Naringenin Increases Insulin Sensitivity and Metabolic Rate: A Case Study

Abstract

Our studies in primary human adipocytes show that naringenin, a citrus flavonoid, increases oxygen consumption rate and gene expression of uncoupling protein 1 (UCP1), glucose transporter type 4, and carnitine palmitoyltransferase 1β (CPT1β). We investigated the safety of naringenin, its effects on metabolic rate, and blood glucose and insulin responses in a single female subject with diabetes. The subject ingested 150 mg naringenin from an extract of whole oranges standardized to 28% naringenin three times/day for 8 weeks, and maintained her usual food intake. Body weight, resting metabolic rate, respiratory quotient, and blood chemistry panel including glucose, insulin, and safety markers were measured at baseline and after 8 weeks. Adverse events were evaluated every 2 weeks. We also examined the involvement of peroxisome proliferator-activated receptor α (PPARα), peroxisome proliferator-activated receptor γ (PPARγ), protein kinase A (PKA), and protein kinase G (PKG) in the response of human adipocytes to naringenin treatment. Compared to baseline, the body weight decreased by 2.3 kg. The metabolic rate peaked at 3.5% above baseline at 1 h, but there was no change in the respiratory quotient. Compared to baseline, insulin decreased by 18%, but the change in glucose was not clinically significant. Other blood safety markers were within their reference ranges, and there were no adverse events. UCP1 and CPT1β mRNA expression was reduced by inhibitors of PPARα and PPARγ, but there was no effect of PKA or PKG inhibition. We conclude that naringenin supplementation is safe in humans, reduces body weight and insulin resistance, and increases metabolic rate by PPARα and PPARγ activation. The effects of naringenin on energy expenditure and insulin sensitivity warrant investigation in a randomized controlled clinical trial.

Introduction

In recent years, polyphenols, particularly flavonoids, have emerged as a class of natural products shown to have antiobesity and insulin sensitizing effects.¹ Naringenin is a flavonoid found mainly in citrus fruits and tomato.^2,3 Naringenin exemplifies the term “phytopharmaceutical,” which refers to its potential for alleviating the effects of disorders such as the metabolic syndrome.¹ In obese humans, 1/2 grapefruit (49 mg naringenin) three times daily for 8 or 12 weeks reduced body weight and waist circumference compared to the placebo group.^4,5 Rodent studies show that naringenin reduces diet-induced weight gain and improves glucose and lipid metabolism.^6–9 In mice, fed with a high-fat diet supplemented with naringenin, increases in energy expenditure and activation of brown fat have been demonstrated.^7–10

Our in vitro studies in differentiated human subcutaneous adipose-derived stem cells from overweight and obese female donors show that naringenin increases gene expression of uncoupling protein 1 (UCP1), carnitine palmitoyltransferase 1β (CPT1β), glucose transporter type 4 (GLUT4), carbohydrate responsive element-binding protein (ChREBP), and peroxisome proliferator-activated receptor gamma coactivator 1-α/β, (PGC-1α/β).⁶ The regulation of these genes are important determinants of thermogenesis, whole-body insulin sensitivity, and glucose homeostasis.^11,12

Flavonoids occur naturally as glycosides, which means that they are bound to different sugars. Hydrolysis of the sugar moiety by colonic bacteria releases the aglycone naringenin. Therefore, the aglycone form rarely occurs in significant amounts in natural foods.¹³ Exploring the therapeutic potential of naringenin in humans has been hindered by previous studies demonstrating that following ingestion of citrus juices or fruits, the circulating concentrations of naringenin are low. Pharmacokinetic studies of orange juice and fruit have produced serum concentrations of <1 μm, whereas cell culture and animal studies have determined that 1–200 μm is needed to elicit a physiologic response.

The aglycone release by colonic microbiota is the rate-limiting step in the absorption of naringenin.¹⁴ We have previously shown that an aqueous and ethanolic extract of whole sweet oranges (Citrus sinensis) containing naringenin, the free aglycone form, can be readily absorbed from the small intestine and is present in human serum at concentrations sufficient to elicit a physiologic response.¹⁵ We hypothesized that naringenin supplementation for 8 weeks would increase resting metabolic rate (RMR) and insulin sensitivity in humans. We also explored the mechanisms by which naringenin mediates thermogenesis and glucose metabolism in differentiated human adipose-derived stem cells (hADSCs). The main goal of this study was to determine the safety of multiple dosing of naringenin and its effects on energy expenditure and glucose metabolism in a single subject with untreated diabetes.

