Human bioartificial muscles (HBAMs) are tissue engineered by suspending muscle cells in collagen/MATRIGEL, casting in a silicone mold containing end attachment sites, and allowing the cells to differentiate for 8 to 16 days. The resulting HBAMs are representative of skeletal muscle in that they contain parallel arrays of postmitotic myofibers; however, they differ in many other morphological characteristics. To engineer improved HBAMs, i.e., more in vivo-like, we developed Mechanical Cell Stimulator (MCS) hardware to apply in vivo-like forces directly to the engineered tissue. A sensitive force transducer attached to the HBAM measured real-time, internally generated, as well as externally applied, forces. The muscle cells generated increasing internal forces during formation which were inhibitable with a cytoskeleton depolymerizer. Repetitive stretch/relaxation for 8 days increased the HBAM elasticity two- to threefold, mean myofiber diameter 12%, and myofiber area percent 40%. This system allows engineering of improved skeletal muscle analogs as well as a nondestructive method to determine passive force and viscoelastic properties of the resulting tissue.
- muscle hypertrophy
- collagen gel
- force transducer
during embryogenesis,the development of human skeletal muscle is coordinated with the development of many different organ systems, especially the skeletal system. Many scientists believe bone elongation (∼2 mm/wk during human embryogenesis) (2) directs the initial phases of muscle organogenesis (34). The continuous passive tension applied to skeletal muscle by bone growth during both embryogenesis and neonatal development influences muscle weight and length (34) and myofilament organization (19). A role for mechanical forces on skeletal muscle growth continues into adulthood. Stretch and exercise are responsible for many in vivo muscle adaptations after early development, including the regulation of protein synthesis and degradation rate (17, 18), total RNA and DNA content (14, 17, 34), and protein accumulation (17, 34). Increased work load or exercise changes mature muscle morphology by increasing both the number and diameter of muscle fibers (10, 20, 21, 23, 28). In addition, mechanical regulation of skeletal muscle responses to growth factors such as insulin sensitivity (12), heat shock protein 72 gene expression (11), and nitric oxide release (35) occur in vivo.
It has been shown that mechanical stimulation of tissue-cultured skeletal muscle cells causes changes similar to those seen when in vivo skeletal muscle is mechanically loaded. Studies have indicated that mechanical stimulation of monolayer skeletal muscle cultures affects gene regulation (5), endogenous protein expression (14, 32, 34), protein accumulation (39), protein localization (41), and metabolic activity (22) in a manner consistent with changes that occur in vivo. Mechanical stimulation also increases cellular proliferation (39), myofiber organization (6, 36, 41), and extracellular matrix (ECM) composition (41). Repetitive stretch/relaxation of avian and rodent skeletal muscle cells by patterns that simulate in vivo skeletal growth and exercise alters the structural organization of the constructs. Unidirectional stretch (36–40 h) followed by repetitive stretch (2–3 wk) of avian bioartificial muscles was shown to lengthen, orient, and organize the myofibers compared with static cultures (41). Stress loading also improves the orientation and density of rodent myofibers cultured in collagen gels (31).
Although these studies have shown improvements in structure and function of cultured avian and rodent muscle with mechanical stimulation, similar studies have not been reported for human skeletal muscle cells. Primary human skeletal muscle cells have been isolated from healthy adults and diseased populations, expanded in tissue culture, and statically tissue engineered into human bioartificial muscles (HBAMs) (M. Nackman, A. Lee, S. McGuire, J. Shansky, C. Powell, J. Hennessey, L. Krivickas, B. Biller, and H. Vandenburgh, unpublished observation; Ref. 33). These HBAMs are representative of in vivo skeletal muscle in that they contain a parallel array of fused postmitotic muscle fibers expressing sarcomeric contractile proteins. However, they fall short of actual skeletal muscle in many aspects, including small diameter muscle fibers, low myofiber density, and excessive ECM. To be used for structural/functional skeletal muscle repair or replacement, HBAMs need to be engineered with properties more similar to in vivo skeletal muscle. In an attempt to improve the structure of tissue engineered HBAMs, we designed a novel computer-controlled device to repetitively, mechanically load HBAMs during both formation and long-term development while recording real time forces in the micronewton (μN) to millinewton (mN) range. This report describes the new device, passive forces measured within tissue-engineered human skeletal muscle, and the morphological and viscoelastic changes observed with repetitive mechanical stimulation of the constructs.
