Abstract
Abstract
This review highlights the application of lipid nanoparticles (Solid Lipid Nanoparticles, Nanostructured Lipid Carriers, or Lipid Drug Conjugates) as effective drug/gene delivery systems for retinal diseases. Most drug products for ocular disease treatment are marketed as eye drop formulations but, due to ocular barriers, the drug concentration in the retina hardly ever turns out to be effective. Up to this date, several delivery systems have been designed to deliver drugs to the retina. Drug delivery strategies may be classified into 3 groups: noninvasive techniques, implants, and colloidal carriers. The best known systems for drug delivery to the posterior eye are intravitreal implants; in fact, some of them are being clinically used. However, their long-term accumulation might impact the patient's vision. On the contrary, colloidal drug delivery systems (microparticles, liposomes, or nanoparticles) can be easily administered in a liquid form. Nanoparticular systems diffuse rapidly and are better internalized in ocular tissues than microparticles. In comparison with liposomes, nanoparticles have a higher loading capacity and are more stable in biological fluids and during storage. In addition, their capacity to adhere to the ocular surface and interact with the endothelium makes these drug delivery systems interesting as new therapeutic tools in ophthalmology. Within the group of nanoparticles, those composed of lipids (Solid Lipid Nanoparticles, Nanostructred Lipid Carriers, and Lipid Drug Conjugates) are more biocompatible, easy to produce at large scale, and they may be autoclaved or sterilized. The present review summarizes scientific results that evidence the potential application of lipid nanoparticles as drug delivery systems for the retina and also as nonviral vectors in gene therapy of retina disorders, although much more effort is still needed before these lipidic systems could be available in the market.
Introduction

Layers of the retina.
The impulse received by nerve cells is originated in photoreceptors. These cells are classified in cones and rods; the former are located in the center of the retina forming the macula and are responsible for color vision; the latter are in the periphery of the retina and are associated with black and white vision. Photoreceptors are in contact with RPE, a monolayer of cells belonging to the blood–retinal barrier, and have several functions related to the absorption of excess of light, epithelial transport, phagocytosis, secretion, or visual cycle. These 2 layers of cells form the outer retina.
Ocular diseases include minor troubles, such as conjunctivitis, but also vision-threatening disorders typically affecting the retina, in the posterior segment of the eye. These include glaucoma, age-related macular degeneration, diabetic retinopathy, or inherited retinal degenerations, known as Leber's congenital Amaurosis or retinitis pigmentosa.
Most drug products for ocular disease treatment are marketed as eye drop formulations designed as therapies to the anterior segment of the eye. 1 The topical administration of these formulations hardly ever results in drug concentrations in the posterior eye segment, due to several ocular barriers. 2 Therefore, there is a need for effective delivery systems to target drugs to the retina.
In addition, the rise of new therapeutic molecules [oligonucleotides, genes, short interfering RNA (siRNA), antibodies, and growth factors], together with the need for chronic therapies for most retina diseases, has led to the development of an increasing number of ocular drug delivery systems. Among these, nanoparticulate systems seem to be especially useful due to their submicron size, their capacity to improve corneal absorption, and the increased bioavailability in the eye of lipophilic and hydrophilic drugs formulated in these particles. 3
This review highlights the application of lipid nanoparticles [solid lipid nanoparticles (SLNs), nanostructured lipid carriers (NLCs), or lipid drug conjugates (LDCs)] as effective drug/gene delivery systems for retinal diseases. The ocular hurdles that active molecules have to overcome to reach the retina will be summarized, as well as the strategies usually employed. In this regard, major attention will be focused on the systems based on lipid nanoparticles (NLCs and SLNs) and their application to the treatment of retinal disorders.
Ocular Barriers for Drug Delivery
The access of drugs to target tissues in the eye is largely limited by the high protection of this organ, provided by its anatomy and physiology. In general, the eye (Fig. 2) is divided into 2 parts: the anterior segment (pupil, cornea, iris, cilliary body, and aqueous humor) and the posterior segment (vitreous humor, retina, choroid, sclera, and optic nerve).4,5

Structure of the eye. The image shows the different parts of each segment (anterior and posterior) of the eye.
After topical instillation, lachrymal secretions and blinking action eliminate drugs from the surface of the eye, and the excess of instilled fluid swiftly drains the compounds through the nasolacrimal ducts.3,6 Most of the administered dose is absorbed into the systemic circulation through capillaries or passes transnasally into the gastrointestinal tract 7 with the subsequent risk of side effects. This reduces the ocular bioavailability to 5%–10%. 8
When drugs are administered by topical instillation, 2 routes of absorption have been described 7 : the corneal route (the drug surpasses the cornea and reaches the intraocular tissues through the aqueous humor) and the noncorneal route (the drug surpasses the conjunctival epithelium, then the sclera, and finally, it reaches the choroid and the retina).
