Activities of the Open-Ended Online Forum (2014-2016)

Dear members of the Online Forum,

Please use this thread to share views, relevant guidance and sources information on “LMOs created through use of dsRNA techniques, engineered to produce dsRNA or dsRNA” and “LMOs containing RNAi”.

Thank you,
Manoela

Dear Colleagues:

We thank the opportunity to contribute to this forum. We have a general caveat regarding to add more “boxes” to the guidance: maybe efforts would be more effective if focused in reaching first a consensus on the more general roadmap text, and leave the discussion of additional guidance applicable to specific cases for latter.

Nonetheless, surely the present forum exchange will be of use; therefore please note the following comments coming from our experience in the biosafety assessment of this kind of LMOs in Argentina:

a) We have found the following study to be very useful:

Dubelman S, Fischer J, Zapata F, Huizinga K, Jiang C, et al. (2014) Environmental Fate of Double-Stranded RNA in Agricultural Soils. PLoS ONE 9(3): e93155. doi:10.1371/journal.pone.0093155

Most of the risk hypothesis related to these products would involve a pathway to harm where the resilience of the RNA molecule in the environment is decisive to assess the validity of the hypothesis.

Although the study was performed with a certain LMO and a certain RNA as a model, we consider the conclusions to be of general application because, in contrast with proteins, changes in sequence would not be of significant impact on dsRNA chemistry and the processes leading to its degradation in the environment.

b) Where only a ss or dsRNA molecule is generated from a transgene, obviously no protein is generated. Therefore, any general guidance* regarding toxicity (other than the RNA interference effect) or allergenicity, would not be applicable to the assessment of risks derived from that transgene.
* Much of the general guidance on toxicity and allergenicity in the Guidance draft seems to be out of scope and would belong to the discipline of food safety assessment of rDNA-derived foods under Codex Guidelines.

c) In cases where the dsRNA sequence is designed for controlling an insect pest, bioinformatics studies may be warranted; and potentially significant results would be those with at least a stretch of 21nt perfect continuous sequence homology against sequences from non-target relevant arthropod organisms. These bioinformatics studies may be taken into account, in a tiered approach, to assess the need and amount of information eventually required from laboratory or field studies regarding risks to NTOs.

d) In early examples of crops approved for cultivation (in other countries) where the trait is due to RNAi from the transgene, the safety assessment was performed in a moment where knowledge on RNAi phenomena was not established and the genetic engineering actually was designed to generate the traits through different mechanisms. Nevertheless, determination of the exact mechanism was not decisive for performing the risk assessment and the LMOs proved to be safe after several years of continued use. This is an important precedent to establish the line between “need to know” for doing a regulatory risk assessment and “nice to know” in terms of scientific curiosity. This is especially important in the light of the vast amount of information and methods related to RNAi that are nowadays available, yet are not all of them essential for performing the risk assessment.

Thanks once more to the facilitators and the Secretariat for the opportunity to exchange on this issue. We are also very interested in learning from your practical experiences in performing risk assessments for this kind of LMOs.

Best Regards,
Martin Lema
Argentina

Dear colleagues,

Thanks to Francisca for moderating this forum.

We have been asked by the moderator, among others, to mention in which part of the Roadmap the key ideas, concepts, examples etc. on each specific topic should be incorporated.
However, it is not clear to me how and in which form these topics will be included in the text of the improved Roadmap. I support Martin Lema [#7580] in that it would be more efficient in reaching first a consensus on a more general Roadmap. Also my understanding was that the improved Roadmap would be written in such a way that it would be applicable to all LMOs, including “LMOs created through use of dsRNA techniques, engineered to produce dsRNA or dsRNA”, “LMOs containing RNAi”. Only in case there will be LMOs that do not longer fit under the improved general Roadmap, further guidance (mentioned in boxes or in separate guidance) on specific topics may be considered. 

Sorry to be of no further help,

Kind regards,
Boet Glandorf,
GMO Office, the Netherlands

Dears,

Thank you for the opportunity to contribute to the discussion.

To be useful, the ERA considerations given in the guidance/road map should be applicable to all LMPs. I therefore support the views of Boet Glandorf [#7586] and Martin Lema [#7580] that the guidance/road map should be sufficiently robust to capture the full spectrum of LMPs, and that the addition of boxes for each peculiar group of LMPs is not an ideal solution. Instead of highlighting specificities in various boxes, it would perhaps be worthwhile to use RNAi-based LMPs as a case study, in order to test the robustness of the guidance/road map and to identify commonalities and differences with LMPs that express novel proteins. Based on the outcome of this exercise, specific ERA considerations could be fine-tuned to account for the specificities of RNAi-based LMPs, where appropriate. The below references, which have not been shared previously, could be useful to support such an exercise.

