- Email: [email protected]
Control of cytochromes P450 expression in Gunn rat liver: Implication of the intracellular heme pool
ARCHIVES
OF BIOCHEMISTRY
Vol. 290, No. 2, November
AND
BIOPHYSICS
1, pp. 407-410,199l
Control of Cytochromes P450 Expression in Gunn Rat Liver: Implication of the Intracellular Heme Pool Chantal
Celier’
INSERM
U75, 156 rue de Vaugirard,
and Thierry
Cresteil F-75730 Paris Cedex 15 France
Received April 10, 1991, and in revised form July 10, 1991
The absence of changes in the overall hepatic cytochrome P450 content after administration of 3-methylcholanthrene (MC) to congenitally jaundiced Gunn rats is believed to be related to a limited heme availability in this strain of rat. The amount of available heme, estimated by tryptophan pyrrolase activity, shows a substantial decrease in control Gunn versus control Wistar rats. This reduction is moderately enhanced by MC treatment in Gunn rats but is abolished after phenobarbital administration. Heme oxygenase activity is diminished in Gunn rats and consequently is not responsible for the decrease in the hepatic heme availability. These data point out that the depletion of the intracellular heme can lead to a limitation in the synthesis of cytochrome P450 isoenzymes in the MC-induced Gunn rat. o 1991 Academic
Press,
Inc.
A family of hemoproteins collectively called cytochrome P450 (EC 1.14.14.1) is the major user of the heme synthesized in hepatocytes. Some isoenzyme(s) of cytochrome P450 are constitutively expressed in the liver, while most of them are expressed at an extremely low level under normal conditions but could be massively induced by the administration of a variety of xenobiotics, with phenobarbital (PB)’ and 3-methylcholanthrene (MC) as major prototypes. We have recently reported that a congenitally icteric strain of rats (Gunn rats) exhibited a response to MC administration different from that of Wistar rats (1): in male Gunn rat the overall cytochrome P450 content of hepatic microsomes is unchanged after administration of MC and this apparent lack of induction is attributed to a balance between the rise in P450 IAl/IAB (MC-inducible isoenzymes) and the decrease in the male constitutive P450 IICll, as shown by immunoquantitation. In i To whom correspondence should be addressed. ’ Abbreviations used: PB, phenobarbital; MC, 3-methylcholanthrene; TPO, tryptophan pyrrolase. 0003.9861/91 $3.00 Copyright 0 1991 by Academic Press, All rights of reproduction in any form
contrast, the response to PB administration was similar in Gunn and Wistar rats (i.e., increase in total P450 and in specific isoenzymes IIBl/IIBS). We have postulated a possible regulatory effect of the level of intracellular heme on the synthesis of cytochrome P450 isoenzymes. To assess this hypothesis, the present study was undertaken to compare the influence of inducer administration on heme availability in livers from Wistar and Gunn rats. According to Badawy and co-workers (2, 3), tryptophan pyrrolase (TPO) is believed to be in dynamic equilibrium with the free heme pool in such a way that minor fluctuations in this pool are reflected by modifications in enzymatic activity of the hemoprotein. We have investigated the effects of variations in heme availability (subsequent to xenobiotic treatment or in viva injection of heme) on TPO activity in Wistar and Gunn rats. Since heme availability seems to be reduced in Gunn rat liver, we have further examined its catabolic pathway into biliverdin through the microsomal heme oxygenase. MATERIALS
AND
METHODS
Two-month-old homozygous male Gunn Pretreatment of animals. rats (CSEAL-CNRS, France, inbred in our animal unit), heterozygous male Wistar/Gunn rats (male Gunn X female Wistar, inbred in our animal unit) and male Wistar rats (IFFA-CREDO, France) were given ip MC (20 mg/kg), PB (80 mg/kg), or dexamethasone (200 mg/kg). Animals were killed 18 h after a single injection and livers were excised, perfused, and immediately frozen in liquid nitrogen for TPO activity determinations. For the measurement of cytochrome P450 and heme oxygenase, animals were treated daily for 3 days. Hematin hydrochloride (70 mg/kg ip) was dissolved in dimethylformamide and animals were killed 3 h after injection. All the experiments were performed with fed animals. Enzyme assays. The activity of tryptophan pyrrolase (tryptophan 2,3-dioxygenase EC 1.13.11.11) was determined in liver homogenates either in the absence (holoenzyme activity) or in the presence (total enzyme activity) of in uitro-added (2 FM) hem&in according to the procedure of Badawy and Evans (2). Homogenates were incubated up to 75 min and tryptophan pyrrolase activity was calculated from the increase in kynurenine absorbance at 365 nm during the linear phase. Microsomal fractions were prepared as previously described (1). Cytochrome P450 content was determined from the CO difference spectrum 407
Inc. reserved.
