Title. Impact of maternal obesity on placental transcriptome and morphology associated 1 with fetal growth restriction in mice 2 3 Running title. Placental transcriptome and morphology in obesity 4 5 Authors and affiliations. *Daniela de Barros Mucci1,2,3, *Laura C. Kusinski1, Phoebe 6 Wilsmore1, Elena Loche1, Lucas C. Pantaleão1, Thomas J. Ashmore1, Heather L. Blackmore1, 7 Denise S. Fernandez-Twinn1, Maria das Graças T. do Carmo2 and Susan E. Ozanne1 8 1 University of Cambridge Metabolic Research Laboratories and MRC Metabolic Diseases 9 Unit, Wellcome Trust-MRC Institute of Metabolic Science, Cambridge, UK 10 2 Nutritional Biochemistry Laboratory, Institute of Nutrition Josué de Castro, Federal 11 University of Rio de Janeiro, Rio de Janeiro, RJ, Brazil 12 3 Nutritional Epidemiology Observatory, Institute of Nutrition Josué de Castro, Federal 13 University of Rio de Janeiro, Rio de Janeiro, RJ, Brazil 14 * Joint corresponding authors: 15 Daniela de Barros Mucci. Address: Avenida Carlos Chagas Filho, 373 - Bloco J, 2º andar - 16 Rio de Janeiro/RJ, Brazil - 21941-902. Telephone number: +55 21 991960867. e-mail: 17 danimucci@gmail.com. 18 Laura C. Kusinski. Address: University of Cambridge Metabolic Research Laboratories and 19 MRC Metabolic Diseases Unit, Level 4, Institute of Metabolic Science, Addenbrookes 20 Hospital Cambridge, CB2 0QQ UK. Telephone number: +44 1223 336784. e-mail: 21 lck34@medschl.cam.ac.uk. 22 * These authors contributed equally to this work. 23 24 Competing Interests. This work was supported by the Biotechnology and Biological Sciences 25 Research Council (BBSRC—BB/M001636/1) and an MRC Metabolic Diseases Unit award 26 (MC_UU_12012/4). DBM was the recipient of a FAPERJ sandwich doctorate scholarship 27 (Carlos Chagas Filho Research Support Foundation—FAPERJ—Brazil—E-26/ 28 200.090/2016). PW is a recipient of a Wellcome Trust studentship (Wellcome - 29 215242/Z/19/Z). LCP was the recipient of a CNPq Science Without Borders Post-Doctoral 30 Fellowship (National Council of Technological and Scientific Development—CNPq—31 Brazil—PDE/204416/2014-0). The authors declare no competing financial interests. 32 Abstract 33 Background: In utero exposure to obesity is consistently associated with increased risk of 34 metabolic disease, obesity and cardiovascular dysfunction in later life despite the divergence 35 of birth weight outcomes. The placenta plays a critical role in offspring development and 36 long-term health, as it mediates the crosstalk between the maternal and fetal environments. 37 However, its phenotypic and molecular modifications in the context of maternal obesity 38 associated with fetal growth restriction remain poorly understood. 39 Methods: Using a mouse model of maternal diet-induced obesity, we investigated changes in 40 the placental transcriptome through RNA-seq and Ingenuity Pathway Analysis (IPA) at 41 embryonic day (E) 19. The most differentially expressed genes (FDR < 0.05) were validated 42 by qPCR in male and female placentae at E19. The expression of these targets and related 43 genes was also determined by qPCR at E13 to examine whether the observed alterations had 44 an earlier onset at mid-gestation. Structural analyses were performed using 45 immunofluorescent staining against Ki-67 and CD31 to investigate phenotypic outcomes at 46 both time points. 47 Results: RNA-seq and IPA analyses revealed differential expression of transcripts and 48 pathway interactions related to placental vascular development and tissue morphology in 49 obese placentae at term, including downregulation of Muc15, Cnn1 and Acta2. Pdgfb, which 50 is implicated in labyrinthine layer development, was downregulated in obese placentae at 51 E13. This was consistent with the morphological evidence of reduced labyrinth zone size, as 52 well as lower fetal weight at both time points irrespective of offspring sex. 53 Conclusions: Maternal obesity results in abnormal placental labyrinth zone development and 54 impaired vascularization, which may mediate the observed fetal growth restriction through 55 reduced transfer of nutrients across the placenta. 56 57 Introduction 58 The prevalence of obesity has nearly tripled since 1975 [1]. As a consequence, the 59 number of women who are classified as overweight or obese during pregnancy has risen 60 substantially, estimated at 38.9 million in 2014 [2]. This is especially concerning as offspring 61 born to obese mothers are more likely to have poor neonatal outcomes [3] and to develop 62 obesity, insulin resistance, hypertension and dyslipidemia later in life [4]. Interestingly, 63 maternal obesity leads to divergent birth weight outcomes; while it is often shown to increase 64 the risk of macrosomia [5, 6], a higher incidence of low birth weight is also documented [6-65 8]. Although both are similarly associated with metabolic disease in later life, distinct 66 placental alterations seem to mediate these contrasting offspring phenotypes [9]. 67 Changes in placental function are thought to be pivotal in the development of 68 pregnancy complications [10, 11] and could also be a key link between the maternal and 69 intrauterine milieu and long-term health of the offspring [12, 13]. Alterations in placental 70 function and structure in response to obesity and their underlying molecular mechanisms 71 have been explored both in humans and in animal models [9, 14-17]. Yet, even though fetal 72 growth restriction (FGR) is recognized as a placenta-related disorder [18], the impact of 73 maternal obesity on the placental transcriptome in this context remains largely unknown. 74 It has been shown that placentae from high fat diet-fed obese mouse dams exhibit 75 altered expression of epigenetic machinery genes at term, which could alter the placental 76 epigenome and lead to FGR [19]. High fat diet-induced obesity has also been found to alter 77 the transcriptome of placenta progenitor cells at early stages of development and is associated 78 with later changes in placental function resulting in FGR [17]. In our mouse model of 79 maternal diet-induced obesity, in which dams are fed a hypercaloric Western-like diet, we 80 have shown that maternal hyperinsulinemia is strongly associated with offspring insulin 81 resistance and excess placental lipid deposition and hypoxia [20]. However, a clear 82 understanding of the molecular mechanisms behind these findings is still lacking and 83 warrants further investigation. 