Irreversible perinatal imprinting of adult expression of the principal sex-dependent drug-metabolizing enzyme CYP2C11

Abstract

We proposed to determine whether, like other sexual dimorphisms, drug metabolism is permanently imprinted by perinatal hormones, resulting in its irreversible sex-dependent expression. We treated newborn male rats with monosodium glutamate (MSG), a total growth hormone (GH) blocker, and, using cultured hepatocytes, examined expression of adult CYP2C11, the predominant cytochrome-P450 expressed only in males, as well as the signal transduction pathway by which episodic GH solely regulates the isoform's expression. In addition, adolescent hypophysectomized (hypox) male rats served as controls in which GH was eliminated after the critical imprinting period. Whereas renaturalization of the masculine episodic GH profile restored normal male-like levels of CYP2C11, as well as CYP2C12, in hepatocytes from hypox rats, the cells derived from the MSG-treated rats were completely unresponsive. Moreover, GH exposure of hepatocytes from hypox rats resulted in normal induction, activation, nuclear translocation, and binding to the CYP2C11 promoter of the signal transducers mediating GH regulation of CYP2C11 expression, which dramatically contrasted with the complete unresponsiveness of the MSG-derived hepatocytes, also associated with hypermethylation of GH-response elements in the CYP2C11 promoter. Lastly, neonatal MSG treatment had no adverse effect on postnatal and adult testosterone levels. The results demonstrate that the sexually dimorphic expression of CYP2C11 is irreversibly imprinted shortly after birth by a hormone other than the customary testosterone, but likely by GH.—Das, R. K., Banerjee, S., Shapiro, B. H. Irreversible perinatal imprinting of adult expression of the principal sex-dependent drug-metabolizing enzyme CYP2C11.

Sexual dimorphisms are among the most commonly observed traits in nearly all living organisms. Although the roles of sexual dimorphisms in reproduction are almost intuitive, their expression appears to penetrate every organ system in the body. It is apparent why sexual reproduction would require sex differences in reproductive organs, accessory tissues, and sexual behaviors. The need, however, for sexual dimorphisms in what would appear to be nonreproductive functions, such as cardiovascular biology, brain chemistry, metabolism, and nonreproductive behaviors (1–4) is obscure. It should be prefaced that most of our in-depth knowledge about sexual dimorphisms are derived from laboratory and domesticated animals, as well as humans. From these studies, we know that the sex chromosomes (genetic factors) play a seminal, but limited, role in the development and expression of sexual differences (5). Ultimately, factors culminating from the expression of these genetic instructions are responsible for the actual phenotypic appearance of sex differences. In mammals, the fetal testes secrete testosterone which, along with its immediate metabolite dihydrotestosterone, directs the undifferentiated reproductive anlagen to organize into the male reproductive structures by a process known as somatic “imprinting” or “programming.” (In the case of all mammals, normal masculinization requires androgen imprinting, whereas feminization is basically the default genetic program.) The process can be considered “immediate,” as the masculinization of the reproductive anlagen begins as soon as the fetal testes secrete androgens (5–7). However, in the case, for example, of sexual behaviors, responsible centers in the brain are programmed during development, but the expressions of these behaviors are delayed, requiring the activation by pubertal and postpubertal hormones (1, 7). These so-called latent expressed sexual dimorphisms extend to nonreproductive behaviors (e.g., cognition, foraging behaviors, taste preferences) and numerous multiorgan enzyme systems (1, 2, 4–6, 8). Regardless of whether the imprinting is immediate or latently expressed, they share two immutable characteristics: the imprinting can only occur at a critical developmental period (i.e., perinatal), and the results are permanent and irreversible. Accordingly, the imprinted male, irrespective of adult hormone treatment, can no longer develop a vagina or a uterus or display normal feminine sexual behavior, nor can the differentiated female ever develop a penis or a prostate or display normal masculine sexual behavior.

Thus, the test as to whether a reproductive or nonreproductive function is sexually imprinted requires that the dimorphism be established during a limited developmental period, and the results must be permanent and irreversible. In this regard, the cytochrome P450 (CYP) monooxygenases are an ancient family of heme enzymes that catalyze a large variety of essential metabolites, which include the synthesis and the deactivation of adrenal, gonadal, and thyroid hormones, prostanoids, bile acids, and fatty acids, as well as the detoxification of innumerable drugs and environmental chemicals—hence, their ubiquitous presence throughout all species, where hundreds of isoforms have evolved to catalyze a countless number of substrates (9, 10). In the case of vertebrates, the CYPs are most prominent in the liver, where they can be expressed as dozens of different isoforms in each species. Those species examined, including trout, chickens, ferrets, dogs, goats, pigs, hamsters, cattle, mice, rats (cf. ref. 11), and not surprisingly humans (12, 13), all exhibit varying degrees of sex differences in adult drug metabolism. Studies extended to the molecular level have shown that the sexual dimorphisms in adult drug metabolism are due to the existence of multiple forms of sex-dependent hepatic CYPs whose sexually dimorphic expression is regulated by growth hormone (GH; refs.11, 14–16). In fact, the only endogenous factor known to maintain sexually dimorphic expression of hepatic CYPs is GH (11, 15). More specifically, it is not the amount of circulating GH per se, but rather the sexually dimorphic ultradian rhythms in plasma GH that regulate sex-dependent isoforms of CYP. Although there are some variations between species, in general, the adult male GH profile is referred to as “episodic” and is characterized by several daily bursts of hormone separated by lengthy undetectable or barely detectable GH concentrations. In contrast, the adult female GH profile is considered “continuous,” as there are far more secretory bursts, often at lower amplitudes, of the hormone separated by briefer interpulse periods, often containing measurable levels of GH. The fundamental regulatory difference between male and female GH profiles in all species examined is the much longer GH-devoid interpulse period observed in adult males (11, 15, 17–19) which, in addition to sex differences in CYP expression, is responsible, at least in part, for sex differences in growth rates and lean body mass; cardiovascular, bone, adipose, and muscle function; protein, carbohydrate, lipid and electrolyte metabolism (20–25); and expression levels of hepatic insulin-like growth factor 1 (IGF-1), IGF-binding protein, and GH-binding protein (21, 22, 26–29).

