Truth: Genetic engineering is different from natural breeding and poses special risks

Myth at a glance

GMO proponents claim that genetic engineering is just an extension of natural plant breeding. But genetic engineering is technically and conceptually different from natural breeding and entails different risks. The difference is recognized in national and international laws.

GMO proponents claim that genetic engineering is just an extension of natural plant breeding. They say that genetically modified (GM) crops are no different from naturally bred crops, apart from the deliberately inserted foreign GM gene (transgene) and the protein it is intended to make.

But GM is technically and conceptually different from natural breeding and poses different risks. This fact is recognized in national and international laws and agreements on genetically modified organisms (GMOs). For example, European law defines a GMO as an organism in which “the genetic material has been altered in a way that does not occur naturally by mating and/or natural recombination” and requires the risks of each GMO to be assessed.1

The Cartagena Protocol on Biosafety,2 an international agreement signed by 166 governments worldwide that seeks to protect biological diversity from the risks posed by GM technology, and the United Nations food safety body, Codex Alimentarius, agree that GM differs from conventional breeding and that safety assessments should be required before GM organisms are used in food or released into the environment.3,4

In 1999 the UK’s Advertising Standards Authority ruled that Monsanto’s advertisements about GM foods and crops were misleading in claiming that genetic modification was an extension of traditional breeding methods.5

Today, few public comment forums on GM crops and foods are complete without claims from GMO promoters to the effect that “We’ve been genetically modifying crops for millennia”. This conveys essentially the same message as Monsanto’s advertisements and seems to have the same intent: to reassure the public that nothing radical or new is being done to their food. This message is scientifically inaccurate and misleading.

Indeed, industry tries to play both sides in its presentation of GMOs. It tells patent offices worldwide that the GM process is totally different from natural breeding and so the generation of a GM crop constitutes a non-obvious “inventive step”, thus making the GM crop patentable. On the other hand, it tells the public that the GM process is little different from natural breeding and that therefore GM foods are as safe as non-GM foods.

Both arguments cannot be correct. And technically speaking, the GM transformation process is radically different from natural breeding.

Natural breeding can only take place between closely related forms of life (cats with cats, not cats with dogs; wheat with wheat, not wheat with tomatoes or fish). In this way, the genes that carry information for all parts of the organism are passed down the generations in an orderly way.

GM, in contrast, is an artificial laboratory-based technique that is specifically designed to enable the transfer of genes between unrelated or distantly related organisms. It even enables the introduction of synthetic DNA into the genome of living organisms.

In an attempt to reassure the public and regulators about GMO safety, GMO developer companies are now focusing on transferring genes from a related organism or the same organism (so-called “cisgenesis”). For example, a gene from one potato may be inserted into another variety of potato. However, even in cisgenesis, a new GM gene unit may contain genetic elements from other organisms, including bacteria or viruses. Cisgenesis also involves the same laboratory methods that are used in genetic engineering and thus carries the potential for unexpected knock-on effects (see Myth 1.4).


The steps of genetic modification

The steps by which GM crops are created make it clear that genetic engineering is not an extension of natural breeding. It is not natural, as the particular combinations of genes put together in the GM gene cassette and the manner in which it is inserted into the host organism would never occur in nature.  

1. Isolation of the gene of interest

Genetic engineering confers a new trait on an organism by introducing the gene for a trait into the genome of that organism. The first step in that process is to identify the gene for the trait of interest and to isolate it. Using existing knowledge about the genome of a given organism, the gene of interest encoding the desired trait is identified and “cloned”. That means the gene is physically isolated and propagated in a GM bacterium as part of a DNA molecule known as a plasmid. The vast majority of currently commercialized GMOs are engineered to tolerate being sprayed with one or more herbicides or to produce one or more insecticides.

2. Cutting and splicing – generation of the GM gene cassette for introduction into the plant

Before being used to produce a GM plant, the gene of interest must be joined up with appropriate genetic control elements that will allow it to be switched on within its new plant host, so that it will efficiently produce the protein that it encodes. Other elements are also spliced into or around the gene for various purposes. Most prominent among the genetic control elements that are spliced to the gene of interest are “promoter” and “termination” sequences.

