5. Objective #5. The potential for recombination between plant viruses and plant-encoded noncapsid viral genes (e.g. replicase), especially for those viruses in supergroup B (carmovirus, tombusvirus, luteovirus, sobemovirus). Such studies should identify recombination potentials and, if demonstrated, define frequencies and effect on symptom expression. Comparisons with recombination frequencies between naturally occurring viral sequences are encouraged.

a. Reviewers' comments

Most of the reviewers agreed that this objective is an important issue. However, several comments questioned the scope of the potential problem and the consequent need for such a targeted area for coverage in a program designed for evaluation of risk. As mentioned previously, the conflict between advancing basic science and targeting funds to address potential regulatory issues seemed apparent through some of these comments.

• "Objectives five and six, which concern risks from transgenic plants with genes from plant viruses, should remain a high priority. Discussion of these risks is now largely based on educated guesses rather than data."

• The 'potential for recombination between plant viruses and plant-encoded, noncapsid viral genes' is a researchable and quantifiable topic. There is obvious risk posed by such occurrences to plant crops."

• "This objective is just as coherent scientifically as Objective 4, but the nature of the biotechnological risk is less obvious. It should be stated."

• "Both (objectives 5 and 6) address extremely low-frequency events, which are going to be very hard to quantify (i.e., without huge confidence intervals). A more useful approach would involve the design of monitoring schemes that could detect undesirable recombination events or host shifts in nature."

• "The emphasis on viruses in super group B has been unchanged over the three year period. The emphasis on luteoviruses is understandable, as this family contains what are arguably the most important plant viral pathogens, and transformation of important crops (potatoes, cereals) has been accomplished. However, the other families mentioned contain viruses which are not important pathogens, except perhaps in rare cases, and hence are not likely to attract commercial development and/or widespread development. On the other hand, viruses in, for example, the cucumovirus and potyvirus families are widespread and damaging. A number of crops have already been transformed with genes from these viruses. Perhaps consideration should be given to redirecting the emphasis to these and/or other more relevant viruses. The 1996 version does additionally encourage comparison with recombination frequencies between 'naturally occurring viral sequences'. This is a highly desirable addition which should also be targeted to the viruses most likely to be the source of genes for plant transformation."

• "Risk in this area seems to be low. However, studies should be encouraged which characterize the underlying mechanisms by which recombination of viral proteins, including the replicases, occur in nature. The gemini viruses which attack tomatoes and peppers in Mexico and the Middle East offer an attractive model system in which to investigate the probability of such recombination. Further, the deployment of defective replicases in plants would appear to minimize the probabilities of recombination between plants and viral species."

• "This is a fairly specific topic of both theoretical and practical importance. In fact, it is so important and obvious that I suspect there already are numerous research projects funded and in progress through USDA National Research Initiative grants and NSF supported research. In other words, this basic biology area may be less appropriate for funding in a risk assessment context than some of the other areas, which are less likely to be included in other programs. If Supergroup B virus genes (or some other subset) are being neglected in this context, perhaps this is the right program for such funding after all."

• "I thought that areas 5 and 6 were narrow and intimately interrelated. They should be combined."

• "The call for proposals under research area 5 is narrow. I lack the perspective to judge the appropriateness of targeting interest in viral supergroup B."

• One reviewer commented that, as with objective #4, there has been a substantial amount of work that has "amply addressed" the issues raised in objective #5. This reviewer went on to note that "The occurrence of transgenic RNA-virus RNA recombination is a given and cannot be prevented under field conditions. The important questions are: what will be the outcome of such a recombination and will the outcome be significant when viewed against the background of naturally-occurring recombination in plants infected with two or more viruses? In addition, are the recombination events significant in the environmental context and what is the risk compared to the potential benefit? Most recombinants can be expected to be less virulent than either of the parental viruses, based on the properties of the bulk of the recombinant viruses produced in the laboratory to date."

This reviewer reflected that "Changing agricultural practices and modifications of natural environments can be expected to bring existing viruses into contact with new potential hosts (as they always have) and provide new opportunities for recombination in dual and multiple virus infections. No planned experiment is likely to provide guidance as to the likelihood of significant effects from the transgenic expression of a virally-derived RNA molecule in the context of background recombination events."

