Molecular activation—deactivation of xanthine oxidase in human milk

Molecular activation—deactivation of xanthine oxidase in human milk

r~ BB Biochi ~mic~a et Biophysica A~ta j_. ELSEVIER Biochimica et Biophysica Acta 1245 (1995) 248-254 Molecular activation-deactivation of xanthi...

788KB Sizes 0 Downloads 71 Views

r~

BB Biochi ~mic~a et Biophysica A~ta

j_.

ELSEVIER

Biochimica et Biophysica Acta 1245 (1995) 248-254

Molecular activation-deactivation of xanthine oxidase in human milk Anna-Marie Brown, Mustapha Benboubetra, Michael Ellison, Deborah Powell, John D Reckless, Roger Harrison * Department ~f Biochemistry, Unit'ersit3"¢~(Bath, Bath, BA2 7AY, UK Received 7 February 1995: accepted 9 May 1995

Abstract

Enzymic activity and protein levels of xanthine oxidase were measured in serial samples of breast milk donated by each of 14 mothers, starting, in all but two cases, within 7 days following parturition. Enzyme activity varied widely, usually reaching peak values during the first 15 days and falling thereafter, by as much as 98%, to basal levels that were subsequently largely maintained. Corresponding changes in xanthine oxidase protein levels were not observed and, consequently, the specific activity of xanthine oxidase followed the above pattern. The capacity of human xanthine oxidase to undergo activation-deactivation cycles at the molecular level has important implications, not only for its role in breast milk, but also for its potential as a source of reactive oxygen species in other human tissues. Keywords: Molecular activation: Activation-deactivation cycle; Xanthine oxidase; Reactive oxygen species; (Human milk)

I. Introduction

Xanthine oxidase is a widely distributed molybdoenzyme that, in mammals, is largely localised to epithelium and to capillary endothelial cells in a range of tissues [ 1,2]. Its generally accepted metabolic role is in purine catabolism, catalysing the oxidation of hypoxanthine to xanthine and of xanthine to uric acid. The enzyme has, however, a wide specificity which, together with its rather specific localisation in tissues, suggests other functions. In particular, it can act as a source of reactive oxygen species, passing electrons to molecular oxygen to generate superoxide anion and hydrogen peroxide [3]. This facility has been implicated in anti-microbial defence [1,2] and as a destructive factor in ischaemia-reperfusion damage in a range of pathological states [4]. The latter implication has stimulated a great deal of clinically targetted research activity, which has been based largely on the well-known properties of xanthine oxidase from bovine milk [3] and, to a lesser extent, on enzymes from chicken [5] and rat [6] livers. Despite this research activity, little is known about human xanthine oxidase. In 1986, Krenitsky and coworkers [7] reported a preparation from post-mortem human liver that, though heavily proteolysed, showed cat-

* Corresponding author. Fax: + 4 4 1225 826449. 0304-4165/95/$09.50 © 1995 Elsevier Science B.V. All rights reserved SSDI 0 3 0 4 - 4 1 6 5 ( 9 5 ) 0 0 0 9 3 - 3

alytic properties similar to those of the well-studied cows' milk enzyme. More recently, however, we described the purification of human milk xanthine oxidase with surprisingly different characteristics. This electrophoretically homogeneous enzyme has a subunit molecular weight and a UV-visible spectrum essentially identical to those of bovine milk or rat liver xanthine oxidases, but shows extremely low activity towards conventional reducing substrates such as hypoxanthine or xanthine [8]. NADH, on the other hand, is oxidised as efficiently as by the bovine milk enzyme. These properties of human milk xanthine oxidase can be explained [8] in terms of a predominance of 'inactive' enzyme, known to occur, in relatively minor proportions, in enzymes from cows' milk [3], chicken [9] and rat [10] livers. Such 'inactive' xanthine oxidase has been shown to comprise 'demolybdo' and 'desulfo' forms of the enzyme, which respectively lack molybdenum or contain molybdenum double-bonded to oxygen, rather than to sulfur [3]. Intriguingly, evidence has been presented that, under certain physiological conditions, the 'desulfo' form of chicken liver xanthine oxidase can be activated by sulfur transfer Ill]. Purified human milk xanthine oxidase oxidised xanthine with rates some two orders of magnitude less than those shown by cows' milk enzyme, and the breast milk, used as a source of human enzyme, showed similarly low activity, compared with cows' milk [8]. Indeed, the low xanthine