The cornea is the outer layer of the eye. This layer consists of the epithelium, the endothelium, and the stroma. The epithelium limits the entry of hydrophilic drugs and macromolecules, while the stroma and the endothelium act as barriers for lipophilic molecules. 9 The cornea may be overpassed in some conditions, such as glaucoma and conjunctivitis, or by using permeation enhancers or mucoadhesives. 3
The conjunctiva is a thin transparent mucous epithelial barrier that contributes to the protection of the ocular surface, the production of the tear film, drug clearance into the systemic circulation, or drug transport to the posterior segment of the eye. 10 This layer may be the channel for topically administered drugs to the back of the eye. 11 Its high surface area, together with a leaky vessel system, makes the conjunctiva an alternative administration route for drugs (including macromolecules) to the retina, although suitable delivery systems are still needed. 1
The blood–retinal barrier (BRB) constitutes a major obstacle for systemic drug delivery to treat retinal diseases. 12 It is composed of the rigid walls of the retinal vessels and the RPE. 13 Its physiological function is the defense of the retina and the vitreous humor from harmful substances 11 ; therefore, the BRB limits the passage of active molecules from systemic circulation to the retina.
Drug Administration Routes to the Retina
Due to the above-mentioned ocular barriers, topical administration does not result in an effective route to reach therapeutic concentrations of drugs in the back of the eye. 2 When the objective is the delivery of active molecules to tissues, such as the retina, systemic, intraocular, or periocular administration routes are needed. 5
Systemic administration
Drugs administered by a systemic route reach the retina through the blood circulation. However, hydrophilic molecules are not able to overpass the BRB, unless it is interrupted. 14 Lipophilic drugs are permeable through retinal capillaries and through the RPE, but the blood flow to the posterior segment of the eye is limited. 15 Consequently, large systemic doses are necessary and there is a significant risk of adverse effects; therefore, this approach is not acceptable for potent drugs with a narrow therapeutic range. 1 Corticosteroids, immunosuppressive agents, and antibiotics are examples of drugs administered orally for the treatment of posterior segment eye diseases, such as, severe infections. 16 The use of drug delivery systems or alternative administration routes, such as intravitreal and periocular routes, may be useful to deliver active molecules to the posterior segment of the eye,17,18 including the retina.
Intravitreal administration
This technique involves the injection of drugs directly into the vitreous, in the form of solutions, particles, suspensions, or implants. Intravitreal injection offers higher drug levels in the retina and reduces side effects of the drugs. Intravitreal injections of monoclonal antibodies have been recently evaluated to treat retinopathy of prematurity, 19 or choroidal neovascularization. 20
The pharmacokinetics of drugs administered intravitreally is affected by several factors. Drugs are eliminated from the vitreous humor mainly via the anterior camera or across the retinal surface. The geometry of the eye and the speed of drug diffusion through the vitreous humor influence the elimination and distribution kinetics. Larger molecules can be retained in the vitreous humor for weeks, but drugs <500 Da administered in a solution show a smaller retention half life of approximately 72 h. 2 Therefore, frequent administrations for an indefinite period of time are required. This is related to short-term adverse effects, such as retinal detachment, endophthalmitis, vitreous hemorrhage, and an increased risk of cataract development.21,22 To avoid these problems, sustained release formulations are a good alternative to reduce the number of injections.
Periocular route
Periocular injections, which include subconjunctival, sub-Tenon, peribulbar, and retrobulbar routes, enable the deposition of drugs next to the external surface of the sclera. This approach is considered the least painful and the most efficient route of drug delivery to the posterior segment of the eye.13,23–25 This administration route offers the opportunity of achieving a localized and sustained release of drugs, 5 as it is safer and less invasive, since periocular routes minimize the risk of adverse effects associated with intravitreal injections.26,27
The sclera is a fibrous tissue with a large surface area, more permeable to macromolecules,13,27–30 although this also depends on tissue hydration and intraocular pressure. However, the vitreous bioavailability of drugs administered via periocular routes is usually low due to the loss of drug from the periocular space, through the BRB and choroidal vessels, and the drug binding to proteins or membrane transporters. 2
Drug/Gene Delivery Systems to the Posterior Segment of the Eye
Up to this date, several delivery systems have been designed to deliver drugs to the retina. Some of these systems aim to improve bioavailability, while the goal of other systems is to achieve the controlled release of the drugs. Moreover, some of these systems can be used for drug targeting, increasing effectiveness, reducing the dose, and the number of administrations and, consequently, improving safety. 31
Drug delivery strategies to the back of the eye may be classified into 3 groups: noninvasive techniques, implants, and colloidal carriers. Tables 1–3 summarize drug delivery systems in the 3 groups, all considered as hopeful approaches for the treatment of disorders of the posterior segment of the eye.
