Wild-type (WT; C57BL/6J) mice were purchased from the Jackson Laboratory. Breeding pairs of Bmal1-myeloid conditional knockout (Bmal1^Mye−/− ; Bmal1FloxP/FloxP;LysMCre) mice were created by breeding B6.129S4(Cg)-Bmal1^tm1Weit/J (JAX007668) with B6.129P2-Lyz2^tm1(cre)Ifo/J (JAX004781) and kept at the National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) animal facility. In total, 74 WT and 70 Bmal1^Mye−/− mice (all female) were used for the experiments. Each dot in the figures reporting animal experiments represents an individual. All mice were kept under specific pathogen-free conditions, and all procedures were performed in accordance with NIH guidelines and approved by the NIAMS Animal Care and Use Committee (ACUC) (protocol numbers are A019-05-03 and A022-08-04). All animals were euthanized in a CO₂ chamber, followed by cervical dislocation once they stopped breathing. No anesthesia was used for the procedures.

Imiquimod (IMQ) cream (0.125 g of 5%; Fougera or Perrigo) was applied on the dorsal side of both ears of 8–10-week-old female mice, three times per week for 6 weeks, as previously described (8). The baseline body weight of the mice was around 20 g. Treatment was done in the evening (around Zeitgeber time (ZT) 11). Blinding or randomization for treatment was not performed. Peripheral blood, spleen, kidney, and ear skin samples were harvested at euthanasia. Mouse euthanasia was done in the morning (ZT 1–4) for all experiments. Animals who died before completing a 6-week treatment period were excluded from the analysis except for survival analysis.

Mouse kidney, ear pinna, and sternum bone marrow were harvested and fixed with 10% formalin at euthanasia. Samples were stored in 10% formalin until slides were prepared. Samples were stained with hematoxylin and eosin (H&E) for ear pinna or H&E and periodic acid–Schiff stain (PAS) for kidneys. The severity of inflammation was semiquantitatively scored by a veterinary pathologist (VH) blinded to background information and sample identity. The number of infiltrating neutrophils was counted on five randomly selected high-power fields of H&E-stained slides. Scoring was performed as follows: for kidneys, glomeruli ware scores on PAS-stained slides. The total scores were generated by adding up the severity and distribution score for hypercellularity, mesangial matrix expansion, and inflammation. The definition of distribution was as follows: 0 = <1%, 1 = 1%–25%, 2 = 25%–75%, 3 = >75%. The definition of severity was as follows: 0 = no significant findings (within normal limits), 1 = mild (focal segmental), 2 = moderate (multifocal segmental), 3 = severe (global) (9). The detailed explanation of each score was provided in Supplementary Table 1 .

For ear pinna, the severity and distribution in the epidermis and dermis were scored. For epidermis, the score is as follows: 0 = normal, 1 = mild (hyperplasia two to three layers thick), 2 = moderate (acanthosis with an occasional increase in thickness of the granular layer), 3 = severe (dyskeratosis, diffuse increase in granular layer, hyperkeratosis rete peg formation). For dermis, the score is as follows: 0 = normal, 1 = mild (slight increase in cellularity; perivascular cuffs), 2 = moderate (definite increase in cellularity; fibrotic foci, folliculitis), 3 = severe (pustules; extensive fibrosis, vacuoles/clefting associated with basement membrane).

For bone marrow immunohistochemistry (IHC), slides were treated with H₂O₂ to block endogenous peroxidases and incubated with Polyclonal Rabbit anti−Human TNFSF13/APRIL Antibody (aa151−200) (LS Bio) at 10 μg/mL. April was visualized with anti-rabbit-HRP secondary antibody and 3,3′-diaminobenzidine (DAB). Finally, slides were counterstained with hematoxylin, dehydrated, and cleared in graded alcohols and xylene, respectively. Images were obtained with a NanoZoomer-XR digital slide scanner (Hamamatsu). Whole bone marrow images were exported as TIFF files, and April-positive area was quantified by ImageJ ver. 1.54i 03.

