Beneficial effect of dietary geranylgeraniol on glucose homeostasis and bone microstructure in obese mice is associated with suppression of proinflammation and modification of gut microbiome

Abstract

Geranylgeraniol (GGOH) is found in edible oils such as olive, linseed, and sunflower oils, which have favorable metabolic effects. However, it is unknown whether these physiological benefits are mediated through the gut microbiome. Thus, the purpose of this study was to test the hypothesis that GGOH supplementation would improve glucose homeostasis and benefit the bone microstructure in obese mice through suppression of inflammation and modification of gut microbiota composition. Thirty-six male C57BL/6J mice were divided into 3 groups: a low-fat diet, a high-fat diet (HFD), and an HFD supplemented with 800 mg GGOH/kg diet (GG) for 14 weeks. Glucose and insulin tolerance tests were measured at baseline and end of study. The concentrations of adipokine cytokines (resistin, leptin, monocyte chemoattractant protein-1, interleukin-6) were measured via ELISA. Bone microarchitecture and quality were measured by micro-CT. Microbiome analysis was performed using 16S rRNA amplicon sequencing on cecal content. Relative to the HFD group, the GG group: (1) improved glucose tolerance and insulin sensitivity; (2) reduced production of pro-inflammatory adipokines, (3) increased serum procollagen I intact N-terminal propeptide (bone formation marker) concentrations, while decreasing serum collagen type 1 cross-linked C-telopeptide (bone resorption marker) levels, and (4) increased stiffness at both femur and LV-4 and cortical thickness at femoral midshaft. Compared to the HFD group, the GG group had an increased abundance of Butyricicoccus pullicaecorum and decreased Dorea longicatena in the cecal microbiome. Collectively, GGOH improves glucose homeostasis and bone microstructure in obese mice, probably via suppression of pro-inflammation and modification of microbiome composition.

1. Introduction

The prevalence of obesity and obesity-associated disorders including type 2 diabetes mellitus (T2DM) is increasing at an alarming rate. Emerging evidence suggests that obesity and T2DM produce detrimental effects on bone health, such as bone architecture deterioration, impaired bone biomechanical properties, poor bone quality, increased risk of bone fracture, and impaired bone healing [1]. Obesity and bone health are interrelated because osteogenesis and adipogenesis are derived from a common mesenchymal stem cell pool and some signaling molecules stimulating adipogenesis inhibit osteogenesis [2,3]. Accumulation of visceral fat deposition, advanced glycation end-products formation in the bone, and circulating pro-inflammatory adipokines seem to promote bone resorption [2] and accelerate bone loss [1], while anti-inflammatory cytokine, such as adiponectin stimulates osteogenesis [4]. However, the regulation of bone health is complex and modified by various factors, such as the presence of obesity and metabolic diseases compared to under healthy conditions [4], as well as the gut microbiome [5].

The gut microbiome composition and function are linked to the progression of T2DM and osteoporosis [6]. Gut microbiome dysbiosis may reshape intestinal barrier functions as well as host metabolic and signaling pathways, which are directly or indirectly associated with insulin resistance in T2DM [7]. Various natural products and their bioactive components exhibited anti-diabetic activity by modulating intestinal microbiota [8]. In addition, intestinal microbiota regulates bone metabolism by influencing the host metabolism, immune function, hormone secretion, and endocrine environment [6]. Manipulation of microbiota by either the colonization of germ-free mice, treatment with antibiotics, or probiotic supplementation significantly alters bone remodeling, bone development, and growth, as well as bone mechanical strength [9]. Eaimworawuthikul et al. reported that the addition of prebiotics, probiotics, and synbiotics to diet-induced obese-insulin resistant rats not only alleviates osteoclast-related bone resorption but also potentiates bone-formation activities through improving metabolic disturbance and reducing systemic inflammation, supporting the importance of balanced gut microbiota in bone health [6].

Among dietary bioactive components, geranylgeraniol (GGOH) is considered a good candidate to evaluate its effects on T2DM-related complications, in terms of hyperglycemia and bone remodeling. GGOH is a C₂₀ isoprenoid found in fruits, vegetables, and grains and it can be extracted from annatto, a plant that is used as a source for natural food coloring [10]. GGOH has been shown to play an important role in the management of glucose homeostasis [11] and enhance insulin sensitivity by regulating peroxisome proliferator-activated receptor-γ (PPARγ), a major transcription factor regulating adipogenesis, lipid metabolism, as well as osteogenesis [12].

