Our prior study revealed that CB modulates IFN-α production by virus-activated pDC (14) and inflammatory factor release from monocytes (43). Given the importance of tonsillar tissue in immunological responses, and its role as the primary site of pathogen exposure and immune activation, we next tested CB activity in a tonsillar mononuclear cells (TMCs) model. The experimental setup and key tests conducted are summarized in a schematic ( Figure 1A ). We exposed TMCs from seven healthy donors to increasing concentrations of histamine (HA) or CB, followed by stimulation with R848. Both HA and CB prevented interferon (IFN) production, with CB being more potent than HA ( Figure 1B ). Importantly, HA or CB treatments did not significantly affect cell viability, except for TMCs treated with the highest CB concentration (50 µM) ( Figure 1C ). CB significantly reduced the production of several inflammatory cytokines, including IFN-α2, IFN-β, IFNλ1, IFN-λ2/3, and IFN-γ ( Figure 1D ) without toxicity ( Figures 1E, F ). To investigate CB’s impact on interferons whole blood samples from 7 healthy donors were treated with a range of CB doses followed by R848 stimulation. The levels of IFN-α within culture supernatants were evaluated using Single Molecule Array (SIMOA) ( Figure 1G ). This analysis revealed CB’s potent prevention of IFN-α production with an IC₅₀ of 6 µM. To this end, pDCs were selected from whole blood in flow cytometry through marking cell specific proteins with fluorophore coupled antibodies followed by the displayed gating strategy ( Figure 1H ). Subsequently, IFN-α production was assessed in R848 pre-treated pDCs upon exposure to CB ( Figures 1I, J ). This approach distinctly demonstrated CB’s capacity to prevent IFN-α production within the selected pDCs sourced from whole blood of healthy donors.

To investigate the effects of CB on the interferon production, peripheral blood mononuclear cells (PBMCs) from three healthy donors were exposed to increasing concentrations of CB, followed by stimulation with R848. The experimental setup and key tests conducted are summarized in a schematic ( Figure 2A ). The subsequent production of interferons in the supernatants was quantified using a luminescence-based assay, employing the twINNE reporter cell line (44). This cell line, engineered to express the interferon-stimulated response element (ISRE) linked to luciferase, facilitates the direct measurement of interferon activity ( Figure 2B ). CB treatment showed negligible effects on PBMC viability. However, it exhibited a potent prevention of R848-mediated interferon production. We then investigated the effects of CB on PBMCs when incubated either simultaneously with R848 or after the stimulation (1h). This approach allowed us to assess how the timing of CB administration influences its effectiveness in modulating the immune response triggered by R848. We observed that when PBMCs were exposed to CB and R848 simultaneously, CB reduced ISRE stimulation, albeit less effectively than with a 1-hour pre-incubation. In contrast, when cells were first activated by R848 and then treated with CB one hour later, no reduction in ISRE activation was observed ( Figure 2C ). Further analysis explored CB’s impact on the mRNA levels of several inflammatory cytokines, interferons, and interferon-stimulated genes (ISGs) in PBMCs, assessed via quantitative PCR (q-PCR) ( Figures 2D, E ). Our findings unveil that CB exerts prevention of transcription for key inflammatory markers, including IFN-α, IFN-β, IFN-γ, and several interferon-stimulated genes (ISGs), underscoring its broad immunomodulatory effects. To specifically address CB’s effects on pDCs, we isolated these cells from PBMCs of healthy donors and treated them with CB at 5 and 20 µM concentrations, followed by R848 activation ( Figures 2F, G ). The production of thirteen inflammatory cytokines was measured in the cell culture supernatants using the LEGENDplex assay, revealing that CB significantly prevents the production of various cytokines, notably IFN-α2 and IFN-γ.

Since the production of IFN-α is mediated by the phosphorylated form of IRF7, we aimed to investigate the state of IRF7 in the presence of CB. pDCs were isolated from PBMCs, treated with CB, and stimulated with R848 for 30 minutes to detect the early induction of pIRF7 ( Figures 3A, B ). CB significantly prevents the early phosphorylation of IRF7 in pDCs. To examine the effect of CB on the later induction of pIRF7, PBMCs were treated with CB and subsequently stimulated with R848 for 24 hours. This experiment demonstrated that CB effectively blocks the late phosphorylation of IRF7 in pDCs isolated from PBMCs ( Figure 3C ).

