Sex differences in the epigenetic regulation of chronic visceral pain following unpredictable early life stress

Abstract

Background:

We previously reported that early life stress (ELS) dysregulated glucocorticoid receptor (GR) and corticotrophin-releasing hormone (CRH) expression in the central nucleus of the amygdala (CeA). Epigenetic modifications serve as memories of adverse events that occurred during early life. Therefore, we hypothesized that epigenetic mechanisms alter GR and CRH expression in the CeA and underlie chronic visceral pain after ELS.

Methods:

Neonatal rats were exposed to unpredictable, predictable ELS, or odor only (no stress control) from postnatal days 8 to 12. In adulthood, visceral sensitivity was assessed or the CeA was isolated for Western blot or ChiP-qPCR to study histone modifications at the GR and CRH promoters. Female adult rats underwent stereotaxic implantation of indwelling cannulas for microinjections of garcinol (HAT inhibitor) into the CeA. After 7 days of microinjections, visceral sensitivity was assessed or the CeA was isolated for ChIP-qPCR assays.

Results:

Unpredictable ELS increased visceral sensitivity in adult female rats, but not in male counterparts. ELS increased histone 3 lysine 9 (H3K9) acetylation in the CeA and H3K9 acetylation levels at the GR promoter in the CeA of adult female rats. After unpredictable ELS, H3K9 acetylation was increased and GR binding was decreased at the CRH promoter. Administration of garcinol in the CeA of adult females, that underwent unpredictable ELS, normalized H3K9 acetylation and restored GR binding at the CRH promoter.

Conclusion:

Dysregulated histone acetylation and GR binding at the CRH promoter in the CeA are an important mechanism for “memorizing” ELS events mediating visceral pain in adulthood.

1 |. INTRODUCTION

Chronic visceral pain is a common complaint in patients with irritable bowel syndrome (IBS), a gastrointestinal disorder that predominately affects women. IBS patients are also two to four times more likely to report a history of early life stress (ELS), such as child abuse, poverty, or trauma, when compared to healthy control subjects.¹ Women often exhibit stronger connections to stressors that occurred in early life, such as physical or sexual abuse, and this may partially explain the gender disparity observed in IBS patients where female IBS patients outnumber males by 2:1.² Stress during critical periods of enhanced neural plasticity in early life can have a profound impact on the formation of neural circuits in the brain and consequently lead to psychopathologies in adulthood, such as depression and/or anxiety disorders, which are often comorbid with IBS.³ Several regions of the limbic brain, implicated in depression and anxiety disorders,⁴ are also key components of in the communication between the brain-gut axis.⁵ In this way, ELS-induced developmental abnormalities can lead to psychopathologies and abnormal communication along the brain-gut axis, which in turn can facilitate abnormal visceral pain reporting.⁶

The hypothalamic-pituitary-adrenal (HPA) axis is an important mediator in brain-gut communication. The HPA axis is the neuroendocrine system that mounts the body’s response to stress. When stress signals are integrated in the paraventricular nucleus (PVN) of the hypothalamus, corticotrophin-releasing hormone (CRH) will be secreted from these cells onto the anterior pituitary. The activation of the HPA axis leads to the release of the glucocorticoid cortisol (CORT) from the adrenal glands into the blood stream. CORT then regulates the HPA axis activity via the high-affinity mineralocorticoid receptor (MR) and the low-affinity glucocorticoid receptor (GR). CORT binding in the PVN and the hippocampus induces a negative feedback loop to turn off the stress response, whereas CORT that binds on the amygdala triggers the release of CRH, which facilitates the stress response. ELS can alter the activity of the hippocampus and amygdala, leading to decreased inhibition or increased facilitation of the HPA axis.^7,8 Interestingly, IBS patients often exhibit abnormal activation or increased amygdala-mediated facilitation of the HPA axis due to the altered GR expression and signaling or increases in CRH secretion from the amygdala.^9–11

Although visceral pain is a key feature of IBS symptomatology, it is mechanistically still unclear how ELS can lead to visceral hypersensitivity in IBS patients. We have previously shown that female rats, exposed to unpredictable stress as neonates, exhibit visceral hypersensitivity in adulthood. Interestingly, female rats exposed to predictable stress and male rats exposed to both forms of conditioned ELS show no long-term abnormality in visceral sensitivity. Based upon our experimental evidence, we attributed the long-term change in visceral nociception, in response to neonatal unpredictable ELS, to increased CRH expression levels in the central nucleus of the amygdala (CeA).¹² As visceral hypersensitivity was caused by long-term gene expression alterations, it is reasonable to postulate that this mechanism is controlled by epigenetic adaptations. Epigenetic changes, such as DNA methylation, histone modifications, and RNA interference, are essential modifiers of gene expression.¹³ As epigenetic modifications are usually long-lasting and do not require the initial environmental trigger in order to remain in place,¹⁴ they are widely implicated in the regulation of long-term adaptations of early life experiences and stress. For instance, evidence suggests that the chronicity of visceral pain in response to neonatal maternal separation was embedded by epigenetic dysregulation in the spinal cord and extended beyond the initial exposure of the stressor.¹⁵

