Read our COVID-19 research and news.

Researchers are using metagenomics to identify new bacteria, fungi, and viruses.

Genomics comes full circle

This special feature is brought to you by the Science/AAAS Custom Publishing Office

As the field of microbiomics evolves, researchers are turning their attention toward culturing and characterizing novel organisms they’ve known only from sequence data.

In the beginning, there was the culture. Without knowing the nature of the molecules involved, early microbiologists relied on identifying heritable traits and chemical activities in microorganisms, then analyzing them through clever laboratory experiments. The identification of DNA as the genetic material, and the subsequent boom in DNA sequencing technology, brought microbiology into the genomic era. Scientists rushed to sequence not only genomes of individual microorganisms, but also metagenomes, encompassing the DNA of all of the microbes in entire ecosystems.

The fields of metagenomics and microbiomics grew exponentially. Biologists had long suspected that Earth was awash in many more life forms than they could see or culture, and now they had proof. From the human gut to the bottom of the ocean, complex biomes of microorganisms shape every aspect of life. The trove of new data has revealed tantalizing correlations, linking different microbiomes to everything from human intestinal health to climate change. What’s missing now are the phenotypes; what do all of these DNA sequences actually do, and how do they do it?

Using a variety of techniques, researchers are now starting to probe those questions. Some are trying to isolate and culture organisms so far known only from their DNA sequences, while others are pushing metagenomic sequencing in new directions to catalog the viruses, fungi, and plasmids that earlier efforts missed.

Researchers have been studying commensal organisms for decades, but now the technology is starting to help us ask different kinds of questions.

Joseph Petrosino

Cultures of innovation

One of the first environments microbiome sequencers sampled was the biggest: the ocean. The results revealed immense populations of bacteria that had never been characterized before. While molecular biologists began mining the data to understand the fundamental biochemistry of these organisms, pharmaceutical chemists saw an abundant source of potential new drugs.

“We thought unusual locations would lead to unusual diversity, and that would lead us to unusual chemistry with unusual activity, and that seems pretty much to have paid off as a strategy,” says Marcel Jaspars, chair of the Department of Chemistry at the University of Aberdeen in Aberdeen, United Kingdom. Jaspars and his colleagues initially worked on the PharmaSea project, a multi-institution collaboration that focused on sequencing marine microbiomes and searching for novel metabolic products from these microbes. Those compounds, in turn, could prove useful as leads in drug development programs. The PharmaSea project ended in 2017, but Jaspars’ group has continued to explore the ocean depths.

“For us, it’s cool to look at the genomes and to look for biosynthetic pathways that we’re interested in for natural products,” says Jaspars. For example, many antibiotics have originally been found in bacteria, and researchers working on the PharmaSea project have also discovered compounds with antioxidant and other promising pharmaceutical properties. The genomic data reveal numerous metabolites the organisms could produce. Proteins and peptides are easy to identify this way, but some of the most interesting compounds come from complex metabolic pathways that aren’t obvious from sequence data alone. To address that, the team also tries to grow the bacteria in culture to study their metabolisms.

It’s a challenging process. “We often have to try multiple different conditions to see if we can get the expression of [interesting] metabolites,” says Jaspars. Because he focuses on marine organisms, his lab also has to replicate the extreme conditions of the deep ocean. By tinkering with culture media and using special equipment, such as pressure chambers that can subject samples to the equivalent of several kilometers of ocean depth, he’s managed to grow numerous organisms that were previously considered unculturable. In collaboration with Advanced Chemistry Development in Toronto, Ontario, Canada, Jaspars and his colleagues then identify the metabolites the organisms produce.

In an alternative approach, the team works with microbiologists to express the unculturable microbes’ genes in more lab-friendly hosts. “When you have the gene cluster sitting there, you can often manipulate it much more easily in an alternative host rather than in the original host,”
explains Jaspars.

Getting novel organisms to grow in culture isn’t just a useful tool for laboratory research. It will also be an essential step in getting microbiome-based treatments into the clinic. Preliminary clinical studies have shown, for example, that transplanting fecal bacteria from a healthy individual into one with a chronic Clostridium difficile infection can cure the infection. Turning that finding into a treatment that regulatory agencies would approve, however, has been hard; feces is not a defined pharmaceutical product.

“You’re going to make a medicine, you know, you’ve got to be able to grow it,” says Trevor Lawley, faculty group leader in the Host–Microbiota Interactions Laboratory at the Wellcome Trust Sanger Institute in Hinxton, UK. In a seminal 2012 paper, Lawley and his colleagues began to address that problem, successfully culturing 18 species of bacteria from healthy mouse feces and using the defined cultures to cure another group of mice of C. difficile infections (1). “It was pretty big because we were told, ‘You just can’t culture those species,’” says Lawley.

