- Email: [email protected]
Flexor tendon healing in the rat: a histologic and gene expression study
Flexor Tendon Healing in the Rat: A Histologic and Gene Expression Study Wataru Oshiro, MD, Jueren Lou, MD, Xiaoyun Xing, MD, Yizheng Tu, MD, Paul R. Manske, MD, St. Louis, MO
Purpose: To establish a rat flexor tendon laceration and repair model to investigate the molecular mechanisms of flexor tendon healing. Methods: Surgery was performed on rat flexor digitorum longus tendons from both hind feet. Repaired tendons were harvested at 0, 3, 7, 14, 21, 28, 42, 56, and 84 days after surgery. Histologic study (first 84 days) and gene expression study (first 28 days) of several collagens and matrix metalloproteinases (MMPs) were performed. Results: In the histologic study pre-existing collagen bundles were degraded between days 7 to 21. Newly formed collagen fibers crossed the repair site by day 28. Remodeling of the collagen fibers continued until day 84. Gene expression of type I collagen decreased initially and then returned gradually to the initial level by day 28, whereas expression levels of types III, V, and XII collagen were increased after surgery. The expression levels of MMP-9 and MMP-13 peaked between days 7 to 14, whereas MMP-2, MMP-3, and MMP-14 levels increased after surgery and maintained high levels until day 28. Conclusions: The rat tendon laceration model represented the entire tendon healing process. The results of this study suggest that MMP-9 and MMP-13 participate only in collagen degradation, whereas MMP-2, MMP-3, and MMP-14 participate not only in collagen degradation but also in collagen remodeling. (J Hand Surg 2003;28A:814-823. Copyright © 2003 by the American Society for Surgery of the Hand.) Key words: Flexor tendon healing, rat, histology, collagen, matrix metalloproteinase.
The improved understanding of flexor tendon healing has led to advances in tendon repair techniques and rehabilitation protocol, which in turn markedly From the Department of Orthopaedic Surgery, Barnes-Jewish Hospital, Washington University, St. Louis, MO; and St. Louis VA Medical Center, St. Louis, MO. Received for publication February 18, 2003; accepted in revised form June 23, 2003. Supported by Shriners Hospitals for Children, Grant # 8510. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: Jueren Lou, MD, Department of Orthopaedic Surgery, Washington University School of Medicine, One Barnes Hospital Plaza, STE 11300, St. Louis, MO 63110. Copyright © 2003 by the American Society for Surgery of the Hand 0363-5023/03/28A05-0017$30.00/0 doi:10.1053/S0363-5023(03)00366-6
814
The Journal of Hand Surgery
have improved the clinical results of zone II flexor tendon injury.1 Recent developments in genomic and molecular biology research provide the opportunity to investigate the molecular events of tendon healing. Although the rat has been used as an experimental animal in various investigative fields it has been used infrequently for in vivo flexor tendon research.2,3 There are many advantages to using the rat as the experimental animal in the investigation of the molecular mechanisms of the tendon healing process. These include the variety of reagents to the rat that are commercially available; the rat genome sequence, which has been elucidated; and a large amount of biologic information that is available from literature databases. In this study a complete flexor tendon laceration
Oshiro et al / Flexor Tendon Healing in the Rat
815
each harvest time (Fig. 1B). The tendon specimens were approximately 4 mm in length including the repair site and did not include the proximal laceration site. We carefully excluded any fibrous adhesions if present along with the vincula structures from the specimens. The animal protocol was approved by the Animal Studies Committee at Washington University School of Medicine.
Histologic Evaluation Figure 1. Surgical procedure. (A) Arrows indicate the distal and the proximal lacerations. The distal laceration is repaired immediately with one simple stitch (Ethilon; Ethicon). (B) The tendon specimen including the repair site is harvested at each time point.
model in the rat was investigated. The histologic findings of the repaired rat tendon during the first 84 days are presented. Gene expression levels of several types of collagen and matrix metalloproteinases (MMPs) during the first 28 days also are examined by reverse-transcription polymerase chain reaction (PCR). The rat tendon injury model described in this study provides an option for investigating the mechanisms of the tendon healing process.
Methods Surgical Procedure Seventy male Lewis rats aged 7 to 10 weeks old (Charles River Laboratory, Wilmington, MA) were anesthetized with ketamine (Ketaset; Fort Dodge Animal Health, Fort Dodge, IA) (75.0 mg/kg of body weight injected intraperitoneally) and medetomidine hydrochloride (Domitor; Animal Health, Exton, PA) (0.5 mg/kg of body weight injected intraperitoneally). Under sterile conditions surgery was performed on the second through fifth toes of both hind feet using a surgical microscope (Fig. 1A). A midline incision was made on the plantar surface of each toe. The flexor sheath was opened between the proximal annular pulley and the distal interphalangeal joint. The flexor digitorum longus tendon was exposed. A complete transverse laceration was made in zone II. The laceration was repaired immediately with one simple 9-0 stitch (Ethilon; Ethicon, Somerville, NJ). A second transverse laceration was made in the tendon 3.0 mm proximal to the repaired laceration to reduce tensile forces across the repair site. The flexor sheath was not repaired and the skin was closed with a continuous running 9-0 suture (Ethilon). After surgery rats were allowed to walk freely in cages without restriction. Healing tendons were harvested at
Flexor tendons from 2 rats (16 tendons) were harvested at each time point of 0 (nonlacerated normal tendon), 3, 7, 14, 21, 28, 42, 56, and 84 days after surgery. The macroscopic appearances of the repaired tendons were observed by microscope at a magnification of 20 times. Tendon specimens then were fixed in 10% buffered formalin, embedded in paraffin, sectioned longitudinally at 6 m thickness, and evaluated with standard hematoxylin-eosin staining.
Total RNA Extraction At least 30 flexor tendons from 4 rats were harvested at each time point including days 0 (nonlacerated normal tendon), 3, 7, 14, 21, and 28 after surgery. A greater number of the tendons were needed for day 0 (80 tendons from 10 rats) than the other harvest time to obtain enough total RNA. Tendon specimens were immersed immediately in lysis solution (4 mol/L guanidinium thiocyanate, 25 mmol/L sodium citrate, pH 7.0, 0.5% sodium N-lauroyl sarcosine, and 0.1 mol/L mercaptoethanol). Tendon specimens in lysis solution were frozen and stored in liquid nitrogen until use. Pooled specimens were homogenized (Polytron PT3000; Brinkmann Instruments, Inc., Westbury, NY) in 4.0 mL of lysis solution. Four milliliters of buffer-saturated phenol (pH 4.3) and 0.8 mL of chloroform-isoamyl alcohol were added to each of the homogenates. The samples were mixed, placed on ice for 15 minutes, and centrifuged at 15,000 ⫻ g for 20 minutes at 4°C. The aqueous phase was collected and added with the same volume of isopropanol and kept at ⫺20°C overnight. The RNA was precipitated by centrifuge at 15,000 ⫻ g for 20 minutes at 4°C. The precipitated RNA was washed with 70% ethanol, dissolved in diethyl pyrocarbonate-treated water, and stored at ⫺80°C until use.
