Biqchimica e~,Biophysica Acta, 1117 (1992) 25-32
25
© 1992 Elsevier Science Publishers B,V. All rights reserved 0304-4165/92/$05.00
BBAGEN 23689
Purification and partial characterization of xanthine oxidase from human milk Shahla Abadeh, Joanne Killacky, Mustapha Benboubetra and Roger Harrison Biochemistry Department, UnA'ersity of Bath, Bath (UK)
(Received 2 March 1992)
Key words: Xanthine oxidase; Milk; Enzyme characterisation; Enzyme purification; (Human) Xanthine oxidase was purified from human milk in yields comparable with those obtained from bovine milk. The freshly purified enzyme appeared homogeneous in gel permeation FPLC and SDS-PAGE, consistent with its being a homodimer with total M r 290 000 _+6000. The ultraviolet/visible absorption spectrum differed only slightly from that of bovine milk enzyme and showed an A2so/A45 o ratio of 5.13 + 0.29, indicating a high degree of purity. Xanthine oxidase activities of purified enzyme varied with batches of milk, ranging between 3 and 46 m U / m g protein; values that are some two to three orders of magnitude smaller than those shown by the most highly purified samples of bovine milk enzyme. Direct comparison with commercially-available bovine milk enzyme showed that activities involving xanthine as reducing substrate were 1-6% that of the bovine enzyme, whereas those involving NADH, in contrast, were of the same order for the two enzymes. Anaerobic bleaching experiments indicated that less than 2% of the human enzyme was present as a form active with xanthine. These findings, together with the activity data, are consistent with a very high content, possibly greater than 98%, of demolybdo- and/or desulpho-forms of human enzyme, both of which occur, to a lesser extent, in bovine xanthine oxidase. Molybdenum assay indicated that demolybdo-enzyme could only account for some 26% of this inactive component, suggesting that desulpho-enzyme may account for the remainder.
Introduction Xanthine oxidase is a molybdoenzyme that catalyses the oxidation of hypoxanthine to xanthine and of xanthine to uric acid [1]. In mammalian systems, it occurs as partially interconvertible Type D (dehydrogenase) and Type O (oxidase) forms, reducing, respectively, N A D + or oxygen. In such systems, the term 'xanthine oxidase' commonly refers to both Type D and Type O forms and, unless otherwise specified, will do so here. The mammalian enzyme is found primarily in mammary epithelial cells and capillary endothelial cells [2]; a particular distribution that is hard to explain in terms of a general housekeeping role in purine catabolism, although, other, more specialised functions have been proposed for the enzyme, such as anti-microbial defence [2]. The latter role of the enzyme depends on its ability to generate free radicals; an ability that has led to its becoming a focus of interest as an initiator of tissue damage in a range of pathological states, includ-
Correspondence to: R. Harrison, Biochemistry Department, University of Bath, Bath, BA2 7AY, UK. Abbreviations: SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography.
ing ischaemia-reperfusion damage of intestine [3], myocardium [4] and joints [5]. The proposed mechanism of such damage involves the following sequence [6]: during ischaemia, A T P is catabolised to hypoxanthine, which accumulates in the tissues. As a result of the low energy state, Ca 2+ passes into the cells and triggers proteolytic conversion of the naturally-occurring form (Type D) of xanthine oxidase to Type O. Reperfusion then reintroduces molecular oxygen, which is reduced by Type O enzyme, in the presence of hypoxanthine, to superoxide anion and hydrogen peroxide. These reactive species, together with hydroxyl radicals derived from them, then cause tissue damage via a range of mechanisms. Despite the many questions concerning its normal and pathological roles, human xanthine oxidase has been little studied in purified form, although a proteolysed preparation from liver has been described [7]. Bovine milk xanthine oxidase, in contrast, has been studied extensively [1] as have, to a lesser extent, chicken [8] and rat [9] liver enzymes. Interest in human xanthine oxidase has been further stimulated by the recent observation [10] that levels of anti-(xanthine oxidase) antibodies, present in most human sera, are elevated in patients who have suffered a
26 myocardial infarction. It is clearly important to determine the origin of these human antibodies, which could arise in response to ingested bovine milk or to endogenous enzyme, and comparisons of the interactions of bovine and human enzymes with the antibodies should help to clarify this issue. In view of all the above questions concerning human xanthine oxidase, we now describe a simple method for its purification from breast milk and outline some properties of the purified enzyme. A preliminary account of some of this work has been published [11]. Materials and Methods
Materials Collection of human breast milk from local donors was coordinated by Mrs Julia Reid with the kind cooperation of the nursing staff of the Special Baby Care Clinic, Royal United Hospitals, Bath and the Health visitors of Bath and Chippenham areas. Milk was stored at 4°C and normally processed within 12 h of expression in batches combined from one to seven donors. Bovine milk xanthine oxidase was purchased (XO-2, non-pancreatin treated) from Biozyme Laboratories, Blaenavon, Wales, or prepared as described by Nakamura and Yamazaki [12]. Bovine milk fat globule membrane was prepared by using fresh uncooled milk from Friesian cows in mid-lactation, as described by Mather and Keenan [13]. Similar milk was used for determinations of xanthine oxidase activity. Unless otherwise specified, other reagents were purchased from Sigma, Poole, Dorset, UK.
