1.

Introduction

Vaccination is one of the most important instruments to fight infectious diseases. Vaccines are mostly applied by intramuscular injection. Injections, however, bare the risk of spreading infectious diseases such as HIV or hepatitis in case of improper usage.¹ According to the World Health Organization (WHO), the number of injections amounts to 12 billion per year. One-third of all injections is applied with improperly sterilized sets of instruments.² In Africa, more than 80% of disposable syringes are used more than once.^{3, 4} This holds a high risk of spreading infectious agents, especially in populations with a high prevalence of infectious diseases.⁵

Oral application of vaccines is a noninvasive alternative strategy for vaccination but reveals insurmountable hurdles. It is for this reason that the oral application against polio was short-lived, because most of the vaccines lost their function after becoming metabolized and worse, many of them functioned as tolerogens.⁶ Therefore, the WHO gave precedence to the development of topical vaccination.^{7, 8}

The skin forms a barrier, protecting our body against environmental hazards. It contains a tight network of immune cells, particularly antigen-presenting Langerhans cells and dendritic cells with the ability to induce a specific immune response. Topical vaccination, in contrast to intramuscular or intradermal application of the vaccine, could therefore be an efficient way to activate effector-T-cells and induce an immune response.^{9, 10}

Topically applied vaccines have first to pass the skin barrier to reach intraepithelial T-cells and Langerhans cells.¹¹ Hair follicles were shown to be a promising target for transcutaneous vaccination,¹² as they are rich in immune-competent cells in the infundibular part of the root sheath and around the excretory duct of the sebaceous gland.¹¹ Hair follicles offer a considerable penetration pathway and potential intracutaneous reservoir for topically applied substances,¹³ which makes them an important target for topical vaccination.

For the development of a topical vaccination strategy, a suitable drug delivery system is essential. Nanoparticles appeared to address this issue excellently. Toll investigated the follicular penetration of nanoparticles at different diameters.¹⁴ Particles with a diameter of $600\mathrm{nm}$ showed deeper penetration into the hair follicles than larger sized particles. Lademann, compared the penetration behavior of fluorescent dyes into porcine skin, comparing nanoparticles $\left(320\mathrm{nm}\right)$ to a nonparticle formulation.¹⁵ Massage, subsequent to application, resulted in a five-fold increase in follicular penetration depth of the fluorescent dye nanoparticles compared to its nonparticle counterpart. This observation was verified by in vivo studies on human skin. Additionally, the in vivo studies showed a remarkable follicular reservoir of the dye when applied as nanoparticles. This reservoir was still detectable $10\text{days}$ after application.¹⁵ A reasonable explanation for this effect is the gear-pump mechanism, whereby the hair shaft acts like a pump. The particle is pushed deeper into the follicle by the movement of the hair and trapped between the cuticula cells of the hair shaft and the inner root sheath. The mechanism works ideally for particles with a diameter similar to the cuticle cells, which is measured to be approximately $500\mathrm{nm}$ in human hair.^{15, 16, 17}

An ideal drug delivery system into hair follicles would therefore be constituted by nanoparticles ranging between 100 and $500\mathrm{nm}$ ^{15, 17} that can penetrate deeply into the hair follicle and release their cargo, i.e., the vaccine, close to the immune cells. Liposomes as vesicular nanoparticles are an excellent vehicle system for vaccines and therefore a promising drug delivery system for topical vaccination.

The present study addressed the issue of cutaneous and follicular penetration of liposomes. Because of its high resemblance to human skin, regarding structure and distribution of the hair follicles in particular, the experiments were carried out on pig ear skin. Investigated liposomes, which were different in size and other properties such as lipid composition and surface charge, had been prepared from special lipid mixtures, exerting a beneficial effect on their loading rate, membrane stability, cell transfection rate, and membrane surface charge potential. In particular, one of the studied liposome types represents a novel class of liposomes which has the possibility to shift its surface charge depending on the environmental pH. Confocal laser scanning microscopy of intact skin and of histological longitudinal sections was used to determine follicular penetration depth.

2.

Materials and Methods

Five different types of liposomes containing a fluorescent dye were provided by Novosom AG (Halle, Germany). Whereas liposome types 1, 2, 4, and 5 enclosed the hydrophilic fluorophore carboxyfluoresceine in their inner hydrophilic compartment, liposome type 3 was prepared with curcumin. Because of its intense hydrophobic character, curcumin was inserted in the membrane of the liposome instead of being enclosed in the inner hydrophilic compartment.