In the past decade, the effects of GGOH supplementation on bone health gained ample attention due to its ability to restore mevalonate pathway functions in bisphosphonate-related osteonecrosis of the jaw [13]. Nitrogen-containing bisphosphonates prevent the incorporation of mevalonate into prenylated (farnesylated and geranylgeranylated) protein in osteoclasts largely due to the loss of geranylgeranylated proteins in osteoclasts [14], resulting in the development of jaw osteonecrosis [15]. GGOH administration has been shown to rescues the negative effects of bisphosphonates in osteogenic cells [16], in vitro apoptosis of bone cells [17], human alveolar osteoblasts [18], primary human osteoclasts [19], and growing male rats [15]. Further, GGOH inhibits osteoclast formation via suppression of the receptor activator of NF-κB ligand expression, and hence abolishes the disruption of bisphosphonate-induced osteoclastic activity [20]. The previous studies strongly suggest that GGOH supplementation would benefit glucose homeostasis and bone health, probably in part through modulating the mevalonate pathway.

To date, no studies investigated the effects of GGOH supplementation on any T2DM-related disorders, including dysglycemia, insulin resistance, bone deterioration, inflammation, and gut dysbiosis. Therefore, in the present study, we investigated the effects of dietary GGOH supplementation on glucose homeostasis, bone parameters (bone turnover biomarkers, microarchitecture, and mechanical property), and adipokines in high-fat-diet-induced obese mice. In addition, we examined GGOH’s effects on the composition and function of the gut microbiome. We hypothesized that GGOH supplementation would improve glucose homeostasis and benefit the bone microstructure in obese mice probably via suppression of inflammation and modification of gut microbiota composition.

2. Methods and materials

2.1. Animals and treatments

Thirty-six 5-week-old male C57BL/6J mice were obtained from Jackson Laboratory (Bar Harbor, ME, USA) and acclimatized to their new environment for 5 days with a standard laboratory chow diet and distilled water ad libitum. Mice were housed (3 mice per cage) and maintained at a controlled temperature of 21 ± 2°C with a 12-hour light-dark cycle. Each treatment group had 4 cages of mice. Body weight, food intake, and water consumption were recorded weekly. All conditions and handling of the animals were approved by the Texas Tech University Health Sciences Center Institutional Animal Care and Use Committee (IACUC #15003). All experiments were performed in accordance with the relevant guidelines and regulations.

After 5-day acclimation, mice were weighed, randomly stratified by body weight, and assigned to 3 groups (n = 12/group): low-fat diet (LFD group, 10% of calories from fat, catalog number D12450J, Research Diets, Inc., New Brunswick, NJ, USA), high-fat diet (HFD group, 60% of calories from fat, catalog number: D12492, Research Diets, Inc., New Brunswick, NJ, USA), and HFD supplemented with GGOH at 800 mg/kg diet group (GG group) (Table 1 for respective diet composition). LFD group was used to serve as the control group. Animals were observed for 14 weeks period during which they had free access to water and food. Body weight, food intake, and water consumption were recorded weekly. GGOH (GG-Gold, GGOH extract from annatto, 85% purity with remaining 15% mainly fatty acids, plant terpenoids, sterols, and waxes) was a gift from American River Nutrition, Hadley, MA, USA.

Table 1 –

The composition of the experimental diet (g/kg)

Ingredient	LFD	HFD	GG
Casein, 80 Mesh	194.2	267.1	267.1
L-Cystine	2.9	4.0	4.0
Corn starch 10	491.4	0	0
Maltodextrin 10	121.4	166.9	166.9
Sucrose	66.8	91.9	91.9
Cellulose, BW 200	48.5	66.8	66.8
Soybean oila	33.38	33.38	33.38
Lard	19.4	327.2	327.2
Mineral mixb, S10026	9.7	13.4	13.4
Dicalcium phosphate	12.6	17.4	17.4
Calcium carbonate	5.3	7.3	7.3
Potassium citrate, 1 H2O	16.0	22.0	22.0
Vitamin mixc, V13401 (without E)	9.7	13.4	13.4
Vitamin E acetate (500 IU/gm)	0.1	0.1	0.1
Choline bitartrate	1.9	2.7	2.7
GGOHd	0	0	0.94

HFD, GG, a high-fat diet supplemented with GGOH at 800 mg/kg diet; high-fat diet control group; LFD, low-fat diet.

^aSoybean oil tocopherol stripped (catalog number: 404365, Dyets Inc., Bethlehem, PA, USA)

^bMineral mix provides (g/kg diet): sodium chloride (39.3% Na, 60.7% Cl), 259; magnesium oxide (60.3% Mg), 41.9; magnesium sulfate,7H₂O (9.87% mg, 13.0% S), 257.6; chromium potassium sulfate,12H₂O (10.4% Cr, 12.8% S, 7.8% K), 1.925; cupric carbonate (57.5% Cu), 1.05; sodium fluoride (45.2% Fl, 54.8% Na), 0.2; potassium iodate (59.3% I), 0.035; ferric citrate (17.4% Fe), 21; manganous carbonate (47.8% Mn), 12.25; sodium selenite (45.7% Se), 0.035; zinc carbonate (52.1% Zn), 5.6; ammonium molybdate,4H₂O (54.3% Mo), 0.3; dextrose (monohydrate), 399.105.

