Ten years ago, chimeric antigen receptors (CAR) promised to change the landscape of cancer therapy forever (1). With the help of genetic engineering, a person’s own immune cells could get retrofitted with a surface receptor that recognized cancer antigens. This would muster an immune response against cancer cells, training the body to defend itself from malignancy.

Even with a handful of FDA-approved CAR therapies — largely using a patient’s own T cells to go after blood cancers — drastically changing the treatment landscape for some of the most challenging conditions, the field is now reckoning with the limitations of the approach. “One of the caveats of CAR T cell therapy is that this is a living drug,” said Daniel Powell, Jr., a cancer biologist at the University of Pennsylvania. “Once it's administered to patients, we have no means to control the activity.”

Acknowledging that the original architecture may not have been flexible or controllable enough for an effective cancer treatment, researchers are now going back to the drawing board to rethink CAR designs and capabilities. A new paradigm is the “universal CAR,” a deconstructed CAR that researchers can customize to go after a nearly unlimited range of antigens. 

In 2023, Jason Lohmueller, a synthetic biologist at the University of Pittsburgh developed a new way to engineer universal CAR: SNAP-CAR (2). These modular CAR provide a way to make immunotherapy less toxic while keeping pace with rapidly evolving cancers. Alongside Coeptis Therapeutics, a biotechnology company that has licensed the technology, Lohmueller hopes that bringing this approach to the clinic will make CAR therapies a more effective option for more patients.

A universal solution?

In the early 2010s, the first conventional CAR-based cell therapies raced toward the clinic. The CAR T cells that achieved early successes in clinical trials for treating blood cancers were part of the second generation, wherein engineered cells expressed a cancer-antigen-specific receptor attached to a T cell surface protein, alongside a costimulatory molecule such as CD28 or 4-1BB to make the cells even more potent (1). However, this approach wasn’t flawless. 

The first FDA-approved CAR T cell therapy, Kymriah, targeted a B cell surface protein called CD19 (3). B cells proliferate wildly in patients with cancers such as acute lymphoblastic leukemia, and these CAR T cells used CD19 to find and kill these cells to control the cancer. But if even a few cancerous B cells don’t express CD19, those cells can evade the CAR T cell therapy and drive cancer relapse. This process, called antigen loss or antigen escape, limits the long-term effectiveness of CAR therapies; in CD19 CAR clinical trials, as many as 50 percent of participants ultimately relapsed (4).

Hopefully it could be a very personalized therapeutic way down the line. You can screen a patient's tumor, figure out what antigens they have, and figure out what [antibody] adaptors to use, all using the same CAR T cell. 
- Jason Lohmueller, University of Pittsburgh

CAR therapies can also have nasty side effects. CD19, for example, is on the surface of both cancerous and healthy B cells. When anti-CD19 CAR T cells kill healthy B cells, patients require transfusions of immune molecules to maintain their defenses against infections. For other organs, it isn’t as easy to compensate for killing healthy cells.

Powell suspected that these limitations of CAR T cells could be addressed with a relatively simple change: uncoupling the internal molecular pathways mediating the immune response from the external receptors that recognize a cancer antigen. By separating the functional elements of a CAR T cell, physicians could turn it on or off or redirect it toward a new antigen by modifying just its external components.

In 2012, Powell’s team first proposed the idea of universal CAR in a paper presenting a modular system based on a pair of molecules, biotin and avidin, that bind to each other (5). The researchers engineered the T cells to express avidin on their surfaces, attached to internal proteins that could trigger an immune attack. They also engineered biotin-labeled antibodies that targeted an antigen of interest. By administering the T cells alongside the antibodies, the antibodies’ biotin tags would bind to the T cell’s avidin molecules to essentially reconstruct a functional CAR T cell with more flexibility to determine its target. 

Powell hoped that this would help avoid antigen escape. Rather than just targeting CD19, the researchers could administer antibodies that bind multiple leukemia antigens so that the CAR T cells have multiple ways to identify and kill cancer cells.

“You're responding to the evolution of the cancer by creating additional agents that allow you to target new antigens,” he said. “That's probably the single biggest benefit.”

Better building blocks

Lohmueller heard about Powell’s universal CAR when he was a graduate student at Harvard University, and he was intrigued by its potential. “This is much better than the alternative of making two different CAR T cell products that recognize different antigens because it's double the work,” he said. “[With this] incredible technology, you can have plug-and-play targeting of CAR T cells.”

