Data released on September 01, 2014
For this study, six term pigs were delivered close to full term (115–116 days gestation, by cesarean section) from two pregnant sows (Large White × Duroc, Askelygaard, Denmark), serving as the 0d-term group. Within 6 hours of birth, these pigs were euthanized (sodium pentobarbitone, 200 mg/kg, intravenous), and mid small intestine was dissected out, snap frozen in liquid nitrogen and stored at -80°C.
Under anesthesia, six preterm pigs were obtained from five pregnant sows (Large White × Duroc; Askelygaard, Denmark) by cesarean delivery at 90- 92% (104–107 days) gestation. Six of these pigs were euthanized within 6 hours of birth without any feeding (0d-preterm group). The intestine was also dissected out and frozen at -80°C. The remaining 6 preterm pigs were reared individually in infant incubators (Air-Shields, Hatboro, PN) and fitted with orogastric catheters for PN and oro-gastric feeding tube. The pigs were given PN solution at a rate of 4 mL/(kg × h) on d1 and 6 mL/(kg × h) on d2 as well as minimal enteral nutrition: 24 mL/(kg × d) on d1 and 40 mL/(kg × d) on d2 with infant formula. The PN solution was based on a commercially available product (Kabiven, Fresenius Kabi) and adjusted in nutrient composition to meet the requirement of pigs. On d3, they were transferred to full enteral nutrition (EN) [120 mL/(kg × d)] for two more days before sacrifice. The mid intestine was also dissected out and frozen at -80°C.
Pigs were continuously monitored for clinical symptoms of Necrotizing enterocolitis (NEC) and at tissue collection the severity of NEC-like intestinal lesions was evaluated using scoring system. Six pigs remained healthy at the time of tissue collection on day 4 (4d-preterm) while six other pigs had significant NEC lesions identified in their intestines. Experimental and animal management procedures followed a high standard of veterinary care and were conducted in accordance with approval of institutional and national policies and guidelines.
For sanger sequencing, the processed sequence was aligned to the pig reference regions (all PCR regions) by the blast software. For HiSeq sequencing, the Illumina reads were post-processed and aligned to the pig reference regions (all PCR regions) using SOAP aligner (Version 2.01) with default parameters that excluded reads with more than five mismatched bases. Multiple reads mapping to the same position were counted only once to remove potential bias from PCR. Then we calculated the methylated level based on the effective data.