Selenium supplementation suppresses immunological and serological features of lupus in B6.Sle1b mice

Abstract

Systemic lupus erythematosus (SLE) is a debilitating multi-factorial immunological disorder characterized by increased inflammation and development of anti-nuclear autoantibodies. Selenium (Se) is an essential trace element with beneficial anti-cancer and anti-inflammatory immunological functions. In our previous proteomics study, analysis of Se-responsive markers in the circulation of Se-supplemented healthy men showed a significant increase in complement proteins. Additionally, Se supplementation prolonged the life span of lupus prone NZB/W-F1 mice. To better understand the protective immunological role of Se in SLE pathogenesis, we have investigated the impact of Se on B cells and macrophages using in vitro Se supplementation assays and the B6.Sle1b mouse model of lupus with an oral Se or placebo supplementation regimen. Analysis of Se-treated B6.Sle1b mice showed reduced splenomegaly and splenic cellularity compared to untreated B6.Sle1b mice. A significant reduction in total B cells and notably germinal center (GC) B cell numbers was observed. However, other cell types including T cells, Tregs, DCs and pDCs were unaffected. Consistent with reduced GC B cells there was a significant reduction in autoantibodies to dsDNA and SmRNP of the IgG2b and IgG2c subclass upon Se supplementation. We found that increased Se availability leads to impaired differentiation and maturation of macrophages from mouse bone marrow derived progenitors in vitro. Additionally, Se treatment during in vitro activation of B cells with anti-CD40L and LPS inhibited optimal B cell activation. Overall our data indicate that Se supplementation inhibits activation, differentiation and maturation of B cells and macrophages. Its specific inhibitory effect on B cell activation and GC B cell differentiation could be explored as a potential therapeutic supplement for SLE patients.

1. Introduction

Systemic lupus erythematosus (SLE) is characterized by the formation of autoantibodies (autoAbs) to nuclear antigens such as double-stranded DNA (dsDNA), followed by immune complex deposition onto the tissues, causing inflammation and damage such as glomerulonephritis [1, 2]. The development of SLE is accompanied by the activation of both innate and adaptive immune systems, which cooperatively maintain the vicious circle of the disease [3]. B cells [3–6] and myeloid cells (macrophages, dendritic cells (DCs) and plasmacytoid dendritic cells (pDCs)) [7–9] are the most well understood players in the pathogenesis of SLE. B cells have emerged as key players in toll like receptor (TLR)-mediated systemic autoimmune responses, especially in SLE [10]. More importantly, spontaneously developed germinal centers (Spt-GCs) play a significant role in generating somatically hypermutated and class-switched pathogenic autoAbs, which cause SLE pathology, indicating dysregulation at the level of the GC tolerance checkpoint [10–14].

Considering the pivotal role of B cells in SLE pathogenesis, current clinically approved therapeutics are centered on B-cell-targeted interventions. The B-cell activating factor (BAFF) blocking human monoclonal antibody “belimumab” has been approved by the FDA and the EMA as a treatment for SLE following two successful phase III trials (BLISS-52 and BLISS-76) [15, 16]. However, clinical trials of another anti-BAFF agent, tabalumab (ILLUMINATE-1 and ILLUMINATE-2) [17, 18] were unsuccessful, as the treatment group did not show sufficient benefit compared to the placebo group. Similarly, an antagonist of BAFF, blisibimod, showed limited success in phase II trials as a treatment for SLE [19]. These data from clinical trials suggest that the consequences of blocking BAFF signaling in SLE patients are variable and need further understanding. Additionally, most patients receive supplemental concurrent administration of corticosteroids that have several adverse effects, including infections, hypertension, hyperglycemia, osteoporosis, cataracts, glaucoma and cognitive impairment [20, 21]. Therefore, effective treatment of SLE with minimal side effects requires newer approaches.

Selenium (Se) is an essential trace element and micronutrient that is required for several aspects of human health. Sodium selenite (Sel) and selenate are inorganic forms found in the soil that are converted in plants to organic forms, including selenomethionine (SM) and selenocysteine and numerous other related compounds (Institute of Medicine U.S. Panel on Dietary Antioxidants and Related Compounds, 2000). Sel and selenate are also components of dietary supplements [22]. SM represents the major form of Se in cereals and other plant crops, as well as in other supplements. Se-methylselenocysteine (MSC), another organic form of Se, can be found in broccoli, garlic and onions, especially when grown under Se-rich conditions [23], methylseleninic acid (MSeA) is a second generation oxidized monomethylated form of Se. All these Se forms have mainly been studied thus far for their potent anti-tumor activity by inhibiting cell proliferation of several cancers [24–29].

