Scientists from the University of Colorado at Boulder and the Centers for Disease Control and Prevention (CDC) have developed an inexpensive “gene chip” test based on a single influenza virus gene that could allow scientists to quickly identify flu viruses, including avian influenza H5N1.

The researchers used the MChip to detect H5N1 in samples collected over a three-year period from people and animals in geographically diverse locales. In tests on 24 H5N1 viral isolates, the chip provided complete information about virus type and subtype in 21 cases and gave no false positive results, report the scientists. They say the MChip could provide a significant advantage over available tests because it is based on a single gene segment that mutates less often than the flu genes typically used in diagnostic tests. As a result, the MChip may not need to be updated as frequently to keep up with the changing virus.

The research was led by University of Colorado scientist Kathy L. Rowlen, Ph.D., and funded by the National Institute of Allergy and Infectious Diseases (NIAID), part of the National Institutes of Health. A paper describing the work, now available online, is scheduled to appear in the December 15 issue of the American Chemical Society’s journal Analytical Chemistry.

“Concerns about a possible influenza pandemic make it imperative that we continue to devise reliable and easy-to-use diagnostic tests for H5N1 that can be employed on-site where outbreaks are suspected,” says NIAID Director Anthony S. Fauci, M.D. “The MChip developed by Dr. Rowlen and her colleagues performed extremely well in initial tests and has the potential to be a valuable tool in global influenza surveillance efforts.”

The MChip has several advantages over the FluChip, a flu diagnostic previously developed by the same research team, says Dr. Rowlen. While the FluChip is based on three influenza genes - hemagglutinin (HA), neuraminidase (NA) and matrix (M) - the MChip is based on one gene segment. Unlike HA and NA, which mutate constantly and thus are technically difficult to use to develop gene chip diagnostic tests, the M gene segment mutates much less rapidly, Dr. Rowlen explains. “The M gene segment is much less of a moving target than the HA or NA gene. We believe that a test based on this relatively unchanging gene segment will be more robust because it will continue to provide accurate results even as the HA and NA genes mutate over time. The work summarized in our paper strongly supports that idea,” she says.

Another potential advantage is that the MChip would, for the first time, create a way to simultaneously screen large numbers of flu samples to learn both the type and subtype of virus present. Current real-time tests provide information about the type of virus (type A or B) in a sample, but additional tests must be run to determine the virus subtype (for example, H5N1 subtype.)

Working in biosafety-level-3-enhanced labs in Atlanta, CDC scientists, including Catherine B. Smith, M.S., extracted H5N1 genetic material from virus samples derived from human, feline and multiple avian hosts, including geese, chickens and ducks. The samples represented infections that had occurred between 2003 and 2006 over a vast geographic area, including Vietnam, Nigeria, Indonesia and Kazakhstan. Six of the human viral isolates were taken from an Indonesian family in which human-to-human H5N1 virus transmission was suspected. The virus diversity in the samples is important, explains Dr. Rowlen, because any diagnostic tool designed for eventual use on a rapidly changing virus, such as H5N1, must be able to detect as many variants as possible.

Dr. Rowlen and her colleagues tested the ability of the MChip to correctly identify 24 different H5N1 viral isolates, and distinguish those from seven non-H5N1 isolates. The MChip accurately identified and gave complete subtype information (identifying the samples as H5N1) for the 21 out of 24 strains of H5N1. Importantly, notes Dr. Rowlen, the test gave no false positives, meaning that the chip never indicated the presence of H5N1 when none was present. Following exposure to a viral isolate, the MChip displays results as a pattern of fluorescent spots. To automate the process of interpreting this pattern - thus eliminating the possibility of human error - the researchers developed an artificial neural network trained to recognize the distinctive pattern indicative of H5N1. Automating the interpretation of MChip results could allow it to be used more readily by health workers at the site of possible flu outbreaks, notes Dr. Rowlen.

“This new technology, once manufactured and distributed, could have the potential to revolutionize the way laboratories test for influenza,” says Nancy J. Cox, Ph.D., director of the CDC’s influenza division. “The MChip could enable more scientists and physicians, possibly even those working in remote places, to more quickly test for H5N1 and to accurately identify the specific strain and its features. This would greatly increase our ability to learn more about the viruses causing illness and take the best steps to respond.”