84 It is recognized that the impact of stressors on placental function and offspring health 85 is closely linked to the stage of tissue development, the type of insult and the sex of the 86 conceptus [21]. Thus, the aim of this study was to identify global changes in the placental 87 transcriptome and related pathways in response to maternal obesity near term at embryonic 88 day (E) 19. Furthermore, we investigated whether the significant transcriptional alterations 89 detected in obese placentae were manifested earlier, i.e. in mid-gestation (E13), and if these 90 alterations translated into a structural phenotype in male and female placentae. 91 92 Methods 93 Animals and diets 94 All experimental protocols were approved by the University of Cambridge Animal 95 Welfare and Ethical Review Board and were carried under the Home Office Animals 96 (Scientific Procedures) Act 1986. The model has been described in detail previously [20, 22]. 97 Briefly, female C57BL/6J mice, proven breeders, were randomly assigned either a standard 98 chow RM1 diet [7% simple sugars, 3% fat (wt/wt), 10.74 kJ/g] or an energy-rich highly 99 palatable obesogenic diet [10% simple sugars, 20% animal lard (wt/wt), 28.43 kJ/g] 100 supplemented with sweetened condensed milk [55% simple sugar, 8% fat (wt/wt); Nestle, 101 Croydon, UK] and fortified with mineral and vitamin mix AIN93G. Both diets were fed ad 102 libitum and purchased from Special Dietary Services (Witham, UK). Body composition was 103 monitored (TD-NMR, Bruker Minispec) and females were set up to breed if body fat was 104 between the thresholds of 10-12% or 35-40% for Control and Obese dams, respectively. 105 After mating for the second time with RM1 fed males, dams were killed at either E13 or E19 106 by rising CO2 concentration. Fetal and placental weights were recorded. Placentae for 107 molecular analysis were immediately snap frozen on dry ice and stored at -80°C. For 108 morphological assessment samples were fixed in 10% formalin for 48h, stored in 70% 109 ethanol and then embedded in wax. 110 The sex of the fetuses at E19 was determined by visual inspection of anogenital 111 anatomy. At E13, DNA extracted from tail tips was used for PCR sexing as described by 112 McFarlane et al. (2013), using the SX primer pair [23]. Amplicons were loaded on 2% 113 agarose gels and submitted to electrophoresis together with a 1 Kb DNA ladder. Bands were 114 visualized with SYBR™ Safe DNA gel stain (Thermo Fisher Scientific, Rochford, UK) 115 under UV-illumination and the genomic sex of each sample was determined according to the 116 number of bands and amplicon size. 117 118 RNA extraction 119 Placenta aliquots were homogenized in 700 uL Qiazol using TissueRuptor (Qiagen, 120 Manchester, UK). Total RNA was isolated with miRNeasy Mini Kit (Qiagen) according to 121 the manufacturer's instructions and including the optional step of DNA digestion with RNase-122 Free DNAse Set (Qiagen). Extracted RNA was quantified by spectrophotometry (Nanodrop™ 123 Thermo Fisher Scientific) and stored at -80 °C. 124 125 RNA sequencing and Ingenuity® Pathway Analysis 126 Total RNA was extracted from E19 male placentae (Control n = 2; Obese n = 3), as 127 previously outlined. One µg of total RNA was depleted of ribosomal RNA and PolyA tails of 128 coding RNAs were captured by treatment with Oligo-dT beads. Complementary DNA 129 (cDNA) libraries were generated after an amplification step, according to the TruSeq 130 Stranded Total RNA Sample Preparation Guide (Illumina, San Diego, CA, USA), and 131 quantified using KAPA Library Quantification Kit. Multiplex single-read sequencing was 132 performed using Illumina HiSeq 2500 (Illumina). Sequence reads were demultiplexed using 133 the CASAVA pipeline (Illumina) and then aligned to the Mus Musculus genome (GRCm38) 134 using TopHat version 2.0.11. Raw read counts and fragments per kilobase of transcript per 135 million mapped reads (FPKM) were generated using Cufflinks version 2.2.1. A quality check 136 of mapped reads was executed using the R package CummeRbund. Databases were trimmed 137 for exclusion of very low detection or undetectable genes. The resulting data were analyzed 138 using edgeR by calculating the likelihood ratio, and by adjusting P-values via Benjamini and 139 Hochberg’s method to control the false discovery rate. Ingenuity Pathway Analysis (IPA) 140 was applied to identify biological pathways related to the genes that were differentially 141 expressed between Control and Obese E19 male placentae. The placenta RNA sequencing 142 (RNA-seq) data has been deposited in GEO database under the accession number 143 GSE140013. 144 145 cDNA synthesis and qPCR 146 Total RNA was extracted from male and female placentae of Control and Obese dams 147 at E13 (n = 10/group from separate litters) and E19 (n=9/group from separate litters). All 148 samples used in the validations were different to those used in the RNA-seq and therefore 149 represent biological replicates. Sample size was based on previous datasets/power 150 calculations. cDNA was generated from 1 μg RNA using High Capacity cDNA Reverse 151 Transcription Kit (Applied Biosystems, Foster City, CA, USA). Quantitative real time PCR 152 (qPCR) was performed on QuantStudio 7 Flex Real-Time PCR System (Applied 153 Biosystems), using 200 nM specific primers (Sigma-Aldrich, Gillingham, UK), 1× 154 SYBR® Green JumpStart™ Taq ReadyMix (Sigma-Aldrich) and cDNA samples at a final 155 dilution of 1:60. For primer sequences see Supplementary Table S1. NormFinder software 156 was used to select the best combination of two out of four reference genes [24]. qPCR results 157 were normalized to the geometric mean of the reference genes Gapdh and Sdha for E19 158 placentae, and Gapdh and Pmm1 for E13 placentae, expression of which did not change 159 between groups. Data was expressed in arbitrary units relative to Male Control average (2-160 ΔΔCq). 