A Review of Solute Encapsulating Nanoparticles used as Delivery Systems with Emphasis on Branched Amphipathic Peptide Capsules

Abstract

Various strategies are being developed to improve delivery and increase the biological half-lives of pharmacological agents. To address these issues, drug delivery technologies rely on different nano-sized molecules including: lipid vesicles, viral capsids and nano-particles. Peptides are a constituent of many of these nanomaterials and overcome some limitations associated with lipid-based or viral delivery systems, such as tune-ability, stability, specificity, inflammation, and antigenicity. This review focuses on the evolution of bio-based drug delivery nanomaterials that self-assemble forming vesicles/capsules. While lipid vesicles are preeminent among the structures; peptide-based constructs are emerging, in particular peptide bilayer delimited capsules.

The novel biomaterial— Branched Amphiphilic Peptide Capsules (BAPCs) display many desirable properties. These nano-spheres are comprised of two branched peptides— bis(FLIVI)-K-KKKK and bis(FLIVIGSII)-K-KKKK, designed to mimic diacyl-phosphoglycerides in molecular architecture. They undergo supramolecular self-assembly and form solvent-filled, bilayer delineated capsules with sizes ranging from 20 nm to 2 microns depending on annealing temperatures and time. They are able to encapsulate different fluorescent dyes, therapeutic drugs, radionuclides and even small proteins. While sharing many properties with lipid vesicles, the BAPCs are much more robust. They have been analyzed for stability, size, cellular uptake and localization, intra-cellular retention and, bio-distribution both in culture and in vivo.

Introduction

A goal of pharmaceutical research is to deliver drugs at appropriate dosages to target cells or tissues in a safe and reproducible manner.¹ While tremendous progress has been made in defining fundamental biological processes and discovering compounds that interact with these molecules, translation of these findings into advanced therapies has lagged behind. Limitations in delivering therapeutic moieties to selective tissues with minimal off-target damage are often cited for this lapse.^1,2 Novel molecular pharmacological agents with proven biological activities but limited aqueous solubilities or a short circulating half-lives in vivo will face substantive barriers. These delivery impediments limit or reduce therapeutic activity and disfavor potential pharmaceutically interesting compounds from entering clinical drug trials.¹ Conventional modes of drug administration such as pills, eye drops, intravenous solutions, ointments, inhalers etc., fall short in meeting these expectations. The oral route for example is one of the most commonly employed and preferred modes for drug administration due to its minimally invasive nature. However the adequate delivery of peptide and protein drug candidates fuelled by the recent advances in recombinant biotechnology,² are ineffective via this route.^{3, 4} Overcoming numerous barriers including — the acid environment of the human stomach, proteolytic degradation, limited intestinal uptake and first pass effects of the liver . All of these factors reduce, alter and block absorption of almost all biomacromolecular therapeutics.^5,6

Self-assembling nano-carriers (Fig. 1) show promise as delivery vehicles that overcome many of the poor absorption issues that plague certain classes of drugs. Both hydrophobic and hydrophilic small molecules, proteins, and nucleic acids can be bound to or encapsulated within the compartments of these nanomaterials and delivered in vitro or in vivo. The self-assembling carriers can be classified into a number of categories depending on whether they are solid or hollow and the materials that comprise them. They can be further categorized depending on whether they contain natural or synthetic polymers. Polymeric vesicles offer advantages over their natural counterparts (Liposomes) in terms of stability, storage and tune-ability. Drug delivery through polymeric vesicles increases the stability of the drug, extends the circulating time in the blood and can be designed for controlled release. Most polymersomes and peptide amphiphiles have low “critical assembly concentrations” (CAC), a value similar to the critical micelle concentration (CMC), used to describe the minimum concentration needed to form lipid vesicles. Their CAC’s fall in the range of 10⁻⁶ – 10⁻⁷ M, which is 1000 times lower than most surfactants (10⁻³ – 10⁻⁴ M).^7,8 This allows for the retention of payloads for longer periods as well as improved delivery to distal areas of the body. A review of the literature reveals a plethora of compounds encapsulated and successfully delivered in vitro and in vivo using polymeric vesicles. These include various anti-cancer drugs like Doxorubicin and Paclitaxel, proteins like myoglobin, hemoglobin and albumin, fluorescent molecules, plasmids and siRNA. The review by Levine et al. (2008)⁹ summarizes the literature on polymersome research related to cancer diagnosis and therapeutics, giving a more detailed account of the types of polymersomes and cargo that have been developed to that point.

Figure 1.

Classification of Nanocarrier Systems for Drug Delivery

Polymeric vesicles are also useful as diagnostic tools and in optical imaging when they encapsulate fluorescent agents. One example is the encapsulation of a porphyrin-based near infrared (NIR) fluorophore that is able to generate a signal even through a 1 cm solid tumor.¹⁰ When such nanovesicles are injected into an animal, their biodistribution can be monitored using non-invasive NIR optical methods eliminating the need to sacrifice the animal.

The surface of these nanovesicles can be functionalized with various ligand molecules specific for up-regulated surface receptors on diseased cells. Such site-specific delivery reduces potential systemic side-effects observed using traditional delivery routes. Some of the ligand molecules attempted include various antibodies, transactivator of transcription (TAT) peptide and anti-HER-2/neu peptide mimetic (AHNP) peptides.

Nano-vectors have been shown to improve the pharmacological characteristics of these drug moieties. The utilization of nanotechnology for drug delivery has been shown to enhance the delivery of poorly soluble drugs, facilitate drug targeting in a cell/tissue-specific manner and enabled the co-delivery of two or more drugs as well the intracellular delivery of larger macromolecular drugs. By enhancing the efficacy of these drugs, new candidate drugs are advancing in clinical trials with improved safety and effectivness.¹¹ The application of nanotechnology to drug delivery is expected to alter the landscape of pharmaceutical and biotechnology industries in the foreseeable future.^{11,12 ,13,14} As of 2013, almost 250 nanotechnology based medicines had been approved or were in various stages of clinical investigation.^15,16,17

Nanoparticulates are colloidal macromolecular nanocarrier systems ranging from 10–200 nm and deemed as potentially attractive candidates for the entrapment and encapsulation of hydrophobic drugs. Solid nanoparticulates can either be defined as nanocapsules - where the active drug molecule is encapsulated within the carrier; or matrix based nanospheres - where the drug molecules are adsorbed and dispersed throughout the nanomaterial.¹⁸ Nanoparticulates can be engineered using ‘top down’ or ‘bottom up’ methods.¹⁹ In the former, a larger material is broken down into smaller particles whereas the bottom up approach involves the thermodynamically regulated, multi-step synthesis of the nanomaterial in a controlled reaction. ²⁰

A competent nano-carrier functions safely, selectively, reliably delivering a therapeutic at the required dosage to the target site in the appropriate time frame.^3,13,20,21 Ideal nano-carriers possess several desirable attributes: they should reduce the concentrations of the active compound needed since the drugs are no longer systemically distribution²² and the pharmacokinetics of biodistribution are enhanced due to improved tissues and organ targeting.^12,23,24,25 This preferential delivery along with the particle’s ability to release drugs in a sustained or stimuli triggered manner will reduce off-target effects and decrease cytotoxicity. Utilization of nano-carriers has the added advantage of improving the delivery of hydrophobic drugs in water; thus enhancing their delivery through parenteral administration. Nano-carrier based delivery systems have also been shown to improve the half-lives of a wide variety of hydrophobic moieties and peptide drugs.^26,27,28 Moreover, nanotechnology based delivery vehicles composed of biocompatible molecules ^29,30,31,32 are projected as safer alternatives to existing vehicles that have been known to cause peripheral neuropathy and hypersensitivity.^33,34

The search for nano-delivery systems involves numerous design elements. The large repertoire of available nano-carriers, apart the ones that are in development, present nano-systems with a variety of structural, functional and physiochemical characteristics that translate into case specific advantages and/or limitations. These include (i) biocompatibility of constituent material(s); (ii) simple and robust assembly processes; (iii) functionalization / pre-functionalization capability; (iv) intracellular stability and biodegradability; (v) long circulating half-life; (vi) a size compatible with facile cellular uptake, charge density, surface hydrophilicity and flexibility; and (vii) negligible immunogenicity.

It has been indicated that the complexity involved in nanoparticle fabrication and functionalization causes batch to batch variations leading to quality and purity concerns.³ The development of targeted nano-carriers via a single step synthesis mediated by the self-assembly of pre-functionalized biomaterials would serve to alleviate these concerns by simplifying the optimization and the scalable manufacture of these systems.^35,36,37,38

Despite recent advancements in nano-medicine, several challenges remain – 1) overcoming physicochemical and biological hurdles such as low stability, low permeability, short half-life, enzymatic susceptibility, targeting and 2) immunogenicity.⁵ Although a majority of these nano-therapeutic products have improved the pharmaceutical efficacy of clinically approved drugs, nanotechnology as a whole has not yet generated entirely new therapies.¹¹

1. STRUCTURES OF SELF-ASSEMBLING NANOPARTICLES

The phenomenon of bio-molecular self-assembly is a common in nature. Assembled multi-protein complexes are required for a number of cellular functions. Diacylglycerol-phospholipid assembly is responsible for the structure of the various cell membranes. In the context of lipids and peptides, the self-assembling molecules are amphiphilic and are comprised of both hydrophobic and hydrophilic domains. These domains can be spatially separated either along the length of the molecule or along a defined face of the folded molecule.³⁹ The hydrophilic segments can either be uncharged (polar) or charged (cationic, anionic or zwitterionic)⁴⁰. While such segments are spatially segregated in lipids, in amphipathic peptides the primary sequences contain alternating or random segments of hydrophobic and hydrophilic residues. Upon folding into either helices or β-sheets, the final functional structures have continuous faces of hydrophobic or hydrophilic residues that allow them to interact with solvent, a membrane, a ligand or an adjacent molecule.