Tissue specimens were put in RNAlater (Thermo Fisher Scientific) at euthanasia of mice and stored at −80°C until they were processed. They were frozen in liquid nitrogen, ground into a fine powder by pestle and mortar, and lysed by TRIzol reagent (Thermo Fisher Scientific). Cell samples were lysed by TRIzol reagent immediately after isolation and stored at −80°C. RNA was purified by using Direct-zol RNA kits (Zymo Research) following the manufacturer’s instructions and quantified by NanoDrop Microvolume Spectrophotometer (Thermo Fisher Scientific). The quality of RNA samples was evaluated by the ratio of absorbance at 260 to 280. cDNA was synthesized by iScript reverse transcription supermix (Bio-Rad). Quantitative PCR (qPCR) was performed on a CFX96 real-time PCR detection system (Bio-Rad) with TaqMan gene expression assays (Thermo Fisher Scientific). Expression of each gene was normalized to Gapdh, and fold change was calculated using the ΔΔCt method. Cxcl1 and Cxcl2 expression in ear skin were evaluated as they are known chemoattractants of neutrophils. Irf7 and Isg15 in kidney were evaluated as IFN-regulated genes and based on our previous reports using the same murine lupus model (10, 11). Bone marrow neutrophil Baff, April, and Ifna4 were evaluated based on previous reports (2).

Mouse bone marrow cells were harvested from bilateral femurs and tibias. Neutrophils and monocytes were isolated from bone marrow cells using MACS-negative selection with a mouse neutrophil isolation kit and mouse monocyte isolation kit (BM) (Miltenyi Biotec), respectively. Purity of CD11b⁺Ly6G⁺ neutrophils was assessed using flow cytometry and was found to be higher than 95%.

The NET-forming capacity of bone marrow neutrophils was assessed by the IncuCyte S3 Live-Cell Analysis System (Sartorius), as previously described (12, 13). In brief, nuclei of neutrophils were stained with NUCLEAR-ID Red DNA stain (Enzo), and 15,000/well of neutrophils were plated onto a 96-well tissue culture-treated plate (Corning). Cells were then incubated in serum-free RPMI-1640 media (Thermo Fisher Scientific) with 1:25,000 diluted SYTOX Green (Thermo Fisher Scientific) for 16 hours at 37֯°C with 5% CO₂. For stimulation, 25 nM phorbol 12-myristate 13-acetate (PMA; MilliporeSigma) or 2.5 µM calcium ionophore A23187 (MilliporeSigma) was used. SYTOX Green-positive cells are defined as NET-forming neutrophils and percentage was assessed every 20 min by time-lapse imaging. In order to minimize the influence from inter-assay variability, equal numbers of samples from each group were included in the same plate and evaluated at the same time.

Immunofluorescent staining of NETs was performed as we previously reported (14). Bone marrow neutrophils were seeded on poly-L-lysine–coated coverslip and incubated for 5 h in RPMI without serum at 37°C in a CO₂ chamber. They were then fixed with 4% paraformaldehyde overnight at 4°C. Subsequently, coverslips were washed by PBS twice and blocked by 0.2% gelatin for 1 h at room temperature, incubated with 1:100 diluted rabbit polyclonal anti-histone H3 (citrulline R2 + R8 + R17) antibody (Abcam), washed three times, and stained with 1:400 diluted Alexa Fluor 555-conjugated donkey anti-Rabbit IgG (H+L) Antibody (Thermo Fisher Scientific). After washing, cells were counterstained with 1:1,000 diluted Hoechst (Thermo Fisher Scientific) and washed twice, and then visualized using a Leica DMI4000 B inverted microscope (Leica Microsystems). Specificity of the primary antibody was confirmed by staining only with the secondary antibody.

Serum citrullinated histone H3-DNA and neutrophil elastase–DNA complex were measured by sandwich ELISA as previously described (15). In short, high-binding 96-well ELISA plates (Corning) were coated by either 75 µL/well of 2.5 μg/mL rabbit anti-citrullinated histone H3 antibody (Abcam) or rabbit anti-neutrophil elastase antibody (Abcam). After overnight incubation at 4°C, the plates were blocked by 100 μL/well of 1% BSA-PBS at room temperature for 1 h. Subsequently, mouse serum samples (1:100 diluted by 1% BSA-PBS) were applied to each well and incubated overnight at 4°C. Next, plates were washed by 200 μL/well PBS containing 0.05% Tween 20 (PBS-T) three times; they were incubated with mouse anti-ds DNA monoclonal antibody (100 μL/well, 1:100 diluted by 1% BSA-PBS; EMD Millipore) for 1 h at room temperature. The plates were washed three times with 200 μL/well PBS-T, and HRP-conjugated anti-mouse antibody (100 μL/well, 1:10,000 diluted by 1% BSA-PBS; Bio-Rad) was added to each well. Finally, the plates were washed five times with 200 μL/well PBS-T, and ELISA was visualized with tetramethylbenzidine and stop solution.

