- Email: [email protected]
Recombinant basic fibroblast growth factor accelerates wound healing
JOURNAL
OF SURGICAL
Recombinant
RESEARCH
45, 145-153 (1988)
Basic Fibroblast
Growth
Factor Accelerates
Wound Healing’
GREGORY S. MCGEE, M.D., JEFFREY M. DAVIDSON, PH.D.,* ANNE BUCKLEY, PH.D.,* ANDREAS SOMMER, PH.D.,? STEPHENC. WOODWARD, M.D.,*
ANTONIOM.AQUINO,M.D.,*RONALDBARBOUR,M.D., AND ACHILLES A. DEMETRIOU, M.D., PH.D. Departments of Surgery and *Pathology, Vanderbilt University School of Medicine, and VA Medical Center, Nashville, Tennessee 37232, and TSynergen, Inc., Boulder, Colorado Presented at the Annual Meeting of the Association for Academic Surgery, Orlando, Florida, November l-4, 1987 Basic fibroblast growth factor (bFGF) stimulates extracellular matrix metabolism, growth, and movement of mesodennally derived cells. We have previously shown that collagen content in polyvinyl alcohol sponges increased after bFGF treatment. We hypothesized that bFGF-treated incisional wounds would heal more rapidly. After intraperitoneal pentobarbital anesthesia, male, 200- to 250-g, Sprague-Dawley rats (n = 27) each underwent two sets of paired, transverse, dorsal incisions closed with steel sutures. On Day 3 postwounding, 0.4 ml of bFGF (recombinant, 400 ng, Synergen) or normal saline was injected into one of each paired incisions. Animals were killed with ether on postwounding Days 5, 6, and 7 and their dorsal pelts were excised. Fresh or formalin-fixed wound strips were subjected to tensile strength measurements using a tensiometer. Breaking energy was calculated. Wound collagen content (hydroxyproline) was measured in wound-edge samples following hydrolysis using high-performance liquid chromatography. There was an overall significant increase in fresh wound tensile strength (13.7 + 1.06 vs 19.1 -t 1.99 g/mm, P < 0.01) and wound breaking energy (476 f 47 vs 747 f 76 mm’, P < 0.001) in bFGF-treated incisions. There was an increase in wound collagen content which was not statistically significant and there was no difference in fixed incisional tensile strength. Histologic examination showed better organization and maturation in bFGF wounds. Recombinant bFGF accelerates normal rat wound healing. This may be due to earlier accumulation of collagen and fibroblasts and/or to greater collagen crosslinking in bFGF-treated wounds. 0 1988 Academic
Press, Inc
More recently, Mustoe et al. [8] and our group [9] demonstrated accelerated incisional wound healing following topical application of recombinant transforming growth factor-p (TGFP). Basic fibroblast growth factor (bFGF) has been shown to stimulate extracellular matrix metabolism, growth, and migration of mesodermally derived cells [ 10, 1 I]. We hypothesized that recombinant bFGF would accelerate incisional wound healing and we carried out experiments utilizing a rat skin incision experimental model [ 121 to test our hypothesis.
INTRODUCTION
Wound repair is the result of complex interactions and well-coordinated biologic processes. Several polypeptide cytokines (growth factors) have been identified which participate in various phases of wound healing. Epidermal growth factor (EGF) is a potent mitogen for mesodermal and ectoderma1 cells in vitro [ 1] and has been shown to stimulate epithelial proliferation in fetal tissues [2] and to accelerate connective tissue organization in subcutaneously implanted polyvinyl alcohol sponges in adult rats [3-51. Platelet-derived growth factor (PDGF) has been shown to stimulate inflammatory cells and promote fibroblast migration [6, 71. ’ Supported in part by USPHS, NIH Grants AGO6528 and lRo1 Dk38763-02, the Veterans Administration, and Synergen, Inc.
MATERIALS
AND METHODS
Adult, male, Sprague-Dawley rats (200250 g) were acclimatized to our laboratory conditions for 5 days prior to use. They were individually housed in stainless-steel mesh 145
0022-4804/88 $1.50 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
146
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY
cages at constant temperature and relative humidity with a 12-hr night/day cycle, and fed commercial rat chow and water ad
libitum. Operative Procedure Animals were anesthetized with sodium pentobarbital (30 mg/kg, intraperitoneally) and their dorsal hair was clipped. The dorsal skin surface was painted with providone iodine and two pairs of 3-cm, full-thickness, transverse dorsal incisions were made with a scalpel using sterile technique. Hemostasis was achieved with gentle pressure, and the incisions were closed using simple interrupted 5-O stainless steel sutures (Ethicon) placed at 0.5~cm intervals. Rats were placed in individual cages and allowed to recover. On the third postwounding day, rats were anesthetized with ether and their incisions were carefully inspected. All wounds were injected with 0.4 ml solutions using a No. 25 gauge needle. Paired wounds were randomized to receive either 0.9% NaCl or bFGF (400 ng; recombinant; Synergen, Inc.). Rats were killed on post wounding Days 5 (N = 9), 6 (N = 9), and 7 (N = 9) using ether overdosage, and wound-edge samples (50 mg) were immediately taken from each incision for assessment of collagen metabolism. Sutures were removed, and three strips of skin with centrally located wounds were fashioned across each incision (Fig. 1). One strip (1 cm wide) was used in fresh tensile strength assessment. The second strip (1 cm wide) was placed in 10% buffered formalin for 72 hr to be used for evaluation of fixed tensile strength. The remaining strip (5 mm wide) was placed in formalin for histologic examination.
1988
Corp., Canton, MA) and a force was applied across the incision at a constant speed (crosshead speed, 5 cm/min). This force was graphically recorded (chart speed 50 cm/ min), and the point of maximal stress before wound separation was used as the incisional breaking strength (Fig. 2). These values were then normalized for incisional width, which varied from 8.5 to 11.0 mm. The resulting value was termed tensile strength (expressed as g/mm incisional width). Fixed weights were used for standards.
Breaking Energy Using the method of Gottrup [ 131, the breaking energy (BE) of fresh incisions was calculated. The area under the breaking strength curve to the point of incision disruption was calculated by computerlinked planimetry using a HIPAD digitizer (Houston Instruments, Austin, TX) interfaced with an Apple II+ Computer. BE was expressed as millimeters squared and normalized for incisional width (Fig. 2).
