A first-principles exploration of the conformational space of sodiated di-saccharides assisted by semi-empirical methods and neural network potentials.

Previous exploration of the conformational space of sodiated mono-saccharides using a random search algorithm leads to ∼103 structurally distinct conformers covering an energy range of ∼150 kJ mol-1. Thus, it is reasonable to expect that the number of distinct conformers for a given disaccharide would be on the order of 106. Efficient identification of distinct conformers at the first-principles level has been demonstrated with the assistance of neural network potential (NNP) with an accuracy of ∼1 kJ mol-1 compared to DFT. Leveraging a local minima database of neutral and sodiated glucose (Glc), we develop algorithms to systematically explore the conformation landscape of 19 Glc-based sodiated disaccharides. To accelerate the exploration, the NNP method is implemented. The NNP achieves an accuracy of ∼2.3 kJ mol-1 compared to DFT, offering a comparable quality to that of DFT. Through a multi-model approach integrating DFTB3, NNP and DFT, we can rapidly locate low-energy disaccharide conformers at the first-principles level. The methodology we show here can be used to efficiently explore the potential energy landscape of any di-saccharides when first-principles accuracy is required.

A first-principles exploration of the conformational space of sodiated pyranose assisted by neural network potentials.

Sampling the conformational space of monosaccharides using the first-principles methods is important and as a database of local minima provides a solid base for interpreting experimental measurements such as infrared photo-dissociation (IRPD) spectroscopy or collision-induced dissociation (CID). IRPD emphasizes low-energy conformers and CID can distinguish conformers with distinct reaction pathways. A typical computational approach is to engage empirical or semi-empirical methods to sample the conformational space first, and only selected minima are reoptimized at first-principles levels. In this work, we propose a computational scheme to explore the configurational space of 12 types of sodiated pyranoses with the assistance of a neural network potential (NNP). We demonstrated that it is possible to train an NNP based on the density functional calculations extracted from a previous study on sodiated glucose (Glc), galactose (Gal), and mannose (Man). This NNP yields a better description of the other five types of aldohexoses than the four types of ketohexoses. We further show that such a discrepancy in the accuracy of NNP can be resolved by an active learning scheme where the NNP model is engaged in generating the data and has itself updated. Through this iterative process, we can locate more than 17 000 distinct local minima at the B3LYP/6-311+G(d,p) level and an NNP with an accuracy of 1 kJ mol-1 was created, which can be used for further studies.

Switchable Lipids: From Conformational Switch to Macroscopic Changes in Lipid Vesicles.

It has been shown that the use of conformationally pH-switchable lipids can drastically enhance the cytosolic drug delivery of lipid vesicles. Understanding the process by which the pH-switchable lipids disturb the lipid assembly of nanoparticles and trigger the cargo release is crucial to optimize the rational design of pH-switchable lipids. Here, we gather morphological observations (FF-SEM, Cryo-TEM, AFM, confocal microscopy), physicochemical characterization (DLS, ELS), as well as phase behavior studies (DSC, 2H NMR, Langmuir isotherm, and MAS NMR) to propose a mechanism of pH-triggered membrane destabilization. We demonstrate that the switchable lipids are homogeneously incorporated with other co-lipids (DSPC, cholesterol, and DSPE-PEG2000) and promote a liquid-ordered phase insensitive to temperature variation. Upon acidification, the protonation of the switchable lipids triggers a conformational switch altering the self-assembly properties of lipid nanoparticles. These modifications do not lead to a phase separation of the lipid membrane; however, they cause fluctuations and local defects, which result in morphological changes of the lipid vesicles. These changes are proposed to affect the permeability of vesicle membrane, triggering the release of the cargo encapsulated in the lipid vesicles (LVs). Our results confirm that pH-triggered release does not require major morphological changes, but can result from small defects affecting the lipid membrane permeability.

A self-adapting first-principles exploration on the dissociation mechanism in sodiated aldohexose pyranoses assisted with neural network potentials.

Understanding the mechanism of collision-induced dissociation (CID) in mono-saccharides with density functional theory (DFT) is challenging because of many possible reaction paths that originate from their high structural diversity. To search for the transition state (TS) from the huge number of conformers, we propose a three-step search scheme with the assistance of neural network potential (NNP). The search starts from a cross-checking of sugars, to a global search of all possible channels, and in the end, an exhaustive exploration around the low-lying channels. The cross-checking step quickly adapts the NNP from the studied molecules to the target ones. The other two steps utilize the adapted NNP to find the available pathways via random sampling of the structures. The study of the CID reactions in all eight types of aldohexose pyranoses was applied using the search scheme. The DFT calculations on AH-0 (Glc, Gal, and Man) in the previous study were utilized to construct an NNP and provide the TS structure database for searching AH-1 (All, Alt, Gul, Ido, and Tal). In total, we identified around 5200 TSs in AH-0 and AH-1, and the final NNP covers an energy range of more than 500 kJ mol-1 with a mean absolute error of energy less than 4 kJ mol-1. The search scheme is useful not only for saccharides but also for highly flexible bio-molecules.