A.-M. Brown et al. / Biochimica et Biophysica Acta 1245 (1995) 248-254

oxidase activity of human milk has long been known [12], but has been attributed to low content of xanthine oxidase, rather than to enzyme of low activity, as suggested by Abadeh et al. [8]. The question arises as to the role of xanthine oxidase in milk, particularly in human milk. It has been proposed that the enzyme, possibly together with peroxidase [13,14] has a bactericidal function, serving to sterilize either the mammary gland or the neonatal gut. The generally very low activity of human milk xanthine oxidase toward conventional reducing substrat~s clearly poses a problem. It was, accordingly, of interest that xanthine oxidase purified from different batches of breast milk showed widely variable, although always relatively low, activities toward xanthine [8]. This observation, coupled with published reports of higher xanthine oxidase activity in colostrum, compared to that in subsequent milk samples [15], suggested the possibility that the true specific activity of xanthine oxidase in human milk, like that in chicken liver [11], might be subject to control. We therefore undertook a systematic study of serial breast milk samples, measuring enzymic activity towards hypoxanthine or pterin and also xanthine oxidase protein with a view to detecting variations in molecular activity of the enzyme with time post partum.

2. Materials and methods 2.1. Milk and reagents

Breast milk was kindly donated by volunteer mothers in the Bath and Chippenham areas. From one to four samples (approx. 1 ml) were expressed each day at random times, not necessarily when feeding, and stored at - 20°C prior to assay. All donors were free from infection and drug administration apart from one isolated incident (Mother 1, see Discussion). Purified human milk xanthine oxidase was prepared as previously described [8]. Bovine milk was obtained fresh, from Friesian cows in mid-lactation. Unless otherwise specified, other reagents were purchased from Sigma, Poole, Dorset, UK. 2.2. Assays f o r xanthine oxidase activity

The radiometric assay involved oxidation of [J4C]hypoxanthine to xanthine plus uric acid, essentially as described for liver homogenates by Ghezzi et al. [16]. Briefly, milk was allowed to thaw (if necessary) and samples (1-10 /zl) were diluted (to 20/xl) with 0.1 M sodium pyrophosphate (pH 8.3), and incubated for 5 - 6 min at 20°C with [8~4C]hypoxanthine (5 /xl, 9 /xCi/ml; Amersham International, Amersham, UIO. For measurements of total (oxidase plus dehydrogenase) activity, 300 /xM NAD + was added prior to incubation. The enzymic reaction was stopped by addition of 1 M perchloric acid (10 /zl) and the mixture was centrifuged (3000 X g, 5 rain) to remove precipitated

