Pharmacological TLR4 Antagonism Using Topical Resatorvid Blocks Solar UV-Induced Skin Tumorigenesis in SKH-1 Mice

Abstract

An urgent need exists for the development of more efficacious molecular strategies targeting nonmelanoma skin cancer (NMSC), the most common malignancy worldwide. Inflammatory signaling downstream of Toll-like receptor 4 (TLR4) has been implicated in several forms of tumorigenesis, yet its role in solar UV-induced skin carcinogenesis remains undefined. We have previously shown in keratinocyte cell culture and SKH-1 mouse epidermis that topical application of the specific TLR4 antagonist resatorvid (TAK-242) blocks acute UV-induced AP-1 and NF-κB signaling, associated with downregulation of inflammatory mediators and MAP kinase phosphorylation. We therefore explored TLR4 as a novel target for chemoprevention of UV-induced NMSC. We selected the clinical TLR4 antagonist resatorvid based upon target specificity, potency, and physicochemical properties. Here, we confirm using ex vivo permeability assays that topical resatorvid can be effectively delivered to skin, and using in vivo studies that topical resatorvid can block UV-induced AP-1 activation in mouse epidermis. We also report that in a UV-induced skin tumorigenesis model, topical resatorvid displays potent photochemopreventive activity, significantly suppressing tumor area and multiplicity. Tumors harvested from resatorvid-treated mice display reduced activity of UV-associated signaling pathways and a corresponding increase in apoptosis compared with tumors from control animals. Further mechanistic insight on resatorvid-based photochemoprevention was obtained from unsupervised hierarchical clustering analysis of protein readouts via reverse-phase protein microarray revealing a significant attenuation of key UV-induced proteomic changes by resatorvid in chronically treated high-risk SKH-1 skin prior to tumorigenesis. Taken together, our data identify TLR4 as a novel molecular target for topical photochemoprevention of NMSC. Cancer Prev Res; 11(5); 265-78. ©2018 AACRSee related editorial by Sfanos, p. 251.

Figure 1. Topical pharmacokinetics of resatorvid enable the effective modulation of UV-induced biomarkers and AP-1 luciferase activity in mouse skin

Resatorvid (chemical structure as shown) dissolved in acetone was applied to ex-vivo SKH-1 mouse skin on the upper chamber of a Franz cell apparatus and transdermal penetration was quantified over time (n = 3). For quantification of total cutaneous resatorvid delivery (after removal of stratum corneum from live skin), epidermal and dermal drug contents were analyzed separately and combined.(A). The UV stability of resatorvid in acetone or water was examined in UV-permeable glass vials exposed to one or two doses of SSL (50 kJ UVA/m² and 2.4 kJ UVB/m²; dose 1×), followed by quantitative HPLC (B). Chemical stability of resatorvid in aqueous solutions of increasing pH and in acetone was examined (64 days, 25°C) (C). The ability of resatorvid to block UV-induced stress signaling in vivo was examined using transgenic SKH-1 AP-1 luciferase mouse models (luciferase expression under control of a 4×TPA-response element). The ears of the mice were treated with acetone (vehicle) or 10 mM resatorvid 24 hr and 1 hour prior to acute UVB. Mice were sacrificed 48 hr later and fold induction was determined by dividing the post-UV luciferase activity by the pre-UV luciferase activity of ear punches from each mouse. N.S.: not significant (D). Epidermal lysates from SKH-1 back skins post-treated with 14 mM resatorvid after acute SSL exposure were examined via western blot (p38 MAPK and p65 subunit of NFκB phosphorylation), quantified using Image J software (loading control: β-actin) (E, F). * : p < 0.05.

