Immune characterization of lupus nephritis patients undergoing dialysis

Abstract

Systemic lupus erythematosus (SLE) activity decreases in some patients with end-stage kidney disease (ESKD). The impact of ESKD on the immune cell profile of SLE patients and lupus activity remains unclear. In this study, we aimed at characterizing immunologically inactive and active SLE patients undergoing dialysis therapy. Based on multi-parametric flow cytometry assays, an extensive immunophenotyping was performed on blood samples from 47 SLE patients undergoing hemodialysis, 10 non-dialyzed SLE patients with active lupus nephritis (aLN), 6 non-dialyzed patients with a history of LN currently in remission (rLN), and 20 healthy volunteers (HV) as controls (ClinicalTrials.gov Identifier: NCT03921398). The hemodialysis group was composed of 16 SLE patients with inactive disease (iHD), 22 with sustained low disease activity with a non-renal SLEDAI ≤4 (aHD≤4), and 9 highly active SLE patients (aHD>4). A factorial discriminant analysis was performed to validate the association between immune cell signatures and lupus activity. By compiling 12 cellular variables, we describe immune profiles related to highly active SLE patients or associated with both inactive and low-disease activity groups. As non-dialyzed active SLE patients, active patients undergoing hemodialysis showed a specific combination of increased numbers of circulating CD19^hi CD27^- "atypical naive" B cells, plasmablasts, CD16⁺ inflammatory monocytes and a basopenia. This study brings a comprehensive overview of immune cell signatures observed in SLE patients undergoing dialysis. We propose a simple immunophenotypic approach for the assessment of lupus activity that may provide help to data-driven personalized medicine in hemodialyzed SLE patients.

Keywords: Dialysis; Disease activity; End stage kidney disease; Immunophenotyping; Lupus nephritis; Systemic lupus erythematosus.

Fig. 1

Identification and characterization of CD19^hi CD27^– ‘’atypical naive’’ B cells in SLE patients. (A) Identification of B cell subsets using CD19 and CD27 markers (represented in SSC^lo CD16^– CD14^– leukocytes). Flow cytometry plots are representative of HV (healthy volunteers), aLN (active SLE patients with lupus nephritis (LN)), rLN (SLE patients with a history of LN currently in remission), aHD (active patients undergoing hemodialysis (HD)) and iHD groups (inactive patients undergoing HD). Proportions (%) of subsets among CD19⁺ B cells are indicated in green: CD19⁺ CD27^– conventional naive, blue: CD19⁺ CD27^– memory, red: CD19^hi CD27^– ‘’atypical naive”, orange: CD19^lo CD27^hi plasmablasts). (B) Representative flow cytometry plots of IgD and CD24 expression in CD19^hi CD27^– ‘’atypical naive’’ B cells (red gate with proportions among CD19⁺ B cells) and other B cells (black gate), in CD15^– CD16^– CD14^– Peripheral Blood Mononuclear Cells (PBMCs). Proportions (%) of IgD⁺ CD24^– activated naive (acN) and IgD^– CD27^– double negative (DN) within ‘’atypical naive’’ (middle red panel) or other (right black panel) B cells are indicated above gates. (C) DN and acN among CD19^hi CD27^– B cells (%). (D) Flow cytometry histograms of CD95, CXCR5, CD11c, IgD and CD24 expression in CD19^hi CD27^– (red line) and other B cells (black line). (E) Phosphorylation levels of PLCγ2 (Y759) and SYK (Y348 and Y525-526) measured by flow cytometry in 4 B cell populations of active lupus patients. Filled grey: isotype control. (F) Relative fluorescence intensity of PLCγ2 and SYK phosphorylations in the 4 B cell subsets in aHD patients (n = 31), mean +s.e.m. Arbitrary Unit (A.U.): ratio between geometric mean fluorescence intensity of specific staining on isotype controls. (B–D) Data are representative of on the analysis of PBMCs from 6 active SLE patients. (C) Data are presented as individual values in truncated violin plots showing median (thick line) and quartiles (thin lines). ∗∗∗:P ≤ 0.001 by unpaired t-test. (F) Data are presented as mean +s.e.m. ∗P ≤ 0.05, ∗∗P ≤ 0.01, ∗∗∗P ≤ 0.001, ∗∗∗∗P ≤ 0.0001, by Kruskal-Wallis test followed by Dunn's post-tests.

Fig. 2

Quantification of 4 relevant B cell populations in SLE patients. (A) B cell counts in HV (n = 20), aLN (n = 10), rLN (n = 6), aHD (n = 31) and iHD (n = 16) groups. (B-D,F) Frequencies (%) of subsets defined in Fig. 1A among B cells in HV, aLN, rLN, aHD and iHD. (E) CD19^hi CD27^– (%) among B cells in aHD≤4 (aHD with nrSLEDAI ≤4) (n = 22), aHD>4 (nrSLEDAI >4) (n = 9) and iHD (n = 16) patients. (G,H) Correlation (and linear regression with 95 % confidence intervals (dotted lines)) between CD19^hi CD27^– B cells and plasmablasts absolute numbers in iHD (G, Pearson's r = 0.1621, P = 0.54, n = 16) and aHD patients (H, Pearson's r = 0.4580, P = 0.0096, n = 31). (A–F) Data are presented as individual values in truncated violin plots showing median (thick line) and quartiles (thin lines). ∗P ≤ 0.05, ∗∗P ≤ 0.01, ∗∗∗P ≤ 0.001, ∗∗∗∗P ≤ 0.0001, ns = not significant, by one-way ANOVA followed by Tukey's post-tests (B,C), or by Kruskal-Wallis tests followed by Dunn's post-tests (A,D-F). (D) # (aLN vs iHD): P = 0.0737, # (aHD vs iHD): P = 0.0794. Pairwise comparisons not displayed on graphs had P values > 0.1.