Subject

A 53-year-old African American female subject was recruited. The subject was a nonsmoker, had a self-reported history of diabetes, was not taking prescription medications, and did not regularly consume citrus fruits. The subject met the inclusion criteria of fasting blood glucose between 126 and 200 mg/dL. She had been prescribed metformin but had discontinued the medication due to gastrointestinal intolerance. Exclusion criteria consisted of known allergy to citrus fruit. The study was approved by the Pennington Biomedical Research Center (PBRC) Institutional Review Board. The participant provided written informed consent. All procedures were in accordance with PBRC's ethical standards.

Methods

We conducted an 8-week case study. The subject completed three visits to the clinic. Visits were performed in the morning after an overnight fast for at least 8 h where only water was permitted. At the baseline visit, upon arrival, weight and vital signs (blood pressure and heart rate) were measured. The subject completed a medical questionnaire and had a physical examination. Blood was collected for a chemistry panel (glucose, insulin, creatinine, potassium, uric acid, albumin, calcium, magnesium, creatine phosphokinase, alanine aminotransferase, alkaline phosphatase, iron, total cholesterol high-density lipoprotein cholesterol [HDL-C], low-density lipoprotein cholesterol [LDL-C], and triglycerides) and a complete blood count.

RMR was measured over 5 h at baseline and at the end of 8 weeks using the ventilated hood to collect expirated gases. The test required the subject to lie quietly in bed after ingesting a capsule containing 150 mg naringenin. To evaluate the subject's metabolic rate, oxygen consumed and the carbon dioxide given off were measured during the last 30 min of each hour. Following the baseline testing, the subject was given 100 capsules each containing 150 mg naringenin with instructions to take one orally three times daily.

At weeks 2 and 6, the subject received a phone call from the study coordinator and was asked about her progress and compliance. Changes in medications and adverse events were also evaluated. She was encouraged to comply with the study protocol. At week 4, the subject returned to the clinic and her weight, blood pressure, and heart rate were measured. Any remaining capsules were collected and compliance was assessed. Naringenin (100 capsules) for the next 4 weeks was dispensed. Week 8 marked the subject's last visit and involved repeating the baseline testing. At this visit, the subject returned any remaining capsules, and compliance was assessed.

Whole orange extract

Whole sweet oranges (Citrus sinensis) were subjected to an aqueous and ethanolic extraction process, dried, milled, and provided in a powder form by Green Chem/Gencor Lifestage Solutions (Irvine, CA, USA). The quantification of naringenin in the extract, and the safety and pharmacokinetics of the extract in humans are previously described.¹⁵ The extract contained 28% naringenin. Therefore, each capsule prepared by the PBRC pharmacist contained 536 mg of the extract.

Human subcutaneous adipocyte cell culture

Abdominal adipose tissue stem cells from three female subjects (BMI: 27, 32, and 36 kg/m²) were obtained from LaCell, LLC (New Orleans, LA, USA) and as a generous gift from Dr. Frank Lau, Louisiana State University Health Sciences Center (New Orleans, LA, USA). Cells were seeded into culture plates and differentiated into mature adipocytes as previously described.⁶

The whole orange extract was dissolved in dimethyl sulfoxide (DMSO) at 15 mM (based on 28% naringenin content) and added to cells at a dilution of 1:500 to achieve a final concentration of 30 μM. The inhibitors added to the naringenin-treated cells included peroxisome proliferator-activated receptor α (PPARα) antagonist GW6471 and peroxisome proliferator-activated receptor γ (PPARγ) antagonist GW9662 (Cayman Chemical, Ann Arbor, MI, USA) at 10 μM concentration, protein kinase A (PKA) inhibitor H89 (Millipore Sigma, Burlington, MA, USA) at 20 μM concentration, and protein kinase G (PKG) inhibitor Rp-8-pCPT-ck (Biolog Life Science Institute, AXXORA, LLC, Farmingdale, NY, USA) at 50 μM concentration.

GW6471 and GW9662 were dissolved in DMSO at 1000 × , H89 was dissolved in PBS at 500 × , and Rp-8-pCPT-cGMPS was dissolved in PBS at 100 × before dilution in cell culture medium. Following a 2-day treatment, the cells were harvested with TRIzol Reagent. RNeasy Mini Kit (Qiagen, Germantown, MD, USA) was used to isolate total RNA following manufacturer's protocol. The expression of UCP1, CPT1β, and GLUT4 was quantified with one-step reverse transcriptase PCR using the reverse PCR primer to prime cDNA synthesis as previously described.⁶

Statistical analyses of mRNA expression

A general linear model was used to perform analysis of variance. The primary outcomes which were differences from the control and inhibitors were analyzed after Welch's test of homogeneity of variances. The assumption of normality was assessed using the Shapiro–Wilk test. Significance was set at P < .05. Outcomes are summarized as means ± SEM. All analyses were performed using SAS 9.4 (SAS Institute, Cary, NC, USA).

