Let's spread the word about Immunotherapy! Click to share this page with your community.

Building a Single-Cell Atlas with Dr. Ansuman Satpathy

May 27, 2020

Over the past decade, cancer immunotherapy has transformed our ability to treat more than a dozen types of advanced cancer. The immune system’s power may very well enable us to conquer all forms of the disease ultimately, but first we must improve our understanding of how it works in the context of cancer. To that end, Cancer Research Institute (CRI) scientists are seeking answers to critical questions, like: “Why do some people respond to immunotherapy but not others?”

Dr. Ansuman SatpathyAnsuman Satpathy, M.D., Ph.D., at Stanford University, is a former CRI postdoctoral fellow who is now a CRI Technology Impact Award recipient. He is building a single-cell atlas to map the genetic factors that contribute to immunotherapy’s success or failure when treating cancer patients. Ultimately, using state-of-the-art genome editing technology, Dr. Satpathy aims to figure out how we might improve cancer patient responses to treatment with immunotherapy.

Recently, we spoke with Dr. Satpathy to learn more about his progress.

Arthur N. Brodsky, Ph.D.:

Right now, the most widely used type of immunotherapy is checkpoint immunotherapy. Most of the most effective types target the PD-1/PD-L1 pathway and are designed to prevent cancer-fighting T cells from being prematurely shut down—so that they can finish the job of eliminating cancer.

In your current CRI-funded work, you’re analyzing T cells from patients who didn't respond to checkpoint immunotherapy and trying to figure out which mechanisms might be responsible. Ultimately, your goal is to use all of this information to build what you call a single-cell atlas. What is that and how might it help us improve our understanding of immunotherapy?

Ansuman Satpathy, M.D., Ph.D.:

To build the single-cell atlas, we really need to build the technologies that enable us to analyze samples taken directly from patients who are getting checkpoint blockade immunotherapy.

A major problem in the past when it comes to understanding some aspects of the molecular wiring of a cell is that existing technologies can be used to study that wiring in cells in the lab or in a large collection of cells, but you really couldn't do it in cells taken from a patient’s samples, and certainly not in subtypes of cells, like T cells.

Our first step for us was to develop a genomic toolkit, a bunch of technologies to understand the molecular wiring of each individual cell, including each T cell, in a patient’s tumor. This wiring comprises several layers of information, too, which we can then integrate together.

Arthur N. Brodsky, Ph.D.:

This could then give you information on the identity of individual cells, instead of analyzing a bunch of different cells at the same time?

Ansuman Satpathy, M.D., Ph.D.:

Right. In the past, we’ve really only been able to look at tumors as a whole. We take the biopsy tissue and mix it all together without separating its individual components, and then do some analysis on what comes out the other end. But the problem with that is that cells are different—heterogeneous—so you get sort of a population average of all of those cells instead of the individual pieces that make up that mix.

With T cells, if you have a dysfunctional T cell population in a tumor, by definition they are not going to proliferate as much and will come to represent a relatively small fraction of the T cell pool within that tumor. So, with this bulk measurement, you might miss all of that signature, because you’re just getting sort of the average of all of the functional cells and missing the key piece of the puzzle.

With a single-cell atlas, we hope to be able to be able to map the wiring of these different T cell populations, no matter how rare they may be within the tumor.

Arthur N. Brodsky, Ph.D.:

Right now, you’re using this technology to analyze T cells from patients with a type of skin cancer called basal cell carcinoma who were treated with checkpoint immunotherapy. What have you learned so far about why some of these patients might not be responding?

Ansuman Satpathy, M.D., Ph.D.:

So, we picked this type of skin cancer because it responds well to checkpoint blockade, relatively speaking. About 50 percent of patients who get PD-1 immunotherapy respond, and about 30 percent experience long-term relief. We've been trying to ask what's different about the immune response in patients who respond versus those who don't. One of the things we've found is that, like in other cancers, patients who don't respond have T cells that are exhausted and dysfunctional, which can limit how well PD-1 immunotherapy works.

Arthur N. Brodsky, Ph.D.:

One way that you’re trying to understand the T cells in these patients, and what causes their dysfunction, is by characterizing them at the epigenetic level—how genes are expressed. Can you explain how this might help improve immunotherapy results?

Ansuman Satpathy, M.D., Ph.D.:

The way a cell regulates which genes it expresses is based on two things. The first is the actual DNA sequence of the genes. If you have certain genetic mutations, you might have high expression or a low expression of a gene. The second is through what we call epigenetics. And what that means is how the cell actually reads and expresses that genetic information.

Every cell in the body has the same genetic information, but obviously every cell is not the same, right? In your body, you have neurons and skin cells and T cells and all kinds of other cells. What differentiates how precursor cells become those different cell types is not the genetic information. That's the same in each cell. It's how each cell is actually reading that information and turning genes on and off.

This regulation of gene expression is what we call epigenetics. It’s controlled at several layers, including whether the DNA portion for a certain gene is physically accessible, which is required for it to be turned on and expressed.

