Attacking Cancers Using Novel Methods of Targeted Protein Degradation

By Paul D. Rennert, SugarCone Biotech LLC.

Cell surface degraders are weird but cool.

There is an emerging class of therapeutics, brilliant really, that make use of cell surface proteins that will respond to antibody binding by moving quickly inside the cell and then shuttling over to lysosomes for degradation. The trick is bind to these “internalizer proteins” with a bispecific antibody or fusion protein so they drag a second bound targeted protein with them. This works because everything that is bound up with the internalizing protein complex gets degraded. There is a well-known version of this trick played by the bispecific antibody MCLA-158 that binds to LRG5 and EGFR. MCLA-158 induces rapid degradation of EGFR that is dependent on LRG5 binding – LGR5 internalizes and is degraded and the bispecific binding ensure that EGFR is dragged along with it. Not only is degradation rapid but MCLA-158 has a prolonged effect on EGFR expression, through 72 hours in cell culture models (see https://www.nature.com/articles/s43018-022-00359-0). Thus binding to LRG5 - a rather innocuous protein that mediates cell:cell adhesion in colonic ”crypts” - has a pronounced effect on EGFR-mediated colon cancer growth and metastasis. It of course is helpful that both LRG5 and EGFR are overexpressed in colon and other cancers, allowing the bispecific antibody to bind avidly.  

It is important to be clear that this is a unique degrader strategy and not at all like ProTAC or ProMAB technologies that targets proteins by binding them to the E3-ligase controlled degradation pathway. That is a complex field that has struggled to bring therapies to approval (see https://www.oncologypipeline.com/apexonco/esmo-2024-degraders-disappoint-again).

What’s weird here is there is no good reason that I can think of for LRG5 to functionally co-localize in the cell membrane with EGFR. They do seem to respond to each other: EGFR stimulation can upregulate LRG5 expression, but why this should happen is unclear. Regardless, the bispecific antibody is sufficient to grab both proteins on the cell surface. But what’s also weird is the apparent specificity. Co-internalization by LRG5 does not appear to occur with the EGFR-related receptor Her3 for example. Thus, it is not enough for 2 proteins to each be expressed on the same cell surface, other features such as protein mobility in the cell membrane must play a role. I suspect this is an area of active research.

Happily, the screen used to identify the MCLA-158 antibody was agonistic to such biases: bispecific antibodies were screened for those that targeted cell surface signaling receptors, like EGFR, and also targeted proteins like LRG5 that are activated by the Wnt-pathway, the canonical driver of colorectal cancers. From the EGFR degradation screen the effect of the LGR5/EGFR dual-binding combination was revealed. 

Merus (NASDAQ: MRUS) the company developing MCLA-158, now called ‘petosemtamab’, is targeting not just colorectal cancer but also head and neck squamous cell carcinoma (HNSCC) and of course there are even bigger patient populations in indications like non-small cell lung cancer. Of note, data presented at ASCO2024 in HNSSC showed that petosemtamab plus the anti-PD-1 antibody Keytruda produced a promising response rate in a (relatively small) sample of patients (https://www.fiercebiotech.com/biotech/asco-early-peak-merus-bispecific-cancer-data-lives-investors-expectations).

Another well-known example comes from the UCSF group headed by Jim Wells. The class of cell-surface protein degraders called KineTACs use natural cytokines engineered onto a bispecific antibody scaffold. The cytokine binds naturally to one or more internalizing cytokine receptors while the antibody arm targets and binds to a cell surface protein. Using the cytokine CXCL12 to bind the internalizing receptor CXCR7 at high affinity, various proteins could be internalized and degraded including PD-L1, Her2 and EGFR (https://www.nature.com/articles/s41587-022-01456-2). The extent of degradation was variable, from 51% (Her2 in MCF7 cells) to 86% (EGFR in Hela cells). The degree of degradation may be related to the amount of targeted protein versus the amount of internalizing receptor CXCR7 expressed on different cell types. This happens because both bound proteins are degraded. Essentially, you run out of internalizing protein CXCR7 faster than you run out of targeted protein, eg. Her2. This study also showed that the antibody binding site on the target proteins Her2 and EGFR impacted the degree of degradation observed, with those binding sites that were further away from the cell surface inducing less degradation. Some of the UCSF team (and patents) were spun into EpiBiologics, a venture-funded biotech (https://www.epibiologics.com/).

This is old news, what’s new?

Yesterday I stumbled across two papers and a pharma program, enough to send me down this rabbit hole. The transferrin receptor is a well-understood “internalizer”. The transferrin receptor facilitates the uptake of iron into cells from transferrin, a protein found in the bloodstream. This system is essential for the development of red blood cells as one key example of its function, indeed, iron is required by all cells throughout the body. The transferrin receptor is rapidly and repeatedly internalized: once the iron is released from transferrin inside the endosome, the receptor is then recycled back to the cell membrane to repeat the process, allowing for efficient iron uptake. 