^cVitamin mix provides (g/kg diet): vitamin A acetate (500,000 IU/g), 0.8; vitamin D3 (100,000 IU/g), 1.0; vitamin K1 (menadione sodium bisulfite, 62.5% menadione), 0.08; biotin (1%), 2.0; cyanocobalamin (B12) (0.1%), 1.0; folic acid, 0.2; nicotinic acid, 3.0; calcium pantothenate, 1.6; pyridoxine-HCl, 0.7; riboflavin, 0.6; thiamine HCl, 0.6; sucrose, 988.42.

^dGGOH (GG-Gold, GGOH extract from annatto, 85% purity with remaining 15% mainly fatty acids, plant terpenoids, sterols, and waxes) was a gift from American River Nutrition, Hadley, MA, USA.

2.2. Glucose and insulin tolerance tests

Twelve weeks after treatments were started, mice were fasted for 4 hours and injected intraperitoneally with glucose (2 mg/g body weight) or insulin (1 U/kg body weight; Humulin, Abbott, Chicago, IL) for glucose tolerance test (ipGTT) or insulin tolerance test (ipITT), respectively. For both ipGTT and ipITT, blood glucose levels were measured in the blood drawn from the tail vein prior to (0 minutes) and at 15, 30, 60, and 120 minutes after mice were intraperitoneally injected with glucose or insulin. The total area under the curve (AUC) was calculated using the trapezoidal method.

2.3. Sample collection

At the end of the experiment, animals fasted for 4 hours and blood was collected from mice euthanized with isoflurane. Pancreases were stored at −80°C prior to insulin extraction for ELISA or fixed in Z-fix (AnaTech Ltd., Battle Creek, MI, USA) at room temperature and embedded in paraffin for histological assessment. Femur and lumbar vertebrae-4 (LV-4) were harvested and cleaned of adhering soft tissues and stored in 70% ethanol at 4°C for later analyses. Blood samples were centrifuged at 1,500 g for 20 minutes and the serum samples were obtained and kept at −80°C until analyzed. The cecum content (fecal) samples were collected from the cecum and frozen at −80°C for later microbiome analyses.

2.4. Insulin and histological assessment of pancreatic tissue

Serum insulin levels were quantified using a mouse insulin ELISA kit (EMD Millipore Co., Billerica, MA, USA). Total pancreatic insulin was extracted from the pancreas by acetic acid extraction and insulin content was determined using the same mouse insulin ELISA kit. Pancreatic tissue sections were immunostained, as described previously [21], for insulin and glucagon. Briefly, antigen retrieval was performed by heating slides for 15 minutes in 0.01 M sodium citrate buffer (pH 6.0) in a microwave at full power. Slides were then quenched with hydrogen peroxide, blocked with 10% goat serum, and incubated with guinea pig anti-insulin (diluted 1:1000; Dako Agilent Pathology Solutions, Santa Clara, CA, USA) or mouse anti-glucagon (diluted 1:5000; Sigma) primary antibodies. Tissue sections were then incubated with biotinylated secondary antibodies (Vector Laboratories, Burlingame, CA, USA), avidin-biotin-enzyme complex (Vector Laboratories), and diaminobenzidine as chromogen (BioGenex, Fermont, CA, USA). Tissue sections were counterstained with hematoxylin.

2.5. Serum bone biomarker analyses

The concentrations of procollagen type 1 N-terminal propeptide (P1NP) and C-terminal telopeptide of type I collagen (CTX) in serum were quantified using respective kits (P1NP and CTX kits from Immunodiagnostic System Ltd, Scottsdale, AZ, USA). The sensitivity for P1NP and CTX was 0.7 ng/mL and 2.0 ng/mL, respectively. Intra-assay and inter-assay coefficients of variance were 6.4% and 0.2% for P1NP as well as 5.6% and 10.5% for CTX.

