Dendritic polyglycerol and N-isopropylacrylamide based thermoresponsive nanogels as smart carriers for controlled delivery of drugs through the hair follicle

Abstract

Nanoparticles with a size of several hundred nanometers can effectively penetrate into the hair follicles and may serve as depots for controlled drug delivery. However, they can neither overcome the hair follicle barrier to reach the viable cells nor release the loaded drug adequately. On the other hand, small drug molecules cannot penetrate deep into the hair follicles. Thus, the most efficient way for drug delivery through the follicular route is to employ nanoparticles that can release the drug close to the target structure upon exposure to some external or internal stimuli. Accordingly, 100–700 nm sized thermoresponsive nanogels with a phase transition temperature of 32–37 °C were synthesized by the precipitation polymerization technique using N-isopropylacrylamide as a monomer, acrylated dendritic polyglycerol as a crosslinker, VA-044 as an initiator, and sodium dodecyl sulphate as a stabilizer. The follicular penetration of the indodicarbocyanine (IDCC) labeled nanogels into the hair follicles and the release of coumarin 6, which was loaded as a model drug, in the hair follicles were assessed ex vivo using porcine ear skin. Confocal laser scanning microscopy (CLSM) enabled independent tracking of the nanogels and the loaded dye, although it is not as precise and accurate as standard analytical methods. The results showed that, unlike smaller nanogels (<100 nm), medium and larger sized nanogels (300–500 nm) penetrated effectively into the hair follicles with penetration depths proportional to the nanogel size. The release of the loaded dye in the hair follicles increased significantly when the investigation on penetration was carried out above the cloud point temperature of the nanogels. The follicular penetration of the nanogels from the colloidal dispersion and a 2.5% hydroxyethyl cellulose gel was not significantly different.

---

1. Introduction

Nanoparticles have been extensively studied as advanced drug delivery devices due to their unique features, such as capability to protect therapeutic compounds, targeted delivery, increased bioavailability, versatility to control the release profile of loaded drugs and tunable surface properties.^1,2 A number of studies also indicated that nanoparticles can enhance the penetration and permeation of drugs through the skin.^3–5 However, the mechanism by which the nanoparticles enhance drug penetration through the stratum corneum is not well explained and the transfollicular pathway was mentioned as the major penetration pathway.^6,7 Some papers also reported the penetration of nanoparticles into the deeper layers of a skin whose barrier is using microneedles or a razor blade,^8,9 but not through the intact stratum corneum. A recent interesting report demonstrated that the use of a low frequency sonophoresis, which was utilized as a penetration enhancer, reduced the percutaneous penetration of drugs due to the partial plugging of the hair follicles.¹⁰

On the other hand, studies indicated that nanoparticles with a size of several hundred nanometers penetrate effectively into the hair follicles and stay in the follicles for several days.¹¹ This tendency of nanoparticles makes them good candidates for the sustained and extended delivery of drugs through the skin. The follicular route might also be used to target distinct cell populations such as stem cells (e.g. nestin expressing follicular bulge cells).^12–14 However, the release of the encapsulated drug from the nanoparticles within the confined environment remains an issue¹¹ and there must be a mechanism to trigger drug release to ensure the desired targeting and sustaining effect of nanoparticles intended for transfollicular drug delivery. As a result, the formulation of smart nanocarriers which respond to various stimuli such as temperature, pH, and ionic strength has recently been given due attention.

Thermoresponsive nanoparticles, such as thermoresponsive nanogels^1,15,16 are mostly prepared by using thermoresponsive polymers that are capable of undergoing conformational changes from an extended/hydrophilic coil to a globular/hydrophobic state upon heating above a certain temperature known as the lower critical solution temperature (LCST).¹⁷ This behavior is assumed to be attributed to changes in intramolecular and intermolecular hydrogen bonding and hydrophobic interactions, and makes them attractive as smart tools in materials and biomedical sciences.¹⁸

Nanogels are loosely crosslinked polymeric chains that are arranged in a three-dimensional network. Thermoresponsive nanogels with great potential as drug delivery systems are commonly prepared from biocompatible hydrophilic thermoresponsive polymers such as poly(N-isopropylacrylamide) (PNIPAm), poly(glycidyl methyl ether-co-ethyl glycidyl ether), poly(N-vinylcaprolactam), poly(oligo(ethylene glycol)-methacrylate), and poly(N-dimethylacrylamide), which undergo reversible volume-phase transitions near the physiological temperature.^15,18,19 Other promising nanogels based on oligo ethylene glycol as the thermoresponsive unit and dendritic polyglycerol (dPG) as a crosslinker are also reported.^20–22

PNIPAm is a well-studied thermoresponsive polymer. Its LCST can be increased or decreased by using hydrophilic or hydrophobic comonomers and copolymers, respectively.^18,19 Consequently, a number of PNIPAm-based thermoresponsive nanogels that were synthesized by using different crosslinkers and synthesis methods have been reported.^5,16,19 dPG on the other hand is an efficient crosslinker which has good aqueous solubility, high biocompatibility and multi-functionality.^5,23,24

The synthesis of 100–200 nm PNIPAm thermoresponsive nanogels prepared by using acrylated dendritic polyglycerol as a macro-crosslinker was reported by our group.^5,16 These nanogels even showed promising results in enhancing the stability and permeation of proteins across the stratum corneum.⁵ Nonetheless, the nanogels are too small for optimal follicular penetration, while nanoparticles as big as 600 nm showed optimum penetration into the hair follicles.^7,11 Therefore, the objective of this study was to synthesize and characterize thermoresponsive dPG-PNIPAm nanogels of different sizes using the precipitation polymerization method and evaluate their potential as follicular drug delivery devices.

