Lupus IgA1 autoantibodies synergize with IgG to enhance pDC responses to RNA-containing immune complexes

Abstract

Autoantibodies to nuclear antigens are hallmarks of the autoimmune disease systemic lupus erythematosus (SLE) where they contribute to pathogenesis. However, there remains a gap in our knowledge regarding how different isotypes of autoantibodies contribute to disease, including the production of the critical type I interferon (IFN) cytokines by plasmacytoid dendritic cells (pDCs) in response to immune complexes (ICs). We focused on IgA, which is the second most prevalent isotype in serum, and along with IgG is deposited in glomeruli in lupus nephritis. Here, we show that individuals with SLE have IgA autoantibodies against most nuclear antigens, correlating with IgG against the same antigen. We investigated whether IgA autoantibodies against a major SLE autoantigen, Smith ribonucleoproteins (Sm/RNPs), play a role in IC activation of pDCs. We found that pDCs express the IgA-specific Fc receptor, FcαR, and there was a striking ability of IgA1 autoantibodies to synergize with IgG in RNA-containing ICs to generate robust pDC IFNα responses. pDC responses to these ICs required both FcαR and FcγRIIa, showing a potent synergy between these Fc receptors. Sm/RNP IC binding to and internalization by pDCs were greater when ICs contained both IgA1 and IgG. pDCs from individuals with SLE had higher binding of IgA1-containing ICs and higher expression of FcαR than pDCs from healthy control individuals. Whereas pDC FcαR expression correlated with blood ISG signature in SLE, TLR7 agonists, but not IFNα, upregulated pDC FcαR expression in vitro. Together, we show a new mechanism by which IgA1 autoantibodies contribute to SLE pathogenesis.

Figure 1.. IgA and IgG RNA-associated anti-nuclear antibodies in SLE serum.

(A) Autoantibody profiling of serum from 24 SLE subjects. Heatmaps of log-scaled antibody binding for IgA (left) and IgG (right) isotype antibodies to 16 RNA-associated nuclear antigens. Each row corresponds to an antigen and each column corresponds to an individual subject. Color bar on top indicates SLEDAI score for each donor at time of blood draw. Clustering of subjects and samples is based on similarity in scaled IgA levels, with the same clustering applied to the IgG heatmap. (B) Correlation between IgA and IgG for anti-Sm/RNP (left) and dsDNA (right) nuclear antigens. Each symbol represents an individual subject. (C) Serial dilution curves of immune complexes (ICs) containing both IgA and IgG in serum from 26 SLE subjects as measured by ELISA. Of the 26 SLE subjects 15% (4/26) had no detectable IgA/IgG ICs. Each line represents an individual subject. (D) Histology and immunofluorescence of representative glomeruli from class IV lupus nephritis showing Jones silver staining (left), and IgA (middle) and IgG (right) staining of serial sections. (B) r = Pearson correlation coefficient.

Figure 2.. Human blood pDCs express FcαR and FcγRIIa and bind IgA and IgG.

Surface staining and flow cytometry analysis performed on PBMCs from healthy control (HC) donors. (A) Representative histograms of FcαR (red) and FcγRIIa (blue) expression on pDCs compared to mIgG1 isotype control (grey), with detection by the same secondary antibody. (B) Paired analysis between monocyte and pDC FcαR (left) and FcγRIIa (right) surface staining (n=12 HC donors). (C) Binding of IgA and IgG to pDCs was assessed using biotinylated human heat-aggregated IgA or heat-aggregated IgG and fluorescently labeled streptavidin for visualization by flow cytometry, Streptavidin (SA) alone (grey), heat-aggregated IgA (red) and heat-aggregated IgG (blue). (D) Direct ex vivo binding of IgA and IgG to pDCs was assessed by staining with anti-IgA (red) or anti-IgG antibody (blue) directly ex vivo. Control (grey) has no anti-IgA or anti-IgG detection antibody. (A–D) Staining was performed on thawed PBMCs and pDCs gated as shown in Supplementary Fig. 2B. (B) Ratio paired t-test (**** p<0.0001).

