Ethics, Law and Science of Using New Genetic Technology in Medicine and Agriculture

Darryl R. J. Macer, Ph.D. Eubios Ethics Institute 1990

Copyright1990, Darryl R. J. Macer. All commercial rights reserved. This publication may be reproduced for limited educational or academic use, however please enquire with the author.

9. Regulations for GMOs

pp.152-175 in Shaping Genes: Ethics, Law and Science of Using New Genetic Technology in Medicine and Agriculture, D.R.J. Macer (Eubios Ethics Institute, 1990).
It should be standard practice that all proposals to release new organisms into the environment, in every country, should be considered by some independent regulatory authority. The factors and form of the regulations are a matter of debate, and different features are found in different reports on the subject. After consideration of these we are in a better position to judge which factors should be included in regulations.

Some developed nations have regulations, but rely primarily on an expert advisory committee (Australia and the U.K.). The USA has elaborate regulations (Kingsbury 1988, OTA 1988b), and some European countries have no regulations (Ager 1988). There are some scientists who feel that the work is being over regulated because the fears are overstated (Davis 1989). While larger nations can enact and staff their own release program, it would be very costly for developing nations. Small and/or developing nations may need the aid of an international review or biosafety committee which could act on the request of the national leaders.

The actual process of genetic engineering is allowed to be pursued, subject to containment procedures, without any general limits. This idea is supported by some recent reports on the subject (OECD 1986, OSTP 1986, LRCV 1989). There are exceptions, such as work on particular pathogenic organisms, and on offensive biological weapons research. The exceptions include pathogenic organisms that have not spread to the geographic region.

In Britain, the proposing institution must complete a preliminary risk assessment themselves, from a safety committee (HSE 1989). By following a standard procedure it makes the process quicker to review, and decisions will be made within 90 days, and in some cases, such as with a GMO with a previous release, a considerably shorter time. The procedures have many similarities, but as will be evident by the end of this chapter, different countries do have different policies and procedures.

Viewing the Product Only

There have been various definitions of GMOs. One definition from the US Office of Science and Technology Policy uses the term intergeneric organism for GMO. "Intergeneric organisms are those organisms deliberately formed to contain an intergeneric combination of genetic material. Excluded are organisms that have resulted from the addition of intergeneric materials that are well-characterised and contain only non-coding regulatory regions, such as operators, promoters, origins of replication, terminators and ribosome binding regions" (OSTP 1986). This is representative of earlier approaches at defining the type of organisms that are required to be notified.

In the USA the Environmental Protection Agency (EPA) drafted new regulations in mid 1989 which do not distinguish whether the new organism is genetically modified but focus on the properties of any organism. In September, 1989, the US National Academy of Science with the National Academies of Engineering and Medicine, which form the National Research Council, published a report which stressed two themes (NAS 1989). There is no conceptual difference between altering an organism by classical breeding techniques or by gene splicing and regulators should evaluate field tests of GMOs on the basis of the potential hazard of the product itself, rather than the molecular techniques by which they are made. This reaffirmed the existing processes, and states that the applications can be evaluated scientifically. They argue that a broad definition of GMOs would encourage the use of genetic engineering techniques, over the traditional breeding techniques. If only organisms produced by genetic engineering are reviewed, it will discourage researchers to use the new techniques. The Ecological Society of America Report (Tiedje et al. 1989) also considers that the phenotype of a transgenic organism, not the process used to produce it, is the primary consideration for regulation.

This is in contrast to the U.K. Royal Commission which considered that GMOs can not be treated simply as products (HMG 1989b). The Ecological Society of America report does note that because many novel genetic combinations can be achieved only by molecular and cellular techniques, products of these techniques might be subjected to greater scrutiny than the products of traditional techniques. Organisms with novel combinations of traits are considered more likely to play novel ecological roles than other organisms, on average. It is important to note that phenotypic changes in plants have been observed after protoplast tissue culture, which may be because of "cryptic" transposons. This "natural" gene change has produced some variants, so designed single gene changes are not especially novel events (Yang 1988).

The United States Recombinant DNA advisory board (RAC) recently redefined the term "deliberate release" as "the planned introduction of recombinant DNA-containing microorganisms, plants, or animals into the environment". The criteria adopted meant that in large scale fermentation procedures using genetically-modified bacteria, companies only need take the same steps as applicable to an unmodified organism. This is because these bacteria have been found to be safe to handle, and the judgement is based on past experimental experience. If the engineered organism duplicates the phenotype (appearance and function) and community relationships of naturally occurring organisms or those produced with conventional methods; if the novel genetic information will not be transferred to any other organism; and if the introduced genes will have no adverse effects on the environment (OTA 1988b), then it will require little review. Also in the lower risk category would be organisms produced by recombinant DNA techniques that are functionally identical to organisms that could be produced by other methods (e.g. mutagenesis and selection) which are not now subject to review, though possibly there should be some review of these organisms also. Organisms that contain no genetic material derived from a potential pathogen, or organisms identical in function to those already approved are also subject to less review.

There have been various definitions of GMOs, and the technology is changing. In most cases organisms made by conventional breeding techniques are excluded, even though various reports (NAS 1989, Tiedje et al. 1989) consider that recombinant DNA GMOs should be considered the same as those made by traditional techniques. Both old and new technology can change. What is the most important concern is the product, however it was made. The different attributes of the product can be considered as a part of a broad risk evaluation.

Exemptions for GMOs with gene deletions?

The RAC has also exempted from review virtually all deliberate release experiments involving gene deletions within organisms. This exemption has been extended to experiments involving single nucleotide changes and gene rearrangements within a given species of microorganism (bacteria, viruses and fungi). This is because these alterations all occur naturally, and thus have already occurred in the ecosystem. If they were to present a major hazard, then we would have already seen it. This exemption for microorganisms has not been extended to experiments on plants and animals, even though microorganisms, because of their rapid reproduction and ease of dispersal, have more uncertainties attached. This means that in those cases researchers must still apply individually to the RAC for permission to conduct the experiment. This does seem a useful precaution to retain, as it requires further experience to test the possible side affects, though deletions should generally be safe.

The Biotechnology Science Coordinating Committee (BSCC), is trying to coordinate USA policy. They have issued a list of recommended exemptions (Fox 1990). It includes many organisms that are GMOs, and is a step backwards. A broad category exemption means that potential hazardous experiments may miss being reviewed. It may also lead breeders to use exempted techniques, even if they are not the best technique, to avoid regulatory control.

Another exception is the testing of new vaccines which, because it is of high medical relevance, is excluded from compulsory review in the USA. Also the Commission of Enquiry in West Germany (GB 1987), which recommended a moratorium on the release of viruses, considered the possibility of exceptions for vaccines. The benefits and risks are balanced in different countries by different people. The German report also recommended a 5 year ban on the release of genetically engineered microorganisms containing foreign genes. There have been similar proposals considered in the European Parliament. However, the approach of the Royal Commission in Britain (HMG 1989b) has been more cautious. They argue that because of ignorance we should not immediately categorise any type of release as sufficiently free of risk not to require individual scrutiny by the Committee. Each stage of the experiment should be subject to approval and licensing.

