An ultrafiltration assay for lysyl oxidase

ANALYTICAL

BIOCHEMISTRY

185,359-362

An Ultrafiltration David

R. Shackleton

(19%))

Assay for Lysyl Oxidase’ and David

J. S. Hulmes

Department of Biochemistry, University of Edinburgh Medical School,Hugh RobsonBuilding, GeorgeSquare,Edinburgh EH8 9XD, United Kingdom

Received

August

31,1989

A modification of the original microdistillation assay for lysyl oxidase is described in which Amicon C-10 microconcentrators are used to separate, by ultrafiltration, the 3H-labeled products released from a [4,6-‘HIlysine-labeled elastin substrate. Enzyme activity is determined by scintillation counting of the ultrafiltrate, after subtraction of radioactivity released in the presence of &aminopropionitrile, a specific inhibitor of the enzyme. Conditions are described which optimize both the sensitivity and the efficient use of substrate. The assay shows linear inhibition of activity in up to 1 M urea; hence, as the enzyme is normally diluted in the assay, samples in 6 M urea can be assayed directly, without prior dialysis, and corrected for partial inhibition. Comparable results are obtained when enzyme activity is assayed by ultrafiltration or microdistillation. The assay is simple and convenient and, by using disposable containers throughout, it eliminates the need for time-consuming decontamination of radioactive glassware. 0 1990 Academic Press. Inc.

Lysyl oxidase (protein-lysine 6-oxidase, EC 1.4.3.13) is a copper-dependent enzyme of connective tissue that initiates crosslinking in collagens and elastin by oxidative deamination of certain lysine (or hydroxylysine) residues (9,14). The original assay for lysyl oxidase (14) involves incubation of enzyme with a collagen or elastin substrate that has been labeled biosynthetically with either [6-3H]lysine or, more recently, [4,5-3H]lysine. As a result of the reaction, tritium exchanges with water and the 3HH0 formed is separated by microdistillation (4,5,7,10,13,16-19). In a modification of this assay, the 3HH0 is separated by ion-exchange chromatography (2,12). Other methods, using synthetic polypeptide substrates (6), following oxygen uptake with an oxygen electrode (11) or measuring H202 release by a fluorometric ’ Supported

by the Arthritis

and Rheumatism

0003~2697/90 $3.00 Copyright 0 1990 by Academic Press, All rights of reproduction in any form

Inc. reserved.

Council.

peroxidase-coupled reaction (8), are useful for studies on enzyme specificity and mechanism, but are inappropriate for routine assays. Microdistillation is cumbersome and requires thorough decontamination of the equipment following each assay. This limits the number of assays which can be performed simultaneously and the frequency with which the apparatus may be used. The ion-exchange method (12) is more convenient, but relatively large amounts of substrate are required due to dilution of reaction products and quenching by TCA.2 Here we describe a new assay for lysyl oxidase. The assay is similar to most others in the initial incubation with a 3H-labeled substrate, but the 3HH0 formed is separated by ultrafiltration using Amicon C-10 microconcentrators. No dilution occurs, and since all containers used are disposable, the need for decontamination is eliminated. MATERLALS

AND

METHODS

Materials Fertile hen eggswere supplied by Ross Breeders, Newbridge, Scotland. [4,5-3H]Lysine (1.18 TBq/mol) was purchased from Amersham International, Little Chalfont, England, C-10 microconcentrators were from Amicon, Stonehouse, England, Ultrafree-MC filter units (10,000 nominal MW limit cutoff) were from Millipore, Watford, England; BAPN (fumarate salt) was from Sigma Chemicals, Poole, England; MEM Select-Amine kit was from GIBCO, Paisley, Scotland; DEAE-Sephaccl and Sephacryl S-200 were from Pharmacia, Milton Keynes, England; all other chemicals (analytical grade) were from BDH, Poole, England. Preparation of [3H]Ekzstin Substrate Lathyritic [4,5-3H]lysine-labeled elastin was prepared as described elsewhere (5,17), from cultured 17-day ’ Abbreviations acetic acid; MEM,

used: BAPN, @-aminopropionitrile; minimum essential medium.

