Effect of silibinin on endothelial dysfunction and ADMA levels in obese diabetic mice

All procedures were carried out in accordance with the Italian Guidelines for the Care and Use of Laboratory Animals and with the ARRIVE guidelines [25]. Eight-week-old male BKS.Cg-m+/+ Leprdb/J (db/db) obese mice and eight-week-old male heterozygous db/m lean control mice were purchased from Charles River Lab (Calco, Italy). Animals were housed at constant room temperature (23°C) (n = 3 per cage) under 12 hours light/dark cycles with ad libitum access to water and were fed a standard diet. Silibinin dihydrogen disuccinate disodium (Madaus, Köln, Germany) was dissolved in saline which was also used as vehicle for the placebo-treated animals. Eighteen mice were distributed in 3 groups: group I (db/m + vehicle) included six heterozygous db/m mice treated with saline; group II (db/db + vehicle) comprised six db/db mice treated with saline; group III (db/db + silibinin) included six db/db mice treated with silibinin (20 mg/Kg i.p. injection, daily). At 12 weeks of age, after an overnight fast, animals were anesthetized and sacrificed; blood, aorta and liver samples were obtained for further analysis.

Blood glucose was measured by Accu-chek (Roche Diagnostics, Milan, Italy). Serum insulin was determined using an automated enzyme immunoassay analyzer from Tosoh (Tokio, Japan). HOMA-IR was calculated as follows: [fasting glucose (mmol/L) × fasting insulin (mU/L)]/22.5. Plasma and tissue ADMA concentration was determined by using a commercially available enzyme-linked immunosorbent assay kit (DLD Diagnostika GmbH, Hamburg, Germany) according to the manufacturer's instructions.

Aortas were immersed in physiological salt solution (composition, mmol/L: NaCl, 118; KCl, 4.6; NaHCO₃, 25; MgSO₄, 1.2; KH₂PO₄, 1.2; CaCl₂, 1.2; glucose, 10; EDTA, 0.025; pH 7.4 at 37°C), dissected out of surrounding connective tissue, cut in segments (2 mm length) and mounted in a wire myograph (610 M, Danish Myo Technology, Aarhus, Denmark), by using 40 μm-diameter stainless steel wire, for isometric record of contractile force. After mounting, each preparation was equilibrated unstretched, for 30 min, in physiologic salt solution, maintained at 37°C and aerated with a gas mixture 95% O₂ - 5% CO₂. The normalized passive resting force and the corresponding diameter were then determined for each preparation from its own length-pressure curve, according to Mulvany and Halpern [26]. Contractile responses were recorded into a computer, by using a data acquisition and recording software (Myodaq and Myodata, Danish Myo Technology). After normalization and 30-min equilibration in physiological solution, the preparations were constricted with isotonic depolarizing KCl solutions, in which part of NaCl had been replaced by an equimolar amount of KCl (composition, mmol/L: NaCl, 22.6; KCl, 98.8; NaHCO₃, 25; MgSO₄, 1.2; KH₂PO₄, 1.2; CaCl₂, 1.2; glucose, 10; EDTA, 0.025, pH 7.4 at 37°C). After washout and 30-min recovery, the preparations were exposed to cumulative concentrations of phenilephrine (PE, 1 nmol/L - 1 μmol/L). Once the vasoconstriction to PE had reached steady state, cumulative concentrations of acetylcholine (ACh, 1 nmol/L - 1 μmol/L) were added to the organ chamber. Relaxing responses were expressed in percentage of pre-existing contractile tone, induced by PE. After wash out and 30 min recovery, preparations were exposed to 5 nmol/L PE, which induced a negligible (< 2% of K⁺100) vasoconstriction in lean, but a stronger one in db/db (17-18% of K⁺100); once PE-induced tone had reached a steady state, preparations were challenged with cumulative concentrations of asymmetric N^G, N^G-dimethyl-L-arginine (ADMA, 300 nmol/L - 300 μmol/L). Preparations were then further constricted by 1 μmol/L PE in the presence of 100 μmol/L N^G-nitro-L-arginine (LNNA), once PE-induced tone had reached a steady state, preparations were challenged with cumulative concentrations of sodium nitroprusside (SNP, 1 nmol/1 L - 1 μmol/L), to assess endothelium-independent vasodilatation. PE, ACh, SNP, LNNA and ADMA were from Sigma (St. Louis, MO, U.S.A.); 10 mmol/L stock solutions were prepared in water and further diluted in physiological salt solution as required, except for ADMA, which was dissolved in water as 30 mmol/L stock solution. Concentration-response curves to vasoactive agonists were plotted as response (for vasoconstrictors, in % of vasoconstriction induced by KCl in the same arteries; for vasodilators in % of residual tone) against log molar concentration of drug. Each set of data points was curve-fitted by a non-linear regression, best-fit, sigmoid dose-response curve with no constraints, with the use of Graph Pad Prism (Graph Pad Software, San Diego, CA). Pharmacological parameters (concentration producing 50% of maximum effect or EC₅₀ and maximum effect or E_max) were calculated from these non-linear fits. Each curve represents 12 preparations from 6 different mice.

Statistics were aided by Graph Pad Prism. All results were expressed as mean ± standard error of mean (S.E.M.). P values less than 0.05 were considered significant. Biochemical data were analyzed by one-way ANOVA with Bonferroni Post-Hoc analysis. Concentration-response curve from isolated aortas were compared by two-way ANOVA.

All db/db mice weighted more than lean controls at week 8, before starting the treatment; silibinin administration did not significantly change body weight or food intake (data not shown). Untreated db/db were insulin resistant, as shown by HOMA-IR; silibinin decreased fasting glucose and insulin, reversing insulin resistance (Table 1).

Plasma ADMA levels were higher in db/db than in db/m (Figure 1A). Silibinin reduced plasma ADMA in db/db to a level lower than in db/m animals. These differences in ADMA levels were even more pronounced in aorta, where, again, ADMA was higher in db/db compared to db/m and was reduced by silibinin in db/db mice, by more than 50% (Figure 1B). Liver ADMA was slightly lower in db/db as compared to lean controls, although the difference was not statistically significant, whereas silibinin led to raised ADMA in hepatic tissue (Figure 1C). Plasma and aorta ADMA levels were positively correlated (Figure 1D); by contrast, liver ADMA was inversely correlated with both plasma and aorta ADMA concentrations (Figure 1E).

Contractile tone induced by high K⁺ was higher in aorta preparations from db/db than in db/m mice and was not influenced by silibinin treatment (db/m + vehicle, 3.42 ± 0.25 mN/mm; db/db + vehicle, 4.33 ± 0.35 mN/mm; db/db + silibinin, 4.36 ± 0.27 mN/mm; db/m + vehicle versus all db/db, P < 0.05).

Vasoconstriction to PE, either expressed as raw values (mN/mm, not shown) or normalized in % of K⁺-induced contraction (Figure 2A and Table 2) was also higher in db/db; in particular concentration-contraction curve to PE was displaced to the left in db/db mice, in a parallel manner. Vasoconstriction to PE was not modified by silibinin administration. Endothelium-dependent vasodilatation to ACh and, to a less extent, endothelium-independent vasodilatation to SNP were lower in db/db mice as compared to lean animals (Figure 2B, C and Table 2); vasodilatation to ACh was significantly increased by silibinin, whereas endothelium-independent vasodilatation to SNP was not modified by the treatment. In vitro challenge with ADMA induced a stronger contractile tone in aorta from db/db mice than in db/m (Figure 2D); of note, the tone induced by applying ADMA in vitro in aorta from db/db mice, was not influenced by in vivo silibinin.