BAFF system expression in double negative 2, activated naïve and activated memory B cells in systemic lupus erythematosus

Introduction B cell activating factor (BAFF) has an important role in normal B cell development. The aberrant expression of BAFF is related with the autoimmune diseases development like Systemic Lupus Erythematosus (SLE) for promoting self-reactive B cells survival. BAFF functions are exerted through its receptors BAFF-R (BR3), transmembrane activator calcium modulator and cyclophilin ligand interactor (TACI) and B cell maturation antigen (BCMA) that are reported to have differential expression on B cells in SLE. Recently, atypical B cells that express CD11c have been associated with SLE because they are prone to develop into antibody-secreting cells, however the relationship with BAFF remains unclear. This study aims to analyze the BAFF system expression on CXCR5- CD11c+ atypical B cell subsets double negative 2 (DN2), activated naïve (aNAV), switched memory (SWM) and unswitched memory (USM) B cells. Methods Forty-five SLE patients and 15 healthy subjects (HS) were included. Flow cytometry was used to evaluate the expression of the receptors in the B cell subpopulations. Enzyme-linked immunosorbent assay (ELISA) was performed to quantify the soluble levels of BAFF (sBAFF) and IL-21. Results We found increased frequency of CXCR5- CD11c+ atypical B cell subpopulations DN2, aNAV, SWM and USM B cells in SLE patients compared to HS. SLE patients had increased expression of membrane BAFF (mBAFF) and BCMA receptor in classic B cell subsets (DN, NAV, SWM and USM). Also, the CXCR5+ CD11c- DN1, resting naïve (rNAV), SWM and USM B cell subsets showed higher mBAFF expression in SLE. CXCR5- CD11c+ atypical B cell subpopulations DN2, SWM and USM B cells showed strong correlations with the expression of BAFF receptors. The atypical B cells DN2 in SLE showed significant decreased expression of TACI, which correlated with higher IL-21 levels. Also, lower expression of TACI in atypical B cell DN2 was associated with high disease activity. Discussion These results suggest a participation of the BAFF system in CXCR5- CD11c+ atypical B cell subsets in SLE patients. Decreased TACI expression on atypical B cells DN2 correlated with high disease activity in SLE patients supporting the immunoregulatory role of TACI in autoimmunity.


Introduction
Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by hyperactive B cells and autoantibody production (1).BAFF (B cell-activating factor) is a type II membrane-bound protein that has a relevant role in SLE pathogenesis promoting self-reactive B cells survival.BAFF can be cleaved and released as a soluble BAFF homotrimer by furin proteases.High levels of soluble BAFF (sBAFF) have been found in SLE patients, correlated with disease activity and anti-dsDNA antibodies (2)(3)(4).BAFF effects are performed through its receptors: BR3 (BAFF-R, BAFF receptor), TACI (transmembrane activator calcium modulator and cyclophilin ligand interactor) and BCMA (B cell maturation antigen).BAFF induces B cell survival, Tindependent class switch and plasma cells (PC) maintenance; it also can act as a negative regulator of B cell numbers (5).The expression of BAFF receptors on B cells in SLE patients have been evaluated previously showing a decreased expression of BR3 and BCMA associated with higher disease activity (6), highlighting the BAFF receptor's role in the regulation of B cell homeostasis.
Recently, new B cell subpopulations have been described that are prone to differentiate into PC in an extrafollicular manner and are expanded in SLE (7,8).These autoimmunity-associated B cells (ABC) could be key in SLE pathogenesis because they are prone to develop into antibody-secreting cells (ASC).Although, T-cell interaction is necessary to initiate this extrafollicular B cell subset, once it is formed, cytokine stimulation is sufficient to drive its development into ASC (8)(9)(10)(11).Different expression patterns have been identified in these ABCs like the low levels of CD21 and high levels of CD11c, negative regulators as FcRL5 and transcriptional factors as T-bet, Zeb2 and Blimp-1 (7,8).Previous studies have classified these ABCs as atypical memory B cells for the presence of somatic hypermutation although they had mixed expression of CD27 (12)(13)(14).Other studies have used the CXCR5 and CD11c markers to identify these ABCs as double negative 2 (DN2) B cells (CD27 -IgD -CXCR5 -CD11c + ) and activated naïve (aNAV) B cells (CD27 -IgD + CXCR5 -CD11c + ).aNAV B cells are precursors of DN2 B cells and differentiate into PC through IL-21, IFNg and TLR7 stimulation (7,15).Although it has been described the important role of BAFF in B cell development, just a few studies have evaluated the presence of the BAFF and their receptors in CXCR5 - CD11c + atypical B cell subsets in mice and humans (16,17).This study aims to analyze the BAFF system expression on CXCR5 -CD11c + atypical B cell subpopulations in SLE patients.