Reverse-Transcription Polymerase Chain Reaction Gene expression levels of types I, III, V, and XII collagen and MMP-2, -3, -9, -13, and -14 were ex-
816
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Table 1. Primer Sequences and PCR Conditions of Collagen and MMP Genes
Products GAPDH
GenBank Accession # M17701
Collagen alpha1 type I (Col I) Z78279 Collagen alpha1 type III (Col III) M21354
Primer Sequences
Annealing Number Size Temperature of (bp) (°C) Cycles
Forward 5⬘-ACCACAGTCCATGCCATCAC-3⬘ Reverse 5⬘-TCCACCACCCTGTTGCTGTA-3⬘
452
55
24
Forward 5⬘-CATGTCTGGTTTGGAGAGAG-3⬘ Reverse 5⬘-TCCATTCCGAATTCCTGGTC-3⬘
404
50
21
Forward 5⬘-GATGGATCAAGTGGACATCCAGGTCCCATT-3⬘ Reverse 5⬘-ATCTTGCAGCCTTGGTTAGGATCAACCCAG-3⬘
650
64
25
344
62
27
530
62
34
328
64
24
462
53
26
458
62
24
479
62
23
440
64
25
Collagen alpha1 type V (Col V) AJ005394 Forward 5⬘-CATGCCCAGAGAGAACAAAGGGAAAGAG-3⬘ Reverse 5⬘-TCCATCGGAAAGGCACATGTGGAATGAC-3⬘ Collagen alpha1 type XII (Col XII) U57362 Forward 5⬘-CAGAGTCGGTTCATCCACAAGTGCTCCTAT-3⬘ Reverse 5⬘-CACAAACATGCAATTCCATCTCGGCAGCAG-3⬘ MMP-2 U65656 Forward 5⬘-ATACAGGTGTGCCAAGGTGGAAACCAGAGA-3⬘ Reverse 5⬘-CCCATGGGGAACTGTTAAAGGGAGAAGCAA-3⬘ MMP-3 X02601 Forward 5⬘-TGAAACCGTCCAGAAGATCG-3⬘ Reverse 5⬘-AGCAGTGCTTCTGAATGTCC-3⬘ MMP-9 U24441 Forward 5⬘-GCTGCACCACCTTACCGGCCCTTTTATTTA-3⬘ Reverse 5⬘-TGGTTATCCTTCCACTGAGGGATCATCTCG-3⬘ MMP-13 M60616 Forward 5⬘-GATGTCAGGCACTAAAGGAAGGGGATAACC-3⬘ Reverse 5⬘-GCACCAAGTGTCAGTCACTAAGGAAAGCAG-3⬘ MMP-14 X83537 Forward 5⬘-CTTGCTGTCTTCTTCTTTAGACGCCATGGG-3⬘ Reverse 5⬘-TGCCAACTGGGTACACTTGGTACGTATAGG-3⬘
amined using reverse-transcription PCR. One microgram of total RNA was reverse transcribed (1stStrand cDNA Synthesis Kit; Roche, Indianapolis, IN) in 20 L of reaction volume, containing 1 mmol/L of dNTP, 10 mmol/L of Tris, 50 mmol/L of KCl (pH 8.3), 5 mmol/L of MgCl2, 50 U of RNase inhibitor, 1.6 g of oligo-p(dT)15 primer, and 20 U of avian myoblastosis virus reverse transcriptase. Incubation was performed for 60 minutes at 42°C. One microliter of 1st-Strand cDNA was amplified in 20 L of reaction volume containing PCR buffer (10 mmol/L Tris-HCl, 1.5 mmol/L MgCl2, 50 mmol/L KCl, pH 8.3), 0.2 mmol/L of each dNTP, 1.25 units of Taq DNA polymerase (Roche), and 0.25 mol/L of forward and reverse gene-specific primer set. Polymerase chain reaction amplification was done (DNA Thermal Cycler 480; PerkinElmer, Inc., Wellesley, MA) under the following conditions: preheating for 5 minutes at 94°C, then for each cycle denaturation for 15 seconds at 94°C, annealing for 45 seconds, extension for 1 minute at 72°C, with final extension for 2 minutes at 72°C. The annealing tem-
perature and the number of PCR cycles were arranged according to gene-specific primer sets (Table 1). Aliquots of PCR products were separated on 1.5% agarose gel and visualized with ultraviolet light after ethidium bromide staining. Gel images were scanned by a digital camera and saved in tiff file formats. Densitometric measurement of gel images was performed (Scion Image version beta 4.0.2; Scion Corporation, Frederick, MD). Band density of PCR products was normalized to that of glyceraldehyde3-phosphate-dehydrogenase (GAPDH).
Results Histologic Evaluation Macroscopic findings. The tendon repair site gradually healed throughout the 84 days after surgery without any obvious gap formation at the repair site (Fig. 2). The surfaces of the tendon specimens were smooth and had a glistening appearance throughout the time course. In all tendon specimens there was no indication of necrotic changes in the tendon segment between the 2 lacerations. Hypertrophic changes in
Oshiro et al / Flexor Tendon Healing in the Rat
817
Figure 2. Macroscopic appearance of the repaired tendons. Stitches (Ethilon; Ethicon) indicate the repair site. No obvious gap formations are noted. The surfaces of the tendon specimens are smooth and have a glistening appearance throughout the time course. Hypertrophic changes in the repair site first were observed on day 7, peaked at day 21, and gradually reduced.
the repair site first were observed on day 7, peaked at day 21, and gradually reduced. Mild to moderate adhesion formation was observed between the repair site and the surrounding tissues during days 7 to 84, and appeared to peak at day 21. The repair site was dissected easily, however, throughout the entire time course. The adhesion formation between the proximal cut tendon end (second laceration site) and surrounding tissues was more severe than those at the repair site.
Microscopic findings. Day 0 (nonlacerated normal tendon). The surface of the flexor tendon was characterized by a 1- to 2-cell–thick layer of the epitenon fibroblasts, which had spindle or slendershaped nuclei. In the deep region the endotenon fibroblasts, which had round to oval-shaped nuclear and clear cytoplasm, were aligned linearly between the longitudinal mature collagen bundles. Multiple sections of nonlacerated normal tendons failed to reveal any distinct vascular structures in the deep region. The vascular structures were noted only in the epitenon, mostly on the dorsal surface (Fig. 3A).
Day 3. In the superficial regions near the repair site there was a thickening of the epitenon to 3- to 4-cell–thick layers. In the repair site the lacerated tendon ends remained in close approximation. In the region near the repair site the cellularity of the endotenon fibroblasts was increased (Fig. 3B). There was no evidence of newly formed vascular structures at the repair site.
Day 7. The lacerated tendon was covered superficially by a several-cell–thick layer of epitenon (Fig. 3C). The repair site was united superficially at this time point. In the deep region of the repair site the severed ends of the collagen bundles began to degrade into thin fibers. The cellularity of the endotenon fibroblasts further increased in the region close to the repair sites. In the repair site there was no sign of cell migration from the surface region toward the deep region. In the areas within the repair site newly formed capillary vessels were observed (Fig. 3D).
Day 14. The collagen bundles were degraded massively in the large area of the repair site, replaced by randomly arranged fine strands of collagen fibers. Particular note was given to numerous large round cells (indicated by arrows on Fig. 4B), which were observed within the tendon (Fig. 4A and B).
Day 21. The collagen bundles were degraded almost completely in the entire region of the repair site. The original collagen bundles remained only in the distal region near the tendon insertion into bone. The number of large round cells increased. In the region subjacent to the tendon surface the longitudinally aligned cells and newly synthesized fine collagen fibers were observed to cross the repair site in a few tissue sections (Fig. 4C). It was difficult to define the site of the laceration clearly.
Day 28. In the repair site there was increased concentration of collagen fibers. The repair site was
818
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Figure 3. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 0 –7). Arrowheads indicate the site of the repaired laceration. (A) Day 0 (original magnification, 400⫻). Epitenon fibroblasts (small arrow) have spindle-or slender-shaped nuclei. Endotenon fibroblasts (large arrows) have round to oval-shaped nuclei and clear cytoplasm. (B) Day 3 (magnification, 100⫻). The lacerated tendon ends remain in close approximation. In the region near the repair site the cellularity of the endotenon fibroblasts has increased. (C) Day 7 (magnification, 100⫻). The repair site is united superficially. Asterisks indicate where the collagen bundles degrade into thin fibers. (D) Day 7 (magnification, 400⫻). Newly formed capillary vessels (arrows) are observed.
crossed completely by parallel alignment of fine collagen fibers and interposed cells. The large round cells began to change shape into more elongated slender cells, suggestive of mature endotenon fibroblasts (Fig. 4D). The cellularity and the vascularity of the repair site peaked at this time point.
