JOURNAL OF BONE AND MINERAL RESEARCH Volume 7, Number 1,1992 Mary Ann Liebert, Inc., Publishers

Transforming Growth Factor b1 Stimulates Type I1 Collagen Expression in Cultured Periosteum-Derived Cells TOSHIHIRO IZUMI,'.' SEAN P. SCULLY,',3 AHLKE HEYDEMANN,1,4 and MARK E. BOLANDERL,*

ABSTRACT Chondrogenesis can occur during a bone repair process, which is related to several growth factors. Transforming growth factor 0,(TGF-0,) downregulates the expression of type I1 collagen by chondrocytes in vitro, but injection of TGF-0, into the periosteum in vivo increases type I1 collagen mRNA levels and initiates chondrogenesis.'l) We examined the effect of TGF-0, on collagen gene expression in a bovine periosteum-derived cell culture system to evaluate its direct effect on the periosteum. Cultured cells expressed alkaline phosphatase and collagen proal(I) and procrl(II) mRNAs. A low level of type I1 collagen synthesis was demonstrated by immunoprecipitation. TGF-0, had no effect on periosteal cell proliferation. Expression of collagen proal(I) mRNA did not change with TGF-0, treatment, but alkaline phosphatase mRNA showed a dose-dependent decrease. Expression of collagen proa,(II) mRNA was stimulated 2.7-fold by TGF-PI. TGFp1also caused a 2.6-fold increase in type I1 collagen synthesis by immunoprecipitation. These findings indicate that TGF-0, is an enhancer of the expression of the chondrocyte phenotype of the periosteal cells and suggest that TGF-0, is important in initiating and promoting cartilage formation in vivo.

INTRODUCTION that support osteoblast differentiation and bone formation during development, growth, and repair.(zJ) In addition to bone, however, cartilage is seen in the reparative tissue that forms adjacent to the periosteum during fracture healing,c4)and free periosteal transplants form new cartilage in various noncartilagenous tissues in vivo. ( 5 . 6 . ) In situ hybridization of the early fracture callus suggests that the periosteum contains cells expressing type I and type I1 collagen genes. (') These findings suggest that the periosteum may also contain chondrocyte progenitor cells. ( 3 . n ) However, factors that stimulate chondrogenesis in the periosteum have not been identified. Many growth factors are involved in bone and cartilage m e t a b o l i ~ m ( ~and - ~ ~in ) the fracture-healing process. (13.14)

P

ERIOSTEUM CONTAINS PROGENITOR CELLS

Transforming growth factor PI (TGF-p,) is a multifunctional growth factor thought to be important in bone formation during remodeling(1S.16) and during embryonic development. (I7) TGF-b increases type I collagen in osteob l a s t ~ ( ' ~ . 'and ~ ) type I1 collagen in embryonic muscle cells(2o) but decreases type I1 collagen in chondrocytes. (l*.zl) Analogous to its multifunctional role in soft tissue wound healing,(z2-z4~ studies of TGF-PI during fracture healing suggest complex functions in regulating chondrogenesis, intramembranous ossification, and endochondral ossification.(25) Recent in vivo studies suggest a potential role of TGF-PI in inducing type I1 collagen gene expression and in initiating the differentiation of chondrocytes from periosteal precursors during bone formation and fracture healing. ('1 The present studies were undertaken to examine the effect of TGF-0, on periosteal cells in vitro and to determine if

'Orthopaedic Research Unit, National Institute of Arthritis, Musculoskeletal, and Skin Diseases, National Institutes of Health, Bethesda, Maryland. 'Present address: Department of Orthopedic Surgery, Mayo Clinic, Rochester, Minnesota. 3Pre~entaddress: Department of Orthopedics, Duke University, Durham, North Carolina. 'Present address: Laboratory of Chemoprevention, National Cancer Institute, National Institutes of Health, Bethesda, Maryland.

115

116

IZUMI ET AL.

