GENET archive


2-Plants: New GE plants exhibit "seed lethality"

------------------------------- GENET-news -------------------------------

SOURCE: IBS News Report, USA, by Johann P. Schernthaner
DATE:   July 2003

--------------------- archive:


Gene flow between related commercial crops, weeds, and wild relatives is
a recognized risk in the commercialization of GM crops. Given an increase
in acreage and variety of GM crop cultivation in the future, it is
evident that containment systems will be required that restrict the
unwanted movement of transgenes. This is of particular concern for crops
where seed is saved, since seed formed by cross-pollination from
varieties carrying novel genes can lead to populations in which the novel
gene becomes established. A number of solutions has been proposed to
alleviate this problem, but apart from male sterility, no application so
far has found its way to the farmers' fields1. One of the most promising
containment techniques still remains the transformation of chloroplasts,
which restricts the inheritance of transgenes to the female part of the
plant reproduction organs, thus eliminating pollen as carrier of
transgenes. While this approach can be very useful, it does not, on the
other hand, generally solve the problem of unwanted outcrossing because
the movement of transgenes is not always restricted to the pollen alone.
It is generally agreed that involuntary seed dispersal also has to be
taken into account, which makes it useful to seek additional containment
solutions that focus on seed sterility rather than on maternal
inheritance in order to prevent GM plants from establishing themselves
outside their intended agricultural setting.

 We have recently proposed a method of gene containment that relies on
the maintenance of a GM variety in which the presence of two genetic
components is required to assure continuous propagation of the variety2.
These components are: i) a seed-lethal (SL) gene (that can be linked to a
novel trait), which prevents a seed containing this element from
germinating; and ii) a repressor (R) element, introduced by crossing,
that inhibits the transcription of the SL gene. The combination of the
two elements leads to repression of the SL-gene, thus allowing seed
propagation of the plant by selfing. However, pollen-mediated in- or out-
crossing will result in the separation of the two elements, and the
subsequent activation of the SL-gene in the seed embryo causing the
arrest of germination. Under optimized conditions (single SL- and R-loci
at the same location of both parental chromosomes), such a method would
be simple and efficient, and would not require any intervention under
managed conditions (Fig.1).

Figure 1. Principle of a single repressor containment system based on the
site-specific insertion of a seed-lethal gene linked to a novel trait
(SL-NT) and a repressor gene (R) into the same parental chromosomes. When
the two parent plants are crossed, offspring with the genotype (SL-NT/R)
produce viable seeds. Upon outcrossing, the two alleles will be separated
and when gametes carrying the SL-NT allele are introduced into a non-GM
plant, the resulting seed will not germinate. In this setup, the R allele
is not controlled.

In the example presented, seed lethality is achieved by the embryo-
specific overproduction of auxin mediated by the products of
Agrobacterium tumefaciens gene 1, tryptophan-2-monooxyge-nase (iaaM), and
gene 2, indole-3-acetamide hydrolase (iaaH). The combined activity of
both enzymes results in the production of the auxin, indole-3-acetic acid
(IAA). The embryo-specific phaseolin promoter from French bean, Phaseolus
vulgaris, mediates seed-specific expression of SL. Repression of the SL-
construct is conferred by the DNA-binding repressor element tetR that
acts on its binding site tetO, which was introduced into the core
promoter region of the phaseolin promoter. Both elements, tetR and tetO,
are derived from E. coli, where they are part of an operon regulating
tetracycline resistance.

 For the actual experiment, two constructs, P-SL (Phaseolin promoter - SL
genes) and P-TOP-SL (Phaseolin promoter - tetO binding element - SL
genes), were prepared in order to evaluate the effect of gene 1 and gene
2 on tobacco seed development as well as the ability to repress P-TOP-SL
with the tet repressor. All transformed tobacco plants containing SL-
constructs showed normal vegetative development and seed production.
Flowers of SL-plants were self-pollinated (selfed) or cross-pollinated
(crossed) with either untransformed tobacco (backcrossed) or with plants
carrying the tetR repressor gene (35S promoter - tetR gene)3. The
resulting seeds were germinated on selective medium to determine the SL-
phenotype (inhibition of germination) as well as the number of active SL-
loci. The inheritance of the SL- and/or R- genotype was determined by
duplex PCR using gene-specific primer pairs for gene 1 and tetR.

