Effect of in vivo Hydroxychloroquine and ex vivo Anti-BDCA2 mAb Treatment on pDC IFNα Production From Patients Affected With Cutaneous Lupus Erythematosus

Abstract

Objective: Plasmacytoid dendritic cells (pDCs) are a major source of Type-I Interferon (IFN-I), a key driver in cutaneous lupus erythematosus (CLE). Currently evaluated in Phase II clinical trial, 24F4A (BIIB059) is an antibody targeting BDCA2, an inhibitory receptor expressed on pDCs. Given that Hydroxychloroquine (HCQ), a widely-used CLE therapy, and 24F4A are both able to inhibit pDC-derived IFN-I production; this study aimed to determine whether 24F4A would show an additional inhibitory effect on pDC response after ex vivo or in vivo treatment with HCQ. Methods: The effect of 24F4A on pDC-derived IFNα was measured from peripheral blood mononuclear cells (PBMC) either from healthy donors in presence or absence of HCQ or from CLE patients clinically exposed to various levels of HCQ. TLR7, TLR7/8, and TLR9 agonists (ssRNA, R848, and CpG-A) were used for pDC stimulation. Results: PDCs were the only producers of IFNα in response to CpG-A, R848, and ssRNA stimulation in PBMC cultures. CLE patients with higher levels of blood HCQ showed lower ex vivo pDC responses to CpG-A, but not R848 or ssRNA. In contrast, 24F4A reduced the amount of IFNα produced by pDCs from CLE patients in response to all TLR agonists, irrespective of the blood HCQ level. Conclusion: Our findings reveal that clinically-relevant HCQ concentrations partially inhibit the pDC response to TLR9 and weakly affect the response to TLR7/8 stimulation. 24F4A robustly inhibits pDC responses even in the presence of HCQ, highlighting its unique potential to disrupt pDC disease relevant biology, which could provide additional therapeutic benefit for CLE patients.

Keywords: BDCA2 blood dendritic cell antigen 2; BIIB059; cutaneous lupus erythematosus (CLE); hydroxychloroquine; interferon; plasmacytoid dendritic cell; systemic lupus erythematosus (SLE); toll like receptor (TLR).

Figure 1

Effect of HCQ and BIIB059 on IFNα release from human whole blood stimulated with CpG-A, R848, or ssRNA. Whole blood from human healthy donors was stimulated or not with CpG-A (10 μM), R848 (1 μM), or ssRNA (4 μg/ml) complexed with pARG in presence or absence of HCQ (550 ng/ml), isotype control mAb (10 μg/ml), 24F4A (10 μg/ml) or a combination of both HCQ and 24F4A. Secreted IFNα was measured 18 h after stimulation in the serum using ELISA. Same color and/or shape-coded data points represent data obtained from a given donor (n = at least 6 donors). Statistical tests were performed using paired one-way Anova (ns, non-significant p ≥ 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Figure 2

HCQ-mediated inhibition of IFNα production from pDCs in response to CpG-A and ssRNA, but not to R848, can be detected after isolation of PBMC from whole blood pre-treated with HCQ. Whole blood samples from healthy donors were treated with HCQ (1000 ng/ml) or not for 1 h prior to PBMC isolation and then stimulated with CpG-A (10 μM), R848 (1 μM) or ssRNA (4 μg/ml) for 6h. IFNα was analyzed by intracellular cytokine staining. (A) Representative dot plots of IFNα+ cells within a BDCA4+ and CD123+ gate. (B) Percentages of IFNα-producing pDCs detected as shown in A) after CpG-A stimulation (n = 5 donors), R848 stimulation (n = 3 donors), or ssRNA (n = 5 donors) from PBMC isolated from whole blood pre-treated with HCQ or not. Statistical significance was assessed using two-tailed paired Student's t-test (*p < 0.05, ****p < 0.0001).

Figure 3

The percentage of IFNα-producing pDCs upon stimulation with CpG-A, but not R848 or ssRNA, is negatively correlated with whole blood HCQ concentrations from CLE patients. (A) Association of the percentage of IFNα-producing pDCs identified by flow cytometry from PBMC after CpG-A (n = 22 donors), R848 (n = 22 donors), or ssRNA (n = 24 donors) stimulations with the HCQ concentrations in whole blood from CLE patients. Patients with non-detectable blood HCQ levels are shown at 0 ng/ml blood HCQ (B) Association of the change (ratio) between 3-month apart clinical visits in the percentage of IFNα-producing pDCs identified by flow cytometry from PBMC after CpG-A (n = 21 donors), R848 (n = 21 donors), or ssRNA (n = 22 donors) stimulation with the change in the blood HCQ concentration of CLE patients. Data points from patients with increase in blood HCQ concentration >600 ng/ml between the two visits are highlighted in red. Data points from patients with decrease in blood HCQ concentration >500 ng/ml between two visits are highlighted in green. Statistical association was assessed using Pearson's (P) and Spearman's rank (S) correlations.

Figure 4

24F4A further reduces pDC IFNα production after CpG-A, R848 or ssRNA stimulations of PBMC isolated from CLE patients regardless blood HCQ levels. (A) Concentrations of HCQ in whole blood from the CLE patient cohort studied (n = 30). (B) Representative dot plots of IFNα+ cells within a BDCA4+ and CD123+ gate after CpG-A, R848 or ssRNA stimulation with or without pre-treatment with 24F4A (10 μg/ml 30 min) from CLE patient without detectable blood HCQ (top panel) or a CLE patient with a blood HCQ concentration of more than 500ng/ml (bottom panel). (C) Effect of 24F4A on the percentage of IFNα-producing pDCs induced by CpG-A, R848 and ssRNA stimulations and detected by flow cytometry in PBMC from CLE patients without detectable level of blood HCQ (n ≥ 6 donors), with blood HCQ concentrations lower than 500 ng/ml (n ≥ 8 donors) or >500ng/ml (n ≥ 6 donors). Statistical significance was assessed with a two-tailed paired Student's t-test using log2-transformed values (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Figure 5

24F4A further reduces pDC IFNα production after CpG-A, R848 or ssRNA stimulations of PBMC isolated from CLE patients with blood IFN-high signature or concomitant SLE diagnostic. (A) Distribution of the blood IFN signature scores in healthy individuals (n = 52) and CLE patients (n = 30). (B) Comparison of SLEDAI2-K disease scores (left) or CLASI skin scores (right) between CLE patients with blood IFN-low and blood IFN-high signature scores. Statistical significance was assessed using Mann-Whitney test (*p < 0.05). (C) Effect of 24F4A on the percentage of IFNα-producing pDCs induced by CpG-A, R848 and ssRNA stimulations and detected by flow cytometry in PBMC from CLE patients with blood IFN-low (n = 4 donors) or IFN-high signature (n ≥ 17 donors). Statistical significance was assessed with a two-tailed paired Student's t-test using log2-transformed values (*p < 0.05, **p < 0.01, ****p < 0.0001). (D) Effect of 24F4A on the percentage of IFNα-producing pDCs induced by CpG-A, R848 and ssRNA stimulations and detected by flow cytometry in PBMC from CLE patients with or without concomitant SLE diagnostic (n ≥ 7 donors and n ≥ 14 donors, respectively). Statistical significance was assessed with a two-tailed paired Student's t-test using log2-transformed values (**p < 0.01, ***p < 0.001, ****p < 0.0001).