RNA recognition by human TLR8 can lead to autoimmune inflammation

Abstract

Studies on the role of the RNA receptor TLR8 in inflammation have been limited by its different function in human versus rodents. We have generated multiple lines of transgenic mice expressing different levels of human TLR8. The high copy number chimeras were unable to pass germline; developed severe inflammation targeting the pancreas, salivary glands, and joints; and the severity of the specific phenotypes closely correlated with the huTLR8 expression levels. Mice with relatively low expression levels survived and bred successfully but had increased susceptibility to collagen-induced arthritis, and the levels of huTLR8 correlated with proinflammatory cytokines in the joints of the animals. At the cellular level, huTLR8 signaling exerted a DC-intrinsic effect leading to up-regulation of co-stimulatory molecules and subsequent T cell activation. A pathogenic role for TLR8 in human diseases was suggested by its increased expression in patients with systemic arthritis and the correlation of TLR8 expression with the elevation of IL-1β levels and disease status. We found that the consequence of self-recognition via TLR8 results in a constellation of diseases, strikingly distinct from those related to TLR7 signaling, and points to specific inflammatory diseases that may benefit from inhibition of TLR8 in humans.

Figure 1.

huTLR8 expression level strictly correlate with decrease survival in chimeric mice. (A) Mice were generated with a varying number of copies of huTLR8. Mice were from clone 6 (CL; n = 14), clone 23 (n = 8), clone 12 (n = 44), and clone 8 (n = 18). (B) Data from spleens was collected from moribund animals and TaqMan was used to evaluate huTLR8 expression. huTLR8 expression was plotted against survival time. 26 spleens (n = 22 from clone 12; n = 4 from clone 23) are shown. (C) PBMCs from huTLR8Tg chimeras were assessed by flow cytometry for expression of H2b (C57BL/6; expressing the transgene) or H2d haplotypes (BALB WT). From 25 mice (n = 21 from clone 12; n = 3 from clone 23; n = 1 from clone 6). (D) Human TLR8 expression level in blood from huTLR8Tg chimeras (total number of mice 26; n = 21 from clone 12; n = 3 from clone 23; n = 2 from clone 6) and from TLR8TgCL8 (n = 36) was evaluated by TAQMAN. Data are expressed as fold compared with huTLR8 expression level in clone 8 (E). 5 × 10⁵ PBMCs from huTLR8Tg chimeras or C57BL/6 WT animals were stimulated with a specific RNA-based TLR8 agonist, ORN-8L, and 24 h later supernatants were harvested and assayed for TNF levels by ELISA; data are cumulative of n = 4 mice (mean ± SEM), 2 mice from TLR8TgCL12, 1 mice from TLR8TgCL6 and 1 mice from TLR8TgCL23. Data are from a representative experiment out of three independent experiments. (F and G) 5 × 10⁵ PBMCs from human healthy donors (F; n = 3 donors; mean ± SEM), from TLR8TgCL12 chimeras, or from WT mice were stimulated with ORN-8L or LPS with or without bafilomycin A1 for 18 h. Stimulation with TLR4 ligand LPS was used as control (G; n = 4 mice; mean ± SEM).

Figure 2.

Multiorgan inflammation in huTLR8 chimera mice. (A) Organs from WT (II, IV, VI, VIII) or chimeric huTLR8Tg (I, III, V, VII) were stained with hematoxylin and eosin. Pancreas (I, II), kidney pelvis (III and IV), kidney (V and VI), and liver (VII and VIII). Arrows indicates normal glomerular structure in kidney sections from WT animals (VI) and evident glomerular abnormalities in kidney from huTLR8Tg animals (V). Sections of liver are represented in panel VII (huTLR8Tg) and panel VIII (WT). Additional pathology details in Fig. S2. Scale bar is 100 µm for all panels except V and VI which is 50 µm. (B) Cumulative data of the pathological score in the affected organs. n = 28 (n = 18 from clone 12; n = 4 from clone 23 and n = 6 from clone 6) and 14 for WT C57BL/6; (mean ± SEM). (C) Cytokines concentration in sera from TLR8TgCL12 (n = 15) and CTRL WT C57BL/6 mice (n = 18) was measured by Milliplex assay. (D) Pancreases from TLR8TgCL12 or CTRL WT C57BL/6 mice (total number of mice = 6/group) were harvested 40–60 d after birth, and the levels of proinflammatory genes were evaluated by TaqMan. Data are expressed as fold increase compared with CTRL mice (mean ± SEM). (E) huTLR8Tg animals were sacrificed when moribund (age ranged from 44 to 140 d). The titers of double-stranded DNA specific antibody (anti-ds DNA) and ribonucleoprotein (anti-RNP) specific antibody in the serum of TLR8TgCL12 chimeras (n = 27) and of age matched WT animals C57BL/6 (n = 37) were determined. *, P ≥ 0.05; **, P ≥ 0.01; ***, P ≥ 0.001.

Figure 3.

