|Home | About | Journals | Submit | Contact Us | Français|
Fibrin is a substance formed through catalytic conversion of coagulation constituents: fibrinogen and thrombin. The kinetics of the two constituents determines the structural properties of the fibrin architecture. We have shown previously that changing the fibrinogen and thrombin concentrations in the final three-dimensional (3D) fibrin matrix influenced cell proliferation and differentiation. In this study, we further examined the effect of changing fibrinogen and thrombin concentrations in the absence or presence of fibroblasts on the structural modulus or stiffness of 3D fibrin matrices. We have prepared fibroblast-free and fibroblast-embedded 3D fibrin matrices of different fibrinogen and thrombin formulations, and tested the stiffness of these constructs using standard mechanical testing assays. Results showed that there was a corresponding increase in stiffness with increasing thrombin and fibrinogen concentrations; the increase was more notable with fibrinogen and to a lesser degree with thrombin. The effect of fibroblasts on the stiffness of the fibrin construct was also examined. We have observed a small increase in the stiffness of the fibroblast-incorporated fibrin construct as they proliferated and exhibited spreading morphology. To our knowledge, this is the first comprehensive report detailing the relationship between fibrinogen and thrombin concentrations, cell proliferation, and stiffness in 3D fibrin matrices. The data obtained may lead to optimally design suitable bioscaffolds where we can control both cell proliferation and structural integrity for a variety of tissue engineering applications.
Biomaterials are essential for the advancement in tissue engineering because they can function as delivery vehicles for bioactive substances or cells. The preferred materials should have mechanical and biochemical properties that are biocompatible, bioabsorbable, and biodegradable; furthermore, the materials should easily be manipulated and easily reproducible. Fibrin has long been recognized as a biopolymer possessing all of the above-mentioned properties.1–4 Physiologic fibrin, as in a fibrin clot, is formed subsequent to tissue injury, and its formation is critical to the initiation of wound healing and to the remodeling process of the wound site. The fibrin clot is formed by enzymatic cleavage and reorganization of two key plasma-derived coagulation components, fibrinogen and thrombin. In the presence of transglutaminase factor XIIIa and calcium ions, a meshwork of cross-linked polyfibrils is formed producing a provisional matrix. Similar to physiologic fibrin, biologic fibrin can be polymerized through fibrinogen and thrombin purified from human or animal source. The polymerized fibrin can be reproduced externally and can be modified to create three-dimensional (3D) porous structures of different size, shape, viscosity, porosity, and tensile strength for a variety of tissue engineering applications: from tissue sealants5 to delivery vehicles for drugs,6 growth factors,7 and cells.8–11
The relationships between fibrin microstructure, rate of polymerization, and structural stiffness have been studied extensively12–14 in a cell-free fibrin clot system; however, the relationship between fibrin microstructure, structural stiffness, and cell biocompatibility in a cell-incorporated system is still unclear. The stiffness of the surface to which cells adhere can affect cell morphology, protein expression, and migration. Yeung et al. have shown that fibroblasts and endothelial cells develop a spread morphology and actin stress fiber expression when grown on matrix surface with a stiffness of at least 2kPa; whereas, stiffness has little effect on neutrophils embedded within the matrix.15 Similarly, we have shown that changing the concentration of fibrinogen and thrombin has significant effect on monocyte adhesion and proliferation.16 It has been established that the modulation of the intrinsic factors such as fibrin constituents, fibrinogen, and thrombin can have profound effect on the mechanical properties of the fibrin microstructure. Modifying the thrombin concentration may affect the rate of polymerization, fiber size, and porosity.17–19 Similarly, modifying the fibrinogen concentration has also been shown to affect fiber size and porosity.20 It is still unclear whether extrinsic factors such as presence of cells can affect the stiffness of the fibrin microstructure. Most studies have focused on how the different fibrinogen to thrombin formulations may have an effect on cell morphology and survival.21–23 However, the relationship between fibrinogen and thrombin with respect to structural stiffness and cellular response still needs to be addressed. In this study, we have taken a comprehensive approach to construct 3D fibrin matrices using a full range of fibrinogen to thrombin formulations to compare final structural stiffness and to determine how stiffness may correlate to cell proliferation.
