Background/Purpose: Tumor necrosis factor-a (TNF), a potent proinflammatory cytokine, activates two receptors: TNFR1 and TNFR2. Anti-TNF biological agents neutralize TNF, thus preventing both TNFR1 and TNFR2 activation. These anti-TNF agents are used in treating inflammatory diseases including psoriasis and psoriatic arthritis. However, potentially life threatening adverse effects that include reactivation of tuberculosis and malignancies are associated with their long-term use. The adverse effects are associated with inhibition of TNFR1, not TNFR2, signaling. It is not clear whether inactivation of both TNF receptors is necessary for inhibiting psoriatic diseases. Here, we used a mouse model to identify the contributions of TNFR1 and TNFR2 in psoriasis.

Methods: We applied imiquimod (IMQ) (4 mg/day/mouse for 4 days) to 10-12-week-old wild type (WT), TNFR1 knockout (KO), and TNFR2KO mice (n = 6 each). Erythema, scaling, and thickness of the IMQ- or vehicle-treated area was assessed by PASI (Psoriasis Area and Severity Index). Epidermal thickness was measured using ImagePro software on H&E stained skin. We compared the psoriasis-related gene expression in the untreated vs. IMQ-treated skin using qPCR and immunohistochemistry. Fluorescence-activated cell sorting (FACS) was used to compare immune cell expansion in the lymph nodes.

Results: In WT, the cumulative score of skin inflammatory phenotype (erythema + scaling + thickness) peaked at day 5 with a cumulative score of 7.0 ±1.3. TNFR2 knock-out (TNFR2KO) did not show any sign of skin inflammation. TNFR1KO mice exhibited low level of inflammation with a cumulative score of 2.5±0.8. Histology revealed significant keratinocyte hyperplasia and leukocyte infiltration in the WT upon IMQ application. These inflammatory responses were significantly blunted in TNFR2KO, but not in TNFR1KO, mice. As expected, the IL-17 family of genes and genes encoding IL-22 and IL-23 are upregulated in IMQ-treated skin compared to untreated WT skin. Compared to WT mice, the IMQ induction of these genes was reduced significantly in TNFR2KO mice. Fold increases in mRNA levels upon IMQ:  IL-17A (WT, 22.3±4.1 vs TNFR2KO, 15.6±2.3), IL-17C (WT, 31±3.3 vs TNFR2KO,11.2±2.2), IL-17F (WT, 51.8±3.8 vs. TNFR2KO, 8.9±1.7), IL-22 (WT, 33.6±3.3 vs TNFR2KO, 9.4±1.2) and IL-23 (WT, 101.8±12.5 vs TNFR2KO, 14.2±3.2).  The effect of TNFR1 inactivation on IMQ induction of these genes was minimal. Importantly, our FACS analysis showed IMQ-induced CD11c+ and CD11b+ dendritic cell expansion, a critical component in T_H na•ve cell to T_H17 differentiation,  is inhibited in TNFR2KO, but not in TNFR1KO, mice. Immunohistochemistry studies showed reduced number of infiltrated T_H17 cells upon IMQ in TNFR2KO skin compared with WT or TNRIKO mice.

Conclusion: Psoriatic inflammation in mice is critically dependent on TNFR2 via T_H17 cell expansion and IL-23/IL-17 induction. Selective inhibition of TNFR2 while retaining TNFR1 may block the psoriasis and psoriatic arthritis, as both share many common inflammatory pathways, and may lead to improved safety compared to the current anti-TNF therapy.