Transforming growth factor-beta 3(Tgf-╬▓3) in a collagen gel delays fusion of the rat posterior interfrontal suture in vivo.код для вставкиСкачать
THE ANATOMICAL RECORD 267:120 –130 (2002) Transforming Growth Factor-Beta 3 (Tgf-␤3) in a Collagen Gel Delays Fusion of the Rat Posterior Interfrontal Suture In Vivo LYNNE A. OPPERMAN,1,2* AMR M. MOURSI,3 JENNIFER R. SAYNE,1 AND ANA MARIA WINTERGERST1 1 Department of Biomedical Sciences, Baylor College of Dentistry, Texas A&M University System Health Science Center, Dallas, Texas 2 Center for Craniofacial Research and Diagnosis, Baylor College of Dentistry, Texas A&M University System Health Science Center, Dallas, Texas 3 Department of Pediatric Dentistry, College of Dentistry, Ohio State University, Columbus, Ohio ABSTRACT Postnatal expansion of the intramembranous bones of the craniofacial skeleton occurs as bone growth at sutures. Loss of the bone growth site occurs when the suture fails to form, or when the newly formed sutures become ossified, resulting in premature obliteration. Previous experiments demonstrated that removal of dura mater from fetal rat coronal sutures, or neutralizing transforming growth factor-beta 2 (Tgf-␤2) activity using antibodies resulted in premature obliteration of the suture in vitro. Conversely, addition of Tgf-␤3 to coronal sutures in vitro rescued them from osseous obliteration. To examine whether Tgf-␤3 rescues sutures from obliteration in vivo, a collagen gel was used as a vehicle to deliver Tgf-␤3 to the normally fusing rat posterior interfrontal (IF) suture. Surgery was done on postnatal day 9 (P9) rats, in which collagen gels containing 0, 3, or 30 ng Tgf-␤3 were placed above the IF suture, underneath the periosteum for 2 weeks. By P24, 75–100% of animals in control unoperated, sham-operated, and collagen gel-only groups had fused IF sutures. In contrast, 40% of sutures exposed to 3 ng Tgf-␤3 remained open, while sutures exposed to 30 ng Tgf-␤ were similar to controls. By immunohistochemistry, sutures rescued from obliteration by Tgf-␤3 had the same Tgf-␤ receptor type II (T␤r-II) distribution as controls. However, Tgf-␤3-treated sutures had altered Tgf-␤2 and T␤r-I distribution compared to controls. Anat Rec 267:120 –130, 2002. © 2002 Wiley-Liss, Inc. Key words: cranial suture morphogenesis; rat; collagen gel; Tgf-␤3; craniosynostosis Development and growth of the calvaria is closely coordinated with the development and growth of the underlying brain (Kreiborg, 2000). The major sites of bone growth in the calvaria are the sutures, which form when the expanding bones of the craniofacial skeleton either butt up against one another or overlap (Kokich, 1986). Failure of sutures to form during embryonic development, or failure of newly formed sutures to remain unossified, results in premature obliteration of sutures, termed craniosynostosis. This leads to loss of the bone growth site, often resulting in severe facial dysmorphology (Enlow, 1986, 1989, 2000). Several genetic mutations have been linked to craniosynostosis, including mutations in genes for fibroblast growth factor receptors (FGFRs)1 (Jabs et al., 1994; Muenke et al., 1994, 1997; Reardon et al., 1994; Rutland et al., 1995; Wilkie et al., 1995; Bellus et al., 1996; ), MSX2 (Jabs et al., 1993), and TWIST (el Ghouzzi et al., 1997, 1999). There are no known mutations of transforming growth factor betas (TGF-␤s) or their receptors (T␤R-I and T␤R-II) that are associated with craniosynostosis. How© 2002 WILEY-LISS, INC. Portions of this work were presented in abstract form at the 30th Annual Meeting of the American Association of Dental Research, Chicago, 2001. Grant sponsor: NIH/NIDCR; Grant number: DE11978. *Correspondence to: Lynne A. Opperman, Ph.D., Department of Biomedical Sciences, Baylor College of Dentistry, Texas A&M University System Health Science Center, P.O. Box 660677, Dallas, TX 75266-0677. Fax: (214) 828-8951. E-mail: firstname.lastname@example.org Received 12 December 2001; Accepted 5 March 2002 DOI 10.1002/ar.10094 Published online in Wiley InterScience (www.interscience.wiley.com). 1 In line with new nomenclature, all abbreviations of genes will be italicized, while all abbreviations of proteins will be regular type. All abbreviations of human genes and proteins will be in upper case, while abbreviations of animal genes and proteins will have the first letter in upper case and all subsequent letters in lower case. TGF-␤3 REGULATION OF SUTURE CLOSURE IN VIVO ever, mutations in the TGF-␤-related NOGGIN are associated with proximal synphalangism and multiple synostoses syndrome, while mutations in the TGF-␤-related CDMP1 are associated with Hunter Thompson-type acromesomelic chondrodysplasia and Grebe-type chondrodysplasia (see Massague et al., 2000, for review). More importantly, mutations in the Smad transcriptional corepressor TGIF have been found to result in holoprosencephaly (Muenke and Beachy, 2000). In vitro data have shown that Tgf-␤2 and Tgf-␤3 are potent regulators of suture patency (Opperman et al., 1998, 1999, 2000). Tgf-␤2 added to fetal rat calvaria in culture will induce the normally patent coronal sutures to fuse, while Tgf-␤3 will rescue these sutures from fusion in vitro. Because of this antagonistic behavior, it is possible that the expression of one of these proteins directly affects the expression or presence of the other. Alternately, these factors could antagonistically affect expression of other factors or their receptors involved in regulating suture patency. If so, the relationship between Tgf-␤2 and Tgf-␤3 in regulating suture patency is of pathological and therapeutic importance. Manipulation of either Tgf-␤2 or Tgf-␤3 could be used to regulate the activity of the other, or regulate expression of other growth factors or their receptors whose activity is affected by genetic mutations. Prenatal diagnosis of craniosynostosis is difficult, and