Normalizing glycosphingolipids restores function in CD4+ T cells from lupus patients

Abstract

Patients with the autoimmune rheumatic disease systemic lupus erythematosus (SLE) have multiple defects in lymphocyte signaling and function that contribute to disease pathogenesis. Such defects could be attributed to alterations in metabolic processes, including abnormal control of lipid biosynthesis pathways. Here, we reveal that CD4+ T cells from SLE patients displayed an altered profile of lipid raft-associated glycosphingolipids (GSLs) compared with that of healthy controls. In particular, lactosylceramide, globotriaosylceramide (Gb3), and monosialotetrahexosylganglioside (GM1) levels were markedly increased. Elevated GSLs in SLE patients were associated with increased expression of liver X receptor β (LXRβ), a nuclear receptor that controls cellular lipid metabolism and trafficking and influences acquired immune responses. Stimulation of CD4+ T cells isolated from healthy donors with synthetic and endogenous LXR agonists promoted GSL expression, which was blocked by an LXR antagonist. Increased GSL expression in CD4+ T cells was associated with intracellular accumulation and accelerated trafficking of GSL, reminiscent of cells from patients with glycolipid storage diseases. Inhibition of GSL biosynthesis in vitro with a clinically approved inhibitor (N-butyldeoxynojirimycin) normalized GSL metabolism, corrected CD4+ T cell signaling and functional defects, and decreased anti-dsDNA antibody production by autologous B cells in SLE patients. Our data demonstrate that lipid metabolism defects contribute to SLE pathogenesis and suggest that targeting GSL biosynthesis restores T cell function in SLE.

Figure 1. Altered GSL profile in T cells from patients with SLE.

(A) Scheme showing GSL biosynthesis pathways, indicating some of the enzymes controlling biosynthesis. Cellular lipids were isolated from negatively selected CD4⁺ T cells from 40 SLE patients and 15 healthy donors by chloroform-methanol extraction. The total cellular GSL profile was analyzed by HPLC following glycanase digestion to release the GSL sugar head groups. (B) Representative qualitative HPLC plots showing the position of known GSL standards and GSL species in 1 healthy control and 2 SLE patients. (C) Cumulative quantitative data for each GSL species identified on the HPLC plots. GSL expression was calculated from the peak HPLC areas after applying an experimentally derived response factor (18) by relating the area of the HPLC peak to the cell number of the sample. Two-tailed Mann-Whitney U test; **P = 0.008; *P ≤ 0.05. Expression of surface GSL was determined in ex vivo PBMCs from 58 SLE patients, 36 healthy donors, and 10 patients with OADs (Sjögren’s syndrome and RA). Cells were stained using fluorescently labeled antibodies against CD4-v450, LC-PE-Cy5, Gb3-FITC, or CTB-FITC and analyzed by flow cytometry. (D) Representative flow cytometric dot plots showing staining with appropriate controls (percentage of CD4⁺GSL^hi T cells and GSL MFI of total CD4⁺ T cells is shown). Cumulative data of percentage of CD4⁺GSL^hi T cells (E) and GSL MFI in total CD4⁺ T cells (F). One-way ANOVA; *P ≤ 0.05; **P ≤ 0.007; ***P = 0.0006.

Figure 2. Increased GSL expression is associated with defective GSL homeostasis in T cells from SLE patients.

PBMCs from 8 healthy donors and 8 SLE patients were stimulated for 24 and 72 hours with or without 1 μM GW3965 (GW). Cells were stained for CD4-APC and CTB-FITC or LC-FITC. Representative flow cytometric plots for (A) CTB binding and (B) LC expression after a 24- and 72-hour culture and (C) cumulative data. MO, medium only. One-way ANOVA, *P ≤ 0.05; 2-tailed Student’s t test, **P ≤ 0.003. RNA extracted from negatively isolated CD4⁺ T cells from 10 SLE patients and 6 healthy controls was assessed by qPCR for the expression of LXRB, LXRA, NPC1, NPC2, ABCA1, ABCG1, and SREBP2 genes. (D) Cumulative results are shown in relative units comparing the gene of interest with a GAPDH control. Two-tailed Student’s t test; *P ≤ 0.05; **P ≤ 0.001. Ex vivo PBMCs from 13 SLE patients and 6 healthy donors were surface stained for CD4-APC followed by intracellular staining for LXRβ before analysis by flow cytometry. (E) Cumulative data show the mean. Two-tailed Student’s t test; **P = 0.007.

