American Scientist

I groaned as I looked through the microscope. I was relying on the cells I was looking at to measure the virus concentration in my sample. Normally, cells with a low concentration of virus would be happily growing and dividing, while cells with a high concentration would be sickly and ruptured. Instead of seeing this contrast between healthy and infected cells, all the cells were destroyed by contamination. After more than 10 hours of careful cell cultivation—and after attentively depositing the cells into thousands of separate growing wells and painstakingly applying my virus samples to those wells so that they would infect the cells—I could not see any pattern of viral infection. I grumpily threw out my sick cells: I would have to restart the experiment that had taken me an entire week to prepare.

My frustrating experience is all too common for researchers who measure viral activity. Scientists who work with viral vectors and vaccines measure viral activity, the concentration of infectious viruses in a sample, to make sure their product will be effective as a therapy after its design, production, purification, and packaging. In my case, studying virus activity helps me understand whether my viruses are intact and survived a purification process. Since viruses that are used in vaccines and gene therapy are grown in cell culture, they must be separated from cellular proteins and DNA for patient safety. The trick to good purification is removing these impurities without losing too much of your desired product in the process. Whole virus vaccines are always inactivated or weakened before distribution to patients.

Bradley Priem, a researcher at Johns Hopkins University, studies flu vaccine growth, which is upstream from purification. Priem notes how viruses are both delicate and time-consuming to measure. He says, “The current ways to measure virus [can be] very difficult to perform correctly.” If an initial virus measurement test fails, it could take multiple weeks to finally gather a successful result. Thankfully, these difficulties are being met with promising new techniques that will speed up research and development of vaccines and other virus-based therapies.

Viruses are much harder to detect than proteins or other biomolecules because they are often 1,000-fold less concentrated. Traditionally, the best way to determine virus concentration is to see how completely they can infect and kill a cell culture. These processes usually take 7–10 days and can be easily disrupted by issues like the contamination that I encountered. Moreover, cell-based experiments yield results that are more variable and thus more error-prone. Cells are living creatures; there are inherent differences in how they grow, divide, and weather viral infection. To account for variability, at least three replicates of the same test must be performed to confirm the result. Therefore, validating a single result can take about a month, at best, considering the time it takes for three rounds of virus growth. Contamination at any point in the process can make the process take even longer.

Using high-tech cameras can avoid some of the time-consuming labor in front of a microscope. A detection test for adenovirus, a cold virus that was instrumental to AstraZeneca’s COVID vaccine, is under investigation at McMaster University. Researchers traditionally measure adenovirus activity by dyeing the virus and manually estimating the virus concentration in a cell culture. Using the new method, researchers measure adenovirus activity by imaging the live cells that have been infected and watching the progression of adenovirus growth. An automated camera inside the cell’s incubator recognizes changes in the viral cultures more quickly than the naked eye and translates those changes into a measure of virus activity.

Claire Velikonja, the graduate student leading the live cell imaging project, hopes that the lab’s method will greatly speed up manufacturing adenovirus therapies. Her manufacturing experiments have produced more than 100 samples in one run, a number that would have been overwhelming to analyze using the traditional manual method. The automated camera watching cells for signs of infection allowed Velikonja’s team to “reduce the manpower associated with the process,” as she puts it. In the future, the research group is planning to increase the capacity of their cell imager, which can currently process 12 samples at one time, with the potential to further expedite manufacturing time. Because the cell imager tracks the virus’s effect on cell culture, it is giving a measure of virus activity. Even this method is vulnerable to setbacks from contamination, though.

While Velikonja’s project detects viruses according to their effects on cells, another method called biolayer interferometry analysis (BLI) detects viruses based on their attachment to a surface that is “sticky” to the virus. By avoiding cell culture, this method has fewer problems with contamination. BLI estimates virus concentration by measuring the thickness of the virus film that builds up on the surface. In Sartorius’s BLI sensor, this thickness is measured by comparing the reflection of white light through the film with a clean reflection. This method is currently being expanded to work for a diverse set of virus types.

Last year, the Instituto de Biologia Experimental e Tecnológica in Portugal published a BLI method to detect rotavirus A, which can cause dangerous infections in children. Traditionally, the rotavirus particles used to produce the vaccine take more than a week to measure. This time constraint significantly slows the manufacturing of the vaccine and its availability to patients. Using the new BLI method, the concentration of rotavirus particles can be determined in a matter of minutes by measuring the virus film buildup on a sensor specific to the rotavirus surface. This real-time rotavirus measurement could significantly speed up vaccine production and lower the cost associated with making the vaccine. One caveat to this method is that it measures virus particles, not virus activity. Not all the viruses may be active or intact to infect cells. BLI could help the process by quickly measuring large numbers of samples, but in many cases an infectivity test to relate virus particle concentration to virus activity would be necessary.

Sofia Carvalho, who led the project, notes that adapting the method from one virus to the next is not straightforward. “One of the most surprising things about virus research is the diversity that exists,” she says. “What works for one virus does not necessarily work for another.” However, Carvalho expects this diversity to yield more breakthroughs. “Viruses have a wide range of applications in various fields, including gene and cancer therapies, vaccines, and environmental biotechnology,” she says. “By understanding [the virus’s] natural mechanisms, we can harness their potential.”

Though cell infection imaging and BLI hold promise for speeding up virus research, the cost of buying the equipment required is currently a barrier. My own lab is experimenting with using BLI to measure a gene therapy vector’s particle concentration on a small scale. As these methods become more common, I hope to see virus researchers spending less time on repetitive tasks such as cell-based infectivity tests and more time on innovation that will produce new vaccines and viral therapies.