CNTF, ciliary neurotrophic factor.
VIP, vasoactive intestinal peptide.
Iontophoresis and other noninvasive techniques, such as microneedles, are interesting strategies for ocular drug delivery that have been further studied in the treatment of diseases in the anterior chamber of the eye,56–58 although published data related to these techniques are not recent. The best known systems for drug delivery to the posterior eye are intravitreal implants; in fact, some of them are at present in clinical use.35,36 However, their long-term accumulation might impact the patient's vision. 59 On the contrary, colloidal drug delivery systems can be easily administered in a liquid form. 15 Colloidal systems include suspensions of microparticles, liposomes, or nanoparticles. A suitable particle size is recommended to reduce irritation and improve bioavailability and compatibility with ocular tissues. 60 Intravitreal injected microparticles act as a reservoir, although in vitrectomized eyes their life is shorter, whereas drug delivery systems in the nanometer range diffuse rapidly and are internalized in ocular tissues.18,61 Liposomes are small artificial vesicles produced from natural phospholipids and cholesterol. Nanoparticles are particles with a diameter of less than 1 μm, composed by biodegradable polymers or lipids. 62 Nanoparticles have a higher loading capacity than liposomes, and the former are more stable in biological fluids and during storage. 63 Furthermore, nanoparticles will probably be an important part of the new therapeutic armamentarium in ophthalmology because of their intrinsic capacity to adhere to the ocular surface and their interaction with the epithelium. 64
Lipid Nanoparticles
Among nanoparticular systems, lipid nanoparticles (SLNs, NLCs, and LDCs) are more biocompatible, easy to produce at large scale, and they may be autoclaved or sterilized.3,65
Solid lipid nanoparticles
Since the development of SLNs in the 90's,66,67 a large number of researchers have been focusing their work on the application of SLNs by several routes (parenteral, 68 oral, 69 pulmonary, 70 topical, 71 rectal, 72 and ocular 73 ).
SLNs are similar in composition to nanoemulsions, but they replace the inner liquid lipid with a solid lipid in such a way that they are made from solid lipids, where active molecules are stabilized, 74 surrounded by a layer of surfactants in an aqueous dispersion. Various core solid lipids and surfactants have been employed to design SLNs.75–78
Due to the structure of SLNs, both hydrophobic (i.e., cyclosporine A, 73 timolol, 79 saquinavir, 80 and carvedilol 81 ) and hydrophilic (i.e., tobramycin, 82 cisplatin, 83 digoxin, 84 and acyclovir 85 ) drugs have been entrapped in this kind of nanoparticles. The promissing application of SLNs as carriers of hydrophilic molecules is also supported by the preclinical development of Ocusolin™ from AlphaRx, a gentamicin-loaded SLN product in the form of an ophthalmic solution. 86 Drug loading ability depends on the solubility of the drug in the lipid matrix and on the structure and polymorphic state of the lipid matrix. 87 Three models have been described to explain the incorporation of drugs in SLNs65,88: (1) the homogeneous matrix of a solid solution, (2) the drug-enriched shell, and (3) the drug-enriched core (Fig. 3). In the first model, the drug is homogeneously dispersed in the lipid matrix, so that the drug is released by diffusion and/or by degradation of the lipid matrix. In the drug-enriched shell model, the drug is concentrated in the outer shell of the nanoparticles, probably because during the preparation process, the lipid matrix precipitates faster than the drug, and a core with a less drug content is formed. When this occurs, the drug is released fast. 89 Finally, the third model has been proposed to account for cases when the drug precipitates before the lipid; the drug is retained in the core and is released slower from the nanoparticles. 90

Drug incorporation models in solid lipid nanoparticles.
Active molecules must be dissolved in the core lipid melted before particle formation. If the drug concentration in the melted lipid is too high, it might lead to immediate drug expulsion during the cooling process. Moreover, perfect crystal matrices in SLNs cannot contain large amounts of drug and the increasing crystal structure during storage leads to drug expulsion. 91 Hydrophilic drugs are usually adsorbed on the surface of the nanoparticles, resulting in a burst effect: the release of a major fraction of loaded drug in a short time. 92
Nanostructured lipid carriers
The redesign of SLNs at the beginning of the 2000's to overcome their drawbacks, led to the development of a second generation of these particles, the NLCs.65,74,93 NLCs are produced by addition of a spatially incompatible liquid lipid to the solid lipid, leading to special nanostructures with a capacity to accommodate larger quantity of drugs and release properties.74,93 Lipid choice is essential for the stability of the drug in NLCs. 91 Although the liquid lipid in NLCs is up to 30% of the composition, the final carriers are in the solid state. 3 Depending on the structure of the lipid matrix, NLCs may be classified as follows: imperfect type, multiple type, and amorphous or structureless type.
