Dear experts
As subchair of the Stacked Genes working group I like to ask for your comments and additions to the draft additional guidance on Stacked genes in order to work out an improved version. In addition I would appreciate if you  could indicate if you think there should be additional or new references included.
Thank you very much for the support
regards
Beatrix Tappeser

Dear forum participants,

In the framework of the specific guidance on Stacked genes, it should
maybe be useful to explicit the steps of the RA that require
experimental data to allow to draw scientifically sound conclusions. (
This remark could actually be valuable for the general roadmap also ) .

Indeed, the assessment is largely based on comparisons with and
potential interactions between the single parental transformation
events.

It should also be taken into account that the parental lines can already
themselves be stacked genes of a lower order.

The question of the comparators is a particularly difficult one in this
case; it should be good to tackle it somehow. 

With best regards.


Lucette Flandroy



Disclaimer : http://www.health.belgium.be/eportal/disclaimer/index.htm

Dear forum participants.

I would like to bring into your consideration my remarks on the risk assessment of LMO with stacked events.

The first consideration is if Parties should do any risk assessment at all for stacked events. The rationale is that stacks combine the properties of the single events and also the risks and therefore the risk assessment would be just the simple sum of the two (or more) prior risk assessments. Conventional crossing can indeed lead to new unanticipated traits, but they would be observed during the stringent agronomic selection of the stacked lines. Moreover, compositional studies are also regularly made for all new plant cultivars, transgenic or not. If the crossing results in a normal plant with normal composition and with the added new characteristics conferred by the stacked transgenes, no new risk assessment would be necessary. Moreover, natural crossings will happen anyway in the fields, both unintentional and deliberately. This approach should be considered by the Parties and explicitly declared as a possible decision in the present guidelines.

Going to the ISSUES TO BE CONSIDERED IN THE RISK ASSESSMENT, first item (Assessment of sequence characteristics at the insertion sites and genotypic stability), in the rationale (that I think would be equivalent to the problem formulation), it is stated: “… the combination of transgenic traits via cross breeding may further change the molecular characteristics of the inserted genes/gene fragments at the insertion site and/or influence the regulation of the expression of the transgenes”. In my understanding, this is a highly implausible issue. Genes or controlling regions are very seldom changed because of natural crossing. Even if that happens, it would most possibly lead to phenotypic changes readily detected in the agronomic evaluation of the many crossings needed to choose the elite stacked plant. Therefore, we can safely consider this risk as negligible.

Following the same paragraph it is further stated: “In addition, changes to the molecular characteristics may influence the ability to detect the LMO, which may be needed in the context of risk management measures” The above considerations are also very speculative. Anyway, PCR experiments, DNA blots and other simple experiments are usually performed just to show that the stacked genes are there were they should be and that they are detectable by the regular techniques.

Following the first citations, the AHTEG texts says: “The reappraisal of the molecular sequence at the insertion sites, and the intactness of the transgenes may be confirmative to the molecular characteristics of the parental LMOs, but may also be a basis for assessing any intended or unintended possibly adverse effects on the conservation and sustainable use of biological diversity in the likely potential receiving environment”. Reappraisal of molecular sequences at the insertion sites are definitely not a basis to foreseen any intended or unintended possible adverse effect, as far as I know. The literature does not support such claim.

In “Assessment of potential interactions between combined events and the resulting phenotypic effects”, the Rationale says: “The combination of two or more TraEvs resulting in a StaEv may influence the expression level of each of the transgenes and there may be interaction between the genes and the expressed products of the different transgenes”. The first hypothesis is possible if promoters for both transgenes are the same and if both are inserted in the nuclear genome or both in chloroplasts. Otherwise it has no support from experimental data. It is important to keep in mind that one of the basic evaluations for stacks regularly done by the manufacturer is the quantification of expression levels for the different transgenes in the relevant plant tissues. On the other hand, most of the gene products expressed by transgenes do not interact with each other simply because they are from different metabolic pathways and from different organisms. Nevertheless, a theoretical consideration should be enough to discard this second hypothesis or to points toward the need for further evaluation. As for a supposed direct transgene interaction, as far as I know it would be very unlikely if not just impossible.

Just below the above commented text, it is said: “In addition, the stacked transgenes may alter the expression of endogenous genes”. This is highly implausible. If this happens, it should have been noticed in the single event(s) by careful scrutinizing all the data that is needed to get the commercial approval of the single parental events. It makes no sense to further insist on that. The assessment suggestions that follow this last hypothesis can be, therefore reduced to the ones I considered before.

As for the item “Assessment of combinatorial and cumulative effects of stacked event LMOs on the conservation and sustainable use of biological diversity in the likely potential receiving environment, taking also into account potential adverse effects to human health”, as I stressed above, most stacks will be just the expected regular crossing and therefore the new stacked plants will have just the combined risks of both parental lines. If protein interactions do occur, they may (but not necessarily will) prompt the regulator to ask for some new experimental data. In my opinion, the example given in the text of a stack with two or more insecticidal proteins does not immediately support the claim that a whole set of environmental assessment endpoints should be evaluated. The impact on non-target insects, for example, would certainly be broader than that of the individual events, but would most possibly be the sum of these impacts, with no extraordinary unanticipated impact.

Finally, in item “Development of specific methods for distinguishing the combined transgenes in a stacked event from the parental LMOs” it is said that “the methods used to detect the transgene in the parental lines may not be sensitive or specific enough to differentiate between single parental transformation events and the same event being part of a stacked event. A special problem may arise particularly in the cases where the StaEv contains multiple transgenes with similar DNA sequences. Therefore, the detection of each and all individual transgenes in a StaEv may become a challenge and need special consideration”. As far as single grains (or any plant tissue) are available, as they usually are for the risk management, the statement is wrong: even if the construction sequences are very similar (which will be seldom true), the flanking sequences are different, since the insertion sites are random. It may be a challenge in the future, however, when techniques for site-specific insertions reach the market.

In conclusion, I think that the problem formulation for stacks is wrong, being too much speculative. When too many speculative hypotheses are included in the problem formulation, the rest of the risk assessment will be very hard work and very expensive. The precautionary approach does not ask for that.

The fundamental paper on this subject is, in my opinion, De Schrijver A, Devos Y, Van den Bulcke M, Cadot P, De Loose M, Reheul D, Sneyers M (2006) Risk assessment of GM stacked events obtained from crosses between. Trends in Food Science and Technology, 1-9. This paper is cited on the corresponding literature sub-page, but its suggestions were not followed.

Kindly,
Paulo Andrade

Dear Paulo,
Replying on your first paragraph, the rationale for stack genes goes beyond mere combination of the properties of the single event and therefore the risk assessment should not be just the simple sum of the two (or more) prior risk assessments.
There could be genes interaction, or the stacks interaction with the recipient genomes to enhance the biosynthetic capacity of host cells (this is particularly true in bacteria, to reply your fifth paragraph), or the stacks products interaction may be significant. Genes (including stacks) can act in trans or cis. Hence, the stack genes working group should take this into consideration and all these warrant RA for stack genes LMO. While acknowledging the risks associated with stack genes, it is anticipated that the draft is not, in my opinion, out of the scope of precautionary principle.