Fas^lpr gene dosage tunes the extent of lymphoproliferation and T cell differentiation in lupus

Abstract

Sle1 and Fas^lpr are two lupus susceptibility loci that lead to manifestations of systemic lupus erythematosus. To evaluate the dosage effects of Fas^lpr in determining cellular and serological phenotypes associated with lupus, we developed a new C57BL/6 (B6) congenic lupus strain, B6.Sle1/Sle1.Fas^lpr/+ (Sle1^homo.lpr^het) and compared it with B6.Fas^lpr/lpr (lpr^homo), B6.Sle1/Sle1 (Sle1^homo), and B6.Sle1/Sle1.Fas^lpr/lpr (Sle1^homo.lpr^homo) strains. Whereas Sle1^homo.lpr^homo mice exhibited profound lymphoproliferation and early mortality, Sle1^homo.lpr^het mice had a lifespan comparable to B6 mice, with no evidence of splenomegaly or lymphadenopathy. Compared to B6 monogenic lupus strains, Sle1^homo.lpr^het mice exhibited significantly elevated serum ANA antibodies and increased proteinuria. Additionally, Sle1^homo.lpr^het T cells had an increased propensity to differentiate into Th1 cells. Gene dose effects of Fas^lpr were noted in upregulating serum IL-1α, IL-2, and IL-27. Taken together, Sle1^homo.lpr^het strain is a new C57BL/6-based model of lupus, ideal for genetic studies, autoantibody repertoire investigation, and for exploring Th1 effector cell skewing without early-age lymphoproliferative autoimmunity.

1. Introduction

Systemic lupus erythematosus (SLE) is a chronic, prototypic autoimmune disorder which could result in damage to a wide range of organs, including skin, kidneys, and the neural system [1]. Besides heterogeneous clinical manifestations, the high complexity of SLE is also reflected by a variety of risk factors. Among them, multiple genetic factors have been associated with the predisposition to SLE [2]. Earlier studies revealed that certain MHC alleles are consistent high-risk factors in SLE patients, which could be related to MHC restriction for T cell autoantigens [3]. Further studies to link single-gene mutations with SLE susceptibility have identified serval high-risk polymorphisms, including those that cause dysregulations in DNA clearance, type I interferon (IFN), or lymphocyte activation pathways [2]. It pinpoints that self-antigen generation and hyper-activation of innate and adaptive immunity are the major players in SLE pathogenesis.

In addition to clinical studies, congenic murine models bearing lupus susceptibility loci have been widely employed to dissect the contributions of different genetic aberrations to manifestations of SLE [4]. Sle1 and Fas^lpr are two lupus susceptibility loci associated with the well-researched spontaneous lupus models, NZM2410 and MRL/lpr, respectively [5–7]. Subsequently, these two loci were reported to result in dysfunctions of the adaptive immunity and/or FAS/FASL-mediated apoptosis pathway, respectively [8, 9]. However, individually introducing (the NZM2410 allele of) Sle1 and Fas^lpr onto a lupus-resistant C57BL/6 (B6) background only led to mild autoimmune phenotypes. For example, renal pathology changes are largely negligible in B6.Fas^lpr/lpr (lpr^homo) mice before 9 months (mos) age [10, 11]. Previous studies further demonstrated the epistatic interaction of Sle1 and Fas^lpr results in severe lupus on the B6 background. High levels of anti-nuclear and anti-glomerular autoantibodies can be detected in 3 mos old B6.Sle1.Fas^lpr/lpr (Sle1^homo.lpr^homo) mice. This strain exhibits massive lymphosplenomegaly and glomerulonephritis as early as 2-3 mos of age, suggesting that Sle1 and Fas^lpr impact non-redundant pathways leading to lymphoproliferative autoimmunity [12]. While the Sle1^homo.lpr^homo strain has the potential to mirror human SLE by driving lymphoproliferative autoimmunity, most SLE patients exhibit mild or no clinical evidence of lymphoproliferation [13]. Given the dominant impact of the Fas/FasL axis on lymphocyte apoptosis [14, 15], we investigated whether Fas^lpr gene dosage on the B6.Sle1^homo background might play a role in the extent of lymphoproliferation in lupus models.

To address this, we comprehensively characterized lymphoproliferation status and cellular and serological phenotypes associated with lupus in a new congenic strain, B6.Sle1/Sle1.Fas^lpr/+ (Sle1^homo.lpr^het). Our findings suggest that the Sle^homo.lpr^het strain constitutes a novel C57BL/6-based model of lupus, characterized by middle-age onset and Th1 skewing, but without overt early-age lymphoproliferative autoimmunity. The requirement for only two genetic loci renders this novel strain useful for molecular genetics studies, unlike the more commonly used lupus strains such as BWF1 and MRL/lpr.

2. Materials and methods

2.1. Mice

C57BL/6 (B6) mice and B6 mice homozygous Fas^lpr/lpr were obtained from The Jackson Laboratory (#000482, Jackson Laboratory, Bar Harbor, ME). B6 congenic strains bearing the NZM2410-derived Sle1 gene interval and Fas^lpr were generated as previously described [9] and subsequently bred in a specific pathogen-free barrier facility at the University of Houston. Mice were handled following protocols approved by the Institutional Animal Care and Use Committees. Equal numbers of male and female mice were used, and any observed sex differences were indicated.