Pathology induced by huTLR8 expression is due to its activation in the hematopoietic compartment. (A) B6.SJLmice were transplanted with bone marrow from huTLR8Tg chimeras and monitored for survival. n = 45 for TLR8TgCL12>B6.SJL; n = 12 for TLR8TgCL6>B6.SJL; n = 12 for TLR8TgCL23>B6.SJL; and n = 12 for C57BL/6>B6.SJL. Data are cumulative of two independent experiments. (B) Organs from mice in A were harvested when mice were moribund. TLR8TgCL12>B6.SJL n = 19; n = 3 for TLR8TgCL6>B6.SJL; n = 3 for TLR8TgCL23>B6.SJL; and n = 9 for C57BL/6>B6.SJL (mean ± SEM). (C) Cellular subsets were isolated from spleens of B6.SJLmice transplanted with bone marrow from TLR8TgCL12 chimeras. Gene expression of huTLR8 was evaluated by TaqMan. Cumulative data from three independent experiments; n = 5–10. Mean ± SEM.

Figure 4.

High expression level of huTLR8 leads to the intrinsic activation of DCs in mice. (A) Peripheral blood and spleens were harvested from high expressing huTLR8Tg chimeras (n = 14–17) or age-matched C57BL/6 (n = 17) mice, and cell numbers were assessed. (B) Absolute white cell numbers in blood of chimera mice from TLR8TgCL12 (n = 14) and WT C57BL/6 (n = 18). A and B show cumulative data of four independent experiments. (C) Spleens and peripheral blood cells from high expressing huTLR8Tg chimeras or WT controls were harvested, and co-stimulatory molecule expression was analyzed on CD11c⁺ cells in spleens (top) and blood (middle) and splenic CD11b⁺ cells (bottom). CD11c were gated based on the expression of MHC-class I. CD11b⁺ cells were gated on CD11c^neg cells and on the expression of MHC class I. H2b (C57BL/6)-positive cells expressing huTLR8 (black line) were compared with H2b-negative cells (BALB/c, red line) not expressing huTLR8 in the same animal. Blue line, WT animal. CD11b^pos CD11c^neg monocytes in the spleen were gated on the expression of MHC-class I. H2b (C57BL/6)-positive cells expressing huTLR8 (black line) are compared with H2b-negative cells (BALB/c, red line) not expressing huTLR8. Plots are representative of at least three independent experiments. (D) Expression of co-stimulatory molecules in CD11c from spleens of TLR8TgCL12>B6.SJL (black line) or C57BL/6 (red line) chimeric mice. Plots are representative of at least three independent experiments. (E) CD11c cells were purified from spleens of TLR8TgCL12>B6.SJL (TLR8DC) or C57BL/6>B6.SJL (WTDC) chimeric mice. The CD11c⁺ cells were loaded with Ova_323–339 and used to stimulate OT-II T cells at the indicated DC:T cells ratio. One representative experiment of two is shown. *, P ≥ 0.05; **, P ≥ 0.01.

Figure 5.

T cell compartment is highly activated in huTLR8Tg mice. (A and B) Splenocytes from either high expressing TLR8TgCL12 chimeras in which at least 95% of the hematopoietic compartment was H2b^pos (C57BL/6) or C57BL/6 WT mice were stimulated in vitro for 2 h with PMA and ionomycin. Percentages of CD4 and CD8 T cells producing TNF and IFN-γ were assessed by flow cytometry. (A) Representative dot plots. (B) Quantitation of data in A from one representative of three independent experiments (n = 3 mice; mean ± SEM). (C and D) TLR8TgCL12 chimeras were sacrificed and spleens were harvested. (C) CD3⁺CD4⁺ and (D) CD3⁺CD8⁺ T cells were gated based on the expression of MHC class I. (D) Right panel shows representative dot plot of CD3⁺CD4⁺ and CD3⁺CD8⁺ from C57BL/6 WT mice from three independent experiments. (E) Quantitative intramouse data showing naive, effector memory, and effector T cells population for both haplotypes (huTLR8Tg in red, WT in blue). **, P ≥ 0.01

Figure 6.

Human TLR8 can function in absence of mouse TLR7, but it regulates its expression levels. (A) 5 × 10⁵ PBMCs from TLR7^+/+ TLR8TgCL8 line and from TLR7^−/− TLR8TgCL8 were stimulated with a specific RNA-based TLR8 agonist, ORN-8L, supernatants were harvested at 24 h and assayed for cytokines levels by ELISA; One representative experiment of three is shown. n = 4; mean ± SEM. (B) Mouse TLR7 expression was evaluated by TaqMan in spleens of TLR8TgCL12 (n = 22) and TLR8TgCL23 (n = 4) chimeras and WT C57BL/6 (n = 30; mean ± SEM). (C) Cellular subsets were isolated from spleens of TLR8TgCL12>B6.SJL mice or with C57BL/6>B6.SJL, as described in the Materials and methods. Gene expression of mouse TLR7 was evaluated by TaqMan. (B and C) Cumulative data from three independent experiments; n = 5–10; mean ± SEM; ***, P ≥ 0.001.

Figure 7.

