Computational prediction of complex cationic rearrangement outcomes.

Recent years have seen revived interest in computer-assisted organic synthesis1,2. The use of reaction- and neural-network algorithms that can plan multistep synthetic pathways have revolutionized this field1,3-7, including examples leading to advanced natural products6,7. Such methods typically operate on full, literature-derived 'substrate(s)-to-product' reaction rules and cannot be easily extended to the analysis of reaction mechanisms. Here we show that computers equipped with a comprehensive knowledge-base of mechanistic steps augmented by physical-organic chemistry rules, as well as quantum mechanical and kinetic calculations, can use a reaction-network approach to analyse the mechanisms of some of the most complex organic transformations: namely, cationic rearrangements. Such rearrangements are a cornerstone of organic chemistry textbooks and entail notable changes in the molecule's carbon skeleton8-12. The algorithm we describe and deploy at https://HopCat.allchemy.net/ generates, within minutes, networks of possible mechanistic steps, traces plausible step sequences and calculates expected product distributions. We validate this algorithm by three sets of experiments whose analysis would probably prove challenging even to highly trained chemists: (1) predicting the outcomes of tail-to-head terpene (THT) cyclizations in which substantially different outcomes are encoded in modular precursors differing in minute structural details; (2) comparing the outcome of THT cyclizations in solution or in a supramolecular capsule; and (3) analysing complex reaction mixtures. Our results support a vision in which computers no longer just manipulate known reaction types1-7 but will help rationalize and discover new, mechanistically complex transformations.

Tuning sterol extraction kinetics yields a renal-sparing polyene antifungal.

Decades of previous efforts to develop renal-sparing polyene antifungals were misguided by the classic membrane permeabilization model1. Recently, the clinically vital but also highly renal-toxic small-molecule natural product amphotericin B was instead found to kill fungi primarily by forming extramembraneous sponge-like aggregates that extract ergosterol from lipid bilayers2-6. Here we show that rapid and selective extraction of fungal ergosterol can yield potent and renal-sparing polyene antifungals. Cholesterol extraction was found to drive the toxicity of amphotericin B to human renal cells. Our examination of high-resolution structures of amphotericin B sponges in sterol-free and sterol-bound states guided us to a promising structural derivative that does not bind cholesterol and is thus renal sparing. This derivative was also less potent because it extracts ergosterol more slowly. Selective acceleration of ergosterol extraction with a second structural modification yielded a new polyene, AM-2-19, that is renal sparing in mice and primary human renal cells, potent against hundreds of pathogenic fungal strains, resistance evasive following serial passage in vitro and highly efficacious in animal models of invasive fungal infections. Thus, rational tuning of the dynamics of interactions between small molecules may lead to better treatments for fungal infections that still kill millions of people annually7,8 and potentially other resistance-evasive antimicrobials, including those that have recently been shown to operate through supramolecular structures that target specific lipids9.

Automated iterative Csp3-C bond formation.

Fully automated synthetic chemistry would substantially change the field by providing broad on-demand access to small molecules. However, the reactions that can be run autonomously are still limited. Automating the stereospecific assembly of Csp3-C bonds would expand access to many important types of functional organic molecules1. Previously, methyliminodiacetic acid (MIDA) boronates were used to orchestrate the formation of Csp2-Csp2 bonds and were effective building blocks for automating the synthesis of many small molecules2, but they are incompatible with stereospecific Csp3-Csp2 and Csp3-Csp3 bond-forming reactions3-10. Here we report that hyperconjugative and steric tuning provide a new class of tetramethyl N-methyliminodiacetic acid (TIDA) boronates that are stable to these conditions. Charge density analysis11-13 revealed that redistribution of electron density increases covalency of the N-B bond and thereby attenuates its hydrolysis. Complementary steric shielding of carbonyl π-faces decreases reactivity towards nucleophilic reagents. The unique features of the iminodiacetic acid cage2, which are essential for generalized automated synthesis, are retained by TIDA boronates. This enabled Csp3 boronate building blocks to be assembled using automated synthesis, including the preparation of natural products through automated stereospecific Csp3-Csp2 and Csp3-Csp3 bond formation. These findings will enable increasingly complex Csp3-rich small molecules to be accessed via automated assembly.

Minimizing higher-order aggregation maximizes iron mobilization by small molecules.