Self-assembly in bio-macromolecules is usually facilitated by weak non-covalent interactions that can be any one or a mixture of hydrogen bonding, Van der Waal forces, ionic, and electrostatic interactions. When such amphiphilic molecules are introduced into an aqueous environment, the hydrophobic segments anneal to exclude water thereby providing the driving force for assembly aided by inter- and intra-molecular associations. The hydrophilic segments of the molecule remain exposed to the aqueous solvent. Much of our understanding of self-assembly of amphiphiles comes from studies with lipids. Attempts to mimic lipid amphiphiles have led to the design and development of many different types of molecules that retain the amphiphilic character with relatively short acyl-chains.^41,42

Lipid and peptide amphiphiles assemble into a number of nano- and micro-sized structures such as micelles, vesicles or molecular gels⁴³ composed of rods and tubules,⁴⁴ fibrils and fibers⁴⁵ with ordered structures. Micelles adopt spherical, worm like or disk shaped structures.^{46,47,48,49,50,51} Vesicles when composed entirely of lipids are called liposomes, but other non-lipid vesicle types exist: niosomes, proniosomes, polyhedral niosomes, polymersomes and vesicles formed by the self-assembly of peptide amphiphiles (peptosomes).⁵² Finally, structures that do not fall into either of these common categories can be as diverse as, bipyramids, lamellar, hexagonal phase, cubic phase, icosahedral, cage like, high axial ratio microstructures (HARM), ^53,54,55 and myelin fibrils.⁴⁰ The energetics, geometric constraints and strengths of the intra- and inter-interacting forces of the individual molecules determine the final architecture of the assembly. Other factors influencing the final assembly are pH, temperature, ionic strength of the solvent, as well as the concentration of the monomer. Using the packing parameter of the monomer one can predict the favored structure of the assembly.⁵⁶ The packing parameter depends on the area of the head group , molal volume of the hydrocarbon chain or chains ν and the extended chain length l₀ of the monomer:

$$\text{Packing}\text{number}=\frac{\nu }{a_0{l}_0}$$

While the head group of the amphiphiles plays a critical role in predicting the overall shape and size of the assemblies, the tail region directs the formation of spherical micelles, rod like micelles and spherical bilayer vesicles.⁴³ While there are numerous review articles focused on the different morphologies adopted by self-assembling molecules, this review is concerned with studies on spherical aggregates, namely micelles and vesicles.

1.1 MICELLAR ASSEMBLIES

1.1.1 Monomeric Micelles

Micelles are the simplest amphiphilic assembly. They are useful in encapsulating hydrophobic molecules within their hydrophobic core (Fig. 2).

Figure 2.

Illustration of a micelle.

Detergent micelles comprised of sulfated acyl chains (i.e. sodium dodecyl sulfate, SDS) adopt this structure. At concentrations below their CMC, the amphiphiles exist as monomers but as the monomer concentration exceeds the CMC, they begin to assemble into stable spherical structures. These structures are in a dynamic equilibrium with the monomers present in solution. The CMC is affected by further increases in concentration of the amphiphile that can lead to a change in the overall shape (discoidal, rod like or liquid crystalline phase), thus allowing one to tune the overall shape and size of the final assembly. Lipid micelles range from 10–100 nm in diameter and exhibit a core-shell architecture, in which the interior, being hydrophobic, entraps lipophilic drugs.⁵⁷ The CMC and structural features exhibited by micelles are a function of the nature of their hydrophilic and hydrophobic block constituents. Yu and Eisenberg⁵⁸ generated a variety of micellar conformers such as, spheres, rods, vesicles, tubules and lamellae depending on solvent conditions and the relative size of hydrophobic and hydrophilic segments. These different structures when tested for their ability to deliver drugs displayed varied pharmacokinetic properties.^{59,60,61,62,63}

1.1.2 Copolymeric micelles

Copolymeric micelles are relatively new constructs that incorporate biodegradable polymers prepared from block copolymers. Block copolymers are linear repeating units of hydrophobic and hydrophilic residues covalently linked to impart amphilicity to the molecule. K₆₀L₂₀ is one such example of a diblock copolymer where a polypeptide comprised of 60 lysine residues is covalently linked to a polypeptide composed of 20 leucine residues. The block of leucine residues form the hydrophobic segment of the block copolymer while the positively charged lysine residues form the more hydrophilic segment.⁶⁴ Triblock copolymers are similar to diblock copolymers but differ in having an extra hydrophobic or hydrophilic segment that can be fused to either of the existing segments. In general, a polymeric micelle consists of two parts: a core shell composed of insoluble hydrophobic segments and the outer corona composed of the soluble hydrophilic segments (Fig. 3). While simple spherical micelles with a core particle and corona are common, more complex multi-lamellar structures have been recently reported that resemble onion-like micelles,⁶⁵ shell cross-linked micelles,⁶⁶ star shaped Janus micelles that allow for two different types of chemical interactions in the same molecule,^67,68,69 schizophrenic diblock copolymer micelles,⁷⁰ multi-compartmental hamburger micelles.⁷¹

Figure 3. Polymeric micelle.

Self-assembly and polymerization of a block copolymer yields a shell cross-linked knedel polymer assemblies, where the cross-linked outer shell can be decorated with biologically relevant ligands [Reprinted from Adv Drug Deliv. Rev, 56(11), Tu and Tirrell, Bottom-up design of biomimetic assemblies, 1537–1563, (2004)⁴² with permission from Elsevier]

1.1.2.1. PEGylated Micelles

Numerous copolymers have been used to produce micelles, but biocompatibility and biodegradability issues have limited their use in therapeutic applications.⁷² Usually, the preferred choice for the hydrophilic block has been polyethylene glycol (PEG). In most micellar assemblies, the molecular weight of PEG tends to exceed that of the hydrophobic core-forming block.⁷³ This outer hydrophilic corona region tends to become hydrated resulting in a splayed appearance, giving rise to various structures, such as polymer brushes.⁷⁴ These conformations imbue the micelles with new properties that suppress binding of serum proteins and phagocytic attack in blood thereby decreasing clearance by the reticuloendothelial system (RES).⁷⁵ PEGylated co-polymer micelles also have lower CAC’s than traditional surfactants resulting in reduced cytotoxicity.⁷⁶ The highly hydrated corona and the hydrophobic core generate a dielectric gradient that aids in the solubilization of a range of non-polar compounds of varied hydrophobicities.⁷⁷ Micellar delivery systems, like many other ones, positively affect the biodistribution and pharmacokinetics of drugs by means of increased circulating half-lives and increased tumor accumulation.⁷⁸

1.1.2.2. Poly(L-amino acid) Micelles

Poly(L-amino acid) micelles are formed from poly(L-Histidine)-PEG block co-polymers combined with poly(L-lactic acid)-PEG (PLLA-PEG). These micelle have been investigated as pH sensitive drug carriers for cancer therapy.^79,80,81 Cancerous cells - due to a higher rate of metabolism - tends to have intracellular pH values < 7.2. ^82,83,84 The imidazole side-chain of histidine has a pKa in this range; which leads to an increase in histidine hydrophobicity, destabilization of the micelle and subsequent drug release. The pH sensitivity of Poly(L-amino acid) micelles can be modulated by varying the %wt. of the poly(L-lactic acid)-PEG composition. Moreover, Poly(L-histidine) (PLHS) appears to have fusogenic activity with endosomes facilitating cytosolic drug release in cancer cells. More complex systems that incorporate phenylalanine ([poly(His-co-Phe)]-PEG) and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-polyethylene glycol-2000(DSPE-PEG) have recently been formulated to decrease the pH sensitivity of the pure poly L-histidine and also act as pH sensitive nanocarriers for cytosolic drug delivery.⁸⁵ Drugs have been conjugated to these poly(amino acid) copolymers via chemical modifications of the drugs, but not without concerns relating to drug decoupling at the target site.^86,87,88 Also, immunogenicity is a potential concern with these systems upon the increase in the number and diversity of amino acids used.⁷¹

1.1.2.3 Micelles composed of mixtures of cationic and anionic surfactants

The preparation of micelles using two surfactants of opposite net charge is well established.^89,90,91 Charge repulsion of the head groups decreases the free energy of assembly.⁹² The size and net charge of the detergent pair determines the CMC as well as the size of the resulting micelles.⁹³ In a single report the mixing of certain oppositely charged surfactants can lead to the spontaneous formation of single walled vesicles.⁹⁴ The size, charge and permeability of such vesicles was controlled by adjusting the molar ratios of the two surfactants. One example of such a charge balancing mixture is cetyl trimethylammonium tosylate (CTAT) and sodium dodecylbenzene sulfonate (SDBS).

