The catalytic mechanism of the RNA methyltransferase METTL3.

The complex of methyltransferase-like proteins 3 and 14 (METTL3-14) is the major enzyme that deposits N6-methyladenosine (m6A) modifications on messenger RNA (mRNA) in humans. METTL3-14 plays key roles in various biological processes through its methyltransferase (MTase) activity. However, little is known about its substrate recognition and methyl transfer mechanism from its cofactor and methyl donor S-adenosylmethionine (SAM). Here, we study the MTase mechanism of METTL3-14 by a combined experimental and multiscale simulation approach using bisubstrate analogues (BAs), conjugates of a SAM-like moiety connected to the N6-atom of adenosine. Molecular dynamics simulations based on crystal structures of METTL3-14 with BAs suggest that the Y406 side chain of METTL3 is involved in the recruitment of adenosine and release of m6A. A crystal structure with a BA representing the transition state of methyl transfer shows a direct involvement of the METTL3 side chains E481 and K513 in adenosine binding which is supported by mutational analysis. Quantum mechanics/molecular mechanics (QM/MM) free energy calculations indicate that methyl transfer occurs without prior deprotonation of adenosine-N6. Furthermore, the QM/MM calculations provide further support for the role of electrostatic contributions of E481 and K513 to catalysis. The multidisciplinary approach used here sheds light on the (co)substrate binding mechanism, catalytic step, and (co)product release, and suggests that the latter step is rate-limiting for METTL3. The atomistic information on the substrate binding and methyl transfer reaction of METTL3 can be useful for understanding the mechanisms of other RNA MTases and for the design of transition state analogues as their inhibitors.

The catalytic mechanism of the RNA methyltransferase METTL3.

The complex of methyltransferase-like proteins 3 and 14 (METTL3-14) is the major enzyme that deposits N6-methyladenosine (m6A) modifications on mRNA in humans. METTL3-14 plays key roles in various biological processes through its methyltransferase (MTase) activity. However, little is known about its substrate recognition and methyl transfer mechanism from its cofactor and methyl donor S-adenosylmethionine (SAM). Here, we study the MTase mechanism of METTL3-14 by a combined experimental and multiscale simulation approach using bisubstrate analogues (BAs), conjugates of a SAM-like moiety connected to the N6-atom of adenosine. Molecular dynamics simulations based on crystal structures of METTL3-14 with BAs suggest that the Y406 side chain of METTL3 is involved in the recruitment of adenosine and release of m6A. A crystal structure with a bisubstrate analogue representing the transition state of methyl transfer shows a direct involvement of the METTL3 side chains E481 and K513 in adenosine binding which is supported by mutational analysis. Quantum mechanics/molecular mechanics (QM/MM) free energy calculations indicate that methyl transfer occurs without prior deprotonation of adenosine-N6. Furthermore, the QM/MM calculations provide further support for the role of electrostatic contributions of E481 and K513 to catalysis. The multidisciplinary approach used here sheds light on the (co)substrate binding mechanism, catalytic step, and (co)product release catalysed by METTL3, and suggests that the latter step is rate-limiting. The atomistic information on the substrate binding and methyl transfer reaction of METTL3 can be useful for understanding the mechanisms of other RNA MTases and for the design of transition state analogues as their inhibitors.

Structure-Based Design of Inhibitors of the m6A-RNA Writer Enzyme METTL3.

Methyltransferase-like 3 (METTL3) and METTL14 form a heterodimeric complex that catalyzes the most abundant internal mRNA modification, N6-methyladenosine (m6A). METTL3 is the catalytic subunit that binds the co-substrate S-adenosyl methionine (SAM), while METTL14 is involved in mRNA binding. The m6A modification provides post-transcriptional level control over gene expression as it affects almost all stages of the mRNA life cycle, including splicing, nuclear export, translation, and decay. There is increasing evidence for an oncogenic role of METTL3 in acute myeloid leukemia. Here, we use structural and dynamic details of the catalytic subunit METTL3 for developing small-molecule inhibitors that compete with SAM. Starting from a hit identified by high-throughput docking, protein crystallography and molecular dynamics simulations were employed to guide the optimization of inhibitory activity. The potency was successfully improved by 8000-fold as measured by a homogeneous time-resolved fluorescence assay. The optimized compound is selective against the off-targets RNA methyltransferases METTL1 and METTL16.