Several biological effects of Se occur through its incorporation into selenoproteins, many of which are involved in the activation, proliferation and differentiation of innate and adaptive immune cells [30, 31]. Other important cellular pathways regulated by selenoproteins include but are not limited to the signaling pathways linked to calcium flux [32] and oxidative stress [31]. Previous findings have demonstrated that supplementation with selenized-yeast in healthy men results in reduction of activated T and natural killer (NK) cells and related cellular pathways [33, 34] and decline in markers of oxidative damage [35]. In addition, supplementation with Se-enriched broccoli showed enhanced immune response in healthy individuals [36]. Interestingly, it was shown that adequate dietary Se supplementation in C57BL/6 mice is necessary for maintaining optimal B cell numbers in circulation [31]. Notably, lower levels of zinc (Zn) and Se were observed in SLE patients compared to healthy controls [37], while Se supplementation of lupus prone NZB/NZW-F1 female mice significantly increased their life span compared to untreated mice [38]. Intriguingly, we observed that selenized-yeast supplementation of healthy men increased the prevalence of several complement proteins, which are protective in SLE [39].

Based on the important role of Se in homeostatic immune cell functions and evidence that Se may play a protective role in murine and possibly in human SLE, in this study we determined the effect of orally supplemented Se (as MSeA) in murine SLE using B6.Sle1b mouse model of lupus. The B6.Sle1b mouse model of lupus harbors the signaling lymphocyte activation molecules (SLAM) family genes from the lupus-prone NZM2410 strain [13, 40–42]. This region in the telomeric part of chromosome 1 in mice is syntenic to the human chromosome 1 region 1q22–25, which is also associated with human SLE with a female gender bias [13, 42]. B6.Sle1b lupus model is specifically suited to study lupus-associated pathogenic, class-switched, hypermutated autoAbs with high affinity for self-antigens like dsDNA and Smith/Ribonucleoprotein (SmRNP), which are primarily generated through the germinal center (GC) B cells [13, 43–45]. We observed a significant reduction in total B cells; specifically GC B cells in Se-treated mice. In accordance, there was a significant reduction in class switched autoAbs to dsDNA and SmRNP in Se-treated B6.Sle1b mice compared to oral PBS-treatment. Unlike B cells, total T cells, T-regulatory cells (Tregs), DCs and pDCs were unaffected. In vitro experiments indicate that supplementation of Se during activation of B cells with anti-CD40L and lipopolysaccharide (LPS) inhibits optimal B cell activation. Similarly, differentiation and maturation of macrophages from mouse bone marrow-derived progenitors were impaired in the presence of Se. Overall, our data indicate that increased Se suppresses GC B cell responses in vivo and is inhibitory to myeloid cell maturation and differentiation in vitro. These observations indicate a promising therapeutic potential for Se in SLE and could be explored as a supplement to current SLE therapies like belimumab.

2. Materials and Methods

2.1. Mice

B6/129SF2/J (stock number 101045) and C57BL/6 (stock number 000664) mice were originally purchased from The Jackson Laboratory (Bar Harbor, ME) and bred in house. B6 mice congenic for the Sle1b sub-locus (B6.Sle1b) were generated as described previously [46–48] and bred in house. All the mice used in this study were females of 2–3 months (mo) of age and were housed in a specific pathogen free facility. All procedures were performed in accordance with Pennsylvania State University Institutional Animal Care and Use Committee (IACUC) guidelines.

2.2. Selenium compounds and treatments

2.2.1. In vitro treatments:

The following Se compounds with indicated dose range were utilized for in vitro treatments. Se-methylselenocysteine (MSC, 5 – 100 μM); selenomethionine (SM, 5 – 100 μM); sodium selenite (Sel, 0.5 – 10 μM); and methylseleninic acid (MSeA 1.25 – 10 μM) (Sigma-Aldrich, St. Louis, MO).

2.2.2. In vivo treatments:

2-months-old B6.Sle1b female mice of an average body weight of 20 ± 2 gm were used for the study. MSeA was dissolved in PBS (pH 7.2) to yield a concentration of 3 ppm Se and stored at −20°C in aliquots until use. A fresh vial was thawed for every use. MSeA was delivered orally five days a week (Monday–Friday) at a similar time of the day. The methodology and dosing frequency have been described previously [28, 49]. Briefly, MSeA was delivered to the back of the tongue of mice in a volume of 10 μl via a disposable sterile plastic tip from a P20 micropipette, such that on average each mouse received a dose of 3 mg per kg body weight. This method was preferred over gastric gavages to rule out any esophagus irritation or damage due to the high frequency of drug administration. These mice ingested the small volume easily without physical irritation. The untreated (UT) control mice received a placebo treatment of 10 μl PBS to the back of the tongue as described above. The mice had unrestricted access to irradiated Teklad 2918 diet pellets (Harlan Teklad, Madison, WI) and water during the entire treatment regimen. The diet contained 0.23 mg/kg Se (https://www.envigo.com/resources/data-sheets/2018-datasheet-0915.pdf). The oral treatments were continued for 3 mo, during which mice were weighed every week.