The Ecological Society of America would not recommend the complete exemption of specific organisms or traits from regulatory oversight (Tiedje et al. 1989). There is a need to minimise unnecessary regulatory burdens, but there are other ways to do this. There are some deletions that would be potentially more harmful, such as the deletion of regulatory sequences. However, the U.S. guidelines (OSTP 1986) may exempt alterations in regulatory sequences. Also once in the environment, there is a low but distinct chance that genes could be interchanged between different organisms. Recent studies that have shown that DNA can be transferred between yeast and bacteria suggests that other eucaryotes including plants and animals may be able to transfer DNA to bacteria, and bacteria could mediate a low level of horizontal gene transfer in eucaryotes (Stachel & Zambryski 1989).

In Britain transgenic animals that "can be recalled with certainty", that are introduced to a securely fenced field but are outside strict containment, are not regarded as intentionally introduced to the environment (HSE 1989). If the animal is not native, a permit must be issued. There should be consideration to the possible interbreeding of transgenic animals with wild animals. While large animals pose very little risk in terms of gene transmission, they should still be subject to risk assessment and review prior to release. While permission may be easily granted, it still seems desirable to monitor releases by the submission of an application for release of GMOs.

Case by Case Risk Modelling

Because the different organisms can behave in very different ways, and pose different potential problems, it is generally considered preferable to use a case-by-case assessment (OTA 1988b, MOE 1988, HMG 1989b, Tiedje 1989). What then becomes important are the risk assessment methods. Our understanding of the risks presented is much greater than several years ago, but scientists can not be complacent (Sussman et al. 1988). Some opponents of field trials claim that the onus is on scientists to establish that any real risks can be managed. There are some debates on the numbers of microorganisms that constitute the term "environmental release" (Strauss 1987), but with their rapid growth rate, any number can soon multiply to fill any environmental niche offered. It is because of this that there are calls for a moratorium on such releases (Wheale & McNally 1988). It is possible to sort planned introductions into broad categories for different levels of review. There are different criteria which can be used to determine whether an application is safe.

There are methods being developed to model the potential risks (Fiksel & Covello 1986), but still they are difficult to construct and use. Risk assessment is the use of scientific data to estimate the effects of exposure to hazardous materials or conditions. Risk management is a different activity. It is the process of weighing alternatives to select the most appropriate regulatory strategy or action. It integrates the results of risk assessment of different alternatives. When examining proposals for release of GMOs on an experimental level, risk assessment is needed. The first part of risk assessment is risk identification, after which comes risk estimation (OTA 1988b). Only after the results are known can the wider release of the GMO be considered against other alternatives, the process of risk management. Benefits are part of risk management, whereas they are not part of risk assessment.

There are many examples of small groups of closely related species in which one of the group has become established in a new environment, while others have not, despite apparently equal opportunities. It is also very difficult to predict whether a species will become a pest or not. The five main criteria for evaluating environmental impact include (OTA 1988b):

* the potential for negative effects
* the survival of the organism
* the reproductive mechanism
* the transfer of genetic information
* the transport or dissemination of the organism

A procedure for estimating the risks of each organism has been developed by the Royal Commission on Environmental Pollution (HMG 1989b) which has been called "Genhaz" (Watts 1989). It is a laid out procedure to simplify and regulate the environmental assessment of each new organism. The rules will impede some research, but there are reasons to be cautious. It points out the biggest brake on the acceleration of the number of releases would be a case of serious damage caused by slack regulations, thereby justifying close examination of each case. It also recommends that carefully monitored environmental releases will make a greater contribution to safety than a moratorium (HMG 1989b). A problem is that we may not even know which experiments are particularly hazardous and may not know what the risks are until the experiment has been attempted. We can examine each previous case to assess the potential hazards (Mantegazzini 1986), but there are many experiments which have no precedents.

There are different components of the risks. The probability of each component occurring must be multiplied together to give the likelihood of harm. The components include (Mantegazzini 1986):

* incorporation of gene for hazardous trait into an organism
* chance of release into natural environment
* survival of the organism there
* multiplication of the organism in the environment
* gene exchange or dissemination
* chance that this will be harmful

If the likelihood of the occurrence of any step is zero, the final outcome will be zero, or no harm. In the case of deliberate release the components of the equation are less, but a more detailed analysis is necessary.

One model proposed for evaluating the risk of invasion by a transgenic plant is to consider positive and negative factors in an equation (Crawley 1989). The rate of increase of the transgenic plant in a given environment will be the sum of the positive factors such as the plant development rate, its seed production, survival of vegetative parts, the immigration old transgenic seed from other sites and the establishment of transgenic plants from dormant seeds, minus the negative factors such as ability for cross-species hybridisation, effects of cross-species hybridisation, effects of herbivores, fungi and plant disease and those of mutualists.

The type of scaling of risk that can be made has been discussed, and suggestions made (see esp. Tiedje et al. 1989). The difficulty with all these models is to assign values to each component in the equation. A more detailed type of categorisation is illustrated in Table 9-1, which is useful for guidance.

The level of risk can be minimised by the selection of organisms that have favourable attributes. Good choice of organism can lower any risk. The attributes of the GMO can be different to the original organisms. An example from New Zealand came from efforts to enhance the nitrogen fixing capability of a pine tree by genetically modifying a fungus that inhabited the tree. Two normally nonpathogenic microorganisms were combined, using the technique of protoplast fusion, but some of the new strains may have been pathogenic. In one case, all the tree seedlings to which the organism was applied died (Giles & Whitehead 1977). It was not clearly established that the cause of the trees dying was the new strain. However if it was, it is fortunate that seedlings were affected. If the disease only affected mature trees, it may not have been detected in this trial, and might have been released. This example illustrates the importance of laboratory testing prior to trials.

The results of previous field trials are important in assessing the risk. While the number of these is small, analysis of risks will be dependent upon information from contained experiments, knowledge of the parent and related organisms, and an understanding of ecology and other biological principles. Only with experimentation will adequate theories of risk assessment be able to be developed (Simonsen & Levin 1988). Until there has been more field trials, each new GMO will need to be developed as outlined in the conclusion of the last chapter, gradually expanding.

It is possible to sort out planned introductions of GMOs into broad categories for which low, medium, or high levels of review are appropriate (OTA 1988b). Some of the criteria that can be used to determine whether an application is inherently safe include:

* The GMO duplicates the phenotype and community relationships of naturally occurring organisms
* The GMO will not survive or reproduce after release
* The genetic material will not be transferred to any other organism
* The gene functions will have no adverse effects on the environment
* No genetic material derived from any pathogens
* Past experience with a similar organism or GMO
* Microorganisms present more uncertainties than do macroorganisms

Public Participation

The U.K. Royal Commission argues that individuals and organisations applying for licences to carry out field trials with GMOs should have to announce their proposals through advertisements in the local press. If they want a licence to sell or supply a GMO in a commercial product they should advertise "in an appropriate national newspaper", and in the London Gazette. Members of the public can comment to the licensing authority within 30 days, and all applications will be open to the public (HMG 1989b).

In New Zealand a public notice must be placed in the newspapers of the main cities when an application is received by the committee. The public have six weeks to submit there comments to the committee (FRWP 1987). It also recommends notices in local newspapers near the proposed test site. However, the test sites do not need to be physically marked once approval is given. In Japan the test site for the first field release was required to be well protected by a boundary, and in other countries some barrier to prevent people entering the site is required. In view of the public protests this is warranted. In a sparsely populated country the best protection against such protest is to put the test in an anonymous private field. If approval is given for a field trial at all, it should not present a harm to people anyway, so there may be little need to inform the public of the site. The most critical time is during the process of assessing the proposal.

The US Department of Agriculture introduced draft guidelines which do not provide for adequate public participation. The public does want a role in the process. The draft guidelines also emphasised the protection of proprietary information so that not all details will be released to the public. It will be interesting if the final document alters this.

