Maternal obesogenic diet during pregnancy and its impact on fetal hepatic function in baboons

Abstract

Objective:

Maternal obesity (MO) increases the risk of later life liver disease in offspring, especially in males. This may be due to impaired cytochrome P450 (CYP) enzyme activity driven by an altered maternal-fetal hormonal milieu. MO increases fetal cortisol concentrations that may increase CYP activity; however, glucocorticoid receptor (GR)-mediated signalling can be modulated by alternative GR isoform expression. We hypothesised MO induces sex-specific changes to GR isoform expression and localisation that contribute to reduced hepatic CYP activity.

Methods:

Non-pregnant nulliparous female baboons were assigned to either ad libitum control or high fat, high energy diet (HF-HED) 9 months pre-pregnancy. At 165d gestation (term=180d) fetal liver samples were collected (n=6/sex/group). CYP activity was quantified using functional assays; GR was measured using qRT-PCR and Western blot.

Results:

CYP3A activity was reduced in HF-HED, whereas CYP2B6 activity was reduced in HF-HED males only. Total GR was increased in HF-HED. Relative nuclear expression of the antagonistic GR isoform, GRβ, was increased in HF-HED males only.

Conclusions:

Reduced CYP activity in HF-HED males may, in part, be driven by dampened hepatic-specific glucocorticoid signalling via altered GR isoform expression. These findings highlight targetable mechanisms that may reduce sex-specific disease risk in later life.

INTRODUCTION

The prevalence of obesity during pregnancy is increasing at an alarming rate(1). Maternal obesity (MO), often the result of a high fat, high energy diet (HF-HED), increases offspring risk of liver-specific dysfunction and disease in later life(2), but knowledge of how HF-HED impacts the function of factors involved in fetal liver function, specifically hepatic cytochrome P450 (CYP) enzymes, is poorly understood. CYP enzymes play a role in the initial oxidation of lipophilic chemicals (either endogenous or exogenous) to more polar metabolites that can undergo further metabolism and elimination(3). Altered CYP enzyme activity can reduce or increase rates of endogenous (steroids and vitamins) or exogenous (drugs and other xenobiotics) metabolism and clearance(4). Altered CYP function may disrupt normal fetal growth and development, thus programming an increased risk of adverse health outcomes in later life, including metabolic disease(5). Postnatal Western diet impacts CYP activity(6), and HF-HED-induced MO reduces the activity of CYPs involved in hormone metabolism in male fetuses only(7); however, it is unclear how HF-HED impacts the function of CYPs involved in exogenous drug metabolism. Understanding the impact of HF-HED on fetal CYP activity and the upstream pathways involved in their regulation is necessary to advance knowledge of the developmental origins of liver disease.

The pathophysiology of MO contributes to dysfunction of several pathways that are involved in regulation of CYP expression and/or function. MO increases serum inflammatory cytokine levels including interleukin (IL)-6, IL-1β, and tumour necrosis factor α (TNFα)(8). These inflammatory cytokines can alter expression and activity of certain CYP isoenzymes(9), but it is unclear whether this heightened state of maternal inflammation impacts fetal-specific CYP activity. Likewise, HF-HED increases circulating fetal cortisol concentrations(10). Increased cortisol binding to the glucocorticoid receptor (GR) is not only involved in development of metabolic disease, but also modulates expression of certain CYP isoenzymes including CYP2B6 and the fetal-specific CYP3A isoenzyme, CYP3A7(11). This GR-mediated mechanism of action may be complicated by alternative GR splice variations and translational isoforms. Different intracellular GR protein isoform expression patterns are known to impact canonical glucocorticoid-mediated signalling(12). However, no study has investigated the impact of maternal HF-HED on fetal hepatic GR isoform profiles and how this may impact CYP activity regulation.

