Experimentally induced cartilage degeneration treated by pulsed electromagnetic field stimulation; an in vitro study on bovine cartilage

Background Osteoarthritis (OA) is the final result of progressive alterations to articular cartilage structure, composition and cellularity, followed by an increase in the concentration of pro-inflammatory cytokines in joint synovial fluid. Even though the effect of pulsed electromagnetic field (PEMF) stimulation in counteracting OA progression and inflammation is of increasing interest, because of its anabolic and anti-inflammatory properties, the present study aimed to improve the knowledge on cartilage extracellular matrix (ECM) and chondrocyte changes related to the exposure of PEMF, from a histological and histomorphometric point of view. Methods An in vitro OA model was realized, culturing bovine cartilage explants with a high dose of interleukin 1β (IL1β, 50 ng/ml) at different experimental times (24 h, and 7 and 21 days). The effects of PEMFs (75 Hz, 1.5 mT) were evaluated in cartilage explants treated with IL1β or not (control), in terms of cartilage structure, cellularity and proteoglycans, glycosaminoglycans, collagen II and transforming growth factor β1 synthesis by using histology, histomorphometry and immunohistochemistry. Results Making a comparison with control cartilage, IL1β-treated explants showed a decrease in cartilage matrix, structure and cellularity parameters. PEMFs were able to counteract the progression of OA acting on both cartilage cellularity and ECM in cartilage previously treated with IL1β. Normal distribution (Kolmogroc-Smirnov test) and homoscedasticity (Levene test) of data were verified, then, the non-parametric Kruskal Wallis test followed by Mann-Whiteny U test for pairwise comparisons were performed. The p-value was adjusted according to the Dunn-Sidak correction. Conclusions These results, obtained by culturing and treating cartilage explants from two different joints, confirmed that PEMF stimulation can be used as adjuvant therapy to preserve cartilage from detrimental effects of high inflammatory cytokine levels during OA.