Rapid ex vivo reverse genetics identifies the essential determinants of prion protein toxicity.

The cellular prion protein PrPC mediates the neurotoxicity of prions and other protein aggregates through poorly understood mechanisms. Antibody-derived ligands against the globular domain of PrPC (GDL) can also initiate neurotoxicity by inducing an intramolecular R208 -H140 hydrogen bond ("H-latch") between the α2-α3 and ß2-α2 loops of PrPC . Importantly, GDL that suppresses the H-latch prolong the life of prion-infected mice, suggesting that GDL toxicity and prion infections exploit convergent pathways. To define the structural underpinnings of these phenomena, we transduced 19 individual PrPC variants to PrPC -deficient cerebellar organotypic cultured slices using adenovirus-associated viral vectors (AAV). We report that GDL toxicity requires a single N-proximal cationic residue (K27 or R27 ) within PrPC . Alanine substitution of K27 also prevented the toxicity of PrPC mutants that induce Shmerling syndrome, a neurodegenerative disease that is suppressed by co-expression of wild-type PrPC . K27 may represent an actionable target for compounds aimed at preventing prion-related neurodegeneration.

1,4,9-Triazaspiro[5.5]undecan-2-one Derivatives as Potent and Selective METTL3 Inhibitors.

N6-methyladenosine (m6A) is the most frequent of the 160 RNA modifications reported so far. Accumulating evidence suggests that the METTL3/METTL14 protein complex, part of the m6A regulation machinery, is a key player in a variety of diseases including several types of cancer, type 2 diabetes, and viral infections. Here we report on a protein crystallography-based medicinal chemistry optimization of a METTL3 hit compound that has resulted in a 1400-fold potency improvement (IC50 of 5 nM for the lead compound 22 (UZH2) in a time-resolved Förster resonance energy transfer (TR-FRET) assay). The series has favorable ADME properties as physicochemical characteristics were taken into account during hit optimization. UZH2 shows target engagement in cells and is able to reduce the m6A/A level of polyadenylated RNA in MOLM-13 (acute myeloid leukemia) and PC-3 (prostate cancer) cell lines.

METTL3 Inhibitors for Epitranscriptomic Modulation of Cellular Processes.

The methylase METTL3 is the writer enzyme of the N6 -methyladenosine (m6 A) modification of RNA. Using a structure-based drug discovery approach, we identified a METTL3 inhibitor with potency in a biochemical assay of 280 nM, while its enantiomer is 100 times less active. We observed a dose-dependent reduction in the m6 A methylation level of mRNA in several cell lines treated with the inhibitor already after 16 h of treatment, which lasted for at least 6 days. Importantly, the prolonged incubation (up to 6 days) with the METTL3 inhibitor did not alter levels of other RNA modifications (i. e., m1 A, m6 Am , m7 G), suggesting selectivity of the developed compound towards other RNA methyltransferases.

Atomistic and Thermodynamic Analysis of N6-Methyladenosine (m6A) Recognition by the Reader Domain of YTHDC1.