2.6. Bone microarchitecture measurement with μCT and mechanical properties evaluation with micro-finite element analysis (micro-FE)

The lumbar vertebrae-4 (LV-4) and right femur were scanned using micro-computed tomography (μCT) (Model: Scanco μCT 40; SCANCO Medical AG, Switzerland) following the procedures of Cao et al. [22] and the recommended guidelines for μCT scanning [23]. The trabecular bone of the LV-4 and distal femur was scanned. The volume of interest comprised the secondary spongiosa in 100 cross-sectional slices of the distal femur beginning about 0.1 mm (~6 slices) from the growth plate region. The entire trabecular bone of LV-4 was evaluated. For assessment of cortical indices, a scan of 100 slices at the femoral midshaft was performed. All scans were performed in a 1024 × 1024 matrix resulting in an isotropic voxel resolution of 16 μm³. An integration time of 150 milliseconds per projection was used. Trabecular parameters in both LV-4 and femur included trabecular bone volume per unit bone area/total volume (BV/TV, %), number (Tb.N, 1/mm), separation (Tb.Sp, mm), and thickness (Tb.Th, mm), structure model index (SMI) and trabecular connectivity density (Conn.Dn, mm⁻³). Cortical parameters in femur included bone area/total area (B.Ar/T.Ar, mm²), medullary area (Me.Ar, mm²), and cortical thickness (Ct.Th, mm). The operator performing the scans and analysis was blinded to treatments. To determine the stiffness of the femur midshaft, virtual compression tests by micro-FE simulations were performed (Scanco Medical AG, Version 1.13). Three-dimensional images of the femur midshaft were used to generate micro-FE models and to estimate the local mechanical stimuli [24]. The compression simulation tests were done within linear elasticity and in the z-direction.

2.7. Adipokine cytokine measurements

Total tissue lysates from white adipose tissue were used to determine the levels of leptin, monocyte chemoattractant protein-1 (MCP-1), and interleukin (IL)-6 using a commercial multiplexing system (Luminex-MagPix, Luminex Corporation, Austin, TX, USA). The adipokines were normalized to total protein content.

2.8. 16S rRNA gene amplicon sequencing, and gut microbiome analysis

After collecting the stool samples from mice cecum, microbial DNA was isolated using PowerFecal DNA isolation kit (Qiagen Inc., Germantown, MD, USA) following the manufacturer’s instructions. 16S rRNA marker gene amplification and sequencing were performed as previously described in our published work [25] at the Center for Biotechnology and Genomics, Texas Tech University, Lubbock, TX, USA. Raw sequencing data were deposited under BioProject accession numbers PRJNA551310 (LFD and HFD samples) and PRJNA637806 (GG samples) in the National Center for Biotechnology Information BioProject database. The 16S rRNA gene database, Greengenes database version 13.8, was used for taxonomy assignment [26]. Relative abundance of individual taxa was compared among all groups using the non-parametric Kruskal–Wallis test followed by Dunn’s test and was regarded as significant when P-value < .05. This was done using GraphPad Prism 7.0 (GraphPad Prism, San Diego, CA, USA).

2.9. Statistical analysis of physiological data

Results were presented as a mean ± standard error of the mean (SEM). SigmaStat software version 14.0 (Systat Software, Inc., San Jose, CA, USA) was used for data analysis of various parameters by t-test or one-way or repeated-measures analysis of variance followed by post-hoc Fisher’s Least Significant Difference tests. A significance level of P-value < .05 applies to all statistical tests.

3. Results

3.1. Body weight

Fig. 1 presents the mean body weight for different treatment groups. At baseline, there was no difference in body weight among all treatment groups. Starting 8 weeks to 14 weeks of intervention (end of study), the HFD group had greater body weight than the LFD group. Throughout the study period, there was no statistical difference in body weight (Fig. 1), food intake (data not shown), and water consumption (data not shown) between the HFD group and the GG group.

Fig. 1 –

Body weight changes in mice with different dietary treatments. Male C57Bl/6J mice were fed a LFD, HFD, or HFD supplemented with GGOH at 800 mg/kg diet for 14 wks. * indicates significant differences between HFD (pink) or GG (blue) vs LFD, *P < .05; while * (underlined asterisk) indicates significant differences between GG (blue) vs HFD, * P < .05

3.2. Glucose homeostasis and pancreatic morphological analysis

As expected, mice in the HFD group had significantly impaired glucose tolerance and reduced insulin sensitivity compared to those in the LFD group, characteristic of T2DM (Fig. 2A–D). For ipGTT, GGOH supplementation significantly improved glucose tolerance compared to the HFD group. At both 60 and 120 minutes after glucose administration, GGOH supplementation resulted in a significant decrease in blood glucose levels (Fig. 2A) (P < .05) and overall improved glucose tolerance of HFD-fed mice as indicated by a decreased ipGTT AUC (Fig. 2B). In terms of ipITT results, relative to the HFD group, GGOH supplementation improved overall insulin sensitivity during the 2-hour ipITT as shown by the ipITT AUC results (Fig. 2D).

