In search for cross-reactivity to immunophenotype equine mesenchymal stromal cells by multicolor flow cytometry

Authors

  • Catharina De Schauwer,

    Corresponding author
    1. Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    • Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Sofie Piepers,

    1. Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Gerlinde R. Van de Walle,

    1. Department of Comparative Physiology and Biometrics, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Kristel Demeyere,

    1. Laboratory of Biochemistry, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Maarten K. Hoogewijs,

    1. Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Jan L. J. Govaere,

    1. Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Kevin Braeckmans,

    1. Laboratory of General Biochemistry and Physical Pharmacy, Faculty of Pharmaceutical Sciences, Ghent University, Harelbekestraat 72, Ghent 9000, Belgium
    Search for more papers by this author
  • Ann Van Soom,

    1. Department of Reproduction, Obstetrics and Herd Health, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author
  • Evelyne Meyer

    1. Laboratory of Biochemistry, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, Merelbeke 9820, Belgium
    Search for more papers by this author

Abstract

During recent years, cell-based therapies using mesenchymal stem cells (MSC) are reported in equine veterinary medicine with increasing frequency. In most cases, the isolation and in vitro differentiation of equine MSC are described, but their proper immunophenotypic characterization is rarely performed. The lack of a single marker specific for MSC and the limited availability of monoclonal antibodies (mAbs) for equine MSC in particular, strongly hamper this research. In this study, 30 commercial mAbs were screened with flow cytometry for recognizing equine epitopes using the appropriate positive controls to confirm their specificity. Cross-reactivity was found and confirmed by confocal microscopy for CD45, CD73, CD79α, CD90, CD105, MHC-II, a monocyte marker, and two clones tested for CD29 and CD44. Unfortunately, none of the evaluated CD34 clones recognized the equine epitopes on positive control endothelial cells. Subsequently, umbilical cord blood-derived undifferentiated equine MSC of the fourth passage of six horses were characterized using multicolor flow cytometry based on the selected nine-marker panel of both cell surface antigens and intracytoplasmatic proteins. In addition, appropriate positive and negative controls were included, and the viable single cell population was analyzed by excluding dead cells using 7-aminoactinomycin D. Isolated equine MSC of the fourth passage were found to be CD29, CD44, CD90 positive and CD45, CD79α, MHC-II, and a monocyte marker negative. A variable expression was found for CD73 and CD105. Successful differentiation towards the osteogenic, chondrogenic, and adipogenic lineage was used as additional validation. We suggest that this selected nine-marker panel can be used for the adequate immunophenotyping of equine MSC. © 2012 International Society for Advancement of Cytometry

The considerable therapeutic potential of equine mesenchymal stem cells (MSC) in regenerative medicine has generated a markedly increasing interest in this research area (1, 2). In equine veterinary medicine, MSC are used experimentally for the treatment of tendon, ligament, and cartilage injuries (3). Currently, no medical treatments are available to reverse cartilage injuries (1). For tendon injuries, the scar tissue formed during the repair is functionally deficient, which has tremendous consequences for the horse in terms of reduced performance and a considerable risk for reinjury (4). For example, Pacini et al. (5) demonstrated in a case–control study that nine out of 11 Italian racehorses treated with MSC derived from bone marrow, successfully returned to their athletic level before injury. Moreover, during a 2-year follow-up period, no reinjury of the superficial digital flexor occurred in the treated group in contrast with the control group horses, which were all reinjured (5).

However, before any type of stem cell can be applied in practice, its unequivocal characterization by a set of specific functional or phenotypic markers is crucial (6). In contrast to the criteria of the International Society for Cellular Therapy (ISCT), which were defined to identify human MSC (7), there is a lack of uniformity to characterize equine MSC in veterinary medicine (8). For human MSC, it has been defined that these cells must be plastic adherent and be capable of differentiating toward the osteogenic, chondrogenic, and adipogenic lineage. Furthermore, they must express CD29, CD44, CD73, CD90, and CD105 and lack expression of CD14, CD34, CD45, CD79α and MHC II. Although there have been several reports on the isolation and in vitro differentiation of equine MSC, few research groups have attempted to identify a set of immunophenotypic markers to characterize these cells (9–13). The lack of a single marker specific for MSC and the currently limited availability of monoclonal antibodies (mAbs) for immunophenotyping equine cells are major factors complicating the progress of this type of research. To our knowledge, commercial mAbs that are directed against equine epitopes are only available for CD44 and MHC II (Serotec, Düsseldorf, Germany). Consequently, for other equine MSC markers, candidate's nonequine mAbs should be evaluated in search for cross-reactivity.

