Transcription Profiles of the Ductus Arteriosus in Brown-Norway Rats

Advance Publication by-J-STAGE Circulation Journal Official Journal of the Japanese Circulation Society http://www.j-circ.or.jp Transcription Profiles of the Ductus Arteriosus in Brown-Norway Rats With Irregular Elastic Fiber Formation Yi-Ting Hsieh, BSc; Norika Mengchia Liu, BSc; Eriko Ohmori, BSc; Tomohiro Yokota, PhD; Ichige Kajimura, MD; Toru Akaike, MD, PhD; Toshio Ohshima, MD, PhD; Nobuhito Goda, MD, PhD; Susumu Minamisawa, MD, PhD

Background: Patent ductus arteriosus (PDA) is one of the most common congenital cardiovascular defects in children. The Brown-Norway (BN) inbred rat presents a higher frequency of PDA. A previous study reported that 2 different quantitative trait loci on chromosomes 8 and 9 were significantly linked to PDA in this strain. Nevertheless, the genetic or molecular mechanisms underlying PDA phenotypes in BN rats have not been fully investigated yet.

Methods and Results: It was found that the elastic fibers were abundant in the subendothelial area but scarce in the media even in the closed ductus arteriosus (DA) of full-term BN neonates. DNA microarray analysis identified 52 upregulated genes (fold difference >2.5) and 23 downregulated genes (fold difference <0.4) when compared with those of F344 control neonates. Among these genes, 8 (Tbx20, Scn3b, Stac, Sphkap, Trpm8, Rup2, Slc37a2, and RGD1561216) are located in chromosomes 8 and 9. Interestingly, it was also suggested that the significant decrease in the expression levels of the PGE2-specfic receptor, EP4, plays a critical role in elastogenesis in the DA.

Conclusions: BN rats exhibited dysregulation of elastogenesis in the DA. DNA microarray analysis identified the candidate genes including EP4 involved in the DNA phenotype. Further investigation of these newly identified genes will hopefully clarify the molecular mechanisms underlying the irregular formation of elastic fibers in PDA.

Key Words: Congenital heart disease; DNA microarray; Elastic fiber; Patent ductus arteriosus; Vascular remodeling

he ductus arteriosus (DA) between the pulmonary ar- as patent ductus arteriosus (PDA) and is frequently observed tery (PA) and the descending aorta is essential for in preterm infants.4,5 In the case of premature infants, devel- T maintaining fetal circulation. The DA begins to close opmental prematurity of the DA is an important determinant shortly after birth via a combination of 2 events: functional and of PDA. In contrast, PDA that occurs in term infants is usu- anatomical closure. Functional closure happens within the first ally associated with a structural abnormality. The prevalence few hours after birth. Subsequent to functional closure is ana- of PDA in term infants is approximately 2–8 per 10,000 live tomical closure, which refers to the structural remodeling and births, and it represents 5–7% of all cases of congenital heart fibrosis of the DA leading to permanent closure. Progressive diseases in term infants.5 In some of these cases, a genetic intimal thickening that appears prominently at late gestation background underlies an impaired structural remodeling of the represents the structural remodeling of the DA. This physio- DA; the intimal cushion formation is abnormal or absent, and logical intimal thickening is characterized by: (a) the detach- the endothelium remains attached to the IEL or to an addi- ment of the endothelium from the internal elastic lamina (IEL); tional subendothelial elastic lamina; additionally, invagination (b) the fragmentation of the IEL and loss of elastic fibers in the of the endothelium or migration of SMCs from the media are medial layer; (c) the deposition of an extracellular matrix not observed.6,7 (ECM) in the subendothelial area; and (d) the migration of To understand the molecular mechanisms underlying PDA smooth muscle cells (SMCs) into the subendothelial space.1–3 associated with impaired structural remodeling, researchers This series of events contributes to the reorganization and have made use of animal models as a powerful experimental permanent closure of the DA. A failure of DA closure is known tool. The Brown-Norway (BN) inbred rat strain is characterized

Received August 19, 2013; revised manuscript received January 25, 2014; accepted January 27, 2014; released online March 19, 2014 Time for primary review: 21 days Department of Life Science and Medical Bioscience, Waseda University, Tokyo (Y.-T.H., N.M.L., E.O., T.Y., T.O., N.G., S.M.); Depart- ment of Cell Physiology, Jikei University, Tokyo (I.K., T.A., S.M.), Japan Mailing address: Susumu Minamisawa, MD, PhD, Department of Cell Physiology, Jikei University School of Medicine, 3-25-8 Nishishimbashi, Minato-ku, Tokyo 105-8461, Japan. E-mail: [email protected] ISSN-1346-9843 doi: 10.1253/circj.CJ-13-1029 All rights are reserved to the Japanese Circulation Society. For permissions, please e-mail: [email protected] Advance Publication by-J-STAGE HSIEH YT et al.

Figure 1. Dysregulated elastic fiber formation in the ductus arteriosus (DA) of Brown-Norway (BN) rats. To determine the boundary line of intimal cushion formation (ICF, 2-way arrow) and internal elastic lamina (IEL, arrowhead), tissue sections of the DA 1 h after birth were stained with Elastica van Gieson. (a and b) In the comparable longitudinal sections of the DA, the lumen of the DA is almost closed with ICF and the fragmented IEL in a F344 neonate (a), whereas the DA was widely open with minimal ICF and thick IEL in a BN neonate (b). (c and d) In the comparable oblique sections of the DA, the lumen of the DA is closing in both F344 (c) and BN neonates (d). The F344 DA exhibits the fragmented IEL and apparent ICF (c), whereas the DA of the BN neonate has thick IEL and less ICF (d). (e and f) In the comparable cross-sections of the DA, the lumen of the DA is closed in both F344 (e) and BN (f) neonates. The F344 DA exhibits the fragmented IEL (e), whereas the DA of BN neonate has thick IEL and the scarcer elastic lamellae in the media (f). Scale bar: 200 μm.

