Ortiz de Montellano Paul R.(Ed.) Cytochrome P450. Structure, Mechanism, and Biochemistry

Подождите немного. Документ загружается.

Hormonal Regulation of Liver Cytochrome P450 Enzymes

355

cell surface GHR abundance may, at least in part,

be due to differential effects of intermittent vs

continuous GH stimulation^^^ and could play a

role in the activation of distinct intracellular sig-

naling pathways by chronic (female) as compared

to intermittent (male) GH stimulation.

4.2.3.1.

Significance of GH pulse

fi^equency. Studies have been carried out to deter-

mine which of the three descriptive features of a

GH pulse—^namely, GH pulse duration, GH pulse

height, and GH pulse frequency—is required for

proper recognition of a GH pulse as "masculine."

Direct measurement of the actual plasma GH

profiles achieved when GH is administered to

hypophysectomized rats by twice daily s.c. GH

injection

(i.e.,

the intermittent GH replacement pro-

tocol most commonly used to stimulate CYP2C11

expression) has revealed broad peaks of circulating

GH, which last as long as 5-6 hr^^. These sustained

GH "pulses" are effective in stimulating expression

of the male-specific CYP2C11, provided that they

are not administered in close succession. It is thus

apparent that physiological GH pulse duration (<2

hr) is not required to elicit a male CYP response.

Studies carried out in GH-deficient rat models

(either dwarf rats or rats depleted of adult circulat-

ing GH by neonatal MSG treatment) demonstrate

that GH pulse height is also not a critical factor for

stimulation of CYP2C11 expression^^^'

^^^.

This

finding can be understood in terms of the K^ of

the GH-GHR complex, which at 10"^^ M (~2

ng/ml)^^^,

is only 1% of

the

peak plasma hormone

level in adult male rats. In contrast, GH pulse fre-

quency is a critical determinant for GH stimulation

of a male pattern of liver P450 expression, as

shown in hypophysectomized rats given physiolog-

ical replacement doses of GH for 7 days by inter-

mittent intravenous injections at frequencies of 2,4,

6, or 7 times per

day^^.

Analysis of liver CYP2C11

levels in these rats revealed a normal male pattern

of liver CYP2C11 gene expression in response to 6

GH pulses per day (which approximates the normal

male plasma GH pulse frequency), as well as in

response to GH pulses given at lower frequencies, 2

or 4 times per day (e.g., Figure 9.3). However,

hypophysectomized rats are not masculinized by

7 daily GH pulses, indicating that the hepatocyte

does not recognize the pulse as "masculine" if GH

pulsation becomes too frequent. Hepatocytes thus

require a minimum GH off time (—2.5 hr in the

hypophysectomized rat model used in these

studies), which implies the need for an obligatory

recovery period to effectively stimulate CYP2C11

expression. This condition is not met in the case of

hepatocytes exposed to GH continuously (female

hormone profile). This recovery period may serve

to reset the cellular signaling apparatus, for exam-

ple,

by replenishing GHRs at the cell surface (see

below).

4.2.3.2. RoleofGH receptor (GHR). The

effects of GH on hepatocytes and other responsive

cells are transduced by GHR, a 620 amino acid

cell surface transmembrane protein^^^ belonging

to the cytokine receptor superfamily^^"^. GHR is

comprised of a 246 amino acid extracellular

domain that binds GH, a single transmembrane

segment, and a 350 amino acid intracellular

domain that participates in the intracellular signal-

ing events stimulated by GH^^^'

^^^.

X-ray crystal-

lographic and other studies establish that a single

molecule of GH binds in a stepwise manner to

two GHR molecules to yield receptor dimers:

+GHR

GH + GHR -> GH-GHR -> GH-(GHR)2

^^^^

^^7

(Figure 9.4).

The four helix bundle protein GH is proposed

to initially bind to a single receptor molecule by

contacts that involve amino acids comprising

GH's Site 1, followed by binding of a second

molecule of GHR, which interacts with Site 2

on the GH molecule to give a heterotrimeric

GH-(GHR)2 complex. GH-induced receptor

dimerization is reversible, and the equilibrium

may be shifted in favor of monomer formation

in the continued presence of excess GH:

