Eur. J . Biochem. 176, 39-45 (1988) :I FEBS 1988

Microsomal and cytosolic epoxide hydrolases, the peroxisomal fatty acid P-oxidation system and catalase Activities, distribution and induction in rat liver parenchymal and non-parenchymal cells

Pablo STEINBERG ', Ludwig SCHLADT', Hans Peter DIENES', Christopher TIMMS' and Franz OESCH' ' Institute of Toxicology, University of Mainz

Institute of Pathology, University of Mainz

(Received February 4/May 18, 1988) ~ EJB 88 0149

A number of structurally unrelated hypolipidaemic agents and certain phthalate-ester plasticizers induce hepatomegaly and proliferation of peroxisomes in rodent liver, but there is relatively limited data regarding the specific effects of these drugs on liver non-parenchymal cells. In the present study, liver parenchymal, Kupffer and endothelial cells from untreated and fenofibrate-fed rats were isolated and the activities of two enzymes associated with peroxisomes (catalase and the peroxisomal fatty acid P-oxidation system) as well as cytosolic and microsomal epoxide hydrolase were measured.

Microsomal epoxide hydrolase, cytosolic epoxide hydrolase and catalase activities were 7 - 12-fold higher in parenchymal cells than in Kupffer or endothelial cells from untreated rats; the peroxisomal fatty acid f3-oxidation activity was only detected in parenchymal cells. Fenofibrate increased catalase, cytosolic epoxide hydrolase and peroxisomal fatty acid P-oxidation activities in parenchymal cells by about 1 .5 , 3.5- and 20-fold, respectively. The induction of catalase (2 - 3-fold) and cytosolic epoxide hydrolase (3 - 5-fold) was also observed in Kupffer and endothelial cells ; furthermore, a low peroxisomal fatty acid P-oxidation activity was detected in endothelial cells. Morphological examination by electron microscopy showed that peroxisomes were confined to liver parenchymal cells in untreated animals, but could also be observed in endothelial cells after administration of fenofibrate.

Peroxisomes are ubiquitous cytoplasmic organelles pre- sent in plant and animal cells. They have both catabolic and anabolic functions [ 11 ; the former include respiration and the P-oxidation of fatty acids [2, 31. Respiration is catalyzed by oxidases which form H 2 0 2 and catalase which decomposes it. The peroxisomal P-oxidation reactions are the same as those of mitochondria1 f3-oxidatioq with the exception of the first reaction, in which electrons are transferred to oxygen by an oxidase in the peroxisomes, instead of to the electron transport chain, as in mitochondria.

A small quantity of peroxisomes is observed in mam- malian liver parenchymal cells [4]. After administration of several structurally unrelated hypolipidaemic drugs, or certain phthalate-ester plasticizers, a marked hepatomegaly and a large increase in both number and size of peroxisomes in liver parenchymal cells occurs [5]; furthermore, the activities of several enzymes in the liver (e.g. catalase, peroxisomal fatty acid P-oxidation system, epoxide hydrolase) are enhanced

The mammalian liver is composed of several different cell types. Parenchymal cells constitute about 90% of the total cell mass, but only represent 65% of the total cell number [8]; the remainder corresponds to non-parenchymal cells, mainly Kupffer and endothelial cells. There is limited data regarding the specific effects of peroxisome proliferators on liver non-

[5 - 71.

Correspondence to P. Steinberg, Institut fur Toxikologie der Universitat Mainz, Obere Zahlbacher StraDe 67, D-6500 Mainz, Fed- eral Republic of Germany

Enzymes. Catalase (EC 1.1 1 .I .6); epoxide hydrolase (EC 3.3.2.3); and acyl CoA oxidase (EC 1.3.3.6).

parenchymal cells. Therefore, we isolated parenchymal, Kupffer and endothelial cells from untreated and fenofibrate- fed rats and measured the activities of catalase, microsomal and cytosolic epoxide hydrolases, and the peroxisomal fatty acid P-oxidation system in all three cell types. In this study, the peroxisome proliferator used was fenofibrate, a relatively potent hypolipidaemic agent structurally related to clofibrate [9- 121.