The natural product hinokitiol mobilizes iron across lipid bilayers at low concentrations and restores hemoglobinization in iron transporter protein-deficient systems. But hinokitiol fails to similarly mobilize iron at higher concentrations, limiting its uses in chemical biology and medicine. Here we show that at higher concentrations, hinokitiol3:Fe(III) complexes form large, higher-order aggregates, leading to loss of transmembrane iron mobilization. Guided by this understanding and systematic structure-function studies enabled by modular synthesis, we identified FeM-1269, which minimally aggregates and dose-dependently mobilizes iron across lipid bilayers even at very high concentrations. In contrast to hinokitiol, FeM-1269 is also well-tolerated in animals at high doses for extended periods of time. In a mouse model of anemia of inflammation, FeM-1269 increases serum iron, transferrin saturation, hemoglobin and hematocrit. This rationally developed iron-mobilizing small molecule has enhanced potential as a molecular prosthetic for understanding and potentially treating iron transporter deficiencies.

Small-molecule ion channels increase host defences in cystic fibrosis airway epithelia.

Loss-of-function mutations in the cystic fibrosis transmembrane conductance regulator (CFTR) compromise epithelial HCO3- and Cl- secretion, reduce airway surface liquid pH, and impair respiratory host defences in people with cystic fibrosis1-3. Here we report that apical addition of amphotericin B, a small molecule that forms unselective ion channels, restored HCO3- secretion and increased airway surface liquid pH in cultured airway epithelia from people with cystic fibrosis. These effects required the basolateral Na+, K+-ATPase, indicating that apical amphotericin B channels functionally interfaced with this driver of anion secretion. Amphotericin B also restored airway surface liquid pH, viscosity, and antibacterial activity in primary cultures of airway epithelia from people with cystic fibrosis caused by different mutations, including ones that do not yield CFTR, and increased airway surface liquid pH in CFTR-null pigs in vivo. Thus, unselective small-molecule ion channels can restore host defences in cystic fibrosis airway epithelia via a mechanism that is independent of CFTR and is therefore independent of genotype.

A small molecule redistributes iron in ferroportin-deficient mice and patient-derived primary macrophages.

Deficiencies of the transmembrane iron-transporting protein ferroportin (FPN1) cause the iron misdistribution that underlies ferroportin disease, anemia of inflammation, and several other human diseases and conditions. A small molecule natural product, hinokitiol, was recently shown to serve as a surrogate transmembrane iron transporter that can restore hemoglobinization in zebrafish deficient in other iron transporting proteins and can increase gut iron absorption in FPN1-deficient flatiron mice. However, whether hinokitiol can restore normal iron physiology in FPN1-deficient animals or primary cells from patients and the mechanisms underlying such targeted activities remain unknown. Here, we show that hinokitiol redistributes iron from the liver to red blood cells in flatiron mice, thereby increasing hemoglobin and hematocrit. Mechanistic studies confirm that hinokitiol functions as a surrogate transmembrane iron transporter to release iron trapped within liver macrophages, that hinokitiol-Fe complexes transfer iron to transferrin, and that the resulting transferrin-Fe complexes drive red blood cell maturation in a transferrin-receptor-dependent manner. We also show in FPN1-deficient primary macrophages derived from patients with ferroportin disease that hinokitiol moves labile iron from inside to outside cells and decreases intracellular ferritin levels. The mobilization of nonlabile iron is accompanied by reductions in intracellular ferritin, consistent with the activation of regulated ferritin proteolysis. These findings collectively provide foundational support for the translation of small molecule iron transporters into therapies for human diseases caused by iron misdistribution.

Segmental Dynamics of Membranous Cholesterol are Coupled.

Cholesterol promotes the structural integrity of the fluid cell membrane and interacts dynamically with many membrane proteins to regulate function. Understanding site-resolved cholesterol structural dynamics is thus important. This long-standing challenge has thus far been addressed, in part, by selective isotopic labeling approaches. Here we present a new 3D solid-state NMR (SSNMR) experiment utilizing scalar 13C-13C polarization transfer and recoupling of the 1H-13C interactions in order to determine average dipolar couplings for all 1H-13C vectors in uniformly 13C-enriched cholesterol. The experimentally determined order parameters (OP) agree exceptionally well with molecular dynamics (MD) trajectories and reveal coupling among several conformational degrees of freedom in cholesterol molecules. Quantum chemistry shielding calculations further support this conclusion and specifically demonstrate that ring tilt and rotation are coupled to changes in tail conformation and that these coupled segmental dynamics dictate the orientation of cholesterol. These findings advance our understanding of physiologically relevant dynamics of cholesterol, and the methods that revealed them have broader potential to characterize how structural dynamics of other small molecules impact their biological functions.

