Interferon-γ signal drives differentiation of T-bethi atypical memory B cells into plasma cells following Plasmodium vivax infection

For development of a long-lasting protective malaria vaccine, it is crucial to understand whether Plasmodium-induced memory B cells (MBCs) or plasma cells develop and stably contribute to protective immunity, or on the contrary the parasite suppresses antibody responses by inducing MBC dysfunction. The expansion of T-bethi atypical MBCs is described in chronic Plasmodium falciparum-exposed individuals. However, it remains unclear whether accumulation of T-bethi atypical MBCs is indicative of a protective role or rather an impaired function of the immune system in malaria. Here, the phenotypic and functional features of T-bethi atypical MBCs were studied in P. vivax patients living in an area of low malaria transmission. During P. vivax infection, the patients produced a twofold higher frequency of T-bethi atypical MBCs compared to malaria non-exposed individuals. This distinct atypical MBC subset had a switched IgG phenotype with overexpression of activation markers and FcRL5, and decreased Syk phosphorylation upon BCR stimulation. Post-infection, expansion of T-bethi IgG+ atypical MBCs was maintained for at least 3 months. Further studies of the contribution of T-bethi atypical MBC function to humoral immunity showed that synergizing IFN-γ with TLR7/8 and IL-21 signals was required for their differentiation into plasma cells and antibody secretion.


Increased proportion of T-bet hi atypical MBCs during acute P. vivax infection.
To observe the acquisition of T-bet hi atypical MBC responses in natural P. vivax infection in the setting of low-level transmission, the frequency of T-bet + B cells was determined during acute malaria. Most (85%) subjects possessed a significantly higher frequency of these cells than did healthy controls (HCs) (Fig. 1a, b and Supplementary  Table S1). As the T-bet + MBC population is heterogeneous 17,28,29 , to understand their function as contributors to antibody responses, T-bet profiling (based on low, intermediate, and high expression on CD19 + B cells and among MBC subsets) was determined (Fig. 1c). Of T-bet hi B cells, 69.31%, 15.64% and 3.55% exhibited atypical, activated and classical MBC phenotypes, respectively. T-bet int B cells made up 38.33% of atypical, 15.36% of activated, and 12.50% of classical MBCs. In addition, among T-bet neg B cell subsets, 13.52% were atypical, 3.95% were activated, and 10.85% were classical MBCs (Fig. 1d).  Table S1). There was a higher percentage (62%) of T-bet hi switched (IgM +/− IgD − ) than unswitched (IgM + , IgD + ) MBCs (Fig. 2a). We next addressed whether the T-bet hi switched MBCs were to IgM or IgG. Significantly more IgG than IgM was detected in this population (Fig. 2b). Further analysis of T-bet hi switched (IgM or IgG) MBCs (from CD21, CD27 expression) showed that the phenotypic markers were mainly on atypical MBCs (T-bet hi IgM; average 59.63%, SEM 4.58%; T-bet hi IgG; average 69.07%, SEM 3.97%) and activated MBCs (T-bet hi IgM; average 31.16%, SEM 4.62%; T-bet hi IgG 27.10%, SEM 4.17%) (Fig. 2c, d).

Expansion of T-bet hi atypical MBCs was stably maintained post-infection.
We observed kinetic responses of T-bet hi B cells and their subsets during and after P. vivax infection. Of the 26 patients, 9 with higher frequencies of T-bet hi B cells than did HCs (2.34%; average + 2SD of HCs) were recruited as a cohort for tracking changes in the cell frequencies during acute malaria (AC) and after infection (Supplementary Table S1). During AC, all patients in this cohort presented higher total T-bet hi B cells (average, 4.42%; SEM 0.73%) compared to the percentage in HCs (average 1.27%; SEM 0.18%). Post-infection, the population of T-bet hi B cells expanded further as found at the 1-month sampling (average 2.59%; SEM 0.65%). However, the frequency of these cells was significantly decreased at the 3-month post-infection time-point compared to AC (Fig. 3a). Four and 2 patients maintained elevated T-bet hi levels 3 and 6 months after infection, respectively (Fig. 3a).

T-bet hi atypical MBCs of P. vivax-infected patients greatly reduced Syk BCR signaling.
As noted above, T-bet hi atypical MBCs upregulated both activation and inhibition markers. To observe whether the upregulated markers on these MBCs are related to upstream BCR signaling events, the phosphorylation of BCR signaling molecules was investigated in patients. We found a significant decrease of Syk phosphorylation in T-bet hi atypical compared to T-bet hi activated MBCs following BCR cross-linking, while slight reductions of pPLCγ2 and pBLNK BCR molecules were observed in these MBCs (Fig. 5a, b). In addition, BCR-induced pSyk in T-bet hi atypical MBCs of P. vivax patients was lower than in HCs, whereas no difference was observed in T-bet hi activated nor classical MBCs (data not shown). Further investigation of functional capacity of these T-bet hi atypical MBCs to produce antibody is needed to determine whether similar or different signals from classical MBCs are required for their effector function.

