Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Aug 1.
Published in final edited form as: Free Radic Biol Med. 2022 Jun 17;188:175–184. doi: 10.1016/j.freeradbiomed.2022.06.223

Extracellular Biomolecular Free Radical Formation During Injury

Madeline R Hines 1, Jessica E Goetz 1, Piedad C Gomez-Contreras 1, Samuel N Rodman III 1, Suryamin Liman 1, Elise L Femino 1, Paige N Kluz 2, Brett A Wagner 1, Garry R Buettner 1, Eric E Kelley 3, Mitchell C Coleman 1
PMCID: PMC9725094  NIHMSID: NIHMS1853609  PMID: 35724853

Abstract

Objective: Determine if oxidative damage increases in articular cartilage as a result of injury and matrix failure and whether modulation of the local redox environment influences this damage.

Osteoarthritis is an age associated disease with no current disease modifying approaches available. Mechanisms of cartilage damage in vitro suggest tissue free radical production could be critical to early degeneration, but these mechanisms have not been described in intact tissue. To assess free radical production as a result of traumatic injury, we measured biomolecular free radical generation via immuno-spin trapping (IST) of protein/proteoglycan/lipid free radicals after a 2 J/cm2 impact to swine articular cartilage explants. This technique allows visualization of free radical formation upon a wide variety of molecules using formalin-fixed, paraffin-embedded approaches. Scoring of extracellular staining by trained, blinded scorers demonstrated significant increases with impact injury, particularly at sites of cartilage cracking. Increases remain in the absence of live chondrocytes but are diminished, thus appear to be a cell-dependent and -independent feature of injury. We then modulated the extracellular environment with a pulse of heparin to demonstrate the responsiveness of the IST signal to changes in cartilage biology. Addition of heparin caused a distinct change in the distribution of protein/lipid free radicals at sites of failure alongside a variety of pertinent redox changes related to osteoarthritis. This study directly confirms the production of biomolecular free radicals from articular trauma, providing a rigorous characterization of their formation by injury.

Keywords: DMPO, Osteoarthritis, Posttraumatic, Trauma, Redox, SOD3, Cartilage, Free Radical, Radical, Oxidative Stress, Oxidation, Injury

Graphical Abstract

graphic file with name nihms-1853609-f0001.jpg

INTRODUCTION:

Post traumatic osteoarthritis (PTOA) involves progressive erosion and destruction of articular cartilage extracellular matrix (ECM) following a traumatic joint injury. Our group has shown that intracellular reactive oxygen species (ROS) and mitochondrial dysfunction increase in chondrocytes immediately after cartilage impact injury and initiate degenerative joint disease in a swine model of intra-articular fracture[1–4]. A wide variety of studies have demonstrated oxidation acutely in cartilage; however, most of these studies were directly focused upon intracellular biochemistry. Thorough description of the sources and targets of oxidation in the extracellular space of articular cartilage after injury is challenging given the tissue’s strength and relative opacity. To explore the formation of extracellular free radicals after injury, we investigated the potential of a novel technique termed immuno-spin trapping (IST), which permits visualization of free radicals in formalin-fixed, paraffin-embedded tissue[5, 6].

Though the existing literature on the topic is limited, thorough benchtop studies of extracellular free radical formation have focused on individual, specific chemical mechanisms of ROS-mediated cartilage damage like superoxide (O2•-)[7, 8]. These in vitro studies demonstrated collagen fiber cleavage at proline residues by O2•- levels as low as 1 nM via xanthine/xanthine oxidase (a well characterized source of O2•-) and indirectly by hydroxyl radical (HO•) through transition metal reactions[7, 9]. HO• is highly reactive and known to cleave collagen fibers[10] in proximity of generation[11]. Extracellular superoxide dismutase (SOD3), a key antioxidant enzyme present in the extracellular compartment of articular cartilage, inhibits this in vitro collagen cleavage[7]. These studies have established that extracellular free radical chemistry can influence collagen fiber damage but have not defined key loci for these reactions.

Detailed mechanistic descriptions of extracellular redox chemistry in vivo after trauma have not been made, though nitric oxide has otherwise been an extensive focus of the field and some focus on SOD3 is present within the literature. It has been shown that loss of the SOD3 heparin binding domain, causing widespread loss of cartilage SOD3 content, led to decreased proteoglycan as well as increased cartilage erosion at the articular surface after prolonged exercise in a murine model[12]. In larger tissue, with more expansive ECM, human samples obtained from total knee arthroplasties show decreased SOD3 and increased 3-nitrotyrosine-modified protein (3-NT) in the extracellular space relative to healthy controls, suggesting the possibility of increases in O2•- concentration and nitric oxide in the ECM of diseased cartilage[13]. Thus, SOD3 and free radical chemistry are crucial to extracellular cartilage function, but the sources and locations of extracellular free radicals in cartilage remain unclear.

Rather than concentrate on the generation of specific radical species, we chose to apply the IST method targeting any radical formed directly on proteins/proteoglycans/lipids of the tissue. Prior studies of non-cartilage tissues have described how biomolecular free radical formation provides a valuable indication of oxidative distress, since these larger biomolecules are often protected by the intracellular antioxidants[5]. In most other tissues, the presence of these free radicals could be rigorously interrogated using electron paramagnetic resonance (EPR); however, the stiffness and high water content of cartilage preclude applying this technique. In lieu of this, we decided to investigate biomolecular free radical formation in the ECM of injured cartilage using the IST technique that detects labeled biomolecular free radicals in histological sections of paraffin embedded tissue[5, 6, 14–17].

