI is the initial rate of reaction observed in observed. Similar partial reversal of the
t,he presence of inhibitor and Vc the initial rate inhibitory effects of cupferron (to 98% of
observed in a simultaneous control assay from the control rate) and bipyridyl (to 87% of
which the inhibitor was omitted. the control rate) was effected by added zinc.
The inhibition of leucine aminopeptidase
E$ect of metal-complexiny ayents on activity by various concentrations of OP at two sub-
of ‘kative” leucine aminopeptidase. The ef- strate concentrations is plotted (Dixon,
fects of a variety of metal-chelating or com- 1958) in Fig. 2. The data are consistent with
plexing agents on the rate of hydrolysis of competitive inhibition. The KI is approxi-
L-leucine-p-nitroanilide by leucine amino- mately 6 X 10-4.
peptidase are shown in Table III. The left- When a solution containing 1 mg of en-
hand column indicates the molar concentra- zyme per ml was dialyzed against lo+ M
tion of each agent in the reaction mixture, OP for 48 hr at 4”, and then against 0.1 M
and the right-hand column the ratio of the Tris-HCl buffer, pH 8.6, for another 4X hr
rate of hydrolysis observed in the presence to remove the OP, full activity was main-
of each agent to that observed in its absence. tained. Preincubation of the enzyme with
All of the agents, except EDTA and sodium 10s2M OP for 4 hr at 40” before assay caused
fluoride, listed inhibited enzyme activity. only the degree of inhibition expected from
The most marked inhibitions were observed the concentration of OP carried over from
with reagents that complex group IIb metals the preincubation mixture with the enzyme.
but have little or no affinity for the alkaline At lower pH values, irreversible loss of
earth metals. enzyme activity with time of preincubation
600 HIMMELHOCH

"I ",X 10
/

.2 4 .6 ,8 1.0

[Z"]/[OPl (nlole/rnole)
FIG. 1. Reversal of the inhibition of leucine
aminopeptidase by orthophenanthroline em
upon the addition of zinc ion. Each cuvette
contained 3 ml of 0.1 M Tris-HCl buffer, pH 7.8,
containing 0.8 X 10v3 M leucine p-nitroanilide and
7 X 10-4 M orthophenanthroline (sufficient to give
507, inhibition when compared to a control
FIG. 2. Inhibition of leucine aminopeptidase
reaction run without orthophenanthroline). The
by orthophenanthroline. Plotted according to
reaction was started by the addition of the stand-
Dixon (10). Substrate concentration of 0.83 X 10v3
ard aliquot of enzyme after the appropriate
M (0); substrate concentration of 1.66 X 10e3 M
amount of zinc chloride had been added to the
(Xl.
medium from a concentrated stock solution. For
the definition of VI and VC see Table III.
TABLE IV
EFFECT OF ZINC REMOVAL BY DIALYSIS AT Low
did occur, but at a rate close to that found
pH ON THE ACTIVITY OF LEUCINE
in the absenceof the chelating agent at the AMINOPEPTIDASE
same pH.
Effect of dialysis of leucine aminopeptidase Zinc specitic
cli~~~h~~~) activity
at pH 6. When preparations of leucine M/g g-atom/mole (C’)’
aminopeptidase were dialyzed against 0.05
0 1320 6.6 79
M Tris acetate buffer, pH 6.0, a progressive
12 990 5.0 58
loss of enzyme activity and zinc was ob- 24 82O 4.1 36
served. This is illustrated in Table IV where 48 650 3.3 40
the enzymic activity and Zn content of ali- 72 370 1.8 2.1
quots of leucine aminopeptidase are plotted 240 ND 0 0
as a function of time of dialysis against 0.05 ((5)
M Tris acetate buffer, pH 6.0, at 4”. The
0 Measured as described in Ref. 5, except that
rate of loss of activity was proportional to
no preincubation with Mn2+ was carried out.
the rate of loss of zinc, the activity of the
enzyme being completely abolished when the
zinc content of the enzyme reached unde- 0.001 M ZnClz and then against 0.05 M Tris-
tectable levels. Attempts to replace zinc by HCl buffer, pH 7.5, to determine if inactiva-
dialysis of “apo-” leucine aminopeptidase tion of leucine aminopeptidase at pH 6.0
(i.e., leucine aminopeptidase from which the could be prevented by an excess of ionic
zinc had been removed by dialysis at pH 6.0) zinc. Neither the lossof enzyme activity nor
against a variety of concentrations of zinc the lossof protein-bound zinc was prevented
chloride at pH values ranging from 6 to 8.6 by this maneuver.
were unsuccessful. Metal-exchange experiments. The effect of
Other aliquots were dialyzed against 0.05 dialyzing leucine aminopeptidase at 25’
M Tris acetate buffer, pH 6.0, containing against 0.04 M Tris-HCl buffer, pH 7.8,
 ZINC IN LEUCINE AMINOPEPTIDASE 601

