Issue 1, August 2003

Engineering & Applied Sciences

Baseline Creep Characterization of Collagen Fiber Scaffolds

Darryl Athos Dickerson
Tulane University
Advisors: Kay C. Dee and Glen A. Livesay
Tulane University

Abstract

High ligament injury rates and lack of a sufficient repair method has led to interest in ligament tissue engineering. Several protocols exist that outline attempts to engineer ligaments in vitro using various scaffolding materials, crosslinking methods, sterilization techniques, cell seeding, and other necessary procedures. The first step in such applications is to determine baseline properties of the scaffold. The baseline properties of a collagen fiber scaffold were studied for future use in ligament tissue engineering. This study assessed the efficacy of two sterilization methods for the scaffolds and determined the effects of sterilization and time under cell culture conditions on the creep properties of these scaffolds. Both ultraviolet irradiation and ethanol immersion proved to be effective in keeping the collagen fibers sterile throughout the culture period. It was found that the creep properties of collagen fiber scaffolds were not significantly affected by either sterilization or time under cell culture conditions. The creep properties of scaffolds did not resemble those of natural ligaments. These results provide a standard method for collagen fiber scaffold sterilization and give baseline creep characterization that will allow future development of a tissue-engineered ligament from collagen fiber scaffolds.

Introduction

Soft tissue injuries account for 45% of the 33 million musculoskeletal injuries that occur each year . Healing of such injuries varies greatly with the site and severity of the injury. One of the most commonly occurring soft tissue injuries is rupture of the anterior cruciate ligament (ACL) of the knee, which displays little if any inherent capacity for healing. It is estimated that 200,000 ACL injuries occur in the United States annually . Because the ACL is critical for knee joint movement and stability, untreated ACL injury can lead to meniscal damage, osteoarthritis, and other injuries that further compromise motion of the knee . Currently, surgical reconstruction using autografts of the patellar tendon is the most prescribed method of ACL repair (Hubbell and Schwartz 2001). However, concerns regarding donor site morbidity and long-term joint stability have motivated interest in tissue-engineered analogues.

Tissue engineering is a diverse field that utilizes engineering, chemistry, and life sciences methods in an effort to replicate the functions of living tissues . The goal of tissue engineers is to grow cells from a patient within a selected scaffolding material to form in vitro ligamentous tissue that can then be implanted as a replacement for injured tissue. Tissue engineering scaffolding materials should be biocompatible with and have a structure similar to that of the native tissue of interest, promote cell growth and proliferation, and allow the eventual development of mechanical properties similar to the tissue. Collagen fibers are a logical scaffold choice for tissue engineering of ligament tissue as collagen is the major structural protein of mammalian tissues. It has a triple-helical molecular structure that forms a fibrillar network and is the major structural component of ligaments. Collagen can be extracted from varied tissue sources and processed into fibers. Such reconstituted fibers have been examined as scaffolds for ligament tissue engineering . Previous studies have tested for biocompatibility and some mechanical properties of reconstituted fibers; however, full mechanical characterization of these collagen fiber scaffolds is yet to be undertaken.

Implementation of any tissue-engineered ligament analogue requires rigorous characterization of its material and mechanical properties. Elastic and viscoelastic properties are key determinants of the performance of materials under loading. One principal parameter of viscoelasticity is creep, the time-dependent elongation behavior exhibited by a material when subjected to constant loading. Tensile creep characterization of a tissue analogue is important since creep behavior of a ligament contributes to basic joint performance and overall joint stability. Excessive ligament creep, for example, can decrease the load-bearing ability of the ligament, causing an increase in knee laxity and abnormal knee motion. Processes necessary to the development of an engineered tissue (e.g., material sterilization) might affect creep properties of the tissue. The major objective of this study was to determine the effects of sterilization and maintenance under standard cell culture conditions on the creep behavior of fabricated collagen fiber scaffolds.

