71
’
BRL?NORGAA.!IC CHEMISTR Y 7,7 l-86 ( 1977)
Concerning the Metalloenzyme Ascorbate Oxidase
KENNETH G. KRUL* and CHARLES R. DAWSON Department of Chemisny, Columbia University. New York, NY 10027
ABSTRACT Apoascorbate oxidase has been shown to have a molecular weight of 137,000 i: 3,000 and essentially the same gross quaternary conformation as native ascorbate oxidase. The apoenzyme, however, lacks much of the conformational stabihty of the native enzyme. The removal of the copper from the oxidase protein, and the simultaneous reduction of the disultide bonds results in an apoenzyme of lower structural stability than the native oxidase. The aging of apoascorbate oxidase has been found to involve a loss of ionizable tyrosine residues and a dissociation to subunits and component polypeptide chains, which was not observed with the more stable native and holoenzymes. The molecular weight of holoascorbate oxidase has been determined of 9.79 has been determined for the holoenzyme. to be 285,000. An s&U, Holoascorbate oxidase has been shown to have an electrophoretic mobility on polyacrylamide gels that is 23% lower than either the native or apoenzyme. Furthermore, electrophoresis of the holoenzyme in buffers containing dodecyl sulfate, and also isoelectric focusing of the holoenzyme. produce patterns of greater similarity to those of apoascorbate oxidase than the native enzyme. KEY WORDS Apoascorbate dase.
Osidase,
Ascorbate
Oxidase,
Holoascorbate
Oxidase,
Oxi-
INTRODUCTION
Ascorbate oxidase (L-Ascorbate: Oa oxidoreductase, EC 1.10. 3.3) is a copper-containing enzyme that catalyzes the aerobic oxidation of Vitamin C. Experimental evidence indicates that this enzyme has a molecular weight of 140,000 [I] and is made up of two identical subunits [2] _ Each subunit has a molecular weight of about 70,000 and is believed to consist of two polypeptide chains
of molecular
*Present address Hoechst Corporation,
weights
approximating
30,000
and address for correspondence: Somerville, NJ 08876.
and 40,000. Behring
Diagnositcs,
Q Elsevier North-Holland,
American
inc., 1977
K. G. KRUL AND C. R. DAWSON
72
Three distinct variants of ascorbate oxidase have been isolated, or produced, native ascorbate oxidase, apoascorbate oxidase, and holoascorbate oxidase. N~rive ascorbclte oxidase is the name given to that variant of the enzyme that has been isolated directly from its plant source. It is enzymatically active and contains 8-10 atoms of copper per molecule [3] _ The native oxidase is characterized by copper content, and by disulfide bond content, that have not been purposefully exposed (for experimental purposes) to conditions that result in copper loss or disulfide reduction. The native oxidase is stable over extended periods of time_ Dialysis against cyanide converts the active, copper-containing, native ascorbate oxidase to its inactive, copper-free apoenzyme, upoascorbate oxidase [4] _ Chang [S] has devised a practical method for the restoration of copper content and enzymatic activity to the apoenzyme. This restored variant of ascorbate oxidase is referred to as ~roloascorbare oxidase. It has been the purpose of the present investigation to examine and compare the properties and the structural features of these three variants of ascorbate oxidase.
in these laboratories:
MATERIALS
AND METHODS
I _ Enzyme Purification Ascorbate
oxidase was isolated
from the green zucchini
squash (Cucurbira
pepo nteduliosa) and purified as described by Lee and Dawson [3] _ 2. Chemicals acrylarnide (electrophoresis Sodium dodecyl sulfate, 2-mercaptoethanol, grade), N,N’-methylene-bisacrylamide, Tris and glycine (ammonia free) were obtained from Eastman Organic Chemical Company_ Bromphenol blue, Buffalo Black NBR, citric acid, cupric sulfate and ascorbic acid were purchased from Fisher Chemical Company. Coomassie Brilliant Blue R-250 was purchased from Schwarz/Mann. Arnpholine carrier ampholytes (40% in water) were purchased from LKB Produktur. BioGel A-I Sm and A-5m (loo200 mesh) were purchased from BioRad Laboratories, and Sephadex G-l 50 was obtained from Pharmacia Fine Chemicals. All aqueous solutions were prepared using water which had been filtered through a glass wool falter and passed through a Culligan D-25 mixed-bed deionizer (Nelson, Phillips and Co., Inc.)_ The copper content of this water was less than 0.005 ppm_ All chemicals were used without further purification.
CONCERNlNG
THE METALLOENZWE
ASCORBATE
OXIDASE
73
3. Protein Determination
The protein concentrations of all enzyme solutions were determined by the Lee modification [6] of the method of Lowry et al. [7] _ This modification is specific for the protein moiety of ascorbate oxidase. 4. Copper Determination Enzyme samples were treated with Chelex-100 (BioRad Laboratories) for 45 min to 1 hr and/or exhaustively dialyzed against copper-free buffer prior to the determination to remove essentially all nonenzymatic copper. Copper was then determined by the method of Stark and Dawson [8] _ 5, Activity Determination Activity was measured by the initial rate of oxygen uptake during the enzymatic oxidation of L-ascorbic acid, essentially as defined by Dawson and Magee [9]. Oxygen uptake was measured using a Gilson KM Oxygraph in a reactive system as defined by Strothkamp and Dawson [2] _ 6. Preparation of ApoAscorbate Apoascorbate
Oxidase
oxidase was prepared as described by Penton and Dawson [4] -
7. Preparation of HoloAscorbate
Oxidase
Fresh apoascorbate oxidase samples were restored to the active holoenzyme by the method of Chang [S], followed by centrifugation of the samples at 17,500 X g for 15-25 min, at 6-10 O. (The temperatures throughout are given in “C.) 8. Gel Filtration All columns were run at room temperature (25-27 “). Sephadex G-150 and BioGel A-1.5m (100-200 mesh) were used in experiments involving standard buffer conditions_ BioGel A-5m (100-200 mesh) was used in all experiments involving buffers containing sodium dodecyl sulfate_ Standard curves and molecular weight determinations were calculated in accordance with Determann [lo] _ 9. Sedimentation
Velocity Determination
The sedimentation coefficients of holoascorbate oxidase samples were determined on a Beckman Model E uhracentrifuge. The sedimentation constant was determined by 20 p with a rotor speed of 60,000 r-pm. 10. Gel Electrophoresis Polyacrylamide disc electrophoresis experiments were carried out essentially in accordance with the method of Ornstein and Davis [1 1] _ Experiments in
K_ G_ KRUL AND C_ R_ DAWSON
74 these laboratories
varied from the published procedure
in that sample gels were
never used, and, in instances where the volume of the protein sample was small, stacking gels were omitted. 11. Treatment
of Protein with Dodecyl Sulfate
Protein standards and samples of ascorbate oxidase, which were later examined by either gel electrophoresis or fel filtration in buffers containing sodium dodecyl sulfate, were incubated prior to analysis according to Strothkamp and Dawson [2] _ 12. Molecular Weight Determinations
by Gel Electrophoresis
Molecular weight determinations of the oxidase samples were made essentially by Weber and Osbom [12] _ Rather than the published procedure, instead, 7.5% acrylamide 13. Isoelectric
dissociation patterns of ascorbate the electrophoretic technique of 10% acrylamide gels used in the gels were used in this investigation_
Focusing
Isoelectric focusing experiments procedure of Carrel et al. [ 13]_ 14. Tyrosineflryptophan
were
carried
out
accord
with
the
Molar Ratio Determination
The TyrosinelTryptophan Molar Ratios of ascrobate determined by the method of Bencze and S&mid [ 14]_ RESULTS
in
oxidase
samples were
AND DISCUSSION
I _ Gel Filtration
and Sedimentation
Experiments
Fresh apoascorbate oxidase samples (lo- 15 mg), prepared from homogeneous native enzyme preparations. were subjected to gel filtration on Sephadex G-l 50. Figure 1 shows a typical elution pattern for the apoenzyme. As can be seen, the apoenzyme (second peak) was indicated to be largely homogeneous, with some evidence of the presence of minor dissociation components_ The molecular weight (Icr,) of the apoenzyme has been calculated from gel filtration data to be 137,000 t 3,000, which is in good agreement with that calculated for native ascorbate oxidase [ 1,2] _ Equal amounts of native and apoascorbate oxidase (1 O- 15 mg) were mixed and subjected to gel filtration on Sephadex G-150. Native ascorbate oxidase binds its copper tightly and there is no exchange of copper ions between the native and apoenzymes [2]. Figure 2 shows a typical elution pattern for this experiment_ As can be seen, native and apoascorbate oxidase migrate together_ The M, calculated for the peak representative of the two ascorbate oxidase
CONCERNiNG
THE METALLOENZYME
ASCORBATE
OXIDASE
0.6-
-O.IZ
OS-
r
B 0 :0.4-
0.10
-
0.06
-
0.06
Y
-
0.04
Y 2 E
-
0.02
E -
*z.onm
_---
2 “, 0.3
-
*x60
I
-
nm
z
4’ P 5: 0.2-
g
3
t 0.1
-
I
I 20
1
‘ 30
I 40
I
t
FRACTION
I
L 60
I 60
I
I
I
70
’
NUMBER
FIG. 1. A typical elution pattern of fresh apoascorbate oxidase in 0.2 M phosphate buffer, pH 8.0, on a Sephadex G-l 50 column_ The first peak (from left to right) is that of blue dextran, the second is the apoenzyme, and the third is that of dinitrophenylalanine.
