All posts in Journal Club

Hit-to-Lead: Turning Hits into Promising Drug Candidates

This month’s article, authored by a group of researchers from multiple institutions collaborating with the European Federation for Medicinal Chemistry and Chemical Biology (EFMC) (https://doi.org/10.1002/cmdc.202400931), explores the critical transition from initial drug discovery hits to optimized lead compounds, highlighting key challenges and strategies in this pivotal phase of drug development.

The Hit-to-Lead (H2L) stage is a critical phase in drug discovery, where initial hits identified through high-throughput screening, virtual screening, or small molecule/fragment-based approaches undergo systematic refinement to yield lead compounds with improved pharmacological and physicochemical properties. This process marks a key decision-making juncture, ensuring that only the most promising chemical series advance into lead optimization while mitigating liabilities related to potency, selectivity, pharmacokinetics, and toxicity. While these principles have been extensively applied to small molecules, they are equally relevant in peptide drug discovery, where additional challenges—such as proteolytic stability, permeability, and half-life extension—must be addressed. Strategies like backbone modifications, non-natural amino acid incorporation, and macrocyclization help optimize peptide leads, paralleling the rational design approaches used for small molecules.

A fundamental aspect of H2L is the prioritization and selection of chemical matter based on a comprehensive assessment. While potency is an essential parameter, it must be balanced with selectivity to minimize off-target interactions that could lead to adverse effects. Physicochemical properties such as lipophilicity, solubility, metabolic stability, and permeability play a crucial role in drug-likeness and should be optimized early. Excessive lipophilicity often leads to poor metabolic stability, rapid clearance, and off-target binding, whereas low permeability may result in poor bioavailability.

Lead optimization in H2L relies on iterative structure-activity relationship studies to enhance desirable attributes while addressing potential liabilities. Scaffold modifications, bioisosteric replacements, and strategic incorporation of hydrogen bond donors/acceptors help fine-tune potency, selectivity, and PK properties. The application of ligand efficiency and lipophilic ligand efficiency metrics supports the selection of compounds that achieve potency gains without excessive increases in molecular weight or lipophilicity. Computational modeling and AI-driven predictive analytics further enhance the efficiency of H2L by providing insights into molecular behavior, protein-ligand interactions, and potential liabilities before extensive experimental validation.

A crucial consideration in H2L is early pharmacokinetic and pharmacodynamic evaluation, as in vitro potency alone does not guarantee in vivo efficacy. Pharmacokinetic profiling—including metabolic stability, plasma protein binding, permeability, and clearance—identifies compounds with favorable ADME properties. In parallel, early toxicological assessments help identify potential safety concerns before significant resources are invested in lead optimization.

The decision to progress or discontinue a chemical series is a key turning point in the H2L process. If a series fails to achieve an acceptable balance between potency, selectivity, and drug-like properties, it may be prudent to deprioritize it in favor of alternative scaffolds. Advanced analytics, including machine learning-guided predictive modeling, are increasingly utilized to streamline decision-making and reduce risk before committing to extensive medicinal chemistry efforts.

In conclusion, the Hit-to-Lead stage is a data-driven, multidisciplinary process that integrates medicinal chemistry, computational modeling, structural biology, and pharmacology to refine initial hits into viable lead compounds. Researchers can improve the likelihood of success in lead optimization and clinical development by optimizing key parameters such as potency, selectivity, pharmacokinetic properties, and synthetic feasibility.

To provide further insights, the authors have organized a webinar on H2L strategies, where experts will discuss best practices and real-world case studies. You can access the webinar at (https://www.efmc.info/hit-to-lead).

Open access article: https://doi.org/10.1002/cmdc.202400931

The European Federation for Medicinal Chemistry and Chemical Biology (EFMC) Best Practice Initiative: Hit to Lead, J. Quancard, A. Bach, C. Borsari, R. Craft, C. Gnamm, S. M. Guéret, I. V. Hartung, H. F. Koolman, S. Laufer, S. Lepri, J. Messinger, K. Ritter, G. Sbardella, A. Unzue Lopez, M. K. Willwacher, B. Cox, R. J. Young, ChemMedChem 2025, e202400931.

Sepideh Afshar, Ph.D.
Head of Peptide Therapeutics - Senior Director, Genentech
Member, BPF Scientific Advisory Board
linkedin.com/in/sepideh-afshar-bab3824

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club

AI Meets Peptides: Revolutionizing Therapeutic Discovery with Machine Learning

This month, we want to draw your attention to a review article by Goles et al. covering the field of artificial intelligence applied to peptide therapeutics (https://doi.org/10.1093/bib/bbae275).