IST uses 5, 5-dimethyl-1-pyrroline N-oxide (DMPO), a common EPR reagent used to prolong free radical signals through stable, covalent bonds with protein/lipid radicals in vivo or in vitro[5, 6, 15]. DMPO adducts on protein/lipid radicals form stable nitrones which are retained though formalin fixation. These nitrones can then be identified using an antibody raised against DMPO adducts that binds to the nitrone structure, normally a structure not found in vivo [16]. Thus, IST is an ideal means of exploring specific loci of extracellular free radical production and oxidative injury in articular cartilage after mechanical trauma. We hypothesized that a controlled and well-characterized impact injury to articular cartilage would increase formation of biomolecular free radicals at sites of ECM failure as indicated by increased DMPO staining. We then tested a prototypic modulation of biomolecular free radical formation from injury by adding heparin to the articular cartilage of skeletally mature Yucatan minipigs. This induced a variety of biological changes reminiscent of arthritic tissue and demonstrates the responsiveness of this signal to changes in cartilage biochemistry[12]. Our results demonstrate that free radical formation is associated with ECM failure and that this signal arises from both cell-dependent and -independent mechanisms. Finally, changes in free radical formation at sites of failure after heparin treatment suggest there may be distinct responses to injury in normal and unhealthy tissue. This novel application of a variety of custom-developed and interdisciplinary techniques demonstrates the value of a broader view of free radical biochemistry for exploring redox biology within challenging physiological niches.

METHODS:

Tissue Culture

This study used two sources of swine stifle (knee) joints. One source was from agricultural animals obtained fresh from a local abattoir (Bud’s Custom Meats, Riverside, IA). The agricultural animals’ sexes were unknown and assumed random. The other source was Yucatan minipig stifles obtained postmortem from an ongoing collaboration in the Department of Orthopedics and Rehabilitation at the University of Iowa. The Yucatan minipigs were an even mix of males and females greater than 2 years of age and thus skeletally mature. From both groups, cylindrical osteochondral plugs, 8 mm diameter, were taken from the load bearing portion of the medial and lateral femoral condyle. Medial tibias of agricultural swine were cut into two separate 1 cm3 explants. After harvest, the explants and plugs were placed in normal growth medium (45% Dulbecco’s modified Eagle’s medium, 45% F-12, 10% Fetal Bovine Serum (FBS) (Gibco), with penicillin-streptomycin and amphotericin B). Media was stored in a humidified chamber at 5% oxygen (O2) and 5% carbon dioxide at 37°C to maintain a low O2 concentration. Media and all samples were stored in the humidified chamber during the duration of the experiment.

Impact

Samples were leveled and potted using polycaprolactone (Sigma Aldrich, St. Louis) then impacted using a stainless steel flat impermeable platen. A previously characterized custom drop tower was used to deliver an impact of 2 J/cm2, and explants were returned to the incubator for 24 h[18–21]. Impact experiments were performed in dim lighting to limit free radical production by photochemistry[22].

DMPO

DMPO stock solution was prepared by diluting DMPO (Dojindo, D048, Kumamoto) to 1 M using distilled water bubbled with argon, then aliquoted and stored at −20°C. For use, the 1 M aliquoted solution was diluted in low O2 medium to the desired concentration. To determine the appropriate concentration of DMPO in articular cartilage, we evaluated the DMPO staining of porcine plugs pre-exposure with a range of DMPO concentrations (5 mM −100 mM). Three serial sections from each sample were assessed for reproducible DMPO staining and each group was compared to determine if the change in DMPO concentration caused a change in staining. DMPO concentrations of 50 mM and below had little to no DMPO staining. DMPO concentrations above 50 mM DMPO staining was present, but the increased concentration of DMPO did not reproduce in appearance until 70 mM and above, Supplemental figure 1.

The reason for using such an elevated concentration of DMPO is due to a direct competition between the spin trap, DMPO, and molecular O2 for free radicals formed on lipids and proteins/proteoglycan subsequent to reaction with reactive species. O2 will react with these protein and lipid radicals at very fast rates (~108 mol/s) whereas DMPO’s affinity for carbon, sulfur and nitrogen-centered biomolecular free radicals are significantly less (~105-107 mol/s)[11, 17]. As such, considerably greater concentrations of DMPO compared to O2 enables DMPO to effectively compete with O2 for these biomolecular free radicals before they form hydroperoxides (ROOH) and are no longer reactive with the spin trap. After this determination of the optimum DMPO concentration (70 mM), we treated the remaining porcine osteochondral samples for 1 h prior to impact with medium containing 70 mM or 0 mM DMPO and cultured in DMPO medium for 24 h after impact, unless otherwise specified.

To generate a positive control, samples were incubated in low O2 media with or without 100 μM each of myoglobin, iron sulfate, and hydrogen peroxide (H2O2). Myoglobin generates a flux of hydrogen peroxyl radical and excess iron facilitates Fenton chemistry and the generation of the HO•. This creates a strong enough flux of free radicals to lead to macromolecule free radical formation. DMPO was added to this solution and samples were co-incubated in that solution for 24 h, prior to fixation and histological processing.