1.4
J
l-
0.35
7
O-B-
0.05
1 30
I 35
1 40
1 46 FRACTION
8 50
8 65
1 60
1 65
70
NUMBER
FIG. 2. A typical elution pattern of a 1: 1 mixture of native and apoascorbate oxidase in 0.2 M phosphate buffer, pIi 8.0, on a Sephadex G-150 column. The first peak (from left to right) is that of blue dextran, and the dotted peak is that of dinitrophenylalanine. The peaks falling between the blue dextran and dinitrophenylalanine are due to the ascorbate oxidaselapoascorbate oxidase mixture.
K. G. KRUL
76
AND C. R. DAWSON r - O-8
0-B
~ 35
50
45
FRACTION
FIG. 3. A typical phosphate buffer,
elution
pattern
55
60
65
70
75
NUMBER
of apoascorbate 1% in dodecyl sulfate,
oxidase
in 0.2
M sodium
on a BioGel A-Sm column. The first peak (from left to right) is that of blue dextran, and the last is that of dinitrophenylalanine. The peaks occurring between these are due to the six dissociation species of the apoenzymepH 7.0,
variants migrating together is 143,000 t 1 ,OOO_The three minor peaks eluted after the native/apoenzyme mixture have calculated M,‘s of 108,000, 89,000, and 68,000 t 2,000. These peaks can be accounted for on the basis of the dissociation of the quaternary structure of the protein moiety of ascorbate oxidase, as described by Strothkamp and Dawson [2] _The total presence of these peaks, however, comprises less than 10% of the protein applied to the column. While the minor peaks indicate the lesser quatemary stability of apoascorbate oxidase (they are not observed with native ascorbate oxidase alone[2] ), they do not substantiate any major conformational difference between the native and apoenzyme molecules_ In order to better observe the dissociation of apoascrobate oxidase, samples of the apoenzyme were incubated according to the method of Strothkamp and Dawson [2], and subsequently subjected to gel filtration on BioGel A-5m. Figure 3 shows a typical elution pattern for the treated apoascorbate oxidase. The protein peaks correspond to M,‘s of 138.000; 80,000; 70,000; 57,000; 47,000; and 28,000, respectively. All of these peaks can be accounted for on the basis of simtilar patterns resulting from-like treatment and dissociation of the quatemary structure of native ascorbate oxidase [2] _ Thus it can be concluded that, on the basis of the above gel filtration experiments_ apoascorbate oxidase, while somewhat less stable than the native enzyme, maintains the same molecular weight and essentially the same gross quaternary conformation as native ascorbate oxidase.
CONCERNING
THE METALLOENZYME
ASCORBATE
OXIDASE
0.6 -
77 -
0.12
-
0.10
: s
-
0.06
* s Y
-
0.06
,”
-
0.04
5 ,”
-
0.02
2
a
1
I 20
1
30
1
40 FRACTION
1
50
1
60
L
I
70
NUMBER
FIG. 4. The elution pattern of holoascorbate oxidase in 0.2 M McIlvaine’s Buffer, pH 5.6, on a BioGel A-1Sm column. The first peak (from left to right) is that of blue dextran, the second is that of the holoenzyme, and the third is dinitrophenyIaianine. The holoenzyme solutions were concentrated by ultrafiltration to approximately 0.75 ml prior to loading on the column. GeI fihration of fresh holoascorbate oxidase was carried out on a BioGel A-1Sm cohrmn. Figure 4 shows a typical ehrtion pattern observed in the above experiment_ The molecuIar weight (M,) of the holoenzyme has been calculated from the gel filtration data to be 285,000. This is approximately twice the molecular weight of either native or apoascorbate oxidase. The sedimentation constants of hoIoascorbate oxidase samples were determined on a Beckman Model E analytical ultracentrifuge. The &!o,~ for the holoenzyme was found to be 9.79. This may be compared with the ~go,~ for the native ascorbate oxidase of 7.52 161. The increase in the sedimentation constant of the hoIoenzyme over that of the native enzyme supports the finding that the M, of hoIoascorbate oxidase, as determined by gel filtration, is twice that of either the native or apoenzyme. Apparently the reconstitution process causes the protein molecuIes of ascorbate oxidase to dimerize, thus forming a tetramer of subunits. As indicated by the symmetry of the peaks resulting during gel filtration and ultracentrifugation, this new, higher molecular weight variant is a homogeneous entity. Although the quatemary structure of the holoenzyme is different from that of the native enzyme, it is still catalyticaliy active. On comparing the specific activities of numerous holoascorbate oxidase samples with those of their corresponding native ascorbate oxidase samples (See Table l), it has always been observed that the holoenzyme is more active than the native enzyme. Thus, the distinct character of holoascorbate oxidase is clearly apparent_
K. G. KRUL AND C. R. DAWSON
78 TABLE
1
Specific Activities of Native Ascorbate Oxidase Samples and Their Corresponding Holoenzymes
Sample
Spec. activity (U/mg protein)
Spec. activity (Uhg
copper)
NatAO-1 HoloAO-I
1,368 1,890
305 475
NatAO-2 HoloAO-2
2,175 3,154
733
NatAO-3 NoloAO-3
1,476 4.220
591 839
NatA HoloAO4
2,900 4,326
439 886
Nat AO-5 HoloAO-5
2,490 3,956
838 1,049
NatAO-6 HoIoAO-6
2,490 3,809
838 1,021
NatAO-7 HoIoAO-7
2,490 3,568
838 1,017
NatAO-8 HoloAO-8
2,490 3,231
838 1,099
NatAO-9 HoloAO-9
2,490 3,375
838 1,099
573
2. Gel Electrophoresis
Samples of native, apo- and holoascorbate oxidase were examked by disc electrophoresis. The electrophoretic mobilities of the samples versus an internal standard (bromphenol blue) were calculated as described by Weber and Osbom [I21 _ Figure 5 illustrates the typical gel patterns demonstrated with the three variants of ascorbate oxidase.
CONCERNING
THE METALLOENZYME
I 1
ASCORBATE
-
-
2
3
OXIDASE
FIG. 5. Showing the results of polyacrylamide gel electrophoresis of native (Gel 11, apo- (Gel 2) and holoascorbate oxidase (Gel 3).
