Saturated Fatty Acids and Omega-3 Polyunsaturated Fatty Acids Improve Metabolic Parameters in Ovariectomized Female Mice

Abstract

In menopausal and postmenopausal women, the risk for obesity, cardiovascular disease, osteoporosis, and gut dysbiosis are elevated by the depletion of 17β-estradiol. A diet that is high in omega-6 polyunsaturated fatty acids (PUFAs), particularly linoleic acid (LA), and low in saturated fatty acids (SFAs) found in coconut oil and omega-3 PUFAs may worsen symptoms of estrogen deficiency. To investigate this hypothesis, ovariectomized C57BL/6J and transgenic fat-1 mice, which lower endogenous omega-6 polyunsaturated fatty acids, were treated with either a vehicle or estradiol benzoate (EB) and fed a high-fat diet with a high or low PUFA:SFA ratio for ~15 weeks. EB treatment reversed obesity, glucose intolerance, and bone loss in ovariectomized mice. fat-1 mice fed a 1% LA diet experienced reduced weight gain and adiposity, while those fed a 22.5% LA diet exhibited increased energy expenditure and activity in EB-treated ovariectomized mice. Coconut oil SFAs and omega-3 FAs helped protect against glucose intolerance without EB treatment. Improved insulin sensitivity was observed in wild-type and fat-1 mice fed 1% LA diet with EB treatment, while fat-1 mice fed 22.5% LA diet was protected against insulin resistance without EB treatment. The production of short-chain fatty acids by gut microbial microbiota was linked to omega-3 FAs production and improved energy homeostasis. These findings suggest that a balanced dietary fatty acid profile containing SFAs and a lower ratio of omega-6:omega-3 FAs is more effective in alleviating metabolic disorders during E2 deficiency.

Decline of ovarian estrogens, especially 17β-estradiol (E2), occurs in women who experience natural physiological transition from perimenopause to menopause or in women who undergo surgical removal of reproductive organs at earlier ages (1). E2 deficiency results in greater risks for obesity, osteoporosis, cardiovascular diseases, and gut dysbiosis (2‐4). Hormone replacement therapy (HRT) has been used as the standard care to mitigate menopausal and postmenopausal symptoms (5). HRT is generally recommended for the first 5 years of postmenopause to alleviate bone loss and menopausal symptoms such as hot flashes and night sweats (6). Use of HRT for more than 5 years or in women older than 65 years is not advised because of elevated risks of breast cancer, thromboembolic events, cholecystitis, and deficits in cognitive functions (5, 7). Options with fewer potential harms are needed to improve health outcomes in E2-deficient women.

Postmenopausal women exhibit higher fat mass and fat percentage than premenopausal women (8). Although body mass index are similar, postmenopausal women have higher risk of developing abdominal obesity. Excessive abdominal fat is linked with unfavorable metabolic alterations such as glucose intolerance, insulin resistance, and hyperlipidemia (9). The caloric energy from dietary fatty acids (FAs) cannot explain the growing obesity epidemic because FAs' profile has been dramatically changed (10). Daily consumption of omega-6 polyunsaturated FA (PUFA), in particular linoleic acid (LA), has replaced saturated FAs (SFAs) and accounts for approximately 6% of total dietary energy (11). A high-fat diet (HFD) rich in LA has been proved to be more obesogenic and impair blood glucose disposal compared with HFD rich in coconut oil (12). Because coconut oil is composed of 93% SFAs, it has a unique profile of 10% to 15% caprylic (C8:0), 10% to 15% capric (C10:0) acid, and 45% to 50% lauric acid (C12:0) (13). These medium-chain SFAs are more readily absorbed in the proximal small intestines and then converted into energy in the liver rather than being deposited as adiposity (14). Coconut oil has been associated with improved osteoporosis and gut microbial profile in rats (15). A beneficial role of dietary supplementation of omega-3 FAs have been implicated in alleviation of depression, cognitive decline, and cardiovascular health in postmenopausal women (16, 17). Human beings can only synthesize a very small amount of omega-3 FAs and mainly rely on dietary consumption of fish or fish oil. Western diets, containing excessive LA, are insufficient in omega-3 FAs with the ratio of omega-6/omega-3 as 15/1 to 16.7/1 (18, 19). Lower omega-6/omega-3 ratio (2-3/1) has been associated with suppressed inflammation in patients with rheumatoid arthritis (20). Omega-6 and omega-3 FAs have differential effects on brain-gut-adipose axis and inflammation, which eventually affect adipocyte differentiation and fat accumulation (21, 22). Scientific evidence of a balanced ratio of dietary omega-6/omega-3 and omega-6/omega-3/SFAs is needed for health improvement in E2-deficient status.

The transgenic fat-1 mouse carries the Caenorhabditis elegans fat-1 gene encoding omega-3 FAs desaturase and can produce omega-3 FAs from omega-6 FAs (23). Endogenous production of omega-3 FAs in fat-1 model can avoid confounding issues such as dose and duration of treatment. In this study, we used ovariectomized wild-type (WT) and fat-1 mice fed HFDs with varying amounts of coconut oil SFAs or omega-6 LA to study how dietary FAs, including coconut oil medium-chain SFAs, omega-6, and omega-3 FAs, differentially affect metabolic biomarkers in association with the gut microbial changes.

Materials and Methods

Animals

Animal studies were performed in accordance with protocols approved by Rutgers University Institutional Animal Care and Use Committee. WT C57BL/6 mice and transgenic fat-1 mice on the C57BL/6 background (purchased from Jackson Laboratory, Bar Harbor, ME; stock #020097) were selectively bred in-house and maintained under controlled conditions (23 °C and 12/12 hours light/dark cycle) with free access to food and water ad libitum (Lab Diet 5V75). All WT and fat-1 female mice, as defined by an estrous cycle and ovaries, were weaned and ear-tagged at postnatal day 21. At age 12 weeks, female mice underwent ovariectomy (OVX) via dorsal incision, as previously described (12).

Steroid Treatment and Diets

Steroid treatment and dietary intervention started at age 13 to 14 weeks approximately one week after recovery from OVX. Both WT and fat-1 mice were perorally administered with vehicle (Veh) as sesame oil or estradiol benzoate (EB, 300 µg/kg) by mixing Veh or EB sesame oil with dried peanut butter every other day until euthanasia (24). HFDs (45% kcal fat) purchased from Research Diets (New Brunswick, NJ, USA) included: (1) an HFD rich in coconut oil SFAs (1% LA, 31% SFA), referred to as 1% LA diet and (2) a HFD rich in omega-6 PUFA LA (22.5% LA, 8% SFA), referred to as 22.5% LA diet. See Supplementary Table 1 (25) for nutritional composition of experimental diets. Veh- and EB-treated WT and fat-1 mice were fed with 1% LA diet or 22.5% LA diet for 14 to 15 weeks. The 8 experimental groups were: WT-Veh-1% (n = 12), WT-EB-1% (n = 10), WT-Veh-22.5% (n = 10), WT-EB-22.5% (n = 10), fat-1-Veh-1% (n = 10), fat-1-EB-1% (n = 10), fat-1-Veh-22.5% (n = 10), and fat-1-EB-22.5% (n = 10).

Metabolic Phenotyping

Body composition

Weekly body weight and food intake was monitored for the first 8 weeks of treatment (Veh or EB) and dietary intervention. Body composition was measured at week 9 on diets using an EcoMRI 3-in-1 Body Composition Analyzer.

