Journal of Orthopaedic Reseurch 8:193-198 Raven Press, Ltd., New York 0 1990 Orthopaedic Research Society
Attachment of Epiphyseal Cartilage Cells and 17/28 Rat Osteosarcoma Osteoblasts Using Mussel Adhesive Protein *§John P. Fulkerson, ?Louis A. Norton, *Gloria Gronowicz, $.Paul Picciano, *§J. Mathieu Massicotte and *Carl W. Nissen Departments of *Orthopaedic Surgery and forthodontics, University of Connecticut School of Medicine, and jBiopolymers, Inc., Farmington, Connecticut, and §Marine Biological Laboratory, Woods Hole, Massachusetts, U.S.A.
Summary: These experiments show that mussel adhesive protein (MAP) enhances the attachment of osteoblasts and epiphyseal cartilage cells to plastic culture dishes and Vitallium. When MAP was applied to culture plate surfaces, there were two- to fivefold increases in the numbers of cells attaching compared to control surfaces (no MAP). Results were confirmed using two different cell attachment assay techniques. Osteoblast replication and culture on MAP is possible, suggesting that MAP is not toxic to cells. MAP also holds applied cells to surfaces as initially attached. Key Words: OsteoblastsAd hesive-Mu ssels-Surfaces.
Underwater adhesion is essential for survival of the common blue mussel, Myfilus edulis. This saltwater mollusc has four glands, including a phenol gland that produces a polyphenolic protein adhesive of molecular weight of 130,000. The adhesive (1 l), originally isolated and characterized by Waite in 1980, enables the blue mussel to attach itself to rocks and other surfaces underwater in order to resist tides and currents. Now that it is commercially available, it may offer important practical advantages in surgical and dental practice, particularly the ability to create adhesion in a saline environment without toxicity for cells and tissues. Although orthopaedic implant fixation is sometimes enhanced by using textured surfaces (14,13), porous metal coatings can loosen (9). Enhanced attachment of osteoblasts to metal, therefore, might permit improved implant fixation. Chondrocyte and osteoblast attachment to plastic
culture dishes is important in connective tissue research. These studies, therefore, provide an early look at mussel adhesive protein (MAP)-enhanced attachment of chondrocytes and osteoblasts to surfaces relevant to orthopaedic surgeons and connective tissue scientists. MATERIALS AND METHODS
Mussel adhesive protein (Celltak) was obtained in acid solution from Biopolymers, Inc. (Farmington, CT, U.S.A.) and applied sparsely to 35 mm Falcon plastic petri plates (30 pg/plate) and small Vitallium (Howmedica, Inc., Rutherford, NJ, U.S.A.) metal discs (7&140 pg/disc). A very small volume of Celltak was applied, 3-14 pl (10 pg/pl), using a small plastic spatula, and the adhesive was allowed to air dry in a laminar flow hood. Evaluation of Cell Attachment to MAP-Treated and Control Surfaces
Received October 13, 1988; accepted April 12, 1990. Address correspondence and reprint requests to Dr. J. P. Fulkerson at Department of Orthopaedic Surgery, University of Connecticut School of Medicine, Farmington, CT 06032, U.S.A.
Mussel adhesive protein (Celltak) was applied to one-half of all metal and plastic surfaces using the 793
794
J . P . FULKERSON ET AL.
following technique: First, each surface was covered with sterile 0.1 M sodium bicarbonate (pH 8.0). Celltak was diluted to make a 1 mg/ml solution and then mixed on the surface with buffer by pipetting several times. After 20 min, excess solution was aspirated off the surface and the surface was washed twice with sterile water. Sixteen- to 20-day-old chick embryo articular and epiphyseal cartilage was obtained from lower extremity long bones, and chondrocytes were released from matrix by incubation (37°C) of the tissue with collagenase and hyaluronidase (1 mg/ml) for 2 h using magnetic stirring pellets, after mincing the tissue thoroughly with scalpels. Heterogeneous populations of mixed epiphyseal and articular chondrocytes and populations of epiphyseal chondrocytes alone give similar results. Similar experiments were performed also with ROS 17/28 rat osteosarcoma osteoblasts (6), a well-characterized cell culture line. Cells were washed twice with F-12 medium (Gibco, Grand Island, NY, U.S.A.), suspended in F-12 medium, and counted on a hemocytometer. A known concentration of cells (usually about 2-5 x lo5 celldml) was placed over each of the MAP-treated surfaces and comparable untreated surfaces. Short incubation periods were chosen to see if rapid cell attachment would permit applications in a surgical solution. Following 5 to 15 min incubations, cell suspensions were agitated and removed from the surfaces being studied. Cells remaining in suspension were counted using a hemocytometer or Coulter counter and the percent of cells lost from suspension determined by subtracting the cell concentration remaining from that originally placed over the surface. These data were confirmed by examining all surfaces by light microscopy to correlate the numbers of cells attached to the cells lost from suspension. Cell shape was compared to controls to see if the adhesive caused any change. All surfaces were examined for evidence of MAP disruption by culture medium or agitation. Other surfaces were treated with isobutylcyanoacrylate (super glue) and examined after similar culture medium treatment and manual agitation. Selected surfaces were examined by scanning electron microscopy after careful dehydration and sputter-coating. In addition, four plastic and four Vitallium surfaces were treated with chick epiphyseal chondrocyte suspension over surfaces that were only half MAP-treated. This permitted corre-
J Orthop Res, Vol. 8, N o . 6, 1990
lation, by scanning electron microscopy, of cell attachment to MAP-treated and untreated surfaces.