In our study, unpredictable ELS only caused visceral hypersensitivity in adult female rats, and therefore, it is plausible that a sexual dimorphic epigenetic mechanism is responsible for these permanent increases in CRH expression. One epigenetic mechanism that is associated with increases in gene expression is histone acetylation, which is mediated by histone acetyltransferases (HAT).¹⁶ We hypothesized that ELS increases histone acetylation at the CRH promoter, causing the upregulation of CRH in the CeA, which leads to visceral hypersensitivity. Furthermore, we investigated whether inhibiting HATs in the CeA normalized histone acetylation levels and consequently CRH expression, in order to ameliorate visceral hypersensitivity in female rats previously exposed to unpredictable neonatal ELS.

2 |. MATERIALS AND METHODS

2.1 |. Animals

Twenty-four nulliparous female timed-pregnant (embryonic day 9 on arrival) Long-Evans rats were purchased from Charles River Laboratories (Wilmington, MA). Dams were housed at 23°C on a 12-hour light/dark cycle (7:00 AM:7:00 PM) with food (5053 Irradiated PicoLab Rodent Diet; Dietlab) and water available ad libitum. Over the course of the study, 242 (119 male) pups were born. On postnatal day (PND) 1, litters were sexed, cross-fostered, and culled to a minimum of 8 up to a maximum of 12 pups per litter (male:female ratio was 1:1). Rat pups were weaned 2 pups per cage according to treatment on PND 22. All protocols were in accordance with the Guide for the Care and Use of Laboratory Animals and approved by the Oklahoma City VAMC Institutional Animal Care and Use Committee (IACUC) (1403–001, 1507–001, 1806–003), and the University of Oklahoma Health Sciences Center IACUC (17–090-SFAHI) approved all animal procedures. All animals were acclimated to the laboratory and experimental areas for 2 weeks before adult experiments.

2.2 |. Induction of early life stress via odor-shock conditioning

Long-Evans pups were housed with dams on aspen wood shaving bedding from birth (PND 0) to PND 22. A total of 185 pups were subjected to the odor-shock conditioning paradigm described by Sullivan et al (2000), which occurred during lights on between 7:00 AM and 7:00 PM.^17,18 Pups were transported from the home cage to the conditioning room (held at a constant temperature of 28°C) on aspen wood shaving bedding. Pups were allowed to acclimate to the conditioning room for 5 minutes before conditioning began. Conditioning took place under red light. Briefly, from PND 8 to 12 pups were conditioned by predictable odor-shock presentation, unpredictable odor-shock presentation, or odor-only presentation as controls. Conditioning was performed using 11 trials of 30-second peppermint odor presentation with 4 minutes between odor presentations. Peppermint oil (Fisher Scientific) was vaporized at a 1:10 concentration of odor and administered at 2 L/min using a flow dilution olfactometer (Med Associates). Predictable odor-shock pups received a 0.5 mA shock (Coulbourn Instruments) to the base of the tail during the final second of odor presentation, eliciting a learned preference for the conditioned odor.¹⁷ Unpredictable odor-shock pups received the 0.5 mA shock to the base of the tail 2 minutes after odor presentation. Behavioral activation was observed during each of the 11 conditioning trials each day and scored on a scale of 0 to 5, where 0 represented no movement and five represented the movement of all four limbs and the head. After conditioning, the skin/fur of the pups was marked with a permanent marker to distinguish treatment groups. A Y-maze was used to verify learned odor preference following the conditioning protocol. On PND 13, pups were subjected to five trials, 1 minute each, and given the choice between fresh aspen bedding or the conditioning peppermint odor. Twelve pups from the predictable group did not show a preference for the peppermint odor (3 or more out of 5 trials) and 2 pups from the unpredictable group that did show a preference for the peppermint odor were removed from the study and euthanized at weaning.

