Day 7 images of wounds treated with labeled fibroblasts. Photomicrographs are overlays of fluorescence microscopy over bright field microscopy. Each photograph is representative of those wound samples whose fluorescent values fall within the specified range. The number of samples that fall within each range is indicated below each photograph. i Indicates relative intensity.
Histogram of fluorescence data for each cell type, vehicle, and control at 7, 14, and 28 days. Statistically significant relationships are indicated by the horizontal lines above each graph. i Indicates relative intensity.
Photomicrograph of cutaneous wound treated with CM-DiI–labeled fibroblasts (CellTracker CM-DiI; Molecular Probes Inc, Eugene, Ore) (day 14, adult rabbit fibroblast–treated wound). Cells can be seen migrating into normal wound margins as the wound begins to contract. Image was obtained at original magnification ×10.
Histogram of tensiometry data for each cell type, vehicle, and control at 7, 14, and 28 days. Statistically significant relationships are indicated by the horizontal lines above each graph.
Sandulache VC, Zhou Z, Sherman A, Dohar JE, Hebda PA. Impact of Transplanted Fibroblasts on Rabbit Skin Wounds. Arch Otolaryngol Head Neck Surg. 2003;129(3):345-350. doi:10.1001/archotol.129.3.345
To determine the effect of injected fibroblasts on full-thickness cutaneous wounds and to distinguish between the properties of fetal vs adult fibroblasts.
Full-thickness cutaneous wounds were created by incision in the skin of adult New Zealand white rabbits and treated with fluorescently labeled autogenic, allogenic, and xenogenic fetal and adult fibroblasts. Wound healing was evaluated by histologic analysis and tensiometry over time.
A total of 9 New Zealand white rabbits with 24 wounds per animal were examined in this study. Groups of 3 animals were treated with different fibroblasts and euthanized at 7, 14, and 28 days after manipulation.
Fibroblasts were labeled with a fluorescent dye (CM-DiI) and suspended in a hyaluronic acid gel. The cell-gel mix was used to treat full-thickness incisional wounds in rabbit skin. Imaging of CM-DiI determined the quantity and the migratory patterns of the transplanted fibroblasts present in the wounds. Tensiometry characterized the mechanical properties of the healed connective tissue.
Transplanted fibroblasts exhibited good survival and migration patterns. Cell therapy increased the tensile strength of the wounds. Allogenic fetal and autogenic adult fibroblasts achieved similar effects.
Cell therapy is a viable approach to significantly affect the course of normal cutaneous wound healing, and cell lines from genetically unrelated donors do not appear to be disadvantaged by a host immune response compared with autogenic cell lines.
CELL-CELL AND cell-matrix interactions are important in overall body homeostasis. Understanding and characterizing these interactions can serve to elucidate the pathways by which tissue heals, dies, or regenerates. The process of wound healing is complex, involving numerous extracellular components, cell types, and mediators. Inflammation, angiogenesis, and reepithelialization, are important components of this process. However, it is fibroplasia that allows a healing tissue to regain form and function similarly to how it functioned prior to the injury.1 Fibroblast migration into the wound area is followed by production of matrix components and remodelers. This function is modulated by chemical messengers secreted by macrophages, platelets, and lymphocytes.2 The collection of these signals directs fibroblasts to fill in the wound with cells and extracellular matrix.
It has been thoroughly established that not all connective tissue is alike in its wound healing response. Fetal skin responds to injury differently than adult tissue; extracellular matrix components such as collagen show differential patterns of expression between the 2 tissue types.3- 6 This raises the question of whether fetal wound healing is different because of the components of the tissue or the wound environment that surrounds them. This issue has been previously addressed, and it was concluded that the properties associated with fetal wound healing are intrinsic to the tissue.7- 9 Because fibroblasts retain their phenotype in culture and can generate varied wound-healing responses, we hypothesize that transplanting a quantity of cells into a wound may modulate the healing process. We believe that the addition of exogenous fibroblasts into the wound environment will increase wound tensile strength due to earlier deposition and remodeling of the extracellular matrix. We also expect that the transplanted fetal fibroblasts will allow the wound to display a more regenerative healing pattern.
