Drug delivery across intact and disrupted skin barrier: Identification of cell populations interacting with penetrated thermoresponsive nanogels

Accepted Manuscript Research paper Drug delivery across intact and disrupted skin barrier: identification of cell populations interacting with penetrated thermoresponsive nanogels F. Rancan, M. Giulbudagian, J. Jurisch, U. Blume-Peytavi, M. Calderón, A. Vogt PII: DOI: Reference:

S0939-6411(16)30823-2 http://dx.doi.org/10.1016/j.ejpb.2016.11.017 EJPB 12352

To appear in:

European Journal of Pharmaceutics and Biopharmaceutics

Received Date: Revised Date: Accepted Date:

30 April 2016 14 November 2016 15 November 2016

Please cite this article as: F. Rancan, M. Giulbudagian, J. Jurisch, U. Blume-Peytavi, M. Calderón, A. Vogt, Drug delivery across intact and disrupted skin barrier: identification of cell populations interacting with penetrated thermoresponsive nanogels, European Journal of Pharmaceutics and Biopharmaceutics (2016), doi: http:// dx.doi.org/10.1016/j.ejpb.2016.11.017

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Drug delivery across intact and disrupted skin barrier: identification of cell populations interacting with penetrated thermoresponsive nanogels

F. Rancana*, M. Giulbudagianb, J. Jurisch a, U. Blume-Peytavia, M. Calderónb and A. Vogta

a

Clinical Research Center of Hair and Skin Science, Department of Dermatology and

Allergy, Charité-Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin, Germany b

Freie Universität Berlin, Institute for Chemistry and Biochemistry, Takustrasse 3,

14195 Berlin, Germany

* Correspondence to: Fiorenza Rancan, Clinical Research Center for Hair and Skin Science, Klinik für Dermatologie, Venerologie und Allergologie, Charitéplatz 1, 10117 Berlin, Germany, Tel. + 49 30 450518347 e-mail: [email protected]

1

Abstract

Nanoscaled soft particles, such as nanogels, can be designed to incorporate different types of compounds and release them in a controlled and triggered manner. Thermoresponsive nanogels (tNG), releasing their cargo above

a defined

temperature, are promising carrier systems for inflammatory skin diseases, where the temperature of diseased skin differs from that of healthy skin areas. In this study a polyglycerol-based tNG with diameter of 156 nm was investigated for penetration and release properties upon topical application on ex vivo human skin with intact or disrupted barrier. Furthermore, temperature-triggered effects and the internalization of tNG by skin cells upon translocation to the viable skin layers were analyzed. The investigated tNG were tagged with indodicarbocyanine and loaded with fluorescein, so that fluorescent microscopy and flow cytometry could be used to evaluate simultaneously particle penetration and release of the fluorochrome. Topically applied tNG penetrated into the SC of both intact and disrupted skin explants. Only in barrierdisrupted skin significant amounts of released fluorochrome and tNG penetrated in the epidermis and dermis 2 h after topical application. When a thermal trigger was applied by infrared radiation (30 s 3.9 mJ/cm2), a significantly higher penetration of tNG in the SC and release of the dye in the epidermis were detected with respect to non-triggered samples. Penetrated tNG particles were internalized by skin cells in both epidermis and dermis. Only few CD1a-positive Langerhans cells associated with tNG were found in the epidermis. However, in the dermis a significant percentage of cells associated with tNG were identified to be antigen presenting cells, i.e. HLA-DR+ and CD206+ cells. Thus, tNG represent promising carrier systems for the treatment of inflammatory skin diseases, not only because of their improved penetration and controlled release properties, but also because of their ability to effectively reach dermal dendritic cells in barrier-disrupted skin.

2

Keywords: nanocarrier; skin penetration; drug release; human excised skin; dendritic cells; cellular uptake; infrared radiation; barrier disruption; tape stripping

Abbreviations:

dPG,

dendritic

polyglycerol;

FL,

fluorescein;

IDCC,

indodicarbocyanine, IR, infrared; LCs, Langerhans cells; MFI, mean fluorescence intensity; relMFI, relative mean fluorescence intensity; SC, stratum corneum; Tcp, cloud point temperature; tNG, thermoresponsive nanogels; tPG, thermoresponsive polyglycerol; TS, tape stripping

3

1

Introduction

Several characteristics of inflamed skin suggest that carrier-based drug delivery systems could help improve the efficacy and selectivity of topical therapies. For instance, inflammatory skin diseases are associated with impaired skin barrier function and stratum corneum (SC) integrity [1]. Moreover, inflammatory stimuli result in dynamic regulation of epidermal tight junctions as reported for several chronic inflammatory diseases [2]. The fact that these two key compartments of the skin barrier are profoundly disturbed compared to healthy skin suggest a facilitated penetration of large molecules and nanoparticulate formulations into lesional skin, and an increased accessibility of underlying inflamed tissue for carrier-based drug delivery system. Last but not least, this also includes facilitated access to immune cells. Increased uptake capacities of antigen-presenting cells in an activated state [3] encourage the exploration of cell targeting strategies including resident antigenpresenting cells, but also inflammatory infiltrates. The uptake of nanoparticles by Langerhans cells (LCs) and keratinocytes, as the first cell types encountering penetrating compounds, has been described by various groups in mice [4, 5] as well as in human skin [6, 7], mostly in the context of transcutaneous vaccination approaches [8]. However, the possibility of particle uptake by other cell types, e.g. epidermal T-cells, or by cells in the dermal compartment is less well explored. We recently obtained first evidence for penetration and cellular uptake of nanogels within the different skin layers [9]. Nanogels refer to aqueous dispersions of polymers which are chemically or physically cross-linked to form nanometer sized particles [1012]. Their attractiveness as drug delivery systems originates from the high loading capacity, long-term stability, and their responsiveness to different stimuli [13]. The considerable amount of water in the swollen nanogels enables the loading of biologically active molecules by electrostatic, van der Waals, or hydrophobic interactions, and release of the therapeutic payload in a controllable fashion. Thermoresponsive nanogels (tNG) exhibit a volume phase transition in response to temperature [14]. In aqueous solutions, polymers which have a lower critical solution temperature, are hydrated below the cloud point temperature (Tcp) and cause the tNG to swell. Above the Tcp the polymer becomes insoluble in water and cause the 4

