Molecular Design Method Using a Reversible Tree Representation of Chemical Compounds and Deep Reinforcement Learning.

Automatic design of molecules with specific chemical and biochemical properties is an important process in material informatics and computational drug discovery. In this study, we designed a novel coarse-grained tree representation of molecules (Reversible Junction Tree; "RJT") for the aforementioned purposes, which is reversely convertible to the original molecule without external information. By leveraging this representation, we further formulated the molecular design and optimization problem as a tree-structure construction using deep reinforcement learning ("RJT-RL"). In this method, all of the intermediate and final states of reinforcement learning are convertible to valid molecules, which could efficiently guide the optimization process in simple benchmark tasks. We further examined the multiobjective optimization and fine-tuning of the reinforcement learning models using RJT-RL, demonstrating the applicability of our method to more realistic tasks in drug discovery.

Towards universal neural network potential for material discovery applicable to arbitrary combination of 45 elements.

Computational material discovery is under intense study owing to its ability to explore the vast space of chemical systems. Neural network potentials (NNPs) have been shown to be particularly effective in conducting atomistic simulations for such purposes. However, existing NNPs are generally designed for narrow target materials, making them unsuitable for broader applications in material discovery. Here we report a development of universal NNP called PreFerred Potential (PFP), which is able to handle any combination of 45 elements. Particular emphasis is placed on the datasets, which include a diverse set of virtual structures used to attain the universality. We demonstrated the applicability of PFP in selected domains: lithium diffusion in LiFeSO4F, molecular adsorption in metal-organic frameworks, an order-disorder transition of Cu-Au alloys, and material discovery for a Fischer-Tropsch catalyst. They showcase the power of PFP, and this technology provides a highly useful tool for material discovery.

Lateral access mechanism of LPA receptor probed by molecular dynamics simulation.

G-protein-coupled receptors (GPCR) are a family of membrane receptors that play important roles in the regulation of various physiological phenomena. LPA receptors (LPA1-6) are members of the class A GPCRs, which transduce a lysophosphatidic acid (LPA) signal across the cell membrane and evoke various responses, including cellular survival, proliferation, differentiation, and migration. The crystal structure of LPA6 revealed a gap between its transmembrane helices (TMs), which is opened toward the membrane side. This led to the proposal of the "lateral access model," in which its lipophilic ligand directly enters the binding pocket through the gap structure at the membrane. In this study, we performed molecular dynamics (MD) simulations and Markov state model (MSM) analyses of LPA6 and LPA, to elucidate the long timescale dynamics of the ligand binding process. The results from the 71.4-µs MD simulation suggested that the flexibility of the TMs constituting the gap structure enables the lateral entrance of the ligand, and the key interactions between the receptor and ligand facilitate the transition state of the ligand binding process.

Time-resolved serial femtosecond crystallography reveals early structural changes in channelrhodopsin.

Channelrhodopsins (ChRs) are microbial light-gated ion channels utilized in optogenetics to control neural activity with light . Light absorption causes retinal chromophore isomerization and subsequent protein conformational changes visualized as optically distinguished intermediates, coupled with channel opening and closing. However, the detailed molecular events underlying channel gating remain unknown. We performed time-resolved serial femtosecond crystallographic analyses of ChR by using an X-ray free electron laser, which revealed conformational changes following photoactivation. The isomerized retinal adopts a twisted conformation and shifts toward the putative internal proton donor residues, consequently inducing an outward shift of TM3, as well as a local deformation in TM7. These early conformational changes in the pore-forming helices should be the triggers that lead to opening of the ion conducting pore.

Cryo-EM structure of the volume-regulated anion channel LRRC8D isoform identifies features important for substrate permeation.

Members of the leucine-rich repeat-containing 8 (LRRC8) protein family, composed of the five LRRC8A-E isoforms, are pore-forming components of the volume-regulated anion channel (VRAC). LRRC8A and at least one of the other LRRC8 isoforms assemble into heteromers to generate VRAC transport activities. Despite the availability of the LRRC8A structures, the structural basis of how LRRC8 isoforms other than LRRC8A contribute to the functional diversity of VRAC has remained elusive. Here, we present the structure of the human LRRC8D isoform, which enables the permeation of organic substrates through VRAC. The LRRC8D homo-hexamer structure displays a two-fold symmetric arrangement, and together with a structure-based electrophysiological analysis, revealed two key features. The pore constriction on the extracellular side is wider than that in the LRRC8A structures, which may explain the increased permeability of organic substrates. Furthermore, an N-terminal helix protrudes into the pore from the intracellular side and may be critical for gating.