The goal of this study was to determine the effect of injected fibroblasts on full-thickness cutaneous wounds and to distinguish between the properties of fetal vs adult fibroblasts, autogenic (same animal) vs allogenic (same species) vs xenogenic (different species). To accomplish this, we first examined the fate of transplanted fibroblasts, observing survival and migration. Second, we measured the impact of transplanted fibroblasts on the wound healing response using tensiometry. Our findings provide proof of the feasibility of using nonautogenic transplanted fibroblasts for future development of cell therapies for regulating postnatal wound healing, potentially toward a more regenerative healing response.
The following donor cells were cultured according to existing protocols.10 Allogenic fibroblast groups included fetal rabbit, neonatal rabbit, and adult rabbit. Xenogenic fibroblast groups included fetal pig on day 75 of gestation and neonatal pig fibroblasts. Autogenic fibroblasts were also prepared from each of the experimental animals as follows. Full-thickness biopsy specimens (1.5 × 0.5 cm) were obtained from the rabbit hind leg skin after the skin was prepared with 10% povidone-iodine and 70% alcohol swabs. Briefly, fibroblast cultures were established as follows. The tissue was excised aseptically then washed in Dulbecco's modified Eagle's medium (DMEM) (Invitrogen, Grand Island, NY), 100 U/mL of penicillin, 100 U/mL of streptomycin sulfate, and 25 µg/mL of amphotericin C and kept at 4°C overnight. Using aseptic technique, skin sections measuring approximately 1 mm2 were partially digested with collagenase type I (Sigma Chemical Co, St Louis, Mo) and were placed eccentrically in 75-mm2tissue culture flasks and covered with DMEM containing 10% fetal calf serum, 10 mM HEPES buffer, and antibiotics. The cultures were left undisturbed for 1 week to promote tissue explant attachment and incubated in a humidified carbon dioxide incubator maintained at 37°C. Then the medium was changed twice a week and cultures examined for fibroblast outgrowth. Fibroblasts were harvested from primary culture after approximately 21 days and either subcultured immediately or cryopreserved and stored in liquid nitrogen for future experiments.
Carbocyanine dye (CellTracker CM-DiI; Molecular Probes Inc, Eugene, Ore) was used to label the cultured fibroblasts according to the manufacturer's standard protocol as previously described.10 The cells were counted and suspended in Hanks balanced salt solution (Life Technologies Inc, Rockville, Md) followed by mixing using extrusion from a syringe with hyaluronic acid non–cross-linked polymer (Hylan A gel; Genzyme Biosurgery, Cambridge, Mass) until a homogeneous mixture was obtained containing approximately 1.5 × 106/mL of cells.
A total of 9 New Zealand white adult rabbits, weighing 3 to 4 kg (Covance, Denver, Pa), were used in the experiments. All work conducted on these animals was consistent with standards approved by the Animal Research and Care Committee of Children's Hospital of Pittsburgh. The animals were housed individually in standard cages, in a room with controlled temperature and light in the Animal Facility at Rangos Research Center, Pittsburgh (Animal Welfare Assurance No. A3617-01, accredited by the Association for the Assessment and Accreditation of Laboratory Animal Care International).
The rabbits were acclimated to their environment for 5 days before biopsy specimens were obtained for primary cell culture work. They were given standard rabbit diet and water ad libitum. For both the biopsy and wounding procedures, the rabbits were anesthetized using intramuscular administration of ketamine hydrochloride (35 mg/kg) and xylazine (5 mg/kg). For all procedures, the animals were shaved and the skin was examined to ensure integrity and lack of any infection. The experimental area was thoroughly cleansed with 10% povidone-iodine and 70% alcohol swabs before manipulation. Aseptic technique was maintained throughout all procedures. The animals were fully anesthetized throughout the entire procedure and were observed every 8 hours for the next 48 hours and on a daily basis throughout the rest of the 28-day experimental period. Postprocedure behavior patterns did not change from the preprocedure patterns. No changes in appetite or behavior were observed at any point in the study. All animals were healthy and alive at the end of the experimental period, without weight loss. Some animal noncompliance was observed, consisting of attempts to remove the occlusive dressing covering the wounds. No injury or infection resulted from this behavior.
