SUMMARY

This paper gives an outline of the legislative framework for the release of genetically modified organisms (GMOs) in the UK and describes the recommended approach to risk assessment for the release of GM plants. It discusses the Department of the Environment's GMO Research Programme which commissions research to obtain information on risk assessment and to develop policy on the release of GM plants.

INTRODUCTION

Advances in modern biotechnology and the application of recombinant DNA technology to produce genetically modified organisms (GMOs) have been rapid over the last decade. Controls have been developed and they have undergone a number of changes to keep pace.

In Great Britain, two main sets of regulations control work on GMOs. Parallel but separate regulations apply in Northern Ireland. One of these regulations focuses on the contained use of GMOs and the other controls the release of GMOs into the environment. This paper will confine itself in considering the regulatory, research and policy aspects for the release of GMOs in the UK.

The main purpose of these controls is a requirement for an assessment of risks before a GMO is released into the environment. This has meant developing approaches to risk assessment of living organisms with research playing an important role in identifying the potential hazards and in the management of any risks.

CONTROLS FOR THE RELEASE OF GMOS

The release and marketing of GMOs are regulated by The Genetically Modified Organisms (Deliberate Release) Regulations 1992 and 1995. These regulations and Part VI of the Environmental Protection Act 1990 implement the European Community (EC) Directive 90/220/EEC on the deliberate release into the environment of genetically modified organisms. Adopting a precautionary approach, the legislation sets out risk based controls to ensure that harm to human health and the environment is prevented or minimised. The legislation is kept under review in line with increasing knowledge and experience.

The Department of the Environment (DOE) operates the GMO deliberate release regulations. This means that anyone proposing to release or market a GMO should submit an application for consent to the DOE. Following a review of the application, a consent may be given by Secretary of the State for the Environment with the agreement from other Government Departments depending upon the purpose and location of the release.

A statutory advisory committee of independent experts, the Advisory Committee on Releases to the Environment (ACRE), gives advice to the Secretary of State on all applications for consent. ACRE comprises of scientific experts and representatives of various interests such as industry and environmental groups, and they advise specifically on whether or not a consent to release or market GMOs should be granted as well as conditions which should attach to any consent granted. The decision to grant, vary, revoke or attach conditions to a consent and take enforcement action rests with the Secretary of State.

Other aspects of the regulations require that a proposed release is advertised in a newspaper and certain information on the release is placed on public registers so that information is made available to the public. Also, a summary of the proposed release is sent to the EC Member States for the purposes of information exchange and comment.

Since the regulations came in force on 1 February 1993, and for the period up to 31 May 1996, the DOE have issued 103 consents for the experimental release of GMOs (Table 1). Some of these consents are for programmes of work covering multiple releases over several years. Ninety-six consents have been for the release of GM plants and the remainder (7) have been for the release of GM microorganisms. The organisms and main purpose of genetic modification are outlined in Table 1. These figures show that most of the research activity is with respect to the release of GM plants, in particular oilseed rape (Brassica napus), sugar beet (Beta vulgaris) and potatoes (Solanum tuberosum) genetically modified for herbicide tolerance, pest resistance, hybrid development and/or quality improvement which include modification of storage compounds.

Two consents for the marketing of GM plants have also been issued by the DOE; hybrid oilseed rape with tolerance to glufosinate ammonium herbicides and soya beans with tolerance to glyphosate herbicides. One other marketing consent for a GM plant has been issued in France for tobacco with tolerance to oxynil herbicides. As consents for marketing of GMOs apply in all EC Member States, the French consent for the tobacco is valid in the UK.

In accordance with the regulations, the main requirement for an applicant is to supply specified technical information about the proposed release and to make a statement evaluating the impacts and risks posed to human health and the environment by the release. Preparation of such a statement requires for the risks to be properly addressed and supported by the risk assessment which should indicate that harm to the environment and harm to human health and safety is prevented or minimised.

RISK ASSESSMENT FOR THE RELEASE OF GM PLANTS

There are a number of possible approaches to risk assessment, but in general, any approach should comprise of an interrogative system designed to identify cases of GMO releases where there may be reasons for concern.