Fibrin Sealant (Tisseel® Sealant Kit) containing human fibrinogen, thrombin, tris-glycine buffer, calcium chloride, and aprotinin (fibrinolysis inhibitor) was generously provided by Baxter Bioscience (Westlake, CA). Normal human dermal fibroblasts (lot number 5F1351) were purchased from Lonza (Walkersville, MD). Live/Dead® Viability Assay Kit and alamarBlue® Cell Proliferation Assay Kit were purchased from Invitrogen (Carlsbad, CA). Mechanical compressive indentation measurements of 3D fibrin matrices were performed on an Instron 5564 Universal Material Testing machine equipped with a load cell capacity of 1 kN and a Bluehill2®-integrated material testing software (Instron, Norwood, MA).
Fibrinogen (100mg/mL) was reconstituted in 40mM tris-glycine containing 3000KIU/mL of aprotinin at 37°C. Thrombin (500IU) was reconstituted in 40mmol/mL of CaCl2; whereas, 1IU of thrombin is defined as the activity contained in 0.0853mg of the First International Standard of Human Thrombin. The reconstituted fibrinogen was prediluted to various concentrations ranging from 2 to 50mg/mL. Thrombin was prediluted to various concentrations ranging from 2 to 100IU. The resulting formulations between the two components were indicated in Table 1. Three-dimensional fibrin matrices were formed in individual wells of a 24-well tissue culture plate. Two milliliters of 3D fibrin matrices was prepared by quickly mixing equal volume of the appropriately diluted fibrinogen and thrombin solutions. A dual-barrel syringe (Discus Dental, Culver City, CA) was used to mix each matrix (Fig. 1A). The components were allowed to polymerize undisturbed at room temperature for 1h. The resulting polymerized 3D fibrin matrices were covered with 1mL of 40mmol CaCl2 and incubated at 37°C. The 3D fibrin matrices measured 18mm in diameter and 5mm in height with average volume of 1271mm3.
Normal human dermal fibroblasts of early passage (<10) were counted and combined with the fibrinogen and thrombin components to yield desired fibrin formulations and cell densities (ranging from 50,000 (50k) to 5,000,000 (500k) cells). Fibroblast-embedded 3D fibrin matrices and corresponding control acellular 3D fibrin were allowed to polymerize for 1h followed by submersion in DMEM (Mediatech, Manassas, VA) supplemented with aprotinin (3000KIU/mL), 10% FBS, 100units/mL penicillin, and 100μg/mL streptomycin (Fisher Scientific, Pittsburg, PA). The 3D fibrin matrices were maintained at 37°C and 5% CO2 in a humidified incubator until processed.
Compressive indentation measurements on acellular 3D fibrin matrices and fibroblast-embedded 3D fibrin matrices were taken using an Instron 5564 Universal Material Testing Machine. Briefly, excess liquid was removed from individual 3D fibrin matrix at the time of testing leaving the fibrin matrix attached to the walls of the well. A cylindrical flat-ended stainless steel indenter with a diameter of 3.0mm was attached to the testing machine and was programmed to press into the 3D fibrin matrix surface at a controlled displacement rate of 5mm/min until a final displacement of 2mm was achieved by the indenter (Fig. 1B). It should be noted that during testing, the area and the volume of the fibrin matrix remained unchanged by matrix adhesion to the well surface; in addition, the 3.0mm indenter diameter was specifically chosen to rule out total matrix volume displacement during testing. The indenter only displaces the matrix at a focal indentation point, whereas a larger diameter indenter would displace a greater surface area and volume of the matrix of which may affect our measurements. We have tested displacement rate from 1 to 10mm/min and have determined that a 5mm/min displacement rate was a reliable strain rate for these 3D fibrin matrix formulations of varying concentrations. The force displacement data were then analyzed for the mechanical stiffness of both the acellular and fibroblast-embedded fibrin matrix groups. Specifically, a nominal Young's modulus, E, was estimated based upon linearizing portions of the load displacement data and then analyzing the slope of the line in accordance with the Boussinesq elastic solution for the flat punch indentation,24 where
where a is the punch radius (1.5mm), υ is the Poisson's ratio of the fibrin matrix, F is the applied load force, and U is the displacement. A Poisson's ratio of 0.25 was assumed for the fibrin matrix. The Young's modulus, E, was calculated by fitting the equation to the experimental data obtained from displacements between 0.0 and 1.0mm.