Like Jaspars, Lawley combined sophisticated data analysis with old fashioned trial and error. “We made a lot of mistakes,” Lawley says. The team progressed from one culture medium to the next, checking and sequencing each culture to identify which bacteria were still present from the original fecal sample and which had gone missing. Then they tinkered with the conditions and tried again.

Lawley’s lab eventually developed a set of media formulations for culturing human microbiomes. “Now we can take a stool sample from you, and within four weeks, we can have 90%–95% of species ultrapurified and archived ... with a taxonomic ID,” he says, adding that “it allows us to make the medicines and enables phenotypic biology.” Microbiotica, a biotechnology company in Cambridge, UK is working to take the resulting microbiomic therapies into the clinic.

While making medicines was the original goal, the work has had a major side benefit for basic research. “One thing I didn’t appreciate, which actually turns out to be massive, is I think it’s transforming the way we do our metagenomic analysis,” says Lawley. By growing much of the human intestinal microbiome in culture, his team has been able to assemble complete reference genomes for the previously uncultured species. Investigators can now compare those reference genomes to metagenomic sequences they obtain in other samples, vastly improving their data.

Funded by the UK’s Natural Environment Research Council, the PolyExESS (poly-extremophile environmental stimulation system), used by Marcel Jaspars’ team at the University of Aberdeen to sequence ocean microbes, can take six 2-L microbial fermentation vessels to pressures of 200 MPa (equivalent to about 20,000 m of sea water—in excess of the world’s deepest trench at 11,000 m). It also has temperature control to take it to low (below freezing) and moderately high temperatures.

The fungus among us

Bacteria aren’t the only organisms that make their homes in the human gut, though one could be forgiven for getting that impression from the early microbiomics literature. “Much of the research focused on the bacterial component because it was sort of the low-hanging fruit, and also thought to be the biggest contributor to health and disease,” says Joseph Petrosino, chair of the Department of Molecular Virology and Microbiology at Baylor College of Medicine in Houston, Texas.

Petrosino’s lab worked on the Human Microbiome Project (HMP), which began in 2005 by sampling the microbiomes of 150 healthy individuals over the course of a year, then sequencing the bacteria in them. More recently, the researchers began to wonder about the fungi. “Many people consider them pathogens, and oftentimes they are, but ... what does [fungal diversity] look like in a healthy individual?”

Focusing on a genome feature common to yeasts and molds, the researchers searched for fungi in the HMP fecal samples, yielding a preliminary look at the normal human mycobiome. The results revealed that “the fungal communities are much less diverse than bacterial communities, and they’re less stable over time,” says Petrosino. Since that work, published in 2017, other labs have begun looking at the mycobiome in both healthy and diseased guts (2). “It’s probably a few to 5 years or so behind the bacterial microbiome field,” says Petrosino.

One of the challenges of studying mycobiomes is the relative scarcity of fungi in samples. In a fecal specimen, for example, the sheer quantity of bacterial DNA can eclipse the fungal sequences. Meanwhile, biopsy samples often contain large numbers of host cells. “There’s been some method development that’s helped, but it really hasn’t made the paradigm-shifting step yet,” says Petrosino.

Nonetheless, the field has yielded some tantalizing results. One clinical study showed that individuals placed on a strict diet lacking bread, beer, mushrooms, and other sources of fungi eventually develop fungus-free gut microbiomes. “Then the question is ... what are we doing by eating these types of foods, and do they benefit health, or are we hurting ourselves?” says Petrosino. In another project, researchers in Petrosino’s lab are using 3D cell cultures derived from gastrointestinal biopsies to study the interactions between various fungi and the intestinal epithelium. Petrosino hopes to combine those models with additional mycobiome sequencing data to develop a complete map of interactions between bacteria, fungi, and host cells.

Going viral

Though challenging, efforts to catalog previously uncharacterized bacteria and fungi can nonetheless rely on common features of those organisms. All bacteria, for example, carry highly conserved 16S ribosomal RNA sequences that change slowly over evolutionary time. A common tactic in microbiomic studies is to use PCR primers that target these sequences, then classify the bacteria in the sample based on their differences.