The promoter marks the beginning of the gene. It attracts and binds multi-protein complexes, called the gene expression machinery. This machinery reads the DNA sequence of the gene and synthesizes a complementary messenger RNA (mRNA) copy of the gene sequence. The termination element, as the name implies, marks the end of the gene and causes the synthesis process to stop.

Promoter and termination elements must be sourced from organisms that will allow them to work in the GM plant. These can be from either plants or, more frequently, plant viruses such as the cauliflower mosaic virus (CaMV). Promoters from plant viruses are usually preferred because they are more potent than plant gene promoters, allowing the GM gene to be expressed at higher levels and hence allowing higher production of the GM protein.

If the gene of interest is not from a plant (for example, if it is from a bacterium or animal), it is typically modified in other ways as well, to make it more compatible with the gene expression machinery of the recipient plant cells.

Genetic engineers use a variety of enzymes to cut DNA into specific sequences and to splice the various pieces of DNA into the plasmid that carries the cloned gene or gene of interest. The result of many cutting and splicing steps is the complete genetically engineered construct, called the gene cassette.

For example, the gene of interest in first-generation GM Roundup® Ready soy, maize, cotton and canola encodes an enzyme (CP4 EPSPS), which confers tolerance to Roundup herbicide. The CP4 EPSPS gene was isolated from a naturally occurring soil bacterium. In order to ensure that the CP4 EPSPS gene is switched on appropriately in plants, it is linked to the CaMV 35S promoter, which is derived from the cauliflower mosaic virus. The CP4 EPSPS gene is also linked at its leading end to a gene fragment called a signal sequence, obtained from the petunia, a flowering plant. This is to ensure that the CP4 EPSPS enzyme localizes to the right place within the plant cells. Finally, a sequence that functions to terminate mRNA synthesis is spliced to the end of the CP4 EPSPS gene. This termination sequence is taken from a second bacterial species, Agrobacterium tumefaciens (A. tumefaciens).

Therefore the first-generation Roundup Ready GM tolerance GM gene cassette combines gene sequences from four diverse organisms: two species of soil bacteria, a flowering plant, and a plant virus. These all end up in the genetically engineered agricultural crop. This graphically illustrates the extreme combinations of genetic material that can be brought about by the GM process. This is something that would never occur naturally.

In addition to the gene(s) that confer traits relevant to the final crop, another gene unit is often included in the gene cassette along with the gene of interest. This additional gene unit functions as a selectable marker, meaning that it expresses a function that can be selected for. Typically this is survival in the presence of an antibiotic or herbicide. The GM gene itself can be used as a surrogate marker gene if it encodes resistance to a herbicide. When the marker gene (along with the other gene(s) in the cassette) is successfully engineered into the genome of the recipient plant cells, those cells are protected from the antibiotic or herbicide. The genetic engineer can then separate the cells that have integrated the GM gene cassette from the majority of other cells in the culture by exposing the culture to the antibiotic or herbicide. Only the cells that have been successfully engineered and are therefore resistant to the antibiotic or herbicide survive exposure.

3. GM gene cassette insertion into cultured plant cells

To introduce the GM gene cassette into the genome of the recipient plant, millions of cells from that species are subjected to the GM gene insertion (transformation) process. This is done by growing cells from the recipient plant or pieces of tissue from the plant in culture in dishes, tubes, or flasks, a system known as “tissue culture”, and then using methods described below to insert the gene cassette into the recipient plant cells. This results in one or more of the GM gene cassettes being inserted into the DNA of some of the plant cells present in the tissue culture. The inserted DNA is intended to re-programme the cells’ genetic blueprint, conferring completely new properties on the cell.