The Live/Dead Viability/Cytotoxicity Kit (Invitrogen) employs two color fluorescent dyes: Calcein-AM that produces an intense green fluorescent (ex/em ~495/~515nm) for live cells and EthD-1 that produces an intense red fluorescent (ex/em ~495/~635nm) for dead cells. Direct visual assessment of fibroblast viability inside the 3D fibrin matrices was conducted using this assay kit. Briefly, at various time points of interest, fibroblast-embedded 3D fibrin matrices were washed once with 1mL of Dulbecco's phosphate buffer saline (D-PBS). The wash buffer was removed and replaced with 1mL of fresh serum-free DMEM containing 5μM of Calcein-AM and 5μM of EthD-1. A 3h incubation period was required to ensure that the reagents had diffused through the cell-embedded fibrin matrix. After incubation, the fibroblast-embedded 3D fibrin matrices were washed three times with 2mL of D-PBS at 10min intervals to remove unbounded reaction products. The bicolor labeling was viewed through a Leica IRB inverted microscope (McBain Instruments, Los Angeles, CA) equipped for fluorescent detection and fitted with an Optronics® digital CCD camera for image capturing (Optronics, Goleta, CA). A Bioquant® Image Analysis System (Bioquant, Nashville, TN) was used for image analysis.
Fibroblast proliferation inside 3D fibrin matrices of varying formulations was also determined using alamarBlue Cell Proliferation/Cytotoxicity Kit (Invitrogen, Irvine, CA). alamarBlue is designed to provide a rapid and sensitive measurement of cell proliferation and cytotoxicity in various human and animal cell lines. alamarBlue is a redox indicator that yields a colorimetric change and a fluorescent signal in a response to cellular metabolic activity. The compound is nontoxic, soluble, and stable in tissue culture medium. Fibroblast-embedded 3D fibrin matrices were incubated with 1mL of fresh DMEM and 10% FBS containing alamarBlue diluted 1:10. At various time intervals, the redox reaction, in which alamarBlue is reduced by the cells, was measured by fluorescent reading at excitation wavelength of 535nm and emission wavelength of 595nm. The fluorescent measurements were translated into cell number against a standard curve using different cell densities. The fluorescent measurements were performed on an Infinite® 200 series microplate reader (Tecan, San Jose, CA).
To assess whether fibroblast contraction may exert an effect of matrix stiffness, fibroblast-embedded fibrin matrices were treated with or without 0.5μM of staurosporine (Sigma, St. Louis, MO). The fibroblast-embedded matrices were cultured up to 14 days followed by Live/Dead assessment and mechanical indentation testing.
Data were presented as mean±standard deviation from three or more experiments with each experimental parameter done in triplicates. The statistical significance was assessed by paired Student's t-test, and a p-value of less than 0.05 was considered statistically significant, while a p-value of 0.01 was considered highly significant. Some data were analyzed by linear regression analysis, and statistical significance was assessed by two-way ANOVA with a Bonferroni posttesting. Statistical analysis was performed using GraphPad Prism® 4 analysis software (GraphPad Software, San Diego, CA).
To determine how changing the fibrinogen or thrombin concentrations may affect the structural modulus of the 3D fibrin construct, we have prepared broad range of matrices comprising of different fibrinogen and thrombin formulations (Table 1). The structural modulus of the 3D fibrin constructs herein shall be loosely defined as an index of stiffness.
Modulating the fibrinogen and thrombin concentrations represent two separate variables that we have evaluated for the stiffness correlation of the 3D fibrin matrices: (i) by comparing different concentrations of fibrinogen (ΔF) with respect to a constant thrombin concentration (~T) to obtain a stiffness versus changing fibrinogen concentrations correlation; (ii) by comparing different thrombin concentrations (ΔT) with respect to a constant fibrinogen concentration (~F) to obtain a stiffness versus changing thrombin concentrations correlation. For ΔF test, we have prepared six groups of 3D fibrin matrices with each group comprising 2, 5, 10, 20, 50, or 100IU/mL of thrombin. For each group, the fibrinogen concentrations varied from 2, 5, 17, 34, and 50mg/mL. For the ΔT test, we have constructed five groups of 3D fibrin matrices with each group comprising 2, 5, 17, 34, or 50mg/mL of fibrinogen, and the thrombin concentrations were varied within each group (2–100IU/mL).