 Of 53 primary transformants, we recovered 41 plants containing the P-
TOP-SL construct and nine plants containing the P-SL construct,
respectively, that segregated seed-lethal phenotypes. Most importantly
was also the demonstration that the introduction of the tetO binding
sequence into the core phaseolin promoter did not affect tissue-
specificity; that both SL-constructs had no observable effect on pollen
viability, and hence normal male fertility could be observed. The
germination analysis of seeds resulting from (SL x R) crosses showed that
repression of the SL-genes by TETR occurred to a varying degree in 21 out
of 37 lines.

 PCR analysis of normally-developing seedlings from (SL x R) lines
revealed only (SL; R) and (R; R) genotypes, but no (SL; SL) genotype was
detected, indicating that the SL-loci remained functional in the next
generation. Since a silenced or non-functional SL-gene can also explain
the normal development of a plant with a (SL; R) genotype, 21 (SL; R) F1
lines were propagated to follow the segregation of the SL-phenotype. All
F2 lines, selfed and backcrossed, produced seedlings with observable SL-
phenotypes as a result of segregation of the SL-gene from the repressor
gene. This is an indication that the repression, as opposed to the
silencing, of the SL-gene generated the (SL; R) plants in the previous
generation. In total, 18 out of the 21 lines produced normal (SL; R)
plants in the F2 generation. The control crosses performed with the P-SL
plants that contain the phaseolin promoter without the tetO-binding
element provided further evidence that TETR is responsible for the loss
of the SL-phenotype. Of the nine P-SL-lines that were crossed with R-
lines, only one line produced normal (SL; R) plants. That line, however,
contained several SL-loci and showed a high number of less severe SL-
phenotypes in the progeny of the selfed T0 generation. It is possible
that the segregation of these weak SL-loci in the F1 generation gave rise
to the normal phenotype observed.

 While the repressible seed-lethal system described here is effective
under experimental conditions, it is obvious that for successful field
application the system has to be optimized. The most critical parameter
is to ensure a tight repression. In an ideal system, only one R-locus and
one SL-locus should be present; therefore, the efficacy of the repressor
and the activity of the seed-specific promoter have to be optimized to
function reliably in combination. The goal is to vary the functional
components, such as seed-specific promoters, repressor binding sites,
promoters driving the repressor, and even the repressor itself, to such a
degree that a watertight repression of the SL-genes can be assured. In
addition, in order to achieve a true containment of all transgenes, the
repressor construct itself would have to be controlled as well. This
could be accomplished by using a double repression (DR) system and
associating an additional lethal component to the R-construct (Fig. 2).

Figure 2. Principle of a double repressor containment system. Two
different repressors (R1 and R2) located on opposite sites of the same
parental chromosomes control each SL gene. Repressor 1 binds to binding
site B1 while repressor 2 binds to B2.When the two parent plants are
crossed, both SL-genes are repressed and viable seeds are produced. Upon
outcrossing with a non-GM plant, the two alleles will be separated and
the resulting seeds will be sterile. A site-specific insertion prevents
co-segregation of the two alleles.

How applicable is the system? In order to attain true containment, the
double-repressor (DR) approach is required (Fig. 2). For practical
considerations, a viable seed production method for seed producers has to
be offered. In the method presented, hemizygous SL-plants were used for
crossing with R-plants; for seed production, however, it would be more
practical to have homozygous parent plants that can be maintained. To
achieve this, additional components would be necessary. Homozygous DR
lines could be used in hybrid seed production where male sterility is
available. If hybrid seed production is not an option, this containment
system could be valuable for GM crops for which seed is not harvested,
e.g., vegetable crops, sugar beet, tobacco, fruit crops, trees, molecular
pharming, etc. Where applicable, seed-lethality could be triggered at an
early stage of seed development to prevent seed formation in cross-
pollinated crops and address the problem of non-GM food contamination
with GM material.

 In summary, we have shown that gene 1 and gene 2 from Agrobacterium in
combination with the tetO/tetR components, regulating a phaseolin
promoter, can be used to promote and repress seed-lethality. The system
described here could be used as a model to address the issue of transgene
management and preservation of specific genetic combinations in general.

1. Daniell H. (2002) Nat Biotech 20: 581-586.
2. Schernthaner JP, Fabijanski SF, Arnison PG, Racicot M, and Robert LS.
(2003) Proc Natl Acad Sci USA 100: 6855-6859.
3. Gatz C, Frohberg C, and Wendenburg R. (1992) Plant J 2: 397-404.

Johann Schernthaner
Agriculture and Agri-Food Canada
Ottawa, Ontario


European NGO Network on Genetic Engineering

Hartmut MEYER (Mr)
Kleine Wiese 6
D - 38116 Braunschweig

phone:  +49-531-5168746
fax:    +49-531-5168747
mobile: +49-162-1054755
email:  genetnl(at)