Product Development

What’s next for targeted protein degradation

As the technology goes mainstream, next-gen molecules will use new enzymes, tap new pathways, and address PK/PD challenges

As the technology goes mainstream, next-gen molecules will use new enzymes, tap new pathways, and address PK/PD challenges.

As targeted protein degradation grows up, the field is going after harder targets and new mechanisms of action.

Over the last year, a series of Phase I/II readouts from front-runner Arvinas Inc. (NASDAQ:ARVN), including the May 13 unveiling of the technology’s first efficacy signal, have taken targeted protein degradation from preclinical white space to clinically viable modality. More efficacy data are expected by year-end.

The company has shown its two lead candidates are well-tolerated and orally bioavailable, and that its androgen receptor degrader ARV-110 induced confirmed PSA declines of >50% in two out of eight castration-resistant prostate cancer (CRPC) patients previously treated with multiple lines of therapy; one of the patients showed a confirmed RECIST partial response. As of May 29, follow-up data had not found responses in four additional patients.

“It was a big de-risking event for the entire field,” Arvinas CSO Ian Taylor told BioCentury. Arvinas raised $238 million through its Sept. 2018 IPO and Nov. 2019 follow-on.

Almost all the major pharmas and big biotechs have entered the field via deals or through internal research programs, and at least ten newcos have been created in the last three years. At least eight other biotechs that were founded without an explicit focus on targeted protein degradation have disclosed they too are building programs in the space. 

Targeted degraders work by bringing together a target protein and an enzyme that can trigger its degradation.

Unlike small molecule inhibitors and blocking mAbs, targeted degraders stop all of a protein’s functions, instead of only those carried out at a specific site. And unlike antisense, RNAi and CRISPR, which gradually reduce the amount of a target protein by preventing new copies from being made, targeted degraders rapidly deplete the existing protein pool as well as counter new protein synthesis.

The technology frees drug developers from having to hit a specific functional site on a target, which opens up targets historically considered “undruggable” because of their hard-to-reach or lack of active sites.

But targeted degraders can still only go after what they can bind. “The challenge is, and will continue to be, finding ligands for the proteins you want to eliminate,” said Arvinas founder and Yale University professor Craig Crews.

Advances in ligand discovery technologies are paving the way for companies to chase a wider range of therapeutic targets. The same platforms are also giving rise new degradation mechanisms.

The field’s first candidates work via a handful of well-characterized E3 ubiquitin ligases, which mark targets for proteasomal degradation by tagging them with a ubiquitin group. In the next wave, several companies are tapping lesser-known E3 ligases, or engaging entirely different protein degradation pathways such as endolysosomal trafficking or autophagy.

A challenge remains how to predict PK/PD relationships for targeted degraders across different cellular systems in the body, where drug exposure and protein expression differences could have an impact on their biological activity.

“That’s where companies will succeed or fail,” Kymera CEO Nello Mainolfi told BioCentury.

Ways to degrade

The primary split among targeted protein degraders has been whether they are composed of a single ligand that binds both the target and E3 ligase, or separate ligands for each, connected by a linker.

The former are known as molecular glues. The latter are categorized as heterobifunctional degraders, and often referred to as proteolysis targeting chimeras (PROTACs), a term coined by the Crews lab and used by Arvinas to describe its compounds.

Molecular glues are generally smaller and more “drug-like,” but are more challenging to discover; heterobifunctional degraders are typically larger, but their modularity makes it easier to build a compound that engages both the target and ligase. Crews thinks the distinction between categories is blurring as the linkers in heterobifunctional degraders get smaller.

The space is also being diversified by an influx of new protein degradation strategies.

“People are saying, I can use the same concept to get different degradation endpoints,”  Stew Fisher, CSO of C4 Therapeutics Inc. (NASDAQ:CCCC), told BioCentury.

One strategy gaining traction is engaging E3 ligases outside of the three that are most commonly used for targeted degradation: cereblon, vHL and IAP.  Companies including Arvinas, Kymera Therapeutics Inc. (NASDAQ:KYMR), Nurix Therapeutics Inc. (NASDAQ:NRIX), Amgen Inc. (NASDAQ:AMGN) and Cullgen Inc. have disclosed plans to explore alternatives among the approximately 600 E3 ligases in the human genome.