Imperfect type NLCs result from the mixture of a small amount of liquid lipid and solid lipids with very different fatty acids chain lengths. The fatty acid chains of the main lipid core are more distanced, which leads to imperfections in the lipid matrix and a larger amount of drug load. 94 Drug expulsion is reduced with respect to SLNs, but not completely; after production, during the transition from a molten to a solid state, crystallization may occur. 95 The mixing in excess of liquid lipids with solid lipids has been proposed as a strategy to avoid drug expulsion (multiple type NLCs). During the cooling process, the solubility of the liquid lipid in the solid lipid is exceeded and the liquid lipid precipitates in form of oily nanocompartments into the solid lipid matrix, where the drug is better accommodated. 96 Another approach is the use of special liquids that upon cooling solidify, but do not crystallize (i.e., isopropylmyristate hydroxyoctacosanylhydroxystearate), 95 and undesired drug expulsion is avoided. This strategy leads to the formation of amorphous type NLCs.
Lipid drug conjugates
An alternative to improve the encapsulation of water-soluble drugs are LDC. Hydrophilic drugs are converted in lipophilic drugs by conjugating with a lipid (by salt formation or covalent linkage). The lipophilic conjugate may be used alone or processed in the same way as SLNs or NLCs. 97 Even though this strategy is advantageous, few articles about the application of LDCs have been reported in the literature. LDCs have shown hopeful outcomes in the treatment of some parasitic diseases98,99 and also as an oral delivery system for methotrexate. 100 To the best of our knowledge, LDCs have never been used for ocular diseases.
Methods for the Preparation of Lipid Nanoparticles
Lipid nanoparticles have been exploited as delivery and targeting systems of cosmetic and dermal active substances,101,102 conventional drugs,83,103–106 and new pharmaceutical entities, such as peptides and proteins,107,108 antisense oligonucleotides, 109 plasmid DNA,55,110–114 and siRNA.115–118 The encapsulation of complex molecules has led to the development of many preparation techniques 119 as an alternative to the classical ones [high-pressure homogenization (HPH),65,120 microemulsion dilution,66,121 or the solvent emulsification–evaporation technique 122 ], such as the emulsification–diffusion technique, 123 supercritical fluid (SCF) technology, 124 coacervation, 125 the spray-drying technique, 126 the membrane contactor-based method, 127 and the electrospray technique. 128
High-pressure homogenization
In the HPH technique, the inner lipids are melted and dispersed in an aqueous surfactant solution by stirring. The liquid mixture obtained is pushed using high pressure through a narrow gap (in the range of a few microns) at a high speed in a short distance, disrupting the particles to the submicron size. 129
Lipid nanoparticle production by HPH can be obtained by 2 general approaches: hot homogenization or cold homogenization. Figure 4 depicts the differences among these 2 processes.

Comparison of preparation of lipid nanoparticles by Hot high-pressure homogenization (HPH) or Cold HPH methods.
Hot homogenization
The hot homogenization technique is performed following these steps:
1. Dissolving, dispersion, or solubilization of the drug in the melted lipids. 2. Dispersion of the melted lipid containing the drug in a hot aqueous solution of surfactant. 3. Preparation of a pre-emulsion by high-speed stirring of the previous dispersion. 4. Application of HPH to obtain a hot nanoemulsion. 5. Solidification by cooling at room temperature to obtain lipid nanoparticles.
These are some of the advantages of this technique: the low particle size and polidispersity index, possible application for medium or large scale, 130 low capital cost, 131 and absence of solvents. Among the disadvantages, one should consider the rate of active degradation due to elevated temperatures 132 and the difficulties in loading hydrophilic molecules efficiently due to the burst effect. 133
Cold homogenization
In the cold homogenization technique, the drug is not subjected to such high temperatures. The process would be as follows:
1. Dissolving, dispersion, or solubilization of the drug in the melted lipid/lipid blend. 2. Solidification of the drug lipid mixture by rapid cooling using liquid nitrogen or dry ice. 3. Milling of the drug containing solid lipid to reduce the size to the micron range. 4. Preparation of a presuspension by dispersion of the lipid phase in a cold aqueous solution of surfactant. 5. Application of HPH at/bellow room temperature to obtain lipid nanoparticles.