CTH

GH-(GHR)2^2GH-GHR. Other, more recent

studies suggest that GH may bind to, and thereby

activate a preformed receptor dimer at the cell sur-

face ^^^' ^^^. Independent of whether the receptor

dimer is preformed, however, it is clear that the

conformational changes that accompany forma-

tion of the active GH-(GHR)2 complex are neces-

sary, and probably sufficient, for stimulation of

GH-induced intracellular signaling events ^^^

(Figure 9.4). Although GHR dimerization is thus

required for most GH responses, some GH

responses might not require receptor dimeriza-

tion^^ ^ and indeed, might be mediated by

monomeric GH-GHR complexes. Conceivably,

the distinct patterns of liver P450 gene expression

induced by continuous plasma GH (female GH

pattern; CYP2C12 expression) as compared to

356

David J. Waxman and Thomas K.H. Chang

GHR

Physiological [GH]

Intermediate

High [GH]

Signaling via

Protein Tyrosine

Phosphorylation

Figure 9.4. Activation of

GHR

by GH-induced sequential dimerization mechanism. GHR is shown localized in

the plasma membrane, and sites

and 2 of

the

GH molecule (see text) are as indicated. The active receptor dimer

dominates at physiological GH concentrations. Model based on data presented in ref

[128].

pulsatile GH (male GH pattern; CYP2C11

expression) might, in part, arise from distinct GH

signaling pathways perhaps stimulated by

monomeric (GH-GHR) as compared to dimeric

(GH-(GHR)2) hormone-receptor complexes. GH

mutants and analogs that bind GHR without

effecting fiinctional receptor dimerization'^^~^^^

could be used to test this hypothesis.

In intact male rat liver, GHR internalizes to an

intracellular compartment coincident with its

stimulation by plasma GH pulses, and then reap-

pears at the cell surface at the time of the next hor-

mone pulse'^^' '^^. Other studies suggest that GHR

undergoes endocytosis constitutively, that is, in a

ligand-independent manner. GHR internalization

involves coated vesicles and ultimately takes the

receptor to lysosomes for degradation. GHR endo-

C5^osis and degradation both require (a) an intact

ubiquitin conjugation system, which targets a spe-

cific 10 amino acid-long cytoplasmic GHR tail

sequence, and (b) 26S proteasome activity, as evi-

denced by the inhibitory effects of the proteasome

inhibitor IVIG132. Interestingly, although cellular

ubiquitination activity is required for receptor

endocytosis, GHR itself does not need to undergo

ubiquitination, as shown using a mutant GHR

devoid of

its

cytoplasmic lysine residue targets for

ubiquitination'^''' '^^. Thus, the ubiquitin-protea-

some system is a major regulator of intracellular

GHR trafficking.

4.2.4.

Role of STAT5b in Sex-Dependent

CYP Expression

4.2.4.1.

GH signaling pathways involving

STAT transcription factors. How does GH

impart sex-dependent transcriptional regulation to

liver P450 genes? To answer this question, we may

consider the following hypotheses: (a) that cell sur-

face GHRs can discriminate between the two tem-

porally distinct patterns of plasma GH stimulation,

and (b) the receptors can then transduce this infor-

mation to the nucleus, where both pulse and

Hormonal Regulation of Liver Cytochrome P450 Enzymes

357

continuous GH-stimulated transcriptional events

occur. Presumably, GH-activated GHR accom-

plishes this by activating two distinct pathways of

intracellular signaling, one in response to GH pulses

and the other in response to continuous GH stimu-

lation (Figure 9.5). Studies of GH-induced signal

transduction pathways^^^^^^ have highlighted the

importance of the GH-bound receptor dimer in acti-

vating JAK2, a GHR-associated tyrosine kinase that

initiates several downstream pathways of intracellu-

lar protein tyrosine phosphorylation (Figure 9.4).

Other studies have shown that a distinct pattern of

nuclear protein tyrosine phosphorylation is

induced in rat liver when the effects of the male

pulsatile plasma GH hormone profile are com-

pared to those of the female pattern of continuous

GH stimulation^^^. This led to the discovery that

an intracellular signaling protein and transcription

factor, termed STATSb, is present in its nuclear,

tyrosine phosphorylated form at a substantially

higher level in male rat liver than in female rat

liver^^^. STATs are latent cytoplasmic transcrip-

tion factors that are activated by tyrosine phos-

phorylation induced by a variety of cytokines and

growth factors, and were first discovered as signal

mediators that carry transcription signals into the

nucleus in the interferon signaling pathway^"^^.

