New Sequencing Trick Catches Hidden Plasmids Spreading Antibiotic Resistance in Wastewater

Beneath the churning aeration tanks of every wastewater treatment plant, an invisible marketplace of genetic exchange is in constant operation. Bacteria swap small circles of DNA called plasmids, and the ones that move between cells through conjugation—the process often described as bacterial sex—can carry antibiotic resistance genes across species boundaries with alarming efficiency. These conjugative plasmids are among the most important engines of antimicrobial resistance spread, and treatment plants, where dense microbial communities meet constant chemical pressure, are recognized hotspots for their traffic. Yet scientists have long struggled with a stubborn technical problem: how do you actually read the plasmids circulating in such a complicated community? A new study published in Nature Water by Yuxi Yan, Yu Xia and colleagues at the Southern University of Science and Technology and their collaborators offers a strikingly practical answer.

The core difficulty is that recovering conjugative plasmids from complex communities has always depended on methods that introduce bias. Plasmid extraction—the standard laboratory route—pulls DNA out of cells chemically, but it captures whatever plasmids happen to be present, whether or not they can actually move between cells, and extraction efficiency varies wildly with plasmid size, copy number and host physiology. Cultivation-based approaches invert the problem: they only recover plasmids that can replicate in the specific host and under the specific conditions chosen by the experimenter. Both routes, in different ways, show researchers a distorted portrait of the true mobile plasmidome. For a field trying to assess environmental resistance risk, that distortion matters enormously.

The new method, which the team calls filter mating coupled with nanopore selective sequencing, or FM-NSS, sidesteps extraction entirely. The first stage borrows a classic microbial genetics technique. In filter mating, an environmental sample is placed on a membrane filter together with a laboratory recipient strain—in this case an Escherichia coli recipient—so that cells are forced into intimate contact. Any conjugative plasmid in the community that is compatible with the recipient can transfer into that strain during incubation. Crucially, this means that only genuinely transferable, recipient-compatible plasmids are captured, which is precisely the fraction most relevant to resistance spread. The researchers applied ampicillin selection to isolate recipients that had acquired plasmids, concentrating the mobile genetic material in a genetically defined pool.

The second stage is where modern sequencing technology takes over. Rather than extracting and sequencing all plasmid DNA indiscriminately, the team exploited nanopore selective sequencing, an adaptive sampling approach in which the sequencer itself makes real-time decisions about which DNA molecules to pass through its pores and which to reject. Molecules are identified in silico as they begin translocating, and unwanted ones—here, reads aligning to the E. coli recipient genome—are actively ejected, allowing only plasmid-derived molecules to accumulate in the output. After computationally depleting reads that mapped to the recipient genome, the workflow recovered a total of 1.10 million plasmid reads from a full-scale wastewater treatment plant, without a plasmid extraction step anywhere in the pipeline.

The yield from that single plant is substantial. De novo assembly of the selectively sequenced reads produced 622 non-redundant plasmid contigs, giving the researchers a catalog of plasmid sequences drawn directly from the treatment system. The most consequential comparison came when the team set this catalog against the results of conventional extraction-based plasmid sequencing performed on the same conjugant pools. Of the 622 contigs, 441 had not been detected by the extraction-based method. That figure—more than seventy percent of the recovered diversity invisible to the standard approach—illustrates just how much of the transferable plasmid pool traditional workflows have been missing. The technique does not merely offer an alternative view; it expands the capturable plasmid repertoire by a wide margin.

Because filter mating imposes a biological filter before sequencing, the recovered plasmids represent the fraction capable of entering a compatible recipient, which is exactly the mobile frontier of the resistome. The team then probed how the recovered plasmidome was distributed across the treatment system, and the patterns they observed were anything but uniform. Activated sludge—the dense, aerated microbial community at the heart of biological treatment—displayed a distinctly different recovered plasmid profile compared with other treatment stages. This distinction was associated with an elevated estimation of community replication rates, suggesting that the faster-growing communities in activated sludge may both host and disseminate a different, and possibly more active, set of conjugative plasmids than slower compartments elsewhere in the plant.

