Leaking From The Lab

2. INTRODUCTION

The genetic engineering of crops and foods has become a controversial issue over recent years and public awareness is high. However, genetic engineering is also being used in other areas, some of which have received much less attention. One of these is the use of genetically modified micro-organisms (GMMs), such as bacteria, yeasts, fungi and viruses, both in public and private research laboratories and in commercial production facilities. This use is referred to as ‘contained use’ to distinguish it from other uses (in agricultural crop production, for example) where the genetically modified organism (GMO) is released deliberately into the environment.

GMMs are being discharged into the environment either accidentally or incidentally

Micro-organisms were the first organisms to be genetically engineered. In the early 1970s, key scientific developments allowed the function of individual genes to be identified; genes to be cut out from the genome using molecular ‘scissors’ called restriction enzymes; genes to be copied (cloned); and the transfer of ‘foreign’ DNA into bacteria, using vectors such as phages (infectious agents of bacteria) and mobile loops of bacterial DNA (plasmids) to transfer DNA. Together, these techniques form the basis of recombining genetic material from different species - so-called recombinant DNA technology or genetic engineering.

The scientists conducting the ground breaking experiments in the early 1970s were concerned about the potential for harmful impacts that might arise, such as the potential to create new pathogens. In 1975, the Asilomar conference in the USA and earlier deliberations of expert committees led scientists to introduce a voluntary moratorium on some laboratory experiments with genetically engineered micro-organisms until guidelines and regulations on their use were put in place. In the USA these took the form of guidelines, whereas in the UK voluntary controls were replaced by statutory regulations in 1978¹.
Since that time, the use of GMMs has become widespread both in university and industrial research laboratories and commercially to produce a wide array of enzymes (particularly for use in food processing and detergents) and drugs such as human insulin. GMMs are certainly being discharged into the environment either accidentally or incidentally through the breakdown of containment facilities or through routine discharges if the GMM is deemed ‘safe’. Although the products of GMMs, such as drugs and enzymes for use in detergents, tend to be viewed with less hostility than some other products of genetic engineering, the impact of the living organism is of concern.
This report reviews the potential environmental and health risks of the escape of GMMs from both research and commercial facilities. The present regulations are described together with a description of GeneWatch’s findings about how GMMs are being used and monitored in the UK. The European Directive intended to ensure the safe ‘contained’ use of GMMs (the Contained Use Directive, 90/219/EEC) has recently been revised and the UK has just (May 1999) published its plans to implement it. Therefore, this an important time to review the current status of GMMs in the UK, the risks involved and how these could be best avoided.

3. RISKS OF GENETICALLY MODIFIED MICRO-ORGANISMS

Escaped GMMs could cause illness, disrupt natural microbial ecosystems and alter other species in unpredictable ways

GMMs could cause harm in several ways. Firstly, if they are pathogenic (able to cause disease) in humans or animals, they could cause illness in the people working with them or more widely if they escape from the laboratory. Secondly, they could survive in the environment and disrupt natural microbial ecosystems. If they continued to produce a certain product (such as an enzyme or antibiotic), they could be directly damaging to organisms. Thirdly, the foreign DNA could move into other species, altering them in unpredictable ways. Because DNA from dead cells can be taken up into living cells, even so-called ‘naked’ DNA (DNA which is not contained in a cell) has the potential to have effects.

Table 1 summarises the questions that are thought relevant to the assessment of the effects of releases of GMMs into the environment. Three of the most important questions in determining the environmental effects of a release of a GMM are its characteristics, whether it is likely to survive outside the laboratory or factory environment, and whether foreign genetic material can be transferred to other organisms. If an organism can survive and/or transfer genetic material, questions arise about the implications of this.

Table 1: Data requirements to predict the effect of the release of a GMM to the environment (adapted from Doyle et al (1995)^2)

QUESTION	FACTORS AFFECTING OUTCOME
Survival of GMM – establishment and multiplication.	Nature of organism, such as its ability to cause disease.
	Effect and nature of the genetic modification – does it give a competitive advantage?
	Scale and frequency of release.
	Receiving environment – including biological and physical characteristics.
Dispersal of GMM to other sites.	Characteristics of environment.
Transfer of foreign genetic material to other organisms.	Survival, establishment and multiplication of organisms.
Ecological impacts of GMM and foreign DNA.	Interactions with other organisms and effect of product(s) of GMM.
Potential for containment, decontamination and mitigation if adverse effects detected.	Nature of receiving environment and scale of effects.