In hypophysectomized rat liver, where there is

no endogenous GH signaling, there is little or no

tyrosine phosphorylated STATSb protein in the

nucleus; essentially all of the STAT5b protein is

found in the C3^osolic fraction, where it resides in a

3.5 hr

Male-specific CYP

Steroid OHases

(CYP2C11)

Female-specific CYP

Steroid OHases

(CYP2C12)

Figure 9.5. Distinct intracellular signaling pathways induced by GH are proposed to be activated by plasma GH

pulses, leading to male-specific CYP expression (left), and by continuous GH stimulation, leading to female-specific

CYP expression (right).

358 David J. Waxman and Thomas K.H. Chang

latent, inactive (non-tyrosine phosphorylated)

form. However, when a hypophysectomized rat is

injected with a single pulse of GH, STATSb protein

appears in the nucleus in its active, tyrosine phos-

phorylated state within 10-15 min^^^'

^"^^

This tyro-

sine phosphorylation reaction occurs on STATSb

tyrosine residue 699, enabling two STATSb mole-

cules to dimerize via mutual interactions between

the phosphotyrosine residue on one STATSb mole-

cule and the SH2 domain (a protein module that

recognizes and binds specifically to phosphotyro-

sine residues) on a second STATSb molecule. The

STATSb-STATSb dimer that is thus formed quickly

enters the nucleus, where it binds with high affinity

to DNA sites upstream of genes that are transcrip-

tionally activated in response to the initial GH

stimulus (Figure 9.6).

STATSb is not present in the nucleus at all

times in male rat liver. Rather, STATSb is repeat-

edly activated in concert with the onset of each

male plasma GH pulse. It thus undergoes repeated

cycles of translocation from the cytoplasm into the

nucleus, and then back out to the cytoplasm^^^' ^'^^.

For example, if the liver is excised from a rat killed

at the peak of

plasma GH pulse, then STATSb is

tyrosine phosphorylated and localized to the

nucleus, whereas if the liver is excised from a rat

killed at a time point between successive plasma

GH pulses, STATSb is inactive and cytoplasmic.

Moreover, in female rats there is generally a much

lower level of active, nuclear STATSb protein

(~S-10% that of the peak male liver level)

^'^^.

This

close temporal linkage between plasma GH pattern

and the state of liver STATSb activation has been

confirmed in intact male rats killed at times shown

to be specifically associated with spontaneous

peaks or troughs of the plasma GH

rhythm^"^"^.

The

key difference between male and female rat liver is

that STATSb is repeatedly, and efficiently, acti-

vated by plasma GH pulses in the case of the

Figure 9.6. Role of complex formed by GH, GH receptor, and the tyrosine kinase JAK2 in activation of STATSb

by tyrosine phosphorylation. JAK2 tyrosine phosphorylates itself and multiple tyrosines on the cytoplasmic tail of

GHR. Several of these sites serve to recruit STATSb to the receptor-kinase complex. STATSb is then tyrosine

phosphorylated, whereupon it dimerizes, translocates to the nucleus, and binds to DNA regulatory elements upstream

of target genes.

Hormonal Regulation of Liver Cytochrome P450 Enzymes

359

males,

but is inefficiently activated by the more

continuous GH profile of the females. The rela-

tively weak STATSb activation pathway in female

rat liver appears to reflect a partial desensitization

of this signaling pathway in response to the chronic

presence of GH in plasma (Figure 9.7)^'*^.

4.2.4.2. STATSb gene knockout mouse

model. The proposal that STATSb is a critical

factor in mediating GH-regulated liver P450 gene

expression is strongly supported by studies carried

out in mice that are deficient in STAT5b (STATSb

knockout mouse model) ^"^^ (see ref [147] for a

review). Disruption of the STAT5b gene results in

two striking phenotypes (Figure 9.8). These

phenotypes are seen in STATSb-deficient males

but not in corresponding females. First, there is a

global loss of GH-regulated, male-specific liver

gene expression, including male-specific P450

Activation of STATSb by Plasma GH Pulses

Proposed Role in Male-specific Liver Gene Transcriptionl

time

Plasma

Membrane

Cytosol

Nucleus

rm:

Transcriptional Activation:

Male-specific LiverCVP Genes

Figure 9.7. Proposed cycle for STATSb activation and deactivation in response to sexually dimorphic plasma GH

profiles. GH pulses, but not continuous GH, induce robust STAT5b tyrosine phosphorylation (pTyr). Nuclear

STATSb is dephosphorylated by a tyrosine phosphatase and then returned to the cytoplasm, where it may undergo

a subsequent round of activation/deactivation.

