Circulating clonotypic B cells in multiple myeloma and monoclonal gammopathy of undetermined significance

Abstract

The B-cell compartment in which multiple myeloma stem cells reside remains unclear. We investigated the potential presence of mature, surface-membrane immunoglobulin-positive B lymphocytes clonally related to the tumor bone marrow plasma cells among different subsets of peripheral blood B cells from ten patients (7 with multiple myeloma and 3 with monoclonal gammopathies of undetermined significance). The presence of clonotypic immunoglobulin heavy chain gene rearrangements was determined in multiple highly-purified fractions of peripheral blood B-lymphocytes including surface-membrane IgM⁺ CD27⁻ naïve B-lymphocytes, plus surface-membrane IgG⁺, IgA⁺ and IgM⁺ memory CD27⁺ B cells, and normal circulating plasma cells, in addition to (mono)clonal plasma cells, by a highly-specific and sensitive allele-specific oligonucleotide polymerase chain reaction directed to the CDR3 sequence of the rearranged IGH gene of tumor plasma cells from individual patients. Our results showed systematic absence of clonotypic rearrangements in all the different B-cell subsets analyzed, including M-component isotype-matched memory B-lymphocytes, at frequencies <0.03 cells/μL (range: 0.0003–0.08 cells/μL); the only exception were the myeloma plasma cells detected and purified from the peripheral blood of four of the seven myeloma patients. These results indicate that circulating B cells from patients with multiple myeloma and monoclonal gammopathies of undetermined significance are usually devoid of clonotypic B cells while the presence of immunophenotypically aberrant myeloma plasma cells in peripheral blood of myeloma patients is a relatively frequent finding.

Introduction

A major challenge in the pathogenesis of both multiple myeloma (MM) and monoclonal gammopathy of undetermined significance (MGUS) has been the identification of the cell of origin of both neoplasias. Monoclonal plasma cells (MPC) from patients with MM usually show a low proliferative capacity as assessed by labeling indices of <0.5%;1 this finding supports the notion that pre-PC compartments may contain the proliferative M-PC progenitor. In line with this hypothesis, circulating peripheral blood cells clonally identical to bone marrow M-PC have been recurrently detected in MM.72 Apart from the presence of the malignant IGH gene rearrangement, these peripheral blood cells have also been found in some patients to carry karyotypic alterations such as trisomy 11, del(17p)-, and oncogenic proteins (e.g. IGH-FGFR3, IGH-MMSET) expressed by CD138 M-PC.86 Moreover, intraclonal heterogeneity, as assessed through the analysis of the IGHV gene sequence9 or through the investigation of cytogenetic alterations10 and gene mutations evaluated by whole exome sequencing,11 has been reported within the M-PC population, suggesting that the tumor PC compartment could be continuously repopulated by more than one stem cell.

There is a growing body of evidence showing the presence of circulating aberrant PC not only in patients with MM but also in those with MGUS,1612 and recent studies have demonstrated the presence of cells sharing stem cell properties within the compartment of M-PC.17 However, the question that remains to be answered is whether or not clonotypic cells are also identifiable in peripheral blood B-cell subsets, which would be the “earliest clonotypic cells”. Clonotypic cells were first identified in a sorted CD19 fraction2 and the pattern of somatic hypermutation in the VH regions of the IGH gene suggested that the malignant cell had passed through the germinal center.18 This was further supported by the detection of peripheral blood clonotypic memory B cells in MM patients62 and the finding that in vitro clonogenicity was reduced when memory B cells were removed.19 In contrast, screening for cells with a pre-PC phenotype in cell lines,21194 and analysis of the engraftment of peripheral blood B cells from MM patients into immunodeficient mice2322193 have provided inconsistent findings. In parallel, depletion of B cells by rituximab therapy has not shown clear beneficial clinical effects in MM, except in a small fraction of patients with CD20 PC,2524 and evaluation of the clonal hierarchy in light chain-secreting myeloma did not detect clonotypic circulating cells, except in one patient showing peripheral blood infiltration by malignant PC.26 One explanation for these discrepant results is a simple technical pitfall: the B cells analyzed could be contaminated in some cases by circulating PC. In fact, to the best of our knowledge, no highly sensitive molecular analysis of different compartments of highly purified circulating B cells specifically devoid of contaminating circulating M-PC has been performed so far, to confirm or rule out the presence of clonotypic B cells in MM and MGUS.

