A novel chemopreventive mechanism for a traditional medicine: East Indian sandalwood oil induces autophagy and cell death in proliferating keratinocytes

Abstract

One of the primary components of the East Indian sandalwood oil (EISO) is α-santalol, a molecule that has been investigated for its potential use as a chemopreventive agent in skin cancer. Although there is some evidence that α-santalol could be an effective chemopreventive agent, to date, purified EISO has not been extensively investigated even though it is widely used in cultures around the world for its health benefits as well as for its fragrance and as a cosmetic. In the current study, we show for the first time that EISO-treatment of HaCaT keratinocytes results in a blockade of cell cycle progression as well as a concentration-dependent inhibition of UV-induced AP-1 activity, two major cellular effects known to drive skin carcinogenesis. Unlike many chemopreventive agents, these effects were not mediated through an inhibition of signaling upstream of AP-1, as EISO treatment did not inhibit UV-induced Akt or MAPK activity. Low concentrations of EISO were found to induce HaCaT cell death, although not through apoptosis as annexin V and PARP cleavage were not found to increase with EISO treatment. However, plasma membrane integrity was severely compromised in EISO-treated cells, which may have led to cleavage of LC3 and the induction of autophagy. These effects were more pronounced in cells stimulated to proliferate with bovine pituitary extract and EGF prior to receiving EISO. Together, these effects suggest that EISO may exert beneficial effects upon skin, reducing the likelihood of promotion of pre-cancerous cells to actinic keratosis (AK) and skin cancer.

Keywords: Autophagy; Cell death; Sandalwood oil; Ultraviolet light.

Figure 1. MTS assay to determine functional dose range of EISO in HaCaT cells

(A) HaCaT keratinocytes grown in a 96-well plate were serum-starved for 24 hr and then treated with a range of EISO from 0.0001-0.01% diluted in DMSO in triplicate for a period of 24 hr. (B) HaCaT cells were plated at varying cell densities so that the initial number of cells per well ranged from 1,563 to 50,000 cells using serial 1:1 dilutions of cells into DMEM. Again, cells were serum starved for 24 hr before treatment with 0.001% or 0.002% EISO in triplicate for an additional 24 hr. In all experiments, cell growth medium was spiked with MTS substrate for a period of 15 minutes and then an OD was obtained using a wavelength of 490 nm. Triplicate values were averaged and means expressed as a percentage of the reading for DMSO-treated control cells. In (A), triplicate wells were treated with DMSO, and in (B) a triplicate cell control was run for each density plated.

Figure 2. Apoptosis/cell death by annexin V/PI staining and cell cycle analysis in EISO-treated HaCaT cells

(A)) HaCaT keratinocytes were serum-starved for 24 hr and then treated with EISO (0.0005% or 0.001% in DMSO) for 1 hr prior to UVB irradiation, after which the treatments were added back to the cells. After a 6.5 hr rest the cells were incubated with Alexa 488-conjugated anti-annexin V (AnnV) antibodies and propidium iodide. Labeled cells were counted by flow cytometry, collecting a total of 10,000 data points per treatment condition. Data are expressed as percent of cells AnnV-/PI- (Live cells), AnnV+/PI- (Apoptotic cells) or PI+ regardless of AnnV status (Late apoptotic/necrotic). Data represent means±SD of three independent samples for each treatment condition. Asterisks indicate significant differences compared to control/mock cells (P value < 0.05). (B-C) Cell cycle analysis was performed on serum-starved cells treated with EISO for 24 hr or on cells first challenged with bovine pituitary extract and epidermal growth factor for 3 hr to initiate cell cycle entry before EISO treatment. Floating and adherent cells were collected and pooled. After spinning, cell pellets were fixed overnight in 70% EtOH and treated with RNase and PI for 15 minutes. Flow cytometry was used to collect 10,000 data points and data is expressed as percent of cells in G₁, S, or G₂/M phase based on relative cellular DNA content per cell.

Figure 3. EISO inhibited UV-induced AP-1 activity without affecting signaling pathways upstream of AP-1 known to be induced by UV

(A) HCL-14 cells, a HaCaT cell line stably expressing a portion of the human collagenase I gene (containing an AP-1 binding site) driving a firefly luciferase reporter gene, were treated with EISO for 1 hr prior to UV irradiation. After exposure to UV, media containing the same pretreatment (DMSO or EISO at the indicated concentration) was added back to the cells for 12 hr before cell lysates collected. Luciferase activity was measured and normalized to total protein content. Data are expressed as percent of the UV-induced luciferase activity in DMSO (vehicle)-treated control cells. (B) HaCaT cells were pretreated with EISO for 1 hr, irradiated with UVB and then returned to media containing pretreatments for 0.5 or 1 hr. Cell lysates were subjected to Western analysis using antibodies against phospho-Akt, phospho-JNK, phospho-ERK 1/2, and phospho-p38. Total ERK 1/2 was also analyzed to show that equal protein was loaded on the gel across all treatments. (C) FL-30 cells, a HaCaT cell line that stably expresses a full-length human c-fos promoter driving a firefly luciferase reporter gene, were pretreated with EISO for 1 hr before irradiation with UV and then returned to media containing pretreatments. Cell lysates were collected 6 hr post UV-irradiation and analyzed for luciferase activity, which was normalized to total protein.

Figure 4. Proliferating HaCaT cells were more sensitive to EISO-induced plasma membrane damage than quiescent cells

HaCaT cells were starved in serum-free DMEM for 24 hr. Fresh media supplemented with bovine pituitary extract and epidermal growth factor (A, C, E, G) or not supplemented (B, D, F, H) was added to the cells for 3 hr prior to treatment with EISO. At the indicated times, cells were detached by trypsinization and pooled with floating cells. Cells were pelleted, resuspended in PBS and incubated with fluorescein diacetate and propidium iodide for 15 minutes before analysis by flow cytometry. Using control cells loaded with individual dyes, regions of cells were identified as live cells with uncompromised membranes (R2), live cells with compromised membranes (R3), or severely compromised cells. Serum-starved EISO-treated cells (6 hr; panel D) displayed 2 distinct PI-stained populations of actively cycling cells in G₁ and G₂ phases (R7) that was used to confirm the population of cells with compromised membranes (evidenced by reduced fluorescein retention) but that were still alive.

Figure 5. Induction of autophagy in HaCaT cells treated with low doses of EISO

HaCaT cells plated on glass chamber slides were serum starved and treated with EISO for 24 hr. HaCaT cells were fixed and stained using antibodies against LC3 to demonstrate the induction of autophagy. Nuclear and perinuclear LC3 staining was observed in EISO-treated HaCaT cells, even at low doses.

Figure 6. UV-induced LC3 processing and expression increased and PARP cleavage decreased in EISO-treated proliferating and quiescent HaCaT cells

Serum-starved and BPE/EGF-treated HaCaT cells were treated with EISO for 16 hr. Adherent and floating cells were pooled and lysed in RIPA buffer for analysis of cellular proteins by Western blotting. PARP cleavage was induced by UV-irradiation, as expected, but EISO treatment blocked UV-induced apoptosis in both serum-starved and BPE/EGF treated cells. LC3 expression and processing were increased by UV in both BPE/EGF treated cells and serum-starved cells. EISO treatment increased LC3 processing in both treatment groups, and augmented the UVB response. Proliferating cells produced the most LC3 II after UVB treatment compared to controls.