Identification of Potent In Vivo Autotaxin Inhibitors that Bind to Both Hydrophobic Pockets and Channels in the Catalytic Domain.

Autotaxin (ATX, also known as ENPP2) is a predominant lysophosphatidic acid (LPA)-producing enzyme in the body, and LPA regulates various physiological functions, such as angiogenesis and wound healing, as well as pathological functions, including proliferation, metastasis, and fibrosis, via specific LPA receptors. Therefore, the ATX-LPA axis is a promising therapeutic target for dozens of diseases, including cancers, pulmonary and liver fibroses, and neuropathic pain. Previous structural studies revealed that the catalytic domain of ATX has a hydrophobic pocket and a hydrophobic channel; these serve to recognize the substrate, lysophosphatidylcholine (LPC), and deliver generated LPA to LPA receptors on the plasma membrane. Most reported ATX inhibitors bind to either the hydrophobic pocket or the hydrophobic channel. Herein, we present a unique ATX inhibitor that binds mainly to the hydrophobic pocket and also partly to the hydrophobic channel, inhibiting ATX activity with high potency and selectivity in vitro and in vivo. Notably, our inhibitor can rescue the cardia bifida (two hearts) phenotype in ATX-overexpressing zebrafish embryos.

Crystal structure of Drosophila Piwi.

PIWI-clade Argonaute proteins associate with PIWI-interacting RNAs (piRNAs), and silence transposons in animal gonads. Here, we report the crystal structure of the Drosophila PIWI-clade Argonaute Piwi in complex with endogenous piRNAs, at 2.9 Å resolution. A structural comparison of Piwi with other Argonautes highlights the PIWI-specific structural features, such as the overall domain arrangement and metal-dependent piRNA recognition. Our structural and biochemical data reveal that, unlike other Argonautes including silkworm Siwi, Piwi has a non-canonical DVDK tetrad and lacks the RNA-guided RNA cleaving slicer activity. Furthermore, we find that the Piwi mutant with the canonical DEDH catalytic tetrad exhibits the slicer activity and readily dissociates from less complementary RNA targets after the slicer-mediated cleavage, suggesting that the slicer activity could compromise the Piwi-mediated co-transcriptional silencing. We thus propose that Piwi lost the slicer activity during evolution to serve as an RNA-guided RNA-binding platform, thereby ensuring faithful co-transcriptional silencing of transposons.

Author Correction: Structural basis for the drug extrusion mechanism by a MATE multidrug transporter.

An Amendment to this paper has been published and can be accessed via a link at the top of the paper.

Structural biology of the multidrug and toxic compound extrusion superfamily transporters.

Xenobiotic and metabolite extrusion is an important process for the proper functions of cells and their compartments, including acidic organelles. MATE (multidrug and toxic compound extrusion) is a large family of secondary active transporters involved in the transport of various compounds across cellular and organellar membranes, and is present in the three domains of life. The major substrates of the bacterial MATE transporters are cationic compounds, including clinically important antibiotics, and thereby MATE transporters confer multi-drug resistance to pathogenic bacteria. The plant MATE transporters are important for the accumulation of various metabolites in organelles, including vacuoles. The human MATE transporters are expressed in the brush-border membrane of the kidney, and are involved in the clearance of cationic drugs from the body. During the past decade, progress in structural biology has clarified the transport mechanism of these MATE transporters in atomic detail. The present review summarizes the reported structures of MATE family transporters, along with their structure-guided functional analyses. This integrated view of the structures of MATE transporters provides novel insights into their transport mechanism.

Cryo-EM structure of the human L-type amino acid transporter 1 in complex with glycoprotein CD98hc.

