Pathway Analysis of Gene Expression in Murine Fetal and Adult Wounds This abstract has been presented at the 8th Annual Academic Surgical Congress on February 5–7,2013 in New Orleans,Louisiana and the 26th Annual Meeting of the Wound Healing Society on April 23–27,2014 in Orlando,Florida.

Animals

Six-week-old wild-type BALB/c mice were purchased from Charles River Laboratories (Wilmington, MA) and housed at 22°C with a 12-h day/12-h night cycle. All animals received water and normal chow ad libitum. For timed gestations, the mice were bred overnight and the day of vaginal plug was considered E0.5 day of gestation. Animals were maintained in the Stanford Animal Care Laboratory and all procedures were conducted in accordance with university-approved protocols according to National Institutes of Health (NIH) guidelines.

Fetal mouse wounding

Fetal wounding was performed as previously described. ¹¹ Briefly, pregnant mice (gestational age E17) were utilized. After induction of anesthesia, midline laparotomy was performed using microsurgical scissors. The uterus and fetus selected for surgery were gently exposed using cotton-tip applicators. The surgical field was irrigated with warm (38°C) phosphate-buffered saline (PBS). The fetus was carefully positioned to allow access to the dorsum. A purse-string stitch using a 7-0 nylon suture was passed through the uterus overlying the site of intended dorsal wounding. A 3 mm incision was made through the uterine wall and amniotic sac in the center of the purse-string stitch. Using microsurgical scissors, a single full-thickness excisional wound, ∼1 mm in diameter, was created on the dorsum of the fetus. The wound was marked with India ink. Warm (38°C) PBS was injected into the amniotic sac with a blunt-tipped 10-guage syringe as the purse string was closed and the syringe was carefully retracted. One fetus of unknown gender per litter was wounded. The peritoneal cavity was irrigated with warm (38°C) PBS. The peritoneum and abdominal skin were everted, reapproximated, and stapled closed. Prior studies from our laboratory have shown that the histology of these wounds reveal complete regeneration within 48 h. Hematoxylin and eosin and trichrome staining reveal rapid reepithelialization and normal collagen architecture. ¹² Thus, time points less than 48 h were chosen for analysis of wounds. The pregnant mice were sacrificed and fetal wounds were harvested at 1, 12, and 24 h (n = 3 for each time point) following wounding by excising the wound with a 2 mm rim of surrounding tissue using microsurgical scissors, as previously described. ⁶ An equivalent amount of tissue from the contralateral dorsum of the same fetus was collected in the same manner for normalization.

Adult mouse wounding

For adult wounding, 2 mm excisional wounds were generated on the back of 3-week-old BALB/c mice of mixed gender using punch biopsy after induction of anesthesia and preparation for aseptic surgery, as described above for pregnant mice. Mice were sacrificed and wounds were harvested at 1, 12, and 24 h (n = 3 for each time point) following injury for comparison to fetal wounds. Wounds were harvested by excising the wound along with a 2 mm rim of normal tissue by punch biopsy for reproducibility, as previously described. ⁶ An equivalent amount of tissue from the contralateral dorsum of the same fetus was collected in the same manner for normalization. The larger size wound, which is a fourfold increase in area, accounts for an approximately fourfold size discrepancy between the fetal and postnatal wound. However, despite the differences in wound size, the relative wound to body size in the fetal and postnatal mouse is not the same and may confound our data.

RNA extraction and amplification

RNA from fetal and adult wounds at various time points was extracted using the Trizol protocol (Invitrogen, Carlsbad, CA) as per manufacturer's instructions and as previously described. ¹³ One microgram of RNA from each experimental sample (n = 3 per group per time point) was amplified. One-microgram aliquots of universal mouse RNA were amplified in individual reaction mixtures and utilized as internal amplification controls.

Preparation of fluorescent complementary DNA probes

Fluorescent complementary DNA (cDNA) probes were prepared as previously described. ¹³

Pretreatment of microarray chips

The Stanford Microarray Database Center was used to print mouse microarray chips with 42,000 specific cDNAs printed onto each lysine-coated slide. These cDNAs represent single accession numbers from Genbank (Sequences and accession numbers of the cDNAs can be found on http://genome-www5.stanford.edu//index.shtml). Before hybridization, microarray chips were rehydrated, snap-dried, and crosslinked as previously described. ¹³

Microarray hybridization

Microarray hybridization was performed as previously described. ¹³

Microarray data analysis

Scanned images were analyzed using the Genepix Pro 4.0 software (Molecular Devices), as previously described. ¹³ Significance analysis of microarrays (SAM) was used to select genes with significant expression differences between the E17 fetal and adult wound transcriptomes for each time point. Genes that had at least a twofold expression difference with false discovery rate <2 were selected.

