Following our panel discussion ‘Can Organic GMOs ever be a ‘thing’? held at Natural and Organic Products Europe in London in April, two of our panellists continued their discussion and eventually agreed to publish an exchange of views and ideas on some of the points raised or arising from the session.
Huw Jones*, Professor of Translational Genomics for Plant Breeding at Aberystwyth University kicked off the exchange with a series of questions about organic for Lawrence Woodward, Director of Beyond GM and co-founder and former Director of the UK’s Organic Research Centre. Woodward responded to these questions and posed some of his own.
We present that exchange below so that readers can consider both sides of what is a very big and very current discussion. We encourage readers to enter into and expand on the discussion on our Facebook page.
Huw: There are many seed varieties and heritage breeds currently available to organic growers. Does organic agriculture need new plant and animal breeding at all?
Lawrence: Yes, there are many (conventionally bred) seed varieties available to organic farmers – if they have been grown in an organic system and, by derogation, if conventionally grown. There are also heritage breeds available (although there is not an agreed definition of ‘heritage’) accessible on the same terms. However, there is no doubt that organic farming does need to develop new plant and animal breeding to keep improving and adapting to environmental changes. Far from being stuck in the past, as some agro-chemical industry PR likes to portray, organic farming has always been progressive and innovative albeit within ecological limits.
Huw: Does organic agriculture currently embrace genetic innovation in breeding such as F1 hybrids and current precision breeding methods?
Lawrence: Biodynamic producers have a conceptual objection to F1 hybrids. However, these are used by many other types of organic producers although there is a pronounced move towards open-pollinated varieties (OPs) amongst some vegetable producers. At times this is accompanied by a discomforting “F1s are bad narrative”, but a significant number of organic producers believe that F1s have many benefits, whilst OPs have others and that both should be available. There are no technical reasons (but several economic ones) why F1 parent lines cannot be maintained in an organic system and why, therefore, organic F1 hybrids can’t be produced. Under organic regulations conventionally produced F1s are used in organic systems by derogation.
EU organic regulations are framed to move towards a situation where all seeds used in organic production will, in time, be produced in an organic system and, in due course, they will all be “organically bred”.
This means that, at present, varieties bred through “traditional” or “random mutagenesis” are allowed for use in organic farming on the terms of derogation outlined above. However, this has long been a contentious issue within the organic sector and a decision was made several years ago to phase out their use; a process which is being hastened by the recent European Court of Justice (ECJ) judgement.
The ECJ made it clear that although some of these methods are exempt from the EU GMO Directive on the grounds of a “long safety record”, they “are GMOs within the meaning of the GMO Directive, in so far as the techniques and methods of mutagenesis alter the genetic material of an organism in a way that does not occur naturally.”
In your presentation at Natural and Organic Products Europe you argued that a) naturally occurring, spontaneous cuts in DNA happen the whole time and these cuts are essentially the same as those made by plant breeders, either indirectly through traditional mutagenesis or directly through genome edited targeted mutagenesis; b) genome editing leads to mutations that are identical to naturally occurring ones or to ones induced by traditional mutagenesis; and c) therefore it doesn’t matter how the cuts are made as long as the plant’s natural repair mechanisms are engaged and no foreign DNA is left at the end of the process. Are you referring here to the whole gamut of genome editing techniques or to some and not others?
Huw: I was not referring to editing techniques such as ODM (oligonucleotide directed mutagenesis) that by definition use a template of additional DNA. I was referring to those editing methods that generate double-stranded breaks in DNA using a guided endonuclease (such as Crispr-Cas9, Talens and ZFNs). Also, irrespective of whether they were delivered as nucleic acids or as proteins. These make cuts in genomic DNA that are synonymous with the background random ‘trauma’ that DNA undergoes during exposure to mutagens found in the environment (e.g. UV light) or to cellular events (such as transcription or DNA replication) or indeed, those imposed by more intense sources during the initial steps of mutation breeding. I was also referring to ‘simple’ editing where cutting DNA was the only intervention (i.e. there was no attempt to direct the repair using a band aid-type template).
There is already a divergence of organic certification standards regarding breeding methods (such as for F1 hybrids) and, as you say, by derogation, mutation-bred seeds can be used in organic agriculture. Therefore, it is logical that seed varieties possessing targeted and simple mutations generated by gene editing (that cannot be distinguished from natural mutation) should also be used in some forms of organic agriculture. Do you agree?