SLE patients and healthy subjects
Forty-five SLE patients were included from the Rheumatology department of the Hospital General de Occidente, Guadalajara, Mexico.All patients fulfilled the EULAR/ACR 2019 classification criteria for SLE (18).The Mexican version of the Systemic Lupus Erythematosus Disease Activity Index (MexSLEDAI) score (19) was determined in SLE patients at the time of inclusion.All patients were stratified according to the MexSLEDAI index as follows: 0-2pts Non-Active, 3-5pts Low Disease Activity (LDA) and ≥6pts High Disease Activity (HDA) (20).Also, 15 healthy subjects (HS) with no infections, allergy, or chronic diseases, were included as controls.The ethics in research committee of the hospital approved this study with the registration number CEI-147/21.All the subjects signed an informed consent according to the regulations of the General Health Law on Research for Health in Mexico.The study was conducted according to the ethical principles for medical research involving human subjects established in the Declaration of Helsinki.

Flow cytometry
Human peripheral blood mononuclear cells (PBMCs) were isolated from fresh peripheral blood with Histopaque-1077 (Merk KGaA; Darmstadt, Germany) by density gradient, and stained using standard flow cytometry methodology with the antibodies described in the Supplementary Table 1.Flow cytometric analyses were performed with an Attune Nxt instrument (Thermo Fisher Scientific Inc; Waltham, MA, USA).The compensation was performed with the UltraComp eBeads (CAT: 01-2222-41, Thermo Fisher Scientific Inc; Waltham, MA, USA).Fluorescence minus one (FMO) control was used to adjust the background fluorescence and gates.The histograms of these controls are included in the Supplementary Figure 1.The acquisition data was obtained with Attune NxT Software version 3.1.2.The results are reported as percentage (%) of expression and geometric mean fluorescence intensity (MFI).Data were analyzed using FlowJo v.10 (BD, Franklin Lakes, NJ, USA).Instrument quality control was performed once a week during the experiment period using Attune Performance Tracking Beads (CAT: 449754, Thermo Fisher Scientific Inc; Waltham, MA, USA) and settings were maintained for all experiments.

Cytokine determination
Serum was separated from the total blood of each subject at the time they were involved in the study and were stored at -20°C.Serum cytokine levels of BAFF and IL-21 were quantified in SLE patients and HS by Enzyme-Linked immunosorbent Assay (ELISA) according to the manufacturer's instructions.Soluble BAFF was measured with Quantikine Human BAFF/BLyS/TNFSF13B (CAT: PDBLYS0B; R&D Systems, MN, USA) and IL-21 was measured with MAX Deluxe Set Human IL-21 (CAT: 433804; BioLegend, San Diego, CA, USA).

Data analysis
Data were analyzed using GraphPad Prism v.9.Categorical variables are presented as absolute values.Continued variables are presented as medians with 25th-75th percentiles.Normality tests were performed, and data didn't have normal distribution.Statistical differences between for 2 groups were calculated with Mann-Whitney U test and for ≥3 groups with Kruskal-Wallis's test and Dunn's test as post hoc.Correlations were calculated with Spearman test.Statistical significance was represented as follows: ns was p>0.05, *p<0.05,**p<0.01,***p<0.001and ****p<0.0001.