Fig. 2 –

Effects of geranylgeraniol (GGOH) supplementation on glucose tolerance and insulin levels. Male C57Bl/6J mice were fed a LFD, HFD, or HFD supplemented with GGOH at 800 mg/kg diet. IPGTT (A), AUC for IPGTT (B), IPITT (C), AUC for IPITT (D), serum insulin (E) and pancreatic insulin (F) are shown. * indicates significant differences between HFD (pink) or GG (blue) vs LFD, * P < .05

The HFD group had significantly lower serum insulin than that of the LFD group; and (ii) dietary supplementation with GGOH resulted in increased serum insulin of obese mice, although there was no statistical difference between the GG group and the LFD group (P = .085) or between the GG group and the HFD group (P = .077) (Fig. 2E). The LFD group had the highest pancreas insulin among the three dietary groups (P = .015), but there was no difference between the HFD group and the GG group in the pancreatic insulin of obese mice (Fig. 2F). Interestingly, regardless of treatment, histological analysis of the pancreas revealed normal exocrine and endocrine tissue with the normal arrangement of the glucagon producing alpha cells along the outer edge of the islets (glucagon, Fig. 3D–F) and insulin-producing beta cells throughout the islets (insulin, Fig. 3A–C).

Fig. 3 –

Immunohistochemical assessment of pancreatic tissue in mice after geranylgeraniol (GGOH) supplementation. Pancreatic tissue sections collected from male C57Bl/6J mice fed an LFD (A and D), HFD (B and E) or an HFD supplemented with GGOH at 800 mg/kg diet (C and F) were immunostained for insulin (A–C) or glucagon (D–F). Tissue sections were counterstained with hematoxylin.

3.3. Serum bone turnover biomarkers

The LFD group had the highest values of serum P1NP among the three treatment groups (P = .008) with no significant differences between the HFD group and the GG group (Fig. 4A). The HFD group had the highest serum CTX concentrations among all groups (P = .002). Relative to the HFD group, the GG group significantly decreased CTX levels to near the LFD group (Fig. 4B).

Fig. 4 –

Effects of geranylgeraniol (GGOH) supplementation on serum P1NP (A) and CTX (B) levels. Group assignment included a low-fat diet group (LFD group), a high-fat diet group (HFD group), and an HFD with GGOH supplementation at 800 mg/kg diet (GG group). * indicates significant differences between groups, * P < .05.

3.4. Trabecular bone of LV-4 and femur

The effects of GGOH supplementation on the trabecular bone of LV-4 and femur are listed in Table 2. GGOH supplementation resulted in higher values for trabecular bone volume/total volume (BV/TV) at both LV-4 and the femur, in the groups with the order: LFD group = GG group > HFD group. We observed GGOH supplementation increased trabecular number (Tb.N) at the femur in HFD-fed mice, but not at LV-4. GGOH supplementation decreased trabecular separation (Tb.Sp) and SMI at both LV-4 and the femur in obese mice. There was no difference in Tb.Sp and structural model index (SMI) at both LV-4 and femur between the LFD and the GG group. Furthermore, compared to the HFD group, GGOH supplementation increased connectivity density (Conn.Dn) and decreased SMI at LV-4, while there was no statistical difference at both parameters between the GG group and the LFD group.

Table 2 –

Effect of geranylgeraniol (GGOH) on the trabecular bone of LV-4 and femur of obese male mice

Parameter	LFD	HFD	GG	P-value
LV-4
BV/TV (%)	19.5 ± 0.7	17.5 ± 0.6a	19.2 ± 0.56b	.046
Tb.N (b/mm)	5.54 ± 0.10	5.36 ± 0.03	5.59 ± 0.12	.174
Tb.Th (mm)	0.05 ± 0.001	0.05 ± 0.001	0.04 ± 0.001	.139
Tb.Sp (mm)	0.18 ± 0.003	0.19 ± 0.002	0.17 ± 0.003b	.018
Conn.Dn (mm−3)	205 ± 16	168 ± 10a	231 ± 10b	.003
SMI	1.43 ± 0.12	1.70 ± 0.10	1.28 ± 0.08b	.015
Femur
BV/TV (%)	9.36 ± 0.62	6.28 ± 0.53a	9.15 ± 0.87b	.005
Tb.N (b/mm)	3.99 ± 0.14	3.39 ± 0.14a	3.81 ± 0.13b	.011
Tb.Th (mm)	0.05 ± 0.001	0.05 ± 0.001	0.05 ± 0.002	.741
Tb.Sp (mm)	0.25 ± 0.01	0.30 ± 0.01a	0.26 ± 0.01b	.007
Conn.Dn (mm−3)	59.1 ± 10.0	28.3 ± 6.6a	50.1 ± 7.3	.039
SMI	2.65 ± 0.07	3.06 ± 0.09a	2.63 ± 0.10b	.003

Data are expressed as mean ± standard error of the mean (SEM), n = 12 per group. Group assignment includes a low-fat diet group (LFD group), a high-fat diet group (HFD group), and an HFD with GGOH supplementation at 800 mg/kg diet (GG group). BV/TV, percent trabecular bone volume Tb.N, trabecular number; Tb.Th, trabecular thickness; Tb.Sp, trabecular separation; Conn.Dn, connectivity density; SMI, structure model index.

