Prevention of carcinogen and inflammation-induced dermal cancer by oral rapamycin includes reducing genetic damage

Abstract

Cancer prevention is a cost-effective alternative to treatment. In mice, the mTOR inhibitor rapamycin prevents distinct spontaneous, noninflammatory cancers, making it a candidate broad-spectrum cancer prevention agent. We now show that oral microencapsulated rapamycin (eRapa) prevents skin cancer in dimethylbenz(a)anthracene (DMBA)/12-O-tetradecanoylphorbol-13-acetate (TPA) carcinogen-induced, inflammation-driven carcinogenesis. eRapa given before DMBA/TPA exposure significantly increased tumor latency, reduced papilloma prevalence and numbers, and completely inhibited malignant degeneration into squamous cell carcinoma. Rapamycin is primarily an mTORC1-specific inhibitor, but eRapa did not reduce mTORC1 signaling in skin or papillomas, and did not reduce important proinflammatory factors in this model, including p-Stat3, IL17A, IL23, IL12, IL1β, IL6, or TNFα. In support of lack of mTORC1 inhibition, eRapa did not reduce numbers or proliferation of CD45(-)CD34(+)CD49f(mid) skin cancer initiating stem cells in vivo and marginally reduced epidermal hyperplasia. Interestingly, eRapa reduced DMBA/TPA-induced skin DNA damage and the hras codon 61 mutation that specifically drives carcinogenesis in this model, suggesting reduction of DNA damage as a cancer prevention mechanism. In support, cancer prevention and DNA damage reduction effects were lost when eRapa was given after DMBA-induced DNA damage in vivo. eRapa afforded picomolar concentrations of rapamycin in skin of DMBA/TPA-exposed mice, concentrations that also reduced DMBA-induced DNA damage in mouse and human fibroblasts in vitro. Thus, we have identified DNA damage reduction as a novel mechanism by which rapamycin can prevent cancer, which could lay the foundation for its use as a cancer prevention agent in selected human populations.

Figure 1. eRapa prevents DMBA/TPA-induced dermal neoplasia and malignant degeneration

WT BL6 mice on Eudragit or eRapa given DMBA/TPA for 24 weeks. A. Tumor prevalence is the proportion of mice with at least 1 visible tumor (>1 mm) lasting > 1 week. p-value, log-rank test. B. Tumor multiplicity (average papilloma number/mouse). p-value, 2-way ANOVA. C. Total tumor burden (∑tumor area/mouse). p-value, unpaired t test. D. Area per tumor is calculated as total tumor burden/number of tumors/mouse. p-value, unpaired t test. Mice without tumors were excluded. E. Number of mice with squamous cell carcinoma (SCC) lesions, n=28–29/group. p-value, Fisher’s exact test. F. Representative photomicrographs from Eudragit-fed mice. Original magnification, 40x.

Figure 2. eRapa protection occurs without local mTORC1 suppression or high skin rapamycin concentrations

WT BL6 mice on Eudragit or eRapa given DMBA/TPA for 24 weeks. A. Epidermal lysates blotted for indicated proteins. Five mice for Eudragit and eRapa. The numbers refer to the number of tumors per mouse for each lane. B. Western blot quantification, N=5/group, from A. p-value, unpaired t test. C. Skin papilloma lysates blotted for indicated proteins. Five mice for Eudragit and eRapa. D. Western blot quantification, N=5/group, from C. p-value, unpaired t test. E. Whole skin and epidermis assayed for rapamycin by liquid chromatography/tandem mass spectrometry, N=5/group. p-value, Mann Whitney Test. Dashed line: assay detection limit.