Lawrence: I think many in the organic sector would disagree and some background in organic concepts may be helpful in understanding why.
The development of the concepts of organic plant breeding and seed has been primarily driven by ideas rooted in, anthroposophical, holistic, socially focussed agro-ecological perspectives. From these emerge the precept to “respect the genome and the cell as an indivisible functional entity” and to “follow the concept of respecting integrity of life.”
The consequence of this precept for organic plant breeding is that: “any technical or physical invasion into the isolated cell is refrained from and plant specific crossing barriers are respected, irrespective of potential benefit risk assessments.”
Until recently this might have been dismissed as “unscientific”, woolly or even “esoteric” but, arguably, it is now finding support in the concept of omnigenics.
According to Dr Michael Antoniou of the Gene Expression and Therapy Group at Kings College, University of London, “complex traits have at their basis the integrated functioning of the entire genome of an organism” which means that “the entire complement of genes (the “genome”) is involved in delivering complex characteristics in the organism.”
“It’s possible”, he writes “that a set of ‘core’ genes may be at the basis of complex traits, but omnigenics reveals that their function is augmented by all the other genes that are expressed in a given cell or tissue. Crucially, omnigenics suggests that genes in a cell should be viewed as a network.”
For these reasons, and others, there is currently no place in organic for this type of breeding. Having said that, the organic sector is not monolithic, there is a great deal of overlapping thinking and there are grounds for thinking that a differentiated perspective might emerge on the question of “targeted mutagenesis”.
It is likely, though, that the relative importance of this for both the organic sector and the wider sustainability movement would be significantly influenced by: a) what risk and safety assessments are being proposed for genome editing; b) the degree of transparency in respect of both labelling the final product and decision-making about the use of public money; and c) whether genome editing will lead to further corporate control (e.g. through patents etc), industrialisation and intensification of the food and farming sector.
Huw: OK, so not all organic concepts are based on science and logic and the process of breeding is as important as the product. However, new breeding techniques are making this position increasing difficult to maintain. You say that conventional mutation varieties continue to be acceptable in organic agriculture by derogation. However, gene editing can produce an identical DNA change, albeit using a slightly different process. It follows that organic principles would distinguish between two plants possessing identical mutations, allowing the one made by untargeted mutation breeding methods to be cultivated organically but not the one made by more targeted gene editing. This makes no logical sense to me and I can see no way to put this into practice, especially when even DNA sequencing cannot always tell them apart.
You mention corporate control. How do you believe corporate ownership and intellectual property rights influence availability of seeds and breeds for organics? For instance, if a new gene-edited variety with qualities ideally suited for organic cultivation were distributed freely and allowed reuse of farm-saved seed etc., would this be acceptable?
Lawrence: There is no doubt that corporate ownership and intellectual property rights influence the availability of seeds and breeds for the whole of agriculture, not just organic. These factors narrow the available gene pool; access can be limited, directed and controlled without transparency and accountability; and production of “minority varieties” which may be important to the development of alternative farming and food systems can be ignored or deliberately starved of resources. There are clearly other factors and influences involved, but the goals of corporate ownership and intellectual property ownership (including through academic institutions) is a pivotal factor. I suppose that it is possible for a corporation or a corporately driven research institution make available an organically suitable variety to be distributed freely and for the seed allowed to be saved. But it would most likely be as one off loss leader or a PR sop and such largesse could be withdrawn at any time. The essential objection is about the systemic control.
In contrast, organic plant breeding is a holistic approach where the process of breeding, including technical, socio-economic and ethical aspects, is equally important as the final product (the cultivar) with its characteristics.
It follows, therefore, that seeds and breeds with targeted mutations from genome editing should not be used in organic agriculture and the current system, based on corporate ownership and intellectual property rights, is wholly out of tune with organic breeding and seed production and distribution.
Furthermore, the socio economic aspects sought in organic plant breeding and seed production are characterised by “promoting free exchange of germplasm, transparency of the breeding process, open pollinated varieties instead of F1-hybrids, participatory breeding involving farmers and the value chain and a plurality of breeding initiatives to enable a more diverse and sustainable agriculture.”