IFN-γ synergizes with TLR7/8 and IL-21 to drive T-bet hi atypical MBC differentiation into plasma cells.
As T-bet hi atypical MBCs showed reductions of upstream BCR signaling events, their differentiation into plasma cells and their secretion of antibody were further assessed in the acute P. vivax patients.
We initially stimulated T-bet hi atypical MBCs (atypical MBCs) under conditions which readily drove classical MBCs into plasma cells, stimulation in vitro with TLR7/8 (R848) and B-cell activating factors (IL-2 and BAFF) 21 .
Our results showed that classical MBCs could differentiate into plasma cells in both patients (average 23.94%; SEM 4.83%) and HCs (average 30.60%; SEM 3.13%). For atypical MBCs, 6.17% of plasma cells were detected in HC cultures, whereas fewer (3.22%) were found in P. vivax patients, indicating that atypical MBCs might require more signals to generate plasma cells ( Fig. 6a, b). Our data showed that T-bet hi atypical MBCs displayed upregulated IL-21R during acute P. vivax malaria (Fig. 4). We next observed an ability of IL-21, which are T follicular helper (Tfh)-derived signals, to promote the differentiation of these cells in vitro. The combination of IL-21 with TLR7/8, IL-2 and BAFF resulted in approximately 2.5-fold higher induction of atypical MBCs into plasma cells (average 8.82%; SEM 0.84%) compared to the same condition without IL-21. Compared with HCs, atypical MBCs from patients had a slightly lower ability to generate plasma cells (HC; average 12.90%; SEM 1.36%) (Fig. 6a, b), suggesting that the IL-21 signal could help in their development. Since IFN-γ signal was reported to drive double negative 2 (DN2) (CD27 − , IgD − , CXCR5 − and CD11c + ) B cell differentiation into plasma cells in autoimmune diseases following stimulation with TLR7/8 and IL-21 22, 27 , we next included IFN-γ, Th1-derived signals, in atypical MBC culture to enhance their differentiation. In the presence of IFN-γ, atypical MBCs from P. vivax patients and HCs generated twofold more plasma cells compared to the same condition without IFN-γ stimulation (Fig. 6a, b), indicating that atypical MBCs required signal from Th1 cells to promote differentiation into plasma cells.
In addition, the functional ability of atypical MBCs to secrete antibody was observed upon stimulation with various signals. Upon TLR7/8, IL-2 and BAFF stimulation, total IgG was greatly detected in classical MBC culture supernatant, whereas these signals did not induce antibody secretion from atypical MBCs (Fig. 6c). Addition of T-cell-derived cytokines into atypical MBC cultures (TLR7/8, IL-2 and BAFF) showed that IL-21 increased total IgG antibodies and IFN-γ significantly increased IgG levels in supernatant (Fig. 6c). Importantly, P. vivax antigen-specific antibody was specifically detected in classical MBC cultures, from 2 of 5 patients. The undetectable level of anti-P. vivax antibody upon TLR7/8, IL-21 and IFN-γ stimulation of atypical MBCs was observed due to the low frequency of antigen specific atypical MBCs in blood circulation (Fig. 6d).

