Leaking From The Lab

6. GMMS IN THE ENVIRONMENT

According to officials in the HSE, when waste is said to be inactivated, this does not mean that all organisms will be killed

There is no explicit legal requirement to kill all GMMs to be disposed of in waste if they have been deemed ‘safe’ and are considered to have a limited ability to survive in the environment (Group I organisms). The HSE have said that, in their experience, large scale facilities routinely inactivate all GMMs prior to disposal. However, according to officials in the HSE, when waste is said to be inactivated, this does not mean that all organisms must be killed although the majority should be. The waste from research facilities may be heat treated in autoclaves or chemically treated before disposal by methods such as incineration, but there is no independent verification that this is undertaken. Therefore, Group I organisms will be entering the environment in waste from both industrial and research facilities using these organisms.

In their applications to the HSE to register uses of GMMs, companies have acknowledged that releases will take place even if cultures are claimed to have been inactivated:

In their 1993 notification of large scale use, Zeneca BioProducts at Billingham expected releases of 10^4-10^6 organisms per millilitre of a GM E.coli K12 producing the enzyme, xylanase. According to their environmental risk assessment, the GMMs were expected to be released into the following sites: “Terrestrial, research and production site. Water drainage/sewage system”.

In another 1993 notification of large scale use, Zeneca BioProducts at Billingham expected treatments to reduce levels in waste to around 100 organisms per millilitre of a GM yeast, S. cerevisae, which produces human serum albumin.

In a 1994 notification of large scale use, SmithKline Beecham Pharmaceuticals in Irvine, Ayrshire, acknowledged that GM Penicillium chrysogenum was ‘rarely’ released via air and effluent. Air sampling is conducted weekly.

In various notifications, SmithKline Beecham Pharmaceuticals at Worthing, West Sussex, claim the likelihood of release of GMMs used at the site will be ‘low’ as they are killed prior to disposal by a method they decline to disclose because of commercial confidentiality.

In commercial situations, the economic value of GMMs gives an incentive to limit leakage through the process plant. GMMs are grown in large vessels called ‘fermenters’ and the end product extracted from them. Following this, the remaining GMM waste is treated in a kill tank to inactivate organisms before discharge. Quality control, especially in pharmaceutical production, will help limit escape at all stages and so will the desire to protect the intellectual property of the company – they do not want competitors to have access to their organisms. This has led to an improvement in the containment practices compared to the production of traditional fermentation products, such as brewing or vinegar production, which are often carried out in large open vessels⁶⁰. However, leaks, especially in the form of aerosols (droplets in the air) but also as fluids, can occur as a result of:

• 
  structural damage or failure of a seal in the fermentation vessel or pipe work – aerosol leaks;
  
• 
  overpressure in fermentation tank leading to large scale release caused by failure of a safety device – aerosol leaks;
  
• 
  leakage during inoculation, addition to or sampling of the fermentation tank – aerosol leaks;
  
• 
  handling of the product after fermentation – aerosol leaks;
  
• 
  leakage during processing post fermentation, e.g. during centrifuging or filtration – aerosol leaks;
  

• 
  effluent disposal following ‘kill tank’ treatment – fluid discharge.
  

The numbers of organisms involved in a release, either in an aerosol or fluid discharge, will not be trivial in the industrial setting

At the present time, there is no independent monitoring of containment and releases of GMMs by the HSE or any other authority

Fermenters are sealed and pressure tested, normally operate at comparatively low pressures and are fitted with bursting discs designed to fail if the vessel is significantly over pressure. Fermenters rarely fail but it can happen⁶¹. Kill tanks, where organisms are ‘inactivated’ before release to waste by either heat or chemical treatment are never 100% effective in killing organisms and if this system fails it could lead to huge numbers of live organisms being released in effluent.

The numbers of organisms involved in a release, either in an aerosol or fluid discharge, will not be trivial in the industrial setting. Fermenters can range in size from 10 to 10,000 litres in capacity with up to 10¹⁴ or 10¹⁶ organisms in larger fermenters. Even the release of 1% or 0.1% will involve many millions of organisms.
In Denmark, which is home to one of the largest enzyme producers using GMMs, Novo Nordisk, rather than obscuring the fact that the release of GMMs takes place, a different approach is taken. The Danish regulations accept that, when used on a large scale, some GMMs can be released so companies are given release limits which they must monitor and reportError: Reference source not found.

6.1 Monitoring for Releases

At the present time there is no independent monitoring of containment and releases of GMMs by the HSE or any other authority. Any monitoring that takes place is undertaken by the user and the results are not available to the public.

In 1996, the HSE sent a questionnaire to 49 companies using GMMsError: Reference source not found. The companies were asked what sampling methods they used to monitor process organisms inside and outside the workplace. Responses were received from 33 companies (see Table 5). Less than half (14 out of 33) of those who responded carried out regular monitoring. All seven in production scale use monitored, most but not all (9 out of 11) pilot scale plants carried out monitoring, but less than half (10 out of 28) of the laboratory scale operations included any monitoring at all. These smaller scale users tended to rely on ‘good microbiological practice’ to ensure absolute containment. All but one of the companies that did not monitor were small scale operations in university or research council laboratories.
The majority of the monitoring took place inside the workplace with only five companies monitoring outside as well as inside the workplace.



The aim should be to obtain results very rapidly because if there is a leak, it needs to be identified and stopped immediately

Table 5: HSE Questionnaire Results (from Crook and Cottam (1996)) Error: Reference source not found

QUESTION	GMO	NON GMO	TOTAL
Nature of the process organism used at the time the questionnaire was administered	24	9	33
Companies at which regular workplace monitoring was taking place	12	2	14
Companies at which monitoring was being done both within and outside the workplace	4	1	5

There are no standard methods for monitoring micro-organisms and no guidelines from the HSE as to appropriate methods to use.

