ELSEVIER Journal of Orthopaedic Research 21 (2003) 798-804

Journal of Orthopaedic

Research www.elsevier.com/locate/orthres

Gamma interferon as an antifibrosis agent in skeletal muscle

William Foster a, Yong Li a, Arvydas Usas a, George Somogyi b, Johnny Huard a Growth and Development L‘aborutory, Department of Orthopuedic Surgery, 4151 Rungos Research Center, Children’s Hospital of Pittsburgh

and University of Pittsburgh, 3705 Fifth Avenue, Pittsburgh. PA 15213-2583, USA Pharmacology, University of Pittsburgh, Pittsburgh, PA 15213, USA

Department of’ Molecular Genetics und Biochemistry, University qf Pittsburgh, Pittsburgh, PA 15213, US.4

Received 23 December 2002; accepted 24 January 2003

Abstract

Muscle injuries are a common problem in sports medicine. Skeletal muscle can regenerate itself, but the process is both slow and incomplete. Previously we and others have used growth factors to improve the regeneration of muscle, but the muscle healing was impeded by scar tissue formation. However, when we blocked the fibrosis process with decorin, an antifibrosis agent, we improved the muscle healing. Here we show that yinterferon (y1NF)-a cytokine that inhibits the signaling of transforming growth factor PI (TGFPI), a fibrotic stimulator-reduces fibrosis formation and improves the healing of lacerated skeletal muscle. With yINF treatment, the growth rate of muscle-derived fibroblasts was reduced and the level of fibrotic protein expression induced by TGFPl (including TGFPI, vimentin, and a-smooth muscle actin) was down-regulated in vitro. In a mouse laceration model, the area of fibrosis decreased when yINF was injected at either 1 or 2 weeks after injury. More importantly, the injection of yINF at either 1 or 2 weeks post-injury was found to improve muscle function in terms of both fast-twitch and tetanic strength. This study demonstrates that yINF is a potent antifibrosis agent that can improve muscle healing after laceration injury. 0 2003 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved.

Keywords: Muscle injuries; Fibrosis; Muscle regeneration; Transforming growth factor PI ; Gamma interferon

Introduction

Muscle injuries occur frequently and account for up to 30% of all injuries seen in sports medicine [S]. Although skeletal muscle is able to regenerate itself, the healing process is both slow and incomplete [ 1 1,13,15,18]. Clinically, these injuries are treated with the RICE principle (i.e., Rest, Ice, Compression, and Elevation) [ 171 and by non-steroidal anti-inflammatory drugs (NSAID) [23,31], but still there is a high rate of injury recurrence [4,25,34].

In cases of mechanical trauma, the integrity of myofibers’ basal lamina is disrupted, which eventually leads to necrosis [11,18]. Macrophages invade the af- flicted area to remove necrotic myofibers, while lym- phocytes secrete various growth factors and cytokines into the injured area [11,18]. Satellite cells then are ac- tivated, proliferate, and fuse into/with multinucleated myotubes in order to regenerate the skeletal muscle

*Corresponding author. Tel.: +412-692-782217807; fax: 412-692-

E-mail uddress: jhuard+@pitt.edu (J. Huard). 7095.

[ 10,11,18]. Unfortunately muscle is incapable of suffi- cient regeneration following severe injuries, partially because scar tissue develops in the injured area and in- creases over time [6,7,11,18].

Several approaches have been utilized to improve the muscle healing process, but they have not been suc- cessful. The use of NSAID to reduce inflammation has not been shown to improve the strength of injured muscle beyond one day after injury [23]. Similarly, at- tempts to improve muscle regeneration through appli- cation of growth factors [e.g., leukemia inhibitory factor (LIF), hepatocyte growth factor (HGF), basic fibroblast growth factor (bFGF), nerve growth factor (NGF), and insulin-like growth factor I and I1 (IGF-1 and -2)] have not been shown to elicit full recovery of injured skeletal muscle [10,11,14-16,18,21]. One of the most promising growth factors, IGF- 1, increases the proliferation of myoblasts in injured normal muscle, but the degree of muscle regeneration is still limited by the accumulation of scar tissue [14,21]. Another possible reason for which the use of growth factors to accelerate muscle regener- ation has not resulted in complete muscle recovery after injury is that some myogenic cells, including myofibers,

0736-0266/$ - see front matter 0 2003 Orthopaedic Research Society. Published by Elsevier Ltd. All rights reserved. doi: 10.1016/S0736-0266(03)00059-7

W. Foster et al. / Journal of Orthopaedic Research 21 (2003) 798404 799

have been shown to differentiate into fibrotic cells during this phase [20]. Consequently, increasing the number of myoblasts through growth factor stimulation also aug- ments the number of cells that could potentially differ- entiate into fibrotic cells.