79
at pH ‘3.5
While native and apoascorbate oxidase were observed to have essentially the same mobility (0.298 5 O-003), that of the holoenzyme was observed to be significantly (23%) lower (0229 + 0.002). Furthermore, holoascorbate oxidase exhibited one, uniform electrophoretic band, similar to that of the native enzyme. Both of these gel patterns were unlike the diffuse, multibanded and indistinct gel pattern of apoascorbate oxidase that has been observed in this investigation and previously [S] _ The single-band, electrophoretic pattern of the holoenzyme further attests to its homogeneity. The lower electrophoretic mobility of holoascorbate oxidase, as compared with its native and apoenzymes, is entirely consistent with the finding of increase molecular weight and sedimentation constant of holoenzyme preparations_ Samples of the native, fresh apo- and holoenzymes were treated with dodecyl sulfate and 2-mercaptoethanol, and analyzed by gel electrophoresis in buffer containing dodecyl sulfate- Table 2 and Fig. 6 illustrate the results of these experiments_ As can be seen, hoioascorbate oxidase exhibits a dissociation pattern more closely resembling that of the apoenzyme than that of the native enzyme_ This can be taken as an indication that the holoenzyme, it its two-fold native conformation, maintains its quatemary structure less firmly than native ascorbate oxidase, and more on the order of the apoenzyme. Additionally, the sulfhydryl
K. G. KRUL AND C. R. DAWSON
80 TABLE 2
Molecular Weight Analyses of Native, APO- and HoloAscorbate Native
71*** 59 50.5 40.5
Oxidase”
Apo-
Holo-
107 81** 69 62 ss* 53.5 47 39.5
108 83.5** 70.5 65 59.5* 56 47.8 41.5
a Values given are those of the molecular weights of bands appearing on the polyacryIamide gels (See Fig_ 6). Asterisks are used to denote the relative intensities of the bands on the gels (*** more than **, ** more than *) and the intensities are relative to the other bands on that particular vertical column. Values not suffixed by an asterisk were considered to be of minor importance due to their weak intensities.
-
1. -
f! -
-
2
3
FIG. 6_ Showing the results of polyacrylamide gel electrophoresis of native (Gel I), apo- (Gel 2) and holoascorbate oxidase (Gel 3), in the presence of dodecyl sulfate and 2-mercaptoethanol. The lower three bands exhibited on Gel I are extremely faint, and are considered insignificant.
CONCERNING THE METALLOENZYME TABLE
ASCORBATE OXIDASE
81
3
Dissociation Species of ApoAscorbate
OxidaseO
(Molecular weights in thousands)
Day I
87-O***
Day II
s3_0***
Day 15
87-o*** 78.0
Day 20
97.5 87-O*** 71.5
53.0*** 42-b**
43-o***
46.0***
29.2***
29-o*:**
31.0***
16.5
16.0
Day 83 113 102 94.8 94.2* 72.0***
66-o*** 58.8 43.0* 38.5 29.1* 24.2 20.2 16.2
L2Results of a time study using polyacrylamide gel electrophoresis in the presence of dodecyi sulfate to observe the stability of aging ApoAscorbate Oxidase. Asterisks are used to denote the relative intensities of the band (*** more than **, ** more than *), and the intensities are relative to the other bands on that particular vertical column. Values not suffixed by an asterisk were considered to be of minor importance due to their weak intensities_
groups of the holoenzyme appear to be more easily accessible to reduction by 2mercapto-ethanol than in the native enzyme, as evidenced by the significance of bands having moIecuIar weights other than that of the major subunit, i.e., approximately 70,000. While the gel patterns of native and holoascorbate oxidase were essentially the same in all samples treated with dodecyl sulfate, the gel patterns of the dodecyl sulfate-treated apoenzyme changed with the age of the apoenzyme sample. The aging pattern of apoascorbate oxidase was therefore investigated by this technique_ Samples of apoascorbate oxidase were stored aerobically, under refrigeration (4 “) in 0.2 M McIlvaine’s Buffer, pH 5.6, for periods of up to 80 days. At various iime intervals, ahquots of the apoenzyme were taken and examined by ge1 electrophoresis in buffer containing dodecyl sulfate. Table 3 shows the molecular weight variations exhibited at significant points in the aging of apoascorbate oxidase. The consistently reoccurring bands indicate that the most stable dissociation species of apoascrobate oxidase have molecular
82
K. G. KRUL AND C. R. DAWSON
I
P
FIG. 7_ The suggested quaternary structures of ascorbate oxidase as supported by evidence in this investigation (I), and as previously proposed (II). A is a polypeptide chain of fir, approximately 30,000 and B is a polypeptide chain of Mr approximating 40,000_ The disulfide bonds shown (-SS-) are the minimum number necessary to explain the experimental data. Electrostatic forces and hydrogen bonds are represented by dashed lines (----).
weights that average 85,600,43,000, and 29,900. In extremely fresh apoenzyme samples, only two species are observed, having molecular weiets of 87,000 and 53,000_ Wirh the aging of the apoenzyme samples, the gel patterns grow increasingly complex. This increased compIexity was not due to bacterial growth, becauSe native ascorbate oxidase samples, stored similarly for the same periods of time, do not show this effect. The results of these experiments indicate that the quaternary composition of the apoenzyme is the same as that of the native enzyme [2]. However, the data presented here tend to support an asymmetrical structure (Fig_ 7, I) rather than the symmetrical structure (Fig. 7, Ii) as proposed by Strothkamp [IS] _ As has been observed previously [2J, and confirmed in this investigation, native ascorbate oxidase yields only one major band (Mr approximating 70,000) during gel electrophoresis in buffer containing dodecyl sulfate_ When the disulfide bonds of the protein were reduced after dissociation of the subunits by dodecyl sulfate, the gel pattern showed two significant bands of M,‘s approximating 30,000 and 40,000. Structure II (Fig. 7) is entirely consistent with this data. In the preparation of the apoenzyme, however, the disulfide bridges were initially reduced by cyanide. The dissociation by use of dodecyl sulfate takes place sllbsequent to the cyanide reduction. it is thus possible to observe dissociation species corresponding to dimers of like polypeptide chains, i.e., M,‘s approximating 60,000 and 80,000. As interaction between like polypeptide
CONCERNING
THE METALLOENZYME
ASCORBATE
OXIDASE
83
chains, resulting in such “dirnerization” would appear to be maximized by their juxtaposition, Structure I (Fig. 7) would appear to be a better model for the quatemary arrangement of the subunits of ascorbate oxidase. The asymmetrical model is consistent with the previous data 12,151 and that obtained in this investigation_ It is interesting to note that, at the end of 83 days, the half-molecular weight subunit (kfr = 72,000) is the predominant species present (See Table 3). This is particularly interesting in light of the fact that in younger samples of apoascorbate oxidase it is not present, or present only in insignificant quantities on the gels. The apparent recombination of polypeptide chains and the reformation of the disulfide bonds are indicative of the thermodynamic stability of the halfmolecular weight subunit moiety, while the increasing complexity of the gei patterns indicate the decreasing stability of the apoprotein moiety as a whole. 3. Isoelectric
Focusing
Samples of native, fresh ape-, and holoascorbate oxidase were repeatedly analyzed by isoelectric focusin g on polyacrylamide gel columns_ In each case, enough protein solution was added to the gel mixtures, such that the final concentrations of the various enzyme variants were approximately equal in terms of protein. The ampholyte-induced pH gradient ranged from pH 3 to pH 10 at 15 ’ _ Figure 8 illustrates the results obtained in a typical experiment. It can be seen that the isoelectric focusing pattern of the hoIoenzyme (Gel 3) more cIosely parallels that of the apoenzyme (Gel 2) than that of the native enzyme- This is particularly interesting due to the fact that isoeIectric focusing separates moIecules soIeIy on the basis of their isoelectric point (PI). The gel serves merely to stabilize the ampholyte-induced pH gradient, and does not function as a molecular seive [ 16,17]_ Native ascorbate oxidase, shown to be homogeneous by gel electrophoresis, was indicated to contain two very similar components in respect to pi by isoelectric focusing_ The secondary band was observed to be very close to the primary band (See Fig. 8, Gel 1) and may be an artifact. Such artifacts have been observed by others studying different proteins [18,19]_ Or, since its pi is essentially that of the native enzyme, the secondary component may well be simpIy some native protein that has lost some of its copper during the poIymerization of the ge1. It is also possible that the secondary component seen by isoelectric focusing may be a small amount of dissociated, half-molecuIar weight subunit (approximately 70,000 M,)_ The similarity of the holo- and apoenzyme gel patterns in Fig. 8 is obvious_ MuItipIe distinct bands are present in each, and the original geIs revealed that they can be separated into two groups of similar intensities. WhiIe one couId speculate as to the specific nature of each of the bands in terms of the dissocia-
K. G. KRUL AND C. R. DAIi’SON
84
0”
LO
LH
3
p”
10 ~ 3
FIG. 8. Showing the isoelectric focusing patterns of native (Gel l), apo- (Gel 21 and holoascorbate oxidase (Gel 3) on polyacrylamide gel columns.