Day 42. By this time point the tendon repair site was not distinguishable and only the suture materials suggested the region of the repair. The collagen fibers crossing the repair site showed maturation to collagen bundles (Fig. 5A). The slender-shaped endotenon fibroblasts were apparent between the mature collagen bundles (Fig. 5B).
Days 56, 84. Collagen bundles progressively matured to normal thickness. The cellularity of the endotenon fibroblasts gradually decreased to a normal
level by day 84 (Fig. 5C and D). The endotenon fibroblasts were more slender and less plump than that of the normal tendon (day 0). The amount of vascular structures in the repair site showed a gradual decrease, however, they were still visible on day 84.
Gene Expression Study Using ReverseTranscription Polymerase Chain Reaction Agarose gel images and the relative expression levels (RELs) of selected genes against GAPDH are shown in Fig.ure 6. The REL of type I collagen (Col I) gene was initially high (day 0), decreased to approximately half that level on day 3, and gradually returned to the initial level by day 28. The RELs of types III, V, and XII collagen (Col III, V, XII, respectively) genes were initially low, started to increase during days 3 to 7, peaked at day 14, and
Oshiro et al / Flexor Tendon Healing in the Rat
819
Figure 4. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 14 –28). Arrowheads indicate the site of the repaired laceration. (A) Day 14 (original magnification, 100⫻). Collagen bundles are massively degraded (asterisk). (B) Day 14 (magnification, 400⫻, the region of asterisk on A). Collagen bundles are replaced by randomly arranged newly synthesized fine collagen fibers. Large round cells are observed (arrows). (C) Day 21 (magnification, 400⫻, repair site subjacent to tendon surface). Longitudinally aligned cells and fine collagen fibers cross the repair site. (D) Day 28 (magnification, 400⫻). Repair site is crossed completely by parallelly aligning fine collagen fibers and interposed cells, which have changed shape from larger and round to more slender (arrows).
gradually decreased, but still remained at a higher level on day 28 than the day 0 level. There were 2 MMP gene expression patterns. The RELs of all examined MMP genes were initially low. In the first pattern the gene expression levels of MMP-9 and MMP-13 were very high on days 7 and 14, and decreased by day 21. In the second pattern the RELs of MMP-2, MMP-3, and MMP-14 genes increased after surgery and maintained high levels until day 28.
Discussion Although several animal experimental models including the dog, rabbit, and chicken have been designed for in vivo flexor tendon research, there are few reports using rat flexor tendons.2,3 Because of their small size (1-mm width), the surgery may be
difficult to perform and postoperative immobilization may be difficult to attain. The rat tendon laceration and repair model developed in this study, however, allowed surgery on rat flexor tendons to be performed without difficulty using the surgical microscope. Furthermore postoperative immobilization was obviated by cutting the tendon proximal to the laceration site. On the other hand there are several advantages to using the rat as an experimental animal for tendon research, particularly in investigating the molecular mechanisms of the repair process. First, rats are less expensive and easier to handle compared with the other animal species. Second, many important and different types of reagents to the rat are now available. Finally, the rat genome sequence has been elucidated and currently can be obtained online (www. ncbi.nlm.nih.gov/genome/guide/rat).
820
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Figure 5. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 42– 84). Arrowhead indicates the site of the repaired laceration. (A) Day 42 (original magnification, 100⫻). The collagen fibers mature to collagen bundles. (B) Day 42 (magnification, 400⫻). The slender-shaped endotenon fibroblasts are present between mature collagen bundles (arrows). (C) Day 56 (magnification, 100⫻). Collagen bundles are more mature compared with day 42 (Fig. 4A). (D) Day 84 (magnification, 400⫻). The collagen bundles and the cellularity of the endotenon fibroblasts are similar to normal. Endotenon fibroblasts (arrows) are more slender and less plump than that of the normal tendon.
Histology This rat model presented the entire tendon healing process, which includes the superficial repair of the repair site, collagen degradation, synthesis of new collagen fibers, collagen remodeling, and neovascularization. This healing process is for the most part consistent with the results of previous reports using other animal species.4 – 6 At day 14 the repair site was massively replaced by degraded collagen fibers and large round cells. These large round cells increased until day 28 and then decreased gradually. These cells changed shape into more slender endotenon fibroblasts, which were located between the mature collagen bundles. These cells seemed to play the central role as reparative cells in the rat tendon healing process, including the degradation of pre-existing collagen, synthesis, and remodeling of new collagen fibers. The origin of this reparative cell was unclear. Investigators of previous studies have had various opinions about their origin.
With a canine flexor tendon healing model, Potenza4 concluded that the tendon cells played no active role in the repair process and that repair was completed entirely by the cellular activity of the sheath and surrounding tissue. With a chicken model, Lindsay and Thomson5 showed that tendon sheath contributed little to the tendon repair; however, all of the connective tissue cells that make up the tendon participated in the repair process. Umeda6 investigated the repair process of chicken tendons within an intact portion of the synovial sheath and suggested that the main origin of fibroblasts of the granulation tissue were perivascular mesenchymal cells. From the observations of the divided canine tendon, followed by controlled mobilization, Gelberman et al7 showed that epitenon fibroblasts migrated to the region with moderate gaps; however, when the tendon ends were held in close approximation there was minimal response from the epitenon and a greater initial response from endotenon fibroblasts. In this study the
Oshiro et al / Flexor Tendon Healing in the Rat
821
Fig.ure 6. Agarose gel images and the relative gene expression levels against GAPDH (x axis, days after surgery; y axis, relative expression levels). Col I gene expression level is initially high, temporally decreases on day 3, and gradually recovers to a normal level. The expression levels of Col III, V, and XII genes peak at day 14 and gradually decrease but still remain at a higher than normal level on day 28. The gene expression levels of MMP-9 and MMP-13 were high on days 7 and 14. Expression levels of the MMP-2, MMP-3, and MMP-14 genes increased after surgery and maintained high levels until day 28.
repaired tendon ends remained in close approximation in almost all tendon specimens, and no migrating epitenon cells could be identified. Additionally the findings on day 3 revealed the increased cellularity of endotenon fibroblasts in the region adjacent to the repair site without neovascularization. This indicated an early response of pre-existing endotenon fibroblast proliferation and/or migration toward the repair site. Thus we speculate that a principal component of the reparative cells noted in this study comes from pre-existing endotenon fibroblasts.
Gene expression— collagen. Although many investigators point out the up-regulation of type III collagen in the early tendon healing stage,8 –10 marked up-regulation of types V and XII collagen gene expression has not been reported. Type V collagen has been found in vascular walls of the tendon,11 immature chick embryonic tendon,12 and degenerated human adult tendons.13 It is postulated that type V collagen is associated closely with the vascular generation and may play a role in tendon growth and healing. Recently type XII collagen has been noted to have functional interaction with type I collagen and proteoglycans.14,15 Type XII collagen may
serve as a molecular bridge between fibrils, as well as between fibrils and other extracellular matrix (ECM) constituents.14 Although the role of this molecule in tendon healing remains to be defined further, a possible mechanism of type XII collagen is to stabilize the reparative tissue components by keeping the newly synthesized type I collagen fibrils together and linking between these fibrils and proteoglycans.