In stem cell research in general and MSC research in specific, immunophenotyping is preferably performed by multicolor flow cytometry to simultaneously demonstrate the coexpression of specific MSC markers and the absence of hematopoietic antigen expression (7, 14). However, many flow cytometric techniques have been developed for analyzing mature, nonadherent leukocytes, and therefore, some refinements are required when using stem cells (15). For example, the use of gating strategies is important not only to select the population of interest based on the selected markers but also to ensure an accurate analysis of the obtained data by excluding aggregates. Furthermore, isotype controls and/or unstained cells are imperative to make a clear distinction between fluorescent positive and negative populations (15).

In this study, 30 commercially available mAbs were first validated for recognizing equine epitopes using equine mononuclear cells (MNC), equine lymphocytes, or equine endothelial cells as appropriate positive control cells. If required, additional experiments on human cells as reference positive control were performed. Confocal microscopy validated the flow cytometric results for all cross-reacting mAbs. Subsequently, equine umbilical cord blood (UCB) MSC of six horses were characterized by the selected panel of nine mAbs, based on their cross-reactivity with equine epitopes as determined in the first part of the study. Cells from the fourth passage were used to perform these immunophenotyping experiments.

Materials and Methods

Isolation of Human and Equine Cells

Equine peripheral blood was obtained from 10 horses between 8 and 12 years old, all females and healthy. Human peripheral blood was obtained from three healthy male humans between 29 and 39 years old with informed consent. Equine MSC were isolated from six mares between 5 and 18 years old that had foaled in the Reproduction Clinic of the Faculty of Veterinary Medicine with informed consent of the owner. The study was approved by the Ethical Committee of the Faculty of Veterinary Medicine of Ghent University (EC2010/147).

After collecting whole blood into a vacuum blood tube, human and equine MNC were isolated using a Percoll® (density 1.080—GE Healthcare, Little Chalfont, UK) gradient, whereas human and equine lymphocytes were isolated using Ficoll-Paque® (density 1.077—GE Healthcare), according to the manufacturer's instructions. Cells were resuspended in Dulbecco's modified Eagle's medium (DMEM) + 1% bovine serum albumin (BSA; Invitrogen, Gent, Belgium). To isolate primary equine vascular endothelial cells from the Arteria carotis of healthy horses, collagenase (Type II; Sigma, Bornem, Belgium) treatment was used as described previously (16, 17). To separate the endothelial cells from smooth muscle cells and fibroblasts, cultures were labeled with 10 μg/mL low density lipoprotein 1,1′-dioctadecyl-3,3,3′,3′-tetramethylindo-carbocyanide perchlorate (Biomedical Technologies, Stoughton, MA) for 4 h at 37°C. After trypsinization, endothelial cells were washed, resuspended in media, and sorted by fluorescence-activated cell sorting (17).

Equine MSC derived from UCB were isolated and cultured as previously described (18). Briefly, MNC were isolated from the UCB using a Percoll (GE Healthcare) gradient and cultured at a concentration of 4 × 106 cells/mL in uncoated T-25 culture flasks. The isolated cells were incubated at 38.5°C in a humidified atmosphere containing 5% CO2. Cells were passaged as soon as confluency exceeded 80% using 0.083% trypsin–ethylenediaminetetraacetic acid (EDTA; Sigma). For this purpose, the adherent MSC were washed with Hank's buffered salt solution without Ca/Mg (Invitrogen) for 5 min, and subsequently incubated with trypsin–EDTA for 5 min at 37°C. Cold culture medium containing fetal bovine serum (FBS) was added to block the action of the trypsin after which the cell suspension was centrifuged for 8 min at 300g. Finally, the cell pellet was resuspended in culture medium, and concentration and cell viability were determined using trypan blue exclusion (19).