To examine the possibility that the sexually dimorphic expression of hepatic CYP isoforms are imprinted, the actual physiological circulating GH profiles of the opposite sex were renaturalized in male and female rats. Using adult GH-ablated rats, in addition to human and rat primary hepatocytes, it was found that replication of the normal masculine episodic GH profile could not stimulate female hepatocytes, either in vivo or in vitro, to express male-like levels of CYPs (16, 30–32), nor could restoration of the physiological feminine continuous GH profile induce in male hepatocytes, in vivo or in vitro, female-like levels of CYPs (16, 33–36), indicating an irreversible sexual dimorphism, and by implication the occurrence of developmental imprinting. This incomplete sex-reversal of CYP isoforms is reminiscent of another latently expressed imprinted sexual dimorphism; i.e., sexual behavior, which can be only partially reversed by administering the sex hormones of the opposite sex (1, 7, 37).

In the present study, we demonstrate for the first time that similar to other sexual dimorphisms, sex-dependent CYP expression in the rat, and by extension, all mammals, is irreversibly imprinted, but unlike other investigated sexual dimorphisms, not by perinatal testosterone.

MATERIALS AND METHODS

Animals

Rats were housed in the University of Pennsylvania Laboratory Animal Resources facility under the supervision of certified laboratory animal medicine veterinarians and were treated according to a research protocol approved by the Institutional Animal Care and Use Committee of the University. Housing conditions, as well as breeding and treatment protocols, were reported previously (38, 39). Basically, newborn male Sprague-Dawley rats [Crl:CD(SD)BR] were treated with monosodium glutamate (MSG), 4 mg/g body weight (BW) (Sigma Chemical, St. Louis, MO, USA) on alternate days, starting within 24 h of birth for a total of 5 s.c. injections. Controls received an equivalent amount of 1.97 M NaCl diluent (12 μl/g BW) or no treatment at all. Additional male rats [Crl:CD(SD)BR] were either hypophysectomized (hypox) or sham hypox by the vendor (Charles River Laboratories, Wilmington, MA, USA) at 8 wk of age and observed in our facility for 6–7 wk to allow sufficient time for any residual pituitary tissue to regenerate. The effectiveness of the surgery was verified by the lack of weight gain over this period and the absence of pituitaries or fragments at necropsy shortly after euthanizing the rats. The Lee index and hexobarbital sleep times were determined as described previously (38, 39).

Catheter implantation, serial blood collection, and GH determination

Repetitive blood samples (40 μl) were obtained at 15-min intervals from unrestrained, unstressed, and completely conscious hypox, neonatally MSG-treated, and NaCl-treated adult rats outfitted with our mobile catheterization apparatus (40) for 8 continuous hours. Circulating plasma GH profiles were determined by a sensitive sandwich ELISA method according to Steyn et al. (41) with modifications by us (42).

Primary hepatocyte culture

Preparation of hepatocytes from control and neonatally MSG-treated adult rats, as well as long-term hypox rats, were performed with minor modifications (30, 31) by in situ perfusion of collagenase through the portal vein of anesthetized rats following standard protocol (43). The viability of the initial cell suspension of hepatocytes was typically between 80 and 90% (with trypan blue). The hepatocytes were plated at a density of 3 × 10⁶ viable cells per well in 6-well plates previously coated with Matrigel. Medium and culture conditions were reported previously (30, 31).

Hormonal conditions

To replicate the masculine episodic GH profile, hepatocytes were exposed to recombinant rat GH (0.2 ng/ml; National Hormone and Peptide Program, Torrance, CA, USA) for 30 min, followed by 2 careful washings with GH-free medium that remained in the wells for 11.5 h, at which time the cells were again washed and exposed to the next 30-min pulse of GH (30, 31). On the 6th day, cells were harvested 30 min after the addition of the last GH pulse. To replicate the feminine continuous GH profile, hepatocytes were constantly exposed to a 2 ng/ml concentration of the rat GH for 12 h, after which time the cells were washed and exposed for another continuous 12 h of the hormone (34, 35). Cells were harvested on the 6th day, 30 min after the final GH treatment.

Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR)

CYP2C11 and CYP2C12 gene expression was determined by qRT-PCR using the TaqMan assay (Rn01502203_m1 and Rn00755856_m1; Life Technologies, Grand Island, NY, USA) and proteasome (prosome, macropain) 26S subunit (PSMC4; Rn00821599_g1) as the housekeeping gene on an Applied Biosystems step-one plus quantitative PCR instrument (Life Technologies) as per the usual manufacturer's recommended protocol. Total RNA from freshly isolated hepatocytes, as well as hepatocyte cultures, were isolated using TRIzol reagent (Life Technologies) purified with the Qiagen RNeasy mini kit (Qiagen, Valencia, CA, USA) and treated with DNase I in order to remove any trace of genomic DNA using the RNase-free DNase set (Qiagen), according to the manufacturer's protocol. RNA concentrations and purity were determined by UV spectrophotometry (A_260/280>1.8 and A_260/230>1.7), and integrity was verified by the intensities of 28S and 18S rRNA bands on a denaturing agarose gel visualized on a FluorChem IS-8800 Imager (Alpha Innotech, San Leandro, CA, USA). cDNA synthesis was completed using the high-capacity RNA-to-cDNA kit (Life Technologies), as per instructions with appropriate no-reverse transcription (−RT) and nontemplate controls.

PCR array

Changes in the expression of genes involved in the Janus kinase(JAK)/signal transducers and activators of transcription (STAT) signaling pathway in control and MSG-treated rat hepatocytes were determined using a real-time PCR array (Rat RT² Profiler PCR Array; SA Biosciences; Qiagen). Isolation, concentration, and purity of total RNA from rat liver was performed as mentioned above. The first cDNA strand was synthesized using an RT² First-Strand Kit (SA Biosciences). The cDNA was then applied to the RT² Profiler PCR Array for the JAK/STAT pathway (PARN-039C; SA Biosciences), according to the manufacturer's guidelines. The qRT-PCR reactions were run on the Step-one plus qPCR instrument (Life Technologies), and the data were analyzed using the online RT² Profiler PCR Array data analysis (SA Biosciences) program as per the manufacturer's instructions.

Western blot and immunoprecipitation

Whole-cell lysates and nuclear fractions were extracted from freshly isolated hepatocytes, as well as cultured primary hepatocytes, 30 min after the last GH pulse, and the protein concentrations of the cell lysates were measured by using the Bio-Rad protein assay reagent (Bio-Rad Laboratories, Hercules, CA, USA). Whole-cell lysate protein was electrophoresed and electroblotted onto nitrocellulose membranes for immunoblotting. Accordingly, the blots were probed with antibodies against CYP2C11 (Detroit R&D, Detroit, MI, USA), anti-rat CYP2C12 (a gift from Dr. Marika Rönnholm, Huddinge University Hospital, Huddinge, Sweden), anti-rat GH receptor (GHR), and anti-suppressors of cytokine signaling 2 (SOCS2; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Signals were normalized to the expression of β-actin (Sigma Chemical). Nuclear fractions were immunoprecipitated with STAT5b antibodies (Santa Cruz Biotechnology), as described previously (31). Next, the immunoprecipitates were probed with antiphosphotyrosine (EMD Millipore, Billerica, MA, USA). The protein signals were scanned, and the densitometric units were obtained as integrated density values quantitated by using a FluorChem IS-8800 Imager (Alpha Innotech) software supplied with the gel documentation system.

Chromatin immunoprecipitation (ChIP) assays

A ChIP assay was performed on cultured hepatocytes 30 min after the last episodic hormone pulse on the 6th day in culture as described previously (31, 36) with slight modifications. Briefly, for the cross-linking reaction, hepatocytes were treated with 1% formaldehyde for 10 min at room temperature, and the reaction was stopped by adding glycine to a final concentration of 0.125 M. The cells were then harvested and washed 3 times with ice-cold Dulbecco's PBS buffer containing 5 mM EDTA. The nuclei were subsequently isolated and lysed. The lysate was sonicated to generate DNA fragments with an average length of 100–1000 bp. After removal of cell debris by centrifugation, the chromatin concentration was measured, and ∼10% of the chromatin was kept as an input, and the rest of the chromatin was diluted 3-fold. Equal concentrations of chromatin were precleared with protein A agarose beads in the presence of 1 mg/ml BSA and 2 μg of sonicated salmon sperm DNA to reduce the nonspecific background. After removal of beads by centrifugation, 2 μg of STAT5b-specific antibody (Santa Cruz Biotechnology) was added and kept at 4°C overnight on a rotary platform. The immune complexes were collected by centrifugation after adding protein A agarose beads and kept at 4°C for 1 h. The immunoprecipitates were washed sequentially and eluted. To reverse the formaldehyde cross-linking, eluates were heated at 65°C for 6 h. DNA fragments were purified by QIAquick Spin kit (Qiagen, Germantown, MD, USA). DNAs were analyzed by semiquantitative PCR as previously reported (31) by specific and nonspecific primers of the STAT5b binding region of the CYP2C11 promoter.

DNA isolation and methylation-specific PCR

DNA was extracted from freshly isolated hepatocytes of adult rats neonatally treated with either MSG or its diluent using ZR Genomic DNA-Tissue MiniPrep (Zymo Research Corp., Irvine, CA, USA), according to the kit directions. DNA (400 ng) was then used for sodium bisulfate modification using the EZ DNA methylation kit (Zymo Research), according to the manufacturer's protocol. Similarly, a universal methylated human DNA standard (Zymo Research) was also modified to assess the efficiency of the bisulfite-mediated conversion of DNA. A ∼2.6-kb sequence upstream from the transcription start site of the CYP2C11 gene (X79081.2) was analyzed with two sets of primers [methylation-specific primers (MSP 1 and 2), as well as unmethylation-specific primers (USP 1 and 2)] designed using MethPrimer (44) to detect the methylation status of CpG sites in the CYP2C11 gene. The primers were as follows: primer set 1 (MSP 1): methylated forward (MF 1) GACGTTGGTTAGTTAGTAGATACGT, methylated reverse (MR 1) TACTAAAAAATTTAACCTTCACGAT, unmethylated forward (UF 1) GATGTTGGTTAGTTAGTAGATATGT, unmethylated reverse (UR 1) TACTAAAAAATTTAACCTTCACAAT; primer set 2 (MSP 2): methylated forward (MF 2) ATTGAGGAAGATTCGGAAATTTC, methylated reverse (MR 2) TTCAAATAATAATAACTCATACCGAC, unmethylated forward (UF 2) AGATTGAGGAAGATTTGGAAATTTT, unmethylated reverse (UR 2) TCAAATAATAATAACTCATACCAAC.