Confirmation Attachment Study Using [3H]Leucine-LabeledOsteoblasts In a confirmatory series of experiments, [3H]leucine (Amersham C o p . , Arlington Heights, IL, U.S.A.) (l@ mCi/ml)-labeled ROS 17/28 osteoblasts (40-120,000/cm2) were placed in suspension over MAP-coated and uncoated culture plate (n = 5 ) and Vitallium disc ( n = 3) surfaces. Cells were labeled by incubating in F-12 medium with 5% fetal calf serum and 5 mCi of 13H]leucinefor 24 h. Following 15 min incubations, comparable coated and uncoated discs were placed in Econoflour scintillation fluid and counted in a Beckman scintillation counter. Cells were scraped from culture plate surfaces to determine cells attached with and without MAP.
ROS 17/28 Rat Osteosarcoma Osteoblast Growth on MAP
In another series ( n = 4) of experiments (three to four trials per experiment), ROS 17/28 rat osteosarcoma osteoblasts were attached, as described above, to MAP-coated culture plates or Vitallium discs and then incubated for 8-10 days. In an equal number of control experiments, cells were attached to bare plastic or metal. Vitallium discs were placed in standard 35 mm culture plates and covered with F-12 medium containing 5% fetal calf serum. After 8-10 days, medium was removed and the surfaces were stained for examination by light microscopy using direct or reflected light to evaluate cell growth on each surface. In two experiments, ROS 17/28 osteoblasts were attached to small MAP-coated “spots” on culture plates such that much of the plastic surface was still not coated with MAP. Original cell suspension was removed after a 15 min cell attachment time, and then F-12 medium with 5% fetal calf serum was applied and the samples incubated for 8-10 days. The experiment was designed to see if cells would remain where originally attached to MAP, or whether they would move to other areas, on the same surface, not coated with MAP.
795
ATTACHMENT OF CELLS USING ADHESIVE PROTEIN RESULTS
Evaluation of Cell Attachment to MAP-Treated and Control Surfaces Early research by Picciano and Benedict (7) and others has shown that mussel adhesive protein works well for attachment of endothelial cells and other cell lines to standard culture plates. As with other cell types, articular and epiphyseal chondrocytes appear to attach better to culture dishes when mussel adhesive protein is applied to the surface (Table 1). In our initial series of experiments, 83.6% of the cells in suspension were lost from suspension when mussel adhesive protein had been applied to the surface, whereas only 30% of the cells in suspension were lost to culture plate surfaces without adhesive. The percent of cells “lost” from suspension is a rough indicator of cells attached to the underlying surface. Results with Vitallium discs were similar (Table 2). In a series of eight experiments, 78% of epiphyseal chondrocytes in suspension were lost to Vitallium coated with mussel adhesive protein within 15 min. Without the adhesive protein, 59% of the cells were lost from suspension. Results on culture plate and Vitallium surfaces were confirmed using ROS 17/28 rat osteosarcoma osteoblasts. Scanning electron micrographs showed attachment of cells to uncoated, control plastic and Vitallium. However, more cells attached to MAP-coated Vitallium or plastic than to either surface alone, grossly confirming the attachment assay findings. This was particularly evident on surfaces that were halftreated by MAP. Figure 1 shows an epiphyseal chondrocyte attached to MAP-coated metal. The TABLE 1. Summary of data: percentage of chick epiphyseal chondrocytes (16-20 day) attached to 35 mm Petri plates after 15 min of incubation in F12 medium at 3PC
+ MAP(%)
Control (%)
76 85 89 96 79 77 u = 83.6% n = 6
35 38 34 23 15 35 u = 30% n = 6
These results are statistically significant at the 0.025 and 0.01 levels using the Kruskal-Wallis nonparametric method for comparing matched pairs (8).
