Journal of Clinical Immunology

, Volume 32, Issue 4, pp 729–735

Persistence of a Large Population of Exhausted Monoclonal B cells in Mixed Cryoglobuliemia After the Eradication of Hepatitis C Virus Infection

  • Marcella Visentini
  • Valentina Conti
  • Maria Cagliuso
  • Giulia Siciliano
  • Carolina Scagnolari
  • Milvia Casato
  • Massimo Fiorilli

DOI: 10.1007/s10875-012-9677-0

Cite this article as:
Visentini, M., Conti, V., Cagliuso, M. et al. J Clin Immunol (2012) 32: 729. doi:10.1007/s10875-012-9677-0



Functionally exhausted and mostly autoreactive B-cells with a peculiar CD21lowCD11c+ phenotype accumulate in several human immunological disorders including common variable immunodeficiency, HIV infection and rheumatoid arthritis. In HCV-associated mixed cryoglobulinemia (MC) there is accumulation of exhausted clonal B cells expressing a VH1-69-encoded cross-reactive idiotype; these cells are phenotypically heterogeneous, displaying either a CD21lowCD11c+ or a marginal zone (MZ)-like (IgM+CD27+CD21+CD11c-) phenotype. Irrespective of their phenotype, VH1-69+ B-cells are unresponsive to the stimulation of Toll-like receptor 9 (TLR9). We investigated the fate of these cells after the eradication of HCV.


Fourteen MC patients were studied before and after antiviral therapy. VH1-69+ B-cells were identified using the G6 monoclonal antibody and their phenotype and responsiveness to the stimulation of TLR9 were investigated.


In seven virological non-responders, cryoglobulin levels and the number and phenotype of VH1-69+ B cells remained substantially unchanged. By contrast, in sustained viral responders cryoglobulinemia subsided and the number of VH1-69+ B cells declined. However, high proportions of MZ-like VH1-69+ B cells retaining unresponsiveness to TLR9 stimulation persisted for several months in these patients.


Clonal expansion of CD21low VH1-69+ B cells may depend on continual stimulation by HCV, whereas their MZ-like counterparts may persist for years after the eradication of infection. Prolonged survival of exhausted MZ-like B cells after withdrawal of the initial inciting stimulus may contribute to the accumulation of autoreactive B cells in immunological disorders.


B-cells exhaustion HCV cryoglobulinemia Toll-like receptor VH1-69 


Type 2 mixed cryoglobulinemia (MC) is a benign monoclonal B cell lymphoproliferative disorder caused, in the large majority of cases, by chronic infection with hepatitis C virus (HCV) [1]. Clonal B cells of MC commonly secrete a natural antibody bearing an idiotype, encoded by the VH1-69 variable gene, which has rheumatoid factor activity and cross-reacts with the E2 protein of HCV [1]. The VH1-69-expressing B cells of MC patients can be subdivided into two major populations [2, 3, 4, 5], one with the phenotype of normal human marginal zone (MZ) B cells (IgM+IgD+CD27+CD21+), and one with the phenotype of the CD21low innate-like B cells that are expanded in common variable immunodeficiency [6, 7, 8], in rheumatoid arthritis [7] and in human immunodeficiency virus infection [9].

The CD21low B cells expanded in HIV infection and in common variable immunodeficiency are functionally exhausted, as they fail to proliferate in response to the stimulation of the B cell receptor (BCR) and of CD40 as well as of the innate Toll-like receptors (TLR) 9 and 7 [6, 7, 8, 9]. Recently, the CD21low VH1-69-expressing B cells expanded in HCV-associated MC have been reported to be functionally anergic, and it was suggested that their counterparts could be normally functioning [3, 4]. However, we found that also the MZ-like CD21high VH1-69+ B cells of MC patients are also functionally exhausted and fail to proliferate in response to TLR ligands [5]. These observations are consistent with the view that both the expansion and the exhaustion of VH1-69+ B cells are caused by chronic antigenic stimulation by HCV and that MZ-like cells could be the precursors of CD21low cells [5]. To untangle the role of continual antigenic stimulation on the maintenance of exhausted VH1-69+ B cells of MC patients, we investigated the fate of these cells the eradication of HCV by antiviral therapy.