Given this scenario, an alternative approach was de- veloped to improve muscle healing by blocking the fi- brosis pathway. The overproduction of transforming growth factor Pl (TGFPl) is a major cause of tissue fibrosis after injury or disease in various tissues [2]. In previous studies, we have shown that TGFPl is ex- pressed in injured skeletal muscle [20] and that a TGFPl inhibitor, decorin, can effectively prevent fibrosis in in- jured skeletal muscle and thereby improve muscle heal- ing to near-complete functional recovery [6]. These studies have demonstrated that inhibition of TGFPl can block the fibrosis process and thus improve the healing of injured skeletal muscle. Recently yinterferon (yINF), a TGFPl pathway inhibitor, has demonstrated the ca- pacity to disrupt TGFPl signal transduction [32]. yINF already had been shown to reduce fibrosis in vari- ous tissues, including the liver, lung, kidney, and skin [24,30,36]. The primary goals of this study were to de- termine whether yINF can effectively block fibrosis and enhance subsequent healing in injured skeletal muscle.

Methods and materials

In iCtro

Cell culture Primary muscle-derived fibroblasts (preplate # I ) were isolated via

the preplate technique as previously described [6,20,27]. In brief, after the muscle underwent a series of enzymatic digestions, the cells were plated into a 25 cm2 collagen-coated flask. Two hours later the su- pernatant was removed; the remaining cell population, termed ‘pre- plate #I,’ was considered to be fibroblasts because the cells were >90% vimentin positive [20,27].

To test the ability of yINF to inhibit TGFpl in myoblasts, C2C12 cells were transfected with a pMAMneoTGFPl plasmid containing the human TGFbl gene under the control of the MMTV-LTR promoter and enhanced by RSV-LTR [19]. The cells then were selected for 2 weeks in 500 pg/ml of G418 in growth medium [lo% Fetal Bovine Serum, 10% Horse Serum, I%I Penicillin/Streptomycin, 0.5% Chicken Embryo Extract in Dulbecco’s Modified Eagle Medium (GibcoBrl)] and were termed ‘CT cells.’

Cell growth Primary muscle-derived fibroblasts and CT cells were used for the

cell growth studies. Fibroblasts and CT cells were grown in low se- rum medium [2Yn Fetal Bovine Serum and 1% Penicillin/Streptomycin in Dulbecco’s Modified Eagle Medium (GibcoBrl)] to reduce the ef- fects of growth factors and cytokines from the serum. These cells were plated into 6-well plates with different concentrations of yINF [0, 100, 500, or 1000 yINF U/ml (Intermune Pharmaceuticals U/ml)]. The cells then were trypsinized and counted on a hematocytometer [6]. Statistical analysis was performed via a Student’s t-test.

Western blot CT cells were plated in 25 cm’ flasks and separated into four

treatment groups with yINF stimulation for 0, 6, 24, or 48 h (1000 yINF Ulml). The cells were subsequently lysed (950 p1 Laemmli buf- fer + 50 PI bmercaptoethanol, BioRad), and separated by SDS poly-

acrylamide gel, which was transferred to a nitrocellulose membrane. Membranes subsequently were incubated with primary antibodies [anti-pactin 1:8000 (Sigma), ctSMA 1:lOOO (Sigma), vimentin 1:lOOO (Sigma), or TGFBl 1:lOOO (Novocastra)] for 2 h. Then the membranes were incubated with secondary antibodies antimouse IgG-conjugated HRP 1: 10000 (Pierce), antigoat IgG-conjugated HRP 1: I0000 (Pierce), or antirabbit IgG-conjugated HRP 1:5000 (Chemicom) for 1.5 h. The membranes then were developed (Supersignal West Pic0 Chemilumi- nescent Substrate, Pierce) and detected on X-ray film.

