A microassay for heme oxygenase activity using thin-layer chromatography

ANALYTICAL

BIOCHEMISTRY

200,27-30

(1992)

A Microassay for Heme Oxygenase Activity Thin-Layer Chromatography’ Esteban

E. Sierra2

Department

Received

May

and Louise

M. Nutter3

of Pharmacology, University of Minnesota, Minneapolis, Minnesota 55455

9, 1991

A sensitive and facile assay for heme oxygenase (HO) has been developed. The basis of the assay is the detection of [ ‘“Clbilirubin formation in a coupled enzyme assay involving HO and biliverdin reductase actions, respectively. Separation of substrate from product is accomplished by thin-layer chromatography with subsequent quantitation by liquid scintillation counting of radioactive material present on chromatograms. As little as 20 pg of total cellular protein derived from cells growing in a standard 25-cm2 culture flask is sufficient for detection of HO enzyme activity using this assay. The reaction is inhibited by tin-protoporphyrin (10 pM final concentration), a specific inhibitor of HO. The linearity of the enzyme reaction with respect to incubation time and amount of protein used was established. Comparison of the new HO assay with a spectrophotometric assay was made, and good agreement of the results from both methods was found. The assay described here should facilitate measurements of this important hemedegrading enzyme in tissue culture studies and cases where limited amounts of material are available. o 1992 Academic

Press,

Using

Inc.

Heme oxygenase (H04; EC 1.14.99.3) is the rate-limiting enzyme in the oxidative catabolism of heme (reviewed in Ref. 1). It catalyzes the degradation of heme to biliverdin IX, carbon monoxide, and iron in a reaction requiring molecular oxygen, NADPH, and NADPH-cyi This work was supported by Grants INS-637 from the American Cancer Society and IR29CA52618-OlAl from the NCI, National Institutes of Health. * E.E.S. is the recipient of a postdoctoral fellowship from the Ford Foundation. 3 To whom correspondence should be addressed at Department of Pharmacology, 435 Delaware Street, S.E., University of Minnesota, Minneapolis, MN 55455. ’ Abbreviations used: HO, heme oxygenase; PBS, phosphate-buffered saline; PMSF, phenylmethylsulfonyl fluoride. 0003-2697192 $3.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

tochrome P450 reductase. In the presence of biliverdin reductase, biliverdin IX is rapidly converted to bilirubin IX. A variety of agents, including heme, heavy metals, and ultraviolet radiation have been shown to induce the expression of the HO gene. In our laboratory we have been studying HO induction by menadione, a redox-cycling napthoquinone which possesses a broad spectrum of anticancer activity in vitro (2). The standard assay for measuring HO activity consists of the spectrophotometric detection of bilirubin, produced in the presence of biliverdin reductase, NADPH, NADPH-cytochrome P-450 reductase, and a source of HO, usually microsomes (3,4). The production of bilirubin is detected either by taking a difference spectra between a control (in the absence of NADPH) and the complete experimental reaction or by extracting the bilirubin produced into chloroform. Bilirubin shows an absorption peak around 460 nm (extinction coefficient in chloroform: 53 mM-’ cm-‘) and may be quantitated using visible spectroscopy. Although this method is appropriate for measurement of HO activity in tissue samples, where relatively large amounts of HO are available, it is rather insensitive for the measurement of HO activity in tissue culture samples. For this reason we have developed a highly sensitive coupled assay that utilizes [14C]heme as a substrate for HO and separates the [“C]bilirubin product from the substrate utilizing thin-layer chromatography. This method is very sensitive, allowing for the detection of HO activity from cells grown in 25cm2 tissue-culture flasks. In this communication we describe and characterize this new HO assay. MATERIALS

AND

METHODS

Chemicals. Hemin (Fe-protoporphyrin IX), biliverdin, bilirubin, and NADPH were purchased from Sigma Chemical Co. (St. Louis, MO). Tin-protoporphyrin was obtained from Porphyrin Products (Logan, Utah). [14C]heme (specific activity: 54 Ci/mol and judged to be 27

Inc. reserved.

