9
Development of fibroblast-seeded ligament analogs for ACL reconstruction Michael G. Dunn,',* Janice B. Liesch,' Moti L. Tiku,' and Joseph P. Zawadsky' 1 Orthopaedic Research Laboratory, Division of Orthopaedic Surgery, and 'Department of Medicine, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, New Jersey 08903 We fabricated "ligament analogs" in vitro by seeding high- strength resorbable collagen fiber scaffolds with intraartic- ular (anterior cruciate ligament, ACL) or extraarticular (pa- tellar tendon, PT) rabbit fibroblasts. Fibroblasts attached, proliferated, and secreted new collagen on the ligament analogs in vitro. Fibroblast function depended on the tissue culture substrate (ligament analog vs. tissue culture plate) and the origin of the fibroblasts (ACL vs. PT). PT fibroblasts proliferated more rapidly than ACL fibroblasts when cul- tured on ligament analogs. Collagen synthesis by ACL and PT fibroblasts was approximately tenfold greater on liga- ment analogs than on tissue culture plates. The composi- tion, structure, and geometry of the collagen fiber scaffolds may promote collagen synthesis within ligament analogs in uitro. Ligament analogs roughly approximate the structure and strength of native ligament tissue. Ongoing in vivo studies suggest that autogenous fibroblast-seededligament analogs remain viable after implantation into the knee joint. With further development, ligament analogs may be useful as implants for ACL reconstruction surgery. 0 1995 John Wiley & Sons, Inc. INTRODUCTION Severe injury to the anterior cruciate ligament (ACL) can cause knee instability, meniscal damage, and osteoarthritis.' Because the ACL heals poorly fol- lowing primary repair, surgical reconstruction is rec- ommended to improve joint function in young active patients.lT2Patellar tendon (PT) autografts and al- lografts are widely used but are not ideally suited for this purpose.14 Problems associated with PT au- tografts include lengthy rehabilitation and persistent patellar paina5 PT allografts carry the risk of disease transmission,6 and their procurement is labor inten- sive and co~tly.~ Both autografts and allografts may become necrotic and weak4 shortly after implanta- tion, and the knee must be protected from high me- chanical loads while the graft gradually gains strength. Permanent synthetic ACL prostheses may perform satisfactorily in the short term, but tend to break down and fail in the long Currently, no prosthesis is approved by the United States Food and Drug Administration for primary ACL reconstruc- tion. Resorbable scaffolds seeded with cells are potential alternatives to biological grafts or permanent prosthe- *To whom correspondence should be addressed. ses. This tissue engineering" strategy has been used by others to repair large defects in skin" and carti- lage.'' Our previous studies suggest that the ACL can be regenerated using a similar approach. We showed that acellular collagen scaffolds can induce neotendon and neoligament formation in rab- bit Achilles tendon'517 and ACL." In our ACL re- construction study, however, nearly half of the col- lagen scaffold implants did not induce ingrowth of functional neoligament tissue. Tissue ingrowth into ACL prostheses and grafts is inconsistent and diffi- cult to control. We hypothesize that autogenous fibroblast seeding of the collagen scaffolds prior to implantation might im- prove neoligament formation in the reconstructed ACL. Therefore, our objective was to fabricate "Iiga- ment analogs" by seeding high-strength resorbable collagen fiber scaffolds with viable fibroblasts from intraarticular (ACL) or extraarticular (PT) tissues. Fi- broblast attachment, morphology, proliferation, and collagen synthesis varied as a function of the culture substrate and the origin of the fibroblast. Collagen synthesis was tenfold greater on ligament analogs than on tissue culture plates. Ligament analogs roughly approximate the structure and strength of native ligament tissue. With further development, Journal of Biomedical Materials Research, Vol. 29, 136S1371 (1995) 0 1995 John Wiley & Sons, Inc. CCC 0021-9304/95/111363-09

Development of fibroblast-seeded ligament analogs for ACL reconstruction

Embed Size (px)

Citation preview

Page 1: Development of fibroblast-seeded ligament analogs for ACL reconstruction

Development of fibroblast-seeded ligament analogs for ACL reconstruction

Michael G. Dunn,',* Janice B. Liesch,' Moti L. Tiku,' and Joseph P. Zawadsky' 1 Orthopaedic Research Laboratory, Division of Orthopaedic Surgery, and 'Department of Medicine, University of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School, New Brunswick, New Jersey 08903

We fabricated "ligament analogs" in vitro by seeding high- strength resorbable collagen fiber scaffolds with intraartic- ular (anterior cruciate ligament, ACL) or extraarticular (pa- tellar tendon, PT) rabbit fibroblasts. Fibroblasts attached, proliferated, and secreted new collagen on the ligament analogs in vitro. Fibroblast function depended on the tissue culture substrate (ligament analog vs. tissue culture plate) and the origin of the fibroblasts (ACL vs. PT). PT fibroblasts proliferated more rapidly than ACL fibroblasts when cul- tured on ligament analogs. Collagen synthesis by ACL and PT fibroblasts was approximately tenfold greater on liga-

ment analogs than on tissue culture plates. The composi- tion, structure, and geometry of the collagen fiber scaffolds may promote collagen synthesis within ligament analogs in uitro. Ligament analogs roughly approximate the structure and strength of native ligament tissue. Ongoing in vivo studies suggest that autogenous fibroblast-seeded ligament analogs remain viable after implantation into the knee joint. With further development, ligament analogs may be useful as implants for ACL reconstruction surgery. 0 1995 John Wiley & Sons, Inc.

