Berberine improves mesenteric artery insulin sensitivity through up‐regulating insulin receptor‐mediated signalling in diabetic rats

Abbreviations

AMPK

AMP‐activated protein kinase

eNOS

endothelial NOS

IRS

insulin receptor substrate

InsR

insulin receptor

MTT

3‐[4,5 dimethylthiazol‐2‐yl]‐2,5‐diphenyl tetrazolium bromide thiazolyl blue

PE

phenylephrine

PI3K

phosphatidylinositol 3‐kinase

SNP

sodium nitroprusside

Tables of Links

TARGETS
Catalytic receptors a	Enzymes b
InsR	Akt (PKB)
	AMPK
	eNOS
	PI3K

LIGANDS
ACh	Palmitic acid (palmitate)
Insulin glargine	Wortmannin
Phenylephrine (PE)

These Tables list key protein targets and ligands in this article which are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, the common portal for data from the IUPHAR/BPS Guide to PHARMACOLOGY (Pawson et al., 2014) and are permanently archived in the Concise Guide to PHARMACOLOGY 2015/16 (^a,bAlexander et al., 2015a, 2015b).

Introduction

Type 2 diabetes characterized by insulin resistance is associated with impaired endothelium‐dependent vasodilatation (Steinberg et al., 1996). Berberine is a natural isoquinoline alkaloid, which is present in medicinal herbs Coptidis rhizome (Huanglian in Chinese). Recent clinical studies have shown that berberine reduces blood pressure and increases systemic insulin sensitivity in patients with metabolic syndrome (Perez‐Rubio et al., 2013). Berberine has been shown to lower blood glucose and modulate lipid metabolism by activating AMP‐activated protein kinase (AMPK) (Ko et al., 2005; Chang et al., 2013). Berberine has also vasodilator effects (Ko et al., 2000; Lau et al., 2001), preventing hyperglycaemia‐induced endothelial injury via AMPK (Wang et al., 2009). In a previous study, we showed that berberine protects the diabetic heart from ischaemic/reperfusion injury through activating the Akt‐eNOS signalling pathway and inhibiting the apoptosis of myocardium cells in rats (Chen et al., 2014). Therefore, we hypothesized that berberine may improve insulin sensitivity in diabetes by up‐regulating insulin receptor‐mediated signalling. The present study aimed to investigate the effects of berberine on insulin‐induced vasodilatation and its underlying mechanism in a streptozotocin (STZ)‐treated, high‐fat diet‐fed type 2 diabetes animal model. We found that berberine significantly ameliorated impaired endothelial function and improved insulin‐induced vasodilatation in diabetic mesentery arteries mainly through up‐regulating insulin receptor‐mediated signalling.

Methods

Male Sprague–Dawley rats, 120–150 g, were housed in a temperature controlled room with a 12 h light dark cycle. Rats were fed a casein‐cornstarch‐sucrose‐based diet containing a high‐fat content consisting of 20% fat, 62% carbohydrate and 17% protein, ad libitum, with free access to dwater. After 1 week of adaptation, rats were fasted overnight and then injected i.p. with 30 mg·kg⁻¹ body weight of STZ (Sigma‐Aldrich, Shanghai, China) dissolved in citrate buffer (pH 4.5) to induce diabetes. Normal control animals (Control) were deprieved of food (fasted) overnight and injected with the citrate buffer vehicle. The fasting blood glucose of STZ‐treated rats was measured 3 days later, and a second injection was applied if rats were not diabetic. Fasting blood glucose was measured again 3 days after the first STZ injection. Rats with fasting blood glucose levels over 7.0 mmol·L⁻¹ were randomly divided into two groups and continued to be fed the high‐fat diet for 8 weeks. These two groups were matched for body weight and blood glucose levels. One group was used as the diabetic control (Diabetes) and the other was treated with berberine chloride (gavage) dissolved in 0.9% saline solution at a dose of 200 mg·kg⁻¹·day⁻¹(Chen et al., 2014). The diabetic rats and the normal control rats were treated with the vehicle of 0.9% saline solution (gavage) as in our previous study (Chen et al., 2014).

At the end of the study, animals were fasted overnight and killed by administration of an overdose of anaesthetic, sodium pentobarbital (30 mg·kg⁻¹, i.p.). All experimental procedures were performed in accordance with International Guidelines for the Care and Use of Laboratory Animals and were approved by the Animal Care and Research Ethics Committee of the Fourth Military Medical University Research Council. All studies involving animals are reported in accordance with the ARRIVE guidelines for reporting experiments involving animals (Kilkenny et al., 2010; McGrath & Lilley, 2015).