intervention is not currently feasible. To correct the synostosis and preclude the deformational changes that occur due to craniosynostosis, early postnatal surgical intervention is currently used (Marchac and Renier, 1982; Marsh and Vannier, 1985; Persing et al., 1981, 1989; Jane and Persing, 1986; Posnick, 1992, 1996a,b). A certain amount of success has been achieved from the use of this technique; however, calvariae often show rapid bone growth within the suture, leading to resynostosis. This necessitates further surgical intervention (Moss, 1959; Marchac and Renier, 1982; Marsh and Vannier, 1985; Jane and Persing, 1986; Ousterhout and Vargervik, 1987; Persing et al., 1989; Hassler and Zentner, 1990; Fatah et al., 1992; Drake et al., 1993). To avoid the problem of reossification, many techniques have been employed, with varying success, to reduce the osteogenic potential of the calvaria (Persing et al., 1989; Cohen, 1993; Salyer, 1999). However, to date no soluble growth factors have been used to inhibit reossification of the suture site in vivo. To test the hypothesis that Tgf-␤3 plays a role in regulating suture patency in vivo, a well characterized model utilizing the normally fusing posterior interfrontal (IF) suture of rats was used to examine the ability of Tgf-␤3 to regulate suture patency in vivo (Moss, 1960; Opperman et al., 1997; Roth et al., 1997). A degradable collagen gel was used as a vehicle for transporting and localizing Tgf-␤3 above IF sutures in young rats prior to the onset of suture fusion. When the rats reached the age at which fusion normally occurs, the sutures were examined for fusion. Using immunohistochemical techniques, Tgf-␤2, T␤r-I, and T␤r-II expression in Tgf-␤3-rescued IF sutures was compared to their expression in fusing IF sutures and nonfusing sagittal sutures. This was done to test the hypothesis that Tgf-␤3 rescues sutures by regulating expression of the receptors used by both Tgf-␤2 and Tgf-␤3, rather than regulating Tgf-␤2 protein expression in sutures. 121 MATERIALS AND METHODS Preparation of Collagen Gels A highly purified, bovine collagen type I gel prepared as described by Rosenblatt et al. (1989), and generously provided by NeuColl Inc. (Campbell, CA), was used as a carrier for keeping Tgf-␤3 localized above the suture region. Under sterile conditions, 100-l aliquots of the gel (65 mg/ml) were extruded from a 1-ml syringe. Each aliquot was mixed with Tgf-␤3 to a final concentration of 0, 9, or 90 ng per gel aliquot. The gel was then molded into a 3-mm-long cylinder and allowed to stand overnight in an incubator at 37°C with 5% CO2 in air. Prior to insertion into the animals, the gels were cut into 1-mm segments, such that each segment contained 0, 3, or 30 ng Tgf-␤3. One 1-mm-long segment was sufficient to cover the IF suture of each animal. Animals and Surgery Thirty-three Sprague-Dawley rats were used in this study. The protocol for surgery and postsurgical observation and care followed the guidelines dictated by the Baylor College of Dentistry IACUC. All animals for this study were housed in the Animal Research Unit at the Baylor College of Dentistry. The results of two separate experiments were combined for data analysis. Nine days after birth (P9), rat pups were anesthetized using 3.25% isofluorane (Butler, Dallas, TX) and 900 cc/min O2 inhalant anesthetic delivered by a Foregger F500 anesthesia system via a nose cone. A coronal incision through the scalp was made from ear to ear, the skin was reflected to expose the coronal and IF sutures, and the periosteum was loosened above the IF suture by use of a periosteal elevator. Animals were randomly assigned to four groups. Group 1 was used as a sham surgical control. Animals in this group had no further manipulation prior to closing the incision. Group 2 had collagen gels with sterile phosphate-buffered saline (PBS) placed on the exposed IF suture, beneath the periosteum. Group 3 received collagen gels with 3 ng Tgf␤3, and group 4 received collagen gels with 30 ng Tgf-␤3. The two edges of the scalp were reapproximated and closed using 6-0 silk suture (DG, New York, NY), without damage or movement of the gel. Fifteen days after surgery (P24), all rats from the four groups were killed by Halothane (Butler) overdose. Group 5 comprised pups killed at P24, which served as nonsurgical controls. All animals were decapitated and the scalps were removed. The heads were fixed overnight in 4% paraformaldehyde and decalcified for 7 days with 0.5 M EDTA. Following decalcification, the heads were cut along the coronal suture, separating each into cranial regions containing frontal bones with intervening IF sutures and parietal bones with intervening sagittal sutures. Samples were post-fixed overnight in 4% paraformaldehyde and processed for paraffin embedding and sectioning. Finally, 6-m sections through tissues containing either IF or sagittal sutures were cut and stained with hematoxylin and eosin (H&E) or used for immunohistochemical analysis. Immunohistochemistry An indirect immunoperoxidase procedure was carried out as described previously (Opperman et al., 1997). Briefly, endogenous peroxidase activity was removed by pre-incubation in 3% hydrogen peroxide in methanol. PBS containing 2% normal rabbit serum (Sigma, St. Louis, 122 OPPERMAN ET AL. TABLE 1. Results of histological analysis of interfrontal suture patency Sagittal Interfrontal Group Treatment Open Fused Open Fused Total sample (n) 1 2 3 4 5 Sham surgical control Collagen gel only Gel with 3ng TGF-␤3 Gel with 30ng TGF-␤3 Non-surgical control 7 5 4 6 8 1 1 1 0 0 2 1 3 2 0 6 5 2 4 8 8 6 5 6 8 MO) was used to block nonspecific binding of secondary antibody. Primary polyclonal antibodies to Tgf-␤2 (R & D Systems, Minneapolis, MN), T␤r-I, or T␤r-II (Santa Cruz Biotechnologies, Santa Cruz, CA), company-tested for specificity by western blotting procedures, were used at a 1:250 