Potentially promiscuous pollen from corn tassels by circulating via Flickr.

Many people, including me, are concerned about potential harm to crop biodiversity from gene flow. Most people’s concern focuses on transgenics. There is a certain probability, albeit small, that transgenes will end up in the progeny of non-transgenic plants, weedy relatives of the crop, or wild relatives that grow nearby due to pollen flow. Transgenes can also be moved from place to place by accidental or purposeful movement of seeds.

How much transgene flow is actually happening is a subject of some controversy, but what about gene flow between non-transgenic plants?

There is potential for problems whenever plants that aren’t supposed to cross stray from their intended mates. Some things to think about include how gene flow happens at the field and genetic levels and what characteristics of the genes themselves can affect permanence of contaminating genes once they get into a variety they shouldn’t be in.

Gene flow with transgenes can help us to think about gene flow of “regular” genes

In the 2004 paper Gene Flow from Cultivated Rice (Oryza sativa) to its Weedy and Wild Relatives, Li Juan Chen showed that a marker gene “flowed” in their test field from transgenic cultivated rice  to weedy rice at rates between 0.011 and 0.046 % and to wild rice at rates between 1.21 and 2.19 %. The marker gene Chen used is called bar, which is easy to screen for because it makes plants resistant to the antibiotic and herbicide biaphalos. Just spray the progeny and you’ll know if they’ve got the gene. Chen confirmed presence of the bar gene with PCR. These rates seem pretty low, but rice is mostly a self-pollinator, and the pollen is very short lived. If out-cross rates in rice reach 2.19 % we could expect to see rates even higher in other species. This tells us that transgenes can be passed to weeds, but also, more broadly, tells us that any gene can be passed from cultivated rice to weed rice.

Gene flow could be a problem in the opposite direction as well. In the 2009 paper Gene flow from weedy red rice (Oryza sativa L.) to cultivated rice and fitness of hybrids, Vinod Shivrain showed that progeny of a cross between weedy red rice and cultivated rice were more successful if their mother was the cultivated plant. These hybrid grains can fall back to the field on accident or be collected and planted the following year with the regular seed. Either way, the rice farmer now has rice plants that don’t have all of the desired characteristics of the cultivated rice. The plants will have at least some genes from the weedy rice that could help it out compete the desired rice plants but produce less grain. This paper shows that gene flow from weeds to crops can happen, and that it can be a problem.

Maize, unlike rice, is a promiscuous out-crosser. The pollen is heavy and still fairly short lived, so mostly pollinates plants that are nearby, but wind-carried pollen and stray seed can carry transgenes away from their intended fields. The story of transgenes in landraces of maize is summed up beautifully in the 2007 paper Gene flow from transgenic maize to landraces in Mexico: An analysis (pdf). Kristin Mercer tells us that research on the subject has had mixed results. Transgenes likely do exist in landraces in Mexico, but the extent of the “contamination” is not as wide as some researchers have proposed. Some of Kristen’s other research focuses on how crop alleles move in wild sunflower populations. The sum of her research is that we can expect gene flow back and forth between any compatible plants: wild, weedy, cultivated, transgenic, landrace.

Gene flow’s effect on biodiversity

Image of corn plant by University of Nebraska Lincoln, adapted by Anastasia Bodnar. All other images in this post by Anastasia Bodnar.

Understanding the impact of gene flow on biodiversity (or more appropriately, crop diversity) requires some understanding of what happens at the genetic level. I like to sit down and draw pictures to help me think about genetics. I hope it helps some Biofortified readers as well!

The image to the right shows two hypothetical varieties of corn. On the left is a modern inbred variety. All the plants are identical. There is no or low genetic variability within the inbred, because there is only one version of each gene present in the variety. On the right is a landrace or heirloom variety. All the plants are different from each other to some degree. There is high genetic variability within the landrace because there can be many versions of each gene present in the variety.

Below  is a (very) simplified view of what happens at the chromosomal level when an inbred is crossed with a landrace (in a hypothetical crop with one chromosome). Note: a hybrid or even an open pollinated variety could be substituted for inbred here, it was just easier to use an inbred. Similarly, a wild variety could contaminate a landrace. One landrace can contaminate another. One inbred could contaminate another. Weedy relatives can contaminate crops. Crops can contaminate wild varieties… you get the idea.

