In Vitro, Ex Vivo, Penetration (EpiDermTM) and in Vivo Dermatokinetics of Ketoconazole Loaded Solid Lipid Nanoparticles for Topical Delivery

The study focused to optimize, evaluate and investigate mechanistic perspective of ketoconazole (KTZ) loaded solid lipid nanoparticles (KTZ-SLNs) for enhanced permeation across rat skin. KTZ-SLNs were evaluated for size, distribution, zeta potential (ZP), percent entrapment eciency (%EE), drug release, morphology, thermal behavior (DSC), compatibility (FTIR) and solid state characterization (X-ray diffraction, XRD). Moreover, ex-vivo permeation and drug deposition into rat skin were conducted using Franz diffusion cell. Mechanistic evaluations were conrmed using confocal laser scanning microscopy and vibrational ATR methods using EpiDerm TM model. An in vivo dermatokinetics study was performed to ensure KTZ access to the dermal region. Accelerated and photostability studies were conducted at different temperatures (0, 30, and 40 °C) for 12 months. The spherical optimized KOF1 showed optimal particle size (291 nm), and high negative ZP (-27.7 mV). Results of FTIR, DSC, and XRD conrmed compatibility of KTZ with excipients, purity of KTZ & dissolved KTZ in lipid matrix, and amorphous nature of KTZ-SLNs. In-vitro release was found to be slow and sustained whereas ex vivo permeation parameters were signicantly high in KTZ-SLNs as compared to drug suspension and marketed product. Drug retention was 10- and -5 fold higher than KTZ-SUS and marketed product, respectively. Pharmacokinetic parameters were improved by SLNs formulation. Confocal raman spectroscopy experiment showed that KTZ-SLNs could penetrate beyond the human stratum corneum into viable epidermis. Fluorescent microscopy conrmed improved penetration of KTZ-SLNs was through human follicular pathway. KTZ-SLNs stable over 12 months under set conditions.


Introduction
Super cial fungal infections are common worldwide and constitute one-fourth of common skin infections caused by dermatophytes. The dermatophytes are responsible to invade the keratinized tissue of stratum corneum (SC) [1]. These have severe psychological, social and nancial consequences for patients [2]. Most fungal infections reside in the lower epidermal and dermal regions and the vellus hair.
These regions are inaccessible to the conventional antifungal formulations due to inadequate dose of the drug reaching to the target site. Most fungal cells have specialised e ux system resulting in frequent drug resistant. Sub-optimal concentration of antifungal agents at the site of infection, poor compliance to therapy, poor delivery modalities, and active antifungal e ux are responsible for frequent recurrence, chronic fungal infections and most seriously to drug resistance. Dermatophytes causing intracellular are critical to treat due to long term therapy and patient incompliance [3]. Topically administered KTZ is clinically recommended to control dermal fungal infections. It is a broad-spectrum antifungal agent to treat skin related candidiasis, tinea, and related dermal infections. Moreover, the drug is potential to control cutaneous super cial (Candida, and Malassezia), localized, and secondary infections such as androgenic alopecia, leishmaniasis, and yeast induced blepharitis [3].
KTZ is lipophilic drug (log P = 4.74) with poor aqueous solubility (0.04 mg/ mL) [4]. The drug is sensitive to light and possessed photo degradation. KTZ is reported to cause several undesirable effects [5] (nausea, vomiting, gastrointestinal disturbance and hepatotoxicity) after oral administration.
Formulation method to load ketoconazole in solid lipid nanoparticles (KTZ-SLNs) Preliminary study was carried out to select excipients and their levels (low and high). Based on benchtop stable product, the higher and the lower levels of factors were decided to feed as input parameters in experimental design software (Design Expert). The lipid organic phase was composed of CATO, PEG-600, and KTZ ( xed amount) which was heated to melt at 75°C. Similarly, the aqueous phase contains tween-80 and P90G previously set at the same temperature. The aqueous phase was stirred at high speed (1000 rpm) using stirrer (WiseTis, HG-15D, Daihan, Korea). The hot organic phase was slowly added to the aqueous phase under constant stirring to result in a coarse emulsion. The prepared coarse emulsion was passed through a high pressure homogenizer (HPH) (EmulsiFlex-C3, Avestin, Canada), at 1000 bar pressure for 7 cycles. The formed o/w emulsion was cooled to room temperature to achieve KTZ loaded SLNs (2% w/v). Thus, several batches of formulations were formulated as per dictated in experimental design (central composite design). In case of uorescein sodium dye probed KTZ-SLNs, the same procedure was adopted and except dye was dissolved in aqueous phase.