Viruses offer no such shortcut. Instead, researchers probing the virome rely on bioinformatics. “There’s multiple ways in which you can kind of try to enrich computationally for viral sequences,” says Stephen Francis, assistant professor of neurological surgery at the University of California, San Francisco. In a massive study of the viral diversity in samples from 150,000 women in China, Francis and his colleagues performed high-throughput sequencing, then subtracted all of the sequences that matched a reference human genome. They then searched for matches between the remaining sequences and a database of known viral sequences.

The results revealed not only the population-wide distribution of well-known pathogens, such as human papilloma and hepatitis B viruses, but also intriguing differences in endogenous retrovirus sequences that have integrated into patients’ genomes (3). “We’ve been constantly bombarded by viruses, and sometimes when a virus endogenizes into our genome, it can provide either some type of functional benefit or detriment to the species as a whole, so I think that event is a really important one to understand,” says Francis. Though these ancient integrated viral sequences were once considered “junk DNA,” more recent analyses suggest they have contributed to everything from the evolution of live birth to variations in disease risk across populations.

Francis plans to scale up his viromic analysis to cover 1 million women, all participants in an immense Chinese effort to collect and catalog samples taken during noninvasive prenatal testing. He argues that “large population-based studies are really going to contribute a huge amount to our understanding of these relationships” between viruses and health. Because the samples come from pregnant women, the investigators can look for links between specific viruses and birth outcomes. “In utero infection is clearly an important aspect to understand, especially for shaping of immune responses later in life,” says Francis. He adds that other large population studies, such as the UK Biobank effort, provide additional stores of samples for viromics researchers to mine.

Links on a chain

Bacteriologists are also starting to reach beyond 16S ribosomal RNA sequencing. Many have now moved on to metagenomic efforts, which involve sequencing all of the DNA in a sample and attempting to assemble complete bacterial genomes from the data. While those projects have made impressive progress in recent years, another important aspect of bacterial life has remained difficult to study: plasmids.

As small circular DNA elements that maintain themselves apart from the bacterial genome, plasmids provide a deep pool of genes that often confer traits such as antibiotic resistance, novel metabolic capabilities, and virulence on their hosts.

“You can use regular sequencing to get an idea of the kinds of plasmids that are in a microbial community, but what you cannot do is link those plamids to the bacterial hosts that they’re in,” says Eva Top, professor of biological sciences at the University of Idaho in Moscow. Many plasmids are promiscuous, transferring easily between different species of bacteria, so even if scientists discover a particular plasmid associated with one species, it doesn’t prove that it’s absent from others. Meanwhile, some strains of a bacterial species may carry a plasmid while others may not, and the distribution of the plasmid can change over time within a microbial ecosystem.

To address that, Top and her colleagues adopted a technique called proximity ligation, which was originally developed to link antibodies together in immunoassays. By modifying the method, the researchers are able to bind plasmids to their host genomes before lysing the bacteria for sequencing. They can then obtain sequence data on both the host chromosome and its associated plasmids, which has allowed them to analyze the dynamics of plasmid movement in different microbiomes.

“We’re interested of course in the basic biology of plasmids, but also because of the problem of antibiotic resistance and the role that plasmids play in the spread of that resistance,” says Top. So far, her team’s analyses have revealed that certain groups of bacteria are especially adept at picking up plasmids with antibiotic resistance genes, possibly serving as reservoirs for resistance that could then spread to pathogenic species (4). The researchers have also started analyzing how newly introduced plasmids spread through a microbiome, such as when antibiotic resistance–carrying plasmids in animal manure encounter the soil microbiome in agricultural fields.

Regardless of their specific approaches, the scientists pushing genomics in new directions are optimistic about the future. As Petrosino explains, “Researchers have been studying commensal organisms for decades, but now the technology is starting to help us ask different kinds of questions.”


  1. T. D. Lawley et al., PLOS Pathog. 8, e1002995 (2012).
  2. A. K. Nash et al., Microbiome 5, 153 (2017).
  3. Liu et al., Cell 175, 347–359.e14 (2018). T. Stalder, M. O. Press, S. Sullivan, I. Liachko, E. M. Top, ISME J. 13, 2437–2446 (2019).

Submit your new product press release/description or product literature information to Visit Science New Products for more information.

Newly offered instrumentation, apparatus, and laboratory materials of interest to researchers in all disciplines in academic, industrial, and governmental organizations are featured in this space. Emphasis is given to purpose, chief characteristics, and availability of products and materials. Endorsement by Science or AAAS of any products or materials mentioned is not implied. Additional information may be obtained from the manufacturer or supplier.

Alan Dove

Alan Dove is a science writer and editor based in Massachusetts.

Search Jobs

Enter keywords, locations or job types to start searching for your new science career.