concentration. Control slides were incubated in the absence of primary antibody. Secondary antibodies were peroxidase-conjugated rabbit anti-goat IgG (1:250; Jackson Immunoresearch Laboratories, West Grove, PA) with diaminobenzidine kit (Vector Laboratories, Burlingame, CA) as chromagen. Scoring and Analysis of Tissues Depending on the suitability of the tissue for scoring, between 24 and 36 sections through the mid-section of the sutures were examined. Sections through sutures were scored as open or fused by two observers blinded to tissue grouping. Suture patency was scored as fused or open according to the presence or absence of bony bridges across the suture matrix. Sutures were designated as open if less than 10% of sections had bony bridging, while sutures with greater than 10% of sections with bony bridges across sutures were scored as fused. Sagittal sutures, which remain open throughout the life of the animal, were used as control sutures. Tgf-␤2 immunoreactivity in sagittal sutures was scored as absent, light, or dark. Immunoreactivity in IF suture tissues was scored as either lighter or darker than Tgf-␤2 immunoreactivity in sagittal sutures. T␤r-I or T␤r-II immunoreactivity was scored as present in most (⬎70%) cells, or present in few (⬍30%) cells within sutural tissues. No tissues had populations of cells with a mixture of immunoreactive and nonimmunoreactive cells in the range of 30 –70% for either T␤r-I or T␤r-II immunoreactivity. RESULTS Effect of Tgf-␤3 on IF Suture Patency In Vivo IF and sagittal sutures were examined. Sagittal and IF sutures are butt sutures, with bone fronts that overlap slightly but never become as interdigitated or complex as the overlapping coronal and lambdoid sutures. Sagittal sutures remain patent throughout the life of the rat, and in all nonsurgical control animals the sagittal suture remained patent (Table 1). The suture was visible as fibrous, cellular tissue between the parietal bones (Fig. 1A and B). One animal in each of groups 1, 2, and 3 had sagittal sutures in which more than 10% of sections through the suture were scored as fused. A bony bridge between the parietal bones was seen spanning the suture (Fig. 1C and D). The remaining animals in each group had open sagittal sutures, similar to controls shown in Figure 1A and B. All sagittal sutures in group 4 animals were scored as open, similar to unoperated controls (not shown). All IF sutures in group 5 animals were fused (Table 1). The bone fronts on either side of the IF suture appeared thickened compared to the bone further away from the suture region (Fig. 2A and B). Suture fusion appeared as a complete bridging of the frontal bones across the dural surface, with residual suture matrix visible above the fused bones. In control groups 1 and 2, the majority of animals (67% and 80%, respectively) had bony bridges across the suture (Fig. 2C and D) or showed typical fusion across the dural surface, similar to that seen in nonsurgical controls. In group 3, 60% of the animals had sutures that were rescued from obliteration (Table 1). The rescued sutures appeared widely patent, with the bone fronts separated by a fibrous, cellular matrix similar to that seen in normal sagittal sutures (Fig. 3A and D). Those sutures that were not rescued varied from partial obliteration of the suture by bony bridging (Fig. 3B and E) to complete fusion, as in untreated controls (not shown). The majority of sutures in group 4 were not rescued from fusion (Table 1). In those that did remain patent, the bone fronts were thicker than bones in untreated sutures, and the suture was markedly narrowed (Fig. 3C and F). The small sample size precluded a statistical analysis of the data. Effect of Tgf-␤3 on Tgf-␤2 Immunoreactivity in IF Sutures Sagittal sutures, used as normal nonfusing controls, showed little to no immunoreactivity for Tgf-␤2 within the cells or matrix of the suture proper, but showed intense levels of Tgf-␤2 immunoreactivity in osteoblasts lining the bone fronts on either side of the suture (Fig. 4A and E). Similarly, fusing IF sutures showed intense levels of Tgf-␤2 immunoreactivity within the osteoblasts lining the bone fronts of the sutures, and in the cells and matrix of the surrounding periosteum and dura mater (Fig. 4B and F). In contrast to sagittal sutures, the residual IF suture matrices and cells were highly immunoreactive for Tgf-␤2. Sutures rescued from obliteration by Tgf-␤3 were found to have strong immunoreactivity for Tgf-␤2 in the osteoblasts lining the bone fronts, similar to that seen in both fusing IF sutures and patent sagittal sutures (Fig. 4C and G). However, immunoreactivity within the matrix and cells of sutures rescued from obliteration by Tgf-␤3 was low to absent, similar to that seen in nonfusing sagittal sutures, and in contrast to that seen in fusing IF sutures. Sections used as immunohistochemical controls and that were not exposed to primary antibodies showed no immunoreactivity in any tissues (Fig. 4D and H). TGF-␤3 REGULATION OF SUTURE CLOSURE IN VIVO 123 Fig. 1. Micrographs of H&E-stained sections through the sagittal sutures of P24 rats. A: Low-power image showing parietal bones (pb) on either side of the fibrous suture (s) are thicker close up to the suture and appear narrower distal to the suture site. B: High-power image of the suture region. C: Low-power image through the fused sagittal suture, showing greatly thickened bones. D: High-power image showing residual suture (asterisk) and fusion of parietal bones across the suture (arrow). d, dura mater, p, periosteum. Original magnification: (A and C) 20⫻, bar ⫽ 30 m; (B and D) 40⫻, bar ⫽ 60 m. Fig. 2. Micrographs of H&E-stained sections through IF sutures of P24 rats. A and B: Low- and high-power images of a nonsurgical control suture, showing residual suture matrix (asterisk) above the fused frontal bones (arrow). C and D: Low- and high-power images of a fused