Recent success in the treatment of congenital blindness demonstrates the potential of ocular gene therapy as a therapeutic approach. The eye is a good target due to its small size, minimal diffusion of therapeutic agent to the systemic circulation, and low immune and inflammatory responses. Currently, most approaches are based on viral vectors, but efforts continue towards the synthesis and evaluation of new nonviral carriers to improve nucleic acid delivery. Our objective is to evaluate the efficiency of novel cationic retinoic and carotenoic glycol phospholipids, designated C20-18, C20-20, and C30-20, to deliver DNA to human retinal pigmented epithelium (RPE) cells. Liposomes were produced by solvent evaporation of ethanolic mixtures of the polyene compounds and coformulated with 1,2-dioleoyl-
Gene therapy is a promising treatment of several pathologies [
Successful gene therapy requires nucleic acid carriers of minimal toxicity allowing long and stable gene expression, such as adenovirus and lentivirus [
Nonviral carriers are typically cationic polymers or liposomes. Cationic liposomes can easily be produced in large scale [
Cationic lipids interact with negatively charged nucleic acids through a combination of electrostatic and hydrophobic interactions, resulting in lipoplexes of multilamellar structures [
The transfection efficiency and biocompatibility of cationic lipid gene carriers can be altered by modifications of the constituent parts of the lipids, specifically, the lipid backbone (often glycerol), the hydrophilic head group, and the interconnecting linker. Less diversity is reported for the hydrophobic part, which is known to play a significant role in the morphology of the lipoplex [
Dietary carotenoids are known to enhance visual performance and to reduce the risk of AMD progression [
C20-18, C20-20, and C30-20 cationic amphiphilic glycol polyene phospholipids. The color of the polyene chain indicates approximately the visually appearance of the compounds. Glycol backbone green, hydrophilic part blue.
Our objective was to evaluate the gene delivery efficiency and cytotoxicity of the cationic polyene glycol lipids against the reference glycerol lipids 3
Reference compounds. DC-Chol and cationic glycerophosphospholipid EPC with two C14:0 chains. Glycerol backbone green, hydrophilic parts blue.
The polyene lipids were formulated with either cholesterol (Chol) or 1,2-dioleoyl-
Colipids. Neutral cholesterol (Chol) and zwitterionic glycerophosphospholipid DOPE with two C18:1 chains. Glycerol backbone green, hydrophilic parts blue.
The ARPE-19 cell line, a human retinal pigment epithelial cell line, was kindly provided by Dr. Franscisco Ambrósio (Center for Neuroscience and Cell Biology, University of Coimbra, Portugal); Dulbecco’s modified eagle medium (DMEM) culture medium and Dulbecco’s modified eagle medium F12 ham (DMEM F12 Ham) culture medium, fetal bovine serum (FBS), trypsin, glutamine, penicillin-streptomycin solution, dichloromethane, thiazolyl blue tetrazolium bromide (MTT), 0.04 N HCl in 2-propanol constituents, dd-water, tris-acetate-EDTA (TAE) constituents, Avertin anesthesia constituents (tribromoethanol), Triton X-100, Mowiol mounting media constituents, sucrose (saccharose), paraformaldehyde, eosin, acetic acid, dibutyl phthalate xylene (DPX), sodium phosphate, potassium phosphate, and goat serum stock were purchased from Sigma-Aldrich (Portugal). Absolute ethanol was purchased from Merck Millipore (Portugal).
Plasmid DNA containing the reporter gene green fluorescent protein (GFP) was kindly provided by Dr. Jean Bennett (University of Pennsylvania, USA). Agarose was purchased from Invitrogen (Portugal). GreenSafe and GeneJuice were purchased from NZYTech (Portugal). For the phosphate buffered saline (PBS), sodium chloride, potassium chloride, 4′,6-diamidino-2-phenylindole (DAPI), and optimal cutting temperature (OCT) cryostat embedding solution were purchased from VWR (Portugal). Primary antibody Iba1, a microglia marker, was purchased from Wako Pure Chemical Industries (Japan) and secondary antibody Alexa 594 was purchased from Invitrogen (Portugal). Oxalic acid, potassium permanganate (KMnO4), and hematoxylin were purchased from Merck (Portugal). Control cationic lipids 3
Three cationic polyene lipids, C20-18, C20-20, and C30-20, were synthesized from glycol, retinoic acid (C20:5) or C30-acid (C30:9), and choline precursors as reported elsewhere [
Ethanolic stock solutions were made separately by first dissolving each cationic lipid and colipid in dichloromethane. The dichloromethane was removed under reduced pressure with a rotary evaporator, and then the residue dissolved in absolute ethanol to a final lipid concentration of 1 mM and subsequently stored at −80°C.
