Transforming growth factor- β 1: Induction of bone morphogenetic protein genes expression during...

Gmwih F r r < i o , T . Vol. IS, pp. 259-277 Rcprint? available directly froin the puhlirher Photocopying permitted by license only

@ 1998 OPA (Overseas Pubhhers Acwciation) N.V. Publiahed by license under

the Harwoud Academic Publishers imprint, pan uf The Gordon and Breach Publishing Group.

Printed in Malaysia.

Transforming Growth Factor-/?1: Induction of Bone Morphogenetic Protein Genes Expression During

Endochondral Bone Formation in the Baboon, and Synergistic Interaction with Osteogenic Protein-1

(BMP-7) NICOLAAS DUNEAS, JEAN CROOKS and UGO RIPAMONTI*

Bone Research Laboratory, Medical Research Council/UniversiQ ofthe Witwatersrand, Medical School, Parktown 2193, Johannesburg, South Africa

(Received 4 September 1997; In final form 10 November 1997)

Certain members of the bone morphogenetic protein (BMP) and transforming growth factor-b (TGF-R) families are inducers of endochondral bone formation in vivo. TGF-Rs, however, do not initiate bone formation when implanted in heterotopic (extraskeletal) sites of rodents. Here we show that platelet-derived porcine TGF-I31 (pTGF-D1) induces endochondral bone in heterotopic sites of the baboon (Pupio ursinus) at doses of 5 pg per 100 mg of guanidinium-inactivated collagenous bone matrix as carrier, with an inductive efficiency comparable to 5 and 25 pg of recombinant osteogenic protein-1 (hop-1, BMP- 7), a well characterized inducer of bone fomiation. We further demonstrate that pTGF-BI and hOP-1 interact synergistically to induce large ossicles in the rectus abdominis of the primate as evaluated by key parameters of bone formation on day 14 and 30. Tissue generated on day 30 by 5 pg pTGF-Dl or 25 pg hOP-1 induced comparable expression levels of OP-1, BMP-3 and type IV collagen mRNA transcripts, whereas TGF-Bl and type I1 collagen expression was 2 to 3 fold higher in pTGF-Bl-treated implants, as determined by Northern analysis. In ossicles generated by 25 pg hOP-1 in combination with relatively low doses of pTGF-I31 (0.5, 1.5 and 5 pg), type I1 collagen expression increased in a pTGF-I31 dose-dependent manner, whilst type IV collagen was synergistically upregulated with a 3 to 4 fold increase compared to ossicles generated by a single application of 5 pg pTGF-81 or 25 pg hOP-l. Morphogen combinations (5 pg pTGF-bl with 20 pg hOP-I, and 5 and 15 pg pTGF-j3l with 100 pg hOP-I per g of collagenous matrix as carrier) induced exuberant tissue formation and greater amounts of osteoid than hOP-1 alone when implanted in calvarial defects of the baboon as evaluated

*Corresponding author: Bone Research Laboratory, MRCRJniversity of the Witwatersrand, Medical School, 7 York Road, Parktown 2193, Johannesburg, South Africa. Tel. and Fax: 27 11 647-2300. E-mail: [email protected]

259

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

260 N. DUNEAS ef al.

on day 30 and 90, with displacement of the temporalis muscle above the defects. Since a single application of TGF-Bl in the primate did not induce bone formation in calvarial defects, whilst it induces endochondral bone differentiation in heterotopic sites, our data indicate that the bone inductive activity of TGF-Bl is site and tissue specific. mRNA expression of multiple members of the TGF-B superfamily suggests complex autocrine and paracrine activities of the ligands and different signalling pathways on responding cells during the cascade of endochondral bone formation in the primate. The present findings may provide the basis for synergistic molecular therapeutics for cartilage and bone regeneration in clinical contexts.

Keywords: Bone morphogenetic proteins, molecular therapeutics, morphogenesis, primates, synergy, tissue regeneration

INTRODUCTION

The transforming growth factor-Bs (TGF-Bs) and the bone morphogenetic proteins (BMPs) are pleiotropic members of the TGF-B superfamily, soluble medi- ators of tissue morphogenesis, and powerful regulators of cartilage and bone differentiation during embryonic bone development and regeneration of cartilage and bone in postnatal life (Wozney, 1992; Reddi, 1994; Centrella et al., 1994; Reddi, 1997). The morphogenetic cascade seen in embryonic bone development can be recapitulated by implanting demineralized bone matrix in rodent heterotopic (extraskeletal) sites, resulting in the initiation of bone formation by induction (Urist, 1965; Reddi and Huggins, 1972). The natural milieu of the extracellular matrix of bone is home to multiple members of the TGF-I3 superfamily, including BMPs in low abundance (Urist et al., 1983; Wang et al., 1988; Luyten et al., 1989; Sampath et al., 1990), and TGF-I31 and -2 in abundant levels (Centrella et al., 1994). Whether the inherent bone inductive activity of intact demineralized bone matrix resides within a combinatorial action of a plurality of morphogens or to a specific group of proteins has been resolved by purification to homogeneity and expression cloning of BMP family members (BMP-2 through BMP-6 and osteogenic protein-1 and -2, OP-1 and OP-2) (Wozney et al., 1988; Celeste et al., 1990; Ozkaynak et al., 1990; 1992), providing evidence that recombinant human (h)BMP-2, BMP-4 and OP-1 (BMP- 7) singly initiate bone formation in heterotopic sites of the rat (Wang et al., 1990; Hammonds et al., 1991; Sampath et al., 1992). This functional

prerogative, originally ascribed solely to BMPs/OPs, has now been extended to other TGF-j3 superfamily members, including DPP and 60A gene products, expressed early in Drosophila development, and more recently, growth/differentiation factor-5 (GDF- 5 ) following demonstration of its osteoinductive activities in the subcutaneous space of the rat (Sam- path et al., 1993; Hotten et al., 1996). In contrast to BMPdOPs, TGF-O1 and -2, purified either from platelets or produced recombinantly, have to date failed to induce heterotopic bone formation in the in vivo bioassay in rodents, eliciting instead a fibrovascular response (Roberts et al., 1986; Sampath et al., 1987; Hammonds et al., 1991). The fact that TGF-Bs are not inductive in rodents whilst BMPs/OPs singly initiate bone formation may not exclude the require- ment and interaction of both morphogens with other growth factors during the complex and coordinated molecular events controlling cell proliferation and differentiation culminating in the organogenesis of bone complete with marrow (Centrella et al., 1994; Reddi, 1994). The relationship and interaction between osteogenic and non-osteogenic members of the TGF-P superfamily are not understood. Although TGF-Bs share limited homology with members of the BMP/OP family, BMP-4 and OP- 1 do not have overlapping receptor binding speci- ficity with TGF-ps (ten Dijke e ta l . , 1994; Rosen- zweig et al., 1995; Liu et al., 1995). Information on gene expression induced by BMPs/OPs and TGF-Bs could be critical for translation in clinical contexts, and may help to design novel therapeutic strategies based on interactions between osteogenic and non- osteogenic members of the TGF-B superfamily.

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-BI AND OP-l 261

Using the calvaria of the adult baboon (Papio ursinus) as a model for tissue morphogenesis and regeneration, we have recently shown that a single application of hOP- 1, in conjunction with a collagenous matrix as carrier, induced complete regeneration of calvarial defects (Ripamonti et al., 1996a), whereas a single application of hTGF-R1 resulted in limited chondro-osteogenesis confined to the margins of the defects only (Ripamonti et al., 1996b). Here we show that, in marked contrast to previous heterotopic studies in rodents, platelet-derived TGF-B 1 induces endochondral bone formation in heterotopic sites, and interacts synergistically with hOP-1 resulting in the organogenesis of large ossicles both heterotopically in the rectus abdominis and orthotopically in calvarial defects of the baboon. We further demonstrate by Northern analysis of heterotopically induced tissue by hOP- 1 and TGF-B1, singly or in combination, the expression of multiple morphogenetic markers of bone formation, including BMP-3, OP-1 and TGF-,5 mRNA species, deployed during the cascade of endochondral bone formation in the primate.

METHODS

Materials

hOP- 1, prepared as mature glycosylated 36-kD ho- modimer of 139 amino acid residue chain (Sam- path et ul., 1992), was kindly provided by Cre- ative BioMolecules Inc. (Hopkinton, MA). cDNA for y-actin, type I1 collagen and type IV collagen were kind gifts of W. de Wet (University of the Witwatersrand, Johannesburg). Type I1 collagen cDNA was a 2.2 kb fragment of the gene, coding from amino acid 343 of the triple helix to 70 amino acids inside the C-propeptide (Baldwin et ul., 1989), and cDNA for type IV collagen alpha 2 was most of the collagenous domain and part of the non-collagenous domain (Hostikka and Tryggva- son, 1988). cDNA for BMP-3 containing a 1508 bp EcoRI fragment including the full coding region of hBMP-3 (Wozney et al., 1988) was kindly provided

by A. H. Reddi (Johns Hopkins University, MD). cDNA for OP-1 (Ozkaynak et al., 1990), compris- ing a 679 bp fragment of hOP-1 (pO320) covering amino acids 63-263 of the pro-region and the first 25 amino acids of the N-terminal part of the mature polypeptide (Helder et al., 1995), and cDNA for TGF-B1 (plasmid pRK5-BlE) (Arrick et al., 1992) containing the unmodified wild type TGF-B1 precursor (Derynk et al., 1985), were gifts of S. Vukicevic (University of Zagreb, Croatia).

