Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Dec 23.
Published in final edited form as: Inflamm Res. 2023 Apr 15;72(5):1083–1097. doi: 10.1007/s00011-023-01731-1

CX3CR1 modulates SLE-associated glomerulonephritis and cardiovascular disease in MRL/lpr mice

Xavier Cabana-Puig 1, Ran Lu 1,2, Shuo Geng 2, Jacquelyn S Michaelis 3, Vanessa Oakes 1, Caitlin Armstrong 1,2, James C Testerman 1, Xiaofeng Liao 1, Razan Alajoleen 1, Michael Appiah 1, Yao Zhang 2, Christopher M Reilly 4, Liwu Li 2, Xin M Luo 1
PMCID: PMC10748465  NIHMSID: NIHMS1951314  PMID: 37060359

Abstract

Objective

Patients with systemic lupus erythematosus (SLE) often develop multi-organ damages including heart and kidney complications. We sought to better define the underlying mechanisms with a focus on the chemokine receptor CX3CR1.

Methods

We generated Cx3cr1-deficient MRL/lpr lupus-prone mice through backcrossing. We then employed heterozygous intercross to generate MRL/lpr littermates that were either sufficient or deficient of CX3CR1. The mice were also treated with either Lactobacillus spp. or a high-fat diet (HFD) followed by assessments of the kidney and heart, respectively.

Results

Cx3cr1−/− MRL/lpr mice exhibited a distinct phenotype of exacerbated glomerulonephritis compared to Cx3cr1+/+ littermates, which was associated with a decrease of spleen tolerogenic marginal zone macrophages and an increase of double-negative T cells. Interestingly, upon correction of the gut microbiota with Lactobacillus administration, the phenotype of exacerbated glomerulonephritis was reversed, suggesting that CX3CR1 controls glomerulonephritis in MRL/lpr mice through a gut microbiota-dependent mechanism. Upon treatment with HFD, Cx3cr1−/− MRL/lpr mice developed significantly more atherosclerotic plaques that were promoted by Ly6C+ monocytes. Activated monocytes expressed ICOS-L that interacted with ICOS-expressing follicular T-helper cells, which in turn facilitated a germinal center reaction to produce more autoantibodies. Through a positive feedback mechanism, the increased circulatory autoantibodies further promoted the activation of Ly6C+ monocytes and their display of ICOS-L.

Conclusions

We uncovered novel, Cx3cr1 deficiency-mediated pathogenic mechanisms contributing to SLE-associated glomerulonephritis and cardiovascular disease.

Keywords: SLE, CX3CR1, Glomerulonephritis, Gut microbiota, Atherosclerosis, High-fat diet

Introduction

Systemic lupus erythematosus (SLE) is a prototypic autoimmune disease for which there is no known cure. It is characterized by systemic inflammation that damages numerous organs of the body, including kidney and heart [1]. The SLE-associated inflammation of the kidney, known as lupus nephritis, affects more than half of SLE patients and represents one of the leading causes of morbidity and mortality [1, 2]. Among the more severe manifestations of SLE, cardiovascular events are also a leading cause of death for SLE patients, who have a relative risk of 1.98 of symptomatic cardiovascular events compared to non-SLE controls based on a 2021 meta-analysis [3]. A recent longitudinal prospective study involving 342 patients with SLE have found that 20% of these patients experienced major adverse cardiovascular events, among which a quarter of them died [4]. Current standard-of-care treatments for SLE are primarily nonselective immunosuppressants and not specific for treating or preventing cardiovascular events [2]. Although current therapies can treat acute symptoms and reduce the risk of renal failure associated with SLE, side effects are a major cause of concern. For example, patients taking long-term immunosuppressants are prone to higher incidence of severe infections [5]. There is an imperative need to develop new treatment strategies against the development of SLE-associated kidney and heart disease, for which a thorough understanding of the pathogenesis is required.

CX3CR1 [6] and its ligand CX3CL1 [7, 8] were both discovered in 1997. They were initially found to mediate lymphocyte and monocyte chemotaxis. CX3CL1 is induced upon LPS or TNF/IL-1 stimulation through NF-κB signaling indicating a role of the receptor/ligand in inflammation [9]. CX3CR1 is highly expressed in human CD16+ and mouse Ly6C− non-classical monocytes [10]. CX3CR1 and its ligand are found to be elevated in the circulation of SLE patients (receptor in the peripheral blood mononuclear cells/PBMCs [11] and soluble ligand in the serum [12]). Studies of human kidney biopsies have shown that renal infiltrates of CX3CR1+ cells are T cells and monocyte/macrophages [13]. In mouse models of SLE, elevations of the receptor and ligand have been shown to be associated with proliferative lupus nephritis [14]. In lupus-prone MRL/lpr mice, CX3CL1 antagonism can attenuate lupus nephritis [15]. In the anti-glomerular basement membrane glomerulonephritis model, antibody neutralization of CX3CR1 attenuates disease severity [16]. We have also shown that renal infiltrates of CX3CR1+ cells are CD11b+ conventional dendritic cells (cDCs) in MRL/lpr mice [17]. In fact, CX3CR1 mediates the homing of cDCs [18] and monocytes/macrophages [19] to the kidney. Furthermore, the CX3CL1-CX3CR1 axis is implicated in chronic kidney disease (CKD) [20], and CX3CR1 is expressed in the developing kidney [21]. Therefore, many have hypothesized that CX3CR1+ cells contribute to the development of kidney inflammation in SLE.

In contrast, the role of CX3CR1 in SLE-associated cardiovascular disease is significantly understudied, although there are many CX3CR1 studies on atherosclerosis unrelated to SLE. In general, CX3CR1 polymorphisms are risk factors for atherosclerosis [22]. This is with the exception of homozygous CX3CR1 M280 mutation, which leads to loss-of-function and thus protection against atherosclerotic cardiovascular disease [23]. Consistent with a pathogenic role of human CX3CR1 in atherosclerosis, Cx3cr1 deficiency in mice leads to rapid turnover (survival disadvantage) of myeloid cells including monocytes within atherosclerotic plaques, suggesting that CX3CL1-CX3CR1 signaling provides a survival signal for monocytes and/or foam cells [9, 22]. However, it has also been shown that CX3CR1 is important for T cell survival [24] and mediates regulatory T cell (Treg) migration to the aorta in atherosclerosis to improve disease [25, 26]. As mentioned earlier, how CX3CR1 modulates SLE-associated cardiovascular disease is unknown.

People have resorted to loss-of-function genetic approaches to elucidate the putative pathogenic role of CX3CR1 in several autoimmune disease models. Previous studies of Cx3cr1 deficiency in autoimmunity had been focused on inducible models. For example, Cx3cr1 deficiency led to more severe disease in C57BL/6 (B6) mice upon induction of experimental autoimmune uveitis due to increased recruitment of monocytes to the retina [27]. The deficiency also accelerated retinopathy in streptozotocin (STZ) induced type 1 diabetes in B6 mice [28], and injection of CX3CR1 antagonist induced cognitive deficits in STZ-treated B6 mice [29]. Notably, these results are contradictory to those presented earlier, where blockade of CX3CR1 signaling tends to attenuate autoimmunity with kidney involvement. Likewise, while Cx3cr1 deficiency led to more severe disease in an experimental autoimmune encephalomyelitis (EAE) mouse model of multiple sclerosis (MS) with accumulation of DCs in the brain [30], inhibition of CX3CR1 attenuated EAE in a myelin oligodendrocyte glycoprotein-induced rat model of MS [31]. The role of CX3CR1 in autoimmunity is thus complex, possibly due to the fact that it is expressed by various immune cell types. In humans, there are two common single nucleotide polymorphisms (SNPs) in CX3CR1, V249I and T280M, with the latter conferring loss-of-function for the gene. Both SNPs have been recently shown to be associated with increased risk of CKD [32].

In this study, we have generated global knockout of Cx3cr1 in lupus-prone MRL/lpr mice and directly assessed the roles of this chemokine receptor in the pathogenesis of glomerulonephritis and SLE-associated heart disease. Our results indicate exacerbation of both in the absence of Cx3cr1, suggesting CX3CR1-mediated control of kidney and cardiovascular disease in SLE. In addition, we have uncovered the regulatory mechanisms that account for the effects of Cx3cr1 deficiency on these two clinically relevant SLE manifestations.