In uiuo

Fibrotic area and muscle regeneration Eighteen mice (C57BLIOJ+/+; 6 weeks of age, 15-20 g, three mice/

group) were used for this experiment. The policies and procedures of the animal facility are in accordance with those detailed by the United States Department of Health and Human Services. The Animal Re- search and Care Committee of the authors’ institution approved the research protocol for these experiments (protocol no. 5/01). A previ- ously reported muscle laceration model was utilized for these experi- ments: both the left and right gastrocnemius muscles were lacerated [6,20-221. The mice were anesthetized with 0.03 ml ketamine and 0.02 ml xylazine by intraperitoneal injection. Lacerations were performed through 50% of the width and IOO‘l/u of the thickness at 60‘!4 of the muscle length. After the laceration was performed, the skin was closed with 4-0 silk suture. After conducting pilot experiments to determine the optimum concentration, 250 U of yINF in 10 111 of PBS was in- jected into the lacerated area by microsyringe. At the same time the contralateral (control) lacerated area was injected with 10 1.11 PBS only. The mice were separated into two groups with the injections performed at either I or 2 weeks after laceration. All mice were sacrificed 4 weeks after laceration. The muscles were isolated, mounted, and frozen in liquid-nitrogen precooled 2-methylbutane and cryostat-sectioned (8 pm) for histological and immunohistochemical analysis.

In order to determine the area of fibrosis, total collagen staining and vimentin staining were performed. For collagen staining, sections were fixed in lo‘%) formalin for 10 min. After washing in deionized water, Masson modified IMEB trichrome staining was performed ac- cording to the manufacturer’s protocol (IMEB, Inc.). This histological technique produced a black color for nuclei, a red color for muscle, and a blue color for collagen. Vimentin was used as a marker of fibrosis as previously described [3,6,11,18]. Sections were fixed in 10% formalin for 10 min, and then underwent 45 min of blocking in 5% Horse Serum (Vector). The sections were incubated with anti-vimentin antibody (1:500, Sigma) conjugated with Cy3. The vimentin-positive area was also used to measure the fibrotic area as previously described [14,16,21]. Five fields within the injured area (three micdgroup) were measured using an Olympus Provis epifluorescence microscope (Olympus Optical Co., Ltd., Tokyo, Japan) and a Sony 970 chip CCD camera (Sony, Tokyo, Japan). Images then were digitized using a Coreco frame grabber board (Coreco Imaging. St Laurent, Que., Canada) and rendered to monochrome. Using Northern Eclipse ver- sion 6.0 (Empix Imaging, Mississauga, ON, Canada), the absolute area of fibrosis (vimentin) was measured for each field.

To determine the degree of muscle regeneration, muscle sections were stained with hematoxylin and eosin. Muscle regeneration was assessed by counting the number and diameter of centronucleated myofibers [6,16,21]. For every sample, five fields from the injured area were compared among the groups as described above. Northern Eclipse software again was used to count the number and diameter of centronucleated myofibers, and a Student’s 1-test was used for statis- tical analysis between groups.

Physiological testing Twelve mice were utilized for physiological strength testing. The left

gastrocnemius was lacerated as above, while the right leg was not in- jured and served as a control. The mice were separated into three in- jection groups [250 yINF U at I week after laceration. 250 U yINF at 2 weeks after laceration, and a laceration control (sham PBS injection) group]. Four weeks after injury, both gastrocnemius muscles were carefully removed and fixated at the bone tendon junction both proxi- mally and distally. Using a previously reported protocol, we tested the fast-twitch and tetanus strength of the gastrocnemius muscle [6,14,16,21]. The fast-twitch and tetanus strength of the muscles then

were corrected by the cross-sectional area to yleld force per unit cross- sectional area. Statistical analyses were performed via a Student's t-test.

c.) L 8 -

{ 7 -

o 6 - 8 5 - ; 4 - ; - 3 -

: 1 - ii! 2 -

* o

CT 6hr 12hr 24hr

*

TGF-81 Results

In vitro

yINF iizlzihition of mu~c.le-d~riven'~bro~lusts and CT cell growth (Fig 1 )