Purification of xanthine oxidase from human milk The procedure is slightly modified from that described by Nakamura and Yamazaki [12] for bovine milk. Human breast milk (100 ml) was supplemented with 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate. It was then cooled to 4°C, centrifuged at 3000 × g for 20 min and the upper layer (cream) was collected. An equal volume of 0.2 M K2HPO4, containing 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate was added to the cream, followed by slow addition of cold ( - 2 0° C) butanol to give 15% (v/v) and then 15% (w/v) solid (NH4)2SO 4 (relative to aqueous volume), with constant stirring. The resultant slurry was stirred for 1 h and then centrifuged at 13 000 × g for 20 min. The aqueous lower phase was separated and 20% (w/v) solid (NH4)2SO 4 was added, with stirring. The suspension was allowed to stand at 4°C for 2 h, after which the separated top phase was collected and centrifuged at 10000 × g for 30 min. The top phase was again collected and suspended in a small volume of buffer A
(0.2 M Na2HPO4, containing 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate, adjusted to pH 6.0 with HC1) which was then dialysed against the same buffer (2 1) overnight. The precipitate was removed by centrifugation at 20 000 × g for 1 h, giving a supernatant of 'crude enzyme'. 'Crude enzyme' was applied to a column (1.5 cm × 15 cm) of calcium phosphate gel (Type II: Neutral (Brushite), Sigma) previously equilibrated in buffer A. The column was washed with this buffer until no more protein (A2~o) was eluted. Bound xanthine oxidase was then eluted from the column by using buffer A, containing, additionally, 5% (w/v) (NH4)2SO 4. Protein-containing fractions were combined and dialysed overnight against suitable buffer.
Enzyme assays Fluorimetric assays were carried out by using a Perkin Elmer LS-5B luminescence spectrometer. Spectrophotometric assays were done by using a Pye Unicam SP6-450 spectrophotometer. Throughout the purification of human enzyme and for assay of bovine milk, xanthine oxidase activities were measured fluorimetrically in terms of the production of isoxanthopterin from pterin,in the presence (Type D + O) and absence (Type O) of methylene blue, as described by Beckman et al. [14] but at 22°C (1 mU = 1 nmol isoxanthopterin per rain). As reported by Beckman et al, proportions of Type D and Type O activities, determined by these means, were found (when practicable, e.g., with purified enzyme) to be essentially identical with those obtained by spectrophotometric assay of urate production (see below). Kinetic assays of purified enzymes were carried out aerobically at 22°C in 0.05 M K3PO 4, adjusted to pH 7.8 with HC1, containing 0.1 mM EDTA in a final volume of 1.0 ml. Xanthine oxidase activities were assayed in the presence (Type D + O) and absence (Type O) of 500 p~M NAD +, by following the production of urate at 292 nm from 100 /xM xanthine. 1 mU = 1 nmol urate per min., E292 12.2 mM-1 cm-1. K m values were determined, in the presence of 500 /~M NAD +, by using xanthine concentrations ranging from 2-100 /xM. Xanthine: methylene blue oxido-reductase was assayed, as above, for xanthine oxidase, except that NAD + was replaced by 20 ~M methylene blue. Xanthine : cytochrome c and NADH : cytochrome e oxido-reductases were assayed in the presence of 100 /xM xanthine or 50 ~M NADH, respectively, by measuring the reduction of cytochrome c (25 /xM) at 550 rim. 1 mU = 1 nmol cytochrome c per min, %5o = 21.0 raM- 1 cm- ~. NADH oxidase and NADH : methylene blue oxido-reductase assays were carried out in the presence of 50/~M NADH and, for the latter assay, 20 /~M methylene blue, at 340 nm. 1 mU = 1 tzmol NADH per min, 6340 6.22 m M - l cm-1. =
=
27
Absorption spectra and bleaching experiments Absorption spectra were obtained by using a Cecil CE 6600 spectrophotometer at 22°C. Bleaching experiments were carried out in a 1 ml quartz spectrophotometer cell, closed with a rubber seal, essentially as described by Hughes et al. [15]. Xanthine oxidase in 50 mM Na+-Bicine buffer (pH 8.2), (1 ml) at 22°C, was made anaerobic by repeated evacuation and flushing with argon using syringe needles inserted through the seal. The absorption spectrum was scanned and xanthine or Na2S204, in anaerobic buffer (10 ~1), was added as above, to a final concentration of 0.1 mM. The spectrum was scanned 2 min after each addition.