After two passages, approximately one million undifferentiated MSC were used to differentiate toward the osteogenic, chondrogenic, and adipogenic lineage and thus confirmed the MSC identity, as previously described (18). Briefly, after 20 days of culture in osteogenic medium, osteogenic differentiation was evaluated using the Alizarine Red S and the Von Kossa histological staining, as well as by detecting alkaline phosphatase activity (Millipore®, Overijse, Belgium; Fig. 1A–1C). Chondrogenic differentiation was evaluated by the Alcian blue histological staining after 3 weeks of culture in chondrogenic medium using a micromass culture system (Fig. 2A). Finally, the adipogenic differentiation was assessed using Oil Red O histological staining after four cycles of 72-h culturing in the adipogenic induction medium and 24-h of culturing in the adipogenic maintenance medium, followed by five consecutive days of culturing in adipogenic maintenance medium (Fig. 3A). For the three lineages, noninduced cells in expansion medium were used as negative controls (Figs. 1D–1F, 2B, and 3B).

Figure 1.

The osteogenic differentiation potential of the equine UCB-derived MSC was confirmed using the Alizarine Red S (A) and the Von Kossa histological staining (B), as well as by detecting alkaline phosphatase activity (C). Noninduced cells in expansion medium were used as negative controls (DF). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 2.

The chondrogenic differentiation potential of the equine UCB-derived MSC was evaluated by the Alcian blue histological staining (A). Putative MSC that were considered as not been differentiated into chondrocytes are shown as negative control (B). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Figure 3.

The adipogenic differentiation potential of the equine UCB-derived MSC was confirmed using the Oil Red O histological staining (A). Noninduced cells in expansion medium were used as negative controls (B). [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Monoclonal Antibodies

The mAbs used in this study to test the cross-reactivity were directed against CD29, CD34, CD44, CD45, CD73, CD79α, CD90, CD105, MHC-II, and a monocyte marker. A full list of all clones which were tested, the species and the companies can be found in Table 1. Secondary antibodies (Abs) included R-Phycoerythrin (RPE)-conjugated sheep antimouse IgG (Sigma) and Alexa 647-conjugated goat antimouse IgG (Invitrogen). The isotype controls in this study included rat IgG2b, mouse IgG2a, mouse IgG1 (all from BioLegend, Uithoorn, The Netherlands), and mouse IgM (Becton Dickinson, Erembodegem, Belgium; Table 2).

Table 1. Overview of the primary mAbs used in this study and their cross-reactivity
HostImmunoepitopeAntibodiesCloneCompanyConcentration (μg/mL)Dilutions testedEquine positive control cellsCross-reactivity
  1. Each mAb was tested using the appropriate equine positive control cells. All data were compensated and corrected for autofluorescence as well as for nonspecific binding. Mo, mouse; Ho, horse; Hu, human; L, lymphocytes; MNC, mononuclear cells; EC, endothelial cells.