Table 9-1: Attributes of GMOs and Environments for Risk Evaluation (extended from that of Tiedje et al. 1989)

Level of Consideration, from Less on the left to More (not to scale)

Attributes of Genetic Alteration
Characterization Full Poor
Genetic stability High (e.g.chromosomal) Low
Number of genes Gene deletion One Multiple
Function None Regulation Only New product
Source of Insertion Same species Related Unrelated
Vector None Non-transmissible Self-transmissible
Source of vector Same species Related Unrelated
Vector DNA inserted Absent Non-functional Functional
Monitoring of spread Easy Difficult

Attributes of Parent Organism (Wild-type)
Domestication Totally dependent Some Wild Self-propagating
Ease of control Agents available Not known
Origin Indigenous Exotic
Habit Free-living Symbiotic Symbiotic & Pathogenic
Pest status No relatives pest Relatives pest Pest
Survival in wild Short term Seeds Longterm
Geographic range Narrow Broad
Gene exchange in wild None Frequent
Turnover rate Low High
Mobility Stationary High

Phenotypic Comparison of GMO to Parent Organism
Fitness Reduced irreversibly Reduced Increased
Infectivity Reduced Increased
Host range Unchanged Shifted
Substrate resource Unchanged Expanded
Environmental limits Narrowed, but not shifted Shifted
Disease resistance Decreased Unchanged Increased
Susceptability to control Increased Unchanged Decreased
Expression of trait Independent of environment Dependent
Previous similar release Identical Similar Dissimilar

Attributes of Environment
Selection pressure Absent Present
Wild relatives close Absent Present
Vectors for dispersal Absent Controllable Present
Direct ecol. involvement None Marginal Key
Alternative symbionts Absent Present
Environmental range Very restricted Broad
Simulation of test Easy Difficult
Public access to site None Little control Uncontrolled
Monitoring Effective Untested
Endangered species Absent Present
Geographically isolated Yes No
Important Crops in Area Absent Present

International Regulations

In Europe the GMOs field testing regulations vary between countries. Experiments are underway without any control in some countries, like Italy, but had been banned completely in Denmark, and in West Germany (Klingmuller 1988). European industry formed a group called the European Biotechnology Coordinating Group which has been developing its own regulations and trying to maintain a safe standard of practice (Poole et al. 1988). Their role is being replaced as government regulations become coordinated, but it is still useful to raise the consciousness over the risks of introductions.

There have been several German biotechnology companies that have decided to build new laboratories outside of Germany to avoid prohibitive local regulations, such as BASF and Bayer. The German biotechnology companies regard the laws and regulations as being very difficult. It is necessary for companies planning to produce recombinant products to seek public approval, which makes planning unpredictable. The protest movement uses scientific "experts" to formulate objections to the proposals. This had led to young scientists leaving the country if they want to work in this area (Dickson 1989b). It had also stopped much foreign investment in biotechnology in Germany. If countries have prohibitive regulations they will lose economic benefits of this industry.

The German Gene Technology Act (Gentechnikgesetz) was passed on May 11th, 1990 (GT 1990). All proposals must be published in the official gazette and in daily newspapers, and the public have one month to submit comments. A committee is established to give a report to the competent authority, within six weeks. The committee consists of fifteen members, consisting of ten scientists, and one representative from each of Unions, Industry Safety Organisations, Management, Environmental Protection, and Organisations which represent the interests of research. Among the ten scientists there should be at least six from the area of recombinant DNA, and at least two ecologists. The bill sets up a competent authority which considers the advice of the committee, and must explain if they do not follow the recommendation of the committee. The competent authority is comprised of government ministers in related portfolios, and must report within three months after public comments have been submitted. The state, rather than federal authorities, can decide who will be the competent authority for that state, but all open air experiments and commercial products must be approved by the federal health office. This state regulation of other classes of trial could lead to multiple standards within Germany.

The first proposals for release under the new framework are already under consideration, how successful they will be is another question. The first field trial since the law began on the 14th May, with the planting of 30,000 red petunias that had the red colour inserted by genetic techniques. There were still protesters. There has also been approval for the construction of industrial genetic engineering factories, which will produce human proteins using Escherichia coli K-12 strain (Dickman 1990b).

Denmark was the first country in the world to enact comprehensive legislation designed specifically to regulate genetic engineering and the production, importation and release of GMOs. The Danish Environment and Gene Technology Act 1986, restricted genetic engineering research to classified laboratories, and prohibited deliberate release of GMOs, except for specific exemptions by the Minister for Environment. In mid 1989 after a parliamentary debate, Denmark announced that they had authorised the first field trial or transgenic plants. The plants will be sugarbeet with either tolerance to the herbicide Roundup, or resistance to a viral disease, rhizomania (Newmark 1989). As well as the safety considerations, manufacturers need to demonstrate "usefulness" or "social need" for GMO release.

There are legal directives covering the whole of the EEC. There have been several bills put forward to the European Parliament, with the Green and left parties trying to impose a moratorium on commercial release, and the right wing parties stressed economic freedom. The new directive, approved in April 1990, gives member states eighteen months to comply with guidelines (EP 1989). The European Commission laid out minimum standards for control, with options for individual countries to impose stricter laws. The wording is where regulations "effect in member states do not provide adequate protection of human health and the environment everywhere in the community", the minimum standards must be used. The regulations include details of the required information for any proposal. Most of the information will be open to the public for comment. No deliberate release shall be authorised by the authority unless it is proven and verified to have no negative impact on the environment and humans. The objective of the release must be demonstrated to be socially desirable. The current European proposal has an extensive list of exclusions from regulation, so is open to criticism. Any proposal for experimental release should be notified to other countries in the Community, who can comment, but not prevent the release. This should be applicable to other countries who permit the release of GMOs when the GMOs may travel across national boundaries or into international waters. The European Parliament recommends that a single-market economy means that a product approved in one EEC country should not be restricted in others. However, this does ignore the different geographical environments in Europe.

Holland enacted regulations for GMO release in March, 1990. The procedures basically follow the European Commission's suggestions. GMOs are subject to similar procedures as other new modified organisms. The public must be notified of the location of trials, as recommended by the U.K. Royal Commission. This is despite sabotage to some GMO field trials in Holland and Belgium.

In the U.K., researchers are required to notify the Health and Safety Executive (HSE) of their plans to release GMOs, and the proposals are subject to the approval of the HSE's advisory committee. They were made the agents of new legislation in 1989. All institutions, or individuals outside of institutions must notify the HSE of their work. The definition of genetic manipulation given is "the propagation of combinations of heritable material by the insertion of that material, prepared by whatever means outside a cell or organism, into a cell or organism in which it does not occur naturally, either (a) directly; or (b) into a virus, microbial plasmid or other vector system which can then be incorporated in the cell or organism." (HSE 1989).Work must be notified at least 30 days prior to intended commencement, or in the case of environmental introduction, 90 days prior to introduction. A genetic manipulation safety committee should be set up in each institute, who will prepare risk assessments of the proposed genetic manipulation. This system is based on encouraging cooperation of scientists, and in addition the HSE provide a range of information packages to those undertaking such work, for example on how to construct safe gene vectors, and to use different types of GMOs. This information is very useful, and may be just as important as regulations themselves.

