To evaluate the pathogenic role of macrophage-derived MIF, we employed a Cre-loxP strategy to generate mice with a specific deletion of MIF within their macrophages. To confirm the efficiency of MIF deletion, bone marrow was isolated from MIF^f/f-lysM-cre mice, control MIF^f/f mice, and MIF KO littermates. The isolated bone marrow cells were cultured and induced to differentiate into macrophages. As shown in Figure 1A , the deficiency of macrophage-derived MIF led to a notable reduction in MIF secretion by BMDM ( Figure 1A ). Additionally, MIF and CD74 mRNA expression were significantly lower in macrophages of MIF^f/f-lysM-cre mice compared with MIF^f/f mice under basal conditions ( Figures 1B, C ). Furthermore, TNF-a treatment for 24 hours significantly increased MIF protein expression in macrophages derived from MIF^f/f mice but not in macrophages from MIF^f/f-lysM-cre mice ( Figure 1D ). These results demonstrated the successful deletion of MIF from macrophages in MIF^f/f-lysM-cre mice.

To examine the involvement of macrophage-derived MIF in experimental anti-GBM CGN, we conducted experiments using both MIF^f/f and MIF^f/f-lysM-cre mice. The mice were induced to develop anti-GBM CGN, and the renal injuries were evaluated. Notably, the MIF^f/f-lysM-cre mice showed a significant inhibition in renal injuries such as segmental glomerular capillary necrosis and crescent formation compared to the anti-GBM GN MIF^f/f mice ( Figures 2A, B ). Renal dysfunction such as the urine albumin/creatinine ratio ( Figure 2C ), serum creatinine levels ( Figure 2D ), and creatinine clearance ( Figure 2E ) were also significantly improved in MIF^f/f-lysM-cre mice compared to the anti-GBM CGN MIF^f/f mice. These findings indicate that macrophage-derived MIF plays a pathogenic role in anti-GBM CGN.

We next examined macrophage-derived MIF on cellular immune response during anti-GBM CGN. Immunohistochemistry detected that a massive F4/80+ macrophages infiltrating the anti-GBM CGN in MIF^f/f mice, which was largely inhibited in MIF^f/f-lysM-cre mice ( Figures 3A-C ). Moreover, the infiltration of glomerular and interstitial CD3+ T cells in MIF^f/f-lysM-cre GN mice was also significantly lower than in MIF^f/f GN mice ( Figures 3D-F ). These findings demonstrate that selective depletion of MIF in macrophages suppresses the infiltration of macrophages and T cells in the kidney during anti-GBM GN.

Immunofluorescence was used to assess the glomerular deposition of sheep anti-mouse GBM antibody, mouse IgG, and complement component C3 in MIFf/f and MIFf/f-lysM-cre mice ( Figure 4A ). Interestingly, there was no significant difference in the glomerular deposition of these markers between the two groups, indicating that macrophage-specific MIF depletion did not affect immune complex deposition in inflamed glomeruli. However, as demonstrated in Figure 4B , the serum levels of mouse anti-sheep IgG antibodies were notably reduced in MIF^f/f-lysM-cre GN mice compared to MIF^f/f GN mice. This was associated with significant reduction in serum MIF levels in MIF^f/f-lysM-cre GN mice compared to MIF^f/f GN mice ( Figure 4C ). These results indicate that macrophage-specific MIF depletion reduced systemic MIF levels and selectively inhibited the antigen-specific antibody production without influencing the immune complex deposition in the inflamed glomeruli.

To investigate the impact of macrophage-derived MIF on macrophage polarization in the kidneys of mice with anti-GBM CGN, we utilized flow cytometry to determine the populations of M1 (F4/80+CD86+) and M2 (F4/80+CD206+) macrophages. Notably, MIF^f/f-lysM-cre GN mice exhibited a significantly decreased M1 macrophages while increasing the M2 macrophages compared to MIF^f/f GN mice ( Figures 5A, B ; Supplementary Figure S1 ). Furthermore, real-time PCR analysis revealed that selective MIF depletion in macrophages led to a significant inhibition of pro-inflammatory cytokines including MCP-1 and IL-1β while increasing the anti-inflammatory cytokine IL-10 in CGN mice ( Figures 5C–E ). These findings indicate that selective depletion of MIF in macrophages results in a shift in macrophage polarization from the pro-inflammatory M1 phenotype towards the anti-inflammatory M2 phenotype in experimental anti-GBM CGN. Moreover, this shift is associated with reduced pro-inflammatory cytokine expression and increased anti-inflammatory cytokine expression.

