Identification and purification of a non-ceruloplasmin ferroxidase of human serum.

Abstract A non-ceruloplasmin ferroxidase (ferroxidase-II) protein was isolated from human serum and completely resolved from ceruloplasmin by DEAE-Sephadex A-50 chromatography. This protein was isolated and purified approximately 500-fold by utilization of the Cohn IV-1 fraction, DEAE-Sephadex A-50 chromatography, and gel filtration on Agarose A-15m, followed by gel filtration on Agarose A-50m. Ferroxidase-II differed from ceruloplasmin in many respects. Three of the most important differences were that ferroxidase-II (a) was yellow rather than blue as ceruloplasmin, (b) it was not inhibited by azide, and (c) it exhibited no p-phenylenediamine oxidase activity. Ferroxidase-II also differed significantly from the known plasma amine oxidases. The ferroxidase activity of ferroxidase-II was not lost by dialysis or ultrafiltration, but was inactivated by heat treatment and was proportional to the protein concentration at all stages of purification. Oxygen consumption with ferroxidase-II was observed simultaneously with iron oxidation. The ferroxidase-II activity in Wilson's disease serum represented a much larger percentage of the total ferroxidase activity than in normal serum. Although ferroxidase-II was decreased in Wilson's disease serum, it was reduced to a lesser extent than the ceruloplasmin ferroxidase activity. Thus ferroxidase-II may account for the lack of correlation of ferroxidase activity with p-phenylenediamine oxidase activity of Wilson's disease serum and may be responsible for the maintenance of near normal iron metabolism despite the low levels of ceruloplasmin.

A non-ceruloplasmin ferroxidase (ferroxidase-II) protein was isolated from human serum and completely resolved from ceruloplasmin by DEAE-Sephadex A-50 chromatography. This protein was isolated and purified approximately 500-fold by utilization of the Cohn IV-1 fraction, DEAE-Sephadex A-50 chromatography, and gel filtration on Agarose A-15m, followed by gel filtration on Agarose A-50m.
Ferroxidase-II dBered from ceruloplasmin in many respects. Three of the most important dserences were that ferroxidase-II (a) was yellow rather than blue as ceruloplasmin, (b) it was not inhibited by azide, and (c) it exhibited no p-phenylenediamine oxidase activity.
Ferroxidase-II also differed significantly from the known plasma amine oxidases. The ferroxidase activity of ferroxidase-II was not lost by dialysis or ultrafiltration, but was inactivated by heat treatment and was proportional to the protein concentration at all stages of purification.
Oxygen consumption with ferroxidase-II was observed simultaneously with iron oxidation. The ferroxidase-II activity in Wilson's disease serum represented a much larger percentage of the total ferroxidase activity than in normal serum.
Although ferroxidase-II was decreased in Wilson's disease serum, it was reduced to a lesser extent than the ceruloplasmin ferroxidase activity. Thus ferroxidase-II may account for the lack of correlation of ferroxidase activity with p-phenylenediamine oxidase activity of Wilson's disease serum and may be responsible for the maintenance of near normal iron metabolism despite the low levels of ceruloplasmin.
The catalysis of the oxidat'ion of Fe(I1) to Fe(II1) by ceruloplasmin results in an increased rate of formation of transfer& from apotransferrin (l-3). This observation led to the proposal for a biological role of ceruloplasmin in promoting the rate of iron satur&ion of apotransferrin and in stimulating iron utilization (4). Later Osaki and Johnson (5), Osaki et al. (6), Frieden and Osaki (7), and Rngan et nl. (8), provided strong evidence obtained from both systems in viva and in vitro which supported * This work was supported in part, by United States Public Health service Grant f1E 08341 from the National Heart Institute. This is Paper No. 37 from this laboratory on Copper Bio-Systems.
: National Institutes of Health postdoctoral research fellow on Grant 1 F02 ITE 44,946.01. the idea that ceruloplasmin functions as a serum ferroxidase in promot'ing Fe(III)-transferrin formation. If iron metabolism is dependent on ceruloplasmin, we might expect a disturbance in iron metabolism in Wilson's disease, a disorder characterized by low plasma ceruloplasmin and the accumulation of copper in the liver and brain. Ho\vever, most Wilson's disease subjects have been found to have low normal or normal levels of iron t'ransport (7). In attempt,ing to correlate the aryldiamine oxidase activity with t,he fcrroxidasc activity of sera from Wilson's disease patients, it, ins observed that these sera had more ferroxidase than expected (7,9). The oxidat,ion of Fe(I1) by serum components other than ceruloplasmin has been previously observed (10, 11). Lee et al. (11) have proposed citrate as an alternative source of ferrous ironoxidizing activity in low ceruloplasmin serum. Friedcn and Osaki (7) suggested the possibility of an alternative prot>cin ferroxidase, quite different from ceruloplasmin, which might substitut'e for the normal enzyme in Wilson's discaqc serum.