Under the new UK Environment Protection Bill, the Department of Environment will also require notification. In some cases positive consent will be required, not from the committee, but from the Secretary of State for the Environment. The first reading of this bill was criticised for a lack of mention of the subject of public notification, and for a failure to make the advisory committee a statutory body (Maddox 1990). However, the Parliamentary Committee rejected amendments to the Bill to force scientists to make public details of their proposals for field trials of GMOs. Rather the Secretary of State will decide what information will be made public regarding field trials or products containing GMOs (Watts 1990b). This is despite the fact that most information from other areas of pollution considered in this Bill will be open to the public. If researchers fail to obtain consent they will not only face destruction of any GMOs, but a maximum five years imprisonment for noncompliance. The new controls will not apply to techniques that involve only naturally occurring processes of reproduction including selective breeding techniques and in vitro fertilisation. They do cover the release of organisms made by any technique for the modification of any genes or other genetic material by the recombination, insertion or deletion of, or of any component parts of, that material from its previously occurring state.

The control of the release of GMOs in the USA rests with one or a number of regulatory agencies, depending on the nature of the proposed release and the type of organism (OTA 1988b). The policies of the agencies were published in 1986, and may be changing. Before any environmental release of GMOs in the USA a formal environmental assessment of the experiment must be conducted. The situation in the USA may be complicated by individual states imposing regulations in addition to federal regulations. Several states already have enacted regulations, including Hawaii, Maine and Maryland. Other states have rejected proposed bills. The concerns are that when there is existing federal legislation, if the states have extra regulations it may be duplicative and confusing. However, some local communities see the federal regulations as inadequate. Several local communities in New Jersey, Massachusetts and California, have passed ordinances preventing the dissemination of GMOs in their areas (Mossinghoff 1989). The first proposed release of ice-nucleating bacteria in Monterey, California was prevented by a local ordinance preventing the release (Bessette 1988).

The U.S. Congress is considering several draft bills designed to help simplify the procedures. One is known unofficially as the "Biotechnology Regulation and Research Integration Act". It would set out uniform practices for permitting deliberate release of GMOs over the next seven years. An interagency management board would be created so that the EPA, USDA and FDA officials could assist each other in review and enforcement. The bill preempts states from prohibiting experiments that have federal approval, though the states can participate in the review by submitting comments. The states may still impose their own regulations for commercial use of GMOs (Fox 1990b). However, it is too early to know which bill, and what revisions will be accepted. It is certainly better for the experimenters to be sure of what criteria they need to meet and what data they need to present for authorization of experiments, rather than being challenged by local court cases. The lesson for other countries is to decide on good federal legislation and to settle for only one layer of regulation. In Britain, the initial assessment committee for a GMO, from the host institute, should invite the local Environmental Health Officer to join it. This allows some immediate local input, prior to central review.

In 1986 the Ministry of Agriculture, Forestry and Fisheries of the Japanese government proposed their first draft guidelines on biotechnology. There has been much work on a case-by-case level with testing in restricted and controlled test areas to avoid the spread of transgenic plants. Up to 1988 there had been no proposals to release genetically engineered microorganisms (Uchida 1988). There have been concerns expressed over the release of GMOs (Watanabe 1988). There has been more work on transgenic plants. The first field trial was approved in 1990, after a 2.1m high fence was built around the field to keep out animals and people. A committee from the Ministry of Education drafted new recombinant DNA regulations, covering contained use of GMOs and field trials in August 1990, but these will not be available for public disclosure until approved in 1991. There is still debate on whether legislation is desirable.

Canada has an Environmental Protection Act 1988, which controls biotechnology products not subject to other legislation. Biotechnology products developed for environmental application must be subject to approved field trials before commercial manufacture.

There are several different controls in Australia. There are administrative guidelines laid down. The initial committee was the Recombinant DNA Monitoring Committee (RDMC) set up in 1981, and this was replaced after 5 years by the Genetic Manipulation Advisory Committee (GMAC). This committee advises research institutes and the government (Federal and State) on safety procedures and possible genetic hazards. Scientists proposing to release new organisms are expected to comply with the RDMC's guidelines (RDMC 1985). The details are submitted to the committee with the usual type of required information, and the proposer must establish the safety of the experiment. The committee is advisory only - it can not approve the release itself (Skene 1988). There are several pieces of environmental legislation which could be applied to GMOs, and there have been reports by law reform groups on implementing a more statutory basis for the committee.

There was a 1989 report by the Victoria Law Reform Commission (LRCV 1989) that recommended specific legislation. This would be the simplest regulatory system, especially if enacted by the Federal (Commonwealth of Australia) Government. The legislation would make it mandatory for all releases of GMOs to be notified, that an environmental assessment must be conducted prior to release, public advertisements and information concerning the release would be made available for comment, and that releases would be subject to approval from the appropriate Government agencies. The existing GMAC would be used.

Regulation of the release of GMOs in New Zealand is based on the case-by-case assessment of proposals by a regulatory committee. A background paper on the New Zealand situation was published in February 1987 by the field release working party (FRWP). After circulating a draft document, public submissions were received. Most favoured the establishment of a statutory committee which would make it obligatory for all proposals for the release of a GMO to be notified (FRWP 1987). At the time of writing, in early 1990, the committee is still without a statutory basis, but is functioning until one is established (possibly in 1990). The Interim Assessment Group (IAG) is serviced by the Ministry for the Environment. The IAG has been given regulatory control as well as performing an advisory role. All researchers in the public sector are obliged to submit all proposals to field test or release GMOs in New Zealand to the IAG for their advice. The advice is also recommended to private researchers. Their recommendations are made to the Minister for the Environment, who may pass it on to the appropriate Minister for government departments, or to the University or private company. The information contained in a proposal will be open to the public with the exception that any sensitive, potentially commercially important, information will be protected if requested.

One thing is for sure, we do need experience, which involves small scale trials, to learn more of the interchange of genetic information in different ecosystems, before launching into the technology full scale. We ought to consider ecological "boundaries", areas of limited genetic exchange as at least provisional warning signs of potential danger zones for the casual transfer of genes between species (Suzuki & Knudtson 1989). Those trials have been underway for several years, and gradually we will begin to see some of the many benefits of the technology, and see whether any negative consequences emerge. It is important that the benefits reach the farmers and the people who need them, rather than being solely a commercial gain for another technological industry.

The harsh rules in Western Europe may lead biotechnology companies to conduct field tests in countries that have liberal laws. It will be cheaper, and easier as there are less controls. Groups of scientists have visited the USSR to examine whether it is easier to conduct field trials in collaboration with Soviet scientists, in the USSR (Watts 1990a). The testing of recombinant vaccinia viruses made in the USA, in Argentina, indicates the use of countries with less regulations. It was performed without the knowledge of local authorities, and the viruses were introduced in the diplomatic baggage of UN authorities, against Argentinean law (Torres 1988). The experiment itself was not the problem, but the way it was planned and was implemented. India has recently developed their own set of recombinant DNA guidelines (Jayaraman 1990). The International Institute for Cooperation in Agriculture (IICA) is trying to encourage biotechnology in Latin America, and in 1988 distributed regulatory guidelines to avoid the "dumping" of uncontrolled genetic engineering experimentation from developed nations. The OECD is trying to formulate guidelines, which may be harmonised with EEC guidelines. The ultimate aim is to harmonise worldwide regulations. Ecological effects and geographic ranges of organisms transcend political boundaries, so it is essential to promote international coordination of risk assessment and regulation.