Figure 3.. Reagents for generating RNA-containing Sm/RNP immune complexes

(A) Schematic showing reagents used to generate immune complexes (ICs) that contained both IgG and IgA (I.), mainly IgG (II.) or only IgA1 (III.) from SLE donors with IgG and IgA anti-Sm/RNP antibodies. IgA1 depleted serum (SLE serum ΔIgA1) and purified SLE IgA1 were generated from SLE serum by using an IgA1 binding substrate column and collecting the flow through and eluted protein, respectively. ICs were made by mixing the different antibody reagents with RNA-containing Sm/RNPs. (B) Levels of IgA anti-Sm/RNP (left) and IgG anti-Sm/RNP (right) antibodies in SLE serum (purple circles), SLE serum ΔIgA1 (blue open squares) and SLE IgA1 (red diamonds), as measured by ELISA. Representative data for one serum donor is shown.

Figure 4.. IgA1 in Sm/RNP immune complexes enhances IFNα production by pDCs

pDCs enriched from PBMCs freshly isolated from healthy control (HC) donor blood were incubated with immune complexes (ICs) generated with Sm/RNP RNA-containing nuclear antigens. (A) pDC IFNα secretion after incubation with Sm/RNPs alone or with Sm/RNP ICs generated with SLE serum, SLE serum ΔIgA1 or SLE IgA1. Each symbol shape represents a different pDC donor, mean and range shown. (B) Paired analysis of pDC IFNα secretion after incubation with Sm/RNP ICs generated with purified SLE IgG (135 μg) and purified SLE IgA1 (10 μg) or purified SLE IgG alone (135μg). (C) Paired analysis of pDC IFNα secretion after incubation with Sm/RNP ICs generated with either SLE serum or SLE serum ΔIgA1 from 8 SLE serum donors. (D) Paired analysis of pDC IFNα secretion after incubation with Sm/RNP ICs generated with either SLE serum or SLE serum ΔIgA1 from the same SLE serum donor and the same HC pDC donor assessed at 4 times over 14 months. (E) Paired analysis of pDC IFNα2 (left) and TNF (right) secretion after incubation with SLE serum or SLE serum ΔIgA1 ICs. (F) pDC IFNα secretion after pDCs were incubated with PBS, mIgG1 isotype control, anti-FcγRIIa, anti-FcαR, aggregated IgG or aggregated IgA for 2 hours prior to the addition ICs generated from SLE serum. Data are shown as a percent of PBS control with mean and range (raw data shown in Supplemental Figure 4E). (A) One way ANOVA, repeated measures, Bonferroni correction for multiple comparisons between all groups. Only significant comparisons are shown (*** p<0.001 and ** p<0.01). (B-D) Paired t-test (** p<0.01 and * p<0.05). (F) One way ANOVA, repeated measures, Bonferroni correction for comparing all blocking conditions to the isotype control (*** p<0.001). SLE serum donors; n=1 (A, B D and F), n=4 (E), n=8 (C). HC pDC donors; n=1 (C-E), n=3 (F), n=4 (A and B).

Figure 5:. IgA1 autoantibodies contribute to IC association with pDCs

(A) pDCs were incubated with Sm/RNP-AF647 alone (bottom row) or Sm/RNP-AF647 ICs generated from SLE serum (top row) or SLE serum ΔIgA1 (middle row), for 3–20 hours. Representative flow cytometry plots show the percent of Sm/RNP-AF647⁺ pDCs. The Sm/RNP-AF647⁺ gate was determined by selecting where binding of Sm/RNP-AF647 control was approximately 1% or less (bottom row). (B) Percent IC⁺ pDCs at 12 hours after incubation with ICs generated with SLE serum or SLE serum ΔIgA1 from 4 SLE donors. (C) Correlation between IC⁺ pDCs (%) at 12 hrs and IFNα secretion at 20 hours (ng/mL) after incubation with Sm/RNP-AF647 ICs generated from SLE serum (solid lavender symbols) or SLE serum ΔIgA1 (open blue symbols). Plot shows the same IFNα secretion data from (B) with each symbol shape corresponding to the SLE serum donor. (D) Correlation of 12 hr IC⁺ pDCs (%) and normalized 20 hr IFNα production (% max) for combined experiments in which pDCs were incubated with Sm/RNP-AF647 ICs generated with SLE serum or SLE serum ΔIgA1. (B) Ratio paired t-test (* p<0.05) (C, D) r = Pearson’s correlation coefficient. SLE serum donors; n=1(A), n=4 (B, C, D). HC pDC donors; n=1 (A, B, C), n=4 (D).