This would seem to pose insurmountable obstacles to acceptance by the organic sector of genome editing-based plant breeding and seed supply and a system characterised by corporate ownership and patents.
I note that the recent Australian government decision on gene technology has made a distinction between genome editing methods:
It will “deregulate” the use of tools in which proteins cut DNA at a specific target site – as long as the tools allow the host cell to repair the break naturally, rather than using a template containing genetic material to direct the repair process. But will continue to regulate “gene-editing technologies that do use a template, or that insert other genetic material into the cell”.
Is this an approach you agree with and would support in the UK and the EU? If so, would you label or in some other way inform end users and the public which varieties or plant material have been produced with which method?
Huw: The Australian decision reflects exactly my view for a workable and logical approach to identify a trigger for regulation. It reflects the position outlined in my earlier answer and deregulates products of editing methods that generate double-stranded breaks in DNA using a guided endonuclease (such as Crispr-Cas9, Talens and ZFNs). The Australian approach would consider mutations made using a DNA template as conventional GMOs and regulate them as such.
I don’t think this is a perfect solution, however, as directed DNA repair using a template or ODM approach could also generate mutations equivalent to those found naturally. Also, without modification, the GMO risk assessment process would be disproportionate for these types of mutations, but, it seems a practical and pragmatic way forward at the present time.
However, it would be illogical to enforce a legal requirement for edited mutations that are not included in the GMO legislation to be labelled when far more disruptive ones, possessed by mutation bred varieties, or other types of plant breeding are not labelled. There is a great opportunity here. I would prefer a voluntary code of conduct and transparency that informs consumers about plant breeding and how new varieties are developed.
Lawrence: I know you take issue with the precautionary approach of the ECJ ruling and that our views on what is ‘natural’ might be different. But I was interested to see Jack Heinemann, at the University of Canterbury, Christchurch, in New Zealand point to experiments which have shown “that some areas of genomes are less susceptible to mutations by radiation or chemicals, because of the way the DNA is compacted and protected by proteins and other molecules.”
He argues that “the way the molecular scissors in gene editing are guided to particular sites on the genome may force changes to these sites, which could otherwise be protected.”
He also recognises that; “scientists are currently working to reduce CRISPR’s known ‘mistake’ rate“, but fears that “speedy gene editing could rapidly introduce cumulative unintended changes that, if not easily observed, may not be picked up before an animal or plant was commercialised.”
Do you accept that there is some validity in his argument?
Huw: In his work Jack Heinemann makes several points. He points to areas of uncertainty regarding the relative mutation susceptibility of different genomic regions and I accept this uncertainty. Some genes certainly evolve faster than others, but even if some regions of the genome are harder for gene editing to access, it does not diminish its utility.
He and others also talk about the dangers of ‘unintended’ changes as if these were absent from other breeding approaches. Of course, the original edited plant may well have off-target effects, like the progeny of an initial wide cross in a conventional breeding programme.
Plant breeders, conventional and organic, deal with genetic changes (without calling them intended or unintended) all the time and will select the edit alongside the other characteristics they aim for in that specific breeding scheme.
Lawrence: It seems to me that the logic of Prof Heinemann and Dr Antoniou’s arguments leads away from the proposition of reduced regulatory evaluation and oversight of genome editing and towards a reformed system which includes the relatively new “omics” technologies – such as transcriptomics, proteomics, metabolomics, epigenomics and mirnomics – which can provide molecular compositional profiling and therefore a greater insight into potential food safety issues.
Is this something you would agree with? If not, could you explain why not?
Huw: On the contrary, I see nothing in their arguments that calls for new breeding techniques to be evaluated more rigorously than mutation breeding or other conventional methods which have been producing the wide range of varieties currently cultivated organically or otherwise. I would support any regulatory system for breeding that is logical, consistent and proportionate to the hazards. I would certainly not advocate using ever more sophisticated methods of analysis without a clear idea of what to look for.
Generating massive omics datasets would not only be hugely expensive but more importantly, merely kick the hazard characterisation steps further down the road.