Background
The presence of pro-inflammatory cytokines in the joint microenvironment favors cartilage degeneration and osteoarthritis (OA) progression with alterations of proteoglycans (PGs) and collagen fibers, prevalently collagen type II (Coll II) [1]. Among pro-inflammatory cytokines, interleukin1β (IL1β) plays a pivotal role in inducing degradation of extracellular matrix (ECM) components and synthesis of other pro-inflammatory cytokines, chemokines and proteolitic enzymes. Since cartilage has a poor intrinsic reparative capability if not treated, OA leads to progressive disability and requirement of joint arthroplasty [2].
Current therapeutic strategies to prevent OA include the use of nonsteroidal anti-inflammatory drugs (NSAIDs) [3], that act at articular and not cellular levels, intraarticular injection of hyaluronic acid (HA) and physical exercises [4], that act only in the muscle of the joint. These treatments relieve pain and inflammation and improve functionality, but do not resolve the pathological process once triggered, thus the outcomes are inconclusive and remain as long as these therapies are administered. For this reason, orthopedic research investigates chondroprotective treatments able to reduce the local inflammatory microenvironment and to favor damaged articular cartilage anabolic activity.
Pulsed electromagnetic field (PEMF) stimulation has been already studied and proposed for the regeneration of musculoskeletal tissues such as cartilage [5], bone [6,7], tendon [8] and ligament [9]. As far as cartilage tissue is concerned, PEMFs positively affect chondrocyte proliferation and ECM component synthesis [10]. However, the existing studies, performed with different PEMF physical parameters and exposure times, gave contradictory results [11][12][13][14][15], in terms of chondrocyte viability, DNA content, glycosaminoglycans (GAGs), Coll II, PGs and aggrecan gene expression and production. In a previous in vitro study, on bovine cartilage explants, we conducted a doseresponse study to find the optimal dosage, in terms of PEMF frequency and exposure intensity, able to stimulate significantly PG synthesis [16]. Furthermore, the stimulatory effect of PEMFs on PG synthesis in presence of IL1β and in combination with insulin-like growth factor 1 (IGF-I) was observed [17,18]. PEMF and IGF-I showed an additive effect on PG synthesis in human OA cartilage explants cultured in the absence or presence of IL1β [19]. In in vivo studies, PEMF stimulation was able to limit the progression of OA of increasing severity [20][21][22], to reduce IL1β and tumor necrosis factor-α (TNF-α) concentrations and to increase transforming growth factor-β1 (TGFβ1) in the synovial fluid of sheep treated with autologous osteochondral grafts [23]. In clinical studies, PEMF stimulation was successfully used to treat pain, knee swelling and functionality, after total knee arthroplasty [24,25] and in early knee OA patients [26]. In other studies, beneficial symptomatic effects [27,28] or additional effects to physical treatments [29] were not observed in OA patients.
Even if PEMFs act on the joint microenvironment and have anabolic and chondroprotective actions, the mechanisms by which PEMFs exert their effects on biological systems remain not completely understood. To our knowledge, no histological and histomorphometric evaluations have assessed the effects of PEMF stimulation in an inflammatory OA microenvironment obtained in vitro with the addition of a high IL1β dose.
The present study aimed to analyze whether PEMF stimulation (75 Hz, 1.5 mT), applied to bovine cartilage explants, derived from two different joints of the same animals, was able to counteract the catabolic effect of a high dose of IL1β. In comparison to 2D monolayer chondrocyte cultures, the cartilaginous 3D explants cultured with IL1β are a well-accepted method to measure and evaluate the effect and mechanism of action of a therapy, by simultaneously evaluating chondrocytes and cartilage matrix and by closely mimicking the clinical situation of OA joint, in which a high dose of IL1β is found in the early stages of OA [30] and its concentration can be further increased by surgical intervention [2].
It was hypothesized that the PEMF stimulation alone, without the addition of other biological stimuli, might improve the production of the most important cartilage ECM components (PGs, GAGs, COLL II) and anabolic factor (TGF-β1) and preserve cartilage structure, thus controlling OA development. To demonstrate this hypothesis it was adopted an already set and validated in vitro model of OA [18,31], which used a high dose of IL1β (50 ng/ml) to create an inflammatory OA microenvironment. After 24 h, 7 and 21 days of culture, PEMF effect, in combination or not with IL1β, was evaluated through histological, histomorphometric and immunohistochemical analyses. In this study it was observed that PEMFs were able to counteract the progression of OA acting on both cartilage cellularity and ECM in cartilage previously treated with IL1β.

Cartilage explant cultures and treatment conditions
Full-thickness explants of bovine articular cartilage were aseptically dissected, by using a 4 mm dermal punch (Stiefel Laboratories, Milan, Italy), from the metacarpophalangeal(MC) and metatarsophalangeal (MT) joints of 14-18-month-old animals (Limousine breed), as previously described [16][17][18]. The abattoir gave permission for the use of cartilage explants in this study and no ethical approval, from a recognized ethics committee, was required for this study. The cartilage was explanted shortly after the bovine slaughter and cartilage discs (three discs in each well) were cultured in 0.5 ml culture medium in multiwells (Nunc, Denmark, 1.6 cm the diameter of each well). Before PEMF exposure and IL1β administration, all explants were allowed to equilibrate in culture for 48 h in DMEM/F12 (Life Technologies, Monza, Italy) supplemented with 10 % FBS and antibiotics (penicillin 100 units/ml, streptomycin 0.1 mg/ml) (complete medium) and for an additional 48 h in medium without serum, at 37°C in an atmosphere of 5 % CO 2 . During the experiments, explants were cultured in complete medium in the absence and presence of IL1β (50 ng/ml) [18]. Half cultures were exposed to PEMF throughout the entire culture period (24 h, 7 and 21 days). Medium was changed at the beginning of the exposure (time 0) and every 3 days.
For each experimental time, 4 experimental groups were set up and each group was composed of 3 explant discs from MC and 3 more from MT joints: -Control group (CTR): cartilage explants; -IL1β group: cartilage explants submitted to IL1β administration; -PEMF group: cartilage explants exposed to PEMFs; -IL1β + PEMF group: cartilage explants submitted to IL1β and exposed to PEMFs.