Materials and methods

Animals

MRL/lpr [MRL/Mp-Faslpr/lpr, stock number 000485] and b6/Cx3cr1gfp/gfp [B6.129P2(Cg)- Cx3cr1tm1Litt, stock number 005582] mice were purchased from The Jackson Laboratory. All mice were maintained in a specific pathogen-free environment under the requirements of Institutional Animal Care and Use Committee (IACUC) at Virginia Polytechnic Institute and State University. MRL/lpr-Cx3cr1gfp/gfp (or Cx3cr1−/− MRL/lpr) mice were obtained after backcrossing the loss-of-function Cx3cr1gfp/gfp locus to MRL/lpr for 12 generations, reaching at least 99.6% similarity to MRL/lpr background, which was confirmed with genome scanning using mouse SNP panels. For all experiments, we first crossed MRL/lpr to Cx3cr1−/− MRL/lpr to obtain Cx3cr1+/− MRL/lpr. The heterozygous mice were then intercrossed to generate littermate experimental animals of 3 different genotypes (+/+, +/−, −/−). The Lactobacillus strains, L. reuteri (CF48-3A), L. oris (F0423), L. johnsonii (135–1-CHN) and L. gasseri (JV-V03) and L. rhamnosus (LMS201), were obtained from BEI Resources. All 5 strains were freshly cultured every week individually, then mixed, washed, and resuspended in PBS, and finally orally administered to female Cx3cr1−/− MRL/lpr and Cx3cr1+/+ MRL/lpr littermates, weekly at 2 × 108 CFU/Lactobacillus strain per dose, from 3 weeks of age till euthanasia at 17 weeks of age. In experiments involving diet modulations, female Cx3cr1−/− MRL/lpr and Cx3cr1+/− MRL/lpr littermates were fed with a high fat diet (HFD; Envigo, TD88137) or a control diet (CD; Envigo, TD05230) for 10 weeks starting at 4 weeks of age, followed by sample collection at 14 weeks of age. All mice were continuously monitored every week for body weight and proteinuria. Upon euthanasia, the body and organ weights were recorded.

Analyses of renal function

All fixed kidney tissues were paraffin-embedded, sectioned, and stained for Periodic Acid-Schiff (PAS) at the Histopathology Laboratory at Virginia-Maryland College of Veterinary Medicine. Kidney histopathology was graded on a scale of 0–3 each for glomerular nephritis (GN; including cellularity, mesangial matrix, necrosis, percentage of sclerotic glomeruli, and presence of crescents), tubulointerstitial nephritis (TI), and perivascular infiltration (PV). Scores were assigned by a board-certified veterinary pathologist in a blinded fashion [33]. Proteinuria was determined with a Pierce Coomassie Protein Assay Kit (Thermo Fisher Scientific, Waltham, MA). Urine samples were also used for the measurement of urinary albumin and creatinine with a mouse Albumin ELISA kit (Bethyl Laboratories/Fortis Life Sciences, Waltham, MA), and a Creatinine Assay kit (Cayman Chemical, Ann Arbor, MI), respectively, according to the instructions provided in the kits. The Albumin to Creatinine Ratio (ACR, mg/g) was then calculated.

Immunohistochemistry

Kidneys and spleens were embedded in Tissue-Tek OCT Compound (Sakura Finetek, Torrance, CA) and rapidly frozen in a freezing bath of dry ice and 2-methylbutane. Frozen OCT samples were cryosectioned and unstained slides were stored at − 80 °C. Frozen slides were warmed to room temperature and let dry for 30 min, followed by fixation in − 20 °C cold acetone at room temperature for 10 min. After washing in phosphate buffered saline (PBS), slides were blocked with PBS containing 1% bovine serum albumin (BSA) and anti-mouse CD16/32 (1:100, BioLegend, San Diego, CA) for 40 min at room temperature. Slides were then incubated with fluorochrome-conjugated antibody mixture overnight at 4 °C in a dark humid box. Slides were mounted with Prolong Gold containing DAPI (Life Technologies, Carlsbad, CA). The following anti-mouse antibodies were used in immunohistochemical analysis: complement C3-APC (1:200, Cedarlane, Cat# 1850362A); IgG2a-PE (1:200, BioLegend, Cat# 407107); Ly6C-APC (1:200, eBiosciences, Cat# 17-5932-80); ICOS-L-PE (1:200, BioLegend, Cat# 107405); CD3-PE (BD Pharmingen, Cat# 555275); and GL7-AF647 (1:100, BioLegend, Cat# 144606). Slides were read and pictured with KEYENCE BZ-X810 Fluorescence Microscope (KEYENCE Corporation of America, Itasca, IL) and a 20 × objective.

16S rRNA sequencing and analysis

Fecal pellets were collected weekly directly from the anus for each mouse to maintain sterile conditions. Samples were stored immediately at − 80 °C until processed. Samples were homogenized, cell lysed, and DNA was extracted using a phenol–chloroform method as previously described [34–38]. DNA was amplified by PCR and amplicons were purified and sequenced bidirectionally (V4 region) on an Illumina MiSeq at Argonne National Laboratory using 150 bp PE chemistry. Reads were quality trimmed using Trimmomatic [39], specifying ILLUMINACLIP:TruSeq3- PE.fa:2:30:10 SLIDINGWINDOW:4:20 MAXINFO:140:0.9 MINLEN:130. Amplicon sequence variants (ASVs) were generated using DADA2 [40] in R. Reads were quality trimmed and filtered using the command fastqPairedFilter using parameters truncLen = c(140,140), maxEE = c(2,2), rm.phix = TRUE, maxN = 0, compress = TRUE, multithread = FALSE. DADA2 was used to learn error rates, perform sample inference, dereplicate and merge paired end reads, and construct a sequence table. Taxonomy was assigned using the SILVA 138 ribosomal RNA (rRNA) database training set (https://doi.org/10.5281/zenodo.4587946) using the DADA2 functions, assignTaxonomy and addSpecies.

Samples were analyzed using the R package phyloseq [41] v 1.34.0. A total of 3698 ASVs were detected in 140 total samples. ASVs seen fewer than three times in at least 20% of samples and samples with fewer than 1000 reads were removed from the dataset, resulting in 139 samples and 286 ASVs used for downstream analyses. ASVs were aggregated at the genus level using the phyloseq function tax_glom. For each comparison, ASVs seen fewer than three times in at least 20% of samples were removed from the analyses. Counts were used for alpha diversity and differential abundance tests, while proportions were used to calculate Bray–Curtis dissimilarity. Differentially abundant and variable taxa between groups were identified using the function differentialTest in corncob [42] v 0.2.0 and significance was assessed using a Wald test with an FDR cutoff of 0.05. Bray–Curtis distances were calculated using the phyloseq function ordinate, specifying “method = ”NMDS”, distance = ”bray”, trymax = 1000, k = 3”. Significance was assessed using the adonis test in the vegan [43] package v 2.5.7 with 999 permutations. Figures were generated using the functions ordiplot3d and ordiellipse in the library vegan3d.

Flow cytometry

Spleen and mesenteric lymph nodes (MLN) were collected and mashed in 70-μm cell strainers with C10 media (RPMI-1640 containing l-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 1% 100X MEM non-essential amino acids, 55 μM 2-mercaptoethanol, 10% fetal bovine serum, 100 U/ml penicillin–streptomycin, all from Life Technologies). For splenocytes, red blood cells were lysed with RBC lysis buffer (eBioscience/Thermo Fisher Scientific). For cell staining, cells were blocked by anti-mouse CD16/32 (1:100, BioLegend), stained with fluorochrome-conjugated antibodies, and analyzed with BD FACS Aria Fusion flow cytometer (BD Biosciences, San Jose, CA). For intracellular staining, Foxp3/Transcription Factor Staining kit was used (eBioscience/Thermo Fisher Scientific). Dead cells were stained with Zombie Aqua (1:100, Biolegend, Cat# 423101). Antimouse antibodies used in this study include: CD3-APC (1:80, BioLegend, Cat# 100235); CD3-APC/Cy7 (1:200, BD Pharmingen, Cat# 559596); CD4-PE/Cy7 (1:160, BioLegend, Cat# 100421); CD8-PE/Dazzle 594 (1:300, BioLegend, Cat# 126621); PD-1-APC/Cy7 (1:300, BioLegend, Cat# 135223); CXCR5-BV605 (1:160, BioLegend, Cat# 145513); CD25-BV421 (1:200, BioLegend, Cat# 102033); CD25-Pacific Blue (1:160, BioLegend, Cat# 102201); Foxp3-PE (1:80, BioLegend, Cat# 320007); CD44-PerCP/Cy5.5 (1:160, BDbiosciences, Cat# 560570); CD62L-APC/Cy7 (1:300, eBioscience, Cat# 25-0621-82); RORγT-PE (1:300, eBioscience, Cat# 12-6981-82); CD3-APC/Cy7 (1:200, BD Pharmingen, Cat# 559596); B220-PE (1:160, BioLegend, Cat# 103207); CD138-PerCP/Cy5.5 (1:200, BioLegend, Cat# 142509); CD19-AF700 (1:200, eBioscience, Cat# 56-0193-82); CD38-PE/Cy7 (1:160, BioLegend, Cat# 102717); GL7-AF647 (1:100, BioLegend, Cat# 144606); XCR1-BV650 (1:200, BioLegend, Cat# 148220); CD11c-APC (1:80, eBioscience, Cat# 17-0114-82); MHC-II-PE (1:80, BioLegend, Cat# 107607); CD172a-PE/Dazzle 594 (1:200, BioLegend, Cat# 144015); CD21/35-APC/Cy7 (1:200, BioLegend, Cat# 123417); F4/80-BV421 (1:200, BioLegend, Cat# 123131); Siglec-H-PerCP/Cy5.5 (1:200, BioLegend, Cat# 129614); CD138-PerCP/Cy5.5 (1:200, BioLegend, Cat# 142509); GL7-AF647 (1:100, BioLegend, Cat# 144606). Flow cytometry data were analyzed with FlowJo.