Indirect calorimetry

During week 9 of treatment and dietary intervention, the Oxymax Comprehensive Lab Animal Monitoring System (Oxymax/CLAMS, Columbus Instruments, Columbus, OH, USA) was used to assess indirect calorimetry and locomotor activities during a 72-hour run, including 24-hour acclimation and 48-hour data collection. As previously described (26), measurements of volume of oxygen consumption and carbon dioxide production were used for calculation of respiratory exchange ratio and energy expenditure (EE). Physical activity was evaluated by wheel running and total and ambulatory infrared-beam interruption. The CLAMS system was configured with triple axis (X-, Y-, and Z-) detection of animal movement, determined as number of times the infrared photocell beams are interrupted in each of the 3 planes. X-total, Y-total, and Z-total indicate the total number of beam interruptions in each plane, whereas X-ambulatory and Y-ambulatory indicate the number of beam interruptions when the mouse traverses the cage. Subtraction of ambulatory counts from the total counts indicates the number of beam interruptions because of repetitive movements such as grooming and scratching.

Hemodynamic measurements

During week 10 of treatment and dietary intervention, noninvasive tail-cuff CODA instrument (Kent Scientific, Torrington, CT, USA) was used to evaluate the effects of EB and different dietary fatty acids on hemodynamic measurements, including systolic, diastolic, mean blood pressure, and heart rate. Mice were acclimated to the mouse restrainer and measuring method for 20 to 30 minutes per day for 4 consecutive days. One the test day (day 5), 10 cycles for acclimation followed by 20 cycles were performed for data collection. Measurements were determined by averaging more than 3 values deemed acceptable by the instrument for each run (27).

Insulin tolerance test

An insulin tolerance test was performed during weeks 12 and 13 of the treatment and dietary intervention as previously described (12). Mice were fasted for 4 hours (9 Am to 1 Pm) with ad libitum access to water and then IP injected with 0.75 units/kg insulin in 0.9% saline solution (Humulin R; Lilly, Indianapolis, IN, USA). Blood samples were collected from tail pricks before injection (t = 0 minutes) and 15, 30, 60, 90, and 120 minutes after injection.

Oral glucose tolerance test

After recovery from insulin tolerance test (~5 days), mice were fasted for 5 hours with ad libitum access to water in clean cage (9 Am to 2 Pm) and perorally gavaged with glucose (2 g/kg body weight) as detailed previously (26). Blood glucose measurements were collected from tail pricks at baseline (t = 0 minutes), and then 15, 30, 60, 90, and 120 minutes after glucose challenge. After a timepoint of 120 minutes, all mice were returned to their home cages.

Tissue Collection

Mice were euthanized at 27 to 28 weeks of age after sedation with ketamine (100 µL of 100 mg/mL) and 2 hours fasting. Tissues including liver, intestinal segments, and gonadal white adipose tissue (GWAT) were harvested, snap frozen in liquid nitrogen, and stored at −80 °C freezer for future analysis. The brain was removed from the skull and the basal hypothalamus was cut using a brain slicer (Ted Pella, Inc., Redding, CA, USA) into 1-mm thick coronal rostral and caudal blocks corresponding to plates 42 through 47 and plates 48 through 53, respectively (12). The rostral and caudal parts of the arcuate nucleus (ARC) were dissected from slices under a dissecting microscope. Dissected ARC tissue was stored at −80 °C until gene expression analyses. Mouse carcasses were stored frozen at −80 °C until bone densitometry analyses.

Bone Densitometry

Whole body and individual bone samples including femur, tibia, humerus, and lumbar spine (L1-L5) were used for assessing areal bone mineral density (BMD) and bone mineral content (BMC) using dual-energy X-ray absorptiometry (GE-Lunar PIXImus mouse densitometer software version 2.10.41) as detailed previously (28).

Hepatic, GWAT, and Arcuate Nucleus Quantitative PCR

Total RNA was extracted from liver and GWAT tissues using NucleoSpin RNA extraction Kits (Macherey-Nagel, Inc.) and from arcuate nucleus using RNAqueous Micro RNA Extraction Kits (Ambion/ThermoFisher) followed by quantification and qualification using Nanodrop (Thermo Fisher Scientific Inc.) and RNA 6000 Nano Kit (Agilent Technologies, Santa Clara, CA, USA), respectively. As previously described (29), cDNA was synthesized from 500 ng of total RNA and RT-quantitative (q)PCR was performed on a CFX-Connect Real-time PCR instrument (BioRad, Inc.) using PowerSYBR Green master mix (Life Technologies) or Sso Advanced SYBR Green (BioRad, Inc.). Primers (Supplementary Table 2 (25)) were designed using Clone Manager 5 software (Sci Ed Software, Cary, NC, USA) and synthesized by Life Technologies, Inc or purchased as primer assays from BioRad (Acadl, Acadm, Cpt1a, Aox1). The qPCR conditions were as follows: at 95 °C for 10 minutes (PowerSYBR) or 3 minutes (Sso Advanced) followed by 40 cycles at 94 °C for 10 seconds (denaturation), 60 °C for 45 seconds (annealing), and completed with a dissociation step for melting point analysis with 60 cycles of 95 °C for 10 seconds, 65 °C to 95 °C (in increments of 0.5 °C) for 5 seconds, and 95 °C for 5 seconds. The Cq geometric mean of reference genes, Hprt, Actb, and Gapdh, were used to calculate the relative gene expression data using the ΔΔCT method.

Ileum Tissue qPCR

Total RNA from ileum tissues (30-35 mg) were extracted using the Qiagen Universal RNA extraction kit and RNA concentrations were quantified by Nanodrop (Thermo Fisher Scientific Inc.). cDNA was synthesized from 5 µg RNA and RT-qPCR was performed (with technical duplicates) on a thermocycler (Quantstudio 3, Thermo Fisher Scientific Inc) using Taqman Fast Universal PCR master mix using conditions: 20 seconds at 95 °C followed by 40 cycles of 95 °C for 1 second (denaturation) and 60 °C for 20 seconds (annealing and extension). TaqMan assay primers (Life Technologies) used are summarized in Supplementary Table 3 (25). Housekeeping gene Hmbs was used to calculate gene expression using the ΔΔCT method.

Histology and Immunohistochemistry

Distal colonic segment with a whole fecal pellet were harvested: the first 4 to 6 cm from the pyloric sphincter was collected. Tissue was fixed overnight at 4 °C in 3% paraformaldehyde and 2% sucrose then embedded in paraffin. For histological analysis, 5-μm intestinal tissue sections were stained with hematoxylin and eosin (Histopathology Core Facility, New Brunswick, NJ, USA). Goblet cells along the crypts and surface of the epithelium were counted from images (5.5×) that were hematoxylin and eosin stained. Goblet cell parameters for each group were determined using an arbitrary (450 × 450 pixels) square region of interest (ROI) as previously described (30, 31). Number of goblet cells within the ROI were recorded. Total length and average width of mucosa analyzed in each ROI was recorded. ImageJ software was used for data acquisition (https://imagej.nih.gov/ij). For histomorphometry, all tissues sections were blindly scored by a board-certified pathologist.

For immunohistochemical studies, intestinal sections were deparaffinized, rehydrated, and subsequently blocked with 10%, 50%, or 100% normal goat serum at room temperature for 2 hours. The tissue sections were then incubated overnight at 4 °C with primary rabbit affinity-purified polyclonal antibodies against structural components of the intestine as described later, or with rabbit IgG (RRID:AB_1940281, ProSci Inc., Poway, CA) as a control. Colon segments were stained for cyclooxygenase-2 (Cox2, 1:200, RRID:AB_2085144, Abcam, Cambridge, MA), mucin-2 (Muc-2, 1:2000, RRID:AB_2888616, Abcam), and inducible nitric oxide synthase (1:500, RRID:AB_2927640, Abcam) expression. Tissue sections were then incubated for 30 minutes with a biotinylated goat anti-rabbit secondary antibody (RRID:AB_916366, Vector Labs, Burlingame, CA). Antibody binding was visualized using a DAB Peroxidase Substrate Kit (Vector Labs). Tissue sections were photographed using the VS120-S5 System (Olympus, Center Valley, PA). For histomorphometry, all tissues sections were blindly scored by a board-certified pathologist. Slide staining was reported using a dual number system (#X#). The first number was the intensity of the stain and the second number was the amount of stain present in the specimen. The intensity was graded on a 1 to 4 scale and the amount indicated as follows: 1, <10%; 2, 11% to 40%; 3, 41% to 60%; and 4, >60%. The total score was obtained by multiplying the 2 numbers.