Characteristics of PEMFs and exposure conditions
The PEMF generator system was the same as that used in previous studies [9, 16-18, 32, 33]. It consisted of a pair of circular Helmoltz coils of copper wire placed opposite to each other and in a signal generator (Igea S.p.A., Carpi, Italy). The multiwell plates were placed between this pair of Helmoltz coils, so that the plane of the coils was perpendicular to the multiwell plates, and the direction of the induced electric field was perpendicular to the direction of the magnetic field. The power generator produced a pulsed signal with the following parameters: the pulse duration was 1.3 ms and the frequency was 75Hz, yielding a duty cycle of 1/10. The intensity peak of the magnetic field was 1.5mT and the induced electric field, as detected with a standard coil probe (50 turns, 0.5 cm internal diameter of the coil probe, 0.2 copper diameter), was 0.07 mV/cm. The intensity of the magnetic field was detected between two coils from one side to the other, by the Hall probe of the Gaussmeter (LE, Gaussmeter DG500, USA), with a reading sensitivity of 0.2 %. Inside this area, where the multiwell plate was placed, the magnetic field was uniform. Cartilage explants were exposed to continuous PEMF for 24 h, 7 or 21 days. PEMF unexposed cultures were placed inside the same incubator at a distance where no difference from background magnetic field was observed when the PEMF generator was turned on. In each experiment, controls and treatments were performed in triplicate wells.

Histology
At the end of the experimental times, cartilage explants, cultured in the above mentioned conditions, were routinely processed for paraffin embedding procedures. Briefly, samples were fixed for 24 h in 10 % neutral buffered formalin solution in PBS, extensively rinsed in distilled water, dehydrated in graded alcohol solutions (70, 95 %, two times and 100 %, three times, one hour for each step), cleared in xylene e finally paraffin embedded. Sections (5 ± 1 μm) were cut along the longitudinal axis of the samples by a semi-automated microtome (Microm H340E, Germany) and stained with Toluidine Blue, for the PG quantification and Safranin O-Fast green staining, for the GAG quantification. Three non-consecutive sections for each sample were evaluated, by 2 independent histologists through a semi-quantitative score (modified O'Driscoll score) [34]. The modified O'Driscoll score (Table 1: minimum 0 = osteoarthritic cartilage -maximum 11 = normal intact cartilage) analyzed 4 cartilage aspects: Safranin O staining, surface regularity, cellularity and chondrocyte clustering.

Immunohistochemical stainings (TGFβ1 and Coll II)
Cartilage sections were dewaxed in decreasing graded ethanol solutions until PBS rinsing for 20 min and then immunostained for TGFβ1 and Coll II. Briefly, after fixation, sections were extensively rinsed in PBS and permeabilized by incubation in 0.3 % hydrogen peroxide in PBS solution for 15 min. For Coll II immunostaining, sections were pre-treated for antigen unmasking with 0.2 % Pronase (P-8811, Sigma, Mo, USA) solution in PBS for 30 min at 37°C. Then, 10 % normal serum was added for 1 h at room temperature to block nonspecific antibody binding and the primary antibodies (rabbit polyclonal antibody anti TGFβ1, sc-146 and mouse monoclonal antibody anti collagen II, sc-52658, 1:50 dilutions, Santa Cruz Biotechnology, CA, USA) were applied and incubated overnight at 4°C. After rinsing in PBS, slides were incubated with appropriate biotinylated secondary antibody and with horseradish peroxidasestreptavidin complex for 1 h each (ABC Staining System, Biotechnology, CA, USA). Sample reaction was developed with 3,3-diaminobenzidine substrate and permanently mounted. Negative controls, by omitting the primary antibody, were included to check proper specificity and performance of the applied reagents.  Table Transforms) based on selected ranges of RGB channels to measure the following parameters expressed as percentage: -GAGs, as the ratio between the Safranin O stained areas and the total ROI areas; -PGs, as the ratio between the Toluidine Blue stained areas and the total ROI areas; -TGFβ1, as the ratio between the immunopositive stained areas and the total ROI areas; -Coll II, as the ratio between the immunopositive stained areas and the total ROI areas.