Bone marrow (BM) and spleen were harvested from Cx3cr1+/− MRL/lpr mice and Cx3cr1−/− MRL/lpr mice fed with HFD or CD. BM and spleen samples were disassociated with mechanical processes and then filtered through 70-μm cell strainers to obtain single-cell suspension. Red blood cells were lysed with ACK buffer (Thermo Fisher Scientific). The samples were incubated with anti-CD16/32 antibodies (1:100, BD Biosciences) to block Fc receptors followed by staining with fluorochrome-conjugated CD11b-APC/Cy7 (1:200, BioLegend, Cat# 101226), Ly6C-PE/Cy7 (1:200, BioLegend, Cat# 128018), Ly6G-PerCP/Cy5.5 (1:200, BioLegend, Cat# 127616), CCR2-APC (1:200, BioLegend, Cat# 150,628), CD14-APC (1:200, BioLegend, Cat# 123312), CD24-PE (1:200, BioLegend, Cat# 138503), CD38 (1:200, BioLegend, Cat# 102712), CD86-APC (1:200, BioLegend, Cat# 159203), CD200R-APC (1:200, BioLegend, Cat# 123916), ICAM-1-PE (1:200, BioLegend, Cat# 116108), and PD-L1-APC (1:200, BioLegend, Cat# 124312) antibodies. The surface phenotype of Ly6G+ neutrophils, Ly6G−CD11b+Ly6C++ monocytes, Ly6G−CD11b+Ly6C+ monocytes, and Ly6G−CD11b+Ly6C− monocytes was examined using FACSCanto II (BD Biosciences). The data were analyzed with FlowJo.

Analysis of atherosclerotic lesions

Histological analyses of murine atherosclerotic lesions were performed as previously described [44, 45]. Briefly, freshly frozen and OCT-embedded proximal aortic sections were fixed in 4% neutral buffered formalin, followed by Oil Red O staining was performed using a kit (Newcomer Supply, Middleton, WI). The samples were observed under a light microscope, and the percentages of total lesion area were calculated.

ELISAs

Serum was separated from blood after clotting at room temperature for 2 h, then collected and stored at − 80 °C. Anti-dsDNA IgG was measured as previously described [46]. Serum total IgG concentration was determined with mouse IgG ELISA kit (Bethyl Laboratories). Blood endotoxin was measured by using a Pierce LAL Chromogenic Endotoxin Quantitation kit (Thermo Fisher Scientific, Cat# A39552).

BM cell isolation and in vitro stimulation

BM cells isolated from C57BL/6 (B6) mice were cultured in complete RPMI 1640 medium supplemented with M-CSF (10 ng/mL) as described previously [44, 45]. BM cultures were incubated with serum (5% of total volume) collected from B6 mice, Cx3cr1+/− MRL/lpr mice, or Cx3cr1−/− MRL/lpr mice. Fresh serum was added every 2 days. After 5 days, BMMs were harvested, filtered through 70-μm cell strainers, and incubated with anti-CD16/32 antibodies (1:100, BD Biosciences, Cat# 553141) to block Fc receptors. The cells were then stained with fluorochrome-conjugated CD11b-APC/Cy7 (1:200, BioLegend, Cat# 101226), Ly6C-PE/Cy7 (1:200, BioLegend, Cat# 128018), and ICOS-L-PE (1:200, BioLegend, Cat# 107405) antibodies. The samples were examined using FACSCanto II, and the data were analyzed using FlowJo.

Statistical analysis

For the comparison of two groups, unpaired student’s t-test was used. For the comparison of three or more groups, one-way ANOVA was used. Two-way ANOVA was used to reveal time- and group-dependent effects. Results were considered statistically significant when p < 0.05. All analyses were performed with the GraphPad Prism software.

Results

Cx3cr1 deficiency exacerbates glomerulonephritis in female MRL/lpr mice

To directly elucidate the role of CX3CR1+ cells in spontaneous SLE, we backcrossed the Cx3cr1-gfp/gfp locus from B6 background (The Jackson Laboratory/JAX Stock Number 005582) to lupus-prone MRL/lpr mice (JAX Stock Number 000485). The Cx3cr1-gfp/gfp mice have EGFP knock-in replacing the first 390 bp of the coding exon of the Cx3cr1 gene. After 12 generations of speed congenic backcrossing [17, 47], we confirmed with genome scanning using mouse SNP panels that the purity of MRL background reached at least 99.9%. Using heterozygous intercross, we generated littermate MRL/lpr mice that were either wildtype (+/+) or global knockout (−/−, or gfp/gfp knock-in) for the gene Cx3cr1. Littermates were separated upon weaning, and those with the same genotype were housed together.

We had hypothesized that the deficiency of Cx3cr1 would attenuate disease in MRL/lpr mice given its potentially pathogenic role especially for kidney inflammation as discussed earlier. Surprisingly, Cx3cr1 deficiency exacerbated lupus-like disease in both MRL/lpr females (Fig. 1 and Suppl. Fig. S1A, B) and males (Suppl. Fig. S1C, D). Female MRL/lpr mice lacking Cx3cr1 exhibited significantly worse splenomegaly and lymphadenopathy (Fig. 1A), as well as significantly exacerbated proteinuria (Fig. 1B), compared to Cx3cr1−/− MRL/lpr mice. Of note, Cx3cr1+/+ MRL/lpr mice did not have a significant elevation of proteinuria even at 17 weeks of age due to the loss of this phenotype in both JAX and our colonies [48]. Heterozygous mice (+/gfp) mostly exhibited an intermediate phenotype between Cx3cr1+/+ and Cx3cr1−/− mice (Suppl. Fig. S1A–C).

Fig. 1.

Fig. 1

Open in a new tab

Cx3cr1 deficiency exacerbates lymphoproliferation and glomerulonephritis in female MRL/lpr mice. A Lymphoid tissue-to-body weight ratios at 17 weeks of age. MLN mesenteric lymph node, A-C LN axillary-cervical lymph nodes. *p < 0.05. B Level of proteinuria over time. n = 6, **p < 0.01. C Urinary albumin-to-creatinine ratio at late disease. D Histopathological scores at late disease stage. GN glomerulonephritis. Note that three Cx3cr1−/− MRL/lpr mice had noticeably higher GN scores. E Representative images showing a normal glomerulus (1) vs. an abnormal glomerulus (2,3), the latter from a Cx3cr1−/− MRL/lpr mouse. 2: Glomerular necrosis. Green arrows indicate cytoplasmic fragmentation; Orange arrows indicate nuclear pyknosis and karyorrhexis. 3: Glomerular sclerosis and crescent formation. Sclerosis is indicated by red arrows and surrounds the entire glomerulus; Crescent formation is indicated by the orange brackets and is segmental. F Representative images showing exacerbated renal deposition of pathogenic IgG2a. G Intensity ratio of IgG2a to complement C3 as determined with ImageJ. Note that three Cx3cr1−/− MRL/lpr mice had noticeably higher ratios leading to an overall trending increase with Cx3cr1 deficiency (p = 0.14)

Consistent with the increase of proteinuria, the urine albumin-to-creatinine ratio (ACR) at the late disease stage was significantly higher in Cx3cr1−/− MRL/lpr mice (Fig. 1C), indicating more severe glomerulonephritis. We next performed kidney histopathological analysis to characterize glomerulonephritis (GN), tubulointerstitial nephritis (TI), and perivascular infiltration (PV). While no statistical significance was observed, 50% of Cx3cr1−/− MRL/lpr mice (3 out of 6), compared to none of Cx3cr1+/+ MRL/lpr mice, exhibited severe glomerulonephritis (Fig. 1D) that was characterized by glomerular necrosis, sclerosis, and formation of crescents (Fig. 1E). TI nephritis and PV infiltration scores, on the other hand, were largely comparable between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice (Suppl. Fig. S2A). Furthermore, we performed immunohistochemical analysis of renal deposition of complement C3 as well as IgG2a, the pathogenic Ig isotype in MRL/lpr mice. While the deposition of C3 was comparable among all images taken, renal deposition of IgG2a was noticeably higher for the three Cx3cr1−/− MRL/lpr mice with worse GN scores (Fig. 1F), leading to an overall trending, but not significant, increase of the IgG2a/C3 intensity ratio with Cx3cr1 deficiency (Fig. 1G). Other measurements, including the serum endotoxin level, the ratio of anti-double stranded (ds)DNA IgG to total IgG in the blood, and serum levels of typical pro-inflammatory and anti-inflammatory cytokines, were comparable between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice (Suppl. Fig. S2B-D).

