Using a Radial Diffusion Method to Investigate the Role of Plasmin Degradation of Fibrin in a Physical Model of an Early-phase Wound

Background. Fibrin clot formation, which acts to stabilize a wound following injury, is among the key early aspects of dermal wound healing. This preliminary matrix is eventually degraded via a process known as fibrinolysis and replaced with a collagen-rich matrix that continues to be remodeled to minimize scarring. Disruptions in these carefully coordinated events lead to certain undesirable conditions such as fibrosis and the formation of abnormal scars that are associated with excess amounts of collagen. The hypothesis proposed herein is that the presence of collagen (and potentially other molecules) in an early-phase model of healing alters fibrinolysis and that this effect can be attenuated with mediators of the process.Materials and Methods. Laboratory in vitro experiments were conducted using agarose-fibrin gel systems with and without collagen to study fibrinolysis caused by plasmin (a serine protease that degrades fibrin) and the effects of aprotinin (a serine protease inhibitor) and bromelain (an extract from pineapple) on fibrin clot breakdown. The extent of fibrinolysis was monitored at various times (0.5, 1, 2, 4, 8, 12, 24, 48, and 72 hours) by measuring the size of rings of fibrinolysis following the diffusion of plasmin. The data obtained at 0.5, 12, and 24-hour time points were considered (because there was no difference found in the data collected for closer intermediates nor for the longer times beyond 24 hours) and were compared using the nonparametric Mann-Whitney U statistical significance test. Results. The results obtained showed aprotinin significantly inhibited fibrinolysis in systems containing collagen, while bromelain improved fibrinolysis. In general, the presence of increasing amounts of collagen in the system decreased the extent of fibrinolysis. Conclusions. These findings support the notion that early-phase deposition of collagen contributes to disrupted fibrinolysis, which could lead to impaired healing as well as potentially facilitate control of fibrinolysis.

Acute and chronic wounds represent a suite of conditions that result in serious public health issues affecting millions of people around the world each year. Wounds come in various forms, affect a variety of tissues, and may result from sports, accidents, gunshots, surgeries, burns, diabetes, and many other means that result in a disruption of normal tissue homeostasis. Repairing wounds, despite the causes, engages a complicated series of coordinated, dynamic, interactive stages involving cells, exchanges between dissimilar tissues, cytokines, growth factors, soluble biochemical intercessors, the extracellular matrix (ECM), and other physiologic processes.^1-3 One important feature of injury repair is fibrinolysis, which is a proteolytic degradation of a temporary fibrin matrix by plasmin generated from plasminogen.

Under normal conditions, the process of wound healing involves 4 imbricating stages: hemostasis, inflammation, proliferation, and remodeling.^4-10 While this tissue repair process varies among dissimilar tissue types with more resemblances than distinctions among them, several clinically significant factors including hypoxia,^1,11,12 influences of necrotic tissue and infection,¹² poor nutrition,^1,12-15 metabolic disorders such as diabetes mellitus,^1,12 microbial contamination,¹⁶ genetic susceptibility of patients,^13,17,18 and certain medications are known to impede desired wound healing^1,19; hence, a conducive physiologic environment must be present in order to achieve preferred tissue repair and regeneration. Fibrin deposition followed by its degradation and ensuing replacement with collagen that is made and arranged by fibroblasts recruited to the wound site after hemostasis represents a vital aspect of the tissue repair process,^20,21 but the inability of the body to degrade the provisional fibrin matrix has been associated with a number of fibrotic conditions.^20,22 Although studies have been conducted on wounds and despite the vast knowledge regarding the wound healing process, healing complications and abnormalities are still observed. Examples of these abnormalities are observed in diabetic ulcers, where multiple biochemical and physiological defects can result in compromised tissue repair, and also in fibroproliferative wound healing abnormalities (eg, hypertrophic scars and keloids), which are categorized by excess accumulation of collagen at the wound site. As tissue and organ fibrosis (a process characterized by excessive deposition of collagen at the wound area during healing) involve disproportionate collagen production or limited collagenase action, molecules that support the production of collagen or prevent its breakdown must be involved.²⁰