Western Blot

Samples from the femur of agricultural swine were cultured in low O2 media supplemented with FastGro synthetic, animal free chemically defined FBS (MPbio, 2640049), instead of FBS. This was to prevent FBS for obscuring proteins from the media and dominating the signal observed. The samples were incubated with DMPO and impacted as described above. The samples were cultured for 24 h after impact in groups of six samples per experimental group. We included negative control (impact, no DMPO), positive control (100 μM myoglobin/H2O2/iron sulfate), impact DMPO, no impact DMPO. After incubation, supernatants from impacted plugs were collected, total proteins were quantified by BCA method (Thermo, 23221). The protein samples were denatured and reduced by addition of LDS sample buffer (Invitrogen NP0007) and DTT reducing agent (Invitrogen, NP0009) and heated at 70°C by 10 min. The amount of 50 μg of protein was loaded per well and electrophoresed through a 10% Bis-Tris NuPage acrylamide gel (Invitrogen, NP0315), at 120 V constant by 2 h with MES-SDS running buffer (Invitrogen, NP0002). After the electrophoresis, the gel proteins were transferred to a 0.45 μM nitrocellulose membrane (BioRad, 1620145), at 300 mAmp constant by 2 h with 20% methanol 0.1% SDS transfer buffer (Invitrogen, NP0006–1). The blotted membrane was stained using Ponceau S, as a loading and protein quality control. The membrane was blocked by 30 min with a goat serum-gelatin solution 5%. Poly clonal chicken anti-DMPO primary antibody (1:1000), a generous gift from Dr. Ronald Mason (NIEHS), diluted in goat serum-gelatin 2.5%, was incubated by 1 h RT, secondary ab was an HRP Goat anti Chicken (1:6000) diluted in goat serum-gelatin 2.5% incubated by 1h RT (Abcam). After rinsing with TBST, the protein signal was developed and visualized using Super Signal West Femto (Thermo, 34095) and Amercham Hyperfilm ECL system (GE Healthcare, 28906839) film.

Non-articular Collagen Failure

Rat tails were used as a source of type I collagen to provide a comparison with distinct anatomy from articular cartilage. The tails were degloved and incubated for 1 h in media with either 0 mM DMPO or 70 mM DMPO. Following incubation, rat tails were either cut cleanly via scalpel or torn, manually, then returned to media for 24 h. After 24 h incubation in DMPO media samples were fixed and processed as described below.

Chondrocyte Viability

To investigate the staining contributions of live cells, agricultural swine samples were placed in 70% ethanol for 2 h. After the 2 h treatment, the lethality of ethanol was confirmed using parallel samples stained using calcein AM (Thermofisher, 65-0853-39) for live cells and ethidium homodimer (Thermofisher, E1169) for dead cells. The stained cartilage was imaged using the Olympus FV1000 confocal laser scanning microscope (Olympus America).

Heparin Treatment

Articular cartilage is an avascular tissue which is comprised mostly of ECM components such as collagen and proteoglycans, leading to a dense and negatively charged tissue with limited diffusion by size and charge[23, 24]. Evidence supports proteins greater than 70 KDa have limited diffusion in the ECM, thus preventing direct supplementation with enzymes that might modulate the IST for this study[25, 26]. To modulate the redox environment surrounding the injuries under study, we applied low molecular weight heparin, which liberates proteins like SOD3 and others. We confirmed these changes via immunohistochemistry (IHC). Heparin was prepared by dissolving the desired concentration in low O2 (5% O2) media. Samples were treated with either a high concentration (2 mg/mL) of heparin (Sigma Aldrich #H3149–10KU), a low concentration (1 mg/mL) of heparin, or in media control. Samples were incubated overnight then co-incubated for 1 h with DMPO prior to impact. After impact injury, the samples were returned to their respective treatment media for 24 h incubation.

Fixation, Processing, and Embedding

Samples were fixed under vacuum for 4 h at 56 kPa in 0.3% Safranin-O (saf-O) in 10% neutral buffered formalin (NBF) to maintain chondrocyte and tissue morphology[27]. The saf-O NBF solution was replaced with regular NBF after 4 h, and the samples were fixed up to one week total. Samples were decalcified using daily changes of 5% formic acid until decalcification was complete. Samples were bisected though the center of impact site to reveal the central cross-sectional plane of the specimen, and then processed and embedded in paraffin. Samples shown were sectioned at 5 μm for IHC.

Immunohistochemistry

Slides were deparaffinized at 55°C to remove moisture trapped in the sample. The slides were then rehydrated via a series of washes: 100% xylene, 100% ethanol, 95% ethanol, 80% ethanol, finally distilled water. Slides were placed in citric acid antigen retrieval (0.1 M citric acid monohydrate (18 mL), 0.1 M trisodium citrate dehydrate (82 mL), deionized water (900 mL)) pH 6, overnight, up to 16 h, at 55°C. The samples were cooled to room temperature, rinsed in distilled water, and endogenous peroxidases were quenched using a 3% hydrogen peroxide solution. Samples were then blocked using normal goat serum blocking solution without bovine serum albumin. The primary anti-DMPO antibody was incubated overnight (1:150) at 4°C. The secondary antibody (1:200) was incubated 30–40 min at room temperature. While the secondary antibody was incubating, the ABC reagent Elite Standard (Vector laboratories, Cat. # PK-6100) was mixed and left to incubate at room temperature for 30 min. Samples were then incubated in ABC reagent for 30 min. The samples were developed using DAB reagent (Vector laboratories, Cat. # SK-4100) which was incubated for 5 min. No counterstain was used.