Modified DNA was used as a template for MSP. All PCRs were performed using ZymoTaq PreMix (Zymo Research) and were hot started at 95°C for 10 min (initial denaturation) followed by 37 cycles of denaturation for 30 s at 95°C, annealing for 30 s at 57°C, extension for 30 s at 72°C, and final extension for 7 min at 72°C for both methylated and unmethylated primer set 1. The procedure for both methylated and unmethylated primer set 2 was the same as for primer set 1, with the exception that the reaction was carried out for 40 cycles, and the annealing temperature was 56°C for 30 s. Similarly to primer set 2, the PCR was also performed for bisulfite-converted universal methylated human DNA using the hMLH1 control primer set (Zymo Research) serving as a positive control. The negative control for the PCR reaction was performed without DNA but using water. PCR products were electrophoresed on 2% agarose gels, stained with ethidium bromide, and directly visualized under UV illumination. The density of each band was calculated by using a FluorChem IS-8800 Imager (Alpha Innotech) with software supplied with the gel documentation system. The relative methylation level in each sample was determined using the following formula: MSP ratio (%) = MSP × 100/(MSP + USP). The data were confirmed by visual examination and scanning. A MSP ratio of 5.0% was used as a cutoff value to assess whether samples were MSP positive or negative (45).

Testosterone measurement

Plasma testosterone concentrations in postnatal and adult MSG-treated and control rats were measured by a sensitive competitive ELISA method (R&D Systems, Minneapolis, MN, USA), according to the manufacturer's recommended protocol.

Statistics

All data were subjected to ANOVA. Significant differences were determined with t statistics and the Bonferroni procedure for multiple comparisons.

RESULTS

Control treatments

The study includes 3 control treatment groups: neonatal rats treated with MSG vehicle, neonatal rats that received no treatment, and 8-wk-old sham-hypox rats. Since there were no significant differences in the measured parameters between the 3 groups, results from the former, i.e., MSG-vehicle treated, were chosen as controls in all figures.

Lee index

The Lee index is a measure of obesity; the higher the value, the greater the obesity. At 8 wk of age, the Lee index of the neonatally MSG-treated rats became significantly greater than controls, with this trend continuing through the conclusion of the study at 22 wk of age (Fig. 1). While the cause of obesity can be very complex, including genetic, hormonal, metabolic, dietary, and environmental factors; GH deficiency can be a contributory factor (46).

Figure 1.

Postweaning maturational Lee index profile of male rats neonatally treated with MSG. MSG-treated pups received 4 mg/g BW on postnatal days 1, 3, 5, 7 and 9. Control males were injected with equivalent doses of vehicle. Results are presented as the means ± sd of ≥10 rats/data point. *P < 0.01 vs. controls.

Circulating GH profiles

Plasma GH profiles are presented as schematic representations of the actual circulating profiles (Fig. 2A). The control rats secreted the typical masculine profile (26, 47) characterized by episodic pulses containing amplitudes of ∼200 ng/ml with durations of ∼1 h, separated by interpulse periods of 2 to 3 h devoid of any measurable hormone. In contrast, none of the adult neonatally MSG-treated rats had detectable plasma GH concentrations (>0.125 ng/ml) during the 8 continuous hours of serial sampling. Here, we added an additional control group: male hypox rats devoid of GH from adolescence, the transitional period between puberty and adulthood when both the masculine GH profiles and hepatic CYP expression levels are apparent, but not fully mature (47, 48). The circulating GH profiles, or lack thereof, in the hypox rats were identical to those of MSG-treated rats (Fig. 2A).

Figure 2.

A) Plasma GH profiles in individual, representative adult male rats that were either neonatally administered MSG or vehicle or hypox in adolescentce. Neonatally treated pups were injected with either MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7, and 9. Plasma levels of circulating GH were obtained from individual undisturbed catheterized rats for 8 continuous hours at 15-min intervals. Similar results were obtained from 4 to 5 additional animals in each treatment group. B) Hexobarbital-induced sleeping times in adult male rats that were either neonatally administered MSG or vehicle or hypox in adolescence. Sleeping times were determined following an i.p. injection of hexobarbital (150 mg/kg BW). Results are presented as means ± sd of ≥10 rats/group. *P < 0.01 vs. MSG-treated rats.

Sleeping times

Hexobarbital-induced sleeptime is an in vivo measure of drug metabolism competency (14, 49). When given the same dose of hexobarbital per kilogram BW, adult rats neonatally treated with MSG slept twice as long as control rats, indicating a deficiency in the CYP isoforms contributing to hexobarbital metabolism (Fig. 2B). The hypox rats with surgically induced GH depletion also exhibited prolonged hexobarbital-induced sleeptimes that were even longer than those of the MSG-treated rats.