Bimodal brush-functionalized nanoparticles selective to receptor surface density.

Nanoparticles or drug carriers which can selectively bind to cells expressing receptors above a certain threshold surface density are very promising for targeting cells overexpressing specific receptors under pathological conditions. Simulations and theoretical studies have suggested that such selectivity can be enhanced by functionalizing nanoparticles with a bimodal polymer monolayer (BM) containing shorter ligated chains and longer inert protective chains. However, a systematic study of the effect of these parameters under tightly controlled conditions is still missing. Here, we develop well-defined and highly specific platforms mimicking particle-cell interface using surface chemistry to provide a experimental proof of such selectivity. Using surface plasmon resonance and atomic force microscopy, we report the selective adsorption of BM-functionalized nanoparticles, and especially, a significant enhanced selective behavior by using a BM with longer protective chains. Furthermore, a model is also developed to describe the repulsive contribution of the protective brush to nanoparticle adsorption. This model is combined with super-selectivity theory to support experimental findings and shows that the observed selectivity is due to the steric energy barrier which requires a high number of ligand-receptor bonds to allow nanoparticle adsorption. Finally, the results show how the relative length and molar ratio of two chains can be tuned to target a threshold surface density of receptors and thus lay the foundation for the rational design of BM-functionalized nanoparticles for selective targeting.

Capturing the potential energy landscape of large size molecular clusters from atomic interactions up to a 4-body system using deep learning.

Exploring the structure and properties of molecular clusters with accuracy using the ab initio methods is a resource intensive task due to the increasing cost of the ab initio methods and the number of distinct conformers as the size increases. The energy landscape of methanol clusters has been previously explored using computationally efficient empirical models to collect a database of structurally distinct minima, followed by re-optimization using ab initio methods. In this work, we propose a new method that utilizes the database of stable conformers and borrow the fragmentation concept of many-body-expansion (MBE) methods in ab initio methods to train a deep-learning machine learning (ML) model using SchNet. Picking 684 local minima of (CH3OH)5 to (CH3OH)8 from the existing database, we can generate ∼51 000 data points of one-body, two-body, three-body and four-body molecular systems to train an ML model to reach a mean absolute error (MAE) of 3.19 kJ mol-1 (in energy) and 2.48 kJ mol-1 Å-1 (in forces) tested against ab initio calculations up to (CH3OH)14. This ML model is then used to create a database of low energy isomers of (CH3OH)n (n = 15-20). The proposed scheme can be applied to other hydrogen bonded molecular clusters with an accuracy of first-principles methods and computational speed of empirical force-fields.

Collision-induced dissociation of Na+-tagged ketohexoses: experimental and computational studies on fructose.

Collision-induced dissociation tandem mass spectrometry (CID-MSn) and computational investigation at the MP2/6-311+G(d,p) level of theory have been employed to study Na+-tagged fructose, an example of a ketohexose featuring four cyclic isomers: α-fructofuranose (αFruf), ß-fructofuranose (ßFruf), α-fructopyranose (αFrup), and ß-fructopyranose (ßFrup). The four isomers can be separated by high-performance liquid chromatography (HPLC) and they show different mass spectra, indicating that CID-MSn can distinguish the different fructose forms. Based on a simulation using a micro-kinetic model, we have obtained an overview of the mechanisms for the different dissociation pathways. It has been demonstrated that the preference for the C-C cleavage over the competing isomerization of linear fructose is the main reason for the previously reported differences between the CID-MS spectra of aldohexoses and ketohexoses. In addition, the kinetic modeling helped to confirm the assignment of the different measured mass spectra to the different fructose isomers. The previously reported assignment based on the peak intensities in the HPLC chromatogram had left some open questions as the preference for the dehydration channels did not always follow trends previously observed for aldohexoses. Setting up the kinetic model further enabled us to directly compare the computational and experimental results, which indicated that the model can reproduce most trends in the differences between the dissociation pathways of the four cyclic fructose isomers.

Survivin silencing improved the cytotoxicity of carboplatin and melphalan in Y79 and primary retinoblastoma cells.

Survivin stands out as one of the most specific cancer targets discovered to date. Although single inhibition, e.g. through small interfering RNA (siRNA), has shown modest results in clinical trials, its combination with drugs holds promise to sensitize cancer cells to chemotherapeutics. In this study, we propose a sequential treatment of siRNA survivin followed by chemotherapy. Firstly, we demonstrated that siRNA-loaded switchable lipid nanoparticles (siLNP) silence survivin in a panel of cancer cell lines. Subsequently, we selected retinoblastoma (RB) as our model to screen four chemotherapeutic agents: carboplatin, topotecan, melphalan or teniposide. The effect of drugs on survivin expression and caspase-3 was investigated by RT-qPCR. The best drug combination was selected measuring the viability, survivin expression and the selectivity of the treatment. Our stepwise method revealed that siRNA delivery by switchable LNP sensitized Y79, but not the healthy APRE-19 cell line, to carboplatin and melphalan cytotoxicity. This ability was validated on primary human RB cells. Finally, the distinct behavior of the drugs demonstrated that a diligent screening of drugs should be envisioned when looking for synergy with survivin. Our sequential approach highlighted carboplatin and melphalan as agents to be investigated in future survivin-associated in vivo testing to tackle RB.