The L-type amino acid transporter 1 (LAT1 or SLC7A5) transports large neutral amino acids across the membrane and is crucial for brain drug delivery and tumor growth. LAT1 forms a disulfide-linked heterodimer with CD98 heavy chain (CD98hc, 4F2hc or SLC3A2), but the mechanism of assembly and amino acid transport are poorly understood. Here we report the cryo-EM structure of the human LAT1-CD98hc heterodimer at 3.3-Å resolution. LAT1 features a canonical Leu T-fold and exhibits an unusual loop structure on transmembrane helix 6, creating an extended cavity that might accommodate bulky amino acids and drugs. CD98hc engages with LAT1 through the extracellular, transmembrane and putative cholesterol-mediated interactions. We also show that two anti-CD98 antibodies recognize distinct, multiple epitopes on CD98hc but not its glycans, explaining their robust reactivities. These results reveal the principles of glycoprotein-solute carrier assembly and provide templates for improving preclinical drugs and antibodies targeting LAT1 or CD98hc.

Structural basis for oligomerization of the prokaryotic peptide transporter PepTSo2.

Proton-dependent oligopeptide transporters (POTs) belong to the major facilitator superfamily (MFS) and transport dipeptides and tripeptides from the extracellular environment into the target cell. The human POTs PepT1 and PepT2 are also involved in the absorption of various orally ingested drugs. Previously reported structures revealed that the bacterial POTs possess 14 helices, of which H1-H6 and H7-H12 constitute the typical MFS fold and the residual two helices are involved in the cytoplasmic linker. PepTSo2 from Shewanella oneidensis is a unique POT which reportedly assembles as a 200 kDa tetramer. Although the previously reported structures suggested the importance of H12 for tetramer formation, the structural basis for the PepTSo2-specific oligomerization remains unclear owing to the lack of a high-resolution tetrameric structure. In this study, the expression and purification conditions for tetrameric PepTSo2 were optimized. A single-particle cryo-EM analysis revealed the tetrameric structure of PepTSo2 incorporated into Salipro nanoparticles at 4.1 Šresolution. Furthermore, a combination of lipidic cubic phase (LCP) crystallization and an automated data-processing system for multiple microcrystals enabled crystal structures of PepTSo2 to be determined at resolutions of 3.5 and 3.9 Å. The present structures in a lipid bilayer revealed the detailed mechanism for the tetrameric assembly of PepTSo2, in which a characteristic extracellular loop (ECL) interacts with two asparagine residues on H12 which were reported to be important for tetramerization and plays an essential role in oligomeric assembly. This study provides valuable insights into the oligomerization mechanism of this MFS-type transporter, which will further pave the way for understanding other oligomeric membrane proteins.

Structural basis for the promiscuous PAM recognition by Corynebacterium diphtheriae Cas9.

The RNA-guided DNA endonuclease Cas9 cleaves double-stranded DNA targets bearing a protospacer adjacent motif (PAM) and complementarity to an RNA guide. Unlike other Cas9 orthologs, Corynebacterium diphtheriae Cas9 (CdCas9) recognizes the promiscuous NNRHHHY PAM. However, the CdCas9-mediated PAM recognition mechanism remains unknown. Here, we report the crystal structure of CdCas9 in complex with the guide RNA and its target DNA at 2.9 Å resolution. The structure reveals that CdCas9 recognizes the NNRHHHY PAM via a combination of van der Waals interactions and base-specific hydrogen bonds. Moreover, we find that CdCas9 exhibits robust DNA cleavage activity with the optimal 22-nucleotide length guide RNAs. Our findings highlight the mechanistic diversity of the PAM recognition by Cas9 orthologs, and provide a basis for the further engineering of the CRISPR-Cas9 genome-editor nucleases.

Crystal structure of plant vacuolar iron transporter VIT1.

The iron ion is an essential cofactor in several vital enzymatic reactions, such as DNA replication, oxygen transport, and respiratory and photosynthetic electron transfer chains, but its excess accumulation induces oxidative stress in cells. Vacuolar iron transporter 1 (VIT1) is important for iron homeostasis in plants, by transporting cytoplasmic ferrous ions into vacuoles. Modification of the VIT1 gene leads to increased iron content in crops, which could be used for the treatment of human iron deficiency diseases. Furthermore, a VIT1 from the malaria-causing parasite Plasmodium is considered as a potential drug target for malaria. Here we report the crystal structure of VIT1 from rose gum Eucalyptus grandis, which probably functions as a H+-dependent antiporter for Fe2+ and other transition metal ions. VIT1 adopts a novel protein fold forming a dimer of five membrane-spanning domains, with an ion-translocating pathway constituted by the conserved methionine and carboxylate residues at the dimer interface. The second transmembrane helix protrudes from the lipid membrane by about 40 Å and connects to a three-helical bundle, triangular cytoplasmic domain, which binds to the substrate metal ions and stabilizes their soluble form, thus playing an essential role in their transport. These mechanistic insights will provide useful information for the further design of genetically modified crops and the development of anti-malaria drugs.