Porcine TGF-Bl and Immunoblot Analysis

TGF-R 1 was purified from 50 g of lyophilized porcine platelets (Zymbio as., SnLa, Norway) as described (Assoian, 1987) with the modification that the gel filtration step on BioGel P-60 was replaced by chromatography of the crude extract onto Sephacryl S-1 00 (Pharmacia, Uppsala, Sweden), eluted with 1 M acetic acid. For improved resolution of high loads the proteins were passaged twice through the same column in a single run. Fractions (50 ml) containing TGF-I31 as determined by ELISA assay were pooled, concentrated and exchanged into 8 M urea, 1 M acetic acid and re-chromatographed in the latter solvent as described above. The purity of the morphogen following chromatography on a Vydac reverse phase C18 HPLC column (Assoian, 1987) was determined by densitometric analysis (Biomed Instruments Inc., CA) of silver stained 15% SDS-polyacrylamide gels (Laemmli, 1970). Reduced pTGF-Bl migrated to a position corresponding to approximately 12 500 D, and to a position of 25 000 D on non-reduced gels (not shown). For ELISA, pTGF-R1 in 20 mh4 sodium acetate pH 5.0 was bound to 96-well Polysorp plates (Nunc, Roskilde, Denmark) overnight at 37 "C and block- ing achieved with 5% bovine serum albumin in PBS, 0.05% TweenTM 20 (Merck, Schuchardt, Germany). Plates were incubated sequentially with chicken anti- pTGF-R1 IgG (R&D Systems, MN), biotinylated rabbit anti-chicken IgG and streptavidin alkaline phosphatase (Amersham, Buckinghamshire, UK). Detection was with 3 mM p-nitrophenyl phosphate

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

262 N. DUNEAS et al.

in 0.1 M diethanolamine pH 9.2 buffer containing 5 mM MgC12 on a microplate reader at 414 nm. The yield of the morphogen, as quantitated"against a standard curve constructed using pTGF-I31 (R&D Systems, MN) was 5.6 pg pTGF-pl/g of porcine platelets, with a total yield of 280 pg. For Western blotting, proteins were resolved by 15% SDS-PAGE and electrophoretically transferred (Hoeffer Scien- tific Instruments, CA) to a polyvinyldene difluoride membrane (Amersham, UK). The membrane was blocked with 5% skim milk powder in PBS, 0.1% TweenTM, incubated with chicken anti-pTGF-01 IgG (R&D Systems, MN), followed by biotinylated rabbit anti-chicken IgG and streptavidin alkaline phosphatase (Amersham, UK). Detection was with the 4-nitro blue tetrazolium chloride and S-bromo-4-chloro-3- indolyl-phosphate (Boehringer Mannheim GmbH, Germany) colorimetric method (Leary et al., 1983).

Preparation of pTGF-/?l and hOP-1 Implants

From a therapeutic perspective, a carrier matrix is required for local delivery of BMPs/OPs and TGF-Bs to obtain a desired morphogenetic response. The extracellular matrix of bone is an optimal substratum for cell recruitment, attachment, proliferation and differentiation (Sampath and Reddi, 1981; Reddi, 1994), as well as a delivery system for BMPs/OPs, both naturally-sourced or in recombinant form (Sampath and Reddi, 1981; Sampath et al., 1992; Ripamonti et al., 1996a). In addition, since both OP and TGF-B were originally extracted and isolated from bone matrix (Wozney et al., 1992; Reddi, 1994; Centrella et al., 1994), the carrier used in this study to deliver pTGF-p 1 and hOP- 1, singly or in combination, was insoluble collagenous bone matrix, the inactive residue obtained after dissociative extraction of baboon bone matrix with 4 M guanidinium-HC1 (Sampath and Reddi, 1981; Ripa- monti et al., 1991). Implants for heterotopic implantation in the rectus abdominis were prepared in ster- ile 5 ml capacity polypropylene tubes by adding 5 , 25 and 125 pg hOP-l in 50 p1 of 5 mM HCI to 100 mg of insoluble collagenous matrix as carrier.

For the combination studies, pTGF-01, dissolved in 50 pl of 20 mM sodium acetate was added to 25 pg hOP-1 samples at doses of 0.5, 1 .5 and 5 pg. For the early time course examination (day 14), pTGF-I31 was tested singly at 0.5, 1.5 and 5 pg per 100 mg of collagenous matrix. For the 30 days study, pTGF-Bl was tested singly at 5 pg per 100 mg of collagenous matrix. For the preparation of samples suitable for calvarial implantation in the baboon, the required amounts of morphogen in 500 pl of respective vehi- cle was combined with 1 g of collagenous matrix per implant and lyophilized. Three morphogen combinations were tested: 100 pg hOP-1 and 5 pg pTGF-pl, 100 pg hOP- 1 and 15 pg pTGF-B 1, and 20 pg hOP- 1 and 5 pg pTGF-Bl. hOP-1 was tested singly at 20 pg per g of matrix, representing a 5 fold reduction of the lowest dose of the recombinant morphogen so far tested in the primate calvarial model (Ripamonti et al., 1996a).

Primate Models for Tissue Induction

Sixteen clinically healthy adult Chacma baboons (Papio ursinus), with a mean weight of 28.1 !c

6.1 kg, were selected from the primate colony of the University of the Witwatersrand, Johannesburg. Criteria for selection, housing conditions and diets were as previously described (Ripamonti, 1991). Research protocols were approved by the Ani- mal Ethics Screening Committee of the university. Research was conducted according to the Guidelines for the Care and Use of Experimental Animals prepared by the university, and in compliance with the National Code for Animal Use in Research, Educa- tion and Diagnosis in South Africa. The heterotopic and orthotopic models of tissue induction and morphogenesis by native and recombinant morphogens in the adult baboon have been described in detail (Ripamonti et al., 1991; 1992; 1996a; 1996b). For the 14 days study, morphogen combinations were implanted in single pellets in ventral intramuscu- lar pouches created by sharp and blunt dissection in the right rectus abdominis of 4 baboons. hOP-1 and pTGF-PI, tested singly for the dose response curves, were implanted in the left rectus muscle,

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-BI AND OP-l 263

duplicate hOP-1 implants in 2 animals, and duplicate pTGF-/31 implants in the remaining 2 animals. For the 30 days heterotopic study, lyophilized pellets of collagenous matrix containing pTGF-Bl and hOP- 1 combinations were implanted ipsilaterally in duplicate in the rectus abdominis of 4 baboons, scheduled for the additional calvarial study and for euthanasia on day 30. Contralateral pouches (across the midline) in 2 animals were implanted with doses of 25 pg hOP-l per 100 mg of collagenous matrix and with insoluble collagenous matrix without morphogens as control for molecular analysis studies. In the remaining 2 animals, contralateral pouches were implanted with collagenous matrix containing 5 pg pTGF-PI (n = 8) and with additional implants of collagenous matrix without either morphogens as control. hOP-I was tested singly for the dose response study (0, 5 , 25 and 125 pg) after implantation of duplicate pellets in each animal in an additional 4 baboons.

The calvariae of the 8 baboons allocated to the orthotopic study, were exposed and, on each side of the calvaria, 2 full-thickness defects, 25 mm in diameter, each separated by 2 . 5 3 cm of intervening calvarial bone, were created with a craniotome under saline irrigation (Ripamonti et al., 1996a; 1996b). An ipsilateral design was used to allocate the position of the implants in the 32 calvarial defects. In each animal, 2 ipsilateral defects were implanted with the higher dose of hOP-1 in combination with either 5 or 15 pg pTGF-PI. Contralateral defects were implanted with the lower dose of hOP-1 (20 pg per g of collagenous matrix), singly or in combination with 5 pg of pTGF-PI. The treatment allocation was alternated anterio-posteriorly, to achieve a balanced distribution between anterior and posterior defects.