An interesting and emerging aspect of fetal programming is sexual dimorphism of disease risk. Indeed, feto-placental sex differences in response to certain pregnancy challenges, including HF-HED-induced MO, have been reported, with some studies indicating males are more sensitive to disrupted metabolic homeostasis than females(13). Thus, these sex-specific adaptations to in utero challenges induced by HF-HED may dysregulate factors required for normal fetal CYP function and exacerbate the risk, severity, and prevalence of later life liver disease for males. Indeed, in a guinea pig model of life-long Western diet consumption we showed that CYP1A2 activity was reduced in males only, whereas activity of CYP3A was reduced in response to Western diet consumption in both sexes(6). More recently, our group showed reduced activity of testosterone-metabolising hepatic CYP enzymes in male fetuses exposed to HF-HED whereas no change was reported in females(7). Despite this, it is unclear whether these prenatal sex differences in hepatic dysfunction due to a HF-HED influence the activity of CYPs involved in exogenous chemical clearance.

There are several established animal models of MO during pregnancy and its impact on fetal metabolic outcomes. In the current study we have used a well-characterised baboon model(14) to investigate the impact HF-HED-induced MO has on fetal hepatic CYP function and determined if there are fetal sex dependent effects. We hypothesised that a HF-HED results in sex-specific changes in factors known to regulate CYP expression, resulting in altered activity of fetal hepatic CYPs, and that the greatest perturbation occurs in males.

METHODS

All animal procedures were approved by the Texas Biomedical Research Institute (TX Biomed) Institutional Animal Care and Use Committee (IACUC) and conducted in an Association for Assessment and Accreditation of Laboratory Animal Care (AAALAC) international-approved facility. The experimental design was informed by the ARRIVE guidelines and the 3Rs (https://www.nhmrc.gov.au/research-policy/ethics/animal-ethics/3rs).

Animal model

Baboons (Papio hamadryas spp.) were maintained in groups up to 16 in custom-built outdoor facilities allowing full socialisation and free movement. Healthy, non-pregnant female nulliparous baboons of similar weight, body dimensions and age were randomly assigned to either an ad libitum regular/normal diet (CD; Purina 5LEO; Purina LabDiets, St Louis, MO, USA; total energy content 2.98 kcal/g; n=12) or an ad libitum combination of the CD and HF-HED (Purina 5045–6, Purina LabDiets, St Louis, MO, USA; total energy content 4.03 kcal/g; n=12) with free access to a high fructose drink at least 9 months pre-pregnancy(15).

Caesarean sections and tissue collection

Pregnant CD and HF-HED baboons underwent Caesarean section at 165 days gestation (term=180 days) using standard sterile surgical techniques(16). Briefly, baboons were pre-medicated with ketamine hydrochloride (Fort Dodge Animal Health, Fort Dodge, USA; 10 mg/kg intramuscularly), and maintained at a surgical plane of anaesthesia with isoflurane (2%). After hysterectomy, the fetus was exsanguinated under general anaesthesia as approved by the American Veterinary Medical Association Panel on Euthanasia(14). Morphometric measurements were collected, and tissue samples obtained immediately. The right lobe of the fetal liver was snap frozen in liquid nitrogen and stored at −80°C.

Microsome extraction

Fetal liver microsomes were extracted using differential centrifugation as previously described(17, 18). Briefly, frozen tissue samples of ~250mg were homogenised (Tissue Lyser, Qiagen, Switzerland) in 600 μl of homogenising buffer (1.15% KCl, 1mM EDTA, pH 7.4). The homogenate was centrifuged at 9000 g for 20 min at 4°C. The supernatant was then transferred and centrifuged at 16,000 g for 60 min at 4°C. Supernatant was then discarded and the microsomal pellet was resuspended in buffer (100mM potassium phosphate buffer, 20% glycerol, pH 7.4). The amount of protein in each extraction was determined using a Micro Bicinchoninic Acid (BCA) Protein Assay Kit (Pierce, Thermo Fisher Scientific Inc, Rockford, USA) with bovine serum albumin (2 mg/ml) to generate a standard curve. Microsomal extractions were stored at −80°C until enzymatic assays.

