Wet granulation is an interfacial reconstruction process mediated by liquid binders. However, the mechanisms underlying the distinct granulation behaviors of drugs with different solubilities remain poorly understood due to a lack of molecular-level insight. Here, we elucidate these mechanisms using density functional theory (DFT) calculations and molecular dynamics (MD) simulations. Three model drugs with distinct solubilities(?)aminophylline (AMI), metformin hydrochloride (MET), and diclofenac sodium (SD)(?)were selected. The results show that differences in granulation behavior cannot be attributed solely to solubility; rather, they arise from variations in interfacial interactions may be governed by crystal structure. Molecular simulations further reveal that crystal structure contributes to macroscopic granulation outcomes by regulating the dynamics of interfacial water molecules.

Non-destructive quality control strategies are essential to enable production of personalised printlets (3D printed tablets). The prediction of dissolution behaviour from printlet-specific data, such as spectroscopic, remains underexplored. In this study, we demonstrate the use of multimodal data consisting of printlet near-infrared (NIR) spectra, timepoint and geometric (surface area-to-volume ratio; SA/V) information, to predict in vitro dissolution profiles of dose-personalised, extended-release (ER) printlets manufactured by fused deposition modelling. Printlets varying in theophylline, hydroxypropyl cellulose (HPC), and D-mannitol were produced in three distinct volumes and SA/Vs for training and three distinct volumes of equal SA/V for testing. Partial least squares (PLS) and Random Forest (RF) regression were investigated through prediction of single or multiple dependent release variables. RF models using a single response variable with features Min-Max scaled outperformed multi-response RF and PLS approaches. Further feature reduction based on spectral variance enhanced the predictive capability, with retraining using the 10 most informative features achieving good predictive performance (R[2] = 0.994, RMSEtrain = 1.80%, RMSEtest = 2.75%). Time and SA/V were the most dictating features, while the retained NIR wavelengths, corresponding to characteristic vibrations from theophylline and HPC, offered model fine-tuning. The predicted dissolution profiles were equivalent to experimentally determined profiles for all test printlets, based on f1 and f2 similarity criteria. These results support the use of NIR spectroscopy with geometric parameters through multivariate modelling as a non-destructive approach for predicting drug release from 3D printed ER tablets to advance quality assurance for personalised pharmaceutical manufacturing.

We employed advanced quantum chemical methods to investigate the electronic properties of five widely consumed psychostimulants: nicotine, cotinine, caffeine, theophylline, and theobromine. Molecular geometries were optimized in an aqueous environment using density functional theory (DFT) with the B3LYP exchange-correlation functional and an extensive basis set. The calculated structures showed excellent agreement with experimentally determined solid-state structures deposited in the Cambridge Crystallographic Data Centre (CCDC). Full vibrational analyses enabled the determination of thermodynamic properties, including Gibbs energies. For the neutral ground states, we calculated dipole and quadrupole moments, dipole polarizabilities, zero-point vibrational energies, and entropic contributions. Adiabatic ionization energies and electron affinities were obtained from optimized cationic and anionic geometries and used to derive molecular electronegativities, chemical hardnesses, and electrophilicity indices. Reaction Gibbs energies for ionization and reduction were converted into absolute oxidation and reduction potentials. To improve accuracy, we additionally employed the highly correlated DLPNO-CCSD(T) method for ionization energies and electron affinities in water. Finally, the electronic and global properties of biologically relevant protonated forms nicotinium, cotinium, caffeinium, theophyllinium, and theobrominium were investigated, providing a description of these compounds in aqueous solution. This approach enabled comparison of their structural, thermodynamic, electronic, and redox properties across the investigated molecular series. In addition to the implicit Conductor-like Polarizable Continuum Model (CPCM), an explicit solvation approach involving several water molecules was also tested for the nicotine-water aggregate.

We present an integrated molecular simulation framework that combines molecular dynamics, enhanced sampling, and constant chemical potential simulations to uncover growth and dissolution mechanisms of molecular cocrystals and predict equilibrium solubility, using the theophylline-aspirin (THE-ASP) cocrystal as a prototypical system. Solubility is determined by sampling the growth and dissolution of ASP and THE molecules at a dimeric kink site on the crystal surface and computing the free-energy differences between their dissolved and crystalline states as a function of solution composition. These free-energy differences are interpreted as effective chemical potentials of the respective molecular species, and the solubility is identified as the solution composition at which the sum of the effective chemical potentials of ASP and THE equals zero. To address the large conformational and orientational freedom inherent to these molecules, we introduce collective variables that incorporate atomic density, molecular orientation, and solvent coordination, which, when used in Well-tempered Metadynamics simulations, enable efficient sampling of growth and dissolution processes and the calculation of well-converged free-energy surfaces. The calculated free energy profiles reveal asymmetric thermodynamic driving forces, with aspirin favoring incorporation into the crystal and theophylline preferentially stabilizing in solution, providing molecular-level insight into cocrystal growth dynamics. The predicted equilibrium solubility of the THE-ASP cocrystal, χ* = 0.00063 ± 0.00014, lies between the reported aqueous solubilities of the individual components (ASP ≈ (3-5) × 10-4; THE ≈ (6-8) × 10-4 at 25 °C).

