Osteoarthritis is a leading cause of decreased athletic performance in horses. Results of multiple studies1–3 indicate that bone marrow–derived MSCs are beneficial for the treatment of various degenerative, inflammatory, and immune-mediated conditions, including the early stages of osteoarthritis. Contemporary evidence indicates that the regenerative effects of MSCs are the result of their secretion of anti-inflammatory cytokines and chemokines, which alter the recipient environment and facilitate tissue regeneration.4–6 Although the effect of MSCs on the recipient environment is fairly well documented, less is known about how the environment might influence MSC function.
Allogeneic MSCs have the potential to provide a more reliable, readily available uniform product and are not subject to variation caused by patient age as are autologous MSCs. Traditionally, MSCs were considered immune privileged because they were reported to be negative for MHC class II and presumed incapable of eliciting an immune response.7 However, that concept has since been challenged.8–15 Results of an in vitro study13 involving equine bone marrow–derived MSCs indicate that cell surface expression of MHC class II is variable and upregulated when the MSCs are exposed to IFN-γ, which suggests the acquisition of antigenicity, particularly in an inflamed environment. Major histocompatibility complex genes code for antigen-presenting molecules and have a critical role in immune system activation. Specifically, MHC class II molecules interact with CD4+ T cells, triggering both a local inflammatory response and systemic antibody production.16
The dynamic nature of MHC class II expression is a serious concern for allogeneic transplantation, particularly if MHC class II expression is affected by the recipient. In vitro exposure of MSCs to inflammatory signals such as lipopolysaccharide or IL-1β increases production of proinflammatory cytokines (TNF-α, IL-1β, IL-6, and IL-8).17 It is unknown whether similar stimulation increases MHC class II expression. Regarding the use of allogeneic MSCs for the treatment of joint disease in horses, it is currently unknown how an inflamed joint environment affects MHC class II expression. The aim of the study reported here was to determine whether expression of MHC class II in equine MSCs changes when those cells are exposed to a proinflammatory environment similar to that of a joint in the early stages of osteoarthritis. The hypothesis was that MHC class II expression by equine MSCs would increase after exposure to proinflammatory cytokines (IL-1β and TNF-α), which are commonly present in inflamed joints.
Materials and Methods
The study was divided into 3 parts. In part 1, the use of qRT-PCR analysis as a surrogate measure of MHC class II cell surface marker expression was evaluated. In part 2, an IL-1β–stimulated synoviocyte-cartilage coculture model was used to generate a conditioned medium that was designed to mimic the environment of an inflamed joint. In part 3, the response of equine MSCs to the conditioned medium and individual proinflammatory cytokines was examined. All study procedures involving animals were approved by the Cornell University Institutional Animal Care and Use Committee.
Comparison of flow cytometry and a qRT-PCR analysis for MHC class II evaluation (part 1)
Previously cryopreserved, passage 3–4 bone marrow-derived MSCs from 4 horses were thawed and maintained in MSC medium as described13 until they reached 70% confluence. Cells were passaged and treated with MSC medium with or without IFN-γa (100 ng/mL).13,18 Mesenchymal stem cells were collected immediately before (baseline) and after 24, 48, and 96 hours of incubation for flow cytometry and qRT-PCR analysis. The assigned medium was replaced 48 hours after initiation of incubation to adequately maintain the MSCs that were collected after 96 hours of incubation. Expression of MHC class I and MHC class II was analyzed by flow cytometry as described.13 For the qRT-PCR analysis, RNA was extracted from MSCs by use of a commercially available RNA extraction kitb in accordance with the manufacturer's instructions. Then, 200 ng of extracted RNA was used as a template to synthesize cDNA with a commercially available synthesis kit.c All cDNA samples were checked for genomic contamination, with the presence of a single band of product indicating no genomic contamination on the results of a β2-microglobulin PCR assay. A SYBR green-based qRT-PCR assay was performed by use of a PCR instrumentd in accordance with the manufacturer's instructions. Relative gene expression of MHC class I and MHC class II following INF-γ stimulation was analyzed with SCAMP3 as a housekeeping gene; the primers used for the MHC class I,19 MHC class II,e and SCAMP320 genes have been previously described. All samples were run in duplicate, and the mean CT value for each sample was calculated. Data were analyzed with the ACT method, where fold change is expressed as 2−-ΔΔCT.21 The CT values for SCAMP3, MHC class I, and MHC class II were also compared with previously validated standards to determine absolute gene copy numbers.
Generation and verification of conditioned media (part 2)
Generation of conditioned media—Cartilage and synovium were collected from the femoropatellar, femorotibial, and scapulohumeral joints of a 1-year-old horse that was euthanized for reasons unrelated to lameness. Synoviocyte-cartilage cocultures were established as described22 with minor modifications. Synovium from the femoropatellar joints was collected and processed to isolate synoviocytes, which were then added to the wells of cell culture plates (1.6 × 106 synoviocytes/well). The plates were incubated at 37°C and 90% humidity in a 5% CO2 environment for 48 hours prior to addition of cartilage explants.