Statistical analysis
Statistical analysis was performed using the IBM SPSS Statistics v.21.0 software (IBM Corp, USA). Firstly, normal distribution (Kolmogroc-Smirnov test) and homoscedasticity (Levene test) of data were verified, then, the non-parametric Kruskal Wallis test followed by Mann-Whiteny U test for pairwise comparisons were performed. The P-value was adjusted according to the Dunn-Sidak correction [adj p = 1 -(1 -p) k , where k is the number of pairwise comparisons]. Since the preliminary Kruskal-Wallis test by distinguish MC and MP joints did not show significant differences between the two harvesting sites within each type of treatment, data were analyzed when merged.
Because of the IL1β effect, the O'Driscoll score progressively worsened by 50-60 %, during the experimental times, in comparison to CTR (at 24 h p = 0.0005, at 7 days p = 0.0005 and at 21 days p = 0.04) (Fig. 6). By analyzing each score parameter (Table 2), it was noted
The O'Driscoll score results of the PEMF-treated group did not differ from those of the CTR group at each experimental time (Fig. 6). Taking into account each O'Driscoll score parameter (Table 2), PEMF stimulation improved chondrocyte clustering at 7 days (p = 0.0005, 100 %) and Safranin-O staining at 7 (p = 0.0005, 100 %) and 21 days (p = 0.03, 200 %) compared to that of the CTR group.

