It’s been nearly two years since a group of Nobel laureates published a letter supporting “Precision Agriculture (GMOs)” and, more specifically, Golden Rice, the genetic engineering project first embarked upon in the mid-1990’s to “reduce or eliminate much of the death and disease caused by a vitamin A deficiency (VAD), which has the greatest impact on the poorest people in Africa and Southeast Asia.” And although Golden Rice with its potential benefit is still not available to people suffering from VAD, a couple of significant technical and other events in the project’s development have occurred over the last couple of years that showcase some of the imprecision and potential risks inherent in the technology used to create this genetically engineered (GE) rice.
The potential benefit of Golden Rice has yet to be realized
The prospect of reducing or eliminating much of the death and disease caused by VAD is certainly an intended potential benefit of Golden Rice. But before that potential benefit might be realized, there are still many issues that remain to be addressed.
For one thing, we still don’t know whether Golden Rice will be effective in treating VAD. According to the International Rice Research Institute (IRRI), the organization developing Golden Rice “for cultivation and use in human food in certain south and southeast Asian countries,” “bioefficacy studies to determine the effect of Golden Rice consumption on various measures of vitamin A status” in humans still need to be carried out, and “appropriate regulatory authorization and institutional review board clearances” for such studies have not been obtained as of yet.
Additionally, regulatory “authorization for the unconfined environmental release of GR2E Golden Rice” and the “use of GR2E Golden Rice in food and feed, or for processing,” in the Philippines or any other country in Asia, Africa and Latin America where VAD is a serious problem, has not yet been obtained.
On the other hand, IRRI has received what it refers to as a “positive food safety evaluation” of Golden Rice GR2E from the US FDA (as well as from Food Standards Australia New Zealand and Health Canada). The short letter from FDA conveying this “positive food safety evaluation” is odd for several reasons, first just as an example of how voluntary consultation about newly developed GMOs works at that US agency. The safety-related content in FDA’s letter to IRRI is comprised primarily of “information IRRI has provided” and conclusions “it is [FDA’s] understanding” IRRI has reached. Non-administrative, safety-related FDA input is limited to the agency’s usual consultation-closing statement: “Based on the information IRRI has presented to FDA, we have no further questions concerning human or animal food derived from GR2E rice at this time. However, as you are aware, it is IRRI’s continuing responsibility to ensure that foods marketed by the firm are safe, wholesome, and in compliance with all applicable legal and regulatory requirements.” (One can see why IRRI does not refer to FDA’s position on Golden Rice as an “approval.”)
Another odd thing is that IRRI informed FDA that “GR2E rice is not currently intended for cultivation or marketing in the United States.” Which begs the question: why, since “according to IRRI, GR2E rice is intended for cultivation and use in human food in certain south and southeast Asian countries…for populations in which vitamin A deficiency is common,” did IRRI go through the “regulatory” process with the US FDA in the first place? IRRI’s answer? “…GR2E rice, or human and animal food products derived from GR2E rice, may enter the U.S. food supply via imports from countries of production.” It’s not clear whether IRRI means that the US may knowingly import Golden Rice or that its non-GMO rice imports might be contaminated with Golden Rice or both; perhaps IRRI is just trying to cover all its bets. In any case, this possibility of Golden Rice being imported into the US explains why 3 of the 6 paragraphs in FDA’s “consultation closure letter” to IRRI deal with the issue of GMO labeling.
The most striking point made in the FDA letter, however, is that the “concentration of β-carotene in GR2E rice is too low to warrant a nutrient content claim.” The whole idea of Golden Rice is to provide a necessary dietary nutrient to people at risk of, or suffering from, vitamin A deficiency…and the levels are too low to warrant a nutrient content claim with FDA?
The explanation of this paradox lies partly in the amounts of rice that Americans versus Filipinos (for example) consume. According to IRRI, to meet the FDA requirement for a nutrient claim “10 percent or more of the recommended daily intake (RDI) for vitamin A must be present as beta-carotene” in the 45 grams of rice that Americans, on average, consume daily; and the amount of beta-carotene in Golden Rice is less than that. (If the 10% threshold had been reached, I believe premarket review and/or official approval by FDA would have been required.) But the average Filipino consumes more than 6 times that much rice, and IRRI expects that if all of that was Golden Rice, “30-50% of the estimated average requirement for vitamin A” could be met in the “subgroups at highest risk of vitamin A deficiency.”
