Metagenomics-based monitoring method improves epidemic monitoring via sewage

Under the leadership of the DTU National Food Institute, researchers from 11 European universities, institutions and knowledge organizations have developed a new method for analyzing data from wastewater monitoring. The method can help identify whether disease-causing bacteria, viruses, and antimicrobial resistance come from humans, animals, industry, or the environment.

Potentially, thousands of threats can be detected simultaneously, including antimicrobial resistance and cholera bacteria, which could help prevent disease outbreaks from escalating into epidemics. The research has been published in Nature Communications.

The researchers analyzed samples collected over three years from seven wastewater treatment plants in five major European cities: Bologna, Budapest, Copenhagen, Rome, and Rotterdam.

“Untreated wastewater is increasingly becoming a vital source for anonymous health and disease surveillance in large urban populations. However, extracting valuable data from it is not straightforward, as the wastewater contains both known and unknown bacteria from a variety of sources, such as humans, plants, animals, rainwater, dishwashing, etc.,” says corresponding author of the research paper, Assistant Professor Patrick Munk from DTU National Food Institute.

Additionally, the contents of the wastewater can vary due to seasonal temperature changes.

These challenges are what the researchers are beginning to overcome using a new computer program.

“Our research shows significant potential in metagenomics-based wastewater monitoring. While this method is more expensive than PCR testing, which proved highly effective during the COVID-19 pandemic, PCR only screens for one threat at a time. Metagenomics-based wastewater monitoring can assess thousands of threats simultaneously.

“Additionally, the value of each individual sample increases the more samples are collected over time, as historical data enhances the value of new analyses,” says Professor Frank Aarestrup, who leads the Research Group for Genetic Epidemiology at DTU National Food Institute and co-authored the article.

A monitoring system could be envisioned that combines metagenomics-based wastewater surveillance with PCR tests for specific threats that authorities deem likely to emerge.

The study is particularly relevant because an EU directive mandates that all major European cities begin monitoring antimicrobial resistance in wastewater. In Denmark, Statens Serum Institut is leading a large European collaboration on the implementation of this wastewater monitoring.

Cholera in Copenhagen

Hidden within the pipes leading to the Avedøre Wastewater Treatment Plant are some bacteria that the researchers did not expect to find: cholera bacteria. Although the amounts were very small, it was a big surprise for the researchers as they investigated the bacteria in wastewater treatment plants across five major European cities, including the three large plants in Copenhagen: Avedøre Wastewater Treatment Plant, Lynetten Wastewater Treatment Plant, and Damhusåen Wastewater Treatment Plant.

One can imagine that the bacteria were brought to the Avedøre facility’s local area by a person from a part of the world where cholera is still infecting people. This individual had the bacteria in their body and contributed feces to the sewage system, after which the bacteria settled in the pipes near the treatment plant and began replicating there. The researchers have observed that the bacteria have remained near the facility week after week but cannot be found further upstream.

Therefore, they suggest that the bacteria are not continuously coming from people who are currently ill but are residing in the biofilm on the pipes. There have been no recorded cases of cholera in Denmark for 150 years, and the bacteria hasn’t spread into the environment. However, higher temperatures could affect the geographic spread of cholera and other potentially dangerous microbes.

The new method of the study can detect where certain bacteria originate from, and although the DNA of the bacteria in the three Copenhagen plants is almost identical, there are still small differences that give each plant its own unique signature.

The presence of cholera bacteria near the Avedøre facility is described in a separate scientific article, which also stems from the present research and was published in the journal Microbial Ecology.

Software arranges vast datasets into mysterious groupings

Over a three-year period, from January 2019 to November 2021, 278 wastewater samples were taken from the inlet of the seven wastewater plants and sent to DTU. The researchers then analyzed billions of DNA sequences from the samples, assembling them into genomes from thousands of bacterial species, 1,334 of which were previously unknown.

The data was analyzed using software developed by the project’s Italian partner at the University of Bologna. This program identifies species that behave similarly over time and groups them.

“In the analyses, we could see that the bacteria in the wastewater clustered into very distinct groups. We began to wonder why and how the groups were formed. Initially, we thought the clusters might represent microbes collaborating with each other, but that was a dead end. Then, we investigated whether some of the groups might consist of bacteria from human feces, and that’s when we hit the mark,” says Munk.

Other groups turned out to be bacteria from the environment, and one group present in all the countries’ treatment plants likely comes from biofilms growing on the pipes leading to the facilities.

Once the researchers identified some of the groups using the analysis software, the task became easier.

“The principle is quite simple—certain bacteria always come from humans, and the bacteria that follow their sequences in the analysis likely come from humans as well. In this way, we can identify groups of species that follow each other over time,” says Munk.

What is a metagenome?

All living organisms have genetic material (a genome) made of DNA. Wastewater and other samples contain many different species of microbes, including bacteria and viruses. When you extract the mixed DNA from these species, you don’t just have one genome, but a metagenome. If each species’ genome is like a jigsaw puzzle, then the metagenome is like a whole bunch of mixed-up jigsaw puzzles.

Metagenomes can answer questions about which organisms were present and how common they were, making them a valuable tool for monitoring disease-causing bacteria and the genes that make them resistant to antibiotics. From each sample millions of DNA fragments are read, and a lot of samples can be analyzed by a supercomputer.

New method significantly improves success rate

The researchers have previously analyzed metagenomes but not as effectively as with the new method.

“In this new study, we identified 1,334 previously unknown bacterial species in the wastewater. Typically, when analyzing a metagenome consisting of 100 million small pieces of DNA, we could only identify the origins of about 10% of the DNA. However, in this new study, we’ve increased that to nearly 70% of the DNA assigned to the species from which we recovered a genome,” says Munk.

The ability to detect new bacteria is essential, as these bacteria may carry previously unknown antimicrobial resistance genes, and this method could potentially reveal new sources of antimicrobial resistance.

This is an observational study where the researchers worked with data based on the bacteria that were present in the samples from the untreated wastewater, but they did not themselves adjust any variables that can affect the frequency of specific bacteria. This introduces some uncertainty, and even though many human-associated bacteria cluster together, it doesn’t always happen.

The next step is to create a synthetic dataset where the researchers know which bacterial species are present and actively change the conditions to observe the outcomes.

“We don’t have a final success rate for this method yet, but it’s clear that we’re onto something significant. We need to optimize the method further to improve its accuracy,” says Munk.