Enhancing Drug Discovery and Development through the Integration of Medicinal Chemistry, Chemical Biology, and Academia-Industry Partnerships: Insights from Roche's Endocannabinoid System Projects.

The endocannabinoid system (ECS) is a critical regulatory network composed of endogenous cannabinoids (eCBs), their synthesizing and degrading enzymes, and associated receptors. It is integral to maintaining homeostasis and orchestrating key functions within the central nervous and immune systems. Given its therapeutic significance, we have launched a series of drug discovery endeavors aimed at ECS targets, including peroxisome proliferator-activated receptors (PPARs), cannabinoid receptors types 1 (CB1R) and 2 (CB2R), and monoacylglycerol lipase (MAGL), addressing a wide array of medical needs. The pursuit of new therapeutic agents has been enhanced by the creation of specialized labeled chemical probes, which aid in target localization, mechanistic studies, assay development, and the establishment of biomarkers for target engagement. By fusing medicinal chemistry with chemical biology in a comprehensive, translational end-to-end drug discovery strategy, we have expedited the development of novel therapeutics. Additionally, this strategy promises to foster highly productive partnerships between industry and academia, as will be illustrated through various examples.

Geometric deep learning-guided Suzuki reaction conditions assessment for applications in medicinal chemistry.

Suzuki cross-coupling reactions are considered a valuable tool for constructing carbon-carbon bonds in small molecule drug discovery. However, the synthesis of chemical matter often represents a time-consuming and labour-intensive bottleneck. We demonstrate how machine learning methods trained on high-throughput experimentation (HTE) data can be leveraged to enable fast reaction condition selection for novel coupling partners. We show that the trained models support chemists in determining suitable catalyst-solvent-base combinations for individual transformations including an evaluation of the need for HTE screening. We introduce an algorithm for designing 96-well plates optimized towards reaction yields and discuss the model performance of zero- and few-shot machine learning. The best-performing machine learning model achieved a three-category classification accuracy of 76.3% (±0.2%) and an F 1-score for a binary classification of 79.1% (±0.9%). Validation on eight reactions revealed a receiver operating characteristic (ROC) curve (AUC) value of 0.82 (±0.07) for few-shot machine learning. On the other hand, zero-shot machine learning models achieved a mean ROC-AUC value of 0.63 (±0.16). This study positively advocates the application of few-shot machine learning-guided reaction condition selection for HTE campaigns in medicinal chemistry and highlights practical applications as well as challenges associated with zero-shot machine learning.

RG7774 (Vicasinabin), an orally bioavailable cannabinoid receptor 2 (CB2R) agonist, decreases retinal vascular permeability, leukocyte adhesion, and ocular inflammation in animal models.

Introduction: Preclinical studies suggest that cannabinoid receptor type 2 (CB2R) activation has a therapeutic effect in animal models on chronic inflammation and vascular permeability, which are key pathological features of diabetic retinopathy (DR). A novel CB2R agonist, triazolopyrimidine RG7774, was generated through lead optimization of a high-throughput screening hit. The aim of this study was to characterize the pharmacology, absorption, distribution, metabolism, elimination, and toxicity (ADMET) profile of RG7774, and to explore its potential for managing the key pathological features associated with retinal disease in rodents. Methods: The in vitro pharmacology of RG7774 was investigated for CB2R binding and receptor activation using recombinant human and mouse CB2R expression in Chinese hamster ovary cells, and endogenous CB2R expression in human Jurkat cells, and rat and mouse spleen cells. The ADMET profile was evaluated and the effects of RG7774 on retinal permeability, leukocyte adhesion, and choroidal neovascularization (CNV) were investigated in rodent models of retinal disease. Pharmacokinetic (PK) parameters and the exposure-response relationship were characterized in healthy animals and in animals with laser-induced CNV. Results: RG7774 was found to be a potent (EC50: 2.8 nM and Ki: 51.3 nM), selective, and full CB2R agonist with no signs of cannabinoid receptor type 1 (CB1R) binding or activation. The ligand showed a favorable ADMET profile and exhibited systemic and ocular exposure after oral delivery. Functional potency in vitro translated from recombinant to endogenous expression systems. In vivo, orally administered RG7774 reduced retinal permeability and leukocyte adhesion in rodents with lipopolysaccharide (LPS)-induced uveitis and streptozotocin (STZ)-induced DR, and reduced lesion areas in rats with laser-induced CNV with an ED50 of 0.32 mg/kg. Anatomically, RG7774 reduced the migration of retinal microglia to retinal lesions. Discussion: RG7774 is a novel, highly selective, and orally bioavailable CB2R agonist, with an acceptable systemic and ocular PK profile, and beneficial effects on retinal vascular permeability, leukocyte adhesion, and ocular inflammation in rodent animal models. Results support the development of RG7774 as a potential treatment for retinal diseases with similar pathophysiologies as addressed by the animal models.

