Skip to main content
Premium Trial:

Request an Annual Quote

Metagenomic Sequencing of Sputum DNA Shows Potential for Future Tuberculosis Dx


NEW YORK (GenomeWeb) – Anticipating a future in which metagenomics becomes the norm in detecting and diagnosing infectious diseases like tuberculosis, researchers from the University of Warwick in the UK have published a study demonstrating that direct metagenomic sequencing of DNA in sputum samples can accurately detect TB infection.

In the study, published this week in the open access journal PeerJ, the team led by microbial genomics professor Mark Pallen tested eight sputum samples as an initial proof of principle for its shotgun metagenomics approach, which involves sequencing DNA from a biological sample without culture- or target-specific amplification or capture, and then filtering out human reads to detect the presence of pathogens or other sequences.

Using this approach, the researchers detected sequences from M. tuberculosis in all eight known-TB-positive sputum samples they sequenced and were able to assign the bacteria to a known lineage in seven.

Pallen told CSN in an email that while there are numerous PCR and other amplification-based assays for detecting TB and other pathogens, as well as sequencing methods applicable to cultured samples that have recently been shown to provide important lineage and drug sensitivity epidemiological information – metagenomics offers the possibility of making disease detection and diagnosis part of a single testing strategy that could also include drug sensitivity detection, resolve mixed infections, and more.

"The final goal, shimmering on the horizon, is that we might be able to extract information from all the macromolecules in a sample (DNA, RNA, proteins) so that we get a readout of what pathogens are there, what virulence or resistance genes are being exposed, what host responses are switched on …. and also maybe detect cancerous or pre-cancerous changes," Pallen said.

"You might be able to recover genomes from other respiratory pathogens, viral, bacterial or parasitic, from the sample. This also means that you can use a unified one-size-fits-all approach to diagnosis that does away with the need for an onerous workflow that includes culture on many different kinds of media in many different atmospheric conditions," he added.

In their initial study to show that sequencing DNA straight from a biological sample without capture or target-specific amplification could detect TB, Pallen and his team tested eight smear- and culture-positive sputum samples collected from patients in The Gambia.

The group's shotgun metagenomics approach begins with a differential-lysis protocol followed by a kit-based DNA extraction, and then sequencing performed on the Illumina MiSeq. According to the authors, the number of sequence reads in each of their sputum-derived metagenomes ranged from 989,442 to 2,818,238. The proportion of reads in each metagenome mapping against the human genome ranged from 20 percent to 99 percent.

The team was able to detect sequences from the M. tuberculosis complex in all eight samples. The researchers were also able to assign seven of eight metagenome-derived genomes to a species and lineage within the M. tuberculosis complex by analyzing the distribution of large sequence polymorphisms and SNPs.

Using their fairly "out-of-the-box" sequencing protocol, Pallen wrote, depth of coverage was not sufficient to yield drug sensitivity or resistance data. But the group is now attempting to enhance the approach to dig deeper.

"We can [achieve greater depth of coverage] in two ways," Pallen said in his email. "We can lower the proportion of reads coming from the human genome and other bacteria or we can try to increase the recovery of mycobacterial sequences or both."

The team's approach involves a differential lysis protocol that lyses human and fragile bacterial cells first, then uses DNAse to remove those sequences before lysing the remaining mycobacterial cells. "We can probably make the DNAse work more effectively. We could also try capturing mycobacterial cells or DNA, although that might compromise the open-endedness of the technique," Pallen wrote.

Any step that might narrow the potential scope of metagenomic sequencing is something the group hopes to avoid.

"Pulling out the DNA from the pathogen first means you only get back what you are looking for, i.e. you only find out about TB. The great advantage of shotgun metagenomics is that it is to a large degree open-ended, detecting things you did not know or even suspect were there," Pallen explained. "[Metagenomics] allows you to make precise epidemiological and evolutionary inferences, while also detecting mixed infections … Plus it can also detect ecological changes in a microbiome in conditions that may not be caused by a single headline pathogen, but by disruptions or differences in microbial communities."

Moving forward, the researchers are planning to work on a larger number of sputum samples. Pallen said that Emma Doughty, the study's first author, is hoping within a few months to return to The Gambia, where the study samples were collected.

Before the investigators can expand their sample cohort though, Pallen said they first intend to work on optimizing their DNA extraction protocols. "We were pleasantly surprised that the protocol worked 'out of the box', but we are confident that we can improve things so we get fewer human DNA sequences and more mycobacterial sequences from each sample," he wrote.

The timeline for advancing shotgun metagenomics into clinical disease detection may also depend on advances in sequencing technology, Pallen wrote. "It is hard to know precisely how quickly we will get there. Perhaps five years from now? … If nanopore sequencing lives up to the hopes for it, it may be quicker than that."

He added that his group has had some discussions with Oxford Nanopore about using its technology in this application as a trial.

Another avenue Pallen said he'd like to explore in terms of clinical applications of metagenomics is analyzing the gut microbiome in critically ill patients, but the team will need to seek separate funding for that.

Meantime, Pallen and his colleagues have also applied shotgun metagenomics to other pathogens, mainly in the context of investigating disease in ancient samples, rather than modern clinical diagnosis. In a recent study, for example, the group sequenced stool samples from a 2011 German Escherichia coli outbreak.

The team was able to identify previously unknown co-infection with Clostridium difficile in some of the E. coli-positive patients, and in one case recovered sequences from an emerging diarrhea pathogen, Campylobacter concisus, when the clinical lab had diagnosed simply C. difficile-associated diarrhea. According to Pallen, the group's German collaborators on that study have since gone on to use the approach to diagnose a case of chlamydial pneumonia using shotgun metagenomics on sputum.

Pallen's team also published a study this year describing its detection of a 700-year-old Brucella melitensis genome from the skeleton of a 14th-century middle-aged male.

In 2013, the researchers used metagenomic sequencing to uncover a pair of co-infecting tuberculosis strains related to contemporary TB outbreaks in lung tissue from a 215-year-old mummy found in a Hungarian crypt.