New methods for the safety testing of transgenitic food

Project no. QLK1-CT-1999-00651 (New methods for the safety testing of transgenic food)

Introduction
The overall objective of this project is to develop and validate the scientific methodology for assessing the safety of foods from genetically modified plants.

The project will examine a two step safety testing procedure for GM plant food, having chemical analysis and short term in vivo and in vitro testing in the first step and a 90-day rat study in the second step. The test design will be evaluated through the testing of 3 transgenic rice varieties (gene inserts: snowdrop lectin (GNA), kidney bean lectin (PHA-E), and a toxin from Bacillus thuringiensis (BtCry1Ab)).

Introduction of recombinant protein for further testing in in vitro and animal studies

Scaling up and optimisation of GNA and PHA-E production

A series of high-cell density fermentation were carried out and optimised.

The use of the glyceraldehyde-3-phosphate dehydrogenase promoter (pGAP) provided significant increase of GNA and PHA-E expression levels in shake flasks compared to the pPICZ constructs (based on the alcohol oxidase promoter). However, despite a rapid and efficient growth, under fermentation conditions the P. pastoris strains carrying the pGAP constructs failed to deliver an improved yield of recombinant protein compared to the strains carrying the pPICZ constructs. It was therefore obvious that the use of the constructs based on the alcohol oxidase promoter was the best option.

In order to produce the amount of recombinant GNA and PHA-E required for the future animal studies it was decided to use another facility, which was commercially provided by the Institut Meurice in Belgium. Fermentations there were successfully scaled-up to a 200-litres fermenter using the optimal conditions. Yields obtained during fermentations were similar or slightly lower than yields previously observed at the 2 and 25-litres scale.

Downstream processing of the supernatant was carried out following the protocols in-use at the Institut Meurice and complementary information from the SCRI. During that step, losses of proteins were observed that reduced the final amount of recombinant GNA and PHA-E obtained.

Scaling up and optimisation of Cry1A(b) production by Escherichia coli

Optimisation of production was carried out in two phase: 1) Optimisation of growth and gene expression by investigating media compositions, growth conditions and their effects on the yields of recombinant Cry1A(b), and 2) Optimisation of Cry1A(b) extraction protocols. A routine production of Cry1A(b) using shake-flasks fermentations under optimal conditions was installed. Approximately 0.15 to 0.2 grams of pure recombinant endotoxin could be obtained per litre of culture.

Optimisation of recombinant GNA, PHA-E and Cry1A(b) purification

Recombinant GNA, PHA-E and Cry1A(b) were purified free of contaminants using a combination of ion-exchange, affinity and hydrophobic-interaction chromatography.

Purity was assessed by SDS-PAGE analysis.

Protocols were revised and optimised to meet the requirements of large-scale purification.

Activity of recombinant and standard PHA-E and GNA was assessed by agglutination assay. Biological function of recombinant Cry1A(b) was demonstrated by insect bioassay. Trypsin-activated recombinant Cry1Ab was shown to exert toxicity towards mid-instar larvae of Plutella xylostella (Diamondback moth) in a preliminary bioassay where the toxin, at various concentration levels, was topically applied to artificial diet. An optimised bioassay system using P. xylostella larvae on artificial diet with different concentrations of incorporated purified Cry1Ab toxin is currently in progress.

Since two bands were present in the final preparations of recombinant PHA-E and Cry1A(b), N-terminal sequencing of both bands were carried out.

The haemagglutination activity of recombinant GNA and PHA-E was shown to be similar to the native lectins from plant. The N-terminal sequence of both recombinant lectins was identical to the native lectins from plant.

Production of rice material for further testing (Chemical/compositional analysis, in vitro testing and animal studies)

Plant transformation and production

Transformation of rice (EYI105) with DNA construct pUbiPHA-E by the whole plasmid procedure has been carried out by John Innes Centre). Regenerated transformed plants were tested for expression of PHA-E lectin by immunoblotting analysis. Three of 11 transgenics tested were shown to express the lectin at low levels of between 0.01 and 0.02% (w/w) of total soluble protein extracted from leaf tissue. Further lines are in the process of being analysed. One line was found to express PHA-E at up to 0.1 % after 4 months.