Materials and Methods


A total of 14 patients (10 female, 4 male, median age 59,5 range 38–78 years) with HCV-associated MC were studied. All patients presented clinical signs of small vessel vasculitis. The cryocrit values ranged from 3% to 50% (median 8.5%). All patients were treated with pegylated interferon-α2b (Peg-IFN; Pegintron, Schering-Plough, Kenilworth, NJ, USA) 1.5 μg/kg weekly, plus ribavirin 800 mg or 1,000 mg daily. Duration of treatment ranged from 6 to 24 months, according to HCV genotype and to previous suggestions for the treatment of patients with refractory MC [10]. Sustained virological response (SVR) was defined as persistently undetectable HCV RNA by qualitative real-time PCR throughout the follow-up. All subjects provided informed consent, in accordance with the Institutional Review Board of the Sapienza University of Rome and with the Declaration of Helsinki.


B cell immunophenotyping and the identification of VH1-69-expressing B cells using the G6 monoclonal antibody (kindly provided by Roy Jefferis, Birmingham, UK) were done by flow cytometry as previously described [2, 5]. The proliferative responses of VH1-69+ and VH1-69- B cells to the stimulation of TLR9 with CpG (Sigma Genosys; 2.5 μg/ml) were measured by the CFSE (Invitrogen) dilution method, as previously described [5]. Briefly, at day 5 of culture cells were permeabilized and stained with antibodies to CD20, IgM and VH1-69, and the VH1-69+ and VH1-69- B cells were electronically gated. The number of cells entering proliferation (percent of divided cells) and the mean number of cell divisions performed by proliferating cells (proliferation index) [11] were calculated using the FlowJo (Tree Star, Inc.) software.

The HCV RNA load in serum was determined by qualitative real-time PCR (COBAS® Amplicor HCV Test, v2.0; Roche) with a limit of detection of 50 IU/mL. Blood samples were processed at 37°C to avoid the possible loss of HCV virions caused by co-precipitation with cryoglobulins.

For determining the presence of HCV in peripheral blood mononuclear cells (PBMC), RNA was extracted from cells using Trizol (Invitrogen) and reverse transcribed; 5 μL of cDNA were used as a template for amplification with NS5B sense (5′ TGATACCCGCTGYTTTGACTC3′) and antisense (5′ GTACCTGGTCATAGCCTCCGTG3′) primers. The amplification products were run on 2% agarose gel. The detection limit of the assay was 100 HCV RNA copies/106 cells.


Comparisons were done with the Wilcoxon paired differences test. A p value less than 0.05 was considered significant.


Patients’ Characteristics and Response to Therapy

We correlated the virological and clinical outcomes with changes of the phenotype of circulating B cells in 14 MC patients, investigated before antiviral therapy and 22 to 41 (median 27) months after its withdrawal except one patient who was studied one month after a 24-mo course of therapy. Seven patients were SVR and 7 were non-responders; there were no appreciable differences between the two groups as to age, sex, pre-therapy viral load, HCV genotype distribution or clinical signs of vasculitis. Clinical signs of vasculitis regressed and the cryocrit values decreased significantly in all SVR but not in non-responders (Fig. 1a).
Fig. 1

Changes of cryocrit and of circulating VH1-69+ B cells in MC patients after antiviral therapy. a cryocrit values, percentages (among B cells) and absolute numbers (μL) of VH1-69+ B cells in sustained virological responders (SVR) and non-responders (NR) before therapy (BT) and after therapy (AT). Asterisks denote statistical significance: * p < 0.01; ** p < 0.001. b changes of the cryocrit level (closed circles) and of the absolute number of circulating VH1-69+ B cells (open squares) during and after antiviral therapy (shaded area) in patients 1, 2 and 3. The symbols (+) and (−) denote, respectively, positive or negative HCV-RNA in serum at the indicated time-points

B cell immunophenotyping was done in all patients before and after antiviral therapy but, except in three patients, proliferative responses to CpG were studied only after treatment, because pre-therapy viable frozen PBMCs were not available. The three patients in whom both B cell phenotype and proliferation studies were prospectively investigated (henceforth indicated as patients 1 to 3) were two SVR and one non-responder; their main immunological and virological findings by are summarized in Fig. 1b.