The imprecision of genetic engineering and its effects on Golden Rice
But another reason why the concentration of beta-carotene in today’s Golden Rice is relatively low (~7.31 micrograms/gram, ~10% of the level found naturally in carrots-from safety assessment documents submitted to the government of the Philippines) is that what had been the most advanced variety of Golden Rice in IRRI’s product development pipeline, a variety with substantially higher levels of beta-carotene, GR2-R1, turned out to be a genetically engineered mutant with no market potential.
GR2-R1 had been a very promising Golden Rice “event,” with levels of beta-carotene more than twice as high those in GR2E (Paine et al. 2005). It was so promising that it was being bred into a popular rice variety called Swarna in preparation for its use by farmers in India. But the plants produced as a result of this breeding effort turned out to be stunted, pale green and have drastically low yields. To determine what was causing these problems a group of scientists in New Delhi (Bollinedi et al. 2017) carried out a molecular characterization of GR2-R1 that revealed that the foreign genes responsible for production of beta-carotene in GR2-R1 rice grains had been inserted into a rice gene called OsAux1, mutating it. The protein product encoded by this gene provides for proper function of a master growth regulator (auxin) in rice that is “involved in every aspect of plant growth and development” (Bollinedi et al. and references therein); mutating it in GR2-R1 rice caused various abnormalities in plant growth and development.
I am not at all surprised that the genetic engineering process caused this mutation. After all, we’ve known for more than 25 years that the Agrobacterium-based methods of genetic engineering cause frequent insertional mutations in recipient plants; Bollinedi et al. (2017, and references cited therein) state that 35-58% of insertions of foreign DNA sequences disrupt recipient plant genes.
What I am surprised about is that no one had (apparently) ever checked to make sure that this GE rice plant did not contain such a GE mutation prior to it being readied for farmers’ fields. Genetic engineering and associated technologies provide genetic engineers with the ability to identify such insertional mutations. Genetic engineers should be doing so early in the development of products intended for commercial release and eliminating any identified mutants from their product development pipelines. And regulatory agencies should ensure that developers of GMOs have carried out this necessary, straightforward screen of potential products.
Another source of abnormalities in GR2-R1 rice related to genetic engineering
Bollinedi et al. (2017) identified an additional problem with GR2-R1 rice, a problem that could affect other GR2 Golden Rice events as well, like the GR2E rice being readied for commercial release.
The foreign genes responsible for increasing levels of beta-carotene in all of the GR2 “events” of Golden Rice are meant to be expressed only in rice grains because expressing them in other parts of the plant could unintendedly disturb other biochemical pathways that share a common substrate, such as those involved in the biosynthesis of chlorophyll or various plant growth regulators. To accomplish this specificity, genetic engineers at Syngenta (Paine et al. 2005) spliced the “promoter” region of a rice gene called Glutelin I, which is only expressed in grain (the promoter is responsible for turning a gene on or off), to the protein-coding regions of the genes involved in beta-carotene synthesis that they wanted to insert into rice plants; then they inserted the resulting chimeric genes into rice plants via genetic engineering.
But in characterizing the abnormal GR2-R1 Golden Rice plants (produced by Syngenta as mentioned above), Bollinedi et al. found that expression of the foreign chimeric genes was not grain-specific as had been intended, the foreign genes were expressed in leaves as well. These scientists hypothesized that this “leaky” expression of the chimeric foreign genes was responsible for the pale leaves and (some of the) growth regulator-related abnormalities in GR2-R1 rice plants (because biosynthesis of green chlorophyll pigments and expression of plant growth regulators had been disrupted due to competition with the GE enzymes for beta-carotene biosynthesis). They documented dramatic alterations in the expression of four key plant growth regulator genes in support of their hypothesis.
The future of Golden Rice
In light of these problems with GR2-R1 Golden Rice that are related to the process of genetic engineering, IRRI is now concentrating on developing GR2E Golden Rice for commercial release. IRRI (in conjunction with the Philippine Rice Research Institute) submitted an application for a biosafety permit for the “direct use in food, feed, or for processing of GR2E Golden Rice,” (http://irri.org/golden-rice/faqs/what-is-the-status-of-the-golden-rice-project) to the Department of Agriculture-Bureau of Plant Industry (DA-BPI) in the Philippines in early 2017. (A link to the collection of Golden Rice data files can be found on this website: http://biotech.da.gov.ph/Decision_docs_jdc_direct.php, about 2/3 of the way down the page.)
Information in the submission package appears to indicate that GR2E Golden Rice has been checked and that a mutation was not created at the site in the rice genome where the foreign genes were inserted. I’m not as convinced by the submitted data that the “leaky” endosperm promoter issue has been adequately analyzed.