Together, these results suggest exacerbation of lupus-like signs, including glomerulonephritis and lymphoproliferation (splenomegaly, lymphadenopathy), in MRL/lpr mice lacking Cx3cr1 globally.

Exacerbation of glomerulonephritis with Cx3cr1 deficiency is gut microbiota-dependent

It is known that Cx3cr1 deficiency compromises the gut barrier [49]. Indeed, Cx3cr1 deficiency causes bacterial translocation to mesenteric lymph nodes (MLN) and more severe DSS-induced colitis through an IL-17A-dependent mechanism [50]. We have previously shown that a mixture of 5 Lactobacillus spp. can decrease intestinal permeability and enhance gut barrier function while reshaping the gut microbiota of MRL/lpr mice [35]. Therefore, we tested the hypothesis that the same 5 species of lactobacilli (L. reuteri, oris, johnsonii, gasseri and rhamnosus) [35] could render the phenotypes of Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice indistinguishable by repairing the leaky gut caused by Cx3cr1 deficiency.

We first mapped the composition of the gut microbiota using 16S rRNA sequencing with or without Lactobacillus treatment. As anticipated, the probiotic bacteria were able to neutralize the gut microbiota differences between female Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice (Fig. 2A, B). Analysis of amplicon sequence variants revealed ten differentially abundant taxa between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice (Fig. 2C). In contrast, only three genera were minimally different in abundance with Lactobacillus treatment (Fig. 2D), confirming that Lactobacillus equalized the gut microbiota differences between the two mouse genotypes. Interestingly, the five genera enriched in Cx3cr1−/− MRL/lpr mice in the absence of Lactobacillus spp., namely Lachnospiraceae unclassified, Lachnospiraceae FCS020 group, Desulfovibrio, Butyricicoccus and Romboutsia, all have commensal species capable of producing short-chain fatty acids (SCFAs). However, preliminary testing showed no difference in fecal levels of acetate, butyrate, and propionate between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice (data not shown). This suggests that the bacterial species enriched in the gut microbiota of Cx3cr1−/− MRL/lpr mice may not be primary producers of SCFAs. Future studies involving metagenomic shotgun sequencing will reveal the identities of these bacterial species.

Fig. 2.

Fig. 2

Open in a new tab

Exacerbation of glomerulonephritis with Cx3cr1 deficiency is gut microbiota-dependent, but exacerbation of lymphoproliferation is not. A, B Non-metric multidimensional scaling (NMDS) ordination of gut microbiota communities in female Cx3cr1+/+ (WT) and Cx3cr1−/− (KO) MRL/lpr littermates treated with either PBS (A) or Lactobacillus spp. (B). Data from fecal samples collected every 2 weeks from weaning till 15 weeks of age are shown (n = 5/group/time point). For the PBS groups the differences were statistically significant (Adonis test, p < 0.001). C, D Differentially abundant taxa at the genus level between WT and KO MRL/lpr mice upon PBS (C) or Lactobacillus treatment (D). E Lymphoid tissue-to-body weight ratios at 17 weeks of age following Lactobacillus treatment. MLN mesenteric lymph node, A-C LN axillary-cervical lymph nodes. *p < 0.05; ****p < 0.0001. B Level of proteinuria over time (n = 6/group) following Lactobacillus treatment. The difference was not statistically significant

Surprisingly, however, oral administration of Lactobacillus spp. did not affect the degree of splenomegaly and lymphadenopathy in female Cx3cr1−/− MRL/lpr mice (Fig. 2E). This observation suggests that a gut microbiota-independent mechanism is responsible for the CX3CR1-mediated exacerbation of splenomegaly and lymphadenopathy in the MRL/lpr mouse. In contrast, treatment with Lactobacillus reversed the exacerbated renal disease (proteinuria levels) caused by Cx3cr1 deficiency (Fig. 2F). Given that Lactobacillus spp. equalized the gut microbiotas between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice, this finding suggests a gut microbiota-dependent mechanism that Cx3cr1 deficiency exacerbates glomerulonephritis in the female MRL/lpr mouse.

Together, these results suggest a novel concept of dissociation between glomerulonephritis and lymphoproliferation in MRL/lpr mice, where gut dysbiosis in the lupus-prone mouse contributes to the development of renal inflammation but not to lymphoproliferation.

Significant differences in splenic populations are equalized by Lactobacillus treatment

To elucidate the cellular mechanisms by which Cx3cr1 deficiency contributes to the development of glomerulonephritis, we analyzed the splenocytes of Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice with or without Lactobacillus treatment. Subsets of B cells, T cells and myeloid cells were quantified with flow cytometry. Three subsets of immune cells were significantly modulated by Cx3cr1. They were CD3+CD4−CD8− double-negative T (DNT) cells and two populations of MHC-II+ cells (CD11chlgh and CD11clow). DNT cells were significantly expanded in Cx3cr1−/− MRL/lpr mice, which accounted for over 50% of total splenocytes (~70% of T cells were DNT cells as shown in Fig. 3A, whereas ¾ of splenocytes were T cells). With the spleens of Cx3cr1−/− MRL/lpr mice being significantly larger, this represented a large number of DNT cells. The two populations of MHC-II+ cells, on the other hand, were significantly decreased in frequency (Fig. 3C). Interestingly, all these differences were equalized with Lactobacillus treatment (Fig. 3B and D). This may have contributed to the same level of proteinuria in Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice with Lactobacillus treatment. A fraction of the two populations of MHC-II+ cells express a specific marker of marginal zone macrophages (MZM), SIGN-R1 [51], which were almost completely depleted in Cx3cr1−/− MRL/lpr mice (Fig. 3E). Interestingly, both CD11chlghSIGN-R1+ and CD11clowSIGN-R1+ MZM cells were restored in Cx3cr1−/− MRL/lpr mice in the presence of Lactobacillus spp. (Fig. 3F). Together, these results suggest that Cx3cr1 deficiency exacerbates lupus-like disease in MRL/lpr mice via suppression of MZM cells that leads to the generation of DNT cells, thus exacerbating SLE.

Fig. 3.

Fig. 3

Open in a new tab

Significant differences in several splenic populations between Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice are equalized by Lactobacillus treatment. A Analysis of DNT cells without Lactobacillus. Left: representative FACS plots. Right: frequency of DNT cells in total T cells. **p < 0.01. B Analysis of DNT cells with Lactobacillus. C Analysis of MHC-II+ cells without Lactobacillus. Left: representative FACS plots. Right: frequencies of CD11chighMHCII+ and CD11clowMHCII+ cells in total splenocytes. ***p < 0.001. D Analysis of MHC-II+ cells with Lactobacillus. E Analysis of MZM cells without Lactobacilluss. Left: gating strategy to identify SIGN-R1+ MZM cells. Right, frequencies of CD11chigh & CD11clow MZM cells in total splenocytes. **p < 0.01. F Analysis of MZM cells with Lactobacillus

A high-fat diet exacerbates cardiovascular inflammation in Cx3cr1−/− MRL/lpr mice

To define the role of CX3CR1 in SLE-associated cardiovascular disease, we used a high-fat diet (HFD) to accelerate the progression of atherosclerotic inflammation [45]. In this experiment, we employed the heterozygous Cx3cr1+/− MRL/lpr mice that co-expressed GFP with CX3CR1. As anticipated, the HFD increased the body weight of both Cx3cr1+/− and Cx3cr1−/− MRL/lpr mice (Suppl. Fig. S3A-B). However, HFD did not exacerbate glomerulonephritis (Suppl. Fig. S3C) or the enlargement of lymphoid tissues (Suppl. Fig. S3D-E). Strikingly, while the HFD was able to slightly increase the plaque size in Cx3cr1+/− MRL/lpr mice over the control diet (CD), the atherosclerotic plaque size was significantly larger when HFD was given to Cx3cr1−/− MRL/lpr mice (Fig. 4A), suggesting CX3CR1 as a suppressor of cardiovascular inflammation in SLE. Notably, the plaques sizes in mice of MRL background are small as independently shown by others [52], and the plaque area was only around 2% even in the HFD group with Cx3cr1 deficiency.

