Abstract
Tendons are commonly injured connective soft tissues characterized by an ineffective healing response that results in scar formation and loss of functional and structural properties. Naturally occurring extracellular matrix (ECM) constructs have become a promising therapeutic for tendon injuries due to their capacity to harness a complex biological environment. However, in tendon, the ECM properties needed for improved healing remain unknown. Interestingly, we have determined that the improved tendon healing response of the scarless-healing MRL/MpJ is driven by intrinsic properties with therapeutic potential to modulate the proliferative and morphological behavior of cells derived from a canonically healing model in vitro. We hypothesize that a distinct composition of ECM deposited during the early healing response of the MRL/MpJ will harnesses the biological cues to stimulate improved structure and function in vivo of canonically healing B6 mice. Accordingly, MRL/MpJ and B6 patellar tendons were injured via midsubstance punch defects. Healing tendons were isolated after 3 or 7 days and encapsulated in PEG-4MAL hydrogels to develop ECM-derived therapeutic constructs. Constructs were then introduced into B6 mice as a treatment following full thickness midsubstance-punch injuries. Treatment with ECM-derived constructs from MRL/MpJ tendons after 7-days post-injury (M7) resulted in improved matrix alignment, tissue stiffness, decreased collagen III content and improved cell morphology in B6 tendons after 6 weeks post-injury. Furthermore, proteomic analysis showed that M7 contained a unique compositional profile rich in glycoproteins, thereby elucidating a valuable naturally-derived platform for the treatment of tendon injuries. Overall this work highlights promising targets for future therapeutic development and tissue engineering applications.
Keywords: extracellular matrix, scarless healing, tendon therapeutics
1 |. INTRODUCTION
Tendons are comprised of a host population of elongated tenocytes, as well as numerous other cell types such as chondrocyte-like cells, and tendon progenitor cells residing in a highly aligned extracellular matrix (ECM) that is responsible for providing mechanical integrity during loading.1 Following injuries, mammalian adult tendons heal by scar deposition at the injury site, represented by a disorganized structural matrix, altered tissue composition, loss of mechanical integrity, increased cellularity and a shift in cell morphology towards a rounded pathological phenotype.2–4 To improve the aberrant healing response following tendinopathies, therapeutics that are based on principles from embryonic healing, mechanical stimulation, specific growth factor delivery and synthetic tissue engineering have been utilized to identify potential targets to encourage tissue regeneration.5–9 However, while these treatments have provided valuable insight into the tendon healing response, their interrogation of isolated components necessary for a healthy tendon environment do not adequately emulate the biochemical complexity of the ECM necessary to stimulate improved healing.
Utilizing naturally occurring ECM-derived therapeutics has become an attractive approach because of their ability to integrate complex networks of proteins into the injury site, thereby promoting recovery of the structural and compositional properties of native tissues.10 Accordingly, ECM-derived scaffolds have been used to improve acute injuries in vivo in skeletal muscle and heart.11,12 Still, while ECM scaffolds have become a promising tool in regenerative medicine, the use of these constructs to treat tendon injuries has been limited, and therefore the optimal composition to develop ECM derived therapeutics remains to be elucidated.
Interestingly, the Murphy Roths Large (MRL/MpJ) mouse has recently been identified as a model of adult mammalian scarless tendon healing.13–15 More specifically, MRL/MpJ tendon healing is characterized by an early deviation in ECM composition post-injury followed by improved mechanical, structural and cellular behavior in vivo, compared to canonically healing C57Bl/6 (B6). Identification of the improved healing capacity of MRL/MpJ tendons provides a model wherein the synergistic behavior of multiple biochemical factors that lead to regeneration could be interrogated. In this manner, we have previously utilized organ culture to determine that the innate tendon properties of the MRL/MpJ act as a driver of its improved healing response even when isolated from the systemic environment.16 Additionally, we have identified that the compositional properties of the ECM obtained from MRL/MpJ tendons after 7-days post-injury (M7), harness the necessary biochemical cues to improve the proliferative and morphological behavior of canonically healing cells in vitro.16 However, while these findings highlight the potential role of the innate MRL/MpJ ECM in modulating cell behavior of a non-regenerative model following tendinopathies; the ability of these constructs to stimulate improved tendon healing in vivo remains unknown.
Moreover, while we have identified the ability of M7 constructs to modulate the behavior of scar-mediated healing B6 cells in vitro, the ECM deposited during the beginning of the proliferative stage of healing, after 3-days post-injury, contains a bioactive cocktail of growth factors, cytokines, and glycosaminoglycans (GAGs) that are transiently remodeled throughout the healing response, towards the deposition of larger structural proteins seen after 7 days.17,18 Thus, to assess the in vivo therapeutic potential of these two distinct ECM environments, we developed MRL/MpJ and B6 decellularized ECM-constructs from matrix deposited after 3 and 7-days post-injury. We hypothesize that the unique composition of MRL/MpJ ECM-derived therapeutics harnesses the biological cues to drive an improved structural and functional healing response in B6 mice by 6-weeks post-injury. Consequently, the objectives of this study are to (1) identify an ECM-derived therapeutic platform that leads to the best improvement in the structural, compositional, cellular and mechanical function of scar mediated healing and (2) identify the compositional profile responsible for providing the biochemical cues that harness the ability to improve tendon healing in vivo.
To interrogate our hypothesis, we assessed the long-term structural and mechanical integrity of B6 tendons treated with our MRL/MpJ and B6 derived ECM constructs following a 1-mm acute midsubstance punch injury. We utilized tendon matrix alignment and mechanical properties to identify the most effective ECM construct and subsequently isolated this treatment group for further analysis of its therapeutic effect on healing tendon composition of collagen III and GAGs, which are significantly increased during the scar mediated canonical healing response.18,19 Moreover, cell elongation, which is characteristic of healthy tendon cells, was also assessed and compared to injured-untreated tendons. Lastly, to elucidate the unique compositional cocktail responsible for the improved tendon healing response following treatment, the proteomic profile was obtained for the most and least effective therapeutic constructs. Comparison of the protein composition of the most effective ECM with the one found to be least effective will further inform a compositional template for future development of tissue engineering therapeutics.
