Synthesis of meta-substituted arene bioisosteres from [3.1.1]propellane.

Small-ring cage hydrocarbons are popular bioisosteres (molecular replacements) for commonly found para-substituted benzene rings in drug design1. The utility of these cage structures derives from their superior pharmacokinetic properties compared with their parent aromatics, including improved solubility and reduced susceptibility to metabolism2,3. A prime example is the bicyclo[1.1.1]pentane motif, which is mainly synthesized by ring-opening of the interbridgehead bond of the strained hydrocarbon [1.1.1]propellane with radicals or anions4. By contrast, scaffolds mimicking meta-substituted arenes are lacking because of the challenge of synthesizing saturated isosteres that accurately reproduce substituent vectors5. Here we show that bicyclo[3.1.1]heptanes (BCHeps), which are hydrocarbons for which the bridgehead substituents map precisely onto the geometry of meta-substituted benzenes, can be conveniently accessed from [3.1.1]propellane. We found that [3.1.1]propellane can be synthesized on a multigram scale, and readily undergoes a range of radical-based transformations to generate medicinally relevant carbon- and heteroatom-substituted BCHeps, including pharmaceutical analogues. Comparison of the absorption, distribution, metabolism and excretion (ADME) properties of these analogues reveals enhanced metabolic stability relative to their parent arene-containing drugs, validating the potential of this meta-arene analogue as an sp3-rich motif in drug design. Collectively, our results show that BCHeps can be prepared on useful scales using a variety of methods, offering a new surrogate for meta-substituted benzene rings for implementation in drug discovery programmes.

Optimization of a Series of 2,3-Dihydrobenzofurans as Highly Potent, Second Bromodomain (BD2)-Selective, Bromo and Extra-Terminal Domain (BET) Inhibitors.

Herein, a series of 2,3-dihydrobenzofurans have been developed as highly potent bromo and extra-terminal domain (BET) inhibitors with 1000-fold selectivity for the second bromodomain (BD2) over the first bromodomain (BD1). Investment in the development of two orthogonal synthetic routes delivered inhibitors that were potent and selective but had raised in vitro clearance and suboptimal solubility. Insertion of a quaternary center into the 2,3-dihydrobenzofuran core blocked a key site of metabolism and improved the solubility. This led to the development of inhibitor 71 (GSK852): a potent, 1000-fold-selective, highly soluble compound with good in vivo rat and dog pharmacokinetics.

Artificial neural network based isotopic analysis of airborne radioactivity measurement for radiological incident detection.

Responders need tools to rapidly detect and identify airborne alpha radioactivity during consequence management scenarios. Traditional continuous air monitoring systems used for this purpose compute the net counts in various energy windows to determine the presence of specified isotopes, such as 235U, 239Pu, and 241Am. These calculations rely on having a well-calibrated detector, which is challenging in low-background environments. Here, an alternative approach of using artificial neural networks to classify alpha spectra is presented. Two network architectures, fully connected and convolutional neural networks (CNNs), were trained to classify alpha spectra into four categories: background and background plus the three isotopes above. Sources were injected into measured background at various fractions of the derived response level (DRL) corresponding to early-phase Protective Action Guides. The convolutional network identifies all sources at 1% of the DRL with average probability of detection of 95% and false alarm probability of 1%. Further, the network identifies sources ranging between 0.25% and 1% of the DRL with higher than 80% probability of detection and lower than 7% false alarm probability. Most significantly, the network performance improves in low-count background conditions, increasing its minimum probability of detection to 93% and reducing the false alarm probabilities to lower than 0.25%. These results show that, once trained on datasets representing a range of detection scenarios, artificial neural networks can accurately identify alpha isotopes of interest without the need for detector calibration.

Dynamics of H+ + CO at E(Lab) = 30 eV.