MATERIALS AND METHODS

Chemical

Pronase E was purchased from Merck (Darmstadt, FRG); collagenase from Boehringer (Mannheim, FRG); and Nycodenz from Molter (Bammentdl, FRG). Fenofibrate was supplied by Laboratoires Fournier (Dijon, France). [G-3H]Benzo [alpyrene 4,5-oxide [13], ci~-[2-~H]stilbene oxide [14] and tran~[2-~H]stilbene oxide [14] were synthesized as previously described. All other chemicals used were of the highest purity available.

Animals and induction

Male Sprague-Dawley rats (170 - 190 g body mass) were purchased from Suddeutsche Versuchstierfarm (Tuttlingen, FRG), housed four to a cage, and allowed free access to both water and a standard rodent diet (Altromin) until used. Induction was performed by feeding rats for 10 days with 0.05% (by mass) fenofibrate in the food. This diet was pre- pared by dissolving the appropriate amount of fenofibrate in

40

a volume of acetone corresponding to 10% of the mass of the food and then thoroughly mixing this solution with the food pellets. The solvent was allowed to evaporate for 48 h in a ventilated hood before use. Control animals were given food treated in the same way with acetone alone, and had the same levels of all four enzyme activities measured in animals fed with untreated pellets.

Isolation of rat liver parenchymal, Kupffer and endotheiial cells

Parenchymal and non-parenchymal cells were isolated from different animals, in order to obtain higher yields of non-parenchymal cells [I 5,161. The two non-parenchymal cell types (Kupffer and endothelial cells) were isolated from the same livers.

Total liver cell suspensions were prepared using a collagenase perfusion method described by Glatt et al. [17] and parenchymal cells were isolated from the total liver cell suspension by differential centrifugation [ 181. Non- parenchymal cell suspensions were prepared as described by Knook and coworkers [15, 16, 191 by using pronase E, an enzyme which selectively destroys parenchymal cells [20]. Kupffer and endothelial cells were separated from the non- parenchymal cell suspensions by centrifugal elutriation [19]. Eight isolations of Kupffer and endothelial cells from untreated and fenofibrate-fed rats were performed. For the enzyme assays, the eight samples of Kupffer and endothelial cells from each group were split into four pairs (n = 4). All determinations were performed in cell homogenates obtained by sonicating the cells for 30 s at 60% duty cycle on a Branson cell disruptor (model B-15).

Viability, yield, purity and characterization of the isolated rat liver cells

Cell viability was estimated by the capacity of the cells to exclude 0.25% trypan blue. Cells were counted with a hemocytometer. Kupffer cells, endothelial cells and lympho- cytes present in the different cell fractions were distinguished from each other by light microscopy after they had been stained for peroxidase and non-specific esterase; in the rat, peroxidase is found exclusively in Kupffer cells [21] while, although esterase activity is present in all liver cells, lympho- cytes do not exhibit the latter activity [22]. Peroxidase activity was cytochemically determined by incubating the fresh isolat- ed cells with 3,3’-diaminobenzidine and HzOz for 30 min at 37°C [21]. For esterase staining, cells were first fixed in ice cold 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer (pH 7.4) and then incubated with 1-naphthylacetate for 10 min at 37°C [22]. Fat-storing cells were identified by fluo- rescence microscopy [23].

For electron microscopy, the cells were pelleted and fixed with 2.5% glutaraldehyde in 0.1 M sodium cacodylate buffer. After extensive washing in 0.01 M phosphate buffer, the cells were postfixed in 1 YO Os04 for 30 min and again subjected to several washings. Then dehydration in graded solutions of ethanol was performed, followed by embedding in Epon. Ultrathin sections were cut and examined, after staining with uranylacetate and lead citrate, under a Philips EM 301 elec- tron microscope.