2 |. METHODS
2.1 |. Patellar tendon punch injury and sample dissection
Under IACUC approval, 16-week-old mice (n = 227 Jackson Laboratories, ME) (n = 57 bred in-house with sire and dame from Jackson Laboratories, ME) were placed under isoflurane, (2% by volume, 0.3 L/min). Briefly, a skin incision exposed the left patellar tendon of MRL/MpJ (n = 52) and B6 (n = 232) mice. A medial and lateral incision on the tendon edges was introduced utilizing an 11-blade scalpel to allow for a polyurethane coated stainless-steel backing to be positioned below the patellar tendon. Following, a 1-mm biopsy punch defect was created in the tendon midsubstance.13 The backing was removed and skin sutured utilizing 6–0 Prolene sutures. Analgesic was administered, (Buprenorphine, 0.2 mg/kg) and mice resumed cage activity (12-hour light cycle, unlimited access to food/water). Mice were kept under standard housing conditions and sacrificed at 3-days, 1-week, or 6-weeks post-injury via CO2.13
2.2 |. Development of therapeutics and PEG-4MAL hydrogel delivery system
MRL/MpJ and B6 mice were sacrificed after 3 (n = 24,24) or 7 days (n = 24,24) post-injury. The midsubstance region from their left injured and right uninjured patellar tendons was collected. Following dissection, tissue was frozen in liquid nitrogen and stored at −80°C. All ECM samples were decellularized utilizing an adapted previously described protocol.16,20 Briefly, samples were submerged in a 50 nM Latrunculin B (BioVision, Milpitas, CA) bath for 2 hours, then submitted to a progression of 0.6 M KCl followed by 1.0 M KI solutions for 2 hours each. 30-minute deionized (DI) H2O washes occurred between all KCl and KI baths. A 12-hour wash was then performed in DIH2O, after-which samples repeated the KCl, DIH2O, and KI progression. A final 48-hour PBS + 1 kU/mL Pierce Universal Nuclease (Thermo Fisher Scientific, Waltham, MA) wash was performed. Samples were then lyophilized for 48 hours. All incubations were done under agitation (450 rpm) and all solutions contained 1X Halt™ protease inhibitor. Lyophilized tendons were pulverized utilizing a bead-ruptor device using 2.8-mm in diameter ceramic beads for 4 cycles at 15 seconds/cycle. Following, 1.4-mm in diameter ceramic beads were utilized for 4 additional cycles at 15 seconds/cycle before being placed in −80°C.
Based on previous studies, 20% (w/v) PEG-4MAL (Laysan Bio Inc, Arab, AL) hydrogels were formed.21–23 Briefly, PEG-4MAL was crosslinked with non-degradable dithiothreitol crosslinker and MMP-degradable GCRDVPMSMRGGDRCG peptides (AAPPTec, LLC, Louisville, KY) (50% DTT, 50% VPM crosslinker ratio). 200 ± 13 μg of pulverized, B6–3-day-provisional-ECM (B3), B6–7-day-provisional-ECM (B7), MRL/MpJ-3-day-provisional-ECM (M3) or MRL/MpJ-7-day-provisional-ECM (M7) was suspended in the cross-linker solution and combined with the PEG-4MAL to create 2-μl hydrogel-therapeutic complexes. To assess whether innate differences in therapeutic potential between naïve MRL/MpJ and B6 ECM extended to the in vivo healing outcome, uninjured MRL/MpJ (MU) and B6 (BU) ECMs were also included in this study and prepared as described above. Hydrogel-therapeutic complexes were inserted into B6 left patellar tendons through the previously described lateral incision introduced to create the 1 mm punch defect. A subset of B6 tendons were left untreated post-injury to comprise the injured-untreated control group, and another was treated with PEG-4MAL delivery system alone to comprise the vehicle-only group. Right limbs were utilized as uninjured controls. Tendons healed for 1 (n = 56) or 6 weeks (n = 124), before sacrifice, at which point tissues were dissected, flash frozen in liquid nitrogen, and placed at −80°C until the time of respective analysis. Finally, following sacrifice, mice treated with our ECM-derived constructs were randomly assigned towards structural or mechanical assessment (Figure 1).
FIGURE 1.
Schematic of therapeutic development and treatment. A, ECM-derived construct development involved patellar tendon isolation, decellularization, and tissue pulverization. Pulverized tissue was then encapsulated within a PEG-4MAL hydrogel and delivered into B6 hosts immediately following a 1-mm midsubstance punch injury. B, ECM-derived constructs were developed from MRL/MpJ or B6 uninjured ECM, 3-day-provisional ECM, or 7-day-provisional ECM. C, Injured B6 tendons treated with ECM-derived constructs were submitted to structural, mechanical or histological analysis at 1 or 6 W to assess the efficacy of these constructs at modulating the canonical tendon healing response
2.3 |. Structural assessment
As previously described, B6 tendons (n = 7–8/group) were fixed utilizing a zinc-buffered formalin solution before embedding in paraffin wax and sectioning (10 μm/section).16 To analyze the auto-fluorescent structural properties of the tendon matrix, unstained sections were imaged via fluorescent microscopy, at ×4 magnification.16 A custom MATLAB script was used to identify matrix disorganization at the tendon midsubstance for uninjured, treated, and injured-untreated samples.13,16
2.4 |. Mechanical testing
Treated B6 tendons, injured-untreated controls, and uninjured controls (n = 8/group) were clamped and submerged in a 1X PBS bath at room temperature. Tendons were subjected to an initial load of 0.15 N (Bose ElectroForce 3200 with 10 lb load cell, Bose Corporation, Eden Prairie, MN). Preconditioning was then performed (15 cycles at 1% strain, 1 Hz). A 5% strain was applied and held for 300 seconds to assess stress relaxation. Tendons then recovered for 300 seconds under no load. Preconditioning was repeated prior to applying a strain to failure at a rate 0.1% strain/second to assess ultimate load and stiffness.16 Mechanical testing data was collected every 0.05 seconds.
2.5 |. Selection of the best and worst performing ECM-derived therapeutic constructs
Late stage structural and mechanical assessment at 6-weeks post-injury, were utilized to select the most and least effective ECM-derived therapeutics. The treatment group resulting in the highest improvement compared to injured-untreated controls was deemed the most effective. Due to the high number of groups that did not result in enhanced matrix alignment, stiffness or ultimate load, the group with the lowest overall improvement, as determined by the sample means, was deemed the least effective therapeutic.
Samples treated with the most effective therapeutic were further analyzed to assess deviations in tendon composition and cell behavior compared to injured-untreated controls. Additionally, to identify the compositional profile of the ECM-derived construct that led to improved healing and compare it to the one that had the most detrimental outcomes, the best and worst performing constructs were submitted to proteomic analysis.
2.6 |. Immunohistochemistry and histology
Paraffin sections (10 μm/section) (n = 7–8/group) were stained with collagen III antibody according to manufacturer’s protocol. Briefly, samples were de-paraffinized, rehydrated, placed in Pro-K solution for 3 minutes and blocked (Dako North America, Inc, Carpinteria, CA) for 30 minutes at 37°C. Samples were incubated with a collagen III antibody for 1-hour at 37°C (#AB1832; Chemicon, Temecula, CA; 1:500 dilution). Vector-Anti Rabbit secondary (Vector Laboratories, Burlingame, CA) was applied for 30 minutes. Diaminobenzidine (DAB) (Sigma-Aldrich, St. Louis, MO) was added to visualize the stain. Samples were submerged in Toluidine Blue for 3 minutes as a counter stain. DIH2O washes were performed between all steps.