2. Materials and methods

2.1. Materials

dPG (Nanopartica GmbH, Berlin, Germany), acryloyl chloride, triethylamine, dry dimethylformamide, N-isopropylacrylamide (NIPAm, 97%), sodium dodecyl sulphate (SDS, 98%), 2,2′-azobis[2-methyl-N-(2-hydroxyethyl)propionamide] (VA-086), 2,2′-azobis[2-(2-imidazolin-2-yl)propane]dihydrochloride (VA-044), 2-iminothiolane, coumarin 6, potassium persulfate, phosphotungstic acid (Sigma-Alrdrich, Schnelldorf, Germany), maleimide functionalized indodicarbocyanine (IDCC) dye (Lumiprobe GmbH, Hannover, Germany), adhesive solution (Marabu GmbH & Co. KG, Bietigheim-Bissingen, Germany), hydroxyethyl cellulose (HEC, Euro OTC Pharma GmbH, Bönen, Germany), cryo spray (Solidofix®-Cryo spray, Carl Roth GmbH + C0. KG., Karlsruhe, Germany), and tissue freezing medium (Leica Biosystems Nussloch GmbH, Nussloch, Germany) were used as received. Milli-Q water was used throughout.

2.2. Methods

2.2.1. Synthesis of 9% acrylated dendritic polyglycerol (dPG-Ac). dPG with an average molecular weight of 10 kDa was synthesized according to a previously reported method.²⁵ 2 g of dPG (27.03 mmol OH equivalents) was dried in a 50 mL volumetric flask at 80 °C for 1 h under vacuum and was dissolved in 16 mL dry dimethylformamide at 60 °C. The solution was cooled by stirring in an ice bath for 10 min and 3.5 mmol triethylamine was added to the cold solution. Then 2.7 mmol acryloyl chloride was added dropwise and the reaction was allowed to run overnight. Finally the reaction was quenched by adding a small amount of water and dialyzed (dialysis membrane with molecular weight cut-off = 2 kDa, Sigma-Alrdrich, Schnelldorf, Germany) in water for 3 days.

2.2.2. Synthesis of IDCC labeled dPG-Ac. To a solution of dPG-(NH₂)₃% (100 mg, 0.04 amine equivalents, 3% of total OH groups converted to NH₂) synthesized according to a previously reported method,²⁶ 2-iminothiolane (4.12 mg, 0.015 mmol) was added. The mixture was stirred for 20 min and maleimide functionalized IDCC (19.23 mg, 0.015 mmol) was added. The mixture was allowed to react overnight to yield the dPG-IDCC conjugate. The product was obtained by dialysis against water (molecular weight cut-off = 1 kDa, Carl Roth GmbH, Karlsruhe, Germany) and further purification via a Sephadex S25 column. The dPG-IDCC conjugate was acrylated further as described under section 2.2.1.

2.2.3. Synthesis of dPG-PNIPAm nanogels. The given amounts of NIPAm, dPG-Ac, SDS, and VA-044 were dissolved in Milli-Q water in a flask and stirred at 500 rpm for 15 min under argon. The polymerization process was initiated by putting the mixture into a hot oil bath (50 or 68 °C) and the reaction continued overnight under stirring. The synthesized nanogels were purified by dialysis (dialysis membrane molecular weight cut-off = 50 kDa, Carl Roth GmbH, Karlsruhe, Germany) in water for three days. For the synthesis of the labeled nanogels, 25% of the dPG-Ac was replaced by an IDCC labeled dPG-Ac and the same synthetic procedure was followed but only for 3 h at 50 °C because of a stability issue associated with the dye. When necessary, coumarin 6 was loaded in the labeled nanogels by adding an excess amount of the dye (10 mg mL⁻¹) and stirring it overnight. The excess suspended dye was then filtered out using a 0.45 μm PTFE filter (for smaller nanogels) or a 1.2 μm glass microfiber filter (for larger nanogels). The yield was calculated by dividing the actual mass of the nanogel obtained after lyophilization by the theoretical mass of the compounds used.

2.2.4. NMR spectroscopy. The ¹H NMR spectra of the acrylated dPG and dPG-PNIPAm nanogels were obtained at 25 °C at 400 MHz using a Bruker DRX 400 NMR spectrometer (Bruker ECX 400, Bruker Corporation, Berlin, Germany). Before the measurement, 10–15 mg of the samples were dissolved in deuterated water. All the recorded spectra were analyzed with the MestReNova software. Accordingly, the integral of the signals of the core glycerol proton 3.2–4.6 (m, 5H) with respect to the signals of the 3 alkene protons 5.76–5.90 (m, 1H), 6.02–6.18 (m, 1H) and 6.30–6.48 (m, 1H) was used to determine the degree of modification of the OH groups into acryloyl groups. The integral of the signals at 3.2–4.4 (5H, dPG + 1H, NIPAm) with respect to the signals of the polymeric backbone 2.03 (1H) and 1.47 (2H) and the isopropyl group of NIPAm 1.16 (6H) were used to determine the monomer crosslinker ratio.

