NIR Light Activated Hollow Gold Nanoshell Structures Based on Layersome Template

In biomedical applications, nanocarriers provide specific advantages such as delivery and sensing. These nanocarriers can be made from different sources including biopolymers and lipids. On the other hand, stimuli-responsive nanocarriers provide a safe method for drug delivery applications. These carriers use different triggers (including temperature) to release the drug. Light can be used as an external trigger to increase the temperature and finally releasing the drug. This thesis describes the development of functionalized liposomes via using different polyelectrolyte coatings to template the synthesis of light activated nanocarriers. In the first part, layer-by-layer coating of liposomes with strong biopolyelectrolytes was examined to have a better understanding of structure-property relationship of polyelectrolytes and layersome behaviors. The stability study of the structures in NaCl solutions with different concentrations demonstrate that the stability behavior depended on the outer layer coating. Samples with positive outer layer show more stability in salt solutions compared to the samples with negative outer layer. For the second part, nearinfrared (NIR) active gold nanostructures on hollow spherical soft templates were prepared by using wet chemistry method. Light scattering, spectroscopy and imaging techniques were used to examine morphology and NIR activity of hollow gold nanoshell structures based on layersome template. The results show by using layersome, small hollow gold nanoshell structures with NIR activity could be formed. All in all, these studies show a method to change the lipid ordering through coating process, alter the stability just by changing the outer layer of liposome and also an easy procedure to make a substrate for stimuli-responsive nanocarriers.


Abstract
Layer-by-layer deposition of polyelectrolytes (PEs) onto self-assembled liposomes represents an alternative to PE deposition on solid particles for the formation of hollow nanoscale capsules. PE-coated liposomes (referred herein as layersomes) reported in the literature display the typical charge inversion behavior that accompanies the deposition of sequential oppositely charged PE layers. However, liposomes are soft, dynamic templates that can be distorted or disrupted by adsorbing PEs. In this work we show that sequential deposition of dextran sulfate-sodium salt (DxS -) and poly-L-arginine (PA + ) onto cationic liposomes does yield the expected charge inversion, however, cryogenic transmission electron microscopy (cryo-TEM) results show that the layersomes formed (up to ~200 nm), and their PE coatings, were heterogeneous. This was due to the formation of PE complexes (PECs) when an even number of layers were deposited (PA + onto DxS -). Some of the PECs desorbed from the layersome surfaces, while many remained attached as patches and were coated by the next PE layer forming layersome-PEC clusters. This behavior was confirmed through fluorescence anisotropy measurements of liposome (bilayer) fluidity, where PA + counteracted the ordering effects of DxSon the lipid bilayer through charge neutralization and PEC desorption. With increased charge screening, DxSdesorbed from the layersomes, while the layersomes terminating in PA + retained their PE coatings. To our knowledge this is the first time such layersome structures have been observed with biopolyelectrolytes.