Spontaneous arthritis develops in high expressing huTLR8Tg mice. (A) Chimera mice from TLR8TgCL12 (n = 6) and C57BL/6 CTLR animals (n = 6) were euthanized 90 d after the birth, and paws and joints were fixed, sectioned, and stained with toluene blue. Representative sections of paws from WT mice (i) and TLR8TgCL12 mice (iii and v) and joints from WT (ii) and TLR8TgCL12 mice (iv and vi). (B) The histological changes, degree of inflammation, and cartilage damage were evaluated by a pathologist in a blinded fashion and scored 1–5 as followed: 1, minimal; 2, mild; 3, moderate; 4, marked; 5, severe. Two paws and two ankles were evaluated for each animal and the scores were summed to obtain the histological disease score. ***, P ≥ 0.001. Bar, 300 mm.

Figure 8.

huTLR8Tg mice have increased susceptibility to arthritis. (A) TLR8TgCL8 line and age-matched WT controls (left) were immunized with collagen and disease score was measured. Total arthritis score (left) and percentage of mice with clinical score equal or more than four (right) in the two strains is shown. One representative experiment (n = 10 mice per group) of 4 is shown (40 mice total per group). Statistical significance was evaluated with a two-way ANOVA test. (B– D) Representative sections (hematoxylin and eosin stained) of normal (B), WT (C), and TLR8TgCL8 (D) joints, sacrificed 80 d after CIA induction. (E) Cumulative data from histopathological evaluation in B–D. Animals were sacrificed at day 80 after CIA induction and joints of hind paws were fixed, sectioned, and stained with H&E. The four histological changes (inflammation, pannus formation, cartilage, and bone damage) were evaluated by a pathologist in a blinded fashion as described in the Materials and methods. Table shows cumulative data from 2 independent experiments, TLR8TgCL8 (n = 16) or WT (n = 15); mean ± SEM; a = P ≤ 0.05. (F) Expression of F4/80 and TNF in the joints of TLR8TgCL8. The two front joints of each animal (WT or huTLR8Tg) were used to prepare RNA and F4/80, and TNF expression was evaluated by TaqMan. Animals were sacrificed 80 d after CIA induction. Number of joints 46 per group (mean ± SEM); data are cumulative of two independent experiments. P-values were calculated using a nonparametric correlation test (Spearman). *, P ≥ 0.05; ***, P ≥ 0.001. Bar, 100 µm.

Figure 9.

TLR8 signaling induces a discrete panel of proinflammatory genes in humans. (A) Cellular subsets were isolated from whole blood of healthy donors and expression of TLR7, 8, and 9 was analyzed by TaqMan assay. Relative Ct of the genes is shown in the table. Cumulative data from at least four independent donors is shown (mean relative CT ± SEM). (B and C) 4 × 10⁵ PBMCs from four healthy donors were stimulated for 6 h with or without TLR7L (CL264; 5 µg/ml), TLR8L (ORN8L; 200 µg/ml), and TLR9L (C274; 0.3 µM). Stimulated samples from each donor were normalized to their own unstimulated control (in medium only). Transcripts over- and underexpressed at least twofold were selected. In B, representative genes are grouped by family; mean fold up-regulation is shown for each ligand. In C, Venn diagram depicts the overlap between TLR7, 8, and 9 ligands for all genes differentially expressed in the three conditions. (C) List of genes specifically up-regulated by TLR8 ligation (right), TLR7 ligation (left), and TLR9 ligation (middle). Notably, a class of TLR8-specific genes (highlighted in orange) belongs to a family coding for highly conserved RNA molecules component of RNP, which constitute potent TLR8 activators (Vollmer et al., 2005).

Figure 10.

The TLR8 pathway is dysregulated in systemic arthritis patients. (A) Expression of TLR8 in the blood of systemic-onset juvenile arthritis patients (SoJIA; two cohorts; left; n = 33) and Still’s disease patients (right; n = 31). The level of huTLR8 in each of the cohorts was normalized to the level of the gene in healthy volunteer cohorts (HV). Cohort 2 of the SoJIA patients was treated with Anakinra (IL-1 receptor antagonist), and levels of huTLR8 gene expression was compared before and after treatment (red dots). (B) Expression levels of TLR8 were plotted against blood levels of IL-1β of SoJIA (cohort 1 and 2 combined) and Still’s disease. (C) Similarly to C levels of TLR7 and TLR9 are shown for SoJIA (cohort 1 and 2 combined). In B and C, p-values were calculated using a nonparametric correlation test (Spearman). (D) TLR8 expression levels in purified monocytes from SoJIA patients with active disease. Human PBMCs (E) or mouse PBMCs from huTLR8TgCL8 mice (F) were cultured for 16 h with 1 ng or 5 ng/ml (for human PBMCs) or 5 ng/ml (mouse PBMCs) of the indicated cytokines, and levels of huTLR8 expression were assessed by TaqMan. The experiments were conducted twice with a total of eight human donors (E) and nine mice (F). Shown is mean ± SEM. *, P ≥ 0.05; **, P ≥ 0.01; ***, P ≥ 0.001.