The present paper describes the identification ant1 isola,tion of an alternative protein ferroxidase of normal human serum which differs greatly from ceruloplasmin.
The isolation, purificatioll, and partial characterization of this non-cerulol~lns~~~jn ferroxidase (ferroxidase-IQ1 protein from the Cohn IV-I fraction of strum is described.

EXPERIUENTAL PROCEIlURE
SeruTn-Fresh samples of 5 ml of whole blood from many donors were collected and allowed to clot. The strum was separated by ccntrifugation, stored at 4", and ust,cl within scvcral hours. Alternatively, outdated blood way obtniucd from the Leon County Blood Bank, Tallahassee, Florida.
The plasma was separated by centrifugation and the citrate which had been added to prevent clotting was removed by dialysii against 0.1 in acetate buffer, ~1-1 6.0. -5ny precipitate formed during the dialysis procedure was removed by centrifugation. These samples had essentially the same ferroxidase-II activity as fresh serum samples.

Jfaterials-DEAESephadex
A-50 (Pharmacia) was prepared with acetate as the count'er ion and was equilibrated with 0.05 11 acetate buffer, pH 5.5. Agarose A-15m and Agarose A-50m (Calbiochem), obtained in fully hydrated form, were equilibrated with 0.05 M acetate buffer, pH 5.5, prior to use.

Enzymic
Assays-The ferroxidase assays were carried out spcctrophotometrically in a manner similar to that previously described (9). Quartz spectrophotometric cuvettes, 1.5 ml in capacity and 1 cm in path length, cont'ained 0.340 ml of a 0.6 N acetate buffer, pH 6.0; 0.250 ml of the 2y0 (w/v) apotransferrin solution; 0.300 ml of the 1 X lop4 ~1 ferrous ammonium sulfate solution, 10 to 100 ~1 of distilled, doubly deionized water, and 10 to 100 ~1 of the appropriate enzyme preparation so that the final reaction volume was 1.0 ml. The final acetate buffer concentration in the reaction mixture was 0.2 M and the final Fe(H) concentration was 30 pM.
The time course of absorbance change at 460 nm resulting from the formation of transferrin from apotransferrin and Fe(II1) was recorded with a Cary 15 spectrophotomet,er equipped w&h a 0.1 absorbance scale and a constant temperature cell at 30.0 f 0.1". A control containing no enzyme solution was included with each assay to determine the nonenzymic rate of Fe(I1) oxidat'ion in the reaction mixture. This value was always subtracted from the observed enzrmic rate to obtain the t'rue cnzymic rate. In all assays, the value for the nonenzymic rate of iron oxidat'ion was less than 5 yc of the enzymic rate.
Since 1.0 ml1 azicle inhibits cerulol~lnsn~in more than 98y0 and does not inhibit ferroxidase-II (see "Results"), the ferroxidase activity of each component of human serum and the Cohn IV-1 extract could be determined by conduct'ing the ferroxidase assa,ys in the pre?cnce and absence of 1 ml azide. Two microliters of a 0.50 JI xaS3 solution were added prior to the addition of the enzyme preparation t,o the reaction mixture. hlternatively, the enzyme preparation and nll other components of the reaction mixture except substrate, Fe(II), were initially incubated for 5 mill with azide. The reaction was initiated by the addition of substrate.
No difference in the rate of Fe(H) oxidation was observed with either order of addition.
Disc Gel a& I,?~,,zunoelectrophoresis-Disc gel electrophoresis was done with a 32 large pore polyacrylamide gel as previously described (13). The gels were subjected to 3 ma per gel for 90 min at pH 8.8, 0.025 nr Tris glycine buffer. hfter electrophoresis the gels were stained with Cooma3>ie Blue as described by Chrambach et al. (14).
Immunoclectrophoresis was carried out as previously described (15) except for one modification.