MIDA- and TIDA-Boronates Stabilize α-Radicals Through B-N Hyperconjugation.

Multifunctional organoboron compounds increasingly enable the simple generation of complex, Csp3 -rich small molecules. The ability of boron-containing functional groups to modify the reactivity of α-radicals has also enabled a myriad of chemical reactions. Boronic esters with vacant p-orbitals have a significant stabilizing effect on α-radicals due to delocalization of spin density into the empty orbital. The effect of coordinatively saturated derivatives, such as N-methyliminodiacetic acid (MIDA) boronates and counterparts, remains less clear. Herein, we demonstrate that coordinatively saturated MIDA and TIDA boronates stabilize secondary alkyl α-radicals via σB-N hyperconjugation in a manner that allows site-selective C-H bromination. DFT calculated radical stabilization energies and spin density maps as well as LED NMR kinetic analysis of photochemical bromination rates of different boronic esters further these findings. This work clarifies that the α-radical stabilizing effect of boronic esters does not only proceed via delocalization of radical character into vacant boron p-orbitals, but that hyperconjugation of tetrahedral boron-containing functional groups and their ligand electron delocalizing ability also play a critical role. These findings establish boron ligands as a useful dial for tuning reactivity at the α-carbon.

Machine Learning May Sometimes Simply Capture Literature Popularity Trends: A Case Study of Heterocyclic Suzuki-Miyaura Coupling.

Applications of machine learning (ML) to synthetic chemistry rely on the assumption that large numbers of literature-reported examples should enable construction of accurate and predictive models of chemical reactivity. This paper demonstrates that abundance of carefully curated literature data may be insufficient for this purpose. Using an example of Suzuki-Miyaura coupling with heterocyclic building blocks─and a carefully selected database of >10,000 literature examples─we show that ML models cannot offer any meaningful predictions of optimum reaction conditions, even if the search space is restricted to only solvents and bases. This result holds irrespective of the ML model applied (from simple feed-forward to state-of-the-art graph-convolution neural networks) or the representation to describe the reaction partners (various fingerprints, chemical descriptors, latent representations, etc.). In all cases, the ML methods fail to perform significantly better than naive assignments based on the sheer frequency of certain reaction conditions reported in the literature. These unsatisfactory results likely reflect subjective preferences of various chemists to use certain protocols, other biasing factors as mundane as availability of certain solvents/reagents, and/or a lack of negative data. These findings highlight the likely importance of systematically generating reliable and standardized data sets for algorithm training.

Transition between Nonresonant and Resonant Charge Transport in Molecular Junctions.

Efficient long-range charge transport is required for high-performance molecular electronic devices. Resonant transport is thought to occur in single molecule junctions when molecular frontier orbital energy levels align with electrode Fermi levels, thereby enabling efficient transport without molecular or environmental relaxation. Despite recent progress, we lack a systematic understanding of the transition between nonresonant and resonant transport for molecular junctions with different chemical compositions. In this work, we show that molecular junctions undergo a reversible transition from nonresonant tunneling to resonant transport as a function of applied bias. Transient bias-switching experiments show that the nonresonant to resonant transition is reversible with the applied bias. We determine a general quantitative relationship that describes the transition voltage as a function of the molecular frontier orbital energies and electrode Fermi levels. Overall, this work highlights the importance of frontier orbital energy alignment in achieving efficient charge transport in molecular devices.

Digitizing Chemical Synthesis in 3D Printed Reactionware.

Chemistry digitization requires an unambiguous link between experiments and the code used to generate the experimental conditions and outcomes, yet this process is not standardized, limiting the portability of any chemical code. What is needed is a universal approach to aid this process using a well-defined standard that is composed of syntheses that are employed in modular hardware. Herein we present a new approach to the digitization of organic synthesis that combines process chemistry principles with 3D printed reactionware. This approach outlines the process for transforming unit operations into digitized hardware and well-defined instructions that ensure effective synthesis. To demonstrate this, we outline the process for digitizing 3 MIDA boronate building blocks, an ester hydrolysis, a Wittig olefination, a Suzuki-Miyaura coupling reaction, and synthesis of the drug sulfanilamide.

