Scar formation is the result of excessive wound-healing and leads to a poor functional outcome after trauma and surgery. In hand surgery, flexor tendon repair can lead to formation of adhesions between the tendon and its surrounding fibro-osseous sheath, decreasing the postoperative range of motion and hand function. The ability to modulate the tendon-healing process, thereby reducing adhesions and increasing the postoperative range of motion while not compromising the strength of the repair, would be of great clinical value.
Biochemical manipulation of flexor tendon adhesion formation has led to few clinical applications thus far1-5. Use of synthetic materials such as nylon and Teflon to create artificial sheaths has not proved successful6. In addition, agents such as antihistamines, steroids, dimethyl sulfoxide, beta-aminoproprionitrile, and hyaluronic acid have not significantly decreased the amount of postoperative adhesions7-12.
In the past decade, wound-healing research has led to the characterization of growth factors and the description of their role in tissue repair13,14. Transforming growth factor-beta (TGF-β) is a cytokine with numerous biologic activities related to wound-healing, including fibroblast and macrophage recruitment, stimulation of collagen production, downregulation of proteinase activity, and increases in metalloproteinase inhibitor activity15. There are three mammalian isoforms of TGF-β: TGF-β1, TGF-β2, and TGF-β3. All three isoforms are potentially produced by most cells active in wound-healing, with platelets being a major contributor. Three distinct classes of TGF-β cell-membrane receptors have been identified. Two are receptor-signaling molecules of the serine/threonine kinase mechanism, termed receptor I and II. Betaglycan, also known as TGF-β receptor III, is a membrane proteoglycan that presents ligand to TGF-β receptor II16.
Excessive production of TGF-β has been linked to fibrotic diseases. For example, enhanced TGF-β1 expression and an associated increase in the production of collagen I, III, and VI have been documented in tissues of patients with systemic sclerosis, postburn hypertrophic scar tissue, and keloids17-20. Conversely, inhibition of TGF-β has been shown to decrease collagen deposition and scarring. For example, the application of neutralizing antibodies to TGF-β in rat incisional wounds successfully reduced cutaneous scarring21.
In this study, we investigated the effects of two potential natural inhibitors of TGF-β, decorin and mannose-6-phosphate. These natural inhibitors have structural similarity to beta-glycan (TGF-β type-III receptor) and may competitively bind TGF-β. Decorin belongs to a family of proteoglycans called small leucine-rich proteoglycans. Border et al. were able to inhibit scarring in the rat kidney with use of decorin22. Recent studies have indicated that collagen-bound decorin may sequester TGF-β in the extracellular matrix, thus reducing its effects23. The infusion of decorin has also reduced levels of scarring at the site of brain injury24. Mannose-6-phosphate, a carbohydrate molecule, has been shown to reduce activation of latent TGF-β25. In another experiment, mannose-6-phosphate reduced scarring in incisional wounds in rats26.
We tested these inhibitors of TGF-β in previously established in vitro and in vivo models of flexor tendon wound-healing. Our in vitro system has been used to demonstrate enhanced collagen-I production by independent tendon sheath, epitenon, and endotenon cell lines in response to TGF-β as well as reduced collagen production with the introduction of neutralizing antibodies to TGF-β27,28. Expression of TGF-β1 and TGF-β receptors has been demonstrated during tendon-healing in our in vivo rabbit model29,30. This rabbit model has also been used to show an improved range of motion after flexor tendon repair with intraoperative infiltration of neutralizing antibody to TGF-β31.
Materials and Methods
In Vitro Studies
All rabbit experiments were performed according to protocols established by our institution and guidelines from the National Institutes of Health. An adult male New Zealand White rabbit (weighing 4.0 to 4.5 kg) was killed with an intravenous injection. The flexor digitorum profundus equivalents were identified and isolated in the forepaws. The tendons, along with the tendon sheaths, were transected and excised with use of 3.5× loupe magnification.