Patient 1 was a 43-year old female infected with HCV strain 3a (1.86 × 106 copies/mL) and received a 12-month course of therapy; patient 2 was a 70-year old male infected with strain 2a2c (2.14 × 106 copies/mL), and was treated for 6 months; patient 3 was a 32-year old female infected with strain 3 (0.55 × 106 copies/mL) and, because of her extremely severe clinical manifestations, was treated with rituximab 375 mg/m2 weekly for four weeks in addition to a 24-month course of antivirals. HCV-RNA in serum became negative at month 2 of therapy in patient 1, at month 5 in patient 2 and at month 3 in patient 3. Serum HCV-RNA remained negative throughout the follow-up in patients 1 and 2; in patient 3 HCV RNA was undetectable during antiviral therapy and immediately thereafter, but viremia relapsed 8 months after the completion of therapy. HCV-RNA sequences in PBMC were searched at months 3 and 12 after the completion of therapy in patient 1, at month 6 in patient 2, and at month 24 after beginning of therapy (last month of treatment) in patient 3, and tested negative in all instances. Cryoglobulins decreased steadily after the beginning of therapy in patient 1 and became negative 20 months after the completion of treatment; in patient 2, cryoglobulins decreased more rapidly and became negative 6 months after therapy. These patients remain aviremic and without cryoglobulinemia 34 and 37 months after the completion of therapy, respectively. In patient 3 cryoglobulins decreased rapidly with therapy, but cryoglobulinemia recurred at month 12 of treatment.

Changes of VH1-69+ B Cells After Eradication of HCV

Before antiviral treatment there were no significant differences in the percentages and absolute numbers of circulating VH1-69+ B cells in sustained virological responders and in non-responders (Fig. 1a). VH1-69+ B cells had, before treatment, a heterogeneous phenotype, with CD21low cells representing 30% to 79% (median 60%) of VH1-69+ B cells. After antiviral therapy, the percentage and the absolute number of VH1-69+ B cells declined in patients with SVR, but not in virological non-responders (Fig. 1a). Nevertheless, in virological responders both the absolute number and the percentage of VH1-69+ B cells of SVR remained, on average, above the values observed in healthy donors (less than 7/μL and less than 3% of B cells) for several months after the eradication of HCV infection. In virological non-responders the relative proportions of CD21low and MZ-like VH1-69+ B cells remained substantially unchanged after treatment while, interestingly, the VH1-69+ B cells persisting in patients with SVR were almost completely represented by MZ-like B cells (Fig. 1b).

Changes of cryocrit levels, VH1-69+ B cell numbers and HCV viremia in the three patients studied prospectively for the proliferative responses to CpG are shown in Fig. 1b. Before antiviral treatment, VH1-69+ B cells constituted 87% of circulating B cells in patient 1, 63% in patient 2 and 85 in patient 3. Patient 1 had a more intense monoclonal B cell lymphocytosis with 892/μL VH1-69+ B cells compared to 156/μL and 374/μL in patients 2 and 3, respectively. Following antiviral therapy, the absolute number of circulating VH1-69+ B cells decreased in all patients (Fig. 1).

The absolute number of circulating VH1-69+ B cells of patients 1 and 2 decreased steadily, but significant numbers of these cells persisted for more than two years after the completion of therapy (Fig. 1b). Even more strikingly, at two years post-therapy VH1-69+ B cells represented about 30% of total B cells compared to less than 3% in healthy donors (Fig. 2). The phenotype of VH1-69+ B cells changed dramatically after the eradication of HCV. In fact, while before therapy the VH1-69+ population was predominantly made of CD21low B cells, early after the eradication of HCV infection the large majority of VH1-69+ B cells had a CD21+ MZ-like phenotype (Fig. 2). Before therapy, the VH1-69+ CD21low B-cells had increased expression of the activation markers CD86 and CD95, whereas VH1-69+ B cells CD21high did not express these surface receptors either before or after the eradication of HCV infection (electronic supplementary material Figure 1).
Fig. 2

Phenotype and function of VH1-69+ B cells in patient 1 (a) and in patient 2 (b) before and at different time points after antiviral therapy. Dot plots illustrate the frequency of VH1-69+ B cells among total B cells (left panels) and the frequency of CD21lowCD11c+ cells among VH1-69+ B cells (right panels). Histograms illustrate the proliferative responses to CpG (percent of cells with reduced CFSE fuorescence) of VH1-69+ and VH1-69- B cells electronically gated at the end of cultures

In the patient (# 3) who had recurrence of cryoglobulinemia during therapy despite negative HCV viremia, the absolute number of circulating VH1-69+ B cells declined (Fig. 1b) and the percentage of these cells decreased from 85% of total B cells before therapy to 20% at month 24 of treatment. However, unlike in patients with sustained responses, the relative distribution of MZ-like and of CD21low VH1-69+ B cells remained substantially unchanged in this patient, with 60% of CD21low B cells (among VH1-69+ B cells) before therapy and 74% at month 24 (electronic supplementary material Figure 2).