Gut Microbiota Analysis

Fresh fecal pellets were collected before sacrifice and stored at −80 °C until analysis. Genomic DNA was extracted using QIAmp Power Fecal DNA kit (QIAGEN, Germantown, MD, USA) according to the manufacturer's instructions. The hypervariable region V4 of the 16S rRNA gene was amplified using the 515F and 806R primers (32, 33) and sequenced using the Ion GeneStudio S5 (ThermoFisher Scientific). Primers were trimmed from the raw reads using Cutadapt in QIIME 2. Amplicon sequence variants (ASVs) were obtained by denoising using the dada2 denoise-single command in QIIME 2 (34), with the parameters -p-trim-left 0 -p-trunc-len 215. A phylogenetic tree of ASVs was built using the QIIME 2 commands including alignment mafft, alignment mask, phylogeny fastree, and phylogeny midpoint-root. Taxonomy assignment was performed using the q2-feature-classifier plugin in QIIME 2, based on the silva database (release 132) (35). The data were rarified to 20 000 reads/sample for subsequent analyses.

Overall gut microbiota structure was evaluated using alpha diversity indices (Shannon index, observed ASVs, and Faith's phylogenetic diversity) and a beta diversity distance metric. Beta diversity metrics analysis of variance was determined using ADONIS and 10 000 × permutation analysis in the vegan package within R Studio v.4.1.2 (R Studio Software, Boston, MA, USA). Differences of gut bacteria at the phylum and genus levels were determined by nonparametric Kruskal–Wallis test followed by 2-stage step-up method of Benjamini, Krieger, and Yekutieli with false discovery rate-adjusted P value, q < 0.05.

Statistical Analysis

Data were tested for normality before using parametric or nonparametric tests. The Grubbs' test was used to detect and remove outliers, if any. All data were analyzed using 3-way ANOVA with genotype (WT or fat-1), treatment (Veh or EB), and diet (1% LA or 22.5% LA) as factors followed by Holm-Sidak multiple comparison using GraphPad Prism 9.4 (GraphPad Software, La Jolla, CA, USA) and expressed as mean ± SEM. Values of P < .05 were considered significant.

Results

Omega-3 PUFAs and EB Treatment Attenuate Weight Gain

Starting at age 13 to 14 weeks, ovariectomized WT and fat-1 mice were fed 1% LA diet or 22.5% LA diet and administered with Veh or EB for 14 to 15 weeks until sacrifice. Compared with Veh-treated WT mice fed a 1% LA diet, Veh-treated fat-1 mice fed a 1% LA diet gained less weight during weeks 3 to 8 on diets (Fig. 1A and Supplementary Fig. 1A (25)) and had a lower final body weight (Fig. 1C). This suggests that conversion of LA to omega-3 PUFAs in HFD rich in coconut oil SFAs can protect from obesity. EB treatment significantly alleviated HFD-induced weight gain regardless of genotypes and diets (Fig. 1B and C, Supplementary Fig. 1B (25)). EB treatment reduced liver weights in WT mice but not in fat-1 mice, and liver weights of fat-1 mice fed a 1% LA diet was lower than WT mice fed a 1% LA diet (Fig. 1D). In terms of body composition, EB-treated mice had lower percent fat mass as well as higher percent lean mass compared with Veh-treated mice on the same diet (Fig. 1E and F). Again, Veh-treated fat-1 mice fed 1% LA diet had significantly lower fat accumulation than their WT counterparts (Fig. 1E). In Veh- and EB-treated mice, fat-1 mice fed 1% LA diet consumed less food compared with WT mice fed 1% LA diet from week 5 or 6 on diets, respectively (Fig. 1G and H and Supplementary Fig. 1C-D (25)). Veh-treated fat-1 mice fed 22.5% mice had lower cumulative food intake at weeks 6 to 8 on diets compared with Veh-treated WT mice fed the same diet (Fig. 1G and Supplementary Fig. 1C (25)).

Figure 1.

Body weight, liver weight, body composition and food intake. Cumulative weight gain of (A) Veh-treated and (B) EB-treated groups for weeks 1 to 8 of treatment and dietary intervention. (C) Final body weight, (D) liver weight, (E) percent fat mass, and (F) percent lean mass measured at weeks 14 to 15 on treatment and diets. (G, H) Cumulative food intake of (A) Veh-treated and (B) EB-treated groups for weeks 1 to 8 of treatment and dietary intervention. Data are presented as mean ± SD. Sample sizes were 10 to 12 per group. Data were analyzed by 3-way ANOVA followed by Holm–Sidak multiple comparisons (*P < .05, **P < .01, ***P < .001, and ****P < .0001) between and within genotype, treatment and diet for panels C through F. Two-way ANOVA followed by Holm-Sidak multiple comparisons between and within genotype and diet were used for panels A, B, G, and H. Different letters (b, c) denote significant difference between diet groups under the same treatment, b for WT-1% vs fat-1-1% and c for WT-22.5% vs fat-1-22.5%. Significantly different ANOVA results for panels C-F are: (C) genotype: F(1, 72) = 23.80, P < .0001; diet: F(1, 72) = 7.101, P = .0095; treatment: F(1, 72) = 92.24, P < .0001; (D) genotype: F(1, 70) = 13.40, P = .0005; diet: F(1, 70) = 11.15, P = .0013; treatment: F(1, 70) = 34.42, P < .0001; (E) genotype: F(1, 74) = 11.09, P = .0014; treatment: F(1, 74) = 250.4, P < .0001; genotype × treatment: F(1, 74) = 4.363, P = .0402; (F) genotype: F(1, 74) = 5.875, P = .0178; treatment: F(1, 74) = 214.7, P < .0001; genotype × treatment: F(1, 74) = 4.348, P = .0405.

EB Treatment and Omega-3 PUFAs Regulate Energy Expenditure and Locomotor Activity

EB treatment significantly increased volume of oxygen consumption and volume of carbon dioxide production during day and night phases (Fig. 2A and B). Independent of diet or E2 status, mean respiratory exchange ratio values for all groups ranged from 0.7 to 0.8, indicated that FAs were burned as primary substrate for energy (Fig. 2C). Normalized to body weight, EB treatment increased EE in both day and night phases, regardless of dietary FAs (Fig. 2D). EB-treated WT and fat-1 mice fed 22.5% LA diet showed increased EE compared with their EB-treated counterparts fed 1% LA diet during night phase, suggesting that higher amount of omega-3 PUFAs can improve energy efficiency in the E2-sufficient status (Fig. 2D).

Figure 2.