Enzyme assays

In all assays the amount of product was linear with both time and protein concentration. The activities of cytosolic

Table 1. Effects of dietary fenofibrate on food intake, body m a n and liver mass Rats were fed with a control diet or a diet containing 0.05% (by mass) fenofibrate for 10 days. Each experimental group consisted of 12 animals. * Indicates significantly different from the control value (P < 0.05, Student’s t-test)

Diet Food intake/ Body mass at Liver mass rat during 10 days Odays 10days

g

Normal 229 5 11 2 0 6 i 7 2 7 3 5 12 13.9k0.9 Fenofibrate-

supplemented 234 14 209 f 10 274 f 14 19.5 1.3 *

epoxide hydrolase [24], microsomal epoxide hydrolase [13], catalase [25] and the peroxisomal palmitoyl-CoA oxidation system [26] were determined as previously described. Protein concentrations were estimated by the method of Lowry et al. [27] using bovine serum albumin as standard.

RESULTS

Effects of dietary,fenojibrate on food intake, body mass and liver mass

Food consumption (about 230 g/rat during 10 days) and body mass gain (about 65 g/lO days) were similar in untreated and fenofibrate-fed rats (Table 1). However, in the latter group a marked hepatomegaly was observed ; the relative liver mass expressed as g liver/100 g body mass increased substantially from 5.0 f 0.4% in untreated rats to 7.1 k 0.5% in fenofibrate-pretreated rats.

While peroxisomes were few in number in liver parenchymal cells isolated from untreated rats (Fig. l), a marked increase in the number of peroxisomes, and a signifi- cant decrease in the surface area occupied by the rough endo- plasmic reticulum, were seen in parenchymal cells isolated from fenofibrate-fed rats (Fig. 2). Furthermore, the electron microscopic analyses showed that peroxisomes, not visible in endothelial cells from untreated rats (Fig. 3), were present in liver endothelial cells from fenofibrate-fed rats (Fig. 4).

Characteristics of the isolated liver cell populations

The yield, viability, purity and protein content of liver parenchymal, Kupffer and endothelial cells isolated from untreated rats are shown in Table 2. The total yield of parenchymal cells obtained after collagenase digestion of the liver was (227 f 30) x lo6 cells per rat liver (n = 4), which included 2 f 1 YO endothelial cells, 2 k 1 YO Kupffer cells and 3 f 2% fat-storing cells. The endothelial cell fractions were contaminated with 9 f 2% lymphocytes and 4 k 2% Kupffer cells. The Kupffer cell preparations were found to contain 10 -t 3% endothelial cells, 3 f 2% fat-storing cells and 0.1 f 0.1% parenchymal cells.

Pretreatment of the animals with fenofibrate did not affect the yield, viability and purity of the isolated liver parenchymal, Kupffer and endothelial cell fractions. How- ever, fenofibrate increased the protein concentration in the parenchymal cell population by about go%, while that in Kupffer and endothelial cells remained unchanged.

41

Fig. I . Liver. ptrr.c,,r(,/??,r?/cr/ cell i.so/uted,/rorn un untrruted rut. Upper, x 6400; lower, abundant rough endoplasmic reticulum is obscrvcd ( x 29410)

Effi2ct.s (?f'(lietcrr?,fL.nofihrcite on thp uctivities of' the pcro.uisonzul.f~itty w i d /j-o.uidation system, c u t c i k r s c ~ , cytosolic opoxide hydroluw ( i t i d niicrosotmil cyw.uidcl hydrolcisc. in liver crl ls

Table 3 shows the activities of the peroxisomal fatty acid /{-oxidation system, catalase, cytosolic epoxide hydrolase and inicrosomal epoxide hydrolase in liver parenchymal, Kupffer and endothelial cells from untreated and fenofibrate-pre- treated rats.