INTRODUCTION

Severe injury to the anterior cruciate ligament (ACL) can cause knee instability, meniscal damage, and osteoarthritis.' Because the ACL heals poorly fol- lowing primary repair, surgical reconstruction is rec- ommended to improve joint function in young active patients.lT2 Patellar tendon (PT) autografts and al- lografts are widely used but are not ideally suited for this purpose.14 Problems associated with PT au- tografts include lengthy rehabilitation and persistent patellar paina5 PT allografts carry the risk of disease transmission,6 and their procurement is labor inten- sive and c o ~ t l y . ~ Both autografts and allografts may become necrotic and weak4 shortly after implanta- tion, and the knee must be protected from high me- chanical loads while the graft gradually gains strength. Permanent synthetic ACL prostheses may perform satisfactorily in the short term, but tend to break down and fail in the long Currently, no prosthesis is approved by the United States Food and Drug Administration for primary ACL reconstruc- tion.

Resorbable scaffolds seeded with cells are potential alternatives to biological grafts or permanent prosthe-

*To whom correspondence should be addressed.

ses. This tissue engineering" strategy has been used by others to repair large defects in skin" and carti- lage.'' Our previous studies suggest that the ACL can be regenerated using a similar approach. We showed that acellular collagen scaffolds can induce neotendon and neoligament formation in rab- bit Achilles tendon'517 and ACL." In our ACL re- construction study, however, nearly half of the col- lagen scaffold implants did not induce ingrowth of functional neoligament tissue. Tissue ingrowth into ACL prostheses and grafts is inconsistent and diffi- cult to control.

We hypothesize that autogenous fibroblast seeding of the collagen scaffolds prior to implantation might im- prove neoligament formation in the reconstructed ACL. Therefore, our objective was to fabricate "Iiga- ment analogs" by seeding high-strength resorbable collagen fiber scaffolds with viable fibroblasts from intraarticular (ACL) or extraarticular (PT) tissues. Fi- broblast attachment, morphology, proliferation, and collagen synthesis varied as a function of the culture substrate and the origin of the fibroblast. Collagen synthesis was tenfold greater on ligament analogs than on tissue culture plates. Ligament analogs roughly approximate the structure and strength of native ligament tissue. With further development,

Journal of Biomedical Materials Research, Vol. 29, 136S1371 (1995) 0 1995 John Wiley & Sons, Inc. CCC 0021-9304/95/111363-09

Page 2: Development of fibroblast-seeded ligament analogs for ACL reconstruction

1364

I +

these ligament analogs may be useful as implants for ACL reconstruction.

I

MATERIALS AND METHODS

Fibroblasts were harvested from rabbit anterior cru- ciate ligament (ACL) and patellar tendon (PT), grown in culture, and seeded onto tissue culture plates and collagen fiber scaffolds, creating ligament analogs (Fig. 1). Fibroblast attachment, morphology, prolifer- ation, and collagen synthesis were measured in vitro as a function of substrate and fibroblast origin.

Fabrication of fibrous collagen

Freeze-dried acid-insoluble bovine dermal collagen was ground in a Wiley mill and dispersed at a con- centration of 1% (wt/vol) in pH 3.0 HCl in a blender for 3 min at 10,000 rpm. The dispersion was degassed under vacuum for 4 h and stored in a 30 cc plastic syringe at 4°C. Collagen “fibers” (60 pm average dry diameter) were produced by extrusion of the collagen dispersion into fiber formation buffer, using a syringe pump and polyethylene tubing with an inner diam- eter of 580 p.m. The buffer was composed of 135 mM NaC1, 30 mM TES [N-tris(hydroxymethyI)methyl-2-

INS OL U B L E CO LL AG E N FIBERS ARE EXTRUDED AND

LSSEMBLED INTO SCAFFOLDS.

FIBROBLASTS FROM RABBIT ~ ACL OR PT ARE EXPLANTED

AND GROWN IN CULTURE. 1 1 I

FIBROBLAST-SEEDED LIGAMENT ANALOG

Figure 1. Schematic representation of ligament analog fabrication. Collagen fiber scaffolds were seeded with rab- bit ACL or PT fibroblasts. Fibroblasts attached, proliferated, and secreted collagen on the ligament analogs in vitro.

DUNN El’ AL.

aminoethane sulfonic acid] and 30 mM sodium phos- phate dibasic, at pH 7.4 and 37°C. In this buffer, the pH, temperature, and salt concentration approach physiological values, and the acid-swollen collagen fibrils and fiber fragments within the dispersion de-swell and aggregate within the extruded collagen “fibers .”

After 45 min in the buffer, the extruded fibers were transferred to an isopropanol bath for 4 h, then washed for 20 min in distilled water. Fibers were dried overnight under tension (their own weight) to improve collagen orientation along the longitudinal fiber axis. Fibers were crosslinked using dehydrother- ma1 treatment*’ to avoid cytotoxic byproducts associ- ated with chemical crosslinking agents. Fibers were placed in an oven at 110°C under a vacuum of less than 1 millitorr for three days.