CYP activity assay

To quantify the activity of CYP isoenzymes, the rate of phenacetin to paracetamol (CYP1A2), bupropion to hydroxybupropion (CYP2B6), omeprazole to 5’-hydroxyomeprazole (CYP2C19), amodiaquine to N-desethyl amodiaquine (CYP2C8), torsemide to hydroxytorsemide (CYP2C9), dextromethorphan to dextrorphan (CYP2D6), chlorzoxazone to 6’-hydroxychlorzoxazone (CYP2E1), and midazolam to 1-hydroxymidazolam (CYP3A) was determined by adding a predetermined concentration of substrate to 70 μg of microsomal protein, 10 mM NADPH, and assay buffer (50 mM phosphate buffer, 2 mM magnesium chloride, pH 7.4) to a total reaction volume of 100 μl as previously validated in sheep and guinea pigs(6, 17, 19). Reactions were stopped by adding 100 μl acetonitrile containing 100ng/ml of: hydroxyomeprazole-d3; midazolam-d6; and dextrorphan-d3 (Toronto Research Chemicals, Toronto, Canada). Incubation mixtures were centrifuged at 12,000 g for 10 min at 4°C. Supernatant was removed, evaporated to dryness using Genevac EZ-2 Plus Evaporating System (Genevac, UK) and reconstituted in water prior to subsequent analysis.

Validation of CYP activity in fetal liver microsomes

To identify species-specific activity of CYPs, initial time course experiments were conducted using 70 μg pooled microsomal protein isolated from fetal liver samples. Fetal liver microsomes were incubated with each parent drug for up to 240 min. Negative controls (no NADPH or no microsomes added) were performed in parallel. Subsequent inhibition studies were conducted on microsomal protein isolated from fetal liver to validate substrate-specificity for each CYP of interest. In brief, 70 μg pooled fetal liver microsomal protein was pre-incubated for 30 min with increasing concentrations of CYP-specific inhibitors (0–100 μM; CYP2B6, ticlopidine; CYP2C8, montelukast; CYP2C9, sulfaphenazole; CYP2C19, ticlopidine; CYP2D6, quinidine; CYP2E1, diethyldithiocarbamate; CYP3A, CYP3Cide)(19). After pre-incubation, CYP activity reactions were conducted as aforementioned.

Liquid chromatography tandem mass spectrometry (LCMS/MS)

Quantitation of CYP-specific metabolites was determined using liquid chromatography tandem mass spectrometry (LC-MS/MS; SCIEX 4500 Triple-Quad (SCIEX, US) with Shimadzu Nexera XR (Shimadzu, Japan)), as previously described(17, 19). Prepared samples were injected onto an ACQUITY BEH C18 Column 130Å, 1.7 μm, 2.1 mm × 100 mm (Waters Corp, US). Mobile phases were 0.05% Formic acid in water (A) and 0.1% Formic acid in 100% acetonitrile (B). Flow rate was 0.2 ml/min and mobile phase B was initially 20% for 1 min, increased linearly to 90% over 6 min and then held at 90% for 30 sec, after which it returned to 10% over 30 sec prior to injection of the next sample.

Western blot

Tissue subcellular fractionation and Western blotting was performed as previously described(7). Membranes were blocked in 5% BSA in TBS with 1% Tween (TBS-T) for 1 h at room temperature, and then incubated with affinity purified polyclonal rabbit anti- GR (1:1000, Bethyl Laboratories, Montgomery, USA, Cat no. A303–491A) or rabbit anti- GRβ (1:500, Thermo Fisher Scientific, USA, Cat no. PA3–514) antibodies. The appropriate secondary antibody (goat anti-rabbit; 1:20000) was applied for 1 h. Membranes were subsequently probed with anti-β actin (1:4000, Bethyl laboratories, USA Cat no. A300–491A). SuperSignal West Pico Chemiluminescent Substrate (Thermo Scientific) was used to detect reactive bands by enhanced chemiluminescence. Western blots were imaged using ImageQuant LAS 4000 (GE Healthcare) and protein abundance was determined by densitometry using Image Quant software (GE Healthcare) and initially normalised to their respective loading control. To allow for comparisons between subcellular GR isoform expression across membranes, expression data were normalised to the subcellular expression of matched GR isoforms in pooled samples that were included on each membrane(7).

Quantification of fetal liver mRNA expression

RNA was extracted and cDNA synthesized from fetal liver tissue samples (~50 mg) as previously described(20, 21). Target gene expression was measured by quantitative reverse transcription polymerase chain reaction (qRT-PCR) as previously described(20, 21).