Indirect calorimetry and locomotor activity. (A) Volume of oxygen consumed (VO₂), (B) volume of carbon dioxide produced (VCO₂), (C) respiratory exchange ratio (RER), (D) energy expenditure (EE) normalized to body weight (kg). (E) X-plane total and ambulatory activity. (F) Y-plane total and ambulatory activity and (G) wheel running. Data are presented as mean ± SD. Sample sizes were 10 to 11 per group. Data were analyzed by 3-way ANOVA followed by Holm-Sidak multiple comparisons (*P < .05, **P < .01, ***P < .001, and ****P < .0001) between and within genotype, treatment and diet for daytime or nighttime. Significantly different ANOVA results are: (A) daytime, diet: F(1, 74) = 6.505, P = .0128; treatment: F(1, 74) = 105.7, P < .0001. Nighttime, F(1, 73) = 6.285, P = .0144; treatment: F(1, 73) = 353.3, P < .0001; (B) daytime, F(1, 74) = 5.719, P = .0193; treatment: F(1, 74) = 92.02, P < .0001. Nighttime, treatment: F(1, 73) = 272.6, P < .0001; (D) daytime, treatment: F(1, 74) = 117.0, P < .0001; diet × treatment, F(1, 74) = 5.495, P = .0218. Nighttime, diet: F(1, 72) = 36.16, P < .0001; treatment: F(1, 72) = 228.5, P < .0001; diet × treatment: F(1, 72) = 4.057, P < .0001. (E) X-total nighttime, genotype: F(1,74) = 4.661, P = .0341; treatment: F(1, 74) = 18.55, P < .0001. X-ambulatory nighttime, genotype: F(1, 73) = 6.683, P = .0117; diet: F(1, 73) = 5.005, P = .0283; treatment: F(1, 73) = 42.24, P < .0001. (F) Y-total nighttime, genotype: F(1, 73) = 4.520, P = .0369; diet: F(1, 73) = 4.677, P = .0338; treatment: F(1, 73) = 36.54, P < .0001. Y-ambulatory nighttime, genotype × treatment: F(1, 69) = 6.187, P = 0.0153; treatment × diet: F(1, 69) = 5.704, P = .0197. (G) genotype: F(1, 74) = 4.158, P = .0450; diet: F(1, 74) = 6.747, P = .0113; treatment: F(1, 74) = 87.24, P < .0001.

Mice are nocturnal and more active during night phase rather than day phase. Compared with Veh-treated counterparts, EB-treated WT mice fed a 1% LA diet had higher X-total and ambulatory activity (Fig. 2E). EB-treated fat-1 mice fed a 1% LA diet had higher X-ambulatory activity compared with Veh-treated counterparts (Fig. 2E). EB treatment elevated Y-total activity in WT mice fed a 22.5% LA diet and fat-1 mice fed a 1% LA diet compared with their corresponding Veh-treated counterparts (Fig. 2F). For EB-treated mice, fat-1 mice fed a 1% LA diet were more active compared with both WT mice fed a 1% LA diet and fat-1 mice fed a 22.5% diet in terms of Y-total activity (Fig. 2F). Veh-treated fat-1 mice fed a 22.5% LA diet exhibited higher Y-ambulatory activity compared with Veh-treated WT mice fed 22.5% LA diet (Fig. 2F). There were no differences in the total number of beam interruptions of Z plane (data not shown). Additionally, EB treatment increased activity as measured by wheel-running in comparison between Veh- and EB-treated mice fed the same diet (Fig. 2G). EB-treated WT mice fed a 22.5% LA diet did more wheel-running compared with EB-treated WT mice fed a 1% LA diet (Fig. 2G).

EB Treatment or Dietary FAs did not Affect Blood Pressure

Using a CODA system, systolic, diastolic, mean blood pressure, and heart rate were measured during week 10 of dietary intervention. Treatment had main effects on systolic blood pressure (F (1, 63) = 5.097, P = .0274) (Supplementary Fig. 2A (25)). The interaction of genotype × diet had main effects on diastolic blood pressure (F (1, 63) = 4.296, P = .0423) and the interaction of genotype × treatment had main effects on heart rate (F (1, 63) = 9.927, P = .0025) (Supplementary Fig. 2B and D (25)). However, post hoc multiple comparisons did not reach significance.

Glucose Tolerance is Improved by EB, Coconut oil SFAs and Omega-3 PUFAs

Glucose and insulin tolerance tests were performed to evaluate the interaction of E2 and different FAs. Compared with Veh-treated mice, EB-treated mice showed improved oral glucose tolerance (Fig. 3A-C). Consistent with higher blood glucose readings at 15 and 30 minutes after oral glucose challenge (Fig. 3A), Veh-treated WT mice fed a 22.5% LA diet had higher blood glucose area under the curve (AUC) compared with Veh-treated WT mice fed a 1% LA diet (Fig. 3C), suggesting that 1% LA diet can improve oral glucose tolerance. Consistent with higher blood glucose readings at 15, 30, and 60 minutes after oral glucose challenge (Fig. 3A), Veh-treated WT mice fed 22.5% LA diet showed higher blood glucose AUC compared with Veh-treated fat-1 mice fed a 22.5% LA diet, suggesting that conversion of omega-6 to omega-3 PUFAs can alleviate glucose intolerance. In both WT and fat-1 groups, EB-treated mice fed 1% LA diet had lower blood glucose AUC compared with Veh-treated counterparts after insulin injection (Fig. 3F), indicating that effectiveness of EB on insulin sensitivity is significant in HFD rich in coconut oil SFAs. Notably, the blood glucose AUC was similar between Veh- and EB-treated fat-1 mice fed 22.5% LA diet, suggesting that omega-3 PUFAs enhance insulin sensitivity in the state of E2 deficiency (Fig. 3F).

Figure 3.

Oral glucose tolerance test (OGTT) and insulin tolerance test (ITT). (A) Blood glucose curve of Veh-treated and (B) EB-treated groups obtained for OGTT and (C) area under the blood glucose curve (AUC). (D) Blood glucose curve of Veh-treated and (E) EB-treated groups obtained for ITT and (F) AUC. Data are presented as mean ± SD. Sample sizes were 10 to 11 per groups. AUC data were analyzed by 3-way ANOVA followed by Holm-Sidak multiple comparison (*P < .05) between and within genotype, treatment, and diet groups. Two-way ANOVA followed by Holm–Sidak multiple comparisons between and within genotype and diet were used for each time point of blood glucose curve. Different letters (a, b, c) denote significant difference between diet groups under the same treatment, a for WT-1% vs WT-22.5%, b for WT-1% vs fat-1-1%, c for WT-22.5% vs fat-1-22.5% and d for fat-1-1% vs fat-1-22.5%. Significantly different results of ANOVA tests of AUC were obtained for: (F) genotype: F(1, 68) = 7.197, P = .0092; treatment: F(1, 68) = 28.88, P < .0001; treatment × diet F(1, 68) = 13.96, P = .0004; (C) genotype: F(1, 72) = 18.82, P < .0001; treatment: F(1, 72) = 149.0, P < .0001; genotype × treatment: F(1, 72) = 3.337, P = .0719; genotype × treatment × diet: F(1, 72) = 6.769, P = .0113.

EB Rescued OVX-induced Bone Loss

Bone quality was assessed after 14 to 15 weeks of EB or Veh treatment and dietary intervention. In WT mice, BMD of individual bones, including femur, tibia, humerus, and spine, increased because of EB treatment, regardless of different FAs although improvement of whole-body BMD did not reach significance (Table 1). Increase in BMC of the femur and humerus was observed in EB-treated WT mice compared with Veh-treated mice fed the same diet, whereas EB-treated WT mice fed 1% LA diet had higher tibia BMC compared with their Veh-treated WT counterparts (Table 1). In fat-1 mice, tibia and humerus BMDs were improved by EB treatment independent of dietary FAs and an increase of spine BMD was observed in EB-treated mice fed a 22.5% LA diet compared with their Veh-treated counterparts (Table 1). Compared with Veh-treated fat-1 mice fed a 1% LA diet, EB-treated fat-1 mice fed a 1% LA diet showed improvement in BMC of individual bones, including the femur, tibia, and humerus (Table 1). The main effects of genotype and interaction of genotype × treatment was observed on BMD and BMC of individual bones, suggesting omega-3 PUFAs attenuate bone loss, although no significant differences were shown in post hoc comparisons (Table 1).

