Hit Identification Driven by Combining Artificial Intelligence and Computational Chemistry Methods: A PI5P4K-ß Case Study.

Computer-aided drug design (CADD), especially artificial intelligence-driven drug design (AIDD), is increasingly used in drug discovery. In this paper, a novel and efficient workflow for hit identification was developed within the ID4Inno drug discovery platform, featuring innovative artificial intelligence, high-accuracy computational chemistry, and high-performance cloud computing. The workflow was validated by discovering a few potent hit compounds (best IC50 is ∼0.80 µM) against PI5P4K-ß, a novel anti-cancer target. Furthermore, by applying the tools implemented in ID4Inno, we managed to optimize these hit compounds and finally obtained five hit series with different scaffolds, all of which showed high activity against PI5P4K-ß. These results demonstrate the effectiveness of ID4inno in driving hit identification based on artificial intelligence, computational chemistry, and cloud computing.

New Class of Heterogeneous Helical Peptidomimetics.

A new class of unnatural heterogeneous foldamers is reported to contain alternative α-amino acid and sulfono-γ-AA amino acid residues in a 1:1 repeat pattern. Two-dimensional NMR data show that two 1:1 α/sulfono-γ-AA peptides with diverse side chains form analogous right-handed helical structures in solution. The effects of sequence length, side chain, N-capping, and temperature on folding propensity were further investigated using circular dichroism and small-angle X-ray scattering.

Sulfono-γ-AApeptides as a new class of nonnatural helical foldamer.

Foldamers offer an attractive opportunity for the design of novel molecules that mimic the structures and functions of proteins and enzymes including biocatalysis and biomolecular recognition. Herein we report a new class of nonnatural helical sulfono-γ-AApeptide foldamers of varying lengths. The crystal structure of the sulfono-γ-AApeptide monomer S6 illustrates the intrinsic folding propensity of sulfono-γ-AApeptides, which likely originates from the bulkiness of tertiary sulfonamide moiety. The two-dimensional solution NMR spectroscopy data for the longest sequence S1 demonstrates a 10/16 right-handed helical structure. Optical analysis using circular dichroism further supports well- defined helical conformation of sulfono-γ-AApeptides in solution containing as few as five building blocks. Future development of sulfono-γ-AApeptides may lead to new foldamers with discrete functions, enabling expanded application in chemical biology and biomedical sciences.

Design, synthesis and characterization of fMLF-mimicking AApeptides.

The tripeptide N-formyl-Met-Leu-Phe (fMLF) is a potent neutrophil chemoattractant and the reference agonist for the G protein-coupled N-formylpeptide receptor (FPR). As it plays a very important role in host defense and inflammation, there has been considerable interest in the development of fMLF analogues in the hope of identifying potential therapeutic agents. Herein we report the design, synthesis, and evaluation of AApeptides that mimic the structure and function of fMLF. The lead AApeptides induced calcium mobilization and mitogen-activated protein kinase (MAPK) signal transduction pathways in FPR-transfected rat basophilic leukemic (RBL) cells. More intriguingly, at high concentrations, certain AApeptides were more effective than fMLF in the induction of calcium mobilization. Their agonistic activity is further supported by their ability to stimulate chemotaxis and the production of superoxide in HL-60 cells. Similarly to fMLF, these AApeptides are much more selective towards FPR1 than FPR2. These results suggest that the fMLF-mimicking AApeptides might emerge as a new class of therapeutic agents that target FPRs.

Synthesis of AApeptides.

The creation and development of nonnatural peptidomimetics has become an area of increasing significance in bioorganic and chemical biology. A wide range of new peptide mimics with novel structures and functions are urgently needed to be explored in order to identify potential drug candidates and targeted probes, and to study protein functions. AApeptides are a new class of peptide mimics based on chiral PNA backbone. They are resistant to proteolytic degradation and have limitless potential for diversification. They have been found to have a wide variety of biological applications including cellular translocation, disruption of protein-protein interactions, formation of nanostructures, antimicrobial activity, etc. The synthesis of AApeptides is modular and straightforward. In this chapter, methods for the synthesis of AApeptides (including different subclasses) are described.

AApeptides as a new class of antimicrobial agents.

Antibiotic resistance is an increasing public health concern around the world, and is recognized as one of the greatest threats facing humankind in the 21(st) century. Natural antimicrobial peptides (AMPs) are small cationic amphiphilic peptides found in virtually all living organisms, and play a key role in the defense against bacterial infections. Compared with conventional antibiotics, which target specific metabolic processes, AMPs are able to adopt globally amphipathic conformations, and kill bacteria through disruption of their membranes. As such, AMPs do not readily induce drug-resistance. However, AMPs are associated with intrinsic drawbacks such as low-to-moderate activity, susceptibility to enzymatic degradation, and inconvenience for optimization. Recently, we have developed a new class of peptidomimetics termed "AApeptides". Such peptide mimics are highly resistant to protease degradation and are straightforward for chemical diversification and development. Our current studies show that AApeptides with globally amphipathic structures can mimic the bactericidal mechanism of AMPs, and display potent and broad-spectrum activity against both Gram-positive and -negative multi-drug-resistant bacteria. In this review, we summarize our current findings of antimicrobial AApeptides, and discuss potential future directions on the development of more potent and specific analogues.