Collagen fiber scaffolds (Fig. 2) were prepared by aligning 200 crosslinked collagen fibers (length = 15 cm) in parallel, coating the fibers with a 1% (wthol) collagen dispersion, rinsing extensively in distilled water, and drying overnight.

Establishment of rabbit fibroblast cultures

Tissue samples for primary explants were removed from mature male New Zealand white rabbits using general anesthesia and sterile surgical procedures. NIH guidelines for the care and use of laboratory an- imals were observed,20 and all procedures were IACUC approved. Samples of rabbit ACL and patel- lar tendon were obtained, cut into 1-2 mm pieces, and placed in polystyrene tissue-culture plates. The culture media (referred to as media I) used in these procedures and for cell propagation was Dulbecco’s

Figure 2. The collagen scaffolds consisted of 200 ex- truded, dehydrothermally crosslinked collagen fibers aligned in parallel. Dry fiber diameters were approximately 50-70 km (Bar = 100 km).

Page 3: Development of fibroblast-seeded ligament analogs for ACL reconstruction

LIGAMENT ANALOGS FOR ACL RECONSTRUCTION 1365

modified Eagle's medium, 10% fetal calf serum, 2 mM glutamine, 100 @mL penicillin, 100 pg/mL strepto- mycin, and 0.25 pg/mL amphotericin B. Media I (1 mL) was carefully added to the plates incubated at 37°C in a humidified atmosphere of 5% C02. The media volume was gradually increased over the ini- tial 2-5 days to 5 mL. After approximately 7 days, the explant pieces were discarded and the outgrown cells were removed with trypsin-EDTA for subculture. Cells were maintained in culture using standard pro- cedures and used at passage 4.

Ligament analogs: Collagen scaffolds seeded with fibroblasts

Fibroblasts were released from culture plates by trypsin-EDTA treatment (terminated by the addition of 1 mg/mL soybean trypsin inhibitor) followed by washing and resuspension in media I and 10 mM HEPES pH 7.4. Cells were resuspended to a final con- centration of lo6 cells/mL. Cell aliquots of 0.1 mL were added to 24-well sterile tissue-culture plates containing 0.1 mL of media I and 10 mM HEPES pH 7.4, with a 1 cm length of sterile collagen scaffold on the bottom of the plate. After 24 h, the seeded colla- gen scaffolds (referred to hereafter as ligament ana- logs) were removed and placed in 96-well tissue- culture treated polystyrene plates with 0.2 mL fresh media I and 10 mM HEPES pH 7.4.

Determination of initial attachment of fibroblasts

Initial fibroblast attachment to tissue culture plates and ligament analogs was measured using 51Cr- labeling. Labeled fibroblasts (lo5 per well) were al- lowed to attach to the substrate for l , 2, 4, or 24 h. After those time periods, the ligament analogs and tissue-culture plates were rinsed extensively so only adherent cells remained, and the total radioactivity (51Cr counts per minute) was measured using a liquid scintillation counter.21

Determination of fibroblast morphology

Fibroblast morphology was examined using phase contrast microscopy and ultraviolet light microscopy for fluorescently labeled fibroblasts. Fibroblast mem- branes were labeled using the fluorescent lipophilic dye PKH2-GL (Zynaxis Cell Science, Malvern, PA). Diluent and PKH2-GL dye (0.5 mL each, with final dye concentration of 5 pMolar) were added to each pellet containing lo6 cells. Fibroblasts were incubated for 10 min at room temperature. Dye incorporation

was terminated by the addition of 1 mL of media I. Fibroblasts were separated from excess dye by cen- trifugation through serum, washed 3 times with me- dia, and resuspended to a concentration of 1 x lo6 cells/mL in media I. The PKHZGL dye has an exci- tation wavelength of 488 nm and emits at 525 nm. This fluorescent dye is incorporated and retained in the plasma membrane of viable cells for more than 200 days.*l During mitosis, the fluorescent mem- brane dye is transferred equally to progeny by shar- ing the parent cell membrane. Therefore, the dye is a specific marker for viable cells and their progeny.

Determination of fibroblast proliferation rate and collagen synthesis

Fibroblast proliferation rate (at days 2, 3, 4, and 7) and collagen synthesis (at day 4) were determined by radioactive labeling studies. Ligament analogs were fabricated by initially seeding collagen scaffolds with lo5 fibroblasts. At the designated time periods, the ligament analogs were removed from the wells and placed in new wells with 0.18 mL fresh media I. For proliferation measurements, 3[H] thymidine (1 pCi/ well, 5 pCi/mL, 0.02 mL) was added to each well, and the cells were incubated for 4 h at 37 "C in 5% CO,. For collagen synthesis measurements, 3[H] proline (1 pCi/well) was added to each well, and cells were in- cubated for 24 h.

After the incubation period, ligament analogs were removed from the wells, washed in HBSS, and placed in 0.3 mL 20% cold TCA for 30 min at 4 "C. Samples were centrifuged at 19,000 rpm for 10 min, and the supernatant was discarded. Two more times the pel- let was rewashed with cold 10% TCA, centrifuged, and the supernatant discarded. For proliferation mea- surements, perchloric acid (2N, 0.4 mL) was added, samples were incubated at 60 "C for 30 min and al- lowed to cool. For collagen synthesis measurements, 0.4 mL of 0.3 N NaOH with 1% sodium lauryl sulfate was added, and the samples were incubated at room temperature for 30 min. Ready Safe (4.6 mL) was added, and radioactivity (counts per min) was mea- sured by liquid scintillation counting. Data were nor- malized to radioactive counts per minute per cell.