Tissue Processing, Histology and Biochemical Analysis

Anesthetized animals were killed with an intra- venous overdose of sodium pentobarbitone, 4 animals on day 14, 8 animals on day 30 and 4 animals on day 90. Generated specimens on day 14 and 30

in the rectus abdominis were dissected free of soft tissues, weighed, and processed for undecalcified histology and serial sections, cut at 4 pm (Polycut- S, Reichert, Heidelberg, Germany), were stained, free-floating, with Goldner’s trichrome. Consecu- tive sections were mounted and stained with 0.1% toluidine blue in 30% ethanol. Additional tissue from duplicate specimens was homogenized in 2 ml of ice-cold 3 mM sodium bicarbonate containing 0.15 M NaC1, pH 7,s. After centrifugation, the alkaline phosphatase activity of the supernatant was used as index of bone formation by induction (Reddi and Huggins, 1972; Ripamonti et al., 199 1). Remaining tissue from the large specimens harvested on day 30 was processed for Northern analysis as described below. Orthotopic specimens with surrounding calvaria were cut along one fourth of the implanted defects (Ripamonti et al., 1992; 1996a; 1996b), dehydrated in ascending grades of ethanol and embedded, undecalcified, in resin (K-Plast, Medim, Buseck, Germany). Undecalcified serial sections were cut at 7 pm and stained, free floating, with the Goldner’s trichrome or, after mounting, with toluidine blue.

Morphornetric Analysis

Calvarial sections were analyzed at 40x with a Provis AX70 research microscope (Olympus Opti- cal Co., Japan), superimposing a calibrated Zeiss Integration Platte I1 with 100 lattice points over five sources (Parfitt et al., 1987) selected for histomorphometry for determination, with the point counting technique (Parfitt, I993), of mineralized bone, osteoid and residual collagenous matrix volumes (in %) (Ripamonti et al., 1992; 1996a; 1996b). Sections generated from the heterotopic specimens were evaluated by superimposing the Zeiss gratic- ule over corticalized outer levels, and transversing to internal regions of the induced ossicles. The cross sectional area (in mm2) of newly generated bone tissue (mineralized bone, osteoid and bone marrow) (Parfitt et al., 1987) in both heterotopic and orthotopic specimens was measured using a computerized image analysis system (Flexible Image

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


Processing System@, CSIR, South Africa). Whilst on day 14 only generated bone tissue was calculated, on day 30 other fibrovascular tissue and residual collagenous matrix, when surrounded by newly formed bone, were included in the analysis (Ripamonti et al., 1996a). Morphometry (volumes and areas) was performed on 2 sections per heterotopic implant, representing 2 levels approximately 800 pm from each other, and on 2 sections per orthotopic implant, representing 2 sagittal levels, approximately 2 mm apart from each other (Ripa- monti et al., 1996a; 1996b).

Northern Blot Analysis

Samples in the range of 80 to 100 mg from replicate specimens of harvested tissue on day 30 from the rectus abdominis were pooled, crushed, and total RNA was isolated according to Nemeth et al. (1989) with minor modifications. Following density gradi- ent centrifugation through CsCl, RNA pellet was purified using RNA purification kits (RNeasy total RNA kit, Qiagen, Hilden, Germany). The quality of the RNA was judged by the A260/A280 ratio of absorbance, and by y-actin signals on Northern blots, and scrutiny of the quality of the ribosomal RNA bands visualised on agarose gels. Samples of 5 and 20 pg RNA were resolved on 1.2% agarose gels containing 2.2 M formaldehyde and transferred to nylon membrane filters (Hybondm-Nf, Amer- sham, UK). cDNA probe inserts were excised from vectors, resolved on 2% agarose gel, purified using a gel extraction kit (Qiagen), and radiolabelled to high specific activity with u ~ ~ P ~ C T P by the random prime method using a commercially supplied kit (RPN 1606, Amersham, UK). Blots probed with type IV collagen were washed under higher strin- gencies (68°C) with 0 . 1 ~ SSC, 0.1% SDS, to min- imize binding to ribosomal RNA (Hostikka and Tryggvason, 1988). Membranes were exposed to Kodak film (Biomax MS) with intensifying screens for 12 h to 96 h. Signals were quantitated relative to their respective y-actin signals on the same blots by densitometric analysis (Biomed Instruments Inc., CA), and adjusted onto a scale of percent of maximum effect.

Statistical Analysis

The data were analyzed with the Statistical Analysis System (1989). An F test was performed using the General Linear Models procedure for an analysis of variance with multiple interactions. For each treatment modality, 4 to 8 specimens were available for morphometry and biochemistry, representing 4 to 8 alkaline phosphatase activity determinations, and 8 to 16 morphometric observations (derived from 2 serial sections from each specimen) per group. Comparison of mean values was obtained using Sheffe’s multiple comparison procedure on the dependent variables included in the analysis.

RESULTS

pTGF-B1 and hOP-1 Induce Bone Differentiation in the Rectus Abdominis

To determine tissue induction and morphogenesis by hOP-l and pTGF-Bl, morphogens were added singly to 100 mg of collagenous matrix as carrier and implanted in the rectus abdominis of the baboon. As expected, 5 ,25 and 125 pg hOP-1 induced bone formation by day 14 (albeit limited) and 30 in the rectus abdominis (Figures 1B and 2A). In marked contrast to previously published negative results in rodents, histological analysis of specimens of collagenous matrix combined with 1.5 and 5 pg pTGF- B1 showed islands of endochondral bone formation as early as day 14 (Figure lA), and almost complete ossicles were generated by 5 pg pTGF-Bl on day 30 (Figure 2B). The inductive efficiency of 5 pg pTGF-Bl was comparable to doses of 5 and 25 pg hOP-l as determined by bone tissue area on day 14, generated tissue area on day 30, mineralized bone and osteoid volumes, and tissue alkaline phosphatase activity (Figures 3 and 4). Tissue generated on day 30 by 5 pg pTGF-Bl or 25 pg hOP-1 induced comparable expression levels of OP- 1, BMP-3 and type IV collagen mRNA transcripts (Figure 5). A striking difference, however, was the finding that pTGF-Bl-treated implants induced a 2

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-I31 AND OP-l 265

FIGURE I (See Colour Plate at back of' issue.) Tissue morphogenesis on day 14 by pTGF-fiI and hOP-l implanted singly, and synergistic induction of bone formation by binary morphogen applications in the rectus abdominis of the baboon. a. bone induction by 1.5 pg pTGF-BI. b, endochondral bone formation by 125 pg hOP-I. c and d, synergistic bone induction by combinations of 25 pg hOP-1 with 1.5 (c) and 5 pg pTGF-fiI (d). Extensive induction of mineralized bone and osteoid, and areas of chondrogenic tissue when using 5 pg pTGF-p1 (d). (Goldner's trichroine stain, original magnification: a and b, x75; c and d, x30).

FIGURE 2 (See Colour Plate at back of issue.) Tissue morphogenesis and synergistic activity of hOP-l and pTGF-fi1 implanted singly or in combination in the rectus abdominis of the baboon and harvested on day 30. a and b, bone induction by a single administration of 25 bg hOP-l (a) and 5 pg pTGF-BI (b) delivered by 100 pg of guanidinium-inactivated collagenous matrix. c and d, large ossicles generated by 25 pg hOP-l combined with 1.5 ( c ) and 5 pg pTGF-Bl (d). Areas of chondrogenesis (arrows in c ) at the periphery of the newly generated tissue and corticalization of the ossicles (original magnification, x4).

to 3 fold increase in TGF-Pl and type I1 collagen expression when compared to 25 pg hOP-I, as determined by Northern analysis (Figure 5). High expression levels of type I1 collagen was predicted

by the histological evidence of chondrogenesis in pTGF-PI-treated specimens. Implants of collagenous matrix without either morphogen (control) had no bone differentiation activity (Figures 3 and 4) as

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


FIGURE 3 Synergistic interaction of pTGF-BI and hOP-l on day 14. Values are mean & SEM of 4 to 8 specimens per group. Levels of significance were determined using the Statistical Analysis System (1 989) General Linear Models procedure with correction for unbalanced data as described in Methods. a, computerized analysis of newly formed bone tissue (in mm’) (mineralized bone, osteoid and marrow) (Parfitt et al., 1987) induced by doses of morphogens, singly or in binary applications. b, mineralized bone and osteoid volume (in %) generated by doses of pTGF-BI and hOP-I, singly or in combination. c, effect of pTGF-BI on alkaline phosphatase activity in the newly generated tissues harvested on day 14. The alkaline phosphatase activity of the supernatant after homogenization of implants was determined with 0.1 M p-nitrophenyl phosphate as substrate at pH 9,3 and 37°C for 30 min. One unit of enzyme liberates one micromole of p-nitrophenol under the assay conditions. Alkaline phosphatase is expressed as units of activity per mg protein (Reddi and Huggins, 1972; Ripamonti et al., 1991). Protein concentration in the supematant was measured by the method of Lowry ef al. (1951). a indicates significant differences with morphogen combinations, with the exclusion of the binary application of 0.5 pg pTGF-PI with 25 pg hOP-1 in panel c.