408
CELIER
Pretreatment WlS,cIl
holo
ws
w*rtar
total
AND
01 rats 0
Gunn
holo
u
Gunn
total
FIG. 1. Tryptophan pyrrolase activities in Wistar and Gunn male rat liver homogenates. W, Wistar; G, Gunn; UT, untreated; MC, 3-methylcholanthrene-treated; PB, phenobarbital-treated; holo, holoenzyme activity determined without in vitro addition of hematin; total, total enzyme activity measured after in vitro addition of 2 pM hematin. *P < 0.001 compared to UT W (comparison between holoenzyme activities and total enzyme activities, respectively). OP < 0.001 compared to UT G (comparison between holoenzyme activities and total enzyme activities, respectively).
of dithionite-treated microsomes as described by Omura and Sato (4) and protein concentrations were measured according to Lowry et al. (5). Heme oxygenase (EC 1.14.88.3) was assayed by the method of Tenhunen et al. (6) by determining the formation of bilirubin by microsomes in the presence of an excess of exogenous biliverdin reductase.
RESULTS
TPO is classically assayed in crude homogenates to reflect the amount of heme available in the tissue (holoenzyme) but also after in vitro addition of exogenous hemin to fully saturate the apoprotein and to determine its total activity. In preliminary experiments, we have assayed TPO activities in homogenates prepared from freshly excised liver and in homogenates prepared from frozen samples of the same liver. No difference was noticed between both types of preparation (1.11 + 0.14 in frozen versus 1.25 f 0.20 pmol . g liver-‘. h-l in fresh samples). Furthermore, the effect of an in vitro addition of hemin was strictly similar in both types of preparations (1.44 + 0.12 in frozen versus 1.47 f 0.21 prnol. g liver-‘. h-’ in fresh livers). These results validate our experimental procedure. As shown in Fig. 1, TPO activity was significantly lower (by ~50%) in control Gunn rats compared to Wistar rats. The decrease is similar in both holoenzyme and total enzyme activities and consequently the percentage of saturation remains unchanged (77% in Wistar versus 76% in Gunn rats). The addition of a higher concentration of hemin (4 versus 2 PM) in the incubation mixture does not increase total TPO activity in Gunn rats and indicates that the existing protein is fully saturated under our incubation conditions and represents the total amount of TPO (data not shown). To exclude a possible interference
CRESTEIL
of bilirubin in the TPO activity determination, we have checked that the in vitro addition of exogenous bilirubin to Wistar liver homogenates (then mimicking the Gunn situation), does not modify TPO activity determinations (data not shown). Administration of MC does not significantly affect TPO activity in Wistar rats, whereas a moderate decrease is observed in Gunn rats. PB treatment elicits a significant rise in TPO activity in Gunn rats, leading to a level of activity similar in both strains. Finally, corticosteroids have been reported to increase the synthesis of apotryptophan pyrrolase. In our experiments we have used dexamethasone for treatment of rats and Table I shows that TPO activity is significantly increased in both strains of rats. The percentage of induction is higher in Gunn rats although the activity remains less intense than in Wistar rats. In viva hematin injection at a dose of 35 mg/kg is known to result in a surge in TPO activity (2). In our experiments we have used a dose of 70 mg/kg to obtain clear-cut results with Gunn rats, given that 35 mg/kg produced inconstant effects in this strain. The enzyme is fully saturated in uiuo with endogenous and injected heme and the level of activity is increased in both homozygous animals and heterozygous Wistar/Gunn (W/G) animals (Table II). This clearly demonstrates that the enzyme remains inducible in the Gunn strain by a large excess of heme. To assess the possibility that an accelerated catabolism of heme could be the primary cause in the reduction of the heme pool in Gunn rats, the microsomal heme oxygenase activity was examined. Results (Table III) show that in untreated rats, activity is less intense in Gunn than in Wistar rats (-50%). MC or PB treatment is as-
TABLE
I
Effect of Dexamethasone Pretreatment on Tryptophan Pyrrolase Activities in Wistar and Gunn Rat Livers
In vitro heme addition Wistar Control DEX-treated Gunn Control
+ +
Tryptophan pyrrolase (Fmol * h-i. g-i liver)
Percentage of saturation
1.11 1.44 3.12 4.08
f f f f
0.14 0.12 0.35* 0.32*
77
0.49 0.64 2.68 2.98
f f f f
0.23* 0.25* 0.12** 0.12**
76
76
~ +
DEX-treated +
90
Note. Results are shown as means f SEM for three animals in each group. +, in vitro added hemin (2 pM). * P < 0.001 when compared to control Wistar rat. ** P < 0.001 when compared to control Gunn rat.