Introduction
Layer-by-layer (LbL) deposition is a versatile technique for creating multilayer micro-and nano-structured materials. LbL technique is based on sequential surface deposition of opposite charged macromolecules, typically polyelectrolytes (PEs), on to a charged substrate to create a self-assembled coating [1].By using particle templates, it is possible to create multifunctional polymeric capsules with tailored surface functionality and barrier properties. [2][3][4][5][6][7] The advantages of using LbL technique to create capsules include the ability to control the chemical, physical, and mechanical properties of the capsules by using different materials in capsule wall [2]; the ability to tailor the capsule wall charge and morphology by varying the terminal layer or the assembly conditions (for example, temperature, pH, and salt concentration); [8][9][10] and the ability to encapsulate macromolecules. [11,12] Capsules prepared by LbL technique are typically formed by depositing PEs on solid or porous inorganic particles as sacrificial templates that can be dissolved under acidic conditions. An alternative capsule template is liposomes, which have been used extensively in the areas of biomedicine, particularly drug delivery. [13,14] Liposomes are self-assembled phospholipid vesicles that have a bilayer membrane structure with an internal aqueous phase core. They can encapsulate both hydrophilic and hydrophobic compounds and, when used as capsule templates, liposomes are layered with PEs (the structures are referred to as layersomes) and remain a functional component of the capsule wall. [15][16][17] This approach has been used to create layersomes for drug delivery. [18,19] When formed with biologically-based PEs, including naturally derived PEs [20,21] and polypeptides, [15,16,22] layersomes can provide a nanoscale colloidal capsule that is biocompatible and biodegradable, and are often more stable than to bare liposome. It is generally observed that adding each PE layer leads to charge reversal and increases layersome size, capsule wall thickness and stiffness, and capsule barrier properties that resist spontaneous leakage. [23,24] A key aspect to creating layersome capsules and understanding their colloidal stability is determining how competition between PE-liposome and inter-PE interactions affects layersome structure. Volodkin et al. [22] investigated the interaction of poly-Llysine (PLL) coated dipalmitoylphosphatidylcholine/dipalmitoylphosphatidylglycerol (DPPC/DPPG) liposomes with polyanions of varying pK a and charge density including (poly-(4-styrenesulfonate), PSS; poly-L-glutamic acid, PGA; hyaluronic acid, HA). PSS with a low pK a (below 1) and high charge density led to complete PLL desorption, and PGA with a high pK a (around 5) and a lesser charge density than PSS led to partial PLL desorption. Partial PLL desorption led to surface charge heterogeneity and layersome aggregation, which was described based on patch-charge attraction and was used to create stable clusters of single-layered layersomes in electrolyte solutions. [25][26][27][28] In contrast, HA with a lower charge density yielded stable layersome capsules because PLL-liposome interactions were stronger than PLL-HA interactions. The competition between PE-liposome and inter-PE interactions has not been thoroughly investigated for other PEs or multilayered structures. An understanding of this competition could be used to predict and control layersome behavior, and may provide new routes for designing unique layersome structures.
In this work layersome structures were formed using cationic liposomes composed of dioleoylphosphatidylcholine (DOPC) and dioleoyltrimethylammonium-propane (DOTAP), coated with alternating layers of dextran sulfate (DxS -) and poly-L-arginine (PA + ) (Figure 1-1). DxSand PA + were chosen because they are FDA approved and have been used to create capsules for therapeutic applications via LbL deposition on solid particles. DxS -/PA + microcapsules prepared using calcium carbonate particle templates are reported to be biodegradable and biocompatible in vivo [29], and are capable of encapsulating proteins. [12] They have also been shown to activate pulmonary antigen presenting cells [30] and achieve immune-activity by targeting antigens to dendritic cells. [31]    and diphenylhexatriene (DPH) were purchased from Sigma-Aldrich Company (Missouri, US). The manufacturer's specification for the average number of sulfate (SO 3 -) groups per glucose unit was 2.3. All materials were used as received.

Liposome preparation
Liposomes were prepared in deionized (DI) water at 10 mM total lipid at a DOPC/DOTAP ratio of 1:1 using a rotary evaporator. The lipids in chloroform were

Layersome preparation
A washless method was used to form the layersomes using polyelectrolyte solutions prepared in DI water at 0.05% w/w. This method has been used previously for solid particles. [34] Layersomes were formed by sequentially titrating the liposomes (layer During each titration, the zeta potential was continuously measured as a function of the polyelectrolyte concentration. The point at which the zeta potential began to plateau was taken as the final polyelectrolyte concentration for layersome formation ( Figure 1-8).

Dynamic light scattering (DLS)
Hydrodynamic diameter (d h ) and zeta potential (ζ) were measured using a Malvern interpretations were based on layersome diffusion assuming spherical particles.
Layersome shape and the conformation of the PE coatings can strongly affect these measurements by altering the slip plane. Hence, d h and ζ reflect the average values for equivalent spheres.