MATERIALS AND METHODS
The mechanical cell stimulator.
A mechanical cell stimulator device, version 4 (MCS4), was designed and built to mechanically load six HBAMs grown in silicone rubber tissue molds positioned in the wells of commercially available sixwell (35-mm diameter) tissue culture dishes (Fig.1 A). Each well contains two 2-mm-diameter stainless steel pins suspended ∼1 mm off the bottom of each well and 20 mm apart, acting as artificial tendons for HBAM attachment. One pin in each well attaches to a stepper motor, allowing for direct mechanical stimulation of HBAMs, and the other pin is either immobile (four wells) or attaches to a force transducer (two wells). A total of 5 mm of mechanical stretching, i.e., 25% change in length, is possible before the end of the mold is reached. National Instruments hardware (NU Drive 4SX-411 4-Axis Step ValueMotion Drive, PC-STEP-4CX 4-Axis Closed Loop Step Controller) drives a size 17 1.8° linear actuator stepper motor (Eastern Air Devices) with linear accuracy of ∼50 μm. LabVIEW software, National Instruments′ graphics-based programming tool, is used to coordinate instrument control, data acquisition, and data presentation. A LabVIEW virtual instrument allows the user to define unidirectional stretch amplitude and velocity, repetitive stretch amplitude (which also may be combined with unidirectional stretch amplitudes), velocity (up to 0.5 mm/s), pattern of stretch (number of stretch/relaxation per set and number of sets/cycle), and duration of rest periods. An additional feature of the MCS4 is that it contains sensitive force transducers [Omega Scientific ultra-low capacity bending beam load cell, model LCUB-002G, 2-g (20 mN) full-scale capacity, ∼5 mg (50 μN) sensitivity] attached to one end of two of the six HBAMs (LC in Fig.1 A). The force transducer is also controlled with LabVIEW software and a National Instruments data acquisition board (PCI-6023/24/25E) and signal conditioner (SCXI-1000). These force transducers were calibrated by applying known loads to the transducers with a force generator (8), reading the outgoing voltage, and generating a calibration curve. Real-time passive force measurements were made during HBAM formation and during mechanical stimulation at a rate of up to 100 samples/s, averaging 100 samples/data point.
The individual components of the MCS4 were cleaned and sterilized before use. The MCS4 was assembled in a sterile biosafety laminar flow hood <12 h before use and incubated in a 37°C incubator to equilibrate the force transducers to the warm, 80% humidified environment. Equilibration of the transducer was usually complete after 6 h in this environment, as indicated by stationary force readings (data not shown).
Dynamic tissue engineering of HBAMs.