2.2. Anti-Fas induced apoptosis assay

Single-cell suspensions from spleen tissues were prepared and depleted of red blood cells by ACK lysing buffer (#A1049201, ThermoFisher Scientific, Waltham, MA). Isolated splenocytes were stimulated with 100ng/ml of anti-Fas antibody (#152803, BioLegend, San Diego, CA) in the presence of 50ng/ml of PMA (#P1585, Sigma-Aldrich, St. Louis, MO) and 1μg/ml of Ionomycin (#I0634, Sigma-Aldrich) for 4 hours (hrs). After stimulation, cells were stained with antibodies targeting CD4, CD8, CD19, and cleaved caspase-3. Cleaved caspase-3 percentages in CD4⁺, CD8⁺ T cells and B cells were determined by flow cytometry (Details in section 2.7).

2.3. Serum collection and analyses

100μl of serum samples were collected from experimental mice at different time points and stored at −80°C for batch analyses to determine anti-double-stranded DNA (anti-dsDNA) autoantibody and cytokine/chemokine levels. The anti-dsDNA ELISA assays were performed by using anti-mouse IgG (H+L) antibody (#115-035-003, Jackson ImmunoResearch Inc., West Grove, PA) or anti-mouse IgG1 antibody (#A10551, ThermoFisher Scientific), as previously described [9]. To measure cytokine/chemokine levels in serum collected from congenic lupus strains, a Luminex assay using the ProcartaPlex^™ Mouse Cytokine & Chemokine Panel 1 36-plex kit (#EPXR360-26092-901, ThermoFisher Scientific) was performed according to manufacturer’s instructions. A Luminex 200 Xmap analyzer (Luminex Corporation, Austin, TX) was used to determine the serum levels of tested soluble factors. Blood urea nitrogen (BUN) was also determined in the same serum samples using a urea colorimetric assay kit (#MAK006, Sigma-Aldrich).

2.4. Autoantigen screens

The autoantigen array study was carried out at the Houston Omics Collaboratives located at the University of Houston. Serum samples were collected from various strains and interrogated using a 128 autoantigens planar array, as previously described [16]. Briefly, serum samples were diluted at 1:50. The diluted samples were then added to the autoantigen array that includes 128 autoantigens. Detection of autoantibodies was performed using Cy3-labeled anti-mouse IgG (#115-166-072, Jackson ImmunoResearch Inc., West Grove, PA) and AF633-labeled anti-mouse IgM (#A21046, ThermoFisher Scientific) with a GENEPix 4300 microarray reader. Raw files and median fluorescent intensity were generated using GENEPix 7.3.0.0 software. The antigens were printed in duplicate, and mean value of each antigen was calculated for each sample. The data for each sample were adjusted by subtracting the negative control values, and these adjusted values were used for further statistical analyses.

2.5. Urine sample collection and analyses

Experimental mice were individually housed in mouse metabolic cages (#MM030505-10R, Lenderking Caging Products, Millersville, MD) for 24 hrs to collect urine samples. The 24-hr urine volume was recorded, and urine samples were stored at −80°C and analyzed to determine urinary protein concentration by a Bradford assay using the Pierce detergent compatible kit (#23246, ThermoFisher Scientific). The 24-hr urine protein amount was calculated based on 24-hr urine volume and total urinary protein concentration.

2.6. Histopathology

Mice were sacrificed to collect kidneys. Kidney samples were Formalin-Fixed and Paraffin-Embedded (FFPE). FFPE tissue slides were prepared and stained with Hematoxylin and Eosin (H&E). Sections were analyzed using a light microscope and scored for evidence of glomerulonephritis, and tubulo-interstitial disease, as detailed previously [17]. Kidneys were critically analyzed for any signs of cell proliferation and infiltration, thickening, and fibrous changes.

2.7. Flow cytometry analyses

For flow cytometric analyses, splenocytes and cells from inguinal lymph nodes were washed twice with staining buffer and incubated with a cocktail of antibodies targeting surface markers at 4°C for 30 min. Cells were then fixed and permeabilized using either the Foxp3/transcription factor staining buffer set (#00-5523-00, ThermoFisher Scientific) or the BD fixation/permeabilization solution kit (#554714, BD BioSciences, Franklin Lakes, NJ) according to the manufacturer’s protocols, and were incubated with a cocktail of antibodies against intracellular markers. The antibodies used for staining included anti-CD3-PE (#555275, BD Biosciences), anti-CD4-eFluor 450 (#48–0042–82, eBioscience, San Diego, CA), anti-CD8-PE/Cy7 (#60–0081, TONBO Biosciences, San Diego, CA), anti-CD25-PerCP (#65-0251, TONBO Biosciences), anti-Foxp3-PE (#12-5773-82, ThermoFisher Scientific), anti-CD19-APC (#550992, BD Biosciences), anti-CD19-PerCP (#65-0193, TONBO Biosciences), anti-CD19-FITC (#35-0193 TONBO Biosciences), anti-Ki67-FITC (#11-5698-82, ThermoFisher Scientific), anti-Ki67-PE/Cy7 (#25-5698-82, ThermoFisher Scientific), anti-CD62L-PerCP (#65-0621, TONBO Biosciences), anti-CD44-FITC (#35-0441, TONBO Biosciences), anti-CXCR5-APC/Cy7 (#145525, BioLegend), anti-NK1.1-APC (#550627, BD Pharmagen, San Diego, CA), anti-CD11b-eFluor 450 (#48-0112-82, eBioscience), anti-CD11c-APC (#20-0114, TONBO Biosciences), anti-F4/80-PE (#50-4801, TONBO Biosciences), anti-Ly6G-PE/Cy7 (#60-1276, TONBO Biosciences), anti-Ly6C-PerCP (#45-5932-82, ThermoFisher Scientific), anti-CD5-PE (#50-0051, TONBO Biosciences), anti-CD21-APC (#123412, BioLegend), anti-CD23-PB (#101616, BioLegend), anti-CD93-PE (#136504, BioLegend), anti-IgM-FITC (#406506, BioLegend) and, anti-cleaved caspase-3-Alexa Fluor 647 (#560626, BD Biosciences). An LSRFortessa X-20 cell analyzer (BD Biosciences) was used for acquisition. The gating strategies for major immune subsets are listed in the Supplementary figure 1.