360 David J. Waxman and Thomas K.H. Chang

Loss of

STATSb

Direct

Impaired

Pulsatile

GH Signaling

in Liver

Indirect

Loss of Male-Characteristic

Body Weight Gain

Suppression of Male-

Specific Gene Expression

Stimulation of Female-Specific

Liver Gene Expression

II Impaired

Feedback-Inhibition

of Pituitary GH Release

Pertubation of

Sexually Dimorphic

Plasma GH Pattern

Figure 9.8. Impact of STAT5b loss in knockout (KO) mice. Shown in the box at right are the major effects of

STAT5b deficiency. These effects all are a direct result of

the

loss of GH-induced liver STAT5b signaling (model I),

rather than an indirect response to impaired feedback-inhibition of pituitary GH release (model

11),

as demonstrated

by the lack of responsiveness of hypophysectomized STATSb-deficient male mice to GH pulses^^^.

gene expression. Thus, in the absence of STATSb,

the liver does not express the male-specific P450s.

Moreover, the expression of female-specific, GH-

regulated liver P450s is increased to near-normal

female levels in STAT5-deficient male mice.

Thus,

the overall pattern of sexually dimorphic

liver gene expression is critically dependent on the

presence of STATSb. The elevated expression in

STATSb-deficient males of female-specific P4S0s

indicates that STATSb can serve as a negative reg-

ulator of female-specific liver P4S0s, in addition

to its positive regulatory effects on male-specific

P4S0 genes.

A second phenot3^e exhibited by STATSb-

deficient mice is that the male pubertal growth

spurt is absent^"^^. This growth deficiency does

not emerge until the beginning of puberty, and

is not seen in STATSb-deficient females. A male-

enhanced pubertal growth spurt is characteristic

of all mammals, including humans. In the case of

rodents, the pubertal growth spurt is augmented

by the strong growth stimulatory effects of the

male pattern of pulsatile GH secretion, which

explains why this growth response is enhanced in

the males. The two major phenotypes that charac-

terize STATSb knockout mice are also seen in

STATSa/STATSb double knockout

mice^"^^

but are

not seen in mice where the disruption is limited to

the STATS a gene^'*^' ^^^, whose protein-coding

sequence is ~90% identical to that of STATSb^^^

Hypophysectomy and GH pulse replacement stud-

ies establish that both major phenotypes of

STATSb-deficient mice are a direct response to

the loss of STATSb-dependent GH signaling in the

Hver, as opposed to indirect effects of the loss of

STATSb on the overall pattern of

secretion by

the pituitary gland'^^ (Figure 9.8).

4.2.4.3. Interaction of GH-responsive

CYP promoters with GH-activated STATSb.

The strong, repeated pulses of nuclear GH-

activated STATSb that occur in adult male rats

have been proposed to induce binding of STATSb

directly to STATS response elements found in pro-

moters of

STATS

target genes, which may include

sex-dependent P4S0 genes, and thereby medi-

ate GH pulse-stimulated gene transcription^ ^^.

Consistent with this hypothesis, STATS response

elements matching the consensus sequence

TTC-NNN-GAA have been found upstream of

several male-specific rat liver P4S0 genes, includ-

ing CYPs 2C11, 2A2, and 4A2 (ref. [1S2]). GH-

stimulated CYP promoter-luciferase reporter

activity has been demonstrated using the corre-

sponding isolated STATS response elements,

although the magnitude of the GH- and STATSb-

dependent gene induction is small, generally only

~2-3-fold'^^' '^^. Moreover, although pulsatile

STATSb signaling is first seen in young male rats

at ~S weeks of

age,

when liver CYP2C11 expres-

sion is first detected, precocious activation of

Hormonal Regulation of Liver Cytochrome P450 Enzymes

361

STAT5b, achieved in 3-week-old male rats given

pulsatile GH injections, does not lead to preco-

cious CYP2C11 gene induction^'^^. These and

other findings suggest that STAT5b may in part

act in an indirect manner, by regulating the tran-

scriptional activity of liver-enriched transcription

factors (HNFs, hepatocyte nuclear factors) that

cooperate with STAT5b to control the expression

of sexually-dimorphic liver P450 genes^^^'

i^^-^^e

4.2.4.4. Interactions between STAT5b and

liver transcription factors regulating sex-

specific CYPs. Liver-enriched transcription

factors are key determinants of the liver-specific

expression of many hepatic P450s (ref. [157]),

including sex-dependent P450s. Functional

analysis of sex-specific liver CYP proximal pro-

moters has led to the identification of the liver

transcription factors HNFla and HNF3p as

strong ^ra«5-activators of CYP2C11 (ref. [152]).