Assessment of biomolecular free radicals used polyclonal chicken anti-DMPO (1:150) primary antibody as stated above[16]. To evaluate off target staining polyclonal chicken IgY (R&D systems, AB-101-C) 1:150 isotype control was used and a no primary control. The secondary antibody used was a goat anti-chicken HRP (Abcam, ab97135) (1:200).

The SOD3 content was evaluated using a mouse anti-SOD3 antibody (1:50) (Santa Cruz, sc-101338) and goat anti-mouse secondary (1:200) (Thermofisher, 31430). Non-specific binding was assessed using an isotype control (Abcam, ab37355) and a no primary control. Immunofluorescence of 3-NT was done using a rabbit anti-nitrotyrosine antibody (Millipore, 06–284, 1:150) and goat anti-rabbit cy5 fluorescent secondary (Roche, 760–238) on a Discovery Ultra (Roche, Switzerland).

Image Analysis and Statistics

Images from the IHC slides were scored in a blinded manner for the intensity of the extracellular space on a scale from 0 (no staining) to 3 (high intensity staining) with graders provided specific instructions to ignore intracellular staining. Using Graph Pad Prism 8, the scores were compared, and a non-parametric one-way ANOVA (Kolmogorov-Smirnov test) was used for statistical analysis with significance at p < 0.05 and post hoc analysis using Mann-Whitney. Outliers were determined using ROUT (Q=1%)[28]. Finally, the interrater reliability was evaluated by calculating the Fleiss’ kappa (κ)[29]. Impacted samples without ECM failure were excluded from the study.

To determine the staining intensity along sites of ECM failure in the heparin treated groups in greater detail and with more precise quantitation, we applied a custom designed algorithm to IST images. Three serial sections from each sample were stained. Images were taken from the same site of ECM failure and each crack was evaluated with replicate measures if appearing in multiple sections. Cracks only appearing in one of the sections were analyzed as a single data point. This resulted in algorithmic analysis of 17 cracks from 4 plugs in the DMPO group and 27 cracks from 5 plugs in the heparin DMPO group. The cracks were traced manually in MATLAB to generate masks. From these masks the intensity was assessed in 5 μm intervals along a perpendicular ray from the crack, yielding an average intensity at each distance grouping of 0–5, 5–10, 10–15, and 15–20 μm outward from a given crack. Maximum intensity in these distance groups was also determined. Analysis was restricted to the tissue, and slide background regions such as above the articular surface and wide fissures were not included. From these values an ANOVA was run to determine statistical significance between the groups, p < 0.05, outliers were removed using ROUT (Q=1%) with post-hoc t-test.

RESULTS:

Injury Increases Extracellular Biomolecular Free Radical Formation

To first demonstrate the anti-DMPO antibody specificity, the DMPO and antibody controls were evaluated. To control for DMPO in the sample, an impacted sample receiving no DMPO was stained with an anti-DMPO antibody. To assess non-specific staining, sections were taken from a DMPO exposed sample with impact injury. One slide was used as a no primary control and the other was stained with anti-DMPO. The no DMPO and the no primary had little to no staining compared to the anti-DMPO sample, Figure 1a. As a positive control for free radical damage to the ECM, we used media containing DMPO with either 100 μM myoglobin and H2O2 or 100 μM each of myoglobin, H2O2, and iron sulfate. Iron sulfate was added to enable a greater flux through the Fenton reaction, which generates HO•. By increasing the flux of possible free radicals generated with this highly reactive primary radical, we were able to cause biomolecular free radical formation on the ECM without impact, Figure 1b.

Figure 1: Increased DMPO staining in ECM after injury suggests increased extracellular biomolecular free radical formation after injury.

Figure 1:

Open in a new tab

a) Representative sections from no DMPO sample stained with anti-DMPO antibody, and sections from a DMPO sample with no primary antibody and with anti-DMPO antibody. The samples without DMPO or without primary antibody had minimal staining compared to the anti-DMPO stained sample. Scale 200 μm. b) Samples were co-incubated with DMPO, H2O2, and myoglobin alone or myoglobin and iron acetate. The open arrows denote chondrocytes. The pink chondrocytes are negative for DMPO in the no primary sample whereas the brown chondrocytes seen in the treated samples are positive for DMPO (arrows). The myoglobin and iron treated samples generated ECM free radicals without impact, a positive control of free radical damage to the ECM. Scale 50 μm. c) Assessment of stain reproducibility by using serial sections of a 2 J/cm2 impacted femoral explant with 70 mM DMPO. Scale 100 μm. d) Representative images of extracellular DMPO staining demonstrate increased DMPO staining after the 2 J/cm2 impact compared to non-impacted tissue. Scale 100 μm. e) Blinded scoring of IST images demonstrated increased extracellular DMPO staining between the impacted femur and tibia. IST stain scoring was higher in the femur compared to the tibia. *p < 0.05 Kolmogorov-Smirnov test, κ = 0.59, n = 4. f) DMPO staining is observed in ECM of collagen fibers from rat tails after injury. Collagen fibers which were cut with a scalpel had increased staining compared to the uninjured control. The samples which were manually torn had the most intense staining in all the groups. These results support biomolecular free radical formation as a result of ECM failure is not restricted to articular cartilage injury. Scale 200 μm.