2.8. T cells differentiation

Naïve CD4⁺ T cells were isolated from single suspensions of spleen tissues by negative selection using the EasySep Mouse CD4⁺ T cell Isolation Kit (#19852, StemCell Technologies, Cambridge, MA) and differentiated into different T helper (Th) subsets. Purified naïve CD4⁺ T cells were stimulated with 2.5μg/ml of anti-CD3 (#70-0031, TONBO Bioscience) and 1μg/ml of anti-CD28 (#553294, BD Biosciences) antibodies in different types of culture medium for T cell differentiation and proliferation [18–20]: (a) Th1 condition: MEMα medium (#32-571-101, ThermoFisher scientific) supplemented with 10% FBS (#S11150, R&D Systems, Minneapolis, MN), 100U/ml of IL-2 (#NDC 65483-116-07, Prometheus Laboratories, San Diego, CA), 4ng/ml of IL-12 (#210-12, PeproTech, East Windsor, NJ), and 10μg/ml of anti-IL-4 mAb (#BE0045, BioXCell, Lebanon, NH); (b) Th2 condition: MEMα medium supplemented with 10% FBS, 100U/ml of IL-2, 10ng/ml of IL-4 (#214-14, PeproTech), and 10μg/ml of anti-IFN-γ mAb (#BE0055, BioXCell); (c) Th17 condition: MEMα medium supplemented with 10% FBS, 100U/ml of IL-2, 30ng/ml of IL-6 (#216-16, PeproTech), 20ng/ml of IL-1β (#211-11B, PeproTech), 10μg/ml of anti-IFN-γ mAb, 2.5ng/ml of TGF-β1 (#763104, BioLegend).

To determine the percentage of differentiated Th1, Th2, and Th17, differentiated T cells were restimulated with 50ng/ml of PMA and 1μg/ml of Ionomycin for 1 hr, with blockade of cytokine secretion being achieved by simultaneous treatment with 4μl of GolgiStop (#51-2092KZ, BD Biosciences) for every 6ml of cell culture for the last 3 hrs. Restimulated T cells were stained with panels of antibodies for flow cytometric analysis as described previously. Antibodies used for this experiment included anti-IFNγ-APC (#20-7311, TONBO Bioscience), anti-IFNγ-PE (#50-7311, TONBO Bioscience), anti-IL-4-PE (#504101, BioLegend), anti-IL-4-PB (#48-0042-82, eBioscience), and anti-IL-17A-FITC (#11-7177-81, eBioscience). To measure cytokine production by differentiated T cells, T cells were re-seeded in a new 96-well plate and incubated overnight with fresh culture medium in the presence of 50ng/ml of PMA. The concentration of IFN-γ, IL-4 and IL-17 in the culture supernatant was determined using ELISA kits from R&D systems DY485, DY404, and DY421, respectively.

2.9. Statistical analyses

Summary statistics (e.g., mean, SEM) of the data are reported. Assessments of differences in continuous measurements between the two groups were made using the two-sample t-test. Either one-way or two-way ANOVA was performed to determine the significance in various subsets. The Kaplan-Meier method and log-rank test were used to compare survival between groups. p-value of less than 0.05 was considered significant. Data acquisition and analysis from the cell analyzer were executed on FACSDiva and FlowJo softwares (BD Biosciences). Graph generation and statistical analyses were performed using GraphPad Prism software.

3. Results

3.1. Sle1^homo.lpr^homo mice, but not Sle1^homo.lpr^het mice, exhibited lymphosplenomegaly with reduced survival.

To further characterize the interaction between the two lupus-associated loci, Sle1 and Fas^lpr, on regulating autoimmune responses in the B6 background, B6.Sle1/Sle1 (Sle1^homo) mice were crossed with B6.Fas^lpr/lpr (lpr^homo) mice over two generations and selected for B6.Sle1/Sle1.Fas^lpr/+ (Sle1^homo.lpr^het) and B6.Sle1/Sle1.Fas^lpr/lpr (Sle1^homo.lpr^homo) mice. Although our focus in this report is the gene dose effect of Fas^lpr on autoimmunity in the B6.Sle1^homo background, we included the lpr^homo strain whenever samples are available for future broader interests. As illustrated in Figures 1A and 1B, weights and cell numbers of spleens and inguinal nodes in Sle1^homo.lpr^homo mice at 2-3 mos of age were significantly higher than those in B6 and other tested B6 congenic mice. We observed 100% enlarged lymph nodes in Sle1^homo.lpr^homo mice during their lifespan (Fig 1C). All tested Sle1^homo.lpr^homo mice achieved primary experimental endpoint before or at 10 mos of age, marked primarily by severe lymphadenopathy (long diameter of any lymph node >15mm; Fig 1D). In contrast, enlarged spleens and lymph nodes were not noted in Sle1^homo.lpr^het, lpr^homo and Sle1^homo mice. The survival curves from these two B6 congenic strains were comparable to the wild-type B6 strain. These results indicate that the complete loss of Fas function is required to elicit lymphosplenomegaly and accelerated mortality in B6 mice bearing the Sle1 locus.

