A novel cell-based screen identifies chemical entities that reverse the immune-escape phenotype of metastatic tumours

Abstract

Genetic and epigenetic events have been implicated in the downregulation of the cellular antigen processing and presentation machinery (APM), which in turn, has been associated with cancer evasion of the immune system. When these essential components are lacking, cancers develop the ability to subvert host immune surveillance allowing cancer cells to become invisible to the immune system and, in turn, promote cancer metastasis. Here we describe and validate the first high-throughput cell-based screening assay to identify chemical extracts and unique chemical entities that reverse the downregulation of APM components in cell lines derived from metastatic tumours. Through the screening of a library of 480 marine invertebrate extracts followed by bioassay-guided fractionation, curcuphenol, a common sesquiterpene phenol derived from turmeric, was identified as the active compound of one of the extracts. We demonstrate that curcuphenol induces the expression of the APM components, TAP-1 and MHC-I molecules, in cell lines derived from both metastatic prostate and lung carcinomas. Turmeric and curcumins that contain curcuphenol have long been utilized not only as a spice in the preparation of food, but also in traditional medicines for treating cancers. The remarkable discovery that a common component of spices can increase the expression of APM components in metastatic tumour cells and, therefore reverse immune-escape mechanisms, provides a rationale for the development of foods and advanced nutraceuticals as therapeutic candidates for harnessing the power of the immune system to recognize and destroy metastatic cancers.

Keywords: antigen processing machinery; curcuphenol; drug discovery; high throughput cell-based assay; major histocompatibility complex class I; metastatic tumours; natural products.

FIGURE 1

Flow cytometry analysis of IFN-γ stimulated unsorted LMD pTAP-1 metastatic prostate cancer cells expressing GFP under the TAP-1 promoter. LMD pTAP-1 cells were stimulated with 100 ng/mL of IFN-γ and analyzed by flow cytometry. Histogram shows GFP negative (black) and GFP positive (red) LMD pTAP-1 cells. Cells expressing high GFP levels (cells with fluorescence intensity above where the arrow is indicating) were sorted individually into 96-well plates to obtain clonal populations for further analysis.

FIGURE 2

Selection of LMD pTAP-1 clone. The intensity of GFP fluorescence for clones 14, 15 and 20 is shown after stimulation of the LMD pTAP-1 cells with a range of IFN-γ concentrations. Clone 15 showed the highest TAP-1 promoter activity measured as GPF intensity after IFN-γ stimulation and was chosen for further experiments.

FIGURE 3

High-throughput cell-based screening assay to identify marine invertebrate extracts and pure compounds that are able to induce TAP-1 expression. (A) Images of the DNA staining and TAP-1 promoter-induced GFP expression in LMP pTAP-1 cells are shown. Image acquisition, segmentation and analysis of 96-well plates were carried out using the Cellomics Arrayscan VTI imager. Segmentation to delineate the nuclei based on the DNA staining fluorescence intensity was performed to identify individual objects and create a cytoplasmic mask around the nuclei in which total GFP fluorescence is measured. Average GFP fluorescence intensity and total number of cells per well were determined. (B) IFN-γ induces high level of GFP expression in APM-deficient LMD pTAP-1 cells. Cells were treated with 10 ng/mL of IFN-γ or 1% DMSO vehicle control. Images were taken with the same exposure time. Lines indicate the average GFP intensities. (C) Induction of GFP expression is TAP-1 promoter dependent. Quantitation of GFP expression in LMD pEGFP cells (LMD cells transfected with a promoterless pEGFP-1 vector) and LMD pTAP-1 EGFP cells treated with 10 ng/mL of IFN-γ or 1% DMSO vehicle control. Lines indicate the average GFP intensities. Representative graph of the Z′-factor which was calculated to assess the quality of the screening assay.

FIGURE 4

Results from high-throughput screening assay of 480 marine invertebrate extracts. The percentage of activity for each extract was calculated in comparison to the GFP fluorescence intensity of the IFN-γ positive control as described in Materials and Methods. The dots denote the results of each individual extract. Extracts with greater then 40% activity and with low cytotoxicity (cell viability within one standard deviation of 1% DMSO negative control represented by dashed vertical lines) were selected as candidates for further analysis. The red dots denote the extracts that fulfilled these criteria.

FIGURE 5

Graph of percentage of extract activity versus cell number for Extract 1 (A), Extract 2 (B), Extract 3 (C), Extract 4 (D), Extract 5 (E), Extract 6 (F) and Extract 7 (G). Selected active marine extracts were serially diluted and further tested in LMD pTAP-1 cells using the Cellomics Arrayscan. Percentage of extract activity was plotted against the average cell number to evaluate the extract dilutions that provide the highest extract activity with lowest cytotoxicity. GFP intensity measurements were not provided in wells treated with extract concentrations that showed high cytotoxicity (low cell numbers) and therefore are not shown in the graphs. Four extracts showed dilutions which were detected as active (depicted as red dots).

FIGURE 6

Dose-response curves for Extract 1 (A), Extract 2 (B), Extract 3 (C), Extract 4 (D), Extract 5 (E), Extract 6 (F) and Extract 7 (G). Selected active marine extracts were serially diluted and further tested in LMD pTAP-1 cells using the Cellomics Arrayscan.

FIGURE 7

Identification of two selected marine extracts (Extracts 2 and 5) with the ability to induce MHC-I surface expression in metastatic cancer cells. (A) MHC-I surface expression in LMD pTAP1 metastatic cancer cells was quantified using flow cytometry upon stimulation for 48 h with Extracts 2 and 5 at varying concentrations. (B) Extracts 2 and 5 were fractionated to identify the components inducing the expression of MHC-I. The fractionated extracts were tested for their ability to induce MHC-I surface expression in LMD pTAP1 cells 48 h after treatment using flow cytometry. Mean fluorescence intensity (MFI); Trichostatin A (TSA).

FIGURE 8

NMR spectra of the natural product curcuphenol identified in Extract 2 and synthetic curcuphenol. (A) Comparison of the 1H NMR spectra of the natural product (left) and synthetic curcuphenol (right) in DMSO-d6 at 600 MHz. (B) Comparison of the 13C NMR spectra of the natural product and synthetic curcuphenol in DMSO-d6 at 150 MHz. (C) Molecular structure of natural curcuphenol.

FIGURE 9

Curcuphenol induces MHC-I and TAP-1 mRNA expression in APM-deficient A9 metastatic lung cancer cells. A9 cells were stimulated with 75uM (16.37 µg/mL) of curcuphenol for 48 h. Relative quantification (RQ), including upper and lower limits, were calculated by the 7,500 Fast RT-PCR system software. Values were normalized to SDHA106 housekeeping gene and calculated with 1% DMSO treatment as reference where cells treated with 1% DMSO were normalised to 1.