as a novel animal model of PDA.8,9 Bokenkamp et al. demonstrated that the elastin lamellae were virtually absent in the Methods media but had accumulated in the intima of the DA in the BN Animals rat, resulting in the formation of a subendothelial elastic lamina Timed-pregnant BN and Fisher 344 (F344) rats were pur- and an inhibition of intimal SMC migration.8 These histologi- chased from Japan SLC, Inc. (Shizuoka, Japan). F344 rats are cal features were also found in human and canine PDA.6,7 The commonly used as the control strain when compared with BN BN rat also develops several elastin-related arterial impair- rats. After pregnant rats were anesthetized with tribromoetha- ments such as ruptures of the IEL in the abdominal aorta, and nol (Avertin®) on the 21st day of gestation (full-term), neonates aortic elastin deficit in adults.10 Therefore, these findings sug- were delivered by cesarean section and confirmed to be breath- gest that the inbred BN rat strain exhibits systemic elastin-re- ing. The average number of offspring from pregnant BN rats lated impairments that might cause PDA. The genetic or mo- was 3~4 per mother, which was lower than the approximately lecular mechanisms underlying elastin impairment and PDA 10 neonates obtained from a F344 pregnant rat. All animals phenotypes in this rat strain have not yet been fully investi- were cared for in compliance with the American Physiological gated, although a study of a genome-wide scan with linkage Society. The experiments were approved by the Ethical Com- analysis in BN rats reported that 2 different quantitative trait mittee on Animal Experiments of Waseda University. loci (QTL) on chromosomes 8 and 9 were significantly linked to PDA in this strain.10 Furthermore, our recent findings dem- Histological Analyses onstrated that prostaglanding E2 (PGE2) and its receptor type After delivery, the neonates were placed into a 37°C incubator 4 (EP4, or Ptger4) play a critical role in impaired elastogenesis for 1 h. After 1 h the neonates were decapitated and incised of the DA via degradation of lysyl oxidase (Lox), a key en- through the sternum. The thoracic cavity containing the aorta, zyme that catalyzes elastin cross-links in DA SMCs, and that PA and DA was dehydrated and embedded in paraffin (BN: EP4 knockout mice displayed the arterialized PDA,11 making n=6; F344: n=6). The paraffin-embedded blocks were cut into EP4 as a possible candidate for the PDA phenotype of BN rats. 6-μm-thick sections and placed on glass slides. To determine In the present study, we investigated the transcription profiles the boundary line of intimal cushion formation, tissue sections of the BN rat DA at birth to understand the genetic and/or were stained with Elastica van Gieson, as recommended by the molecular mechanisms underlying the structural abnormali- manufacturer (Muto Pure Chemicals). To determine the extent ties, especially the irregular elastic fiber formation. of EP4 expression in DA tissues, the tissue sections were incubated with anti-EP4 antibody (1:200 dilution; #LS-A3890, Advance Publication by-J-STAGE Transcription Profiles of the DA in BN Rats

Table 1. Upregulated Genes in the Ductus Arteriosus (DA) of Brown-Norway (BN) Rats Gene BN/F344 DA/Aorta Chromosome Probe Set ID mRNA-Description Symbol DA Aorta F344 Location Aorta-Dominant Group 10725778 Nuclear protein, transcriptional regulator, 1 Nupr1 7.39 1.86 0.15 1q36 10829046 Glucagon-like peptide 1 receptor Glp1r 4.87 1.11 0.34 20p12 10909328 Sodium channel, voltage-gated, type III, beta Scn3b 4.73 1.01 0.3 8q22 10880627 Grainyhead-like 3 (Drosophila) Grhl3 4.24 1.12 0.18 5q36 10820613 Synaptic vesicle glycoprotein 2c Sv2c 4.17 1.71 0.31 2q12 10802416 Cell death-inducing DFFA-like effector a Cidea 3.99 1.44 0.29 18q12 10813563 Similar to CG10225-PA Ranbp3l 3.62 1.1 0.14 2q16 10886806 Delta-like 1 homolog (Drosophila) Dlk1 3.33 1.36 0.37 6q32 10929458 SPHK1 interactor, AKAP domain containing Sphkap 3.17 0.86 0.37 9q34 10940654 Secreted phosphoprotein 1 Spp1 3.15 1.38 0.2 14p22 10815369 Periostin, osteoblast specific factor Postn 3.14 1.22 0.27 2q26 10938635 solute carrier family 7 (Cationic amino acid transporter, y+ system), Slc7a3 3.13 0.99 0.16 Xq31 member 3 10920588 Similar to protein Stac Stac 2.99 1.52 0.37 8q32 10817881 Calsequestrin 2 (cardiac muscle), nuclear gene encoding mito- Casq2 2.92 1.21 0.29 2q34 chondrial protein 10842657 GNAS complex locus, transcript variant 4 Gnas 2.92 1.01 0.13 3q42 10891548 Similar to echinoderm microtubule associated protein like 5 isoform 2 Eml5 2.58 1.35 0.27 6q32 10745408 Coronin 6 Coro6 2.57 1 0.32 10q26 10750505 Purkinje cell protein 4 Pcp4 2.54 2.46 0.26 11q12 Equivalent Group 10893424 Zinc finger protein 347 Zfp347 4.84 2.2 0.57 7q11 10925228 Transient receptor potential cation channel, subfamily M, member 8 Trpm8 4.34 4.73 0.49 9q35 10940473 CD36 antigen (CD36), mRNA Cd36 4.09 1.83 0.64 4q11 10855925 Similar to multimerin 1 Mmrn1 3.86 1.39 0.65 4q24 10708021 Aggrecan Acan 3.67 1.45 1.28 1q31 10876208 Chemokine (C-C motif) ligand 21b Ccl21b 3.66 1.22 0.83 5q22 10871002 Similar to CG31019-PA LOC689771 3.56 5.5 1.47 5q35 10733723 Glutamate receptor, ionotropic, AMPA 1 Gria1 3.51 1.38 0.47 10q22 10915817 T-box 20 Tbx20 3.48 1.14 1.72 8q13 10724895 Lymphatic vessel endothelial hyaluronan receptor 1 Lyve1 3.43 1.12 0.47 1q33 10844331 Lipocalin 2 Lcn2 3.4 2.54 1.39 3p11 10788889 HtrA serine peptidase 4 Htra4 3.4 2.44 0.63 16q124 10784334 Similar to tumor necrosis factor receptor superfamily, member 19 RGD1564996 3.35 1.95 0.84 15p12 10845051 Low density lipoprotein-related protein 1B (deleted in tumors) Lrp1b 3.34 1.78 1.63 3q12 10789670 Ligase IV, DNA, ATP-dependent Lig4 3.27 2.96 0.91 16q12.5 10772986 Peroxisome proliferator-activated receptor gamma, coactivator 1 alpha Ppargc1a 3.23 1.02 1.07 14q11 10711396 Erythroid associated factor Eraf 3.22 2.88 1.46 1q36 10794444 MI0000833 let-7f-1 stem-loop Mirlet7f-1 3.21 1 0.57 17p13 10789477 ADP-ribosylhydrolase like 1 Adprhl1 3.17 1.79 0.8 16q12.5 10892939 DNA primase, p49 subunit Prim1 3.12 2.02 0.71 7q11 10731075 GDNF family receptor alpha 1 Gfra1 3.12 3.88 0.78 1q55 10775968 Albumin Alb 3.11 3.43 1.16 14p22 10880095 Serine incorporator 2 Serinc2 3.1 2.53 1.15 5q36 10889213 Visinin-like 1 Vsnl1 3.03 1.93 1.28 6q14 10855387 GTPase, IMAP family member 4 Gimap4 2.89 1.35 0.5 4q24 10934754 Kelch-like 4 (Drosophila) Klhl4 2.88 1.32 0.64 15q21 10856497 Leucine rich repeat transmembrane neuronal 4 Lrrtm4 2.86 1.38 0.7 4q34 10874887 Cysteine-rich secretory protein LCCL domain containing 1 Crispld1 2.85 2.41 0.59 5q11 10869555 SWI/SNF related, matrix associated, actin dependent regulator of Smarca2 2.84 1.32 0.69 1q51 chromatin, subfamily a, member 2 10938654 Interleukin 2 receptor, gamma Il2rg 2.69 1.05 0.45 Xq31 10878068 Acrosome formation associated factor Afaf 2.68 2.1 0.89 5q33 10765335 Cellular repressor of E1A-stimulated genes 1 Creg1 2.6 1.41 0.6 13q23 10825925 Glutathione S-transferase mu 1 Gstm1 2.57 1.17 0.66 2q34 10827438 Rab geranylgeranyltransferase, beta subunit Rabggtb 2.56 2.28 0.91 2q45 Fifty-two genes were expressed less than 2.5-fold in the DA of BN than in that of the F344 neonates (P<0.05). These 52 genes were then divided into 3 groups: DA-dominant (2.5-fold higher in the DA than in the aorta), aorta-dominant (0.4-fold lower in the DA than in the aorta), and the remaining equivalent groups on basis of the data from F344 neonates. No gene belongs to the DA-dominant group. Advance Publication by-J-STAGE HSIEH YT et al.