Principal sex-dependent forms of CYP

CYP2C11 is the principal male-dependent form, comprising >50% of the total CYPs in male rat liver (50). In contrast, CYP2C12 is the principal female-dependent form, compromising >40% of the total CYPs in female rat liver (51). With the exception of measurements using highly sensitive detection procedures, e.g., qRT-PCR, CYP2C11 is basically undetectable in female rat liver, and CYP2C12 is similarly undetectable in male rat liver, explaining their designation as sex-specific (14, 15, 18, 19). In this regard, the levels of CYP2C12 mRNA in the freshly isolated hepatocytes of males appear to be considerable, but only because they are based on the 100% arbitrary value of the controls, which actually represent <5% of the amount of CYP2C11 mRNA (Fig. 3).

Figure 3.

Relative expression levels (mRNA and protein) of the two paramount sex-specific CYP isoforms in freshly isolated hepatocytes from adult male rats that were either neonatally administered MSG or vehicle or hypox in adolescence. Neonatally treated pups were injected with either MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7 and 9. Results are presented as a percentage of mRNA or protein in hepatocytes from control rats arbitrarily designated 100%. Values are presented as means ± sd of ≥7 rats/data point. ND, not detected. *P < 0.01 vs. control rats.

As freshly isolated hepatocytes, CYP levels in vitro reflect in vivo concentrations. Accordingly, expression of GH-dependent, male-specific CYP2C11 was profoundly and equally reduced in hepatocytes from both the MSG-treated and hypox adult male rats (Fig. 3). Noting that the comparative levels of CYP2C12 are only a small fraction of that of CYP2C11, none of the treatments altered the characteristic masculine pattern in which female-specific CYP2C12 is suppressed. These results were not unexpected, since the control males secrete episodic GH (Fig. 2A) required for CYP2C11 expression (15, 17, 19), while neither the MSG-treated nor hypox males secrete any measurable GH (Fig. 2A). Moreover, none of the rats secrete GH in the continuous feminine profile, explaining the basic absence of continuous GH-dependent CYP2C12 (11, 15, 18) in all treatment groups. Next, we determined the plasticity of the CYP expression.

Restoration of CYP expression

After 6 d of exposure to the male-like episodic GH profile, CYP2C11 expression levels in primary hepatocytes derived from control rats more than doubled above baseline concentrations (Fig. 4A). Exposure to the female-like continuous GH profile was either ineffective or suppressive. The response of hepatocyte CYP2C11 from the hypox males, GH deficient for their entire adult life, was nearly identical to that of hepatocytes from control rats. In contrast, the inability of both the masculine and feminine GH profiles to induce CYP2C11 expression in hepatocytes from MSG-treated males indicated an irreversible repression of the enzyme.

Figure 4.

GH-responsive male-specific CYP2C11 (A) and female-specific CYP2C12 (B) expression in cultured primary hepatocytes derived from adult male rats that were either neonatally administered MSG or vehicle or hypox in adolescence. Neonatally treated pups were injected with either MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7 and 9. Hepatocytes from the same animal were exposed to either the episodic male-like GH profile or the hormone's vehicle, or the continuous female-like GH profile or vehicle for 6 d. Results are presented as a percentage of mRNA or protein of hepatocytes from control, MSG-treated, or hypox rats exposed to vehicle administered in an episodic male-like profile, each value arbitrarily designated 100%. Values are expressed as means ± sd of ≥6 rats/data point. *P < 0.01 vs. hepatocytes exposed to the episodic male-like vehicle profile from rats of the same preadult treatment group.

Whereas CYP2C12 is female-specific, it can be induced in male hepatocytes, albeit at much lower levels than observed in female hepatocytes (34, 35). Indeed, hepatocytes obtained from both control and hypox males exhibited a modest, but significant induction in CYP2C12 levels when exposed to the female-like continuous GH profile, whereas hepatocytes from MSG-treated males were completely unresponsive (Fig. 4B). Exposure to the male-like episodic GH profile was a more effective inducer of CYP2C12 expression than the feminine profile in control and hypox male-derived hepatocytes, whereas hepatocytes from the MSG-treated rats were still completely unresponsive. It is worth noting again that, where observed, significant changes in CYP2C12 induction levels were statistically demonstrated (P<0.01), but the actual amounts were nominal when compared to the levels of CYP2C11 expression or to that of expression levels reported in hepatocytes from female rats (34, 35).

Signal transducers mediating GH regulation of CYP2C11

GH regulation of CYP2C11 requires the activation of the JAK/STAT signaling pathway (52). Accordingly, we initially performed a rat JAK/STAT signaling pathway PCR array in order to identify potential factors that could be involved in the refractory response to GH of the MSG-derived hepatocytes. We compared freshly isolated (i.e., preplated) hepatocytes from control and MSG-treated rats (data not presented). Realizing that the JAK/STAT signaling pathway mediates the activities of far more ligands than GH, we selected only those affected mRNAs known to be involved in GH induction of CYP2C11 for further investigation. Analysis of the array findings indicated that two factors required for GH induction of CYP2C11, GHR, and SOCS2 were down-regulated in the preplated hepatocytes from the MSG-treated rats. Accordingly, we cultured primary hepatocytes from adult control, neonatally MSG-treated, and adolescently hypox male rats in the presence of episodic GH or episodic vehicle alone (Fig. 5). After 6 d of treatment, the cells were harvested 30 min following the last pulse of GH. Whereas cellular concentrations of GHR and SOCS2 proteins from the control and hypox male hepatocytes were highly responsive to the masculine GH profile (∼2-fold increase), hepatocytes from the MSG-treated rats were completely unresponsive to the episodic GH profile.

