Interview with Chiara Bonini,MD

Editor’s note: Chiara Bonini, MD is a Professor and Group Leader of the Experimental Hematology Unit at Università Vita-Salute San Raffaele and was the recipient of the 2024 ESGCT Outstanding Achievement Award.

Thomas Gallagher: Could you take me through your background and how you came into the gene therapy field?

Chiara Bonini: Sure. I have been in the field for about 30 years. I started as an undergraduate student, and I was fortunate enough to have Professor Claudio Bordignon, who is one of the pioneers in gene therapy, giving a lecture in a class of immunology. He had just treated the first kid with adenosine deaminase-severe combined immunodeficiency—ADA-SCID—with gene therapy, administering autologous stem cells and T cells engineered with the retroviral vector encoding ADA. He illustrated to us the project starting from the concept, the possibility of finally having good vectors for transducing cells, and all the preclinical data, and then he showed us the results of the first patient treated. It was there at that point that I recognized exactly what I wanted to do; I had it in my mind that I wanted to try to adapt it to treat cancer patients. I asked him at the end of the lesson to give me the chance to learn in his lab. He wanted me to take the exam of immunology, and so a few months later I went back and said, “look, I got a good score in Immunology; now I’m here to try again,” and he opened the door of the lab.

OPEN IN VIEWER

Chiara Bonini, MD

At that time, the early 1990s, gene therapy was not really known that much; it was the dawn of this field. I graduated with a medical degree in 1994, and I remember telling my colleagues that my thesis was going to be in gene therapy and hearing them ask me, “is this something that has anything to do with medicine?” I have experienced the ups and downs, the enthusiasm and safety concerns, the successes such as with ADA-SCID and other diseases, and also the difficulty of encountering adverse events. Today, it is really a great moment to be a scientist in the field because the number of patients successfully treated is really high. Patients affected not just by rare but also by common diseases such as lymphoma, for example, or multiple myeloma, are starting to really benefit from having registered gene therapy-based products.

T.G.: Your work is specifically in gene therapy for cancer applications; are there any areas of this field that are particularly exciting to you?

C.B.: Today there are consolidated clinical results developed within both the ex vivo and the in vivo gene therapy approaches. Among the ex vivo gene therapy projects, the most advanced ones involve engineered T cells, and in particular, CAR (chimeric antigen receptor)-T cells. For the generation of CAR-T cells, cells from our immune system, most frequently (but not only) T lymphocytes, are genetically engineered with the purpose of exploiting their natural properties while conferring a given antigen specificity. In nature, T lymphocytes are able to proliferate, expand, and kill the target when they see their specific antigen. After antigen encounter, a fraction of T cells differentiates in a memory state and thus persist long-term. This is their own physiological nature, which is usually very effective in protecting us from infections. We try to exploit this nature and manipulate it through genetic engineering to adapt it for cancer treatment. For example, we can express a receptor that makes the cells able to recognize cancer cells, such as a CAR, and most of the T-cell-based cellular products currently in the market for cancer treatment are engineered lymphocytes expressing a CAR. These products reached the market because they showed their ability to successfully treat very aggressive diseases. These results were obtained at the end of a long scientific journey, with several lessons learned, often also from failures.

These are not the only gene transfer/genome editing-based treatment options, and I would say that I don’t expect CAR-T cells to be the cure for every tumor. They have some clear pros that we need to exploit and use, but also some minuses. Among the pros, there is their ability to recognize cancer cells in an HLA (human leukocyte antigen)-independent way, meaning that the same CAR can be used to treat a large number of cancer patients. However, CAR-T cells recognize only antigens expressed on the cell surface of cancer cells, and that can be a minus. If you look at which CAR-T became registered products, you see that four of them use CD19. CD19 is not a tumor antigen; it is not a molecule uniquely expressed by tumor cells; it is a lineage antigen that is expressed by B cell tumors such as leukemia and some lymphomas. But it is also expressed by healthy B lymphocytes, so with these CAR-T cells, we kill the cancer cells, but we also kill all the healthy B cells. And we can do that; we can live without B cells, so it’s a price that we are all happy to pay, but if you start to think of using the same approach for other cancers, you see that you cannot really kill also the healthy counterparts. For a disease such as acute myeloid leukemia, for example, the healthy counterpart is the hematopoietic stem cells, which we cannot live without. So, you cannot really use the same approach. I think that CAR-T cells represent a milestone for the field, but we need to adapt the lessons learned to make engineered T cells work also for other tumors.

