CD200 in acute myeloid leukemia: marked upregulation in CEBPA biallelic mutated cases

Abstract

CD200 is a glycoprotein that binds with its receptor CD200R, providing immunosuppressive signals to T and NK cells. CD200 is expressed by normal stem cells and progenitors committed to B-lymphopoiesis and myeloid development. CD200 biological relevance in acute leukemias is only partially understood.The study included a consecutive series of four hundred thirty-one patients with acute myeloid leukemia (AML). Immunophenotype was established by multiparametric flow cytometry, and the genetic diagnosis was performed by PCR-based methods and a targeted resequencing method covering 42 genes.66% of AML patients expressed CD200 being significantly associated with CD34 reactivity. The frequency of CD200 positivity was higher in cases with core-binding factor genetic lesions such as RUNX1-RUNX1T1 (81.3%) fusions and CBFB-MHY11 (63.2%) rearrangements and also with biallelic CEBPA mutations (100%). The molecular AML group with the lowest CD200 reactivity (19.1%) corresponded to AML with NPM1 mutations. RNA seq showed no uniform pattern of infiltrating cells in CEBPA mutated AML. Deconvolution analysis may be used to assess the immunoregulatory mechanisms of AML.CD200 expression could help identify the more immature compartment and, combined with other markers, single out CEPA-mutated AML.

Keywords: CEBPA -NPM1; FLT3; Acute myeloid leukemia- RUNX1; CD200.

Fig. 1

CD200 positivity patterns. Arbitrary positivity patterns were established based on the presence of negative cells and MFI. Pattern 0 or negative (P0) 0–20% expression of CD200 in blast CD45 + gate; Pattern 1 or partially positive (P1) 20–50% expression of CD200; Pattern 2 or fully positive (P2) > 50% expression of CD200

Fig. 2

CD200 expression in AML with CBF genetic lesions. Immunophenotype from one AML patient with a RUNX1/RUNX1T1 rearrangement (UPN 213) which is associated with P1 (partially positive), expressed 44.18% CD200-positive cells. Red: Leukemic cells expressing CD200

Fig. 3

CD200 expression in AML with CBF genetic lesions. Immunophenotype obtained from one AML patient with a CBFB-MYH11 rearrangement (UPN: 347) which is associated with P1 (partially positive), expressed 28.53% CD200-positive cells. Red: Leukemic cells expressing CD200

Fig. 4

CD200 expression in AML with biallelic CEBPA mutation. Immunophenotype from one AML patient with biallelic CEBPA mutation (UPN: 102; c.68delC // P23fsX137) which is associated with P2 (fully positive), expressed 80.08% CD200-positive cells. Red: Leukemic cells expressing CD200

Fig. 5

CD200 expression by RT-PCR. CD200 expression comparing 3 groups: t(8,21), inv(16) and CEBPA-m vs. NPM1-m vs. Other AML. p < 0.001 (***); p < 0.05 (*)

Fig. 6

Immune cell deconvolution of AML biallelic CEBPA patients, TCGA-AML and BEAT patients. (A) CD200 expression (log2 normalized) in CEBPA mutated patients and WT patients in BEAT and TCGA cohorts. P-values from Wilcoxon rank sum test. (B) Deconvolution results of the RNA-seq data from 13 AML cases with biallelic CEBPA (top left panel), TCGA-AML CEBPA-mutated patients (top center panel) and TCGA-AML CEBPA WT (top right panel), BEAT CEBPA-mutated patients (bottom left panel) and BEAT CEPBA-WT (bottom right panel). Cell type proportions are shown in the y axis. Samples were clustered with euclidean distance and complete method. (C) Cell types with significantly different proportions between CEBPA-mutated and CEBPA-WT in the meta-analysis of the TCGA and BEAT cohorts. Random effects models p-value: B cells naive (p-value = 0.0867), B cells memory (p-value = 3.53E-5), dendritic cells activated (p = 0.0117) and Monocytes (p-value = 0.074). Meta-analysis plots can be found in Suppl. Figure 4.1. (D) Correlation of CD200 expression (log2 normalized) with B cells naïve and memory. Pearson correlation coefficient (R) and p-values are shown separately for CEBPA mutated and WT patients

Fig. 7

Pathway analysis on bulk RNAseq experiments from CEBPA mutated AML. Gene network plot of CD200-CD200R1 related genes. Genes analyzed were previously identified as WGCNA modules that significantly correlated with CD200 expression in CEBPA mutated patients. CD200, CD200R1, DOK2 and RASA1 from the CD200-CD200R pathway were included in the correlation analysis. Only genes connected to CD200 and/or CD200R1 were retained for visualization. A correlation threshold of 0.75 was set for the network