Hypoxia, AMPK activation and uterine artery vasoreactivity

Abstract

Genes near adenosine monophosphate-activated protein kinase-α1 (PRKAA1) have been implicated in the greater uterine artery (UtA) blood flow and relative protection from fetal growth restriction seen in altitude-adapted Andean populations. Adenosine monophosphate-activated protein kinase (AMPK) activation vasodilates multiple vessels but whether AMPK is present in UtA or placental tissue and influences UtA vasoreactivity during normal or hypoxic pregnancy remains unknown. We studied isolated UtA and placenta from near-term C57BL/6J mice housed in normoxia (n = 8) or hypoxia (10% oxygen, n = 7-9) from day 14 to day 19, and placentas from non-labouring sea level (n = 3) or 3100 m (n = 3) women. Hypoxia increased AMPK immunostaining in near-term murine UtA and placental tissue. RT-PCR products for AMPK-α1 and -α2 isoforms and liver kinase B1 (LKB1; the upstream kinase activating AMPK) were present in murine and human placenta, and hypoxia increased LKB1 and AMPK-α1 and -α2 expression in the high- compared with low-altitude human placentas. Pharmacological AMPK activation by A769662 caused phenylephrine pre-constricted UtA from normoxic or hypoxic pregnant mice to dilate and this dilatation was partially reversed by the NOS inhibitor l-NAME. Hypoxic pregnancy sufficient to restrict fetal growth markedly augmented the UtA vasodilator effect of AMPK activation in opposition to PE constriction as the result of both NO-dependent and NO-independent mechanisms. We conclude that AMPK is activated during hypoxic pregnancy and that AMPK activation vasodilates the UtA, especially in hypoxic pregnancy. AMPK activation may be playing an adaptive role by limiting cellular energy depletion and helping to maintain utero-placental blood flow in hypoxic pregnancy.

Figure 1. Phosphorylated AMPK (brown staining) is present in uterine arteries (UtA) and placentas from normoxic mice and increased by hypoxia in the UtA and the placental labyrinthine (LZ) but not junctional (JZ) or the decidua (DEC) zones

For the placenta specimens, panels A and B are with and without AMPK antibody, respectively, at 5× and panels C and D are with AMPK antibody at 20× magnification.

Figure 2. mRNA expression relative to beta‐actin, in human and murine placentas

Upper panel, mRNA for both AMPK α‐subunits and LKB1 is present at higher levels in human placentas from 3100 m vs. sea level. Lower panel, murine placentas also contain LKB1 and both AMPK α‐subunits at sea level, with expression levels for the α2‐subunit being greater in the labyrinthine than junctional zone. *P < 0.05, **P < 0.01, ****P < 0.0001. Data are means ± SEM.

Figure 3. Murine uterine artery relaxation in response to the pharmacological AMPK agonist A769662

A, uterine arteries pre‐constricted with a submaximal dose of PE from normoxic (n = 5) and hypoxic (n = 7) pregnant (day 19) mice relax in response to increasing concentrations of the AMPK agonist, A769662. B, the area under the curve (AUC), representing nitric oxide (NO)‐dependent and ‐independent components did not differ in the normoxic and hypoxic groups. Data are means ± SEM.

Figure 4. Murine uterine artery constriction to phenylephrine (PE) in control or A769662 treated vessels, with and without l‐NAME

Incubation with A769662 (1 × 10⁻⁴ m, open circles, dotted lines) decreased contractile sensitivity to phenylephrine (PE) in UtA from both normoxic (A, n = 8) and hypoxic (B, n = 6–7) pregnant mice compared to values obtained prior to incubation with A769662 (open circles, continuous lines). This decrease was significant in both groups when expressed as the area under the curve (AUC) (C, P < 0.05, two way ANOVA and post‐hoc Tukey's test). The AUC, representing the change in sensitivity to PE following incubation with A769662, was significantly greater in the hypoxic compared to the normoxic animals (C, P < 0.05, unpaired t test). Treatment with l‐NAME (1 × 10⁻⁵ m, filled squares and dashed lines) decreased the contractile response to PE in A769662‐treated vessels from normoxic (A) and hypoxic (B) pregnant mice, indicating a significant NO‐dependent component (both P < 0.05), but did not fully restore the contractile response. The contributions of both nitric oxide (NO)‐dependent and ‐independent mechanisms as well as the total relaxation response were greater in the UtA from hypoxic compared with normoxic animals (D, P < 0.05, unpaired t test). *P < 0.05. Data are means ± SEM.