1.2 NANOPARTICULATE ASSEMBLIES

1.2.1 Solid Lipid Nanoparticles

Solid Lipid Nanoparticles (SLNs) are members of a class of nano-scale carriers consisting of a central solid hydrophobic core comprised of physiological lipids emulsified with an aqueous surfactant.⁹⁵ Hydrophobic drugs are dissolved in the solid hydrophobic core.⁹⁶ These nanoparticles due to their narrow size range (100 – 200 nm) evade the RES and cross the blood-brain barrier.⁹⁷ The biodegradable properties of these lipids minimize their cytotoxicity. SLNs also demonstrate stable drug entrapment, especially in case of very hydrophobic drugs and provide controlled release lasting several weeks. SLN can be conjugated to hydrophilic polymers and/or surfactants to minimize their hepatic uptake and improve bioavailability.⁵⁸ SLNs can be stabilized with stearic acid - PEG 2000, however cytotoxic effects have been observed upon the release of free stearic acid.⁹⁸ By choosing lipids with varying chain lengths and degrees of unsaturation the loading capacity of SLNs can be controlled. Overall, SLNs appear to be safe delivery systems for hydrophobic drugs especially to the brain and production is scalable with excellent reproducibility.^99,100

1.2.2 Amphiphilic Dendrimers

Dendrimers are 1–10 nm sized hyper-branched synthetic polymers comprised of well-defined branched oligomers attached to a central core.^101,102.103 Dendrimers are synthesized using convergent or divergent approaches.¹⁰³ The former approach capitalizes on the symmetric nature of the dendrimer with synthesis commencing at the periphery and terminating at the core; while the divergent approaches builds out from the core.¹⁰⁴ Dendrimers display low poly-dispersity despite their large molecular mass. The branching of the dendrimers generates semi-globular/globular structures that are easily functionalized.¹⁰³ Glycol-, peptide- and silicon-capped dendrimers have been synthesized using carbohydrates, peptides and silicon.^39,105 Drugs can associate with dendrimers through physical entrapment within the void spaces or by attachment of the pro-drug to functional groups positioned on the dendrimer surface.⁹⁴

Amphiphilic dendrimers (Janus dendrimers) are capable of self-assembling into stable bilayer vesicles, referred to as dendrimersomes¹⁰⁶. A Janus dendrimer is formed by covalently linking two distinct dendritic building blocks. They can be made amphiphilic by attaching a non-polar block (e.g. alkoxy gallate ether to a polar block such as aliphatic arborol).¹⁰⁷ A review by Rosen et al. (2009)¹⁰⁸ details the self-assembly and disassembly of dendrons. Figure 4, is an illustration reproduced from the above review.

Figure 4. Convergent and divergent synthesis of dendrimers and dendrons.

Reprinted (adapted) with permission from Rosen et al. (2009)¹⁰⁸, Chem Rev., 109(11), 6275–6540. Copyright 2009 American Chemical Society.

Dendrimers have been studied for the delivery of fluorouracil¹⁰⁹ and indomethacin.¹¹⁰ Dendrimer size has been exponentially correlated with the duration of extravasation across the endothelium, with larger dendrimers indicating faster extravasations.¹¹¹ The positive charges on the polyamine and/or polyamide linkages used in some dendrimers pose a potential risk for cellular toxicity and immune activation. Partial derivatization of the dendrimer surface with PEG and/or fatty acids is known to mitigate these concerns by shielding the positive charges on the surface.^112,113

1.3 VESICULAR ASSEMBLIES

In contrast to the different micellar constructs discussed above where drugs are trapped within the micellar matrix, the ability of vesicles to encapsulate solutes offers increased versatility. Their hollow interiors permit encapsulation of higher concentrations of cargo, thus making them potentially better drug delivering vehicles. Vesicles are colloidal aggregates that are most often spherical. In biology, lipid vesicles are bilayer-delimited chambers that entrap solvent during the self-assembly process. The solvent is water containing dissolved solutes. In terms of drug delivery, solutes including peptides, proteins, genetic material, or bioactive compounds like anticancer drugs can be encapsulated. It can also trap non-polar or hydrophobic substances within their membrane, thereby increasing their utility. Such versatility has made them the most attractive choice for use in biology. Apart from its biological/medical uses, vesicles can be used as micro-chemical reactors and have found application in the cosmetic and food industries.¹¹⁴ Lipid based vesicles are the most extensively studied drug carriers used today. Until recently, lipids were the only choice for generating artificial vesicles. However with the introduction of many new self-assembling molecules, one can choose from a variety of molecules. Figure 1, classifies the various polymeric vesicles described to date. It is difficult to precisely classify all of the various polymeric vesicles due to the issue of multi-functionality. This is particularly true with the introduction of hybrid polymers, where two or more different materials are used to make an amphiphile such as combining the synthetic polymer polyethylene oxide (PEO) to either a polypeptide or a diacylglycerol phospholipid. These hybrid polymers can improve solubility, increase circulating half-lives in the bloodstream and allow for the surface functionalizing with desired targeting ligands. Such a molecule could be classified as a block

Copolymer, lipid vesicle or peptide vesicle depending on the proportions of the individual components and/or which component is driving the assembly.¹⁰³ Employing natural biopolymer vesicles as drug carriers over synthetic and semi-synthetic polymer vesicles improves their biocompatible and biodegradable. Lipids, polysaccharides and peptides are all capable of self-assembling into vesicles and other nano- and micro-sized structures.¹¹⁵

1.3.1 Liposomes

Liposomes or lipid vesicles are possibly the oldest and most widely studied drug delivery systems and have been in use since the 1960s.^116,117 These artificially produced vesicles range from 30 nm to several micrometers in size and consist of an aqueous core surrounded by uni- or multi-lamellar membranes formed from the supramolecular-assembly of primarily phospholipids (Fig. 5) and cholesterol. Numerous studies have promoted the use of these structures.^118,119,120

Figure 5. Illustration of a Liposome.

By courtesy of Encyclopaedia Britannica, Inc., copyright 2007; used with permission. [http://www.britannica.com/EBchecked/media/92244/Phospholipids-can-be-used-to-form-artificial-structures-called-liposomes].

Liposomes can be prepared from a variety of diacyl phospholipids including: 1,2-Dimyristoyl-sn-glycero-3-phosphocholine (DMPC), 1,2 Dipalmitoyl-sn-glycero-3-choline (DPPC), and 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC), palmitoyl-2-oleoylphosphatidylethanolamine (POPE) among others. As an example we show POPC:POPE vesicles generated in our laboratory as well as their resizing after extrusion. (Fig.6). Most of our knowledge on vesicle assembly, dynamics and physical properties come from studies on these assemblies. Since then, many different and diverse vesicles have been designed and developed. Unlike polypeptides and polysaccharides, lipids are not polymers. Liposomes have found uses in many biological, pharmaceutical and cosmetic applications. They are employed in studies related to cell physiology, as model cell membranes and drug delivery vehicles.

Figure 6. Electron microscopy pictures of POPC:POPE (6:4) liposomes made using extrusion.

Seabra, M.B. (2006) Studies of a Channel-Forming Peptide Inserted into Liposomes formed by POPC:POPS and POPC:POPE, (M.S. Thesis Kansas State University).

Their utility lies in their ability to fuse with cellular membranes¹²¹ and lipid bilayers membranes or enter cells through clathrin-mediated endocytosis.^122,123, The properties of liposomes have been modulated to produce variation in size, lipid composition, surface charge and other characteristics.⁷⁷ The aqueous core of the liposomes is known to encapsulate a large payloads of hydrophilic and moderately hydrophobic drugs; and their ability to naturally associate with tumors and the EPR effect has led to the development of numerous FDA approved drugs based on the liposome platform. ^72,100,124

Numerous studies have been conducted to surface modify classical liposomes in an effort to increase their targeting capabilities and circulating half-lives. These include the introduction of linear dextrans,¹²⁵ gangliosides containing sialic acid,¹²⁶ lipid derivatives of hydrophilic polymers such as polyvinyl alcohol,¹²⁷ polyethylene glycol^128,129 and poly-N-vinylpyrrolidones.¹³⁰ These modifications stabilize and protect them from uptake by the mononuclear phagocytic system (MPS). Targeted therapy has also been demonstrated using liposomes conjugated to monoclonal antibodies via a PEG linker,¹³¹ and protease-sensitive polymer-caged liposomes have been developed to enable selective targeting and drug release at the cancerous site by exploiting the natural tendency of affected cells to produce cancer associated proteases to destabilize the liposome.¹³²