Isometric force measurement

After the rats had been killed, the mesenteric artery (inner diameter 100 to 150 μm) was dissected out, any adhering connective tissues were removed, and then it was cut into ring segments (1 mm in length). Rings were suspended on a micromyograph in organ baths containing oxygenated (95% O₂; 5% CO₂) physiological saline solution (composition in mmol·L⁻¹: NaCl 119, KCl 4.7, KH₂PO₄ 1.18, MgSO₄ · 7H₂O 1.17, CaCl₂ 2H₂O 2.5, NaHCO₃25, EDTA 0.03 and d‐glucose 5.5, pH 7.4, 37°C). Following equilibration for 90 min under a resting tension of 0.25 g, rings were twice contracted with KCl (60 mmol·L⁻¹) (Xing et al., 2013a, 2013b). All rings were initially contracted by addition of 10 nmol·L⁻¹ phenylephrine (PE) (to evaluate the contractility) and then relaxed to 1 mmol·L⁻¹ ACh (to assess the endothelial integrity). Then a sustained contraction was induced by PE (10^{− 5} mol·L⁻¹), and ACh was added cumulatively to evoke relaxations. Contraction to PE (10^{− 5} mol·L⁻¹) and subsequent vasodilatation to insulin (10^{−1 0} to 10^{− 6} mol·L⁻¹), ACh (10^{− 9} to 10^{− 4} mol·L⁻¹) and sodium nitroprusside (SNP, 10^{−1 0} to 10^{− 5} mol·L⁻¹) were tested in all vessels. Subsequently, we evaluated the vasodilator responses of mesentery artery rings preconstricted with PE; vasorelaxation was evoked by increasing doses of insulin (10^{−1 0} to 10^{− 6} mol·L⁻¹). The maximal contraction induced by PE was considered as the baseline and vasorelaxant responses are expressed as % reduction in contraction. We obtained concentration‐response curves to ACh, insulin and SNP on different segments from the same artery.

Cell culture

Human artery endothelial cell line EA.hy 926 (Summers et al., 2014) and HUVECs (obtained from American Type Culture Collection) were used. Endothelial cells at passages 4–8 were cultured on gelatin‐coated flasks in Dulbecco's minimum essential medium supplemented with 10% FBS (Invitrogen, Carlsbad, CA, USA) under endotoxin‐free conditions (Armani et al., 2014). Cells were cultured in an atmosphere of 5% CO₂ at 37°C. One day before treatment, cells were trypsinized and allowed to grow to about 80% confluence. Afterwards, fresh media supplemented with vehicle (DMSO) or berberine were added to the cells as indicated for 1 h. We used HG/HF (25 mmol·L⁻¹ glucose + 500 μmol·L⁻¹ palmitate)‐treated endothelial cells to simulate the in vivo diabetic cell condition. For the insulin stimulation experiments, 1 or 100 nmol·L⁻¹ of bovine insulin was added to the culture medium (Tanigaki et al., 2009). Cells were treated for an additional 60 min before harvest. Berberine was dissolved in DMSO and was added to the medium at the indicated concentrations.

Cell viability

The 3‐[4,5 dimethylthiazol‐2‐yl]‐2,5‐diphenyl tetrazolium bromide thiazolyl blue (MTT) assay was conducted as described previously (Hao et al., 2011). Briefly, the MTT was dissolved in PBS at a concentration of 5 mg·mL⁻¹ and sterilized by passage through a 0.22 μm filter. EA.hy 926 cells were seeded in wells of a 96‐well plate containing 200 μL of the culture medium, and 20 μL MTT stock solution was then added to each well. After incubation for 4 h at 37°C, 150 μL of a DMSO solution were added to all of the wells and mixed thoroughly to lyse the cells and dissolve the dark blue crystals. After 10 min, the absorbance was read on a microplate reader (Beckman DU640, Brea, CA, USA) at a wavelength of 570 nm.

Transmission electron microscopy

HUVECs were quickly fixed with 2.5% glutaraldehyde in 0.1 mol·L⁻¹ phosphate buffer (pH 7.4) overnight at 4°C. After fixation, cells were then immersed in 1% osmium tetroxide for 2 h, dehydrated in graded ethanol and then embedded in epoxy resin. After that, cells were post‐stained with uranyl acetate and lead citrate, and then examined under a JEM‐1230 transmission electron microscope (JEOL, Tokyo, Japan).