sp3 -Rich Heterocycle Synthesis on DNA: Application to DNA-Encoded Library Production.

DNA encoded library (DEL) synthesis represents a convenient means to produce, annotate and store large collections of compounds in a small volume. While DELs are well suited for drug discovery campaigns, the chemistry used in their production must be compatible with the DNA tag, which can limit compound class accessibility. As a result, most DELs are heavily populated with peptidomimetic and sp2 -rich molecules. Herein, we show that sp3 -rich mono- and bicyclic heterocycles can be made on DNA from ketochlorohydrin aldol products through a reductive amination and cyclization process. The resulting hydroxypyrrolidines possess structural features that are desirable for DELs and target a distinct region of pharmaceutically relevant chemical space.

Comparing auditory distance perception in real and virtual environments and the role of the loudness cue: A study based on event-related potentials.

The perception of the distance to a sound source is relevant in many everyday situations, not only in real spaces, but also in virtual reality (VR) environments. Where real rooms often reach their limits, VR offers far-reaching possibilities to simulate a wide range of acoustic scenarios. However, in virtual room acoustics a plausible reproduction of distance-related cues can be challenging. In the present study, we compared the detection of changes of the distance to a sound source and its neurocognitive correlates in a real and a virtual reverberant environment, using an active auditory oddball paradigm and EEG measures. The main goal was to test whether the experiments in the virtual and real environments produced equivalent behavioral and EEG results. Three loudspeakers were placed at ego-centric distances of 2 m (near), 4 m (center), and 8 m (far) in front of the participants (N = 20), each 66 cm below their ear level. Sequences of 500 ms noise stimuli were presented either from the center position (standards, 80 % of trials) or from the near or far position (targets, 10 % each). The participants had to indicate a target position via a joystick response ("near" or "far"). Sounds were emitted either by real loudspeakers in the real environment or rendered and played back for the corresponding positions via headphones in the virtual environment. In addition, within both environments, loudness of the auditory stimuli was either unaltered (natural loudness) or the loudness cue was manipulated, so that all three loudspeakers were perceived equally loud at the listener's position (matched loudness). The EEG analysis focused on the mismatch negativity (MMN), P3a, and P3b as correlates of deviance detection, attentional orientation, and context-updating/stimulus evaluation, respectively. Overall, behavioral data showed that detection of the target positions was reduced within the virtual environment, and especially when loudness was matched. Except for slight latency shifts in the virtual environment, EEG analysis indicated comparable patterns within both environments and independent of loudness settings. Thus, while the neurocognitive processing of changes in distance appears to be similar in virtual and real spaces, a proper representation of loudness appears to be crucial to achieve a good task performance in virtual acoustic environments.

[Influence of overtones and undertones on melody recognition with a cochlear implant with SSD]. / Einfluss von Ober- und Untertönen auf die Melodieerkennung mit einem Cochlea-Implantat bei SSD.