Liposomes also appear to be the preferred carriers for purposes of radionuclide based targeting for cancer therapy. A variety of liposomes such as multi-vesicular liposomes (MUVEL) - small vesicles containing radionuclides trapped into large liposomes, polymer coated long circulating liposomes - with low bilayer permeability and low lipid exchange, sterically stabilized liposomes (SSL) - with high load capacity and tumor affinity, etc. have been developed exclusively for this purpose.^{75,133,134,135} The therapeutic efficacy of targeted radiotherapy is due to the tumor’s absorption of alpha (α) or beta (β) radiation emitted by the radionuclide. β-emitters such as ⁹⁰Y, ³²P, 89Sr, ¹⁸⁶Re, ¹⁵³Sm, ¹⁷⁷Lu, and ¹³¹I are by far the most widely utilized radionuclides as therapeutics and alleviation of bone pain.¹³⁶ β-electrons have low linear energy transfer (LET) values and long path-lengths. They can pass through tissue, and interact with atoms via energy loss causing ionization, generating free radicals thus causing DNA damage by inducing single strand breaks. On the other hand α-particles have high LET values and shorter path lengths, and are used to generate more localized cellular effects with high chromosomal damage during mitosis and irreparable double strand DNA breaks. Short half-life α-particles emitters such as ²²⁵Ac, ²¹¹At and ²¹³Bi are commonly employed for targeted alpha particle therapy (TAT).^137,138 Naturally varying considerations exist in the selection of liposomal carriers based on whether they are employed for beta- or alpha- radiation therapy. For instance studies relating to the effect of surface charge of the liposomes on the radionuclide delivery demonstrated that the use of neutral lipids such as DMPC-Cholesterol in liposomal preparation substantially increased the effective maximum absorbed dosage of beta emitters such as ³²P, ⁶⁷Cu, ⁹⁰Y and ¹³¹I at the tumor site as opposed to cationic DC-cholesterol lipids. On the other hand a study performed on cholesterol stabilized PEGylated liposomes with ²²⁵Ac and ²¹³Bi showed high retention of the radionuclide and daughter isotopes only where larger cationic liposomes were involved.¹³⁹

Generally speaking, due to the low LET values exhibited by beta electrons, considerations of liposomal rupture due to beta-emissions are not as critical as in the case of alpha emitters where the high LET values of alpha particles coupled with the high energy recoil generated during the formation of daughter nuclides can damage and rupture the liposomal membrane. Moreover, the greater availability of beta-emitters for radiotherapy enables a greater choice in the selection of radionuclides as opposed to those employed in alpha-particle therapy. The ²²⁵Ac used in most alpha particle therapies has a chemistry that is not well understood; and that precludes the effective utilization of these emitters in chelate complexes and from integration in larger compounds for the purposes of liposomal encapsulation. Time is yet another factor to be considered in case of radiotherapy. Unlike conventional drugs, radioactive moieties can rapidly decay over time, thus losing therapeutic potency. This is especially true in cases of radionuclides such as ²¹³Bi (t_1/2 = 45.7 min), where the long times taken for liposomal preparation come at the expense of decreasing therapeutic dosage.

Despite their versatility, biocompatibility, relative non-toxicity and wide application platform, liposome preparation is lengthy and tedious, and preparation steps have to be very carefully monitored to ensure reproducibility in size and entrapment efficiency.¹²⁸ Moreover, liposomes have been demonstrated to alter the pharmacokinetic properties of the drugs and are prone to systemic leakage.¹⁴⁰ All of these parameters must be factored in while selecting liposomes as candidates for therapeutic delivery.

1.3.1.1 Polysaccharide vesicles

This class of vesicles is less common. Dextran, chitin, chitosan and their derivatives have been investigated for their use as drug carriers. Most of the vesicles in this category are modified polysaccharides adducted with either small (trimethyl-) ^141,142 or larger (3-pentadecylphenol-) ^143,144 hydrophobic groups to render them amphipathic. Chitosan derived polymer vesicles are the most studied in this category. Due to their membrane-penetrating ability, these vesicles, palmitoyl glycol chitosan vesicles, are a good choice for intranasal, or oral administration of gut-labile molecules.^145,146

1.3.2 Niosomes

Niosomes or Non-ionic surfactant vesicles (NSVs) are nanoscopic lamellar structures that self-assemble into closed bilayers upon hydrating a preparation of non-ionic surfactants such as alkyl or dialkyl polyglycerol ethers; cholesterol and a charge-inducing agent.¹⁴⁷ They were first reported in the 1970s as a feature of the cosmetic industry.^148,149 Niosomes were developed and patented by L’Oreal in the 1970s and 1980s and introduced in 1987 by Lancôme.^150,151 Examples of non-ionic surfactants used to generate niosomes include, C₁₆ monoalkyl glycerol ether, sorbitan esters, and others.

Niosomes are capable of entrapping hydrophobic and hydrophilic solutes.¹⁵² They can be unilamellar or multilamellar depending on the method of preparation. In many ways Niosomes are liposome analogs and are prepared much like liposomes through non-spontaneous processes involving the input of energy in the form of heat, ultrasound, physical agitation, application of pressure or a combination thereof.¹⁵³ Consequently, just like liposomes, most NSV preparation methods involve some hydration of the non-ionic surfactant at an elevated temperature followed by an optional size reduction to obtain the colloidal suspension.¹⁵⁴ Negatively charged molecules such as dicetyl phosphate (DCP) and phosphatidic acid, or positively charged molecules such as cetylpyridinium chloride and stearylamine (SA) can be added (2.5 – 5 mol%) to prevent aggregation and stabilize the niosomal bilayer.^152,153

There are some fundamental distinctions. While liposomes are made up of neutral or charged double chained phospholipids; niosomes are made up of uncharged single chain surfactants. And since non-ionic surfactants are more stable to enzymatic attack and resistant to air-oxidation than phospholipids, preparing quality niosomes is quite simple.¹⁴⁸ Liposome preparations are expensive owing to their need for special storage and handling due to their predisposition to breakdown when exposed to air.

When a water-soluble carrier biomolecules such as sorbitol, sucrose stearate, maltodextrin, etc. are coated with a thin film of dry non-ionic surfactant, the resulting preparation is termed ‘proniosomes’.¹⁵⁵ These proniosomes when hydrated generate niosomes. Since proniosomes are obtained as a dry powder; they can be formulated to make beads, capsules or tablets with convenient unit dosing and demonstrate reduced aggregation, fusion, leakage and increased drug entrapment efficiency.^150,156

Niosomes, due to the presence of lipophilic and hydrophilic domains can incorporate drugs with a wide range of solubility, and indicate potential applications for the delivery of numerous pharmacological agents.¹⁵⁰ Niosome preparations have been investigated for the purposes of drug, gene and vaccine delivery through parenteral, oral, ocular, pulmonary and transdermal routes. The breadth of these studies precludes them from comment in this thesis. Marianecci et al.¹⁵⁴ provides an excellent and comprehensive review on the numerous potential applications of niosomes. Another interesting property of niosomes is their ability to deliver drugs via topical application. Dermal / transdermal drug delivery is a less invasive route for the localized delivery of high concentrations of drugs, bypassing the limitations associated with systemic circulation and/or gastrointestinal degradation. The stratum corneum of the epidermis works as a barrier for permeation and consequently severely restricts the administration of most drugs through transdermal routes.¹⁵⁶ Niosomes loaded with drugs demonstrate enhanced permeation characteristics and niosomes formulations demonstrated greater skin permeation of enoxacin and higher stability of tretinoin when compared to their respective liposomal counterparts.^{157, 158} Acid sensitive niosomes generated with Span 60, cholesterol and cholesteryl hemisuccinate (CHEMS) were proposed for the topical delivery of ibuprofen and demonstrated a significant increase in drug penetration through skin.¹⁵⁹

Despite the numerous advantages niosomes have over liposomes, niosomal preparation is generally exhaustive, utilizes numerous moieties, involves the application of non-spontaneous energy consuming processes and takes considerable time.

1.3.3 Polymeric nanocapsules (Polymersomes)

Block copolymers are synthetically linked segments of hydrophobic and hydrophilic polymeric units to generate the amphipathic molecule. The segments of block copolymers can be entirely synthetic, semi-synthetic or composed of biopolymers like polypeptides. Like phospholipids, these amphiphiles self-assemble into various ordered structures.^58,160,161, Vesicles formed from block copolymers are commonly referred to as polymersomes and are among the best characterized of the polymeric vesicles. By controlling the formula weight of the blocks, polymersomes with different properties are generated, including ones that vary in- elasticity, permeability and mechanical stability. By their very nature, polymeric building blocks allow for greater chemical diversity than lipids.¹⁶² Due to their higher molecular weights, polymersomes are inherently thicker, tougher and more stable than conventional liposomes. It has been shown that polymersome membranes sustain dilatational strains up to 40–50% when compared with ~5% or less for lipid membranes.¹⁶³ A coarse grain molecular dynamics simulation supports the experimental observation that while lipid membranes require only about ~ 5% strain rate for rupture, while it takes 20% or more to rupture copolymer membranes.¹⁶⁴ Also polymersomes (Fig. 7) can be designed to be responsive to environmental factors such as pH, temperature, redox potential, magnetic field, light and ultrasound.¹⁶² There are a number of comprehensive reviews on this topic.^165,166 Numerous variants of the block polymers have been tested in vitro for their ability to encapsulate and deliver cargo.

Figure 7. Polymersomes derived from asymmetric block copolymers.