In the inbred (red), the two sister chromatids for each chromosome are identical to each other. There is only one version of each gene in the inbred, also known as two copies of the same allele. In the landrace (blue), the two sister chromatids are different from each other. For each gene in the landrace, there can be two different alleles. The different shades of blue indicate different alleles for some genes on the sister chromatids.

Imagine a situation where a field with the inbred is right next to a field with the landrace. Pollen will flow between the fields (to some degree – depending on weather conditions, pollen size, and tons of other factors). If the inbred and the landrace are crossed (whether pollen from the inbred fertilizes the landrace or vice versa), each of the offspring will have about half of the genetic information from the inbred and half from the landrace. Since the two chromatids are the same for the inbred, none of the information from the inbred is lost in any individual. Since the two chromatids in the landrace individual are different, each of the offspring only receive half of the genetic information from the landrace.

When those offspring make gametes, recombination often occurs which results in chromatids that contain some alleles from each grandparent. Crossing over, shown here, is one type of recombination. If those gametes then combine with the inbred, their progeny will only have about 1/4 of its genes from the landrace grandparent.

Genetic diversity can be lost in certain situations. For example, if a farmer growing a landrace finds plants in the field that have positive traits, the farmer will choose to plant those seeds for the next year. If those beneficial traits are due to genes from the inbred, the farmer could effectively select for plants with one or more genes with the inbred and against plants that don’t contain any genes from the inbred. If pollen from the inbred is reintroduced year after year, the farmer could plant seeds from those plants that contain more and more alleles from the inbred variety, and alleles from the landrace could be lost over time.

On the other hand, if the farmer chooses seeds from plants that look more like the landrace, then alleles from the inbred could be lost fairly quickly. If pollen or seeds from the inbred are introduced infrequently, the landrace would maintain a low level of alleles from the inbred, with those alleles eventually disapearing.

Of course there are many situations in between, and those depend greatly on what effect each gene or allele has on the plants they have contaminated.

Once it’s in there, how long will it stay?

Transgene or not, wild or cultivated, all of the genetic material goes into a big mixing pot to be stirred by random mating and natural selection in the case of wild plants or by breeding and artificial selection in the case of cultivated plants. One of Kristen’s points in Gene flow from transgenic maize to landraces in Mexico: An analysis (pdf) is that the permanence of transgenes in a non-transgenic population depends a lot on what the transgene is exactly, and the same idea applies to non-transgenic alleles.

To break it down: Any given transgene or any allele of a gene can have one of three effects on the plant: positive, neutral, and negative. The effect depends on what plant the allele is contaminating and what trait is conferred by the allele. Finally, how long a contaminating allele stays in a population depends on all of these factors.

Positive

Some alleles would be beneficial in almost any situation. Herbivore resistance, including genetically engineered Bt toxin and increased expression of non-transgenic chemical defenses, would help both cultivated and non-cultivated plants escape damage from susceptible herbivores. These types of transgenes and alleles would be likely to persist in any population they contaminated. These would definitely be bad traits to have in weeds. They could be desirable in a landrace from a farmer’s point of view.

Neutral

A gene that increases the size and number of fruits produced by a plant is desirable from an agricultural perspective, but could have a negative effect a wild plant, because the plant would have less resources to devote to other needs like herbivore defense and drought tolerance. These types of alleles will not persist in a wild population, but could persist in a landrace if it is seen as desirable to farmers.

Negative

Alleles or genes that are specific for certain farming systems won’t persist in wild populations, weeds, or landraces unless they are exposed to those farming conditions. These include genetically engineered genes like glyphosate tolerance and the non-transgenic allele for Clearfield tolerance. If these alleles or genes contaminate a population but that population is never sprayed with the chemical, there is no selection pressure to keep the trait.

Of course these are just three examples of different traits and there are thousands if not millions of traits out there that might have different effects, but you get the idea.

Every day, pollen blows and seed is moved. Every day, genes and alleles are transferred from one plant population to another, no matter if they are transgenes or not. Those naughty plants just won’t keep to themselves! If we are truly concerned about gene flow, we really should be considering gene flow from all sources, not just transgenic crops.

