Stemness of B-cell progenitors in multiple myeloma bone marrow

Abstract

Purpose: In myeloma, B cells and plasma cells show a clonal relationship. Clonotypic B cells may represent a tumor-initiating compartment or cancer stem cell responsible for minimal residual disease in myeloma.

Experimental design: We report a study of 58 patients with myeloma at time of diagnosis or relapse. B cells in bone marrow were evaluated by multicolor flow cytometry and sorting. Clonality was determined by light chain and/or immunoglobulin chain gene rearrangement PCR. We also determined aldehyde dehydrogenase activity and colony formation growth. Drug sensitivity was tested with conventional and novel agents.

Results: Marrow CD19+ cells express a light chain identical to plasma cells and are therefore termed light chain restricted (LCR). The LCR B-cell mass is small in both newly diagnosed and relapsed patients (≤ 1%). Few marrow LCR B cells (~10%) are CD19+/CD34+, with the rest being more differentiated CD19+/CD34- B cells. Marrow LCR CD19+ B cells exhibit enhanced aldehyde dehydrogenase activity versus healthy controls. Both CD19+/CD34+ and CD19+/CD34- cells showed colony formation activity, with colony growth efficiency optimized when stroma-conditioned medium was used. B-cell progenitors showed resistance to melphalan, lenalidomide, and bortezomib. Panobinostat, a histone deacetylase inhibitor, induced apoptosis of LCR B cells and CD138+ cells. LCR B cells are CD117, survivin, and Notch positive.

Conclusions: We propose that antigen-independent B-cell differentiation stages are involved in disease origination and progression in myeloma. Furthermore, investigations of myeloma putative stem cell progenitors may lead to novel treatments to eradicate the potential reservoir of minimal residual disease.

Figure 1. Phenotype of bone marrow B cell populations in multiple myeloma

(A) Multicolor flow cytometric analysis of marrow cells gated based on SSC/FSC properties to avoid granulocytes and monocytes contamination (left top panel). CD14+ cells were excluded to avoid additional monocytes contamination (Supplement Figure 1). Plasma cells were identified as CD138+ (left middle panel). CD138−/low SSC were further analyzed based on CD19 and CD34 expression, with the same gating strategy applied to peripheral blood (PB) cells (right panels). Kappa and lambda light chain (LC) expression levels in cell surface were tested in bone marrow subpopulations (right panels) compared to fluorescence minus one (FMO) controls. (B) Immunoglobulin (Ig) gene rearrangement of a representative lambda positive patient. DNA extracted from sorted CD138+/lambda+ and CD138−/CD19+/lambda+ cells followed by PCR amplification using primers targeting all three Ig heavy chain (Ig HC) frameworks (FR) and Ig lambda light chain. PCR products were separated and detected by capillary electrophoresis. Peaks shown fall within acceptable ranges for each primer (FR1: 290–360 bp, FR2: 235–295 bp, FR3: 69–129 bp, and lambda:135–170bp). As a negative control, we used polyclonal DNA, per manufacturer’s instructions. (C) Percentages of light chain restricted (LCR) CD138−/CD19+ clonotypic cells in whole marrow in patients at time of diagnosis (n = 23) or in patients at time of relapse (n = 21) in kappa (left) and lambda LCR myeloma (right). Statistical analysis was performed with Student t test. (D) Light chain restriction of CD19+ cells in bone marrow was determined based on the kappa-to-lambda ratio. Graph compares kappa restricted (left) and lambda restricted (right) patients at diagnosis or at relapse compared to healthy marrow cells.

