FellowRNDr., PhD. Jana Jakubikova
Project NameClonal architecture in development of MM and therapy induced intra-clonal heterogeneity in MM progression
Host organisationBiomedical Research Center
Duration of the project01.09.2015 - 31.12.2018

Tumor heterogeneity is likely, from a Darwinian-selection perspective, to be the essential feature of clonal evolution, disease progression and relapse. The existence of intra-clonal heterogeneity resulting from clonal selection has been recently reported in many cancer types. Multiple myeloma (MM), as a prototype disease of intra-clonal heterogeneity resulting from clonal selection, is a B cell malignancy characterized by clonal proliferation of malignant plasma cells in the bone marrow. MM remains incurable despite improved survival after development of novel therapies. The overall objective of this proposal is to better understand clonal architecture of primary patient-derived BM samples during the development of MM and therapy induced intra-clonal dynamics during progression of MM. A central component of these studies is evolution of bone marrow samples from MGUS, SMM, and MM patients before, during and after anti-MM therapies. By combining genetic, molecular and phenotypic approaches, the impact of chemotherapy and immunotherapy on the dynamic nature of the clonal composition together with the role of the tumor microenvironment on clonal selection in MM will be evaluated. To reveal insights into intra-clonal heterogeneity in MM, the study will provide the framework for development of more personalized diagnostic criteria and novel therapeutic strategies against coexisting persistent subclones resulting in a more individualized targeted therapy to either maintain long-term remission or completely eradicate MM disease.

Project Summary with Interim Results

To characterize dynamics of tumor intra/inter-clonal heterogeneity during the development and progression of multiple myeloma (MM) based on cellular and molecular complexity provided in the Sp. Aim 1, I used time-of-flight mass cytometry (CyTOF) in order to analysis limited quantity of bone marrow (BM) samples from MM patients on the single cell level. CyTOF is a novel high-dimensional technology, enables simultaneous evaluation of up to 40 parameters on the cell surface (immunophenotype) and intracellularly (signaling/regulatory molecules and transcriptional factors) as well.

I have designed 3 panels of 40 markers (approximately 100 markers due to the some repetitions) in which the antibodies were labeled with rare stable earth elemental metals (lanthanides). In general, the impurity and oxidation contributions for CyTOF metals are quite low when compared to the average spillover from fluorescence in flow cytometry. However, designing functional panels represent important step to determine channels to which a given metals may contribute and which one may receive background signal. Based on these knowledges specific antibodies were conjugated with appropriate metals (lanthanides).  After antibody-metal conjugations, I evaluated and validated their efficacy using negative and positive controls (either based on cell lines with positive expression of particular marker vs. negative or by stimulation with drug/cytokine of the specific marker) by CyTOF analyses and then, compared to fluorescent-based flow cytometry.

            To evaluate B cell lineage hematopoiesis in MM, I design 2 panels in which I used B cell-stage indicators CD10, CD19, CD20, CD22, CD27, CD34, CD38, CD45, IgA, IgD, IgG and IgM to define the spectrum of maturation of B cell precursors from hematopoietic stem cells to mature B lymphocytes as well as clonal plasma cells by CyTOF technology. To detect malignant plasma cells, I consistently used analyses for expression of a 6 marker phenotype: CD19/CD38/CD138/CD45/cyt-kappa/cyt-lambda in all panels. The resulting high-dimensional data are delineated by spanning-tree progression analysis of density-normalized events (SPADE) or/and t-Distributed Stochastic Neighbor Embedding (t-SNE) algorithm (viSNE) analyses. By clustering SPADE analyses using cytobank software, I was able to define main B maturation stages such as pro-B, pre-B I, pre-B II, immature B cells, naïve and transitional B cells, together with memory non-switched and memory switched B cells, plasmablasts and plasma cells (as shown in Fig.1). To complete analyses of B cell lineage by functional regulatory molecules in different B signaling pathways, my panels also were comprised of cell surface specific markers (CD25, CD28, CD47, CD52, CD56, CD81, CD117, CD184, CD200, CD221, CD243, CD319, CD338, etc.) and intracellular markers (sXBP1, IRF4, MYD88, FGFR3, Notch1, Sox-2, c-Myc, Bcl-6, Blimp-1, Pax5, RARa2 etc.) in order to study tumor heterogeneity.