The electrophoresis was con-

Ident$cation
and Isolation of Ferroklase-II from Iluman Serum Differential sensitivity to 1 rnM azide was chosen as an initial probe to determine whether any ferroxidaac-II activity might, be present in normal human and Wilson's disease serum. Fcrroxidase-II activit,y was found to be present in these sern (' Table I). Ferroxidase-II act'ivit'y was not affect'ed by eit,her type of dialysis. However, the ceruloplasmin activity was considerably decreased by dialysis at low buffer concentrations at pH 5.5. Kilson's disease serum also contained ferroxidnse-II. Although both ferroxidase activities are lowvcr in these \1?lson's disease Sara, the total ferroxidase activity was decrensed by a factor of 17 to 18, whereas that of ferroxidase-II was decreased by a factor of only 3 to 4. Thus in Wilson's disease serum ferroxidasc-II accounts for a larger percentage (-30%) of the total ferroxidasc activity, about a 5-fold increase over fresh undialyzed serum from normal subjects.
Sane of these fractions contained either ferrosidnse acGvity. chloride (Fig. 1). Ferroxidase-II protein was eluted from the column immediately followin, w the void volume with 0.05 M acetate buffer, pH 5.5, containing no sodium chloride. This ferroxidase-II protein showed no difference in activity when tested in the presence or absence of 1 m&r azide. No ceruloplasmin was eluted from the column until the sodium chloride concentration in the buffer was increased to 0.20 M. Most of the ceruloplasmin was eluted when the sodium chloride concentration KVRS increased to 0.3 nr, a result typically observed in the preparation of ccruloplasmin (17). The ferroxidase activity exhibited by the ceruloplasmin-containin g fractions IT-as virtually completely ( >98%) inhibited by 1 mhr &de.
Thus the two fer-roxidase proteins were completely resolved by DEAE-Sephadex A-50 chromatography of human serum and were further distinguished by the inability of azide to inhibit fcrroxidase-II. Various lyophilized Cohn fractions of human serum were examined to determine whether the ferroxidase-II activity was present (Table II).
Two fractions contained this activity, Cohn IV-l and Cohn III-O.
No loss of ferroxidnse-II activity from either of these fractions was observed on extensive dialysis against 0.10 RI acetate buffer, pH 6.0. Cohn IV-l contained the greatest percentage of the ferroxidase-II activity relative to the total ferroxidase activity and was used as the starting material for the purification of ferroxidase-II. Cohn IV-1 powder (6 g) was extracted with 60 ml of 0.05 M aceMe buffer, PI-I 5.5, for 4 hours at 4" with stirring. The resulting extract was centrifuged at 25,000 x g for 30 min to remove any undissolved materials.
The supernatant from this centrifugation was decanted carefully and dialyzed in the cold for 24 hours with a large volume of 0.05 XI ncet'ate buffer, pH 5.5. The buffer was changed several times during the course of the dialysis.
The dialyzed Cohn IV-1 extract was recentrifuged (25,000 x g, 30 min) to remove any material which precipitated during the dialysis. The Cohn IV-1 fractionation and lyophilization apparently resulted in a 70.fold purification of ferroxidase-II as compared to whole human serum (Table III).
Of this extract, 2 ml were chromatographed on a DEAE-Sephadex A-50 column (1.5 x 6.5 cm). The clution pattern of the ferroxidase activities 1~s identical with that of fresh whole human serum. The ferroxidase-II protein was clukd immediately following the void volume with 0.05 JI ncctatc buffer, pH 5.5, containing no sodium chloride.
The ferrosidase activity of this component from rohn IV-1 also was not inhibited by 1 111~ azide. Ceruloplasmin activity began to appear in fractions containing 0.2 M Sac1 in the acetate buffer; howcvcr the major portion was found in the fractions containing 0.3 ar NaCl. Ceruloplasmin-containing fractions were inhibited by more than 98y0 with 1 mM xzide. Thus t,he two fcrroxidase proteins of the Cohn IV-1 fraction of serum could be completely resolved when chromatographed on DEXE-Sephadex h-50 just as with fresh human serum.
For preparative work, t,he entire dialyzed extract of c'ohn IV-1 (approximately 45 ml following the final cent rifugation) was chromatographed on a DEAE-Sephadex A-50 column (4.5 x 8.5 cm) with an identical elution pattern of protein and fcrroxidase activity.
All the ferroxidase-II activity was elutetl in the first 10 10-1111 fr&iOnS, 100 ml Of the 0.05 M XChte h&X, $1 5.5, containing no NaCl. The ceruloplnsmin remained as a well defined blue-green band at the very top of the preparative DEAE-Sephadex A-50 column until the sodium chloride concentration in the acetate buffer K~S incrcnsed to 0.2 &I.