• One reviewer offered further revisions of the topical parameters for proposals submitted to this Program area: "Grants should be given for research to investigate the role in recombination of: (1) noncoding as well as coding sequences; (2) nonfunctional protein mutants as well as wild type proteins; (3) consider first the replicase/polymerases since plants with such sequences are likely to reach market first; studies of other proteins should follow as development progresses to near-market situations. Studies should encourage multi-cropping situations as well as multiple crops in succession in the same field or area. Studies of the effects of recombination under nonselective conditions in transgenic plants are necessary. It is also important to include in all tests the impacts of recombination in transgenic plants by single as well as multiple viruses, and if possible, the impacts of insect transmission of the recombinant on the derivation of novel viruses and virus strains. While there is also a need to consider the effects of transgenes on the virulence of surviving virus in a resistant field and the durability of a transgene, this is an important question for the seed company more than the USDA grants program."

b. Expert Panel's comments

See Expert Panel comments under objective 3.

6. Objective #6. Changes in viral host ranges or the types of viral vectors as a result of the use of transgenic plants expressing viral genes.

a. Reviewers' comments

Many of the comments about this objective were similar to those about objective 5. Overall, the comments were favorable, at least in terms of the scientific importance of the questions to be addressed. Several reviewers noted the similarities between objectives five and six and made suggestions either to combine the two areas or apply similar criteria. Examples of this observation include the following:

• "Research area six is related to area five and addresses the potential for changes in viral host ranges or viral vectors as a result of the expression of viral genes in transgenic plants. This is an important area, particularly with the recognition of the pleiotropic effects of viral genes. The information concerning this area has not changed over the three year period. Once again, perhaps solicitations based upon the criteria discussed above for area five should be considered."

• "It is conceded by all that recombination between a virus and a transgene can occur, but the main consideration is whether the outcome is bad. To date, there has been no suggestion that new, and more deleterious strains or viruses have arisen. The same argument can be made with respect to the broadening of the host range. I feel that the only grants which should be considered in the plant virus area in the future should deal with large scale field trials to see if there is credence to the genesis of bad viruses or expanded host ranges. This may well be done for us by the entry of virus resistant plants into commerce."

Additional comments about objective #6 include:

• "The statement of this point is a bit too broad to serve as a useful objective. Its authors must certainly have some idea what sort of risks they have in mind, and it would be helpful to mention them here."

• While acknowledging the importance of this general area of research one reviewer wrote, "The wording of this topic is somewhat ambiguous. What is meant by 'types of viral vectors?' Is it changes in types of vectors? Intent is not clear."

• "Both (objectives 5 and 6) address extremely low-frequency events, which are going to be very hard to quantify (i.e., without huge confidence intervals). A more useful approach would involve the design of monitoring schemes that could detect undesirable recombination events or host shifts in nature."

• "This research area is broader than area 5, and seems an appropriate area of concern for environmental risk assessment."

• "As to objective #6, the risk of transcapsidation is extremely low and not worthy of attention. The arguments provided for objective #5 apply even more strongly here. It is even less likely that RNA-RNA recombination would present a risk to agriculture or the environment."

• "...such studies should include studies of plants that are under inoculum pressure by multiple viruses since such conditions may influence the acquisition of virus by an insect, or make it possible for some insects to transmit that did not normally transmit in the case of single infection. All such studies should be well controlled with non-transgenic plants that are similarly treated. The impact of resistance on selection, as well as the degree of resistance, i.e., near-immune conditions versus highly or moderately resistant plant lines should be included in studies. It is hoped that studies are not carried out only with members of the Solanaceae, but also with crops such as corn and soybeans where the pest problems are different and the acreages of planting likely to be much greater than for horticultural crops.

  Furthermore, the contents of alkaloids and other secondary compounds are different with different crops and may affect insect feeding and transmission of disease agents."

  b. Expert Panel's comments

  See Expert Panel comments under objective 3.