Statistical analyses

For the initial attachment studies, a total of 48 mea- surements were made: 4 groups (ACL and PT fibro- blasts, cultured on plates and on ligament analogs) X

12 measurements per group. For the proliferation studies, a total of 48 measurements were made: 4 groups (as above) x 12 measurements per group. For

Page 4: Development of fibroblast-seeded ligament analogs for ACL reconstruction

1366 DUNN ET AL.

collagen synthesis studies, a total of 12 measure- ments were made: 4 groups x 3 measurements per group. Analysis of variance was performed using Statgraphics software to determine the effects of fi- broblast type (ACL vs. PT) and substrate (ligament analogs vs. culture plates) on fibroblast behavior. These effects were considered significant for p values less than 0.05.

RESULTS

We fabricated ligament analogs resembling native ligament tissue by seeding ACL and PT fibroblasts onto collagen fiber scaffolds in vitro. ACL and PT fi- broblasts attached, proliferated, and secreted new matrix on ligament analogs in vitro. Fibroblast func- tion was dependent on both the origin of the fibro- blasts (ACL vs. PT) and the substrate on which the fibroblasts were seeded (ligament analogs vs. tissue culture plates).

Initial attachment of fibroblasts

The number of fibroblasts attached through the first 24 h to ligament analogs and tissue culture plates is shown in Figure 3 (51Cr counts per minute). No significant differences existed as a function of time (comparing cell numbers at 1, 2, 4, and 24 h, data not shown), so data for these 4 time periods were pooled.

Initial attachment of fibroblasts was significantly

INITIAL FIBROBLAST ATTACHMENT

TISSUE CULTURE PLATE LIGAMENT ANALOG

Figure 3. Initial attachment (radiolabeled 51Cr counts) of ACL and PT fibroblasts to ligament analogs and tissue cul- ture plates in vitro (mean values with standard deviations). ACL and PT fibroblast attachment data were not signifi- cantly different. However, significantly fewer (p < 0.05) fibroblasts initially were attached to ligament analogs (*)

influenced by the substrate but not by the origin of the fibroblast. The total number of attached fibro- blasts (not normalized per area) was significantly greater on tissue culture plates than on ligament an- alogs for both ACL and PT fibroblasts.

Fibroblast morphology

Fibroblast morphology was dependent on both the tissue culture substrate and the origin of the fibro- blast. ACL fibroblasts were well spread on tissue cul- ture plates and were ovoid in shape [Fig. 4(A)]. No preferred orientation was observed. PT fibroblasts on tissue culture plates [Fig. 4(B)] appeared longer and thinner than ACL fibroblasts, with increased cell-cell connections compared to ACL fibroblasts. Again, no preferred orientation was noted.

On ligament analogs, ACL fibroblasts remained plump and were arranged in columns along the col- lagen fibers in some areas [Fig. 4(C)]. Many PT fibro- blasts, in contrast, became bipolar and aligned their long axis with the long axis of the collagen fibers [Fig. 4(D)j. Fibroblast sizes and shapes on the ligament analogs in vifro were consistent with previous de- scriptions of ACL and PT fibroblasts in viv~.*~

Fibroblast proliferation

The rate of fibroblast proliferation (3H-thymidine incorporation per cell, Fig. 5) was influenced both by the substrate and the origin of the fibroblast. On tis- sue-culture plates [Fig. 5(A)], ACL fibroblasts prolif- erated at an increased rate compared to PT fibro- blasts through day 3. On ligament analogs (Fig. 5B), the trend was reversed; PT fibroblasts proliferated more rapidly than ACL fibroblasts. PT fibroblast pro- liferation rate was similar on both substrates, but ACL fibroblast proliferation rate was decreased on ligament analogs compared to tissue-culture plates. Fibroblast proliferation rates decreased sharply by day 4 on tissue culture plates, and by day 7 on liga- ment analogs.

Collagen synthesis

Collagen synthesis (at day 4, after seeding in vitro) was significantly influenced by the substrate but not by the origin of the fibroblast (3H-proline incorpora- tion per cell, Fig. 6). On both substrates collagen syn- thesis by ACL and by PT fibroblasts was similar. - -

than to tissue culiure plates. However, collagen synthesis was significantly in-

Page 5: Development of fibroblast-seeded ligament analogs for ACL reconstruction

LIGAMENT ANALOGS FOR ACL RECONSTRUCTION 1367

Figure 4. Morphology of anterior cruciate ligament (A, C) and patellar tendon (B, D) fibroblasts on tissue culture plates (A, B) and ligament analogs (C, D) in vitro. A) ACL fibroblasts were well spread and ovoid-shaped on tissue culture plates. Bar = 20 km. B) PT fibroblasts on tissue culture plates were longer and thinner than ACL fibroblasts, with more cell-cell connections. Bar = 20 pm. 4C) Ligament analog consisting of collagen scaffold seeded with ACL fibroblasts. The fibroblast cell membranes were stained with a fluorescent green dye to monitor their viability, location, and morphology. In some areas the ACL fibroblasts were arranged in linear clusters (arrow) along the side of collagen fibers (CF). Bar = 100 pm. D) PT fibroblast-seeded ligament analog. PT fibroblasts tended to become bipolar and oriented along the long axis of the collagen fibers (CF). Bar = 100 pm.

