Loss of endothelial cell BH4 during pregnancy caused vascular dysfunction uterine arteries from pregnant Gch1fl/flTie2cre mice.
We next determine how loss of endothelial cell Gch1 impacts on pregnancy induced vascular adaptations of the uterine artery. Firstly, the lumen diameter of uterine arteries, as determined by the length-tension relationship, was significantly increased in pregnancy in both genotypes with no difference between genotypes observed in either pregnancy or non-pregnancy mice (Pregnant WT: 258 ± 12.6 μm, PregnantGch1fl/fl Tie2cre: 276 ± 12.7 μm; Nonpregnant WT 155 ± 6.6 μm, nonpregnant Gch1fl/fl Tie2cre: 153 ± 3.1; P<0.05 ) (Figure 2A). This was accompanied by a greater KCl response (~2-fold increased) in uterine arteries from pregnant wild-type andGch1fl/fl Tie2cre mice (Pregnant WT: 5.8 ± 0.7 mN, Pregnant Gch1fl/fl Tie2cre: 5.1 ± 0.4 mN, nonpregnant WT: 2.6 ± 0.3 mN, nonpregnantGch1fl/fl Tie2cre: 2.3 ± 0.4;P<0.05 ) (Figure 2A), indicating that a normal pregnancy is associated with increased lumen size and media thickness of the uterine artery. However, loss of endothelial cell Gch1 in pregnancy lead to an increase in uterine artery stiffness, with increased wall tension in response to increasing stretch observed in uterine arteries from pregnancyGch1fl/fl Tie2cre mice compared with wild type controls (Supplementary Figure 1).
In contrast to the aorta, uterine arteries from pregnant wild type mice had a significantly attenuated contractile response, with a corresponding increase in the EC50 and decrease in Emax (% maximum contraction) to the TxA2 receptor against, U46619, (Figure 2B, C, and D). Incubation with the NOS inhibitor L-NAME lead to a significant augmentation of the contractile response (Figure 3E), indicating that the pregnancy induced attenuation was due in part to increased NOS-derived NO.
However, uterine arteries from pregnantGch1fl/fl Tie2cre mice showed incomplete adaptation to pregnancy with a significantly greater contractile response observed compared with arteries from pregnant wild type mice, as demonstrated by an increased maximum constriction and reduced EC50 (Figure 2B, C, and D). This difference appeared to be driven in part by a reduced NOS dependent production of vasodilators in uterine arteries from pregnantGch1fl/fl Tie2cre mice as in contrast to wild type mice. L-NAME did not alter the vasoconstrictor response in uterine arteries from pregnant Gch1fl/fl Tie2cre mice (Figure 2E).
In wild-type uterine arteries, endothelium-dependent vasodilatation to ACh was significantly enhanced in pregnancy compared to nonpregnant wild-type controls (Figure 2F; P<0.05 ). This was accompanied by a significantly enhanced endothelium-independent vasodilatation in response to nitric oxide donor, sodium nitroprusside (SNP), indicating increased vascular smooth muscle sensitivity to downstream NO signalling pathway during pregnancy (Figure 2I).
In contrast, pregnancy induced enhanced endothelium-dependent vasodilatations was blunted in pregnantGch1fl/fl Tie2cre uterine arteries with no difference in endothelial cell mediated dilation between pregnancy and non-pregnant uterine arteries from knock out mice (Figure 2F, G and H). This blunted pregnancy induced vascular remodelling was observed despite the presence of pregnancy induced enhanced endothelium-independent vasodilatation in response to nitric oxide donor, SNP (Figure 2I).