Modeling pyranose ring pucker in carbohydrates using machine learning and semi-empirical quantum chemical methods.

Pyranose ring pucker is a key coordinate governing the structure, interactions and reactivity of carbohydrates. We assess the ability of the machine learning potentials, ANI-1ccx and ANI-2x, and the GFN2-xTB semiempirical quantum chemical method, to model ring pucker conformers of five monosaccharides and oxane in the gas phase. Relative to coupled-cluster quantum mechanical calculations, we find that ANI-1ccx most accurately reproduces the ring pucker energy landscape for these molecules, with a correlation coefficient r2 of 0.83. This correlation in relative energies lowers to values of 0.70 for ANI-2x and 0.60 for GFN2-xTB. The ANI-1ccx also provides the most accurate estimate of the energetics of the 4 C1 -to-1 C4 minimum energy pathway for the six molecules. All three models reproduce chair more accurately than non-chair geometries. Analysis of small model molecules suggests that the ANI-1ccx model favors puckers with equatorial hydrogen bonding substituents; that ANI-2x and GFN2-xTB models overstabilize conformers with axially oriented groups; and that the endo-anomeric effect is overestimated by the machine learning models and underestimated via the GFN2-xTB method. While the pucker conformers considered in this study correspond to a gas phase environment, the accuracy and computational efficiency of the ANI-1ccx approach in modeling ring pucker in vacuo provides a promising basis for future evaluation and application to condensed phase environments.

Constructing 3-Dimensional Atomic-Resolution Models of Nonsulfated Glycosaminoglycans with Arbitrary Lengths Using Conformations from Molecular Dynamics.

Glycosaminoglycans (GAGs) are the linear carbohydrate components of proteoglycans (PGs) and are key mediators in the bioactivity of PGs in animal tissue. GAGs are heterogeneous, conformationally complex, and polydisperse, containing up to 200 monosaccharide units. These complexities make studying GAG conformation a challenge for existing experimental and computational methods. We previously described an algorithm we developed that applies conformational parameters (i.e., all bond lengths, bond angles, and dihedral angles) from molecular dynamics (MD) simulations of nonsulfated chondroitin GAG 20-mers to construct 3-D atomic-resolution models of nonsulfated chondroitin GAGs of arbitrary length. In the current study, we applied our algorithm to other GAGs, including hyaluronan and nonsulfated forms of dermatan, keratan, and heparan and expanded our database of MD-generated GAG conformations. Here, we show that individual glycosidic linkages and monosaccharide rings in 10- and 20-mers of hyaluronan and nonsulfated dermatan, keratan, and heparan behave randomly and independently in MD simulation and, therefore, using a database of MD-generated 20-mer conformations, that our algorithm can construct conformational ensembles of 10- and 20-mers of various GAG types that accurately represent the backbone flexibility seen in MD simulations. Furthermore, our algorithm efficiently constructs conformational ensembles of GAG 200-mers that we would reasonably expect from MD simulations.

Switchable proline derivatives: tuning the conformational stability of the collagen triple helix by pH changes.

(4S)-Aminoproline is introduced as a pH-sensitive probe for tuning the conformational properties of peptides and proteins. The pH-triggered flip of the ring puckering and the formation/release of a transannular H bond were used to switch the formation of collagen triple helices on and off reversibly.

Consequences of incorporating thiaproline and its oxidized derivatives into collagen triple helices.

(2R)-4-thiaproline (Thp) is an analog of proline, replacing Cγ in the pyrrolidine ring with sulfur. Its thiazolidine ring easily interconverts between endo and exo puckers due to a small energy barrier, which leads to destabilize polyproline helices. Collagen, composed of three polyproline II helices, mainly consists of X-Y-Gly triplets, where X is often proline and Y is frequently (2S,4R)-hydroxyproline. In this study, we incorporated Thp into either position-X or position-Y to investigate the consequences of such a replacement on the triple helix. Circular dichroism and differential scanning calorimetry analyses showed that the Thp-containing collagen-mimetic peptides (CMPs) can fold into stable triple helices, in which the substitution at position-Y exhibits a larger destabilization effect. Additionally, we also prepared the derivative peptides by oxidizing Thp in the peptide to N-formyl-cysteine or S,S-dioxide Thp. The results showed that the oxidized derivatives at position-X only slightly affect collagen stability, but those at position-Y induce a large destabilization effect. The consequences of incorporating Thp and its oxidized derivatives into CMPs are position dependent. Computational results suggested that the ease of interconversion between exo and endo puckers for Thp and the twist conformation of S,S-dioxide Thp may cause the destabilization effect at position-Y. We have revealed new insights into the impacts of Thp and its oxidized derivatives on collagen and demonstrated that Thp can be used to design collagen-related biomaterials.

Efficient Construction of Atomic-Resolution Models of Non-Sulfated Chondroitin Glycosaminoglycan Using Molecular Dynamics Data.

Glycosaminoglycans (GAGs) are linear, structurally diverse, conformationally complex carbohydrate polymers that may contain up to 200 monosaccharides. These characteristics present a challenge for studying GAG conformational thermodynamics at atomic resolution using existing experimental methods. Molecular dynamics (MD) simulations can overcome this challenge but are only feasible for short GAG polymers. To address this problem, we developed an algorithm that applies all conformational parameters contributing to GAG backbone flexibility (i.e., bond lengths, bond angles, and dihedral angles) from unbiased all-atom explicit-solvent MD simulations of short GAG polymers to rapidly construct models of GAGs of arbitrary length. The algorithm was used to generate non-sulfated chondroitin 10- and 20-mer ensembles which were compared to MD-generated ensembles for internal validation. End-to-end distance distributions in constructed and MD-generated ensembles have minimal differences, suggesting that our algorithm produces conformational ensembles that mimic the backbone flexibility seen in simulation. Non-sulfated chondroitin 100- and 200-mer ensembles were constructed within a day, demonstrating the efficiency of the algorithm and reduction in time and computational cost compared to simulation.