The complexities of tomorrow’s cell therapies: The future is gene editing
In all areas of biotech and pharma, there is a never-ending quest to improve the safety, potency, and effectiveness of drug therapies by harnessing the latest scientific and technological advances. Nowhere is this more evident than in the rapidly growing area of cell and gene therapy (CGT) products. These products are becoming increasingly complex as our knowledge of disease biology expands and CGT platforms evolve, offering insights for developing better CGT products and the technologies needed to make those improvements.
In vivo gene therapies provide a simple example of this evolving complexity. This approach typically begins with an adeno-associated viral (AAV) vector: a gene construct (a designed segment of DNA) is engineered into the vector, and the vector is injected directly into the patient. Each subtype (serotype) of AAV targets a different cell and / or tissue type; specific AAV vectors are therefore selected according to the target cell or tissue in the body in which the gene therapy is intended to go.
However, most AAV serotypes target tissue types beyond the gene therapy’s desired therapeutic target. For these therapies, the next level of complexity involves addressing one or both of two possible solutions: narrowing the target range of the AAV vector both by reducing its targeting for non-desirable tissues and enhancing its targeting for the desired tissue; and engineering the vector’s genetic payload for expression only in the desired cell type, so that it is not expressed in non-target tissues after delivery.
Any of these capabilities can be engineered into the viral vector. This is the present state of complexity for most in vivo gene therapies.
In contrast, the complexities are more layered for ex vivo engineered T cell or other immune cell-based therapies, which are in development for cancer and autoimmune disease. These layers include selection of a disease-specific antigen receptor; delivery of the receptor construct into the cell; other engineering to resist the immunosuppressive tumor microenvironment, to enable long term persistence and potency once transferred back to the patient; and large-scale commercial manufacturing of the cell therapy in a GMP-compliant manner.
Moreover, ongoing research in cancer has revealed the variety of mechanisms tumors use to evade the immune system. These advances point the way of improving cell therapies; but they also highlight the need for more sophisticated tools to make those improvements. The increasing interest in developing allogeneic cell therapies or in vivo engineering of antigen receptors and other components – a concept known as “in vivo cell therapy” – adds yet another layer of complexity.
A company aiming to develop and commercialize a cell therapy must address all layers of complexity to be successful. It’s certainly not easy. It’s also expensive, because most drug developers won’t have unfettered access to the full array of technologies needed to develop the product on its own. This lack of access is especially true for gene editing, a technology essential to addressing the multi-layered complexities of tomorrow’s cell therapies.
Continued advancements in the science of T cell therapies
The earliest iterations of cell therapies for cancer were autologous; that is, they began with T cells derived from the patient. The cells were either tumor-infiltrating lymphocytes (TILs), already presumed to contain some T cells specific for tumor antigens; or they were peripheral blood lymphocytes that could be stimulated with tumor antigens ex vivo to become antigen-specific. In either case, the cells were cultured and expanded to make them modestly active, without any genomic engineering or editing, then administered back to the original patient.
However, for various reasons this approach didn’t work as well as expected. It turns out that the antigen specificities of TILs found in certain tumor types, such as breast and colorectal, are poorly defined; and some TILs populations may include T cells that have immunosuppressive, rather than antitumor, activity.
The evolution of the science, from these early iterations of T cell therapies to the current state of the art, has introduced three fundamental requirements:
1
The first is an antigen receptor, such as a chimeric antigen receptor (CAR) or T cell receptor (TCR), enabling the T cell to become highly specific in targeting the tumor.
For a CAR, the identification of receptor candidates is relatively straightforward: the CAR is designed and engineered to target a known tumor antigen, such as CD19 expressed on B cell cancers.
For a TCR, identification of the best receptor candidate is not always so straightforward, because the TCR is selected from the individual patient’s repertoire of T cells and the immune system’s response to cancer (and for that matter, to infection) is usually polyclonal. This means the immune system randomly generates a large set of receptor variants against tumor antigens, but not all of these variants will be tumor-specific. Instead, activation by tumor antigens selects for variants with tumor specificity – a real-time process of natural selection, in the Darwinian sense – and thus determines the repertoire of anti-tumor receptors T cells will have.