This article provides a comprehensive exploration of the role of machine learning (ML) and artificial intelligence (AI) in advancing peptide drug discovery, with a strong focus on de novo design and optimization. The paper highlights several advanced ML techniques, including classifier methods for activity prediction, numerical property estimations, and deep generative models (DGMs) like generative adversarial networks (GANs) and variational autoencoders (VAEs) for sequence generation. Deep generative models are central to the de novo design of therapeutic peptides. VAEs, for instance, allow the encoding of peptide data into latent spaces to generate novel sequences resembling training data, while GANs iteratively refine the quality of generated sequences through adversarial training. Newer approaches, such as diffusion models, offer even higher fidelity in sequence generation by leveraging iterative, reversible transformations to fine-tune peptide designs. The use of predictive models alongside these generative techniques facilitates the assessment of physicochemical properties and binding affinities, streamlining the identification of promising candidates before experimental testing.

A unified AI-driven pipeline proposed in the article integrates several critical steps to optimize therapeutic peptide discovery. The process begins with functional classification models, which categorize peptide activity and evaluate properties. This is followed by affinity prediction models that assess peptide interactions with target proteins using deep learning architectures like graph convolutional networks. Generative methods then explore peptide sequence space, creating candidates with specific biological activities and favorable structural properties. The peptide properties are then optimized in silico for stability, toxicity, and pharmacokinetics. By
incorporating iterative validation and reinforcement learning, the pipeline allows continuous refinement of predictive and generative models using experimental feedback​.

Finally, the authors address the significant challenges that remain in advancing the field of predictive peptide design. One primary issue is the lack of centralized, high-quality peptide datasets, as fragmented and inconsistent data often lead to misclassification and hinder model training. Additionally, post-translational modifications (PTMs), critical to many of natural peptides' functional roles in vivo, are frequently overlooked in current design pipelines. Moonlighting effects, where peptides exhibit multiple biological activities, further complicate the prediction of specific properties. Addressing these challenges will enable the rapid and precise design of highly specialized peptides that can tackle a wide range of indications with unprecedented efficiency and specificity.

Open access article: https://doi.org/10.1093/bib/bbae275

Goles M., Daza A., Cabas-Mora G., Sarmiento-Varón L., Sepúlveda-Yañez J., Anvari-Kazemabad H., Davari M.D., Uribe-Paredes R., Olivera-Nappa A., Navarrete M.A., Medina-Ortiz D., Peptide-based drug discovery through artificial intelligence: towards an autonomous design
of therapeutic peptides, Briefings in Bioinformatics, Volume 25, Issue 4, July 2024.

Ewa Lis, Ph.D.
Founder/ CEO, Koliber Biosciences
Member, BPF Scientific Advisory Board
https://linkedin.com/in/ewa-lis-6b99528

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club

A Sustainable Solid-Phase Peptide Synthesis Approach

This month, we want to draw your attention to the topic of sustainability in peptide synthesis. The article “In Situ Fmoc Removal – A Sustainable Solid-Phase Peptide Synthesis Approach” (Kumar et al., 2022), while a few years old, remains highly relevant. The authors pave the way for the concept of peptide sustainability and green chemistry by introducing a clever method to reduce solvent waste.

Until green solvents for solid-phase peptide synthesis (SPPS) become widely accepted in the industry, it's crucial to work towards minimizing the environmental impact of existing SPPS practices. This article addresses one of the largest sources of solvent consumption in solid peptide synthesis: the solvents used for resin washes.

The authors propose a method that combines acylation and Fmoc deprotection by directly adding neat piperidine to the spent solution containing residual amino acids and oxyma. This approach effectively telescopes the acylation reaction into the deprotection step, which they refer to as in situ Fmoc deprotection. This practical solution is exactly what the industry needs to tackle the short-term issues related to process mass intensity (PMI) in SPPS. The authors demonstrate that this method does not compromise the final product; they achieve peptide materials of equal purity compared to standard SPPS.

Another reason we selected this article is that its corresponding author and inspiration is Professor Fernando Albericio.

Professor Albericio has been a prominent figure in solid-phase peptide chemistry for the past 50 years, developing numerous reagents that are essential for peptide chemists in their daily syntheses. Recently, he has made significant contributions to green chemistry, focusing on environmentally friendly methods for peptide synthesis. Given the rising interest in peptides as potential drugs, such as GLP-1, his work is increasingly important. His dedication to replacing hazardous solvents like DMF, DCM, and NMP with safer, less toxic alternatives is consistently evident in his numerous recent publications on this topic.

Professor Albericio was recently honored with the Meienhofer Award for excellence in peptide sciences at the Boulder Peptide Symposium in Boulder, CO. His lecture was well received by attendees. His research emphasizes scaling up peptide synthesis for pharmaceutical applications, enabling the production of large quantities of peptides with minimal environmental impact. He has pioneered the use of greener solvents, such as water, achieving high yields and purity, and has conducted detailed studies on the properties of green solvents in SPPS. Additionally, he is exploring new protecting groups that are easily removed and generate less waste.