Table 2. Downregulated Genes in the Ductus Arteriosus (DA) of the Brown-Norway (BN) Rats Gene BN/F344 DA/Aorta Chromosome Probe Set ID mRNA-Description Symbol DA Aorta F344 Location DA-Dominant Group 10892499 Gamma-2a immunoglobulin heavy chain IgG-2a 0.12 0.60 8.58 6q32 10769086 Similar to pappalysin-2 precursor LOC304903 0.17 1.24 5.36 13q22 10916264 Similar to spermatogenesis associated glutamate (E)-rich protein 4b RGD1561216 0.20 0.68 7.66 8q22 10892509 Immunoglobulin heavy chain (gamma polypeptide) Ighg 0.25 0.76 4.32 6q32 10770117 Gremlin 2, cysteine knot superfamily Grem2 0.30 0.70 5.73 13q24 10909009 3’ non-translated beta-F1-ATPase mRNA-binding protein mRNA Rup2 0.33 1.06 6.27 8q21 10847118 Olfactory receptor 650 Olr650 0.34 1.03 4.45 3q24 10727806 Putative retrovirus-related gag protein mRNA – 0.35 1.05 6.33 – 10830279 Similar to double homeobox 4c RGD1560095 0.35 1.15 6.80 20q12 10916167 Similar to solute carrier family 37 (glycerol-3-phosphate trans- Slc37a2 0.36 0.92 2.72 8q21 porter), member 2 10703308 Similar to serine/threonine-protein kinase MARK2 RGD1561706 0.37 1.17 10.43 1q12 10748763 MI0000968 miR-297 stem-loop Mir297 0.38 1.08 2.71 ? 10750108 Hypothetical protein LOC685437 0.39 0.76 4.55 11q11 10769106 Tenascin N (Tnn) Tnn 0.39 0.82 5.09 13q22 10869877 Similar to interferon alpha 8/6 precursor; IFNa8/6 RGD1565911 0.39 0.80 2.84 5q32 Equivalent group 10908861 Lactate dehydrogenase A Ldha 0.02 0.05 1.79 1q22 10930616 ATP synthase F0 subunit 8, Mitochondrion ATP8 0.22 0.67 2.34 mitochondria 10782187 Integrin, beta-like 1 Itgbl1 0.24 0.74 1.00 15q25 10837138 Integrin, alpha 4 Itga4 0.32 0.31 0.68 3q24 10812922 Similar to ELOVL family member 7, elongation of long-chain fatty Elovl7 0.34 0.62 2.22 2q14 acids gene:ENSRNOG00000010450 10821689 Prostaglandin E receptor 4 (subtype EP4) Ptger4 0.37 0.54 2.41 2q16 10716939 Similar to G protein-coupled receptor 126 Gpr126 0.39 0.59 2.41 1p13 10818708 Coagulation factor III (thromboplastin, tissue factor) F3 0.40 0.53 1.08 2q41 Twenty-three genes were expressed less than 0.4-fold in the DA of BN than in that of the F344 neonates (P<0.05). These 23 genes were then divided into 3 groups: DA-dominant (2.5-fold higher in the DA than in the aorta), aorta-dominant (0.4-fold lower in the DA than in the aorta), and the remaining equivalent groups on the basis of the data from F344 neonates. No gene belongs to the aorta-dominant group.

LifeSpan Biosciences, Inc.) and were analyzed as previously The differences of mRNA expression levels between BN an described.11–13 F344 rats were compared.

RNA Extraction and Quantitative Real-Time PCR DNA Microarray Pooled tissues from the aorta and the DA in neonatal rats (1 h Total RNA was converted to cDNA using an Ambion® Whole after birth) were collected separately. Total RNA was extracted Transcript (WT) Expression Kit for Affymetrix GeneChip® from these pooled tissues using TRIzol reagent (Invitrogen, La WT Expression Arrays. cDNA was then biotin-labeled and Jolla, CA) and then reverse-transcribed to cDNA using a hybridized to a GeneChip Rat Gene 1.0 ST Array (Affymetrix, High-Capacity cDNA Reverse Transcription Kit (Applied Santa Clara, CA). Briefly, a total of 100 ng of total RNA was Biosystems, Foster City, CA, USA), as recommended by the reverse-transcribed to cDNA, which was subsequently used as manufacturer. a template for an in vitro transcription reaction. Sense-strand For quantitative real-time-polymerase chain reaction (RT- cDNA containing 2’-deoxyuridine 5’-triphosphate (dUTP) was PCR) analysis, sequences for PCR primers are listed in Table S1. synthesized by amplified cRNA. The Affymetrix GeneChip® For each amplification, 6μl of cDNA equivalent to 15 ng of WT Terminal Labeling Kit (Affymetrix, Santa Clara, CA, USA) total RNA was mixed with 0.4 μl of each primer (forward and was used to recognize the dUTP and to fragment the cDNA reverse) and 10 μl SYBR® Green PCR Master Mix (Applied with uracil-DNA glycosylase and apurinic/apyrimidinic endo- Biosystems) (for TaqMan® Gene Expression Assay, 1 μl prim- nuclease 1. These fragmented cDNAs were then labeled through er and 10 μl of 2×TaqMan Universal Master Mix) for a final a terminal deoxy-transferase reaction and hybridized to the volume of 20 μl. Triplicate reactions were performed for each Affymetrix GeneChip® Rat Gene 1.0 ST Array (Affymetrix). sample. The reactions were performed in 48-well plates on the The arrays were incubated in a 45°C hybridization oven, at Step One™ Real-Time PCR System (Applied Biosystems). 60 rpm, for 17 h. After incubation, the arrays were washed, For each RT-PCR experiment, an RT negative control was in- stained and scanned using an Affymetrix GeneChip® Scanner. cluded, and no amplification was confirmed in all reactions. The data were analyzed with Affymetrix® GeneChip® Command For data analysis, the mRNA levels of interest were normalized Console® Software (AGCC) version 2.0, and then exported. to rat 18 S ribosomal RNA. The resulting data were shown as The hybridization experiments were performed in duplicate, mean ± SD. One-way ANOVA followed by the Turkey Multiple and the intensities in the aorta and the DA were averaged. If Comparison Test was used for statistical analysis of the results. the ratio of signal intensities of the DA in BN and F344 rats in Advance Publication by-J-STAGE Transcription Profiles of the DA in BN Rats

