Differentiation stage of myeloma plasma cells: biological and clinical significance

Abstract

The notion that plasma cells (PCs) are terminally differentiated has prevented intensive research in multiple myeloma (MM) about their phenotypic plasticity and differentiation. Here, we demonstrated in healthy individuals (n=20) that the CD19-CD81 expression axis identifies three bone marrow (BM)PC subsets with distinct age-prevalence, proliferation, replication-history, immunoglobulin-production, and phenotype, consistent with progressively increased differentiation from CD19+CD81+ into CD19-CD81+ and CD19-CD81- BMPCs. Afterwards, we demonstrated in 225 newly diagnosed MM patients that, comparing to normal BMPC counterparts, 59% had fully differentiated (CD19-CD81-) clones, 38% intermediate-differentiated (CD19-CD81+) and 3% less-differentiated (CD19+CD81+) clones. The latter patients had dismal outcome, and PC differentiation emerged as an independent prognostic marker for progression-free (HR: 1.7; P=0.005) and overall survival (HR: 2.1; P=0.006). Longitudinal comparison of diagnostic vs minimal-residual-disease samples (n=40) unraveled that in 20% of patients, less-differentiated PCs subclones become enriched after therapy-induced pressure. We also revealed that CD81 expression is epigenetically regulated, that less-differentiated clonal PCs retain high expression of genes related to preceding B-cell stages (for example: PAX5), and show distinct mutation profile vs fully differentiated PC clones within individual patients. Together, we shed new light into PC plasticity and demonstrated that MM patients harbouring less-differentiated PCs have dismal survival, which might be related to higher chemoresistant potential plus different molecular and genomic profiles.

Figure 1

Bone marrow (BM) normal plasma cell (PC) subsets according to the CD19 − CD81 expression axis. (a) Age-related changes in the distribution of BM normal PC subsets. The percentage of the CD19⁺CD81⁺, CD19⁻CD81⁺ and CD19⁻CD81⁻ subsets within total BM PCs from each healthy donor (n = 20) was determined, and median values per subset for each age decade are represented by light, intermediate and dark blue areas, respectively. Linear regression between individuals’ age and the respective percentage for each PC subset is also shown. (b) Proliferative potential of the different BM normal PC subsets. The percentage of the CD19⁺CD81⁺ (light blue), CD19⁻CD81⁺ (intermediate blue) and CD19⁻CD81⁻ (dark blue) subsets within total BM PCs from healthy donor (n = 5) in G₀/G₁ and S-phase/G₂M phases of the cell cycle is shown. (c) Quantification of the replication history of progressively maturing BM B cell and PC subsets from healthy individuals using κ-deleting recombination excision circles. The line in the middle and vertical lines correspond to the median value and both the 10th and 90th percentiles, respectively, for the ΔCT between the coding joint and the signal joint in FACS-sorted B-cell precursors, transitional, naïve and memory B-cells, CD19⁺ and CD19⁻ PCs from BM samples of healthy individuals (n = 5). (d) Immunoglobulin (Ig) heavy chain isotype distribution of the different BM normal PC subsets. After PC identification according to their bright CD38 and CD138 expression and unique scatter characteristics, cyIgG⁺ PCs were defined as those showing reactivity in the PE channel (cyIgA+cyIgG) but not in the FITC channel (cyIgM+cyIgA), whereas cyIgA⁺ PCs were defined as those showing (diagonal) double-staining in the FITC+PE channels; cyIgM⁺ PCs were defined by reactivity in the FITC channel but not in PE. The percentage of cytoplasmic IgG, IgA and IgM is shown within the respective CD19⁺ CD81⁺, CD19⁻CD81⁺ and CD19⁻CD81⁻ subsets. (e) Immunophenotypic protein expression profiles of the different BM normal PC subsets. Due to the existence of five parameters measured in common for each aliquot (CD38, CD45, CD19, forward light scatter – FSC and sideward light scatter – SSC), it was possible to define the PC compartment in each aliquot and fuse the different data files corresponding to the four different eight-colour MoAb combinations studied per sample into a single data file containing all information measured for that sample, using the merge function of the Infinicyt software. For any single PC in each eight-colour MoAb combination, this included data about those antigens that were measured directly on it and antigens that were not evaluated directly (‘missing values’) for that cell in the corresponding tube it was contained in. Then, the calculation function of the Infinicyt software was used to fill in the ‘missing values’, based on the ‘nearest neighbour’ statistical principle, defined by the unique position of individual PCs the multidimensional space created by the five common (backbone) parameters (FSC, SSC, CD38, CD45 and CD19). Ultimately, the expression of all 23 phenotypic markers could be analysed at the single PC level, and compared between PCs clustering into the specific CD19⁺CD81⁺, CD19⁻CD81⁺ and CD19⁻CD81⁻ subsets. Markers differentially expressed between the CD19⁺CD81⁺ (light blue), CD19⁻CD81⁺ (intermediate blue) and CD19⁻CD81⁻ (dark blue) subsets within BM normal PCs from healthy individuals (n = 10). Notched boxes represent the 25th and 75th percentile values of the amounts of antigen mean fluorescence intensity expression per BM PCs; the line in the middle and vertical lines correspond to the median value and both the 10th and 90th percentiles, respectively.