tion of the quatemary structure of ascorbate o*dase, such speculation is not justified at this time_ The muitiple bands of the apo- and holoenzymes’ isoeIectric focusing patterns clearly indicate that these two variants of ascorbate oxidase maintain a similar subunit bonding nature, which is weaker than that of the native enzyme. As this analytical technique separates on the basis of pI alone, it is apparent that strong forces are not necessary to partially dissociate the subunit structure of both apoand holoascorbate oxidase. 4. Determina tion of the Tyrosineflryptophan of Native, Apo- and HoloAscorbate Ox&se
Molar Ratio
The Tyr/Trp MoIar Ratios of native, apo- and holoascorbate oxidase sampIes stored aerobically, at 4 O in 0.2 M Mcilvaines’s Buffer, pH 5.6, were determined as previously described_ Various samples of the enzyme were examined by this method while aging for up to 3 days_While the TyrlTrp Molar Ratios of native and holoascorbate oxidase remained constant and equal to 1.0, all apoenzyme samples showed a decline in this molar ratio with time, indicating a loss of ionizable tyrosine residues. The rate of decline, however, varied with the sample. Later assays (6 days) showed essentially no change from the lowest levels reached after 3 days. Figure 9 shows the differing rate of Tyr/Trp Molar Ratio decline for various apoascorbate oxidase sampIes, stored identically, as compared with the same molar ratio for the native enzyme_ As determined by Lee [6] for the native enzyme, 100% = a TyrlTrp Molar Ratio of 1 .O. In order to determine whether or not the treatment with cyanide (involved in the apoenzyme preparation) resulted in any chemical changes of the tyrosine residues avaiiable for ionization, bovine serum albumin samples were prepared in 0.2 M McIlvaine’s Buffer, pH 5.6. Half of the samples were treated wfth cyanide
CONCEFWING
THE METALLOENZYME
ASCORBATE
OXIDASE
85
FIG. 9. Showing the rates of Tyrosine/Tryptophan Molar Ratio decline with various samples of apoascorbate oxidase, as compared with the native enzyme (100% = a Tyr/Trp Molar Ratio of 1.0). as in the preparation of apoascorbate oxidase. The remainder were stored as controls, aerobically at 4 O_ After 3 days, fresh albumin samples were prepared_ The Tyr/Trp MoIar ratios of the aged, cyanide treated albumin; the aged, untreated albumin; and the fresh albumin solutions were then determined_ All samples were found to have essentially the same Tyr/Trp Molar Ratio. Thus, cyanide itself does not have an effect on the loss of ionizable tyrosine. Since there is cIearIy a loss of ionizable typrosine residues, a further possibility was explored. It is known that tryosine phenolic groups often become masked or buried by conformational changes in a protein, precluding their ionization [20,2 1] _ Samples of apoascorbate oxidase were prepared, ang their average Tyr/Trp Molar Ratio determined to be 1.0 1_ The samples were then stored aerobically at 4 o in 0.2 M McIIvaine’s Buffer, pH 5.6. After 3 days, the samples were reassayed, and their average Tyr/Trp Molar Ratio determined to be 0.70. The samples were then made 1% in dodecyl sulfate, and incubated for 2 hr at 40 O. At the end of the incubation period, the average Tyr/Trp Molar Ratio (as measured against a blankof 0.1 N NaOH/l% dodecyl sulfate in 0.2 M McIlviane’s Buffer, pH 5.6) was found to be 1.04. The molar ratio was completely restored by treatment with dodecyl sulfate, indicating that conformational changes occur
K. G. KRUL AND C. R. DAWSON
86
during the aerobic aging of apoascorbate oxidase which mask the previously ionizable tyrosine groups_ Furthermore, the data indicate that in the native and hoioenzymes, and in the fresh apoenzyme, the tryosine residues are externally located and availabIe to solvent. The authors wish to express their apprecian-on to Dr. Joseph Stauffer and Dr. Lu-ku Li for their help in obtaining the sedimentation data_
REFERENCES 1. 2. 3.
4
5_ 6_ 7. 8. 9. 10. 11..
12. 13. 14. 15. 16. 17. 18.
19. 20. 21.
C. R. Dawson, in Biochemisrry of Copper (J. Peisach, P. Aisen. and W. E. Blumberg, Eds.j. Academic Press, New York (1966?. pp- 305337. K. G. Strothkamp and C. R. Dawson, Biochem 13,434-440 (1974). &I. H. Lee and C. R. Dawson, J. BioL Chem 248,6596-6602 (1973). 2. G. Penton and C. R. Dawson, in Oxidases and Related Redox Sysrems (T- E. King, H. S. Mason, and M_ hforison, Eds.j. John Wiley and Sons, New York (1965). pp_ 22% 239_ H. T. Chang, Ph.D. Dissertation, Columbia University (1970). M. H. Lee. Ph_D. Dissertation, Columbia University (1968). 0. H. Lowry, N. 1. Rosebrough. A. L. Farr, and R. J. Randall, J. BioL Chem. 193, 265-272 (19.51). G. R_ Stark and C. R. Dawson, Anal. Chem 30,19I-I94 (1958). C. R. Dawson and R- J. hlagee, Methods Enzymol. 2,831-839 (19.57). H. Determann, Gel Chromatography. 2nd edn., Springer-Verlag, Inc., New York (1969), pp_ 105-119. L. Otnstein and B. J. Davis.Ann. IV. Y. Acad. Sci. 121,321-349;404-427 (1964). K. Weber and hi. Osbom, J. BioL Chem 244,4406-4412 (1969). S-Carrel. L.Theilakis, S. Skwuii. and S. Barandun,J. Chromatog. 45,483-488 (1969). W_ L_ Benae and K. Schmid, AnaL Chem 29,1193-l 196 (1957). K. G. Strothkamp, Ph.D. Dissertation, Columbia University (1973). H- Svensson, Aeta Chem Stand_ X5,325-334 (1966). 0. Vesterberg and H. Svensson, Acta Chem Stand_ 20,820-824 (1966). J. Brewer. Science lS6,256-258 (1967). H. K. Fantes and 1. G. S. Furminger, Nature (London) 215,750-753 (1967). J. Steinhardt and N. Stocker, Biochem 12,1789-l 797 (1973 j. J. Steinhardt and N- Stocker, Biochem 12,2798-2802 (1973).