^aP < .05 where a, indicates significant differences between HFD or GG vs LFD.

^bP < .05 where b, indicates significant differences between HFD and GG.

3.4. Cortical bone and stiffness

Table 3 lists the impacts of GGOH supplementation on cortical bone and stiffness of the femur midshaft. Relative to the LFD group, the HFD group had lower values of bone volume/total volume (BV/TV), bone area (B.Ar), cortical thickness (Ct.Th), and stiffness at midshaft of mice. Relative to the HFD group, the GG group had higher values of BV/TV and stiffness of the femoral midshaft bone. There was no difference between the GG group and the LFD group in BV/TV and stiffness. Among the three treatment groups, the HFD group had the lowest Ct.Th values, which resulted in the group order: HFD group < the GG group < the LFD group.

Table 3 –

Effect of geranylgeraniol (GGOH) on cortical bone and strength of femur midshaft in obese male mice

Parameter	LFD	HFD	GG	P value
Cortical bone
BV/TV (%)	49.2 ± 0.56	47.4 ± 0.42a	49.2 ± 0.51b	.031
B.Ar (mm2)	1.05 ± 0.03	0.96 ± 0.03a	1.02 ± 0.03	.047
Me.Ar (mm2)	1.07 ± 0.02	1.07 ± 0.01	1.04 ± 0.01	.539
Ct.Th (mm)	0.223 ± 0.004	0.20 ± 0.001a	0.21 ± 0.003ab	<.001
Bone strength
Stiffness (N/mm)	129 ± 35a	18 ± 7ba	119 ± 45ab	.035

BV/TV, bone volume/total volume; B.Ar, cross-sectional bone area; Ct.Th, cortical thickness; Me.Ar, cross-sectional marrow area. Data are expressed as mean ± SEM, n = 10–12 per group. Group assignment includes a low-fat diet group (LFD group), a high-fat diet group (HFD group), and HFD with GGOH supplementation at 800 mg/kg diet (GG group).

^aP < .05 where a, indicates significant differences between HFD or GG vs LFD.

^bP < .05 where b, indicates significant differences between HFD and GG.

3.5. Fat pad weight and serum adipokine production

Fig. 5 presents the effects of GGOH supplementation on fat pad weight and serum adipokines production. As expected, the HFD group had significantly greater fat pad weights compared to the LFD group, and there was no difference in fat pad weight between the HFD group and the GG group (Fig. 5A). Relative to the LFD group, the HFD group had relatively higher levels of leptin (Fig. 5B), MCP-1 (Fig. 5C), and IL-6 (Fig. 5D) in adipose tissue. In addition, compared to the HFD group, the GG group had relatively lower levels of leptin (Fig 5B), MC.P-1 (Fig. 5C), and IL-6 (Fig. 5D) in adipose tissue.

Fig. 5 –

Alteration in fat pad weight and adipocytokine production caused by geranylgeraniol (GGOH) supplementation. LFD, low-fat diet group; HFD group, high-fat diet group; GG group, HFD + GGOH at 800 mg/kg diet. * indicates significant differences between groups, *P < .05.

3.7. Microbiome composition and function

Among the cecal samples, a median of 200,144 reads was generated. Using DADA2, we observed 2,017 exact amplicon sequence variants. The median frequency of these amplicon sequence variants among all samples was 16,334. In general, the effect of GGOH supplementation on the gut microbiome was minor as it did not alter the alpha diversity of the gut microbiome nor the abundance of the majority of bacterial species, however, its effect was observed on two species (Fig. 6). GGOH supplementation increased the relative abundance of B. pullicaecorum in comparison to LFD and HFD (Fig. 6A). Unlike B. pullicaecorum that was at minimal abundance in both LFD and HFD, the relative abundance of D. longicatena was slightly higher in HFD than in LFD but did not reach statistical significance (P = .1597, Fig. 6B). In comparison to the HFD group, GGOH supplementation decreased the abundance of D. longicatena to a comparable level similar to the LFD group (P = .0305, Fig. 6B).