tNG to collapse by repealing the inner solvent. The expulsion of water could be accompanied by the release of loaded hydrophilic or hydrophobic drugs upon shrinkage of the tNG. Such triggered drug release can be achieved after accumulation of the tNG in target tissues with locally elevated temperature, e.g. due to infection or inflammation, as well as during tumor hyperthermia or by artificially elevating the temperature of the region to be treated using an external thermal trigger [15]. In this study, we investigated a tNG based on dendritic polyglycerol (dPG) and linear thermoresponsive polyglycerol (tPG) consisting of poly(glycidyl methyl ether - co ethyl glycidyl ether). The unique characteristics of dPG make these molecules extremely relevant for biomedical applications [16, 17]. The multifunctional surface of free hydroxyl groups enable chemical modifications of highest precision to control their properties in terms of chemical reactivity, solubility, their role as macromolecular crosslinker and indodicarbocyanine (IDCC) dye labeling. tPG has a structural similarity to the biocompatible PEG and allows to tune the Tcp in a broad range of temperatures depending on the ratio of the copolymerized monomers. For these reasons, it was chosen as thermoresponsive unit in the crosslinked polymer network [18]. The penetration and release properties of tNG were investigated after topical application on human skin explants with intact and disrupted barrier. Also, the effects of an external thermal trigger on these properties was explored using barrierdisrupted skin samples. Furthermore, the internalization of tNG by cells of both, epidermis and dermis, was analyzed. Particular attention was given to skin immunocompetent cells, since these cells are prone to take-up particulate material, and they have been shown to play a central role in both initiation and maintenance of skin conditions like psoriasis and atopic dermatitis [19].

5

2

2.1

Material and Methods

tNG synthesis and characterization

The tNGs were synthesized according to a previously reported method [18]. Briefly, for the synthesis of tPG based nanogels (tPG_tNG), dPG functionalized with (1R,8S,9s)-bicyclo[6.1.0]non-4-yn-9-ylmethyl carbonate (dPG-BCN8%) (10 mg) and di-azide functionalized tPG (tPG-(N3)2) (20 mg) were mixed in 1 mL of dimethylformamide (DMF), cooled in an ice bath and injected with a syringe into 20 mL of water at 45 °C. The mixture was stirred for 3 h and the unreacted alkynes were quenched with IDCC azide . The product was purified by dialysis membrane with a molecular weight cut-off (MWCO) of 50 kDa in water, for at least two days. For the encapsulation, nanogels were let to swell in a solution of sodium fluorescein (FL) (50 wt. %). The mixture was stirred overnight under exclusion of light. Following, the tNG with encapsulated dye were separated from the free dye by washings five times with a centrifugal filtering device. tNG were characterized by

1

H-nuclear magnetic

resonance (1H-NMR), dynamic light scattering (DLS), and UV-Vis spectroscopy. The hydrodynamic diameter size was found to be 156.04 nm (PDI = 0.120). The Tcp was determined by temperature-dependent UV-Vis transmission measurements to be 34.0 °C. The amounts of loaded dyes were 0.099 and 0.23 wt. % for IDCC and FL, respectively.

2.2

Skin samples

Excised human skin was obtained from informed, healthy donors who underwent plastic surgery. The study was conducted with consent of the subjects, after approval by the Ethics Committee of the Charité – Universitätsmedizin Berlin and in accordance with the Declaration of Helsinki guidelines (approval EA1/135/06, renewed on July 2015). Skin was used within 3 and 24 h after surgery. Skin was stretched on a Styrofoam block and 2 cm² large areas, free from injuries or redness, were cleaned with phosphate buffer saline (PBS). In order to disrupt the SCbarrier 50 TS were performed. This treatment has widely be used to remove part of the SC and 6

thereby simulate the skin barrier impairment typical of certain skin conditions [20]. Thereafter, 20 µL/cm2 of the tNG suspension (2 mg/mL, 40 µg/cm2) or FL (0.1 µg/cm2) were applied on skin with intact and disrupted barrier. For each experiment, a control was prepared applying 20 µL/cm2 of a sterile saline solution (NaCl, 0.9%). To avoid samples' dehydration, the incubation took place in humidified chambers. After 2 h incubation at 37 °C or at room temperature (RT) (see next section), non penetrated material was removed from skin surface using cotton swabs. A skin area of 0.5 x 0.5 cm was cut and shock-frozen in liquid nitrogen prior to cryo-sectioning. The remaining tissue (1.5 x 1.5 cm) was further processed to isolate skin cells.

2.3

Irradiation experiments

To visualize the effects of tNG thermoresponsivity on skin penetration and drug release, an external thermal trigger was applied on skin samples with topically applied tNG. First, 1 cm² large areas were treated with 50 TS and tNG (40 µg/cm2) were applied. Then, skin was irradiated for 30 s with an infrared (IR)-lamp (Philips Infrared RI 1521, Germany), with broad irradiation spectrum and a power density in skin surface of 116.8 mW/cm², at a distance of 40 cm from skin surface. This corresponded to a light doses of 3.9 mJ/cm2. During this short time of irradiation the skin surface temperature reached a maximum of 40 °C, as measured by means of an IR thermometer (Basetech, Germany). Samples treated with tNG but not irradiated were prepared for comparison. Samples were then kept in the dark and incubated for 2 h at RT. Thereafter, the non penetrated material was removed, skin was cut in four blocks of 0.5 x 0.5 cm and snap-frozen for cryosectioning.