Functional analysis of differentially expressed genes

To identify functional connections among significantly regulated genes, both network and pathway analyses of the probes filtered by microarray were performed using Ingenuity Pathways Analysis (Ingenuity Systems, Redwood City, CA), as previously described. ¹³

Differential gene expression between adult and fetal wounds

Wound microarray data were normalized to age-matched unwounded control skin (taken from the contralateral dorsum of the same fetus/adult during wound collection) data sets. Normalized transcriptomes from E17 fetal wounds were directly compared to normalized transcriptomes from adult wounds. SAM identified 471 differentially expressed genes with greater than twofold difference between E17 fetal and adult wounds at 1, 12, and 24 h postinjury. At 1 h following wounding, 178 genes were upregulated in E17 fetal wounds, whereas 13 genes were downregulated when compared to adult wounds (Table 1). At 12 h following injury, E17 fetal wounds upregulated 112 genes, whereas adult wounds upregulated 141 genes (Table 2). Twenty-four hours postwounding, 16 genes were downregulated in E17 fetal wounds and 11 genes were upregulated versus adult wounds (Table 3).

Functional pathway analysis

Analysis of the 178 genes found to be upregulated in E17 fetal wounds at 1 h postinjury resulted in the identification of 20 functional pathways (Fig. 1B). The top five pathways were associated with the following: proline biosynthesis, arginine degradation, mevalonate pathway, and Parkinson's signaling. Utilizing the 13 downregulated genes in E17 fetal wounds, a list of 20 functional pathways were again identified (Fig. 1C). The top five pathways were as follows: interleukin (IL)-8 signaling, mechanistic target of rapamycin signaling, glucocorticoid receptor signaling, regulation of IL-2 expression in activated and anergic T lymphocytes, and T cell receptor signaling.

OPEN IN VIEWER

Figure 1.

Microarray analysis of E17 fetal and adult wounds at 1 h postwounding. (A) Hierarchical clustering of differentially regulated genes from E17 fetal and adult wounds at 1 h postwounding. Individual genes are clustering according to the dendrogram on the left, and expression levels are represented in the heatmap on the right. White/lighter gray and darker gray indicate upregulation and downregulation, respectively. (B) Canonical pathways significantly enriched for among genes whose expression was significantly upregulated in E17 samples compared to adult. (C) Canonical pathways significantly enriched for among genes whose expression was significantly downregulated in E17 samples compared to adult. E, embryonic day.

At 12 h following injury, out of 112 genes upregulated in E17 fetal wounds, 20 pathways were identified (Fig. 2B). The top five pathways were related to apoptosis, role of osteoblasts, osteoclasts and chondrocytes in rheumatoid arthritis, Ras-related nuclear protein signaling, and protein ubiquitination pathway. Conversely, of 141 genes upregulated in adult wounds at 12 h, 20 functional pathways (Fig. 2C) reveal the top five to include the thioredoxin pathway, CXC chemokine receptor (CXCR)4 signaling, IL-1 signaling, Cdc42 signaling, and vitamin C transport.

OPEN IN VIEWER

Figure 2.

Microarray analysis of E17 fetal and adult wounds at 12 h postwounding. (A) Hierarchical clustering of differentially regulated genes from E17 fetal and adult wounds at 12 h postwounding. Individual genes are clustering according to the dendrogram on the left, and expression levels are represented in the heatmap on the right. White/lighter gray and darker gray indicate upregulation and downregulation, respectively. (B) Canonical pathways significantly enriched for among genes whose expression was significantly upregulated in E17 samples compared to adult. (C) Canonical pathways significantly enriched for among genes whose expression was significantly downregulated in E17 samples compared to adult.