DN2 and aNAV B cells have a decreased expression of TACI in SLE patients
The CXCR5 -CD11c + atypical B cells DN2 and aNAV are described as ABC cells with an important role in the development of SLE for their capacity to develop into antibody-secreting cells.We found that these subpopulations have a decreased expression of TACI in SLE patients compared to HS, DN2 (28.25% vs 44.0%, p = 0.0153) and aNAV (39.55% vs 53.8%, p = 0.0020) (Figures 3I, J).The expression of mBAFF, BR3, TACI and BCMA in the atypical SWM (CXCR5 -CD11c + ) and USM (CXCR5 -CD11c + ) B cells (Figures 3K,  L) was similar between SLE patients and HS.The expression of BCMA on aNAV was decreased in SLE patients compared with HS (MFI= 1,939 vs 4,075, p = 0.0315) (Supplementary Figure 2J).

sBAFF and IL-21 are correlated with atypical B cells DN2 and aNAV
Soluble levels of BAFF and IL-21 were compared in SLE patients and HS.Although sBAFF didn't show differences between SLE patients and HS (1281 vs 1076 pg/mL, p = 0.1322), soluble levels of IL-21 were higher in SLE patients compared to HS (161 vs 31 pg/ mL, p = 0.0388) (Figures 4A, B).Spearman correlation test was made for sBAFF and IL-21 with the frequency of the B cell subpopulations and the BAFF system expression.sBAFF showed a positive correlation with the frequency of aNAV B cells (r = 0.3630, p = 0.0154) (Figure 4C) and negative correlation with the expression (MFI) of BR3 in aNAV B cells (r = -0.3144,p = 0.0400) (Figure 4D).Additionally, IL-21 had a negative correlation with the expression (MFI) of TACI in DN2 B cells (r = -0.3335,p = 0.0270) (Figure 4E).Furthermore, there was a positive correlation of IL-21 with the MexSLEDAI score (r = 0.3207, p = 0.0360) (Figure 4F).
The frequency of atypical B cell subsets DN2, CXCR5 -CD11c + SWM and CXCR5 -CD11c + USM B cells is correlated with the expression of the BAFF system in SLE patients.
We evaluated correlations among mBAFF, BR3, TACI and BCMA MFI expression and the frequency of DN2 B cells (Figure 5A) and CXCR5 -CD11c + USM B cells (Figure 5B).Also, correlations among the percentage's expression of mBAFF, BR3, TACI and BCMA and the frequency of CXCR5 -CD11c + SWM B cells (Figure 5C) and CXCR5 -CD11c + USM B cells (Figure 5D) were analyzed.A negative correlation between the frequency of DN2 and CXCR5 -CD11c + USM B cells with the MFI expression of BR3 was observed.Also, we found a negative correlation between the frequency of CXCR5 -CD11c + SWM and CXCR5 -CD11c + USM B cells with the percentage's expression of TACI.