I agree with Jack Heinemann [#7595] that relevant terms should be defined clearly and that it would be beneficial to dedicate necessary efforts to reach agreement/consensus on definitions/interpretation of terms. However, in my view, it would be best to define relevant terms in the glossary instead of boxes that will be scattered throughout the core text. In the core text of the guidance/road map, one could easily cross-refer to the glossary, where appropriate.

Just a few thoughts.

With thanks,
Yann

-Casacuberta JM, Devos Y, du Jardin P, Ramon M, Vaucheret H, Nogué F (2015) Biotechnological uses of RNA interference in plants: risk assessment considerations. Trends in Biotechnology, 33: 145-147, http://www.cell.com/trends/biotechnology/pdf/S0167-7799(14)00252-2.pdf

-European Food Safety Authority (EFSA) (2014) International scientific workshop ‘Risk assessment considerations for RNAi-based GM plants’ (4–5 June 2014, Brussels, Belgium). EFSA supporting publication 2014:EN-705, 38 pp, http://www.efsa.europa.eu/sites/default/files/corporate_publications/files/705e.pdf

-Ramon M, Devos Y, Lanzoni A, Liu Y, Gomes A, Gennaro A, Waigmann E (2014) RNAi-based GM plants: food for thought for risk assessors. Plant Biotechnology Journal, 12: 1271-1273, http://onlinelibrary.wiley.com/doi/10.1111/pbi.12305/epdf

-Roberts AF, Devos Y, Zhou X, Lemgo G (2015) Biosafety research for non-target organism risk assessment of RNAi-based GE plants. Frontiers in Plant Science, 6:958.doi:10.3389/fpls.2015.00958, http://journal.frontiersin.org/article/10.3389/fpls.2015.00958/full

Dear Francisca, the Secretariat and Online Forum participants

We have one additional review article for this forum;
Shah et al. RNA Interference for Insect Pest Management – Recent Developments: A Review. Journal of Cell and Tissue Research, 14(3): 4601-4608 (2014).

Further, we would like to mention that the Trans-Tasman food regulator – Food Standards Australia and New Zealand (FSANZ) - has issued approvals for food derived from;
- Potatoes modified to produce RNA against a virus. Lines: RBMT21-129, RBMT21-350, RBMT22-82 (NZ BCH entry 108394) and RBMT15-101, SEMT15-02, SEMT15-15 (NZ BCH entry 108397)
- Soyabean modified with RNAi sequence to increase the level of oleic acid. Lines: DP-305423-1 (NZ BCH entry 108410) and MON87705 (NZ BCH entry 108458)
- Lucerne modified with RNAi sequence to reduce the level of lignin. Line KK179 (NZ BCH entry 108449)
- Corn modified to produce dsRNA against corn rootworm. Line MON-87411-9 (NZ BCH entry 108883)
The FSANZ food safety risk assessments can be accessed via the New Zealand BCH Profile. Environmental safety assessment information for some of the cultivated lines can be accessed via ISAAA (the International Service for the Acquisition of Agri-Biotech Applications).

Thanks, Stacy

Dear colleagues,
thank you all for the very interesting discussion and the opportunity to participate in this forum.
I support Martin Lema [#7580], Boet Glandorf [#7586] and Yann Devos [#7610] and share their view that the roadmap should be written with the purpose to cover the full range of LMOs and therefore should keep a robust and general character. Explicit addition of boxes for RNAi based LMOs to the roadmap with links to steps in the RA process would consequently entail the addition of similar boxes for other groups of LMOs. In consideration of the rapid technical development in the field of biotechnology this approach would delimit the scope of the guidance. All textboxes included so far support the understanding of the risk assessment process in general.

It is rather necessary, and in this point I explicitly support the proposal of Yann Devos [#7610] and Boet Glandorf [#7586], to test the robustness and applicability of the guidance in terms of RNAi-based (or other novel) LMOs. Further guidance on specific topics should be considered only if LMOs do not fit under the general roadmap.

I think Vincenza Ilardi [#7591] has made a good point in questioning the wording in the topic of this discussion “LMOs created through use of dsRNA techniques, engineered to produce dsRNA or dsRNA” and “LMOs containing RNAi”. I agree to her alternative text since it is better suited to describe what needs to be covered by the guidance.