Here, we investigated the presence of circulating B cells which would be clonally related to the M-PC in patients with MM and MGUS, using an allele-specific oligonucleotide polymerase chain reaction (ASO-PCR) technique aimed at sensitive and specific detection of IGHVDJ monoclonal gene sequences unique to the tumor cells of individual patients. This analysis was performed in multiple, different, highly-purified, maturation-associated subsets of peripheral blood B cells (_SmIgM/CD27 naïve, isotype non-switched _SmIgM+ and isotype-switched _SmIgG and _SmIgA memory B-lymphocytes, and PC) from a group of seven patients with MM and three with MGUS.

Methods

Patients and samples

Ten patients (7 with MM and 3 with MGUS) diagnosed with monoclonal gammopathies27 were studied. In all cases, paired bone marrow and peripheral blood samples were collected at diagnosis for further immunophenotypic characterization, isolation, and molecular analysis of M-PC and other B-cell subsets, after informed consent was given by each individual. The study was approved by the Ethics Committee of the Cancer Research Center of the University of Salamanca, (Salamanca, Spain) and it followed the Helsinki Declaration protocol. The most relevant clinical and biological characteristics of the patients studied are summarized in Online Supplementary Table S1.

Multiparameter flow cytometry immunophenotypic studies and cell purification

Multiparameter flow cytometry immunophenotypic studies were performed as recently described3128 (and detailed in the Online Supplementary Methods) in both bone marrow and peripheral blood samples. Fluorescence-activated cell sorting (FACS) was used to purify bone marrow M-PC as well as peripheral blood B-cell subsets and normal PC (N-PC); whenever present, circulating M-PC were also purified. The purity of the sorted M-PC and the multiple peripheral blood normal B-cell subsets was systematically ≥98%, and only those peripheral blood B-cell fractions which showed no contamination (<0.01%) by circulating peripheral blood M-PC were included in the study; all other fractions were not tested (Table 1).

Table 1.Frequency and absolute counts of clonotypic cells detected in different B-cell and PC subsets present in the peripheral blood of patients with MGUS (n=3) or MM (n=7).

Analysis of IGH gene rearrangements of bone marrow monoclonal plasma cells and CDR3 sequencing

The malignant IGH gene rearrangement was identified in genomic DNA according to the BIOMED-2 protocol3332 slightly modified to amplify the specific monoclonal complete VH-JH and incomplete DH-JH. PCR amplified products were identified by high-resolution capillary electrophoresis in an automated ABI PRISM 3130 Avant sequencer, using GENEMAPPER 3.1 software (AB). Clonal PCR products were purified, sequenced and identified as previously reported.34 Once the segments had been identified, the N-region was highlighted for ASO-primer design due to its high specificity for each individual rearrangement, as previously described.35

A qualitative ASO-PCR with the CDR3 specific primer and its respective 6-FAM-labeled IGH JH primer, was carried out on purified bone marrow M-PC from each patient; purified peripheral blood B cells from healthy donors were studied in parallel as negative controls. If a single (clonal) peak was detected in a sample, it was sequenced to confirm that the sets of primers were really specific. In order to assess the sensitivity of each patient-specific set of primers used in the ASO-PCR assay (e.g.: the lowest number of cells detected), ASO-PCR analyses were carried out in multiple tubes containing serial dilutions of one or more previously purified bone marrow M-PC from individual patients (see Online Supplementary Methods for a more detailed description).

Allele-specific oligonucleotide polymerase chain reaction analysis of purified peripheral blood subpopulations of B cells and plasma cells

DNA was extracted from pre-defined numbers of highly-purified peripheral blood B-cell and PC subsets. ASO-PCR conditions were as described above; all purified cells from each cell compartment from each patient were analyzed by ASO-PCR (7 to 10 ASO-PCR replicates/patient). Clonotypic B cells were defined as being present within a specific B-cell or PC compartment when two or more replicates were found to be positive. For every sample, DNA quality was evaluated by a control according to the BIOMED-2 protocols.32

Results

Distribution of different subsets of B-lymphocytes and plasma cells in the peripheral blood of patients with multiple myeloma and monoclonal gammopathy or undetermined significance

