- Top of page
- Materials and Methods
- Literature Cited
Myelodysplastic syndrome (MDS) is a kind of clonal stem-cell disorder in which aberration within a hematopoietic stem cell (HSC) gives rise to the entire disease as in acute myeloid leukemia (AML). Studies have showed that contrasting normal stem cells, AML stem cells express CD96 and CD123, but lack of CD90, although both of them reside within the CD34+CD38− population. So far, little is known about expression of the markers on MDS HSC. In this study, we analyzed the immunophenotypic characteristics of CD34+CD38− bone marrow (BM) cells by multicolor flow cytometry in 38 patients with MDS and 10 control patients. We found that CD34+CD38− BM cells coexpressed CD13, CD33, CD117, CD133, and HLA-DR almost in all patients, but in MDS they expressed higher amounts of CD13 (79% ± 16% vs. 36% ± 13%, P < 0.05) and CD133 (66% ± 20% vs. 25% ± 13%, P < 0.05). CD90 was expressed in all control patients but just in 63% of patients with MDS. No control patients had an expression of CD2, CD5, CD7, CD44, CD96, and CD123, which expressed variable amounts in 17–53% of patients with MDS. The level of CD13 in RCMD (89% ± 7%), RAEB-1 (88% ± 11%), and RAEB-2 (81% ± 13%) were obviously higher than that of RA (63% ± 16%, P < 0.05). CD2, CD5, and CD7 were more frequently observed in RAEB or INT and HIGH-R cases. Taken together, we demonstrate MDS stem cells display deranged phenotypic abnormalities that may make them particularly difficult to eradicate using therapies targeted against surface antigens, and the percentage of cells expressing CD13 is notably higher in patients with high-grade MDS that may be a potential prognostic indicator of MDS in the future. © 2010 International Society for Advancement of Cytometry
Cancer stem cells (CSCs), defined as a small subset of cells within a cancer that have the exclusive ability to self-renew and to differentiate into the heterogeneous lineages of cancer cells that comprise the tumor (1), have attracted considerable attention in cancer research in recent years. So far, CSCs have already been confirmed in many kinds of cancers such as leukemia, multiple myeloma, and solid organ malignancies including brain, breast, prostate, colon, and pancreatic cancer (2). These rare cells may play an important role in tumor formation, metastasis, therapeutic resistance, and recurrence (2). So, it is crucial to functionally identify molecular targets in CSCs and to define “stem cell expression profiles” for various neoplasms, and this may provide insights into the development of predictive and prognostic markers and specific therapeutic interventions.
Myelodysplastic syndromes (MDSs) are a heterogeneous group of clonal malignant hematopoietic disorders, which is characterized by ineffective hematopoiesis and frequent progression to acute myeloid leukemia (AML) (3). Recent studies suggest that MDS is a stem-cell disorder in which aberration within a hematopoietic stem cell (HSCs) gives rise to the entire disease just as in AML (4, 5). Previous data have confirmed that similar to normal HSCs, human AML stem cells reside within the CD34+CD38− compartment of the leukaemic clone (6). However, contrasting normal stem cells, AML stem cells express substantial amounts of CD96 (7) and CD123 (8), which are thought to be leukemic stem cell-specific markers, but lack the expressions of CD90 (9). Other antigens that have been detected on the surface of AML stem cells include CD13, CD33, CD44, CD133, and CD117 (10–13). However, whether these markers including CD123 are specific for AML stem cells remains a matter of controversy (11–16). As for MDS stem cells, so far, only a few data about their characteristics have become available. It is now commonly viewed that the MDS stem cells have a similar phenotype of CD34+CD38− CD90+ as normal HSCs (17), whereas abnormal expression of CD7, CD13, CD33, HLA-DR, and CD117 have also been found on dysplastic CD34+ myeloid precursors (18–20) but they are not certain.
In this study, we analyzed the immunophenotypic characteristics of CD34+CD38− bone marrow (BM) cells from patients with MDS compared to normal/reactive BM. Our aim was to define the profile of immunoreactive cell surface target antigens expressed on CSCs in patients with MDS, which could contribute to distinguish MDS stem cells from normal/reactive HSCs, and to find out whether there was a relationship between the antigen level and the risk of MDS. Overall, our results showed that most patients with MDS displayed complex and aberrant phenotypic abnormalities in the immature CD34+CD38− BM cells.