Applications to Release GMOs

There need to be different factors included in applications to release GMOs, and from the desirable extent of analysis, the applications can be lengthy. The various guidelines have common features, and it is desirable that greater international agreement be reached on the information to be provided for regulatory consideration. The key elements that need to be considered include (HMG 1989b):

* identity of personnel involved (qualifications etc.)
* Objectives of release
* location of proposed release, geographic and environmental information
* descriptions of parent organism, vector, GMO (see table 1)
* description of the manipulations used to produce GMO
* arrangements for release, preparations, timing, method, decontamination
* potential environmental effects
* monitoring arrangements
* contingency plans in case of unexpected events
* results of prior local consultation and assessment

The evidence should provide convincing evidence that the proposer has carried out a thorough risk assessment of the proposed release. A handbook will be produced in Britain, based on a system for risk assessment called Genhaz, which should help to identify environmental impacts which might otherwise be overlooked.

Different committees requirements for information vary, as do their criteria for what is a GMO. Some information is open to the public, while some information that is industrially sensitive, such as that which if made public would invalidate a patent application, needs to be protected. The expertise that can deal with the release of a microorganism may be different to that required for human applications. The members of the committee need to be flexible, and to be able to seek expert consultant advice. Some examples of specific questions that need to be answered in a proposal should include:

* Genetic Characteristics of the Organism
Parent Organism, identification, pathogenicity, taxonomy, source, variety, capacity for gene transfer and reproductive cycle.
Source of the vector DNA, and properties, capacity for gene transfer.
Source and function of inserted DNA, similar information to that for parent organism of the GMO.
Genetic structure of inserted DNA (vector and gene), and method used.
Verification of the number of copies of the insert in the GMO, and the genetic structure if possible (laboratory results).
Laboratory, and researchers who modified the organism, and precautions used in laboratory handling.
Results of laboratory experiments on stability, and expression of trait.
Does the trait involve any animal suffering?

* Environmental Properties
Parent and donor organism's habitats and distribution.
Susceptability to temperature, humidity, desiccation, Ultraviolet light etc.
Survival of these organisms in similar conditions to field trial.
Does the organism produce spores, seeds or longterm survival bodies?
Biological processes that the organism is involved in, ecological niche.
Other organisms closely dependent upon this organism in the habitat, such as competitors, and symbionts.
Will the GMO modify the abundance of other organisms, and what are target species? Both the predicted effects, the effects of laboratory studies, and the parent and donor organisms actions should be known.
Are there any possible consequences for human health, agricultural production, the ecosystem and pollution in that area?
What are the consequences of long persistence of the GMO?
Consequences (and probability) of gene transfer to other organisms?

* Details of proposed site
Location of the site, nature of surrounding area, facilities, ownership, proximity to farms, roads, people's houses, other crops.
Details of supervision, will there be fences, signs etc?
The contingency plans for the unexpected, such as fire, floods, animal invasion, public protests.
Production of the organism prior to release, and transport to test site.
Elimination methods after experiment, or in event of necessary termination.
Results of relevant prior field tests on same or related GMO.
Details of any target organisms, pests etc. Ecological assessment.

* Future Goals
Why is this GMO of potential benefit, and being tested?
What would the scale of larger trials be if this trial is approved and successful?
Some evaluation of the social and economic benefits of the GMO compared to viable alternatives.

In the ideal case, the application for release of a GMO should be processed swiftly. In small countries it should be possible for a committee to provide information and advice to the researchers. If they omit some information the procedure should await the necessary information. The best system is one where there is feedback between the committee and the researchers so that they can advice the proposer. They may require preliminary experiments to be done. A legalistic approach, where a proposal may be rejected if it omits some details and requires reapplication is undesirable. In many cases the people who know the most about an organism may be those doing the experiment, so they should be actively involved in the procedure.

As well as a role for public submissions, and public representation on the committee, there should be consideration of the social benefits of such a release. Much public anxiety can be averted if the decisions are made in public. Some experiments may be merely to attain information on gene transfer, which is of general value. Most will also be of potential economic benefit. Rather than only considering the short-term questions, we need to consider the long term goals and consequences of such releases. This is especially necessary before any commercial use of GMOs, but is also applicable to experimental size trials. For example, there may be little benefit seen in developing herbicide resistance to a nonbiodegradable herbicide, and in view of the potential risks of gene transfer to weeds, it may not be justifiable even as a field trial however, we may see advantages for a crop resistant to a biodegradable herbicide which will also reduce herbicide use.

Challenges Facing Regulators

As discussed above, there are some dilemmas facing regulators, and it may be useful to summarise some of these:

* Definition of organisms that are to be included. The definition of a GMO may be ambiguous. To make it safer, all new organisms should be considered. A simple, definition of a GMO is an organism that is genetically different from the starting organism as a result of human manipulation. However, there may be too many organisms for the regulators to consider in many countries.

* Risk Evaluation is difficult, becoming simpler only after past experience. There should be a repository of test results.

* To create a favourable climate for research and economic development, excessive regulations should be avoided.

* Consistent regulation of different classes of GMOs. This is easier in a small country , but can be very difficult in a country like the USA which has different regulatory authorities responsible for different GMOs.

* The regulations should be statutory, with little room for challenge by petty court cases.

* The same agencies in the USA both fund and regulate research.

* Consistent penalties for violators need to be assigned.

* Public opinion in the local area, as well as over the whole country, needs to be considered.

Financial Liability for Damage

There is uncertainty over the actual risks of ecological or crop damage from deliberate release of GMOs. The risks can be characterised as low probability/high consequence type. As experience with GMOs grows there will be a standard of acceptable risk set by regulations, which insurers, as well as economists, will have to assess (Fleischer 1989). The average risk to the environment from products of genetic engineering may be similar for quite different kinds of products. However, the variability in risk among successive individual cases may be higher for GMOs than other products. This produces a broader risk profile (Fiskel & Covello 1986). There are various types of risk also, to health, food safety, ecosystem disruption, pollution and technology failure.

One way to put a cost to the risks is via insurance brokers. Those releasing transgenic organisms should accept product liability and insure themselves. Lloyds of London view deliberate release as an insurable proposition (Hodgson 1989a). The risk assessment procedures will require expert advice, but this has been done in other technological risks. However, we should not leave it to the market to decide such important questions. The costs of an accident are unknown. Under the new German Gene Technology Act the German federal government will assume liability on approved experiments for damages caused, up to DM160 million..

It is difficult to assess all the benefits that will come from the introduction of new organisms, though many direct benefits can be identified. The benefit needs to be considered to the nation as a whole, or the particular benefits to farmers, or a section of the community. The benefits can be assessed in monetary terms, but other factors are important. It is difficult to value parts of the environment in a monetary way.

Food Safety & Consumer Rejection

Novel Foodstuffs

At recent conferences on GMOs the concern has been switching somewhat from the environmental issues to the issue of safety of the end product for human consumption. There are worries about genetically engineered crops, and many will soon be under scrutiny in the USA for Food and Drug Administration (FDA) approval. The regulatory authorities have been slow to consider the issues of genetically modified foods, and the increasingly larger scale plantings of crops should force them to consider these issues.

The use of recombinant DNA technology should be aimed at protecting the natural environment. For this to happen, new technologies that minimise erosion, desertification, salination and reduce the use of chemicals, pesticides or drugs should be encouraged. If we consider the large amount of antibiotics given to farm animals (half the USA production is used for animals), which results in bacterial resistance to the antibiotics so that they are no longer effective for medical use, this is a much more serious problem. However, new crops must be safe, not just better than the alternatives that are accepted today.

