Healthy donors were recruited from medical staff at the Second Xiangya Hospital.

C57BL/6 mice were purchased from Slack Company (shanghai, China). The mouse forage was purchased from Dyets, Inc (Wuxi, China). For the LID mouse model, 3 weeks old female C57BL/6 mice were fed with low iron diet (5mg/kg) for 5 weeks, age-matched female B6 mice fed with normal iron diet (ND, 50mg/kg) were served as the controls. The nutrient elements of LID forage were consistent with the ND group except for the content of trace element iron. All mice were maintained in specific pathogen-free conditions. After 5 weeks of LID treatment, the spleen and dLNs (isolated from the inguinal lymph nodes) were collected for flow cytometric analysis, and CD4⁺T cells were isolated from the spleen for qPCR analysis.

For the pristane-induced lupus mouse model, 8-weeks old female mice were i.p. injected with 500 μl pristane(Sigma, catalog P9622) and fed with ND or LID for 6 months. After 6 months of pristane stimulation, mice were sacrificed for analysis. The urine protein was detected by a colorimetric assay strip (URIT). The spleen and dLNs (isolated from the inguinal lymph nodes) were collected for flow cytometric analysis, and the renal tissue was fixed in formalin and embedded in paraffin for histological analysis. Serum was collected for auto-antibody analysis.

Hematoxylin and eosin (H&E) staining were used to evaluate the morphological changes of the kidney. Based on the criterion of the previous study, renal damage can be scored from 0-3: 0, normal; 1, slight cell proliferation and cell infiltration; 2, mesangial proliferation and a lobular structure formation; 3, crescent formation with hyalinosis (15).

To determine the immune complex deposition in the kidney, paraffin-embedded renal sections were incubated with anti-mouse C3 antibody (Abcam, catalog ab200999) for mouse C3 staining and anti-mouse IgG antibody (Abcam, catalog ab205724) for mouse IgG staining. Fluorescence was labeled using Opal 7-color Manual IHC Kit (Perkin Elmer, catalog NEL811001KT). The image was captured by Perkin Elmer and analyzed by the Mantra system. The deposition of C3 and IgG was evaluated by Mean Fluorescence Intensity (MFI) using Fiji software (16).

For surface marker, cells were incubated with fluorochrome-conjugated antibodies against surface markers at 4°C for 30 min in the dark. For cytokine, cells were stimulated with PMA and ionomycin with the addition of GolgiPlug (BD Biosciences, catalog 550583) at 37°C and 5% CO2 for 6hr. For intracellular markers, cells were fixed and permeabilized with Cytofix/Cytoperm TM Fixation/Permeabilization Solution Kit (BD Biosciences, catalog 554714) or Foxp3/Transcription Factor Staining Buffer Set (eBioscience, catalog 00-5523-00), and then incubated with fluorochrome-conjugated antibodies at 4°C for an additional 30 min. For ROS detection, cells were loaded with DCFH-DA probe (Beyotime, catalog S0033S) at 37°C for 30 min protected from the light and washed twice with 1×PBS. For cell apoptosis detection, Annexin V Apoptosis Detection Kit I (BD bioscience, catalog 556547) was used according to the manufacturer’s instructions. Annexin V⁺ PI⁺ cells were determined as apoptotic cells. For intracellular iron detection, cells were loaded with FerroOrange (DOJINDO, catalog F374) at 37°C for 30 min in the dark. The following antibodies were used: FITC anti-mouse CD4 (Biolegend, catalog 100406), PE anti-mouse CD25(Biolegend, catalog 102007), AF488 anti-mouse CD25 (Invitrogen, catalog 53-0251-82), APC anti-mouse FOXP3 (Invitrogen, catalog 17-4776-42), APCCY7 anti-mouse IL-17A (BD Pharmingen, catalog 560821), PERCP-CY5.5 anti-mouse IFN-γ (BD Pharmingen, catalog 560660), PERCP-CY5.5 anti-mouse TNF-α (Biolegend, catalog 506321), BV786 anti-mouse T-bet (BD Pharmingen, catalog 564141), PE anti-mouse RORγt (BD Pharmingen, catalog 562607), PECY7 anti-human CD4 (Biolegend, catalog 317414), APC anti-human CD127 (Invitrogen, catalog 17-1278-42), BB700 anti-human CD127 (BD Pharmingen, catalog 566398), APC anti-human CD25(BD Pharmingen, catalog 555434), PE anti-human CD25 (Invitrogen, catalog 12-0259-42), APC anti-human FOXP3 (Invitrogen, catalog 17-4776-42).