TCA,

trichloro-

359

360

SHACKLETON

chick embryo aortas, but using culture medium that was supplemented with the most abundant amino acids in elastin (15), alanine, glycine, proline, and valine. Briefly, 120 aortas were preincubated for 1 h at 37°C in 125 ml Nalgene flasks (30 aortas per flask) containing sterile MEM (50 ml per flask) without lysine but supplemented with 50 mg/liter alanine, 50 mg/liter glycine, 50 mg/liter proline, 50 mg/liter valine, 50 mg/liter ascorbic acid, 50 mg/liter BAPN, lo5 units/liter penicillin, 100 mg/liter streptomycin, and adjusted to pH 7.4. The aortas were then incubated on an orbital shaker for 20 h in 200 ml of the above medium (50 ml per flask) containing 0.56 MBq/ml [4,5-3H]lysine at 37°C in an atmosphere of 5% CO,/95% air. The following procedures were at 0-4°C. The aortas were homogenized in 50 ml 0.15 M NaCl and then centrifuged 5 min at 10,OOOg. The pellet was retained and this extraction step was repeated. To inactivate endogenous lysyl oxidase (9), the pellet was further homogenized in 1 N HCl and centrifuged as above, and this step was repeated. The pellet was then homogenized in assay buffer (0.1 M Na2B407, 0.15 M NaCl, pH 8.0) and centrifuged as above, and this step repeated. Finally the pellet was resuspended in assay buffer to give a concentration of 3 X lo5 dpm in 100 ~1. A typical yield from 10 dozen embryos was 2.4 X lOa dpm, which was sufficient for 800 assays. The yield was approximately twice that obtained without amino acid supplements.

Preparation

HULMES

g ;

1000

-

ewme

I= 3 5

enzyme

+ BAPN

d

4

0

8 TIME

12

16

12

16

(h)

1500)

E s

.b

0

4

8 TIME

(h)

FIG. 1.

Time dependence of total radioactivity in the ultrafiltrate. [3H]Elastin substrate (300,000 dpm) was incubated with enzyme at 37°C in an assay volume of 900 ~1 for up to 16 h. After centrifugation and ultrafiltration of the supernatant, 3H radioactivity in 400 pl of the ultrafiltrate was measured (a) with and without 50 pg/ml BAPN in the assay. (b) Corrected lysyl oxidase activity, obtained by subtracting radioactivity with enzyme + BAPN from radioactivity with enzyme alone. Error bars are +SD (n = 3).

of Lysyl Oxidase

All procedures were at 0-4°C. Piglet skins (approximately 450 g wet wt) were cut into small pieces with scissors and then homogenized in a Waring Blendor in 900 ml 0.1 M sodium phosphate, 0.15 M NaCl, pH 7.8 (Buffer A). The homogenate was centrifuged at 10,OOOg for 20 min and the pellet was rehomogenized and centrifuged as above. After washing the pellet three times in 900 ml 10 mM sodium phosphate, pH 7.8 (Buffer B), the pellet was extracted three times, by homogenization and centrifugation as above, in 900 ml 6 M urea, 10 mM sodium phosphate, pH 7.8 (Buffer C). Pooled supernatants from the 6 M urea extraction were filtered through Whatman No. 1 filter paper and lysyl oxidase was partially purified from this crude extract by ion-exchange chromatography (1) on DEAE-Sephacel. Enzyme was further purified by selective binding to Sephacryl S-200 in Buffer B and elution with Buffer C. Further details of the enzyme preparation will be presented elsewhere (21).

RESULTS

Development

AND

of the Assay

Optimum assay conditions were determined using a range of incubation times and substrate amounts. In a

typical assay, 100 ~1 [3H]elastin substrate suspended in assay buffer was placed in a 1.5-ml microcentrifuge tube. To this, 700 ~1 assay buffer and 100 ~1 enzyme were added. The reaction mixture was vortexed and then incubated at 37°C for various times. Following incubation, the reaction was stopped by placing the microcentrifuge tube on ice. The bulk of the elastin substrate was then pelleted by centrifugation at 15,600g for 5 min at 2°C. From the supernatant, 700 ~1 was transferred to an Amicon C-10 microconcentrator. The microconcentrator was centrifuged at 5000g for 1 h at 2°C in a Beckman JA-20 fixed angle rotor, using adaptors of 17 X 100 mm internal diameter and length, respectively. During centrifugation, low molecular weight material passed through the ultrafiltration membrane (molecular weight cutoff 10,000) and collected in the lower chamber of the microconcentrator. To measure tritium release, 400 ~1of the ultrafiltrate was mixed with 3.6 ml BDH Cocktail “T” and then counted in a liquid scintillation counter for 10 min. Figure la shows the results obtained when a fixed amount of elastin substrate (3 X 10’ dpm) was incubated with a fixed amount of enzyme for 0 to 18 h. Badioactivity in the ultrafiltrate increased with time. Also shown in Fig. la are control data with enzyme but in the pres-

ULTRAFILTRATION

ASSAY

enzyme

“0

200 VOLUME

400 600 OF SUBSTRATE

LYSYL

361

OXIDASE

BAPN controls as described earlier. Since control radioactivity increased linearly while lysyl oxidase activity approached saturation, the ratio of lysyl oxidase-released radioactivity to control radioactivity (signal-tonoise) decreased with increasing amounts of substrate (Fig. 2b). This favored the use of low substrate concentrations, but the particulate nature of the elastin substrate gave rise to pipetting errors at low substrate concentrations. Use of 100 ~1 substrate containing 300,000 dpm reduced this variability to less than 3% and gave acceptable ratios of signal-to-noise. Hence this amount of substrate was used routinely.