2.4

Preparation of cryosections and fluorescence microscopy

The skin blocks were placed in tissue freezing medium (Leica Microsystems, Germany). Cryosections of 6 μm thickness were prepared using a microtome (2800 Frigocut-N, Reichert-Jung, Heidelberg, Germany). Sections were observed by means of a fluorescence microscope (BX60, Olympus). Fluorescence of samples and controls were observed using filter combinations of various wavelength; BP = 545 7

580 nm, LP >610 nm for IDCC and BP = 470 – 490 nm, LP >550 nm for FL. Pictures (magnification of x200) of at least 20 randomly chosen skin sections per donor and skin sample were taken using the same camera settings. The mean fluorescence intensity (MFI) of each area was calculated using the ImageJ software. Mean values for SC and epidermis of three different donors were calculated. For each donor, the relative mean values were calculated (MFI sample / MFI control = rel. MFI) in order to take off skin background fluorescence. Averages and standard deviations were reported in the diagrams using Microsoft Excel. ANOVA and student-T tests were used for data statistical analysis.

2.5

Isolation of skin cells and flow cytometry analysis

After incubation and removal of tNG suspension from skin surface, samples were cut in small pieces (0.2 x 0.2 cm). The cutting was performed from the dermis towards the epidermis in order to avoid that particles are transferred in the skin by the cutting blade. Thereafter, the sliced skin was incubated with 2.4 U/mL dispase solution (RocheApplied Science, Germany) for 16-20 h (overnight) at 4 °C in order to separate the epidermis from the dermis. Trypsin digestion of the epidermis sheets (0.25% trypsin, 1 M CaCl 2) was then performed over 10 minutes at 37 °C. Dermis was treated for 2 h at 37 °C or 16 h at 4 °C with an enzyme cocktail (3 mg/mL collagenase, 1.5 mg/mL hyaluronidase, 10 µg/mL DNase). After tissue digestion, cells were harvested by repeated pipetting and filtering through a 70 µm cell strainer (FalconTM, Becton Dickinson, Germany). Cells were collected by centrifugation (300 g, 10 minutes), fixed with 2 % PFA (Sigma-Aldrich, Germany), and stored at 4°C. Cells were then washed with 1 mL PBS, re-suspended in a protein blocking solution (5 % BSA, 1 mM EDTA in PBS) and incubated with 10 µL antibody for 20 min at RT under constant stirring. For epidermal cells, a monoclonal mouse anti-human CD1a antibody (Immunostep, Germany) tagged with FITC was used to visualize LCs and a FITC mouse antihuman CD3 antibody (BD Biosciences TM, Germany) was used to visualize T-cells. For dermal cells staining, a FITC mouse anti-human HLA-DR antibody (BD 8

BiosciencesTM, Germany) was used to visualize antigen presenting cells and Alexa Fluor 488 anti-human CD206 antibody (BioLegend, Germany) was used to label macrophages and non-activated dendritic cells. After incubation, cells were added with 5 mL PBS, collected by centrifugation (160 g, 10 min), and stored at 4 °C until analysis by flow cytometry (FACS Calibur, BD, Germany). At least 20.000 events were collected in the selected gate. The analysis and evaluation of the measured data was performed with the software FCS Express version 3.1 (De Novo Software, USA). The results are represented using histograms and dot plots. The mean fluorescent intensity (MFI) of samples was normalized to that of the respective controls to obtain relative MFI (relMFI) values.

9

3

3.1

Results and Discussion

Fluorochrome and nanogel penetration on skin with intact and disrupted barrier

The release properties of topically applied tPG-based NGs were investigated on human excised skin with intact barrier and in skin pre-treated with 50 tape-stripping (TS) in order to disrupt the SC barrier. Skin sections (6 µm) were prepared and observed by means of fluorescence microscopy to detect skin penetration of the released dye (FL) and of the tNG (IDCC). Figure 1 shows the analysis of skin sections from four independent experiments. In Figure 1a a representative image of an overlay of green and red fluorescence pictures from a skin section is shown. The image shows that, after a 2 h incubation, tNG localized predominantly in the upper part of the SC, whereas FL was localized in the SC as well as in the lower skin layers. To carry out a statistical analysis of FL release, the mean fluorescent intensity (MFI) of different skin areas of four different donors was determined using the ImageJ software. Normalized MFI values (sample MFI/ control MFI = rel. MFI) are plotted in figure 1b-d. In Figure 1b the penetration in barrier disrupted skin of free FL was compared with the penetration of FL released by tNG (n=1). Higher FL intensities than untreated control were detected for both FL and tNG samples. Interestingly, whereas in the free FL sample similar amounts of FL were detected in SC, viable epidermis, and dermis, in the tNG sample a decreasing FL concentration gradient was observed between the three analyzed skin layers. In all skin layers, higher fluorescence intensities were detected of tNG samples with respect to FL samples, showing that tPG_tNG create a depot for FL in the SC and enhance the amounts of substance delivered to the epidermis and dermis. In further three independent experiments, tPG_tNG were applied on skin with intact (0x TS) and disrupted (50x TS) SC and the penetration of tNG and FL was investigated. As for tNG' associated fluorescence, the highest signals were detected in the SC (Fig. 1c). Significant higher signals with respect to controls were detected also in the viable epidermis of all three donors but only in barrier disrupted skin. In general it is assumed, that only moderately lipophilic drugs with a molecular weight 10