Fig. 4.

Fig. 4

Open in a new tab

A high-fat diet exacerbates cardiovascular inflammation in Cx3cr1−/− MRL/lpr mice. Cx3cr1+/− and Cx3cr1−/− MRL/lpr mice were fed either a 42% fat high-fat diet (HFD) or its control diet (CD; 6% fat) starting 4 weeks of age. Mice were euthanized at 15 weeks of age for analyses. A Measurement of vasculitis. Left, representative Oil Red O (ORO) images. Right, quantification of plaque size. ***p < 0.001, ****p < 0.0001. B Frequency of splenic neutrophils. **p < 0.01, ****p < 0.0001. C Frequencies of splenic monocytes. From left to right, representative FACS plot and frequencies of Ly6C−, Ly6C+, and Ly6C++ monocytes. *p < 0.05, ***p < 0.001. D Mean fluorescence intensity (MFI) of CD38 on total monocytes. **p < 0.01. E MFI of ICAM-1 on total monocytes. *p < 0.05. One-way ANOVA was performed for all statistical analyses

As both neutrophils [53] and monocytes [45] are capable of releasing pro-inflammatory mediators that contribute to the development of atherosclerotic plaques, we assessed the frequencies of these two types of myeloid cells in the bone marrow and spleen. While no significant differences were observed among treatment groups in the bone marrow (Suppl. Fig. S4), in the spleen, neutrophils were expanded with HFD; but the expansion was comparable between Cx3cr1+/− and Cx3cr1−/− MRL/lpr mice (Fig. 4B). In contrast, the frequency of a subpopulation of monocytes expressing a high level of Ly6C (Ly6G−CD11b+Ly6C++ classical monocytes) was the highest in the spleen of Cx3cr1−/− MRL/lpr mice fed the HFD and significantly higher than their Cx3cr1+/− MRL/lpr counterparts (Fig. 4C). This suggests that the expansion of splenic Ly6C++ classical monocytes may be involved in the exacerbation of cardiovascular inflammation observed in Cx3cr1−/− MRL/lpr mice with HFD. Moreover, the splenic monocytes of HFD-fed Cx3cr1−/− MRL/lpr mice expressed a significantly increased level of CD38, an activation marker of enhanced inflammation [54], over those in mice treated with CD (Fig. 4D). Furthermore, the adhesion molecule ICAM-1 was significantly increased on splenic monocytes isolated from Cx3cr1−/− MRL/lpr mice with HFD compared with CD-fed Cx3cr1−/− MRL/lpr mice (Fig. 4E), suggesting the potential of these monocytes to undergo trans-endothelial migration/trafficking towards sites of inflammation such as the atherosclerotic plaques [45].

We also analyzed the frequencies of multiple T cell subsets (Suppl. Fig. S5A), B cell subsets (Suppl. Fig. S5B) and macrophage and dendritic cell subsets (Suppl. Fig. S5C). With the exception of macrophages and plasmacytoid dendritic cells, HFD did not significantly alter the frequencies of these cell populations in the spleen of Cx3cr1−/− MRL/lpr mice compared to CD-fed Cx3cr1−/− MRL/lpr mice or Cx3cr1+/− MRL/lpr mice.

Together, these results suggest Cx3cr1 deficiency facilitated the activation and migration of Ly6C++ classical monocytes leading to the exacerbation of cardiovascular inflammation in lupus-prone MRL/lpr mice.

Monocytes promote a positive feedback cycle of inflammation that involves germinal center reaction and autoantibody production in Cx3cr1−/− MRL/lpr mice fed a high-fat diet

To elucidate the mechanism by which splenic monocytes contribute to the exacerbation of atherosclerotic inflammation in Cx3cr1−/− MRL/lpr mice with HFD, we first investigated the formation of germinal centers, as monocytes are known to promote ICOS+ follicular T helper (Tfh) cells by expressing ICOS-L [55, 56]. Indeed, we observed on splenic sections the co-localization of Ly6C with ICOS-L, which was most evident in Cx3cr1−/− MRL/lpr mice fed a HFD (Fig. 5A). Consequently, the frequency of CXCR5+PD1+ Tfh cells (Fig. 5B) and that of Tfh cells expressing a high level of ICOS (Fig. 5C) were significantly higher in HFD-fed Cx3cr1−/− MRL/lpr mice than their CD-fed counterparts, suggesting the promotion of germinal center reaction, evidenced by the expression of GL-7 (Fig. 5D), that involves the interaction between ICOS-L expressed by monocytes and ICOS expressed by Tfh cells.

Fig. 5.

Fig. 5

Open in a new tab

ICOS-L expressing Ly6C+ monocytes promote germinal center reaction in Cx3cr1−/− MRL/lpr mice fed a high-fat diet. A Immunohistochemical analysis showing co-expression of Ly6C and ICOS-L. B Frequency of splenic CXCR5+PD1+ cells in CD4+ T cells. Left, representative FACS plot pre-gated on CD4+ T cells. *p < 0.05, **p < 0.01. (C) Frequency of splenic CXCR5+PD1+ cells in ICOShigh CD4+ T cells. Left, representative FACS plot pre-gated on T cells. Middle, FACS plot within the CD4+ICOShigh T cell gate. *p < 0.05. One-way ANOVA was performed for all statistical analyses. D Immunohistochemical analysis showing GL7+ germinal centers in relationship to CD3+ T cells

The enhanced germinal center reaction led to a significantly higher level of anti-dsDNA IgG autoantibodies in the circulation (Fig. 6A). Therefore, we asked if the upregulation of autoantibodies had an impact on monocyte activation. To do this, we established in vitro bone marrow cultures treated with the sera from B6 mice, Cx3cr1+/− MRL/lpr mice and Cx3cr1−/− MRL/lpr mice, respectively, and quantified the expansion of Ly6C−, Ly6C+ and Ly6C++ monocyte populations (Fig. 6B) as well as their expression of ICOS-L. Strikingly, the serum from Cx3cr1−/− MRL/lpr mice was the most potent in converting Ly6C− monocytes to Ly6C-expressing (Ly6C+ and Ly6C++) monocytes (Fig. 6C). In addition, the serum from Cx3cr1−/− MRL/lpr mice significantly induced the expression of ICOS-L on all 3 monocyte populations over the B6 and Cx3cr1+/− MRL/lpr controls (Fig. 6D). This observation, coupled with the greater level of autoantibodies in the serum of Cx3cr1−/− MRL/lpr mice, suggests autoantibody-mediated expansion of Ly6C+ICOS-L+ monocytes.

Fig. 6.

Fig. 6

Open in a new tab

The high-fat diet induces anti-dsDNA IgG in Cx3cr1−/− MRL/lpr mice that in turn promotes monocyte expression of ICOS-L. A Levels of anti-dsDNA IgA, total IgG, and ratio of anti-dsDNA/total IgG. *p < 0.05. B–D Total bone marrow cells from B6 mice were incubated with the serum from B6, Cx3cr1+/− MRL/lpr, or Cx3cr1−/− MRL/lpr mice for 5 days in the presence of M-CSF. B Representative FACS plot showing gating strategy for Ly6C−, Ly6C+ and Ly6C++ populations. C Frequencies of the 3 monocyte populations. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. D MFI of ICOS-L expression for the 3 monocyte populations. *p < 0.05, ***p < 0.001, ****p < 0.0001. One-way ANOVA was performed for all statistical analyses

Together, these results suggest a positive feedback cycle of inflammation with Cx3cr1 deficiency in MRL/lpr mice, where Ly6C+ monocytes facilitate germinal center reaction via ICOS—ICOS-L interaction leading to upregulation of circulatory autoantibodies, which in turn promote the activation of Ly6C+ monocytes and their expression of ICOS-L, thereby further activating ICOS-expressing Tfh cells. This cycle is then exacerbated by HFD leading to aggravated cardiovascular inflammation in Cx3cr1−/− MRL/lpr mice.

Discussion

In this study, we generated global Cx3cr1 knockout on MRL/lpr background to model the role of the genetic deficiency on autoimmune lupus. Our results showed exacerbation of both glomerulonephritis and HFD-facilitated atherosclerotic inflammation with the deficiency, suggesting that the chemokine receptor CX3CR1, which was originally discovered to function in chemotaxis, also plays an important role in controlling peripheral inflammation associated with SLE. Mechanistically, we uncovered that Cx3cr1 deficiency-mediated exacerbation of renal disease was gut microbiota-dependent, as the difference in phenotype was abrogated when the gut microbiotas of Cx3cr1+/+ and Cx3cr1−/− MRL/lpr mice were equalized by Lactobacillus treatment. In addition, we provided evidence to support downregulation of immunosuppressive MZM cells with Cx3cr1 deficiency, which then allowed for the generation of DNT cells known to be pathogenic by infiltrating the kidney in both human and mouse SLE. Moreover, we found that HFD-facilitated exacerbation of cardiovascular inflammation in Cx3cr1−/− MRL/lpr mice was mediated by a feed-forward mechanism involving monocyte activation and expression of ICOS-L, which stimulates ICOS-expressing Tfh cells leading to increased germinal center reaction, followed by enhancement of autoantibody production which in turn promotes further activation of monocytes and their expression of ICOS-L.