8.3) at respective flow rates of 0.3 ml/min and 1.0 ml/min. Molecular weights of eluted proteins were determined by comparison with protein standards (MW-GF-1000 Kit, Sigma Chemical). Molybdenum was determined colorimetrically with toluene-3,4-dithiol, after wet ashing with HCIO4/ H2504, as described by Hart et al. [18] and FAD was determined fluorimetrically, as described by Burch [19]. Both values are expressed per subunit (M r 145000) determined by protein assay. Protein determinations were carried out by the method of Bradford [20]. Results
Resulphuration experiments Human xanthine oxidase (approx. 1 mg protein) in 0.1 M NaaP207, adjusted to pH 7.8 with HC1, containing 0.1 mM EDTA (1 ml) was made anaerobic as described for the bleaching experiments, and anaerobic solutions of 1 M Na2S and 0.1 M Na2S204 were added (10 p.l aliquots) by syringe as above. After incubation at 22°C for 40 min, the solution was gel filtered on Sephadex G25 and used for activity assays. This procedure follows that described by Wahl and Rajagopalan [16] for resulphuration of the bovine enzyme.
Analytical procedures SDS-PAGE was performed in 10% acrylamide gels, according to Laemmli [17]. Molecular weights of bands on the gels were determined by comparison with standard proteins (MW-SDS-200 Kit, Sigma Chemical). FPLC was performed on a Pharmacia LKB system, using Superose 6 HR10/30 and Hiload Superdex 200 prepacked columns (Pharmacia, Uppsala, Sweden). Samples were loaded in 200 /zl and 500 /~1 volumes, respectively, and eluted in 0.2 M Tris-HCl buffer (pH
Xanthine oxidase activity in human milk was too low to be measured by conventional spectrophotometric assay of urate production (Materials and Methods) but could be assayed fluorimetrically in terms of isoxanthopterin production from pterin (Materials and Methods). The activity of 0.52 mU/ml, shown in Table I, is very low compared with 70 m U / m l similarly determined for fresh bovine milk. Purification of human milk was monitored by the fluorimetric assay, giving patterns of activities typified by that of Table I, which shows an overall yield of purified enzyme of 18% relative to milk. Ten separate preparations yielded 1.5-6.0 mg purified enzyme per 100 ml milk. The percentage of Type D activity in untreated milk varied between 20% and 55% and commonly rose by some 20% during purification in the presence of dithioerythritol (e.g., Table I). Omission of dithioerythritol from the media led to 100% Type O enzyme. Freshly purified human xanthine oxidase showed a single major band on SDS-PAGE, corresponding to an M r of approx. 150 000. After storage at 4°C for several days, a second band, corresponding to approximately
TABLE I
Purification of xanthine oxidase from human milk The fractionation procedure was as described in the Materials and Methods section. Enzymic activity was assayed throughout in terms of production of isoxanthopterin from pterin, by measuring fluorescence at 390 nm. Types D + O and Type O activities were determined in the presence and absence, respectively, of methylene blue, as described by Beckman et al. [14] (see Materials and Methods). Fraction
Volume (ml)
Types D + O activity (mU/ml)
Type D activity (%)
Types D + O activity (total, mU)
Protein (mg)
Specific activity mU/mg
Recovery (%)
Milk 1st (15% w/v) (NI-I 4)2 SO 4 subnatant 'Crude enzyme' Purified enzyme
140
0.52
53
72.8
1988
0.04
100
22
1.23
70
27.1
55
0.50
37
12
1.91
70
22.9
17.5
1.30
32
2.03
70
13.2
3.1
4.30
18
6.5
28 140 000 appeared. This behaviour is very similar to that shown by freshly-prepared bovine milk enzyme (Fig. 1). As for bovine enzyme, storage of human xanthine oxidase for more than 3 weeks at 4°C led to intensification of the minor band, at the expense of the major, and the appearance of further bands, corresponding to those shown by commercial bovine enzyme (Fig. 1). Purified human enzyme showed a single symmetrical peak in gel filtration FPLC, using either Superose 6 or Superdex 200 columns. Comparison with protein standards run on the latter column (Fig. 2) gave an M r of 290 000 + 6000 (mean + S.E., n = 4), consistently higher than the value (275 000 _+ 5500 (mean + S.E.; n = 4)) obtained for the single peak shown by bovine enzyme, whether freshly prepared from cows' milk or obtained commercially (Biozyme Laboratories). Storage of either human or bovine enzyme in the absence of dithioerythritol led to multiple peaks, in FPLC, corresponding to aggregates, having M r values up to 1000 000. The ultraviolet/visible absorption spectrum of purified human enzyme (Fig. 3a) is similar to that of bovine enzyme (Fig. 3b), apart from consistently lower absorbance in the 300-330 nm region, and shows an Azso/A45 o ratio of 5.13 +_ 0.29 (mean + S.E.; n = 8). Purified enzyme contained 0.74-0.95 (n = 4) atoms of molybdenum and 0.98 mol FAD per 145 000 subunit.