Quantification of fetal liver glucocorticoids

Tissue hormone concentrations were determined by liquid chromatography (LC; Shimadzu Nexera XR, Shimadzu, Japan) coupled to a SCIEX 6500 Triple-Quad system (MS/MS; SCIEX, US) using an adapted protocol(17). Prepared samples were injected onto an ACQUITY UPLC BEH C18 Column 130Å, 1.7 μm, 2.1 mm × 100 mm (Waters Corp, US). Mobile phases were 0.1% Formic acid in water (A) and 0.1% Formic acid in acetonitrile (B). Flow rate was 0.3 ml/min and mobile phase B was initially 10% and increased linearly to 90% over 10 min and then held at 90% for 2 min, after which it returned to 10% over 3 min prior to injection of the next sample. Hormone concentrations were calculated via integration with a standard curve that ranged from 0.05 ng/ml to 100 ng/ml.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 8 (GraphPad Software, Inc., USA). Data were analysed using two-way ANOVA (factors = diet and sex) with Tukey’s post-hoc analysis. Potential outliers were detected using Grubbs’ test. Data are presented as mean ± SEM, unless stated otherwise. The alpha level was 0.05.

RESULTS

Maternal and fetal morphometric measurements

Some characteristics of the cohort, including maternal and fetal weight and fetal organ weight, have previously been described(7) and can be found in Table 1. Maternal weight in female-bearing pregnancies increased across gestation and was higher in HF-HED compared to CD (P_Diet=0.0125, P_Time=0.0010; data not shown); however, in male-bearing pregnancies, maternal weight did not change between groups or across gestation (P_Diet=0.1224, P_Time=0.1711; data not shown). Male fetal body weight and male fetal weight relative to maternal weight was increased compared with females irrespective of diet (P=0.0492 and P=0.0494, respectively; Table 1).

Table 1 –

Maternal and fetal morphometric measurements.

	Female		Male
	CD (n=6)	HF-HED (n=6)	CD (n=6)	HF-HED (n=6)	PDiet	PSex	PIntx
Maternal measurements
Maternal weight (kg; 30 d preconception)	17.1 ± 0.4	19.6 ± 1.0	17.8 ± 0.6	20.1 ± 0.6	0.003	0.395	0.896
Maternal weight (kg; 160 d GA weight)	18.3 ± 0.5	21.3 ± 0.7	19.1 ± 1.2	20.3 ± 0.9	0.023	0.884	0.307
Maternal weight change (kg)	1.2 ± 0.5	1.7 ± 0.4	1.7 ± 0.9	0.2 ± 0.7	0.449	0.407	0.120
Maternal glucose (mg/dl)	57.00 ± 9.74	85.50 ± 18.54	59.67 ± 12.49	137.33 ± 32.90	0.017	0.198	0.244
Fetal measurements
Fetal body weight (BW; g)	776.5 ± 41.4	746.4 ± 23.9	802.93 ± 40.90	855.50 ± 15.22	0.732	0.049	0.216
Fetal:maternal BW	0.04 ± 0.00	0.03 ± 0.00	0.04 ± 0.00	0.04 ± 0.00	0.064	0.049	0.052
Fetal heart weight (g)	4.52 ± 0.44	4.98 ± 0.30	5.28 ± 0.30	5.42 ± 0.23	0.367	0.081	0.624
Fetal liver weight (g)	23.89 ± 1.52	26.35 ± 0.77	24.09 ± 1.37	24.87 ± 0.98	0.200	0.609	0.502
Fetal brain weight (g)	83.61 ± 3.39	76.87 ± 2.24	86.06 ± 4.24	84.70 ± 2.99	0.233	0.135	0.423
Placental weight (PW; g)	193.12 ± 16.19	209.13 ± 13.48	213.91 ± 13.79	229.78 ± 6.27	0.253	0.142	0.996
Relative brain weight	0.11 ± 0.00	0.10 ± 0.00	0.11 ± 0.01	0.10 ± 0.00	0.087	0.540	0.630
BW:PW	4.12 ± 0.29	3.64 ± 0.23	3.78 ± 0.24	3.72 ± 0.14	0.260	0.596	0.394
Relative liver weight	0.03 ± 0.00	0.04 ± 0.00	0.03 ± 0.00	0.03 ± 0.00	0.094	0.003	0.014
Brain:liver	3.54 ± 0.14	2.93 ± 0.11	3.60 ± 0.16	3.43 ± 0.07	0.046	0.008	0.107
Fetal glucose (mg/dl)	49.40 ± 13.16	50.67 ± 7.77	50.33 ± 4.87	35.33 ± 15.60	0.504	0.484	0.430

GA = gestational age. Statistical analysis: two-way ANOVA (factors: diet and sex; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM.