Discussion
Expansion of T-bet + B cells has been described in the context of various chronic antigenic stimulations, including aging 30,31 , autoimmunity 27, 32 , vaccination 16,23 , and infectious diseases such as HIV, HCV, and P. falciparum malaria 16,17,29 . This cell population expressed the phenotype of as atypical MBCs (CD21 − CD27 − ), and also expressed CD11c, CXCR3 and FcRL5 17,25,26 . However, the function of high expression of T-bet in atypical MBCs in these models, as well as the triggers of their differentiation into plasma cells, is unclear. Our results revealed an increase of T-bet hi B cells during acute P. vivax infection. Most of the responses were T-bet hi atypical MBCs with IgG class switching, and were stably maintained the responses in the convalescence phase for at least 3 months. These T-bet hi atypical MBCs overexpressed CD11c, CD69, CD86 and IL-21R, as well as FcRL5 and CD95. Functional analysis of these MBCs exhibited diminished BCR signaling following cross-linking. To drive the functional role of T-bet hi atypical MBC differentiation into plasma cells and antibody secretion, stimulation with a combination of TLR-7/8 and T cell-derived cytokines (IL-21 and IFN-γ) was required. Together, our data suggest that P. vivax infection triggered an expansion of T-bet hi atypical MBCs and that these expanded cells played a role in anti-malarial humoral immunity under TLR7/8 ligands, and Tfh-and Th1-derived signals.
A more detailed understanding of the generation and persistence of the atypical MBCs associated with malaria infection is useful for vaccine development. In P. falciparum malaria, an increase of T-bet hi atypical MBCs occurs in children and adults living in moderate and high transmission settings. These expanded cells were commonly found to express CD11c + , CXCR3 + , FcRL5 + phenotypes and exhibited reduced BCR signaling 17,21,24 . Moreover, the number of atypical MBCs is higher in children with chronic asymptomatic P. falciparum infection 17,21 or previous exposure to the parasite compared to having a primary infection 25   In addition, they potentially participate in isotype switching to IgG and persist after parasite clearance 17,33,35 . The lower expression of the GC-positioning receptors CXCR5 and CCR7 on these cells might explain their migration outside GCs to play a crucial role in humoral immunity. The overexpression of FcRL5 on IgG + T-bet hi atypical MBCs in vivax malaria might be part of a feedback mechanism which downregulates the hyperactivated state of atypical MBCs, or they may be markers of durable and robust responses to re-infection as previously proposed in P. falciparum studies 36 . In-depth investigations regarding these pivotal signals of the immune mechanisms that drive T-bet hi atypical MBCs to play effector function are still needed for improved malaria vaccine design. The role of T-bet hi atypical MBCs in the immune system remains elusive. To date, the signals inducing high expression of T-bet in atypical MBCs, as well as the function of this cell population in the immunity, require more understanding. In malaria, it has been reported that P. falciparum-infected erythrocytes drive expansion of T-bet in B cells 37 , and that IFN-γ is the key driver for T-bet hi atypical MBCs differentiation 17,20 . With these findings, a functional role of T-bet hi atypical MBCs might be immunosuppression, as suggested by the upregulation of multiple inhibitory receptors with reduced BCR-mediated signaling and limited antibody production 17,20,21 . Up to now, there is no report in P. vivax malaria which differs from P. falciparum since it has a dormant stage in the human liver (hypnozoites) and a relapsing fever. Moreover, different malaria transmission settings may influence the functional roles of atypical MBCs. Here, we found that T-bet hi atypical MBCs from P. vivax-infected patients had significantly reduced Syk BCR signaling upon cross-linking. This indicates that these cells might have less activation in response to BCR signaling or require additional signals for transition from resting to a highly active state. Our functional analysis focused on antibody responses demonstrated that, unlike classical MBCs, atypical MBCs did not differentiate into plasma cells and secrete IgG antibodies upon TLR7/8, IL-2 and BAFF stimulation. These results suggested that atypical MBCs require different or more signals for plasma cell generation. Previous reports documented that T cell-derived cytokines (IL-21 and IFN-γ) were significantly increased in sera from patients with autoimmune diseases 38,39 and P. falciparum malaria infection 40,41 . Of note, we found upregulation of IL-21R on these cells. The addition of IL-21 to atypical MBCs increased the capacity to induce plasma cell generation, indicating that IL-21 is one of the most potent cytokines derived from Tfh cells in regulating atypical MBCs. Furthermore, we observed their functional capacity by adding Th1-derived cytokines. www.nature.com/scientificreports/ Our results showed that plasma cell differentiation was enhanced by IFN-γ, as it showed a higher frequency of plasma cells in atypical MBC culture compared to the condition without IFN-γ. Collectively, the findings indicated that atypical MBCs required multiple stimuli including TLR7/8, IL-21 and IFN-γ signals, to induce plasma cell generation. Of note, anti-P. vivax specific antibodies were found in classical MBC cultures of some patients, but not in atypical MBC cultures, likely reflecting the low level of P. vivax specific atypical MBCs in circulating blood. Thus, there is a need to develop high sensitivity techniques for detection of the rare population of malarial specific MBCs in producing anti-malarial specific antibodies as well as their protective roles in further studies.
For the studies of MBC responses in malaria, the complex life cycle, antigenic variation/polymorphism, and its ability to establish chronic infection could well influence development and longevity of MBC responses. P. vivax parasites have unique biological features compared to P. falciparum. The main differences of P. vivax parasite include: (i) dormant hypnozoite in liver, which often escapes immune response and leads to latent state 42, 43 , (ii) recurrent parasitemia implicating in relapsing symptoms 44 , (iii) preference for invading reticulocytes and increasing their deformability and fragility leading to evasion of immune system 45,46 . Given these different characteristics, the detailed knowledge of atypical MBC function in P. falciparum infection might not explain in P. vivax-exposed individuals. In comparison of T-bet hi atypical MBC responses in malaria, these cells were increased in frequency and activated by showing upregulation of activating markers (CD11c, CD69) and costimulatory molecules (CD86 and IL-21R) in both P. falciparum and P. vivax infection 17,20,47 . For the contribution www.nature.com/scientificreports/ of atypical MBCs in humoral immunity, the BCR signaling (PLCγ2, pBLNK and Syk) of P. falciparum-associated atypical MBCs was impaired 21 , thereby requiring Staphylococcal enterotoxin B and Tfh cell interaction for plasma cell differentiation 47 . In P. vivax, only greatly reduced Syk BCR signaling was detected in T-bet hi atypical MBCs from infected patients, resulting in requirement of additional signals from TLR7/8 and T cell-derived cytokines (IL-21 and IFN-γ) for driving into plasma cell differentiation and antibody secretion. Future exploration is required to determine the role of Tfh cells in contribution to activation and differentiation of T-bet hi atypical MBCs into plasma cells. Limitation of this study is a small sample size for tracking kinetic responses of expanded T-bet hi atypical MBCs in cohort analysis. Nine of total 26 who had higher T-bet hi atypical MBC frequencies than HCs (average + 2 SDs of HCs) were recruited in a 6-month cohort study. Thus, this observation should be considered as preliminary result to demonstrate the maintenance of T-bet hi IgG atypical MBCs in response to P. vivax infection. More screening of expanded T-bet hi atypical MBCs samples at acute disease and cohort study design would help to ensure their kinetic responses after parasite clearance as well as to clarify their functional roles in malaria immunity.
In summary, P. vivax infection was found to be associated with T-bet + B cell activation and expansion. The predominant response was of T-bet hi atypical MBCs with an IgG phenotype. These cells showed a profile of upregulated activation markers and FcRL5, and exhibited reduced responsiveness to pSyk following cross-linking. Notably, the signals from TLR7/8 and T cell-derived cytokines (IL-21 and IFN-γ) appeared to drive their function in anti-malarial humoral immunity. A better understanding of the regulators and functions of these expanded cells, in settings of chronic stimulation by various antigens, will provide further insight into the mechanism of their differentiation, and may improve vaccine strategies and therapeutic interventions.