2.2.5. Particle size and zeta potential. The particle size, size distribution and zeta potential of the nanogels were measured in Milli-Q water or phosphate buffer at 25 °C using a Zetasizer (Malvern Zetasizer-Nano ZS, Malvern Instruments Limited, Worcestershire, UK) after dilution of the samples and temperature equilibration for 60 s. The intensity size distribution and the PDI values were recorded.

2.2.6. Cloud point temperature (T_cp). T _cp determinations were carried out in three complete heating and cooling cycles using a Cary 100 Bio UV-visible spectrophotometer (Lambda 950 UV/Vis/NIR spectrometer, Perkin Elmer Life and Analytical Sciences, Connecticut, USA), which was equipped with a temperature-controlled six-position sample holder. During the measurement, the temperature of a 1 mg mL⁻¹ nanogel in Milli-Q water was raised from 20 °C to 60 °C at a rate of 1 °C min⁻¹ while measuring the transmittance at 450 nm (1 cm path length, ramp = 1, slit = 2.0, gain = 0). The T_cp was defined as the temperature at 50% transmittance on the normalized transmittance vs. temperature curve.

2.2.7. Transmission electron microscopy (TEM). TEM samples were prepared by applying a droplet (5 μL) of the sample solution (10 mg mL⁻¹ in Milli-Q water) on a hydrophilised (60 s low discharging at 8 W using a BALTEC MED 020 device) carbon-coated copper grid (400 meshes, Quantifoil Micro Tools GmbH, Großlöbichau, Germany) for 60 s. The supernatant fluid was removed by blotting with filter paper. Then a droplet (5 μL) of 1% (w/v) uranyl acetate solution was applied and kept for another 60 s. The excess contrasting material was removed by means of filter paper and the sample was allowed to dry in air. The measurements were carried out using the TEM mode of a Hitachi Scanning Electron Microscope (SU8030, Hitachi High-Technologies Corporation, Tokyo, Japan) (20 kV) at different magnifications.

2.2.8. Atomic force microscopy (AFM). The aqueous solution (1 mg mL⁻¹) of the nanogels was spin-coated on a mica sheet at 90 rps for 5 min. AFM readings were obtained using a MultiMode 8 AFM equipped with a Nanoscope V controller (Veeco Instruments, Santa Barbara, California) in the tapping mode using a long cantilever (NCL-W) fitted with Nano World tips at a resonance frequency of 190 kHz and a force constant of 48 N m⁻¹. The data were analyzed using the NanoScope Analysis 1.3 software and the statistical analysis was performed in a 3 μm × 3 μm image.

2.2.9. Confocal laser scanning microscopy (CLSM). Histological sections of porcine ear skin containing hair follicles, obtained after follicular penetration experiments, were examined under a CLSM (LSM 700, Carl Zeiss, Jena, Germany) using a Plan-Apochromat 5×/0.16 M27 objective. The penetration of coumarin 6 (at an excitation wavelength of 488 nm, laser intensity 2%, pinhole 90, Gain 650, digital offset −12800, digital gain 1.2 and emission wavelengths of 488–630 nm) and the IDCC labeled nanogels (at an excitation wavelength of 639 nm laser intensity 5%, pinhole 90, Gain 750, digital offset −51200, digital gain 1.5 and emission wavelengths of 640–630 nm) was independently but simultaneously tracked in the hair follicles. Skin samples without nanogel application were used as negative controls for configuration of the microscope settings.

2.2.10. Ex vivo follicular penetration study. For the ex vivo follicular penetration study fresh porcine ear skin, which was obtained from a local abattoir and stored in the refrigerator for less than 3 days, was used. The skin was thoroughly washed with distilled water, dried, and the hairs were trimmed using a pair of scissors. To prevent lateral diffusion of the colloidal dispersion, an adhesive solution was put around the application area (2 cm × 3 cm) and dried for 90 min. Then 50 μL of the labeled nanogels, or an equivalent amount of the HEC gel, was uniformly spread over the application area (n = 6), massaged for 2 min using a mini massager (Rehaforum Medical GmbH, Elmshorn, Germany), and incubated at a specific temperature (RT, 35 or 37 °C) for a specified period of time (1 or 4 h). Then cryo spray was applied and biopsies of about 0.6 cm × 0.6 cm were cut out using a scalpel. The biopsies were put in an Eppendorf tube, frozen in liquid nitrogen and stored at −20 °C. Finally, the biopsies were mounted in a frozen tissue freezing medium and 10 nm histological sections containing hair follicles (n = 10) were cut out using a cryostat (Microm HM 560, Microm GmbH, Walldorf, Germany) and observed under a CLSM to record the penetration depth of the dye and the nanogel. Approval to conduct the ex vivo experiments using the porcine ear skin was obtained from the Veterinaeramt Berlin.

2.2.11. Statistical analysis. Unless stated otherwise, all the measurements were conducted in triplicates and the data were presented as the mean ± standard deviation. One-way ANOVA with Tukey's test was used to compare means at the statistical significance level (P) ≤ 0.05.

3. Results and discussion

3.1. Synthesis and characterization of dPG-PNIPAm nanogels

3.1.1. Synthesis of dPG-PNIPAm nanogels. The synthesis of about 100–200 nm thermoresponsive dPG-PNIPAm nanogels was previously reported.^5,16 However, the follicular penetration of nanoparticles is highly dependent on the particle size and nanoparticles in the order of 600–700 nm exhibit optimal follicular penetration.^7,11 Therefore, an attempt was made to synthesize larger dPG-PNIPAm nanogels.