Mouse ear pinna was snap frozen in an OCT compound (Tissue-Tek). Frozen sections were blocked with 1% BSA-PBS for 1 h at room temperature. Sections were incubated overnight at 4°C with rabbit polyclonal anti-histone H3 (citrulline R2 + R8 + R17) antibody (Abcam), which was 1:500 diluted using 1% BSA-PBS. After washing five times, they were stained with 1:400 diluted Alexa Fluor 555-conjugated donkey anti-Rabbit IgG (H+L) Antibody (Thermo Fisher Scientific) for 1 h at room temperature. Slides were then counterstained with Hoechst (Thermo Fisher Scientific) for 10 min at room temperature, washed, mounted by Prolong Gold (Thermo Fisher Scientific), and visualized using a Leica DMI4000 B inverted microscope (Leica Microsystems).

Mouse kidney was snap-frozen in an OCT compound (Tissue-Tek). Frozen sections were fixed with cold acetone for 10 min at −20°C, washed, and blocked by 5% BSA-PBS at room temperature for 1 h. Subsequently, they were stained with Alexa Fluor 594-conjugated goat anti-mouse IgG (1:100 diluted by PBS; Thermo Fisher Scientific), FITC-conjugated goat anti-mouse C3 antibody (1:100 diluted by PBS; Immunology Consultants Laboratory), and Hoechst (1:1,000 diluted by PBS; Thermo Fisher Scientific) at room temperature for 1.5 h. After washing, they were mounted with Prolong Gold (Thermo Fisher Scientific) and visualized using a Leica DMI4000 B inverted microscope (Leica Microsystems).

For quantification of immune complex deposition, three photos per sample were taken to include four glomeruli. Integrated density was calculated by ImageJ 2.3.0. Average integrated density of three photos from each sample was compared.

Serum autoantibodies, total IgG, April, cell-free DNA, and DNase I activity were quantified by commercially available ELISAs or plate assay kits. Mouse serum anti-dsDNA, anti-RNP, and anti-SS-A/Ro antibodies were assessed by Mouse anti-dsDNA IgG-specific ELISA Kit, Mouse Anti-nRNP Igs (total (A+G+M)) ELISA Kit, and Mouse Anti-SSA/Ro60 Ig’s (total) ELISA Kit, respectively (all from Alpha Diagnostic). Total serum IgG concentrations were determined by IgG (Total) Mouse Uncoated ELISA Kit (Thermo Fisher Scientific). Serum April levels were tested by Mouse TNFSF13/APRIL ELISA Kit (LS Bio). Cell-free DNA in mouse peripheral blood was analyzed with the Circulating DNA Quantification Kit (Abcam). Serum DNase I activity was tested with the DNase I assay kit (Abcam).

Mouse urine samples were collected at euthanasia and frozen at −80°C until used. Urine albumin and creatinine were quantified by Albuwell M and the Creatinine Companion (both from Ethos Biosciences). The severity of proteinuria was assessed by dividing urine albumin level by urine creatinine level.

Splenocytes were harvested at euthanasia and immediately incubated with TruStain FcX (BioLegend) for 15 min at 4°C. Subsequently, fluorochrome-conjugated antibodies were added, and cells were incubated for 30 min in 4°C. Splenocytes were then washed with PBS and stained with LIVE/Dead Fixable Aqua (Thermo Fisher Scientific) for 30 min in the dark at room temperature, in accordance with the manufacturer’s instructions. After washing, they were promptly analyzed on a BD FACSCelesta flow cytometer (Becton, Dickinson & Company). Data were analyzed using FlowJo 10.8.1 (Becton, Dickinson & Company). Gating strategies were established based on previous reports (11, 16). The number of cells was counted using CountBright Plus Absolute Counting Beads (Thermo Fisher Scientific). A detailed information of antibodies and gating strategy used are provided in a Supplementary Excel File .

