Elevated serum autoantibodies against co-inhibitory PD-1 facilitate T cell proliferation and correlate with disease activity in new-onset systemic lupus erythematosus patients

The study included 90 consecutive, untreated, new-onset patients with SLE who fulfilled the American College of Rheumatology (ACR) classification criteria for the diagnosis of SLE [17], and 50 patients with primary Sjogren’s syndrome (pSS), 50 with rheumatoid arthritis (RA) and 25 with ankylosing spondylitis (AS) according to the standard diagnostic criteria [18‐20]. Finally, samples from 80 sex- and age-matched healthy donors with neither autoimmune nor infectious diseases were collected as healthy controls. Demographic data, clinical characteristics, and laboratory findings of SLE patients were collected. All of the sera samples were stored at -80 °C until use. Disease activity was measured using the SLE disease activity index 2000 (SLEDAI-2 K) [21]. The study was performed in accordance with the Declaration of Helsinki and the principles of Good Clinical Practice. Biological samples were obtained under a protocol approved by the Institutional Research Ethics Committee of Ruijin Hospital (ID: 2016-62), Shanghai, China. All subjects signed written informed consent.

The extracellular domain of the human PD-1 gene (NP_005009.2) Leu24-Gln167 was synthesized by General Biosystems Inc. (Morrisville, NC, USA). The PD-1 extracellular domain gene was fused to the human immunoglobulin (Ig)G1 Fc region at the N-terminus and the 6xHis tag at the C-terminus using the OE-PCR method. The PCR product was digested with BamHI and XhoI, and cloned into the pCDNA3.4-Topo vector (Invitrogen, Grand Island, NY, USA) under the CMV promoter and mouse Ig-kappa signal peptide for high-level secretion. The recombinant PD-1-nFc-His6 was expressed in Expi293 cells by transient transfection using ExpiFectamine reagent (Life Technologies, Carlsbad, CA, USA). After transfection for 3 days, the culture supernatant was collected and the recombinant protein was purified using a protein A column (GE Healthcare, Milwaukee, WI, USA). As a result of glycosylation, the purified PD-1-nFc-His6, as a disulfide-linked homodimer, migrated as an approximately 120 kDa protein in nonreducing SDS-PAGE. The homodimer was incubated with recombinant enterokinase (Novoprotein, Shanghai, China) at 37 °C for 24 h, passed through the protein A column to remove the Fc region, and exchanged to a PBS buffer for storage at 4 °C. The concentration of recombinant PD-1-His6 was determined by bicinchoninic acid (BCA) protein assay (Pierce Biotechnology, Rockford, USA). The recombinant PD-1-His6 consisted of 156 amino acids and, due to glycosylation, migrated as an approximately 35 kDa protein in nonreducing SDS-PAGE.

Antibodies against human PD-1 in the sera of SLE patients, pSS patients, RA patients, AS patients and healthy controls were determined by using an enzyme-linked immunosorbent assay (ELISA). Ninety-six-well high-binding plates (Corning, Corning, NY, USA) were coated with recombinant human PD-1 extracellular domain (1 μg per well) overnight at 4 °C in 0.05 mol/L carbonate buffer sodium (PH = 9.6). The antigen-coated wells were washed three times with PBST (PBS plus 0.05% Tween-20) and blocked with PBST containing 5% bovine serum albumin (BSA) for 2 h at 37 °C. The blocking buffer was tapped off and washed as described above before the addition of 100 ul of serum sample (1:100 diluted in 1% BSA). The PD-1 monoclonal antibody (Novoprotein, Summit, NJ, USA) was used as a positive control. Human sera were incubated for 2 h at room temperature followed by incubation with HRP-conjugated goat anti-human IgG and IgM (Abcam, Cambridge, UK) for another 1 h at room temperature. Then, the plates were washed and 100 ul of tetramethylbenzidine (TMB) substrate solution was added. The color development was stopped by the addition of 50 ul of 0.5 M H₂SO₄. Absorbance was measured at a wavelength of 450 nm in a microplate reader (Bio-Rad Laboratories, Richmond, CA, USA). All samples and standards were run in duplicates and the readings were performed in duplicates.

IgG was isolated from the sera of new-onset SLE patients with high levels of anti-PD-1 IgG according to the ELISA assay by protein A agarose (Pierce, Rockford, IL, USA). The isolates were examined for endotoxin contamination using a limulus amoebocyte lysate (LAL) assay (A&C Biological Ltd., Zhanjiang, China). The total protein content was estimated using a BCA assay.