How does this polyclonal response complicate the identification of an antigen-specific TCR? The immune system may respond to, say, 10 different antigens expressed by a tumor; then, the immune system may generate 10 sets of 100,000 T cells, one set for each antigen. But among those one million T cells, some may be therapeutically beneficial while others have no effect on the cancer. The challenge lies in isolating what might be a very small number of beneficial T cells, then identifying the receptor they express that confers their antitumor activity.
2
The second requirement is engineering the T cells ex vivo to introduce the selected antigen receptor.
The workhorses for delivering antigen receptor constructs such as CARs and TCRs into a cells are lentiviral vectors, not AAV vectors. This is because lentiviruses integrate into the genome of the cell, which enables long-term expression of the receptor construct and transmission of that construct to daughter cells. But lentivirus is not without its drawbacks. It is complex and expensive to manufacture: dose-for-dose, lentivirus and AAV cost about the same, but lentivirus has lower yields per batch. However, recent improvements in manufacturing technologies for lentiviral vectors are overcoming these limitations.
3
The third requirement is the need for process development, analytical development and QC, and manufacturing capabilities to produce the engineered T cells under GMP conditions, in batches large enough for human therapeutic applications.
To do this at a commercial scale, a company first has to solve the manufacturing challenges of large-scale biology, which entails defining the critical quality attributes of the product through proper analytical development and QC during manufacturing scale-up. This process is significantly cost- and labor-intensive for a small biotech company to develop for one or two products, especially if they have never done so before.
This is the present level of complexity in cell therapies, but the field isn’t standing still and the landscape continues to evolve.
Tomorrow’s cell therapies depend on gene editing
The evolution of cell therapies brings up new ideas, and with them come new dimensions in functionality to maximize the cells’ effectiveness. To incorporate these functions, cell therapy innovators will need to access new technologies.
For example: it’s long been known that lentiviral vectors insert themselves at random places in the genome. But recent research has shown that insertion of an antigen receptor construct at one specific locus, instead of randomly across the genome, can result in T cells with greater potency. So, improving T cell therapies means employing an alternative to lentivirus for delivering the receptor construct – and that alternative is gene editing.
Next, a great deal is now known about how tumors, especially solid tumors, evade the immune system.
Blood cancers are more responsive to T cell therapies because the cancer cells circulate in the blood where T cells can easily find them. By contrast, solid tumors have many types of defenses – collectively known as the immunosuppressive tumor microenvironment – to keep out T cells. To get a T cell therapy past these outer defenses and into the tumor tissue, the T cell has to be engineered with some form cell trafficking – an address code, so to speak.
Even after the T cell gets past the defenses and infiltrates the tumor, the tumor can deploy countermeasures specifically designed to weaken or destroy the T cells, such as secreting the immune checkpoint protein, PD-L1. To avoid these countermeasures, the T cell needs defenses of its own, such as the deletion of the gene encoding PD-1, the receptor for PD-L1.
Lentiviral vectors cannot be engineered to deliver all of the gene constructs needed to make the foregoing changes to a T cell, precisely where they’re needed in the genome. Instead, the changes have to be made by gene editing of the T cell itself.
Similarly, allogeneic therapies – cells derived from one donor that are engineered to treat many patients – also require technology capable of making multiple edits to the genome.
The CGT space has made significant progress in developing allogeneic therapies derived from hematopoietic stem cells (HSCs), mesenchymal stem cells (MSCs) and induced pluripotent stem cells (iPSCs), with the EMA approving the first allogeneic cell therapy in the EU in late 2022 and FDA approving the first in the U.S. in 2023. To prevent rejection and graft-versus-host disease (GvHD) in the recipient, an allogeneic T cell must be HLA-matched to the recipient. This is a tall order, given that the human genome encodes six HLA genes, each with thirty or more variants (alleles). Alternatively, GvHD risk can be reduced by using gene editing to delete the cell therapy’s TCRαβ receptor that enables T cells to distinguish “self” from “non-self” cells, or to block expression of one or more of the HLA genes.