249

protein. The supernatant (2-5 /xl) was applied to cellulose thin-layer sheets (Sigma) and subjected to ascending chromatography in a closed tank containing butanol, methanol, water and ammonia solution (sp. gr. 0.88) in the proportions 60:20:20:1. The sheets were allowed to dry and the individual lanes were cut apart. Areas corresponding to hypoxanthine (R F approx. 0.36) and xanthine plus uric acid (R F approx. 0.10-0.16) were separately counted for radioactivity in a liquid scintillation counter (TRI-CARB 1600 TR, Packard), with 96% efficiency, by using Optiphase scintillant (5 ml) and averaging counts over 5 min. Xanthine oxidase activity is expressed as mU, where 1 mU is defined as 1 nmol hypoxanthine converted to xanthine plus uric acid per min. All assays were done in duplicate or triplicate and results were within 2-5% of each other. Assays were shown to be linear with both time and volume of milk. All activity was shown to be oxidase type (no increase in rate on inclusion of NAD +) and could be totally abolished by addition of 60 /zM allopurinol. In addition to most of the human milk samples (Table 1), bovine milk was assayed by this method. Xanthine oxidase activity was shown to be stable (i.e., to decrease < 5%) after up to five freeze-thaw cycles in seven samples taken from different mothers. The fluorimetric assay for xanthine oxidase activity involved oxidation of pterin to isoxanthopterin, essentially as described for tissue homogenates by Beckman et al. [17]. Milk was allowed to thaw and a sample (50 /xl) was diluted (to 500 /xl) in 50 mM potassium phosphate (pH 7.4), containing 0.1 mM EDTA (assay buffer), in an 800 /xl quartz cuvette at 20°C. The cuvette was placed in an LS-5B Perkin-Elmer Luminescence spectrophotometer with excitation wavelength 345 nm and emission wavelength 390 nm and using a 5 mm slit width. When the baseline was stable, 1 mM pterin in assay buffer (5 /zl) was added and the rate of increase of fluorescence was monitored for 2-10 min. For measurements of total (oxidase plus dehydrogenase) activity, 1 mM Methylene blue in assay buffer (5 /xl) was added at this stage and the rate of increase of fluorescence was again followed for 2-10 min. 0.01 mM isoxanthopterin in assay buffer (5 /xl) was then added to the reaction mixture and the increase in fluorescence was used to calibrate the assay (the exact concentration of isoxanthopterin was determined spectrophotometrically by using 1~336 13.0 mM -t c m - t ) . Enzyme activities are expressed as mU where 1 mU is defined as 1 nmol isoxanthopterin formed per min. All assays were done in duplicate or triplicate and the results were within 2-5% of each other. All activity was shown to be of oxidase type (no increase in rate on addition of Methylene blue) and could be totally abolished by addition of 5 mM allopurinol in the assay buffer (5 /xl). Although different substrates are oxidised in the radiometric and fluorimetric determinations, these assays are directly comparable. Application of both assays to serial dilutions of a single sample of human xanthine oxidase =

250

A.-M. Brown et al. / Biochimica et Biophysica Acta 1245 (1995) 248-254 layer was combined with Fraction A, above, and vortexmixed. The mixed suspension (0.5-1.0 ml) was added to an equal volume of Sample buffer (62.5 mM Tris-HC1 (pH 6.8) containing 6% ( w / v ) SDS, 50% ( v / v ) glycerol, 0.1% 2-mercaptoethanol and 0.005% ( v / v ) Bromophenol blue) and samples (100 /zl) were subjected to polyacrylamide gel electrophoresis in 7% and 10% gels as described by Laemmli [19]. Intensity of the Coomassie Blue-stained band corresponding to xanthine oxidase (Fig. l a) was measured by an LKB 2202 Ultrascan laser densitometer. Varying concentrations of pure human milk xanthine oxidase [8], showing essentially a single 150 kDa band on SDS-PAGE, were run on every gel to enable construction of a standard curve from which the concentration of enzyme in the milk extract could be calculated. The validity of using Coomassie blue staining of the 150 kDa band in this way was supported by parallel experiments with mouse monoclonal anti-(human xanthine oxidase) antibody. A gel with the above components was subjected to Western blotting with H3G7 anti-(human xanthine oxidase) antibody as described by Abadeh et al. [20] except that diaminobenzidine was used as final substrate. The nitrocellulose sheet was made transparent by incubation for 1 min at room temperature in xylene and the intensity of the stained 150 kDa band (Fig. l b) was used, as before, to estimate the concentration of xanthine oxidase in the human milk extract. Values obtained were within