The purification of specific antibodies was performed according to a previous report [22]. In brief, recombinant human PD-1 was spotted onto a nitrocellulose filter and incubated with the purified IgG from SLE patients. After washing with PBST, the antibodies were eluted with 100 mM of glycine (pH = 2.8). The eluted antibodies were immediately neutralized with 1 M of Tris-HCl (pH = 8.0). Normal human IgG was purchased from Millipore (Darmstadt, Germany). The purified PD-1 antibodies were concentrated with a final concentration of 1 mg/ml. After sterile filtration, the autoantibodies were stored as small aliquot package in -80 °C.

Peripheral blood monocytes (PBMCs) were isolated from the blood of three healthy volunteers using Ficoll density gradient centrifugation (GE Healthcare). The CD14⁺ PBMCs were isolated by positive selection using CD14 microbeads (Miltenyi Biotec, Auburn, CA, USA) according to the manufacturer’s instructions. The selected cells were cultured in RPMI 1640 containing 10% fetal calf serum (FCS; Gibco, Grand Island, NY, USA), 2 mM L-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin and supplemented with 20 ng/ml of granulocyte-macrophage colony-stimulating factor (GM-CSF) and 20 ng/ml of interleukin (IL)-4 (R&D Systems, Inc., Minneapolis, MN, USA) in a humidified 5% CO₂ incubator. On day 7, DCs were stimulated with 50 ng/ml of lipopolysaccharide (LPS) for 24 h. The matured DCs were labeled with anti-human CD80, CD86, and HLA-DR (BD Biosciences Pharmingen, San Diego, CA, USA) and analyzed using a FACS Canto II instrument (BD Biosciences Pharmingen). More than 90% of the cultured cells were triple positive for CD80, CD86, and HLA-DR.

Autologous T cells were isolated from the PBMC of three healthy volunteers on day 8. The CD3⁺ T cells were isolated by positive selection using CD3 microbeads (Miltenyi Biotec) according to the manufacturer’s instructions. CD3⁺ T cells were cultured in RPMI 1640 containing 10% FCS, 2 mM L-glutamine, 100 units/ml penicillin, 100 μg/ml streptomycin and supplemented with 20 U/ml of IL-2 (R&D Systems, Inc.) and 0.5 ug/ml of mouse anti-human CD3 (BD Biosciences Pharmingen) for 3 d. Activated T cells were stained with PE-labeling anti-human PD-1 antibody (BD Biosciences Pharmingen) and examined using a FACS Canto II instrument.

For the proliferation assay, 1 × 10⁵ T cells were labeled with 10 μM of CFSE (Sigma-Aldrich, St. Louis, MO, USA) in 100 μl pre-cold PBS, followed by incubation in 37 °C for 10 min and neutralization with pre-cold fetal bovine serum according to the manufacturer's instructions (Molecular Probes, Leiden, Netherlands). CFSE-labeled T cells were incubated with fresh medium containing purified human PD-1 antibodies from SLE patients or normal human IgG for 1 h at 37 °C. After being co-cultured with 1 × 10⁴ DCs in round-bottom 96-well tissue culture plates (Corning) for 72 h, cells were collected. PerCP Cy5.5-conjugated anti-human CD4 antibody was used for CD4 T cell staining according to the manufacturer’s protocols. Each experiment was preformed, duplicated and analyzed using a FACS Calibur cytometer and analyzed using FlowJo software (Tree Star Inc., Ashland, OR, USA).

For continuous variables, comparisons were carried out using an independent t test for two independent samples, a Mann-Whitney U test for non-normal data, and a paired t test for matched samples. Categorical variables were compared by the chi-square test or Fisher’s exact test. Correlations between groups were evaluated by the Spearman test. Results are expressed as the mean ± standard deviation (SD) of three independent experiments. Multivariate analysis was performed in order to identify the independent factors that are positively associated with the dependent variable. To control for possible confounding parameters, various multivariate logistic regression models were designed. The magnitude of each factor was expressed as an odds ratio (OR). A p value less than 0.05 was considered statistically significant. Graphs were drawn using Graphpad Prism (version 6, Graphpad Software, San Diego, CA, USA). Data were analyzed using the IBM SPSS software package for Windows (version 23.0; IBM Corp., Armonk, NY, USA).

A total of 90 patients diagnosed with SLE, 50 patients with pSS (mean age ± SD, 32.39 ± 7.6 years; female/male, 41/9), 50 patients with RA (mean age ± SD, 30.87 ± 8.1 years; female/male, 42/8), 25 patients with AS (mean age ± SD, 30.19 ± 7.1 years; female/male, 20/5) and 80 healthy donors (mean age ± SD, 31.45 ± 8.9 years; female/male, 68/12) were included in this study. There were no significant differences between the groups with regard to age and gender ratio. SLE disease activity was evaluated using the systemic lupus erythematosus disease activity index 2000 (SLEDAI-2 K) [21]. The history, manifestation, and autoantibody profile of SLE was collected for each patient. Demographic and clinical data are summarized in Table 1.