At this stage the combined preparation (100 ml) was concentrated to a volume of 30 ml without loss of ferroxidase-II activity, either by desiccation against Sepl~atlcx  to an Agarose A-&n column (4.5 x 53 cm). Protein was eluted with 0.05 M acetate buffer, pH 5.5. Fractions of I3 ml were collected.
Two bands of ferroxidase-II activity were eluted (Fig. 2). The first of these bands, containing 10 to 12y0 of ferroxidase-II, (Band I, Fig. 2) MS yellowish and very turbid and was eluted immediately after the void volume.
The major portion (-90%) ferroxidase-II activity was retained on this column and eluted in Fractions 24 to 46. These fractions exhibited a yellow color but were not turbid.
The ferroxidase-II preparation isolated from fresh human serum by DEAE-Sephadex A-50 chromatography ( Fig. 1) Fas also applied to the column of Agarosc A-15,.
A single band of ferroxidase-II was observed with a mobility identical with the second major band shon-n in Fig. 2 (Band II).
Thus, the first minor band of ferroxidase-II activity (Band I, Fig. 2 This aggregate may rcprcsent partially denatured, insoluble, or lipid-occluded, and still active ferroxidase-II, which may have resulted during the Cohn fractionation or lyophilizat'ion procedure. Fractions 26 to 42 (Band ZZ, Fig. 2) from the Agarose A-15m gel filtration were combined.
The combined fractions (-200 ml) represented an 87.37, recovery of the ferroxidase-II activity and a further purification of 2-fold (Table III). Again 1 m&c azide did not inhibit the ferroxidase-II preparation.
This preparation was concentrated to a final volume of 30 ml by either desiccation against Sephadex G-200 or by ultrafiitrat,ion. Neither procedure resulted in a loss of ferroxidase-II activity.
Concentrated Band ZZ was divided into two equal 15.ml portions.
Each portion was chromatographed on a column of Xgnrose A-50m (3 X 45 cm). A single band of ferroxidasc-II activity was obtained (Fig. 3), which eluted with the major protein band. The immunoelectrophoresis was performed as described under "Methods." In the top well was placed 2 11 (5.8 pg) of the purified ferroxidase-II preparation.
In the middle well was placed 2 ~1 of whole human serum.
In both troughs was placed 100 11 of whole anti-human serum.
The disc gel elcctrophoresis was performed with a 3iy0 large pore polynrrylamide gel as described under "Methods." To the gels, 50 pll (38.6 ~g) of the purified ferroxidase-II preparation were applied. 9 minor, more slowly moving band of protein with no ferroxidase activity was also observed. Fractions 15 to 22 (Fig. 3) from this column were combined yielding a 54% recovery of ferroxidase-II and a further small purification. Again, no inhibition of the ferroxidase activity of this preparation was observed with 1 mM azide. The total purification procedure routinely yielded ferroxidase-II of 400-to 600-fold purification (Table III).
At the final &age of purification the enzyme was stable when stored at 4' for at least 2 weeks. hfter this time period, some slow precipitation and loss of activity was observed.

Properties of Puri$ed Ferroxidase-II Disc Gel and Immunoelectrophoresis of Ferroxidase-II after Final Purifkation
Step--When the purified ferroxidase-II preparation was subjected to gel filtration on a column of Sephadex G-200 (2.5 x 45 cm), all of the ferroxidase activity was excluded from the column and eluted immediately following the void volume. However, the same preparation was retained on Agarose A-15m. This indicated that the protein must have a large molecular weight, probably greater than 800,000, the exclusion limit of Sephadex G-200 for globular proteins. Accordingly for disc gel electrophoresis, a 3% large pore polyacrylamide gel was chosen. Electrophoresis of purified ferroxidase-II indicated a single major band of protein which migrated approximately one-half the distance of the tracking dye (Fig. 4). The maximum molecular weight a protein can possess and still migrate into this gel is approximately 2,000,OOO (18). Thus from the gel filtration dat.a and the disc gel electrophoresis it would appear that ferroxidase-II must have a molecular weight between 800,000 to 2,000,000, which is much larger than the value of 160,000 for ceruloplasmin (19).
The purified ferroxidase-II was also subjected to immunoelect.rophoresis. A single major precipitation arc was observed when cross-reacted with whole anti-human serum and stained with buffalo black (Fig. 4) or oil red, a stain for lipoproteins (15). The ferroxidase-II protein migrated only a small distance toward the anode and it has a mobility similar to that of serum lipo- of initial velocity and protein concentration of ferroxidase-II following the Agarose A-50m step of purification.
