**1. Introduction**

In the United States, women typically enter perimenopause in their mid-to-late forties and reach menopause at the average age of 51 [1]. Decline in ovarian 17β-estradiol (E2) levels during the menopause transition results in uncomfortable vasomotor symptoms (e.g., hot flushes) that affect up to 80% of women and can persist for 5–13 years [2–7]. In humans and rodents, the loss of E2 has also been associated with cardiometabolic disease and bone loss [5,6,8] due to reduced gu<sup>t</sup> barrier integrity leading to chronic low-grade inflammation [5,6,9,10]. During the menopause transition, hormone replacement therapy (HRT) is typically prescribed to relieve vasomotor symptoms and HRT was initially believed to mitigate chronic disease [11]. After several clinical studies, HRT appears to be most useful for managemen<sup>t</sup> of vasomotor symptoms in women under 60 with reported benefits for the cognitive function and decreased risk of bone fracture from osteoporosis [12]. HRT dose and treatment regimen should be individualized according to medical history [13] and the lowest effective dose is generally recommended for no more than 5 years to reduce chronic disease risk [12]. HRT use in women over 60 is associated with increased risk of cancer, heart disease, and stroke [12] leaving this population with few options for symptom managemen<sup>t</sup> and chronic disease prevention.

**Citation:** Bacalia, K.M.A.; Tveter, K.M.; Palmer, H.; Douyere, J.; Martinez, S.; Sui, K.; Roopchand, D.E. Cannabidiol Decreases Intestinal Inflammation in the Ovariectomized Murine Model of Postmenopause. *Biomedicines* **2023**, *11*, 74. https://doi.org/10.3390/ biomedicines11010074

Academic Editor: Wesley M. Raup-Konsavage

Received: 1 December 2022 Revised: 21 December 2022 Accepted: 23 December 2022 Published: 28 December 2022

**Copyright:** © 2022 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

Cannabidiol (CBD) is a non-psychotropic phytocannabinoid derived from the industrial hemp plant (*Cannabis sativa* L.). In preclinical studies, CBD was shown to reduce inflammation [14–16], improve gu<sup>t</sup> barrier integrity [17], and protect against bone loss [18–20]. Currently, CBD (i.e., Epidiolex®) is a FDA-approved for treatment of epilepsy-related disorders in both children and adults [21–24]. Although there are currently no studies for the effectiveness of CBD for menopausal or postmenopausal symptoms, a recent survey of 258 perimenopausal and postmenopausal women reported that 86% used medical cannabis and 79% endorsed its use to relieve menopause symptoms [25]. The effects of perorally administered CBD (25 mg/kg/day, 5 days/week for 18 weeks) or vehicle (VEH) treatment (i.e., sesame oil and peanut powder) in the ovariectomized (OVX) mouse model of postmenopause along with sham surgery (SS) controls were recently investigated [26]. Compared to VEH-treated OVX mice, the CBD-treated OVX group had improved oral glucose tolerance, increased energy expenditure, improved bone phenotypes, and the decreased markers of inflammation in bone and intestinal tissues [26]. CBD-treated OVX and SS mice had altered gu<sup>t</sup> microbial communities and BA profiles [26]. Notably, the CBD-treated OVX group, but not the CBD-treated SS group, developed a bloom in *Lactobacillus* species [26]. Studies sugges<sup>t</sup> that probiotic *Lactobacilli* bacteria improve gu<sup>t</sup> barrier integrity leading to the lower levels of inflammatory cytokines in the gut, circulation, and bone [27–31]. A recent meta-analysis of randomized controlled clinical studies cautiously concluded that supplementation with probiotics could increase lumbar bone mineral density in postmenopausal women [32]. BAs modulate gu<sup>t</sup> bacteria [33], inflammation [34], glucose metabolism [35], and bone turnover [36]. Compared to SS groups, several ileal BAs were increased in the VEH-treated OVX group while in CBD-treated OVX mice these BAs were normalized to levels comparable to the SS groups [26].

Oral CBD bioavailability is 6% and increases 4-fold if consumed with fat [37]. Since most ingested CBD would be in the intestinal tract, CBD may improve metabolic health by directly modulating the gu<sup>t</sup> microbiota and/or decreasing intestinal inflammation. In this present study, samples collected from a previously described murine study [26] was used to investigate the response of the colonic transcriptome to CBD or VEH treatment in OVX and SS mice. Targeted metabolomics was performed in order to profile BAs in colon content and liver tissue. Finally, the direct effect of CBD on intestinal inflammation was tested in ileal gu<sup>t</sup> organoids.