With this technique, the burst effect and the thermal exposure of the drug are reduced, but the particle size and polidispersity index are higher compared to the ones in the hot technique. 131
Microemulsion dilution
The microemulsion dilution technique basically consists in the addition of a microemulsion to a high volume of cold water, so that the inner lipids precipitate and lipid nanoparticles are formed (Fig. 5). The general steps of this process are the following:
1. Melting of the lipids in a recipient. 2. Heating of the aqueous solution of surfactants and cosurfactants (if necessary) to the same temperature as the lipids. 3. Preparation of the microemulsion by adding the aqueous solution to the lipids under soft stirring. 4. Dispersion of the clear thermodynamically stable system (obtained in the previous step) in cold water to precipitate lipid nanoparticles with reduced particle size.

Preparation of lipid nanoparticles by the microemulsion dilution technique.
As a final step, washing with distilled water and posterior membrane filtration are useful to remove undesirable bigger particles. 134
This process requires no special equipment, but features some disadvantages. For example, high temperatures are needed, although Koziara et al. 135 modified this method by using an emulsifying wax with a lower melting point (37°C–55°C) to obtain a microemulsion that is dispersed in water at room temperature. This process allows the formulation of lipid nanoparticles at mild operating temperatures. Besides, the suspensions obtained have a very low particle concentration, which is a major obstacle for industrial application, 130 although the excess of water may be removed by freeze-drying. 134
Solvent emulsification–evaporation
In the solvent emulsification–evaporation method, an emulsion is initially prepared by high-speed homogenization of a lipidic phase (composed of the lipids dissolved in an organic solvent) and an aqueous phase containing surfactants (Fig. 6). These are the steps to be followed:
1. Dissolving the lipids in a water immiscible organic solvent. 2. Homogenization of the organic phase with an aqueous solution of surfactants by high stirring to obtain an emulsion. 3. Evaporation of the organic solvent by stirring at room temperature to precipitate the lipid and form the dispersion of lipid nanoparticles.

Preparation of lipid nanoparticles by the solvent emulsification–evaporation technique.
This technique avoids thermal stress, and it is appropriate for the encapsulation of thermolabile compounds. However, the use of organic solvents is a limitation.
Solvent emulsification–diffusion
As the method just mentioned, the solvent emulsification–diffusion process is also characterized by the preparation of an emulsion precursor. The difference lies in the solvent used: in the present method, the solvent must be partially miscible with water (e.g., ethyl acetate, isopropyl acetate, and benzyl alcohol). The critical steps are listed below:
1. Mutual saturation of both the solvent and the water. 2. Dissolution of the lipid and drug in a water-saturated solvent. 3. Emulsification of the previous solution with an aqueous solution containing a stabilizer. 4. Dilution of the system in water in ratios from 1:5 to 1:10, to ensure solvent diffusion into the continuous phase, thus allowing nanoparticle formation by lipid precipitation. 5. Elimination of the diffused solvent by vacuum distillation or lyophilization.
This approach is advantageous 129 because of its versatility, the absence of high-energy sources (thus making the scaling up easy) or the narrow and reproducible size of the nanoparticles formed. On the contrary, characteristics, such as the need to clean up and to concentrate nanoparticles, and the low drug entrapment, 136 limit the application of this technique.
Solvent injection
The basis of the solvent injection technique is similar to the solvent emulsification–diffusion method, as summarized in the following steps
137
:
1. Dissolution of the solid lipid and the drug in a water-miscible solvent or a water-miscible solvent mixture. 2. Injection of the previous solution into a stirred aqueous phase with a surfactant. 3. Formation of droplets at the site of injection and formation of lipid nanoparticles by solvent diffusion. 4. Removal of excess lipid by filtration of the dispersion through a filter paper.
Supercritical fluid
The basis of the SCF technique is the use of SCF. These gases show unique thermophysical properties: the density of the gas increases as the pressure raises, while the viscosity does not notably increase, but the ability to dissolve compounds is improved. By careful control of the temperature and pressure, the ability of a gas to dissolve a compound may be altered. The most used SCF for this technique is the carbon dioxide (CO2). Among the different processes using the SCF technology, SCF extraction of emulsions (SFEE) and gas-assisted melting atomization (GAMA) have been developed to produce lipid nanoparticles. 138 Figure 7 shows the schemes of these processes.

Comparison of supercritical fluids (SCF) extraction of emulsions (SFEE) and gas-assisted melting atomization (GAMA), as methods to prepare lipid nanoparticles.