T.G.: Can you share anything that is being done in your lab or what you see in the field that might be moving toward the market in the future?

C.B.: My first experience in the lab was with a project that reached the market, and that was really an important step in the development of gene therapy for cancer. In this first project, we wanted to insert a suicide gene into lymphocytes to make the cells sensitive to a prodrug. We used donor lymphocytes, not lymphocytes harvested from the intended recipient of a hematopoietic stem cell transplant. For some diseases such as acute myeloid leukemia, for example, in very high-risk forms, allogeneic hematopoietic stem cell transplant represents the only possible cure. Allogeneic hematopoietic stem cell transplantation was born with a regenerative purpose. With this therapeutic approach, we aim at killing the entire burden of leukemia with such an intense chemo-radiotherapy that we’re going to kill also the healthy hematopoietic stem cells. So we need to substitute the stem cells with those from a healthy donor. However, several clinical observations questioned the regenerative role of the transplant and showed that a large part of the efficacy of hematopoietic stem cell transplantation stems from the activity of donor lymphocytes infused with the transplant itself. Indeed, the immune system of the donor, and lymphocytes in particular, is able to recognize and kill residual leukemic cells of the patients. This immunological effect is named graft versus leukemia. However, these are donor cells and are thus also able to recognize several healthy tissues in the host, such as the skin, lung, and liver, giving rise to a very bad complication that we are still not always able to control: graft versus host disease (GVHD). So, we know that donor lymphocytes are very important to kill leukemia, but they can also be toxic, and the idea we had in the 1990s was to insert a suicide gene with a retroviral vector, which was the thymidine kinase (TK) gene from herpes simplex virus. We engineered our donor lymphocytes to express TK, and we then infused the cells in transplanted patients to induce the graft versus leukemia effect and also to protect the patients from infections. Then, when GVHD occurred, we could kill these cells because TK makes them sensitive to the prodrug Ganciclovir. So by giving patients Ganciclovir, we can kill only those donor cells that are mediating the complication without touching the rest of the immune system and all of the cells in the hematopoietic system that is engrafting and growing in the transplanted patient. This was a very simple concept, but at the time it was really pioneering and took 24 years from the concept to the market. Although transplant procedures changed dramatically in these years, making the TK approach redundant in the current setting, it remains a milestone from which we learned so much on how to sort the cells, how to expand them, and how to genetically modify them. At that time we used retroviral vectors; today, several alternative tools have been developed, including lentiviral vectors and genome editing tools.

For example, we learned that not every T cell comes with the same properties. T cells arise from the thymus as naïve cells, and, after antigen encounter, they differentiate into effectors and memory cells. In the first clinical applications, we used to administer cell therapy products highly enriched in effectors. By studying patients treated with TK cells, we learned that if we produce and engineer cells with early memory phenotypes, such as stem memory or central memory phenotypes, these lymphocytes will live longer in the patients and possibly patrol the host for cancer cells for a long time, and this information is now used to generate cancer-specific cells, such as CAR-T cells, with an early memory phenotype.

Recently, we applied the concepts and expertise acquired in the field of hematology to develop new therapeutic approaches for solid tumors. We are currently working on advanced colorectal cancer and pancreatic cancer. And not only with CARs but also with T-cell receptors (TCR) which are an alternative to CARs. TCRs are the natural molecules that usually allow T lymphocytes to recognize and kill their targets. TCRs recognize peptides that are presented by HLA molecules. Of course there are pros and cons of using a TCR versus a car: While CARs recognize only surface targets, a TCR can also recognize peptides that derive from intracellular molecules so we can dig harder inside the cancer cell to find the target. This also allows us to find antigens expressed by cancer cells, and not expressed on the healthy counterpart, a major pro of TCRs. Furthermore, by increasing the number of potential antigens, we can often also choose targets relevant for cancer cell survival and growth, thus reducing the probability of cancer immune evasion and relapse. Another pro is the fact that TCRs are highly sensitive, and they are the molecules that physiologically provide survival signals to the lymphocytes. Thus, we might have a more persistent response, a more persistent therapy with TCR engineering than with the CAR-T cells. But TCR are HLA-restricted, so once we have a TCR, we will be able to use it only for HLA-matched patients, and we shall need a large number of TCRs, restricted by several HLA molecules, to treat all possible patients, and this is a cons. An additional con of TCRs is that all T lymphocytes express their own TCRs. If we want to insert a tumor-specific TCR in a lymphocyte, we need to substitute the endogenous one with the tumor-specific one; we cannot just add the TCR against our tumor antigen, but we also have to take care that the endogenous T cell receptor is shut down. This has been an issue and a con for TCR-engineered lymphocytes compared to CARs for a while.