We evaluated the ability of CB to modulate immune responses in cells from SLE patients, aiming to assess CB’s efficacy in a disease-relevant context. The experimental setup and key tests conducted are summarized in a schematic ( Figure 4A ). PBMCs from four SLE patients were isolated and then treated with CB and subsequently stimulated with R848. The production of 13 inflammatory cytokines was quantified in culture supernatant using the LEGENDplex assay ( Figure 4B ). This analysis revealed that CB effectively prevents the production of inflammatory cytokines, particularly IFNs, in PBMCs from SLE patients, highlighting its potential as a modulator of disease-associated inflammation. Immune cells from the bone marrow of a patient experiencing an active SLE flare were subjected to CB along with HCQ, the standard SLE treatment. Following R848 stimulation, the level of IFNα was quantified in the culture supernatants using Single Molecule Array (SIMOA) technology ( Figure 4C ). The results demonstrated that CB prevented IFN-α production similarly to HCQ in the bone marrow-derived immune cells of an SLE patient in flare, without inducing toxicity ( Figure 4D ). To further characterize the immune cells isolated from the bone marrow of the SLE patient, phenotyping was performed using flow cytometry ( Figure 4E ). Bone marrow sample contained 0.3% pDCs and within these selected pDCs, cells double-positive for IFNα and TNF-α production were detected under conditions of R848 stimulation, an effect reversed by CB and HCQ treatment ( Figure 4F ).

The current study further explores the mechanism of action of CB, with particular emphasis on its interaction with CXCR4. Utilizing an in silico docking approach, we compared the binding sites of the well described minor pocket ligand IT1t (45) and CB. The analysis revealed distinct binding preferences. IT1t binds into the minor pocket with an overlap with the major pocket targeted by other antagonists such as AMD3100 (plerixafor), an FDA/EMA-approved CXCR4 antagonist ( Figure 5A ). In contrast, CB targets the minor pocket deeper than IT1t, and did not show overlapping with the major pocket ( Figure 5B ), suggesting a unique mode of interaction with the receptor. To further elucidate the functional implications of these binding characteristics, we used a NanoBRET binding assay, using CXCL12-AF647 as a fluorescent probe, to assess the capacity of IT1t and CB to inhibit the binding of the endogenous ligand of CXCR4, CXCL12 ( Figure 5C ). IT1t and AMD3100 effectively inhibit binding of CXCL12-AF647 with pKi values of 7.64 ± 0.12 and 7.59 ± 0.14, respectively. Notably, CB blocks CXCL12-AF647 binding with a 1000-fold lower potency (pKi = 4.73 ± 0.25). However, this indicates that CB very weakly antagonizes CXCL12 while targeting CXCR4. We investigated whether CB has antagonistic activity towards CXCL12 using a G protein activation assay ( Figure 5D ). CXCL12 activates Gαi2 with a pEC₅₀ of 8.73 ± 0.29, consistent with previous findings (46, 47). This activation of CXCL12 can be reversed by the simultaneous addition of IT1t with a pIC₅₀ of 6.88 ± 0.26. On the other hand, CB appears to have only a small influence on CXCL12 G protein activation, observable primarily at high concentrations (100 µM). To further assess whether the anti-IFN properties of CB are CXCR4 dependent, PBMCs were incubated with the known CXCR4 antagonists AMD3100 ( Figure 5E ) or MSX-130 ( Figure 5F ) prior to the CB/R848 treatment. AMD3100 and MSX-130 negate CB’s anti-IFN effect in a dose-dependent manner. To further investigate whether CB’s anti-IFN activity is CXCR4-dependent and the role of IRF7, the expression of CXCR4 was reduced via the transfection of siRNA in isolated pDCs ( Figure 5G ). On pDCs treated with a control siRNA (CTR), CB prevented the mRNA level encoding for IFN-α ( Figure 5H ) and IRF7 ( Figure 5I ), whereas on pDCs treated with a CXCR4 siRNA, CB loses its ability to modulate the response ( Figures 1I, J ). All these data show that CB directly interacts with CXCR4 and mediates its anti-IFN properties through blocking IRF7 phosphorylation potentially by reducing the amount of total IRF7.