Many cochlear implant (CI) users have difficulties recognising pitches and melodies because pitch transmission is blurred and shifted. This study investigates whether postlingually deafened adult CI users recognize melodies better when overtones are removed or undertones are added.Fifteen unilaterally postlingually deafened CI users (single sided deafness = SSD) were included aged 22 to 73 years (MW 52, SD 11.6) with CI hearing experience between 3 and 75 months (MW 33, SD 21.0) with varying MED-EL devices. Three short piano melodies were presented to them firstly to the normal-hearing ear and then in modified overtone or undertone variants and the original variant to the CI ear. These variants should be identified as one of the three original melodies. In addition, musical experience and ability were assessed by the Munich Music Questionnaire and the MiniPROMS music tests.The CI users showed the best melody recognition in the fundamental frequency variant. The overtone variant with the third overtone was as good as the original variant with all overtones with regard to melody recognition (p=1). However, the undertone variant with the first undertone was recognised significantly worse than the fundamental version (p=0.032). Furthermore, there was no correlation between musical experience or musical ability and the number of melodies recognised (p>0.1).Since a reduction of overtones did not worsen the melody recognition, overtone reduction should be considered in future music processing programs for the CI. This could reduce the energy consumption of the CI.

Enabling late-stage drug diversification by high-throughput experimentation with geometric deep learning.

Late-stage functionalization is an economical approach to optimize the properties of drug candidates. However, the chemical complexity of drug molecules often makes late-stage diversification challenging. To address this problem, a late-stage functionalization platform based on geometric deep learning and high-throughput reaction screening was developed. Considering borylation as a critical step in late-stage functionalization, the computational model predicted reaction yields for diverse reaction conditions with a mean absolute error margin of 4-5%, while the reactivity of novel reactions with known and unknown substrates was classified with a balanced accuracy of 92% and 67%, respectively. The regioselectivity of the major products was accurately captured with a classifier F-score of 67%. When applied to 23 diverse commercial drug molecules, the platform successfully identified numerous opportunities for structural diversification. The influence of steric and electronic information on model performance was quantified, and a comprehensive simple user-friendly reaction format was introduced that proved to be a key enabler for seamlessly integrating deep learning and high-throughput experimentation for late-stage functionalization.

Investigations into mRNA Lipid Nanoparticles Shelf-Life Stability under Nonfrozen Conditions.

mRNA LNPs can experience a decline in activity over short periods (ranging from weeks to months). As a result, they require frozen storage and transportation conditions to maintain their full functionality when utilized. Currently approved commercially available mRNA LNP vaccines also necessitate frozen storage and supply chain management. Overcoming this significant inconvenience in the future is crucial to reducing unnecessary costs and challenges associated with storage and transport. In this study, our objective was to illuminate the potential time frame for nonfrozen storage and transportation conditions of mRNA LNPs without compromising their activity. To achieve this goal, we conducted a stability assessment and an in vitro cell culture delivery study involving five mRNA LNPs. These LNPs were constructed by using a standard formulation similar to that employed in the three commercially available LNP formulations. Among these formulations, we selected five structurally diverse ionizable lipids─C12-200, CKK-E12, MC3, SM-102, and lipid 23─from the existing literature. We incorporated these lipids into a standard LNP formulation, keeping all other components identical. The LNPs, carrying mRNA payloads, were synthesized by using microfluidic mixing technology. We evaluated the shelf life stability of these LNPs over a span of 9 weeks at temperatures of 2-8, 25, and 40 °C, utilizing an array of analytical techniques. Our findings indicated minimal impact on the hydrodynamic diameter, zeta potential, encapsulation efficiency, and polydispersity of all LNPs across the various temperatures over the studied period. The RiboGreen assay analysis of LNPs showed consistent mRNA contents over several weeks at various nonfrozen storage temperatures, leading to the incorrect assumption of intact and functional LNPs. This misunderstanding was rectified by the significant differences observed in EGFP protein expression in an in vitro cell culture (using HEK293 cells) across the five LNPs. Specifically, only LNP 1 (C12-200) and LNP 4 (SM-102) exhibited high levels of EGFP expression at the start (T0), with over 90% of HEK293 cells transfected and mean fluorescence intensity (MFI) levels exceeding 1. Interestingly, LNP 1 (C12-200) maintained largely unchanged levels of in vitro activity over 11 weeks when stored at both 2-8 and 25 °C. In contrast, LNP 4 (SM-102) retained its functionality when stored at 2-8 °C over 11 weeks but experienced a gradual decline of in vitro activity when stored at room temperature over the same period. Importantly, we observed distinct LNP architectures for the five formulations through cryo-EM imaging. This highlights the necessity for a deeper comprehension of structure-activity relationships within these complex nanoparticle structures. Enhancing our understanding in this regard is vital for overcoming storage and stability limitations, ultimately facilitating the broader application of this technology beyond vaccines.