The degree of acrylation and molecular weight of dPG affect the nanogel size and various other nanogel characteristics.^16,20 In our case, better results were obtained when 10 kDa dPG, with 9% of its hydroxyl groups acrylated, was used as a crosslinker and SDS was used as a stabilizer. The use of different initiators, namely potassium persulfate and the water soluble azo initiators VA-086 and VA-044, was considered and larger nanogels were obtained with the azo initiators. Similar results were also reported, where the azo initiator azobisisobutyronitrile gave significantly bigger styrene nanoparticles than potassium persulfate.²⁷ VA-044 was chosen for the synthesis of the intended nanogels because it can be activated at a relatively lower temperature than VA-086, without the need for UV light irradiation.

3.1.2. Nanogel size optimization. There are numerous factors that affect the size and size distribution of nanoparticles synthesized by the precipitation polymerization technique. These include the type and concentration of the surfactant, crosslinker, and initiator used; monomer-to-crosslinker ratio; precursor concentration; and reaction conditions such as the stirrer speed and reaction temperature.²⁸ Besides, the addition of electrolytes and other compounds that provide charges such as 2-aminoethylmethacrylate hydrochloride may also change the resulting size of the nanoparticles.^27,28 Preparation of dPG-PNIPAm-based nanoparticles showed that the degree of acrylation of the dPG also affected the size and size distribution of the nanogels.¹⁶ In our case, based on preliminary investigations, the monomer-to-crosslinker ratio, precursors, surfactants, and initiator concentrations, as well as the reaction temperature were the most significant factors and their effects were systematically assessed.

NG-1 to NG-6 (Table 1) were synthesized to investigate the effects of the crosslinker percentage and initiator concentration on the particle size and T_cp. The particle size increased significantly with decreasing the percentage of the crosslinker (Fig. 1a). This can be attributed to the formation of a more compact and smaller nanogel due to the high degree of crosslinking.¹⁶ However, for further investigations, the crosslinker concentration was maintained at or above 20% as the percentage of the crosslinker affected the T_cp (Fig. 3b) and other properties of the nanogels.

Fig. 1 Effect of the (a) percentage of the crosslinker and (b) initiator concentration on the nanogel size (labels represent the corresponding PDI values).

Table 1 Precursor feed and yield of dPG-PNIPAm nanogels synthesized at 68 °C in 6 mM SDS solution to study the effect of various factors on the nanogel size and other characteristics

Nanogel	NIPAm + dPG-Ac (mg mL^−1)	Wt% of dPG-Ac in feed precursor	VA-044 (wt%)	Yield ± SD (wt%)
NG-1	40	33	1	83.9 (1.3)
NG-2	40	20	1	88.1 (1.9)
NG-3	40	10	1	87.3 (3.3)
NG-4	40	33	0.5	85.8 (1.3)
NG-5	40	33	2	82.9 (2.7)
NG-6	40	33	3	84.5 (2.1)
NG-7	20	20	2	90.3 (2.0)
NG-8	40	20	2	86.1 (1.3)
NG-9	46	20	2	86.4 (2.5)
NG-10	49	20	2	86.5 (2.5)
NG-11	52	20	2	85.1 (3.1)
NG-12	20	20	3	92.1 (1.5)
NG-13	40	20	3	87.4 (2.0)
NG-14	46	20	3	86.3 (1.3)
NG-15	49	20	3	87.6 (1.4)
NG-16	52	20	3	87.1 (1.1)

---

Mostly, increasing the initiator concentration results in increased primary radicals which in turn results in an increased particle number and reduced particle size.²⁹ Surprisingly, in this case, the particle size increased significantly with increasing initiator concentration (Fig. 1b). The results in Fig. 2a and b also showed a similar effect where significantly larger nanogels were obtained at 3% VA-044 than at 2%. Factors like multiple polymer branching due to chain transfer or self-destruction of the primary radicals might contribute to this. However, considering the magnitude of the change in particle size, the more plausible explanation would be an occurrence of the coagulation of primary particles to form bigger nanogels during the polymerization process. That is, the increased initiator concentration may increase the primary radicals and primary particle number, which further enhances the extent of particle coagulation. The same effect was observed when the water soluble initiators KPS and AIBA were used for the synthesis of styrene polymer nanoparticles using the emulsion polymerization technique.²⁷ Thus, further investigations, including the determination of the change in the particle number as a function of monomer conversion, should be conducted for a clear understanding of the process.²⁷

Fig. 2 Effect of the (a) surfactant and (b) precursor concentrations at two levels of the initiator and the reaction temperature on the size and size distribution of dPG-PNIPAm nanogels (labels represent the corresponding PDI values).

NG-8 (2% initiator) and NG-13 (3% initiator) were synthesized at 5 different concentrations of SDS (2, 3, 4, 5 and 6 mM) to investigate the effect of the surfactant concentration on the particle size (Fig. 2a) and T_cp (Fig. 3). The particle size decreased significantly with increasing concentration of the surfactant. This is expected as a higher concentration of the surfactant has a better stabilization effect by forming smaller pockets, in which the insoluble polymer is accommodated.³⁰ However, at a higher concentration of the precursor a lower SDS concentration resulted in significantly higher polydispersity (results not shown), and further synthesis of the nanogels was carried out at 6 mM SDS.