161 162 Structural analyses 163 E13 and E19 formalin-fixed placentae from males and females were cut into 5µm 164 sections. Three serial sections close to the midline of each placenta were selected for staining. 165 Antigen retrieval was performed with pH9 Target Retrieval buffer [S2375 (Dako Agilent, 166 Stockport, UK)] and non-specific binding sites were blocked with 3% goat serum. Sections 167 were incubated overnight at 4oC with rabbit polyclonal primary antibodies against Ki67 168 [1:200; ab15580 (Abcam, Cambridge, UK)] or CD31 [1:500; ab124432 (Abcam)]. 169 Primary antibody binding was visualized by incubation at room temperature for 1 170 hour with a fluorescent-conjugated goat polyclonal anti-rabbit IgG [1:1000; A11008 171 (Invitrogen, Warrington, UK)]. In negative control slides, placental sections were incubated 172 in 1.5% goat serum in TBS-T instead of primary antibodies (Supplementary Figure S1). All 173 sections were incubated with 1:2500 DAPI for 10 minutes at room temperature to stain 174 nuclei. Autofluorescence was quenched by incubation with Vector TrueVIEW 175 Autofluorescence Quenching Kit [SP8400 (Vector Laboratories, Peterborough, UK)]. 176 Placental sections were imaged on an AxioScan Slide Scanner (Zeiss, Cambridge, 177 UK) and blindly analyzed using HALO analysis software (Indica Labs, Corrales, NM, USA). 178 The DAPI channel image was used to define nuclear outlines using the CytoNuclear FL 179 module. For the analysis of Ki67 staining (Supplementary Figure S2), nuclear outlines were 180 transposed onto the 488nm channel image and the proportion of nuclei positive for Ki67 181 across the whole section was recorded. Placental sections stained for the endothelial cell 182 marker CD31 were used to investigate labyrinth zone (LZ) size and structure 183 (Supplementary Figure S3). The border of the LZ was outlined manually at 40X 184 magnification (depicted in yellow) and total area was recorded using HALO analysis 185 software after canals were excluded. The boundary with the junctional zone was determined 186 as the interface between the phenotypically distinct spongiotrophoblast of the junctional zone 187 and the fetal capillaries of the LZ. The rest of the boundary was either at the edge of the 188 tissue image or at the interface with the chorionic plate, which is structurally distinct from the 189 LZ, characterized by smaller nuclei and the absence of CD31-positive endothelial cells. 190 Indica Labs' Tissue Classifier module was used to differentiate between fetal blood vessels 191 (lumen bound by CD31-positive endothelium) and other tissue of the LZ. 192 193 Statistical analyses 194 Benjamini-Hochberg multiple testing correction [25] was applied to the RNA-seq 195 differential expression data and only genes with False Discovery Rate (FDR) < 0.05 were 196 considered significantly different between the two experimental conditions. For qPCR 197 validation of RNA-seq differentially expressed genes, comparisons were made between 198 Control and Obese placentae of the same sex, by Student's t test. 199 Anthropometric parameters and qPCR of selected targets were analyzed by two-way 200 analysis of variance (ANOVA) followed by Tukey's multiple comparisons test to estimate the 201 effects of maternal obesity and fetal sex at each time point. 202 For the morphological analyses, the effects of gestational age (E13 or E19), offspring 203 sex (male or female) and maternal diet (regular chow or obesogenic diet) on placental 204 phenotype were investigated by three-way ANOVA, and backwards stepwise elimination was 205 used to come to a minimal model. Three-way ANOVA of the proportional area of the LZ 206 that was fetal capillaries data revealed no significant effect of gestational age, maternal diet 207 or fetal sex. However, there was a borderline significant interaction between maternal diet 208 and fetal sex (P = 0.055). In order to identify if a maternal diet effect was only present in one 209 sex, these data were separated and sex-specific two-way ANOVAs (gestational age / maternal 210 diet) were performed. 211 Only one sample from each litter was used for each analysis, except in the case of 212 offspring and placental weights, in which each litter’s average was used as a single data 213 point. Data are presented as mean ± standard error of the mean (SEM), and the threshold for 214 significance was set at P< 0.05, unless stated otherwise. Statistical analyses were performed 215 using R (R Core Team 2017) or Prism 6 (GraphPad Prism, La Jolla, CA, USA). 216 217 Results 218 Fetal and placental measurements 219 Fetal and placental weights were reduced in response to obesity and female placentae 220 were smaller than those of males at both stages of gestation. There was no significant 221 difference in the ratio of fetal to placental weight, although there was a trend (P = 0.05) 222 towards higher placental efficiency in females at E19 (Table 1). 223 224 RNA-seq, Ingenuity® Pathway Analysis and qPCR at term placentae 225 The RNA-seq at E19 detected a total of 350 transcripts differentially expressed in 226 placentae of Obese compared to Control males considering a significance threshold of 227 P < 0.05 (Figure 1A, Supplementary Table S2). However, only 9 genes remained 228 significantly altered after correction for multiple testing (FDR < 0.05) (Figures 1A and 1B). 229 Ingenuity® Pathway Analysis (IPA) was used with a less stringent threshold (P < 0.05) to 230 identify global changes in pathways and biological functions promoted by maternal obesity. 231 The most significant diseases and bio functions are shown in Figure 1C. 232 Genes identified as significantly changed in response to obesity by RNA-seq (FDR < 233 0.05) were validated in a larger number of samples, all from independent litters, by qPCR 234 (Figure 2) and results were confirmed in 8 out of the 9 genes in E19 male placentae (i.e. 235 Pi15, Gabrd, Sez6l, Nup210, Acta2, Rnf222, Muc15 and Cnn1). These genes were also 236 examined in female placentae, however, only Pi15, Nup210, Acta2, Rnf222 and Muc15 were 237 significantly modulated by obesity (Figure 2). 