established that 1 mU in the fluorimetric assay corresponds to 12.2 mU in the radiometric assay. Accordingly, for purposes of comparison, activity values obtained fluorimetrically (samples from Mothers 13, 14) are converted throughout to the radiometric assay equivalents. 2.3. Assay for alkaline phosphatase actiuib' Alkaline phosphatase activity was measured in terms of the hydrolysis of para-nitrophenyl phosphate at 32°C, exactly as described by Kitchen [18]. Activity is expressed in U where 1 U is defined as 1 /zmol para-nitrophenyl phosphate hydrolysed per min. 2.4. Assay for xanthine oxidase protein Xanthine oxidase protein was determined by densitometric measurement of the 150 kDa band on SDS-PAGE patterns of total milk extracts. Milk was allowed to thaw to room temperature and was centrifuged (3000 X g, 15 min) to yield an aqueous phase (Fraction A) and a cream supernatant. The cream layer was mixed with 15% ( v / v ) butanol and shaken at 35°C for 3 min. Centrifugation (3000 X g, 15 min) yielded a proteinaceous layer and a butanol-oil supernatant, which was discarded. The protein layer was mixed with 15% ( v / v ) butanol, shaken at 35°C for 10 min and centrifuged as above. The lower protein

1 5 0 KD " ' ' ' ~

123 ao

45

67

1 234567 b,

Fig. I. SDS-Polyacrylamidegels of human milk xanthine oxidase Lanes 1-5, purified human enzyme (1-5 /zg, respectively);lanes 6,7, extracts from human milk samples from Mother I (samplesS1 and $5 respectively(Fig. 2d)). (a) Coomassieblue stained; (b) Western blots (Materials and methods).

A.-M. Brown et al. / Biochimica et Biophysica Acta 1245 (1995) 248-254

251

Table 1 Xanthine oxidase activities in serial samples of breast milk taken from different mothers Mother:

1

2

Period of collection (days post partum): 2-28 2-9 Xanthine oxidase activity ( m U / m l ) : mean 0.40 0.87 maximum 1.71 3.20 minimum 0.01 0.07

3

4

5

6

7

8

9

10

11

12

13 *

14 *

2-15

2-13

6-15

3-14

3-8

6-18

4-8

12-38

7-18

29-87

6-17

1-25

3.47 8.33 0.20

0.64 1.40 0.15

0.84 1.83 0.13

0.23 0.65 0.05

0.97 2.65 0.15

0.52 1.42 0.13

0.84 1.38 0.17

0.02 0.21 0.01

0.65 1.84 0.10

0.03 0.18 0.01

1.40 5.42 0.22

1.79 7.69 0.24

All samples were assayed radiometrically using labelled hypoxanthine (Materials and methods) apart from those marked with an asterisk *, which were assayed fluorimetrically by oxidation of pterin, in which cases the quoted values are the radiometric assay equivalents (Materials and methods).

1-2% of those obtained by using direct staining with Coomassie blue. Using the Coomassie blue staining for quantification of xanthine oxidase activity in milk extracts, intraplate (7 samples) and interplate (10 samples) coefficients of variation were 4% and 6%, respectively.

by radiometric assay of fresh bovine milk ( 8 0 _ 5.0 m U / m l (mean + S.E., n = 3)), varied greatly with time post partum. A general pattern emerged, whereby enzymic activities, to hypoxanthine or pterin, reached a peak within the first days post partum. Typical examples of this pattern are shown in Fig. 2. Activities remained relatively high for several days before falling, by as much as 98%, to basal levels that were subsequently largely maintained, although occasional smaller peaks, with no clear pattern, did occur during this latter period (e.g., Fig. 2d, but see Discussion). Basal levels were usually achieved by day 15. In one case (Fig. 2b), however, the whole pattern occurred relatively late, with activities peaking between days 10-20. In several cases (e.g., Fig. 2c,d), the first samples measured

3. Results

Xanthine oxidase activity was determined in serial samples of breast milk from 14 individual mothers, taken over periods of 4-58 days and starting, in all but two cases, within the first week after parturition (Table 1). Levels of enzyme activity, while very much lower than those shown

300

300 ~. ~. E E o o N ~

a

250

250

200

200

--~. E ;

15o .~

150'

1800

-1800

•o . IQ . 1600

b • 1600 • 1400

-~ 12oo.

