Topical Administration of SLN-Based Gene Therapy for the Treatment of Corneal Inflammation by De Novo IL-10 Production

One of the main challenges in gene therapy is the issue of delivery, and it is especially relevant for the success of gene therapy in the cornea. In the present work, eye drops containing biocompatible non-viral vectors based on solid lipid nanoparticles (SLNs) as gene delivery systems to induce the expression of interleukin 10 (IL-10) were designed to address the treatment of corneal inflammation. Two kinds of SLNs combined with different ligands (protamine, dextran, or hyaluronic acid (HA)) and formulated with polyvinyl alcohol (PVA) were prepared. SLN-based vectors were characterized in terms of size, adhesiveness, viscosity, and pH, before topical administration to wild type and IL-10 knock out (KO) mice. The formulations showed a homogenous particle size below 400 nm and a positive surface charge to favor bioadhesion; the incorporation of PVA improved the corneal penetration. After three days of treatment by topical instillation, SLN-based vectors mainly transfected corneal epithelial cells, HA-formulations being the most effective ones. IL-10 was capable of reaching even the endothelial layer. Corneal sections showed no histological change and formulations seemed to be well tolerated after repeated topical administration. These promising results highlight the possible contribution of non-viral gene augmentation therapy to the future clinical approach of corneal gene therapy.


Transmission electronic microscopy (TEM) images
Visualization of SLNC was performed using electron microscopy negative staining. For that purpose, 10 μl of the sample was adhered onto glow discharged carbon coated grids for 60 s. Then the remaining liquid was removed by blotting on filter paper, and the samples were stained with 2% uranyl acetate for 60 s. SLNC were visualized using a Philips EM208S TEM and digital images were acquired on an Olympus SIS purple digital camera. Technical and human support for TEM was provided by the General Service (SGIker) of Analytical Microscopy and High Resolution in Biomedicine at the University of the Basque Country UPV/EHU. TEM images of the SLNEE were previously published [1].
TEM photograph of the SLNC ( Figure S1) showed the spherical shape of the nanoparticles.

Ability of the vectors to bind, protect and release the plasmid DNA
In order to evaluate DNA binding efficacy of the SLNC, as well as their protection and release capacity, a 0.7% agarose gel electrophoresis containing Gel RedTM was employed ( Figure S2). Assessment of the ability of the vectors to bind electrostatically pcDNA3-EGFP and pUNO1-hIL10 plasmids was performed by preparing the complexes at a final concentration of 0.03 µg/µl of DNA in MilliQTM water. This concentration of DNA was also subjected to 1 U DNase I/2.5 µg DNA during 30 minutes at 37ºC to study the protection capacity of the complexes. Finally, the release of DNA from the vectors was performed with a SDS solution (4%) to a final concentration of 1%. As control, naked pcDNA3-EGFP or pUNO1-hIL10 plasmid and 1 kb DNA ladder from NIPPON Genetics Europe (Dueren, Germany) were added. For the analysis of the gel, an Uvitec Uvidoc D-55-LCD-20M Auto transilluminator was used as previously reported [34]. DX-SLNEE and HA-SLNEE were assessed in previous studies [2,3]. Figure S2. Capacity of the vectors to bind, protect and release the pcDNA3-EGFP plasmid (A) and the pUNO1-hIL10 plasmid (B). To study the release of the plasmid, the samples were treated with SDS, and for the protection assay, the samples were mixed first with DNase I and later with SDS to release the plasmid DNA. MW: molecular weight; SDS: sodium dodecyl sulphate; DNase I: deoxyribonuclease I.
As can be seen in lanes 3 and 4, DNA was completely bound to the vectors, since no band was observed on the correspondent lanes, and the plasmid was detected on the loading wells.
The release of the plasmid was evaluated after treating the complexes with SDS for 5 minutes. The DNA bands in lanes 8 and 9 indicate that both vectors were able to release the plasmids, although not completely, since the plasmids were also partially detected in the loading wells, especially in the case of the vectors prepared with the IL-10 plasmid ( Figure S2B).

Rheology results with the plasmid pUNO1-hIL10
Table S1 and Figure S3 summarizes the results obtained in the rheological studies with the plasmid pUNO1-hIL10 and the vectors prepared with it. The results are consistent with those obtained with the plasmid pcDNA3-EGFP. Table S1. High coefficient of determination (R 2 ), viscosity (mPa·s) at shear rate of 10 and 500 s -1 , consistency coefficient (k; Pa·s n ) and flow behaviour index (n) values of vectors and plasmid solutions with and without PVA.

R 2
Viscosity 10 s -1 (mPa·s)  The plasmid solutions, with and without PVA, and HA-SLNC with PVA showed a flow behaviour similar to water (n index near 1), indicating Newtonian behaviour, whereas the values of n were lower in all the SLNEE-based vectors, with and without PVA, and the vector HA-SLNC without PVA This is indicative of a pseudoplastic behaviour.