Targeted protein degradation is an emerging therapeutic modality, which has the potential to tackle diseases in a unique way. The process of protein degradation induces selective protein elimination without the requirement for genetic modification to cells, which would be a more expensive approach.

There are up to three components to a targeted protein degrader: an E3 ubiquitin ligase ligand; a linker; and a ligand for a target protein of interest. Where the three components are distinct structural subunits, the molecules are known as bifunctional targeted protein degraders (BTPDs). By simultaneously binding to both an E3 ligase and target protein, the molecules form a ternary complex, inducing subsequent degradation. This concept of ubiquitin recruitment was first described in patent literature in 1999,¹ but has not progressed since. A paper coined the term ‘proteolysis targeting chimera’ (PROTAC) two years later in 2001, but the first generation of molecules synthesized were peptide-based and not cell-permeable. The first small molecule protein degraders were described in 2008,¹ and since then, a wide range have been synthesized, recruiting different E3 ligase enzymes, including cereblon (CRBN), von Hippel Lindau (VHL)² and inhibitor of apoptosis (IAP).³ Research has also demonstrated a wide variety of potential protein targets.

BTPDs are one of many approaches that, because of their rather large molecular structure, occupy a chemical space that is different from commercial drugs and other protein degraders such as molecular glues. BTPDs have a unique and complex molecular structure, with multiple hydrogen bond acceptors, hydrogen bond donor groups, and hydrophobic moieties, which could allow them to form intramolecular interactions. When comparing the molecular properties of BTPDs such as ionization, molecular size, lipophilicity, and polarity, it is important to consider their impact on molecule solubility, stability, and permeability. These factors are key to determining oral bioavailability, as well as metabolism. A BTPD’s chameleonicity is the molecular property that allows the compounds to adapt their conformations to the environment they encounter. Chameleonicity is believed to be an important quality that facilitates drugs outside of Lipinski’s rule of five (Ro5) space to display satisfactory cell permeability and oral absorption, and can be quantified as ChamelogD,⁴ which is the difference between two lipophilicity index determiners (BRlogD and ElogD), in two different systems.

Oral delivery of bifunctional targeted protein degraders

Delivering BTPDs orally tends to be challenging, particularly when factoring in their oral bioavailability that is compounded by their large molecular size. Additionally, their make-up can vary by degrader type, for example, molecules that recruit the VHL enzyme have a different structure to those that recruit the CRBN enzyme for degradation, due to the larger size of the VHL recruiting region. Nonetheless, there have been many instances where oral delivery has proven feasible for molecules with properties that are beyond the Ro5, and this also extends to BTPDs. BTPD molecules with unfavorable properties such as higher lipophilicity can be orally bioavailable, and this was reported by a team at AstraZeneca in 2020.⁴

Due to the limited guidance on how best to achieve bioavailability in the beyond Ro5 space, most innovators rely on drug development and delivery experts to overcome these challenges using enabling formulation technologies. A major misconception is that poor solubility is the only liability to unlocking bioavailability, and that using an enabling formulation technology to fix a solubility issue to progress into pharmacokinetic (PK) and toxicology studies for example, is sufficient. However, this approach frequently leads to the later discovery of other problems with the molecule that were put aside while trying to resolve poor solubility, leading to delay, which could potentially take months to overcome.

Involve experts early

Having input from a drug development expert who is familiar with the application of enabling formulations at the lead optimization stage is beneficial, as they can provide vital input into key elements of any program to maximize the chances of success. Firstly, their experience from developing other similar molecules will help guide candidate selection and ensure that the molecule chosen for progression is feasible for oral delivery. Additionally, they will be able to influence the choice of the appropriate formulation technology and determine whether an enabling technology would be needed or not. Finally, they can provide insight into the right dose form and manufacturing process for future clinical supplies and advise on the development path needed to achieve this. This might not be immediately crucial if the first study is in healthy volunteers, but there should be a plan in place for the transition to patients.

Often the molecules are locked in before they reach the formulation development stage, and there can be no influence on their choice, so, for a formulator must work with what has been chosen. This can lead to making compromises about which formulation tools and enabling technologies can be used with the selected API form.

Putting this into the context of development milestones, while it is common to engage a formulator at preclinical to IND enabling stage, it can be more advantageous to think about the concept of a “developability assessment” as part of the discovery phase. In the best-case scenario, this assessment will potentially include a review of a handful of lead candidates and a discussion as to whether an enabling technology will be required for any or all of them. In the longer term, this could prove preferable to adapting the lead candidate without engaging a formulator who will later be asked to provide an opinion on what can be done with a single candidate where salt or free form have already been decided, and where enabling technologies may be incompatible with salt selection.

Developability assessment

When embarking on such a developability assessment, it is essential to begin by listing all the liabilities that can be observed in a molecule to guide the formulation decision. These liabilities can include, but are not limited to, solubility, permeability, stability, and dose range, while limitations of available API may also discount some approaches. Not all issues and risks necessarily need to be addressed at that moment, but it is important to be aware of potential future manufacturing issues such as shelf life and processability during dosage form design and optimization.

In the case of selecting orally developable BTPDs, a four-stage rational approach, supported by class-specific toolkits may be employed.