Optimization process
Experimental protocol was designed to evaluate the critical factors and their signi cant levels to get the most robust formulation with optimal content of lipid (CATO) and tween 80 (surfactant). A central composite design (CCD) with α=1.414 was run in the Design Expert (version 8.0.1 Stat-Ease Inc. USA) [7].
CATO (X 1 ) and tween 80 (X 2 ) were selected as independent variables (factors). Similarly, mean particle size (Y 1 ), %EE (Y 2 ) and total drug content (Y 3 ) were responses (dependent variables). In optimization process, total 13 runs were experimented at 5 levels (-α, -1, 0, +1, +α). A general polynomial mathematical quadratic equation was generated to quantify and establish a correlation between the independent (X) and dependent variables (Y): Where Y is dependent variable with two coe cients (β 1 & β 2 ) of factors (X 1 & X 2 ). β 0 is an intercept. β 3 is a coe cient of interaction between factors X 1 and X 2 , whereas β 4 and β 5 are the coe cients of quadratic terms "X 1 " and "X 2 ", respectively. Positive and negative signs indicate synergistic and antagonistic effect of factors on the response, respectively. ANOVA (analysis of variance) provides parameters (F, p, and r 2 values) to validate the model applied for optimization process using the experimental design.
Characterization of KTZ loaded suggested SLNs formulations Measurement of particle size, polydispersity index, and zeta potential Particle size, size distribution and surface charge are critical factors to control in-vitro and in-vivo performance of product. Particle size and PDI were measured using photon correlation spectroscopy (PCS) technique which is based on the principle of light diffraction phenomenon. The sample was previously diluted with water (50 fold) for analysis (Beckman Coulter, Delsa™ Nano C, USA). Zeta potential of KTZ-SLNs dispersion was measured without sample dilution (Beckman Coulter, Delsa™ Nano C zetasizer, USA) at 25ºC and the electric eld strength of 23.2 V/cm. Experiments were replicated for mean and standard deviation (n = 3). Percent total drug content (%TDC) and entrapment e ciency (%EE) KTZ-SLNs formulations contain 2%w/v of KTZ. Formulation (1 mL) was dissolved in chloroform: methanol mixture (2:1). The mixture of organic solvents were able to dissolve and disrupt solid lipid of tailored KTZ-SLNs. The mixture was ltered and the content of KTX was quanti ed using validated HPLC method. % EE was determined by dialyzing KTZ-SLN dispersion (1 mL) in a dialysis membrane (14K Da MW cut-off) immersed in 50 mL ethanol and stirred magnetically. After 1h KTZ-SLNs were removed from the bag, disrupted with suitable quantity of chloroform: methanol mixture (2:1) and amount of drug was determined by HPLC. The dialysate was decided based on the assumption that the entire quantity of unentrapped can dissolve in a suitable quantity (50 mL) in an appropriate time (≥1h) to accurately determine amount of unentrapped drug.

Desirability function
The desirability parameter was used to identify and evaluate the optimized formulation by experimental design. Mathematically, this is a numerical function parameter to identify possible interaction between factors. Moreover, it depends upon the set conditions of optimization process such as goal and importance given to each dependent and independent variables. The value of desirability function varies from zero to unity. Zero indicates the model is not t and out of optimization whereas the value approaching to unity indicate the best t of the model applied for optimization. The signi cant terms (p < 0.05) were chosen for nal equations. The model was considered to be the best t when the actual correlation coe cient (r 2 ) value was close to the adjusted correlation coe cient (adjusted r 2 ). Selected formulations of KTZ-SLNs were prepared from the design space and used as checkpoints to assess the prognostic behaviour of the developed mathematical model.