The process of inserting the GM gene cassette is carried out in one of two ways. The first way is with a “gene gun”, which randomly shoots microscopic gold or tungsten nanoparticles coated in GM DNA into the plant cells in a process called particle bombardment or biolistics. In a few instances, the nanoparticles end up in the nucleus of the plant cells and in an even smaller number of cases, the DNA on the particles gets incorporated into the DNA of the plant cell. This is a completely random process that genetic engineers have no ability to control. They do not fully know what processes are involved in the DNA insertion process and have no control over when it occurs or where in the DNA of the plant cell it will occur.

The second mechanism of gene insertion is by infection of the cultured cells with the soil bacterium A. tumefaciens. In its natural form, A. tumefaciens infects plants at wound sites, causing grown gall disease, a type of tumour. The infection process involves the actual insertion of DNA from A. tumefaciens into the DNA of the infected plant. The genetic engineer uses the natural ability of A. tumefaciens to insert DNA into the genome of infected plants to insert the GM gene cassette into the DNA of plant cells in culture. This is done by first linking the GM gene cassette to a piece of A. tumefaciens DNA called the Ti plasmid. This modified DNA is then introduced back into A. tumefaciens. Then the plant cells in culture are infected with the A. tumefaciens that contains the GM gene cassette-Ti plasmid DNA complex.  A small fraction of the plant cells exposed to the A. tumefaciens are successfully infected and incorporate the GM gene cassette into their own DNA. As with biolistics, the A. tumefaciens insertion process is random and the genetic engineer has no way of controlling where in the plant cell genome the GM gene cassette will be inserted. It is hit or miss.

At this point in the process, the genetic engineer has a tissue culture consisting of millions of plant cells. Some will have picked up the GM gene cassette, whilst the vast majority will not have done so. The genetic engineer now needs to select out the cells that have not picked up the GM genes and discard them from the process.

4. Selection of the modified plant cells

Depending on the type of marker genes that are part of the GM gene cassette (herbicide-tolerant or antibiotic-resistant), the plant tissue culture that has undergone the GM transformation process is treated with either a herbicide or an antibiotic, to kill all cells except those that have successfully incorporated the GM gene cassette into their own DNA and switched it on. Only the cells that have incorporated the marker gene into their genome and are expressing it will be resistant to the chemical and survive exposure.

Only a small percentage of GM gene cassette insertion events result in expression of the GM genes in the plant cells.

5. Hormone treatment

The few plant cells that have successfully incorporated the GM gene cassette and survived the chemical treatment are then further treated with plant hormones. The hormones stimulate the genetically modified plant cells to proliferate and differentiate into small GM plants that can be transferred to soil and grown to maturity.

6. Verification of the GM transformation

Once the GM plants are growing, the genetic engineer examines them and discards any that are deformed or do not seem to be growing well. The remaining plants are tested so as to identify one or more that express the GM genes at the desired high levels and locations within the plant. Out of many hundreds or thousands of GM plants produced, only a few may fit this requirement. These are selected as candidates for commercialization.

Each of these GM plants carries the same GM gene cassette, but it will be inserted at a different location in the genome of the plant. The GM gene will express at different levels in different GM plants and even in different parts of the same GM plant.

At this stage the GM plants have not been assessed for health and environmental safety or nutritional value. This part of the process is described in later chapters.

The GM transformation process is highly inefficient

The GM transformation process is a complex multistep process in which each step needs to work as intended in order to produce the desired result. The GM gene cassette must be successfully inserted and the gene of interest switched on so that it produces the protein it encodes, while at the same time, all other properties of the plant, including fertility, must be preserved.

This is a very inefficient process. The process of GM gene insertion into the plant cell DNA occurs only rarely. Most inserted GM genes fail to function, either due to integration into regions of the plant genome that are not permissive for gene activation or to natural plant defence mechanisms that silence or switch off of the “invading” foreign gene.

GM gene cassettes currently used by genetic engineers do not possess any elements that are able to overcome these limitations of the transformation process. Therefore obtaining GM plants that are good candidates for taking forward for potential commercialization is a long, arduous, labour-intensive, and expensive process6,7 (see Myth 6.4).

How unnatural is genetic engineering and does it matter?

