Identifying Small Molecules which Inhibit Autophagy: a Phenotypic Screen Using Image-Based High-Content Cell Analysis

Autophagy plays an important role in cancer and it has been suggested that it functions not only as a tumor suppressor pathway to prevent tumor initiation, but also as a pro-survival pathway that helps tumor cells endure metabolic stress and resist death triggered by chemotherapeutic agents, including acquired resistance. We aimed to identify small-molecule autophagy inhibitors using a HTS/HCA approach through a phenotypic, cell image-based assay, in order to screen multiple biological targets simultaneously and to screen compounds in a physiologically relevant environment. LC3 is a component of the autophagosome, which undergoes a cytoplasmic redistribution from diffuse to punctate dots during autophagy. We employed HeLa cells stably expressing EGFP-LC3 in a primary phenotypic screen. As a first step, a “Validation Library” of about 8,000 pre-selected compounds, about 25% of which had known biological activity and the others representing a range of chemical structures, was run in duplicate both to assess screening suitability and likely hit rate, and to give a valuable preview of possible active structures or biological targets. The primary screen of about 0.25 million compounds yielded around 10,500 positive compounds. These were tested in a suite of further cellular assays designed to eliminate unwanted positives, together with the application of chemi- and bioinformatics to pick out compounds with known biological activity. These processes enabled the selection of compounds that were the most promisingly active and specific. The screening “tree” identified, amongst others with as yet unidentified targets, chemical series active against autophagy-relevant biological targets ULK or Vsp34, validating the phenotypic screening methods selected. Finally, about 400 compounds were fully qualified after following this triage. The development of the assays, compound screening process and the compound triage is described.

Synthesis and identification of [1,3,5]triazine-pyridine biheteroaryl as a novel series of potent cyclin-dependent kinase inhibitors

On the basis of previous studies, we identified pyrazine-pyridine A as a potent vascular endothelial growth factor inhibitor and pyrimidine-pyridine B as a moderately potent cyclin dependent kinase (CDK) inhibitor. A proposed combination of CGP-60474 and compound B led to the discovery of [1,3,5]triazine-pyridine as a new series of potent CDK inhibitors. Palladium-catalyzed C-C bond formation reactions, particularly the Negishi coupling reaction, were used to assemble various triazine-heteroaryl analogues effectively. Among them, compound 20 displayed high inhibitory potency at CDK1 (IC(50) = 0.021 microM), CDK2, and CDK5 and submicromolar potency at CDK4, CDK6, and CDK7. Compound 20 also displayed high potency at GSK-3beta. It demonstrated potent antiproliferative activity on various tumor cell lines, including HeLa, HCT-116, U937, and A375. When 20 was administered intraperitoneally at 150 and 125 mg/kg to nude mice bearing human A375 xenografts, the compound produced a significant survival increase. Molecular docking studies were conducted in an attempt to enhance the understanding of the observed structure-activity relationship.

Flow cytometric measurement of cytoplasmic pH: a critical evaluation of available fluorochromes

Three pH-sensitive fluorochromes-4-methyl-umbelliferone(4MU),2, 3-dicyano-hydroquinone (DCH), and 2',7'-bis(carboxyethyl)-5,6-carboxy fluorescein (BCECF)--were evaluated for their resolution, range, and stability of cellular fluorescence. Flow cytometric techniques for determining cytoplasmic pH (pHi) have been fully described for 4MU and DCH; BCECF has previously been used for fluorimetric estimation of pHi, and was adapted to flow cytometry. For each fluorochrome, the ratio of fluorescence intensity at two wavelengths gives a measure of pHi, which may be calibrated by obtaining the fluorescence ratios for cells suspended in buffers of varying pH in the presence of a proton ionophore. Reliable calibration proved difficult using 4MU, partly because of poor retention within cells. Both DCH and BCECF could be calibrated using a fluorescence ratio and had resolutions of 0.2 and 0.4 pH units, respectively. The fluorescence of DCH is so strongly pH dependent that there were practical difficulties in its use over a wide pH range; however, pHi measurements are possible between pH 6.0 and pH 7.5 using either DCH or BCECF. Substantial dye leakage was found for 4MU and, to a lesser extent, DCH, while BCECF was retained by cells for up to 2 hours. Despite its lower resolution BCECF had a usable range of more than 1.5 pH units and this coupled with its stable fluorescence and excitation at 488 nm rather than UV suggests a wide application.