Preparation of ketoconazole suspension (KTZ-SUS)
KTZ suspension (KTZ-SUS) was prepared by method described before with slight modi cation [8]. An accurately weighed amount of KTZ was dispersed in water containing 1% w/v of tween 80 as surfactant and sodium salt of carboxymethyl cellulose (0.1%w/v) as suspending agent. The drug was rigorously stirred for 60 min in the aqueous phase to obtain a stable suspension with optimal consistency. Final strength of the suspension was equivalent to commercial product (2% w/v). This product was used in the further studies as control.
Thermal behaviour of the formulations The thermal behaviours (fusion temperature and fusion enthalpy) of pure and formulations were assessed using a differential scanning calorimeter (DSC). A weighed amount (2 mg) of the samples (Lipid, KTZ, KTZ-SLNs, Blank SLNs) was placed in an aluminium pan and heated at a xed heating rate (10 º C/min) till 300°C using DSC (821e Mettler Toledo, Switzerland). The generated thermograms were analysed, and marked for the values of any signi cant shift or disappearance/appearance of new peaks. The calorimeter was calibrated by pure Indium (melting point) for nitrogen ow and heating rate. Nitrogen gas was used at a purging rate of 50 mL /min. Compatibility study using Fourier Transform Infra-red (FT-IR) To negate any chemical interaction of the drug with explored excipients, the sample alone (KTZ) and formulations (KTZ-SLNs, and placebo SLNs) were subjected for FT-IR analysis. The FT-IR spectrometer (Agilent Technologies 630 Cary) was run for the sample using pellet method. A small amount of the sample was physically mixed with KBr followed by pellet formation. The pellet was processed for characteristic peaks using Micro Lab software. The samples were scanned over the range of 4000-400 cm -1 .
Solid state behaviour using powder X-ray diffraction (XRD) method The prepared SLNs formulations were solid in nature and considered for improved solubility of the drug in solid matrix. In general, crystalline materials exhibit characteristic peaks in XRD graph. Therefore, it was required to assess solid state behaviour of the developed formulation. This was con rmed by analysing the nature of formulated nanoparticles using XRD (XPERT-PRO, PANalytical, Netherlands). KTZ-SLNs and blank SLNs dispersions were lyophilized prior for the analysis. The test sample was exposed to CuKα radiation (45 kV, 40 mA) with scanning angle ranged between 5° and 50°. The values of 2θ and scanning step time were 0.017° and 25s, respectively. Pure drug, lyophilized KTS-SLNs, blank SLNs (without drug) were analysed.

Surface morphology analysis
Surface morphology of prepared solid lipid nanoparticles (KTZ-SLNs) was examined by high resolution transmission electron microscopy (HR-TEM) and eld emission scanning electron microscopy (FE-SEM). Prior to observation under HR-TEM and FE-SEM, KTZ-SLNs were diluted (50 X) with distilled water. The procedure for FE-SEM observations including placing the KTZ-SLNs dispersion on Nucleopore Track-Etch membrane and drying at room temperature. Dried membrane was attached to the silicon wafer using double sided carbon tape followed by sputter coating with gold under FE-SEM (FE-SEM SU8000, Hitachi, Japan). For HR-TEM KTZ-SLNs was stained (0.2% w/v of phosphotungstic acid) during 5 min in phosphate buffer at pH 6.8. Then, the excess phosphotungstic acid was removed using a lter paper. The stained sample of KTZ-SLNs was spread over carbon coated copper grid and was observed under HR-TEM (H-7500, Hitachi, Japan) at a voltage of 200 kV, for morphology (shape and size).