- Top of page
- Materials and Methods
- Literature Cited
Previous studies have indicated that both human normal and AML stem cells reside within the CD34+CD38− subpopulation, and these cells are capable of initiating long-term growth of normal myelopoiesis (normal stem cells) or leukemias (leukemic stem cells) in NOD/SCID mice (6, 23, 24). We infer that MDS stem cells may reside within the CD34+CD38− population, although we have not proven this yet by transplantation experiments. In this study, we have analyzed expression of various molecular antigens on the surface of CD34+CD38− BM cells in patients with MDS in an effort to define the phenotype of CSCs in patients with MDS and find potential therapeutic targets.
CD34 is a marker of HSC. In AML, the positivity of leukemic cells for CD34 is obviously different in various FAB subtypes (25, 26), and in MDS CD34+ cell clusters were found in 23% of patients and were associated with WHO categories with excess of blasts and poor-risk cytogenetics (27, 28). Consistent with these reports, our study showed that both percentages of CD34+ and CD34+CD38− cells on MNCs by definition increased with more advanced stage of MDS (and greater than in normal); and they were related to subtype, risk and WBC counts in MDS.
CD90, which is expressed on 5–25% of normal CD34+ cells but lack on AML stem cells (9), has now been found to be expressed on CD34+CD38− MDS stem cells (17). In line with these studies, our results showed that CD90 was expressed on normal stem cells in all control patients but only in 63% of patients with MDS. Furthermore, the expression amounts between two groups was significant different (P < 0.05). More recently, several leukemic stem cell-specific markers have been proved to express on CD34+CD38− cells, such as CD123 (8) and CD96 (7). CD123 has been reported to be expressed on stem cells in AML but not in normal BM (8). However, since important control experiments have not yet been reported, and high CD123 expression on non-AML-regenerating BM CD34+CD38− cells in five such cases have been found, some researchers think this issue needs serious attention (14). In our research, we detected no expression of CD123 on CD34+CD38− cells in any control patient, but we found CD123 was expressed in 53% of patients with MDS (54% ± 23% in 19 of 36 cases), which confirmed Jordan CT's report (8). CD96 has been demonstrated to be expressed on the majority of CD34+CD38− AML cells in many cases (74.0% ± 25.3% in 19 of 29 cases), whereas weakly expressed on only a few (4.9% ± 1.6%) cells in the normal HSC-enriched population (Lin-CD34+CD38−CD90+) (7). We detected CD96 on a few part of CD34+CD38− MDS stem cells in about 34% of patients (43% ± 19% in 13 of 38 cases), and negative on normal stem cells. According to the expression of CD90, CD96, and CD123 on MDS CD34+CD38− cells, we might speculate it is probable that normal HSC (with a phenotype of CD90+CD96−CD123−) and leukemic stem cells (CD90−CD96+CD123+) coexist in MDS stem cell pool.
CD33 is highly expressed on normal CD34+CD38− stem cells (14), whereas Hauswirth et al find CD33 is specially expressed on AML stem cells but not on normal ones (12). Abnormal expression of CD33 on dysplastic CD34+ myeloid precursors is also detected in MDS (18). We found both normal and MDS CD34+CD38− BM cells coexpressed CD33. The majority of studies have shown that the CD34+CD38− AML stem cells coexpress CD13, CD117, and CD133 (11, 29), but usually lack of HLA-DR (16). Nevertheless, abnormal expression of CD13, CD117, CD133, and HLA-DR on dysplastic CD34+ myeloid precursors or BM myeloblasts have been reported in MDS (18, 30, 31). CD34+CD38− subset in MDS had a phenotype of HLA-DR+CD13+CD33+ (32). Likewise, we detected CD13, CD117, CD133, and HLA-DR on CD34+CD38− cells with variable amounts both in MDS and in control patients. It is noteworthy that the CD34+CD38− cells in MDS expressed obviously higher amounts of CD13 and CD133 compared with controls. The difference in the percentage of cells expressing CD13 may be due to increased abnormal myeloid progenitors that ultimately lead to progression of the dysplasia. CD133 is expressed primarily on a small subset of stem cells (CD34+) (33), and increased numbers of CD133 positive cells are also found to be present in the majority of patients with MDS by other groups (34). The level of these two molecules might be a potential measure standard to discriminate normal and MDS stem cells.