SCF extraction of emulsions
In the SFEE method, the suspension of lipid nanoparticles is obtained by SCF extraction of the organic solvent of an emulsion
139
:
1. Preparation of the emulsion. 2. Introduction of the emulsion into an extraction column from the top and countercurrently supercritical CO2 from the bottom. 3. The solvent is extracted into the supercritical CO2 and lipid-drug material precipitates in the form of nanoparticles.
The main advantage of this technique is the solvent extraction efficiency, in comparison with the above-mentioned methods.
Gas-assisted melting atomization
When the GAMA method is chosen, the following steps must be performed
140
:
1. Melting of lipids in contact with supercritical CO2, in a mixing chamber. 2. Forcing the lipid-saturated mixture through a nozzle, thus inducing a rapid depressurization. The high supersaturation induces particle precipitation. 3. Collection of nanoparticles. 4. Dispersion of nanoparticles in water by vortexing and ultrasound treatment.
Coacervation
The coacervation technique is based on a phase transformation from micellar solution into fatty acid solid particles by acidification
83
as follows:
1. Preparation of a micellar solution of a sodium salt of a fatty acid where the drug is incorporated. Emulsifiers, such as sodium dioctylsulfosuccinat, may be required to solubilize hydrophilic molecules. 2. Slow addition of an acid solution (coacervating solution) to the previous micellar solution, in the presence of an appropriate amphiphilic polymer as a stabilizing agent. 3. Precipitation of the lipid nanoparticles due to the low pH.
Thermolabile drugs may be encapsulated by this method, without solvents or very complex equipment use.
Spray-drying
The spray-drying technique is a one-step process, which transforms a liquid feed into a spray dryer into dried particles. It may be considered an alternative to the lyophilization process.
The liquid, an organic solvent solution, is first atomized to a spray as it is fed into the spray dryer. The spray is immediately put in contact with a hot gas, thus evaporating the solvent to form dried nanoparticles. These are separated from the gas by typically using a cyclone, although other systems, such as an electrostatic precipitator or a bag filter, may be used. 141
Membrane contactor-based method
The membrane contactor technique has been developed for large-scale production of lipid nanoparticles. The lipid phase is subjected to pressure at a temperature above the melting point, through the membrane pores, thus forming small droplets (Fig. 8). These droplets are detached from the membrane surface by stirring the aqueous phase continuously and tangentially. Lipid nanoparticles are formed after cooling the dispersion below the lipid melting point. 127 The particle size may be modified by controlling several parameters, such as the temperature of the lipid fusion or the lipid phase pressure. 142

Scheme of the lipid nanoparticle preparation by the membrane contactor technique. (Modified from Charcosset et al. 127 ).
The scaling up ability, the simplicity and the control of particle size are clear advantages of this technique. 143
Electrospray
Electrospraying is a novel technique proposed to prepare monodisperse lipid-based micro- and nanoparticles for drug delivery in a single step. An electrostatic atomizer is used for this purpose (Fig. 9).

Schematic drawing of the electrospray technique for lipid nanoparticles preparation.
The lipid dissolved in an organic solvent is loaded in a syringe. The solution is directed into a nozzle connected to a high-voltage power supply to be atomized. A metal foil collector is placed opposite the nozzle as a counter electrode. The droplets are formed due to the electrical field; and after the evaporation of the organic solvent, the lipid nanoparticles are obtained. The flow rate and the voltage applied can be modulated to modify the size range. 128
Application of Lipid Nanoparticles in the Retina
An ocular drug delivery system based on lipid nanoparticles may prolong drug residence time on the ocular surface and conjunctival sac, in comparison with drug solution drops. 3 Cavalli et al. 82 prepared fluorescent SLNs containing tobramycin by a warm o/w microemulsion technique. After topical administration in the eye of male New Zealand albino rabbits, the presence of SLNs in the eyes was observed to over 1h. In addition, ocular bioavailability of encapsulated tobramycin increased with respect to the administration of the drug solution. SLNs produced a Cmax increase (1.5-fold), a tmax increase (8-fold), and a 4-fold AUC increase with respect to the reference solution. The authors attributed these results to the entrapment of SLNs in the mucin layer over the epithelium, the sustained release of tobramycin, and the enhancement of penetration induced by soya phosphatidyl choline, employed in the preparation of the nanoparticles.
Despite the apparent advantages of the lipid nanoparticles as a drug delivery to the eye, few research groups have done research on this issue. The number of publications is even scarcer, when these systems are applied to deliver drugs to the posterior segment of the eye.
