American Scientist

The human immune system is a marvel of complexity, coordinating the activities of many different cell types to try to protect the body from invaders. There are cases when it might overlook hostile forces, such as cells that have become cancerous, or when it overzealously attacks tissues that it should recognize as belonging to the body, as happens in diseases that range from rheumatoid arthritis to diabetes to Parkinson’s. Alessandro Sette, Head of the Division of Vaccine Discovery and the Center for Infectious Disease and Vaccine Research at La Jolla Institute for Immunology in California, has spent decades unraveling the immune system’s functions and uncovering ways to harness it for better disease prevention and treatment. One of his focuses has been on epitopes, the specific fragments and structures that the immune system recognizes. He has overseen the design and curation efforts of the national Immune Epitope Database (IEDB). Sette is the recipient of Sigma Xi’s 2025 William Procter Prize for Scientific Achievement. He spoke with editor-in-chief Fenella Saunders about his work. (This interview has been edited for length and clarity.)

How did you get interested in studying immunology?

I really fell in love with the study of the immune system for a couple of different reasons. In my university studies in math and chemistry, I was very reductionistic, I wanted to go down to the molecule. And immunology was, of all the biological sciences, one of the disciplines that was really starting to describe events in biochemical terms. You could measure very accurately antibody-antigen interactions. We were starting to understand the genes linked to antibody rearrangement, and also the whole mystery at that time of what T cells recognize and how they recognize it. The other big thing is that it had real potential for applied science. I’d always been very interested in pursuing avenues with my research that could have an impact, and immunology was a science that was clearly very close to human health.

How has the study of immunology advanced since the start of your career?

Things have changed so much that it’s hard for someone today to appreciate. Back when I was starting, to sequence a gene or synthesize a peptide was a big deal. Forget about having genomes. I would literally synthesize peptides by hand, one residue at a time, come back in two hours, add another residue, and in a few weeks you’d have a peptide. Just the amount of data that can be generated, the amount of granularity, has increased by orders of magnitude.

Can you explain how different parts of the immune system interact?

The immune system is made of different components that work together in a synergistic way, but they are very different in the ways they recognize a substance, an antigen. Antibodies recognize proteins on the outside of a virus or a bacteria, and they bind to it to neutralize its activity, so they are great at preventing infections.

But what if the microbe makes its way inside our cell? Then it becomes invisible to antibodies. That’s where the T cells come in, because they have an uncanny capacity to recognize fragments, or epitopes, of a virus or a bacteria. What happens is that when a virus or a bacteria gets inside the cell, it gets broken down into pieces. These pieces end up being bound to certain proteins called HLA [human leukocyte antigen] molecules. These proteins, with a fragment bound to them, end up on the surface of a cell, and that’s how a T cell can tell that that particular cell is infected. The T-cells kill the infected cell and secrete other proteins called cytokines that amplify and potentiate the immune response to favor the production of antibodies.

T cells are not particularly good at preventing an infection, but they are what is required to terminate an infection and prevent disease. Antibodies are great to prevent an infection. So that’s why they work together.

How are antibodies generated?

The antibodies are made by B cells, but B cells become activated if the antibodies that they express on their cell surface bind the protein or the surface of the virus. Then that particular B cell that bound the target starts to proliferate, rearrange its genes, secrete antibodies, and mount an immune response.

How can you develop therapies inspired by immune responses?

In whatever disease we are interested in, we start from the organism. It may be someone who has been unlucky enough to get the flu. But then we zoom in and say, okay, something is going on in the lung. In the lung, which cells of immune systems are at play? Which cell is infected? Which cell is responding? What is being recognized as a virus, and of that virus, which particular protein? And of that particular protein, which specific fragment? So we go all the way down to a fragment of a protein with a sequence of a few amino acids. Once we have that, then the journey starts in reverse. Now we can use this fragment as bait, to fish out the different cells that are recognizing it. And then we ask: What do they do? What kind of activation? Are they exhausted? Are they proliferating? Are they secreting one particular mediator or another?

We are interested in a whole series of diseases from viruses and bacteria. Those are instances in which you benefit, most of the time, from the immune response. Those responses are amplified by vaccines. Another instance in which you benefit from immune response is when you react against a cancer.

“You would like to induce the type of immune response that is associated with smelling flowers all day, as opposed to an asthma attack.”