238 239 Placental gene expression at different gestational ages 240 All 9 genes found differentially expressed in obese male placentae at E19 were then 241 investigated in E13 placentae (Figure 3A). Pi15, Nup210 and Sez6l were upregulated by 242 maternal obesity in both sexes at mid-gestion. Gabrd mRNA levels, which were upregulated 243 in Obese male placentae at E19, was downregulated by maternal obesity at E13. Muc15, 244 Nup210 and Acta2 expression was higher in females compared to males at E13. Rnf222 and 245 Cnn1 were not differentially expressed in either sex at E13, however, their transcript levels 246 were very low compared to E19. 247 Since genes involved in spiral artery remodeling (Muc15 and Cnn1) and labyrinthine 248 pericytes (Acta2) that were found to be dysregulated in obese placentae at E19 were not 249 affected at E13, we additionally measured the expression of candidate genes recognized as 250 relevant for these processes at mid-gestation. Both Hand1 and Pdgfb were downregulated by 251 maternal obesity in both sexes, but no differences were observed in Prl2c2 (Figure 3B). 252 253 Immunofluorescent staining of placental morphology 254 Phenotypic analyses were also conducted by immunofluorescent staining of targets 255 within pathways identified by IPA. The marker Ki67 was used to investigate cellular growth 256 and proliferation. Cellular movement, assembly and organization was assessed through 257 analyses of LZ size and fetal vasculature structure, using CD31 as a marker of fetal 258 endothelial cells. 259 E19 placentae had fewer cells (P < 0.05, Figures 4A-C) and a lower proportion of 260 proliferating cells across the whole placenta (P < 0.01, Figures 4D-H) compared to E13. 261 These parameters were not affected by offspring sex or maternal diet. 262 The size of the LZ was significantly reduced in placentae from obese dams (P < 0.01, 263 Figures4I-M). There was a significant increase in LZ size from mid-gestation to term (P < 264 0.01, Figure 4I). Labyrinthine vascular organization was analyzed by the proportion of the 265 total LZ area that was fetal blood vessels. A reduced model considering males and females 266 separately by two-way ANOVA (gestational age / maternal condition) detected a reduction in 267 area bound by fetal capillaries within the LZ in female placentae of obese dams (P < 0.05, 268 Figures 4N-P). Representative images of male placentae are shown in Supplementary 269 Figure S4. 270 271 Discussion 272 The RNA-seq analysis revealed a total of 350 transcripts differentially expressed in 273 Obese male placentae at term. Among the top downregulated transcripts, Muc15, Cnn1 and 274 Acta2 were of particular significance as these genes are required for appropriate development 275 of placental vasculature and related to key pathways identified by IPA such as cellular 276 movement, assembly and organization. Previous studies have shown that Muc15 suppresses 277 the migration/invasion of trophoblast like-cells in vitro, a process implicated in blood vessel 278 remodeling in the maternal–fetal interface [26]. Cnn1 is largely expressed by smooth muscle 279 cells [27] which line the uterine blood vessels and are lost in the normal remodeling of 280 maternal spiral arteries during placental development [28]. Acta2 is a marker of pericytes 281 which surround fetal endothelial cells during blood vessel development in the mouse LZ [29]. 282 These data together suggest that exposure to maternal obesity affects the remodeling 283 of maternal spiral arteries and the development of fetal blood vessels within the LZ, both of 284 which are crucial for adequate nutrient and oxygen transfer across the placenta [28] and 285 possibly linked to the uteroplacental hemodynamic alterations present in pre-eclampsia and 286 intrauterine growth restriction [30-32]. Obese women are two to three times more likely to 287 develop pre-eclampsia [33] and hypertensive obstetric complications are generally associated 288 with small-for-gestational age neonates [34]. Here we see significant growth restriction in the 289 fetus which may result from poor utero-placental perfusion in addition to placental hypoxia 290 previously suggested in this model [20]. 291 Furthermore, the RNA-seq analysis identified several genes that have not been 292 functionally described in placental tissue thus far, but are conserved in humans and rodents. 293 Pi15 encodes a peptidase inhibitor that may regulate extracellular matrix modifications [35] 294 and has been implicated in vascular defects in rat aorta [36], though its role in placental 295 vascularization is unknown. Gabrd gene encodes the delta subunit of gamma-aminobutyric 296 acid type A receptor (GABAA). GABAA activation impacts stromal cell proliferation and 297 apoptosis during decidualization [37] and increased expression of its pi subunit (GABRP) has 298 been detected in preeclamptic placentae [38]. NUP210 is a major component of the nuclear 299 pore complex and is required for regulation of gene expression during differentiation and cell 300 fate determination, as demonstrated in myoblasts and embryonic stem cells [39]. Although 301 the function of SEZ6L is not well understood, it has been shown both in mice and in vitro 302 that this protein is almost exclusively processed by β-site APP cleaving enzyme (BACE) 303 [40]; BACE1 and BACE2 are abundantly expressed in human placentae and are up-regulated 304 in pregnancies complicated with preeclampsia [41]. Lastly, Rnf222 is also a protein-coding 305 gene, however no functional description has been found. 306 Although information on these genes in placentae is currently limited, it must be 307 noted that a large number of placental genes and related phenotypes remain uncharacterized. 308 Recent efforts to systematically identify the genes required for normal embryogenesis are still 309 unravelling many previously underappreciated placental defects [42]. Thus, our findings 310 might represent novel targets that could be implicated in the pathophysiology of maternal 311 obesity and associated adverse outcomes in the offspring. Moreover, additional genes might 312 have been identified if a larger sample size was used in the RNA-seq analysis. 