You are here: UNIL > Department of Fundamental Microbiology > Research units > van der Meer > Projects > Adaptation & Evolution

Genetic adaptation and evolution of catabolic pathways

Horizontal gene transfer | The clc element | Plasmid metagenomes | Regulation of 2-hydroxybiphenyl degradation

Horizontal gene transfer

Our group has long been interested in the evolutionary mechanisms which enable bacteria to acquire new functions for the degradation of toxic and xenobiotic substances in the environment. We were approaching this question for two reasons. From an environmental viewpoint it would be important to know if bacteria would be capable of degrading chemical substances which they had not seen before, because it would result in more or less spontaneous disappearance of pollutants from the environment. From an evolutionary point of view, challenging bacterial communities with new and large amounts of chemical substances would pose a selective pressure for any new bacterial variant capable of using the compounds as carbon and or energy source for growth. This would represent a method for selective enrichment of spontaneous natural variants which should have visible traces of the evolutionary events allowing their adaptation.

The approach, which in a similar form was taken by other groups as well, was relatively successful and allowed us to propose that not accumulation of small mutations within one species (vertical evolution) was the most important driving force for the development of new catabolic pathways, but gene exchange and recombinations of existing genetic material in microbial populations within one microorganism (horizontal evolution). Mechanistically, the process of horizontal evolution seemed to be carried out by transposable elements, (unknown) recombinatorial processes and plasmid transfer between different cells. Much to our surprise, in two cases of natural adaptation the transfer process did not involve plasmids but another type of conjugative element. We could prove that this type of conjugative element was a mobile region on the chromosom. Such regions are now more widely known as genomic islands, among which pathogenicity islands are found as well, although most genomic islands are (no longer) mobile (see above).


The clc element

The genomic island we identified in Pseudomonas sp. strain B13, the first bacterium described to degrade 3-chlorobenzoate, has been called the clc element (or ICEclc) because of its ability to provide chlorocatechol degradation to the host. The clc element has a size of 105 kb and we know of one natural variant which is very closely related. This clc like element was present in a Ralstonia sp. degrading chlorobenzenes and was isolated from contaminated groundwater in the United States. We and others showed that the clc element is self-transferable in laboratory 'matings' at high frequencies, but also transfers in more complex communities and technical systems from strain B13 to other Beta- and Gammaproteobacteria, such as Ralstonia, Pseudomonas putida, Burkholderia sp. and P. aeruginosa. The work on Ralstonia demonstrated that the clc element is not a laboratory artifact but present in the natural environment and can acquire gene fragments from other related species.ICEclc is completely capable to self-transfer from the strain B13 to an intermediate recipient and from there to another species. Details on the transfer process are still mostly lacking.

Depending on the host, the clc element can be present in one, two, or more copies. Also in strain B13 two copies are present. In all hosts examined, the clc element resides site-specifically in the chromosome at the most 3' 18-bp sequence of a tRNAGly gene. At the other end of the element, a copy of this 18-bp sequence is found. A critical feature of the clc element's life-style is formed by the integrase, the gene of which (intB13) is found close to the tRNAGly insertion site and oriented away from it. A cloned intB13 integrase in E. coli carries out the site-specific recA independent integration of a plasmid with the 18-bp repeat sequence into a second plasmid carrying the tRNAGly gene. This showed that the integration process into the chromosome is mediated by the IntB13 integrase enzyme. The reverse process, excision, has not been reproduced in E. coli, but small amounts of the excised clc element can be detected in cultures of strain B13 itself by Southern hybridization and PCR.

Sequencing of this excised form showed that it consists of a closed DNA molecule in which both 18-bp repeats have recombined again to one. The current hypothesis therefore is that the integrated clc element can excise from its chromosomal location and form a closed 'circular' intermediate. The closed intermediate can either reintegrate or be transferred to a new recipient cell, in which it can again integrate. Reintegration of the clc element has been observed in single cells which had an additional integration site engineered to induce GFP fluorescence expression when inserted by the clc element. ICEclc is behaving rather strangely: it becomes only active for transfer in a few percent of all cells in a population. This phenomenon is called bistability. Bistability of ICEclc can be observed by using transcriptional fusions of the integrase promoter to a gene for an autofluorescent protein.