below 500 Dalton are able to cross the intact skin barriers through the lipophilic, apolar route of SC extracellular matrix [21]. However, skin penetration studies on deformable carrier systems [22] and recent work on the penetration of model compounds using superresolution microscopy [23] give a more differentiated view on this paradigm. Based on such studies, we believe that tPG_tNG (156 nm in diameter) were capable of penetrating to deeper layers of the SC and, in small amounts, also in the viable epidermis and dermis, not only because skin barrier was disrupted but also as a result of their deformability, cross-linked nature, and high water content, which might act as penetration enhancer. Obviously, the skin barrier disruption largely contributed to the enhanced penetration. This treatment removes the superficial layers of the SC and compromises the integrity and stability of the deeper corneocyte layers, probably by disrupting corneodesmosomes and hook-like structures between corneocytes. As for FL, the highest signals were detected in the SC. However, in barrier disrupted skin samples, a clearly higher penetration of FL was observed in the viable epidermis as well as in the dermis (Fig. 1d). Also in these three further donors, a gradient of FL signal intensity could be observed between SC, viable epidermis, and dermis. Overall, the penetration of tNG in SC correlated well with the delivery of FL in the skin viable layers, indicating that, when skin barrier integrity is altered, a significantly higher penetration of low molecular weight molecules, like FL, as well as macromolecular nanocarriers, like tNG, can be achieved.

3.2

Penetration enhancement upon thermo-triggered phase transition of topically applied nanogels

To visualize thermoresponsive effects upon topical application of tPG_tNG on barrier disrupted human skin explants, skin penetration of tNG and FL was analyzed on tape-stripped skin samples that had been treated with tPG_tNG and then irradiated for 30 s with an IR lamp or kept in the dark previous further incubation for 2 h at RT (Fig. 2). This short irradiation time and the further 2 h incubation at RT ensured that the measured effects are due mainly to the thermoresponsive property of the nanogel 11

and not to an increased tissue permeability at higher temperature. Fig. 2a represents a typical image of a skin samples where tNG were topically applied and incubated for 2 h at RT, whereas in Fig. 2b a section is shown from skin where tNG conformational change was triggered by IR radiation previous incubation at the same conditions. These two representative images show that in the irradiated sample higher fluorescence intensity was recorded with respect to the non-irradiated sample; higher tNG signal was observed in the SC, whereas higher FL signal was noticeable in the epidermis. Figure 2c shows the analysis of at least 20 sections per sample for tNG skin penetration. In the SC, higher fluorescence intensity was detected in the IRtreated sample (rel. MFI 2.02  0.6) with respect to the non-irradiated sample (rel. MFI 1.4  0.3), indicating that the conformational change induced by the thermal trigger favors the penetration of tNG in the SC. Figure 2d reports the skin penetration of FL released from the tNG. In both samples (with or without IR radiation), a clearly high fluorescent signal with respect to control skin was detectable in the SC as well as in the viable epidermis. With a rel. MFI of 34.3  7, the temperature triggered release of FL into the SC was not significantly higher than that of non-irradiated sample (rel. MFI 31.2  8.5). However, in IR-treated samples significant higher fluorescence intensity with respect to non-irradiated skin samples was detected in the viable epidermis as well as in the dermis (rel. MFI of 9.6  5 and 2.0  0.7, respectively) as compared to samples kept in the dark (rel. MFI of 3.6  2 and 1.1  0.3, respectively). These results show that FL was released by simple diffusion when the tNG were incubated on skin at RT. However, when the external thermal trigger was applied, tNG underwent a reversible change in conformation that caused a further triggered release of the encapsulated FL. Thus, the external thermal trigger enhanced not only tNG penetration in the SC (Fig. 2c) but resulted also in improved delivery of FL to the epidermis (Fig. 2d). This correlated well with in vitro data showing conformational change, size reduction, increase of hydrophobicity and release of encapsulated cargo when the temperature is set above the Tcp of the investigated tNG [24, 25].

12

3.3

Cellular uptake of tNG upon topical application on barrier-disrupted skin

In previous studies [9, 25], we showed that small amounts of tNG penetrated across the intact SC barrier and were internalized by epidermal cells. Also in this study, the analysis of skin sections incubated for 2 h at 37 °C showed a fluorescent signal indicating the presence of tNG also in the viable epidermis, especially for skin samples with disrupted skin barrier. Thus, further analyses were carried out to find out if the investigated tNG can be taken-up by skin cells. First, in vitro experiments were performed using primary keratinocytes and the keratinocyte cell line HaCaT (Fig. S1, Support Information) to test if tNG can be internalized by epidermis cells. Confocal laser scanning microscopy of primary keratinocytes showed that tNG were internalized when cells are incubated for 2 h at 37 °C (Fig. S1a). tPG_tNG were localized in vesicles in the perinuclear region, indicating an endocytotic process. FL released by the tNG was also detected within the cells but with a more homogeneous distribution pattern than the tNG. HaCat cells were incubated with tNG at 4 and 37 °C (Fig. S1b) and analyzed by flow cytometry. No significant fluorescence intensity, with respect of control cells, was detected for cells incubated with the free dye FL at 4 °C and even at 37 °C (Fig. S1c). This is due to the fact that the free dye was washed out during the wash steps after incubation. On the contrary, a clear shift of the whole cell population was observed for samples incubated at 37 °C. Because of their size, tNG can be only taken-up by the energydependent process of endocytosis, whereby they are internalized in vesicles and thus cannot be washed out as the free dye [26, 27]. In addition, we could show that the internalization of tNG could be detected by flow cytometry. In order to find out if the tNG were also able to cross the SC and be internalized by skin cells, excised human skin from four different donors was treated with the investigated tNG and epidermis and dermis cells were isolated and analyzed by flow cytometry (Fig. 3). Figure 3a,b shows examples of flow cytometry histograms where the fluorescence intensity of isolated epidermis (Fig. 3a) and dermis (Fig. 3b) cells is plotted against cell counts. It is visible that the histograms of cells from skin samples with disrupted barrier are shifted with respect to control histograms (see insets). Figure 3c shows the averages of normalized MFI values from three different donors. 13