Preventing and treating tissue repair abnormalities such as hypertrophic scars and keloids are difficult due to lack of knowledge about their origin,¹⁰ but Gauglitz et al²² have reported that abnormal scarring could form following irregularities in the tissue repair process. For normal tissue regeneration and repair to occur, ensuring a suitable physiological environment and good overall health of the patient are necessary to prevent the accumulation of early-phase excess collagen,^1,22 since the tissue repair process entails a sufficient provision of blood and nutrients for the synthesis and degradation of ECM in the damaged tissue. Extracellular matrix degradation is a very important aspect of tissue remodeling, which involves a completely structured equilibrium between the formation and breakdown of ECM components necessary for natural development, growth, repair, and maintenance of proper tissue architecture.²³ The removal of wound debris as well as the temporary fibrin base components to facilitate formation of new cells and to guide the production of new blood vessels during tissue repair is crucial, and this process is facilitated by the actions of serine proteases. Plasmin, a serine protease produced from plasminogen by the action of plasminogen activators, functions to degrade many blood plasma proteins, including those in fibrin matrices,²⁴ and it has been attributed to controlling tissue remodeling during tissue repair. Substantial evidence suggests plasmin plays an important role in triggering suppressed growth factors such as transforming growth factor and matrix metalloproteinases (MMPs), especially MMP-1.^24,25

An objective of this study has been to design, build, test, modify (as necessary), and model a cell-free system for assessing fibrinolysis and the influence of various mediators of this process with and without collagen in the model. As plasmin is known to break down fibrin, and since an alteration in the balance of collagen in a healing wound is associated with adverse outcomes,²⁶ the hypothesis tested herein is that the presence of collagen in a fibrin-rich physical model decreases fibrin degradation (a phenomena which in vivo could also be the result of poor nutrition,¹ microbial contamination,¹⁶ and/or genetic susceptibility of patients¹⁷) and that this effect can be attenuated with mediators of the process. Ultimately, appropriate utilization of such mediators might be useful in improving outcomes of wound healing.

Protein samples and preparations. All proteins used in this experiment were of bovine origin (unless otherwise stated) and of the highest grade available from the authors’ commercial sources. Purified fibrinogen (> 98% purity, > 75% clottable protein, Australian source) depleted of fibronectin and plasminogen was obtained from Akron Biotech (Boca Raton, FL) as a lyophilized powder. The fibrinogen protein was dissolved in 0.2 M Tris-hydrochloride (pH 7.4) buffer containing 0.15 M sodium chloride at 37°C and then held at 4°C for 24 hours to enable the lyophilized fibrinogen to fully dissolve in the saline buffer solution before using it for experiments. Bovine alpha-thrombin (8.1 mg/mL in 50% glycerol/water by volume, specific activity of 3487 units/mg) was obtained from Haematologic Technologies, Inc (Essex Junction, VT). Alpha-thrombin was diluted (2 µL/mL) in 0.2 M Tris buffer (pH 7.4) and then used to cleave fibrinogen for the formation of fibrin clots. Plasmin (> 95% purity) at a concentration of 1.8 mg/mL was purchased from Innovative Research, Inc (Novi, MI) as a frozen liquid. Aliquots were prepared at 0.072, 0.144, 0.216, 0.288, 0.36, 0.72, 1.08, 1.44, and 1.80 mg/mL by combining 0.4, 0.8, 1.2, 1.6, 2.0, 4.0, 6.0, 8.0, and 10.0 µL, respectively, of this stock with an appropriate volume of 0.2 M Tris buffer (pH 7.4) containing 0.15 M sodium chloride. These aliquots were then stored at 4°C until use. Collagen (> 98% type 1 collagen from bovine achilles tendon) was obtained as a lyophilized powder from Sigma-Aldrich Corporation (St. Louis, MO). The insoluble collagen powder was reconstituted as discussed below as a suspension in 0.2 M Tris buffer (pH 7.4) containing 0.15 M sodium chloride at 37°C. Aprotinin from bovine lung (approximately 4 inhibitor units/mg, 6236 kallikrein inhibitor units/mg specific activity) was purchased from MP Biomedicals (Solon, OH) as a lyophilized white powder. This water-soluble serine protease inhibitor was prepared for use by dissolving 10 mg of aprotinin powder in 1 mL of 0.2 M Tris buffer (pH 7.4) containing 0.15 M sodium chloride at 37°C. Bromelain from pineapple (specific activity 1100–1500 gelatin digestion unit/g) also was purchased as a powder from MP Biomedicals. Bromelain stock was prepared by dissolving 10 mg of bromelain powder in 1 mL of 0.2 M Tris buffer (pH 7.4) containing 0.15 M sodium chloride at 37°C. The sodium chloride and agarose (< 7% moisture) used in gel preparation were purchased from Fisher Scientific (Pittsburgh, PA). Tris reagents were procured from Promega Corporation (Fitchburg, WI), and the hydrochloric acid solution was from Science Kit & Boreal Laboratories (Tonawanda, NY).