The British Government recently approved the production of a food product containing a live GMO. The GMO in question is a genetically manipulated strain of bakers' yeast (Saccharomyces cerevisiae). The yeast is manufactured by Gistbrocades, the world's biggest yeast producer. The yeast will be used commercially at the end of this year in U.K. bakeries, and permission is being sort also in The Netherlands, France, West Germany, Portugal, Morocco, Japan and the USA. The maltose permease and maltase genes from the yeast were combined with new promoters from another strain, of the same species. There are also some short pieces of synthetic DNA, such as stop codons, included (Aldhous 1990). The genes will be constitutive, that is, continually active, rather than inducible only in the presence of maltose and absence of glucose. The yeast should take up and digest maltose more efficiently, making bread rise more quickly. The product will not be required to carry a label indicating it was made using genetic manipulation. It has taken four years from development to approval.

Another example that may soon be approved is the use of genetically modified yeast for beer production. In late 1986, a USA Company, BioTechnia, arranged for trials at a UK Brewery for the production of low calorie beer. The yeast, Saccharomyces uvarum, contains a gene from Asperigillus niger for glucoamylase (Palca 1986). This allows faster brewing, and the beer requires no additives to remove starch as the alternatives require. If the unpasteurised beer was provided, it would contain live yeast.

It is expected that approval for limited trial human consumption of a genetically engineered tomato produced by ICI, will be given during 1990 by British authorities. It will be longer before approval is given for general public consumption. There have been taste trials of genetically modified crops in many laboratories around the world, and most scientists would be happy to eat such foodstuffs themselves, as there is no substantial difference compared to crops bred by traditional methods. It should be one of the first plant crops to be approved for human consumption.

Plants that are genetically modified may need some preliminary testing to ensure that no secondary toxic product has been produced after the manipulation, as it has been found that some plant defenses against pesticides involve the synthesis of carcinogens, cancer-causing agents. These plants may need approval before being sold for human consumption. There are concerns that there could be harms from high levels of some toxins, which are probably of low risk, but there was the case in the 1960's over a new strain of potato called "lenape" which had high levels of a usually trace level toxin, solanine, and caused illness after eating. The range of solanine in potatoes varies 40 fold. There are unknown affects on people's allergies. The concerns also cover grains or food that can be given to animals as feed (Wickelgren 1989), though the effect of any toxin is unlikely to be passed on. There is the potential to make more nutritious plants, such as by increasing the level of the amino acid methionine in soyabeans.

There are some products that have a history of previous use. The B.thuringiensis insecticidal protein gene has been incorporated into plants for insect resistance. This protein has been licensed in various formulations since 1962 (OTA 1988b). It is available in a number of formulations in over 400 products in the USA. There have been very few instances of harm being noted, even though hundreds of thousands of tons of the protein have been administered. One harmful effect observed was an association with corneal ulcers in humans (Samples & Buettner 1983). It will be important to clarify this before approving the consumption of transgenic plants that contain this toxin.

In many European countries, if novel food proteins are regarded legally as food, they are not subject to special legislative approval. They are different to food additives which require safety appraisal. The term novel food protein can include sources of protein not previously exploited for human consumption, including novel enzymes from GMOs (Mantegazzini 1986). There does need to be further regulatory control of food proteins in this category, as it can include novel toxins. The volume of protein consumed will be much greater than if the substances were just food additives. To observe any effects due to minor constituents of food may take a considerable length of monitoring following consumption.

There have been many natural toxic substances described in plants, including existing food crops. Some act in subtle ways, such as alkaloids and teratogens, which can cause birth defects (Cheeke & Shall 1985). The actual number of naturally occurring toxicants detected in a major survey of plants was 148, with 14 found in normal diets (IFBC 1989). Compared to the huge number of food constituents, this is very small. Many substances are yet to be described. For example, some kinds of beans need cooking to make them safe to eat, and potatoes and rhubarb have poisonous parts which we need to know about to avoid eating. Some cases should be easier to predict. For example transgenic plants expressing the TMV coat protein gene are found to have 0.01% - 0.5% capsid protein per total protein. This is well below the levels found in plants infected with this endemic virus. This should facilitate registration and commercialisation of the virus-resistant plants (Gasser & Fraley 1989). It is likely that virus-resistant plants will be among the first to be approved for human consumption, however it is very difficult to predict what regulatory authorities will do with new foodstuffs.

In 1988, the International Food Biotechnology Council (IFBC 1989) was formed with the aim of identifying the issues and assembling a set of scientific criteria to evaluate the safety of food derived from plants and microorganisms resulting from the applications of biotechnology. They did not consider animal foodstuffs. The membership of the Council is comprised of approximately 30 companies, who set up committees to look at scientific, legal/regulatory and policy/public relations aspects.

They discuss the variability of composition inherent in foods and food ingredients, such as the nutrients and toxicants. There are several vitamins (A & D), certain trace minerals (Fluorine, Iodine, Copper, Selenium) and other essential nutrients that are consumed safely only within a narrow range. Intake below that range results in deficiency or disease, and above that range in toxicity. There are many food toxicants that are already accepted at low levels in foods. For intentional introductions a safety factor of 100 is commonly used. They surveyed the range of toxicants and nutrients in traditional foods as a basis for comparison with new foods, and as the standard for defining food that is considered safe.

They recommend that the regulation of food from GMOs be directly patterned on the existing law. This is also the recommendation of the Victorian Law Reform Commission (LRCV 1989). There are existing legal requirements for food safety which can be used (OSTP 1986). Also the possibilities of financial liability and legal suits in the USA will make companies very cautious. If the purpose of the modification is to introduce as an expression product of the transferred gene, a functional chemical entity that, if introduced exogenously, would be regulated in the GMO as a food additive or GRAS substance, then the new food would be treated as such. They proposed decision trees for evaluating the relative safety of food derived from GMOs (IFBC 1989).

The IFBC recommend that the following types of genetic elements be considered acceptable for use in food:

* Uncharacterised genetic material presently consumed in food, that was introduced from non-food species used as sources of genetic variation in developing and improving foods using traditional methods of genetic modification and for which documentation of safe food product is available.

* Fully characterised genetic material derived from nontoxic, nonpathogenic microorganisms that are not intentionally consumed as food but are commonly found in or on food and accordingly have an established record of safe use.

* Fully characterised noncoding DNA from sources that are not traditional foods. Since noncoding DNA can not encode any protein then only the intrinsic properties need be considered. The only concern is a quantitative one: large quantities of nucleic acids can cause gout.

* Coding DNA from nonfood species that have already been used as sources of genetic variation in developing and improving foods using traditional methods of genetic modification and for which documentation of safe food product use is available.

There is some possibility of secondary metabolic effects from the inserted genes. There will probably be no pleiotrophic effects as the biochemical mode of action of each gene to be inserted is known. However, because there are many metabolic routes inside cells, the excesses or absence of different chemicals caused by the inserted genes, may have effects in what is called secondary metabolism. For example the herbicide tolerant plants often involve the synthesis of some amino acids, and there may be secondary effects, that are difficult to predict. The control of the metabolic pathways is still poorly understood, and the only way to determine them is to analyse the products of such plants.

If plants are made in a more uncontrolled many, by insertional mutagenesis there may be further unknown effects. Pieces of DNA important for regulation of different genes may be either turned off or turned on, and this may affect other genes. Because there are some natural toxins in most plants at low level, it is important to ensure that these toxins remain only at low level and are not accidentally increased in a new variety.