Mechanical indentation for the ΔF tests (Fig. 2A) showed that the structural modulus of the 3D fibrin matrices increased with increasing fibrinogen concentration. This effect was measured in all groups of thrombin concentrations from 2 to 100IU/mL (Fig. 2A). Paired Student's t-test analysis showed statistical significance (*p<0.001) of increasing stiffness in response to ΔF with matrices having lower kPa values at 2mg/mL of fibrinogen and higher kPa values as the fibrinogen concentration increased up to 50mg/mL. This was observed in all groups of ~T. In addition, there were significant differences (**p<0.001) in the change in stiffness between ΔF at 2 and 5mg/mL between 5 and 17mg/mL, between 17 and 34mg/mL, and between 34 and 50mg/mL (Fig. 2A). The percent difference in stiffness corresponding to ΔF was also evaluated (Fig. 3A). At ~T of 2IU/mL, changing the fibrinogen concentration from 2 to 5mg/mL resulted in a fivefold increase in the stiffness of the matrices. At 17mg/mL, the increase was 12-fold; at 34mg/mL, the increase was 24-fold; and at 50mg/mL, the increase was 46-fold. Similar increases in stiffness were measured in all groups of thrombin (2, 5, 10, 20, 50, and 100IU/mL) (Fig. 3A). To evaluate whether the corresponding increases were linear, we have performed a linear regression analysis against the data obtained (data not shown). Results showed that the increase in the structural modulus of the matrices linearly corresponds to increasing fibrinogen concentrations at ~T. The R2 values were greater than 0.97, and two-way ANOVA analysis showed that fibrinogen had the same significant effect (p<0.001) at all values of thrombin.
Evaluating the data with respect to ΔT and at ~F showed that the structural modulus of the 3D fibrin matrices increased when the thrombin concentration was increased (Fig. 2B), but in the lowest fibrinogen formulation of 2mg/mL, the stiffness varied with increasing thrombin concentration. At ~F of 5mg/mL, the stiffness steadily increased as the thrombin concentration increased (from 2 to 20IU/mL); however, at thrombin concentrations of 50 and 100IU/mL, the stiffness appeared to have declined compared to the 20units/mL of thrombin formulation. Similarly, at ~F of 17, 34, and 50mg/mL, changing the thrombin concentrations showed significant increases in the stiffness, with decreasing values measured in matrices comprising 100IU/mL of thrombin. Paired Student's t-test showed statistical significance where p<0.05 for all ΔT values compared to lower 2IU/mL concentrations. In addition, statistical significance was observed when compared 2 to 5IU/mL, 5 to 10IU/mL, 10 to 20IU/mL, 20 to 50IU/mL, and 50 to 100IU/mL. Evaluating the percent differences of the stiffness corresponding to ΔT showed that at ~F of 2mg/mL, the percent differences varied with at various thrombin values (Fig. 3B). Increasing the thrombin concentrations, from 2 to 5IU/mL, resulted in a 3-fold difference; at 10IU/mL, a 4.6-fold difference; at 20IU/mL, 2.3-fold; at 50IU/mL, 4-fold; and at 100IU/mL, 3-fold. This variability in values was diminished as the fibrinogen concentration increased. At ~F of 5mg/mL, a 1.3-fold difference was observed when the thrombin concentration was increased from 2 to 5IU/mL; 10IU/mL resulted in a 1.6-fold difference; 20IU/mL resulted in a 1.8-fold difference; 50IU/mL resulted in a 1.2-fold difference; and 100IU/mL resulted in a 1.5-fold difference. Similar result was observed at ~F of 17, 34, and 50mg/mL. Linear regression showed that increasing the thrombin concentration at lower ~F of 2mg/mL does not affect the gel stiffness linearly, where R2<0.5 for most values and that rate of change or differences varied greatest at lower 2mg/mL of fibrinogen; however, in matrices of higher fibrinogen concentrations, >5mg/mL, the effect of thrombin on the matrices' stiffness appeared to follow a more linear trend. With the exception at the 100IU/mL of thrombin formulation, the stiffness decreased compared to the 20IU/mL formulation. Two-way ANOVA showed that thrombin does not have the same effect at all values of fibrinogen, and that the effect is only significant in matrices consisted of fibrinogen concentrations >5mg/mL. In summary, fibrinogen accounted for 84% of the total variance, while thrombin accounted only 5.8% of the total variance. The data analysis showed that changing the fibrinogen concentrations had a greater effect on the structural modulus of the 3D matrices than changing the thrombin concentrations.
