Messenger RNA differential display. Complementary DNAs generated by differential display reverse transcription–polymerase chain reaction from nonwounded (N) and wounded (W) skin samples of fetal and adult rabbits 12 hours after incisional wounding were run on a 6% polyacrylamide gel. This section of the gel shows a differentially expressed complementary DNA fragment.
Sequence homology between a clone sequence (352 base pairs) and a rabbit prostaglandin E2 EP4 receptor messenger RNA (3338 base pairs, GenBank accession No. L47207).12 Flanking primer sequences are underlined. Mismatched nucleotides are shown in lowercase. A polyadenylation signal (AATAAA) is shown in boldface.
Reverse RNA dot blot. An equal amount of the EP4 clone complementary DNA (100 ng) from nonwounded (N) and wounded (W) skin of fetal and adult rabbits was hybridized with original differential display reverse transcription–polymerase chain reaction samples labeled with [α-33P]-deoxyadenosine triphosphate ([α-33P]-dATP; specific activity, 1 × 105-6cpm/mL).
Semiquantitative reverse transcription–polymerase chain reaction. One complementary DNA (452 base pairs [bp]; arrow) was differentially amplified by means of gene-specific primers (EP4B) in fetal and adult nonwounded (N) and wounded (N) skin at 30 cycles. NC is a reverse transcription–polymerase chain reaction negative control. The level of 18S ribosomal RNA (rRNA) in each sample was assessed as a control.
Reverse transcription–polymerase chain reaction of the EP4 receptor from the nonwounded adult rabbit skin with the use of gene-specific primers. Three amplified complementary DNAs were size-fractionated on a 1.5% agarose gel (EP4A, 449 base pairs [bp]; EP4B, 458 bp; and EP4C, 447 bp). The complementary DNAs were digested by unique restriction enzymes (SmaI, DraII, and SacII), and DNA fragments with expected sizes were produced.
Li H, Hebda PA, Kelly LA, Ehrlich GD, Whitcomb DC, Dohar JE. Up-regulation of Prostaglandin EP4 Receptor Messenger RNA in Fetal Rabbit Skin Wound. Arch Otolaryngol Head Neck Surg. 2000;126(11):1337-1343. doi:10.1001/archotol.126.11.1337
Scar formation and subglottic stenosis often cause health problems in surgical otolaryngology. However, fetal wounds demonstrate scarless healing. The underlying mechanism remains poorly understood. We isolated differentially expressed genes by comparison between nonwounded with wounded skin of fetal and adult rabbits.
Skin incisional wounds were made in fetal (21 to 23 days' gestation) and adult rabbits. Nonwounded and wounded skin were harvested 12 hours after surgery. Total RNA was extracted. By means of messenger RNA differential display, differentially expressed complementary DNA fragments were isolated, cloned, and sequenced. The expressed transcripts were verified by reverse RNA dot blot and semiquantitative reverse transcription and polymerase chain reaction.
One complementary DNA tag that was induced in fetal skin wounds and repressed in adult skin wounds was isolated. The sequence of this complementary DNA (352 base pairs) encodes the messenger RNA for the E-prostanoid (EP) 4 receptor for prostaglandin E2 (PGE2). The truly differential expression of the transcript was confirmed. In normal skin, the EP4 receptor messenger RNA levels were higher in adults than in fetuses. Twelve hours after wounding, the EP4 receptor transcript was remarkably induced in fetal skin wounds but repressed in adult skin wounds.
Our study demonstrates the differential expression of the EP4 receptor messenger RNA in fetal and adult skin before and 12 hours after wounding. Our results suggest that prostaglandin E2 is involved in the differential cellular responses and in the regulation of the intracellular signal transduction through its binding to EP4 receptor during fetal wound repair.
SCAR FORMATION and subglottic stenosis often cause structural, functional, growth, and cosmetic problems in children.1,2 Previous studies have demonstrated that fetal wounds in skin and airway heal in a regenerative pattern without scar formation.3- 6 Distinct differences between the fetal and adult wound healing response have been found in humans and in animal models.6- 8 A unique aspect of skin and airway mucosal wounds made in the second or early third trimester of gestation, in contrast with adult wounds, is that they heal more rapidly and without scar formation.
