Abstract
Abstract 2558
Recent evidence has implicated the bone marrow microenvironment directly in the pathogenesis of preleukemic bone marrow disorders (Raaijmakers MHGP et al, Nature 2010) with potential to transform to AML. Moreover, the bone marrow microenvironment is critical to AML survival (Garrido et al 2001, Meads et al 2008). We sought to investigate aspects of the bone marrow microenvironment which may contribute to the pathogenesis and persistence of AML by direct analysis of primary bone marrow MSCs isolated from AML patients in comparison with primary bone marrow MSCs from normal subjects. Our analyses included (1) a comparison of cytokine elaboration between normal and AML bone marrow MSCs (2) immunophenotyping of normal and AML bone marrow MSCs (3) characterisation of binding by AML cells to their autologous stroma and (4) gene expression profiling of normal and AML bone marrow MSCs and (5) cytogenetic analysis AML bone marrow MSCs.
We have been able to derive confluent cultures of mesenchymal stromal cells from 80% of AML patient marrow samples. Fresh or cryopreserved bone marrow samples were plated in non-hematopoietic expansion media (Miltenyi) under reduced oxygen conditions. After 48 hours of culture, nonadherent cells are removed, and over a period of 1–2 weeks, cultures of spindle shaped cells are derived, that can be sustained in culture for up to 5 passages. Similar cultures were derived from normal bone marrow. Gene expression profiling of bone marrow MSCs was performed by whole genome analysis using Illumina's BeadChip microarray platform. Samples included mRNAs isolated from confluent cultures of AML bone marrow MSCs, normal bone marrow MSCs, and the normal bone marrow stromal cell lines HS27a and HS5.
Comparison of cytokine elaboration showed a statistically significant (p = 0.037) 5-fold decrease in stromal MCP-1 production by AML bone marrow MSCs compared with normal bone marrow MSCs (327 ± 169 vs. 1669 ± 570 pg/mL, mean ± SE). Normal and AML MSCs showed no statistically significant differences in the production of G-CSF, GM-CSF, M-CSF, IL6, IL12, SCF, TNFα, MCP1 and SDF1β. Like their normal counterparts, AML bone marrow MSCs strongly express CD90, CD29 (β1 integrin), CD73, CD105, CD146, and CD44. The normal bone marrow derived stromal cell lines HS27a and HS5 demonstrated moderate expression of CD324/E-cadherin (28.4% and 37.9% respectively). E-cadherin expression proved highly variable among normal bone marrow MSCs (1.9%-54.9%) and similarly variable in AML bone marrow MSCs (7.8%–56.5%). AML binding to autologous MSCs primarily dependent on β1 integrin, L-selectin and VCAM-1. In contrast, prior data (Basu et al., ASH 2010 Abstract 2756) demonstrated AML binding to the normal bone marrow stromal cell line HS27a as primarily dependent on β1 integrin, CXCR4, and E-cadherin. Gene expression profiling demonstrated no significant differences between 6 AML and 5 normal bone marrow MSCs. AML bone marrow MSCs, as expected, demonstrated marked differences from AML bone marrow mononuclear cells, expressing higher levels of connective tissue growth factor (128 fold), tropomyosin 1 (84.4 fold), collagen type 1 α1 (194 fold), collagen type 1 α2 (137 fold), collagen type 4 α1 (128 fold), collagen type 5 α1 (119 fold), transgelin (111 fold), cadherin 11 (21 fold), biglycan (137 fold), IGF binding protein 6 (18 fold), and procollagen-lysine 2-oxoglutarate 5-dioxygenase 2 (74 fold). In our cytogenetic analyses of MSCs to date, bone marrow MSCs derived from one of the complex karyotype AML patients demonstrated normal cytogenetics. In contrast, bone marrow MSCs derived from a second AML patient shared a common t(2;11) translocation present in the AML cells but demonstrated an abnormal clone with del(4q) which lacked the t(6;9) also present in the AML cells [i.e.-MSC karyotype: t(2;11), del(4q); AML karyotype: t(2;11), t(6;9)]. These results suggest that in some patients AML cells and their autologous MSCs may share the same clonal origin, while in other cases, the MSCs may have a distinct origin.
AML and normal bone marrow MSCs demonstrate only subtle differences, providing an explanation of the ability of AML bone marrow MSCs to support normal hematopoiesis after leukemic debulking (e.g. via induction chemotherapy or allogeneic stem cell transplant).
No relevant conflicts of interest to declare.
Author notes
Asterisk with author names denotes non-ASH members.