Toward Closing the Gap between Hexoses and N-Acetlyhexosamines: Experimental and Computational Studies on the Collision-Induced Dissociation of Hexosamines.

Motivated by the fundamental difference in the reactivity of hexoses and N-acetylhexosamines under collision-induced dissociation (CID) mass spectrometry conditions, we have investigated the CID of two hexosamines, glucosamine (GlcN) and galactosamine (GalN), experimentally and computationally. Both hexosamines undergo ring-opening and then dissociate via the 0,2A and the 0,3A (0,3X) cross-ring cleavage channels. The preference for the ring-opening is similar to the behavior of N-acetylhexosamines and explains why the two anomers of the same sugar give the same mass spectrum. While the spectrum for GlcN is dominated by the 0,2A signal, the signal intensities for both 0,2A and the 0,3A (0,3X) dissociation channels are comparable for GalN, which allows GlcN and GalN to be distinguished easily. Calculations at MP2 level of theory indicate that this is related to the differences in the relative barrier heights for the 0,2A and the 0,3A (0,3X) cross-ring cleavage channels. This, in return, reflects the circumstance that the 0,2A cross-ring cleavage barriers are different for the two sugars, while the barriers of all other dissociation channels are comparable. While the mechanisms of the cross-ring dissociation channels of hexoses are well described using the retro-aldol mechanism in the literature, this study proposes a new mechanism for the 0,3A (0,3X) cross-ring cleavage of hexosamines that involves the formation of an epoxy intermediate or a zwitterionic intermediate.

Unexpected Dissociation Mechanism of Sodiated N-Acetylglucosamine and N-Acetylgalactosamine.

The mechanism for the collision-induced dissociation (CID) of two sodiated N-acetylhexosamines (HexNAc), N-acetylglucosamine (GlcNAc), and N-acetylgalactosamine (GalNAc), was studied using quantum-chemistry calculations and resonance excitation in a low-pressure linear ion trap. Experimental results show that the major dissociation channel of the isotope labeled [1-18O, D5]-HexNAc is the dehydration by eliminating HDO, where OD comes from the OD group at C3. Dissociation channels of minor importance include the 0,2A cross-ring dissociation. No difference has been observed between the CID spectra of the α- and ß-anomers of the same HexNAc. At variance, the CID spectra of GlcNAc and GalNAc showed some differences, which can be used to distinguish the two structures. It was observed in CID experiments involving disaccharides with a HexNAc at the nonreducing end that a ß-HexNAc shows a larger dissociation branching ratio for the glycosidic bond cleavage than the α-anomer. This finding can be exploited for the rapid identification of the anomeric configuration at the glycosidic bond of HexNAc-R' (R' = sugar) structures. The experimental observations indicating that the dissociation mechanisms of HexNAcs are significantly different from those of hexoses were explained by quantum-chemistry calculations. Calculations show that ring opening is the major channel for HexNAcs in a ring form. After ring opening, dehydration shows the lowest barrier. In contrast, the glycosidic bond cleavage becomes the major channel for HexNAcs at the nonreducing end of a disaccharide. This reaction has a lower barrier for ß-HexNAcs as compared with the barrier of the corresponding α-anomers, consistent with the higher branching ratio for ß-HexNAcs observed in experiment.

Collision-induced dissociation of sodiated glucose, galactose, and mannose, and the identification of anomeric configurations.

Collision-induced dissociation of sodiated α-glucose, ß-glucose, α-galactose, ß-galactose, α-mannose, and ß-mannose was studied using electronic structure calculations and resonance excitation in a low-pressure linear ion trap. We made an extensive search of conformers and transition states in calculations to ensure the transition state with the lowest barrier height for each dissociation channel could be located. The major dissociation channels, in addition to desodiation, are cross-ring dissociation and dehydration. Cross-ring dissociation starts with H atom transfer from the O1 atom to the O0 atom, followed by the cleavage of the C1-O0 bond. Dehydration of the anomer with O1 and O2 atoms in the cis configuration involves the transfer of an H atom from the O2 atom to the O1 atom, followed by the cleavage of the C1-O1 bond. In contrast, dehydration of the anomer with O1 and O2 atoms in the trans configuration mainly occurs through H atom transfer from the O3 or O2 atom to the O1 atom for glucose, from the O3 or O4 atom to the O1 atom for galactose, and from the O4 or O2 atom to the O1 atom for mannose, followed by the cleavage of the C1-O1 bond. The dehydration barrier heights are lower than those of cross-ring dissociation for cis anomers, but higher than those of cross-ring dissociation for trans anomers. The relative barrier heights from calculations are consistent with the experimental measurements of branching ratios. Both computational and experimental results show that the branching ratio of dehydration can be generalized as a simple rule for rapidly identifying the anomeric configurations of these monosaccharides.