Epidermal wound healing occurs by quite different mechanisms in embryos and adults, but both heal essentially by regeneration, ie, original form and functions are restored. Wounding of the connective tissue in the adult elicits bleeding and inflammation, which stimulate subsequent steps of healing, including granulation tissue formation and angiogenesis, fibroplasia, collagen deposition, and remodeling.3 The end results are fibrosis and scar formation. In contrast, wounding of fetal connective tissue, in both skin and airway mucosa, occurs in the absence of inflammation and granulation tissue formation, and results in restoration of the connective tissue, often referred to as regenerative or scarless wound healing.3- 5
The possible mechanisms associated with differential fetal wound healing have been extensively investigated. The fetal environment, sterile and surrounded by amniotic fluid, differs markedly from that of the adult.9 Amniotic fluid is rich in growth and trophic factors, such as insulin-like growth factor II, and in matrix components, such as adhesion molecules (fibronectin and tenascin) and hyaluronic acid, which stimulate cell attachment and facilitate cell migration, leading to more rapid wound epithelialization.9 Fetal tissues and wounds have distinct cytokine profiles.6,10 Fetal wounds have higher levels of collagens III and V and a greater proportion of type III to type I collagen than adult wounds.6 The higher type III collagen level of fetal wounds may allow better approximation of normal tissue architecture in the fetus.
The fetal environment is not the sole determinant of scarless repair. A transition from scarless healing to healing with scar formation correlates with gestational age–related changes in the cytokine response to wounding, the complexity of dermal and subdermal tissue architecture, and the ability of the fetus to generate an inflammatory response. The ability of the fetal fibroblast to induce formation of epidermal appendages (mesenchymal induction) decreases with advancing gestational age. Loss of ability to regenerate hair follicles precedes the development of adult-type scarring.6 Because the onset of scarring is immediately preceded by a decrease in the inductive ability of mesenchyme, it is possible that epithelial-mesenchymal signaling molecules are key factors of scarless fetal healing. Therefore, intrinsic properties of fetal and adult cells, rather than the surrounding environment, determine the outcome of healing.
Moreover, scarless fetal wound healing is not a universal phenomenon, and some fetal organs heal with scar formation (eg, gut and diaphragm).6,11 These findings support the argument that the key to scarless fetal wound healing is the responding cell, namely, the fetal fibroblast. The fetal environment may be a contributing factor to the outcome of wound healing, but the phenotype of the fetal cells is critical. A key issue for understanding the molecular mechanisms of differential wound repair is to identify genes specifically expressed in tissue-, age-, and time-dependent manners during wound healing. Therefore, differential gene expression was used to better understand the mechanism of scarless wound healing and, by systematic comparison, the contrasting mechanism of fibrotic wound repair that leads to scarring.
The purpose of this study is to isolate and clone key molecules that are specifically expressed in fetal wounds. An understanding of gene regulation governing the absence of scarring in the fetus may lead to strategies for controlling both normal and pathologic scarring. In our present study, incisional wounds in rabbit skin were made on the basis of our previous experience.5,12 Tissues were harvested 12 hours after wounding because a peak of the early inflammatory response occurs at about 12 to 24 hours in adult wounds.11 Differential display of messenger RNA (mRNA) was used to isolate differentially expressed transcripts by comparing nonwounded and wounded skin in fetal and adult rabbits.12 One of the complementary DNA (cDNA) fragments preferentially induced in wounded fetal skin was isolated, cloned, and sequenced. Its cDNA sequence encodes a rabbit prostaglandin E2 (PGE2) E-prostanoid (EP) 4 receptor mRNA, suggesting potential roles for PGE2 and this receptor in directing the wound healing response in the fetal skin.
