KR-12 Derivatives Endow Nanocellulose with Antibacterial and Anti-Inflammatory Properties: Role of Conjugation Chemistry

This work combines the wound-healing-related properties of the host defense peptide KR-12 with wood-derived cellulose nanofibrils (CNFs) to obtain bioactive materials, foreseen as a promising solution to treat chronic wounds. Amine coupling through carbodiimide chemistry, thiol-ene click chemistry, and Cu(I)-catalyzed azide-alkyne cycloaddition were investigated as methods to covalently immobilize KR-12 derivatives onto CNFs. The effects of different coupling chemistries on the bioactivity of the KR12-CNF conjugates were evaluated by assessing their antibacterial activities against Escherichia coli and Staphylococcus aureus. Potential cytotoxic effects and the capacity of the materials to modulate the inflammatory response of lipopolysaccharide (LPS)-stimulated RAW 245.6 macrophages were also investigated. The results show that KR-12 endowed CNFs with antibacterial activity against E. coli and exhibited anti-inflammatory properties and those conjugated by thiol-ene chemistry were the most bioactive. This finding is attributed to a favorable peptide conformation and accessibility (as shown by molecular dynamics simulations), driven by the selective chemistry and length of the linker in the conjugate. The results represent an advancement in the development of CNF-based materials for chronic wound care. This study provides new insights into the effect of the conjugation chemistry on the bioactivity of immobilized host defense peptides, which we believe to be of great value for the use of host defense peptides as therapeutic agents.

The distribution of the KR-12 derivatives on the KR12-CNF materials was investigated using scanning electron microscopy (SEM) imaging together with energy dispersive spectroscopy to detect nitrogen. The KR12-CNF suspensions were air-dried, placed on a carbon stub, and coated with a thin layer of gold/palladium with a sputter coater Polaron SC7640 sputter coater (Thermo VG Scientific) to be imaged using a Zeiss LEO 1550 SEM with SE2 detector and an energy dispersive detector EDS (Carl Zeiss Microscopy, Oberkochen Germany).

Molecular Dynamics Simulations: Model and simulation
System preparation. To prepare the linkers for molecular dynamics simulations, models (maleimide, triazole and carboxymethyl units) were firstly built by GaussView software and optimized using the Gaussian 16 package.
[1] Frequency analyses were also performed to ensure that each structure was a true minimum (N imag = 0). Then, the conformational space was evaluated for each structure using the MacroModel module included in the Maestro software. [2] For this purpose, the MMFFs force field [3] was applied including water as solvent. An extended level of torsion sample options was selected and a maximum of 20 structures were saved for each fragment.
Each conformation was subsequently optimized at B3LYP or HF/6-31G* level of theory. Diffuse functions were employed for charged fragments. This procedure allows to overcome issues about atomic partial charges associated to conformation dependency. Two different schemes were accounted to be consistent with AMBER and GLYCAM [4] parametrizations. For the carboxymethyl unit, the employed scheme involves the calculation of the atomic partial charges at HF/6-31G* level and subsequent one-stage RESP fitting (CHELPG method) with a charge restraint weight of 0.01 thus making equivalent the aliphatic protons. For triazole and maleimide fragments, a two-stage fitting was employed also assigning the same partial charge on aliphatic protons (atomic charge equilibration predicted by atom paths). These fitting procedures were carried out based on LIBRETA library implemented in Multiwfn software. [5] Finally, parameters for these linkers were generated following an in-house-developed protocol for combining GLYCAM_06j-1 and gaff2 force fields in a semiautomatic manner System equilibration and MD simulations. The Amber20 software [6] with GLYCAM_06j-1, ff14SB [7] and gaff2 [8] force field parameters were applied for cellulose, peptides and linkers, respectively. The initial structures were neutralized with either Na + or Cl − ions and set at the center of a cubic TIP3P [9] water box with a buffering distance between the solute and box of 10 Å. Three steepest descent/conjugate gradient minimizations (maximum 5000 steps) were carried out on the initial structures of all systems. In the first minimization, the positions of cellulose, peptides and ions were restrained using Harmonic potential restraints with a force constant k of 100 kcal mol -1 Å -2 . Subsequent minimizations were performed without any restraints. After minimization, the system was gradually heated from 0 to 300 K during 100 ps (NVT assemble) under the Andersen temperature coupling [10] scheme. Assuming periodic boundary conditions (PBC), the particle mesh Ewald method was employed for modelling long-range electrostatic interactions whereas the cut-off for non-bonded interactions was set to 8 Å. [11] The SHAKE algorithm [12] was also used to constraint covalent bonds involving hydrogen atoms. Then, the system was equilibrated for 2 ns at a constant volume and temperature of 300 K. The final snapshot of this equilibration was used as jumping-off point for MD production under the same conditions. Two consecutive MD simulations (100 ns each) was run to obtain the final production trajectory. Three independent replicas were carried out for each system.
Data analysis. Solvent Accessible Surface Area (SASA) was calculated along the simulation by using the cpptraj module. [13] Raw data were referenced to the maximum value (in Å 2 ) obtained for the free ligand in a water box (100% SASA). Then, a bisquare smoothing algorithm with a sampling proportion of 0.1 was applied for the final plot representation. The secondary structure for each peptide was calculated using the DSSP algorithm [14] as implemented in the cpptraj module.        whereas C is represented as blank space). Letter codes: "G", 3-10 helix; "H", alpha helix; "I", πhelix; "E", β-sheet; "B", β-bridge; "S", bend; "T", helix turn and "C", coil.