The approach recommended in the UK consists of several steps (DOE/ACRE Guidance Note 1, 1993). The first step is to identify any hazards and whether they can be realised in the receiving environment. This is followed by an estimate for the magnitude of harm which may be caused by realisation of the hazard.

Hazards are characteristics of an organism which have the potential for harm, or potential to cause an adverse effect. It does not necessarily mean that hazards are always harmful. For example, transfer of an inserted gene from a plant to a wild relative can be considered as a potential hazard. However, that hazard may not be realised in the absence of wild relatives or if it is realised, it might not result in significant harm. It is the consequence of the realisation, or rather the magnitude of the realisation, of the hazard that determines harm. For example the expressed product of an inserted gene in a plant might result in a reduction of a microbial population in soil by 10% constituting an insignificant consequence, whereas a 10% death rate in a protected bird species would be of greater consequence. Therefore, the level of harm could be dependent upon changes in the numbers of one or more species of other organisms, including beneficial and endangered species and the time over which effects might be expected including the reversibility of any impacts on the ecosystem.

The other estimate is of the probability or likelihood that harm could result. It may be that the hazardous consequence could be negligible but there is a high probability of that being manifested; in such a case the risk is likely to be low. The qualitative estimate of risk can be obtained by combining the consequence with the likelihood of the hazard being realised. It may be possible to get a quantitative measure for the likelihood or probability, but because of the complexity of the ecosystems and living organisms, it is very difficult to get anything other than a qualitative measure for the consequence at present.

Once the risks have been estimated, the assessment should consider whether any management procedures are required to control the identified risks so that harm is prevented or minimised. For example, if there is a risk of harm to the environment associated with the spread of the inserted gene through pollen dispersal, then pollen compatible species in the dispersal area could be removed to obtain biological isolation. Alternately, measures such as preventing flower development, removal of wild relatives from the dispersal area, crop rotations, agronomic practices and/or appropriate monitoring could be adopted as measures for risk management.

THE GMO RESEARCH PROGRAMME

The main purpose of the research programme is to provide information on the hazards and risks to the environment when releasing GMOs in accordance with the legislative framework. Research focuses on issues relevant to the current and foreseeable situation in the UK, concentrating on plants and microorganisms. The outputs from the research are used to provide technical information for risk assessment and decision making on the release of GMOs, to develop guidance addressing particular aspects of risk assessment and monitoring, and to anticipate both policy and technical developments concerning GMOs.

In the late 1980's, it was not clear in which areas recombinant DNA technology would be widely applied. GM microorganisms were expected to be the early candidates for release and therefore research focused on detecting, identifying and monitoring dispersal of microorganisms in the environment. Although there have not been many releases of GM microorganisms (see Table 1), information from these initial research studies formed the basis of two guidance notes on risk assessment and monitoring of GM microorganisms and GM baculoviruses (DOE/ACRE Guidance Note 5, 1994; DOE/ACRE Guidance Note 6, 1995). The information from this early research has also been used in the formulation of various documents at the international level, for example in the development of European standards on sampling and monitoring of GMOs, and in the OECD initiatives on mutual acceptance of data for the release of GMOs.

By the early 1990's, and as indicated in Table 1, it became clear that the majority of applications for consent were for the release of GM plants. Therefore a change in focus was required for the research programme to consider hazard and risks of GM plants. In particular, the research has focused on release of GM oilseed rape and GM sugar beet which are increasingly modified for herbicide tolerance and pest resistance. For GM potatoes, most releases are for improvements in quality through modification of carbohydrate metabolism and these are considered to be of low risk.

The two most important potential hazards for the release of GM plants are the capacity of the plants to survive, establish and spread, and the potential for gene transfer to wild relatives. Other potential hazards include the effect of the expressed products of the inserted gene, phenotypic and genetic stability, pathogenicity to other organisms, effects upon non-target organisms, and other potential adverse effects including any harm to human health.

In line with this, research studies on the release of GM plants have been commissioned in the following areas with the aim of obtaining further information on risk assessments and to assist in making informed decisions on applications for consent:

"Genetically modified crops and their wild relatives" - This study considered the likelihood and consequence of plants escaping and surviving outside the agricultural environment as well as the likelihood of gene transfer from GM plants to wild relatives and where appropriate the consequence of such gene transfer (DOE Research Report 1, 1994).