Full-thickness cartilage samples were removed from the trochlear ridges and femoral and humeral condyles. The cartilage samples were cut into 4 × 4-mm pieces and divided into transwell insertsf so that each insert contained approximately 0.4 g of cartilage (range, 0.38 to 0.42 g). Cartilage explants were cultured in standard synoviocyte medium as described22 for 24 hours, after which the medium was replaced with medium containing 2% FBS with or without recombinant equine IL-1βa (10 ng/mL). The explants were cultured for an additional 24 hours and then added to the synoviocyte cultures (4 explants/well). The medium for the cocultures was replaced with fresh Dulbecco modified Eagle medium containing 2% FBS with or without IL-1β (10 ng/mL). Cartilage explants previously exposed to IL-1β were again treated with IL-1β, whereas those not previously exposed to IL-1β were treated with medium without IL-1β. The concentration of IL-1β (10 ng/mL) used was selected because that is the concentration frequently used to model osteoarthritis in vitro.17 The cocultures were incubated for 24 hours. The medium was replaced with Dulbecco modified Eagle medium with 10% FBS, and the cocultures were incubated for an additional 48 hours, at which time the medium was collected from each coculture. Thus, there were 2 types of coculture media collected: a control-conditioned medium that was obtained from synoviocyte-cartilage cocultures that were never exposed to IL-1β, and an IL-1β-conditioned medium that was obtained from synoviocyte-cartilage cocultures that were treated with IL-1β. The media samples were frozen and stored at −80°C. Samples collected from each joint were kept separate throughout the experiments.
Conditioned media verification—To confirm that control-conditioned and IL-1β-conditioned media were generated, chondrocytes were used to assess the functionality of the conditioned media. Chondrocytes in chondrocyte medium23 were added to individual wells (200,000 chondrocytes/well) of a 24-well plate.f The plate was incubated for 24 hours, after which the medium of each well was replaced with 1 of 4 treatments (chondrocyte medium with 2% FBS, chondrocyte medium with 2% FBS and IL-1β [10 ng/mL], control-conditioned medium, or IL-1β–conditioned medium). All samples were run in duplicate, and conditioned media isolated from 2 different joints were used to verify uniformity between coculture samples. The plate was incubated for another 24 hours, then the chondrocytes were lysed and RNA was isolated. The concentration of RNA in each sample was determined by use of a spectrophotometer.g For each sample, the RNA was then diluted to a concentration of 5 ng/mL and mixed with prediluted primers and fluorescent probesd (MMP-13 and 18S) as described.24 Gene expression of MMP-13 was analyzed by use of a real-time PCR assayd with 18S as the housekeeping gene. Each sample was assayed in duplicate, and the mean CT value was calculated and used for analysis. The data were analyzed with the ACT method as described,21 where fold change was expressed as 2−-ΔΔCT.
MHC class II gene expression after exposure to coculture media (part 3)
Passage 3–4 MSCs from each of 8 horses were inoculated (1.5 × 105 MSCs/well) with standard MSC medium into duplicate wells of a cell culture plate and incubated for 24 hours, after which the medium in each well was replaced with 1 of 6 treatment media (standard MSC medium [untreated control treatment], control-conditioned medium [control-conditioned treatment], IL-1β-conditioned medium [IL-1β-conditioned treatment], MSC medium with IL-1β [10 ng/mL; IL-1β treatment], MSC medium with TNF-α [10 ng/mL; TNF-α treatment], or MSC medium with IFN-γ [100 ng/mL; IFN-γ treatment]). The cytokine concentrations used in the treatment media were selected on the basis of the cytokine concentrations used in similar experiments.10,13 The MSCs were incubated with the treatment medium for 48 hours, the treatment medium was replaced, and the cells were incubated for an additional 48 hours. Thus, the MSCs were incubated with the assigned treatment medium for 96 hours.
Following incubation, the MSCs from each well were collected for RNA analysis. The RNA was extracted from each sample, cDNA was created, and a qRT-PCR assay was performed as previously described for part 1. Previously validated standards for SCAMP3, MHC class I, and MHC class II were also assayed to allow for comparisons of gene expression among horses. Cytokines (IL-6 and IL-8) and chemokines (CCL2 and CXCL10) associated with proinflammatory MSCs were also analyzed.
Relative gene expressions of MHC class I, MHC class II, IL-6, IL-8, CCL2, and CXCL10 for each treatment, compared with those for the untreated control treatment, were analyzed with SCAMP3 used as a housekeeping gene and reported as previously described. The primers used for SCAMP3,20 MHC class I,19 MHC class II,e IL-6,25 IL-8,26 CCL2,27 and CXCL1027 have been described. Additionally, gene expression of MHC class I and MHC class II was compared with that for standards to determine the absolute gene copy numbers in each sample.