Table 3. Elastic Fiber Formation and Degradation-Related Genes BN/F344 DA/Aorta Probe Set ID mRNA-Description Gene Symbol DA Aorta BN F344 10757726 Elastin Eln 0.94 0.94 0.84 0.84 10849327 Fibrillin 1 Fbn1 1.09 1.06 0.71 0.69 10940545 Fibrillin 2 Fbn2 1.10 1.35 0.93 1.15 10898315 Fibulin 1 Fibl1 0.58 0.70 1.04 1.24 10857278 Similar to Fibulin-2 precursor isoform 2 Fibl2 1.25 0.92 0.86 0.63 10774638 Fibulin 3 Fibl3 1.32 0.94 0.98 0.70 10713061 Fibulin 4 Fibl4 0.78 0.95 0.87 1.06 10891780 Fibulin 5 Fbln5 1.04 1.03 0.95 0.95 10804463 Lysyl oxidase Lox 1.23 1.04 0.86 0.73 10917883 Lysyl oxidase-like 1 Loxl1 0.94 0.88 0.55 0.51 10781304 Lysyl oxidase-like 2 Loxl2 0.98 1.05 0.80 0.86 10856545 Lysyl oxidase-like 3 Loxl3 0.90 0.92 0.75 0.76 10715416 Lysyl oxidase-like 4 Loxl4 0.83 0.92 0.99 1.09 10888424 Latent transforming growth factor beta binding protein 1 Ltbp1 0.88 1.17 1.17 1.54 10891165 Latent transforming growth factor beta binding protein 2 Ltbp2 1.03 0.82 0.60 0.48 10713123 134 kDa protein Ltbp3 0.83 0.94 0.72 0.82 10720000 Latent transforming growth factor beta binding protein 4 Ltbp4 1.18 1.09 0.93 0.87 10888953 Elastin microfibril interfacer 1 Emilin1 0.81 0.96 0.86 1.02 10930428 Similar to elastin microfibril interfacer 2 Emilin2 0.93 0.96 0.87 0.90 10851402 Elastin microfibril interfacer 3 Emilin3 0.98 1.10 1.06 1.20 10785063 Bone morphogenetic protein 1 BMP1 1.13 1.01 0.92 0.82 10809540 Matrix metallopeptidase 2 Mmp2 0.61 1.12 0.81 1.49 10907945 Matrix metallopeptidase 7 Mmp7 0.96 0.90 1.04 0.98 10842239 Matrix metallopeptidase 9 Mmp9 1.16 1.01 1.13 0.98 10907904 Matrix metallopeptidase 10 Mmp10 0.81 1.11 0.92 1.26 10907869 Matrix metallopeptidase 12 Mmp12 1.06 1.37 0.59 0.76 Because the formation of elastic fibers and laminae was dysregulated in the ductus arteriosus (DA) of Brown-Norway (BN) rats, the genes involved in elastic fiber formation and degradation are listed in Table 3.

a given gene was equal to or greater than 2.5-fold, that gene National Center for Biotechnology Information (NCBI). was considered to be upregulated in the BN rat. In contrast, a downregulated gene was identified when the ratio was equal to or less than 0.4-fold. The scattered plots of fold changes in Results BN/F344 ranged primarily from 0.4 to 2.5. Therefore, we de- Irregular Elastic Laminae in the DA of BN Rat Neonates fined a 2.5-fold change or a 0.4-fold change as the cut-off for Six BN and six F344 neonates were prepared for histological a robust analysis. The DNA microarray data was submitted to analysis. All neonates survived at least 1 h in a 37°C incubator the Gene Expression Omnibus and approved as accession num- after delivery. The lumen of the DA was completely or almost ber GSE40534. closed with intimal cushion formation in all F344 neonates (Figures 1a,c,e). In the DA of F344 neonates, the endothelial DNA Sequencing cells separated from the IEL, and a widened subendothelial The rat EP4 genome was obtained from the UCSC Genome region was formed. SMCs in the media migrated through the Browser Home (Ptger4: NM_032076) based on UniProt, Ref- fragmented IEL into the subendothelial region. The elastic Seq and GeneBank mRNA. The genome size is 11,246 base fibers in the media were sparse and disorganized. These histo- pairs, and there are 3 exons. The coding exons are Exon 2 and logical changes were consistent with our previous findings in 3. Genomic DNA was extracted from the tail of neonatal BN the DA of Wistar rat neonates.14 In contrast, in 6 BN rat neo- rats using a NucleoSpin® Tissue Kit. In order to design the nates 1 h after birth, the DA was widely open in 2 neonates primers for sequencing, the exons were divided into Exon 1-1, (Figure 1b), 2 had PDA on the aortic end with either a closed 2-1, 2-2, 2-3, 3-1 and 3-2. The primer for each part of an exon or narrowed PA end (Figure 1d), and the DA was closed in (forward and reversed) was designed and its efficiency was the remaining 2 (Figure 1f). In the 4 BN rats with PDA, inti- tested (Table S2). The DNA products were amplified by Phu- mal cushion formation was either delayed or had not occurred, sion High-Fidelity DNA Polymerase (Finnzymes, Finland) and migration of SMCs from the media into the subendothe- based on the procedure outlined in the Phusion protocol. The lial region was obscure (Figures 1b,d). In the DA of all 6 BN PCR products were then purified by QIAquick PCR Purifica- neonates, including those with a closed DA, the IEL was tion Kits. The PCR products were applied for DNA sequenc- thicker and the elastic lamellae in the media were scarcer than ing. The results were analyzed using an Applied Biosystems those of the DA in the F344 rats (Figures 1b,d,f). These re- Sequence Scanner (version 1.0). The homology of the sequence sults indicated that irregular elastic fiber formation was a was determined using BLAST, a program supported by the common feature of the DA in the BN rat strain. Advance Publication by-J-STAGE HSIEH YT et al.