creased (about tenfold) on ligament analogs com- pared to tissue-culture plates.

plications for Achilles tendon1517 and ACL recon- struction." Extruded acid-insoluble collagen fibers have wet tensile strengths of 40 MPa or greater, de- pending on the crosslinking method.14

DISCUSSION

We have developed viable fibroblast-seeded liga- ment analogs that are potentially useful as implants for ACL reconstruction and as tissue models to study fibroblast function in vitro. Previously, Huang et al.23 fabricated "ligament equivalents" by seeding dermal fibroblasts onto aligned collagen gels. However, the tensile strength of their gels (0.14 MPa) was 2 orders of magnitude lower than that of ligament tissue (38 MPa),24 rendering their constructs unsuitable for load-bearing applications. In contrast, we have used acellular collagen fiber scaffolds in load-bearing ap-

Effect of collagen fiber substrate on fibroblast function

Collagenous coatings, sheets, sponges, gels, and tissues have been used as tissue culture sub- strates.25zs The source, composition, solubility, de- gree of crosslinking, and physical form of the colla- gen can influence cell function. Collagen coatings or sheets provide only a two-dimensional substrate.26 Three-dimensional lattices (gels,27 sponges28) more closely approximate the extracellular matrix of tis-

Page 6: Development of fibroblast-seeded ligament analogs for ACL reconstruction

DUNN ET AL. 1368

PROLIFERATION RATE ON CULTURE PLATE

A DAYS

PROLIFERATION RATE ON LIGAMENT ANALOG

DAYS B Figure 5. Proliferation rate (radiolabeled thymidine incor- porati.on) of ACL and PT fibroblasts on (A) tissue culture plates, and (B) ligament analogs in vitro (mean values with standard deviations). (A) On tissue culture plates, ACL fi- broblast proliferation rate was slightly increased compared with that of PT fibroblasts. By day 4, the proliferation rate decreased significantly for both fibroblast types. (B) ACL fibroblast proliferation rate decreased on ligament analogs compared with tissue culture plates while PT fibroblast pro- liferation rate remained the same as that on tissue culture plates. By day 7, the proliferation rate decreased signifi- cantly for both fibroblast types.

sues, such as dermis, and have been developed into ”skin equivalents.” Collagen fiber scaffolds, unlike sponges or gels, approximate the structure of normal tendon or ligament, with continuous, aligned ”fi- bers’’ analogous to fascicles or fiber bundles in native tissue.22

In this study, fibroblast function was strongly in- fluenced by the use of extruded collagen fibers as a substrate. Varying the substrate (tissue culture plate vs. ligament analogs) was more influential than vary- ing fibroblast origin (ACL vs. PT). Fibroblast mor- phology on ligament analogs was similar to native fibroblast morphology in vivo.22 Many PT fibroblasts were bipolar and aligned with the long axis of the

collagen fibers. ACL fibroblasts were more plump, and in some areas were arranged in columns along- side the collagen fibers. Contact guidance, or cell elongation and orientation in response to substrate morphology,29r3o may promote fibroblast orientation on these collagen fibers. Fibroblasts tend to orient along the long axis of cylindrical substrates with radii less than 100 In addition, because these fi- bers are fabricated from acid-insoluble collagen, the quaternary structure of collagen may be maintained during processing and recognized as native by fibro- blast integrins. 32

Preliminary data suggest that collagen synthesis was about tenfold greater on ligament analogs than on tissue culture plates. Similarly, Berthod et al.33 found that collagen synthesis by skin fibroblasts was doubled in porous collagen sponges compared to monolayer culture. In contrast, they reported colla- gen synthesis was decreased when fibroblasts were cultured within collagen gels. In monolayer culture, at confluency, collagen synthesis may be limited by cell-cell contacts. In collagen gels, collagen synthesis may be limited by ”biochemical confinement” or by entrapment of newly synthesized collagen near the fibroblast membrane.33 These inhibitory mechanisms apparently are inoperative in the ligament analogs. Furthermore, maintenance of the quaternary struc- ture in acid-insoluble collagen fibers may provide a substratum more conducive to collagen synthesis. Regardless of its cause, increased collagen synthesis by seeded fibroblasts in vitvo may improve the strength and biocompatibility of ligament analog im- plants.

COLLAGEN SYNTHESIS

Figure 6 . Collagen synthesis (radiolabeled proline incor- poration at day 4) by ACL and PT fibroblasts on tissue culture plates and ligament analogs in vitro (mean values with standard deviations). Collagen synthesis on ligament analogs was significantly greater ( p < 0.05) than on tissue culture plates. ACL and PT fibroblasts synthesized similar amounts of new collagen.

