Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Maternal Obesity Affects Fetal Neurodevelopmental and Metabolic Gene Expression: A Pilot Study

  • Andrea G. Edlow ,

    Affiliation Mother Infant Research Institute, Tufts Medical Center, Boston, Massachusetts, United States of America

  • Neeta L. Vora,

    Address Current address: Department of Obstetrics and Gynecology, Division of Maternal-Fetal Medicine, University of North Carolina-Chapel Hill, NC, United States of America

    Affiliation Mother Infant Research Institute, Tufts Medical Center, Boston, Massachusetts, United States of America

  • Lisa Hui,

    Address Current address: Department of Perinatal Medicine, Mercy Hospital for Women, Heidelberg, VIC, Australia

    Affiliation Mother Infant Research Institute, Tufts Medical Center, Boston, Massachusetts, United States of America

  • Heather C. Wick,

    Affiliation Department of Computer Science, Tufts University, Medford, Massachusetts, United States of America

  • Janet M. Cowan,

    Affiliation Department of Pathology and Laboratory Medicine, Tufts Medical Center, Boston, Massachusetts, United States of America

  • Diana W. Bianchi

    Affiliation Mother Infant Research Institute, Tufts Medical Center, Boston, Massachusetts, United States of America

Maternal Obesity Affects Fetal Neurodevelopmental and Metabolic Gene Expression: A Pilot Study

  • Andrea G. Edlow, 
  • Neeta L. Vora, 
  • Lisa Hui, 
  • Heather C. Wick, 
  • Janet M. Cowan, 
  • Diana W. Bianchi



One in three pregnant women in the United States is obese. Their offspring are at increased risk for neurodevelopmental and metabolic morbidity. Underlying molecular mechanisms are poorly understood. We performed a global gene expression analysis of mid-trimester amniotic fluid cell-free fetal RNA in obese versus lean pregnant women.


This prospective pilot study included eight obese (BMI≥30) and eight lean (BMI<25) women undergoing clinically indicated mid-trimester genetic amniocentesis. Subjects were matched for gestational age and fetal sex. Fetuses with abnormal karyotype or structural anomalies were excluded. Cell-free fetal RNA was extracted from amniotic fluid and hybridized to whole genome expression arrays. Genes significantly differentially regulated in 8/8 obese-lean pairs were identified using paired t-tests with the Benjamini-Hochberg correction (false discovery rate of <0.05). Biological interpretation was performed with Ingenuity Pathway Analysis and the BioGPS gene expression atlas.


In fetuses of obese pregnant women, 205 genes were significantly differentially regulated. Apolipoprotein D, a gene highly expressed in the central nervous system and integral to lipid regulation, was the most up-regulated gene (9-fold). Apoptotic cell death was significantly down-regulated, particularly within nervous system pathways involving the cerebral cortex. Activation of the transcriptional regulators estrogen receptor, FOS, and STAT3 was predicted in fetuses of obese women, suggesting a pro-estrogenic, pro-inflammatory milieu.


Maternal obesity affects fetal neurodevelopmental and metabolic gene expression as early as the second trimester. These findings may have implications for postnatal neurodevelopmental and metabolic abnormalities described in the offspring of obese women.


Maternal obesity is a major public health problem in the United States. Sixty percent of reproductive age women are overweight at conception and one third are obese [1], [2]. There has been a parallel rise in childhood obesity and metabolic syndrome. This has coincided with an increased interest in the impact of the intrauterine environment on fetal gene expression and development [3]. Offspring of obese parents are significantly more likely to be obese and to have metabolic syndrome [4], [5]. Importantly, maternal body mass index (BMI) is more significantly associated with offspring obesity than paternal BMI, suggesting that the in utero environment plays an important role [6].

Maternal obesity also appears to have intergenerational health consequences beyond childhood metabolic syndrome and obesity. Data from large epidemiologic studies suggest an association with adverse neurodevelopmental outcomes in offspring, including lower general cognitive capabilities [7], [8], [9], and an increased incidence of autism spectrum disorders [10], developmental delay [11], and attention deficit hyperactivity disorder [12]. The molecular mechanisms by which maternal obesity might result in an increased risk for childhood obesity, metabolic syndrome, and adverse neurodevelopmental outcomes are currently unknown.

Amniotic fluid supernatant (AFS) offers unique advantages in studying real-time human fetal physiology and development. The analysis of cell-free fetal RNA (cffRNA) in AFS makes use of a readily available, typically discarded human biofluid. Prior work by our laboratory has demonstrated that fetal gene expression patterns in AFS vary according to gender, gestational age, and disease state [13], [14], [15], [16], [17]. Cell-free fetal nucleic acids are present in significantly higher concentrations in amniotic fluid, and arise from a distinct pool, compared to cell-free nucleic acids in maternal serum [18], [19], [20], [21]. While cell-free fetal DNA and RNA in maternal serum are known to arise from the placenta [22], [23], [24], [25], epigenetic studies and gene expression microarrays of cell-free fetal nucleic acids in amniotic fluid demonstrate relatively little contribution from the placenta [20], [25]. Thus, cell-free nucleic acids in amniotic fluid provide real-time information about fetal development.

Characterization of the normal second trimester amniotic fluid core transcriptome has demonstrated cffRNA transcripts in mid-trimester amniotic fluid reflecting the development of multiple organs including the fetal thyroid, liver, lung, pancreas, blood, and brain [14]. In prior work from our group, twenty-three highly organ-specific transcripts were identified, one-third of which mapped to the fetal brain [14]. This unexpected and novel finding has been substantiated in later studies [17], and suggests that amniotic fluid supernatant may be used to obtain neurodevelopmental information from living fetuses. These nucleic acid transcripts may pass into amniotic fluid via transport through fetal membranes in the fontanelle, nose, or ear; via shedding through urine; the trachea; fetal blood, or other mechanisms.