Table 4. TGF Superfamily BN/F344 DA/Aorta Probe Set ID mRNA-Description Gene Symbol DA Aorta BN F344 10856948 Transforming growth factor alpha Tgfa 0.89 1.08 0.93 1.13 10705213 Transforming growth factor, beta 1 Tgfb1 1.07 0.88 0.94 0.77 10770577 Transforming growth factor, beta 2 Tgfb2 1.14 0.84 1.12 0.82 10891303 Transforming growth factor, beta 3 Tgfb3 1.10 1.07 0.84 0.82 10766248 Transforming growth factor, beta 4 Tgfb4 0.86 0.95 0.94 1.04 10705213 Transforming growth factor, beta 5 Tgfb5 1.07 0.88 0.94 0.77 10797138 Transforming growth factor, beta induced Tgfbi 0.68 1.03 2.88 4.37 10868923 Transforming growth factor, beta receptor 1 Tgfbr1 1.48 1.23 0.95 0.79 10920745 Transforming growth factor, beta receptor II Tgfbr2 1.15 0.92 0.62 0.50 10771070 Transforming growth factor, beta receptor III Tgfbr3 1.18 1.16 0.73 0.72 10927680 Transforming growth factor, beta receptor associated protein 1 Tgfbrap1 1.14 0.96 0.98 0.83 10930411 TGFB-induced factor homeobox 1 Tgif1 0.91 0.86 1.07 1.00 10840138 Bone morphogenetic protein 2 BMP2 0.78 1.02 1.02 1.33 10775573 Bone morphogenetic protein 3 BMP3 1.86 1.35 1.11 0.81 10782891 Bone morphogenetic protein 4 BMP4 1.19 0.83 2.21 1.54 10911711 Bone morphogenetic protein 5 BMP5 0.71 1.08 0.87 1.33 10797949 Bone morphogenetic protein 6 BMP6 1.28 0.96 1.01 0.76 10852106 Bone morphogenetic protein 7 BMP7 1.54 1.04 1.11 0.75 10790481 Bone morphogenetic protein receptor, type 1A BMPR1A 1.23 1.21 0.96 0.95 10827068 Bone morphogenetic protein receptor, type 1B BMPR1B 1.76 1.02 0.68 0.39 10923687 Bone morphogenetic protein receptor, type 2 BMPR2 1.29 1.09 1.06 0.90 10806981 SMAD1 Smad1 1.05 0.98 1.05 0.98 10802741 SMAD2 Smad2 1.05 1.13 0.87 0.93 10918169 SMAD3 Smad3 1.03 1.02 1.02 1.01 10805165 SMAD4 Smad4 1.10 0.99 0.96 0.86 10797127 SMAD5 Smad5 1.06 0.94 1.01 0.89 10918186 SMAD6 Smad6 1.37 0.85 0.90 0.56 10802734 SMAD family member 7 (Smad7) Smad7 1.08 0.83 0.91 0.70 10815436 SMAD8 Smad8 1.60 1.08 0.88 0.59 10815436 SMAD9 Smad9 1.60 1.08 0.88 0.59 Because the transforming growth factor (TGF) signals regulate elastic fiber formation, the genes involved in the TGF signals are listed in Table 4. DA, ductus arteriosus; BN, Brown-Norway.

Gene Expression Analysis by DNA Microarray mation and degradation to be of great interest (Table 3). Con- In order to reveal the genes involved in these histological sistent with sparse elastic fibers in the DA when compared with changes, we carried out gene expression analysis of the DA the aorta, the expression levels of elastin, fibrillin 1, fibulin-2, and the aorta from full-term BN and F344 neonates. When we fibulin-3, Lox, lysyl oxidase-like 1 (Loxl-1), Loxl-3, and latent compared the gene expression levels of the DA of BN rats to transforming growth factor beta binding protein 2 (Ltbp2) were those of F344 rats, the DNA microarray analysis identified 52 lower in the DA than in the aorta in F344 neonates. Although upregulated genes (fold difference >2.5) and 23 downregu- there was no significant statistical difference in these genes lated genes (fold difference <0.4). In addition, the expression between the 2 strains, the mRNA expression levels of fibulin-2, levels of transcripts between the DA and the aorta were com- fibulin-3, and Lox were increased in the DA of BN neonates pared in BN and F344 rats. On the basis of the comparison when compared to F344 neonates. In contrast, fibulin-1, Ltbp1, from the F344 rats, 52 upregulated or 23 downregulated genes matrix metallopeptidase 2 (Mmp2), and Mmp10, of which were then divided into 3 groups: DA-dominant (2.5-fold high- mRNA expression levels were higher in the DA than in the er in the DA than in the aorta), aorta-dominant (0.4-fold lower aorta in F344 neonates, were lower in the DA of BN than F344 in the DA than in the aorta), and the remaining equivalent neonates. groups. Interestingly, none of the upregulated genes belonged Because TGF, especially TGFβ1, is known to regulate elas- to the DA-dominant group (Table 1). On the contrary, most togenic activities, we listed the TGF superfamily members in of the downregulated genes belonged to the DA-dominant Table 4. TGFβ family members including TGFβ1 did not show group, and none of them belonged to the aorta-dominant group an obvious expression difference in the DA between BN and (Table 2). F344 neonates. The TGFβ-induced transcript (Tgfbi) was the only gene for which the expression level was considerably Genes Involved in Elastic Fiber Formation decreased in the DA of BN rats compared to that of F344 rats As our histological analysis revealed that the formation of in this list (fold change=0.68). elastic fibers and laminae was dysregulated in the DA of the BN rat, we considered the genes involved in elastic fiber for- Advance Publication by-J-STAGE Transcription Profiles of the DA in BN Rats