Cap-specific terminal N 6-methylation of RNA by an RNA polymerase II-associated methyltransferase.

N 6-methyladenosine (m6A), a major modification of messenger RNAs (mRNAs), plays critical roles in RNA metabolism and function. In addition to the internal m6A, N 6, 2'-O-dimethyladenosine (m6Am) is present at the transcription start nucleotide of capped mRNAs in vertebrates. However, its biogenesis and functional role remain elusive. Using a reverse genetics approach, we identified PCIF1, a factor that interacts with the serine-5-phosphorylated carboxyl-terminal domain of RNA polymerase II, as a cap-specific adenosine methyltransferase (CAPAM) responsible for N 6-methylation of m6Am. The crystal structure of CAPAM in complex with substrates revealed the molecular basis of cap-specific m6A formation. A transcriptome-wide analysis revealed that N 6-methylation of m6Am promotes the translation of capped mRNAs. Thus, a cap-specific m6A writer promotes translation of mRNAs starting from m6Am.

Structural Basis of H+-Dependent Conformational Change in a Bacterial MATE Transporter.

Multidrug and toxic compound extrusion (MATE) transporters efflux toxic compounds using a Na+ or H+ gradient across the membrane. Although the structures of MATE transporters have been reported, the cation-coupled substrate transport mechanism remains controversial. Here we report crystal structures of VcmN, a Vibrio cholerae MATE transporter driven by the H+ gradient. High-resolution structures in two distinct conformations associated with different pHs revealed that the rearrangement of the hydrogen-bonding network around the conserved Asp35 induces the bending of transmembrane helix 1, as in the case of the H+-coupled Pyrococcus furiosus MATE transporter. We also determined the crystal structure of the D35N mutant, which captured a unique conformation of TM1 facilitated by an altered hydrogen-bonding network. Based on the present results, we propose a common step in the transport cycle shared among prokaryotic H+-coupled MATE transporters.

Crystal structure of the Agrobacterium tumefaciens type VI effector-immunity complex.

The type VI secretion system (T6SS) comprises needle-shaped multisubunit complexes that play a role in the microbial defense systems of Gram-negative bacteria. Some Gram-negative bacteria harboring a T6SS deliver toxic effector proteins into the cytoplasm or periplasm of competing bacteria in order to lyse and kill them. To avoid self-cell disruption, these bacteria have cognate immunity proteins that inhibit their toxic effector proteins. T6SS amidase effector protein 4 (Tae4) and T6SS amidase immunity protein 4 (Tai4) are a representative of the toxic effector-immunity pairs of the T6SS. Here, the three-dimensional structures of Tai4 and the Tae4-Tai4 complex from Agrobacterium tumefaciens are reported at 1.55 and 1.9 Šresolution, respectively. A structural comparison with other Tae4-Tai4 homologs revealed similarities and differences in the catalytic and inhibitory mechanisms among the Tae4 and Tai4 family proteins.

Structural insights into cGAMP degradation by Ecto-nucleotide pyrophosphatase phosphodiesterase 1.

ENPP1 (Ecto-nucleotide pyrophosphatase phosphodiesterase 1), a type II transmembrane glycoprotein, hydrolyzes ATP to produce AMP and diphosphate, thereby inhibiting bone mineralization. A recent study showed that ENPP1 also preferentially hydrolyzes 2'3'-cGAMP (cyclic GMP-AMP) but not its linkage isomer 3'3'-cGAMP, and negatively regulates the cGAS-STING pathway in the innate immune system. Here, we present the high-resolution crystal structures of ENPP1 in complex with 3'3'-cGAMP and the reaction intermediate pA(3',5')pG. The structures revealed that the adenine and guanine bases of the dinucleotides are recognized by nucleotide- and guanine-pockets, respectively. Furthermore, the structures indicate that 2'3'-cGAMP, but not 3'3'-cGAMP, binds to the active site in a conformation suitable for catalysis, thereby explaining the specific degradation of 2'3'-cGAMP by ENPP1. Our findings provide insights into how ENPP1 hydrolyzes both ATP and cGAMP to participate in the two distinct biological processes.

Free Energy Landscape for the Entire Transport Cycle of Triose-Phosphate/Phosphate Translocator.

Secondary active transporters translocate their substrates using the electrochemical potentials of other chemicals and undergo large-scale conformational changes. Despite extensive structural studies, the atomic details of the transport mechanism still remain elusive. We performed a series of all-atom molecular dynamics simulations of the triose-phosphate/phosphate translocator (TPT), which exports organic phosphates in the chloroplast stroma in strict counter exchange with inorganic phosphate (Pi). Biased sampling methods, including the string method and umbrella sampling, successfully reproduced the conformational changes between the inward- and outward-facing states, along with the substrate binding. The free energy landscape of this entire TPT transition pathway demonstrated the alternating access and substrate translocation mechanisms, which revealed that Pi is relayed by positively charged residues along the transition pathway. Furthermore, the conserved Glu207 functions as a "molecular switch", linking the local substrate binding and the global conformational transition. Our results provide atomic-detailed insights into the substrate transport mechanism of the antiporter.

Vibrational and Molecular Properties of Mg2+ Binding and Ion Selectivity in the Magnesium Channel MgtE.

Magnesium ions (Mg2+) are crucial for various biological processes. A bacterial Mg2+ channel, MgtE, tightly regulates the intracellular Mg2+ concentration. Previous X-ray crystal structures showed that MgtE forms a dimeric structure composed of a total of 10 transmembrane α helices forming a central pore, and intracellular soluble domains constituting a Mg2+ sensor. The ion selectivity for Mg2+ over Ca2+ resides at a central cavity in the transmembrane pore of MgtE, involving a conserved aspartate residue (Asp432) from each monomer. Here, we applied ion-exchange-induced difference FTIR spectroscopy to analyze the interactions between MgtE and divalent cations, Mg2+ and Ca2+. Using site-directed mutagenesis, vibrational bands at 1421 (Mg2+), 1407 (Mg2+), ∼1440 (Ca2+), and 1390 (Ca2+) cm-1 were assigned to symmetric carboxylate stretching modes of Asp432, involved in the ion coordination. Conservative modifications of the central cavity by Asp432Glu or Ala417Leu mutations resulted in the disappearance of the Mg2+-sensitive carboxylate bands, suggesting a highly optimized geometry for accommodating a Mg2+ ion. The dependency of the vibrational changes on Mg2+ and Ca2+ concentrations revealed the presence of a two different classes of binding sites: a high affinity site for Mg2+ ( Kd ≈ 0.3 mM) with low Ca2+ affinity ( Kd ≈ 80 mM), and a medium affinity site for Mg2+ ( Kd ≈ 2 mM) and Ca2+ ( Kd ≈ 6 mM), tentatively assigned to the central cavity and the sensor domain, respectively. With the aid of molecular dynamics simulation and normal-mode analysis by quantum chemistry, we confirm that changes in carboxylate bands of the high affinity binding site originate from Asp432 in the central cavity.

An Atomistic Model of a Precursor State of Light-Induced Channel Opening of Channelrhodopsin.

Channelrhodopsins (ChRs) are microbial light-gated ion channels with a retinal chromophore and are widely utilized in optogenetics to precisely control neuronal activity with light. Despite increasing understanding of their structures and photoactivation kinetics, the atomistic mechanism of light gating and ion conduction remains elusive. Here, we present an atomic structural model of a chimeric ChR in a precursor state of the channel opening determined by an accurate hybrid molecular simulation technique and a statistical theory of internal water distribution. The photoactivated structure features extensive tilt of the chromophore accompanied by redistribution of water molecules in its binding pocket, which is absent in previously known photoactivated structures of analogous photoreceptors, and widely agrees with structural and spectroscopic experimental evidence of ChRs. The atomistic model manifests a photoactivated ion-conduction pathway that is markedly different from a previously proposed one and successfully explains experimentally observed mutagenic effects on key channel properties.