Statistical analysis
For each gene evaluated (MHC class I, MHC class II, IL-6, IL-8, CCL2, and CXCL10), the difference in the mean expression relative to that for the untreated control treatment (log[2−-ΔΔCT]) was determined by use of a linear mixed-effects model in which horse was included as a random effect followed by a Tukey honest-significant-difference test for multiple comparisons. Values of P < 0.05 were considered significant for all analyses. All calculations were performed with statistical software.h
Results
Part 1
Stimulation of equine MSCs with IFN-γ resulted in a significant increase in cell surface expression of MHC class II after 24, 48, and 96 hours of incubation as determined by both flow cytometry and qRT-PCR analysis in all samples (Figure 1). Interferon-γ stimulation resulted in an increase in the percentage and mean fluorescent intensity of MSCs positive for MHC class II over time (Table 1), with the greatest percentage of MCH class II–positive cells and highest mean fluorescent intensity recorded after 96 hours of incubation. On the basis of the qRT-PCR assay results, the time at which the maximum increase in MHC class II expression was recorded varied among horses. For example, maximum MHC class II expression was recorded at 48 hours after initiation of IFN-γ stimulation for 2 horses, and MHC class II expression did not differ from baseline until 96 hours after initiation of IFN-γ stimulation for 1 horse. Because MHC class II expression was significantly increased from baseline in all horses at 96 hours after initiation of IFN-γ stimulation, that time point was used for analysis for all further experiments.
Cell surface expression of MHC class II on equine MSCs at 24, 48, and 96 hours after incubation in MSC medium with (solid line) or without (dashed line) IFN-γ (100 ng/mL) as determined by flow cytometry (A) and a qRT-PCR assay (B; part 1). Passage 3–4 bone marrow–derived MSCs from each of 4 horses were assayed in duplicate with and without IFN-γ. For each sample with and without IFN-γ stimulation, the results of flow cytometry were reported as the mean percentage of MHC class II–positive cells, and the results of the qRT-PCR assay were reported as the mean MHC class II gene copy number. The overall mean percentage of MHC class II–positive cells and mean MCH class II gene copy number for MSCs that were stimulated with IFN-γ were significantly (P < 0.05) greater than the corresponding values for MSCs that were not stimulated with IFN-γ at all times. *Within a treatment, value is significantly greater (P < 0.05) than the values at the preceding times. Results indicated that results of the qRT-PCR assay were comparable to those obtained by flow cytometry; therefore, qRT-PCR assay results could be used as an accurate measure of MHC class II cell surface expression.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.867
Flow cytometry and qRT-PCR assay results for cell surface expression of MHC class II on equine MSCs immediately before (baseline) and at 24, 48, and 96 hours after incubation in MSC medium with IFN-γ (100 ng/mL; part 1).
| Flow cytometry | qRT-PCR assay | |||
|---|---|---|---|---|
| Time (h) | Positive cells (%) | Mean fluorescence | Relative gene expression (2−ΔΔCT) | Gene copy No. |
| Baseline | 8.53 (2.7–14.1) | 1,202 (769–5,275) | 1 | 2.7 (0–907.1) |
| 24 | 50.2 (6.9–76.3) | 1,852 (758–2,976) | 190.2 (56.7–4,349.4) | 54,133.7 (7,193.1–227,462.8) |
| 48 | 36.2 (14.1–67.3) | 2,269 (1,215–3,027) | 541.5 (153.8–64,644.7) | 266,923.7 (54,091.9– 1,429,963.1) |
| 96 | 97.6 (49.4–99.3) | 6,176 (3,333–7,265) | 6,793.8 (98.3–46,697.1) | 881,756.6 (2,066.3–13,509,981.1) |
Values represent the median (95% confidence interval); the absence of a 95% confidence interval indicates that the result was the same for all MSC samples evaluated. Passage 3–4 bone marrow–derived MSCs from each of 4 horses were assayed in duplicate; the mean was calculated for each horse and used for analysis. The treatment medium was replaced every 48 hours.
Part 2
Mean chondrocyte MMP-13 expression increased 11.4-fold (range, 9.4- to 13.1-fold) from baseline following culture with the IL-1β–conditioned medium, which indicated successful generation of a proinflammatory conditioned medium. There was an unanticipated 2.5-fold (range, 2.3- to 2.7-fold) increase in mean chondrocyte MMP-13 expression for cells cultured in the control-conditioned medium (control-conditioned treatment), compared with cells cultured in untreated control medium (standard chondrocyte medium23). The magnitude of MMP-13 upregulation did not differ significantly among chondrocytes that originated from different joints when cultured in either control-conditioned medium or IL-1β–conditioned medium.
Part 3
MHC class I expression—Mean relative gene expression of MHC class I for the TNF-α and IFN-γ treatments was significantly increased, compared with the mean relative gene expression of MHC class I for the untreated control treatment (Table 2). However, the mean relative gene expression of MHC class I for the control-conditioned, IL-1β-conditioned, and IL-1β treatments did not differ significantly from that for the untreated control treatment.