Alternative research is being performed at Dental Research centres in the USA. The bacteria responsible for tooth decay are being genetically manipulated so that they do not produce acids that causes the tooth decay. The new strain could be applied in toothpaste, or it could be applied once for life if competitive with other mouth bacteria (Goel 1989). This will pose interesting regulatory problems, as it will involve the continual exposure of users to the bacteria.

There are less toxins known to be present in animal foods, but care must also be taken. The tolerance for vitamins can vary, for example the level of Vitamin A in polar bear liver is 500 times higher than found in cat liver, and some dogs have been poisoned from eating polar bear liver. It is unlikely that such differences will be found between two varieiteies of the same animal, but monitorring is essential. There was recently a controversy in Australia regarding accusations that 53 genetically altered pigs were sold for human consumption without proper approval. The pigs came from a company, Metrotec, associated with the University of Adelaide. They have been trying to breed leaner pigs since 1982. The work involves pigs with inserted growth hormone genes. In 1988, Metrotec transported 53 pigs, who were the offspring of a genetically altered pig but did not express the growth hormone gene, to the abatoir. They had been given permission by the South Australia Health Commission and the National Health and Medical Research Council, as they did not express the hormone (Anderson 1990). One would not envisage any harm from consumption of these pigs.

One of the key legal points is whether the gene should be regarded as a food additive or not. When naturally occuring food is altered, there are two broad classes. In the USA, Federal Food, Drug, and Cosmetic Act, the clauses for the two classes read "in case the substance is not an added substance such food shall not be considered adulterated under this clause if the quantity of such substance in such food does not ordinarily render it injurious to health" and "a food shall be deemed to be adulterated if it bears or contains any added poisonous or added deleterious substance". There are five categories of added substance that are listed, including unavoidable by-products of food manufacturing, pesticides, food and colour additives and new animal drugs (Jones 1989). The added gene will be physically made in the same way as the rest of the DNA, so is not an added substance in that sense. However, the information contained in the gene is an added substance in that reading of the definition. In the case of animal gene transfer in the USA the exclusion for animal drugs may be applied to transferred genes. However, as discussed here, and in length in the IFBC Report, most transgenic crops should present no health risk and should be approved without considering the new gene products as manmade substances, such as pesticides.

Public Acceptance of New Foodstuffs

It is important to confirm that the products used are not only harmless to the animals but are harmless to human consumers, unlike the case of anabolic steroids which have been used to increase animal growth rate but remain in the meat possibly affecting human consumers, a reason why EEC countries ban the import of such meat from the USA. There have been earlier cases with more widespread use of steroids. In Italy in 1980 it was found that diethylstilboestrol, which enhances the growth of calves was present in the veal to be used for baby food, causing infant cancers and the onset of secondary sexual characteristics (Wheale & McNally 1988).

There have been some controversial results of treating animals with recombinant bovine somatotropin (BST), as it can increase the milk yield (claimed to be about 10-20%) with improved feed-conversion efficiency without any sign of changes in meat composition. It also results in leaner lambs and pigs, which means healthier meat, as it alters the metabolism in favour of net protein gain (Lamming 1988). BST is the first product of genetic engineering to be offered to farmers. It has been approved in the USA by the FDA as presenting no risks to human health (Sun 1989), but they are still considering whether it is harmful to cows and should decide in early 1991. It will probably be rejected in Europe. The decision will be taken in late 1990. Many small farmers and the Green party do not want to use BST, and there is an immediate question of why it is needed when there has been considerable money spent to control the overproduction of milk already. In several studies using BST the cows have shown increased mastitis and stress, increased incidence of infectious diseases, reduced fertility and heat intolerance which make it easier for large scale farmers who have cheaper veterinary help (MacKenzie 1989).

There has also not been adequate food testing performed on the milk produced from cows treated with BST. One of the BST proteins contains an additional nine amino acids compared to the normal protein. There are legitimate concerns over the safety of it. The research was dominated by the industrial companies, and they have not disclosed much information from their tests. The US Animal Health Institute is supporting the use of BST, but has failed to mention a wide range of adverse effects in about half the trials that have been conducted (Epstein 1989). Stressed cows produce more adrenalin and other steroids, which may be present in the milk or meat. Insulin-like growth factor-1, a hormone that affects postnatal growth in humans is found in BST milk, and the total level of BSTs is 50% higher than in normal cows' milk. BST is known to be biologically active in a wide range of species, including goats, sheep, pigs, mice and fish. There have been studies showing that partially-digested BST is biologically active in humans, inducing nitrogen retention. Nitrogen retention has been linked with abnormalities in the human central nervous system. Until further studies are done, industry can not say that BST is inactive in humans (Epstein 1989, MacKenzie 1989). This test case has certainly not been handled well by the agencies concerned.

Even if BST is safe, the consumer objection to milk produced may be considered enough to warrant a ban in Europe. The European Court ruled that consumer objection was sufficient reason to ban the use of steroids to fatten meat animals, so it may do the same with BST. In the USA there is also consumer rejection of BST-milk, and requirements that milk be labelled if it is used. It should certainly be necessary to label products made using BST treated animals, and the public can then be involved in the decisions regarding the use of it. A survey of California milk producers in late 1987 indicated that many farmers would not use BST, or would wait to see the results. Their concerns were whether there would be negative consumer reaction, as well as the effects on the cows health. Some creameries will not accept BST-milk (Zepeda 1989). This consumer pressure which has resulted in bans on BST milk in some U.S. states (e.g. Wisconsin), and in Europe, is a much broader issue. More than milk is at stake, rather the way that decisions on future agriculture are made. Even if BST has no effect on humans, it may still be banned.

Consumer objection has delayed the introduction of the technology of food irradiation (Murray 1990). Irradiated food has been tested widely and found that if radiation levels are controlled it is safe. It has the significant advantage that it sterilises food, which means the food lasts longer, and it may lower the rate of food poisoning. It has been supported by the World Health Organisation for a decade, but consumer groups have opposed it, in the UK and in New Zealand. It is legal in Holland, Belgium and France, and will only be used for up to 5% of the food. There is currently much public objection to it in the UK as the government debates a bill to allow it (Coghlan 1990).

In the recent public opinion survey towards genetic engineering performed in New Zealand, the issue of public acceptance of foodstuffs produced by GMOs was addressed (Couchman & Fink-Jensen 1990). Not only is food safety an issue but as stated the consumer opinion can be influential. About 75% of the public were aware that GMOs could be used to produce food and medicines. All the respondents were asked the question "If any of the following were to be produced from GMOs, would you have any concerns about eating them?". If the respondents said they had concerns, they were asked what concerns they would have. Some of the results are presented in Tables 9-2 and 9-3.

Table 9-2: Concern expressed by the general public in NZ about using various products produced with, or from, GMOs (%) (Couchman & Fink-Jensen 1990)

Responses(%) to the following types of Foodstuffs or Products: Dairy Food; Vegetables; Meat; Medicines;

Concerned 42.8 38.4 48.3 34.1
Not concerned 57.2 61.6 51.7 65.9

Reasons for concern (% who included as reasons):

Unnatural 26 32 27 14
Unknown effects 22 20 21 23
Not sure what eating 14 12 15 10
Product safety 12 10 9 15
Need information 11 10 9 12
Side effects 7 5 5 12
Unknown area 4 4 4 5
Disease causing 3 2 3 1
Product quality 3 4 4 1
Animal cruelty 1 0 5 1
Anti-medicine - - - 7
Don't know 11 11 12 14

Table 9-4: The concerns over products made using GMOs by different sectors of the New Zealand public (Couchman & Fink-Jensen 1990).