Plasmid metagenomes

In this new project we are studying the plasmid metagenome of wastewater treatment plants. This research is motivated by the concept that wastewater treatment plants are probably the biggest open mixed continuous cultures of microorganisms on our Planet. WWTP receive different kinds of wastewater, from households, industry or hospitals, and depending on all this a community of microbes developes in the system. Since it is known that plasmids contribute to a large extent to the adaptation potential of microbial communities, we are very interested to see what gene functions are encoded on plasmids in WWTP and how this evolves over time and as a function of WWTP operation.

In collaboration with JGI-DOE in the US two libraries of plasmid purified DNA from the WWTP in Morges, Canton Vaud, were sequenced by classical Sanger technology and 454 Titanium. This 'pool' of metagenomic sequences in currently being assembled, annotated and compared to other microbial and viral metagenomes.


Regulation of 2-hydroxybiphenyl degradation

Most of the catabolic pathways, which allow bacteria to use aromatic compounds as sole carbon and energy source, are controlled at the transcriptional level. This allows the cells to express the necessary enzymes only when the substrates for the pathway are present. Transcriptional regulation is usually mediated by a single regulatory protein.

Different bacterial strains are able to use hydroxybiphenyls as sole carbon and energy sources. One of these, Pseudomonas azelaica HBP1, degrades 2-HBP and 2,2'-dihydroxybiphenyl through a so-called meta-cleavage pathway. The first steps of 2-HBP metabolism are initiated by three enzymes encoded by the hbp genes, hbpCA and hbpD.

Bacteria such as Pseudomonas azelaica do not synthesize the enzymes for metabolism of 2-HBP when there is no 'need' for it. How does the microorganism know when there is a need? Usually, the bacteria can sense the presence of the chemical they might be able to metabolize. P. azelaica has developed a rather sophisticated system to detect 2-HBP in the cell and subsequently tune the expression of the hbp genes to the appropriate level to make sufficient of the enzymes for 2-HBP breakdown.

Detection of 2-HBP in the cells of P. azelaica is achieved by a specialized transcriptional regulatory protein, called HbpR. As a matter of fact, the gene for the HbpR protein is encoded just upstream of the gene hbpC. This transcriptional regulator is an amazing protein. First of all, it is capable of sensing the presence of 2-HBP. It is not known how this process is performed although we know which parts of the protein are involved. The HbpR protein is also capable of binding the DNA at a few very specific positions. We have some idea about which sequences HbpR likes to bind to and which it doesn't. Next, HbpR can activate RNA polymerase to start transcription. This is also very special, because this form of RNA polymerase contains the sigma factor s54, a form which is unable to start transcription by itself. This is in contrast to RNA polymerase containing s70. How HbpR is doing this is not known. It is only known that the process requires ATP hydrolysis and requires binding of 2-HBP.

As far as is known from biochemical studies with HbpR and other related proteins, these transcriptional regulators have dedicated structural protein domains for the roles outlined above and form a symmetrical hexameric protein complex when binding to the DNA. Very peculiar, indeed!

HbpR is activating the expression of the hbp genes, but from two separate locations. The first binding site (and promoter for the s54-RNA polymerase) is in front of hbpC. A transcript is formed that contains both hbpC and hbpA, but after hbpA it stops. A second promoter and HbpR binding site is in front of hbpD. Transcription here results in a single cistronic mRNA containing only hbpD. Also curious, since there is no particular reason why hbpCA and hbpD would not be synthesized as a single polycistronic mRNA. We think that this is just a relict of some evolutionary event.


 in this site:














Bistability of the ICEclc in Pseudomonas knackmussii, observed with a transcriptional fusion of the integrase promoter to egfp. Picture: Marco Minoia 








Biophore - CH-1015 Lausanne  - Switzerland  -  Phone +41 21 692 56 00  -  Fax  +41 21 692 56 05