The idea is to identify E3 ligases that are upregulated in tissues or disease states of interest, or that are particularly efficient at degrading certain classes of proteins. Increasing the specificity of where targeted degraders act can increase their therapeutic window. “Imagine how many programs could benefit from a lack of pharmacology in the bone marrow,” said Kymera’s Mainolfi.

Companies are not disclosing the alternative ligases they are pursuing, but Kymera is collaborating with academic researchers to build an E3 Ligase Whole-Body Atlas, with the aim of mapping the expression patterns of all known human E3 ligases in both disease and healthy contexts.

UC Berkeley professor and  Frontier Medicines Corp. founder Daniel Nomura said the rise of targeted protein degraders coincided with advances in chemoproteomics ligand discovery tools and DNA-encoded libraries (DELs), which have been key to discovering ligands for new ubiquitin ligases and challenging targets.

“I think all of these have come together in the last few years in a way that you can apply them toward real problems, not just tool compounds,” he said.

Nomura is also director of the Novartis-Berkeley Center for Proteomics, a collaboration with researchers at Novartis Institutes for BioMedical Research (NIBR) that is exploring how covalent chemoproteomics could support the development of targeted protein degraders.

Nomura said his lab’s chemoproteomics work has identified ligandable sites across more than 90% of the E3 ligases in the genome, and has published studies on compounds that recruit the ligases RNF114, RNF4 and KEAP1.

Vividion Therapeutics Inc. is using chemoproteomics to develop its targeted degraders, which connect non-covalent ligands that bind therapeutic targets to covalent ligands that bind undisclosed new E3 ligases.

Nurix CEO Arthur Sands said his company has “invested heavily” in its DELigase discovery platform, which uses DELs to identify compounds that can harness or inhibit new E3 ligases. Amgen and Plexium Inc. have also made DELs a focal point of their discovery platforms to find compounds that engage E3 ligases.

Other companies such as Bristol-Myers Squibb Co. (NYSE:BMY), C4 and Monte Rosa Therapeutics Inc. have told BioCentury they are open to working with new ligases, but are primarily focusing on cereblon-based degraders because of the extensive chemical libraries and knowledge built up around that ligase. 

“I’m not saying we shouldn’t look at other ligases, but for 90% of the targets we’re going after, we can find hits with cereblon, and optimize them quickly,” said Fisher. “The question is, do you see a need for another ligase?”

A handful of companies are tapping targeted protein degradation mechanisms that don’t involve E3 ligases or the proteasome at all.

Lycia Therapeutics Inc. is developing lysosomal targeting chimeras (LYTACs), which target cell surface and secreted proteins for degradation via the endosome and lysosome.

Frontier Medicines and Casma Therapeutics Inc.  are each developing autophagy degrader platforms, which recruit the autophagosome to degrade targets too large for the proteasome, such as dysfunctional organelles.

LYTACs and autophagy targeting chimeras (AUTACs) are also within the scope of Amgen’s targeted protein degradation work via its Induced Proximity Platform unit, launched in 2019.

Kinetic energy

Understanding how a targeted degrader’s PK and enzymatic kinetics intersect with the way its target is expressed and functions in relevant tissues will be key for success in the clinic, said C4’s Fisher and Kymera’s Mainolfi.

Both companies have developed quantitative systems pharmacology models to help optimize their targeted degraders by predicting how they’ll behave in vivo.

Fisher said that in addition to parameters like a target’s resynthesis rate target, such models need to account for the target’s biological function in the pathway of interest to determine the extent and duration of degradation needed.

“How deep do you need to go, for how long, to have a catastrophic effect on a tumor cell?” he said. “Some targets only need a small perturbation to have a large effect.” 

An Oct. 13 presentation at the third annual Targeted Protein Degradation Summit by Kymera VP of Preclinical Development Haojing Rong showed a case study of the PK/PD relationship for the company’s STAT3 degrader KTX-201. 

Mainolfi  said it’s important to collect data on and to model a compound’s effects across different disease-relevant cell types, which for example might have different starting levels of target protein expression. 

“Simplifying things by monitoring the distribution and degradation profile in one particular cell type, or in the tumor on one side of mouse, that’s the risk the field should not be taking,” he said.

The next targets

Although the first disclosed protein degradation programs have had little overlap in targets and indications, they have generally gone after well-validated targets for which small molecule inhibitors were already available or in development, which provide the most straightforward path to validation.