• Facebook
• Twitter

There are other instances in which the immune response is not so welcome. One example is allergies, where you violently react to smelling a flower and inhaling that pollen, and it’s really no harm to you. Other instances of an unwelcome immune response are diabetes or rheumatoid arthritis, where your body somehow has an immune response directed against components of your own body. An area where we are working a lot these days is neurodegenerative diseases, such as Parkinson’s, Alzheimer’s, and ALS [amyotrophic lateral sclerosis]. Again, you have an inflammatory phenomenon that can maybe trigger the disease.

In all these different cases what we try to do is to use the fragment to ask questions. What’s the difference between good and bad outcomes? Why is that person reacting that way and not in another way? You would like to induce the type of immune response that is associated with smelling flowers all day, as opposed to having an asthma attack and ending up in the emergency room.

How can this research help to stop an overactive inflammatory response?

For neurdegenerative diseases such as Parkinson’s and Alzheimer’s, the disease takes a long time to take hold. The clinical effect of Parkinson’s, the shaking and the impairment of cognitive function, is a result of a certain area of the brain called substantia nigra, which makes dopamine, having been destroyed. But that process takes decades. It’s very similar to what happens in diabetes: Unwanted activity of T cells gradually destroys the cells in the pancreas that make insulin. If we could recognize this process, then that inflammation, that immune response, could be shut down.

There’s some interesting data on people who have Crohn’s disease or irritable bowel disease and are treated with monoclonal antibodies that essentially shut down a certain inflammation mediator. These medications have been used in thousands of people over the course of decades. People who get this treatment when they’re younger don’t get Parkinson’s. That means that if you can shut down this inflammation, you can stop Parkinson’s in its tracks. Now, that doesn’t mean you’re going to put everybody that turns 50 on these medications: they have side effects, and it’s expensive. But if you had a test to recognize people that are at risk, that would be a call to arms to intervene.

What are the prospects for applying your results to personalized medicine?

The process of cancer epitope identification can be very personalized. Your normal cells accumulate mutations, and those mutated fragments are not normal parts of your body, so they are potential candidates for your immune system to recognize. The particular HLA molecules that present this fragment to the immune response are different from one individual to the next. And particular mutations most often are different from one cancer to the next in one individual. So that’s why people are interested in making personalized cancer vaccines or targeted immunotherapy.

On the opposite end of the spectrum, if I want to develop a reagent to monitor the type, the magnitude, or the quality of immune response induced by a vaccine, it’s not convenient to try to identify the specific epitope on an individual basis, because what you want is something that works in the population that has been immunized with a vaccine. The other complication is that these HLA molecules can be different in different ethnic groups. The U.S. population is very ethnically diverse, so you want to make sure that your reagents will work in all the different ethnic groups that are represented in the population.

The other issue is the issue of conservation across relative species. And that actually is also an important consideration that can drive the immune response. One of the first times that was clearly apparent to us was when we were studying allergic reaction to pollens. Obviously there’s a lot of plants and a lot of different types of pollens, but many plants are related to one another. If someone is allergic to grass, there’s a lot of different types of grass. What happens is, a person is breathing ragweed or timothy grass or another grass, and it happens that certain fragments are going to be shared, or be very similar, between timothy grass and other grasses. If you have an immune response that generically sees timothy grass, but then you breathe another grass, your immune system will get a boost for both fragments that are shared between the two different types of grass. Over time your T cell response will be skewed in recognizing things that are similar or shared in different grasses. The same thing happens for bacteria and viruses. To a certain extent, this reaction also generates some kind of a background, preexisting immune reactivity that is often, in the context of infectious diseases, helpful. In the case of SARS-CoV-2, there’s a lot of data that showed that people who had been recently exposed to other coronaviruses had less severe disease manifestation. This effect is particularly important when there is an infectious agent that is completely new. That, for example, is widely believed to have been the case in the 1918 flu pandemic. This flu was something totally new that the human population had no pre-existing immunity against, and that’s one of the reasons why it was as damaging as it was.

What has the role of databases been in advancing this research?

We have been running the Immune Epitope Database (IEDB) since 2003. We also have launched a sister database called CEDAR, and that is specifically hosting all data related to cancer epitopes published in the scientific literature.

This freely available resource catalogs all known immune responses against viruses, bacteria, antigens, allergens, everything. That has been a passion of mine and my colleagues, making this data freely available. If it’s published in the scientific literature, we extract that data and we curate it. The key, very powerful thing is that the data is put in a context that is computer-readable. That way you can collate a thousand different studies and look for trends.

You can ask, is there a general rule that a certain position in this fragment is more immunogenic than another one, or are there particular regions of a protein that are more prone to have mutations that are recognized by the immune system? You can do a lot more, if you can collate data from tens of different independent studies. That also gives you more confidence that that result is reliable.