313 When comparing both time points, most transcripts exhibited a different expression 314 pattern, including Muc15, Cnn1 and Acta2 which were not affected by obesity at E13. This 315 could be due to low functional relevance of these transcripts at mid-pregnancy rather than 316 absence of alterations in related cellular processes, as illustrated by low Cnn1 mRNA levels 317 in our analysis at E13 compared to E19 (data not shown). In fact, the mouse placenta 318 undergoes a transcriptome transition from the “development phase” of organogenesis to the 319 “mature phase” at mid-pregnancy [43]. 320 Thus, we next used the IPA data to identify genes which were previously shown to be 321 both highly expressed at mid-pregnancy and pivotal to spiral artery remodeling and formation 322 of fetal blood vessels in the LZ. Hand1, which is required for trophoblast giant cell (TGC) 323 differentiation [44], was downregulated in placentae of obese dams. However, maternal 324 obesity had no effect on the expression of Prl2c2, a marker of TGC that line maternal blood 325 canal spaces and spiral arteries in the definitive placenta [45]. Considering the complexity of 326 spiral artery remodeling and the limited number of transcripts that were analyzed here, it 327 remains to be established whether the alterations occur only later in development or if other 328 mechanisms are involved. 329 On the other hand, the growth factor Pdgfb was downregulated in response to obesity 330 at E13 and could be a relevant link to other molecular and phenotypic observations in our 331 model. It has been shown that Pdgfb-deficient placentae exhibit defective labyrinthine 332 development, with alterations in fetal blood vessel structure and reduced numbers of 333 pericytes from mid-pregnancy until term, leading to growth restriction in PDGFB -/- embryos 334 [29]. Here, lower expression of the pericyte marker Acta2 was detected by RNA-seq and 335 confirmed by qPCR in obese placentae at E19. In addition, defects in LZ morphology and 336 FGR were observed in response to maternal obesity at E13 and persisted until E19. 337 As shown by our immunofluorescence staining, male and female placentae from 338 Obese dams exhibited reduced LZ area, which is the primary site of gas, nutrient and waste 339 exchange between the maternal and fetal circulations in the mouse [28], and a decrease in the 340 proportion of fetal blood vessels within the LZ was also evident in females. This is further 341 corroborated by recent evidence of lower vascularity in placentae of high fat diet-fed dams 342 both at mid-pregnancy and near term, which was associated with placental transcriptome 343 alterations in early stages of development and FGR [17]. In addition, it has been suggested 344 that defects in placental villi vasculature seen in obese human pregnancies could be partly 345 due to obesity-associated tissue hypoxia [46], which is also consistent with our model [20]. 346 Next, we used IPA to investigate the mechanism behind this reduction in LZ area. 347 Cellular growth and proliferation were pointed out as the main molecular and cellular 348 functions affected by obesity. Surprisingly, however, maternal obesity had no effect on the 349 number of Ki67-positive cells in the placenta. Abnormalities in placental size are often 350 associated with disruption of cellular growth and/or apoptosis [47-49]. Thus, it is possible 351 that other mechanisms such as cell death could explain our results, although limitations to our 352 analysis, which was not zone-specific, cannot be discounted. In this regard, it has been shown 353 in a mouse model of high fat diet-induced obesity through phosphohistone H3 staining that 354 the proliferating cells in placenta are mostly restricted to the labyrinthine layer and appear 355 reduced in response to obesity within this region [50]. 356 Despite these morphological disturbances, changes in tissue structure that are 357 expected to occur from mid-pregnancy until term seemed preserved in obese placentae. The 358 LZ is well-reported to expand as pregnancy progresses, so that the transport capacity of the 359 placenta can meet the nutrient demands of the growing fetus [50-52]. Accordingly, we found 360 that labyrinth area increased between E13 and E19, irrespective of maternal diet. 361 We also observed significant differences in placentae which are specific to fetal sex. 362 Female placentae were smaller than male counterparts at all time points and maternal 363 conditions, which is consistent with observations from both human cohorts [53] and studies 364 in mice [19, 54]. Moreover, we found sex differences in a subset of genes, with females 365 exhibiting slightly increased expression. Similarly, global transcriptomic analysis in normal 366 full-term human placentae revealed higher overall mRNA levels in females compared to 367 males [55]. Sexual dimorphism in the context of developmental programming is increasingly 368 commonly reported [56]. How these relate to sex-specific responses of the placenta to a 369 suboptimal environment remain to be determined. 370 Overall, we have shown through genome-wide analysis that maternal obesity induces 371 a dysregulation of transcripts and pathway interactions related to placental vasculature and 372 structure. Fetal growth restriction, as well as changes in placental morphology and a gene 373 expression signature associated with impaired labyrinthine development, were detectable at 374 mid-pregnancy, suggesting an enduring negative effect of maternal obesity over these 375 processes. The LZ is the exchange region of the murine placenta and reductions in its size 376 and vasculature may impair the transport of nutrients from the maternal circulation to the 377 developing fetus, thus restricting its growth. Disruption of placental structure could thus 378 represent an important factor contributing to the development of FGR in pregnancies 379 complicated by maternal obesity. Moreover, novel targets were revealed by our RNA-seq 380 analysis. Characterizing their functional roles in the placenta will help us better understand 381 the processes mediating the effects of maternal obesity on offspring outcomes and potentially 382 inform suitable interventions. 