To evaluate CB’s therapeutic potential for SLE, we used the pristane-induced SLE mouse model, which closely replicates the human condition. The experimental setup and key tests conducted are summarized in a schematic ( Figure 6A ). Indeed, pristane-induced SLE aligns closely with the criteria established by the Systemic Lupus Erythematosus International Collaborating Clinics (SLICC) for classifying human lupus. This alignment is evident through the existence of an IFN signature and the presence of anti-dsDNA antibodies (35).

The mice were treated over a 10-week period with PBS (as a placebo), the reference SLE treatment prednisolone at 15 mg/kg, or CB at 3, 10, or 30 mg/kg (n=8 mice per condition). To assess the general condition of the mice, daily weight measurements were recorded throughout the treatment period ( Figure 6B ). Notably, the diverse treatments had no apparent impact on the mice’s weight. Additionally, all treated mice exhibited normal behaviors and showed no signs of distress or illness throughout the study, indicating that the treatments were well tolerated. Key inflammatory cytokines, including IL-17 ( Figure 6C ), IL-1β ( Figure 6D ), TNF-α ( Figure 6E ), and TRAIL ( Figure 6F ), were measured in the blood at both 4- and 10-weeks post-treatment initiation. CB at all tested dosages reduced the levels of these cytokines.

Docking was performed using FlexX in LeadIT Version: 2.3.2 using default settings. The extended binding pocket in receptor 3ODU was defined around ligand IT1t by selecting residues within a distance of 16A. Ligand preparation was done using Datawarrior 5.5.0. Docking was performed with 1000 maximum number of solutions per iteration and 200 maximum number of solutions per fragmentation. The 10 best scored poses were kept after docking. Protein overlay and poses visualization was performed using Discovery Studio Visualizer v21.1.0.20298.

Blood samples were collected from healthy donors through the French Blood Establishment (“Etablissement Français du Sang”) under the authorization number 07/CABANEL/106, in Paris, France. The procurement of materials from SLE patients was approved by the French Ethical Committee (“Comité de Protection des Personnes”), with the approval number EudraCT: 2018-A01358-47. All experimental protocols involving human blood were in strict compliance with the European Union regulations and the Helsinki Declaration. Consent was duly obtained from every participant, encompassing healthy subjects as well as SLE patients. A summary of the clinical information of the SLE patients is presented in Table 1 . The in vitro studies utilized human mononuclear cells from peripheral blood or synovial fluid, isolated through centrifugation over a density gradient medium (STEMCELL Technologies). Isolation of human pDCs was achieved by negative selection with the EasySep Human Plasmacytoid DC enrichment kit (STEMCELL Technologies). Peripheral blood mononuclear cells (PBMC), and pDCs were incubated in RPMI 1640 medium (Invitrogen, Gaithersburg, MD), enriched with 10% heat-inactivated fetal bovine serum and 1mM glutamine (Hyclone, Logan, UT).

The procurement of tonsils received approval from the Ethical Committee (“Comité de Protection des Personnes”), with the identification number ID-RCB/EUDRACT: 2018-A01358-47. Tonsils removed from children diagnosed with obstructive sleep apnea during partial tonsillectomy procedures were processed. These tissues were preserved in 1× phosphate-buffered saline (PBS) for a duration not exceeding two hours, subsequently diced into smaller fragments, and mechanically disrupted with a glass pestle against a 60-mesh steel grid. The resultant Tonsil Mononuclear Cells (TMCs) were then strained through a 70-μm filter and rinsed with PBS. A Ficoll gradient centrifugation was employed to eliminate remaining epithelial cells and debris. The isolated TMCs were maintained in R10 medium. The sample donors were cataloged sequentially from numbers 1 to 7.