CD44 has recently been described as a target on leukemic CD34+CD38− stem cells; however, it is also weakly expressed on normal CD34+CD38− cells, more differentiated hematopoietic cells and many different tissues (12). Our data showed CD44 had minor expression on the CD34+CD38− cell compartment in MDS (9/30 cases, 35% ± 15%) and no expression in normal ones.
So far, the reports about expression of lymphoid markers on CD34+CD38− cell compartment in AML or MDS patients are rare. CD7 expression was found to be altered on immature precursors in almost half (41%) of the patients with MDS (19), and CD7 positivity of enriched blast cells was an independent variable for a poor prognosis in MDS (32). Our findings indicated that not only CD7 but also CD5 and CD2, which were found no expression on normal stem cells, were expressed on CD34+CD38− MDS BM cells in a relatively small percentage of cases (20%, 20%, and 17%, respectively). Besides, CD10 was detected on CD34+CD38− subset in both groups.
As it is known, WHO subtype and risk stratification of MDS are in close relation with prognosis of individuals suspected of MDS and the evolution to AML. So we analyzed the phenotypic abnormalities present in different WHO subtypes or risk groups of CD34+CD38− cells from MDS versus normal/reactive BM. Consistent with Matarraz's study that CD13 expression was found to be decreased in the very early phases of the disease (RA) (19), we found the levels of CD13 on CD34+CD38− cells in RCMD, RAEB-1 and RAEB-2 groups were notably higher than that of RA group (P < 0.05). Matarraz also found altered levels of CD33 and CD117 among CD34+ precursors, consisting of an abnormally low expression of CD33 in low-grade MDS (RA, RCMD, INT-1, and LOW-R cases) and significantly increased amounts in RAEB-2 cases, and increased expression of the CD117 being more frequently observed among patients with high- versus low-grade MDS (19). However, we did not see any markedly differences of these myeloid-associated markers except CD13 among subtypes or risk groups. One reason for this discrepancy may be that our phenotypic data were referred to overall population of BM CD34+CD38−cells, and what they analyzed were CD34+ neutrophil precursors. An alternative explanation would be that the numbers of patients included in every subtype or risk group were too small.
In AML, patients with positive CD7 on blasts have poor prognosis. In MDS, HIGH-R cases more frequently show increased percentages of CD7+ cells on immature precursors (19). Our data displayed that lymphoid markers such as CD2, CD5, and CD7 were more frequently observed in RAEB subtypes or INT and HIGH-R cases, although the level of these markers was rather low. We may infer that the lymphoid markers presented on MDS stem cells probably imply a high-grade MDS. Another remarkable observation was a progressively higher positive rate of CD96 and CD123 and lower positive rate of CD90 were observed from low- to high-grade MDS, but these alterations could not show statistical difference since there were not enough cases included in every subtype or risk group. Whatever, this tendency appeared to show a relationship between positive rate of CD90, CD96, and CD123 and MDS subtype or risk.
All in all, our study may demonstrate that MDS stem cells display a deranged phenotype different from both normal stem cells and AML stem cells, and this may make them particularly difficult to eradicate using therapies targeted against surface antigens. The percentage of cells expressing CD13 is notably higher in patients with high-grade MDS may be a potential prognostic indicator of MDS in the future. However, this indicator should be used together with a flow cytometric scoring system (35) or gene expression profiles (36) and data on the sensitivies and specificities is required. Since very recently Bonnet's group showed that some AML stem cells (defined by in vivo assay) are CD34− (37), to define the phenotype of the entire stem cells population in MDS may require further research not only on CD34+CD38− subset but also on CD34− subset.