Discovery of potent inhibitors of α-synuclein aggregation using structure-based iterative learning.

Machine learning methods hold the promise to reduce the costs and the failure rates of conventional drug discovery pipelines. This issue is especially pressing for neurodegenerative diseases, where the development of disease-modifying drugs has been particularly challenging. To address this problem, we describe here a machine learning approach to identify small molecule inhibitors of α-synuclein aggregation, a process implicated in Parkinson's disease and other synucleinopathies. Because the proliferation of α-synuclein aggregates takes place through autocatalytic secondary nucleation, we aim to identify compounds that bind the catalytic sites on the surface of the aggregates. To achieve this goal, we use structure-based machine learning in an iterative manner to first identify and then progressively optimize secondary nucleation inhibitors. Our results demonstrate that this approach leads to the facile identification of compounds two orders of magnitude more potent than previously reported ones.

A Relationship between the Structures and Neurotoxic Effects of Aß Oligomers Stabilized by Different Metal Ions.

Oligomeric assemblies of the amyloid ß peptide (Aß) have been investigated for over two decades as possible neurotoxic agents in Alzheimer's disease. However, due to their heterogeneous and transient nature, it is not yet fully established which of the structural features of these oligomers may generate cellular damage. Here, we study distinct oligomer species formed by Aß40 (the 40-residue form of Aß) in the presence of four different metal ions (Al3+, Cu2+, Fe2+, and Zn2+) and show that they differ in their structure and toxicity in human neuroblastoma cells. We then describe a correlation between the size of the oligomers and their neurotoxic activity, which provides a type of structure-toxicity relationship for these Aß40 oligomer species. These results provide insight into the possible role of metal ions in Alzheimer's disease by the stabilization of Aß oligomers.

Using Generative Modeling to Endow with Potency Initially Inert Compounds with Good Bioavailability and Low Toxicity.

In the early stages of drug development, large chemical libraries are typically screened to identify compounds of promising potency against the chosen targets. Often, however, the resulting hit compounds tend to have poor drug metabolism and pharmacokinetics (DMPK), with negative developability features that may be difficult to eliminate. Therefore, starting the drug discovery process with a "null library", compounds that have highly desirable DMPK properties but no potency against the chosen targets, could be advantageous. Here, we explore the opportunities offered by machine learning to realize this strategy in the case of the inhibition of α-synuclein aggregation, a process associated with Parkinson's disease. We apply MolDQN, a generative machine learning method, to build an inhibitory activity against α-synuclein aggregation into an initial inactive compound with good DMPK properties. Our results illustrate how generative modeling can be used to endow initially inert compounds with desirable developability properties.

A pulsar in a binary with a compact object in the mass gap between neutron stars and black holes.

Some compact objects observed in gravitational wave events have masses in the gap between known neutron stars (NSs) and black holes (BHs). The nature of these mass gap objects is unknown, as is the formation of their host binary systems. We report pulsar timing observations made with the Karoo Array Telescope (MeerKAT) of PSR J0514-4002E, an eccentric binary millisecond pulsar in the globular cluster NGC 1851. We found a total binary mass of 3.887 ± 0.004 solar masses (M⊙), and multiwavelength observations show that the pulsar's binary companion is also a compact object. The companion's mass (2.09 to 2.71 M⊙, 95% confidence interval) is in the mass gap, indicating either a very massive NS or a low-mass BH. We propose that the companion formed in a merger between two earlier NSs.

Exploration and Exploitation Approaches Based on Generative Machine Learning to Identify Potent Small Molecule Inhibitors of α-Synuclein Secondary Nucleation.

The high attrition rate in drug discovery pipelines is an especially pressing issue for Parkinson's disease, for which no disease-modifying drugs have yet been approved. Numerous clinical trials targeting α-synuclein aggregation have failed, at least in part due to the challenges in identifying potent compounds in preclinical investigations. To address this problem, we present a machine learning approach that combines generative modeling and reinforcement learning to identify small molecules that perturb the kinetics of aggregation in a manner that reduces the production of oligomeric species. Training data were obtained by an assay reporting on the degree of inhibition of secondary nucleation, which is the most important mechanism of α-synuclein oligomer production. This approach resulted in the identification of small molecules with high potency against secondary nucleation.