Chen LJ, Lee DS, Song ZP, Suh HS, & Lu BR (2004). Gene flow from cultivated rice (Oryza sativa) to its weedy and wild relatives. Annals of botany, 93 (1), 67-73 PMID: 14602665

Shivrain VK, Burgos NR, Gealy DR, Sales MA, & Smith KL (2009). Gene flow from weedy red rice (Oryza sativa L.) to cultivated rice and fitness of hybrids. Pest management science, 65 (10), 1124-9 PMID: 19530257

Mercer, K., & Wainright, J. (2008). Gene flow from transgenic maize to landraces in Mexico: An analysis Agriculture, Ecosystems & Environment, 123 (1-3), 109-115 DOI: 10.1016/j.agee.2007.05.007

Volker Brendel, professor of bioinformatics at Iowa State, spoke at the Maize Genetics Conference about the need for a better system of community annotation of the maize genome. The genome of the popular maize inbred line B73 is sequenced, but we don’t actually know what a lot of the code stands for. It’s going to take a lot of collaborative effort to discover and annotate (explain) the function of each gene and to put all of that information in one place so it will be useful.
Volker reminds us that the Arabidopsis 2010 funding is running out, so we need to assess the plant genetics situation. How many genes do we know the function of? There is still much to learn.
Maize is uniquely positioned to replace Arabidopis as a focus for basic plant research due to the many resources that are already established, the most important of which is the extensive maize genetics community (he didn’t say it, but there is another reason why maize is a better choice than Arabidopsis right now – all of our major grains are very closely related, so work on maize applies to rice, wheat, sorghum, and more). The community needs to work together in the annotation process, assigning functions to the genes that have been sequenced, putting the data from a variety of sources together to make a bigger picture. Each researchers has a favorite gene (pathway, organelle, etc) – how can each of the researchers contribute to the annotation process?
PlantGDB is a comparative genomics site funded by NSF has information on 14 species, including maize, which is very useful. However, no matter how clever the computer programs are, the human touch is still needed. Filling in information on any of these species helps us to better understand all of them. On the site, community members can flag genes for which the models don’t seem to fit, and can contribute alternative explanations. The final goal is to have every gene model approved by the relevant community member(s). When a person annotates a gene, the PlantGDB committee reviews it, approves it, and the information is shortly available on the site. Annotating the genes you are working on is your civil duty, something you owe due to public funding you receive.
After Volker’s talk, the attendees discussed what is the public’s role in the attenuation process should be. There are a lot of cases where the the gene model can be checked without any lab work, simply by looking at the sequences. Some members of the community think we should harness the brainpower of thousands of biology undergraduate students by assigning annotations for class. I like the idea of getting students involved, and hope they follow through.Diversity of people to represent the maize genetics community.
A panel discussion followed, where a lot of great new ideas for annotation were brought up (unfortunately I don’t have the names of some of the people that spoke).
One panel member said we need “Zeazomics” – a collection of information including genomics, metabolimics, proteomics, and whatever else we can come up with – to fill in gaps in our knowledge. being able to link all of this information together will lead to stronger explanations of the phenotypes we see. He said this process will not be definitive, it will create a series of hypothesis that will lead to more hypotheses. The hypothesis testing will lead to functional biolgoy, from physiology to biochemistry to cell biology and more. Additional genome sequencing is necessary to capture the entire diversity of maize. Maize is the model for grasses, for crops, for future applications like biofuels. Now is the time to push maize research to a much higher level.
To accomplish all this, we’ll need to take care of a few things, as the other panel members and members of the community brought up:

• Need to have reciprocal links from genes from MaizeGDB to NCBI Entrez Gene. Currently, about 20,000 NCBI Entrez Genes need links back to MaizeGDB.
• To help with annotation, Lisa Harper, curator of MaizeGDB, will do a movie that shows the common problems of using the databases, including how the genome changes over time as the contigs are reordered, etc. This is needed because people are often working off of older copies of the information for a given gene, as it might not be updated frequently enough.
• There is also a need to integrate microarray data into the databases. Particularly complicated are those microarrays that are specific to a particular tissue and/or developmental stage. Volker says that this problem is common and new technologies with new ways to visualize data are necessary.
• MaizeGDB needs a forum such that people working on the same genes can coordinate their work.
• iPlant is organizing a workshop in St. Louis in June to help coordinate the various genome annotation groups.

• There is a plan to create outreach information that any member of the maize community will be able to download and use to communicate the needs and accomplishments to the public and to government officials.

