Contribution of NO, Prostacyclin and Endothelium-derived Hyperpolarizing factors (EDHF) in uterine arteries from pregnant mice deficient in endothelial cell BH4.
We next determined the relative contributions of the endothelium-derived vasodilators NOS, prostacyclin and EDHF in uterine arteries of pregnant mice lacking endothelial cell BH4. Firstly, we found that in the presence of L-NAME, endothelium-dependent vasodilatation was significantly inhibited in uterine arteries from pregnant and non-pregnant wild-type and non-pregnantGch1fl/fl Tie2cre mice, but was unaltered in uterine arteries from pregnant Gch1fl/fl Tie2cre mice (Figure 3A and B), adding further evidence to the loss of eNOS-mediated vasodilator function in uterine arteries of pregnantGch1fl/fl Tie2cre mice.
In the presence of L-NAME, addition of indomethacin had minimal effect on endothelium-dependent vasodilation in uterine arteries from either pregnant or non-pregnant wild-type andGch1fl/fl Tie2cre mice (Figure 3A and B) . In contrast, addition of EDHF blockers (combination of apamin and charybdotoxin) totally inhibited the non-NOS mediated component of endothelium-dependent vasodilation in uterine arteries from both non-pregnant and pregnant mice of both genotypes with a significantly greater augmentation observed in uterine arteries fromGch1fl/fl Tie2cre mice (Figure 3A, B, and C) . Systematic quantification contribution of the vasodilator responses revealed a marked reduction that in contrast to wild-type mice, uterine arteries from pregnantGch1fl/fl Tie2cre mice revealed a striking loss of NOS-derived NO and a significant increase in the EDHF component (Figure 3C), suggesting loss of endothelial cell Gch1 /BH4 leads to a compensatory upregulation of EDHF inGch1fl/fl Tie2cre uterine arteries. However, this compensatory upregulation was insufficient to fully correct for the loss of NOS mediated dilators.
To further investigate which specific components of the EDHF response is affected in Gch1fl/fl Tie2cre mice, endothelium-dependent vasodilatation was assessed in the presence of L-NAME and indomethacin with either charybdotoxin or apamin. First, we found that in a normal pregnancy in wild-type mice, intermediate and large Ca2+-activated K+ channels (charybdotoxin-sensitive component) make up the majority of the EDHF component (Figure 4A and B) with the apamin-sensitive component accountable for approximately a third of the EDHF response (Figure 4A and B). In contrast to wild-type mice, loss of endothelial cellGch1 leads to a significant reduction in the apamin-sensitive component with the charybdotoxin-sensitive component dominating. This effect was seen even when the order of the inhibitors used was changed (Supplementary Figure 2). Taken together, these observations indicated that loss of endothelial cell BH4 impacts on the ability of endothelial cells to upregulate small Ca2+-activated K+ channel mediated EDHF responses.