Cholesterol and DOPE were employed as neutral colipids [
The addition of predetermined volumes of negatively charged plasmid DNA to positively charged liposomes resulted in lipoplex formulations of defined nitrogen-to-phosphorus (N/P) molar charge ratios. DNA contained the reporter gene GFP. Lipid concentrations, derived from a 2 mM hydrated stock, were 0.243 mM, 0.162 mM, and 0.081 mM, which correlates with the molar charge ratios of 1.5 : 1, 1 : 1, and 0.5 : 1, respectively. Equal volumes of DNA solution in DMEM or DMEM/F12 Ham culture medium and lipids were mixed and incubated at room temperature for 30 min to allow lipoplex formation.
A gel retardation assay was used to evaluate lipoplex formation, as a neutral or net positive charge associated with the lipid-DNA complex retards migration through agarose gel. Lipoplexes with varying lipid : DNA ratios were prepared as described above in a total volume of 20
Lipoplexes were prepared in phosphate buffered solution (PBS), for subsequent determination of size and zeta potential by dynamic light scattering (DLS) and laser Doppler anemometry, respectively, using a Zetasizer Nanoseries ZS (Malvern Instruments) by diluting the sample 50x with dd-water.
The size and size distribution of lipoplexes are indicated as displayed by the instrument. Measurement parameters were as follows: a laser wavelength of 633 nm and a scattering angle of 173° (fixed). The samples were loaded into polystyrene cuvettes and three measurements were performed, for which the mean result was recorded. Particles with PdI > 0.7 and
D407 cells were maintained in DMEM culture medium supplemented with 5% FBS, 1% glutamine, and 1% penicillin-streptomycin solution at 37°C with 5% CO2. ARPE-19 cells were maintained in DMEM F12 HAM culture medium supplemented with 10% FBS, 1% glutamine, and 1% penicillin/streptomycin at 37°C with 5% CO2. For the cytotoxicity assays, cells were seeded at 35,000 cells per well (D407) or 65,000 cells per well (ARPE-19) in 24-well plates in complete medium and incubated for 24 h at 37°C and 5% CO2. Cell seeding numbers differ with cell line due to differences in cell doubling rate. After 24 h, cells were washed with PBS and 250
Transfection was determined using cells seeded at 100,000 cells per well (D407) or 250,000 cells per well (ARPE-19) in 6-well plates in complete growth medium and incubated at 37°C with 5% CO2 for 24 h. Cell seeding numbers differ with cell line due to differences in cell doubling rate. Immediately prior to transfection, cell monolayers were washed with PBS. The lipoplex formulations at various N/P molar charge ratios were added to each well in culture medium up to 2 mL and incubated for 4 h at 37°C, 5% CO2. After this period, the medium containing the lipoplexes was removed, the cells were washed with PBS, and complete growth medium was added and further incubated for 72 h at 37°C, 5% CO2. Subsequently, the cells were suspended using trypsin, washed with PBS, and resuspended in 500
The
All figures are representative of at least 3 separate experiments. For comparison of multiple sets of data one-way ANOVA, including the Dunnet post-test, was performed. Results are expressed as mean +/− SEM. Statistical analysis was performed using GraphPad Prism 6 Software with
The size and surface charge of lipid-DNA complexes composed of the polyene lipids were determined using Dynamic Light Scattering (DLS) (Table
Size and surface charge (as measured by zeta potential, ZP) of C20-18, C20-20, C30-20, DC-Chol, and EPC lipoplexes in PBS at varying N/P molar charge ratios with either Chol or DOPE as colipids.