The following subsets of peripheral blood B cells and PC were identified:3129 naïve (CD10 CD19 CD20 CD27 CD38) and memory (CD10 CD19 CD20 CD27 CD38) B-lymphocytes plus normal circulating PC (CD 10 CD19 CD20 CD27 CD38). Peripheral blood memory B-lymphocytes were further subdivided into switched SmIgA, _SmIgG, and non-switched _SmIgM cells. In four of the seven MM patients (cases #3, #5, #6 and #7), peripheral blood circulating M-PC with restricted CyIgλ or CyIgκ expression were also identified as clearly different from N-PC based on differential expression of CD38, CD19, CD45, and light scatter.2926

Overall, no statistically significant differences (P>0.05, Mann-Whitney U test) were observed in the distribution of circulating B cells and PC between MM and MGUS patients (Figure 1). This was the case for total B cells (113±15 versus 83±71 cells/μL), naïve (83±11 versus 53±56 cells/μL) as well as total memory B cells (30±10 versus 22±16 cells/μL) and their _SmIgM+ (9±4 versus 9±6 cells/μL), _SmIgG (14±2 versus 4±2 cells/μL), and _SmIgA+ (12±1 versus 6±4 cells/μL) subsets, and also N-PC (0.7±0.4 versus 1.2±1.5 cells/μL) (Figure 1). In contrast, M-PC were found at frequencies >8.0 cells/μL (range: 0.08–29.8 cells/μL) in four of the seven MM patients, but not in the three MGUS cases (although we had detected them in up to 21% of a larger series of MGUS cases16) (P<0.05; Figure 1). Of note, all FACS-purified peripheral blood B-cell subsets (naïve and memory B-lymphocytes) and N-PC showed a normal protein expression profile, including a normal polyclonal Igκ versus Igλ light chain ratio, while the M-PC compartment systematically showed an aberrant immunophenotype associated with monoclonal light chain restriction (either CyIgκ or CyIgλ) profile.

Figure 1.Distribution of B-lymphocytes and plasma cells and their subsets in the peripheral blood of patients with MGUS (n=3, white boxes) and MM (n=7; black boxes). N-PC: normal plasma cells; M-PC: (mono)clonal plasma cells. Boxes extend from the 25^th to the 75^th percentiles; the line in the middle and the vertical lines represent median values and both the 10^th and 90^th percentiles, respectively. The Mann–Whitney U test was used to estimate the statistical significance of differences observed. *P<0.05 vs. MGUS.

Sensitivity of the molecular assays

The sensitivity assays performed were aimed at defining the minimum number of clonotypic M-PC required for a positive result and at establishing the sensitivity of the ASO-PCR assay over a polyclonal background. The first set of experiments consistently showed the ability to systematically detect a (clonal) patient-specific IGH gene rearrangement peak whenever ≥5 clonotypic M-PC were present in the sample, for all ten patients analyzed; in three of the seven patients tested (cases #MM1, #MM2, #MGUS3) the sensitivity was even higher since patient-specific IGH CDR3 gene rearrangements were detected once a single MPC was present in the sample. Regarding the ability to detect clonotypic M-PC over a background of polyclonal B cells, sensitivity from dilution experiments was systematically <10, ranging between <10 to <10 (from 0.2% to 0.001% M-PC cells/total B cells) (Online Supplementary Table S2), as illustrated in Figure 3 for one of the cases studied.

The sensitivity of the assay varied per patient for the different cell fractions from <0.08 to <0.0003 cells/μL of peripheral blood (i.e. <80 to <0.3 cells/μL), according to: the sensitivity and reproducibility of the assay; the number of purified peripheral blood B cells and PC analyzed; and the distribution of the subpopulation (e.g. absolute count) in the peripheral blood of each individual patient at the moment the sample was collected (Table 1).

Frequency of clonotypic cells in highly-purified peripheral blood subsets of circulating B cells and plasma cells

The search for the presence of clonotypic cells was performed on highly purified peripheral blood naïve B-lymphocytes, switched _SmIgA and _SmIgG, as well as non-switched _SmIgM memory B cells and N-PC, present at variable numbers in the peripheral blood of each of the ten patients with monoclonal gammopathies (Table 1). In addition, circulating M-PC from four MM patients in whom these cells were detected at frequencies of 0.79%, 0.09%, 0.03% and 0.17% of all peripheral blood nucleated cells (absolute M-PC count of 29.8, 0.7, 0.08, and 6.1 cells/μL, respectively) were also purified and analyzed. Overall, no cells carrying clonotypic IGH gene rearrangements were detected among the phenotypically normal B-cell subsets or N-PC from any of the MM (0/7) or MGUS (0/3) patients analyzed (Table 1). For Ig-switched memory B cells sharing the same Ig heavy chain isotype as bone marrow M-PC, a sensitivity of between <6 cells/mL and <24 cells/mL of peripheral blood (IgG MM) and of <9 cells/mL (IgA MM) of peripheral blood was reached. As expected, analysis of peripheral blood M-PC confirmed the presence of clonotypic IGH gene rearrangements identical to those identified in the corresponding (paired) bone marrow M-PC, in all replicates of purified M-PC analyzed (Table 1 and Figure 2).

