You are here: Agriculture > Transgene Plants > Forum 1995 > Horizontal Genetransfer

Horizontal gene transfer from transgenic plants into plant associated microorganisms and soil microorganisms

Table of Contents:

• Introduction
• Persistence of transgenic plant DNA in soil
• Detection of horizontal gene transfer from plants to soil bacteria
• Occurrence of nptII in environmental bacteria
• Conclusion

Author: K. Smalla

Introduction

Bacterial antibiotic or herbicide resistance genes are still frequently used as selectable markers for transgenic plants. Due to well-known problems caused by antibiotic resistant pathogens, concerns arose about the large-scale use of transgenic plants containing antibiotic resistance genes. Therefore, horizontal transfer of such resistance genes from plants to microorganisms has often been discussed as potential unwanted effect of transgenic plants on the soil microbiota. However, there is no clear evidence for gene transfer from plants to microorganisms to date. Biosafety research on horizontal gene transfer from transgenic plants to microorganisms (bacteria and fungi) targets two main aspects: the scientifically challenging question of whether in principle gene transfer from plants to microorganisms is possible and detectable under field conditions, and the assessment of ecological consequences. Studies on both points under field conditions are impeded by the fact that only a minor fraction of the microbial community is accessible to cultivation techniques. In view of the expected scarcity of horizontal gene transfer from plants to microorganisms representative soil sampling and a high sensitivity of the detection methods applied are of particular importance to reduce the probability of gene transfer events remaining undetected. To avoid false positives detection methods used have to be highly specific allowing the unequivocal differentiation of the construct from naturally occurring resistance genes.
The objective of this contribution is to describe studies tackling the question of whether gene transfer from plants to microorganisms does occur, and to which extent. Experimental approaches and difficulties impeding the detection of such gene transfer events under field conditions will be discussed. Furthermore, the results of a study on the natural occurrence of the nptII gene, the most frequently used selective marker for transgenic plants, will be summarized.

Persistence of transgenic plant DNA in soil

Transgenic plant DNA can be released into the environment e.g. from senescent or rotting plant material. The persistence of free DNA released in soil is of great importance with respect to potential gene transfer by transformation. If the released free transgenic plant DNA is immediately degraded by extracellular nucleases, transgenic DNA is unlikely to be taken up by competent soil bacteria. On the other hand, long-term persistence of DNA might enhance the likelihood of transformation-like processes to occur. Numerous reports have shown that free DNA can persist in soil for a longer period and still remains transformable. Obviously, DNA adsorbed to soil minerals is to some degree protected against the attack of nucleases. Several abiotic factors such as the concentration of bivalent cations, the mineral content or pH of the soil and the DNA conformation will affect the degree of the adsorption of the DNA to the soil surface. Therefore, persistence of released transgenic plant DNA is supposed to be different in different agricultural soils and depending of the microbial activity and composition of the respective soils. Persistence of transgenic plant DNA in soil was investigated for transgenic tobacco (aacC1, Paget et al., 1993), transgenic petunia (NOS-nptII, Becker et al., 1994) and transgenic sugar beets (bar/TR1, TR2/nptII, 35S/BNYVV-cp, Smalla et al., 1995). The persistence of the construct in soil was tracked by direct DNA extraction from soil followed by PCR-based amplification of the construct. Appropriate primer selection allowed an unequivocal detection of the transgenic construct besides the naturally occurring genes. With this methodology the presence of the construct could be detected but no information was gained on its localization inside rotting plant material, as free DNA adsorbed to soil surface, in metabolically active, dormant or dead cells. The sensitivity for detection of the construct in total soil DNA extracts should be determined since it will vary with the soil type, the lysis and purification protocol and PCR conditions. Limits of detection were determined for the construct used in transgenic sugar beets with three different primer sets (bar/TR1;TR2/nptII, 35S-BNYVV-cp) to be around 10² target sequences per gram soil (Gebhard et al., unpublished). Transgenic sugar beet DNA was detectable in soil samples taken from a disposal site 6, 12 and 18 months after transgenic sugar beets had been ploughed into soil. Whereas for soil sampled 6 months after shredding and incorporation of transgenic sugar beets into the soil PCR products were obtained with all 3 primer sets, no PCR products were detected in DNA extracted from soils sampled 12 or 18 months after incorporation using the 35S/cp primer system. Becker et al. (1994) reported that DNA from transgenic petunia could be only detected in 3 soil samples taken 2 months after the petunia plants were ploughed into the soil. Paget et al. (1993) found that transgenic tobacco DNA remained detectable for more than one year after harvest of the transgenic plants. To compare investigations on the persistence of transgenic plant DNA in agricultural soils respective studies should contain information on the sampling procedure, on relevant soil parameters (mineral content, pH, content of organic matter), the soil DNA extraction and purification procedure, the determined limits of detection. Although only a few studies investigated the persistence of transgenic plant DNA under soil conditions long-term persistence of the constructs in soil could be demonstrated unequivocally. However, none of the studies proved whether the persisting DNA was naked DNA adsorbed to mineral surfaces, or still covered in rotting plant material.

