NEW YORK (GenomeWeb) – Researchers at Oxford Nanopore Technologies have cultured difficult-to-grow bacteria in their natural environment — a stream running behind the company's office — and assembled two bacterial genomes from sequencing data generated on the MinIon device.
"This approach allows rapid, inexpensive, and straightforward discovery, and genomic analysis, of previously uncultured prokaryotic genomes," they wrote.
The study was published on the BioRxiv preprint server last week and represents the first formal publication by Oxford Nanopore's applications group. The team included researchers from Oxford Nanopore's UK headquarters and its New York laboratory, co-located with the New York Genome Center, and also involved recent Oxford Nanopore data-analysis spinout Metrichor. Up until now, the firm had largely relied on customers to publish research results gained through its technology.
Led by Dan Turner, Oxford Nanopore's senior director of applications, the team harnessed a technique for growing "uncultivable" microbes from environmental samples in high throughput that relies on a so-called isolation chip, or iChip, developed by Slava Epstein and colleagues at Northeastern University and published five years ago in Applied and Environmental Microbiology.
The plastic chip, which the researchers manufactured using a 3D printer, has several hundred small diffusion chambers that are each seeded with a single cell from an environmental sample, in this case water from the stream. The researchers then left the chip in the stream for two weeks, allowing the bacteria to grow in their natural habitat. After that, they grew the bacteria for another four days in the laboratory, using filtered and sterilized water from the stream, before they picked and grew individual colonies and extracted their DNA.
After library preparation, they sequenced the DNA from a number of isolates on the MinIon and tried to identify the species in real time while the run was ongoing. Despite the fact that they could assign a species group to each culture, the results were "fairly inconclusive," they wrote, likely because the genomes of the subspecies in their sample were not in the database.
Next, they chose two isolates — which they identified as strains of Pseudomonas and Klebsiella oxytoca — for whole-genome sequencing, generating about 180 megabases of data for each. After error-correcting the reads, they assembled them using the Celera assembler, followed by manual finishing. Each assembly resulted in a circular contig, one 4.4 megabases and the other 6 megabases long. They then compared the genomes with the database to find the most closely related bacterial strains.
In a company blog post describing their study, the researchers wrote that they are currently analyzing sequence data from additional bacteria from the same experiment. They also plan to collaborate with others to correlate the genomic information with phenotypic traits and to explore the possible use of the bacteria in biotechnology. In addition, "we also plan to use this combination of iChip and MinIon sequencing to explore other environments, and possibly to investigate the effects on ecosystems resulting from environmental changes," they wrote.
A long-term goal is to completely avoid microbial culture, which is currently the slowest step in the process, and to sequence and assemble several bacteria directly from a sample. DNA extraction, library preparation, and sequencing on the MinIon, they noted, took less than a day.
"We anticipate that this kind of assay will have a significant impact on the way bacterial infections are diagnosed, especially where the infectious bacteria are unculturable under laboratory conditions, where misdiagnoses are not uncommon," they wrote, adding that other groups are already working on a rapid diagnostic test from urine that does not require cell culture.
The company expects to present posters with additional work from its applications group later this month at the American Society of Human Genetics annual meeting in Baltimore.