ISIS Report 23/3/2004
Bt Toxins in Genetically Modified Crops: Regulation by Deceit
Prof. Joe Cummins reviews the
impacts of Bt toxins and Bt crops and points to a fundamental flaw in their
regulatory assessments - toxicity testing based, not on the toxins in Bt crops
themselves, but on surrogate toxins. There is, furthermore, evidence that some
Bt toxins are toxic to mammals.
If you wish to see the complete document
with references, please consider becoming a member or friend of ISIS. Full details here
Bacillus thuringiensis (Bt) toxin genes inserted into genetically modified
(GM) crops are, along with herbicide tolerance, the leading modifications of
food crops. Bt crops were planted on over 62 million hectares worldwide as of
Bt bacteria store multiple toxin proteins as crystals in spores. The individual
toxin genes have been isolated and cloned; each of the toxin genes and proteins
are related, but differ in the range of insects that each poisons. The main
crystal toxins are designated Cry, then individual toxins are designated Cry1,
Cry2 etc. A particular toxin such as Cry1 may have alternate forms, designated
Cry1A or Cry1B, which differ significantly in gene sequence. Finally, small
differences in gene sequence may reflect significant difference in specificity
and the final designation is Cry1Aa, cry1Ab, etc.
Each toxin that modifies a crop is normally modified in its DNA sequence from
the natural toxin by the introduction of regulatory sequences such as introns,
polyA signals, promoters and enhancers. The DNA sequence for the toxin is altered
from the natural gene to make the gene more active in the crop and in many instances
the amino acid sequence of the toxin is altered to make the toxin more soluble
in the plant cell .
Each toxin in a GM crop must be evaluated separately from other toxin genes
and proteins, making regulatory evaluations complex. But, in every case, the
Bt crops released in North America have been evaluated based on the toxicity
to mammals and to the environment of the natural toxins, not the product
of the synthetic altered genes in the GM crops. Regulators have simply assumed
that the toxins produced using the altered synthetic genes are equivalent to
the natural gene toxin so long as the altered toxins contain domains for insect
toxicity and they had an immunological relationship to the natural toxin .
Therefore, the actual toxins in the GM crops have not been tested.
This is because the cost of isolating the toxins from the GM crops was considered
Toxicity to mammals
The Cry toxins have a common mode of action in insects. The toxin proteins
bind to cell membranes of the cells of the insect gut. The receptors for the
toxins have been identified as membrane-anchored aminopeptidase enzymes and
cadherin-like proteins. Cry toxins form ion channels that cause an efflux of
potassium ions from the insect gut cells, leading to cell lysis (the cell breaking
open) . The actual aminopeptidase binding sites for the Cry toxin is glycosylated
(short carbohydrate molecules are added to the protein) and recognized by a
lectin-like protein domain on the toxin [5-8]. Lectins are a class of proteins
that bind to carbohydrates associated with proteins. They are usually recovered
from plants and many are known to affect mammalian cell growth while others
are toxic to plant predators such as insects.
The toxicity to mammals of relatively few of the numerous Cry toxins has been
reported in the scientific literature. Senior scientist Dr. Arpad Pusztai has
prepared a superior review of the health risks of GM food, which included a
comprehensive section on Bt toxins. Areas covered in the review included an
earlier report from an Egyptian laboratory showing the Cry1 toxin, either fed
alone or in transgenic potatoes to mice, led to hypertrophic and other changes
in gut ultrastructure. Pusztai pointed out the need for fuller and much more
extensive animal feeding studies on GM crops .
Dr. Mae-Wan Ho has reviewed recent findings on the mammalian toxicity of Bt
toxins. Her reports include observations on the death of cows fed GM fodder,
survival of transgenic DNA during digestion and binding of Bt toxin to the intestine
of mice [10-12]. Some of the studies in those reports are mentioned below.
Cry1Ac toxin was observed to bind to the cell surface proteins of the mouse
small intestine and caused changes in the physiological state of the intestine
. Vaginal and intraperitoneal immunization with Cry1Ac toxin elicited antibody
response at several mucosal sites, including the vagina. At the large intestine,
the antibody response changed during the oestrus cycle, while the vaginal response
did not change throughout the reproductive cycle . Intranasal, rectal and
intraperitoneal immunization with Cry1Ac toxin induced serum, intestinal, vaginal
and pulmonary (lung) immune response in mice . Cry1Ac toxin was a potent
immunogen, more potent than cholera toxin [16, 17].
These few studies have made important breakthroughs on the impact of Bt toxins
but are seldom followed up vigorously, a serious mistake considering the widespread
consumption of unlabeled foods containing Bt toxins. Furthermore, the adverse
findings seem to be seldom mentioned in regulatory reviews.
The behaviour of transgenes and toxins in the mammalian digestive system is
crucial to evaluating their impact on the animal. Cattle were fed maize silage
containing Cry1Ab toxin. After four weeks, the contents of their digestive system
and faeces were analysed. The low-copy Bt genes could not be quantified in the
digestive system, but the Bt toxin protein was detected in the digestive system
and faeces of the cattle .
Pigs fed maize containing Cry1Ab were found to have quantities of the Cry toxin
genes and toxin protein; and Cry1Ab protein was not totally degraded in the
digestive system . Pigs fed StarLink (Cry9c) maize were found to have about
a quarter of the ingested Cry genes in their rectal material, showing that the
genes were only partly degraded during digestion .