At 72 h, the control group (0 yINF U/ml) contained more cells (i.e., muscle-derived fibroblasts) than all other groups. In the groups treated with 100 and 500 U/ml there was a slight decrease in cell number when com- pared to the control group, and the 1000 U/ml group contained a significantly lower number of cells than the control group ( P < 0.05). Similarly, CT cells showed a dose-dependent inhibition of growth: high concentra- tions of yINF (1000 U/m1) significantly inhibited CT growth ( P < 0.05).

y INF treufnzent ci'owiz-reguluted Jlhrosis-reluted protein expression it? nzyogeiiic cell.$ expressing TGFPl (Fig. 2 )

CT cells expressed TGFPl due to TGFPl gene transfer. Western blot rekealed that the CT cells also expressed the proteins vimentin and aSMA, which are

Muscle Derived Fibroblast growth with 7-INF 2.5

T 1 *

0 100 500 1000

units of YlNF

CT cell growth with YlNF g 10 8 9 1 T

0 100 500 1000

units of YlNF

Fig. I. Effects of ylNF on the growth of muscle-derived fibroblasts and CT cells in vitro. Cultures grown for 72 h with 100 or 500 y l N F Uiinl showed a decrease in the number of fibroblasts, and those grown with 1000 yINF U/ml showed a significant decrease in the number of fi- broblasts or CT cells relative l o the control cultures ( ' P < 0.05).

rC Vimentin

4 a S M A

+ pactin

Fig. 2. yINF-mediated inhibition of myofibroblastic protein expression induced in myogenic cells by TGFBI. Western blot analysis for CT cells expressing TGFP1 grown with y I N F for different amounts of time (6. 12, or 24 h) revealed a decrease in TGFBI, vimentin. and XSMA protein expression.

myofibroblast markers that are associated with scar tissue formation [I ,3,20,26,28,29,33]. The CT cells then were treated with ylNF for 6, 12, or 24 h. We found that increasing the time of ylNF treatment resulted in de- creased expression of TGFPl, with longer times of in- cubation resulting in even lower TGFPl expression. Furthermore, we found that none of the ylNF-treated CT cells expressed a detectable amount of aSMA pro- tein. Vimentin bands decreased following 6 and 12 h of treatment, and after 24 h of treatment vimentin bands were undetectable. These findings underscore the ability of yINF to block the fibrotic effects of TGFPl on skeletal muscle cells in a time-dependent manner.

In uivo

The ,fibrosis urea decreased while muscle regeneration incrrmsed in injured muscle treated )2-ith yINF

The mice were sacrificed 4 weeks after laceration and various histological analyses were performed [e.g., vi- mentin immunohistochemistry, Masson modified tri- chrome (total collagen staining), and hematoxylin and eosin staining]. Masson trichrome staining revealed a large collagenous area in the control that decreased with injections of yINF at either 1 or 2 weeks after injury (Fig. 3A-F). A large vimentin-positive area also was found in the gastrocnemius of the untreated controls. Both injection time points, 1 or 2 weeks post-injury, resulted in a significant decrease in the vimentin-positive area when compared to that of the control ( I 12 849 and 132098 pm2 versus 354074 gm2, respectively, P < 0.01; Fig. 3G-J). In hematoxylin and eosin-stained muscles, the number of centronucleated myofibers significantly increased with yINF treatment ( P < 0.05), while the dia- meter of the centronucleated myofibers was unaffected.

W. Fo;ostw rt al. I Journul of Orthoppnedic Riwarrli 21 (2003) 798-804 80 I

J Area of Fibrosis 450000

3 400000

p 350000 .5 300000

250000 C

.- CI

g 200000

150000

2 100000 $ 0

50000

p= o, Lacerated Control 1 wk injection 2 wk injection

Fig. 3. Masson modified trichrome and vimentin staining of lacerated muscles injected with yINF. A large area of collagen deposition (blue) was visible in the lacerated control muscles (A,D), but the muscles injected with yINF at either I (B,E) or 2 weeks (C,F) after injury revealed a smaller amount of collagenous scar tissue. Four weeks after laceration, a large area of fibrosis (vimentin) was visible in the lacerated control muscles (G), whereas muscles injected with yINF at 1 (H) or 2 (I) weeks post-injury displayed a reduced amount of fibrosis at the same time point. Assessment of the area of fibrosis via Northern Eclipse software revealed a significant reduction in the vimentin-positive area of the injected muscles versus that of the lacerated control muscles (J) (*P < 0.01) (panels A-C, l00x magnification; panels D-I, 200x magnification).