Gene expression of E17 fetal and adult wounds at 24 h after injury reveal 27 differentially expressed genes. Of the 16 downregulated in E17 fetal wounds, 18 pathways were distinguished (Fig. 3B). Eumelanin biosynthesis, transfer RNA (tRNA) charging, glial cell line-derived neurotrophic factor family ligand–receptor interactions, neurotrophin/tropomyosin receptor kinase (TRK) signaling, and platelet-derived growth factor signaling comprise the top five pathways. From the 11 genes upregulated in the fetal wounds, 11 pathways were identified to be relevant to fetal wound healing (Fig. 3C). The top five functional pathways include dolichol and dolichyl phosphate biosynthesis, intrinsic prothrombin activation pathway, chronic myeloid leukemia signaling, IL-6 signaling, and atherosclerosis signaling.

OPEN IN VIEWER

Figure 3.

Microarray analysis of E17 fetal and adult wounds at 24 h postwounding. (A) Hierarchical clustering of differentially regulated genes from E17 fetal and adult wounds at 24 h postwounding. Individual genes are clustering according to the dendrogram on the left, and expression levels are represented in the heatmap on the right. White/lighter gray and darker gray indicate upregulation and downregulation, respectively. (B) Canonical pathways significantly enriched for among genes whose expression was significantly downregulated in E17 samples compared to adult. (C) Canonical pathways significantly enriched for among genes whose expression was significantly upregulated in E17 samples compared to adult.

Updated discoveries in differential gene expression

Using microarray analyses similar to those previously described by Colwell et al., we found 191, 253, and 27 genes that were differentially expressed between E17 fetal and adult wounds at 1, 12, and 24 h postinjury, respectively. E17 was the earliest gestational age where scarless healing occurs, ^6,12 and that allowed reproducible and reliable survival after surgery. The results represent an overall increase from the 175, 134, and 19 differentially expressed genes, at each respective time point, which were identified in the previous study. With the updated gene database utilized in this new study, we were able to detect not only a more accurate but also expanded set of genes that may serve as potential candidates for further study in scarless regeneration. Furthermore, improved pathway analysis abilities provided an objective, unified picture of associated functional pathways rather than solely individual genes, thereby further elucidating possible mechanisms in fetal wound healing. Interestingly, our results reveal some pathways that corroborate earlier hypotheses by Colwell et al., such as the upregulation of protein ubiquitination in fetal wounds at 12 h, while providing additional context for understanding specific prior findings, such as the earlier discovery of greater expression of angiomotin and our new association of the intrinsic prothrombin activation pathway in fetal wound healing 24 h postinjury.

The phenotype of fetal scarless healing is skin regeneration that cannot be differentiated from unwounded skin. Collagen deposition is unchanged and dermal appendages are present after fetal scarless repair. Although our analyses did not solely focus on the phenotypic sequelae, we identified myriad genes and pathways that may play a role in regeneration. Below, we discuss in greater detail some of the most relevant pathways found to be differentially activated in E17 fetal versus adult wounds at various time points following injury. Although the potential roles of these pathways are discussed individually, they may interact together. Such potential interactions are yet unknown and not discussed in this article.

Proline biosynthesis I

The proline biosynthesis I pathway in humans cumulates with the synthesis of l-proline. Proline, a major component of collagen and hydroxyproline, found in only a few other proteins in vertebrates, is essential to the stabilization of the collagen triple helix. ¹⁴ As such, the proline biosynthetic pathway can correspondingly contribute to the modulation of wound healing and scar formation. An hour after wounding, our results demonstrated markedly increased activation of the proline biosynthesis pathway in E17 wounds compared to adult wounds. While traditionally excess collagen has been associated with a number of fibrotic disorders, such as lung cirrhosis, excessive scar formation, and cirrhosis of the liver, ¹⁵ our results suggest that early upregulation of synthesis of a core component of collagen may actually contribute to scarless regeneration. This may be due to the effect of early upregulation of proline synthesis on the type of collagen deposited or the metabolism of collagen, especially as poly(l-proline) has been found to be an effective competitive inhibitor of vertebrate type I collagen prolyl 4-hydroxylase. ¹⁶ However, the exact mechanism underlying this observation remains to be delineated and more studies are required to elucidate means by which early upregulation of proline synthesis contributes to scarless repair.