Fig. 3 Effects of the (a) initiator, (b) crosslinker, (c) precursor, and (d) surfactant concentrations on the volume phase transition temperature of dPG-PNIPAm nanogels.

NG-7 to NG-16 were synthesized at 50 and 68 °C to investigate the effect of the precursor concentration and reaction temperature on the particle size. The nanogel size increased significantly with increasing precursor concentration and reaction temperature (Fig. 2b). Above a precursor concentration of 40 mg mL⁻¹, the increase in size was exponential and above 60 mg mL⁻¹, the nanogels aggregated to form a gel at all levels of the initiator concentrations and reaction temperatures considered. In principle, increasing the initiation temperature should also increase the primary radicals and particle number and tends to result in a reduced particle size.²⁹ Thus, the increase in particle size as a function of temperature is also a significant indicator of particle coagulation to form bigger nanogels. This is also supported by a recent finding by Liu et al.²⁷ The nanogels also had acceptable PDI values with acceptable standard deviations (Fig. 2). Therefore, at a high temperature of the synthesis and precursor and initiator concentrations, nanogels of the desired sizes of 600–700 nm, with an acceptable level of polydispersity, were obtained. The yield was also relatively high (82.9–92.1) (Table 1).

3.1.3. Nanogel characterization. T _cp values increased with an increase in the percentage of the crosslinker (Fig. 3). This is attributed to the hydrophilic nature of dPG, which was added at the cost of PNIPAm, the precursor that has a transition temperature of about 32 °C and is responsible for the thermoresponsive behavior of the nanogel.^16,31 The T_cp value also increased with the SDS concentration but significantly decreased with the initiator and precursor concentrations.

Representative T_cp values for small and large nanogels are shown in Fig. 4a. Generally, at a given SDS concentration, factors that had a positive effect on the particle size had a negative effect on T_cp. Consequently, bigger nanogels resulted in lower T_cp and vice versa. The T_cp value of the different nanogels was plotted against their size (Fig. 4b) and the same relationship was observed.

Fig. 4 T _cp values of dPG-PNIPAm nanogels: (a) normalized transmittance vs. temperature curves of a big (NG-15) and a small (NG-12) nanogel, (b) a general trend depicting the relationship between particle size and T_cp.

The TEM images of NG-15 and NG-12 (Fig. 5a and c) were by about 3–5 factors smaller than the hydrodynamic diameters of the nanogels obtained by dynamic light scattering (NG-15: 576.8 ± 52.7 nm; PDI = 0.229 ± 0.065 and NG-12: 128.2 ± 3.5 nm; PDI = 0.157 ± 0.010). Water contributes to a significant proportion of the nanogels’ mass and shrinking of the nanogels in TEM can be attributed to the drying process during sample preparation.^31,32 However, irrespective of the volume contraction, the difference in size between the bigger (Fig. 5a) and smaller (Fig. 5c) nanogels was apparent.

Fig. 5 Different microscopy images of selected nanogels obtained at two different magnifications: (a) TEM and (b) AFM images of NG-15, (c) TEM and (d) AFM images of NG-12.

The AFM images of NG 15 and NG-12 are also shown in Fig. 5b and d, respectively, and the average, minimum and maximum sizes of the nanogels are given in Table 2. Compared to TEM, relatively larger nanogels were obtained with AFM, which can be attributed to the minimal degree of water removal with AFM compared to TEM. Interestingly, the minimum sized nanogels obtained with NG-15, which was prepared at a higher concentration of the initiator, were smaller than the minimum sized nanogels obtained with NG-12 and the results again substantiate that a high degree of agglomeration of the primary particles occurred at a higher concentration of the initiator and feed concentration. In addition, the average nanogel sizes obtained with AFM are still smaller than the hydrodynamic radius obtained by DLS and this can partly be attributed to the differences in the methods and partly to the partial drying of the sample during the preparation for the AFM analysis.

Table 2 The average, minimum and maximum sizes of selected nanogels obtained by AFM measurements

Nanogel	Mean size (nm)	Minimum size (nm)	Maximum size (nm)
NG-12	118.3	68.7	172.3
NG-15	165.6	33.7	564.0

---

Besides, both TEM and AFM images showed that the bigger the particle size the more irregular was the nanogel. It can also be taken as evidence that significant agglomeration of the primary particles occurred during the polymerization process to form bigger nanogels. The low zeta potential of the nanoparticles (Table 3) also substantiates the possibility of agglomeration of the primary nanoparticles.

Table 3 Zeta potentials of selected dPG-PNIPAm nanogels at 25 and 45 °C

Formulation	Reaction temp. (°C)	Zeta potential (mV) in H2O		Zeta potential (mV) in phosphate buffer
		25 °C	45 °C	25 °C	45 °C
NG-15	68	2.69	−3.38	−0.0493	−0.248
NG-15	50	1.65	−4.42	−0.609	−0.995
NG-12	68	−0.661	−9.76	−1.89	−4.45
NG-12	50	−1.28	−9.72	−2.56	−4.03

---

3.2. Ex vivo follicular penetration study

3.2.1. Assumptions and descriptions. The follicular penetration study was conducted using porcine ear skin, which is a better alternative than the human skin for ex vivo follicular penetration studies because excised human skin contracts after surgical removal and the hair follicles remain closed.³³ Besides, different reports indicated that the pig skin is the most suitable model for human skin for ex vivo and in vivo percutaneous absorption and penetration experiments, where the degree of penetration of many compounds through the human and porcine skins was well correlated.^7,34–39

The nanogels were labeled with IDCC (λ_max excitation = 650 nm) prior to the experiment to enable their independent tracking from the loaded dye coumarin 6 (λ_max excitation = 444 nm). Three different sizes of the labeled nanogels were synthesized and characterized (L-76, L-396 and L-508, Table 4) to investigate the effect of the nanogel size on the depth of follicular penetration. L-508 was also incorporated into a 2.5% HEC gel to assess the effect of the formulation of the nanogels into the final dosage form on the follicular penetration of the nanogels.