In livers of untreated animals peroxisomal /j-oxidation activity was only detected in parenchymal cells. After administration of fenofibrate parenchymal cells showed a 20-fold increase of this enzyme activity, while endothelial cells displayed an activity similar to that observed in parenchymal cells of untreated rats.

Catalase activity was 8 - 12-fold higher in parenchymal cells than in Kupffer and endothelial cells derived from untreated animals. Pretreatment of the rats with fenofibrate led to an increase in catalase activity in parenchymal cells (1 .Sfold) as well as in Kupffer (2-fold) and endothelial cells

Cytosolic epoxide hydrolase activity, as determined using trans-stilbene oxide as substrate, was higher in parenchymal cells than in Kupffer and endothelial cells (7- and 11-fold differences). Administration of fenofibrate enhanced cytosolic epoxide hydrolase activity to a similar extent (3 - 5-fold) in all three cell types.

Microsomal epoxide hydrolase activity was measured with benzo[n]pyrene 4,5-oxide and cis-stilbene oxide a s substrates, and similar results were obtained in all cases with the two substrates. N o increase in microsomal epoxide hydrolase ac-

(3.5-fold).

42

Fig. 2. L i w r /~crr .c , /7c. / i~. / i ier l c,cll isolcitcdfrorn (I /2&o/;ihrcitc-/cd rut. Upper, x 5500; lower, note thc high number 01' pcroxisomcs (1') and the paucity of thc rough cndoplasmic reticulum ( x 16 150)

t ivi ty w;is observed in parenchymal, Kupffer or endothelial cells isolated from fenofibrate-fed rats when compared to untreated animals. The activity of microsomal epoxide hydro- lase actually decreased in liver parenchymal cells prepared fro in induced an i ma1 s.

DISCUSSION

Peroxisomes were sparse in liver parenchymal cells from untreated rats; morphometric studies [4] have previously shown that this organelle only accounts for about 2% of the parenchymal cell volume and that i t is absent from Kupffer and cndotheliul cells in rat liver. The observation that in untreated uiimals peroxisomal fatty acid [)-oxidation activity could only be detected in parenchymal cells is consistent with the above data. Furthermore, we were able to measure catalase activity in parenchymal as well as Kupffer and endo- thelial cells isolated from untreated livers, thus confirming previous reports [28, 291 which suggested the presence of this enzyme in rat liver non-parenchymal cells.

In the rat liver, at least three different forms of epoxide hydrolases exist, which can be distinguished by their substrate specificities. The most investigated form of microsomal epoxide hydrolase metabolises the epoxides of polycyclic aro- matic hydrocarbons, steroids and other spatially bulky mol- ecules, with a preference for cis-epoxides (e.g. benzo[rr]pyrcnc 4 , h x i d e and cis-stilbene oxide) [30, 311. In contrast, the cytosolic enzyme tnetabolises spatially smaller molecules and exhibits a preference for fmns-epoxides (e.g. trcms-stilbene oxide) [32]. In addition to the main microsomal form, a further activity has been described, for which the only substrates discovered until now are the 2- and [i-epoxides of endogenous d '-steroids, e.g. cholesterol 5 2 , 62-oxide [33, 341. Further- more, i t has recently been observed that rat liver peroxisomes possess an epoxide hydrolase with the same substrate speci- ficity as the cytosolic epoxide hydrolase (C. W. Timms, un- published observations). In the present study, the activities of the main form of microsomal epoxide hydrolase and cytosolic epoxide hydrolase have been shown to be present not only in parenchymal cells but also in Kupffer and endothelial cells. In contrast, it has not been possible to detect microsomal

43

Fig. 3. Endot/io/iu/ wl1.s i.soltitrd/rorn un untreated rat. Upper, x 5500; lower, part of the cytoplasma a t a higher magnilication ( x 8500)