To evaluate how fibrin composition may affect cell proliferation, we seeded normal human fibroblasts (200,000 (200k) cells) in 3D fibrin matrices of different fibrinogen formulations: 5, 17, and 34mg/mL at ~T of 2IU/mL. To exclude the degradation effect, 3000KIU/mL of aprotinin was added to the matrices, and the cells were allowed to grow in the 3D fibrin matrices for 1 to 10 days in normal culture media supplemented with aprotinin. At various time points, Live/Dead assessment of fibroblasts viability and mortality was visualized by Calcein-AM/EthD-1 labeling. For proliferation studies, cellular metabolic activity of fibroblast embedded in 3D fibrin matrices was determined by alamarBlue. Absorbance reading from the assay was interpreted as the relative number of cellular activity or cell number. Results showed that in the 3D fibrin matrices consisted of 5mg/mL of fibrinogen, fibroblast continued to proliferate from day 1 through day 10 in culture (Fig. 4A). In the 17mg/mL formulation, the proliferation values increased at day 7; by day 10, the values had remained the same as that of day 7; conversely, in the 34mg/mL formulation, fibroblast proliferation did not increase with respect to day 1; rather, there was a slight decrease at both days 7 and 10 (Fig. 4A). Paired Student's t-tests showed that there was significant fibroblast growth inhibition in 3D fibrin constructs consisted of fibrinogen concentration >5mg/mL; further, the effect was also significant as time progressed. Fibroblast proliferation in the 5mg/mL formulation increased linearly by as much as 80% after 7 days in culture, and at day 10, the proliferation was greater than 100%. In the 17mg/mL formulation, the proliferation was less linear and only increased by 24% by 7 days and 26% by 10 days. Proliferation was not observed in formulation of 34mg/mL; in fact, fibroblast growth inhibition was observed (Fig. 4B). The lack of EthD-1 labeling for dead cells in all fibroblast-embedded 3D fibrin matrices suggested that cell death was not the cause of growth inhibition as observed in higher fibrin formulation, where they displayed abnormal morphology. Morphological assessment showed that fibroblasts formed densely packed cell network and extensions (cell spreading) in lower 5mg/mL fibrin formulation; whereas, in the 17 and 34mg/mL formulations, the cells appeared spherical and cell spreading was minimal (Fig. 4C).
To examine whether mixing cells with the fibrin formulation components would affect the kinetic rate of fibrin polymerization and, ultimately, the microstructure and structural modulus of the 3D matrices, fibroblasts, from low to high cell densities (0–5000k cells), were incorporated into 3D fibrin matrices comprising ~F of 5mg/mL and ~T of 5IU/mL. Mechanical indentation tests showed that there were insignificant differences (<5%) between the structural modulus of the acellular 3D fibrin matrices (no cells) and that of those matrices with fibroblasts embedded in them (Fig. 5). There were little differences in the stiffness of 3D fibrin matrices containing 200k fibroblasts and that of the constructs with high cell density (5000k cells).
To find the optimum cell seeding density for the 3D fibrin matrices, we incorporated fibroblasts at various densities into matrices of ~F=5mg/mL and ~T=2IU/mL. AlamarBlue assessment showed that there was a linear increase in fibroblast proliferation from day 0 up to day 21. This increase was observed in all cultures of different seeding densities, and the paired Student's t-test showed significance, p<0.05, when compared day 1 culture to all other time points (Fig. 6A). The percent change in cell number was also evaluated (Fig. 6B). After 7 days, in the 50k cell density group, fibroblast number increased significantly by 91%, by 23% in the 100k cell density group, by 105% in the 200k cell density group, and by 15% in the 400k cell density group. After 14 days, the 50k cell density group had a 650% increase in cell number, 319% in the 100k cell density group, 183% in the 200k cell density group, and 16% in the 400k cell density group. By 21 days, the 50k cell density group had increased by 1000%, the 100k cell density group by 660%, the 200k cell density group by 274%, and the 400k cell density group by only 61%. From the evaluation, it was clear that there was a correlation between cell seeding density and cell proliferation over time. In this case, seeding fibroblasts at 50,000 cells appeared to be the optimal density for incorporation into 3D fibrin matrices, where cells showed continued growth for up to 21 days and beyond. In comparison, in matrices where the initial seeding density was at 400,000 cells, fibroblast proliferation was minimal.