Detection of horizontal gene transfer from plants to soil bacteria

The most probable mechanism for gene transfer from plants to microorganisms is natural transformation requiring the uptake of free DNA by naturally competent soil bacteria and the integration of the foreign DNA into the bacterial genome (Stewart, 1989). In order to detect gene transfer from plants to microorganisms under field conditions, there must not only be mechanisms to allow uptake and replication in the new host but perhaps most importantly, a selection for the host expressing a new trait. Even though competent soil bacteria are known, there are only a few reports that could show that natural transformation does occur in nonsterile soils (Lorenz et al., 1992, Gallori et al., 1994). Detection of horizontal gene transfer events can be performed by analysing bacteria after a prior cultivation step. In order, to get information on the presence of the construct in nonculturable bacteria the bacterial fraction recovered directly from soil can be analysed for the transgenic DNA. However, screening of cultivated bacteria for the presence of the construct has the advantage that putative transconjugants can be further characterised. Provided that the transferred resistance genes are expressed in their new hosts selective cultivation can be applied. In a project accompanying the field release of rhizomania resistant transgenic sugar beets containing the nptII and the bar gene under the control of the TR1/TR2 promotor and the BNYVV coat protein gene under the control of the 35S promotor soil bacteria were screened for horizontal gene transfer events. Since the level of expression of the nptII gene in different soil bacteria was unknown, resuspended soil samples were plated onto nutrient media containing 100 æg/ml, 10 æg/ml or no kanamycin. Approximately 3,000 soil bacteria isolated from different samplings were randomly picked and analysed by cell hybridisation using the construct as probe. Several positive colonies giving slight positive signals were further confirmed by PCR. None of the colonies contained the construct. To improve the sensitivity of the detection total DNA was extracted from bacterial lawns growing at the lowest dilution, analysed by PCR and hybridisation. Until now the construct was never detectable in the fraction of bacteria grown on plate count agar. To get information on gene transfer into the nonculturable fraction of soil bacteria DNA the bacterial fraction was obtained from soil samples by different blending and centrifugation steps. DNA extracted from the bacterial fraction was analysed by PCR. However, positive PCR signals are difficult to interpret since it is almost impossible even after several DNase treatment steps to ensure the absence of free DNA. Especially in soils with a high content of clay particles a complete separation of bacteria from clay particles is difficult, and DNA adsorbed to clay minerals will resist DNase attacks to a certain extent.

Under laboratory conditions Becker et al. (1994) analysed hygromycin and kanamycin resistant bacteria isolated from a soil microcosm 7 days after inoculation with transgenic tobacco homogenates containing the Tn5-nptII or Tn5-hph. None of the isolates contained the construct. Transformation of competent Bacillus spp. and Pseudomonas spp. with transgenic plant DNA failed. Only when the host (Acinetobacter calcoaceticus) contained homologous sequences the stable integration of a linearized plant transformation vector could be shown. Broer et al. (1994) reported that transformation experiments of competent Agrobacterium tumefaciens with linear single stranded DNA or transgenic plant DNA did not result in any transformants. While the transformation frequencies with circular plasmid DNA carrying the same construct as the transgenic plant DNA was about 2 x 10-7. The analysis of gentamycin resistant colonies isolated from Agrobacterium-caused tumors revealed no horizontal gene transfer. The authors concluded that the induction of virulence genes and the presence of plant material did not stimulate gene transfer. Recently investigations on horizontal gene transfer from different transgenic plants carrying the hph gene to coinoculated Aspergillus niger were published by Hoffmann et al. (1994). With one exception the foreign DNA seemed to be not stable in the putative Aspergillus niger transformants. The mechanism of transfer and integration remained unclear.

Occurrence of nptII in environmental bacteria

For an adequate evaluation of selective marker genes used in transgenic plants data on the natural dissemination of antibiotic resistance genes used in transgenic plants should be known. Data on the prevalence of the kanamycin resistance phenotype in different habitats such as soil, sewage, river water, manure slurries and the proportion of nptII carrying, kanamycin resistant bacteria were recently published by Smalla et al. (1993, 1995). In addition, total community DNA extracted from the different environments was tested by PCR for the presence of nptII. The proportion of kanamycin resistant bacteria was highest in manure slurry samples (38%), whereas for soils up to 1% of the total colony forming units showed a high level Km-resistance. Most Km resistant bacteria carrying the nptII were isolated from sewage, fewer strains from manure slurries and river water. A high proportion of Km-resistant isolates with nptII homology belonged to the Enterobacteriaceae. However, up to now more than 3,000 Km-resistant soil bacteria were screened for the presence of nptII, but none of the Km-resistant soil bacteria showed nptII homology. PCR amplifification of total DNA extracts with nptII-specific primers revealed the presence of nptII in some soil samples.

Conclusion

Recently developed molecular approaches allow a sensitive and specific tracking of transgenic DNA in soil. However, detection of gene transfer into bacteria not accessible to cultivation still remains complicated. Although transgenic plant DNA is detectable in soil for a longer period horizontal gene transfer from plant to bacteria was not detectable under laboratory and field conditions until now. The present results of investigation focussed on horizontal gene transfer from plants to microorganisms allow to conclude that such transfer events, in case they occur at all, will be rare. Furthermore, the ongoing environmental release of nptII containing bacteria with sewage or manure slurries into agricultural soils did not result in a dissemination of the nptII gene amongst soil bacteria. This fact might serve as an argument to ease concerns on the possible impediment of medical therapies through the use of the nptII gene in transgenic plants. Since alternative approaches such as positive selection or are expected to substitute antibiotic resistance marker genes in transgenic plants soon, the topic of horizontal gene transfer from plants to microorganisms will get less and less public attention.

This article is from 10/1995.