Validation of CYP isoenzyme activity in the baboon fetal liver

Activity of CYP2B6, 2C8, 2C9, 2C19, 2D6, 2E1, and 3A isoenzymes were detected in isolated fetal liver microsomes (Figure 1A); however, when inhibition studies were performed, activity of only CYP2B6, 2C8, 2D6, and 3A showed species specificity (Figure 1B).

Figure 1. Quantification and validation of CYP activity in isolated fetal liver microsomes.

A) Activity of CYP2B6, CYP2C8, CYP2C9, CYP2C19, CYP2D6, CYP2E1 and CYP3A was measured over 6 hour time course periods. B) Inhibition studies were performed by pre-incubating fetal liver microsomes with increasing concentrations (0–100 μM) of CYP-specific inhibitors. Time course data are presented as the ratio of the area under the curve (AUC) of analyte to internal standard and inhibition data are presented as the % of activity relative to baseline.

MO impacts fetal liver CYP activity in a sex-specific manner

CYP2B6 activity was lower in HF-HED pregnancies in males only (85% reduction, P = 0.0091; Figure 2). CYP3A activity was reduced in fetuses from HF-HED pregnancies (62% reduction in males, 33% reduction in females; P = 0.0093; Figure 2). Fetal liver CYP2C8 and CYP2D6 activity were not affected by diet or sex (Figure 2).

Figure 2. Effect of diet and sex on fetal liver cytochrome P450 activity.

Activity of CYP2B6, CYP2C8, CYP2D6 and CYP3A was measured by quantifying the amount of metabolite produced per mg microsomal protein per minute of incubation (nmol/mg/min). Functional activity was measured in female and male fetal liver samples from control diet (CD; open data points) and high fat, high energy diet (HF-HED; closed data points) pregnancies. Limit of quantitation for each assay was 0.05ng/ml. Statistical analysis: two-way ANOVA (factors: diet and sex; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM.

Gene expression of transcription factors involved in CYP2B6 and CYP3A regulation are altered by maternal diet

Of the transcription factors involved in CYP2B6 and CYP3A regulation (constitutive androstane receptor (CAR), hepatic nuclear factor 4 alpha (HNF4A), hypoxia inducible factor 1 alpha (HIF1A), estrogen receptor (ESR1), and glucocorticoid receptor (GR)), HF-HED increased fetal liver HNF4A and GR expression, irrespective of sex (Figure 3).

Figure 3. Normalised mRNA expression of transcription factors involved in CYP2B6 and CYP3A regulation in the fetal liver.

Transcript expression was measured in female and male fetal liver samples from control diet (CD; open data points) and high fat, high energy diet (HF-HED; closed data points) pregnancies. Statistical analysis: two-way ANOVA (factors: diet and sex; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM.

Expression of genes involved in glucocorticoid signalling and inflammatory pathways are altered by maternal diet and sex

11β hydroxysteroid dehydrogenase type 1 (HSD11B1) was increased in the liver of female fetuses irrespective of maternal diet (Figure 4A). No other gene involved in glucocorticoid metabolism was changed by maternal diet or sex. Of the inflammatory markers measured, transforming growth factor beta 1 (TGFB1) increased in females; however, TGFB1 expression was decreased in response to HF-HED in both sexes (Figure 4B).

Figure 4. Normalised mRNA expression of A) glucocorticoid and B) inflammatory pathways in the fetal liver.

Transcript expression was measured in female and male fetal liver samples from control diet (CD; open data points) and high fat, high energy diet (HF-HED; closed data points) pregnancies. Statistical analysis: two-way ANOVA (factors: diet and sex; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM.

GR isoform expression is altered by maternal diet and sex

Six known GR isoforms were identified in cytoplasmic and nuclear fractions of baboon fetal liver including GRαA (94 kDa), GRβ (91 kDa), GRαC (81 kDa), GRP (74 kDa), GRA (65 kDa), and GRαD1–3 (50–55 kDa) (Figure 5B). Additional immunoreactive protein bands with molecular weights of 88, 68/69, 58 and 48 kDa were identified in both subcellular fractions, but have not been confirmed as GR isoforms.

