Placenta 35 (2014) 737e747

Contents lists available at ScienceDirect

Placenta journal homepage: www.elsevier.com/locate/placenta

Spatial and temporal changes of Decorin, Type I collagen and Fibronectin expression in normal and clone bovine placenta zelin a, O. Sandra a, I. Hue a, D. Le Bourhis a, M. Guillomot a, *, E. Campion a, A. Pre C. Richard b, F.H. Biase c, C. Rabel c, R. Wallace d, H. Lewin c, e, J.-P. Renard a, H. Jammes a INRA, UMR1198 Biologie du D eveloppement et Reproduction, F-78350 Jouy-en-Josas, France INRA, UE1298 Unit e Commune d'Exp erimentation Animale, F-91630 Leudeville, France c Institute for Genomic Biology, Department of Animal Sciences, University of Illinois at Urbana-Champaign, Urbana, IL 61820, USA d College of Veterinary Medicine, University of Illinois at Urbana-Champaign, USA e Department of Evolution and Ecology and The Genome Center, University of California, Davis, CA 95616, USA a

b

a r t i c l e i n f o

a b s t r a c t

Article history: Accepted 13 June 2014

Introduction: Alteration of expression of various genes including extracellular matrix components, have been suggested to play major role in the placental pathologies after somatic cloning in mammals. The objectives of the present study were to analyze pattern of expression (mRNA and protein) of the small leucine-rich proteoglycan, Decorin in association with Type I Collagen and Fibronectin in bovine placental tissues from normal and clone pregnancies. Methods: Genotyping and allelic expression of Decorin were determined by Sanger sequencing. The expression patterns of Decorin, Type I collagen and Fibronectin 1 were analyzed by quantitative RT-qPCR and combined in situ hybydization (ISH) and immunohistochemistry (IHC) in endometrial and placental tissues from D18 to term from artificially inseminated and somatic cloning pregnancies. Results: The expression levels of DCN increased in the AI endometrial stroma and chorionic mesenchyme during implantation and declined during placentome growth until term. Combined ISH and IHC revealed an unexpected discrepancy mRNA and protein tissue distribution. Moreover, Decorin was maintained in the placentome tissues from SCNT pregnancies while both mRNA and protein were absent in AI derived placenta. Discussion: In bovine, the pattern of expression of Decorin exhibits significant changes during placental formation. Downregulation of Decorin is associated with proliferation, remodeling and vascularization of placental tissues. These observations reinforces the putative role of Decorin in these processes. Conclusions: These observations suggest that Decorin is involved in placental growth and that dysregulation of its expression is associated with placental abnormalities in SCNT derived pregnancy. © 2014 Elsevier Ltd. All rights reserved.

Keywords: Extracellular matrix Cloning Imprinting Placenta Ruminants

1. Introduction Placental morphogenesis and development require coordinated processes of proliferation and differentiation of the endometrium and chorionic membranes. In ruminants, the maternal part of the cotyledonary placenta originates from the pre formed endometrial caruncles whose size and shape change dramatically during pregnancy. On the fetal side, chorionic villi facing the uterine caruncles develop and invade the mazy endometrial caruncular crypts giving rise to placentomes dispersed along the uterine horns and the

* Corresponding author. E-mail address: [email protected] (M. Guillomot). http://dx.doi.org/10.1016/j.placenta.2014.06.366 0143-4004/© 2014 Elsevier Ltd. All rights reserved.

chorionic sac. In cow, progressive establishment of the placentomes starts after implantation (day 20) and lasts until around day 60 of pregnancy. Thereafter, the placentomes increase in size and weight throughout pregnancy until the third trimester. The cloning of mammals by somatic cell nuclear transfer (SCNT) has been most successful in cattle. However, the rate of viable offspring is still very low (6e15% of transferred embryos) when compared with embryos produced by artificial insemination (AI) or in vitro fertilization (IVF) (45e60%) [1]. A high frequency of gestational losses occurs during the whole course of pregnancy [2]. Since first SCNT experiments, placental deficiencies have been depicted as one of the major cause of fetal losses during the mid- to late pregnancy in several species, including mouse [3,4] pigs [5] sheep [6] and cow [1,7]. In cattle, main placental defaults observed in cloned pregnancies are a delay