For example, Arvinas’ lead programs ARV-110 and ARV-471 aim to overcome tumor resistance to androgen receptor and HER2 inhibitors, respectively. Kymera’s KT-474  is one of several IRAK4 inhibitors in development for autoimmune or inflammatory diseases, and C4 and Nurix are each leading with hematological cancer programs that hit IKAROS and AIOLOS; Nurix’s NX-2127 also hits Btk,  the target of multiple marketed inhibitors.

But groups are starting to look to targets that have proven challenging to drug via traditional modalities.

On Oct. 14, Arvinas announced it was pursuing early-stage programs targeting the transcription factors BCL6 and MYC, the oncogenic signaling molecule KRAS — which only a handful of small molecules have recently shown signs of addressing — and HPK1, an immuno-oncology target whose scaffolding role is considered to be an important part of its immunosuppressive function. The company is also developing degraders that selectively target mutant, but not wild-type, HTT for Huntington’s disease.

BMS’s lead cereblon E3 ligase modulators (CELMoD) compounds are CC-92480 and CC-99282, which seek to overcome resistance to “imids” through more potent degradation of IKAROS and AIOLOS. The molecules are in Phase I testing for multiple myeloma (MM) and non-Hodgkin lymphoma (NHL), respectively.

The company also has CC-90009, which promotes degradation of a more recently discovered cereblon substrate GSTP1, in Phase I testing for acute myelogenous leukemia (AML).

BMS is also developing therapies that target non-natural cereblon substrates, starting with the androgen receptor, but with an eye toward “classically undruggable” targets like RAS, Kristen Hege, SVP of early clinical development, oncology/hematology and cell therapy, told BioCentury.

Targeted protein degraders are also being deployed outside of oncology. “We will see an increase in indications,” said Crews. “My personal favorite will be going after aggregates.”

Besides its Huntington’s disease program, Arvinas’ neurology pipeline includes a program degrading pathogenic forms of tau in frontotemporal lobar degeneration (FTLD), progressive supranuclear palsy (PSP) and Alzheimer’s disease (AD), and another degrading pathogenic forms of α-synuclein in multiple system atrophy (MSA) and Parkinson’s disease (PD).

Kymera, whose internal pipeline includes one program in inflammation and two in oncology, is also pursuing six other programs with partner Vertex Pharmaceuticals Inc. (NASDAQ:VRTX) that span different diseases, said Mainolfi.

Nurix has early discovery-stage programs in virology. “Virology will be an interesting area for targeted protein degradation, because certain ligases are part of the strategy for viral replication,” said CEO Arthur Sands.

In addition to populating pipelines, targeted protein degraders are also increasingly being used as tools for target discovery. Their rapid timescale means they are less likely to induce compensatory biological mechanisms than RNAi or CRISPR, with the added benefit of being reversible.

“It’s the ultimate conditional gene knockout,” said C4’s Stew Fisher.

C4 has partnered with Tocris Holdings Ltd. to develop a protein degradation-based tool system, and tools developed by the Crews lab are sold by Promega Corp.

TARGETS
AIOLOS (IKZF3) – IKAROS family zinc finger 3 
BCR-ABL – BCR-ABL tyrosine kinase 
Btk – Bruton's tyrosine kinase
Cereblon (CRBN)
CRBN- Cereblon
GSTP1 – Glutathione S-transferase pi 1 
HER2 (EGFR2; ErbB2; neu) – Epidermal growth factor receptor 2 
HPK1 (MAP4K1) – Mitogen-activated protein kinase kinase kinase kinase 1 
HTT – Huntingtin 
IAP – Inhibitor of apoptosis 
IKAROS (IKZF1; LYF1) – IKAROS family zinc finger 1 
IRAK4 – Interleukin-1 receptor-associated kinase 4 
KEAP1 – Kelch-like ECH-associated protein 1 
KRAS (K-Ras) – KRAS proto-oncogene, GTPase 
MYC (c-Myc) – v-myc myelocytomatosis viral oncogene homolog 
STAT3 – Signal transducer and activator of transcription 3
vHL – von Hippel-Lindau tumor suppressor 

As targeted protein degradation grows up, the field is going after harder targets and new mechanisms of action.

Over the last year, a series of Phase I/II readouts from front-runner Arvinas Inc. (NASDAQ:ARVN), including the May 13 unveiling of the technology’s first efficacy signal, have taken targeted protein degradation from preclinical white space to clinically viable modality. More efficacy data are expected by year-end. ...