2.3.

Preparation of the Test Areas and Application of Samples

Two test areas ( $4\times 6$ and $4\times 2{\mathrm{cm}}^2$ ) per formulation tested were marked by a barrier (ART-GLASSs by WACOs; peelable dye; Heinrich Wagner, Zurich, Switzerland) on the dorsal side of the pig ears. The barrier was necessary to prevent lateral spreading of the applied formulations. A $0.02\mathrm{mL}$ sample of the liposome stem solutions, $2\mathrm{mg}$ of the standard formulation, respectively, was applied per $1{\mathrm{cm}}^2$ of skin. After application, the samples were spread homogeneously with a spatula. Skin biopsies $(5\times 5\mathrm{mm})$ were taken at $1\text{hour}$ , three, five, and seven days after application, each from a different subdivision of the larger test area 1. Another subdivision of this area also served for investigation by noninvasive confocal laser scanning microscopy.^{21, 22, 23, 24} The smaller test area 2 was used to take cyanoacrylate skin surface biopsies to measure the amount of liposomes in the hair follicle infundibulum. Cyanoacrylate skin surface biopsies were analyzed by laser scanning microscopy. Figure 1a illustrates the arrangement of the test areas on the pig ear.

Fig. 1

Experimental procedure. (a) Schematic representation of the arrangement of the application areas on the pig ear. (b) Schematic representation of the preparation of histological sections from biopsies taken $1\mathrm{h}$ after application. I. Five biopsies were taken from the application area. II. From the biopsies histological sections were prepared. III. The histological sections were collected on a glass slide.

3.

Results

Noninvasive measurements of the cutaneous penetration process by confocal laser scanning microscopy was carried out for liposome types 1, 4, and 5, and for the standard formulation, respectively, using intact skin. As shown in Fig. 2, $1\mathrm{h}$ after application of the three liposome samples, the depth of the laminar fluorescence, which marks the plane of highest sample concentration relative to the measurement tip of the microscope, amounted to $80\mu \mathrm{m}$ . This depth was used as a reference marked by the dashed line in Fig. 2 to assess whether cutaneous penetration had occurred. Whereas at day 3 all formulations showed similar values of the relative depth of the laminar fluorescence compared to the reference values $1\mathrm{h}$ after application (dashed line), the liposome samples clearly showed increased depths at day 5 which ranged between 20 and $40\mu \mathrm{m}$ . Finally, at day 7 after application, both the liposome samples and the standard formulation showed a similar and significant increase in their relative penetration depth, ranging between 35 and $45\mu \mathrm{m}$ compared to the reference value (dashed line).

Fig. 2

Cutaneous penetration. Depicted are the relative depths of LT1, LT4, and LT5, and of the standard formulation, respectively, at which a laminar fluorescence signal occurred. Measurements were carried out at $1\mathrm{h}$ , three, five, and seven days after application. Relative depth values at $1\mathrm{h}$ and $5\text{days}$ after application of the standard formulation were not determined.

The present study focused mainly on the investigation of the follicular penetration of the applied liposomes. Twenty-four thousand histological longitudinal sections were prepared. Eight hundred and seventy-one of these sections contained a complete parallel section of a hair follicle, which were analyzed by confocal laser scanning microscopy. Fluorescence was detected in 84% $(=735)$ of the hair follicles. The remaining 136 sections did not show a fluorescence signal in the hair follicle but in the stratum corneum. The access to the follicle depicted in these two figures seems to be impeded by structures at the top of the infundibulum [Figs. 3a and 3b ].

Fig. 3

Fluorescence imaging. (a) (b), Superimposed images depict histological sections of hair follicles taken at day 3 after application of LT3, measured by fluorescence and light microscopy. (c) Fluorescence image of a histological section taken at day 7 after application of LT4. (d) Superimposed image depicts closeup of a hair follicle of a histological section taken three days after application of LT2. (e) (f) Fluorescence images of histological sections taken at day 7 after application of LT4. $\mathrm{Bars}=100\mu \mathrm{m}$ .