2.3 |. Bilateral cannulation implantation and drug administration

The procedure for cannula implantation has been described in detail previously.^19,20 Cannulas were implanted between 8:00 AM and 5:00 PM. After anesthetizing the rat with 2% isoflurane (Henry Schein Animal Health), the rat received a subcutaneous injection of Carprofen (5 mg/kg). Bilateral 26-gauge cannulas (Plastics One, Inc) were implanted on the dorsal margin of the CeA based on the stereotaxic coordinates obtained from Paxinos and Watson (bregma −2.5 mm, medial/lateral ± 4.2 mm, and anteroposterior −7.0 mm from dura). An incision was made down the midline, and the skull was cleared from the underlying fascia. Two stainless steel mounting screws were placed on opposite side of the cannula. Each guide was secured using dental cement (Stoelting) and closed using a dummy cannula. Before closing the incision with wound clips (9 mm, Kent Scientific Corporation), 2–3 drops of 0.5% bupivacaine were applied in each quadrant. Bupivacaine was removed after 2 minutes. Rats recovered for 1 week, during which Carprofen (0.1 mL/kg) was administered subcutaneous the first 2 days postsurgery. Garcinol (1 ng/nL; Tocris) or vehicle (50% dimethyl sulfoxide in artificial cerebrospinal fluid; Tocris) was administered once daily for a total of 7 days. For drug administration, the rats were transported from the animal facility to the laboratory between 8:00 AM and 12:00 PM. Rats were anesthetized with 2% isoflurane and positioned within the anesthesia mask so that the cannula was accessible. The dummy cannula was unscrewed, and the matched injector was placed into the cannula. A total volume of 0.5 μL of garcinol or vehicle was administered through each cannula at a rate of 0.1 μL/min for a total of 5 minutes using a microsyringe (Hamilton) and UMP3 injection pump (World Precision Instruments). The injection cannula was left in place for an additional 2 minutes to ensure complete diffusion, before it was removed and replaced by the dummy cannula. Rats were transported back to the animal facility after they had recovered from anesthesia.

2.4 |. Visceral sensitivity assessment

Visceromotor responses (VMR) to colorectal distention (CRD) were quantified as the number of abdominal contractions in response to graded pressured of isobaric CRD (0–60 mmHg) in freely moving rats as previously described.^21,22 On the day before VMR, rats were placed in a fasting cage (standard housing cage where the bedding was replaced with a wire mesh bottom, ad libitum water access) around 3:30 PM for overnight fasting. Rats were transported from the animal facility to the laboratory between 8:00 and 9:00 AM. VMRs were conducted between 10:00 AM and 2:00 PM to reduce diurnal variations. Rats were anesthetized by 2% isoflurane inhalation, and in both male and female rats, a 5-cm latex balloon catheter was inserted up to 11 cm from the anus and secured to the tail using surgical tape. Rats were then allowed to recover for 30 min in a standard housing cage with normal bedding. Catheters were connected to a Distender Series IIR Barostat (G&J Electronics Inc) for delivery of controlled, isobaric inflation of the balloon, and CRD at randomized distentions of 0, 20, 40, and 60 mmHg. During each 10 minutes of distention, the number of abdominal muscle contractions was counted. An abdominal contraction consisted of longitudinal stretching of the body and visible contraction of the abdominal cavity. A 10-minute recovery period was allowed between subsequent distention pressures. Rats were euthanized after the VMR.

2.5 |. Tissue collection

Tissue was collected from rats that did not undergo behavioral testing. These animals were either unmanipulated or had received vehicle or garcinol infusions, in which case tissue was collected 24 hours after the final infusion. Rats were anesthetized with isoflurane and decapitated in order to isolate the brain. After extraction, brains were placed in a precision coronal brain matrix (Braintree Scientific Inc) and a 2-mm slice targeting the CeA was removed. With 1-mm precision micropunches (Braintree Scientific Inc), the CeA were collected. Tissue punches were flash-frozen in liquid nitrogen and stored at −80°C until processed.

2.6 |. Nuclear protein extractions

Nuclear proteins were extracted from the tissue samples using the NE-PER Nuclear and Cytoplasmic Extraction Kit (Thermo Fisher Scientific) in the presence of protease and phosphatase inhibitor cocktails (Thermo Fisher Scientific) according to the manufacturer’s protocol. Protein quantification of the extraction product was performed using BCA protein assay kit (Thermo Fisher Scientific). Following quantification, the samples were stored at −80°C for subsequent analysis.

2.7 |. Western blot

Quantification of proteins was done via Western blot. One μg of nuclear protein extract was solubilized in Laemmli buffer supplemented with 2-mercaptoethanol and denatured at 95°C for 5 minutes. The samples were resolved on a 4%−20% gradient Tris-Glycine polyacrylamide gel (Bio-Rad) using sodium dodecyl sulfate polyacrylamide gel electrophoresis and transferred onto a PVDF membrane (Bio-Rad). The membranes were blocked with 5% milk in TBS for 1 hour. Blots were incubated overnight at 4°C with primary antibodies: rabbit anti-acetyl-H3K9 (1:500, #9649, Cell Signal Technologies) and rabbit anti-H3 (1:1000, #4499, Cell Signal Technologies), which was used for normalization. Following antibody incubation, the blots were washed in three changes of TBS-T and incubated for 1 hour with HRP-conjugated secondary anti-rabbit antibodies (1:2000, #7074, Cell Signal Technologies). After three more washes in TBS-T, bands were visualized with ECL Western Blot Detection Kit (Bio-Rad) and imaged using Chemidoc (Bio-Rad). Densitometry was performed using the ImageLab software (Bio-Rad).