Reprinted (adapted) with permission from Meng et al. (2009) Biomacromolecules, 10(2), 197–209. Copyright 2009 American Chemical Society. ¹⁵⁶

1.3.3.1 Peptide containing vesicles

Peptides derivatives linked to synthetic molecules form many different nanostructures: tubes, fibers, rods, micelles, vesicles, doughnuts, bilayers and others.^{167,168,169,170,171,172,173}

Vesicular structures, composed entirely of pure peptides, are relatively rare. Block copolymer monomers composed entirely of polypeptides (i.e. Ac-V_mK_n-NH₂, Ac-G_mD_n-OH, Ac-V₆D-OH, or Ac-KA₆-OH) have been reported to form vesicles.^140,174,175 Upon surveying the published works describing peptides that form vesicles, few were found compared to the many synthetic polymers capable of self-assembling into vesicles. The field of peptide vesicles is still in its infancy but growing at a rapid pace. Vauthey et al. (2002)¹⁷² were the first to show that a simple 7–8 residue amphiphilic peptide is capable of self-assembling into nanotubes and nanovesicles. They called these peptides, “Surfactant-Like Peptide”. Santoso et al. (2002),¹⁷⁴ designed similar peptides as the ones by Vauthey et al. (2002)¹⁷² with glycine and aspartic acid capable of self-assembling into nanotubes and nanovesicles.

Peptide amphiphiles capable of self-assembling into nanovesicles can be a simple polypeptide sequence, or a more complex molecule, characterized by the addition of an alkyl chain or with a hydrophobic amino acid segment joined to a hydrophilic sequence (Fig. 8).

Figure 8. Peptide amphiphile structure.

The basic structural components of peptide nanoparticle components: (1) a hydrophobic tail, (2) lower, rigid part, including 4 cysteine residues (3) upper, flexible part, including 3 glycines, a phosphoserine group and an integrin-binding motif, RGD. "Reprinted (adapted) with permission from Tsonchev et al. (2004).¹⁷³

The associating and stabilizing forces that direct lipid vesicle formation also apply to other polymeric vesicles comprised of long acyl chains attached to peptide amphiphiles. Some examples of designer peptide amphiphiles that form nano structures include G_nD₂, A₆D, V₆D, V₆D₂, A₆K, I₃K, A₆K₂, GAVILRR, A₂V₂L₃WE_{(2 or 7)} and similar peptides. In addition, amphiphilic block macrocyclic-copolypeptides have been used to generate bioactive nanovesicles that incorporate the arginine rich human immunodeficiency virus type-1 Rev protein to target the assembled vesicles. The macrocyclic protein sequence also includes tryptophan residues that add stability potentially through Pi-Pi stacking interactions.¹⁷⁶ The Sherman group employed a host-guest approach to drive self assembly using the macrocycle Cucurbit[8]uril (CB[8]). The ternary host complex formed by CB[8] bound two guest pyrene ligands yielding a tight binding constant with a CAC of 12 μM. This assembly generated larger vesicles ~150–250 nm in diameter.¹⁷⁷ Subsequently the same group was able to load the vesicles with basic fibroblast growth factor, retain it and subsequently release it for uptake by fibroblasts¹⁷⁸. A third group was able to prepare bilayer-delimited vesicles in the 60 nm size range using a simple peptide Ac-Ala-Ala-Val-Val-Leu-Leu-Leu-Trp-Glu₂-COOH under acidic conditions¹⁷⁹. The SA2 peptide has a CAC of around 50 μM and was able to entrap calcein.¹⁸⁰ These vesicles not only carry the advantage of being more biocompatible and biodegradable than other synthetic and semi-synthetic polymer vesicles and they where shown to be more stable than lipid and polysaccharide vesicles¹⁷⁹. Recently another group studied the same SA2 assemblies using biophysical and computational tools to produce a visual model of the vesicles.¹⁸¹

Reviews by Zhao et al. (2010)¹⁷⁵ and Zhang, S (2002)¹⁸² summarize the various peptide amphiphiles that form nano structures and discusses a more exhaustive list of the various peptide amphiphiles. Some studies reported the use of oligo valines to promote inter-digitation, the mixing of oppositely charged peptides to reduce the critical aggregation concentration (CACs), and hydrophobic sequences containing complimenting small and large amino acids.¹⁸³ The hydrophobic chain length plays a critical role in the final assembly. Typically, the size of the hydrophobic tail in such peptide amphiphiles is about 3–9 hydrophobic residues. The length of the hydrophobic tail can determine the thickness of the outer layer of such nano structures but with longer chain lengths comes the issue of solubility and conversely, shorter chain lengths decrease the propensity of aggregation. Hence a balance of chain length is desired for the desired property. This is achieved by increasing or decreasing the length of the hydrophilic segments of the molecule.

1.3.3.1.1 Branched amphiphilic peptide capsules (BAPCs)

These structures represent a novel class of nanocarriers. First described by Gudlur et al.¹⁸⁴ in 2012, they are self-assembling structures where a water filled cavity is surrounded by an amphiphilic branched peptide bilayer. They are essentially stable, biocompatible nanocarriers that are structurally similar to liposomes and polymersomes. The properties, applications and biophysical characterization of these carrier systems will be discussed at length in the next sections.

2. PEPTIDE VESICLES STABILIZED BY BETA-SHEET SECONDARY STRUCTURES

Most self-assembling peptide based polymeric vesicles either lack a defined secondary structure or adopt helical conformations in their assembled structures. Recently, however, vesicular^185,186,187 and micellar^160,188,189 assemblies stabilized by β-sheets have been reported. Gebhardt et al. studying poly(butadiene) -poly(L-lysine) co-block polymers elucidated the effects of secondary structure on the morphology of vesicles.¹⁸⁷ In the case of poly(butadiene)₁₀₇-poly(L-lysine)₂₇ coblock polymers, the transition from α-helix to β-sheet takes place at a pH above the pKa (>10.5) of the lysine side chains. This transition resulted in a slight increase in the hydrodynamic radii of the vesicles and the overall assembly remained intact as determined by dynamic light scattering (DLS) and static light scattering (SLS). The formation of parallel β-sheets between the corona chains at the vesicular interface served to release interfacial curvature by creating a flatter interface. A molecular dynamics simulation study on the self-assembly of peptide amphiphiles¹⁹⁰ showed an interplay between hydrophobic interaction between the alkyl chains and hydrogen bonding between the peptide blocks, resulting in assemblies that were spherical β-sheets, micelles with β-sheets in the corona, and long cylindrical fibers (Fig. 9).

Figure 9. Snapshots from molecular simulations of peptide amphiphiles.

“(a) The spherical micelle, (b) the micelle with β-sheets on the outside forming the corona, (c) the β-sheets, and (d) the fiber aggregate”. [Reprinted (adapted) with permission from Velichko et al. (2008)¹⁹⁰, J. Phys Chem B., 112(8), 2326–34. Copyright 2008 American Chemical Society.]

3. STUDIES ON BRANCHED AMPHIPHILIC PEPTIDE CAPSILES

3.1 ORIGIN OF THE BAPC SEQUENCES

The concept of building diacyl phospholipid orthologs made up of just amino acids is not particularly new, however identifying a functional sequence(s) is. The sequences used for the self-assembling branched amphipathic peptide vesicles evolved from earlier studies on several adhesive peptides. Previously, several peptide constructs were identified that produced nano-fibrils with mechanical adhesive properties due to entanglement.^191,192 The hydrophobic core sequence used in the adhesives occurs in nature as an internal fragment of the human dihydropyridine sensitive L-type calcium channel segment, CaIVS3 (DPWNVFDFLIVI-GSIIDVILSE). The underlined segment is a portion of one of the transmembrane helices that contribute to the central water-filled pore of this channel.^193,194 Lyophilization of the synthesized CaIVS3 peptide resulted in insoluble clumps resistant to suggested that strong cohesive forces were driving this association. An α to β transition was observed that could be controlled predictably by adding charged amino acid segments to both ends of the underlined sequence (h₉) (Table 1) and by mechanical disruption. This property suggested that strong cohesive forces drove this association. An α to β transition was observed that could be controlled predictably by adding charged amino acid segments to both ends of the underlined sequence (h₉) (Table 1) and by varying the pH.^191,192 An assay was used that measured the strength (MPa) required to shear (pulled at 180°) apart two glued cherry wood strips treated with a 4% w/v solution of the peptide dissolved in water and then adjusted to the indicated pH values before hot pressing the samples (Fig. 10). The sequences with the flanking tri-lysine or tri-diaminopropionic acid (DAP) segments performed best at pH 12.0, conditions in which the amino groups of the Lys and DAP residues are mostly deprotonated thereby increasing the hydrophobicity of the peptides. The DAP amino acid is a shortened lysine analog with just one methyl group on its side chain as compare to the four methyl groups in lysine. There were no statistical differences between the two different cationic residues or increasing the number of h₉ hydrophobic segments.

Table 1.