When he tried to use existing universal CAR approaches to develop CAR T cells targeting mucin 1 (MUC1), an antigen found on many types of cancer, it wasn’t as effective as he had hoped. He turned his attention to designing more potent universal CAR. This required increasing the affinity between the T cell and the antibody. At first, he just tried tweaking avidin and biotin to make them bind more tightly, but he realized that the CAR activity would always be limited by intermittent binding of avidin and biotin (6). 

To maximize activity, he needed to permanently attach the antibody to the T cell with a covalent bond — something that Powell’s team had also begun to incorporate into their universal CAR (7). “It mimics the natural traditional CAR [fusion proteins] in being attached,” he said. “By using that covalent bond, we can achieve more traditional CAR-like activity with a universal system.”

To accomplish this, his team used the SNAP-tag system that had been previously developed to make fusion proteins held together by a covalent bond (8). SNAP is a self-labeling enzyme. When it sees its target molecule, benzylguanine, it attaches itself to the target. SNAP could take the place of avidin in the recipe for universal CAR, and benzylguanine could replace biotin. When benzylguanine-tagged antibodies encountered SNAP on a T cell’s surface, they covalently attached to the T cell to form a functional CAR that could recognize a target antigen and trigger the T cell to release cell-killing molecules.

In a 2023 study, Lohmueller’s team found that SNAP-CAR T cells could successfully fight a tumor in a mouse model (2). Importantly, however, SNAP-CAR also offered a path to tuning the potency of the therapy by increasing or decreasing the levels of the antigen-targeting antibody. “If we titrate the amount of [antibody] adaptor, we can get different levels of tumor cell killing,” Lohmueller said. “It's really nice to have that extra control.”

Lohmueller also noted that the SNAP-CAR system accommodates multiple antibodies to target a wider range of antigens. The limit of how many antibodies can be used for one patient is still unknown. However, he proposed that these antibodies could be administered all at once to target multiple antigens simultaneously or in sequence, depending on how the cancer evolves during treatment. After all, the antibody building blocks of these CAR T cells are dynamic. After being infused into the patient, they float around for weeks on average, and even the antibodies that bind to the T cells only remain active there for a few days. This gives physicians the opportunity to continually update the mixture of antibodies available to form CAR T cells in the patient.

“It could be a very personalized therapeutic way down the line,” Lohmueller said. “You can screen a patient's tumor, figure out what antigens they have, and figure out what [antibody] adaptors to use, all using the same CAR T cell.”

New frontiers for SNAP

Even before the study was published, Lohmueller’s SNAP-CAR technology attracted Coeptis Therapeutics’ attention. In 2022, the company, which focuses on developing cell therapies, entered into an agreement with the University of Pittsburgh to license the technology for the development of SNAP-CAR therapeutics. “There's a clear attraction to therapies that could be broadly applicable,” said Colleen Delaney, chief scientific and medical officer of Coeptis Therapeutics and an oncologist at the University of Washington.

In particular, Coeptis Therapeutics is interested in making off-the-shelf allogeneic SNAP-CAR therapies that can be used for any patient. Typically, cells are extracted from a patient, engineered to express the CAR, and infused back into the patient to avoid immune rejection of cells from a different donor. But this autologous approach is time-consuming and expensive. An allogeneic therapy could be manufactured in bulk in advance and distributed to hospitals or pharmacies for many different patients to use. “The therapy is really waiting for a patient as opposed to a patient waiting for the therapy,” Delaney said. 

SNAP-CAR add another dimension to the flexibility of an allogeneic therapy: Now, one therapy can be used for many patients with many diseases, as well, since the CAR wouldn’t have a pre-defined target. 

There are still many hurdles to making allogeneic T cell therapies because T cells have receptors that can trigger graft-versus-host disease (GVHD) when infused into a different person (9). While Lohmueller’s team continues to engineer T cells, Coeptis Therapeutics focuses on a different type of immune cell: natural killer (NK) cells. These cells are easier to transplant from one person into another without triggering GVHD. Phase 1/2 clinical trials have already shown that CAR NK cells can target cancer with minimal side effects, and SNAP-CAR could make these therapies even more versatile, Delaney said (10).

“In the future, you could create master lots of NK cells that are expressing universal CAR and utilize them to treat tens or hundreds of patients,” Powell said.

Coeptis Therapeutics is currently optimizing SNAP CAR for NK cells. They start with stem cells that the researchers differentiate into NK cells, a process that is modeled after an allogeneic cell therapy platform that Delaney previously developed at the University of Washington.  