28

SIERRA

AND NUTTER

>95% pure by HPLC) was purchased from Leeds Radioporphyrins (Dr. Stanley Brown, Leeds, UK). The porphyrins were dissolved in a small volume of 0.05 M NaOH and the pH was adjusted to 8.4 with HCl. Solutions of the desired concentration were then diluted with distilled water. After preparation the samples were aliquoted in small volumes and stored at -70°C. Cell culture. Human MCF-7 (breast adenocarcinoma) cells were used as a source of enzymatic activity for these studies. The cell line was maintained in Roswell Park Memorial Institute (RPMI) 1640 media supplemented with 10% fetal calf serum (all cell culture reagents were from GIBCO, Grand Island, NY). Preparation of cellular extract. MCF-7 cells were grown in 25cm2 tissue-culture flasks to approximately 80% confluence and harvested. After washing twice with ice-cold phosphate-buffered saline (PBS) the cells were transferred to a 1.5-ml Eppendorf tube and resuspended in 50 ~1 of hypotonic cell lysing buffer containing the proteinase inhibitor phenylmethylsulfonyl fluoride (PMSF) (10 mM potassium phosphate buffer (pH 7.4), 50 PM PMSF), incubated in ice for 10 min, vortexed vigorously, and freeze-thawed three times. This treatment effectively disrupts the cellular membrane. The resulting suspension was centrifuged for 1 min at 14,000 rpm in a microcentrifuge. The supernatant contains the heme oxygenase, biliverdin reductase, and NADPH-cytochrome P450 reductase activities required for the enzyme assay and will be referred to as the 14K supernatant herein. Microsomes were prepared by differential centrifugation of the cellular extract. Briefly, the cellular homogenate, in 0.25 M sucrose, is sequentially centrifuged at 6OOg, 16,OOOg, and 105,OOOg to pellet nuclei and debris, mitochondria, and microsomes, respectively. The microsomes were resuspended in 0.1 M potassium phosphate buffer (pH 7.4). Biliverdin reductase was partially purified from rat liver through the 40-60% ammonium sulfate cut (step 3), as described by others (5). Heme oxygenase assay. The standard reaction mixture (10 ~1) consisted of 11.2 pM [14C]heme (specific activity 54 Ci/mol), 1 InM NADPH, 2 mM glucose-6-phosphate, 0.1 units of glucose-6-phosphate dehydrogenase, and 50-100 pg of total protein. The reaction was initiated by the addition of the heme, incubated at 37°C for the indicated period of time, and stopped by the addition of excess nonradioactive heme and bilirubin as carriers followed by placement in an ice-water bath. Two microliters of the reaction mixture were spotted twice onto a silica gel thin-layer chromatography sheet (Kodak). The chromatogram was developed using a chloroform:acetic acid (2O:l) solvent. The spots corresponding to heme and bilirubin were cut and placed in lo-ml scintillation fluid. The radioactivity associated with each

sample was quantitiated in a Beckman scintillation counter to determine the conversion of [14C]heme into [14C]bilirubin. The data was then expressed as picomoles of bilirubin formed/mg protein/h. For autoradiography, the TLC plates were sprayed evenly with EN3HANCE spray (New England Nuclear), placed directly against Kodak X-Omat X-ray film, and exposed at -70°C for 24-72 h. RESULTS Separation

AND DISCUSSION of [“C]Bilirubin

from [14C]Heme

The method described here offers a simple, sensitive assay for measuring heme oxygenase activity. The assay is based on the detection of [“Clbilirubin production from [14C]heme. In the presence of excess biliverdin reductase activity [‘4C]biliverdin is converted to [‘“Clbilirubin. Crude extracts from MCF-7 cells (14K supernatant) contain the HO activity and the excess biliverdin reductase activity used for the experiments in this report. We have employed thin-layer chromatography (TLC) utilizing silica gel plates and a chloroform:acetic acid (20~1) solvent system to separate [14C]heme from the [14C]bilirubin product of the coupled reaction of heme oxygenase and biliverdin reductase. The choice of a chloroform-acetic acid solvent system was selected after numerous experiments using other solvent systems and authentic heme and bilirubin standards. In the chloroform-acetic acid chromatographic system, heme and biliverdin remain close to the origin while bilirubin migrates with an &of 0.56 + 0.01 (Fig. 1). When [14C]heme is incubated with 120 pg of the 14K supernatant from MCF-7 cells at 37°C for 1 h, approximately 19% is converted to a radioactive product that comigrates with the same R, as a bilirubin standard (Fig. lA, lane 2), strongly suggesting that it is [14C]bilirubin. This reaction is strongly inhibited by tin-protoporphyrin, a competitive inhibitor of heme oxygenase ((6); Fig. lA, lane 3) and is not catalyzed by boiled cellular extract (Fig. lA, lane 4). Further addition of partially purified biliverdin reductase from rat liver did not result in increased heme oxygenase activity (i.e., 0.4 and 0.32 nmol of bilirubin produced/mg protein/h in control and biliverdin reductase supplemented reactions, respectively), thus demonstrating that the endogenous biliverdin reductase activity is sufficient to convert the [14C]biliverdin to [14C]bilirubin. Direct measurement of the biliverdin reductase activity in the 14K supernatant yields values of 35-45 nmol/mg protein/h, which is well in excess of the average HO activity in MCF-7 cells (0.1-0.5 nmol/mg protein/h). These results demonstrate that HO activity is rate-limiting in the 14K supernatant from MCF-7 cells. However, this should be evaluated for each experiment since some treatments may result in decreased biliverdin reductase activity.