Small interfering (si) RNA design and transfection

The cognate siRNA against insulin receptors (InsR‐homo‐3428) was designed as described previously and purchased from Genepharm (Shanghai, China) along with a scrambled control siRNA (siCONTROL non‐targeting siRNA, scrambled) (Xing et al., 2013a, 2013b). HUVECs were seeded and transfected with siRNA to a final concentration of 20 nmol·L⁻¹ using RNA mate (Genepharma, Shanghai, China) when cells reached 30–50% confluence. Protein knockdown efficiencies were assessed at 72 h after transfection.

Western blot

HUVECs were washed twice with ice‐cold 1× PBS solution and lysed in ice‐cold lysis buffer and proteins were extracted as described previously (Dong et al., 2008). Protein samples were subjected to SDS‐PAGE and transferred to polyvinylidene difluoride membranes by a semi‐dry transfer cell (Bio‐Rad, Hercules, CA, USA). For detection of InsR‐β‐1150, Ser⁴⁷³‐phosphorylated Akt and total Akt, LC3B and p62 were detected by specific rabbit polyclonal antibodies (Cell Signaling Technology, Danvers, MA, USA). Anti‐phospho‐eNOS (Ser¹¹⁷⁷), anti‐eNOS antibodies, anti‐phospho‐AMPK and anti‐AMPK were obtained from Abcam (San Francisco, CA, USA) with GADPH as internal control.

Berberine, PE, acetylcholine, insulin, sodium palmitic acid, MTT and PI3K inhibitor wortmannin were obtained from Sigma‐Aldrich.

Statistical analysis

All values are presented as mean ± SEM. Differences were compared by ANOVA followed by Bonferroni correction for post hoc t‐test, where appropriate. Probabilities of <0.05 were considered to be statistically significant. All of the statistical tests were performed with the graphpad Prism software, version 5.0 (GraphPad Software, San Diego, CA, USA). The data and statistical analysis comply with the recommendations on experimental design and analysis in pharmacology (Curtis et al., 2015).

Results

Effects of berberine on blood glucose in diabetic rats

To examine the effects of berberine on blood glucose responses to an oral glucose challenge, the oral glucose tolerance test (OGTT) was performed during the last week of berberine treatment. Diabetic rats (Diabetes) showed significantly increased blood glucose with fasting glucose level of 15.3 ± 1.9 mmol·L⁻¹ and 2 h glucose level in the OGTT of 24.7 ± 1.3 mmol·L⁻¹. Compared with the diabetic control group, the berberine‐treated diabetic rats showed a significant improvement in oral glucose tolerance. The fasting glucose level of the berberine‐treated diabetic rats group was 4.0 ± 0.2 mmol·L⁻¹, and 2 h glucose level was 9.5 ± 1.6 mmol·L⁻¹. Berberine administration significantly lowered the blood glucose in diabetic rats, especially the fasting glucose level and the 2 h glucose level in OGTT (n = 10–15, P < 0.01).

Berberine improved ACh‐induced vasodilatation of mesenteric arteries in diabetic rats

Berberine is known to have endothelial cell‐ and smooth muscle cell‐dependent vasodilator effects (Ko et al., 2000; Lau et al., 2001). Previous studies also showed that berberine prevented hyperglycaemia‐induced endothelial injury and enhanced vasodilatation via AMPK and eNOS(Wang et al., 2009). In the present study, as shown in Figure 1, the relaxation response induced by ACh in PE‐precontracted, mesenteric artery rings from diabetic rats was significantly impaired compared with that of the control group. The Emax fell to 19.10 ± 8.56% in diabetic rats compared with 93.94 ± 2.10% in that of the control. Four weeks treatment with berberine (200 mg·kg⁻¹·day⁻¹) (Cui et al., 2009; Chen et al., 2014) significantly improved ACh‐induced relaxation in the mesenteric arteries from diabetic rats. The Emax reached 71.09 ± 12.77% compared with that of the Diabetes group (n = 12, P < 0.01) (Figure 1A). Berberine (200 mg·kg⁻¹·day⁻¹) treatment had no effect on ACh‐induced relaxation in the control group. Moreover, SNP‐induced endothelium‐independent relaxation was not different among the four groups (Figure 1B). These results indicate that berberine improves ACh‐induced vasodilatation of mesenteric arteries in diabetic rats.