HBAMs were tissue engineered in a manner similar to that previously described (33). Cryopreserved primary human skeletal muscle cells isolated by needle biopsy from the vastus lateralis (33) of healthy adults ages 35–50 were thawed and expanded in 10-cm tissue culture dishes with myoblast growth medium [MGM: SkGM (Clonetics, catalog no. CC-3161) plus 15% (vol/vol) FBS (Sigma, catalog no. F2442)]. Biopsies were performed on volunteers according to procedures approved by the Institutional Clinical Review Board. Cells from four different biopsies were used in these studies and averaged 60–80% myogenic based on desmin positive staining. The cells were always used before 22 population doublings, a time after which growth and differentiation potential diminishes (33). After expansion, the cells were suspended by using trypsin/EDTA, pelleted by centrifugation at 800 g, and resuspended in chilled MGM containing 0.8 mg/ml type I rat tail collagen (Collaborative Biomedical Products, Bedford, MA, catalog no. 354236) neutralized with 10% (vol/vol) 0.15 N NaOH (1 × 106 cells/770 μl solution). MATRIGEL (Collaborative Biomedical Products, catalog no. 354234), a complex mixture of ECM proteins, growth factors, and glycosylaminoglycans produced by murine carcinoma cells, was added to the cell/collagen suspension at a ratio of 1:6 (vol/vol) MATRIGEL:collagen. The new cell/gel mixture was then carefully transferred into the silicone rubber molds of the MCS4 (900 μl/mold). After the HBAMs were cast, the MCS4 was placed in a humidified 37°C, 5% CO2 incubator, and, after 6 h, the wells were flooded with MGM (9 ml/well). The HBAMs were maintained in MGM for 3 days, fusion medium [FM: high glucose DMEM (GIBCO Life Science, catalog no. 11995-040) supplemented with 2% HS (Sigma, catalog no. H1270), pen (10 U/ml)] for 5 days, and then maintenance medium [MM: high glucose DMEM supplemented with 10% (vol/vol) HS, 5% (vol/vol) FBS, and pen (10 U/ml)] for the remainder of the experiment. Medium changes were made every 2–3 days for the length of the experiments (8–16 days). ECM gels without cells were used as controls.
Within 24 h after casting, the cell/gel mix had contracted and detached from the mold and was held in place only at the two attachment sites (Fig. 1 B). In those experiments that included mechanical stimulation, unidirectional stretch began at 24 h, acting to simulate the bone growth that occurs during embryogenesis. The HBAMs were stretched 500 μm/day for 4 days. The stretch occurred in step intervals of ∼3.5 μm every 10 min (total unidirectional stretch over 4 days of 10% of initial length). A similar unidirectional pattern aligned fusing avian myoblasts grown on an elastic substratum into organized myofibers (41). The HBAMs were held at the new extended length for 3 days. Repetitive stretch/relaxation similar to a pattern found to cause skeletal muscle hypertrophy in monolayer avian skeletal muscle cultures (39) then began on day 8 postcasting. The stretch pattern consisted of three sets of five stretch/relaxations. Each set was separated by ∼30 s of rest, with 28 min of rest after the third set; thus each cycle lasted ∼30 min (Fig. 1 C). The HBAMs were stretched at 5% strain (1 mm) for 2 days (days 8–10), 10% strain (2 mm) for 2 days (days 10–12), and 15% strain (3 mm) for 4 days (days 12–16). HBAMs that were not repetitively strained were used as controls.
Measurement of cell-generated passive force.
The sensitive force transducer connected to the pin around which the HBAM formed was used for isometric force measurements. HBAMs were cast in the MCS4 silicone rubber tissue molds, and real-time sampling of the passive force within nonmechanically stimulated and mechanically stimulated HBAMs was recorded for up to 16 days. ECM gels without cells were used as controls. To determine whether the passive tension in HBAMs was related to the cell's cytoskeleton, HBAMs were treated with an actin depolymerizer (cytochalasin D). Previous work in other laboratories (7, 25) has shown that passive fibroblast force generation in a collagen matrix is attenuated by disrupting the actin filaments. Because this attenuation was reversible when the pharmacological agent was washed out, we also studied reversibility. Nine days after casting, control (nonmechanically stimulated) HBAMs were rinsed with serum-free medium and incubated with 2–5 μM cytochalasin D (Sigma, catalog no. C8273) in serum-free maintenance medium. An injection port in the cover of the MCS4 device allowed for solution exchange without removing the device from the incubator or exposing the force transducers and HBAMs to changes in temperature and humidity. When the cytochalasin D HBAM response plateaued (∼3 h), the HBAMs were rinsed with Earle's balanced salt solution (EBSS), fresh serum-free medium without cytochalasin D was added for 1 h, and then serum-containing maintenance medium was added for 24 h. Passive forces were recorded every 10 min by using the Omega force transducers and LabVIEW software.