#### **2. Materials and Methods**

#### *2.1. Chemicals and Materials*

Experiments were performed using Optima LC-MS grade formic acid, methanol, acetonitrile, and water purchased from Fisher Scientific (Hampton, NH, USA). Oasis Prime HLB 1cc Cartridges (30 mg) were purchased from Waters (Milford, MA, USA) for filtration of phospholipids from liver samples. Corning Costar Spin-X centrifuge tubes with 0.22 μm nylon membrane (Corning, NY, USA) were used to filter BA extracts prior to injecting. CBD isolate was purchased from Bluebird Botanicals (Louisville, CO, USA). Twelve unconjugated, 9 taurine conjugated, and 6 glycine conjugated BAs were ordered from Sigma-Aldrich Inc. (St. Louis, MO, USA), Cayman Chemical (Ann Arbor, MI, USA), or Steraloids Inc. (Newport, RI, USA). Deuterated internal standards of BAs were purchased from Cayman Chemical. BA and CBD information are detailed in Table S1.

#### *2.2. Animals and Treatment*

Animal study protocols were approved by Rutgers institutional animal care and use committee under protocol# PROTO201900041. The details of the mouse study were previously described [26]. Briefly, at age 12 weeks, female wild-type C57BL/6J mice were ovariectomized (OVX) or underwent sham-surgery (SS). After recovery, at age 14 weeks, OVX and SS mice were subdivided to receive either CBD isolate or vehicle (VEH) treatment (*n* = 10 mice/group) for 18 weeks (5 days per week excluding weekends). A calculated volume of VEH (sesame oil) or CBD (25 mg/kg, dissolved in sesame oil, 10 mg/mL) was

mixed with 100 mg of powdered peanut butter to prepare individual peroral doses, which were consumed within 1 min of offering. Mice were euthanized by CO2 asphyxiation and liver, and intestinal tissues were collected as previously described [26]. Individual colon segments were flushed with ice-cold, sterile PBS (pH 7.4) to collect luminal contents. Samples were placed in cryogenic tubes, snap frozen in liquid nitrogen, and stored at −80 ◦C until analysis.

#### *2.3. Ribonucleic Acid Sequencing (RNA-Seq)*

Total ribonucleic acid (RNA) from colon tissue (*n* = 4 mice/group) was extracted using RNeasy Plus Universal Mini Kit (Catalog#73404, QIAGEN, Germantown, MD, USA). The concentration, quality, and integrity of total RNA was determined using a Nano Drop 2000 spectrophotometer (Thermo Scientific, Wilmington, DE, USA) and RNA samples were provided to Azenta (South Plainfield, NJ, USA) for Standard RNA-Seq processing. Pairedend sequencing was performed using Illumina HiSeq 2 × 150 bp paired-end configuration yielding 124,739 Mbases and 415,802,210 reads.

#### *2.4. RNA Sequencing Data Analysis*

Raw reads were pre-processed using FastQC 0.11.9 and Java 14.0.1 software. Phred quality scores of individual sequences were checked before merging paired-end sequences with 20 bp overlap. Trimmomatic-0.39 data analysis software was used to trim adapters of paired-end sequences and assess sequence quality. HISAT2 2.1.0 was used to map reads to a reference mouse genome (grcm38 with Ensembl annotation v38.102) [38]. Differentially expressed genes (DEGs) were analyzed using DESeq2 in R (R Studio v4.2.0) [39]. Gene ontology (GO) enrichment analysis was performed using ShinyGO v0.76.3 (South Dakota State University, Brookings, SD, USA) [40] with the application of FDR correction to generate the list of pathways affected using gene ontology biological processes (GOBP) and Kyoto Encyclopedia of Genes and Genomes (KEGG) databases. The FDR q-value cut-off criteria were assigned as 0.05. The inflammatory response pathway genes were identified in the GOBP network while the bile secretion pathway genes were found in the KEGG network.

PCA plots to compare similarity/difference among RNA-Seq datasets for biological replicates within group and between groups were generated using Metaboanalyst 5.0 [41]. The fold change of DEGs was converted to logCPM (Z-score) values and heat maps of DEGs for each pathway were generated using Euclidean clustering between samples and genes using Origin Pro 2023 software (Origin Lab Corp., Northampton, MA, USA). Venn diagrams were also generated using Origin Pro.