Results

Case study

Weight and vital signs measured at baseline, week 4, and week 8 are presented in Table 1. Body weight decreased by 2.3 kg over the 8-week period. Body weight during a menstrual cycle should not vary more than 1.5 kg, so the weight reduction of 2.3 kg is greater than the variability expected at a stable weight. Insulin decreased by 2.3 μU/ml (18%), and this change was greater than the 4.7% coefficient of variation (CV) of the assay. Serum glucose increased by 2 mg/dL (1.9%), which falls just outside the CV of the assay (1.6%). The homeostatic model assessment of insulin resistance (HOMA-IR) reduced from 3.4 at baseline to 2.8 (17.6%) at the end of the study.

Table 1.

Anthropometric Measurements and Vital Signs Obtained at Baseline and Week 4 and 8 Clinic Visits

Measures	Baseline	Week 4	Week 8
Body weight (kg)	109.4	108.8	107.1
Height (cm)	166.7	—	—
BMI (kg/m2)	39.4	39.2	38.5
Blood pressure (mmHg)a	138/71	134/78	132/78
Heart rate (bpm)	80	80	70

^aSystolic/diastolic.

HDL-C decreased by 3.1 mg/dL (6.8%) which is greater than the assay CV of 1.6%. As it was expected, with the fall in HDL-C, there was a reciprocal change in triglycerides. The low triglyceride baseline value of 53 mg/dL led to an increase by 32% to 70 mg/d which is within the recommended value of <150 mg/dL. LDL-C decreased by 10.3 mg/dL, but this was a calculated value using the Friedewald equation and does not have a CV. The total cholesterol decreased by 10 mg/dL (6.5%), which is greater than the CV (1.2%) of the assay. The serum values of all the blood markers assessed at baseline and after 8 weeks of naringenin treatment are presented in Tables 2 and 3.

Table 2.

Serum Concentrations of Chemistry Panel Markers at Baseline and After 8 Weeks

Variable	Baseline	Week 8
Insulin (μU/ml)	12.8	10.5
Blood glucose (mg/dL)	107	109
HDL-C (mg/dL)	45.0	41.9
LDL-C (mg/dL)	98.4	88.1
Blood urea nitrogen (mg/dL)	13	16
Creatinine (mg/dL)	0.6	0.8
Potassium (mmol/L)	3.94	4.08
Uric acid (mg/dL)	6.4	6.6
Albumin (g/dL)	3.9	4.1
Calcium (mg/dL)	9.1	9.2
Magnesium (mg/dL)	2.0	2.1
Total bilirubin (mg/dL)	0.6	1.0
CPK (IU/L)	151	149
AST (IU/L)	23	26
ALT (IU/L)	28	30
ALK (IU/L)	56	53
Iron (UG/dL)	58	98
Total cholesterol (mg/dL)	154	144

ALK, alkaline phosphatase; ALT, alanine aminotransferase; AST, aspartate aminotransferase; CPK, creatine phosphokinase; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol.

Table 3.

Results of the Complete Blood Count Obtained at Baseline and Week 8

Variable	Baseline	Week 8 (units)
Hemoglobin	12.7	12.7 (g/L)
Hematocrit	38.9	38.6 (%)
Mean cell volume	81.6	80.7 (fL)
Platelets	283	290 ( × 103 cells/μL)
White blood cells	6.4	6.5 ( × 103 cells/μL)
Granulocyte	3.8	4.1 ( × 103 cells/μL)
Neutrophils	3.6	3.8 ( × 103 cells/μL)
Eosinophils	0.2	0.2 ( × 103 cells/μL)
Basophils	0.0	0.1 ( × 103 cells/μL)

The RMR peaked at 3.5% above baseline at 1 h (Fig. 1) and the maximal respiratory quotient increase was 1.2% above baseline at 1 h. The CV of the metabolic rate for the metabolic cart used in this case study was 2.5% and the CV for the respiratory quotient was 1.5–2%. Thus, the metabolic rate increase was greater than the variability of the assay while the respiratory quotient did not change. No adverse events were reported. There were no clinically significant changes in blood safety markers (Table 2).

FIG. 1.