Received 9 April 1976; revised 7 July I9 76
’
BRL?NORGAA.!IC CHEMISTR Y 7,7 l-86 ( 1977)
Concerning the Metalloenzyme Ascorbate Oxidase
KENNETH G. KRUL* and CHARLES R. DAWSON Department of Chemisny, Columbia University. New York, NY 10027
ABSTRACT Apoascorbate oxidase has been shown to have a molecular weight of 137,000 i: 3,000 and essentially the same gross quaternary conformation as native ascorbate oxidase. The apoenzyme, however, lacks much of the conformational stabihty of the native enzyme. The removal of the copper from the oxidase protein, and the simultaneous reduction of the disultide bonds results in an apoenzyme of lower structural stability than the native oxidase. The aging of apoascorbate oxidase has been found to involve a loss of ionizable tyrosine residues and a dissociation to subunits and component polypeptide chains, which was not observed with the more stable native and holoenzymes. The molecular weight of holoascorbate oxidase has been determined of 9.79 has been determined for the holoenzyme. to be 285,000. An s&U, Holoascorbate oxidase has been shown to have an electrophoretic mobility on polyacrylamide gels that is 23% lower than either the native or apoenzyme. Furthermore, electrophoresis of the holoenzyme in buffers containing dodecyl sulfate, and also isoelectric focusing of the holoenzyme. produce patterns of greater similarity to those of apoascorbate oxidase than the native enzyme. KEY WORDS Apoascorbate dase.
Osidase,
Ascorbate
Oxidase,
Holoascorbate
Oxidase,
Oxi-
INTRODUCTION
Ascorbate oxidase (L-Ascorbate: Oa oxidoreductase, EC 1.10. 3.3) is a copper-containing enzyme that catalyzes the aerobic oxidation of Vitamin C. Experimental evidence indicates that this enzyme has a molecular weight of 140,000 [I] and is made up of two identical subunits [2] _ Each subunit has a molecular weight of about 70,000 and is believed to consist of two polypeptide chains
of molecular
*Present address Hoechst Corporation,
weights
approximating
30,000
and address for correspondence: Somerville, NJ 08876.
and 40,000. Behring
Diagnositcs,
Q Elsevier North-Holland,
American
inc., 1977
K. G. KRUL AND C. R. DAWSON
72
Three distinct variants of ascorbate oxidase have been isolated, or produced, native ascorbate oxidase, apoascorbate oxidase, and holoascorbate oxidase. N~rive ascorbclte oxidase is the name given to that variant of the enzyme that has been isolated directly from its plant source. It is enzymatically active and contains 8-10 atoms of copper per molecule [3] _ The native oxidase is characterized by copper content, and by disulfide bond content, that have not been purposefully exposed (for experimental purposes) to conditions that result in copper loss or disulfide reduction. The native oxidase is stable over extended periods of time_ Dialysis against cyanide converts the active, copper-containing, native ascorbate oxidase to its inactive, copper-free apoenzyme, upoascorbate oxidase [4] _ Chang [S] has devised a practical method for the restoration of copper content and enzymatic activity to the apoenzyme. This restored variant of ascorbate oxidase is referred to as ~roloascorbare oxidase. It has been the purpose of the present investigation to examine and compare the properties and the structural features of these three variants of ascorbate oxidase.
in these laboratories:
MATERIALS
AND METHODS
I _ Enzyme Purification Ascorbate
oxidase was isolated
from the green zucchini
squash (Cucurbira
pepo nteduliosa) and purified as described by Lee and Dawson [3] _ 2. Chemicals acrylarnide (electrophoresis Sodium dodecyl sulfate, 2-mercaptoethanol, grade), N,N’-methylene-bisacrylamide, Tris and glycine (ammonia free) were obtained from Eastman Organic Chemical Company_ Bromphenol blue, Buffalo Black NBR, citric acid, cupric sulfate and ascorbic acid were purchased from Fisher Chemical Company. Coomassie Brilliant Blue R-250 was purchased from Schwarz/Mann. Arnpholine carrier ampholytes (40% in water) were purchased from LKB Produktur. BioGel A-I Sm and A-5m (loo200 mesh) were purchased from BioRad Laboratories, and Sephadex G-l 50 was obtained from Pharmacia Fine Chemicals. All aqueous solutions were prepared using water which had been filtered through a glass wool falter and passed through a Culligan D-25 mixed-bed deionizer (Nelson, Phillips and Co., Inc.)_ The copper content of this water was less than 0.005 ppm_ All chemicals were used without further purification.
CONCERNlNG
THE METALLOENZWE
ASCORBATE
OXIDASE
73
3. Protein Determination
The protein concentrations of all enzyme solutions were determined by the Lee modification [6] of the method of Lowry et al. [7] _ This modification is specific for the protein moiety of ascorbate oxidase. 4. Copper Determination Enzyme samples were treated with Chelex-100 (BioRad Laboratories) for 45 min to 1 hr and/or exhaustively dialyzed against copper-free buffer prior to the determination to remove essentially all nonenzymatic copper. Copper was then determined by the method of Stark and Dawson [8] _ 5, Activity Determination Activity was measured by the initial rate of oxygen uptake during the enzymatic oxidation of L-ascorbic acid, essentially as defined by Dawson and Magee [9]. Oxygen uptake was measured using a Gilson KM Oxygraph in a reactive system as defined by Strothkamp and Dawson [2] _ 6. Preparation of ApoAscorbate Apoascorbate
Oxidase
oxidase was prepared as described by Penton and Dawson [4] -
7. Preparation of HoloAscorbate
Oxidase
Fresh apoascorbate oxidase samples were restored to the active holoenzyme by the method of Chang [S], followed by centrifugation of the samples at 17,500 X g for 15-25 min, at 6-10 O. (The temperatures throughout are given in “C.) 8. Gel Filtration All columns were run at room temperature (25-27 “). Sephadex G-150 and BioGel A-1.5m (100-200 mesh) were used in experiments involving standard buffer conditions_ BioGel A-5m (100-200 mesh) was used in all experiments involving buffers containing sodium dodecyl sulfate_ Standard curves and molecular weight determinations were calculated in accordance with Determann [lo] _ 9. Sedimentation
Velocity Determination
The sedimentation coefficients of holoascorbate oxidase samples were determined on a Beckman Model E uhracentrifuge. The sedimentation constant was determined by 20 p with a rotor speed of 60,000 r-pm. 10. Gel Electrophoresis Polyacrylamide disc electrophoresis experiments were carried out essentially in accordance with the method of Ornstein and Davis [1 1] _ Experiments in
K_ G_ KRUL AND C_ R_ DAWSON
74 these laboratories
varied from the published procedure
in that sample gels were
never used, and, in instances where the volume of the protein sample was small, stacking gels were omitted. 11. Treatment
of Protein with Dodecyl Sulfate
Protein standards and samples of ascorbate oxidase, which were later examined by either gel electrophoresis or fel filtration in buffers containing sodium dodecyl sulfate, were incubated prior to analysis according to Strothkamp and Dawson [2] _ 12. Molecular Weight Determinations
by Gel Electrophoresis
Molecular weight determinations of the oxidase samples were made essentially by Weber and Osbom [12] _ Rather than the published procedure, instead, 7.5% acrylamide 13. Isoelectric
dissociation patterns of ascorbate the electrophoretic technique of 10% acrylamide gels used in the gels were used in this investigation_
Focusing
Isoelectric focusing experiments procedure of Carrel et al. [ 13]_ 14. Tyrosineflryptophan
were
carried
out
accord
with
the
Molar Ratio Determination
The TyrosinelTryptophan Molar Ratios of ascrobate determined by the method of Bencze and S&mid [ 14]_ RESULTS
in
oxidase
samples were
AND DISCUSSION
I _ Gel Filtration
and Sedimentation
Experiments
Fresh apoascorbate oxidase samples (lo- 15 mg), prepared from homogeneous native enzyme preparations. were subjected to gel filtration on Sephadex G-l 50. Figure 1 shows a typical elution pattern for the apoenzyme. As can be seen, the apoenzyme (second peak) was indicated to be largely homogeneous, with some evidence of the presence of minor dissociation components_ The molecular weight (Icr,) of the apoenzyme has been calculated from gel filtration data to be 137,000 t 3,000, which is in good agreement with that calculated for native ascorbate oxidase [ 1,2] _ Equal amounts of native and apoascorbate oxidase (1 O- 15 mg) were mixed and subjected to gel filtration on Sephadex G-150. Native ascorbate oxidase binds its copper tightly and there is no exchange of copper ions between the native and apoenzymes [2]. Figure 2 shows a typical elution pattern for this experiment_ As can be seen, native and apoascorbate oxidase migrate together_ The M, calculated for the peak representative of the two ascorbate oxidase
CONCERNiNG
THE METALLOENZYME
ASCORBATE
OXIDASE
0.6-
-O.IZ
OS-
r
B 0 :0.4-
0.10
-
0.06
-
0.06
Y
-
0.04
Y 2 E
-
0.02
E -
*z.onm
_---
2 “, 0.3
-
*x60
I
-
nm
z
4’ P 5: 0.2-
g
3
t 0.1
-
I
I 20
1
‘ 30
I 40
I
t
FRACTION
I
L 60
I 60
I
I
I
70
’
NUMBER
FIG. 1. A typical elution pattern of fresh apoascorbate oxidase in 0.2 M phosphate buffer, pH 8.0, on a Sephadex G-l 50 column_ The first peak (from left to right) is that of blue dextran, the second is the apoenzyme, and the third is that of dinitrophenylalanine.