Adhesive peptides sequences.¹⁹¹

Peptide	Sequence	MW (Daltons)
Eh9E	EEEFLIVIGSIIEEE	1,748.9
Kh9E	KKKFLIVIGSIIEEE	1,746.1
Eh9K	EEEFLIVIGSIIKKK	1,746.1
Kh9K	KKKFLIVIGSIIKKK	1,743.3
(Kh9)2K	KKKFLIVIGSIIKKKFLIVIGSIIKKK	3,084.1
(Kh9)3K	KKKFLIVIGSIIKKKFLIVIGSIIKKKFLIVIGSIIKKK	4,428.8
Xh9X	XXXVLIVIGSIIXXX	1,490.8

Abbreviations: X = diaminopropionic acid

Figure 10. Shear strength of peptide adhesives measured in Mega Pascals (MPa) with different tri-residue flanking the h₉ sequence.

E, K and X represent the tripeptides (Glu)₃, (Lys)₃ and (diaminopropionic acid)₃, respectively. Peptide solutions-- 4% (360 μL) were spread onto marked 8.0 cm by 2.0 cm areas on one side of two separate strips of wood. The coated strips sat for 15 minutes at RT, before being pressed together for 5 minutes at 130 °C and 1.4 MPa • cm⁻¹.using a Hot Press Model 3890 Auto M (Carver Inc., Wabash, IN). The glued strips were then conditioned for 3 days at 50% relative humidity and 23 °C before being cut into test sized pieces. Reproduced from Shen et al. ¹⁹¹

Table 2 summarizes the shear strength and hydrophobicity for a number of FLIVIGSII derived sequences. The two longer sequences simply incorporated the naturally occurring C- or N-terminal hydrophobic residues present in the CaIVS3 sequence. The shorter sequences represent truncated versions of the h₉ sequence and in some cases include single amino acid substitutions. There are three h₅ sequences (alpha, mezzo and omega) derived by frame shifting a five-residue window within the h₉ sequence. In addition, four unrelated β-forming sequences were tested. The first three Aβ_16–21, Aβ_30–35, and IAPP_22–28 were reported nucleation sites for amyloid fiber formation and have sequences that closely resemble that of K₃h_5αK₃. The last sequence, KAE₁₆-IV, contains both basic and acidic residues and should not display any pH dependence.

Table 2.

Correlation of shear strength and hydrophobicity for a number of peptides showing adhesive properties.¹⁹²

Peptides	Sequences	Shear Strength (MPa)	Gavg (kcal/mol)
K3h9K3	KKKFLIVIGSIIKKK	3.0 ± 0.1	−0.46
K3h11K3	KKKVFFLIVIGSIIKKK	2.9 ± 0.2	−0.48
K3h12K3	KKKFLIVIGSIIVILKKK	2.2 ± 0.1	−0.50
K3h7K3	KKKFLIVIGSKKK	2.6 ± 0.2	−0.42
K3h5αK3	KKKFLIVIKKK	3.3 ± 0.2	−0.62
K3h5MK3	KKKIVIGSKKK	3.4 ± 0.2	−0.37
K3h5ΩK3	KKKIGSIIKKK	3.2 ± 0.2	−0.31
K2h5αK2	KKFLIVIKK	2.8 ± 0.1	−0.53
K2h5ΩK2	KKIGSIIKK	2.4 ± 0.2	−0.31
K3h5MLK2	KKKIVLGSKKK	3.4 ± 0.2	−0.36
K3h5MAK2	KKKIVAGSKKK	2.7 ± 0.1	−0.35
Aβ16–21	KKKLVFFAKKK	3.4 ± 0.3	−0.50
Aβ30–35	KKKAIIGLMKKK	3.4 ± 0.1	−0.39
IAPP22–28	KKKIGGAILKKK	2.9 ± 0.4	−0.30
KAE16-IV	(KA)4(EA)4-CONH2	1.4 ± 0.1	+0.62

Samples were prepared as described in Figure 10.

With the exception of KAE₁₆-IV, all of the sequences showed adhesive properties when dry pressed at 130 °C. The shorter h_5α containing sequence required the most force to break the adhesive. It was determined that this shorter sequence had the lowest free energy for the packed assembly and was the most hydrophobic. While reducing the size of the hydrophobic core sequence increased strength, reducing the number of flanking lysines from three to two had a negative effect on adhesive strength. All of the CaIVS3 derived sequences acted as adhesives even when the dry press temperature was reduced, although with decreased strength. However the amyloid derived sequences displayed no adhesive properties if heating was omitted during pressing. Several of the sequences showing strengths above 3 MPa were tested for their wet strength after emersion for 48 h (Table 3).

Table 3.

Comparison between dry and wet strength for select peptide adhesives. ¹⁷⁶

Peptide Sequence	Dry strength (MPa)	Wet strength (MPa)
K3h5αK3	3.3 ± 0.2	1.13 ± 0.3
K3h5ΩK3	3.2 ± 0.2	0.66 ± 0.2
K3h5MK3	3.4 ± 0.2	0.0
K3h9K3	3.0 ± 0.2	0.67 ± 0.1

All samples were used at 4% w/v and hot-pressed at 130°C. Glue samples were soaked in water for 48 h before wet strength test. Adapted from Mo et al. 2008.¹⁹²

Structural analyses were conducted on both soluble peptides (CD) and dry pressed peptide (FTIR) processed at pH 12.0. All of the dry pressed peptides showed β-structure. Only K₃h_5αK₃ and K₃h₉K₃ showed β-structure in water. K₃h_5ΩK₃ and K₃h_5MK₃ showed β-structure only in ethylene glycol. No structure was observed for the amyloid forming peptides with either solvent.

In parallel, transmission electron microscopy was used to visualize the assemblies being generated at pH 12.0 (Fig 11). The images taken at pH 12.0 show entangled nano-fibers. Regarding the adhesive properties the peptides, they only bonded with rough surfaces. When the peptides were used to glue smooth glass slides together shear strengths of just ~1.5 MPa were required to separate the two pieces. If the slides were siliconized the peptides showed no effect in joining the glass slides.

Figure 11. TEM images of K₃FLIVIK₃ that were dried from peptide solutions prepared at different pH values.

The left panel is peptide suspended at pH 12.0 and imaged at 70,000 x magnification. The center panel is peptide imaged at 34,000 x. The right panel is peptide dissolved at pH 7.0 and imaged at 70,000 x.

Table 4 compares the different properties of the nano-fiber forming sequences. It is clear that the hydrophobic core sequences of FLIVIGSII and FLIVI have unique structural properties, in particular their ability to associate and assemble as a β-structure in water and their wet adhesive strength. Even though the K₃FLIVIK₃ sequence appeared unstructured in water at pH 2.0 and 7.0, transmission electron microscopy images were taken these sequences that were applied as solutions to the TEM grids and allowed to dry (Fig. 11).

Table 4.

Comparison of structural and adhesive properties of test sequences.

FLIVIGSII, FLIVI	IVIGS, IGSII	IFGAIL, KLVFF, AIIGL
random structure at pH 2 and 7	random structure at pH 2 and 7	random structure at pH 2 and 7
β-structure at pH 12 in water	β-structure at pH 12 in ethylene glycol	no structure at pH 12 in any solvent
β-structure when dried from water	β-structure when dried from water	β-structure only when dry pressed
assembles in water pH > 9	no assembly at pH 12	no assembly at pH 12
adhesive after dry press, RT	adhesive after dry press, 130 °C	adhesive after dry press, 130 °C
measurable wet adhesive strength	no wet adhesive strength	no wet adhesive strength
forms entangled nano-fibrils	forms entangled nano-fibrils	forms entangled proto-fibrils

All sequences were tested in parallel for adhesive strength under different dry press conditions. Assembly properties were measured by analytic ultracentrifugation.

The images show two distinct structures-- at pH 7.0 the peptides associate to form a soot-like fractal while at pH 2.0 they resemble lipid micelles. Upon viewing the pH 2.0 structures and recalling that linear lipid detergents or soap amphipaths make similar micellar structures and that branched lipid amphipaths such as diacylglycerol-phospholipids assemble into liposomes, the idea of altering the sequence from a linear to a branched one with all of the charges or one side was born.

3.2 SHARED PROPERTIES BETWEEN BAPCS AND LIPID VESICLES

To generate an all peptide bilayer delimited vesicle, the original design included 5 lysines at the C-terminus with the N-terminal lysine used to generate the branch point. The two hydrophobic segments were added simultaneously using standard Fmoc chemistries. It was decided from the beginning to include a mixture of two different sized peptides such that the shorter sequence would compensate for any strain due to curvature. These vesicles were formed from the two 15- and 23-residue branched amphiphilic peptides:bis(FLIVI)-K-K₄ and bis(FLIVIGSII)-K-K₄ commonly referred to as bis(h₅)-K-K₄ and bis(h₉)-K-K₄, respectively (Fig. 12). In 2012, Gudlur et al., first described the formation of these structures at 25 °C (RT) (Fig. 13).¹⁸⁴

Figure 12. Graphic representation of the branched and linear amphipathic sequences initially tested.

The color coding is added to signify those residues that are cationic (yellow), non-polar (blue) and hydrophilic (violet) along with the branch point (pink).

Figure 13. Transmission electron microscopy (TEM) obtained from the synthesis of nanocapsules solution (1.5 mM).

The image shows the process of fusing the nanocapsules, resulting in larger and more elongated structures (scale bar: 200 nm). Reproduced from Gudlur et al. (2012).¹⁸⁴

The linear (h₉)-K-K₄ sequence (Fig. 12) didnot form any detectable structures. Henceforth these structures will be referred to as branched amphiphilic peptide capsules (BAPCs) to avoid confusion with lipid vesicles.