Figure 2.Example of the results of molecular analysis of clonotypic B cells and PC in purified B-cell and PC subsets from the peripheral blood (PB) of a MM patient (case #MM3). As shown, circulating (mono)clonal plasma cells (M-PC) are the only compartment of PB B cells from this patient that showed a clonal CDR3 gene rearrangement signature similar to that of bone marrow M-PC from the same patient (A). The quality of the DNA samples from all FACS-purified PB B-cell and PC subsets of each individual patient was analyzed in parallel for a control gene with PCR primers directed to the human thromboxane synthase gene (TBXAS1) (B).

Figure 3.Example of a dilution experiment performed and analyzed by GeneScanning for one representative MM patient (case #MM6). To evaluate the sensitivity of the ASO-PCR assay for this patient, DNA from M-PC obtained from the bone marrow (BM) was diluted into polyclonal DNA obtained from FACS-purified peripheral blood (PB) B cells from a healthy donor. Numbers between brackets indicate the M-PC:normal B-cell dilution factor.

Discussion

Since the pioneering experiments which used anti-idiotypic antibodies to track the oncogenic event in MM throughout B-cell ontogeny,3736 controversial results have been reported in the literature regarding the B-cell compartment in which the potential myeloma stem cell resides. For more than a decade now, it has been generally accepted that circulating CD19 B-lymphocytes from MM patients could contain cells that carry clonotypic rearrangements identical to those of MM-PC.72 However, the frequency and nature of such clonally-related cells remains a matter of debate. While some groups suggested that the majority of circulating B-lymphocytes would belong to the malignant clone,32 others could not confirm these preliminary observations and reported that, if present, circulating clonotypic B cells would constitute a minor population of all peripheral blood leukocytes.3938

Important advances have been achieved in molecular as well as in cell purification techniques (e.g. improved exclusion of cell doublets and sorting efficiency, standard usage of ≥8-color experiments for better definition of the cell subpopulations of interest and exclusion of unwanted cells). Based on such technological advances we decided to revisit in the present study the detection of clonally related B cells in the peripheral blood of patients with monoclonal gammopathies. For this purpose we used a highly-sensitive approach. First, an eight-color multiparameter flow cytometry technique was used for the identification of the distinct subsets of circulating peripheral blood B cells and PC, in combination with multiparameter sorting of highly purified, non-contaminated, cell fractions corresponding to such subsets of peripheral blood naïve and memory B-lymphocytes (including _SmIgM, _SmIgA and _SmIgG memory B cells), in addition to N-PC and, whenever present, also M-PC. In a second step, molecular detection of clonotypic B cells was performed on DNA from the purified B-cell subsets and PC using a highly-sensitive ASO-PCR technique specifically designed for the detection of clonotypic B cells of individual patients.

Our results showed systematic absence of clonotypic IGH gene rearrangements in all different B-cell subsets analyzed, including M-component isotype-matched memory B-lymphocytes; the only exception were the M-PC detected and purified from the peripheral blood of four MM patients. In line with these observations, no altered patterns of protein expression were found in any of the purified fractions of naïve and memory B-lymphocytes or N-PC, all such cell subsets also showing a normal polyclonal Igκ versus Igλ light chain ratio, except M-PC.

Considering the sensitivity of the assay used, the number of cells analyzed and the volume of blood investigated, it seems unlikely that circulating clonotypic B-lymphocytes are present in MM and MGUS patients at frequencies between >6 and 30 cells/mL (<1 cell in between 33 and 166 μL of peripheral blood). Even more, when we specifically consider the absence of clonotypic B cells among the compartment of memory B cells (as the most susceptible to containing the myeloma stem cell) the frequency would go down to between <4.1 and <37 cells/mL.