Cclltypc Number 10 '' Viability Purity Protein 01'riits x Yield content

of cellsirat liver

'Yo pg 10' cell\

P,ireiichyiii,il 4 2 2 7 k 3 0 8 7 k 8 9 3 k 3 1651 k 1 0 9 Kupffer 8 4 2 + 7 9 2 i 3 8 7 + 4 1 1 7 + 21 Endothelial 8 X6+11 9 3 * 4 8 7 + 5 7 4 + 1 3

epoxide hydrolase i n rat liver Kupffer and endothelial cells irninunohistocheinicailly [35, 361. A possible explanation for this discrepancy is that only a fraction (about 10%) of the total number of the two non-parenchymal cell types present in the rat liver possess this enzyme (P. Steinberg, unpublished observations).

As expected, fenofibrate caused hepatomegaly and a marked proliferation of peroxisomes in liver parenchymal cells. The liver enlargement is the result of an initial mitotic burst, followed by hypertrophy [37]. Peroxisome proliferation in liver parenchymal cells was associated with increases in the activities of the peroxisomal fatty acid /]-oxidation system (20-fold), cytosolic epoxide hydrolase (3.5-fold) and catalase (1.5-fold) in these cells. On account of the only moderate induction of catalase, as compared to peroxisomal /Loxi- dation, an imbalance between H 2 0 2 production and degra- dation could lead to H202-mediated oxidative damage, and eventually to carcinogenesis, in livers of animals treated with peroxisome proliferators [5, 381. The cytosolic epoxide hydro- lase, which is able to catalyze the hydrolysis of several epoxides derived from unsaturated fatty acids such a s oleic acid [39] or arachidonic acid [40], could be simultaneously induced to metabolize potentially toxic lipid epoxides formed during lipid peroxidation [41 - 431. Surprisingly, after administration of fenofibrate, the activity of microsomal epoxide hydrolase, which is expressed per mass of cellular protein, decreased in liver parenchymal cells. This may be due to the massive induction of other proteins (e.g. peroxisomal) and should not be regarded as an inhibitory effect of the

44

Fig. 4. E t h r l i d i t d cd/ , f i -om u /iwofihrure-fi.d ruf. A few peroxisomes (P) are observed ( x 17000)

Table 3 . Tlic, trc.t ivific, .s o f t h e pcro.vi.somul, fuff.y ucid /?-oxidution sys fem, cutuluse, und of'the microsomul und c,yosolic c~po.~ir/c lij~/roltr.sc~.s in l i w r ~ ~ ~ i r i ~ t i ~ ~ / i ~ ~ t ~ ~ t r l , Kirp,ffi,r trnd i~ti~loflii~liul cc1I.s isolufed from ruts fi.d wifh u normal or u.fenof~hruti~-.suppIementrd diet All ewynie assays were carried out using broken cell preparations. Each value is expressed as mean f SD of four rats per group i n the c;isc o f parcnchynial cells, while eight samples of Kupffcr and endothelial cells from each group were split into four pairs ( t i = 4). Pcroxisonial /I-oxidation was measured by reduction of NAD'. Catalase activity is presented as its first-order rate constant. Cytosolic and inicrosoni;iI cpoxide hydrolase activities were determined as products formed from (a) truns-stilbcnc oxide, (b) benzo[u]pyrenc. and (c) c.i.r-stilbenc oxide. * Indicatcs values which are significantly different from controls ( P < 0.05, Student's t-test)

Cell type Fcnofibrate Specific enzyme activity

pcroxisomal ca talase [I-oxida t i on

cytosolic microsomal m icroso ma I cpoxide cpoxide cpoxidc hydrolase hydrolase hydrolase (a) (b) ( c )

nmol min mg-'

Parenchymal ~ 6 f 1 + 124f21"