Arthur N. Brodsky, Ph.D.:

You’re also characterizing the T cell receptor repertoire in these patients. In other words, trying to figure out which tumor-associated markers the T cells are designed to target and bind to, right?

Ansuman Satpathy, M.D., Ph.D.:

Yes, and this ability to analyze T cell receptors, or TCRs, is really valuable because it tells us two things about a T cell: which antigen that T cell recognizes and responds to, and how to trace that cell’s lineage. The breadth of possible T cells is so vast that any two cells that have exactly the same TCR sequence can be assumed to come from the same cell after it was activated by an antigen.

By combining this information from all of a person’s T cells, we can determine the overall diversity of their T cell pool. And if you look at the same patient both before and after therapy you can follow the T cell responses over time. So, if we see the same TCRs in both samples, we know that that T cell existed before therapy, and we can ask how these T cells looked and whether they were functional or dysfunctional. We can also ask if they target an antigen that is expressed by the tumor.

Then we can look at the T cells again after treatment to understand how checkpoint immunotherapy or how is the tumor affecting the function of this T cell over time. That's one of the really important pieces that we’re trying to uncover, and we also want to pair that information with the epigenetics.

In the basal cell carcinoma patients who were treated with immunotherapy, for example, we analyzed the T cells in each person's tumor before and after treatment and were able to identify populations, based on their behavior and by what their T cell receptors target as well as their epigenetic state.

Ultimately, we found T cell populations that were exhausted pre-therapy, and then we looked post-therapy to see if PD-1 immunotherapy reactivated them or not. Interestingly, the T cells were still exhausted, so PD-1 really didn't rescue their function.

So, then where's the response coming from? What we saw, at least in basal cell carcinoma, is that this immunotherapy instead recruits a huge new army of T cells to come into the tumor. That’s what appears to be responsible for patient responses, rather than reinvigorating the exhausted T cells.

That’s an example of what we can do. With all of these pieces of information at the single-cell resolution, we can start to understand the fundamental mechanisms of how a T cell is wired and how it responds to checkpoint immunotherapy.

Arthur N. Brodsky, Ph.D.:

Is it possible that some of these insights might help improve how patients are treated with current approaches?

Ansuman Satpathy, M.D., Ph.D.:

That's a good question. I do think the answer is yes. One of the ways that you can envision doing that is by understanding the early immune response after a given therapy. If we think that PD-1 immunotherapy works by recruiting new T cells, we can use these technologies to track that T cell response in a patient and determine if we’re getting the optimal response or not. Maybe it tells us that we need to change the dose or stagger it in some way or combine it with another existing therapy with complementary immune action.

Arthur N. Brodsky, Ph.D.:

By enabling us to learn more about the intricacies of T cells and how their wiring affects their behavior, this technology may also help guide the development of the next generation of immunotherapies. To highlight this potential, you’ll be using genome engineering to re-wire T cells to hopefully make them more effective. How might this capability improve future immune-based approaches?

Ansuman Satpathy, M.D., Ph.D.:

Up until now we've talked about profiling the cell and understanding the existing molecular wiring that occurs in T cells within the tumor, especially as it relates to T cell dysfunction or exhaustion. One of the longer-term goals of this project, though, is to use this technology to guide genome engineering strategies to reverse that dysfunction.

Once we’ve defined the distinct epigenetic state of a given type of T cell, like exhausted T cells, we can use that knowledge to manipulate those cells to see if we can reverse that cell state and the exhausted behavior. And once we identify each of the important factors in the epigenetic profile, we aren’t forced to examine their roles one by one. We can ask these questions simultaneously using a high-throughput, single-cell platform. Now, we can perturb or disrupt a thousand or five thousand different factors or genes, and then measure what's happening at the epigenetic level and how it affects the overall behavior of the T cells from these patients.

You can imagine how this could also be useful when it comes to improving our ability to design combination strategies involving checkpoint immunotherapy and could also aid the design of the next generation of cell therapies, like CAR T cells.

Arthur N. Brodsky, Ph.D.:

Absolutely. There definitely seems to be a very bright future ahead. Of course, none of this would have been possible if not for decades of investments in basic immunology research, much of which has been supported by CRI.

How important has CRI’s support been, both for you personally and for the historical advancement of the field of cancer immunology?

Ansuman Satpathy, M.D., Ph.D.:

I completely agree. The wonderful thing about CRI is that they have supported fundamental research in areas long before they were in vogue. This was definitely the case with cancer immunology, and I think this is true in my case as well. CRI was the first to fund my postdoctoral work, several years before it was apparent that genome technologies would become useful for understanding the immune response to cancer. I am very grateful once again that they have funded my independent laboratory to go after the next hurdle. This support will continue to allow us to think outside of the box to envision what could be possible for the future of cancer immunotherapy. 

Support cancer immunotherapy research

Be in the know with the latest immunotherapy news

*Immunotherapy results may vary from patient to patient.

Top