I hope Golden Rice will be successful in reducing vitamin A deficiency. But we need to remember that “Precision Agriculture (GMOs)” is not as precise as some would have you believe. There is a lesson to be learned from GR2-R1 Golden Rice, and it is to be more vigilant about the less precise aspects of crop genetic engineering.
Biotech Salon 
You are right that the FDA letter you mention is peculiar. But a fuller elaboration of the reasoning behind what in effect amounts to an approval of Golden Rice as safe may be found in the FDA summary document on which the letter was based, which is available at:
Click to access ucm607450.pdf
The penultimate paragraph states: “FDA evaluated IRRI’s submission to determine whether GR2E rice raises any safety or regulatory issues with respect to its uses in human or animal food. Based on the information provided by the company AND OTHER INFORMATION AVAILABLE TO THE AGENCY, FDA did not identify any safety or regulatory issues under the FD&C Act that would require further evaluation at this time.” [Emphasis added.]
Note this says that the “FDA did not identify any safety or regulatory issues.” This means the FDA found the rice as safe.
With regard to the FDA statement that “the concentration of Beta-carotene in GR2E rice is too low to warrant a nutrient content claim,” you may be interested in the following response provided to me by Adrian Dubock, secretary of the Golden Rice Humanitarian Board: “The FDA comments are regarding the US population, their average consumption of rice (45g daily) , and with reference to the Recommended Daily Allowance of beta-carotene. The RDA includes enough beta-carotene to maintain 3 months liver stores. Such are not required to combat VAD. RDA is 2.5X standard deviations larger than the Estimated Daily Intake. And nutritionists consider that 30% to 40% of the latter consumed daily is sufficient to prevent the nutritionally acquired immune deficiency syndrome which is VAD, and causes 23 -34% of all under 5 years mortality globally as well as being the leading cause of childhood blindness.”
All of your comments regarding the Bollinedi paper are to a GR event that has been discarded long since, GR2R. All of the GR approvals to date, and all current work on Golden Rice, is on event GR2E. Haritha Bollenedi, author of the paper you mention, was a post-doc working in the lab of Ashok Kumar Singh, head of the IARI (Indian Agricultural Research Institute) Division of Genetics. In a personal communication to me, Dr. Singh has said: “The anti-GM lobby has misinterpreted our findings as a failure of GR technology. One must understand that the problem with the R event is event-specific and nothing to do with GR technology per se.”
Thanks for the additional information and for your opinion as to what conclusion of the voluntary consultation process with FDA really means. Based on the stated purposes of the FDA summary document you cite, i.e. “The purposes of this evaluation are (1) to assess whether the developer has introduced into human or animal food a substance requiring premarket approval as a food additive and (2) to determine whether use of the new plant variety in human or animal food raises other regulatory issues under the Federal Food, Drug and Cosmetic Act (FD&C Act),” I’d say that FDA concluded–based on the current system for “regulating” GMOs in the US–not that it approves GR2E rice but rather that GR2E doesn’t require the agency’s approval.
I am very aware of the differences between Golden Rice events GR2R1 and GR2E, and I was careful to identify these different events in this post. I agree with Dr. Singh that the problem with the R event–at least the problem of its being an insertional mutant–is event-specific. But the technology, per se, has everything to do with that fact since the mutation was caused by the genetic engineering process itself. My point is not that GR2R1 is indicative of a failure of GR technology but that, because the risk of causing insertional mutations is inherent in the process of genetic engineering, events like GR2R1 should be identified and eliminated very early in a product development pipeline–much earlier than the mutant GR2R1 was identified (which was during the process of introgressing the GE GR2R1 trait into a rice variety favored by Indian farmers). Drs. Singh and Bollinedi apparently agree with me about this as they conclude their paper thusly: “The study conclusively demonstrates the importance of event characterization and event selection before adopting the transgenic germplasm into introgression breeding.”
And the problem of the leaky endosperm-“specific” promoter may not be event-specific but rather construct-specific, in which case the foreign genes in the GR2E event (and other GR2 events) could be expressed in vegetative tissues as well.
Re: “the risk of causing insertional mutations is inherent in the process of genetic engineering,…”
In fact, the risk of causing genomic mutations (even if not “insertional” in the technical sense in which you are using the term) is inherent in all forms of plant breeding, not just genetically engineered plants. A mutation, whether caused by molecular or conventional methods, is a kind of unintended effect, an alteration of the genome.
