NEW YORK (GenomeWeb) – Using a combined genomic-proteomic approach, Stanford University researchers have identified a group of neoantigen peptides in human mantle-cell lymphoma (MCL).
Published this week in Nature, the study looked at tumor neoantigen peptides presented by the major histocompatibility complex (MHC) class I and class II molecules in 17 patients and found that all the neoantigen peptides they discovered came from the immunoglobulin variable regions and, with one exception, were presented by MHC class II molecules.
This finding suggests that, at least in the case of MCL, there could be commonalities in neoantigen production and presentation across patients that could aid in the development of immunotherapies for the disease, said Ash Alizadeh, an assistant professor of medicine at Stanford and senior author on the paper.
He noted that the study results run counter to other recent investigations into neoantigen production, which have found little overlap among patients. For instance, a study published last month in Genome Medicine by researchers from Foundation Medicine and Roche Pharma Research & Early Development concluded that a universal cancer vaccine using tumor neoantigens to stimulate an immune response would be effective in only around 0.3 percent of the population.
"But what we found is that in every patient [the Nature study looked at] you recover [neoantigen] peptides from the immunoglobulin gene," Alizadeh said. "That's never been described before."
MHC molecules present peptides from foreign proteins to T cells, which then triggers the body's immune response. Researchers are working to use this system for cancer therapy by identifying cancer-specific neoantigens that they can use to stimulate an immune response to a patient's tumor.
If, as the Genome Medicine paper suggested, cancer patients' MHC molecules typically display different repertoires of tumor neoantigens with little overlap across patients, then clinicians will likely have to develop highly patient-specific therapies. If, on the other hand, there are neoantigens that are commonly present across patients — as the Stanford research indicates could be the case in MCL — more universal treatments could be possible.
One challenge facing the field is identifying neoantigens that are presented by a patient's MHC molecules. Typically, this has been done through sequencing, with researchers first identifying mutations and then using informatics approaches to predict which of these mutations would likely lead to neoantigen peptides that the immune system would recognize as foreign.
These prediction approaches, however, "are not that accurate," Alizadeh said, adding that researchers' ability to identify which DNA sequences would go on to produce a peptide presented by an MHC molecule was "just a little better than tossing a coin."
Rather than rely on prediction from DNA sequences, the Stanford team used a proteogenomic approach in which they isolated MHC-bound peptides and analyzed them using mass spec on a Thermo Fisher Scientific Orbitrap Elite instrument while at the same time using exome sequencing to generate patient-specific sequence databases for making their peptide identifications.
While still not commonplace, this sort of workflow has seen increasing uptake among proteomics researchers in recent years, especially in the cancer proteomics space where initiatives like the National Cancer Institute's Clinical Proteomic Tumor Analysis Consortium (CPTAC) have applied it to in-depth tumor profiling.
It remains relatively novel for neoantigen profiling, though, Alizadeh said, noting that so far as he knew, the Stanford group was the first to try it when they began the project several years ago. Since then, several other teams have taken a similar approach, but, he said, it is limited by the relatively large amount of tumor tissue required to identify low-abundance neoantigens by mass spec.
Alizadeh said that his group had access to large amounts of MCL samples, "which gave us a unique opportunity in the tissue we looked at." He added that other groups have managed to take a similar approach to the study of neoantigens in melanoma, which, he said, "suggests that it could scale to other [cancer] types."
In perhaps an indication of the challenge of predicting informatically the DNA sequences that will produce MHC-presented peptides, Alizadeh and his colleagues found in their study that mutations in non-Ig genes, including in genes known to be commonly mutated in MCL were not presented as neoantigen peptides. "Among expressed genes bearing coding mutations, we found 46 percent with at least one peptide presented by either class I or class II MHC," the authors wrote. However, they added, with the exception of the Ig-derived molecules, these presented peptides came from the non-mutated portions of the gene.
With regard to the presented Ig peptides, they varied across patients, but, Alizadeh said, "there are clearly some hotspots."
"And in those hotspots, the sequence is not identical in every patient, but we can identify shared features," he said. "The next step is, can we design immunogens that would be shared across patients. Or can we find a T-cell receptor that will not just attack one person's antibody but another person's antibody that happened to share similar features."
Alizadeh and his colleagues are also investigating whether other forms of lymphoma and other cancers outside of lymphoma exhibit similar behavior around neoantigen presentation.
They are also looking into whether the findings might represent a function of the standard biology of B cells (which express MHD class II), as opposed to tumor biology, Alizadeh said. "Is what we're observing a cancer phenomenon or is it a B-cell phenomenon and how a B cell needs to deal with its immunoglobulin gene to survive?"
Either way, the phenomena could be useful therapeutically, he said. "We use relatively blunt instruments in treating patients with lymphomas that wipe out all these cells. So even if this mechanism is something that is shared by B cells, I think it could still be therapeutically very interesting. And immunologically it would be a cool thing to figure out."