Cells from the flexor tendon sheath, epitenon, and endotenon were isolated from the procured specimens and cultured as previously described27. The intact flexor tendons and tendon sheaths were separated by dissection under magnification. The tendon sheaths were digested with 0.5% collagenase (Sigma, St. Louis, Missouri) in 20 mmol/L of HEPES buffered Hanks salt solution with 100 U of penicillin and 100 mg/mL of streptomycin for ten minutes at room temperature. The tendon sheath fibroblasts were plated and cultured in Ham's F12 medium (Gibco, Rockville, Maryland) supplemented with 10% fetal calf serum.
The intact tendons were treated with 0.25% trypsin at 37°C for twenty minutes to release the epitenon tenocytes. The epitenon tenocytes were plated and cultured in Ham's F12 medium. The remaining tendons were treated with 0.5% collagenase to release endotenon tenocytes, which were similarly plated and cultured in Ham's F12 medium. All three cell types were grown to confluence at 37°C in a humidified tissue culture chamber supplemented with 5% CO2. At confluence, cells were passaged by washing with phosphate-buffered saline solution and detached with trypsin/EDTA.
Cells from each of the three culture groups (flexor tendon sheath, epitenon, and endotenon) (4 × 104 per well) were grown overnight in forty-eight-well plates supplemented with 10% fetal calf serum. This optimal cell density was determined by a dose curve (data not shown). On the subsequent day, the culture media were changed to serum-free media (Gibco) supplemented with 0.2% Lac albumin (Sigma). The cells were then divided into two groups. The first was the control group without supplementation. The second group was supplemented with 1 ng/mL of TGF-β (R and D Systems, Minneapolis, Minnesota), which included isoforms 1, 2, and 3. The TGF-β groups were further divided according to the concentration of mannose-6-phosphate added. One subgroup received no mannose-6-phosphate. The other three subgroups received 1, 10, or 100 μmol of mannose-6-phosphate. The mannose-6-phosphate was added simultaneously with the TGF-β. In a similar fashion, decorin was added in increasing concentrations of 1, 10, and 50 nmol. Therefore, a total of eight experimental conditions were created for each of the three cell-culture groups (sheath, epitenon, and endotenon): control, TGF-β only, TGF-β and 1 μmol of mannose-6-phosphate, TGF-β and 10 μmol of mannose-6-phosphate, TGF-β and 100 μmol of mannose-6-phosphate, TGF-β and 1 nmol of decorin, TGF-β and 10 nmol of decorin, and TGF-β and 50 nmol of decorin. The cells were subsequently incubated for three days and then fixed with 1% formalin and used to measure collagen-I production with enzyme-linked immunosorbent assay (ELISA).
Quantification of Collagen-I Production with ELISA
Anti-collagen-I primary antibody (Sigma) was added to the fixed cells at 100 μL per well, and this was followed by the addition of a peroxidase-conjugated secondary antibody (Sigma) at 100 μL per well. Each antibody was incubated for one hour and was washed at the end of each step. The optimal concentration of both the primary and the secondary antibodies was determined with titration assays (data not shown). Substrate solution, 3,3′,5,5′-tetramethyl-benzidine (Sigma), was then added to the cells and allowed to incubate for thirty minutes. The reaction was stopped by the addition of 100 μL of 0.5M H2SO4. The reaction solution was then transferred to a ninety-six-well plate at 100 μL per well and read on a microtiter plate reader (Bio-Tek Instruments, Winooski, Vermont) at OD450 (optical density 450) nm.
Collagen production by each cell line was calculated as the mean OD450 and the standard error of the mean. Significance of differences in the decorin and mannose-6-phosphate groups was calculated with use of one-way analysis of variance. For the post-test comparison, baseline collagen production by each cell type was compared with collagen production after administration of TGF-β to confirm significant upregulation of collagen. The separate dosing regimen groups were then compared with the TGF-β group to determine the significance of collagen downregulation. Post-test comparison was performed with use of the Bonferroni correction to adjust for multiple comparisons with a p value of <0.05 to determine significance.
In Vivo Studies
Rabbit Model of Zone-II Flexor Tendon Repair
All rabbit experiments were done according to protocols established by our institution and guidelines from the National Institutes of Health. All operations were performed by one surgeon. Forty adult New Zealand White rabbits (weighing 4.0 to 4.5 kg) were anesthetized with an intramuscular injection of acepromazine (0.01 mg/kg), xylazine (5 mg/kg), and ketamine (50 mg/kg). A single dose of Baytril (enrofloxacin antibiotic) was administered before the left forepaw was shaved and prepared with alcohol and povidone-iodine solution. Continuous intraoperative anesthesia was administered with isoflurane gas through a face mask.