Figure 1. Phenotype of bone marrow B cell populations in multiple myeloma

(A) Multicolor flow cytometric analysis of marrow cells gated based on SSC/FSC properties to avoid granulocytes and monocytes contamination (left top panel). CD14+ cells were excluded to avoid additional monocytes contamination (Supplement Figure 1). Plasma cells were identified as CD138+ (left middle panel). CD138−/low SSC were further analyzed based on CD19 and CD34 expression, with the same gating strategy applied to peripheral blood (PB) cells (right panels). Kappa and lambda light chain (LC) expression levels in cell surface were tested in bone marrow subpopulations (right panels) compared to fluorescence minus one (FMO) controls. (B) Immunoglobulin (Ig) gene rearrangement of a representative lambda positive patient. DNA extracted from sorted CD138+/lambda+ and CD138−/CD19+/lambda+ cells followed by PCR amplification using primers targeting all three Ig heavy chain (Ig HC) frameworks (FR) and Ig lambda light chain. PCR products were separated and detected by capillary electrophoresis. Peaks shown fall within acceptable ranges for each primer (FR1: 290–360 bp, FR2: 235–295 bp, FR3: 69–129 bp, and lambda:135–170bp). As a negative control, we used polyclonal DNA, per manufacturer’s instructions. (C) Percentages of light chain restricted (LCR) CD138−/CD19+ clonotypic cells in whole marrow in patients at time of diagnosis (n = 23) or in patients at time of relapse (n = 21) in kappa (left) and lambda LCR myeloma (right). Statistical analysis was performed with Student t test. (D) Light chain restriction of CD19+ cells in bone marrow was determined based on the kappa-to-lambda ratio. Graph compares kappa restricted (left) and lambda restricted (right) patients at diagnosis or at relapse compared to healthy marrow cells.

Figure 1. Phenotype of bone marrow B cell populations in multiple myeloma

(A) Multicolor flow cytometric analysis of marrow cells gated based on SSC/FSC properties to avoid granulocytes and monocytes contamination (left top panel). CD14+ cells were excluded to avoid additional monocytes contamination (Supplement Figure 1). Plasma cells were identified as CD138+ (left middle panel). CD138−/low SSC were further analyzed based on CD19 and CD34 expression, with the same gating strategy applied to peripheral blood (PB) cells (right panels). Kappa and lambda light chain (LC) expression levels in cell surface were tested in bone marrow subpopulations (right panels) compared to fluorescence minus one (FMO) controls. (B) Immunoglobulin (Ig) gene rearrangement of a representative lambda positive patient. DNA extracted from sorted CD138+/lambda+ and CD138−/CD19+/lambda+ cells followed by PCR amplification using primers targeting all three Ig heavy chain (Ig HC) frameworks (FR) and Ig lambda light chain. PCR products were separated and detected by capillary electrophoresis. Peaks shown fall within acceptable ranges for each primer (FR1: 290–360 bp, FR2: 235–295 bp, FR3: 69–129 bp, and lambda:135–170bp). As a negative control, we used polyclonal DNA, per manufacturer’s instructions. (C) Percentages of light chain restricted (LCR) CD138−/CD19+ clonotypic cells in whole marrow in patients at time of diagnosis (n = 23) or in patients at time of relapse (n = 21) in kappa (left) and lambda LCR myeloma (right). Statistical analysis was performed with Student t test. (D) Light chain restriction of CD19+ cells in bone marrow was determined based on the kappa-to-lambda ratio. Graph compares kappa restricted (left) and lambda restricted (right) patients at diagnosis or at relapse compared to healthy marrow cells.

