Scientists from the National Institute of Allergy and Infectious Diseases have developed a mathematical algorithm that can identify broadly neutralizing antibodies in blood samples.
The antibodies, called bNAbs for short, can combat infection by the majority of HIV strains. As such, the researchers believe that their tool could help speed up efforts to develop a vaccine for HIV as well as other viral diseases.
Details of the method and its application to HIV-1, a subtype of the virus, are described in a recent Science paper. That study was intended to show that the "epitope specificities" — epitopes are the part of an antigen that antibodies recognize — of the HIV-1-neutralizing antibodies in a given blood sample could be determined by simply looking at similarities between the virus neutralization patterns of the sample and known HIV-1 antibodies.
The paper explains that the so-called serologic algorithm analyzes HIV strains that are neutralized in the blood samples to determine what types of HIV bNAbs are present. The algorithm works by first generating a database of neutralization patterns or "fingerprints" of known antibodies — a measurement of which virus strains the antibodies can block and with what potency — and comparing them to the patterns of the antibodies in the sample.
"The main idea is that the virus varies a lot in terms of sequence and so the antibodies that [target] different parts of the virus will be affected in different ways [as a result] of the sequence variation," Ivelin Georgiev, a staff scientist in NIAID's Structural Bioinformatics core and one of the paper's authors, explained to BioInform. That also means that the "neutralization of the different viral strains will be affected by the variation within a given epitope," he added. "But antibodies targeting a different site should not be affected by the variation within that epitope."
Building on that premise, "we took neutralization data for a large set of antibodies against diverse viral strains … and we analyzed the diversity of neutralization patterns [or] neutralization fingerprints for the antibodies," he said. "What we found was that the relative potency with which a given viral strain is neutralized by a given antibody correlated between antibodies targeting the same epitope and didn't correlate between antibodies targeting different epitopes." Simply put, antibodies that neutralize viral strains in the same way have similar targets, he explained.
Comparing the neutralization patterns of the antibodies observed in samples to antibodies with known fingerprints helps "[us] establish what fraction of the neutralization by the given serum can be attributed to a particular antibody type," Georgiev said. This is necessary because blood samples usually contain mixtures of antibodies.
The team believes that their approach may prove useful in vaccine-development efforts for HIV and viral diseases such as influenza and hepatitis C because it "allows us to quickly and more accurately look at [infected] patients' or vaccinees' serum," Georgiev explained.
In the case of a vaccine, "if we have a number of vaccinees for which samples were taken at different time points after vaccination … then we can look through these samples and quickly analyze what kinds of antibodies developed in response to the vaccine … when they developed, [and] how they evolved at those different time points," he said. It could be used in a similar fashion to analyze blood from infected patients, he added.
In the long run, he said, the method could potentially be used to develop personalized vaccines that target the specific antibodies that are observed in a particular patient's blood sample.
The developers also claim that their method improves on current efforts to identify bNAbs. Some methods work by introducing point mutations into the HIV virus that knock out certain regions, Georgiev said. Others introduce antibodies into the samples that have known epitope targets, analyze their effect, and compare it with the effects of antibodies that were already present in the sample.
However, these methods do not easily yield specific information about the HIV bNAbs present or the parts of the virus they targeted, according to the NIH researchers.
"In some cases, other methods may not be as specific due to inherent limitations of the mapping experiments," Georgiev said. "For example, mapping experiments that rely on introducing point mutations within a given epitope may affect antibodies of several different specificities that target different epitopes within the same general site on the virus."
In contrast, "the serologic algorithm appears to provide specific information about the viral epitopes targeted by antibodies in serum," he said. "In cases where the samples primarily contain antibody types of known specificity, the serologic algorithm can be an effective alternative to traditional mapping methods."
The serologic algorithm is also useful in cases where only small amounts of blood are available. Other methods typically require relatively larger quantities of blood for bNAb analysis, the researchers said.