Outcrossing and spread of cultivated plants genes

“Gene flow is a biological principle and does not constitute damage”

Using mathematical models, American scientists have calculated that genes from cultivated plants can spread to wild populations within a few generations, even if these genes offer no survival advantage to the plant. This has served to further fuel the debate surrounding the risk assessment of transgenic plants. GMO Safety discussed the results of the study with Detlef Bartsch from the Robert Koch Institute in Berlin.

Dr. Detlef Bartsch ist stellvertretender Leiter des Zentrum Gentechnologie am Robert Koch Institut Berlin (inzwischen Bundesamt für Verbraucherschutz und Lebensmittelsicherheit; BVL)

Wild beet (beta maritima) in the demonstration garden at the Max Planck Institute for Breeding Research in Cologne. In the Italian Pianura Padana region, where sugar beet seed is propagated, wild relatives of sugar beet are found. Even genes that confer no advantage on the plants, such as those which ensure that the beets flower only after two years, have established themselves in the wild population. So far, however, there has been no decrease in the population.

GMO Safety: The fact that genes which confer no advantage on the plant eventually disappear from the gene pool is a familiar argument in the risk assessment of vertical gene transfer to related wild species. The authors of the study are now emphasising the fact that even genes that offer no advantage to the wild plants can become established in wild populations. In what conditions is this possible?

Bartsch: It depends on the extent of the gene flow. If ‘neutral’ genes are involved, in other words genes which offer neither advantages nor disadvantages for survival, even low transfer rates can quickly lead to long-term establishment.

I like to use the “bath tub model” as a simple illustration: if you equate the total amount of DNA for a population with the quantity of water in a full bath, the colour of the water will slowly change if water containing red dye, for example, is slowly added from a tap. But clear water is also flowing in from many other feed pipes (these represent the gene flow from neighbouring wild plants), so the water may turn pink. At the same time, water is draining away, since the total quantity of water (for water read DNA) remains the same. Both neutral and advantageous genes are retained in the bathtub population, as it were. In the model by Haygood et al. the other feed pipes are discounted in order to simplify matters.

GMO Safety: In slightly exaggerated terms, then, the result of the mathematical model is that the wild populations sooner or later decline and are ‘infiltrated’ by foreign influences from the cultivated species. What happens in reality?

Bartsch: ‘Foreign infiltration’ is not a good description. First the genetic material is thoroughly mixed, resulting in an ‘enrichment’ of the DNA. If a very strong and dominant gene flow takes place from the cultivated to the wild forms, then in extreme cases displacement of rare genes in the wild population is to be expected. It also depends to a large extent on the reproduction mechanism of the plant. Self-fertilising plants are generally immune to gene flow through pollen, but could become displaced by highly competitive cultivated plants. The latter case is highly unlikely, since cultivated plants are poorly adapted to life in the wild. Cross pollinators, on the other hand, generally have a distinct tendency to mix genetic material. Provided that the influx of cultivated genes is not too high, these cultivated genes will simply be absorbed.

In the bath tub model, the water would slowly turn pink without necessarily impairing the water quality. In population genetics the random loss of rare genes is referred to as “genetic drift”. If genes are continually coming in from outside, this effect could increase and could indeed lead to the extinction of the wild population in extreme cases. This has been observed in some regions of the world in gene flow from conventional cultivated rice to wild rice. But in practice we also know of examples where even an extremely strong gene flow from sugar beet to wild beet has resulted in an enrichment of the genetic diversity rather than foreign infiltration. Generalisations are therefore barely feasible and mathematical models must always be checked against reality.

GMO Safety: The authors stress that their mathematical model applies both to conventionally bred and genetically modified plants. Is this equation valid or are there in fact some differences?

Bartsch: Although they are largely similar, there is one small yet subtle difference. With conventional plants, for every allele (gene locus) in the donor plant there is a corresponding allele in the recipient plant. Advantageous alleles can assert themselves. This can result in a genuine exchange that may eventually culminate in displacement. In transgenic organisms on the other hand, the newly inserted DNA has no real counterpart in the wild plant. This is referred to as a ‘null allele’ in the wild plant. Consequently, in strong cross-pollinators there is really nothing to displace initially. It’s a different matter, however, if linked gene groups that cannot be separated from the transgene are located in the immediate vicinity of the transgenic site on the genome. Then, in certain circumstances, a few wild genes could disappear. It all depends on the balance between genetic recombination and selection. At this point it becomes very complicated and we are moving away from the model referred to in the publication.

GMO Safety: If transgenes manage to become established in the gene pool, how should that be evaluated in your view?

Bartsch: In January 2003 we had a large international congress in Amsterdam on this topic. The speakers agreed that in many cases transgenes will become established naturally, because gene flow is a fundamental biological principle and does not constitute damage. The consequences must be considered on a case-by-case basis. In my view there is nothing to support a wholesale negative assessment.