Here, we performed a discovery-driven research study to test the hypothesis that fetuses of obese women have different gene expression patterns than fetuses of lean women. We used cell-free RNA isolated from second trimester amniotic fluid supernatant, gene expression microarrays, and publicly available bioinformatics resources to identify differentially expressed genes, and their functions and interactions. In so doing, we have identified mechanisms that may be associated with an increased risk of neurodevelopmental and metabolic morbidity in offspring of obese pregnant women.

Materials and Methods

Ethics statement

Samples were collected with approval from the Tufts Medical Center Institutional Review Board from June 2011 through April 2012 (IRB protocol # 5582). Subjects signed informed consent for amniocentesis, which was performed for standard clinical indications.

Recruitment and sample collection

This was a prospective pilot study of women with singleton fetuses without structural anomalies undergoing second trimester (14–24 weeks) genetic amniocentesis at Tufts Medical Center. Subjects signed informed consent for amniocentesis, which was performed for standard clinical indications (advanced maternal age, ultrasonographic soft markers of aneuploidy, abnormal serum screening results, or maternal request). Women with a BMI≥30 (obese, n = 14) or <25 (lean, n = 23) at the time of amniocentesis were enrolled. Per protocol, access to the medical record was limited to clinical information available at the time of amniocentesis (i.e. indications for the procedure, presence/absence of fetal anomalies, standard maternal demographic data), and cytogenetic results. Fetuses later found to have an abnormal karyotype, or those with RNA or cDNA of insufficient quality or quantity to hybridize to microarrays, were subsequently excluded. We aimed for a target of at least eight samples per group, based on the demonstration that near-maximal levels of statistical stability are obtained with between eight and 15 biological replicates in microarray studies [26]. Figure 1 depicts the flow of subjects through the study. Samples were matched for gestational age and fetal sex, both of which have been previously shown to influence fetal gene expression [13], [27]. The amniotic fluid samples were centrifuged at 165× g for 10 minutes at room temperature to separate amniocytes for subsequent diagnostic testing. The residual AFS was stored at −80°C until RNA extraction.

Figure 1. Flow of Subjects Through the Study.

Process by which the final subjects were selected for microarray analysis.

RNA extraction, processing, and hybridization to microarrays

Cell-free fetal RNA was extracted from five to ten milliliters of AFS within six months of collection, according to previous protocols developed by our group to maximize RNA yield [28]. RNA was stored at −80°C for less than 6 months, to maximize RNA integrity [29]. Subjects were matched for gestational age, to avoid any confounding effects of gestational age on cell-free RNA quantity [13], [30]. RNA was extracted, purified, converted to cDNA, and hybridized to whole genome expression arrays (Affymetrix GeneChip® Human Genome U133 Plus 2.0, Affymetrix Inc.), using previously described protocols that have been established in our laboratory [31], [32].

Gene expression data analysis

The gene expression array results have been uploaded to the Gene Expression Omnibus repository (GSE48521). The microarray data were normalized using the three-step command from the affyPLM package in Bioconductor [33], using ideal-mismatch background-signal adjustment, quantile normalization, and the Tukey biweight summary method [34]. This summary method includes a logarithmic transformation to improve the normality of the data. The mas5calls function from the Bioconductor affy package was used to obtain detection calls consistent with those produced by the Affymetrix 5.0 software. We used paired t-tests to identify differentially regulated probe sets in the fetuses of obese women compared to lean controls. The Benjamini-Hochberg (BH) correction was applied to the normalized expression data for all probe sets on the microarray, to adjust for multiple comparisons in order to limit the false discovery rate [35]. Throughout the manuscript, the terms “BH p-value” and “false discovery rate” are used interchangeably. A principal component analysis was performed using R (version 2.13.1) to identify dominant sources of variation in the gene expression data [36]. Boxplots were generated in the R software environment (version 2.13.1) to examine the distribution of normalized gene expression data across samples, and in obese versus lean study subjects.

Functional genomic analysis

We considered genes that were up- or down-regulated in eight of eight pairs, and associated with BH-p values <0.05, to be significantly differentially regulated. The Affymetrix gene probe IDs, with corresponding median fold change and BH-p values, were uploaded to Ingenuity® Pathways Analysis (IPA, Ingenuity® Systems, Redwood City, CA version 9.0, content version 12710793). IPA utilizes a manually curated database containing biological interactions and functional annotations to identify differentially represented biological functions and/or diseases in a data set. IPA uses a right-tailed Fisher's exact test to calculate a significance score for each association between genes in the experimental dataset and a biological function. To reduce false positive results, we considered IPA pathways to be significant only if they were associated with a BH-p value <0.05. Within these pathways, we considered functional annotations associated with right-tailed Fisher's exact p-values <0.01 to reflect a statistically significant, non-random association. This cutoff is more stringent than the recommended threshold of <0.05 [37]. False-discovery rates, and when calculated, bias-corrected Z-scores (predicting the impact of gene expression on a particular function), were reported separately for each association. Only those functional annotations associated with three or more genes in the dataset were considered. Enriched pathways for the subcategories “Molecular and Cellular Functions,” and “Physiological System Development and Function” were reported separately.

The Upstream Regulator Analysis feature of IPA was utilized to predict the activation or inhibition of transcriptional regulators based on the direction of gene expression changes in our data set. We defined upstream regulators as significantly activated or inhibited if the bias-corrected Z-score was ≥2.0 or ≤2.0, respectively, in accordance with recommended thresholds [38].

We used a publicly-available gene expression atlas ( to determine whether genes that were significantly dysregulated in fetuses of obese women had tissue-specific expression [39]. We chose the BioGPS atlas due to its coverage of normal adult and fetal tissues, compatibility with the Affymetrix microarray platform, and good correlation between transcript levels and protein abundance [40]. We considered genes to be highly organ-specific if they corresponded to a single organ with an expression value >30 multiples of the median (MoM), consistent with previously-established stringency thresholds [41].


