The Lack of Thrombospondin-1 (TSP1) Dictates the Course of Wound Healing in Double-TSP1/TSP2-Null Mice

Animal Models

Homozygous TSP1-null (129SvJ) and TSP2-null mice (129SvTer) were generated as previously described. 6,7 Single-null mice were mated and the resulting heterozygotes were bred to obtain wild-type and the three null genotypes on a 129SvJ/129SvTer background. All experiments performed with these mice have been approved by the Institutional Animal Care and Use Committee at the University of Washington.

Analysis of Genomic DNA by Polymerase Chain Reaction (PCR) and Southern Blot

PCR for TSP1 and Southern analyses for TSP1 and TSP2 were performed as described previously. 6,7 Briefly, genomic DNA was digested with StuI for Southern analysis of TSP1, and with BamHI for analysis of TSP2. For PCR analysis of TSP2, tail DNA was digested with proteinase K (100 ng/μl) at 55°C overnight, followed by isopropanol precipitation. The DNA was used with forward TSP2 primer (5′-CTGGTGACCACGTCAAGGACACTTCAT-3′) and reverse TSP2 primer (5′-ATGCACCTTTGGCCACGTACATCCTGC-3′) for the wild-type allele; and forward Neomycin primer (5′-ATGACTGGGCACAACAGACAATCGGCT-3′) and reverse Neomycin primer (5′-CCGCATTGCATCAGCCATGATGGATAC-3′) for the mutant allele. DNA was amplified in 100-μl reactions containing 100 to 300 ng DNA, 20 pmol of each forward and reverse primers, 1.0 mmol/L MgCl₂, 200 μmol/L of each dNTP, and 2.5 U of TaqDNA polymerase (Promega, Madison, WI). PCR amplification was performed using the following conditions: 30 cycles at 94°C for 1 minute, 60°C for 1 minute, and 72°C for 2 minutes, followed by a final cycle at 72°C for 10 minutes. The wild-type and mutant alleles yield PCR products of 539 bp and 900 bp, respectively.

Cell Culture

Skins were taken from the backs of 3-month-old mice. After removal of hair, specimens were digested with collagenase (Sigma Chemical Co., St. Louis, MO), and processed as previously reported. 13 After two passages, the cell population appeared, by light microscopy, to be composed almost entirely of fibroblasts.

Western Blot Analysis

Fibroblast monolayer cultures were grown in Dulbecco’s modified Eagle’s medium supplemented with 10% (v/v) fetal bovine serum. On reaching confluence, the medium was replaced with 5 ml of serum-free medium and the cells were incubated for 18 to 20 hours. The conditioned medium was collected, concentrated with a Centricon-10 (Amicon, Danvers, MA), and protein concentrations were determined with the BCA assay (Bio-Rad, Hercules, CA). Twenty μg of total protein was subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis; after electrophoresis the proteins were transferred to a nitrocellulose membrane. Blots were then incubated with antibodies against murine TSP1 (a kind gift of Dr. D. Mosher, University of Wisconsin, Madison, WI) and TSP2 (developed by Dr. L. Armstrong in our laboratory).

Wounds

Animals were anesthetized and 6-mm excisional wounds were made as described previously. 14 All mice used in these experiments were ∼3 months of age and were sex-matched. All wounded animals were housed individually, and a total of four mice/genotype/time point were used.

Tissue Processing and Histology

Excised wounds were fixed in 10% zinc-formalin buffer (Z-fix; Anatech, Battle Creek, MI), paraffin-embedded, and sectioned at 5 μm thickness, as previously described. 14 Lungs were processed similarly. Sections were subsequently stained with hematoxylin and eosin (H&E) and Masson’s trichrome according to standard procedures.

Immunohistochemistry

Paraffin-embedded sections were stained with antibodies against PECAM-1 (Pharmingen, San Diego, CA), TSP1, and TSP2 as described previously. 14 Macrophage analysis was performed by staining sections with anti-Mac3 antibodies (Pharmingen) according to the supplier’s instructions. For each wound section, the number of macrophages was counted in five to eight random high-power fields with the aid of an optical grid. Data were analyzed for significance by a one-way variance (analysis of variance). All examinations were performed with the aid of an Eclipse 800 microscope (Nikon, Tokyo, Japan).

Quantification of Neovascularization

Images were captured with a digital camera and computer-assisted morphometric analysis was performed using Metamorph software (Universal Corp., West Chester, PA). Neovascularization of granulation tissue was measured as described previously. 14,15 PECAM-1-positive vascular profiles in the stained sections were measured and numbers of blood vessels and vessel diameters per high-power field (0.04 mm²) were determined. The collected data were used to determine vessel size distribution. Four sections per wound were analyzed.