Some aspects of plant genetic engineering are unique to the GM process and do not occur in other types of plant breeding. They include the artificial construction of the GM gene cassette, which contains new synthetic genes and combinations of gene control elements that have never existed before in nature.

Also, genetic engineering enables genes to be transferred not only between different species but also between different kingdoms – for example, from animals or humans into plants. Therefore genetic engineering evades natural barriers between species and kingdoms that have evolved over millennia. Moreover, genetic engineering can introduce purely synthetic genes, thus, for better or worse, expanding the range of possible genes to the frontiers of the human imagination.

The fact that the GM transformation process is unnatural and artificial does not automatically make it undesirable or dangerous. It is the consequences of the procedure, combined with the current lack of systematic assessment of potential risks, that give cause for concern, as detailed in subsequent sections.

Horizontal gene transfer – should we worry?

The movement of genetic material between unrelated species through a mechanism other than sexual reproduction is called horizontal gene transfer, or HGT. Genetic engineering can be seen as intentional horizontal gene transfer. Reproduction, in contrast, is known as vertical gene transfer, because the genes are passed down through the generations from parent to offspring.

GM proponents argue that horizontal gene transfer occurs spontaneously in nature and that therefore genetic engineering is only speeding up a natural process, or making it more precise.

It is true that horizontal gene transfer occurs in lower organisms relatively frequently – for example, between different species of bacteria.8 HGT has evolutionary benefits from the perspective of microorganisms.

However, in higher organisms HGT occurs only under special circumstances. An example is infection with viruses, resulting in the development of endogenous retroviruses (ERVs). These are viruses that write themselves into the host’s own DNA. When they do this to a germline cell – a cell involved in reproduction (a sperm or egg cell) the genes for that virus are passed down to the offspring and become a permanent part of the genome of the descendants.

Human endogenous retroviruses (HERVs), the inherited remnants of past retroviral infections in our ancestors, are estimated to make up as much as 8% of the human genome.9

The fact that HGT has taken place does not mean that infection with such retroviruses is safe or desirable. Nor does this in any way justify commercializing GMOs without testing their impacts on health and the environment. All we know is that some people survived these retroviral infections, which changed their DNA, and that we are descended from the survivors. Virtually all of these HERVs are not expressed: that is, cellular mechanisms have silenced any effect that they might have on cellular or organismic functioning. However, the silenced HERV sequences have been passed down the generations and any side-effects due to the presence of those sequences remain unknown. It may well be that the only people who survived HERV insertion were those whose cells had the capacity to silence HERV gene expression.

The existence of HERV sequences in the human genome is evidence that horizontal gene transfer events do occur on an evolutionary timescale. But the fact that they occur does not provide evidence that HGT is “normal”, harmless, or beneficial, particularly in the short timescale relevant to direct genome changes via genetic engineering.

Another example of HGT happening in nature is infection with A. tumefaciens, a bacterium with a natural ability to carry and transfer part of its DNA to the cells of the plants that it infects, thereby causing crown gall disease, a type of plant tumour. For this reason, A. tumefaciens is a valued tool of genetic engineers.

It is important to note that the above examples of “natural” HGT into higher organisms are pathogenic processes. They illustrate the fact that in nature, the HGT process often causes disease in the infected organism. The result of the HGT process is the introduction into the host organism of a retrovirus that can play a role in cancer development (in the case of HERVs) or tumour-causing DNA sequences (in the case of A. tumefaciens infection of a plant). This is evidence that such processes cannot be assumed to be benign and may be harmful. So these examples are not an argument in favour of genetic engineering of our food supply, but rather an argument counselling against its use.

It is also important to note that unlike the GM transformation process, HGT by A. tumefaciens does not modify the germ cells of the plant and so does not affect future generations of the infected plant.

In nature, the question of whether any given example of horizontal gene transfer is beneficial or harmful is answered over long periods of co-evolution and natural selection. It cannot be answered based on the limited knowledge of the genetic engineer or under the limited timescales in which GMO introduction takes place. Neither can it be answered by the inadequate “safety assessment” regimes that are currently used in GMO regulatory processes around the world.