A longitudinal incision was made on the volar surface between the metacarpophalangeal and proximal interphalangeal joints of the middle digit. Tissues were carefully dissected under loupe magnification until the flexor sheath was identified. The sheath was sharply opened in the midline. The flexor digitorum profundus tendon was isolated between the A2 and A4 pulleys and sharply transected. An immediate tendon repair was performed with 5-0 Prolene (polypropylene) suture in the modified Kessler fashion with two strands crossing the repair site.
With use of a tuberculin syringe, 100 μL of the control substance (phosphate-buffered saline solution) or test substance (decorin or mannose-6-phosphate) was applied to the tendon repair site and the surrounding tissue and allowed to infiltrate for one minute. The skin was then reapproximated with a running 4-0 Prolene suture. Antibiotic ointment was applied to the wound, and the forepaw was immobilized in an above-the-elbow cast with the paw and elbow joints in flexion as previously described32. The cast was worn for six weeks in order to protect the tendon repair site.
The animals were fed standard laboratory chow and allowed to roam unrestricted in their cages. At six weeks postoperatively, the animals were again anesthetized, the casts were removed, and the animals were again allowed to roam free in their cages. At eight weeks postoperatively, the animals were killed with an intravenous injection. The left forepaw of each rabbit was removed in preparation for range-of-motion and tensiometry analyses. The paws were grossly observed for overall skin quality and the character of incisional wound-healing.
Commercially available bovine cartilage decorin and mannose-6-phosphate (Sigma) were used in this experiment. Each experimental rabbit received a single dose of decorin or mannose-6-phosphate administered in 100 μL of sterile phosphate-buffered saline solution. The decorin doses were 0.5 mg/100 μL (low-dose) or 1 mg/100 μL (high-dose). The mannose-6-phosphate doses were 1 mg/100 μL (low-dose) or 2 mg/100 μL (high-dose). The decorin doses were based on the work of Fukui et al.33. The mannose-6-phosphate doses were then extrapolated from the in vitro data. Control rabbits received 100 μL of phosphate-buffered saline solution alone. Eight rabbits were used for each experimental and control group, for a total of forty rabbits.
Under loupe magnification, the flexor digitorum profundus tendon was isolated in the proximal portion of the paw and dissected up to its point of entry into the fibro-osseous tendon sheath without disturbing the peritendinous adhesions. The paw was then mounted on the range-of-motion apparatus in a manner that blocked metacarpophalangeal joint rotation. A 1.2-N weight was applied to the proximal part of the flexor digitorum profundus tendon of the middle digit and allowed to hang freely. Angular rotation of the proximal interphalangeal and distal interphalangeal joints was immediately calculated in total degrees of motion.
In order to test the strength of the tendon repair, the flexor digitorum profundus insertion onto the distal phalanx was isolated and transected. The flexor digitorum profundus tendon was then dissected from the proximal and distal directions to the point of the tendon repair. The specimen was kept moist with saline-solution-soaked gauze, and then clamps were firmly affixed to the proximal and distal ends. The clamps and intervening tendon were then mounted on an Instron tensiometer (Instron, Canton, Massachusetts). Under displacement control, the tendon was loaded longitudinally along the axis of its fibers by distracting the clamped ends at a rate of 20 mm/min for continuous uniaxial distraction across the tendon until failure occurred. The location of the failure (either the repair site or the tendon substance) was noted. The ultimate load to failure was recorded in newtons.
The range of motion in each group was calculated as the mean absolute degrees and standard deviation. The breaking strength for the tendons in each group was calculated as the mean force in newtons and the standard deviation. The significance of differences in the range of motion and breaking strength was calculated with analysis of variance. Post-test comparisons of range-of-motion data were performed with use of the Bonferroni correction to adjust for multiple comparisons with a p value of <0.05 to determine significance. Post-test power was calculated to determine the level of beta error in the breaking-strength data. The sample size for all of the in vivo statistical analyses was thirty-two rabbits after exclusion of eight rabbits before they were killed (as described in the Survival and Gross Inspection subsection in the In Vivo Studies section).