PBMCs were initially plated at a density of 2×10⁶ cells per milliliter, and pDCs derived from PBMCs were plated at 1×10⁶ cells per milliliter. Before undergoing a 16-hour stimulation with R848 at a dose of 5 µg/mL or another indicated concentration, cells received a 1-hour pretreatment with AMD3100 (Sigma-Aldrich) or MSX-130 (MedChemExpress), and/or CB (Sigma-Aldrich) at 20 µM or another specified concentration. Following this, cells were analyzed using flow cytometry, or supernatants were collected for cytokine detection. Brefeldin A (BFA) was introduced to the cultures 30 minutes post-stimulation and maintained for 5 hours to facilitate intracellular staining.

Cells were first washed with PBS and then treated with Zombie Aqua viability dye (Biolegend) for 30 minutes at 37°C. After another wash, they were resuspended in PBS containing 2% FCS and 2mM EDTA. For surface marker identification, cells were stained with anti-CD14, anti-CD19, anti-CD56, anti-CD3, anti-CD14 anti-CD123 and anti-BDCA4 antibodies (clone REA) from Miltenyi Biotec at a dilution of 1:100. This involved fixing the cells with 250 µL of for 20 minutes at room temperature, followed by washing and staining with 100 µL of Inside Perm solution (Miltenyi) mixed with anti-IFN-α, anti-TNF-α and/or anti-pIRF7 antibodies at a dilution of 1:50 for 30 minutes at room temperature. Data acquisition was performed on a Canto II flow cytometer and analyzed with Diva software (BD Biosciences, San Jose, CA) and FlowJo software (Treestar, Ashland, OR).

In the CXCR4 knock-down study, pDCs were plated at a concentration of 10⁵ cells per 100 µl in 96-well plates and incubated at 37°C. Two types of siRNA were prepared: control siRNA (Qiagen) and CXCR4-specific siRNA (SMARTPool, Dharmacon), each diluted in DOTAP (1,2-dioleoyl-3-trimethylammonium-propane; Roche Applied Sciences). This mixture was gently stirred and left to incubate at room temperature for 15 minutes. After this incubation, the mixture was introduced to the cells to reach a final siRNA concentration of 160 nM. Following this addition, the cells were incubated for an additional 24 hours at 37°C before proceeding with further treatment and stimulation protocols.

Cellular RNAs were extracted using the RNeasy Mini kit (Qiagen) following the manufacturer′s instructions. RNA concentration and purity were evaluated by spectrophotometry (NanoDrop 2000c, Thermo Fisher Scientific). A maximum of 500 ng of RNA were reverse transcribed with both oligo dT and random primers using a PrimeScript RT Reagent Kit (Perfect Real Time, Takara Bio Inc.) in a 10 µL reaction. Real-time PCR reactions were performed in duplicate using Takyon ROX SYBR MasterMix blue dTTP (Eurogentec) on an Applied Biosystems QuantStudio 5 (Thermo Fisher Scientific). Transcripts were quantified with the following program: 3 min at 95°C followed by 40 cycles of 15 s at 95°C, 20 s at 60°C and 20 s at 72°C.

Primers used for quantification of transcripts by real time quantitative PCR are the following:

To assess cytokine production, supernatants were analyzed using the LEGENDplex Antivirus Human Panel bead assay (Biolegend, San Diego, USA), executed in accordance with the instructions provided by the manufacturer.

To quantify the release of active IFN from TMCs, we employed a biological assay involving a stable cell line (STING37) in which a luciferase reporter gene is controlled by five IFN-stimulated response elements. Initially, supernatants from TMCs were collected following a 24-hour stimulation and stored at -20°C. These TMCs supernatants were subsequently distributed into the culture wells of a 96-well plate, with each well containing 35,000 STING37 cells. After an additional 24-hour incubation, luciferase activity was assessed by introducing 50 μl of Bright-Glo reagents (Promega) to the culture wells and measuring bioluminescence using a luminometer.

A Simoa digital ELISA designed for the specific detection of IFN-α2 was established using the Quanterix Homebrew Assay (74). The BMS216C (eBioscience) antibody clone was used as a capture antibody after coating on paramagnetic beads (0.3 mg/mL), and the BMS216BK already biotinylated antibody clone was used as the detector at a concentration of 0.3ug/mL. The SBG revelation enzyme concentration was 150 pM. Recombinant IFN-α2c was used as calibrator. The limit of detection was determined by calculating the mean value of all blank runs plus three standard deviations (SDs), resulting in a limit of detection of 0.23 fg/mL.