3.1 Survival of GMMs in the Environment

A great range of organisms have been genetically modified, including viruses, bacteria and yeasts. Some of this work involves organisms able to cause disease in humans, animals or plants. Other work uses organisms which, in their natural state, are not harmful.

In theory, many of the GMMs used in contained facilities have lost their ability to survive in the natural environment….

….yet there is evidence that ‘disabled’ organisms can survive outside the laboratory

In theory, many of the GMMs used in contained facilities have either been bred in laboratories over many generations and lost their ability to survive in the natural environment or have had specific sequences inserted or deleted to reduce their ability to survive. For example, the Bacillus organisms used by the Danish enzyme company, Novo Nordisk, to produce protease and amylase have had genes removed making them asporogenous, so only one cell in 10 million is able to form a spore. Spore forming ability, when an organism develops a protective coat and can survive longer in the environment, is an important characteristic of the organism.

Eschericia coli (E.coli) K12 is another of the most commonly used bacteria in research and is a disabled strain which it is assumed cannot survive outside the laboratory. The E.coli K12 strain has probably been engineered and manipulated by human beings more than any other strain of bacteria. It was originally isolated in 1922 from the faeces of a diphtheria patient at Stanford Medical School and has been maintained under laboratory conditions since then. Other commonly used species include the bacteria Bacillus sp; Streptomyces sp; Kluyveromyces sp; Trichoderma sp; Klebsiella sp; the yeast, Saccharomyces cerevisiae and the fungus, Apsergillus niger.
Particular attention has been given to the ability of E.coli K12 to survive and colonise because some strains of E.coli can be pathogenic and cause intestinal disease. There are a great number of K12 derivatives with various mutations which should make them unlikely to survive or compete well. However, there is evidence that these disabled organisms can survive outside the laboratory, although the length of survival depends on a variety of factors related to the organism and the environment. In the intestines of experimental animals, various strains of E.coli K12 which had been genetically engineered to produce bovine somatotrophin (BST) or human growth hormone (HGH) and were resistant to one or more antibiotics survived for up to 7-14 days but did not appear to colonise the intestine even in the presence of selective pressure in the form of the relevant antibiotic^3,4,5,6. Similarly, various other strains of E.coli K12 survived for around 4-6 days in the human intestine but did not colonise longer term^7,8.
Although E.coli is an organism which is normally found in the intestines of animals, it can survive in the wider environment. The Health and Safety Executive (HSE) guidelines on risk assessment state that E.coli K12 can survive for 7 days in external environments⁹. However, other research indicates this may be an underestimate in some circumstances although there is great variation between studies, probably related to differing experimental conditions. For example, Tschäpe¹⁰ showed that E.coli K12 could survive in a small sludge unit - although the E.coli could not be detected for 12 days, it eventually ‘reappeared’ having acquired an additional plasmid which appeared to confer no competitive advantage. Other research has shown that a genetically engineered E.coli K12 strain survived for at least 35 days in a non-sterile silt loam soil¹¹. In contrast, in other studies, a BST strain of E.coli K12 was eliminated from sewage sludge over 5-6 days following a single, high dose innoculum¹².
E.coli K12 can also survive in river and sea water for periods of well over 2 months if the water is sterile but only for periods of about 2-18 days if the water is untreatedError: Reference source not found^,13. This is thought to be due to competition with other organisms in non-sterile conditions. Survival times are much longer at lower temperatures.




GMMs may not only survive in water, soil or air, they may also be ingested by invertebrates which could affect their survival and their distribution

There is less published information on the survivability of many of the other strains used in genetic modification experiments although some organisms used in laboratory work are quite robust. For example, Pseudomonas putida UWC1 survived for 8 weeks in a sewage activated sludge unit¹⁴ although other strains may have shorter or longer survival periods.

