Dr. Govinda Sharma is a postdoctoral fellow working in the laboratory of Dr. Rob Holt at BC Cancer and the Michael Smith Genome Sciences Centre. During his graduate studies in the same lab, he worked to develop a new platform technology for the identification of functional T cell receptor epitopes. This platform is now the basis of a spin-out company called Immfinity Biotechnologies. We sat down with Dr. Sharma to discuss his research and the launch of the company.
What are CD8+ cytotoxic T cells, and what is their mechanism of action?
CD8+ cytotoxic T cells are a subdivision of T cells, which themselves are a subdivision of the adaptive immune system. We’re interested in studying them because they’re especially able to directly kill cells in the body that have gotten infected by an intracellular pathogen or are harbouring mutations that could lead to cancer. They do this by using T-cell receptors (TCRs), which are structurally unique proteins expressed by individual T cell clones, to inspect short peptide antigens that are derived from normal turnover of proteins inside the cell and presented on major histocompatibility (MHC) molecules on the outside of the cell. If a patrolling CD8+ T cell recognizes one of these peptide antigens as “non-self”, they engage the granzyme-perforin pathway. This means that the T cells release the contents of their lytic granules, which are little organelle sacks filled with apoptosis-initiating proteases (called granzymes), very specifically into the desired target cell while sparing healthy bystander cells.
What role do CD8+ cytotoxic T cells play in cancer?
T cell receptors are amazingly tuned to detect even single amino acid changes in normal self-proteins, which makes them very well suited for seeking out and destroying cells containing mutations. In cancer, CD8+ cytotoxic T cells are the members of the immune system that correlate most strongly with positive outcomes. In other words, if you have a lot of these cells infiltrating your tumour masses, you tend to live longer. To this end, many strategies have been devised to try and boost the T-cell response and help it overcome the growth of malignant tumours. This is the basis of a number of different types of immunotherapy, including checkpoint blockade drugs, vaccines, immunomodulatory agents, and cell therapies.
Can you explain the process of T-cell therapy in more detail?
The therapy works by extracting T cells from a patient’s body and genetically modifying them in the lab by adding an antigen receptor specific for a tumour target. At the moment, the most common type of gene-modified T cell therapy in clinical trials, and, indeed, the only approved cell therapy for cancer treatment, is called CAR-T therapy. In this case, a chimeric antigen receptor (or CAR), which is a type of synthetic immune receptor, is used to target the therapeutic to the tumour. However, other types of antigen receptor genes can be used to modify T cells too, including TCRs just like the ones that already exist inside of every person – this format is called TCR-T therapy. The key advantage of TCRs over CARs in the context of cell therapy is that they can recognize intracellular antigens, which means that they can “see” a wider range of potential targets than CARs can, but it’s difficult to profile TCR antigens so there hasn’t been as much focus on TCR-T therapies as CAR-T therapies in recent years. In either case, these receptors act as the guidance system for these cytotoxic killers. Once they have been propagated in the lab, this huge population of T cells can then be reinfused back into the patient with the goal of “surprising” and overwhelming the tumour.
Is cell therapy FDA approved?
Anti-CD19 CAR T-cell therapy is the only approved therapy, which is available for a few different indications in leukemia, and it has shown fantastic results in the patients who are eligible to receive it. Unfortunately, this is just a subset of a subset of cancer patients. This has motivated groups around the world to develop new cell therapies in order to make these “living medicines” available for a larger chunk of the patient population. In 2018 alone, more than 10 billion dollars was invested worldwide into developing new cell therapies for cancer treatment. There are currently over 1,000 cell therapies currently under development globally, but we’re still waiting for the next ones beyond anti-CD19 CAR-T to be approved.
How does this relate to your research?
In 2012, I began working on my PhD under the supervision of Dr. Rob Holt at BC Cancer and the Michael Smith Genome Sciences Centre. At the time, the lab was doing a lot of sequencing of TCRs. But we soon realized that knowing the sequence of a receptor doesn’t tell us much about what antigen it recognizes. In silico prediction tools can allow you to shortlist possible T cell antigens to a few hundred or a few thousand to test using classical methods, which is basically the practical limit of these techniques, but when you consider how many possible T-cell antigens could exist, it’s clear that we’re not searching deeply enough. The other issue is that in silico prediction looks at the binding of peptides to MHC molecules as a proxy for immunogenic activity; the historical wisdom is that the stronger a peptide binds to MHC, the more likely it will trigger a TCR – but this isn’t totally true. There’s a lot of complicated biophysics and signaling pathways behind T cell activation, so directly functionally testing T cells against real cells that are processing and presenting antigens is essential. We wanted to overcome the challenge of assessing antigen space deeply but doing it without losing all that complex cell biology behind the TCR-peptide-MHC interaction.