Cryogenic Transmission Electron Microscopy (cryo-TEM)
Cryo-TEM samples were prepared at 25 °C using a Vitrobot (FEI Company), which is

Bilayer fluidity measurements
DPH anisotropy of the samples as a function of temperature was measured by using a Perkin Elmer LS 55 fluorescence spectrometer. The excitation wavelength and the detection wavelength were set at λ 23 =350 nm and λ 24 =452 nm, respectively, and the excitation and emission slit widths were set at 10 nm. Anisotropy was calculated from the following equation < r > = I 99 − I 9; I 99 + 2I 9; where I is the fluorescence emission intensity, and subscripts V and H represent the vertical and horizontal orientation, respectively, of the excitation and emission polarizers. [36]

Layersome formation and characterization
Unilamellar cationic DOPC/DOTAP liposomes were used as the initial template for layersome formation. The layer number for layrsome structures is denoted as Ln where n is 1 to 4; L1 and L3 were terminated with DxS -, and L2 and L4 were terminated with PA + . The weight percent of layersomes compounds and also charge ratio for the samples are shown in Table 1  Layersomes were formed in DI water with alternating coatings of DxSand PA + via electrostatic attraction. A schematic of the process reflecting the structures formed is shown in Figure 1    and that the layersome charge also persisted over 30 days (Figure 1-1d layersomes. The first DxSlayer (L1) led to a two-fold increase in <r> due to lipid ordering driven by electrostatic attraction between DxSand the lipids, notably DOTAP + (Figure 1-4).
The deposition of PA + (L2) counteracted the effects of DxSon lipid ordering due to charge neutralization which reduced the strength of the DxS --liposome interaction.
Also due to PECs desorption from liposome surface, some parts of layered structure do not have coating which leads to lower <r> value compared to layer 1 structure.
Layers 3 and 4 followed this same pattern where DxS -(L3) increased lipid ordering and PA + (L4) decreased it. An even number of layers with strong inter-PE interactions coincide with the formation of PEC patches. As shown by cryo-TEM, patch formation and also PECs desorption from liposome surface exposed regions of uncoated bilayer that would exhibit similar <r> values to bare liposomes.
In this case <r> reflected the average lipid ordering of bare regions ('low' <r>) and regions with bound PECs ('high' <r>). These results provide further evidence that PECs formed as a result of inter-PE interactions.

Layersome response to increasing ionic strength
To probe layersome stability, the zeta potential and the hydrodynamic diameter of each layersome was examined as a function of NaCl concentration (Figure 1-5).
Layersomes terminating in PA + retained a high |ζ| compared to DxS -, similar to the liposome template. Layersomes terminating with DxSshowed a significant reduction    (Figure 1-9a). The time lag was significant and independent of NaCl concentration because NaCl was in excess relative to DxS -. In contrast, the correlogram for PA + was less sensitive to NaCl concentration and, compared to DxS -, a modest shift in the time lag was observed (Figure 1-9b). We can conclude that DxSlayers condensed Na + and that condensation was responsible for

Conclusions
We have shown that a washless method can be used to create multilayered layersomes with DxSand PA + polyelectrolytes. In deionized water, electrostatic 'coupling' between PEs, when PA + was the terminal layer, led to the formation of nanoscale PECs. Most of these PECs remained anchored to the liposome surface and some of them were desorbed from liposome surface. This appears to be a feature of the strong PE pair and to our knowledge such structures have not been reported for layersomes prepared by LbL deposition. We assumed that due to stronger affinity between the polyelectrolytes compared to liposome-polyelectrolytes, the patchy structures are formed and so some parts of layersomes are without polyelectrolytes so those parts Fluorescence anisotropy results suggest that PEC formation occurred when the adsorption of PA + weakened the interaction between DxSand the liposome template.
In the presence of NaCl the interactions between the layers and the liposome, and the stability of the layersomes, was dependent upon the terminating PE layer. Layersomes terminating in PA + exhibited zeta potentials >20 mV, and charge screening in high [NaCl] likely drove aggregation. For these layersomes, the interaction between the layers and the liposome was only modestly affected by [NaCl], indicating that the layers remained intact on the liposome surface. Additional work is needed to determine if the PEC patches remained present in NaCl. In contrast, for layersomes terminating in DxS -, counterion condensation led to very low zeta potentials, probable DxSdesorption, and significant aggregation.