TABLE 2. Summary of data: percentage of chick epiphyseal chondrocytes (16-20 day) attached to Vitallium discs (2.8 cm2) after 15 min of incubation in F-12 medium at 37°C
+ MAP (%)
Control (%)
64 79 75 83 87 76 93 72 u = 78.6% n = 8
41 30 60 77 79 56 88 44 u = 59% n = 8
These results are statistically significant at the 0.025 and 0.01 levels using the Kruskal-Wallis nonparametric method for comparing matched pairs (8).
scanning electron microscope (SEM) studies were qualitative only, and quantitative cell counts were not done. They confirmed, however, a pattern of cell attachment without significant spreading immediately after cell suspension application. Culture medium will not disrupt MAP from the surfaces tested, further supporting its efficacy as an underwater adhesive. Manual agitation of culture medium over MAP-coated surfaces caused no visual disruption of the applied MAP. High pressure fluid flow, however, was not applied over these MAP-coated surfaces. Furthermore, when culture medium was placed over cyanoacrylate adhesive (n = 3), there was notable and consistent blistering of the cyanoacrylate from the underlying culture plate surface within 15 min. This does not occur when MAP is used as an adhesive under culture medium. Additionally, MAP-coated surfaces were incubated with F-12 medium and 5% fetal calf serum for periods up to 2 weeks without altering subsequent cell attachment capability. ROS 17/28 Rat Osteosarcoma Osteoblast Growth on MAP
When ROS 17/28 rat osteosarcoma osteoblasts were attached to MAP initially, they tended to stay on the MAP-treated surface during culture growth and cell replication. Conversely, when cells were attached to culture plates, there was “seeding” to other uncoated parts of the metal or plastic during the 8-10 days of culture growth to confluence. In other words, MAP appears to hold cells and prevent spread to other areas (see Figs. 2A and 2B).
J Orthop Res, Vol. 8, No. 6 , 1990
796
J . P . FULKERSON ET AL.
FIG. 1. Chick embryo epiphyseal cell attached to mussel adhesive protein coated metal (magnification of ~6,000).
ROS 17/28 rat osteosarcoma osteoblasts were noted to achieve confluence on plastic culture plates or Vitallium with or without MAP. MAP appeared to encourage confluence somewhat earlier, perhaps because of a greater number of cells attached initially. Confluent cell growth was achieved 2 4 4 8 h sooner on MAP-coated surfaces when compared to uncoated surfaces. In our 8-10 day culture experiments, MAP provided an excellent surface for cell culture growth, with no evidence of cell toxicity, shape change, or disruption. Cells released
from MAP behaved similarly to cells that had not been attached with regard to subsequent confluent cell growth. Confirmatory Attachment Studies Using [3H]LeucineLabeled Osteoblasts Known concentrations of [3H]leucine-labeled ROS 17/28 cells in suspension were applied over MAP-coated and uncoated culture plates and vitallium discs. Three of four experiments demonstrated
FIG. 2. (A) Osteoblasts have "seeded" to other parts of a plastic culture plate after initial attachment to bare plastic. (8) Osteoblasts remain where initially attached after initial attachment to an area coated with MAP.
J Orthop Res, Vol. 8, N o . 6, 1990
ATTACHMENT OF CELLS USING ADHESIVE PROTEIN
improved attachments of cells to MAP-coated surfaces when 40,000 cells/cm were applied. When 80,000 and 120,000 cells/cm2 were applied, there was a consistent increase in the number of cells attached to MAP-coated culture plate surfaces vs. surfaces without MAP (Fig. 3). In experiments with [3H]leucine-labeled ROS 17/28 cells in suspension over MAP-coated and uncoated Vitallium, there was some enhancement of cell attachment (as determined by [3H]leucine counts) onto MAP-coated Vitallium but less striking enhancement than noted on culture plates. Overall, 9% more of the cells in suspension attached to Vitallium with MAP than to Vitallium alone. Enhancement of cell attachment with MAP w a s readily confirmed using t h e [3H]leucine tracer. DISCUSSION
Mussel adhesive protein (MAP) seems well suited for attaching cells to surfaces. These studies have shown that cartilage and osteosarcoma osteoblast attachment to synthetic surfaces can be enhanced
eJxp.dele
FIG. 3. Percent of [3H]leucine labelled ROS 17/28 rat osteosarcoma osteoblasts attaching to plastic culture plates with and without MAP, as determined by scintillation counter.
