Validation of an in vitro screening test for predicting the tumor promoting potential of chemicals based on gene expression

Can a 12-gene expression signature predict the cell transforming potential of tumor promoting agents in Bhas 42 cells?

To date, long-term rodent carcinogenesis assays are the only assays recognized by regulators to assess non-genotoxic carcinogens, but their reliability has been questioned. In vitro cell transformation assays (CTAs) could represent an interesting alternative to animal models as it has the advantage of detecting both genotoxic and non-genotoxic transforming chemicals. Among them, Bhas 42 CTA uses a cell line that has been transfected with the oncogenic sequence v-Ha-ras. This sequence confers an "initiated" status to these cells and makes them particularly sensitive to non-genotoxic agents. In a previous work, transcriptomic analysis revealed that the treatment of Bhas 42 cells with transforming silica (nano)particles and 12-O-tetradecanoylphorbol-13-acetate (TPA) commonly modified the expression of 12 genes involved in cell proliferation and adhesion. In the present study, we assess whether this signature would be the same for four other soluble transforming agents, i.e. mezerein, methylarsonic acid, cholic acid and quercetin. The treatment of Bhas 42 cells for 48 h with mezerein modified the expression of the 12 genes of the signature according to the same profile as that of the TPA. However, methylarsonic acid and cholic acid gave an incomplete signature with changes in the expression of only 7 and 5 genes, respectively. Finally, quercetin treatment induced no change in the expression of all genes but exhibited higher cytotoxicty. These results suggest that among the transforming agents tested, some may share similar mechanisms of action leading to cell transformation while others may activate different additional pathways involved in such cellular process. More transforming and non-transforming agents and gene markers should be tested in order to try to identify a relevant gene signature to predict the transforming potential of non-genotoxic agents.

Predictive early gene signature during mouse Bhas 42 cell transformation induced by synthetic amorphous silica nanoparticles

Synthetic amorphous silica nanoparticles (SAS) are used widely in industrial applications. These nanoparticles are not classified for their carcinogenicity in humans. However, some data still demonstrate a potential carcinogenic risk of these compounds in humans. The Bhas 42 cell line was developed to screen chemicals, as tumor-initiators or -promoters according to their ability to trigger cell-to-cell transformation, in a cell transformation assay. In the present study, we performed unsupervised transcriptomic analysis after exposure of Bhas 42 cells to NM-203 SAS as well as to positive (Min-U-Sil 5® crystalline silica microparticles, and 12-O-tetradecanoylphorbol-13-acetate) and negative (diatomaceous earth) control compounds. We identified a common gene signature for 21 genes involved in the early stage of the SAS- Min-U-Sil 5®- or TPA-induced cell transformation. These genes were related to cell proliferation (over expression) and cell adhesion (under expression). Among them, 12 were selected on the basis of their potential impact on cell transformation. RT-qPCR and western blotting were used to confirm the transcriptomic data. Moreover, similar gene alterations were found when Bhas 42 cells were treated with two other transforming SAS. In conclusion, the results obtained in the current study highlight a 12-gene signature that could be considered as a potential early "bio-marker" of cell transformation induced by SAS and perhaps other chemicals.

Carcinogenicity risk assessment supports the chronic safety of dapagliflozin, an inhibitor of sodium-glucose co-transporter 2, in the treatment of type 2 diabetes mellitus

INTRODUCTION

Dapagliflozin is a selective inhibitor of the sodium-glucose co-transporter 2 (SGLT2) that increases urinary glucose excretion to reduce hyperglycemia in the treatment of type 2 diabetes mellitus. A robust carcinogenicity risk assessment was undertaken to assess the chronic safety of dapagliflozin and SGLT2 inhibition.

METHODS

Genotoxicity potential of dapagliflozin and its metabolites was assessed in silico, in vitro, and in vivo. Dapagliflozin was administered daily by oral gavage to mice, rats, and dogs to evaluate carcinogenicity risks, including the potential for tumor promotion. SGLT2(-/-) mice were observed to evaluate the effects of chronic glucosuria. The effects of dapagliflozin and increased glucose levels on a panel of human bladder transitional cell carcinoma (TCC) cell lines were also evaluated in vitro and in an in vivo xenograft model.

RESULTS

Dapagliflozin and its metabolites were not genotoxic. In CD-1 mice and Sprague-Dawley rats treated for up to 2 years at ≥100× human clinical exposures, dapagliflozin showed no differences versus controls for tumor incidence, time to onset for background tumors, or urinary bladder proliferative/preneoplastic lesions. No tumors or preneoplastic lesions were observed in dogs over 1 year at >3,000× the clinical exposure of dapagliflozin or in SGLT2(-/-) mice observed over 15 months. Transcription profiling in Zucker diabetic fatty rats showed that 5-week dapagliflozin treatment did not induce tumor promoter-associated or cell proliferation genes. Increasing concentrations of glucose, dapagliflozin, or its primary metabolite, dapagliflozin 3-O-glucuronide, did not affect in vitro TCC proliferation rates and dapagliflozin did not enhance tumor growth in nude mice heterotopically implanted with human bladder TCC cell lines.

CONCLUSION

A multitude of assessments of tumorigenicity risk consistently showed no effects, suggesting that selective SGLT2 inhibition and, specifically, dapagliflozin are predicted to not be associated with increased cancer risk.