738

M. Guillomot et al. / Placenta 35 (2014) 737e747

of placentome formation around days 35e60 and enlarged edematous placentomes associated with hydrallantois and fetal overgrowth, known as the large offspring syndrome (LOS), during the second and third trimester [8,9]. Despite a grossly normal histologic appearance, placental development and structures are modified. Stereological analysis of placentome sections show that the proportion of fetal mesenchymal compartment increased and maternal epithelial height decreased [8]. Causes of these gestation failures are still largely unknown and might be due to multiple factors. LOS associated with bovine clone pregnancies resembles abnormalities reported with alterations of imprinted gene expression in the human syndrome Beckwith-Wiedemann [10]. The placenta expresses many of these imprinted genes and all these genes have been shown to regulate placental growth and differentiation in mouse mutants [11,12]. Comparisons of the transcriptomes of chorionic, endometrial or placental tissues from bovine AI and IVF vs. SCNT pregnancies have evidenced various differentially expressed sets of genes controlling various cell and tissue functions such as extracellular matrix (ECM) formation [13e16]. The processes of implantation and placentation involved remodeling of the endometrial and chorionic ECM [17e19]. The organization of the ECM is a complex process which involves interactions between the ECM components and extracellular proteoglycans such as members of small leucine-rich proteoglycans family (SLRPs). The interactions of SLRPs with ECM components play important roles during growth and development [20,21]. One of the most studied members of SLRPs is Decorin. The Decorin gene (DCN) belongs to the imprinted gene family in mouse [22] but is not imprinted in human and pig placenta [23,24] nor in bovine tissues [25]. Decorin is expressed and secreted in the ECM of a wide range of connective tissues [21,26] including the placenta in human, bovine, mouse, rat and pig species [24,27e31]. Decorin binds to fibrillar collagens and delays the formation of thinner fibrils [32,33]. In the endometrium of Decorin-deficient mice, the assembly of collagen fibrils is disturbed, resulting in thickening of collagen matrix [34]. In mouse endometrium, Decorin and associated SLRPs are finely regulated by ovarian steroids during the estrous cycle and pregnancy [35-37]. Besides its primary role in collagen assembly, Decorin interacts with various factors to control cell proliferation or migration [38] and vasculogenesis [39] and thus, is now recognized as a potent anticancer agent [40,41]. The possibility that Decorin could be associated with placental development in AI and SCNT pregnancies in cattle, has initiated the present study. Thus, we have analyzed the temporal and spatial expression (mRNA and protein) of Decorin in parallel with two associated ECM components, Type I collagen and Fibronectin, in endometrium and placenta obtained from AI and SCNT pregnancies and we have performed a new analysis of DCN imprinting in all extraembryonic tissues using other SNP than used by Khatib [25].

uterine horn. The embryos and fetuses were dissected from the extra-embryonic membranes and measured to evaluate their status of development. Endometrial caruncles and intercaruncular endometrium were sampled separately. The placentomes from D62 to term, were dissected away from the uterine mucosa. Pregnant cows at term (n ¼ 3) were delivered by caesarean section (C-section) 24 h after induction with dexamethasone and one placentome was collected. The chorionic villi of some placentomes were separated from the uterine caruncles. At all stages studied, all tissues samples were either fixed in 4% paraformaldehyde and embedded in paraffin or snap frozen in liquid nitrogen and stored at 80
C until further processing.