FAM210B is an erythropoietin target and regulates erythroid heme synthesis by controlling mitochondrial iron import and ferrochelatase activity.

Erythropoietin (EPO) signaling is critical to many processes essential to terminal erythropoiesis. Despite the centrality of iron metabolism to erythropoiesis, the mechanisms by which EPO regulates iron status are not well-understood. To this end, here we profiled gene expression in EPO-treated 32D pro-B cells and developing fetal liver erythroid cells to identify additional iron regulatory genes. We determined that FAM210B, a mitochondrial inner-membrane protein, is essential for hemoglobinization, proliferation, and enucleation during terminal erythroid maturation. Fam210b deficiency led to defects in mitochondrial iron uptake, heme synthesis, and iron-sulfur cluster formation. These defects were corrected with a lipid-soluble, small-molecule iron transporter, hinokitiol, in Fam210b-deficient murine erythroid cells and zebrafish morphants. Genetic complementation experiments revealed that FAM210B is not a mitochondrial iron transporter but is required for adequate mitochondrial iron import to sustain heme synthesis and iron-sulfur cluster formation during erythroid differentiation. FAM210B was also required for maximal ferrochelatase activity in differentiating erythroid cells. We propose that FAM210B functions as an adaptor protein that facilitates the formation of an oligomeric mitochondrial iron transport complex, required for the increase in iron acquisition for heme synthesis during terminal erythropoiesis. Collectively, our results reveal a critical mechanism by which EPO signaling regulates terminal erythropoiesis and iron metabolism.

Solid-State NMR of highly 13C-enriched cholesterol in lipid bilayers.

Cholesterol (Chol) is vital for cell function as it is essential to a myriad of biochemical and biophysical processes. The atomistic details of Chol's interactions with phospholipids and proteins is therefore of fundamental interest, and NMR offers unique opportunities to interrogate these properties at high resolution. Towards this end, here we describe approaches for examining the structure and dynamics of Chol in lipid bilayers using high levels of 13C enrichment in combination with magic-angle spinning (MAS) methods. We quantify the incorporation levels and demonstrate high sensitivity and resolution in 2D 13C-13C and 1H-13C spectra, enabling de novo assignments and site-resolved order parameter measurements obtained in a fraction of the time required for experiments with natural abundance sterols. We envision many potential future applications of these methods to study sterol interactions with drugs, lipids and proteins.

Peridinin Is an Exceptionally Potent and Membrane-Embedded Inhibitor of Bilayer Lipid Peroxidation.

Antilipoperoxidant protein dysfunction is associated with many human diseases, suggesting that bilayer lipid peroxidation may contribute broadly to pathogenesis. Small molecule inhibitors of this membrane-localized chemistry could in theory enable better understanding and/or treatment of such diseases, but currently available compounds have important limitations. Many biological questions thus remain unanswered, and clinical trials have largely been disappointing. Enabled by efficient, building block-based syntheses of three atypical carotenoid natural products produced by microorganisms that thrive in environments of extreme oxidative stress, we found that peridinin is a potent inhibitor of nonenzymatic bilayer lipid peroxidation in liposomes and in primary human endothelial cells. We also found that peridinin blocks monocyte-endothelial cell adhesion, a key step in atherogenesis. A series of frontier solid-state NMR experiments with a site-specifically 13C-labeled isotopolog synthesized using the same MIDA boronate building block-based total synthesis approach revealed that peridinin is completely embedded within and physically spans the hydrophobic core of POPC membranes, maximizing its effective molarity at the site of the targeted lipid peroxidation reactions. Alternatively, the widely used carotenoid astaxanthin is significantly less potent and was found to primarily localize extramembranously. Peridinin thus represents a promising and biophysically well-characterized starting point for the development of small molecule antilipoperoxidants that serve as more effective biological probes and/or therapeutics.

The Molecular Industrial Revolution: Automated Synthesis of Small Molecules.

Today we are poised for a transition from the highly customized crafting of specific molecular targets by hand to the increasingly general and automated assembly of different types of molecules with the push of a button. Creating machines that are capable of making many different types of small molecules on demand, akin to that which has been achieved on the macroscale with 3D printers, is challenging. Yet important progress is being made toward this objective with two complementary approaches: 1) Automation of customized synthesis routes to different targets by machines that enable the use of many reactions and starting materials, and 2) automation of generalized platforms that make many different targets using common coupling chemistry and building blocks. Continued progress in these directions has the potential to shift the bottleneck in molecular innovation from synthesis to imagination, and thereby help drive a new industrial revolution on the molecular scale.