Table 4 Compositions and characteristics of coumarin 6 loaded IDCC labeled nanogels synthesized at 50 °C for ex vivo follicular penetration studies

No	Nanogel	NIPAM + 20% crosslinker^a (mg mL^−1)	VA-044 (wt%)	SDS (mM)	Size (±SD) (nm)	PDI (±SD)	T cp (±SD) (°C)
a IDCC labeled dPG-Ac : dPG-Ac = 1 : 3.
1	L-76	20	2	6	76.7 (3.4)	0.255 (0.016)	36.0 (0.0)
2	L-396	46	2	6	396.3 (3.7)	0.316 (0.037)	34.3 (0.3)
3	L-508	49	3	6	508.9 (39.2)	0.332 (0.043)	34.0 (0.0)

---

The PDI values of the labeled nanogels were slightly higher than their non-labelled equivalents. This is assumed to occur due to the dye effect on the crosslinker hydrophilicity during the precipitation polymerization process. The TEM image of L-508 (Fig. 6) indicated particle shrinkage due to water loss.

Fig. 6 TEM image of L-508.

Typical CLSM fluorescence emission images of the dye and the nanogel (L-508), which were taken independently but simultaneously, are shown in Fig. 7a and b, respectively. The light transmission image of the histological section was also obtained (Fig. 7c). Fig. 7d represents the superimposed images of the three and shows that the penetration depth of the nanogel (shown by the small arrow drawn perpendicular to the hair follicle) was different from the dye (shown by the line drawn parallel to the hair follicle). Fig. 7i to iv show the change in the fluorescence emission intensities of coumarin 6 and the nanogel as a function of the follicular penetration depth. Accordingly, the effect of various factors on the nanogel penetration depth and dye release was assessed.

Fig. 7 Typical CLSM images of a histological section of a treated porcine ear skin containing a hair follicle: (a) green fluorescence emission images of the dye obtained at an excitation wavelength of 488 nm, (b) red fluorescence emission images of the nanogel obtained at an excitation wavelength of 639 nm, (c) the image of the hair follicle obtained in the transmittance mode, (d) superimposed images of the dye, the drug and the hair follicle; (i–iv) fluorescence emission as a function of the depth of the hair follicle.

3.2.2. Effects of various factors on the follicular penetration of the nanogels and dye release. The effect of the nanogel size, viscosity/consistency of the colloidal dispersion, the temperature at which the penetration and release experiment was carried out (incubation temperature) and the time of incubation at a given temperature (incubation time) on the ex vivo follicular penetration of the nanogels and the release of the loaded dye was assessed (Fig. 8).

Fig. 8 The effect of the nanogel size, viscosity/consistency of the dispersing media, incubation temperature, and incubation time on the follicular penetration of the nanogel and release of the loaded dye. Numbers in nanogel identifiers show the nanogel size, Coll = colloidal dispersion/nanogel, Gel = 2.5% HEC gel of the nanogel, RT = room temperature.

The follicular penetration depth of the nanogels significantly increased with nanogel size. This is in line with the previous reports where optimal penetration of nanoparticles into the hair follicles was attained with particles sized at about 600 nm.¹¹ However, L-76 (76 nm in diameter) failed to penetrate to any appreciable extent. In addition, the penetration depth of L-396 and L-508 was relatively shallow when compared with the previously reported solid nanoparticles of similar sizes.¹¹ These phenomena can be attributed to the significant shrinking of the nanogels due to rapid water loss during application. The TEM and AFM images also confirmed the shrinking of the nanoparticles due to water loss. Therefore, unlike other compact nanoparticles, the penetration depth of nanogels could depend on the water evaporation kinetics and, to compensate for the volume contraction, it might be advantageous to synthesize even bigger nanogels for deeper follicular penetrations.

Apart from the results shown in Fig. 8, representative histological sections showing the effect of incubation temperature and time on nanogel and dye penetration are shown in Fig. 9 (green and red designate the dye and the nanogel and the parallel and perpendicular lines running along the hair follicle designate the depth of penetration of the dye and the nanogel, respectively). As can be seen from the two figures, there was no significant difference in nanogel penetration and dye release at RT and 35 °C after 1 h of incubation and the dye and the nanogel traveled almost the same distance showing the lack of significant dye release. In principle, the T_cp of the polymer was determined to be 34 °C and above this temperature a significant increase in dye release and diffusion deep into in the hair follicle was expected. However, T_cp is a temperature at which only 50% of the polymer responds to the change in temperature in a very dilute colloidal dispersion and the temperature at which all the particles respond is higher (Fig. 4a). Interestingly, unlike nanogel penetration, dye release and penetration increased significantly at 37 °C or when the incubation time was increased to 4 h. At 37 °C, increasing the incubation time to 4 h further increased the dye release and penetration significantly (2× the penetration at 35 °C and 1 h). This is attributed to the thermoresponsive nature of the nanogel. Thus, the temperature deep in the hair follicle is expected to be close to the body temperature of 37 °C, and a significant drug release is expected from the nanogels in vivo.