Table 5. Prostanoid-Related Genes BN/F344 DA/Aorta Probe Set ID mRNA-Description Gene Symbol DA Aorta BN F344 10835817 Prostaglandin-endoperoxide synthase 1 (Cox-1) Ptgs1 1.37 1.05 0.81 0.62 10764551 Prostaglandin-endoperoxide synthase 2 (Cox-2) Ptgs2 0.83 0.84 0.75 0.75 10810280 Prostaglandin E receptor 1 (subtype EP1) Ptger1 0.94 0.99 1.09 1.14 10779553 Prostaglandin E receptor 2 (subtype EP2) Ptger2 0.88 1.29 0.75 1.09 10819905 Prostaglandin E receptor 3 (subtype EP3) Ptger3 0.74 1.13 0.35 0.53 10821689 Prostaglandin E receptor 4 (subtype EP4) Ptger4 0.37 0.54 1.64 2.41 10844223 Prostaglandin E synthase Ptges 0.97 1.20 1.04 1.28 10834005 Prostaglandin F receptor Ptgfr 1.03 1.11 0.79 0.85 10704565 Prostaglandin I2 (prostacyclin) receptor (IP) Ptgir 0.90 1.04 0.93 1.08 10750966 Prostaglandin I2 (prostacyclin) synthase Ptgis 1.10 0.87 0.98 0.77 10900212 Thromboxane A2 receptor Tbxa2r 0.85 0.89 0.85 0.89 Because the prostaglandin E signal plays a critical role in intimal cushion formation, the genes related to prostanoids are listed in Table 5. EP4 is lower in the DA of BN than in F344 rats. Abbreviations as in Table 4.

Figure 2. Immunohistochemical analysis of the EP4 expression in the ductus arteriosus (DA). Arrows indicate positive brown staining by the reaction with anti- EP4 antibody. The signals are clearly observed in the endothelium and the smooth muscle layers in the longitudinal section of the DA from F344 neonates (a,b). In contrast, the EP4 signals are very weak in the cross-section of the DA from BN neonates (c,d). Both (b) and (d) show enlarged images of the rectangle indicated in (a) and (c), respectively. Scale bar: 100 μm.

EP4 was Downregulated in the DA of BN Rats tion of 2q16, because Kota et al. indicated that the phenotype Our previous study showed that EP4 plays an essential role in of thoracic aortic elastic content was linked to 2 loci on chro- DA remodeling by promoting hyaluronan-mediated neointi- mosome 2, although PDA was significantly linked to 2 differ- mal formation at birth14 and by inhibiting Lox-mediated elas- ent QTL on chromosomes 8 and 9.10 After searching its ho- togenesis.11 Furthermore, EP4-deficient mice exhibit a PDA mology using BLAST by NCBI, the identities of the nucleotide phenotype.15,16 Therefore, we considered the expression of compositions of 3 exon parts were found to be 100% between genes associated with prostanoids to also be of great interest the BN rat genome and the EP4 gene in Rattus norvegicus (Table 5). None but EP4 showed a lower expression level in chromosome 2 genome scaffolds (NW_001084796.1). the DA of BN rats compared with that of F344 rats (fold change=0.37). We confirmed the downregulation of EP4 Extracellular Matrix mRNA in BN neonates by quantitative RT-PCR analysis Accumulation of ECM such as hyaluronan is a hallmark of (Figure S1B). In addition, immunohistological analysis revealed intimal thickness in the DA. Therefore, the major ECM genes that the EP4 expression was very weak in the DA of BN rats are listed in Table 6. Tenascin-N (Tnn) transcript, which was when compared with that of F344 rats (Figure 2). predominantly expressed in the DA compared to in the aorta, We then performed sequencing on the EP4 gene to find a was significantly decreased in the DA of BN rats compared to mutation of the gene in the BN rat with a chromosome loca- that of F344 rats (fold change=0.39). On the contrary, aggre- Advance Publication by-J-STAGE HSIEH YT et al.

Table 6. Extracellular Matrix BN/F344 DA/Aorta Probe Set ID mRNA-Description Gene Symbol DA Aorta BN F344 10769106 Tenascin N (Tnn) Tnn 0.39 0.82 2.41 5.09 10764947 Tenascin R Tnr 0.79 1.10 1.04 1.44 10831384 Tenascin XA (Tnxa) Tnxa 0.80 0.97 1.07 1.29 10935890 Biglycan Bgn 0.97 0.97 1.01 1.00 10708021 Aggrecan Acan 3.67 1.45 3.25 1.28 10820282 Versican Vca 1.18 1.13 1.33 1.27 10895075 Decorin Dcn 1.26 1.03 0.92 0.76 10787630 Neurocan Ncan 0.88 1.12 0.95 1.21 10824226 Brevican Bcan 0.86 0.98 0.91 1.05 10793909 Testican Spock1 0.84 1.05 1.04 1.29 10873031 Perlecan Hspg2 0.90 0.86 0.84 0.80 10764050 Fibromodulin Fmod 1.07 1.61 1.45 2.17 10895083 Lumican Lum 0.84 1.12 0.85 1.13 10925859 Laminin A1 LAMA1 0.69 0.58 2.54 2.12 10702096 Laminin A2 LAMA2 1.12 1.15 0.58 0.60 10800237 Laminin A3 LAMA3 1.03 1.06 1.02 1.05 10830328 Laminin A4 LAMA4 1.32 1.32 0.65 0.65 10852270 Laminin A5 LAMA5 0.99 0.89 1.05 0.95 10884080 Laminin B1 LAMB1 1.35 0.98 0.68 0.50 10793140 Laminin B2 LAMB2 0.70 0.85 1.10 1.33 10766809 Laminin B3 LAMB3 0.95 0.95 0.95 0.96 10768668 Laminin C1 LAMC1 0.89 0.97 0.98 1.06 10768642 Laminin C2 LAMC2 1.08 1.10 1.07 1.08 10835420 Laminin C3 LAMC3 0.94 1.09 0.96 1.11 10827231 CCN1 Cyr61 0.77 0.92 1.21 1.43 10717233 CCN2 Ctgf 0.73 0.93 0.92 1.17 10896541 CCN3/Nov Nov 0.94 0.81 1.04 0.89 10928761 Fibronectin Fn1 0.86 0.95 0.96 1.06 10718344 Hyaluronan synthase 1 Has1 0.95 1.00 0.90 0.95 10903825 Hyaluronan synthase 2 Has2 1.68 1.13 1.84 1.24 10807576 Hyaluronan synthase 3 Has3 1.00 1.22 0.91 1.11 Because extracellular matrix (ECM) regulates intimal cushion formation, the genes encoding ECM are listed in Table 6. Abbreviations as in Table 4. can (Acan) was upregulated in the DA of BN rats compared levels in the DA were not different between BN and F344 rats, to that of F344 rats (fold change=3.67). Although hyaluronan is only a putative transcriptional regulator for the EP4 pro- is known to play a critical role in intimal thickness in the DA, moter. The putative transcription factors for the EP4 gene of there were no significant differences in the expression levels the 5’ regulatory region (~ −1,000 bp) are listed in Table S4. of all hyaluronan synthases between BN and F344 rats. In- stead, lymphatic vessel endothelial hyaluronan receptor 1 Quantitative RT-PCR (Lyve1) was upregulated in the DA of BN rats compared to To confirm the DNA microarray analysis, we examined the that of F344 rats (fold change=3.43) (Table 1). We also found expression levels of genes of interest by qRT-PCR. The expres- that integrin beta-like 1 (Itgbl1) and integrin alpha 4 (Itga4) sion levels of the genes examined were essentially similar to were downregulated at less than a 0.4-fold change in the DA the results obtained by DNA microarray analysis (Figure S1). of BN rats (Table 2). These genes are involved in cell-to-cell or cell-matrix adhesion. Discussion Transcription Regulation PDA that occurs in term infants is usually associated with Transcription factors play an important role in determining the structural abnormalities: the intimal cushion is less formed or phenotypic morphology of the developing arteries, including absent, and the endothelium remains attached to the IEL or to the DA, although a limited number of transcription factors an additional subendothelial elastic lamina; additionally, in- such as Tfap2β, Hif2α, and myocardin have been known to act vagination of the endothelium or migration of SMCs from the in a transcriptional network during ductal smooth muscle de- media are not observed.6,7 Bokenkamp et al. demonstrated that velopment.17,18 We therefore investigated the expression of the elastin lamellae were virtually absent in the media but had genes associated with transcription factors involved in forma- accumulated in the intima of the persistent DA in the adoles- tion of the cardiovascular system (Table S3). Among the tran- cent BN rat, which resulted in the formation of a subendothe- scription factors listed in Table S3, Arnt, of which expression lial elastic lamina and the inhibition of intimal SMC migration.8 Advance Publication by-J-STAGE Transcription Profiles of the DA in BN Rats