It is clear that Cry toxin genes and proteins are not entirely digested in
animals fed GM maize. The impact of the genes and toxin proteins on the animals
deserves much fuller study. As GM foods are not labelled in North America, it
has not been possible to determine whether or not feeding humans and animals
has had an adverse impact. Clearly, the DNA and toxin proteins studied in the
feeding experiments are the ‘real thing’, not the bacterial proteins used as
surrogates in the toxicity testing approved by regulatory agencies.
Toxicity to non-target organisms
The impact of Bt toxins and the genes determining them on non-target organisms
in the environment has been studied to some limited extent, but nowhere as much
as is needed. The soil around the GM crop may accumulate toxins, if these are
released to the soil.
Bt toxin is found released to the soil by maize plants both in the laboratory
and in the field . The toxin is released in root exudates from a number
of maize hybrids expressing three different transformation events . Bt toxin
released from exudates is bound to soil particulates and is active for at least
180 days . Bt toxin is not taken up from the soil by plants, not even from
hydroponic growth media . It has been reported that Bt maize exudates had
no effect on earthworms, nematodes, protozoa, bacteria and fungi in soil .
However, a recent study showed that the litter from Bt maize, while not fatal
to earthworms, caused a large weight loss in worms exposed for over 200 days
to the litter containing Bt toxin .
Some Bt crystal proteins have been reported to target nematodes while the toxins
targeting Lepadoptera or Coleoptera insects do not appear to target nematodes
, but further investigation is needed. Soil nematodes include both plant
pathogenic species and species that eat insect pests (entomopathogenic). The
destruction of the latter by Bt crops would have disastrous economic consequences.
Beneficial predators may be destroyed by Bt crops and their loss would be costly.
The green lacewing is an important predator of the insect pests of maize. Insect
herbivores feeding on Bt maize were fed to green lacewings. Insect herbivores
that ingested little Bt Cry1Ab toxin did not affect lacewing survival, while
herbivores that ingested a quantity of Cry1Ab toxin caused low survival in lacewings
and delayed development among survivors .
To test the concern that consuming herbivore insects may have produced “indirect
toxicity to the lacewing predator”, the herbivores were treated with high levels
of the natural Cry1Ab toxin in the laboratory, then fed to green lacewing predators.
The lacewings that were fed insects treated with natural Cry1Ab toxin turned
out to be far less severely affected than the lacewings fed insects feeding
on Cry1Ab transgenic maize [28, 29]. The researchers stated that their procedure
was to “give evidence that Bt-maize poses no threat to this predator” .
But they failed to indicate that the natural Cry1Ab toxin was not identical
to the Cry1Ab toxin synthesized from a synthetic gene in Bt maize, and the Bt
maize proved to be far more toxic. Whether it is the Bt toxin in Bt maize that
is responsible for the toxicity, or something in the transgenic process itself
is not known.
The negative impact of Bt maize pollen on the survival of the monarch butterfly
has been extensively reviewed and will not be discussed further here .
The manner in which mammalian toxicity and environmental impact of Bt crops
is evaluated is spelt out in the United States Department of Agriculture (USDA)
Animal and Plant Health Inspection Service (APHIS) reports on the deregulation
of GM crops that had been field tested. With both Bt insect resistant crops
and herbicide tolerant crops, approval was not based on the Bt toxin proteins,
nor on the bacterial enzymes providing herbicide tolerance in the crop, but
on Bt toxin proteins or enzymes isolated from bacterial cultures.
The Bt toxins in bacterial cultures were produced using genes that differed
from those used in the GM crops. The proteins were significantly altered in
amino acid sequence from those in GM crops. The regulatory agencies and their
advisory committees argued that so long as the bacterial products retained their
active domains as toxins or enzymes and had similar immune profiles to the proteins
produced in GM crops, they were “substantially equivalent” to the proteins produced
in GM crops.
For example, Cry1Ab toxin gene in maize was tested using a toxin produced in
E. coli bacteria that differed from the protein produced in the GM
crop . Maize altered with Bt gene Cry3Bb1 was similarly tested using the
bacterial protein . Maize modified with Cry1F and a gene for herbicide tolerance
was approved based on testing the bacterial - not the crop - proteins for safety
towards mammals and non-target organisms . Cotton modified with Bt toxin
gene Cry2Ab was approved based on studies of the surrogate product produced
in bacteria . Potato modified with Bt toxin gene was tested, as in the cases
above, using the bacterial surrogate protein, not the protein in the crop .
These examples are representative of all risk assessments for Bt crops. The
practice of testing surrogates for toxins and enzymes produced in GM crops is
unsound in the light of the millions of people and animals being exposed to
the products. Such careless procedures have been made possible by the absence
of labels on GM food and feed, which makes tracing impacts of the products difficult,
if not impossible.
The second report of the UK’s GM Science Review Panel (2004) commented, “Many
of the genes introduced into GM plants are based on bacterial gene sequences,
but are synthesized de novo in the laboratory to include more appropriate
codon usages for more efficient expression in plants” . However, the report
failed to mention that the evaluations of mammalian safety and environmental
impacts (particularly the impact on non-target organisms) have been done using
a bacterial surrogate for the proteins produced in GM crops. Consequently, GM
crops produced in the United States, Canada and other countries are untested
and unknown for toxicities, and the failure to label the GM foods produced from
the GM crops has obscured any impact on humans and animals.
Apparently, this makes the approval of the GM crops illegal.