More Options

We Are All GMOs

Matan Shelomi

Skeptical Inquirer Volume 39.6, November/December 2015

Implications of New Research on Horizontal Gene Transfer

Humans have been modifying the genomes of plants and animals for millennia with techniques ranging from cross-breeding to irradiation. But these methods and their products do not seem to raise the same level of public outcry as transgenesis, which is the ability to directly move genes between organisms without sexual reproduction. An explanation for the unequal opposition to such genetically modified organisms (GMOs) is the idea that transgenesis is more “unnatural” (Pivetti 2007). Splicing genes from a flounder into a tomato (Hightower et al. 1991) or from a bacterium into rice (Tang et al. 2015) is seen as posing unspecified risks beyond those of conventional breeding because, while plant and even some animal hybrids exist in nature (Zhang et al. 2014), cross-kingdom
hybrids do not. Surveys suggest the more “dissimilar” the organisms whose genes are being shared (Dragojlovic and Einsiedel 2013), and the more artificial the methods used to move the genes (Kronberger et al. 2014), the more unnatural the food will appear and the more reluctant consumers will be to eat it, especially if one of the organisms is an animal (Frewer et al. 2013).

Some of this fear is supported by concepts taught in secondary school– and college-level biology textbooks (Zeigler 2011; Alberts et al. 2013). Among animals, cross-species breeding is taught to generally fail or produce infertile hybrids such as mules; among plants, it only works for closely related organisms (Morris et al. 2013). Horizontal gene transfer (HGT), or the natural movement of genes from one organism to another without sexual reproduction (sound familiar?), is taught as being “rare among eukaryotes [animals, plants, fungi, and protists] but common among bacteria” (Alberts et al. 2013). When HGT between domains of life is taught, it is often limited to the ancient, one-time transfers of mitochondrial and chloroplast genes from bacteria to eukaryotes (Reece et al. 2012; Morris et al. 2013). Even when HGT among eukaryotes is noted, the emphasis is still on unicellular organisms (Fitzpatrick 2012). The assumption is that the evolution of separate, distinct sex cells in multicellular life posed unambiguous restrictions on HGT (Doolittle 2000). Both laypersons and many scientists thus believe that the sharing of genes over large phylogenetic distances, especially between animals, is rare or even impossible and that HGT is purely in the domain of unicellular life. Since genetic engineering is artificial HGT between two organisms of seemingly any evolutionary distance, it is accordingly seen as an unnatural violation of these laws: the creation of something that could not arise naturally even in theory.

Yet what if these basic premises are wrong? Analyses of the ever-rising number of published genomes and transcriptomes for multicellular life have revealed unexpected numbers of functional genes whose sudden appearance in a clade (a group of organisms assumed to share a common ancestor) suggests they jumped across great phylogenetic distances rather than evolving from a common ancestor: in other words, HGT. Each new genome can be a source and/or recipient of new genes, meaning the accuracy and power of HGT searches improves with each new sequenced organism (Paganini et al. 2012; Boto 2014). This growing body of research suggests horizontal gene transfer across domains is far more common than previously thought, including in metazoans (Dunning Hotopp 2011; Syvanen 2012; Schönknecht et al. 2014), where HGT’s prevalence in vertebrates and invertebrates does not differ significantly (Crisp et al. 2015). The suggestion is that HGTs may be ubiquitous rather than uncommon in all forms of life, even animals.

Yes, animals includes us: you, me, and the heads of Monsanto and Greenpeace alike. A study published this March conservatively estimated that primates have on average thirty-two foreign genes per species (there are thirty-nine in humans), with maximum estimates up to 109 in primates and 145 in humans from sources including plants, fungi, protists, bacteria, archaea, and even viruses (Crisp et al. 2015). These jumped genes may even have shaped our own evolution: the ABO-blood group system in humans is hypothesized to have been acquired horizontally (Brew et al. 2010), along with some fat mass and obesity-associated proteins found only in humans and algae (Robbens et al. 2008). Within our DNA are genes we did not inherit but that were acquired laterally from organisms across vast distances of evolutionary space. We are transgenic organisms—and everything else might be too.