Release behaviour and mechanism
In-vitro release pattern of optimized formulation was studied using a dialysis membrane as per reported method [9][10]. A xed volume (1 mL) of KTZ-SLNs and KTZ-SUS containing 20 mg of KTZ was loaded in the dialysis membrane (molecular weight cut-off of 12KDa). The dialysis membrane was soaked in water for 12 h before experiment. The dialysis membrane containing sample was suspended in a release medium (phosphate buffer solution, pH 7.4). Sink condition was maintained using dimethyl sulfoxide (DMSO). Sampling (2 mL) was carried out at various time points (1,2,4,8,12,16,24,48, and 72 h). The withdrawn volume was replaced with fresh release medium (equal volume) at each time points. The sample withdrawn was ltered and analysed using validated HPLC method at λ max of 210 nm. Analysis was replicated for mean and standard deviation (n=6). Finally, various mathematical models were applied to investigate release mechanism (zero order, rst order release, Higuchi model and Korsmeyer-Peppas model).
Drug permeation and deposition studies: Ex vivo performance across rat skin Permeation potential and drug deposition were carried out using Franz diffusion cells as per reported method [11]. The optimized SLNs formulation was compared against drug suspension and marketed product. For this, rat skin (abdominal) was made free of hairs using digit trimmer without making any surgical cuts or injury. The excise and trimmed skin was placed between two chambers of Franz diffusion cell such that the upper layer faces the formulation and inner layer towards the receptor medium (PBS, pH 7.4). The receptor chamber was lled of release medium (30 mL) and set at 32±1°C under constant stirring using te on coated magnetic bead [12]. The test sample (0.5 mL containing 10 mg of KTZ) was placed over exposed skin (available surface area of 2.1 cm 2 ). Three samples (KTZ-SLNs, KTZ-SUS and KTZ-MKT) were studied separately under similar experimental conditions. The donor chamber was properly covered with para n lm to avoid loss of solvent or dryness of the sample. The sampling was performed at different time points (0.5, 1, 2, 4, 6, 8, 12, 24, 48 and 72 h) followed by replacing equivalent volume of withdrawn sample with fresh release medium. Notably, DMSO (5%) was added to the receptor medium to maintain sink condition. The withdrawn sample was ltered using membrane lter (0.2 µm). The permeated amount of the drug across the skin was estimated using HPLC. Several permeation parameters (cumulative amount of drug permeation, permeation ux, and enhancement ratio) were calculated [13]. After completion of permeation study, the skin samples were removed and washed with running water for drug deposition (retention) study.

Skin retention studies
This study is an extension of the skin permeation study. After completion of the permeation study, the skin was washed with running water to remove adhered treated sample. Then, the skin was sliced into small pieces using surgical scissor and placed in a solution containing methanol and chloroform (2:1 ratio) [14]. The drug deposited or retained in the skin was extracted over 12 h under constant stirring at room temperature. Then, they were ltered and the ltrate was centrifuged (9000 rpm) for 15 min to get the supernatant. The obtained supernatant was analysed for the permeated amount of the drug using HPLC. The study was repeated for mean and standard values.

Dermatokinetics: An in-vivo study
The animal study was carried out in Wistar albino rats weighing about 250-300 g of both sexes. They were get issued from Institutional animal house UIPS (University Institute of Pharmaceutical Sciences) (Approved as regd. No. 45/GO/ReBiBt/S/99/CPCSEA). All of the animals were housed in conditioned room with free access of food and water as per guideline. Animals were randomly selected and grouped (n=6 per group) as per treatment schedule. The body surface of rats was properly inspected for any possible injury and abnormality. The dorsal surface was used to locate a site of application by making three areas (3 cm 2 ) free of hairs. Each group received all three formulations at labelled location on the dorsal site. After 24 h of shaving, formulations (KTZ-SLNs, KTZ-SUS and KTZ-MKT), were applied with equivalent concentration and dose strength. Rat was ethically sacri ced at varied time points (2, 6, 12, 24, 48 and 72 h) for dermatokinetic study. Three skin samples were excised from the applied site and washed water to remove adhered content. Then, the skin samples were sliced into small pieces to extract the drug content by dispersing in a mixture of chloroform and methanol (2:1). The mixture was homogenized after 8 h and ltered. The ltrate was centrifuged to get a supernatant. The supernatant was used to estimate the extracted amount of the drug by HPLC method. The data obtained was tted into one compartment open model. For dermatokinetics pro le, the drug concentration versus time pro le was estimated presented as a graph using a PK solver (version 1.1). Several dermatokinetics parameters such as area under the curve (AUC 0 -72 and AUC 0-∞ ), the maximum drug concentration reached in the skin layer (C max ), the time required to attain C max as T max were assessed.