"Rapid cycling Brassica as a model for risk assessment in relation to GM plants" - This study investigated the dispersal of pollen from oilseed rape to wild relatives using rapid cycling Brassica as a model rather than GM plants. It showed that inter-specific hybrids occurred at a frequency of approximately 4% in intimate mixtures and that pollen could be dispersed up to 200m from a large source of oilseed rape.

"Gene flow in natural populations of Brassica and Beta" - This study considered gene transfer through pollen dispersal and found that isolation for large scale releases was not practical for oilseed rape and sugar beet because of wild relatives and feral populations. It recommended that risk assessment should focus on the consequence of such gene transfer (DOE Research Report 8, 1996).

"Selectable markers in GMOs" - This study reviewed the use of selectable markers and identified potential hazards such as spread of antibiotic resistance. It considered alternatives and suggested use of markers such as for sugar utilisation or light emission (DOE Research Report 9, 1996).

"Generic answers to safety questions on GM plants" - This study considered the feasibility for generic answers to the technical information requirements for consent applications to release GM plants. It showed that generic answers could be developed for certain questions on the recipient and the GM plant.

"Investigations of feral oilseed rape populations" - This study is investigating the weediness of oilseed rape and the composition and distribution of feral oilseed rape populations, and to identify traits likely to influence spread of oilseed rape beyond the agricultural environment (to be completed).

"Gene flow between forage crops and wild relatives" - This study is determining the hazards and risks associated with GM forage grasses including survivability and gene transfer by pollen dispersal with a view to developing confinement strategies for risk management (to be completed).

"Environmental impact of GM plants" - This study is investigating the environmental impact by considering the consequence of gene transfer between crops and their feral populations and wild relatives in particular with respect to pest resistance in oilseed rape and sugar beet (to be completed).

"Monitoring releases of GM crops" - This study is carrying out monitoring of large scale commercially released GM plants with the aim of confirming the risk assessment and making recommendations on the feasibility and effectiveness of such monitoring (to be completed).

The programme has also commissioned research on risk assessment systems (DOE Research Report 2, 1994) and risk assessment of bacteria (DOE Research Reports 5, 6 and 7, 1996), viruses and plant viral inserts, dispersal of microorganisms (DOE Research Report 4, 1995), nematodes, microbial biopesticides, insects and fish (DOE Research Report 3, 1994). These preliminary studies tend to be literature reviews to produce reports on the current state of play on a particular topic so that the policy can anticipate future developments and research can be commissioned where necessary.

In addition to the publications, the other key deliverables from the research has included the development of guidance notes on the risk assessment and monitoring of GM plants (DOE/ACRE Guidance Note 4, 1994). Also, the findings from the research have considerably increased our understanding of GMOs. It has contributed to the steadily growing body of evidence to show that some of the risks initially envisaged to be posed by genetic modification were misperception based on lack of scientific evidence. The majority of the existing GM plant releases pose low risk; and where risks tend to be greater, management can be applied to reduce the level of risk.

Future research will continue to investigate whether GMOs of various kinds affect the environment. In particular, studies should further investigate the consequence of gene transfer and the possibility of risk ranking of GM plants as well as the environmental impact of gene stacking and development of multiple tolerance. Studies should also consider hazard identification and risk assessment of unusual and non-arable types of GM plants such as perennials, trees and plants modified for production of novel compounds.

DEVELOPMENT OF POLICY

The outputs from the research have played an important role in underpinning and development of policy. In particular, the regulatory framework for the release of GM plants has been streamlined without compromising safety.

For example, research established which plant species did not have any wild relatives or did not hybridise with wild relatives, or had very low pollen and seed dispersal, or were susceptible to low temperature and were therefore unlikely to survive in the environment. Certain traits which were not likely to cause an adverse effect were also identified. From this information, fast track procedures with generic advice were developed for certain low hazard and low risk release of GM plants (DOE/ACRE Guidance Note 2, 1994). This has meant that about 75% of the applications are processed and issued with a consent within 30 days instead of the 90 day statutory time limit.