Histologic Preparation Formalin-fixed strips were stained using hematoxylin and eosin and Masson triDorsal
Rat Pelf
oCo”frol Compartsons
A~-P, Eo. ,:}
hfh
.bFGF
ternlIe *t,C”pth
Be-Be FomF.} filed t*n*i,*
Tensile Strength Assessment Fresh and fixed tensile strengths were assessed using techniques previously described [ 121. Briefly, strips with centrally located incisions were placed within a tensiometer (Model No. 1130, series No. 1900; Instron
werv$h
Co-0 hlltolo~y Go-G. 1 cdloqe” Confen, on0 Do-D*} *y”lhetlC ac+w,y HO-H.
Fxc. 1. This figure depicts the bilateral, transverse dorsal incisions used in this model. Use of this design eliminated variations in skin thickness and allowed each animal to serve as its own control.
MCGEE ET AL.: bFGF ACCELERATES
WOUND
Maximal
147
HEALING
load = breaking
q
Breokmg
strength
energy
gms
Time FIG. 2. The breaking strength of the incisions was defined graphically as the maximal force applied across the incisions before disruption. Using computer planimetry, the area under the breaking strength curve was calculated as an index of breaking energy.
chrome stains and examined by a single pathologist in a blind coded fashion. Wounds were examined for collagen organization, epidermal thickness, and vascularity.
Collagen Content and Synthetic Activity Wound-edge samples were prepared and assayed for hydroxyproline content by the method of Buckley et al. [ 141. Briefly, tissue was hydrolyzed in vacua in constantly boiling 6 N HCl (110°C). The hydrolysate was dried, taken up in 1 ml redrying solution (ethanol/water/TEA, 2:2:1) and 100 ~1 was redried in a nitrogen vacuum. A derivatization agent (20 ~1, ethanol/water/TEA/PITC, 7: 1: 1: 1) was added and the reaction was allowed to proceed for 20 min at room temperature. The samples were then lyophilized during centrifugation (Savant Speed Vat) to remove the reagents. Samples were diluted to 1 ml with buffer (5% acetonitrile in 5 mA4 disodium phosphate, pH 7.4), and a solution containing amino acid standards was dried and prepared in the same manner. Reversephase high-performance liquid chromatography (HPLC) was then performed on an Altex Ultrasphere ODS 5-mm column (4.6 by 15 cm) with a Beckman HPLC. The absorbance detector was set at 254 nm and connected via a Nelson Analytical 760 series interface to a IBM PC-AT computer, programmed with the Nelson 3600 series chromatography software.
In a similar experiment, 2 pg of bFGF was injected into rat incisions on the third postwounding day and tissue samples were removed at later time points for analysis of collagen synthetic activity. This was done using the method of Buckley et al. [14]. Minced tissue fragments were incubated in Dulbecco’s modified Eagle’s medium plus 1% dialyzed bovine serum, 50 mg/ml Na+ ascorbate, 44 mM NaHCOs, 25 mM Hepes, 100 U/ml penicillin, 100 mg/ml streptomycin, 64 pg/ml fi-aminoproprionitrile, and 0.25 rig/ml fungizone. After 30 min of preincubation at 37°C in a Gyrotary shaker bath, media were discarded and the minced fragments were rinsed and resuspended in fresh media plus 20 clCi/ml [3H]proline. Explants were incubated at 37°C in 95% air, 5% CO* for 3 hr. Media were aspirated off and the minced tissue fragments were washed with cold (4°C) phosphate-buffered saline (PBS) several times. After centrifugation, the tissue was resuspended in 3-5 ml of 0.5 M acetic acid containing 1 pg/ml pepstatin and 10 mM iodoacetic acid. Suspensions were homogenized at 4°C and the homogenates were extracted by slow rotation of sealed tubes along their long axis for 14 hr at 4°C. Aliquots of the homogenate were removed for amino acid analysis using HPLC; the remaining homogenate was clarified by centrifugation (25,OOOg, 4°C 20 min) and the supernatant was dialyzed against 0.1 M acetic acid. An aliquot of the supematant was used
148
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY
TT
q
T
p5.01
DS.001
Cl
I-IS
FGF Contrc ps.07
T
Fixed Tensile Strength
Tensile Strength
1988
Repair
Breaking Energy
Collagen Content
Parameter
FIG. 3. Effect of Recombinant FGF on Surface Wounds. This bar graph depicts the results when data were anlayzed as ratios, FGFcontrol, from each pair of incisions. Statistically significant results were seen in the tensile strength and breaking energy comparisons. Analysis was carried out with a one-tailed Mann-Whitney U test.
to quantitate collagen synthesis on the amount of radioactivity released into trichloroacetic acid-soluble material after limited digestion with bacterial collagenase. All animal studies were approved by our Institutional Animal Care Committee. Anima1 care complied with the Principles of Laboratory Animal Care (National Society for Medical Research) and the Guide for thk Care and Use of Laboratory Animals (National Academy of Sciences, NIH Publication No. 80-23, revised 1978). Statistical Analysis A one-tailed Mann-Whitney U test was used to assessdifferences from unity of ratios TABLE CUMULATIVE
of experimental/control data. Student’s paired t test was used when analyzing cumulative as well as daily paired data files. Results were expressed as means f standard errors of the means @EM) unless otherwise indicated. RESULTS
Fresh tensile strength, dependent upon both total collagen content and the amount of mature (crosslinked) collagen in wounds, was noted to be significantly greater in bFGF incisions than in saline-treated incisions (cumulative data, P < 0.01, Fig. 3; experimenml/control ratio, P < 0.0 1, Table 1). Analysis of data from separate days revealed the dif1 DATA
Tensile strength (g/mm)
bFGF saline P
Fresh
Fixed
Breaking energy (mm2)
Hydroxyproline content (pmoles/mg)
19.1 f 1.99 13.7 f 1.06 0.01
118 f 14.3 104 + 12.8 0.14
147 iI 15.9 416 zk 46.1 0.001
68.9 f 1.78 63.9 + 2.42 0.07
MCGEE ET AL.: bFGF ACCELERATES
WOUND
149
HEALING
ooy Day
5
FIG. 4. This bar graph depicts the fresh tensile strength results when analyzed on a daily basis.