REGULATION TABLE
OF P450 EXPRESSION
II
Tryptophan Pyrrolase Activities in Wistar, Gunn, and Wistar/Gunn Rat Livers with or without in. Viuo Injection of Hematin In uiuo heme injection Wistar
+
Gunn + Wistar/Gunn +
Tryptophan pyrrolase (pm01 . h-i * g-i liver) 1.11 4.48 0.49 3.93 0.96 4.40
f 0.14 f 0.22 + 0.23* k 0.87 IL 0.42 f 0.48
Percentage of saturation 71 100 76 100 70 100
Note. Results are shown as mean values + SEM for three animals in each group. +, animals injected i.p. with hematin hydrochloride (70 mg/ kg) and killed 3 h after injection. * P < 0.001 when compared to control Wistar rat without heme injection.
sociated with a significant reduction of heme oxygenase activity in Wistar but not in Gunn rats. To confirm the involvement of hepatic unconjugated bilirubin in the reduction of the heme pool and to exclude another genetic defect, we have used heterozygous Wistar/ Gunn rats of first generation which present a normal bilirubinaemia. After MC administration, their hepatic content of cytochrome P450 is induced by the same extent as in Wistar rats (1.80 k 0.05 in MC-treated versus 1.10 & 0.02 nmol . mg -’ protein in control W/G rats). TPO activity in W/G rats is closer to that of homozygous Wistar than to homozygous Gunn (0.96 + 0.42 for holoenzyme and 2.14 k 0.37 pmol . g liver ‘. hem1for total enzyme activity) and definitively does not exhibit intermediate values. Finally, as shown in Table III, heme oxygenase activity of W/G rats is identical to that of Wistar rats. All together, these data suggest that in homozygous Gunn rats, the accumulation of bilirubin is associated with a lower heme availability and leads to a limitation in the synthesis of cytochrome P450.
409
BY HEME
tained when hemolysis occurs after thawing of liver samples containing residual traces of blood and (ii) there is no difference in the saturation index between perfused and nonperfused livers (see Results section). The examination of results shows that TPO activity is selectively reduced in homozygous Gunn compared to Wistar and heterozygous W/G rats in terms of activity, while the saturation of enzyme is unchanged. This could be interpreted as a reduced availability of heme and we have designed several experiments to confirm this hypothesis: (i) This limitation can be reversed by an in uiuo injection of hematin. A low dose of hematin is suitable to fully saturate the existing apoprotein (8) while a higher but nontoxic dose produces a surge in total activity, switching on the synthesis of the apoprotein (2). This could explain that the synthesis of apoprotein is adjusted to long-term variations of heme flow, like in the Gunn liver situation where the low heme availability down-regulates the apotryptophan pyrrolase synthesis. (ii) This assumption supposes that the TPO apoprotein synthesis could be stimulated in Gunn as in Wistar rat liver by other modulators. Cortisol and other steroids induce TPO apoprotein (2, 9, 10). Similar conclusions can be drawn for dexamethasone: the protein synthesis is stimulated in both strains and is likely associated with an enhancement of heme synthesis to maintain the saturation percentage at similar levels in Wistar and Gunn rats. All together, these data definitively validate the assumption that a lower amount of heme is available in
TABLE
III
Heme Oxygenase Activities in Wistar, Gunn, and Wistar/Gunn Rat Liver Microsomes Heme oxygenase (nmol f mini’ . rng-’ protein) Wistar
DISCUSSION
The validity of this approach is based on the assumption that tryptophan pyrrolase reflects the fluctuations of the heme availability (2) although some technical artifacts can make the interpretation questionable. In a first attempt, we delineated the accuracy of TPO determinations (see above). Assay conditions were acceptable even if some parameters differ from those in the literature: for example, in our hands, the percentage of heme saturation is higher than those previously reported (2). A possible explanation could be that heme released from contaminating hemoglobin or methemoglobin is incorporated into apotryptophan pyrrolase (7). However, this explanation is unlikely because (i) the concentration of hemoglobin injected in these experiments is considerably higher than that ob-
Control PB-treated MC-treated
0.133 t 0.023 0.056 t 0.017* 0.051 f 0.002* Gunn
Control PB-treated MC-treated
0.066 f 0.015” 0.046 f 0.011 0.066 f 0.002 Wistar/Gunn
Control
0.128 f 0.003
Note. Results are shown as mean values +- SEM for three animals in each group. * P < 0.001 compared to control. ** P < 0.001 compared to Wistar rats.