Proliferative Anergy of VH1-69+ B Cells

We previously reported that the MZ-like, as well as the CD21low, VH1-69+ B cells of untreated MC patients are functionally exhausted, since they fail to proliferate in response to the stimulation of TLR9 with CpG, [5]. Accordingly, before therapy the VH1-69+ B cells of all three patients described in this report were unresponsive to CpG, whereas autologous non-clonal VH1-69- B cells proliferated robustly (Fig. 2). Interestingly, the MZ-like VH1-69+ B cells that remained in the blood of patients 1 and 2 after the eradication of HCV infection were also unresponsive to CpG, although low-level proliferation was seen in patient 1 (Fig. 2); similarly, the VH1-69+ B cells detected at month 24 of treatment in patient 3 were unresponsive to CpG (not shown). We also investigated the proliferative responses of VH1-69+ B cells persisting in five additional patients with SVR. Pre-treatment cells were not available in these patients, and therefore they were studied only at one time-point (24 to 40 months) after the termination of antiviral therapy. In all of these patients the persisting MZ-like VH1-69+ B cells had reduced proliferative responses to CpG compared to autologous VH1-69-negative B cells; in fact, the percent of divided cells (mean ± SD) was 9 ± 4,1% in VH1-69+ B cells and 69,9 ± 4,4% VH1-69-negative B cells. Finally, unresponsiveness of VH1-69+ B cells to CpG was also confirmed in post-therapy samples of six virological non-responders (data not shown) and one relapser (patient # 3).

Collectively, our findings indicate that a relatively large population of functionally exhausted VH1-69+ B cells with a MZ-like phenotype, initially expanded in MC patients under the pressure of HCV, persists long after the inciting viral stimulus had withdrawn.


Persistence of significant numbers of circulating clonal B cells was already described in HCV-associated MC patients at the end of therapy with Peg-IFN and ribavirin, but follow-up data were not reported [12]. Our observation of the selective persistence of VH1-69+ B cells with a MZ phenotype is particularly intriguing. It is unlikely that a phenotypic switch of VH1-69+ B cells was due to the immunomodulatory effects of IFN or ribavirin, since these changes were not observed in virological non-responders (not shown) or in the patient described here with on-therapy relapse of cryoglobulinemia. The persistence of VH1-69+ B cells could be sustained by an occult HCV infection, due to the persistence of small quantities of HCV-RNA in the liver or in PBMC after therapy [13]. Indeed, although HCV RNA was negative in PBMC, occult infection in patient # 3 was consistent with the early viral relapse after discontinuation of therapy and with the persistence of cryoglobulinemia. By contrast, occult infection in the other two patients described here is made very unlikely by the negative testing for HCV viremia and for cryoglobulins at long-term follow-up. Thus, while occult HCV infection may be associated with the persistence of CD21low B cells it is unlikely to be the cause of the persistence of MZ-like VH1-69+ B cells. Interestingly, also in HIV infection the accumulation of exhausted CD21low B cells appears to depend on ongoing viremia, since it was observed only in HIV-viremic but not in anti-retroviral-treated aviremic individuals [9]. It would be of interest to investigate whether in aviremic HIV patients there is a population of exhausted MZ-like B cells.

It is possible that, after the eradication of HCV, the MZ-like CD21+ cells survive longer then their CD21low counterparts or that they more efficiently self-renew in response to stimuli other than HCV. It has been suggested that murine MZ B cells are extremely long-lived [14], but this concept has been subsequently challenged [15]. On the contrary, it has been shown that anergic B cells have a shorter lifespan than normally reactive mature B cells [16]. Reduced half-life is, at least in part, dependent on the poor competition of anergic B cells with non-anergic B cells for limited ambient B-cell-activating factor (BAFF) [16], which is increased in patients with HCV-associated MC [17]. Thus, BAFF made available by the reduction of the total B cell numbers in MC patients following eradication of HCV could selectively sustain the survival of MZ-like VH1-69+ B cells.