Notably, the results of this study are surprising in that several previous reports have shown the opposite effects of CX3CR1 signaling in SLE, where it contributes to disease pathogenesis and that, conversely, its inhibition attenuates lupus nephritis in mice [14–16]. In addition, CX3CR1+ cells have been shown to infiltrate the kidney in both mice [17–19] and humans [13]. However, none of the previous studies had completely knocked out the gene; instead, antagonism or antibody neutralization was used, which may not completely remove the protein. This suggests a potential dose-dependent effect of CX3CR1 in SLE, with exacerbation of disease at a high dose, whereas a low dose may actually be protective. The observation that heterozygous female mice had the largest renal lymph nodes (Suppl. Fig. S1A)—which may suggest more immune cells in the draining lymph node and fewer infiltrates into the kidney—supports this hypothesis. Male mice did not have this phenomenon (Suppl. Fig. S1C), suggesting a sex-dependent difference that we will investigate in the future.

Interestingly, our results showed that the exacerbation of glomerulonephritis in Cx3cr1−/− MRL/lpr mice was reversed following treatment with Lactobacillus spp., whereas the exacerbation of splenomegaly and lymphadenopathy was not affected. This suggests distinct regulatory mechanisms leading to glomerulonephritis and lymphoproliferation: one depends on the gut microbiota while the other does not. The next question is how the gut microbiota from Cx3cr1−/− MRL/lpr mice facilitates the exacerbation of glomerulonephritis. Interestingly, the conversion of CD8+ T cells to DNT cells is increased in the presence of antigens [57]. These antigens can be protein antigens like ovalbumin, or self-antigens provided by apoptotic cells. It is possible that gut bacteria can provide such antigens. Therefore, we hypothesize that the gut microbiota of Cx3cr1−/− MRL/lpr mice provides antigens necessary for the stimulation of pathogenic lymphocytes (e.g., DNT cells) which lodge the kidneys to exacerbate glomerulonephritis.

DNT cells are T cell receptor (TCR)-αβ+, IL-17 producing T cells well characterized in both human SLE and murine lupus models [57–63]. Peripheral DNT cells can be derived directly from thymocytes, or from activated CD8+ T cells that lose the expression of CD8 [59]. In type 1 diabetes (T1D), DNT cells are considered regulatory cells [59]. In SLE, however, these cells promote the activation of B cells to produce autoantibodies, represent a major source of IL-17A that is known to be pathogenic in SLE, and infiltrate the kidney to exacerbate lupus nephritis [59]. Therefore, it is logical to interpret that the increase of splenic DNT cells in Cx3cr1−/− MRL/lpr mice may be the cause of disease exacerbation in these mice.

Contrary to DNT cells, the potential function of losing CD11chighMHC-II+ and CD11clowMHC-II+ cells in Cx3cr1−/− MRL/lpr mice was initially difficult to interpret. If they were antigen presenting cDCs that stimulate T cells, why would their change be opposite to that of DNT cells? In addition, neither of the MHC-II+ populations expressed a significant amount of XCR-1 or CD172a (data not shown), respective markers of cDC1 and cDC2 cells [64]. Nor did they express Siglec-H, a marker of plasmacytoid DCs. Therefore, we questioned if they were indeed DCs, as markers of DCs and macrophages are not mutually exclusive [64–68]. While the expression of F4/80 was low, we postulated that the two MHC-II+ populations were F4/80−/low splenic macrophages. A subset of such macrophages has been described, namely splenic MZM cells [69]. MZM cells surround splenic follicles to clear dying cells through the expression of scavenger receptors MARCO, SR-A and SIGN-R1 [51]. Importantly, these cells can induce tolerance [51] and have been recently described to possess the capability to control the generation of DNT cells by producing TGFβ [57]. Indeed, we were able to detect SIGN-R1+ MZM cells in both CD11chighMHC-II+ and CD11clowMHC-II+ cells, which were almost completely removed in the spleen of Cx3cr1−/− MRL/lpr mice unless mice were given Lactobacillus spp. This would explain why their change (decrease) was opposite to that of DNT cells (increase).

A recent publication from George Tsokos’ group has described an elegant mechanism of SLE pathogenesis involving MZM and DNT, where the removal of TGFβ-producing MZM leads to conversion of CD8+ T cells to IL-17 producing DNT cells, thus promoting SLE [57]. Our observation of decreased MZM cells (found in two cell populations, CD11chighMHC-II+ and CD11clowMHC-II+) and increased DNT cells in Cx3cr1−/− MRL/lpr mice is consistent with this mechanism. Interestingly, when probing the cell types expressing CX3CR1 using Cx3cr1+/gfp mice, which co-express GFP with the receptor, we found that CD11chighMHC-II+ and CD8+ T cells express CX3CR1, but CD11clowMHC-II+ cells and DNT cells do not (data not shown). It is possible that CD8+ T cells downregulate CX3CR1 when losing CD8 to become DNT cells, whereas in MZM cells, the downregulation of CD11c coincides with the loss of CX3CR1. The loss of CX3CR1 in MZM cells may have led to decreased production of TGFβ from these cells, as the expression of CX3CR1 has been shown to serve as a marker of TGFβ production from macrophages [70, 71]. Finally, decreased TGFβ would allow the generation of pathogenic DNT cells leading to exacerbation of SLE [57]. Taken together (Suppl. Fig. S6A), our observations and the current literature suggest the hypothesis that Cx3cr1 deficiency exacerbates lupus-like disease in MRL/lpr mice via suppressing the production of TGFβ from MZM cells leading to gut microbial antigen-facilitated conversion of CD8+ T cells to IL-17 producing DNT cells, thus exacerbating SLE.

Our data further complement emerging studies demonstrating a close connection between SLE and cardiovascular complications [3, 4]. Mechanistically, our findings reveal that Cx3cr1 deletion may contribute to the expansion of Ly6C++ monocytes which uniquely elevate the expression of ICOS-L both in vitro and in vivo. In turn, the expanded Ly6C++ monocytes may facilitate the expansion of Tfh cells and the increase of autoantibody-expressing B cells. This positive feedback circuitry (Suppl. Fig. S6B) may underlie the mechanistic connection between SLE and elevated cardiovascular risks observed in patients.

In conclusion, we present here a mechanistic study detailing the roles of CX3CR1 in controlling SLE-associated kidney and cardiovascular inflammation. Future studies will include identifying the pathobionts in the gut microbiota of Cx3cr1−/− MRL/lpr mice and examine the role of these pathobionts in exacerbating glomerulonephritis, elucidating how Cx3cr1 deficiency suppresses MZM cells while promoting monocyte activation, as well as delineating the mechanisms by which Ly6C++ monocytes and anti-dsDNA IgG autoantibodies contribute to SLE-associated cardiovascular disease. Together, these novel mechanisms will lead to identification of new pathogenic pathways and thus therapeutic targets in the treatment of two severe manifestations, lupus nephritis and lupus-associated cardiovascular disease, that account for most of the morbidity and mortality in SLE patients.

Supplementary Material

Supplementary File

Acknowledgements

We thank to Sarah Owens and Argonne National Laboratory for assistance on Illumina MiSeq sequencing. In addition, we thank Melissa Makris for flow cytometry analysis. This work was supported by NIH Grants AR067418 and AR073240 (Luo) and HL163948 (Li).

Footnotes

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s00011-023-01731-1.

Conflict of interest There are no competing financial interests in relation to the work described.

Availability of data and materials

The 16S rRNA sequences are available in the National Institutes of Health SRA database (PRJNA835508).