The relationship between initial seeding density, proliferation over time, and effect on stiffness was examined. After 24h in culture, the structural modulus of the matrices was measured. Compared to the acellular matrices, the 50k cell density and the 100k cell density group had lower kPa values than those of the acellular matrices, whereas the 200k cell density and the 400k cell density group had high kPa values. Throughout the course of 21 days in culture, all cell groups showed increases in stiffness. This increase was most significant (p<0.05) at day 21. Comparing stiffness values to the acellular matrices, seeding cells within the fibrin matrices does affect the stiffness over time. The effect by cells was further investigated by comparing the effect on stiffness by cell densities. Results showed that there were definite differences between the 50k cell density group and the other higher cell density groups. Most notably, at 21 days, the stiffness of the 400k cell density group was greater than the stiffness of the 50k cell density group (Fig. 7A).
In comparison to time in the fibrin matrices, the 50k cell density group and the 100k cell density group started with lower stiffness values; however, as the culture time increased, the percent change in stiffness of the matrices correspondingly increased with significance at day 21. The percent change was calculated by dividing the stiffness values of the day 7-21 groups to that of the day 1 groups. This would indicate the degree of change in the stiffness every 7 days of culture time up to 21 days for various cell seeding densities. The 200k and 400k groups had small increases in stiffness from day 7 and reaching significant increases after 21 days (Fig. 7B).
The percent change in cell proliferation was compared to the percent change in the stiffness of the matrices (Fig. 8A). After 7 days of culture, the 50k cell density group had an average increase of 91% in cell number; correspondingly, the average change in stiffness for the matrices was recorded at 45%, thereby yielding a change in proliferation to change in stiffness index ratio of 2:1. At 100k cell density, the ratio was 0.8:1; at 200k cell density, the ratio was 5.6:1; and at 400k cell density, the ratio was 6.8:1 (Fig. 8B). This index ratio would vary among all groups, with the 50k cell density group recording a continued increase in the change in proliferation to change in stiffness index ratio at days 14 and 21. The 100k cell density and the 200k cell density groups recorded fluctuations in the change in proliferation to change in stiffness index ratio at those time points. Whereas in the 400k cell density group, the change in proliferation to change in stiffness index ratio decreased to 1.3:1 at day 14 and 0.8:1 at day 21. The larger index ratio value suggested that the disparity between the change in cell proliferation and the change in stiffness was smaller. As time progressed, the 50k cell density groups had increasing ratio values, suggesting that the large increase in fibroblast proliferation within the 3D fibrin matrices affected the structural modulus of the matrices minimally. In the 100, 200, and 400k cell density groups the change in proliferation to change in stiffness ratios appeared to be decreasing as culture time progressed to 21 days. Most notably, the 400k cell density had the smallest ratio after 21 days in culture. This would suggest that by day 21, the 400k density groups had decreased proliferation, while the corresponding structural modulus continued to increase, thereby suggesting that some other factor besides cellular proliferation is contributing to the increase in the structural stiffness. We have observed that by day 14, the fibroblast-embedded fibrin matrices appeared compacted with decreasing size and volume and that the compaction was more significant in the 400k density groups.
As the culture time increased, the 3D fibrin matrices decreased in volume (Fig. 9A). Morphological assessment of the fibroblasts within the matrices revealed that over time, the cells proliferated and formed densely packed networks by cell spreading. We next tested whether fibroblast proliferation had an effect on matrix contraction, thereby affecting the structural modulus. Fibroblast cell spreading was inhibited by adding staurosporine to the culture media. After 14 days, Live/Dead labeling showed that the staurosporine had no toxic effect on the cells, and that cell spreading was inhibited (Fig. 9C). We have shown that the 3D fibrin matrices of the 50k cell density group did not contract. The effect of staurosporine in this cell density group was not clear; however, in the other higher cell density groups, the effect of staurosporine clearly showed that cell spreading does contribute to the stiffness of the matrices. The 100k group recorded stiffness differential of 39% in presence of staurosporine; similarly, the 200k group recorded 16.7%, and the 400k cell density group recorded 34.4% (Fig. 9B).
