Reversion of anergy signatures in clonal CD21^low B cells of mixed cryoglobulinemia after clearance of HCV viremia

Mixed cryoglobulinemia (MC) is a lymphoproliferative disorder caused by hepatitis C virus (HCV) and is characterized by clonal expansion of CD27⁺IgM⁺ B cells producing a rheumatoid factor often encoded by the V_H1-69 and V_k3-20 genes.^1,2  It is widely believed³  that the clonal expansion of V_H1-69⁺ B cells in MC and in HCV-associated non-Hodgkin lymphomas (NHLs) is triggered by the protracted stimulation by an HCV antigen that, however, remains elusive.⁴ 

Clonal B cells of patients with HCV⁺ MC display peculiar phenotypic and functional features. In fact, they commonly express low levels of CD21 (CD21^low B cells), express an array of inhibitory and apoptosis-related genes and a distinctive pattern of homing receptors (including CD11c⁺), fail to flux calcium upon B-cell receptor (BCR) triggering and proliferate in response to the stimulation of BCR or of Toll-like receptor 9 (TLR9), and are prone to die by apoptosis.^5-7  Identical CD21^low B cells are expanded in patients with common variable immunodeficiency (CVID)⁸  and in HIV-infected individuals.⁹  These CD21^low B cells have been defined “exhausted”⁹  for their similarity with virus-specific exhausted T cells.¹⁰  For some of their characteristics, human CD21^low B cells recall the murine “aged B cells” (ABCs) increased in aged mice,^11,12  which are CD21^lowCD11c⁺ B cells expressing the T-box expressed in T cells (T-bet) transcriptional factor and are important for the control of viral infections.¹³  However, ABCs lack many signatures of human CD21^low exhausted B cells, such as the fact that they robustly respond to the stimulation of TLR7 or TLR9.^11-13 

CD21^low B cells of HCV⁺ MC⁷  and CVID patients⁸  display high constitutive expression of the active phosphorylated form of extracellular signal–regulated kinase (pERK); this signature, together with reduced calcium flux and proneness to apoptosis, makes them closely resemble murine B cells made anergic by continual BCR engagement by antigen.¹⁴  In murine anergic B cells, constitutive ERK signaling and reduced lifespan are reversed by dissociation of self antigen from BCR using hapten competition, indicating the need for continual BCR occupancy for maintaining anergy.¹⁵  Here, we exploited the newly available direct-acting antivirals (DAAs), which rapidly suppress HCV viremia in HCV⁺ MC patients¹⁶  and lack the immunomodulatory properties of interferon, to untangle the effects of BCR disengagement in a human model of virus-driven anergy and exhaustion.

Twenty-four patients with HCV⁺ MC were treated with guideline-tailored DAA therapy as described previously¹⁶ ; all patients had negative HCV viremia at the end of treatment and sustained virologic responses (SVRs) through a follow-up of 12 (SVR12) to 24 (SVR24) weeks after the end of treatment. All patients were investigated for phenotypic changes in clonal B-cell populations after the cure of infection; 10 patients also underwent functional studies of ERK signaling, apoptosis, and proliferative responses to stimuli. Patients, B-cell phenotyping, and functional assays are described in detail in supplemental Methods, available on the Blood Web site. This study was approved by the local institutional ethics committee, and informed consent was obtained from all patients.

Clonal B cells of most patients with untreated HCV⁺ MC had high basal levels of pERK (Figure 1A), while the relative increment (fold increase) of pERK upon BCR stimulation was reduced (Figure 1B); uncoupling of constitutive and BCR-induced ERK activation was previously observed in CD21^low B cells of HCV⁺ MC⁷  and CVID⁸  patients. Furthermore, patients’ B cells showed increased apoptosis upon in vitro culture without stimuli (Figure 1C). In addition to these anergy-like features,^14,15  clonal B cells displayed the features of exhausted B cells,^5-7  as they were mostly CD21^lowIgM⁺CD27⁺CD11c⁺FCRL4⁺ and failed to proliferate in response to BCR or TLR9 stimulation (not shown).

Four weeks after the beginning of DAA therapy, constitutive ERK phosphorylation was significantly reduced compared with pretherapy levels (Figure 1D), although BCR-induced ERK phosphorylation increased only slightly (Figure 1E). At this time point, spontaneous in vitro apoptosis was also significantly reduced (Figure 1F). The possibility that early normalization of constitutive ERK activation and apoptosis was due to replacement of clonal B cells by normal B cells contrasts with the fact that the proportions of CD21^low B cells in these patients were unchanged at week 4 of therapy (54.9% ± 25% vs 53.8% ± 23% pretherapy).

