In recent years, intracellular targeted protein degradation (iTPD) has become an important new paradigm in small molecule drug development. The first class is known as proteolysis-targeting chimeras (PROTACs), which construct bifunctional small molecules to recruit the protein of interest (POI) and the E3 ligase for proteasomal degradation. The second class is known as molecular glues, which enhance the interaction between E3 ligases and POIs, leading to their ubiquitination and proteasomal degradation.
Compared with traditional inhibitors, iTPD has significant advantages. First, PROTACs or molecular glues can theoretically bind to any site on the POI, whereas inhibitors generally need to bind to active or allosteric sites. Second, the degradation of the target removes the entire protein, including its scaffold function, so the therapeutic activity of the drug should last longer and be closer to the results of gene knockout experiments. Third, these molecules work catalytically, unlike classic pharmacological inhibitors, which work stoichiometrically, allowing for lower doses to achieve the same therapeutic effect.
The progress of iTPD has inspired the biological and chemical communities to target extracellular proteins, known as extracellular TPD (eTPD). There are at least three significant differences between iTPD and eTPD, aside from the location of the target POI. First, cells recycle proteins through two main pathways: the proteasomal and lysosomal pathways. iTPD almost exclusively uses the proteasomal pathway, which is typically used to process intracellular proteins. eTPD involves bispecific biomolecules or small molecules that recruit membrane-bound or secreted POIs to membrane-bound cyclic receptors and deliver the POIs to the lysosome, which is the typical pathway for extracellular protein degradation.
Second, these two protein hydrolysis destruction modes have different kinetics and use different proteases. iTPD is faster—usually within minutes to hours—because all components are within the cell. In contrast, eTPD is generally slower, typically taking 6–48 hours, because it involves vesicular transport through early and late endosomes from the membrane, eventually fusing with the lysosome to result in protein degradation.
Third, almost all iTPD systems use only a few E3 ligases, mainly cereblon (CRBN) or von Hippel–Lindau (VHL), creating challenges in finding E3 ligase binders. These ligases are widely expressed in tissues, limiting tissue-selective targeting. In contrast, eTPD can use a variety of degradation systems, allowing for more specific tissue selectivity. Finally, compared with typical small molecules, antibodies have a very long pharmacokinetic profile, which results in a lower frequency of administration. Therefore, eTPD represents an emerging new paradigm in the development of biopharmaceuticals.
For existing biopharmaceutical models, there seems to be an ample selection in the field of biologics, such as antibodies, antibody-drug conjugates (ADCs), bispecific T cell engagers, or chimeric antigen receptor (CAR)-T cells. eTPD is unlikely to replace efficient methods for killing cancer cells. However, the existing models all have limitations, such as the collateral damage caused by the toxic payloads in ADCs, or cytokine storms induced by CAR-T cells, along with challenges in production and administration. Additionally, these therapies are limited to membrane-bound POIs and are primarily used for acute, life-threatening diseases in oncology.
The closest comparison to eTPD is naked antibodies that function by functionally blocking POIs. Both antibodies and eTPD can act on soluble or membrane-bound targets. However, eTPD may have a significant advantage. First, compared with antibodies that block function by stoichiometric binding, eTPD can catalytically degrade the target. Therefore, in principle, eTPD could reduce the required dosage, which could be a significant advantage when POIs are abundant. Second, eTPD and other degradation mechanisms can not only completely block the function of a single active site of the POI but also fully block the entire protein's function, ensuring that any potential functions are eliminated. In fact, eTPD eliminates the targeted functional site, opening up targeting space for amyloid proteins and the like.
Another possibly less obvious advantage is that the random binding of long half-life antibodies may produce low levels of toxic POIs, especially for soluble POIs. For example, the therapeutic antibody bevacizumab has been shown to significantly increase the circulating half-life of soluble target vascular endothelial growth factor (VEGF). Similarly, antibodies that bind to IL-6 and several other cytokines or growth factors have demonstrated the same effect. Even small buffering effects can produce contradictory consequences. In one early example, it was found that the naturally shed extracellular domain of the human growth hormone receptor (hGHbp) significantly increased the half-life of hGH and enhanced its activity. Therefore, compared to simple binding that leads to functional blockage, eTPD degradation of pathogenic POIs may have significant pharmacological advantages.