Figure 1.

Berberine improved ACh‐induced vasodilatation of mesenteric arteries in diabetic rats. Mesenteric artery rings were isolated from control and diabetic rats treated with berberine for 4 weeks. (A) Berberine improved ACh‐induced vasodilatation of mesenteric arteries in diabetic rats. ACh‐induced relaxation in age‐matched control, berberine‐treated control, diabetic rats and berberine‐treated diabetic rat mesenteric arteries. (B) Sodium nitroprusside (SNP)‐induced relaxation in mesenteric artery. Relaxation is expressed as a percentage relative to the contraction induced by PE. Data are expressed as mean ± SEM (n = 12). Control, control rats; Control + BBR, control rats treated with berberine (200 mg·kg⁻¹·day⁻¹, gavage), Diabetes, diabetic rats; Diabetes + BBR, diabetic rats treated with berberine (200 mg·kg⁻¹·day⁻¹, gavage). ^** P < 0.01 versus Control; ^## P < 0.01 versus Diabetes.

Berberine improved insulin‐induced vasodilatation in mesenteric arteries of diabetic rats through PI3K/Akt pathway

Insulin mainly combines with the insulin receptor to stimulate the intracellular signalling pathway and induce vasorelaxation of vessels (Zeng et al., 2000; Vicent et al., 2003). Interestingly, insulin‐evoked vasorelaxation was significantly improved in isolated mesentery arteries from diabetic rats treated with berberine (200 mg·kg⁻¹·day⁻¹) for 4 weeks. This indicates that berberine significantly improved the insulin‐mediated vasodilatation in diabetic rats. Furthermore, in mesenteric arteries isolated from the diabetic rats treated with both berberine and wortmannin (Wm, 15 mg·kg⁻¹, i.p.), the inhibitor of PI3K/Akt, for 4 weeks, insulin‐evoked vasorelaxation almost disappeared (Figure 2), indicating that the effect of berberine on insulin‐induced vasorelaxation was mediated through the PI3K/Akt pathway.

Figure 2.

Berberine increased insulin‐induced vasodilatation in mesenteric arteries in diabetic rats via the PI3K/Akt pathway. Insulin‐induced relaxation was improved in berberine‐treated diabetic rat's mesenteric arteries. Relaxation is expressed as a percentage relative to the contraction induced by PE. Data are expressed as mean ± SEM (n = 12). Control, control rats; Diabetes, diabetic rats; Diabetes + BBR, diabetic rats treated with berberine (200 mg·kg⁻¹·day⁻¹, gavage); Diabetes + BBR + Wm, diabetic rats treated with berberine (200 mg·kg⁻¹·day⁻¹, gavage) and Wm (15 mg·kg⁻¹·day⁻¹, i.p.) for 4 weeks. ^** P < 0.01 versus Control; ^## P < 0.01 versus Diabetes; ^†† P < 0.01 versus Diabetes + BBR.

Berberine combined with insulin alleviated the impaired vasodilatation in isolated mesenteric arteries of diabetic rats ex vivo

We performed ex vivo experiments to further determine the effect of berberine on vasodilatation in isolated mesenteric arteries from normal, control, and diabetic rats. Berberine (from 2.5 to 10 μmol·L⁻¹) was directly added to the bath solution of artery rings. It has been shown that berberine (from 2.5 to 10 μmol·L⁻¹) elicits a vasodilator response in PE‐preconstricted mesenteric artery rings (Wang et al., 2009). As shown in Figure 3A, berberine (2.5 μmol·L⁻¹) did not produce any significant constriction or relaxation in control and diabetic mesenteric arteries, whereas it caused significant relaxation at concentrations of 5 μmol·L⁻¹, 76.11 ± 4.04% relaxation, and at 10 μmol·L⁻¹, 89.05 ± 6.06% relaxation, in the control group. In contrast, in the diabetic group, the vasodilator effects of berberine were significantly decreased (P < 0.01, n = 5); it induced a relaxation response of 18.45 ± 4.60% at 5 μmol·L⁻¹, and 62.10 ± 14.08% at 10 μmol·L⁻¹. The vasodilatation evoked by berberine was concentration‐dependent in both the control and diabetic rat mesenteric arteries, as shown in Figure 3A.

Figure 3.