Change in RMR following 8 weeks of naringenin treatment. Change from baseline in the 5-h RMR following 8 weeks of naringenin treatment at 150 mg three times daily. RMR, resting metabolic rate.

Gene expression in human adipocytes

The mRNA expression of UCP1 and CPT1β increased by more than threefold compared to control. The addition of PPARα and PPARγ inhibitors reduced the induction of UCP1 and CPT1β (P < .001) over the course of the 2-day treatment period (Fig. 2). PKA and PKG inhibition did not affect the mRNA expression of UCP1 or CPT1β. We have previously demonstrated an increase in GLUT4 mRNA expression over a 7-day treatment of human adipocytes with naringenin.⁶ However, over a 2-day period, there was no increase in GLUT4 mRNA expression with naringenin treatment, which precluded determination of the effect of the inhibitors.

FIG. 2.

qRT-PCR in human adipocytes treated with naringenin and inhibitors. qRT-PCR assays for mRNA expression conducted in duplicates following naringenin treatment of hADSCs for 2 days compared to naringenin treatment+inhibitors and untreated hADSCs (Control): UCP1 and CPT1β. The results are presented as mean ± SEM. Naringenin increased UCP1 and CPT1β mRNA induction compared to control (P < .001). PPARα and PPARγ inhibition reduced UCP1 and CPT1β mRNA expression (P < .001). Each experiment was conducted with three biological replicates and four technical replicates. CPT1β, carnitine palmitoyltransferase 1β; hADSCs, human adipose-derived stem cells; PPARα, peroxisome proliferator-activated receptor α; PPARγ, peroxisome proliferator-activated receptor γ; UCP1, uncoupling protein 1.

Discussion

This is the first study in humans to investigate the effect of naringenin supplementation for 8 weeks on energy expenditure and glucose metabolism. Body weight decreased by 2.3 kg over the 8-week period. Serum total cholesterol and insulin concentrations reduced; however, HDL-C also decreased. The change in glucose was not clinically significant. However, there was a reduction in HOMA-IR by 17.6%. The RMR increased above the baseline value at 1 h, but there was no change in the respiratory quotient. In human adipocytes treated with naringenin at physiologically attainable doses,¹⁵ UCP1 and CPT1β mRNA expression were downregulated following inhibition of PPARα and PPARγ, but were not affected by PKA or PKG inhibition.

Reduction in body weight occurs with a reduction in energy intake or an increase in energy expenditure or both. The majority of the studies of naringenin supplementation for 12 weeks in obese mouse models of metabolic dysfunction found a reduction in body weight without a decrease in food intake as measured in standard caging, and improvements in insulin sensitivity and lipid metabolism.^7,16,17 When measured in metabolic cage studies, naringenin supplementation in mice results in a small but significant increase in energy intake.^18,19 However, energy expenditure increases, which may explain the reduction in body weight.^7,16–19 Consistent with the rodent studies, the subject in the present study maintained her usual food intake, but her body weight reduced and metabolic rate increased with 8 weeks of naringenin supplementation. Although there was no change in serum glucose concentrations, the reduction in fasting insulin and HOMA-IR suggests improvements in insulin sensitivity.

A reduction in HOMA-IR of 0.13 is associated with a diet-induced weight loss of 1 kg.²⁰ Therefore in our study, the 2.3 kg weight loss would predict a 0.3 reduction in HOMA-IR. We showed a 0.6 reduction in HOMA-IR without any dietary restriction, which suggests that naringenin acts largely through factors independent of weight loss. Our in vitro data provide evidence of the effects of naringenin on upregulation of ChREBPβ and GLUT4 expression in human adipocytes, which in rodent models is associated with regulation of whole-body glucose homestasis.^6,11,21 In humans, expression of ChREBP in white adipose tissue correlates with insulin sensitivity.^11,22,23 In our case study, the subject lost weight and there was a perceptible change in her metabolic rate. Therefore, the improvement in insulin sensitivity may be attributed to the effects of naringenin on body weight, metabolic rate, and upregulation of target genes, but their individual contribution remains to be determined.