Besides, as a tobramycin delivery system, 82 SLNs have also been studied as delivery systems of other poorly absorbed ocular drugs, such as diclofenac, 103 timolol, 144 or cyclosporine A.73,145 Attama et al. developed SLNs for ocular delivery of both diclofenac 103 and timolol, 144 and observed that, regardless of the drug studied, the modification of the SLNs surface with phospholipids provided a higher loading efficiency without a burst effect, which is ideal to achieve a sustained release of the drug over an extended period of time. Moreover, the same authors showed that drugs efficiently crossed the cornea, in a permeation study using a human cornea construct model. Gokce et al. 73 evaluated in vitro and ex vivo sterile SLNs containing cyclosporine A. The lipid nanoparticles enhanced penetration of cyclosporine A in the cornea, which was related to the internalization of SLNs in the cornea and in the corneal epithelial cell lines. Later, Basaran et al. 145 prepared cyclosporine A-loading cationic SLNs, by adding a positively charged lipid, octadecylamine, to the melted lipid containing the drug. After autoclaving, the nanoparticle suspension experienced a decrease of the pH in only 1 month, which was attributed to fatty acid degradation. Cyclosporine A has also been incorporated into NLCs systems. 146 The authors of this work observed that incorporation of liquid lipids significantly increased drug loading. In vitro release studies showed a fast release of cyclosporine A on the first 12 h, followed by sustained release. In addition, the higher the liquid lipid content was, the faster the release rate; higher concentrations of liquid lipid were also related to increased cellular uptake. After in vitro characterization, cyclosporine A-loaded NLCs were administered to rabbit eyes. The formulation was well tolerated and provided a prolonged ocular surface retention and corneal penetration. As a strategy to improve the efficiency of this system, Shen et al. 147 modified the NLCs with cysteine polyethylene glycol monostearate to obtain thiolated NLCs. This nanoparticular system prolonged precorneal residence time, and high cyclosporine A levels were obtained in the eye. NLCs have been also proposed as promising and effective systems for ocular delivery of the anti-inflammatory flurbiprofen. 148 In this regard, NLCs were modified with different types of chitosans.149,150 Due to their mucoadhesive and antimicrobial properties, chitosans have been considered for ophthalmic application. The incorporation of chitosans to the surface of NLCs improved the efficiency of these systems by enhancing transcorneal penetration.149,150
All the studies mentioned above are focused on the delivery of drugs to the anterior segment of the eye. However, their results and conclusions are very useful for advances in the treatment of diseases of the posterior segment of the eye and more specifically, of the retina, an area where lipid nanoparticles have been less studied. An example is the use of lipid nanoparticles to administer triamcinolone acetonide in the posterior segment of the eye. Intravitreal injection of ophthalmic suspensions containing triamcinolone acetonide has become increasingly popular to treat a broad spectrum of retinal diseases. 151 However, this kind of administration is not ideal, due to the risk associated with intravitreal injection: retinal detachment, endophthalmitis, vitreous hemorrhage, and cataract development.21,22 To reduce these systemic adverse effects, to reach higher concentration of drug at the site of action during an extended period of time, and to avoid the complications involved in intravitreal injection, Araújo et al.54,152 developed a formulation based on NLCs containing triamcinolone acetonide for ocular instillation. In a first study, 152 the authors produced NLCs by HPH using Precirol®ATO5, as solid lipid, Squalene®, as liquid lipid, and Lutrol®F68 as surfactant. The carriers showed a particle size range of 100–300 nm, with a low polidispersity index, a high zeta potential, and an entrapment efficacy of approximately 95%. When the formulation was administered in the conjunctival sac of male albino New Zealand rabbits, no irritation was detected. After this preliminary study, Araújo et al. 54 administered the formulation in female adult CD1 mice onto the eye surface. NLCs were detected in the retina, reaching a peak 40 min after administration, decreasing thereafter, and almost disappearing 160 min after administration. NLCs were longer retained on the ocular surface, which was related to the absorption of the drug and its appearance in tissues. Taken together, these studies show that lipid nanoparticles are a promising approach to provide a selective and prolonged drug concentration in the posterior segment of the eye and to avoid the side effects associated to the intravitreal injection.
Lipid nanoparticles and retinal gene therapy
Gene therapy consists in the introduction of foreign nucleic acids into target cells, and induce the expression of a protein that is altered in a certain disease. A gene delivery system is needed to facilitate cellular uptake and intracellular processing of the exogenous active molecule. The eye is an interesting organ for gene therapy because of its well-defined anatomy, accessibility, transparency, and immunoprivilege. 153 The clinical potential of gene therapy in eye diseases has been demonstrated by stable reversal of blindness in phase I clinical trials with patients suffering from Leber`s congenital amaurosis.154–156 In these clinical trials, patients were treated with adenoviral vectors. In spite of the promising results, viral vectors present important drawbacks due to their immunogenicity and oncogenicity. 157 In addition, the potential persistence of viral vectors in the brain after intravitreal injection has been documented. 158 These limitations have led to the development of nonviral systems, including lipid nanoparticles.