Page 7: Development of fibroblast-seeded ligament analogs for ACL reconstruction

LIGAMENT ANALOGS FOR ACL RECONSTRUCTION 1369

Effect of fibroblast origin on fibroblast function

Recent studies have shown that all tendon and iig- ament fibroblasts do not behave the same. For exam- ple, Amiel et al.3436 have correlated poor healing in the ACL with decreased migration,34 pr~liferation,~~ and collagen synthesis36 in ACL fibroblasts compared to medial collateral ligament (MCL) fibroblasts. Be- cause the ACL is often reconstructed using PT tissue, we are interested in comparing the behavior of ACL and PT fibroblasts. Furthermore, to optimize the function of ligament analogs, we need to understand the effects of fibroblast type, substrate, and other fac- tors on fibroblast attachment, proliferation, and col- lagen synthesis.

During the seeding process, similar numbers of ACL and PT fibroblasts attached to the ligament an- alogs. Afterwards, PT fibroblasts proliferated more rapidly than ACL fibroblasts on ligament analogs. Similarly, ACL fibroblasts migrate and proliferate slowly compared to MCL fibroblast^.^^ Based on pre- vious re orts comparing ACL and MCL collagen syn- the~is,~'we also expected that collagen synthesis might be decreased in ACL fibroblasts. However, ACL and PT fibroblasts synthesized similar amounts of collagen per cell. ACL and PT fibroblast behavior in vitro may be influenced by specific interactions with the substrate or by media-soluble growth factors. Fur- ther studies are required to determine which vari- ables most strongly influence fibroblast behavior on ligament analogs in vitro.

Potential applications of ligament analogs

Ligament analogs were designed primarily for use as viable, tissue-engineered implants for ACL recon- struction. Although this approach already is under investigation for skin" and cartilage'' repair, implan- tation of viable cells within a resorbable matrix is a novel concept for ligament reconstruction. Ligament analogs also may be useful to study fibroblast func- tion, wound healing, and fibroblast response to chemical and physical stimuli in vitro.

Preliminary implantation studies (unpublished) suggest that autogenous fibroblast-seeded ligament analogs remain viable for at least 24 h after implan- tation in the knee joint (see Fig. 7). We plan to mea- sure ligament analog viability as a function of fibro- blast type and implantation time in the knee. PT fi- broblast-seeded ligament analogs may not be capable of long-term survival in the knee joint because PT allografts37 and autografts (free3' and vas~ularized~~) undergo necrosis shortly after implantation. Liga-

Figure 7. Light microscopic examination of an autogenous fibroblast-seeded ligament analog after 24 h implantation in the knee joint of a rabbit. (A) Cross section shows intact collagen fibers (CF) surrounded by seeded ACL fibroblasts and matrix deposited in vituo. Hematoxylin and eosin stain. Bar = 50 pm. (€3) Unfixed collagen fiber teased from ex- planted ligament analog and examined under fluorescent light microscopy. The seeded fibroblasts remained viable (arrow) and remained attached to the collagen fiber (CF) scaffold. Bar = 50 km.

ment analogs seeded with fibroblasts of intrasynovial origin, however, may survive and proliferate in the knee joint since they are accustomed to a relatively "fragile" blood supply.4o In addition, synovial fluid may provide nutrition to implanted ligament analogs in the early postoperative period. Synovial fluid has been shown to provide nutrition to fibroblasts in vitro4' and in vivo during repopulation of autografts prior to revascularization.38~4z4

SUMMARY

We developed viable ligament analogs potentially useful for ACL reconstruction and for in vitro studies

Page 8: Development of fibroblast-seeded ligament analogs for ACL reconstruction

1370 DUNN ET AL.

of fibroblast function. Collagen fiber scaffolds are nat- ural, three-dimensional substrates that support fibro- blast attachment and proliferation and promote col- lagen synthesis. We currently are optimizing liga- ment analogs for ACL reconstruction by varying the collagen crosslinking method,45 fibroblast type, fibro- blast seeding density, incubation conditions, and in- cubation time in vitro. Based on recent studies in our l ab~ra to ry ,~~ We now use ultraviolet irradiation (254 nm) to crosslink collagen fibers.

This work was supported by grants to MGD from the National Institutes of Health (R29-AR42230), the Whitaker Foundation (90-0195), the Musculoskeletal Transplant Foundation, and the New Jersey Center for Biomaterials and Medical Devices. The authors thank Lisa Bellincampi for technical assistance and Kevin S. Weadock, Ph.D., for helpful discussions.