Figure 5.

Relative levels of GHR and SOCS2 proteins in cultured primary hepatocytes derived from adult male rats that were either neonatally administered MSG or vehicle or hypox in adolescence. Neonatally treated pups were injected with either MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7, and 9. Hepatocytes from the same animal were exposed to either the episodic male-like GH profile or the episodic profile containing GH vehicle alone for 6 d. Results are presented as a percentage GHR or SOCS2 of hepatocytes from control-treated rats exposed to the episodic vehicle profile, arbitrarily designated 100%. Values are expressed as means ± sd of ≥6 rats/data point. *P < 0.01 vs. hepatocytes exposed to the episodic vehicle profile from rats of the same preadult treatment group.

STAT5b

The terminal transducer in the signaling transduction pathway mediating episodic GH induction of CYP2C11 is STAT5b. This transcription factor must be activated (i.e., phosphorylated), homodimerized, and translocated to the nucleus, where it binds to a promoter site initiating transcription of the CYP2C11 gene (52). We cultured primary hepatocytes from control, MSG-treated, and hypox male rats in the presence of the masculine episodic GH profile or episodic GH vehicle alone. The cells were harvested 30 min after the last GH pulse. In agreement with the transduction requirements for GH-induced transcription of CYP2C11, the concentrations of activated nuclear phospho STAT5b (pSTAT5b) increased 3- to 4-fold in GH-exposed hepatocytes from control and hypox rats, whereas the same hormone treatment produced a 50% decline in nuclear levels of the pSTAT5b in hepatocytes from MSG-treated rats (Fig. 6A). Using identically treated hepatocytes, the binding of pSTAT5b to the CYP2C11 promoter was evaluated by ChIP assays. In agreement with the expected sequence of signaling events, binding of pSTAT5b to the promoter occurred in hepatocytes from control and hypox rats (Fig. 6B). In contrast, none of episodic GH-exposed hepatocytes from the MSG-treated rats exhibited detectable binding of the activated transcription factor to the CYP2C11 promoter. PCR amplification of a negative control using primers flanking the CYP2C11 promoter at a genomic sequence not including the STAT5b binding site was undetectable, demonstrating no measurable nonspecific binding (not presented).

Figure 6.

A) Nuclear levels of pSTAT5b in cultured primary hepatocytes derived from male rats that were either neonatally administered MSG or vehicle or hypox in adolescence. Neonatally treated pups were injected either with MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7, and 9. Hepatocytes from the same animal were exposed to either the episodic male-like GH profile or the episodic profile containing GH vehicle alone for 6 d. Results are presented as a percentage of nuclear pSTAT5b of hepatocytes from control-treated rats exposed to the episodic vehicle profile, arbitrarily designated 100%. Values are expressed as means ± sd of ≥6 rats/data point. *P < 0.01, ^†P < 0.05 vs. hepatocytes exposed to the episodic vehicle profile from rats of the same preadult treatment group. B) Representative ChIP blots demonstrating episodic GH regulation of pSTAT5b binding to the CYP2C11 promoter in hepatocytes derived from adult male rats neonatally treated with either MSG or vehicle (control) or hypox in adolescence. As described above, the cells were exposed to either the episodic male-like GH profile (GH) or the vehicle-containing profile (vehicle). ChIP assays were performed on 5 rats of each treatment group with the same results as that presented in the representative blots. C) Methylation status of 2 upstream GH-activated regulatory elements of the CYP2C11 gene (primer set 1, nt −951 to −1106; primer set 2, nt −725 to −905) examined in freshly isolated hepatocytes from adult male rats neonatally treated with either MSG or diluent. Figure is representative of 7 animals/treatment group. MK, marker; +VE, chemically methylated DNA positive control; −M, methylated primer without DNA negative control; −U, unmethylated primer without DNA negative control; U, unmethylation-specific primer; M, methylation specific primer.

Methylation status of the CYP2C11 enhancer

The methylation status of 2 upstream GH-activated regulatory elements of the CYP2C11 gene was examined in freshly isolated hepatocytes from adult male rats neonatally treated with either MSG or diluent (Fig. 6C). The region located between nt −951 to −1106 (primer set 1) contains binding sites for hepatocyte nuclear factor 6, STAT6, and, of particular relevance to the present study, STAT5b (MatInspector; Genomatix, Munich, Germany; ref. 53). The methylation frequency of this region of the CYP2C11 gene was 25% greater (P<0.02) in the MSG-treated rats (69.8±8.3%) than controls (56.3±7.6%). We also measured the methylation status of the upstream region located between nt −725 to −905 (primer set 2) containing GH-responsive binding sites that include nuclear factor κ, STAT3, and peroxisome proliferator-activated receptor (53). The methylation frequency for this region of the CYP2C11 gene was dramatically different (P<0.01) between the two treatments. The methylation frequency for the hepatic gene in the MSG-treated rats was 47.4 ± 5.2% compared to 14.1 ± 2.9% for the controls.