Figure 2. Resatorvid modulates UV responses in chronically exposed mouse epidermis

SKH-1 female mice were exposed to SSL three times weekly for 15 weeks, during which time they were topically treated with either vehicle (acetone) or resatorvid (10 mM, 1 hr prior to SSL). After week 15, SSL treatments stopped, but topical acetone (Control group), or resatorvid (Prevention group) continued. Another cohort (Intervention group) received topical acetone during the SSL exposures, but switched to topical resatorvid after completion of the SSL regimen until the end of the experiment (week 25). To examine the effects of UV and resatorvid in the chronic setting but before visible tumorigenesis (high-risk skin), mice (n = 4) were sacrificed at week 14 and compared to mice treated with agent but not UV (n =5; dropout mice). The remaining mice (n = 20/group) were maintained on the protocol for the tumorigenesis study (A). Chronically exposed skin was examined for epidermal immune infiltration using H&E stained tissue (400×) (representative images, left panel). Skin was treated with vehicle (acetone) for 14 weeks (I), SSL + vehicle (II, arrows indicate lymphocytic infiltrates), or SSL + resatorvid (III). Histopathological quantification (cells/field) and morphological assignment of lymphocytic infiltrates was performed by a pathologist (right panel) (B). Epidermal lysates from the chronically-exposed dropout mice were subjected to proteomic RPPA analysis, generating an unsupervised two-way hierarchical clustering heatmap of protein/phosphoprotein expression. Each group clustered as predicted (top to bottom): Acetone No UV, Resatorvid No UV, Acetone +UV and Resatorvid +UV (C). Selected RPPA biomarkers were interrogated for resatorvid-specific inhibition of chronic UV-induced phosphorylation/expression [Stat3 (Y705), PDK1 (S241), IL-6, and p62/SQSTM1] (D). RPPA identification of a cluster of proteins (yellow box highlighted in C; partially annotated) significantly suppressed by resatorvid irrespective of chronic UV exposure, e.g. NF-κB (p65 S536) and p53 (S15) (E). * : p < 0.05. *** : p < 0.001.

Figure 3. Topical resatorvid blocks UV-induced skin tumorigenesis in SKH-1 mice

Female SKH-1 mice were exposed to SSL three times a week for 15 weeks as described (Fig. 2A). Chronic exposure to topical resatorvid did not cause overt toxicity, as evidenced by similar weight gain between all treatment groups (A). The Resatorvid Prevention group had significantly increased tumor latency compared to acetone controls as revealed by Kaplan-Meier survival analysis (B). At sacrifice (week 25) the Resatorvid Prevention group displayed a significant reduction in average tumor burden versus the Acetone control group; the Resatorvid Intervention group displayed reduced tumor burden that was not statistically significant (C). Similarly, at sacrifice the Resatorvid Prevention group displayed significantly reduced tumor multiplicity compared to Acetone controls, while reductions in tumor multiplicity in the Resatorvid Intervention group did not reach the level of statistical significance (D). * : p < 0.05. *** : p < 0.001

Figure 4. UV-induced tumors from resatorvid-treated mice display attenuated p38/Akt phosphorylation and increased apoptosis

Protein lysates from mouse cSCCs (Fig. 3) were subjected to Western blot analysis (A) followed by Image J-based quantification of p-Akt (normalized to total p38, B), p-p38 (normalized to total p38, C) and p21 (normalized to β-actin, D). Analogous analysis was performed using epidermal lysates from non-tumor-bearing high-risk mice (also used in Fig. 2) (E), followed by Image J-based quantification of TLR4 expression normalized against β-tubulin (F). Immunohistochemistry was performed on sections from the same tumors described in A to probe for cleaved caspase-3 expression (400×, G), followed by quantification using ImagePro Plus software (H). * : p < 0.05.

Figure 5. Resatorvid antagonizes TLR4 dependent cell signaling upstream of skin photocarcinogenesis

Resatorvid blocks TLR4 interaction with its downstream adaptor proteins containing the TIR-domain (including MyD88, TRAM/TRIF, TIRAP), thereby causing sustained suppression of multiple TLR4-dependent effector pathways, including IRF3, NF-κB and AP-1. Our findings indicate that topical pharmacological TLR4 suppression in the epidermis antagonizes chronic UV-induced inflammatory signaling and photocarcinogenesis, with potentiation of keratinocyte apoptosis.