suture in a gel-only treated animal, showing a bony bridge through the center of the suture (arrow). Original magnification: (A and C) 10⫻, bar ⫽ 15 m; (B, and D) 20⫻, bar ⫽ 30 m. Effect of Tgf-␤3 on T␤r-II Immunoreactivity in IF Sutures surfaces of the bone away from the suture site. In contrast, periosteal fibroblasts were nonreactive for T␤r-II in all groups. In nonfusing sagittal sutures, most of the osteoblastic cells lining the bone fronts on either side of the In all groups, T␤r-II immunoreactivity was present in most of the osteoblasts lining the dural and periosteal 124 OPPERMAN ET AL. Fig. 3. Micrographs of H&E-stained sections through IF sutures of P24 rats. A and D: Low- and high-power images through a suture treated with 3 ng Tgf-␤3. The suture (s) remains widely patent between frontal bones (fb). The suture can be seen to be closely associated with and indistinguishable from the underlying dura mater (d) and overlying periosteum (p). B and E: Low- and high-power images of a suture treated with 3 ng Tgf-␤3, in which the suture fails to remain patent. Small bony bridges occur across the dural surface of the suture (arrow), while the remainder of the suture remains widely patent. C and F: Low- and high-power images through a suture treated with 30 ng Tgf-␤3. Highly thickened frontal bones (fb) are separated by a narrow but patent suture (s). Original magnification: (A–C) 10⫻, bar ⫽ 15 m; (D–F) 20⫻, bar ⫽ 30 m. suture and most of the fibroblastic cells within the suture matrix were immunoreactive for T␤r-II (Fig. 5A and D). Similarly, in both fusing IF sutures (Fig. 5B and E) and IF sutures rescued from obliteration by Tgf-␤3 (Fig. 5C and F), T␤r-II immunoreactivity was noted in most of the osteoblastic cells lining the bone fronts, and in most fibroblastic cells within the matrix of the sutures. Effect of Tgf-␤3 on Intensity of T␤r-I Immunoreactivity in IF Sutures Similar to the immunoreactivity seen for T␤r-II, most of the osteoblasts lining the dural and periosteal surfaces of the bone, as well as those lining the bone fronts on either side of the suture, were immunoreactive for T␤r-I in all groups. Fibroblasts within the periosteum above the su- TGF-␤3 REGULATION OF SUTURE CLOSURE IN VIVO Fig. 4. Micrographs of immunoreactivity for Tgf-␤2 in P24 rat (A and E) sagittal and (B–H) IF sutures. A and E: Low- and high-power micrographs of the nonfusing sagittal suture. Tgf-␤2 immunopositive cells can be seen lining the bone fronts on either side of the suture (large arrows). Most of the suture cells within the suture matrix are negative for Tgf-␤2 immunoreactivity (small arrows). B and F: Low- and high-power micrographs of normally fusing IF sutures. Osteoblastic cells lining the bone fronts are immunoreactive for Tgf-␤2 (large arrows), as are cells within the sutures (small arrows), along with the suture matrix. C and G: Low- 125 and high-power micrographs of an IF suture rescued from obliteration by Tgf-␤2. Immunoreactive osteoblasts can be seen lining the bone fronts (large arrows), with nonreactive cells within the suture matrix (small arrows). D and H: Low- and high-power micrographs of a sagittal suture used as a negative control. Note complete absence of immunoreactivity in osteoblasts (large arrows) and suture cells (small arrows), as well as in the matrix. Original magnification: (A–D) 10⫻, bar ⫽ 15 m; (E–H) 80⫻, bar ⫽ 120 m. 126 OPPERMAN ET AL. Fig. 5. Micrographs of immunoreactivity for T␤r-II in sections through the (A and D) sagittal and (B–F) IF sutures of P24 rats. A and D: Low- and high-power micrographs of sections through sagittal sutures. Darkly immunoreactive osteoblasts can be seen lining the bone fronts on either side of the patent sagittal suture (large arrows). Most of the fibroblastic cells within the suture matrix are immunoreactive for T␤r-II (small arrows). B and E: Low- and high-power micrographs through fusing IF sutures. Frontal bones can be seen to be fused underneath the residual suture. Osteoblasts lining the residual suture are highly immunoreactive for T␤r-II (large arrows), as are fibroblastic cells within the suture matrix (small arrows). C and F: Low- and high-power micrographs of sutures rescued from obliteration by 3 ng Tgf-␤3. Note immunoreactivity for T␤r-II in the osteoblasts lining the bone (large arrows) and in the fibroblastic cells within the suture matrix (small arrows). Original magnification: (A–C) 30⫻, bar ⫽ 45 m; (D–F) 80⫻, bar ⫽ 120 m. Fig. 6. Micrographs of immunoreactivity for T␤r-I in sections through the (A and D) sagittal and (B–F) IF sutures of P24 rats. A and D: Low- and high-power micrographs of sections through sagittal sutures. Immunoreactive osteoblasts (large arrows) can be seen around the leading edges of bone fronts on either side of the suture. Suture fibroblasts (small arrows) are not immunoreactive. B and E: Low- and high-power micrographs of sections through fusing IF sutures. Note immunoreactive osteoblasts (large arrows) lining the bone fronts, and immunoreactive suture fibroblasts (small arrows) within the residual suture matrix. C and F: Low- and high-power micrographs of sections through IF sutures rescued from fusion by 3 ng Tgf-␤3. Note a distribution of osteoblasts (large arrows) similar to those seen in nonfusing sagittal sutures (described in A and D). Within the suture matrix, note suture fibroblasts (small arrows) that lack T␤r-I immunoreactivity. Original magnification: (A–C) 30⫻, bar ⫽ 45 m; (D–F) 80⫻, bar ⫽ 120 m. tures were negative for T␤r-I immunoreactivity in all groups, similar to that seen for T␤r-II immunoreactivity. Most suture fibroblasts within the matrix of nonfusing sagittal sutures were negative for T␤r-I immunoreactivity (Fig. 6A and D). In contrast, all fibroblastic cells within the residual matrix of fusing IF sutures were immunoreactive for T␤r-I (Fig. 6B and E). Unlike fibroblasts within fusing IF sutures, but