His career continues to inspire many students, and his commitment to green chemistry is a testament to his dedication to the field of peptide chemistry.

Kumar, A., Sharma, A., de la Torre, B.G., and Albericio, F. (2022). In situ Fmoc removal – a sustainable solid-phase peptide synthesis approach. Green Chemistry, 24, 4887-4896. Link to publication

Professor Pravin Kaumaya, Ph.D.
Indiana University School of Medicine
Member, BPF Scientific Advisory Board

Matteo Villain, PharmD.
Member, BPF Scientific Advisory Board
linkedin.com/in/matteo-villain

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club

Treating Fungal infections with synthetic peptide mimics

Modern medicine often relies on invasive medical interventions or drugs that can compromise the patient’s immune system. An unfortunate consequence of these undeniably successful treatments for life-threatening diseases like cancer is the occurrence of severe infections caused by opportunistic fungal pathogens, including species of Candida, Aspergillus, Cryptococcus, and Pneumocystis. More recently, increasing numbers of opportunistic fungal infections caused by Aspergillus, Mucorales, and Candida species have been observed in COVID-19 patients with severe respiratory syndromes in intensive care units. These factors, among others, have resulted in over 2 million invasive fungal infections annually worldwide, with alarmingly high mortality rates, causing more than 1.5 million deaths.

Candida species are the most common cause of nosocomial, invasive fungal infections and are associated with mortality rates above 40%. Despite the increasing incidence of drug-resistance, the development of novel antifungal formulations has been limited. In a recent publication by Schaefer, S., Vij, R., Sprague, J.L. et al., they highlight the enormous potential of synthetic peptide mimics to be used as novel antifungal formulations as well as adjunctive antifungal therapy.

Schaefer, S., Vij, R., Sprague, J.L. et al. A synthetic peptide mimic kills Candida albicans and synergistically prevents infection. Nat Commun 15, 6818 (2024). Link to publication

Trishul Shah
Global Director Sales and Marketing, PolyPeptide Group
Member, BPF Scientific Advisory Board
www.linkedin.com/in/trishul-shah-bb92a2/

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club

A Knotty Challenge:  Novel cystine-knot peptide inhibitors of HTRA1

The July BPS journal club features an article showcasing the power of cystine-knot peptides (CKPs) as high-affinity and high-specificity protein inhibitors. A team from the laboratories of Rami Hannoush and Daniel Kirchhofer at Genentech created diversity libraries based on the carboxypeptidase A1 inhibitor and Ecballium elaterium trypsin inhibitor II CKPs. They performed phage display and affinity maturation studies to select potent inhibitors of a different serine protease, HTRA1 (https://doi.org/10.1038/s41467-024-48655-w)¹. The authors then characterized their modes of binding in structural studies and exploited differences in this binding pocket between serine protease family members to increase selectivity to the target. Specificity is an especially challenging problem in the serine protease family due to the large number of related family members with a fairly conserved binding pocket, including four highly related members of the HTRA family, HTRA1-4.

HTRA1, or high-temperature requirement A serine peptidase 1, is a trimeric serine protease implicated in several disorders, including arthritis, osteoporosis, age-related macular degeneration, Alzheimer’s, and chemoresistance. Both antibody-based (https://doi.org/10.1073/pnas.1917608117)² and small molecule, peptidomimetic-like inhibitors of HTRA1 (https://doi.org/10.1016/j.bmcl.2024.129814)³ have been identified, and clinical studies have been initiated with a Fab inhibitor of HTRA1 for age-related macular degeneration (NCT03972709).

Cystine-knots are structural motifs in proteins comprising three disulfide bridges between cysteine amino acids, where one strand or loop passes through another loop, forming a rotaxane structure that is protease resistant and shows both chemical and thermal stability. The core portions of CKPs have relatively low molecular weights (generally around 30 amino acids and less than 4 kDa) and are naturally occurring in animals, plants, and insects. Some common examples include nerve growth factor, transforming growth factor-beta, and various venom peptides from snails, spiders, and scorpions. They can also be chemically synthesized or manufactured using biological expression systems, making them excellent candidates for therapeutics.

The authors generated phage peptide libraries by introducing amino acid randomization and length variation into the surface-exposed loops of the CKP scaffolds and screened the resulting phage for binding to HTRA1 constructs by phage ELISA. Hits from the initial screen were modified through affinity maturation studies, and several inhibitors were identified with single-digit nanomolar affinities that also showed IC₅₀s for HTRA1 activity on a substrate in the same range. Biochemical studies were then used to map the approximate binding regions for the compounds on HTRA1, followed by structural studies that demonstrated that the CKP inhibitors bind a cryptic pocket near the active site and stabilize a structure that renders HTRA1 non-competent for protease activity. This cryptic pocket showed some differences from other members of the HTRA protease family, and the authors used a combination of structure-based design and further affinity maturation to exploit these differences and generate inhibitors with both high potency and high selectivity for HTRA1, even when compared to HTRA2, 3, and 4.

