Manufacturing CAR T cell therapy—one of the most powerful weapons against certain blood cancers—has to date involved an elaborate process, through which doctors first extract a patient’s immune cells and ship them to a specialized facility where the cells are genetically reprogrammed to fight cancer. The engineered cells must then be shipped back, so they can be infused into the patient’s bloodstream. The process can take weeks and may cost hundreds of thousands of dollars, placing it out of reach for many of the patients who need it most.

Scientists at UC San Francisco and collaborators have now reported on the development of a method to precisely reprogram these cancer-fighting cells directly inside the body, potentially eliminating the manufacturing process, cost, and waiting time.

It is the first time that scientists have integrated a large sequence of DNA at a specific site in human T cells without removing the cells from the body. “To our knowledge, this is the first demonstration of programmable, site-specific integration of a large DNA payload into T cells in vivo,” Justin Eyquem, PhD, an associate professor of medicine at UCSF, told GEN. “While others have pursued in vivo CAR T cell generation using either lipid nanoparticles for transient mRNA delivery or engineered lentiviral vectors for random integration, we achieved a combination of cell specificity, locus specificity, and stable integration of a large transgene in a living organism.”

Crucially, in reported tests the new, targeted approach was found to outperform the standard method of randomly integrating DNA using viruses. And in experiments using mice with humanized immune systems, the researchers used the new technology to successfully treat aggressive leukemia, multiple myeloma, and even a solid tumor. “… the most exciting, and honestly the most surprising, finding was how well the approach worked in solid tumors,” Pierre-Louis Bernard, PhD, a postdoc in the Eyquem lab, acknowledged to GEN.

“There has been a broadly held assumption in the field, shared by our reviewers, that engineering a small number of T cells circulating in the blood and expecting them to traffic to a solid tumor, encounter antigen only upon arrival, and then expand sufficiently to control the tumor would be an extremely tall order. We proved it was possible. That result, more than any other, convinced us that the platform has therapeutic reach well beyond hematological malignancies.”

Bernard, is co-first author, and Eyquem senior and corresponding author of the team’s published paper in Nature titled, “In vivo site-specific engineering to reprogram T cells.”

CAR T cell therapy works by giving T cells—the immune system’s disease fighters—new genetic instructions to recognize and destroy cancer cells. These instructions come in the form of chimeric antigen receptors (CARs), molecules that protrude from the surface of T cells. The CAR binds to a target on a cancer cell surface, triggering the T cell to attack and kill the tumor cell. Seven CAR T cell therapies are currently approved by FDA for use in blood cancers. “Engineered T cells, reprogrammed to express chimeric antigen receptors (CAR) or T cell receptors (TCR), have transformed cancer treatment and are being explored as therapeutics for autoimmune and infectious diseases,” the authors commented.

However, accessing these therapies, which cost $400,000 to $500,000, has proven difficult for many patients. The manufacturing process requires specialized facilities and takes weeks. Before receiving the engineered cells, patients must also undergo intensive chemotherapy that some patients can’t tolerate.

“It’s become a global access issue; many patients who would benefit from CAR T cells either can’t afford them or can’t get them fast enough,” Eyquem said. Re-engineering immune cells in the body—in vivo manufacturing—could completely bypass the costly and complex ex vivo manufacturing, while also eliminating the need for preparatory chemotherapy. Eyquem told GEN, “In vivo CAR T cell generation aims to engineer T cells directly within the body … Done without lymphodepleting preconditioning, this approach has the potential to dramatically reduce cost, shorten timelines, and broaden global access to a treatment that remains out of reach for the vast majority of patients who might benefit from it.”

Dual-vector system

For their newly reported research, Eyquem and collaborators, including scientists at the Gladstone Institutes, Duke University, and Innovative Genomics Institute, designed a dual-particle system to carry CRISPR-Cas9 ribonuclear protein (RNP) gene-editing machinery—the molecular scissors required to alter genes—directly to T cells circulating in the body. “Using CRISPR–Cas9 and adeno-associated virus (AAV)-mediated homology-directed repair (HDR), we targeted CAR integration into the endogenous human TCR alpha locus (TRAC),” the authors summarized in their paper.

William Nyberg, PhD, co-first author and former postdoc in the Eyquem lab, further explained, “We developed a two-vector system to deliver CRISPR–Cas9 ribonucleoproteins and a DNA donor template, using enveloped delivery vehicles and adeno-associated viruses, respectively … We optimized both vectors for T cell-specific delivery and gene-targeting efficiency.”

The goal, Eyquem explained, was to remain both cell-specific and locus-specific, using CRISPR to integrate the CAR transgene at a defined site in the T cell genome. “That required delivering two distinct components to the same T cell in vivo,” he told GEN, “… the CRISPR machinery to cut at a precise location, and a DNA template encoding the CAR to be inserted there … Because these two payloads have fundamentally different delivery requirements, we developed a two-vector system, with each vector specialized for one component.”

Eyquem expanded on the process for developing the two vectors. “The first vector—for delivery of the CRISPR–Cas9 machinery—was developed by Jennifer Hamilton in the Doudna laboratory. We call it an Enveloped Delivery Vehicle, or EDV. It packages a fully formed Cas9–gRNA complex inside a particle that outwardly resembles a lentivirus, but instead of delivering DNA it delivers the editing protein directly. By engineering its surface with a mutated fusogen and an anti-CD3 antibody fragment, we direct it selectively to T cells in vivo.

