Search

Introduction: Intrahepatic cholestasis of pregnancy (ICP) is characterised by elevated total serum bile acid (TSBA) concentrations and adverse fetal outcomes such as preterm labour (PTL) and intrauterine death (IUD); the risk of which is directly associated with the degree of maternal TSBA derangement. Assays for measurement of TSBA concentration are not readily available at every clinical laboratory nor are results provided immediately. The exact mechanisms of ICP-associated PTL and IUD are unknown, however the current hypotheses are that elevated bile acids increase sensitivity to oxytocin in the myometrium and induce a sudden cardiac arrythmia in the fetus, resulting in the increased rate of PTL and IUD respectively. ICP is currently commonly treated using ursodeoxycholic acid (UDCA) pharmacotherapy. Although there is some experimental and observational evidence that UDCA is cardioprotective, data from randomised controlled trial data suggests that it does not improve a composite of adverse fetal outcomes. We aimed to improve the monitoring of ICP via the development of a novel electrochemical bile acid biosensor that can function in a point-of-care setting. We also aimed to improve the mechanistic understanding of ICP-associated PTL and IUD via exposure of bile acids to a myometrial cell model and an observational study of the fetal cardiac phenotype affected by ICP. Materials and Methods: An electrochemical biosensor was developed using commercially available carbon screen-printed electrodes (CSPEs) that had been manually modified with the electron mediators meldola's blue (MDB) or methylene blue (MEB). Pooled human serum was spiked with 0-100 µmol/L of the bile acid taurocholic acid (TCA) along with the bile acid-specific enzyme 3α-hydroxysteroid dehydrogenase (3α-HSD) and oxidised coenzyme nicotinamide adenine dinucleotide (NAD+). The redox current generated from reduction of NAD+ (NADH) was measured via chronoamperometry. Myometrial biopsies were taken from women undergoing elective caesarean sections and the primary cells were subsequently isolated and cultured until passage 5 or 6. Cells were incubated with the intracellular calcium ([Ca2+]i) dye Fluo4-AM prior to administration of 10 nmol/L oxytocin (OT) (n=3), 100 µmol/L taurocholic acid (TCA) (n=3), a combination of OT and TCA (n=3), 10 nmol/L prostaglandin E2 (PGE2) (n=4) or F2α (PGF2α) (n=7) with or without concomitant administration of 100 µmol/L UDCA using HBSS as the vehicle. Optical recording of [Ca2+]i transients was performed via time-lapse imaging and the amplitude, time to peak, duration and time to decay of the [Ca2+]i transients was measured. Pregnant women who were ≥20 weeks of gestation who had either uncomplicated pregnancies (controls, n=43), untreated ICP (n=26) or UDCA-treated ICP were recruited for fetal ECG recordings (n=22). A separate cohort of controls (n=15), untreated ICP (n=36) and UDCA-treated ICP (n=40) cases provided umbilical venous blood samples at delivery. Overnight fetal ECG recording was conducted in the first cohort. A 2-hour period from this recording was taken and the cardiac time intervals (PR and QT interval length) was measured. Time-domain measures of fetal heart rate variability (RMSSD and SDNN) were analysed in the context of fetal behavioural state 1F (quiet sleep) or 2F (active sleep). Assays to measure the concentration of NT-proBNP and individual bile acids, and the hydrophobicity index were performed on the umbilical venous serum. Results: MEB was found to be a more effective electron mediator than MDB. The current generated from the reduction of NAD+ to NADH did not have a linear relationship with TCA concentration. Measurements had a relative standard deviation between 20.4 and 47.9%; sensitivity of the biosensor was found to be 1.053 µA per µmol/L of TCA. Oxytocin administration resulted in a significant increase in the amplitude of [Ca2+]i transients when compared to the control (p=0.003). Co-administration with TCA also resulted in an increase in amplitude compared to the control (p < 0.0005). Subsequent concomitant UDCA administration resulted in a significant reduction of amplitude (p=0.002). Co-administration of UDCA with PGE2 resulted in a decrease in the PGE2-induced increase in the time to decay of [Ca2+]i transients (p = 0.043). A positive correlation was found between fetal NT-proBNP concentrations and maternal TSBA concentrations (p = 0.026) fetal TSBA concentrations (p = 0.019), their hydrophobicity index (p = 0.039), glycocholic acid (p = 0.007) and taurocholic acid (p = 0.039). No significant correlations between bile acid concentrations and NT-proBNP were observed in the UDCA-treated cohort. Fetal PR interval length positively correlated with maternal TSBA in untreated (p = 0.027) and UDCA-treated ICP (p = 0.026). Fetal RMSSD values in active sleep (p = 0.028) and SDNN values in quiet sleep (p = 0.013) and active sleep (p = 0.003) were significantly higher in untreated ICP cases than controls. UDCA treatment was associated with a significant reduction in RMSSD values in active sleep (p = 0.030). Discussion: Substantial preliminary work has been done towards development of a bile acid biosensor, however further advancement is required with the aim of increasing sensitivity at the pertinent TCA concentration of 100 µmol/L. MEB has been established as a suitable electron mediator. TCA acts via OT to increase the amplitude of [Ca2+]i transients in human myometrial cells, which supports previous data suggesting that ICP increases sensitivity to OT via upregulation of the OT receptor. UDCA has shown some protective attributes against contractile agonists in this model, suggesting further investigations are required to assess whether it is beneficial for ICP-associated PTL. Untreated ICP is associated with an abnormal fetal cardiac phenotype which is closely associated with maternal and fetal TSBA concentrations and fetal bile acid hydrophobicity; this data supports the hypothesis that ICP-associated IUD is due to a fetal arrythmia. This phenotype was partially ameliorated in UDCA-treated cases, implying that UDCA treatment may be cardioprotective in a certain subset of ICP cases. However, further investigations are required to elucidate the clinical impact of this work.