Fluorescence microscopy study on human dermatome skin (EpiDerm TM )
This was conducted to visualize the permeated KTZ tailored in SLNs using EpiDerm TM as a skin model (MatTek Corporation, Ashland, USA). Fluorescein probed SLNs (F-SLNs) was prepared as per method discussed before. Approximately 30 μL of 2 % aqueous solution of uorescein (aqueous solution) was taken as control and topically applied to the surface of EpiDerm TM . Fluorescein was excited at 470 nm and the uorescent emission was detected at 515 nm. Several representative images of the treated skin were visualized for mechanistic evaluation using uorescent microscopy (IX71 Olympus Inverted Microscope, Olympus, Tokyo, Japan) at 2 h and 24 h with a 10X magni cation.

Vibrational spectroscopic imaging techniques in human skin
Skin treatment procedure Flash-frozen human skin with thickness 4cm 2 (T-SKN-FF2CM) purchased from licensed supplier (ZenBio Inc, USA) was used for this study. All of the skin samples used in this study were from the same donor. 2.5 cm x 2.5 cm piece of skin was cut and cleaned. Formulations (KTZ-SLNs and KTZ-SUS) were applied topically on the skin surface in excess. Product was massaged on the skin using a glass rod and allowed to sit for 5 min. Skin was placed on a Franz diffusion cell for 3 and 24 h at 32˚C. After 3 and 24 h, the excess product on the skin surface was gently blotted with a wet kimwipe. To evaluate product penetration inside the skin, sample preparations were used. Transverse skins sections (8 µm) were obtained using cryo-microtome and scanned by ATR-FTIR imaging to visualize product penetration inside the different skin layers. Skin cross-sections (8µm) were cut using a cryostat. These cross-sections were scanned by ATR-FTIR imaging to evaluate product penetration inside the epidermis. ATR-FTIR images of the cross sections were recorded with a Spotlight 400 System (Perkin Elmer Instruments, Shelton, Conn., USA), consisting of a FTIR spectrometer with a mercury-cadmium-telluride (MCT) focal plane array detector. Images were collected in re ective mode at a spectral resolution of 4 cm -1 and 4 scans accumulations in the mid-infrared (MIR) region between 4000 and 750 cm -1 with a spatial resolution of 6.25 x 6.25 µm at room temperature (24 º C). The ATR imaging accessory used a germanium crystal placed directly in contact with the skin samples. All the data were processed (baseline correction, generation of spectroscopic parameters) using GRAMS/AI (Thermo Fisher Scienti c) or ISys software from Spectral Dimensions (Olney, MD).

Confocal Raman spectroscopy imaging
Skin was also scanned by confocal Raman spectroscopy to evaluate product penetration inside the stratum corneum and beyond in the epidermis. Human skin was treated for 24 h at 34 • C. After incubation, skin was placed in a home-built brass cell. Confocal Raman images were acquired with a WITec Alpha-3000R plus confocal Raman microscope (UIm, Germany) equipped with a 532nm laser. XZ images were taken for each sample. The XZ depth image was typically 50x30µm 2 covering SC and upper viable epidermis (VE) region with 4 µm steps and a 20 second exposure time.
Stability studies Photostability It was conducted for KTZ-SLNs and KTZ-SUS as per ICH guidelines Q1B [15]. The freshly prepared samples were packed in amber coloured clear glass vial, labelled and recorded for further process. Each batch was exposed to illumination light of 1.2 million lux h and an integrated near UV energy (200 watt h/m 2 ) for 10 days in a photostability chamber (Binder Gmbh, Germany).

