Survival and dispersal strategies of genetically modified bioluminescent rhizobia in the field

(1996 - 2000) Bielefeld University, Chair of Genetics, Bielefeld; Federal Agricultural Research Centre (FAL), Institute of Agroecology, Braunschweig; TÜV Südwest Dienstleistungs-GmbH, Biological Safety, Freiburg; Friedrich-Alexander University Erlangen-Nuremberg, Institute of Microbiology, Biochemistry and Genetics, Erlangen

Topic

In order to investigate their behaviour, transgenic rhizobia with a luciferase marker gene were released. This gene from glow worms causes the bacteria to produce the enzyme luciferase. The luminescent effect enables colonies of the genetically modified rhizobia to be identified.

Rhizobia (nodule bacteria) live in symbiosis with legumes. They fix nitrogen from the air by converting it into a form that the host plant can utilise.

As part of a long‑term monitoring scheme of genetically modified rhizobia in the soil, the project aimed to address the following questions:

How do the released strains disperse and survive?
Are the native (endogenous) bacterial populations influenced by the released bacteria?
Does horizontal gene transfer take place from the native bacteria to the released Rhizobium strains?
Are the released rhizobia genetically stable?
Does the plant species cultivated (alfalfa and rye) affect the survival ability (persistence) of the released genetically modified rhizobia?
Do the transgenic rhizobia accumulate on non-host plants?

To this end, the field trials started in the previous joint project were continued and extended to an additional site.

Two transgenic Rhizobium meliloti strains (recA+ und recA-) carrying the luciferase marker gene were released. In the recA- strain, the luc gene is integrated into a recombinase gene (recA), which makes this strain recombination-defective. This has the effect of prohibiting the natural repair of gene mutations .

Further information:

SiFo joint project: Investigating possible effects of genetically modified rhizobia on soil ecology

Summary

Dispersal/Survival. Contrary to expectations, the recA gene does not appear to be essential for survival in the soil, since both strains survive equally well. However, the distribution of the recA-defective strain is limited.

Effect on endogenous bacterial populations. The impact on the endogenous bacterial population of released bacteria of the same species is judged to be slight and only temporary.

Horizontal gene transfer to the Rhizobium strains. No evidence of horizontal gene transfer to the released strains was demonstrated.

Stability of the released rhizobia. The marker gene proved stable in both transgenic strains.

Effect of alfalfa and rye on the persistence of the transgenic rhizobia. The released bacteria can become established in the soil when their host plant is cultivated.

Experiment description and Results

Soil samples were takes four times a year from the plots of the previous release experiment (joint project). Alfalfa was planted at the second site before the release in order to build up a native (endogenous) Rhizobium population there. Each plot was inoculated with one of the two transgenic strains, with the wild type strain. An uninoculated plot served as a control.

Three plants were dug up from each type of plot at each sampling. The bacteria attached to the roots were removed and separated and their DNA was extracted and purified.

Persistence

Survival: To study the persistence, the number of transgenic bacteria in the soil samples was determined by means of the bioluminescence.

Dispersal: The soil was analysed to a depth of 65 centimetres once a year. Soil samples were also taken from outside the trial plots.

At the second site there was heavy rainfall on the day following the release. Samples were therefore taken immediately after the release from areas with clearly discernable water traces.

Airborne micro-organisms were also ‘captured’ there during the release.

Survival: At the first site, both strains became established in the plots inoculated with them. At the second site, the released bacteria fell by about half.

Dispersal: Both strains were found only down to a depth of 25 centimetres. Isolated recA+ strains were found outside the trial plots on alfalfa plants at the first site.

At the second site transgenic bacteria were initially found only in the immediate vicinity of the water traces. Later, isolated incidences of luminous bacteria were found.

During the release itself, the airborne micro-organisms doubled. However, only 0.1 per cent were the released strains, which means that the dispersal of these strains by air is judged to be very low.

At the second site some soil was transported by mice. Here too samples were taken. No transgenic bacteria were found.

Effect on endogenous bacterial populations

To analyse the competition between endogenous and released rhizobia in colonising alfalfa root nodules, the composition of the rhizobia isolated from the nodules was investigated.

At the second site, the rhizosphere population was also assessed using various culture-independent methods.

At the first site almost all nodules were colonised by the released strain, since no endogenous population had been present before the release. After three years an endogenous population had become established, but it did not displace the released strains. The two transgenic strains did not differ noticeably.

By contrast, at the second site an endogenous Rhizobium population had been present before the release, so the released strains were found in only 21 per cent of nodules; one year later they were found in only five to ten per cent of nodules. There was no difference in the endogenous Rhizobium population between the various plot types.

In general, a high diversity was found with a large proportion of rhizobia.

Because of physical influences (weather) and biological influences (plant growth) clear fluctuations were observed in the population, although these could not be attributed to the release.

Horizontal gene transfer

To assess gene transfer from the endogenous bacterial population to the released strains, over 2000 reisolated genetically modified rhizobia were tested by PCR and selection for uptake of plasmids that confer mercury resistance.

The stability of the reporter gene (luc) was investigated using PCR and bioluminescence.

No horizontal gene transfer of mercury-resistant plasmids to the transgenic strains was found.

Stability: The luc gene is expressed stably in both strains.

Impact of alfalfa and rye

At the second site, part of the inoculated and uninoculated plots were mown and dug up. Rye was sown on these parts following usual agricultural practice. Rye was also sown in microcosms (closed containers) with the soils of both sites. The survival, spread and possible accumulation of the rhizobia in the rhizosphere and soil were determined as a function of the plant species and type of soil.

No differences in the number of transgenic rhizobia in the soil of alfalfa and rye plots were found, i.e. the absence of the host plant produced no effect on the persistence of the released rhizobia in the space of a year. There was a greater accumulation of rhizobia in the rhizosphere of alfalfa than in the rhizosphere of rye in both soils. The soil type had a much lesser impact than the type of plant.

Soil investigations

Soil ecology parameters, like nitrate content, organic carbon content, microbial biomass and the physiological profile of the microbial soil community, were also analysed.

The soils differed in carbon content (higher at the second site than at the first). The nitrogen content was the same. The released strains could not be detected in the physiological profile of the soils.

Summary. The impact of released bacteria on the native bacterial population at a site is judged to be slight and only temporary. The released bacteria can, however, become established in the soil if their host plants are cultivated.

The recombination-defective bacterial strain proved to have limited effectiveness as an additional safety measure in the field since, like the recA+ strain, it could still be detected after more than 3.5 years. However, its horizontal dispersal in the soil was limited.