2.3. Ex vivo B cell activation assay

B cells were isolated from mice by anti-CD43 (Ly-48) negative selection using MACS purification (Miltenyi Biotech, San Diego, CA). Isolated B cells were either left unactivated (UA) or activated with 10 μg/ml anti-CD40 (UCSF) and 1μg/ml E.coli LPS (Sigma-Aldrich, St. Louis, MO) for 24h in the presence or absence of MSeA at indicated concentrations. Cells were harvested after 24h, washed and stained for indicated markers and analyzed by flow cytometry.

2.4. Macrophage differentiation and treatments

Bone marrow cells harvested from WT mice were cultured in DMEM containing 10% FBS along with L929 culture supernatant in a 1:1 ratio for two days in the presence or absence of various Se compounds at varying concentrations (described earlier). Floating cells were washed off and the adherent cells were further cultured for 6 days in 10% DMEM with L929 culture supernatant in a 4:1 ratio, along with same dosage and Se compounds. The Se compounds were supplemented again on the 4^th and 6^th day of culture. After the 7^th day, differentiated bone marrow derived macrophages (BMDMs) were trypsinized and analyzed for the expression of F4/80, CD11b, MHC-II, CD11c and CD86 etc., by flow cytometry.

2.5. Flow cytometry reagents

Single cell suspensions were prepared from harvested spleens by mechanical disruption. Red blood cells were lysed by incubation with Tris Ammonium Chloride. Splenocytes, purified B cells, or in vitro differentiated macrophages were stained with combinations of the following antibodies as indicated: Pacific Blue-anti-B220 (RA3–6B2); Alexa Fluor-700-anti-CD4 (RM4–5); PE-anti-PD-1 (29F.1A12); PeCy7-anti-MHC-II (M5/114.15.2); FITC-anti-F4/80 (BM8); APC-anti-Ly6C (HK1.4); BV711-anti-CD115 (AFS98); APC-anti-CD44 (IM7); Biotin-anti-CD23 (B3B4); PeCy7-anti-CD62L (MEL-14); PeCy7-anti-CD86 (GL-1); PE-anti-CD80 (16–10A1); BV605-anti-CD69 (H1.2F3); and PE-Cy5-streptavidin (SA), which were purchased from BioLegend (San Diego, CA). Biotin-anti-CXCR5 (2G8); PeCy7-anti-CD95 (FAS, Jo2); FITC-anti-CD11c (HL3); APC-anti-CD5 (53–7.3); FITC-GL7; FITC-anti-CD11c (HL3) and AF700-anti-CD11b (M1/70), which were purchased from BD Pharmingen (San Diego, CA). Purified anti-SIGN-R1 (hamster Ab eBio22D1), PE-anti-PDCA1 (eBio927) and PE-anti-CD21/35 (eBio8D9) were purchased from eBioscience (San Francisco, CA) (Biotinylated Abs were detected by staining with Streptavidin-PeCy5 or SA-PE (BioLegend, San Diego, CA). Hamster antibodies were detected using PE-anti-hamster Ab (eBioscience, San Francisco, CA). Dead cells were excluded from analysis or quantitated by staining with Fixable Viability Dye (APCCy7) (eBioscience, San Francisco, CA). The following antibodies specific to humans were used: FITC-anti-human HLA-DR, DP, DQ (Tu-39, BioLegend, San Diego, CA), APC-anti-hMer (R&D Systems, Minneapolis, MN), PeCy7-anti-hCD86 (FUN1, BD Biosciences, San Jose, CA), APC-Cy7-anti-hCD11b (ICRF44, BioLegend, San Diego, CA). Data were acquired on a BD LSRII flow cytometer (BD Biosciences, San Jose, CA). Data were analyzed using FloJo software vs. 9 and 10 (Tree Star, San Carlos, CA).

2.6. ELISA

Immulon 4HBX high binding plates (ThermoFisher, Rockford, IL) were used for ELISA assays. For ANA ELISAs, plates were coated with a 1:10 dilution of poly-L-lysine (Sigma-Aldrich, St. Louis, MO), followed by coating with salmon sperm dsDNA (Sigma-Aldrich, St. Louis, MO) or Smith/Ribonucleoprotein SmRNP (Arotec Diagnostics, Wellington, New Zealand). Plates were blocked with 4% non-fat dry milk in PBS and coated with diluted serum samples with serial double dilution. Detection of antibody subtypes was performed using the following combinations of primary and secondary detection antibodies: primary- anti-IgG-biotin (Jackson ImmunoResearch, West Grove, PA), anti-IgG2b-biotin (Southern Biotech, Birmingham, AL), anti-IgG2c-alkaline phosphatase (Southern Biotech, Birmingham, AL) and secondary- streptavidin-alkaline phosphatase (Vector Laboratories, Burlingame, CA). Plates were developed with PNPP (p-nitrophenyl phosphate, disodium salt) (ThermoFisher, Rockford, IL) substrate and quantitation was performed as previously described [50]. Anti-DNA or anti-SmRNP IgG-subclass titers were determined by calculating arbitrary units by using serial dilution of the serum from a positive animal as a standard, and were expressed as units per volume.