RNA was determined by spectrophotometric absorption at 260 nm. For normalization, equal amounts of total RNA were electrophoresed under denaturing conditions and MATERIALS AND METHODS stained with ethidium bromide to confirm the integrity and loading amount of RNA in each lane. RNA was transPeriosteum was stripped from the diaphysis of metacar- ferred to nylon membranes (Hybond-N; Amersham, pal bones of 4- to 6-month-old calves. Areas of dissection Arlington Heights, IL) using standard methods.(28)After were distant from cartilage tissue and periochondrium in prehybridization, the membranes were hybridized with rathe region of the epiphysis and articular surfaces. Perios- diolabeled cDNA probes. ( 2 9 ) The hybridization signals teum was digested with 1.5 mg/ml of collagenase (CLS2, were visualized using x-ray films (X-Omat AR; Kodak, Worthington, Freehold, NJ) in modified BGJb media (Fit- Rochester, NY). Relative levels of mRNA expression were ton-Jackson modification; GIBCO, Grand Island, NY), determined by counting the radioactivity of hybridized supplemented with 1 mM sodium pyruvate (GIBCO), 2% probes with a beta scanner (Betagen, Waltham, MA). For immunoprecipitation for type I1 collagen, the periantibiotic-antimycotic (containing penicillin, streptomycin, and amphotericin B; GIBCO), 0.02% bovine serum albu- osteal cells were treated with TGF-PI for 24 h. Cells were min, and 25 mM HEPES (pH 7.4; GIBCO) in humidified then incubated for an additional 24 h in media containing 5% CO, and 95% air at 37°C overnight. The isolated cells TGF-P, supplemented with [14C]proline (10 pCi/ml, 310 were collected by centrifugation at 500 x g for 10 minutes, mCi/mmol; New England Nuclear) and 50 pg/ml of Pwashed, resuspended in modified BGJb containing 10% aminopropionitrile. Following incubation, both media and fetal bovine serum (FBS; GIBCO), and seeded on tissue cells were collected, mixed with lysis buffer (10 mM Tris, culture plates (Falcon 3046; Becton Dickinson Labware, pH 7.2; 0.15 M NaCI; 1% Triton X-100; 0.1% sodium doLincoln Park, NJ) at a density of lo5 cells per cm2. After decyl sulfate; 1Vo aprotinin; and 1070 deoxycholate), soni24 h of incubation, attached cells were washed by changing cated, and centrifuged. The total protein concentrations the media. Cells were cultured for 4 days with 10% FBS were determined (Bio-Rad protein assay; Bio-Rad, Richand then 2 days in serum-free media. At that time, cells mond, CA). Samples containing equivalent protein conwere almost confluent. Media were then supplemented centrations were incubated with anti-type I1 collagen antiwith 10% FBS or TGF-0, (0.05-10 ng/ml) for 24 h. Cells body for 16 h at 4"C, followed by incubation with protein were harvested for RNA extraction or metabolically la- A-Sepharose in PTA (phosphate-buffered saline, pH 7.2, 0.5% Tween 20, 0.05% sodium dodecyl sulfate, 0.1% bobeled for immunoprecipitation as described later. For histologic examination, isolated periosteal cells were vine serum albumin, and 0.02% sodium azide). The samresuspended in modified BGJb medium containing 10% ples were centrifuged, washed with lysis buffer, and exFBS and grown in tissue culture chambers for 4 days (cham- tracted with sample buffer (62.5 mM Tris, pH 6.8, 10% ber slide #177445, Lab-Tek; Naperville, IL). Cells on the glycerol, 2% sodium dodecyl sulfate, and 2.5% 2-mercapchamber slides were stained with alcian blue (1010, pH 2.5). toethanol) at 100°C for 3 minutes. Immunoprecipitates DNA synthesis was determined by [3H]thymidine incor- were electrophoresed on sodium dodecyl sulfate-polyacrylporation. Cells were treated for 24 h with TGF-PI or FBS, amide gel electrophoresis (SDS-PAGE) gels (8% precast and methyl-[3H]thymidine (5 pCi/ml, 83.7 Ci/mmol; New gels; Novex, Encinitas, CA). Gels were incubated with EnEngland Nuclear, Boston, MA) was added to each culture lightning (New England Nuclear), dried, and exposed to for the last 4 h. The cells were washed twice, scraped from x-ray films for 4 weeks. Relative levels were obtained by the culture wells, and digested with 0.05% proteinase K scanning fluorograms with a densitometer (UltroScan XL; (Boehringer-Mannheim, Indianapolis, IN). DNA was pre- LKB, Bromma, Sweden). TGF-0, was a gift from Drs. A. Roberts and M. Sporn, cipitated with 10% trichloroacetic acid using bovine serum albumin as a carrier. Samples were dissolved in 0.25 M National Cancer Institute, NIH. cDNA probes for bovine NaOH and counted on a scintillation counter (Packard crl(II)-procollagen, rat aI(I)-procollagen, and alkaline 2500TR; Downers Grove, IL). Aliquots of proteinase K-di- phosphatase were gifts from Dr. M. Sobel (National Cangested samples were used to determine DNA content by cer Institute, NIH), Dr. D. Rowe (University of Connectithe method of Labarca and Paigen(16)using bis-benzimide cut, Farmington), and Drs. G. Rodan and M. Thiede (Hoechst dye 33258). Incorporated radioactivity was nor- (Merck, Sharp and Dohme, West Point, PA), respectively. A monoclonal antibody to type I1 collagen was a gift from malized to total DNA. Total cellular RNA was extracted by acid guanidine Dr. T. Linsenmayer (Massachusetts General Hospital, thiocyanate-phenol-chloroform extraction.'") Briefly, Boston, MA). Except as noted, all chemicals were obwashed cells were incubated in 1 vol denaturing solution (4 tained from Sigma (St. Louis, MO). Data are expressed as mean and standard error of the M guanidine thiocyanate and 25 mM sodium citrate, pH 7.0; 0.5% sarkosyl, and 0.1 M 2-mercaptoethanol), 0.1 vol mean (SEM). Data for various doses of TGF-/3, were anaof 2 M sodium acetate, pH 4.0, 1 vol phenol, and 0.2 vol lyzed by analysis of variance (ANOVA) (Statview 11; chloroform. Then RNA was precipitated with isopropanol, Abacus, Berkeley, CA) and are reported with the p value. washed with ethanol, and resuspended in diethylpyrocar- Data for FBS were compared to controls by Student's fbonate (DEPC)-treated water. The concentration of total test and are reported with the p value. TGF-P, has the ability to increase type I1 collagen expression in cells derived from adult tissue.