Neonatal testosterone

The latent suppression of adult hepatic CYP expression in the MSG-treated rats is not due to the direct action of the amino acid on the developing liver (54). Rather, it is a result of disruption of the developing hypothalamic-pituitary axis and presumably the subsequent depletion, distinctively of the GH-dependent axis (55, 56). However, all known mammalian sexual dimorphisms, including those expressed latently (e.g., sexual behavior, gonadotropin secretory profiles, and taste preferences), are permanently imprinted or programmed by perinatal testosterone or one of its metabolites (1, 5, 7, 37). Accordingly, we measured plasma testosterone concentrations during the critical developmental period of hypothalamic-pituitary-hepatic differentiation when the pups were injected with MSG (Fig. 7). Plasma testosterone concentrations during the first two weeks of life were similar in MSG-treated and control pups, and in agreement with levels previously reported (57, 58). In fact, testosterone levels in adult, neonatally MSG-treated rats were somewhat higher than control rats of the same age, though both reflecting the normal fold increase observed in adulthood (59, 60) and both within the reported adult range (61).

Figure 7.

Postnatal plasma testosterone concentrations in male rats neonatally administered either MSG or vehicle. Pups were injected with either MSG (4 mg/g BW) or equivalent doses of vehicle (control) on postnatal days 1, 3, 5, 7, and 9. Values are expressed as means ± sd of ≥7 rats/data point. *P < 0.05 vs. control rats of the same age.

DISCUSSION

Hormonal imprinting refers to a biological process in which the target tissue becomes responsive to the hormone. During the initial exposure, the hormone irreversibly reprograms the development of the affected tissue so as to permanently alter some functional aspect normally responsive to the hormone (62). Moreover, the tissue is programmable for only a brief developmental period, after which time the tissue becomes permanently unresponsive to imprinting (1, 5–7). Imprinting alone, however, is generally insufficient to ensure expression of the reprogrammed functions. Our findings from rats and humans imply that expression of adult patterns of CYPs require both imprinting and activation, which is a reversible process, but required to express the imprinted function. For example, perinatal testosterone or its metabolites are required to permanently “wire” or imprint the male brain to exclusively express masculine sexual behavior. However, the brain has to be stimulated or activated in adulthood by the same hormones to elicit the imprinted male sexual behavior. Both imprinting and activation are required for normal masculine sexual behavior (1, 7, 37), which explains why females, not hormonally imprinted, are unable, regardless of adult testosterone treatment, to exhibit normal masculine sexual behavior (7, 37). In this regard, we previously reported that adult male rats and humans cannot be induced (regardless of treatment) to express the normal female profile of hepatic CYPs (16, 33–36), nor can adult female rats or humans be induced to fully express the masculine CYP profile (16, 30–32). If CYP enzymes were not imprinted, then irrespective of sex, the same adult treatment should produce the same CYP expression levels in males and females. Since this is not the case, we have concluded that the sex differences in adult expression profiles of CYPs are permanent and irreversible, one of the two conditions needed to demonstrate imprinting. The other requirement for imprinting is that it occurs during a limited developmental period. Exposure of the responsive tissue to the hormone any time before this narrow developmental window or thereafter would be ineffective.

Since adult tissues responsive to androgens are developmentally imprinted by androgens (1, 5, 7), it seemed reasonable to assume that the sex-dependent expression of hepatic CYPs by GH is accordingly imprinted by GH. In the absence of any completely effective antagonist of GH action or synthesis, we treated neonates with MSG, a total inhibitor of GH secretion (38, 39). To investigate the specific requirement for perinatal GH in the imprinting process, we included a control group, in which the hormone was eliminated after the critical developmental period, but before adulthood, i.e., our hypox rats. In this regard, we found a complete absence of circulating GH in both neonatally MSG-treated and hypox adult male rats. Moreover, both GH-null groups exhibited dramatically prolonged hexobarbital-induced sleep times indicative of suppressed CYP expression: CYP2C11, in particular (38, 39). Finally, freshly isolated hepatocytes from both the MSG-treated and hypox adult rats expressed only a small fraction of CYP2C11 (mRNA and protein) as compared to GH-secreting controls. When the hepatocytes were cultured for 6 d in the absence of GH, cells from all groups expressed low baseline concentrations of CYP2C11, characteristic of in vivo GH depletion. Restoration of the masculine episodic-like GH profile more than doubled male-specific CYP2C11 expression levels in controls, as it did in cultured hepatocytes derived from the hypox rats, in which neonatal plasma hormone levels were normal. In contrast, CYP2C11 mRNA and protein in hepatocytes from the MSG-treated rats were completely unresponsive to the inductive effects of the masculine GH profile, indicating an irreversible suppression (i.e., demasculinization) of the isoform. Our observation that restoration of the feminine continuous-like GH profile in culture had a suppressive effect on CYP2C11 suggests a similar degree of defeminization in all groups. Whereas female-specific CYP2C12 expression in hepatocytes from female rats exhibits a robust response to the feminine GH profile, and a minimal response to the masculine GH profile (34, 35), the male hepatocytes from the control and hypox rats exhibited the opposite response (albeit at only a fraction of the actual concentrations induced in female cells) in which the episodic profile was more inductive than the continuous GH profile characteristic of masculine cells that are now programmed to be refractory to the feminine continuous GH profile (33–35). CYP2C12 levels in hepatocytes derived from MSG-treated rats were unresponsive to both sexually dimorphic GH profiles, again suggesting a demasculinization of GH-responsive CYPs in the absence of a concomitant feminization.