Table 1.

Bone parameters

	Bone mineral density (g/cm3)					Bone mineral content (g)
	Whole body	Femur BMD	Tibia BMD	Humerus	Spine	Whole body	Femur	Tibia	Humerus	Spine
WT-Veh-1%	0.0516 ± 0.005	0.0560 ± 0.003	0.0465 ± 0.002	0.0442 ± 0.001	0.0548 ± 0.003	0.4953 ± 0.129	0.0256 ± 0.001	0.0212 ± 0.002	0.0125 ± 0.001	0.0485 ± 0.003
WT-Veh-22.5%	0.0499 ± 0.002	0.0553 ± 0.002	0.0473 ± 0.002	0.0434 ± 0.002	0.0532 ± 0.003	0.4606 ± 0.073	0.0251 ± 0.000	0.0220 ± 0.001	0.0117 ± 0.001	0.0474 ± 0.007
WT-E2-1%	0.0555 ± 0.003	0.0637 ± 0.005a	0.0515 ± 0.002a	0.0507 ± 0.002a	0.0621 ± 0.005a	0.4990 ± 0.041	0.0293 ± 0.001a	0.0249 ± 0.001a	0.4343 ± 0.002a	0.0546 ± 0.006
WT-E2-22.5%	0.0544 ± 0.003	0.0613 ± 0.003a	0.0510 ± 0.002a	0.0494 ± 0.002a	0.0608 ± 0.004a	0.4651 ± 0.003	0.0279 ± 0.002a	0.0234 ± 0.002	0.0274 ± 0.043a	0.0494 ± 0.005
fat-1-Veh-1%	0.0529 ± 0.004	0.0592 ± 0.003	0.0485 ± 0.001	0.0453 ± 0.001	0.0564 ± 0.003	0.4613 ± 0.06	0.0260 ± 0.002	0.0228 ± 0.002	0.0130 ± 0.007	0.0470 ± 0.005
fat-1-Veh-22.5%	0.0517 ± 0.003	0.0579 ± 0.003	0.0477 ± 0.003	0.0441 ± 0.001	0.0550 ± 0.005	0.4729 ± 0.087	0.0259 ± 0.002	0.0230 ± 0.002	0.0123 ± 0.001	0.0440 ± 0.006
fat-1-E2-1%	0.0557 ± 0.002	0.0631 ± 0.003	0.0528 ± 0.003a	0.0488 ± 0.002a	0.0612 ± 0.005	0.4893 ± 0.022	0.0290 ± 0.001a	0.0206 ± 0.003a	0.0144 ± 0.001a	0.0510 ± 0.007
fat-1-E2-22.5%	0.0549 ± 0.003	0.0618 ± 0.003	0.0527 ± 0.003a	0.0486 ± 0.002a	0.0601 ± 0.004a	0.4739 ± 0.028	0.0277 ± 0.002	0.0240 ± 0.002	0.0134 ± 0.001	0.0480 ± 0.006
Significant ANOVA results	Treatment (P < .0001)	Treatment (P < .0001) Genotype × treatment (P = .032)	Genotype (P = .008) Treatment (P < .0001)	Treatment (P < .0001) Diet (P = .030) Genotype × treatment (P = .006)	Treatment (P < .0001)	−	Diet (P = .030) Treatment (P < .0001)	Genotype (P = .024) Treatment (P < .0001) Treatment × diet (P = .017)	Treatment (P < .0001) Diet (P = .001) Genotype × treatment (P = .012)	Treatment (P = .001) Diet (P = .010)

Values are means ± SD. Data were analyzed by a 3-way ANOVA followed by Holm–Sidak multiple comparison.

Abbreviations: BMC, bone mineral content; BMD, bone mineral density; Veh, vehicle; WT, wild-type.

^a Denotes significant difference between Veh- and E2-treated mice fed the same diet within each genotype (P < .05).

Dietary FAs and EB Treatment Regulated ARC Neuropeptide and Hormone Receptor

To determine the interaction of E2 and different FAs on appetite and energy homeostasis regulated by ARC neurons, gene expression levels of neuropeptides and hormone receptors were assessed. The anorexigenic neuropeptide proopiomelanocortin (POMC), encoded by Pomc, and cocaine-and-amphetamine-regulated transcript (CART), encoded by Cart, are negatively associated with appetite and food consumption (36). EB treatment reduced Pomc expression in WT mice regardless of dietary FAs and decreased Cart expression in fat-1 mice fed 22.5% LA diet (Fig. 4A and B), which was not consistent with the observed reduction in body weight gain and food intake (Fig. 1C, G and H). Reduction of Pomc expression was also observed in fat-1 mice compared with their corresponding WT counterparts (Fig. 4A), which was not consistent with the observed decrease in cumulative food intake in fat-1 mice (Fig. 1G and H). The orexigenic peptides neuropeptide Y (NPY) and agouti-related peptide regulate feeding behavior and promote deposition of triglycerides within adipose tissue (37). In fat-1 mice, EB-treated mice fed a 22.5% LA diet had higher NPY mRNA levels compared with EB-treated mice fed 1% LA diet and Veh-treated mice fed a 22.5% LA diet, separately (Fig. 4C). The opposite change of Cart and Npy gene expression between EB- and Veh-treated fat-1 mice fed a 22.5% LA diet suggests EB along with omega-3 modulate appetite and energy storage (Fig. 4B and C). No significant differences were observed on gene expression of Agrp among all groups (Fig. 4D).

Figure 4.

Arcuate gene expression of neuropeptides and hormone receptors. Regulation of the following genes: (A) Pomc, (B) Cart, (C) Npy, (D) Agrp, (E) Lepr, (F) Ghsr, (G) Insr, (H) Glp1r, (I) Pparg, and (J) Esr1. Data are presented as mean ± SD. Sample sizes were n = 8 per group. Differences determined by 3-way ANOVA followed with Holm–Sidak post hoc comparison (*P < .05) between and within genotype, treatment, and diet groups. Significant ANOVA results are shown under the graph.

The gene expression levels of ARC hormones receptors involved in energy regulation were assessed. Leptin receptor (LEPR), encoded by Lepr, collects and transmits signals from leptin to regulate food intake. Veh-treated WT mice fed a 1% LA diet had higher Lepr expression compared with Veh-treated WT mice fed a 22.5% LA and fat-1 Veh-treated mice fed 1% LA, suggesting that 1% LA diet rich in coconut oil may increase leptin receptor sensitivity in the state of E2 deficiency (Fig. 4E). Ghrelin, primarily secreted by the stomach, binds to GH secretagogue receptor, encoded by Ghsr, to drive food intake in response to hunger. EB treatment augmented Ghsr expression regardless of dietary FAs (Fig. 4F). EB-treated fat-1 mice fed a 1% LA diet had increased level of Ghsr compared with EB-treated WT mice fed a 1% LA diet, whereas EB-treated WT mice fed a 1% LA diet showed reduced Ghsr compared with fat-1 EB-treated mice fed a 1% LA diet, suggesting that addition of omega-3 PUFAs into HFD rich in coconut oil SFAs can enhance Ghsr sensitivity (Fig. 4F). We have recently demonstrated that the increase in Ghsr in the arcuate by E2 is largely driven by expression in arcuate kisspeptin neurons (24, 38). Gene expression of insulin receptor, encoded by Insr, was upregulated by EB treatment in both WT and fat-1 mice fed a 1% LA diet (Fig. 4G), which was consistent with improved insulin sensitivity (Fig. 3F). EB treatment enhanced the gene expression of Glp1r (Fig. 4H), encoding glucagon-like peptide 1 receptor, which binds to glucagon-like peptide 1 to inhibit food intake and limit hepatic glucose production. In addition, EB-treated fat-1 mice fed a 1% LA diet had higher Glp1r mRNA compared with EB-treated WT mice fed a 1% LA diet (Fig. 4H), suggesting omega-3 PUFAs together with coconut oil-rich diet can improve Glp1r expression. Compared with Veh-treated WT mice fed a 1% LA diet, Veh-treated fat-1 mice fed a 1% LA diet had reduced mRNA level of Pparg, encoding peroxisome proliferator-activated receptor gamma, which is associated with decreased weight gain and fat accumulation (39) (Fig. 4I, 1C and D). Estrogen receptor α, encoded by Esr1, plays a pivotal role in regulating metabolism (40). fat-1 mice had significantly reduced mRNA level of Esr1 compared with their WT counterparts on the same treatment and diet (Fig. 4J), whereas Esr2 expression was not altered by any treatment (Supplementary Fig. 6 (25)).

