RETRACTED ARTICLE: PNIPAM nanoparticles for targeted and enhanced nose-to-brain delivery of curcuminoids: UPLC/ESI-Q-ToF-MS/MS-based pharmacokinetics and pharmacodynamic evaluation in cerebral ischemia model

Abstract

Stroke is a one of the leading causes of disease and deaths worldwide, which causes irreversible deterioration of the central nervous system. Curcuminoids are reported to have a potential role in the amelioration of cerebral ischemia but they exhibit low serum and tissue levels due to low solubility and poor absorption. Curcumin (CUR), demethoxycurcumin (DMC) and bisdemethoxycurcumin (BDMC)-loaded PNIPAM nanoparticles (NPs) were prepared by free radical polymerization and characterized for particles size, entrapment efficiency, zeta potential, in vitro release and ex vivo permeation study. Optimized CUR, DMC and BDMC-loaded NPs had the mean size of 92.46 ± 2.8, 91.23 ± 4.2 and 94.28 ± 1.91 nm; zeta potential of −16.2 ± 1.42, −15.6 ± 1.33 and −16.6 ± 1.21 mV; loading capacity of 39.31 ± 3.7, 38.91 ± 3.6 and 40.61 ± 3.6% and entrapment efficiency of 84.63 ± 4.2, 84.71 ± 3.99 and 85.73 ± 4.31%, respectively. Ultra-performance liquid chromatography/electrospray ionization quadrupole time-of-flight mass spectroscopy based bioanalytical method was developed and validated for pharmacokinetics, biodistribution, brain-targeting efficiency and brain drug-targeting potential studies post-intranasal (i.n.) administration which showed enhanced bioavailability of curcuminoids in brain as compared to intravenous administration. Improved neurobehavioural activity (locomotor and grip strength) and reduced cytokines levels (TNF-α and IL-1β) was observed in middle cerebral artery occlusion induced cerebral ischemic rats after i.n. administration of curcuminoids NPs. Finally, the toxicity study was performed which revealed safe nature of developed NPs.

Keywords:

• Bisdemethoxycurcumin
• brain-targeting efficiency
• brain drug-targeting potential
• curcumin
• demethoxycurcumin
• intranasal

View retraction statement:

Statement of Retraction

INTRODUCTION

Stroke is a global health concern with astronomical financial repercussions on health systems worldwide (Allen & Bayraktutan, 2009; Ashafaq et al., 2012). Cerebral ischemia typically causes irreparable wearying of central nervous system (CNS) (Kim et al., 2007). Inception of cerebral ischemia leads to activation of inflammatory processes that trigger the acceleration of cerebral damage, morbidity and mortality (Berner et al., 2005).

Oxidative stress has always been associated with ischemia–reperfusion injury consequently producing surplus amounts of reactive oxygen species (ROS) (Raza et al., 2011). ROS although play a role in normal physiology, is also found to be initiator of various disease processes. Various studies relate the production of ROS and subsequent oxidative damage to the pathogenesis of ischemia–reperfusion (Sun et al., 2009). Brain tissues are mainly vulnerable to oxidative damage, so pharmacological modification of oxidative damage can be promising for stroke therapy.

A number of antioxidants drugs (thioperamide, ropinirole and thymoquinone) are reported to reduce ROS-mediated reactions and rescue the neurons from reperfusion-induced neuronal damage in animal models (Iida et al., 1999; Badary et al., 2003; Akhtar et al., 2008). Recent in vitro and in vivo studies reports the use of curcumin (CUR) (chiefly found in Curcuma longa) in cerebral ischemia (Thiyagarajan et al., 2004a,b; Jayaprakasha et al., 2006). Other curcuminoids i.e. demethoxycurcumin (DMC) and bisdemethoxycurcumin (BDMC) are also reported to have potential role in the amelioration of cerebral ischemia (Sandur et al., 2007; Somparn et al., 2007; Anand et al., 2010). Curcuminoids show low serum and tissue levels due to its low aqueous solubility, poor absorption, extensive metabolism and rapid elimination (Yang et al., 2007; Anand et al., 2010). Therefore, therapeutic efficacy of curcuminoids is restricted to its short systemic retention in circulation and in tissues. Intranasal (i.n.) administration of curcuminoids can be very effective in this regard through which blood–brain barrier can be bypassed and greater degree of drug concentration can be attained in the cortex, caudate putamen and hippocampus (Hanson et al., 2008).