Our results support the idea that, as in the case of antigen-induced reversible B-cell anergy in mice,¹⁵  reduction in the lifespan of B cells of HCV⁺ MC patients depended on chronic stimulation and could be reversed by disengagement of the BCR. Although it is believed that reduced lifespan of murine anergic B cells is largely due to unfavorable competition for B-cell–activating factor,¹⁷  the rapid increase of lifespan after BCR disengagement suggests that chronic BCR signaling is sufficient to initiate apoptosis.¹⁵  Thus, we investigated whether reduced lifespan of MC B cells was related to ERK signaling; treating MC B cells with the MEK/ERK inhibitor U0126, as previously described,⁸  failed to reduce spontaneous in vitro apoptosis (not shown), suggesting that ERK signaling is not directly involved in their proapoptotic pathway.

The proportions of CD21^low B cells declined steadily up to SVR24 (Figure 1G), although they remained on average higher than in healthy donors (mean ± SD, 5.8% ± 2.5%). However, the proportions of V_H1-69-expressing B cells remained stable up to SVR24 in 3 out of 5 patients (Figure 1H), suggesting that phenotypic changes occurred in clonal B cells after eradication of HCV, as previously observed in HCV⁺ MC patients treated with interferon.¹⁸  Indeed, in these 5 patients, the proportions of V_H1-69⁺CD21^low cells decreased from 84.8% ± 11.1% pretherapy to 36.4% ± 11.8% at SVR24 (P = .043) (supplemental Results), indicating a reversion of the CD21^low phenotype. Figure 1I shows representative cytograms from 1 patient. Despite phenotypic changes, long-lived clonal B cells failed to restore their capacity to proliferate in response to TLR9 stimulation. Indeed, the number of V_H1-69⁺ cells entering division (precursor cohort) in 3 patients (supplemental Results) was 2.8 ± 1.2 pretherapy and 5.3 ± 2.4 at SVR24, whereas it was 35.6 ± 6.5 and 32.6 ± 8.5, respectively, in autologous V_H1-69⁻ cells. Figure 1J shows representative cytograms from 1 patient. This indicates that HCV-driven B-cell exhaustion makes use of durable reprogramming mechanisms irrespective of the CD21^low phenotype, as suggested by previous studies.^18,19 

Recently, T-bet⁺CD21^lowCD11c⁺ B cells, similar to murine ABCs, were found increased in patients with chronic hepatitis C or cirrhosis without MC, and, importantly, eradication of HCV with DAAs led to their decrease, indicating dependence on infection.²⁰  Although we did not investigate and cannot exclude T-bet expression in clonal B cells of MC patients, their CD27⁺IgD⁺IgM⁺ B cells^1,2  are phenotypically distinct from the T-bet⁺ B cells of HCV-infected patients without MC, which are mostly CD27⁻IgD⁻IgM⁻IgG⁺ class-switched cells.²⁰  Interestingly in this regard, while the T-bet⁺CD27⁻IgG⁺ B cells of HCV-infected patients decrease after antiviral therapy, the few T-bet⁺IgD⁺IgM⁺ B cells appear to increase.²⁰ 

In summary, we show that clonal B cells of HCV⁺ MC display signatures of anergy induced by continual BCR occupancy and of exhaustion driven by chronic viral infection. Anergy features (pERK overexpression and accelerated apoptosis) rapidly revert after disengagement from HCV, while phenotypic and functional features of exhaustion persist for several months. The rapid clearance of HCV viremia with DAAs, which unlike interferon do not act as immunological modulators, offers a unique model for untangling the interplay of virus-driven anergy and exhaustion in human B cells.

This work was supported by the Intramural Research Program of Sapienza University of Rome (C26A15AKF7) (M.V.). M.V. was supported by Fondazione Roma, Rome, Italy (Prot. 287/AI).

M.D.P. and S.C. are PhD candidates at Sapienza University of Rome, and this work is submitted in partial fulfillment of the requirement for the PhD.

Contribution: M.D.P., M.V., and M.F. designed the research; M.D.P., M.M., R.M., and L.T. performed experiments; M.C., S.C., and M.V. recruited and followed up the patients; M.D.P. and M.V. analyzed results; and M.D.P., M.C., M.V., and M.F. wrote the paper.

Conflict-of-interest disclosure: The authors declare no competing financial interests.

Correspondence: Marcella Visentini, Department of Clinical Medicine, Sapienza University of Rome, Viale dell’Università 37, 00185 Rome, Italy; e-mail: marcella.visentini@uniroma1.it.