We next examined whether disrupted macrophage MIF influences T cell immunity as it is well-established that Th1 and Th17 are pathogenic whereas Treg is protective in anti-GBM CGN (23–26). In light of this, we investigated the impact of selective MIF depletion in macrophages on the immune differentiation of CD4+ T cells, specifically focusing on Th1 (CD4⁺IFNγ⁺), Th2 (CD4⁺IL-4⁺), Th17 (CD4⁺IL-17a⁺) and Treg (CD4⁺CD25⁺FoxP3⁺) subpopulations. Flow cytometry analysis revealed that selective MIF depletion in macrophages led to a significant reduction in Th1 (CD4⁺IFNγ⁺) and Th17 (CD4⁺IL-17a⁺) cells in the anti-GBM CGN kidney ( Figures 6A, C ; Supplementary Figures S2A, C ). Conversely, there was an increase in Treg population (CD4⁺CD25⁺FoxP3⁺), however, the Th2 (CD4+IL-4+) immune response remained unaffected ( Figures 6D, B ; Supplementary Figure S3 , Supplementary Figure S2B ). These findings indicate that selective MIF depletion in macrophages enhances the Treg immune response while inhibiting the Th1 and Th17 immune responses in experimental anti-GBM CGN. This shift in immune cell differentiation may contribute to the amelioration of renal injury observed in this context.

We next examined the mechanisms through which specific deletion of macrophage MIF inhibits anti-GBM GN induced in MIF ^f/f and MIF ^f/f-lysM-cre mice. Western blot analysis revealed that there was a marked upregulation of CD74 and activation of NF-κB/p65 and p38 MAPK signaling and expression of iNOS in the diseased kidney of MIF ^f/f mice ( Figure 7 ). In contrast, selective MIF depletion from macrophages significantly inhibited the expression of CD74 and phosphorylation of NF-κB/p65 and p38 MAPK, as well as expression of iNOS in MIFf/f-lysM-cre GN mice ( Figure 7 ). All of these findings indicated that MIF may promote anti-GBM GN by activating M1 macrophages through the CD74/NF-kB/p38 MAPK signaling. This was further demonstrated in vitro in cultured BMDM from MIF WT and MIF KO mice. We found that addition of TNF-α largely promoted iNOS-producing M1 macrophages in MIF WT BMDM by activating CD74/NF-κB/p38 MAPK signaling, which was blocked in BMDM lacking MIF ( Figure 8 ). Thus, macrophage-derived MIF may mediate anti-GBM GN by promoting M1 macrophage activation via the CD74/NK-κB/p38 MAPK-dependent mechanism.

To generate mice with myeloid-specific MIF deletion (MIF^{f/f−lysM−cre}), we utilized C57BL/6 mice carrying MIF genes with homozygous loxP-flanked regions (MIF^f/f), which were previously described (53). Lysozyme M promoter-driven cre (lysM-cre) mice were obtained from the Jackson Laboratory in the USA. The MIF^f/f mice were crossed with lysM-cre mice to obtain the desired genotype. The genotypes of the resulting littermates were confirmed using PCR with specific primers recommended by the Jackson Laboratory. All mice used in the study were maintained under specific pathogen-free conditions at a temperature of 25°C and a 12-hour light-dark cycle. They were housed in our animal facility and provided with standard food and water ad libitum.

To culture bone marrow-derived macrophages (BMDMs), bone marrow cells were isolated from the tibias and femurs of MIF^f/f, MIF^{f/f−lysM−cre}, and MIF knockout (KO) mice. These cells were then cultured in Dulbecco modified Eagle medium (DMEM) supplemented with 50 ng/ml macrophage colony-stimulating factor (M-CSF) for 7 days, following a previously established protocol (54). After the 7-day culture period, BMDMs were stimulated with TNF-α (10 ng/ml) for 24 hours and proteins were collected for western blot analysis.

Total RNA was extracted from either diseased kidney tissue or cultured cells using the RNeasy Isolation Kit (Qiagen, Valencia, CA) following the manufacturer’s protocol. RT-PCR was conducted following established methods (44, 55). The mRNA expression levels of the target genes were normalized to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) as an internal control. The primer sequences for mouse MIF, CD74, MCP-1, IL-1β, IL-10, and GAPDH were previously reported (44, 55, 56).

The mouse model of anti-glomerular basement membrane crescentic glomerulonephritis (anti-GBM cGN) was established using male MIF^f/f and MIF^{f/f−lysM−cre} mice (8–12 weeks old), following a well-established protocol (44). Herein, the procedure involved the following steps: Firstly, groups of MIF^f/f and MIF^{f/f−lysM−cre} mice received a flank subcutaneous immunization with normal sheep IgG mixed with Freund’s complete adjuvant (Sigma Aldrich, St. Louis, Missouri, USA) five days prior to the initiation of the experiment. Anti-GBM cGN was induced by administering sheep anti-mouse GBM IgG via tail vein injection at a dose of 60 μg/g of body weight (referred to as day 0). On day 14, the mice were sacrificed using a lethal dose of ketamine and xylazine mixture. Age-matched normal male MIFf/f and MIFf/f−lysM−cre mice (n = 5 per group) were included as normal control counterparts. All animal experiments were conducted in compliance with the guidelines approved by the Animal Experimentation Ethics Committee at the Chinese University of Hong Kong.

Analysis of proteinuria: Urinary samples were collected at various time points, including before and after the induction of anti-GBM disease on days 0, 1, 3, 7, and 14. Proteinuria analysis was performed following the manufacturer’s protocols, as described previously (44). Urinary albumin excretion was quantified as total urinary albumin/creatinine ratio (expressed as micrograms per milligram). Measurement of urinary and serum creatinine levels was conducted using an enzymatic kit (Stanbio Laboratory, Boerne, USA).