Figure 1. Sle1^homo.lpr^homo, but not Sle1^homo.lpr^het mice, exhibit prominent splenomegaly, lymphadenopathy, and reduced survival.

(A) Representative image of spleens and inguinal lymph nodes from B6 mice and B6 congenic mice at the age of 3 months (mos). The B6 and NZM240 alleles of Sle1 are indicated as the b and z alleles, respectively. The numbers of Lpr alleles in Fas loci are also listed for each B6 congenic strain. (B) Weight and total cell numbers in the spleens and inguinal lymph nodes. (C) Lymphadenopathy percentage. The lymph node size and health conditions of experimental mice were monitored at least twice per week. Mice with visible lymph nodes were counted as having lymphadenopathy. The endpoints include any lymph node >1.5 cm or severe health issues. (D) Survival data. Tested B6 congenic mice include Sle1^homo, lpr^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo. N=4-6/group; Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

3.2. Partial loss of Fas function is sufficient to lose B cells tolerance and accelerate autoimmune responses in B6.Sle1 congenic mice.

We first investigated whether the loss of Fas function would alter the sensitivity of lymphocytes to Fas-induced apoptosis in various lupus-prone strains. Splenocytes from B6 (WT control), Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo mice were stimulated with anti-Fas antibody in the presence of PMA and Ionomycin for 4 hrs, followed by flow cytometry analysis to determine apoptosis rate in lymphocytes. Our results demonstrated that the rate of Fas-induced lymphocyte apoptosis in WT and Sle1^homo are similar (Fig 2A). As expected, a significant reduction was observed in the rate of Fas-induced apoptotic activity of CD8⁺ T cells and B cells in both Sle1^homo.lpr^het and Sle1^homo.lpr^homo mice when compared to B6 WT and Sle^homo strains (Fig 2A). These results indicate that Fas dysfunction in CD8⁺ T cells and B cells in the B6.Sle1^homo background requires only heterozygosity for the Fas mutation. Interestingly, reduced apoptotic activity in CD4⁺ T cells was only noted in Sle1^homo.lpr^homo mice, and not in other strains (Fig 2A). This suggests that the threshold of Fas expression required for Fas-induced apoptosis might be lower in CD4⁺ T cells compared to CD8⁺ T cells and B cells. In other words, the Sle1^homo.lpr^het and Sle1^homo.lpr^homo strains may both have a loss of B cell tolerance, but only the latter experiences a loss of CD4⁺ T cell tolerance.

Figure 2. Impact of Fas^lpr zygosity on the rate of Fas-induced apoptosis in lymphocytes and the levels of serum autoantibody.

(A) Splenocytes were collected from various strains and stimulated with anti-Fas antibody in the presence of PMA plus Ionomycin for 4 hours (hrs). After incubation, cells were stained with a panel of antibodies to detect CD4⁺, CD8⁺, B cells, and cleaved caspase-3. Cleaved caspase-3 percentages in CD4⁺, CD8⁺ T cells and B cells were determined by flow cytometry. N=3-4/group. (B) Levels of total IgG autoantibodies to dsDNA in the serum of B6 and B6 congenic mice. Serum were collected from experimental mice at three different ages: 2-3 mos, 4-8 mos, and 10-12 mos (Except for the Sle1^homo.lpr^homo mice including only 2-3 mos) and examined for anti-dsDNA IgG by ELISA. N= 5-6/group. (C-D) IgG and IgM autoantibodies levels in the serum of B6 and B6 congenic mice at 2-3 mos of age were measured using autoantigen arrays. (C) Serum IgG autoantibody levels against ssDNA, genomic DNA, AQP4, TIF1-γ, and TLR4. (D) Serum IgM autoantibodies levels against TLR4. Tested B6 congenic mice include lpr^homo, Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo. N=3-5/group. Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

Next, we examined whether complete loss of Fas function is required to promote autoimmune responses in B6 mice bearing the homozygosity of the Sle1 locus. Anti-nuclear antibody (ANA) levels in B6 and B6 congenic mice serum were examined at three different ages: 2-3 mos, 4-8 mos, and 10-12 mos. Given that the mortality in Sle1^homo.lpr^homo mice older than 4 months is high, and this strain has been well characterized previously [12], we only included 2-3 mo-old Sle1^homo.lpr^homo mice in comparative studies. Compared with age-matched B6 controls, 2-3 mo-old Sle1^homo.lpr^homo mice exhibited significantly elevated serum anti-dsDNA of total IgG and IgG1 levels, while marginal increases were found in Sle1^homo and Sle1^homo.lpr^het mice (Fig 2B, Supplementary Fig 2, and Supplementary Tab 1). Moreover, serum ANA levels in Sle1^homo.lpr^het mice at the age of 4-8 mos were significantly higher than those in Sle1^homo and B6 controls, and Sle1^homo mice also started to show elevated serum ANA levels at 10-12 mos of age (Fig 2B and Supplementary Fig 2). Our results suggest that partial loss of Fas function promotes autoimmune response and accelerates the earlier emergence of serum ANA in B6 mice congenic for Sle1 locus.