Full-thickness incisional wounds were created in the dorsal skin of the animals. A grid was made using a permanent marker, leaving a 2-cm margin between wounds, to prevent the risk of fibroblast migration to adjacent wounds. Each animal received 24 wounds, with 3 wounds assigned to each of the 6 experimental fibroblast lines, 3 wounds as vehicle controls containing only hyaluronic acid gel, and 3 wounds as untreated controls containing neither fibroblasts nor vehicle.
A 5-mL syringe with a 20-gauge needle was then used to place approximately 50 µL of the hyaluronic acid–fibroblast mixture to each wound. The wounds were left untouched for 5 to 10 minutes after which a coating of liquid bandage (NewSkin; Medtech Products Inc, Jackson, Wyo) was applied to every treated and control wound individually and allowed to dry for 5 minutes. The entire experimental area was then covered with a sterile transparent occlusive dressing (Tegaderm; 3M, Minneapolis, Minn) and secured with a nontoxic degradable adhesive (Mastisol; Ferndale Laboratories Inc, Ferndale, Mich).
The 9 animals were euthanized in groups of three, 7, 14, and 28 days after transplantation. The rabbits were heavily anesthetized using the ketamine-xylazine mixture, followed by intracardiac administration of euthanasia solution (pentobarbital sodium and phenytoin sodium) (1.5 mL). The dorsal skin was removed using the aseptic technique, followed by dissection of each individual wound. The samples were then placed at −80°C overnight and the next day embedded in frozen section embedding medium (Fisher Scientific, Springfield, NJ). Thin 6-µm sections were cut on a cryostat at −20°C, mounted on glass microscope slides, stored in moisture-proof slide boxes at −20°C, and protected from light until evaluation by fluorescence and light microscopy.
The wound sections were analyzed using an imaging system consisting of a microscope (Nikon Inc, Melville, NY), imaging software (Metamorph; Universal Imaging Corporation, Downington, Pa), and a printer (FujiX Pictography 3000; Fuji Photo Film, Edison, NJ). Fluorescence data were collected using a rhodamine filter (TRITC). All calculations were conducted on images collected at original magnification ×10. The data were then analyzed using a computer database program (Microsoft Excel; Microsoft Corporation, Bellevue, Wash) and a statistical software package (STATISTICA; StatSoft Inc, Tulsa, Okla). The thresholding function in Metamorph was used to measure the total fluorescence area (transplanted cell density) present in a cross-sectional slide obtained at the center of each wound. The fluorescence value was then stored in a Microsoft Excel database. Time and cell-type effects were determined using statistical computer software. This statistical package was used to draw conclusions about statistically significant relationships for the entire data sample consisting of 216 wounds. Individual comparisons between the different cell types were determined using means and SEMs generated by the database.
Tensiometry data were collected using a tensiometer and data collection software (custom built). The excised wounds (approximately 1-cm wide) were trimmed of subcutaneous fascia. Each wound was cut into sections 3 to 5 mm in width. The cross-sectional area of each section was measured with calipers. Then the section was clamped in the tensiometer, and force was exerted to the breaking point. Measurements were recorded by a computer and tensile strength was calculated using the following formula:
Maximum Tensiometer Reading (Converted to Grams) ÷ Cross-sectional Area (mm2) = Tensile Strength (g/mm2).
The results for individual sections from one wound were combined for each wound specimen to determine an average tensile strength per wound. The tensile strength per wound was tabulated for each group at each time point, and the means and SDs were determined using database software.