ferences in fresh tensile strength between paired bFGF and control incisions to approach, though not exceed, significant levels on each day (Fig. 4). No significant differences in formalin-fixed tensile strength, a measure of total scleroprotein (crosslinked collagen) wound content, were found (Figs. 3 and 5, Table 1). Breaking energy, a measure of the extensibility of the incisions, was noted to be significantly greater in wounds injected with bFGF than after saline injection (cumulative data, P 6 0.001, Table 1; experimental/control ratio, P < 0.001, Fig. 3; separate day analysis-Day 5, P < 0.05; Day 6, P < 0.05; Day 7, P < 0.05, Fig. 6). No significant differences in total collagen (hydroxyproline) content were found between bFGF and saline-treated incisions, except on Day 5 of the separate day analysis (P < 0.05, Fig. 7).
FIG. 5. This bar graph depicts results of formalin-fixed tensile strength when analyzed on individual days.
7
FIG. 6. This figure demonstrates the significant increases in breaking energy seen in bF6F-treated incisions on Days 5, 6, and 7.
Collagen synthetic activity, assessed in a separate study using 2 pg bFGF injections, was not increased in bFGF-treated wounds on Days 7, 14, and 21 when compared to saline-treated incisions (Table 2). Histologic examination revealed the bFGF-treated wounds to show greater collagen organization, better realignment of the fibers of panniculous comosus, and a more mature epidermal layering than the control wounds. Representative photomicrographs of saline- and bFGF-treated wounds are shown (Fig. 8 and Fig. 9). DISCUSSION
Fibroblast growth factor, first isolated by Gospodarowiz in 1974 [ 151, is found in both an acidic and a basic form [ 161. Basic fibroblast growth factor, the predominant and more potent form, is a single-chain polypep
FIG. 7. This figure depicts the total collagen content of wound-edge samples analyzed using HPLC.
150
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. I, JULY
TABLE 2 EFFECT OF
Day 7 14 21 Note. x = 1.14
bFGF ON OHPro SYNTHESIS Ratio FGFxontrol 1.15 1.29 1.01
kSEM 0.41 0.31 0.85
k 0.43, ns.
tide found in macrophages and other mesodermally derived cells [ 17- 191. Participants in wound repair such as fibroblasts, vascular smooth muscle cells, and endothelial cells, proliferate upon exposure to bFGF in vitro [ 10, 201. Neovascularization of wound margins could be augmented by exposure to bFGF, which has been shown to have potent angiogenic activity [2 11. Studies by Davidson et al. [3] show increased granulation tissue formation in subcutaneously implanted polyvinyl alcohol sponges after injection with bFGF, possibly resulting from bFGF’s known ability to augment the production of extracellular matrix components (collagen, proteoglycans, and fibronectin) [ 1 I]. These studies strongly suggested that bFGF may enhance incisional wound healing. The present study demonstrated that a single intraincisional injection of bFGF accelerated wound healing, as manifested by an increase in wound tensile strength and breaking energy. Tensile strength has often been defined as a force per unit area. In our experiment, cross-sectional areas of wound edges were not calculated-rather, breaking strength measurements were corrected for incisional length. Thus, tensile strength was defined as force (breaking strength, g) per unit length (mm). The significant increases in fresh wound strip tensile strength in bFGF wounds, noted in analyses of both cumulative data and experimental/control ratios, could reflect increased collagen maturation (crosslinking) and/or increased total collagen content in bFGF-treated incisions. Analysis of fresh wound tensile strength on each specific day demonstrated that, although bFGF-
1988
treated incisions always had tensile strength greater than that of untreated wounds, the values approached but did not exceed the 95% confidence level when analyzed by Student’s paired t test. This finding may be due to the relatively wide standard deviations in the data derived from the bFGF-treated incisions, possibly resulting from uneven distribution of the injected bFGF and subsequent variations in local tissue concentrations. Formalin-fixation of tissue initiates cross linking of protein. The tensile strength of formalin-fixed incisions could thus be used as an index of total collagen present in these early wounds. No significant difference in fixed tensile strength was noted between experimental and control incisions. This finding, and the lack of a significant increase in collagen content in bFGF-treated wounds, suggests that the most tenable explanation for the increased tensile strength in bFGFtreated wounds must be primarily the presence of more mature, crosslinked collagen in the incisions. A small increase in wound collagen content, however, cannot be excluded; such an increase would be difficult to detect because of sampling errors. Only minor differences in incisional collagen synthetic activity were appreciated on Days 7, 14, and 2 1 in a similar experiment using a 2-pg bFGF injection. This finding was surprisingstudies have shown increased synthetic activity of fibroblasts after exposure to bFGF [22, 231. However, our result may be due to local variation in tissue bFGF concentrations, to tissue sampling errors, or to the high background collagen synthetic activity of normal skin which would make changes induced by bFGF at the wound margins difficult to detect. Experiments are needed to examine directly wound collagen synthesis using labeled precursors. Gottrup, in studying incisional wounds in rat stomach, suggested that breaking energy (BE) calculations from breaking strength curves provided a more accurate and perhaps more physiologic assessment of wound strength than tensile strength analysis alone [ 131. We thus determined the BE of incisions
FIG. 8. This photomicrograph shows the effect of saline injection on incisional healing at Day 7. FIG. 9. This photomicrograph shows a more mature epidermal layer and a more organized dermal collagen response after bFGF injection.