Use of potent plant promoters in GM gene cassettes attempts to override host plant gene regulatory mechanism

The random insertion of the GM gene cassette at the vast majority of locations within the plant cell DNA results in little or no expression of the transgene. This “silencing” of the GM gene cassette, including any associated antibiotic selectable marker gene present, is in part due to the plant’s natural response to invasion by foreign DNA, as occurs, for example, in the case of infection by viruses. This silencing occurs despite that fact that in most cases plant genetic engineers use the powerful 35S cauliflower mosaic virus (CaMV) promoter or similar powerful promoters in an effort to overcome GM gene inactivation.

Consequently the selection procedure of plants via the GM transformation process actively selects for purely fortuitous events in which the GM gene cassette, plus any associated antibiotic marker gene, has inserted into those rare sites within the plant’s DNA that allow it to function. These rare sites are by definition regions within the plant cell DNA where active host genes and their control elements are located. In other words, the GM plants contain GM gene cassette insertions into regions of the DNA where their own genes are active (gene regions represent only a tiny fraction of the total genome). This fact maximizes the chances that the plants’ host gene function will be disturbed – with unexpected downstream consequences to their biochemistry and performance.

In addition, the use of potent plant promoters such as the CaMV to switch on GM genes has other potential downsides. The CaMV promoter functions in all the different types of cells within the plant. Such ubiquitous expression is necessary in cases such as when the GM crop is engineered to tolerate being sprayed with a herbicide, to ensure that the plant survives.

But in other situations, ubiquitous GM gene expression is not so desirable. For example, GM maize engineered with the insecticidal Bt toxin gene obtained from bacteria aims to target either the corn borer or rootworm pest. Therefore the GM Bt toxin gene only needs to be expressed in stems, corn cobs, and roots, in order to ensure protection from these pests. However, the use of the CaMV promoter to drive expression of the Bt toxin transgene unit (as is the case in all current GM crops) results in the presence of this insecticide in all plant structures, not just the stems, cobs, and roots. This in turn increases the possibility of toxic effects on non-target insect populations that may feed on the pollen of these Bt GM crops, such as bees and butterflies. Thus valuable pest predator or pollinator insect populations may be harmed when feeding on Bt GM crops.

In conclusion, the use of ubiquitous promoters such as the CaMV in an effort to override the host plant’s gene regulation systems and force expression of the GM gene at high levels may have undesirable effects on plant biochemistry, crop performance and the surrounding environment.

In contrast, in natural breeding and even in mutation breeding (mutagenesis), which exposes plants to radiation or chemicals to induce genetic mutations (inheritable changes), the plants’ own gene regulation systems remain active.

In other words, scientists use genetic engineering to bypass the plants’ natural gene regulation systems and to re-programme their genetic functioning. Natural breeding, on the other hand, uses the inherent genetic potential in plants and does not deliberately disrupt their gene regulation system.

Muddying the waters with imprecise terms

GMO proponents often use the terminology relating to genetic modification incorrectly, blurring the line between genetic modification and conventional breeding.

For example, they claim that conventional plant breeders have been “genetically modifying” crops for centuries by selective breeding and that GM crops are no different. But this is incorrect. The term “genetic modification” is recognised in common usage and in national and international laws as referring to the use of laboratory techniques, mainly recombinant DNA technology, to transfer genetic material between organisms or modify the genome in ways that would not take place naturally, bringing about alterations in the genetic makeup and properties of the organism.

The term “genetic modification” is sometimes wrongly used to describe marker-assisted selection (MAS). MAS is a relatively uncontroversial branch of biotechnology that can speed up conventional breeding by identifying natural genes that confer important traits. MAS does not involve the risks and uncertainties of genetic modification. It is supported by organic and sustainable agriculture groups worldwide, with objections mostly focusing on patenting issues.