MoHuCD29-Alexa 488TS2/16Biolegend5001:16.6, 1:25, 1:50, 1:100, 1:200MNC+
MoHuCD29-PE4B4Beckman CoulterAscites1:25, 1:50, 1:100MNC+
MoHuCD34-Alexa 6474H11BiolegendAscites1:10, 1:20EC
MoHuCD34-PE581Beckman CoulterAscites1:3, 1:6, 1:10EC
RatMoCD34-FITCMEC14.7SerotecAscites1:10EC
MoHuCD34-RPEAC136MiltenyiAscites1:5, 1:10EC
MoHuCD34-PE8G12Becton Dickinson251:2, 1:5EC
MoHoCD44CVS18SerotecAscites1:10MNC+
RatMoCD44-APCIM7Becton Dickinson2001:10, 1:20, 1:40, 1:80MNC+
MoHuCD45-PE5B1MiltenyiAscites1:5, 1:10MNC
MoHuCD45-FITC35-Z6Santa Cruz2001:3.3, 1:5, 1:10, 1:20, 1:40MNC
MoHuCD45B-A11AbcamAscites1:33, 1:50, 1:100, 1:200, 1:400MNC
MoHuCD45-Alexa488F10-89-4SerotecAscites1:2, 1:2.5, 1:5MNC+
RatMoCD45-FITC30-F11Becton DickinsonAscites1:25MNC
MoHuCD45-FITCHI30Becton DickinsonAscites1:2.5, 1:5MNC
MoHuCD45-APC-H72D1Becton DickinsonAscites1:10, 1:20MNC
MoHuCD734G4Hycult1001:6.25, 1:12.5, 1:25, 1:50L
RatMoCD73-PE496406R&D251:0, 1:2, 1:4L
MoHuCD73-PEAD2BiolegendAscites1:5, 1:12.5, 1:25L
MoHuCD7310f1AbcamAscites1:5, 1:10, 1:50, 1:100, 1:200L+
MoHuCD737G2Abcam5001:25L
MoHuCD79α-Alexa647HM57SerotecAscites1:2.5, 1:5MNC+
MoDogCD90DH24AVMRDAscites1:33.3, 1:66.6, 1:100, 1:133.3, 1:266.6MNC+
MoHuCD105-PESN6SerotecAscites1:25, 1:50EC+
MoHuCD105-Alexa 48843A3BiolegendAscites1:10, 1:20EC
RatMoCD105-Alexa488MJ7/18Biolegend5001:11.1, 1:16.6, 1:25, 1:33.3, 1:50, 1:100, 1:200EC
MoHuCD105266Becton Dickinson5001:16.6, 1:25, 1:50, 1:100, 1:200EC
MoHuCD10535/CD105Becton Dickinson2501:8.3, 1:12.5, 1:25, 1:50, 1:100EC
MoHoMHC IICVS20SerotecAscites1:50, 1:100MNC+
MoHuMonocytesmarker- Alexa488MAC387SerotecAscites1:2.5, 1:5MNC+
Table 2. Overview of the marker panels of primary mAbs and 7-AAD selected in this study to immunophenotype viable equine UCB-derived MSC using multicolor flow cytometry
 SubsetMarkerCloneSecondary AbDilution
  1. In addition, the relevant isotype controls as well as the secondary antibodies for the indirectly labeled markers are also provided with their corresponding fluorochrome.

Multicolor FCM1CD29-Alexa488TS2/16 1:20
MHC IICVS20Antimouse RPE1:50
7-AAD   
CD44-APCIM7 1:20
2CD105-RPESN6 1:10
7-AAD   
CD90DH24AAntimouse Alexa6471:100
3CD45-Alexa488F10-89-4 1:5
CD7310f1Antimouse RPE1:5
7-AAD   
4Monocyte-Alexa488MAC387 1:2.5
CD79α-Alexa647HM57 1:2.5
Secondary Ab1, 3Sheep antimouse IgG-RPE  1:20
2Goat antimouse IgG-Alexa647  1:200
Isotype controls1, 4Mouse IgG1-Alexa488  1:20
1–3Mouse IgG1-RPE  1:10
1Rat IgG2b-APC  1:20
2Mouse IgM Antimouse Alexa6471:50
3Mouse IgG2a-Alexa488  1:20
4Mouse IgG1-Alexa647  1:100

Single-Color Flow Cytometry

To screen for cross-reactivity, approximately 2 × 105 cells per tube were centrifuged in DMEM + 1% BSA and incubated for 15 min at 4°C in the dark with each of the primary mAbs (Table 1). After two washing steps, cells that were incubated with nonlabeled primary mAbs were incubated with the RPE-conjugated sheep antimouse IgG secondary Ab for 15 min at 4°C in the dark. After three washing steps, cell pellets were finally resuspended in 400 μL phosphate-buffered saline (PBS) and analyzed after 10-min incubation with 7-aminoactinomycin D (7-AAD), a viability dye, which is excluded by viable cells but can penetrate cell membranes of dying or dead cells. For intracellular antigen detection, that is, when using the mAb directed against CD79α or the monocyte marker, cells were fixed and permeabilized first using FIX & PERM® (Caltag, Invitrogen) according to the manufacturer's instructions. In addition, an incubation step for 15 min at room temperature (RT) in the dark with 10% horse serum was included to block nonspecific binding of these mAbs to equine epitopes. No 7-AAD staining was performed on these fixed and permeabilized cells.

For all tubes, at least 10,000 cells were analyzed using a FACSCanto flow cytometer (Becton Dickinson Immunocytometry Systems, Erembodegem, Belgium) equipped with two lasers, a 488 nm solid state laser and a 633 nm HeNe laser, and FACSDiva software. All data were corrected for autofluorescence as well as for unspecific bindings using either secondary Ab or isotype-matched negative controls. All isotypes were matched to the immunoglobulin (Ig) subtype, conjugated to the same fluorochrome and used at the same fluorescence/protein concentration as the corresponding Ab.

