Transcriptomic comparison of cyanotoxin variants in a human intestinal model revealed major differences in oxidative stress response: effects of MC-RR and MC-LR on Caco-2 cells.

Microcystins (MCs) are cyclic hepatotoxins produced by various species of cyanobacteria. Their structure includes two variable amino acids (AA) giving rise to more than 90 MC variants, however most of the studies to date have focused on the most toxic variant: microcystin LR (MC-LR). Ingestion is the major route of human exposure to MCs and several in vivo studies have demonstrated macroscopic effects on the gastro-intestinal tract. However, little information exists concerning the pathways affected by MC variants on intestinal cells. In the current study, we have investigated the effects of MC-RR and MC-LR on the human intestinal cell line Caco-2 using a non-selective method and compared their response at the pangenomic scale. The cells were incubated for 4h or 24h with a range of non-toxic concentrations of MC-RR or MC-LR. Minimal effects were observed after short term exposures (4h) to either MC variant. In contrast, dose dependent modulations of gene transcription levels were observed with MC-RR and MC-LR after 24h. The transcriptomic profiles induced by MC-RR were quite similar to those induced by MC-LR, suggestive of a largely common mechanism of toxicity. However, changes in total gene expression were more pronounced following exposure to MC-LR compared to MC-RR, as revealed by functional annotation. MC-LR affected two principal pathways, the oxidative stress response and cell cycle regulation, which did not elicit significant alteration following MC-RR exposure. This work is the first comparative description of the effects of MC-LR and MC-RR in a human intestinal cell model at the pangenomic scale. It has allowed us to propose differences in the mechanism of toxicity for MC-RR and MC-LR. These results illustrate that taking into account the toxicity of MC variants remains a key point for risk assessment.

Effects of DDT and permethrin on rat hepatocytes cultivated in microfluidic biochips: Metabolomics and gene expression study.

Dichlorodiphenyl-trichloroethane (DDT) and permethrin (PMT) are amongst most prevalent pesticides in the environment. Although their toxicity has been extensively studied, molecular mechanisms and metabolic effects remain unclear, including in liver where their detoxification occurs. Here, we used metabolomics, coupled to RT-qPCR analysis, to examine effects of DDT and PMT on hepatocytes cultivated in biochips. At 150 µM, DDT caused cell death, cytochrome P450 induction and modulation of estrogen metabolism. Metabolomics analysis showed an increase in some lipids and sugars after 6 h, and a decrease in fatty acids (tetradecanoate, octanoate and linoleate) after 24 h exposure. We also found a change in expression associated with genes involved in hepatic estrogen, lipid, and sugar metabolism. PMT at 150 µM perturbed lipid/sugar homeostasis and estrogen signaling pathway, between 2 and 6 h. After 24 h, lipids and sugars were found to decrease, suggesting continuous energy demand to detoxify PMT. Finally, at 15 µM, DDT and PMT appeared to have a small effect on metabolism and were detoxified after 24 h. Our results show a time-dependent perturbation of sugar/lipid homeostasis by DDT and PMT at 150 µM. Furthermore, DDT at high dose led to cell death, inflammatory response and oxidative stress.

A simple protocol for attenuating the auto-fluorescence of cyanobacteria for optimized fluorescence in situ hybridization (FISH) imaging.

Cyanobacteria contain pigments, which generate auto-fluorescence that interferes with fluorescence in situ hybridization (FISH) imaging of cyanobacteria. We describe simple chemical treatments using CuSO4 or H2O2 that significantly reduce the auto-fluorescence of Microcystis strains. These protocols were successfully applied in FISH experiments using 16S rRNA specific probes and filamentous cyanobacteria.

Multiparametric temporal analysis of the Caco-2/TC7 demonstrated functional and differentiated monolayers as early as 14 days of culture.

Reducing the differentiation period for obtaining an in vitro intestinal barrier model is required to reduce the duration and cost for drug screening assays. In this frame, the Caco-2/TC7 subclone differentiation state was investigated from day 0 (D0) to day 32 (D32). As such, the expression of 45 genes (including cell junction, cell polarization, cell functionality, drug transport and metabolism genes) was followed throughout the 32 days. In parallel, the monolayer polarization and the formation of the cellular junctions were characterized by the immuno-staining of occludin, claudin-1 and actin proteins. The cell monolayer permeability was analyzed via transepithelial electric resistance measurements and paracellular transport of Lucifer Yellow. The P-gp efflux efficiency was assessed by rhodamine 123 transport. Alkaline phosphate activity was quantified to assess the cell differentiation. Three stages of differentiation were observed using the clustering of principal component analysis of the RTqPCR data and the overall assays. From D0 to D10, cells were in a proliferation stage and under-differentiated; from D14 to D21 a stable differentiation stage was reached; from D25 to D32 the epithelium seemed to enter into a post-differentiated stage. This study demonstrates that Caco-2/TC7 cells are functional and ready for use in drug screening permeability assays from 14 days in culture when compared with conventional 21 days for Caco-2 cells. In addition, this study provides a refined set of data allowing temporal and multi scale investigations, due to the intracellular kinetics and mRNA levels that can be correlated with membrane protein kinetics and functional extracellular activities. Therefore, shorter time in culture combined with a better knowledge of the cells during the time in culture will in turn help to improve the quality and cost of Caco-2/TC7 assays for drug development.

Development of a new microfluidic platform integrating co-cultures of intestinal and liver cell lines.

We developed a new biological model to mimic the organ-organ interactions between the intestine and the liver. We coupled polycarbonate cell culture inserts and microfluidic biochips in an integrated fluidic platform allowing dynamic co-cultures (called IIDMP for Integrated Insert in a Dynamic Microfluidic Platform). The intestinal compartment was simulated using Caco-2 TC7 cells and the liver one by HepG2 C3A. We showed that Caco-2 TC7 viability, barrier integrity and functionality (assessed by paracellular and active transport), were not altered during co-cultures in the bioreactor in comparison with the conventional insert Petri cultures. In parallel, the viability and metabolism of the HepG2 C3A cells were maintained in the microfluidic biochips. Then, as proof of concept, we used the bioreactor to follow the transport of phenacetin through the intestinal barrier and its metabolism into paracetamol by the CYP1A of the HepG2 C3A cells. Our results demonstrated the performance of this bioreactor with cell co-cultures compared to static co-culture controls in which weak biotransformation into paracetamol was detected. Our study illustrated the interest of such a bioreactor combining the advantages of a cell culture barrier and of liver microfluidic cultures in a common framework for in vitro studies.