Wenting Luo:
The Soil Mobilome: A direct approach to study plasmid functional responses to long-term copper contamination

Date: 17-02-2015    Supervisor: Søren J. Sørensen, Leise Riber & Lars H. Hansen

Soil is considered a reservoir of diverse bacterial cellular functions, of which resistance mechanisms towards biological antimicrobial agents are of substantial interest to us. Previous findings report that the long-term accumulation of copper in an agricultural soil significantly affects the build-up of antibiotic resistance in culturable bacterial communities. Due to the difficulty of natural degradation, copper might represent a persistent selective pressure towards antibiotics even in the absence of antibiotics perturbation, and thereby copper and antibiotic resistance genes might be co-selected for among natural bacterial populations. One possible explanation is the horizontal transfer of resistance genes among soil bacteria mediated by mobile genetic elements, such as plasmids, integrons, transposons and bacteriophages, of which copper and antibiotic resistance genes can be linked on the same mobile elements. To test this hypothesis, we collected non-polluted and CuSO4- contaminated soil samples and attempted to describe the co-selection of plasmid-encoded copper and antimicrobial resistance via both an endogenous plasmid isolation approach as well as a plasmid metagenomic approach. The plasmid metagenomic approach, derived from the chromosomal metagenome, was named as the mobilome. It mainly focuses on plasmids and mobile genetic elements (MGEs), such as transposons, integrons and phages at a bacterial community level as described in manuscript 5.

In manuscript 1, we recovered a fraction of the culturable bacterial community by using the artificial soil extract medium. 16S rRNA gene sequencing showed that the proportion of Cu sensitive and resistant isolates shifts significantly, but that the bacterial community composition is highly resilient following the introduction of Cu. The Cu tolerance and antibiotic resistance assays revealed that bacterial copper tolerance did indeed select for resistance towards kanamycin, nalidixic acid and tetracycline. This might be explained by an observed positive response of bacterial plasmids to copper in copper tolerant bacteria. Besides, we successfully annotated a candidate plasmid harboring both Cu and multidrug resistance genes indicating the possible role of plasmid-encoded multidrug efflux pump systems in shaping the bacterial antibiotic resistance profile in soil.

Then, we applied a home-established mobilome approach in soil and evaluated the soil mobilome in manuscript 2. In contrast to a well-characterized soil metagenome, the soil mobilome shows the enrichment of plasmids, transposable elements and phages and approximately one eighth of the gene set was of plasmid-intrinsic traits, such as plasmid conjugation, mobilization and stability. The composition of plasmid replication protein families was significantly different from a previously reported wastewater and rat cecum mobilome. In addition, we detected antibiotic resistance determinants to aminoglycoside, beta-lactam and glycopeptide as well as multi-drug functions in the soil mobilome indicating that a substantial fraction of the soil resistome is plasmid-encoded and potentially mobilizable. This application provides an access to plasmid-encoded genetic traits in the soil environment as well as in other distinct ecological niches.

Manuscript 3 presented the comparative mobilome study between non-polluted and copper polluted soil. It reported the phylogenetic and functional response of the plasmid community towards environmental perturbation. The increased abundance of both antibiotic and heavy-metal resistance determinants in polluted soil indicated that copper contamination launched a new functional-resilient plasmid community in soil. Multidrug efflux pumps might be the main contributor to the heavy metal and antibiotic resistance in the bacterial plasmid community, and the persistent contamination triggers the frequency of horizontal gene transfer events, which accelerates the dissemination of resistance determinants in the community.

However, the mobilome approach contains one inevitable bias introduced by the phi29 polymerase during multiple displacement amplification (MDA), with which we guarantee the sufficient amounts of plasmid DNA for high-throughput sequencing. More specifically, the population of small circular DNA elements is enriched significantly on behalf of larger-sized plasmids. Thus, an improved method recovering larger-sized plasmids was described in manuscript 4. One electroelution step separating plasmids of a size larger than 23 Kbp was added before the MDA step. A wastewater mobilome spiked with the ratio 1:1 of two model plasmids of various sizes (~5 Kbp versus 56 Kbp) indicated the complete recovery of the larger plasmid using the modified protocol. The efficiency of the modified protocol was also revealed by the enrichment of conjugation associated functions, such as type IV secretion systems and T4SS-relaxosome coupling, which was detected in the wastewater mobilome without spiking of the two model plasmids. The test of the modified protocol in an even more complex sample, such as soil, will subsequently be a good supplement to further verify the efficiency of this modified protocol.