2.7. HEp-2 ANA Assay

HEp-2 diagnostic slides were purchased from Antibodies Incorporated (Davis, CA). The assay protocol was followed according to manufacturer’s instructions. Briefly, serum was diluted 1:50 in PBS and incubated on HEp-2 slides. Detection of binding was performed by incubation with anti-kappa-FITC (H139–52.1) and slides were imaged on a Leica DM4000 fluorescent microscope and fluorescence intensity was quantitated in arbitrary units (Gray value) using the Leica application suite software.

2.8. REDD1 levels in human macrophages

Macrophages that had been treated with MSeA were incubated in RIPA buffer for 30 min on ice and processed for Western Blotting as described earlier [29]. Equal amounts of protein (50 μg) were separated on 10% SDS-PAGE gels and transferred to PVDF membranes. Primary antibodies against REDD1 (Proteintech, Chicago, IL) and β-actin (Santa Cruz Biotechnology, Santa Cruz, CA) were reacted at 1:1000 with blots. The HRP-conjugated anti-rabbit and anti-goat secondary antibodies (Cell Signaling, Danvers, MA) respectively were incubated with the blots at a dilution of 1:3000. Band expressions were developed using Pierce ECL reagents (ThermoFisher, Rockford, IL).

2.9. Statistical Analysis

Comparisons between multiple groups were performed by one-way ANOVA with multiple comparisons analysis, Tukey test. Comparisons between two groups were performed by Student’s t-test with Mann-Whitney analysis. p values less than or equal to 0.05 were considered to significant and significance was assigned according to the following breakdown: p <0.05= * p <0.01= ** p <0.001= *** p <0.0001= ****. GraphPad Prism 7 (Sand Diego, CA) was used for all statistical analysis.

3. Results

3.1. Selenium supplementation reduces B cell responses in lupus prone B6.Sle1b mice

In a previous study using the NZB/NZW-F1 female mouse model of SLE, it was observed that the life span of Se-supplemented mice was increased significantly compared to untreated mice [38]. However, the cellular responses affected by Se supplementation were not thoroughly studied.

To determine the type of Se compound and concentration that are effective in vitro and which might be used for in vivo administration, we tested a panel of Se compounds previously used in cancer studies [29, 49]. Bone marrow cells from B6/129SF2/J or C57BL/6 mice were cultured with L929 culture supernatant to allow their differentiation into macrophages over a period of 7 days, in the presence of various concentrations of Sel, SM, MSC and MSeA. The survival of mouse bone marrow-derived macrophages was assessed thereafter.

All the compounds tested did not adversely affect the survival of macrophages even after a continuous 7-day treatment (Supplementary Fig.1A, B). Amongst various Se compounds, MSeA has previously been described as a potent reactive inhibitor of breast and prostate cancer growth in vitro as well as in vivo with higher efficacy, compared to other forms such as Sel and SM [49]. Therefore, we further tested the effect of MSeA at various concentrations on the survival of activated B cells. We found that up to a concentration of 5.0 μM, the survival of activated B cells from WT mice was not adversely affected by MSeA treatment (Supplementary Fig.1C, D).

To evaluate the effect of Se on lupus prone B6.Sle1b mice, we began the in vivo oral supplementation of 2 mo old female B6.Sle1b mice with MSeA, or PBS (placebo) as described in the methods. PBS/ untreated B6.Sle1b (UT-B6.Sle1b) or Se-treated B6.Sle1b (Se-B6.Sle1b) mice were weighed at regular intervals during the three months of Se administration. In the beginning of the experiment Se-B6.Sle1b mice did not gain weight compared to the UT-B6.Sle1b. However, after about 4–5 weeks of treatment the body weight of Se-B6.Sle1b mice was similar to UT-B6.Sle1b mice (Fig. 1A). After 3 months of treatment mice were sacrificed and analyzed. UT-B6.Sle1b mice had greater spleen weight compared to 5 mo old WT mice, while the mean spleen weight of Se-B6.Sle1b mice was similar to WT mice, indicating a trend in the reduction of splenomegaly upon Se treatment (Fig. 1B). In accordance, we observed a significant reduction in the total number of splenocytes in Se-B6.Sle1b mice (Fig. 1C). Although not statistically significant, a decrease in total number of F4/80⁺CD11b⁺ splenic macrophages in Se-B6.Sle1b mice compared to UT-B6.Sle1b mice was observed (Fig. 1D). Most notable was a significant reduction in the total number of B cells (Fig. 1E), and B220⁺GL-7⁺CD95^hi splenic GC B cells (Fig. 1F and G), in Se-treated mice compared to UT controls. Unlike the splenic B cells and macrophages, no significant reduction in percent splenic monocytes (Fig. 2A), cDCs (Fig. 2B), pDCs (Fig. 2C), CD4⁺Tregs (Fig. 2D), neutrophils (Fig. 2E), eosinophils (Fig. 2F) or in the number of blood monocytes (Fig. 2G) was observed. Altogether, the in vivo analysis of Se-B6.Sle1b mice indicated a specific inhibitory effect of Se-treatment on the total numbers of B cells, GC B cells and macrophages, while T cells and other myeloid cells were unaffected.