M r
x10-3
205
........
116 97.4
......... .......
66
..........
5
9
• ............
..........
1
2
3
4
5
6
Fig. 1. SDS-PAGE of xanthine oxidase preparations. Conditions were as described in the Materials and Methods Section. Lanes: 1, bovine enzyme [12]; 2, human enzyme; 3, mol.wt, markers; 4, bovine enzyme (Biozyme); 5, bovine enzyme [12]; 6, bovine milk fat globule membrane [13].
401 A
30~B BXO DO'~o
E
10-
45
50
55
6.0
- log M Fig. 2. Determination of the molecular weights of human (HXO) and bovine (BXO) xanthine oxidases by FPLC on Superdex 200, as described under Materials and Methods. V~ and Vo are elution and void volumes, respectively. Standard proteins were carbonic anhydrase (A, 29000), bovine serum albumin (B, 66000), yeast alcohol dehydrogenase (C, 150000), /3-amylase (D, 200000), apoferritin (E, 443 000) and thyroglobulin, (F, 669000).
Specific xanthine oxidase activities (Types D + O) of purified human enzyme, as measured by urate production in the presence of NAD + (Materials and Methods), ranged between 3 and 46 m U / m g protein (n = 10), depending on the batch of milk. Specific activities for oxidation of xanthine and N A D H in the presence of different electron acceptors are shown, for one preparation of purified human enzyme, in Table II. Comparison is made with values obtained for commercially available bovine enzyme of similar dehydrogenase content. Activities involving xanthine as substrate are only 1 - 6 % of corresponding values for bovine milk enzyme. K m for xanthine, determined in the presence of oxygen and NAD+, (Materials and Methods) for the same human enzyme preparation (18 txM) was somewhat higher than that (2.5/xM) for the bovine enzyme. In contrast to the great disparities seen with xanthine, specific activities obtained with N A D H as reducing substrate are of the same order for human and bovine enzymes (Table II). In order to investigate the possibility that the human enzyme might contain a high proportion of enzyme that is inactive to xanthine (see Discussion), anaerobic bleaching experiments were carried out, as described in the Materials and Methods section. Under anaerobic conditions, the rapid fall in absorbance at 450 nm brought about by addition of xanthine was compared with the maximum fall caused by N a 2 S 2 0 4. In studies with bovine milk enzyme, this ratio is accepted as giving an indication of the percentage of active enzyme
29 [23,24] and, indeed, in our hands, yielded a value of 59% for the commercially-available enzyme (Fig. 4a). Addition of xanthine to the human enzyme, in con-
0.4-
0.3.