Gene expression—MMP. MMPs degrade several substances in the ECM and are considered integral mediators of ECM remodeling that takes place during tissue development, morphogenesis, and repair.16 Although MMPs were reported to be present in the human degenerated supraspinatus17 and Achilles tendons,18 as well as in the torn rabbit supraspinatus tendon,19 little is known of their behavior in the flexor tendon healing process within the digital sheath. Gene expression study revealed 2 different gene expression patterns of MMPs during the first 28 days. MMP 13 (collagenase 3) functions to cleave the triple helix of fibrillar collagens, types I, II, and III.16,20 MMP-9 (gelatinase B) is derived predominantly from inflammatory cells and degrades denatured collagens as well as other ECMs.20 In degrad-
822
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
ing the pre-existing collagen bundles, MMP-9 may follow the function of MMP-13 by further degrading fibrillar collagens.16,20,21 As noted in the histologic findings the degradation of pre-existing collagen bundles first was observed on day 7 and almost completed by day 21; reparative fibroblast proliferation and realignment of the newly synthesized collagen fibers began on day 21. It is possible that the primary roles of the MMP-9 and MMP-13 are accomplished by day 21 and they contribute minimally to the subsequent remodeling phase. The exhaustion of the target substrates, namely, damaged helical collagens, might trigger the down-regulation of MMP-9 and MMP-13 gene expression. The decrease of MMP-9 producible inflammatory cells might be another possible reason for the down-regulation of MMP-9 gene expression. MMP-2 (gelatinase A) has a similar function to MMP-9 but it generally is secreted from fibroblasts. MMP-3 (stromelysin 1) has a broad substrate range and activates other MMPs and growth factors. MMP-3 also is thought to have the potential to play a major role in ECM remodeling.18 MMP-14 (membrane-type MMP-1) exists on cell membranes and has been implicated as a key player in a proteinase cascade involving MMP-2, MMP-9, and MMP-13.16 Our results suggest that these 5 MMPs are involved in tendon ECM degradation: MMP-9 and MMP-13 make little contribution to the subsequent ECM remodeling; conversely MMP-2, MMP-3, and MMP-14 might have the potential to play key roles during the remodeling phase of tendon healing. It should be noted that this study focused only on MMP gene expression level, which does not represent the enzyme activity. All MMPs are synthesized as inactive pro-MMPs,16 which then are converted to an active form by various factors.20 Nevertheless this study potentially increases our understanding of the molecular mechanisms involved in the tendon repair process. Further studies using substrate gel electrophoresis and fluorogenic substrate22 are needed to evaluate the enzyme activities of MMPs. Although our rat model presented the entire healing process without gap formation it is important to note that the model did not entirely represent the clinical condition. Tensile stress across the repair site was absent because of the second laceration proximal to the repair site. Also the progressive adhesion formation at the more proximal laceration site may have prevented passive gliding of the repaired tendons; however, passive gliding still might have been
achieved, at least in the initial stage, by allowing the rat to cage walk without restriction. Nevertheless this model does provide an opportunity to study the biologic processes of tendon healing.
References 1. Strickland JW. Development of flexor tendon surgery: twenty-five years of progress. J Hand Surg 2000;25A:214 – 235. 2. Greenlee TK Jr, Pike D. Studies of tendon healing in the rat. Remodeling of the distal stump after severance. Plast Reconstr Surg 1971;48:260 –270. 3. Lane JM, Black J, Bora FW Jr. Gliding function following flexor-tendon injury. A biomechanical study of rat tendon function. J Bone Joint Surg 1976;58A:985–990. 4. Potenza AD. Tendon healing within the flexor digital sheath in the dog. An experimental study. J Bone Joint Surg 1962; 44A:49 – 64. 5. Lindsay WK, Thomson HG. Digital flexor tendons: an experimental study. Part I. The significance of each component of the flexor mechanism in tendon healing. Br J Plast Surg 1960;12:289 –316. 6. Umeda HT. An experimental study on the intrinsic healing mechanism of digital flexor tendons. J Jpn Orthop Assoc 1978;52:917–929. 7. Gelberman RH, Vandeberg JS, Manske PR, Akeson WH. The early stages of flexor tendon healing: a morphologic study of the first fourteen days. J Hand Surg 1985;10A:776 – 784. 8. Sakai H, Koibuchi N, Ohtake H, Tamai K, Fukui N, Oda H, et al. Type I and type III procollagen gene expressions in the early phase of ligament healing in rabbits: an in situ hybridization study. J Orthop Res 2001;19:132–135. 9. Hiranuma K, Suzuki K, Hirata K, Nakamura H, Higashi K, Hirano H. Extracellular matrices in peritendinous connective tissue after surgical injury to the chicken flexor tendon. Arch Orthop Trauma Surg 1996;115:63– 67. 10. Williams IF, Mccullagh KG, Silver IA. The distribution of types I and III collagen and fibronectin in the healing equine tendon. Connect Tissue Res 1984;12:211–227. 11. Jo´ zsa LG, Kannus P. Human tendons: anatomy, physiology and pathology. Champaign: Human kinetics, 1997:68 –76. 12. Birk DE, Mayne R. Localization of collagen types I, III and V during tendon development. Changes in collagen types I and III are correlated with changes in fibril diameter. Eur J Cell Biol 1997;72:352–361. 13. Goncalves-Neto J, Witzel SS, Teodoro WR, Carvalho-Junior AE, Fernandes TD, Yoshinari HH. Changes in collagen matrix composition in human posterior tibial tendon dysfunction. Joint Bone Spine 2002;69:189 –194. 14. Olsen BR, Ninomiya Y: Collagens. In: Kreis T, Vale R, eds. Guidebook to the extracellular matrix, anchor, and adhesion proteins. 2nd ed. New York: Oxford University Press, 1999: 380 – 407. 15. Thomopoulos S, Hattersley G, Rosen V, Mertens M, Galatz L, Williams GR, et al. The localized expression of extracellular matrix components in healing tendon insertion sites: an in situ hybridization study. J Orthop Res 2002;20:454 – 463.
Oshiro et al / Flexor Tendon Healing in the Rat 16. Sternlicht MD, Werb Z. ECM proteinases. In: Kreis T, Vale R, eds. Guidebook to the extracellular matrix and adhesion proteins. 2nd ed. New York: Oxford University Press, 1999: 505–563. 17. Gotoh M, Hamada K, Yamakawa H, Tomonaga A, Inoue A, Fukuda H. Significance of granulation tissue in torn supraspinatus insertions: an immunohistochemical study with antibodies against interleukin-1, cathepsin D, and matrix metalloprotease-1. J Orthop Res 1997;15:33–39. 18. Ireland D, Harrall R, Curry V, Holloway G, Hackney R, Hazleman B, et al. Multiple changes in gene expression in chronic human Achilles tendinopathy. Matrix Biol 2001;20: 159 –169. 19. Choi H-R, Kondo S, Hirose K, Ishiguro N, Hasegawa Y,
823
Iwata H. Expression and enzymatic activity of MMP-2 during healing process of the acute supraspinatus tendon tear in rabbits. J Orthop Res 2002;20:927–933. 20. Armstrong DG, Jude EB. The role of matrix metalloproteinases in wound healing. J Am Podiatr Med Assoc 2002;92: 12–18. 21. Salo T, Ma¨ kela¨ M, Kylma¨ niemi M, Autio-Harmainen H, Larjava H. Expression of matrix metalloproteinase-2 and -9 during early human wound healing. Lab Invest 1994;70: 176 –182. 22. Riley GP, Curry V, DeGroot J, van El B, Verzijl N, Hazleman BL, et al. Matrix metalloproteinase activities and their relationship with collagen remodelling in tendon pathology. Matrix Biol 2002;21:185–195.