Modulation of transferrin receptor expression by inhibitors of nucleic acid synthesis

We investigated the effects of the iron chelator desferrioxamine on the expression of transferrin receptors (TfR) by CCRF-CEM human T-cell leukaemia and B16 mouse melanoma cells growing in tissue culture. Desferrioxamine (DFOA) enhanced TfR expression when added in the dose range of 10(-5)-10(-4) to CCRF-CEM cells, but was toxic to these cells, the lower concentrations producing a slowing of cell growth with a build up in S-phase, while higher concentrations caused cell death with a block at the G1/S-phase interface. These toxic effects are compatible with its previously reported inhibition of the non-haem iron containing (M2) subunit of ribonucleotide reductase. In marked contrast, DFOA caused the growth of B16 melanoma cells to arrest in G1, without loss of cloning efficiency, and resulted in a fall in TfR expression to approximately 50% of control values. These results suggested that the effects of DFOA on TfR expression were linked to DNA synthesis rather than to a more generalised inhibition of iron-dependent cellular processes. It was subsequently found that inhibition of the M2 subunit of ribonucleotide reductase in CCRF-CEM cells with 5 X 10(-5) M hydroxyurea, which is not an iron chelator, also enhanced TfR expression, as did thymidine and cytosine arabinoside, which have different enzyme targets. By measuring cellular DNA and RNA content simultaneously it was shown that all of these agents caused unbalanced growth, i.e., inhibited DNA synthesis more than RNA synthesis. In contrast, 6-thioguanine was more inhibitory to RNA synthesis, and treatment with this drug caused a fall in TfR expression. Thus, although CCRF-CEM cells treated with DFOA show enhanced TfR expression, similar effects are also seen with other inhibitors of DNA synthesis, provided that RNA synthesis is allowed to continue. These results provide further evidence that the regulation of TfR expression by proliferating cells is specifically linked to DNA synthesis rather than to the iron requirements of other cellular processes.

Transferrin receptor expression during exponential and plateau phase growth of human tumour cells in culture

Transferrin receptor expression by the human tumour cell lines CCRF-CEM leukaemia and PMC-22B melanoma was studied, measuring the specific binding of fluorescein isothiocyanate (FITC)-labelled transferrin using a fluorescence-activated cell sorter. By measuring the fluorescence of cells stained at subsaturating concentrations of conjugate it was possible to calculate the average numbers of receptors per cell and the binding affinity by Scatchard analysis. These values (1.9 X 10(5) binding sites/cell, KA 1.2 X 10(9) M-1 for CCRF-CEM during exponential growth and 6.9 X 10(4) binding sites/cell, KA 1.4 X 10(-9) M-1 for PMC-22B) are in close agreement with previously published data obtained using radiolabelled transferrin. The present method, however, allowed the transferrin receptor expression of individual cells within a population to be measured and thus it has been possible to test the hypothesis that transferrin receptor is a marker for cycling cells. Frequency-distribution histograms of transferrin receptor showed a wide range of values for both cell lines during exponential growth. When the extreme ranges were sorted and the cells examined for cellular DNA content it was found that those with the highest transferrin receptor expression were enriched with cells in S, G2, and M phases of the cell cycle, whereas those with low transferrin receptor expression were mainly in G1. However, two-parameter-correlated dot plots of transferrin receptor expression versus DNA content showed there was considerable overlap between the ranges of receptor expression for the different cell cycle compartments. Using a stathmokinetic method we have measured the proportion of quiescent cells in fed plateau phase cultures. Transferrin receptor expression was downgraded under these growth conditions but, contrary to expectation, the decline affected the population uniformly, without the emergence of a distinct, transferrin receptor-negative subpopulation corresponding to the increasing proportion of quiescent cells. Thus, although transferrin receptor expression bears some relation to cell cycle phase and reflects the proliferative activity of populations of cells, it is incapable of identifying individual cells which are out of cycle.

Flow cytometric and morphological studies of ovarian carcinoma cell lines and xenografts

Human ovarian cancers of four different histological types have been cultured in vitro and in nude mice. Nineteen tumor specimens (11 solid tumors and eight malignant effusions) were obtained from 14 patients. Tumor lines from ten of these patients were established after several subpassages, and six xenograft lines have been grown, all of them from tumors of which a cell line exists in vitro. In all, 14 lines have been successfully cultured from the 19 tumor specimens. The morphology (studied by light and electron microscopy) of the established lines in vitro and in vivo resembles that of the original tumors in all cases. Flow cytometric studies of DNA content of the original tumor specimens and the cell culture and xenograft lines were performed. In all lines in both culture systems, aneuploid cells became predominant after the first to fourth passages, despite an aneuploid peak having been evident in only nine of the 19 initial specimens. Four of the original tumor specimens contained measurable estrogen receptors and five progesterone receptors, but none of the established cell lines expressed these hormone receptors. These results indicate that, while morphological features are similar in the initial tumor specimens and in the established lines in vitro and in vivo, flow cytometric and steroid hormone receptor data suggest selection of aneuploid and receptor-negative cells.