Etanercept decreases the innate immune wounding response in psoriasis

Abstract

Cathelicidin is increased when normal skin is injured and in psoriasis lesions where it has been suggested to play a pivotal role in inflammation through interactions with self-DNA and toll-like receptor 9 (TLR-9) in keratinocytes and plasmacytoid dendritic cells. Because of etanercept’s success in treating psoriasis, we hypothesized that etanercept may suppress TLR-9 and cathelicidin induction. Examination of experimentally induced wounds of psoriatic lesional and non-lesional skin, and comparison with wounded normal skin, shows that the induction of cathelicidin and TLR-9 is greatly enhanced in lesional psoriatic skin. Six weeks of etanercept appears not to affect the baseline expression of cathelicidin or TLR-9, but does blunt the induction of cathelicidin in psoriasis with wounding. These findings support the role of cathelicidin in the enhancement of local inflammation in psoriasis and may partially explain one of the mechanisms enabling TNF-α inhibitors to successfully treat this disorder.

Background

Cathelicidin is an antimicrobial peptide that is highly expressed in psoriatic lesional skin (1–3), and induced with wounding in normal skin (4,5). Cathelicidin’s role as a participant, rather than a bystander in psoriasis, was advanced with the discovery that the cathelicidin peptide LL-37 can complex with self-DNA to induce IFN-α in plasmacytoid dendritic cells (pDCs) through toll-like receptor-9 (TLR-9) (6,7). Because etanercept is so successful in treating psoriasis, and because it has been shown to suppress Th17 driving genes such as IL-12B, IL-23A and IL-1β, Th-17 cell products, and downstream molecules IL-17, and IL-22 (8–11), we hypothesized that etanercept would have a negative effect on TLR-9 and cathelicidin induction since IL-17 and TNF-α can induce keratinocytes to produce cathelicidin (9–11).

Questions addressed

1. Does the cathelicidin and TLR-9 response to wounding differ in psoriatic versus normal skin?

2. Does etanercept alters cathelicidin or TLR-9 expression before or after wounding?

Methods

Patients

Two millimetre punch biopsies were obtained from 10 normal subjects and from lesional and non-lesional skin of 10 psoriatic subjects. Three days later, a 4-mm punch biopsy was performed at the same site to assess the wounding response. A second cohort of 10 psoriatic subjects treated with etanercept also was biopsied to assess the wounding response, and in a subset of five subjects, a 2 mm pretreatment biopsy was also taken. This study was approved by the Human Research Protection Program at the University of California, San Diego, and performed in accordance with the Declaration of Helsinki. Samples were stored at −80°C until analysed by quantitative RT-PCR normalized to GAPDH (Data S1).

Statistics

Statistical analysis was performed using PRISM Software, San Diego, CA, USA. Results are expressed as the median of means of duplicate samples normalized with the mean of normal skin or as boxplot displays. P values comparing cathelicidin and TLR9 mRNA expression pre- and postetanercept treatment, and pre- and postwounding, were analysed by Wilcoxon signed rank test. P values comparing cathelicidin and TLR9 mRNA expression in normal, psoriasis non-lesional or psoriasis lesional skin were analysed by Mann–Whitney U test. P < 0.05 was considered significant (See Data S2).

Results

As shown previously in psoriasis, mRNA for the cathelicidin gene CAMP is significantly higher in lesional skin than non-lesional skin (Fig. 1a, P = 0.0002). In this study, normal subjects had slightly higher baseline levels than non-lesional psoriatic skin. This finding may be attributed to the small number of normals and the variance within our population (Fig. 1a). Examination of the wounding response revealed induction with wounding appeared most exaggerated in psoriatic lesional skin compared with non-lesional psoriatic skin (Fig. 1b, P = 0.04).

Figure 1.

Cathelicidin (a–c) and TLR-9 (d–f) mRNA expression. In all images, N, non-lesional skin; PS, psoriatic skin; W, postwounding; L, psoriatic lesional skin; E, postetanercept treatment for 6 weeks. (a) Cathelicidin mRNA expression at baseline. Lesional psoriatic skin showed a statistically significant increase in cathelicidin expression in comparison with non-lesional psoriatic skin P = 0.002. Sample size is as follows: Normal, n = 8; PS-N, n = 10; PS-L, n = 10. Bar shown is median. (b) Cathelicidin induction results. Cathelicidin induction after wounding is greatest in lesional psoriatic skin, P = 0.04. Data are represented in a boxplot display showing the median, quartiles and range. Sample sizes are as follows: Normal, n = 8; PS-N, n = 8; PS-L, n = 10. Bar shown is median. (c) Cathelicidin mRNA expression in psoriatic patients pre- and postetanercept treatment. No change in cathelicidin expression was noted after 6 weeks of etanercept. Sample sizes are as follows: PS-N, n = 5; PS-N-E, n = 5; PS-L, n = 5; PS-L-E, n = 4. Bar shown is median. (d) Baseline TLR-9 mRNA expression. Sample sizes are as follows: Normal, n = 10; PS-N, non-lesional psoriatic skin n = 10; PS-L, n = 10. No significant differences were noted between the three groups. Bar shown is median. (e) TLR-9 induction after wounding. TLR-9 induction after wounding is greatest in lesional psoriatic skin, P = 0.05. Data are represented in a boxplot display showing the median, quartiles and range. Sample sizes are as follows: Normal n = 10, PS-N, n = 8; PS-L, n = 10. Bar shown is median. (f) TLR-9 expression pre- and postetanercept treatment. There was no significant difference in TLR-9 expression pre- or post-treatment with etanercept. Sample sizes are as follows: PS-N, n = 5; PS-N-E, n = 5; PS-L, n = 5; PS-L-E, n = 5. Bar shown is median.