Another wave of the future is the concept of using iPSCs to produce “universal cell therapies”, in which functional expression of the six HLA genes is prevented via knockout of B2M, a shared subunit necessary for surface expression of these molecules, effectively eliminating the risk of rejection or GvHD. The hurdle here is that clinical-grade iPSC lines – those that comply with both Good Tissue Practices (GTP) and GMP – aren’t easy to obtain or produce.
However, introducing this array of genetic changes into a single T cell with traditional CRISPR-based gene editing systems is problematic, because there is a risk of jumbling the genome and a high cost to accessing the technology.
Gene editing nucleases like Cas9 and its orthologs used in CRISPR make a double-stranded DNA break in a target gene, temporarily resulting in two chromosome segments that are readily reassembled. If the nuclease happens to make more than one double-stranded break, temporarily resulting in four chromosome segments, the odds are still good that the segments will reassemble correctly – but those odds are not 100%. Incorrect reassembly produces chromosomal abnormalities, such as translocations and inversions, that adversely affect cellular processes and can even cause diseases. This risk is associated with every edit to the cell, and so places limits to the numbers of edits per cell that Cas9 and related nucleases can safely make.
Base editing technology gets around the potential reassembly problems posed by Cas9 and orthologous nucleases. A base editing enzyme does not create a double-stranded DNA break; instead, it cuts or “nicks” one strand of DNA, then converts one nucleotide base in the target genetic sequence to another base type. Because it carries no risk of jumbling the genome, base editing allows multiple gene edits per cell – exactly what’s necessary to make T cell therapies as potent and effective as possible.
Finally, the concept of in vivo cell therapy, which involves generating CAR or TCR T cell therapies by gene editing T cells in the patient’s body, adds the latest layer of complexity. In addition to meeting all of the requirements described above, these multi-gene editing constructs must be packaged and delivered specifically to T cells in vivo – for example, by encapsulating the cargo in nanoparticles conjugated to a CD3-targeting antibody. The first in vivo cell therapy products are approaching the clinic; as a class, they may be competitive with allogeneic T cell therapies and the space could evolve quickly. In either case, the next wave of cell therapy’s future is already upon us.
Not an impasse: A solution
To recap, then: A company aiming to create the complex immune cell therapies of tomorrow will need at least five, and possibly six, key components: (1) the antigen receptor; (2) a means of delivering that receptor, preferably with gene editing instead of a lentiviral vector; (3) GMP manufacturing for commercialization; (4) GTP- and GMP-compliant iPSC lines, preferably gene edited to be “universal cells” as starting points for allogeneic cell therapies; (5) base editing technology to engineer multiple features into the cells; and (6) for in vivo cell therapies, a means of delivering the gene editing constructs to T cells in the patient.
If a company doesn’t have access to all five (or six) components – three of which involve gene editing – it won’t be able to make and commercialize an ex vivo (or in vivo) cell therapy product. This situation might appear to leave a company with an innovative cell therapy at an impasse, but this is not quite so.
ElevateBio has technology platforms for all six components: protein, AAV vector, lentiviral vector and cell engineering platforms to identify antigen receptors and deliver them into cells; a large-scale biology approach to GMP manufacturing of CGTs at our BaseCamp facility; a bank of clinical-grade, GTP- and GMP-compliant iPSC lines in development to power regenerative medicine and allogenic cell therapies; the gene editing technology for making multiple changes to cell therapies, including those derived from iPSCs; and a range of delivery vehicles, including lipid nanoparticle (LNPs), for gene editing constructs.
Life Edit Therapeutics adds a highly innovative gene editing platform to our expertise in the discovery and development of new CGTs. Life Edit’s large and diverse array of RNA-guided nucleases (RGNs) and base editors can access virtually any region of the genome and enable the entire spectrum of edits, from small insertions or deletions to rewriting DNA sequences.
By offering this complete suite of CGT technologies to its strategic partners and portfolio companies, ElevateBio aims to develop CGTs at scale in a cost-effective and efficient manner. This is at the core of “ElevateBio’s DNA”; that is to say, ElevateBio has built this into its business model for a large number of partners, thereby significantly reducing the risk and cost of CGT development and commercialization. We believe that if the industry has more gene editing tools we can “democratize” access to these essential technologies required to make the cell therapies of tomorrow – ultimately improving medicines for patients around the globe.
