Introduction
G protein-coupled receptors (GPCRs) are one of the most important classes of drug targets in modern medicine, with approximately 40% of all prescription drugs targeting these receptors (1). However, they have long been a thorn in the side of drug discovery researchers.
This large family of transmembrane proteins play a critical role in cellular communication, facilitating the coordination of activities across cells and tissues, and is involved in many major illnesses such as cardiovascular disorders and cancer. Yet, despite their significance, GPCRs remain notoriously difficult to target therapeutically: the unstable, complex structures and hard-to-reach binding sites pose significant obstacles for traditional methods of drug discovery, particularly those using small molecules. Consequently, many diseases linked to GPCRs remain ‘undrugged’ – at the time of writing, there are over 200 GPCRs (2) within known diseases where effective therapies remain elusive.
To tackle this issue, the binding mechanisms between GPCRs and small molecules and antibodies must be considered. Antibodies offer several key advantages over small molecules, such as enhanced target selectivity, reduced off-target binding, and the ability to stabilise GPCR conformations. However, despite these clear benefits, developing effective antibody therapies for GPCRs remains complex due to size and binding difficulties.
With its growing adoption in drug discovery, generative AI (GenAI) offers a potential solution, not only enabling precise antibody design for these previously intractable targets but also expediting the drug discovery process from years to months. Here, Murat Tunaboylu, CEO and founder of Antiverse, considers how innovative applications of GenAI can help overcome the longstanding challenges surrounding drug design for GPCRs.
One of drug design’s biggest challenges: why are GPCRs so hard to target?
With such a high prevalence of ‘undrugged’ GPCR-related diseases, there is a huge, unmet patient need. Addressing this chasm is critical to improving access to treatment and population health. Currently, GPCRs’ reputation as elusive targets for drug design is due to several factors:
• Complex and dynamic structures: GPCRs exhibit shape-shifting behaviour; they are not static or fixed in nature and undergo conformational changes as part of the ligand binding process. This means a drug that binds effectively to one conformation may not bind to others.
• Hard to reach binding sites: GPCR binding sites are not easily accessible – typically embedded deep within the protein’s molecular structure or difficult-to-reach regions. Therefore, effective drug candidates must possess both sufficient size to access the binding site and high specificity to ensure selective binding.
• Specificity: Specificity is a critical factor, ensuring that drug targets bind effectively to the intended GPCR target and not to other receptors. Non-specific binding can lead to off-target side effects and reduced treatment efficacy.
• Transmembrane nature: GPCRs are embedded in the cell membrane, which holds its shape and structure; however, functional structure is lost when the GCPR is removed, which makes them difficult to study in isolation. Effective treatment options cannot be designed in this disordered state, so drug design requires an understanding of the 3D structure of a target protein.
These well-documented challenges of GPCRs pose many limitations in traditional drug design methods, which have historically relied heavily on the use of small molecules.
Roadblocks to progress: the limitations of small-molecule drug design
The longstanding dominance of small molecules in the drug design process stems from the well-established development pathways and relative ease of synthesis. However, for complex and dynamic targets like GPCRs, small molecules are not well suited for several reasons.
Low selectivity
This distinctive feature makes achieving target exclusivity difficult, particularly for receptor families that share structural homology like GPCRs. As a result, off-target binding is not uncommon, and subsequent activation of cellular signalling pathways can lead to unwanted side effects. Additionally, the dynamic and conformational flexibility of GPCRs, coupled with small molecules’ limited surface area, makes it challenging to achieve the specificity required to effectively target GPCRs.
Chemical property limitations
Small molecules are limited in size and often designed to be lipophilic, allowing them to pass through cell membranes. However, this preference for hydrophobic environments conflicts with complex binding requirements necessary for GPCRs. High lipophilicity also increases the risk of off-target interactions, potentially leading to adverse side effects and clinical failure.
The advantages and challenges of antibody-based therapeutics
Advantages
Antibodies present several advantages over small molecule therapeutics, particularly regarding specificity, duration of action, and functional versatility. The high-target selectivity and low propensity for off-target binding compared to small molecules contribute to a more favorable safety profile (3). This superior specificity is largely due to the ability to recognise and effectively bind to large, complex, and often unique epitopes – binding sites that are typically inaccessible to small molecules.