Acknowledgements
This work was supported by the National Science Foundation under Grant No. CBET-1337061. We would like to thank Dr. Richard Kingsley for his assistance with cryo-TEM imaging.

Introduction
Photothermal therapy refers to the methods that use electromagnetic radiation to treat diseases such as cancer by increasing the temperature of the cancerous cells [1].
Metallic nanostructures have been widely used in the development of external triggered drug delivery systems [2], [3]. Among these nanostructures, plasmon resonant nanostructures have been used for drug release using localized heat via light exposure [4]. Plasmon resonant nanostructures are made of noble metals and show enhanced absorption and scattering due to the collecting oscillation effect. This is a phenomenon that results from the interaction between electric field of a specific wavelength light and the free electron of a metal that leads to metal free electron oscillation and subsequent absorption and scattering [5].
Plasmon resonant gold nanostructures are attractive for a range of biomedical applications due to their biocompability, stability and light absorption properties [6]- [8]. The absorption behavior of these nanostructures, which include nanoparticles, nanorods, nanoshells, nanocages, and nanostarts, depends on shape, size, aspect ratio, core material, shell thickness and particle diameter [9]. One advantage to the use of plasmon resonant nanostructures in this application is their potential to enable precise heating, thereby avoiding damage to nearby healthy cells.
Liposomes are self-assembled phospholipid vesicles that have a bilayer membrane structure with an internal aqueous core. These structures are suitable vectors for both hydrophilic and hydrophobic entities [10], [11]. The merging of the liposome with the gold nanostructure is an attractive concept for light-triggered drug delivery systems as well as cancer heat treatment systems.
The combination of the NIR active gold nanostructures with liposomes, have attracted much interest for drug delivery applications. The main release mechanisms are liposome phase transition by way of a temperature increase or liposome rupture due to nanobubble formation induced by pulsed laser [12] [13]. These two release mechanisms can be used in conjunction. Four different configurations exist for the construction of these types of delivery systems.

Gold-liposome suspensions
In this configuration, the therapeutic is loaded inside of the liposome. Polymer and non-attached NIR light absorbent gold nanoparticles are also present in suspension.

Gold nanostructures-liposome composites
In these structures, NIR-sensitive gold nanostructures (gold clusters, nanoshells, nanorods, etc) are attached to the liposome surface. By NIR light irradiation of these structures, the liposome content (payload) will be released (Figure 2-2). Similar to main differences between this configuration and the previous one are the absence of temperature sensitive polymer and also direct attachment of gold nanostructures to liposome surface.

Encapsulated gold nanostructures in liposome core
This type of structure is similar to gold nanostructures-liposome composites but with one difference. In this configuration, the NIR-sensitive gold nanostructures, including hydrophilic gold nanorods, gold nanoparticles and hollow gold nanoshells, are inserted into liposome core [17] [18]. So in the case of loading inside the liposome core, direct contact between gold and payload is possible. Despite the other configurations that the gold nanostructures are formed in a separate process, in this configuration, gold is formed directly on liposome surface as a coating.

Gold nanoshell
These structures combine optical properties of the coating (plasmon resonant shell) and biodegradability and encapsulation properties of liposome. Due to the fact that the gold is attached to liposome surface, the generated heat by light can be transferred to liposome in an efficient way [19] [20].
Since in all the configurations, formed gold structure is not continuous and they are formed on liposome surface which have leaky nature, the idea of using layersome as a coated liposome to improve leak resistant material for the substrate of gold nanoshell structure was considered [21]- [23].
In this work, by using poly-L-histidine as a FDA-approved polyelectrolyte, the efficacy of using a layersome as a template for NIR light active gold coated hollow soft structures, gold nanoshells, was investigated and the formation of gold nanoshells with NIR absorption capability on two different substrates, liposome and layersome, was evaluated. On the other hand, there are only few studies about in-situ gold nanoshell formation on soft structures, so to have a better understanding of this process and dominant parameters, this study was done. To achieve the goals, three sets of samples (L1, L2 and L3) were considered. For the first series (L1), the main hypothesis was to use PLH affinity for gold through histidine-gold binding to make gold nanoshells. This affinity plus charge attraction between PLH and gold ions help in gold ion attraction to the liposome surface. For the second series (L2), regarding the strong affinity of gold to sulfur, the idea of using PEG-thiol with liposome to absorb gold was considered. It was mentioned that the strong covalent between gold and thiol group could lead dense gold nanoparticles network formation [33], [34]. PEG-thiol also helps bind gold and improve steric stabilization after formation. For the third series (L3), the idea of applying PEG-thiol in the presence of PLH was considered.
The hypothesis behind this idea was to use the charge attraction between PLH and PEG-thiol to have more PEG-thiol on the layersome surface and to determine the synergistic effect of PEG-thiol and PLH on gold shell formation.