Discussion
The present study aimed to evaluate the effects and mechanisms of action of PEMF stimulation (75 Hz, 1.5 mT) in preserving cartilage from OA structural deterioration. The in vitro effects were investigated on bovine cartilage explants harvested from two different joints. To simulate the OA microenvironment, a high IL1β concentration (50 ng/ ml) was administered in explant cultures, as set in an our previous in vitro study [30], for a longer period of 21 days and the changes in the structure and cellularity of CTR and IL1β-treated cartilage explants, either exposed or not to PEMFs, were analyzed. The concentration of IL-1β found in the synovial fluid of patients affected by OA is between 0.068 and 0.33 pg/ml [35]. However, most of the in vitro literature studies, that aimed to re-create an inflammatory microenvironment, added 10 ng/ml of IL-1β in the medium of cartilage explants [36][37][38], while there are other authors that employed 50 or 100 ng/ml [39][40][41]. The so high concentration of IL-1β used in the present study was chosen to employed an intermediate cytokine dosage and to evaluate if PEMF stimulation is able to restore cartilage ECM biological and morphological properties also with a huge concentration of IL-1β than the most employed one (10 ng/ml). The current data confirmed the hypotheses of the study. The addition of a high dose of IL1β to the culture medium induced a significant reduction, not only in structural parameters, but also ECM components and TGFβ1 synthesis, a decline observed at all experimental times until 21 days. These results are in agreement with those of previous in vitro studies, that observed GAGs, PGs and Coll II loss with the addition of IL1β in the culture medium of human, bovine and horse cartilage explants [30,42,43]. Although the joint inflammatory microenvironment of OA consists of different inflammatory mediators, such as matrix metalloproteinases (MMPs), the disintegrin and metalloproteinases with thrombospondin motifs (ADAMTs) and other pro-inflammatory cytokines, TNF-α and Interleukin-6 (IL-6), data suggested that IL1β might have the strongest effect. Indeed, with the addition of IL1β in the culture medium, the most important components of cartilage ECM and TGFβ1 synthesis were reduced and other cartilage structural parameters (matrix and cellularity) were Secondly, the stimulation with PEMFs, applied to healthy explants, did not modify PG content or the structural parameters at any of the experimental times, and moreover, increases in GAGs and Coll II were observed after 24 h of culture. Instead, PEMFs improved all the analyzed ECM proteins and the structural parameters in IL1β-treated explants, also at 21 days of culture. Similar results were published by Brighton et al., in 2006 and 2008 using capacitively coupled electric fields (CCEF). They observed an increase in Coll II and PG production in human and bovine cartilage explants, treated or not with IL1β, after stimulation [44,45]. Finally, in this study it was also observed that IL1β and PEMFs have an opposite effect on TGFβ1, the anabolic growth factor that induces matrix production, mesenchymal stem cell chondrogenic differentiation and chondrocyte proliferation [46].
This study allowed a more complete view of the role of PEMFs that has been already investigated in previous our in vitro studies [18,31]. The strengths of the study were: 1) the histological and histomorphometric evaluations on in vitro cartilage explants, stimulated with PEMF parameters already used in patients suffering from knee pain due to cartilage degeneration [47]; 2) the long-term culture of cartilaginous explants (21 days) and 3) the comparative analysis in two different joints. A limitation of the study was the absence, in the culture explants model, of the subchondral bone, that is recognized to play a role in OA development. It is noteworthy that the score employed in this study is a modification of the score proposed by O' Driscoll SW, et al., in 1988 [34]: only the parameters that regard the matrix, the structure and the chondrocyte hypocellularity and clustering could be evaluated.
In comparison to 2D monolayer chondrocyte cultures and in vivo situation, the 3D cartilage explants, adopted in this study, are a more homogenous model to evaluate cartilage metabolism. Chondrocytes are maintained in their physiological microenvironment, the de-differentiation phenomenon is prevented and the inter-individual variables are reduced [48]. In addition, the use of cartilage explants is a more suitable in vitro method than chondrocytes embedded in alginate cultures to study PEMF and IL1β effects, because the chondrocytes are embedded in their native ECM. In fact the changes in ECM composition may play an important role in maintaining osteoarthritis damage and the alterations in hyaluronate degradation can be an alternative mechanism for the regulation of proteoglycan release from cartilage following IL1β stimulation [49]. 3D explants are reliable even at long experimental times because the effects of IL1β treatment and PEMF stimulation were evident also at 21 days of culture. Many in vitro culture studies evaluated bovine cartilage because of the advantageous easy availability and similar thickness to human cartilage [50]. Moreover, the metabolic activity and structural composition of bovine cartilage is reported to be quite similar to those of humans [50].

Conclusions
This study shows that the addition of IL1β to the culture medium has a detrimental effect on cartilage explants by reducing PG, GAG, Coll II and TGFβ1 quantification and Chondrocyte clustering 2 (1-2) 0.5 (0-1)** 2 (2-2) 1 (1-2) Chondrocyte clustering O'Driscoll modified score parameters. PEMF stimulation is able to counteract these catabolic effects both on ECM and chondrocyte features. It was also found that PEMFs have a greater effect on cartilage explants compromised by IL1β in comparison to normal tissues. These results confirm the role of PEMFs in chondroprotection and, as shown by the long-term results, they suggest their ability to limit OA progression.

Competing interests
The authors declare that they have no competing interests. Stefania Setti is employed in IGEA, that developed the PEMF generator.
Authors' contributions FV has been involved in acquisition, analysis and interpretation of data. MF has made substantial contributions to conception and design of the study and she has been involved in revising the manuscript critically for important intellectual content. GG has made substantial contributions to analysis and interpretation of data and carried out statistical analysis. AO has made substantial contributions to conception and design of the study and was involved in the in vitro culture of cartilage explants. MDM has made substantial contributions to conception and design of the study and revised the manuscript. AP has made substantial contributions to acquisition and analysis of data. SS provided the PEMF generator and has made substantial contributions to analysis and interpretation of data. MT has made substantial contributions to analysis and interpretation of data and has been involved in revising the manuscript critically for important intellectual content. All authors have been involved in drafting the manuscript, have given final approval of the version to be published and agree to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.