N6-Methyladenosine (m6A) is the most frequent modification in eukaryotic messenger RNA (mRNA) and its cellular processing and functions are regulated by the reader proteins YTHDCs and YTHDFs. However, the mechanism of m6A recognition by the reader proteins is still elusive. Here, we investigate this recognition process by combining atomistic simulations, site-directed mutagenesis, and biophysical experiments using YTHDC1 as a model. We find that the N6 methyl group of m6A contributes to the binding through its specific interactions with an aromatic cage (formed by Trp377 and Trp428) and also by favoring the association-prone conformation of m6A-containing RNA in solution. The m6A binding site dynamically equilibrates between multiple metastable conformations with four residues being involved in the regulation of m6A binding (Trp428, Met438, Ser378, and Thr379). Trp428 switches between two conformational states to build and dismantle the aromatic cage. Interestingly, mutating Met438 and Ser378 to alanine does not alter m6A binding to the protein but significantly redistributes the binding enthalpy and entropy terms, i.e., enthalpy-entropy compensation. Such compensation is reasoned by different entropy-enthalpy transduction associated with both conformational changes of the wild-type and mutant proteins and the redistribution of water molecules. In contrast, the point mutant Thr379Val significantly changes the thermal stability and binding capability of YTHDC1 to its natural ligand. Additionally, thermodynamic analysis and free energy calculations shed light on the role of a structural water molecule that synergistically binds to YTHDC1 with m6A and acts as the hub of a hydrogen-bond network. Taken together, the experimental data and simulation results may accelerate the discovery of chemical probes, m6A-editing tools, and drug candidates against reader proteins.

Hitting a Moving Target: Simulation and Crystallography Study of ATAD2 Bromodomain Blockers.

Small molecule ligand binding to the ATAD2 bromodomain is investigated here through the synergistic combination of molecular dynamics and protein crystallography. A previously unexplored conformation of the binding pocket upon rearrangement of the gatekeeper residue Ile1074 has been found. Further, our investigations reveal how minor structural differences in the ligands result in binding with different plasticity of the ZA loop for this difficult-to-drug bromodomain.

Small-Molecule Inhibitors of METTL3, the Major Human Epitranscriptomic Writer.

The RNA methylase METTL3 catalyzes the transfer of a methyl group from the cofactor S-adenosyl-L-methionine (SAM) to the N6 atom of adenine. We have screened a library of 4000 analogues and derivatives of the adenosine moiety of SAM by high-throughput docking into METTL3. Two series of adenine derivatives were identified in silico, and the binding mode of six of the predicted inhibitors was validated by protein crystallography. Two compounds, one for each series, show good ligand efficiency. We propose a route for their further development into potent and selective inhibitors of METTL3.

Selectively Disrupting m6A-Dependent Protein-RNA Interactions with Fragments.

We report a crystallographic analysis of small-molecule ligands of the human YTHDC1 domain that recognizes N6-methylated adenine (m6A) in RNA. The 30 binders are fragments (molecular weight < 300 g mol-1) that represent 10 different chemotypes identified by virtual screening. Despite the structural disorder of the binding site loop (residues 429-439), most of the 30 fragments emulate the two main interactions of the -NHCH3 group of m6A. These interactions are the hydrogen bond to the backbone carbonyl of Ser378 and the van der Waals contacts with the tryptophan cage. Different chemical groups are involved in the conserved binding motifs. Some of the fragments show favorable ligand efficiency for YTHDC1 and selectivity against other m6A reader domains. The structural information is useful for the design of modulators of m6A recognition by YTHDC1.

Flexible Binding of m6A Reader Protein YTHDC1 to Its Preferred RNA Motif.

N6-Methyladenosine (m6A) is the most prevalent chemical modification in human mRNAs. Its recognition by reader proteins enables many cellular functions, including splicing and translation of mRNAs. However, the binding mechanisms of m6A-containing RNAs to their readers are still elusive due to the unclear roles of m6A-flanking ribonucleotides. Here, we use a model system, YTHDC1 with its RNA motif 5'-G-2G-1(m6A)C+1U+2-3', to investigate the binding mechanisms by atomistic simulations, X-ray crystallography, and isothermal titration calorimetry. The experimental data and simulation results show that m6A is captured by an aromatic cage of YTHDC1 and the 3' terminus nucleotides are stabilized by cation-π-π interactions, while the 5' terminus remains flexible. Notably, simulations of unbound RNA motifs reveal that the methyl group of m6A and the 5' terminus shift the conformational preferences of the oligoribonucleotide to the bound-like conformation, thereby facilitating the association process. The binding mechanisms may help in the discovery of chemical probes against m6A reader proteins.