Discussion
SLE is an autoimmune systemic disease, characterized by autoantibody production and diverse clinical manifestations.The role of the BAFF system has been associated with the pathogenesis of SLE and disease activity by its capacity to promote B cell survival and proliferation.Recently, CD11c + B cells have been described as potential self-reactive B cells prone to develop into ASC in an extrafollicular manner.Jenks et al. has characterized some of these B cell subpopulations by the markers CXCR5 -CD11c + (7).Our work aims to analyze the BAFF system expression in the CXCR5 -CD11c + atypical B cells in order to elucidate its role in the SLE pathogenesis.
Firstly, we found a decreased frequency of unswitched memory B cells (CD27 + IgD + ) in SLE patients.This decreased USM frequency was reported to be related with the increase of DN B cells (CD27 -IgD -) after TLR9 stimulation (13).Although we found an increase number of DN B cells, the difference wasn't significant as reported in previous studies (13,21).Like ABCs, the CD27 + CD21 -memory B cells have been characterized before in malaria- exposed donors as "activated memory B cells" with high expression of Blimp-1 and IRF4 (22).As expected, the CXCR5 + CD11c -B cell subsets have the highest frequency in all B cell subpopulations.However, we found a decreased frequency of these CXCR5 + CD11c - B cell subsets in SLE patients.The above could be related with the elevated frequency of CXCR5 -CD11c + B cells in SLE patients, which correspond to atypical B cells that resemble autoimmunityassociated B cells (ABCs).ABCs have been described previously by the expression of CD11c and T-bet (8,11), afterward by absence of CXCR5 marker as: CXCR5 -CD11c + atypical DN2 and aNAV B cells (7,23).
On the other hand, we found an increased expression of mBAFF in the classic B cell subpopulations (DN, NAV, SWM and USM) as well as in all the CXCR5 + CD11c -B cell subsets (DN1, rNAV, CXCR5 + CD11c -SWM and USM B cells).This elevated expression of mBAFF could be a consequence of B cells activation related to SLE pathogenesis.The expression of membrane-bound BAFF (mBAFF) is null in resting B cells; however, in vitro studies have shown that B cells can express mBAFF when they are activated by TLR9 through MyD88 (24, 25).B cells of SLE have higher expression of TLR9 ( 26) and its effect is related with the B cells activation and extrafollicular subpopulations development (27).CXCR5 -CD11c + atypical B cell subsets also had elevated expression of mBAFF compared with CXCR5 + CD11c -B cells, in both SLE and HS patients, suggesting the importance of this cytokine in the maintenance and likely in ABCs development.
The low expression of BR3 on B cells has been reported previously in SLE patients and even the lower BR3 expression was correlated with the disease activity (6,28).In this study we found a decreased expression of BR3 on DN and SWM B cells of SLE patients.It is known that BR3 can be shed by ADAM10 as consequence of BAFF binding and enhanced by TLR9 activation in the presence of TACI (29).Therefore, the low expression of BR3 in these cells in part could be explained by the release of soluble BR3 (sBR3) (30).Other studies have demonstrated that the decreased expression of BR3 measured by flow cytometry could be associated to the receptor occupancy (31).Also, in ABCs (CD11c + T-bet + ), a lower BR3 expression has been previously reported (8).In this study, we also found a decreased expression of BR3 in ABCs although it didn't show significant differences.Another study carried out in mice report an ABC subset (CD21 -CD23 -) that have similar expression of BR3 with the follicular B cells and have the same capacity to bind BAFF but doesn't rely on the BAFF/BR3 axis to survive (16).This could reflect an increase in the ABCs subpopulation at the expense of conventional follicular B cells frequency.This is interesting to analyze, since it could reflect a phenomenon of competition for niches, in which ABCs sequester the available BAFF by being more avid.The remarkable low expression of BR3 in atypical B cell subpopulations in addition to the increased frequency of these cells in SLE could explain the decreased expression of BR3 on CD19 + cells in SLE patients reported previously (6,28,31,32).In addition, our work showed that SLE patients had an increased expression of BCMA on rNAV, CXCR5 + CD11c -SWM and in the classic B cell subpopulations (DN, NAV, SWM and USM B cells).The elevated expression of BCMA on B cells have been reported in previous studies.Also, this elevated expression of BCMA could be related with TLR9 stimulation (25,33).In this study the expression of BCMA was found to increase in CXCR5 - CD11c + atypical