Renal pathology assessment was conducted on methyl Carnoy’s fixed, paraffin-embedded tissue sections (4μm thick). The sections were stained with periodic acid Schiff (PAS) to visualize the renal structures. Glomerular crescents and necrosis were quantified by examining 50 glomeruli per diseased kidney section and calculating the percentage of affected glomeruli. Immunohistochemistry was performed on paraffin sections stained with monoclonal an anti-F4/80 antibody (Serotec, Oxford, UK) to detect macrophages, and a rabbit anti-mouse polyclonal CD3+ antibody (SP7) (Abcam, Cambridge, UK) to identify total T cells. The number of positive cells for CD3 and F4/80 was counted in 20 glomeruli and expressed as cells per glomerular cross-section (gcs). In the tubulointerstitium, positive cells were counted under high-intensity fields (400× magnification) using a 0.0625 mm² graticule fitted in the microscope eyepiece. The cell count was then reported as cells per square millimeter (mm²⁾.

To assess the presence of MIF-expressing macrophages in the kidney, immunofluorescence staining was performed. Acetone-fixed bone marrow-derived macrophages (BMDMs) were cultured with the MIF antibody (sc-20121; Santa Cruz) followed by incubation with a fluorescein isothiocyanate anti-rabbit secondary antibody, as previously described (54). To evaluate immune deposition in the glomeruli, direct immunofluorescence was conducted using FITC-conjugated polyclonal antibodies specific to sheep IgG, mouse IgG, and complement C3, following established protocols (38).

Plasma levels of mouse anti-sheep IgG were quantified using a method previously described (56). The concentration of serum MIF was determined using ELISA kits (R&D Systems, Minneapolis, USA) according to the manufacturer’s instructions.

Kidney single cells were prepared and subjected to flow cytometry analysis following a previously described protocol (44, 56). Single-cell suspensions were treated with IC Fixation Buffer and Permeabilization Buffer (eBioscience) to allow intracellular staining. The suspended kidney cells were then incubated with specific conjugated antibody cocktails in the dark for 30 minutes on ice. Negative controls included cells treated with irrelevant antibodies (isotype). Cells were also incubated with only one specific conjugated antibody. The antibodies used in this study were as follows: F4/80-Pacific blue (BioLegend, Catalog: 123124), CD86-APC (BioLegend, Catalog: 10512), CD206-Alexa 647 (Serotec, Catalog: MCA2235), CD4-FITC (eBioscience, Catalog: 11–0042-86), IFNγ-APC (eBioscience, Catalog: 17–7311-82), IL-4-PE (eBioscience, Catalog: 12–7041-82), IL-17a-PE (eBioscience, Catalog: 12–7177-81), CD25-PE (eBioscience, Catalog: 12–0251-83), and Foxp3-APC (eBioscience, Catalog: 12–0251-83). Flow cytometry analysis was performed using a FACS Calibar instrument and analyzed using the CellQuest Pro Analysis software (BD Biosciences, Franklin Lakes, New Jersey, USA).

Western blotting was performed as described previously (55, 57–59). Proteins from BMDMs and the kidney cortex were extracted with RIPA lysis buffer. After blocking nonspecific binding with 5% BSA, membranes were incubated overnight at 4 ° C with the primary antibodies against rabbit anti-iNOS (Abcam ab-15323), rabbit anti-iNOS (Abcam ab-178945), goat anti-CD74 (Santa Cruz, sc-5438), mouse anti-CD74 (Santa Cruz, sc-6267), goat anti-CD74 (Santa Cruz, sc-5438), rabbit anti-phosphorylated NF-kB p65 (Cell Signaling, #3031), rabbit anti-phosphorylated NF-kB p65 (Cell Signaling, #3033s), mouse anti-NF-kB p65 (Cell Signaling, #6965), rabbit anti-NF-kB p65 (Cell Signaling, # 8242S), rabbit anti-pp38 (Cell signaling, #9211), rabbit anti-pp38 (Cell signaling, # 4631L), rabbit anti-p38 (Cell signaling, #9212), rabbit anti-p38 (Cell signaling, # 8690S), mouse anti-ß-actin (Santa Cruz, sc-69879), mouse anti-GAPDH(proteintech,#60004–1-Ig).Then the membranes were incubated with IRDye800-conjugated secondary antibody (Rockland Immun- chemicals). Signals were scanned using the Odyssey IR imaging system (LI-COR Biosciences). Image J software (National Institutes of Health) was used for quantitative analysis of images.

Statistical analysis was conducted using Prism 9.0 GraphPad Software (GraphPad Software, La Jolla, California, USA). Data obtained from the study were presented as the mean ± standard error of the mean (SEM). Two-group comparisons were assessed using an independent sample t-test. Multiple group comparisons were performed using one-way analysis of variance (ANOVA) followed by Tukey’s post hoc tests. A p-value less than 0.05 was considered statistically significant in this experiment.