The roles of the conserved tyrosine in the ß2-α2 loop of the prion protein.

Prions cause neurodegenerative diseases for which no cure exists. Despite decades of research activities the function of the prion protein (PrP) in mammalians is not known. Moreover, little is known on the molecular mechanisms of the self-assembly of the PrP from its monomeric state (cellular PrP, PrP(C)) to the multimeric state. The latter state includes the toxic species (scrapie PrP, PrP(Sc)) knowledge of which would facilitate the development of drugs against prion diseases. Here we analyze the role of a tyrosine residue (Y169) which is strictly conserved in mammalian PrPs. Nuclear magnetic resonance (NMR) spectroscopy studies of many mammalian PrP(C) proteins have provided evidence of a conformational equilibrium between a 3(10)-helical turn and a type I ß turn conformation in the ß2-α2 loop (residues 165-175). In vitro cell-free experiments of the seeded conversion of PrP(C) indicate that non-aromatic residues at position 169 reduce the formation of proteinase K-resistant PrP. Recent molecular dynamics (MD) simulations of monomeric PrP and several single-point mutants show that Y169 stabilizes the 3(10)-helical turn conformation more than single-point mutants at position 169 or residues in contact with it. In the 3(10)-helical turn conformation the hydrophobic and aggregation-prone segment 169-YSNQNNF-175 is buried and thus not-available for self-assembly. From the combined analysis of simulation and experimental results it emerges that Y169 is an aggregation gatekeeper with a twofold role. Mutations related to 3 human prion diseases are interpreted on the basis of the gatekeeper role in the monomeric state. Another potential role of the Y169 side chain is the stabilization of the ordered aggregates, i.e., reduction of frangibility of filamentous protofibrils and fibrils, which is likely to reduce the generation of toxic species.

Structure-based drug design identifies polythiophenes as antiprion compounds.

Prions cause transmissible spongiform encephalopathies for which no treatment exists. Prions consist of PrP(Sc), a misfolded and aggregated form of the cellular prion protein (PrP(C)). We explore the antiprion properties of luminescent conjugated polythiophenes (LCPs) that bind and stabilize ordered protein aggregates. By administering a library of structurally diverse LCPs to the brains of prion-infected mice via osmotic minipumps, we found that antiprion activity required a minimum of five thiophene rings bearing regularly spaced carboxyl side groups. Solid-state nuclear magnetic resonance analyses and molecular dynamics simulations revealed that anionic side chains interacted with complementary, regularly spaced cationic amyloid residues of model prions. These findings allowed us to extract structural rules governing the interaction between LCPs and protein aggregates, which we then used to design a new set of LCPs with optimized binding. The new set of LCPs showed robust prophylactic and therapeutic potency in prion-infected mice, with the lead compound extending survival by >80% and showing activity against both mouse and hamster prions as well as efficacy upon intraperitoneal administration into mice. These results demonstrate the feasibility of targeted chemical design of compounds that may be useful for treating diseases of aberrant protein aggregation such as prion disease.

Evolutionary conserved Tyr169 stabilizes the ß2-α2 loop of the prion protein.

Experimental evidence indicates that the primary structure of the ß2-α2 loop region (residues 165-175) in mammalian prion proteins (PrP) influences the conversion from the cellular species (PrP(C)) to the ß-sheet-rich aggregate. Here, we captured the transition of the ß2-α2 loop from 310-helical turn to ß turn by unbiased molecular dynamics simulations of the single-point mutant Y169G. Multiple conformations along the spontaneous transition of the mutant were then used as starting point for sampling of the free-energy surface of the wild type and other single-point mutants. Using two different methods for the determination of free energy profiles, we found that the barrier for the 310-helical turn to ß turn transition of the wild type is higher by about 2.5 kcal/mol than for the Y169G mutant, which is due to favorable stacking of the aromatic rings of Y169 and F175, and a stable hydrogen bond between the side chains of Y169 and D178. The transition of the ß2-α2 loop to ß turn increases the solvent-exposure of the hydrophobic stretch 169-YSNQNNF-175. The simulations indicate that the strictly conserved Y169 in mammalian prion proteins stabilizes the 310-helical turn in the ß2-α2 loop, thus hindering the conversion to an aggregation-prone conformation.