410
CELIER
AND
Gunn rat liver. This fluctuation of heme availability cannot be attributed to an accelerated degradation of heme since heme oxygenase activity is reduced in homozygous Gunn liver, but might rather result from a decrease in heme synthesis. In heterozygous W/G rats, there is no accumulation of hepatic bilirubin, and heme availability and heme oxygenase activity are comparable to that of Wistar rats as is the induction of cytochrome P450 by MC. Like for TPO, the synthesis of active cytochrome P450 requires the coordination of two multienzymatic pathways: the apoprotein synthesis on one hand and the heme biosynthesis on the other hand. In untreated Gunn rats, the level of the hepatic heme pool, although low, is sufficient to maintain a P450 concentration comparable to that of Wistar rats (1). The response to induction depends on the molecule administered: PB produces an important rise in heme synthesis (11) which results in an increase in available heme leading subsequently to an enhancement in the TPO activity and to the insertion of heme into the newly synthesized molecules of cytochrome P450 IIBl and -1IB2. MC produces a comparable requirement for heme but without activation of its synthesis (11): this results in a diminution of the free heme and a decrease in TPO activity. In Gunn rats, heme availability is reduced by MC and becomes insufficient to allow the insertion of heme in all the molecules of apocytochrome P450 synthesized. Several authors have demonstrated that apocytochrome P450 could be rapidly degraded in the liver when deprived of heme (12, 13). This explains why MC failed to increase the overall P450 content, as well as the
CRESTEIL
immunochemical level of cytochrome treated homozygous Gunn rats. The pairment of heme biosynthesis remains cidated but might involve the inhibition ogen I synthase by bilirubin (14).
P450 proteins, in origin of the imto be further eluof uroporphyrin-
REFERENCES 1. Celier, C., and Cresteil, T. (1989) Biochem. Pharmacol. 38, 28252832. 2. Badawy, A. A. B., and Evans, M. (1975) Biochem. J. 150,511-520. 3. Badawy, A. A. B., Morgan, C. J., and Davis, N. R. (1986) in Porphyrins and Porphyrias (Nordmann, Y., Ed.), Colloque INSERM, Vol. 134, pp. 69-81, John Libbey Eurotext, Montrouge. 4. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239, 2370-2378. 5. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 6. Tenhunen, R., Marver, H. S., and Schmid, R. (1969) J. Biol. Chem. 244, 6388-6394. 7. Wyman, J. F., Gollan, J. L., Settle, W., Farrel, G. C., and Correia, M. A. (1986) Biochem. J. 238, 837-846. 8. Welch, A. N., and Badawy, A. A. B. (1980) Biochem. J. 192, 403410. 9. Feigelson, P., and Greengard, 0. (1962) J. Biol. Chem. 237,37143747. 10. Schimke, R. T., Sweeney, E. W., and Berlin, C. M. (1965) J. Biol. Chem. 240, 322-331. 11. Baron, J., and Tephly, T. R. (1970) Arch. Biochem. Biophys. 139, 410-420. 12. Dwarki, V. J., Francis, V. N. K., Bhat, G. J., and Padmanaban, G. (1987) J. Biol. Chem. 262, 16,95&16,962. 13. Bonkovsky, H. L., Sinclair, J. F., Healey, J. F., Sinclair, P. R., and Smith, E. L. (1984) Biochem. J. 222, 453-462. 14. Kohashi, M., Tse, J., and Piper, W. N. (1984) Life Sci. 34, 193196.