The number of centronucleated myofibers in the 1-week post-injury yINF injection group was greater than the number in the control, and the number of centronucle- ated myofibers was even greater in the 2-week post- injury injection group (Fig. 4).

yINF injections improved the physiologicul properties o j injured skeletul muscle

At 4 weeks post-laceration, yINF displayed a benefi- cial effect in promoting the functional recovery of the

muscle. The fast-twitch strength of the muscle injected with yINF 1 week post-injury was slightly higher than the PBS-injected, lacerated control. However, when the muscle was injected with yINF 2 weeks post-injury, the muscle was significantly stronger than the control injured muscle ( P < 0.05), and nearly indistinguish- able from the normal non-injured control in terms of strength. The results of tetanic strength testing mirrored the results of the fast-twitch testing; yINF-injected muscle was stronger than PBS-injected muscle with

802 W. Foster et al. I Journal of Orthopaedic Research 21 (2003) 798404

D Number of Centronucleated Myofibers E Diameter of Centronucleated Myoflbers

Irce~t.d Control 1 w INF InjecUon 2 w INF lnjntlon l0eeRt.d Control 1 w INF Injection 2 w INF Injution

Fig. 4. Hematoxylin and eosin staining of lacerated muscle injected with yINF. Hematoxylin and eosin staining revealed a large area of scar tissue in the lacerated controls (A) containing very few centronucleated myofibers, whereas the muscles injected with yINF at 1 (B) or 2 weeks (C) post-injury contained a smaller area of fibrosis and an increased number of centronucleated myofibers. Although the diameter (E) of centronucleated myofibers was unaffected by y INF treatment, the number (D) of centronucleated myofibers significantly increased with treatment of yINF when compared to a sham control, with later injection time points resulting in higher numbers of centronucleated myofibers than did earlier injection time points ( 'P < 0.01) (panels A-C, lOOx magnification).

significant differences observed between the group re- ceiving injections 2 weeks after injury and the control group (P < 0.05). In summary, yINF improved both the fast-twitch and the tetanus strength of the injured muscle (Fig. 5).

Discussion

Muscle injuries occur by a variety of mechanisms, including direct injuries (e.g., contusion and laceration) and indirect injuries (e.g., strain, ischemia, and neuro- logical dysfunction) [6,7.11,13,15,16,18]. Usually, the muscle is able to regenerate itself after injury, but the process tends to be slow and incomplete [11,13-16,181. Following skeletal muscle injury, scar tissue forms and impedes muscle regeneration [7,11,15,18,20].

yINF is known to inhibit fibrosis in various tissues, including the liver, lung, kidney, and skin [ 12,24,30,36]. yINF has been shown to not only down-regulate en- dogenous collagen expression, but also to effectively block TGFPl -mediated increases in collagen protein levels [9]. Furthermore, yINF inhibits TGFPl signaling by inducing expression of SMAD 7, which participates in a negative feedback loop in the TGFPl signal transduction pathway [32]. These findings demon- strate yINF's ability to inhibit the actions of TGFPl. Blocking TGFP 1 signal transduction is important be- cause TGFPl promotes the myofibroblast phenotype as well as up-regulation of collagen and ECM pro- duction [5,33,35]. TGFPl has been implicated in the