The lean and obese study groups did not differ significantly with respect to maternal or gestational ages, but did vary significantly with respect to maternal BMI (Table 1). Clinical indications for amniocentesis were the same in both groups (advanced maternal age, abnormal serum screening result, and/or the presence of soft markers for aneuploidy on Level II ultrasound examination, Table 2). Mean array hybridization efficiency was similar, at 42.13%±3.54% in obese and 41.85%±4.92% in lean subjects.

The Affymetrix HG U133 Plus 2.0 array contains more than 54,000 probe sets, allowing for comprehensive analysis of whole genome expression. IPA analysis demonstrated differential regulation of 205 genes in fetuses of obese compared to lean women. One hundred and fourteen of these were significantly up-regulated, and 91 were significantly down-regulated (Table S1). These 205 genes comprise approximately one percent of the total number of unique genes interrogated by the Affymetrix whole genome array [42]. Apolipoprotein D (APOD) was the most up-regulated gene in fetuses of obese women (nine-fold). The top ten most up- and down-regulated genes in fetuses of obese women and their functions are listed in Tables 3 and 4.

Table 3. Top ten most up-regulated genes in fetuses of obese versus lean women in the second trimester.

Table 4. Top ten most down-regulated genes in fetuses of obese versus lean women in the second trimester.

Box-and-whisker plots were generated to examine the distribution of normalized gene expression data across samples (Figure S1), and to investigate the expression of genes of particular interest in obese versus lean study subjects (Figure 2). These plots demonstrate that the differences between groups are not driven by a single sample pair or by sample normalization, but indeed reflect differential expression between the obese and lean cohorts as a whole (Figure S1). Figure 2 depicts the distribution of log-normalized expression data for genes of interest, including genes tissue-specific for the central nervous system (APOD, CA11, PCLO), as well as those implicated in apoptotic pathways (BCL2, BCL2L11, BCL3, STK24). These boxplots provide further confirmation of the differences in gene expression in the obese and lean study populations for the CNS-specific genes, as well as for apoptosis-related genes, with clearly different median normalized gene expression values, no overlap in the interquartile ranges, and few outliers.

Figure 2. Central Nervous System-Specific and Apoptosis-Related Gene Expression Values in Obese versus Lean Subjects.

Each box encompasses the interquartile range (IQR) of log-normalized gene expression values for the gene of interest in all obese (shaded box) and all lean (white box) subjects. The dark horizontal lines represent the median gene expression value for the obese or lean subjects. The whiskers represent values within 1.5 times the interquartile range greater than or less than the upper or lower quartile, respectively. The open circles represent values greater than 1.5 times the interquartile range.

Functional analysis of differentially regulated genes in fetuses of obese women

Molecular and cellular functions.

Molecular and cellular functions that were significantly differentially regulated in fetuses of obese women are presented in Table 5A. Of these, cell death was the most significant due to multiple genes involved in apoptosis, including B-cell CLL/lymphoma 2 (BCL2), B-cell CLL/lymphoma 3 (BCL3), BCL2-like 11 (BCL2L11), BCL2-like 1 (BCL2L1), serine/threonine kinase 24 (STK24), and caspase 9 (CASP9). Anti-apoptotic genes were significantly up-regulated in fetuses of obese women (BCL2, BCL3, BCL2L1), while pro-apoptotic genes (STK24, CASP9) were down-regulated.

Table 5. Significantly Differentially Regulated Biological Functions and Systems in Fetuses of Obese versus Lean Women.

Within the IPA category of “Cell Death,” many functional annotations associated with the central nervous system (CNS) were significantly dysregulated in fetuses of obese women. Apoptosis of cerebral cortex cells, sympathetic neurons, cortical neurons, and neuroblastoma cell lines, cell viability of dendritic cells and hippocampal neurons, and cell death of hippocampal cells were all significantly dysregulated. To investigate the association between maternal BMI and CNS apoptosis, a post hoc principal component analysis (PCA) was performed utilizing all the genes annotated by IPA to CNS apoptosis (BCL2, BCL2L1, BCL2L11, CASP9, P4HB, REST, HMGA1, XBP1). The PCA demonstrated that within our dataset, genes implicated in CNS apoptosis segregated primarily by maternal BMI (Figure 3).

Figure 3. Principal Component Analysis of Genes Implicated in Central Nervous System Apoptosis.

Figure demonstrates the results of the principal component analysis. Obese subjects are represented by red spheres and lean subjects are represented by black spheres. The results suggest that gene expression segregates on the basis of maternal BMI. On the x-axis is principal component (PC) 1, maternal body mass index (BMI), which accounts for the greatest proportion of variance in the gene expression data (21%). On the y-axis is PC 2, which accounts for the second greatest proportion of variance (14%).

Gene expression patterns suggestive of dysregulated apoptosis in fetuses of obese women were not limited to the nervous system. The IPA analysis predicted a significant decrease in gastrointestinal tract apoptosis, and significant dysregulation of liver cell and pancreatic beta islet cell apoptosis (Table S2). Cell death of B and T lymphocytes, embryonic stem cells, and hematopoietic progenitor cells was also significantly dysregulated (Table S3).

Physiological systems.

Significantly differentially regulated physiological systems in fetuses of obese women are presented in Table 5B. Within these categories, functional annotations related to apoptosis, gonadogenesis and gametogenesis, and survival/cell viability of B and T lymphocytes were significantly affected. Detailed results are provided in Table S3.

Upstream regulators.

In fetuses of obese women, the IPA upstream regulator analysis predicted the activation of estrogen receptor (ESR1/2) (activation Z-score 2.01), FBJ murine osteosarcoma viral oncogene homolog (FOS) (activation Z-score 2.17), and signal transducer and activator of transcription 3 or STAT3, also called acute-phase response factor (activation Z-score 2.16). These transcriptional regulators are of particular interest because of their involvement in hormonal and inflammatory signaling pathways, leptin regulation, glucose homeostasis, and hepatic steatosis (Table 6).

Table 6. Activated Upstream Regulators in Fetuses of Obese versus Lean Women.