797
with MAP. Cells can multiply and colonize on these surfaces, indicating that MAP is nontoxic. There are other adhesives available for applications in surgery. Fibrin preparations and cyanoacrylates (super glue) have been tried most commonly. Fibrin adhesives have not provided a strong enough union of tissues to warrant widespread applications and immunogenicity of blood products raises some concern about widespread use of fibrin adhesives. Isobutyl-2-cyanoacrylate (“super glue”) has been effective in the repair of cortical bone fragments (12). Also, cyanoacrylates have been used for repair of corneal perforations (10). Little is known, however, about the long-term effects of these “super glue” adhesives on cells, and cell toxicity has been noted with cyanoacrylates (5). Culture medium and serum d o not disrupt MAP grossly, but they do cause blistering of cyanoacrylate on surfaces. At the time of total joint implantation, it is most desirable to achieve bone “ingrowth” onto the implant metal. This helps assure stable fixation and minimizes the risk of implant loosening. Application of bone cells directly to metal at the time of implantation might increase the chance of solid bone fixation since bone is formed directly on the implant itself. The main application of this work will likely be in the field of total joint implantation. Currently, there is interest in “coating” implants with hydroxyapatite or methylmethacrylate to enhance fixation either directly to bone or to a “filling” layer of methylmethacrylate. We believe that direct fixation to bone, by maximizing osteoblast presence on a prosthesis, will ultimately prove to be more effective than interposing other layers of material between prosthesis and bone. To the best of our knowledge, this study is the first to introduce the concept of adhesive-enhanced cell coating on a prosthesis. A nonreactive, nontoxic adhesive would enable the total joint replacement surgeon to assure adhesion of osteoblasts directly onto an implant surface. We believe that implants might eventually be coated with a film of adhesive and subsequently treated with a population of autogenous osteoblasts immediately prior to insertion. In this way, bone could begin growing directly on a prosthesis surface and ultimately incorporate with bone growing from the patient. We believe that the future of joint implant surgery may well include adhesives. These studies show consistent increases in numbers of cells attached to metal and plastic using
J Orthop Res, Vol. 8, N o . 6, 1990
J . P . FULKERSON ET AL.
798
MAP. It is also impressive, however, that cells attach well to the Vitallium alone (59% at 15 min, see Table 2) even though not quite as well (79% at 15 min) as with MAP. This observation raises the possibility of prosthesis treatment by bone cells prior to implantation, even without MAP. The results of these studies suggest that mussel adhesive protein is effective at the level of cellular adhesion. MAP may be helpful in cell culture research requiring specific or enhanced cell attachment. Further laboratory research with MAP will undoubtedly reveal other applications for this new adhesive. Acknowledgment: The authors wish to thank the Arthroscopy Association of North America, Howmedica, Inc., and the University of Connecticut Research Foundation for funding this research. The authors also wish to thank Marilyn Folcik, R.N., M.P.H. and Robert Rippey, Ph.D. for assistance with statistical analysis, Kathy Black and Susan Philo for help with manuscript preparation, and Anjali Saini for laboratory assistance. The authors are particularly indebted to Harry R. Gossling, M.D. for his ongoing support of these research efforts. In addition, the authors wish to thank Biopolymers, Inc., particularly Chris Benedict, for technical advice periodically during the course of these experiments and Mark Fontana for technical assistance.
REFERENCES 1 . Bobyn JD, Cameron HU, Abdulla D, Pilliar RM, Weatherly GC: Biologic fixation and metal modeling with an uncon-
J Orthop Res, Vol. 8, No. 6, 1990
strained canine total knee prosthesis. Clin Orthop 166:301312, 1982 2. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC: The optimum pore size for the fixation of porous-surfaced metal implants by the ingrowth of bone. Clin Orthop 150:263-270, 1980 3. Cameron HU, Pilliar RM, MacNab I: The rate of bone ingrowth into porous metal. J Biomed Muter Res 10:295-302, 1976 4. Cook SD, Walsh KA, Haddad RJ: Interface mechanics and bone growth into porous Co-Cr-Mo alloy implants. Clin Orthop 193:271-280, 1985 5. Lehman R, Hayes G, Leonard F: Toxicity of alkyl 2 cyanoacrylates. Arch Surg 93:441446, 1966 6. Majeski RJ, Rodan SB, Rodan GA: Parathyroid hormoneresponsive clonal cell lines from rat osteosarcoma. Endocrinology 107:1494, 1980 7. Picciano PT, Benedict CV: Mussel adhesive protein: a new cell attachment factor. In Vitro Cell Dev Biol22:24A, 1986 8. Rolf F, Sokal R: Statistical Tables, San Francisco, W. H. Freeman, 1969 9. Rosenquist R, Bylander B, Knutson K, Rydholm U, Rooser B, Egund N, Lidgren L: Loosening of the porous coating of bicompartmental prostheses in patients with rheumatoid arthritis. J Bone Joint Surg [Am] 68538-542, 1986 10. Soong M, Wolter R, Wolter J: Fistula excision and peripheral grafts in the treatment of persistent limbal wound leaks. Ophthalmology 95:31, 1988 1 1 . Waite JH, Tanzer M: Polyphenolic substance of Mytilus edulis: novel adhesive-containing L-Dopa and hydroxyproline. Science 212:1938-1940, 1981 12. Walker AM, Tomlinson JL: Use of isobutylcyanocrylate adhesive in osteosynthesis: a preliminary study. Vet Surg 13:257-262, 1984 13. Welsh RP, Pilliar RM, MacNab I: Surgical implants. The role of surface porosity in fixation to bone and acrylic. J Bone Joint Surg [Am] 53:963-977, 1971
Attachment of Epiphyseal Cartilage Cells and 17/28 Rat Osteosarcoma Osteoblasts Using Mussel Adhesive Protein *§John P. Fulkerson, ?Louis A. Norton, *Gloria Gronowicz, $.Paul Picciano, *§J. Mathieu Massicotte and *Carl W. Nissen Departments of *Orthopaedic Surgery and forthodontics, University of Connecticut School of Medicine, and jBiopolymers, Inc., Farmington, Connecticut, and §Marine Biological Laboratory, Woods Hole, Massachusetts, U.S.A.