Determination of Bleeding Time

Mice were restrained and the tip of the tail (0.4 cm) was cut with a razor blade. The cessation of bleeding was measured as described previously. 7

Determination of Monocyte Chemoattractant Protein (MCP-1) Levels in Wounds

Wound tissues (n = 6) were prepared as described previously. 16 Briefly, wound tissue was homogenized with a Polytron homogenizer in 1 ml of phosphate-buffered saline containing 2 mmol/L of phenylmethyl sulfonyl fluoride and 1 μg/ml each of aprotinin, leupeptin, and pepstatin A (Sigma Chemical Co.) and sonicated for 1 minute on ice. The debris was removed by centrifugation and the protein contents of wound extracts were determined by the BCA assay according to the manufacturer’s instructions (Bio-Rad). MCP-1 protein levels were determined in 50- to 100-μl samples using a commercially available enzyme-linked immunosorbent assay kit (Bio Source International, Camarillo, CA). Background levels were subtracted from all measurements, and data were analyzed for significance by a one-way variance (analysis of variance).

Determination of Total and Active TGF-β1 Protein Levels in Wounds

Wound tissues (n = 6) were prepared as described above and levels of active TGF-β1 protein in 100-μl samples were determined using a commercially available enzyme-linked immunosorbent assay kit (R & D Systems, Minneapolis, MN). To measure levels of total TGF-β1, wound extracts were heated at 80°C for 10 minutes to activate endogenous latent TGF-β1. Background levels were subtracted from all measurements, and data were analyzed for significance by a one-way variance (analysis of variance).

Generation and Characterization of Double-TSP1/TSP2-Null Mice

TSP1- and TSP2-null mice, derived on a 129SvJ/129SvTer genetic background, were crossed and their progeny were bred to produce double-TSP1/TSP2-null animals. Double-null mice were obtained at approximately half the frequency expected from Mendelian ratios. This finding is in agreement with the reduced viability of TSP1-null mice. 6 On the other hand, TSP2-null mice were generated with the expected frequency. 7 On Southern analysis, a 4.4-kb StuI fragment was detected for the mutant TSP1 allele, and a 5.8-kb fragment for the wild-type allele. For TSP2, a 4.8-kb BamHI fragment was detected for the mutant allele and a 6.0-kb fragment for the wild-type allele (Figure 1, A and B) ▶ . PCR analysis produced fragments with the expected sizes of 400 bp for TSP1 mutant allele and 700 bp for the wild-type allele. For TSP2, the product sizes were 900 bp for the mutant TSP2 allele and 539 bp for the wild-type allele (Figure 1, C and D) ▶ . Western blot analysis of conditioned media from dermal fibroblasts confirmed the absence of proteins encoded by the targeted alleles (Figure 1, E and F) ▶ . The difference in intensity of the bands in lanes 1 and 3 of Figure 1E ▶ was not reproducible.

Double-TSP1/TSP2-null mice appeared to be normal on superficial examination, but displayed a mild spinal lordosis that was also noted previously in TSP1-null animals. Histopathological examinations of the heart, kidney, and spleen revealed no major abnormalities, but the lungs showed evidence of acute and chronic inflammation marked by patchy consolidation and prominent macrophage infiltration, as had been reported for TSP1-null mice 6 (Figure 2) ▶ . Consistent with our earlier description of TSP2-null animals, 7 a moderate fragility of the skin and ductility of the tail were observed in double-null animals (data not shown). Finally, the bleeding time, defined as the time required for cessation of bleeding after excision of the terminal 0.4 cm of the tail, was prolonged in double-null animals (8 ± 3 minutes, n = 5, in comparison with a bleeding time of 3 ± 1 minute, n = 5, in wild-type mice). This finding was similar to that previously reported for TSP2-null mice. 7 Thus, in unchallenged mice the apparent phenotype of double-null mice is the sum of the phenotypes of TSP1- and TSP2-null animals.

Histological Analysis of Excisional Skin Wounds

We chose to examine wound healing in TSP1/TSP2 double-null mice because previous studies had shown that both TSP1 and TSP2 are expressed during the course of wound healing in wild-type animals. We had previously determined that although the rate of re-epithelialization in TSP2-null excisional wounds was the same as in controls, scab loss and wound closure were accelerated. On the other hand, wound healing is delayed when TSP1 is down-regulated using an antisense strategy. 16 Gross examination of healing wounds revealed that, similar to TSP1-null mice but in contrast to TSP2-null animals, double-null mice exhibited delayed healing, as indicated by prolonged inflammation and the persistence of scabs. Examination of H&E-stained day 14 wound sections revealed prominent differences in the granulation tissue of all three mutant mice, in comparison with that in wild-type mice. Thus, the fibers in the wound bed of TSP1-, TSP2-, and double-null animals lacked the parallel orientation to the epidermis that was evident in wild-type wounds (data not shown). To visualize the orientation of collagen fibers more specifically, sections of day 14 wounds were also stained with Masson’s trichrome. As observed in Figure 3 ▶ , the collagen fiber pattern in double-null animals resembled that in TSP1-null mice and was less dense than that in either control or TSP2-null animals. The patterns in wild-type and TSP2-null wounds also differed in that collagen fibers in the latter were disorganized and lacked a parallel orientation. Abnormalities in collagen-staining patterns in TSP1-null and double-null wounds could result from the reduced inflammatory response and subsequent reduction in cytokines, such as TGF-β1, that promote matrix synthesis in these mice. The basis for the disorganization of collagen fibers in TSP2-null wounds is still not understood.

