Interferon-β inhibits toll-like receptor 9 processing in multiple sclerosis

Abstract

Objective: Viral infections have been implicated in the pathogenesis of multiple sclerosis (MS). Plasmacytoid dendritic cells (pDCs) are present in peripheral blood, cerebrospinal fluid, and brain lesions of MS patients. pDCs sense viral DNA via Toll-like receptor 9 (TLR9), which has to be cleaved from the N-terminal to become functional (TLR9 processing). pDCs activated with TLR9 agonists promote T-helper type 1 (Th1)/T-helper type 17 (Th17) responses. In the animal model of MS, TLR9 agonists can induce disease. We hypothesized that pDCs are inhibited by disease-modifying therapy such as interferon (IFN)-β, consequently decreasing the frequency of MS attacks.

Methods: We separated pDCs from healthy subjects and patients diagnosed with relapsing-remitting MS and clinically isolated syndrome. Cytokine secretion by pDCs activated with TLR9 agonists was measured by enzyme-linked immunosorbent assay and multianalyte profiling. TLR9 gene and protein expression was studied by DNA microarrays and western blot.

Results: In untreated patients, pDCs activated with TLR9 agonists produced increased levels of IFN-α, a Th1-promoting cytokine, as compared to healthy subjects. In IFN-β-treated patients, activated pDCs had decreased ability to produce both IFN-α and the proinflammatory cytokines interleukin (IL)-6 and tumor necrosis factor α as compared to untreated patients. pDCs separated from IFN-β-treated patients had significantly reduced levels of the processed TLR9 protein but normal levels of the full-length TLR9 protein and TLR9 gene expression as compared to untreated patients.

Interpretation: This finding represents a novel immunomodulatory mechanism of IFN-β: inhibition of TLR9 processing. This results in decreased activation of pDCs by viral pathogens and, thus, may affect the frequency of MS exacerbations.

Figures

Figure 1

IFN-alpha secretion by activated pDCs. pDCs were separated from healthy donors (HD), untreated patients (MS: No Rx), and IFN-beta-treated RR MS patients (MS: IFN-beta). pDCs were activated in vitro with TLR9 agonist for 16 hours. IFN-alpha was measured in culture supernatants by ELISA, as described in Patients and Methods. IFN-alpha production was elevated in untreated patients (mean ± SEM = 8,479 ± 1,898 pg/ml, n = 11) as compared to healthy donors (3,659 ± 891 pg/ml, n = 17), p = 0.0164. MS patients treated with IFN-beta had decreased IFN-alpha production (2,826 ± 1,350 pg/ml, n = 12) as compared to untreated RR MS patients, p = 0.027. No detectable IFN-alpha was seen in control pDC cultures without TLR9 agonist. ELISA = enzyme-linked immunosorbent assay; SEM = standard error of the mean.

Figure 2

IFN-beta treatment inhibits IFN-alpha, IL-6, and TNF-alpha secretion by activated pDCs. pDCs were separated from the same MS/CIS patients before (pretreatment) and after a 3-month course of treatment (IFN-beta treatment) with recombinant human IFN-beta-1b (subcutaneous, n = 8) or recombinant human IFN-beta-1a (intramuscular, n = 6). The average age of patients was 36.3 years; 9 females and 5 males. pDCs were activated in vitro with TLR9 agonist for 16 hours. (A) IFN-alpha2, (B) IL-6, and (C) TNF-alpha were measured in the same cell supernatants by MILLIPLEX Multi-Analyte Profiling as described in Patients and Methods. IFN-beta-treated patients had decreased levels of IFN-alpha2 (mean ± SEM = 1,498 ± 372 pg/ml vs 2,917 ± 535; p = 0.0246), IL6 (33 ± 6 pg/ml vs 74 ± 11; p = 0.0012), and TNF-alpha (117 ± 17 pg/ml vs 317 ± 63; p = 0.0021) when compared to untreated patients. SEM = standard error of the mean.

Figure 3

TLR9 processing in pDCs and B cells. PBMCs, pDCs, and B cells were obtained from a healthy subject. The levels of the full-length TLR9 with the expected molecular weight of 115kDa (which may be higher due to protein glycosylation) and the C-terminal of TLR9 (processed TLR9) with the expected molecular weight of 65–80kDa were determined by western blot. The beta-actin level in the same sample was used for protein normalization. Note that more PBMCs and B cells were used for each experiment (1 × 106) than pDCs (1 × 105). The results are representative of 3 independent experiments with 3 different healthy subjects.

Figure 4

IFN-beta treatment inhibits TLR9 processing in MS. Plasmacytoid dendritic cells (pDCs) were separated from healthy donors (HD), nontreated MS/CIS patients (MS: No Rx), and MS/CIS patients treated with IFN-beta–based medications (MS: IFN-beta) for at least 3 months. The patient’s characteristics are presented in Table 2. The level of the (A) full-length and (B) processed TLR9 protein expression was determined in the same patients by western blot, as described in Patients and Methods. IFN-beta-treated CIS/MS patients had a similar level of the full-length TLR9 (mean ± SEM = 0.108 ± 0.025 relative units, n = 12) compared to untreated patients (0.112 ± 0.015 relative units, n = 14). However, IFN beta-treated patients had a significantly decreased level of the processed TLR9 C-terminal when compared to untreated patients (mean ± SEM = 0.924 ± 0.063 vs 1.445 ± 0.106 relative units, respectively; p = 0.0005). The level of the TLR9 C-terminal in healthy subjects was 1.189 ± 0.1385 relative units (n = 9), which was less than, but not statistically different from, untreated patients. SEM = standard error of the mean.