Molecular characterisation (Southerns, Northerns) studies have been initiated on GNA and Bt rice. GNA rice, together with its parental line ASD16, have been bulked up in the glasshouse. Seed material (500 g each of transgenic and parental) from these plants have been sent to China for bulking up in the field by Prof. Q. Shu. Three successive generations (R₂ to R₄) of GNA rice seed are now available for comparison of levels of transgene product accumulation so as to assess stability of transgene expression within the seed. Quantities of Bt rice expressing Cry1Ab (KMD1), together with its parental line Xiushui 11, have now been bulked in the field in China by Prof. Q. Shu. The bulked-up Bt rice seed (approx. 120 - 140 kg each of transgenic and parental) has been dispatched for rat feeding trials. Three successive generations (R₉ to R₁₁) of Bt rice seed are ready for comparison of levels of transgene product accumulation so as to assess stability of transgene expression within the seed.

Compositional analysis
Near Infrared Spectroscopy (NIRS) is being evaluated as a non-destructive methodology for fingerprinting as well as for rapid determination of single compounds in rice flour.

In the reporting period development and optimisation of a calibration method to determine phytic acid in brown rice by means of NIR spectroscopy has been set up.

Brown rice samples (55) of different origin and grown in different seasons were analysed for their phytic acid contents using the photometric reference method.

For metabolite profiling the following steps have been taken: 1) validation of the developed GC-FID/MS profiling method, 2) comparative investigation of a set of conventional brown rice samples, and 3) development and validation of tools for automated analysis of chromatographic data.

Regarding the protein pattern in rice the following steps have been taken: 1) screening of prolamin patterns of different rice varieties, 2) investigation of prolamin pattern of one rice variety grown at different locations, 3) optimisation of the extraction and separation of albumins/globulins, and 4) screening of albumin/globulin patterns of different rice varieties.).

For the targeted analysis of specific constituents in rice, a number of methods have been developed.

g-Oryzanol comprises a mixture of phytosteryl ferulates located in rice bran. A method for rapid analysis of g-oryzanol in rice lipids by on-line coupling liquid chromatographic preparation with capillary gas chromatography (on-line LC-GC) has been set up and identification of the major constituents of g-oryzanol in rice separated by the developed on-line LC-GC method are done.

In addition methods for determination of Vitamin B₆ Niacin, and Vitamine B₁ (thiamine) has been developed and validated.

In vitro studies
The biodegradibility i.e. proteolytic degradation of the proteins, GNA and PHA-E, was determined in simulated gastric fluid for up to 4 hrs at 37^oC and in simulated intestinal fluid for up to 5-8 hrs at 37^oC. In addition, the digestibility of the proteins was analysed in SGF. Upon incubation, the digestion mixtures were analysed by SDS-PAGE and Western blotting.

All three proteins are resistant to trypsin at pH 7.5 for up to 5-8 hrs of incubation

PHA-E and Cry1Ab are rapidly degraded by pepsin at pH 2.0 (nearly no protein can be detected by Western blot after 90-120 min incubation). Incubation at pH 1.2 speeds up the degredation of PHA-E and Cry1Ab.

GNA is stable during at least 4 hrs of incubation, irrespective of the condition used. Thus, even at pH 1.2 GNA is resistant to pepsin digestion.

A new method for bioavailabilty studies in rats has been optimised and combines two techniques using isolated perfused gut segments and portal vein canulated rats. In this "two-in one" approach, perfusion of the jejunum-ileum part of the gut takes place in vivo and not in isolated segments.

Dose-response experiments were performed to determine whether plant-derived GNA and PHA-E and E. coli-produced Cry1Ab would be cytotoxic for Caco-2 cells. Preliminary results indicate that GNA and Cry1Ab were not cytotoxic to differentiated Caco-2 cells at concentrations up to 1000 µg/ml and 250 µg/ml, respectively. However, PHA-E was cytotoxic at concentrations of 250 µg/ml and higher whereas 100 µg/ml did not have an effect on the parameters under examination.

Microarrays containing 3000 well-characterised rat cDNAs has been constructed.

Animal studies
A small-scale preliminary animal study in rats was carried out to investigate the toxicity of the recombinant PHA-E protein. Acute toxic effects on the small intestine were seen with a few days at a dose level of 10 mg/rat/day, which are information to be used in the 28-day study. It is essential to state that rPHA-E protein and PHA-E expressing rice seed are only being generated and used as positive controls in the animal studies. The project has no intention for them to be used in any other way. The 28-day study will be initiated in Spring 2002, and the 90-day study in 2003.