To validate the reproducibility of IST resulting from impact injury, serial sections from the same injured specimen were stained. Following a 70 mM DMPO pre-exposure, 2 J/cm2 impact injury, and return to DMPO media for 24 h incubation, the DMPO staining pattern along sites of ECM damage was remarkably consistent between serial sections, Figure 1c. This supports that positive staining is not a product of artifactual stain retention at damaged sites. Throughout the impacted specimens, we observed increased DMPO staining indicative of increased biomolecular free radical formation at sites of visible cracks or tears in the ECM, Figure 1d. While cartilage from both the femur and the tibia showed increased DMPO staining after 2 J/cm2 impact, representative images and scoring support more intense and extensive staining in the ECM failure in the impacted femur than the impacted tibia, Figure 1e. DMPO staining was reproduced after injury in rat tails, abundant in type I collagen, Figure 1f. These increases in DMPO staining after injury support the hypothesis that ECM damage increases biomolecular free radical production in the extracellular space in tissue.

To assess how widespread free radical formation might be among different extracellular proteins after injury, culture media was collected 24 h after impact and analyzed via Western blotting for anti-DMPO reactive, IST products. Media from six osteochondral plugs per group were pooled in the following groups: impacted no DMPO, impact DMPO myoglobin (myo)/H2O2/iron sulfate, impact DMPO, and no impact DMPO. All groups had similar amounts of protein visualized by Ponceau S staining, Figure 2a. The anti-DMPO western blot showed little-to-no staining in the impact without DMPO lane supporting the selectivity of the antibody for DMPO. The impact DMPO myoglobin (myo)/H2O2/iron sulfate lane had strong staining as expected. The impact DMPO also demonstrated much darker staining than the no impact DMPO, Figure 2a. This signal from a variety of proteins in the culture media 24 h after injury supports the hypothesis that protein radicals are formed by cartilage injury.

Figure 2: Extracellular biomolecular free radical production at the time of impact compared to the subsequent 24 hours.

Figure 2:

Open in a new tab

a) Anti-DMPO western blot of culture media 24 h after impact injuries. The impacted DMPO treated samples had more staining than the unimpacted samples, supporting ECM radical formation increases with impact injury. b) Schematic of how exposures were arranged. By staggering the timing of DMPO incubation prior or post injury the timing for biomolecular free radical formation can be determined. c) DMPO incubation prior to injury has increased DMPO staining at sites of ECM failure (black arrows) compared to DMPO incubation after injury. Scale 200 μm. d) Semi-quantitative scoring demonstrated the pre-exposed samples did have more DMPO staining than the control and post exposure. Due to the low n value significance was not reached, p = 0.057. *p < 0.05 Kolmogorov-Smirnov test, κ = 0.15 n = 4.

To determine if biomolecular free radicals are produced at the time of impact injury or in the subsequent 24 h, we treated samples with DMPO prior to impact or immediately following impact, Figure 2b. By pre-exposing samples, the DMPO is available to react with free radicals generated at the time of impact, whereas exposure after impact captures free radicals formed in the subsequent 24 h. Increased DMPO staining was observed in the pre-exposed groups, 1 h and 24 h pre-exposure, Figure 2c and d, while post-exposure produced staining similar to controls receiving no DMPO. This suggests that the impact injury itself contributes the greatest proportion of the biomolecular free radicals DMPO staining observed compared to contributions taking place after impact.

Injury Causes Extracellular Free Radical Formation by Cell-Dependent and -Independent Mechanisms

To determine whether chondrocytes were contributing directly to the DMPO staining, some specimens were incubated in 70% ethanol for 2 h to induce chondrocyte death. The ethanol media was removed, and all the samples received fresh media and were incubated overnight prior to injury to ensure no residual ethanol remained prior to impact. Chondrocyte death was confirmed using confocal microscopy, Supplemental Figure 2. Specimens were pre-exposed with DMPO for 1 h prior to impact. As anticipated, viable chondrocytes (open arrow Figure 3a) had DMPO staining whereas the ethanol treated lacked cellular staining. The extracellular DMPO staining was evaluated in the control impact group compared to the ethanol treated impact group, Figure 3a. Ethanol treatment decreased the DMPO staining by 50%, Figure 3b. These findings suggest that both cell-dependent and -independent sources of biomolecular free radicals are present after injury to articular cartilage.

Figure 3: Extracellular DMPO staining is associated with viable chondrocytes.

Figure 3:

Open in a new tab

a) Representative images illustrate increased extracellular staining (black arrows) in the viable chondrocyte sample compared to the ethanol-treated samples. This suggests that viable chondrocytes (open arrows) contribute to DMPO staining at sites of ECM failure. Scale 200 μm. b) Histological scoring of the images indicates viable chondrocytes contribute to increased DMPO staining at sites of failure. The scores also suggest that since removing the cells does not return the staining to baseline, the impact itself is also a contributor to the DMPO staining intensity. *p < 0.05 Kolmogorov-Smirnov test, κ = 0.33, n = 4.