By contrast, HNFSp and HNF6 serve as strong

^ra«^-activators of CYP2C12 (refs

[155],

[158]),

with synergistic interactions between these two liver

transcription factors leading to an overall ~300-fold

activation of CYP2C12 promoter activity^ ^^. Other

studies indicate a role for C/EBPa^^^' ^^^ and

HNF4^56 in the regulation of CYP2C12 trans-

cription. Interestingly, two of the factors regulating

CYP2C12 expression, HNF6 and HNF3P, require

GH for maximal expression in liver, with HNF6

being expressed at 2-3-fold higher levels in female

compared to male rat liver^^^. This sex difference in

HNF6 levels is too small to account for the

>50-fold higher level of CYP2C12 in female com-

pared to male liver. Similarly, the STAT5b-respon-

siveness of male CYP promoter-derived regulatory

elements^^^' ^^^ is too weak to account for the

high degree of male-specificity (> 100-fold) that

characterizes male-specific CYP expression.

Accordingly, additional factors and GH regulatory

mechanisms are likely to be required for the sex-

specific pattern of gene expression seen in vivo.

One such mechanism may involve an as yet

uncharacterized "regulator of sex-limitation"

("Rsl"), which has been defined genetically and

appears to repress certain GH-regulated, sex-

dependent liver CYP genes in female mouse liver

in a manner that is independent of GH action^^^.

Another possible mechanism for GH-regulated

sex-specific CYP expression involves interactions

between STAT5b and HNFs. For example, studies

of the CYP2C12 promoter have shown that

STAT5b can substantially block the induction

of promoter activity by HNF3P and HNF6

(ref. [152]). This inhibitory action of male GH

pulse-activated STAT5b on HNF3P- and HNF6-

stimulated CYP2C12 transcription is consistent

with the de-repression (i.e., upregulation) of cer-

tain female-specific, GH-regulated mouse P450

genes seen in male STAT5b-deficient mice^"^^' ^^^.

The synergistic interaction between HNF6 and

HNF3P in stimulating CYP2C12 promoter activ-

ity illustrates how small differences in the levels

of these factors in liver cells in vivo (cf. the 2-3-

fold higher expression of HNF6 in females) might

lead to much larger differences in CYP2C12 gene

expression, particularly when coupled to the

inhibitory action of GH pulse-activated STAT5b

in males. Other studies point to additional com-

plexities, based on the potential of STAT5b to

upregulate CYP2C12 gene transcription via an

upstream pair of

STAT5

response elements^^^, and

the potential role of an inhibitory, COOH-terminal

truncated STAT5 form^^^

4.2.4.5. Downregulation of hepatic

STAT5b signaling. Other issues relating to GH

and the STAT5b signaling pathway that are of cur-

rent research interest include how the cycle of

STAT5b activation is turned off at the conclusion

of each GH pulse, and how STAT5b is subse-

quently returned to the cytoplasm in an inactive

form, where it apparently waits for ~2-2.5 hr until

it can be re-activated by the next pulse of GH

(Figure 9.7)^^^. Another important question is why

robust activation of the STAT5b signaling path-

way is not achieved in female liver. The answer to

these questions may, in part, involve an intriguing

family of inhibitory proteins, referred to as SOCS

and CIS proteins, which turn off signals to various

hormones and cytokines, including GH^^^'

^^^.

the case of GH signaling, SOCS proteins bind to

the GHR-JAK2 tyrosine kinase complex,

enabling them to inhibit GH signaling by a com-

plex series of interrelated mechanisms ^^^. One of

these inhibitory factors, CIS, may be induced to a

higher level by the continuous (female) GH pat-

tern than by the pulsatile (male) GH pattern. This

inhibitor of

STAT5b

signaling has been implicated

in the downregulation of GH-induced STAT5b

signaling in liver cells exposed to the female GH

pattem^^^

362 David J. Waxman and Thomas K.H. Chang

4.3.

Regulation by Thyroid

Hormone

4.3.1.

Cytochromes P450

Although GH is the major regulator of specific

liver P450s, thyroid hormone also plays a critical

role.

The major thyroid hormones,

and

T4,

pos-

itively regulate some^^' ^^ but not alP^ of the

female-predominant liver P450 enzymes, while

they negatively regulate several of the male-

specific enzymes^^'^"^ (Table 9.1). These effects

of thyroid hormone are operative at the level of

mRNA expression, and are independent of the

indirect effects that thyroid hormone has on liver

P450 levels as a consequence of its effects on liver

GHRs^^^ and its stimulation of GH gene tran-

scription and GH secretion by the pituitary^^^.

