Subjects

In this nested cross-sectional analysis, subjects between the ages of 18 and 40 years who had singleton live births were included after written consent was obtained. Placental samples were collected from women at term (n = 43; term labor 14 and term not in labor 29) with PTB and intact fetal membranes (n = 27). Cases (PTB) were defined as the presence of regular uterine contractions (2 contractions/10 minutes) with documented cervical changes followed by delivery at < 360 / 7 weeks’ gestation. Gestational age was determined by last menstrual period and was corroborated by ultrasound dating. Subjects with multiple gestations, preeclampsia, preterm premature rupture of the membranes, placental previa, fetal anomalies, cerclage, gestational diabetes mellitus, pregestational diabetes mellitus, poly-and oligohydramnios, intrauterine growth restriction, and other complications such as surgeries during pregnancy were excluded. Control subjects consisted of women who had delivered at term (≥370 / 7 weeks’ gestation) with no medical or obstetric complications during pregnancy. We used noncontiguous gestational ages to define cases and control subjects to minimize overlap of phenotypes. Only pregnancies with gestational ages > 22 weeks’ gestation were included in the study. Demographic data were collected from patient interviews, and clinical data were abstracted from the medical records of the patients. Data that were collected included age, socioeconomic status (education, yearly income, insurance status, and marital status), behavioral status (smoking, alcohol, and drug use during pregnancy), and a complete medical and obstetric history. Histologic examination of the placenta and umbilical cord was performed in all cases. Histologic evidence of inflammation was defined as a dense polymorphonuclear leukocyte/neutrophil infiltration of the amniochorionic membrane excluding decidua. Clinical chorioamnionitis was defined as any 3 of the following events: increased white blood cell count > 15,000/dL, increased C-reactive protein levels (≥0.8 units), fever ≥102.0°F, abdominal pain and/or uterine tenderness, and foul-smelling vaginal discharge. Details of this cross-sectional study and samples can be found in our other publications.

Clinical sample collection

Fetal membranes were collected after placental delivery and washed in phosphate-buffered saline solution (PBS). Fetal membranes were dissected from the mid-zone portion (away from the placental bed and those overlying the cervix) and cleansed of blood clots and decidua with the use of cotton gauze and PBS and stored at –80°C.

Selection of ASD candidate genes

A thorough literature search was performed in PubMed with the following search key words: “Autism Spectrum Disorder,” “gene,” “epigenetics,” and “methylation.” A total of 494 articles were obtained and evaluated carefully by the primary author for genes for which methylation patterns have been associated in ASD. This search resulted in 14 candidate genes: OXTR , SHANK3 , BCL2 , RORA , Engrailed-2 ( EN2 ), Reelin, Methyl-CpG-binding protein 2, Autism susceptibility candidate 2, Neuroligin, Neurexin 1, Solute carrier family 6, member 4, Ubiquitin protein ligase E3A, GABA receptor subunits, and AF4/FMR2 Family, Member 2. The association between PTB and DNA methylation status of these 14 genes in fetal leukocyte DNA was determined based on our previous methylation study. Four of the 14 genes had CpG sites whose methylation levels were associated with PTB: OXTR , SHANK3 , BCL2 and RORA . Thus, we focused on these 4 candidate genes.

Genomic DNA extraction

Total genomic DNA was extracted from fetal membranes (amniochorion) with the DNeasy Blood & Tissue Kit (catalog no: 69504 and 69506; Qiagen, Valencia, CA) according to manufacturer’s specifications.