The Fc region of antibodies plays a crucial role in antibody recycling and immune cell activation. The neonatal Fc receptor (FcRn) is responsible for recycling internalized antibodies, sending them out of the cell before they reach lysosomal degradation. Enhancing the binding of antibodies to FcRn can extend their serum half-life in humans.
By clever design, engineered antibodies can deliver POIs to acidic endosomes through pH-switchable FcRn, releasing the POIs for lysosomal degradation. Tocilizumab (Tcz), a humanized antibody approved for treating rheumatoid arthritis, binds to the IL-6 receptor (IL-6R) in a pH-dependent manner. By redesigning Tcz into a clearing antibody, its affinity to FcRn was enhanced using known mutations, and by mutating the CDR, the affinity to IL-6 was reduced by about 20-fold at pH 6.0, while maintaining affinity at pH 7.4. The modified antibody can transport IL-6R to the lysosome while maintaining its connection to FcRn and cycle back to the plasma membrane to collect more IL-6. This extended the serum half-life in monkeys from one week to one month.
Studies have shown that the pH-switchable antibody requires 20 times less dose to clear IL-6R compared to the parent antibody, and it allows for subcutaneous injection instead of intravenous infusion. The pH-switchable Tocilizumab was approved in 2020 for treating neuromyelitis optica spectrum disorder, representing the first approved antibody in eTPD.
Using glycan-targeted recycling receptors, such as cation-independent mannose-6-phosphate receptor (CI-M6PR) or asialoglycoprotein receptor (ASGPR), can promote lysosomal degradation of membrane and soluble POIs. One method involves the bioconjugation of multiple glycan ligands of CI-M6PR with antibodies targeting POIs, known as lysosome-targeting chimeras (LYTACs). The LYTAC–POI complex is internalized upon binding to CI-M6PR, leading to POI degradation in the lysosome. Another method, called MoDE or ASGPR-targeting chimeras (ATACs), uses bispecific small molecules developed for ASGPR-based approaches.
CI-M6PR is a 300 kDa dimeric type I receptor that is a pH-switchable receptor, binding to ligands at neutral pH and releasing them in acidic endosomes before fusion with lysosomes. CI-M6PR has been used to shuttle exogenous lysosomal enzymes carrying M6P to lysosomes for the treatment of lysosomal diseases. One study conjugated 6–8 copies of mannose-6-phosphate (M6Pn)-containing, non-hydrolyzable glycans to the lysine residues of POI-targeting antibodies to create potential eTPD. The study showed that glycan-labeled mCherry or ApoE4 antibodies were internalized 10 to 100 times faster than non-specific controls.
ASGPR is highly expressed on hepatocytes and can rapidly clear asialo-glycoproteins. A study demonstrated that approximately 10 copies of tri-GalNac bioconjugated with POI-binding antibodies could be used to degrade a variety of therapeutically interesting proteins. Up to 70% of EGFR in HEPG3 cells was degraded after 24–48 hours, comparable to the results from CI-M6PR-based LYTACs.
In conclusion, glycan-based eTPD degradation approaches primarily utilize ASGPR or CI-M6PR. They have very different tissue distributions; ASGPR is highly expressed in the liver, while CI-M6PR is more widely expressed but mainly in immune cells. In fact, the affinity of glycans for degradation arms is inherently lower than antibodies, requiring multiple sugar units and high levels of conjugation. This may present significant challenges in CMC. Nevertheless, these two glycan-based degradation systems provide exciting ways to remove membrane and soluble proteins.
In addition to the 600–700 intracellular E3 ligases, there is also a family of over 30 transmembrane E3 ligases. Bispecific antibodies targeting members of this family and a POI can degrade target proteins, a method known as AbTACs or PROTABs. One of the most characterized members is RNF43, a 783-residue type I protein containing a 200-residue extracellular domain, a transmembrane domain, and an intracellular RING domain with E3 ligase activity. Another homolog, ZNRF3, has a different tissue distribution from RNF43.
One study constructed an AbTAC targeting PD-L1, using the Fab domain of atezolizumab for the POI arm and the Fab of RNF43's extracellular domain for the degradation arm. This Atz-AbTAC employed the classic KIH format, inducing degradation of PD-L1 with a DC50 of 3.4 nM and a maximum degradation (Dmax) of approximately 63% at 24 hours. Whole-cell proteomics revealed no significant overall protein changes or cytotoxicity.