References

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

F. H. Fu, C. D. Harner, D. L. Johnson, M. D. Miller, and S. L.-Y. Woo, ”Biomechanics of knee ligaments,” I. Bone It. Surg, 75A, 1716-1727 (1993). R. J . Johnson, B. D. Beynnon, C. E. Nichols, and F. H. Renstrom, ”The treatment of injuries of the an- terior cruciate ligament,” J. Bone Jt. Surg., 74A, 140- 151 (1992). J. Gillquist, ”Repair and reconstruction of the ACL: Is it good enough?,” Arthroscopy, 9, 68-71 (1993). P. 0. Newton, S. Horibe, and S. L.-Y. Woo, “Exper- imental Studies on Anterior Cruciate Ligament Au- tografts and Allografts. Mechanical Studies,” in Knee Ligaments. Structure, Function, Injury, and Repair, D. Daniel, W. Akeson, J . O’Connor (ed.), Raven Press, New York (1990), pp. 389-399. S. J. O’Brien, R. F. Warren, H. Pavlov, R. Panariello, and T. L. Wickiewicz, ”Reconstruction of the chroni- cally insufficient anterior cruciate ligament with the central third of the patellar ligament,” J. Bone J t . Surg., 73A, 278-286 (1991). J. A. Nemzek, S. P. Arnoczky, and C. L. Swenson, ”Retroviral transmission by the implantation of con- nective-tissue allografts,” J. Bane Jt. Surg., 76A, 1036- 1041 (1994). B. E. Buck, T. I. Malinin, ”Human bone and tissue allografts. Preparation and safety,” Clin. Orthop. Rel. Res., 303, 8-17 (1994). F. H. Silver, A. J . Tria, J. P. Zawadsky, and M. G. Dunn, ”Anterior cruciate ligament replacement: A re- view,” J. Long-Term Effects M e d . Impl., 1, 135-154 (1991). M. G. Dunn, “Biomechanics of Ligament Reconstruc- tion,” in Ligaments of the Knee, A. J. Tria, Jr., (ed.), Churchill Livingstone, New York, 1995, pp. 275-286. R. Langer and J. P. Vacanti, “Tissue engineering,” Science, 260, 920-926 (1993). S. T. Boyce, D. G. Greenhalgh, R. J. Kagan, T. Hous- inger, J . M. Sorrell, C. P. Childress, M. Rieman, and G. D. Warden, “Skin anatomy and antigen expression after burn wound closure with composite grafts of cul- tured skin cells and biopolymers,” Plast. Reconstr. Surg., 91, 632-641 (1993). C. A. Vacanti, R. Langer, B. Schloo, and J. P. Vacanti, ”Synthetic biodegradable polymers seeded with chon-

13.

14.

15.

16.

17.

18.

19.

20.

21.

22.

23.

24.

25.

26.

27.

28.

drocytes provide a template for new cartilage forma- tion,” Plast. Reconstr. Surg., 88, 75S759 (1991). Y. P. Kato, D. L. Christiansen, R. A . Hahn, S. J . Shieh, J. D. Goldstein, and F. H. Silver, ”Mechanical properties of collagen fibers: A comparison of recon- stituted and rat tail tendon fibers,” Biomaterials, 10, 38-42 (1989). M. G. Dunn, P. N. Avasarala, and J. P. Zawadsky, ”Optimization of extruded collagen fibers for ACL re- construction,“ J. Biomed. Muter. Res. , 27, 1545-1552 (1993). J. D. Goldstein, A. J. Tria, J. P. Zawadsky, Y. P. Kato, D. Christiansen, and F. H. Silver, ”Development of a reconstituted collagen tendon prosthesis,” J. Bone Jt. Surg., 71A, 1183-1191 (1989). Y. P. Kato, M. G. Dunn, J. P. Zawadsky, A. J. Tria, and F. H. Silver, ”Regeneration of Achilles tendon with a collagen tendon prosthesis: Results of a one year implantation study,“ 1. Bone It. Surg., 73A, 561- 574 (1991). A. J. Wasserman, Y. P. Kato, D. Christiansen, M. G. Dunn, and F. H. Silver, ”Achilles tendon replacement by a collagen fiber prosthesis: Morphological evalua- tion of neotendon formation,” Scanning Microscopy, 111, 118>1200 (1989). M. G. Dunn, A. J . Tria, J . R. Bechler, R. S. Ochner, J . P. Zawadsky, Y. P. Kato, and F. H. Silver, “Ante- rior cruciate ligament reconstruction using a compos- ite collagenous prosthesis. A biomechanical and his- tologic study in rabbits,“ A m . 1. Sports Med. , 20, 507- 515 (1992). K. S. Weadock, R. M. Olson, and F. H. Silver, “Eval- uation of collagen crosslinking techniques,” Biomat. Med. Dev. Artif. Organs, 11, 293-318 (1984). NIH Guide for the Care and Use of Laboratory Ani- mals, NIH Publication 85-23, National Institutes of Health, Bethesda, MD (1985). A. Ramachandrula, K. Tiku, and M. L. Tiku, “Tripep- tide RGD-dependent adhesion of articular chondro- cytes to synovial fibroblasts,” J. Cell Sci., 101, 859-871 (1992). D. Amiel, C. Frank, F. Harwood, J. Fronek, and W. H. Akeson, “Tendons and ligaments: A morpho- logical and biochemical comparison,” J. Orthop. Res., 1, 257-265 (1984). D. Huang, T. R. Chang, A. Aggarwal, R. C. Lee, and H. P. Ehrlich, “Mechanisms and dynamics of me- chanical strengthening in ligament-equivalent fibro- blast-populated collagen matrices,” Ann. Biomed. Eng., 21, 289-305 (1993). F. R. Noyes and E. S. Grood, “The strength of the anterior cruciate ligament in humans and rhesus mon- keys,” J. Bone I t . Surg., 58A, 1074-1082 (1976). F. H. Silver and G. Pins, “Cell growth on collagen: A review of tissue engineering using scaffolds contain- ing extracellular matrix,“ J. Long-Term Effects Med. lm- plants, 2, 67-80 (1992). T. Elsdale and J. Bard, ”Collagen substrata for studies on cell behavior,” I. Celi Biol., 54, 626637 (1972). E. Bell, H. P. Ehrlich, S. Sher, C. Merrill, R. Sarber, B. Bull, T. Nakatsuji, D. Church, and D. J. Bittle, “De- velopment and use of a living tissue equivalent,” Plast. Reconstr. Surg., 67, 386391 (1981). T. Maruguchi, Y. Maruguchi, S. Suzuki, K. Matsuda, K .4 . Toda, and N. Isshiki, “A new skin equivalent: Keratinocytes proliferated and differentiated on colla- gen sponge containing fibroblasts,” Plast. Reconstr. S U Y ~ . , 93, 537-544 (1994).