I addition to her argument, I think the starting line “LMOs created through use of dsRNA techniques” might be misleading. I am not aware that “the use of dsRNA techniques” can create a “living organism that possesses a novel combination of genetic material” as would be required according to Article 3 (g) of the Protocol to create a LMO. I think the focus in the definition of the term LMO lies clearly in the novel combination of genetic material. Therefore the modification of gene regulation or the metabolism of an organism via dsRNAs does not create a LMO if no novel combination of the genetic material of the organism is created. Rather dsRNAs are already covered by the current guidance document as gene products (cf. “Use of terms”) of certain LMOs and should be treated as such throughout the risk assessment process.

Having said this, it is clear that I also support Yann’s [#7610] call that relevant terms should be defined clearly and efforts to reach agreement/consensus on definitions/interpretation of terms are beneficial. I would like to advocate to stick to definitions already agreed on wherever possible. I think we should be very conservative with expanding interpretation of terms that are well understood.

Thank you
Kind regards

Werner

Dear all,

In the discussion of this specific topic I understand that we are talking about a transgenic plant using the RNAi mechanism to obtain the characteristic we are interested and those are the examples we have in the market to test if the Road Map can be used to assess the risk.

The topic use of dsRNA or the use of dsRNA techniques are not in the scope of CPB, according to the LMO definition – In this case I would like to agree with Dr. Werner [#7628] about the fact that “the modification of gene regulation or the metabolism of an organism via dsRNAs does not create a LMO if no novel combination of the genetic material of the organism is created”.

In the case of transgenic plants obtained through the RNAi mechanism I´d like to present the case of RA of GM common bean approved in Brazil expressing virus resistance. The GM bean was assessed using our regulatory framework, using different risk hypothesis, and no new issue regarding risk expressed by RNAi was identified concluding that existing guidance/regulations as much as available scientific literature can be used in the RA of this organism. The same rationality can be applied to the RoadMap – the Elements for Consideration should be enough for a risk assessor to build the risk hypothesis related with a RNAi modified organism without the necessity of new sections or boxes.

Some references for GM common bean are available at:

http://www.ctnbio.gov.br/index.php/content/view/17813.html
http://www.ctnbio.gov.br/upd_blob/0001/1750.pdf (dossier in portuguese)
Aragão, F. et al. Journal of Biotechnology v. 166, p. 42– 50. 2013.
Faria, J. et al. Plant Breeding v.133 (5), p. 649-653. 2014.
http://www.fundacaopetermuranyi.org.br/main.asp?pag=2004

Thank you.

Best regards,

Luciana P. Ambrozevicius
Ministry of Agriculture / Brasil

Dear all,

I would like to support the information and opinion expressed by Dr Luciana [#7651].
I would like to add that I took part in the development of the risk assessment formulation and testing (as part of the Embrapa´s team) and we found no evidence of risks to environment neither to human health for the RNAi- common- bean under consideration. I was confirmed by the Brazilian CTNBio when they approved this bean for commercial release.

Thanks Dr Luciana for providing the documents to this Forum

Deise Capalbo

Dear colleagues
Many thanks for the lively discussions and abundance of thoughtful interventions. It is somewhat daunting to proffer any new insights when faced with such a plethora of diverse and worthy ideas. Instead, I would like to add an observation. There appears to be a divide between risk assessment as an academic exercise and risk assessment as part of regulatory decision making. I feel that that the pendulum has swung in favour of referencing many academic treatises, whose publication favour intellectual boldness and innovation. This may present difficulties for emerging regulatory systems in need of decisions based on pragmatic, understandable processes. Therefore, I would support greater referencing of existing regulatory decision documents that have already confronted these box bound suggestions, such as those referenced by Stacy [#7614, see also #7615] and Luciana {#7651]. Let us tease out the specific reference to RNAi or dsRNA techniques in these regulatory decision documents and how they were accommodated in the risk assessment. Unfortunately, the Australian experience may not be of much use as risk assessments that specifically cover RNAi [e.g. http://www.ogtr.gov.au/internet/ogtr/publishing.nsf/Content/DIR131] do not give special status to this technique, but treat it the same as any other type of genetic modification, where greater importance in the risk assessment is given to expected effects at the phenotypic level.
Best regards
Paul Keese

As this is my first posting, I would first like to sincerely thank the Secretariat and Fran for all their hard work.

Due to lack of time, I will not post much on this topic, except to say that I thank Sarah Agapito-Tenfen [#7594] and Jack Heinemann [#7595] for their extensive and well-referenced comments on the matter. 

I support their comments.

Dear colleagues,

I think Yann Devos stated my view more clearly that I could, namely that "the guidance/road map should be sufficiently robust to capture the full spectrum of LMPs, and that the addition of boxes for each peculiar group of LMPs is not an ideal solution".