Confocal Immunofluorescence Microscopy

The specificity of the cross-reactive mAbs was confirmed by confocal fluorescence microscopy. Briefly, the staining procedure was performed as described earlier for the single-color flow cytometry, with the exception of the use of the RPE-conjugated rabbit antimouse Ig as secondary Ab, which was replaced by fluorescein isothiocyanate (FITC)-conjugated rabbit antimouse Ig (Dako, Glostrup, Denmark). Subsequently, cells were fixed using Celfix® (Becton Dickinson) and incubated at 4°C for 20 min in the dark. After centrifugation and resuspension in 300 μL PBS, propidium iodide (PI; 10 μg/mL) was added to visualize the cell nuclei and incubated for 20 min at 4°C in the dark. Following centrifugation and a washing step, the cells were resuspended in 100 μL PBS and after cytocentrifugation (Shandon, Southern Products, Runcorn, UK), the stained cells were screened using confocal fluorescence microscopy (Nikon BeLux, Brussels, Belgium).

Multicolor Flow Cytometry

For the multicolor flow cytometry, undifferentiated equine MSC from the fourth passage were incubated with following combinations of marker panels: CD29-Alexa488/MHC II-RPE/CD44-APC/7-AAD (Subset 1), CD105-RPE/CD90-Alexa647/7-AAD (Subset 2), CD45-Alexa488/CD73-RPE/7-AAD (Subset 3), and the monocyte marker-Alexa488/CD79α-Alexa647 (Subset 4). To identify the percentage of viable cells, 7-AAD was used in Subsets 1–3 but not in Subset 4, as cells in the latter subset were permeabilized. A detailed description of mAb clones and dilutions used can be found in Table 2.

In the Subsets 1–3 for the cell surface markers, approximately 2 × 105 cells per tube were centrifuged to pellet in DMEM + 1% BSA and incubated for 15 min at 4°C in the dark with following nonlabeled primary mAbs: MHC II (Subset 1), CD90 (Subset 2), and CD73 (Subset 3), respectively. After two washing steps, cells that were incubated with these nonlabeled primary mAbs were incubated with a secondary Ab conjugated with a relevant fluorochrome for 15 min at 4°C in the dark (Table 2). Cell pellets were washed twice to remove the excess secondary Ab and subsequently treated with a 15-min blocking step using 10% mouse serum to exclude nonspecific binding of the directly labeled primary mAbs on the secondary Ab. Next, these directly labeled primary mAbs, that is, CD29 and CD44 (Subset 1), CD105 (Subset 2), and CD45 (Subset 3) were incubated for 15 min at 4°C in the dark. After three washing steps, cell pellets were finally resuspended in 400 μL PBS and analyzed after 10 min incubation with 7-AAD for all three subsets of markers. For the intracellular antigen detection in Subset 4, cells were first fixed and permeabilized using FIX & PERM® (Caltag, Invitrogen) according to the manufacturer's instructions. Subsequently, cells were preincubated with 10% horse serum for 15 min in the dark at RT as a blocking step, after which the CD79α and the monocyte marker primary mAbs were incubated for 15 min at 4°C in the dark. Finally, the pellet was resuspended in 400 μL PBS after three washing steps.

For all tubes, at least 10,000 cells were analyzed using a FACSCanto flow cytometer (Becton Dickinson Immunocytometry Systems) equipped with two lasers, a 488 nm solid state laser and a 633 nm HeNe laser, and FACSDiva software. All data were compensated and corrected for autofluorescence as well as for unspecific bindings using both secondary Ab and isotype negative controls. Compensation for spectral overlap between fluorochromes was performed using an automatic calibration technique (FACSDiva software, Becton Dickinson) and subsequently evaluated individually with a matrix.

Gating Strategy

A primary gate was placed on the area versus width signal of the forward scatter (FSC-A/FSC-W) dot plot, after which this population was visualized on the area versus width signal of the side scatter (SSC-A/SSC-W) dot plot to discriminate for doublets and clumps. The single cell population was identified by defining the gated population on a side scatter area signal versus a forward scatter area (SSC-A/FSC-A) signal dot plot. The final gate for analysis was a Boolean gate on the single cell population and the 7-AADneg cells, enabling the analysis of a viable single cell population.