DNA methylation assays

To investigate the role of epigenetic modifications on our selected ASD candidate genes, we explored the methylation status of the CpG islands within the promoter of the 4 selected ASD candidate genes in fetal membranes with the use of the predesigned methylation-sensitive, restriction-enzyme-based EpiTect Methyl II PCR assays (SABioscieneces, Frederick, MD). Briefly, input genomic DNA was aliquoted into 4 equal portions and subjected to mock (no enzyme), methylation-sensitive, methylation-dependent, and double (methylation-sensitive and methylation-dependent) restriction endonuclease digestion. After digestion, the enzyme reaction was mixed directly with real-time polymerase chain reaction (PCR) master mix and was dispensed into a PCR Array plate that contained prealiquoted primer mixes. The remaining DNA after digestion was quantified by real-time PCR with the use of primers that flank the region of interest. The OXTR assay covers a 172–base pair (bp) region centered around chr3:8811193; the SHANK3 assay covers a 247-bp region centered around chr22:51112164; the BCL2 assay covers a 150-bp region centered around chr18:60986747, and the RORA assay covers a 251-bp region centered around chr15:60884703. The PCR cycling protocol for each assay was 1 cycle (95°C for 10 minutes). 3 cycles (99°C for 30 seconds; 72°C for 1 minute) and 40 cycles (97°C for 15 seconds; 72°C for 1 minute). With the use of the SABiosciences Data Analysis Template, methylation percentages were calculated for each sample based on the ratio of methylated-to-unmethylated products determined by the delta Ct values.

RNA isolation, complementary DNA preparation, and reverse transcription PCR

RNA was extracted from tissues with the use of the Direct-zol RNA Mini Prep (Zymo-Research, Irvine, CA), according to the manufacturer’s instructions. Genomic DNA contamination was removed with DNAse I treatment during RNA extraction. The quality and concentration of extracted total RNA were measured with the use of Biotek Synergy H4 Hybrid Reader (Biotek Instruments Inc, Winooski, VT). RNA samples (0.1 mg/mL) were reverse transcribed with the High-Capacity cDNA Archive Kit (Applied Biosystems, Carlsbad, CA), in accordance with the manufacturer’s instruction.

SYBR green real-time PCR was performed with an ABI 7500 Fast Real-Time PCR System (Applied Biosystems) for the following assays: OXTR (Hs001687m1), SHANK3 (Hs01586468), BCL2 (Hs00236808) and RORA (Hs00536545). We tested primer specificity by reverse transcription PCR (RT-PCR) and confirmed it by melting (dissociation) curve analysis. Glyceraldehyde-3-phosphate dehydrogenase, a housekeeping gene, was used as an internal control. Amplification was performed under the following conditions: denaturation for 30 seconds at 95°C, followed by 40 cycles of denaturation for 15 seconds at 95°C, and annealing/extension for 1 minute at 60°C. All reactions were performed in duplicate, and no template controls were included in each run. The comparative 2 ^-ΔΔ Ct method was used to calculate relative quantification of gene expression, and data are presented as fold changes among groups.

Hydroxymethylation assays

Cytosine hydroxymethylation (5-hmc) provides additional complexity to gene regulation. Hydroxymethylation replaces hydrogen at the C ⁵ -position with a hydroxyl-methyl group in cytosine. To investigate the role of this epigenetic modification, we used the Quest 5-hmC Detection Kit (Zymo Research Corp. Irvine, CA). Equal amounts of DNA from samples were divided into 2 separate reactions. One group of DNA samples were modified with glycosylation. The second group of DNA sample did not receive treatment. An internal control sample was included for reference. The reaction mixtures were incubated at 37°C for >2 hours. Both DNA sample groups were digested with 30 units MspI, a glucosyl-5hmC sensitive restriction endonuclease that cleaves DNA when cytosine, 5-methylcytosine, or 5-hydroxymethylcytosine is located at the inner C position with its recognition site (ie, C C GG). Glucosylation of the base inhibits cleavage by the enzyme. The samples were incubated at 37°C for >4 hours. A low elution spin column was used for DNA clean-up before locus-specific quantitative reverse transcription polymerase chain reaction. This was performed only for OXTR with the following primers: 5′ TGC ACT ATC GCA CGG GTC 3′ (forward) and 5′ GTT GGT CCC AGA GCC TTA ACA 3′ (reverse). Test DNA that has not been processed was used as an internal control. Amplification was performed under the following conditions as suggested by manufacturer’s protocol: denaturation for 1 minute at 95°C followed by 40 cycles of denaturation for 30 seconds at 95°C, and annealing for 15 seconds at 50°C followed by elongation for 15 seconds at 72°C. Followed by 40 cycles of denaturation at 95°C. Final elongation was for 5 minutes at 72°C. Percentage of hydroxymethylation was obtained by the following formula: % = [(–control) – (+5hmC)/(–control) – (no treatment)] × 100, where “–control” is unglucosylated but MspI digested, “+5hmC” is glucosylated and MspI digested, and no treatment is glycosylated or MspI digested.