trast, showed a detectable fall in absorbance at 450 nm that was very much smaller, and accordingly, more difficult to quantify. This fall represented some 2% or less (after correction for dilution effects) of that brought about by addition of NazSzO 4 (Fig. 4b). With a view to examining the possible presence of desulpho-enzyme in human xanthine oxidase, resulphuration was attempted. A sample of purified human enzyme, with specific xanthine oxidase (Types D + O) activity of 5.0 m U / m g protein was incubated anaerobically with NazS and Na2S204, as described by Wahl and Rajagopalan [16]. After gel filtration on Sephadex G25, the specific activity was found to have increased to 29 m U / m g protein. Discussion
c
.~ 0.2
0.1
360
460
560
e6o
Wavelength (rim)
0.6-
0.5-
0.4-
G m 0.3 .Q .Q ,
25
© 1992 Elsevier Science Publishers B,V. All rights reserved 0304-4165/92/$05.00
BBAGEN 23689
Purification and partial characterization of xanthine oxidase from human milk Shahla Abadeh, Joanne Killacky, Mustapha Benboubetra and Roger Harrison Biochemistry Department, UnA'ersity of Bath, Bath (UK)
(Received 2 March 1992)
Key words: Xanthine oxidase; Milk; Enzyme characterisation; Enzyme purification; (Human) Xanthine oxidase was purified from human milk in yields comparable with those obtained from bovine milk. The freshly purified enzyme appeared homogeneous in gel permeation FPLC and SDS-PAGE, consistent with its being a homodimer with total M r 290 000 _+6000. The ultraviolet/visible absorption spectrum differed only slightly from that of bovine milk enzyme and showed an A2so/A45 o ratio of 5.13 + 0.29, indicating a high degree of purity. Xanthine oxidase activities of purified enzyme varied with batches of milk, ranging between 3 and 46 m U / m g protein; values that are some two to three orders of magnitude smaller than those shown by the most highly purified samples of bovine milk enzyme. Direct comparison with commercially-available bovine milk enzyme showed that activities involving xanthine as reducing substrate were 1-6% that of the bovine enzyme, whereas those involving NADH, in contrast, were of the same order for the two enzymes. Anaerobic bleaching experiments indicated that less than 2% of the human enzyme was present as a form active with xanthine. These findings, together with the activity data, are consistent with a very high content, possibly greater than 98%, of demolybdo- and/or desulpho-forms of human enzyme, both of which occur, to a lesser extent, in bovine xanthine oxidase. Molybdenum assay indicated that demolybdo-enzyme could only account for some 26% of this inactive component, suggesting that desulpho-enzyme may account for the remainder.
Introduction Xanthine oxidase is a molybdoenzyme that catalyses the oxidation of hypoxanthine to xanthine and of xanthine to uric acid [1]. In mammalian systems, it occurs as partially interconvertible Type D (dehydrogenase) and Type O (oxidase) forms, reducing, respectively, N A D + or oxygen. In such systems, the term 'xanthine oxidase' commonly refers to both Type D and Type O forms and, unless otherwise specified, will do so here. The mammalian enzyme is found primarily in mammary epithelial cells and capillary endothelial cells [2]; a particular distribution that is hard to explain in terms of a general housekeeping role in purine catabolism, although, other, more specialised functions have been proposed for the enzyme, such as anti-microbial defence [2]. The latter role of the enzyme depends on its ability to generate free radicals; an ability that has led to its becoming a focus of interest as an initiator of tissue damage in a range of pathological states, includ-
Correspondence to: R. Harrison, Biochemistry Department, University of Bath, Bath, BA2 7AY, UK. Abbreviations: SDS-PAGE, sodium dodecyl sulphate-polyacrylamide gel electrophoresis; FPLC, fast protein liquid chromatography.
ing ischaemia-reperfusion damage of intestine [3], myocardium [4] and joints [5]. The proposed mechanism of such damage involves the following sequence [6]: during ischaemia, A T P is catabolised to hypoxanthine, which accumulates in the tissues. As a result of the low energy state, Ca 2+ passes into the cells and triggers proteolytic conversion of the naturally-occurring form (Type D) of xanthine oxidase to Type O. Reperfusion then reintroduces molecular oxygen, which is reduced by Type O enzyme, in the presence of hypoxanthine, to superoxide anion and hydrogen peroxide. These reactive species, together with hydroxyl radicals derived from them, then cause tissue damage via a range of mechanisms. Despite the many questions concerning its normal and pathological roles, human xanthine oxidase has been little studied in purified form, although a proteolysed preparation from liver has been described [7]. Bovine milk xanthine oxidase, in contrast, has been studied extensively [1] as have, to a lesser extent, chicken [8] and rat [9] liver enzymes. Interest in human xanthine oxidase has been further stimulated by the recent observation [10] that levels of anti-(xanthine oxidase) antibodies, present in most human sera, are elevated in patients who have suffered a
26 myocardial infarction. It is clearly important to determine the origin of these human antibodies, which could arise in response to ingested bovine milk or to endogenous enzyme, and comparisons of the interactions of bovine and human enzymes with the antibodies should help to clarify this issue. In view of all the above questions concerning human xanthine oxidase, we now describe a simple method for its purification from breast milk and outline some properties of the purified enzyme. A preliminary account of some of this work has been published [11]. Materials and Methods
Materials Collection of human breast milk from local donors was coordinated by Mrs Julia Reid with the kind cooperation of the nursing staff of the Special Baby Care Clinic, Royal United Hospitals, Bath and the Health visitors of Bath and Chippenham areas. Milk was stored at 4°C and normally processed within 12 h of expression in batches combined from one to seven donors. Bovine milk xanthine oxidase was purchased (XO-2, non-pancreatin treated) from Biozyme Laboratories, Blaenavon, Wales, or prepared as described by Nakamura and Yamazaki [12]. Bovine milk fat globule membrane was prepared by using fresh uncooled milk from Friesian cows in mid-lactation, as described by Mather and Keenan [13]. Similar milk was used for determinations of xanthine oxidase activity. Unless otherwise specified, other reagents were purchased from Sigma, Poole, Dorset, UK.