2. Materials and methods

2.6. Analysis of DCN imprinting in placental tissues

2.1. Animals

Genomic DNA extraction from 38 different bovine tissues was performed using a standard proteinase K/RNase/phenol/chloroform extraction protocol. Exonic single nucleotide polymorphisms (SNPs) were chosen based on their validation from the Ensembl Genome Browser (DCN-201 ENSBTAT00000004562). Genotyping was performed on two SNPs rs210096509 and rs109465157, located in exon 7 on chromosome 5. Primers were designed to amplify fragment spanning more than exon 7 (419 bp; Supplementary Table 3). Individuals with a heterozygosity were selected for the cDNA genotyping. Total RNA from fetal tissues (brain, heart, liver and intestine), extraembryonic tissues (cotyledon, extracotyledonary chorion and amnion) and maternal endometrial caruncles, collected at D62 of gestation, were extracted as described above. The primers were designed to cross introneexon boundaries to exclude the possibility of mistyping as a result of gDNA contamination in the RT-PCR reactions (PCR product of 231 bp; Supplementary Table 3). RT-PCR was carried out on individual cDNA samples. Amplifications of gDNA and cDNA were performed using 100 ng of gDNA or 200 ng cDNA, 1X Taq buffer, 2 mM MgCl2, 0.2 mM final of each primer, 0.2 mM of dNTPs, 2.5 U of Taq Platinum (Invitrogen, Cergy-Pontoise, France) in 50 ml of final volume. The PCR amplification was subjected to 35 cycles

Conceptuses from day 18 to term from AI pregnancies and all NT conceptuses were produced at the experimental farm of Bressonvilliers, Institut National de la Recherche Agronomique (INRA), France. All experimental procedures were in accordance with the rules and regulations of the INRA ethical committee (COMETHEA Jouy-en-Josas n
12/122). Additional gestations producing AI-derived conceptuses at day 34 were established and processed at the University of Illinois at UrbanaeChampaign, USA with the approval of the Institutional Animal Care and Use Committee. Total number of animals used per stage is given in Supplementary Table 1. 2.2. Production of AI pregnancies Estrous synchronized Holstein cows were inseminated artificially with frozen Holstein bull semen at day 0 (D0) then slaughtered at known days of pregnancy. To collect D18 conceptuses, the uterine horns were flushed with PBS. By D34 of pregnancy, placental tissues were collected by incision through the wall of the ipsi-lateral

2.3. Production of SCNT pregnancies The cloned embryos were produced following a previously published protocol [42]. Pregnancy diagnosis was performed first at D21 by circulating progesterone assay, followed ultrasonography according to routine protocol in our laboratory [8]. All SCNT recipients positively diagnosed for onset of hydrallantois from D150 to 270 (n ¼ 11) were slaughtered (n ¼ 7) or delivered by C-sections (n ¼ 4) as described above. All SCNT recipients at term (n ¼ 12) were delivered by C-section as described above. After slaughter or C-section, all tissues were collected and processed as described above for AI pregnancies. 2.4. Total RNA extraction Total RNA were isolated from tissue samples with the use of Trizol® Reagent (Invitrogen, Cergy-Pontoise, France) according to the manufacturer's instructions. 2.5. Gene expression quantification Due to putative cross contamination by chorionic membranes within uterine caruncles, expression levels in the uterine compartment were not considered after D62 of pregnancy and only the fetal part of placentomes were used for quantitative analysis. RNA messenger levels were determined for the genes of interest by reverse transcription followed by real-time quantitative PCR (RT-qPCR). First-strand cDNAs were synthesized from 2 mg of total RNA (treated by DNAse RNAse-free, 20 ng/ml, 37
C, 15 min) in the presence of oligo-(dT) 12e18, 400 nM dNTPs and 200 U Superscript™ II Rnase H Reverse Transcriptase, according to the manufacturer's instruction (Invitrogen, Cergy-Pontoise, France). All primer sets were designed using Primer Express software Primer3 v. 0.4.0, encompassing two or three exons to eliminate any amplification from genomic DNA contamination (Supplementary Table 2). Decorin (DCN) and Biglycan (BGN) gene exhibit a high sequence identity (60%) which results from the duplication of an ancestral gene. Thus, the primers used for DCN and BGN expression analysis were designed to eliminate any risk of cross reactivity. Under these conditions, BGN expression was found dramatically low in all analyzed tissues (data not shown). COL1A1 and COL1A2 genes encode the alpha-1 and alpha-2 chains of type I collagen, respectively and FN1 gene encodes type 1 Fibronectin protein. Real-time-qPCR analyses were carried out with the Mesa Blue qPCR mix according to the manufacturer's instructions (Eurogentec, Angers, France), using an ABI PRISM 7300 apparatus (Applied Biosystems®, Life Technology, Saint-Aubin, France). Each reaction was carried out in a final volume of 15 ml, in duplicate. For each gene analyzed and each run, a standard curve was designed from a series of 10 successive dilutions of cDNAs to determine primer set efficiency. The presence of a unique and specific PCR product was verified by an ABI Prism generated melting curve profile. Controls lacking reverse transcriptase were performed and consistently yielded no amplification below 40 cycles. 2.5.1. Normalization Six genes, EIF4A2, ACTB, GAPDH, RPLP0, RPL19 and SDHA (primer list in Supplementary Table 2) have been tested as reference genes in GeNorm software [43]. Based on the geometric mean data, the RT-qPCR results were normalized with the normalized factor obtained for RPLP0, RPL19 and SDHA quantifications.