383 384 Acknowledgments 385 The authors would like to thank Ania Wilczynska and Martin Bushell from the 386 University of Leicester for their helpful analysis on the RNA-seq data and Claire Custance 387 for technical assistance. We also thank BBSRC, BHF, MRC Metabolic Diseases Unit, 388 Wellcome Trust, FAPERJ and CNPq for the financial support. 389 390 Competing Interests 391 This work was supported by the BBSRC (BB/M001636/1), the BHF 392 (PG/13/46/30329) and the MRC Metabolic Diseases Unit [MC_UU_12012/4]. DBM was the 393 recipient of a FAPERJ scholarship (E-26/200.090/2016). PW was the recipient of a 394 Wellcome Trust studentship (Wellcome-215242/Z/19/Z). LCP was the recipient of a CNPq 395 Post-Doctoral Fellowship (PDE/204416/2014-0). The authors declare that no conflict of 396 interest exists. 397 398 Supplementary information is available at International Journal of Obesity’s website. 399 References 400 1. WHO. World Health Organization: Obesity and overweight. Fact sheet Nº 311. 2018 401 [updated 16 February 2018]. Available from: http://www.who.int/news-room/fact-402 sheets/detail/obesity-and-overweight. 403 2. Chen C, Xu X, Yan Y. Estimated global overweight and obesity burden in pregnant 404 women based on panel data model. PloS one. 2018;13(8):e0202183. 405 3. Marchi J, Berg M, Dencker A, Olander EK, Begley C. Risks associated with obesity 406 in pregnancy, for the mother and baby: a systematic review of reviews. Obesity reviews : an 407 official journal of the International Association for the Study of Obesity. 2015;16(8):621-638. 408 4. Godfrey KM, Reynolds RM, Prescott SL, Nyirenda M, Jaddoe VW, Eriksson JG, et 409 al. Influence of maternal obesity on the long-term health of offspring. The lancet Diabetes & 410 endocrinology. 2017;5(1):53-64. 411 5. Yu Z, Han S, Zhu J, Sun X, Ji C, Guo X. Pre-pregnancy body mass index in relation 412 to infant birth weight and offspring overweight/obesity: a systematic review and meta-413 analysis. PloS one. 2013;8(4):e61627. 414 6. Rode L, Nilas L, Wojdemann K, Tabor A. Obesity-related complications in Danish 415 single cephalic term pregnancies. Obstetrics and gynecology. 2005;105(3):537-542. 416 7. Radulescu L, Munteanu O, Popa F, Cirstoiu M. The implications and consequences of 417 maternal obesity on fetal intrauterine growth restriction. Journal of medicine and life. 418 2013;6(3):292-298. 419 8. Rajasingam D, Seed PT, Briley AL, Shennan AH, Poston L. A prospective study of 420 pregnancy outcome and biomarkers of oxidative stress in nulliparous obese women. 421 American journal of obstetrics and gynecology. 2009;200(4):395 e391-399. 422 9. Howell KR, Powell TL. Effects of maternal obesity on placental function and fetal 423 development. Reproduction. 2017;153(3):R97-R108. 424 10. Flenady V, Koopmans L, Middleton P, Froen JF, Smith GC, Gibbons K, et al. Major 425 risk factors for stillbirth in high-income countries: a systematic review and meta-analysis. 426 Lancet. 2011;377(9774):1331-1340. 427 11. Moran MC, Mulcahy C, Zombori G, Ryan J, Downey P, McAuliffe FM. Placental 428 volume, vasculature and calcification in pregnancies complicated by pre-eclampsia and intra-429 uterine growth restriction. European journal of obstetrics, gynecology, and reproductive 430 biology. 2015;195:12-17. 431 12. Dimasuay KG, Boeuf P, Powell TL, Jansson T. Placental Responses to Changes in the 432 Maternal Environment Determine Fetal Growth. Frontiers in physiology. 2016;7:12. 433 13. Lewis RM, Demmelmair H, Gaillard R, Godfrey KM, Hauguel-de Mouzon S, 434 Huppertz B, et al. The placental exposome: placental determinants of fetal adiposity and 435 postnatal body composition. Annals of nutrition & metabolism. 2013;63(3):208-215. 436 14. Altmae S, Segura MT, Esteban FJ, Bartel S, Brandi P, Irmler M, et al. Maternal Pre-437 Pregnancy Obesity Is Associated with Altered Placental Transcriptome. PloS one. 438 2017;12(1):e0169223. 439 15. Saben J, Kang P, Zhong Y, Thakali KM, Gomez-Acevedo H, Borengasser SJ, et al. 440 RNA-seq analysis of the rat placentation site reveals maternal obesity-associated changes in 441 placental and offspring thyroid hormone signaling. Placenta. 2014;35(12):1013-1020. 442 16. Saben J, Lindsey F, Zhong Y, Thakali K, Badger TM, Andres A, et al. Maternal 443 obesity is associated with a lipotoxic placental environment. Placenta. 2014;35(3):171-177. 444 17. Stuart TJ, O'Neill K, Condon D, Sasson I, Sen P, Xia Y, et al. Diet-induced obesity 445 alters the maternal metabolome and early placenta transcriptome and decreases placenta 446 vascularity in the mouse. Biology of reproduction. 2018;98(6):795-809. 447 18. Kingdom JC, Audette MC, Hobson SR, Windrim RC, Morgen E. A placenta clinic 448 approach to the diagnosis and management of fetal growth restriction. American journal of 449 obstetrics and gynecology. 2017. 450 19. Panchenko PE, Voisin S, Jouin M, Jouneau L, Prezelin A, Lecoutre S, et al. 451 Expression of epigenetic machinery genes is sensitive to maternal obesity and weight loss in 452 relation to fetal growth in mice. Clinical epigenetics. 2016;8:22. 453 20. Fernandez-Twinn DS, Gascoin G, Musial B, Carr S, Duque-Guimaraes D, Blackmore 454 HL, et al. Exercise rescues obese mothers' insulin sensitivity, placental hypoxia and male 455 offspring insulin sensitivity. Scientific reports. 2017;7:44650. 456 21. Rosenfeld CS. Sex-Specific Placental Responses in Fetal Development. 457 Endocrinology. 2015;156(10):3422-3434. 458 22. Samuelsson AM, Matthews PA, Argenton M, Christie MR, McConnell JM, Jansen 459 EH, et al. Diet-induced obesity in female mice leads to offspring hyperphagia, adiposity, 460 hypertension, and insulin resistance: a novel murine model of developmental programming. 461 Hypertension. 2008;51(2):383-392. 462 23. McFarlane L, Truong V, Palmer JS, Wilhelm D. Novel PCR assay for determining the 463 genetic sex of mice. Sexual development : genetics, molecular biology, evolution, 464 endocrinology, embryology, and pathology of sex determination and differentiation. 