Purification of xanthine oxidase from human milk The procedure is slightly modified from that described by Nakamura and Yamazaki [12] for bovine milk. Human breast milk (100 ml) was supplemented with 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate. It was then cooled to 4°C, centrifuged at 3000 × g for 20 min and the upper layer (cream) was collected. An equal volume of 0.2 M K2HPO4, containing 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate was added to the cream, followed by slow addition of cold ( - 2 0° C) butanol to give 15% (v/v) and then 15% (w/v) solid (NH4)2SO 4 (relative to aqueous volume), with constant stirring. The resultant slurry was stirred for 1 h and then centrifuged at 13 000 × g for 20 min. The aqueous lower phase was separated and 20% (w/v) solid (NH4)2SO 4 was added, with stirring. The suspension was allowed to stand at 4°C for 2 h, after which the separated top phase was collected and centrifuged at 10000 × g for 30 min. The top phase was again collected and suspended in a small volume of buffer A
(0.2 M Na2HPO4, containing 2.5 mM dithioerythritol, 1 mM EDTA and 1.25 mM sodium salicylate, adjusted to pH 6.0 with HC1) which was then dialysed against the same buffer (2 1) overnight. The precipitate was removed by centrifugation at 20 000 × g for 1 h, giving a supernatant of 'crude enzyme'. 'Crude enzyme' was applied to a column (1.5 cm × 15 cm) of calcium phosphate gel (Type II: Neutral (Brushite), Sigma) previously equilibrated in buffer A. The column was washed with this buffer until no more protein (A2~o) was eluted. Bound xanthine oxidase was then eluted from the column by using buffer A, containing, additionally, 5% (w/v) (NH4)2SO 4. Protein-containing fractions were combined and dialysed overnight against suitable buffer.
Enzyme assays Fluorimetric assays were carried out by using a Perkin Elmer LS-5B luminescence spectrometer. Spectrophotometric assays were done by using a Pye Unicam SP6-450 spectrophotometer. Throughout the purification of human enzyme and for assay of bovine milk, xanthine oxidase activities were measured fluorimetrically in terms of the production of isoxanthopterin from pterin,in the presence (Type D + O) and absence (Type O) of methylene blue, as described by Beckman et al. [14] but at 22°C (1 mU = 1 nmol isoxanthopterin per rain). As reported by Beckman et al, proportions of Type D and Type O activities, determined by these means, were found (when practicable, e.g., with purified enzyme) to be essentially identical with those obtained by spectrophotometric assay of urate production (see below). Kinetic assays of purified enzymes were carried out aerobically at 22°C in 0.05 M K3PO 4, adjusted to pH 7.8 with HC1, containing 0.1 mM EDTA in a final volume of 1.0 ml. Xanthine oxidase activities were assayed in the presence (Type D + O) and absence (Type O) of 500 p~M NAD +, by following the production of urate at 292 nm from 100 /xM xanthine. 1 mU = 1 nmol urate per min., E292 12.2 mM-1 cm-1. K m values were determined, in the presence of 500 /~M NAD +, by using xanthine concentrations ranging from 2-100 /xM. Xanthine: methylene blue oxido-reductase was assayed, as above, for xanthine oxidase, except that NAD + was replaced by 20 ~M methylene blue. Xanthine : cytochrome c and NADH : cytochrome e oxido-reductases were assayed in the presence of 100 /xM xanthine or 50 ~M NADH, respectively, by measuring the reduction of cytochrome c (25 /xM) at 550 rim. 1 mU = 1 nmol cytochrome c per min, %5o = 21.0 raM- 1 cm- ~. NADH oxidase and NADH : methylene blue oxido-reductase assays were carried out in the presence of 50/~M NADH and, for the latter assay, 20 /~M methylene blue, at 340 nm. 1 mU = 1 tzmol NADH per min, 6340 6.22 m M - l cm-1. =
=
27
Absorption spectra and bleaching experiments Absorption spectra were obtained by using a Cecil CE 6600 spectrophotometer at 22°C. Bleaching experiments were carried out in a 1 ml quartz spectrophotometer cell, closed with a rubber seal, essentially as described by Hughes et al. [15]. Xanthine oxidase in 50 mM Na+-Bicine buffer (pH 8.2), (1 ml) at 22°C, was made anaerobic by repeated evacuation and flushing with argon using syringe needles inserted through the seal. The absorption spectrum was scanned and xanthine or Na2S204, in anaerobic buffer (10 ~1), was added as above, to a final concentration of 0.1 mM. The spectrum was scanned 2 min after each addition.