Previous in vitro experiments were used to determine the optimum experimental procedure described herein. Adult rabbit fibroblasts were labeled with CM-DiI dye and grown in culture. The cells were imaged at 7 and 14 days. The fluorescence persisted through day 14, although the intensity was reduced due to dilution of the label with multiple cell divisions. Imaging of these living fibroblasts revealed that the dye has a microsomal distribution pattern that excludes the nucleus. Previous in vivo experiments have shown that the dye-labeled cells are still present and detectable in unwounded skin at 28 days.10
A second in vitro experiment was used to test the viability of cells mixed with hyaluronic acid gel. Adult rabbit fibroblasts were labeled with CM-DiI dye and mixed into a volume of hyaluronic acid gel according to the procedure discussed herein. The cell-gel mixture was cultured for 7 days. The labeled fibroblasts grew, multiplied, and exhibited the same fluorescence pattern as labeled fibroblasts grown under standard culture conditions. The hyaluronic acid gel appeared to solubilize into the medium within 2 hours of incubation.
Day 7 wounds were used to determine the success rate of the cell delivery technique. This was based on the assumption that on day 7 cell presence was more indicative of survival than ability to multiply. Fluorescence measurements were used to approximate the amount of cells present in each wound. Figure 1 contains a numerical analysis of the 52 samples collected at day 7, as well as photographic companions, which can be used to determine the significance of the numerical values. As shown, the absolute success rate for the delivery technique was 92% (number obtained by counting all the samples in which cell delivery succeeded, regardless of specific amount). Fluorescent label was detected in neither the untreated nor the vehicle control wounds. Therefore, although fibroblasts did migrate into the margins of the wound, they did not migrate into adjacent wounds.
Image analysis of all wound samples revealed several interesting trends. There was a statistically significant time effect (F test, P<.001). Figure 2 illustrates the decrease in cell number for each individual cell type from day 7 to day 28. Since only 3 time points were included, the true rate of decay cannot be determined. Statistical analysis of the fluorescence data revealed an additional cell-type effect (F test, P<.01). However, in this case, there was no clear pattern that persisted throughout the entire experimental period. The xenogenic neonatal fibroblasts seemed to exhibit optimal survival at least through the first 14 days compared with the other categories. However, the reproducibility of this result is unknown due to some variability in the delivery technique. Individual statistically significant differences are illustrated in Figure 2. Additional analysis revealed no significant differences between fetal and adult fibroblast treatment groups. There were no significant differences among autogenic, allogenic, and xenogenic fibroblasts, suggesting that fibroblasts in general are immunologically tolerated. Previously, we reported that xenogenic adult fibroblasts did not survive well in host tissue,10 so this cell type was excluded from further study.
Histologic analysis of the various wound samples revealed additional information about the distribution of transplanted fibroblasts. As shown in Figure 1, wounds collected at days 7 and 14 revealed a cell pattern concentrated in the immediate wound area (data for day 14 not shown). A large percentage of the labeled fibroblasts were observed at the epidermal-dermal interface. However, by day 14 transplanted fibroblasts were observed migrating along the border between the dermis and the underlying fascia. As shown in Figure 3, there appears to be a clear migration of cells from the initial wound area into the surrounding normal tissue. In addition, detailed histologic examination of all wound samples revealed no inflammation as characterized by the presence of lymphocytes and macrophages.
Tensiometry measurements were collected for all 3 time points. Statistical analysis revealed 2 significant effects: time (F test, P<.01) and cell type (F test, P<.001). Figure 4 illustrates the changes in tensile strength for each cell type and the vehicle and untreated control wounds throughout the 28 days of the experiment. In general, tensile strength increased with increasing time.
At day 7, wounds containing xenogenic fetal fibroblasts, neonatal allogenic fibroblasts, and xenogenic neonatal fibroblasts all exhibited higher tensile strength than control wounds. At day 14, wounds containing allogenic neonatal fibroblasts and xenogenic neonatal fibroblasts still showed higher tensile strength than controls, along with wounds containing allogenic adult fibroblasts. However, by day 28, no statistically significant differences were observed between any cell type and control wounds, indicating that normal wound healing was "catching up" with cell therapy–increased healing by this time. Control and vehicle wounds yielded similar tensile strength results at all time points. Autogenic fibroblasts and allogenic fetal fibroblasts showed similar tensiometry values throughout.