MICROASSAY A

FOR

C

BR

Origin 1

2

3

4

1

2

FIG. 1. Autoradiograms of the TLC separation of bilirubin from heme using chloroform:acetic acid (2&l; A and B) and heptane: methyl ethyl ketone:acetic acid (l&5:1; C). (A) 120 pg of 14K supernatant from MCF-7 cells was incubated with the standard reaction mixture for 1 h and processed as described under Materials and Methods. Lane 1 (control), the reaction was stopped at 0 min; lane 2, the reaction proceeded for 1 h; lane 3, the reaction mixture included 10 pM tin-protoporphyrin, an inhibitor of HO activity; lane 4, the reaction mixture was incubated with 120 pg of boiled cellular extract. In this system, radioactive heme remains at the origin. (B) The standard reaction mixture was incubated for 1 h with 30 pg of purified microsomes from MCF-7 cells and 60 pg of partially purified biliverdin reductase from rat liver. Lane 1, no NADPH was added, lane 2, complete reaction mixture. (C) The standard reaction mixture was scaled-up to 100 ~1 and incubated with 1 mg of 14K supernatant for 30 min. The reaction was extracted with 200 rl of chloroform, the chloroform was evaporated, and the extracted material was spotted in a silica gel TLC sheet. The chromatogram was developed using a heptane:methyl ethyl ketone:acetic acid (10&l) solvent system. Since heme is not readily extracted into chloroform no radioactive heme is detected in (C).

For example, in reconstituted systems, 120 PM CoCl,, a heme oxygenase inducer, inhibits biliverdin reductase activity by 45% (8). Since heme oxygenase is a microsomal protein and it requires NADPH stoichiometrically for its function we explored whether purified microsomes would generate the radioactive product. For this experiment exogenous biliverdin reductase is required, since it is a cytosolic protein. In the presence of NADPH and biliverdin reductase (from rat liver cytosol), microsomes isolated from MCF-7 cells catalyze the formation of [‘“Clbilirubin (Fig. lB, lane 2), but not in the absence of NADPH (lane 1). In order to confirm that the identity of the radioactive product was indeed bilirubin, a second solvent system, employed by others to separate heme and bilirubin was used (7). The 14K supernatant was incubated with [14C]heme for 30 min and the bilirubin was extracted into chloroform; heme is not extracted into chloroform. The chloroform layer was evaporated and its constituents spotted on a silica gel plate. The chromatogram was developed using heptane:methyl ethyl ketone:acetic acid (1051). Under these conditions bilirubin migrates with a reported &of 0.67 (7). Figure 1C shows an autoradiograph of the chromatogram. A single radioactive spot comigrated with the bilirubin standard with an R, of 0.64. Based on the chromatographic behavior of the radioactive spot in two solvent systems, the NADPH

HEME

29

OXYGENASE

requirement, the tin-protoporphyrin inhibition of the reaction, and the lack of activity in boiled extracts, we conclude that the spot represents [14C]bilirubin. Clear demonstration of its identity would require isolation of the radioactive product and further analyses. The large amounts of material required to do this makes it impractical; in fact, this assay was developed to be able to detect small amounts of heme oxygenase activity. However, the comigration of the radioactive product with bilirubin standards under two different TLC systems and the properties of its formation strongly suggest its identity with bilirubin. Linearity

of the Assay

In order to determine the conditions under which the reaction rate is linear, the effects of increasing times of incubation and protein concentrations were examined. A time course of the reaction, using 90 pg of 14K supernatant, is shown in Fig. 2. The reaction rate is linear and maximal for the first 20 min of the reaction. By 30 min, the rate has decreased to 85% of the maximal rate, and after 60 min the reaction proceeds very slowly. The effect of protein concentration in the reaction was studied by adding increasing amounts of 14K supernatant to the reaction mixture and allowing it to react for 30 min. The reaction is linear with protein concentration up to approximately 10 pg protein/jd assay mix (Fig. 3). Comparison with Spectrophotometric Method The activity of heme oxygenase in MCF-7 cells, as measured by TLC separation of [i4C]bilirubin from [ 14C]heme, was compared with that measured using the

0

20

40

time

60

60

(min)

FIG. 2. Time course of the conversion of [“Clheme to [“Clbilirubin. The standard reaction mixtures, described under Materials and Methods, contained 90 pg of 14K supernatant from MCF-7 cells. At the indicated periods of time the reactions were stopped by the addition of excess nonradioactive heme and quenching in an icewater bath. The values shown are the averages of two experiments.