RNA-mediated phenotypes occur in LM and non-LM plants.  It is not the mechanism of the RNA-mediation that needs attention from the risk assessor, but rather a consideration of the phenotype of the organism, as well as potential interactions that might occur in light of the type of genetic modification and the resultant phenotype.

At the heart of the matter, I still cannot improve upon Yann's statement I quoted above. 

Best regards,

David Heron
Biotechnology Regulatory Services, USDA-APHIS

Dear colleagues,

Persuaded by our moderator’s warm plea to use these final hours of the online debate, I share the following thoughts with you.

My point of departure is the general observation that I made earlier that I believe that the discussions and decisions of MOP7 advise us to first aim at improving the guidance on the basis of the test results and to discuss the revised guidance in MOP8, before new topics are considered.

In relation to the specific topics of RNAi and dsRNA, I share the following thoughts.

First of all, I fully agree with the points made by Werner Schenkel, Luciana P. Ambrozevicius and Vincenza Ilardi that the opening line “LMOs created through use of dsRNA techniques” needs clarification, bearing in mind that the essence of the LMO definition is the creation of novel combinations of genetic material.

Second, I support the notion put forward by Martin Lema, Boet Glandorf, Yann Devos, Werner Schenkel, Robert Andrews and others that the improved Roadmap would be written in such a way that it would be applicable to all LMOs, and that only in case there will be LMOs that do not longer fit under the improved general Roadmap, further guidance on specific topics may need to be considered. 

I particularly like the suggestion of Yann Devos to use RNAi-based LMPs as a case study to test the robustness and applicability of the guidance/road map. Such testing should preferably be conducted by a group of scientists with substantial experience in this field.

To facilitate such testing, I paste below some observations shared by scientists with substantial experience in this field.

1) With regard to RNAi:
Taking the example quoted by Vincenza Ilardi, i.e. the GM plum resistant to Plum Pox Virus (PPV): what needs to be considered is the specific sRNA and what type of scrutiny a specific sRNA may require.   For example, with PPV we know that the sRNA species produced by the GM PPV resistant plum are a small subset of what is produced in PPV-infected plums.  Many more species are produced in infected plums and the species that are the same are produced in much greater abundance in infected plums.  This brings us back to the need for a case by case approach, based upon scientifically valid hypotheses of what might be the potential harm.  It is clear that when small RNAs in a transgenic plant are produced from a plant virus and that plant virus has been present in the environment, and people have been ingesting virus infected fruit for at least a century, it is difficult to hypothesize the potential harm.

2) With regard to dsRNA:
Concern has been expressed [#7594] over the ability of dsRNA to be dispersed by animals that feed on an LMO, and the fact that dsRNA can affect non-target organisms through its ability to induce heritable epigenetic changes. As far as the ability of dsRNA to induce heritable epigenetic changes in mammals, only a few examples of RNA mediated transgenerational epigenetic effects are known (Heard and Martienssen 2014). These are primarily associated with the Piwi-interacting RNAs (piRNAs). The piRNAs have only been found in animals, are germline specific in expression (Klattenhoff, Theurkauf (2008), and are of a distinct class and structure (26-30 nt in length) from small RNA species found in plants (Holoch and Moazed 2015). The piRNAs and piRNA pathway are distinct from RNA-directed DNA Methylation (RdDM) in plants that leads to epigenetic changes, and which occurs via 24 nt small-interfering RNAs (siRNAs). The RdDM pathway is plant-specific and involves multiple enzymes that exist only in plants (Luo and Hall 2007, Ream, Haag et al. 2009, Tucker, Reece et al. 2010, Matzke, Kanno et al. 2015).  Therefore, there is no obvious mechanism through which a plant dsRNA can trigger heritable, epigenetic changes in a mammal. Last but not least, context is important.  Plants contain a considerable amount of endogenous dsRNA. For example, in soybean, there is on average 1.64 ng of endogenous small RNA per 1 µg of total cellular RNA (Ivashuta, Petrick et al. 2009). A large number of endogenous plant dsRNA shows sequence identity to the human and animal genomes and in theory is capable of Watson-Crick pairing to human transcripts (see reference list provided by message #7603). In rice, there are 4759 small RNAs showing perfect sequence identity to the human genome and 270 with perfect homology to the human transcriptome (Ivashuta, Petrick et al. 2009). If plant-derived dsRNA was capable of having a significant impact on the epigenome with lasting, even transgenerational effects, then this begs the question as to why we have not regularly observed such effects.

Best regards to all!

Piet

Dear members of the forum,

Thank you Francisca for moderating this round of discussions and thank you all for the good discussion. I am happy to have the opportunity to participate in this forum.