Whereas epidermis cells isolated from skin with intact barrier had rel. MFI values of 1.2 ± 0.5, i.e. fluorescence intensities similar to control cells, epidermal cells isolated from skin with impaired SC had higher relMFI values (1.7 ± 0.9). Also cells isolated from the dermis of barrier disrupted skin had higher MFI values (2.2 ± 0.6) than those of skin with intact barrier (1.7 ± 0.6). Whereas a clear shift was observed in the in vitro experiments, more moderate changes of cell fluorescence intensity were detected in the ex vivo experiments. Nevertheless, the results were reproducible and indicate that small amounts of cells isolated from the viable skin layers were associated with the investigated tNG. To draw comparisons between topically applied tNG and the dye used to tag the nanocarrier (IDCC), cells were analyzed after tNG and IDCC were applied on skin with disrupted barrier (Fig. 3d). Cells from skin treated with tNG showed higher fluorescence signals with rel. MFI values of 4.75 (epidermis) and 1.84 (dermis) compared to the samples treated with free dye (rel. MFI of 2.16 and 1.11 for epidermis and dermis, respectively). Despite the fact that the concentration of IDCC was the same in the two samples, cells from skin incubated with IDCC only had less dye because this was washed out during the isolation process. On the contrary, high molecular weight compounds like nanocarriers can be internalized or released only by energy-dependent processes, i.e. endocytosis and exocytosis, and cannot be simply washed out. Thus, these results show that tNG can penetrate the SC as a whole moiety and can be taken-up by skin cells. One step of the isolation process is the enzymatic treatment of dermis connective tissue, which is normally conducted at 37 °C. During this step, further uptake of penetrated tNG might occur. To find out if this was occurring, in this experiment the enzymatic digestion was performed at 4 °C, temperature at which all endocytosis processes are inhibited. The fact that also at these conditions similar rel. MFI were found confirmed that the up-take had occurred during tNG incubation with skin and not during the isolation process. We can summarize that, already after 2 h of incubation on skin surface, the investigated tNG could cross a disrupted SC and were taken-up by cells in both epidermis and dermis.

14

3.4

Identification of cell populations with internalized tNG

In order to investigate which populations of epidermal and dermal cells have internalized the penetrated tNG, cells were stained with tagged antibodies specific for CD markers expressed by selected epidermis and dermis cell populations. Among the epidermal cells, keratinocytes and LCs are clearly most likely to internalize particles once they come in contact, both from the biological point of view and with regard to their distribution within the skin. Recently, however, increasing studies report on the role of epidermal T cell populations as well as on their interaction with skin dendritic cells [28, 29]. Even carrier-based targeting strategies of lymphocytes are increasingly being reported [30, 31]. Thus, epidermal cells were stained with CD1a antibodies specific for LCs [32] and with CD3 antibodies to visualize epidermal T-cells [33]. Within dermis cells, antigen presenting cells were marked using HLADR-antibodies specific for class II major histocompatibility complex [34] and CD206antibodies specific for the C-type 1 lectin receptor, which is predominantly expressed in macrophages and immature dendritic cells [35, 36]. The percentages of cells associated with the tNG signal were calculated and plotted in Figure 4. Two main populations of cells could be distinguished in the epidermis: positive or negative for the chosen CD marker. Small amounts of cells were also positive for tNG (Fig. 4 a, right quadrants). The data from three different donors showed that, in most of the cases, CD1a- and CD3-positive cells were not involved in the uptake of tNG. A significant uptake was detected mainly for CD1a- and CD3-negative cells from skin with disrupted barrier. The only skin sample that resulted positive for tNG uptake also without previous tape-stripping treatment (open squares and circles in Fig.4 c,d) was prepared with skin that had been transported in a isotonic solution after surgery. This might have caused hydration of SC and swelling of the epidermis with consequent alteration of skin barrier permeability. This was also the only sample in which a clear uptake of tNG by LCs was detected (Fig. 4c, open black circle). In the case of dermis, depending on the used gate different cell populations were observed (Fig. S2). When almost all cells

were gated (Fig. S2a, gate b), three

different cell populations were identified, with one population having low side and forward scatter and low auto-fluorescence, corresponding mainly to T-cells, and two 15

cell populations with higher auto-fluorescence that had different levels of HLA-DR and CD206 expression (Fig. S2b). When only cells with high forward and side scatter were gated (Fig. S2a, gate c), the T-cell population was excluded allowing the analysis of the other two cell populations (Fig. S2c). For dermal T-cells, no association with tNG was recorded, whereas within the other cell types, uptake of tNG was detected, with higher percentages in case of samples with disrupted skin barrier (Fig. S2d,e). When cells with higher side and forward scatter were investigated (Fig. 4 b), it was visible that both HLA-DR-negative and HLA-DR-positive cells had internalized tNG (Fig. 4e), but mainly for samples with disrupted or altered skin barrier. In fact, for skin samples without the TS pre-treatment, only the samples that were transported in isotonic solution (empty black circles) resulted positive for nanocarrier uptake. Similar findings could be observed for CD206 stained cells (Fig. 4f). Thus, when skin barrier was altered, tNG were able to penetrate to both, viable epidermis and dermis. After 2 h of incubation, LCs in the epidermis were not principally involved in the uptake of tNG. On the other side, in the dermis, part of the cells associated with tNG were HLA-DR and CD206 positive cells, i.e. antigen presenting cells. In previous results of our group, we could detect uptake of both polylactic acid (PLA) and polystyrene (PS) particles by LCs after skin barrier disruption, however after 16 h of incubation [7]. Clearly LCs have the capacity to migrate to the dermis, and in our study, the damage of skin barrier by repetitive TS can induce such event due to cytokine induction and LCs activation. Still, the time window for this generally observed process lies within several hours, e.g., the migratory activity of LCs in mice treated with topical FITC application peaked at 18-24 h [37]. In detailed studies on the uptake process of topically applied biotin into the dendrites of activated LCs, Kubo et al. found positive results at 16 h[3], which is in line with our own previous observations of nanoparticle uptake across disrupted skin barrier [4, 6, 7]. Current investigations using longer incubation time (16 h) on tapestripped skin, showed the presence of a small populations of LCs with internalized tNG (data not shown). However, these cells were found rather in the dermis and not in the epidermis. This suggest that a small number of LCs can internalize tNG and migrate to the dermis after longer incubation time. 16