Collagen III imaging was performed at ×4 magnification via brightfield microscopy. As previously described, RGB images were imported into ImageJ and decomposed into Red, Green and Blue channels.24 The Blue channel, representing the IHC signal, was isolated for further analysis. Images were pooled into a stack and a pixel intensity histogram of the merged signal was obtained to establish a threshold for the detection of “on/stained” pixels. The peak of the histogram signal was identified and utilized as this threshold. Values above the threshold intensity were labeled as “on/stained” to reduce the effect of background and shared high intensity pixels from analysis. The midsubstance region of the tendon was isolated and percent stain per area (“on”/total pixels) was measured by a blinded user.
For GAG analysis, additional sections (n = 7–8/group) were stained with Toluidine Blue (Electron Microscopy Sciences, Hatfield, PA) for 20 minutes. Imaging protocol was performed as described above. Mean intensity measurements of this colorimetric stain were taken to obtain semiquantitative analysis of GAG content. Toluidine stained sections were then imaged at ×40 magnification.13 A square grid was superimposed onto each image utilizing ImageJ, and a random number generator was utilized to select five regions within the grid for cell analysis. Nuclear aspect ratio (major/minor axis) was measured by a blinded user (n = 6–8/group).
2.7 |. Proteomic analysis of M7 and BU composition
ECM-derived constructs were homogenized as described (n = 4/group). A solution of 400 μL PBS pH 7.4, 6 M Urea, 2 M Thiourea, 10 mM DTT was added for initial digestion. Samples were vortexed (1200 rpm, at room temperature) for 2.5 hours, centrifuged (12 000g at room temperature) for 5 minutes and supernatant was removed. The insoluble pellet was additionally digested with 400 μL PBS pH 7.4, 6 M Guanidine-Hydrochloride (Gdn-HCL), and 10 mM DTT for 1-hour. Centrifugation and supernatant collection were performed as described. All fractions were concentrated to a volume of 90–100 μL. Sample concentration was determined via Bradford assay, and further quantified by running a precast NOVEX 10% Bis/Tris mini-gel (Invitrogen, Carlsbad, CA). Based on the gel quantitation, 4 μg protein for the Urea fractions and 0.8 μg protein for the Gdn-HCL fractions were used for in-solution digestion. Sample volumes were adjusted with 10 mM DTT reducing agent to a final volume of 16ul for Urea Fractions and 65ul for Gdn-HCL fractions. The digested fractions for each sample were reconstituted in 20-μL of 0.5% FA for nano-LC-ESI-MS/MS analysis (Orbitrap Fusion™ Tribrid™, Thermo Fisher Scientific, San Jose, CA). Proteome Discoverer 2.3 (Thermo Fisher Scientific, Bremen, Germany) was utilized to perform database search using Sequest HT searching engine against MusMusculus Uniprot 2016 database. Data from Urea and Gdn-HCL digestions were pooled for analysis of each individual sample. Samples with at least two unique peptides and ≥2-fold change between groups were included for statistical analysis.25–27 Data obtained from the Proteome Discoverer 2.3 Software was used to identify protein classes.
2.8 |. Statistical analysis
Data is shown as mean ± standard deviation. To assess the effectiveness of ECM-derived constructs at improving the canonical tendon healing response, a One-way analysis of variance (ANOVA) with Dunnett’s multiple comparisons test was utilized to assess structural and mechanical properties of treated groups compared to injured-untreated controls. Additionally, to assess the effectiveness of the ECM-derived constructs at restoring the structure and mechanics of treated tendons to pre-injury levels, groups were compared to uninjured controls using a One-way ANOVA with Dunnett’s multiple comparisons test. For histological and cell behavior assessment, individual unpaired student t test were performed between the best performing therapeutic and injured-untreated controls. For proteomic analysis comparing the concentrations of individual proteins between the best and worst performing construct, a non-parametric Kruskal Wallace ANOVA was performed on proteins that were expressed in at least three samples per group. The false discovery rate was controlled post-hoc via a false discovery rate (FDR) corrected method of Benjamini and Hochberg (*q < 0.05, #q < 0.1). All statistical analysis was performed on raw data prior to normalization.
3 |. RESULTS
Surgery was tolerated and all animals survived until the designated time-point.
3.1 |. Structural and functional assessment of healing tendons following treatment
Analysis of tendon structure showed that by 6-weeks post-injury matrix alignment was improved for the M7 (P = .0205) treated groups compared to injured-untreated controls. No differences were found between treated groups and injured-untreated controls at 1-week post-injury (Figure 2).
FIGURE 2.
Matrix disorganization in B6 injured tendons at (A) 1 week (1 W) and (B) 6 weeks (6 W) following treatment with PEG-4MAL vehicle alone or PEG-4MAL in addition to ECM-derived therapeutic treatment from MRL/MpJ Uninjured ECM (MU), MRL/MpJ 3-day-provisional-ECM (M3), MRL/MpJ 7-day-provisional-ECM (M7), B6 Uninjured ECM (BU), B6 3-day-provisional-ECM (B3), or B6 7-day-provisional-ECM (B7). Tendons treated with M7 showed improvements in matrix alignment compared to injured-untreated controls at 6 W. Dashed line represents mean matrix disorganization for injured-untreated controls. #P ≤ .1, *P ≤ .05
Mechanical assessment showed that only the M7-treated group displayed higher stiffness compared to injured-untreated controls (P = .0779) by 6 weeks (Figure 3A). Surprisingly, PEG-4MAL delivery system treatment alone showed significantly lower stiffness compared to injured-untreated controls (P = .0259). No differences in ultimate load were found between treated groups and injured-untreated controls(Figure 3B). No differences were found in the stress relaxation of B6treated tendons compared to injured-untreated controls (Figure 3C). Lastly, all treated groups exhibited significantly higher levels of matrix disorganization, and significantly lower stiffness and ultimate load than uninjured controls at 6 W (Table 1) (P < .05 for all groups compared to uninjured).
FIGURE 3.
Mechanical assessment of injured B6 patellar tendons after 6-weeks (6 W) post-treatment with vehicle alone or vehicle with ECM-derived therapeutic constructs. A, M7-treated tendons showed increased stiffness while treatment with PEG-4MAL alone exhibited decreased stiffness compared to injured-untreated controls. B, No differences were found in the ultimate load of treated tendons compared to injured-untreated controls. C, No differences were found in the stress relaxation of treated tendons compared to injured-untreated controls. Dashed line represents mean mechanical data for injured-untreated controls. #P ≤ .1, *P ≤ .05. B3, B6 3-day-provisional-ECM; B7, B6 7-day-provisional-ECM; BU, B6 ininjured ECM; M3, MRL/MpJ 3-day-provisional-ECM; M7, MRL/MpJ 7-day provisional ECM; MU, MRL/MpJ uninjured ECM
TABLE 1.