GeneWatch wrote to all those facilities listed as using GMMs on a large scale to ask about their monitoring techniques and frequencies. None of them supplied any details of their monitoring.

6.2 Where Monitoring is Necessary and the Difficulties Involved

Designing and developing systems which are reliable and effective is not easy but is essential if the assumptions behind risk assessments are to be tested and compliance with containment is to be determined.

6.2.1 Where to Monitor
Monitoring must be comprehensive if it is to detect leaks and monitor routine discharges. The sites where detection methods are needed include:

Within the factory or research laboratory for the early detection of leaks in equipment, thus preventing damage to employees’ health, ensuring no economic loss and minimising the risk of environmental release.

At factory or laboratory outlets to monitor continually, if possible, the numbers of GMMs released into the environment. If monitoring here is accurate, it can provide a measure of how many live organisms are released into the environment. This is particularly necessary at air and gas release outlets, because once organisms have moved into the wider atmospheric environment they will be massively diluted.

External environmental monitoring: a) to detect any surviving GMMs, b) to detect horizontal gene transfer, c) to look for any effects the released organisms might have.

In principle, the aim should be to obtain results very rapidly because if there is a leak it needs to be identified and stopped immediately.

6.2.2 Difficulties in Monitoring




It may only be through evidence of their impact (such as an outbreak of human, animal or plant disease) that GM viruses will be detected

The main method used to detect GMMs is by culture. However, viruses are especially difficult to identify as they require cell culture techniques for their isolation. These are both time consuming (weeks not days) and technically demanding. In reality, there are no practicable ways of monitoring for GM viruses yet developed. It may only be through evidence of their impact (such as an outbreak of human, animal or plant disease) that they will be detected.

For bacteria, the situation is somewhat easier. Culture techniques are routine and take 2-3 days, but because many of the cells have been disabled, traditional cultivation techniques are not applicable. In addition, when exposed to environmental stress, many bacteria may remain viable but become non-culturable (VNC) and cannot be counted on agar plates. These methods also do not detect ‘naked’ DNA. Furthermore, effluent leaving a plant or laboratory will contain a mixture of killed and living organisms and DNA, making monitoring for live cells difficult.
As well as issues related to the organism itself, the environment in which the organism is found can also cause difficulties. Monitoring in the environment is more difficult than in the process plant or laboratory, and air, water and soil each have their own specific problems. Monitoring of air requires incredibly sensitive techniques due to the massive dilution that takes place. However, compared with other media, air contains a relatively small variety of micro-organisms. In water, organisms can be hugely diluted and difficult to identify. In soil, there is less chance of dilution, but the number of other micro-organisms is massive (approximately 10⁹ g^-1) and the ecology of soils very complex.

6.2.3 Methodologies
Despite the difficulties involved in monitoring, these are not insurmountable. There are various methods which can be used in different circumstances. None are ideal for every situation and a combination will be needed. In Denmark, Novo Nordisk carry out their external monitoring by the use of selective media, and then colonies with the same phenotype are isolated and PCR used to identify the modified strain.
Marker genes – these are inserted into the GMM in order to give it an easily distinguishable feature. Antibiotic resistance is not useful in this context because many soil micro-organisms often have resistance to at least one antibiotic⁶². Other approaches include the incorporation of genes which code for bioluminescent or fluorescent molecules which can then be measured, or the inclusion of genes for enzymes which catalyse a colour change. There are drawbacks to these approaches. For example, the production of luminescence can be reduced by environmental stress and, in aquatic environments, the presence of indigenous luminescent bacteria will distort counts.
Selective media – because microorganisms have different nutritional and environmental requirements for growth, special media can be used to limit the range of micro-organisms grown on it. It can be sensitive and simple but the organism must be viable and culturable. Results can take up to 4 days which means this method is not appropriate for continuous monitoring or to check for accidental releases from a process plant. Other similar organisms will also grow and this becomes a problem in identifying micro-organisms which are present in low quantities.



There is clearly an urgent need to develop methods which are effective, rapid and accurate in identifying GMMs….

…Until such tests are developed, it is difficult to understand how releases of GMMs can be justified

DNA probes – perhaps the most specific method of detecting both GMMs and naked DNA is through the use of DNA probes. These ‘probes’ recognise and bind to the foreign DNA inserted in the organism, are very specific and, if combined with techniques such as PCR, can detect low levels. Because they do not indicate if an organism is viable they may need to be combined with other methods.
Immunological techniques – using antibody detection systems and radio-immuno assay. These do not detect whether organisms are living but are sensitive to specific changes in the GMM and so can assist in their identification against a background of other organisms.
Sampling – different sampling techniques are needed according to the medium and organisms to be isolated and the identification method used. A full review is beyond the scope of this report, but techniques include sampling on plates, collection for other forms of detection and more general methods such as aerosol detection systems to alert for leaks inside the workplace.
There is clearly an urgent need to develop methods which are effective, rapid and accurate in identifying GMMs. Military interest in biological weapons is leading to systems which can rapidly identify biological agents in air. The task for users of GMMs should be much easier as they know which organisms to look for, and knowledge derived from military research could advance civilian detection systems enormously. Until now, there has been little incentive to develop monitoring methods as there are no legally specified requirements. Demands for the detection of GM ingredients in food over the past six months have led to the rapid emergence of accurate and sensitive tests. If such pressures were brought to bear over the need to detect GMMs, tests are likely to emerge just as rapidly. Until such tests are developed, it is difficult to understand how releases of GMMs can be justified.