In mice placed on a weight cycling protocol, weight regain was accompanied by a significant reduction in energy expenditure. During the high-fat diet cycle, metabolomics studies showed that naringenin, its metabolite apigenin, and bile acids were depleted and did not return to normal levels during the weight loss, despite recovery of other metabolic derangements. Administration of naringenin to mice during the high-fat diet cycle, attenuated weight gain, increased energy expenditure, and upregulated the gene expression of UCP1, the key regulator of thermogenesis in brown adipose tissue.¹⁰ The upregulation of UCP1 gene expression has also been shown in white adipose tissue of mice treated with naringenin, but the results are not consistent.^17,18

The mechanisms by which naringenin exerts its effects on energy expenditure, lipid metabolism, and insulin sensitivity are not completely understood. In hADSCs and primary human white adipose tissue, we have previously shown a robust increase in the genes involved in thermogenesis and glucose metabolism with naringenin treatment. Naringenin stimulated mRNA expression of UCP1, GLUT4, PGC-1α/β (the nuclear receptor coactivators involved in thermogenesis), adipose triglyceride lipase and CPT1β (key enzymes necessary for fat oxidation), and adiponectin (insulin-sensitizing adipokine), in addition to increasing oxygen consumption rate.⁶

In rodents, the browning of adipose tissue is largely stimulated by sympathetic activation of the β3 adrenergic receptor (AR) that activates a signaling pathway involving cyclic adenine monophosphate (cAMP) and PKA.²⁴ However, human adipocytes predominantly express β1- and β2-ARs, which may also have a role in thermogenic activity.^25,26 Although β3 AR agonist treatment has been shown to increase metabolic activity in humans,²⁷ direct evidence of substantial contribution of the β3-AR to the thermogenic program is currently lacking.^28–30

Bypassing ARs by treating with forskolin, a direct activator of the cAMP/PKA pathway, increases the expression of UCP1 in hADSCs.³¹ Natriuretic peptides can act through PKG in human adipocytes to phosphorylate the same targets as the β-ARs do when acting through PKA to stimulate energy expenditure.³² In addition, PPAR ligands have been shown to promote the conversion of human white adipocytes to the brown-like phenotype, decrease body fat, and increase expression of the genes required for fat oxidation.^31,33,34 Thus, PPAR activators have relevance to human physiology.

To identify the signaling pathways activated by naringenin in human adipocytes, we investigated the effect of PPARα and PPARγ inhibition as well as PKA and PKG inhibition. Consistent with evidence to suggest that naringenin is an activator of PPARα and PPARγ,³⁵ inhibition of these nuclear receptor proteins reduced mRNA expression of UCP1 and CPT1β. Inhibition of PKA or PKG did not have a significant effect on naringenin-stimulated induction of UCP1 or CPT1β mRNA, suggesting that naringenin may not be acting through the adrenergic signaling pathway. Mirabegron, a β3 AR agonist produces a marked increase in metabolic rate in humans, but is accompanied by cardiovascular side effects.^27,36 Naringenin supplementation for 8 weeks did not induce any adverse events and may be a safe way to stimulate energy expenditure, but the effects warrant investigation in a randomized controlled trial.

In epidemiologic studies, high flavanone intake from oranges and grapefruits is associated with a cardioprotective effect, especially, a reduction in the risk of ischemic stroke,^37–39 and human intervention trials show that grapefruit consumption improves body composition, insulin sensitivity, blood pressure and circulating lipids.^4,5 Grapefruit has been shown to increase the bioavailability of orally administered drugs.

Theoretically, naringenin being polyphenolic and high in electrons can inhibit cytochrome P450 enzymes and enhance the bioavailability of medications, including statins. However, unlike in the rodent and in vitro studies, the results of in vivo studies in humans suggest that naringenin is not the main inhibitory compound in grapefruit.^40,41 The clinically active constituents responsible for the inhibition are the furanocoumarin derivatives in grapefruit.⁴² While grapefruit contains a high content of furanocoumarin derivatives, sweet oranges contain a very small amount of these derivatives.⁴³ Therefore, we used an extract of sweet oranges which is low in the inhibitors of cytochrome P450 enzymes.

The main limitation of this study is that a single subject participated; therefore, the data generated prevented a clear generalized statement about the role naringenin played over 8 weeks. However, this study can provide preliminary evidence for the claim that naringenin can cause an increase in metabolic rate and thus the rodent studies can be extended to humans. In conclusion, naringenin reduces body weight, increases metabolic rate and improves insulin sensitivity in a human subject with untreated diabetes, without a change in food intake. Our in vitro studies suggest that naringenin may act through PPARα and PPARγ agonism in humans. The results of this case study warrant investigation in a randomized, controlled trial to evaluate the effects of naringenin on energy expenditure and glucose metabolism in humans.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This work was supported, in part, by a grant (U54 GM104940) from the National Institute of General Medical Sciences of the National Institutes of Health (NIH), which funds the Louisiana Clinical and Translational Science Center.