Dr Jens Freitag, office manager of the German Plant Genome Research Programme GABI.

Debate

"Epigenetic effects occur with every hybridisation."

Epigenetics: New findings – new uncertainty for genetically modified plants?

Questions for Jens Freitag

GMO Safety: The "epigenetic effects" scenario is receiving a lot of attention in the discussion about potential risks of transgenic plants. What lies behind this catchword?

Jens Freitag: All effects that are not based on the information stored in the DNA but are still passed on to subsequent generations are described by molecular biologists as epigenetic. One example is gene silencing – switching off genes by transcribing DNA into RNA (post-transcriptional gene silencing). According to our current level of understanding, this is based primarily on quantitative processes. In other words, the quantity of transcript formed determines the regulatory mechanisms that lead to gene silencing. This can be expressed phenotypically, for instance, in a different flower colour. These effects – called transcriptional or post-transcriptional effects – can be caused e.g. by means of methylation at DNA level.

But small, interfering RNA molecules are also involved in these processes. These so-called siRNAs have radically changed the face of molecular biological research over recent years. Probably no discovery has shaped the face of molecular biology as much as these RNA interference (RNAi) mechanisms since the discovery of PCR . Their effect was first mentioned by Dr Richard Jorgensen. In 1998 he published a study on differences in the gene expression of transgenic petunia flowers and the molecular mechanisms involved. Today we know that interfering RNA molecules occur in most natural organisms. In the area of human medicine in particular, researchers are hoping that these small molecules will lead to new treatments for complex diseases such as cancer.

GMO Safety: Not all that long ago it was assumed that a gene coded for just one protein. Now we are aware of a whole range of mechanisms that are increasingly casting doubt on the general applicability of the old "one gene one protein" rule. To what extent are new findings from epigenetics already being incorporated into plant research and plant breeding?

Jens Freitag: We have known that the number of genes is different from the number of specific proteins at least since the complete sequencing of the human genome in 2000. With the knowledge that 30,000 genes probably code for considerably more than 100,000 proteins it was impossible to maintain the view of a linear relationship between gene and gene products. Today we recognise that the expression processes of a gene’s basic genetic matrix can function in different ways. We call this “alternative splicing”.

Increasing knowledge about complex biological processes at molecular level, including epigenetic effects, means that it is now increasingly possible to carry out plant breeding as a "rational" process. This knowledge will not replace the quantitative genetics practised since Mendel, but underpins and gives greater precision to our approach. Aspects of practical plant breeding which are today reminiscent of trial and error are being steered towards knowledge-based approaches.

GMO Safety: Have genetic engineering methods been refined on the basis of the new findings?

Jens Freitag: They certainly have. Many of the newly discovered processes, such as DNA methylation and siRNAs can now be verified and used. Some phenomena from the early days of plant genetic engineering can now be explained in molecular biological terms. For instance, we now know that gene silencing takes place at transcript rather than DNA level. This is why the oft-cited risk scenario involving the position effect is no longer justified. It says that untargeted integration of the transgenes into the plant genome could unintentionally switch some of the plant’s own genes on or off.

GMO Safety: Epigenetic processes manifest themselves in altered plant characteristics, e.g. changes to the morphology, agronomic characteristics or plant substances. Do such effects go unnoticed in the numerous tests and inspections that transgenic lines pass through before they are authorised? Are there gaps in the current risk assessment concept?

Jens Freitag: Epigenetic effects are naturally occurring mechanisms. Plant breeders use them – wittingly or unwittingly – in hybrids: the progeny of plants from widely disparate gene pools demonstrate increased vigour, e.g. biomass formation, which is clearly above the average of the two parents (heterosis effect) . The resultant broadening of the genetic base has led to what we today call adapted cultivated plants. This means that epigenetic effects are not confined to genetic engineering. Every new breed produced by humans is subject to epigenetic effects – but so are all crosses that take place without human intervention. Take the example of resistance breeding, which has to use very disparate crosses in order to transfer resistance characteristics from wild forms to our cultivated plants. This results in changes to the methylation pattern and gene activity. No one was interested at this point in whether people were aware that they were interfering in molecular regulatory mechanisms.

At the development stage, but especially when it comes to authorising a new plant variety, the focus is on detecting early on any unintended effects – including effects that might have epigenetic causes. An example would be undesired changes to the composition of plant substances. If such an effect is detected, the authorisation process is halted. The same applies to the authorisation of genetically modified plants, except that these are analysed in a much more targeted and comprehensive manner. Transgenic plants are examined at transcript, proteome and metabolome level (totality of proteins and metabolites) and in terms of external appearance. So if we want to talk about "safe" plants then we are particularly justified in doing so in the case of genetically modified plants. The safety assessment of genetically modified plants also focuses on epigenetic effects, and in a much more targeted manner than with conventional crosses in which entire genomes are mixed. With each new variety – whether produced conventionally or using genetic engineering methods – the assessment/authorisation concept involves years of studies at different sites. It also looks at the stability of expression. Undesired epigenetic effects can therefore be ruled out to a large extent.

GMO Safety: What do new findings about epigenetic effects mean for the safety of transgenic plants?

Jens Freitag: We know that molecular biological processes are much more complex than was thought a few years ago. But does this mean that the consequences of genetic modifications are less predictable? Jens Freitag: Every increase in knowledge is positive. It enables new insights into nature and allows us to try new approaches with greater awareness and in a more targeted manner. Our growing understanding about epigenetic factors has enabled us to open a new book in the study of molecular regulatory mechanisms. I find that above all extremely exciting and scientifically very pleasing. However, I fail to understand why knowledge about the complexity of natural processes should go hand in hand with an increase in risk. Just because I cannot analyse such complex processes in natural hybridisations does not make them safer or better per se than a much more specific genetic modification. Turning this argument around would mean ignorance is bliss. I imagine that the motivation for research and innovation is somewhat different.

GMO Safety: Epigenetic effects are not restricted to transgenic plants. They also occur in conventionally bred plants. In other words, progeny can have characteristics that are not directly derived from those of the parents. Do the new findings also affect classic plant breeding?

Jens Freitag: Plant breeding has for a long time been making use of epigenetic phenomena – unknowingly of course. The aim of the current research is to understand these phenomena better in their complexity. Knowledge gain is the basis for targeted application in the breeding process.

At the moment plant research is undergoing a renaissance in public consciousness. Some of the buzzwords are climate change, renewable raw materials and safer, healthier food. However, even now we can see greater competition for land emerging between the production of energy and material resources on the one hand and food production on the other. Yield increases and yield stability in crop production are also based on advances in plant breeding. For modern plant breeding focused on new goals, molecular biological research – and therefore genetic engineering – have for years been unavoidable, integral tools.

Scientifically, transgenic plants are perfect for the specific investigation of epigenetic effects. Systematic studies on model plants, such as thale cress (Arabidopsis thaliana) and rice provide the basis for this research. This kind of research would currently be possible only at immense cost, if at all, using conventional crop plants. The key is for this knowledge to be transferred successfully from the model systems to the large number of different crop plants. Research into epigenetic effects is an excellent example of the potential offered by a combination of fundamental research and application-based research into crop plants.

GMO Safety: Thank you.

Debate

"Epigenetic effects occur with every hybridisation."

More from GMO Safety