fibrosis of many tissues, including the lung and kidney [2]. Recently, we have found that TGFPl is expressed during muscle injury [20]. A TGFPl inhibitor, decorin, has been shown to histologically and functionally im- prove lacerated-injured skeletal muscle healing [6]; these findings confirm the detrimental role played by TGFPl in the skeletal muscle healing process. Previ- ously it was determined that yINF is capable of in- hibiting the TGFP1-induced myofibroblast phenotype in palatal fibroblasts and in animal models of liver and kidney fibrosis [1,24,35]. Here we have shown that yINF can block the effects of TGFPl in myogenic cells. We used a clonal cell line CT (C2C12 cells transfected with a TGFPl plasmid) [I91 and treated the cells with yINF. Western blot analysis revealed that the myo- genic cells stimulated with TGFPl expressed the myo- fibroblastic proteins vimentin and USMA [26,28,29,33], in addition to the TGFPl protein. We further observed that the CT cells expressed decreased amounts of the TGFPl protein upon yINF treatment. This reduction in TGFPl production subsequently resulted in a de- crease in the expression of vimentin and aSMA. Al- though the myofibroblast phenotype at early time points post-injury is believed to play a beneficial role in the healing process, the long-term persistence of this cell phenotype is associated with scar tissue formation [26, 28,291. By inhibiting myofibroblasts and the effects of TGFP 1, the post-injury fibrosis process should be limited.

yINF is known to promote inflammation; however, we injected the mice with yINF at 1 or 2 weeks post- laceration and thereby avoided the inflammatory stage,

W. Fosier et ul. I Journal oj Orthopaedic Research 21 (2003) 798404

A Fast-twitch Strength B Tetanic Strength

803

5 9- 4.5 8-

0 4

x 3 2 2.5 € 2 t 3-

E 2- $ 1.5

1 0.5 1 -

0 0,

g 3.5 8 ;: :: .

Noimnl liserated I wk INF 2 wk INF Normal Lacrralrd I wk INF 2 wk INF Mu\cle cuntrul iiijcctioii injectiun Muwle cuiilrul iiijectioii injecrion

Fig. 5. Physiological effects of yINF injection on muscle strength. The fast-twitch (A) and tetanus strength (B) of muscles treated with yINF 1 week post-injury was greater than that of the control muscles, and the difference reached significance with yINF injections at the 2-week time point ( * P < 0.05). The 2-week yINF-injected muscle was virtually identical in strength to the normal non-injured control.

which diminishes within a week after muscle injury [11,18]. In both the 1- and the 2-week yINF injection groups (sacrificed at 4 weeks), there was a significant ( P < 0.01) decrease in the fibrotic area, which was de- termined by the total collagen and vimentin-positive area. Given this decrease in the fibrotic area, muscle regeneration was presumably less restricted. Although the diameters of the muscles treated at either 1 or 2 weeks post-injury were the same as the diameters of the lacerated control muscles, the number of centronucle- ated myofibers in the yINF-injected muscles was sig- nificantly greater (P < 0.05) than the number observed in the lacerated control muscles. Animals in the 2-week post-injury injection group had a greater number of centronucleated cells in the injured area than did ani- mals in the I-week group. We theorize that the lack of physical obstruction by scar tissue was responsible for the increase in muscle regeneration within the injured area.

Additionally, we found that yINF was able to im- prove the function of injured skeletal muscle. At 4 weeks post-laceration, we tested the fast-twitch and tetanic strength of the muscles that had been treated with injections of yINF at either 1 or 2 weeks post- injury. We found that injection of yINF 1 week after laceration resulted in modest increases in the fast- twitch and tetanus strength of these muscles as com- pared to the strength of PBS sham-injected muscles. More importantly, the muscles that were injected with yINF 2 weeks after injury displayed significant (P < 0.05) increases in strength compared to the lacerated control muscles. Muscle recovery was considered al- most complete based on physiological testing, since the strength of the lacerated muscle injected with yINF 2 weeks after injury was indistinguishable from that of the contralateral non-injured control muscles. These findings were supported by our histological results, which revealed that the 2-week injections of yINF re- sulted in increased numbers of centronucleated myofi- bers and an absence of fibrosis. In previous studies we have reported that the optimum injection time point of decorin, another antifibrosis agent, was also 2 weeks

post-injury [6], concurrent with the onset of the fibrosis process [ 1 1,181.