Fig. 6 –

Cecal microbiome changes associated with geranylgeraniol (GGOH) supplementation. LFD, low-fat diet group; HFD group, high-fat diet group; GG group, HFD + GGOH at 800 mg/kg diet. Statistical significance was determined using Kruskal–Wallis test followed by Dunn’s test (* P < .05). Data distribution is represented by a violin plot showing replicates as points and lines indicate the median and quartiles.

4. Discussion

In this study, we presented the beneficial effects of dietary bioactive component, GGOH, on T2DM-associated disorders, namely, hyperglycemia, insulin resistance, bone deterioration, inflammation, and altered gut microbiome composition. The findings of this study support our hypotheses that GGOH supplementation into a high-fat diet: (i) improved glucose homeostasis, (ii) mitigated bone microarchitecture deterioration and improved bone quality, (iii) reduced pro-inflammatory adipokine, and (iv) modified taxonomic profiles.

Even though the HFD mice exhibited impaired glucose tolerance and insulin resistance there was no increase in serum insulin concentration. The lack of such an increase indicates the pancreatic beta cells were not responding by increasing insulin secretion, and suggests beta cell failure due to depletion of insulin stores. Beta cell failure to compensate for insulin resistance has been reported previously in C57Bl/6 HFD mice [27] and was associated with altered lipid metabolism. While the reason for beta cell dysfunction in our study remains to be determined, we have reported impaired lipid metabolism in our HFD mice [36]. Clinically patients with T2DM can also develop beta cell failure and decreased serum insulin level as the disease progresses. Although GGOH had no effect on pancreatic insulin content, there was a significant improvement in glucose homeostasis and serum insulin levels were not different from LFD mice.

Improved glucose tolerance and insulin sensitivity in mice with hyperglycemia and insulin resistance by supplementing GGOH might be through regulating PPAR action, but not through regulating target genes, such as adiponectin. PPARγ is mainly expressed in adipose tissues and the immune system, but it is also expressed in the liver [28]. Previously [29], we showed no changes in serum adiponectin levels in HFD-induced insulin-resistant mice. Further, adipocyte hypertrophy induced by HFD was significantly decreased with tocotrienol and metformin administration without altering serum adiponectin levels [29]. As Lewis et al. [4] mentioned in their recent review article, the role of adiponectin may be dysregulated by diseases to modify existing disease status in contrast to healthy conditions. Thiazolidinediones, anti-diabetic drugs, increased insulin sensitivity through PPARγ activation. Supporting this, GGOH treatment in adipocyte linage cells enhanced PPARγ expression and adiponectin marker genes including fatty acid binding protein 4 (Fabp4), adiponectin (Adipoq), and CCAAT/enhancer-binding protein alpha (Cebpa) in the presence of the anti-diabetic drug, rosiglitazone [30]. In contrast, GGOH downregulates PPARγ expression with suppression of the adipogenic genes and proteins including FABP4 and CEBPα in adipocyte lineage cells [31]. These conflicting results suggest that GGOH could act as PPARγ agonist or antagonist depends on environmental cues, such as the presence of obesity or T2DM [28].

PPARγ is also involved in regulating bone mass by modulating osteoblast differentiation and adipogenesis in bone marrow mesenchymal stem cells [2]. Metformin, the first drug of choice for T2DM has been shown to benefit bone metabolism by promoting osteoblast differentiation of mesenchymal stem cells through inhibiting PPARγ [32]. In contrast, T2DM patients, treated with thiazolidinediones, PPARγ agonist, improved insulin sensitivity by modulating PPARγ expression in other target tissues, such as liver, but increased bone fracture risk [33,34]. Thus, the balance between osteoblastogenesis and adipogenesis could have potential implications in treating T2DM-associated with bone loss.

Osteopenia or osteoporosis is a process of imbalance between inadequate bone-building osteoblasts and excessive bone-resorptive osteoclasts. Excessive ROS and inflammatory responses have been shown to stimulate osteoclastic differentiation and induce osteoblastic apoptosis [17]. Previous studies have shown that GGOH suppresses inflammation or oxidative stress via either inhibiting NF-kB activation [35,36] or activating the mevalonate pathway [37]. The anti-inflammatory actions of GGOH could be contributing to its anti-resorptive effect, as shown by decreased bone resorption biomarker, such as CTX. Our findings of GGOH’s anti-inflammatory action on HFD-induced adipokine production, as shown in reduced MCP-1 and IL-6 levels, are further corroborated with previous work in human macrophage-like THP-1 cells [35,38], monocytes [38], and animals [36]. The decrease of P1NP in HFD-induced obese mice compared to the normal chow fed mice has been reported previously [39]. Decreased bone formation and increased bone resorption in HFD mice have also been reported [40]. Decreased bone formation in HFD is likely due to the expansion of bone marrow adipocytes [39] since osteoblasts and adipocytes are differentiated from the same mesenchymal stem cell pool. Bone is a dynamic organ that undergoes constant turnover which involves bone resorption by osteoclasts followed by bone formation by osteoblasts. Therefore, the decrease in bone formation P1NP in GGOH is likely through a different mechanism, i.e., reduced bone turnover as a result of decreased bone resorption.