In Vitro Studies
Effects of Decorin on TGF-β-Induced Collagen-I Production
Analysis-of-variance testing of all in vitro data showed a significant difference among groups (p = 0.0001). The addition of TGF-β significantly increased collagen production above baseline by 76%, 56%, and 68% for sheath cells, epitenon cells, and endotenon cells, respectively (p < 0.05, Bonferroni correction). The addition of decorin reduced TGF-β-induced collagen-I production in both the sheath-cell and epitenon-cell groups (Fig. 1). The entire experiment was repeated with cells from a separate rabbit, and similar results were obtained.
Effects of Mannose-6-Phosphate on TGF-β-Induced Collagen-I Production
Analysis of variance testing of all in vitro data showed a significant difference among groups (p = 0.0001). Addition of TGF-β significantly increased collagen production above baseline by 76%, 43%, and 91% for the sheath cells, epitenon cells, and endotenon cells, respectively (p < 0.05, Bonferroni correction). Mannose-6-phosphate also reduced the collagen-I-stimulating effects of TGF-β in all three cell groups (sheath, epitenon, and endotenon) (Fig. 2). The entire experiment was repeated with cells from a separate rabbit, and similar results were obtained.
In Vivo Studies
Survival and Gross Inspection
Eight of the original forty experimental rabbits were excluded from the study because they lost the leg cast before the six-week time-point. These excluded rabbits were distributed among the experimental groups, with three in the high-dose decorin group, two in the low-dose mannose-6-phosphate group, one in the high-dose mannose-6-phosphate group, and two in the control group). Therefore, at the time of tensiometry and range-of-motion testing, eight rabbits remained in the low-dose decorin group; five, in the high-dose decorin group; six, in the low-dose mannose-6-phosphate group; seven, in the high-dose mannose-6-phosphate group; and six, in the control group.
Postoperatively, the decorin-treated, mannose-6-phosphate-treated, and control paws did not appear different on gross inspection. Furthermore, no difference in peritendinous adhesions was observed between any of the groups under 3.5× loupe magnification.
Postoperative Range of Motion
Analysis of variance of the in vivo range of motion showed a significant difference among all groups (p = 0.0033). When compared with the control (treated with phosphate-buffered saline solution), neither the group treated with a low dose (0.5 mg/100 μL) of decorin or the group treated with a high dose (1.0 mg/100 μL) had a significant change in the range of motion (Fig. 3) (p > 0.05, Bonferroni correction). Intraoperative application of low-dose mannose-6-phosphate (1 mg/100 μL) significantly increased the range of motion of the operatively treated digits (by 136%; p < 0.05, Bonferroni correction). There was no significant change in the range of motion after application of high-dose mannose-6-phosphate (2 mg/100 μL; p > 0.05, Bonferroni correction) (Fig. 3).
Postoperative Breaking Strength
On tensiometry testing of tendon breaking strength, all tendons ruptured either proximal or distal to the repair site; none ruptured at the repair site. Analysis of variance of in vivo breaking strength showed no difference among groups (p = 0.308) (Fig. 4). Post hoc power testing produced a beta value of 0.28.
Flexor tendon injuries are a difficult challenge in hand surgery. Despite meticulous surgical technique and aggressive postoperative therapy, stiffness and decreased hand function often develop as a result of adhesion formation. The normal wound-healing cascade responsible for tendon repair also produces scarring between the tendon and sheath. Therefore, the strength and durability of a flexor tendon repair often comes at the expense of tendon gliding mechanics. Molecular modulation of the wound-healing response may help to reduce the formation of peritendinous adhesions without compromising the strength of the tendon repair.