Furthermore, we examined the levels of IgG and IgM antibodies targeting 128 autoantigens in B6 WT control, lpr^homo, Sle1^homo, Sle1^homo.lpr^het and Sle1^homo.lpr^homo mice. At the age of 2-3 mos, the Sle1^homo.lpr^homo strain exhibits increased IgG/IgM serum levels targeting 111/106 of tested autoantigens, respectively (Supplementary Tab 2 and 3). In contrast, the Sle1^homo.lpr^het strain demonstrates statistically elevated serum levels of 7 specificities of IgG and 2 specificities of IgM autoantibodies only(Supplementary Tab 2 and 3). However, the autoantibodies levels in the Sle1^homo strain are largely comparable to those in B6 WT mice at the same age (Supplementary Tab 2 and 3). Consistent with the ANA profiles, we observed a gene dose effect of Fas on the IgG levels of autoantibodies against AQP4, TIF1-γ, and TLR4 and the level of IgM autoantibodies against TLR4 at the age of 2-3 mos (Fig 2C and 2D). These results further support our conclusion that both Sle1^homo.lpr^het and Sle1^homo.lpr^homo strains exhibit an early onset of autoimmune responses, with a more profound loss of tolerance in Sle1^homo.lpr^homo mice, once T cell tolerance is compromised.

3.3. Complete loss of Fas is required for severe renal pathology in the B6.sle1^homo background.

Next, we compared renal functional changes in B6 and B6 congenic mice based on urine protein levels and serum BUN concentrations in the above-mentioned three age groups. Increased proteinuria was observed in Sle1^homo.lpr^homo, Sle1^homo.lpr^het, and Sle1^homo mice starting at the age of 2-3 mos, 4-8 mos, and 10-12 mos, respectively (Fig 3A and Supplementary Tab 4). Higher levels of BUN were found in all B6 congenic strains than the B6 controls in all three age groups, with the phenotype being more prominent in Sle1^homo.lpr^homo (Fig 3B and Supplementary Tab 5). Detailed renal pathology changes in B6 WT, lpr^homo, Sle1^homo, and Sle1^homo.lpr^het at 4-8 mos of age are summarized in Supplementary Table 6. Within this age group, Sle1^homo and lpr^homo mice have no signs or have mild mesangial cellular proliferation (Fig 3C and Supplementary Tab 6). However, mesangial cellular proliferation, foci of inflammatory cell infiltrates, and fibrous tissue formation in glomerular and peri-glomerular regions can be observed in Sle^homo.lpr^het mice (Fig 3C and Supplementary Tab 6). Consistent with previous studies [12], Sle1^homolpr^homo mice at 2-3 mos old display segmental glomerulosclerosis, mesangial expansion and proliferation, and hypercellular capillary loops (Fig 3C). These findings confirm that homozygous or complete loss of Fas, leading to loss of T cell tolerance, is required for more severe renal pathology.

Figure 3. Impact of Fas^lpr zygosity on the levels of urine protein, BUN, and renal pathology.

(A) Total urine protein amount. The 24-hr urine volume and protein concentration were measured and used to calculate the total amount of 24-hr urine protein. N=5-6/group. (B) Blood urea nitrogen (BUN) levels in B6 and B6 congenic mice. Tested B6 congenic mice include Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo at three different ages: 2-3 mos, 4-8 mos, and 10-12 mos (Except for the Sle1^homo.lpr^homo mice including only 2-3 mos). N=4-14/group. Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0. 0001. (C) Representative images of renal tissue slides with H&E staining from B6 and B6 congenic strains. Kidney tissues from 4-8 mo-old B6 control, lpr^homo, Sle1^homo and Sle1^homo.lpr^het mice, and 2-3-mo-old Sle1^homo.lpr^homo mice are shown.

3.4. Profound alterations in the lymphocyte compartment of peripheral lymphoid organs are observed in Sle1^homo.lpr^homo mice, but not Sle1^homo.lpr^het mice.