similar to those in nonfusing sagit- TGF-␤3 REGULATION OF SUTURE CLOSURE IN VIVO tal sutures, fibroblasts in IF sutures rescued from obliteration by Tgf-␤3 were not immunoreactive for T␤r-I (Fig. 6C and F). DISCUSSION In vitro, Bmp2, Bmp4 (Kim et al., 1998), and Tgf-␤2 (Opperman et al., 1997) have been shown to induce suture fusion. Furthermore, Tgf-␤3 (Opperman et al., 2000) and neutralizing antibodies to Tgf-␤2 (Opperman et al., 1999) rescued sutures from fusion in vitro. In vivo, Tgf-␤2 (Roth et al., 1997), and Fgf2 (Iseki et al., 1999) have both been reported to induce or accelerate suture fusion. To date, however, no soluble factors have been shown to inhibit reossification of the suture site in vivo. The potential of using growth factors to inhibit reossification of the suture in conjunction with current initial surgery, or, preferably, with more limited and localized surgery, has great importance for reducing surgical risk and improving outcome. In the rat, all of the cranial sutures remain patent throughout the life of the animal, with the exception of the posterior region of the IF suture (Moss, 1954). This region of the suture undergoes osseous obliteration early, becoming completely obliterated by P21 (Moss, 1960). This pattern of early postnatal fusion is similar to that found in the metopic suture in humans, which becomes obliterated by 2–3 years after birth (Cohen, 2000). The rat IF suture, as a model of naturally occurring suture fusion, was used to test whether Tgf-␤3 can prevent or delay suture fusion. A low dose of Tgf-␤3 rescued IF sutures from obliteration in vivo, although not all IF sutures in animals from the low-dose group were rescued. A high dose of Tgf-␤3 failed to rescue sutures from obliteration, similar to vehicle alone or sham operation. These findings are similar to those reported in vitro (Opperman et al., 2000). Tgf-␤3 is known to be a biphasic regulator of osteoblast activity, with lower concentrations of Tgf-␤3 stimulating cell proliferation, while higher concentrations inhibit cell proliferation and stimulate production of extracellular matrix and markers of cellular differentiation (ten Dijke et al., 1990). It is therefore possible that the higher dose of Tgf-␤3, which did not rescue sutures from obliteration, instead stimulated osteogenesis, resulting in suture obliteration. This is not easy to demonstrate, as these sutures are destined to fuse, so markers of osteoblast differentiation would be present both in the absence of Tgf-␤3 and in the presence of the high dose of Tgf-␤3. One way to address this problem would be to place gels containing low and high doses of Tgf-␤3 above the sagittal suture, which remains patent. If the high dose of Tgf-␤3 induced fusion of the sagittal sutures, then it would be interesting to look for markers of osteoblast differentiation within these fusing sutures. Some fusion of sagittal sutures from animals undergoing sham and vehicle-only procedures was noted, suggesting that the surgical procedure may result in sporadic bony bridging within the suture. In support of this idea, it has been noted anecdotally in clinical cases (Genecov, personal communication) that intermittent bony bridging within sutures can occur in association with surgical procedures. Alternately, since collagen gels were used as a vehicle, it is possible that most of the Tgf-␤3 was released from the gels early and hence the rescue effect of Tgf-␤3 127 disappeared, with the sutures reentering the fusion process. However, since most of the sutures remained unfused, this is unlikely. Various types of collagen biomaterials have been tested as drug delivery vehicles (Friess, 1998), and delivery of growth factors using this particular gel has been evaluated (Schroeder-Tefft et al., 1997; Bentz et al., 1998). In a mink lung cell bioassay, Tgf-␤2 admixed with gel retained a level of activity similar to that of Tgf-␤2 alone (Bentz et al., 1998; Schroeder-Tefft et al., 1997). In vivo, Tgf-␤2 in collagen gel induced a typical subcutaneous fibroblastic response, indicating that the growth factor retained its biological activity in the gel (Schroeder-Tefft et al., 1997). Although highly purified, this formulation retains the native fibrillar conformation of collagen, as opposed to other commercially available degraded collagen gels. The antigenic N- and C-terminal domains have been removed, greatly reducing the likelihood of an immune response (DeLustro et al., 1986; Ellingsworth et al., 1986). There are a variety of clinical applications for this collagen gel. It is most widely used as an intradermal injection for soft-tissue augmentation, in which it resorbs in approximately 6 months to a year (Pachence, 1996). Its retention at other sites and in other species is not well known. However, studies by Moursi and coworkers (unpublished data) indicated that biotinylated collagen was visualized histologically up to 63 days postimplantation in vivo at cranial sutures in rabbits. There was no evidence of an inflammatory response, immune reaction, or fibrosis. In addition, the injectable material is semiviscous at room temperature and gels at 37°C, making it ideal for placement into surgical sites. Together these studies support the use of this collagen gel as a vehicle for the delivery of growth factors to cranial sutures. Before growth factors can be used therapeutically for rescuing sutures from obliteration, it is important to understand the mechanism by which growth factors regulate suture patency. Tgf-␤3 reduced Tgf-␤2 immunoreactivity in matrix and cells of IF sutures when compared to untreated IF sutures, with immunoreactivity appearing more similar to that of nonfusing sagittal sutures. This finding is different from the effect seen in vitro, wherein Tgf-␤2 immunoreactivity was unchanged in response to Tgf-␤3 treatment, remaining as intense in the suture matrix of Tgf-␤3-rescued sutures as it was in