Assessment of B6 tendons treated with ECM-derived constructs compared to uninjured controls
| Uninjured | PEG-4MAL | MU | M3 | M7 | BU | B3 | B7 | |
|---|---|---|---|---|---|---|---|---|
| Matrix disorganization (%) | 5.94 ± 5.03 | 78.76 ± 14.9* | 66.8 ± 16.68* | 64.95 ± 28.72* | 44.95 ± 11.7* | 73.75 ± 9.62* | 51.74 ± 16.34* | 71.18 ± 23.08* |
| Ultimate load (N) | 7.93 ± 1.68 | 3.13 ± 0.92* | 4.25 ± 1.19* | 5.03 ± 1.13* | 5.13 ± 1.08* | 4.86 ± 1.33* | 4.60 ± 1.10* | 5.36 ± 1.02* |
| Stiffness (N/mm2) | 43.98 ± 12.67 | 15.36 ± 3.04* | 23.37 ± 4* | 28.13 ± 6.73* | 30.51 ± 6.88* | 21.06 ± 5.67* | 25.88 ± 3.82* | 25.01 ± 5.46* |
| Stress relaxation (%) | 24.7 ± 4.80 | 25.41 ± 4.23 | 26.9 ± 4.56 | 24.05 ± 4.51 | 28.11 ± 5.82 | 27.25 ± 8.21 | 24.98 ± 5.31 | 25.13 ± 5.67 |
Note: All groups exhibited significantly higher matrix disorganization, and lower stiffness and ultimate load than uninjured controls at 6W regardless of treatment.
Abbreviations: B3, B6 3-day-provisional-ECM; B7, B6 7-day-provisional-ECM; BU, B6 Uninjured ECM; M3, MRL/MpJ 3-day-provisional-ECM; M7, MRL/MpJ 7-day-provisional-ECM; MU, MRL/MpJ uninjured ECM.
Significant difference from uninjured control.
3.2 |. Selection of the best and worst performing ECM-derived therapeutic constructs
M7 was identified as the most effective therapeutic since tendons treated with these constructs were the only group that exhibited both improved stiffness (P = .0779) and matrix alignment (P = .0205) compared to injured-untreated controls. Furthermore, BU was identified as the least effective therapeutic, since treatment with this construct resulted in the highest degree of matrix disorganization and lowest stiffness based on group means (Table 2).
TABLE 2.
Selection of the least effective therapeutic based on group means of treatment groups (MU, M3, BU, B3, and B7) that did not show statistical improvements compared to injured-untreated controls in matrix alignment, stiffness and ultimate load after 6-weeks post-treatment
| ECM-derived constructs | Matrix disorganization (%)* | Stiffness (N/mm2)* | Ultimate load (N)* |
|---|---|---|---|
| B3 | 51.74 | 25.88 | 4.60 |
| M3 | 64.95 | 28.13 | 5.03 |
| MU | 66.8 | 23.37 | 4.25 |
| B7 | 71.18 | 25.01 | 5.36 |
| BU | 73.75 | 21.06 | 4.86 |
Note: BU was found as the least effective treatment.
Abbreviations: B3, B6 3-day-provisional-ECM; B7, B6 7-day-provisional-ECM; BU, B6 uninjured ECM; M3, MRL/MpJ 3-day-provisional-ECM; MU, MRL/MpJ uninjured ECM.
Mean values per group.
3.3 |. Compositional and cell behavior assessment of healing tendons following treatment with M7 construct
There was a decrease in the percent area of the injured region that stained positive for collagen III content in the healing M7-treated group at both 1 week (P = .062) and 6 weeks (P = .0003) compared to injured-untreated controls (Figure 4).
FIGURE 4.
Assessment of collagen III content. A,B, M7-treated B6 tendons exhibited a decrease in the percent area of the injured region that stained positive for collagen III content compared to injured-untreated controls at both 1-week (1 W) and 6-weeks (6 W) post-treatment. C, Representative analysis of histological sections. #P ≤ 0.1, *P ≤ .05. (B6 tendons treated with MRL/MpJ 7-day-provisional-ECM [M7]). Scale bar = 200 μm
GAG content, as assessed by colorimetric Toluidine blue staining, was lower in M7-treated tendons after 6 weeks compared to injured-untreated controls (P = .0436). No differences in GAG content between M7-treated tendons and injured-untreated controls was found after 1 week (Figure 5).
FIGURE 5.
Glycosaminoglycan (GAG) content assessment via toluidine blue staining. A, No differences were found after 1 week (1 W). B, M7-treated tendons exhibited a decrease in GAG content compared to injured-untreated controls at 6-weeks (6 W) post-treatment. C, Representative histological sections at 1 W and 6 W. *p ≤ .05. (B6 tendons treated with MRL/MpJ 7-day-provisional-ECM [M7]). Scale bar = 200 μm
Samples treated with M7 exhibited increased cell aspect ratio compared to injured-untreated controls at both 1 (P = .0852) and 6 weeks (P = .0118) post-injury (Figure 6).
FIGURE 6.
Assessment of cell morphology. A,B, M7-treated tendons exhibited a more elongated cell morphology compared to injured-untreated controls at 1-week (1 W) and 6-weeks (6 W) post-treatment. C, Representative histological sections at 1 W and 6 W. #P ≤ .1, *P ≤ .05. (B6 tendons treated with MRL/MpJ 7-day-provisional-ECM [M7]). Green arrows point to elongated cells, while red arrows point to more rounded cells. Scale bar = 50 μm
3.4 |. Compositional protein profile of the most (M7) and least (BU) effective therapeutics
Proteomic assessment of M7 and BU identified a total of 176 proteins present combined between both groups. A total of 11 proteins differed in concentration between M7 and BU (P < .1). From these proteins two were glycoproteins, one was an ECM associated protein, and eight were cell processes related proteins. All proteins found to be different between these groups were higher in M7 compared to BU (Table 3). Furthermore, 74 proteins were identified only in M7, two of these were glycoproteins, 12 were ECM associated proteins, and 60 were cell processes related (Table 4).
TABLE 3.
Proteomic assessment of most and least effective therapeutics
| Protein name | Accession | MW, kDa | BU abundance | M7 Abundance | Q value |
|---|---|---|---|---|---|
| Glycoproteins | |||||
| Fibrinogen β chain | Q3TGR2 | 54.7 | 3.08E + 06 ± 2.01E + 06 | 1.40E + 08 ± 1.74E + 08 | 0.0723 |
| Fibronectin | B9EHT6 | 250.1 | 1.24E + 07±9.59E + 06 | 2.90E + 08 ±2.72E + 08 | 0.0723 |
| ECM associated | |||||
| Myosin-9 | Q8VDD5 | 226.2 | 2.14E + 06±2.95E + 06 | 7.53E + 07 ± 6.07E + 07 | 0.0723 |
| Cell processes | |||||
| α-Globin 1 | Q91VB8 | 15.1 | 1.56E + 06±6.22E + 05 | 9.12E + 07 ± 9.30E + 07 | 0.0723 |
| ATP synthase subunit β, mitochondrial | P56480 | 56.3 | 2.69E + 06± 2.18E + 06 | 9.43E + 07 ±4.50E + 07 | 0.0723 |
| Glyceraldehyde-3-phosphate dehydrogenase | A0A0A0MQF6 | 38.6 | 2.14E + 06±1.69E + 06 | 5.93E + 07 ± 3.79E + 07 | 0.0888 |
| Histone H2A type 3 | Q8BFU2 | 14.1 | 7.80E + 06± 8.97E + 06 | 3.71E + 08 ±1.57E + 08 | 0.0723 |
| Histone H3.2 | P84228 | 15.4 | 1.76E + 07±2.05E + 07 | 3.97E + 08 ±1.86E + 08 | 0.0723 |
| Histone H4 | B2RTM0 | 11.4 | 2.39E + 07±2.39E + 07 | 7.59E + 08 ±4.12E + 08 | 0.0723 |
| Tubulin β−5 chain | P99024 | 49.6 | 2.07E + 06±1.86E + 06 | 1.91E + 08 ±1.04E + 08 | 0.0723 |
| Vimentin | P20152 | 53.7 | 4.54E + 07± 5.81E + 07 | 8.30E + 08 ± 1.88E + 08 | 0.0723 |
Note: A total of 11 proteins were higher in the M7 therapeutic compared to the BU construct. Abbreviations: BU, B6 uninjured ECM; M7, MRL/MpJ 7-day-provisional-ECM.