To characterize the immune profiles of peripheral lymphoid organs in B6 and B6 congenic strains, spleens and inguinal lymph nodes from 3-mo-old female and male mice for each strain were collected and examined. As lpr^homo mice have been well characterized [10, 21] and have no sign of glomerulonephritis at the age of 4 mos (Supplementary Tab 6), we focused on B6 congenic strains with homozygous Sle1 locus for immune profiling. Phenotypic analysis of splenocytes revealed significant alterations in the cellular composition of major immune cells only in Sle1^homo.lpr^homo mice, including a reduced percentage of CD8⁺ T cells and B cells, when compared to the B6 controls (Fig 4A, 4B and Supplementary Fig 3). Whereas the major changes in absolute numbers in spleens from Sle1^homo.lpr^homo mice included significantly increased CD4⁺ T cells and other non-lymphoid cells (Supplementary Tab 7). The changes in the absolute numbers of immune cells in Sle1^homo.lpr^het mice were minor or not statistically significant (Supplementary Tab 7). Consistent with published literature [12], Sle1^homo.lpr^homo mice exhibited a dramatic increase in splenic CD3⁺ CD4⁻CD8⁻ double negative (DN) T cells (Fig 4A and Supplementary Fig 3A). However, the percentage of DN T cells in Sle^homo.lpr^het mice is comparable to WT mice. These data suggest that homozygous loss of Fas, leading to loss of T cell tolerance, is required for the emergence of DN T cells. Furthermore, we examined subsets within splenic CD8⁺ T cell, CD4⁺ T cell, and B cell populations. Both CD8⁺T and CD4⁺T cells in Sle1^homo.lpr^homo spleens were skewed towards effector and memory subsets with three-fold lower naïve T cell percentage compared to the B6 controls (Fig 4A). In addition, Sle1^homo.lpr^homo mice exhibited significantly increased Tfh percentages in spleen tissues (Fig 4A). There were no major changes in other CD4⁺ T cell subsets among the remaining B6 congenic strains and B6 controls (Fig 4A). Unlike T cells, alterations in the B cell compartment in Sle1^homo.lpr^homo mice were limited to slightly reduced B1 and immature B cells (Fig 4B). In contrast, the composition of B cell subsets in Sle1^homo.lpr^het was largely comparable with B6 controls and Sle1^homo mice (Fig 4B). Moreover, we determined the proliferation status of different immune cells based on the expression of Ki67, a proliferation marker. Both Sle1^homo.lpr^homo and Sle1^homo.lpr^het spleens revealed an increased percentage of proliferative CD8⁺ and CD4⁺ T cells with more prominent phenotypes being observed in Sle1^homo.lpr^homo spleens (Fig 4C and 4D). However, no changes in the percentage of proliferative B cells were observed (Fig 4E). The changes in non-lymphoid cell compartments are largely limited in Sle1^homo.lpr^homo mice showing increased proliferation (Fig 4F and 4G). Furthermore, homozygosity in both Sle1 and Fas^lpr resulted in similar changes in the T cell, B cell, and non-lymphoid cell compartments in the lymph nodes, just as those observed in the spleen, while the Sle1^homo.lpr^het mice consistently displayed mild phenotypic changes in all immune cells in the lymph nodes (Fig 5A–G; Supplementary Fig 3B and Supplementary Tab 8). Interestingly, unlike moderately increased T cell proliferation observed in spleen tissues, the proliferation of CD8⁺ and CD4⁺ T cells in lymph nodes from Sle1^homo.lpr^het remain unchanged compared to the B6 controls (Fig 5C and 5D). In contrast, T and B cells in lymph nodes from Sle1^homo.lpr^homo mice displayed significantly increased proliferative status (Fig 5C–E).

Figure 4. Immune profiling of splenocytes from B6 (Control), Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo mice.

Splenocytes were isolated from spleens of B6 and B6 congenic mice at 3 mos of age and stained with a panel of fluorescence-conjugated antibodies. Flow cytometry was used to characterize the abundance and proliferation of immune cells. (A) Abundance of T cell subsets, (B) abundance of B cell subsets, (C) proliferation of CD8⁺ T cells, (D) proliferation of CD4⁺ T cells, (E) proliferation of B cells, (F) abundance of non-T/B immune subsets, and (G) proliferation of non-T/B immune subsets among different strains are depicted. Tested B6 congenic mice include Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo. B6 mice were used as a control. N=4-5/group. Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0. 0001.

Figure 5. Immune profiling of lymph nodes from B6 (Control), Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo mice.

Single-cell suspension samples were prepared from inguinal lymph nodes of B6 and B6 congenic mice at 3 mos of age and stained with a panel of fluorescence-conjugated antibodies. Flow cytometry was used to characterize the abundance and proliferation of immune cells. (A) Abundance of T cell subsets, (B) abundance of B cell subsets, (C) proliferation of CD8⁺ T cells, (D) proliferation of CD4⁺ T cells, (E) proliferation of B cells, (F) abundance of non-T/B immune subsets, and (G) proliferation of non-T/B immune subsets among different strains are depicted. Tested B6 congenic mice include Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo. B6 mice were used as a control. N=4-5/group. Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0. 0001.

3.5. Gene dose effects of Fas^lpr regulates serum cytokine levels.

Given that several cytokines, such as IL-6, IL-10, IFN-γ and TNF-α, have been implicated in SLE pathogenesis and are associated with SLE clinical activity and serological changes [22], we next determined the serum levels of 36 cytokines in B6 and B6 congenic mice at the age of 2-3 mos. Among the tested cytokines, serum levels of 32 cytokines were significantly higher in Sle1^homo.lpr^homo mice than those in B6 controls (Supplementary Tab 9). Unlike Sle1^homo.lpr^homo mice, Sle1^homo.lpr^het mice exhibited fewer significantly elevated serum cytokine levels, including IL-1α, IL-2, IL-22, IL-23, IL-27, IL-4, IL-9, IP-10, ENA-78, MIP-2 and RANTES relative to the B6 controls (Supplementary Tab 9). While the majority of these serum elevations in Sle1^homo.lpr^het are shared by Sle1^homo and/or lpr^homo mice, increased IL-1α, IL-2 and IL-27 are only observed in B6 mice with both Sle1 and Fas^lpr loci, with homozygosity at Sle1 and Fas^lpr resulting in more profound elevations (Fig 6A). Furthermore, we also observed another group of cytokines whose levels were primarily affected by homozygosity at Fas^lpr. Specifically, lpr^homo and Sle1^homo.lpr^homo mice, but not Sle1^homo and Sle1^homo.lpr^het mice showed significantly increased IL-3, IL-10, IL-18, IL-28, and IL-31(Fig 6B and Supplementary Tab 9). Homozygosity at Sle1 results in more profound elevations in this group of cytokines. The interaction effect of Fas^lpr and Sle1 in serum cytokine profiles provides further evidence that Fas^lpr and Sle1 have non-redundant roles in promoting autoimmune responses.