Summary: These experiments show that mussel adhesive protein (MAP) enhances the attachment of osteoblasts and epiphyseal cartilage cells to plastic culture dishes and Vitallium. When MAP was applied to culture plate surfaces, there were two- to fivefold increases in the numbers of cells attaching compared to control surfaces (no MAP). Results were confirmed using two different cell attachment assay techniques. Osteoblast replication and culture on MAP is possible, suggesting that MAP is not toxic to cells. MAP also holds applied cells to surfaces as initially attached. Key Words: OsteoblastsAd hesive-Mu ssels-Surfaces.
Underwater adhesion is essential for survival of the common blue mussel, Myfilus edulis. This saltwater mollusc has four glands, including a phenol gland that produces a polyphenolic protein adhesive of molecular weight of 130,000. The adhesive (1 l), originally isolated and characterized by Waite in 1980, enables the blue mussel to attach itself to rocks and other surfaces underwater in order to resist tides and currents. Now that it is commercially available, it may offer important practical advantages in surgical and dental practice, particularly the ability to create adhesion in a saline environment without toxicity for cells and tissues. Although orthopaedic implant fixation is sometimes enhanced by using textured surfaces (14,13), porous metal coatings can loosen (9). Enhanced attachment of osteoblasts to metal, therefore, might permit improved implant fixation. Chondrocyte and osteoblast attachment to plastic
culture dishes is important in connective tissue research. These studies, therefore, provide an early look at mussel adhesive protein (MAP)-enhanced attachment of chondrocytes and osteoblasts to surfaces relevant to orthopaedic surgeons and connective tissue scientists. MATERIALS AND METHODS
Mussel adhesive protein (Celltak) was obtained in acid solution from Biopolymers, Inc. (Farmington, CT, U.S.A.) and applied sparsely to 35 mm Falcon plastic petri plates (30 pg/plate) and small Vitallium (Howmedica, Inc., Rutherford, NJ, U.S.A.) metal discs (7&140 pg/disc). A very small volume of Celltak was applied, 3-14 pl (10 pg/pl), using a small plastic spatula, and the adhesive was allowed to air dry in a laminar flow hood. Evaluation of Cell Attachment to MAP-Treated and Control Surfaces
Received October 13, 1988; accepted April 12, 1990. Address correspondence and reprint requests to Dr. J. P. Fulkerson at Department of Orthopaedic Surgery, University of Connecticut School of Medicine, Farmington, CT 06032, U.S.A.
Mussel adhesive protein (Celltak) was applied to one-half of all metal and plastic surfaces using the 793
794
J . P . FULKERSON ET AL.
following technique: First, each surface was covered with sterile 0.1 M sodium bicarbonate (pH 8.0). Celltak was diluted to make a 1 mg/ml solution and then mixed on the surface with buffer by pipetting several times. After 20 min, excess solution was aspirated off the surface and the surface was washed twice with sterile water. Sixteen- to 20-day-old chick embryo articular and epiphyseal cartilage was obtained from lower extremity long bones, and chondrocytes were released from matrix by incubation (37°C) of the tissue with collagenase and hyaluronidase (1 mg/ml) for 2 h using magnetic stirring pellets, after mincing the tissue thoroughly with scalpels. Heterogeneous populations of mixed epiphyseal and articular chondrocytes and populations of epiphyseal chondrocytes alone give similar results. Similar experiments were performed also with ROS 17/28 rat osteosarcoma osteoblasts (6), a well-characterized cell culture line. Cells were washed twice with F-12 medium (Gibco, Grand Island, NY, U.S.A.), suspended in F-12 medium, and counted on a hemocytometer. A known concentration of cells (usually about 2-5 x lo5 celldml) was placed over each of the MAP-treated surfaces and comparable untreated surfaces. Short incubation periods were chosen to see if rapid cell attachment would permit applications in a surgical solution. Following 5 to 15 min incubations, cell suspensions were agitated and removed from the surfaces being studied. Cells remaining in suspension were counted using a hemocytometer or Coulter counter and the percent of cells lost from suspension determined by subtracting the cell concentration remaining from that originally placed over the surface. These data were confirmed by examining all surfaces by light microscopy to correlate the numbers of cells attached to the cells lost from suspension. Cell shape was compared to controls to see if the adhesive caused any change. All surfaces were examined for evidence of MAP disruption by culture medium or agitation. Other surfaces were treated with isobutylcyanoacrylate (super glue) and examined after similar culture medium treatment and manual agitation. Selected surfaces were examined by scanning electron microscopy after careful dehydration and sputter-coating. In addition, four plastic and four Vitallium surfaces were treated with chick epiphyseal chondrocyte suspension over surfaces that were only half MAP-treated. This permitted corre-
J Orthop Res, Vol. 8, N o . 6, 1990
lation, by scanning electron microscopy, of cell attachment to MAP-treated and untreated surfaces.