8.3) at respective flow rates of 0.3 ml/min and 1.0 ml/min. Molecular weights of eluted proteins were determined by comparison with protein standards (MW-GF-1000 Kit, Sigma Chemical). Molybdenum was determined colorimetrically with toluene-3,4-dithiol, after wet ashing with HCIO4/ H2504, as described by Hart et al. [18] and FAD was determined fluorimetrically, as described by Burch [19]. Both values are expressed per subunit (M r 145000) determined by protein assay. Protein determinations were carried out by the method of Bradford [20]. Results
Resulphuration experiments Human xanthine oxidase (approx. 1 mg protein) in 0.1 M NaaP207, adjusted to pH 7.8 with HC1, containing 0.1 mM EDTA (1 ml) was made anaerobic as described for the bleaching experiments, and anaerobic solutions of 1 M Na2S and 0.1 M Na2S204 were added (10 p.l aliquots) by syringe as above. After incubation at 22°C for 40 min, the solution was gel filtered on Sephadex G25 and used for activity assays. This procedure follows that described by Wahl and Rajagopalan [16] for resulphuration of the bovine enzyme.
Analytical procedures SDS-PAGE was performed in 10% acrylamide gels, according to Laemmli [17]. Molecular weights of bands on the gels were determined by comparison with standard proteins (MW-SDS-200 Kit, Sigma Chemical). FPLC was performed on a Pharmacia LKB system, using Superose 6 HR10/30 and Hiload Superdex 200 prepacked columns (Pharmacia, Uppsala, Sweden). Samples were loaded in 200 /zl and 500 /~1 volumes, respectively, and eluted in 0.2 M Tris-HCl buffer (pH
Xanthine oxidase activity in human milk was too low to be measured by conventional spectrophotometric assay of urate production (Materials and Methods) but could be assayed fluorimetrically in terms of isoxanthopterin production from pterin (Materials and Methods). The activity of 0.52 mU/ml, shown in Table I, is very low compared with 70 m U / m l similarly determined for fresh bovine milk. Purification of human milk was monitored by the fluorimetric assay, giving patterns of activities typified by that of Table I, which shows an overall yield of purified enzyme of 18% relative to milk. Ten separate preparations yielded 1.5-6.0 mg purified enzyme per 100 ml milk. The percentage of Type D activity in untreated milk varied between 20% and 55% and commonly rose by some 20% during purification in the presence of dithioerythritol (e.g., Table I). Omission of dithioerythritol from the media led to 100% Type O enzyme. Freshly purified human xanthine oxidase showed a single major band on SDS-PAGE, corresponding to an M r of approx. 150 000. After storage at 4°C for several days, a second band, corresponding to approximately
TABLE I
Purification of xanthine oxidase from human milk The fractionation procedure was as described in the Materials and Methods section. Enzymic activity was assayed throughout in terms of production of isoxanthopterin from pterin, by measuring fluorescence at 390 nm. Types D + O and Type O activities were determined in the presence and absence, respectively, of methylene blue, as described by Beckman et al. [14] (see Materials and Methods). Fraction
Volume (ml)
Types D + O activity (mU/ml)
Type D activity (%)
Types D + O activity (total, mU)
Protein (mg)
Specific activity mU/mg
Recovery (%)
Milk 1st (15% w/v) (NI-I 4)2 SO 4 subnatant 'Crude enzyme' Purified enzyme
140
0.52
53
72.8
1988
0.04
100
22
1.23
70
27.1
55
0.50
37
12
1.91
70
22.9
17.5
1.30
32
2.03
70
13.2
3.1
4.30
18
6.5
28 140 000 appeared. This behaviour is very similar to that shown by freshly-prepared bovine milk enzyme (Fig. 1). As for bovine enzyme, storage of human xanthine oxidase for more than 3 weeks at 4°C led to intensification of the minor band, at the expense of the major, and the appearance of further bands, corresponding to those shown by commercial bovine enzyme (Fig. 1). Purified human enzyme showed a single symmetrical peak in gel filtration FPLC, using either Superose 6 or Superdex 200 columns. Comparison with protein standards run on the latter column (Fig. 2) gave an M r of 290 000 + 6000 (mean + S.E., n = 4), consistently higher than the value (275 000 _+ 5500 (mean + S.E.; n = 4)) obtained for the single peak shown by bovine enzyme, whether freshly prepared from cows' milk or obtained commercially (Biozyme Laboratories). Storage of either human or bovine enzyme in the absence of dithioerythritol led to multiple peaks, in FPLC, corresponding to aggregates, having M r values up to 1000 000. The ultraviolet/visible absorption spectrum of purified human enzyme (Fig. 3a) is similar to that of bovine enzyme (Fig. 3b), apart from consistently lower absorbance in the 300-330 nm region, and shows an Azso/A45 o ratio of 5.13 +_ 0.29 (mean + S.E.; n = 8). Purified enzyme contained 0.74-0.95 (n = 4) atoms of molybdenum and 0.98 mol FAD per 145 000 subunit.