Figure 3. eRapa protection does not require suppression of known pro-inflammatory factors or significant reductions in keratinocyte stem cell numbers, proliferation, or epidermal hyperplasia

WT BL6 mice on Eudragit or eRapa given DMBA/TPA for 24 weeks. A. Epidermal lysates, N=8/group, assayed by Luminex. p-value, unpaired t test. B. Epidermal lysates Western blotted for indicated proteins. Five representative mice for Eudragit and eRapa groups in Western blot on top. Western blot quantification (N=10/group, bottom). C. Absolute numbers (left) and proliferation (right) by BrdU staining of basal (CD34⁺CD49f^hi and suprabasal (CD34⁺CD49f^mid) keratinocyte stem cells by flow cytometry, N=8/group. p-value, unpaired t test. Data representative of 3 experiments. Flow cytometry gating on left. p-value, unpaired t test. D. Epidermal cells stained for Ki-67 and assessed by flow cytometry, N=8/group. p-value, unpaired t test. E. >300 epidermal thickness measurements from hematoxylin and eosin sections, N=13–14/group. p-value, unpaired t test.

Figure 4. eRapa mediates skin neoplasia prevention in the absence of T cells

WT BL6 mice on Eudragit or eRapa and given DMBA/TPA for 24 weeks. Digested epidermis assayed by flow cytometry for A. frequency (left) and absolute numbers (right) of αβ T cells (CD45+β TCR⁺) and B. frequency (left) and absolute numbers (right) of γδ T cells (CD45⁺γδ TCR⁺), N=14–15/group. p-value, unpaired t test. C. T cell receptor knockout (βδ TCR KO, lacking all T cells) and syngeneic WT BL6 (data from Fig. 1 as control) mice on Eudragit or eRapa given DMBA/TPA for 24 weeks. Left: tumor multiplicity. p-value, 2-way ANOVA. Right: total tumor burden. p-value, unpaired t test. D. Number of mice with squamous cell carcinoma (SCC, n=8–9/group). p-value, Fisher’s exact test.

Figure 5. eRapa protection against tumorigenesis is associated with reduced skin DNA damage and reduction in vivo of the specific hras mutation that is carcinogenic in this model

A. Tumor multiplicity (left) and SSC prevalence (right) in WT BL6 mice fed Eudragit or eRapa one week after DMBA (post-DMBA), then promoted with TPA. p-value, 2-way ANOVA for tumor multiplicity and Fisher’s exact test for SCC prevalence. SSC, squamous cell carcinoma. B. Digested epidermis assayed by flow cytometry for p-H2AX. Post-DMBA data compared to WT BL6 mice fed Eudragit or eRapa for 1 month before DMBA (pre-DMBA). p-value, unpaired t test. C. Epidermal genomic DNA from post-DMBA mice assayed by quantitative PCR for the CAA → CTA mutation in codon 61 of hras gene. Normalized to naïve (no DMBA/TPA) epidermis, N=4–5/group. p-value, unpaired t test. D. Representative immunofluorescence staining of whole skin from pre-DMBA mice using anti phospho-H2AX (p-H2AX) and DAPI. Secondary antibody: anti-mouse IgG Alexa Fluor 488. Three eRapa or Eudragit mice representative of N=10/group for immunofluorescence studies are shown in the bottom 3 rows. E. Epidermal lysates blotted for indicated proteins. Five mice shown for Eudragit and eRapa (left). Western blot quantification (right, N=5/group). p-value, unpaired t test. F. Epidermal genomic DNA from pre-DMBA mice assayed by quantitative PCR for the CAA → CTA mutation in codon 61 of hras gene. Normalized to naïve (no DMBA/TPA) epidermis, N=4/group. p-value, unpaired t test.

Figure 6. Rapamycin in vitro at levels achievable in skin protects from DNA damage by DMBA

A. Mouse fibroblasts (3T3) and B. human foreskin fibroblasts (HFF) pre-treated with indicated doses of rapamycin for 24 hours before DNA damage was induced with 16.4 µg/ml DMBA, then assessed by flow cytometry for phospho-H2AX (p-H2AX) staining. p-value, unpaired t test. C. 3T3 mouse fibroblasts pre-incubated with DMBA for 24 hours to induce DNA damage, then treated with 200 pM rapamycin without continuing DMBA for 24–72 hours. DNA damage assessed by flow cytometry p-H2AX staining. p-value, unpaired t test. D. Digested epidermis assayed by flow cytometry for prevalence (left) and numbers (right) of Langerhans cells (CD11b⁺Langerin⁺). p-value, unpaired t test.