Researchers in Beijing, Toronto, and the United States have developed a protein array that can be used to test blood for the severe acute respiratory syndrome coronavirus.
SARS is typically diagnosed by using an antibody-based ELISA test on blood, or by doing an RT-PCR test on sputum samples.
The new array, which harbors proteins from the SARS coronavirus and five other coronaviruses, is advantageous over ELISA in that it requires much less sera, and it can detect SARS infection after about 10 days compared with 14 days for the ELISA test, according to Heng Zhu, the first author of the study, published March 7 in the Proceedings of the National Academy of Sciences of the United States of America.
The coronavirus array has been patented, and Snyder is currently working with various undisclosed companies to try to commercialize it.
In an interview with ProteoMonitor this week, Snyder said that the chip is more powerful than the ELISA assay because it relies on five different proteins to classify SARS, whereas an ELISA relies on just one protein.
"The power of the protein chip is that we're working with multiple classifiers," he said. "We have five different proteins we can use, and that is a lot more reliable than any one marker."
The array is also advantageous over the RT-PCR method because it can be used on serum samples, instead of sputum samples from the throat, which can be dangerous to collect.
"I was in China at that time, and I felt people were really desperate. Because the virus was really new to all of us, there wasn't a very good diagnostic method to efficiently and rapidly detect the infectious agent. ... That triggered my desire to develop a non-invasive method for detection."
"To get enough virus to be detected by the RT-PCR method, the clinician has to go deep into the patient's throat to get the sample," said Zhu. "This is dangerous in many cases, the clinician got infected."
The RT-PCR method is not used with blood because the SARS virus can be detected in blood only in late stages, Zhu said.
Though the RT-PCR method is very sensitive, it can be difficult to deal with because trace amounts of viral DNA can lead to contamination of samples, Zhu added.
Zhu said he decided to construct the coronavirus array in 2003 after SARS broke out in China.
"I was in China at that time, and I felt people were really desperate," said Zhu. "Because the virus was really new to all of us, there wasn't a very good diagnostic method to efficiently and rapidly detect the infectious agent. ... That triggered my desire to develop a non-invasive method for detection."
Instead of putting only the SARS coronavirus proteins on the array, Zhu and his colleagues decided to put proteins from five additional coronaviruses on the array as well.
"Coronaviruses have very conserved proteins. The nuclear capsid protein, which is the most antigenic one used in the ELISA assay, is very conserved across different coronavirus species. It was very important to us to test whether our assay could differentiate common cold patients from SARS patients," Zhu explained. "That's why we added the other proteins from five related coronavirus species.
To develop a SARS test based on their coronavirus array, Zhu and his colleagues screened 400 sera from patients in Toronto, where SARS was also detected in 2003. The Canadian samples, which came with clear patient histories, were used as a training set to develop the test, which is based on five biomarkers.
Once the test was developed, it was used to screen for SARS in 206 sera from Chinese fever patients. The test divided the patients into two groups: those with antibodies to the SARS virus, and those without.
When compared with ELISA results, the coronavirus array results were 94 percent in agreement with ELISA, Zhu said. There was no cross-diagnosis of SARS with other coronavirus-related diseases, he added.
Because clinical diagnostic data was missing from the Chinese patient samples, the researchers could not tell if patients who tested positive for SARS using the array truly had SARS.
"Strictly speaking, the ELISA results were not 100 percent accurate," said Zhu. "We don't know which patients are truly SARS patients because the clinical diagnosis data is missing."
During the SARS outbreak in 2003, many patients who showed symptoms of fever or cold were treated as SARS patients, though they didn't truly have SARS, Zhu said. The treatment has a lot of side effects, and could be damaging to patients, he added.
"The treatment is a high dose of hormone," said Zhu. "If you're not infected with SARS, you really don't want it. The high dose of hormone really has a lot of side effects, even for patients with SARS. The patients' bones become very fragile easy to break and they gain weight."
There is another SARS protein array that was developed by a group of Chinese researchers, Zhu noted. That array is different in that it includes only proteins form the SARS coronavirus, and not proteins from other coronaviruses, he said.
"Just to make sure you know what you're looking at, it's better to have the additional coronaviruses," said Zhu.
Tien-Shun Lee ([email protected])