117

TGF-P, STIMULATES TYPE I1 COLLAGEN EXPRESSION

RESULTS Morphology of cultured periosteal cells Histologic evaluation of the periosteal tissue used showed fibrous and cambial layers of the periosteum, only occasional bone, and no cartilage tissue (data not shown). After 6 days in culture, we were able to identify two distinct types of cells in confluent cultures of bovine periosteal cells. Cells attached to the surface of the plate had a pleomorphic, fibroblast-like shape; other cells were round or polygonal and formed densely packed nodules (Fig. 1). Some nodules stained with alcian blue, indicating the presence of proteolglycans, whereas the cells on the surface of the plate apart from the nodules did not show alcian blue staining.

DNA synthesis Treatment with TGF-@,for 24 h did not affect the proliferation of the cultured periosteal cells as evaluated by thymidine incorporation (Fig. 2). Cultured periosteal cells were able to proliferate, however, as thymidine incorporation was stimulated fivefold by 10% FBS.

FIG. 1. Appearance of cultured periosteal cells. Isolated periosteal cells were cultured in chamber slides and stained with alcian blue dye. Before treatment with TGF-PI, two types of cells were seen, round cells and fibroblastic cells. Round cells formed nodules. Some nodules showed alcian blue staining. Magnification, x 130.