TABLE 4.
Proteomic assessment of most and least effective therapeutics
| Accession | Description | MW, kDa |
|---|---|---|
| Glycoproteins | ||
| Q62009 | Periostin | 93.1 |
| Q61554 | Fibrillin-1 | 312.1 |
| ECM associated proteins | ||
| Q3ULT2 | Actinin α4 | 104.9 |
| P10107 | Annexin A1 | 38.7 |
| P48036 | Annexin A5 | 35.7 |
| P14824 | Annexin A6 | 75.8 |
| P11499 | Heat shock protein HSP 90-β | 83.2 |
| F8WJ05 | Inter-a-trypsin inhibitor heavy chain H1 | 101.6 |
| Q61704 | Inter-a-trypsin inhibitor heavy chain H3 | 99.3 |
| P17742 | Peptidyl-prolyl cis-trans isomerase A | 18 |
| P50543 | Protein S100-A11 | 11.1 |
| P27773 | Protein disulfide-isomerase A3 | 56.6 |
| Q9Z204 | Heterogeneous nuclear ribonucleoproteins C1/C2 | 34.4 |
| P18760 | Cofilin-1 | 18.5 |
| Cell processes | ||
| Q5M9K7 | 40S ribosomal protein S10 | 18.9 |
| P14131 | 40S ribosomal protein S16 | 16.4 |
| Q5M9L7 | 40S ribosomal protein S17 | 15.5 |
| Q561N5 | 40S ribosomal protein S18 | 17.7 |
| P62267 | 40S ribosomal protein S23 | 15.8 |
| P62855 | 40S ribosomal protein S26 | 13 |
| P97351 | 40S ribosomal protein S3a | 29.9 |
| Q545X8 | 40S ribosomal protein S4 | 29.6 |
| D3YYM6 | 40S ribosomal protein S5 (Fragment) | 20.4 |
| Q5BLK1 | 40S ribosomal protein S6 | 28.7 |
| P14206 | 40S ribosomal protein SA | 32.8 |
| P63038 | 60kDa heat shock protein, mitochondrial | 60.9 |
| Q9CXW4 | 60S ribosomal protein L11 | 20.2 |
| P47963 | 60S ribosomal protein L13 | 24.3 |
| E9Q5A0 | 60S ribosomal protein L13a | 48.3 |
| Q9CR57 | 60S ribosomal protein L14 | 23.5 |
| Q58EW0 | 60S ribosomal protein L18 | 21.6 |
| Q9CQM8 | 60S ribosomal protein L21 | 18.6 |
| P62830 | 60S ribosomal protein L23 | 14.9 |
| Q8BP67 | 60S ribosomal protein L24 | 17.8 |
| P14115 | 60S ribosomal protein L27a | 16.6 |
| P27659 | 60S ribosomal protein L3 | 46.1 |
| P62889 | 60S ribosomal protein L30 | 12.8 |
| Q9D1R9 | 60S ribosomal protein L34 | 13.3 |
| P14148 | 60S ribosomal protein L7 | 31.4 |
| P12970 | 60S ribosomal protein L7a | 30 |
| P47738 | Aldehyde dehydrogenase, mitochondrial | 56.5 |
| Q8C2Q8 | ATP synthase subunit gamma | 30.2 |
| Q9DB20 | ATP synthase subunit O, mitochondrial | 23.3 |
| Q542X7 | Chaperonin subunit 2 (β), isoform CRA_a | 57.4 |
| Q8BMK4 | Cytoskeleton-associated protein 4 | 63.7 |
| Q99LC5 | Electron transfer flavoprotein subunit α, mitochondrial | 35 |
| Q58E64 | Elongation factor 1-α | 50.1 |
| Q3UAD6 | Endoplasmin | 92.4 |
| E9Q561 | Eukaryotic initiation factor 4A-II | 36.1 |
| Q8BTM8 | Filamin-A | 281 |
| P68040 | Guanine nucleotide-binding protein subunit β−2-like 1 | 35.1 |
| P63017 | Heat shock cognate 71 kDa protein | 70.8 |
| P02089 | Hemoglobin subunit β−2 | 15.9 |
| Q9Z2X1 | Heterogeneous nuclear ribonucleoprotein F | 45.7 |
| Q8C2Q7 | Heterogeneous nuclear ribonucleoprotein H | 51.2 |
| B2RVP5 | Histone H2A | 13.5 |
| P70168 | Importin subunit β−1 | 97.1 |
| Q9D646 | Keratin, type I cuticular Ha4 | 44.5 |
| P08249 | Malate dehydrogenase, mitochondrial | 35.6 |
| P35700 | Peroxiredoxin-1 | 22.2 |
| P09411 | Phosphoglycerate kinase 1 | 44.5 |
| P60335 | Poly(rC)-binding protein 1 | 37.5 |
| O54724 | Polymerase I and transcript release factor | 43.9 |
| Q922R8 | Protein disulfide-isomerase A6 | 48.1 |
| J3QMG5 | Protein Gm5786 | 31.2 |
| D3YWG1 | Protein Gm7964 | 50.9 |
| F6SVV1 | Protein Gm9493 | 21.9 |
| P52480 | Pyruvate kinase PKM | 57.8 |
| A2AVJ7 | Ribosome-binding protein 1 | 158.3 |
| P19324 | Serpin H1 | 46.5 |
| P38647 | Stress-70 protein, mitochondrial | 73.4 |
| P68369 | Tubulin α−1A chain | 50.1 |
| Q7TMM9 | Tubulin β−2A chain | 49.9 |
| Q922F4 | Tubulin β−6 chain | 50.1 |
Note: A total of 74 proteins were identified in M7 that were not present in BU.
Abberviations: BU, B6 uninjured ECM; M7, MRL/MpJ 7-day-provisional-ECM.