To determine the internal stresses elicited by step stretching, control (nonmechanically stimulated) HBAMs were stretched to 5, 10, and 15% strain and held at this strain for 10 min before returning to the culture length (0% strain). After 10 min at the culture length, the strain routine was repeated for a total of four step stretches. Force measurements were recorded immediately after the stretch and every minute for the duration of the experiment.
Measurement of HBAM viscoelastic properties.
The MCS4 device was used to measure HBAM nonfailure material properties of HBAMs mechanically conditioned from day 8 to day 16 postplating by the pattern shown in Fig. 1 C. After 0, 2, 6, and 8 days of repetitive mechanical conditioning, the stretch pattern was paused and the HBAMs were stretched at a rate of 5% strain/min until their tension increased to 2,500 μN and then relaxed to their original length. Passive force was recorded every 250 ms during the testing. This pattern was repeated five to six times with a 10-min rest between tests. After material testing, the repetitive conditioning stretch/relaxation pattern was restarted and applied until the next testing day. The average elastic modulus was calculated by averaging the slope at each point along the ascending portion of the stress/strain hysteresis and compared with time-matched static control HBAMs.
For whole mount sarcomeric tropomyosin staining, HBAMs were fixed with 4% formalin followed by Dent's fixative (1 part DMSO/4 parts 100% ethanol). Localization of sarcomeric tropomyosin was visualized by an overnight incubation with mouse antitropomyosin (Sigma, catalog no. T9283), followed by biotinylated anti-rabbit IgG secondary antibody, avidin-biotinylated reagent coupled to horseradish peroxidase (Vector Laboratories, Burlington, CA, catalog no. PK-6101) and developed with 3,3′-diaminobenzidine (DAB) (Sigma, catalog no. D5905). HBAMs were also stained and visualized in cross section to determine the effect of mechanical stimulation on HBAM morphology. HBAMs were fixed in 4% formalin, and a portion of the formalin fixed HBAM was paraffin embedded and sectioned in 5-μm cross sections. The paraffin cross sections were deparaffinized and then gradually rehydrated from 100% ethanol to dH2O. The slides were stained with an antibody to myosin heavy chain (Sigma, catalog no. M4276) followed by incubation with a biotinylated anti-mouse IgG secondary antibody, avidin-biotinylated reagent coupled to horseradish peroxidase, and developed with DAB. Sections not incubated with biotinylated secondary antibody were used as negative controls. These procedures are detailed in Powell et al. (33).
Stained sections were photographed at various magnifications using a Sony charge-coupled device (CCD) color video camera (DXC-970MD) attached to a Zeiss microscope. The Zeiss KS300 3.0 Image Analysis Software package was used to measure HBAM cross-sectional area, myofiber diameter, and percent HBAM cross-sectional area occupied by myofibers. The software calculated myofiber area percent by dividing the area occupied by myofibers from the total cross-sectional area.
Each experiment was performed with at least three HBAMs in each experimental group and repeated in at least two independent experiments. All data are presented as means ± SE. We analyzed the data using the Student's t-test (SigmaStat software, Jandel Scientific). Differences were statistically significant withP < 0.05.
Morphological changes with mechanical stimulation.
Primary human skeletal muscle cells were initially tissue engineered into HBAMs by suspending the cells in a collagen/Matrigel ECM solution and casting into silicone rubber molds. Internal longitudinal tensions developed within the cell/gel mixture as it dehydrated, causing the formation of a cylindrical HBAM (Fig. 1 B). Gel dehydration and detachment from the molds did not occur in the absence of cells (data not shown). The internal longitudinal tensions were adequate to align the human myoblasts into parallel arrays which then fused into aligned multinucleated myofibers (Fig.2 A). Sarcomeric myosin staining of these statically formed HBAMs indicated that only 2–10% of their cross-sectional area comprised muscle fibers (Fig.2 B) with fiber diameters <10 μm by 16 days in tissue culture. In an attempt to improve our tissue-engineering techniques, we mechanically stimulated the HBAMs, simulating forces associated with muscle organogenesis and exercise. Mechanical conditioning of HBAMs improved HBAM myofiber diameter and area percentage (Fig.3). After 8 days of repetitive stimulation, the mean myofiber diameter significantly increased by 12% from 6.4 to 7.1 μm (P < 0.05, n = 3–4 HBAMs, >2,000 myofibers measured), whereas the myofiber area percent increased by 40% from 7.8 to 10.9% (P < 0.05, n = 3–4 HBAMs, 8–10 cross sections).