Collagen gene expression After treatment of cultured periosteal cells with either various concentrations of TGF-0, or 10% FBS, cells were harvested for RNA extraction and northern analysis. Type I1 collagen gene expression was determined by counting the radioactivity of a bovine al(II)-procollagen probe that hybridized to a 5.5 kb message. a,(II)-procollagen mRNA was expressed at low levels in periosteum-derived cells before treatment with TGF-@,or FBS. This finding is consistent with investigations reporting type I1 collagen mRNA in periosteum by in situ hybridization.‘”) After treatment with TGF-P, (0.1 ng/ml), proa,(II) gene expression increased 2.7-fold (Fig. 3A). This increase was reduced to 1.7-fold as the concentration of TGF-/3, was increased to 10 ng/ml. A 2-fold increase in type I1 collagen gene expression resulted from treatment with 10% FBS. Type I collagen gene expression was determined by counting the radioactivity of a proa,(I) cDNA probe that hybridized to a 4.7 kb message on northern blots. Although a,(I)-procollagen mRNA levels showed slight differences among the various doses of TGF-0, and 10% FBS (Fig. 3B), there were no statistically significant dose-dependent changes in type I collagen gene expression after treatment with TGF-&.

Immunoprecipitation of type 11 collagen Immunoprecipitation was undertaken to determine if the increase in type I1 collagen mRNA levels were accompanied by an increase in synthesis of type I1 collagen. Cultures treated with TGF-0, for 48 h were labeled with [“C]proline for the last 24 h. Media and cell lysates were combined and type I1 collagen was precipitated with monoclonal antibodies against type I1 collagen. Treatment of

800

-

600 -

400 -

200

0

0.1

TGF-I3

1.0

10

FBS

(ng/ml)

FIG. 2. Effect of TGF-/3, on thymidine incorporation. Periosteal cells were treated with TGF-0, or 10% FBS for 24 h. Cells were labeled with [’Hlthymidine for the last 4 h of the incubation period. TGF-0, had no significant effect on thymidine incororation by periosteal cells (ANOVA p = 0.11); FBS stimulated a 5-fold increase (p = 0.OOOl). Data are mean k SEM for three observations.

IZUMI ET AL.

118

B

A

li

Q) -5.5 kb I

0 0.05 0.1 1.0 10

3.0

T

c

FBS

-5.7 kb -4.7 kb

T

2.0 -

.-

0

B5 2.0

c

T

.-0 3

w

c?P 3

a3

.-c>

m 2

0 0.05 0.1 1.0 10

FBS

T

1.0

.-%

1.0

-ma3 4-

a

0.0

0.0

0 0.05 0.1 1.0 10

FBS

0 0.05 0.1 1.0 10

TGF-P (ng/ml)

FBS

TGF-fi (ng/ml)

C

0 0.05 0.1 1.0 10

FBS

TGF-P (ng/ml) FIG. 3. Gene expression by TGF-0,. Periosteal cells were treated with TGF-0, or 10% FBS for 24 h. Total RNA was extracted and analyzed by northern blots. Expression levels of collagen genes were quantitated with a beta scanner. (A) Expression of type I1 collagen mRNA. (Top) The northern blots in one series of observations. (Bottom) mRNA expression relative to nonstimulated cultures (0 ng/ml of TGF-&). Relative expression was determined by counting the radioactivity of a or,(II)-procollagen probe that hybridized to a 5.5 kb band. Treatment with TGF-@, stimulated maximal type I1 collagen gene expression at lower doses (ANOVA, p = 0.OOOl). FBS induced a twofold increase (p = 0.08). Data SEM; three to five observations were made at each TGF-0, dose in two independent experiments. (B) Exare mean pression of type I collagen mRNA. (Top) The northern blots in one series of experiments. (Bottom) mRNA expression relative to nonstimulated cultures (0 ng/ml of TGF-P,). Relative expression was determined by counting the radioactivity of a proor,(I) cDNA probe that hybridized to a 4.7 kb band. Treatment with TGF-0, (ANOVA, p = 0.45) or FBS (p = 0.64) had no significant effect on type I collagen gene expression. Data are mean f SEM for three observations. (C) Northern blot of alkaline phosphatase mRNA. Alkaline phosphatase mRNA was detected by hybridization to the cDNA probe for alkaline phosphatase. Northern analysis suggests a decrease in alkaline phosphatase gene expression with increasing TGF-P, concentrations.

*

periosteal cells with TGF-P, resulted in a 2.6-fold increase in synthesis of type I1 collagen during the labeling period (Fig. 4).