NanoLuc binding assays were slightly modified from (White et al.) (75). In brief, HEK293T cells were transiently transfected with 100 ng/well Nanoluc-CXCR4 and seeded at 10,000 cells/well into poly-D-lysine coated, white flat bottom 96-well plates 48h before the experiments and incubated at 37°C/5% CO₂. On the day of the experiment, cells were washed once with HBSS + 0.1% BSA before unlabelled compounds or buffer and 10 µM furimazine were added for 5 min and read twice on a Varioskan Lux using a 460 nm (80 nm bandpass) and a >610 nm (longpass) filter. Subsequently CXCL12-AF647 was added, and luminescence was monitored in the same manner for further 36 min every 3 min. The BRET ratio was calculated by dividing the 610 nm emission by the 460 nm emission.

G protein activation assays are performed as described in (Schihada et al., 2021) (76). HEK293T cells were transiently transfected with 50 ng/well CXCR4 WT and 50 ng/well G_i2 construct and seeded at 30,000 cells/well into poly-D-lysine coated, white flat bottom 96-well plates 48h before the experiments and incubated at 37°C/5% CO₂. On the day of the experiment, cells were washed once with HBSS + 0.1% BSA buffer before IT1t or CB and 10 µM furimazine were added for 5 min and the plate was read three times on a Varioskan Lux using a 460 nm (80 nm bandpass) and a 530 nm (30 nm bandpass) filter. Subsequently CXCL12 was added, and luminescence was monitored in the same manner for further 30 min every 3 min. The BRET ratio was calculated by dividing the 530 nm emission by the 460 nm emission.

Animal experiments were conducted under blind conditions by Washington Biotechnology, Inc. The animals were housed in a controlled environment with regulated temperature and humidity, following a 12-hour light/dark cycle, and had unrestricted access to food and water. The Animal Care and Use Committee at Washington Biotechnology, Inc. reviewed and approved all mouse experiments (IACUC no. 17-006). Animal welfare and experimental protocols strictly adhered to the principles outlined in the Guide for the Care and Use of Laboratory Animals and complied with ethical guidelines and regulations.

For the pristane injection procedure, each mouse in all experimental groups received an intraperitoneal injection of 0.5 ml of pristane (catalog no. P9622, Sigma-Aldrich, BioReagent). This was used to induce a systemic lupus erythematosus (SLE) model in female BALB/c mice (20-25g, ENVIGO, R# 3859), aged 8-10 weeks. The mice were quarantined, tagged, and monitored to ensure good health before the start of the study. The choice of female mice reflects the higher prevalence of SLE in females, making the model more representative of human disease.

Treatment with test compounds, including CB, began on the same day as the pristane injection, with the first dose administered one hour after the injection. Prednisolone (4.5 mg; catalog no. P4153, Sigma-Aldrich) was used as a positive control and was dissolved in 3 ml of PBS to create a solution with a concentration of 1.5 mg/ml. In group 2, all mice were administered daily oral gavage of prednisolone at a dose of 15 mg/kg. CB was dissolved in PBS and administered via intraperitoneal injection at daily doses of 3 mg/kg, 10 mg/kg, or 30 mg/kg, in groups 3 to 5.

Body weights and general health conditions were monitored throughout the 10-week study to assess any potential toxicity, with body weight loss kept under 20%. Blood samples were collected via the orbital sinus at multiple time points (weeks 4, 8, and 10) and processed to serum for analysis. Anti-dsDNA antibodies were measured as a primary readout, along with cytokine levels (TNF-α, IL-1β, IL-17, and TRAIL) by ELISA, to evaluate the efficacy of CB. Spleen and kidney samples were collected at the end of the study for further analysis.

For ELISA assays, the following kits were used in accordance with the manufacturer’s instructions: TNF-α ELISA kit (catalog no. MTA00B, R&D Systems), IL-1β ELISA kit (catalog no. MLB00C, R&D Systems), IL-17 ELISA kit (catalog no. M1700, R&D Systems), and TRAIL ELISA kit (catalog no. ELM-TRAIL-1, RayBiotech).

mg/kgmg/kgmg/kg