Purpose: To establish a rat flexor tendon laceration and repair model to investigate the molecular mechanisms of flexor tendon healing. Methods: Surgery was performed on rat flexor digitorum longus tendons from both hind feet. Repaired tendons were harvested at 0, 3, 7, 14, 21, 28, 42, 56, and 84 days after surgery. Histologic study (first 84 days) and gene expression study (first 28 days) of several collagens and matrix metalloproteinases (MMPs) were performed. Results: In the histologic study pre-existing collagen bundles were degraded between days 7 to 21. Newly formed collagen fibers crossed the repair site by day 28. Remodeling of the collagen fibers continued until day 84. Gene expression of type I collagen decreased initially and then returned gradually to the initial level by day 28, whereas expression levels of types III, V, and XII collagen were increased after surgery. The expression levels of MMP-9 and MMP-13 peaked between days 7 to 14, whereas MMP-2, MMP-3, and MMP-14 levels increased after surgery and maintained high levels until day 28. Conclusions: The rat tendon laceration model represented the entire tendon healing process. The results of this study suggest that MMP-9 and MMP-13 participate only in collagen degradation, whereas MMP-2, MMP-3, and MMP-14 participate not only in collagen degradation but also in collagen remodeling. (J Hand Surg 2003;28A:814-823. Copyright © 2003 by the American Society for Surgery of the Hand.) Key words: Flexor tendon healing, rat, histology, collagen, matrix metalloproteinase.
The improved understanding of flexor tendon healing has led to advances in tendon repair techniques and rehabilitation protocol, which in turn markedly From the Department of Orthopaedic Surgery, Barnes-Jewish Hospital, Washington University, St. Louis, MO; and St. Louis VA Medical Center, St. Louis, MO. Received for publication February 18, 2003; accepted in revised form June 23, 2003. Supported by Shriners Hospitals for Children, Grant # 8510. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. Reprint requests: Jueren Lou, MD, Department of Orthopaedic Surgery, Washington University School of Medicine, One Barnes Hospital Plaza, STE 11300, St. Louis, MO 63110. Copyright © 2003 by the American Society for Surgery of the Hand 0363-5023/03/28A05-0017$30.00/0 doi:10.1053/S0363-5023(03)00366-6
814
The Journal of Hand Surgery
have improved the clinical results of zone II flexor tendon injury.1 Recent developments in genomic and molecular biology research provide the opportunity to investigate the molecular events of tendon healing. Although the rat has been used as an experimental animal in various investigative fields it has been used infrequently for in vivo flexor tendon research.2,3 There are many advantages to using the rat as the experimental animal in the investigation of the molecular mechanisms of the tendon healing process. These include the variety of reagents to the rat that are commercially available; the rat genome sequence, which has been elucidated; and a large amount of biologic information that is available from literature databases. In this study a complete flexor tendon laceration
Oshiro et al / Flexor Tendon Healing in the Rat
815
each harvest time (Fig. 1B). The tendon specimens were approximately 4 mm in length including the repair site and did not include the proximal laceration site. We carefully excluded any fibrous adhesions if present along with the vincula structures from the specimens. The animal protocol was approved by the Animal Studies Committee at Washington University School of Medicine.
Histologic Evaluation Figure 1. Surgical procedure. (A) Arrows indicate the distal and the proximal lacerations. The distal laceration is repaired immediately with one simple stitch (Ethilon; Ethicon). (B) The tendon specimen including the repair site is harvested at each time point.
model in the rat was investigated. The histologic findings of the repaired rat tendon during the first 84 days are presented. Gene expression levels of several types of collagen and matrix metalloproteinases (MMPs) during the first 28 days also are examined by reverse-transcription polymerase chain reaction (PCR). The rat tendon injury model described in this study provides an option for investigating the mechanisms of the tendon healing process.
Methods Surgical Procedure Seventy male Lewis rats aged 7 to 10 weeks old (Charles River Laboratory, Wilmington, MA) were anesthetized with ketamine (Ketaset; Fort Dodge Animal Health, Fort Dodge, IA) (75.0 mg/kg of body weight injected intraperitoneally) and medetomidine hydrochloride (Domitor; Animal Health, Exton, PA) (0.5 mg/kg of body weight injected intraperitoneally). Under sterile conditions surgery was performed on the second through fifth toes of both hind feet using a surgical microscope (Fig. 1A). A midline incision was made on the plantar surface of each toe. The flexor sheath was opened between the proximal annular pulley and the distal interphalangeal joint. The flexor digitorum longus tendon was exposed. A complete transverse laceration was made in zone II. The laceration was repaired immediately with one simple 9-0 stitch (Ethilon; Ethicon, Somerville, NJ). A second transverse laceration was made in the tendon 3.0 mm proximal to the repaired laceration to reduce tensile forces across the repair site. The flexor sheath was not repaired and the skin was closed with a continuous running 9-0 suture (Ethilon). After surgery rats were allowed to walk freely in cages without restriction. Healing tendons were harvested at
Flexor tendons from 2 rats (16 tendons) were harvested at each time point of 0 (nonlacerated normal tendon), 3, 7, 14, 21, 28, 42, 56, and 84 days after surgery. The macroscopic appearances of the repaired tendons were observed by microscope at a magnification of 20 times. Tendon specimens then were fixed in 10% buffered formalin, embedded in paraffin, sectioned longitudinally at 6 m thickness, and evaluated with standard hematoxylin-eosin staining.
Total RNA Extraction At least 30 flexor tendons from 4 rats were harvested at each time point including days 0 (nonlacerated normal tendon), 3, 7, 14, 21, and 28 after surgery. A greater number of the tendons were needed for day 0 (80 tendons from 10 rats) than the other harvest time to obtain enough total RNA. Tendon specimens were immersed immediately in lysis solution (4 mol/L guanidinium thiocyanate, 25 mmol/L sodium citrate, pH 7.0, 0.5% sodium N-lauroyl sarcosine, and 0.1 mol/L mercaptoethanol). Tendon specimens in lysis solution were frozen and stored in liquid nitrogen until use. Pooled specimens were homogenized (Polytron PT3000; Brinkmann Instruments, Inc., Westbury, NY) in 4.0 mL of lysis solution. Four milliliters of buffer-saturated phenol (pH 4.3) and 0.8 mL of chloroform-isoamyl alcohol were added to each of the homogenates. The samples were mixed, placed on ice for 15 minutes, and centrifuged at 15,000 ⫻ g for 20 minutes at 4°C. The aqueous phase was collected and added with the same volume of isopropanol and kept at ⫺20°C overnight. The RNA was precipitated by centrifuge at 15,000 ⫻ g for 20 minutes at 4°C. The precipitated RNA was washed with 70% ethanol, dissolved in diethyl pyrocarbonate-treated water, and stored at ⫺80°C until use.