I actually started working on this in the mid-2000s; maybe it was 2006. I was at a Keystone meeting in Banff, and I had heard two wonderful lectures. One was from Steve Rosenberg, the pioneer of cancer immunotherapy with tumor-specific T cells, and he showed us how we could engineer lymphocytes. But with the TCR, he also showed that the expression of both endogenous and tumor-specific TCRs in the same cell could lead to mispairing because the TCR is made of two chains, the alpha and beta chains, so we ended up with two alpha and two beta chains in every possible combination. So he showed us what the hurdles were. Then, after his talk, there was a talk from Carl June, who showed us how he had used genome editing technologies to knock down a gene encoding CCR5 (C-C chemokine receptor type 5), a co-receptor for HIV (human immunodeficiency virus), and by doing this in T lymphocytes, he made T lymphocytes resistant to the CCR5-dependent HIV strain. I saw these two talks and reasoned that we needed to use genome editing technologies to knock down the endogenous TCR in the T cells. I was at this Keystone conference together with a colleague, Luigi Naldini, another pioneer of gene therapy, and I knew that he was starting to use genome editing applied to hematopoietic stem cells, so I asked him to collaborate on the project. Actually, I asked him if he thought that this idea was too crazy or something that might work. He listened to me, and he said, “I think it’s crazy, and so it’s worth doing it.” And then we started working on it together. I didn’t have a tumor-specific TCR in the lab, so I asked for a collaboration to Phil Greenberg, a great mentor for me and a pioneer of the use of antigen-specific T cells to treat infections and cancer. He generously accepted to give us the first TCR construct and to collaborate on the project. At that time, there was no CRISPR, so we worked with a biotech company, Sangamo Biosciences, who was a leader in the field, and we ended up preparing the first protocol of genome editing of T cells to knock down the endogenous TCR and insert the tumor-specific TCR. Today, with CRISPR, it is much easier to do this, and everybody does it, and you can even do a CAR T without a TCR, but at that time it was really a lot of work and a lot of fun, I have to say. This is where we’re going now.

T.G.: Are there any particular challenges in the field that you think will be overcome in the near future?

C.B.: Well, there are many. One of the major challenges I think that has potential in our ongoing work is the fact that no matter how we redirect the specificity of T cells against cancer antigens, no matter if we choose a CAR or TCR or maybe both, at the end of the day, we’ll have a lymphocyte that will have to be able to reach the tumor site and still work. However, at the tumor site, these engineer lymphocytes will probably find the same immunosuppressive signals that natural lymphocytes find. We all know how cancer is good at orchestrating the tumor microenvironment to change the behavior of all the surrounding cells to make its growth and life easier and also to exhaust, anergize, and kill potential harmful cells such as T lymphocytes.

Indeed, in tumors rich in T lymphocytes, namely “hot tumors,” these lymphocytes are very often exhausted, and in these specific cases, we can give checkpoint inhibitors, drugs that counteract exhaustion, and revitalize T cells. These drugs free the lymphocytes from the exhausted phenotype, making them fit again so that they will kill the tumor. This incredible discovery, which led to the Nobel Prize in 2018, allows us to cure tumors rich in tumor-specific lymphocytes. If the tumor is cold, there are too few lymphocytes, which is the majority of cases, checkpoint inhibitors cannot work. By gene engineering, we can build these T cells. However, we might expect that our engineered CAR or TCR will probably find in the tumor the same signals leading to exhaustion.

So, I think that we need to identify, possibly for each tumor type, what are the major pathways that anergize T cells or lead to their exhaustion, and we should possibly try to make our engineer cells resistant to those negative signals. We can do this with a combination of genome editing tools, such as CRISPR, base or prime editing tools, and viral or nonviral vectors. All of these innovative biotechnological tools allow us to add and delete more than one gene in the same cell, so we can make cells specific for our selected tumor antigen but also resistant to the inhibitory pathways active in the tumor that we are trying to treat. This approach is already being tested in some initial trials and will be more and more useful in the near future.