151
152
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY 1988
in this study, and noted highly significant increases in BE in wounds treated with bFGF. The differences in BE could have resulted from increased wound “elasticity” and extensibility, as a function of increased extracellular matrix accumulation in bFGFtreated wounds. Increased elastin content of experimental wounds could also explain the BE differences. We have presented data demonstrating that a single application of bFGF can augment wound repair. However, the physiologic role of bFGF in normal wound healing has not yet been established. Studies are in progress to determine whether treatment of wounds with polyclonal anti-bFGF antiserum will inhibit accumulation of cells, DNA, and collagen and result in impaired wound healing. In summary, we demonstrated a modest but significant enhancement of normal skin incisional wound healing in rats, manifested by an increase in fresh wound tensile strength, wound breaking energy, and evidence of accelerated histologic wound maturation. This effect appears to be primarily due to an increase in collagen maturation (cross linkage). Experimental studies are in progress to further define the mechanism of the bFGF effect on wound healing and to examine the effect of bFGF in animal models of impaired wound healing (i.e., diabetes, steroids, and injury). REFERENCES 1. Carpenter, G., and Cohen, S. Epidermal growth factor. Annu. Rev. Biochem. 48: 193, 1979. 2. Sundell, H., Serenius, R. S., Barthe, P., Friedmen, A., Kanarek, K. S., Escobdo, M. B., Orth, D. N., and Stahlman, M. T. The effect of EGF on fetal lamb lung maturation. Pediatr. Rex 9: 37 1, 1975. 3. Davidson, J. M., Klagsbrun, M., Hill, K. E., Buckley, A., Sullivan, R., Brewer, P. S., and Woodward, S. C. Accelerated wound repair, cell proliferation, and collagen accumulation are produced by a cartilage-derived growth factor. J. Cell Biol. 100: 12 19, 1975. 4. Laats, M., Niinikosk, J., Lebel, L., and Gerdin, B. Stimulation of wound healing by epidermal growth factor. Ann. Surg. 203: 379, 1985.
5. Buckley, A., Davidson, J. M., Kamerath, C. D., and Woodward, S. C. Epidermal growth increases granulation tissue formation dose dependently. J. Surg. Rex 43: 322, 1987. 6. Senior, R. M., Griffin, G. L., Huang, J. S., Waltz, D. A., and Deuel, T. F. Chemotactic activity of platelet (Y granule proteins for fibroblasts. J. Cell Biol. 96: 382, 1983. 7. Bauer, E. A., Cooper, T. W., Huang, J. S., Altman, J., and Deuel, T. F. Stimulation of in vitro human skin collagenese expression by platelet-derived growth factor. Proc. Nat/. Acad. Sci. USA 82: 2, 1985. 8. Mustoe, T. A., Pierce, G. F., Thomason, A., Gramates, P., Spom, M. B., and Deuel, T. F. Accelerated healing of incisional wounds in rats induced by transforming growth factor-& Science 237: 1333, 1987. 9. McGee, G. S., Davidson, J. M., Buckley, A.. Aguino, A. M., Woodward, S. C., and Demetriou, A. A. Recombinant 6 transforming growth factor accelerates incisional wound healing. Submitted for publication. 10. Gospodarowicz, D., Massoglia, S., Cheng, J., Lui, G. M., and Bohlen, P. Isolation of pituitary fibroblast growth factor by fast protein liquid chromatography (FGLC). Partial chemical and biological characterization. J. Cell. Physiol. 122: 323, 1985. 11. Vlodavsky, I., Johnson, L. K., Greenburg, G., and Gospodarowicz, D. Vascular endothelial cells maintained in the absence of libroblast growth factor undergo structural and functional alterations that are incompatible with their in vivo differentiated properties. J. Cell Biol. 83: 468, 1983. 12. Levenson, S. M., Crowley, L. V., Geever, E. F., Rosen, H., and Berard, C. W. Some studies of wound healing: Experimental methods, effect of ascorbic acid, and effect of deuterium oxide. J. Trauma 4: 534, 1964. 13. Gottrup, R. Healing of incisions, wounds in stomach and duodenum. Amer. J. Surg. 146: 296, 1980. 14. Buckley, A., Hill, K. E., and Davidson, J. M. Methods for the study of collagen metabolism. In Methods in Enzymology, Academic Press, San Diego, in press. 15. Gospodarowicz, D. Localization of a fibroblast growth factor and its effect alone and with hydrocortisone and T cell growth. Nature (London) 249: 123, 1974. 16. Gospodarowicz, D., Weseman, J., and Moran, J. Presence in brain of a mitogenic agent promoting proliferation of myoblasts in low density culture. Nature (London) 256: 216, 1975. 17. Esch, F., Baird, A., Lung, N., et al. Primary structure of bovine pituitary basic libroblast growth factor (FGF) and comparison with the amino-terminal sequence of bovine brain acidic FGF. Proc. Natl. Acad. Sci. USA 85: 6507, 1985.
MCGEE
ET AL.: bFGF ACCELERATES
18. Baird, A., Mormede, P., and Bohlen, P. Immunoreactive fibroblast growth factor in cells of peritoneal exudate suggest its identity with macrophage-derived growth factor. Biochem. Biophys. Res. Commun. 126: 358, 1985. 19. Gospodarowicz, D., Neufeld, G., and Schweigerer, L. Fibroblast growth factor. Mol. Cell. Endocrinol. 46: 187, 1986. 20. Gospodarowicz, D., and Moran, J. S. Mitogenic effect of fibroblast growth factor on early passage cultures of human and marine fibroblasts. J. Cell Biol. 66: 451, 1975. 21. Gospodarowicz, D., Chone, J., Lui, G. M., Baird, A., and Bohlen, P. Isolation of brain fibroblast growth factor by heparin-spharose affinity chro-
WOUND
HEALING
153
matograph: Identity with pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA 81: 6963, 1984. 22. Buntrock, P., Jentzsch, K. D., and Heder, G. Stimulation of wound healing using brain extract with fibroblast growth factor activity. I. Quantitative and biochemical studies into the formation of granulation tissue. Exp. Pathol. 21: 46, 1982. 23. Davidson, J. M., Buckley, A., Woodward, S. C., Nichols, W. K., McGee, G. S., and Demetriou, A. A. Mechanisms of accelerated wound repair using epidermal growth factor and basic fibroblast growth factor. In Barbul, A., Pines, E., Caldwell, M., and Hunt, T. K. (Eds.), Growth Factors and Other Aspects of Wound Healing: Biological and Clinical Implications, Alan R. Liss, Inc., New York, 1987. pp. 63-75.