465 2013;7(4):207-211. 466 24. Andersen CL, Jensen JL, Orntoft TF. Normalization of real-time quantitative reverse 467 transcription-PCR data: a model-based variance estimation approach to identify genes suited 468 for normalization, applied to bladder and colon cancer data sets. Cancer research. 469 2004;64(15):5245-5250. 470 25. Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and 471 Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B 472 (Methodological). 1995;57(1):289-300. 473 26. Shyu MK, Lin MC, Shih JC, Lee CN, Huang J, Liao CH, et al. Mucin 15 is expressed 474 in human placenta and suppresses invasion of trophoblast-like cells in vitro. Human 475 reproduction. 2007;22(10):2723-2732. 476 27. Kumar A, D'Souza SS, Moskvin OV, Toh H, Wang B, Zhang J, et al. Specification 477 and Diversification of Pericytes and Smooth Muscle Cells from Mesenchymoangioblasts. 478 Cell reports. 2017;19(9):1902-1916. 479 28. Adamson SL, Lu Y, Whiteley KJ, Holmyard D, Hemberger M, Pfarrer C, et al. 480 Interactions between trophoblast cells and the maternal and fetal circulation in the mouse 481 placenta. Developmental biology. 2002;250(2):358-373. 482 29. Ohlsson R, Falck P, Hellstrom M, Lindahl P, Bostrom H, Franklin G, et al. PDGFB 483 regulates the development of the labyrinthine layer of the mouse fetal placenta. 484 Developmental biology. 1999;212(1):124-136. 485 30. Prefumo F, Sebire NJ, Thilaganathan B. Decreased endovascular trophoblast invasion 486 in first trimester pregnancies with high-resistance uterine artery Doppler indices. Human 487 reproduction. 2004;19(1):206-209. 488 31. Khong TY, De Wolf F, Robertson WB, Brosens I. Inadequate maternal vascular 489 response to placentation in pregnancies complicated by pre-eclampsia and by small-for-490 gestational age infants. Br J Obstet Gynaecol. 1986;93(10):1049-1059. 491 32. Brosens IA, Robertson WB, Dixon HG. The role of the spiral arteries in the 492 pathogenesis of preeclampsia. Obstet Gynecol Annu. 1972;1:177-191. 493 33. Wang Z, Wang P, Liu H, He X, Zhang J, Yan H, et al. Maternal adiposity as an 494 independent risk factor for pre-eclampsia: a meta-analysis of prospective cohort studies. 495 Obesity reviews : an official journal of the International Association for the Study of Obesity. 496 2013;14(6):508-521. 497 34. Allen VM, Joseph K, Murphy KE, Magee LA, Ohlsson A. The effect of hypertensive 498 disorders in pregnancy on small for gestational age and stillbirth: a population based study. 499 BMC pregnancy and childbirth. 2004;4(1):17. 500 35. Gibbs GM, Roelants K, O'Bryan MK. The CAP superfamily: cysteine-rich secretory 501 proteins, antigen 5, and pathogenesis-related 1 proteins--roles in reproduction, cancer, and 502 immune defense. Endocrine reviews. 2008;29(7):865-897. 503 36. Falak S, Schafer S, Baud A, Hummel O, Schulz H, Gauguier D, et al. Protease 504 inhibitor 15, a candidate gene for abdominal aortic internal elastic lamina ruptures in the rat. 505 Physiological genomics. 2014;46(12):418-428. 506 37. Luo W, Liu Z, Tan D, Zhang Q, Peng H, Wang Y, et al. Gamma-amino butyric acid 507 and the A-type receptor suppress decidualization of mouse uterine stromal cells by down-508 regulating cyclin D3. Molecular reproduction and development. 2013;80(1):59-69. 509 38. Lu J, Zhang Q, Tan D, Luo W, Zhao H, Ma J, et al. GABA A receptor pi subunit 510 promotes apoptosis of HTR-8/SVneo trophoblastic cells: Implications in preeclampsia. 511 International journal of molecular medicine. 2016;38(1):105-112. 512 39. D'Angelo MA, Gomez-Cavazos JS, Mei A, Lackner DH, Hetzer MW. A change in 513 nuclear pore complex composition regulates cell differentiation. Developmental cell. 514 2012;22(2):446-458. 515 40. Pigoni M, Wanngren J, Kuhn PH, Munro KM, Gunnersen JM, Takeshima H, et al. 516 Seizure protein 6 and its homolog seizure 6-like protein are physiological substrates of 517 BACE1 in neurons. Mol Neurodegener. 2016;11(1):67. 518 41. Buhimschi IA, Nayeri UA, Zhao G, Shook LL, Pensalfini A, Funai EF, et al. Protein 519 misfolding, congophilia, oligomerization, and defective amyloid processing in preeclampsia. 520 Science translational medicine. 2014;6(245):245ra292. 521 42. Perez-Garcia V, Fineberg E, Wilson R, Murray A, Mazzeo CI, Tudor C, et al. 522 Placentation defects are highly prevalent in embryonic lethal mouse mutants. Nature. 523 2018;555(7697):463-468. 524 43. Knox K, Baker JC. Genomic evolution of the placenta using co-option and 525 duplication and divergence. Genome research. 2008;18(5):695-705. 526 44. Scott IC, Anson-Cartwright L, Riley P, Reda D, Cross JC. The HAND1 basic helix-527 loop-helix transcription factor regulates trophoblast differentiation via multiple mechanisms. 528 Molecular and cellular biology. 2000;20(2):530-541. 529 45. Simmons DG, Fortier AL, Cross JC. Diverse subtypes and developmental origins of 530 trophoblast giant cells in the mouse placenta. Developmental biology. 2007;304(2):567-578. 531 46. Dubova EA, Pavlov KA, Borovkova EI, Bayramova MA, Makarov IO, Shchegolev 532 AI. Vascular endothelial growth factor and its receptors in the placenta of pregnant women 533 with obesity. Bulletin of experimental biology and medicine. 2011;151(2):253-258. 534 47. Hung TH, Chen SF, Lo LM, Li MJ, Yeh YL, Hsieh TT. Increased autophagy in 535 placentas of intrauterine growth-restricted pregnancies. PloS one. 2012;7(7):e40957. 536 48. Belkacemi L, Kjos S, Nelson DM, Desai M, Ross MG. Reduced apoptosis in term 537 placentas from gestational diabetic pregnancies. Journal of developmental origins of health 538 and disease. 2013;4(3):256-265. 539 49. Kenchegowda D, Natale B, Lemus MA, Natale DR, Fisher SA. Inactivation of 540 maternal Hif-1alpha at mid-pregnancy causes placental defects and deficits in oxygen 541 delivery to the fetal organs under hypoxic stress. Developmental biology. 2017;422(2):171-542 185. 543 50. Kim DW, Young SL, Grattan DR, Jasoni CL. Obesity during pregnancy disrupts 544 placental morphology, cell proliferation, and inflammation in a sex-specific manner across 545 gestation in the mouse. Biology of reproduction. 2014;90(6):130. 