Increased adiposity, inflammation, metabolic disruption and dyslipidemia in adult male offspring of DOSS treated C57BL/6 dams

Evidence indicates that obesity can be promoted by chemical ‘obesogens’ that drive adiposity, hunger, inflammation and suppress metabolism. Dioctyl sodium sulfosuccinate (DOSS), a lipid emulsifier and candidate obesogen in vitro, is widely used in processed foods, cosmetics and as stool softener medicines commonly used during pregnancy. In vivo testing of DOSS was performed in a developmental origins of adult obesity model. Pregnant mice were orally administered vehicle control or DOSS at times and doses comparable to stool softener use during human pregnancy. All weaned offspring consumed only standard diet. Adult male but not female offspring of DOSS-treated dams showed significantly increased body mass, overall and visceral fat masses, and decreased bone area. They exhibited significant decreases in plasma adiponectin and increases in leptin, glucose intolerance and hyperinsulinemia. Inflammatory IL-6 was elevated, as was adipose Cox2 and Nox4 gene expressions, which may be associated with promoter DNA methylation changes. Multiple significant phospholipid/sterol lipid increases paralleled profiles from long-term high-fat diet induced obesity in males. Collectively, developmental DOSS exposure leads to increased adult adiposity, inflammation, metabolic disorder and dyslipidemia in offspring fed a standard diet, suggesting that pharmaceutical and other sources of DOSS taken during human pregnancy might contribute to long-term obesity-related health concerns in offspring.

CpG2 percent methylation correlated with circulating IL-6 levels, C) Cox2 promoter 22 methylation and D) Cox2 CpG1 percent methylation (n = 12/group). Two-Way Anova 23 was used to assess significant changes in DNA methylation based on treatment across the 24 whole promoter region with Sidak's post-hoc test for individual CpG sites differences 25 (*p<0.05 treated vs. control, Two-Way Anova). Graph bars represent means and standard 26 deviations. Individual sites were then also compared using Sidak's post-hoc test 27 (#p<0.05). Linear regression was used to determine correlations between CpG2 percent 28 methylation and circulating IL-6 levels. 29

Supp Figure 4. Effects of DOSS treatment on promoter methylation and 30
relationships to gene expression. 31 Targeted bisulfite sequencing was used to assess promoter methylation in IWAT tissue 32 for genes associated with increased gene expression or circulating protein levels upon 33 DOSS treatment (n = 12/group). Results were obtained for: A) IL-6 CpG2 percent 34 methylation correlation with IL-6 IWAT gene expression, B) Cox2 CpG1 percent 35 methylation correlated with Cox2 IWAT expression, C) adiponectin region 1 promoter 36 methylation, D) adiponectin region 2 promoter methylation, and E) adiponectin region 2 37 CpG3 percent methylation correlated with AdipoQ IWAT gene expression. Two-Way 38 Anova was used to assess significant changes in DNA methylation based on treatment 39 across the whole promoter region with Sidak's post-hoc test for individual CpG sites 40 differences (*p<0.05 DOSS treated vs. control, Two-Way Anova). Graph bars represent 41 means and standard deviations. Individual sites were then also compared using Sidak's 42 post-hoc test (#p<0.05). Linear regression was used to determine correlations between 43 percent methylation and gene expression.

A B
Supp. Figure 2. Supp. Figure 3.  The column was washed 1x with 500uL Buffer AW1, then 1X with 500uL Buffer AW2. Care was taken to ensure the column was completely free of ethanol before eluting DNA with 75uL Buffer AE. DNA concentrations were quantified using Nanodrop ND 1000 (ThermoFisher Scientific).

Bisulfite Conversion and PCR Amplification
Genomic DNA (500ng) from IWAT of 16 week-old mice was bisulfite converted for methylation analysis using the Methylamp DNA Modification Kit (Epigentek) following the manufacturer's protocol. Eight control samples and 12 treated samples were used and selected based on their DNA quality. Target genes for PCR were selected based on previous studies that identified methylation or expression changes in these genes as a result of obesity, glucose intolerance and diabetes, or obesogen exposure (Supp Table 2). Studies from both human and mice were utilized. For targets where primers were listed in publications, these primers were used with conditions optimized in-house. For new targets, primers were designed using MethPrimer in promoter regions of genes flanking CpG islands. For PCR amplification, 12.5 ng of converted DNA was used with EpiMark HotStart Taq DNA polymerase following the manufacturer's recommendations (Supp Table 2). Amplification of single products was validated using gel electrophoresis. PCR products were purified using GenCatch PCR Clean Up Kit (Epoch) and quantified using Nanodrop equipment (Thermo Fisher Scientific). All products from a single individual were then pooled using equal nanomoles with volumes adjusted based on nM concentration (based on lowest concentration) for sequencing.

Illumina Library Prep, Next Generation Sequencing and Analysis
The MUSC Cancer Genomics core was used for next generation sequencing of products.
The sequencing library was prepared using the TruSeq Kit, following the manufacturer's protocol beginning at "End Repair." Adapters were ligated so each individual was barcoded.
Samples were then paired end sequenced on an Illumina MiSeq. All data analysis was performed in BaseSpace. Raw sequences were trimmed using Trimmomatic 1 . Methylation analysis was performed in Methyl Seq v1.0 (Illumina, Inc.). Methyl Seq aligns generated sequences to a bisulfite converted genome using Bowtie 2 2 . Sequences were aligned to Mouse mm9 using a targeted manifest specifically for targeted amplicons based on chromosomal locations for start and end sites. Since this was a targeted analysis, the differences between mm9 and mm10 likely would not impact differences in sequence alignment. CpG methylation status was obtained for each CpG in an amplicon using BisMark 3 . If a sample did not have at least 12 reads per CpG site it was dropped from analysis. No correlation was observed between read count and methylation status (data not shown).
Supplemental Tables: Table 1. Genes and corresponding primers used for qRT -PCR analysis