Figure 1. Phenotype of bone marrow B cell populations in multiple myeloma

(A) Multicolor flow cytometric analysis of marrow cells gated based on SSC/FSC properties to avoid granulocytes and monocytes contamination (left top panel). CD14+ cells were excluded to avoid additional monocytes contamination (Supplement Figure 1). Plasma cells were identified as CD138+ (left middle panel). CD138−/low SSC were further analyzed based on CD19 and CD34 expression, with the same gating strategy applied to peripheral blood (PB) cells (right panels). Kappa and lambda light chain (LC) expression levels in cell surface were tested in bone marrow subpopulations (right panels) compared to fluorescence minus one (FMO) controls. (B) Immunoglobulin (Ig) gene rearrangement of a representative lambda positive patient. DNA extracted from sorted CD138+/lambda+ and CD138−/CD19+/lambda+ cells followed by PCR amplification using primers targeting all three Ig heavy chain (Ig HC) frameworks (FR) and Ig lambda light chain. PCR products were separated and detected by capillary electrophoresis. Peaks shown fall within acceptable ranges for each primer (FR1: 290–360 bp, FR2: 235–295 bp, FR3: 69–129 bp, and lambda:135–170bp). As a negative control, we used polyclonal DNA, per manufacturer’s instructions. (C) Percentages of light chain restricted (LCR) CD138−/CD19+ clonotypic cells in whole marrow in patients at time of diagnosis (n = 23) or in patients at time of relapse (n = 21) in kappa (left) and lambda LCR myeloma (right). Statistical analysis was performed with Student t test. (D) Light chain restriction of CD19+ cells in bone marrow was determined based on the kappa-to-lambda ratio. Graph compares kappa restricted (left) and lambda restricted (right) patients at diagnosis or at relapse compared to healthy marrow cells.

Figure 1. Phenotype of bone marrow B cell populations in multiple myeloma

(A) Multicolor flow cytometric analysis of marrow cells gated based on SSC/FSC properties to avoid granulocytes and monocytes contamination (left top panel). CD14+ cells were excluded to avoid additional monocytes contamination (Supplement Figure 1). Plasma cells were identified as CD138+ (left middle panel). CD138−/low SSC were further analyzed based on CD19 and CD34 expression, with the same gating strategy applied to peripheral blood (PB) cells (right panels). Kappa and lambda light chain (LC) expression levels in cell surface were tested in bone marrow subpopulations (right panels) compared to fluorescence minus one (FMO) controls. (B) Immunoglobulin (Ig) gene rearrangement of a representative lambda positive patient. DNA extracted from sorted CD138+/lambda+ and CD138−/CD19+/lambda+ cells followed by PCR amplification using primers targeting all three Ig heavy chain (Ig HC) frameworks (FR) and Ig lambda light chain. PCR products were separated and detected by capillary electrophoresis. Peaks shown fall within acceptable ranges for each primer (FR1: 290–360 bp, FR2: 235–295 bp, FR3: 69–129 bp, and lambda:135–170bp). As a negative control, we used polyclonal DNA, per manufacturer’s instructions. (C) Percentages of light chain restricted (LCR) CD138−/CD19+ clonotypic cells in whole marrow in patients at time of diagnosis (n = 23) or in patients at time of relapse (n = 21) in kappa (left) and lambda LCR myeloma (right). Statistical analysis was performed with Student t test. (D) Light chain restriction of CD19+ cells in bone marrow was determined based on the kappa-to-lambda ratio. Graph compares kappa restricted (left) and lambda restricted (right) patients at diagnosis or at relapse compared to healthy marrow cells.

Figure 2. Stem cell-like phenotype of clonotypic B cell progenitors in multiple myeloma bone marrow

(A) Representative contour plots of ALDH activity of CD138−/CD34+/CD19− multipotent progenitors, comparing CD138− cells gated first based on light chain (kappa or lambda) similar to plasma cells of each individual patient, thus termed light chain restricted (LCR) and CD138+ cells (top panels). As a negative control, an aliquot of aldefluor-stained cells was immediately quenched with a specific ALDH inhibitor (DEAB) (bottom panels). Scatter plot shows percentages of cells with increased ALDH activity within bone marrow subpopulations (n=7). Statistical analysis was performed with Student t test. (B) Representative contour plots showing FACS labeling with Notch-1, survivin, and CD117 (c-Kit) expression gated based on FMO controls (not shown).

Figure 2. Stem cell-like phenotype of clonotypic B cell progenitors in multiple myeloma bone marrow

(A) Representative contour plots of ALDH activity of CD138−/CD34+/CD19− multipotent progenitors, comparing CD138− cells gated first based on light chain (kappa or lambda) similar to plasma cells of each individual patient, thus termed light chain restricted (LCR) and CD138+ cells (top panels). As a negative control, an aliquot of aldefluor-stained cells was immediately quenched with a specific ALDH inhibitor (DEAB) (bottom panels). Scatter plot shows percentages of cells with increased ALDH activity within bone marrow subpopulations (n=7). Statistical analysis was performed with Student t test. (B) Representative contour plots showing FACS labeling with Notch-1, survivin, and CD117 (c-Kit) expression gated based on FMO controls (not shown).

