Persistent polyclonal B-cell lymphocytosis: extensively proliferated CD27+IgM+IgD+ memory B cells with a distinctive immunophenotype

Persistent polyclonal B-cell lymphocytosis (PPBL) is a rare disorder, which mostly affects middle-aged women and is associated with smoking.1 It is characterized by the persistent expansion of CD27+IgM+IgD+ B cells, the presence of circulating binucleated lymphocytes and increased IgM serum levels.1, 2, 3 Furthermore, PPBL is associated with HLA-DRB1*07 carriership and a supernumerary isochromosome +i(3q) in a small proportion of lymphocytes.3 Although the disorder is usually benign,4 10% of patients present with mild splenomegaly and a few patients subsequently develop a mature B-cell malignancy, such as diffuse large B-cell lymphoma (DLBCL) or splenic marginal zone lymphoma (MZL).4 Despite an increasing number of described cases, the nature of the persistent lymphocytosis and its relation to mature B-cell malignancies remain poorly understood. Therefore, we subjected PPBL patients and healthy controls to immunoglobulin (Ig) repertoire and replication history analyses, and gene expression profiling (detailed Materials and Methods online).

Five patients (4 women, 1 man) were included with a moderate lymphocytosis of CD27+IgM+IgD+ B cells that was persistent for >1 year (Supplementary Table 1, Supplementary Figure 1a). These cells showed a normal Igκ/Igλ ratio, but abnormally low CD38 and CD21 expression levels (Supplementary Figure 1b). IGH-CDR3 immunospectratyping revealed polyclonal IGH rearrangements with a normal length distribution and in-frame selection (Supplementary Figure 1c). This was confirmed by detailed sequence analyses of 375 in-frame IGH gene rearrangements from sorted CD27+IgM+IgD+ cells. Similar to healthy controls, each clone yielded a unique sequence, and diverse IGHV, IGHD and IGHJ genes were used without signs for preferential pairing (Supplementary Figure 2). The IGHV1-2*04 allele that is associated with splenic MZL was used in only four rearrangements from three patients.5 Considering that 4/5 patients carried an IGHV1-2*04 allele in their germline DNA, it did not seem preferentially rearranged in PPBL cells. Finally, the IGH-CDR3 size distribution in IGH gene rearrangements of PPBL was similar to control natural effector and IgM-only memory B cells (Supplementary Figure 2f). Thus, PPBL cells showed a diverse Ig gene repertoire with normal IGH-CDR3 size selection and without obvious selection for IGHV gene use.

To study the nature of the lymphocytosis, we quantified the in vivo replication history with Ig kappa-deleting recombination excision circles.6 The observed proliferation of 16 cell divisions was significantly higher (P<0.05) than the replication histories of natural effector and IgM-only B cells of controls (Figure 1a). Despite the increased number of cell divisions, PPBL showed low levels of SHM, both in an Igκ CDR1 hotspot region and in IGHV genes (Figures 1b and c). These low SHM frequencies did not seem to result from impaired activity of activation induced cytidine deaminase or the error-prone Polymerase η, because targeting of RGYW/WRCY and WA/TW DNA sequence motifs in IGHV genes was normal (Supplementary Table 2).7 To study signs of antigenic selection, we determined the replacement/silent mutation ratios for framework (FR) and complementary determining regions (CDR), and compared these with the expected ratios using the BASELINe program.8 The FR of PPBL showed negative selection for replacement mutations similar to controls. In contrast, the CDR of PPBL did not show the positive selection of replacement mutations that was found in natural effector and IgM-only B cells from controls (Figure 1d). Thus, when compared with memory B cells from healthy controls, PPBL cells proliferated extensively, but showed unusually low SHM levels and limited signs of antigen-driven selection.

Figure 1

Molecular analysis of purified CD27+IgM+IgD+ PPBL cells. (a) Replication history in PPBL cells and control naive and IgM+ memory B cells, as measured with the kappa-deleting recombination excision circles assay.6, 9 Here and in panel (b), bars represent mean values with s.e.m., and dashed lines represent values for tonsil-derived centrocytes.6, 9 In two out of five analyzed patients, replication history of PPBL cells was above the detection limit of the assay. Differences between subsets were statistically analyzed with the Mann–Whitney test. *P<0.05; ****P<0.0001 (b) Frequency of mutated IGKV3-20 genes as measured with the IgκREHMA assay. (c) Frequency of mutated nucleotides in rearranged IGHV genes. This panel includes previously published data points for transitional, natural effector (73/148) and IgM-only (68/161) B-cell subsets.9 All individual data points are shown as gray dots with red lines indicating median values. (d) Selection of replacement mutations in rearranged IGHV genes calculated with the BASELINe program.8 Solid lines represent CD27+IgM+IgD+ cells of PPBL patients, and dashed lines represent IgM+ memory B-cell subsets of healthy controls. The red lines represent the selection strength for CDR and the blue lines for FR. (e) Supervised hierarchical clustering (complete linkage) of the gene expression profiles from PPBL cells (n=4), and control naive and IgM+ memory B-cell subsets (each subset: n=3) using 1-correlation as a distance measure based on the 643 (threshold at log2 value 4) probe sets that showed the most variation between any two samples. Here, and in panel f clustering analyses was performed without bias for known genes. (f) Unsupervised hierarchical clustering (complete linkage) of PPBL cells and control IgM+ memory B-cell subsets based on the 2342 (threshold 2.5) probe sets that showed the most variation between any two samples.

To obtain more insight into the origin and maturation status of PPBL cells, we performed genome-wide gene expression profiling in purified PPBL cells of four patients as well as five B-cell subsets from three healthy controls. Both normal and PPBL B-cell subsets showed high expression of genes encoding pan-B markers, including CD19, CD20, CD22, CD79A, CD79B and BAFF-R (Supplementary Table 3). The gene expression patterns for markers used to isolate the subsets (CD5, CD38, IgM and IgD) correlated well with protein expression. PPBL and control B-cell subsets were out of cell cycle as revealed by low MKI67 gene expression and absence of nuclear Ki67 protein (not shown). Furthermore, PPBL cells showed high expression of the genes encoding CD27, CD80, CD86, CD180 and TACI fitting with an activated memory B-cell profile.9 Supervised clustering of the subsets with the probe sets that showed the highest signal variation over all arrays (threshold 4; 643 probe sets) confirmed that PPBL B cells were more similar to control IgM+ memory B-cell subsets (distance, 0.25) than control naive B cells (distance, 0.55; Figure 1e). Still, PPBL B cells showed a distinct gene expression profile from control IgM+ memory B cells as revealed with unsupervised clustering (Figure 1f; Supplementary Table 4).

The expression levels of oncogenes that are typically involved in immature and mature B-cell malignancies showed large variation between individual donors of PPBL and control B-cell subsets, and no single oncogene was upregulated in all PPBL cases (Supplementary Table 3; Supplementary Figure 4). Two patients showed high BCL2 levels, but these were not caused by t(14,18), because these were only present in 1 out of 1.3 × 103 PPBL cells (Supplementary Figure 3).

We did not confirm previously observed deregulation of the TGFβ signaling pathway and the activator protein 1 complex.10, 11 Moreover, multiple genes encoding Fas signaling proteins were upregulated (Supplementary Figure 5). These findings do not support impaired apoptosis in PPBL. The discrepancy with previous studies is likely due to the fact that we analyzed purified B-cell subsets, providing a higher resolution to study differences in gene expression levels.

To identify novel immunophenotyping markers for PPBL cells, we selected eight differentially expressed genes between PPBL and control memory B cells that encoded membrane proteins and for which monoclonal antibodies were available (Supplementary Figure 6a). Of these, CD62L and CD73 expression levels were the most discriminative between PPBL and control memory B cells with flow cytometry (Supplementary Figure 6b). Although both markers showed a bimodal expression pattern on control memory B-cell subsets, their expression levels were uniformly low on all PPBL cells. Importantly, CD62L and CD73 were not downregulated in heavy smokers that were otherwise healthy (Figure 2a). Thus, the low levels of CD62L and CD73 seemed markers of PPBL rather than of heavy smoking.

Figure 2

Flow cytometric immunophenotyping of PPBL cells. (a) CD73 and CD62L surface expression levels on B cells of PPBL patients, and on IgM+ memory B-cell subsets of healthy non-smoking controls and otherwise healthy smokers. Each histogram represents merged data obtained for 4 PPBL patients, four smokers or four healthy donors. Gray shaded histograms represent isotype controls. (b) Principal component analysis of flow cytometric data performed with INFINICYT software.12 Analyzed subsets were separated by an automatic population separator (APS) based on the expression levels of IgD, CD38, CD73 and CD62L. 2D projection of the first and the third principal component provided the best visual separation between the three subsets. Each dot represents the mean value for the analyzed subset of an individual. The contributions of each of the markers to both the principal components are depicted with markers that contributed >10% indicated in bold. PC denotes principal component. (c) Cell surface expression levels of TLR1, TLR6 and TLR10 on PPBL cells, and CD27+IgM+IgD+ and CD27+IgM+IgD- B cells of healthy controls. (d) PPBL cells (purple) were immunophenotyped with the previously published B-CLPD antibody panel, analyzed with INFINICYT software and compared with mature B-cell malignancies.14 Maximum separation between PPBL and each of the analyzed malignancies was achieved with the APS, and visualized in 2D projections of the first and second principal components.12 Inner and outer continuous lines represent 1 and 2 s.d., respectively, of the indicated mature B-cell malignancy or PPBL. The following abbreviations are used: MZL (marginal zone lymphoma), MCL (mantle cell lymphoma), LPL (lymphoplasmacytoid lymphoma), DLBCL (diffuse large B-cell lymphoma), CLL (chronic lymphocytic leukemia), BL (Burkitt lymphoma), FL (follicular lymphoma), and HCL (hairy cell leukemia).

CD73 and CD62L could not be used independently to fully discriminate PPBL cells from control B-cell subsets. Therefore, we applied principal component analysis to study whether the combined information of CD62L, CD73, CD38 and IgD expression levels would be sufficient for discrimination of PPBL from controls.12 In a 2D projection, PPBL was separated from controls mainly by CD73 and CD62L (principal component 1), whereas control natural effector and IgM-only B cells were separated mainly by IgD and CD38 (principal component 3; Figure 2b). Concluding, we showed that PPBL cells and control IgM+ memory B cells can be separated based on a limited set of membrane markers: CD38, IgD and the newly identified CD62L and CD73.

MyD88/TIRAP/IRAK4 signaling, probably downstream of Toll-like receptor 10 (TLR10), has been shown to be critical for homeostatic maintenance of CD27+IgM+IgD+ B cells.13 PPBL cells showed normal TLR1, and TLR6 expression, and significantly upregulated TLR10 expression (Figure 2c, Supplementary Table 3, Supplementary Figure 6a). Moreover, MYD88, IRAK1 and TLR-signaling target gene TNF were upregulated in PPBL compared with control IgM+ memory B cells (Supplementary Table 3, Supplementary Figure 4). These increased levels of TLR10 and downstream signaling components could make PPBL B cells more receptive to exogenous signals and thereby contribute to the polyclonal expansion.

To study how closely the PPBL phenotype resembles that of a mature B-cell malignancy, we performed immunophenotyping in four PPBL cases with the EuroFlow B-CLPD antibody panel.14 Multiparameter analysis was subsequently performed to compare the PPBL profile with those of nine diagnostic subgroups of mature B-cell malignancies. Principal component analysis revealed close clustering of all PPBL samples, implying that this is a truly unique biological entity. PPBL cells clustered separately from Burkitt lymphoma, follicular lymphoma, CD10+ DLBCL, hairy cell leukemia and chronic lymphocytic leukemia. Still, in the automatic population separator view, the 1 s.d. line of PPBL overlapped with lymphoplasmacytoid lymphoma (LPL) and CD10-DLBCL, and to a lesser extent with MZL and mantle cell lymphoma. The separation between PPBL and LPL was based mainly on the lower expression of CD200 and higher expression of CD305 (LAIR-1) in PPBL. Separation between PPBL and CD10-DLBCL was mainly based on the high expression levels of CD79b and IgM in PPBL. In contrast to healthy controls, CD62L contributed little to separation between PPBL and any of the analyzed malignancies (<10%). Thus, discrimination of PPBL from mature B-cell malignancies is based on different markers than healthy controls and is possible with currently implemented EuroFlow immunophenotyping protocols.14 This may be of value to discriminate between PPBL and rare biclonal malignancies with a seemingly normal Igκ/Igλ ratio.15 Interestingly, in a few described cases, the diagnosis of PPBL was complicated by DLBCL and splenic MZL, but not any other hematological malignancy.4 These observations suggest an origin from the same cell compartment and support the hypothesis that PPBL might precede B-cell malignancy development in rare cases.4

In this study, we showed for the first time that CD27+IgM+IgD+ cells in PPBL have undergone extensive proliferation, presumably in lymphoid tissues, such as the spleen. The high Ig diversity with absence of stereotypes and the distinctive immunophenotype of PPBL are suggestive of a different pathophysiology than in mature B-cell malignancies. Still, individual cases show some deregulated oncogene and tumor-suppressor expression, and might be prone to develop into a malignancy. The newly identified cell surface markers can prove valuable for diagnosis of PPBL and discrimination from B-cell malignancies.


  1. 1

    Gordon DS, Jones BM, Browning SW, Spira TJ, Lawrence DN . Persistent polyclonal lymphocytosis of B lymphocytes. N Engl J Med 1982; 307: 232–236.

    CAS  Article  Google Scholar 

  2. 2

    Himmelmann A, Gautschi O, Nawrath M, Bolliger U, Fehr J, Stahel RA . Persistent polyclonal B-cell lymphocytosis is an expansion of functional IgD(+)CD27(+) memory B cells. Br J Haematol 2001; 114: 400–405.

    CAS  Article  Google Scholar 

  3. 3

    Troussard X, Cornet E, Lesesve JF, Kourel C, Mossafa H . Polyclonal B-cell lymphocytosis with binucleated lymphocytes (PPBL). Onco Targets Ther 2008; 1: 59–66.

    CAS  Article  PubMed Central  Google Scholar 

  4. 4

    Cornet E, Lesesve JF, Mossafa H, Sebahoun G, Levy V, Davi F et al. Long-term follow-up of 111 patients with persistent polyclonal B-cell lymphocytosis with binucleated lymphocytes. Leukemia 2009; 23: 419–422.

    CAS  Article  Google Scholar 

  5. 5

    Bikos V, Darzentas N, Hadzidimitriou A, Davis Z, Hockley S, Traverse-Glehen A et al. Over 30% of patients with splenic marginal zone lymphoma express the same immunoglobulin heavy variable gene: ontogenetic implications. Leukemia 2012; 26: 1638–1646.

    CAS  Article  Google Scholar 

  6. 6

    van Zelm MC, Szczepanski T, van der Burg M, van Dongen JJ . Replication history of B lymphocytes reveals homeostatic proliferation and extensive antigen-induced B cell expansion. J Exp Med 2007; 204: 645–655.

    CAS  Article  PubMed Central  Google Scholar 

  7. 7

    Longo NS, Lugar PL, Yavuz S, Zhang W, Krijger PH, Russ DE et al. Analysis of somatic hypermutation in X-linked hyper-IgM syndrome shows specific deficiencies in mutational targeting. Blood 2009; 113: 3706–3715.

    CAS  Article  PubMed Central  Google Scholar 

  8. 8

    Uduman M, Yaari G, Hershberg U, Stern JA, Shlomchik MJ, Kleinstein SH . Detecting selection in immunoglobulin sequences. Nucleic Acids Res 2011; 39 (Web Server issue): W499–W504.

    CAS  Article  PubMed Central  Google Scholar 

  9. 9

    Berkowska MA, Driessen GJ, Bikos V, Grosserichter-Wagener C, Stamatopoulos K, Cerutti A et al. Human memory B cells originate from three distinct germinal center-dependent and -independent maturation pathways. Blood 2011; 118: 2150–2158.

    CAS  Article  PubMed Central  Google Scholar 

  10. 10

    Roussel M, Roue G, Sola B, Mossafa H, Troussard X . Dysfunction of the Fas apoptotic signaling pathway in persistent polyclonal B-cell lymphocytosis. Haematologica 2003; 88: 239–240.

    CAS  PubMed  Google Scholar 

  11. 11

    Lawrie CH, Shilling R, Troussard X, Cattan H, Mossafa H, Pezzella F et al. Expression profiling of persistent polyclonal B-cell lymphocytosis suggests constitutive expression of the AP-1 transcription complex and downregulation of Fas-apoptotic and TGFbeta signalling pathways. Leukemia 2009; 23: 581–583.

    CAS  Article  Google Scholar 

  12. 12

    Pedreira CE, Costa ES, Lecrevisse Q, van Dongen JJ, Orfao A . On behalf of the EuroFlow C. Overview of clinical flow cytometry data analysis: recent advances and future challenges. Trends Biotechnol 2013; 31: 415–425.

    CAS  Article  Google Scholar 

  13. 13

    Weller S, Bonnet M, Delagreverie H, Israel L, Chrabieh M, Marodi L et al. IgM+IgD+CD27+ B cells are markedly reduced in IRAK-4-, MyD88- and TIRAP- but not UNC-93B-deficient patients. Blood 2012; 120: 4992–5001.

    CAS  Article  PubMed Central  Google Scholar 

  14. 14

    van Dongen JJ, Lhermitte L, Bottcher S, Almeida J, van der Velden VH, Flores-Montero J et al. EuroFlow antibody panels for standardized n-dimensional flow cytometric immunophenotyping of normal, reactive and malignant leukocytes. Leukemia 2012; 26: 1908–1975.

    CAS  Article  PubMed Central  Google Scholar 

  15. 15

    Hsi ED, Hoeltge G, Tubbs RR . Biclonal chronic lymphocytic leukemia. Am J Clin Pathol 2000; 113: 798–804.

    CAS  Article  PubMed Central  Google Scholar 

Download references


We thank I Pico and AA Tarique for technical support, S de Bruin-Versteeg for assistance with preparing the figures, Drs AW Langerak, VHJ van der Velden and M van der Burg for fruitful discussions and critical reading of the manuscript and the EuroFlow Consortium for kindly providing the EuroFlow tools and the data on mature B-cell malignancies. This work was supported by a grant from the Erasmus University Rotterdam (EUR-Fellowship to MCvZ).

Author information



Corresponding author

Correspondence to M C van Zelm.

Ethics declarations

Competing interests

The authors declare no conflict of interest.

Additional information

Author Contributions

MAB, JJMvD and MCvZ designed the experiments. CG-W, HJA and DdR provided conceptual advice. MAB and CG-W performed and analyzed most of the experiments. DdR and MCvZ contributed to data analyses. HJA, KPM-O, HJA, SB, AO provided material necessary for performing experiments and data analysis. MAB and MCvZ wrote the manuscript. CG-W, HJA, DdR, KPM-O, HJA, SB, AO and JJMvD commented on the manuscript.

Supplementary Information accompanies this paper on the Leukemia website

Supplementary information

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Berkowska, M., Grosserichter-Wagener, C., Adriaansen, H. et al. Persistent polyclonal B-cell lymphocytosis: extensively proliferated CD27+IgM+IgD+ memory B cells with a distinctive immunophenotype. Leukemia 28, 1560–1564 (2014).

Download citation

Further reading