  ---

  7. Objective #7. The potential for nontarget effects of introduced plant-defense compounds expressed in genetically modified plant associated microorganisms (e.g., compounds in phyllosphere or rhizosphere inhabiting bacteria) or in plants (e.g., Bacillus thuringiensis delta-endotoxin), especially in regard to persistence of the organisms and material in the environment.

  a. Reviewers' comments

  The general consensus was favorable towards this objective albeit, as one reviewer said, "Perhaps some fine tuning is in order." Only a few of the reviewers offered specific comments directed towards this Program area:

  • "This is clearly stated and well circumscribed... although it appears to assume that persistence of engineered organisms in the environment is inherently a risk. The use of Bacillus thuringiensis as an example of non-target effects overlooks the widely accepted view that the greatest risk associated with the delta-endotoxin is that their injudicious use will result in rapid insect resistance and hence loss of an otherwise low-risk strategy of pest control to higher risk chemical alternatives."

  • "Studies regarding the impacts of Bt on the soil ecosystem are desperately needed-- but, the impacts will depend on soil properties, and therefore, this effort should make sure impacts are evaluated in a wide variety of soil types."

  • "This is a specific, important area for research in the context of risk assessment, and intent is clear."

  • "Characterization of non-target pesticidal effects of genetically modified microbes or plants is timely and appropriate. This is justified by well-documented declines in valued non-target populations following the widespread adoption of chemical pesticides."

  • This same reviewer offered this overview of objectives 4-7: "Taken collectively, research areas 4-7 call for sophisticated experimental designs quantifying risks associated with narrowly defined impact pathways. This portion of the Program's call for proposals underlines the notion that knowledge of risk pathways for genetically modified plants is much more advanced than for other sectors of agricultural biotechnology. In contrast, risk assessment for genetically modified microbes, arthropods, and aquatic organisms is in an exploratory stage."

  • The following refinements were offered by another reviewer who wrote: "There are persistent and lingering questions related to the impacts of transgene-derived proteins on non-target organisms in the environment, including the rhizosphere and soils that contain crop residues, and leaf tissues. One of the most nagging questions is related to the survival of genes that encode resistance to antibiotics, since such genes are generally included in transformation/selection protocols. In the selection of grants for funding at least one should be for the study of survival of gene sequences in animal digestive systems. Although many of us consider the likelihood of horizontal transfer between organisms to be extremely unlikely (even fantasy) the public has a concern that has not yet been carefully addressed." The reviewer went on to suggest that, "the proteins that are selected for study should include those that are likely to be included in crops, including PR proteins and phytoallexins, as well as proteins that control specific insects and microbes."

  b. Expert Panel's comments

  The Expert Panel regarded this as an important objective because of its focus on microbial issues. They concluded that the question of antibiotic resistance should be given a high priority and suggested that it is so important that the BRARGP might consider upgrading it to the status of an objective (rather than one issue within objective 7).

  ---

  8. Objective #8. "Identification of genes which can confer additional pathogenicity to animal pathogens."

  a. Reviewers' comments

  Of all the objectives, this one was the focus of some of the more strident comments from the reviewers. As seen in the comments listed below, the preponderance of the invective is directed at a perceived lack of clarity in the Program description.

  • One reviewer offered this explanation for the lack of specificity of this Program area: "Research area 8, addressing the possibility that introduced genes could recombine with or otherwise increase the pathogenicity of animal pathogens, corresponds to the plant-oriented research areas 5 and 6. The broad generality of research area 8, compared to the specificity of areas 5 and 6, emphasizes how much less we know about the risk pathways and animal pathogens at issue."

  • Another reviewer noted: "It is unclear if genes of the host or the pathogen are of interest. If the former, it might be clearer to use wording such as 'increased susceptibility to pathogens,' Perhaps an example would increase clarity."

  • Objective #8 was described by one reviewer as "...very concise, to the point that some individuals did not understand the nature of the area being proposed. If included in subsequent announcements, the nature of the 'animal pathogens' should be clarified. For example, are insects 'animals?' Or, are some other animals pathogenic? Or, perhaps the announcement meant pathogens of animals?"

  However, the bulk of the commentary on this objective was less magnanimous.

