Following regional ethics approval, written informed consent was obtained from all participants. The pre-eclamptic group included 117 women with singleton pregnancies, blood pressure of ≥140/90 mmHg (twice on separate occasions >6 hours apart) and proteinuria (>300 mg/l protein excreted; +/+ dipstick urinalysis, 24 hours apart; International Society for the Study of Hypertension in Pregnancy definition). The control group (n = 146) had uncomplicated singleton term pregnancies (>37 weeks). Both groups were exclusively primigravid white Europeans and fitted the highly stringent and extensive inclusion/exclusion criteria defined by GOPEC (for details, please refer to www.gopec.org). Exclusion criteria for both groups included essential hypertension, diabetes and autoimmune disease. The control group’s median maternal age was 30 years (range = 16–40 years), gestational age 39.5 weeks (range = 37–42 weeks) and infant weight 3440 g (range = 2430–5370 g). The pre-eclamptic population’s median maternal age was 29 years (range = 16–42 years), gestational age 35.6 weeks (range = 25–42 weeks) and infant weight 2230 g (range = 430–5520 g). There were no significant differences in any demographic parameters other than gestational age and infant birthweight, which were significantly lower in the pre-eclamptic group (P < 0.05).
Power calculations (α= 0.05, power 80%), based on variant allelic distributions from related women of identical ethnicity from control and preterm labour populations, were IL-4 SNP 0.12 and 0.27 (135 controls and 108 pre-eclamptics, respectively), and MMP-9 SNP 0.10 and 0.24 (139 controls and 112 pre-eclamptics, respectively). For the TLR-2 SNP, the investigation was largely exploratory; the very low incidence of the variant allele meant that this study was not adequately powered for this SNP. Blood was collected from the antecubital vein and DNA extracted using QIAmp DNA mini kits (Qiagen, Crawley, West Sussex, UK). This analysis was retrospective and used stored DNA samples. Genotyping involved polymerase chain reaction–restriction fragment length polymorphism approaches using the following primers (Invitrogen, Paisley, UK): TLR-2 +2258 (G>A) (forward: 5′-TATGGTCCAGGAGCTGGAGA-3′ and reverse: 5′-TGACATAAAGATCCCAACTAGACAA-3′), MMP-9 −1562 (C>T) (forward: 5′-GCCTGGCACATAGTAGGCCC-3′ and reverse: 5′-CTTCCTAGCCAGCCGGCATC-3′) and IL-4 −590 (C>T) (forward: 5′-TAAACTTGGGAGAACATGGT-3′ and reverse: 5′-TGGGGAAAGATAGAGTAATA-3′). IL-4 amplifications introduced a point mutation using a mismatch primer designed to create a restriction sequence comprising the polymorphic site. IL-4 −590 (C>T) cycling conditions were as follows: 1 minute at 94°C denaturation, 40 cycles of 94°C, 48°C and 72°C for 45 seconds each, and a 5-minute final extension at 72°C. TLR-2 +2258 (G>A) cycling conditions were as follows: 1 minute at 94°C denaturation, 35 cycles of 94°C for 30 seconds each, 58°C for 1 minute and 72°C for 1 minute, and a 5-minute final extension at 72°C. Cycling conditions for MMP-9 −1562 (C>T) were as follows: denaturation at 94°C for 10 minutes, 35 cycles of 94°C, 59°C and 72°C for 30 seconds each, and a 10-minute final extension at 72°C. Overnight (16–18 hours) restriction digests were carried out at 37°C using PstI [TLR-2 +2258 (G>A)], SphI [MMP-9 −1562 (C>T)] and AvaII [IL-4 −590 (C>T)]. Genotype was assessed following 2% agarose (Bioline, London, UK) gel electrophoresis and visualised on a transilluminator (GelDocR; Bio-Rad Laboratories, Hemel Hempstead, UK). Genotyping procedures were confirmed independently by sequencing of randomly selected samples at the time of protocol validation prior to starting the study. Unclear genotype calls were all resubmitted to genotyping. Allelic distributions were calculated and data compared by chi-square analysis (using Yates’s correction) (SPSS, version 14.1; SPSS Inc., Woking, UK); contingency tables were used to determine odds ratios and 95% confidence intervals. All genotype distributions were assessed for Hardy–Weinberg equilibrium.