Dysbiosis of gut microbiota may lead to metabolic deregulation, including obesity, increased insulin resistance, and inflammation which are key risk factors in the development of T2DM [41]. Accumulating evidence suggests that gut microbiota composition plays an important role in bone remodeling [6]. In this study, we demonstrated that GGOH’s beneficial impact on anti-obesity, glucose homeostasis, and bone remodeling seems to be associated with the favorable modification of inflammation-associated gut microbiota, namely B. pullicaecorum and D. longicatena, in HFD-fed mice, particularly at the species levels, since lower taxonomic ranks are better correlated with metabolic consequences [42]. B. pullicaecorum (a clostridial cluster IV strain and butyrate-producing bacteria) has a high implication on anti-inflammation. B. pullicaecorum is an intrinsic acid and conveys bile acid tolerance in terms of viability and metabolic activity in patients suffering from inflammatory bowel disease [43]. Patients with Crohn’s disease, an inflammatory bowel disease, have a lower number of B. pullicaecorum in their stools [44]. Oral administration of B. pullicaecorum to rats with colitis mitigates intestinal inflammation production, as shown in decreased intestinal myeloper-oxidase, tumor necrosis factor-α, and interleukin-12 histologically [44]. Oral administration of B. pullicaecorum in mice with colorectal cancer regressed clinical outcomes by activating the short-chain fatty acid transporter and receptors [45]. These results suggest that B. pullicaecorum could be a promising probiotic candidate for people suffering from inflammatory bowel disease [44] and colorectal cancer [45]. In addition, D. longicatena, a glutamate-producing bacterium, was shown to be positively associated with obesity-associated metabolic dysfunction [46,47] and negatively correlated with markers for insulin resistance [48]. Liu et al. reported that (i) D. longicatena is highly enriched in obese individuals than those in lean individuals, (ii) positively correlated with circulating leptin, and (iii) negatively correlated with circulating adiponectin, indicating that D. longicatena may constitute potential biomarker linking gut microbiota and metabolic status [46]. Supporting this notion, in a case-control study, Allin et al. reported that the abundance of D. longicatena was associated with greater BMI and waist circumference of individuals with prediabetes [47]. Moreover, D. longicatena was associated with obesity in humans [49]. Thus, we cautiously speculate that the observed decreased in leptin production in the white fat pad of obese mice by GGOH supplementation is associated with the reduction in D. longicatena abundance. However, further investigation is required to establish this link.

This study has some limitations to be considered. The microbiome consists of hundreds to thousands of different species. Our analysis showed that only two of them (B. pullicaecorum and D. longicatena) are influenced by the GGOH treatment. We acknowledge that this association doesn’t imply causation, but we present this association as a promising area of research to pursue possible links between gut microbiota and glucose homeostasis as well as bone health. We believe our findings shall provide the promising future research direction to determine whether these species would provide a new potential mechanism for GGOH’s anti-diabetes benefits and osteoprotection, similarly to Akkermansia muciniphila has been widely used for treating obesity and T2DM associated disorders [50].

In summary, our present findings show that GGOH not only improves T2DM-associated glucose homeostasis but also improves bone microstructure, along with decreased inflammation and favorable gut microbiome composition in obese mice. Our study suggests that dietary GGOH supplementation may have potential in the management of diabetes. Future studies are warranted to investigate the direct correlation among glucose homeostasis, bone microstructure, and gut microbiota in obese mice.

Abbreviations:

AUC

area under the curve

B.Ar

cross-sectional bone area

BV/TV

percent trabecular bone volume

Conn.Dn

connectivity density

Ct.Th

cortical thickness

GGOH

geranylgeraniol

GG

HFD + 800 mg GGOH/kg diet

GTT

glucose tolerance test

HFD

high-fat diet

IL-6

interleukin-6

ITT

insulin tolerance test

LFD

low-fat diet

MCP-1

monocyte chemoattractant protein-1

Me.Ar

cross-sectional marrow area

RDA

redundancy analysis

SMI

structure model index

T2DM

type 2 diabetes mellitus

Tb.N

trabecular number

Tb.Th

trabecular thickness

Tb.Sp

trabecular separation