1.4
J
l-
0.35
7
O-B-
0.05
1 30
I 35
1 40
1 46 FRACTION
8 50
8 65
1 60
1 65
70
NUMBER
FIG. 2. A typical elution pattern of a 1: 1 mixture of native and apoascorbate oxidase in 0.2 M phosphate buffer, pIi 8.0, on a Sephadex G-150 column. The first peak (from left to right) is that of blue dextran, and the dotted peak is that of dinitrophenylalanine. The peaks falling between the blue dextran and dinitrophenylalanine are due to the ascorbate oxidaselapoascorbate oxidase mixture.
K. G. KRUL
76
AND C. R. DAWSON r - O-8
0-B
~ 35
50
45
FRACTION
FIG. 3. A typical phosphate buffer,
elution
pattern
55
60
65
70
75
NUMBER
of apoascorbate 1% in dodecyl sulfate,
oxidase
in 0.2
M sodium
on a BioGel A-Sm column. The first peak (from left to right) is that of blue dextran, and the last is that of dinitrophenylalanine. The peaks occurring between these are due to the six dissociation species of the apoenzymepH 7.0,
variants migrating together is 143,000 t 1 ,OOO_The three minor peaks eluted after the native/apoenzyme mixture have calculated M,‘s of 108,000, 89,000, and 68,000 t 2,000. These peaks can be accounted for on the basis of the dissociation of the quaternary structure of the protein moiety of ascorbate oxidase, as described by Strothkamp and Dawson [2] _The total presence of these peaks, however, comprises less than 10% of the protein applied to the column. While the minor peaks indicate the lesser quatemary stability of apoascorbate oxidase (they are not observed with native ascorbate oxidase alone[2] ), they do not substantiate any major conformational difference between the native and apoenzyme molecules_ In order to better observe the dissociation of apoascrobate oxidase, samples of the apoenzyme were incubated according to the method of Strothkamp and Dawson [2], and subsequently subjected to gel filtration on BioGel A-5m. Figure 3 shows a typical elution pattern for the treated apoascorbate oxidase. The protein peaks correspond to M,‘s of 138.000; 80,000; 70,000; 57,000; 47,000; and 28,000, respectively. All of these peaks can be accounted for on the basis of simtilar patterns resulting from-like treatment and dissociation of the quatemary structure of native ascorbate oxidase [2] _ Thus it can be concluded that, on the basis of the above gel filtration experiments_ apoascorbate oxidase, while somewhat less stable than the native enzyme, maintains the same molecular weight and essentially the same gross quaternary conformation as native ascorbate oxidase.
CONCERNING
THE METALLOENZYME
ASCORBATE
OXIDASE
0.6 -
77 -
0.12
-
0.10
: s
-
0.06
* s Y
-
0.06
,”
-
0.04
5 ,”
-
0.02
2
a
1
I 20
1
30
1
40 FRACTION
1
50
1
60
L
I
70
NUMBER
FIG. 4. The elution pattern of holoascorbate oxidase in 0.2 M McIlvaine’s Buffer, pH 5.6, on a BioGel A-1Sm column. The first peak (from left to right) is that of blue dextran, the second is that of the holoenzyme, and the third is dinitrophenyIaianine. The holoenzyme solutions were concentrated by ultrafiltration to approximately 0.75 ml prior to loading on the column. GeI fihration of fresh holoascorbate oxidase was carried out on a BioGel A-1Sm cohrmn. Figure 4 shows a typical ehrtion pattern observed in the above experiment_ The molecuIar weight (M,) of the holoenzyme has been calculated from the gel filtration data to be 285,000. This is approximately twice the molecular weight of either native or apoascorbate oxidase. The sedimentation constants of hoIoascorbate oxidase samples were determined on a Beckman Model E analytical ultracentrifuge. The &!o,~ for the holoenzyme was found to be 9.79. This may be compared with the ~go,~ for the native ascorbate oxidase of 7.52 161. The increase in the sedimentation constant of the hoIoenzyme over that of the native enzyme supports the finding that the M, of hoIoascorbate oxidase, as determined by gel filtration, is twice that of either the native or apoenzyme. Apparently the reconstitution process causes the protein molecuIes of ascorbate oxidase to dimerize, thus forming a tetramer of subunits. As indicated by the symmetry of the peaks resulting during gel filtration and ultracentrifugation, this new, higher molecular weight variant is a homogeneous entity. Although the quatemary structure of the holoenzyme is different from that of the native enzyme, it is still catalyticaliy active. On comparing the specific activities of numerous holoascorbate oxidase samples with those of their corresponding native ascorbate oxidase samples (See Table l), it has always been observed that the holoenzyme is more active than the native enzyme. Thus, the distinct character of holoascorbate oxidase is clearly apparent_
K. G. KRUL AND C. R. DAWSON
78 TABLE
1
Specific Activities of Native Ascorbate Oxidase Samples and Their Corresponding Holoenzymes
Sample
Spec. activity (U/mg protein)
Spec. activity (Uhg
copper)
NatAO-1 HoloAO-I
1,368 1,890
305 475
NatAO-2 HoloAO-2
2,175 3,154
733
NatAO-3 NoloAO-3
1,476 4.220
591 839
NatA HoloAO4
2,900 4,326
439 886
Nat AO-5 HoloAO-5
2,490 3,956
838 1,049
NatAO-6 HoIoAO-6
2,490 3,809
838 1,021
NatAO-7 HoIoAO-7
2,490 3,568
838 1,017
NatAO-8 HoloAO-8
2,490 3,231
838 1,099
NatAO-9 HoloAO-9
2,490 3,375
838 1,099
573
2. Gel Electrophoresis
Samples of native, apo- and holoascorbate oxidase were examked by disc electrophoresis. The electrophoretic mobilities of the samples versus an internal standard (bromphenol blue) were calculated as described by Weber and Osbom [I21 _ Figure 5 illustrates the typical gel patterns demonstrated with the three variants of ascorbate oxidase.
CONCERNING
THE METALLOENZYME
I 1
ASCORBATE
-
-
2
3
OXIDASE
FIG. 5. Showing the results of polyacrylamide gel electrophoresis of native (Gel 11, apo- (Gel 2) and holoascorbate oxidase (Gel 3).