In this mouse laceration model, muscle healing after yINF treatment-measured both histologically and physiologically-was nearly complete. However, clinical injuries tend to be more severe than these lacerations and involve much larger muscles than those in the mouse. Further studies are necessary to investigate the use of gene therapy and cell therapy to deliver antifi- brosis substances and promote their highly persistent, localized expression in larger injured muscles and in other muscle injury models. Hopefully, such studies also will confirm the utility of antifibrosis agents for the treatment of other common muscle injuries.

In this study, yINF improved the healing of injured muscles by blocking the fibrosis process. yINF was proficient at reducing fibroblast cell growth. In addition to the inhibition of myofibroblast phenotype, yINF also was able to decrease the amount of protein produced by TGFPl stimulation in vitro. In vivo, yINF reduced the amount of fibrosis in lacerated skeletal muscle, and thereby fostered enhanced muscle regeneration. The strength of the lacerated skeletal muscle treated with yINF was restored to near-normal levels. Our results indicate that yINF is capable of promoting near-com- plete healing of lacerated skeletal muscle by reducing scar tissue formation, thus enabling enhanced muscle healing.

Acknowledgements

The authors thank Dr. Yi-Sheng Chan, Dr. Kenji Sato, Dr. Takashi Horaguchi, Marcelle Pellerin and Jing Zhou for technical assistance and Thomas Payne, Dr. Victor Prisk, Ryan Sauder, and Jim Cummins for their careful reading and editing of this manuscript. This work was supported by grants to Dr. Johnny Huard from the National Institutes of Health (NIH# 1 R01 AR47973-Ol), the Orris C. Hirtzel and Beatrice Dewey Hirtzel Memorial Foundation, the William Jean Donaldson Endowment at Children’s Hospital of

804 W. Fostrr cc ul. I Journul of Orthopuediic Research 21 (ZOOS) 798 804

Pittsburgh, and the Henry J. Mankin, M.D., Chair of Orthopaedic Surgery Research at the University of Pittsburgh.

References

Baroni G. D'Ambrosio L, Curto P, et al. Interferon gamma decreases hepatic stellate cell activation and extracellular matrix deposition in rat liver fibrosis. Hepatology 1996:23: 1189-99. Border W, Noble N. Transforming growth factor beta in tissue fibrosis. N Engl J Med 1994;331:1286-92. Brocks L, Jap PH, Ramaekers FC, et al. Vimentin and desmin expression in degenerating and regenerating dystrophic murine muscles. Virchows Arch B Cell Pathol Incl Mol Pathol 1991: 61 :89-96. Croisier J. Forthomme B, Namurois M, et al. Hamstring muscle strain recurrence and strength performance disorders. Am J Sports Med 2002;30: 199-203. Friedman S. Molecular regulation of hepatic fibrosis, an inte- grated cellular response to tissue injury. J Biol Chem 2000; 275:2247-50. Fukushima K, Badlani N, Usas A, et al. The use of an antifibrosis agent to improve muscle recovery after laceration. Am J Sports Med 2001:29:394-402. Garrett Jr W, Seaber AV. Boswick J, et al. Recovery of skeletal muscle after laceration and repair. J Hand Surg 1984: 9A:683-92. Garrett W. Muscle strain injuries. Am J Sports Med 1996;24:S2-8. Ghosh A, Yan W, Mori Y, et al. Antagonistic regulation of type 1 collagen gene expression by interferon-y and transforming growth factor-(J. J Biol Chem 2001:276:11041-8. Grounds M. Muscle regeneration: molecular aspects and thera- peutic implications. Curr Opin Neurol 1999;12:53543. Huard J, Li Y, Fu F. Muscle injuries and repair: current trends in research. J Bone Joint Surg Am 2002;84-A:822-32. Hunzelmann N, Anders S, Fierlbeck G, et al. Double-blind, placebo-controlled study of intralesional interferon gamma for the treatment of localized scleroderma. J Am Acad Dermatol 1997: 36:433-5. Jarvinen T, Kaarianen M, Jarvinen M, et al. Muscle strain injuries. Ciirr Opin Rheumatol 2000;12:155-6l.

[I41 Kasemkijwattana C. Menetrey J , Bosch P, et al. Use of growth factors to improve muscle healing after strain injury. Clin Orthop 20003 70:272-8 5,

[I51 Kasemkijwattana C, Menetrey J, Day C, et al. Biologic interven- tion in muscle healing and regeneration. Sports Med Arthrosc Rev 1998;6:95-102.