Fig. 3

Granulocytes in SLE patients undergoing hemodialysis. (A) Representative flow cytometry plots of eosinophils (Eos.) and neutrophils (Neutro.) selection by flow cytometry, based on side scatter (SSC-A) parameter and expression of CD16 molecule. Proportions (%) of Eos. and Neutro. among leukocytes are indicated for HV, aLN, rLN, aHD and iHD individuals. Proportions (%) and absolute numbers of neutrophils (B) and eosinophils (C) among leukocytes in HV (n = 20), aLN (n = 10), rLN (n = 6), aHD (n = 31) and iHD (n = 16) groups. Neutrophil (B, right panel) and eosinophil (C, right panel) counts in iHD (n = 16), aHD≤4 (n = 22), and aHD>4 (n = 9) individuals. (D) Representative flow cytometry plots of basophil gating based on CRTH2 and FcεRIα expression (in CD14^–, CD16^–, CD19^– leukocytes). (E) Proportions (%) among leukocytes and absolute numbers of basophils in HV, aLN, rLN, aHD and iHD groups. Right panel: basophil counts in iHD, aHD≤4, and aHD>4 patients. (B,C,E) Data are presented as individual values in truncated violin plots showing median (thick red line) and quartiles (thin red lines). ∗P ≤ 0.05, ∗∗P ≤ 0.01, ∗∗∗P ≤ 0.001, ns = not significant by Kruskal-Wallis tests followed by Dunn's post-tests. (E) # (HV vs aLN): P = 0.0747. Pairwise comparisons not displayed on graphs were ns.

Fig. 4

Analysis of monocyte subsets in healthy volunteers and lupus patients. (A) Representative flow cytometry plots of CD14 and CD16 expression in CD19^–, CD4^–, CD8α^– and not SSC^hi leukocytes. Proportions (%) of Non-Classical Monocytes (NCM), Intermediate Monocytes (IM) and Classical Monocytes (CM) among monocytes are indicated next to the corresponding gates for HV, aLN, rLN, aHD and iHD individuals. Proportions (%) among monocytes (B) and absolute numbers (C) of CM, IM and NCM in HV (n = 20), aLN (n = 10), rLN (n = 6), aHD (n = 31) and iHD (n = 16) individuals. Absolute numbers (D) and proportions among monocytes (E) of CM and CD16⁺ monocytes in iHD (n = 16), aHD≤4 (n = 22), and aHD>4 (n = 9) SLE patients. (F) TNF, IL1B and TNFSF13B mRNA expression levels relative to GAPDH mRNA (2^–ΔCt) measured by real-time PCR in sorted classical (CM, white bars) and CD16⁺ (grey bars) monocytes from iHD (n = 7) and aHD>4 (n = 7) patients (mean +s.e.m). (B-E) Data are presented as individual values in truncated violin plots showing median (thick red line) and quartiles (thin red lines). Statistical analyses were by Kruskal-Wallis tests followed by Dunn's post-tests (B-D), or by one-way ANOVA followed by Tukey's post-tests (E). (F) Two-way ANOVA followed by uncorrected Fisher's tests. ∗P ≤ 0.05, ∗∗P ≤ 0.01, ∗∗∗P ≤ 0.001, ∗∗∗∗P ≤ 0.0001, ns = not significant. (B) # (HV vs aHD) P = 0.0629, (C) # (aLN vs rLN): P = 0.0931, (E) # (iHD vs aHD>4): P = 0.0826. Pairwise comparisons not displayed on graphs were ns.

Fig. 5

Immunophenotyping based discriminant analysis of healthy controls and lupus patients. (A) Factorial Discriminant Analysis (FDA) was performed using the following 12 quantitative variables: absolute numbers of basophils, neutrophils, eosinophils, CD8α⁺ T cells, CD4⁺ T cells, CD19⁺ B cells, proportions among CD19⁺ cells of conventional naive B cells, memory B cells, “atypical naive” B cells and plasmablasts, and proportions among monocytes of classical (CM) and CD16⁺ monocytes. F1 and F2 factors show the efficiency of FDA in discriminating HV (n = 20, blue dots), iHD (n = 16, green dots), aHD≤4 (n = 22, grey dots), aHD>4 (n = 9, orange dots), rLN (n = 6, black dots) and aLN (n = 10, red dots) individuals. The sizes of the dots reflect their distribution on the F1 axis. (B) 12 quantitative variables (A–L) used to perform FDA and their correlation with F1 and F2 axis. Related groups column summarizes the association between quantitative variables and their efficiency in stratifying patients depending on the predefined group membership.