Figure 1. Selenium supplementation reduces germinal center B cells in B6.Sle1b mice.

B6.Sle1b mice were administered PBS (UT) or MSeA orally starting at 2 mo of age. At 5 mo of age, mice were sacrificed and flow cytometry analyses were performed. (A) Determination of total body weight at indicated time points. (B) Analysis of spleen weight at 5 mo of age. (C) Total number of cells per spleen. (D) Comparative cytometric analysis of total macrophage (CD19^-CD3^-Ly6G^-Ly6C^-SSC^loCD11b⁺F4/80⁺) numbers per spleen (E) Cytometric analysis of total CD19⁺ B220⁺ B cells in the spleen (F) Representative flow plots showing gating strategy of B220⁺ GL-7^hi CD95^hi GC B cells. (G) Total number of B cells in GCs per spleen. Each symbol represents value in a mouse and horizontal lines indicate mean values. (NS, not significant; *, p<0.05).

Figure 2. Selenium supplementation does not affect monocytes, dendritic cells and granulocytes in B6.Sle1b mice.

B6.Sle1b mice were orally administered PBS (UT) or MSeA starting at 2 mo of age. At 5 mo of age mice were sacrificed and the percentages of cells were analyzed by flow cytometry. (A) CD11b⁺ SSC^loLy6C⁺ Ly6G^- Monocytes. (B) CD11c^hi MHC-II^hi-DCs (C) CD11c^int B220⁺PDCA1⁺-pDCs. (D) CD4⁺ CD25⁺ FOXP3⁺ Tregs. (E) CD11b⁺ SSC^hiLy6C⁺ Ly6G⁺ Neutrophils. (F) CD11b⁺ SSC^hiLy6C⁺ Ly6G^- Eosinophils. (G) CD11b⁺ SSC^loLy6C^hi Ly6G^- naïve monocytes and CD11b⁺ SSC^loLy6C^lo Ly6G^- - inflammatory monocytes in the blood.

3.2. Selenium supplementation reduces anti-dsDNA and anti-SmRNP IgG2c titers in B6.Sle1b mice.

To understand the effect of Se-treatment on autoimmunity in vivo, we bled the control and Se-treated B6.Sle1b mice every 30 days after the start of the experiment and analyzed their sera for anti-nuclear antibodies (ANAs) using the HEp-2 cell line-based ANA-assay. B6.Sle1b mice showed a perinuclear and nuclear staining pattern by HEp-2 ANA analysis, a characteristic-staining pattern associated with lupus (Fig. 3A). We observed a gradual increase in HEp-2 staining intensity with increasing age in UT-B6.Sle1b mice, however this increased intensity in Se-B6.Sle1b mice was only incremental with passage of time after treatment (Fig. 3A). After 3 mo of Se-treatment, we found an overall reduction in ANA responses in B6.Sle1b mice, although the difference was not statistically significant (Fig. 3B). To further address the reduction in autoantibody responses, we quantitated the anti-dsDNA and anti-SmRNP IgG titers in UT-B6.Sle1b vs. Se-B6.Sle1b mice. We observed a slight, statistically insignificant decrease in total anti-dsDNA (upper panel) and anti-SmRNP (lower panel) IgG titers (Fig.3C). This observation was similar to the earlier report [38].

Figure 3. Selenium supplementation reduces the production of autoantibodies in B6.Sle1b mice.

(A) Representative microscopy images of HEp-2 ANA analysis using sera from B6.Sle1b mice, untreated (UT) or at 1 mo, 2 mo and 3 mo after MSeA treatment. (B) Quantification of total ANA fluorescence intensity from 5 mice per indicated group. Each sample was analyzed in duplicate and average values were plotted. (C) Analysis of serum titers of total anti-dsDNA IgG (upper panel) and anti-SmRNP IgG (lower panel) from UT or MSeA-treated B6.Sle1b mice by ELISA. (D-G) Analysis of serum for anti-dsDNA IgG2b (D), anti-dsDNA IgG2c (E), anti-SmRNP IgG2b (F) and anti-SmRNP IgG2c (G) from UT or MSeA treated B6.Sle1b mice by ELISA, at the indicated time after MSeA treatment. Each symbol represents value in a mouse and horizontal lines indicate mean values for panels D-G.

Next, we assessed if the overall reduction in GC B cell numbers in Se-B6.Sle1b mice led to a reduction in the production of class switched autoAbs against dsDNA and SmRNP compared to untreated mice. Indeed, we observed that in accordance with a significant decrease in GC B cell responses there was a significant reduction in anti-dsDNA IgG2c, however, anti-DNA IgG2b did not show a significant reduction (Fig. 3 D, E). Notably, a highly significant reduction in anti-SmRNP IgG2b and IgG2c (Fig. 3F, G) autoAb titers was observed in Se-B6.Sle1b mice compared to control mice. These data indicate that a low dose oral supplementation of Se in lupus prone mice can lower the production of disease-associated class-switched autoAbs.