The first step is looking at a series of candidates and preparing a risk-ranked assessment based on each of the molecules’ descriptors. This can be undertaken in silico, but partnering with formulation experts early on has the advantage of ensuring that the required dataset for each molecule is aligned and understood to enable a full evaluation.

The next step is to expand the physicochemical screen to define the challenges and parameters in the development of an oral formulation, which should include assays that are specific to bifunctional protein degraders.

It is then important to assess available data, such as in vitro and in vivo evaluations which can guide the initial ranking of a molecule in the Developability Classification System (DCS). The DCS is a useful early tool to screen several molecules and determine whether each molecule is likely to be permeability or solubility limited for it to become systematically bioavailable and provides insights into the appropriate formulation technology approach for each.

The final step is then to build an enhanced physiologically based pharmacokinetic (PBPK) model to highlight challenges that might limit human oral bioavailability, and subsequently refine this based on new data from assays and testing as they become available.

PBPK modeling

Using PBPK modeling helps guide the developability assessment for BTPDs and is useful in looking at parameters that may be most critical in limiting oral bioavailability.

The primary piece of information to be determined is the fraction absorbed (F_a), which is the fraction of an administered drug that permeates the first biological barrier, in the oral case this is the gut wall. If the F_a is not very high, an enabling formulation will most likely be needed. There is also the fraction escaping gut metabolism (F_g) and the fraction escaping liver metabolism (F_h) and multiplying these three fractions together gives an estimation of bioavailability.

Further information that can be determined through assays to improve the model includes permeability, efflux, and metabolism. Either a Caco-2 or MDCK-MDR-1 assay can determine the in vitro permeability of a compound, with the MDR-1 transfected MDCK cells being preferred for degraders, as it overexpresses permeability glycoprotein 1. This provides an understanding of the compound’s efflux rate.

Metabolism studies should preferably be carried out in hepatocytes because degraders are limited with regards to permeability, and so an intact cell must be used to recreate the need to permeate into the hepatocytes that are to be degraded. Using microsomes, for example, could give clearance data that overestimate the true in vitro situation.

It is crucial to use measured data when modeling BTPDs. ADME calculators and predictors are trained on small molecules that fit within the Ro5, making them inappropriate for larger molecules.

Potential formulation solutions

Having proposed a developability assessment to review physicochemical data and used molecular descriptors and ADME pharmacokinetic data to build a PBPK model, it becomes possible to identify the factors that limit bioavailability beyond the DCS assessment. This information can then be used to make informed recommendations regarding formulation.

There are a few typical outcomes: if human absorption suggests no solubility or permeability issues, a conventional dosage form such as a tablet; API or powder blend in capsule; API or powder blend in bottle; or liquid in bottle may be appropriate. If there is a solubility limitation or a solubility/permeability interplay, enabling technologies such as amorphous dispersions, lipid formulations or particle size reduction may assist. However, if oral dosing is hindered by extremely low fraction absorption, to the extent that effective oral dosing is impossible, an injectable formulation may be the only feasible option. For expedited Phase 1 trials in healthy volunteers, it is becoming more common to utilize extemporaneous preparations made at the clinical research organization rather than invest in lengthy finished dosage form and process development at a specialist CDMO.

Other development challenges

Chameleonicity could make molecules more difficult to crystallize as a powdered API, which increases the risk of changes during early scale-up.

Additionally, as three-part molecules have more functional groups present on them, stability and complexity can lead to a lack of stability that might not become evident in short-term studies.

High potency and toxicity may also be issues, and where the E3 ligase recruitment employs an imide-based structure and mechanism of action, these BTPDs will require enhanced safety handling precautions. Any formulation option, and especially where enabling technologies will be needed, should always be carried out with the appropriate level of containment and safety handling protocols.

Conclusion

Developing oral drug products with targeted protein degraders presents unique challenges that go well beyond those of the Ro5. However, taking a practical approach to development and engaging drug delivery experts from the lead optimization stage can be highly beneficial. Additionally, creating a customized PBPK model can help determine the feasibility of oral delivery for a given shortlist of bifunctional targeted protein degrader molecules. The approach of molecule developability assessment has enormous potential, as it allows for a comprehensive understanding of its potential challenges and opportunities at the time of candidate selection. Undoubtedly, the expertise of drug development experts remains indispensable as they play a critical role in bringing these molecules to market and making them available to patients.

References
1. Mullard. Article. Nature Reviews Drug Discovery 18, 237-239 (2019).
2. Wang C et al. Recent advances in IAP-based PROTACs (SNIPERs) as potential therapeutic agents. J Enzyme Inhib Med Chem. 2022 Dec;37(1):1437-1453.
3. Ermondi et al. “Updating the portfolio of physicochemical descriptors related to permeability in the beyond the rule of 5 chemical space” European Journal of Pharmaceutical Sciences, Volume 146, 2020, 105274.
4. Pike et al. “Optimizing proteolysis-targeting chimeras (PROTACs) for oral drug delivery: a drug metabolism and pharmacokinetics perspective” Drug Discovery Today 2020 Oct;25(10):1793-1800.