Dietary FAs and/or EB Treatment Altered Markers of Fat and Glucose Metabolism in Liver and GWAT

We next investigated effects of EB treatment and dietary fat on markers of glucose, fatty acids, and lipid metabolism in liver tissues and GWAT. Inhibition of forkhead box transcription factor O1, encoding by Foxo1, reduces hepatic gluconeogenic gene expression (41). In liver, phosphoenolpyruvate carboxykinase and glucose 6-phosphatase are rate-limiting enzymes for gluconeogenesis and glycogenolysis, respectively. Compared with WT mice, fat-1 mice had lower mRNA levels of Foxo1 and Pepck, regardless of EB treatment, indicating that omega-3 PUFAs suppress hepatic gluconeogenesis via inhibition of Foxo1 (Fig. 5A and B); however, there was no difference in G6pc (Fig. 5C) or Srebp1c (Supplementary Fig. 7 (25)) gene expression between groups. mRNA level of insulin receptor, encoding by Insr, was augmented by EB treatment in WT and fat-1 mice fed a 1% LA diet as well as WT mice fed a 22.5% LA diet, suggesting EB treatment increases insulin sensitivity (Fig. 5D). However, fat-1 mice had lower Insr mRNA compared with WT counterparts on the same treatment and diet, suggesting that fat-1 mice may have increased insulin production or blunt insulin sensitivity (Fig. 5D). Fatty acid synthase encoded by Fasn is responsible for the synthesis of long-chain saturated fatty acid. Compared with EB-treated WT mice fed a 22.5% LA diet, EB-treated fat-1 mice fed a 22.5% LA diet had a higher Fasn mRNA level (Fig. 5E). Gene expression of acetyl-CoA carboxylase 1, encoded by Acc1, was decreased in Veh-treated fat-1 mice fed 1% LA diet compared with WT counterparts (Fig. 5G), in association with reduced fat accumulation, whereas gene expression levels of hepatic lipogenesis enzymes such as ATP citrate lyase (Acly) and Acc2 were not regulated (Fig. 5F and H). Diacylglycerol acyltransferase 2, encoded by Dgat2, catalyzes triglyceride synthesis (42). Compared with WT counterparts, fat-1 mice had elevated Dgat2 mRNA, regardless of treatment, suggesting that dietary omega-3 FAs may induce triglyceride production, whereas EB treatment in fat-1 mice induced reduction of Dgat2 mRNA (Fig. 5I). No differences were shown in gene expression of 2 fatty acid transport proteins FATP2 and FATP5, encoded by Fatp2 and Fatp5 (Fig. 5J and K).

Figure 5.

Liver and GWAT tissue quantitative PCR. Gene expression of markers of lipid and glucose metabolism pathways in (A-K) liver samples and (L-O) gonadal white adipose tissue (GWAT) samples. Data are presented as mean ± SD. Sample sizes were n = 8 per group. Significance was determined by 3-way ANOVA followed with Holm–Sidak post hoc comparison (*P < .05) between and within genotype, treatment, and diet groups. Significant ANOVA results are shown under the graph.

Activation of peroxisome proliferator-activated receptor alpha and gamma, encoded by Pparα and Pparγ, has been linked to improved insulin resistance, reduced inflammation, and adipocyte hypertrophy (43, 44). Compared with their WT counterparts, fat-1 mice fed a 1% LA diet had increased mRNA levels of both Pparα and Pparγ in GWAT (Fig. 5L and M). Expression of Pgc1a, a coactivator of PPAR receptors, was only altered by EB in fat-1 females fed the 22.5% LA diet (Supplementary Fig. 8 (25)). In addition, EB-treated fat-1 mice fed a 22.5% LA diet showed higher Pparα mRNA compared with EB-treated WT mice fed a 22.5% LA diet and Veh-treated fat-1 mice fed a 22.5% LA diet (Fig. 5L). However, reduced mRNA of Pparγ was observed in Veh-treated fat-1 mice fed a 22.5% LA diet compared with Veh-treated fat-1 mice fed a 1% LA diet (Fig. 5L). Gene expression levels of Lepr or Lpl, encoding leptin receptor and lipoprotein lipase, were similar between intervention groups (Fig. 5N and O). Examination of adipose genes involved in fatty acid oxidation and metabolism found that EB treatment increased expression of Acadm and Aox1 only in fat-1 females fed the 22.5% LA diet (Supplementary Fig. 8 (25)).

Differential Effects of EB Treatment and Dietary FAs on Markers of Inflammation, Glucose and Lipid Metabolism in Colon and Ileum Tissues

In colon tissues, hematoxylin and eosin staining revealed that morphology was similar among all 8 groups (Supplementary Fig. 3A (25)). Quantified by Alcian blue-periodic acid–Schiff staining, there were no differences in numbers of goblet cells from treatment or dietary FAs (Supplementary Fig. 3B (25)). In both WT and fat-1 groups, EB-treated mice fed a 1% LA diet had higher Cox2 expression than Veh-treated mice on the same diet (Supplementary Fig. 3C (25)). Compared with EB-treated WT mice fed a 22.5% LA diet, EB-treated fat-1 mice fed a 22.5% LA diet had lower expression level of Cox2 (Supplementary Fig. 3C (25)). These data suggest that omega-3 PUFAs can reduce inflammatory in the status of E2 sufficiency.

In ileum samples, relative mRNA levels of genes related to inflammation, gut barrier integrity and glucose metabolism were assessed in ileum samples. Gcg encodes preproglucagon, which is cleaved by PC1/3 protease (encoded by Pcsk1) to yield the incretin peptides Glp1 and Glp2, which regulates intestinal hexose transport (45). Compared with EB-treated WT mice fed a 22.5% LA diet, EB-treated WT mice fed a 1% LA diet had reduced mRNA level of Gcg, but there were no differences in Pcsk1 mRNA (Supplementary Fig. 4 (25)). Compared with Veh-treated WT mice fed a 1% LA diet, both EB-treated WT mice fed a 1% LA diet and Veh-treated WT mice fed a 22.5% LA diet had lower gene expression of carbohydrate transporter (Glut2) (Supplementary Fig. 4 (25)). No changes were observed in the gene expression levels of inflammatory markers such as Il6, Tnf, and Nos2 as well as tight junction protein 1 (Tjp1) and occludin (Ocln) (Supplementary Fig. 4 (25)). Gene expression levels of markers for lipid uptake such as fasting-induced adipose factor, fatty acid receptor CD36, and fatty acid binding proteins 2 and 4 were similar among all 8 groups (Supplementary Fig. 4 (25)).