Confirmation Attachment Study Using [3H]Leucine-LabeledOsteoblasts In a confirmatory series of experiments, [3H]leucine (Amersham C o p . , Arlington Heights, IL, U.S.A.) (l@ mCi/ml)-labeled ROS 17/28 osteoblasts (40-120,000/cm2) were placed in suspension over MAP-coated and uncoated culture plate (n = 5 ) and Vitallium disc ( n = 3) surfaces. Cells were labeled by incubating in F-12 medium with 5% fetal calf serum and 5 mCi of 13H]leucinefor 24 h. Following 15 min incubations, comparable coated and uncoated discs were placed in Econoflour scintillation fluid and counted in a Beckman scintillation counter. Cells were scraped from culture plate surfaces to determine cells attached with and without MAP.
ROS 17/28 Rat Osteosarcoma Osteoblast Growth on MAP
In another series ( n = 4) of experiments (three to four trials per experiment), ROS 17/28 rat osteosarcoma osteoblasts were attached, as described above, to MAP-coated culture plates or Vitallium discs and then incubated for 8-10 days. In an equal number of control experiments, cells were attached to bare plastic or metal. Vitallium discs were placed in standard 35 mm culture plates and covered with F-12 medium containing 5% fetal calf serum. After 8-10 days, medium was removed and the surfaces were stained for examination by light microscopy using direct or reflected light to evaluate cell growth on each surface. In two experiments, ROS 17/28 osteoblasts were attached to small MAP-coated “spots” on culture plates such that much of the plastic surface was still not coated with MAP. Original cell suspension was removed after a 15 min cell attachment time, and then F-12 medium with 5% fetal calf serum was applied and the samples incubated for 8-10 days. The experiment was designed to see if cells would remain where originally attached to MAP, or whether they would move to other areas, on the same surface, not coated with MAP.
795
ATTACHMENT OF CELLS USING ADHESIVE PROTEIN RESULTS
Evaluation of Cell Attachment to MAP-Treated and Control Surfaces Early research by Picciano and Benedict (7) and others has shown that mussel adhesive protein works well for attachment of endothelial cells and other cell lines to standard culture plates. As with other cell types, articular and epiphyseal chondrocytes appear to attach better to culture dishes when mussel adhesive protein is applied to the surface (Table 1). In our initial series of experiments, 83.6% of the cells in suspension were lost from suspension when mussel adhesive protein had been applied to the surface, whereas only 30% of the cells in suspension were lost to culture plate surfaces without adhesive. The percent of cells “lost” from suspension is a rough indicator of cells attached to the underlying surface. Results with Vitallium discs were similar (Table 2). In a series of eight experiments, 78% of epiphyseal chondrocytes in suspension were lost to Vitallium coated with mussel adhesive protein within 15 min. Without the adhesive protein, 59% of the cells were lost from suspension. Results on culture plate and Vitallium surfaces were confirmed using ROS 17/28 rat osteosarcoma osteoblasts. Scanning electron micrographs showed attachment of cells to uncoated, control plastic and Vitallium. However, more cells attached to MAP-coated Vitallium or plastic than to either surface alone, grossly confirming the attachment assay findings. This was particularly evident on surfaces that were halftreated by MAP. Figure 1 shows an epiphyseal chondrocyte attached to MAP-coated metal. The TABLE 1. Summary of data: percentage of chick epiphyseal chondrocytes (16-20 day) attached to 35 mm Petri plates after 15 min of incubation in F12 medium at 3PC
+ MAP(%)
Control (%)
76 85 89 96 79 77 u = 83.6% n = 6
35 38 34 23 15 35 u = 30% n = 6
These results are statistically significant at the 0.025 and 0.01 levels using the Kruskal-Wallis nonparametric method for comparing matched pairs (8).