  • "It is unclear what types of projects are anticipated to involve genes that can confer additional pathogenicity to animal pathogens, although certain of those described above might be placed in this category." This reviewer went on to propose the following revisions to this objective: "It goes without saying that such studies should involve livestock animals and poultry, etc., rather than rodents and other "model" animals that are not part of the (U.S.) diet. Other genes that might impact pathogenicity of animal pathogens might include proteins that affect freezing tolerance, or that encode proteases, toxins and other proteins that might impact pathogenicity of microorganisms. However, because the likelihood that any gene recombinations or other types of gene transfer is extremely low (fantasy) it will be important to fund only those proposals that are of the highest quality by the most suited of research teams. The team must include plant biologists, microbiologists, animal nutritionists, and population biologists/geneticists, and should be funded for 3-5 years, with good preliminary results available to the research and regulatory agencies within 6-18 months. If such a project is to be funded, it should be done correctly so that the results will leave few doubts of their significance and applicability to biotechnology (i.e., no model systems)."

  • "Among the more specific research areas outlined in points 4-9, point 8 is particularly weak. As written, it is not clear what value this point has to either basic science or to risk assessment and I would recommend eliminating it altogether."

  • "Sounds like a fishing expedition here; what's the point? Our pathogens aren't pathogenic enough? this like Objective 6. One suspects that it is code for something more specific, that should be spelled out here."

  • "I do not feel this is worth much effort -- of course, genes that enhance virulence exist -- but so what? I do not see where this research effort will lead us to what is practically useful."

  • "Research area eight is about as vague and unfocused as possible; any gene, any pathogen, and animal. At least the previous version identified organisms of interest to APHIS, again presumably because these pathogens are currently being engineered for vaccine production or other uses. There seems to be less outside input into potential risks in animals systems as opposed to plant systems. Considerable thought needs to be given to what types of research proposals are needed in this area."

  • "Do such genes exist? Have there been any applications in this group?"

  • "I do not understand why objective eight...is a priority. This may reflect poor understanding on my part -- I have little scientific training with warm blooded vertebrates. Nevertheless, at the very least, more context for this objective would be helpful!"

  • "My view of virulence and pathogenicity in both animals and plants is that the fields are too immature to make any sensible projections of practical work ('identification of genes which can confer additional pathogenicity...') worthwhile."

  b. Expert Panel's comments

  The Expert Panel considered the criticisms of objective 8 to be valid and concluded that it should not be continued in its current form.

  ---

  9. Objective #9. Environmental risk analysis of large scale deployment of genetically engineered organisms; especially commercial uses of such organisms with special reference to consideration that may not be revealed through small scale evaluations and tests."

  a. Reviewers' comments

  Conceptually, reviews of this Program area were, for the most part, favorable. With the exception of a few comments about the need for clarifying examples, the comments were supportive.

  • "...I would like to see an emphasis on experiments in Category 9. I feel we need this information, which is the ultimate in risk assessment."

  • "Research targeted under area 9 is absolutely essential for fostering environmentally safe commercialization of agricultural biotechnology. A salient example of an important problem is scaling up of production of crops expressing Bt endotoxin from the experimental stage (where plant pests encounter this pesticide only rarely) to the commercial scale (where widespread use would cause sufficient selective pressure for resistance that Bt's future as an effective pesticide would prove limited). Other risk mechanisms involving other products of agricultural biotechnology also might become important only at the landscape scale. Results of studies supported under this research area could prove critically important for strategic development of agricultural biotechnology, in terms of realizing its potential while minimizing negative environmental impacts."

  Some reviewers offered suggestions for clarification of the objective:

  • "Research area nine was introduced in the solicitation for the 1995 Program. It would have helped my understanding if a 'for example' was provided, but perhaps the originators were leaving it to the applicants to make the case. If so, hopefully there have been submissions that may help the Program administrators to phrase the statement in such a way as to be able to hone in on the subject."

  • An additional comment about the need for more clarity was offered by the reviewer who wrote, "The objective is very vague. The objective asks for an environmental risk analysis of large scale deployment of genetically engineered organisms. It seems obvious that this objective was written with some circumstance in mind; it would have been better to share that circumstance as an example for potential respondents."