Fig. 9 Representative CLSM images of histological sections of porcine ear skin containing a hair follicle obtained after the application of the coumarin 6 loaded IDCC labeled nanogels (L-508) at varying incubation temperatures and times: (a) RT; 1 h, (b) 35 °C; 1 h, (c) 37 °C; 1 h, (d) 35 °C; 4 h, and (e) 35 °C; 4 h.

Looking at nanogel penetration from another perspective, above the T_cp, theoretically, the penetration depth of the nanogel into the hair follicle should decrease due to particle shrinking. However, interestingly, nanogel penetration at RT and 35 °C, and even at 37 °C when only incubated for 1 h, was not significantly different (Fig. 8). This is most likely attributed to the rapid penetration of the nanogels into the hair follicles, which need some seconds to few minutes,^14,40 before undergoing any significant phase transition. This might also be attributed to the tendency of the nanogels to aggregate above their transition temperature to form bigger aggregates. Thus, a better understanding of the follicular penetration needs careful investigations of the nanogel penetration, shrinking and aggregation kinetics.

Generally, semisolid formulations are preferred for applications to the skin and one way of preparing nanogels into a semisolid dosage form is by formulating them into gels using gelling agents. Thus, the effect of the incorporation of the gelling agent HEC (2.5%) into the nanogels on their follicular penetration was investigated (Fig. 8), and no significant differences in nanogel penetration or dye release were observed.

The CLSM images also enabled the visualization of the distribution of the labeled nanogel and the loaded dye on the skin surface and showed that the nanogels and the loaded dye did not penetrate through the skin surface. Although CLSM clearly showed that the main penetration pathway for the nanogels is through the hair follicle, it is a semi-quantitative method and it is difficult to exactly quantify the amount of drug that penetrated at different depths of the skin. Thus, this method should be complemented with other sensitive, precise and accurate analytical methods if quantification of the amount of drug at different depths of the skin is required.

4. Conclusions

Various sizes of dPG-NIPAm based thermoresponsive nanogels that are as big as 600–700 nm were synthesized by the precipitation polymerization technique, by controlling the various synthetic conditions. The nanogels exhibited T_cp values of 32–37 °C, which are ideal for skin applications. Ex vivo follicular penetration investigations showed that the depth of nanogel penetration was proportional to their sizes. Temperature dependent dye release from the thermoresponsive nanogels in the hair follicle was also investigated ex vivo for the first time and there was a significant increase in dye release above the T_cp of the nanogels. Interestingly, the formulation of the nanogels into gels did not affect their follicular penetration.

Abbreviations

dPG-Ac	Acrylated dendritic polyglycerol
AFM	Atomic force microscopy
T cp	Cloud point temperature
CLSM	Confocal laser scanning microscope
dPG	Dendritic polyglycerol
HEC	Hydroxyethyl cellulose
NIPAm	N-Isopropylacrylamide
PNIPAm	Poly N-isopropylacrylamide
SDS	Sodium dodecyl sulphate
TEM	Transmission electron microscopy

Acknowledgements

This work has been partly supported by the collaborative research center 1112 (http://www.sfb1112.de), Projects A04 and C05 of the DFG. We greatly acknowledge the financial support provided to Fitsum F. Sahle by the Alexander von Humboldt Foundation. Marcelo Calderón also acknowledges the Bundesministerium für Bildung und Forschung (BMBF) through the NanoMatFutur award (13N12561, Thermonanogele). Julian Bergueiro acknowledges the Dahlem Research Center for a Dahlem International Network PostDocs fellowship. We would also like to gratefully acknowledge Dr Alexa Patzelt for her consultation on some aspects of the work, Heike Richter for her technical support, and Dr Fanny Knorr for proofreading the manuscript.