Similarly, “genetic modification” is sometimes wrongly used to describe tissue culture, a method that is used to select desirable traits or to reproduce whole plants from plant cells in the laboratory. In fact, while genetic modification of plants as carried out today is dependent on the use of tissue culture, tissue culture is not dependent on GM. Tissue culture can be used for many other purposes, including some safe and useful ones.

Using the term “biotechnology” to mean genetic modification is inaccurate. Biotechnology is an umbrella term that includes a variety of processes through which humanity harnesses biological functions for useful purposes. For instance, fermentation in wine-making and breadmaking, composting, the production of silage, marker-assisted selection (MAS), tissue culture, and even agriculture itself, are all biotechnologies. GM is one among many biotechnologies.

GM proponents’ misleading use of language may be due to unfamiliarity with the field, or may represent deliberate attempts to blur the line between controversial and uncontroversial technologies in order to win public acceptance of GM.

Contained and uncontained use of GM technology

GM technology is used in both contained and uncontained systems. “Contained use” means that the use of GM technology does not result in the deliberate release into the environment of a living GMO that is capable of reproducing and spreading.

In Europe, all laboratory and industrial uses of GM technology are regulated by the Contained Use Directive.10 Containment can be physical, in the form of barriers preventing escape, chemical, or biological (by genetically crippling the GMO so that it cannot reproduce).

Contained medical uses of GM technology include diagnosis of disease and manufacture of pharmaceuticals and GM viruses used to deliver somatic (non-germline and thus non-inheritable) gene therapy. Contained uses of GM technology in plant breeding are confined to the laboratory and include identification of genes of interest and study of their functions and protein products under normal and disease conditions.

We oppose non-contained uses of GM technology but support contained use, as long as containment is effective. There is always risk of escape during contained use, either due to physical or biological “leakiness”. However, for most current and envisioned applications, the benefits outweigh the risks when strong and well-designed containment strategies are employed.


Genetic engineering is different from natural/conventional plant breeding and poses special risks, as is recognized in national and international biosafety laws. The genetic engineering and associated tissue culture processes are highly mutagenic, leading to unpredictable changes in the DNA and proteins of the resulting GM crop that can lead to unexpected toxic, allergenic and nutritional effects.


  1. European Parliament and Council. Directive 2001/18/EC of the European Parliament and of the Council of 12 March 2001 on the deliberate release into the environment of genetically modified organisms and repealing Council Directive 90/220/EEC. Off J Eur Communities. 2001:1–38.
  2. Secretariat of the Convention on Biological Diversity. Cartagena Protocol on Biosafety to the Convention on Biological Diversity. Montreal; 2000. Available at:
  3. Codex Alimentarius. Foods derived from modern biotechnology (2nd ed.). Rome, Italy: World Health Organization/Food and Agriculture Organization of the United Nations; 2009. Available at:
  4. Codex Alimentarius. Guideline for the conduct of food safety assessment of foods derived from recombinant-DNA plants: CAC/GL 45-2003; 2003.
  5. GeneWatch UK. ASA rules that Monsanto adverts were misleading: GeneWatch UK complaints upheld [press release].[cid]=492860&als[itemid]=507856. Published August 10, 1999.]
  6. Phillips McDougall. The cost and time involved in the discovery, development and authorisation of a new plant biotechnology derived trait: A consultancy study for Crop Life International. Pathhead, Midlothian; 2011.
  7. Goodman MM. New sources of germplasm: Lines, transgenes, and breeders. In: Martinez JM, ed. Memoria Congresso Nacional de Fitogenetica. Univ Autonimo Agr Antonio Narro, Saltillo, Coah, Mexico; 2002:28–41. Available at:
  8. Doolittle WF. Lateral genomics. Trends Cell Biol. 1999;9(12):M5-8.
  9. Hughes JF, Coffin JM. Evidence for genomic rearrangements mediated by human endogenous retroviruses during primate evolution. Nat Genet. 2001;29:487-9. doi:10.1038/ng775.
  10. European Parliament and Council. Directive 2009/41/EC of the European Parliament and of the Council of 6 May 2009 on the contained use of genetically modified micro-organisms. 2009. Available at: