Aleta’s Novel T Cell Engagers

Aleta Bio recently announced that their scientists have created, characterized and validated novel classes of Immune Cell Engager therapeutics including T Cell Engagers (TCE) and CAR T Cell Engagers (Ref. 1). In this post we describe the technology in more detail.

We’ve published extensively on the CAR-T Engager class, exemplified by our Phase 1/2 clinical program that is evaluating the safety and efficacy of ALETA-001, a therapeutic that binds CD19-directed CAR-T cells to a second antigen, CD20 (Ref. 2). That program is designed to improve the rate of durable response in patients who have received CAR-T cells - we will update that exciting program as the clinical data continue to come in.

Today, however, we turn to the TCE therapeutic class. These agents simultaneously bind to a tumor-associated protein on cancer cells and the CD3 protein on T-cells, facilitating a targeted anti-cancer immune response. These therapeutics are most commonly directed towards blood cancers like Non-Hodgkin Lymphoma (NHL) and Multiple Myeloma (MM) and include approved therapeutics like blinatumomab, glofitamab, teclistamab and talquetamab that target CD3 on T cells and the proteins CD19, CD20, BCMA, and GPRC5d, respectively, on cancer cells. There is one approved TCE for a solid tumor indication, the DLL3-targeting therapeutic tarlatamab, approved for patients with previously treated small-cell lung cancer (SCLC).

These therapeutics share some characteristic measures of clinical efficacy, including high initial overall response rates but a low rate of complete responses and limited duration of response. Consequently, these drugs have limited ability to keep patients disease-free and limited impact on patient survival. Our goal therefore is to safely improve these types of therapeutics for longer-lasting clinical efficacy.

When we look at how TCEs interact with targeted cancer cells we can quickly see how cancers respond to T Cell-based therapies. The most common response is quite expected, which is that the cancer cells escape from TCE therapy by downregulating expression of the targeted cancer protein (Ref. 2). In MM, loss of BCMA expression appears to be the most common cause of poor responses to BCMA-targeted TCEs, and diminished BCMA expression also causes relapses after an initially favorable response therapy (Ref. 3). Other mechanisms that cancer cells use to escape from TCE treatment are less expected. CD58 is a protein that interacts with the CD2 protein that is expressed on T cells to upregulate T cell activation. Downregulation of CD58 by cancer cells is observed after TCE therapy even though the TCE mechanism of action, which is to bind to the CD3 protein complex, does not require CD2 signaling (Ref. 4). Similarly, downregulation of the MHC protein complex on cancer cells has been demonstrated in response to TCE therapy. This is surprising in that this implies either an interaction between the TCE-bound T Cells and MHC expressed on the targeted cell or implies that the TCE triggers an expanded immune response that would utilize normal MHC display of antigen. Both explanations are plausible, but the relevant clinical data are limited.

These kinds of results – loss of tumor-expressed proteins that might otherwise interact with the T cell population that is stimulated by the TCE – tell us that those signals are important for clinical efficacy. This is not gone unnoticed of course. CD58 downregulation led companies like Cargo Therapeutics to overexpress CD2 proteins in CAR T cells and Novartis to put an anti-CD2 agonist antibody domain into a TCE. Similarly, T cell bispecific companies like Regeneron and Roche have incorporated anti-CD28 antibody domains into TCE format – this is an attempt to recapitulate the B7/CD28 protein interaction that is costimulatory for activated T cells.

In related work, investigators like Michael Dustin, Misty Jenkins, Manuel Izquierdo and many others have been drilling down on the structure of T cell interactions with immune cells – and with cancer cells (Refs. 5,6,7). What has been uncovered gives us a blueprint for designing better TCE therapeutics and better CAR T Engagers. The blueprint reveals the structure of a productive T cell immune synapse cell surface and the proteins involved (Figure 1).

Figure 1. An immune synapse forms between a T cell and a cell to which the T cell binds such as an antigen presenting cell (APC) or any cell that has been infected by a pathogen or is otherwise abnormal, like a cancer cells. When it forms, the synapse segregates different types of proteins into zones, including the initial contact zone, a zone of costimulatory proteins interactions, and a zone of adhesion protein interactions.

Because the mechanism by which TCE therapeutics activate a T cell – by binding to the CD3 protein rather than binding through the normal T Cell Receptor complex (TCR) – a naturally structured immune synapse fails to form. The T cell is still activated via the TCE, and potent cytotoxic activity is triggered, but the differentiation of the activated T cells into costimulation-dependent “central memory” T cell subsets is not effectively supported by the TCE design. Efforts to trigger differentiation have focused on binding to and activating the costimulatory CD28 protein using antibody domains. While the anti-CD28 strategy can trigger costimulatory signaling, the impact on anti-cancer efficacy is unpredictable and has not been routinely effective. These observations are further complicated by the wide diversity of bispecific engager design, ranging from the relatively simple BiTE format to the more complicated “full-antibody” formats. We expect further complexity in observations as the field moves from B cell cancers to solid tumors. This is because B cells cancers, being derived from normal B cells, can express the CD58 and B7 proteins with which the T cell would normally interact. Thus TCE interaction may induce an immune synapse-like array on the surface of NHL and MM cells. On the other hand, solid tumor cells can express CD58 but never express B7 proteins. Therefore, we would expect the interaction of a TCE with a solid tumor cell to produce a suboptimal T cell response to the tumor. Finally, as noted earlier, loss of CD58 by any targeted cancer cell would disrupt binding to CD2, and impair proper immune synapse formation.