With regard to the data presented in this study, however, the remarkable insight is that, despite looking at this early time point, we clearly have uptake of tNG by dermal antigen-presenting cell populations. These results are in line with the reported motile behavior of skin dendritic cells. In fact, using intravital microscopy Ng et al. had demonstrated that LCs appeared as rather sessile cells with a cycling habitude of extension and retraction kwon as dSEARCH, while dermal dendritic cells were found to traffic in the interstitial tissue with significant higher velocities, effectively internalizing injected parasites in uninflamed skin within 2-3 h [38]. Taken altogether the existing data on uptake behavior of those cell populations and the results presented for penetration and cellular uptake of topically applied nanogels suggest that, especially in inflamed skin, particle-based delivery of compounds is feasible even to cells in the dermis. In addition, it suggests that not only epidermal processes but also cellular pathologies with dermal cell infiltrates could effectively be reached by topical means. The fact that dendritic cells of barrier-disrupted skin can internalize tNG suggest the possibility to use these nanocarriers to target such cell populations. For instance, in skin inflammatory diseases such as atopic dermatitis or psoriasis, skin barrier is disrupted and population of dendritic cells were found to play a crucial role in the initiation and maintenance of the inflammatory process [19, 39].

17

4

Conclusion

Many inflammatory skin diseases are characterized by a disturbance of the SC integrity and immune cell activation. Therefore, the understanding of nanocarrier interactions with disrupted skin barrier and immune

cells is an important step

towards the development of innovative carrier systems for the topical treatment of skin inflammatory diseases [39, 40]. One limitation of many topical treatments for inflammatory skin conditions is that the applied drugs permeate to the lower dermis from where they can diffuse to the blood compartment. This rapid clearance reduces the concentration of drug at the site of action and is often associated with systemic toxicity. In this study, it was shown that tPG-based tNG have advantageous delivery properties with respect to the free dye formulation, especially in skin with altered barrier integrity. A good correlation between the penetration of tNG in the SC and the delivery of FL in the skin layers was detected. tNG created a depot in the SC and released locally their cargo creating a gradient in the different skin layers. Further, IR irradiation experiments put in evidence that enhancement of tNG penetration and drug delivery to skin can be triggered by an external thermal stimulus. This suggests also the use of tNG able to sense the temperature of inflamed skin, in order to enhance the selective delivery of drugs to diseased skin regions. Finally, different cell populations of epidermis and dermis were identified which are mainly involved in the interaction with penetrated tNG. Interestingly, within dermis, significant percentages of antigen presenting cells with internalized tNG were detected, suggesting the possibility to use tNG to target drugs to specific cell population of the dermis. These observations could further be developed for specific targeting approaches using nanogels functionalized with targeting units to increase the selectivity for selected cell populations.

Acknowledgments We gratefully thanks financial support by the German Research Foundation (DFG) within SFB 1112, projects C04 and A04, and the Bundesministerium für Bildung und Forschung (BMBF) through the NanoMatFutur award (13N12561, Thermonanogele). 18