Overall, the sensitivity of the approach used here is far beyond that of the techniques used in most previously reported studies in which the presence of clonotypic B-lymphocytes was reported. In this regard, the discrepancies observed between our and other studies cannot be due to selection of patients12 since almost all our MM patients had symptomatic MM and they had not received prior therapy for their disease. We did confirm the presence of circulating M-PC at relatively high numbers in the peripheral blood of the majority of MM patients (range: 80 to 29,800 cells/mL), in line with previous observations by our and other groups.1612 This frequency is lower than that reported in earlier studies for circulating clonotypic B cells (range: 10,000 to 610,000 cells/mL).3 It is worth noting that, in a substantial number of these previous studies, total peripheral blood B cells were used to search for clonotypic IGH gene rearrangements and contamination by M-PC (e.g. due to doublet formation or altered time delay in sorting experiments) was not systematically excluded. In this regard, we have recently confirmed contamination of peripheral blood B cells by M-PC through re-analysis of a frozen sample from a patient who had been previously reported to carry clonotypic non-switched B cells.26 This could contribute to explain, at least in part, the discrepancies observed between our and other studies, even when our approach dramatically increased the sensitivity and specificity of the analysis. In contrast, our results fully support recent findings by Kim et al.21 in SCID mice previously engrafted subcutaneously with fetal human bone, showing that only the MM-PC compartment was able to give rise to clonally-related, patient-specific MM cells in (two) serial transplants; although plasmablasts and B cells were also able to engraft in those mice, their capacity was limited to the first xeno-transplantation and these compartments could not give rise to any cell with the same patient-specific MM molecular signature.21 Alternatively, it is possible that the tumor precursor would circulate only at the earlier stages of disease to fill/feed the bone marrow niches, giving rise to tumor PC. However, this seems unlikely since different peripheral blood B-cell compartments from MGUS patients analyzed here were also found to be devoid of clonotypic cells with the same IGH gene rearrangement as bone marrow M-PC.

Altogether, our results suggest that if peripheral blood clonal B-lymphocytes exist in MM and MGUS, they are present at extremely low frequencies in most patients, and would not explain the dissemination throughout the bone marrow during active/progressive disease, as observed in many symptomatic MM patients. Conversely, since circulating M-PC were frequently detected at rather higher numbers, it could be hypothesized that the myeloma clonogenic cell could be part of the malignant PC compartment which is able to re-circulate and spread the disease. In line with this latter hypothesis recent findings indicate that circulating peripheral blood M-PC from both MM and MGUS patients display a slightly more immature immunophenotype with lower expression of CD138,16 a phenotypic feature of M-PC that has been associated with increased clonogenic potential and other myeloma stem cell features.194 Furthermore, circulating peripheral blood M-PC are also frequently detected in early phases of the disease including MGUS and smoldering MM.1612 Further studies, in which the clonogenic potential of peripheral blood M-PC is compared with that of bone marrow tumor cells, are required to validate this hypothesis.

In summary, based on a sensitive and well-controlled approach, in this study we failed to demonstrate the presence of circulating clonotypic B cells in the peripheral blood of patients with MM or MGUS, outside the compartment of immunophenotypically aberrant clonal PC detected in a substantial fraction of MM patients. These findings suggest that tumor dissemination in monoclonal gammopathies is more likely to be related to the peripheral blood compartment of circulating M-PC than to other populations of less differentiated naïve and memory B-lymphocytes.

Footnotes

• The online version of this article has a Supplementary Appenix.
• Funding This work was supported by grants from European Union FP6 STREP MSCNet (N. E06005FF), Cooperative Research Thematic Network on Cancer (RTICs; RTICC RD06/0020/0035-FEDER, RD06/0020/0006, RD12/0036/0048, RD12/0036/0069 and G03/136), Instituto de Salud Carlos III/Subdirección General de Investigación Sanitaria Ministerio de Sanidad y Consumo (FIS: PI060339; 02/0905; 01/0089/01-02;PS09/01897, and PI06/0824-FEDER), Asociacion Española Contra el Cancer-AECC (GCB120981SAN) and Gerencia Regional de Salud de Castilla y León; Ayuda de Excelencia de Castilla y León, Consejería de Educación (EDU/894/2009, GR37) Junta de Castilla y León, Valladolid, Spain. LST received a CAPES/Ministério da Educação scholarship from the Brazilian Government.
• Authorship and Disclosures Information on authorship, contributions, and financial & other disclosures was provided by the authors and is available with the online version of this article at www.haematologica.org.

• Received June 6, 2013.
• Accepted July 17, 2013.