Assays to assess respiratory function are included in the ICH S7A guidance for Safety Pharmacology, which identifies the core in vivo studies required to evaluate safety prior to First in Human (FIH) dosing. This requirement is most commonly met using dedicated stand-alone rodent respiratory studies that require additional studies and increased resources. More importantly, the demonstrated transitional value of dedicated respiratory endpoints to predict risk and improve FIH safety is lacking. As such, alternative approaches that utilize commonly collected signals from cardiovascular assessment studies, such as electrocardiograms (ECG), to calculate respiration rate (RespRate) as a biomarker of potential adverse effects are becoming increasingly relevant. In this study, we implement three algorithms developed to estimate ECG-derived RespRate (EDR) in humans on preclinical ECG data collected from dogs. We then compare EDR to the benchmark direct measure of RespRate derived from telemetry intra-thoracic pressure (ITP) sensors in 19 recordings. EDR obtained using the moment-based and principal component-based algorithms was not significantly different from the benchmark, based on a TOST (two one-sided tests) analysis with an equivalence margin of five breaths/min, in 100% and 89% of animals respectively. Furthermore, in a positive control study administering Theophylline, the PCA method detected a drug-induced increase in RespRate also present in the ITP data. Thus, EDR may provide a potential alternative to stand-alone respiratory assays, but further evaluation across a broader range of respiration changes, subjects, and ECG hardware is needed to establish the robustness of these algorithms.

Therapeutic drug monitoring and targeted toxicological assessment require quantitative methods capable of covering chemically diverse drugs over clinically relevant concentration ranges. We developed and validated a liquid chromatography-tandem mass spectrometry method for the simultaneous determination of 54 analytes in human serum, including antiepileptic drugs, antidepressants, antipsychotics, benzodiazepines, antifungal agents, theophylline, and selected clinically relevant metabolites. Serum samples were prepared by protein precipitation with perchloric acid. Chromatographic separation was achieved on an ACQUITY UPLC BEH C18 column using water/methanol mobile phases containing 2 mM ammonium acetate. Detection was performed in polarity-switching selected reaction monitoring mode using two transitions per analyte and 28 isotope-labeled or structurally related internal standards. The total run time was 8.5 min. Analyte-specific calibration models and validated concentration ranges covered 0.1-22,500 ng/mL, depending on the analyte. The method demonstrated acceptable selectivity, calibration performance, accuracy, precision, matrix effects, and stability under routine handling and storage conditions for analytes with available data. Within-run and between-run precision did not exceed 14.3% and 13.9%, respectively. Matrix effects ranged from -8.35% to +10.25%. Application to 120 remnant patient serum samples supported the feasibility of multiplex measurement of drugs from several therapeutic classes in a single analytical run. The method is suitable for routine serum-based therapeutic drug monitoring and for targeted toxicological interpretation of a predefined analyte panel.

The automated discovery of governing kinetic laws from observational data is a critical challenge in systems biology, pharmacology, and chemical engineering. In these experimental domains, data-driven mechanism discovery is complicated by sparse sampling, measurement noise (e.g., blood assays), and the prevalence of nonlinear saturation limits (Michaelis-Menten kinetics). Existing symbolic regression methods, such as SINDy, rely on polynomial basis functions that fundamentally fail to extrapolate physical saturation, and their reliance on numerical differentiation degrades catastrophically in high-noise regimes (>5%). In this work, we introduce ChemODE, a physics-informed gray-box framework tailored for biochemical and pharmacokinetic (PK) networks. ChemODE integrates (1) Constrained Rational Layers to autonomously enforce physical saturation laws, (2) On-the-fly Feature Normalization to resolve magnitude bias in multicompartment scale separations, and (3) a noise-robust integral-based optimization scheme. We validate ChemODE on fundamental biological benchmarks, including the nonlinear Brusselator limit cycle and a synthetic 3-compartment pharmacokinetic model. To demonstrate clinical applicability, ChemODE successfully extracts physiological absorption and elimination rates from real-world human theophylline pharmacokinetic assays. Furthermore, rigorous uncertainty quantification (10-seed ensembles) proves that ChemODE recovers enzyme kinetics with high precision (V max error ≈ 1%), significantly outperforming unconstrained baselines. This work establishes that integral-based minimization, combined with rational architectural constraints and a neural residual paradigm, provides a robust surrogate modeling path for interpretable mechanism discovery in noise-dominated biochemical settings.