Overall, in vivo oral supplementation of Se as MSeA in B6.Sle1b mice preferentially reduces the GC B cell numbers and the titers of IgG2c anti-DNA and anti-SmRNP autoAbs without any other adverse effects on the health of the B6.Sle1b mice.

3.3. Selenium treatment decreases B cell activation in vitro.

To understand how Se supplementation reduces B cell responses including GC B cell differentiation and ANA-production, we tested the effect of MSeA on B cell activation in vitro. B cell activation is an essential first step for B cell dependent antigen presentation and B cell-T cell interactions, which in turn are necessary for GC and extra follicular B cell dependent production of pathogenic autoAbs. Purified B cells from WT mice were activated with anti-IgM and anti-CD40 in the presence or absence of MSeA and 24h later, cells were analyzed by flow cytometry for measurement of viability and the expression of surface activation markers. These included the surface co-stimulatory molecules CD69, CD80 (B7–1) and CD86 (B7–2) that play roles in regulating B and T-cell interactions and T-cell functional maturation [51].

As shown earlier, we did not observe a reduction or change in the overall B cell viability after a 24h Se-treatment. Additionally, no significant differences were observed in IgD (Fig. 4A) or MHC-II (Fig. 4B) expression on UT or Se-treated activated B cells. Upon activation there was a significant increase in the expression of CD69, CD80 and CD86 on B cells and a dose dependent decrease in the surface expression of these markers with increasing concentration of MSeA (Fig. 4C–D). At a 2.5μM concentration of MSeA (also administered in vivo), we observed very low expression levels of CD69, CD80 and CD86 on B cells, indicating that the overall activation of B cells is compromised upon Se-supplementation. To evaluate if suppression of B cell activation also inhibited antigen-presentation, we performed an in vitro antigen-presentation assay by co-culturing pre-activated ovalbumin (OVA)-loaded WT B cells as APCs and OVA-specific CFSE-labeled OT-II T cells, in the presence or absence of Se. The proliferation of OT-II T cells was quantified by flow cytometry after 72h of co-culture, as a measure of the antigen-presentation capacity of B cells. Interestingly, we did not observe a significant impact of Se on antigen-presentation by B cells. UT or Se-treated B cells were able to present OVA-antigen similarly to OT-II T cells (data not shown).

Figure 4. Selenium suppresses B cell activation in vitro.

Purified WT B cells were left unactivated (UA) or activated with anti-IgM and anti-CD40 for 24h in the absence or presence of indicated MSeA concentrations, after which the cells were analyzed by flow cytometry. gMFI of surface marker: (A) IgD, (B) MHC-II, (C) CD69, (D) CD80 and (E) CD86 expression on B cells. Data are representative of three independent experiments with similar results.

Overall, the data indicate that Se decreases B cell activation, which potentially contributes to the reduced GC responses and subsequent effects on antibody production in Se-B6.Sle1b mice.

3.4. Selenium treatment impairs optimal differentiation and maturation of macrophages in vitro.

The oral supplementation of B6.Sle1b mice with MSeA caused a measurable decrease in the number of macrophages in vivo (Fig. 1D), while there was no effect on number of DCs, pDCs and monocytes.

To understand the suppressive effect of Se-treatment on macrophages we differentiated WT BM cells into macrophages in vitro (BMDMs), in the absence or presence of various Se compounds at varying concentrations as indicated in Fig. 5. We observed that the viability and number of macrophages differentiating from BM cells were similar between untreated and Se-treated samples (data not shown). However, there were phenotypic differences in the differentiated macrophages upon Se-treatment. Macrophages present antigens and produce inflammatory cytokines when activated by lupus-associated immune complexes and in turn activate T cells, amplifying the inflammatory signals [52]. F4/80 and CD11b are expressed on a variety of myeloid cells including macrophages and increased expression of CD11b and several polymorphisms in CD11b have been associated with SLE [53–57]. CD86 is involved in proinflammatory responses and is reportedly upregulated in SLE [58, 59]. Similarly, MHC-II upregulation promotes increased antigen-presentation to T cells and is prominent molecule regulating the pathogenesis of SLE [60]. MSeA-treatment reduced the expression of CD11b, F4/80, CD86 and MHC-II in a dose-dependent manner (Fig. 5A). SM and MSC reduced the expression of CD86 and MHC-II at higher concentrations (Fig. 5B and C). However, we did not see an effect of Sel on the expression profile of any of the markers (Fig. 5D). A 24h treatment of BMDMs with various Se-compounds in the presence of 1μg/ml of LPS did not have a significant dose dependent effect on the survival or phenotype of macrophages (data not shown). These data indicate that certain Se compounds may reduce macrophage-dependent amplification of autoimmune responses by impairing macrophage differentiation and functional maturation.

Figure 5. Effect of various selenium compounds on macrophage differentiation and maturation.