Structural Analysis of the Binding of Type I, I1/2, and II Inhibitors to Eph Tyrosine Kinases.

We have solved the crystal structures of the EphA3 tyrosine kinase in complex with nine small-molecule inhibitors, which represent five different chemotypes and three main binding modes, i.e., types I and I1/2 (DFG in) and type II (DFG out). The three structures with type I1/2 inhibitors show that the higher affinity with respect to type I is due to an additional polar group (hydroxyl or pyrazole ring of indazole) which is fully buried and is involved in the same hydrogen bonds as the (urea or amide) linker of the type II inhibitors. Overall, the type I and type II binding modes belong to the lock-and-key and induced fit mechanism, respectively. In the type II binding, the scaffold in contact with the hinge region influences the position of the Phe765 side chain of the DFG motif and the orientation of the Gly-rich loop. The binding mode of Birb796 in the EphA3 kinase does not involve any hydrogen bond with the hinge region, which is different from the Birb796/p38 MAP kinase complex. Our structural analysis emphasizes the importance of accounting for structural plasticity of the ATP binding site in the design of type II inhibitors of tyrosine kinases.

Structured water molecules in the binding site of bromodomains can be displaced by cosolvent.

Bromodomains are α-helical bundles of approximately 110 residues that recognize acetylated lysine side chains mainly on histone tails. Bromodomains are known to play an important role in cancer and inflammation, and as such, significant efforts are being made to identify small-molecule inhibitors of these epigenetic reader proteins. Here, explicit solvent molecular dynamics (MD) simulations of two bromodomains (BAZ2B and CREBBP) are used to analyze the water molecules that seem to be conserved at the bottom of the acetyl-lysine binding site in most crystal structures of bromodomains. The MD runs suggest that the occupancy of the structured water molecules is influenced by conformational transitions of the loop that connects helices Z and A. Additional simulations in the presence of 50 molecules of cosolvent (i.e., 440 mM of dimethylsulfoxide, methanol, or ethanol) indicate that some of the structured water molecules can be displaced transiently. The residence time in the acetyl-lysine binding site is calculated to be about 1 ns, 2-5 ns, and 10-30 ns for methanol, ethanol, and dimethylsulfoxide, respectively, while the affinity of the three cosolvents for BAZ2B and CREBBP is in the range of 50-500 mM. The results described have implications for ligand design, suggesting that only structured water molecules that do not exchange with cosolvent should be maintained in crystal structures used for docking campaigns, and that hydroxy substituents should be incorporated in the ligand so as to map the structured water molecules replaced by (m)ethanol.

Does bromodomain flexibility influence histone recognition?

Bromodomains are protein modules that selectively recognize histones by binding to acetylated lysines. Here, we have carried out multiple molecular dynamics simulations of 20 human bromodomains to investigate the flexibility of their binding site. Some bromodomains show alternative side chain orientations of three evolutionarily conserved residues: the Asn involved in acetyl-lysine binding and two conserved aromatic residues. Furthermore, for the BAZ2B and CREBBP bromodomains we observe occlusion of the binding site which is coupled to the displacement of the two aromatic residues. In contrast to available structures, the simulations reveal large variability of the binding site accessibility. The simulations suggest that the flexibility of the bromodomain binding site and presence of self-occluded metastable states influence the recognition of acetyl-lysine on histone tails.

Optimization of inhibitors of the tyrosine kinase EphB4. 2. Cellular potency improvement and binding mode validation by X-ray crystallography.