• 1200

"-- I ooo,

.1000

.~, =

800.

800

o o

400200-

400

100

~ ~

50

5o.

600 x

X

0

0 1

200

0

1'0 1'5 Time post portum (days)

Time past partum (days) 160~o Q.O.

120-

.~ ~

loo-

~--~>- c

• ':

.

:" ........."

~ :

i

i

/i :

.100 i

~g

lOO-

"

-~ -~ 8o. ~

60-

'---"

120

q ',..

,

, 4bq ~=Ji == '

g

100 80

e

8O

60

60-

60

,o. O

120.

E E

120

.,

~:/

80-

140

"c .140

¢ /

E E

25

140

.160

,

140-

2'0

a.

10

12

1,

Time post portum (day=)

(6

40

>o ~

40-

- 40

20

o

20-

-20

0 18

0

3

5

7 9 11 13 15 17 19 21 23 25 27 29 Time post porfum ( do)'= )

0

Fig. 2. Xanthine oxidase (XO) activities ( 0 - - 0 ) , xanthine oxidase protein levels ( • • ) and alkaline phosphatase (AP) activities ( • • ) in serial milk samples from individual donors. Values are expressed as percentages of the initial values which were as follows: (a) Mother 3 (Table 1) 3.10 m U / m l , 0.10 m g / m l , 20.90 U / m l ; (b) Mother 14, 0.48 m U / m l , 0.17 m g / m l ; (c) Mother 13, 5.42 m U / m l , 0.09 m g / m l ; (d) Mother 1, 1.70 m U / m l , 0.16 m g / m l . S 1 - $ 5 refer to individual samples used for SDS-PAGE (SI, $5; Fig. 1) or for checking the possible presence of inhibitor (S1-$4; see Discussion). XO activities are all quoted in terms of the radiometric assay (Materials and methods).

252

A.-M. Brown et aL / Biochimica et Biophysica Acta 1245 (1995) 248-254

showed the maximum activities, possibly because, for practical reasons, earlier samples were not available. Alkaline phosphatase, like xanthine oxidase, is a constituent of the milk fat globule membrane and its activity was monitored as an indicator of total membrane content [18]. In contrast to xanthine oxidase activity, which was seen to vary as much as 50-fold (e.g., see Fig. 2a), alkaline phosphatase activity, monitored in 10 of the 14 mothers, never varied by more than 40% from that of the first sample. A typical example is shown in Fig. 2a. In no case was there any correlation with xanthine oxidase activity. Despite the low activity of xanthine oxidase in human milk, the levels of enzyme protein (0.1 + 0.04 m g / m l (mean + S.E. of means from 10 mothers)) were similar to those (0.1 + 0.03 m g / m l (mean + S.E., n = 5)) found in bovine milk. Human xanthine oxidase protein levels never varied by more than 25% from that of the first sample and did not correspond to the peaks of total activity (Fig. 2). In consequence, profiles of specific activity for xanthine oxidase (Fig. 3) are very similar to those of total activity (Fig. 2), showing higher levels during the first 10 (or, in one case, 20) days post partum and falling thereafter.

0-

100.

~

so-

4. D i s c u s s i o n

Xanthine oxidase activities of human milk determined in this study (Table 1) are similar to values (0.16-9.2 m U / m l ) observed by Zikakis et al. [15], who used a similar radiometric method to assay milk samples taken, at widely varying stages of lactation, from 59 donors. Such levels are from one to two orders of magnitude less than those determined in fresh bovine milk, which, nevertheless, appears to contain similar levels of xanthine oxidase protein. These observations further support the contention that the low activity of purified human milk xanthine oxidase towards xanthine or pterin is an intrinsic property of the human milk enzyme and not a preparation artefact [8]. Xanthine oxidase activity in breast milk, whether assayed radiometrically or fluorimetrically, showed great variation in the first weeks post partum, reaching peak values usually within the first 10 days, and falling to basal levels thereafter. Peak values could be as much as 50-fold higher than basal levels. Such variations have been little documented, although Bradley and Gunther [21] briefly noted similar results from milk samples of three mothers in

o

[3.