Western blot analysis

OXTR protein levels were assessed by Western blot. Fetal membranes were homogenized in RIPA buffer with protease inhibitors using the bullet blender (Next Advance, Averill Park, NY). Protein quantification was done with the Pierce BCA kit (Thermo Scientific, Rockford, IL). Samples that contained 45 μg of protein were separated by sodium dodecylsulfate-gel electrophoresis (Bio-Rad, Hercules, CA) according to the manufacturer’s suggestions; proteins were transferred to polyvinyl difluoride membranes using the iBlot dry blotting system (Life Technologies, Grand Island, NY). The membranes were blocked for 2 hours in 5% fat-free milk in TBS-Tween-20 and incubated with primary antibody to OXTR (catalog no. ABN297; EMD Millipore, Billerica, MD) overnight at dilution of 1:400 respectively. Blots were washed and incubated with secondary antibody for 1 hour. After the final washings, the blots were incubated with Pierce ECL2 chemiluminescence detection reagent (1:100 dilution; catalog np. 8019; Thermo Scientific) followed by autoradiography. Detected bands were analyzed densitometrically with the Image J software (National Institutes of Health, rsbweb.nih.gov/ij), and results were normalized to beta-actin (Sigma A5441; Sigma-Aldrich, St. Louis, MO).

Immunohistochemistry

Sample preparation and immunohistochemistry were performed at Histopathology Special Procedures Laboratory of the University of Texas Medical Branch at Galveston Pathology Department, following a standard protocol. Briefly, 3-5 μm-thick sections from formalin-fixed, paraffin-embedded tissue blocks were cut and mounted on positively charged slides. Sections were deparaffinized, and endogenous peroxides were quenched by soaking sections in 2 changes of aqueous hydrogen peroxide for 5 minutes each. After deparaffinization, slides were exposed to 10 mmol/L citrate at pH of 6.0 and heated for 20 minutes and then cooled down for 20 minutes for antigen exposure. Tris-buffered saline solution was used to rinse the sections between each of the immunohistochemistry steps. The primary antibody (catalog no. ABN297, EMD; Millipore, Billerica, MD) was diluted in the antibody dilutent (1:200 dilution) and applied for 1 hour. After being washed in PBS, the sections were incubated in labeled streptavidin-biotin kit (LSAB2), universal secondary antibody (Dako, Glostrup, Denmark) for 15 minutes, followed by LSAB2 label (Dako) for 15 minutes. Color was developed with the use of 3, 30-diaminobenzidine tetrahydrochloride (DAB; Dako) was applied for 5 minutes. The slides were counterstained with Harris hematoxylin (Fisher Scientific, Waltham, MA) and cover-slipped with synthetic glass and permount mounting media. Sections were examined and photographed with an Olympus light microscope (BX43 with a URFL-T digital camera; Olympus, Center Valley, PA) and the software QCapture Pro IMicropublisher (version 6.0; QImaging, Burnaby, BC, Canada) were used for analysis and reproduction of immunohistochemical images. We did not quantitate histochemical staining because the primary purpose of this work was to localize site of production in fetal membranes.

Statistical analysis

For continuous and categoric variables, respectively, Student t tests and X ² tests were performed to determine the statistical significance of any differences in the distribution of baseline characteristics between PTB cases and control subjects. The association between OXTR methylation and case control status was evaluated with a linear regression model with adjustment for race and fetal sex. Analysis of variance (ANOVA) was used to compare messenger RNA and protein expression data among PTB, term labor (TL), and term not in labor (TNIL) groups, followed by Tukey-Kramer pairwise comparisons between groups. A probability value of < .05 was considered significant.