fusing sutures (Opperman et al., 2000). The difference between the in vivo and the in vitro effects could be related to the fact that the tissue culture environment is different from that found in vivo. However, this difference demonstrates that Tgf-␤3-induced changes in Tgf-␤2 immunoreactivity are probably not primarily responsible for the ability of Tgf-␤3 to rescue sutures from fusion in vivo. It is important in any study examining growth factor effects to determine whether receptors for the growth factors are present, and whether the cells that express them respond to the growth factors. Both T␤r-I and T␤r-II receptors are present on suture fibroblasts and osteoblasts. No change in the number of T␤r-II immunoreactive cells in sutures was found in response to Tgf-␤3 treatment. This does not preclude a change in the number of receptors per cell; however, immunoreactivity was not considered sensitive enough to gauge differences in receptor numbers per cell. In contrast, Tgf-␤3 dramatically reduced the numbers of suture fibroblasts immunoreactive for T␤r-I in IF sutures rescued from obliteration, when com- 128 OPPERMAN ET AL. pared to fusing IF sutures. This is similar to the findings of Opperman et al. (2002), who reported down-regulation of T␤r-I in IF suture fibroblasts in vitro in response to Tgf-␤3 treatment. Interestingly, Opperman et al. (2002) reported that nonfusing coronal sutures had many T␤r-Ipositive fibroblasts and osteoblasts within the suture, whereas the data presented here indicate that nonfusing sagittal sutures have few T␤r-I immunoreactive fibroblasts. This may reflect developmental differences between these sutures, which are also shown to have differences in Msx2 and Shh expression (Liu et al., 1995; Kim et al., 1998). While it has been shown that Tgf-␤2 and Tgf-␤3 have antagonistic effects in regulating suture morphogenesis (Opperman et al., 2000), they have been shown to have distinctly different sequential regulatory effects during epithelial to mesenchymal transformation in embryonic heart (Boyer et al., 1999), and differential effects on medial edge epithelium during fusion of the palatal shelves (Taya et al., 1999). It is intriguing that Tgf-␤2 and Tgf-␤3 have such opposite effects on cell function, when they share the same set of transmembrane receptors and the same intracellular Smads (Smad2, Smad3, and perhaps Smad5 (Centrella et al., 1996)). These Smads interact with the ubiquitous Smad4 after Tgf-␤ receptor stimulation and then translocate to the nucleus, where they interact with other transcription factors such as SP1 (Inagaki et al., 2001; Poncelet and Schnaper, 2001), c-Jun/cFos (Zhang et al., 1998), and Runx2 (Leboy et al., 2001) to activate Tgf-␤-inducible transcription (Massague and Wotton, 2000). Of note, it now appears that Tgf-␤ signaling may interact with the MAP kinase signaling pathways of tyrosine kinase receptors. Tgf-␤ can signal via the Ras/ Mek pathway to phosphorylate Smad1 (Yue et al., 1999), Tgf-␤1 activation of Atf-2 is dependent upon Smad4 activation of p38 (Simeone et al., 2001), and blocking the Mek1/Erk pathway inhibits Tgf-␤1-induced ␣1(I) collagen expression (Hayashida et al., 1999). Furthermore, there is now good evidence that Tgf-␤s can signal through these other pathways, independently of Smads. For example, they can activate Jnk signaling in a Rho-dependent, Smad-independent manner to activate Ap-1 (Engel et al., 1999), c-Jun and Atf-2 function (Hocevar et al., 1999; Rousse et al., 2001), and mobilization of the actin cytoskeleton (Edlund et al., in press), suggesting that alternate Tgf-␤ signaling pathways from the Smad pathway exist. This could account for differences in cell function seen in response to Tgf-␤1, Tgf-␤2, and Tgf-␤3 during suture (Opperman et al., 1999, 2000), palate (Taya et al., 1999), and heart morphogenesis (Boyer et al., 1999). While different Tgf-␤s can have distinct effects using the same receptors, it appears that different Tgf-␤ receptors can also have distinct activities in response to growth factor binding (Boyer and Runyan, 2001). This means that Tgf-␤s likely can regulate the activity of each other, either by regulating different signaling pathways or by regulating access to the Tgf-␤ receptors. It is now apparent that Tgf-␤3 can regulate suture patency in vivo, maintaining the suture in its unossified state. However, it is of critical importance to determine the long-term effects of Tgf-␤3 on suture patency and to determine whether high doses of Tgf-␤3 are deleterious to nonfusing sutures. It is also necessary to develop a better understanding of how Tgf-␤s interact with one another and their receptors to regulate developing su- tures, before these factors can be considered to be clinically useful. ACKNOWLEDGMENTS The authors thank Mr. Phillip Winnard for advice concerning the technical aspects of gel preparation, and Mr. Greg Cooper for doing the surgical procedures and tissue harvesting. We also thank Drs. J. Pascal Mbiene, Kathy K. Svoboda, and Bob Hinton, and Mr. Joseph T. Rawlins for critical reading of the manuscript, and NeuColl Inc. (Campbell, CA) for generously donating the gel used in the experiments. LITERATURE CITED Bellus GA, Gaudenz K, Zackai EH, Clarke LA, Szabo J, Francomano CA, Muenke M. 1996. Identical mutations in three different fibroblast growth factor receptor genes in autosomal dominant craniosynostosis syndromes. Nat Genet 142:174 –176. Bentz H, Schroeder JA, Estridge TD. 1998. Improved local delivery of TGF-beta2 by binding to injectable fibrillar collagen via difunctional polyethylene glycol. J Biomed Mater Res 394:539 – 548. Boyer AS, Ayerinskas II, Vincent EB, McKinney LA, Weeks DL, Runyan RB. 1999. TGFbeta2 and TGFbeta3 have separate and sequential activities during epithelial-mesenchymal cell transformation in the embryonic heart. Dev Biol 2082:530 –545. Boyer AS, Runyan RB. 2001. TGF␤ type III and TGF␤ type II receptors have distinct activities during epithelial-mesenchymal cell transformation in the embryonic heart. Dev Dynam 2214: 454 – 459. Centrella M, Ji C, Casinghino S, McCarthy TL. 1996. Rapid flux in transforming growth factor-beta receptors on bone cells. J Biol Chem 27131:18616 –18622. Cohen Jr MM. 1993. Sutural biology and the correlates of craniosynostosis. Am J Med Genet 475:581– 616. Cohen Jr MM. 2000. Suture biology. In: Cohen Jr MM, MacLean RE, editors. Craniosynostosis, diagnosis, evaluation and management. 