The survey of the public was face-to-face, the rest were written questionnaires. The total number of respondents were; public 2034, biology teachers 277, farmers 200, scientists 258. Type of Foodstuff or Product using GMOs. The Occupation of Respondents: Public; Teachers; Farmers; Scientists;

Dairy Products
Concerned 42.8 13.0 41.0 24.0
Not concerned 57.2 87.0 59.0 76.0

Concerned 38.4 9.7 30.5 21.7
Not concerned 61.6 90.3 69.5 78.3

Concerned 48.3 13.7 26.0 24.4
Not concerned 51.7 86.3 74.0 75.6

Concerned 34.1 9.7 29.0 19.8
Not concerned 65.9 90.3 71.0 80.2

The lowest level of concern was for medicines, and the highest was for meat. Of the people who had general worries about genetic engineering, the highest proportion had concerns about meat, and lowest about vegetables that were genetically modified. Of those who expressed concern about using each type of product, about half had no worries about genetic engineering in general. There was greater worry about all types of products by women than men. The level of concern generally increased with age. Therefore there may be special concerns about food made by GMOs. The most frequent reason for concern was that it was unnatural, or that it might have unknown effects, and they would not know what they were eating. This may be an important feeling to overcome, though it should be remembered that at least half the people were not concerned. This concern was shared by farmers. Scientists were more concerned about product safety.

People's awareness of food made using biotechnology such as biosweetener, biowine, and biovegetables was also surveyed in the Japanese science magazine Newton (1989), that was refered to in chapter 3. Some were worried, though most did not understand why it might be dangerous. 23% thought it was good to use sweetener made in bioreactors, 23% did not care, but 41% were unsure and 5% worried. With wine made using cell fusion, 28% thought it was good, 30% did not care but 30% were unsure and 3% worried. 83% knew new vegetables are being made by cell culture. 38% thought it was good, but 34% were a little unsure and 5% were worried about this. By law, food made using biotechnology is treated the same as other food; 35% thought it was no problem to treat it the same if it was safe.

Many people think it is useful to have biomedicine or biodetergent. 46% thought that it was good that medicine made by genetic recombination might be sold soon, but 33% were a little worried, with 3% worried. Another question pointed out that recently the size of washing machine detergent packets had became smaller. 74% knew that this was because they use stronger enzymes made using biotechnology. 45% thought this was good, 21% did not care but 25% were a little worried, and 4% worried. In response to the comment that "biotechnology is starting to be used for daily life"; 55% thought that this is good, 34% were a little worried and 5% were worried.

The readership of this magazine are people with some interest in science in general, and are on average more educated than the general public. In Japan the prefix "bio-" is used to attract the public to new products, made using biotechnology. In this sense the public may be attracted by new products seen to be better. The general public may not be aware of the concerns about these products. In the American OTA survey it was found that education did not correlate with degree of concern about biotechnology, but in the New Zealand survey there was some correlation; the more educated may be aware of more risks even though they may not reject the products because they are seen to be "unnatural". Further study of this will highlight the way education may be used to inform people of the actual similarity of food made using any technique, there is far more variation among different foodstuffs that people eat. 48% of the New Zealand school biology teachers thought that the material available for the teaching of genetics and genetic engineering was unsatisfactory. This must be improved.

Better Products?

Preparations of bacterially produced "human" insulin have been available since the early 1980's. They were assumed to be better than porcine insulin, and most diabetics in Britain have been switched to the human insulin. However, there are serious doubts whether it is actually any better. One of the concerns over porcine insulin was that a few people develop allergies to it, but it is also possible to develop an allergy to the preparations of human insulin. It has been observed that those people using "human" insulin do not receive symptoms of hypoglycemia (a low blood sugar level) as they do if they use porcine insulin, which can have serious consequences. The reason for the widespread switching was commercial promotion of a new product, not necessarily a better one. There are now 17 different brands of human insulin sold in Britain, yet there has been no clinical advantage found (Lesser 1989).

Tissue plasminogen activator (TPA) is a recombinant DNA product that was developed by Genentech, as a blood clot dissolving agent. The 1989 sales were worth US$ 200 million. The U.S. Government decided that TPA was too expensive for the Medicare scheme, in 1988. A recent study has found that TPA may be no more effective than streptokinase, which is one tenth of the price, and used in Europe (Gershon 1990b). Streptokinase is derived from streptococcal bacteria, and commands a two thirds share of the market for these agents in the USA. In view of the results of this study, it will reduce the proportion of the market that TPA hoped to fill.

Protein pharmaceuticals produced by recombinant DNA technology require approval. Part of this process is the purity analysis of the product. Analytical methods have improved as the technique develops. The approval of any pharmaceutical relies upon a convincing demonstration by the manufacturer of the safety and efficacy of the product. Before human trials, analysis must be made. There are a variety of impurities that are possible, including endotoxin, host cell and media proteins, monoclonal antibodies, DNA, and infectious agents. Contaminants can be the same substances accidentally present. Impurities can have immunological and/or biological effects (Anicetti et al. 1989). Testing is also required during the production of each batch. For example, there are approximately 750 separate tests performed in the production of human growth hormone. Recombinant DNA techniques combined with new purification methods have produced the highest purity proteins that have ever been available for human therapeutic applications (Davies 1988). Some proteins, such as human albumen, may be made using this technology because of the purity possible, and there is no risk of virus transmission, as there is with blood products.

In August 1990 the controversy over the safety of products produced by recombinant DNA technology resurfaced with the association between a batch of the amino acid tryptophan produced by Showa Denko, a large Japanese chemical company, and a disease called eosinophilia-myalgia syndrome (EMS). Tryptophan is part of a dietary supplement used to treat insomnia, depression or premenstrual tension. The outbreak of EMS in USA during 1989 led to 27 deaths, and can be traced to a particular batch of this amino acid. The batches concerned were made with lower amounts of carbon which was used to filter out contaminants, and were made using a new strain of bacteria. The actual cause has not been determined, but opponents of genetic engineering are calling for a moratorium on the approval of further products of genetic engineering. The probable cause of the EMS is a contaminant, so rather than the genetic modification being responsible, the purification procedure was to blame.

The range of products that can be produced include vaccines, hormones, and genetic probes for genetic screening. The quality of the products is regulated according to their intended use, not their method of manufacture. There are existing laws to ensure safety, and quality of these products. In the USA the FDA has a Biological Response Modifiers Advisory Committee, which assesses the results of clinical trials of new drugs. There needs to be clear evidence of a therapeutic effect, with no harmful effects, before a drug will be permitted to be more widely used. In the case of a longterm disease, the therapeutic effect may be difficult to assess, but this problem existed before the advent of drugs or proteins made using recombinant DNA techniques. The regulators must attempt to introduce useful drugs as sson as possible, in cases where there is no alternative therapy. Some proteins must be used very carefully, by experienced clinical teams, but the drug should still be approved if the criteria of use is restricted to use by such clinics.

The list of approved products is expanding. According to the 1990 annual survey by the U.S. Pharmaceutical Manufacturers Association, there are 104 different genetically engineered medicines being tested in human clinical trials or being reviewed by the FDA. About half are for cancer related conditions, and 15 are being tested for treating HIV or AIDS related conditions. Of the 104 drugs, only 11 have so far been approved by the FDA for physicians to prescribe. At the time of writing another 18 have completed clinical trials, and 14 more are in the final stages of human trials.

Please send comments to Email < >.

To Shaping Genes contents
To Shaping Genes chapter 10
To Eubios Ethics Institute home page