Additionally, researchers explored VHHs that bind to transmembrane E3 ligase family members including RNF43, ZNRF3, RNF128, RNF130, and RNF167. They called these REULRs (Receptor Elimination Using Ligase Recruitment). Some REULRs with cetuximab-Fab arms inhibited up to 75% proliferation of A431 cells. A new startup, InDuPro, is currently focusing on REULR technology.
In summary, bispecific molecules like AbTACs, PROTABs, and REULRs targeting transmembrane E3 ligases (e.g., RNF43, ZNRF3, RNF128, RNF130, and RNF167) can selectively degrade key membrane proteins like EGFR, HER2, IGF1R, PD-L1, and PD1 for therapeutic purposes.
It is well known that many cytokines are internalized and degraded via lysosomes by their receptors. This provides another opportunity to harness endogenous eTPD mechanisms using a method known as KineTAC (Cytokine Receptor Targeting Chimeras).
The first example used the decoy receptor CXCR7, which internalizes its ligand CXCL12 without downstream signaling. A study created an atezolizumab-KineTAC using CXCL12. In vitro, PD-L1 levels were reduced by ~70% at 24 hours and up to 84% at 48 hours. Similar results were observed with a trastuzumab-KineTAC targeting HER2, achieving 60% degradation on MDA-MB-175VII cells in 24 hours.
Like AbTACs, KineTACs can degrade a variety of membrane proteins across cell types. Similar to LYTACs, they can also degrade soluble proteins since trafficking is ubiquitin-independent. KineTACs appear to act catalytically as CXCR7 levels remained unchanged during degradation. Moreover, even with high POI transcript levels, effective degradation still occurred.
Inspired by using integrin αVβ3 to deliver anticancer drugs, an integrin-based eTPD system was developed. These adhesion molecules, often overexpressed in cancers, serve as mediators between cells and the extracellular matrix. Integrins are attractive targets because they bind to ligands containing a simple Arg-Gly-Asp (RGD) motif and transport them to lysosomes.
As proof of concept, a biotin-based chimera targeting biotin-binding protein NeutrAvidin was created using a cyclic RGD motif (cRGD) covalently linked to biotin. NeutrAvidin was successfully internalized in A549 cells within 20 hours. The internalization co-localized with LysoTracker and was blocked by lysosomal inhibitors. A PD-L1 binding small molecule (BMS-8) was conjugated with cRGD; in xenograft mouse models, this chimera reduced body weight loss, tumor growth, and spleen metastasis significantly compared to BMS-8 alone. Overall, this integrin-based system shows exciting potential for tumor-selective degradation of membrane or soluble proteins.
Receptor tyrosine kinases (RTKs) have been targeted by PROTACs, which link RTK inhibitors to CRBN or VHL E3 ligase binders. C4 Therapeutics developed an EGFR degrader (L858R), CFT8919, which has received IND approval from the FDA.
Compared to inhibition alone, degradation of receptors offers several advantages. PROTACs targeting MET were more effective at reducing cell viability and exhibited prolonged responses post-washout, indicating durable effects. Additionally, degradation induced less kinase reprogramming—a common resistance mechanism to classical RTK inhibitors. The degradation was entirely proteasome-dependent and, surprisingly, clathrin-independent.
In conclusion, using PROTACs that bind VHL or CRBN can degrade RTKs intracellularly. However, the degradation kinetics are slower, and the compounds are less potent than biologics in terms of IC50. Their efficacy is also limited by challenges in cell permeability needed for entering and reaching the intracellular ligases. Despite being in early stages, this approach shows promise in targeting intracellular domains of membrane proteins.
The field of eTPD is still in its infancy, with many key questions yet to be answered. For which targets does eTPD provide clear advantages? What are the most critical POIs, and in which therapeutic areas? Is it more useful for chronic or acute diseases? Does eTPD primarily target soluble or membrane POIs—or both? How important is tissue selectivity for safe and effective eTPD drugs? What resistance mechanisms might emerge, and are they more likely than with other drug modalities?
In some ways, the development of eTPD resembles the evolution of aviation—early designs included many creative aircraft, from human-powered and motorized birds to multi-wing and fixed-wing planes. Though few early designs survived, their diversity was crucial to advancing the technology and achieving widespread commercialization. Similarly, eTPD is at an exciting stage, with many discoveries anticipated in the future.
Wells, J.A., Kumru, K. Extracellular targeted protein degradation: an emerging modality for drug discovery. Nat Rev Drug Discov 23, 126–140 (2024). https://doi.org/10.1038/s41573-023-00833-z