Tissue origin of differentially regulated genes

Of the 205 differentially regulated genes in fetuses of obese women, 12 were tissue-specific. Three of these 12 genes were highly expressed in CNS tissues, including APOD, CA11, and PCLO. Of the remaining nine genes, three mapped to the placenta, three mapped to lymphocyte populations (T- and B-cells, NK cells) or myeloid, monocyte, and dendritic cells, and one mapped to the liver. A list of the tissue-specific genes and their corresponding functions are presented in Table 7.


In this study we demonstrated that significant differences in fetal gene expression in obese pregnant women are detectable as early as the second trimester. The two major and unexpected findings in fetuses of obese pregnant women, overexpression of APOD and gene expression patterns consistent with decreased brain apoptosis in the mid-trimester, suggest two potential mechanisms by which maternal obesity may lead to adverse neurodevelopmental outcomes in offspring. While prior work has demonstrated that mid-trimester amniotic fluid is enriched for brain-specific transcripts [14], [17], neither of these studies has described the types of neurological gene abnormalities demonstrated here in fetuses of obese pregnant women.

APOD is highly and specifically expressed in the central nervous system (CNS), and is synthesized and secreted by oligodendrocytes and astrocytes [43]. The tissue-specific mapping of APOD to the CNS is strongly conserved across multiple species, including human, rat and mouse [44], [45]. ApoD, the protein product, has been demonstrated to exert neuroprotective and neurotrophic effects in cell culture and in a rodent model of excitotoxic brain injury [46], [47]. However, ApoD is thought to play a role in the pathophysiology of schizophrenia [48]. Increased expression of APOD in the human prefrontal cortex during critical developmental periods is associated with increased susceptibility to schizophrenia [49]. Elevated levels of ApoD have also been noted in human plasma during a first psychotic episode [50], as well as in patients with bipolar disorder, Parkinson's and Alzheimer's diseases [51], [52], [53]. Cortical ApoD expression has been demonstrated to increase six- to eight-fold between the neonatal period and adulthood, with increased expression correlating with genetic and biochemical markers of oxidative stress [54]. In a rodent model, maternal high fat diet was associated with increased oxidative stress and inflammatory signaling in the brains of offspring [55], Thus, the presence of excess ApoD during brain development could itself be deleterious, or could represent a response to a harmful or neurodegenerative process.

Our results also suggest decreased apoptosis of cerebral cortex cells, and dysregulation of cell death and cell survival in multiple areas of the fetal brain. The differential gene expression observed in fetuses of obese women, with significant upregulation of BCL2 and BCL2L11 and downregulation of CASP9 suggests decreased signaling through the internal or mitochondrial apoptotic pathways [56]. Animal models have similarly demonstrated differential brain gene expression and differences in brain apoptosis in offspring of obese females [57], [58]. Apoptosis plays a critical role in normal neurodevelopment [59], [60], so decreased brain apoptosis may be a potential mechanism by which maternal obesity adversely influences offspring neurodevelopment.

The IPA upstream regulator analysis predicted significant activation of the estrogen receptor (ESR1/2), FOS, and STAT3. Prior experience with diethylstilbestrol, a synthetic estrogen, and recent data regarding in utero exposure to estrogenic compounds such as bisphenol A, suggest that developmental exposure to excess estrogen signaling may result in obesity, earlier sexual maturation in girls, and increased risk for breast and other reproductive tract cancers [61], [62]. FOS, also called cFOS, is a proto-oncogene involved in inflammatory signaling, appetite regulation, and regulation of cell proliferation and apoptotic cell death [63], [64]. FOS has been implicated in the pathogenesis of atherosclerosis and heart disease [65], [66]. STAT3, also called acute-phase response factor, has been implicated in inflammatory signaling, inhibition of apoptosis, glucose homeostasis, hyperleptinemia, and hepatic steatosis [67], [68], [69], [70].

The gene APOD was nine-fold up-regulated in fetuses of obese women. APOD encodes the protein product apolipoprotein D, a member of the lipocalin family of transporter proteins and a component of high-density lipoprotein [71]. This gene is of specific interest, given its known involvement in lipid homeostasis [71], insulin resistance [72], obesity [73], and hypothalamic regulation of food intake [74]. There is mounting concern that maternal obesity, which has been associated with both intrauterine growth restriction and large-for-gestational age fetuses, may contribute to the epidemic of childhood obesity and metabolic disorders via fetal programming [75], [76]. Up-regulation of APOD, in conjunction with gene expression patterns consistent with increased estrogen and inflammatory signaling, suggest potential molecular mechanisms underlying the increased risk for obesity and metabolic syndrome in offspring of obese women.

The use of human amniotic fluid, rather than an animal model, to examine the effects of maternal obesity on fetal gene expression was a significant strength of this study. Other strengths include that all samples were processed by the same clinical cytogenetics and research laboratories, and that the length of sample storage was standardized, limiting any variability in specimen handling or processing techniques. Case and control samples were matched for variables known to influence gene expression, including fetal sex and gestational age. Rigorous statistical criteria were applied to identify significantly differentially regulated genes in fetuses of obese women.