2.8 |. Chromatin immunoprecipitation

Chromatin immunoprecipitation (ChIP) assays were adapted to MAGnify ChIP assay protocol (Thermo Fisher Scientific). Chromatin was isolated from CeA micropunches and immunoprecipitated using one of the following antibodies: rabbit anti-acetyl-H3K9 (1:25, #9649, Cell Signal Technologies), rabbit anti-GR (1:25, #3360, Cell Signal Technologies), or normal non-immune rabbit IgG antibody (1:100, Thermo Fisher Scientific). One-tenth of the lysate was reserved as an input control. After reverse cross-linking, the rat CRH (forward primer: 5′-TCAGTATGTTTTCCACACTTGGAT-3′ and reverse primer: 5′-TTTATCGCCTCCTTGGTGAC-3′) promoter region was subjected to real-time PCR amplification using QuantiFAST SYBR Green PCR Mastermix (Qiagen) on a StepOne Plus System (Thermo Fisher). The thermocycler protocol involved an initial denaturation cycle (5 minutes, 95°C), 40 cycles of denaturation (10 seconds, 95°C), annealing, and extension (30 seconds, 60°C) and finally held at 4°C. All reactions were carried out in triplicate. Relative quantification of binding was calculated by normalizing the immunoprecipitated DNA C(t) values to the input DNA C(t) values [ΔC(t)] and transformed [2^ΔC(t)] to show relative quantities.

2.9 |. Experimental design

Neonatal rats, that had learned the correct association during the ELS paradigm, were divided into four cohorts in adulthood. The first cohort consisted of female and male rats that underwent ELS. In this cohort, visceral sensitivity was assessed in adulthood (starting at PND90) via a graded response to colorectal distention (CRD) and quantified as the number of abdominal contractions during the 10-minute distention period. The second cohort also consisted of female and male rats that underwent ELS. The brain of these rats was isolated for tissue collection and subsequent molecular analyses. The third cohort consisted of female adult rats, previously exposed to ELS. In adulthood (starting around PND90), these rats underwent stereotaxic implantation of an indwelling cannula for bilateral microinjections into the CeA of a non-specific HAT inhibitor (garcinol) or the vehicle control for 7 days. Twenty-four hours after the final infusion, visceral sensitivity was assessed. The fourth cohort consisted of animals that received the same ELS paradigm, stereotactic surgery, and infusion scheme. Twenty-four hours after the final infusion, the CeA was micro-dissected from these animals for molecular analyses.

2.10 |. Statistical analysis and experimental rigor

Data are represented as the mean ± SD. Animal numbers for the behavioral analysis and molecular experiments were based on prior experimental experience rather than a sample size calculation. Rat pups were randomly assigned to ELS conditioning paradigms (odor only, predictable, or unpredictable ELS). Uneven group numbers were caused by the fact that not all rat pups learned the association correctly during their conditioning. Stereotaxic surgery was performed on adult conditioned rats, which were again randomly assigned to treatment groups (vehicle or garcinol infusions). Behavioral testing was performed by an experimenter blinded to treatment groups. After behavioral assessment, rat brains were isolated and the CeA was checked for cannula placement and to verify that garcinol infusions did not cause abnormal apoptosis. For molecular analysis, the CeA was isolated from rats that had not undergone colorectal distention in order to avoid any potential confounding influence from the behavioral assays. Molecular analyses were performed by an experimenter who was not blinded to treatment groups. A two-way repeated measurement analysis of variance (ANOVA) was used to analyze the differences in VMR responses between ELS animals. A two-way repeated measurement analysis of variance (ANOVA) was performed to analyze differences in VMR responses between ELS animals after treatment. All ANOVAs were followed by a Bonferroni post hoc test. One-way ANOVAs with Tukey post hoc test (where applicable) were performed to analyze Western Blot and ChIP-qPCR data. ChIP-qPCR after garcinol treatment was analyzed with an unpaired t test. All data were analyzed with GraphPad Prism 8.0.

3 |. RESULTS

3.1 |. Unpredictable ELS-induced visceral hypersensitivity in adult females

We observed a main effect of ELS (F_(2,12) = 8.992, P = .0041), a main effect of pressure (F_(3,36) = 118.4, P < .0001), and a significant ELS X pressure interaction (F_(6,36) = 13.86, P < .0001) in female rats. As shown in Figure 1C, after Bonferroni post hoc corrections, the number of abdominal contractions in response to visceral distention was significantly increased in female rats exposed to unpredictable ELS (n = 4), when compared to female rats from the predictable ELS (n = 5, P < .0001) or odor-only group (n = 5, P < .0001). In male rats, exposure to predictable or unpredictable ELS had no effect on the number of abdominal contractions in response to visceral distention (F_(2,10) = 0.3611, P = .7057), as the number of abdominal contractions of male rats previously exposed to predictable (n = 5) or unpredictable ELS (n = 5) did not differ significantly from odor-only male controls (n = 5, P > .2139) (Figure 1B).