Poly-N-isopropylacrylamide (PNIPAM) is a familiar thermoresponsive polymer (Hsiue et al., 2002; Li et al., 2009) having a lower critical solution temperature (LCST) of ∼32 °C in aqueous medium (Xu et al., 2004; Li et al., 2009) which is formed by copolymerization of N-isopropylacrylamide (NIPAM) with vinyl pyrrolidone (VP) and acrylic acid (AA) through free radical polymerization (Naha et al., 2009, 2010). PNIPAM has been widely used in the preparation of thermoresponsive systems for biomedical applications e.g. controlled release of drugs and in tissue engineering (Hsiue et al., 2002; Xu et al., 2004; Zhang et al., 2005; Xu et al., 2006). Recently, the effect of NIPAM/BAM copolymer nanoparticles (NPs) of varying size and varying copolymer ratios on adsorption of proteins from plasma and the potential implications for biological interactions has been reported (Cedervall et al., 2007) to impede or even reverse the fibrillation of amyloid-β, the protein involved in Alzheimer's disease (Cabaleiro-Lago et al., 2008). In another study, PNIPAM NPs have been effectively applied to target human keratinocyte (HaCaT) and colon cells (SW-480) (Naha et al., 2009).

In this study, we explored the development, characterization and comparative PK/PD analysis of CUR, DMC, BDMC-loaded PNIPAM Nps in middle cerebral artery occlusion (MCAO)-induced focal cerebral ischemia thorough i.n. administration in Wistar rats.

MATERIAL AND METHODS

Materials

Chemicals CUR, DMC and BDMC (assigned purity >96%; molecular weight 368.38, 338.36 and 308.33; melting point 183, 168 and 223–224 °C, respectively) were procured from LGC Promochem India Pvt Ltd., Bangalore, Karnataka, India. Nimesulide (internal standard, IS) was obtained as a gift sample from Jubilant Clinsys Clinical Research Limited (Noida, Uttar Pradesh, India). NIPAM was purchased from ACROS Chemicals, NJ; VP, AA, 2,3,5-Triphenyltetrazolium chloride (TTC) and EDTA were purchased from Sigma-Aldrich, MO. Rat TNF-α and IL-1β Immunoassay Kits were purchased from eBioscience, SDG & BD Bioscience, NJ. All solvents used for ultra-performance liquid chromatography–mass spectroscopy (UPLC-MS) were of UPLC grade.

Methods

Preparation of polymeric NIPAM (PNIPAM) NPs

NIPAM (85 mg), VP (10 μL) and AA (5 μL) were dissolved in 10 mL water and polymerization was activated by adding 20 μL of ferrous ammonium sulphate and 30 μL of ammonium persulphate solution. About 20 μL of N,N′-methylene-bis-acrylamide (MBA) (20 mg/mL) was eventually added for crosslinking during the polymerization reaction. The reaction was continued for 28 h at 32 °C in a N₂ atmosphere. Post-polymerization, aqueous dispersion was dialysed for 24 h using a spectropore membrane dialysis bag (cutoff 12 kDa) and freeze dried. Curcuminoids (CUR, DMC and BDMC) were physically entrapped in the hydrophobic core of PNIPAM micelles by vortexing and sonication. Briefly, 10 mg PNIPAM lyophilized powder was dispersed in 2 mL of water and ethanolic solution (1 mg/mL) of CUR, DMC and BDMC were separately and gradually added. The mixture was vortexed and sonicated (1 min, 35% amplitude) to give a clear NP suspension. The different drug-loaded polymeric NPs (named CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM) were finally lyophilized to get fine powder.