Pipturus albidus (commonly known as māmaki) is an endemic Hawaiian plant traditionally consumed as a medicinal herbal tea; however, its comprehensive potential for phytochemical composition and biological activities remains underexplored. In this study, we present the first comprehensive phytochemical characterization of the ethanol extract of P. albidus, identifying twelve compounds through chromatographic and spectroscopic analyses. The isolated compounds were subjected to a series of antioxidant assays, including 2,2-diphenyl-1-picrylhydrazyl (DPPH) radical scavenging, ferric reducing antioxidant power (FRAP), and oxygen radical absorbance capacity (ORAC). Cinchonain Ia (2), vitexin (6), and (7S,8R)-dihydrodehydro diconiferyl alcohol (11) demonstrated strong free radical-scavenging and reducing capacities. Evaluation of Interleukin-8 (IL-8) expression in TNF-α-stimulated human aortic endothelial cells (HAECs) revealed that catechin-(7,8-bc)-4a-(3,4-dihydroxyphenyl)-dihydro-2(3H)-pyranone (3) and trans-ferulic acid (9) significantly increased IL-8 mRNA expression (p < 0.05), indicating potential pro-inflammatory effects of these isolated constituents. In contrast, neither the crude extract (PE) nor vitexin (6) produced a statistically significant difference in IL-8 mRNA levels compared to the TNF-α control (p < 0.05). Quantitative LC-MS/MS analysis confirmed that caffeine, theobromine, theophylline, and EGCG were below the limit of detection in P. albidus leaf extracts, indicating that P. albidus possesses a phytochemical constituent distinct from that of green tea. In summary, P. albidus leaves contain unique phytochemicals with cell-free antioxidant activity, while cell-based assays indicate potential pro-inflammatory risks for compounds 3 and 9 and no significant anti-inflammatory suppression by PE or vitexin (6).

Methylliberine (1,O2,7,9-tetramethyluric acid; 1,2,7,9-MU) and theacrine (1,3,7,9-tetramethyluric acid; 1,3,7,9-MU) are purine alkaloids found in a variety of botanicals and products marketed as dietary supplements. Despite their widespread use, extensive data on their human health effects are lacking. Therefore, methylliberine and theacrine were profiled using multiple new approach methods (NAMs): in silico metabolite predictions, in silico blood-brain barrier (BBB) permeability predictions, and in vitro assays for adenosine receptor binding (A1, A2A), functional activity (A2A), cytochrome P450 inhibition (CYP1A2), and phosphodiesterase inhibition (PDE3A, PDE4B, PDE4D). Like caffeine, methylliberine and theacrine were predicted to be metabolized via desmethylation by multiple in silico models. In silico predictions and cluster analysis also suggested that methylliberine and theacrine may permeate the BBB like caffeine. In vitro assays revealed that methylliberine and its metabolite 1,2,9-MU were CYP1A2 inhibitors (IC50, 9.6 and 27 μM, respectively) that agreed with clinical observations and could lead to interactions with drugs and xenobiotics. Both methylliberine and theacrine were A2A receptor antagonists, suggesting a potential for stimulant effects, although their potencies (52-62% inhibition at 100 μM) were weaker than theophylline, paraxanthine, and caffeine (IC50, 4.9-23.3 μM). Overall, this study highlights integration of multiple NAMs for profiling two purine alkaloids relevant to human health.

Dipeptidyl peptidase IV (DPP IV) inhibition prolongs incretin activity and represents an established therapeutic strategy for the management of type 2 diabetes mellitus. To investigate the underexplored combination of benzyl groups at the N7 position and a piperazine-linked arylthiazole moiety at the C8 position of the xanthine scaffold, 15 novel compounds (6a-6o) were synthesised. The synthetic route involved bromination at the C8 position of theophylline, introduction of benzyl groups at the N7 position, replacement of the bromine atom at the C8 position with piperazine, and subsequent cyclocondensation with phenacyl thiocyanate derivatives. All final compounds were characterised by [1]H and [13]C NMR spectroscopy, high-resolution accurate-mass mass spectrometry, and analytical HPLC. All derivatives inhibited DPP IV in the low-nanomolar range, with compounds 6c and 6b showing the highest activities, with IC50 values of 3.95 ± 0.53 and 4.04 ± 0.65 nM, respectively. NIH3T3 cytotoxicity IC50 values ranged from 28.52 to 397.63 µM, indicating substantial preliminary concentration margins relative to DPP IV inhibition. Molecular docking, 100 ns molecular dynamics simulations, interaction-fingerprint analyses and trajectory-based MM/PBSA calculations supported persistent interactions with key DPP IV active-site residues. These results identify the synthesised derivatives, particularly 6b and 6c, as promising lead structures for further optimisation.

Synthetic auxotrophs are a useful means for developing genetically encoded biocontainment systems. Current methods for developing synthetic auxotrophs are complicated or expensive, limiting access and adoption of biocontainment technologies. To address this gap, we developed a simple and modular platform for creating synthetic auxotrophs based on ligand-dependent translational gating of essential genes. By inserting a theophylline-responsive riboswitch upstream of essential genes in Escherichia coli, we created strains whose viability depends on the presence of theophylline. We systematically applied this approach to 29 essential genes, obtained ligand-dependent growth phenotypes for 19 targets, and found that 18 of these essential genes yielded stringent live-die synthetic auxotrophs. These strains exhibited robust theophylline dependence with escape frequencies ranging from 1 × 10-5 to 1 × 10-6 (most below limit of detection). Our modular design allows for the rapid (<1 week), low-cost, and reliable generation of synthetic auxotrophs. This work introduces ligand-dependent translational control as a new mechanism for engineering synthetic auxotrophy and provides an accessible platform that expands the biocontainment toolkit.