Viruses That Mimic Human Proteins May Be More Common than Previously Thought

Distinguishing between self and non-self is a critical ability of the immune system. Some pathogens have evolved proteins that resemble those of their host, a mechanism called molecular mimicry, in an attempt to evade this surveillance system. However, immune training isn’t perfect, so host cells that recognize self-proteins occasionally slip into the milieu. Their presence doesn’t immediately raise any alarms, but if a pathogen with this same mimicked protein invades the body, it can trigger an immune response against the same sequence on the body’s own cells.

Previous work on molecular mimicry focused on the 3D structure of proteins, which is important for antibody-mediated immune recognition, but T cells typically recognize short, linear peptide sequences.¹ Studies of mimicry in these short sequences often focus on a subset of viruses or suspected proteins.^2,3 Now, researchers have performed a more comprehensive analysis, screening more than 100 human-infecting viruses for mimicry in short, linear peptide sequences.⁴ They demonstrated that this phenomenon is widespread across the virome. These findings, published in Nature Communications, could help researchers explore novel autoimmune triggers and potential interventions.

Esther Melamed, a physician scientist and neuroimmunologist at the University of Texas at Austin, and her group study how molecular mimicry impacts the development of autoimmune diseases. One in particular that her group is interested in is multiple sclerosis (MS), which has a strong association with Epstein Barr virus (EBV).⁵ Melamed said that her team was interested in exploring the contribution of other viruses.

When the COVID-19 pandemic forced their group to work from home, it introduced a perfect opportunity to explore this question since these sequences are available online. Cole Maguire, a neuroscience graduate student in Melamed’s group, first screened the SARS-CoV-2 virus for molecular mimicry. “Once we had it working, we're like, ‘Well, what if we just run it for every virus that we can’,” he said.

“We don't necessarily know why viruses gain [mimicry] and what advantages they have from gaining it,” Maguire explained. “This paper kind of takes the first step at looking at that, which might be useful in predicting future mutations.”

Using consensus sequences from 134 human-infecting viruses and the annotated human genome, the team looked for viral mimicry to linear sequences of eight, 12, or 18 amino acid residues that had no more than three mismatches in these epitopes. They found that viruses from the same family shared mimicked epitopes and that viruses that cause chronic infections demonstrated more mimicry compared to viruses that cause acute infections. Two families, Herpesviridae and Poxviridae, which include viruses like herpes simplex virus 1 and Mpox, demonstrated some of the highest rates of molecular mimicry.

“Even if you look at other autoimmune diseases [beyond MS], they're really associated with a variety of herpes viruses. So, seeing this trend at a family level was interesting because it might point to one reason why herpes viruses often associate with autoimmunity,” Maguire said.

“We wanted to really look at the biological effects and try to figure out how molecular mimicry fits into the human body as a whole,” Melamed said. To do this, the team analyzed the pathways that these mimicked proteins came from and found that many were involved in cellular replication and inflammation, but that there wasn’t a preference for a specific cell or tissue type.

Then, they investigated which chromosomes encoded the mimicked proteins. “One of the most interesting things to us was that, when we looked at chromosomes across the genome, viruses tended to avoid mimicking proteins encoded on the Y chromosome,” Melamed said, adding that this finding may indicate evolutionary pressure on the virus to mimic more universal proteins.

Because T cell immune training against self-proteins occurs in the thymus, the team compared the genes of the viral-mimicked proteins to those sequences that are expressed by thymic cells. They observed a large overlap of viral-mimicked sequences and those of thymic genes.

Finally, although MS has traditionally been characterized as a T cell mediated disorder, B cells have recently been shown to contribute to the autoimmune condition.⁶ The team explored the prevalence of MS autoantibodies targeting peptide sequences that are mimicked by EBV. They showed that the majority of the most common MS autoantibodies included an eight-residue sequence that matched an EBV sequence. “So, for the antibodies that do exist, EBV actually plays a big role in contributing to the generation of those antibodies,” Melamed said.

Ana Beatriz DePaula-Silva, an immunologist at the University of Utah who studies how viral infections contribute to neurological disorders and who was not involved with the study, said that it will be important to follow up this work with functional studies to determine if all of the mimicked sequences play a role in autoimmune disease development. However, she said that it was interesting to see the extent of mimicry in viruses. “It brings a lot to the field,” she said.

“I think the scale of what we found was that there are so many chronic viruses that can cause molecular mimicry really suggests that there's a larger problem at hand that we don't really think about when we see patients clinically with autoimmune diseases,” Melamed said.

References