4 |. DISCUSSION
The role of the ECM as a tool to treat pathological conditions has been supported by studies in the heart, kidneys, and esophagus of non-regenerative models.28–30 In tendon, however, a hurdle towards the optimization of ECM-derived therapeutics has been the lack of knowledge regarding the effective matrix environment necessary to stimulate improved healing. Interestingly, we have shown that decellularized-ECM constructs from MRL/MpJ provide the necessary biochemical cues to modulate canonically healing B6 cell behavior in vitro towards the regenerative phenotype of MRL/MpJ cells during in vivo healing.16 Therefore, in this study we harnessed the compositional properties of the MRL/MpJ healing environment and assessed the ability of its ECM to improve scar-mediated tendon healing in vivo.
Following injury, the compositional properties of the tendon ECM continuously remodel over time, providing different cues that guide cell behavior throughout the stages of healing.17,18 Therefore, the unique biological contributions of ECMs harnessed from different early healing timepoints give rise to distinct therapeutic potentials depending on the time of ECM acquisition. Supporting this notion, we evaluated the therapeutic capabilities of ECM from tendons after 3 and 7-days post-injury and identified that MRL/MpJ-ECM after 7-days harnesses the potential to improve structural, mechanical, compositional, and cellular behavior of scar-mediated healing tendons.
The loss of structural and mechanical integrity resulting from the ineffective canonical tendon healing response, leads to high incidences of re-injury and tendon rupture during physical activity.2–4 Interestingly, our analysis showed that M7 treatment possesses the ability to improve the tendon healing response by stimulating the deposition of matrix with enhanced structural qualities. Still, the main physiological role of tendons is to transfer high levels of loading from muscles to bones during locomotion.18,19,31 Therefore, while matrix alignment is a key structural characteristic of healthy tissues, improvements in this metric, without functional assessment, are not sufficient to completely determine therapeutic efficiency.
Analysis of tissue stiffness showed that M7 treatment was capable of improving the functional outcome of the healing tendons. The increased stiffness following treatment could result in deviations to the mechanical environment experienced by cells compared to the injured-untreated environment throughout healing.32 In turn, it is possible that cells respond to these changes in mechanical stimulus by depositing higher quality matrix that allows tendons to more effectively resist deformation during sub-rupture loading. Interestingly, mechanical analysis revealed a lack of improvement in ultimate load regardless of treatment. While surprising, this lack of improvement could be due to the ineffective integration of newly deposited matrix at the injury site with the native tendon core, thereby resulting in the formation of stress concentrations at this interface that are susceptible to failure during high levels of loading. Additionally, no changes were identified in the stress relaxation of our treatment groups compared to injured-untreated, or uninjured controls. This finding was not surprising, as previous studies assessing the mechanical properties of canonically healing tendons have identified that structural and compositional changes following injury do not always translate to changes in the bulk tissue stress relaxation. For instance, others have identified that introduction of an incisional injury to patellar or flexor digitorum longus tendons did not result in changes to the stress relaxation parameter after three weeks of healing compared to uninjured controls.33 Furthermore, studies utilizing genetically modified mice showed that the absence of biglycan, a pro-teoglycan that modulates tissue stiffness and ultimate load, does not lead to changes in stress relaxation compared to a wildtype B6.34 However, while changes in the stress relaxation of our treated tendons are not being manifested at the tissue level, future work assessing cell level mechanics or changes in fiber sliding throughout the relaxation period could further elucidate the effect of treatment with our ECM-derived constructs on the tendon viscoelastic properties. Nevertheless, its ability to modulate the tendon healing response towards both improved structure and function, labeled M7 as the most effective therapeutic.
To further interrogate our hypothesis and elucidate the compositional differences that lead to improved versus ineffective healing, we performed proteomic assessment of our M7 construct and compared it to the least effective BU therapeutic. Glycoproteins such as fibronectin and fibrinogen were higher in M7 compared to BU. These proteins are an essential stimulus for cell migration, adhesion and integrin receptor interactions.35–37 Our findings showed that treatment with the glycoprotein-rich environment of M7, led to early and late decreases in collagen III content and increases in cell elongation from untreated canonically healing tendons. Thereby, early introduction of these components via M7 treatment can result in changes to the mechan-otransduction environment that could entail a possible mechanism for the observed improvements in cell morphology and matrix deposition. While our findings suggest a beneficial role for the exogenous addition of these glycoproteins during early healing, retention of a high GAG content into the remodeling phase has been identified as a marker of the scar-mediated healing response of canonically healing tendons.19,38
On the other hand, studies have shown that a hallmark of the scarless-healing MRL/MpJ model is a decrease in GAG content compared to B6 between 4- and 8-weeks post-injury.14 Similarly, we identified a decrease in GAG content by the 6-week timepoint following M7 treatment compared to injured-untreated controls, elucidating signs of beneficial matrix remodeling and further highlighting the potential of this therapeutic to encourage scarless behavior in a non-regenerative model.
Moreover, the M7 constructs contained 74 proteins that are not present in BU. Interestingly, out of these proteins, Fibrillin-1 is associated with the regulation of insulin-like growth factor (IGF) transport and uptake.39 Studies have shown that IGF signaling is necessary for proper tendon growth and maturation in adults, thereby highlighting a promising role for this glycoprotein in modulating the growth factor environment necessary for the improved healing response following M7 treatment.39,40 Additionally, heat shock protein 90, was identified in M7 but not BU. Interestingly, studies have elucidated this anti-apoptotic protein as a promising target for tendon therapeutics, as it has been implicated in a natural attempt for tendons to protect cells from increased oxidative stresses and prevent programmed cell death following injury.41 It is possible that a transient induction of these proteins through our therapeutic, beneficially modulates the apoptotic response of cells by keeping tenocytes alive to encourage tissue remodeling and effective matrix deposition at the injury site. However, due to the limited number of studies that have interrogated the role of HSP 90 in tendons, future work is necessary to assess the specific role of this protein during scarless healing of acute tendinopathies.
Lastly, while the true function of proteins found in M7 and not BU such as 40S ribosomal protein S6, pyruvate kinase PKM, guanine nucleotide-binding protein subunit β−2-like 1 and 60S ribosomal proteins remains unknown, these have been implicated in cell processes such as regulation of mRNA translation, metabolism, growth, apoptosis and proliferation.42–44 Therefore, while identification of this protein profile is a necessary first step to identify the compositional players involved in the outcomes observed in this study, future work is necessary to understand the specific roles of these proteins in tendon healing.
This study is not without limitations. Despite the improvements of M7-treated tendons compared to injured-untreated controls, this ECM-derived therapeutic did not restore treated tendon’s structure and mechanics to pre-injury levels. To overcome this limitation, undergoing studies are assessing the role of specific proteins within the MRL/MpJ ECM after 7 days post-injury (M7) at modulating B6 cell behavior towards a scarless phenotype. This new analysis, coupled with an optimized delivery system for individual proteins, could allow us to narrow down the M7 components that could be isolated and translated into more effective protein-based therapeutics that may be beneficial in future clinical applications. Moreover, while this work has identified a beneficial role for therapeutics derived from MRL/MpJ ECM after 7 days of healing, future work elucidating the therapeutic potential of MRL/MpJ ECM harnessed from later timepoints throughout the healing response could help further interrogate the specific compositional profile that leads to improved canonical tendon healing.