These histological features were also found in the neonatal DA relation between the 2 DNA microarray data. However, several of BN rats in the present study (Figure 1). Therefore, the BN genes showed the similar increase or decrease in their expres- inbred rat strain is characterized as a suitable animal model of sion levels, suggesting that these are EP4-dependent genes. For PDA with structural abnormality. It should be noted that even example, coagulation factor III (F3) and Itgbl1 are significantly in the closed DA of BN rats, the IEL was thicker and elastic decreased in the DA of BN rats and EP4 knockout mice. fiber formation of the media was poorer than in F344 rats. The transcriptional regulation of the EP4 is another interest- Therefore, a common molecular mechanism should underlie ing issue in the DA of BN rats. It is known that SP-1 is an the irregular formation of elastic fibers and laminae in the DA important transcription factor in the regulation of EP4 expres- of BN rats. Regarding the genes involved in elastic fiber forma- sion in humans20 and rats.21 However, the expression of SP-1 tion, we found that the expression levels of elastin, fibrillin 1, mRNA was lower (fold change=0.59) in the DA than in the fibulin-2, fibulin-3, Lox, Loxl-1, Loxl-3, and Ltbp2 were lower aorta in F344 rats, but it was higher (fold change=2.03) in BN in the DA than in the aorta in F344 neonates. Interestingly, than in F344 rats in the present DNA microarray analysis. among these genes, the expression levels of fibulin-2, fibulin-3, Therefore, it is unlikely that the increase in SP-1 expression is and Lox mRNAs were increased in the DA of BN neonates. In responsible for the downregulation of EP4 in the DA of BN contrast, fibulin-1, Ltbp1, Mmp2, and Mmp10 were lower in rats. Furthermore, there were no putative transcription factors the DA of BN than F344 neonates, although it did not reach for the EP4 gene of which expression were significantly al- statistical differences in the expression levels of elastic fiber- tered in the DA of BN rats. Further investigation is required to related genes between BN and F344 neonates. These changes understand the mechanism of downregulation of the EP4 gene might contribute to the dysregulation of elastogenesis in the in the DA of BN rats. DA of BN rats. In addition, we used the whole DA tissues for The present study identified the genes that are upregulated DNA microarray analysis in the present study. Although total or downregulated in the DA of BN neonates when compared transcripts fell within a normal range, one may assume that with control F344 rats. Importantly, most of the newly identi- elastogenesis or elastolysis were locally dysregulated in the fied genes have not yet been investigated in the DA. Among DA of BN rats on the basis of our histological observations. these genes, some might play an important role in the struc- Because the endothelium can also produce elastin and its orga- tural irregularity of elastic fibers in the DA of BN neonates, nizing factors, examining the transcription profiles of the endo- which would result in a high prevalence of persistent DA. Kota thelium and smooth muscles separately rather than through et al. reported that 2 different QTL on chromosomes 8 and 9 whole DA tissues would be necessary in a future study. were significantly linked to PDA in this strain using a genome- The present DNA array analysis revealed that the expres- wide scan with linkage analysis in BN rats.10 In this regard, sion levels of EP4 mRNA were decreased in the DA of BN Tbx20 (8q13), Scn3b (8q22), Stac (8q32), Sphkap (9q34), and rats by ~40% when compared with F344 rats, which is consis- Trpm8 (9q35) are the upregulated genes, and Rup2 (8q21), tent with the immunohistological analysis. Both PGE2 and EP4 Slc37a2 (8q21), and RGD1561216 (8q22) are the downregu- play a critical role in not only dilating the DA but also in lated genes in the DA of BN neonates located in chromosomes promoting DA-specific vascular remodeling. Mice generated 8 and 9. Although none of them have been known to play a with a deficiency of EP4 showed PDA, resulting in early neo- role in elastogenesis, these genes are of great interest as candi- natal death.15,16 Our previous study demonstrated that EP4- date genes responsible for the PDA phenotype in BN rats. mediated signals aid DA closure by enhancing hyaluronic acid- The present gene expression analysis revealed several can- mediated intimal cushion formation.14 However, it is unlikely didate genes responsible for persistent DA with structural ab- that hyaluronic acid plays a role in the PDA phenotype of BN normalities in BN rats. Because pharmacological treatment for rats, because all hyaluronan synthases, including hyaluronan patients with these types of PDA is not yet available, it is im- synthase type 2, were increased in the DA of BN rats. In addi- portant to identify the molecular target underlying the structural tion to the regulation of intimal thickening, we recently re- abnormalities from the upregulated or downregulated genes ported that PGE2 and EP4 signals play an important role in including EP4 in the present study. Further investigation will impaired elastogenesis through degradation of the Lox pro- clarify the role of these genes in the persistent DA. tein.11 Although we did not examine the expression levels of the Lox protein in BN rats, we found that the expression of Acknowledgments Lox mRNA was increased in the DA of BN rats. This might This work was supported by grants from the Ministry of Education, Cul- contribute to the thicker IEL of the DA in BN rats. Further- ture, Sports, Science and Technology of Japan (T.A., S.M.), the “High- more, Kota et al. demonstrated that the phenotype of thoracic Tech Research Center” Project for Private Universities: MEXT (T.O., aortic elastic content linked to 2 loci on chromosome 2 where N.G., S.M.), MEXT-Supported Program for the Strategic Research Foun- 10 dation at Private Universities (T.O., N.G., S.M.), the Vehicle Racing the EP4 gene is located. Therefore, EP4 and its related genes Commemorative Foundation (S.M.), and the Shimabara Science Promo- are also of great interest. Although there was no mutation or tion Foundation (S.M.). single nucleotid polymorphism in the coding region of the EP4 genome, we found that the PGE2-specific receptor, EP4, was References significantly downregulated in the DA of BN neonates. Fur- 1. Clyman RI. Mechanisms regulating the ductus arteriosus. Biol Neo- ther investigation is necessary to clarify whether the down- nate 2006; 89: 330 – 335. regulation of EP4 mRNA contributes to the cause of PDA in 2. Yokoyama U, Minamisawa S, Ishikawa Y. Regulation of vascular BN rats, because the PDA phenotype is different in BN rats tone and remodeling of the ductus arteriosus. J Smooth Muscle Res 2010; 46: 77 – 87. and EP4 knockout mice. In EP4 knockout mice, the DA al- 3. de Reeder EG, van Munsteren CJ, Poelmann RE, Patterson DF, most exhibits an elastic artery and the elastic fibers develop Gittenberger-de Groot AC. Changes in distribution of elastin and very well in the whole layers, whereas the IEL was thicker but elastin receptor during intimal cushion formation in the ductus arte- the elastic lamellae in the media were scarcer in the DA of BN riosus. Anat Embryol (Berl) 1990; 182: 473 – 480. 4. Schneider DJ, Moore JW. Patent ductus arteriosus. Circulation 2006; rats. In this regard, it is of great interest to compare the data of 114: 1873 – 1882. the present DNA microarray with the published DNA micro- 5. Hajj H, Dagle JM. Genetics of patent ductus arteriosus susceptibility array data from EP4 knockout mice.19 We found no strong cor- and treatment. Semin Perinatol 2012; 36: 98 – 104. Advance Publication by-J-STAGE HSIEH YT et al.

6. Gittenberger-de Groot AC. Persistent ductus arteriosus: Most prob- tor EP4-deficient mice. Biochem Biophys Res Commun 1998; 246: ably a primary congenital malformation. Br Heart J 1977; 39: 610 – 7 – 12. 618. 17. Ivey KN, Sutcliffe D, Richardson J, Clyman RI, Garcia JA, Srivastava 7. Gittenberger-de Groot AC, Strengers JL, Mentink M, Poelmann RE, D. Transcriptional regulation during development of the ductus arte- Patterson DF. Histologic studies on normal and persistent ductus arteriosus. Circ Res 2008; 103: 388 – 395. riosus in the dog. J Am Coll Cardiol 1985; 6: 394 – 404. 18. Huang J, Cheng L, Li J, Chen M, Zhou D, Lu MM, et al. Myocardin 8. Bokenkamp R, Gittenberger-De Groot AC, Van Munsteren CJ, Grauss regulates expression of contractile genes in smooth muscle cells and RW, Ottenkamp J, Deruiter MC. Persistent ductus arteriosus in the is required for closure of the ductus arteriosus in mice. J Clin Invest brown-norway inbred rat strain. Pediatr Res 2006; 60: 407 – 412. 2008; 118: 515 – 525. 9. Bokenkamp R, DeRuiter MC, van Munsteren C, Gittenberger-de 19. Gruzdev A, Nguyen M, Kovarova M, Koller BH. PGE2 through the Groot AC. Insights into the pathogenesis and genetic background of EP4 receptor controls smooth muscle gene expression patterns in the patency of the ductus arteriosus. Neonatology 2010; 98: 6 – 17. ductus arteriosus critical for remodeling at birth. Prostaglandins Other 10. Kota L, Osborne-Pellegrin M, Schulz H, Behmoaras J, Coutard M, Lipid Mediat 2012; 97: 109 – 119. Gong M, et al. Quantitative genetic basis of arterial phenotypes in 20. Kambe A, Iguchi G, Moon Y, Kamitani H, Watanabe T, Eling TE. the brown norway rat. Physiol Genomics 2007; 30: 17 – 25. Regulation of EP4 expression via the SP-1 transcription factor: Inhi- 11. Yokoyama U, Minamisawa S, Shioda A, Ishiwata R, Jin MH, Masuda bition of expression by anti-cancer agents. Biochim Biophys Acta M, et al. Prostaglandin e2 inhibits elastogenesis in the ductus arte- 2008; 1783: 1211 – 1219. riosus via ep4 signaling. Circulation 2014; 129: 487 – 496. 21. Chien EK, Macgregor C. Expression and regulation of the rat pros- 12. Jin MH, Yokoyama U, Sato Y, Shioda A, Jiao Q, Ishikawa Y, et al. taglandin E2 receptor type 4 (EP4) in pregnant cervical tissue. Am J DNA microarray profiling identified a new role of growth hormone Obstet Gynecol 2003; 189: 1501 – 1510. in vascular remodeling of rat ductus arteriosus. J Physiol Sci 2011; 61: 167 – 179. 13. Ichikawa Y, Yokoyama U, Iwamoto M, Oshikawa J, Okumura S, Supplementary Files Sato M, et al. Inhibition of phosphodiesterase type 3 dilates the rat Supplementary File 1 ductus arteriosus without inducing intimal thickening. Circ J 2012; Table S1. PCR primer sequences 76: 2456 – 2464. 14. Yokoyama U, Minamisawa S, Quan H, Ghatak S, Akaike T, Segi- Table S2. Rat ptger4 (EP4) genome exon 1, 2 and 3 primers for DNA Nishida E, et al. Chronic activation of the prostaglandin receptor ep4 sequencing promotes hyaluronan-mediated neointimal formation in the ductus Table S3. Transcription factors arteriosus. J Clin Invest 2006; 116: 3026 – 3034. 15. Nguyen M, Camenisch T, Snouwaert JN, Hicks E, Coffman TM, Table S4. Putative transcription factors Anderson PA, et al. The prostaglandin receptor EP4 triggers remodel- Figure S1. Quantitative RT-PCR analysis on genes of interest. ling of the cardiovascular system at birth. Nature 1997; 390: 78 – 81. 16. Segi E, Sugimoto Y, Yamasaki A, Aze Y, Oida H, Nishimura T, et Please find supplementary file(s); al. Patent ductus arteriosus and neonatal death in prostaglandin recep- http://dx.doi.org/10.1253/circj.CJ-13-1029