Rational design of peptide agonists of cell-surface receptors

Peptipedia v2.0: a peptide sequence database and user-friendly web platform. A major update

In recent years, peptides have gained significant relevance due to their therapeutic properties. The surge in peptide production and synthesis has generated vast amounts of data, enabling the creation of comprehensive databases and information repositories. Advances in sequencing techniques and artificial intelligence have further accelerated the design of tailor-made peptides. However, leveraging these techniques requires versatile and continuously updated storage systems, along with tools that facilitate peptide research and the implementation of machine learning for predictive systems. This work introduces Peptipedia v2.0, one of the most comprehensive public repositories of peptides, supporting biotechnological research by simplifying peptide study and annotation. Peptipedia v2.0 has expanded its collection by over 45% with peptide sequences that have reported biological activities. The functional biological activity tree has been revised and enhanced, incorporating new categories such as cosmetic and dermatological activities, molecular binding, and antiageing properties. Utilizing protein language models and machine learning, more than 90 binary classification models have been trained, validated, and incorporated into Peptipedia v2.0. These models exhibit average sensitivities and specificities of 0.877±0.0530 and 0.873±0.054, respectively, facilitating the annotation of more than 3.6 million peptide sequences with unknown biological activities, also registered in Peptipedia v2.0. Additionally, Peptipedia v2.0 introduces description tools based on structural and ontological properties and user-friendly machine learning tools to facilitate the application of machine learning strategies to study peptide sequences. Database URL: https://peptipedia.cl/.

Protein Language Models and Machine Learning Facilitate the Identification of Antimicrobial Peptides

Peptides are bioactive molecules whose functional versatility in living organisms has led to successful applications in diverse fields. In recent years, the amount of data describing peptide sequences and function collected in open repositories has substantially increased, allowing the application of more complex computational models to study the relations between the peptide composition and function. This work introduces AMP-Detector, a sequence-based classification model for the detection of peptides' functional biological activity, focusing on accelerating the discovery and de novo design of potential antimicrobial peptides (AMPs). AMP-Detector introduces a novel sequence-based pipeline to train binary classification models, integrating protein language models and machine learning algorithms. This pipeline produced 21 models targeting antimicrobial, antiviral, and antibacterial activity, achieving average precision exceeding 83%. Benchmark analyses revealed that our models outperformed existing methods for AMPs and delivered comparable results for other biological activity types. Utilizing the Peptide Atlas, we applied AMP-Detector to discover over 190,000 potential AMPs and demonstrated that it is an integrative approach with generative learning to aid in de novo design, resulting in over 500 novel AMPs. The combination of our methodology, robust models, and a generative design strategy offers a significant advancement in peptide-based drug discovery and represents a pivotal tool for therapeutic applications.

Development of Peptide Paratope Mimics Derived from the Anti-ROR1 Antibody and Long-Acting Peptide-Drug Conjugates for Targeted Cancer Therapy

Antibody-based targeted therapy in cancer faces a challenge due to uneven antibody distribution in solid tumors, hindering effective drug delivery. We addressed this by developing peptide mimetics with nanomolar-range affinity for Receptor Tyrosine Kinase-Like Orphan Receptor 1 (ROR1) using computational methods. These peptides showed both specific targeting and deep penetration in vitro and in vivo. Additionally, we created peptide-drug conjugates (PDCs) by linking targeting peptides to toxin drugs via various linkers and enhancing their in vivo half-life with fatty side chains for albumin binding. The antitumor candidate II-3 displayed exceptional affinity (KD = 1.72 × 10-9 M), internalization efficiency, anticancer potency (IC50 = 0.015 ± 0.002 μM), and pharmacokinetics (t1/2 = 2.6 h), showcasing a rational approach for designing PDCs with favorable tissue distribution and strong tumor penetration.