Surprisingly, mechanical assessment of our constructs showed that introduction of PEG-4MAL alone resulted in a decrease in stiffness of treated tendons compared to injured-untreated controls. Recently, this delivery mechanism has been effectively utilized as a mechanical pillow to ameliorate damage in a post-traumatic osteoarthritis model.21 However, our results indicate that while this hydrogel may be useful for intra-articular applications, its compatibility with the tendon healing environment may be limited. The mechanical forces experienced by the gel when placed inside the joint are largely compressive compared to the tensile nature of the tendon environment. Therefore, it is likely that the differences in the biomechanical stimulus experienced in the joint cavity compared to tendons, is responsible for the lack of effectiveness of this delivery mechanism as applied to our study. Nevertheless, despite the use of a delivery system that caused further deviations from uninjured controls, treatment with M7 resulted in improved matrix structure, composition and function compared to injured-untreated tendons. Therefore, future work identifying an improved delivery system that is more compatible with the tensile nature of the tendon healing environment, could help further improve the healing response of tendons treated with M7 closer to pre-injury levels.
Furthermore, we have previously shown that both males and females exhibit improved healing compared to B6 in regard to their mechanical outcome.45 Therefore, as a first step to interrogate the therapeutic potential of MRL/MpJ derived constructs we utilized male mice as the focus of this work. However, studies have identified differences in the quantitative trait loci that influence the MRL/MpJ healing response between male and females.46 These genetic differences could manifest in the composition of ECM deposited at the injury site between sexes. Therefore, future work is necessary to identify whether ECM therapeutics derived from female mice exhibit similar potential at improving the tendon healing response. Lastly, tendon area was not calculated in this study. This metric is usually obtained from digital images acquired during mechanical testing, however, when submerged into the water bath, loose connective tissue protruding from the tendon at the sites of dissection limited the ability of accurate cross-sectional area measurements through visual interpretation. Therefore, to preserve the accuracy of our results we did not measure material properties such as Young’s modulus and ultimate stress. Laser-based methods, ultrasound imaging or use of a profilometer could be applied in future studies to increase the accuracy these measurements.47,48
Overall, we have shown the ability of MRL/MpJ ECM-derived constructs to improve the structural, functional and cellular behavior of canonically healing tendons. Additionally, this work elucidates that incorporation of a glycoprotein rich therapeutic during early healing provides a valuable platform for tendon regeneration and highlights promising targets such as Fibrillin-1 and HSP 90 for the development of future tendon therapeutics and tissue engineering applications.
ACKNOWLEDGMENTS
The authors acknowledge Dr Ankur Singh (Cornell University) for providing PEG-4MAL gel formulations and support with gel optimization. We also thank the Proteomics Facility of Cornell University for providing the mass spectrometry data and NIH SIG grant 1S10 OD017992-01 grant support for the Orbitrap Fusion mass spectrometer. This work was supported by US National Institutes of Health (NIH) Grants NIAMS R01-AR068301 and R01-AR052743.
Footnotes
CONFLICT OF INTERESTS
The authors declare that there are no conflict of interests.
REFERENCES
- 1.Screen HRC, Berk DE, Kadler KE, et al. Tendon functional extracellular matrix. J Orthop Res. 2015;33(6):793–799. https://www.ncbi.nlm.nih.gov/pubmed/25640030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lin TW, Cardenas L, Soslowsky LJ. Biomechanics of tendon injury and repair. J Biomech. 2004;37(6):865–877. http://www.sciencedirect.com/science/article/pii/S0021929003004068 [DOI] [PubMed] [Google Scholar]
- 3.Nourissat G, Berenbaum F, Duprez D. Tendon injury: from biology to tendon repair. Nat Rev Rheumatol. 2015;11(4):223–233. 10.1038/nrrheum.2015.26 [DOI] [PubMed] [Google Scholar]
- 4.Andarawis-Puri N, Flatow EL. Promoting effective tendon healing and remodeling. J Orthop Res. 2018;36(12):3115–3124. https://www.ncbi.nlm.nih.gov/pubmed/30175859 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Zhang J, Pan T, Liu Y, Wang JH-C. Mouse treadmill running enhances tendons by expanding the pool of tendon stem cells (TSCs) and TSC-related cellular production of collagen. J Orthop Res. 2010;28(9): 1178–1183. 10.1002/jor.21123 [DOI] [PubMed] [Google Scholar]
- 6.Paredes JJ, Andarawis-Puri N. Therapeutics for tendon regeneration: a multidisciplinary review of tendon research for improved healing. Ann NY Acad Sci. 2016;1383(1):125–138. https://www.ncbi.nlm.nih.gov/pubmed/27768813 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Thomopoulos S, Kim HM, Das R, et al. The effects of exogenous basic fibroblast growth factor on intrasynovial flexor tendon healing in a canine model. J Bone Joint Surg Am. 2010;92(13):2285–2293. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Yang G, Rothrauff BB, Tuan RS. Tendon and ligament regeneration and repair: clinical relevance and developmental paradigm. Birth Defects Res C Embryo Today. 2013;99(3):203–222. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Hitchcock TF, Light TR, Bunch WH, et al. The effect of immediate constrained digital motion on the strength of flexor tendon repairs in chickens. J Hand Surg Am. 1987;12(4):590–595. 10.1016/S0363-5023(87)80213-7 [DOI] [PubMed] [Google Scholar]
- 10.Rijal G The decellularized extracellular matrix in regenerative medicine. Regen Med. 2017;12(5):475–477. 10.2217/rme-2017-0046 [DOI] [PubMed] [Google Scholar]
- 11.Urciuolo A, Urbani L, Perin S, et al. Decellularised skeletal muscles allow functional muscle regeneration by promoting host cell migration. Sci Rep. 2018;8(1):8398. 10.1038/s41598-018-26371-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Chen WCW, Wang Z, Missinato MA, et al. Decellularized zebrafish cardiac extracellular matrix induces mammalian heart regeneration. Sci Adv. 2016;2(11):e1600844. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Paredes J, Shiovitz DA, Andarawis-Puri N. Uncorrelated healing response of tendon and ear injuries in MRL highlight a role for the local tendon environment in driving scarless healing. Connect Tissue Res. 2018;59(5):472–482. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sereysky JB, Flatow EL, Andarawis-Puri N. Musculoskeletal regeneration and its implications for the treatment of tendinopathy. Int J Exp Pathol. 2013;94(4):293–303. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Lalley AL, Dyment NA, Kazemi N, et al. Improved biomechanical and biological outcomes in the MRL/MpJ murine strain following a full-length patellar tendon injury. J Orthop Res. 2015;33(11):1693–1703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Paredes J, Marvin CJ, Vaughn B, Andarawis-Puri N. Innate tissue properties drive improved tendon healing in MRL/MpJ and harnesses cues that enhance behavior of canonical healing cells. FASEB. 2020; 1–16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Molloy T, Wang Y, Murrell G. The roles of growth factors in tendon and ligament healing. Sports Med. 2003;33(5):381–394. [DOI] [PubMed] [Google Scholar]