Serum was collected for autoantibody detection. For anti-dsDNA IgGs, mouse anti-dsDNA IgG ELISA Kit (Alpha Diagnostic International, catalog 5120) was used according to the manufacturer’s instructions.

Total CD4⁺T cells were isolated from the spleen of mice by mouse CD4 microbeads (Miltenyi Biotec, catalog 130-117-043). TRIzol reagent (MRC, catalog TR118) was used to extract total RNA from cells. HiScript III All-in-one RT SuperMix (Vazyme, catalog R333) was used to generate cDNA according to the manufacturer’s instructions. ChamQ Universal SYBR qPCR Master Mix (Vazyme, catalog Q711) was used for selected gene amplification. The relative expression of selected genes was analyzed by ΔΔCt method, which normalized to the house-keeping gene β-actin. The following primers are used: β-actin: FP, GTGACGTTGACATCCGTAAAGA; RP, GCCGGACTCATCGTACTCC. Il2r: FP, TGGCAACACAGATGGAGGAAG; RP, ACAGCCGTTAGGTGAATGCT. Foxp3: FP, CCCCCTCTAGCAGTCCACTT; RP, AAGTTGCCGGGAGAGCTGAA. Tfrc: FP, TTCGCAGGCCAGTGCTAGG; RP, TACAAGGGAGTACCCCGACAG. Fth: FP, CAGACCGTGATGACTGGGAG; RP, TCAATGAAGTCACATAAGTGGGGA.

Naive CD4⁺T cells were isolated from the peripheral blood of healthy donors, and cultured in the presence of anti-CD3 2 μg/ml (Calbiochem, catalog 217570), and anti-CD28 1μg/ml (Calbiochem, catalog 217669). For Treg cell differentiation, recombinant human TGF-β1 5 ng/ml (R&D, catalog 240-B-002), and recombinant human IL-2 10 ng/ml (PeproTech, catalog 200-02-10). Cells were cultured under the Treg cell-polarized conditions supplemented with 10% FBS (HyClone) at 37°C and 5% CO2 for 72 hr. For ROS stimulation, cells were treated with ROSUP (50 ug/mL) (Beyotime, catalog S0033S-2) for 12 hr, and then harvested for subsequent experiments. For iron supplementation, cells were treated with Hemin 100 μM (Selleck, catalog S5645) for 72 hr. For ROS inhibitor rescue, SOD 300U (Beyotime, catalog S0087) was added into cells after 24 hr of Hemin treatment.

SPSS 25.0 software was used for statistical analysis. All statistical results were presented as the mean ± S.E.M. For normally distributed data, two-tailed Student’s t-test was used for evaluating the statistical differences between the two groups. For abnormally distributed data, two-tailed Mann-Whitney U-test was used for assessing the statistical differences between two groups. One-way analysis of variance (ANOVA) with relevant post hoc tests was used for multiple comparisons.

To explore the role of iron deficiency to Treg cell expansion, we treated 3-weeks old female c57/B6 mice with 5 mg/kg low iron diet (LID) for 5 weeks, mice treated with 50 mg/kg normal iron diet (ND) was served as the control group. After 5 weeks of treatment, the weight of LID-treated mice was slightly reduced compared with the ND group ( Supplementary Figure 1A ). We did not observe significant changes in the sizes of dLNs (including cervical, axillar, and inguinal lymph nodes) and spleen between the ND and LID groups ( Supplementary Figures 1B, C ). The percentage and number of total CD4⁺T cells were also similar between the two groups ( Supplementary Figures 2A, B ). Transferrin receptor (Tfrc) is responsible for cellular iron uptake (17). Fth encodes the ferritin heavy polypeptide 1 which controls intracellular iron storage when cells are in the iron-sufficient condition (18). To determine the intracellular iron homeostasis, we detected the expression of Tfrc (Transferrin receptor) and Fth 1 (Ferritin Heavy Chain 1) genes in the splenic CD4⁺T cells. As expected, the gene expression of Tfrc and Fth was significantly reduced in splenic CD4⁺T cells of LID-treated mice ( Figure 1A ). Next, we examined the changes of Treg cells after LID treatment. The expression of Treg cell-related genes Il2r and Foxp3 was significantly increased in splenic CD4⁺T cells of LID mice ( Figure 1B ). Consistent with the mRNA expression, the frequency of CD4⁺CD25⁺Foxp3⁺ Treg cells was significantly elevated in the dLNs (inguinal lymph nodes) of LID-treated mice compared with the ND controls, but it has no significant changes in the spleen ( Figures 1C, D ). However, the cell numbers of Treg cells were increased in both dLNs (inguinal lymph nodes) and spleen of LID-treated mice ( Figure 1E ). These results show that insufficient iron may contribute to Treg cell differentiation and expansion.