+ < 0.5 1:ndothelial - < 0.5

+ 6,2*

~ Kupffer < 0.5

0.95 f 0.10 22.5 f 5 4.51 f 1.21 4.99 f 1.02 1.40 & O.l5* 77 + 1 8 * 2.61 f 0.59 3.44 * 0.58 0.12 * 0.03 3 * 2 0.54 * 0.05 0.44 * 0.04 0.23 f 0.05* 10 f 3* 0.51 f 0.08 0.39 f 0.12

0.60 f 0.21 0.08 f 0.03 2 f 1 0.28 & 0.10* 10 & 5* 0.60 * 0.16 0.57 * 0.30

0.70 f 0.22

45

hypolipidaemic agent. Indeed, it has been previously shown that microsomal epoxide hydrolase activity remains un- changed upon in vitro addition of fenofibrate [44]. In contrast to this, pretreatment of mice with clofibrate [6] or nafenopin [7] resulted in an approximately 2-fold induction of this en- zyme.

Fenofibrate increased the activities of catalase and cytosolic epoxide hydrolase in Kupffer and endothelial cells to the same extent as in parenchymal cells. The finding that peroxisomal fatty acid B-oxidation activity could be measured in liver endothelial cells, isolated from fenofibrate-pretreated rats, is explained by the detection of peroxisomes (although in a low number) in these cells after administration of the hypolipidaemic agent.

This study was supported by the Deutsche Forschungsgemein- schaft (SFB 302) and the Alexander von Humboldt Foundation (P.S.). L.S. is the recipient of a post-doctoral scholarship from the Deutsche Forschungsgemeinschaft. We thank Ms. A. Raczek for her excellent technical assistance and Ms. I. Bohm for typing the manu- script.

REFERENCES 1. 2. 3. 4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

14.

15.

16.

17.

Tolbert, N. E. (1981) Annu. Rev. Biochem. 50, 133-157. de Duve, C. & Baudhuin, P. (1966) Physiol. Rev. 46, 323-357. Lazarow, P. B. (1978) J . Biol. Chem. 253, 1522- 1528. Blouin, A., Bolender, R. P. & Weibel, E. R. (1977) J . Cell Biol.

Reddy, J. K. & Lalwani, N. D. (1983) CRC Crit. Rev. Toxicol.

Hammock, B. D. & Ota, K. (1983) Toxicol. Appl. Pharmacol. 71,

Waechter, F., Bieri, F., Stiiubli, W. & Bentley, P. (1984) Biochem.

Weibel, E. R., Staubli, W., Gnagi, R. H. & Hess, F. A. (1969) J .

Hess, R., Stlubli, W. & Reiss, W. (1965) Nature (Lond.) 208,

Cohen, A. J. & Grasso, P. (1981) Food Cosmet. Toxicol. 19,585- 605.

Bliimcke, S., Schwartzkopff, W., Lobeck, H., Edmondson, N. A., Prentice, D. E. & Blane, G. F. (1983) Atherosclerosis 46,

Gariot, P., Barrat, E., Mejean, L., Pointel, J. P., Drouin, P. &

Schmassmann, H. U., Glatt, H. R. & Oesch, F. (1976) Anal.

Oesch, F., Sparrow, A. J. & Platt, K. L. (1980) J . Labelled Compd

de Leeuw, A. M., Barelds, R. J., de Zanger, R. & Knook, D. L.

Praaning-van Dalen, D. P. & Knook, D. L. (1982) FEBS Lett.

Glatt, H. R., Billings, R., Platt, K. L. & Oesch, F. (1981) Cancer

72, 441 -455.

12, 1 - 58.

254-265.

Phurmacol. 33, 31 - 34.

Cell Biol. 42, 68 -91.

856- 858.

105 - 116.

Debry, G. (1983) Arch. Toxicol. 53, 151 -163.

Biochem. 74, 94 - 104.

Radiopharm. I7 ,93 - 100.

(1982) Cell Tissue Res. 223, 201 -215.

141,229-232.

Res. 41, 270 - 277.

