(637i) Development of Coarse-Grained Forcefields for the Self-Assembly of Ceramide-Based Lipid Membranes

Ceramides are a crucial component of the lipid lamella in the stratum corneum (SC) layer of mammalian skin, where barrier function is known to be localized [1]. Studying the SC lipid membrane can be challenging, as it is composed of at least 14 unique ceramides, along with cholesterol and free fatty acid of various lengths. Even when considering simplified mixtures that contain only a subset of these lipids, as are commonly studied via both experiment and simulation, challenges still exist. While most biological membranes form fluid-like bilayers with high lipid mobility, lipids in the SC instead form dense, gel-like multi-layer structures with very low lipid mobility [2]. As such, accurate modeling of membranes formed by SC lipids requires large simulation sizes to capture multilayer behavior and very long simulation times to allow the lipids to reach their equilibrium morphologies. This is further complicated by the fact that it is generally agreed upon that at least some ceramide lipids span multiple leaflets, making the use of pre-assembled multilayer configurations unlikely to be representative of experiment. In order to capture the larger length and timescales needed to represent the SC lamella, coarse-grained (CG) forcefields that allow for the self-assembly of large multilayer structures are necessary.

Here, we report our work to develop an accurate, transferable CG forcefield for studying the key lipids in the SC [3-7], optimized to match experimental data and the atomistic CHARMM forcefield [8,9]. This work utilizes the multistate iterative Boltzmann inversion (MS-IBI) method [10] combined with simulated wetting simulations to achieve the correct balance between hydrophobic and hydrophilic interactions [3]. MS-IBI is an extension of the original, single-state IBI methodology [11] commonly used to self-consistently optimize CG forcefields by matching the radial distribution function (RDF) of a CG model with that of the target system (typically a system simulated using atomistic models). MS-IBI builds upon the IBI approach by optimizing a single forcefield to match the RDFs from several thermodynamic states simultaneously; this methodology also allows a single forcefield to be optimized using different molecules with shared chemical topologies as target states. By using multiple states and molecules, MS-IBI has been shown to reduce structural artifacts and increase transferability between phases and molecules; this methodology also allows using multiple ensembles simultaneously, e.g., NVT and NPT, enabling the density-pressure relationship to be accurately captured.

Using the MS-IBI approach, forcefields for ceramide NS, free-fatty acid and cholesterol that provide close agreement with key structural properties from atomistic simulations and experiment, when available, have been developed [3-6]. The models are also able to reproducibly self-assemble multilayer structures. We further demonstrate the transferability of the SC forcefield developed using this approach by applying the models optimized for ceramide NS and FFA to other ceramides, namely NP, AP, and AS, finding close agreement with atomistic CHARMM simulations and experiment, without the need to reoptimize the non-bonded parameters [7]. To improve the reproducibility and ability to disseminate the forcefield, forcefield files are prepared that are compatible with the general purpose Foyer atom-typing library [12] as part of the Molecular Simulation and Design Framework (MoSDeF) [13] .

[1] Bouwstra, J.A. and Ponec, M., 2006. The skin barrier in healthy and diseased state. Biochimica et Biophysica Acta (BBA)-Biomembranes, 1758(12), pp.2080-2095.

[2] Moore, T.C., Hartkamp, R., Iacovella, C.R., Bunge, A.L. and McCabe, C., 2018. Effect of ceramide tail length on the structure of model stratum corneum lipid bilayers. Biophysical journal, 114(1), pp.113-125.

[3] Moore, T.C., Iacovella, C.R., Hartkamp, R., Bunge, A.L. and McCabe, C., 2016. A coarse-grained model of stratum corneum lipids: free fatty acids and ceramide NS. The Journal of Physical Chemistry B, 120(37), pp.9944-9958.

[4] Moore, T.C., Iacovella, C.R., Leonhard, A.C., Bunge, A.L. and McCabe, C., 2018. Molecular dynamics simulations of stratum corneum lipid mixtures: A multiscale perspective. Biochemical and biophysical research communications, 498(2), pp.313-318.

[5] Moore, T.C., Iacovella, C.R. and McCabe, C., 2016. Development of a coarse-grained water forcefield via multistate iterative Boltzmann inversion. In Foundations of Molecular Modeling and Simulation (pp. 37-52). Springer, Singapore.

[6] Moore, T.C. Shamaprasad P, Iacovella C.R., Bunge A.L., McCabe C., Multiscale Simulations of Stratum Corneum Lipid Mixtures, in preparation

[7] Frame, C.O., Shamaprasad P, Iacovella C.R., Bunge A.L., McCabe C., Transferrable models for the Simulation of Stratum Corneum lipids, in preparation

[8] Klauda, J.B., Venable, R.M., Freites, J.A., O’Connor, J.W., Tobias, D.J., Mondragon-Ramirez, C., Vorobyov, I., MacKerell Jr, A.D. and Pastor, R.W., 2010. Update of the CHARMM all-atom additive force field for lipids: validation on six lipid types. The journal of physical chemistry B, 114(23), pp.7830-7843.

[9] Guo, S., Moore, T.C., Iacovella, C.R., Strickland, L.A. and McCabe, C., 2013. Simulation study of the structure and phase behavior of ceramide bilayers and the role of lipid headgroup chemistry. Journal of chemical theory and computation, 9(11), pp.5116-5126.

[10] Moore, T.C., Iacovella, C.R. and McCabe, C., 2014. Derivation of coarse-grained potentials via multistate iterative Boltzmann inversion. The Journal of chemical physics, 140(22), p.06B606_1.

[11] Reith, D., Pütz, M. and Müller‐Plathe, F., 2003. Deriving effective mesoscale potentials from atomistic simulations. Journal of computational chemistry, 24(13), pp.1624-1636.

[12] Klein, C., Summers, A.Z., Thompson, M.W., Gilmer, J.B., McCabe, C., Cummings, P.T., Sallai, J. and Iacovella, C.R., 2019. Formalizing atom-typing and the dissemination of force fields with foyer. Computational Materials Science, 167, pp.215-227.

[13] Summers, A.Z., Gilmer, J.B., Iacovella, C.R., Cummings, P.T. and McCabe, C., 2020. MoSDeF, a python framework enabling large-scale computational screening of soft matter: Application to chemistry-property relationships in lubricating monolayer films. Journal of chemical theory and computation, 16(3), pp.1779-1793. https://mosdef.org