“The second vector carries the CAR DNA template. We used an AAV, a vector naturally suited to efficient nuclear DNA delivery … The challenge was that the standard AAV serotype used in this context, AAV6, is poorly suited for in vivo use … In collaboration with Aravind Asokan at Duke, we evolved a new capsid variant (AAV-hT7) that overcomes barriers … Bernard further demonstrated using a genome wide screen that our engineered AAV interacts with CD7.”

While a two-vector system might sound more complex, it carries a critical advantage, Eyquem explained. “… while neither vector alone is entirely T cell-specific, their combination is. Only cells that take up both the EDV and AAV-hT7 undergo CAR integration and that overlap is highly specific to T cells. Engineering precision is preserved through the intersection of the two vectors working together.”

Nyberg, who now has his own lab at the Karolinska Institute in Sweden, was the first to combine both vectors in a humanized mouse model to deliver a CD19 CAR, Eyquem noted. “Remarkably, after a single injection of the two vectors, we observed a large population of CAR T cells and complete B cell aplasia … The day he analyzed the cells, I told him: ‘Will, let’s have a drink, this is going to be BIG.’ It has been working in every system and model we have tested since.”

In their reported study the researchers tested the technology in mice engrafted with aggressive leukemia. A single injection of the dual-particle system cleared all detectable cancer in nearly all the mice within two weeks. The engineered CAR T cells made up as much as 40% of immune cells in some organs and successfully eliminated cancer from both the bone marrow and spleen.

Potential for solid tumor therapy

The approach also worked against multiple myeloma and, strikingly, against a solid sarcoma tumor.  “Our demonstration that in vivo generated CAR T cells can control an aggressive solid tumor model in mice, something many in the field assumed would be beyond reach has given us real confidence that this is worth pursuing seriously,” Eyquem said.

“One of the most important findings was a direct head-to-head comparison with engineered lentiviral vectors, the leading competing approach for in vivo CAR T cell generation,” he continued. “At an equivalent dose, our TRAC-integrated CAR T cells were markedly superior. We believe this comes down to the fundamental difference between site-specific and random integration. With site-specific integration at the TRAC locus, every successfully engineered T cell expresses a functional, physiologically regulated level of CAR. With lentiviral delivery, expression is lower on average and highly heterogeneous, suggesting that only a subset of transduced cells reaches the threshold needed for optimal antitumor function. Precision of integration translates directly into potency.”

Fundamentally different 

The researchers’ system is fundamentally different from other integrative in vivo platforms, he stated. “Other integrative in vivo platforms currently being tested clinically deliver fully independent expression cassettes, meaning the CAR gene carries its own promoter and can be expressed in any cell the vector happens to transduce. This creates real risks: if a tumor cell is accidentally transduced, for example, it will express the CAR and can downregulate the target antigen on its surface, driving antigen-negative relapse and treatment failure.

“We deliver what is called a promoterless cassette, a CAR transgene that carries no promoter of its own and therefore cannot be expressed unless it integrates precisely and in-frame into the right gene in the right cell type. That gene is TRAC, which encodes one of the chains of the T cell receptor and is exclusively expressed in T cells. This provides a level of selectivity that no randomly integrating vector can match.”

Beyond CAR T cell therapy

The choice of TRAC as the integration site offers several advantages beyond simple T cell specificity, he said. “We believe this represents a breakthrough that extends well beyond CAR T cell therapy. The same fundamental strategy, pairing a cell-targeted RNP delivery vehicle with an evolved AAV carrying a site-specific DNA template, could in principle be applied to other cell types and other indications. The TRAC locus and T cells are our proof of concept, but the platform logic is broadly applicable wherever precise, stable transgene integration in a defined cell population is therapeutically desirable.”

Eyquem and collaborators have established Azalea Therapeutics to advance the dual-vector platform. Azalea has exclusively licensed foundational academic IP to support its programs for in vivo site-specific engineering.

The reported mouse data have given the researchers the confidence that the platform is “… ready to be pushed toward patients, and that is now the primary driving force behind everything we do,” Eyquem said. Azalea has already moved with “remarkable speed” since its founding, he stressed. “The team onboarded the technology, reproduced all the key findings from the Nature paper, and has since further developed and scaled the platform for clinical application.”

A critical milestone in that journey was demonstrating efficacy in non-human primates, a pivotal translational step that the mouse data alone could not provide,” he explained to GEN. “In a proof-of-concept non-human primate study evaluating feasibility, pharmacodynamics and tolerability, a single intravenous dose of the EDV and AAV vectors achieved complete B cell aplasia in peripheral blood by day 10, expansion of TRAC-CAR T cells to approximately 35% of all peripheral T cells confirming robust in vivo engineering, complete B cell aplasia in lymph nodes and bone marrow by day 13, and a favorable tolerability profile with no unexpected findings to date.”

These data were presented at the ASGCT Breakthroughs in Targeted In vivo Gene Editing session back in November 2025, he said.

Towards the clinic

Azalea is now “racing toward the clinic, with CD19-directed CAR T cell therapy as the lead program,” GEN learned. “… The goal is clear: to get this technology to patients as fast as responsibly possible … The most likely first indication is a CD19-directed CAR T cell therapy, pursued across two distinct disease areas: B cell malignancies and autoimmune disorders.”

The current plan is to reach first-in-man by the end of 2027, Eyquem suggested. “Given where the platform stands today with compelling mouse data, non-human primate studies already demonstrating robust in vivo CAR T cell generation and deep B cell aplasia with a favorable tolerability profile, and Azalea actively working through the necessary manufacturing and regulatory steps, we believe this is an ambitious but realistic timeline.”

The post Dual-Vector System Generates CAR T Cells <i>In Vivo</i> appeared first on GEN – Genetic Engineering and Biotechnology News.