One limitation of this study was its relatively small sample size. However, given the well-characterized neurodevelopmental and metabolic morbidities of offspring of obese women, and the knowledge gap regarding underlying in utero molecular mechanisms, we believe these pilot data are novel, and will help to focus future studies. In addition, eight matched pairs have been shown to provide near-maximal statistical stability for microarray experiments, and are well above the five subjects per group considered to be an acceptable minimum for microarray experiments [26], [77], [78]. Although maternal age was not significantly different between groups (p = 0.6), there was a three-year difference in mean age between groups (37.1±5 for obese compared to 33.9±5.5 for lean women) that may have achieved significance with larger study numbers. While this is a potential confounder, we are not aware of data suggesting that maternal age alone influences fetal gene expression in the absence of aneuploidy. The fact that all samples came from women having an amniocentesis may itself suggest underlying pregnancy co-morbidities that could theoretically introduce bias. However, amniocentesis is the only way to acquire this biofluid in living fetuses, and we excluded all fetuses with structural anomalies and/or abnormal karyotype, in order to minimize such bias. Several of the amniocenteses were performed for sonographic “soft markers” of aneuploidy, such as choroid plexus cyst and echogenic intracardiac focus. Such soft markers are not structural abnormalities and do not have clinical significance if the fetus has a euploid karyotype. These samples represent a narrow window of gestational age; this is necessarily a reflection of the time during pregnancy in which amniocentesis is routinely performed. Finally, because these amniotic fluid specimens were anonymized after collection per IRB specification, we were only able to analyze the sonographic and standard demographic data available at the time of amniocentesis, the indication for amniocentesis, and the cytogenetic result. Pregnancy outcomes and maternal medical comorbidities of included subjects were therefore unknown. However, the majority of pregnancy-related conditions that could potentially influence fetal gene expression, such as gestational diabetes and preeclampsia, would not yet exist or be diagnosed at the time of second trimester amniocentesis.

In summary, analysis of the amniotic fluid transcriptome in fetuses of obese women demonstrates gene expression patterns suggestive of decreased brain apoptosis; lipid, insulin and appetite dysregulation; and increased estrogen and inflammatory signaling. The molecular mechanisms predisposing offspring of obese women to neurodevelopmental abnormalities and metabolic complications may be initiated as early as the second trimester. While prior work has demonstrated that mid-trimester amniotic fluid is enriched for brain-specific transcripts [17], the ones demonstrated here are novel and may contribute to adverse neurodevelopmental outcomes in fetuses of obese pregnant women. Future experiments are planned to study fetal brain apoptosis, APOD gene expression, and offspring neurocognitive development in a mouse model of maternal diet-induced obesity.

Supporting Information

Figure S1.

Normalized Gene Expression Values for Each Subject. Each box encompasses the interquartile range (IQR) of log-normalized gene expression values for each microarray. Shaded boxes represent obese subjects, white boxes represent lean subjects. The dark horizontal lines represent the median gene expression value for each array. The whiskers represent values within 1.5 times the interquartile range greater than or less than the upper or lower quartile, respectively. The open circles represent values greater than 1.5 times the interquartile range.



Table S1.

Significantly differentially regulated genes in fetuses of obese versus lean women in the second trimester.



Table S2.

Significantly differentially regulated molecular and cellular functions in fetuses of obese versus lean women, with associated functional annotations.



Table S3.

Significantly differentially regulated physiological systems in fetuses of obese versus lean women, with associated functional annotations.



Author Contributions

Conceived and designed the experiments: AGE NLV DWB. Performed the experiments: AGE NLV. Analyzed the data: AGE HCW DWB. Contributed reagents/materials/analysis tools: AGE NLV LH JMC DWB. Wrote the paper: AGE LH DWB. Revised the manuscript critically for important intellectual content: AGE NLV LH HCW JMC DWB.