The team is testing lentivirus and retrovirus vectors to deliver the DNA encoding the SNAP protein and its internal components into the cells. But viral vectors can be an expensive and scarce resources. To reduce the amount and the cost of virus required, Coeptis Therapeutics is transfecting the stem cells early in their differentiation phase when the number of cells is still low. They’ve found that the SNAP-CAR proteins persist in the NK cells as they differentiate, allowing them to create batches of allogeneic SNAP-CAR NK cells that can be repeatedly infused into patients alongside antibodies to point the cells at their targets.

Once they optimize the system and test it in animal models of leukemia and solid tumors, the Coeptis Therapeutics team hopes to apply for Investigational New Drug status within the next 24 months. Delaney thinks Coeptis Therapeutics’ approach will stand out in a crowded field of CAR therapies. “A lot of people are doing this, so what I hope to get across is that we're going to do this in a more efficacious way,” Delaney said.

From bench to bedside

For SNAP-CAR cell therapies to make it to the clinic, researchers will have to show that they can work at least as well as conventional CAR. Lohmueller said that the data already show that, and now he hopes to demonstrate that SNAP-CAR can do even more. For example, conventional CAR can have off-target effects that lead to toxicity, but Lohmueller recently showed that it’s possible to design the covalent link between the antibody and SNAP protein so that it can be severed by an “off switch,” such as UV light or a small molecule (11). 

The therapy is really waiting for a patient as opposed to a patient waiting for the therapy. 
- Colleen Delaney, Coeptis Therapeutics

Similarly, the researchers can build in “on-switches,” for example, so that the antibody only binds to the SNAP protein in the low-oxygen conditions found in a tumor. “We can do all kinds of cool chemistry to make that chemical tag available for binding,” Lohmueller said. Even though allogeneic T cell products are more challenging to develop, he still believes that would be the ideal final product, not only for cancer treatment. “You can have it ready to go to tailor to whatever disease you have,” Lohmueller said. 

Coeptis Therapeutics licensed the SNAP CAR technology for autoimmune indications as well. For example, in lupus nephritis, where B cells attack the patient’s own cells, researchers tested whether CAR that can target CD19 could offer a new treatment avenue. According to Delaney, SNAP CAR could boost these therapies and might even help treat infectious diseases in immunocompromised people. 

This is an area where Delaney has seen the need firsthand as a practicing physician, a role she still holds. “It helps me maintain an understanding of what the unmet need is for these patients,” she said. “My hope is to make these therapies more clinically accessible.”

References

1. Mitra, A. et al. From bench to bedside: the history and progress of CAR T cell therapy. Front Immunol  14, 1188049 (2023).
2. Ruffo, E. et al. Post-translational covalent assembly of CAR and synNotch receptors for programmable antigen targeting. Nat Commun  14, 2463 (2023).
3. Maude, S.L. et al. Tisagenlecleucel in Children and Young Adults with B-Cell Lymphoblastic Leukemia. N Engl J Med  378, 439-448 (2018).
4. Majzner, R.G. et al. Tumor Antigen Escape from CAR T-cell Therapy. Cancer Discov  8, 1219-1226 (2018).
5. Urbanska, K. et al. A universal strategy for adoptive immunotherapy of cancer through use of a novel T-cell antigen receptor. Cancer Res  72, 1844-52 (2012).
6. Lohmueller, J.J. et al. mSA2 affinity-enhanced biotin-binding CAR T cells for universal tumor targeting. Oncoimmunology  7, e1368604 (2018).
7. Minutolo, N.G. et al. Quantitative Control of Gene-Engineered T-Cell Activity through the Covalent Attachment of Targeting Ligands to a Universal Immune Receptor. J Am Chem Soc  142, 6554-6568 (2020).
8. Keppler, A. et al. A general method for the covalent labeling of fusion proteins with small molecules in vivo. Nat Biotechnol  21, 86-9 (2003).
9. Basar, R. et al. Next-generation cell therapies: the emerging role of CAR-NK cells. Blood Adv  4, 5868-5876 (2020).
10. Marin, D. et al. Safety, efficacy and determinants of response of allogeneic CD19-specific CAR-NK cells in CD19+ B cell tumors: a phase 1/2 trial. Nat Med  30, 772-784 (2024).
11. Kvorjak, M. et al. Conditional Control of Universal CAR T Cells by Cleavable OFF-Switch Adaptors. ACS Synth Biol  12, 2996-3007 (2023).