I support  Sarah Agapito-Tenfen [#7594] and Jack Heinemann [#7595] comments and want to add that the importance of GMO using RNAi technology can be expected to increase in the near future as new applications, such as pest control or altered nutrition, are available.
Using dsRNA expression in plants, e.g. for pest control, raise some new questions concerning the safety of such crops for the environment as well as for human and animal health.

RNAi has only a limited degree of specificity. Even small differences in RNAi pathways between taxa may be sufficient to cause considerable variation in organisms. Thus the risk assessment has to take off-target effects into account. RNAi may, for example, sometimes silence the targeted gene in non-target organisms or may bind off-target and silence non-target, unrelated genes. In addition RNAi saturation and immune stimulation by the RNAi should be considered in the RA. 
Compared to Bt crops RNAi based insect resistant crops can be expected to have a larger activity spectrum both for target and non-target species and methods are lacking to test for sublethal effects of RNAi. Because of the mode of action of RNAi and missing genomic data of exposed non-target species the environmental risks of RNAi are unclear.
It is therefore useful to consider RNAi technology and its possible impacts on the protection goals as a special issue.

Regards Birgit

Dear participants of the online forum,

Regarding  the point (b) “LMOs created through use of dsRNA techniques, engineered to produce dsRNA or dsRNA” and “LMOs containing RNAi”.
I suggest the following title:
“LMO possessing a transgene(s) capable to induce an RNA silencing response”
In fact, in several studies the organisms were not directly engineered to produce dsRNA molecules. In these cases the induction of gene silencing was due to the transgene rearrangement that produced an aberrant mRNA or a dsRNA that triggers the silencing of the target gene. An example is the approved event: Name: C-5 Code: ARS-PLMC5-6, a plum resistant to Plum pox virus (PPV). The plum was transformed with the sequence coding for the PPV coat protein (CP) gene. Among transgenic clones obtained, only the plum C5 was resistant to PPV. The resistance was the result of the silencing of the CP transgene. In particular, C5 is characterized by possessing multiple and rearranged CP gene copies, produces a low level of CP mRNA and does not accumulate a detectable amount of CP (Scorza et al. 1994; Scorza et al.  2001; Scorza et al. 2013).
References:
Scorza  et al. 1994. Transgenic plum (Prunus domestica) L.- express the Plum pox virus coat protein gene. Plant Cell Rep; 14:18-22
Scorza R, et al.  2001. Post-transcriptional gene silencing in Plum pox virus resistant transgenic European plum containing the plum pox Potyvirus coat protein gene. Trans Res; 1054:1-9
Scorza et al. 2013. Genetic engineering of Plum pox virus resistance:‘HoneySweet’ plum—from concept to product. Plant Cell Tiss. Organ. Cult. 115, 1–12.doi: 10.1007/s11240-013-0339-6;
For a recent review on biotechnological strategies and tools to obtain  Plum pox virus resistance see:
Ilardi & Tavazza 2015. Biotechnological strategies and tools for Plum pox virus resistance: trans-, intra-, cis-genesis, and beyond. Frontiers in Plant Science, 6.
Thank you.
Best regards
Vincenza Ilardi,
ITALY

Dear members of the forum,

Thank you Francisca for moderating this round of discussions. I also thank the opportunity to participate in this forum.

My comments will follow the structure proposed by Francisca and hope to contribute to the integration of this topic into the Roadmap. The first suggestion is to have a dedicated space in the Roadmap (Part I) and this could be in the format of a text box. In addition, it would be helpful to have it referred/linked to in different sections. For instance:

Step 1 (regarding characterization of the LMO) lines 541 item (i), line 554 (iv), line 559 (c), line 563 (d), and line 569 (e). These items could make reference to potential changes at the mirnome level (set of miRNA and other sRNAs), to that dsRNA and sRNA species should be considered gene products that could cause adverse effects, to the search of potential ORFs for miRNA genes and novel miRNA or dsRNA sequences. Also in Step 1 (regarding the potential adverse effects resulting from the interaction between the LMO and the likely potential receiving environment) lines 612 item (o), line 615 (p), line 618 (q), line 622 (r) and lines 624 (s). Suggested elements linking these items to LMOs or products containing dsRNA are related to the potential od dsRNA being dispersed by animals feeding in the LMO, dsRNA could be transferred without the dsRNA transgene being transferred, dsRNA can be transferred horizontally even without dsRNA transgene being transferred because dsRNA have the potential to produce cross-kingdom gene regulation.