Two parameters of creep were studied: equilibration time and equilibrium strain. Equilibration time was determined to be the time at which the elongation slowed to a point to be almost undetectable, and equilibrium strain was measured as the strain at that time. It was hypothesized that standard sterilization techniques of ethanol immersion and ultraviolet irradiation should be effective in sterilization and should not have a significant effect on the creep parameters. However, the materials in cell culture medium should change their creep properties as the scaffolds spend more time under culture conditions. The information obtained from this study can be used to establish baseline properties of such scaffolds, necessary for comparative studies and future development of collagen fiber-based ligament analogues.

Materials and methods

Collagen Scaffold Fabrication

Insoluble Type I collagen from bovine Achilles’ tendons was used to make collagen fibers by extrusion methods previously established (Kato, 1989 #34; Skok, 2000 #47). A suspension of 1% (w/v) collagen in 0.005 N HCl was mixed in a blender at low speed for 4 minutes. The suspension was allowed to settle for 10 minutes, then had an 4 additional minutes of blending at low speed. The collagen suspension was then centrifuged at 5000 rpm for 5 minutes to remove trapped air, and stored at 8° C for up to 3 days before extrusion.

To create reconstituted collagen fibers, the collagen suspension was extruded through polyethylene tubing (inner diameter = 0.04 in) at a flow rate of 0.07 mL/min into a buffer bath composed of 2.75 g N-tris(hydroxymethyl)methyl-2-aminoethanesulfonic acid, 3.16 g NaCl, 1.70 g Na₂PO₄, and 400 mL of deionized water at pH 7.5 and 37° C. Newly formed fibers were soaked in the buffer for 45 minutes and were then transferred to a room temperature bath of 95% ethanol for 4 hours. The fibers were then quickly rinsed in deionized water and allowed to dry under tension overnight.

The collagen fibers were crosslinked in a 1% (w/v) solution of 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide in deionized water at room temperature for 24 hours. The fibers were rinsed in deionized water for 1 hour and then dried under tension. Scaffolds (approximate length = 51 mm) consisted of 10 individual collagen fibers tied at the ends with surgical silk.

Sterilization and Culture

Collagen scaffolds were sterilized either by immersion in 70% ethanol for 18 hours or by irradiation in ultraviolet (UV) light for 18 hours. After sterilization, the scaffolds were soaked in sterilized deionized water for 10 minutes, vigorously rinsed in sterilized deionized water, and placed individually into the wells of six-well tissue culture plastic plates (Falcon, Becton Dickson, Franklin Lakes, NJ). Three mL of Dulbecco’s Modified Eagle Medium supplemented with 10% fetal bovine serum (Invitrogen Corp, Carlsbad, CA) was added to each of the wells. The plates were then placed in a standard tissue culture incubator (humidified, HEPA-filtered, 5% CO₂/95% air) to mimic cell culture conditions. The culture medium was changed every two days. Control scaffolds (n = 3 per sterilization method) were not placed in culture, but were tested immediately following sterilization. Experimental scaffolds were cultured for 1 day (n = 3 per sterilization technique) or 7 days (n = 3 per sterilization technique), rinsed in sterile deionized water, and transferred into sterile phosphate-buffered saline (PBS) for storage at 37° C for 30 minutes prior to mechanical testing.

Mechanical Testing

Figure 1. Creep-testing device. This custom-designed creep-testing device consists of a Plexiglas base structure with a linear motion slide (A) mounted on the vertical face of the base. At the top of the structure is a two-pulley system (B) leading to a platform (C) where the weight used for the test is rested. The specimen to be tested is clamped at D, below the linear motion slide.

Results

Sterility Under Cell Culture Conditions

Phenol red (a pH indicator) in the culture medium remained red throughout the duration of the study for all collagen scaffolds. Microscopic inspection revealed no infection. Thus, both UV-sterilized and ethanol-sterilized collagen scaffolds maintained sterility under cell culture conditions.

Equilibration Time

The mean T_eq for the UV-sterilized scaffolds was 31.41 ± 0.53 sec (mean ± S.D.), 32.89 ± 0.93 sec, and 31.27 ± 0.42 sec for scaffolds tested immediately after sterilization, after 1 day, and after 7 days under culture conditions, respectively. The mean T_eq for ethanol-sterilized scaffolds was 30.02 ± 1.33 sec, 31.22 ± 1.30 sec, and 31.36 ± 0.92 sec for the same time periods. Time maintenance under cell culture conditions did not affect the T_eq of the scaffolds (Figure 2). ANOVA analysis indicated some difference in the effect of UV sterilization vs. ethanol sterilization (p = 0.0499). However, paired t-tests showed no statistically significant difference in the T_eq of the UV-sterilized scaffolds vs. the ethanol-sterilized scaffolds.