Results And Discussion
Preparation of formulation and pre-optimization Initially, several trial formulations were prepared using the investigated lipid (CATO as lipid), surfactant (tween 80), co-surfactant (PEG600) and stabilizer (P90G) as per Taguchi design (Table 1). Moreover, the design was used to select factors and levels (Table 1). Several trial formulations were prepared by varying run cycles, speed and time for homogenization using HPH as shown in Table 1. The formulation exhibiting benchtop stability for overnight was selected for selecting lower and higher levels of each factor for Design Expert (central composite design) ( Table 2). We selected X 1 (CATO as lipid) and X 2 (tween 80) as independent factors at three levels as shown in Table 2. The set responses were Y 1 , Y 2 , and Y 3 for particle size (nm), %EE, and the drug content (%), respectively. This tool is reliable to obtain the most robust formulation with optimum content of the solid lipid and surfactant under set desired goals. Moreover, the software was used to identify the potential factors affecting dependent variables (Y 1 -Y 3 ) and possible interaction between factors (X 1 and X 2 ). The central composite design (CCD) suggested thirteen formulations under given set of constraints and goal.
The suitability of the model was assessed by analysis of variance (ANOVA). The statistical parameters (p, F, and r 2 values) were carefully examined for the model to be t. All responses (Y 1 -Y 3 ) followed quadratic model and the generated polynomial equations are presented in Table 2. The negative and positive signs associated with each term represent antagonistic and synergistic in uence of individual factor on the investigated responses, respectively.
Particle size (Y 1 ) of SLNs dispersion is a signi cant parameter for e cient in-vitro and in-vivo experiments. The nanoscaled safeguard circumvents direct interaction of KTZ with e ux proteins thus ensuring its entry into the fungal cells. Furthermore, lower particle size reduces the tendency of coalescence which leads to increase in stability and shelf life of KTZ-SLNs formulation. The polynomial equation for Y 1 is given in Table 1 where negative signs of coe cients associated with both factors X 1 and X 2 indicate that these factors need to be reduced to get desired response. The 3-and 2-dimensional contour plots for Y 1 -Y 3 are illustrated in Fig. 1A-B. The result showed that Y 1 increases with increase in X 1 (CATO) which may be due to higher content of lipid (Fig. 1A). Therefore, it should be reduced to an optimal level. In contrast, Y 1 was found to be decreased with decrease in tween 80 content rst then increased on further reduction in X 2 . Thus, both X1 and X2 need to be optimized to get the most robust SLNs. The lower value of p (0.0006) and high F (18.97) value con rmed the best t of the model adopted for analysis of Y 1 . Moreover, the adjusted correlation coe cient (r 2 ) was close to the observed value which suggested good t of the model. The optimum particle size (minimum) was due to relatively higher concentration of surfactant (2.0 g) and optimal lipid content (1.4 g). However, with further increase concentrations of both the factors, there was increase in particle size due to micelles formation. Hence, it may conclude that to obtain optimized formulation, the optimum levels of X 1 and X 2 have to be required.
For Y 2 , the positive signs of X 1 and X 2 associated with rst and second terms of the quadratic equation (Table 2) suggested that the concentrations of X 1 and X 2 should be at high level for the optimized product. Thus, the effect of the concentration of lipid and tween 80 is directly proportional to Y 2 . These two were found to be considerable factors to use in optimal concentration. The result is exhibited in in Compritol®888 ATO is a complex lipid composed of mixtures of mono, di, and triglycerides which form less perfect crystals, and accord space to accommodate drug molecules [17]. Furthermore, Y2 was found to be increased with increasing content of twee 80 (X 2 ). Statistical analysis suggested that the model was the best t for this response as evidenced with high value of F (483.32), low p value (0.0001) and the closeness of r 2 between adjusted (0.9542) and predicted (r 2 = 0.9266) values. The generated quadratic equation is presented in Table 2 with associated coe cients in each terms. As Y 3 , it is required to assess the totoal content of the drug present in the nal formualtion. The total drug content is signi cant parameter for assessing stability and shelf-life of the formulation. It was expetced that the drug content (Y 3 ) shopuld increase with increase in X 1 and X 2 due to drug solubility in lipid and surfactant based improved emulsi cation. The drug content was the highest when both the factors (X 1 and X 2 ) were at axial point (X 1 :+1 and X 2 : +1) (Fig. 1E-F). The mathematical relationship of Y 3 to factors is represented by the quadratic equation obtained and presented in Table 2. When the lipid and tween 80 concentrations were held at central point (X 1 = 2.0 g and X 2 = 1.4 g) total drug content of KTZ loaded in SLNs (KTZ-SLN4) as in Table 2 had shown a maximal value of 96.2%. The model was signi cant (p < 0.00087) and t as evidenced with the correlation coe cient (regular r 2 = 0.95) and adjusted r 2 (0.90).