Reverse-Transcription Polymerase Chain Reaction Gene expression levels of types I, III, V, and XII collagen and MMP-2, -3, -9, -13, and -14 were ex-
816
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Table 1. Primer Sequences and PCR Conditions of Collagen and MMP Genes
Products GAPDH
GenBank Accession # M17701
Collagen alpha1 type I (Col I) Z78279 Collagen alpha1 type III (Col III) M21354
Primer Sequences
Annealing Number Size Temperature of (bp) (°C) Cycles
Forward 5⬘-ACCACAGTCCATGCCATCAC-3⬘ Reverse 5⬘-TCCACCACCCTGTTGCTGTA-3⬘
452
55
24
Forward 5⬘-CATGTCTGGTTTGGAGAGAG-3⬘ Reverse 5⬘-TCCATTCCGAATTCCTGGTC-3⬘
404
50
21
Forward 5⬘-GATGGATCAAGTGGACATCCAGGTCCCATT-3⬘ Reverse 5⬘-ATCTTGCAGCCTTGGTTAGGATCAACCCAG-3⬘
650
64
25
344
62
27
530
62
34
328
64
24
462
53
26
458
62
24
479
62
23
440
64
25
Collagen alpha1 type V (Col V) AJ005394 Forward 5⬘-CATGCCCAGAGAGAACAAAGGGAAAGAG-3⬘ Reverse 5⬘-TCCATCGGAAAGGCACATGTGGAATGAC-3⬘ Collagen alpha1 type XII (Col XII) U57362 Forward 5⬘-CAGAGTCGGTTCATCCACAAGTGCTCCTAT-3⬘ Reverse 5⬘-CACAAACATGCAATTCCATCTCGGCAGCAG-3⬘ MMP-2 U65656 Forward 5⬘-ATACAGGTGTGCCAAGGTGGAAACCAGAGA-3⬘ Reverse 5⬘-CCCATGGGGAACTGTTAAAGGGAGAAGCAA-3⬘ MMP-3 X02601 Forward 5⬘-TGAAACCGTCCAGAAGATCG-3⬘ Reverse 5⬘-AGCAGTGCTTCTGAATGTCC-3⬘ MMP-9 U24441 Forward 5⬘-GCTGCACCACCTTACCGGCCCTTTTATTTA-3⬘ Reverse 5⬘-TGGTTATCCTTCCACTGAGGGATCATCTCG-3⬘ MMP-13 M60616 Forward 5⬘-GATGTCAGGCACTAAAGGAAGGGGATAACC-3⬘ Reverse 5⬘-GCACCAAGTGTCAGTCACTAAGGAAAGCAG-3⬘ MMP-14 X83537 Forward 5⬘-CTTGCTGTCTTCTTCTTTAGACGCCATGGG-3⬘ Reverse 5⬘-TGCCAACTGGGTACACTTGGTACGTATAGG-3⬘
amined using reverse-transcription PCR. One microgram of total RNA was reverse transcribed (1stStrand cDNA Synthesis Kit; Roche, Indianapolis, IN) in 20 L of reaction volume, containing 1 mmol/L of dNTP, 10 mmol/L of Tris, 50 mmol/L of KCl (pH 8.3), 5 mmol/L of MgCl2, 50 U of RNase inhibitor, 1.6 g of oligo-p(dT)15 primer, and 20 U of avian myoblastosis virus reverse transcriptase. Incubation was performed for 60 minutes at 42°C. One microliter of 1st-Strand cDNA was amplified in 20 L of reaction volume containing PCR buffer (10 mmol/L Tris-HCl, 1.5 mmol/L MgCl2, 50 mmol/L KCl, pH 8.3), 0.2 mmol/L of each dNTP, 1.25 units of Taq DNA polymerase (Roche), and 0.25 mol/L of forward and reverse gene-specific primer set. Polymerase chain reaction amplification was done (DNA Thermal Cycler 480; PerkinElmer, Inc., Wellesley, MA) under the following conditions: preheating for 5 minutes at 94°C, then for each cycle denaturation for 15 seconds at 94°C, annealing for 45 seconds, extension for 1 minute at 72°C, with final extension for 2 minutes at 72°C. The annealing tem-
perature and the number of PCR cycles were arranged according to gene-specific primer sets (Table 1). Aliquots of PCR products were separated on 1.5% agarose gel and visualized with ultraviolet light after ethidium bromide staining. Gel images were scanned by a digital camera and saved in tiff file formats. Densitometric measurement of gel images was performed (Scion Image version beta 4.0.2; Scion Corporation, Frederick, MD). Band density of PCR products was normalized to that of glyceraldehyde3-phosphate-dehydrogenase (GAPDH).
Results Histologic Evaluation Macroscopic findings. The tendon repair site gradually healed throughout the 84 days after surgery without any obvious gap formation at the repair site (Fig. 2). The surfaces of the tendon specimens were smooth and had a glistening appearance throughout the time course. In all tendon specimens there was no indication of necrotic changes in the tendon segment between the 2 lacerations. Hypertrophic changes in
Oshiro et al / Flexor Tendon Healing in the Rat
817
Figure 2. Macroscopic appearance of the repaired tendons. Stitches (Ethilon; Ethicon) indicate the repair site. No obvious gap formations are noted. The surfaces of the tendon specimens are smooth and have a glistening appearance throughout the time course. Hypertrophic changes in the repair site first were observed on day 7, peaked at day 21, and gradually reduced.
the repair site first were observed on day 7, peaked at day 21, and gradually reduced. Mild to moderate adhesion formation was observed between the repair site and the surrounding tissues during days 7 to 84, and appeared to peak at day 21. The repair site was dissected easily, however, throughout the entire time course. The adhesion formation between the proximal cut tendon end (second laceration site) and surrounding tissues was more severe than those at the repair site.
Microscopic findings. Day 0 (nonlacerated normal tendon). The surface of the flexor tendon was characterized by a 1- to 2-cell–thick layer of the epitenon fibroblasts, which had spindle or slendershaped nuclei. In the deep region the endotenon fibroblasts, which had round to oval-shaped nuclear and clear cytoplasm, were aligned linearly between the longitudinal mature collagen bundles. Multiple sections of nonlacerated normal tendons failed to reveal any distinct vascular structures in the deep region. The vascular structures were noted only in the epitenon, mostly on the dorsal surface (Fig. 3A).
Day 3. In the superficial regions near the repair site there was a thickening of the epitenon to 3- to 4-cell–thick layers. In the repair site the lacerated tendon ends remained in close approximation. In the region near the repair site the cellularity of the endotenon fibroblasts was increased (Fig. 3B). There was no evidence of newly formed vascular structures at the repair site.
Day 7. The lacerated tendon was covered superficially by a several-cell–thick layer of epitenon (Fig. 3C). The repair site was united superficially at this time point. In the deep region of the repair site the severed ends of the collagen bundles began to degrade into thin fibers. The cellularity of the endotenon fibroblasts further increased in the region close to the repair sites. In the repair site there was no sign of cell migration from the surface region toward the deep region. In the areas within the repair site newly formed capillary vessels were observed (Fig. 3D).
Day 14. The collagen bundles were degraded massively in the large area of the repair site, replaced by randomly arranged fine strands of collagen fibers. Particular note was given to numerous large round cells (indicated by arrows on Fig. 4B), which were observed within the tendon (Fig. 4A and B).
Day 21. The collagen bundles were degraded almost completely in the entire region of the repair site. The original collagen bundles remained only in the distal region near the tendon insertion into bone. The number of large round cells increased. In the region subjacent to the tendon surface the longitudinally aligned cells and newly synthesized fine collagen fibers were observed to cross the repair site in a few tissue sections (Fig. 4C). It was difficult to define the site of the laceration clearly.
Day 28. In the repair site there was increased concentration of collagen fibers. The repair site was
818
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Figure 3. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 0 –7). Arrowheads indicate the site of the repaired laceration. (A) Day 0 (original magnification, 400⫻). Epitenon fibroblasts (small arrow) have spindle-or slender-shaped nuclei. Endotenon fibroblasts (large arrows) have round to oval-shaped nuclei and clear cytoplasm. (B) Day 3 (magnification, 100⫻). The lacerated tendon ends remain in close approximation. In the region near the repair site the cellularity of the endotenon fibroblasts has increased. (C) Day 7 (magnification, 100⫻). The repair site is united superficially. Asterisks indicate where the collagen bundles degrade into thin fibers. (D) Day 7 (magnification, 400⫻). Newly formed capillary vessels (arrows) are observed.
crossed completely by parallel alignment of fine collagen fibers and interposed cells. The large round cells began to change shape into more elongated slender cells, suggestive of mature endotenon fibroblasts (Fig. 4D). The cellularity and the vascularity of the repair site peaked at this time point.
