CSL362 potently and specifically depletes pDCs invitro and ablates SLE-immune complex-induced IFN responses

Abstract

Systemic lupus erythematosus (SLE) is an autoimmune disease with significant morbidity and mortality. Type I interferon (IFN) drives SLE pathology and plasmacytoid dendritic cells (pDCs) are potent producers of IFN; however, the specific effects of pDC depletion have not been demonstrated. We show CD123 was highly expressed on pDCs and the anti-CD123 antibody CSL362 potently depleted pDCs in vitro. CSL362 pre-treatment abrogated the induction of IFNα and IFN-induced gene transcription following stimulation with SLE patient-derived serum or immune complexes. RNA transcripts induced in pDCs by ex vivo stimulation with TLR ligands were reflected in gene expression profiles of SLE blood, and correlated with disease severity. TLR ligand-induced protein production by SLE patient peripheral mononuclear cells was abrogated by CSL362 pre-treatment including proteins over expressed in SLE patient serum. These findings implicate pDCs as key drivers in the cellular activation and production of soluble factors seen in SLE.

Keywords: Health sciences; Immunology; Pathophysiology.

Graphical abstract

Figure 1

Characterization of cell subsets, CD123 expression and CSL362 mediated cell depletion in SLE patient peripheral blood and matched healthy controls Whole blood from 32 healthy donors (HD) and 33 SLE patients was analyzed by flow cytometry to determine absolute counts and expression of CD123. (A) SLE patients have significantly fewer pDCs, basophils, mDCs, CD4 T-cells, CD8 T-cells, NKT-cells, naive B-cells, memory B-cells, monocytes and NK cells. (B) pDCs and basophils express the highest levels of CD123, mDCs express an intermediate amount while all other cell types have a mean receptor expression of less than 1500 ABC. (C) PBMC from 19 to 29 SLE donors and 31–34 healthy controls were cultured with CSL362 (1 μg/mL) for 20 h before the proportion of each cell type was determined by flow cytometry and normalized to no antibody control. (D) PBMC cultures from 6 healthy controls and 7 SLE patients were cultured with CSL362 (0.0005–3 μg/mL) for 20 h before the proportion of pDCs was determined by flow cytometry and normalized to no antibody control. ABC = Antibodies bound per cell as determined by BD quantibrite beads. Bars depict mean ± SEM (Mann-Whitney test ∗∗∗∗p ≤ 0.0001, ∗∗∗p ≤ 0.001, ∗∗p ≤ 0.01, ∗p ≤ 0.05 and n.s p > 0.05. See also Figures S1 and S2.

Figure 2

Effects of CSL362 on IFNα production and gene expression induced by TLR ligands, serum or purified IgG from SLE patients PBMC from 8 SLE donors were treated with 1 μg/mL CSL362 or isotype control for 20 h before the cells were resuspended in various stimuli (0.25 μM CpGc, 10 μg/mL LPS, HD Ig + NCL, SLE Ig + NCL, SLE sera + NCL or HD sera + NCL) and cultured for 24 h. (A) Supernatant was then collected and analyzed for IFNα production by ELISA. IFNα levels for each donor PBMC culture treated with isotype control (circles) or CSL362 (triangles) after stimulation are shown. (B) RNA was also extracted from the cultured cells and analyzed by RNA-seq. Graph shows the IFN gene score for each treatment (isotype control circles, CSL362 triangles) and stimulation determined as the average log2 fold change compared to control (HD PBMC unstimulated and treated with isotype control) of 11 IFN genes (IFI44L, IFIT1, IFIT3, IRF7, ISG15, MX1, MX2, OAS1, OAS2, SERPING1 and XAF1). (C) The genes differentially expressed (absolute log2FC >1, FDR<0.05) between PBMCs stimulated with SLE immune complexes (SLE Ig + NCL) in the presence of CSL362 or isotype control were subjected to pathway analysis with Ingenuity (IPA). The top 5 altered processes, ranked by p value of overlap, are shown. (D) Heatmap of the expression changes of TLR9, TLR4 and TLR3 genes in response to the different stimuli with and without CSL362 treatment. PI:C columns show the expression of n = 3 SLE donors only, stimulated with 10 μg/mL poly I:C (PI:C) all other columns show n = 8 SLE donors. NCL = Necrotic Cell Lysates used at 0.1 mg/mL. See also Figure S3.

Figure 3

Soluble factors transcribed by pDCs when stimulated with TLR7 and TLR9 agonists PBMC from healthy donors (n = 6) were cultured for 18 h with 0.5 μM CpGc (TLR9 agonist) or 0.5 μg/mL Imiquimod (IMQ; TLR7 agonist) before pDCs were isolated by FACS sorting and analyzed by RNA-seq. (A) Transcripts of secreted proteins (NABA secreted protein gene list⁴⁴) are shown. (B) The genes differentially expressed (absolute log2FC >2, FDR <0.05) between sorted pDCs stimulated with CpGc or unstimulated were subjected to pathway analysis with Ingenuity (IPA). The top 5 altered processes, selected based on p value of overlap, are shown.

Figure 4

SLE patients show upregulated expression of IFN genes and this can be correlated with other indicators of disease RNA from SLE samples (n = 30) and HD Samples (n = 29) was analyzed by RNA-seq. (A) A large proportion of the top 25 differentially expressed genes with the highest fold-change differences that were significantly different between SLE donors and matched HD were IFN inducible genes. (B) The genes differentially expressed (absolute log2FC >0.8, FDR <0.05) between whole blood from SLE patients and HD were subjected to pathway analysis with Ingenuity (IPA). The top 5 altered processes, selected based on p value of overlap, are shown. (C) An IFN gene score was determined based on the average log2 fold change of 11 gene transcripts (IFI44L, IFIT1, IFIT3, IRF7, ISG15, MX1, MX2, OAS1, OAS2, SERPING1 and XAF1) with SLE samples showing a significantly increased IFN gene score compared to HD. Bars depict mean ± SEM (Mann-Whitney test ∗∗∗∗ p value <0.0001). (D) SLEDAI-2K, IFN gene score, NK cell number and pDC number correlate with various other parameters (experimental data - IFN gene score, absolute NK cell number per microliter of whole blood, absolute pDC number per microliter of whole blood; clinical parameters - adjusted mean SLEDAI-2K, ratio of anti-dsDNA titer to the upper limit of normal, C3 protein, C4 protein, Erythrocyte sedimentation rate (ESR); and serum protein levels of - BAFF, ITAC/CXCL11 and M-CSF) only correlation coefficients that were significant p = <0.05 are shown, with values closest to 1 having strongly positive correlations and values closest to −1 having strongly negative correlations.

Figure 5

CSL362 reduces production of TLR9-induced proteins including proteins that are significantly elevated in SLE patients sera (A) PBMC from 9 healthy controls and 8 SLE patients were incubated in media alone (untreated; UT) or treated with 1 μg/mL CSL362 or isotype control for 20 h before wells were stimulated with 0.5 μM CpGc for 24 h. Supernatant was collected and assayed for various soluble proteins. Heatmap shows proteins that were significantly upregulated by CpGc stimulation when normalized to unstimulated control (fold change >2 and FDR <0.05). (B) Sera was collected from 33 SLE patients and 34 HD and analyzed for levels of 119 soluble proteins, 27 of which were found to be upregulated in SLE patient sera (see also Table S3). The IFN inducible proteins MCP-2/CCL8, IP-10/CXCL10, ITAC/CXCL11 and MIP-3β/CCL19 were among those significantly upregulated in the sera of SLE patients (Benjamini and Hockberg method, fold change >1.5 and adj p value <0.05).