The University of Minnesota was tapped by federal authorities Thursday to help answer fundamental questions about the antibody response to COVID-19 and whether people develop long-term immunity to the coronavirus that causes the infectious disease.

Labs since early spring have been able to identify antibody proteins in patients after they have recovered from COVID-19, but researchers haven’t established their importance and whether they prevent people from being infected a second time.

“They are interested to not only figure out why people have different levels of response ... but also to try to identify what happens to those levels over time,” said Dr. Amy Karger, leader of the U’s Advanced Research and Diagnostic Laboratory, where much of the institution’s COVID-19 antibody research has taken place.

The National Cancer Institute on Thursday awarded as much as $6.7 million over five years to the U and named it as one of four lead institutions to help build up the availability and accuracy of antibody testing. Along with Arizona State University, the Feinstein Institutes for Medical Research and the Icahn School of Medicine at Mount Sinai in New York, the U will be expected to build capacity to be able to test 5,000 people per week.

The announcement is part of the NCI’s Serological Sciences Network (SeroNet), launched on Thursday as the largest national coordinated effort to understand the immune response to COVID-19. Other centers will focus on the body’s natural innate response to infection or to the role T-cells play in identifying and killing infected cells.

Antibodies are easily identifiable in blood serum as markers for prior COVID-19 infection, but it’s unclear if or how many are needed to provide immunity against reinfection and whether they diminish over time, said Dr. Douglas Lowy, NCI’s principal deputy director. Also unknown is how many people would need to have antibodies for a community to achieve a level of “herd immunity” that would choke the spread of the virus.

“I don’t think we have answers right now, but I am hopeful by the end of this year we will at least have partial answers,” Lowy said.

NCI also is in the process of developing a dashboard that would show the results of local-level antibody testing and the percentage of people in communities who have already been infected. Federal estimates indicate as many as 10% of Americans have already been infected, with some never seeking diagnostic testing because they suffered mild or no symptoms.

The Minnesota Department of Health sought to launch a statewide prevalence study last month through randomized antibody testing of volunteers in 180 communities, but it was halted when federal survey teams reported incidents of racial abuse and intimidation and withdrew from the state.

The U and Mayo Clinic were among the first U.S. institutions to develop COVID-19 antibody testing capability this spring. Karger said one reason the U was selected for SeroNet was that its test was not only capable of identifying antibodies but also of assessing the amount of them in blood serum.

“Most of the commercial methods don’t give any sort of quantitative assessment of antibody level,” Karger said. “Most of them just say positive or negative.”

The U-derived antibody test also has shown 100% accuracy, albeit when validated on samples of people who had symptomatic COVID-19. It’s possible the test could be fooled by people with asymptomatic infections, and Karger said it will be important to study how antibody levels differ in those two populations.

“For me, the most interesting questions are really understanding the degree of disease severity and immunity,” she said, “and how those two are related.”

Case in point was the experience of Dr. Jon Cole, an emergency physician at HCMC, who tested positive for COVID-19 in March and whose wife and four children all suffered the illness as well during their isolation together at home. They sought antibody testing in June, and both parents tested positive but only the child who had the most severe symptoms tested positive as well.

“For us, it just left us with different questions,” Cole said.

First steps at the U will be to evaluate existing antibody tests and detection methods and to determine if other viral infections interfere with the results of some types of tests more than others.

The U will need an array of volunteers, both people who have suffered COVID-19 and those who haven’t, to donate blood serum samples for the studies, Karger said.

The U also is in the beginning stages of research to learn if antibodies can be reliably detected in saliva. It also will be conducting prevalence studies in medical residents and examining differing antibody levels in COVID-19 patients with HIV.