Characterization of drug-loaded PNIPAM NPs

Transmission electron microscopy (TEM) was performed to determine the morphology of the prepared NPs, in which a drop of the NP suspensions were placed on carbon-coated copper grid (Polysciences Inc., Warrington, PA), 2% uranyl acetate was added and dried. Images were acquired by digital imaging software—AMT image capture engine (version 5.42.391) (Advanced Microscopy Techniques, Corp., MA). Surface morphology of the prepared NPs was studied using scanning electron microscopy (SEM) (Zeiss EVO 40; Carl Zeiss SMT Ltd, Cambridge, UK) after gold coating. The particle size, polydispersity index (PDI) and zeta potential were determined by Zetasizer Nano ZS (Malvern Instruments Ltd, Worcestershire, UK). The lyophilized powder was dispersed in aqueous solution and observed at 25 °C with a fixed angle of 90° (Ahmad et al., 2014).

Determination of the loading capacity, encapsulation efficiency and process yield of NPs

In order to quantify encapsulation efficiency (EE) and loading capacity (LC) of PNIPAM-NPs, the lyophilized particles were redispersed in water and were separated from unentrapped free drug using a Millipore UFP2THK24 (100 kDa cutoff), (Millipore Corp., MA) membrane filter. The samples were analysed in triplicate by means of reverse-phase ultra-performance liquid chromatography (UPLC) with Waters ACQUITY UPLC™ BEH C18 (2.1 × 100 mm; 1.7 µm) column (Waters Corp., Milford, MA) using isocratic mobile phase (acetonitrile:10 mM ammonium formate:formic acid::90:10:0.05 v/v/v), at a flow rate of 0.2 mL min⁻¹. The peak detection was performed at specific λ_max of 427, 421 and 417 nm for CUR, DMC and BDMC. The EE and LC was calculated by the following formula: $\text{EE}\left(\%\right)=\left(\text{total drug}-\text{free drug}\right)/\text{total drug}\times \text{100}$ $\text{LC}\left(\%\right)=\left(\text{total drug}-\text{free drug}\right)/\text{weight of NPs}\times 100$ The process yield (PY) was calculated from the weight of dried NPs recovered (W₁) and the sum of the initial dry weight of starting materials (W₂) using the following formula: $\text{PY}\left(\%\right)={\text{W}}_1/{\text{W}}_2\times 100$

NMR spectroscopy

Nuclear magnetic resonance spectroscopy (¹H-NMR) was performed for determination of elucidation of molecular structure and crosslinking of PNIPAM NPs. NPs were dispersed the in D₂O solvent and the ¹H-NMR spectra were taken by using a Bruker Spectrospin Advance 300MHz spectrometer.

Differential scanning calorimetry study

Differential scanning calorimetry (DSC) analysis of pure CUR, DMC, BDMC, physical mixture (PNIPAM + CUR, DMC and BDMC, separately), CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM NPs were carried out using PerkinElmer® 7 DSC (Waltham, MA) calibrated with indium. Sample (5 mg) was placed onto a standard aluminium pan, crimped and heated from 0–300 °C at a heating rate of 5 °C/min. An empty sealed pan was used as reference. All samples were run in triplicate.

In vitro release study

The in vitro release of curcuminoids from developed PNIPAM-NPs was performed using dialysis method. CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM (equivalent to 10 mg of drug) were placed in pre-treated dialysis sacs and immersed into 100 mL of phosphate-buffered solution (PBS), Ph 7.4, at 32 °C and magnetically stirred at 50 rpm. Aliquots (1 mL) were withdrawn at stipulated time intervals from the release medium and replaced with the same amount of phosphate buffer. The samples were analysed by the method described in “Determination of the loading capacity, encapsulation efficiency and process yield of NPs” section.