OF SURGICAL
Recombinant
RESEARCH
45, 145-153 (1988)
Basic Fibroblast
Growth
Factor Accelerates
Wound Healing’
GREGORY S. MCGEE, M.D., JEFFREY M. DAVIDSON, PH.D.,* ANNE BUCKLEY, PH.D.,* ANDREAS SOMMER, PH.D.,? STEPHENC. WOODWARD, M.D.,*
ANTONIOM.AQUINO,M.D.,*RONALDBARBOUR,M.D., AND ACHILLES A. DEMETRIOU, M.D., PH.D. Departments of Surgery and *Pathology, Vanderbilt University School of Medicine, and VA Medical Center, Nashville, Tennessee 37232, and TSynergen, Inc., Boulder, Colorado Presented at the Annual Meeting of the Association for Academic Surgery, Orlando, Florida, November l-4, 1987 Basic fibroblast growth factor (bFGF) stimulates extracellular matrix metabolism, growth, and movement of mesodennally derived cells. We have previously shown that collagen content in polyvinyl alcohol sponges increased after bFGF treatment. We hypothesized that bFGF-treated incisional wounds would heal more rapidly. After intraperitoneal pentobarbital anesthesia, male, 200- to 250-g, Sprague-Dawley rats (n = 27) each underwent two sets of paired, transverse, dorsal incisions closed with steel sutures. On Day 3 postwounding, 0.4 ml of bFGF (recombinant, 400 ng, Synergen) or normal saline was injected into one of each paired incisions. Animals were killed with ether on postwounding Days 5, 6, and 7 and their dorsal pelts were excised. Fresh or formalin-fixed wound strips were subjected to tensile strength measurements using a tensiometer. Breaking energy was calculated. Wound collagen content (hydroxyproline) was measured in wound-edge samples following hydrolysis using high-performance liquid chromatography. There was an overall significant increase in fresh wound tensile strength (13.7 + 1.06 vs 19.1 -t 1.99 g/mm, P < 0.01) and wound breaking energy (476 f 47 vs 747 f 76 mm’, P < 0.001) in bFGF-treated incisions. There was an increase in wound collagen content which was not statistically significant and there was no difference in fixed incisional tensile strength. Histologic examination showed better organization and maturation in bFGF wounds. Recombinant bFGF accelerates normal rat wound healing. This may be due to earlier accumulation of collagen and fibroblasts and/or to greater collagen crosslinking in bFGF-treated wounds. 0 1988 Academic
Press, Inc
More recently, Mustoe et al. [8] and our group [9] demonstrated accelerated incisional wound healing following topical application of recombinant transforming growth factor-p (TGFP). Basic fibroblast growth factor (bFGF) has been shown to stimulate extracellular matrix metabolism, growth, and migration of mesodermally derived cells [ 10, 1 I]. We hypothesized that recombinant bFGF would accelerate incisional wound healing and we carried out experiments utilizing a rat skin incision experimental model [ 121 to test our hypothesis.
INTRODUCTION
Wound repair is the result of complex interactions and well-coordinated biologic processes. Several polypeptide cytokines (growth factors) have been identified which participate in various phases of wound healing. Epidermal growth factor (EGF) is a potent mitogen for mesodermal and ectoderma1 cells in vitro [ 1] and has been shown to stimulate epithelial proliferation in fetal tissues [2] and to accelerate connective tissue organization in subcutaneously implanted polyvinyl alcohol sponges in adult rats [3-51. Platelet-derived growth factor (PDGF) has been shown to stimulate inflammatory cells and promote fibroblast migration [6, 71. ’ Supported in part by USPHS, NIH Grants AGO6528 and lRo1 Dk38763-02, the Veterans Administration, and Synergen, Inc.
MATERIALS
AND METHODS
Adult, male, Sprague-Dawley rats (200250 g) were acclimatized to our laboratory conditions for 5 days prior to use. They were individually housed in stainless-steel mesh 145
0022-4804/88 $1.50 Copyright 0 1988 by Academic Press, Inc. All rights of reproduction in any form reserved.
146
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY
cages at constant temperature and relative humidity with a 12-hr night/day cycle, and fed commercial rat chow and water ad
libitum. Operative Procedure Animals were anesthetized with sodium pentobarbital (30 mg/kg, intraperitoneally) and their dorsal hair was clipped. The dorsal skin surface was painted with providone iodine and two pairs of 3-cm, full-thickness, transverse dorsal incisions were made with a scalpel using sterile technique. Hemostasis was achieved with gentle pressure, and the incisions were closed using simple interrupted 5-O stainless steel sutures (Ethicon) placed at 0.5~cm intervals. Rats were placed in individual cages and allowed to recover. On the third postwounding day, rats were anesthetized with ether and their incisions were carefully inspected. All wounds were injected with 0.4 ml solutions using a No. 25 gauge needle. Paired wounds were randomized to receive either 0.9% NaCl or bFGF (400 ng; recombinant; Synergen, Inc.). Rats were killed on post wounding Days 5 (N = 9), 6 (N = 9), and 7 (N = 9) using ether overdosage, and wound-edge samples (50 mg) were immediately taken from each incision for assessment of collagen metabolism. Sutures were removed, and three strips of skin with centrally located wounds were fashioned across each incision (Fig. 1). One strip (1 cm wide) was used in fresh tensile strength assessment. The second strip (1 cm wide) was placed in 10% buffered formalin for 72 hr to be used for evaluation of fixed tensile strength. The remaining strip (5 mm wide) was placed in formalin for histologic examination.
1988
Corp., Canton, MA) and a force was applied across the incision at a constant speed (crosshead speed, 5 cm/min). This force was graphically recorded (chart speed 50 cm/ min), and the point of maximal stress before wound separation was used as the incisional breaking strength (Fig. 2). These values were then normalized for incisional width, which varied from 8.5 to 11.0 mm. The resulting value was termed tensile strength (expressed as g/mm incisional width). Fixed weights were used for standards.
Breaking Energy Using the method of Gottrup [ 131, the breaking energy (BE) of fresh incisions was calculated. The area under the breaking strength curve to the point of incision disruption was calculated by computerlinked planimetry using a HIPAD digitizer (Houston Instruments, Austin, TX) interfaced with an Apple II+ Computer. BE was expressed as millimeters squared and normalized for incisional width (Fig. 2).