18. Tolleshaug, H., Berg, T., Nilsson, M. & Norum, K. R. (1977) Biochim. Biophys. Acta 499, 73 -84.

19. Steinberg, P., Lafranconi, W. M., Wolf, C. R., Waxman, D. J., Oesch, F. & Friedberg, T. (1987) Mol. Pharmacol. 32, 463- 470.

20. Berg, T. & Boman, D. (1973) Biochim. Biophys. Acta 321, 585- 596.

21. Wisse, E. (1974) J. Ultrastruct. Res. 46, 393-426. 22. Emeis, J. J. & PlanquC, B. (1976) J . Reticuloendothel. Soc. 20,

23. Blomhoff, R., Hoke, K., Naess, L. & Berg, T. (1984) Exp. Cell Res. 150, 185-193.

24. Schladt, L., Worner, W., Setiabudi, F. & Oesch, F. (1986) Bio- chem. Pharmacol. 35,3309-3316.

25. Aebi, H. (1970) in Methoden der enzymatischen Analyse. (Bergmeyer, H. U., ed.) vol. 1, pp. 636-647, Verlag Chemie, Weinheim.

26. Bieri, F., Bentley, P., Waechter, F. & Staubli, W. (1984) Carcinogenesis 5, 1033- 1039.

27. Lowry, 0. H., Rosebrough, N. J., Farr, A. L. & Randall, R. J . (1951) J . Biol. Chem. 193, 265-275.

28. van Berkel, T. J. C. (1974) Biochem. Biophys. Res. Commun. 61,

29. Fahimi, H. D., Gray, B. A. & Herzog, V. K. (1976) Lab. Invest.

30. Oesch, F. & Golan, M. (1980) Cancer Lett. 9, 169- 175. 31. Hammock, B. D. & Hasagawa, L. S. (1983) Biochem. Pharmacol.

32. Gill, S. S., Ota, K. & Hammock, B. D. (1983) Anal. Biochem.

33. Levin, W., Michaud, D. P., Thomas, P. E. & Jerina, D. M. (1983) Arch. Biochem. Biophys. 220,485 -494.

34. Oesch, F., Timms, C. W., Walker, C. H., Guenthner, T. M., Sparrow, A,, Watabe, T. & Wolf, C. R. (1984) Carcinogenesis 5,7-9.

35. Bentley, P., Waechter, F., Oesch, F. & Staubli, W. (1979) Biochem. Biophys. Res. Commun. 91, 1101 - 1108.

36. Kawabata, T. T., Guengerich, F. P. & Baron, J. (1982) Mol. Pharmacol. 20,709 - 714.

37. Price, S. C., Hinton, R. H., Mitchell, F. E., Hall, D. E., Grasso, P., Blane, G. F. & Bridges, J. W. (1986) Toxicology 41, 169- 191.

38. Tomaszewski, K. E., Agarwal, D. K. & Melnick, R. L. (1986) Carcinogenesis 7, 1871 - 1876.

39. Gill, S. S. & Hammock, B. D. (1979) Biochem. Biophys. Res. Commun. 89,965 -971.

40. Chacos, N., Capdevilla, J., Falck, J. R., Manna, S., Martin Wixtrom, C., Gill, S. S., Hammock, B. D. & Estabrook, R. W. (1983) Arch. Biochem. Biophys. 223, 639-648.

11-19.

204 - 209.

34, 192-201.

32, 1155-1164.

131,273-282.

41. Kadis, B. (1 978) J . Steroid Biochem. 9, 75 - 81. 42. Sevanian, A,, Mead, J. F. & Stein, R. A. (1979) Lipid.7 14, 634-

43. Sevanian, A., Stein, R. A. & Mead, J. F. (1980) Biochim. Biophys.

44. Schladt, L., Hartmann, R., Timms, C., Strolin-Benedetti, M., Dostert, P., Worner, W. & Oesch, F. (1987) Biochem. Phurmacol. 36,345 - 351.

643.

Acta 614,489-500.