Adult pregnant and nonpregnant Pasteurella-free New Zealand white rabbits were obtained from an approved supplier (Hazelton Research Products, Denver, Pa). Fetal rabbits at gestational day 21 to 23 (term is 31 days) were used for the model of fetal-type scarless wound healing. Adult rabbits (>6 months) were used for adult-type wound healing.
Fetal rabbit surgery was performed according to an established technique as previously applied in our laboratory.5,12 Anesthesia was induced with ketamine hydrochloride (35 mg/kg) and xylazine hydrochloride (5 mg/kg). Maintenance anesthesia consisted of halothane (1%-1.5%) with oxygen delivered by spontaneous mask ventilation at a rate of 1 L/min, supplemented with local subcutaneous infiltration of the skin with 2% lidocaine.
The abdominal hair was shaved and the skin was prepared with povidone-iodine. Under aseptic conditions, a lower midline laparotomy incision was made extending from the umbilicus to the most caudal set of nipples. The size, number, and position of the fetuses were determined by palpation, and only half of the fetuses were manipulated to decrease the risk of spontaneous abortion. A purse-string suture was placed through all layers of the uterus and a hysterotomy incision was made within the borders of the suture. The fetal animal was partially delivered through the opening to expose areas intended for wounding (dorsal skin or airway) and then carefully replaced. All skin incisional wounds were made in an identical pattern (approximately 1 cm long).
Alternate manipulated fetuses received upper airway wounds as previously described.5 Briefly, the subglottic region was entered after a vertical midline incision and cricoidotomy. The posterior mucosa was wounded with a scalpel blade. The wounded areas were marked with silk suture and the fetus replaced in the amniotic sac. Isotonic sodium chloride solution was added to replace lost amniotic fluid and the pursestring suture was closed under tension. The abdominal cavity was irrigated with isotonic sodium chloride solution containing ampicillin and closed in layers with running sutures. After 12 hours, tissues were harvested and the mother was killed under anesthesia with an intracardiac injection of pentobarbital sodium (50 mg/kg).
Animal maintenance and experiments were performed according to protocols approved by the Animal Research and Care Committee of the Children's Hospital of Pittsburgh, Pittsburgh, Pa (Animal Welfare Assurance No. A3617-01).
Nonwounded and wounded skin and airway tissues from fetal and adult rabbits were rapidly harvested and flash-frozen in liquid nitrogen. Total RNA was extracted by means of TRIzol reagent (Gibco BRL, Gaithersburg, Md) according to the instructions of the manufacturer. The RNA was digested with DNase I (MessageClean Kit; GenHunter Co, Nashville, Tenn) to reduce chromosomal DNA contamination.
RNA samples from 2 or 3 individual rabbits of each group were used for mRNA differential display reverse transcription–polymerase chain reaction (PCR). Reagents were supplied in RNAimage kits (GenHunter Co). For reverse transcription, a 20-µL reaction included 9.4 µL of water, 4 µL of 5× reaction buffer, 1.6 µL of 250-µmol/L each deoxyribonucleotides (dNTPs), 2 µL of 2-µmol/L oligo(dT) primer, and 2 µL of 0.1-µg/µL DNA-free total RNA. After incubation at 65°C for 5 minutes and 37°C for 10 minutes, 1 µL (100 U) of Moloney murine leukemia virus (M-MLV; GibcoBRL) reverse transcriptase was added. The reaction was incubated at 37°C for 50 minutes and at 75°C for 5 minutes in a thermocycler 480 (Perkin Elmer, Foster City, Calif). A control was included in which no reverse transcriptase was added. The PCR reaction included 4 µL of 10× PCR buffer, 1.6 µL of 25-µmol/L each dNTPs, 2 µL of 2-µmol/L oligo(dT) primer, 2 µL of 2-µmol/L arbitrary primer, 2 µL of cDNA, 0.2 µL of [α-33P]-deoxyadenosine triphosphate ([α-33P]-dATP; 1.11 × 1014 Bq/mmol) (ICN Biomedicals Inc, Costa Mesa, Calif), and 0.2 µL (1 U) of AmpliTaq DNA polymerase (Perkin Elmer, Norwalk, Conn) in a final volume of 20 µL. Low-stringency PCR was performed to randomly amplify cDNAs for 40 cycles with temperature at 94°C for 15 seconds, 40°C for 2 minutes, and 72°C for 2 minutes in a thermocycler 9600 (Perkin Elmer). A final extension was performed at 72°C for 10 minutes. Samples without reverse transcriptase or cDNA added were simultaneously tested as negative controls.