How did you go about developing this method?
We aimed to build a reporter system that could sense granzyme-B delivery, since that is such a specific early marker of T-cell targeting. After some trial and error with different configurations, we settled on a fluorogenic reporter system consisting of Cyan Fluorescent Protein (CFP) and Yellow Fluorescent Protein (YFP) fused together using a granzyme B-cleavable peptide linker. When cells expressing the intact fusion protein are hit with a violet laser, this pair of fluorescent proteins undergoes Förster Resonance Energy Transfer (FRET) to produce yellow light. If the reporter is cleaved by granzyme, the cell switches from yellow to blue, which allows you to sort out cells that are actively being targeted by the T cell population-of-interest using FACS.
If we encode a library of short peptide-coding DNA sequences, which we call minigenes, in tandem with a copy of the reporter gene into host target cells by lentiviral transfer, we can create a bulk population of cells where every cell has a different minigene sequence inside. Once we do this, we can then expose our library of target cells to T cells containing a TCR whose reactivity we’re interested in profiling. Cells that have an antigen that is recognized by the TCR of interest encoded in their minigene are subject to granzyme-B loading and undergo that yellow-to-blue shift in their fluorescence properties. We recover those cells by FACS and do deep amplicon sequencing to identify which minigenes were responsible for triggering the T cell response. So far we’ve built libraries of over a million minigenes and shown that our method can still easily pick out antigens from this super diverse background.
How do you foresee this platform being used to advance cancer cell therapies?
To answer that question, we need to discuss the cross-reactivity of TCRs. T-cell receptors are not like their immune receptor cousins, antibodies, which undergo affinity maturation to become specific for only one antigen. Each individual TCR can recognize many, many different antigens, which is how such a limited number of cells in the body can protect us against such a vast universe of possible threats. What this means for researchers trying to make new cell therapies is that we need to stop thinking about identifying the target of a TCR. Instead, we need to start thinking about profiling the repertoire of targets that a single TCR recognizes. This is what our technology can do. For example, we estimate that to cover all the peptides encoded by the human proteome, we would need to screen a million different sequences in parallel – and we’ve shown that our platform can achieve that scale.
The example of building a library of minigenes that codes for all the peptides in the whole human proteome is especially important in developing TCR-guided cell therapies, because it’s possible that designer therapeutic TCRs that have fantastic reactivity against a desired tumour target could also have really dangerous cross-reactivities with healthy tissues. If you are a company or institute with a number of potential TCRs that you’re interested in bringing to clinical trials, you would likely want to know about these cross-reactivities before investing in the very expensive pre-clinical and clinical studies needed to advance these therapeutics. We think our technology can help cell therapy developers by providing the ability to de-risk or re-prioritize their therapeutic TCR candidates up front.
So these cross-reactivities could be quite dangerous?
Absolutely. Back in 2013, two UK-based pharma companies and a US-based medical centre collaborated to take a TCR that had some known reactivity to a known tumour-associated antigen, and engineered it to enhance their affinity for the antigen. They tested the cross-reactivity of these modified TCRs against a panel of laboratory human cell lines and fresh tissue samples. The T cells passed all the tests, so they went ahead and infused them into two patients. Both patients died within hours. It was incredibly traumatizing. When they investigated the deaths, they found that the modified TCR that they engineered cross-reacted with a naturally expressed protein in the heart (Linette, et al. 2013). Unfortunately, with the tools available to them at the time, they had no way to see this coming.
Looking back, it makes sense that this cross-reactivity wasn’t caught by any of the cell lines or tissue panels. First off, there’s no common laboratory cell line that is derived from heart tissue. Plus, this protein is only expressed in the beating heart, so it wouldn’t have been present in a block of banked heart tissue either. This underscores the obvious problem that it is very difficult to create a tissue panel comprehensive enough to be an adequate stand in for a living human body.
If they had had access to a discovery assay that would have spotted this cross-reactivity, these catastrophic autoimmune reactions might have been avoided. We have been in communication with these two companies, and they are very supportive of our research. We’re actually working on testing the TCR that they used in the 2013 study to see if our platform would have identified the heart protein as an antigen. I’m confident that it will.
When will the company, Immfinity Biotechnologies, be launching?
It’s early days. We’re incorporating now and we still have lots to do before we can ‘go live’. Our goal right now is to continue to assemble our team, refine our business plan, and try to scrape some funding together. It’s an exciting time!
Thank you for taking the time to discuss your research, Dr. Sharma! We wish you and Immfinity Biotechnologies the best of luck!