The physics of lipidic mesophase delivery systems

An ideal drug delivery system should be biodegradable and biocompatible and should incorporate the active pharmaceutical ingredient without losing or altering its activity. To optimize the drug’s therapeutic time window, the system should provide an efficient and controlled release mechanism for a drug at a specific location in vivo. Lipidic mesophases, a special class of materials belonging to the lyotropic-liquid-crystal family, meet those criteria. They thus share characteristics of both crystals and liquid materials: Like crystals, their molecules have short-range positional order, and like liquids, they lack long-range positional order. Lipidic mesophases exhibit distinct structures based on their composition, and they constitute an attractive alternative drug delivery system, capable of incorporating molecules of different polarity and size and protecting them from chemical and physical degradation.

As figure 1 depicts, when lipid molecules are hydrated, they can self-assemble in different arrangements that correspond to the lamellar, hexagonal, cubic, and micellar cubic phases, among others. ¹ The lamellar phase consists of a two-dimensional stack of amphiphilic bilayers separated by aqueous layers. Each bilayer consists of two monolayers packed tail-to-tail to minimize contact between the hydrocarbon chains and water. In that arrangement, water partitions nearly exclusively to the lipid polar heads and forms slabs of water–lipid heads. The lamellar phase is denoted L_α if the lipid tails are disordered in the 2D layers, L_β if they are organized in a lattice. The hexagonal phase consists of an ensemble of cylindrical micelles packed on a hexagonal lattice. The water–lipid heads form the cylinders, whereas the lipid tails fill the continuous matrix.

Figure 1.

The bicontinuous cubic phases are composed of two sets of water channels separated by curved bilayers in the 3D space such that every point on the bilayer midplane surface is a saddle point with a zero-mean curvature. ² Those bilayers encompass a system of aqueous channels and form structured yet flexible networks that are nonbirefringent and optically transparent. Specifically, lipid molecules in bicontinuous cubic phases form a highly curved continuous bilayer that separates the two interpenetrating but nonintersecting aqueous channel networks. The structures can display a double gyroid (Ia3d, with threefold connectivity of aqueous channels), double diamond (Pn3m, with fourfold connectivity), and primitive (Im3m, with sixfold connectivity) symmetry. ³ Finally, the inverse micellar phase comprises spheroidal micelles arranged in Fd3m or Fm3m symmetry, or in a disordered state L₂. Most lipids do not show those phases in pure water and need oil-based components, such as oleic acid or limonene, to form the micellar cubic phase.

The critical packing parameter CPP—defined as the ratio between the hydrophobic chain volume and the hydrophobic chain length in the molten state times the cross-sectional molecular area—can predict qualitatively the structural phase given by each lipid in the presence of water. However, because of the complexity of bicontinuous cubic phase structures, CPP theory cannot fully describe the system. ⁴ Lipids, such as phytantriol (PHT), monoolein (MO), and monolinolein (MLO), and phospholipids are the most well-known macromolecules capable of forming lyotropic liquid crystals in water via self-assembly. MO and MLO, which are generally recognized as safe for human and animal use by the US Food and Drug Administration, are the most widely used materials for encapsulating a broad range of drugs with various sizes and polarities. Their phase diagrams, shown in figure 2, describe their behavior at thermodynamic equilibrium, where the system is at its lowest state of free energy. The lipid–water mixtures show lyotropic-liquid-crystalline behavior: Changing the water content or the temperature results in a different structure. By increasing the water content, the lamellar phase of MO, for example, transforms first to an Ia3d and then to a Pn3m phase, which in turn will swell until reaching its maximum hydration beyond which Pn3m and water coexist. On the other hand, increasing the system’s temperature induces a transition from the cubic phase to an inverse hexagonal phase.

Figure 2.

Once the maximum hydration level has been reached, the Pn3m phase cannot swell anymore. The excess water given to the system remains confined outside the cubic structure, and the Pn3m bicontinuous cubic phase coexists with excess water at thermodynamic equilibrium. The Im3m bicontinuous cubic, hexagonal, and the micellar Fm3m phases show a similar behavior: Excess water coexists thermodynamically with lipidic mesophases.That behavior is a crucial and distinctive feature of lipid mesophases and makes them valuable as drug delivery systems. The thermodynamic equilibrium allows researchers to design nanostructures that remain stable for the entire drug-delivery process, which makes lipid mesophases more advantageous compared to other nanostructured fluids, such as thermodynamically unstable nanoemulsions. ⁵