Berberine combined with insulin alleviated the vascular dysfunction in isolated mesenteric arteries of diabetic rats. Mesenteric artery rings were isolated and the vasodilator responses were assessed. (A) Berberine improved vasodilatation in mesentery arteries in control and diabetic rats in a dose‐dependent (2.5 μmol·L⁻¹, 5 μmol·L⁻¹ and 10 μmol·L⁻¹) manner. Data are expressed as mean ± SEM (n = 5). ^** P < 0.01 versus Control. (B) When combined with insulin (1 nmol·L⁻¹), berberine (5 μmol·L⁻¹) increased the vasodilatation of PE‐precontracted mesenteric artery rings isolated from control and diabetic rats. Data are expressed as mean ± SEM (n = 5). ^* P < 0.05 versus Control; ^## P < 0.01 versus diabetic mesenteric artery rings treated with berberine (5 μmol·L⁻¹).

Insulin at 100 nmol·L⁻¹ induced a vasorelaxation response of 30% in mesenteric arteries, whereas a lower concentration of insulin (1 nmol·L⁻¹) had little effect. In contrast and more importantly, insulin (1 nmol·L⁻¹) combined with berberine (5 μmol·L⁻¹) significantly ameliorated the impaired vascular relaxation of PE‐preconstricted arteries from diabetic rats, as shown in Figure 3B. This suggests that insulin combined with berberine exhibits synergistic beneficial effects on impaired vascular relaxation in diabetic rats.

Berberine increased cell viability and autophagy in HG/HF‐treated endothelial cells

Elevated plasma‐free fatty acid level is a common feature of both poorly‐controlled type 1 and type 2 diabetes, and is associated with obesity and metabolic syndrome (Yan et al., 2013). One of the most common dietary fatty acids is palmitate, a 16‐carbon saturated fatty acid. We used HG/HF‐treated EA.hy 926 cells as the simulated diabetic endothelial cell model and found that HG/HF (25 mmol·L⁻¹ glucose + 500 μmol·L⁻¹ palmitate) decreased cell viability. A low concentration of insulin, 1 nmol·L⁻¹, and berberine at 50 μmol·L⁻¹ had no significant effects on cell viability as assessed by the MTT assay. However, interestingly, exposure of the cells to insulin (1 nmol·L⁻¹) combined with berberine (50 μmol·L⁻¹) resulted in a significant increase in cell viability (101 ± 9.19% vs. 71 ± 8.49%, n = 6, P < 0.01) compared with insulin (1 nmol·L⁻¹) (Figure 4A).

Figure 4.

Berberine increased cell viability and autophagy in HG/HF‐treated endothelial cells. EA.hy 926 cells were incubated for 24 h in medium supplemented with HG/HF (25 mmol·L⁻¹ glucose + 500 μmol·L⁻¹ palmitate) and 0.5% FBS in the presence or absence of insulin (1 nmol·L⁻¹ or 100 nmol·L⁻¹), or berberine (50 μmol·L⁻¹). (A) Berberine increased the viability of HG/HF‐treated endothelial cells. Cell viability was assessed with MTT assay. Data are expressed as mean ± SEM (n = 6). ^* P < 0.05 versus Control; ^# P < 0.05 versus HG/HF; ^† P < 0.05 versus HG/HF + insulin (1 nmol·L⁻¹). (B) Berberine increased autophagy in HG/HF‐treated HUVECs. HUVECs were fixed and examined using transmission electron microscopy for autophagosomes. Control, control culture medium; Control + BBR, control culture medium + berberine (50 μmol·L⁻¹); HG/HF, HG/HF (25 mmol·L⁻¹ glucose + 500 μmol·L⁻¹ palmitate) culture medium; HG/HF + BBR, HG/HF culture medium treated with berberine (50 μmol·L⁻¹). (C) Berberine increased autophagy‐related protein expression as analysed by western blots. Quantification of band intensity was normalized to GAPDH. The relative quantification of the Western blot band intensities are presented as mean ± SEM (n = 5–7). ^* P < 0.05 versus HG/HF.

Autophagy is involved in insulin resistance (Lumeng and Saltiel, 2006; Yang et al., 2010). We assessed whether berberine inhibits autophagy in HG/HF‐treated HUVECs. There was no difference in the autophagosome formation between the control and control + berberine groups, whereas berberine increased the formation of autophagosomes in HG/HF‐treated endothelial cells (Figure 4B). Then western blot analysis was carried out to measure autophagy‐associated biomarkers LC3B (microtubule‐associated protein 1 light chain 3). Conversion of the non‐lipidated form of LC3 (LC3‐I to the autophagosome membrane‐associated lipidated form, LC3‐II) and degradation of the autophagy substrate protein p62 were tested. The results showed that berberine increased autophagic activity, as demonstrated by an increase in LC3B‐II/LC3B‐I and decrease in p62 expression (Figure 4C).

