New hepatitis virus or glassware contaminant?

Next-generation sequencing enables exponentially faster and cheaper sequencing of millions of DNA molecules in a single run through a lab protocol. Tiny amounts of DNA can be used to piece together the genomes, potentially detecting previously unknown pathogens, at unprecedented rates.

However, the very sensitivity of the technology can be problematic as the system can easily pick up contaminants and lead researchers into false conclusions.

Earlier this year scientists from the National Institutes of Health reported finding a new hepatitis virus in their study of blood samples from 92 people from China who had serious cases of hepatitis not caused by any of the five known hepatitis viruses.

Charles Chiu, MD, PhD, director of the University of California, San Francisco  Viral Diagnostics and Discovery Center, discovered the same virus, which they called parvovirus-like hybrid virus (PHV), independently, in a different set of hepatitis patients whose disease was not caused by known viruses.

To further investigate the origin of the virus, Chiu’s UCSF team tracked down its true source by applying next-generation DNA sequencing techniques in a set of carefully controlled experiments, and by referencing the ever-expanding scientific databases that spell out and catalogue viral genomes.

“At first we thought this was a genuine hepatitis virus, but later we found it in data sets from patients with many other diseases and even from animals,” said Chiu, a professor of laboratory medicine at UCSF.

The researchers also found that the virus sequenced from samples around the world strangely exhibited almost no genomic diversity. The research team strongly suspected that PHV was a laboratory contaminant and began probing other databases in search of the true source of the virus.

“We did some data mining of environmental databases and found matching DNA sequences from viruses originating in coastal waters off California and Oregon, but not elsewhere,” Chiu said.

It turns out that the new virus was actually a contaminant present in a type of glassware used in many research labs. The source of the contamination was tiny diatoms, a type of oceanic algae having nothing to do with human disease.

A developer of sampling and testing technologies may have sourced silica from diatoms in the ocean to make a popular glass column used in many studies, Chiu said. The columns are used to centrifugally spin biological samples to extract nucleic acids - DNA and RNA. Viral DNA that may have once infected the diatoms was also likely extracted as a contaminant during the procedure, along with DNA from biological samples, he said.

“The silica used in nearly all commercial spin columns is derived from the cell walls of diatoms,” he said. “We believe that PHV may be a diatom virus that had inadvertently contaminated the silica-based spin columns during manufacture.” The scientists do not know what caused the cases of hepatitis examined in the studies.

A similar viral false-alarm scenario unfolded in recent years with a mouse virus called XMRV, first reported in 2006. The virus initially was thought to be associated with human prostate cancer and chronic fatigue syndrome, but last year Chiu and others, including original XMRV co-discoverer Joseph DeRisi, PhD, Howard Hughes Medical Institute investigator and professor of biochemistry at UCSF, confirmed that XMRV was actually a viral contaminant of laboratory cell cultures and was not present in prostate cancer tissue.

These studies highlight the importance of repeating experiments with good controls to ensure that results are accurate, Chiu said. “Reproducibility is a cornerstone of science, yet far too few studies are validated by follow-up investigation,” he said. Next-generation sequencing is a promising approach to rapidly confirm and validate discoveries of new disease agents, saving investments in time and money that might otherwise be spent pursuing false leads, he added.

Chiu’s work was published online 11 September 2013 in the Journal of Virology.