containing 0.001 M CdClz for various times, TABLE V

and the effect of dialyzing the “Cd”-leucine MANGANESE-ZINC EXCHANGE IN LEUCINE
aminopeptidase thus prepared against the AMINOPEPTIDASE
same buffer containing 0.001 M ZnClz are
illustrated in Fig. 3. In all cases analyses
were performed after excess ionic metal had
been removed by dialysis against metal-free
0 980 4.5 ND ND 1.00
buffer as specified in Methods. The dialysis
12 620 2.8 290 1.6 1.44
of leucine aminopeptidase against CdClz
24 510 2.3 426 2.3 1.68
resulted in a simultaneous and stoichio- 48 421 1.9 495 2.7 1.78
metrically equivalent loss of zinc and gain in 72 400 1.8 258 2.7 .94
cadmium content. A concommitant and pro-
portional loss in enzymic activity was ob- a Calculated assuming a molecular weight of
served. When the enzymic zinc had been 309,090 (5).
reduced to negligible levels, activity had b VE is the activity measured as described in
been completely obliterated. The reversible “Methods” after dialysis for the time specified;
Vc is the activity measured before dialysis was
nature of this exchange is shown by the
begun.
second limb of the curve in Fig. 3 where,
upon dialysis of “Cd’‘-leucine aminopepti-
dase against ionic zinc, a loss of cadmium, a
gain in zinc, and a proportional restoration
Z of activity were observed, reaching close to
iii I I
control values in each case.
When native leucine aminopeptidase was
dialyzed at room temperature against 0.05
M Tris-HCl buffer, pH 8.6, containing 0.005
M Mm?&, a gradual, equivalent loss of en-
zyme-bound zinc, a gain in enzyme-bound
R/Ii?+, and a gain in activity were observed
(Table V). Complete replacement of the
“native” zinc by manganese, however, was
not achieved even after prolonged dialysis.
The maximum exchange was observed by
48 hr, at which point 58% of the “native”
zinc was replaced by manganese, and the
activity of the preparation towards leucine
p-nitroanilide was 1.78 times the control
value. Further prolongation of the dialysis
resulted first in loss of activity without addi-
tional change in zinc or manganese content,
DAYS OF DIALYSIS and eventually in gradual loss of both en-
FIG.3. Reversible exchange of cadmium and zyme-bound zinc and enzyme-bound man-
zinc in leucine aminopeptidase. A sample of ganese, presumably reflecting instability of
leucine aminopeptidase was dialyzed for 12 days the enzyme molecule on prolonged exposure
against 0.001 M CdClz in 0.04 M Tris-HCl buffer, to manganous ion.
pH 7.6, and aliquots were removed at the times
indicated by the points for analysis. The sample DISCUSSION
was then dialyzed against the same buffer con-
The evidence that leucine aminopeptidase
taining 0.001 M ZnCls and aliquots removed in a
similar fashion. Free Cd2+ or Zn2+ ions were re- is a zinc metalloenzyme can be summarized
moved by dialysis of each analytical aliquot for 24 as follows. First, during purification of the
hr against metal-free Tris-HCl buffer, pH 7.8 enzyme, the zinc-to-protein ratio of the frac-
prior to Cd (circles), zinc (triangles), and activity tions rose by a factor of 10, so that the zinc
(crosses) estimation. content of all of the purified preparations
602 HIMMELHOCH

examined was stoichiometrically significant. The specificity of this inhibition is suggested