The astrophysically relevant system H(+) + CO (v(i) = 0) → H(+) + CO (v(f)) at E(Lab) = 30 eV is studied with the simplest-level electron nuclear dynamics (SLEND) method. This investigation follows previous successful SLEND studies of H(+) + H(2) and H(+) + N(2) at E(Lab) = 30 eV [J. Morales, A. Diz, E. Deumens, and Y. Öhrn, J. Chem. Phys. 103(23), 9968 (1995); C. Stopera, B. Maiti, T. V. Grimes, P. M. McLaurin, and J. A. Morales, J. Chem. Phys. 134(22), 224308 (2011)]. SLEND is a direct, time-dependent, variational, and non-adiabatic method that adopts a classical-mechanics description for the nuclei and a single-determinantal wavefunction for the electrons. A canonical coherent-states (CS) procedure associated with SLEND reconstructs quantum vibrational properties from the SLEND classical dynamics. Present SLEND results include reactivity predictions, snapshots of the electron density evolution, average vibrational energy transfers, rainbow angle predictions, total and vibrationally resolved differential cross sections (DCS), and average vibrational excitation probabilities. SLEND results are compared with available data from experiments and vibrational close-coupling rotational infinite-order sudden (VCC-RIOS) approximation calculations. Present simulations employ four basis sets: STO-3G, 6-31G, 6-31G**, and cc-pVDZ to determine their effect on the results. SLEND simulations predict non-charge-transfer scattering and CO collision-induced dissociation as the main reactions. SLEND/6-31G, /6-31G**, and /cc-pVDZ predict rainbow angles and total DCS in excellent agreement with experiments and more accurate than their VCC-RIOS counterparts. SLEND/6-31G** and /cc-pVDZ predict vibrationally resolved DCS for v(f) = 0-2 in satisfactory experimental agreement, but less accurate than their comparable H(+) + CO VCC-RIOS and H(+) + H(2) and H(+) + N(2) SLEND results. SLEND∕6-31G** and ∕cc-pVDZ predict qualitatively correct average vibrational excitation probabilities, which are quantitatively correct for v(f) = 2, but under(over)estimated for v(f) = 0(1). Discrepancies in some H(+) + CO SLEND vibrational properties, not observed in H(+) + H(2) and H(+) + N(2) SLEND results, are attributed to the moderately overestimated SLEND vibrational energy through its effect upon the canonical CS probabilities. Correction of that energy to its experimental values produces a remarkable improvement in the average vibrational excitation probabilities. Ways to obtain more accurate vibrational properties with higher-level versions of electron nuclear dynamics are discussed.

Dynamics of H+ + N2 at E(Lab) = 30 eV.

The H(+) + N(2) system at E(Lab) = 30 eV, relevant in astrophysics, is investigated with the simplest-level electron nuclear dynamics (SLEND) method. SLEND is a time-dependent, direct, variational, non-adiabatic method that employs a classical-mechanics description for the nuclei and a single-determinantal wavefunction for the electrons. A canonical coherent-states procedure, intrinsic to SLEND, is used to reconstruct quantum vibrational properties from the SLEND classical mechanics. Present simulations employ three basis sets: STO-3G, 6-31G, and 6-31G∗∗, to determine their effect on the results, which include reaction visualizations, product predictions, and scattering properties. Present simulations predict non-charge-transfer scattering and N(2) collision-induced dissociation as the main reactions. Average vibrational energy transfer, H(+) energy-loss spectra, rainbow angle, and elastic vibrational differential cross sections at the SLEND∕6-31G∗∗ level agree well with available experimental data. SLEND∕6-31G∗∗ results are comparable to those calculated with the vibrational close-coupling rotational infinite-order sudden approximation and the quasi-classical trajectory method.

Activation of carbon-hydrogen bonds via 1,2-addition across M-X (X = OH or NH(2)) bonds of d(6) transition metals as a potential key step in hydrocarbon functionalization: a computational study.

Recent reports of 1,2-addition of C-H bonds across Ru-X (X = amido, hydroxo) bonds of TpRu(PMe3)X fragments {Tp = hydridotris(pyrazolyl)borate} suggest opportunities for the development of new catalytic cycles for hydrocarbon functionalization. In order to enhance understanding of these transformations, computational examinations of the efficacy of model d6 transition metal complexes of the form [(Tab)M(PH3)2X]q (Tab = tris-azo-borate; X = OH, NH2; q = -1 to +2; M = TcI, Re(I), Ru(II), Co(III), Ir(III), Ni(IV), Pt(IV)) for the activation of benzene C-H bonds, as well as the potential for their incorporation into catalytic functionalization cycles, are presented. For the benzene C-H activation reaction steps, kite-shaped transition states were located and found to have relatively little metal-hydrogen interaction. The C-H activation process is best described as a metal-mediated proton transfer in which the metal center and ligand X function as an activating electrophile and intramolecular base, respectively. While the metal plays a primary role in controlling the kinetics and thermodynamics of the reaction coordinate for C-H activation/functionalization, the ligand X also influences the energetics. On the basis of three thermodynamic criteria characterizing salient energetic aspects of the proposed catalytic cycle and the detailed computational studies reported herein, late transition metal complexes (e.g., Pt, Co, etc.) in the d6 electron configuration {especially the TabCo(PH3)2(OH)+ complex and related Co(III) systems} are predicted to be the most promising for further catalyst investigation.

Performance of the correlation consistent composite approach for transition states: a comparison to G3B theory.

The correlation consistent composite approach (ccCA) was applied to the prediction of reaction barrier heights (i.e., transition state energy relative to reactants and products) for a standard benchmark set of reactions comprised of both hydrogen transfer reactions and nonhydrogen transfer reactions (i.e., heavy-atom transfer, SN2, and unimolecular reactions). The ccCA method was compared against G3B for the same set of reactions. Error metrics indicate that ccCA achieves "chemical accuracy" with a mean unsigned error (MUE) of 0.89 kcal/mol with respect to the benchmark data for barrier heights; G3B has a mean unsigned error of 1.94 kcal/mol. Further, the greater accuracy of ccCA for predicted reaction barriers is compared to other benchmarked literature methods, including density functional (BB1K, MUE=1.16 kcal/mol) and wavefunction-based [QCISD(T), MUE=1.10 kcal/mol] methods.