Co‐treatment with berberine and insulin increased the phosphorylation of InsR, Akt and eNOS in HG/HF‐treated HUVECs

Previous studies have indicated that insulin triggers the phosphorylation of cardiovascular Akt‐eNOS (Gao et al., 2002). In the present study, we determined whether berberine could affect the phosphorylation of the downstream molecules of insulin receptor‐mediated signalling. Non‐stimulated endothelial cells had either no or only a low level of phosphorylated InsR, AMPK, Akt and eNOS. Berberine treatment (50 μmol·L⁻¹) for 1 h increased InsR, AMPK, Akt and eNOS phosphorylation 1.5–2.5 fold compared with their corresponding basal levels, as shown in Figure 5A. Insulin (at both 1 nmol·L⁻¹ and 100 nmol·L⁻¹) increased the phosphorylation of InsR, Akt and eNOS. The combination of insulin (1 nmol·L⁻¹) with berberine (50 μmol·L⁻¹) resulted in a significant increase in phosphorylated InsR, Akt and eNOS, as shown in Figure 5B. These findings suggest that berberine improves insulin receptor‐mediated signalling.

Figure 5.

Co‐treatment with berberine and insulin increased the HG/HF‐induced phosphorylation of InsR, Akt and eNOS in HG/HF‐treated endothelial cells. Berberine increased the protein expression of insulin receptor‐mediated signalling as analysed by western blots. (A) The effects of berberine on the phosphorylation of InsR, AMPK, Akt and eNOS. Quantification of band intensity was normalized to the corresponding total protein levels. The relative quantification of the western blot band intensities are presented as mean ± SEM (n = 5–9). ^* P < 0.05, ^** P < 0.01 versus HG/HF. (B) Co‐treatment with berberine and insulin increased HG/HF‐induced phosphorylation of InsR, Akt and eNOS in HG/HF‐treated endothelial cells. Representative bands are shown. Quantification of band intensity of p‐InsR, p‐Akt and p‐eNOS was normalized to InsR, Akt and eNOS protein levels respectively. The relative quantification of the western blot band intensities are presented as mean ± SEM (n = 5). ^** P < 0.01 versus Control; ^# P < 0.05, ^## P < 0.01 versus HG/HF; ^† P < 0.05, ^†† P < 0.01 versus HG/HF + Ins (1 nmol·L⁻¹).

Berberine‐induced phosphorylation of AMPK, Akt and eNOS were blunted following insulin receptor knockdown using small interfering RNA (siRNA)

To further investigate the mechanism underlying berberine's effect on the insulin receptor, HUVECs were transfected with pre‐designed siRNA for insulin receptors or scrambled control RNA for insulin receptors. When cells had reached 30–50% confluence, they were seeded and transfected with siRNA to a final concentration of 20 nmol·L⁻¹ using RNA mate (Genepharma, Shanghai, China). Then the cells were exposed to HG/HF for 24 h and treated with berberine (50 μmol·L⁻¹) for 1 h. The results show that the expression of InsRs was decreased by ~85% in the insulin receptor knockdown group, as shown in Figure 6. This knockdown of insulin receptors significantly reduced insulin receptor protein expression and blunted berberine‐induced phosphorylation of AMPK, Akt and eNOS, as shown in Figure 6.

Figure 6.

Berberine‐induced phosphorylations of AMPK, Akt and eNOS were blunted following insulin receptor knockdown using siRNA in HG/HF‐treated endothelial cells. Semi‐quantification of p‐AMPK was normalized to the total AMPK protein levels and presented as mean ± SEM (n = 5). ^* P < 0.05 versus siRNA for InsR + BBR.

Discussion

Our findings demonstrate that berberine improves vascular insulin sensitivity of the mesenteric arteries and protects against endothelial dysfunction in diabetic rats mainly through up‐regulation of insulin receptor‐mediated signalling.