For each individual, the optical density (OD) values at 450 nm for anti-PD-1 IgG and IgM were examined using ELISA. As shown in Fig. 1, the OD value of anti-PD-1 IgG in sera from SLE patients was significantly higher than in healthy controls (HC, p < 0.001), RA group (p < 0.001), pSS group (p < 0.01), and AS group (p < 0.001) whereas the IgM isotype of PD-1 antibodies showed no significant difference between any two groups. Next, we defined the positive sample as the ratio of the positive and negative optical densities (P/N) greater than 2.1, according to previous work [23]. Among 90 patients with SLE, the positive rate of anti-PD-1 IgG and IgM was 30 (33.3%) and 18 (5%), respectively.

Table 2 shows the correlation of anti-PD-1 with the clinical parameters of patients with new-onset SLE. The level of anti-PD-1 is closely associated to malar rash (OR = 15.773; 95% CI 3.065–81.186), arthritis (OR = 22.937; 95% CI 4.619–113.9), serositis (OR = 16.008; 95% CI 2.119–120.967), hematological (OR = 35.187; 95% CI 6.459–191.679), renal (OR = 8.306; 95% CI 1.323–52.132), and neurological involvement (OR = 37.282; 95% CI 1.497–928.697). However, no association was observed between anti-PD-1 and the clinical features of discoid rash, oral ulcer, photosensitivity, vasculitis, or Raynaud’s phenomenon (p > 0.05). These findings suggest that anti-PD-1 might play a critical role in the occurrence of systematic clinical features.

Next, in order to evaluate the relationship of serum anti-PD-1 levels and disease activity, we first analyzed the correlation between anti-PD-1 IgG, the SLEDAI score, and laboratory parameters including erythrocyte sedimentation rate (ESR), anti-dsDNA, and antinuclear antibodies (ANA). The results show that the serum levels of anti-PD-1 IgG are positively correlated with SLEDAI score (r = 0.296, p = 0.0046) and ESR (r = 0.2446, p = 0.0201), whereas no statistically significant correlation was observed for anti-PD-1 IgG with anti-dsDNA (r = 0.0398, p = 0.9563), ANA (r = 0.0164, p = 0.8783) (Fig. 2), as well as C-reactive protein (CRP) and complement C3 and C4 levels (p > 0.05, data not shown). Furthermore, we classified SLEDAI scores into four equal groups (group I, SLEDAI = 0–4; group II, SLEDAI = 5–9; group III, SLEDAI = 10–14; group IV, SLEDAI > 14) and found that the serum levels of anti-PD-1 IgG in group I patients were significantly lower than levels in group II (p = 0.04) and IV (p = 0.001), but not group III (p = 0.151) (Additional file 1: Figure S1). Finally, we also analyzed the correlation between anti-PD-1 and other laboratory findings. As shown in Fig. 3, the serum titer of anti-PD-1 IgG is not significantly different between patients with seropositive and seronegative autoantibodies including anti-Sm, anti-U1RNP, anti-SSA, and anti-SSB. These findings indicate that the serum level of autoantibodies against PD-1 might be a potential marker for SLE disease activity.

Given that co-inhibitor PD-1/PD-L1 is crucial in regulating T cell activation [7], we therefore explored the effect of purified anti-PD-1 from SLE patients on T cell proliferation using a DC and T cell co-culture system. It is well known that PD-1 is expressed on activated T cells. We first examined the PD-1 levels on the surface of CD4⁺ T cells activated by anti-CD3 (0.5 ug/ml) for 3 d. The results showed that PD-1 is expressed after the activation of the T cells, while there was no PD-1 detected on the surface of unstimulated T cells (Fig. 4a, b), which is consistent with previous findings [24].

Next, we purified PD-1 autoantibodies from SLE patients with high levels of anti-PD-1 IgG and carried out a proliferation assay using a T and DC cell co-culture system. The results showed that, compared to the IgG control group, T cell proliferation was enhanced after blocking the PD-1 function with PD-1 autoantibodies from SLE patients and commercial monoclonal PD-1 antibody (Fig. 4c and d). These data indicate that self-reactive autoantibodies against co-inhibitor PD-1 may contribute to the dysregulated T cell hyperproliferation observed during SLE pathogenesis.