Passive tension measurements.
At 6–24 h after casting, the collagen gel dehydrated, causing the HBAMs to detach from the silicone rubber tissue mold and become attached only at the two end attachment sites. This creates cell-generated passive tensions within the cell/gel matrix. The Omega ultra-low capacity bending beam load cell measured the passive tension generated by HBAMs through several weeks in culture (Fig.4). Noncell controls did not detach from the molds or generate any passive tension. However, when maintaining the HBAMs with attached force sensors at 37°C in an 80% humidified CO2 incubator, transducer drift jumps in the transducer signal due to changing the tissue culture medium (arrows in Fig. 4), and other unpredictable events, e.g., incubator door openings, made it difficult to completely characterize the passive tension nondestructively in real time. A perfusion system is under development to reduce these problems. The passive tension at the termination of the experiment was determined by detaching the HBAM from the force transducer and allowing the force reading to return to zero; if the transducer reading after HBAM removal was nonzero, this value was subtracted to obtain the HBAM's final passive tension measurement. HBAMs typically generated 500–1,000 μN of passive tension per 106 cells, and no passive tension was generated in the absence of cells in the gel (Fig. 4). To determine the potential involvement of the cytoskeleton in passive tension, 9-day-old HBAMs were exposed to an actin depolymerizer, cytochalasin D (2–5 μM), in serum-free medium. Passive tension decreased 57 ± 2% over a 3-h incubation at 37°C, with the majority of tension decreasing within the first hour (Fig. 5). After the cytochalasin D was rinsed out, the tension quickly increased in serum-containing medium, leveling off at 66 ± 3% of the original tension. No increase in tension was seen after an hour in serum-free medium (Fig. 5). These results suggest that the passive tension development in the HBAMs is in part related to the cell's cytoskeleton and that the reestablishment of that tension after depolymerization is enhanced by undefined serum factors present in the medium.
Passive force measurements during and immediately after mechanical loading allowed us to measure the force imposed during stretching and the response of the HBAM after mechanical loading. Force measurements were taken while the HBAM was held at a strained length to determine the load imposed by 5–15% stains. Results indicate that there is increasing sensed force with increasing strain (Fig.6), which is two to four times greater than static passive tensions. This sensed load decreased as the HBAM was held at the extended position. To determine the internal tension response of the HBAM to repetitive stimulation, HBAMs were mechanically stimulated with three sets of five stretch/relaxations over 2 min (Fig.1 C), and the passive tension was measured at 8, 18, and 28 min after the end of the stimulation. At 8 min after stretch relaxation, the force was lower than beginning passive force (Fig.7). The passive tension then increased asymptotically until the next loading cycle (30 min after the start of the previous one). This initial drop in tension resulted from slack in the HBAM immediately after shortening, followed by reestablishment of its initial passive tension over 30 min.
Material property measurements.
The elastic modulus of HBAMs mechanically conditioned for 8 days fromday 8 to day 16 in vitro vs. control HBAMs cultured statically for 16 days was determined using nonfailure testing methods as outlined in materials and methods. Mechanically conditioned HBAMs were found to be more elastic than static HBAMs (Fig.8). The elastic modulus of control HBAMs increased over time, whereas the elastic modulus of mechanically conditioned HBAMs stayed constant so that, by day 16, there was a two- to threefold difference between the two groups.