The extent of endothelium‐dependent vasodilatation in obese and insulin‐resistant subjects correlates with their individual insulin sensitivity (Steinberg et al., 1996; Rask‐Madsen and King, 2007). Berberine has a wide range of pharmacological actions; it has been shown to have anti‐diarrheic, anti‐inflammatory (Amin et al., 1969; Anis et al., 2001) and anti‐metabolic (Kong et al., 2004; Turner et al., 2008) effects, and protect against hypertrophy (Hong et al., 2003) or ischaemia reperfusion injury (Zeng et al., 2003). In the present study, we aimed to evaluate whether berberine can ameliorate endothelial dysfunction of mesenteric arteries in diabetic rats. The behaviour of mesenteric artery rings is thought to be a useful proxy for the function of vascular beds that contribute more directly to blood pressure. We found that diabetic rats treated for 4 weeks with berberine exhibited substantially improved vasodilatation, particularly vasodilator responses to insulin compared with vehicle‐treated diabetic animals. Notably, the magnitude of this improvement mediated by berberine was blunted by pretreatment with wortmannin. Although wortmannin is not a totally specific inhibitor for PI3K/Akt, its ability to completely abolish the vasodilatation induced by berberine strongly suggests that the vasodilator effects of berberine are mediated via a PI3K/Akt‐dependent mechanism. Previous studies have suggested that berberine protects against endothelial injury and enhances vasodilatation through activation of the AMPK signalling cascade (Wang et al., 2009). Our results indicate that in diabetic rats the PI3K‐Akt signalling pathway is involved in the vasodilator effects of berberine. This discrepancy may be related to differences in the experimental systems studied, including the animal models. In the present study, we used mesenteric artery rings isolated from STZ‐induced diabetic rats instead of high glucose‐perfused artery rings.

To further assess berberine's vasodilator action and its synergistic effects with insulin, we performed an ex vivo experiment in isolated mesenteric arteries and found that infusion of berberine in combination with a low concentration of insulin significantly ameliorated the impaired vasodilatation of mesenteric arteries isolated from diabetic rats (Figure 3), suggesting synergistic effects between insulin and berberine. We found that diabetic rats exhibited a reduced vasorelaxant response to insulin, while berberine treatment enhanced these insulin‐induced endothelium‐dependent vasorelaxant responses, which is regarded as one of the important indicators reflecting vascular insulin sensitivity (Xing et al., 2013a, 2013b). Insulin has important vascular actions; it stimulates the production of nitric oxide from endothelium, leading to capillary recruitment, vasodilatation, increased blood flow and subsequent augmentation of glucose disposal in classic insulin target tissues, while in diabetic animals, the reduced insulin sensitivity often leads to a reduced vasorelaxant effect of insulin (Muniyappa et al., 2007).

In diabetes, the relative ineffectiveness of insulin and hyperglycaemia act together to reduce the activity of InsR/PI3K/Akt signalling, resulting in an impaired InsR/Akt‐mediated endothelial function. Studies have reported that berberine is a potent sensitizer of the insulin signal and is derived from anti‐diabetic plants (Chen et al., 2009; Chen et al., 2010; Di Pierro et al., 2012; Gomes et al., 2012). Our data revealed that phosphorylation of the InsR, AMPK, Akt and eNOS increased in response to a combined treatment of insulin with berberine (Figure 5). Previous study showed that insulin receptor substrates IRS‐1 and IRS‐2 activate the Akt‐mTOR signalling pathway. At the same time, an increased AMP/ATP ratio activates AMPK. Increased AMPK activity further antagonizes mTOR/Raptor‐mediated protein synthesis and up‐regulation of fatty acid oxidation (Long et al., 2011). In our study, the down‐regulation of insulin receptor expression significantly reduced berberine‐stimulated phosphorylation of the insulin receptor, AMPK, Akt and eNOS (Figure 6), suggesting that berberine improves insulin sensitivity by up‐regulating insulin receptor‐mediated signalling in diabetic rats. Many studies have shown that insulin suppresses rather than activates AMPK (Witters and Kemp, 1992; Gamble and Lopaschuk, 1997; Winder and Holmes, 2000), whereas insulin has been shown to activate the AMPK signalling pathway in adipocytes and that this is associated with an elevation of the AMP/ATP ratio (Liu et al., 2010). AMPK is also known to regulate insulin sensitivity by inhibiting IRS‐1 (Jakobsen et al., 2001; Qiao et al., 2002; Ning and Clemmons, 2010). Our data provide novel evidence that berberine improves endothelium‐mediated vasodilatation in diabetes via a mechanism involving both Akt and AMPK activation through the up‐regulation of the insulin receptor.