Bone marrow cells were differentiated into macrophages in vitro, in the absence or presence of various Se compounds at the indicated concentrations. After 7 days of differentiation the macrophages were analyzed by flow cytometry. Effect of (A) MSeA, (B) SM, (C) MSC and (D) Sel on the surface expression of CD11b, F4/80, CD86 and MHC-II measured as geometric mean fluorescence intensity (gMFI). Data are representative of two independent experiments with similar results.

3.5. In vitro differentiated human macrophages show reduced REDD1 levels upon MSeA treatment under hypoxic conditions.

Our in vitro studies with murine BMDMs showed that Se treatment suppressed pro- inflammatory responses from macrophages. We have previously shown that Se-treatment induces apoptosis in prostate cancer cells through down regulation of HIF-1α [29]. HIF-1α is also responsible for inflammation in prostatitis leading to benign prostate hyperplasia [61] and so Se offers partial anti-inflammatory responses through downregulation of HIF-1α in a tumor environment. REDD1 is a downstream target of HIF-1α and is also implicated in inflammation involving macrophages [62]. Recently, it was described that the GC microenvironment is hypoxic [63, 64]. The GC generates high numbers of apoptotic B cells which are efferocytosed in a non-inflammatory manner by tingible body macrophages expressing MerTK and CD68. Inefficient efferocytosis in humans and mice is linked to excessive inflammation, dysregulation of GC responses and SLE [50, 65, 66].

To further understand the molecular basis of Se activity in macrophages, we investigated the effect of Se on human monocyte derived macrophages under hypoxic conditions. We differentiated human macrophages in vitro from CD14⁺ monocytes enriched using magnetic sorting of healthy human PBMCs, as described in methodology (Fig. 6A). The in vitro differentiated human macrophages expressed CD11b, CD86, MHC-II, CD68 and MerTK, which confirmed our optimal differentiation conditions (Fig. 6B). We observed a dose-dependent decrease in the expression of pro-inflammatory protein REDD-1 with increasing Se-concentration (Fig. 6C), suggesting that Se supplementation might reduce inflammation in the hypoxic microenvironment. Se-supplementation may prevent the expansion of dysregulated GCs in B6.Sle1b mice, but this finding needs further investigation.

Figure 6. Selenium suppresses REDD1 levels in human monocyte derived macrophages under hypoxic conditions.

(A) Human monocytes were purified by MACS sorting from PBMCs and differentiated into macrophages in vitro. (B) Histograms show surface expression of CD11b, CD86, MHCII, CD68 and MerTK on in vitro differentiated macrophages. (C) In vitro differentiated macrophages were incubated with MSeA under hypoxia (1% oxygen) and REDD1 protein levels were measured by Western blot. β-actin was used as a loading control for protein content.

4. Discussion

The role of Se has been extensively studied in human cancers, where an inverse relationship between in vivo Se-levels and the risk of developing cancer has been established. Se supplementation in animal models of cancer has shown that it either decreases the incidence of cancers, or improves the survival of mice with various cancers [27–29]. MSeA is a Se-containing synthetically produced compound which has been extensively studied as a chemotherapeutic agent in several cancers. MSeA in vivo is metabolized to produce methylselenol, which interacts with oxygen to generate superoxide. The mechanism of action of MSeA on cancer cells includes apoptosis induction through activation of caspases, ER stress leading to unfolded protein responses, cytochrome c release and PARP cleavage. Notably, orally administered MSeA was found to considerably reduce tumor growth without inducing significant systemic toxicity compared to Sel [49] and SM [67]. We found that MSeA potently inhibited the proliferation of B cells without affecting the survival of B cells. The selection of MSeA for in vivo administration was based on the aforementioned criteria for this study.

Unlike cancer, the role of Se in autoimmunity is not well understood. A previous study investigated the role of Se-supplementation in NZB/NZW-F1 mice and found improved survival of mice when Se supplementation was started at an early age [38]. Interestingly, our study on proteomic analysis of plasma from healthy males supplemented with selenized-yeast revealed that Se-supplementation increases the production of complement proteins [39]. C1q deficiency is one of the strongest genetic factors linked to SLE and complement proteins are essential for the clearance of apoptotic cells, the impairment of which has been reported in several SLE patients and SLE mouse models [68]. These data indicated a potential protective role of Se in SLE pathogenesis.

Self-reactive B cells generated through V(D)J recombination in the bone marrow or through accrual of random mutations during somatic hypermutation (SHM) in the germinal centers of secondary lymphoid tissues are mostly purged or edited to prevent anti-DNA responses [69]. The GC-tolerance checkpoint plays an essential role in maintaining peripheral tolerance, the loss of which leads to positive selection of self-reactive B cells, which contribute to the long-lived autoimmune plasma cells [14]. Long-lived plasma cells are refractory to most B cell therapies including cyclophosphamide, rituximab and anti-BAFF [70, 71]. Therefore, regulation of B cell responses at the GC tolerance checkpoint could be a more targeted way of managing plasma cell production for regulating autoantibodies. One of the aims of this study was to understand if Se supplementation could modulate GC-responses in SLE prone mice. For this purpose, we used B6.Sle1b mice that develop high titers of autoAbs due to an altered GC tolerance checkpoint, but they do not develop overt SLE pathology and inflammation. Thus, the B6.Sle1b mouse model can be used as a prototype to study predominant GC-dependent B cell responses in SLE.