Figure 5. Multiple glucocorticoid receptor (GR) protein isoforms are present in the cytoplasm and nucleus of the baboon fetal liver.

A) Cytoplasmic (C) and nuclear (N) protein extracts of female and male liver samples from control diet (CD) and high fat, high energy diet (HF-HED) pregnancies. Five known GR proteins were detected with molecular weights ranging from 50 to 95kDa. Blots were probed with β-actin as a loading control. B) Relative nuclear expression of GR isoforms of female and male liver samples from CD (open data points) and HF-HED (closed data points) pregnancies. Statistical analysis: two-way ANOVA (factors: diet and sex; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM. ⨂ = lane not analysed.

Relative nuclear GRA, GRαD1, and GRαD3 expression was increased in males compared to females (P=0.0070, P=0.0020, P<0.0001, respectively), independent of diet (Figure 5A). Relative GRβ nuclear expression was increased in females compared to males independent of diet (P=0.0302; Figure 5A); however, in males only relative GRβ nuclear expression was increased in response to HF-HED (P=0.0062). The relative nuclear expression of GRαC and GRP were both increased in response to diet (P=0.0253 and P=0.0313, respectively), independent of sex (Figure 5A). GRαA was not impacted by diet or sex. Of the uncharacterised proteins identified, relative nuclear expression of the 68/69 kDa band was higher in males (P=0.0039) and in response to maternal diet (P=0.0239), whereas the 58 kDa band was higher in males independent of diet (P<0.0001) and the 48 kDa band was increased in response to diet independent of fetal sex (P=0.0252; Data not shown).

Liver glucocorticoid concentrations

Liver concentrations of cortisol, cortisone, 11-deoxycortisol and corticosterone were not affected by maternal diet or fetal sex; however, the ratio of cortisone to cortisol was higher in males compared with females independent of diet (P=0.0331; Figure 6).

Figure 6. Liver glucocorticoid concentrations are not impacted by diet or sex.

Cortisol, cortisone, 11-deoxycortisol and corticosterone concentrations were quantified in fetal liver tissue from control diet (CD; open data points) and high fat, high energy diet (HF-HED; closed data points) and are displayed as ng per mg of tissue used for analysis. Statistical analysis: two-way ANOVA (factors = diet and fraction; Intx = interaction effect) with Tukey’s post-hoc analysis. Data presented as mean ± SEM.

DISCUSSION

Our study characterised for the first time the impact of HF-HED on fetal drug metabolism in the baboon model of human pregnancy. We show that reduced activity of fetal hepatic CYP3A and CYP2B6 in response to HF-HED was more apparent in males compared with females, which supports previous work from our group that found hepatic-specific testosterone metabolism via CYP3A and CYP2B6 was reduced in males from HF-HED pregnancies only(7). This impaired CYP activity in utero may not only reduce testosterone metabolism and thus enhance androgen signalling pathways in males, which may contribute to sex-specific fetal growth differences(22), but also compromise the ability of the male fetus to metabolise and clear medications that are prescribed during pregnancy including antidepressants, antinauseants and antihypertensives. We propose that a greater perturbation to CYP function in males is regulated in part by the differential expression and subcellular localisation of GR protein isoforms that initiate a state of hepatic-specific glucocorticoid resistance. Thus, targeting the glucocorticoid signalling pathway to ameliorate CYP dysfunction may improve male intrauterine outcomes and reduce the likelihood for metabolic disease risk in later life.

We reported a decrease in CYP3A activity in response to HF-HED, which may contribute to programmed liver pathophysiology and disease severity. Cyp3a-null male mice have increased liver triglyceride accumulation and liver weight in response to HFD, whereas Cyp3a-null females do not(23). In a guinea pig model of lifelong Western diet, hepatic CYP3A activity was reduced in both sexes(6). Importantly, our study is first to characterise fetal liver CYP3A activity using a clinically relevant probe, midazolam, within the context of a preclinical model of HF-HED-induced MO. Reduced CYP3A activity in response to HF-HED supports other work showing that the development of liver disease can occur due to a suboptimal intrauterine environment(2), and may therefore contribute to disease progression in later life.

