Biotech company Molecure develops small molecule therapies that can regulate RNA and underexplored protein targets for the treatment of incurable diseases, including cancer and fibrotic and inflammatory diseases.
The company, which was named as one of our 10 biotech companies leading the charge in Poland, told Labiotech that its RNA platform finds small molecules that target the mRNA of genes associated with cancer and inflammatory diseases.
The platform is designed to use small molecules to selectively “lock” the mRNA of interest in a specific conformation, making it unavailable for translation and thus preventing the expression of protein(s) involved in pathogenesis and disease progression.
“Molecure uses a combination of computational and experimental methods to identify and optimize hit compounds and has created a virtual library of 500,000 compounds that are specifically designed to bind to RNA. These compounds have been selected based on their structural and electronic features, which are known to interact with RNA,” said Molecure CEO, Marcin Szumowski.
Szumowski added: “We use bioinformatics methods to identify structurally stable regions within large mRNA molecules. Our computational platform uses quantum-level of theory to accurately represent the physicochemical properties of RNA, including its polarizability. This allows us to predict how small molecules will interact with RNA and how they will affect its structure.
“We start by using experimentally resolved 3D structures of an mRNA target to computationally investigate its dynamical behavior. We look for local structural changes that may lead to the development of pockets that can potentially bind small molecules. If some of these structural changes are too slow to be calculated, we use fine-tuned external forces to speed them up.”
Molecure’s approach: identifying draggable regions
Szumowski said the company always validates the results of its simulations against corresponding data obtained from experimental biological assays, integrating omics, data and information from scientific publications to confirm the functionality of selected regions with stable structures.
“Our approach also involves using bioinformatics methods to assess off-target effects on the human transcriptome. This helps us to identify druggable regions that are unlikely to have harmful side effects,” Szumowski said.
“The hit-to-lead stage of our platform builds on the deep knowledge of medicinal chemistry and the vast synthetic expertise of our team. We further improve our hit mRNA-binders by optimizing their interaction with RNA and by improving their cellular permeability, selectivity, DMPK, and toxicity.”\
Depending on the specific region of interest, screening methods such as DEL, NMR fragment-based screening, or virtual screening are used to identify hits that bind to the target region. The impact of these hits on protein translation is evaluated within cells. Molecule then uses hit optimization procedures by combining medicinal chemistry activities with chemoinformatics analysis as well as binding and cellular assays.
Other approaches
The first widely-used approach to designing small molecule binders to mRNA involved phenotypic screening, an untargeted strategy for the identification of molecules with particular biological effects.
Molecule said the challenge with phenotypic screening is that it can yield hits that do not bind to RNA, as well as other false positives. This is because phenotypic screens are not specific to RNA binders. In fact, almost any screen will generate hundreds or thousands of hits and the chances of finding molecules that bind to RNA are almost zero.
Another bottleneck with phenotypic screening is identifying the RNA that the compound is hitting, Szumowski noted.
“In order to do medicinal chemistry on a compound, one needs to know what RNA it binds to. This can be a difficult task, as phenotypic screens do not provide this information.”
To overcome this challenge, Molecure uses a target-based approach.
“Target-based approaches are more likely to yield hits that are specific to RNA. Our approach focuses on functional sequences on the mRNA of interest, which are more likely to be targeted by RNA binders
Molecure integrates experimental and computational structural biology with bio/chemo-informatics techniques. Two of its small molecules that act on proteins are currently in clinical trials.
“Additionally, we are currently developing a new machine learning approach that integrates properties at the quantum level of small molecules that are potential RNA binders. The results obtained from this model are tested against high-resolution experimental structural data of disclosed and unpublished RNA-ligand complexes. Once fully validated, this new method will be employed at the beginning of our RNA-targeting drug design pipeline, to speed up the virtual screening process of finding hit molecules,” said Szumowski.
The development of an RNA platform requires advanced knowledge in various fields, including RNA biology, genetics, bioinformatics, immunology, chemistry, chemoinformatics, and drug design.
The company said molecular modeling techniques and machine learning methodology as well as access to proprietary software and experimentally resolved at high-resolution 3D structures of RNA targets, preferably with ligands, are crucial for an RNA-targeting computer-aided drug design.
“The success at the hit-to-lead stage also depends on synthetic and medicinal chemists who are able to integrate all aspects of lead parameter optimization right from the start. They must be supported by dedicated databases, and proprietary software but most of all by a direct collaboration with RNA biologists