When mesophases coexisting with excess water have reached their maximum hydration level, a dispersion of bulk phases can be formed by adding energy to the system. For sonication, the most commonly used approach to form dispersions, a lipid film is hydrated in the desired buffer containing a stabilizing polymer, and sound waves agitate the film until the sample becomes a homogenous dispersion. ⁶ The water-dispersed forms of the parent cubic- and hexagonal-phase bulk counterparts are cubosomes and hexosomes. They possess the same nanostructure as the parent bulk phase but exhibit a much lower viscosity, improved fluidity, and larger surface area. Such dispersions are commonly prepared by mixing lipids with a stabilizing polymer that acts as a surfactant and improves the nanostructure’s shelf life. With those dispersions, researchers can deliver poorly water-soluble drugs, selectively target malignant cells, and use molecular detection to identify and deliver radioactive drugs for radiation therapies. ⁷

Phase characterization techniques

Researchers routinely use small‐angle x-ray scattering (SAXS) to determine the lipid phase, and thus construct the phase diagram, for different lipid–water systems. That method relies on constructive interferences in the reciprocal space from many ordered scattering planes that belong to the mesophase. An x-ray beam is directed at the lipid sample, and the resulting scattering pattern gives a characteristic set of rings, or maxima, that correspond to Bragg reflections. Their positions in the reciprocal space depend on the Miller indices of the mesophase scattering planes, and the sequence of Bragg reflections consequently identifies the symmetry of the mesophase studied. SAXS allows for the lattice parameter—the size of the repeat unit cell—to be determined. When the parameter is translated, researchers can reconstruct the entire mesophase in 3D. The lattice parameter and the phase composition enable scientists to determine structural features of the mesophase, such as the width of a lipid bilayer and the diameter of water channels, micelles, and slabs.

Besides SAXS, cryogenic transmission electron microscopy (cryo‐TEM) also characterizes lipidic mesophases. Rather than probing reciprocal space, it uses direct measurements of real-space structures. Using cryo-TEM, scientists have obtained high‐resolution images of bicontinuous cubic mesophases, which enables direct visualization of their internal channels, size, and other important structural information. Since SAXS probes structure in the reciprocal space and cryo-TEM in real space, taking the Fourier transform of the lipid structures in the cryo‐TEM images results in a phase assignment equivalent to that obtained via SAXS. ⁸

Rheology of lipidic mesophases

One of the first features to consider when designing drug delivery systems based on lipidic mesophases is the rheological behavior of the mesophases, which varies tremendously. Measurements of the storage moduli G′ and the loss moduli G″ as a function of the shear frequency ω reveal important information about the rheological behavior of the mesophases. Different viscoelastic regimes emerge because of the diverse topologies of the mesophases, and each regime’s water–lipid interface has its own specific relaxation time. The viscosity of mesophases increases progressively from the less viscous lamellar phase to the viscoelastic hexagonal and the highly elastic bicontinuous cubic phases. ⁹ Specifically, the lamellar phase behaves as a plastic fluid that has extensive energy dissipation mechanisms due to the parallel slip of the lamellae. The hexagonal phase behaves as a viscoelastic material with a relaxation time of about 0.5–1 s. In contrast, the bicontinuous cubic phases show complex rheological behavior simulated by a multiple Maxwell model. Those phases have characteristic times of up to several tens of seconds that are associated with the relaxation of the lipid head group–water interface. The variation of the relaxation time as a function of the temperature or composition captures subtle cubic-to-cubic phase transitions and the more trivial cubic and excess-water transition.

Although the viscous cubic phase has an optimal rheology to act as a drug depot, its high viscosity could make the subcutaneous administration of drugs challenging. Water can overcome that challenge by triggering the in situ formation of gelled lipidic structures, starting from a liquid or less viscous material. ¹⁰ For example, a technology developed by the Swedish company Camurus dissolves the drug in a lipid solution that transforms to gel upon contact with water. After subcutaneous injection, the precursor lipid solution swells from the water naturally present in the tissue and forms the lipid depot in situ.

Connecting mesophase structure and drug release

Mesophases can slowly release hydrophobic or hydrophilic biomacromolecules such as proteins, antibodies, RNA, and DNA. But some small hydrophilic drugs encapsulated in the water channels of a mesophase burst out of it. So far, the most commonly explored approaches to control a hydrophilic drug’s release from lipidic mesophases include ionic interactions between the drug and the lipid polar group, variation of the water-channel dimensions, and the formation of a lipid prodrug, a compound that activates in the body to produce a working drug.

The release processes of hydrophilic molecules depend on the structure of the mesophase. The release phenomenon—a Fickian diffusion process—follows a first-order kinetic profile regulated by the size of the mesophase’s aqueous channels, the size of the drug, and most importantly, the symmetry of the mesophase. To be released, hydrophilic drug molecules must diffuse through the water domains and cross the lipid domains. The ability of the lipids to keep a drug confined in the water channels controls its release. Drug diffusion depends on the dimensionality d of the water domains, defined as the number of orthogonal directions a molecule may explore without leaving the water channel.