2nd ed. New York: Oxford University Press. p 11–23. DeLustro F, Condell RA, Nguyen MA, McPherson JM. 1986. A comparative study of the biologic and immunologic response to medical devices derived from dermal collagen. J Biomed Mater Res 20:109 – 120. Drake DB, Persing JA, Berman DE, Ogle RC. 1993. Calvarial deformity regeneration following subtotal craniectomy for craniosynostosis: a case report and theoretical implications. J Craniofac Surg 42:85– 89; discussion 90. Edlund S, Landstrom M, Heldin CH, Aspenstrom P. TGF-beta-induced mobilisation of the actin cytoskeleton requires signalling by the small GTPases Cdc42 and RhoA. Mol Biol Cell (in press). el Ghouzzi V, Le Merrer M, Perrin-Schmitt F, Lajeunie E, Benit P, Renier D, Bourgeois P, Bolcato-Bellemin AL, Munnich A, Bonaventure J. 1997. Mutations of the TWIST gene in the Saethre-Chotzen syndrome. Nat Genet 151:42– 46. el Ghouzzi V, Lajeunie E, Le Merrer M, Cormier-Daire V, Renier D, Munnich A, Bonaventure J. 1999. Mutations within or upstream of the basic helix-loop-helix domain of the TWIST gene are specific to Saethre-Chotzen syndrome. Eur J Hum Genet 71:27–33. Ellingsworth LR, DeLustro F, Brennan JE, Sawamura S, McPherson J. 1986. The human immune response to reconstituted bovine collagen. J Immunol 1363:877– 882. Engel ME, McDonnell MA, Law BK, Moses HL. 1999. Interdependent SMAD and JNK signaling in transforming growth factor-beta-mediated transcription. J Biol Chem 27452:37413–37420. Enlow DH. 1986. Normal craniofacial growth. In: Cohen MMJ, editor. Craniosynostosis: diagnosis, evaluation and management. New York: Raven Press. p 131–156. Enlow DH. 1989. Normal and abnormal patterns of craniofacial growth. In: Persing JA, Edgerton MT, Jane JA, editors. Scientific foundations and surgical treatment of craniosynostosis. Baltimore: Williams and Wilkins. p 83– 86. TGF-␤3 REGULATION OF SUTURE CLOSURE IN VIVO Enlow DH. 2000. Normal craniofacial growth. In: Cohen MMJ, MacLean RE, editors. Craniosynostosis, diagnosis, evaluation and management. New York: Oxford University Press. p 35–50. Fatah M, Ermis I, Poole M, Shun-Shin G. 1992. Prevention of cranial reossification after surgical craniectomy. J Craniofac Surg 3:170–172. Friess W. 1998. Collagen— biomaterial for drug delivery. Eur J Pharmaceut Biopharmaceut 452:113–136. Hassler W, Zentner J. 1990. Radical osteoclastic, craniectomy in sagittal synostosis. Neurosurgery 27:539 –546. Hayashida T, Poncelet AC, Hubchak SC, Schnaper HW. 1999. TGFbeta1 activates MAP kinase in human mesangial cells: a possible role in collagen expression. Kidney Int 565:1710 –1720. Hocevar BA, Brown TL, Howe PH. 1999. TGF-beta induces fibronectin synthesis through a c-Jun N-terminal kinase-dependent, Smad4-independent pathway. EMBO J 185:1345–1356. Inagaki Y, Nemoto T, Nakao A, ten Dijke P, Kobayashi K, Takehara K, Greenwel P. 2001. Interaction between GC box binding factors and Smad proteins modulates cell lineage-specific alpha 2I collagen gene transcription. J Biol Chem 27619:16573–16579. Iseki S, Wilkie AO, Morriss-Kay GM. 1999. Fgfr1 and Fgfr2 have distinct differentiation- and proliferation-related roles in the developing mouse skull vault. Development 12624:5611–5620. Jabs EW, Muller U, Li X, Ma L, Luo W, Haworth IS, Klisak I, Sparkes R, Warman ML, Mulliken JB. 1993. A mutation in the homeodomain of the human MSX2 gene in a family affected with autosomal dominant craniosynostosis. Cell 753:443– 450. Jabs EW, Li X, Scott AF, Meyers G, Chen W, Eccles M, Mao JI, Charnas LR, Jackson CE, Jaye M. 1994. Jackson-Weiss and Crouzon syndromes are allelic with mutations in fibroblast growth factor receptor 2 [published erratum appears in Nat Genet 1995;94:451]. Nat Genet 83:275–279. Jane J, Persing J. 1986. Neurosurgical treatment of craniosynostosis. Craniosynostosis: diagnosis, evaluation, and management. In: Cohen MMJ, editor. New York: Raven Press. p 249 –320. Kim HJ, Rice DP, Kettunen PJ, Thesleff I. 1998. FGF-, BMP- and Shh-mediated signalling pathways in the regulation of cranial suture morphogenesis and calvarial bone development. Development 1257:1241–1251. Kokich VG. 1986. The biology of sutures. In: Cohen MMJ, editor. Craniosynostosis, diagnosis, evaluation and management. New York: Raven Press. p 81–103. Kreiborg S. 2000. Postnatal growth and development of the craniofacial complex in premature craniosynostosis. Craniosynostosis, diagnosis, evaluation and management. In: Cohen MMJ, editor. New York: Oxford University Press. p 158 –174. Leboy PS, Grasso-Knight G, D’Angelo M, Volk SW, Lian JB, Drissi H, Stein G, Adams SL. 2001. Smad-Runx interactions during chondrocyte maturation. J Bone Joint Surg 83A:S115–S122. Liu YH, Kundu R, Wu L, Luo W, Ignelzi Jr MA, Snead ML, Maxson Jr RE. 1995. Premature suture closure and ectopic cranial bone in mice expressing Msx2 transgenes in the developing skull. Proc Natl Acad Sci USA 92:6137– 6141. Marchac D, Renier D. 1982. Craniofacial surgery for craniosynostosis. Boston: Little Brown & Co. 201 p. Marsh J, Vannier M. 1985. Comprehensive care for craniofacial deformities. St. Louis: C.V. Mosby, Co. 339 p. Massague J, Blain SW, Lo RS. 2000. Tgf␤ signaling in growth control, cancer and heritable disorders. Cell 103:295–309. Massague J, Wotton D. 2000. Transcriptional control by the Tgf-B/ Smad signaling system. EMBO J 198:1745–1754. Moss ML. 1954. Growth of the calvaria in the rat. The