Statistical Analysis

All data were analyzed using the dedicated FACSDiva software (Becton Dickinson) and subsequently exported to Excel (Excel 2007; Microsoft, Redmond, WA) to calculate different parameters such as mean, median, standard error of the mean, and interquartile range (IQR).

Results

Assessment of Ab Cross-Reactivity with Equine Epitopes

Freshly isolated equine MNC, lymphocytes, and primary endothelial cells were used to validate the cross-reactivity of human, murine, and canine mAbs listed in Table 1. Equine MNC were used to demonstrate cross-reactivity of the mAbs directed against CD29, CD44, CD45, CD79α, CD90, MHC-II, and a monocyte marker, whereas equine lymphocytes were used to detect cross-reactivity for CD73. An increase in mean fluorescence intensity (MFI), as compared with the negative controls, indicated positivity and was detected for all screened clones directed against CD29, CD44, CD79α, CD90, MHC-II, and the monocyte marker (Table 1, Fig. 4, and Table S1, Supporting Information). Confocal fluorescence microscopy was successfully used to validate the flow cytometric data (Fig. 4).

Figure 4.

Flow cytometric and confocal fluorescence microscopic analyses of the cross-reacting mAbs based on the appropriate equine positive control cells. Fluorescence channel histograms represent relative numbers of cells versus their MFI. The light gray and black histograms, which systematically overlap, represent the negative controls (i.e., autofluorescence and relevant isotype control, respectively). The dark gray histograms represent the test samples incubated with each of the selected mAbs. Mean MFI ± SEM values were described for each histogram. Isolated equine MNC were positive for CD29, CD44, CD45, CD79α, CD90, MHC-II, and the monocyte marker, isolated equine lymphocytes were positive for CD73, and isolated equine endothelial cells were positive for CD105. To confirm the cellular binding of each cross-reacting mAbs on the appropriate positive control cells, confocal fluorescence microscopy was used. Nuclei are visualized using PI. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

For the CD45 and CD73 markers, only one out of the five antihuman clones tested for each marker recognized the equine epitopes on MNC and lymphocytes, respectively (Fig. 4). However, the percentage of positive equine cells was rather low with an average 24.6% of the equine MNC being positive for CD45 and 16.8% of the lymphocytes being positive for CD73 (Fig. 4). A similar percentage was obtained when excluding the nonviable cells based on their 7-AAD positivity. Moreover, nonspecific binding was minimalized by including appropriate blocking steps and negative controls such as relevant isotype controls for both mAbs. Still, to further evaluate the specificity of this positive signal, the anti-CD45 and anti-CD73 stained equine cells were visualized by confocal fluorescence microscopy. In addition, human MNC and lymphocytes were also used to analyze these two antihuman mAbs by flow cytometry as well as confocal fluorescence microscopy. Hereby, it was found that 65.3% of the human MNC expressed CD45 and 18.5% of the human lymphocytes stained positive for CD73 (Fig. 5 and Fig. S1, Supporting Information). As these percentages of positive cells are in the same range as those for their equine counterparts, it was concluded that these two clones cross-react with equine epitopes, although further research like Western blot or immunoprecipitation analyses might be required to unambiguously confirm cross-reactivity.

Figure 5.

Flow cytometric and confocal fluorescence microscopic analyses of the cross-reacting mouse antihuman CD45-Alexa488 mAb on isolated human MNC and of the mouse antihuman CD73 mAb on isolated human lymphocytes. Fluorescence channel histograms showing the expression of antihuman CD45 on isolated human MNC (A) and antihuman CD73 on isolated human lymphocytes (B), respectively. Histograms represent relative numbers of cells versus their MFI. The light gray and black histograms, which systematically overlap, represent the negative controls (i.e., autofluorescence and isotype control, respectively). The dark gray histogram represents the test sample incubated with the mAb. To confirm the cell surface binding of the CD45 mAb on the human MNC (C) and of the CD73 mAb on the human lymphocytes (D), confocal fluorescence microscopy was used. Nuclei are visualized using PI. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.]

Finally, to assess cross-reactivity of the CD34 and CD105 mAbs, pure populations of equine primary endothelial cells were used. One clone of the five CD105 mAbs tested identified the equine epitope (Table 1 and Fig. 4), whereas none of the five CD34 mAbs tested showed cross-reactivity (Table 1).