546 51. Georgiades P, Ferguson-Smith AC, Burton GJ. Comparative developmental anatomy 547 of the murine and human definitive placentae. Placenta. 2002;23(1):3-19. 548 52. Coan PM, Ferguson-Smith AC, Burton GJ. Developmental dynamics of the definitive 549 mouse placenta assessed by stereology. Biology of reproduction. 2004;70(6):1806-1813. 550 53. Eriksson JG, Kajantie E, Osmond C, Thornburg K, Barker DJ. Boys live dangerously 551 in the womb. Am J Hum Biol. 2010;22(3):330-335. 552 54. Ishikawa H, Rattigan A, Fundele R, Burgoyne PS. Effects of sex chromosome dosage 553 on placental size in mice. Biology of reproduction. 2003;69(2):483-488. 554 55. Sood R, Zehnder JL, Druzin ML, Brown PO. Gene expression patterns in human 555 placenta. Proceedings of the National Academy of Sciences of the United States of America. 556 2006;103(14):5478-5483. 557 56. Dearden L, Bouret SG, Ozanne SE. Sex and gender differences in developmental 558 programming of metabolism. Molecular metabolism. 2018;15:8-19. 559 560 Figure Legends 561 Figure 1. RNA-seq identification of differentially expressed genes between Control (C, n=2) 562 and Obese (O, n=3) male mouse placentae at E19. (A) Volcano plot representing all detected 563 transcripts, distributed according to -log10 P-value in the y-axis and log2 Fold Change in the 564 x-axis, with downregulated genes shifted to the left (P< 0.05 in blue) and upregulated genes 565 shifted to the right (P< 0.05 in red). Significantly altered genes after correction for multiple 566 testing (FDR < 0.05) are depicted with a pink diamond; (B) Heatmap representation of genes 567 significantly regulated by maternal obesity using a cutoff FDR < 0.05, with scaled Z-score 568 color key of normalized counts showing expression levels ranging from blue (lower) to red 569 (higher). Genes are sorted from lowest to highest log2 Fold Change value; (C) Top Diseases 570 and Bio Functions from Ingenuity® Pathway Analysis (IPA) of the RNA-seq data with a 571 threshold of P< 0.05, showing the most significant molecular and cellular functions 572 dysregulated in the placenta by maternal obesity, sorted by P-value. 573 574 Figure 2. Validation of RNA-seq data by qPCR in E19 male and female placentae. qPCR 575 results were normalized to the reference genes Gapdh and Sdha and are expressed as mean ± 576 SEM in arbitrary units relative to Male Controls. * P< 0.05, determined by Student's t test 577 comparing qPCR data of same sex Control vs Obese, n = 9/group. 578 579 Figure 3. qPCR expression in E13 male and female placentae. (A) Validated RNA-seq 580 genes; (B) Hand1, required for trophoblast giant cell (TGC) differentiation; Prl2c2, a marker 581 of spiral artery-associated TGC and canal-associated TGC; Pdgfb, a growth factor that 582 regulates placental labyrinthine layer development. qPCR data were normalized to the 583 reference genes Gapdh and Pmm1. Results are shown as mean ± SEM in arbitrary units 584 relative to Male Control average expression. * denotes Maternal Obesity effect (P< 0.05) and 585 # denotes sex difference (P< 0.05), according to two-way ANOVA, n = 10/group. ª Rnf222 586 and Cnn1 expression levels were low at E13 placentae, with average Cq values above 31 and 587 29, respectively. 588 589 Figure 4. Immunofluorescent staining of targets related to the top three Molecular and 590 Cellular Functions shown in IPA. All analyses were conducted in both male and female 591 placentae of mothers fed either regular chow (C, Control group) or obesogenic diet (Ob, 592 Obese group), at E13 and E19. (A – C) The total number of cells in the placenta decreased 593 between E13 (n = 20) and E19 (n = 19); (D – H) The proportion of Ki67-positive cells across 594 the whole placenta decreased between E13 (n = 20) and E19 (n = 19); (I – M) The size of the 595 labyrinth zone increased between E13 and E19, and was reduced in response to maternal 596 obesity (C n = 10, Ob n = 9, at each time point); (N – P) The proportion of fetal capillaries 597 within the labyrinth zone was decreased by maternal obesogenic diet in females (C n = 10, 598 Ob n = 9). (A, D, I, N) Results are shown as mean ± SEM. Gestational age differences are 599 denoted by *(P< 0.05), **(P< 0.001) or ***(P< 0.0001), and Maternal Obesity effect is 600 indicated by #(P< 0.001), according to three-way ANOVA. § denotes Maternal Obesity 601 effect (P< 0.05), determined by two-way ANOVA analysis of E13 and E19 female placentae 602 only. 603 AC B Relative gene expression 0.00.51.01.52.02.53.03.5 Gabrd Sfrp4 Sez6l Nup210 Cnn1Muc15Pi15 Rf222Acta2 Obese FemalesControl FemalesObese MalesControl Males Relative gene expression Gabrd Sez6l Nup210 Muc15Pi15 Acta2 0.00.51.01.52.02.5A Rf222a Cnn1a PdgfbPrl2c2Hand1 Obese FemalesControl FemalesObese MalesControl MalesRelative gene expression 0.00.51.01.5B Total n cells (x10 )4o E13 E19A Ki67-positive cells (%) E13 E19D Labyrinth zone area (mm )2 E13 E19ObCC ObI Fetal capillaries (%) MaleFemale ObCC ObN Control FemaleControl MaleObese MaleObese Female1 mmB DAPI 1 mmC DAPI500 μm Ki67E 500 μmDAPI/Ki67F 500 μm Ki67G 500 μmDAPI/Ki67HCD31100 μmO CD31100 μmPE13 E191 mm DAPI/CD31J 1 mm DAPI/CD31KC 1 mmL DAPI/CD31 1 mmM DAPI/CD31Ob Control Female Obese Female Table 1. Fetal and Placental weights at E13 and E19 Control Obese P-value Males Females Males Females Maternal Obesity Sex E13 Fetal weight (g) 0.17 ± 0.01 0.16 ± 0.00 0.15 ± 0.01 0.14 ± 0.01 0.003 0.187 Placental weight (mg) 94.0 ± 2.2 88.9 ± 3.5 89.7 ± 2.1 78.7 ± 2.9 0.015 0.008 Fetal: Placental ratio 1.84 ± 0.06 1.85 ± 0.06 1.66 ± 0.08 1.76 ± 0.08 0.079 0.452 E19 Fetal weight (g) 1.23 ± 0.03 1.17 ± 0.03 1.02 ± 0.03 1.03 ± 0.02 < 0.0001 0.339 Placental weight (mg) 93.8 ± 5.1 81.7 ± 5.3 82.8 ± 1.9 73.1 ± 2.7 0.039 0.021 Fetal: Placental ratio 13.45 ± 0.76 14.77 ± 0.92 12.38 ± 0.37 14.28 ± 0.78 0.331 0.051 Values are mean ± SEM. P-values < 0.05 indicated in bold show significant effect of maternal obesity and sex differences in the studied parameters according to two-way ANOVA followed by Tukey's multiple comparisons test, using each litter’s average as a single data (Control Male n = 6 and 9, Control Female n = 6 and 9, Obese Male n = 7 and 7, Obese Female n = 7 and 6, respectively at E13 and E19).