determination of osseous morphology. Am J Anat 94:333–362. Moss ML. 1959. The pathogenesis of premature cranial synostosis in man. Acta Anat 374:351–370. Moss ML. 1960. Inhibition and stimulation of sutural fusion in the rat calvaria. Anat Rec 136:457– 467. Muenke M, Schell U, Hehr A, Robin NH, Losken HW, Schinzel A, Pulleyn LJ, Rutland P, Reardon W, Malcolm S. 1994. A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome. Nat Genet 83:269 –274. Muenke M, Gripp KW, McDonald-McGinn DM, Gaudenz K, Whitaker LA, Bartlett SP, Markowitz RI, Robin NH, Nwokoro N, Mulvihill 129 JJ, Losken HW. 1997. A unique point mutation in the fibroblast growth factor receptor 3 gene FGFR3 defines a new craniosynostosis syndrome. Am J Hum Genet 603:555–564. Muenke M, Beachy PA. 2000. Genetics of ventral forebrain development and holoprosencephaly. Curr Opin Genet Dev 10:262–269. Opperman LA, Nolen AA, Ogle RC. 1997. TGF-beta 1, TGF-beta 2, and TGF-beta 3 exhibit distinct patterns of expression during cranial suture formation and obliteration in vivo and in vitro [see comments]. J Bone Min Res 123:301–310. Opperman LA, Chhabra A, Nolen AA, Bao Y, Ogle RC. 1998. Dura mater maintains rat cranial sutures in vitro by regulating suture cell proliferation and collagen production. J Craniofac Genet Dev Biol 183:150 –158. Opperman LA, Chhabra A, Cho RW, Ogle RC. 1999. Cranial suture obliteration is induced by removal of transforming growth factor TGF-beta 3 activity and prevented by removal of TGF-beta 2 activity from fetal rat calvaria in vitro. J Craniofac Genet Dev Biol 193:164 –173. Opperman LA, Adab K, Gakunga PT. 2000. TGF-B2 and TGF-B3 regulate fetal rat cranial suture morphogenesis by regulating rates of cell proliferation and apoptosis. Dev Dyn 2192:237–247. Opperman LA, Galanis V, Williams AR, Adab K. 2002. Transforming growth factor-beta 3 (Tgf-␤3) down-regulates Tgf-␤ receptor I (T␤r-I) during rescue of cranial sutures from osseous obliteration. Orthodont Craniofac Res 5:5–16. Ousterhout D, Vargervik K. 1987. Aesthetic improvement resulting from craniofacial surgery in craniosynostosis syndromes. J CranioMax-Fac Surg 15:189 –197. Pachence JM. 1996. Collagen-based devices for soft tissue repair. J Biomed Mater Res 331:35– 40. Persing J, Babler W, Winn H, Jane JA, Rodheaver G. 1981. Age as a critical factor in the success of surgical correction of craniosynostosis. J Neurosurg 54:601– 606. Persing J, Jane JA, Edgerton MT. 1989. Surgical treatment of craniosynostosis. In: Persing JA, Edgerton MT, Jane JA, editors. Scientific foundations and surgical treatment of craniosynostosis. Baltimore: Williams & Wilkins. p 87–95. Poncelet AC, Schnaper HW. 2001. Sp1 and Smad proteins cooperate to mediate transforming growth factor-beta 1-induced alpha 2I collagen expression in human glomerular mesangial cells. J Biol Chem 27610:6983– 6892. Posnick J. 1992. Craniosynostosis: diagnosis and treatment in infancy and early childhood. In: Bell WH, editor. Modern practice in orthognathic and reconstructive surgery. Philadelphia: WB Saunders. p 1864 –1867. Posnick J. 1996a. Monobloc and facial bipartition osteotomies: a stepby-step description of the surgical technique. J Craniofac Surg 7:229 –250. Posnick J. 1996b. Unilateral coronal synostosis anterior plagiocephaly: current clinical perspectives. Ann Plast Surg 36:430 – 447. Reardon W, Winter RM, Rutland P, Pulleyn LJ, Jones BM, Malcolm S. 1994. Mutations in the fibroblast growth factor receptor 2 gene cause Crouzon syndrome. Nat Genet 81:98 –103. Rosenblatt J, Rhee W, Wallace D. 1989. The effect of collagen fiber size distribution on the release rate of proteins from collagen matrices by diffusion. J Control Rel 9:195–203. Roth DA, Longaker MT, McCarthy JG, Rosen DM, McMullen HF, Levine JP, Sung J, Gold LI. 1997. Studies in cranial suture biology. I. Increased immunoreactivity for TGF-beta isoforms beta 1, beta 2, and beta 3 during rat cranial suture fusion. J Bone Min Res 123:311–321. Rousse S, Lallemand F, Montarras D, Pincet C, Mazars A, Prunier C, Atfi A, Dubois C. 2001. Transforming growth factor-beta inhibition of insulin-like growth factor binding protein-5 synthesis in skeletal muscle cells involves a c-Jun N-terminal kinase-dependent pathway. J Biol Chem 276:46961– 46967. Rutland P, Pulleyn LJ, Reardon W, Baraitser M, Hayward R, Jones B, Malcolm S, Winter RM, Oldridge M, Slaney SF. 1995. Identical mutations in the FGFR2 gene cause both Pfeiffer and Crouzon syndrome phenotypes. Nat Genet 92:173–176. Salyer KE. 1999. Salyer and Bardach’s atlas of craniofacial and cleft surgery. Vol. I. Craniofacial surgery. New York: LippincottRaven. 415 p. 130 OPPERMAN ET AL. Schroeder-Tefft JA, Bentz H, Estridge TD. 1997. Collagen and heparin matrices for growth factor delivery. J Control Rel 49: 291–298. Simeone DM, Zhang L, Graziano K, Nicke B, Pham T, Schaefer C, Logsdon CD. 2001. Smad4 mediates activation of mitogen-activated protein kinases by TGF-beta in pancreatic acinar cells. Am J Physiol Cell Physiol 2811:C311–C319. Taya Y, O’Kane S, Ferguson MW. 1999. Pathogenesis of cleft palate in TGF-beta3 knockout mice. Development (Suppl) 12617:3869 – 3879. ten Dijke P, Iwata KK, Goddard C, Pieler C, Canalis E, McCarthy TL, Centrella M. 1990. Recombinant transforming growth factor type beta 3: biological activities and receptor-binding properties in isolated bone cells. Mol Cell Biol 109:4473– 4479. Wilkie AO, Slaney SF, Oldridge M, Poole MD, Ashworth GJ, Hockley AD, Hayward RD, David DJ, Pulleyn LJ, Rutland P. 1995. Apert syndrome results from localized mutations of FGFR2 and is allelic with Crouzon syndrome. Nat Genet 92:165–172. Yue J, Hartsough MT, Frey RS, Frielle T, Mulder KM. 1999. Crosstalk between the Smad1 and Ras/MEK signaling pathways for TGF-␤. Oncogene 1811:2033–2037. Zhang Y, Feng XH, Derynck R. 1998. Smad3 and Smad4 cooperate with c-Jun/c-Fos to mediate TGF-beta-induced transcription. Nature 394:909 –913.