50

b

40 Ill

E

6o-

30 --I

E :~

40-

°°

20-

O x o

~

E

v

20 o

10

0 x

.2

o

l

I

#

I

3

5

7

9

=o

Time post parfum ( days )

o

~:

60

®

50.

==

40-

E

E

"~ ...1

E

"-"

30-

"6

20-

_e,

v

o

= n

I

20

25

15

d

10-]

E

g

5-

o o

o x u

I

15

@

70

a.

OI

I

10

Time post padum ( days )

m

o

I

5

10-

o

.2

0 8

10

12

14

Time post parfum ( days )

16

18 a. el

0 3

5

7

9

11

13

15

17

19

Time poet porfum ( days )

Fig. 3. Specific (molecular) xanthine oxidase (XO) activities (radiometric assay values) corresponding to Fig. 2. (a) Mother 3; (b) Mother 14; (c) Mother 13; (d) Mother 1.

A.-M. Brown et al. / Biochimica et Biophysica Acta 1245 (1995) 248-254

which activity peaked at 3 days post partum, falling by up to 80% by day 7, while Zikakis et al. [15] observed that colostrum showed higher activity than did mature milk. Xanthine oxidase is a major protein component of the bovine milk fat globule membrane [22] and the question arises as to whether variations in its enzymic activity simply reflect different contents of membrane. This was shown not to be the case, in that levels of alkaline phosphatase activity, taken as a marker for milk fat globule membrane, varied little over the relevant time scale. More surprisingly, enzyme activity showed no correlation with xanthine oxidase protein levels, which remained relatively constant. The consequence of these observations is that the true specific activity of xanthine oxidase is seen to vary dramatically with time, essentially following the pattern of enzyme activity expressed in terms of milk volume. These findings could be explained by the presence of an endogenous inhibitor or inactivator of xanthine oxidase in those milk samples showing low activity. Indeed, the presence of low-molecular-weight inhibitors has been detected in homogenates of tissues from several species [23]. This possibility was investigated by combining samples of relatively high activity towards hypoxanthine, each with an equal volume of a sample of low activity (Mother 1: S 1 and $2, S I and $3, $2 and $4 (Fig. 2d)). Assay of the combined sample showed in each case activity approx. 50% of that of the higher activity sample (data not shown), a finding inconsistent with the presence of an inhibitor in the low activity sample. In the absence of any evidence for the presence of endogenous inhibitors in the milk samples, the most plausible explanation of our data is that they reflect post-translational activation-deactivation cycles of the enzyme. This demonstration of activation and deactivation in human milk xanthine oxidase may shed light on the question as to why an apparently inactive (to most reducing substrates, apart from NADH [8]) enzyme should occur in breast milk. Although the maximum specific activity detected in the present study (approx. 80 m U / m g enzyme protein) is still an order of magnitude less than values determined in bovine milk, it is nevertheless two orders of magnitude higher than the levels commonly found in breast milk taken later than 3 weeks post partum, and strongly suggests a functional role for the enzyme in the early weeks. Such a role could involve bactericidal protection of the neonatal gut, as previously proposed [13,14]. It is difficult to do other than speculate as to the factors initiating these changes in the molecular activity of milk xanthine oxidase, but response to hormonal variation is clearly a possibility. Levels of progesterone, oestradiol and cortisol fall sharply post partum, while levels of prolactin rise. Maximal changes occur around the 4th or 5th day after delivery and are virtually completed by day 10 [24,25], a time scale very similar to that of most of the enzyme activity changes observed in the present study. In this context, it is worth noting that, in the case of Mother