Time Course of Expression of TSP1 and TSP2

Both TSP1 and TSP2 are induced in response to injury. 16,17 To examine the time course of expression of these two proteins, sections of wounds were stained with antibodies against murine TSP1 and TSP2. Because the antibodies raised against these two proteins probably possess different antigenic affinities, their levels of expression cannot be compared directly. In agreement with previously published reports, TSP1 protein was detected during the early stages of wound healing. 15,16 The highest expression of TSP1 was observed on day 3 after injury, the earliest day examined, during the inflammatory phase of the wound-healing process (Figure 4A) ▶ . Subsequently, TSP1 levels dropped significantly at day 7, and by day 10 they were undetectable. In contrast, TSP2 protein was first detected at day 7 after injury and reached maximal levels by day 10, during the remodeling phase of the healing response. This pattern of deposition in tissues is reflective of the cells that synthesize these two proteins. TSP1 is released from platelets and synthesized by inflammatory cells that are present early at the wound site, whereas TSP2 is produced primarily by fibroblasts that comprise the major cell population at the wound site by day 10. Thus, the physiological roles of these two proteins undoubtedly reflect their synthesis by different cells in different temporal and spatial patterns.

Immunohistochemical analysis revealed that the temporal distributions of TSP1 in TSP2-null wounds, and of TSP2 in TSP1-null wounds were comparable to those present in wild-type wounds (Figure 4, B and C) ▶ . These findings provide strong evidence for the view that neither TSP compensates for absence of the other by altered expression.

Neovascularization in Double-TSP1/TSP2-Null Animals

Wound neovascularization was quantified with the aid of Metamorph imaging software after immunolocation of the endothelial cell marker, PECAM-1. 15 At both time points (7 and 14 days), the number of blood vessels per unit area was markedly higher in TSP2-null than in wild-type wounds (Figure 5A) ▶ , a finding that is consistent with our previous observations. 14 However, the extent of neovascularization in double-null mice was similar to that observed in wild-type and TSP1-null mice (Figure 5A) ▶ . Both the diameter and the size distribution of blood vessels in granulation tissue were comparable among all four genotypes (Figure 5, B and C) ▶ . Thus, in TSP1- and double-null mice the absence of TSP1 did not alter the vascular density of wounds, whereas the lack of TSP2 resulted in increased neovascularization only in TSP2-null animals. Interestingly, the vascular density of wounds in double-null mice was not influenced by the absence of TSP2.

Macrophage Infiltration in Excisional Wounds

TSP1 has been found to be a major chemoattractant for macrophages. 18 In an effort to assess the contribution of this protein to the influx of macrophages in wounds, we quantified the recruitment of macrophages with anti-Mac 3 antibodies. When compared to wild-type or TSP2-null wounds, both TSP1-null and double-TSP1/TSP2-null wounds exhibited a significant reduction in the number of macrophages in the wound bed, as determined by the number of macrophages per high-power field (Figure 6A) ▶ .

MCP-1 is a major chemokine for monocytes, the progenitors of tissue macrophages, and shows a temporal pattern of expression during wound healing that correlates with the infiltration of macrophages at the wound site. 19 Therefore, we assessed the levels of MCP-1 protein in wounds at 12 hours, day 1, and day 3 after injury. No significant differences were observed 12 hours after injury, but the levels of MCP-1 protein in day 1 and day 3 wounds of TSP1-null and double-null wounds were reduced by ∼40% (Figure 6B) ▶ . The amount of MCP-1 in TSP2-null wounds did not differ from that in wild-type wounds at any of the time points (Figure 6B) ▶ .

TGF-β1 Levels in Day 3 Wounds

Because TSP1 is reported to be a major activator of TGF-β1, 20 we determined the level of this cytokine in day 3 wounds. The amounts of both total and active TGF-β1 were significantly reduced in TSP1-null and double-TSP1/TSP2-null wounds compared to those in wild-type and TSP2-null wounds (Figure 7) ▶ , but the ratios of active to total TGF-β1 were similar in all four groups of animals (wild-type, 0.18; TSP2-null, 0.16: TSP1-null, 0.16; double-null, 0.16). Thus, it is likely that the reduced levels of active TGF-β1 in TSP1- and double-null mice result from the presence of correspondingly lower amounts of latent TGF-β1 in the wounds of these mice.