Page 9: Development of fibroblast-seeded ligament analogs for ACL reconstruction

LIGAMENT ANALOGS FOR ACL RECONSTRUCTION 1371

29.

30.

31.

32.

33.

34.

35.

36.

37.

R. Singhvi, G. Stephanopoulos, and D. I. C. Wang, ”Review: Effects of substratum morphology on cell physiology,“ Biotech. Bioeng., 43, 764-772 (1994). G. A. Dunn and J. P. Heath, “A new hypothesis of contact guidance in tissue cells,” E x p . Cell Res., 101,

Y. A. Rovensky and V. I. Samoilov, “Morphogenetic response of cultured normal and transformed fibro- blasts, and epitheliocytes, to a cylindrical substratum surface,” J. Cell Sci., 107, 12551263 (1994). C. E. Klein, D. Dressel, T. Steinmayer, C. Mauch, B. Eckes, T. Krieg, R. B. Bankert, and L. Weber, ”Inte- grin a2pl is upregulated in fibroblasts and highly ag- gressive melanoma cells in three-dimensional collagen lattices and mediates the reorganization of collagen I fibrils,” J. Cell Biol., 115, 1427-1436 (1991). F. Berthod, D. Hayek, 0. Damour, and C. Collombel, ”Collagen synthesis by fibroblasts cultured within a collagen sponge,” Bionzuteriuls, 14, 749-754 (1993). M. H. Geiger, M. H. Green, A. Monosov, W. H. Ake- son, and D. Amiel, “An in nitro assay of anterior cru- ciate ligament (ACL) and medial collateral ligament (MCL) cell migration,” Connect. Tiss. Res., 30, 215224 (1994). C. N. Nagineni, D. Amiel, M. H. Green, M. Ber- chuck, and W. H. Akeson, ”Characterization of the intrinsic properties of the anterior cruciate and medial collateral ligament cells: An in vitro cell culture study,” J. Orthop. Res., 10, 465-475 (1992). M. E. Wiig, D. Amiel, M. Ivarsson, C. Nagineni, C. D. Wallace, and K.-E. Arfors, ”Type I procollagen mRNA levels in normal and early healing of the me- dial collateral ligament and the anterior cruciate liga- ments in rabbits: An in situ hybridization study,“ 1. Orthop. Res., 9, 374-382 (1991). D. W. Jackson, T. M. Simon, P. R. Kurzweil, and M. A. Rosen, “Survival of cells after intraarticular transplantation of fresh allografts of the patellar and

1-14 (1976).

38.

39.

40.

41.

42.

43.

44.

45.

anterior cruciate ligaments. DNA-probe analysis in a goat model,” J. Bone Jt . Surg., 74A, 112-118 (1992). J. B. Kleiner, D. Amiel, F. L. Hanvood, and W. H. Akeson, ”Early histologic, metabolic, and vascular as- sessment of anterior cruciate ligament autografts,” J. Orthop. Res., 7, 235242 (1989). D. L. Butler, E. S. Grood, F. R. Noyes, M. L. Olm- stead, R. B. Hohn, S. P. Arnoczky, and M. G. Siegel, “Mechanical properties of primate vascularized vs. nonvascularized patellar tendon grafts: changes over time,” J. Orthop. Res., 7, 6&79 (1989). S. P. Arnoczky, ”Blood supply to the anterior cruciate ligament and supporting structures, Orthop. CIin. North Am., 16, 15-28 (1985). D. A. Nickerson, R. Joshi, S. Williams, S. M. Ross, and C. Frank, ”Synovial fluid stimulates the prolifer- ation of rabbit ligament fibroblasts in vitro,“ Clin. Or- thop. Rel. Res., 274, 294-299 (1992). L. B. Dahlin, G. Hanff, and R. Myrhage, “Healing of ligaments in synovial fluid,” Scund. 1. Plust. Reconstr. Hand Surg., 25, 97-102 (1991). J. P. Fulkerson, A. Berke, and N. Parthasarathy, ”Col- lagen biosynthesis in rabbit intraarticular patellar ten- don transplants,“ A m . J. Sports M e d . , 18, 249-253 (1990). D. Amiel, W. H. Akeson, S. Renzoni, F. Harwood, and M. Abel, ”Nutrition of cruciate ligament recon- struction by diffusion,” Acta Orthop. Scund., 57, 201- 203 (1986). K. S. Weadock, E. J . Miller, L. D. Bellincampi, J. P. Zawadsky, and M. G. Dunn, ”Physical crosslinking of collagen fibers: Comparison of ultraviolet irradiation and dehydrothermal treatment,” J. Biomed. Muter. Res., 29, 1373-1379, 1995.

Received November 9, 1994 Accepted March 20, 1995