The study also examined the functional content of the recovered plasmidome, exploring the relationship between the plasmids and their host microbial community, the dynamics of the plasmid pool across different biological treatment conditions, and the correlation of mobile relaxase types with antibiotic resistance genes and incompatibility groups. Relaxases are the enzymes that initiate conjugative transfer by nicking the plasmid at its origin of transfer, and their classification provides a molecular fingerprint of how a plasmid moves. Linking relaxase diversity to resistance gene carriage and plasmid incompatibility groups allows researchers to build a more integrated picture of which mobile elements are most likely to shuttle resistance genes onward, and under what conditions.

Methodologically, the work stands on a growing foundation of adaptive sampling research. Nanopore selective sequencing was first proposed as a real-time enrichment tool in 2016, and subsequent developments—including Cas9-assisted targeting and Bayesian adaptive sampling algorithms—have made it increasingly powerful for pulling rare sequences out of complex metagenomes. What the Nature Water team adds is a clever biological pre-enrichment step: by using conjugation itself as the capture mechanism, they recruit the bacteria’s own transfer machinery to isolate the functionally mobile subset of the plasmid pool before a single molecule reaches the pore. The pipeline is also accompanied by publicly available software; the team released their custom Real End Finder framework through GitHub and deposited raw sequence data in the National Genomics Data Center, making the workflow reproducible and auditable by other laboratories.

The practical implications reach well beyond one treatment plant. Antimicrobial resistance is a global health crisis, and wastewater systems sit at a critical interface between human, animal and environmental microbial reservoirs. Surveillance programs currently rely heavily on short-read metagenomics and extraction-based methods that struggle to resolve plasmids—often repetitive, multi-copy circles that defy assembly—let alone determine which of them are conjugative. A workflow that directly captures transferable plasmids and reads them on long, information-rich nanopore reads offers a fundamentally sharper lens on that risk. It could allow regulators and engineers to compare plasmid traffic across treatment technologies, to test whether specific process conditions—redox state, chemical inputs, hydraulic regimes—suppress or stimulate conjugative transfer, and to track emerging resistance platforms before they establish themselves in receiving environments.

There are, of course, boundaries to what the method can see. The captured fraction is defined by the choice of recipient strain and the selection applied, so plasmids incompatible with E. coli or carrying different selective markers will remain outside the frame. Filter mating conditions themselves may favor some transfer rates over others, and the paper’s authors are careful to describe their technique as expanding the capturable fraction of conjugative plasmids rather than exhaustively cataloging all mobility. Even with these caveats, the advance is significant: a workflow that is practical, extraction-free, and demonstrably recovers hundreds of plasmid contigs that conventional approaches miss. As antibiotic resistance continues to migrate from clinical settings into environmental reservoirs and back, tools that can watch the traffic directly—in the tanks where it happens—are exactly what the field has been waiting for. This study suggests the watch has begun.

Subject of Research: Recovery of conjugative plasmids from wastewater treatment systems using filter mating coupled with nanopore selective sequencing

Article Title: Coupling filter mating with nanopore selective sequencing expands capturable conjugative plasmid recovery in biological wastewater treatment systems

Article References: Yan, Y., Wu, Z., Sun, Y., Zhang, M., Zhao, B., Nie, C., Cheng, Z., Yang, Q., Chen, L., Hao, Q., Fu, B., & Xia, Y. (2026). Coupling filter mating with nanopore selective sequencing expands capturable conjugative plasmid recovery in biological wastewater treatment systems. Nature Water, 4(9), 1100-1113. https://doi.org/10.1038/s44221-026-00695-w

Image Credits: AI Generated

DOI: 10.1038/s44221-026-00695-w

Keywords: conjugative plasmids, wastewater treatment, nanopore selective sequencing, filter mating, antibiotic resistance genes, plasmidome, activated sludge, horizontal gene transfer, metagenomics, adaptive sampling, antimicrobial resistance, environmental microbiology