Gene expression of MHC class I and MHC class II on equine MSCs following incubation for 96 hours with each of 6 treatments (standard MSC medium [untreated control], control-conditioned medium, IL-1β–conditioned medium, MSC medium with IL-1β [10 ng/mL], MSC medium with TNF-α [10 ng/mL], or MSC medium with IFN-γ [100 ng/mL]; part 3).
| Gene | Treatment | Relative gene expression (2−ΔΔCT) | Gene copy No. |
|---|---|---|---|
| MHC class I | Untreated controla | 1 | 702,942.1 (28,864.1–13,003,747.8) |
| Control conditioneda | 0.5 (0.3–2.6) | 318,977.8 (11,167.7–8,167,925.2) | |
| IL-1β conditioneda | 0.4 (0.2–2.6) | 220,574.0 (25,110.0–2,427,498.4) | |
| IL-1βa | 0.7 (0.5–13.7) | 836,551.4 (32,847.8–11,803,562.0) | |
| TNF-αb | 4.6 (0.7–45.2) | 1,451,864.8 (216,705.7–25,206,851.8) | |
| IFN-γb | 7.2 (0.5–87.0) | 2,301,583.0 (497,250.9–85,429,304.4) | |
| MHC class II | Untreated controla | 1 | 3,012.0 (275.9–922,588.2) |
| Control conditioneda,b | 0.6 (0.1–1.8) | 1,339.4 (77.2–597,848.4) | |
| IL-1β conditionedb | 0.2 (0–1.5) | 444.7 (54.5–280,435.1) | |
| IL-1βb | 0.1 (0.1–1.4) | 608.9 (33.0–290,120.2) | |
| TNF-αa | 1.4 (0.1–12.9) | 11,331.8 (649.6–688,264.8) | |
| IFN-γc | 95.1 (0.5–1,655.1) | 482,770.4 (142,536.8–4,929,730.7) |
Passage 3–4 bone marrow–derived MSCs from each of 8 horses were incubated in duplicate with each treatment, and MCH class II gene expression was determined with a qRT-PCR assay. The mean for each sample-treatment combination was calculated and used for analysis.
Treatments with different superscript letters differ significantly (P < 0.05).
See Table 1 for remainder of key.
MHC class II expression—Mean relative gene expression of MHC class II for the control-conditioned and TNF-α treatments did not differ significantly from that for the untreated control treatment (Figure 2). The mean relative gene expression of MHC class II for the IL-1β-conditioned and IL-1β treatments was significantly decreased, whereas that for the IFN-γ treatment was significantly increased, compared with the mean relative gene expression of MHC class II for the untreated control treatment. Mean expression of MHC class II did not differ significantly among the control-conditioned, IL-1β-conditioned, and IL-1β treatments.
Box-and-whisker plots for gene expression of MHC class II in equine MSCs that were incubated for 96 hours in each of 6 treatments (standard MSC medium [control], control-conditioned medium, IL-1β-conditioned medium, MSC medium with IL-1β [10 ng/mL], MSC medium with TNF-α [10 ng/mL], or MSC medium with IFN-γ [100 ng/mL]; part 3). Passage 3–4 MSCs from each of 8 horses were incubated in duplicate with each treatment, and MHC class II gene expression was determined with a qRT-PCR assay. The mean for each sample-treatment combination was calculated, and the collective data for each treatment were plotted. For each plot, the horizontal line within the box represents the median; the lower and upper ends of the box represent the first and third quartiles, respectively; and the whiskers delimit the outmost data points that fall within 1.5 times the first and third quartiles, respectively. Dots represent outliers. The dashed horizontal line represents the control treatment (gene expression, 1) to which each treatment group was compared, so points above the line indicate an increase in gene expression and points below indicate a decrease in gene expression relative to the control. Mean gene expression for treatments with different letters differs significantly (P < 0.05).
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.867
Baseline expression of MHC class II varied among the 8 horses; 3 horses had baseline MHC class II expression that was substantially greater (mean, 300-fold) than that for the remaining 5 horses. Regardless of baseline MHC class II expression, all horses had an increase in MHC class II expression when either absolute gene copy numbers or fold changes were used for analysis (Table 2).
IL-6 and IL-8 expression—Mean expression of IL-6 and IL-8 for the IL-1β-conditioned, IL-1β, and TNF-α treatments was significantly greater than that for the untreated control treatment (Figure 3). For the IFN-γ treatment, mean expression of IL-6 but not IL-8 was significantly increased, compared with that for the untreated control treatment. For the control-conditioned treatment, mean expression of IL-8 was significantly increased, whereas mean expression of IL-6 did not differ significantly, compared with that for the untreated control treatment.
Box-and-whisker plots for cell surface expression of IL-6 (A), IL-8 (B), CXCL10 (C), and CCL2 (D) in equine MSCs that were incubated for 96 hours in each of the treatments described in Figure 2. Notice the scale of the y-axis varies among panels. See Figure 2 for remainder of key.