A comprehensive review of the unintended effects problem was produced by a group of fourteen European scientists who in 2004 published in the journal Food and Chemical Toxicology a study, “Unintended effects and their detection in genetically modified crops.”
https://www.ncbi.nlm.nih.gov/pubmed/15123383
In this study, the authors make the following points. First, in regard to the chromosomal rearrangements caused by transgene insertion, there is the fact that integration of an insert into a plant genome is subject to the same natural DNA repair mechanisms that are used by cells to repair normal, untransformed DNA.
“The major source of natural variation and of breeding programmes is the natural molecular mechanisms of DNA exchange and repair,” the authors wrote. “These mechanisms are the same for all crops, irrespective of whether the DNA has been specifically modified by genetic engineering techniques or has been altered by via conventional crossing of different varieties.”
Second is the important fact that unintended effects occur in all forms of plant breeding, whether made by traditional or by molecular methods. In other words, unintended effects was not a new complication introduced by recombinant methodology, but was one that breeders had long been aware of, and for which they had developed and adopted corrective measures. “In classical breeding programmes, extensive backcrossing procedures are applied in order to remove unintended effects.”
Moreover, a body of research evidence exists which shows that genetic engineering methods in fact cause less disruption to the host genome than do traditional plant breeding methods. A 2006 study by a group of British scientists compared the gene expression profiles of transgenic versus conventionally bred lines of wheat.
https://www.ncbi.nlm.nih.gov/pubmed/17177803
By the use of microarray analysis the authors determined that conventional breeding methods introduced greater changes in gene expression than did genetic engineering techniques.
“Differences in observed gene expression in the endosperm between conventionally bred material were much larger in comparison to differences between transgenic and untransformed lines,” the authors wrote. “These results suggest that the presence of the transgenes did not significantly alter gene expression and that, at this level of investigation, transgenic plants could be considered substantially equivalent to untransformed parental lines.”
Re: “events like GR2R1 should be identified and eliminated very early in a product development pipeline–much earlier than the mutant GR2R1 was identified (which was during the process of introgressing the GE GR2R1 trait into a rice variety favored by Indian farmers).”
In fact, the gene that produced the mutations in GR2R had been identified as the Aux1 gene as early as 2006, but was only brought to the attention of the Golden Rice Humanitarian Board in 2013. See the document below, page 79, and more particularly p. 81, in the paragraph starting “In a December 2013 meeting,…”
Click to access Dubock-The_present_status_of_Golden_Rice-2014.pdf
By 2013, however, the R event was already being introgressed into local varieties by IARI and others. A tragicomedy of errors, and a sad outcome given what’s potentially at stake with Golden Rice.
Yes, the risk of causing mutations in crop plants is inherent in traditional breeding methods as well as in genetically engineering methods. But when using genetic engineering methods, developers of new crops can identify–and then eliminate from their product development pipelines–any events with insertional mutations that were unintentionally created during the engineering process. And I think they should (even be required to) do so; case in point: the Golden Rice project would certainly be better off today if developers had thrown away the GR2G and GR2R events back in 2006 when IRRI first received the molecular data on those events from Syngenta.
And yes, in traditional breeding programs “extensive backcrossing procedures are applied in order to remove unintended effects.” Maybe extensive backcrossing procedures should be (required to be) applied when developing GE crops as well.
As for comparing disruptions to genomes caused by traditional breeding methods versus genetic engineering…I think that body of research needs to be greatly expanded upon. We need more examples comparing GE crops vs traditionally bred crops vs their parent lines, in order to improve our understanding of the kinds of differences in each case. As far as I know though, microarray analysis in out of favor these days; as suggested in the most recent National Academy of Sciences report on GMOs, the various “omics” techniques should be employed for such studies.
And thanks for alerting me to Adrian Dubock’s paper “The present status of Golden Rice” (2014). I certainly agree with you that the Golden Rice story is a tragicomedy of human errors (in fact I think it deserves an entire blog post of its own)! But in fact, according to Dubock (see p. 79), the molecular data from Syngenta “was found by IRRI post March 2009” and mentioned in Dubock’s e-mail of December 1, 2009 to the Humanitarian Board; that data indicated that both GR2G and GR2R were insertional mutants…and was thereby known to the Humanitarian Board years before the multi-location field trials in 2012 and 2013 were planted. It makes me wonder why Syngenta even contributed those particular events to the Golden Rice project; aren’t all genetic engineers checking to make sure their potential products aren’t insertional mutants? If they aren’t, the regulatory agencies should require them to.
Pingback: Golden Rice: Of Good Intentions, Insertional Mutants, Human Error, and the Need for Better Regulation of GMOs | Biotech Salon