Histologic Preparation Formalin-fixed strips were stained using hematoxylin and eosin and Masson triDorsal
Rat Pelf
oCo”frol Compartsons
A~-P, Eo. ,:}
hfh
.bFGF
ternlIe *t,C”pth
Be-Be FomF.} filed t*n*i,*
Tensile Strength Assessment Fresh and fixed tensile strengths were assessed using techniques previously described [ 121. Briefly, strips with centrally located incisions were placed within a tensiometer (Model No. 1130, series No. 1900; Instron
werv$h
Co-0 hlltolo~y Go-G. 1 cdloqe” Confen, on0 Do-D*} *y”lhetlC ac+w,y HO-H.
Fxc. 1. This figure depicts the bilateral, transverse dorsal incisions used in this model. Use of this design eliminated variations in skin thickness and allowed each animal to serve as its own control.
MCGEE ET AL.: bFGF ACCELERATES
WOUND
Maximal
147
HEALING
load = breaking
q
Breokmg
strength
energy
gms
Time FIG. 2. The breaking strength of the incisions was defined graphically as the maximal force applied across the incisions before disruption. Using computer planimetry, the area under the breaking strength curve was calculated as an index of breaking energy.
chrome stains and examined by a single pathologist in a blind coded fashion. Wounds were examined for collagen organization, epidermal thickness, and vascularity.
Collagen Content and Synthetic Activity Wound-edge samples were prepared and assayed for hydroxyproline content by the method of Buckley et al. [ 141. Briefly, tissue was hydrolyzed in vacua in constantly boiling 6 N HCl (110°C). The hydrolysate was dried, taken up in 1 ml redrying solution (ethanol/water/TEA, 2:2:1) and 100 ~1 was redried in a nitrogen vacuum. A derivatization agent (20 ~1, ethanol/water/TEA/PITC, 7: 1: 1: 1) was added and the reaction was allowed to proceed for 20 min at room temperature. The samples were then lyophilized during centrifugation (Savant Speed Vat) to remove the reagents. Samples were diluted to 1 ml with buffer (5% acetonitrile in 5 mA4 disodium phosphate, pH 7.4), and a solution containing amino acid standards was dried and prepared in the same manner. Reversephase high-performance liquid chromatography (HPLC) was then performed on an Altex Ultrasphere ODS 5-mm column (4.6 by 15 cm) with a Beckman HPLC. The absorbance detector was set at 254 nm and connected via a Nelson Analytical 760 series interface to a IBM PC-AT computer, programmed with the Nelson 3600 series chromatography software.
In a similar experiment, 2 pg of bFGF was injected into rat incisions on the third postwounding day and tissue samples were removed at later time points for analysis of collagen synthetic activity. This was done using the method of Buckley et al. [14]. Minced tissue fragments were incubated in Dulbecco’s modified Eagle’s medium plus 1% dialyzed bovine serum, 50 mg/ml Na+ ascorbate, 44 mM NaHCOs, 25 mM Hepes, 100 U/ml penicillin, 100 mg/ml streptomycin, 64 pg/ml fi-aminoproprionitrile, and 0.25 rig/ml fungizone. After 30 min of preincubation at 37°C in a Gyrotary shaker bath, media were discarded and the minced fragments were rinsed and resuspended in fresh media plus 20 clCi/ml [3H]proline. Explants were incubated at 37°C in 95% air, 5% CO* for 3 hr. Media were aspirated off and the minced tissue fragments were washed with cold (4°C) phosphate-buffered saline (PBS) several times. After centrifugation, the tissue was resuspended in 3-5 ml of 0.5 M acetic acid containing 1 pg/ml pepstatin and 10 mM iodoacetic acid. Suspensions were homogenized at 4°C and the homogenates were extracted by slow rotation of sealed tubes along their long axis for 14 hr at 4°C. Aliquots of the homogenate were removed for amino acid analysis using HPLC; the remaining homogenate was clarified by centrifugation (25,OOOg, 4°C 20 min) and the supernatant was dialyzed against 0.1 M acetic acid. An aliquot of the supematant was used
148
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY
TT
q
T
p5.01
DS.001
Cl
I-IS
FGF Contrc ps.07
T
Fixed Tensile Strength
Tensile Strength
1988
Repair
Breaking Energy
Collagen Content
Parameter
FIG. 3. Effect of Recombinant FGF on Surface Wounds. This bar graph depicts the results when data were anlayzed as ratios, FGFcontrol, from each pair of incisions. Statistically significant results were seen in the tensile strength and breaking energy comparisons. Analysis was carried out with a one-tailed Mann-Whitney U test.
to quantitate collagen synthesis on the amount of radioactivity released into trichloroacetic acid-soluble material after limited digestion with bacterial collagenase. All animal studies were approved by our Institutional Animal Care Committee. Anima1 care complied with the Principles of Laboratory Animal Care (National Society for Medical Research) and the Guide for thk Care and Use of Laboratory Animals (National Academy of Sciences, NIH Publication No. 80-23, revised 1978). Statistical Analysis A one-tailed Mann-Whitney U test was used to assessdifferences from unity of ratios TABLE CUMULATIVE
of experimental/control data. Student’s paired t test was used when analyzing cumulative as well as daily paired data files. Results were expressed as means f standard errors of the means @EM) unless otherwise indicated. RESULTS
Fresh tensile strength, dependent upon both total collagen content and the amount of mature (crosslinked) collagen in wounds, was noted to be significantly greater in bFGF incisions than in saline-treated incisions (cumulative data, P < 0.01, Fig. 3; experimenml/control ratio, P < 0.0 1, Table 1). Analysis of data from separate days revealed the dif1 DATA
Tensile strength (g/mm)
bFGF saline P
Fresh
Fixed
Breaking energy (mm2)
Hydroxyproline content (pmoles/mg)
19.1 f 1.99 13.7 f 1.06 0.01
118 f 14.3 104 + 12.8 0.14
147 iI 15.9 416 zk 46.1 0.001
68.9 f 1.78 63.9 + 2.42 0.07
MCGEE ET AL.: bFGF ACCELERATES
WOUND
149
HEALING
ooy Day
5
FIG. 4. This bar graph depicts the fresh tensile strength results when analyzed on a daily basis.