Citation: American Journal of Veterinary Research 78, 7; 10.2460/ajvr.78.7.867
CXCL10 and CCL2 expression—Mean expressions of CXCL10 and CCL2 for the TNF-α and IFN-γ treatments were significantly increased, compared with that for the untreated control treatment (Figure 3). Compared with the untreated control treatment, mean expression of CXCL10 for the control-conditioned and IL-1β treatments and CCL2 expression for the IL-1β-conditioned treatments were significantly decreased, whereas mean expression of CCL2 for the control-conditioned treatment did not differ significantly. Media supplemented with IL-1β (IL-1β–conditioned and IL-1β treatments) resulted in a significant decrease in CXCL10 expression and increase in CCL2 expression relative to those for the untreated control medium.
Discussion
Understanding how the environment of an inflamed joint affects the antigenicity of MSCs is critical when considering treatment of joint disease with allogeneic MSCs. If transplanted MSCs elicit an antigenic response, they will likely be eliminated by the host immune system and could generate an inflammatory response, thereby worsening the clinical condition of the patient. Contrary to our hypothesis, exposure of equine MSCs to the proinflammatory cytokine IL-1β resulted in a decrease, rather than an increase, in MHC class II expression.
Exposure of MSCs to IL-1β (IL-1β–conditioned or IL-1β treatments) resulted in a significant decrease in MHC class II expression, which suggested a reduction in the antigenicity of MSCs. The mechanism by which IL-1β decreases MHC class II expression has not been investigated in MSCs; however, it has been studied in other cell lines.28–30 For MHC class II transcription to occur, class II transactivator must be present. In human astroglioma cells, pretreatment with IL-1β prevents IFN-γ–mediated induction of MHC class II by diminishing expression of class II transactivator; however, treatment with IL-1β alone had no effect on MCH class II expression.31 The decrease in MHC class II expression observed for the equine MSCs of the present study indicated that those cells had low expression of MHC class II at baseline (prior to stimulation), which was further suppressed in the presence of IL-1β. That finding was consistent with the gene copy numbers, which likewise indicated a wide range of MHC class II expression among horses at baseline.
The fact that cell surface expression of MHC class II on MSCs decreased in response to IL-1β but did not change significantly in response to TNF-α was unanticipated because results of other studies30,32 indicate that MHC class II expression on MSCs increases in response to inflammatory cytokines. Interferon-γ is commonly used as a positive control for studying changes in MHC class II expression because it reliably increases MHC class II expression in MSCs and other cell types.13,29–31 Prior studies30,32 tested combinations of proinflammatory cytokines, all of which included IFN-γ, whereas cytokines were independently evaluated in the present study. It is likely that the increase in MHC class II expression observed in the previous studies30,32 was primarily influenced by IFN-γ and was not associated with the presence of other cytokines. Although IFN-γ is frequently used as a positive control, it is not detectable in acutely inflamed synovial fluid,33,34 which suggests that conditioned media and media containing individual cytokines without IFN-γ might better model the environment of an inflamed joint for evaluation of alterations in MHC class II expression than media that contain IFN-γ.
The changes in MSC cytokine expression observed following treatment with IL-1β or TNF-α in the present study reflected those reported for MSCs that are in a proinflammatory state,35 despite a decrease in MHC class II expression. Both IL-8 and IL-6 expression increased significantly following MSC treatment with IL-1β-conditioned, IL-1β, or TNF-α medium. Currently, the literature contains conflicting results regarding IL-8 and IL-6 expression in MSCs that are likely to stimulate a proinflammatory response35 or have enhanced immunomodulatory capabilities.36 Evaluation of the immunomodulatory capabilities of treated MSCs was beyond the scope of the present study; therefore, it is unclear how an inflamed joint environment may affect MSC function.
In the present study, although the MSC expression patterns were similar for the cytokines evaluated, the expression patterns differed for the 2 chemokines evaluated. Expression of CXCL10 paralleled that of MCH class II in that MSCs exposed to IL-1β had a decrease in CXCL10 expression. Chemokine (C-X-C motif) ligand-10 is a T-cell-specific chemokine that has an important role in driving T-cell chemotaxis.37 Because T cells primarily interact with MHC class II receptors, it is not surprising that a stimulus that decreases MHC class II expression would similarly affect CXCL10 expression. Conversely, compared with the untreated control treatment, there was an increase in CCL2 expression for all treatments except the control-conditioned treatment. Chemokine (C-C motif) ligand-2 is a potent chemokine for macrophage recruitment and activation,38 and it was expected that CCL2 expression would increase following stimulation with inflammatory mediators. In humans, MSC-mediated recruitment of macrophages directs proinflammatory macrophages toward macrophages with immunosuppressive properties.39 This phenomenon is still being investigated in horses; therefore, it is currently unknown whether increased MSC-mediated recruitment of macrophages will contribute to or diminish inflammation within equine joints.