Figure 6. Impact of Fas^lpr zygosity on serum cytokine/chemokine levels.

Cytokine/chemokine levels in serum from B6 and B6 congenic mice at 2-3 mos of age were measured by Luminex assays. (A) IL-1α, IL-2, and IL-27 levels in serum show phenotype driven by Fas^lpr zygosity, in concert with Sle1 homozygosity. (B) Cytokine levels in serum show phenotype driven by homozygosity of Fas^lpr. Tested B6 congenic mice include lpr^homo, Sle1^homo, Sle1^homo.lpr^het, and Sle1^homo.lpr^homo. N=5-6/group. Mice from both sexes were included. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

3.6. Skewed Th1 differentiation in Sle1^homo.lpr^het mice.

Among the limited cytokines elevated in Sle1^homo.lpr^het mice, IL-2 and IL-27 have been reported to be indicators of Th1-type immune response [23, 24]. Thus, we hypothesized that heterozygosity at Fas^lpr on the B6.Sle1^homo background may alter CD4⁺ Th cell differentiation, which may contribute to autoimmune responses without the profound effect on lymphoproliferation. To test this hypothesis, we isolated naïve CD4⁺ T cells from B6, Sle1^homo, and Sle1^homo.lpr^het mice, stimulated with anti-CD3 and anti-CD28 antibodies, then cultured T cells in different types of medium to polarize towards Th1, Th2, and Th17 differentiation. Phenotypes of differentiated CD4⁺ T cells were determined by the production of IFN-γ, IL-4, and IL-17, key effector cytokines for Th1, Th2, and Th17 cells, respectively. An increased percentage of IFN-γ⁺ cells in T cells under Th1 polarization condition was observed in Sle1^homo.lpr^het mice (Fig 7A and 7B). There were no significant changes in IL-4⁺ cells under the Th2 polarization and no change in IL-17⁺ cells under the Th17 polarization condition among these three strains (Fig 7A and 7B). Consistent with the flow cytometry-based analysis, ELISA results demonstrated that CD4⁺ T cells from Sle1^homo.lpr^het mice in the presence of Th1 polarization medium produced more IFN-γ than counterpart controls from B6 and congenic Sle1^homo mice (Fig 7C). However, IL-17 production by Th17-differentiated in Sle1^homo and Sle1^homo.lpr^het T cells were significantly lower than those from B6 controls (Fig 7C). Elevated serum levels of Th1-related cytokines such as IL-12(p70), IFN-γ and IL-27, but not Th2-related cytokines including IL-4 and IL-5, were noted in Sle^homo.lpr^het mice at the age of 4-8 months compared to WT or Sle^homo mice (Supplementary Fig 4). Together, our data suggest that partial loss of Fas function on the B6.Sle1^homo background preferentially drives differentiation of naïve CD4⁺ T cells towards Th1 cells upon TCR stimulation.

Figure 7. T cell differentiation changes in Sle1^homo and Sle1^homo.lpr^het mice.

Naïve CD4⁺ T cells were isolated from the spleens of B6, Sle1^homo and Sle1^homo.lpr^het and stimulated with 2.5μg/ml of anti-CD3 and 1μg/ml of anti-CD28 with IL-2 under culture conditions for Th1, Th2, and Th17 differentiation. Cultured T cells were restimulated and used to evaluate IFN-γ, IL-4, and IL-17 production by flow cytometry analysis and ELISA. (A) Representative flow cytometry results, (B) Bar plots showing the percentage of IFN-γ⁺, IL-4⁺, and IL-17⁺ T cells in different culture conditions. (C) Levels of IFN-γ, IL-4, and IL-17 in culture supernatant in restimulated T cells from different culture conditions. N=3/group. *p<0.05, **p<0.01, ***p<0.001 and ****p<0.0001.

4. Discussion

A broad range of immunological dysfunctions involving both innate and adaptive immune systems have been identified in SLE patients [2], including abnormal serum levels of cytokines [25], altered complement activation [26], production of pathogenic autoantibodies [27] and hyperactive of T and B cells [28], as recently reviewed [29]. Although many murine models have been generated to study the molecular and cellular mechanism of lupus [7], there is no single experimental model that mirrors all phenotypes present in human SLE. Introducing lupus susceptibility loci on the B6 background has resulted in establishing a series of B6 congenic lupus strains [4, 9, 12, 30–34] with distinct phenotypes. While B6 monocongenic lupus strains exhibit limited or no autoimmune symptoms or renal diseases, introgression of multiple lupus susceptibility loci leads to massive production of autoantibodies and fatal lupus nephritis as previously reported on the B6.sle1.sle2.sle3 and B6.Sle1^homo.lpr^homo strains [12, 32]. However, these two strains display severe hyperproliferation of T and B cells. Therefore, phenotype characterization of mice bearing one or multiple lupus susceptibility loci, particularly in the B6 congenic background, can not only identify the link between genetic alterations and immunological dysfunctions but also provide a series of congenic lupus models to better represent the kaleidoscope of clinical manifestations in human SLE.