Among lipidic systems, SLNs have shown good capacity for transfection in vitro153,159 and in vivo.111,160 Moreover, SLNs have good stability and are subject to be lyophilized, 161 which facilitates large-scale manufacturing and storage.
In the field of retinal diseases, hopeful results have been obtained with SLNs as nonviral vectors. In a first approach, 153 cationic SLNs composed by Precirol® ATO 5 as solid lipid, and DOTAP and Tween 80 as surfactants, were prepared by a solvent emulsification–evaporation method. SLNs carrying a plasmid model for the enhanced green fluorescent protein only transfected 2.5% of the RPE cells treated in vitro as a model of inherited retinal diseases. The study of the intracellular trafficking of these vectors allowed the identification of the main limiting steps for transfection and, as outlined below, the setting of different strategies to develop more efficient transfectants. On the one hand, the cellular uptake of SLNs into RPE cells is carried out predominantly by the clathrin-mediated endocytosis that directs the vectors to lysosomes, where the content can be degraded. In this regard, when SLNs were combined with the cell penetrating peptide SAP, 162 the percentage of transfected ARPE-19 cells in vitro increased. This effect was attributed to the capacity of the peptide to favor caveolae/raft-dependent endocytosis over clathrin endocytosis. Caveolae/raft-mediated endocytosis avoids the lysosomal pathway and its consequent degradation of the vector and, therefore, transfection increases. On the other hand, the low division rate of retinal cells hampers the entrance of DNA into the nucleus. There are 2 mechanisms to overcome the nuclear envelope: the interruption of the nuclear membrane during mitosis or the entrance through the nuclear pore complex. This latter mechanism requires nuclear localization signals, which are used to improve nonviral vector-mediated transfection. 163 With this in mind, SLNs were combined with protamine, 78 a peptide that condenses DNA and presents sequences of 6 consecutive arginine residues, 164 which make it able to translocate molecules, such as DNA from the cytoplasm to the cellular nucleus. Protamine induced a 6-fold increase in the transfection capacity of SLNs in retinal cells, reaching a 29% of transfected cells, which was maintained during at least 7 days. To the best of our knowledge, no other article in the literature had shown a similar transfection level in RPE cells with nonviral vectors. From studies of the internalization mechanism and intracellular disposition of DNA, the authors of this study observed that vectors containing protamine require endocytosis via clathrin for transfection because the lysosomal activity induces the release of the complex protamine–DNA from the SLNs. In a follow-up study, the vector was modified by adding dextran. 55 This polyanion is a biocompatible polysaccharide that is used as a clathrin-mediated endocytosis marker, and, in consequence, it has ability to enhance the cellular uptake by using the clathrin-mediated pathway. 165 The new system, patented for gene therapy application, 166 significantly increased the transfection of RPE cells, reaching levels close to 50% of cells transfected. The vector has been evaluated as a carrier of the therapeutic plasmid pCEP4-RS1, which encodes the protein retinosquisin, whose deficiency is responsible for the X linked juvenile retinoschisis. This is a common cause of male juvenile blindness with a prevalence of 1:5,000 to 1:25,000. 167 The vector composed of SLNs, dextran, protamine, and the plasmid pCEP4-RS1, was able to transfect in vitro ARPE-19 cells, producing significant amounts of retinosquisin. In a preliminary study, the vector was administered to Wistar rats by different ocular routes: intravitreous, subretinal, and topical. Depending on the administration route, the protein expression was detected in different types of cells. There was a good response in retina ganglion cells when intravitreal injection was employed, but protein expression was poor in RPE cells. In contrast, after subretinal injection, the vectors transfected RPE cells as well as photoreceptors. These vectors were also able to transfect corneal cells after topical application. 55
Conclusions
This review has collected evidence of the potential usefulness of colloidal lipid systems in the treatment of diseases of the posterior segment of the eye, and specifically of the retina. There is a critical need for effective drug delivery systems for this group of diseases, including systems to deliver traditional active molecules or new entities, such as, genes.
Research articles reported in the present review evidence the potential application of lipid nanoparticles as drug delivery systems for the retina, and also as nonviral vectors in gene therapy of retina disorders. In spite of the promising results, further studies are still needed to asess the real potential of these lipidic systems before they reach the market.
Footnotes
Acknowledgments
None.
Author Disclosure Statement
No competing financial interests exist.