Gut Microbial Communities Were Altered by EB Status and Dietary FAs

To assess the effects of different dietary FAs and E2 deficiency on fecal microbial communities, 16S rRNA V3-V4 amplicon sequencing was performed. No differences were observed in richness and evenness among all groups (Supplementary Fig. 5A (25)). All groups had similar ASVs, as indicated by Faith's phylogenetic diversity index. Beta diversity analysis was performed to assess the dissimilarity of gut microbial community among samples. Bray-Curtis dissimilarity based on ASV abundance and Jaccard distance based on ASV presence indicated that dietary FAs and EB treatment influenced the separation of gut microbiota along PC1 and PC2. Taking phylogenetic distance between ASVs and their presence/absence into consideration, unweighted unifrac showed 8 groups separated from each other along PC1 and PC2; however, weighted unifrac, accounting for abundance of taxa and phylogenetic distance, showed that only the interaction of dietary FAs and EB treatment affected gut microbial community separation (Supplementary Fig. 5B (25)). There were no differences between groups at the phylum level (data not shown).

Reduction of the relative abundance of Bilophila, Blautia and GCA-900066575 in fat-1 mice were observed compared with their WT counterparts on the same treatment and diet (Fig. 6A-C), which have been reported to inversely link with fat accumulation (46‐48). Compared with corresponding WT counterparts, the relative abundance of Lachnospiraceae UCG-004 and Lachnospiraceae NK4A136 were lower in Veh-treated fat-1 mice fed a 1% LA diet and EB-treated fat-1 mice fed a 22.5% LA diet (Fig. 6D and E), respectively, which is associated with reduced risks of fracture (49). Lower relative abundance of Ruminococcaceae UCG-003 was reported in healthy women (50), and reduction of the relatively abundance of Ruminococcaceae UCG-003 was induced by EB treatment in both WT and fat-1 mice fed 22.5% LA diet (Fig. 6F). As a biomarker of heart failure (51), lower relative abundance of Parasutterella was shown in Veh-treated WT mice fed 22.5% LA diet compared with Veh-treated WT mice fed 1% LA diet (Fig. 6G). Coriobacteriaceae UCG-002 and Eubacterium nodatum, reported as short-chain fatty acid (SCFA) producers (52, 53), were increased because of endogenous omega-3 PUFAs (Fig. 6H and I). However, in WT mice, EB-treatment reduced Coriobacteriaceae UCG-002, regardless of dietary FAs (Fig. 6H). Another SCFA producer (54), Odoribacter maintained higher relative abundance in 3 groups, including Veh-treated WT mice fed a 1% LA diet, EB-treated WT mice fed a 22.5% LA diet, and Veh-treated fat-1 mice fed a 22.5% LA diet (Fig. 6J). It has been reported that the relative abundance of Ileibacterium decreased in omega-3 PUFA-treated HFD-induced insulin-resistant mice (55), whereas we observed that Ileibacterium increased in fat-1 mice (Fig. 6K). Compared with Veh-treated WT mice on the same diet, EB-treated WT mice had reduced relative abundance of Faecalibaculum, which is associated with improved lipid metabolism (56); however, the potential beneficial taxa Dubosiella decreased in the relative abundance of in EB-treated WT mice (Fig. 6M). Barnesiella was shown to be inversely associated with aerobic capacity in OVX mice (57) and was decreased in EB-treated mice is consistent with their higher activity (Fig. 6N and 2E-G). The relative abundance of Mucispirillum was decreased in fat-1 mice fed a 1% LA diet compared with WT mice on the same diet, which is associated with improved mucin degradation and inflammation (58); however, EB-treated fat-1 mice fed a 22.5% LA diet had higher levels of Mucispirillum compared with EB-treated fat-1 mice fed a 1% LA diet (Fig. 6O). Compared with Veh-treated counterparts, decrease in the relative abundance of Clostridium sensu stricto 1 in EB-treated fat-1 mice fed a 1% LA diet (Fig. 6P) is associated with improved dyslipidemia (59).

Figure 6.

Altered relative abundance of gut microbial taxa because of EB treatment and dietary fatty acids. Relative abundance of amplicon sequence variants (ASVs) identified in fecal gut microbiota samples collected after 14 to 15 weeks of dietary intervention (n = 10-11/per group) classified at the genera level. Relative abundance of the following genera: (A) Bilophila, (B) Blautia, (C) GCA-900066575, (D) Lachnospiraceae UCG-004, (E) Lachnospiraceae NK4A136, (F) Ruminococcaceae UCG-003, (G) Parasutterella, (H) Coriobacteriaceae UCG-002, (I) Eubacterium nodatum, (J) Odoribacter, (K) Ileibacterium, (L) Faecalibaculum, (M) Dubosiella, (N) uncultered Barnesiella sp, (O) Mucispriillum, and (P) Clostridium sensu stricto 1. Significant difference was determined by Kruskal–Wallis test followed by the 2-stage step-up method of Benjamini, Krieger, and Yekutieli with false-discovery rate-adjusted P value, *q < 0.05, **q < 0.01.

Discussion

Our results indicate that obesity and glucose intolerance from E2 deficiency can be mitigated by adjusting dietary FA profile. A diet rich in coconut oil SFAs and omega-3 PUFAs rather than omega-6 PUFAs appeared to benefit energy homeostasis and glucose homeostasis with some beneficial alterations to gut bacteria communities. Specifically, we showed that EB treatment can improve energy homeostasis in OVX mice fed HFDs with different FA profiles (Fig. 1B, C, E, and F). Conversion of omega-6 PUFAs to omega-3 PUFAs in HFD rich in coconut oil SFAs conferred greater resilience to E2 deficiency-induced obesity and fat accumulation (Fig. 1A, C, E, and F). A lower ratio of omega-6/omega-3 in HFD enriched with PUFAs increased EE in EB-treated OVX mice and triggered more wheel running in Veh-treated OVX mice (Fig. 2D and G). Except for EB treatment, HFDs rich in coconut oil or with lower omega-6/omega-3 ratio can protect against glucose intolerance (Fig. 3A-C). Apparent improvement of insulin sensitivity was achieved by EB treatment in combination with HFD rich in coconut oil, whereas a FA profile of lower omega-6/omega-3 ratio can protect against insulin resistance without EB treatment (Fig. 3D‐F). Increase of relative abundance of gut microbial taxa as SCFA producers were associated with omega-3 FAs production and improved energy homeostasis (Fig. 6H‐J).

In our previous study, dietary intervention of the 1% LA and 22.5% LA diets had no impact on final body weight in both Veh- or EB-treated OVX mice (12). However, fat-1 mice fed a 1% LA diet showed substantial reduction in final body weight, liver weight, and fat percentage, highlighting that omega-3 plays a pivotal role in metabolic resilience, especially in the status of E2 deficiency (Fig. 1C‐E). Decrease in food intake in EB-treated OVX mice has been linked to estrogen receptor signaling pathways (60, 61). In this study, EB-treated mice consumed similar amounts of food with Veh-treated OVX mice (Fig. 1G and H), whereas assessment of indirect calorimetry and locomotor activity indicate that reduced body weight in EB-treated groups is resulted from elevated metabolic rate. Omega-3 PUFAs supplementation (3 g/day, 12 weeks) in healthy postmenopausal women increased resting metabolic rate by 14% and EE during exercise by 10% (62). Omega-3 PUFAs converted from both 1% and 22.5% LA diets lowered cumulative food intake in Veh-treated OVX mice (Fig. 1G and H), but increased EE because of omega-3 PUFA production was only observed in an E2-sufficient status (Fig. 2D).