References

[1] P.M. Elias, L.F. Eichenfield, J.F. Fowler Jr, P. Horowitz, R.P. McLeod, Update on the structure and function of the skin barrier: atopic dermatitis as an exemplar of clinical implications, Semin. Cutaneous Med. Surg. 32 (2013) 21-24. [2] J. Brandner, M. Zorn-Kruppa, T. Yoshida, I. Moll, L. Beck, A. De Benedetto, Epidermal tight junctions in health and disease, Tissue barriers 3 (2015) e974451. [3] A. Kubo, K. Nagao, M. Yokouchi, H. Sasaki, M. Amagai, External antigen uptake by Langerhans cells with reorganization of epidermal tight junction barriers, J. Exp. Med. 206 (2009) 2937-2946. [4] B. Mahe, A. Vogt, C. Liard, D. Duffy, V. Abadie, O. Bonduelle, A. Boissonnas, W. Sterry, B. Verrier, U. Blume-Peytavi, Nanoparticle-based targeting of vaccine compounds to skin antigen-presenting cells by hair follicles and their transport in mice, J. Invest. Dermatol. 129 (2009) 1156-1164. [5] A. Kolonics, Z. Csiszovszki, E.R. Tőke, O. Lőrincz, D. Haluszka, R. Szipőcs, In vivo study of targeted nanomedicine delivery into Langerhans cells by multiphoton laser scanning microscopy, Exp. Dermatol. 23 (2014) 596-605. [6] A. Vogt, B. Combadiere, S. Hadam, K.M. Stieler, J. Lademann, H. Schaefer, B. Autran, W. Sterry, U. Blume-Peytavi, 40 nm, but not 750 or 1,500 nm, nanoparticles enter epidermal CD1a+ cells after transcutaneous application on human skin, J. Invest. Dermatol. 126 (2006) 1316-1322. [7] F. Rancan, S. Amselgruber, S. Hadam, S. Munier, V. Pavot, B. Verrier, S. Hackbarth, B. Combadiere, U. Blume-Peytavi, A. Vogt, Particle-based transcutaneous administration of HIV-1 p24 protein to human skin explants and targeting of epidermal antigen presenting cells, J. Controlled Release 176 (2014) 115-122. [8] B. Combadière, B. Mahé, Particle-based vaccines for transcutaneous vaccination, Comp. Immunol. Microbiol. Infect. Dis. 31 (2008) 293-315. [9] F. Rancan, M. Asadian-Birjand, S. Dogan, C. Graf, L. Cuellar, S. Lommatzsch, U. BlumePeytavi, M. Calderón, A. Vogt, Effects of thermoresponsivity and softness on skin penetration and cellular uptake of polyglycerol-based nanogels, J. Controlled Release 228 (2016) 159169. [10] A.V. Kabanov, S.V. Vinogradov, Nanogels as pharmaceutical carriers: finite networks of infinite capabilities, Angew. Chem., Int. Ed. 48 (2009) 5418-5429. [11] M. Molina, M. Asadian-Birjand, J. Balach, J. Bergueiro, E. Miceli, M. Calderón, Stimuliresponsive nanogel composites and their application in nanomedicine, Chem. Soc. Rev. 44 (2015) 6161-6186. [12] M. Asadian-Birjand, A. Sousa-Herves, D. Steinhilber, J. Cuggino, M. Calderon, Functional nanogels for biomedical applications, Curr. Med. Chem. 19 (2012) 5029-5043. [13] N. Smeets, T. Hoare, Designing responsive microgels for drug delivery applications, J. Polym. Sci., Part A: Polym. Chem. 51 (2013) 3027-3043. [14] L. Zha, B. Banik, F. Alexis, Stimulus responsive nanogels for drug delivery, Soft Matter, 7 (2011) 5908-5916. [15] M. Malmsten, Soft drug delivery systems, Soft Matter, 2 (2006) 760-769. [16] J. Khandare, M. Calderón, N.M. Dagia, R. Haag, Multifunctional dendritic polymers in nanomedicine: opportunities and challenges, Chem. Soc. Rev. 41 (2012) 2824-2848. [17] M. Calderón, M.A. Quadir, S.K. Sharma, R. Haag, Dendritic polyglycerols for biomedical applications, Advanced Materials 22 (2010) 190-218. [18] M. Giulbudagian, M. Asadian-Birjand, D. Steinhilber, K. Achazi, M. Molina, M. Calderón, Fabrication of thermoresponsive nanogels by thermo-nanoprecipitation and in situ encapsulation of bioactives, Polym. Chem. 5 (2014) 6909-6913.

19

[19] C.-C. Chu, P. Di Meglio, F.O. Nestle, Harnessing dendritic cells in inflammatory skin diseases, Semin. Immunol. 23 (2011) 28-41. [20] S.J. Bashir, A.L. Chew, A. Anigbogu, F. Dreher, H.I. Maibach, Physical and physiological effects of stratum corneum tape stripping, Skin Res. Technol. 7 (2001) 40-48. [21] J.D. Bos, M.M. Meinardi, The 500 Dalton rule for the skin penetration of chemical compounds and drugs, Exp. Dermatol. 9 (2000) 165-169. [22] G. Cevc, U. Vierl, Nanotechnology and the transdermal route: a state of the art review and critical appraisal, J. Controlled Release 141 (2010) 277-299. [23] P. Volz, A. Boreham, A. Wolf, T.-Y. Kim, J. Balke, J. Frombach, S. Hadam, Z. Afraz, F. Rancan, U. Blume-Peytavi, Application of single molecule fluorescence microscopy to characterize the penetration of a large amphiphilic molecule in the stratum corneum of human skin, Int. J. Mol. Sci. 16 (2015) 6960-6977. [24] M. Witting, M. Molina, K. Obst, R. Plank, K.M. Eckl, H.C. Hennies, M. Calderón, W. Frieß, S. Hedtrich, Thermosensitive dendritic polyglycerol-based nanogels for cutaneous delivery of biomacromolecules, Nanomedicine 11 (2015) 1179-1187. [25] M. Asadian-Birjand, J. Bergueiro, F. Rancan, J. Cuggino, R.-C. Mutihac, K. Achazi, J. Dernedde, U. Blume-Peytayi, A. Vogt, M. Calderón, Engineering thermoresponsive polyether-based nanogels for temperature dependent skin penetration, Polym. Chem. 6 (2015) 5827-5831. [26] H. Kettiger, A. Schipanski, P. Wick, J. Huwyler, Engineered nanomaterial uptake and tissue distribution: from cell to organism, Int J Nanomedicine 8 (2013) 3255-3269. [27] S.D. Conner, S.L. Schmid, Regulated portals of entry into the cell, Nature, 422 (2003) 37-44. [28] S.N. Mueller, A. Zaid, F.R. Carbone, Tissue-resident T cells: dynamic players in skin immunity, Front. Immunol. 5 (2014) 00332. [29] J. Seneschal, R.A. Clark, A. Gehad, C.M. Baecher-Allan, T.S. Kupper, Human epidermal Langerhans cells maintain immune homeostasis in skin by activating skin resident regulatory T cells, Immunity, 36 (2012) 873-884. [30] J.J. Glass, D. Yuen, J. Rae, A.P. Johnston, R.G. Parton, S.J. Kent, R. De Rose, Human immune cell targeting of protein nanoparticles–caveospheres, Nanoscale, 8 (2016) 82558265. [31] S.-Y. Li, Y. Liu, C.-F. Xu, S. Shen, R. Sun, X.-J. Du, J.-X. Xia, Y.-H. Zhu, J. Wang, Restoring anti-tumor functions of T cells via nanoparticle-mediated immune checkpoint modulation, J. Controlled Release 231 (2016) 17-28. [32] A. Wollenberg, S. Kraft, D. Hanau, T. Bieber, Immunomorphological and ultrastructural characterization of Langerhans cells and a novel, inflammatory dendritic epidermal cell (IDEC) population in lesional skin of atopic eczema, J. Invest. Dermatol. 106 (1996) 446-453. [33] C.A. Foster, H. Yokozeki, K. Rappersberger, F. Koning, B. Volc-Platzer, A. Rieger, J.E. Coligan, K. Wolff, G. Stingl, Human epidermal T cells predominantly belong to the lineage expressing alpha/beta T cell receptor, J. Exp. Med. 171 (1990) 997-1013. [34] F. Nestle, X.-G. Zheng, C. Thompson, L. Turka, B. Nickoloff, Characterization of dermal dendritic cells obtained from normal human skin reveals phenotypic and functionally distinctive subsets, J. Immunol. 151 (1993) 6535-6545. [35] F. Noorman, E. Braat, M. Barrett-Bergshoeff, E. Barbe, A. Van Leeuwen, J. Lindeman, D. Rijken, Monoclonal antibodies against the human mannose receptor as a specific marker in flow cytometry and immunohistochemistry for macrophages, J. Leukocyte Biol. 61 (1997) 63-72. [36] M. Kato, T.K. Neil, D.B. Fearnley, A.D. McLellan, S. Vuckovic, D.N. Hart, Expression of multilectin receptors and comparative FITC–dextran uptake by human dendritic cells, Int. Immunol. 12 (2000) 1511-1519. [37] S.E. Macatonia, S.C. Knight, A.J. Edwards, S. Griffiths, P. Fryer, Localization of antigen on lymph node dendritic cells after exposure to the contact sensitizer fluorescein isothiocyanate. Functional and morphological studies, J. Exp. Med. 166 (1987) 1654-1667.