The immature neutrophil score was calculated as previously described (20, 21) with some modifications. In brief, RPKM values of seven genes were used for calculation (Mpo, Elane, Bpi, Ctsg, Prtn3, Camp, and Defa4). AZU1 was removed because this gene is not present in mice. The mean value of each gene in untreated WT was subtracted from each value of each gene in each sample. Each value was then divided by the standard deviation of each calculated gene by values in untreated WT. The immature neutrophil score of each sample was calculated by the addition of these numbers. The qPCR version of immature neutrophil score was calculated by the addition of geometric mean of relative expression level of each gene (fold change based on the ΔΔCt method).

Bone marrow neutrophils were isolated from IMQ-treated Bmal1^Mye−/− and WT using the method explained above (MACS-negative selection). The neutrophils were resuspended in RPMI with 10% fetal bovine serum (FBS) at 10 million neutrophils/mL and incubated for 16 h at 37°C with 5% CO₂. Then, culture supernatant was collected and stored in −80°C until they were used.

IgD⁻ Lin (TER-119, F4/80, CD8a, CD4)⁻ CD138⁺ plasma cells were sorted out using FACSAria Fusion (Becton, Dickinson & Company) from spleen cells of untreated WT mice. The gating strategy was decided by referencing a previous report (22). After the sorting, plasma cells were resuspended in RPMI with 10% FBS at 1 × 10⁵ cells/mL and mixed with SYTOX Green and Incucyte Nuclight Rapid Red (Sartorius). Then, 100 µL of the cell suspension was mixed with the same volume of culture supernatant of neutrophils and plated onto a 96-well plate (Corning) with or without 50 μg/mL of anti-mouse April antibody (Apry-1-1, AdipoGen Life Sciences), which selectively inhibits the binding of April to its receptors BCMA and TACI (23, 24). Cell death was quantified using the same method as the quantification of NETs explained above (12, 25).

Patients were recruited at the Clinical Center, NIH, Bethesda, MD, with written informed consent. The study is in compliance with the Helsinki Declaration and was approved by the institutional review board of NIAMS (NIH 94-AR-0066). SLE patients fulfilled the ACR/EULAR 2019 classification criteria for SLE. All blood samples were harvested in the morning (7–11 am). Samples were immediately processed. Neutrophils and peripheral blood mononuclear cells (PBMCs) were isolated by Ficoll-Paque (GE) density gradient and dextran (Alfa Aesar) sedimentation (25). The expression of each gene was normalized to GAPDH, and fold change was calculated using the ΔΔCt method. SLE disease activity index (SLEDAI), anti-dsDNA antibody, and complement C3 and C4 were quantified on the same day when each sample was taken. The positivity of anti-RNP, Sm, SS-A/Ro, and SS-B/La antibodies was based on the medical record review. Plasma APRIL levels were tested by Human APRIL/TNFSF13 DuoSet ELISA kit (R&D Systems).

Statistical analysis was done, and graphs were drawn by GraphPad Prism version 9.3.1. To compare two groups, unpaired t-test and Mann–Whitney test were used for parametric and non-parametric parameters, respectively. Correlation was evaluated with Pearson correlation or Spearman’s rank correlation coefficient. Log-rank test with Kaplan–Meier curve was used for mouse survival analysis. Repeated measures two-way ANOVA was used for plasma cell survival analysis. Adjustment for multiple comparison was not performed except for plasma cell survival assay. In any statistical test, p<0.05 was considered significant.

After 6 weeks of epicutaneous IMQ cream, WT and Bmal1^Mye−/− mice displayed similar induction of splenomegaly, leukocytosis, anemia, thrombocytopenia, dermatitis, and mild glomerulonephritis as previously described for this model ( Figures 1A–D ). As in other studies using mice in C57BL6-background (10, 11), significant proteinuria was not observed ( Figure 1E ) and mortality rates were similar between groups during the treatment ( Figure 1F ).