Contrary to IL-1β, treatment of MSCs with TNF-α did not alter MHC class II expression. That finding was consistent with results of another study30 in which TNF-α variably induced or inhibited MHC class II expression in human promyelocytic cells and differentiated murine MSCs. Although TNF-α did not affect MHC class II expression in the equine MSCs of the present study, it did cause a significant increase in expression of MHC class I, IL-6, IL-8, CXCL10, and CCL2, which suggested that TNF-α may have an important role in influencing MSC function.
Consistent with results of other studies,5,7,13,18 MHC class I expression was high in all MSCs at baseline, and exposure of those cells to TNF-α resulted in a significant increase in MHC class I gene expression. Although the present study focused on the role of MHC class II on the initiation of an immune response, it should be noted that MHC class I can also elicit allorecognition of MSCs by the immune system. For example, allorecognition of MHC class I surface molecules can lead to direct cell lysis, and soluble MHC class I can initiate production of antibodies when cross-presented to antigen-presenting cells.40 Therefore, the changes in MHC class I expression observed in the present study should not be ignored during investigation of the overall antigenic potential of allogeneic MSCs.
A unique feature of the present study was the use of a synoviocyte-cartilage coculture model to generate conditioned media to mimic noninflamed and inflamed joint environments. In a previous study,22 the deleterious effects of IL-1β on cartilage for cocultures of synoviocytes and cartilage were decreased, compared with effects for cartilage monocultures, which suggested that some protective mechanism exists when the 2 cell types interact. The catabolic cascade in joints is thought to be initiated by IL-1β41; thus, IL-1β was chosen as the stimulatory cytokine for the coculture model. Measurement of MMP-13 expression in chondrocytes following exposure to the control-conditioned and IL-1β–conditioned media provided validation that a proinflammatory medium was generated.42 We chose to use a synoviocyte-cartilage coculture model to mimic an inflamed joint environment rather than synovial fluid from horses with clinical joint disease for 2 reasons. First, the model allowed for better control of the composition of the treatment media and eliminated the variability that would be inherent in the composition of synovial fluid from clinically inflamed joints. Second, the intrinsic viscosity of synovial fluid makes it difficult to work with in vitro, and it often has to be diluted, which can lead to false-negative results. A limitation of the coculture model used was that the exact composition of the conditioned media was unknown, and it may have lacked some cytokines and chemokines produced by other cells normally present in inflamed joints. The fact that cartilage and synovium samples were obtained from only 1 horse to generate the reagent for the conditioned media could be a limitation; however, results of a previous study22 suggest that cartilage and synovium samples from different horses respond similarly to IL-1β stimulation. Overall, the coculture model provided a reliable method to mimic the response of synoviocytes and cartilage to IL-1β and thus reflected at least some of the changes that would be observed in acutely inflamed joints.
Consistent with the findings of another study,13 MSC expression of MHC class II at baseline varied substantially among the 8 horses of the present study, with baseline MHC class II expression for 3 horses being approximately 300-fold that for the other 5 horses. That finding reinforces the fact that not all MSCs are negative for MHC class II expression, as was originally reported.7 It also highlights the challenge associated with trying to categorically classify cells as positive or negative for MHC class II expression. Results for part 1 of the present study in which MSCs were exposed to IFN-γ indicated that MSCs that would be classified as negative for MHC class II by flow cytometry had MHC class II gene copy numbers ranging from 0 to 900 as determined by qRT-PCR assay (Table 1). The extent of MHC class II gene transcription necessary for sufficient cell surface marker expression to activate an immune response is unknown. This is important because allogeneic MSCs stimulate antibody formation in mismatched recipients, regardless of whether the cells were classified as MHC class II positive or negative.43 The in vivo mechanisms that lead to antibody formation are complex, but it is likely that variability in what is considered negative MHC class II expression could be an important factor in whether an immune reaction is or is not initiated.
Results of the present study suggested that an inflamed joint environment would not cause upregulation of MHC class II expression, which implies that allogeneic MSCs can be used to treat inflamed joints without the risk of causing an immune reaction. However, it is important to recognize that, although the observed changes in MHC class II expression supported the use of allogeneic MSCs, there are many additional factors involved in determining whether an immune response is initiated and whether the MSCs would survive. Therefore, caution is still recommended regarding the use of allogeneic MSCs for the treatment of joint disease in horses because their efficacy and safety have yet to be determined.
Acknowledgments
Supported by the Harry M. Zweig Memorial Fund for Equine Research and Resident Research Grants Program through the College of Veterinary Medicine at Cornell University.
Presented in abstract form at the 2016 Surgery Summit of the American College of Veterinary Surgeons, Seattle, October 2016.
The authors thank Lynn Johnson for statistical consultation.
ABBREVIATIONS
| CCL | Chemokine (C-C motif) ligand |
| CT | Cycle threshold |
| CXCL | Chemokine (C-X-C motif) ligand |
| FBS | Fetal bovine serum |
| IFN | Interferon |
| IL | Interleukin |
| MHC | Major histocompatibility complex |
| MMP | Matrix metalloproteinase |
| MSC | Mesenchymal stem cell |
| qRT-PCR | Quantitative reverse transcriptase PCR |
| SCAMP | Secretory carrier-associated membrane protein |
| TNF | Tumor necrosis factor |
Footnotes
R&D Systems, Minneapolis, Minn.