There are two transferrin receptors: Trf1 is expressed ubiquitously while Trf2 is more selectively expressed. It has been known the 1980s that Trfs control movement of transferrin across endothelial boundaries in blood vessels and at the blood brain barrier (BBB). Further, anti-Trf antibodies can trigger internalization and even “receptor-mediated transcytosis” to pass through the BBB and other endothelial boundaries. It follows that some bispecific antibodies have been developed to the use an anti-Trf arm to bring a second binder into the CNS. Examples include an anti-TfR/anti-BACE1 bispecific and an anti-Aβ monoclonal antibody fused to an anti-TfR Fab. These are in development for neurodegenerative diseases.

But the papers that caught my attention are using Trf1 bispecifics in a novel and very clever way to develop anti-cancer therapies. Trf1 is highly overexpressed on aggressive cancers whose cells are rapidly proliferating and whose metabolic needs, including for iron, are high. Trf1 recycles to and from the cell surface at a fast rate, in line with its role in iron delivery. Trf1 is a constitutively endocytosed protein. Indeed, if one just binds transferrin to Trf1 it will internalize, release iron and pop back up on the cell surface.

Trf1-targeting bispecific antibodies named TransTACs were developed by researchers at the Dana-Farber Cancer Institute to take advantage of the rapid internalization kinetics. TransTACs were engineered to drive rapid co-internalization of a target protein and TfR1, but then forcing entry of the bound-up complex into the lysosomes for degradation (https://doi.org/10.1038/s41586-024-07947-3). The authors used a few tricks to push the system. An antibody format that used bivalent binding of Tfr1, rather than monovalent binding, enhanced degradation. And to avoid degrading all the Trf1 a protease-cleavable linked was placed between the bispecific antibody components, including the bound target protein, from the bound Trf1. Clipping by a cathepsin protease released Trf1 to return to the cell surface, while the rest of the complex was shuttled into the lysosome for degradation. Very clever. The technology was used to degrade a variety of cell surfaces proteins including CD20, EGFR and PD-L1. Degradation of target proteins was rapid and very deep, up to 80-90%, and proteins reappeared only slowly after the TransTACs were washed out. This is still not a catalytic system, since the protease cleavage that occurs is not reversible, but so long as Trf1 remains available at the cell surface this just becomes a matter of TransTAC bioavailability, which should be easy to dose as needed.

Another Trf1-based degrader technology is visible as a preprint on BioRxiv: this work addresses a method for keeping the degrader cycle catalytic (https://doi.org/10.1101/2024.02.21.581471). The platform is called CYpHER and the paper is co-authored by Cyclera Therapeutics personnel – the company web site is just a link to the paper. The pitch is interesting and takes advantage of a pH change that occurs as proteins move into endosomes, then lysosomes. When proteins are endocytosed the endosome “matures” from an early to a late endosome, during which the interior pH drops to pH 5.5 (thus, acidic). At this pH, transferrin releases its iron but remains bound to Tfr1, and the transferrin/Trf1 complex returns to the cell surface. The process is rapid and repetitive, cycling in and out every 10-20 minutes. This cycle can be engineered by changing the pH sensitivity of the binding components. In this case the CYpHER molecule was engineered to contain a pH-sensitive target protein antibody domain that releases the target from the complex at low pH, allowing the protein target to be taken up by lysosomes, while the antibody-bound Tfr1 returns to the cell surface, kicking off a new cycle.

The targets used to demonstrate this technology were the usual suspects, PD-L1 and EGFR. To alter the pH sensitivity, histidine substitutions were made in the antibody binding domains and, in a further example, an EGF variant. In each case, pH-sensitive release was demonstrated. Thus the target protein, PD-L1 or EGFR, would be degraded and the rest of the molecule returned to the cell surface intact and able to initiate a new cycle. The degree of reduction of EGFR obtained ranged from 55% - 81% across lung cancer cell lines and the reductions were maintained for 3 days. Cancer cell viability was reduced by 50% - 80% after 7 days treatment in vitro, however only limited in vivo data was shown in this paper. The degree to which the catalytic nature of the technology translates into better anti-tumor efficacy remains uncertain.

Intriguingly, Kyowa Kirin Co., Ltd. has developed an anti-EGFR x anti-Trf1 degrader with a very different mechanism of action by picking binding domains that don’t interfere with the binding of natural ligands for the targets, that is, EGF and transferrin/iron complex (https://www.abstractsonline.com/pp8/#!/20272/presentation/1125). Now the entire complex enters the degradation pathway via lysosomes but the goal is not to prevent EGFR expression or activity. Instead, binding to tumor cells is ensured by the EGFR binding domain that only binds EGFR-“bright” cells, meaning they have 10 of thousands of copies of EGFR on the cell. This is 10-fold higher than normal EGFR expression. Because Trf1 is degraded as part of the degradation of the internalized protein complex, these EGFR-bright cells become iron-starved and die as a consequence. They have taken this drug into normal primates and have shown the drug is well tolerated. Wow - a very clever strategy.

There are many examples of cell surface degraders – Nature recently published a review (https://doi.org/10.1038/s41587-022-01594-7) – and there are a number of stealth programs maturing nicely in the background. It’s a hot area. 

So, look, I had fun learning about the Trf1-based degrader technologies as these were new to me. I suspect a bunch of smart labs, companies and a few VCs are looking carefully at cell surface degrader technologies, including those using Trf1. I’m going to keep an eye on this space.

Stay tuned.