Discussion

Collagen fiber prostheses have shown promise for use in ligament tissue engineering . However, before implementation, the elastic and viscoelastic properties of the engineered tissue must be characterized to ensure that these properties mimic those of native tissue. Viscoelasticity is the time- and history-dependent mechanical behavior exhibited by ligaments and other biological tissues . Viscoelastic behavior is described by stress relaxation and creep. While many studies tend to focus on stress relaxation, creep appears to be more relevant to physiological loading conditions. Daily activity produces static and cyclic loading of ligaments . It has been hypothesized that such loading situations generate creep responses in vivo . Thus, in engineering a ligament replacement, it is important that the replacement tissue mimics the creep properties of native ligament tissue.

The specific objective of this study was to evaluate the creep properties of these collagen fiber scaffolds after sterilization and time under culture conditions. Using uniaxial static loading, parameters of creep behavior defined as equilibrium strain and equilibration time were studied. For an engineered ligament to function properly, the equilibrium strain must lie within the normal physiological range; otherwise, the ligament could be too long or too short to properly guide and restrict motion at the joint. Equilibrium strain in these collagen fiber scaffolds did not appear to be affected by either sterilization technique or time under cell culture conditions. Equilibrium strain of sterilized collagen fiber scaffolds appeared to be larger than the range of normal ligament under similar loading, which is approximately 0.03-0.05 . This also seems to correlate well with data that show the collagen fiber scaffolds are not as stiff as normal ligaments.

There was no distinct trend in the data to indicate that short (1 day) or long (7 days) periods under cell culture conditions had any effect on the equilibration time of the scaffolds. Statistical analysis did suggest that there was some difference in equilibration time between the ethanol-sterilized and UV-sterilized scaffolds. On average, ethanol-sterilized scaffolds appeared to reach equilibrium more quickly than UV-sterilized scaffolds. Observed creep of these collagen fiber constructs occurs very rapidly in comparison to actual ligaments. Equilibration time was less than 35 seconds for all scaffolds tested; however, previous studies indicate that creep continues for more than 20 minutes in native ligaments . The relative speed with which the fiber scaffolds reach equilibrium may provide some insight as to the causes of creep behavior in ligaments. It has been suggested that viscoelastic behavior is the result of the interactions of collagen and extracellular matrix components . Other studies support the supposition that creep properties of ligaments are, in part, the result of fiber recruitment . The present study confirms that there must be some causal mechanism for viscoelastic creep behavior in ligaments beyond the presence of collagen fibers.

The results obtained from this experiment give insight into the baseline properties of the collagen scaffolds proposed for use in the development of engineered ligament analogues. Beyond tissue engineering, the relatively simple construction of collagen fiber scaffolds could be used to examine and possibly elucidate causal mechanisms for creep behavior. UV irradiation and ethanol immersion can be used as sterilization techniques for these collagen scaffolds in multiple applications. In addition, the fact that maintenance under cell culture conditions does not affect creep properties of collagen fiber scaffolds indicates that controlled experiments can be performed, manipulating other variables without concern over the influence of scaffold degradation due to culture medium. While the collagen scaffolds alone do not show the viscoelastic creep characteristic of ligaments, they do form fibrous constructs that can be used as a starting point for more advanced tissue engineering efforts. For example, it might be possible to employ cell seeding and extracellular matrix emulation in an effort to produce more ligament-like viscoelastic response in the scaffolds. Fundamental characterization studies remain important to the continued development of tissue-engineered ligaments.

Acknowledgements

The author would like to acknowledge Dr. Henry Bart, Dr. Calvin Mackie, Jannie Price, Dr. Eric Nauman, Eileen Gentleman, Andrea Lay, James Raasch, and Inchan Youn for support of this work. This work was funded by the Louis Stokes Louisiana Alliance for Minority Participation and NSF Career #0093969.