The PCR products were run for 3.5 hours in a 6% denaturing polyacrylamide gel (National Diagnostics, Atlanta, Ga). After electrophoresis, the gel was transferred directly onto Whatman 3 MM paper (Whatman Inc, Clifton, NJ). The gel was covered with a sheet of Sealwrap (AEP Industries Inc, South Hackensack, NJ), dried under vacuum at 80°C, and exposed to film (Kodak XAR-2; Scientific Imaging Systems, Rochester, NY) for 24 to 48 hours.
A gel with bands showing differential expression was aligned with an autoradiogram assisted with Glogos II Autorad Markers (Stratagene, La Jolla, Calif). Differentially expressed bands were excised from the gel and the cDNA was extracted. The cDNA was reamplified by PCR with the use of the same primers and conditions as above, except that 250 µmol/L of each dNTP was used and no isotope was added in a final volume of 40 µL for 30 cycles. The reamplified PCR samples were run on a 1.5% agarose gel for size determination.
A PCR product containing a cDNA of interest with predicted size was purified by means of a Wizard PCR Preps DNA Purification System (Promega, Madison, Wis). The purified PCR product was cloned into pGEM-T vectors and transformed into JM109 competent cells provided in a pGEM-T Vector System (Promega). The colonies were evaluated for inserts by blue-white screening. Positive clones with cDNA inserts were screened by colony lysis PCR with the use of T7 and SP6 primers and verified via HindIII digestion (New England Biolabs Inc, Beverly, Mass). Plasmid DNA was purified with a QIAwell 8 Plasmid Kit (Qiagen, Chatsworth, Calif). DNA sequencing was performed with T7 and SP6 primers and reagents supplied in an ABI Prism BigDye Terminator Cycle Sequencing Ready Reaction Kit (Perkin Elmer) and an ABI PRISM 377 automated sequencer (PE Applied Biosystems, Foster City, Calif) at the Center for Genomic Sciences, University of Pittsburgh School of Medicine. Sequence analysis and restriction mapping were performed with Sequencher 3.1 software (Gene Codes Co, Ann Arbor, Mich). Analysis of the nucleotide sequences was performed through GenBank/EMBL databases (located at: http://www.ncbi.nlm.nih.gov/Genbank/GenbankSearch.html).
The total RNA was used, from which a cDNA of interest was isolated. Differential display reverse transcription PCR was performed with the use of reagents provided in RNAimage kits and labeled with 0.037 MBq of [α-33P]-dATP (1.11 × 1014 Bq/mmol) (ICN Biomedicals Inc). The PCR product was purified to remove unincorporated nucleotides by means of a Sephadex G-50 spin column (Boehringer Mannheim, Indianapolis, Ind). Plasmid DNA containing a cDNA insert of interest was obtained by colony PCR with the use of T7 and SP6 primers. Approximately 100 ng of plasmid DNA was applied in duplicates onto a positively charged nylon membrane (Ambion Inc, Austin, Tex) with the use of a microfiltration system (Bio-Dot; BioRad, Hercules, Calif). The membrane was dried at 80°C under vacuum and cross-linked by UV light (254 nm, 120 mJ/cm2) in a crosslinker (Stratagene).
Duplicate membranes containing the target DNA were hybridized with the 33P-labeled PCR product (specific activity, 1 × 105-6 cpm/mL). Hybridization was performed with a hybridization buffer (Gibco BRL) overnight at 65°C. The membrane was washed 2 times for 30 minutes at room temperature with a buffer containing 2× standard saline citrate–0.1% sodium dodecyl sulfate and 2 times for 30 minutes at 65°C with a buffer containing 0.1× standard saline citrate–0.1% sodium dodecyl sulfate. The membranes were then assembled and exposed for 24 hours to film (Kodak XAR-2; Scientific Imaging Systems) with an intensifying screen. Signal intensities from the autoradiographic films of the blots were determined by image scanning (Agfa NDT, Ridgefield Park, NJ) and hybridization patterns were compared.