Aging is characterized by elevated blood pressure, and the prevalence of hypertension in postmenopausal women is higher than in aged men (63). Replacement of SFAs with PUFAs such as LA have been reported to reduce low-density lipoprotein and higher plasma LA levels was associated with lower total cholesterol (64‐66). A diet rich in extra virgin coconut oil increased HDL in aged coronary artery disease patients (67). A meta-analysis of randomized controlled trials in a general adult population indicated that provision of omega-3 PUFAs reduces systolic and diastolic blood pressure (68); however, in aging populations, effects of omega-3 PUFAs on blood pressure are controversial (69, 70). In this study, there were no differences in blood pressure or heart rate because of dietary FAs in E2-sufficent or E2-difficient status, whereas we observed reduced blood pressure in ovary-intact postmenopausal mice fed a HFD rich in LA compared with their counterparts fed a HFD rich in coconut oil SFAs. Ovarian hormones such as androgens in an ovary-intact model may impact blood pressure and the optimal FA profile for cardiovascular health in different hormonal status need more studies.

E2 levels and estrogen receptor signaling pathways in hypothalamic nuclei regulate food intake and energy expenditure. The expression pattern of POMC neurons is positively related to E2 levels, and decreased Pomc mRNA levels are accompanied in E2-decline in postmenopausal status. Reduced mRNA levels of Npy in mice with E2 replacement is correlated to attenuated diet-induced obesity. The decrease in mRNA levels of Pomc and increase of Npy with E2 replacement (Fig. 4A) is inconsistent with attenuated weight gain and food consumption (Fig. 1). This study confirms previous findings that E2 increases Ghsr expression, which increases KNDy sensitivity to ghrelin and subsequently glutamate release onto ARC NPY neurons to suppress appetite (38, 71). Previously, we have shown that E2 increases M-current activity in ARC NPY neurons to reduce food intake (38, 71). Additionally, in clonal hypothalamic neurons the inhibition of NPY secretion has been modulated by membrane-associated estrogen receptor alpha (72) and the ratio of estrogen receptor alpha to estrogen receptor beta (73). However, we did not find any effects of diets, steroids, or genotype on Esr2 and only genotype differences on Esr1. Transgenic mouse models will be used to explore the regulatory effects of estrogen receptors on NPY secretion and excitability in the context of dietary fatty acids in the future. Upregulation of Glp1r expression levels by EB treatment indicates a novel mechanism for E2 to mitigate the effects of HFDs on energy balance (74). Of note, compared with EB-treated WT mice fed a 1% LA diet, EB-treated fat-1 mice had increased mRNA levels of both Ghsr and Glp1r, which is consistent with observed lower cumulative food intake (Fig. 1H). However, these changes were not observed in 22.5% LA diet-fed groups. Future experiments will be required to understand the interaction of different neurohormones on energy and glucose metabolism.

Ablation of hepatic transcription factor Foxo1, as a component of hepatic insulin signaling, can rescue diabetic phenotypes in mice lacking hepatic insulin receptor substrates 1 and 2 (75). Metabolic abnormalities including high circulating blood glucose and insulin resistance in Insr^+/− mice were reversed in Insr^+/−  Foxo1^+/− double knockout mice, suggesting that Foxo1 haploinsufficiency restores insulin sensitivity (41). Independent on E2 status, omega-3 PUFAs suppressed gene expression of Foxo1 in liver (Fig. 5A), although Insr was downregulated by omega-3 PUFAs (Fig. 5D) (insulin level not collected in this study). These data suggest that supplementation of omega-3 PUFAs potentiate insulin-regulated glucose homeostasis.

Most FA absorption takes place in the small intestines. The impact of omega-3 PUFAs on gut microbial structure may be linked to docosahexaenoic acid in fish oil or converted from dietary linolenic acid. In middle-aged and elderly women (76), omega-3 PUFAs strongly correlated with bloom of family Lachnospiraceaes, 1 of the main taxonomic groups of human gut that functions to degrade polysaccharides to SCFAs. SCFAs have been shown to exert multiple beneficial effects on energy metabolism, bone health, and gut integrity (77, 78). SCFA producers, such as Bifidobacterium, Oscillospira, Roseburia, and Lachnospira species, were increased after 8 weeks of 4 g of omega-3 FAs supplementation in both capsule and drink formulation (79). We observed an increase in relative abundance of gut microbial taxa in associated with SCFA production (Fig. 6H‐J); however, more evidence of a relationship between gut microbiota and omega-3 PUFAs and/or their metabolites are needed. Circulating SCFA levels need to be assessed for validation of bloom of gut bacteria that can produce SCFAs.

The gut microbiota is essential for lipid absorption. Previous reports have demonstrated elevated triglycerides in the feces of germ-free mice compared with conventional mice (80) and reduced lymphatic transport of lipid in antibiotic-treated rats (81). Bile acids are originally synthesized in the liver as primary bile acids and released from the gallbladder into the small intestines to emulsify and break the dietary fats down into small droplets for digestion and absorption. Gut microbial enzymes such as bile acid hydrolase can convert deconjugate primary bile acids to secondary bile acids (82). Mediated by Blautia wadworthia, HFD feeding induces changes in bile acid profiling in the cecum, characterized by significantly higher total bile acids and primary bile acids conjugates (83). Increase of the relative abundance of Ileibacterium is associated with improved lipid metabolism in the EB-treated OVX mice, which was observed in the fat-1 groups regardless of the estrogen states (84). The relative abundance of Ileibacterium is negatively correlated with glycocholic acid and ursodeoxycholic acid in the hyperlipidemic hamster fed (85). However, continued efforts to validate the correlation between gut microbial taxa, bile acid metabolism, and lipid digestive and absorptive pathways are warranted. New insights from this study indicate that medium-chain coconut oil SFAs and long-chain omega-3 rather than omega-6 PUFAs differentially interact with the gut microbiome, leading to alleviation in HFD-induced obesogenesis and glucose intolerance in the E2-deficient status. These data suggest that different dietary FAs regulate markers related to energy metabolism in hypothalamus and liver, which is important for understanding the full scope of the impact of dietary FA metabolic parameters in postmenopausal women.

Abbreviations

ARC

arcuate nucleus

ASV

amplicon sequence variant

AUC

area under the curve

BMC

bone mineral content

BMD

bone mineral density

CART

cocaine-and-amphetamine-regulated transcript

COX2

cyclooxygenase-2

E2

17β-estradiol

EE

energy expenditure

EB

estradiol benzoate

FA

fatty acid

GWAT

gonadal white adipose tissue

HFD

high-fat diet

HRT

hormone replacement therapy

LA

linoleic acid

LEPR

leptin receptor

NPY

neuropeptide Y

OVX

ovariectomized

POMC

proopiomelanocortin

PUFA

polyunsaturated fatty acid

qPCR

quantitative PCR

ROI

region of interest

SCFA

short-chain fatty acid

SFA

saturated fatty acid

Veh

vehicle

WT

wild-type

Funding

This work was supported by a seed grant from the NJ Institute for Food Nutrition and Health, Rutgers University.

Author Contributions

T.A.R., D.E.R., and S.C.C. conceived the study, designed the experiments, received funding, and provided oversight of the project. K.S. and A.Y. performed animal experiments. A.T.M., Z.H.J., S.A.M., and S.E.F. performed quantitative PCR. N.M. and C.R.L. performed intestinal histology with S.C.C.'s guidance. I.S. and B.D.M. performed bone dissections and dual-energy X-ray absorptiometry scans. S.A.S. provided guidance on bone-related endpoints and L.P.Z. provided guidance on gut microbial analysis. G.W. generated pipeline for 16s rRNA analysis and K.M.T. and K.S. performed gut microbial analysis and made figures. K.S. and A.Y. performed statistical analysis and prepared final figures. K.S. drafted the manuscript and T.A.R. and D.E.R. edited. All authors read and approved final manuscript.

Disclosures

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Data Availability

16S amplicon sequences for fecal microbiota analyses were submitted to the database of the European Bioinformatics Institute and can be retrieved with accession number PRJEB60241.