#### *2.5. LC-MS Analysis of Bile Acids*

#### 2.5.1. Preparation of Liver and Colon Samples

BAs were extracted from individual liver samples (n = 9–10/group). Frozen liver tissue was sectioned on dry ice and 50–60 mg of median lobe was transferred into 2 mL bead beating tubes with 4 stainless steel beads (2.8 mm, GBSS 089-5000-11, OPS Diagnostics, Lebanon, NJ, USA) and 300 μL of water. Samples were homogenized using a 1600 MiniG® (SPEX SamplePrep, Metuchen, NJ, USA) for 4 min. Homogenized tissue was transferred to microcentrifuge tubes with 300 μL of pre-dried deuterated internal standards (TCA-d4, DCA-d4, CDCA-d4, and GCDCA-d4 at 1 μg/mL each). Protein was precipitated with 99.9% acetonitrile and 1% formic acid (800 μL) and vortexed for 30 s and then placed on an orbital shaker for 1 h at 4 ◦C. Samples were centrifuged at 13,000× *g* for 10 min at 4 ◦C and supernatant was transferred to glass scintillation vials. Pellets remaining after centrifugation was resuspended in 1 mL 80% methanol and sonicated for 1 min with a Qsonica sonicator Q700 with chiller fitted with cuphorn and 8-tube holder (Cole-Palmer, Vernon Hills, IL, USA); 55% AMP, 30 s on, 59 s off, then a final 30 s on). Sonicated samples were centrifuged at 16,000× *g* for 20 min at 4 ◦C and supernatants were pooled and dried under speed vacuum at room temperature overnight. Samples were resuspended

in 300 μL 50% methanol, placed on orbital shaker for 30 min, vortexed for 2 min, and filtered through Corning® Costar® Spin-X® microcentrifuge tube filters (nylon membrane, pore size 0.22 μm, cat#CLS8169-200ea, Sigma-Aldrich, Darmstadt, Germany) for 5 min at 16,000× *g*. Samples were transferred into sampler vials (Cat# 6PSV9-1PSS Thermofisher" Walthm, MA, USA) with 300 μL inserts (9 mm, C4010-630 Thermofisher, Walthm, MA, USA) for HPLC analysis. Concentrations (μg/mg tissue) were determined by dividing final concentrations by tissue weights used for extraction. For each liver sample, the limit of detection (LOD), the limit of quantification (LOQ), and the coefficient of variance (CV) are presented in Table S2. Recoveries ranged from 66–151% for TCA-d4, 35–76% for DCA-d4, 70–102% for GCDCA-d4, and 53–113% for CDCA-d4.

Individual colon content samples (*n* = 8–9/group) were collected into microfuge tubes by flushing the lumen of the colon with 1× PBS (pH 7.4). The colon content was freezedried in a FreeZone 1.0 L Benchtop lyophilizer (model# 7740020, LABCONCO, Kansas City, MO, USA) overnight to evaporate PBS and dry weight (mg) of colon content was recorded. To subtract weight contributed by salts in PBS, 3 tubes containing 1 mL of 1× PBS were freeze-dried and their mean weight was subtracted from dried colon content weights. Deuterated internal standards (TCA-d4, DCA-d4, GCDCA-d4, and CDCA-d4 at 1 μg/mL each) were resuspended in 50% methanol and 300 μL was added to pre-weighed microfuge tubes and dried in speed vacuum (CentriVap concentration system with cold trap, Model 7810014 and 7460020 Labconco, Kansas City, MO, USA), after which colon content (15–30 mg) was added. Then 600 μL of 90% acetonitrile/9.9% water/0.1% formic acid (*v*/*v*/*v*) was added to internal standards and dry colon content, vortexed for 1 min, and left on a benchtop shaker at 4 ◦C for 1 h. After extraction, samples were centrifuged at 15,000× *g* for 10 min and the supernatant was collected into a clean microfuge tube. For the second round of extraction, 700 μL of 50% methanol/50% water (*v*/*v*) was added to the pellet, vortexed for 2 min, and extract was placed in QSonica sonicator Q700 (with chiller fitted with cuphorn and 8-tube holder, Cole-Palmer, Vernon Hills, IL, USA) at 65% amplitude for 2 min. Samples were placed on shaker at 4 ◦C for 45 min and then centrifuged at 12,000× *g* for 10 min. Supernatants were transferred to microfuge tubes and extraction was repeated with 600 μL of 90% acetonitrile and 0.1% formic acid. Samples were vortexed for 30 sec and placed in QSonica sonicator at 65% amplitude for 1 min. Samples were placed on shaker at 4 ◦C for 45 min and then centrifuged at 12,000× *g* for 10 min. Supernatants from first and second extractions were pooled and solvent was evaporated to dryness using a speed vacuum (CentriVap concentration system with cold trap, Model 7810014 and 7460020 Labconco, Kansas City, MO, USA) and resuspended in 300 μL of 50% methanol. Samples were sonicated at 65% amplitude for 1 min and then filtered using 0.2 μm filters (Corning Costar Spin-x centrifuge tube filters, cat#CLS8169-200ea, Sigma-Aldrich, Darmstadt, Germany). Filtrates were centrifuged at 12,000× *g* for 10 min and transferred to HPLC vials (6PSV9-1PSS, Thermofisher, Waltham, MA, USA) fitted with 300 μL inserts (9 mm, C4010-630, Thermofisher, Waltham, MA, USA).

For each colon content sample, the limit of detection (LOD), the limit of quantification (LOQ), and the coefficient of variance (CV) are presented in Table S2. Recoveries ranged 94–156% for TCA-d4, 63–124% for DCA-d4, 47–131% for GCDCA-d4, and 75–175% for CDCA-d4.