In Step 2, line 667 item (b) and line 689 (g) and in Step 3, line 749 (f). Regarding the persistence and accumulation in the environment, a suggestion for a reference to dsRNA potential to produce heritable changes (through epigenetic transmission) that may result in persistent changes either within cells or entire tissues, and be heritable through reproduction in some animals and other organisms. In fact, the potential to produce secondary miRNA by enzymes called RNA-dependent-RNA polymerase should also be taken into account.

In Step 5, line 847 (b)(elements for consideration related to the risk management strategies and/or monitoring) regarding methods to detect and identify the LMO, and their specificity, sensitivity and reliability in the context of environmental monitoring. I is a suggestion to also make reference to future protocols for transgenic miRNA/dsRNA detection that might be available at the BCH.

Regarding the most relevant elements, I would like to suggest the following:
• Some LMOs are intended to produce new regulatory-RNA molecules, but these may also arise in other LMOs not intended to express them. Therefore, the suggested elements should be studied in all LMOs under approval.
• Sequence-independent and sequence-dependent risks. Sequence independent risks are related to the length of the dsRNA molecule and also to the amount of molecules the organism is exposed to. Larger molecules over 30 base-pairs in length can cause toxicity in mammalians. Also, larger amounts of dsRNA can saturate the RNAi machinery in the recipient’s cells and thus cause toxicity.
• Silencing effect is initiated, the effect may be inherited. The biochemistry of this process varies depending on the organism and remains an area of active research with many unknown aspects. Nevertheless, it is known for example that human cells can maintain the modifications necessary for TGS, creating actual or potential epigenetic inheritance within tissues and organisms. In some cases the dsRNA pathways induce RNA-dependent DNA methylation and chromatin changes (TGS) that persist through reproduction or cell division, and in other cases the cytoplasmic pathways remain active in descendants.
• dsRNAs are remarkably stable in the environment. A property perhaps overlooked based on the relative instability of single stranded species of RNA. Insects and worms that feed on plants that make dsRNA can take in the dsRNA through their digestive system, where it remains intact.
• dsRNA taken up is further amplified. It can cause a secondary reaction that leads to more and different dsRNAs (“secondary” dsRNAs) with unpredictable targets.

Therefore, the stability and transmissibility of dsRNAs suggest the potential for existence of exposure routes that are relevant to human and environmental risk assessments of genetically engineered/modified (GM) organisms.

These elements could be taken into consideration by adding a box after Step 2. In fact, I would also like to suggest for your consideration a potential dedicated piece in Part II.

Please find below the suggestion of a couple of articles that illustrate the elements I have mentioned. One of our papers also presents suggestions to improvements to risk assessments that might be useful to this discussion. Regarding the knowledge gaps, I would like to mention the report of the EFSA meeting on ‘Risk assessment considerations for RNAi-based GM plants’ in which they also present under topic 3.6.2.4. on page 26 the issues that were identified as knowledge gaps and where more research could be warranted.

Thank you again for the opportunity to contribute in this discussion and I am sorry for the long post.

Best regards,
Sarah

Suggested literature:
Heinemann JA, Agapito-Tenfen, SZ, Carman, JA. A comparative evaluation of the regulation of GM crops or products containing dsRNA and suggested improvements to risk assessments. Environ Int 55, 43-55, doi:10.1016/j.envint.2013.02.010 (2013).
Heinemann JA, Agapito-Tenfen SZ, Kurenbach B. Response to "A 28-day oral toxicity evaluation of small interfering RNAs and a long double-stranded RNA targeting vacuolar ATPase in mice.". Regul Toxicol Pharmacol. 2015 Apr;71(3):599-600. doi: 10.1016/j.yrtph.2015.02.003. Epub 2015 Feb 10.
EFSA Workshop Report at: http://www.efsa.europa.eu/sites/default/files/assets/705e.pdf.

Dear members of the forum

I believe that this topic requires at minimum two text boxes, one in the main text of the Roadmap and one in the Terms section, although I agree with Sarah Agapito that the best treatment requires a new entry in Part II of the Guidance.

There is a large proliferation of terminology in the area of dsRNA-mediated effects on gene expression which result in intra- and also sometimes inter-generational RNAi. This is effected by either dsRNA-directed RNA degradation, translation inhibition or de novo methylation changes to DNA.

Sarah Agapito has described where in Part I reference could be made for cases where DNA is used to create an LMO that then produces a form of dsRNA intended to cause RNAi or which may do so as a consequence of the engineering process (also mentioned by Vincenza Ilardi in his post). Sarah has also kindly described risk pathways that result from the transmission of the dsRNA or RNAi manifestation from the intended LMO by means separate from transfer of the initiating DNA.