We need clear hypothesis-led risk assessment. For example, if a plant possesses a known and potent allergen, it is reasonable to ask whether breeding (however that is done) has altered the levels of that substance. To generate billions of untargeted data end points without a hypothesis of harm and with no clear understanding of what levels of specific allergens are safe, would not advance the science and is a waste of valuable risk assessment resources.
Indeed, it would merely show that all breeding methods generate significant differences in composition by shuffling genes and alleles (the whole point of breeding!). Ironically, these proposals probably would be supported by big multinationals who are the only ones that could invest in the hardware and expertise to deliver it as part of an already hugely expensive GMO data package.
As above, generating the gene edit is only one aspect of the breeding that is needed to generate a new variety that is ‘distinct, uniform and stable’. Any potential off-types or unexpected outcomes would be dealt with by breeding practices of back-crossing/selfing and selection.
Lawrence: There is no doubt that genome editing is a powerful technology. You and others have pointed out its potential to be a “game-changer” in food and farming. Don’t you think therefore, that there is a case for comprehensive labelling of its use? If not, do you think there should be some other method of informing consumers and the general public about products and processes where it has been used?
Huw: It is a game-changer because it makes breeding, particularly in vegetative-propagated and hard-to-breed crops, more targeted and predictable. Simple gene editing does not make breeding any riskier. Although I partially addressed this question above, I will reiterate that it would be illogical to label gene edited varieties when they cannot always be detected as such and when other forms of breeding are not labelled. However, I do see an opportunity for making information more available on all plant breeding methods. There is an amazing lack of understanding about how different plant varieties are generated and I would love to see more consumer information about the breeding processes used. Of course, if editing stays within GM regulations, your question is redundant within the EU as it would automatically be labelled as a GMO.
Under current law gene-edited crops cultivated overseas (even those cultivated organically) would need EU GMO authorisation (including the requirement to be labelled) before they can be imported. Although, I do not anticipate anyone doing this due to the significant costs and uncertainty of timescales. For the countries that have put editing into their legal frameworks covering conventional breeding, there is no requirement to label. However, I would support total transparency within the food chain for all breeding approaches.
Lawrence: But with such a powerful and “game-changing” technology don’t you think citizens should have more of a say in how public money is applied?
Huw: Of course, the public should have a say in how their taxes are spent. However, it is private companies that currently do most of the global applied biotechnology research. One way to give citizens more of a say, assuming products are safe, is to get them on the market and allow farmers/consumers to decide whether or not to buy.
I would love to have more dialogue on how public money is spent on research. I am just not sure how that can be done in an informative and practical way. Interestingly, I was closely involved in a similar public dialogue exercise on how my previous employer, the BBSRC-funded Rothamsted Research should engage with industry. It involved only 50 members of the public over two locations in UK so I’m not sure how representative it was of the UK as a whole. I remember it was certainly a huge amount of work! The resulting report is here.
Lawrence: We at Beyond GM feel it is a mistake to try to do this from behind silo walls – especially so for those people from institutions and organisations which purport to have a public good/service remit, which is why we were grateful for your participation in the panel session at Natural and Organic Products Europe and for your willingness to continue the dialogue.
Clearly there are significant differences involved in any discussion about these technologies – and some of them might be fundamental and incompatible. However, that does not mean there is no common ground or that plurality of approaches is not possible and cannot be found.
Achieving this might be the best – possibly the only achievable – way out of the silo of stalemate and conflict. That requires the development of an understanding of all sides and a willingness to accommodate – which requires mature discussion and discourse. Which, I hope, we can continue.
Huw: I absolutely agree. I can input into the discussion from the position of a research scientist with some expertise in risk assessment, but I appreciate that there is a broader discussion to be had with a wide spectrum of stakeholders. The challenges of global agriculture are significant, the issues are complex and safety is paramount. Plant breeding has a long history of safely incorporating new techniques and developing better varieties for horticulture and field. I fear that EU farmers and consumers will be deprived of choice while other areas of the world are taking advantage.
Lawrence: The issues relating to the boundaries of plant breeding, seed maintenance and supply and now the new genome editing methods are overarching and pose challenges across the organic sector – as I am sure they do within the conventional sector where its strives to address sustainability concerns. On the point that we need to get more stakeholders involved we can, at least, both agree!
*The views expressed here by Huw Jones are his own and are not necessarily representative of other organisations and bodies he is associated with.