Fibrin gel preparation. Fibrin gels were prepared using the method employed by Astrup and Mullertz²⁷ with slight modifications. A plasminogen-free, lyophilized bovine fibrinogen with a mass of 50 mg was dissolved in 20 mL of 0.2 M Tris buffer solution (pH 7.4) containing 0.15 M sodium chloride at 37°C. The resulting aqueous solution containing 0.25% (wt/vol) of bovine fibrinogen was cleaved overnight with 2 µL of alpha-bovine thrombin (specific activity 3487 units/mg) diluted in 1 mL of 0.15 M saline buffer solution (pH 7.4) and then mixed with 40 mL of 1% (wt/vol) agarose solution at 70°C for 2 minutes. Then, 2 mL of the resulting hot mixture was carefully pipetted into each well (of diameter 3.81 cm) contained in two 6-well sterile plates (Fisher, Catalog # 08-772-33) and allowed to solidify at room temperature. The procedure was repeated as necessary to produce a sufficient number of gel-containing plates. The result from this approach is that each of the fibrin gels contained ~0.083% wt/vol of cleaved bovine fibrinogen.

Collagen-fibrin gel preparation. The collagen-fibrin gel was prepared using the same procedure described above in preparing the fibrin gels with minor modifications. First, the fibrinogen solution used was prepared as follows: multiple vials of a 0.25% (wt/vol) solution of fibrinogen in 0.2 M of Tris buffer containing 0.15 M sodium chloride were incubated at 37°C overnight (approximately 24 hours) with alpha-thrombin to produce a fibrin clot in each vial. Then, 50 mg of collagen powder was dissolved in 10 mL of 0.2 M Tris buffer solution containing 0.15 M of sodium chloride at 37°C to form a hazy suspension at 0.5% wt/vol. Following this, 10 mL of the resultant suspension and 10 mL of a 0.5% (wt/vol) “processed” fibrinogen solution were mixed with 20 mL of 1% (wt/vol) agarose solution at 70°C. Upon addition of the “processed” fibrinogen solution to the collagen solution and the subsequent addition of this mixture to the heated agarose solution, the resulting gel was expected to simulate a wound with varying levels of fibrin. Two milliliters of the hot mixture were carefully pipetted into each well of 6-well sterile plates and allowed to solidify at room temperature. Similarly, collagen-fibrin plates containing 1% (wt/vol) collagen solution were prepared using this method. The result from this approach is that each of the collagen-fibrin gels contained ~0.083% wt/vol of cleaved bovine fibrinogen and either ~0.083% or ~0.167% wt/vol of collagen.

Determination of fibrinolytic activity in fibrin gels. The fibrinolytic activity was determined at different plasmin concentrations. After cooling to room temperature, a 3.6-mm-diameter well was made at the center of each fibrin (or fibrin-collagen) gel in the 6-well plates by removing a gel plug using a coffee stirrer (referred to as a cannula hereon). Stock solutions of bovine plasmin were diluted to obtain dilutions at the following concentrations: 0.072, 0.144, 0.216, 0.288, 0.360, 0.720, 1.080, 1.44, and 1.80 mg/mL. Then, approximately 10 µL of each diluted plasmin sample was carefully pipetted into these 3.6-mm wells, and the gels were incubated at 37°C with 5% carbon dioxide (to simulate typical tissue culture conditions and maintain desired pH) for 0.5, 1, 2, 4, 8, 12, 24, 48, and 72 hours. Fibrinolytic activity was assessed by measures of the size of the “cleared” zone that was revealed during radial diffusion of the plasmin. This experiment was conducted 5 times for each set of conditions to ensure consistency, and the average diameter of the zones (minus the diameter of the cannula) was taken as the indicator of fibrinolytic activities.