Molecular studies of these effects of thyroid hor-

mone have not been carried out.

animals^'^^. The induction of rat hepatic P450

reductase by thyroid hormone in rats occurs by

transcriptional ^^^ and post-transcriptional mecha-

nisms ^^^ and appears to involve enhanced protein

stability in hyperthyroid rat liver*

^'^.

P450 reductase

levels are also modulated by thyroid hormone sta-

tus in several extrahepatic tissues*^^ Conceivably,

interindividual differences in P450 reductase levels

may occur in response to physiological or patho-

physiological differences in circulating th)Toid hor-

mone levels and could be an important contributory

factor to individual differences in P450 reduc-

tase/c54ochrome P450-catalyzed carcinogen metab-

olism and carcinogen activation reactions.

5. Alteration of Liver P450

Expression by Hormonal

Perturbation

4.3.2.

NADPH-Cytochrome P450

Reductase

Thyroid hormone is also required for full

expression of NADPH-cytochrome P450 reduc-

tase,

a flavoenzyme that catalyzes electron transfer

to all microsomal P450s. P450 reductase is an

obligatory, and oflen rate-limiting electron-transfer

protein that participates in all microsomal P450-

catalyzed drug oxidation and steroid hydroxylase

reactions *^^' '^^. This thyroid hormone dependence

P450

reductase enzyme expression is evidenced

by the major decrease (>80% reduction) in liver

microsomal P450 reductase activity and P450

reductase mRNA levels that occurs following

hypophysectomy'^^ or in response to methimazole-

induced hypothyroidism^'''. It is further supported

by the reversal of this activity loss when thyroxine

(T4),

but not GH or other pituitary-dependent fac-

tors,

is given at a physiological replacement

dose'^^'

^^K

Restoration of liver P450 reductase

activity in vivo by T4 replacement also effects a

substantial increase in liver microsomal P450

steroid hydroxylase activities. A similar effect can

be achieved when liver microsomes isolated from

hypophysectomized rats are supplemented with

exogenous, purified P450 reductase, which prefer-

entially stimulates steroid hydroxylation catalyzed

by microsomes prepared from thyroid-deficient

As discussed earlier in this chapter, many,

although not all, liver P450s are under hormone

regulatory controls. An individual's circulating

hormonal profile can, however, be altered under

certain situations, including drug therapy, exposure

to chemicals found in the environment, and disease

states such as diabetes and liver cirrhosis. The

resultant changes in circulating hormone levels or

alterations in hormone secretory dynamics could,

therefore, influence the expression of specific liver

P450s. The following sections describe some of

the factors that are known to cause hormonal per-

turbation and discuss the impact of these changes

on liver P450 expression and on P450-dependent

drug and xenobiotic metabolism and toxicity.

5.1.

Modulation by Drugs

The anticancer drugs cisplatin'^^' ''^^, cyclo-

phosphamide''^^' '^^, and ifosfamide'^^ alter the

profile of P450 enzyme expression in liver and

perhaps other tissues, at least in part due to the hor-

monal perturbations that these cytotoxic agents

induce. Treatment of adult male rats with a single

dose of cisplatin depletes serum androgen, and this

effect persists for up to 28 days after drug adminis-

tration'^^. Serum androgen depletion by cisplatin is

associated with a feminization of hepatic liver

enzyme expression. Thus, cisplatin-treated male

rats have elevated levels of the female-predominant

Hormonal Regulation of Liver Cytochrome P450 Enzymes

363

CYP2A1,

CYP2C7, and steroid 5a-reductase, but

have reduced levels of

the

male-specific CYP2A2,

CYP2C11,

and CYP3A2 (refs

[175],

[176]). The

effects of cisplatin on androgen levels may result

from the drug's action on the testes^^^'

^^^;

however,

effects on the hypothalamus are also suggested to

contribute, both to the observed depletion of circu-

lating testosterone and the resultant alteration in

liver P450 expression^ ^^. Cisplatin treatment of

adult female rats severely decreases circulating

estradiol levels and significantly reduces the

expression of the estrogen-dependent CYP2A21,

CYP2C7, and CYP2C12 (ref [176]).

Serum testosterone is also depleted in adult

male rats treated with cyclophosphamide^^^'

^^^^^^^

or ifosfamide^^^ and this depletion is associated

with feminization of liver enzyme profiles^^^'

^^^

a manner similar to that produced by cisplatin.