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


FIGURE 4 Effect of pTGF-BI on parameters of heterotopic bone formation induced by hOP-l on day 30. Values are mean f SEM of 4 to 8 specimens per group. a, computerized analysis of newly generated tissue (in mm2) induced by doses of morphogens, singly or in combination. b, histomorphometric results of induced mineralized bone and osteoid volume (in %) in the newly formed ossicles. c, effect of pTGF-BI on tissue alkaline phosphatase activity in the newly generated heterotopic ossicles. a indicates significant differences with morphogen combinations (with the exclusion of the binary application of 0.5 pg pTGF-Bl with 25 pg hOP-l in panel c), and with 5 pg pTGF-j31 (c).

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


FIGURE 5 Northern analysis of OP-1, BMP-3, TGF-PI, and type I1 and type IV collagens mRNA expression of tissues generated by hOP-1 and pTGF-Bl. hOP-1 and pTGF-PI were implanted singly or in combination with collagenous matrix in the rectus abdominis of the baboon, and replicate generated tissues ( n = 4) were harvested on day 30, pooled and total RNA extracted as described in Methods (Nemeth et al., 1989). Samples of total RNA were subjected to Northern analysis for type I1 and type IV collagens ( 5 yg per lane), and for OP-l , BMP-3 and TGF-PI (20 yg per lane). The y-actin mRNA, and the 18s and 28s ribosomal RNA signals were used as standard for size determination. mRNA levels are expressed in relative densitometric units.

previously reported (Ripamonti et al., 1991; 1992), and the detection of OP-1 mRNA in control specimens (Figure 5 ) may signify a possible role of OP-1 in soft tissue repair.

hOP-1 and pTGF-B1 Interact Synergistically During Induction of Endochondral Bone Formation

To determine potential synergistic interactions between hOP-1 and pTGF-B1, morphogen combinations were added to collagenous matrix and

implanted in the rectus abdominis of the baboon. On day 14, binary applications of pTGF-Bl and hOP-l resulted in synergistic induction of tissue morphogenesis as determined by histology (Figures 1C and lD), histomorphometry, and biochemistry (Figure 3). Combinations of 1.5 and 5 pg pTGF-Bl with 25 pg hOP-1 induced extensive amounts of osteoid and mineralized bone with foci of chondrogenic tissue mainly local- ized at the periphery of the newly generated ossicles (Figure 1D). Bone tissue (mm2) and bone and osteoid volumes (%) in specimens generated by

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-I31 AND OP-I 269

morphogen combinations were increased synergistically by several fold when compared to single applications of 25 or even 125 pg hOP-1 (Figures 3A and 3B). On day 30, the addition of comparatively low doses of pTGF-pl (0.5, 1.5 and 5 pg) to 25 pg hOP-l resulted in the generation of large corticalized ossicles displacing the muscle fibers of the rectus abdominis (Figure 2). Though by day 30 binary applications might have reached a plateau of activity, morphogen combinations resulted in a 2 to 3 fold increase in generated tissue area when compared to a single application of both morphogens (Figure 4A). Maximal increase in mineralized bone and osteoid volumes were observed with combinations of 0.5 and 1.5 pTGF-B 1 and 25 pg hOP-1, where osteoid volumes reached a 2 fold increase when compared to a single application of 25 pg hOP-l (Figure 4B). Doses of 1.5 and 5 pg pTGF-B 1 increased the inductive efficiency of 25 pg hOP-1 by raising 2 to 3 fold the alkaline phosphatase activity in ossicles generated by morphogen combinations, though a single application of 5 pg TGF-B1 resulted in substantial alkaline phosphatase activity (Figure 4C). It is noteworthy, however, that binary applications of 1.5 pg pTGF-Bl with 5 fold less of the highest tested dose of hOP- 1 , resulted in key parameters of tissue induction superior (tissue area and bone volume) or comparable (alkaline phosphatase activity) to the 125 pg dose of hOP-l (Figure 4). In ossicles generated by 25 pg hOP- I in combination with pTGF-Bl, type I1 collagen expression increased in a pTGF-BI dose- dependent manner, and approached levels induced by a single application of 5 pg pTGF-Bl (Figure 5) . Type IV collagen mRNA synthesis in implants containing morphogen combinations showed a 3 to 4 fold increase when compared to ossicles generated by a single application of 5 pg pTGF-Bl or 25 pg hOPl (Figure 5) . In contrast, BMP-3, OP- 1 and TGF-PI mRNA synthesis was 3 fold lower in ossicles generated by combinations of 1.5 and 5 pg pTGF-Bl with 25 pg hOP-1 (Figure 5) . In ossicles generated by combinations of 0.5 pg pTGF-Bl and 25 pg hOP-1, expression levels of OP-l and BMP-3 mRNA were identical to OP-1 and BMP-3 expression levels induced by a single application

of pTGF-Bl and hOP-1 (Figure 5 ) . It is noteworthy that the combination of 0.5 pg pTGF-Bl and 25 pg hOP-1 induced high levels of TGF-BI mRNA expression, comparable to levels induced by 5 pg of pTGF-Bl, with a 3 fold increase when compared with TGF-Bl expression induced by a single application of 25 pg of hOP-1 (Figure 5) .

Morphology of Calvarial Regeneration by hOP-1 and pTGF-Bl Singly or in Combination

To determine the efficacy of naturally-sourced or recombinantly produced morphogens as potential therapeutic initiators of osteogenesis, this laboratory has developed an orthotopic model in which large osseous defects are prepared in the calvaria of adult baboons (Ripamonti et nl., 1992; 1996a; 1996b). Using the baboon calvarial model, we have recently shown that a single application of 100 pg of hOP-I, delivered by an identical collagenous matrix preparation as described in the present study, induced complete regeneration of the calvarial defects by day 90 (Ripamonti et al., 1996a). In marked contrast, 5 , 30 and 100 pg of recombinant hTGF-Bl, delivered by the same collagenous matrix, failed to regenerate calvarial defects in the adult baboon, resulting in limited chondro-osteogenesis at the margins of the defects, and in enhanced fibrosis and vascular invasion when compared to controls of collagenous matrix alone (Ripamonti et nl., 1996b). We next administered combinations of different doses of both hOP-1 and pTGF-Bl to investigate the combinatorial action of both morphogens in 32 calvarial defects prepared in 8 adult baboons. At harvest, calvarial specimens treated with combinations of hOP- 1 and pTGF-/31 on day 30 showed displacement of the pericranium and associated temporalis muscle. Cut surfaces were brownish-red in the gross, indicating prominent angiogenesis and vascular invasion. On day 30, treatment with 20 pg of hOP-l initiated sparse bone formation pericranially and in proximity to the defect margins (Figure 6A). The addition of 5 pg pTGF-Bl to 20 pg hOP-l resulted in pericranial and endocranial osteogenesis across

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


FIGURE 6 (See Colour Plate at back of issue.) Tissue induction and morphogenesis of bone in calvarial defects by hOP-1 alone or in combination with doses of pTGF-pl on day 30. a and b, low power photomicrographs of defects treated with 20 pg hOP-1 (a) and with 20 pg hOP-1 combined with 5 pg pTGF-pl (b). c, d, e, and f, calvarial defects treated with 100 pg hOP-I, alone (c) and in combination with 5 (d and e ) and 15 pg pTGF-BI (f). Formation of trabeculae of mineralized bone across the entire defect but within the profile of the original calvaria following treatment with 100 pg hOP-l without pTGF-Bl (c). Extensive pencranial osteogenesis in specimens treated with morphogen combinations, culminating in gross displacement of the pencranial tissues ( e and f). Goldner’s trichrome, original magnification x2.8

the treated defects (Figure 6B). When compared to a single application of 100 pg hOP-1 (Ripamonti et al., 1996a) (Figure 6C), the addition of 5 and 15 pg of pTGF-Bl to 100 pg of hOP-1 resulted in prominent pericranial osteogenesis with pronounced lifting of the newly formed mineralized bone pericranially, grossly displacing the temporalis muscle (Figures 6D, 6E and 6F). The mineralized bone that had formed beneath the temporalis muscle and along the dura was separated by a loose connective tissue, characterized by the presence of an internal, dividing

zone with prominent vascular invasion and scattered remnants of collagenous matrix (Figures 6D, 6E and 6F). Numerous delicate trabeculae of newly formed bone, covered by continuous osteoid seams populated by contiguous osteoblasts, characterized the pericranial zone of osteogenesis. The histological features seen on specimens generated by morphogen combinations were remarkably similar to the histological results obtained using higher doses of hOP-1 (0.5 and 2.5 mg hOP-1 per g of collagenous matrix) as evaluated on day 30 in the adult