Determination of inhibitor effects on fibrinolytic activity in collagen-fibrin gels. Aprotinin from bovine lung was used as an inhibitor of plasmin activity in the gels. Inhibitor stock was prepared by dissolving 10 mg of aprotinin in 1 mL of Tris buffer (pH 7.4), and 2 µL of the inhibitor stock was reconstituted in 1 mL of 0.15 M sodium chloride solution (pH 7.4) and used for the analysis. Serial dilutions of stock bovine plasmin, approximately 10 µL that also contained aprotinin at a concentration of 0.003 pM, were carefully pipetted into these 3.6-mm wells and the systems incubated at 37°C with 5% carbon dioxide for 0.5, 1, 2, 4, 8, 12, 24, 48, and 72 hours. Inhibition of fibrinolysis via aprotinin was illustrated by comparing the size of the cleared lytic zone in the wound model with inhibitor present versus that without inhibitor. This experiment was conducted 5 times to ensure consistency, and the average diameter of the zones (minus the diameter of the cannula) was taken as the indicator of fibrinolytic activities.

Determination of bromelain effects on fibrinolytic activity in collagen-fibrin gels. A measure of plasmin activity promotion (via bromelain from pineapple) in gels was obtained following the same general procedure as discussed above for determining inhibitor effects on fibrinolytic activity in collagen-fibrin gels. Bromelain stock was prepared by dissolving 10 mg in 1 mL of Tris buffer (pH 7.4); 2 µL of the bromelain stock was reconstituted in 1 mL of 0.5 M sodium chloride solution prepared with 0.2 M buffer solution (pH 7.4) and used for the analysis. Serial dilutions of stock bovine plasmin (~10 µL) that also contained bromelain at 0.003 pM was carefully pipetted into the 3.6-mm wells and incubated at 37°C with 5% carbon dioxide for 0.5, 1, 2, 4, 8, 12, 24, 48, and 72 hours. The measure of enhancement of fibrinolysis was illustrated by comparing the size of the cleared zone in the system with bromelain versus that without this proposed enhancer. This experiment was conducted 5 times to ensure consistency, and the average diameter of the zone (minus the diameter of the cannula) was taken as the measure of fibrinolytic activities.

Statistical analysis. The diameters of the lysed area in the in vitro fibrin and fibrin-collagen gel models were measured at different time intervals for a maximum duration of 72 hours. The nonparametric Mann-Whitney U significance test was used to statistically analyze the experimental data of 2 categories for each point in time as summarized in Table. Statistically significant differences were considered at P < .05, and SPSS version 22.0 statistical software (IBM Corporation, Armonk, NY) was used for all statistical evaluations.

Laboratory experiments were conducted to assess the interplay between fibrinolysis in the presence or absence of collagen and inhibitors or promoters of the process. A comprehensive report on the different laboratory trials and a discussion on the probable causes for the behavior of the in vitro systems are provided below.

Measures of Fibrinolysis (Group A). To study the effect of fibrinolysis in the presence of collagen, 2 different in vitro fibrin systems were used; 1 contained 0.083% (wt/vol) type 1 collagen in addition to the fibrin at 0.083% (wt/vol); and the other contained 0.167% (wt/vol) type 1 collagen in addition to the fibrin at 0.083% (wt/vol). The experimental trials (n = 5) were conducted at 37°C and pH 7.4 for each system with 10 µL of various concentrations of plasmin (added to the wells at the center of the gels) were allowed to diffuse for a maximum duration of 72 hours. The diameter of cleared zones (darker-shaded regions around the wells in each gel; eFigure 1) was measured at particular times, and the diameter of the cannula used to create the wells was subtracted from this value to provide the diameter of lysis, which was observed to be increasing with time at different plasmin concentrations as shown in eFigure 2. A Mann-Whitney U statistical analysis was conducted on the data gathered from these systems after 0.5, 12, and 24 hours to determine any statistical significance in the degradation of fibrin within these systems. These 3 time intervals were chosen because no difference in the data collected for closer intermediates nor for the longer times of 48 and 72 hours was found. Results indicated differences in the sizes of the cleared zone were greater at the 2 earlier time points (0.5 and 12 hours) in the system that only contained fibrin, but any early differences between this system and those also containing collagen were diminished within 24 hours. 