Expression of alkaline phosphatase mRNA As periosteum is thought to be a source of osteoblasts during bone growth,(3)the effect of TGF-0, on expression of a gene related to osteoblast function was also evaluated. Alkaline phosphatase expression is associated with osteo-

blast differentiati~n'~~' and bone formation. Northern blots for alkaline phosphatase mRNA suggested decreased expression with increasing concentrations of TGF-0, between 0 and 1.0 ng/ml (Fig. 3C). The effect of TGF-0, on alkaline phosphatase gene expression appeared opposite to the effect on type I1 collagen gene expression.

DISCUSSION The periosteum contains several cell types, including the fibroblasts in the outer layer and precursors in the inner or

TGF-0, STIMULATES TYPE I1 COLLAGEN EXPRESSION

3.0

T

-a1

(11)

2.0

.-> Q)

0 0.1

Y

1.o

0.0

0 0.1

FIG. 4. Effects of TGF-0, on type I1 collagen synthesis. Periosteal cells were exposed to TGF-0, for 48 h and labeled with [“Clproline for the last 24 h. Type I1 collagen was immunoprecipitated from combined media and cell lysates and evaluated by SDS-PAGE. Fluorograms were analyzed by scanning densitometry. Data are mean SEM for three observations. Treatment with 0.1 ng/ml of TGF-0, stimulated a 2.6-fold increase in type I1 collagen synthesis (p = 0.02).

cambial layer that support osteoblast differentiation during bone formation and growth. We were able to identify two distinct types of cells in confluent cultures of bovine periosteal cells: fibroblast-like cells and round cells. These latter cells appeared similar to the cells described in previous reports of cultured periosteal and of cultured embryonic limb bud cells, which also form nodules when cultured to confluence.(34) Despite the reported stimulatory effect of TGF-0 on proliferation and type I collagen synthesis by osteoblasts, ( 1 8 . 1 9 ) TGF-0, treatment did not increase mRNA levels for type I collagen or stimulate proliferation in periosteal cultures, and expression of the alkaline phosphatase gene was decreased. Instead, TGF-0, stimulated type I1 collagen gene expression and synthesis. Increased type I1 collagen mRNA levels were not caused by proliferation of cells expressing type I1 collagen, as TGF-0, did not stimulate thymidine incorporation. TGF-8, stimulates the proliferation of many skeletal cell types, including fetal rat calvarial cells,(3s) osteoblasts,(36) and chondrocytes. (21.37) Depending on the cell type and culture conditions, however, TGF-0 can also inhibit proliferation, that is, osteoblasts from a rat osteo~arcoma.(~*) In other systems, TGF-

119

0, does not have a significant effect on proliferation, that is, fetal calf calvarial cells(39)and rabbit growth plate chondrocytes. 1.O) In this study, there were no significant differences in type I collagen mRNA levels resulting from treatment with TGF-0,. These results contrast with reports that TGF-0, stimulates type I collagen gene expression in most cell types, including epithelial cells,‘41’f i b r o b l a ~ t s , ( ~bone~-~~) derived cells,(19) and osteosarcoma cells.(38)In fetal rat parietal bone cells, however, mRNA levels of type I collagen were decreased at concentrations between 0.15 and 1.5 ng/ml but increased at a concentration of 15 ng/ml.(4s’ The mechanism by which TGF-0, regulates type I and type I1 collagen gene expression in periosteal cells is not known. Different patterns of collagen expression seen after TGF/3 stimulation suggest that regulatory mechanisms in periosteal cells differ from those in other cell types. Cells digested from the periosteum could not be established in culture without serum (data not shown). It is possible that culturing the cells in the presence of serum factors before TGF-0 stimulation could have some effects on periosteal cell differentiation, and indeed, periosteal cells in culture for 5 days have a similar morphology to cultured chondrocytes. Based on their response to TGF-PI, however, periosteal cells are unlike cultured chondrocytes, which decrease type I1 collagen mRNA and protein in response to TGF-/31.(18.21.46) Cultured periosteal cells appeared to respond to TGF-0, similarly to cells from embryonic muscle or from embryonic limb bud, which show increased type I1 collagen expression after stimulation by TGF-/3.(20.47) Previous data suggest that chondrocytes can be distinguished from their precursors by assessing the change in type I1 collagen expression after TGF-0 stimulation.(18.20.21.46.471 The current study suggests that a population of cells in the periosteum of adult animals retain the response to TGF-0, seen in chondrocyte precursors from embryonic limb bud or muscle. Additionally, these data suggest that an increase in type I1 collagen expression after stimulation by TGF-0, predicts, as seen in subperiosteal injection of TGF-0 in viva,(') the future appearance of chondrocytes. It is notable that alkaline phosphatase expression was decreased by TGF-0,. Alkaline phosphatase expression in the periosteum is probably not from fibroblasts or chondrocyte precursors but from osteoblasts and osteoblast precursors.(48) Yamaguchi et showed that TGF-0, decreased alkaline phosphatase activity in immature osteoblast precursor cells. We speculate that osteoblast precursors known to be present in the periosteum are also present in periosteal cell cultures. These cells would then be responsible for the expression of alkaline phosphatase mRNA seen in culture, and we would expect that alkaline phosphatase expression by these cells would be downregulated by TGF-0,. Together with increased type I1 collagen synthesis, suppression of alkaline phosphtase gene expression suggests that TGF-0, stimulated confluent periosteal cells to express the chondrocyte phenotype rather than the osteoblast phenotype. These observations are consistent with the enhancement of type I1 collagen gene expression seen in vivo following subperiosteal TGF-0, injections(’)and indicate that TGF-