Figure 3. Upregulation of LXRβ by oxysterol and TCR stimulation.

(A) CD4⁺ T cells from 3 healthy donors and 3 SLE patients were cultured for 18 hours with GW3965 or CM and analyzed by Western blotting for LXRβ expression. (B) Cumulative data. Paired and 2-tailed Student’s t test; *P ≤ 0.05. (C) CD4⁺ T cells from 5 healthy donors and 5 SLE patients cultured for 18 hours with GW3965 or CM were assessed by qPCR for LXRB expression. Cumulative results comparing LXRB with GAPDH control. Paired and 2-tailed Student’s t tests; *P ≤ 0.05. PBMCs from a healthy donor were cultured for 72 hours (all plus IL-2) with serum from 5 heterologous healthy donors (HC serum), 5 SLE patients (SLE serum), or with CM only (MO). (D) Cumulative data. One-way ANOVA; *P = 0.05. (E) CD4⁺ T cells from 5 healthy donors were cultured for 72 hours with LDL, oxidized LDL (oxLDL), or CM before CTB staining. Representative histograms and cumulative data. One-way ANOVA; *P ≤ 0.01. (F) PBMCs from a healthy donor were cultured with serum from 12 healthy donors, 12 SLE patients, or with CM (all plus IL-2) (for 72 hours and for 10 days) and stained with CTB. Cumulative data showing percentage change from CM. One-way ANOVA; *P ≤ 0.05. (G) PBMCs from 4 SLE patients and 4 healthy donors were cultured for 3 days with GW3965 ± the LXR antagonist 5CPPSS-50 and stained with CTB. One-way ANOVA; *P ≤ 0.05. (H) PBMCs from a healthy donor were cultured for 10 days as described in F with serum from 4 SLE patients ± 5CPPSS-50 and stained with CTB. One-way ANOVA; *P = 0.03.

Figure 4. Increased recycling of GSLs in T cells from SLE patients.

CD4⁺ T cells from 6 SLE patients and 6 healthy donors were incubated with Bodipy-LC on ice, then washed and cultured for up to 30 minutes at 37°C to allow endocytosis before analysis by confocal microscopy and flow cytometry. (A) Representative confocal microscopy images of Bodipy-LC at 1 and 30 minutes. Bodipy-LC was excited at 450 to 490 nm and viewed at 520 to 560 nm (green) or >590 nm (red). Scale bars: 5 μM. (B) Quantitative flow cytometric data showing Bodipy-LC emissions at 520 to 560 nm (left) and >590 nm (right). Paired Student’s t test, *P = 0.05 and **P = 0.005; 2-tailed Student’s t test, ***P ≤ 0.05. (C) Experiment in B was repeated ± dynasore to inhibit endocytosis. Representative histograms. (D) CD4⁺ T cells from 5 healthy donors and 5 SLE patients were labeled with CTB-FITC. Cells were washed and incubated at 37°C for 1, 3, or 5 minutes to allow endocytosis of CTB-stained lipids. Cells were washed in neutral or acidic PBS, and lipid internalization was calculated as a ratio of the two. As a control, cells were preincubated with dynasore (dyn). Two-tailed Student’s t test; *P = 0.04; ***P = 0.0005. (E) Surface GSLs on CD4⁺ T cells from 5 healthy donors and 5 SLE patients were blocked with unconjugated CTB and incubated at 37°C for 5, 10, and 15 minutes. Newly recycled GSLs were detected by surface staining with CTB-FITC and flow cytometry. Two-tailed Student’s t test; *P = 0.03.

Figure 5. Increased LAMP1 expression and altered lipid colocalization to functionally normal lysosomes in SLE CD4⁺ T cells.

(A) PBMCs from 11 healthy donors and 11 SLE patients were surface stained for CD4-APC before fixing, permeabilizing, and staining for intracellular LAMP1-PE (CD107a-PE). Representative dot plots and cumulative data. Two-tailed Student’s t test; ****P = 0.00001. (B–E) CD4⁺ T cells from 6 SLE patients and 3 healthy donors were fixed and permeabilized before staining for LAMP1 and either CTB (B) or LC (D) and analyzed by confocal microscopy. Scale bars: 5 μM (insets show ×2.5 enlargement). Colocalization of LAMP1-CTB (C) or LAMP1-LC (E) assessed by Pearson’s colocalization coefficient (Rr). Cumulative data. Two-tailed Student’s t test, *P = 0.05. (F) PBMCs from 10 SLE patients were labeled with CD4-APC and incubated with LysoSensor Green DND before flow cytometric analysis. Results were correlated with LAMP1 expression measured in F. R² = 0.4754; *P ≤ 0.05.