The unique feature of fibrin is that it can be modified to create microstructures of different porosities and stiffness to simulate an environment that is optimized for cell growth and survival. This would include matching the mechanical properties of the specific tissue type. For cell-incorporated fibrin matrices, porosity and structural stiffness are important mechanical properties to consider. Porosity allows for diffusion of macromolecules and nutrients for cell signaling and proliferation, as well as facilitating cell retention. For example, fibrin matrices with large, opened network of fibrils may result in increased cell loss and faster degradation of the fibrin material, whereas matrices with dense network may retain cells and resist fibrinolysis.25 The porosity of the fibrin matrices can be controlled by modifying the thrombin and fibrinogen concentration.26 Modifying the thrombin and fibrinogen concentrations has been shown to affect the rigidity of the fibrin microstructure.14,15,21
The relationship between structural stiffness and cell processes has been investigated. Lo et al.27 have shown that fibroblasts exhibit spreading morphology and preferentially migrate from a soft to hard surface. Similarly, Yeung et al.15 have reported that fibroblasts developed spreading morphology when grown on surfaces with a stiffness modulus of greater than 2kPa, and that different cell types may prefer different stiffness for growth and survival. Most cells rely on a stiffness-sensing feedback mechanism to which they can interact with the substrate surface.28
In this study, we have explored a wide range of formulations involving different concentrations of fibrinogen and thrombin to create 3D fibrin matrices of varying stiffness. The range in stiffness was measured between 0.058kPa (for lower concentration of fibrinogen/thrombin formulation) and 4kPa (for higher concentration of fibrinogen/thrombin formulation). Much of the increase in the stiffness was directly related to increasing fibrinogen concentration, and to a lesser degree to thrombin concentration. Physiologic fibrinogen and thrombin concentration for clot formation has been measured around 2mg/mL and 0.01 to 0.1units/mL of thrombin.17,29 Our formulations include the low 2mg/mL of fibrinogen and the low 2units/mL of thrombin, and we have extended range of the formulation to a high 50mg/mL of fibrinogen and 100units/mL of thrombin. From this study, we now have a correlation between stiffness and fibrinogen/thrombin formulation; in general, a low stiffness value of 0.05kPa to a high stiffness value of 4kPa could be achieved using this range of formulations. This range falls within the required substrate stiffness of 0.001 to 10kPa for cell attachment as reported in other studies.30,31 Interestingly, in our fibroblast proliferation study within 3D fibrin matrices comprising 5, 17, and 34mg/mL of fibrinogen at constant thrombin of 5units/mL, the respective stiffness was 0.4, 0.7, and 1.5kPa. These values fall short of the 2 to 3kPa value as indicated in Yeung et al.15 for fibroblast mechanosensing and cell spreading. However, our morphological assessment showed that the fibroblast favored lower 5mg/mL concentration, at 0.4kPa, as opposed to the 34mg/mL concentration of 0.7kPa. Clearly, at least in our system, the fibroblast preferred formulations that yielded lower kPa values. Fibroblasts in lower formulation continued to proliferate and expressed spreading morphology, whereas the proliferation was decreased and cell spreading inhibited in higher formulations.
We next asked whether proliferation of fibroblasts within the fibrin matrix could contribute to changes in the structural stiffness. Seeding fibroblasts at various cell densities (50,000 to 400,000 cells per matrix) and culturing them up to 21 days showed that cell proliferation and culture time do not significantly alter the structural stiffness of the matrix. Interestingly, we have observed that an initial seeding density of 50,000 cells was optimal for ensuring continued proliferation within the matrices, whereas with higher initial cell seeding density of 400,000 cells, the rate of proliferation had diminished over time.
Degradation and shrinkage are major factors that can affect the 3D fibrin architecture and, ultimately, its stiffness. Fibrinolysis was inhibited by the addition of aprotinin into the matrices at the time of polymerization. When fibroblasts were incorporated, initial cell seeding density does not alter the stiffness of the matrices; however, as fibroblasts proliferated, matrix compaction or contraction was observed, and correspondingly, there was a significant increase in the structural stiffness of those matrices that have compacted. Inhibition of fibroblast spreading and expression of actin stress fibers through exposure to staurosporine showed that the compaction of the 3D fibrin matrices by fibroblasts may increase the stiffness by as much as 39%.
In conclusion, we have presented a detailed analysis of how changing the intrinsic factors such as fibrinogen and thrombin correlates to change in the structural stiffness of the 3D fibrin matrices; in addition, we have explored various extrinsic factors, such as cell proliferation, cell seeding density, and cell contraction of matrix, and how these can affect the structural stiffness.
This study was supported by the NIH/National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) (AR05325001A).
No competing financial interests exist.