120

IZUMI ET AL.

p, is an important regulator of osteoblast and chondrocyte phenotypes during bone formation and repair. (15,16) TGFp1 may initiate chondrogenesis during fracture repair by stimulating embryonic-like chondrocyte precursors present in the periosteum to increase type I1 collagen gene expression. ACKNOWLEDGMENTS We thank Drs. V. Hascall, H. Kleinman, and A. Roberts for many helpful discussions and suggestions and N. Marinos for expert technical assistance. Part of this work was presented at the 37th Annual Meeting of the Orthopaedic Research Society. This work was supported in part by grants to M. Bolander from the Orthopaedic Research and Education Foundation and from the American Academy of Orthopaedic Surgeons.

REFERENCES 1 . Joyce ME, Roberts AB, Sporn MB, Bolander ME 1990 Transforming growth factor-@and the initiation of chondrogenesis and osteogenesis in the rat femur. J Cell Biol 110: 2 195-2207. 2. Koshihara Y, Kawamura M, Endo S, Tsutsumi C, Kodama H, Oda H, Higaki S 1989 Establishment of human osteoblastic cells derived from periosteum in culture. In Vitro Cell Dev Biol 25:37-43. 3. Wlodarski KH 1989 Normal and heterotopic periosteum. Clin Orthop 241:265-277. 4. McKibbin B 1978 The biology of fracture healing in long bones. J Bone Joint Surg 60-8:150-162. 5 . Poussa M, Ritsila V 1979 The osteogenic capacity of free periosteal and osteoperiosteal grafts. Acta Orthop Scand 50: 49 1-499. 6. Poussa M, Rubak J, Ritsila V 1981 Differentiation of the osteochondrogenic cells of the periosteum in chondrotrophic environment. Acta Orthop Scand 52:235-239. 7. Sandberg M, Aro H, Multimaki P, Aho H, Vuorio E 1989 In situ localization of collagen production by chondrocytes and osteoblasts in fracture callus. J Bone Joint Surg 71-A:69-77. 8. Nakahara H, Bruder SP, Haynesworth SE, Holecek JJ, Baber MA, Goldberg VM, Caplan A1 1990 Bone and cartilage formation in diffusion chambers by subcultured cells derived from the periosteum. Bone 11:181-188. 9. Hauschka PV, Mavrakos AE, lafrati MD, Doleman SE, Klagsbrun M 1986 Growth factors in bone matrix. J Biol Chem 261:12665-12674. 10. Canalis E, McCarthy T, Centrella M 1988 Growth factors and the regulation of bone remodeling. J Clin Invest 81:277281. 1 1 . Wozney JM, Rosen V, Celeste AJ, Mitsock LM, Whitters MJ, Kriz RW, Hewick RM, Wang EA 1988 Novel regulators of bone formation: Molecular clones and activities. Science 242:1528-1534. 12. Vukicevic S, Luyten FP, Reddi AH 1989 Stimulation of the expression of osteogenic and chondrogenic phenotypes in vitro by osteogenin. Proc Natl Acad Sci USA 86:8793-8797. 13. Nemeth GG, Bolander ME, Martin GR 1988 Growth factors and their role in wound and fracture healing. Prog Clin Biol Res 266:l-17.

