Efficiency of Leukemic Stem Cell Separation From Patients with Acute Myeloid Leukemia

Abstract 4997

Leukemic stem cells (LSC) are the source for leukemic disease self-renewal and account for disease relapse after treatment. Therefore LSCs probably represent a critical target for therapeutic options. Xenograft models confirmed repeatedly that LSCs from AML patients reside mainly in CD34+CD38- compartment of leukemic blasts which makes the pure and efficient separation of this population mandatory to identify new therapeutic drugs to target LSC in different AML subtypes.

We separated this subpopulation out of primary AML peripheral blood mononuclear cells (PBMC) samples with fluorescence-activated cell sorting (FACS) and magnetic-activated cell sorting (MACS) and compared the efficiency of both methods. In order to profile gene expression of LSCs and hematopoietic stem cells (HSC) MicroArrays were performed using GeneChip Human Genome U133 Plus 2.0 from Affymetrix.

The CD34+CD38- subpopulation was separated from PBMCs of 12 AML patients and 5 healthy volunteers using FACS. Concerning the 12 primary AML samples, the ratio of CD34+CD38- cells ranges between 0.79% and 86.2% using 1–5×10⁷ PBMC for separation. After sorting, the purity of those AML samples increased to 88.4–98.4% while 2×10⁴-3.6×10⁶ cells were obtained. MACS was used to separate 2 representative samples, in which the CD34+CD38- subpopulation was rather small (sample1: 0.78%) or large (sample2: 86.1%). Those sorted subpopulations were compared to the samples sorted via FACS. In order to evaluate separation efficiency in a standardized manner, we defined the recovery rate: (CD34+CD38- cell number obtained /total CD34+CD38- cell number) × 100%. The total CD34+CD38- cell number was calculated through a pre-sorting FACS analysis. For sample 1, MACS resulted in a recovery rate of 4.2–6.4% with a purity of 86.6–90.3%, which is inferior to the recovery rate of 17% and the purity of 92.1% using FACS. For Sample 2, MACS resulted in a recovery rate of 0.4% with a purity of 98.8%, compared to the recovery rate of 11.6% with a purity of 98.1% by FACS. Comparing both methods it is obvious that the purity doesn't differ a lot, but the yield is much higher using FACS. This could represent a powerful tool, when managing rare samples. Finally, by comparing purity and yield, we showed that FACS is the adequate separation method. At the moment MicroArrays are being performed in order to investigate the gene expression profile for 12–15 AML patients and 5 HVs.

Taken together, we showed a widely efficient method to routinely separate LSCs from patients with different subtypes of AML. Microarrays, that have been performed, represent a method that allows the comparison of the characteristics of LSCs in different AML subtypes and also of LSCs from bone-marrow with LSCs from peripheral blood and with HVs. These array data analyses are ongoing and will be presented.