Figure 3. Colony formation assay of clonotypic B cell progenitors in multiple myeloma bone marrow

(A) Comparison of colony development in methylcellulose (MC) supplemented with PHA stimulated-5% lymphocyte-conditioned medium (PHA-LCM). Bone marrow (BM) cells were gated and sorted as described in Design and Methods. MC cultures were established with CD138−/CD19−/CD34+ (1,000 cells) or LCR CD138−/CD19+ (100,000 cells) and CD138+ cells (300,000 cells). Colonies were scored after 14 days of culture. Graph shows results (mean ± SD) from 5 myeloma patients. Colonies were harvested from each group for morphologic evaluation (hematoxylin-eosin stain, Zeiss Axiovert inverted microscope, magnification x40). (B) Sorted CD138−/CD19−/CD34+ or CD138− LCR CD19+ and CD138+ cells were cultured for 14 days. After induction stage, developing colonies or CD138+ cells that remained in culture were harvested for FACS analysis. Representative contour plots show CD138 expression on input cells before culture (day 0) and on harvested cells (day 14). (C) Colony counts in MC PHA-LCM cultures for 14 days with LCR BM cells, sorted based on phenotype as indicated. Results (mean ± SD) shown colony numbers of 3 myeloma patients/experiment.

Figure 4. Colony formation cell assay of myeloma CD138−/CD34+/CD19− multipotent progenitors or clonotypic B cells in the presence of marrow-conditioned medium

(A) Quantification of hematopoietic progenitors in methylcellulose (MC) colony assays with sorted CD138−/CD34+/CD19− cells (n = 3). Culture was supplemented with 5% PHA-LCM in all culture conditions (baseline); and with either HS5 stroma conditioned medium (stroma-CM) or recombinant human IL-2 IL-6, IL-10, IL-15 and IL-21. Graph represents colony output of triplicate experiments from 3 myeloma patients at 14 days. Representative contour plots show phenotype (CD34, CD19, and CD138) of plucked cells from MC culture. CD138 cells in each culture condition were analyzed for light chain expression. Histogram shows kappa (solid line) and lambda (dotted line) light chain (LC) versus fluorescence minus one (FMO) controls (gray shade). (B) Sorted LCR CD138−/CD19+ cells were grown in MC PHA-LCM (baseline) or in the presence of stroma-CM. Colonies were scored at day 14 of culture. Images show representative colonies grown in each condition (Zeiss Axiovert Inverted Microscope, magnification 5X).

Figure 5. Drug sensitivity of clonotypic B cell progenitors in multiple myeloma bone marrow

(A) Apoptotic response of myeloma patients’ bone marrow (BM) cells after treatment with melphalan (25 μM) or acid-ethanol (control) for 48 hours (n = 3). Multipotent progenitors were sorted based on CD138−/CD19−/CD34+ expression. To isolate CD138+ or LCR CD19+ cells, BM was first gated on surface LCR corresponding to each patient’s plasma cells (kappa or lambda). Apoptosis was determined with annexin V-PE and 7-AAD staining. Percentage of each population is indicated in each quadrant of representative contour plots. (B) Apoptosis responses to treatment of BM subpopulations with bortezomib (10 nM) or DMSO (control) for 48 hours. Graph represents mean ± SD of percent specific apoptosis determined by flow cytometry (n = 3). (C) Apoptosis responses to treatment of BM subpopulations with lenalidomide (10 μM) or DMSO (control) for 48 hours. Graph represents mean ± SD of percent specific apoptosis determined by flow cytometry (n = 3). (D) Apoptosis responses to treatment of BM subpopulation with panobinostat (100 nM) or DMSO (control) for 24 hours. Graph represents mean ± SD of percent specific apoptosis determined by flow cytometry (n = 5).