Inhibition of the tyrosine kinase erythropoietin-producing human hepatocellular carcinoma receptor B4 (EphB4) is an effective strategy for the treatment of solid tumors. We have previously reported a low nanomolar ATP-competitive inhibitor of EphB4 discovered in silico by fragment-based high-throughput docking combined with explicit solvent molecular dynamics simulations. Here we present a second generation of EphB4 inhibitors that show high inhibitory potency in both enzymatic and cell-based assays while preserving the appealing selectivity profile exhibited by the parent compound. In addition, respectable levels of antiproliferative activity for these compounds have been obtained. Finally, the binding mode predicted by docking and molecular dynamics simulations is validated by solving the crystal structures of three members of this chemical class in complex with the EphA3 tyrosine kinase whose ATP-binding site is essentially identical to that of EphB4.

Discovery of tyrosine kinase inhibitors by docking into an inactive kinase conformation generated by molecular dynamics.

Several small molecules that bind to the inactive DFG-out conformation of tyrosine kinases (called type II inhibitors) have shown a good selectivity profile over other kinase targets. To obtain a set of DFG-out structures, we performed an explicit solvent molecular dynamics (MD) simulation of the complex of the catalytic domain of a tyrosine kinase receptor, ephrin type-A receptor 3 (EphA3), and a manually docked type II inhibitor. Automatic docking of four previously reported type II inhibitors was used to select a single snapshot from the MD trajectory for virtual screening. High-throughput docking of a pharmacophore-tailored library of 175,000 molecules resulted in about 4 million poses, which were further filtered by van der Waals efficiency and ranked according to a force-field-based energy function. Notably, around 20 % of the compounds with predicted binding energy smaller than -10 kcal mol(-1) are known type II inhibitors. Moreover, a series of 5-(piperazine-1-yl)isoquinoline derivatives was identified as a novel class of low-micromolar inhibitors of EphA3 and unphosphorylated Abelson tyrosine kinase (Abl1). The in silico predicted binding mode of the new inhibitors suggested a similar affinity to the gatekeeper mutant T315I of Abl1, which was verified in vitro by using a competition binding assay. Additional evidence for the type II binding mode was obtained by two 300 ns MD simulations of the complex between N-(3-chloro-4-(difluoromethoxy)phenyl)-2-(4-(8-nitroisoquinolin-5-yl)piperazin-1-yl)acetamide and EphA3.

How Does Darunavir Prevent HIV-1 Protease Dimerization?

The drug Darunavir (DRV) is a potent inhibitor of HIV-1 protease (PR), a homodimeric essential enzyme of the AIDS virus. Recent experimental data suggest that DRV is able to prevent dimerization of HIV-1 PR, which, together with its high affinity for the mature enzyme, has been linked to the high genetic barrier to the development of viral resistance. The mechanism of dimerization inhibition and the binding mode(s) of DRV to monomeric HIV-1 PR are unknown. Here, multiple molecular dynamics simulations with explicit solvent (for a total of 11 µs with the CHARMM force field and 1 µs with the Amber force field) show that the monomer of HIV-1 PR is structurally stable and reveal a major binding mode of DRV. This binding mode is stabilized by favorable interactions between the apolar groups of DRV and the hydrophobic residues Ile32, Ile47, Ile50, Ile54, Pro79, Val82, and Ile84. The binding mode to monomeric HIV-1 PR identified by molecular dynamics is different from the two binding modes observed in the crystal structure of the complex with dimeric HIV-1 PR. As an example, there are no interactions between DRV and the catalytic Asp25 in the binding mode to monomeric HIV-1 PR revelead by the simulations. In contrast, the simulations show extensive and stable interactions between DRV and the flap (residues 46-55), which are likely to sterically hinder the formation of the flap interface as observed in the dimeric structure. Which of the two mechanisms of inhibition (dimerization inhibition by association with the flap or binding to the active site of the mature enzyme) dominates might depend on the HIV-1 PR mutations, and it is likely that dimerization inhibition is predominant for multiple mutations at the active site in multidrug resistant strains.