In SLE-prone B6.Sle1b mice, we found that Se-supplementation significantly reduced the expansion of Spt-GCs. There was also a modest reduction in class-switched anti-DNA and anti-SmRNP antibodies of IgG2b and IgG2c subclass, which are preferentially produced in this SLE model through the GC-pathway. However, total anti-DNA IgG and anti-SmRNP IgG titers were unaltered, similar to that reported in a previous study where NZB/NZW-F1 mice were orally supplemented with Se in drinking water [38]. These data suggest that the GC-pathway of autoantibody production was specifically inhibited in the presence of Se causing a significant decrease in class-switched antibodies, while total IgG responses were unaltered, which may indicate that the extrafollicular pathway of autoantibodies production was not affected by Se-treatment in B6.Sle1b mice. Unlike GC B cells, the frequencies of myeloid cells and T cells were unaltered upon treatment with Se. The predominant suppressive effect of Se on GC B cells could be explained by the unique microenvironment within the GC and the idiosyncrasy of the GC B cells, which are the fastest proliferative mammalian cells, dividing every 6–12 hours [72]. We observed a suppressive role of Se on the activation of B cells in vitro and on GC B cell expansion in vivo – indicating that Se supplementation might specifically inhibit the proliferative function of GC BC cells in vivo thereby limiting the number of autoreactive B cells in B6.Sle1b mice. Further in vivo studies need to be performed to confirm this hypothesis.

Interestingly, the GC-light zone where the selection of antigen-specific B cells and the deletion of autoreactive B cells occurs has been shown to be hypoxic [63]. The light zone is also the specific subzone within the GC where B cells with reduced antigen affinity or self-reactivity are deleted. The purged B cells are promptly cleared within the GC’s by specialized macrophages expressing Mer tyrosine kinase and CD68, known as tingible body macrophages (TBMs) [50, 65, 66, 73]. TBMs are universally present in mice and human and are essential for the non-inflammatory clearance of apoptotic cells in the GCs. Apoptotic cell clearance deficiency in the GCs leads to SLE in mice and humans [66, 74]. Uncleared apoptotic cells can provide TLR-ligands like DNA and RNA which induce a proinflammatory environment and promote autoreactive B cell expansion. These studies outline the critical role of efferocytotic macrophages in the GC microenvironment. In our study we observed a reduction in REDD1 protein in human macrophages under hypoxic conditions, when Se was supplemented to the macrophages – suggesting a negative regulatory effect of increased Se on inflammation. This could be explained partially through increased glycolysis in macrophages due to reduced REDD1 expression in the GC microenvironment, similar to what has been reported in tumor associated macrophages [75], and which will be a focus of future experiments. Furthermore, our data indicate that Se-supplementation may reduce the oxidative stress in macrophages, thereby suppressing inflammatory responses within the GC microenvironment. This may in turn prevent proliferation and expansion of autoreactive GC B cells.

Overall, our data suggest that Se is inhibitory to differentiation and expansion of GC B cells, and suppresses inflammatory stimulation of macrophages – collectively inhibiting GC responses. This study is the first to show an inhibitory role of Se on GC B cells in vivo. Further studies are warranted to better understand the mechanisms governing these observations and to determine if Se-supplementation can be used alongside the canonical therapeutic interventions in SLE.

Supplementary Material

Supp1

Supplementary Figure 1. Selenium supplementation does not induce cell death in BMDMs and B cells. (A, B) Bone marrow cells were differentiated into macrophages in vitro in the absence or presence of indicated concentration and compound of Se. After 7 days of culture, macrophages were harvested and the cellular integrity of cells was assessed using fixable live/dead dye by flow cytometry. (A) Pseudo-color dot plots show the gating strategy and cellular profile after staining with live dead dye. (B) Quantification of the percentage of live macrophages after indicated treatments with shown Se compounds. (C, D) Purified B cells were left unactivated or activated with or activated with anti-CD40 and E. coli LPS for 24h in the presence or absence of MSeA at indicated concentrations. Cells were harvested after 24h, washed and stained with fixable live/dead dye for the analysis of live/dead cells by flow cytometry. (C) Pseudo-color dot plots show the gating strategy and cellular profile after staining of B cells with live/dead dye. (D) Quantification of the percentage of live B after indicated treatments. Data are representative of two or three independent experiments with similar results.

Non-standard abbreviations:

Spt-GC

spontaneous germinal center

ANA

anti-nuclear antibody

MSeA

methylseleninic acid

SLE

systemic lupus erythematosus

Se

Selenium