Figure 6:. IgA1 autoantibodies enhance pDC internalization of immune complexes

Enriched pDCs from healthy individuals were incubated with Sm/RNP-AF647 alone or Sm/RNP-AF647 ICs generated with either SLE serum or SLE serum ΔIgA1 for 12 hrs (n=2 SLE serum donors). (A) Representative confocal images of the three categories of staining observed with DAPI (blue), CD123 (green) and Sm/RNP-AF647 ICs (magenta). Group I (left), pDCs that had no ICs bound or internalized or very faint IC signaling indicating low level or diffuse ICs bound. Group II (middle), pDCs with either small (top cell) or large (bottom cell) internalized ICs. Group III (right), pDCs with either small (top cell) or large (bottom cell) ICs that were bound but not internalized. Images shown are a composite of collapsed z-stacks, Supplemental Fig. 6 shows examples with full z-stacks to demonstrate the difference between internalized and bound ICs. (B) Donut plots show the proportion of pDCs in each group for each experimental condition: Sm/RNP-AF647 with no serum (left), Sm/RNP-AF647 ICs generated with SLE serum (middle), and Sm/RNP-AF647 ICs generated with SLE serum ΔIgA1 (right). The center of the donut plot shows the number of cells imaged and analyzed in each group. Chi-squared analysis performed between all groups and statistically significant comparisons shown (** p<0.01, ** p<0.01).

Figure 7.. SLE pDCs have increased binding to immune complexes and increased expression of FcαR

(A) PBMCs from individuals with SLE (n=12) and matched healthy control subjects (n=12) were incubated for 3 hours with ICs generated from Sm/RNP-AF647 and SLE serum. Percent of gated pDCs with ICs is shown, determined as described in Figure 5A. (B) PBMCs from the same subjects as in (A) were also stained for surface FcαR and FcγRIIa and correlations between IC⁺ pDCs and FcαR (left) or FcγRIIa (right) are shown. Grey circles are pDCs from healthy control individuals and purple circles those from SLE individuals. (C) Cell surface expression of FcαR (left) and FcγRIIa (right) on gated pDCs from individuals with SLE (n=36) and matched HC subjects (n=37). (D) Cell surface expression of FcαR on gated monocytes from the same SLE and HC subjects in Fig. 7C. (A–D) Each circle represents an individual subject; (A, C and D) Student’s t-tests (* p<0.05 and ***p<0.001). (B) r=Pearson’s correlation coefficient.

Figure 8:. pDC FcαR expression correlates with IFN signature in SLE.

(A) Whole blood RNA-Seq was performed for n=18 donors that were also assessed for FcαR and FcγRIIa expression (Fig. 7C). Heatmap shows the top 50 genes positively correlated to FcαR expression. Scales above the heatmap show FcαR MFI expression (red), FcγRIIa expression (blue) and SLEDAI score (pink). Genes that were found to be part of the Hallmark IFN-alpha response gene set from the Broad’s Molecular Signatures Database are indicated with an asterisk (*). (B) Correlation between cell surface expression of FcαR on gated pDCs and an interferon-stimulated gene (ISG) signature determined by RNA-seq performed on whole blood from 18 of the SLE subjects in (Fig. 7C). (C) Correlation between cell surface expression of FcγRIIa on gated pDCs and ISG signature determined by RNA-seq performed on whole blood from 18 of the SLE subjects in (Fig. 7C). (D) pDC FcαR expression (geometric MFI) after 36 hr incubation with PBS control, TLR7 agonist R848, IFNα or TNF. (D) Healthy control pDC donors n=4–5. (B and C) r = Pearson’s correlation coefficient. (D) One way ANOVA, repeated measures comparing all induction conditions to the PBS control (* p<0.05).