Figure 3c shows one of the 871 hair follicles that contained a fluorescence signal. In this follicle the fluorescence signal reached down to the bulbus pili. As already seen in Figs. 3a and 3b, a significant fluorescence signal in the stratum corneum can also be observed. In cases where the section hit the excretory duct of the sebaceous gland, a continuation of the fluorescence signal inside the duct was found [Fig. 3d].

Figures 3e and 3f show fluorescence images of a hair follicle that was sliced in a slantwise manner. Therefore, one can clearly observe a distribution of the fluorescence signal along the edges of the cuticula.

At least 20 follicles were evaluated for each formulation and at each time point, respectively. The measured absolute penetration depths were normalized by their relation to the full length of the corresponding follicle, resulting in a relative penetration depth value expressed in percent. The relative penetration depth allowed the comparison of the penetration depth within follicles of different lengths.

Figure 4 summarizes the mean values of the relative penetration depth for all tested formulations. One hour after application, all formulations had reached a relative penetration depth ranging between 30 and 40%. Whereas the liposomes showed a significant increase in penetration, the mean relative penetration depth of the standard formulation stagnated during the following six days of monitoring the penetration process.

Fig. 4

Survey of the mean relative penetration depth values of the tested formulations.

The highest relative penetration depths with mean values up to 69% were reached by liposome types 4 and 5. The maximum penetration of these two liposome types was reached at day 5. This was also the case for liposome types 1, which reached a penetration depth of 45%. At day 7, the mean relative penetration depth values of the three liposome types decreased in comparison with the previous measurements. In contrast, liposome type 2 showed a continuous increase in penetration, reaching its maximum of 55% at day 7 after application. Liposome type 3 showed a maximum of the relative penetration depth at day 3 after application, at approx. 48%.

The method of cyanoacrylate surface skin biopsies functioned unsatisfactorily. From 21 biopsies only one vellus hair root with a total length of $550\mu \mathrm{m}$ was gained (Fig. 5 ). However, this hair root clearly showed that liposomes were located between the hair root sheath and the hair shaft along the entire follicle, whereas the cutaneous penetration was restricted to the stratum corneum. The deeper epidermis did not show a fluorescence signal.

Fig. 5

Fluorescence imaging of a cyanoacrylate surface skin biopsy. Depicted is a series of superimposed images all recorded at a different depth starting at (a) the highest depth or the bulbus pili of the hair, respectively. Total length of the hair root amounted to $550\mu \mathrm{m}$ . Sample was taken at day 5 after application of LT1. $\mathrm{Bar}=200\mu \mathrm{m}$ .

4.

Discussion

Liposomes showed a significantly higher penetration at the follicular path compared to a standard formulation. At the cutaneous path, the penetration of both the liposome samples and the standard formulation were of similar dimensions. Although there is an increase in the cutaneous penetration depth, it was restricted to the stratum corneum, as verified by the histological sections (Fig. 3) and the cyanoacrylate surface skin biopsy (Fig. 5). Therefore, the cutaneous penetration path where topically applied samples must cross the stratum corneum would be less effective regarding the development of topical vaccination strategy, because the applied vaccine would never reach the immune cells, which are located in subepidermal skin layers.

As liposomes penetrated deeply into the hair follicles, this penetration path seems to play a major role in the development of a nanosized liposome-based topical vaccination strategy. Dependent on the liposome type, mean relative penetration depths of 50 to around 70% of the full length of the hair follicle were reached. Therefore, nanosized liposomes would be able to reach the immune-competent cells which are accumulated around the hair follicles. However, comparison of the penetration behavior of the tested liposome types revealed considerable differences regarding the time point at which the maximum of penetration was reached, as well as the dimension of the penetration depth.

The highest penetration depths were reached by the amphoteric liposome type (LT4) and one of the cationic liposome types (LT5). The isoelectric point of the amphoteric liposome type lies at a slightly acidic pH, and because porcine skin shows a pH of 6–7,²⁷ its surface was presumably neutral or slightly anionic under experimental conditions. The lowest penetration depth maximum was observed for negatively charged liposomes, like liposome type 1. Therefore, it can be concluded that the surface charge plays an important role in the penetration behavior of a given liposome type.