20

[38] L.G. Ng, A. Hsu, M.A. Mandell, B. Roediger, C. Hoeller, P. Mrass, A. Iparraguirre, L.L. Cavanagh, J.A. Triccas, S.M. Beverley, Migratory dermal dendritic cells act as rapid sensors of protozoan parasites, PLoS Pathog. 4 (2008) e1000222. [39] M.A. Lowes, F. Chamian, M.V. Abello, J. Fuentes-Duculan, S.-L. Lin, R. Nussbaum, I. Novitskaya, H. Carbonaro, I. Cardinale, T. Kikuchi, Increase in TNF-α and inducible nitric oxide synthase-expressing dendritic cells in psoriasis and reduction with efalizumab (antiCD11a), Proc. Natl. Acad. Sci. U. S. A. 102 (2005) 19057-19062. [40] T.W. Prow, J.E. Grice, L.L. Lin, R. Faye, M. Butler, W. Becker, E.M. Wurm, C. Yoong, T.A. Robertson, H.P. Soyer, Nanoparticles and microparticles for skin drug delivery, Adv. Drug Delivery Rev. 63 (2011) 470-491.

21

Figures

Figure 1. Penetration and release properties of tPG_tNG in skin with intact and disrupted barrier (50x TS). After topical application of the investigated tPG_tNG (20 µL, 40 µg/ cm²) and the negative control (NaCl, 0.9 %), skin was incubated for 2 h at 37 °C. Pictures of skin sections were taken using filters specific for IDCC (c) and FL fluorescence (b, d). The analysis was performed on at least 20 section images from each donor. The mean fluorescence intensity of areas in stratum corneum (SC), viable epidermis (E), and dermis (D) was calculated using the ImageJ software. The mean values of samples were then normalized to those of respective controls (rel. MFI). (a) A representative overlay image of a section from skin treated with 50 TS and incubated with the investigated tPG_tNG. (b) Comparison of FL skin penetration when applied as an aqueous solution (FL 0.1 µg/cm 2) or encapsulated in tNG (tNG-FL). (c,d) Comparison of skin penetration of tPG_tNG (c) and FL (d) after application of the nanocarriers on skin with intact (0x TS) and disrupted barrier (50x TS). Skin sample of donor 3 was transported in isotonic solution. This might have altered skin morphology and permeability. *** p < 0.001; ** p < 0.01; * p < 0.5

22

Figure 2. Temperature triggered penetration and release properties of tPG_tNG on barrier disrupted human skin explants. After topical application of tNG (40 µg/cm²), one skin sample was irradiated with an IR lamp (116.8 mW) and the other was kept in the dark followed by previous further incubation at the same conditions (2 h, RT). The fluorescence intensity of skin sections was analyzed with ImageJ. (a) A representative image of skin treated with tNG but not irradiated. (b) Skin section of a sample treated with tNG and irradiated with the IR lamp. (c) Analysis of the penetration of tNG in SC, viable epidermis and dermis. (d) Analysis of FL skin penetration. At least 20 cryosections per sample were measured. The average MFI values of sample are normalized to the MFI values of control skin (0.9% NaCl). *** p < 0.001; ** p < 0.01; * p < 0.5

23

Figure 3 Cellular uptake of tNG by epidermis and dermis cells. Skin with intact (0x TS) and disrupted (50x TS) barrier was incubated with the investigated tNG (40 µg/cm²) for 2 h at 37 °C. Epidermal and dermal cells were then isolated and analyzed by flow cytometry. Representative histogram of cells isolated from epidermis (a) and dermis (b) are shown. In the insets, the scale of the axes was modified to point out the shift of samples with respect to controls. The averages of normalized MFI values are plotted in (c) (n=3). In skin from a fourth donor, the penetration of tNG after barrier disruption was compared to the penetration of free IDCC (d).

24

Figure 4. Analysis of epidermis (a,c,d) and dermis (b,e,f) cell populations with internalized tNG. Cells were isolated from skin samples, with intact (0x TS) and disrupted (50x TS) barrier, 2 h after tNG topical application and analyzed for particle uptake (right quadrants) and expression of CD markers (upper quadrants). The diagrams show the cell percentages of tNG-positive cells stained with CD1a (c), CD3 (d), HLA-DR (e) and CD206 (f).

Graphical Abstract

25

*Graphical abstract