Identifying opportunities for late-stage C-H alkylation with high-throughput experimentation and in silico reaction screening.

Enhancing the properties of advanced drug candidates is aided by the direct incorporation of specific chemical groups, avoiding the need to construct the entire compound from the ground up. Nevertheless, their chemical intricacy often poses challenges in predicting reactivity for C-H activation reactions and planning their synthesis. We adopted a reaction screening approach that combines high-throughput experimentation (HTE) at a nanomolar scale with computational graph neural networks (GNNs). This approach aims to identify suitable substrates for late-stage C-H alkylation using Minisci-type chemistry. GNNs were trained using experimentally generated reactions derived from in-house HTE and literature data. These trained models were then used to predict, in a forward-looking manner, the coupling of 3180 advanced heterocyclic building blocks with a diverse set of sp3-rich carboxylic acids. This predictive approach aimed to explore the substrate landscape for Minisci-type alkylations. Promising candidates were chosen, their production was scaled up, and they were subsequently isolated and characterized. This process led to the creation of 30 novel, functionally modified molecules that hold potential for further refinement. These results positively advocate the application of HTE-based machine learning to virtual reaction screening.

Diligent Design Enables Antibody-ASO Conjugates with Optimal Pharmacokinetic Properties.

Antisense-oligonucleotides (ASOs) are a promising drug modality for the treatment of neurological disorders, but the currently established route of administration via intrathecal delivery is a major limitation to its broader clinical application. An attractive alternative is the conjugation of the ASO to an antibody that facilitates access to the central nervous system (CNS) after peripheral application and target engagement at the blood-brain barrier, followed by transcytosis. Here, we show that the diligent conjugate design of Brainshuttle-ASO conjugates is the key to generating promising delivery vehicles and thereby establishing design principles to create optimized molecules with drug-like properties. An innovative site-specific transglutaminase-based conjugation technology was chosen and optimized in a stepwise process to identify the best-suited conjugation site, tags, reaction conditions, and linker design. The overall conjugation performance was found to be specifically governed by the choice of buffer conditions and the structure of the linker. The combination of the peptide tags YRYRQ and RYESK was chosen, showing high conjugation fidelity. Elaborate conjugate analysis revealed that one leading differentiating factor was hydrophobicity. The increase of hydrophobicity by the ASO payload could be mitigated by the appropriate choice of conjugation site and the heavy chain position 297 proved to be the most optimal. Evaluating the properties of the linker suggested a short bicyclo[6.1.0]nonyne (BCN) unit as best suited with regards to conjugation performance and potency. Promising in vitro activity and in vivo pharmacokinetic behavior of optimized Brainshuttle-ASO conjugates, based on a microtubule-associated protein tau (MAPT) targeting oligonucleotide, suggest that such designs have the potential to serve as a blueprint for peripherally delivered ASO-based drugs for the CNS in the future.