FIGURE 1.

Sexually dimorphic effects of unpredictable early life stress (ELS) on visceral sensitivity and H3K9 acetylation in the CeA of adult rats. (A) Experimental design. (B-C) After ELS, no differences in visceral sensitivity were observed between groups of male rats. However, in female rats, neonatal unpredictable ELS (n = 4) induced visceral hypersensitivity compared with predictable ELS (n = 5) and odor-only controls (n = 5) at 40 mmHg and 60 mmHg. Data represent mean ± SD with significance at ***P < .001, ****P < .0001 by repeated measures two-way ANOVA and Bonferroni post hoc test. (D–E) Representative images and Western blot quantification of acetylated H3K9 (~17 kDa) and H3 (~17 kDa) in the CeA of adult male rats previously exposed to odor only (n = 4), predictable (n = 4), or unpredictable ELS (n = 4), and adult female rats previously exposed to odor only (n = 4), predictable (n = 3), or unpredictable ELS (n = 4). Data shown as mean ± SD.; *P < .05, ****P < .0001 compared with odor only by one-way ANOVA

3.2 |. Importance of histone acetylation in the CeA of adult females previously exposed to unpredictable ELS

An epigenetic mechanism may underlie long-lasting alterations in visceral sensitivity, and we therefore sought to determine functionally relevant histone modifications. We assessed the global levels of acetylated histone 3 lysine 9 (H3K9) in the CeA and observed that acetylated H3K9 was significantly increased (P < .05) in female rats from the unpredictable ELS group (n = 4), when compared to the odor-only group (n = 4) (Figure 1E). ELS had no effect on the levels of H3K9 acetylation in the CeA of male rats (n = 4/group) (Figure 1D). As male animals did not show any phenotype or molecular changes, we continued our remaining experiments in female animals. Since acetylated H3K9 is a marker for transcriptional activation, we performed ChIP-qPCR assays on tissue from the CeA to evaluate whether increased levels of H3K9 acetylation were associated with increased enrichment at the GR exon I7 and CRH promoter in female rats previously exposed to ELS. A one-way ANOVA revealed that there was a significant effect of ELS on GR (F_(2,16) = 15.69, P = .0002) and CRH (F_(2,16) = 6.680, P = .0078) expression levels. Post hoc analysis revealed that acetylated H3K9 was enriched at the GR promoter, respectively, 2.91-fold and 3.12-fold after predictable (n = 5) and unpredictable ELS (n = 6) when compared to odor-only controls (n = 8) (Figure 2B). Further post hoc analysis discovered that acetylated H3K9 was enriched at the CRH promoter, respectively, 2.05-fold only after unpredictable ELS when compared to odor-only controls (Figure 2C).

FIGURE 2.

Unpredictable early life stress (ELS) increases promoter activity of corticotrophin-releasing hormone (CRH) in the CeA of adult female rats. Analysis of the chromatin structure and glucocorticoid receptor (GR) binding at the CRH promoter in the CeA. (A) Experimental design. (B-C) Chromatin was isolated from the CeA and immunoprecipitated with antibodies for acetylated H3K9 followed by real-time PCR analysis for GR and CRH. (D) Chromatin immunoprecipitation with GR antibodies and real-time PCR analysis of CRH promoter. The amount of immunoprecipitated DNA was normalized to input DNA loading control. Data shown are mean ± SD; animals/group: odor only = 8, predictable ELS = 5, and unpredictable ELS = 6; **P < .01, P < .001 compared with odor only by one-way ANOVA

3.3 |. Unpredictable ELS decreases GR binding at the CRH promoter

GR is a known negative regulator of CRH expression in the CeA. We next sought to determine whether ELS caused a decrease in GR binding at the CRH promoter region by using ChIP-qPCR. A one-way ANOVA revealed that there was a significant effect of ELS on CRH expression levels (F_(2,16) = 14.39, P = .0003). Post hoc analysis showed that, when compared to odor-only controls (n = 8), GR binding at the CRH promoter did not change after predictable ELS (n = 5), but was reduced by 64% (P = .0012) after unpredictable ELS (n = 6) (Figure 2D).

3.4 |. HAT inhibition attenuates ELS-induced visceral hypersensitivity

To further investigate the mechanism by which increases in HAT expression/activity influences visceral hypersensitivity, we employed a pharmacological approach and bilaterally administered the HAT inhibitor garcinol directly into the CeA (Figure 3B). In a pilot study, Apoptag analysis revealed that the microinjections into the CeA caused no damage to the structural integrity of the CeA (Figure 3C). Our statistical analysis revealed a main effect of pressure (F_(3,96) = 90.95, P < .0001), drug treatment (F_(5,96) = 14.4, P < .0001), and a significant pressure X drug treatment interaction (F_(15,96) = 2.921, P = .0008) in female rats. Bonferroni post hoc correction showed that after garcinol administration in the CeA, the number of abdominal contractions in response to colorectal distention was decreased in female adult rats, previously exposed to unpredictable ELS (n = 5), and comparable to numbers observed in odor-only controls (n = 5) (Figure 3D–F).