Thus the idea still taught in textbooks—that HGTs almost never happen in multicellular life—is a misconception. HGTs’ perceived higher prevalence in microbes may have been an artifact of the greater number of sequenced microbial genomes: an error that is rapidly being corrected (Keeling and Palmer 2008; Boto 2014). Acceptance among scientists that cross-domain and animal HGT is a natural process “both ancient and ongoing” (Crisp et al. 2015) will not only broaden studies into the field, but it also conveniently erodes the foundation of the argumentum ad naturum against GMOs.

Consider Bt crops, in which an insecticidal gene from the bacterium Bacillus thuringiensis is inserted into plant genomes (Vaeck et al. 1987); or golden rice, which contains bacterial (Erwinia uredovoia) and daffodil genes for beta-carotene (Tang et al. 2015). Consider also the first GMO sold for human consumption, the Flavr Savr tomato, created by inserting a duplicate but antisense copy of the tomato’s own polygalacturonase enzyme (technically making it cisgenesis not transgenesis) into the genome to improve shelf life (Bruening and Lyons 2000). These varieties have been extensively and successfully tested for safety (Potrykus 2012; Helassa et al. 2013; Naranjo 2014), yet their in vitro origin still repels would-be consumers (Knight 2009), with Flavr Savr eventually declared a commercial failure (Bruening and Lyons 2000). We now know, however, that the HGTs responsible for these lab-created strains could just as easily have occurred in nature. B. thuringensis is a widespread microbe found on plant leaf surfaces (de Maagd 2015). Given the association of symbioses with HGT events (Dunning Hotopp et al. 2007), B. thuringiensis genes could theoretically have transferred into a plant on their own given enough time and opportunities. One of the first discovered animal HGTs was of a carotenoid biosynthesis gene transferred from fungi to pea aphids (Moran and Jarvik 2010), a similar HGT in terms of gene family and phylogenetic distance to the ones involved in golden rice (Tang et al. 2015). Finally, while Flavr Savr’s production involved only one species, natural HGTs of polygalacturonases were recently documented between highly dissimilar organisms, such as transfer from fungi to beetles (Kirsch et al. 2014) and proteobacteria to stick insects (Shelomi et al. 2014). In other words, while the techniques that created the aforementioned GMOs were artificial, fully natural versions of these mechanisms exist, and transgenic events covering similar genes over similar or greater evolutionary distances have already happened. The trait transfers induced in many GM crops could theoretically arise without any human intervention at all.

One important caveat is that natural HGT events in eukaryotes mostly still involve microbes or at least fungi. HGT between two animals is rare (for now) and requires organisms be in close contact with each other. An example is the transfer of insect genes to bats either from respective parasite-host or prey-predator interactions (Tang et al. 2015). Thus, even though the evolutionary distance involved need no longer be seen as a problem, an arctic flounder to tomato HGT in nature remains highly unlikely given the ecological distance between the two (Hightower et al. 1991). This does not change the basic facts about GMO safety, though it will surely be a point anti-GMO activists seize upon.

When seen as HGTs, GMOs recover much of the “naturalness” denied to them by their opponents, and a key (though by no means sole or largest) motivator to anti-GMO fear is weakened. Experimental evidence suggests increasing public knowledge of biological facts reduces the role of subjective perceptions of naturalness in public approval of GMOs, changing the discussion toward one about quantitative risk and benefit assessments (Mielby et al. 2012), at which point one can bring up the large and ever expanding body of evidence for GMO safety (Snell et al. 2012; Klümper and Qaim 2014; Nicolia et al. 2014). Thus education about the incredible sources of genetic variation in the natural world, such as the naturalness of HGTs among plants and animals and even humans, may greatly reduce the “unnatural” appearance of transgenesis and improve public acceptance of GM crops . . . though improving awareness among biologists of the ubiquity of HGTs is a necessary first step.


Matan Shelomi

Matan Shelomi's photo

Matan Shelomi, PhD, is a research fellow at the Max Planck Institute for Chemical Ecology in Jena, Germany, studying horizontal gene transfer in insects. He is also a top writer on, which you can follow for skeptical answers to biological questions and plenty of amazing entomological facts.