Genetically modified organisms for the environment: What went wrong?
The concept of using genetically engineered bacteria for environmental release as agents for in situ bioremediation of industrial pollution can be traced to the very beginning of the recombinant DNA technology. As early as 1972, Ananda Chakrabarty, of the University of Illinois in Chicago, made global headlines in his attempt to patent a genetically modified Pseudomonas strain able to degrade a suite of petroleum components and thus holding a potential for dissipating oil spills (Cases and de Lorenzo, 2005). After ten years of litigation, the patent of the first man-manipulated live entity was granted, a seminal event that was to trigger a large number of consequences in many different realms e.g. scientific, legal, ethical, biosafey, biosecurity and social acceptance. In the meantime, the first usable tools for facilitating gene cloning were developed by Boyer and Cohen (Cohen et al., 1973) and the arch-famous Asilomar Conference took place (Berg et al., 1975a; 1975b). Although the patented Chakrabarty’s strain did not really fulfill its promise, the entire case brought about considerable hype on the potential that genetic engineering could have to endow bacteria with a superior capacity to eliminate pollutants in situ. One distinct aspect of such an endeavour is that bacteria tailored for environmental release must be vigorously active rather than attenuated (as was recommended in Asilomar). This posed a fascinating challenge for the genetic engineers of the time, as strains had to be programmed to do their catalytic mission efficaciously while at the same time being safe. The approach proposed by that time was the design of genetic containment and biological containment systems to programme death of the engineered agents once the environmental purpose for which they had been created had been fulfilled (Diaz et al., 1994; Molin et al., 1993; Ramos et al., 1995; Ronchel and Ramos, 2001).
GEMs for in situ catalysis, for biological control and for plant protection have been for nearly 20 years the workhorses in which these early concepts have been tested and their success and failures examined (Cases and de Lorenzo, 2005). The balance is extremely good in having expanded the knowledge base on microbial ecology and biodegradation biochemistry - but clearly disappointing in terms of efficacious applications in the field. Despite some early successes in the engineering of sophisticated GEMs able to consume otherwise recalcitrant compounds (Rojo et al., 1987; Ramos et al., 1987) the reality is that bioaugmentation (i.e. increasing removal of pollutants by inoculating the target sites with catalytic bacteria) is not yet a reliable technology. Alas, this applies not only to GEMs, but to virtually all types of micro-organisms, natural or recombinant, the few exceptions being less than five. One is Dehalococcoides, an anaerobic bacterium able to cause reductive dechlorination of many chloro-organic compounds when inoculated in polluted aquifers (Lovley, 2003). A second one is Geobacter (Amos et al., 2007), which has shown its ability to remediate uranium- contaminated groundwater (Lovley, 2003). The best strains to do the job in both cases occur naturally. Furthermore, many of the toughest recalcitrant molecules (e.g. highly chlorinated aromatics) can be dealt with only by anaerobic bacteria, which are most often not amenable to genetic modification. To finish the less-than-rosy picture for transgenic bacteria, conditional killing circuits were far from achieving a certainty of containment which was hoped for.
On this basis, it is surprising to still see in environmental biotechnology numerous reports that propose engineering this or that bacteria for biodegradation of a target compound for potential use in bioremediation. There is a big gap between the potential and realisation and, for the sake of the field, it is better to accept that basically all early expectations of solving pollution and many other environmental problems through genetic engineering have conspicuously failed (Cases and de Lorenzo, 2005; de Lorenzo, 2009). In contrast, the field has yielded some dividends in the production and application of whole-cell biosensors (Ron, 2007; Vollmer and Van Dyk, 2004; Garmendia et al., 2008; de Las Heras et al., 2008) some of them for in situ application for detection of underground chemicals, as well as bioadsorption an immobilization of heavy ions in engineered bacterial biomass (Valls et al., 2000). These are, however, minor victories in the midst of the debacle that has afflicted the pursuit of superbugs for combating pollution.