Layersome preparation
The washless method was used to form the layersome using polyelectrolyte solution (PLH) prepared in DI water at 0.05% w/w [24]. In this method, the required amount of polyelectrolyte to cover the liposome surface was determined through a titration process and determining the plateau point of zeta-potential. After determining the required amount of polyelectrolyte, it was added to liposome solution and mixed for 5 min at room temperature for coating process.

Gold nanostructure formation
To make gold nanoshell structures (GNSs), three sample series (L1, L2 and L3) were prepared. The main difference between these series is the type of template used to form GNSs. The first template was a liposome with a mixture of DOPC:DOPG (1:1 mole ratio) and the second template was a layersome (PLH layered on DOPC:DOPG liposomes). To make GNSs, different amounts of gold ion solution were added to the substrate solution and mixed using mixing bar for 5 min at room temperature. Then ascorbic acid, as a reducing agent, was added to the above solution and mixed for another 5 min. For samples that included PEG-thiol, before adding gold solution, PEG-thiol was added to liposome or layersome solutions. For all the gold samples, the mole ratio of ascorbic acid to gold ions was constant and equal to 4, but the mole ratios between gold ions and the templates varied between 1 to 7. For control samples (L1-6-control, L2-6-control and L3-6-control), the preparation conditions were the same as L1-6, L2-6 and L3-6 samples, respectively but without any liposome or layersome. In the other words, there was no substrate for the control samples.

Dynamic light scattering (DLS)
Hydrodynamic diameter (d h ) and zeta potential (ζ) were measured using a Malvern

Perkin-Elmer Lambda 1050 spectrophotometer with Deuterium and Tungsten
Halogen light sources was used for the absorbance spectra of the samples. The spectrophotometer was used in the wavelength range from 500 nm to 1300 nm. Quartz cell was used for this measurement. The samples were diluted 10 times.

Cryogenic Transmission Electron Microscopy (cryo-TEM)
Cryo-TEM samples were prepared at 25 °C using a Vitrobot (FEI Company), which is

Field Emission Scanning Electron Microscopy (FE-SEM)
To prepare samples for FE-SEM microscopy, 5 µL of 10 times diluted samples was put on a silicon wafer and dried in vacuum oven during night. Imaging was performed using Sigma VP FE-SEM instrument (ZEISS Inc., Thornwood, NY). The Energy Dispersive Spectroscopy (EDS) analysis was performed using Silicon Drift Detector (OXFORD Inc. Concord, MA). Since the samples including gold are conductive, we decided not to use sputter coating, so only the structures with gold are seen.