TABLE 2. Summary of data: percentage of chick epiphyseal chondrocytes (16-20 day) attached to Vitallium discs (2.8 cm2) after 15 min of incubation in F-12 medium at 37°C
+ MAP (%)
Control (%)
64 79 75 83 87 76 93 72 u = 78.6% n = 8
41 30 60 77 79 56 88 44 u = 59% n = 8
These results are statistically significant at the 0.025 and 0.01 levels using the Kruskal-Wallis nonparametric method for comparing matched pairs (8).
scanning electron microscope (SEM) studies were qualitative only, and quantitative cell counts were not done. They confirmed, however, a pattern of cell attachment without significant spreading immediately after cell suspension application. Culture medium will not disrupt MAP from the surfaces tested, further supporting its efficacy as an underwater adhesive. Manual agitation of culture medium over MAP-coated surfaces caused no visual disruption of the applied MAP. High pressure fluid flow, however, was not applied over these MAP-coated surfaces. Furthermore, when culture medium was placed over cyanoacrylate adhesive (n = 3), there was notable and consistent blistering of the cyanoacrylate from the underlying culture plate surface within 15 min. This does not occur when MAP is used as an adhesive under culture medium. Additionally, MAP-coated surfaces were incubated with F-12 medium and 5% fetal calf serum for periods up to 2 weeks without altering subsequent cell attachment capability. ROS 17/28 Rat Osteosarcoma Osteoblast Growth on MAP
When ROS 17/28 rat osteosarcoma osteoblasts were attached to MAP initially, they tended to stay on the MAP-treated surface during culture growth and cell replication. Conversely, when cells were attached to culture plates, there was “seeding” to other uncoated parts of the metal or plastic during the 8-10 days of culture growth to confluence. In other words, MAP appears to hold cells and prevent spread to other areas (see Figs. 2A and 2B).
J Orthop Res, Vol. 8, No. 6 , 1990
796
J . P . FULKERSON ET AL.
FIG. 1. Chick embryo epiphyseal cell attached to mussel adhesive protein coated metal (magnification of ~6,000).
ROS 17/28 rat osteosarcoma osteoblasts were noted to achieve confluence on plastic culture plates or Vitallium with or without MAP. MAP appeared to encourage confluence somewhat earlier, perhaps because of a greater number of cells attached initially. Confluent cell growth was achieved 2 4 4 8 h sooner on MAP-coated surfaces when compared to uncoated surfaces. In our 8-10 day culture experiments, MAP provided an excellent surface for cell culture growth, with no evidence of cell toxicity, shape change, or disruption. Cells released
from MAP behaved similarly to cells that had not been attached with regard to subsequent confluent cell growth. Confirmatory Attachment Studies Using [3H]LeucineLabeled Osteoblasts Known concentrations of [3H]leucine-labeled ROS 17/28 cells in suspension were applied over MAP-coated and uncoated culture plates and vitallium discs. Three of four experiments demonstrated
FIG. 2. (A) Osteoblasts have "seeded" to other parts of a plastic culture plate after initial attachment to bare plastic. (8) Osteoblasts remain where initially attached after initial attachment to an area coated with MAP.
J Orthop Res, Vol. 8, N o . 6, 1990
ATTACHMENT OF CELLS USING ADHESIVE PROTEIN
improved attachments of cells to MAP-coated surfaces when 40,000 cells/cm were applied. When 80,000 and 120,000 cells/cm2 were applied, there was a consistent increase in the number of cells attached to MAP-coated culture plate surfaces vs. surfaces without MAP (Fig. 3). In experiments with [3H]leucine-labeled ROS 17/28 cells in suspension over MAP-coated and uncoated Vitallium, there was some enhancement of cell attachment (as determined by [3H]leucine counts) onto MAP-coated Vitallium but less striking enhancement than noted on culture plates. Overall, 9% more of the cells in suspension attached to Vitallium with MAP than to Vitallium alone. Enhancement of cell attachment with MAP w a s readily confirmed using t h e [3H]leucine tracer. DISCUSSION
Mussel adhesive protein (MAP) seems well suited for attaching cells to surfaces. These studies have shown that cartilage and osteosarcoma osteoblast attachment to synthetic surfaces can be enhanced
eJxp.dele
FIG. 3. Percent of [3H]leucine labelled ROS 17/28 rat osteosarcoma osteoblasts attaching to plastic culture plates with and without MAP, as determined by scintillation counter.