  1. 1. Flegal KM, Carroll MD, Kit BK, Ogden CL (2012) Prevalence of obesity and trends in the distribution of body mass index among US adults, 1999–2010. JAMA 307: 491–497. doi: 10.1001/jama.2012.39
  2. 2. Kim SY, Dietz PM, England L, Morrow B, Callaghan WM (2007) Trends in pre-pregnancy obesity in nine states, 1993–2003. Obesity (Silver Spring) 15: 986–993. doi: 10.1038/oby.2007.621
  3. 3. Gluckman PD, Hanson MA, Cooper C, Thornburg KL (2008) Effect of in utero and early-life conditions on adult health and disease. N Engl J Med 359: 61–73. doi: 10.1056/nejmra0708473
  4. 4. Boney CM, Verma A, Tucker R, Vohr BR (2005) Metabolic syndrome in childhood: association with birth weight, maternal obesity, and gestational diabetes mellitus. Pediatrics 115: e290–296. doi: 10.1542/peds.2004-1808
  5. 5. Crespo PS, Prieto Perera JA, Lodeiro FA, Azuara LA (2007) Metabolic syndrome in childhood. Public Health Nutr 10: 1121–1125. doi: 10.1017/s1368980007000596
  6. 6. Lawlor DA, Smith GD, O'Callaghan M, Alati R, Mamun AA, et al. (2007) Epidemiologic evidence for the fetal overnutrition hypothesis: findings from the mater-university study of pregnancy and its outcomes. Am J Epidemiol 165: 418–424. doi: 10.1093/aje/kwk030
  7. 7. Tanda R, Salsberry PJ, Reagan PB, Fang MZ (2013) The Impact of Prepregnancy Obesity on Children's Cognitive Test Scores. Matern Child Health J 17: 222–229. doi: 10.1007/s10995-012-0964-4
  8. 8. Heikura U, Taanila A, Hartikainen AL, Olsen P, Linna SL, et al. (2008) Variations in prenatal sociodemographic factors associated with intellectual disability: a study of the 20-year interval between two birth cohorts in northern Finland. Am J Epidemiol 167: 169–177. doi: 10.1093/aje/kwm291
  9. 9. Neggers YH, Goldenberg RL, Ramey SL, Cliver SP (2003) Maternal prepregnancy body mass index and psychomotor development in children. Acta Obstet Gynecol Scand 82: 235–240. doi: 10.1034/j.1600-0412.2003.00090.x
  10. 10. Krakowiak P, Walker CK, Bremer AA, Baker AS, Ozonoff S, et al. (2012) Maternal metabolic conditions and risk for autism and other neurodevelopmental disorders. Pediatrics 129: e1121–1128. doi: 10.1542/peds.2011-2583
  11. 11. Hinkle SN, Schieve LA, Stein AD, Swan DW, Ramakrishnan U, et al. (2012) Associations between maternal prepregnancy body mass index and child neurodevelopment at 2 years of age. Int J Obes (Lond) 36: 1312–1319. doi: 10.1038/ijo.2012.143
  12. 12. Rodriguez A, Miettunen J, Henriksen TB, Olsen J, Obel C, et al. (2008) Maternal adiposity prior to pregnancy is associated with ADHD symptoms in offspring: evidence from three prospective pregnancy cohorts. Int J Obes (Lond) 32: 550–557. doi: 10.1038/sj.ijo.0803741
  13. 13. Larrabee PB, Johnson KL, Lai C, Ordovas J, Cowan JM, et al. (2005) Global gene expression analysis of the living human fetus using cell-free messenger RNA in amniotic fluid. JAMA 293: 836–842. doi: 10.1001/jama.293.7.836
  14. 14. Hui L, Slonim DK, Wick HC, Johnson KL, Bianchi DW (2012) The amniotic fluid transcriptome: a source of novel information about human fetal development. Obstet Gynecol 119: 111–118. doi: 10.1097/aog.0b013e31823d4150
  15. 15. Slonim DK, Koide K, Johnson KL, Tantravahi U, Cowan JM, et al. (2009) Functional genomic analysis of amniotic fluid cell-free mRNA suggests that oxidative stress is significant in Down syndrome fetuses. Proc Natl Acad Sci U S A 106: 9425–9429. doi: 10.1073/pnas.0903909106
  16. 16. Koide K, Slonim DK, Johnson KL, Tantravahi U, Cowan JM, et al. (2011) Transcriptomic analysis of cell-free fetal RNA suggests a specific molecular phenotype in trisomy 18. Hum Genet 129: 295–305. doi: 10.1007/s00439-010-0923-3
  17. 17. Hui L, Slonim DK, Wick HC, Johnson KL, Koide K, et al. (2012) Novel neurodevelopmental information revealed in amniotic fluid supernatant transcripts from fetuses with trisomies 18 and 21. Hum Genet 131: 1751–1759. doi: 10.1007/s00439-012-1195-x
  18. 18. Bianchi DW, LeShane ES, Cowan JM (2001) Large amounts of cell-free fetal DNA are present in amniotic fluid. Clin Chem 47: 1867–1869.
  19. 19. Zhong XY, Holzgreve W, Tercanli S, Wenzel F, Hahn S (2006) Cell-free foetal DNA in maternal plasma does not appear to be derived from the rich pool of cell-free foetal DNA in amniotic fluid. Arch Gynecol Obstet 273: 221–226. doi: 10.1007/s00404-005-0068-0
  20. 20. Lun FM, Chiu RW, Leung TY, Leung TN, Lau TK, et al. (2007) Epigenetic analysis of RASSF1A gene in cell-free DNA in amniotic fluid. Clin Chem 53: 796–798. doi: 10.1373/clinchem.2006.084350
  21. 21. Makrydimas G, Gerovassili A, Sotiriadis A, Kavvadias A, Nicolaides KH (2008) Cell-free fetal DNA in celomic fluid. Ultrasound Obstet Gynecol 32: 594–595. doi: 10.1002/uog.6117
  22. 22. Flori E, Doray B, Gautier E, Kohler M, Ernault P, et al. (2004) Circulating cell-free fetal DNA in maternal serum appears to originate from cyto- and syncytio-trophoblastic cells. Case report. Hum Reprod 19: 723–724. doi: 10.1093/humrep/deh117
  23. 23. Guibert J, Benachi A, Grebille AG, Ernault P, Zorn JR, et al. (2003) Kinetics of SRY gene appearance in maternal serum: detection by real time PCR in early pregnancy after assisted reproductive technique. Hum Reprod 18: 1733–1736. doi: 10.1093/humrep/deg320
  24. 24. Masuzaki H, Miura K, Yoshiura KI, Yoshimura S, Niikawa N, et al. (2004) Detection of cell free placental DNA in maternal plasma: direct evidence from three cases of confined placental mosaicism. J Med Genet 41: 289–292. doi: 10.1136/jmg.2003.015784
  25. 25. Edlow AG, Bianchi DW (2012) Tracking fetal development through molecular analysis of maternal biofluids. Biochim Biophys Acta 1822: 1970–1980. doi: 10.1016/j.bbadis.2012.04.005