GMMs may not only survive in water, soil or air, they may also be ingested by invertebrates which could affect their survival and their distribution. This is an issue which has only recently been addressed and experiments have shown that a genetically modified Pseudomonas fluorescens can survive and multiply in the intestines of the earthworm, Octolasion cyaneum¹⁵, and the woodlouse, Porcellino scaber¹⁶. Because these organisms are consumed by others, GMMs and DNA could move through the food web.
Laboratory techniques may not be able to identify all living organisms in the environment so those experiments which have been done may underestimate survival rates. Some organisms enter what is referred to as a ‘viable, non-culturable’ (VNC) condition^17,18. That is, although an organism may not grow on the culture media used in laboratories, it may still be alive and able to multiply in the correct environmental conditions. This possibility was identified because there are often differences between visual counts of bacteria (based on their ability to take up certain stains, which is thought to indicate metabolic activity) and numbers isolated by culture. Numbers cultured tend to be lower than those considered viable by staining techniques, leading to the hypothesis that bacteria may enter a dormant phase which conceals their viability when cultured on artificial media. This has been challenged on the grounds that staining may not provide an accurate indication of viability¹⁹, but the large amount of literature demonstrating VNC for such a large number of species suggests it is not a spurious observation.
Knowledge of disease transmission has shown that viruses can survive in air and be transported over long distances, a characteristic which is very important in the spread of some viral diseases such as foot and mouth disease. Whether a virus can survive in air depends on its own characteristics, such as coat lipid content, and the physical conditions of the air such as humidity. Viruses are also spread in the environment via faeces or other discharges from infected animals. However, because viruses are much more difficult to isolate than bacteria (see Section 6.3.2) there is much less information about their persistence in the environment.
It is clear, therefore, that even disabled organisms have the potential to survive for many days or weeks in the environment. Because of the VNC condition, it may be difficult to determine survival rates of micro-organisms with confidence. In addition, the potential exists for GMMs to move through the food web if they are ingested by organisms which may, in some cases, improve their likelihood of survival.
For organisms which are known to cause disease, even though they may not survive for long periods, they could still cause harm in the short term if they escape confinement and encounter a susceptible person, animal or plant.

3.2 The Transfer of Genetic Material

Escaped GMMs could either pass their foreign genetic material to other organisms or else acquire the ability to become established from others

Even if GMMs do not become established in the environment in the long term, it is possible that they could either pass their foreign genetic material to other organisms or else acquire the ability to become established from others. This movement of genetic material between organisms is known as ‘horizontal transfer’ to differentiate it from the vertical transfer between one generation and the next. Over the past twenty years, there has been a burgeoning literature about gene transfer between micro-organisms leaving the impression - reinforced by the way in which antibiotic resistance has spread between bacterial species - that it is an extremely important and influential process.

There are three mechanisms by which horizontal gene transfer is thought to take place:

Transformation: The uptake of free ('naked') DNA from the environment and its incorporation into the bacterial genome.

Conjugation: Movement of DNA between bacteria following cell-to-cell contact and effected by plasmids or transposons.

Transduction: The transfer of genetic material from one bacterium to another by a bacteriophage (an infective virus of bacteria).

3.2.1 Transformation
The process of ‘natural genetic transformation’ is restricted to bacteria and involves the uptake of naked DNA (of chromosmal or plasmid origin). For transformation to take place, there must be free DNA in the environment which can be taken up by bacteria and bacteria must be able to take up DNA – a state which is known as ‘competence’²⁰. It has been known for a number of years that extracellular DNA exists in the environment, most of which is of microbial origin. This DNA is released when cells die and start to degrade, but can also be excreted at other times such as during cell growth and during spore germination.
Competence, when bacteria can bind extracellular DNA and take it in, is not present in all species of bacteria and even in those species which show competence it may vary according to environmental conditions. For example, in Neisseria gonorrhoea competence is persistent, in Haemopillus influenzae it can be induced under conditions which inhibit growth, and in E.coli competence is difficult to induce often requiring laboratory techniques such as electroporation²¹.
DNA is broken down at high rates when initially introduced into waste water, seawater, freshwater sediments and soilsError: Reference source not found. There is evidence that this degradation is caused by a mixture of micro-organisms producing DNase (an enzyme that degrades DNA)²². Despite the high level of DNases found in a whole range of environmental samples, extracellular DNA has been found consistently in a variety of habitats. This is partly because DNA is produced continually by micro-organisms, but also because in some circumstances DNA can avoid being degraded.