FIGURE 3.

Infusions of the histone acetyltransferase inhibitor garcinol (GAR) directly into the central nucleus of the amygdala (CeA) attenuate visceral hypersensitivity in female rats previously exposed to neonatal unpredictable early life stress (ELS). (A) Experimental design. (B) Stereotaxic administration of vehicle (VEH) or GAR in the CeA. (C) Apoptosis of the CeA after infusions with vehicle or GAR was verified with Apoptag and revealed that GAR infusions did not cause increased apoptosis in the CeA. (D–F) GAR infusions in the CeA of female rats, exposed to unpredictable ELS, attenuated visceral hypersensitivity, but did not affect visceral sensitivity of female rats exposed to predictable ELS or odor only. Animals/group: Odor Only + VEH = 5, Odor Only + GAR = 5, Predictable + VEH = 5, Predictable + GAR = 4, Unpredictable + VEH = 4, Unpredictable + GAR = 5. Data represented as mean ± SD with significance at *P < .05, ***P < .001 by repeated measures two-way ANOVA and Bonferroni post hoc test

3.5 |. HAT inhibition normalizes H3K9 acetylation and restores GR binding at the CRH promoter

In light of our finding that HAT inhibition reversed visceral hypersensitivity, next we investigated the underlying mechanism. In a new cohort of female rats that underwent unpredictable ELS, we bilaterally administrated vehicle (n = 6) and garcinol (n = 4) into the CeA. After collection of the CeA, we performed additional ChIP-qPCR assays of the CRH promoter region. An unpaired t test revealed that H3K9 acetylation was significantly reduced at the CRH promoter (P = .0252), whereas GR binding at the CRH promoter was significantly enriched (P = .0015) (Figure 4B–C).

FIGURE 4.

Garcinol infusions normalized corticotrophin-releasing hormone (CRH) promoter activity of adult female rats exposed to unpredictable early life stress (ELS). (A) Experimental design. (B) Chromatin was isolated from the CeA treated with vehicle (VEH) or garcinol (GAR) and immunoprecipitated with antibodies for acetylated H3K9 followed by real-time PCR analysis of CRH. Normalization to input DNA loading control. (C) Chromatin immunoprecipitation with glucocorticoid receptor antibodies and real-time PCR analysis of CRH promoter. The amount of immunoprecipitated DNA was normalized to input DNA loading control. Data shown are mean ± SD; animals/group: Unpredictable + VEH = 6, Unpredictable + GAR = 4; *P < .05, P < .01 compared with odor only by unpaired Student’s t test

4 |. DISCUSSION

In this study, we have shown that a sexually dimorphic epigenetic mechanism underlies visceral sensitivity in adult female rats after exposure to neonatal ELS. Both predictable and unpredictable neonatal stress increased H3K9 acetylation at the GR promoter in the CeA of adult female rats. This epigenetic modification induced resilience to visceral hypersensitivity after predictable ELS. In contrast, after exposure to unpredictable ELS female rats exhibit visceral hypersensitivity, and in spite of the increased H3K9 acetylation at the GR promoter in the CeA of adult female rats, we also observed decreased GR binding and increased H3K9 acetylation at the CRH promoter in the CeA of adult female rats. Taken together, the increase in visceral sensitivity following neonatal unpredictable ELS appears to be directly attributable to changes in histone acetylation in the CeA. In support, we demonstrated that stereotaxic microinjections of the HAT inhibitor garcinol directly into the CeA significantly decreased H3K9 acetylation at the CRH promoter, restored GR binding at the CRH promoter, and attenuated visceral hypersensitivity. In contrast, in male rats ELS did not lead to visceral hypersensitivity in adulthood. The lack of epigenetic changes in the CeA of male rats likely accounts, at least in part, to their resilience to neonatal ELS.