ferences in fresh tensile strength between paired bFGF and control incisions to approach, though not exceed, significant levels on each day (Fig. 4). No significant differences in formalin-fixed tensile strength, a measure of total scleroprotein (crosslinked collagen) wound content, were found (Figs. 3 and 5, Table 1). Breaking energy, a measure of the extensibility of the incisions, was noted to be significantly greater in wounds injected with bFGF than after saline injection (cumulative data, P 6 0.001, Table 1; experimental/control ratio, P < 0.001, Fig. 3; separate day analysis-Day 5, P < 0.05; Day 6, P < 0.05; Day 7, P < 0.05, Fig. 6). No significant differences in total collagen (hydroxyproline) content were found between bFGF and saline-treated incisions, except on Day 5 of the separate day analysis (P < 0.05, Fig. 7).
FIG. 5. This bar graph depicts results of formalin-fixed tensile strength when analyzed on individual days.
7
FIG. 6. This figure demonstrates the significant increases in breaking energy seen in bF6F-treated incisions on Days 5, 6, and 7.
Collagen synthetic activity, assessed in a separate study using 2 pg bFGF injections, was not increased in bFGF-treated wounds on Days 7, 14, and 21 when compared to saline-treated incisions (Table 2). Histologic examination revealed the bFGF-treated wounds to show greater collagen organization, better realignment of the fibers of panniculous comosus, and a more mature epidermal layering than the control wounds. Representative photomicrographs of saline- and bFGF-treated wounds are shown (Fig. 8 and Fig. 9). DISCUSSION
Fibroblast growth factor, first isolated by Gospodarowiz in 1974 [ 151, is found in both an acidic and a basic form [ 161. Basic fibroblast growth factor, the predominant and more potent form, is a single-chain polypep
FIG. 7. This figure depicts the total collagen content of wound-edge samples analyzed using HPLC.
150
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. I, JULY
TABLE 2 EFFECT OF
Day 7 14 21 Note. x = 1.14
bFGF ON OHPro SYNTHESIS Ratio FGFxontrol 1.15 1.29 1.01
kSEM 0.41 0.31 0.85
k 0.43, ns.
tide found in macrophages and other mesodermally derived cells [ 17- 191. Participants in wound repair such as fibroblasts, vascular smooth muscle cells, and endothelial cells, proliferate upon exposure to bFGF in vitro [ 10, 201. Neovascularization of wound margins could be augmented by exposure to bFGF, which has been shown to have potent angiogenic activity [2 11. Studies by Davidson et al. [3] show increased granulation tissue formation in subcutaneously implanted polyvinyl alcohol sponges after injection with bFGF, possibly resulting from bFGF’s known ability to augment the production of extracellular matrix components (collagen, proteoglycans, and fibronectin) [ 1 I]. These studies strongly suggested that bFGF may enhance incisional wound healing. The present study demonstrated that a single intraincisional injection of bFGF accelerated wound healing, as manifested by an increase in wound tensile strength and breaking energy. Tensile strength has often been defined as a force per unit area. In our experiment, cross-sectional areas of wound edges were not calculated-rather, breaking strength measurements were corrected for incisional length. Thus, tensile strength was defined as force (breaking strength, g) per unit length (mm). The significant increases in fresh wound strip tensile strength in bFGF wounds, noted in analyses of both cumulative data and experimental/control ratios, could reflect increased collagen maturation (crosslinking) and/or increased total collagen content in bFGF-treated incisions. Analysis of fresh wound tensile strength on each specific day demonstrated that, although bFGF-
1988
treated incisions always had tensile strength greater than that of untreated wounds, the values approached but did not exceed the 95% confidence level when analyzed by Student’s paired t test. This finding may be due to the relatively wide standard deviations in the data derived from the bFGF-treated incisions, possibly resulting from uneven distribution of the injected bFGF and subsequent variations in local tissue concentrations. Formalin-fixation of tissue initiates cross linking of protein. The tensile strength of formalin-fixed incisions could thus be used as an index of total collagen present in these early wounds. No significant difference in fixed tensile strength was noted between experimental and control incisions. This finding, and the lack of a significant increase in collagen content in bFGF-treated wounds, suggests that the most tenable explanation for the increased tensile strength in bFGFtreated wounds must be primarily the presence of more mature, crosslinked collagen in the incisions. A small increase in wound collagen content, however, cannot be excluded; such an increase would be difficult to detect because of sampling errors. Only minor differences in incisional collagen synthetic activity were appreciated on Days 7, 14, and 2 1 in a similar experiment using a 2-pg bFGF injection. This finding was surprisingstudies have shown increased synthetic activity of fibroblasts after exposure to bFGF [22, 231. However, our result may be due to local variation in tissue bFGF concentrations, to tissue sampling errors, or to the high background collagen synthetic activity of normal skin which would make changes induced by bFGF at the wound margins difficult to detect. Experiments are needed to examine directly wound collagen synthesis using labeled precursors. Gottrup, in studying incisional wounds in rat stomach, suggested that breaking energy (BE) calculations from breaking strength curves provided a more accurate and perhaps more physiologic assessment of wound strength than tensile strength analysis alone [ 131. We thus determined the BE of incisions
FIG. 8. This photomicrograph shows the effect of saline injection on incisional healing at Day 7. FIG. 9. This photomicrograph shows a more mature epidermal layer and a more organized dermal collagen response after bFGF injection.