References

  • 1.Tsokos GC. Systemic lupus erythematosus. N Engl J Med. 2011;365:2110–21. [DOI] [PubMed] [Google Scholar]
  • 2.Lech M, Anders HJ. The pathogenesis of lupus nephritis. J Am Soc Nephrol. 2013;24:1357–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Restivo V, Candiloro S, Daidone M, Norrito R, Cataldi M, Minutolo G, et al. Systematic review and meta-analysis of cardiovascular risk in rheumatological disease: Symptomatic and non-symptomatic events in rheumatoid arthritis and systemic lupus erythematosus. Autoimmun Rev. 2021;21(1): 102925. [DOI] [PubMed] [Google Scholar]
  • 4.Skaggs BJ, Grossman J, Sahakian L, Perry L, FitzGerald J, Charles-Schoeman C, et al. A panel of biomarkers associates with increased risk for cardiovascular events in women with systemic lupus erythematosus. ACR Open Rheumatol. 2021;3:209–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Tektonidou MG, Wang Z, Dasgupta A, Ward MM. Burden of serious infections in adults with systemic lupus erythematosus: a national population-based study, 1996–2011. Arthritis Care Res (Hoboken). 2015;67:1078–85. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Imai T, Hieshima K, Haskell C, Baba M, Nagira M, Nishimura M, et al. Identification and molecular characterization of fractalkine receptor CX3CR1, which mediates both leukocyte migration and adhesion. Cell. 1997;91:521–30. [DOI] [PubMed] [Google Scholar]
  • 7.Bazan JF, Bacon KB, Hardiman G, Wang W, Soo K, Rossi D, et al. A new class of membrane-bound chemokine with a CX3C motif. Nature. 1997;385:640–4. [DOI] [PubMed] [Google Scholar]
  • 8.Pan Y, Lloyd C, Zhou H, Dolich S, Deeds J, Gonzalo JA, et al. Neurotactin, a membrane-anchored chemokine upregulated in brain inflammation. Nature. 1997;387:611–7. [DOI] [PubMed] [Google Scholar]
  • 9.Garcia GE, Xia Y, Chen S, Wang Y, Ye RD, Harrison JK, et al. NF-kappaB-dependent fractalkine induction in rat aortic endothelial cells stimulated by IL-1β, TNF-α, and LPS. J Leukoc Biol. 2000;67:577–84. [DOI] [PubMed] [Google Scholar]
  • 10.Hirose S, Lin Q, Ohtsuji M, Nishimura H, Verbeek JS. Monocyte subsets involved in the development of systemic lupus erythematosus and rheumatoid arthritis. Int Immunol. 2019;31:687–96. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Qiu F, Li Y, Zhu Y, Li G, Lei F, Zhang S, et al. CX3CR1 might be a promising predictor of systemic lupus erythematosus patients with pulmonary fibrosis. Scand J Immunol. 2021;94: e13038. [DOI] [PubMed] [Google Scholar]
  • 12.Yajima N, Kasama T, Isozaki T, Odai T, Matsunawa M, Negishi M, et al. Elevated levels of soluble fractalkine in active systemic lupus erythematosus: potential involvement in neuropsychiatric manifestations. Arthritis Rheum. 2005;52:1670–5. [DOI] [PubMed] [Google Scholar]
  • 13.Segerer S, Hughes E, Hudkins KL, Mack M, Goodpaster T, Alpers CE. Expression of the fractalkine receptor (CX3CR1) in human kidney diseases. Kidney Int. 2002;62:488–95. [DOI] [PubMed] [Google Scholar]
  • 14.Nakatani K, Yoshimoto S, Iwano M, Asai O, Samejima K, Sakan H, et al. Fractalkine expression and CD16+ monocyte accumulation in glomerular lesions: association with their severity and diversity in lupus models. Am J Physiol Renal Physiol. 2010;299:F207–16. [DOI] [PubMed] [Google Scholar]
  • 15.Inoue A, Hasegawa H, Kohno M, Ito MR, Terada M, Imai T, et al. Antagonist of fractalkine (CX3CL1) delays the initiation and ameliorates the progression of lupus nephritis in MRL/lpr mice. Arthritis Rheum. 2005;52:1522–33. [DOI] [PubMed] [Google Scholar]
  • 16.Feng L, Chen S, Garcia GE, Xia Y, Siani MA, Botti P, et al. Prevention of crescentic glomerulonephritis by immunoneutralization of the fractalkine receptor CX3CR1 rapid communication. Kidney Int. 1999;56:612–20. [DOI] [PubMed] [Google Scholar]
  • 17.Liao X, Ren J, Reihl A, Pirapakaran T, Sreekumar B, Cecere TE, et al. Renal-infiltrating CD11c+ cells are pathogenic in murine lupus nephritis through promoting CD4+ T cell responses. Clin Exp Immunol. 2017. 10.1111/cei.13017. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Hochheiser K, Heuser C, Krause TA, Teteris S, Ilias A, Weisheit C, et al. Exclusive CX3CR1 dependence of kidney DCs impacts glomerulonephritis progression. J Clin Invest. 2013;123:4242–54. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Liu F, Dai S, Feng D, Qin Z, Peng X, Sakamuri S, et al. Distinct fate, dynamics and niches of renal macrophages of bone marrow or embryonic origins. Nat Commun. 2020;11:2280. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Cormican S, Griffin MD. Fractalkine (CX3CL1) and its receptor CX3CR1: a promising therapeutic target in chronic kidney disease? Front Immunol. 2021;12: 664202. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Grone HJ, Cohen CD, Grone E, Schmidt C, Kretzler M, Schlondorff D, et al. Spatial and temporally restricted expression of chemokines and chemokine receptors in the developing human kidney. J Am Soc Nephrol. 2002;13:957–67. [DOI] [PubMed] [Google Scholar]
  • 22.Landsman L, Bar-On L, Zernecke A, Kim KW, Krauthgamer R, Shagdarsuren E, et al. CX3CR1 is required for monocyte homeostasis and atherogenesis by promoting cell survival. Blood. 2009;113:963–72. [DOI] [PubMed] [Google Scholar]
  • 23.McDermott DH, Fong AM, Yang Q, Sechler JM, Cupples LA, Merrell MN, et al. Chemokine receptor mutant CX3CR1- M280 has impaired adhesive function and correlates with protection from cardiovascular disease in humans. J Clin Invest. 2003;111:1241–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Mionnet C, Buatois V, Kanda A, Milcent V, Fleury S, Lair D, et al. CX3CR1 is required for airway inflammation by promoting T helper cell survival and maintenance in inflamed lung. Nat Med. 2010;16:1305–12. [DOI] [PubMed] [Google Scholar]
  • 25.Bonacina F, Martini E, Svecla M, Nour J, Cremonesi M, Beretta G, et al. Adoptive transfer of CX3CR1 transduced-T regulatory cells improves homing to the atherosclerotic plaques and dampens atherosclerosis progression. Cardiovasc Res. 2021;117:2069–82. [DOI] [PubMed] [Google Scholar]
  • 26.Dong L, Nordlohne J, Ge S, Hertel B, Melk A, Rong S, et al. T cell CX3CR1 mediates excess atherosclerotic inflammation in renal impairment. J Am Soc Nephrol. 2016;27:1753–64. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Dagkalis A, Wallace C, Hing B, Liversidge J, Crane IJ. CX3CR1- deficiency is associated with increased severity of disease in experimental autoimmune uveitis. Immunology. 2009;128:25–33. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Beli E, Dominguez JM 2nd, Hu P, Thinschmidt JS, Caballero S, Li Calzi S, et al. CX3CR1 deficiency accelerates the development of retinopathy in a rodent model of type 1 diabetes. J Mol Med (Berl). 2016;94:1255–65. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Kawamura N, Katsuura G, Yamada-Goto N, Novianti E, Inui A, Asakawa A. Reduced brain fractalkine-CX3CR1 signaling is involved in the impaired cognition of streptozotocin-treated mice. IBRO Rep. 2020;9:233–40. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Mai W, Liu X, Wang J, Zheng J, Wang X, Zhou W. Protective effects of CX3CR1 on autoimmune inflammation in a chronic EAE model for MS through modulation of antigen-presenting cell-related molecular MHC-II and its regulators. Neurol Sci. 2019;40:779–91. [DOI] [PubMed] [Google Scholar]
  • 31.Ridderstad Wollberg A, Ericsson-Dahlstrand A, Jureus A, Ekerot P, Simon S, Nilsson M, et al. Pharmacological inhibition of the chemokine receptor CX3CR1 attenuates disease in a chronic-relapsing rat model for multiple sclerosis. Proc Natl Acad Sci USA. 2014;111:5409–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Yadav AK, Kumar V, Jha V. Association of chemokine receptor CX3CR1 V249I and T280M polymorphisms with chronic kidney disease. Indian J Nephrol. 2016;26:275–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.De Lema GP, Lucio-Cazaña FJ, Molina ANA, Luckow B, Schmid H, de Wit C, et al. Retinoic acid treatment protects MRL/lpr lupus mice from the development of glomerular disease. Kidney Int. 2004;66:1018–28. [DOI] [PubMed] [Google Scholar]