Recent advances in wound-healing research have presented a number of potential targets for biochemical manipulation. In particular, TGF-β has proven to be a key component of the wound-healing response, and its manipulation has allowed modulation of scar formation in a number of models. Earlier studies have shown an upregulation of TGF-β mRNA in both tendon and sheath cells in transected rabbit tendons29. Another study demonstrated increased levels of all three TGF-β receptors in both tendon parenchyma and tendon sheath, suggesting that these cells are susceptible to TGF-β activation30. These results suggest a dual mechanism of intrinsic (tendon) and extrinsic (sheath) repair with respect to TGF-β.
In an in vitro cell culture model of rabbit tendon wound-healing, application of TGF-β has been shown to increase levels of collagen production by all three flexor tendon cell lines: sheath, epitenon, and endotenon27. This model has been used to show a decrease in collagen production by tendon cells with the addition of anti-TGF-β antibodies28. Furthermore, intraoperative addition of anti-TGF-β antibody has improved the postoperative range of motion of rabbit forepaws31. In the present study, we screened two natural inhibitors of TGF-β—decorin and mannose-6-phosphate—in order to find a more clinically suitable and less costly antifibrotic agent than antibodies.
Decorin is a small proteoglycan consisting of a core protein and a single glycosaminoglycan side-chain of chondroitindermatan sulfate. Decorin inhibits all three isoforms of TGF-β, and it has been shown to inhibit TGF-β activity in a number of wound-healing models, including intra-articular adhesions in rabbits33-35. The exact mechanism of inhibition is unknown. It has been suggested that decorin forms a complex with TGF-β and sequesters the growth factor in the extracellular matrix22,23.
Our in vitro results showed a significant decrease in TGF-β-induced collagen production by epitenon cells and sheath cells when they were exposed to decorin. Despite these inhibitory effects, no significant improvement in the postoperative range of motion was found following application of decorin in vivo. It is possible that a different intraoperative dose of decorin may be required to improve the range of motion in the in vivo model. Additional studies with a wider dosing range will be required in order to determine this. Furthermore, other investigators have used continuous infusion of decorin to decrease postoperative adhesion formation and this may be applied to our model in the future33. The decorin used in this study was derived from a bovine source, and cross-reactivity with the rabbit model may have decreased its effectiveness. Additional studies will be required to determine the immunogenicity of bovine decorin and its effects on tendon wound-healing.
Endotenon cells showed no significant change in collagen production in response to decorin, and the absolute breaking strength of the repair was not weakened by the application of decorin in the doses tested. However, the post hoc power calculation produced a beta value of 0.28. Given this level of beta error, additional studies must be performed with a larger sample size in order to determine the effect of decorin on tendon repair strength in this model.
Mannose-6-phosphate is a simple carbohydrate that is also known to inhibit TGF-β. It is unclear how mannose-6-phosphate produces its inhibitory effect. Normally, TGF-β is secreted from cells in an inactive form bound to latent TGF-β-binding proteins (LTBPs). One way in which the LTBP-TGF-β complex becomes activated is by binding to the insulin-like growth factor-2 receptor. Mannose-6-phosphate may competitively bind to this receptor, thereby inhibiting activation of the LTBP-TGF-β complex36.
In our in vitro study, mannose-6-phosphate inhibited the effects of TGF-β-induced collagen production in all three cell lines. In contrast to decorin, low-dose mannose-6-phosphate significantly improved the postoperative range of motion in our in vivo study. High-dose mannose-6-phosphate improved the range of motion, but the result did not reach significance. We concluded that postoperative adhesion formation, more than joint stiffness from immobilization, appears to be the major variable with respect to the differences in the range of motion in this in vivo model.
It is possible that the larger inhibitory effect of man-nose-6-phosphate (when compared with decorin) on sheath cells in vitro correlates with the improved range of motion. If this is the case, then the in vitro model may be a useful tool for screening other inhibitors prior to performing more expensive and time-consuming in vivo experiments. Also, the inhibition of collagen production by endotenon cells by mannose-6-phosphate suggests that intrinsic tendon-healing might be inhibited, but our in vivo data did not show a decrease in the absolute breaking strength of the tendon repair. A number of factors may explain this finding. First, the post hoc beta value previously discussed suggests the possibility of a beta error. Second, the temporal sequence of TGF-β inhibition was not evaluated in vivo. There may be significant differences in tendon repair strength earlier than at the eight-week time-point used for sampling in this study. With the recent popularity of early motion protocols, this possibility could have profound effects on tendon mechanics in the postoperative period. Finally, our biomechanical testing protocol did not analyze tendon stiffness or energy absorption and did not standardize gauge length between samples. The testing protocol will need to be refined and standardized to a greater degree before any conclusions can be drawn regarding tendon repair strength.