OF BIOCHEMISTRY
Vol. 290, No. 2, November
AND
BIOPHYSICS
1, pp. 407-410,199l
Control of Cytochromes P450 Expression in Gunn Rat Liver: Implication of the Intracellular Heme Pool Chantal
Celier’
INSERM
U75, 156 rue de Vaugirard,
and Thierry
Cresteil F-75730 Paris Cedex 15 France
Received April 10, 1991, and in revised form July 10, 1991
The absence of changes in the overall hepatic cytochrome P450 content after administration of 3-methylcholanthrene (MC) to congenitally jaundiced Gunn rats is believed to be related to a limited heme availability in this strain of rat. The amount of available heme, estimated by tryptophan pyrrolase activity, shows a substantial decrease in control Gunn versus control Wistar rats. This reduction is moderately enhanced by MC treatment in Gunn rats but is abolished after phenobarbital administration. Heme oxygenase activity is diminished in Gunn rats and consequently is not responsible for the decrease in the hepatic heme availability. These data point out that the depletion of the intracellular heme can lead to a limitation in the synthesis of cytochrome P450 isoenzymes in the MC-induced Gunn rat. o 1991 Academic
Press,
Inc.
A family of hemoproteins collectively called cytochrome P450 (EC 1.14.14.1) is the major user of the heme synthesized in hepatocytes. Some isoenzyme(s) of cytochrome P450 are constitutively expressed in the liver, while most of them are expressed at an extremely low level under normal conditions but could be massively induced by the administration of a variety of xenobiotics, with phenobarbital (PB)’ and 3-methylcholanthrene (MC) as major prototypes. We have recently reported that a congenitally icteric strain of rats (Gunn rats) exhibited a response to MC administration different from that of Wistar rats (1): in male Gunn rat the overall cytochrome P450 content of hepatic microsomes is unchanged after administration of MC and this apparent lack of induction is attributed to a balance between the rise in P450 IAl/IAB (MC-inducible isoenzymes) and the decrease in the male constitutive P450 IICll, as shown by immunoquantitation. In i To whom correspondence should be addressed. ’ Abbreviations used: PB, phenobarbital; MC, 3-methylcholanthrene; TPO, tryptophan pyrrolase. 0003.9861/91 $3.00 Copyright 0 1991 by Academic Press, All rights of reproduction in any form
contrast, the response to PB administration was similar in Gunn and Wistar rats (i.e., increase in total P450 and in specific isoenzymes IIBl/IIBS). We have postulated a possible regulatory effect of the level of intracellular heme on the synthesis of cytochrome P450 isoenzymes. To assess this hypothesis, the present study was undertaken to compare the influence of inducer administration on heme availability in livers from Wistar and Gunn rats. According to Badawy and co-workers (2, 3), tryptophan pyrrolase (TPO) is believed to be in dynamic equilibrium with the free heme pool in such a way that minor fluctuations in this pool are reflected by modifications in enzymatic activity of the hemoprotein. We have investigated the effects of variations in heme availability (subsequent to xenobiotic treatment or in viva injection of heme) on TPO activity in Wistar and Gunn rats. Since heme availability seems to be reduced in Gunn rat liver, we have further examined its catabolic pathway into biliverdin through the microsomal heme oxygenase. MATERIALS
AND
METHODS
Two-month-old homozygous male Gunn Pretreatment of animals. rats (CSEAL-CNRS, France, inbred in our animal unit), heterozygous male Wistar/Gunn rats (male Gunn X female Wistar, inbred in our animal unit) and male Wistar rats (IFFA-CREDO, France) were given ip MC (20 mg/kg), PB (80 mg/kg), or dexamethasone (200 mg/kg). Animals were killed 18 h after a single injection and livers were excised, perfused, and immediately frozen in liquid nitrogen for TPO activity determinations. For the measurement of cytochrome P450 and heme oxygenase, animals were treated daily for 3 days. Hematin hydrochloride (70 mg/kg ip) was dissolved in dimethylformamide and animals were killed 3 h after injection. All the experiments were performed with fed animals. Enzyme assays. The activity of tryptophan pyrrolase (tryptophan 2,3-dioxygenase EC 1.13.11.11) was determined in liver homogenates either in the absence (holoenzyme activity) or in the presence (total enzyme activity) of in uitro-added (2 FM) hem&in according to the procedure of Badawy and Evans (2). Homogenates were incubated up to 75 min and tryptophan pyrrolase activity was calculated from the increase in kynurenine absorbance at 365 nm during the linear phase. Microsomal fractions were prepared as previously described (1). Cytochrome P450 content was determined from the CO difference spectrum 407
Inc. reserved.