I will extend Sarah’s treatment to suggest where in the text reference could be made when the RNA (and not DNA) is the source initiating the gene silencing effect. This would be the case where species of dsRNA are used for, eg, delivery across the ‘skin’ of a target organism using a pesticide formulation as a vector.

Beginning with the definition of modern biotechnology from Article 3 of the Protocol and definitions of “Transformation Event” (An LMO with a specific modification that is the result of the use of modern biotechnology according to Article 3 (i) (a) of the Protocol) and “Transgene” (Transgene – A nucleic acid sequence in an LMO that results from the application of modern biotechnology as described in Article 3 (i) (a) of the Protocol) in the Guidance, an LMO is created through the use of dsRNA molecules when it results in a modification because dsRNA is 1) a nucleic acid sequence that 2) in an LMO is 3) an application of modern biotechnology because it conforms to the definition of 4) Article 3 (i)(a) of the Protocol which captures any “application of:
a. In vitro nucleic acid techniques, including [but not restricted to] recombinant deoxyribonucleic acid (DNA) and direct injection of nucleic acid into cells or organelles”.
This modification may be heritable in the LMO or may be transmitted through various means to other organisms including non-target organisms.

A definition of dsRNA should be provided and this should include all forms (miRNA, long dsRNA etc) that are relevant to a assessment. A text box describing the many different names of dsRNA species with biological roles would help especially the novice risk assessor to recognize the relevant risk issues.

In addition, it would be helpful to have this topic referred/linked to in different sections.

Step 1
Reference could be made to sections:
(e), (j), (k), (q), (r) and (s).

Step 2
Reference could be made to sections:
(b), (d), (f-h).

Step 3
Reference could be made to sections:
(a)(i)(iv-vi), (d), (f).

Step 4
Reference could be made to sections:
all.

Step 5
Reference could be made to sections:
(b), (d), (h), (i).

Depending on how the text box is composed, references at all the above places (or indeed at places suggested by others) may not be necessary.

For references:
Baier, S.R., Nguyen, C., Xie, F., Wood, J.R. & Zempleni, J. MicroRNAs are absorbed in biologically meaningful amounts from nutritionally relevant doses of cow milk and affect gene expression in peripheral blood mononuclear cells, HEK-293 kidney cell cultures, and mouse livers. J. Nutr. 144, 1495-1500 (2014).

Demonstration of the potential diet-relevance of dsRNA.

Hanning, J. E., Saini, H. K., Murray, M. J., van Dongen, S., Davis, M. P. A., Barker, E. M., Ward, D. M., Scarpini, C. G., Enright, A. J., Pett, M. R. & Coleman, N. Lack of correlation between predicted and actual off-target effects of short-interfering RNAs targeting the human papillomavirus type 16 E7 oncogene. Br. J. Cancer 108, 450-460 (2013).

Demonstration that current bioinformatics tools are not sufficient to produce evidence that can exclude potential activity.

Heinemann, J. A., Agapito-Tenfen, S. Z. & Carman, J. A. A comparative evaluation of the regulation of GM crops or products containing dsRNA and suggested improvements to risk assessments. Environ Int 55, 43-55, doi:10.1016/j.envint.2013.02.010 (2013).

Provides an overview of risk assessment both environmental and with regard to human health.

Liu, S., da Cunha, A. P., Rezende, R. M., Cialic, R., Wei, Z., Bry, L., Comstock, L. E., Gandhi, R. & Weiner, H. L. The host shapes the gut microbiota via fecal microRNA. Cell Host & Microbe 19, 32-43, doi:10.1016/j.chom.2015.12.005 (2016).

Provides a basis for considering microbes as potential non-target organisms.

Lundgren, J. G. & Duan, J. J. RNAi-based insecticidal crops: potential effects on nontarget species. Biosci. 63, 657-665 (2013).

Specific risk-related information for dsRNA-based pesticides.

Lukaski, A. & Zielenkiewicz, P. In silico identification of plant miRNAs in mammalian breast milk exosomes - a small step forward? PLoS ONE 9, e99963 (2014).

Coupled with Baier et al (above), indicates that ingestion sourced dsRNAs may be packaged in diet-relevant ways. Relevant to mammals, not just human health.

Nunes, F. M. F., Aleixo, A. C., Barchuk, A. R., Bomtorin, A. D., Grozinger, C. M. & Simões, Z. L. P. Non-target effects of green fluorescent protein (GFP)-derived double-stranded RNA (dsRNA-GFP) used in honey bee RNA interference (RNAi) assays. Insects 4, 90-103 (2013).

Further evidence that bioinformatics tools do not yet provide sufficient evidence to exclude off-target effects.