79
at pH ‘3.5
While native and apoascorbate oxidase were observed to have essentially the same mobility (0.298 5 O-003), that of the holoenzyme was observed to be significantly (23%) lower (0229 + 0.002). Furthermore, holoascorbate oxidase exhibited one, uniform electrophoretic band, similar to that of the native enzyme. Both of these gel patterns were unlike the diffuse, multibanded and indistinct gel pattern of apoascorbate oxidase that has been observed in this investigation and previously [S] _ The single-band, electrophoretic pattern of the holoenzyme further attests to its homogeneity. The lower electrophoretic mobility of holoascorbate oxidase, as compared with its native and apoenzymes, is entirely consistent with the finding of increase molecular weight and sedimentation constant of holoenzyme preparations_ Samples of the native, fresh apo- and holoenzymes were treated with dodecyl sulfate and 2-mercaptoethanol, and analyzed by gel electrophoresis in buffer containing dodecyl sulfate- Table 2 and Fig. 6 illustrate the results of these experiments_ As can be seen, hoioascorbate oxidase exhibits a dissociation pattern more closely resembling that of the apoenzyme than that of the native enzyme_ This can be taken as an indication that the holoenzyme, it its two-fold native conformation, maintains its quatemary structure less firmly than native ascorbate oxidase, and more on the order of the apoenzyme. Additionally, the sulfhydryl
K. G. KRUL AND C. R. DAWSON
80 TABLE 2
Molecular Weight Analyses of Native, APO- and HoloAscorbate Native
71*** 59 50.5 40.5
Oxidase”
Apo-
Holo-
107 81** 69 62 ss* 53.5 47 39.5
108 83.5** 70.5 65 59.5* 56 47.8 41.5
a Values given are those of the molecular weights of bands appearing on the polyacryIamide gels (See Fig_ 6). Asterisks are used to denote the relative intensities of the bands on the gels (*** more than **, ** more than *) and the intensities are relative to the other bands on that particular vertical column. Values not suffixed by an asterisk were considered to be of minor importance due to their weak intensities.
-
1. -
f! -
-
2
3
FIG. 6_ Showing the results of polyacrylamide gel electrophoresis of native (Gel I), apo- (Gel 2) and holoascorbate oxidase (Gel 3), in the presence of dodecyl sulfate and 2-mercaptoethanol. The lower three bands exhibited on Gel I are extremely faint, and are considered insignificant.
CONCERNING THE METALLOENZYME TABLE
ASCORBATE OXIDASE
81
3
Dissociation Species of ApoAscorbate
OxidaseO
(Molecular weights in thousands)
Day I
87-O***
Day II
s3_0***
Day 15
87-o*** 78.0
Day 20
97.5 87-O*** 71.5
53.0*** 42-b**
43-o***
46.0***
29.2***
29-o*:**
31.0***
16.5
16.0
Day 83 113 102 94.8 94.2* 72.0***
66-o*** 58.8 43.0* 38.5 29.1* 24.2 20.2 16.2
L2Results of a time study using polyacrylamide gel electrophoresis in the presence of dodecyi sulfate to observe the stability of aging ApoAscorbate Oxidase. Asterisks are used to denote the relative intensities of the band (*** more than **, ** more than *), and the intensities are relative to the other bands on that particular vertical column. Values not suffixed by an asterisk were considered to be of minor importance due to their weak intensities_
groups of the holoenzyme appear to be more easily accessible to reduction by 2mercapto-ethanol than in the native enzyme, as evidenced by the significance of bands having moIecuIar weights other than that of the major subunit, i.e., approximately 70,000. While the gel patterns of native and holoascorbate oxidase were essentially the same in all samples treated with dodecyl sulfate, the gel patterns of the dodecyl sulfate-treated apoenzyme changed with the age of the apoenzyme sample. The aging pattern of apoascorbate oxidase was therefore investigated by this technique_ Samples of apoascorbate oxidase were stored aerobically, under refrigeration (4 “) in 0.2 M McIlvaine’s Buffer, pH 5.6, for periods of up to 80 days. At various iime intervals, ahquots of the apoenzyme were taken and examined by ge1 electrophoresis in buffer containing dodecyl sulfate. Table 3 shows the molecular weight variations exhibited at significant points in the aging of apoascorbate oxidase. The consistently reoccurring bands indicate that the most stable dissociation species of apoascrobate oxidase have molecular
82
K. G. KRUL AND C. R. DAWSON
I
P
FIG. 7_ The suggested quaternary structures of ascorbate oxidase as supported by evidence in this investigation (I), and as previously proposed (II). A is a polypeptide chain of fir, approximately 30,000 and B is a polypeptide chain of Mr approximating 40,000_ The disulfide bonds shown (-SS-) are the minimum number necessary to explain the experimental data. Electrostatic forces and hydrogen bonds are represented by dashed lines (----).
weights that average 85,600,43,000, and 29,900. In extremely fresh apoenzyme samples, only two species are observed, having molecular weiets of 87,000 and 53,000_ Wirh the aging of the apoenzyme samples, the gel patterns grow increasingly complex. This increased compIexity was not due to bacterial growth, becauSe native ascorbate oxidase samples, stored similarly for the same periods of time, do not show this effect. The results of these experiments indicate that the quaternary composition of the apoenzyme is the same as that of the native enzyme [2]. However, the data presented here tend to support an asymmetrical structure (Fig_ 7, I) rather than the symmetrical structure (Fig. 7, Ii) as proposed by Strothkamp [IS] _ As has been observed previously [2J, and confirmed in this investigation, native ascorbate oxidase yields only one major band (Mr approximating 70,000) during gel electrophoresis in buffer containing dodecyl sulfate_ When the disulfide bonds of the protein were reduced after dissociation of the subunits by dodecyl sulfate, the gel pattern showed two significant bands of M,‘s approximating 30,000 and 40,000. Structure II (Fig. 7) is entirely consistent with this data. In the preparation of the apoenzyme, however, the disulfide bridges were initially reduced by cyanide. The dissociation by use of dodecyl sulfate takes place sllbsequent to the cyanide reduction. it is thus possible to observe dissociation species corresponding to dimers of like polypeptide chains, i.e., M,‘s approximating 60,000 and 80,000. As interaction between like polypeptide
CONCERNING
THE METALLOENZYME
ASCORBATE
OXIDASE
83
chains, resulting in such “dirnerization” would appear to be maximized by their juxtaposition, Structure I (Fig. 7) would appear to be a better model for the quatemary arrangement of the subunits of ascorbate oxidase. The asymmetrical model is consistent with the previous data 12,151 and that obtained in this investigation_ It is interesting to note that, at the end of 83 days, the half-molecular weight subunit (kfr = 72,000) is the predominant species present (See Table 3). This is particularly interesting in light of the fact that in younger samples of apoascorbate oxidase it is not present, or present only in insignificant quantities on the gels. The apparent recombination of polypeptide chains and the reformation of the disulfide bonds are indicative of the thermodynamic stability of the halfmolecular weight subunit moiety, while the increasing complexity of the gei patterns indicate the decreasing stability of the apoprotein moiety as a whole. 3. Isoelectric
Focusing
Samples of native, fresh ape-, and holoascorbate oxidase were repeatedly analyzed by isoelectric focusin g on polyacrylamide gel columns_ In each case, enough protein solution was added to the gel mixtures, such that the final concentrations of the various enzyme variants were approximately equal in terms of protein. The ampholyte-induced pH gradient ranged from pH 3 to pH 10 at 15 ’ _ Figure 8 illustrates the results obtained in a typical experiment. It can be seen that the isoelectric focusing pattern of the hoIoenzyme (Gel 3) more cIosely parallels that of the apoenzyme (Gel 2) than that of the native enzyme- This is particularly interesting due to the fact that isoeIectric focusing separates moIecules soIeIy on the basis of their isoelectric point (PI). The gel serves merely to stabilize the ampholyte-induced pH gradient, and does not function as a molecular seive [ 16,17]_ Native ascorbate oxidase, shown to be homogeneous by gel electrophoresis, was indicated to contain two very similar components in respect to pi by isoelectric focusing_ The secondary band was observed to be very close to the primary band (See Fig. 8, Gel 1) and may be an artifact. Such artifacts have been observed by others studying different proteins [18,19]_ Or, since its pi is essentially that of the native enzyme, the secondary component may well be simpIy some native protein that has lost some of its copper during the poIymerization of the ge1. It is also possible that the secondary component seen by isoelectric focusing may be a small amount of dissociated, half-molecuIar weight subunit (approximately 70,000 M,)_ The similarity of the holo- and apoenzyme gel patterns in Fig. 8 is obvious_ MuItipIe distinct bands are present in each, and the original geIs revealed that they can be separated into two groups of similar intensities. WhiIe one couId speculate as to the specific nature of each of the bands in terms of the dissocia-
K. G. KRUL AND C. R. DAIi’SON
84
0”
LO
LH
3
p”
10 ~ 3
FIG. 8. Showing the isoelectric focusing patterns of native (Gel l), apo- (Gel 21 and holoascorbate oxidase (Gel 3) on polyacrylamide gel columns.