A deficiency in autophagy leads to a deterioration of whole‐body energy metabolism and the development of metabolic disorders (Zhang et al., 2014a, 2014b), and impaired autophagy in endothelial cells is believed to contribute to endothelial dysfunction and diabetes (Ho et al., 2000; Cifarelli et al., 2011). Recent studies have shown that a chronic high‐fat diet or obesity impairs autophagy in various cells and tissues by blocking the formation of autophagosome and fusion between autophagosome and lysosome (Guo et al., 2013; Armani et al., 2014). Notably, inhibition of autophagy subsequently leads to the development of insulin resistance (Liu et al., 2009; Yang et al., 2010). Zhang et al. have reported that berberine promotes autophagy, subsequently attenuating left ventricular remodelling and cardiac dysfunction after myocardial ischaemia (Zhang et al., 2014a, 2014b). Similar results have also been obtained indicating that berberine attenuates left ventricular remodelling and cardiomyocyte apoptosis through an autophagy‐dependent mechanism in a rat model of cardiac hypertrophy, which is, at least in part, associated with enhanced autophagy through inhibition of the mTOR, p38 and ERK1/2 MAPK signalling pathways (Li et al., 2014). We found that berberine treatment enhanced autophagy in HG/HF‐induced diabetic endothelial cells as demonstrated by the increase in autophagosomes and protein expression of LC3B‐II/LC3B‐I (Figure 4B, C). Autophagy is a complex process that involves numerous signalling molecules and one of the most important positive regulated mechanisms is AMPK signalling in endothelial cells (Goyal et al., 2014). It has also been reported that AMPK phosphorylation directly leads to autophagosome formation (Guo and Ren, 2012).

The mechanism underlying the hypoglycaemic effect of berberine remains largely unclear. Berberine has been found to activate insulin signalling through the inhibition of PTP1B activity (Chen et al., 2010), and has, more recently, been shown to increase insulin secretion from islets of diabetic mice via the activation of AMPK and uncoupling protein‐2 (UCP2) (Liu et al., 2014). Another reasonable explanation for the effects of berberine is through restoration of uncoupling protein‐2 expression in high glucose (HG)‐treated INS‐1E cells and rat islets or in db/db mouse islets, and UCP2 may regulate glucose‐stimulated insulin secretion (Liu et al., 2014). Also, it has been suggested that berberine reduces endothelium‐dependent contractions by activating AMPK, thus inhibiting ER stress and subsequently scavenging ROS leading to COX‐2 down‐regulation in SHR carotid arteries (Liu et al., 2015).

In conclusion, our data show that berberine improves insulin‐mediated vasodilatation of mesenteric arteries in diabetic rats through up‐regulating insulin receptor‐mediated signalling and increasing vascular insulin sensitivity. These findings suggest that berberine has the potential to have a beneficial role in the preventive and adjunctive treatment of vascular complications associated with diabetes (summarized in Figure 7).

Figure 7.

Schematic figure illustrating the enhancement by berberine of vascular insulin‐mediated signalling and vasodilatation. In addition to increasing AMPK‐induced autophagy, berberine up‐regulates the phosphorylation of the insulin receptor, resulting in enhanced activation of insulin‐mediated signalling and subsequent endothelial NO production with resultant vasodilatation in diabetic rats.

Author contributions

F‐H.G., G‐H.L., Z‐J.Z. and Y.Z. carried out the laboratory experiments, collected and analysed the data and interpreted the results. X.Z., P.Z. and M‐Q.D. analysed and interpreted the data. L.D. and F.G. designed the experiments, supervised the study, drafted and revised the manuscript.

Conflict of interest

The authors declare no conflicts of interest.

Declaration of transparency and scientific rigour

This Declaration acknowledges that this paper adheres to the principles for transparent reporting and scientific rigour of preclinical research recommended by funding agencies, publishers and other organizations engaged with supporting research.

Geng, F.‐H. , Li, G.‐H. , Zhang, X. , Zhang, P. , Dong, M.‐Q. , Zhao, Z.‐J. , Zhang, Y. , Dong, L. , and Gao, F. (2016) Berberine improves mesenteric artery insulin sensitivity through up‐regulating insulin receptor‐mediated signalling in diabetic rats. British Journal of Pharmacology, 173: 1569–1579. doi: 10.1111/bph.13466.