- a

e 6000 CT s

FOR

+ BAPN

600 (pi)

Effect of Urea

0’

0

.

’

200

VOLUME

.

’

400

.

’

600

OF SUBSTRATE

’

600

’

(~1)

FIG. 2. Dependence of radioactivity in the ultrafiltrate on substrate amount. Different volumes of elastin substrate, containing 3 X 10s dpm/ml, were incubated in assay buffer, in a volume of 906 pl, with a fixed amount of enzyme, for 16 h at 37°C. After centrifugation and ultrafiltration of the supernatant, 3H radioactivity in 400 ~1 of the ultrafiltrate was measured (a) with or without 50 pg/ml BAPN in the assay. Error bars are +SD (n = 3). (b) Signal-to-noise ratio (ratio of corrected, lysyl oxidase released radioactivity (see legend to Fig. lb) to radioactivity released with enzyme in the presence of BAPN) versus substrate concentration. Each point is the mean of three observations.

ence of 50 pg/ml BAPN, a highly specific inhibitor of lysyl oxidase (20). Radioactivity in the controls was much less than that with enzyme in the absence of BAPN, but there was also an increase with time which leveled off after about 4 h. There was no significant difference between controls without enzyme (blanks, not shown) and with enzyme in the presence of BAPN. When radioactivity in the controls was subtracted from radioactivity with enzyme in the absence of BAPN, the data shown in Fig. lb were obtained. When corrected for nonspecific tritium release in this way, lysyl oxidase activity was linear up to 4 h, but thereafter the reaction proceeded more slowly, perhaps due to instability in the enzyme. Since the ratio of total radioactivity to control radioactivity increased with time, longer incubation times were preferred, especially for samples with low activity, when sensitivity rather than linearity was important. Routinely, an incubation time of 16 h was used. Experiments were also carried out to determine the optimum substrate concentration in the assay. Figure 2a shows total radioactivity in the ultrafiltrate obtained when different amounts of substrate were incubated with a fixed amount of enzyme for 16 h. Also shown are

Urea is usually present during extraction and purification of lysyl oxidase. Furthermore, the enzyme is often stored in urea, in which it is stable. However, urea inhibits lysyl oxidase activity, so assays are usually carried out after removal of urea by dialysis. It has been shown that while lysyl oxidase is totally inhibited by urea concentrations greater than 1.2 M, below 1.2 M the degree of inhibition is proportional to urea concentration (5). Therefore it was of interest to examine the effect of urea on the ultrafiltration assay. Figure 3 shows the results obtained when standard assay conditions (i.e., 300,000 dpm [3H]elastin, 16 h incubation) were used in the presence of different concentrations of urea. For each assay, 100 ~1 urea-free enzyme was used, while different urea concentrations were achieved by adding urea to the assay buffer. When corrected for control radioactivity in the presence of 50 pg/ ml BAPN, the percentage inhibition was found to be linear for urea concentrations up to 1 M. Hence enzyme ac-

1007 ,

0.0

0.3

0.6

0.9

1.2

UREA (M) FIG. 3. Inhibition of lysyl oxidase activity by urea. [3H]Elastin substrate (300,000 dpm) was incubated at 37’C with a fixed amount of urea-free enzyme in an assay volume of 900 ~1 for 16 h. Urea concentrations were varied by addition of urea to the assay buffer. After centrifugation and ultrafiltration of the supernatant, 3H radioactivity in 400 ~1 of the ultrafiltrate was counted. Corrected enzyme activity was obtained by subtracting radioactivity released with enzyme + BAPN from radioactivity with enzyme alone. Data from assays containing urea were calculated as a percentage of activity without urea and then expressed as percentage inhibition compared to those of the urea-free assay. Each point is the mean of three observations.

362

SHACKLETON TABLE

1

Comparisonof the Ultrafiltration and Microdistillation Assays Ultrafiltration kbm) Substrate alone With enzyme Enzyme with BAPN

Distillation (dpm)

1221+ 70 3368 + 205 12585 44

1267 + 135 3379 * 145 10362 92

Note. [3H]Elastin substrate (9 X 10’ dpm) was incubated at 37”C, with or without enzyme, or in the presence of enzyme and 50 ag/ml BAPN, in an assay volume of 2.7 ml, for 16 h. For ultrafiltration, 900 pl of the incubation mixture was taken and treated as described in the text. For microdistillation, 180 pl of 50% TCA was added to the remainder of the incubation mixture, which was then transferred to a conventional microdistillation apparatus (3). Four hundred microliters of both the ultrafiltrate and the distillate was measured by liquid scintillation counting. The radioactivity in the distillate was corrected for the volume increase caused by the addition of TCA. Errors are +SD (n = 3).

tivity in the absence of urea could be calculated knowing the final urea concentration in the assay.