When assessing histology of the IMQ-treated local skin lesions, there was increased neutrophilic infiltration in the Bmal1^Mye−/− mice compared with WT ( Figure 1G ), whereas there were similar degrees of skin inflammation ( Figure 1C ) and tissue chemokine levels ( Figure 1H ). This supports that Bmal1 regulates migratory capacity of myeloid cells in inflammatory milieus (3). Furthermore, IMQ-treated Bmal1^Mye−/− mice displayed significantly higher levels of serum anti-dsDNA ( Figure 1I ) whereas serum total IgG ( Figure 1J ) and other autoantibodies (anti-RNP and anti-SS-A/Ro, Supplementary Figure 1 ) were not significantly different compared with IMQ-treated WT. Moreover, increased renal immune complex deposition was observed in IMQ-treated Bmal1^Mye−/− compared with WT ( Figures 2A–C ). In contrast, expression levels of tissue type I interferon-stimulated genes (ISGs) did not significantly differ between both groups of mice ( Figures 2D, E ).

Next, we performed flow cytometry analysis of splenocytes to assess if Bmal1 regulates immune cell subsets. While no significant differences were found between WT and Bmal1^Mye−/− in percentages of neutrophils, monocytes, conventional dendritic cells (cDCs), plasmacytoid DCs (pDCs), and CD4⁺ and CD8⁺ T cells ( Supplementary Figures 2A–F ), significant increases in the percentage of CD19⁺ B cells was observed in IMQ-treated Bmal1^Mye−/− mice when compared with IMQ-treated WT mice ( Supplementary Figure 2G ), alongside a non-significant trend for increased splenic CD19⁻CD138⁺ plasma cells in Bmal1^Mye−/− ( Supplementary Figure 2H ). There were no significant differences in follicular T helper cells (Tfh) and germinal center B cells ( Supplementary Figures 2I, J ). Furthermore, there were no differences in percentages of naïve CD19⁺ B cells, whereas marginal zone B cells were significantly lower in Bmal1^Mye−/− ( Supplementary Figures 2K, L ). Since percentages of each cell fraction may not always reflect the actual expansion in cell numbers, we checked the total number of splenocytes, and found that it did not show any differences between WT and Bmal1^Mye−/− mice ( Supplementary Figure 3 ).

In summary, although the absence of Bmal1 in myeloid cells did not dramatically modify murine lupus symptoms during the 6 weeks of treatment, it led to increase anti-dsDNA antibody levels, renal immune complex deposition, and enhanced dermal tissue neutrophil infiltration.

We considered that increased autoantigen generation by neutrophils in Bmal1^Mye−/− would contribute to increased autoantibody and immune complex formation. As higher numbers of neutrophils infiltrate, the IMQ-treated local skin lesions of Bmal1^Mye−/− and this gene are known to modulate neutrophil biology (26), we assessed NET formation in the IMQ-treated local skin lesions. However, no significant differences in skin lesion NETs were detected by immunofluorescence ( Figure 3A ). While bone marrow neutrophils from IMQ-treated mice have enhanced spontaneous NET generation, no difference was observed between WT and Bmal1^Mye−/− ( Figures 3B, C ). Responsiveness to PMA or to the A23187 calcium ionophore, both of which are inducers of NET formation, was not affected, although a non-statistically significant trend was seen for PMA-induced NETs ( Figures 3D, E ). Consistent with these findings, serum cell-free DNA and serum NET remnant levels, which can be used as surrogate markers of cell death and NET formation in vivo, were also similar between both groups of mice ( Figures 3F–H ). Since decreased serum DNase I activity has been reported and may contribute to increase autoantigen burden in human SLE patients (27), we tested serum DNase I activity in the mice. There was a statistically significant increase in DNase I activity in IMQ-treated Bmal1^Mye−/− ( Figure 3I ), which may reduce autoantigen half-life in Bmal1^Mye−/− . These results do not support that enhanced autoantibody and immune complex generation in Bmal1^Mye−/− are the result of increased autoantigen generation.