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-I31 AND OP-l 27 1

FlGURE 7 (See Colour Plate at back of issue.) Morphology of calvarial regeneration by hOP-l alone or in combination with doses of pTGF-Pl on day 90. a and b, defects treated with 20 pg hOP-1 (a) and with 20 ~g hOP-l combined with 5 pg pTGF-j31 (b). Optimization of bone formation and augmentation of the regenerate by the addition of 5 pg pTGF-PI, approaching levels of bone tissue formation comparable with regenerates induced by 100 pg hOP-l alone (c and Figure 8B). c and d, remodelling and adaptation of bone tissue within the profile of the original calvaria in a defect treated with 100 pg hOP-l (c), but persistence of distorted pericranial profile and scattered remnants of collagenous matrix within a zone of fibrovascular tissue invasion beneath the pericranial corticalized bone layer in a specimen treated with 100 kg hOP-l combined with 15 pg pTGF-BI (d). Goldner's trichrome stain, original magnification x2.8.

baboon (Ripamonti et al., 1996a). Both treatments (morphogen combinations and high doses of hOP- 1) resulted in a marked proliferative and inductive phase at the periphery of the implanted matrix, par- ticularly at the level of the pericranium and temporalis muscle, followed by an exuberant osteogenic effect that resulted in the loss of cohesiveness between the pericranial and endocranial osteogenic sites (Figures 6D and 6E). On day 90 (Figure 7), defects treated with 20 pg hOP-1 were markedly thinner when compared to regenerates obtained by the combination of 5 pg pTGF-Bl and 20 pg hOP-1 (Figures 7A and 7B). Defects treated with 100 pg hOP-1 in combination with 15 pg pTGF-Bl showed marked and exuberant osteogenesis, persistence of pericranial lifting with temporalis muscle displacement and of zones of fibrovascular tissue influx with scattered collagenous matrix particles (Figure 7D).

Histomorphometry: Effect of pTGF-/?l on Bone Formation by hOP-1

Volume fractions (with level of significance) of bone and osteoid in defects treated with hOP-l and

TGF-B1, singly or in combination, are presented in Table I. Histomorphometric data of the present series of calvarial defects were compared with previously published results using 100 pg of hOP-l (Ripamonti et al., 1996a) and 100 pg of recombinant hTGF-Bl (Ripamonti et al., 1996b), delivered by an identical preparation of collagenous matrix (Table I). On day 30, all doses of hOP-1, either singly or in combination with 5 or 15 pg of pTGF- B1, induced greater amounts of bone and osteoid volumes when compared with specimens treated with 100 pg hTGF-Bl ( P < 0.01, Table I). Spec- imens treated with 20 pg hOP-1 in combination with 5 pg pTGF-Bl induced greater amounts of bone and osteoid when compared with specimens treated with a single application of 20 pg hOP-l ( P < 0.05, Table I). Specimens treated with 100 pg hOP-1 induced equal and greater amounts of bone when compared with 100 pg hOP-l combined with 5 and 15 pg pTGF-B1, respectively (Table I). How- ever, morphogen combinations resulted in greater amounts of osteoid volumes when compared with specimens treated with 100 pg hOP-1 alone ( P < 0.05, Table I). On day 90, on average, greater

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


TABLE I 32 calvarial defects in 8 adult male baboons

Effect of pTGF-I31 on bone induction by hOP-l in

Day Morphogen Bone Osteoid Matrix combinations (pg) % % %

hOP- 1

30 20 20

100" 100 100

0 90 20

20 100' 100 100

TGF-I3 1 *

0 5 0 5

15 100

0 5 0 5

15

18.1 5 1.4 4 . 6 f 0 . 3 33.9% 1.6 24.5 i 1.4z 6.7 i 4.4$ 13.0 f 1 1,7** 29.1 i 0.65 4.5 f 0.2 7.7 f 0.6** 31.7 f 1.45 9.0 i 0.5¶ 7.0 f 0.6** 24.7 f 1.2 7.0 f 0.3¶ 9.2 & 1.4** 8.5 f 1 .5 1 .O % 0.2 30.7 f 2.6 35.5 i 1.9 3.1 f 0 . 2 6.3 f 1.3 49.2 f 1.6$ 4.2 f 0.33: 2.7 i 0.5 5 3 . 3 f 1.5'1 2.6&0.l 0.3ztO.l 46.6 & 1.6 4.5 f 0.2¶ 1.6 k 0.6 56.9 * 1.7l1 6.9 f 0.3¶ 0.7 f 0.2

Single and binary applications of morphogens were applied once at time of surgery delivered by 1 g of baboon collagenous matrix. Operated sites were harvested on day 30 and 90, and serial undecalcified sections, cut at 7 pm, were analyzed by morphometry. Volume fractions of tissue components (in 76) were calculated using a Zeiss Integration Platte I1 with 100 lattice points superimposed over 5 sources (each representing 7.84 mm2) in each of 2 sagittal sections used for analysis as described in Methods. Bone indicates mineralized bone plus osteoid. Matrix indicates the residual collagenous carrier used for local delivery of morphogens. Values are means f SEM of 4 calvarial specimens per group. Levels of significance were determined using the Statistical Analysis System General Linear Models Procedure with multiple interactions on a balanced set of data (1989). *The 5 and 15 pg doses of TGF-PI refer to purified porcine TGF-PI, used in the present experiments, whilst the 100 pg dose refers to previously published data on day 30 in 4 adult male baboons implanted with 100 pg of recombinant human TGF-PI (Ripamonti er al., 1996b). The 100 pg dose of hOP- I alone represents data from 8 calvarial defects on day 30 and 90 (Ripamonti et al., 1996a), implanted in conjunction with an identical preparation of collagenous matrix as used in the present experiments. $f < 0.05 vs 20 pg hOP-1 alone on d 30 and 90. sf < 0.05 vs 100 pg hOP-l and 15 pg pTGF-PI on d 30.

< 0.05 vs 100 pg hOP-l and 5 pg pTGF-PI. ¶ P < 0.05 vs 100 pg hOP-l alone on d 30 and 90. **P < 0.05 vs 20 pg hOP-I and 100 pg hTGF-pI on d 30.

amounts of bone and osteoid volumes were found in specimens treated with 100 pg hOP-1 combined with 15 pg pTGF-B1. This difference, however, was significant only for osteoid volume ( P < 0.05, Table I), indicating substantial osteoid deposition (7.0 and 6.9% on day 30 and 90, respectively) sustained by continuous osteoblastic synthesis in specimens treated with the 15 pg dose of pTGF- p1. Treatment with 100 pg hOP-1 combined with 5 pg pTGF-pl showed significantly less bone but greater osteoid volume when compared with specimens treated with 100 pg hOP-1 alone ( P < 0.05, Table I). The histomorphometric results indicate a 2

200

150

100

50

0

Ipg TGF-I31 I d 30

0 20 100

d 90

100

50

0

0 20 100 hOP-I pg/g collagenous matrix

FIGURE 8 Computerized analysis of area of new tissue (mineralized bone, osteoid, marrow, and other fibrovascular tissue and residual collagenous matrix on day 30 when surrounded by mineralized bone) generated by doses of hOP-l alone or in combination with 5 and 15 pg pTGF-PI revealed a pTGF-PI dose-dependent increase in the cross-sectional areas of newly generated tissue on day 30 (a). On day 90 (b), addition of 5 and 15 pg pTGF-BI to 20 and 100 pg hOP-1 respectively, resulted in a 2 fold increase in newly generated bone tissue. With the exclusion of the treatment combination of 100 pg hOP-1 and 15 pg pTGF-Bl, the exuberant tissue formation shown on day 30 (a) was remodelled rapidly to normal unoper- ated calvaria (60.8 f 3 . 1 mm', inset: Figure 8B). Data represent means f SEM. a and b indicate a significant difference with calvarial specimens treated with morphogen combinations on day 30 (a) and 90 (b), respectively.

fold increase in bone volume between day 30 and 90 in specimens treated with 20 pg hOP-1 alone and in combination with 5 pg pTGF-pl, and in specimens treated with 100 pg hOP-1 in combination with 15 pg pTGF-Bl (Table I). Specimens treated with 100 pg hOP-1 in combination with 5 pg pTGF- p1 showed the least fold increase in bone volume on day 90, in spite of the highest osteoid volume present on day 30 (Table I).

Computer generated data of cross-sectional areas (in mm2) of morphogen-treated calvarial defects are shown in Figure 8. Exuberant bone and fibrovascular

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


tissue fonnation induced on day 30 by morphogen combinations showed a significant increase in tissue area when compared with specimens treated with 20 and 100 pg hOP-1 alone ( P < 0.05, Figure 8A). This exuberant tissue formation was remodelled rapidly to near normal calvariae (mean cross-sectional area: 60.8 f 3.1 mm2) (Ripamonti et al., 1996a) by day 90 (Figure SB). In contrast to the other treatment modalities, specimens treated with 100 pg hOP-l combined with 15 pg pTGF- B1 showed persistence of exuberant tissue formation when compared with the profile of the normal baboon calvaria, in spite of 50% reduction of bone and fibrovascular tissue between the two time peri- ods (Figure 8). It is noteworthy that on day 90 the combination of 20 pg hOP-1 and 5 pg pTGF-81 resulted in a 2 fold increase in bone tissue when compared to specimens treated with 20 pg of hOP- 1 alone, and approached levels of generated tissue area comparable with the higher dose of hOP-1 (Figure 8B).