[David Nothmann, Monsanto’s Soybean Agronomic Trait Lead,] said Roundup Ready 2 Yield technology is based on an advanced gene-mapping and insertion process. “Through gene mapping, Monsanto has identified specific DNA regions in soybeans that have a positive impact on yield,” he explained. “Using these new insertion and selection technologies, the Roundup Ready 2 Yield gene is situated in one of these DNA regions.”

There has undoubtedly been decreased yield in Roundup Ready crops when compared to conventionally bred crops. This has two possible causes: lag “a temporary or transient problem associated with the introduction of a new technology” or drag “an inherent yield reduction associated with the technology itself”. There is a lot of evidence that the problem is in fact lag, but more research must be done. Some of the issues are discussed in Challenges in Comparing Transgenic and Nontransgenic Soybean Cultivars.
I covered the topic of yield lag/drag somewhat in my post Exposed, Indeed.

GM seeds are often “one hit wonders” that excel in one specific trait, but not particularly for increased yield. Non-GM lines, on the other hand, are improved every year, with the best yielding plants being used to produce the next year’s seed. I recently attended a seminar presented by a scientist from Pioneer where he said that they were working to develop better yielding lines that would work in conjunction with their primary transgenic traits. The companies are aware that this is a problem with their products, and are of course working to solve it, to avoid losing sales.

Back to the question at hand – as I understand it, advanced gene mapping is a selection tool.  The companies start with huge experimental fields (much larger than what an academic lab can afford) in multiple locations with different climates that include many varieties of the crop in question. They measure yield and determine genetic markers* for each variety/location combo, using known markers for yield as the starting point. The researchers are then able to see which varieties do and do not have certain markers. They cross varieties that have different markers, with the goal of a super high yielding plant that has all of the markers that are positively correlated with yield and none of the markers that are negatively correlated with yield. They have fields in multiple locations so they can choose the markers that confer an advantage in a variety of climates – ensuring that the plants will perform no matter what the location or conditions. There are a lot of benefits of this method over blind selection, the biggest of which (in my opinion) is that you don’t have to know what’s happening physiologically in the plant. Knowing what each gene does (and what each mutation to each gene does) is nice, but really not necessary for the purpose of breeding bigger better plants.
As for the “insertion process” part, I admit that I’m not 100% sure why positioning the insertion in an area of the genome that is correlated with high yield would matter (any readers who know, feel free to enlighten me!). I can think of a few reasons why the specific position of the transgene insertion does matter, but all of them are part of the normal process running up to a marketable genetically engineered crop. In fact, I’m in the process of some of those stages right now. Once the gene of interest is chosen, a compatible promoter must also be chosen. I imagine that a constitutive promoter (always on, in every cell) would be used for the glyphosate resistance trait. The gene construct is introduced into many plant cells that are then grown into individual adult plants. Each introduction is called an event. Each event is treated separately because the position of insertion is different each time. When the insertion lands in the middle of a gene, it can stop the gene’s normal expression – so many events are investigated to see which ones have the least effect on the plant’s normal gene expression while at the same time producing the desired trait. A video on Monsanto’s website says that they used Agrobacterium instead of biolistic transformation in RR2 because it is “gentler”, causing less damage to the surrounding DNA. They then screened many events using genetic markers to find the best ones – an expensive process that (to my knowledge) has not been done before. They say that having the insertion in one of the areas near a marker for high yield increases yield an additional 7 to 11 percent.
Edit: I don’t know why I was having a mental block on this! What I said in the last paragraph stands but I’ve figured out why having the insertion in a high yield correlated area would matter. If the insertion is near an allele for a gene that is correlated with poor yield, selecting for the trait of interest would bring along the area that you don’t want. Having the insertion in a  “good” area of the genome (assuming that it isn’t actually interrupting any genes) eliminates this problem.
*Here, markers basically correspond to alleles of a gene. Wikipedia has a decent explanation of genetic markers, but unfortunately requires the understanding of much more jargon. If you’d like a more detailed explanation of genetic markers, please let me know, and I will be happy to write a post on the subject.

Raymer, PL and TL Grey (2003). Challenges in Comparing Transgenic and Nontransgenic Soybean Cultivars. Crop Science, 43, 1584-1589.