- 18.Thomopoulos S, Parks WC, Rifkin DB, Derwin KA. Mechanisms of tendon injury and repair. J Orthop Res. 2015;33(6):832–839. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Oakes BW. Tissue healing and repair: tendons and ligaments. Rehab Sports Injuries: Scientific Basis. 2003:28–98. [Google Scholar]
- 20.Gillies AR, Smith LR, Lieber RL, Varghese S. Method for decellularizing skeletal muscle without detergents or proteolytic enzymes. Tissue Eng Part C Methods. 2011;17(4):383–389. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Holyoak DT, Wheeler TA, van der Meulen MCH, Singh A. Injectable mechanical pillows for attenuation of load-induced post-traumatic osteoarthritis. Regen Biomater. 2019;6(4):211–219. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Lee TT, Garcia JR, Paez JI, et al. Light-triggered in vivo activation of adhesive peptides regulates cell adhesion, inflammation and vascularization of biomaterials. Nat Mater. 2015;14(3):352–360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Purwada A, Shah SB, Beguelin W, et al. Ex vivo synthetic immune tissues with T cell signals for differentiating antigen-specific, high affinity germinal center B cells. Biomaterials. 2019;198:27–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Bell R, Boniello MR, Gendron NR, et al. Delayed exercise promotes remodeling in sub-rupture fatigue damaged tendons. J Orthop Res. 2015;33(6):919–925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Yang Y, Thannhauser TW, Li L, Zhang S. Development of an integrated approach for evaluation of 2-D gel image analysis: impact of multiple proteins in single spots on comparative proteomics in conventional 2-D gel/MALDI workflow. Electrophoresis. 2007;28(12):2080–2094. [DOI] [PubMed] [Google Scholar]
- 26.Thomas CJ, Cleland TP, Zhang S, et al. Identification and characterization of glycation adducts on osteocalcin. Anal Biochem. 2017;525: 46–53. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Yang Y, Anderson E, Zhang S. Evaluation of six sample preparation procedures for qualitative and quantitative proteomics analysis of milk fat globule membrane. Electrophoresis. 2018;39(18):2332–2339. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Figliuzzi M, Bonandrini B, Remuzzi A. Decellularized kidney matrix as functional material for whole organ tissue engineering. J Appl Biomater Funct Mater. 2017;15(4):e326–e333. [DOI] [PubMed] [Google Scholar]
- 29.Kc P, Hong Y, Zhang G. Cardiac tissue-derived extracellular matrix scaffolds for myocardial repair: advantages and challenges. Regen Biomater. 2019;6(4):185–199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Riganti JM, Ciotola F, Amenabar A, et al. Urinary bladder matrix scaffolds strengthen esophageal hiatus repair. J Surg Res. 2016;204(2):344–350. [DOI] [PubMed] [Google Scholar]
- 31.Stauber T, Blache U, Snedeker JG. Tendon tissue microdamage and the limits of intrinsic repair. Matrix Biol. 2019;85–86:68–79. [DOI] [PubMed] [Google Scholar]
- 32.Subramanian A, Kanzaki LF, Galloway JL, Schilling TF. Mechanical force regulates tendon extracellular matrix organization and tenocyte morphogenesis through TGFbeta signaling. eLife. 2018:7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Beason DP, Kuntz AF, Hsu JE, et al. Development and evaluation of multiple tendon injury models in the mouse. J Biomech. 2012;45(8): 1550–1553. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Dourte LM, Pathmanathan L, Mienaltowski MJ, et al. Mechanical, compositional, and structural properties of the mouse patellar tendon with changes in biglycan gene expression. J Orthop Res. 2013;31(9): 1430–1437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Dejana E, Lampugnani MG, Giorgi M, et al. Fibrinogen induces endothelial cell adhesion and spreading via the release of endogenous matrix proteins and the recruitment of more than one integrin receptor. Blood. 1990;75(7):1509–1517. [PubMed] [Google Scholar]
- 36.Johansson S, Svineng G, Wennerberg K, et al. Fibronectin-integrin interactions. Front Biosci. 1997;2:d126–d146. [DOI] [PubMed] [Google Scholar]
- 37.Hsiao C-T, Cheng H-W, Huang C-M, et al. Fibronectin in cell adhesion and migration via N-glycosylation. Oncotarget. 2017;8(41): 70653–70668. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Attia M, Scott A, Duchesnay A, et al. Alterations of overused supraspinatus tendon: a possible role of glycosaminoglycans and HARP/pleiotrophin in early tendon pathology. J Orthop Res. 2012; 30(1):61–71. [DOI] [PubMed] [Google Scholar]
- 39.Disser NP, Sugg KB, Talarek JR, et al. Insulin-like growth factor 1 signaling in tenocytes is required for adult tendon growth. FASEB J. 2019;33(11):12680–12695. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40.Provenzano PP, Alejandro-Osorio AL, Grorud KW, et al. Systemic administration of IGF-I enhances healing in collagenous extracellular matrices: evaluation of loaded and unloaded ligaments. BMC Physiol. 2007;7:2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Millar NL, Wei AQ, Molloy TJ, et al. Heat shock protein and apoptosis in supraspinatus tendinopathy. Clin Orthop Relat Res. 2008;466(7): 1569–1576. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 42.Robledo S, Idol RA, Crimmins DL, et al. The role of human ribosomal proteins in the maturation of rRNA and ribosome production. RNA. 2008;14(9):1918–1929. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 43.Du Y, Meng J, Huang Y, et al. Guanine nucleotide-binding protein subunit beta-2-like 1, a new Annexin A7 interacting protein. Biochem Biophys Res Commun. 2014;445(1):58–63. [DOI] [PubMed] [Google Scholar]
- 44.Desai S, Ding M, Wang B, et al. Tissue-specific isoform switch and DNA hypomethylation of the pyruvate kinase PKM gene in human cancers. Oncotarget. 2014;5(18):8202–8210. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45.George N, Bell R, Paredes J, et al. Superior mechanical recovery in male and female MRL/MpJ tendons is associated with a unique genetic profile. J Orthop Res. 2020. 10.1002/jor.24705 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 46.Heber-Katz E, Chen P, Dvm LC, et al. Regeneration in MRL mice: further genetic loci controlling the ear hole closure trait using MRL and M.m. Castaneus mice. Wound Repair Regen. 2004;12(3):384–392. [DOI] [PubMed] [Google Scholar]
- 47.Pokhai GG, Oliver ML, Gordon KD. A new laser reflectance system capable of measuring changing cross-sectional area of soft tissues during tensile testing. J Biomech Eng. 2009;131(9):94504. [DOI] [PubMed] [Google Scholar]
- 48.Moon DK, Abramowitch SD, Woo SL-Y. The development and validation of a charge-coupled device laser reflectance system to measure the complex cross-sectional shape and area of soft tissues. J Biomech. 2006;39(16):3071–3075. [DOI] [PubMed] [Google Scholar]