In the micellar cubic mesophase, the drug is confined in the micelles with water domains of about 2 nm in diameter (d = 0). In a hexagonal phase, the molecules move through cylindrical water channels (d = 1), whereas in a lamellar phase they can move along the planar bilayers (d = 2). In bicontinuous cubic phases, the diffusion is a 3D process (d = 3) in which the mobility of embedded hydrophilic molecules is hindered only by the channel size—about 5 nm in diameter—and its tortuosity. In general, the diffusion coefficient of a hydrophilic drug decreases with the dimensionality of the phase. Therefore, the Pn3m phases (d = 3) show the highest diffusion coefficients, whereas the micellar cubic phases (d = 0) show the lowest, as illustrated in figures 3a and 3b. That the release rate of drugs from micellar cubic phases is low but nonzero highlights that partitioning of the drug within the lipid bilayers contributes to the release kinetics.

Figure 3.

Researchers in 2015 introduced the structural control efficiency index (SCEI). ¹¹ It provides an estimate of the kinetics of drug release for various phases and hydrophobic domains, which have different symmetries and permeability to drugs. The SCEI = log (D_max / D_min) / DS, where D_max and D_min are the highest and lowest diffusion coefficients observed for a given system, and DS is the structural dimension span explored; DS = 3 for MLO-based systems, whereas DS = 2 for phosphatidylcholine-based systems. Figure 3c graphs the slope in the logarithmic plot of the diffusion coefficients versus the structural dimension and visualizes the SCEI. The hydrophobic lipid domain shows high permeability to a hydrophilic drug such as caffeine, which easily penetrates lipidic barriers, as indicated by the partition coefficient. On the other hand, a drug with high hydrophilicity and large size has a highly negative partitioning coefficient, the lowest diffusion coefficient, and optimal structural control.

For a given drug–lipid pair with the same dimensionality, topology controls the diffusion process. Diffusing molecules, for example, interact with the geometrical attributes of the three main symmetries of inverse bicontinuous cubic phases—Pn3m, Ia3d, and Im3m—which make a pivotal contribution in determining the diffusion mechanism. For small hydrophilic molecules such as glucose, at identical water content the diffusion coefficient decreases from a maximum to a minimum when going from Ia3d to Pn3m to Im3m. As a general rule, the less porous Im3m symmetry shows a slower transport efficiency at water content parity, compared with the more porous Pn3m geometry. Moreover, mesophase symmetry, lipid characteristics, and particle features must all be considered when developing systems with optimized release behavior. ¹² In contrast and somewhat surprisingly, the water-channel connectivity does not influence diffusion, which is a result of the properties of the random walk in a lattice. ¹²

Tuning with additives

Additives can modify the structure of the lipid mesophases. For example, adding an increasing amount of hexadecane or vitamin A can tune the phase of the MO–water system. The release profile, and therefore the drug diffusion coefficient, changes with the phase. In particular, the release from the cubic phase is significantly faster than the release from the micellar, lamellar, and hexagonal phases. The geometry of self-assembled mesophases is important in determining the release rate and confirms the hypothesis that the open or closed state of the aqueous channels influences the rate of drug release.

Typical lipidic mesophases with a Pn3m, Im3m, or Ia3d symmetry are characterized by water channels with a diameter of 3–5 nm. That geometric constraint prevents large hydrophilic molecules, such as hydrophilic proteins, hormones, and antibodies, from being included in the mesophase. But that major structural limitation can be overcome by additives that increase the water-channel dimensions, including hydration-modulating agents such as sucrose stearate, phospholipids, and cholesterol. Electrostatic swelling also increases the size of water channels, either by doping lipids with charged lipids capable of swelling the mesophases or by decreasing the chemical potential of water in the nanochannels so that the total free energy of the system requires more water to equilibrate. Electrostatic swelling has been demonstrated by several research groups, ^13–15 and in those studies, the cubic phases have water channels more than five times as large as traditional lipidic mesophases.

Interestingly, researchers use a combination of neutral lipids such as monopalmitolein (MP) and charged ones such as one called DSPG to design ultraswollen cubic phases of double gyroid (Ia3d), double diamond (Pn3m), and double primitive (Im3m) structures. As shown in figure 4, the channels are sufficiently large to allow in-meso crystallization of membrane proteins with gigantic extracellular domains. ¹⁵ In such a case, the increasingly hydrated mesophases reenter the phase diagrams of the type Ia3d → Pn3m → Ia3d and Pn3m → Im3m → Pn3m. The reentrances depend on the amount of phospholipid used, which is unique in lipid self-assembly. An Ia3d phase normally converts to a Pn3m phase upon hydration and will swell until reaching its maximum hydration; similarly, Pn3m swells upon hydration and converts into an Im3m symmetry, which can swell further before reaching its maximum hydration.