Unlike CYP3A, activity of CYP2B6 was reduced only in males exposed to HF-HED when compared with control males. This finding is supported in part by a human study that showed females have increased activity compared with males, and that overall, males were more likely to be poor CYP2B6 metabolisers (indicative of reduced activity)(24). Cyp2b-null male mice, but not female mice, fed HFD are more susceptible to the development of fatty liver disease(25). Thus, fetal CYP2B6 dysfunction may contribute to the pathophysiology of liver disease. Indeed, our findings show that in response to HF-HED, male fetuses have an 85% reduction in liver CYP2B6 activity. Not only could this reduction in activity exacerbate the predisposition to develop liver disease, but may also contribute to altered efficacy and safety of medication use.

CYP3A and CYP2B6 metabolise up to 40% and 7% of clinically prescribed medications, respectively. Many of these medications are used clinically to manage pre-existing conditions such as depression (e.g., bupropion, CYP2B6), pregnancy-related complications including nausea and vomiting (domperidone and ondansetron, CYP3A), or are being studied in both preclinical and clinical trials. For example, CYP3A metabolises the phosphodiesterase inhibitors, including sildenafil and tadalafil. These drugs have been trialled preclinically and clinically as potential therapeutics for fetal growth restriction (26, 27, 28, 29, 30), which is more likely to occur in MO compared with lean pregnancies(31). Comparatively, CYP2B6 is involved in the metabolism of several antidepressants that are associated with adverse neonatal outcomes(32). Within the context of the current study, altered CYP2B6 and CYP3A activity may result in prolonged fetal exposure to CYP-specific substrates, especially in males, and highlights the need to consider fetal sex when managing pregnancy-specific complications such as MO.

There are many key regulators of CYP2B6 and CYP3A expression and activity that are impacted by HFD. For example, HFD reduces HNF4A(33) but increases CAR(34); these transcripts are known to directly regulate CYP2B6 and CYP3A expression(35). Of the transcription factors measured, we showed that HF-HED upregulated fetal liver HNF4A expression in both sexes. Despite this increase in transcription factor expression, hepatic CYP2B6 and CYP3A activity were reduced in response to HF-HED, especially in male fetuses. These findings suggest alternative pathways within the fetal liver may contribute to the observed reduction in CYP activity such as the glucocorticoid-mediated signalling pathway.

Although we reported no change in tissue-specific cortisol concentrations, we did observe an increase in the ratio of cortisone to cortisol in males independent of diet, which coincided with a reduction in 11BHSD1 expression. These data indicate that the hepatic interconversion of cortisone to cortisol may be suppressed in male fetuses, resulting in a dampened glucocorticoid signalling axis. Indeed, the proposed male-specific state of glucocorticoid resistance has previously been described in other organs including the human placenta(36, 37, 38), and may be one mechanism by which male growth in utero is greater than female(22). These previous studies show changes in glucocorticoid bioavailability, metabolism, and signalling are impacted by sex and the complication of certain pregnancy stressors including reduced fetal growth. However, to our knowledge, this is the first study to report changes in hepatic glucocorticoid signalling between the sexes that is further impacted by a maternal HF-HED.

Indeed, the differential expression and subcellular localisation of GR isoforms can modulate the intracellular responsivity to glucocorticoids, which can influence the regulation of downstream target genes. In the current study, we reported an increase in the relative nuclear expression of GRαD1, GRαD3 and GRA, and a decrease in GRβ in males compared with females, independent of diet. The only measured GR isoforms to be impacted by HF-HED were GRαC, GRP and GRβ. Importantly, changes in GRβ expression in response to HF-HED occurred in males only. GRβ forms a non-responsive heterodimer with GRα-A, thereby inhibiting target gene transcriptional regulation(22, 39). Indeed, others have shown obesity increases GRβ mRNA expression in the adult mouse liver and results in hepatic steatosis(41). Collectively, the sexual dimorphic changes in the interconversion of cortisone to cortisol, as well as the relative nuclear expression of GR isoforms, may suggest an overall suppression of canonical glucocorticoid signalling in males that is exacerbated by the presence of a maternal HF-HED. These changes may not only be a contributing factor to the observed suppression of CYP2B6 and CYP3A activity, but may also be involved in the developmental programming of liver dysfunction and/or disease in later life.