As anti-TNF therapy has been shown to down-regulate pro-inflammatory cytokine genes IL-17, IFN-γ, IL-1β and IL-8 in psoriatic plaques (12), and IL-17 and TNF-α have been shown to induce keratinocytes to produce cathelicidin (9,10), we next examined the effect of etanercept on cathelicidin and TLR-9 expression. After 6 weeks of etanercept therapy, there was no change in lesional or non-lesional expression of cathelicidin or TLR-9 (Fig. 1c).

TLR-9 mRNA expression was similar prior to wounding and increased with wounding in all cases, but as with cathelicidin, this change appeared most exaggerated in lesional compared with non-lesional skin (Fig. 1d, e, P = 0.05). As with cathelicidin, there was no change in lesional or non-lesional expression of TLR-9 after etanercept therapy (Fig. 1f).

After 6 weeks of etanercept therapy, cathelicidin induction by wounding was maintained in non-lesional skin (Fig. 2a), but no longer occurred in lesional skin (Fig. 2b). The expression of TLR-9 after wounding did not appear to be affected by etanercept (Fig. 2c, d). However, comparison of the difference between wounded and non-wounded skin before and after etanercept showed a trend towards blunting of the rise in TLR-9 in lesional skin (P = 0.07, Fig. 2e).

Figure 2.

Cathelicidin (a, b) and TLR-9 (c–e) mRNA levels in control psoriasis patients and in psoriasis patients who have had 6 weeks of etanercept. In all images, N = non-lesional skin, PS = psoriatic skin, W = postwounding, L = psoriatic lesional skin, E = postetanercept treatment for 6 weeks, Psor. Control = control psoriasis group not receiving treatment, Psor. ET = psoriatic patients after 6 weeks of etanercept therapy. (a) Cathelicidin mRNA levels pre- and postwounding in non-lesional skin psoriatic control and treatment groups. The wounding response of non-lesional psoriatic skin did not change after 6 weeks of etanercept, P = 0.002, P = 0.002. Sample sizes are as follows: PS-N, n = 10, PS-N-W, n = 10; PS-N-E, n = 10; PS-N-E-W, n = 10. Bar shown is median. (b) Cathelicidin mRNA expression pre- and postwounding in psoriatic lesional skin in the psoriasis control and treatment groups. Induction of cathelicidin after wounding was blunted in lesional psoriatic skin after 6 weeks of etanercept, P = 0.02, P = 0.098, respectively. Sample sizes are as follows: PS-L, n = 10; PS-L-W, n = 10; PS-L-E, n = 9; PS-L-E-W, n = 9. Bar shown is median. (c, d) TLR-9 mRNA expression in normal or lesional skin of psoriatic patients either not being treated for psoriasis or after receiving 6 weeks of etanercept therapy pre- and postwounding. Etanercept did not block the wounding response of TLR-9 in lesional or non-lesional skin of psoriatics. TLR-9 was induced with wounding both pre- and postetanercept in non-lesional and lesional skin of psoriatics, P = 0.02, P = 0.004, P = 0.002 and P = 0.004, respectively. Sample sizes are as follows: PS-N-E, n = 10, PS-N-E-W, n = 10, PS-L-E, n = 10, PS-L-E-W, n = 9. Bar shown is median. (e) Boxplot display of median, quartiles and range of the difference in TLR-9 mRNA expression in normal skin and psoriatic lesional and non-lesional skin pre- and post-treatment with 6 weeks of etanercept. Etanercept did not block but did decrease the induction of TLR-9 after wounding in lesional but not non-lesional skin of psoriatics. There was a statistically significant difference in TLR-9 induction in psoriatic lesional skin in comparison with psoriatic non-lesional skin, P = 0.05. Although not statistically significant, 6 weeks of etanercept appeared to blunt this effect in lesional but not non-lesional psoriatic skin, P = 0.07. Bar shown is median.

Conclusion

In this study, cathelicidin and TLR-9 are both induced with wounding in lesional and non-lesional psoriatic skin but more exaggerated in lesional skin. We postulate that the increase in cathelicidin in lesional skin may contribute in the pathogenesis of psoriasis, possibly enhancing the formation of cathelicidin (LL-37) self-DNA complexes through TLR-9 to induce the production of type 1 IFN by pDCs (6) and keratinocytes (13). This enhanced increase in cathelicidin and TLR-9 after injury of psoriatic skin may therefore be contributing to the perpetuation of inflammation in psoriatic lesions.

Our finding with etanercept suggests that although etanercept down-regulates the pro-inflammatory cytokine genes IL-17 and IFN-γ, there is no direct effect on cathelicidin or TLR-9 production in the absence of wounding (8). The blunting of TLR-9 with wounding may be related to a decreased cathelicidin production by etanercept as shown previously (14). This finding does conflict with our current data that did not detect a decrease in baseline cathelicidin with etanercept therapy, but this may relate to the selection of sites for biopsy after therapy. We propose that the dynamic changes in expression of these genes in response to acute injury are an important new aspect to consider when evaluating the pathogenesis of psoriatic lesions.

In this study, we have shown for the first time in vivo that cathelicidin and TLR-9 are both induced with wounding in psoriatic subjects. Six weeks of etanercept blunted the wounding response for cathelicidin, and to a lesser (but not statistically significant) degree, TLR-9 in lesional skin. This finding supports the role that wounding may play in psoriasis induction and suggests that one factor affecting the efficacy of etanercept may be the blunting of this wounding response. Further studies with larger sample sizes, immunohistochemistry and qRT-PCR data are necessary to confirm this novel preliminary finding.