Day 42. By this time point the tendon repair site was not distinguishable and only the suture materials suggested the region of the repair. The collagen fibers crossing the repair site showed maturation to collagen bundles (Fig. 5A). The slender-shaped endotenon fibroblasts were apparent between the mature collagen bundles (Fig. 5B).
Days 56, 84. Collagen bundles progressively matured to normal thickness. The cellularity of the endotenon fibroblasts gradually decreased to a normal
level by day 84 (Fig. 5C and D). The endotenon fibroblasts were more slender and less plump than that of the normal tendon (day 0). The amount of vascular structures in the repair site showed a gradual decrease, however, they were still visible on day 84.
Gene Expression Study Using ReverseTranscription Polymerase Chain Reaction Agarose gel images and the relative expression levels (RELs) of selected genes against GAPDH are shown in Fig.ure 6. The REL of type I collagen (Col I) gene was initially high (day 0), decreased to approximately half that level on day 3, and gradually returned to the initial level by day 28. The RELs of types III, V, and XII collagen (Col III, V, XII, respectively) genes were initially low, started to increase during days 3 to 7, peaked at day 14, and
Oshiro et al / Flexor Tendon Healing in the Rat
819
Figure 4. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 14 –28). Arrowheads indicate the site of the repaired laceration. (A) Day 14 (original magnification, 100⫻). Collagen bundles are massively degraded (asterisk). (B) Day 14 (magnification, 400⫻, the region of asterisk on A). Collagen bundles are replaced by randomly arranged newly synthesized fine collagen fibers. Large round cells are observed (arrows). (C) Day 21 (magnification, 400⫻, repair site subjacent to tendon surface). Longitudinally aligned cells and fine collagen fibers cross the repair site. (D) Day 28 (magnification, 400⫻). Repair site is crossed completely by parallelly aligning fine collagen fibers and interposed cells, which have changed shape from larger and round to more slender (arrows).
gradually decreased, but still remained at a higher level on day 28 than the day 0 level. There were 2 MMP gene expression patterns. The RELs of all examined MMP genes were initially low. In the first pattern the gene expression levels of MMP-9 and MMP-13 were very high on days 7 and 14, and decreased by day 21. In the second pattern the RELs of MMP-2, MMP-3, and MMP-14 genes increased after surgery and maintained high levels until day 28.
Discussion Although several animal experimental models including the dog, rabbit, and chicken have been designed for in vivo flexor tendon research, there are few reports using rat flexor tendons.2,3 Because of their small size (1-mm width), the surgery may be
difficult to perform and postoperative immobilization may be difficult to attain. The rat tendon laceration and repair model developed in this study, however, allowed surgery on rat flexor tendons to be performed without difficulty using the surgical microscope. Furthermore postoperative immobilization was obviated by cutting the tendon proximal to the laceration site. On the other hand there are several advantages to using the rat as an experimental animal for tendon research, particularly in investigating the molecular mechanisms of the repair process. First, rats are less expensive and easier to handle compared with the other animal species. Second, many important and different types of reagents to the rat are now available. Finally, the rat genome sequence has been elucidated and currently can be obtained online (www. ncbi.nlm.nih.gov/genome/guide/rat).
820
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
Figure 5. Microscopic appearance of the repaired tendons (hematoxylin-eosin, days 42– 84). Arrowhead indicates the site of the repaired laceration. (A) Day 42 (original magnification, 100⫻). The collagen fibers mature to collagen bundles. (B) Day 42 (magnification, 400⫻). The slender-shaped endotenon fibroblasts are present between mature collagen bundles (arrows). (C) Day 56 (magnification, 100⫻). Collagen bundles are more mature compared with day 42 (Fig. 4A). (D) Day 84 (magnification, 400⫻). The collagen bundles and the cellularity of the endotenon fibroblasts are similar to normal. Endotenon fibroblasts (arrows) are more slender and less plump than that of the normal tendon.
Histology This rat model presented the entire tendon healing process, which includes the superficial repair of the repair site, collagen degradation, synthesis of new collagen fibers, collagen remodeling, and neovascularization. This healing process is for the most part consistent with the results of previous reports using other animal species.4 – 6 At day 14 the repair site was massively replaced by degraded collagen fibers and large round cells. These large round cells increased until day 28 and then decreased gradually. These cells changed shape into more slender endotenon fibroblasts, which were located between the mature collagen bundles. These cells seemed to play the central role as reparative cells in the rat tendon healing process, including the degradation of pre-existing collagen, synthesis, and remodeling of new collagen fibers. The origin of this reparative cell was unclear. Investigators of previous studies have had various opinions about their origin.
With a canine flexor tendon healing model, Potenza4 concluded that the tendon cells played no active role in the repair process and that repair was completed entirely by the cellular activity of the sheath and surrounding tissue. With a chicken model, Lindsay and Thomson5 showed that tendon sheath contributed little to the tendon repair; however, all of the connective tissue cells that make up the tendon participated in the repair process. Umeda6 investigated the repair process of chicken tendons within an intact portion of the synovial sheath and suggested that the main origin of fibroblasts of the granulation tissue were perivascular mesenchymal cells. From the observations of the divided canine tendon, followed by controlled mobilization, Gelberman et al7 showed that epitenon fibroblasts migrated to the region with moderate gaps; however, when the tendon ends were held in close approximation there was minimal response from the epitenon and a greater initial response from endotenon fibroblasts. In this study the
Oshiro et al / Flexor Tendon Healing in the Rat
821
Fig.ure 6. Agarose gel images and the relative gene expression levels against GAPDH (x axis, days after surgery; y axis, relative expression levels). Col I gene expression level is initially high, temporally decreases on day 3, and gradually recovers to a normal level. The expression levels of Col III, V, and XII genes peak at day 14 and gradually decrease but still remain at a higher than normal level on day 28. The gene expression levels of MMP-9 and MMP-13 were high on days 7 and 14. Expression levels of the MMP-2, MMP-3, and MMP-14 genes increased after surgery and maintained high levels until day 28.
repaired tendon ends remained in close approximation in almost all tendon specimens, and no migrating epitenon cells could be identified. Additionally the findings on day 3 revealed the increased cellularity of endotenon fibroblasts in the region adjacent to the repair site without neovascularization. This indicated an early response of pre-existing endotenon fibroblast proliferation and/or migration toward the repair site. Thus we speculate that a principal component of the reparative cells noted in this study comes from pre-existing endotenon fibroblasts.
Gene expression— collagen. Although many investigators point out the up-regulation of type III collagen in the early tendon healing stage,8 –10 marked up-regulation of types V and XII collagen gene expression has not been reported. Type V collagen has been found in vascular walls of the tendon,11 immature chick embryonic tendon,12 and degenerated human adult tendons.13 It is postulated that type V collagen is associated closely with the vascular generation and may play a role in tendon growth and healing. Recently type XII collagen has been noted to have functional interaction with type I collagen and proteoglycans.14,15 Type XII collagen may
serve as a molecular bridge between fibrils, as well as between fibrils and other extracellular matrix (ECM) constituents.14 Although the role of this molecule in tendon healing remains to be defined further, a possible mechanism of type XII collagen is to stabilize the reparative tissue components by keeping the newly synthesized type I collagen fibrils together and linking between these fibrils and proteoglycans.