Lipo-γ-AApeptides as a new class of potent and broad-spectrum antimicrobial agents.

There is increasing demand to develop antimicrobial peptides (AMPs) as next generation antibiotic agents, as they have the potential to circumvent emerging drug resistance against conventional antibiotic treatments. Non-natural antimicrobial peptidomimetics are an ideal example of this, as they have significant potency and in vivo stability. Here we report for the first time the design of lipidated γ-AApeptides as antimicrobial agents. These lipo-γ-AApeptides show potent broad-spectrum activities against fungi and a series of Gram-positive and Gram-negative bacteria, including clinically relevant pathogens that are resistant to most antibiotics. We have analyzed their structure-function relationship and antimicrobial mechanisms using membrane depolarization and fluorescent microscopy assays. Introduction of unsaturated lipid chain significantly decreases hemolytic activity and thereby increases the selectivity. Furthermore, a representative lipo-γ-AApeptide did not induce drug resistance in S. aureus, even after 17 rounds of passaging. These results suggest that the lipo-γ-AApeptides have bactericidal mechanisms analogous to those of AMPs and have strong potential as a new class of novel antibiotic therapeutics.

Lipidated peptidomimetics with improved antimicrobial activity.

We report a series of lipidated α-AApeptides that mimic the structure and function of natural antimicrobial lipopeptides. Several short lipidated α-AApeptides show broad-spectrum activity against a range of clinically related Gram-positive and Gram-negative bacteria as well as fungus. Their antimicrobial activity and selectivity are comparable or even superior to the clinical candidate pexiganan as well as previously reported linear α-AApeptides. The further development of lipidated α-AApeptides will lead to a new class of antibiotics to combat drug resistance.

Development of NGR-based anti-cancer agents for targeted therapeutics and imaging.

Besides the common issue of drug-resistance, the conventional approaches for cancer diagnostics and treatment are constantly challenged by poor selectivity and limited access to neoplastic cells, which not only lead to the dose-limiting effect on the tumor region, but also bring side-effects to healthy cells/tissues. In recent years, a novel strategy has arisen to target the vasculature of tumors for drug-delivery and molecular imaging, based on the success of anti-angiogenic therapy. In addition to being easily accessible, the endothelial cells of tumor vasculature are also genetically stable and thus do not develop drug-resistance, making them ideal targets for chemotherapeutics and biomedical imaging. Among various ligands identified so far, the Asn-Gly-Arg (NGR) tripeptide can specifically target the neovasculature via interaction with the aminopeptidase N (APN/CD13) receptor which is highly up-regulated in the membranes of endothelial tumor cells. NGR-directed drug delivery as well as molecular imaging have therefore been undergone development, and appear to be intriguing approaches in current cancer research. Herein we highlight some recent developments of the NGR peptide based cancer therapy including drug-delivery and imaging studies, with future perspectives. Some of these agents have been under clinical trials, indicating promising future for the NGR-based drugs.

Development of self-immolative dendrimers for drug delivery and sensing.

Traditional dendrimers possess unique cascade-branched structural properties that allow for multivalent modifications with drug cargos, targeting/delivery agents and imaging tools. In addition to multivalency, the dendrimer's macromolecular size also brings about the enhanced permeability and retention (EPR) effect, which makes it an attracting agent for drug delivery and biosensing. Similar to other macromolecules, therapeutic application of dendrimers in the human body faces practical challenges such as target specificity and toxicity. The latter represents a substantial issue due to the dendrimer's unnatural chemical structure and relatively large size, which prohibit its in vivo degradation and excretion from the body. To date, a class of self-immolative dendrimers has been developed to overcome these obstacles, which takes advantage of its unique structural backbone to allow for cascade decompositions upon a simple triggering event. The specific drug release can be achieved through a careful design of the trigger, and as a result of the fragmentation, the generated small molecules are either biodegradable or easily excreted from the body. Though still at a preliminary stage, the development of this novel approach represents an important direction in nanoparticle-mediated drug delivery and sensor design, thereby opening up an insightful frontier of dendrimer based applications.

Non-hemolytic α-AApeptides as antimicrobial peptidomimetics.

We report a new class of peptide mimetics, α-AApeptides, that display broad-spectrum activity against both Gram-negative and Gram-positive bacteria and fungi. With non-hemolytic activity, resistance to protease hydrolysis, and easy sequence programmability, α-AApeptides may emerge as a novel class of antibiotics.

Design and synthesis of AApeptides: a new class of peptide mimics.

A new family of peptide mimics termed 'AApeptides', which are oligomers of N-acylated-N-aminoethyl amino acids, was proposed. The design and efficient synthesis of AApeptides are described. As proof-of-the-concept, we show that AApeptides can inhibit p53/MDM2 protein-protein interaction with significant activity (IC(50)=38 µM) and specificity. Preliminary data also demonstrates that AApeptides are resistant to enzymatic hydrolysis. With the ease of synthesis and diversification, potent bioactivity, and resistance to proteolysis, the development of sequence-specific AApeptides may expand the potential biomedical applications of peptidomimetics.