One can further speculate about the effect of surface charge for the penetration of nanoparticles because the skin itself and the hair are negatively charged on their surface. Therefore, the electrostatic interactions with cationic nanoparticles might consequently lead to high local concentrations, whereas negatively charged particles might be rejected. It remains to be analyzed whether the electrostatic has a beneficial or inhibiting effect. Nevertheless, this perspective might be helpful in the interpretation of the superior penetration of the amphoteric liposomes. The electrostatic effect may have drawn the liposome in penetration direction. Another parameter that may affect the penetration behavior of the investigated liposomes is the phase transition temperature of the lipid types used for synthesis. Liposome types 4 and 5, which reached the highest penetration depths, are based on the lipid dipalmitoylphosphatidylcholine, which has a $T_{\mathrm{m}}$ of $23°\mathrm{C}$ . This will constitute a more flexible bilayer structure, whereas the liposome types 1 and 2 may be more rigid. Latter ones comprise large amounts of dipalmitoylphosphatidylcholine with a $T_{\mathrm{m}}$ of $41°\mathrm{C}$ . Higher flexibility may be an advantage within the narrow slot between the hair root sheath and the hair shaft, which depicts the penetration path as verified by the vellus hair gained by cyanoacrylate biopsy (Fig. 5). Penetration of liposomes being of more rigid character may be retarded due to spatial limitations.

No correlation was found between the diameters of the investigated liposome types and their maximum penetration depths. This may be caused by the slight differences in the diameter of the different liposomes. For bigger differences in the diameter of the liposomes, differences in the penetration depths could be expected. Five out of six of the investigated formulations, including the standard formulation, showed a decrease in the penetration depth with time. An explanation for this phenomenon may be a drying process of the skin tissue. As the hair follicles are surrounded by a tight reticulum of connective tissue, the formulation may have been pressed out due to the relaxation of these fibers, which is induced by the decrease in the overall tension of the skin.

All liposomes investigated possess high stability in serum, but nothing is known about their stability in sebum. However, there is evidence to assume that the main portion of the liposomes remained intact during penetration. First of all, the fluorescence pattern in the micrographs is punctuated and uneven (Fig. 3). Secondly, most samples contained carboxyfluorescein, which, in case of release, may not penetrate in sebum. Otherwise, its fluorescence may be quenched because of the low pH of the environment.

The infundibula of human hair follicles ranges between 25 and 33% of the total follicle length.²⁸ Considering the high penetration depth (up to 69%) in pig ear skin and the smaller absolute length of the human hair follicle, all applied liposome formulations had the potency to act as a follicular drug delivery system in human skin. As penetration reached beyond the infundibulum, this could for several reasons be advantageous for the application of substances that are designed to provoke an immune response. First of all, the immune cells are located in the region ranging from below the infundibulum to the bulbus pili. In can be assumed that the deeper the penetration of vaccine containing liposomes, in particular, beyond the extent of the infundibulum, the higher the immune response. Otherwise, liposomes, which are restricted to the infundibulum, may induce a much weaker immune response. Additionally, these liposomes are exposed to the clearance by the continuous stream of sebum from the outlet of the sebaceous gland toward the follicular orifice. This is no longer of interest once the liposomes penetrated below the confluence of the excretory duct. A further point that must be addressed is the observed attachment of liposomes to the cuticula of the hair shaft [Figs. 3e and 3f]. On the one hand, this interaction might slow down the penetration process. On the other hand, the interaction may decelerate the clearance of the liposomes by the sebum stream.

Although penetration is a necessary condition for an effective vaccination, this is not sufficient. More important, the vaccine must reach the antigen-presenting cells. Amphoteric liposomes and cationic liposomes can mediate the intracellular delivery of antigens. Further studies will detail the effect on the immune response.

5.

Conclusion

In summary, in the present study a methodology was established to assess the strategy of topical vaccination using nanosized liposomes. Interfollicular penetration through the stratum corneum seems to play a minor role for the tested liposomes, whereas follicular penetration of the liposomes seems to be the main route. Liposomes penetrate deeper into hair follicles than a standard formulation. Therefore, they could increase the transfollicular drug uptake. The highest efficacy of follicular penetration is exhibited by liposomes owning the following set of attributes: a positively charged or unloaded surface and a phase transition temperature, which is in the range of the skin temperature or below. Liposomes with a constant high anionic surface charge might be less suitable for follicular drug delivery.