IL-8 signaling

In adult wounds, IL-8 has been shown to be primarily responsible for chemotaxis of neutrophils, which contributes to the inflammatory process. ¹⁷ While IL-8 is known to stimulate inflammation in adult wound healing, its role in fetal wound healing remains unclear. Our results demonstrate downregulation of IL-8 signaling pathways in fetal wounds 1 h postinjury in comparison to adult wounds. This diminished IL-8 pathway activation corresponds to a diminished inflammatory response commonly found in fetal wounds as well as the lack of polymorphic leukocyte infiltrate seen in scarless wound healing. ¹⁸ Less inflammatory cell recruitment and diminished cytokine release have been suggested to cause decreased paracrine stimulation of extracellular matrix production, as well as increased fibroblast and epithelial cell migration and proliferation. ¹⁷ Thus, this lack of inflammatory cascade amplification may be crucial to the formation of an environment conducive to scarless wound healing. Last, excessive IL-8 has been found in disease states characterized by excessive fibroplasia, such as pulmonary fibrosis ¹⁹ and psoriasis, ²⁰ further underscoring its contribution to profibrotic processes.

Apoptosis signaling

Apoptosis is critical to the normal progression of wound healing. Especially, as the wound matures, apoptosis plays an important role in the regulation of collagen synthesis and degradation through its regulation of fibroblast and endothelial cell death. Our results indicate that E17 fetal wounds demonstrated upregulation of apoptosis signaling pathways 12 h following wounding in comparison to adult counterparts. This finding is supported by published studies demonstrating a role for increased induction of apoptosis in fibroblast populations as a possible mechanism for promotion of scarless wound healing. ²¹ Furthermore, recent studies have suggested that decreased rates of apoptosis in mice may lead to the formation of hypertrophic scars and keloids. ²² Taken together, these data suggest that initiation of apoptosis following injury may contribute to scarless wound healing through the programmed removal of damaged and unwanted cells at the site of injury.

CXCR4 pathway

CXCR4 encodes the receptor for stromal cell-derived factor-1 (SDF-1), which acts as a potent chemoattractant for lymphocytes and monocytes ²³ and is further responsible for the trafficking of circulating stem and progenitor cells to areas of tissue damage during cutaneous wound repair. ^24,25 SDF-1 is often induced by proinflammatory factors, such as tumor necrosis factor-α (TNF-α) and IL-1, ²⁶ and has also been found in elevated levels in fibrotic disorders. Our results demonstrate upregulation of CXCR4 pathway in adult wounds 12 h following injury in comparison to fetal wounds. This finding is in concordance with recent experimental data suggesting that the activation of SDF-1 signaling by proinflammatory factors recruits cells expressing CXCR4 such as fibrocytes, which contribute to the formation of hypertrophic scars. ²⁶ The regulation of SDF-1 and CXCR4 thus provides an important ligand-receptor target for reducing scarring through inhibition of fibrocyte trafficking.

Neurotrophin/TRK signaling

Neurotrophin is synthesized and released by many skin cells, including keratinocytes, melanocytes, and fibroblasts. ²⁷ In skin, neurotrophins help regulate innervation and act as prosurvival and growth factors. As such, they are believed to play an important role in regulating skin homeostasis in both physiologic and pathologic states. Data from our pathway analysis indicate that neurotrophin and its receptor TRK signaling are upregulated in adult wounds 24 h postinjury in comparison to E17 fetal wounds, and as such may play an important role in scar formation. This finding is supported by studies showing that neurotrophin acts to recruit fibroblasts. In addition, injury and inflammation both act to enhance neurotrophin production, mainly by keratinocytes at the site of injury, supporting the role of neurotrophin signaling during wound repair. ²⁸ Altered levels of neurotrophin signaling may thus be expected to contribute to adult scar formation, although no studies to date have looked at this relationship.

Intrinsic prothrombin activation pathway

Thrombin, which is produced from prothrombin, enhances the production of cytokines such as IL-1, 6, and 8, and TNF-α to upregulate the production of growth factors that in turn induce cellular proliferation at the site of injury. ²⁹ This suggests that prothrombin, thrombin, and their receptors may be responsible for modulating the various phases of scar formation. Our microarray results indicate that the intrinsic prothrombin activation pathway is upregulated in fetal wounds 24 h following wounding in comparison to the adult. Increased levels of thrombin and prothrombin have been found in old scars, suggesting that these proteins play an extended role in wound healing beyond the more immediate coagulation following injury. The unexpected result that this pathway is more active during scarless repair suggests an unknown function during extracellular matrix formation and remodeling. Interestingly, studies have found that the administration of thrombin peptides to incisional wounds in rats accelerates normal wound healing and enhances neovascularization. ³⁰ However, more studies are required to delineate the exact mechanisms by which differential activation of this pathway contributes to scarring versus regeneration.