A multitude of experimental models of ELS is currently employed to recapitulate childhood adversity and to elucidate the underlying central molecular mechanisms that underlie anxiety, depression, and visceral pain in adulthood following a history of ELS.^15,23–27 In the current study, we used the odor-associated learning model of neonatal adversity that mimics early attachment to an abusive caregiver.¹⁷ This model takes advantage of the stress-hyporesponsive period, during which the HPA axis of neonates is suppressed. Adequate maternal behavior is an important mediator of the stress-responsive period seen in human infants and rat neonates.^28,29 Although the neonates are separated from the dam for the odor-shock conditioning, the separation is not long enough to induce changes in maternal care. As a result, the HPA axis and CORT-responsive regions, such as the amygdala, are not activated during the neonatal conditioning paradigm, which allows for the correct development of the HPA axis. In contrast, other rodent models of ELS such as maternal separation and limiting bedding induce profound changes in maternal behavior,^30,31 which results in the premature activation of the HPA axis and other CORT-responsive regions in the neonates.³² Consequently, the neonatal disruption of the HPA axis leads to an increase in basal HPA axis activity in adulthood. The maternal separation model also induces changes in the nutritional status of the rat pups that results in lower weaning weights of the pups, whereas the limited nesting and odor attachment learning models do not.³³ This absence of nutritional changes in the odor attachment learning model is important because nutritional status in early life has a severe impact on the epigenome.³⁴ As the odor-associated model does not induce changes in weaning weights, the epigenetic effects induced by predictable and unpredictable stress are not confounded by nutritional status. Finally, when we compared all three models in the Long-Evans rat strain, we found that only the odor-associated learning model led to visceral hypersensitivity in female rats, whereas maternal separation and limited nesting induced a male phenotype.³³ When investigating which brain regions were involved, we found that the odor-associated learning model led to changes in gene expression in the CeA of adult female rats, whereas this has not yet been validated in the other ELS models.^12,15 Taken together, the construct validity of the odor-associated learning model is preferable over maternal separation or limited nesting as IBS patients are predominantly female, show a hyperactive amygdala in brain imaging studies, and are more likely to be exposed to unpredictable adversity in early life.³⁵

We have previously reported that unpredictable ELS leads to visceral hypersensitivity in adult female rats through increases in the expression of GR and CRH in the CeA.¹² As GR is a negative regulator of CRH, the concomitant increase in both GR and CRH after unpredictable ELS was unexpected. We hypothesized that the increase in GR might have been a compensatory mechanism for the uncoupling of the GR-CRH interaction. In the current study, we provide further evidence for an epigenetic mechanism that supports our earlier reported increase in GR and CRH expression in the CeA. Although we are aware that ELS is likely to dysregulate epigenetic modifications in many genes, we choose to specifically elucidate the epigenetic mechanism at the GR and CRH promoters due to the important roles of amygdala GR and CRH in visceral hypersensitivity. We found that both predictable and unpredictable ELS caused a global increase in H3K9 acetylation in the CeA of adult females. Interestingly, H3K9 acetylation was significantly higher after predictable ELS compared with that seen in rats exposed unpredictable ELS. Several studies have shown that resilient phenotypes are associated with increased transcriptional activity, when compared to susceptible phenotypes.^36–38 Therefore, the resilience of female rats to develop visceral hypersensitivity, after exposure to predictable ELS, might be due to increased H3K9 acetylation in the CeA. However, interpretation of this increase in H3K9 acetylation is complex; one explanation may be elevated HAT activity and increased adding of acetyl groups to histones, whereas another explanation may be lower HDAC activity and decreased removal of acetyl groups of histones. In our model of a female-specific visceral hypersensitivity, a likely mediator of these changes in acetylation might be cycling estrogen. We have previously shown that cycling estrogen is essential for maintaining the visceral phenotype.³⁹ Several studies have indicated that estradiol infusions in the hippocampus can reduce HDAC expression and increase HAT activity.^40,41 Therefore, we postulate that cycling estrogen might be responsible for the increased H3K9 acetylation in the CeA of female rats, previously exposed to ELS. In an attempt to elucidate the gene-specific epigenetic mechanism in the CeA, we found that both predictable and unpredictable ELS elevated H3K9 acetylation at the GR promoter in the CeA of adult females, whereas only unpredictable ELS also led to increased H3K9 acetylation at the CRH promoter. Acetylated H3K9 is a marker for transcriptional activation,^42,43 providing support for the increases in GR and CRH mRNA levels that we reported previously.¹² Our data also revealed that GR binding at the CRH promoter was reduced in the CeA of adult females after unpredictable ELS. In vitro studies have shown that GR attracts HDACs to the CRH promoter which silence CRH gene expression.^44,45 In contrast, when GR is lacking at the CRH promoter, the appropriate HDAC may not be recruited and deacetylation will not take place. As a result, HATs may acetylate H3K9 residues and promote CRH gene expression. Alterations in epigenetic mechanism in response to ELS have also shown that visceral hypersensitivity induced by maternal separation in rats develops through decreases in global H4K12 acetylation in the lumbosacral spinal cord. Moreover, intraperitoneal administration of the histone deacetylase (HDAC) inhibitor suberanilohydroxamic acid (SAHA) reversed the visceral phenotype by increasing in H4K12 acetylation.¹⁵

Interestingly, apart from being sexually dimorphic, our current findings show that the epigenetic mechanism that is activated during neonatal ELS to induce visceral hypersensitivity might be also distinctly different from the mechanism observed to underlie visceral hypersensitivity in response to adult stress. However, an important caveat is the fact that the epigenetic mechanisms underlying adult chronic stress have only been studied in male rats. Future studies aim to investigate whether epigenetic mechanisms are sexually dimorphic. After chronic stress in adult male rats, we observed a global decrease in histone acetylation and specifically at the GR promoter. The decrease in GR expression was mediated through increased HDAC binding at the GR promoter. Administration of HDAC inhibitors in the CeA ameliorated visceral hypersensitivity after chronic stress.⁴⁶ Taken together, our findings might point toward two potentially different epigenetic mechanisms in the CeA that regulate visceral hypersensitivity after unpredictable ELS and chronic adult stress.