253

I, a significant peak of specific xanthine oxidase activity was, surprisingly, observed at day 15 (Fig. 3d), which was subsequently found to coincide with her starting a course of progesterone-containing contraceptive pills. It would be of interest to examine the effects of the above hormones on the specific activity of xanthine oxidase in human mammary epithelial cells in vitro. The enzyme in other species also has been shown to undergo activation and deactivation at the molecular level, albeit to a lesser extent than demonstrated here in human milk. Itoh et al. [i 1] reported that the specific activity of chicken liver xanthine dehydrogenase rose and fell again in response to raising and lowering the protein content of the chicken's diet, while Furth-Walker and Amy [26] showed similar activation of rat liver xanthine oxidase on exposure to a high protein diet. Both groups of workers attributed these changes to variations in the content of inactive isoforms, known to make up 30-50% of the enzyme in each case [9,10]. The activation resulting from switching from normal to high protein diet was about 1.5-fold, for chicken, and 4-fold, for rat liver, in each case markedly less than the values of up to 50-fold observed in the present study. However, the properties of purified human milk xanthine oxidase suggest that, in this case, the content of inactive enzyme could be more than 98% [8], which clearly potentially allows many-fold greater activation. The greater proportion of inactive xanthine oxidases and dehydrogenases is made up of 'desulfo' enzyme [3,810], in which the molybdenum atom is covalently doublebonded to oxygen, rather than sulfur. In the case of chicken liver xanthine dehydrogenase, Itoh et al. [11] proposed that the activation-inactivation cycles reflected, specifically, interconversion of 'desulfo' and 'sulfo' forms of the enzyme. Nishino and co-workers [27] went on to demonstrate that the mitochondrial enzyme, rhodanese, could catalyse the reversible interconversion of 'sulfo' and 'desulfo' bovine milk xanthine oxidase, in the presence of thiosulfate and sulfhydryl reagent. In further studies, Nishino [28] showed that the cytosolic, and hence physiologically more relevant enzyme, mercaptopyruvate sulfur transferase, could catalyse the same interconversions, and proposed mechanisms whereby sulfide ion, made available by sulfur transferases, might interact with 'desulfo' or 'sulfo' xanthine oxidase. Furth-Walker and Amy [26] similarly interpreted their data, obtained with rat liver, in terms of 'desulfo'-'sulfo' enzyme conversion. In support of this interpretation, the latter workers showed that supplementation of the rats' diet with the sulfur-containing amino acid, methionine, was particularly effective in increasing the specific activity of xanthine oxidase in rat intestine. On the basis of data obtained from xanthine oxidase and other enzymes, Coughlan [29] proposed that the reversible incorporation of sulfur might serve to regulate the function of several proteins, and it may well be that human milk xanthine oxidase, in which exceptionally low proportions

254

A.-M. Brown et al. / Biochimica et Biophysica Acta 1245 (1995) 248-254

of active enzyme can occur, constitutes a particularly sensitive substrate for such regulation. The apparent capacity of human milk enzyme to undergo activation-deactivation cycles, at the molecular level, has important implications for the role of xanthine oxidase in human tissues. Immunoaffinity purification of xanthine oxidase from human heart indicates that this enzyme has properties very like those of that from breast milk [30], and it may well be that xanthine oxidases in many other human tissues are similar. These findings extend the possibilities for control of xanthine oxidase activity in vivo with consequent potential for the regulation of reactive oxygen species, known to be involved in both pathological and normal physiological processes.

Acknowledgements We thank the Arthritis and Rheumatism Council for financial support (M.B.) and the Science and Engineering Research Council for research studentships (A-M.B., D.P.)