SIERRA

0

I 4

2 Protein

FIG. 3. conversion formed as heme was supernatant shown are

1 6 concentration

I 0

I 10

AND

I 12

(mglml)

Effect of protein concentration on the rate of [“Clheme to [“‘C]bihrubin. The heme oxygenase assay was perdescribed under Materials and Methods except that 25 PM used. Reactions containing the indicated amounts of 14K from MCF-7 cells were incubated for 20 min. The values the averages of two experiments.

spectrophotometric assay. For this purpose the reaction had to be scaled-up to 100 ~1 in order to be able to detect spectrophotometrically the bilirubin produced. After 30 min the reaction was extracted with chloroform and the bilirubin formed was calculated from the difference in absorption between 460 and 530 nm. The amount of bilirubin produced was 1.1 X 10-l’ mol/mg protein. Quantitation of the same reaction using the method described here yielded a value of 1.0 X 10-l’ mol/mg protein. The good agreement between both methods indicates that quantitation of [‘4C]bilirubin production using TLC yields reliable results using much less amounts of material. Currently, three methods have been described for measuring heme oxygenase activity. The most frequently used method measures bilirubin formation spectrophotometrically (3,4). The main drawback of this method is that it requires relatively large amounts of tissue to prepare microsomes. While this is reasonable for work with organ tissue, it demands large numbers of cells from cell culture, making it less feasible and costly. In addition, high absorptivity of the samples can produce artifactual results caused by light scattering and

NUTTER

chloroform extraction of the sample sometimes does not result in quantitative extraction of the bilirubin (7). A microassay for measuring heme oxygenase activity using an isotope dilution technique has been described by others (9). This method, although sensitive, involves chloroform extraction and crystallization of the bilirubin. A third method involves the detection of the carbon monoxide (CO) produced in the reaction using a gas chromatographer (10). This method is also very sensitive, but it requires access to equipment that is not standard in a biochemical laboratory. The sensitivity of the assay described in this report makes it particularly useful for the detection of heme oxygenase activity in cultured cells and when limited material is available. Using [14C]heme of high specific activity (54 Ci/mol) we can detect activities as low as 5 X lo-i4 mol of bilirubin produced/min, which is lo-fold higher than the sensitivity of the isotope dilution method (i.e., 0.5-l pmol bilirubin produced/min; Ref. 9). In addition, the use of small reaction volumes minimizes the amount of 14K supernatant required for the assay. In summary, the method described above represents a good alternative to the standard assay for heme oxygenase activity, particularly under conditions of limited protein availability and/or massive testing, such as large-scale screening of cells for the effect of different treatments on the activity of heme oxygenase. REFERENCES 1. Maines,

M. D. (1988)

FASEB J. 2, 2557-2568.

2. Nutter, 3.

L. M., Cheng, A.-L., Hung, H.-L., Hsieh, R.-K., Ngo, E. O., and Liu, T.-W. (1991) Biockm. Phurmucal. 41,1283-1292. Yoshida, T., Takahashi, S., and Kikuchi, G. (1974) J. Biochem. 75,1187-1191.

4. Ibraham, N. G., Lutton, J. D., and Levere, R. D. (1982) Br. J. Haematol. 50, 17-25. 5. Tenhunen, R., Ross, M. E., Marrer, H. S., and Schmid, R. (1970) Biochemistry 9, 298-303. 6. Kappas, A., Drummond, G. S., Simionatto, C., and Anderson, K. E. (1984) Heputobgy 4,336-341. 7. Kutty, R. K., and Maines, M. D. (1982) J. Biol. Chem. 257,99449952. a. Maines, M. D., and Kappas, A. (1975) J. Biol. Chem. 250,41714177.

9. Tenhunen, 10. Vreman,

R. (1972) Anal. Biochem. 45,600-607. H. J., and Stevenson, D. K. (1988) 168,31-38.