We have recently discovered that we can leverage specific components of immune synapses to improve our engager technologies. These improvements can be applied to both CAR T Engagers and T Cell Engagers. 

A critical protein of the immune synapse that forms around the natural TCR is CD2. CD2 plays multiple roles in the T cell immune synapse and appears to act as a protein “organizer” within the synapse. Cell surface proteins that engage in synapse formation are mobile – they move across the cell surface. In models of natural T cell activation it has been shown that CD2 is highly labile, first localizing within the T cell contact zone and subsequently segregating into the costimulation zone (Figure 1). If sufficient CD2 is present, interaction with CD58 results in a unique protein rearrangement called the corolla pattern which changes the localization of costimulatory proteins - including CD28. This re-localization supports sustained CD28 signaling (Ref. 8). Thus, the immune synapse can switch from initial signaling via the TCR to costimulatory signaling via CD2 and CD28. The segregation of initial signaling and costimulatory signaling - and the requirement that CD2 and CD28 have cell surface mobility - gives us some hints as to how to best co-opt the immune synapse structure to optimize TCE design (Figure 2).

Figure 2.  

A) A graphical view of a COSTIM-TCE bound a tumor cell surface. A tumor antigen is targeted with a VHH antibody; this positions the anti-CD3 antibody to bind to a T cell. Around the contact zone, CD58 and B7 protein domains create a costimulation zone that will engage CD2 and CD28 on the T cell surface. As the costimulation corolla forms, LFA-1 on the T cell is displaced, forming an adhesion zone that will bind to ICAM-1 expressed by the tumor cell.

Figure 2 illustrates a COSTIM-TCE was designed to create a T cell immune synapse around the point of tumor cell contact – the bullseye – that is created when the anti-CD3 VHH and the anti-IL13Ra2 VHH bind to their targets.

These COSTIM-TCE are robustly better at tumor cell killing than the matched TCE (Figure 3). The increase in T cell activation required two or more CD58 binding domains and also required that they be at a distance from each other within the TCE. We have further shown that the increase in T cell activation is correlated with CD2 binding and was correlated in turn with the level of IFN-γ secretion and the extent of anti-tumor cytotoxicity. In the most potent format, the COSTIM-TCE mediated melanoma and glioma tumor cell killing with an IC50 of 1 pM, equal to approximately 80 pgs/ml – more than a log more potent than the corresponding TCE.

Figure 3.  A TCE (green line) and a COSTIM-TCE (red line) were used in a T cell activation assay using normal human T cells and tumor cells that overexpress that overexpress the solid tumor antigen IL13Rα2.  In this experiment, the COSTIM TCE was designed to include 2 copies of a portion of the CD58 protein that can bind to and activate CD2 (Ref. 9). The blue line is a negative control protein.

Similar improvements have been produced for multi-antigen targeting TCE for diverse cancer indications, for example in a TCE that targets both IL13Rα2 and B7H3. For other indications – MM and AML for example – we’ve developed both advanced CTEs and TCEs, giving us multiple paths forward for the benefit of these patients.

In parallel we have developed COSTIM-TCE that contain small B7-2 protein binding domains that activate CD28 signaling (Ref. 10). In further refinements, we have now combined CD58 and B7-2 binding domains within a TCE to induce canonical T-cell activation. As noted above, we have used these same technologies to create novel CAR T Engagers designed for use with CAR-T cells in diverse indications.

Because the costimulation domains are very compact and stable they can be used alongside other functional designs within a TCE or within a CAR T Engager. As just one example, we have added a monomeric TRAP protein to inhibit TGFbeta-mediated immune suppression within the tumor microenvironment. Finally, these small engager proteins are readily developed as biologics but can also be expressed from cells including T cells. This provides ample flexibility to optimize the treatment modality to attack specific cancers. 

We are interested in forming collaborations and partnerships to advance our COSTIM programs. If you would like to know more about this proprietary technology, please contact me: paul.rennert@aletabio.com. 

————————————————————————————

References.

1. https://www.aletabio.com/news-publications/boosting-car-t-cell-efficacy-to-increase-cancer-cure-rates

2. Su et al. 2022. Oncoimmunology 11: 2111904. doi: 10.1080/2162402X.2022.2111904

3. Singh et al. 2021. Br J Cancer 124: 1037–1048. doi: 10.1038/s41416-020-01225-5

4. van de Donk et al. 2024. Lancet Haem 11: e693-e707. doi: 10.1016/S2352-3026(24)00186-8

5. Shen et al. 2022. J Immunother Cancer 10:e004348. doi: 10.1136/jitc-2021-004348

6. Dustin. 2023. Fac Rev 12: 25. doi: 10.12703/r/12-25

7. Davenport et al. 2018. Proc Natl Acad Sci U S A 115: E2068-E2076. doi: 10.1073/pnas.1716266115

8. Izquierdo et al. 2024. Front Immunol 15: 1497118. doi: 10.3389/fimmu.2024.1497118

9. Demetriou et al. 2020. doi: 10.1038/s41590-020-0770-x

10. Ikemizu et al. 1999. Proc Natl Acad Sci U S A 96: 4289-4294. doi: 10.1073/pnas.96.8.4289

11. Rennert et al. 1997. Int Immunol. 9: 805-813. doi: 10.1093/intimm/9.6.805.