C1q/TNF-related proteins, a family of novel adipokines, induce vascular relaxation through the adiponectin receptor-1/AMPK/eNOS/nitric oxide signaling pathway

Abstract

Objective: Reduced plasma adiponectin (APN) in diabetic patients is associated with endothelial dysfunction. However, APN knockout animals manifest modest systemic dysfunction unless metabolically challenged. The protein family CTRPs (C1q/TNF-related proteins) has recently been identified as APN paralogs and some CTRP members share APN's metabolic regulatory function. However, the vasoactive properties of CTRPs remain completely unknown.

Methods and results: The vasoactivity of currently identified murine CTRP members was assessed in aortic vascular rings and underlying molecular mechanisms was elucidated in human umbilical vein endothelial cells. Of 8 CTRPs, CTRPs 3, 5, and 9 caused significant vasorelaxation. The vasoactive potency of CTRP9 exceeded that of APN (3-fold) and is endothelium-dependent and nitric oxide (NO)-mediated. Mechanistically, CTRP9 increased AMPK/Akt/eNOS phosphorylation and increased NO production. AMPK knockdown completely blocked CTRP9-induced Akt/eNOS phosphorylation and NO production. Akt knockdown had no significant effect on CTRP9-induced AMPK phosphorylation, but blocked eNOS phosphorylation and NO production. Adiponectin receptor 1, but not receptor 2, knockdown blocked CTRP9-induced AMPK/Akt/eNOS phosphorylation and NO production. Finally, preincubating vascular rings with an AMPK-inhibitor abolished CTRP9-induced vasorelaxative effects.

Conclusion: We have provided the first evidence that CTRP9 is a novel vasorelaxative adipocytokine that may exert vasculoprotective effects via the adiponectin receptor 1/AMPK/eNOS dependent/NO mediated signaling pathway.

Figure 1. CTRPs induced vasorelaxation in aortic vascular rings

(A) ACh and NaNO₂ induced similar vasorelaxation in vascular segments with intact endothelial function. (B) Vasorelaxation in cumulative concentrations of CTRP3, 5, 9 and APN. gCTRP9 exhibited most potent vasorelaxive action; gCTRP3 and 5 have comparable vasorelaxive potency as APN. n=5–7 mice/group. ^#indicates P<0.05, ^##indicates P<0.01. (C) gCTRP9 induced aortic vascular ring vasorelaxation with intact endothelium, but failed to do so in vascular rings without endothelium, evidencing CTRP9’s vasorelaxive effects to be endothelium-dependent. (D) gCTRP9-induced aortic vascular ring vasorelaxation is blocked when rings were pre-treated with L-NAME (NO synthesis inhibitor, 100 μmol/L). n=5–7 mice/group. ^#indicates P<0.05, ^##indicates P<0.01.

Figure 2. CTRP9 stimulated NO production, AMPK phosphorylation, and eNOS phosphorylation in time and dose dependent manner

(A) HUVEC incubation with gCTRP9 (0.3, 1, 3 μg/ml) or gAPN (10 μg/ml) for 15 minutes enhanced NO production in concentration-dependent manner. Treatment with L-NAME abolished gCTRP9-induced NO production enhancement. (B) Treatment of HUVECs with gCTRP9 or gAPN (3 μg/ml) for different time periods (5, 15, and 30 minutes) enhanced NO production, significantly increasing after 15 minutes, with increasing trend after 30 total minutes. Treatment with L-NAME abolished gCTRP9-induced NO production enhancement. HUVEC incubation with gCTRP9 (0.3, 1, 3 μg/ml) for 15 minutes enhanced AMPK (C), eNOS (D), and Akt (E) phosphorylation respectively. n=5–6 repeated experiments/condition. ^#indicates P<0.05, ^##indicates P<0.01.

Figure 3. CTRP9-induced eNOS activation and NO production in HUVECs is AMPK dependent

After AMPK siRNA knockdown, HUVECs were incubated with gCTRP9 (3 μg/ml) for 15 minutes. (A) Western blot analysis confirmed siRNA induced significant AMPK protein concentration reduction. AMPK knockdown significantly attenuated gCTRP9-induced Akt (B) and eNOS phosphorylation (C), and gCTRP9-induced NO production (D). n=5–6 repeated experiments/condition. ^##indicates P<0.01.

Figure 4. The role of Akt in CTRP9-induced eNOS activation and NO production in HUVECs

After Akt siRNA knockdown, HUVECs were incubated with gCTRP9 (3μg/ml) for 15 minutes. (A) Western blot analysis confirmed siRNA induced significant Akt protein concentration reduction. Akt knockdown had no significant effect on gCTRP9-induced AMPK phosphorylation (B) but significantly attenuated gCTRP9-induced eNOS phosphorylation (C) and gCTRP9-induced NO production (D). n=5-6 repeated experiments/condition. ^## indicates P<0.01.

Figure 5. The role of AdipoR1 and R2 in CTRP9-induced eNOS activation and NO production in HUVECs

After Adipo R1 and R2 siRNA knockdown, HUVECs were incubated with gCTRP9 (3μg/ml) for 15 minutes. Western blot analysis confirmed siRNA induced significant AdipoR1 and R2 protein concentration reduction (A). Knockdown of both AdipoR1 and R2 or AdipoR1 alone reduced gCTRP9-induced AMPK/Akt/eNOS phosphorylation (B) and NO production (C), while down-regulation of AdipoR2 alone did not have significant effects. (D): Representative co-immunoprecipitation results showing that CTRP9 binds AdipoR1. HUVECs were incubated with (right lane, +) or without (−, left lane) CTRP9 and cross-linked with 3.0 mmol/l dimethyl 3,3′-dithiopropionimidate dihydrochloride (DTBP) for 30 minutes at room temperature. Samples were immunoprecipitated with CTRP9-specific antibody followed by immunoblotting with antibodies against AdipoR1 (top) or CTRP9 (bottom). The experiments were repeated 5–6 times/condition.

Figure 6. CTRP9 promoted endothelial cell tube formation in vitro, and improved endothelium-dependent vasorelaxation in high fat diet (HFD) fed mice when administered in vivo

(A): CTRP9 promoted endothelial cell differentiation into tube-like structures when HUVECs were plated upon Matrigel matrix. ^##indicates P<0.01. In aortic segments from HFD (high fat diet) mice treated with vehicle, concentration-dependent vasorelaxation in response to ACh (B) is significantly attenuated whereas response to acidified NaNO₂ is normal (C), indicating that HFD-induced significant endothelial dysfunction. Treatment of HFD mice with fCTRP9 for 1 week significantly improved endothelial function as evidenced by downward-shift of concentration-dependent curve in response to ACh (B). n=5–7 mice/group. ^#P<0.05 and ^## P<0.01 vs. HFD mice treated with vehicle.