Nontoxic antimicrobials that evade drug resistance.

Drugs that act more promiscuously provide fewer routes for the emergence of resistant mutants. This benefit, however, often comes at the cost of serious off-target and dose-limiting toxicities. The classic example is the antifungal amphotericin B (AmB), which has evaded resistance for more than half a century. We report markedly less toxic amphotericins that nevertheless evade resistance. They are scalably accessed in just three steps from the natural product, and they bind their target (the fungal sterol ergosterol) with far greater selectivity than AmB. Hence, they are less toxic and far more effective in a mouse model of systemic candidiasis. To our surprise, exhaustive efforts to select for mutants resistant to these more selective compounds revealed that they are just as impervious to resistance as AmB. Thus, highly selective cytocidal action and the evasion of resistance are not mutually exclusive, suggesting practical routes to the discovery of less toxic, resistance-evasive therapies.

From synthesis to function via iterative assembly of N-methyliminodiacetic acid boronate building blocks.

The study and optimization of small molecule function is often impeded by the time-intensive and specialist-dependent process that is typically used to make such compounds. In contrast, general and automated platforms have been developed for making peptides, oligonucleotides, and increasingly oligosaccharides, where synthesis is simplified to iterative applications of the same reactions. Inspired by the way natural products are biosynthesized via the iterative assembly of a defined set of building blocks, we developed a platform for small molecule synthesis involving the iterative coupling of haloboronic acids protected as the corresponding N-methyliminodiacetic acid (MIDA) boronates. Here we summarize our efforts thus far to develop this platform into a generalized and automated approach for small molecule synthesis. We and others have employed this approach to access many polyene-based compounds, including the polyene motifs found in >75% of all polyene natural products. This platform further allowed us to derivatize amphotericin B, the powerful and resistance-evasive but also highly toxic last line of defense in treating systemic fungal infections, and thereby understand its mechanism of action. This synthesis-enabled mechanistic understanding has led us to develop less toxic derivatives currently under evaluation as improved antifungal agents. To access more Csp(3)-containing small molecules, we gained a stereocontrolled entry into chiral, non-racemic α-boryl aldehydes through the discovery of a chiral derivative of MIDA. These α-boryl aldehydes are versatile intermediates for the synthesis of many Csp(3) boronate building blocks that are otherwise difficult to access. In addition, we demonstrated the utility of these types of building blocks in accessing pharmaceutically relevant targets via an iterative Csp(3) cross-coupling cycle. We have further expanded the scope of the platform to include stereochemically complex macrocyclic and polycyclic molecules using a linear-to-cyclized strategy, in which Csp(3) boronate building blocks are iteratively assembled into linear precursors that are then cyclized into the cyclic frameworks found in many natural products and natural product-like structures. Enabled by the serendipitous discovery of a catch-and-release protocol for generally purifying MIDA boronate intermediates, the platform has been automated. The synthesis of 14 distinct classes of small molecules, including pharmaceuticals, materials components, and polycyclic natural products, has been achieved using this new synthesis machine. It is anticipated that the scope of small molecules accessible by this platform will continue to expand via further developments in building block synthesis, Csp(3) cross-coupling methodologies, and cyclization strategies. Achieving these goals will enable the more generalized synthesis of small molecules and thereby help shift the rate-limiting step in small molecule science from synthesis to function.

Amphotericin forms an extramembranous and fungicidal sterol sponge.

For over 50 years, amphotericin has remained the powerful but highly toxic last line of defense in treating life-threatening fungal infections in humans with minimal development of microbial resistance. Understanding how this small molecule kills yeast is thus critical for guiding development of derivatives with an improved therapeutic index and other resistance-refractory antimicrobial agents. In the widely accepted ion channel model for its mechanism of cytocidal action, amphotericin forms aggregates inside lipid bilayers that permeabilize and kill cells. In contrast, we report that amphotericin exists primarily in the form of large, extramembranous aggregates that kill yeast by extracting ergosterol from lipid bilayers. These findings reveal that extraction of a polyfunctional lipid underlies the resistance-refractory antimicrobial action of amphotericin and suggests a roadmap for separating its cytocidal and membrane-permeabilizing activities. This new mechanistic understanding is also guiding development of what are to our knowledge the first derivatives of amphotericin that kill yeast but not human cells.