  26. 26. Pavlidis P, Li Q, Noble WS (2003) The effect of replication on gene expression microarray experiments. Bioinformatics 19: 1620–1627. doi: 10.1093/bioinformatics/btg227
  27. 27. Massingham LJ, Johnson KL, Bianchi DW, Pei S, Peter I, et al. (2011) Proof of concept study to assess fetal gene expression in amniotic fluid by nanoarray PCR. J Mol Diagn 13: 565–570. doi: 10.1016/j.jmoldx.2011.05.008
  28. 28. Dietz JA, Johnson KL, Massingham LJ, Schaper J, Horlitz M, et al. (2011) Comparison of extraction techniques for amniotic fluid supernatant demonstrates improved yield of cell-free fetal RNA. Prenat Diagn 31: 598–599. doi: 10.1002/pd.2732
  29. 29. Peter I, Tighiouart H, Lapaire O, Johnson KL, Bianchi DW, et al. (2008) Cell-free DNA fragmentation patterns in amniotic fluid identify genetic abnormalities and changes due to storage. Diagn Mol Pathol 17: 185–190. doi: 10.1097/pdm.0b013e31815bcdb6
  30. 30. Lapaire O, Bianchi DW, Peter I, O'Brien B, Stroh H, et al. (2007) Cell-free fetal DNA in amniotic fluid: unique fragmentation signatures in euploid and aneuploid fetuses. Clin Chem 53: 405–411. doi: 10.1373/clinchem.2006.076083
  31. 31. Hui L, Wick HC, Edlow AG, Cowan JM, Bianchi DW (2013) Global gene expression analysis of term amniotic fluid cell-free fetal RNA. Obstet Gynecol 121: 1248–1254. doi: 10.1097/aog.0b013e318293d70b
  32. 32. Hui L, Wick HC, Moise KJ Jr, Johnson A, Luks F, et al. (2013) Global gene expression analysis of amniotic fluid cell-free RNA from recipient twins with twin-twin transfusion syndrome. Prenat Diagn doi: 10.1002/pd.4150
  33. 33. Gentleman RC, Carey VJ, Bates DM, Bolstad B, Dettling M, et al. (2004) Bioconductor: open software development for computational biology and bioinformatics. Genome Biol 5: R80. doi: 10.1186/gb-2004-5-10-r80
  34. 34. Gentleman R (2005) Reproducible research: a bioinformatics case study. Stat Appl Genet Mol Biol 4: Article2.
  35. 35. Benjamini Y, Hochberg Y (1995) Controlling the false discovery rate: a practical and powerful approach to multiple testing. J Roy Statist Soc Ser B (Methodological) 57: 289–300.
  36. 36. Jolliffee IT (2002) Principal Component Analysis, Second Edition. New York: Springer-Verlag.
  37. 37. IPA website (2013) Calculating and Interpreting the p-values for Functions, Pathways, and Lists in IPA. Available: Accessed 2013 Mar 14.
  38. 38. IPA website (2013) Ingenuity Upstream Regulator Analysis Whitepaper. Available: RegulatorAnalysis Whitepaper.html. Accessed 2013 Jan 1.
  39. 39. Su AI, Wiltshire T, Batalov S, Lapp H, Ching KA, et al. (2004) A gene atlas of the mouse and human protein-encoding transcriptomes. Proc Natl Acad Sci U S A 101: 6062–6067. doi: 10.1073/pnas.0400782101
  40. 40. Kislinger T, Cox B, Kannan A, Chung C, Hu P, et al. (2006) Global survey of organ and organelle protein expression in mouse: combined proteomic and transcriptomic profiling. Cell 125: 173–186. doi: 10.1016/j.cell.2006.01.044
  41. 41. Pennings JL, Kuc S, Rodenburg W, Koster MP, Schielen PC, et al. (2011) Integrative data mining to identify novel candidate serum biomarkers for pre-eclampsia screening. Prenat Diagn 31: 1153–1159. doi: 10.1002/pd.2850
  42. 42. Affymetrix (2012) NA33 release for the HG U133 Plus 2.0, October 30, 2012.
  43. 43. Rassart E, Bedirian A, Do Carmo S, Guinard O, Sirois J, et al. (2000) Apolipoprotein D. Biochim Biophys Acta. 1482: 185–198. doi: 10.1016/s0167-4838(00)00162-x
  44. 44. Wu C, Orozco C, Boyer J, Leglise M, Goodale J, et al. (2009) BioGPS: an extensible and customizable portal for querying and organizing gene annotation resources. Genome Biol 10: R130. doi: 10.1186/gb-2009-10-11-r130
  45. 45. Eurexpress website (2013) Available: Accessed 2013, Jan 15.
  46. 46. He X, Jittiwat J, Kim JH, Jenner AM, Farooqui AA, et al. (2009) Apolipoprotein D modulates F2-isoprostane and 7-ketocholesterol formation and has a neuroprotective effect on organotypic hippocampal cultures after kainate-induced excitotoxic injury. Neurosci Lett 455: 183–186. doi: 10.1016/j.neulet.2009.03.038
  47. 47. Kosacka J, Gericke M, Nowicki M, Kacza J, Borlak J, et al. (2009) Apolipoproteins D and E3 exert neurotrophic and synaptogenic effects in dorsal root ganglion cell cultures. Neuroscience 162: 282–291. doi: 10.1016/j.neuroscience.2009.04.073
  48. 48. Khan MM, Parikh VV, Mahadik SP (2003) Antipsychotic drugs differentially modulate apolipoprotein D in rat brain. J Neurochem 86: 1089–1100. doi: 10.1046/j.1471-4159.2003.01866.x
  49. 49. Choi KH, Zepp ME, Higgs BW, Weickert CS, Webster MJ (2009) Expression profiles of schizophrenia susceptibility genes during human prefrontal cortical development. J Psychiatry Neurosci 34: 450–458.
  50. 50. Mahadik SP, Khan MM, Evans DR, Parikh VV (2002) Elevated plasma level of apolipoprotein D in schizophrenia and its treatment and outcome. Schizophr Res 58: 55–62. doi: 10.1016/s0920-9964(01)00378-4
  51. 51. Thomas EA, Dean B, Pavey G, Sutcliffe JG (2001) Increased CNS levels of apolipoprotein D in schizophrenic and bipolar subjects: implications for the pathophysiology of psychiatric disorders. Proc Natl Acad Sci U S A 98: 4066–4071. doi: 10.1073/pnas.071056198
  52. 52. Ordonez C, Navarro A, Perez C, Astudillo A, Martinez E, et al. (2006) Apolipoprotein D expression in substantia nigra of Parkinson disease. Histol Histopathol 21: 361–366.
  53. 53. Thomas EA, Laws SM, Sutcliffe JG, Harper C, Dean B, et al. (2003) Apolipoprotein D levels are elevated in prefrontal cortex of subjects with Alzheimer's disease: no relation to apolipoprotein E expression or genotype. Biol Psychiatry 54: 136–141. doi: 10.1016/s0006-3223(02)01976-5