My preliminary data of SPADE analysis of CD38 expression of newly diagnosed MM patient with t(11;14) translocation (SPADE tree in Fig.1) showed significant CD38 expression in B cell precursors from pro-B to mature B cells (orange color), with the highest intensity in malignant plasma cells (red color); in contrast, CD38 was only minimally expressed in both memory B cell subsets (blue to green). As an example, to support the idea of intra-clonal heterogeneity, various expressions of surface markers (CD319 and CD56) and intracellular regulatory molecules (sXBP-1, cyto-kappa, IRF-4, FGFR3, Nanog) in clusters of plasma cells and plasmablasts were observed (Fig. 1; in the frame; frame below the SPADE tree) as well as illustrate the use of CyTOF for clonal evolution using cellular and molecular approaches.

            Since MM is characterized by immune dysfunction, the impact of the complex immune system on clonal dynamics in the tumor microenvironment or vice versa during development and progression of MM provided in the Sp. Aim 3 was also determined. Therefore, I have designed a panel of 40 markers to define: natural killer (NK) subsets (such as NK and NKT cells); T cell subsets (such as memory CD4T, naive CD4T, memory CD8T, naive CD8T, T regulatory cells and Tg/d cells); as well as markers for monocytes, myeloid and plasmacytoid dendritic cells, as well as granulocytic, erythroid, and platelet lineages together with tumor cells.

            The high-dimensional data of immune subsets of a newly diagnosed MM patient are shown in SPADE analyses (Fig. 2). In Fig. 2A, expression of CD45 is shown with highest intensity in T and B lymphocytes, NK cells, monocytes (red color); with medium expression of CD45 in granulocytes (orange to yellow); in contrast, CD45 expression was significantly lower in both plasma cells and erythroid lineage (green to blue). The major subsets of T cells were either T helper cells defined by CD3+/CD4+ or T killer cells defined by CD3+/CD8+, together with CD2, CD5 and CD7 expression. NKT cells were defined by CD3+/CD161+ expression in T lymphocytes, whereas NK cells clustered separately from T cells and were defined by CD3-/CD56+/CD16+/CD161+/CD7+ and CD2 expression (Fig. 2B). In Fig. 2C, using manual gating strategies I defined cells by DNA Ir191/193 intercalator (i) which incorporate into DNA and also allowed for removal of doublets (ii) and then, dead cells were removed as well (iii). From all viable CD45+/- (iv) cells and gating on CD3+ cells (v), either CD4+T helper or CD8+ T killer cells were detected. Moreover, both naïve and memory cells as well as Treg cells were identified. Similarly, from CD3+ (v) population I identified CD161+ NKT cells. From all viable CD45+/-(iv) and gating on CD3-/CD19- (vi) population, I was able to identify NK cells by CD56 and CD16 expression (Fig. 2C). Multiple markers were combined to evaluate differences between monocytes, dendritic cells and granulocytes lineages in SPADE analyses (Fig. 2D). Similarly, hierarchic gating strategies from CD3-/CD19- (vi) population showed either: monocytic differentiation with CD13+/HLA-DR+/CD33+ towards mature monocytes with CD14+ expression; or granulocytic differentiation with HLA-DR-/CD33-/CD11b+ towards mature neutrophils with CD16+ expression. Moreover, gating on CD3-/CD19-/CD14-/CD16- population with HLA-DR+ expression, I detected either plasmacytoid dendritic cells (CD123+CD11c-) or myeloid dendritic cells (CD123-CD11c+) vs. HLA-DR-/CD123+ population representing basophils (Fig. 2E). B cells expressed CD45+/CD19+/CD38+/CD138- were either cyto kappa+ or cyto lambda+, whereas malignant plasma cells expressed CD45-/CD19-/CD38++/CD138+/cyto kappa+ vs. lambda- in SPADE analyses (Fig. 2F). Similarly, CD38+ cells with CD138+ expression represent malignant plasma cells with cyto kappa+ clone, while CD38+ cells with CD19+ expression with either cyto kappa or cyto lambda positivity represent B cells (Fig. 2G). Furthermore, erythroid and platelet lineages were detected as well (Fig. 2H).

            My preliminary high-dimensional data of follow up MM patient samples (before and after treatment) analyzed by viSNE algorithm showed significant decrease of malignant plasma cells: newly diagnosed MM patient before (top row) and after Revlimid-Velcade-Dexamethasone treatment (bottom row); however, a minor clone persisted. Moreover, decrease expression of immune cells (B cells, T cells, monocytic and granulocytic lineages) was also observed (Fig. 3). Using both SPADE and viSNE analyses, this allows me to correlate malignant plasma cell clonal composition with overall immune cell status in an individual patient, and its relationship to clonal heterogeneity and treatment at a given point in time.