408
CELIER
Pretreatment WlS,cIl
holo
ws
w*rtar
total
AND
01 rats 0
Gunn
holo
u
Gunn
total
FIG. 1. Tryptophan pyrrolase activities in Wistar and Gunn male rat liver homogenates. W, Wistar; G, Gunn; UT, untreated; MC, 3-methylcholanthrene-treated; PB, phenobarbital-treated; holo, holoenzyme activity determined without in vitro addition of hematin; total, total enzyme activity measured after in vitro addition of 2 pM hematin. *P < 0.001 compared to UT W (comparison between holoenzyme activities and total enzyme activities, respectively). OP < 0.001 compared to UT G (comparison between holoenzyme activities and total enzyme activities, respectively).
of dithionite-treated microsomes as described by Omura and Sato (4) and protein concentrations were measured according to Lowry et al. (5). Heme oxygenase (EC 1.14.88.3) was assayed by the method of Tenhunen et al. (6) by determining the formation of bilirubin by microsomes in the presence of an excess of exogenous biliverdin reductase.
RESULTS
TPO is classically assayed in crude homogenates to reflect the amount of heme available in the tissue (holoenzyme) but also after in vitro addition of exogenous hemin to fully saturate the apoprotein and to determine its total activity. In preliminary experiments, we have assayed TPO activities in homogenates prepared from freshly excised liver and in homogenates prepared from frozen samples of the same liver. No difference was noticed between both types of preparation (1.11 + 0.14 in frozen versus 1.25 f 0.20 pmol . g liver-‘. h-l in fresh samples). Furthermore, the effect of an in vitro addition of hemin was strictly similar in both types of preparations (1.44 + 0.12 in frozen versus 1.47 f 0.21 prnol. g liver-‘. h-’ in fresh livers). These results validate our experimental procedure. As shown in Fig. 1, TPO activity was significantly lower (by ~50%) in control Gunn rats compared to Wistar rats. The decrease is similar in both holoenzyme and total enzyme activities and consequently the percentage of saturation remains unchanged (77% in Wistar versus 76% in Gunn rats). The addition of a higher concentration of hemin (4 versus 2 PM) in the incubation mixture does not increase total TPO activity in Gunn rats and indicates that the existing protein is fully saturated under our incubation conditions and represents the total amount of TPO (data not shown). To exclude a possible interference
CRESTEIL
of bilirubin in the TPO activity determination, we have checked that the in vitro addition of exogenous bilirubin to Wistar liver homogenates (then mimicking the Gunn situation), does not modify TPO activity determinations (data not shown). Administration of MC does not significantly affect TPO activity in Wistar rats, whereas a moderate decrease is observed in Gunn rats. PB treatment elicits a significant rise in TPO activity in Gunn rats, leading to a level of activity similar in both strains. Finally, corticosteroids have been reported to increase the synthesis of apotryptophan pyrrolase. In our experiments we have used dexamethasone for treatment of rats and Table I shows that TPO activity is significantly increased in both strains of rats. The percentage of induction is higher in Gunn rats although the activity remains less intense than in Wistar rats. In viva hematin injection at a dose of 35 mg/kg is known to result in a surge in TPO activity (2). In our experiments we have used a dose of 70 mg/kg to obtain clear-cut results with Gunn rats, given that 35 mg/kg produced inconstant effects in this strain. The enzyme is fully saturated in uiuo with endogenous and injected heme and the level of activity is increased in both homozygous animals and heterozygous Wistar/Gunn (W/G) animals (Table II). This clearly demonstrates that the enzyme remains inducible in the Gunn strain by a large excess of heme. To assess the possibility that an accelerated catabolism of heme could be the primary cause in the reduction of the heme pool in Gunn rats, the microsomal heme oxygenase activity was examined. Results (Table III) show that in untreated rats, activity is less intense in Gunn than in Wistar rats (-50%). MC or PB treatment is as-
TABLE
I
Effect of Dexamethasone Pretreatment on Tryptophan Pyrrolase Activities in Wistar and Gunn Rat Livers
In vitro heme addition Wistar Control DEX-treated Gunn Control
+ +
Tryptophan pyrrolase (Fmol * h-i. g-i liver)
Percentage of saturation
1.11 1.44 3.12 4.08
f f f f
0.14 0.12 0.35* 0.32*
77
0.49 0.64 2.68 2.98
f f f f
0.23* 0.25* 0.12** 0.12**
76
76
~ +
DEX-treated +
90
Note. Results are shown as means f SEM for three animals in each group. +, in vitro added hemin (2 pM). * P < 0.001 when compared to control Wistar rat. ** P < 0.001 when compared to control Gunn rat.