DISCUSSION

A striking and discriminatory feature of BMPs/OPs is their capacity to singly initiate bone formation in heterotopic sites, recapitulating events that occur in the normal course of embryonic bone development (Wozney, 1992; Reddi, 1994), a phe- nomenon so far shared by other members of the TGF-jl superfamily but not by TGF-j31 or -82 (Roberts et al., 1986; Sampath er al., 1987; Ham- monds et al., 1991). As predicted from results obtained in rodents (Sampath et al., 1992), hOP- 1, at doses of 5 , 25 and 125 pg per 100 mg of collagenous matrix as carrier, initiated bone formation by induction in heterotopic sites of the adult baboon. Platelet-derived pTGF-B 1, combined with guanidinium-inactivated collagenous matrix, also induced heterotopic endochondral bone formation at doses of 5 pg, with an inductive efficiency comparable to doses of 5 and 25 pg hOP-1 as determined by key parameters of bone formation. This remarkable result obtained in adult baboons contrasts markedly

with previous studies in rodents, where heterotopic implantation of TGF-81 or -82, combined with either the insoluble collagenous matrix or purified collagen in conjunction with ceramic carriers, induces a fibrogenic response, without signs of cartilage or bone formation (Roberts er d., 1986; Sam- path et al., 1987; Bentz et al., 1989; 1991). To rule out the possibility that endogenous BMPs/OPs may have escaped our extraction procedure, the bone matrix used as carrier for pTGF-81 was exhaus- tively extracted with guanidinium-HC1. However, we cannot completely dismiss the possibility that specific BMPs/OPs, binding to the matrix from systemic circulation or endogenously upregulated during heterotopic wound healing, might have syn- ergized with the exogenously-applied pTGF-8 I . The recent immunolocalization of OP- 1 in basement membranes of epithelia and endothelium in absence of detectable mRNA expression (Vukicevic et al., 1994a), strongly indicates deposition within basement membranes at distance from sites of BMPs/OPs synthesis (Ozkaynak et al., 1992; Vukicevic et al., 1994b; Helder et al., 1995). In the studies presented here, TGF-Bl was purified from porcine platelets, cellular elements known to contain other potent growth factors (Assoian et al., 1984), so that the potency of our purified morphogen could be the result of other activities. To rule out this possibility, we also used recombinant hTGF-81 purified from medium of transfected CHO cells (Derynk et al., 1985), and we found that hTGF-BI induced endochondral bone formation in the rectus abdominis of the baboon comparable to porcine TGF-Bl (Ripa- monti et al., 1997). These results make it extremely unlikely that the bone inductive activity observed in the present experiments is due to impurities in the pTGF-PI preparation.

The presence of multiple molecular forms with osteogenic activity raises important questions about the biological relevance of this apparent redundancy, and suggests multiple interactions during endochondral bone formation (Wozney, 1992; Reddi, 1994; Centrella et al., 1994). It is natural to envision that a plurality of morphogens are required, if not to singly initiate, to promote synchronously and sequentially

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


the cascade of bone formation. Data indicating that factors isolated from bone may act in concert to induce heterotopic osteoinduction were reported using demineralized bone matrix (Carrington et al., 1988), and bone-derived fractions containing BMP- 2 and -3 combined with bone-derived TGF-B2 (Bentz et al., 1989; 1991). In rat experiments, the ratio of cartilage to bone was increased by combination of BMP fractions co-administered with increas- ing doses of bone-derived TGF-B2 (Bentz et al., 1989; 1991). In the experiments presented here, the addition of comparatively low doses of pTGF- /?1 (0.5, 1.5 and 5 pg) to 25 pg hOP-1 resulted in rapid and extensive bone formation by day 14, and in the generation of large corticalized ossicles complete with bone marrow by day 30 in the rectus abdominis of the baboon. On day 14, binary applications of hOP-1 and pTGF-Bl induced bone morphogenesis several fold greater than the sum of the effects of both morphogens implanted singly. Whilst we have not performed a dose-dependent induction curve for pTGF-/?l implanted singly on day 30, the results obtained on day 14 indicate that the two morphogens interact synergistically during the induction of endochondral bone formation in the adult primate. The many-fold increase in type IV collagen mRNA synthesis over a single application of hOP-1 and pTGF-B1, suggests that the two morphogens interact synergistically to induce angiogenesis and vascular invasion. Since angiogenesis is a prerequisite for osteogenesis, enhanced vas- cularization (including sinusoidal capillaries in the marrow compartment), may be part of the mech- anism whereby hOP-1 and pTGF-Bl synergize in heterotopic bone induction. Tissue generated by a single application of either morphogen showed high expression levels of their own mRNAs, indicating an autoinductive effect, which, for TGF-B1, is in agreement with previous results obtained in vitro (Robey et al., 1987) and in vivo (Joyce et al., 1990). It is worth noting that tissue generated by pTGF-Bl showed OP-1 and BMP-3 mRNA synthesis, indicating that bone formation induced by pTGF-Bl in the rectus abdominis requires, at least in part, synthesis of members of the BMP/OP family. The

high levels of type I1 collagen mRNA transcripts found in tissue generated by a single application of pTGF-Bl highlights the well-known chondrogenic potential of TGF-B1 (Centrella et al., 1994). Addi- tion of pTGF-Bl to hOP-1 caused a dose-dependent rise in type I1 collagen mRNA synthesis, approaching levels induced by pTGF-Bl alone. The moderate expression of OP-1 and BMP-3 in tissue generated by morphogen combinations at doses of 1.5 and 5 pg pTGF-Bl may signify a temporal acceleration of the induction cascade, and thus with decreased OP-1, BMP-3 and TGF-B1 mRNA synthesis on day 30.

Evidence that TGF-/?s stimulate local osteogenesis in vivo consists of experiments where repeated applications of the morphogen were delivered sub- periosteally in mice and rats (Noda and Camilliere, 1989; Joyce et al., 1990), and, after a single administration, in calvarial defects of rabbits (Beck et al., 1991; 1993) and in humeral defects of dogs (Sumner et al., 1995). In the primate, however, recombinant hTGF-Bl failed to regenerate defects of the calvarial membranous bone (Ripamonti et al., 1996b). These paradoxical effects of TGF-Bl in the primate (osteoinductive in heterotopic but not in orthotopic sites) should be viewed in the context of the contradictory biphasic effects of TGF-B1 on proliferation and differentiated functions of osteo- progenitor and osteoblastic cells in vitro (Centrella et al., 1994), and more recently in vivo (Cassiede et al., 1996). hOP-1, on the other hand, is capa- ble of inducing bone differentiation in both heterotopic and orthotopic sites, changing the fate of cellular elements irrespective of their tissue loca- tion. In the context of the calvarial wound, hOP-l induces bone cell differentiation and TGF-j3 1, on the other hand, acts synergistically with hOP- 1 to promote and accelerate the inductive events, but does not in itself induce bone formation. These effects are specific, and are evidenced by the histological analysis of the tissue regenerates induced when hOP-1 and TGF-Bl are applied singly to the calvarial wound (Ripamonti et al., 1996a; 1996b). The latter morphogen induces and promotes the influx of mesenchymal elements resulting in exuberant bone

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-81 AND OP-l 275

formation in the presence of hOP- 1, and in a marked dose-dependent fibrogenic response when applied alone (Ripamonti et al., 1996b). Co-administration of pTGF-Bl and hOP-1 caused the induction of exuberant regenerates well above the profile of the normal calvaria with displacement of the temporalis muscle. This synergistic activity is possibly the result of superior cell migration and recruitment at site of implantation as well as enhancement of differentiated functions of osteoblasts, as documented by increased osteoid volumes. The orthotopic study further demonstrates that the co-administration of 5 pg pTGF-01 improved dramatically the regenerates obtained by 20 pg hOP-1, resulting in bone volumes comparable to a single administration of 100 pg hOP-1, and in regenerates with geomet- ric and architectural configurations equal to those obtained by a 5 fold higher amount of hOP- 1.