Optimization and validation parameter
Based on statistical parameters for the responses Y 1 -Y 3 , the most robust formulation was obtained with optimum content of X 1 and higher content of X 2 to achieve the set goal. The predicted and observed values were closely related as evidenced with correlation coe cient value (r 2 0.97) all explored responses (Y 1 -Y 3 ) ( gure 2A-C). There were no observed interaction for Y 1 and Y 2 ( gure 2D-E). However, the response Y3 exhibited a slight interaction between factors ( gure 2F). The value of overall desirability function was obtained as 0.907 close to unity. Thus, the used model was the best t under given set of experimental conditions and importance to each factor and response. The most optimized formulation obtained was KOF1 with the highest desirability function parameter as compared to other suggested formulations ( gure 3).

Post-optimization studies: Evaluation parameters of KOF1
Particle size, PDI and zeta potential The nally optimized formulation KOF1 was comprised of X 1 (2.0 g) and X 2 (1.4 g) with maximum desirability value. The results of particle size (Y 1 ), PDI, and zeta potential were found as 293 ± 6 nm, The lipid is considered to be composed of a mixture of metastable polymorphic form when characteristic peak is obtained at 71.1°C [18]. The results of thermal enthalpy were 269.5 J/g and 486.9 J/g for the drug and the lipid (CATO), respectively. KOF1 showed an endothermic shift to 89.39 °C and a signi cantly lowered heat ow of 92.26 J/g, which indicating incorporation of the drug into the lipid matrix ( Figure   4A). Lower enthalpy for KOF1 indicates solubilized ketoconazole in the lipid matrix having more imperfections in the crystal lattice which can accommodate more drug content in their crystal lattice [19].

Fourier transform infrared spectroscopy (FT-IR)
It was prerequisite to assess compatibility of the drug with excipients used in the study. . These suggested the crystalline nature of the drug and were found to be complying with previous report [23]. Blank formulation revealed amorphous nature due to blend of lipid with surfactant and stabilizer. The lack of characteristic peaks of the drug in KOF1 may be due to solubilized form of KTZ in lipid matrix amalgamated completely or molecularly dispersed state of KTZ or amorphous nature [24] . Moreover, this suggested least unentrapped drug outside the lipid core.
Surface morphology analysis FE-SEM and HR-TEM are two advanced and sophisticated technology to visualize morphological behavior of nanomedicine or nanocarrier at varied resolution and magni cation. The shape and size of the particle are important aspect to assess for several con rmation. The representative images of FE-SEM and HR-TEM are presented in gure 5A-B. The high magni cation (10000X) HR-TEM image illustrated that the lipid core was enclosed with a hydrophilic rm layer of the surfactant. Acquisition of perfectly spherical shape was attributed to the mixture of surfactants mediated effective covering around the SLNs surface during homogenization. The size of SLNs observed by HR-TEM was a relatively lower than the value obtained in DLS. This variation is associated with several factors such as instrumental error of TEM (relative adsorption of smaller size particle by the grid) and the collapsed SLNs after water evaporation during drying process [25]. It is apparently observed that the optimized KTZ-SLNs (KOF1) formulation was homogeneously dispersed, discrete and spherical. In general, spherical particles are relatively stable and suitable for improved permeation across skin. Moreover, spherical particles are considered to sustain their shape upon storage while disc shaped or discoid particles tend to aggregate which lead to gelation upon storage [26] [27].
In-vitro drug release The drug is highly crystalline in nature and poorly soluble in water (0.04 mg/ mL) or buffer. In-vitro release behaviour of the drug suspension and KOF1 (KTZ-SLNs) was determined PBS (pH 7.4) containing 5% of DMSO (maintained sink condition). It observed that the drug was poorly released (38%) across the dialysis membrane from the drug suspension over a period of 2 h and then no release was observed.
There was no release after 2 h due to poor solubility of the drug. Moreover, the optimized formulation exhibited an initial burst release within 2 h due to free drug and subsequent extended release (99.84%) up to 72 h ( gure 6A). KTZ release from SLNs was exponential in rst 2 h followed by sustained release over 72 h which may be attributed to solid lipid matrix amalgamated with the drug. However, 35% of the total loaded drug was released within 2 h and these values closely correspond to the free/unentrapped drug present in the KTZ-SLN formulation. Free/unentrapped KTZ was released from the SLNs at a rate slower than the free drug associated with SLNs. Different kinds of kinetic release models were applied to understand the release mechanism from the lipid matrix. In rst 12 h, the release mechanism followed non-Fickian as evidenced with low 'n' value (0.82) (Korsmeyer-Peppas model) and the applied model was the best t (r 2 = 0.9948). From 24 h to 72 h, zero order release model was the best t model (r 2 =0.9961) for KTZ-SLNs suggesting slow and controlled release from the lipid matrix. This diffusion controlled release was predicted to occur from porous matrices [28]. The release of drug from a matrix can be predicted based on the various drug release models.