HBAMs can be tissue engineered by suspending primary human muscle cells in a collagen/MATRIGEL ECM. These tissue-engineered muscle constructs resemble in vivo muscle in that they contain parallel arrays of postmitotic muscle fibers (33). However, the morphology of static HBAMs is different from adult in vivo skeletal muscle in several ways. Whereas adult skeletal muscle fiber diameters range from 10 to 100 μm (4, 24), static HBAM muscle fiber diameters range from 3 to 25 μm with a mean of ∼5 μm (Fig. 3). Because muscle fiber hypertrophy occurs during development and exercise training (24, 34), we predicted that repetitive stretching in vitro HBAMs might increase myofiber diameter, as was the case with mechanical stimulation of avian myofiber monolayers (39). To improve HBAM morphology, we designed and built a mechanical cell stimulator (MCS4) to add mechanical loading to the three-dimensional tissue-engineering procedure. The MCS4 was designed to stretch BAMs in a manner that stimulates muscle growth by simulating in vivo long bone growth during embryogenesis and the repetitive muscle loading of exercise. We chose to initially concentrate on a pattern of mechanical loading similar to one that stimulated myofiber hypertrophy in avian myofibers (39) and investigated the effect of this stimulus on myofiber diameter and myofiber area percent. Mechanical conditioning was found to increase mean HBAM myofiber diameter by 12% after 8 days of mechanical stimulation (Fig. 3). The resultant 12% myofiber hypertrophy in HBAMs was comparable to the 16% observed in avian BAMs (39) and to the 10–17% fiber hypertrophy after 12 wk of resistance training in healthy adults (28); however, this in vitro hypertrophy did not increase the myofiber diameter to approach the in vivo skeletal muscle range of 10 to 100 μm. One reason for the small myofiber size may be the lack of innervation. In humans, denervated adult myofibers atrophy to an ∼20-μm mean diameter (30) compared with ∼50-μm mean diameter in innervated muscle. Appropriate growth factors/cofactors and other nutrients may be missing from the tissue culture medium and may also contribute to the small size of the tissue-engineered muscle fibers.
The myofiber distribution of static HBAMs also has limited resemblance to adult skeletal muscle. In vivo skeletal muscle contains closed-packed fibers organized in fasciculi distributed throughout the tissue; HBAMs contain widely dispersed myofibers throughout the construct (Fig. 2). Approximately 90% of in vivo muscle is occupied by muscle fibers (28), whereas only 2–15% of our tissue-engineered (control and mechanically stimulated) HBAMs is occupied by muscle fibers (Fig. 3). The remaining area is ECM. It is likely that the HBAM morphology is a result of poor oxygenation and nutrient availability in the HBAM core because they are avascular. It is generally accepted that cells do not survive unless they are within 250–300 μm from their nutrient source (13). Mechanical stimulation improves the area percent myofibers by 40%, possibly by enhancing nutrient diffusion through the ECM. Average HBAM diameter also decreases during the stretching phase (due to increased length and conservation of volume), which may also allow nutrient penetration to the interior of the structure. Nutrients that have penetrated the HBAM may be pushed toward the center of the construct with mechanical perturbation.
With the use of a stretching pattern that stimulates hypertrophy in in vitro avian skeletal muscle, HBAMs demonstrate significant increases in both myofiber diameter and myofiber density. These changes are in the right direction to improve HBAM morphology but are still far removed from in vivo skeletal muscle. Further improvements may be possible by changing the mechanical stimulation pattern or improving the ECM that the constructs are engineered with. Furthermore, skeletal muscle growth and hypertrophy is known to be due to an interaction of nutrition, hormones, innervation, and exercise (37), and, therefore, a combination of these other factors may be critical for inducing more substantial changes and engineering HBAMs with the necessary structural and functional characteristics for use as repair or replacement skeletal muscle. Future experiments will address these factors.