Antibodies also benefit from longer half-lives, enabling prolonged therapeutic activity and reduced dosing frequency. They can also engage with and modulate the body’s immune system to help attack the targeted cells, enhancing therapeutic response.
Perhaps most importantly for complex targets like GPCRs, antibodies can stabilise the conformational states of receptors. They represent critical tools for overcoming the complication of traditional drug design for dynamic targets, freezing or holding the receptor in a particular structural state.
Challenges
Despite these advantages, the development of antibodies in drug design has historically been complex. The large size can restrict access to certain epitopes, particularly those embedded deep within membrane proteins. Furthermore, despite high specificity, antibodies can exhibit cross-reactivity, especially when target proteins have structural and sequence homology. This increases the risk of off-target binding, which can compromise therapeutic selectivity and increase the likelihood of unwanted side effects.
Engineering antibodies to overcome these challenges requires a deep understanding of structural biology, a rational approach to antibody design and high-throughput screening. The increasing adoption of AI-driven modeling and optimisation offers a novel suite of tools to advance and accelerate this process, significantly reducing the time and cost associated with typical drug discovery while improving the likelihood of clinical success.
A technology-driven breakthrough: how GenAI can design antibodies to target GPCRs
Generative AI (GenAI) offers a powerful solution for overcoming the hurdles of drugging challenging targets using traditional drug discovery methods. By analysing structural and sequence data of targets – for example the protein sequences of antibody-antigen complexes, their structures analysed by X-ray crystallography or Cryo-EM, – and exploring design spaces at scale, GenAI enables more precise and efficient development of novel therapeutic candidates. GenAI can not only design highly specific antibodies with an improved likelihood of clinical success, but also accelerate the discovery process to deliver patient impact faster.
It is possible to analyse large volumes of structural and sequence data of challenging targets like GPCRs, using a GenAI platform to design epitope-specific antibody libraries that have high confidence against the target region. Where binders are rare, this comprehensive approach increases the likelihood of generating successful antibody candidates within six months. To progress towards real-world clinical success, these antibodies can then be screened against hyperexpressing stable cell lines to validate their function.
Where traditional drug discovery methods that rely on lengthy random-sampling processes prove inefficient for challenging targets like GPCRs, the application of GenAI offers a rational, data-driven approach to drug design. By allowing us to tackle these targets with greater accuracy, GenAI plays a critical role in reducing discovery timelines, maximising clinical success and meeting huge areas of unmet clinical and patient need.
Conclusion
The path to effective GPCR therapies has been fraught with difficulty, stemming from the intricate nature of these essential protein receptors.
The “undrugged” status of numerous disease-linked GPCRs underscores the pressing need to improve the lives of patients with innovative solutions that overcome these long-standing obstacles.
The rise of GenAI is ushering in a new era of therapeutic possibilities from previously intractable targets. It is now enabling the design of highly specific antibodies by leveraging advanced algorithms and machine learning, tackling the structural limitations that have traditionally plagued GPCR targeting.
The creation of epitope-specific libraries is transforming our approach to drug design, allowing researchers to accelerate the drug discovery process exponentially. While challenges undoubtedly remain, the potential for GenAI to change lives by unlocking effective treatments
for previously untreatable diseases is undeniable.
References and notes
- Insel PA, Tang C-M, Hahntow I, Michel MC. G Protein-Coupled Receptors as Targets for Approved Drugs: How Many Targets and How Many Drugs? Pharmacol Rev. 2018;70(1):3-25. doi:10.1124/pr.117.014894
- Hauser AS, Attwood MM, Rask-Andersen M, Schiöth HB, Gloriam DE. Trends in GPCR drug discovery: new agents, targets and indications. Nat Rev Drug Discov. 2017;16(12):829-842. doi:10.1038/nrd.2017.178
- Gao Y, Wang Y, Chen Y, Wang Y, Wang X, Li Y, et al. The role of gut microbiota in the pathogenesis of Parkinson’s disease. Translational Neurodegeneration. 2022;11(1):62. doi:10.1186/s40035-022-00338-0.