The initial velocity of Fe(I1) oxidation was measured utilizing 100, 75, 50, 25, and 10 ~1 of the purified ferroxidase-II solution (0.772 mg per ml) in a final volume of 1 ml as in Fig. 1. B, heat inactivation of ferroxidase-II following the Agarose A-50m stage of purification.
A sample of 10 ml of the purified ferroxidase-II (specific activity = 4.93) solution was placed in a test tube in a boiling water bath. Samples of 1 ml were withdrawn from this tube at 0, 3,5,10,20,30, and 40 min of heating.
The ferroxidase activity of each sample was assayed as in Fig. 1 with X00 ~1 of enzyme solution from each sample.
proteins. Ceruloplasmin when subjected to immunoelectrophoresis under the same conditions migrated much further toward the anode than the ferroxidase-II protein. The electrophoretic data suggest that the combination of the previously mentioned fractions from the Agarose A-50m column yields a relatively homogeneous ferroxidase-II protein.
Further Evidence for Enzymic Nature of Ferroxidase-II-Several facts previously mentioned all point to the macromolecular nature of ferroxidase-II. Extensive dialysis and concentration by ultrafiltration did not result in any loss of ferroxidase-II activity. Ferroxidase-II was purified by chromatographic methods usually employed for soluble proteins. Additional data also indicated the enzymic nature of the reaction catalyzed by ferroxidase-II. A linear relationship was observed between the initial velocity of the ferroxidase reaction and the concentration of ferroxidase-II prot,ein in the reaction mixture (Fig. 5A). The direct proportionality of initial velocity and protein concentration was observed at all stages of purification as well as with 6. Visible absorption spectra of ferroxidase-II in 0.05 M acetate buffer, pH = 5.5, at 30". One milliliter (3.13 mg) of the purified ferroxidase-II (specific activity = 4.0) was placed in a 1.5-ml quartz ruvette and the visible absorption spectrum recorded. A second spectrum was recorded with t,he enzyme solu- the purified protein.
The specific activity of purified ferroxidase-II in units of AA460 per 10 min per mg of protein was 4.9. The specific activity of purified ceruloplasmin in these same units is 30. However, if the approximate molecular weights of each ferroxidase are taken into consideration, the molar activities are similar. Another indication of the enzymic "oxidase" nature of the ferroxidase-II protein was the fact that oxygen consumption was observed simultaneously with Fe(I1) oxidation measured as transferrin formation.% Substrate Speci$$-p-Phenylenediamine was tested as a possible substrate for ferroxidase-II as described under "Met'hads." Ferroxidnse-II differed from ceruloplasmin in substrate specificity in t,hat it had no p-phenylenediamine oxidase activity (12,20). Another copper-containing plasma enzyme which could possibly serve as the non-ceruloplasma ferroxidase is plasma monoamine oxidase. However, ferroxidase-II did not catalyze the oxidation of benzylamine, a good substrate for human plasma monoamine oxidase (21). Two potent inhibitors of plasma monoamine oxidase, AT- (3 ,4-dichlorophenacyl) nm (460), and 332 nm (1600) (23). The purified ferroxid:lse-II showed none of the intense blue color (610.am absorption) which is characteristic of ceruloplasmin. The 402.nm maximum was shifted in the presence of 1 IY~M azide to a 415.nm maximum; however, the shoulders at 448 and 482 nm were unaffected.
When the visible spectrum was recorded immediately following the addition of Fe(I1) to the enzyme solution, the absorbance at 402 nm was significnntly reduced and the shoulders at 448 and 482 nm disappear.
Tt7hen the spectrum was recorded immediately after the addition of Fe(I1) to the enzyme solution coutaining 1 rnnf azide, the 415-nm peak was significantly reduced and the shoulders at 448 and 482 nm again disappeared.
The iron oxidase activity of fcrrosidase-II thus correlated with reduction of its characteristic risible absorbance spectrum.
Azide appnrent'ly affcctjcd the chromophore responsible for the 402.nm nb~orbnnce by a shift of this maximum to 415 nm; however, t'he azide did not block reduction of the visible absorbance when Fe(H) was nddad. Copper Content of PurQied E'errolidnse-II--Col,I,er remained firmly associated with ferroxidase-11 thro~~ghout the plnification procedure.
Chemical analysts performed by the method of Wharton and Radin (24) on ferroxidasc-II after t)he final step of purification and passage over a Chelrx 100 column to remove nonprotein-bound copper, indicated a copper content of 12 nmoles of copper per mg of protein.