Results and Discussion
As mentioned before, three sets of layersome-templated gold nanoshell samples (L1,  DLS data ( Similarly, the number size distribution graphs don't show layersome peak (near 100 nm) (Figure 2-4). One of the possibilities that in gold nanostructure solution, we can not see the peaks for liposomes is due to the fact that we have gold structures and since gold is heavier, it absorbs more signal in DLS measurement and so bare liposome can not be detected by the equipment. To check this possibility, cryo-TEM was used. Regarding the z-potential, for layersome we have a quite large value (+48.8 mV) and other than control sample, all of them have positive surface charge. Since we used ascorbic acid for reduction process, it was expected to have negative zeta potential for samples. This phenomenon can be explained by PLH absorption on gold structures. By increasing the gold to lipid ratio, the absolute value of zeta potential decreases and this is due to the fact that the amount of added ascorbic acid increases but the PLH remained constant.  To check the substrate in which gold was formed on, EDS analysis for L1-6 sample was done (Figure 2-6). Gold mapping is shown in red and phosphorous, as liposome representative, is shown in yellow. Based on the EDS analysis, it is clear that gold is almost everywhere but it is more concentrated on liposome surface (where the phosphorous exists) and close to liposome. Based on the absorption spectra of this sample series (Figure 2-8), although there is absorption in NIR range (650-900 nm), the wavelengths corresponding to the peak maxima are out of that range. By increasing the Au to lipid ratio, the peak location maxima go to higher wavelength. The obtained absorption spectra is similar to the gold nanoparticles spectra which is another evidence to have gold nanoparticle for L1 series [12], [25]- [28]. There is some absorption in the NIR zone that is due to either clusters of gold nanoparticles and/or gold coated liposomes. The possible reason of having peak out of NIR range is that for this sample series we have more gold nanoparticles and so the absorption of gold nanoparticles is stronger than gold coated liposomes. Although we might have some gold coated liposome in this sample series, but according to DLS data and also absorption spectra, these structures were not dominant so we were trying to modify the process to increase the number of gold coated structures.    DLS data for this series shows bigger structures than L1 series ( Table 2-2). Also the structures are more uniform (lower PDI) and all the samples have single peak in size graphs (Figure 2-10). The zeta potential of this series is all negative which shows that the surface of structures were covered either with ascorbic acid and/or PEG-thiol.  show smaller peaks. For L2-6-control, since we do not have any liposome, we expected this behavior but for L2-6 sample, it seems that we made much more gold nanoparticles compared to gold coated liposomes and that's why we see a bigger peak in smaller size range. Absorption spectra of this series show peaks in NIR range ( Figure 2-11), So based on spectra and DLS data we may have desired shell structures.
Besides L2-1 sample, for all the other samples, we have absorption peak in NIR range and this also includes L2-6-control sample. By adding more gold ions to the system (from L2-2 to L2-6), the peak location moves toward lower wavelength which can be due to increasing in shell thickness. For L2-6-control, we have a very high absorption peak that maybe due to the gold nanoparticle aggregates. To have a better understanding and to check the morphology, SEM technique was used.    To make the L3 series, layersome was used as the substrate. The hypothesis behind this idea was to use charge attraction between PLH and PEG-thiol to have more PEGthiol absorption on liposome (substrate) surface and more gold formation on substrate and a denser gold layer. The schematics of L3 formation process is shown in Figure 2-15. DLS data ( Table 2-3) shows that compared to L2 samples, this series has bigger size and higher PDI but number size distribution size graphs show single peak similar to L2 series (Figure 2-16). This sample series shows positive surface charge that can be due to the existing of extra PLH and coating the final structures.  L3-6-CONTROL Similar to L2 samples, the L3 series shows NIR absorption peak but the peaks shift to higher wavelengths (Figure 2-17). As described elsewhere, this can be due to the fact that for this series, layersomes are covered more densely by gold nanoparticles [12].
One of the interesting phenomenon about this samples series is their absorption peak locations. Besides L3-1, for the rest of samples, the absorption peak location is almost the same. To check the morphology with another technique, cryo-TEM was used for L3-6 sample (Figure 2-19). These images show bumpy structures which is the result of gold formation on layersome surface and the TEM results are in agreement with obtained SEM images.

Conclusions
The idea of using layersome as a substrate for NIR active gold nanoshell formation was investigated. According to this study the electrostatic interaction provided by PLH is not solely sufficient to make desired structures on layersome substrate. By using PEG-thiol, gold nanoshell can be formed on layersome substrate through strong covalent bond. On the other hand, and according to formed structures on liposome with negative surface charge, it seems that the covalent bond role is stronger than electrostatic interaction in this process. By using the applied method, small size NIR active gold nanostructures based on layersome template were formed which have potential application for cancer heat treatment.

Acknowledgements
This work was supported by the National Science Foundation under Grant No. CBET-1337061. We would like to thank Dr. Richard Kingsley for his assistance with cryo-TEM imaging.