[A]rsenic levels in the grain of different varieties of rice in Bangladesh were as high as 1.8 parts per million, compared to levels of just 0.05 parts per million in Europe and the US. Contamination was even greater in leafy vegetables – in amaranthus and spinach, arsenic content can be double or triple the levels found in rice. For drinking water, WHO recommends a maximum arsenic level of 0.01 parts per million, which indicates that for some people, staple food crops such as rice may be an important source of exposure to arsenic.

Until now, the farmers essentially have three options: leave the fields fallow, plant rice and hope it doesn’t have too much arsenic, or attempt to plant a crop that doesn’t need as much water.
Om Parkash (photo and story from Newswise) of the University of Massachusetts Amherst primarily works on bioremediation, which aims to remove pollutants from the soil by binding it up in plants. His recent work branches into the opposite direction, using genetic engineering to produce rice plants that take up less As. The work is in the process of patenting, so technical details are scarce. For now, I’ll have to be content with the following:

“By increasing the activity of certain genes, we can create strains of rice that are highly resistant to arsenic and other toxic metals,” says Parkash, a professor of plant, soil and insect sciences. “Rice plants modified in this way accumulate several-fold less arsenic in their above-ground tissues, and produce six to seven times more biomass, making the rice safer to eat and more productive.” This could help alleviate the current world-wide rice shortage.

I’m really looking forward to learning more about the genetics, and hope that Dr. Parkash is able to move forward with this exciting crop improvement.
While As is actually a necessary mineral in small amounts and only becomes dangerous to health when consumed in high levels (as in Bangladesh), decreasing As in the food supply is definitely a worthy cause. Dr. Parkash says that As can accumulate in all parts of the rice plant, including grain and straw. High As levels in rice not only affect people, but can sicken animals who eat the straw and contaminate their meat (think bioaccumulation). See GreenFacts for a good summary of arsenic as it relates to human health and the environment (incidentally, they also have some of the most levelheaded information on GM crops that I’ve ever seen).
Other recently published work on arsenic levels in rice by Yamily Zavala and John Duxbury of Cornell was reported in the 2 May 2008 ISAAA Crop Biotech Update. For a summary of the articles, see the press release from the American Chemical Society. Disclosure: I wasn’t able to access these two articles themselves as ISU’s library site is down while I write this.
In Arsenic in Rice: I. Estimating Normal Levels of Total Arsenic in Rice Grain, they showed that mean As concentrations in samples of commercial rice in Europe and the US (0.198 mg/kg) were higher than in samples from Asia (0.07 mg/kg). The concentrations varied greatly by region, but not by farming method. Their data confirmed that irrigation with As contaminated groundwater in Bangladesh is correlated with higher As concentrations in grain. In the US, where groundwater is not contaminated with As, the authors suggest that historical contamination of soil is a likely cause. Note: mg/kg and ppm are equivalent units.
In Arsenic in Rice: II. Arsenic Speciation in USA Grain and Implications for Human Health, they showed that the As in some rice varieties accumulates in a less toxic form than inorganic As (inorganic = molecules do not contain carbon). Arsenic in rice grown in the US is bound into mostly into dimethyl arsinic acid (DMA), which . This data is in agreement with previous studies done by Andrew Meharg of the University of Aberdeen in the UK. There is evidence that DMA is safer than inorganic As, which means that US rice may be safer than European or Asian rice. The authors hypothesize that 30 years of breeding in the US for straighthead disorder resistant rice could have caused US varieties to acquire this As metabolic pathway.
Huge phenotypic variance is present in rice grains across varieties. It’s easy to imagine that metabolic pathways vary widely from variety to variety as well.

Good news from Africa – “Scientists and crop researchers at Kenya´s Agricultural Research Institute (KARI) developed the new wheat seeds over the past decade. Through a process called ‘mutation plant breeding’, they applied radiation-based techniques to modify crop characteristics and traits.” In 2001, KARI plant breeders released Njoro-BW1, their first mutant wheat variety. It is drought tolerant, moderately resistant to rust (a fungus), has good yield, and good flour quality. “Kenya´s plant breeders soon will release a second mutant wheat variety, code-named DH4, which shares most of the same good qualities of Njoro-BW1.” [Golden Wheat “Greens” Kenya´s Drylands]