One cDNA tag was isolated and showed a high sequence homology to rabbit PGE2 EP4 receptor mRNA. The gene-specific primers were designed (Table 1) according to published sequence information (GenBank accession No. L47207)13 and synthesized through Gibco BRL. Duplicate RNA samples (0.4 µg) of the nonwounded and wounded skin from fetal and adult rabbits were used for reverse transcription with the same oligo(dT) primer we used for differential display reverse transcription PCR. A PCR reaction included 2 µL of cDNA, 0.4 mmol/L of each dNTP, 0.2 µmol/L of each gene-specific primer, and 1.5 U of AmpliTaq DNA polymerase in a final volume of 25 µL. The PCR was performed at 94°C for 15 seconds, 65°C for 30 seconds, and 72°C for 1 minute in a PE thermocycler 9600. The PCR products were analyzed on a 1.5% agarose gel after 25, 30, and 34 thermal cycles. As an internal control, the mRNA level of a rabbit 18S ribosomal RNA (rRNA) in each sample was assessed by reverse transcription PCR with the use of rabbit 18S rRNA gene-specific primers. The 18S rRNA product was exponentially amplified by PCR for 14, 16, and 18 cycles.
The first-strand cDNA was synthesized from the total RNA isolated from nonwounded adult skin with the use of an oligo(dT)12-18 primer and a SuperScript II reverse transcriptase provided in a SuperScript Preamplification System (Gibco BRL) according to the instructions of the manufacturer. The cDNA was amplified by PCR with gene-specific primers (Table 1). The PCR product was verified by predicted size, unique restriction enzyme digestion, and direct sequencing. Alternatively, the PCR product was subcloned into TA cloning vectors (pGEM-T Vector System, Promega) and sequenced.
At least 51 unique PCR products were identified that were differentially expressed when normal and wounded tissue samples of skin and airway from fetal and adult rabbits were compared. Among them, 26 were potentially induced and 25 were repressed in fetal wound healing. To date, more than 20 differentially expressed cDNA tags have been isolated, cloned, and sequenced and are under further characterization.
One cDNA fragment preferentially induced in wounded fetal skin and repressed in wounded adult skin (Figure 1) was 99% homologous (326/352 base pairs [bp]) with rabbit PGE2 EP4 receptor mRNA at 3′ noncoding region (GenBank accession No. L47207)13 (Figure 2). In contrast to skin, the expression pattern of this fragment in fetal and adult upper airway mucosa was unchanged by wounding and thus did not exhibit differential regulation under the conditions of this experiment.
The differential expression of the EP4 clone was further verified by reverse RNA dot-blot assay (Figure 3). In normal skin, the EP4 receptor mRNA level was higher in adult than in fetal skin. However, the intensity of the signals for the EP4 receptor was increased in fetal wounded skin but relatively decreased in adult wounded skin, vs nonwounded controls, indicating the differential regulation of the EP4 receptor mRNA in the wound healing process between fetus and adult. The differential expression of the EP4 receptor mRNA was also verified by semiquantitative reverse transcription PCR at 30 cycles (Figure 4). The mRNA levels of the rabbit 18S rRNA did not differ significantly in each sample among groups at 18 cycles of PCR amplification (Figure 4).
Three cDNAs with predicted sizes were amplified with EP4 receptor gene-specific primers and verified by appropriate restriction enzyme digestion (Figure 5). Direct sequencing with the use of EP4B primers showed that the sequence (452 bp) was highly homologous to the rabbit EP4 receptor (99%; GenBank accession No. L47207)12 in the coding region (1428-1879 bp), including a seventh transmembrane domain and a C-terminal domain of the rabbit EP4 receptor (data not shown). The results further confirmed that the differentially expressed transcript is a rabbit PGE2 EP4 receptor.