Omega Bio-Tek, Norcross, Ga.
Bio-Rad Laboratories Inc, Hercules, Calif.
Applied Biosystems, Foster City, Calif.
College of Veterinary Medicine, Cornell University, Ithaca, NY: Personal communication, 2016.
Corning Inc, Corning, NY.
Thermo Fischer Scientific, Wilmington, Del.
JMP Pro, version 12.0.1, SAS Institute Inc, Cary, NC.
References
1. Papadopoulou A, Yiangou M, Athanasiou E, et al. Mesenchymal stem cells are conditionally therapeutic in preclinical models of rheumatoid arthritis. Ann Rheum Dis 2012; 71: 1733–1740.
2. Introna M, Rambaldi A. Mesenchymal stromal cells for prevention and treatment of graft-versus-host disease: successes and hurdles. Curr Opin Organ Transplant 2015; 20: 72–78.
3. Vangsness CT Jr, Farr J II, Boyd J, et al. Adult human mesenchymal stem cells delivered via intra-articular injection to the knee following partial medial meniscectomy. J Bone Joint Surg Am 2014; 96: 90–98.
4. Stagg J. Immune regulation by mesenchymal stem cells: two sides to the coin. Tissue Antigens 2007; 69: 1–9.
5. Carrade DD, Lame MW, Kent MS, et al. Comparative analysis of the immunomodulatory properties of equine adult-derived mesenchymal stem cells. Cell Med 2012; 4: 1–11.
6. Paterson YZ, Rash N, Garvican ER, et al. Equine mesenchymal stromal cells and embryo-derived stem cells are immune privileged in vitro. Stem Cell Res Ther 2014; 5: 90.
7. De Schauwer C, Meyer E, Van de Walle GR, et al. Markers of stemness in equine mesenchymal stem cells: a plea for uniformity. Theriogenology 2011; 75: 1431–1443.
8. Eliopoulos N, Stagg J, Lejeune L, et al. Allogeneic marrow stromal cells are immune rejected by MHC class I– and class II–mismatched recipient mice. Blood 2005; 106: 4057–4065.
9. Badillo AT, Beggs KJ, Javazon EH, et al. Murine bone marrow stromal progenitor cells elicit an in vivo cellular and humoral alloimmune response. Biol Blood Marrow Transplant 2007; 13: 412–422.
10. English K, Barry FP, Field-Corbett CP, et al. IFN-γ and TNF-α differentially regulate immunomodulation by murine mesenchymal stem cells. Immunol Lett 2007; 110: 91–100.
11. Zangi L, Margalit R, Reich-Zeliger S, et al. Direct imaging of immune rejection and memory induction by allogeneic mesenchymal stromal cells. Stem Cells 2009; 27: 2865–2874.
12. Pigott JH, Ishihara A, Wellman ML, et al. Investigation of the immune response to autologous, allogeneic, and xenogeneic mesenchymal stem cells after intra-articular injection in horses. Vet Immunol Immunopathol 2013; 156: 99–106.
13. Schnabel LV, Pezzanite LM, Antczak DF, et al. Equine bone marrow-derived mesenchymal stromal cells are heterogeneous in MHC class II expression and capable of inciting an immune response in vitro. Stem Cell Res Ther 2014; 5: 13.
14. Griffin MD, Ryan AE, Alagesan S, et al. Anti-donor immune responses elicited by allogeneic mesenchymal stem cells: what have we learned so far? Immunol Cell Biol 2013; 91: 40–51.
15. Ankrum JA, Ong JF, Karp JM. Mesenchymal stem cells: immune evasive, not immune privileged. Nat Biotechnol 2014; 32: 252–260.
16. Al-Daccak R, Mooney N, Charron D. MHC class II signaling in antigen-presenting cells. Curr Opin Immunol 2004; 16: 108–113.
17. Vézina Audette R, Lavoie-Lamoureux A, Lavoie JP, et al. Inflammatory stimuli differentially modulate the transcription of paracrine signaling molecules of equine bone marrow multipotent mesenchymal stromal cells. Osteoarthritis Cartilage 2013; 21: 1116–1124.
18. Guest DJ, Ousey JC, Smith MR. Defining the expression of marker genes in equine mesenchymal stromal cells. Stem Cells Cloning 2008; 1: 1–9.
19. Tallmadge RL, Lear TL, Antczak DF. Genomic characterization of MHC class I genes of the horse. Immunogenetics 2005; 57: 763–774.
20. Brosnahan MM, Miller DC, Adams M, et al. IL-22 is expressed by the invasive trophoblast of the equine (Equus caballus) chorionic girdle. J Immunol 2012; 188: 4181–4187.