Peptide-based drug discovery through artificial intelligence: towards an autonomous design of therapeutic peptides

With their diverse biological activities, peptides are promising candidates for therapeutic applications, showing antimicrobial, antitumour and hormonal signalling capabilities. Despite their advantages, therapeutic peptides face challenges such as short half-life, limited oral bioavailability and susceptibility to plasma degradation. The rise of computational tools and artificial intelligence (AI) in peptide research has spurred the development of advanced methodologies and databases that are pivotal in the exploration of these complex macromolecules. This perspective delves into integrating AI in peptide development, encompassing classifier methods, predictive systems and the avant-garde design facilitated by deep-generative models like generative adversarial networks and variational autoencoders. There are still challenges, such as the need for processing optimization and careful validation of predictive models. This work outlines traditional strategies for machine learning model construction and training techniques and proposes a comprehensive AI-assisted peptide design and validation pipeline. The evolving landscape of peptide design using AI is emphasized, showcasing the practicality of these methods in expediting the development and discovery of novel peptides within the context of peptide-based drug discovery.

A novel human scFv fragment against TNF-α from de novo design method

Anti-TNF antibody has been an effective therapeutic strategy for the diseases related to aberrant production of TNF-alpha, such as rheumatoid arthritis (RA) and Crohn's disease. The limitations of large molecule inhibitors in the therapy of these diseases prompted the search for other potent novel TNF-alpha antagonists. Antagonistic peptides, derived directly or designed rationally from complementarity-determining regions (CDRs) of neutralizing antibodies against TNF-alpha, have been demonstrated for their ability of inhibiting TNF-alpha. However, their activity is very low. In this study, to increase the affinity and bioactivity, human antibody variable region was used as scaffold to display antagonistic peptides, which were designed on the interaction between TNF-alpha and its neutralizing monoclonal antibody (mAb Z12). Based on the previously designed domain antibody (framework V(H)5), framework V(kappa)1 was used as light chain scaffold. On the basis of computer-guided molecular design method, a novel human scFv fragment (named as TSA1) was designed. Theoretical analysis showed that TSA1 could bind to TNF-alpha with more hydrogen bonds and lower binding free energy than the designed domain antibody. The biological experiments demonstrated that TSA1 could directly bind with TNF-alpha, competitively inhibit the binding of mAb Z12 to TNF-alpha and block the binding of TNF-alpha to TNFR I and TNFR II. TSA1 could also inhibit TNF-induced cytotoxicity on L929 cells and TNF-mediated NF-kappaB activation on HEK-293T cells. The bioactivity of TSA1 was significantly increased over the domain antibody. This study indicated that the framework of antibody variable region could serve as an ideal scaffold for displaying the peptides and provides a novel strategy to design TNF-alpha inhibitors with the ability to block the deleterious biological effects of TNF-alpha.

Fusion protein of CDR mimetic peptide with Fc inhibit TNF-α induced cytotoxicity

The variable regions of antibodies play central roles in the binding with antigens. Based on the model of a tumour necrosis factor-alpha (TNF-alpha) neutralizing monoclonal antibody (named as Z12) with TNF-alpha, heavy chain CDR2 (HCDR2) and light chain CDR3 (LCDR3) of Z12 were found to be the most responsible to bind with TNF-alpha. A mimetic peptide (PT) was designed based on the sequence derived from HCDR2 and LCDR3. Fusion protein PT-Fc was constructed by linking PT with Fc of human IgG1 through a flexible linker (GGGGGS). The primary structural characteristics of Fc and PT-Fc were analyzed, including the flexibility, hydrophilicity and epitopes. It was demonstrated that PT and Fc in the fusion protein possessed bio-function properly and non-interfering with each other. Furthermore, PT-Fc was expressed in Escherichia coli by fusion with thioredoxin (Trx). After trx-PT-Fc was cleaved with recombinant enterokinase, PT-Fc was obtained. The results of in vitro cytotoxic assays showed that both PT and PT-Fc could efficiently inhibit TNF-alpha induced apoptosis on L929 cells. At the same micromole concentration, the inhibition activity of PT-Fc was significantly higher than PT.