While endogenous androgen secretion can be

stimulated in cyclophosphamide-treated rats by

the luteinizing hormone analog chorionic

gonadotropin, the resultant increase in serum

testosterone does not reverse the loss of hepatic

CYP2C11 expression^^^. This observation is anal-

ogous to the earlier finding that the suppression of

CYP2C11 by 3,4,5,3',4',5'-hexachlorobiphenyli82

is not causally related to the associated depletion

of serum testosterone ^^^. Consequently, modu-

lation of liver enzyme expression by cyclo-

phosphamide may involve action at the

hypothalamic-pituitary axis, which establishes the

sex-dependent plasma GH profile that in turn dic-

tates the expression of CYP2C11 and other sex-

dependent liver P450 enzymes, as discussed earlier

in this chapter. CYP2C11 can also be suppressed

by other mechanisms, as demonstrated by the

finding that CYP2C11 levels are suppressed by

the anticancer drug l-(2-chloroethyl)-3-cyclo-

hexyl-1-nitrosourea (CCNU; lomustine) with-

out affecting circulating testosterone levels ^^'^.

Conceivably, CCNU may act directly on the hypo-

thalamic-pituitary axis to alter key signaling ele-

ments in the ultradian rhythm of circulating GH.

Other drugs that suppress hepatic CYP2C11

and CYP3A2 levels include chloramphenicols^^

and cyclosporine^^^' ^^^. The effects of chlo-

ramphenicol are strain-specific, occurring in

Sprague-Dawley rats but not in Fischer 344 rats.

Moreover, this suppression is accompanied by a

modest reduction in plasma levels of thyroxine but

not testosterone s^^. GH does not appear to play a

role in the suppression of

CYP2C11

and CYP3A2

by cyclosporine, a drug that does not alter the

plasma GH peak amplitude, number, or dura-

tion^^l Phenobarbital^^' s^^'

^^o,

dexamethasoneS9\

and S-fluorouracil^^^ also reduce hepatic CYP2C11

expression, but the mechanism(s) by which these

effects occur have not been elucidated.

5.2. Modulation by Polycyclic

Aromatic Hydrocarbons

Exposure of adult male rats to polycyclic

aromatic hydrocarbons, including 3-methylcholan-

threne (3MC)i^2, i90, 193^ 2,3,7,8-tetrachlorodi-

benzo-p-dioxin^^"^, anthracene, benz(a)anthracene,

dibenz(a,c)anthracene, dibenz(a,/i)anthracene, and

7,12-dimethylbenz(a)anthracene^^^, leads to decre-

ases in hepatic CYP2C11 protein and activity levels.

In the case of 3MC, this suppression reflects a

decrease in the rate of CYP2C11 transcription^^^.

The hormonal mechanisms by which polycyclic aro-

matic hydrocarbons modulate CYP2C11 expression

are not known, however, 3MC^^^ and 2,3,7,8-tetra-

chlorodibenzo-/7-dioxin^^^ have been reported to

decrease serum testosterone levels. 3MC may inter-

fere with the stimulation of

CYP2C11

expression by

GH^^^, but in a manner that does not involve

STAT5bi99. Interestingly, the extent of CYP2C11

suppression by polycyclic aromatic hydrocarbons

is correlated with Ah receptor-binding affinity and

Ah receptor transformation potency^ ^^. However, it

is not clear whether the Ah receptor plays a role in

the transcriptional suppression of

CYP2C11

by poly-

cyclic aromatic hydrocarbons^^^.

5.3. Modulation by

Pathophysiological State

5.3.1.

Diabetes

Uncontrolled insulin-dependent diabetes is

accompanied not only by defective carbohydrate

metabolism, which results in hyperglycemia,

hyperlipidemia, and hyperketonemia, but it is also

associated with hormonal perturbation, including

a reduction in circulating testosterone^^ ^~^^^,

th3a'oid hormone, and plasma

Gtf^'*'

^^^. As

described earlier in this chapter, these hormones

regulate many liver P450 enzymes, either directly

or indirectly. Accordingly, the diabetic state is

364 David J. Waxman and Thomas K.H. Chang

associated with profound changes in the levels of

various hepatic P450 enzymes. Whereas diabetes

leads to induction of several rat liver P450 forms,

including CYP2A1 (refs

[205],

[206]), CYP2B1

(refs

[207],

[208]), CYP2C7 (ref [206]), CYP2E1

(refs [209]-[212]), CYP4A2 (ref. [206]), and

CYP4A3 (ref [206]), it suppresses CYP2A2 (ref

[205]),

CYP2C11 (refs

[205], [207], [208],

[212]),

and CYP2C13 (ref. [205]). Changes in the levels

of some of these liver P450s (e.g., CYP2C11 and

CYP2E1) have been shown at the mRNA leveP^^,

208,213,214

^jj^

^j.g reversed by insulin replacement.