Many studies have investigated the internal passive tension of cells plated on or in a collagen matrix. A number of recent studies have found that passive tension can be attributed to the cells' acting on the gel to generate the passive forces (7, 9). The primary human muscle cells used in the present work caused gel contraction and dehydration as previously observed with avian and rodent muscle cells (38, 40). This suggests that HBAMs are under passive tension in their normal tissue-cultured state. Fibroblasts grown in a collagen matrix generate passive tensions which have been quantitated. During the first 6–36 h after plating, the cell/gel mixture rapidly contracts, generating forces in the range of 0.1 and 1.0 g (1,000–10,000 μN) per 106 cells, depending on the cell source (7, 25, 27). The Omega load cell in the MCS4 allowed measurement of the passive tension at intermittent time points over days to weeks in the CO2 incubator; ∼100 mg (1,000 μN) of tension/106 cells were found to develop over the first 3 days after plating (Fig. 4). Perturbations in the force readings corresponding to culture medium changes (possibly due to disconnecting the electrical wires necessary for the force readings, a temporary change in the cells' environment when removed from the incubator, or mechanical perturbation of the force transducer) prevented us from making reliable, real-time, long-term measurements with this set up. Other groups have not reported similar problems, but their data was collected for <4 days, and, therefore, culture medium changes appear to not have been performed. Real-time, long-term measurements were not obtained; however, the tension at the end point of each experiment was obtained by determining the drift of the transducers, allowing us to compare the passive force in HBAMs to those of other tissue-engineered constructs. Although we were able to compare the passive tension in HBAMs to that in other cell/gel constructs based on tension per cell number, we were unable to directly compare it to other tissue-engineered skeletal muscle constructs (i.e., myooids; Refs. 8 and 26) because passive tensions in those studies were not normalized to cell number. HBAMs and myooids do have similar passive tension when normalized for cross-sectional area (∼2 mN/mm2 in HBAMs compared with 0.8 to 6 mN/mm2in rodent myooids; Refs. 8 and 26). A primary goal for future research is to develop more robust force transducers for tissue-engineering purposes.
In 1979, Bell attributed the passive tension in cell/gel structures to cells pulling against the collagen fibrils (1). Recent data more specifically connect the tension in fibroblast-contracted gels to the cellular cytoskeleton (3, 7, 25). Cytoskeleton disruption with pharmacologic agents such as colcemid (7), cytochalasin B (7), and cytochalasin D (25) resulted in 100% decrease in passive tension. When HBAMs were treated with cytochalasin D, the force measurements rapidly decreased ∼70%, supporting a role for the cytoskeleton in passive force. A higher concentration or longer period of incubation with cytochalasin D may be required for complete cytoskeletal disruption and 100% loss of passive tension in the HBAMs. Also, other components in the cells or their matrix (e.g., microtubules, actin-myosin sarcomeric structure, collagen gel, or MATRIGEL) may be responsible for a portion of the HBAM passive tension.
Nonfailure tensile testing of HBAMs showed that mechanically conditioned HBAMs maintained a constant elastic modulus over 8 days of stretching (day 8 to day 16 in vitro), whereas control HBAMs become stiffer (Fig. 8). Girton et al. (15,16) showed that static collagen gels containing smooth muscle cells also become stiffer with time in culture. Their studies attribute the stiffening to collagen cross linking. It is possible that the repetitive stretch/relaxation cycles inhibit collagen cross-linking in HBAMs. It is also likely that fibroblasts in the HBAMs are remodeling the ECM in response to the repetitive mechanical loading in a different fashion to their remodeling in statically grown HBAMs. As the HBAMs become more elastic with mechanical stimulation, it will be necessary to change the stretch amplitude to impose a desired load. Our ability to predict and measure the tissue viscoelastic properties and the load applied to the engineered tissue will allow us to improve our dynamic engineering techniques to engineer HBAMs with improved morphology that will be useful in the future for a number of applications, including structural repair, drug screening, and gene therapy.
This work was supported by grants from the National Institute on Aging R01-AG-15415, R44-AG-14958, National Aeronautics and Space Administration NAG2–1205, and National Institute of Standards and Technology ATP 70NANB9H3011.
Address for reprint requests and other correspondence: H. H. Vandenburgh, Cell Based Delivery, Inc., 4 Richmond Square, 5th Fl., Providence, RI 02906 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
July 17, 2002;10.1152/ajpcell.00595.2001
- Copyright © 2002 the American Physiological Society