What, then, are the defects in the GH-inductive response of CYP2C11 in adult rats neonatally treated with MSG? That is, why cannot the hepatocytes respond to GH? In this regard, GH signaling in liver by the episodic profile (in contrast to the continuous profile) is initiated by hormone binding to and the resulting activation of GHR on the surface of the target cells. This allows for the recruitment and/or activation of two molecules of JAK2, which then cross-phosphorylate each other, as well as phosphorylating the receptor on key tyrosine residues. STAT5b, a latent transcription factor, binds to these phosphorylated receptor docking sites, is, in turn, phosphorylated, and homodimerizes and translocates to the nucleus, where it binds to promoter sites initiating transcription of GH-regulated genes. An important negative regulatory mechanism of GH signaling, the SOCS family of inhibitory proteins, which, depending on the protein, inhibit GH signaling by inactivating JAK2, by competing with STAT5b for common tyrosine-binding sites on the intracellular tail of the GHR, and/or by a proteasome-dependent mechanism that results in the degradation of the (GH-GHR-JAK2)·SOCS complex (52, 63–67). At least 3 components in this signal transduction pathway were permanently altered by neonatal MSG. In contrast to hepatocytes from control and hypox rats, where episodic GH increased total cellular GHR and SOCS2, as well as nuclear pSTAT5b, by 2- to 4-fold, these transducers in hepatocytes from the MSG-treated rats were completely unresponsive to the masculine GH profile. Not only did developmental exposure to MSG permanently suppress STAT5b activation (i.e., phosphorylation) and nuclear translocation, but it completely blocked the binding of the transcription factor to the CYP2C11 promoter, explaining the inability of the masculine GH profile to induce CYP2C11 expression. [The ability of hepatocytes from all treatment groups, both in culture (30, 31) and in situ (17, 19, 32), to express minimal levels of CYP2C11 in the absence of GH may be explained by the presence of a GH·GHR·JAK2·STAT5b-independent mechanism responsible for baseline expression levels (68).] An additional factor that may have contributed to the MSG-induced developmental defect in CYP2C11 expression involves possible epigenetic alterations in the CYP2C11 gene. Neonatal exposure to MSG increased the methylation frequency of two regions of the CYP2C11 promoter containing several binding sites for GH-activated/regulated transcription factors, including STAT5b. Whether any of these binding sites were actually silenced, though possible, is speculative and requires further study. Clearly, then, neonatal exposure to MSG irreversibly suppresses GH activation of CYP2C11-dependent signal transduction at multiple steps. Whether the multiple defects in the signaling pathway are all independently induced or rather the result of an initial upstream defect (i.e., GHR) cascading dysfunctions down the signaling pathway is undetermined.

Although the hypox rat exhibits multiple hormone deficiencies, including adrenal, gonadal, and thyroid hormones, the ability of GH to restore normal levels of CYP2C11, as well as other sex-dependent isoforms, both in vivo (18, 19) and in vitro (30, 34), demonstrates that GH is the sole regulator of these enzymes. Similarly, neonatal treatment with MSG totally blocks GH secretion, and some studies have reported reduced plasma concentrations of other pituitary hormones, as well as gonadal steroids, in adulthood (56, 69, 70). The possibility then exists that a neonatal hormone other than GH is responsible for imprinting hepatic CYP2C11. The obvious candidate, testosterone, is the only hormone (or its metabolites) known to imprint a myriad of sexual structures and functions (1, 5–7). In fact, before the full realization that GH regulates adult CYPs, perinatal testosterone was reported to be responsible for the masculinization (i.e., imprinting) of hepatic sex-dependent CYP isoforms, as neonatal orchidectomy appeared to demasculinize drug metabolism. However, the permanence or irreversibility of the imprinting was either never tested by administering testosterone to neonatal castrates as adults (71, 72) or, in fact, was found to be completely restored when treated with the androgen as adults (73–76). (Androgens, including testosterone, have no direct effect on hepatic CYP expression, but rather can act at the hypothalamic-pituitary axis to regulate the sexually dimorphic secretion of GH, which, in turn, regulates sex-dependent CYPs; refs. 26, 77, 78.) If the effects of neonatal castration on CYP expression can be completely corrected (i.e., reversible), by definition and, despite the researchers' conclusions, androgens do not imprint the sexually dimorphic expression of CYP isoforms. In addition, our observation that neonatal MSG had no inhibitory effects on circulating testosterone levels during the first 15 d of life, inclusive of the critical period of hypothalamic-pituitary-hepatic axis differentiation, is further support to our contention that perinatal androgens are not imprinters of CYP sexual dimorphisms. This, of course, begs the question “if not androgen, then what?” Whereas GH seems the most reasonable choice, as the imprinting hormone is invariably the same hormone that activates the adult function, and neonatal MSG does completely eliminate circulating GH by disrupting the differentiating hypothalamic-pituitary axis and not by directly interfering with liver development (54), there still remains the possibility that other MSG-depressed hormones may contribute to the sexual imprinting of the hepatic CYP family. Regardless, our findings do demonstrate that the sexually dimorphic expression of hepatic CYPs, or at least the predominant male-specific CYP2C11, is irreversibly imprinted shortly after birth by a hormone other than the expected testosterone, but likely GH that permanently programs the responsiveness of CYP-dependent signal transduction to GH activation. Since there are a myriad of sex differences, including growth rates and lean body mass; cardiovascular, bone, adipose, and muscle physiology; protein, carbohydrate, lipid, and electrolyte metabolism all regulated, at least in part, by sex differences in circulating GH profiles (20–29), it may be that these functions are similarly imprinted by neonatal GH and accordingly, irrespective of any treatment, remain permanently sexually dimorphic.