References [1] Jarasch. E.-D., Bruder, G. and Heid, H.W. (1986) Acta Physiol. Scand. Suppl. 548, 39-46. [2] Gossrau, R., Frederiks, W.M. and Van Noorden, C.J.F. (1990) Histochemistry 94, 539-544. [3] Bray, R.C. (1975) in The Enzymes (Boyer, P.D., ed.), 3rd Edn. Vol XI1, pp. 299-419, Academic Press, New York. [4] Sussman, M.S., Schiller, H.J., Buchman, T.G. and Bulkley, G.B. (1992) in Multiple System Organ Failure (Fry, D.E., ed.), pp. 143-165, Mosby Year Book, St. Louis. [5] Nishino, T. and Nishino, T. (1989) J. Biol. Chem. 264, 5468-5473. [6] Amaya, Y., Yamazaki, K.-I., Sato, M., Noda, K., Nishino, T. and Nishino, T. (1990)J. Biol. Chem. 265, 14170-14175.

[7] Krenitsky, T.A., Spector, T. and Hall, W.W. (1986) Arch. Biochem. Biophys. 247, 108-119. [8] Abadeh, S., Killacky, J., Benboubetra, M. and Harrison, R. (1992) Biochim. Biophys. Acta 1117, 25-32. [9] Nishino, T., Itoh, R. and Tsushima, K., (1975) Biochim. Biophys. Acta 403, 17-22. [10] Ikegami, T. and Nishino, T. (1986) Arch. Biochem. Biophys. 247, 254-260. [11] Itoh, R., Nishino, T., Usami, C. and Tsushima, K. (1978)J. B iochem. 84, 19-26. [12] Rodkey, F.L, and Ball, E.G. (1946) J. Lab. Clin. Med. 31,354-356. [13] Bjorck, L. and Claesson, O. (1979) J. Dairy Sci. 62, 1211-1215. [14] Moldoveanu, Z., Tenovuo, J., Mestecky, J. and Pruitt, K.M. (1982) Biochim. Biophys. Acta 718, 103-108. [15] Zikakis, J.P., Dougherty, T.M. and Biasotto, N.O. (1976) J. Food Sci. 41, 1408-1412. [16] Ghezzi, P., Bianchi. M., Gianera, L., Landolfo, S. and Salmona, M. (1985) Cancer Res. 45, 3444-3447. [17] Beckman, J.S., Parks, D.A., Pearson, J.D., Marshall, P.A. and Freeman, B.A. (1989) Free Rad. Biol. Med. 6, 607-615. [18] Kitchen, B.J. (1974)Biochim. Biophys. Acta 356, 257-269. [19] Laemmli, U.K. (1970) Nature 227, 680-685. [20] Abadeh, S., Rogers, A. and Harrison, R. (1993) Biochem. Soc. Trans. 21, 100S. [21] Bradley, P.L. and Gunther, M. (1960) Biochem. J. 74, 15P. [22] Patton. S. and Keenan, T.W. (1975) Biochim. Biophys. Acta 415, 273-309. [23] Terada, L.S., Left, J.A. and Repine, J.E, (1990) Methods Enzymol. 186, 651-656. [24] Harris, B., Lovett, L., Newcombe, P.G., Read, G.F., Walker, R. and Riad-Fahmy, D. (1994) Br. Med. J. 388, 949-951. [25] Martin, M.C. and Hoffman, P.G. (1986) in Basic and Clinical Endocrinology (Greenspan, F.S. and Forsham, P.H., eds.), pp. 476500, Lange, Los Altos, CA. [26] Furth-Walker, D. and Amy, N.K. (1987) J. Nutr. 117, 1697-1703. [27] Nishino, T., Usami, C. and Tsushima, K. (1982) Proc. Natl. Acad. Sci. USA 80, 1826-1829. [28] Nishino, T. (1986) in Purine and Pyrimidine Metabolism in Man, V (Nyham, W.L., Thompson, L.F. and Watts, R.W.E., eds.), pp. 259-262, Plenum Press, New York. [29] Coughlan, M.P. (1981) Biochem. Soc. Trans. 9, 307-308. [30] Abadeh, S., Case, P. and Harrison, R. (1993) Biochem. Soc. Trans. 21, 99S.