For guidance:
FIFRA. RNAi Technology as a Pesticide: Program Formulation for Human Health and Ecological Risk Assessment. (United States Environmental Protection Agency, 2014).

With best regards,
Jack

Dear all,

Thank you for this opportunity to provide input into this discussion.  I particularly wish to thank Francisca for taking on the task of moderating.

My post is in response to the bibliography provided by Jack [#7595] and Sarah [#7594].  It is important to include the following references for a more complete view of this topic:

Petrick, J.S., et al, 2015. A 28-day oral toxicity evaluation of small interfering RNAs and a long double-stranded RNA targeting vacuolar ATPase in mice. Regul. Toxicol. Pharmocol. 71, 8-23.

Laubier, J., et al, 2015 No effect of an elevated miR-30b level in mouse milk on its level in pup tissues. RNA Biol. 12(1): 26-29  COI: 10.1080/15476286.2015.1017212. 

Title, A.C., et al, 2015 Uptake and function studies of maternal milk-derived microRNAs. J. Biol. Chem. 290(39): 23680-23691. 

Petrick et al (see above), was the target of the response cited in Sarah's comment [#7594].  As such the original work should also be cited.  Note further that the authors have responded to the Heinemann et al response.  See:  

Petrick, J.S., Moore, W.M., Heydens, W.F., Koch, M.S., Sherman, J.H., and Lemke, S.L. 2015. Authors’ response to letter to the editor by Heinemann et al. “Response to ‘A 28-Day oral toxicity evaluation of small interfering RNAs and a long double-stranded RNA targeting vacuolar ATPase in mice.’” 2015. Reg. Tox. Pharm. 71, 597-598.

With regard to inclusion of information on dsRNA in the guidance, we should simply recognize that the mechanism of action of many GMOs is based on dsRNA. 

Thanks and best regards to all,
Tom

In addition, Jay Petrick kindly provided me with a relevant reference to a review he published with some colleagues in 2013. 

Petrick, J.S., Brower-Toland, B., Jackson, A.L., and Kier, L.D. 2013. Safety assessment of food and feed from biotechnology-derived crops employing RNA-mediated gene regulation to achieve desired traits: A scientific review. Reg. Tox. Pharm. 66, 167-176.

Thanks,
Tom

Thanks for the opportunity to post in this forum.

We would like to give our support to the points addressed by Sarah Agapito (#7594) on the topic of “LMOs created through use of dsRNA techniques, engineered to produce dsRNA or dsRNA” and “LMOs containing RNAi”, especially when it comes to the points of :
• Apparent stability of  dsRNAs in the environment.
• Ability of dsRNA to be taken up and further amplified.

These points point to the potential of these RNAs to have a more unpredictable mode of action due to their apparent potential for survival in the environment.

Best regards.

I'd like to once again thank Francisca for moderating the forum, the Secretariat for the opportunity to participate and to all of my colleagues who have posted their thoughts and views.

I would like to share a link to a proceedings document resulting from a workshop considering problem formulation for plants with RNAi phenotypes: http://www.cera-gmc.org/files/cera/uploads/pub_08_2011.pdf.  This workshop was sponsored by a grant from USDA's Biosafety Risk Assessment Research Grants Program (#2010-33522-21796), and participants came to several consensus conclusions.  The most relevant for this discussion being that plants with RNAi traits can be assessed using the same paradigm as other genetically engineered (LMO) plants. 

For this reason, I would strongly support the views of Boet Glandorf (#7586), Martin Lema (#7580), and Yann Devos (#7610) that a section or box on this topic is not necessary or useful in the context of the Roadmap.  Ultimately, dsRNA is one of many potentially "novel" substances that might be produced by an LMO.  These should always be assessed with respect to their potential to impact the environment.  It is not necessary or useful for the Roadmap to attempt to address every conceivable novel substance, how it might interact with the environment, how it "works" to elicit a phenotype in the plant etc.  In that case, we might envision a section on the roadmap for globular proteins, structural proteins, membrane bound proteins, transmembrane proteins, enzymes, transcription factors, glycoproteins, lipids, phospholipids, small molecules etc.  This will only serve the purpose of making the Roadmap dated, as new techniques are emerging much faster than work is progressing on this document.

Setting aside for the moment a debate on the validity and relevance of claims being made about  dsRNA, the principles for assessing risk are identical to other novel substances.  Some participants in the forum and the AHTEG seem to feel that the way to improve the Roadmap is to delve into ever more specific and intrusive detail about every conceivable case specific risk assessment.  This is, in my view, counterproductive and contrary to the intended purpose.