The current data obtained in the adult baboon, a primate that shares a remarkably similar bone remodelling with man (Schnitzler et al., 1993), are relevant to therapeutic tissue engineering of bone. The results of the heterotopic and orthotopic studies indicate a synergistic interaction between two related but different members of the TGF-0 superfamily. TGF-B1 may act as chemotactic and mito- genic factor for responding precursor cells for subse- quent induction by hOP-1, as previously suggested for a combination of bovine bone-derived TGF- 0 2 and BMP fractions (Bentz et al., 1989; 1991), and possibly with simultaneous shift and redis- tribution in receptor binding profiles for TGF-B 1 regulated by hOP-1 (Centrella et al., 1995). Ulti- mately, predictable and rapid bone regeneration for the treatment of human skeletal defects will require information concerning the expression and cross- regulation of members of the TGF-0 superfamily during tissue morphogenesis and regeneration induced by single or combined morphogen applications. In this study, the expression of multiple members of the TGF-B superfamily indicates complex autocrine and paracrine activities of the ligands and different signalling pathways on responding cells during the cascade of endochondral bone formation in the primate. This may provide the

scientific basis for synergistic molecular therapeutics for cartilage and bone regeneration in clinical contexts.

Acknowledgements

This work is supported by the South African Med- ical Research Council, the University of the Wit- watersrand, Johannesburg, and by a grant from the Slome Orthopaedic Research and Development Fund, Johannesburg.

References

Anick, B. A,, Lopez, A. R., Elfman, F., Ebner, R., Damsky, C. H. and Derynck, R. (1992) Altered metabolic and adhe- sive properties and increased tumorigenesis associated with increased expression of transforming growth factor beta 1 . J . Cell Biol., 118, 715-726.

Assoian, R. K. (1987) Purification of type-beta transforming growth factor from human platelets. Methods Enzymol., 146,

Assoian, R. K., Grotendorst, G. R., Miller, D. M. and Sporn, M. B. (1984) Cellular transformation by coordinated action of three peptide growth factors from human platelets. Nature (Lond.), 309, 804-806.

Baldwin, C. T., Reginato, A. M., Smith, C.. Jimenez, S. A. and Prockop, D. J . (1989) Structure of cDNA clones coding for human type I1 procollagen. The alpha 1 (11) chain is more similar to the alpha 1 (I) chain than two other alpha chains of fibrillar collagen. Biochem. J . , 262, 521 -528.

Beck, L. S., DeGuzman, L., Lee, W. P., Xu, Y., McFatridge, L. A,, Gillett, N. A. and Amento, E. P. (1991) TGF-PI induces bone closure of skull defects. 1. Bone Miner. Rex , 6 ,

Beck, L. S . , Amento, E. P., Xu, Y., DeGuzman, L., Lee, W. P., Nguyen, T. and Gillett, N. A. (1993) TGF-PI induces bone closure of skull defects: Temporal dynamics of bone formation in defects exposed to rhTGF-BI. I. Bone Miner. Res., 8,

Bentz, H., Nathan, R. M., Rosen, D., Armstrong. R. M., Thomp- son, A. Y. , Segarini, P. R., Mathews, M. C., Dasch, J. R., Piez, K. A. and Seyedin, S. M. (1989) Purification and characterization of an unique osteoinductive factor from bovine bone. J . Biol. Chem., 264, 20805-20810.

Bentz, H., Thompson, A,, Armstrong, R., Chang, R.-J., Piez, K. A. and Rosen, D. M. (1991) Transforming growth factor- B2 enhances the osteoinductive activity of a bovine bone- derived fraction containing bone morphogenetic protein-2 and 3. Matrix, 11, 269-275.

Carrington, J. L., Roberts, A. B., Flanders, K. C., Roche, N. S. and Reddi, A. H. (1988) Accumulation, localization, and coinpartmentation of transforming growth factor during endochondral bone development. J . Cell Biol., 107,

Cassiede, P., Dennis, J. E., Ma, F. and Caplan, A. I. (1996) Osteochondrogenic potential of marrow mesenchymal progenitor cells exposed to TGF-PI or PDGF-BB as assayed in vivo and in vitro. J. Bone Miner. Rex, 11, 1264- 1273.

153-163.

1257- 1265.

753-761.

1969- 1975.

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.


Celeste, A. J., Iannazzi, J. M., Taylor, J. A,, Hewick, R. M., Rosen, V., Wang, E. A. and Wozney, J. M. (1990) Identifica- tion of transforming growth factor I3 family members present in bone-inductive protein purified from bovine bone. Proc. Natl. Acad. Sci. USA., 87, 9843-9847.

Centrella, M., Horowitz, M., Wozney, J. M. and McCarthy, T. L. (1994) Transforming growth factor B (TGF-B) family members and bone. Endocrine Rev., 15, 27-39.

Centrella, M., Casinghino, S., Kim, J., Pham, T., Rosen, V., Wozney, J. and McCarthy, T. L. (1995) Independent changes in type I and type I1 receptors for transforming growth factor induced by bone morphogenetic protein 2 parallel expression of the osteoblast phenotype. Mol. Cell. Biol., 15,

Derynk, R., Jarret, J. A., Chen, E. Y., Eaton, E. H., Bell, J. R., Assoian, R. K., Roberts, A. B., Spom, M. B. and Goeddel, D. V. (1985) Human transforming growth factor4 comple- mentary DNA sequence and expression in normal and trans- formed cells. Narure (Lond.), 316, 701 -705.

Hammonds, R. G., Schwall, R., Dudley, A,, Berkemeier, L., Lai, C., Lee, J., Cunningham, N., Reddi, A. H., Wood, W. I. and Mason, A. J. (1991) Bone inducing activity of mature BMP-2b produced from a hybrid BMP-2d2b precursor. Mol. Endocrinol., 5,,, 149- 155.

Helder, M. N., Ozkaynak, E., Sampath, T. K., Luyten, F. P., Latin, V., Oppermann, H. and Vukicevic, S. (1995) Expres- sion pattern of osteogenic protein- 1 (bone morphogenetic protein-7) in human and mouse development. J. Histochem. Cytochem., 43, 1035- 1044.

Hostikka, S. L. and Tryggvason, K. (1988) The complete pri- mary structure of the alpha 2 chain of human type IV collagen and comparison with the alpha 1 (IV) chain. J. Biol. Chem.,

Hotten, G. C., Matsumoto, T., Kimura, M., Bechtold, R. F., Kron, R., Ohara, T., Tanaka, H., Satoh, Y., Okazaki, M., Shi- rai, T., Pan, H., Kawai, S., Pohl, J. S. and Kudo, A. (1996) Recombinant human growtWdifferentiation factor 5 stimulates mesenchyme aggregation and chondrogenesis responsible for the skeletal development of limbs. Growth Factors, 13, 65-74.

Joyce, M. E., Roberts, A. B., Spom, M. B. andBolander, M. E. (1 990) Transforming growth factor-B and the initiation of chondrogenesis and osteogenesis in the rat femur. J. Cell Biol.,

Laemmli, U. K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature ( h n d . ) ,

Leary, J. J., Brigati, D. J. and Ward, D. C. (1983) Rapid and sensitive colorimetric method for visualizing biotin-labeled DNA probes hybridized to DNA or RNA immobilized on nitrocellulose: Bio-blots. Proc. Natl. Acad. Sci. USA., 80,

Liu, F., Ventura, F., Doody, J. and MassaguC, J. (1995) Human type I1 receptor for bone morphogenetic proteins (BMPs): Extension of the two-kinase receptor model to the BMPs. Mol. Cell. Biol., 15, 3479-3486.

Lowry, 0. H., Rosebrough, N. J. and Farr, A. L. (1951) Protein measurement with the folin phenol reagent. J. Biol. Chem.,

Luyten, F. P., Cunningham, N. S., Ma, S., Muthukumaran, N., Hammonds, R. G., Nevins, W. B., Wood, W. I. and Reddi, A. H. (1989) Purification and partial amino acid sequence of osteogenin, a protein initiating bone differentiation. J. Biol. Chem., 264, 13377-13380.

3273-3281.

263, 19488- 19493.

110, 2195-2207.

227, 680-685.

4045-4049.

193, 265-275.

Nemeth, G. G., Heydemann, A. and Bolander, M. E. (1989) Iso- lation and analysis of ribonucleic acids from skeletal tissues. Anal. Biochem., 183, 301-304.

Noda, M. and Camilliere, J. J. (1989) In vivo stimulation of bone formation by transforming growth factor-B. Endocrinology, 124, 2991-2994.

Ozkaynak, E., Rueger, D. C., Drier, E. A,, Corbett, C., Ridge, R. J., Sampath, T. K. and Oppermann, H. (1990) OP-l cDNA encodes an osteogenic protein in the TGF-I3 family. EMBO (Eur. Mol. Biol. Organ.) J. , 9, 2085-2093.

Ozkaynak, E., Schnegelsberg, P. N. J. and Oppermann, H. (1991) Murine osteogenic protein (OP-1): High levels of mRNA in kidney. Biochem. Biophys. Res. Commun., 179, 116- 123.