  • One reviewer, in support of the objective, offered the following suggestions about criteria for research design and subsequent proposal review: "Large-scale risk assessment should assume both the probability that the gene will migrate into domestic and weedy relatives and, secondly, impact when the gene or genes are transmitted to domestic or weedy relatives. The function of the gene can then be considered as the primary criterion of risk assessment. A gene that modifies seed oil characteristics will likely have different impact on weedy and domestic relatives than herbicide resistance, insect resistance, male sterility, or disease resistance. Large-scale deployment of single transgene resistance to a single insect or disease pest will alter the occurrence of other disease and/or insect pathogens that occur. Market penetrations may change the pattern of diseases or insects if safeguards are not taken. Normally, competition in the marketplace is the preferred safeguard since alternative sources of resistance, both transgenic and nontransgenic, will arise for the most significant pests. Consideration should be given to evaluation of source of transgenes and the genetic background of the plants in which the transgene is deployed. Genetic variation for both factors will minimize risk of new insect and/or disease pest becoming dominant, and also minimize development of resistance a transgene (i.e., Bt loss of effectiveness).

    Consideration might also be given to determining how long negative effects might persist once a transgene is no longer used commercially. Would transgenes that have migrated into domesticated and weedy relatives disappear rapidly after large scale use is curtailed?

    The use of appropriate statistical models, experimental designs, and use of appropriate statistical consultants are very important criteria that should be used to evaluate all proposals. Appropriate statistical treatment of data will be necessary to predict the impact of one study for similar transgenes or species."

  There was, however, some ambivalence expressed about this objective as exemplified by the following comments:

  • "I am a little concerned about the purpose of the 9th area of research to be supported. The risks of large scale releases should and could be addressed in the first two areas of research mentioned in the Announcement. While I support obtaining data from large scale field releases, mentioning this area of research as a separate topic may serve to debase the small scale studies that have been done, and unduly heighten the concern of large scale releases."

  • Consistent with the opinion expressed above was the conclusion that, "All risk analysis should be aimed at large-scale commercial use of transgenic organisms. Hence, this effort does not need an extra item. I suppose the point here is that data from small, short-term field trials cannot address some concerns that emerge when we go ahead on commercialization. In that event, certainly we need to worry about those phenomena that emerge only after commercialization (such as the evolution of insect resistance to Bt endotoxin)."

  • Of interest are the comments of the reviewer who suggested that this Program area be reduced, not because of the lack of a need for information derived from large-scale uses, i.e., introduction of transgenic plants into large acreage commercial applications, but because "It would appear to me that this effort will now become a component of normal agricultural extension activities at the state level. The USDA should implement mechanisms to insure the assignment of this responsibility, thereby decreasing item 9..."

  • One criticism of this objective was voiced within the context of limited funds: "Although this has an admirable sound to it, such projects are scarcely feasible with the restricted program budget."

  • In questioning the ability of this Program to achieve its stated goals one reviewer offered a valuable suggestion: "How realistic it will be to compare small and large scale releases of GMOs in projects funded by this program is doubtful since nothing that could be attempted here could match the large commercial release covering millions of acres. In my view, this is best done through comparing the records of previous releases of conventional organisms and GMOs."

  • The need for collaboration with commercial interests who are active in this areas was reinforced by the reviewer who suggested "...perhaps the most appropriate use of the resources is to work with the providers of the materials, monitoring populations of pests and pathogens, as well as non-target organisms. It is also important to provide support not only for study of macro-organisms (animals, plants, etc.) but also for modified microorganisms. It is anticipated that modified soil and foliar organisms, including Rhizobitan spp., micorhizal fungi, and other beneficial organisms, as well as microbes that provide biological control of pests and pathogens will soon be ready for the market. Given the history of the early Lindow experiments, those of Brill, Cook, and numerous private companies, and the public perception of microbes as involved in disease, such studies are important to the industry associated with agricultural biotechnology."

  b. Expert Panel's comments

  The Expert Panel recognized that because of limitations in funding and amount of resources required to conduct such large scale studies as called for in this objective, more creative approaches are needed to accomplish its worthy goals. The Expert Panel suggested that efforts be made to create effective lines of communication leading to collaboration between commercial enterprises, who have the resources to conduct such studies, and scientists who could design experiments that would successfully examine the risk. The Expert Panel stressed that the endpoints/consequences (e.g., effects of gene transfer, impact on non-target organisms, impact on the ecology, changes in pathogenicity and host range, creation of new pathogens, resistance management, etc.) to be evaluated in such studies be well-defined and focused on the overriding programmatic goal of identification of gaps in knowledge related to regulation of risk.

  ---