Despite the current study highlighting sex-specific differences in CYP isoenzyme activity and GR protein isoform expression in response to maternal HF-HED, there are some limitations. In the current study we were unable to quantify the activity of CYP2B6, CYP2C8, CYP2D6 or CYP3A in the maternal compartment due to biospecimen access limitations. Given previous studies in sheep have shown that changes to CYP activity in either maternal, fetal, or both compartments in response to pregnancy complications can occur(17, 18), measuring activity within the maternal compartment would enable larger profiling of the impact that HF-HED has on drug safety and efficacy. Indeed, the pharmacokinetics of medications during pregnancy is a complex process that is determined by effects in maternal, placental, and fetal compartments(35, 42); altered activity of certain CYP isoenzymes that are functional in the maternal or placental compartments, but not fetal compartments, may contribute to distinct sex-specific risks associated with intrauterine medication exposure. Therefore, future studies would benefit from incorporating all three compartments in vitro to assess the impact HF-HED has on phase 1 drug metabolism during pregnancy.

Although the in vitro assays used in this study are useful for assessing the impact of pathophysiological conditions on initial drug metabolism, this approach cannot measure drug disposition and pharmacokinetics. While there are many strengths for the use of non-human primates(43), there are some limitations that warrant discussion. Indeed, while non-human primates are excellent models of human health and disease, they are a highly resource-demanding model to establish and maintain. This inherently limits the number of animals able to be included in study groups; however, our previous studies in baboons(7) and guinea pigs(6) that examined the effect of sex and diet on hepatic CYP activity used comparable sample sizes and reported similar outcomes. Nonetheless, it would benefit future works to expand on the current findings using larger cohorts, but this is beyond the scope and feasibility of our study. Another limitation of the current model is the inability to chronically catheterise both maternal and fetal systems to deliver drugs and collect serial blood samples. This limits the capacity to perform complex in vivo pharmacokinetic studies; indeed, other large animal models such as the pregnant sheep overcome this limitation(17, 18) and may therefore be an appropriate model for future studies to assess the effect of HF-HED on in vivo drug disposition and pharmacokinetics. By advancing our understanding of how HF-HED impacts pharmacokinetics and how changes may be further impacted by feto-placental sex, future studies may be equipped to assess the efficacy and safety of medication regimens during pregnancy that take into consideration feto-placental sex.

CONCLUSION

In our current preclinical baboon model of HF-HED-induced MO, hepatic CYP activity is more impaired in males than females. These sex-specific changes to CYP activity may be driven by a state of hepatic glucocorticoid resistance in males only. Our findings not only indicate that fetal sex should be considered when prescribing medications during pregnancy, but also identifies targetable pathways that may improve intrauterine CYP activity and thus enhance the efficacy and safety of medication use in pregnancy.

STUDY IMPORTANCE QUESTIONS.

What is already known about this subject?

• Maternal obesity (MO), often the result of a high fat, high energy diet (HF-HED) increases fetal glucocorticoid concentrations and the risk of later life liver disease in offspring, particularly males.

• Altered liver cytochrome P450 (CYP) enzyme activity can exacerbate liver disease risks and impact medication safety and efficacy.

• Changes to glucocorticoid signalling can impact hepatic CYP activity.

What are the new findings in your manuscript?

• CYP3A activity was reduced in male and female HF-HED groups, whereas CYP2B6 activity was reduced in HF-HED males only.

• Males compared with females had dampened liver-specific glucocorticoid signalling, as determined by changes to 11BHSD1 expression and a greater cortisone-to-cortisol ratio.

• HF-HED males had increased relative nuclear expression of the antagonistic glucocorticoid receptor (GR) isoform, GR-β, when compared with control males.

How might your results change the direction of research or the focus of clinical practice?

• Our results highlight distinct sexual dimorphism in fetal hepatic function in response to HF-HED, especially in males; these findings indicate that fetal sex should be considered when prescribing medications during pregnancy to reduce adverse outcomes associated with intrauterine drug exposure.

• Knowledge gained provides novel, targetable pathways that may reduce programmed liver dysfunction and/or disease in offspring from HF-HED pregnancies.