Gene expression—MMP. MMPs degrade several substances in the ECM and are considered integral mediators of ECM remodeling that takes place during tissue development, morphogenesis, and repair.16 Although MMPs were reported to be present in the human degenerated supraspinatus17 and Achilles tendons,18 as well as in the torn rabbit supraspinatus tendon,19 little is known of their behavior in the flexor tendon healing process within the digital sheath. Gene expression study revealed 2 different gene expression patterns of MMPs during the first 28 days. MMP 13 (collagenase 3) functions to cleave the triple helix of fibrillar collagens, types I, II, and III.16,20 MMP-9 (gelatinase B) is derived predominantly from inflammatory cells and degrades denatured collagens as well as other ECMs.20 In degrad-
822
The Journal of Hand Surgery / Vol. 28A No. 5 September 2003
ing the pre-existing collagen bundles, MMP-9 may follow the function of MMP-13 by further degrading fibrillar collagens.16,20,21 As noted in the histologic findings the degradation of pre-existing collagen bundles first was observed on day 7 and almost completed by day 21; reparative fibroblast proliferation and realignment of the newly synthesized collagen fibers began on day 21. It is possible that the primary roles of the MMP-9 and MMP-13 are accomplished by day 21 and they contribute minimally to the subsequent remodeling phase. The exhaustion of the target substrates, namely, damaged helical collagens, might trigger the down-regulation of MMP-9 and MMP-13 gene expression. The decrease of MMP-9 producible inflammatory cells might be another possible reason for the down-regulation of MMP-9 gene expression. MMP-2 (gelatinase A) has a similar function to MMP-9 but it generally is secreted from fibroblasts. MMP-3 (stromelysin 1) has a broad substrate range and activates other MMPs and growth factors. MMP-3 also is thought to have the potential to play a major role in ECM remodeling.18 MMP-14 (membrane-type MMP-1) exists on cell membranes and has been implicated as a key player in a proteinase cascade involving MMP-2, MMP-9, and MMP-13.16 Our results suggest that these 5 MMPs are involved in tendon ECM degradation: MMP-9 and MMP-13 make little contribution to the subsequent ECM remodeling; conversely MMP-2, MMP-3, and MMP-14 might have the potential to play key roles during the remodeling phase of tendon healing. It should be noted that this study focused only on MMP gene expression level, which does not represent the enzyme activity. All MMPs are synthesized as inactive pro-MMPs,16 which then are converted to an active form by various factors.20 Nevertheless this study potentially increases our understanding of the molecular mechanisms involved in the tendon repair process. Further studies using substrate gel electrophoresis and fluorogenic substrate22 are needed to evaluate the enzyme activities of MMPs. Although our rat model presented the entire healing process without gap formation it is important to note that the model did not entirely represent the clinical condition. Tensile stress across the repair site was absent because of the second laceration proximal to the repair site. Also the progressive adhesion formation at the more proximal laceration site may have prevented passive gliding of the repaired tendons; however, passive gliding still might have been
achieved, at least in the initial stage, by allowing the rat to cage walk without restriction. Nevertheless this model does provide an opportunity to study the biologic processes of tendon healing.
References 1. Strickland JW. Development of flexor tendon surgery: twenty-five years of progress. J Hand Surg 2000;25A:214 – 235. 2. Greenlee TK Jr, Pike D. Studies of tendon healing in the rat. Remodeling of the distal stump after severance. Plast Reconstr Surg 1971;48:260 –270. 3. Lane JM, Black J, Bora FW Jr. Gliding function following flexor-tendon injury. A biomechanical study of rat tendon function. J Bone Joint Surg 1976;58A:985–990. 4. Potenza AD. Tendon healing within the flexor digital sheath in the dog. An experimental study. J Bone Joint Surg 1962; 44A:49 – 64. 5. Lindsay WK, Thomson HG. Digital flexor tendons: an experimental study. Part I. The significance of each component of the flexor mechanism in tendon healing. Br J Plast Surg 1960;12:289 –316. 6. Umeda HT. An experimental study on the intrinsic healing mechanism of digital flexor tendons. J Jpn Orthop Assoc 1978;52:917–929. 7. Gelberman RH, Vandeberg JS, Manske PR, Akeson WH. The early stages of flexor tendon healing: a morphologic study of the first fourteen days. J Hand Surg 1985;10A:776 – 784. 8. Sakai H, Koibuchi N, Ohtake H, Tamai K, Fukui N, Oda H, et al. Type I and type III procollagen gene expressions in the early phase of ligament healing in rabbits: an in situ hybridization study. J Orthop Res 2001;19:132–135. 9. Hiranuma K, Suzuki K, Hirata K, Nakamura H, Higashi K, Hirano H. Extracellular matrices in peritendinous connective tissue after surgical injury to the chicken flexor tendon. Arch Orthop Trauma Surg 1996;115:63– 67. 10. Williams IF, Mccullagh KG, Silver IA. The distribution of types I and III collagen and fibronectin in the healing equine tendon. Connect Tissue Res 1984;12:211–227. 11. Jo´ zsa LG, Kannus P. Human tendons: anatomy, physiology and pathology. Champaign: Human kinetics, 1997:68 –76. 12. Birk DE, Mayne R. Localization of collagen types I, III and V during tendon development. Changes in collagen types I and III are correlated with changes in fibril diameter. Eur J Cell Biol 1997;72:352–361. 13. Goncalves-Neto J, Witzel SS, Teodoro WR, Carvalho-Junior AE, Fernandes TD, Yoshinari HH. Changes in collagen matrix composition in human posterior tibial tendon dysfunction. Joint Bone Spine 2002;69:189 –194. 14. Olsen BR, Ninomiya Y: Collagens. In: Kreis T, Vale R, eds. Guidebook to the extracellular matrix, anchor, and adhesion proteins. 2nd ed. New York: Oxford University Press, 1999: 380 – 407. 15. Thomopoulos S, Hattersley G, Rosen V, Mertens M, Galatz L, Williams GR, et al. The localized expression of extracellular matrix components in healing tendon insertion sites: an in situ hybridization study. J Orthop Res 2002;20:454 – 463.
Oshiro et al / Flexor Tendon Healing in the Rat 16. Sternlicht MD, Werb Z. ECM proteinases. In: Kreis T, Vale R, eds. Guidebook to the extracellular matrix and adhesion proteins. 2nd ed. New York: Oxford University Press, 1999: 505–563. 17. Gotoh M, Hamada K, Yamakawa H, Tomonaga A, Inoue A, Fukuda H. Significance of granulation tissue in torn supraspinatus insertions: an immunohistochemical study with antibodies against interleukin-1, cathepsin D, and matrix metalloprotease-1. J Orthop Res 1997;15:33–39. 18. Ireland D, Harrall R, Curry V, Holloway G, Hackney R, Hazleman B, et al. Multiple changes in gene expression in chronic human Achilles tendinopathy. Matrix Biol 2001;20: 159 –169. 19. Choi H-R, Kondo S, Hirose K, Ishiguro N, Hasegawa Y,
823
Iwata H. Expression and enzymatic activity of MMP-2 during healing process of the acute supraspinatus tendon tear in rabbits. J Orthop Res 2002;20:927–933. 20. Armstrong DG, Jude EB. The role of matrix metalloproteinases in wound healing. J Am Podiatr Med Assoc 2002;92: 12–18. 21. Salo T, Ma¨ kela¨ M, Kylma¨ niemi M, Autio-Harmainen H, Larjava H. Expression of matrix metalloproteinase-2 and -9 during early human wound healing. Lab Invest 1994;70: 176 –182. 22. Riley GP, Curry V, DeGroot J, van El B, Verzijl N, Hazleman BL, et al. Matrix metalloproteinase activities and their relationship with collagen remodelling in tendon pathology. Matrix Biol 2002;21:185–195.