In Figure 13, 50–150 nm spherical structures are present after a 2 h incubation period. The BAPCs in this image appear to be fusing. The resolution of these images was insufficient to identify whether or not the capsules were surrounded by a bilayer. Subsequently methyl mercury was added to the bis(FLIVI)-K-K₄ and bis(FLIVIGSII)-K-K₄ peptides in which the thiol containing amino acid cysteine was added to the C-terminal lysine of both peptide sequences.¹⁹⁵ In this representative image (Fig. 14) of capsules formed at 25 ºC, the bilayers are distinct due to the emission of a specific energy x-ray associated with mercury bombarded by electrons. The BAPCs remain water filled and a number of capsules appear to be fusing. Note that unlike many liposome preparations all of the BAPC structures are unilamellar.

Figure 14.

Scanning transmission electron microscopy (STEM) peptides labeled with Me-Hg 24 hours after mixing. Capsules were prepared with the peptides at 0.1 mM and 30% of them marked with Me-Hg. The images were captured using the inverted dark-field mode. Reproduced from Sukthankar et al.¹⁹⁵

Based on computer simulations these peptides behave much like phospholipids. As observed with membrane lipids, the hydrophobic regions of these peptide molecules are poorly soluble in water and in an analogous manner shield themselves by associating to form bilayers in aqueous solutions that appear to be 4 nm in thickness (Fig. 15). The aromatic side chains of phenylalanine residues and other aliphatic residues are buried in the interior of the newly formed membrane.

Figure 15. Atomistic model of the BAPC peptide bilayer.

In the center, the peptides are shown in a transparent surface, with Lys colored in yellow and other residues in blue and water in ball-and-stick mode. The top panel illustrates extensive π-π stacking interactions among Phe residues. The bottom panel shows the average electron density profiles of water (black line), Lys (red) and Phe residues (green line) calculated from last 40 ns of the 100 ns simulation.

In addition to the aromatic and non-polar residues interacting with each other through hydrophobic interactions; CD and FT-IR indicated significant parallel β-like structure, indicative of hydrogen bonding networks. The positively charged lysines reside in the two aqueous phases, on both the inner and outer sides of the bilayer (Fig. 15). Additionally, this model seems to show a networks of hydrogen bonds between adjacent β-structures that help the organized structure stabilize at very low sub-micromolar concentrations - as judged by dilution studies determined by both CD and ITC – conditions where nearly all phospholipid assemblies would be fully dissociated. We have yet to determine the critical assembly concentrations (CAC) of BAPCs.

3.3.1 BAPC assembly

Nanocapsules formation follows nearly the same steps used for the preparation of diacylglycerol phospholipid vesicles (Fig. 16). Initially the two peptides, bis(FLIVI)-K-K₄ and bis(FLIVIGSII)-K-K₄, are dissolved individually in neat 2,2,2-trifluoroethanol (TFE). In this solvent the peptides adopt a helical structure. As helices they are monomeric, thereby ensuring complete mixing of the two sequences when combined. This solvent also ensures that they remain helical upon drying, based on FT-IR studies.

Figure 16.

Assembly protocol for peptides nanocapsules.

The individual dissolved peptide concentrations are determined using the molar absorptivity (ε) of phenylalanyl residues in the sequence (195 M⁻¹ cm⁻¹ at 257.5 nm)¹⁸⁴. The two peptides are then mixed in equimolar ratios to generate desired concentrations and then dried in vacuo. Since neither peptide, by itself, forms stable capsules, all initial experiments limited the ratio to 1:1 to ensure an adequate concentration of the shorter peptide to accommodate the curvature of assembled capsules.

Distilled deionized water or aqueous solution containing solutes for encapsulation are added drop-wise to the dried peptides at the desired temperature. Different secondary structures and capsule sizes are seen when assembly occurs at different temperatures. This will be discussed later.

3.2.2 Fusion and Sizing

Sukthankar et al.¹⁹⁵ followed the assembly process over 2 h (Fig. 17) and showed that within 5 min of hydrating the mercury-labeled peptides, nano-fibrils appeared which gradually began to group together forming the nanocapsules with a 20 nm average size by 30 min. At 60 min the small BAPCs appeared to coalese and by 2 h have fused into larger capsules. By 48 h, micron sized capsules are present.¹⁸⁴ While these STEM images provide a static view into the underlying assembly events, a solution study was designed to demonstrate the dynamic nature of fusion process (Fig. 18A & B).¹⁹⁵

Figure 17. BAPC fusion time course.

Peptides(0.1 mM) bis (FLIVI) and bis (FLIVIGSII) containing 30% Me-Hg were hydrated and samples were removed at the indicated times and dried. The STEM images are displayed in inverted dark field. The scale bar at the bottom of the micrographs are 200 nm, 500 nm, 200 nm, 100 nm, 200 nm and 500 nm, for the intervals of 0, 5, 10, 30, 60, and 120 min, respectively. Reproduced from Sukthankar et al.

Figure 18. BAPC Fusion Study.

Panel A. Salt washed eosin Y trapped BAPCs were mixed with water filled BAPCs at 25 ° C. Fluorescence scans were taken at 5 min intervals for 235 min. The inset shows spectrum of sample stored at 4° C for 6.5 h. The units shown are identical to those in A. Panel B. Measured maximum eosin fluorescence intensity as a function of time during the fusion reaction. The t = 0 represents quenched value of salt washed eosin encapsulate in the capsules (2.2 mM). The data was fitted to a second order exponential with the error bars representing the SEM with n = 3. Reproduced from Sukthankar et al.¹⁹⁵

The dilution of the self-quenching fluorescent dye Eosin Y was recorded during fusion the process. Two populations of BAPCs, harvested and washed 30 min post hydration, containing either 2.2 mM dye (5 parts) or just water (95 parts), were mixed. At the earliest time points the dye containing BAPCs would be expected to fuse with the water filled ones thereby diluting the dye. This dilution was recorded as an increase in fluorescence. With each 5 min scan the intensity of the dye increased without any dye appearing in the 30 kDa cut-off filtrates, thus demonstrating fusion and the fact that the BAPCs are water filled. Lowering the BAPCs temperature to 4 °C at t = 35 min (inset to Fig. 18 A) shows that the fusion process can be stopped. Returning the 4 °C sample to RT did not restore fusion. This phenomenon is addressed later under the stability heading.

Since the BAPCs fuse like liposomes, resizing of larger BAPCs was tested using polycarbonate filters using methods commonly employed to prepare small phospholipid vesicles (SUVs). Starting with a 24 h, 25 °C sample (200–500 nm diamerter BAPCs) the BAPCs were extruded through a 100 nm pore filter, yielding a mixed population, 20 to 60 nm diameter capsules. Few if any 100 nm capsules were seen suggesting this size is disfavored over smaller ones. Those BAPCs were then extruded a second time through a 30 nm filter resulting in a relatively homogeneous population with sizes ranging from 20–30 nm. The 30 nm membrane is the smallest commercially available and attempting to go even smaller does not appear feasible since the BAPCs assemble as 20–30 nm structures. Once resized to 20–30 nm, the BAPCs should be cooled to 4 °C at RT to prevent their re fusion.

The inset to Fig. 18A showed that lowing the temperature to 4 °C after incubating at 25 °C for 35 min resulted in an altered BAPC that was no longer able to fuse. Even increasing the temperature step-wise to 80 °C failed to restart fusion. In addition, the cooled BAPCs were no longer susceptible to dissassembly in 50% TFE/water containing solutions. Based on these results the effect of temperature on BAPC properties were initiated. The results from these experiments indicate that there is a temperature dependence for assembly.

3.3 DISTINCTIVE PROPERTIES OF BAPCs

3.3.1 Temperature effects on BAPC formation

BAPC assembly was monitored at multiple temperatures with their final secondary stuctures determined.¹⁹⁶ To briefly summarize, BAPCs formed at 4 °C and 37 °C yield uniform BAPCs that are 20–30 nm in diameter, do not fuse yet are disassembled in 50% TFE. The 4 °C BAPCs were predominantly unstructured while the 37 °C displayed beta-like structure. The BAPCs formed at 25 °C were the only ones that displayed the fusogenic property and disassembled in 50% TFE. They displayed secondary structure that appear to be a mixture of random coil and beta-like structure. Over time as the BAPCs grew in size they became more beta-like in structure. It was hypothesized that the mixture of random and beta structure is higher energy and meta-stable and that fusion leads to a lower free energy state. Assembly at 25 °C for 30 min followed by 1 h of cooling at 4 °C and then rewarming to tempertures >25 °C makes the BAPCs even more stable and resistant to disassembly in 50% TFE. Their structures changed during the experiement from a combination of random and beta at 25 °C to mostly random at 4 °C and then back to mixed upon rewarming. It was postulated that during the thermal cycling entanglement of the hydrophobic segments occurred during the transition from random coil back to the mixed conformer. This unusual BAPC is referred to as being in a “locked” conformation. This particular BAPC has proven to be useful in delivering DNA to cells (not discussed in this review).