This study provides additional evidence that inhibition of TGF-β may reduce postoperative scar formation in flexor tendons. Our data suggest that a single intraoperative dose of mannose-6-phosphate is more effective than decorin at improving postoperative range of motion. Mannose-6-phosphate, a simple carbohydrate, is ubiquitous, nonimmunogenic, and easily produced, making it an ideal candidate for additional study and possibly clinical application. ▪
In support of their research for or preparation of this manuscript, one or more of the authors received grants or outside funding from the Veterans Administration Merit Review Award. None of the authors received payments or other benefits or a commitment or agreement to provide such benefits from a commercial entity. No commercial entity paid or directed, or agreed to pay or direct, any benefits to any research fund, foundation, educational institution, or other charitable or nonprofit organization with which the authors are affiliated or associated.
Investigation performed at the Division of Plastic Surgery, Stanford University Medical Center, Stanford, and the Section of Plastic Surgery, VA Palo Alto Health Care System, Palo Alto, California
1. , Rank F. Experimental intrinsic healing of flexor tendons based on synovial fluid nutrition. J Hand Surg [Am]. 1978;3: 21-31.
2. . The role of tendon tissues in tendon healing. Plast Reconstr Surg. 1976;57: 39-49.
3. . Tendon healing within the flexor digital sheath in the dog. J Bone Joint Surg Am. 1962;44: 49-64.
4. , Lesker PA. Biochemical evidence of flexor tendon participation in the repair process—an in vitro study. J Hand Surg [Br]. 1984;9: 117-20.
5. , Manske PR, Vande Berg JS, Lesker PA, Akeson WH. Flexor tendon repair in vitro: comparative histologic study of the rabbit, chicken, dog, and monkey. J Orthop Res. 1984;2: 39-48.
6. . Experimental tendon repair within the flexor tunnels: use of polyethylene tubes for improvement of functional results in the dog. Surgery. 1949;26: 181-98.
7. . The effect of the local corticosteroids on the healing and biochemical properties of the partially injured tendon. Clin Orthop Relat Res. 1982;163: 170-9.
8. , Lindsay WK, McCain WG. Effect of beta-aminoproprionitrile on adhesions following digital flexor tendon repair in chickens. Surg Forum. 1970;21: 509-11.
9. , Harvey JS Jr. Dimethyl sulfoxide. Biomechanical effects on tendons. Am J Sports Med. 1982;10: 177-9.
10. , Ishizue K, Billings E Jr, Wiig M, Vande Berg J, Akeson WH, Gelberman R. Hyaluronan in flexor tendon repair. J Hand Surg [Am]. 1989;14: 837-43.
11. , Tuel RJ, Mass DP. Effect of hyaluronic acid on rabbit profundus flexor tendon healing in vitro. J Surg Res. 1993;55: 411-5.
12. , Jones LC, Hungerford DS. Hyaluronic acid and its effect on postoperative adhesions in the rabbit flexor tendon. A preliminary look. Clin Orthop Relat Res. 1986;206: 281-9.
13. , Schultz GS. Growth factors and wound healing: biochemical properties of growth factors and their receptors. Am J Surg. 1993;165: 728-37.
14. , Klagsbrun M. Angiogenic factors. Science. 1987;235: 442-7.
15. , Roberts AB. Transforming growth factor-beta: recent progress and new challenges. J Cell Biol. 1992;119: 1017-21.
16. , Chen YG. Controlling TGF-beta signaling. Genes Dev. 2000;14: 627-44.
17. , Hogg A, Lankat-Buttgereit B, Krieg T. Co-localization of transforming growth factor beta 2 with alpha 1(I) procollagen mRNA in tissue sections of patients with systemic sclerosis. J Clin Invest. 1990;86: 917-22.