As it has been reported that lupus bone marrow neutrophils can affect B-cell development through enhanced type I IFN, BAFF, and APRIL synthesis (2), we evaluated the expression of mRNAs encoding for these molecules in mouse bone marrow neutrophils. We found that April expression was significantly higher in IMQ-treated Bmal1^Mye−/− neutrophils whereas there were no differences in Baff expression ( Figures 4A, B ). There was a non-significant trend for increased Ifna4 in Bmal1^Mye−/− , independent of IMQ treatment ( Supplementary Figure 4 ). The neutrophil April expression level in the bone marrow was significantly associated with serum anti-dsDNA antibody levels in IMQ-treated mice ( Figure 4C ). Bulk RNA sequencing (RNA-seq) analysis of bone marrow neutrophils showed that the number of differentially expressed genes was small when the mice were not treated ( Supplementary Figure 5A ) but was enhanced when the mice were treated with IMQ ( Supplementary Figure 5B ). Pathway enrichment analysis revealed that genes associated with mitosis were upregulated in bone marrow neutrophils from IMQ-treated Bmal1^Mye−/− ( Figure 4D ). The expression level of genes associated with neutrophil differentiation/maturation status (19–21) suggested higher levels of immature neutrophils in the IMQ-treated Bmal1^Mye−/− mice ( Figures 4E, F ; Supplementary Figure 6 ). These observations are consistent with previous findings that reported the expression of APRIL being higher in immature myeloid cells (28). To confirm findings at the protein level, we attempted to quantify serum April levels, but they were under the detection level (data not shown). Bone marrow immunohistochemistry (IHC) showed that more April was detected in Bmal1^Mye−/− bone marrow following IMQ treatment when compared with WT ( Figure 5A ). This increased April expression in immature neutrophils was observed at the protein level ( Figures 5B, C ). In contrast, splenic mRNA expression levels of Baff and April were similar between IMQ-treated WT and Bmal1^Mye−/− ( Supplementary Figure 7 ). The maturation status of bone marrow neutrophils in the absence of IMQ did not differ between WT and Bmal1^Mye−/− ( Supplementary Figure 8 ). Furthermore, the expression levels of April in bone marrow monocytes from IMQ-treated WT and Bmal1^Mye−/− were comparable ( Supplementary Figure 9 ).

We then incubated splenic plasma cells with supernatants from neutrophil cultures obtained from bone marrows from IMQ-treated mice. There was a non-statistically significant trend for IMQ-treated Bmal1^Mye−/− neutrophil supernatants to increase the plasma cell survivals when compared with IMQ-treated WT neutrophil supernatants. A blocking anti-April monoclonal antibody reduced the survival of plasma cells in both IMQ-treated WT and KO mice ( Supplementary Figure 10 ).

Taken together, these observations suggest that Bmal1 regulates neutrophil maturation in the bone marrow in murine models of lupus. Immature neutrophils increase in IMQ-treated Bmal1^Mye−/− mice and produce more April. As such, it is possible that Bmal1 negatively regulates plasma cell dysregulation, at least in part, by controlling neutrophil differentiation and maturation and April synthesis in inflammatory conditions.

We assessed if lupus clinical parameters correlate with BMAL1 expression in peripheral blood neutrophils obtained from SLE patients. The patients’ demographics and clinical characteristics are shown in Supplementary Table 2 . SLE neutrophils showed variability in BMAL1 expression levels, which negatively associated with serum anti-dsDNA levels and positively with C3 levels, whereas SLEDAI and C4 showed no association ( Figures 6A–D ). When BMAL1 expression levels were compared between patients that were negative or positive for other autoantibodies, they were significantly lower in anti-Sm-positive patients and tended to be lower in anti-RNP-positive patients, whereas there was no association with anti-SS-A/Ro- and anti-SS-B/La-positive patients ( Supplementary Figure 11 ). We next assessed the association of BMAL1 expression with corticosteroid dose, but no correlation was detected ( Figure 6E ; mean daily prednisone dose in this cohort was 4.96 ± 3.1 mg/day, mean + SEM). BMAL1 expression levels correlated with neutrophils and PBMCs ( Figure 6F ), suggesting that the circadian clock is well-synchronized in peripheral white blood cells. Plasma levels of APRIL protein and expression levels of APRIL mRNA did not correlate with BMAL1 expression levels in human neutrophils ( Supplementary Figures 12A, B ). This may be because peripheral blood neutrophils are mostly mature. Furthermore, plasma APRIL levels did not correlate with anti-dsDNA levels ( Supplementary Figure 12C ). When SLE and healthy donor APRIL levels were compared, they were higher in SLE, but the difference was not statistically significant ( Supplementary Figure 12D ). Overall, the data in lupus patients is consistent with findings in murine lupus suggesting that BMAL1 may negatively regulate autoantibody production.