Possible reasons for past failures of genetic engineering techniques for creating novel, xenobiotics-degrading bacteria

It was fascinating to read the 2010 paper on the “Synthetic Bug Synthia” produced by the C. Venter¹ group, which triggered further stimulating discussions of novel perspectives in relation to the “Construction of the Super Bug” for bioremediation purposes.² Modified bacteria, which could bioremediate toxic compounds, such as PCBs, halonitroaromatics, fluoroaliphatics, and other persistent and recalcitrant toxic contaminants present in the biosphere, still represent a cost-effective and environmentally friendly alternative to physico-chemical elimination and/or detoxification.

More recent discussions have toyed with the idea of simply incorporating known catabolic sequences into a synthetic chromosome in order to mineralize or detoxify environmental pollutants.

However, these discussions were rapidly imbued with pessimism when it was found that the novel biosynthetic Mycoplasma mycoides species JCVI-syn1.0, with an original but synthetically engineered chromosome of a closely related species,¹ is only productively propagated in an artificial environment and, probably, could survive in the original ruminant host, although the latter possibility has yet be confirmed.

With regard to environmental applications involving techniques such as on-site, in situ, or even off-site bioremediation, highly robust microorganisms are required. These microorganisms are expected to survive over the long-term, perform very effectively, and be resistant under many adverse conditions (high salinity, risk of dehydration, high and low temperatures and/or pH values, etc.). Another important issue is the need for stable insertion of the genetic information into the host chromosome of robust microorganisms. This is necessary because plasmid-based systems, requiring the presence of antibiotics for their maintenance within their host cell,³ are less suitable for environmental applications, if at all.

However, the stable chromosomal integration of potentially promising catabolic gene sequences may constitute a more rational approach. Victor de Lorenzo developed effective gene ferries based on modified Tn5 mini-transposons over 20 years ago⁴ (similar systems are commercially available from Epicenter Biotechnologies). However, positive research findings published on their successful application in the areas of biodegradation and/or bioremediation are extremely scarce.^5,6 One of the reasons for this could be the fact that the published protocol for insertion and transconjugant selection requires selection by an antibiotic marker or by heavy metal resistance. We did not succeed in isolating transconjugants with the desired capabilities by using these protocols. However, after omitting selective screens above for these markers, we directly obtained transconjugants with the desired degradation capabilities in the presence of low concentrations of the novel (halogenated) target compounds. This can be achieved when transferring catabolic gene sequences taken from another potent bacterial donor strain, which were incorporated into the respective gene cassettes.⁴ This in turn allows the new hybrid strain to grow at the expense of a hitherto, non degradable or mineralisable compound such as a halogenated target substrate (chlorobiphenyls for example).⁶ However, for other applications or purposes, novel detection systems may need to be developed. Another important factor that may need to be considered is that the development of the desired novel capabilities is actually an evolutionary process; we very often obtained transconjugants only after prolonged maintenance of the mating mixtures over a period of many weeks or even months on the selected media containing the target substrates. On the other hand, after such a long period of incubation we obtained colonies which grew with isomers of halogenated target substrates such as chlorobenzenesulphonates, chloronaphthalenes, and chloronitrobenzenes, examples of basic building blocks for chemical syntheses. However, the analysis of their DNA by PCR showed clearly that the constructed gene cassettes were neither incorporated chromosomally, nor any more present in the used suicide plasmids for delivery. The positive clones probably developed from spontaneous mutations of catabolic genes, thus leading to the acquisition over time of altered gene sequences. This probable “evolution” enabled these clones to deal with the new halogenated carbon and energy sources (analysis of key genes is currently underway).

In a number of experiments, we totally failed to obtain the desired transconjugants, although the rational design and logical complementation of catabolic gene sequences should be theoretically feasible. One of the possible reasons for this may relate to the enzymatic activity and/or specificity of the host’s restriction-modification systems for the recognition of foreign DNA, resulting in its degradation by known classes of restriction enzymes.⁷ Although hitherto reported only in relation to pathogenic bacteria and archaea, an RNA-based CRISPR immune system acting against foreign DNA elements⁸ might also be present in environmentally relevant, biodegrading bacteria. However, bacterial taxonomy does not help very much to define and/or exclude potential candidates, which may harbour restriction-modification systems, since their numbers, and probably also those of an CRISPR systems, may vary significantly, even within a taxonomically defined species such as Pseudomonas putida. All this may explain, at least partially, the limited success achieved to date with synthetic microbial genetic engineering, as summarized in an article by Cases and de Lorenzo.⁹

Another concern is the selection of suitable host strains used in our “semisynthetic” experiments. When we initiated the procedure using these manipulations, we obtained a derivative of Cupriavidus necator strain H850 which is able to grow at the expense of Aroclors 1221 and 1232.⁶ Later, we manipulated a number of other biodegrading strains, especially of known potent PCB co-oxidizers, although many other strains capable of mineralizing biphenyl and other classes of chemicals of concern could not be manipulated genetically. We were able to achieve relatively satisfactory growth with technical mixtures of polychlorinated biphenyls by the modified strain Burkholderia xenovorans LB400, although growth was reduced on the more halogenated Aroclors 1016 and 1242, and to some extent on Aroclor 1248. These were already totally void of the growth inducing biphenyl required for expression of catabolic pathway genes (work in progress). However, only very limited growth of the similarly modified strains H850 and Pseudomonas pseudoalcaligenes KF707 was observed on Aroclor 1016. The sphingomonads we worked with were not found to be manipulable at all.

Over the medium term, taking into account the amount of work involved, the cost, and the budget limit of approximately US$40 million for the recent creation of “Synthia” on the one hand, and the hitherto limited success with conventional gene ferries on the other, novel broad-host-range (transposon) systems are required, together with new specific genetic systems in order to either rationally reduce or partially destroy the immune system of robust bugs, requiring their catabolic improvement for more efficient environmental applications.