151
152
JOURNAL
OF SURGICAL
RESEARCH: VOL. 45, NO. 1, JULY 1988
in this study, and noted highly significant increases in BE in wounds treated with bFGF. The differences in BE could have resulted from increased wound “elasticity” and extensibility, as a function of increased extracellular matrix accumulation in bFGFtreated wounds. Increased elastin content of experimental wounds could also explain the BE differences. We have presented data demonstrating that a single application of bFGF can augment wound repair. However, the physiologic role of bFGF in normal wound healing has not yet been established. Studies are in progress to determine whether treatment of wounds with polyclonal anti-bFGF antiserum will inhibit accumulation of cells, DNA, and collagen and result in impaired wound healing. In summary, we demonstrated a modest but significant enhancement of normal skin incisional wound healing in rats, manifested by an increase in fresh wound tensile strength, wound breaking energy, and evidence of accelerated histologic wound maturation. This effect appears to be primarily due to an increase in collagen maturation (cross linkage). Experimental studies are in progress to further define the mechanism of the bFGF effect on wound healing and to examine the effect of bFGF in animal models of impaired wound healing (i.e., diabetes, steroids, and injury). REFERENCES 1. Carpenter, G., and Cohen, S. Epidermal growth factor. Annu. Rev. Biochem. 48: 193, 1979. 2. Sundell, H., Serenius, R. S., Barthe, P., Friedmen, A., Kanarek, K. S., Escobdo, M. B., Orth, D. N., and Stahlman, M. T. The effect of EGF on fetal lamb lung maturation. Pediatr. Rex 9: 37 1, 1975. 3. Davidson, J. M., Klagsbrun, M., Hill, K. E., Buckley, A., Sullivan, R., Brewer, P. S., and Woodward, S. C. Accelerated wound repair, cell proliferation, and collagen accumulation are produced by a cartilage-derived growth factor. J. Cell Biol. 100: 12 19, 1975. 4. Laats, M., Niinikosk, J., Lebel, L., and Gerdin, B. Stimulation of wound healing by epidermal growth factor. Ann. Surg. 203: 379, 1985.
5. Buckley, A., Davidson, J. M., Kamerath, C. D., and Woodward, S. C. Epidermal growth increases granulation tissue formation dose dependently. J. Surg. Rex 43: 322, 1987. 6. Senior, R. M., Griffin, G. L., Huang, J. S., Waltz, D. A., and Deuel, T. F. Chemotactic activity of platelet (Y granule proteins for fibroblasts. J. Cell Biol. 96: 382, 1983. 7. Bauer, E. A., Cooper, T. W., Huang, J. S., Altman, J., and Deuel, T. F. Stimulation of in vitro human skin collagenese expression by platelet-derived growth factor. Proc. Nat/. Acad. Sci. USA 82: 2, 1985. 8. Mustoe, T. A., Pierce, G. F., Thomason, A., Gramates, P., Spom, M. B., and Deuel, T. F. Accelerated healing of incisional wounds in rats induced by transforming growth factor-& Science 237: 1333, 1987. 9. McGee, G. S., Davidson, J. M., Buckley, A.. Aguino, A. M., Woodward, S. C., and Demetriou, A. A. Recombinant 6 transforming growth factor accelerates incisional wound healing. Submitted for publication. 10. Gospodarowicz, D., Massoglia, S., Cheng, J., Lui, G. M., and Bohlen, P. Isolation of pituitary fibroblast growth factor by fast protein liquid chromatography (FGLC). Partial chemical and biological characterization. J. Cell. Physiol. 122: 323, 1985. 11. Vlodavsky, I., Johnson, L. K., Greenburg, G., and Gospodarowicz, D. Vascular endothelial cells maintained in the absence of libroblast growth factor undergo structural and functional alterations that are incompatible with their in vivo differentiated properties. J. Cell Biol. 83: 468, 1983. 12. Levenson, S. M., Crowley, L. V., Geever, E. F., Rosen, H., and Berard, C. W. Some studies of wound healing: Experimental methods, effect of ascorbic acid, and effect of deuterium oxide. J. Trauma 4: 534, 1964. 13. Gottrup, R. Healing of incisions, wounds in stomach and duodenum. Amer. J. Surg. 146: 296, 1980. 14. Buckley, A., Hill, K. E., and Davidson, J. M. Methods for the study of collagen metabolism. In Methods in Enzymology, Academic Press, San Diego, in press. 15. Gospodarowicz, D. Localization of a fibroblast growth factor and its effect alone and with hydrocortisone and T cell growth. Nature (London) 249: 123, 1974. 16. Gospodarowicz, D., Weseman, J., and Moran, J. Presence in brain of a mitogenic agent promoting proliferation of myoblasts in low density culture. Nature (London) 256: 216, 1975. 17. Esch, F., Baird, A., Lung, N., et al. Primary structure of bovine pituitary basic libroblast growth factor (FGF) and comparison with the amino-terminal sequence of bovine brain acidic FGF. Proc. Natl. Acad. Sci. USA 85: 6507, 1985.
MCGEE
ET AL.: bFGF ACCELERATES
18. Baird, A., Mormede, P., and Bohlen, P. Immunoreactive fibroblast growth factor in cells of peritoneal exudate suggest its identity with macrophage-derived growth factor. Biochem. Biophys. Res. Commun. 126: 358, 1985. 19. Gospodarowicz, D., Neufeld, G., and Schweigerer, L. Fibroblast growth factor. Mol. Cell. Endocrinol. 46: 187, 1986. 20. Gospodarowicz, D., and Moran, J. S. Mitogenic effect of fibroblast growth factor on early passage cultures of human and marine fibroblasts. J. Cell Biol. 66: 451, 1975. 21. Gospodarowicz, D., Chone, J., Lui, G. M., Baird, A., and Bohlen, P. Isolation of brain fibroblast growth factor by heparin-spharose affinity chro-
WOUND
HEALING
153
matograph: Identity with pituitary fibroblast growth factor. Proc. Natl. Acad. Sci. USA 81: 6963, 1984. 22. Buntrock, P., Jentzsch, K. D., and Heder, G. Stimulation of wound healing using brain extract with fibroblast growth factor activity. I. Quantitative and biochemical studies into the formation of granulation tissue. Exp. Pathol. 21: 46, 1982. 23. Davidson, J. M., Buckley, A., Woodward, S. C., Nichols, W. K., McGee, G. S., and Demetriou, A. A. Mechanisms of accelerated wound repair using epidermal growth factor and basic fibroblast growth factor. In Barbul, A., Pines, E., Caldwell, M., and Hunt, T. K. (Eds.), Growth Factors and Other Aspects of Wound Healing: Biological and Clinical Implications, Alan R. Liss, Inc., New York, 1987. pp. 63-75.