How do we know an inserted gene does only what it is supposed to do?
Genes are like recipes, they tell the cell how to make a particular protein. It is the presence (or absence) of the particular protein (often an enzyme) that gives the plant, animal or microbe a trait. Insulin, for example, is a protein that helps control blood sugar in mammals. The insulin gene recipe is carried in the genome of mammals, but is not present in the genomes of animals lacking blood, nor in plants or microbes, for that matter, as they have no blood to regulate. Since the 1980s, insulin used by diabetics is made by Genetically Modified bacteria into which the human insulin gene recipe was inserted. Although the bacteria have no insulin gene themselves, they were able to read and follow the human gene recipe to make insulin identical to the insulin made by human cells. This bacterial source GM insulin is extracted from the bacterial culture medium and provided to diabetics.
Genetic modification works only because all living things use the same genetic language, so a human gene—such as the insulin gene— transferred to bacteria will work the same way in the bacteria as it does in humans.
Genes are composed of long stretches of DNA, which is composed of the chemical building blocks we abbreviate as a,t,c and g (Adenine, Thymine, Cytosine and Guanine, respectively). Just as in the human language English, in which out thousands of words are composed of specific sequences of 26 letters, in biology all genes in all species are made of specific sequences of the four DNA bases, a,t,c and g. The human insulin gene consists of 4044 of these bases, beginning with …atggccctgtggatgcg… at the start of the insulin recipe.
When the plant, animal or microbe is ‘expressing’ a gene to make the requisite protein, it reads the DNA three letters at a time. Each three letter ‘word’ in DNA specifies a particular amino acid—of which there are twenty different kinds floating around in the cell—, and these amino acids serve as building block ingredients for proteins. As the gene sequence is read by the cell’s kitchen machinery, amino acids are strung together like beads on a string, called a polypeptide chain. Thus, the first part of the DNA recipe, atg gcc ctg tgg atg cgc translates to the amino acids (in order) Methionine, Alanine, Leucine, Tryptophan, Methionine, Arginine, and so on until the entire sequence of 102 amino acids is completed. The string of amino acids is processed in the cell to release functional insulin.
Therefore, when engineers insert a known DNA sequence into a plant animal or microbe, the resulting ‘expressed’ string of amino acids in the protein will correspond to the recipe as provided in the DNA base sequence. This is true regardless of the species source, or the species recipient, as all living things use the same DNA language.
To ensure the inserted gene does what it is supposed to, various molecular, chemical and biological tests are conducted on the GM plant, animal or microbe to check that the inserted DNA sequence is intact and the expressed protein functional before the GMO is commercialized. They are also extensively tested for safety and efficacy for several years prior to commercial release to ensure there are no unexpected or unusual results.
There are no known examples where a specific gene recipe (DNA sequence) was inserted into a cell and produced something other than the predicted protein. What can—and does—go wrong is to have the inserted gene go unexpressed, or only partially expressed, such that the amount of protein is insufficient to prove commercial levels of the desired trait. In those cases, the GMO is destroyed as soon as it is detected, usually in very early testing, and in any case long before commercialization.
Finally, the ‘proof of the pudding’ is in the final product. If a culture batch of bacteria genetically modified with the human insulin gene does not produce insulin, it cannot be commercialized. Similarly, if a corn plant is genetically modified to produce a protein allowing the plant to survive a herbicide treatment, and the gene does something other than what is it designed to do, the GM corn plant will die upon application of that herbicide.