Another possibility for explaining the selective persistence of MZ-like VH1-69+ B cells is that they are more fit to respond to stimuli provided by antigen(s) different from HCV. Immune complexes could provide such a stimulus, because the rheumatoid factor cross-reactivity of the VH1-69-encoded IgM allows the BCR of these cells to bind antigen-bound endogenous IgG [1]. It has been shown that circulating immune complexes are able to activate rheumatoid factor-producing murine B cells, especially in the presence of surrogate T cell help or when they contain nucleosome antigens [18]. We found that VH1-69+ B cells do not proliferate in response to CpG alone or to the cross-linking of the B-cell receptor with anti-Ig alone, but co-stimulation with anti-Ig and CpG can elicit a proliferative response, although of low-level [19]. Thus, immune complexes containing DNA or RNA [20, 21], such as those formed with microbial components or with nucleosomes, might provide a dual BCR/TLR stimulus capable to sustain low-level self-renewal of anergic MZ-like VH1-69+ B cells. Antinuclear antibodies potentially forming DNA-containing immune complexes are frequent in MC patients [22]; in our patients, however, antinuclear antibodies tested negative.

The mechanism(s) by which exhausted CD21low B cells accumulate in some human immunological disorders remain largely elusive. These cells are mostly autoreactive and resemble innate-like B cells [6, 7], and several lines of evidence suggest that they become exhausted after extensive proliferation driven by microbial antigens [6, 9]. Recently, exhausted CD21lowCD11c+ B cells have been found to accumulate in aged female mice, and these cells have been termed aged B-cells (ABCs); these cells are mostly autoreactive and may have a direct role in the development of autoimmunity [23]. The similarity between murine ABCs and human CD21low B cells remains to be clarified. It has been suggested that the functional exhaustion of human CD21low B cells depends at least in part on their expression of an array of inhibitory genes including CD22, CD72, CD32b, CD85j, CD85d, Fc receptor-like 4 (FCRL4) and sialic acid–binding Ig-like lectin 6 (Siglec-6) [9]. The contribution of these inhibitory receptors to the dysfunction of CD21low B cells is supported by the partial recovery of the proliferative capacity and of effector function upon silencing of these genes with siRNA [24]. Thus, lack of expression of these inhibitory receptors by anergic VH1-69+ CD21high MZ-like B cells is intriguing. However, we recently observed [19] that MZ-like VH1-69+ B cells display constitutive over-expression of phosphorylated ERK and attenuated BCR signaling, a molecular signature of the anergy induced by continual BCR occupancy by antigen [25], and over-expression of the antiproliferative transcription factor Stra13.

Collectively, the currently available data suggest that the CD21low B cells of MC patients are part of a broader population of exhausted cells that includes MZ-like B cells, of which the former cells are seemingly descendants. Thus, immune dysfunction associated with expansion of autoreactive CD21low B cells is broader than previously recognized. Our present observations indicate that the MZ-like counterparts of CD21low B cells may persist in an exhausted condition long after the initial driving stimulus has withdrawn. These findings may help to shed light on the pathophysiology of these unusual autoreactive B cells and on the mechanisms of maintenance of autoimmunity.


We thank Dr. Maurizio Carbonari for supervising flow cytometry studies. This work was supported by the Intramural Research Program of the Sapienza University of Rome.

Conflict of interest

The authors declare that they have no conflict of interest

Supplementary material

10875_2012_9677_MOESM1_ESM.pdf (286 kb)
ESM 1(PDF 286 kb)

Copyright information

© Springer Science+Business Media, LLC 2012

Authors and Affiliations

  • Marcella Visentini
    • 1
  • Valentina Conti
    • 1
  • Maria Cagliuso
    • 1
  • Giulia Siciliano
    • 1
  • Carolina Scagnolari
    • 2
  • Milvia Casato
    • 1
  • Massimo Fiorilli
    • 1
  1. 1.Department of Clinical ImmunologySapienza University of RomeRomeItaly
  2. 2.Department of Molecular Medicine, Laboratory of VirologySapienza University of RomeRomeItaly

Personalised recommendations