The information obtained from the research also helped identify the safety questions specific for the release of GM plants (DOE/ACRE Guidance Note 3, 1994) thereby reducing the information requirements for the applications for consent to release GM plants. This is no longer applicable following the amendments of the regulations in 1995 which introduced specific information requirements for the release of GM plants.

The research has also contributed to the development of simplified procedures in the EC for the release of GM plants allowing for a single application for consent to include a programme of work involving several releases of different types of GM plants from a single recipient species at several locations over a specified time period.

These changes have arisen mainly as a consequence of three factors: first, a systematic approach to risk assessment has been introduced for the release of GM plants; second, research activity has been targeted specifically at understanding the risks posed by GM plants; and third, the experience gained through a large number of confined releases has built confidence in the risk management procedures.

There is one fundamental aspect of risk assessment which deserves further consideration, that is risk evaluation. It poses an important question: "what level of risk is reasonable?" It is difficult to address this convincingly because of the differences in perception of risk and the balance between risks and benefits. In most cases, the answer is likely to depend upon the particular circumstances. In future, with increasing knowledge, experience and understanding, it may be possible to judge these complex issues generically.

REFERENCES

DOE/ACRE Guidance Note 1 (1993) The regulation and control of the deliberate release of genetically modified organisms.

DOE/ACRE Guidance Note 2 (1994) Fast track procedures for certain GMO releases.

DOE/ACRE Guidance Note 3 (1994) Reduced information requirements for releases of plants under the GMO regulations.

DOE/ACRE Guidance Note 4 (1994) Guidance for experimental releases of genetically modified plants.

DOE/ACRE Guidance Note 5 (1994) Guidance for experimental releases of genetically modified microorganisms (excluding viruses and similar agents).

DOE/ACRE Guidance Note 6 (1995) Guidance for experimental releases of genetically modified baculoviruses.

DOE Research Report 1 (1994) Genetically modified crops and their wild relatives - a UK perspective.

DOE Research Report 2 (1994) An evaluation of GENHAZ as a risk assessment system for proposals to release genetically modified organisms into the environment.

DOE Research Report 3 (1994) Genetic modification of fish - a UK perspective.

DOE Research Report 4 (1995) Dispersal of genetically modified microorganisms by abiotic factors and soil invertebrates.

DOE Research Report 5 (1996) Risk assessment and the release of genetically modified microorganisms into the phylloplane.

DOE Research Report 6 (1996) Risk assessment and the release of genetically modified microorganisms into the rhizosphere.

DOE Research Report 7 (1996) Risk assessment and the release of genetically modified microorganisms into the environment.

DOE Research Report 8 (1996) Gene flow in natural populations of Brassica and Beta.

DOE Research Report 9 (1996) Selectable marker genes in genetically modified organisms.

Table 1. Consents issued in the UK for the release of GMOs between 1/2/1993 to 31/5/1996.

GMO	Herbicide tolr*	Virus rest	Nematode rest	Fungal rest	Insect rest	Hybrid devl	Quality impr	Selectable mark	Total
Oilseed rape	16+	0-	0-	1-	0-	15+	9-	0+	41
Swede	1-	0-	0-	0-	0-	0-	0-	0+	1
Potato	0+	4-	1+	0-	3-	0-	12-	0+	20
Sugar beet	13+	1-	0-	0-	0-	0-	1-	0+	15
Maize	3+	0-	0-	1-	0-	0-	0-	0+	4
Wheat	1+	0-	0-	1-	0-	0-	0+	2+	4
Tobacco	0-	0-	0-	0-	0-	1-	0-	3+	4
Chicory	0+	0-	0-	0-	0-	2-	0-	0+	2
Eucalyptus	1-	0-	0-	0-	0-	0-	0-	1+	2
Poplar	0-	0-	0-	0-	0-	0-	1-	0+	1
Apple	0-	0-	0-	0+	1-	0-	0-	0+	1
Strawberry	0-	0-	0-	0-	1-	0-	0-	0+	1
Baculovirus									2
Pseudomonad									2
Rhizobium									2
Bacteriophage									1
Total	35	5	1	3	5	18	23	6	103

*Main purpose of genetic modifications and number of consents (+/- additional traits)