Ozkaynak, E., Schnegelsberg, P. N. J., Jin, D. F., Clifford, G. M., Warren, F. D., Drier, E. A. and Oppermann, H. (1992) Osteo- genic protein-2. A new member of the transforming growth factor4 superfamily expressed early in embryogenesis. J. Biol. Chem., 267, 25220-25227.

Parfitt, A. M. (1983) Stereologic basis of bone histomorphometry; theory of quantitative microscopy and reconstruction of the third dimension. In Bone histomorphometry: techniques and interpretation. R. R. Recker, editor. CRC Press, Boca Raton,

Parfitt, A. M., Drezner, M. K., Glorieux, F. H., Kanis, J. A,, Malluche, H., Meunier, P. J., Ott, S. M. and Recker, R. R. (1 987) Bone histomorphometry: Standardization of nomencla- ture, symbols, and units. J. Bone Miner. Rex , 2, 595-610.

Reddi, A. H. and Huggins, C. B. (1972) Biochemical sequences in the transformation of normal fibroblasts in adolescent rats. Proc. Natl. Acad. Sci. USA., 69, 1601-1605.

Reddi, A. H. (1994) Bone and cartilage differentiation. Curr. Opin. Genet. Dev., 4, 737-744.

Reddi, A. H. (1997) Bone morphogenetic proteins: An uncon- ventional approach to isolation of first mammalian morphogens. Cytokine Growth Factor Rev., 8, 11 -20.

Ripamonti, U. (1991) Bone induction in nonhuman primates. An experimental study on the baboon (Papio ursinus). Clin. Orthop., 269, 284-294.

Ripamonti, U., Magan, A., van den Heever, B., Moehl, T. and Reddi, A. H. (1991) Xenogeneic osteogenin, a bone morphogenetic protein, and demineralized bone matrices, including human, induce bone differentiation in athymic rats and baboons. Matrix, II, 404-41 1.

Ripamonti, U., Ma, S., Cunningham, N., Yeates, L. and Reddi, A. H. (1992) Initiation of bone regeneration in adult baboons by osteogenin, a bone morphogenetic protein. Matrix., 12,

Ripamonti, U., van den Heever, B., Sampath, T. K., Tucker, M. M., Rueger, D. C. and Reddi, A. H. (1996a) Complete regeneration of bone in the baboon by recombinant human osteogenic protein- 1 (hop- 1, bone morphogenetic protein-7). Growth Factors, 13, 273-289.

Ripamonti, U., Bosch, C., van den Heever, B., Duneas, N., Mel- sen, B. and Ebner, R. (1996b) Limited chondro-osteogenesis by recombinant human transforming growth factor-p1 in calvarial defects of adult baboons (Papio ursinus). J. Bone Miner. Rex, 11, 938-945.

Ripamonti, U., Duneas, N., van den Heever, B., Bosch, C. and Crooks, J. (1997) Recombinant transforming growth factor- B1 induces endochondral bone in the baboon and synergizes with recombinant osteogenic protein- 1 (bone morphogenetic protein-7) to initiate rapid bone formation. J . Bone Miner. Res., 12, 1584-1595.

pp. 53-87.

369-380.

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

BONE INDUCTION BY TGF-B1 AND OP-l 277

Roberts, A. B., Sporn, M. B., Assoian, R. K., Smith, J. M., Roche, N. S., Wakefield, L. M., Heine, U. I., Liotta, L. A,, Falanga, V., Kehrl, J. H. and Fauci, A. S. (1986) Transforming growth factor type B: Rapid induction of fibrosis and angiogenesis in vivo and stimulation of collagen formation in vitro. Pi-oc. Natl. Acad. Sci. USA., 83,4167-4171.

Robey, P. G., Young, M. F., Flanders, K. C., Roche, N. S., Kon- daiah, P., Reddi, A. H., Termine, J. D., Sporn, M. B. and Roberts, A. B. (1987) Osteoblasts synthetize and respond to TGF-B in vitro. J . Cell Biol., 105, 457-463.

Rosenzweig, B. L., lmanura, T.,Okadome, T.,Cox, G. N., Yama- shita, H., ten Dijke, P., Heldin, C.-H. and Miyazono, K. (1995) Cloning and characterization of a human type 11 receptor for bone morphogenetic proteins. Pruc. Nurl. Acud. Sci. USA., 92,

Sampath, T. K. and Reddi, A. H. (1981) Dissociative extraction and reconstitution of extracellular matrix components involved in local bone differentiation. Proc. Nafl. Acad. Sci. USA., 78,

Sampath. T. K.. Muthukumaran, N. and Reddi, A. H. (1987) Isolation of osteogenin, an extracellular matrix-associated bone-inductive protein, by heparin affinity chromatography. Proc. Natl. Acad. Sci. USA., 84, 7109-71 13.

Sampath, T. K.. Maliakal, J. C., Hauschka, P. V., Jones, W. K., Sasak, H., Tucker, R. F., White, H. K., Coughlin, J. E., Tuc- ker, M. M., Pang, R. H. L., Corbett, C., Ozkaynak, E., Opper- mann, H. and Rueger, D. C. (1992) Recombinant human osteogenic protein- 1 (hOP-1) induces new bone formation in vivo with a specific activity comparable with natural bovine osteogenic protein and stimulates osteoblast proliferation and differentiation in vitro. J. Biol. Chem., 267,

Sampath, T. K.. Rashka, K. E., Doctor, J. S., Tucker, R. F. and Hoffmann, F. M. (1 993) Drosophila TGF-I3 superfamily proteins induce endochondral bone formation in mammals. Proc. Nutl. Acad. Sci. USA., 90, 6004-6008.

Schnitzler, C. M., Ripamonti, U. and Mesquita, J. M. (1993) Histomorphometry of iliac crest trabecular bone in adult male baboons in captivity. Calcif Tiss. Z i z f . , 52, 447-454.

7632-7636.

7599- 7603.

20352-20362.

Statistical Analysis System. (1 989) SAYSTATS User’s Guide, Version 6,4th ed, Sas Institute Inc., Cary, N.C., 2, pp 891 -996.

Sumner, D. R., Turner, T. M., Purchio, A. F., Gombotz, W. R., Urban, R. M. and Galante, J. 0. (1995) Enhancement of bone ingrowth by transforming growth factor-p. J, Bone Joint Surg.,

ten Dijke, P., Yamashita, H., Sampath, T. K., Reddi, A. H., Estevez, M., Riddle, D. L., Ichijo, H., Heldin, C.-H. and Miyazono, K. (1994) Identification of type I receptors for osteogenic protein-] and bone morphogenetic protein-4. J . Biol. Chem., 269, 16985- 16988.

Urist, M. R. (1965) Bone: Formation by autoinduction. Science

Urist, M. R., DeLange, R. J. and Finerman, G. A. (1983) Bone cell differentiation and growth factors. Science (Wash. DC).,

Vukicevic, S., Latin, V., Chen, P., Batorsky, R., Reddi, A. H. and Sampath, T. K. (1994a) Localization of osteogenic protein- I , a bone morphogenetic protein, during human embryonic development. Biochem. Biophys. Rex Commun., 198, 693-700.

Vukicevic, S., Helder, M. N. and Luyten, F. P. (1994b) Devel- oping human lung and kidney are major sites for synthesis of bone morphogenetic protein-3 (osteogenin). J. Histochem. Cytochem., 42, 869-875.

Wang, E. A,, Rosen, V., D’Alessandro, J. S., Bauduy, M., Cor- des, P., Harada, T., Israel, D. I., Hewick, R. M., Kerns, K. M., LaPan, P., Luxenbeg, D. P., McQuaid, D., Moutsatsos, I. K., Nove, J. and Wozney, J . M. (1990) Recombinant human bone morphogenetic protein induces bone formation. Pruc. Nutl.

Wozney, J. M., Rosen, V., Celeste, A. J., Mitsock, L. M., Whit- ten, M. J., Kriz, R. W., Hewick, R. M. and Wang, E. A. (1988) Novel regulators of bone formation: Molecular clones and activities. Science (Wash. DC)., 242, 1528- 1534.

Wozney, J. M. (1992) The bone morphogenetic protein family and osteogenesis. Molec. Reprod. Dev., 32, 160- 167.

77-A, 1135-1147.

( W U S ~ . DC)., 159, 893-899.

220, 680-686.

A c u ~ . Sci. USA., 87, 2220-2224.

Gro

wth

Fac

tors

Dow

nloa

ded

from

info

rmah

ealth

care

.com

by

CD

L-U

C B

erke

ley

on 1

0/28

/14

For

pers

onal

use

onl

y.

Transforming growth factor- β 1: Induction of bone morphogenetic protein genes expression during...

Documents

Transcript of Transforming growth factor- β 1: Induction of bone morphogenetic protein genes expression during...