B cells with no differences between SLE and HS.This increased expression of BCMA on CD11c + B cells could be associated with the capacity of BCMA to promote Ag-presentation through JNK-axis activation (34).T follicular helper cells (Tfh) have been associated with CD11c + B cells because of their interrelationship with IL-21, nevertheless, it has been hypothesized that CD11c + B cells can stimulate Tfh cells through strong Ag-presentation (35,36).Although in the present study we observe a higher percentage of B cells positive for BCMA, some subpopulations of SLE patients showed a lower expression of BCMA by MFI.Particularly for aNAV in which a negative correlation between BCMA expression and disease activity was found.This finding had been previously reported in our research group (6).
Furthermore, we found an increased expression of TACI on rNAV B cells of SLE patients.However, the CXCR5 -CD11c + atypical B cells DN2 and aNAV, showed decreased expression of TACI in SLE patients compared with HS.This finding is interesting because of the paradoxical functions of TACI in B cells (37).It's known that TACI can promote Ig-class switching in T-independent responses (38), but the decreased or null expression of TACI is associated with expanded numbers of B cells (39).In addition, TACI can induce and maintain Blimp-1 expression (40) that promotes B cell differentiation instead of proliferation which emphasizes the control exerted by TACI in the mature B cells pool, positioning TACI as an important receptor for the B cell homeostasis.It is possible that the interaction of CD11c + B cells with Tfh-like cells could be mediated by ICOS-ICOSL because TACI deficiency can increase ICOSL expression on GC B cells in mice (36,41).The immunoregulatory effect of TACI has been tested in mice, describing that TACI can upregulate the expression of Fas and FasL promoting apoptosis in MZ B cells (42).Although TACI is important in the B cell maturation, loss of TACI has been associated with autoimmunity development in mice (43).Also, in humans TACI deficiency was associated with breached immune tolerance in subjects with common variable immune deficiency (44).We hypothesize that the lower expression of TACI in the CXCR5 - CD11c + atypical B cells DN2 and aNAV in SLE could be responsible in part for loss of regulation of these populations, highlighting the BAFF role in the SLE pathogenesis.
Finally, levels of sBAFF were positively correlated with aNAV frequency, whereas the expression of BR3 on aNAV B cells was negatively correlated with sBAFF levels.This correspond with previous reports where the decreased expression of BR3 was correlated with elevated levels of sBAFF (6,28,30).Soluble levels of IL-21 were elevated in SLE and correlated with the disease activity as previously reported (45).It has been described that the atypical B cells are developed through IL-21 signaling (8) in addition to that IL-21 can suppress the expression of TACI in activated B cells (46).This relation could explain that the soluble levels of IL-21 were negatively correlated with the expression of TACI on DN2 B cells.On the other hand, the decreased expression of TACI could be related with the release of the receptor as recently Hoffmann et al. has demonstrated that TACI is shed constitutively by the metalloprotease ADAM10 and the soluble form of TACI acts as a decoy receptor that reduces the NF-kB activation and thus promotes a negative effect on signaling (47).In the context of TACI occupancy, the evidence shows the BAFF 3-mer affinity for TACI is poor compared with the BAFF oligomeric forms (48).Since the prevalent soluble BAFF form is the 3-mer we hypothesize the preferential binding of this form occurs with BR3 in vivo.
We previously have reported that the decreased expression of TACI and BCMA on B cells was related with the disease activity in SLE (6).In this study we found that the expression of TACI on DN2 and BCMA on aNAV B cells were negatively correlated with the disease activity.Furthermore, we found that HDA patients has a decreased (MFI) expression of TACI compared with Non-Active SLE patients in SWM, CXCR5 + CD11c -SWM, DN and DN1 B cells.The above highlights the regulatory role of TACI in B cells homeostasis, which appears to be affected in the autoimmunity context.

Conclusion
These results suggest a participation of the BAFF system in CXCR5 -CD11c + atypical B cell subsets in SLE patients.Decreased TACI expression on DN2 B cells correlated with high disease activity in SLE patients supporting the immunoregulatory role of TACI in autoimmunity.

Data availability statement
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

TABLE 1
Clinical characteristics of SLE patients.