We assessed the disease progression of pristane-induced lupus in mice treated with ND or LID. 8-weeks old female C57/B6 mice were intraperitoneally injected with 500 μl pristane and fed with ND or LID for 6 months. Mice fed with 6 months of LID showed slight weight loss compared with the ND group ( Supplementary Figure 3A ). After 6 months of pristane stimulation, the urine protein level was reduced in mice treated with LID ( Figure 2A ). Morphological analysis also showed improved renal damage in LID-treated mice compared with the ND group ( Figures 2B, C ). Furthermore, the immune-complex deposition was reduced in the kidney of LID-treated mice ( Figure 2D ), and the titer of anti-dsDNA IgGs was also significantly decreased in mice fed with LID ( Figure 2E ). These results indicate that insufficient iron improves the disease progression of pristane-induced lupus.

Since the defects in the differentiation and functions of Treg cells play an important role in the development of lupus, we asked whether insufficient iron can improve lupus by promoting Treg cell expansion. After 6 months of pristane stimulation, mice treated with LID showed an elevated percentage of Treg cells in the spleen ( Figures 3A, B ), and the cell numbers of Treg cells in the dLNs and spleen were increased ( Figure 3C ). To assess whether Treg cell expansion influence the Th17/Treg balance, which contributes to the disease progression of lupus, we determined the proportions of Th17 cells in pristane-treated mice fed with LID. We found that in the pristane-induced lupus mouse model, mice fed with LID showed the fewer frequency and number of CD4⁺ IL17A⁺ Th17 cells ( Figures 3D–F ) and the lower ratio of Th17/Treg cells compared with the ND controls ( Figure 3G ). These results suggest that insufficient iron promotes Treg cell expansion and reshapes the Th17/Treg cell balance in pristane-induced lupus-like diseases. In addition, we found that IFN-γ production was also reduced in the splenic CD4⁺T cells of LID-treated mice ( Supplementary Figure 4A ). However, the production of TNF-α and the expression of T-bet and RORγt were slightly reduced in LID-treated mice but without significant difference ( Supplementary Figures 4B-D ).

Next, we sought to explore the mechanism of how iron deficiency affects the expansion of Treg cells. Previous studies have reported that iron promotes intracellular ROS production, which is harmful for Treg cell differentiation (11, 19, 20). Therefore, we determined the ROS levels in CD4⁺ T cells after LID treatment using DCFH probe (21, 22). We found that 5-weeks of LID treatment reduced the ROS levels in the CD4⁺ T cells of dLNs ( Figures 4A, B ), but there was no significant difference in the spleen ( Figures 4C, D ), which were consistent with the frequencies of Treg cells in the dLNs and spleen after 5 weeks of LID treatment ( Figure 1D ), indicating that Treg cell expansion is sensitive to intracellular ROS production.

To confirm whether ROS affects the differentiation of Treg cells, we used ROSUP to induce the overproduction of ROS during in vitro differentiation process of Treg cells. We isolated naive CD4⁺T cells from the peripheral blood mononuclear cells (PBMCs) of healthy donors. The cells were cultured in Treg cell-polarized conditions for 3 days and stimulated with ROSUP for 12 hr. ROSUP treatment significantly increased the level of ROS in induced Treg cells compared with the control group ( Figure 5A ). The percentage of Treg cells (gated by CD4⁺CD25⁺CD127^- and CD4⁺CD25⁺FOXP3⁺) was also significantly decreased in ROSUP treated group, which suggested that ROS impaired the differentiation of Treg cells ( Figures 5B, C ). ROS is harmful to cell biology, which can induce mitochondrial dysfunction and cell death. Therefore, we determined the cell apoptosis after ROSUP treatment. The result showed that ROSUP significantly elevated the frequency of cell death compared with the control group ( Figure 5D ). Next, we asked whether iron supplementation inhibits Treg cell differentiation by increasing ROS levels. FerroOrange was used to determine the levels of ferrous iron after iron supplementation treatment (23, 24). The results showed that iron supplementation by Hemin significantly increased the levels of ferrous iron and ROS in induced Treg cells ( Figures 5E, F ), and impaired the differentiation of Treg cells in vitro ( Figure 5G ). However, ROS inhibitor SOD rescued the defect of Treg cell differentiation in Hemin-treated cells by reducing ROS levels ( Figures 5H, I ). These data suggest that overproduction of ROS inhibits Treg cell differentiation by promoting cell apoptosis, and iron deficiency might promote Treg cell expansion by diminishing intracellular ROS accumulation.