Abstract

Acute myeloid leukemia (AML) is thought to arise from a rare putative ‘leukemic stem cell’ that is capable of self-renewal and formation of leukemic blasts. Serial xenotransplantation studies in immunodeficient mice have shown that this leukemia-initiating cell resides at very low numbers within CD34(high)-positive CD38-negative AML cells. Thus, immunotherapeutic approaches successfully eradicating this cell compartment should result in cure from disease. The objective of our study was to characterize the immune phenotype of the CD38-negative and CD38-positive subsets of primary CD34(high)-positive AML blasts ex vivo. We obtained therapeutic leukapheresis products from 17 AML patients of FAB M0-M5 subtypes with white blood cell counts exceeding 10^11/L at primary diagnosis. These products were used to purify CD34-positive cells by immunomagnetic microbeads. CD34(high)-expressing cells were subsequently sorted by flow cytometry into CD38-negative and CD38-positive subsets, respectively. Both fractions were then phenotyped with fluorochrome-conjugated monoclonal antibodies for expression of surface markers previously described to differ between leukemic blasts and stem cells, i.e. CD71 (transferrin receptor), CD90 (Thy-1), CD117 (c-kit receptor), CD123 (IL3Ralpha), CD44, and CD11c. We also included markers relevant for recognition by natural killer cells and T cells, namely HLA class I, HLA-DR, CD40, CD54 (ICAM-1), CD58 (LFA-3), CD80 (B7.1), and CD86 (B7.2), as well as the lineage markers CD2, CD3, CD4, CD7, CD8, CD10, CD14, CD19, CD20, and CD56. Our results demonstrated that the CD38-positive and CD38-negative subsets of CD34(high)-positive AML blasts differed considerably in the expression of CD58, CD71, CD86, CD117, and HLA class I. Although these markers were detected on both subsets, the mean fluorescence intensity (MFI) values were lower in the CD38-negative compartment compared to the CD38-positive counterpart (medians: CD58, 1335 versus 1933; CD71, 657 versus 811; CD86, 746 versus 753; CD117, 490 versus 758; HLA class I, 4431 versus 6000). We compared the MFI values of both cell subsets with the Wilcoxon signed-rank test. P-values below 5% were detected for CD58 (p=0.005), CD71 (p=0.003), CD86 (p=0.041), CD117 (p=0.009), and HLA class I (p=0.011), respectively. The CD38-positive and CD38-negative subsets showed comparable intense staining for CD11c, CD44, CD54, CD123, and HLA-DR. In contrast, CD90, CD80, CD40, and the lineage markers were negative in both fractions. We concluded from these results that primary CD34(high)-positive CD38-negative AML blasts containing small numbers of leukemia-initiating cells expressed overall lower levels of the immune recognition molecules CD58 (LFA-3), CD86 (B7.2), and HLA class I compared to their CD38-positive counterparts. However, all CD34(high)-positive CD38-negative AML cells showed detectable HLA class I expression on the cell surface, making them accessible to T-cell based immunotherapies. In line with previous data, CD71 (transferrin receptor) and CD117 (c-kit receptor) were observed at reduced levels on CD34(high)-positive CD38-negative AML cells. Ongoing functional studies explore if the CD38-negative and CD38-positive subsets of CD34(high)-positive AML blasts differ in the immunogenicity for leukemia-reactive CD4 and CD8 T cells, both in vitro as well as in immunodeficient mice in vivo.

Disclosures: No relevant conflicts of interest to declare.

Author notes

Corresponding author