The inability of an organism to survive does not mean that its genetic material could not be transferred to other species


Plasmids are the most widely used tool to introduce new DNA sequences artificially into micro-organisms…

Extracellular DNA has been associated with cellular slime which it is believed may stabilise the DNA structure – up to 40% of the dry matter of cellular slime can be DNAError: Reference source not found. DNA may also be protected through its ability to form complexes with various minerals such as clay, feldspar, heavy metals, and humic substances. Adsorption of DNA to sand or clay particles is thought to protect DNA against DNase activity and, although adsorption slows the process of transformation, uptake of adsorbed DNA does take place and does not require a desorbtion step before it can take place²³. There are a variety of factors which affect the rate and extent of this protective adsorption of DNA by minerals. For example, the type of mineral and its acidity or alkalinity (pH) both have a large effect, although binding can occur over a wide range of pH values. The shape and size of the DNA molecule and general temperature have a much lesser effectError: Reference source not found.

DNA may persist for considerable periods of time. For example, using PCR analysis, which can identify very small quantities of specific DNA, it was found that DNA from a genetically engineered E.coli K12 remained undegraded for at least 40 days in a silt loam soil²⁴.
The evidence from microcosm and other studies that transformation can take place both in aquatic and terrestrial environments involving both chromosomal and plasmid DNA^25,26,27 suggests that transformation may be a significant route of gene transfer between bacterial species. The frequencies of such transformation events may be low, making detection difficult, but the findings show that the inability of an organism to survive does not mean that its genetic material could not be transferred to other species.

3.2.2 Conjugation
Conjugation is the most studied form of gene transfer between bacteria. It involves DNA exchange following cell-to-cell contact and is mediated by some (but not all) plasmids and transposons. Plasmids are circular strands of extra chromosomal DNA and transposons are mobile genetic elements which are capable of integration into both chromosomal or plasmid DNA. Both plasmids and transposons can carry, and are thought to be responsible for, the widespread occurrence of antibiotic resistance genes and both are used in genetic modification techniques. Plasmids are the most widely used tool to introduce new DNA sequences artificially into micro-organisms, but in order for a plasmid to transfer genes between bacteria under natural conditions they require certain characteristics:

the ability to produce pili (thread-like structures which bind the two cells together) and enzymes necessary for replication and transport of DNA;

a sequence of DNA called Tra which will allow conjugative plasmids to move between one cell and another (to be transmissible);

…“there is no such thing as a safe plasmid”

In addition, plasmids have specific host ranges which may be narrow or broadError: Reference source not found so they may be able to transfer DNA between one or two species or across a whole range of unrelated species.

One of the most important safety mechanisms in the production of GMMs is the use of plasmids which are deficient in one or all of these transfer mechanisms and have a restricted host range. However, although such precautions will reduce the risk of transfer, it is possible for the plasmids in such a GMM to acquire the ability to undergo conjugation. For example, E.coli cells containing a non-conjugative recombinant plasmid have been shown to be capable of receiving a conjugative plasmid from another E.coli strain²⁸. If the recombinant plasmid contained the Mob sequence, it was then capable of transferring itself into a third E.coli strain by utilising the structures and enzymatic properties of the conjugative plasmid. Similarly, non-conjugative plasmids in P. putida in activated sludge units acquired the ability to conjugate in the presence of other bacteria. Bacteria isolated from waste water were able to mobilise a recombinant non-conjugative plasmid from E.coli K12²⁹. About 50% of E.coli strains from human volunteers were able to promote the transfer of a recombinant non-conjugative plasmid from E.coli K12³⁰. However, the rate of this transfer was low and the resulting organisms did not colonise the intestinal tract of mice.
In addition, it has been shown that some conjugative transposons can also transfer into plasmids and facilitate mobilisation³¹ which has led to the observation that “there is no such thing as a safe plasmid” ³².
Recent research has shown that gene transfer between the laboratory strains, E.coli K12 and E.coli B can take place in the digestive vacuoles of a protozoan, Tetrahymena pyriformis³³. Such free-living protozoa are widespread in the environment, would normally ingest many released GMMs and many survive the digestive process. If an innocuous GMM was ingested at the same time as a pathogen or an organism that contains a plasmid that could restore conjugative properties to the plasmid of the GMM, the E.coli could acquire such genes.
Earthworms have been shown to increase the distribution through the soil of both a genetically modified P. fluorescens and soil organisms which acquired the plasmid it was carrying by conjugation³⁴.
There is considerable evidence, not least from the spread of antibiotic resistance, that conjugation is an extremely important mechanism for sharing genetic material between bacteria in natural systems. Complete confidence cannot be placed in the steps taken to limit gene transfer by conjugation.