We investigated a novel approach for the delivery of viable labeled fibroblasts into full-thickness cutaneous wounds. Our technique had an absolute success rate of 92%. On the basis of this study, 2 conclusions about the viability of transplanted cell therapy for wound healing can be reached. First, transplanted fibroblasts do not appear to be immunogenic, as shown by the absence of inflammatory cell infiltrates within the wound area where donor cells were present. Second, all of the fibroblasts exhibited similar survival profiles in healing wounds throughout 28 days. The fluorescence data indicate several statistically significant differences in cell survival, most notably, the success of xenogenic neonatal fibroblasts. At day 7, these cells were present in higher numbers than most other cell types. It is interesting that xenogenic fetal and neonatal fibroblasts exhibited significantly different survival, considering the 2 cell types differ in gestational age by only about 5 weeks.
Fetal and adult fibroblasts did not exhibit significantly different survival patterns. Previous studies have shown that fetal tissue heals in a dramatically different manner from adult tissue. The importance of amniotic fluid and the sterile intrauterine environment in this process has been refuted by previous work, and many believe that it is fetal tissue and the fetal cells themselves that possess special regenerative properties.9 Future work will determine whether transplanted fibroblasts can regulate the qualitative outcome of healing toward regeneration, but this study demonstrates the feasibility of such an approach.
The tensiometry data suggest that understanding the dose effect of cell therapy is crucial to completely elucidating the role of exogenous fibroblasts in a wound environment. It is clear that transplanted fibroblasts can affect wound tensile strength. Several of the cell treatment groups exhibited statistically significant increased tensile strength, among them xenogenic neonatal, allogenic neonatal, and allogenic adult fibroblasts. Other groups, autogenic and allogenic fetal fibroblasts, showed a similar trend but did not achieve statistical significance due to small sample size. It is possible that by transplanting fibroblasts into the wound, normal migration processes, including fibroplasia, have been bypassed, allowing for earlier wound repopulation and matrix production, as we had hypothesized. We are currently investigating this hypothesis using immunohistochemical methods. Irrespective of the mechanism of action, the effect of the transplanted cells on the regain of wound tensile strength indicates that it is possible for transplanted cells to have an impact on the host wound environment. However, the increase in tensile strength did not precisely correlate with the number of cells present in the wound. For example, at day 7 allogenic adult fibroblasts were second highest with respect to cell number, yet this did not make a difference on the wound tensile strength. This may be an indication that both cell number and cell type are important parameters.
There are several sources of error, which if corrected should improve future studies. First, a micromixing apparatus that delivers a more precise dose of cells to each wound is under development. Second, variation in wound volume will be controlled by using an adhesive mesh grid to serve as a wound template, as well as using cautery and packing to prevent cell washout due to bleeding. Finally, animal noncompliance will be reduced by increasing the partial sedation of the animals and applying secondary dressings. These sources of error, however, are probably partially offset by the large number of wound samples (>200).
In conclusion, we confirm that fibroblasts transplanted in the wound survive and persist in a time-dependent manner and that they contribute to the wound healing process by expediting regain of tensile strength. The transplanted cells do not appear to incite inflammatory response. Both allogenic and xenogenic and both neonatal and fetal fibroblasts showed promising results that warrant further investigation.
Corresponding author and reprints: Patricia A. Hebda, PhD, Department of Pediatric Otolaryngology, Children's Hospital of Pittsburgh, Rangos Research Center, 3460 Fifth Ave, Pittsburgh, PA 15213 (e-mail: email@example.com).
Accepted for publication July 23, 2002.
This study was supported by grants from Children's Hospital of Pittsburgh and the Pittsburgh Tissue Engineering Initiative, Pittsburgh, Pa.
This study was presented in part at the Annual Meetings of the American Society of Pediatric Otolaryngologists, Boca Raton, Fla, May 13-14, 2002, and the Wound Healing Society Meeting, Baltimore, Md, May 31, 2002.
We thank Beverly Gambrell, BS, for skilled technical assistance with all of the tissue processing for microscopic analysis.