14. Hulth A 1989 Current concepts of fracture healing. Clin Orthop 249:265-284. 15. Carrington JL, Roberts AB, Flanders KC, Roche NS, Reddi AH 1988 Accumulation, localization, and compartmentation of transforming growth factor fl during endochondral bone development. J Cell Biol 107:1969-1975. 16. Centrella M, McCarthy TL, Canalis E 1988 Skeletal tissue and transforming growth factor @.FASEB J 2:3066-3073. 17. Sandberg M, Vuorio T, Hirvonen H, Alitalo K, Vuorio E 1988 Enhanced expression of TGF-fl and c-fos mRNA in the growth plates of developing human long bones. Development 102:461-470. 18. Rosen DM, Stempien SA, Thompson AY, Seyedin SM 1988 Transforming growth factor-@ modulates the expression of osteoblast and chondroblast phenotypes in vitro. J Cell Physiol 134:337-346. 19. Wrana JL, Maeno M, Hawrylyshyn B, Yao K-L, Domenicucci C, Sodek J 1988 Differential effects of transforming growth factor-@on the synthesis of extracellular matrix proteins by normal fetal rat calvarial bone cell populations. J Cell Biol 106:915-924. 20. Seyedin SM, Thomas TC, Thompson AY, Rosen DM, Piez KA 1985 Purification and characterization of two cartilageinducing factors from bovine demineralized bone. Proc Natl Acad Sci USA 82:2267-2271. 21. Horton Jr WE, Higginbotham JD, Chandrasekhai S 1989 Transforming growth factor-@and fibroblast growth factor act synergistically to inhibit collagen I1 synthesis through a mechanism involving regulatory DNA sequences. J Cell Physiol 14i:8-15. 22. Sporn MB, Roberts AB, Wakefield LM, de Crombrugghe B 1987 Some recent advances in the chemistry and biology of transforming growth factor-@. J Cell Biol 105:1039-1045. 23. Rodland KD, Muldoon LL, Magun BE 1990 Cellular mechanisms of TGF-@action. J Invest Dermatol 94:33S-40S. 24. Sporn MB, Roberts AB 1990 The transforming growth factor-betas: Past, present, and future. Ann NY Acad Sci 593:1-6. 25. Joyce ME, Terek RM, Jingushi S, Bolander ME 1990 Role of transforming growth factor-@in fracture repair. Ann NY Acad Sci 593:107- 123. 26. Labarca C, Paigen K 1980 A simple, rapid, and sensitive DNA assay procedure. Anal Biochem 102:344-352. 27. Chomczynski P, Sacchi N 1987 Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal Biochem 162:156-159. 28. Sambrook J, Fritsch EF, Maniatis T 1989 Transfer of denatured RNA to nitrocellulose filters. In: Molecular cloning. A laboratory manual, 2nd ed., Vol. 1 . Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, pp, 7.467.50. 29. Jingushi S, Heydemann A, Kana SK, Macey LR, Bolander ME 1990 Acidic fibroblast growth factor injection stimulates cartilage enlargement and inhibits cartilage gene expression in rat fracture healing. J Orthop Res 8:364-371. 30. Sandberg M, Vuorio E 1987 Localization of type I, 11, and I l l collagen mRNAs in developing human skeletal tissues by in situ hybridization. J Cell Biol 104:1077-1084. 31. Burks JK, Peck WA 1978 Bone cells: A serum-free medium supports proliferation in primary culture. Science 199542544. 32. Uchida A, Kikuchi T, Shimomura Y 1988 Osteogenic capacity of cultured human periosteal cells. Acta Orthop Scand 59: 29-33. 33. Bellows CG, Heersche JNM, Aubin JE 1989 Effects of dexa-

TGF-PI STIMULATES TYPE I1 COLLAGEN EXPRESSION methasone on expression and maintenance of cartilage in serum-containing cultures of calvaria cells. Cell Tissue Res 256: 145-15 1. 34. von der Mark K, Conrad G 1979 Cartilage cell differentiation. Clin Orthop 139:185-205. 35. Canalis E, Lian JB 1988 Effects of bone associated growth

36.