18. , Kahari L, Jaakkola S, Kahari VM, Varga J, Uitto J, Jimenez SA. Evaluation of transforming growth factor beta and type I procollagen gene expression in fibrotic skin diseases by in situ hybridization. J Invest Dermatol. 1990;94: 365-71.
19. , Shen YJ, Scott PG, Gong Y, Tredget EE. Enhanced expression of mRNA for transforming growth factor-beta, type I and type III procollagen in human post-burn hypertrophic scar tissues. J Lab Clin Med. 1993;122: 465-73.
20. , Hsiao LL, Jaakkola S, Sollberg S, Aumailley M, Timpl R, Chu ML, Uitto J. Activation of collagen gene expression in keloids: co-localization of type I and VI collagen and transforming growth factor-beta 1 mRNA. J Invest Dermatol. 1991;97: 240-8.
21. , Foreman DM, Ferguson MW. Neutralizing antibody to TGF-beta 1,2 reduces cutaneous scarring in adult rodents. J Cell Sci. 1994;107: 1137-57.
22. , Noble NA, Yamamoto T, Harper JR, Yamaguchi Y, Pierschbacher MD, Ruoslahti E. Natural inhibitor of transforming growth factor-beta protects against scarring in experimental kidney disease. Nature. 1992;360: 361-4.
23. , Hausser H, Schonherr E, Kresse H. Influence of decorin expression on transforming growth factor-beta-mediated collagen gel retraction and biglycan induction. Matrix Biol. 2000;19: 631-6.
24. , Berry M. Transforming growth factor-beta 1 and basic fibroblast growth factor in the injured CNS. Trends Pharmacol Sci. 1993;14: 337-42.
25. , Rifkin DB. Cellular activation of latent transforming growth factor beta requires binding to the cation-independent mannose 6-phosphate/insulin-like growth factor type II receptor. Proc Natl Acad Sci USA. 1991;88: 580-4.
26. , Sporn MB. Transforming growth factor-beta. In: Clark RA, editor. The molecular and cellular biology of wound repair. 2nd ed. New York: Plenum; 1996. p 275-98.
27. , Yalamanchi N, Pham H, Longaker MT, Chang J. Flexor tendon wound healing in vitro: effects of TGF-beta on tendon cell collagen production. J Hand Surg [Am]. 2002;27: 615-20.
28. , Pham H, Ho F, Teng K, Longaker MT, Chang J. Inhibition of TGF-beta-induced collagen production in rabbit flexor tendons. J Hand Surg [Am]. 2004;29: 230-5.
29. , Most D, Stelnicki E, Siebert JW, Longaker MT, Hui K, Lineaweaver WC. Gene expression of transforming growth factor beta-1 in rabbit zone II flexor tendon wound healing: evidence for dual mechanisms of repair. Plast Reconstr Surg. 1997;100: 937-94.
30. , Pham H, Longaker MT, Chang J. Differential expression of transforming growth factor-beta receptors in a rabbit zone II flexor tendon wound healing model. Plast Reconstr Surg. 2001;108: 1260-7.
31. , Thunder R, Most D, Longaker MT, Lineaweaver WC. Studies in flexor tendon wound healing: neutralizing antibody to TGF-beta1 increases postoperative range of motion. Plast Reconstr Surg. 2000;105: 148-55.
32. , Chang J, Broome RL, Most D. A method for immobilizing the forelimbs of rabbits. Contemp Top Lab Anim Sci. 1998;37: 94-5.
33. , Fukuda A, Kojima K, Nakajima K, Oda H, Nakamura K. Suppression of fibrous adhesions by proteoglycan decorin. J Orthop Res. 2001;19: 456-62.
34. , Mann DM, Ruoslahti E. Negative regulation of transforming growth factor-beta by the proteoglycan decorin. Nature. 1990;346: 281-4.
35. , Schonherr E. Proteoglycans of the extracellular matrix and growth control. J Cell Physiol. 2001;189: 266-74.
36. , Rifkin DB. Cellular activation of latent transforming growth factor beta requires binding to the cation-independent mannose-6-phosphate/insulin-like growth factor type II receptor. Proc Natl Acad Sci USA. 1991;88: 580-4.