Whereas Fas^lpr impairs cell apoptosis and activation-induced cell death, Sle1 breaches B cell tolerance, leading to the emergence of a chromatin-targeted autoantibody repertoire, insufficient to elicit significant renal damage [5, 6, 35–37]. Here, we test whether impairing lymphocyte apoptosis (using the Fas^lpr allele) can transition the humoral autoimmunity seen in Sle1 mice into more severe lupus-like disease states. We found that heterozygosity of Fas mutation in the B6.Sle1^homo background is sufficient to impair Fas-induced apoptosis in CD8⁺ and B cells, but not in CD4⁺ T cells. These results support that Sle1^homo.lpr^het and Sle1^homo.lpr^homo strains may both have a loss of B cell tolerance, but only the latter experiences a loss of CD4⁺T cell tolerance. This profound difference may explain the observed difference in clinical phenotypes. Additionally, Sle1^homo.lpr^het mice start to exhibit elevated ANA and gene dosage effect on various IgG and IgM autoantibodies levels at 2-3 mos of age. Renal pathological analysis of partial loss of Fas further supports that homozygous or complete loss of Fas, leading to loss of T cell tolerance, is required for more severe renal pathology.

There was no evidence of splenomegaly or lymphadenopathy in Sle1^homo.lpr^het mice. Unlike in Sle1^homo.lpr^homo mice, the percentage of proliferative T and B cells in Sle1^homo.lpr^het mice were comparable with B6 controls. Interestingly, immune profiling of immune cells from Sle1^homo.lpr^het mice revealed the effects of Fas^lpr in upregulating serum levels of IL-2 and IL-27, two Th1-associated cytokines, resulting in Th1 skewed T cell differentiation. Additionally, we also observed significantly elevated levels of Th1 cytokines, including IL-12 and IFN-γ in Sle^homo.lpr^het mice at the age of 4-8 mos. These results suggest that the heterozygous absence of Fas interacting with sle1 can augment anti-chromatin autoimmunity by upregulating Th1-based immune responses.

Early studies in mouse models showed that autoimmune responses in murine lupus models can be mitigated by T cell depletion [38, 39]. T cell abnormalities are also frequently observed in SLE patients [40, 41]. While CD8⁺ T cells have been suggested to directly contribute to autoimmune-mediated tissue damage [42, 43], data also reveal that pathogenic CD4⁺ T cells promote chronic inflammatory responses through releasing cytokines and elevate autoantibody levels through helping self-reactive B cells in lupus [44]. Here, our studies showed immune responses mediated by CD4⁺ T cells from Sle1^homo.lpr^het mice are skewed to Th1-type, and increased levels of cytokines associated with Th1 responses are also noted in this strain. Even under appropriate cytokine environments, Sle1^homo.lpr^het naïve CD4⁺ T cells exhibited reduced Th17 production. Results from this new lupus strain are consistent with clinical data indicating increased IFN-γ production in SLE [45]. Therefore, our data not only provide direct evidence to link pathogenic Th1 responses with lupus-like autoimmunity but also support the usage of this strain to study Th1-related mechanistic studies or therapeutic approaches.

In the same vein, we also observed the gene dose effect of Fas on elevating serum IL-27 levels. IL-27, produced mainly by dendritic cells and macrophages, belongs to the IL-12 family [46]. Upon Toll-like Receptor (TLR) stimulation, IL-27 has been reported to induce IFN-γ production via the Jak-Stat pathway and serve as an early initiator of Th1 responses [24, 46]. Development of arthritis was delayed in IL-27R knockout (KO) mice, and IL-27R KO only alters the frequency of IFN-γ⁺ T cells, not IL-4⁺ or IL-17⁺T cells [47]. Recently, the hyperlipidemia-TLR4-IL-27-T_FH cell axis was implicated in promoting atherosclerosis in SLE patients [48], and certain IL-27 gene polymorphisms are noted to be associated with SLE [49]. Perhaps most importantly, the newly generated strain lacks the profound lymphoproliferation described in the traditionally studied Sle1^homo.lpr^homo mice, which are also regarded as a mouse model of chronic lymphocytic leukemia. Since the new strain does not die prematurely due to lymphosplenomegaly, this allows one to age these mice further till autoimmunity with mild renal disease gradually sets in, reminiscent of patients with lupus. Since the Sle1^homo.lpr^het strain bears only one allele of the mutant Fas allele and is on the C57BL/6 genetic background, this strain is ideal for investigating the additive effect of other immune genes already backcrossed onto the same genetic background. More detailed renal histological characterizations such as immunostaining of IgG and C3 deposition in the renal lesions in future studies will contribute to a more comprehensive understanding of the gene dose effect of Fas on rental damages. Taken together, the newly generated Sle1^homo.lpr^het strain constitutes a novel mouse model of lupus, ideal for genetic studies, autoantibody repertoire investigation, and Th1 effector cell skewing.

Abbreviations:

SLE

Systemic Lupus Erythematosus

B6

C57BL/6

lpr^homo

B6.Fas^lpr/lpr

Sle1^homo

B6.Sle1/Sle1

Sle1^homo.lpr^het

B6.Sle1/Sle1.Fas^lpr/+

Sle1^homo.lpr^homo

B6.Sle1/Sle1.Fas^lpr/lpr

IFN

Interferon

H&E

Hematoxylin and Eosin

Anti-dsDNA

Anti-double-stranded DNA

ANA

Anti-Nuclear Antibody

BUN

Blood Urea Nitrogen

FFPE

Formalin-Fixed and Paraffin-Embedded

mos

months

hr

hour

hrs

hours

GN

Glomerulonephritis

KO

Knockout

Th

T helper

TLR

Toll-like Receptor

OD

Optical Density

NS

No Significance

NA

Not Available