37.

38.

39.

40.

41.

42.

43.

factors on DNA, collagen and osteocalcin synthesis in cultured fetal rat calvariae. Bone 9:243-246. Robey PG, Young MF, Flanders KC, Roche NS, Kondaiah P, Reddi AH, Termine JD, Sporn MB, Roberts AB 1987 Osteoblasts synthesize and respond to transforming growth factor-b (TGF-b) in vitro. J Cell Biol 105:457-463. O'Keefe RJ, Puzas JE, Brand JS, Rosier RN 1988 Effect of transforming growth factor-p on DNA synthesis by growth plate chondrocytes: Modulation by factors present in serum. Calcif Tissue Int 43:352-358. Noda M, Rodan GA 1987 Type 0 growth factor (TGFO) regulation of alkaline phosphatase expression and other phenotype-related mRNAs in osteoblastic rat osteosarcoma cells. J Cell Physiol 133:426-437. Globus RK, Patterson-Buckendahl P, Gospodarowicz D 1988 Regulation of bone cell proliferation by fibroblast growth factor and transforming growth factor @. Endocrinology 123:98-105. Kato Y, Iwamoto M, Koike T, Suzuki F, Takano Y 1988 Terminal differentiation and calcification in rabbit chondrocyte cultures grown in centrifuge tubes: Regulation by transforming growth factor 0 and serum factors. Proc Natl Acad Sci USA 85:9552-9556. Silberstein GB, Strickland P, Coleman S, Daniel CW 1990 Epithelium-dependent extracellular matrix synthesis in transforming growth factor-&growth-inhibited mouse mammary gland. J Cell Biol 110:2209-2219. Fine A, Poliks CF, Smith BD, Goldstein RH 1990 The accumulation of type 1 collagen mRNAs in human embryonic lung fibroblasts stimulated by transforming growth factor-p. Connect Tissue Res 24:237-247. Narayanan AS, Page RC, Swanson J 1989 Collagen synthesis

44.

45.

46.

47.

48.

49.

121 by human fibroblasts. Regulation by transforming growth factor-p in the presence of other inflammatory mediators. Biochem J 260:463-469. Quaglino D Jr, Nanney LB, Kennedy R, Davidson JM 1990 Transforming growth factor-8 stimulates wound healing and modulates extracellular matrix gene expression in pig skin. I . Excisional wound model. Lab Invest 63:307-3 19. Centrella M, McCarthy TL, Canalis E 1987 Transforming growth factor B is a bifunctional regulator of replication and collagen synthesis in osteoblast-enriched cell cultures from fetal rat bone. J Biol Chem 262:2869-2874. O'Keefe RJ, Puzas JE, Brand JS, Rosier RN 1988 Effects of transforming growth factor-0 on matrix synthesis by chick growth plate chondrocytes. Endocrinology 122:2953-2961. Kulyk WM, Rodgers BJ, Greer K, Kosher RA 1989 Promotion of embryonic chick limb cartilage differentiation by transforming growth factor-p. Dev Biol 135:424-430. Deren JA, Kaplan FS, Brighton CT 1988 Alkaline phosphatase production by periosteal cells at various oxygen tensions in vitro. Clin Orthop 252:307-312. Yamaguchi A, Katagiri T, Ikeda T, Wozney JM, Rosen V , Wang EA, Kahn AJ, Suda T, Yoshiki S 1991 Recombinant human bone morphogenetic protein-2 stimulates osteoblastic maturation and inhibits myogenic differentiation in vitro. J Cell Biol 113:681-687.

Address reprint requests to: Toshihiro Izumi, M.D. Department of Orthopedic Surgery Mayo Clinic Med. Sci. 3-69 200 First Street, S W Rochester, MN 55905 Received for publication May 7, 1991; in revised form July 1 , 1991; accepted July 26, 1991.