Comparison

with the Distillation

Assay

Table 1 compares the results obtained with the ultrafiltration assay to those obtained by conventional microdistillation. To ensure that only the means of separation were compared, aliquots from the same incubation mixture were assayed by each method. There is no significant difference between the results obtained by ultrafiltration or by microdistillation.

AND

HULMES

trations which may be performed simultaneously, in practice this does not limit the number of assays since, once the reaction has been stopped rapidly by cooling and the bulk of the substrate removed by centrifugation, no measurable change in activity is found. If necessary, the supernatant may be stored at 0 to 4°C for up to 24 h before ultrafiltration, without loss of accuracy. The use of a BAPN control is recommended as an indication of specific lysyl oxidase activity. ACKNOWLEDGMENTS We thank sity of Oulu

Drs. K. I. Kivirikko and R.-M. Mannermaa for advice in the initial stages of this work.

of the Univer-

REFERENCES 1. Burbelo,

P. D., Monckeberg,

A., and Chichester,

C. 0. (1986)

Col-

lagen Relat. Res. 6,153-162. 2. Han, S., and Tanzer, M. L. (1979) J. Biol. Chem. 254, 10,43810,442. 3. Hutton, J. J., Tappel, A. L., and Udenfriend, S. (1966) Anal. Biothem. 16,384-394. 4. Iguchi, H., and Sano, S. (1985) Connect. Tissue Res. 14,129-138. 5. Kagan, H. M., Sullivan, K. A., Olsson, T. A., III, and Crolund, A. L. (1979) Biochem. J. 177,203-214. 6. Kagan, H. M., Tseng, L., Trackman, P. C., Okamoto, K., Rapaka, R. S., and Urry, D. W. (1980) J. Biol. Chem. 255,3656-3659. 7. Kagan, H. M., and Sullivan, K. A. (1982) in Methods in Enzymology (Cunningham, L. W., and Frederiken, 650, Academic Press, San Diego, CA.

Eds.),

8. Kagan,

H. M., Williams, M. A., Williamson, J. M. (1984) J. Biol. Chem. 259,11,203-11,207.

Vol.

82, pp. 631-

P. R., and Anderson,

9. Kagan, H. M. (1986) in Regulation of Matrix Accumulation (Mecham, R. P., Ed.), pp. 321-398, Academic Press, Orlando, FL. H., Savolainen, E-R., and Kivirikko, K. I. (1984) J. 10. Kuivaniemi,

Biol. Chem. 259,6996-7002.

Alternative

Ultrafiltration

System

All results presented here were obtained with Amicon C-10 microconcentrators. More recently, however, we have tested Millipore Ultrafree-MC units as a more economical alternative. Under standard assay conditions, the Millipore units were found to be satisfactory, although their capacity is somewhat less than that of the Amicon units.

13. Misiorowski,

B&hem.

The main advantages of the ultrafiltration assay, compared with other assays (12-14), are the efficient use of substrate, ease of operation, and elimination of timeconsuming radioactive decontamination. Many advantages of the microassay (13) are retained, but the need for purpose built apparatus is eliminated. Although the availability of centrifuges limits the number of ultrafil-

R. L., Ulreich,

B. N. (1977)

J. B., and Chvapil,

M.

(1976)

Anal. Anal.

71,186-192.

14. Pinnell,

S. R., and Martin,

G. R. (1968)

F’roc.

Natl. Acad. Sci. USA

61,708-716. 15. Sandberg,

L. B., Soskel,

N. T., and Leslie,

J. G. (1981)

N.

Engl. J.

Med. 304,556-579. 16. Shieh,

DISCUSSION

Biochem. J. 239,

11. van der Meer, R. A., and Duine, J. A. (1986) 789-791. J., Vianden, G. D. N. E., and Bachra, 12. Melet, B&hem. 77,141-146.

J. J., and Yasunobu,

T. (1976)

Adu. Exp. Med. Bi01. 74,

447-463. R. C., Pinnell, S. R., and Martin, G. R. (1970) Biochemistry 9,4486-4492. 18. Siegel, R. C. (1974) Proc. Natl. Acad. Sci. USA 71,4826-4830. 19. Stassen, F. L. H. (1976) B&him. Biophys. Acta 438,49-60. P. C., and Kagan, H. M. (1983) J. Biol. 20. Tang, S-S., Trackman, Chem. 258,4331-4338. D. R., and Hulmes, D. J. S. (1990) B&hem. J., in 21. Shackleton, 17. Siegel,

press.