Figure 6. Inhibition of GSL biosynthesis modifies GSL recycling in T cells from SLE patients.

(A) PBMCs from 24 SLE patients and 11 healthy donors were TCR stimulated for 72 hours with or without 10 μM NB-DNJ. Cells were surface stained using fluorescently labeled antibodies against CD4-v450, LC-PE-Cy5, or CTB-FITC and assessed by flow cytometry. Cumulative data showing expression of CTB and LC. Two-tailed Student’s t test; *P ≤ 0.05. (B) CD4⁺ T cells from 5 SLE patients and 5 healthy donors were cultured for 72 hours with or without NB-DNJ. Internalization of endogenous GSLs after a 5-minute incubation was assessed as described in Figure 4D. Cumulative data showing the effect of NB-DNJ treatment on CTB internalization. Two-tailed Student’s t test; *P ≤ 0.05. CD4⁺ T cells from 3 SLE patients and 3 healthy donors were cultured for 72 hours with or without NB-DNJ. Uptake of Bodipy-LC was assessed as described in Figure 4, A and B. (C) Line graphs showing uptake of Bodipy-LC by flow cytometry. Paired Student’s t-test for SLE samples; *P ≤ 0.01. (D) Representative confocal images showing merged green and red Bodipy-LC accumulation in T cells after a 30-minute incubation. Scale bars: 5 μM. (E) CD4⁺ T cells isolated from 4 patients with SLE were cultured for 72 hours with or without NB-DNJ. Cells were surface stained for CD4-APC and intracellularly stained for LAMP1-PE. Paired Student’s t test; *P ≤ 0.05.

Figure 7. NB-DNJ modifies TCR-associated defects in T cells from patients with SLE.

PBMCs from 5 SLE patients and 5 healthy donors were cultured for 7 days with NB-DNJ. Cells were labeled with anti-CD3/CD28 (1 μg/ml) and cross-linking anti-mouse IgG on ice before culturing at 37°C for 1 and 5 minutes. Cells were fixed and labeled with fluorescent barcoding reagents and stained for CD4-APC and intracellularly stained for the phosphorylated signaling proteins pTCR-ζ, pERK, pAKT, and pNF-κB. (A) Cumulative results from three separate experiments. Two-tailed Student’s t test; *P ≤ 0.05 or paired t test; **P = 0.001. PBMCs from 24 SLE patients and 11 healthy donors were TCR stimulated for 72 hours with anti-CD3/CD28 (1 μg/ml) with or without 10 μM NB-DNJ. Cell culture supernatants were collected. Cells were surface stained with CD4-v450 and intracellularly stained for Ki67-PE and assessed by flow cytometry. (B) Representative flow cytometric dot plots (C) and cumulative results showing Ki67 expression as the percentage change from unstimulated controls. Two-tailed Student’s t test; *P ≤ 0.05. Cell culture supernatants from 12 SLE patients and 10 healthy donors collected from B were analyzed by CBA. Representative CBA plots (D) and cumulative data (E) are shown. Two-tailed Student’s t test and paired t test; *P ≤ 0.05. (F) CD4⁺ T cells isolated from 4 SLE patients were pretreated with NB-DNJ, then cocultured with autologous B cells for 7 days in the presence of anti-CD3 (1 μg/ml). The production of anti-dsDNA antibodies was detected by ELISA. Paired Student’s t test; *P ≥ 0.05.

Figure 8. Proposed model for altered GSL expression and metabolism in CD4⁺ T cells from SLE patients.

We propose that several components of SLE pathology, including dyslipidemia and T cell hyperactivity (i), elicit upregulation of the nuclear receptor LXRβ, and hence its downstream target genes NPC1 and NPC2 (ii). Enhanced activation of LXRβ induces increased de novo GSL biosynthesis (iii). CD4⁺ T cells from SLE patients are characterized by accelerated GSL trafficking and accumulation of GSL in the plasma membrane (iv) and intracellular compartments (v), despite higher levels of functionally normal lysosomes (vi). Inhibition of GSL biosynthesis with NB-DNJ reduced cellular GSL levels and modified cell function in CD4⁺ T cells from SLE patients (vii). TGN, trans-Golgi network.