Community Research and Development Information Service - CORDIS

Construction of cDNA microarrays and evaluation of data obtained

This result addressed the question whether in vitro experiments combined with new transcriptomics technologies would be useful for the safety testing of transgenic foods. For this feasibility study, the in vitro effect of GNA lectin, PHA-E lectin and Bt toxin on gene expression in intestinal epithelial cell lines from human (Caco-2) and rat (IEC-6) was assessed using prior knowledge obtained from the biodegradation and cytotoxicity experiments. Finally, the gene expression patterns obtained in vitro with PHA-E were compared with those found in vivo (28-day and 90-day feeding studies).

IEC-6 cells and undifferentiated/differentiated Caco-2 cells were exposed to non-cytotoxic concentrations of the lectins, Bt toxin, and peptic-tryptic (PT) digests. Total RNA was isolated and hybridised to microarrays containing either 3136 rat-specific or 2352 human-specific cDNAs. Also RNA from untreated cells (control) and cells exposed to a pepsin-trypsin mixture (digest control) was isolated and hybridised to the cDNA microarrays. Many genes were found to be differentially regulated upon exposure of the cells to the three proteins or corresponding PT digests. The most striking effect was seen in IEC-6 cells and differentiated Caco-2 cells exposed to PHA-E. From the three proteins, PHA-E resulted in the largest number of differentially expressed genes.

Furthermore, the number of genes differentially expressed upon exposure to the PT digest of PHA-E was approximately ten-fold lower. When the results from the human-derived Caco-2 cells and rat IEC-6 cells were compared, only little correlation could be found.

To assess the validity of the gene expression results obtained in vitro, gene expression profiling was performed on jejunal scrapings sampled from two groups of male and female rats in the PHA-E 28-day study. The rats from one group were fed on a purified diet with 60% rice, the rats from the other group were fed on the same diet but supplemented with 0.08% PHA-E. Total RNAs from the scrapings were isolated and hybridised to the same rat-specific microarrays as used for the in vitro experiments with IEC-6. When the data obtained with the male and female PHA-fed rats were compared, the majority of differentially expressed genes, including some immunologically relevant ones, were differentially expressed in a gender-specific fashion.

Furthermore, comparison of the data from the PHA-E 28-day and Caco-2/IEC-6 studies showed poor correlation between in vivo and in vitro. Two possible explanations for this lack of correlation were envisaged: IEC-6 and Caco-2 cells are not appropriate models to mimic the effects on intestinal epithelial cells of the jejunum; or intestinal scrapings consist of so many different cell types that the effects on the target cell type (epithelial cells) are masked.

In order to examine the latter possibility, in the subsequent PHA-E 90-day study epithelial cells were differentially isolated from the jejunal segments of female rats using a procedure that has been proven to produce a sequential isolation of epithelial cells from villus to crypt. Rats were from the control group (purified diet with non-transgenic rice), from the group fed with purified diet containing PHA rice, and from the group fed with purified diet containing PHA rice and 0.1% purified PHA. Total RNA was isolated from samples containing mainly either crypt or villus cells.

The most striking results were found in villus cells from rats fed with the diet containing purified PHA-E. Interestingly, immunologically relevant genes which were differentially regulated in these latter cells were also found to be regulated in a similar fashion in scrapings from female rats fed with PHA-E in the 28-day study. Gender-specific effects of PHA-E on the intestine were also observed upon macroscopic examination of the intestine of rats in the acute and 28-day study. Although, in the 90-day study microarray hybridisations were performed with RNAs from distinct fractions of cells, it did not improve the in vitro-in vivo correlation. Therefore, the combined results suggest that Caco-2 and IEC-6 cells are in vitro culture systems that are too simple to screen for effects of PHA-E on the intestine.

More information on the project can be found at:

Related information

Reported by

RIKILT-Institute of Food Safety (former: State Institute for Quality Control of Agricultural Products)
Bornsesteeg 45
6708 PD Wageningen
See on map
Follow us on: RSS Facebook Twitter YouTube Managed by the EU Publications Office Top