Methods
Ethical statement. Ethical approval was obtained from the Committee on Human Rights Related to Human Experimentation, Mahidol University, Thailand (MUIRB 2012/079.2408). The study participants gave written informed consent to enroll in this study before collecting blood samples. All experiments involving human subject were conducted in accordance with relevant guidelines and regulations.
Study subjects and sampling. Twenty-six patients with acute P. vivax malaria living in an area of low intensity malaria transmission (Kraburi, Ranong Province, a village near the Myanmar border in Southern Thailand), were enrolled in the study for determination of frequency and phenotype of T-bet hi atypical MBCs. Of the 26, the 9 who had T-bet hi atypical MBC frequencies more than the average + 2 standard deviations (SDs) of that in HCs were enrolled in a 6-month cohort sub-study to track the kinetic changes of these cells with follow-up visits at 1, 3 and 6 months after infection. To further explore the profile of T-bet hi atypical MBCs during P. vivax infection, 10 subjects were recruited for analysis of activation and inhibition marker expression. The function of T-bet hi atypical MBCs was analysed for upregulation of BCR signaling molecules, differentiation into total IgG, and anti-P. vivax antigen-specific antibodies after in vitro stimulation (n = 5).
Plasmodium vivax infection was confirmed by microscopic examination and nested PCR of peripheral blood. Patients were evaluated for sub-patent parasitemia every 2 weeks. Data on past malaria infections were obtained from the records of the Vector-borne Disease Unit. Twenty-six malaria non-infected Thai residents who lived in non-endemic areas (Bangkok, Thailand) were recruited as HCs.
IgG ELISA. Culture supernatants were assessed for total IgG and anti-PvDBL-TH2 antibodies by indirect ELISA as previously reported 7 . Briefly, 1 μg/ml of anti-human IgG (clones MT91/145; Mabtech, Sweden) or 2 μg/ ml PvDBL-TH2 were coated on 96-well plates followed by blocking with 5% BSA-PBS. Undiluted supernatant was added to wells followed by detection with goat anti-human IgG conjugated to horseradish peroxidase (HRP)