Evaluation of CD9 as a Candidate Leukemia Stem Cell Marker In Childhood Precursor B Cell Acute Lymphoblastic Leukemia

Abstract 3241

Leukemia cells are believed to arise from leukemia stem cells (LSC). It is also known that LSC are responsible for relapse in certain types of leukemia, such as acute myeloid leukemia (AML). However, the existence and role of LSC in acute lymphoblastic leukemia (ALL) is unclear. CD9 was reported to be a marker for LSC in B-ALL using cell lines (Nishida H. et al., 2009). CD9 is a tetraspanin and is believed to be involved in cell adhesion, motility, and signaling events. It is also involved in metastasis; however, the mechanisms are unknown. Since childhood ALL is a heterogeneous group of diseases and cell lines can be different from primary leukemia cells, we tested the role of CD9 as a candidate LSC marker using primary precursor B (preB) ALL cells from pediatric patients. Two methods, Raman spectroscopy and serial transplantation of sorted leukemia cells in NOD/SCID/IL2R g null (NSG) mice, were used to confirm LSC. Raman spectroscopy is a laser-based technique for the single cell analysis of intrinsic molecular vibrations reflecting cellular biochemical information. It can provide a quantitative assessment of the levels of DNA, RNA, proteins, lipids, and carbohydrates in the cell, as well as molecular-level conformational changes. Previous studies by our group showed that unique Raman fingerprints were identified in normal blood cells, ALL cells, and stem cells, including hematopoietic stem cells and embryonic stem cells. Four preB ALL samples were stained for CD9 and sorted by flow cytometry. ALL samples were obtained from patients at diagnosis or freshly harvested from NSG mice engrafted with primary leukemia samples. All samples showed heterogeneous expression of CD9. CD9 high-positive cells and negative cells were flow sorted. Raman spectra of freshly sorted CD9 high-positive and negative cells were obtained. 10 to 20 cells were analyzed in each sample. CD34 positive cells, which were isolated from normal donors, were also analyzed by Raman spectroscopy as a control. No unique Raman fingerprints were identified to separate CD9 high-positive cells from negative cells using Principal Component Analysis (PCA). Furthermore, CD9 high-positive and negative cells from three preB ALL samples were transplanted into NSG mice via intra-bone marrow injection. Equal cell numbers (5×10⁵ to 1.5×10⁶ cells) were used for positive and negative samples in each injection. The majority of the mice from both groups (transplanted with CD9 high-positive or negative cells) developed leukemia 3 to 4 months after injection. Leukemia phenotype was confirmed to be the same as the original leukemia. In conclusion, although CD9 was shown to be a marker for LSC in B-ALL cell lines, it does not appear to be an LSC marker in primary preB ALL. Since childhood preB ALL is a heterogeneous group of diseases, larger cohorts are necessary to confirm our findings. Raman spectroscopy may be a useful screening tool for analysis of cellular intrinsic markers.