Comparison of Flowcytometric and Immunocytochemistry Analysis of Stem Cell Surface Markers

Departement of Histology, Faculty of Medicine, Universitas Indonesia, Jl. Salemba 6, Jakarta, Indonesia Stem Cell Medical Technology Integrated Service Unit, Cipto Mangunkusumo Central Hospital-Faculty of Medicine Universitas Indonesia, CMU 2 Building, 5th floor, Jl. Diponegoro 71, Jakarta, Indonesia Department of Clinical Pathology, Faculty of Medicine, Universitas Indonesia, Jl. Salemba 6, Jakarta, Indonesia Biomedical Science Master Program, Faculty of Medicine, Universitas Indonesia, Jl. Salemba 6, Jakarta, Indonesia Department of Anatomy, Faculty of Medicine, Universitas Indonesia, Jl. Salemba 6, Jakarta, Indonesia


Introduction
Stem cells are very promising for regenerative medicine, but before the cells can be used to treat patients, they need to be characterized by analyzing their surface markers. Flowcytometry analysis is widely used in characterization of stem cells. Flowcytometry analysis can be done using various numbers of cells. In many instances, 10 5 (Oedayrajsingh-Varma et al., 2006;Choudhery et al., 2014;Li et al., 2013), or 5×10 5 (Mitchel et al., 2006), cells were incubated with a fluorophore-conjugated antibody against a certain surface marker. In our study, we used 10,000 cells and 10,000 gated events to simultaneously analyze the percentage of CD90, CD73 and CD34 in Adipose Tissue-derived Mesenchymal Stem Cells (AT-MSCs) (Pawitan et al., 2013a).
In some tissue culture labs, flowcytometer is not available. Therefore, in areas where a flowcytometry device is not available, characterization of stem cells can be done using immunocytochemistry, which requires less sophisticated equipments. We have developed a simple spot method that was suitable for immunocytochemistry staining (Pawitan et al., 2010), which can be used in stem cells surface marker enumeration. For surface marker enumeration, at least 100 cells should be analyzed and the percentage of positive cells can be calculated. Therefore, to make a simple spot specimen, theoretically less cells are required, though when available we usually use 20,000 cells to make several spots, to be stained by antibodies against various surface markers.
However, there were no data concerning the agreement of stem cell surface marker analysis between flowcytometry and immunocytochemistry staining. Therefore, in this study we compared the results of flowcytometry against immunocytochemistry for the measurement of CD34, CD73 and CD90 in stem cells.

Materials and Methods
This is an experimental descriptive study, which was approved by the Ethical Committee of the Faculty of Medicine Universitas Indonesia (no. 136/PT 02.FK/ETIK/2012) and was done from September 2012 through May 2013.
Isolation of adipose derived stem cells and cell cultures were conducted in the Integrated Laboratory of the Faculty of Medicine Universitas Indonesia as previously described (Pawitan et al., 2013b;Suryani et al., 2013). Spot specimens was made in the Department of Histology, as previously described (Pawitan et al., 2010) and stained by immunocytochemistry, while flowcytometric analysis was done in the Department of Clinical Pathology, Faculty of Medicine, Universitas Indonesia.

Flowcytometric Analysis
Around 10,000 adipose tissue derived stem cells were labelled by three kinds of antibodies i.e., per CP labelled antibody against CD34 (8G12 [BD Biosciences cat. Number BD 340430]), PE labelled antibody against CD73 (BD Biosciences cat. number BD 550257) and APC labelled antibody against CD90 (BD Biosciences cat. Number BD 559869). Another 10,000 cells were labeled by the respective isotypes, i.e., perCP, PE and APC labeled isotypes (BD Biosciences cat. number BD 559425, BD 555749 and BD 555751, respectively). Further the three surface markers were simultaneously analyzed using a BD FACSCalibur flowcytometer (BD Biosciences) with 10,000 gated events.

Immunocytochemistry Staining
The same sample was made into spot specimens and stained for CD34, CD73 and CD90 (Thy-1) (Santa Cruz, Table 1) by immunocytochemistry (SCB, 2014a;2014b;2014c), according to manufacturer instruction. When available, 100 cells were screened for positive results and percentages of positive results for CD34, CD73 and CD 90 were computed.

Data Collection and Analysis
The results of the two methods were noted and tabulated and discrepancies between the two methods were noted for all samples. Discrepancies lower than 3, 11 and 11% for CD34, CD73 and CD90 respectively were regarded as agreement between the two methods. Percentages of agreement in CD34, CD73 and CD90 were calculated.

Results
Some spot specimens did not yield 100 cells and analyses were done on available cells. Percentage of CD positive cells for CD34, CD73 and CD90 in flowcytometry and immunocytochemistry can be seen in Table 2.