Biomolecular Free Radicals at Sites of ECM Failure are Responsive to Manipulations of the Local Redox Environment

Because aging, obesity, and a large number of other factors are associated with modulation of the response to injury and arthritis, we wanted to explore how a controlled modulation of the local redox environment in a manner consistent with previously published observations might alter free radical formation after injury[12, 13]. We treated the culture media with heparin, hypothesizing that a bolus addition of heparin would cause a variety of disruptions in cartilage biology reminiscent of arthritis including: disruption of proteoglycan; liberation of local growth factors and proteinases from the tissue; and disruption of the heparin binding domain of SOD3. These changes are not meant to model the sum total of arthritis in cartilage so much as demonstrate that modulating tissue redox environs will significantly impact the free radical production observed.

To support that heparin was inducing relevant changes in redox biochemistry, IHC analysis of SOD3 after heparin treatment revealed significantly decreased SOD3 compared to controls, Figure 4a. To compare our manipulation with prior results[12, 13], we assessed 3-NT-modified protein staining similarly to the cited studies. Heparin treatment resulted in 3-NT staining of a distinct, localized, punctate pattern in the ECM of unimpacted specimens, Figure 4b compared to impacted tissue in Supplemental Figure 3. This pattern was very similar to published results of SOD3 deficient OA patient samples[13]. After impact, the heparin increased IST at sites of minor ECM damage and at the edges of the cracks themselves, but the tissue further away from the cracks appeared less intensely stained compared to controls, Figure 4c.

Figure 4: Heparin treatment delocalizes SOD3, produces extracellular 3-NT, and increases DMPO staining at site of failure.

Figure 4:

Open in a new tab

a) Representative images demonstrate that SOD3 is present in the articular cartilage of our swine samples (back arrows) and that heparin treatment decreased SOD3 localization compared to untreated in the absence of injury. Scale 100 μm. b) Confirmation of extracellular 3-NT punctate staining (yellow arrows) observed in the ECM of heparin treated samples without injury. Red indicates safranin O counterstain and pink indicates anti-3NT staining. Scale 50 μm. c) The staining pattern surrounding ECM failure was different between the DMPO and heparin DMPO treated samples. The staining appeared sharper around the failure in the heparin DMPO group compared to the DMPO group. Scale 50 μm.

To provide more direct quantitation of the specific extracellular changes observed at sites of cartilage failure, we applied custom-designed computer algorithms that rely upon crack traces, Figure 5a, to determine the intensity profile along the length of each individual crack identified in our experiment. IST images were inverted, Figure 5b, and the cells near the cracks were masked to avoid cells contributing to the signal, arrows. From each traced crack, perpendicular vectors were generated along the length to collect intensity values within distance ranges of 0–5, 5–10, 10–15, and 15–20 μm, Figure 5c. Under normal conditions, IST appears uniform across the different distances after injury but disruption of SOD3 localization concentrates biomolecular free radical formation immediately adjacent to cracks in the tissue, Figure 5d. While heparin is not a specific or direct manipulation of specific free radical species, we note that this supports the hypotheses of prior authors that SOD3 may have a role in preventing damage to cartilage, in this case by preventing accumulation of free radical damage at sites of matrix failure.

Figure 5: Delocalization of SOD3 increases biomolecular free radical formation at the site of failure but decreases the biomolecular free radical formation radiating from the failure.

Figure 5:

Open in a new tab

a) Using a custom MATLAB program, the cracks were traced. b) The image was inverted. The cells near the traced lines were masked (arrows) to avoid cellular contribution to calculations of staining intensity. c) Vectors perpendicular to the traced crack were generated along the length of the crack. Each vector was divided into distance ranges 0–5 (white), 5–10 (blue), 10–15 (red), 15–20 (pink) μm. Each vector had an average and maximum intensity reported. d) The difference in average intensity between the DMPO (blue) and Heparin DMPO (red) treated groups were significantly different at distances greater 5 μm from the crack. This suggests that the delocalization of SOD3 prevents biomolecular free radicals’ formation farther into the tissue. This is supported by the maximum intensity being highest within the first 5 μm and the maximum intensities decreasing farther from the crack. p < 0.05 scale 50 μm. n = 17 DMPO, n = 27 Heparin DMPO.

DISCUSSION:

Though previous research has focused on specific ROS reactions within cartilage, we have chosen the broader indicator of oxidative damage, biomolecular free radical formation evaluated by IST. Here, that lack of specificity to any one free radical is a strength insofar as the large variety of biomolecular free radicals that might be generated by ECM failure is too great to be captured by measuring an individual species. Further, each specific free radical’s individual concentration may be too low to measure accurately. Visualized in the western blot, impact increased a variety of biomolecular free radicals in the media supporting the idea of broad free radical damage to macromolecules occurring from impact. Thus, assessing the formation of all biomolecular free radicals via IST enables a wide view of oxidative injury spatially in tissue after trauma which other means of detection lack. This study revealed that the mechanical trauma itself may be a direct initiator of oxidative damage to tissue. Our results support the hypothesis that areas of increased ECM damage have increased biomolecular free radical formation.

There are a variety of strengths and specific implications from this study demonstrating free radical formation after trauma in cartilage. Serial sectioning revealed a consistent DMPO staining pattern between adjacent sections, supporting the robustness and usefulness of the technique. This demonstrates that the staining does not result from an artifact of stain trapping, which is stochastic and does not reproduce between sections. Reproducible staining increases with ECM failure, anatomic site, and in the presence or absence of treatments suggest an important extracellular biochemistry may be at work to preserve articular cartilage components after mechanical damage.