Figure 4.

The phase reentrances indicate the competing effects at play when electrostatic swelling and increasing hydration occur together. One possible scenario is that increasing the length scales upon swelling by raising the water content decreases the effectiveness of the electrostatic interactions, which moderates the swelling effect at extreme swelling ratios. That hypothesis requires validation, and the reentrances of the phase diagram for electrostatically swollen lipidic mesophases highlight how multicomponent lipidic mesophases continue to reveal fascinating and unexpected surprises. Electrostatic swelling of the mesophase via the addition of charged lipids is a promising tool to generate thermodynamically stable swollen cubic phases. With that approach, one could load large hydrophilic biomacromolecules: The resulting lipidic surface would be decorated by a negatively charged lipid and increase the gel’s adhesion with positively charged inflamed tissues.

Adding even low concentrations of electrolytes to the aqueous phase of water–amphiphile systems can significantly modify the corresponding phase diagrams. MO–water mixtures, depending on the type and amount of electrolyte, swell and shrink the water channels. Sodium chloride or sodium sulfate dehydrates the monoolein head groups at the interface, which reduces their occupied effective area; the result is a shrinking of the water channels. On the other hand, sodium iodide leads to an increase in the hydration of the lipid head groups, expands the interfacial area occupied by a monoolein head group, and swells the water channels of the cubic phase. Low concentrations of Ca²⁺ can shrink the lipidic mesophase water channels, and if the Ca²⁺ concentration increases up to 30 millimolar, it can transform the multilamellar vesicles—formed by MO and a negatively charged lipid such as DOPG—into a cubic phase. ¹⁰

As depicted in figure 3, a drug’s release can also be controlled by including additives that render the delivery system sensitive to internal stimuli—including pH and temperature—and external stimuli, such as light and magnetic fields. With appropriate molecular design, the delivery of therapeutic agents can be spatially and temporally controlled by triggered drug release. Such on–off switching is advantageous when continuous drug release may be toxic, and it should improve the therapy’s effectiveness. Mesophases can host stimuli-responsive lipids without impairing structure and stability and can endow them with specific functionalities.

For example, figure 5a indicates that adding 3% by weight of vitamin A to the MO–water system leads to an efficient temperature-induced release of hydrophilic molecules. The system composed of MO, vitamin A, and water at 30 °C shows a Pn3m phase and a faster drug release ¹⁶ compared with the hexagonal phase obtained at 40 °C. The mesophases composed of monolinolein–water doped with pyridinylmethyl–linoleate show a hexagonal-to-cubic transition induced by the slightly acidic pH typical of cancerous tissues. Figure 5b indicates that for a pH of 5.5 versus 7.4, the phase transition shows a faster release, which means that the hydrophilic anticancer drug doxorubicin, loaded into the engineered mesophase, has an increased efficiency for killing cancer cells at low pH. ¹⁶

Figure 5.

With respect to external stimuli, the application of light offers high spatial and temporal precision. Modern lasers modulate drug release profiles and provide a high level of pharmacological control by regulating the light’s wavelength, intensity, duration, and spot size. Light, therefore, can directly affect the nanocarrier’s assembly and thereby lead to drug release. As figure 5c illustrates, a light-responsive mesophase composed of a mixture of monoolein and a small amount of a light-sensitive lipid produces a sequential light-triggered release when irradiated with UV light and retains an embedded hydrophilic dye when exposed to visible light. ¹⁷

The outlook for lipidic mesophases

To date, inverse bicontinuous cubic and hexagonal phases have both been shown to hold a wide range of active pharmaceutical ingredients. Because researchers can develop and design new lipid molecules capable of self-assembling into lipidic mesophases, the use of mesophases will grow quickly in the near future. Although lipidic mesophases are an interesting drug delivery system, they haven’t been as extensively commercialized as analogous products, such as liposomal delivery systems, emulsions, and nanosuspensions. To bring lipidic mesophase products to market will require a clinical translation process, and research has been more focused on gaining information about the physics of these systems. The pharmacokinetics and pharmacodynamics in the biological environment remain less advanced, and researchers are still uncertain how the kinetics of lipidic mesophase formation in the body affects tissue tolerability.

The Camurus company is developing a new technology using a lipid solution, containing the dissolved drug, that transforms into lipidic mesophases upon contact with water. The Camurus pipeline contains different liquid crystalline systems at various stages of clinical trials. For example, their FluidCrystal technology includes injection depot, topical bioadhesive, and nanoparticle formulations based on lipid self-assemblies. Once duly engineered, lipidic mesophases could outperform the more commonly used micelles, emulsions, and liposome systems in terms of local and systemic tolerability, encapsulation efficiency, physical instability, manufacturing and excipient costs, and more importantly, sustained-release control.