Investigation on the Pathway of Disulfide Bonds Reduction on Human Defensins Using Molecular Dynamics Simulations

Defensins are cationic cysteine-rich small molecules with molecular masses ranging from 3 to 5 kDa. They are a critical part of the innate immune system that provides an initial antimicrobial barrier for mucosal surfaces such as the surface of the eyes, airways, lungs, and skin. They are known as ‘natural antibiotics’ as they have broad spectrum antimicrobial activity against both Gram-positive and Gram-negative bacteria, and even against bacteria that have demonstrated resistance to the currently used antibiotics. Mammalian defensins are classified into a, b, q categories based on their size and disulfide bonding pattern. Human a defensin type 5 (HD5) is mainly produced by Paneth cells in the crypts of the small intestine. The peptide has 6 cysteine residues forming 3 disulfide bridges: Cys3-Cys31, Cys5-Cys20, Cys10-Cys30. Human b defensins (HBD) are produced by epithelial cells, for example in gut and lung. HBD type 3 (HBD-3) is a newly discovered defensin, which also has 6 cysteine residues, and forms three disulfide bridges: Cys11-Cys40, Cys18-Cys33, and Cys23-Cys41.

The disulfide bonds of defensins break at reduced condition, and human defensins show diverse activity dependence on their disulfide bridge structure. For example, for human b defensin type 3 (HBD-3), its disulfide bridge status doesn’t seem to affect its anti-bacterial activity, but does have influences on its chemotactic activity; while for HD-5, reduction of all three disulfide bridges may contribute to its wide variety of antimicrobial, antivirial and immune modulating activities. In order to understand the different effects of disulfide bridges on functional activities of different defensins, it is important to find out the pathway of disulfide bonds reduction at reduced condition.

Since there is only experimental data available for HD5 disulfide bridge reduction pathway, a simulation strategy including performing all atom molecular dynamics simulations using both CHARMM and NAMD programs was carried out on HD5 first. The disulfide bond reduction pathway on HD5 was analyzed, which predicted results consistent with available experimental data. Using the same simulation strategy, the disulfide bond breaking order on HBD-3 was predicted. It was found out that the majority of HD5 unfolds by initial reduction of C5-C20, followed by C10-30 and C3-C31, while under the reduced condition, the HBD-3 disulfide bond on C18-C33 would always break first. Those results can shed light on understanding the different functional activity dependence on the disulfide bridge status for HD5 and HBD-3.