Traditional breeding encompasses all plant breeding methods that do not fall under current GMO regulations.As the European legal framework defines GMOs and specifies various breeding techniques that are excluded from the GMO regulations,we use this framework as a starting point, particularly the European Directive 2001/18/EC on the deliberate release of GMOs into the environment (European Parliament, 2001). Excluded from this GMO Directive are longstanding cross breeding, in vitro fertilization, polyploidy induction, mutagenesis and fusion of protoplasts from sexually compatible plants (European Parliament, 2001).It is indeed good news that Kenyan farmers have these lines of wheat with such improvements over unimproved varieties. However, radiation based so-called mutation plant breeding could have unintended changes in the genome. This technique, widely used in both organic and conventional crops, literally bombards the seeds with radiation. The seeds are allowed to germinate, and interesting mutants are used to create new lines. The problem is that multiple mutations can occur in the same seed, and some of those mutations may go undetected.

A February report entitled “Microarray analyses reveal that plant mutagenesis may induce more transcriptomic changes than transgene insertion” from the National Institute of Health in Portugal indicates that this plant breeding tool may not be the best idea. The last few sentences of their abstract sums it up:

We found that the improvement of a plant variety through the acquisition of a new desired trait, using either mutagenesis or transgenesis, may cause stress and thus lead to an altered expression of untargeted genes. In all of the cases studied, the observed alteration was more extensive in mutagenized than in transgenic plants. We propose that the safety assessment of improved plant varieties should be carried out on a case-by-case basis and not simply restricted to foods obtained through genetic engineering.

Transgenesis is the genetic modification of a recipient plant with one or more genes from any non-plant organism, or from a donor plant that is sexually incompatible with the recipient plant. This includes gene sequences of any origin in the anti-sense orientation, any artificial combination of a coding sequence and a regulatory sequence, such as a promoter from another gene, or a synthetic gene.Trying to regulate GM or non-GM as broad categories are impossible, because each resulting plant variety is going to have its own “quirks”. If DH4 and Njoro-BW1 have been extensively tested for unwanted alteration in gene expression and subsequently released for general use, then they are reasonably safe (remember, nothing is definitive in science). Similarly, if transgenic plants such as Sub1A-1 rice have been tested and released, then they too can be used without worry. However, if plant varieties mutated with radiation are not adequately tested before release, then we might all have something to worry about.

To my knowledge, only Canada requires testing of these crops. We can’t even assume that traditional breeding by cross pollination is 100% safe because of natural mutation and new combinations of genes and alleles. Tomatoes, potatoes, and celery all naturally produce some nasty toxins. We’ve mostly bred them out, but there have been cases where the toxins appeared at higher levels through traditional breeding. These plants have much higher probability of danger for consumers than transgenic plants, but don’t have to be tested at all under current regulations in the US or EU.

Intragenic or cisgenic plants are our best opportunity for safe enhancement of food crops (cis- means same). This is a form of genetic engineering that uses the plant’s own genome as a source for new traits instead of other non-related organisms (has also been called GM-lite). To learn more about the idea, please see www.cisgenesis.com.

Cisgenesis is the genetic modification of a recipient plant with a natural gene from a crossable—sexually compatible—plant. Such a gene includes its introns and is flanked by its native promoter and terminator in the normalsense orientation.Cisgenic plants can harbour one or more cisgenes, but they do not contain any transgenes.Some people, including myself, beleive that cisgenic crops should be regulated differently from transgenic crops that express proteins that don’t normally occur in that species. The applications of cisgenics are more limited than transgenics, but still there is a lot to be done. A great example of cisgenics is gene silencing, which can be used to inactivate unwanted genes, such as those that cause toxins. Examples that are currently being researched are less carcinogenic tobacco and rice that can more easily form hybrids. All of the benefits in KARI’s mutated wheat could have been accomplished with cisgenics.

JR Simplot is a company that is particularly interested in cisgenics, and has produced a lot of literature that essentially says that Monsanto’s way of creating new plant lines is not the right way. I think there’s room for both, but agree that cisgenics are inherently safer. I especially like the idea that cross pollination between cisgenic plants and wild varieties won’t be a problem, since these things could have all happened naturally anyway. The idea of cisgenics has been around for quite a few years now, but scientists need to talk with the public about it, so the public can talk to their government representatives, so the representatives can go about getting the regulations changed. 

Images from “Cisgenic plants are similar to traditionally bred plants: International regulations for genetically modified organisms should be altered to exempt cisgenesis”.