  54. 54. Kim WS, Wong J, Weickert CS, Webster MJ, Bahn S, et al. (2009) Apolipoprotein-D expression is increased during development and maturation of the human prefrontal cortex. J Neurochem 109: 1053–1066. doi: 10.1111/j.1471-4159.2009.06031.x
  55. 55. White CL, Pistell PJ, Purpera MN, Gupta S, Fernandez-Kim SO, et al. (2009) Effects of high fat diet on Morris maze performance, oxidative stress, and inflammation in rats: contributions of maternal diet. Neurobiol Dis 35: 3–13. doi: 10.1016/j.nbd.2009.04.002
  56. 56. Patel MP, Masood A, Patel PS, Chanan-Khan AA (2009) Targeting the Bcl-2. Curr Opin Oncol 21: 516–523. doi: 10.1097/cco.0b013e328331a7a4
  57. 57. Niculescu MD, Lupu DS (2009) High fat diet-induced maternal obesity alters fetal hippocampal development. Int J Dev Neurosci 27: 627–633. doi: 10.1016/j.ijdevneu.2009.08.005
  58. 58. Stachowiak EK, Oommen S, Vasu VT, Srinivasan M, Stachowiak M, et al. (2012) Maternal obesity affects gene expression and cellular development in fetal brains. Nutr Neurosci 16: 96–103. doi: 10.1179/1476830512y.0000000035
  59. 59. Blaschke AJ, Weiner JA, Chun J (1998) Programmed cell death is a universal feature of embryonic and postnatal neuroproliferative regions throughout the central nervous system. J Comp Neurol 396: 39–50. doi: 10.1002/(sici)1096-9861(19980622)396:1<39::aid-cne4>;2-j
  60. 60. Rice D, Barone S Jr (2000) Critical periods of vulnerability for the developing nervous system: evidence from humans and animal models. Environ Health Perspect 108 Suppl 3: 511–533. doi: 10.1289/ehp.00108s3511
  61. 61. Herbst AL, Ulfelder H, Poskanzer DC (1971) Adenocarcinoma of the vagina. Association of maternal stilbestrol therapy with tumor appearance in young women. N Engl J Med 284: 878–881. doi: 10.1056/nejm197104222841604
  62. 62. Newbold RR (2011) Developmental exposure to endocrine-disrupting chemicals programs for reproductive tract alterations and obesity later in life. Am J Clin Nutr 94: 1939S–1942S. doi: 10.3945/ajcn.110.001057
  63. 63. Durchdewald M, Angel P, Hess J (2009) The transcription factor Fos: a Janus-type regulator in health and disease. Histol Histopathol 24: 1451–1461.
  64. 64. Morton GJ, Schwartz MW (2010) Leptin and the central nervous system control of glucose metabolism. Physiol Rev 91: 389–411. doi: 10.1152/physrev.00007.2010
  65. 65. Hastings NE, Simmers MB, McDonald OG, Wamhoff BR, Blackman BR (2007) Atherosclerosis-prone hemodynamics differentially regulates endothelial and smooth muscle cell phenotypes and promotes pro-inflammatory priming. Am J Physiol Cell Physiol 293: C1824–1833. doi: 10.1152/ajpcell.00385.2007
  66. 66. Min W, Bin ZW, Quan ZB, Hui ZJ, Sheng FG (2009) The signal transduction pathway of PKC/NF-kappa B/c-fos may be involved in the influence of high glucose on the cardiomyocytes of neonatal rats. Cardiovasc Diabetol 8: 8. doi: 10.1186/1475-2840-8-8
  67. 67. Greenhill CJ, Rose-John S, Lissilaa R, Ferlin W, Ernst M, et al. (2011) IL-6 trans-signaling modulates TLR4-dependent inflammatory responses via STAT3. J Immunol 186: 1199–1208. doi: 10.4049/jimmunol.1002971
  68. 68. Hirano T, Ishihara K, Hibi M (2000) Roles of STAT3 in mediating the cell growth, differentiation and survival signals relayed through the IL-6 family of cytokine receptors. Oncogene 19: 2548–2556. doi: 10.1038/sj.onc.1203551
  69. 69. Kirk SL, Samuelsson AM, Argenton M, Dhonye H, Kalamatianos T, et al. (2009) Maternal obesity induced by diet in rats permanently influences central processes regulating food intake in offspring. PLoS One 4: e5870. doi: 10.1371/journal.pone.0005870
  70. 70. Clementi AH, Gaudy AM, van Rooijen N, Pierce RH, Mooney RA (2009) Loss of Kupffer cells in diet-induced obesity is associated with increased hepatic steatosis, STAT3 signaling, and further decreases in insulin signaling. Biochim Biophys Acta 1792: 1062–1072. doi: 10.1016/j.bbadis.2009.08.007
  71. 71. Perdomo G, Kim DH, Zhang T, Qu S, Thomas EA, et al. (2010) A role of apolipoprotein D in triglyceride metabolism. J Lipid Res 51: 1298–1311. doi: 10.1194/jlr.m001206
  72. 72. Baker WA, Hitman GA, Hawrami K, McCarthy MI, Riikonen A, et al. (1994) Apolipoprotein D gene polymorphism: a new genetic marker for type 2 diabetic subjects in Nauru and south India. Diabet Med 11: 947–952. doi: 10.1111/j.1464-5491.1994.tb00252.x
  73. 73. Vijayaraghavan S, Hitman GA, Kopelman PG (1994) Apolipoprotein-D polymorphism: a genetic marker for obesity and hyperinsulinemia. J Clin Endocrinol Metab 79: 568–570. doi: 10.1210/jcem.79.2.7913935
  74. 74. Liu Z, Chang GQ, Leibowitz SF (2001) Apolipoprotein D interacts with the long-form leptin receptor: a hypothalamic function in the control of energy homeostasis. FASEB J 15: 1329–1331. doi: 10.1096/fj.00-0530fje
  75. 75. Heerwagen MJ, Miller MR, Barbour LA, Friedman JE (2010) Maternal obesity and fetal metabolic programming: a fertile epigenetic soil. Am J Physiol Regul Integr Comp Physiol 299: R711–722. doi: 10.1152/ajpregu.00310.2010
  76. 76. Armitage JA, Poston L, Taylor PD (2008) Developmental origins of obesity and the metabolic syndrome: the role of maternal obesity. Front Horm Res 36: 73–84. doi: 10.1159/000115355
  77. 77. Romero R, Tromp G (2006) High-dimensional biology in obstetrics and gynecology: functional genomics in microarray studies. Am J Obstet Gynecol 195: 360–363. doi: 10.1016/j.ajog.2006.06.077
  78. 78. Olson NE (2006) The microarray data analysis process: from raw data to biological significance. NeuroRx 3: 373–383. doi: 10.1016/j.nurx.2006.05.005