In conclusion, only 11 out of the 30 mAbs evaluated in the first part of this study recognized the respective equine epitopes and as such, are useful to characterize equine MSC.

Immunophenotyping of Equine UCB-Derived MSC by Multicolor Flow Cytometry

The cross-reacting mAbs identified in the first part of this study were used to immunophenotype equine MSC isolated from the UCB of six horses using multicolor flow cytometry, as outlined in Table 2. A representative example of the gating strategy and the multicolor analysis was shown in Figure 6. On average 92.8% (IQR 90.6–94.7) of the undifferentiated MSC of the fourth passage simultaneously expressed CD29 and CD44 and lacked expression of MHC II (Table 3, Subset 1). Also, on average 94.9% (IQR 89.6–99.6) of MSC were positive for CD90 but had a low and variable expression of CD105 varying between 0.1% and 20.0% (Table 3, Subset 2). The equine MSC lacked expression of CD45 (IQR 98.9–99.5) and displayed a variable expression for CD73 with the proportion of positive MSC ranging between 0.0% and 25.3% (Table 3, Subset 3). Finally, on average 98.1% (IQR 97.2–99.1) of the MSC lacked expression of both CD79α and the monocyte marker (Table 3, Subset 4). Equine MSC from the 10th passage as well as cryopreserved MSC from the fourth passage upon thawing showed a virtually identical phenotype (Tables S2 and S3, Supporting Information).

Figure 6.

Gating strategy to enumerate the equine MSC. After visualizing the population of interest on the FSC-A/FSC-W dot plot (P1), this P1 population was gated on the SSC-A/SSC-W dot plot to discriminate for doublets and clumps (P2). Subsequently, the single cell population was identified by defining P2 on a SSC-A/FSC-A signal dot plot (P3). The final gate for analysis was a Boolean gate on the single cell population and the 7-AADneg cells, enabling the analysis of a viable single cell population. For each subset, this viable single cell population was displayed on the respective fluorescence channel versus SSC-A dot plots. Mean ± SEM values were described for each histogram.

Table 3. Results of the immunophenotypic characterization of equine UCB-derived MSC from the fourth passage, expressed as the percentage (%) of cells either positive or negative for each of the selected nine markers analyzed in four subsets
SubsetMarkerMeanMedianSEMIQRMin.Max.
  1. [n = 6; mean, median, standard error of the mean (SEM), IQR, minimum, and maximum values]. Subset combinations are presented in bold.

 CD29pos98.398.30.20.897.898.8
 MHC IIneg97.398.11.11.59299.7
 7-AADneg97.197.20.71.494.298.9
 CD44pos98.7990.41.597.399.7
1CD29pos + MHC IIneg + 7-AADneg + CD44pos92.8931.24.189.296.6
 CD105pos61.63.49.70.120
 7-AADneg89.988.52.87.780.399.5
 CD90pos94.999.22.910.18599.7
2CD105pos + 7-AADneg + CD90pos4.40.5350.218.5
 CD45neg99.199.30.20.698.299.7
 CD73pos4.60.44.11.1025.3
 7-AADneg93.395.92.84.580.399.3
3CD45neg + CD73pos + 7-AADneg3.60.13.50.6020.9
 Monocyteneg99.299.40.20.898.499.7
 CD79αneg98.799.40.51.996.999.7
4Monocyteneg + CD79αneg98.198.60.61.995.999.3

Discussion

The unequivocal immunophenotyping of equine cells in general and MSC in particular is strongly hampered by the limited availability of mAbs directed against equine epitopes or against epitopes of other species that show cross-reactivity with the horse. In a study of Ibrahim et al. (20), only 14 out of the 379 tested antihuman mAbs, that is, less than 5%, recognized the corresponding epitopes on isolated equine leukocytes. This illustrates the urgent need for additional studies evaluating potential cross-reactivity of xenogenic mAbs against specific molecules present on equine cells. In this study, 30 mAb clones directed against epitopes used as markers to immunophenotype human MSC were evaluated at first (7). Eleven clones showed cross-reactivity with equine epitopes using MNC, lymphocytes, or endothelial cells as positive control cells. Based on the cross-reacting mAb clones identified in the first part of this study, a multicolor marker panel based flow cytometric protocol to immunophenotype equine MSC was subsequently developed. Multicolor detection is an attractive strategy for the identification of MSC, as different antigens on a single cell can be simultaneously detected (21), and the overlapping patterns of the phenotypic markers allow the discrimination of MSC from other cells (6). Although gene expression is also often assessed to evaluate the presence or absence of selected markers and is a complementary parameter to antigen measurement, mRNA is not alwaystranslated. Therefore, it is preferable to examine the presence or absence of the corresponding proteins rather than their mRNA levels (10).