Application of mRNA differential display14 provides a unique and powerful experimental system to study differential gene expression in wound healing between fetus and adult. The major advantages of this method are that a small amount of tissue can be used for generating mRNA profiles and that multiple samples at different developmental stages and under different experimental conditions can be simultaneously compared, a feature of particular importance for our purpose.
One cDNA fragment selectively induced in fetal wounded skin matched with a rabbit PGE2 EP4 receptor. The truly differential expression of the EP4 receptor mRNA was verified by reverse RNA dot blot and semiquantitative reverse transcription PCR. In normal skin, mRNA expression of the EP4 receptor was stronger in adults than in fetuses. However, 12 hours after wounding, the gene expression of EP4 was remarkably induced in fetal wounds but repressed in adult wounds. Our results suggest that the differential expression of the EP4 receptor in wound healing, rather than its mere presence, may be important in differential wound repair between fetus and adult. This has been shown to be the case for transforming growth factor-β, which is selectively up-regulated in adult but not fetal wounds,15 and which is believed to contribute to fibrotic wound repair.
Prostaglandin E2 produces a broad range of biological actions through its binding to specific receptors. One of the possible reasons for the diversity of the actions of PGE2 is the presence of various prostaglandin receptor subtypes coupled to different intracellular signal transduction pathways. The EP receptors of PGE2 have been subclassified into EP1, EP2, EP3, and EP4 receptor subtypes.16 Among these, the EP4 receptor is the most ubiquitously expressed, including lung, peripheral-blood lymphocytes, and vasculature. EP4, the most recently identified subtype, was first detected in porcine saphenous vein, where it was found to mediate PGE2-induced smooth muscle relaxation.17 Rabbit EP4 receptor has been cloned from a rabbit kidney cortex cDNA library.13 The cDNA has 7 transmembrane regions, typical of guanine nucleotide regulatory protein (G protein)–coupled receptors.
The EP4 receptor mediates various PGE2 actions through intracellular increase of adenosine 3′,5′-cyclic monophosphate (cAMP) by stimulation of adenylate cyclase via G protein in many tissues and cells.17 The effect produced by increased cAMP depends on the specific cell type. Tissue distribution and functional characterization of the EP4 receptor support its importance in mediating the effects of PGE2 on renal hemodynamics, intestinal epithelial transport, adrenal corticosteroid secretion, uterine function, and immune function. Despite the wide distribution of the EP4 receptor, it has only recently been identified by pharmacological means, and its function has just started to be investigated.
Although EP4 mRNA appears to be expressed in many cells and tissues, its expression is increased up to 3-fold by exogenous cell stimulators such as serum or bacterial lipopolysaccharide.18 This observation indicates that EP4 is inducible and may be important for establishing a context for determining the specific cellular functions that are triggered by PGE2.18 In our study, differential up-regulation of EP4 in fetal skin was found 12 hours after incisional wounding, indicating that EP4 receptor is also inducible in fetal wound healing. A specific role for the EP4 receptor in wound healing has not yet been described. During the wound healing response, specific cellular responses to prostaglandins are regulated through a complex of organized signal transducing molecules involving prostaglandin receptors, G proteins, and effectors. Recent evidence suggests that the interactions of receptors, G proteins, and effectors are highly organized on plasma membranes.19 EP4 receptor–mediated cAMP formation consists of sequential coupling of the EP4 receptor, G protein, and adenylate cyclase, that is, EP4 receptor–G proteins–adenylate cyclase transduction pathway. In our study, the finding of the up-regulation of EP4 receptor mRNA in fetal wounds indicates that it may be involved in the regulation of fetal scarless wound healing through intracellular signal transduction pathways within responding cells in the wound margin.