Lipid | Colipid |
|
|
Mode (nm) | Pdl | ZP (mV) |
---|---|---|---|---|---|---|
|
|
1180 | 477 | 0.718 |
|
|
|
|
1480 | 455 | 0.840† |
|
|
|
1220 | 516 | 0.611 |
|
||
|
900 | 441 | 0.664 |
|
||
|
|
1310 | 600 | 0.424 |
|
|
|
1290 | 506 | 0.768* |
|
||
|
||||||
|
|
510† | 488 | 0.220† |
|
|
|
|
360 | 380 | 0.058 |
|
|
|
320† | 322† | 0.234† |
|
||
|
1410 | 528 | 0.750 |
|
||
|
|
930 | 625 | 0.366 |
|
|
|
2010* | 656 | 0.862 |
|
||
|
||||||
|
|
320 | 256 | 0.322 |
|
|
|
|
390 | 361 | 0.285 |
|
|
|
1523* | 490 | 0.702* |
|
||
|
550 | 487 | 0.425 |
|
||
|
|
910 | 406 | 0.624 |
|
|
|
1270 | 453 | 0.596 |
|
||
|
||||||
|
|
560 | 548 | 0.130 |
|
|
|
|
570 | 537 | 0.157 |
|
|
|
660 | 537 | 0.257 |
|
||
|
640 | 465 | 0.304 |
|
||
|
|
970 | 500 | 0.155 |
|
|
|
750 | 594 | 0.147 |
|
||
|
||||||
|
|
500† | 355*† | 0.353 |
|
|
|
|
820 | 763 | 0.178 |
|
|
|
710 | 697 | 0.260 |
|
||
|
1500 | 937* | 0.144 |
|
||
|
|
970 | 605 | 0.419 |
|
|
|
1140 | 768 | 0.220 |
|
The surface charge values (as measured by zeta potential, ZP, Table
The cationic liposomes were combined with a GFP-coded plasmid DNA to evaluate relative gene transfer efficiency to RPE cells. The DNA binding efficiency of C20-18, C20-20, C30-20, DC-Chol, and EPC-containing lipoplexes at various N/P molar charge ratios was assessed using a gel retardation assay. The negatively charged plasmid DNA, when interacting with sufficient cationic lipids, forms a neutral or net positive complex that will not migrate through the agarose gel. As seen in Figure
Gel retardation assays with lipoplexes C20-18 (a), C20-20 and C30-20 (b), DC-Chol (c), and EPC (d) with DOPE and Chol as colipids. Molar charge ratios N/P used were 0.5 : 1, 1 : 1, and 1.5 : 1. Retention of DNA increases with increasing molar charge ratios.
ARPE-19 cells were cultured in the presence of C20-18, C20-20, C30-20, DC-Chol, and EPC lipoplexes at varying N/P molar charge ratios, and cell viability was assessed after 24, 48, and 72 h (Figures
DC-Chol lipoplex cytotoxicity. ARPE-19 cell viability for DC-Chol lipoplexes with either Chol or DOPE as colipid at various N/P ratios incubated up to 72 h. Horizontal line at 75% viability represent the threshold according to the ISO standard for
EPC lipoplex cytotoxicity. ARPE-19 cell viability for EPC lipoplexes with either Chol or DOPE as colipids at various N/P molar charge ratios incubated up to 72 h. Horizontal line at 75% viability represents the threshold according to the ISO standard for
C20-18 lipoplex cytotoxicity. ARPE-19 cell viability for C20-18 lipoplexes with either Chol or DOPE as colipid at various N/P ratios incubated up to 72 h. Horizontal line at 75% viability represents the threshold according to the ISO standard for
C20-20 lipoplex cytotoxicity. ARPE-19 cell viability for C20-20 lipoplexes with either Chol or DOPE as colipid at various N/P ratios incubated up to 72 h. Horizontal line at 75% viability represents the threshold according to the ISO standard for
C30-20 lipoplex cytotoxicity. ARPE-19 cell viability for C30-20 lipoplexes with either Chol or DOPE as colipid at various N/P ratios incubated up to 72 h. Horizontal line at 75% viability represents the threshold according to the ISO standard for
Toxicity was concentration dependent (increasing with increased N/P molar charge ratio) and time dependent (duration of the incubation period) as previously described [
Based on the cytotoxicity results, the 0.5 : 1 N/P molar charge ratio was subsequently used in transfection assays and
To evaluate whether the level of toxicity observed
Immunohistochemistry detection of Iba1 in mouse retinas injected with C20-20/DOPE lipoplexes at the N/P molar charge ratio 0.5 : 1. Magnification: 100x and scale bar: 50
Administration of drugs by intravitreal injection can result in morphological disruption of the retinal structure, caused by the injection procedure and tissue inflammation. Hematoxylin and eosin staining (Figure
Hematoxylin and eosin staining of mouse retinas injected with C20-20/DOPE lipoplexes at the N/P molar charge ratio 0.5 : 1, showing the maintenance of the integrity of the retinal layered structure. Magnification: 100x and scale bar: 50
We investigated the
Qualitative assessment of lipoplex transfection. ARPE-19 cells were incubated with lipoplexes containing lipids C20-18, C20-20, or C30-20 against reference lipids DC-Chol and EPC with either Chol or DOPE as colipid at N/P molar ratio 0.5 : 1 for 4 h and GFP-expressing cells evaluated by fluorescence microscopy after 72 h. Magnification: 100x and scale bar: 100
The transfection efficiency of lipoplexes. ARPE-19 cells were incubated with lipoplexes containing lipids C20-18, C20-20, or C30-20 against reference lipids DC-Chol and EPC with either Chol or DOPE as colipid at N/P molar ratio 0.5 : 1 for 4 h and transfection efficiencies, measured by GFP expression, determined by flow cytometry after 72 h. Statistical significance (
The transfection efficiency of C20-20 and C30-20 lipoplexes at N/P molar charge ratio of 0.5 : 1 containing either Chol or DOPE as colipid was also assessed in D407 cells against the positive control, GeneJuice (supplementary Figure S1). As was found with the ARPE-19 cell line, the C20-20/DOPE formulation outperformed the corresponding Chol formulation in the D407 cell line (supplementary Figure S1). In addition, C20-20/DOPE revealed a level of transfection competitive to the control GeneJuice, whereas the C30-20 lipid revealed only a low level of transfection with Chol as colipid.