M r
x10-3
205
........
116 97.4
......... .......
66
..........
5
9
• ............
..........
1
2
3
4
5
6
Fig. 1. SDS-PAGE of xanthine oxidase preparations. Conditions were as described in the Materials and Methods Section. Lanes: 1, bovine enzyme [12]; 2, human enzyme; 3, mol.wt, markers; 4, bovine enzyme (Biozyme); 5, bovine enzyme [12]; 6, bovine milk fat globule membrane [13].
401 A
30~B BXO DO'~o
E
10-
45
50
55
6.0
- log M Fig. 2. Determination of the molecular weights of human (HXO) and bovine (BXO) xanthine oxidases by FPLC on Superdex 200, as described under Materials and Methods. V~ and Vo are elution and void volumes, respectively. Standard proteins were carbonic anhydrase (A, 29000), bovine serum albumin (B, 66000), yeast alcohol dehydrogenase (C, 150000), /3-amylase (D, 200000), apoferritin (E, 443 000) and thyroglobulin, (F, 669000).
Specific xanthine oxidase activities (Types D + O) of purified human enzyme, as measured by urate production in the presence of NAD + (Materials and Methods), ranged between 3 and 46 m U / m g protein (n = 10), depending on the batch of milk. Specific activities for oxidation of xanthine and N A D H in the presence of different electron acceptors are shown, for one preparation of purified human enzyme, in Table II. Comparison is made with values obtained for commercially available bovine enzyme of similar dehydrogenase content. Activities involving xanthine as substrate are only 1 - 6 % of corresponding values for bovine milk enzyme. K m for xanthine, determined in the presence of oxygen and NAD+, (Materials and Methods) for the same human enzyme preparation (18 txM) was somewhat higher than that (2.5/xM) for the bovine enzyme. In contrast to the great disparities seen with xanthine, specific activities obtained with N A D H as reducing substrate are of the same order for human and bovine enzymes (Table II). In order to investigate the possibility that the human enzyme might contain a high proportion of enzyme that is inactive to xanthine (see Discussion), anaerobic bleaching experiments were carried out, as described in the Materials and Methods section. Under anaerobic conditions, the rapid fall in absorbance at 450 nm brought about by addition of xanthine was compared with the maximum fall caused by N a 2 S 2 0 4. In studies with bovine milk enzyme, this ratio is accepted as giving an indication of the percentage of active enzyme
29 [23,24] and, indeed, in our hands, yielded a value of 59% for the commercially-available enzyme (Fig. 4a). Addition of xanthine to the human enzyme, in con-
0.4-
0.3.
trast, showed a detectable fall in absorbance at 450 nm that was very much smaller, and accordingly, more difficult to quantify. This fall represented some 2% or less (after correction for dilution effects) of that brought about by addition of NazSzO 4 (Fig. 4b). With a view to examining the possible presence of desulpho-enzyme in human xanthine oxidase, resulphuration was attempted. A sample of purified human enzyme, with specific xanthine oxidase (Types D + O) activity of 5.0 m U / m g protein was incubated anaerobically with NazS and Na2S204, as described by Wahl and Rajagopalan [16]. After gel filtration on Sephadex G25, the specific activity was found to have increased to 29 m U / m g protein. Discussion
c
.~ 0.2
0.1
360
460
560
e6o
Wavelength (rim)
0.6-
0.5-
0.4-
G m 0.3 .Q .Q ,