These snmc enzyme prc,pnrations exhibited a typical electron pammagnetic rcsonancc signal (Fig. 7) with a g value very close to that observed for many other cupric copper-containing protcius (25). This t,ightlT-bound copper may be responsible at least in part, for the risible abqorption spectrum of the purified ferrosidase-II. DISCUSSION The properties which distinguish ferrositlase-II from ccruloplasmin are summarized in  Particularly interesting was the fact that ferrosidase-II had no aryldiamine oxidase activity.
The specific ferroxidase activities of ceruloplasmin and ferrosidase-II, expressed on a milligrams of protein basis, differ by a factor of 6. However, if the respective molecular weight of each ferroxidase is taken into consideration t,he molar activities are very close. The specific activities per copper atom are also similar.
The last six properties represent the important chemical differences.
It is interesting to note that the immunoelectrophoretic mobility, the oil red-staining property, and the large molecular weight suggest that ferroxidase-II may be a serum lipoprotein (present paper and Reference 10). Ceruloplasmin is approximately 0.28% copper; however ferroxidase-II contained approximat)ely O.OSy, copper.
The copper content of ferroxidase-II is similar to several other copper proteins, e.g. tyrosinase and cytochrome oxidase (24).
Similarly, many properties distinguish ferroxidase-II from human plasma monoamine oxidase. (a) Ferroxidase-II did not catalyze the oxidation of benzylamine.
(b) Ko inhibition of ferroxidase-II was observed with potent inhibitors of plasma monoamine oxidase. (c) Th e molecular weights of the well characterized plasma monoamine oxidases are 260,000 (bovine plasma enzyme) and 190,000 (pig plasma enzyme) (26). Ferroxidase-II appears to have a much larger molecular weight.
(d) Purified human plasma monoamine oxidase preparations were colorless (21) and purified bovine plasma monoamine oxidase preparations were pink (27). Ferroxidase-II preparations are yellow in appearance.
(e) The visible spect'ral features of purified ferroxidase-II preparations correlated wit,h their ferroxidase activity.
Xo correlation between the visible spectral features and the monoamine oxidase activity was observed for purified human plasma monoamine oxidase preparations (21). The iron oxidation reaction cat'alyzed by ferroxidase-II appears to be an "enzymic oxidase" reaction for the following reasons. (a) The initial velocity of the ferroxidase-II reaction was directly proportional to the concentration of protein at all stages of purification.
(b) Inactivation of these ferroxidase-II preparations was achieved by heat treatment and with the purified material after prolonged storage in the cold. (c) The molar act,ivity of ferroxidase-II or specific activity per copper atom is close to that of ceruloplasmin although the specific activities on a milligrams of protein basis were quite different. (4 Oxygen consumpt,ion was observed simultaneously with the catalysis of iron oxidation by the ferroxidase-II. (e) With regard to sub-strates thus far tested, ferroxidase-II apparently exhibits specificity for Fe(I1).
One major objection has been raised to the proposal that the physiological function of ceruloplasmin in blood is to catalyze the oxidation of Fe(I1) in plasma to Fe(III), promoting the iron saturation of transferrin, and thereby stimulating t,he over-all turnover of iron from tissue depots. This objecbion stems from the fact that there appears to be little disturbance in iron metabolism in Wilson's disease, a disorder characterized by low plasma ceruloplasmin.
The finding t,hat ferroxidase-II exists in human serum may resolve this apparent paradox.
Our results indicate that while the ceruloplasmin level of Wilson's disease serum decreases very dramatically, ferroxidase-II is less affected. The total ferroxidase activity of Wilson's disease sera is approximately 5 to 10% of that of normal sera (set "Results" and Reference 7). Osaki et al. (6) have found that only about 10% of the tot.al ferroxidase activit,y of normal serum was necessary in order to observe maximum iron mobilization response from liver. Thus the total ferroxidase of Wilson's disease serum would be enough for normal or low normal iron metabolism.
However, without ferroxidase-II which accounts for 30% of the total ferroxidase in Wilson's disease serum, this would not be possible. Ferroxidase-II also accounts for the fact that p-phenylenediamine oxidase activity did not correlate with ferroxidase activity in Wilson's disease sera and in sera from some other animals (6,7,9).
Further investigations with ferroxidase-II concerned with the kinetics of the iron oxidation and oxygen consumption reactions, the possible role of copper in the enzymic reaction, the physicochemical properties, the substrate specificity, and the physiological significance in Wilson's disease and other species are currently in progress.