Although DMPO adducts are increased after impact, the western blot demonstrates that a variety of proteins are adducted to DMPO supporting the importance of not looking for a single free radical target. ECM damage increased DMPO staining after impact, whereas impact without failure of the tissue did not seem to increase extracellular DMPO staining. Though we have not measured or addressed differences in energies of impact, the presence or absence of ECM damage is likely to dominate extracellular IST signal and higher energy impacts are much more likely to cause ECM failure in this model system. Among the variety of tissues impacted in our model, we noted intensity differences between the femur and tibia that could be driven by the differences in material properties of either site or other physical properties of the tissue like water or proteoglycan content.

Ethanol treated articular cartilage showed decreased DMPO staining at sites of ECM failure compared to the live tissue. The ethanol treated specimens still produced visible DMPO staining, suggesting the impact itself can induce biomolecular free radical formation independent of the presence of live cells. However, some noteworthy diminishment of the signal by ethanol suggests a distinct cellular contribution to the extracellular free radical biochemistry. We hypothesize that the physical breakage of collagen fibers themselves could be the source for the initiation of events leading to cell-independent free radical formation of ECM. This appears to then blend with intracellular redox biochemistry from cell-dependent events to yield the high amounts of radical formation observed.

Though we did not manipulate SOD3 directly, our exploration of disrupting normal redox function during ECM failure is indirectly supportive of previous findings which demonstrated that SOD3 plays a role in cartilage homeostasis. Large animal tissue is required for these extracellular injuries and explorations, limiting our ability to transgenically manipulate SOD3. Further, because of its density and charge, cartilage greatly limits diffusion of large protein complexes like SOD enzymes. Available SOD mimetics participate in other redox reactions and we are also unsure of how the high concentration of DMPO might impact these reagents and determinations. Thus, we chose to use heparin as a means of altering the cartilage environment to mimic features of OA observed by Regan et al, such as decreased SOD3, increased 3-NT in the ECM, and decreases in proteoglycan, Supplemental fig 4. Our results build upon existing chronic literature on SOD3 and imply it could be important to mediating the acute extracellular redox biochemistry observed[12, 13]. Conversely, samples without heparin treatment showed a more uniform distribution of DMPO staining over a wider area, suggesting that oxidative damage may be spread out over a larger area, away from cracks. Whether this change is directly dependent upon specific species or not, our data suggest a concentration of oxidative damage immediately adjacent to cracks in unhealthy tissue with redox characteristics similar to OA cartilage. This high concentration of oxidation might cause accelerated local degeneration if it results in more significant production of ECM breakdown products associated with pro-catabolic cascades. Future studies will explore more specific manipulations of the extracellular redox environment targeting O2•−. We speculate that SOD3 may be serving a protective function, preventing accumulation of high levels of damage at one location.

Damage to the collagen network is slow to repair because collagen synthesis in humans slows down substantially after reaching skeletal maturity[30]. We hypothesize that the damage created by cracking or injury might be a contributor to PTOA. This might operate though damaged pieces of cartilage, creating fibronectin fragments and other well-recognized sources of inflammation to contribute to disease. Chronically, recurring or ongoing damage at any sites like those described here represent an additional contributor to joint degeneration after injury. Further study is required to better understand the role of damage-induced free radical formation in the pathology of OA; nonetheless, this study demonstrates a new form of molecular damage to cartilage from injury resulting in free radicals that can be directly measured in articular cartilage ECM.

Supplementary Material

Supplemental Figures-MRH

ACKNOWLEDGEMENTS:

In addition to our funders at NIAMS, we would like to acknowledge the gift of the anti-DMPO antibody from Ronald Mason, PhD, National Institute of Environmental Health Sciences. The University of Iowa Orthopedics Department for ongoing support, and the Orthopedic Histology Service Center for their outstanding technical support.

Funding Sources:

This work was supported by the National Institute of Arthritis and Musculoskeletal and Skin (AR070914) and the National Institutes of Health [R01 DK124510-01 and R01 HL153532-01A1].

REFERENCES:

  • 1.Brouillette MJ, et al. , Strain-dependent oxidant release in articular cartilage originates from mitochondria. Biomech Model Mechanobiol, 2014. 13(3): p. 565–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Coleman MC, et al. , Injurious Loading of Articular Cartilage Compromises Chondrocyte Respiratory Function. Arthritis Rheumatol, 2016. 68(3): p. 662–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Coleman MC, et al. , Targeting mitochondrial responses to intra-articular fracture to prevent posttraumatic osteoarthritis. Sci Transl Med, 2018. 10(427). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Gomez-Contreras PC, et al. , Intersections Between Mitochondrial Metabolism and Redox Biology Mediate Posttraumatic Osteoarthritis. Curr Rheumatol Rep, 2021. 23(5): p. 32. [DOI] [PubMed] [Google Scholar]
  • 5.Khoo NK, et al. , In Vivo Immuno-Spin Trapping: Imaging the Footprints of Oxidative Stress. Curr Protoc Cytom, 2015. 74: p. 12.42.1–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Mason RP, Using anti-5,5-dimethyl-1-pyrroline N-oxide (anti-DMPO) to detect protein radicals in time and space with immuno-spin trapping. Free Radic Biol Med, 2004. 36(10): p. 1214–23. [DOI] [PubMed] [Google Scholar]
  • 7.Monboisse JC and Borel JP, Oxidative damage to collagen. Exs, 1992. 62: p. 323–7. [DOI] [PubMed] [Google Scholar]
  • 8.Monboisse JC, et al. , Effect of oxy radicals on several types of collagen. Int J Tissue React, 1984. 6(5): p. 385–90. [PubMed] [Google Scholar]
  • 9.Lloyd RV and Mason RP, Evidence against transition metal-independent hydroxyl radical generation by xanthine oxidase. J Biol Chem, 1990. 265(28): p. 16733–6. [PubMed] [Google Scholar]
  • 10.Uchida K, Kato Y, and Kawakishi S, A novel mechanism for oxidative cleavage of prolyl peptides induced by the hydroxyl radical. Biochem Biophys Res Commun, 1990. 169(1): p. 265–71. [DOI] [PubMed] [Google Scholar]
  • 11.Buettner GR, The pecking order of free radicals and antioxidants: lipid peroxidation, alpha-tocopherol, and ascorbate. Arch Biochem Biophys, 1993. 300(2): p. 535–43. [DOI] [PubMed] [Google Scholar]
  • 12.Pate KM, et al. , The beneficial effects of exercise on cartilage are lost in mice with reduced levels of ECSOD in tissues. J Appl Physiol (1985), 2015. 118(6): p. 760–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Regan E, et al. , Extracellular superoxide dismutase and oxidant damage in osteoarthritis. Arthritis Rheum, 2005. 52(11): p. 3479–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Mason RP, Imaging free radicals in organelles, cells, tissue, and in vivo with immuno-spin trapping. Redox Biol, 2016. 8: p. 422–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Mason RP and Ganini D, Immuno-spin trapping of macromolecules free radicals in vitro and in vivo - One stop shopping for free radical detection. Free Radic Biol Med, 2019. 131: p. 318–331. [DOI] [PubMed] [Google Scholar]
  • 16.Summers FA, Mason RP, and Ehrenshaft M, Development of immunoblotting techniques for DNA radical detection. Free Radic Biol Med, 2013. 56: p. 64–71. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Khoo NK, et al. , Obesity-induced tissue free radical generation: an in vivo immuno-spin trapping study. Free Radic Biol Med, 2012. 52(11–12): p. 2312–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Martin JA, et al. , N-acetylcysteine inhibits post-impact chondrocyte death in osteochondral explants. J Bone Joint Surg Am, 2009. 91(8): p. 1890–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Sauter E, et al. , Cytoskeletal dissolution blocks oxidant release and cell death in injured cartilage. J Orthop Res, 2012. 30(4): p. 593–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Wolff KJ, et al. , Mechanical stress and ATP synthesis are coupled by mitochondrial oxidants in articular cartilage. J Orthop Res, 2013. 31(2): p. 191–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Coleman MC, et al. , Differential Effects of Superoxide Dismutase Mimetics after Mechanical Overload of Articular Cartilage. Antioxidants (Basel), 2017. 6(4). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Grzelak A, Rychlik B, and Bartosz G, Light-dependent generation of reactive oxygen species in cell culture media. Free Radic Biol Med, 2001. 30(12): p. 1418–25. [DOI] [PubMed] [Google Scholar]
  • 23.Gao Y, et al. , The ECM-cell interaction of cartilage extracellular matrix on chondrocytes. Biomed Res Int, 2014. 2014: p. 648459. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Lee JI, et al. , Measurement of diffusion in articular cartilage using fluorescence correlation spectroscopy. BMC Biotechnol, 2011. 11: p. 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Leddy HA, Christensen SE, and Guilak F, Microscale diffusion properties of the cartilage pericellular matrix measured using 3D scanning microphotolysis. J Biomech Eng, 2008. 130(6): p. 061002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Nimer E, Schneiderman R, and Maroudas A, Diffusion and partition of solutes in cartilage under static load. Biophys Chem, 2003. 106(2): p. 125–46. [DOI] [PubMed] [Google Scholar]
  • 27.Hunziker EB, Herrmann W, and Schenk RK, Ruthenium hexammine trichloride (RHT)-mediated interaction between plasmalemmal components and pericellular matrix proteoglycans is responsible for the preservation of chondrocytic plasma membranes in situ during cartilage fixation. J Histochem Cytochem, 1983. 31(6): p. 717–27. [DOI] [PubMed] [Google Scholar]
  • 28.Motulsky HJ and Brown RE, Detecting outliers when fitting data with nonlinear regression - a new method based on robust nonlinear regression and the false discovery rate. BMC Bioinformatics, 2006. 7: p. 123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Gibson-Corley KN, Olivier AK, and Meyerholz DK, Principles for valid histopathologic scoring in research. Vet Pathol, 2013. 50(6): p. 1007–15. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Heinemeier KM, et al. , Radiocarbon dating reveals minimal collagen turnover in both healthy and osteoarthritic human cartilage. Sci Transl Med, 2016. 8(346): p. 346ra90. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Figures-MRH
NIHMS1853609-supplement-Supplemental_Figures-MRH.docx (42.1MB, docx)

RESOURCES