Impact of aprotinin on fibrinolysis: comparisons between systems for fibrinolytic activity (groups B, D, and F). The statistical results from the analysis of data from Group D, which compared a 50:50 mix of collagen and fibrin with and without the presence of aprotinin (Table), showed significant differences in the extent of fibrinolysis in the first 30 minutes when aprotinin was used versus when it was not (eFigure 3A) for all plasmin concentrations. After 12 hours, no significant inhibition was observed when 8 µL of plasmin stock was used; however, there was a significant difference for plasmin volumes of 0.4, 0.8, 1.2, 1.6, 2.0, 6.0, and 10.0 µL and 4.0 µL, respectively, at P < .01 and P < .05 (eFigure 3B). Further statistical analysis of data after 24 hours showed no significant inhibition of fibrinolysis when 6 µL of plasmin stock was used, but a significant difference at P < .01 and P < .05 was observed when 0.4, 0.8, 1.2, 1.6, 2.0, 4.0, and 10.0 µL and 8.0 µL of plasmin stock, respectively, were used (eFigure 3C). In general, experimental data obtained from Group F in which “excess” collagen was used (eFigure 4) produced similar results as Group D (eFigure 3). Variations in statistical differences as highlighted above likely relate to difficulty in fully controlling all experimental parameters and in estimates of the lysis zone. Finally, results obtained from the statistical analysis of Group B (Table) that involved systems with different amounts of collagen and fibrin — in other words, some cases without collagen — or equal amounts of total substrate (ie, collagen and/or fibrin) of the 2 combined showed a statistically significant difference (P < .01) for all plasmin concentrations after 30 minutes and 24 hours (eFigure 3A, C), respectively. Results obtained after 12 hours when 0.4, 0.8, 1.2, 1.6, 2, 4.0, 8.0, and 10.0 µL and 6.0 µL were used also showed that aprotinin significantly inhibited fibrinolysis at P < .01 and P < .05, respectively.

Impact of bromelain on fibrinolysis: comparisons between systems for fibrinolytic activity (Groups C, E, and G). The data obtained from these systems were statistically compared to reveal any significant differences due to the role of bromelain in promoting fibrin degradation. The data collected from Group C (Table) showed the presence of bromelain has no statistically significant effect (P < .05) in promoting fibrinolysis after 30 minutes (eFigure 3A) for all plasmin concentrations used. After 12 hours (eFigure 3B), the presence of bromelain showed no statistically significant difference (P < .05) in the rate of fibrinolysis when 0.4, 2.0, 4.0, and 8.0 µL volumes of plasmin stock were used, but a significant difference was observed (P < .01) when 0.8 µL and 10.0 µL volumes were used and (P < .05) when 1.2, 1.6, and 6.0 µL volumes of plasmin stock were used. This variation in statistical significance with respect to plasmin concentration suggested perhaps only a slight effect of bromelain at this time point. The bromelain effect was further monitored after 24 hours (eFigure 3C), and the results obtained from statistical analysis showed a significant difference (P < .01 or P < .05) for each of the plasmin concentrations used.

Analysis of the data from Group E (Table) after 30 minutes showed a statistically significant difference (P <.01) for all plasmin concentrations (eFigure 3A). Although the effects of bromelain after 12 hours showed a statistically significant difference in fibrinolysis (P < .01) when 0.8, 1.2, 1.6, 4.0, 6.0, 8.0, and 10.0 µL volumes were used, and (P < .05) when the 2 µL volume of plasmin stock was used (eFigure 3B), no significant difference was seen when the 0.4 µL volume of plasmin stock was used. After 24 hours, statistical analysis showed the presence of bromelain significantly increased fibrinolysis (P < .01) for the 0.4, 0.8, 1.2, 1.6, 2.0, 4.0, 6.0, and 8.0 µL volumes and (P < .05) for the 10 µL volume of plasmin stock (eFigure 3C).