This article is a fantastic example of collaborative, multidisciplinary work using genetics, biochemistry, and structural biology to identify peptide inhibitors with high affinity and selectivity for a challenging target. Furthermore, it is another example of optimizing naturally occurring structures to develop candidate therapeutics that have the potential to be stable and well-tolerated since there are numerous examples of CKPs in biology. I look forward to seeing the effects of these inhibitors in animal models of diseases related to HTRA1 in the near future, as well as publications using a similar approach to identify cystine-knot peptide inhibitors of other challenging targets.

Dave Garman, PhD
Member, BPF Scientific Advisory Board

¹Li, Yanjie et al.  Cystine-Knot peptide inhibitors of HTRA1 bind to a cryptic pocket within the active site region.  Nature Communications 15:4359 (2024)

²Tom, Irene et al.  Development of a therapeutic anti-HtrA1 antibody and the identification of DKK3 as a pharmacodynamic biomarker in geographic atrophy.  PNAS 117(18):9952-63 (2020)

³Dennis, David G et al.  Identification of highly potent and selective HTRA1 inhibitors.  Bioorganic & Medicinal Chemistry Letters 109, September 2024 (available online ahead of print)

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club

Green Chemistry in Peptide Therapeutics: Leading the Way from the Beginning

In recent years, the field of peptide therapeutics has seen a surge of interest in cyclic peptides, driven by their unique properties and potential as active pharmaceutical ingredients (APIs). Cyclic peptides are particularly appealing due to their ability to be designed with desirable membrane solubility and oral absorption properties, which opens up the opportunity to target intracellular receptors and proteins effectively.

One of the key advantages of cyclic peptides is their rigid structure, which presents a large surface area capable of interacting with protein-protein interfaces. This makes them ideal candidates for disrupting protein interactions that are otherwise challenging to target with small molecules.

The development and screening of complex cyclic peptide libraries, especially those incorporating non-canonical amino acids, have become standard practices in many research laboratories. This advancement is exemplified by companies like Peptidream and RA Pharma, which have demonstrated success using mRNA libraries. The incorporation of non-canonical amino acids significantly expands the diversity and properties of these peptides, overcoming the limitations of the 20 natural amino acids.

However, once a promising lead is identified, scaling up the production of these peptides presents significant challenges. The synthesis of large quantities of non-canonical amino acids required for clinical programs is often complex and expensive, potentially delaying projects and increasing costs. To address this, several major pharmaceutical companies are investing in innovative platforms for the production of non-canonical building blocks that are both scalable and cost-effective.

An excellent example of such innovation is highlighted in an open-access article by Merck & Co. Inc. The article outlines a general approach for producing beta-branched aryl chiral amino acids using biocatalysis. Merck’s approach focuses on three key prerequisites: i) conducting reactions in water, ii) using an enzyme with broad selectivity and high chiral specificity, and iii) utilizing a cheap and readily available substrate as the amino donor.

In their study, Merck successfully used lysine as an amino donor, racemic aryl keto acids as the beta-branched starting materials, and transaminase as the enzyme to achieve high-yield, stereoselective reactions. Remarkably, the reactions proceeded with yields averaging above 70%, consistently producing the S,S configuration.

These beta-branched amino acids are now likely being incorporated into cyclic peptide libraries produced by Merck and its partners. This development is particularly exciting because the ability to lock the conformation of sidechain rotamers is a critical parameter for the success of these libraries.

In our April Journal Club edition, we discussed the necessity for peptide chemistry to become lean and green, emphasizing the adoption of green chemistry early in the API lifecycle. Merck’s efforts exemplify this principle by integrating green chemistry into their lead compound library screening process (we hope), ensuring that the compounds selected can be produced sustainably.

This article from Merck provides a compelling case study of how green chemistry can be implemented effectively in peptide therapeutics, paving the way for more sustainable and efficient drug development processes.

Article Citation: Dunham N, Ray R, Eberle C, Winston M, Newman J, Gao Q, et al. Efficient Access to β-Branched Noncanonical Amino Acids via Transaminase-Catalyzed Dynamic Kinetic Resolutions Driven by a Lysine Amine Donor. ChemRxiv. 2024; doi:10.26434/chemrxiv-2024-qm6pl. This content is a preprint and has not been peer-reviewed. The authors have deposited a ChemRxiv open-access version at https://chemrxiv.org/engage/chemrxiv/article-details/660da6f221291e5d1dfe8287.

Matteo Villain, PharmD.
Member, BPF Scientific Advisory Board
linkedin.com/in/matteo-villain

Read previous editions of the BPF Journal Club series: https://www.boulderpeptide.org/journal-club