The profile of GH secretion in the diabetic male

rat is altered so as to resemble the pattern found in

the normal female

rat^^^.

The induction of CYP2A1

and CYP2C7 in diabetic male rats can therefore be

explained, at least in part, as a response to the more

continuous pattern of GH secretion, which stimu-

lates expression of these P450 forms^^'

3U 63,215

jj^

contrast, this pattern of GH secretion reduces

CYP2A2 and CYP2C13 levels because continuous

GH administration suppresses these two

P450s2^'28'

116

CYP2C11 expression is obligatorily

dependent on the intermittent male pattern of

plasma GH secretion^^. Therefore, the more contin-

uous secretion of GH in diabetic male

rats^^'*

would

be expected to suppress this P450. In the case of

CYP2B1,

GH pulse height is the suppressive

signal

^^^

and accordingly, the reduction in GH peak

concentration in diabetic male rats^^^ leads to

increases in CYP2B1 levels^^^' 208 A GH-inde-

pendent mechanism is likely to contribute to some

of the other effects of diabetes on liver P450

expression. GH, independent of

its

plasma profile,

is suppressive toward hepatic CYP2E1 (ref [62]),

but the levels of this P450 are substantially elevated

in both diabetic male and female rats^^^' 216,217

CYP2E1 induction in diabetes has been attributed

to increased plasma concentrations of ketone bod-

jgg2io,

218

Recently, a role of hypoinsulinemia and

hyperglucagonemia was proposed for the suppres-

sion of

CYP2C11

in diabetes, based on the finding

that treatment of cultured rat hepatocytes with

glucagon decreases CYP2C11 expression in a

cAMP-dependent manner and this decrease can be

reversed by insulin administration^ ^^.

5.3.2. Liver Cirrhosis

Gonadal abnormalities occur in liver cirrhosis.

Adult male rats fed a chronic choline-deficient

diet to induce cirrhosis have enhanced serum

estradiol concentrations^^^ and reduced testicular

weight^^^ and serum testosterone levels^^^. In

association with the perturbation in hormonal

status is a major decline in hepatic CYP2C11 con-

tent^^^, and this decline is not accompanied by

induction of steroid 5a-reductase activity^^^. The

suppression of hepatic CYP2C11 is also evident in

other models of liver cirrhosis, including bile duct

ligation^^^' ^^^ and carbon tetrachloride-induced

cirrhosis^^^' ^^^. Whether the alteration in serum

steroid hormone levels is directly or indirectly

responsible for the apparent demasculinization of

liver P450 remains to be determined.

5.4. Modulation by Ethanol and

Dietary Factors

Adult male rats administered ethanol by a total

enteral nutrition system have reduced hepatic

CYP2C11 and CYP3A2 levels, whereas their

CYP2A1 activity is unaltered^^^. The same ethanol

treatment alters the dynamics of plasma GH secre-

tion by decreasing the GH pulse amplitude and

increasing the GH pulse frequency. The increased

frequency of GH pulses thus explains the reduced

expression of CYP2C11 after chronic ethanol

intake because hepatocytes require a minimum "off

time"

in order to express the male-pattern of GH

secretion that stimulates CYP2CI1 expression^^. In

another study, chronic intragastric infusion of

ethanol-containing diets resulted in suppression of

CYP3A2 while substantially increasing the expres-

sion of CYP3A9 in adult male rats^^^

Dietary vitamin-A deficiency reduces hepatic

CYP2C11 (refs [227]-[229]) and CYP4A2

levels^^^,

while inducing steroid 5a-reductase

activity in adult male rats^"^^. These effects, which

are accompanied by a reduction in plasma testos-

terone levels, can be restored by the inclusion of

all-

trans-TQiinoic

acid in the diet^^^ or by exogenous

administration of androgen^^^'

^^K

Interestingly,

twice daily subcutaneous administration of GH,

which is known to induce CYP2C11 expression

in hypophysectomized male rats^^, is not effective

in restoring the levels of

CYP2C11

or CYP4A2 in

male rats fed a vitamin A-deficient diet^^^.

Dietary trace minerals can also alter liver

expression of sex-dependent P450 enzymes.

Prepubertal male rats fed a zinc-deficient diet