797
with MAP. Cells can multiply and colonize on these surfaces, indicating that MAP is nontoxic. There are other adhesives available for applications in surgery. Fibrin preparations and cyanoacrylates (super glue) have been tried most commonly. Fibrin adhesives have not provided a strong enough union of tissues to warrant widespread applications and immunogenicity of blood products raises some concern about widespread use of fibrin adhesives. Isobutyl-2-cyanoacrylate (“super glue”) has been effective in the repair of cortical bone fragments (12). Also, cyanoacrylates have been used for repair of corneal perforations (10). Little is known, however, about the long-term effects of these “super glue” adhesives on cells, and cell toxicity has been noted with cyanoacrylates (5). Culture medium and serum d o not disrupt MAP grossly, but they do cause blistering of cyanoacrylate on surfaces. At the time of total joint implantation, it is most desirable to achieve bone “ingrowth” onto the implant metal. This helps assure stable fixation and minimizes the risk of implant loosening. Application of bone cells directly to metal at the time of implantation might increase the chance of solid bone fixation since bone is formed directly on the implant itself. The main application of this work will likely be in the field of total joint implantation. Currently, there is interest in “coating” implants with hydroxyapatite or methylmethacrylate to enhance fixation either directly to bone or to a “filling” layer of methylmethacrylate. We believe that direct fixation to bone, by maximizing osteoblast presence on a prosthesis, will ultimately prove to be more effective than interposing other layers of material between prosthesis and bone. To the best of our knowledge, this study is the first to introduce the concept of adhesive-enhanced cell coating on a prosthesis. A nonreactive, nontoxic adhesive would enable the total joint replacement surgeon to assure adhesion of osteoblasts directly onto an implant surface. We believe that implants might eventually be coated with a film of adhesive and subsequently treated with a population of autogenous osteoblasts immediately prior to insertion. In this way, bone could begin growing directly on a prosthesis surface and ultimately incorporate with bone growing from the patient. We believe that the future of joint implant surgery may well include adhesives. These studies show consistent increases in numbers of cells attached to metal and plastic using
J Orthop Res, Vol. 8, N o . 6, 1990
J . P . FULKERSON ET AL.
798
MAP. It is also impressive, however, that cells attach well to the Vitallium alone (59% at 15 min, see Table 2) even though not quite as well (79% at 15 min) as with MAP. This observation raises the possibility of prosthesis treatment by bone cells prior to implantation, even without MAP. The results of these studies suggest that mussel adhesive protein is effective at the level of cellular adhesion. MAP may be helpful in cell culture research requiring specific or enhanced cell attachment. Further laboratory research with MAP will undoubtedly reveal other applications for this new adhesive. Acknowledgment: The authors wish to thank the Arthroscopy Association of North America, Howmedica, Inc., and the University of Connecticut Research Foundation for funding this research. The authors also wish to thank Marilyn Folcik, R.N., M.P.H. and Robert Rippey, Ph.D. for assistance with statistical analysis, Kathy Black and Susan Philo for help with manuscript preparation, and Anjali Saini for laboratory assistance. The authors are particularly indebted to Harry R. Gossling, M.D. for his ongoing support of these research efforts. In addition, the authors wish to thank Biopolymers, Inc., particularly Chris Benedict, for technical advice periodically during the course of these experiments and Mark Fontana for technical assistance.
REFERENCES 1 . Bobyn JD, Cameron HU, Abdulla D, Pilliar RM, Weatherly GC: Biologic fixation and metal modeling with an uncon-
J Orthop Res, Vol. 8, No. 6, 1990
strained canine total knee prosthesis. Clin Orthop 166:301312, 1982 2. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC: The optimum pore size for the fixation of porous-surfaced metal implants by the ingrowth of bone. Clin Orthop 150:263-270, 1980 3. Cameron HU, Pilliar RM, MacNab I: The rate of bone ingrowth into porous metal. J Biomed Muter Res 10:295-302, 1976 4. Cook SD, Walsh KA, Haddad RJ: Interface mechanics and bone growth into porous Co-Cr-Mo alloy implants. Clin Orthop 193:271-280, 1985 5. Lehman R, Hayes G, Leonard F: Toxicity of alkyl 2 cyanoacrylates. Arch Surg 93:441446, 1966 6. Majeski RJ, Rodan SB, Rodan GA: Parathyroid hormoneresponsive clonal cell lines from rat osteosarcoma. Endocrinology 107:1494, 1980 7. Picciano PT, Benedict CV: Mussel adhesive protein: a new cell attachment factor. In Vitro Cell Dev Biol22:24A, 1986 8. Rolf F, Sokal R: Statistical Tables, San Francisco, W. H. Freeman, 1969 9. Rosenquist R, Bylander B, Knutson K, Rydholm U, Rooser B, Egund N, Lidgren L: Loosening of the porous coating of bicompartmental prostheses in patients with rheumatoid arthritis. J Bone Joint Surg [Am] 68538-542, 1986 10. Soong M, Wolter R, Wolter J: Fistula excision and peripheral grafts in the treatment of persistent limbal wound leaks. Ophthalmology 95:31, 1988 1 1 . Waite JH, Tanzer M: Polyphenolic substance of Mytilus edulis: novel adhesive-containing L-Dopa and hydroxyproline. Science 212:1938-1940, 1981 12. Walker AM, Tomlinson JL: Use of isobutylcyanocrylate adhesive in osteosynthesis: a preliminary study. Vet Surg 13:257-262, 1984 13. Welsh RP, Pilliar RM, MacNab I: Surgical implants. The role of surface porosity in fixation to bone and acrylic. J Bone Joint Surg [Am] 53:963-977, 1971