Discussion
Agreement between flowcytometry and immunocytochemistry in this study was low to moderate. This fact might be due to the primary antibody choices for immunohistochemistry, which epitopes were not the same as those from flowcytometry.
Adipose derived MSCs have rather high CD34 level that decrease upon passages (Mitchel et al., 2006) and in our study this trend was better represented by immunocytochemistry result (Table 2). In this study, immunocytochemistry staining of CD34 used mouse monoclonal antibody raised against KG-1 cells (class III CD34 epitope) (Table 1), while flowcytometry used perCP labelled antibody against CD34 (8G12 [BD Biosciences cat. Number BD 340430]), which was derived from 8G12 clone that recognizes two distinct epitopes that are expressed by human KG-1a cell line (BD Biosciences, 2014). CD34 has at least three epitopes, i.e., class I, II and III epitopes. A study showed that different CD34 epitopes gave different results in the detection of CD34 bearing bone marrow and peripheral blood stem cells (Croockewit et al., 1998).    , P5 = 5% PRP containing DMEM-HG, VE = final 10 ng mL −1 VEGF (Invitrogen PHC9394) and 10% human AB serum (Gibco 34005-100) containing DMEM-HG Mitchel et al. (2006) in a flowcytometry study on AT-MSCs that was cultured in 10% Fetal Bovine Serum (FBS) containing Dulbecco's modified Eagle medium/Ham's F12 (DMEM/Ham's F12) used CD34 antibody that was derived from 8G12 clone, as was our CD34 flowcytometry antibody. However, we used different fluorophore, i.e., per CP, while Mitchel et al. (2006) used PE. Mitchel et al. (2006 showed that in early passages (P0-P2), CD34 percentage was high, with high variability especially in passage-2, where the standard deviation was greater than the mean value. CD34 tended to decrease upon passages and the percentage in passage-4 was 1.7+1.0 (Mitchel et al., 2006). Our flowcytometry study showed variable results for CD34, that was in line with Mitchel's results (Mitchel et al., 2006).

-----------------------------------------------
In this study, agreement between flowcytometry and immunocytochemistry for CD73 was moderate, while immunocytochemistry results showed better agrrement to AT-MSCs in term of CD73 level, except for one case in passage-4 that showed CD73 level of 39.1% (Table  2). Mitchel et al. (2006) used the same CD73 antibody as that was used in this study to analyze CD73 expression in AT-MSCs. Their results were in accordance with our results for AT-MSCs that were cultured in MesenCult® and 5 and 10% PRP containing DMEM-HG (Table 1). However, CD73 percentages in adipose derived stem cells that were cultured in VEGF and 10% human AB serum containing DMEM-HG showed much lower percentage compared to those in Mitchel et al. (2006) study. Our study showed that adipose derived stem cells that were cultured in VEGF and 10% human AB serum containing DMEM-HG were not MSCs, but had differentiated into chondrogenic cells, which was shown by Alcian blue staining (Pawitan et al., 2013c).
Immunocytochemistry staining of CD73 used mouse monoclonal antibody (IgG3) raised against human CD73 purified from placenta (Table 1), while flowcytometry used PE labelled antibody against CD73 from AD2 clone (BD Biosciences cat. number BD 550257), which was purified from tissue culture supernatant or ascites (BD Pharmingen, 2014a). The epitope that was recognized by immunocytochemistry was FL(h) (SCB, 2014d), while by flowcytometry was whole CD73, which is ecto-5'nucleotidase, a 70 kDa, Glycosyl Phosphatidylinositol (GPI)-anchored glycoprotein (BD Pharmingen, 2014a). Table 2 showed that for CD90, all flowcytometry results were higher compared to immunocytochemistry, thus were more appropriate for AT-MSC. Immunocytochemistry staining of CD90 used goat polyclonal antibody raised against a peptide mapping near the C-terminus of Thy-1 of human origin (Table 1), while flowcytometry used monoclonal antibody against whole human CD90 from 5E10 clone that was purified from tissue culture supernatant or ascites (BD Pharmingen, 2014b). Mitchel et al. (2006) used FITC labelled CD90 antibody from 5E10 clone as our antibody and showed that CD90 expression in AT-MSCs was high and became higher upon passages. This finding was corroborated by our results in AT-MSCs that were cultured in MesenCult®, which is a special commercial medium for MSCs and 10% PRP containing DMEM-HG.

Conclusion
Agreement between flowcytometry and immunocytochemistry was low for CD 34 and CD90 and was moderate for CD73. The difference might be due to different epitopes that were used to generate the antibodies.