[I61 Kasemkijwattana C . Menetrey J , Somogyi G, et al. Development of approaches to improve the healing following muscle contusion. Cell Transplant 1998;7:585 -98.

[I71 Kellett J. Acute soft tissue iiijuries--a review of the literature. Med Sci Sports Exerc 1986;18:489 -500.

[I81 Li Y , Cummins J. Huard J . Muscle injury and repair. Curr Opin Orthop 2001;12:409-15.

[I91 Li Y, Han B, Li K, et al. TGF-beta 1 inhibits HLA-DR and beta 2-microglobulin expression in HeLa cells induced with y-IFN. Transplant Proc 1999;31:2143-5.

[20] Li Y, Huard J. Differentiation of muscle-derived cells into myofibroblasts in injured skeletal muscle. Am J Pathol 2002; 16 1 :895-907.

[21] Menetrey J, Kasemkijwattana C, Day C. et al. Growth factors improve muscle healing in vivo. J Bone Joint Surg Br 2000232- B:131-7.

[22] Menetrey J, Kasemkijwattana C, Fu F. et al. Suturing versus immobilization of a muscle laceration. Am J Sports Med 1999; 27:222-9.

[23] Obremsky W. Seaber A, Ribbeck B, Garrett W. Biomechanical and histologic assessment of a controlled muscle strain injury treated with piroxicam. Am J Sports Med 1994;22:558-61.

[24] Oldroyd S, Thomas G. Gabbiani G, et al. Interferon-y inhibits experimental renal fibrosis. Kidney Int 1999;56:2116 27.

[25] Orchard J , Best T. The management of muscle strain injuries: an early return versus the risk of recurrence. Cliii J Sport Med 2002;12:3-5. Phan S, Zhang K, Zhang H, Gharaee-Kermani M. The myofi- broblast as an inflammatory cell in pulmonary fibrosis. Curr Top Pathol 1999;93:173-82. Qu Z . Balkir L, Van Deutekom J. et al. Development of approaches to improve cell survival in myoblast transfer therapy. J Cell Biol 1998; 142:1257-67. Sappino A, Schurch W, Gabbiani G. Differentiation repertoire of fibroblastic cells: expression of cytoskeletal proteins as marker of phenotypic modulations. Lab Invest 1990:63:144-61. Serini G, Gabbiani G. Mechanisms of myofibroblast activity and phenotypic modulation. Exp Cell Res 1999;250:273-83. Striz I, Mio T, Adachi Y, et al. Effects of interferons alpha and gamma on cytokine production and phenotypic pattern of human bronchial epithelial cells. Int J Immunopharmacol 2000;22:573- 85. Thorsson 0, Rantanen J, Hurme T, Kalimo H. Effects of nonsteroidal anti-inflammatory medication on satellite cell prolif- eration during muscle regeneration. Am J Sports Med 1998; 26: 172-6. Ulloa L, Doody J, Massague J . Inhibition of transforming growth factor-j3/SMAD signalling by the interferon-ylSTAT pathway. Nature 1999;397:7 lQ3. Vaughan M, Howard E, Tomasek J. Transforming growth factor- (Jl promotes the morphological and functional differentiation of the myofibroblast. Exp Cell Res 2000;257:180 9. Verall G, Slavotinek J, Barnes P, et al. Clinical risk factors for hamstring muscle strain injury: a prospective study with correla- tion of injury by magnetic resonance imaging. Br J Sports Med 200 1 ;35:435-9. Yokozeki M, Baba Y, Shimokawa H, et al. Interferon-y inhibits the myofibroblastic phenotype of rat palatal fibroblasts induced by transforming growth factor-(Jl in vitro. FEBS Lett 1999; 442:6 1 ~ 4. Zhang Li, Mi J, Yu Yizhi, et al. IFN-y gene therapy by intrasplenic hepatocyte transplantation: a novel strategy for reversing hepatic fibrosis in schistosoma japonicum-infected mice. Parasite lmmunol 2001;23: 11-7.