The systems in Group G (Table) were also compared statistically, and the results obtained from data showed no significant difference (P < .05) for any of the plasmin stock volumes used after 30 minutes (eFigure 4A); however, bromelain was found to significantly increase fibrinolysis after 12 hours (P < .01) when plasmin stock volumes of 0.8, 1.2, 1.6, 2.0, 6.0, 8.0, and 10.0 µL were used and (P < .05) for the 0.4 µL volume. No statistically significant difference was found when 4 µL volume of plasmin stock was used (eFigure 4B), but the overall effect across the entire plasmin concentration range suggests bromelain increased fibrinolysis over this 12-hour timeframe. Statistical analysis of data after 24 hours (eFigure 4C) showed a significant difference (P < .01) for all plasmin volumes, thus showing bromelain significantly increases fibrinolysis.

The presence of plasmin inhibitors that could disrupt the normal fibrinolysis process has been implicated in excess collagen synthesis.²⁸ To examine such disruption, in vitro laboratory trials were conducted to observe how aprotinin affects the breakdown of fibrin clots in physical models containing type 1 collagen in addition to fibrin. The results (eFigures 3, 4) are consistent with those from a previous study by Tuan et al²⁰ who showed increased levels of plasmin inhibitors adversely affect fibrin degradation, thus potentially making the tissue repair process more susceptible to collagen accumulation. Further, increased inhibition of fibrinolysis has been a symbol of fibrotic disorders in tissues and organs. A study by Eitzman et al²⁹ revealed increased activities of plasmin inhibitors led to extended accumulation of fibrin in the wound area. Since fibrin degradation products assist in recruiting inflammatory cells, an inhibition of fibrinolysis would be expected to result in tissue fibrosis; this has been implicated in promoting the formation of keloids following a prolonged inflammatory response.²⁰ 

The effect of bromelain in fibrinolysis was also tested in the described in vitro gel systems. The results as illustrated in eFigures 3, 4 confirm those from a previous report³⁰ on the therapeutic applications of bromelain in promoting fibrinolysis, and the associated benefits of bromelain might be enhanced when bromelain is administered in high doses for about a 24-hour timeframe based on the work described herein.

The limitation of this study primarily relates to the relative simplicity of the wound model used, which does not incorporate chemical sensors and cells. Such a model was deemed effective at exploring the interplay of plasmin and promoters or inhibitors of plasmin on fibrinolysis in the absence or presence of collagen. To completely understand how collagen, aprotinin, and bromelain affect fibrinolysis and wound healing in entirety, future studies could utilize expanded versions of the physical model described herein and in vivo models.

This project explored the role of fibrinolysis in in vitro physical models of early-phase wounds. The models incorporated fibrin or a fibrin-collagen mix that was immobilized in an agarose gel. Results revealed a dose- and time-dependent response to plasmin that was allowed to diffuse through the gels. In addition, fibrinolysis was slightly reduced in systems having excess collagen, and greatly reduced when aprotinin, an inhibitor of plasmin, was added to the system. In contrast, fibrinolysis was increased when bromelain was added to the system. These findings support the notion that early excess accumulation of collagen (without a corresponding breakdown) with certain fibrotic disorders has been shown to alter fibrinolysis. Since adverse alteration of fibrinolysis has been reported to affect the inflammatory phase of wound healing,³¹ the feature of early excess collagen during wound healing may fairly rationalize the incidence of fibrotic disorders as seen in abnormal scars. Based on these findings, it is crucial to direct future studies in exploring how the process of fibrinolysis can be controlled to facilitate the wound healing process. Results can also be used to develop mathematical models to help predict efficient pathophysiology of tissue fibrosis and to find solutions that would ultimately prevent fibrotic disorders during tissue repair.

The authors acknowledge insightful contributions and guidance from Dr. John Gunderson. 

From the Tennessee Technological University, Cookeville, TN

Address correspondence to:
Jonathan R. Sanders, PhD
Prescott Hall, Room 214
1020 Stadium Drive
Tennessee Technological University
Cookeville, TN 38505
rsanders@tntech.edu

Disclosure: The authors received financial support for this study from the Office of Research and Graduate Studies of Tennessee Technological University (TTU; Cookeville, TN). This work has been included, in large part, in an electronic thesis prepared by the first author (Mr. Chukwuemeka) that was approved by his MS committee and the Graduate School at TTU for publication in the ProQuest Thesis and Dissertation database (https://search.proquest.com/docview/1552484997), which is a requirement for graduate students at TTU. The authors disclose no financial or other conflicts of interest.