of genes from indigenous bacteria by inoculant strains at long-term release
Bacteria used as agricultural inoculants can be improved by genetic manipulation.
This includes modifications to minimise the transfer of novel genes to
other microbes in the environment. Nevertheless, indigenous bacteria can
transfer their DNA to inoculants, resulting in new, potentially undesirable
gene combinations. Rhizobia, bacteria that form nitrogen-fixing root nodules
in symbiotic association with compatible host legumes, are the most important
plant inoculants in Europe and world-wide. They are potential targets
for genetic modifications, and can be used as a model for modifying other
soil bacteria. There are several mechanisms by which bacteria can exchange
genetic material, and rhizobia are known to carry forms of DNA (conjugative
plasmids and bacteriophages) that can transfer between strains, often
conferring significant novel properties. For example, in many rhizobial
genera, symbiotic genes are known to occur on large plasmids. Plasmids
and bacteriophages may also mediate the transfer of other plasmids, genes
and transposable insertion (IS) elements.
This project investigated gene acquisition by rhizobial inoculants from
native bacteria in the field, to provide baseline information relevant
to biosafety risk assessments.
with GM rhizobia.
Approach and methodology
Genetic elements including plasmid replication origins, IS elements and
bacteriophages, in naturally-occurring rhizobial populations were isolated
and characterised. This facilitated the design of molecular tools for
detection of genetic elements, collection of information on their distribution
within populations, and the search for evidence of transfer to inoculant
strains. In the course of the project, rhizobial inoculants designed to
act as acceptors of indigenous genetic elements, were released. These
strains were re-isolated from the field together with established, previously-introduced
rhizobial populations, and screened for the acquisition of genetic elements
present in the native rhizobial population.
Main findings and outcome
Plasmid replication origins from Rhizobium leguminosarum and Sinorhizobium
meliloti were cloned and sequenced to design PCR primers to amplify
repC from field isolates. Four repC families were found
in R. leguminosarum isolates from Dijon, Bielefeld, Erlangen and
Rothamsted. Two repC groups were identified in S. meliloti,
one was widespread and was also found in R. tropici.
Primers were designed to amplify IS elements from field isolates, and
the distribution of the various elements was assessed in inoculant strains
and field populations. One R. leguminosarum IS isolate from Dijon
was identical to an independently isolated IS from Bielefeld. Similar
sequences were detected in IS from R. leguminosarum and S. meliloti,
and using primers for two S. meliloti IS, amplified related elements
were detected in R. leguminosarum. Five completely new S. meliloti
IS were found, one with 90% sequence similarity to a Methylobacterium
The majority of R. leguminosarum field isolates were lysogenic,
and bacteriophages were isolated from soil. A virulent phage isolated
from a field release site could transduce genes in one inoculant strain.
To measure gene acquisition in the field, an R. leguminosarum
strain that had lost its symbiotic plasmid, had a gusA marker gene
inserted in its chromosome. This strain, CT0370, could form nodules only
after acquiring a new symbiotic plasmid, and the nodules could be readily
identified because of their gusA expression. After field release
at Rothamsted, > 20,000 nodules were screened for gusA but no
expression was found. In addition, 1000 CT0370 re-isolates were screened
after two years, for the acquisition of repC groups I and II using
PCR, but all were negative. Similarly, 500 colonies of a previously-released
inoculant strain re-isolated after 8 years, were screened for ISRm3
but it was not detected. In Dijon, an estimated 125,000 T2 inoculant bacteria
released 7 years previously, were screened for IS and repC. One
re-isolate was repCII-positive and appeared to have a small plasmid
that was lost with subculture, so it could not be verified. In Granada,
27,000 nodules formed by an S. fredii inoculant were screened with
primers for IS which are present only in the native population, but expression
was not detected in the inoculant.
Thus, fewer than 4x10-5S. fredii acquired an IS element.
The frequency of transfer of the symbiotic plasmid CT0370 was less than
5x10-5 per donor, and the Dijon results indicate a possible
frequency of 8x10-6 per T2 recipient.
Genetic elements were found to be ubiquitous in rhizobial populations.
Related elements were present in isolates from soils throughout Europe
and in different genera, providing circumstantial evidence that transfer
occurs in the field, although quantitative data obtained over the relatively
short time periods studied was mostly negative. This has major ramifications
for the safety of introducing genetically modified organisms (GMOs) into
the environment. Another important outcome of this research was the development
of a means of identifying various genetic elements, providing a molecular
toolbox for future investigations.
Villadas P.J., Burgos P., Jording D., Selbitschka W., Puhler A.,
Toro N., Comparative analysis of the genetic structure of
a Rhizobium meliloti field population before and after environmental
release of the highly competitive R. meliloti strain GR4.
FEMS Microbiol Ecol, 21, 1996, p. 37.
Turner S.L., Rigottier-Gois L., Power R.S., Amarger N., Young J.P.W.,
Diversity of repC plasmid replication origins in Rhizobium
Microbiol UK, 142, 1996, p. 1705.
Mazurier S.I., Rigottier-Gois L., Amarger N., Characterization,
distribution, and localization of ISRl2, an insertion sequence element
isolated from Rhizobium leguminosarum bv viciae.
Appl Env Microbiol, 62, 1996, p. 685.
Hirsch P.R., Population dynamics of indigenous and genetically
modified rhizobia in the field.
New Phytol, 133, 1996, p. 159.
Selbitschka W., Jording D., Nieman S., Schmidt R., Puhler A., Mendum
T. & Hirsch P., Construction and characterization of a
Rhizobium leguminosarum biovar viciae strain designed
to assess horizontal gene transfer in the environment.
FEMS Microbiol Lett, 128, 1995, p. 255.
January 1992 - March 1996
University of York (UK)
INRA - Dijon (FR)
Estación Experimental del Zaidín
Universität Bielefeld (DE)