M. Guillomot et al. / Placenta 35 (2014) 737e747 at 94
C for 30 s, 53
C for 30 s and 72
C for 30 s. The sizes of the PCR products were estimated on a 1.5% agarose gel. The PCR products were purified from free nucleotides using Wizard® SV gel and PCR Clean-Up System (Promega). Genotypes and allelic expression were determined by Sanger sequencing (Applied Biosystems, CA). 2.7. In situ hybridization procedure Number of animals used is given in Supplementary Table 1. At least two different tissue samples from the same animal were analyzed in duplicates and 2 to 3 sections per samples were analyzed at the same time. Tissues to be compared either stage to stage or AI vs. SCNT were treated in the same batch of slides. The same sampling protocol was applied for immunohistochemistry. In situ hybridization (ISH) was performed with Digoxygenin-labeled probes prepared from bacterial cloned present in libraries described previously by Everts et al. [15]. The cDNA fragments (Supplementary Table 4) were amplified by PCR using specific couples of primers for cloning vectors pT7T3Pac. Purified cDNA were used as template to synthesize Digoxigenin-UTP (DIG) labeled RNA using RNA T7 or T3 polymerases for sense or anti-sense probes, respectively (Roche Diagnostics, Indianapolis, IN). Serial tissue sections were hybridized following previously published protocols [13,44]. After successive washes in SSC, and RNase treatment, the sections were incubated with sheep anti-DIG-alkaline phosphatase-conjugated antibody (Roche Diagnostics, Indianapolis, IN). Staining was developed in NBT/BCIP solution (Promega, Madison, WI). The staining time was consistent for both sense and antisense probes and for tissues from different stages or status of pregnancy to be compared with each other. Images were captured with a slide scanner (Nanozoomer 2.0-HT, Hamamatsu, Hamamatsu City, Japan) and areas of interest were saved as JPEG file image. 2.8. Immunoperoxidase procedure Paraffin serial tissue sections from the same blocks used for ISH, were treated following previous published immunohistochemical protocol [13]. Tissues sections were incubated with either mouse monoclonal bovine Decorin (DS1, Hybridoma Bank), rabbit polyclonal bovine Type I Collagen or plasma Fibronectin antibodies (Novotec, St Martin La Garenne, France). Incubations with normal mouse or rabbit sera were used as negative controls. Horseradish peroxidase-conjugated goat anti-rabbit or anti-mouse IgG (Jackson ImmunoResearch Inc., Baltimore, PA) were used as secondary antibodies and chromogen detection was performed using 3,30 -diaminobenzidine e H2O2 (SigmaFast™ DAB tablets, SigmaeAldrich, St Louis, MO). On some sections, postISH immunohistochemistry was performed according to the same protocol starting by blocking goat serum incubation after the staining step of the ISH protocol. 2.9. Statistical analysis All statistical analyses were carried out with R package software. Expression data were analyzed by non parametric KruskaleWallis tests. Differences between experimental groups were considered to be significant if p value