Tumour necrosis factor-alpha blockade suppresses murine allergic airways inflammation

Abstract

Asthma is a heterogeneous disease that has been increasing in incidence throughout western societies and cytokines, including proinflammatory tumour necrosis factor alpha (TNF-alpha), have been implicated in the pathogenesis of asthma. Anti-TNF-alpha therapies have been established successfully in the clinic for diseases such as rheumatoid arthritis and Crohn's disease. TNF-alpha-blocking strategies are now being trialled in asthma; however, their mode of action is poorly understood. Based on the observation that TNF-alpha induces lymph node hypertrophy we have attempted to investigate this as a mechanism of action of TNF-alpha in airway inflammation by employing two models of murine airway inflammation, that we have termed short and long models, representing severe and mild/moderate asthma, respectively. The models differ by their immunization schedules. In the short model, characterized by eosinophilic and neutrophilic airway inflammation the effect of TNF-alpha blockade was a reduction in draining lymph node (DLN) hypertrophy, eosinophilia, interleukin (IL)-5 production and immunoglobulin E (IgE) production. In the long model, characterized by eosinophilic inflammation, TNF-alpha blockade produced a reduction in DLN hypertrophy and IL-5 production but had limited effects on eosinophilia and IgE production. These results indicate that anti-TNF-alpha can suppress DLN hypertrophy and decrease airway inflammation. Further investigations showed that anti-TNF-alpha-induced inhibition of DLN hypertrophy cannot be explained by preventing l-selectin-dependent capture of lymphocytes into the DLN. Given that overall TNF blockade was able to suppress the short model (severe) more effectively than the long model (mild/moderate), the results suggest that TNF-alpha blocking therapies may be more effective in the treatment of severe asthma.

Fig. 1

Ovalbumin (OVA)-specific (3 × 10⁶) T cells were transferred adoptively into naive Balb/c recipients, which were then immunized with 100 μg OVA and 1% alum by subcutaneous (s.c.) injection. Ten days later, mice were challenged by intranasal administration of 50 μg OVA plus 2·5 μg lipopolysaccharide. Mice were treated throughout with 50 μg tumour necrosis factor (TNF) receptor p80 Fc fusion protein (sTNFR:Fc) (○) by s.c. injection every 2 days. Control mice were given 50 μg human IgG (▪) instead of sTNFR:Fc (a). Bronchoalveolar lavage (BAL) fluid was sampled and eosinophil (b) and neutrophil (c) content was analysed by cytospin preparation. BAL fluid was sampled and analysed for interleukin-5 production (d). Serum from the same mice was analysed for OVA-specific immunoglobulin E (e). Results are represented as mean ± standard error of the mean (*P < 0·05), n = 6 per group.

Fig. 2

Ovalbumin (OVA)-specific (3 × 10⁶) T cells were transferred adoptively into naive Balb/c recipients, which were then immunized with 100 μg OVA and 1% alum by intraperitoneal injection. Ten days later, mice were challenged by intranasal administration of 50 μg OVA plus 2·5 μg lipopolysaccharide. Mice were treated throughout with 50 μg tumour necrosis factor (TNF) receptor p80 Fc fusion protein (sTNFR:Fc) (○) by subcutaneous (s.c.) injection every 2 days. Control mice were given 50 μg human immunoglobulin (▪) instead of sTNFR:Fc. Draining and non-draining lymph nodes were removed and weighed (a, b), and cellularity was determined (c, d). Results are represented as mean ± standard error of the mean (*P < 0·05), n = 6 per group.

Fig. 3

Ovalbumin (OVA)-specific (3 × 10⁶) T cells were transferred adoptively into naive Balb/c recipients, which were then immunized with 100 μg OVA and 1% alum by subcutaneous (s.c.) injection on three occasions, each a week apart. One week after the final immunization, mice were challenged by intranasal administration of 50 μg OVA. Mice were treated throughout with 50 μg tumour necrosis factor (TNF) receptor p80 Fc fusion protein (sTNFR:Fc) (○ or white bars) by s.c. injection every 2 days. Control mice were given 50 μg human immunoglobulin (▪ or black bars) instead of sTNFR:Fc (a). Bronchoalveolar lavage (BAL) fluid was sampled and eosinophil content was analysed by cytospin preparation (b). Lung sections were scored for inflammation (c). Solid line represents baseline pathology. BAL fluid was sampled and analysed for interleukin-5 production (d). Serum from the same mice was analysed for OVA-specific immunoglobulin E (e). Lung sections were stained with haematoxylin and eosin. Sections show sTNFR:Fc- (f, g) and IgG (h, i)-treated mice at day 4 post-airways challenge at 10× and 20× magnification. Results are represented as mean ± standard error of the mean (*P < 0·05), n = 3–6 per group.

Fig. 4

Ovalbumin (OVA)-specific (3 × 10⁶) T cells were transferred adoptively into naive Balb/c recipients, which were then immunized with 100 μg OVA and 1% alum by subcutaneous (s.c.) injection on three occasions, each a week apart. One week after the final immunization, mice were challenged by intranasal administration of 50 μg OVA. Mice were treated throughout with 50 μg tumour necrosis factor (TNF) receptor p80 Fc fusion protein (sTNFR:Fc) (○) by s.c. injection every 2 days. Control mice were given 50 μg human IgG (▪) instead of sTNFR:Fc. Draining and non-draining lymph nodes were removed and weighed (a, b) and cellularity was determined (c, d). Results are represented as mean ± standard error of the mean (*P < 0·05), n = 6 per group.

Fig. 5

Ovalbumin (OVA)-specific (3 × 10⁶) T cells were transferred adoptively into naive Balb/c recipients, which were then immunized with 100 μg OVA and 1% alum by subcutaneous (s.c.) injection. Ten days later, mice were challenged by intranasal administration of 50 μg OVA plus 2·5 μg lipopolysaccharide. Mice were treated throughout with 50 μg tumour necrosis factor (TNF) receptor p80 Fc fusion protein (sTNFR:Fc) (black) by s.c. injection every 2 days. Control mice were given 50 μg human immunoglobulin (white) instead of anti-TNF-α. Draining lymph nodes (a), lung tissue (b) and non-draining lymph nodes (c) were removed and l-selectin low expression (effector/memory) by OVA-specific T cell numbers were determined by flow cytometry. Results are represented as mean ± standard error of the mean (*P < 0·05), n = 3 per group.