3.3.2 BAPC Stability

The thermal stability of the nanocapsules was also evaluated by Gudlur et al.,¹⁸⁴ using Differential Scanning Calorimetry (DSC) for temperatures ranging from 20–95 °C. This experiment was confirmed over the same temperature range using BAPCs containing the quenched dye Eosin Y (2.1 mM). No fluorescence intensity increase was observed over the full temperature range indicting the BAPCs did not rupture and release contents.

The lack of a phase transition over this temperature range indicates maintenance of the bilayer’s structural and functional integrity. This observation is important because phospholipid membranes comprised of acyl chains of 18 carbons or less with no double bonds generally exhibit phase transition temperatures between 35–55 °C. ¹⁹⁷ At these temperatures, lipid vesicles tend to become permeable and such property can be seen as a disadvantage due to the impossibility of retention of its internal contents when exposed to elevated temperatures.

3.3.3 Cellular uptake

As showed by Sukthankar et al.¹⁹⁸ nanocapsules are formed with an average diameter of 20–30 nm with an internal volume of 4000 nm³ allowing them to be considered as possible candidates for the encapsulation of a large number of different molecules including proteins. For this purpose BAPCs were used to encapsulate TAMRA-labeled TRNase A (RNase A, 13.7 kDa) and TAMRA-labeled cytochrome c (Tcytc, ~12 kDa). These apoptogenic proteins were chosen since their release from the BAPCs would trigger a measurable cytotoxic effect. For the positive delivery control the amphipathic peptide called Pep-1 was employed. This peptide reagent is known to induce cellular uptake of proteins and peptides.^199,200

Both cytochrome c and RNase A were successfully encapsulated in the BAPCs and efficiently delivered into HeLa cells (Fig. 19). In the images obtained we observed a slightly lower signal when comparing the transport of Tcytc using BAPCs with the control Pep-1.

Figure 19. Fluorescence Microscopy images of BAPCs delivering TAMRA labeled protein to HeLa Cells.

A) HeLa cells uptake of BAPCs carrying Tcytc; B) Control with Tcytc with Pep-1; C) BAPCs uptake containing TRNase; D) Control TRNase A with Pep-1. Reproduced from Sukthankar et al.¹⁹⁸

An interesting point observed is that although cytochrome c and RNase A are known for their apoptotic effects.^{201, 202} The BAPC experiments do not record any obvious sign of apoptosis whereas the transport mediated by Pep-1 exhibited the expected effect of cellular cytotoxicity. This study (Fig. 20) also showed that the BAPCs are retained intact in HeLa cells for long periods of time. The images generated with confocal microscopy revealed that even after 14 days, the BAPCs labeled with Rhodamine B persisted inside the HeLa cells and interestingly are transferred without any apparent degradation to the daughter cells formed during mitotic divisions.

Figure 20. Time function of the BAPCs integrity.

Images obtained with confocal microscopy of BAPC uptake by HeLa cells after 2 weeks showing: A) Excitation of Rhodamine in Dark field image; B) DIC image; C) Merge images exhibiting the Rhodamine inside of the BAPCs. Reproduced from Sukthankar et al.¹⁹⁸

3.3.4 In vivo studies- encapsulated alpha-emitting radionuclides

The remarkable stability of BAPCs makes them singularly useful in several applications and ill suited for others. Given their temperature stability and flexibility (based on resizing results) BAPCs were used to encapsulate an alpha-particle emitting radioisotope. Molecular nuclear medicine plays important roles in the diagnosis and treatment of cancer. Depending on the radioisotope’s properties, the compounds can be used for diagnostic and/or therapeutic purposes. Targeted alpha-particles have a significant advantage in targeted radiotherapy because of their high potency, specificity as well as their ability to kill non-proliferating cells and without the need for oxygenation.²⁰³ Current targeted alpha radio-immunotherapy takes advantage of the specificity of antibody-conjugated radionuclides to deliver biocidal radiation to cancer cells.²⁰⁴ Alpha-emitting radionuclides such as Actinium-225 and Bismuth-213 are commonly used for this purpose. This method does have some disadvantages in that the high energy of the ejected alpha particle can collide with and destroy the covalent bond between the chelator molecule and the antibody. Also the energy released during the recoil of the daughter nuclide is sufficient to dislodge the radionuclide from the chelator itself. This release of free radionuclides can cause off-target accumulation and potential damage. ²⁰⁵

For the BAPC encapsulation studies ²²⁵Actinium was tried. This radionuclide has a T½ = 9.92 days while releasing 4 α-particles as it decays to ²⁰⁹Bismuth.²⁰³ Alpha particles are emitted with a velocity of 5% the speed of light. The average energy of the 4 ejected helium nuclei is 6.93 MeV.²⁰³ The average mass of the daughter nuclei must have equal and opposite momenta which accounting for the mass difference calculates to an average energy of 0.13 MeV. As shown in Fig. 21 the BAPCs are able to retain 95% of the radionuclides (parent and daughters) for 7 days. As previously mentioned theoretical calculations by Sofou et al.^134,135 suggest negligible (<0.001%) daughter retention for the last isotope of ²²⁵AC for 100 nm diameter liposomes and 50% retention for liposomes with a diameter larger than 650 nm. Even giant liposomes (1 μm diameter), have retention values < 65% and in fact, the measured last daughter retention for the 650 nm liposomes was found to be substantially lower (11%) than predicted.^134,135

Figure 21. Retention of BAPCs encapsulated with ²²⁵Ac. Encapsulation and retention of ²²⁵Ac within BAPCs over 7 days.

Reproduced from Sukthankar et al.¹⁹⁸

Given the demonstrated robust properties of our peptide capsule design, we have started investigating the delivery of the alpha particle emitting Actinium to cells. To date¹⁹⁸, (in vitro) we have been able to load the BAPCs with ²²⁵Ac that have stably retained ≥95% of their activity over 7 days, and shown that cultured cells readily take-up non-lethal concentrations of encapsulated ²²⁵Ac. In vivo (Fig. 22), free ²²⁵Ac completely cleared from the blood while ²²⁵Ac-loaded capsules continued to circulate due to their small size.

Figure 22. Biodistribution of free and BAPC-encapsulated, ²²⁵Ac and its daughter ²¹³Bi, in CD1 mice.

A) 1 h time point; B) 24 h time point. Reproduced from Sukthankar et al.¹⁹⁸

Free ²²⁵Ac accumulated significantly more in the liver (P=0.03) and in the bone (P=0.02) than the ²²⁵Ac- BAPCs which points to the tight incorporation of ²²⁵Ac inside the BAPCs. At 1 h when both ²²⁵Ac-BAPCs and free ²²⁵Ac were still exiting the peritoneal cavity there was no significant difference in organ uptake between ²²⁵Ac-BAPCs and free ²²⁵Ac except for the bone where free ²²⁵Ac tends to accumulate. The only organ where there was more ²¹³Bi present compared to ²²⁵Ac were kidneys which serve as the “sink” for ²¹³Bi which has been released from any organ in the body.²¹³Bi daughter was present together with ²²⁵Ac indicating the possible retention of the daughter by the BAPCs as well. Together, these preliminary results point to the ability of BAPCs to incorporate and retain ²²⁵Ac and its daughter ²¹³Bi. Being able to selectively deliver this isotope to a location just outside the nucleus using targeting molecule adducted BAPCs should improve therapeutic efficacy while reducing off-target damage and lowering the whole body dose of radiation needed to kill targeted cells.

4. SUMMARY

There are three fundamental factors that go into the development of a competent nano-encapsulating system, viz - characteristics of the diseased state, choice of therapeutics and the nature of the delivery vehicle. The choice of the therapeutic and the nature of its action need to be carefully considered during the selection of a delivery platform. The drug’s site and mechanism of action define what carrier system will achieve optimal delivery. For example; drugs that require intracellular sites of action require intracellular delivery for bioactivity and therefore necessitate a delivery vehicle that enables homogenous tissue penetration.

Fundamental research into the nature of diseased state biomarkers and associated ligands, play vital roles in the design of functionalized delivery vehicles. Even in cases where targeting is not employed, an understanding into the characteristics of the diseased site, tissue accumulation and cellular uptake can help us engineer more efficient non-specific drug delivery systems by modulating the biophysical properties of nano-carriers.^206,207 BAPCs by their very nature represent a new and exciting class of novel encapsulating nano-system for drug delivery, not only because of their versatility, tune-ability, extraordinary stability, biocompatibility or targetability; but also because of the unique structural potentialities of its branched parent sequences that result in the formation of a pure peptide bilayer delimited nano-vesicle similar to liposomes in its organizational architecture. Indeed it could be said that the branched peptide sequences that constitute BAPCs are in themselves a unique class of biomaterial with great potential in a variety of chemical, physical and biological applications. An exhaustive investigation of these nano-materials is currently in progress.

Highlights.

• Review of bio-based drug delivery nanomaterials that self-assemble to form vesicles

• Focus on novel biomaterial— Branched Amphiphilic Peptide Capsules (BAPCs)

• BAPCs overcome some of the key limitations of the current packaging

• The polycationic capsules are readily taken up by epithelial cells in culture

• They escape or evade the endocytic pathway without any apparent degradation.

• Show great potential as safer drug delivery systems for various solutes/drugs