Structure-Based Discovery of Small-Molecule Inhibitors of the Autocatalytic Proliferation of α-Synuclein Aggregates.

The presence of amyloid fibrils of α-synuclein is closely associated with Parkinson's disease and related synucleinopathies. It is still very challenging, however, to systematically discover small molecules that prevent the formation of these aberrant aggregates. Here, we describe a structure-based approach to identify small molecules that specifically inhibit the surface-catalyzed secondary nucleation step in the aggregation of α-synuclein by binding to the surface of the amyloid fibrils. The resulting small molecules are screened using a range of kinetic and thermodynamic assays for their ability to bind α-synuclein fibrils and prevent the further generation of α-synuclein oligomers. This study demonstrates that the combination of structure-based and kinetic-based drug discovery methods can lead to the identification of small molecules that selectively inhibit the autocatalytic proliferation of α-synuclein aggregates.

Polarized x-rays from a magnetar.

Magnetars are neutron stars with ultrastrong magnetic fields, which can be observed in x-rays. Polarization measurements could provide information on their magnetic fields and surface properties. We observed polarized x-rays from the magnetar 4U 0142+61 using the Imaging X-ray Polarimetry Explorer and found a linear polarization degree of 13.5 ± 0.8% averaged over the 2- to 8-kilo-electron volt band. The polarization changes with energy: The degree is 15.0 ± 1.0% at 2 to 4 kilo-electron volts, drops below the instrumental sensitivity ~4 to 5 kilo-electron volts, and rises to 35.2 ± 7.1% at 5.5 to 8 kilo-electron volts. The polarization angle also changes by 90° at ~4 to 5 kilo-electron volts. These results are consistent with a model in which thermal radiation from the magnetar surface is reprocessed by scattering off charged particles in the magnetosphere.

A tau homeostasis signature is linked with the cellular and regional vulnerability of excitatory neurons to tau pathology.

Excitatory neurons are preferentially impaired in early Alzheimer's disease but the pathways contributing to their relative vulnerability remain largely unknown. Here we report that pathological tau accumulation takes place predominantly in excitatory neurons compared to inhibitory neurons, not only in the entorhinal cortex, a brain region affected in early Alzheimer's disease, but also in areas affected later by the disease. By analyzing RNA transcripts from single-nucleus RNA datasets, we identified a specific tau homeostasis signature of genes differentially expressed in excitatory compared to inhibitory neurons. One of the genes, BCL2-associated athanogene 3 (BAG3), a facilitator of autophagy, was identified as a hub, or master regulator, gene. We verified that reducing BAG3 levels in primary neurons exacerbated pathological tau accumulation, whereas BAG3 overexpression attenuated it. These results define a tau homeostasis signature that underlies the cellular and regional vulnerability of excitatory neurons to tau pathology.

A method for partitioning the information contained in a protein sequence between its structure and function.

Proteins employ the information stored in the genetic code and translated into their sequences to carry out well-defined functions in the cellular environment. The possibility to encode for such functions is controlled by the balance between the amount of information supplied by the sequence and that left after that the protein has folded into its structure. We study the amount of information necessary to specify the protein structure, providing an estimate that keeps into account the thermodynamic properties of protein folding. We thus show that the information remaining in the protein sequence after encoding for its structure (the 'information gap') is very close to what needed to encode for its function and interactions. Then, by predicting the information gap directly from the protein sequence, we show that it may be possible to use these insights from information theory to discriminate between ordered and disordered proteins, to identify unknown functions, and to optimize artificially-designed protein sequences.

Relativistic spin precession in the double pulsar.

The double pulsar PSR J0737-3039A/B consists of two neutron stars in a highly relativistic orbit that displays a roughly 30-second eclipse when pulsar A passes behind pulsar B. Describing this eclipse of pulsar A as due to absorption occurring in the magnetosphere of pulsar B, we successfully used a simple geometric model to characterize the observed changing eclipse morphology and to measure the relativistic precession of pulsar B's spin axis around the total orbital angular momentum. This provides a test of general relativity and alternative theories of gravity in the strong-field regime. Our measured relativistic spin precession rate of 4.77 degrees (-0 degrees .65)(+0 degrees .66) per year (68% confidence level) is consistent with that predicted by general relativity within an uncertainty of 13%.