Cationic polyene lipids were evaluated for their potential to carry DNA into retinal cells. Transfection efficiency was correlated with molecular structure, lipoplex size, zeta potential, biocompatibility, and the presence of colipids. DOPE and cholesterol increase DNA delivery and thus were employed as colipids. DOPE, often referred to as a “fusogenic lipid,” could alter the packing of the hydrophobic chains within the lipoplex [
The analysis of the physical properties of the lipoplexes, summarized in Table
Almofti et al. demonstrated for the cationic lipid DC-6-14 that lipoplex size increased lipoplex transfection [
Zeta potential reflects the surface charge and lipoplex stability:
Our lipoplexes presented negative zeta potential values between
Negative ZP values most likely derive from DNA bound to the outer shell of the lipoplexes and from the phosphate groups associated with the solvent (PBS) that remains close to the surface of the lipoplexes.
Successful gene carriers do not only transfect efficiently, but they also need to be biocompatible with the organism and target tissue.
The efficiency of
Although transfection in ARPE-19 cells was higher for the DC-Chol/DOPE and EPC/DOPE lipoplexes, transfection efficiencies of the C20-18/DOPE and C20-20/DOPE lipoplexes are noteworthy when compared to other investigations reporting considerably lower efficiencies [
In an attempt to further compare the relative transfection efficiency of C20-20 and C30-20, we tested the same formulations in D407 cells (supplementary Figure S1), another cell line from the retina, but with a faster mitotic rate than ARPE-19. We observed with these cells a greater transfection efficiency than with ARPE-19 cells, which is most likely due to the higher mitotic rate, since during mitosis the nuclear membrane is disaggregated and therefore nuclear penetration of nucleic acids is facilitated.
Remarkably, C20-20 displayed greater transgene efficiency than C30-20 in both cell lines tested. The retinoic lipids C20-18 and C20-20 may therefor enter the nucleus via a possible retinoid receptor [
C20-18/Chol and C20-20/Chol performed comparably to references DC-Chol and EPC. C20-20/DOPE lipoplexes delivered DNA
Cationic liposomes formulated from polyene lipids C20-18, C20-20, and C30-20 and a neutral colipid were combined with a GFP-coded plasmid DNA for studies in retina cells. Gel electrophoresis was used to confirm the formation of lipoplexes in all formulations containing either DOPE or Chol as colipid. The diameters of lipoplexes were relatively large and variable; however, this did not appear to hinder transfection, since for a given cationic lipid larger lipoplexes gave rise to greater transfection (Table
C20-20 was found to have greater
Overall, C20-20/DOPE gave promising results for ocular gene therapy. Future work is directed to optimize formulations for
The authors declare that there is no conflict of interests regarding the publication of this paper.
The ARPE-19 cell line, a human retinal pigment epithelial cell line, was kindly provided by Dr. Franscisco Ambrósio (Center for Neuroscience and Cell Biology, University of Coimbra, Portugal); Plasmid DNA containing the reporter gene green fluorescent protein (GFP) and D407 cell line, a human retinal pigment epithelial cell line, was kindly provided by Dr. Jean Bennett (University of Pennsylvania, USA). This publication was made possible by grants from The FP7 program under the Marie Curie Scheme PIRG-GA-2009-249314; FCT Portugal (IBB/LA; PEst-OE/EQB/LA0023/2011; SFRH/BD/76873/2011 to Sofia Calado, SFRH/BD/70318/2010 to Ana Vanessa Oliveira, and PTDC/SAU-BEB/0984754/2008 to Gabriela A. Silva) and from the Qatar National Research Fund under the National Priorities Research Program, award NPRP08-705-3-144 (LPI: M. Pungente). Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the Qatar National Research Fund.