Assuming a molecular weight of 300,000 not only by the fact that these agents, taken
(1, 5), the zinc content of the purest prepara- together, share little other than the ability
tion examined corresponded to 6 g-atoms to complex metals but also by the reversal
per mole of protein. Because of uncertainties of the inhibition by the addition of group
in the molecular weight, as well as variation IIb metals (8). The competitive nature of
in the zinc contents of various preparations, the inhibition of leucine aminopeptidase by
however, firm assignment of the “native” metal-complexing agents suggests that zinc
zinc content of the pure enzyme is not pos- is at or near the active site of the protein
sible. While successive fractions in the purifi- (3).
cation become enriched with respect to Hanson and his co-workers have demon-
protein-bound zinc, all other metals orig- strated a close immunological relationship
inally present are progressively removed, so between the leucine aminopeptidase of swine
that no metal other than zinc is present in kidney supernatant fraction and that found
significant (>0.5 g-atoms/mole) quantity in in cattle lens (2). They have found, addition-
any of the finally purified products. An ex- ally, that crystalline enzyme from the cattle
ception was found only in preparation five lens contains 8-12 g-atoms of zinc/326,000
which contained significant quantities of g of protein (9). Since, however, its purifica-
iron. Since, however, the other equally tion requires zinc precipitation it was not
active preparations lacked this metal, it possible to determine whether the zinc had
probably represented a contaminant. been present in the native enzyme or had
Secondly, removal of zinc from the pro- been added during purification. The present
tein by dialysis at pH 6 or exchange with data, taken in conjunction with their dem-
cadmium abolishes activity in proportion to onstration of extensive chemical and im-
the amount of zinc removed. Although this munological similarity between the bovine
might be interpreted as an adventitious and porcine enzymes suggests that the zinc
correlation in the case of inactivation at low found in the bovine protein is indeed a
pH (some other damaging transition occur- native constituent of it.
ring pari passu with loss of the metal), this
is less likely in the case of cadmium exchange, REFERENCES
where double exchange (first replacing the 1. SMITH, E. L., AND SPACKMAN, D. H., J. Biol.
zinc with cadmium and then replacing the Chem. 212, 271 (1955).
cadmium with zinc) results first in loss and 2. HANSON, H., GLASSER, D., LUDEWIG, M.,
then in restoration of activity. MANNSFELDT, H.-G., JOHN, M., AND NES-
Thirdly, the degree of activation of the VADBA, H., 2. Physiol. Chem. 348,689 (1967).
3. VALLEE, B. L., Enzymes 3,225 (1960).
enzyme by manganous ion (1) is correlated
4. HIMMELHOCH, S. R., SOBER, H. A., VALLEE,
not only with an increase in protein-bound
B. L., AND FUWA, K., Biochemistry 6, 2523
manganous ion, but also with a correspond- (1966).
ing loss in protein-bound zinc. This suggests 5. HIMMELHOCH, S. R., AND PETERSON, E. A.,
that the mechanism of manganese activation Biochemistry 7, 2085 (1968).
is by replacement of the “native” zinc by 6. VALLEE, B. L., AND GIBSON, J. G., J. Biol.
manganese. Chem. 176, 435 (1948).
Lastly, the native enzyme, which is active 7. LOWRY, 0. H., ROSEBROUGH, N. J., FARR,
in the absence of any added metal, is in- A. L., AND RANDALL, R. J., J. Biol. Chem.
hibited by a wide variety of agents which 193, 265 (1951).
8. ADELSTEIN, S. J., AND VALLEE, B. L., J. Biol.
share the property of strongly complexing or
Chem. 233, 539 (1958).
chelating group IIb metals but is not in- 9. FITTKAU, S., KETTMANN, U., AND HANSON, H.,
hibited by agents such as sodium fluoride, J. Labelled Comp. 2, 255 (1966).
which form strong magnesium complexes. 10. DIXON, M., Biochem. J. 66, 161 (1953).