  • 34.Zhang H, Liao X, Sparks JB, Luo XM, Schloss PD. Dynamics of gut microbiota in autoimmune lupus. Appl Environ Microbiol. 2014;80:7551–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Mu Q, Zhang H, Liao X, Lin K, Liu H, Edwards MR, et al. Control of lupus nephritis by changes of gut microbiota. Microbiome. 2017;5:73. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Mu Q, Edwards MR, Swartwout BK, Cabana Puig X, Mao J, Zhu J, et al. Gut microbiota and bacterial DNA suppress autoimmunity by stimulating regulatory B cells in a murine model of lupus. Front Immunol. 2020;11: 593353. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Mu Q, Cabana-Puig X, Mao J, Swartwout B, Abdelhamid L, Cecere TE, et al. Pregnancy and lactation interfere with the response of autoimmunity to modulation of gut microbiota. Microbiome. 2019;7:105. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Mu Q, Swartwout BK, Edwards M, Zhu J, Lee G, Eden K, et al. Regulation of neonatal IgA production by the maternal microbiota. Proc Natl Acad Sci. 2021. 10.1073/pnas.2015691118. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for illumina sequence data. Bioinformatics. 2014;30:2114–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Callahan BJ, McMurdie PJ, Rosen MJ, Han AW, Johnson AJ, Holmes SP. DADA2: high-resolution sample inference from Illumina amplicon data. Nat Methods. 2016;13:581–3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.McMurdie PJ, Holmes S. phyloseq: an R package for reproducible interactive analysis and graphics of microbiome census data. PLoS ONE. 2013;8: e61217. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Martin BD, Witten D, Willis AD. Modeling microbial abundances and dysbiosis with beta-binomial regression. Ann Appl Stat. 2020;14:94–115. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, O’Hara RB, Simpson GL, Solymos P, Stevens MHH, Wagner H. Vegan: community ecology package. R Package Version 2.2–0; 2014. http://CRAN.Rproject.org/package=vegan. [Google Scholar]
  • 44.Geng S, Zhang Y, Yi Z, Lu R, Li L. Resolving monocytes generated through TRAM deletion attenuate atherosclerosis. JCI Insight 2021; 6(20):e149651. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Geng S, Chen K, Yuan R, Peng L, Maitra U, Diao N, et al. The persistence of low-grade inflammatory monocytes contributes to aggravated atherosclerosis. Nat Commun. 2016;7:13436. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Liao X, Ren J, Wei CH, Ross AC, Cecere TE, Jortner BS, et al. Paradoxical effects of all-trans-retinoic acid on lupus-like disease in the MRL/lpr mouse model. PLoS One. 2015;10: e0118176. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Reilly CM, Olgun S, Goodwin D, Gogal RM Jr, Santo A, Romes- burg JW, et al. Interferon regulatory factor-1 gene deletion decreases glomerulonephritis in MRL/lpr mice. Eur J Immunol. 2006;36:1296–308. [DOI] [PubMed] [Google Scholar]
  • 48.Cabana-Puig X, Bond JM, Wang Z, Dai R, Lu R, Lin A, et al. Phenotypic drift in lupus-prone MRL/lpr mice: potential roles of microRNAs and gut microbiota. Immunohorizons. 2022;6:36–46. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Schneider KM, Bieghs V, Heymann F, Hu W, Dreymueller D, Liao L, et al. CX3CR1 is a gatekeeper for intestinal barrier integrity in mice: limiting steatohepatitis by maintaining intestinal homeostasis. Hepatology. 2015;62:1405–16. [DOI] [PubMed] [Google Scholar]
  • 50.Medina-Contreras O, Geem D, Laur O, Williams IR, Lira SA, Nusrat A, et al. CX3CR1 regulates intestinal macrophage homeostasis, bacterial translocation, and colitogenic Th17 responses in mice. J Clin Invest. 2011;121:4787–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 51.McGaha TL, Chen Y, Ravishankar B, van Rooijen N, Karlsson MC. Marginal zone macrophages suppress innate and adaptive immunity to apoptotic cells in the spleen. Blood. 2011;117:5403–12. [DOI] [PubMed] [Google Scholar]
  • 52.Qiao JH, Castellani LW, Fishbein MC, Lusis AJ. Immune-complex-mediated vasculitis increases coronary artery lipid accumulation in autoimmune-prone MRL mice. Arterioscler Thromb. 1993;13:932–43. [DOI] [PubMed] [Google Scholar]
  • 53.Geng S, Zhang Y, Lee C, Li L. Novel reprogramming of neutrophils modulates inflammation resolution during atherosclerosis. Sci Adv. 2019;5: eaav2309. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Pradhan K, Yi Z, Geng S, Li L. Development of exhausted memory monocytes and underlying mechanisms. Front Immunol. 2021;12: 778830. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 55.Aicher A, Hayden-Ledbetter M, Brady WA, Pezzutto A, Richter G, Magaletti D, et al. Characterization of human inducible costimulator ligand expression and function. J Immunol. 2000;164:4689–96. [DOI] [PubMed] [Google Scholar]
  • 56.Wikenheiser DJ, Stumhofer JS. ICOS co-stimulation: friend or foe? Front Immunol. 2016;7:304. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Li H, Adamopoulos IE, Moulton VR, Stillman IE, Herbert Z, Moon JJ, et al. Systemic lupus erythematosus favors the generation of IL-17 producing double negative T cells. Nat Commun. 2020;11:2859. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Rodriguez-Rodriguez N, Flores-Mendoza G, Apostolidis SA, Rosetti F, Tsokos GC, Crispin JC. TCR-α/β CD4(−) CD8(−) double negative T cells arise from CD8(+) T cells. J Leukoc Biol. 2020;108:851–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Li H, Tsokos GC. Double-negative T cells in autoimmune diseases. Curr Opin Rheumatol. 2021;33:163–72. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Chen PM, Tsokos GC. The role of CD8+ T-cell systemic lupus erythematosus pathogenesis: an update. Curr Opin Rheumatol. 2021;33:586–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Shivakumar S, Tsokos GC, Datta SK. T cell receptor alpha/ beta expressing double-negative (CD4−/CD8−) and CD4+ T helper cells in humans augment the production of pathogenic anti-DNA autoantibodies associated with lupus nephritis. J Immunol. 1989;143:103–12. [PubMed] [Google Scholar]
  • 62.Crispin JC, Oukka M, Bayliss G, Cohen RA, Van Beek CA, Stillman IE, et al. Expanded double negative T cells in patients with systemic lupus erythematosus produce IL-17 and infiltrate the kidneys. J Immunol. 2008;181:8761–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Mizui M, Koga T, Lieberman LA, Beltran J, Yoshida N, Johnson MC, et al. IL-2 protects lupus-prone mice from multiple end-organ damage by limiting CD4-CD8- IL-17-producing T cells. J Immunol. 2014;193:2168–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Cabeza-Cabrerizo M, Cardoso A, Minutti CM, Pereira da Costa M, Reis e Sousa C. Dendritic cells revisited. Annu Rev Immunol. 2021;39:131–66. [DOI] [PubMed] [Google Scholar]
  • 65.Gottschalk C, Kurts C. The debate about dendritic cells and macrophages in the kidney. Front Immunol. 2015;6:435. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Arnold IC, Mathisen S, Schulthess J, Danne C, Hegazy AN, Powrie F. CD11c(+) monocyte/macrophages promote chronic Helicobacter hepaticus-induced intestinal inflammation through the production of IL-23. Mucosal Immunol. 2016;9:352–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Lu L, Kuroishi T, Tanaka Y, Furukawa M, Nochi T, Sugawara S. Differential expression of CD11c defines two types of tissue-resident macrophages with different origins in steady-state salivary glands. Sci Rep. 2022;12:931. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Cao Q, Wang Y, Wang XM, Lu J, Lee VW, Ye Q, et al. Renal F4/80+ CD11c+ mononuclear phagocytes display phenotypic and functional characteristics of macrophages in health and in adriamycin nephropathy. J Am Soc Nephrol. 2015;26:349–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Borges da Silva H, Fonseca R, Pereira RM, Cassado Ados A, Alvarez JM, D’Imperio Lima MR. Splenic macrophage subsets and their function during blood-borne infections. Front Immunol. 2015;6:480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Helmke A, Nordlohne J, Balzer MS, Dong L, Rong S, Hiss M, et al. CX3CL1-CX3CR1 interaction mediates macrophage-mesothelial cross talk and promotes peritoneal fibrosis. Kidney Int. 2019;95:1405–17. [DOI] [PubMed] [Google Scholar]
  • 71.Kassianos AJ, Wang X, Sampangi S, Afrin S, Wilkinson R, Healy H. Fractalkine-CX3CR1-dependent recruitment and retention of human CD1c+ myeloid dendritic cells by in vitro-activated proximal tubular epithelial cells. Kidney Int. 2015;87:1153–63. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary File
NIHMS1951314-supplement-Supplementary_File.pdf (1.7MB, pdf)

Data Availability Statement

The 16S rRNA sequences are available in the National Institutes of Health SRA database (PRJNA835508).

RESOURCES