In the current study, we next applied a pharmacological approach of directly administering a HAT inhibitor into the CeA in an attempt to reverse the epigenetic mechanism. With this stereotaxic approach, we are able to limit the effects of the inhibitor to the CeA rather than using intracerebroventricular administration which would also affect other brain regions in the stress-pain matrix. We have previously reported that stereotaxic treatment of the CeA with a CRH antisense oligodeoxynucleotide ameliorated visceral hypersensitivity after unpredictable ELS.¹² In the current study, stereotaxic administration of the HAT inhibitor garcinol directly into the CeA attenuated visceral hypersensitivity in adult female rats that underwent neonatal unpredictable ELS. Although garcinol is a non-specific HAT inhibitor, in vitro studies have shown that it is a more potent inhibitor of PCAF than p300/CBP.⁴⁷ Interestingly, PCAF mainly acetylates residues in the tail of H3 and might have been involved in the observed elevated H3K9 acetylation after ELS. Nevertheless, as the exact HATs involved in ELS-induced changes in histone acetylation are still unknown and specific HAT inhibitors are unavailable, we opted for a non-specific inhibitor in order to show a proof-of-principle. After garcinol infusions, H3K9 acetylation at the CRH promoter was reduced in the CeA of adult female rats, previously exposed to unpredictable ELS. Moreover, GR binding at the CRH promoter was also increased. Our data suggest that HAT inhibition is required to restore the binding capacity of GR at the CRH promoter. As a result of decreased transcriptional activity at the CRH promoter, we assume CRH gene expression was reduced in the CeA, which ameliorated visceral hypersensitivity in these rats. Although it is possible that garcinol may also have effects on neuroprotection and microglial inflammation,^48,49 we consider that such effects are unlikely to have led to the decreases in H3K9 acetylation at the CRH promoter observed in the current study. It seems unlikely that the initial stressor, which occurred during early life, triggered or sustained neuronal death or microglial activation in adulthood, when the garcinol administration took place, and the attenuation of the visceral phenotype was observed. If ELS had any effect on neuronal death or microglial activation, it would likely have occurred before adulthood, and garcinol administration adulthood would have been unable to reverse these changes.

We have reported before and we confirm again in the current study that in adult male rats, exposure to predictable and unpredictable ELS did not lead to visceral hypersensitivity. The resilience, observed in male rats, may partially be explained through the lack of increased H3K9 acetylation in the CeA. It is plausible that ELS did not disturb the balance between HAT and HDAC activity in the CeA of male rats. In our previous work, we showed that after ELS, GR, and CRH mRNA expression was unchanged in adult male rats.¹² Because we did not observe changes in global H3K9 acetylation combined with the knowledge that GR and CRH expression was not altered, we decided that further investigations of H3K9 acetylation at the GR and CRH promoter of male rats were not warranted. However, the possibility that ELS did not disturb the epigenetic regulation of other genes in the CeA of male rats cannot be eliminated. However, if these genes contributed to phenotypes other than visceral hypersensitivity, they were out of the scope of the current study.

In summary, our results provide novel insights in the epigenetic regulation of the CeA after ELS. Given the important role in the pain matrix, epigenetic dysregulation in the CeA led to visceral hypersensitivity in female rats. Exposure to ELS increased histone acetylation in the CeA of adult female rats. Unpredictable ELS increased histone acetylation and a decreased GR binding at the CRH promoter. The resulting increase in pro-nociceptive gene expression caused the visceral pain phenotype. Through the stereotaxic administration of a HAT inhibitor in the CeA, we were able to modify the epigenome in one specific region of the pain matrix, which ameliorated visceral hypersensitivity. Our data show that even long-lasting epigenetic modifications are dynamic and potential targets for treatment of visceral pain. Hence, reversing ELS-induced epigenetic marks may become a potential treatment option for chronic pain symptomatology.

Abbreviations:

CeA

central nucleus of the amygdala

ChIP

chromatin immunoprecipitation

CORT

corticosterone

CRD

colorectal distention

CRH

corticotrophin-releasing hormone

ELS

early life stress

GR

glucocorticoid receptor

H3K9

histone 3 lysine 9

HAT

histone acetyl transferase

HDAC

histone deacetylase

HPA

hypothalamic-pituitary-adrenal

IBS

irritable bowel syndrome

MR

mineralocorticoid receptor

PND

postnatal day

qPCR

quantitative polymerase chain reaction

VMR

visceromotor response