21. Genc B, Dogrusoz U. A layout algorithm for signaling pathways. Inf Sci (Ny) 2006; 176: 135–149:
22. Gregg AJ, Fortier LA, Mohammed HO, et al. Assessment of the catabolic effects of interleukin-1β on proteoglycan metabolism in equine cartilage cocultured with synoviocytes. Am J Vet Res 2006; 67: 957–962.
23. Nixon AJ, Lust G, Vernier-Singer M. Isolation, propagation, and cryopreservation of equine articular chondrocytes. Am J Vet Res 1992; 53: 2364–2370.
24. Fortier LA, Motta T, Greenwald RA, et al. Synoviocytes are more sensitive than cartilage to the effects of minocycline and doxycycline on IL-1α and MMP-13-induced catabolic gene responses. J Orthop Res 2010; 28: 522–528.
25. Figueiredo MD, Salter CE, Andrietti ALP, et al. Validation of a reliable set of primer pairs for measuring gene expression by real-time quantitative RT-PCR in equine leukocytes. Vet Immunol Immunopathol 2009; 131: 65–72.
26. Padoan E, Ferraresso S, Pegolo S, et al. Real time RT-PCR analysis of inflammatory mediator expression in recurrent airway obstruction-affected horses. Vet Immunol Immunopathol 2013; 156: 190–199:
27. Wimer CL, Damiani A, Osterrieder N, et al. Equine herpesvirus type-1 modulates CCL2, CCL3, CCL5, CXCL9, and CXCL10 chemokine expression. Vet Immunol Immunopathol 2011; 140: 266–274.
28. Glimcher LH, Kara CJ. Sequences and factors: a guide to MHC class-II transcription. Annu Rev Immunol 1992; 10: 13–49.
29. Ting JP, Trowsdale J. Genetic control of MHC class II expression. Cell 2002; 109(suppl):S21–s33:
30. Watanabe Y, Jacob CO. Regulation of MHC class II antigen expression. Opposing effects of tumor necrosis factor-alpha on IFN-gamma-induced HLA-DR and Ia expression depends on the maturation and differentiation stage of the cell. J Immunol 1991; 146: 899–905.
31. Rohn W, Tang LP, Dong Y, et al. IL-1β inhibits IFN-γ-induced class II MHC expression by suppressing transcription of the class II transactivator gene. J Immunol 1999; 162: 886–896.
32. Barrachina L, Remacha AR, Romero A, et al. Effect of inflammatory environment on equine bone marrow derived mesenchymal stem cells immunogenicity and immunomodulatory properties. Vet Immunol Immunopathol 2016; 171: 57–65.
33. Benito MJ, Veale DJ, FitzGerald O, et al. Synovial tissue inflammation in early and late osteoarthritis. Ann Rheum Dis 2005; 64: 1263–1267.
34. Sokolove J, Lepus CM. Role of inflammation in the pathogenesis of osteoarthritis: latest findings and interpretations. Ther Adv Musculoskelet Dis 2013; 5: 77–94.
35. Bunnell BA, Betancourt AM, Sullivan DE. New concepts on the immune modulation mediated by mesenchymal stem cells. Stem Cell Res Ther 2010; 1: 34.
36. Djouad F, Charbonnier LM, Bouffi C, et al. Mesenchymal stem cells inhibit the differentiation of dendritic cells through an interleukin-6-dependent mechanism. Stem Cells 2007; 25: 2025–2032.
37. Ren G, Zhang L, Zhao X, et al. Mesenchymal stem cell-mediated immunosuppression occurs via concerted action of chemokines and nitric oxide. Cell Stem Cell 2008; 2: 141–150.
38. Vallés G, Bensiamar F, Crespo L, et al. Topographical cues regulate the crosstalk between MSCs and macrophages. Bio-materials 2015; 37: 124–133.
39. Marigo I, Dazzi F. The immunomodulatory properties of mesenchymal stem cells. Semin Immunopathol 2011; 33: 593–602.
40. Curry AJ, Pettigrew GJ, Negus MC, et al. Dendritic cells internalise and re-present conformationally intact soluble MHC class I alloantigen for generation of alloantibody. Eur J Immunol 2007; 37: 696–705.
41. Kapoor M, Martel-Pelletier J, Lajeunesse D, et al. Role of proinflammatory cytokines in the pathophysiology of osteoarthritis. Nat Rev Rheumatol 2011; 7: 33–42.
42. Mengshol JA, Vicenti M, Coon C, et al. Interleukin-1 induction of collagenase 3 (matrix metalloproteinase 13) gene expression in chondrocytes requires p38, c-Jun N-terminal kinase, and nuclear factor κB: differential regulation of collagenase 1 and collagenase 3. Arthritis Rheum 2000; 43: 801–811.
43. Pezzanite LM, Fortier LA, Antczak DF, et al. Equine allogeneic bone marrow-derived mesenchymal stromal cells elicit antibody responses in vivo. Stem Cell Res Ther 2015; 6: 54.