tion of the quatemary structure of ascorbate o*dase, such speculation is not justified at this time_ The muitiple bands of the apo- and holoenzymes’ isoeIectric focusing patterns clearly indicate that these two variants of ascorbate oxidase maintain a similar subunit bonding nature, which is weaker than that of the native enzyme. As this analytical technique separates on the basis of pI alone, it is apparent that strong forces are not necessary to partially dissociate the subunit structure of both apoand holoascorbate oxidase. 4. Determina tion of the Tyrosineflryptophan of Native, Apo- and HoloAscorbate Ox&se
Molar Ratio
The Tyr/Trp MoIar Ratios of native, apo- and holoascorbate oxidase sampIes stored aerobically, at 4 O in 0.2 M Mcilvaines’s Buffer, pH 5.6, were determined as previously described_ Various samples of the enzyme were examined by this method while aging for up to 3 days_While the TyrlTrp Molar Ratios of native and holoascorbate oxidase remained constant and equal to 1.0, all apoenzyme samples showed a decline in this molar ratio with time, indicating a loss of ionizable tyrosine residues. The rate of decline, however, varied with the sample. Later assays (6 days) showed essentially no change from the lowest levels reached after 3 days. Figure 9 shows the differing rate of Tyr/Trp Molar Ratio decline for various apoascorbate oxidase sampIes, stored identically, as compared with the same molar ratio for the native enzyme_ As determined by Lee [6] for the native enzyme, 100% = a TyrlTrp Molar Ratio of 1 .O. In order to determine whether or not the treatment with cyanide (involved in the apoenzyme preparation) resulted in any chemical changes of the tyrosine residues avaiiable for ionization, bovine serum albumin samples were prepared in 0.2 M McIlvaine’s Buffer, pH 5.6. Half of the samples were treated wfth cyanide
CONCEFWING
THE METALLOENZYME
ASCORBATE
OXIDASE
85
FIG. 9. Showing the rates of Tyrosine/Tryptophan Molar Ratio decline with various samples of apoascorbate oxidase, as compared with the native enzyme (100% = a Tyr/Trp Molar Ratio of 1.0). as in the preparation of apoascorbate oxidase. The remainder were stored as controls, aerobically at 4 O_ After 3 days, fresh albumin samples were prepared_ The Tyr/Trp MoIar ratios of the aged, cyanide treated albumin; the aged, untreated albumin; and the fresh albumin solutions were then determined_ All samples were found to have essentially the same Tyr/Trp Molar Ratio. Thus, cyanide itself does not have an effect on the loss of ionizable tyrosine. Since there is cIearIy a loss of ionizable typrosine residues, a further possibility was explored. It is known that tryosine phenolic groups often become masked or buried by conformational changes in a protein, precluding their ionization [20,2 1] _ Samples of apoascorbate oxidase were prepared, ang their average Tyr/Trp Molar Ratio determined to be 1.0 1_ The samples were then stored aerobically at 4 o in 0.2 M McIIvaine’s Buffer, pH 5.6. After 3 days, the samples were reassayed, and their average Tyr/Trp Molar Ratio determined to be 0.70. The samples were then made 1% in dodecyl sulfate, and incubated for 2 hr at 40 O. At the end of the incubation period, the average Tyr/Trp Molar Ratio (as measured against a blankof 0.1 N NaOH/l% dodecyl sulfate in 0.2 M McIlviane’s Buffer, pH 5.6) was found to be 1.04. The molar ratio was completely restored by treatment with dodecyl sulfate, indicating that conformational changes occur
K. G. KRUL AND C. R. DAWSON
86
during the aerobic aging of apoascorbate oxidase which mask the previously ionizable tyrosine groups_ Furthermore, the data indicate that in the native and hoioenzymes, and in the fresh apoenzyme, the tryosine residues are externally located and availabIe to solvent. The authors wish to express their apprecian-on to Dr. Joseph Stauffer and Dr. Lu-ku Li for their help in obtaining the sedimentation data_
REFERENCES 1. 2. 3.
4
5_ 6_ 7. 8. 9. 10. 11..
12. 13. 14. 15. 16. 17. 18.
19. 20. 21.
C. R. Dawson, in Biochemisrry of Copper (J. Peisach, P. Aisen. and W. E. Blumberg, Eds.j. Academic Press, New York (1966?. pp- 305337. K. G. Strothkamp and C. R. Dawson, Biochem 13,434-440 (1974). &I. H. Lee and C. R. Dawson, J. BioL Chem 248,6596-6602 (1973). 2. G. Penton and C. R. Dawson, in Oxidases and Related Redox Sysrems (T- E. King, H. S. Mason, and M_ hforison, Eds.j. John Wiley and Sons, New York (1965). pp_ 22% 239_ H. T. Chang, Ph.D. Dissertation, Columbia University (1970). M. H. Lee. Ph_D. Dissertation, Columbia University (1968). 0. H. Lowry, N. 1. Rosebrough. A. L. Farr, and R. J. Randall, J. BioL Chem. 193, 265-272 (19.51). G. R_ Stark and C. R. Dawson, Anal. Chem 30,19I-I94 (1958). C. R. Dawson and R- J. hlagee, Methods Enzymol. 2,831-839 (19.57). H. Determann, Gel Chromatography. 2nd edn., Springer-Verlag, Inc., New York (1969), pp_ 105-119. L. Otnstein and B. J. Davis.Ann. IV. Y. Acad. Sci. 121,321-349;404-427 (1964). K. Weber and hi. Osbom, J. BioL Chem 244,4406-4412 (1969). S-Carrel. L.Theilakis, S. Skwuii. and S. Barandun,J. Chromatog. 45,483-488 (1969). W_ L_ Benae and K. Schmid, AnaL Chem 29,1193-l 196 (1957). K. G. Strothkamp, Ph.D. Dissertation, Columbia University (1973). H- Svensson, Aeta Chem Stand_ X5,325-334 (1966). 0. Vesterberg and H. Svensson, Acta Chem Stand_ 20,820-824 (1966). J. Brewer. Science lS6,256-258 (1967). H. K. Fantes and 1. G. S. Furminger, Nature (London) 215,750-753 (1967). J. Steinhardt and N. Stocker, Biochem 12,1789-l 797 (1973 j. J. Steinhardt and N- Stocker, Biochem 12,2798-2802 (1973).
Received 9 April 1976; revised 7 July I9 76