A significantly elevated level of PGE2 was observed only at the first day after discectomy.20 Synthesis of PGE2 was increased markedly in response to injury and was proportional to the extent of wounding.21 Platelet-derived growth factor induces phospholipase A2 activity and subsequent arachidonic acid mobilization and production of prostaglandins in wounds.22 The differential expression of PGE2 EP4 receptor mRNA in fetal and adult skin wounds in our study indicates that the synthesis of PGE2 may be regulated in skin epidermal cells through the EP4 receptor. This autocoid may act to inhibit endothelial mitosis and activity of the EP4 receptor in adult wounds. On the other hand, the elevation of PGE2 synthesis may increase mRNA expression of the EP4 receptor in fetal wounds, which we found to occur within 12 hours of injury.
The tissue movement of repair (reepithelialization and connective tissue contraction) is the same in fetuses as in adults, but the cellular responses to the wounding and repairing are very different. The chief contractile cell of adult granulation tissue is the myofibroblast, a special cell that differentiates from a normal wound fibroblast but more closely resembles a smooth muscle cell with expression of α-smooth muscle actin and a capacity for generating strong contractile forces. The presence of myofibroblasts at the wound site correlates with scar formation.3 Fetal skin wounds, however, appear to close by epithelial contraction facilitated through a rapidly assembled circumferential actin purse string.3 EP4 receptor may differentially regulate fetal wound closure by inducing circumferential assembly of the actin cytoskeleton. There is published evidence that PGE2, acting through the EP4 receptor, may play a critical role in another type of tissue closure. The ductus arteriosus of EP4-deficient mice fails to close at birth,23 indicating that the PGE2/EP4 receptor system modulates the developmental state of this major vessel, possibly by affecting smooth muscle contraction and vessel tonicity. It is proposed that PGE2 acting through the EP4 receptor is a key regulator of intracellular cAMP in the late gestation–early neonatal period.23 Our results support this hypothesis by demonstrating an altered expression of EP4 receptor mRNA from the fetal stage to the adult stage of skin wound healing.
It has been speculated that a key difference between scarless repair and scar formation in healing is due to the absence of inflammation and proinflammatory cytokines. There is always an extensive inflammatory response at adult wound sites, but in the fetus inflammation is minimal.4 The prostaglandins in general, and PGE2 in particular, are key mediators of various inflammatory processes, including the acute inflammatory response after injury that serves to regulate subsequent wound healing events. The uncovered fetal rabbit wounds do not contract or exhibit classic signs of inflammation or healing, whereas fetal wounds covered with an occlusive dressing heal with inflammation and contraction.24 It was suggested that the uncovered fetal rabbit wounds are exposed to elevated levels of PGE2 in amniotic fluid, which may be acting as a potent immunosuppressant. The anti-inflammatory activity of PGE2 mediated through the EP4 receptor has been reported.25 Our result that the EP4 receptor is up-regulated in fetal skin shortly after injury supports the hypothesis that endogenous PGE2 is binding to the EP4 receptor to produce a noninflammatory response in uncovered fetal rabbit wounds.
Our results provide preliminary evidence of the involvement of PGE2, acting through the EP4 receptor, in the regulation of fetal wound healing in the skin. Future work will focus on the cellular localization of the EP4 receptor mRNA by means of in situ hybridization and immunohistochemistry. The time course will be taken into account to determine the time-dependent expression of the EP4 receptor mRNA and protein during wound healing processes in the fetus and the adult.
Accepted for publication March 13, 2000.
This study was supported by the Center for Genomic Science and the Departments of Medicine and Otolaryngology, University of Pittsburgh School of Medicine, Pittsburgh, Pa; the Children's Hospital of Pittsburgh; and National Institutes of Health grant DK50236 RO1 (Dr Whitcomb), Bethesda, Md.
Portions of this work were presented at the 22nd Association for Research in Otolaryngology Midwinter Meeting, St Petersburg Beach, Fla, February 13-18, 1999.
Reprints: Joseph E. Dohar, MD, MS, Department of Pediatric Otolaryngology, Children's Hospital of Pittsburgh, ENT Wound Healing Research Program, Rangos Research Center, 3460 Fifth Ave, Pittsburgh, PA 15213.