REGULATION TABLE
OF P450 EXPRESSION
II
Tryptophan Pyrrolase Activities in Wistar, Gunn, and Wistar/Gunn Rat Livers with or without in. Viuo Injection of Hematin In uiuo heme injection Wistar
+
Gunn + Wistar/Gunn +
Tryptophan pyrrolase (pm01 . h-i * g-i liver) 1.11 4.48 0.49 3.93 0.96 4.40
f 0.14 f 0.22 + 0.23* k 0.87 IL 0.42 f 0.48
Percentage of saturation 71 100 76 100 70 100
Note. Results are shown as mean values + SEM for three animals in each group. +, animals injected i.p. with hematin hydrochloride (70 mg/ kg) and killed 3 h after injection. * P < 0.001 when compared to control Wistar rat without heme injection.
sociated with a significant reduction of heme oxygenase activity in Wistar but not in Gunn rats. To confirm the involvement of hepatic unconjugated bilirubin in the reduction of the heme pool and to exclude another genetic defect, we have used heterozygous Wistar/ Gunn rats of first generation which present a normal bilirubinaemia. After MC administration, their hepatic content of cytochrome P450 is induced by the same extent as in Wistar rats (1.80 k 0.05 in MC-treated versus 1.10 & 0.02 nmol . mg -’ protein in control W/G rats). TPO activity in W/G rats is closer to that of homozygous Wistar than to homozygous Gunn (0.96 + 0.42 for holoenzyme and 2.14 k 0.37 pmol . g liver ‘. hem1for total enzyme activity) and definitively does not exhibit intermediate values. Finally, as shown in Table III, heme oxygenase activity of W/G rats is identical to that of Wistar rats. All together, these data suggest that in homozygous Gunn rats, the accumulation of bilirubin is associated with a lower heme availability and leads to a limitation in the synthesis of cytochrome P450.
409
BY HEME
tained when hemolysis occurs after thawing of liver samples containing residual traces of blood and (ii) there is no difference in the saturation index between perfused and nonperfused livers (see Results section). The examination of results shows that TPO activity is selectively reduced in homozygous Gunn compared to Wistar and heterozygous W/G rats in terms of activity, while the saturation of enzyme is unchanged. This could be interpreted as a reduced availability of heme and we have designed several experiments to confirm this hypothesis: (i) This limitation can be reversed by an in uiuo injection of hematin. A low dose of hematin is suitable to fully saturate the existing apoprotein (8) while a higher but nontoxic dose produces a surge in total activity, switching on the synthesis of the apoprotein (2). This could explain that the synthesis of apoprotein is adjusted to long-term variations of heme flow, like in the Gunn liver situation where the low heme availability down-regulates the apotryptophan pyrrolase synthesis. (ii) This assumption supposes that the TPO apoprotein synthesis could be stimulated in Gunn as in Wistar rat liver by other modulators. Cortisol and other steroids induce TPO apoprotein (2, 9, 10). Similar conclusions can be drawn for dexamethasone: the protein synthesis is stimulated in both strains and is likely associated with an enhancement of heme synthesis to maintain the saturation percentage at similar levels in Wistar and Gunn rats. All together, these data definitively validate the assumption that a lower amount of heme is available in
TABLE
III
Heme Oxygenase Activities in Wistar, Gunn, and Wistar/Gunn Rat Liver Microsomes Heme oxygenase (nmol f mini’ . rng-’ protein) Wistar
DISCUSSION
The validity of this approach is based on the assumption that tryptophan pyrrolase reflects the fluctuations of the heme availability (2) although some technical artifacts can make the interpretation questionable. In a first attempt, we delineated the accuracy of TPO determinations (see above). Assay conditions were acceptable even if some parameters differ from those in the literature: for example, in our hands, the percentage of heme saturation is higher than those previously reported (2). A possible explanation could be that heme released from contaminating hemoglobin or methemoglobin is incorporated into apotryptophan pyrrolase (7). However, this explanation is unlikely because (i) the concentration of hemoglobin injected in these experiments is considerably higher than that ob-
Control PB-treated MC-treated
0.133 t 0.023 0.056 t 0.017* 0.051 f 0.002* Gunn
Control PB-treated MC-treated
0.066 f 0.015” 0.046 f 0.011 0.066 f 0.002 Wistar/Gunn
Control
0.128 f 0.003
Note. Results are shown as mean values +- SEM for three animals in each group. * P < 0.001 compared to control. ** P < 0.001 compared to Wistar rats.