The rationale for using the macrophage/monocyte antihuman mAb in this study, instead of the frequently used CD11b or CD14 clones, was based on the fact that its cross-reactivity with equine epitopes had been previously reported (Serotec, Oxford, UK). This mAb recognizes the intracytoplasmic calprotectin molecule L1, which has a restricted distribution within the monocyte-derived cell lineage (22). For CD34, we were not able to confirm the cross-reactivity of any of the five tested clones on equine endothelial cells, although two of these clones were recently used in other studies to characterize equine MSC (9, 11). These apparently contradictory data emphasize again the importance of using proper positive and negative control cells when evaluating cross-reactivity of mAbs.

As revealed in this study, viable equine MSC simultaneously expressed CD29 and CD44, and lacked expression of MHC II. They also simultaneously lacked expression of CD79α and the monocyte marker. Because of the more variable expression of CD73 and CD105 among the six mares, less straightforward results were obtained for the combined expression of both these markers. Regardless, equine MSC clearly express CD90 and lack expression of CD45.

According to the ISCT criteria, the expression of CD105 on human MSC must exceed 95% (7). In this study, however, a low and variable expression was noted for UCB-derived equine MSC. Still, these results are valuable, as it has been described in independent studies that human MSC isolated from UCB show a lower expression of CD105 (21, 23, 24). Moreover, canine MSC isolated from adipose tissue were recently reported to even be CD105neg (25). These are all indications for a variable CD105 expression on MSC originating from either different sources or different species. For equine MSC, only Braun et al. (13) have investigated CD105 using the same clone as the one used in this study, for which they reported a strong positive signal. However, equine MSC were derived from adipose tissue instead of UCB in the latter study, which might explain the marked difference in observed CD105 expression. On the other hand, the detaching agent used could also provide an explanation for the apparent discrepancy. Indeed, in the study of Braun et al. (13), accutase was used, whereas trypsin/EDTA was used in our study. Trypsin is a pancreatic serine protease, whereas accutase exhibits protease and collagenolytic activities (15). Interestingly, it has been reported that trypsin can cause removal and/or functional impairment of certain cell-surface membrane proteins (26, 27). In line with these findings, Hackett et al. (28) recently demonstrated that detaching equine cells with trypsin damaged certain cell surface proteins like CD14, while other markers such as CD90 appeared unaffected. Further research remains relevant to identify which equine epitopes are trypsin labile.

Similar to CD105, the expression of CD73 on human MSC must exceed 95% according to the ISCT (7). Nevertheless, a moderate expression has been reported for human MSC, which were derived from bone marrow and cultured in medium containing FBS (29). In line with the data reported for CD105 on canine MSC, these cells also appear to be CD73neg (25). As the expression of CD73 has not been reported yet for equine MSC, it is difficult to compare our results with other studies or to provide an explanation for the variable CD73 expression observed in this study.

In conclusion, this is the first report that describes a protocol to immunophenotype equine MSC isolated from UCB using multicolor flow cytometry. Hereby, the salient findings were that the equine MSC were CD29pos, MHC-IIneg, CD44pos, CD45neg, CD90pos, CD79αneg, and monocyte markerneg. The intriguing variability in expression of CD73 and CD105 on equine MSC, which is not in accordance with human MSC, warrants further research including potentially critical factors such as the influence of the sources of equine MSC and the sample pretreatment. Furthermore, the application of this proposed multicolor protocol as an isolation tool to sort putative equine MSC from a mixed cell population warrants further study, as this would provide a major added value to this exciting research field.

Acknowledgements

The authors thank Jella Wauters and Dries Vercauteren for the technical assistance with (confocal) fluorescence microscopy.

Ancillary