Ex-vivo skin permeation
The layers for sustained and extended release [31] . However, these nanoparticles may opt follicular route of the drug permeation to serve as depot reservoir [32]. Vibrational spectroscopy imaging of human skin ATR-FTIR imaging on skin cross section Human skin cross-sections were recorded by ATR-FTIR imaging spectroscopy. Figure 8 shows visible images of a skin cross-section and the associated hyperspectral image. The method allowed to investigate and visualize the penetration of KTZ of KTZ-SLNs and other SLN components in the integral epidermis. The FTIR Imaging System recorded hyperspectral images which provide maps showing the colocalization of speci c molecular components (KTZ or SLN components). These images were generated with false colours where the red represent highest values and the blue the lowest values for each parameter investigated. The ratio between 816 cm −1 (C-Cl) to Amide I band area was used to follow the penetration of KTZ while the ratio between 945 cm -1 to Amide I band area was used to follow the penetration of SLN components. Indeed, the band around 816 cm -1 is speci c of the KTZ while the band around 945cm -1 is speci c of the SLN components ( Figure 9).
The SLNs penetration will allow the release of KTZ inside the human skin samples when they will exhibit polymorphosize from β to β ' , a more stable crystalline form. Peak for both KTZ and SLN components is observed deep inside the epidermis (Figure 9). The ATR-FTIR images in the Figure 10 show clearly that SLNs penetrated inside the human skin, the maximum penetration was reached after 24h of treatment. After 24h, the SLNs had penetrated up to the deepest part of the epidermis.
Confocal Raman spectroscopy imaging Page 20/33 The FTIR data were con rmed by Confocal Raman spectroscopy. The Confocal Raman images in the gure 11 clearly show that SLNs penetrated the stratum corneum in the rst 3h and reached the viable epidermis after 24h. There was a nice co-localization of the KTZ

Conclusion
Ketoconazole is a potential established antifungal drug with limited aqueous solubility and poor topical e cacy in conventional dosage form. Commercial cream available is also challenged for limited clinical e cacy on topical application and oral delivery to control resistant and recurrence cases. In literature, various topical and transdermal formulations have been reported for improved e cacy based on overestimated data of in vitro ndings. Therefore, we addressed solid lipid nanoparticle with optimum level of compritol and tween 80 to achieve desired size, % EE and %TDC. Ex vivo permeation and drug deposition ndings were further supported with uorescence microscopy study using human cadaver skin model after topical application. It was important to investigate and assess in human skin model (cadaver) for mechanistic and tangible degree of permeation across the crystalline barrier (stratum corneum). Moreover, in vivo dermatokinetics data assured that the approach was capable to access the drug to the dermal region in substantial concentration with limited transport to systemic circulation. Vibrational imaging spectroscopies techniques used as proof of concept as KTZ-SLNs were con rmed to penetrate up to the viable epidermis human skin. Finally, long term stability and photostability data indicated protective bene ts of compritol used in the formulation. Thus, the compritol based SNLs was promising carrier for improved permeation, drug deposition, penetration, dermatokinetics parameters, and increased stability over long period.

Declarations
Ethics approval and consent to participate All animal experiments are conducted after obtaining ethical approval from the Institutional Animal Ethics Committee of Panjab University, Chandigarh (PU/45/99/CPCSEA/IAEC/2018/150)

Consent for publication
All the authors agreed with the content and gave explicit consent to submit the work.

Availability of data and materials
The data generated during the study is provided in the manuscript and is available from corresponding authors on reasonable request.

Funding
There is no funds given by any funding agency    Typical FTIR spectra recorded with 2% KTZ-SLNs. FTIR bands speci c of SLN contribution are represented by the green rectangles while the bands speci c of the KTZ are shown by the red rectangles in the spectra.