References

1. A. Rahikkala, V. Aseyev, H. Tenhu, E. I. Kauppinen and J. Raula, Biomacromolecules, 2015, 16, 2750–2756.
2. J.-W. Yoo, N. Giri and C. H. Lee, Int. J. Pharm., 2011, 403, 262–267.
3. J. Shim, H. S. Kang, W. S. Park, S. H. Han, J. Kim and I. S. Chang, J. Controlled Release, 2004, 97, 477–484.
4. M. Schafer-Korting, W. Mehnert and H. C. Korting, Adv. Drug Delivery Rev., 2007, 59, 427–443.
5. M. Witting, M. Molina, K. Obst, R. Plank, K. M. Eckl, H. C. Hennies, M. Calderon, W. Friess and S. Hedtrich, Nanomedicine, 2015, 11, 1179–1187.
6. T. Subongkot, N. Wonglertnirant, P. Songprakhon, T. Rojanarata, P. Opanasopit and T. Ngawhirunpat, Int. J. Pharm., 2013, 441, 151–161.
7. M. Schneider, F. Stracke, S. Hansen and U. F. Schaefer, Derm.-Endocrinol., 2009, 1, 197–206.
8. S. A. Coulman, A. Anstey, C. Gateley, A. Morrissey, P. McLoughlin, C. Allender and J. C. Birchall, Int. J. Pharm., 2009, 366, 190–200.
9. E. Kimura, Y. Kawano, H. Todo, Y. Ikarashi and K. Sugibayashi, Biol. Pharm. Bull., 2012, 35, 1476–1486.
10. W. Rangsimawong, P. Opanasopit, T. Rojanarata and T. Ngawhirunpat, Int. J. Nanomed., 2015, 10, 7413–7423.
11. J. Lademann, F. Knorr, H. Richter, S. Jung, M. C. Meinke, E. Rühl, U. Alexiev, M. Calderon and A. Patzelt, J. Innovative Opt. Health Sci., 2015, 08, 1530004.
12. J. Lademann, N. Otberg, U. Jacobi, R. M. Hoffman and U. Blume-Peytavi, J. Invest. Dermatol. Symp. Proc., 2005, 10, 301–303.
13. A. Vogt, N. Mandt, J. Lademann, H. Schaefer and U. Blume-Peytavi, J. Invest. Dermatol. Symp. Proc., 2005, 10, 252–255.
14. C.-L. Fang, I. A. Aljuffali, Y.-C. Li and J.-Y. Fang, Ther. Delivery, 2014, 5, 991–1006.
15. J. Ramos, A. Imaz, J. Callejas-Fernandez, L. Barbosa-Barros, J. Estelrich, M. Quesada-Perez and J. Forcada, Soft Matter, 2011, 7, 5067–5082.
16. J. C. Cuggino, C. I. Alvarez, M. C. Strumia, P. Welker, K. Licha, D. Steinhilber, R.-C. Mutihac and M. Calderon, Soft Matter, 2011, 7, 11259–11266.
17. J. Bergueiro and M. Calderon, Macromol. Biosci., 2015, 15, 183–199.
18. I. Yildiz and B. S. Yildiz, J. Nanomater., 2015, 2015, 1–12.
19. M. A. Ward and T. K. Georgiou, Polymers, 2011, 3, 1215–1242.
20. M. Asadian-Birjand, J. Bergueiro, F. Rancan, J. C. Cuggino, R. C. Mutihac, K. Achazi, J. Dernedde, U. Blume-Peytayi, A. Vogt and M. Calderon, Polym. Chem., 2015, 6, 5827–5831.
21. F. Rancan, M. Asadian-Birjand, S. Dogan, C. Graf, L. Cuellar, S. Lommatzsch, U. Blume-Peytavi, M. Calderón and A. Vogt, J. Controlled Release, 2016, 228, 159–169.
22. M. Giulbudagian, F. Rancan, A. Klossek, K. Yamamoto, J. Jurisch, V. Colombo-Neto, P. Schrade, S. Bachmann, E. Rühl, U. Blume-Peytavi, A. Vogt and M. Calderón, J. Controlled Release, 2016, 243, 323–332.
23. M. Calderón, M. A. Quadir, S. K. Sharma and R. Haag, Adv. Mater., 2010, 22, 190–218.
24. D. Wilms, S.-E. Stiriba and H. Frey, Acc. Chem. Res., 2010, 43, 129–141.
25. R. Haag, H. Tuerk and S. Mecking, German Patent DE10211664, 2003.
26. F. S. Mehrabadi, O. Hirsch, R. Zeisig, P. Posocco, E. Laurini, S. Pricl, R. Haag, W. Kemmner and M. Calderon, RSC Adv., 2015, 5, 78760–78770.
27. B. Liu, Y. Wang, M. Zhang and H. Zhang, Polymers, 2016, 8, 55.
28. T. Still, K. Chen, A. M. Alsayed, K. B. Aptowicz and A. G. Yodh, J. Colloid Interface Sci., 2013, 405, 96–102.
29. I. Capek and P. Potisk, Eur. Polym. J., 1995, 31, 1269–1277.
30. W. H. Blackburn and L. A. Lyon, Colloid Polym. Sci., 2008, 286, 563–569.
31. L.-W. Xia, R. Xie, X.-J. Ju, W. Wang, Q. Chen and L.-Y. Chu, Nat. Commun., 2013, 4, 2226.
32. K. Zielinska, H. Sun, R. A. Campbell, A. Zarbakhsh and M. Resmini, Nanoscale, 2016, 8, 4951–4960.
33. J. Lademann, A. Patzelt, H. Richter, S. Schanzer, W. Sterry, A. Filbry, K. Bohnsack, F. Rippke and M. Meinke, Eur. J. Pharm. Biopharm., 2009, 72, 600–604.
34. G. A. Simon and H. I. Maibach, Skin Pharmacol. Appl. Skin Physiol., 2000, 13, 229–234.
35. I. Boudry, O. Blanck, C. Cruz, M. Blanck, V. Vallet, A. Bazire, A. Capt, D. Josse and G. Lallement, J. Appl. Toxicol., 2008, 28, 645–657.
36. F. P. Schmook, J. G. Meingassner and A. Billich, Int. J. Pharm., 2001, 215, 51–56.
37. A. M. Barbero and H. F. Frasch, Toxicol In Vitro, 2009, 23, 1–13.
38. E. C. Jung and H. I. Maibach, J. Appl. Toxicol., 2015, 35, 1–10.
39. B. Godin and E. Touitou, Adv. Drug Delivery Rev., 2007, 59, 1152–1161.
40. J. Lademann, F. Knorr, H. Richter, S. Jung, M. C. Meinke, E. Rühl, U. Alexiev, M. Calderon and A. Patzelt, J. Innovative Opt. Health Sci., 2015, 08, 1530004.

---

Footnote

† These authors contributed equally to this work.

---