Figure 2

Multiple myeloma (MM) patients’ survival according to the differentiation stage of myeloma PC clones. Panels a and b show progression-free survival (PFS) and overall survival (OS) for the overall series of MM patients (n = 225) grouped according to the differentiation stage of clonal plasma cells (PCs) at diagnosis: more-differentiated (CD19⁻CD81⁻), intermediate-differentiated (CD19⁻CD81⁺) and less-differentiated (CD19⁺CD81⁺). Patients’ treatment consisted of either nine identical induction cycles with bortezomib, melphalan, prednisone (VMP) followed by other nine cycles of lenalidomide plus low-dose dexamethasone (Rd; n = 112), or alternating cycles of VMP and Rd for up to 18 courses (n = 113). The median follow-up of the series was 3 years. Panels c and d show PFS and OS in patients with standard-risk cytogenetics (n = 154; all those cases without t(4;14), t(14;16) and/or del(17p13)).

Figure 3

Therapeutic selection at the MRD stage of myeloma PC subclones defined according to their differentiation stage. (a) Correlation between the percentage of CD19 (black squares) and CD81 (open circles) positive plasma cells (PCs) within total baseline (x axis) vs MRD (y axis) clonal PCs in longitudinal bone marrow samples from 40 multiple myeloma (MM) patients analysed at diagnosis and after therapy. (b) Schema showing the frequency of patients following specific clonal dynamics according to the differentiation stage of myeloma PCs from diagnosis to the MRD stage. Representative bivariate dot plot histograms illustrating the patterns of CD19 vs CD81 expression in clonal PCs at diagnostic (represented by lines corresponding to one and two SD) and at the MRD stage (red dots) corresponding to four out of the eight patients that evolved from baseline more differentiated (that is: CD19⁻CD81⁻) into intermediate-differentiated (that is: CD19⁻CD81⁺) chemoresistant PC clones after therapy, ordered from left to right, denoting high to low MRD levels, are also shown. Twelve out of the 30 patients displaying the same differentiation stage during baseline and MRD monitoring attained CR, three out of the eight cases with fully differentiated phenotypes at diagnosis showing intermediate stage chemoresistant clonal PCs after therapy attained CR, and so did one out of the two patients transitioned from a CD19⁻CD81⁺ into a CD19⁻CD81⁻ phenotype.

Figure 4

Distinct PC differentiation subsets within individual patients show different mutation profiles. Clonal plasma cells (PCs) corresponding to the intermediate- (CD19⁻CD81⁺) and more-differentiated (CD19⁻CD81⁻) subsets were FACS-sorted from patients #1, #2, #4, #5 and #6 (a, b, d, e and f) for mutation analysis using a targeted-sequencing panel covering 77 genes; in patient #3 (c), mutations were investigated in less-differentiated (CD19⁺CD81⁺) vs intermediate- (CD19⁻CD81⁺) and more-differentiated (CD19⁻CD81⁻) PC clones.

Figure 5

The sequential CpGs measured by HumanMethylation450 BeadChip for the CD81 gene. We investigated the expression levels of CD19 and CD81 in a large panel of MM cell lines (RPMI-8226, RPMI-LR5, NCI-H929, OPM-2, JJN3, MM1S, MM1R, MM144, U266, U266-DOX4, U266-LR7, SJR and MGG) and identified five cell lines positive for CD81 (RPMI-8226, RPMI-LR5, NCI-H929, OPM-2, JJN3) in the absence of CD19; all the others exhibited no expression for both CD19 and CD81 (data not shown). Afterward, under the hypothesis that loss of CD81 expression could be due to epigenetic regulation of the CD81 gene, we investigated the DNA methylation profile of CD81 in the NCI-H929, JJN3 and U266 cell lines (the first two positive for CD81 and the third negative). Accordingly, we observed an inverse correlation between DNA methylation levels in the CpG island region of the CD81 gene and the protein (antigen) expression level of CD81 in the three MM cell lines. Interestingly, the DNA methylation levels in the CpG island region were also inversely correlated with the DNA methylation levels in the gene body region of CD81. These results indicate that an epigenetic mechanism of DNA methylation plays an important role in the regulation of CD81 expression. The mean of the DNA methylation levels of the CpGs located in the CpG island or gene body region of CD81 are also shown.