3.2.3 Transduction
Transduction, mediated by phages, may only be important for the exchange of genetic material between closely related species, because phages have a limited host rangeError: Reference source not found.
Phages are infective agents (viruses) of bacteria which are able to pick up, carry and inject DNA into a new host. The DNA may then be integrated into the host genome or into a plasmid where it may persist. A phage could infect a GMM and transfer the foreign DNA to another organism.
Although there is evidence that a large number of phages exist in the environment, there are few data about the frequency of transduction in the wild and thus its significance for GMMs is difficult to assess.

Although safety mechanisms may be built into GMMs, they are by no means foolproof

3.3 The Effect of the Inserted DNA

Although safety mechanisms may be built into GMMs, they are by no means foolproof. The exact nature of the inserted foreign DNA will influence the impact any GMM has if it escapes confinement and particular areas of concern include:

The use of antibiotic resistance marker genes. This is very common practice as a way of identifying when a genetic modification has been successful. The release of GMMs with antibiotic resistance genes could exacerbate the present problems with antibiotic resistance in disease-causing organisms if they spread to other organisms. It is argued that such genes are ubiquitous in nature but the scale, sites and nature of any releases have the potential to increase the risk.

Gene transfers which could alter the host range an organism can infect or, if transferred to other organisms in the environment, could increase their pathogenicity. A single gene transferred from Yersinia pseudotuberculosis to E.coli K12 enabled it to invade mammalian cells in culture^35. Conceivably, so-called ‘pathogenicity islands’, which are regions of DNA that contain a variety of virulence genes^36, could be transferred.

The introduction of genes from vectors (the plasmids, transposons and phages used in genetic modification) which facilitate the transfer of DNA. There are mechanisms which act as obstacles to the transfer of foreign DNA. For example, restriction enzymes can recognise and cut up such DNA so that it is not incorporated. However, by using genes and gene sequences which can overcome these defences, there are fears that gene transfer could increase in frequency and make a harmful effect more likely to occur^37.

3.4 Evaluating the Impacts of GMMs in the Environment

The effects of any obviously pathogenic GMMs which are released into the environment are relatively easy to assess. However, the majority of organisms which are used are not overtly pathogenic and, although the GMM may persist in the environment and transfer foreign genetic material, predicting the impact of this is difficult. This is largely because so little is known about the ecology of micro-organisms in the environment. Nevertheless, the natural microbial flora are unlikely to be unimportant either in ecosystem or human health terms and their disturbance by GMMs could be significant. Issues which need to be addressed include:

Until our basic knowledge of microbial systems improves, ignorance dominates any risk assessment

• 
  The impact of vector systems developed to facilitate gene transfer in the laboratory. The presence of GMMs in the environment containing gene sequences from these systems may pose special risks by increasing the likelihood of gene transfer through overcoming natural barriers.
  
• 
  The impact of antibiotic resistance genes used as markers in GMMs. The presence of increased levels of antibiotic resistance genes could make the treatment of bacterial diseases more difficult if they were to be transferred to disease-causing organisms.
  
• 
  The impact that the products of some GMMs (such as enzymes and drugs) may have on the environment. If a GMM which was designed to produce a drug or enzyme survives in the environment, it may continue to produce the product. This chemical may have effects on other bacteria or other components of the ecosystem.
  

Until our basic knowledge of microbial systems improves, ignorance dominates any risk assessment.