410
CELIER
AND
Gunn rat liver. This fluctuation of heme availability cannot be attributed to an accelerated degradation of heme since heme oxygenase activity is reduced in homozygous Gunn liver, but might rather result from a decrease in heme synthesis. In heterozygous W/G rats, there is no accumulation of hepatic bilirubin, and heme availability and heme oxygenase activity are comparable to that of Wistar rats as is the induction of cytochrome P450 by MC. Like for TPO, the synthesis of active cytochrome P450 requires the coordination of two multienzymatic pathways: the apoprotein synthesis on one hand and the heme biosynthesis on the other hand. In untreated Gunn rats, the level of the hepatic heme pool, although low, is sufficient to maintain a P450 concentration comparable to that of Wistar rats (1). The response to induction depends on the molecule administered: PB produces an important rise in heme synthesis (11) which results in an increase in available heme leading subsequently to an enhancement in the TPO activity and to the insertion of heme into the newly synthesized molecules of cytochrome P450 IIBl and -1IB2. MC produces a comparable requirement for heme but without activation of its synthesis (11): this results in a diminution of the free heme and a decrease in TPO activity. In Gunn rats, heme availability is reduced by MC and becomes insufficient to allow the insertion of heme in all the molecules of apocytochrome P450 synthesized. Several authors have demonstrated that apocytochrome P450 could be rapidly degraded in the liver when deprived of heme (12, 13). This explains why MC failed to increase the overall P450 content, as well as the
CRESTEIL
immunochemical level of cytochrome treated homozygous Gunn rats. The pairment of heme biosynthesis remains cidated but might involve the inhibition ogen I synthase by bilirubin (14).
P450 proteins, in origin of the imto be further eluof uroporphyrin-
REFERENCES 1. Celier, C., and Cresteil, T. (1989) Biochem. Pharmacol. 38, 28252832. 2. Badawy, A. A. B., and Evans, M. (1975) Biochem. J. 150,511-520. 3. Badawy, A. A. B., Morgan, C. J., and Davis, N. R. (1986) in Porphyrins and Porphyrias (Nordmann, Y., Ed.), Colloque INSERM, Vol. 134, pp. 69-81, John Libbey Eurotext, Montrouge. 4. Omura, T., and Sato, R. (1964) J. Biol. Chem. 239, 2370-2378. 5. Lowry, 0. H., Rosebrough, N. J., Farr, A. L., and Randall, R. J. (1951) J. Biol. Chem. 193, 265-275. 6. Tenhunen, R., Marver, H. S., and Schmid, R. (1969) J. Biol. Chem. 244, 6388-6394. 7. Wyman, J. F., Gollan, J. L., Settle, W., Farrel, G. C., and Correia, M. A. (1986) Biochem. J. 238, 837-846. 8. Welch, A. N., and Badawy, A. A. B. (1980) Biochem. J. 192, 403410. 9. Feigelson, P., and Greengard, 0. (1962) J. Biol. Chem. 237,37143747. 10. Schimke, R. T., Sweeney, E. W., and Berlin, C. M. (1965) J. Biol. Chem. 240, 322-331. 11. Baron, J., and Tephly, T. R. (1970) Arch. Biochem. Biophys. 139, 410-420. 12. Dwarki, V. J., Francis, V. N. K., Bhat, G. J., and Padmanaban, G. (1987) J. Biol. Chem. 262, 16,95&16,962. 13. Bonkovsky, H. L., Sinclair, J. F., Healey, J. F., Sinclair, P. R., and Smith, E. L. (1984) Biochem. J. 222, 453-462. 14. Kohashi, M., Tse, J., and Piper, W. N. (1984) Life Sci. 34, 193196.