Ex vivo permeability study

Ex vivo permeation study was carried out using Automated Transdermal Diffusion Cells Sampling System (SFDC 6; LOGAN Instruments Corp., NJ) through excised goat nasal mucosa. Briefly, tissue samples (permeation area of 0.785 cm²) were fixed and 2 mL of PBS (pH 7.4) was added to the receptor chamber. After a pre-incubation time of 20 min, pure drug solution and formulation equivalent to 10.0 mg of CUR, DMC and BDMC were placed separately in the donor chamber in each case. 95:5% O₂:CO₂ was bubbled to the system and temperature was maintained at 37 °C. Two millilitre of samples were withdrawn from the receptor chamber at stipulated time intervals and replaced with an equivalent volume of PBS, for a period of 24 h. The samples withdrawn were filtered and analysed by the method specified in “Determination of the loading capacity, encapsulation efficiency and process yield of NPs” section.

In vivo study

For the in vivo pharmacokinetic, biodistribution and pharmacodynamic studies proper approval (protocol approval no. 847) was taken from Animal Ethical Committee, Jamia Hamdard (New Delhi, India), conforming to National Guidelines on the Care and Use of Laboratory Animals. Wistar rats (300–400 g, 8–10 weeks old) were kept in an environmentally controlled room (temperature: 25 ± 2 °C, humidity: 60 ± 5%, 12-h dark–light cycle) for at least 1 week before the experiments. Animals were fed on a standard pelleted diet and water (ad libitum).

Bioanalytical method development and validation by UPLC/quadrupole time-of-flight MS/MS

Chromatographic separation was performed by Waters ACQUITY UPLC™ system (Waters Corp., MA) equipped with a binary solvent delivery system, an auto-sampler, column manager and a tunable MS detector (SYNAPT; Waters Corp., Manchester, UK) using Waters ACQUITY UPLC™ BEH C18 (2.1 × 100 mm; 1.7 µm) column (Waters Corp.). The mobile phase being acetonitrile:10 mM ammonium formate:formic acid (90:10:0.05 v/v) with flow rate of 0.2 mL min⁻¹ and injection volume of 10 µL. The quadrupole time-of-flight (Q-ToF) SYNAPT MS was operated in V-mode with resolution over 32 000 mass, 1.0 min scan time and 0.02 s inter-scan delay. Collision gas was kept at a pressure of 5.3 × 10⁻⁵ Torr. The transitions occurred at m/z 367.1/217.1, 337.1/173.2, 307.1/187.2 and 307.0/229.0 for CUR, DMC, BDMC and nimesulide (IS) while compound-dependent parameter i.e. trap collision energy (Trap CE) were set to 24.0, 26.0, 29.5 and 29.5 eV, respectively.

Calibration standards for CUR, DMC and BDMC in plasma, lungs (LH) and brain homogenate (BH) with a set of eight non-zero concentrations (1, 2, 25, 210, 420, 640, 850 and 1000 ng mL⁻¹) were prepared for each by spiking 2% aqueous analytes in blank samples of plasma, LH and BH (20 µL aqueous aliquots to 980 µL blank samples). Quality control samples were prepared at three levels: 800 (HQC, high-quality control), 400 (MQC, middle quality control) and 2.9 ngmL⁻¹ (LQC, low quality control). All the solutions were stored at 2–8 °C until use. For the sample preparation, 500 µL of each sample (calibration standard, QC and blank plasma), 50 µL of IS (50 ng mL⁻¹), 200 µL of 10 mM ammonium formate solution was mixed and vortexed at 300 rpm for 5 min. About 5 mL of extraction mixture (ethyl acetate:chloroform::9:1) was added, shaked and centrifuged (4000 rpm, 10 min, 4 °C). Four millilitres of supernatant organic layer was dried and reconstituted in 500 µL of mobile phase.

The validation of developed bioanalytical method was performed according to USFDA guidelines (US Food and Drug Administration, 2011). Linearity was calculated using a regression equation with a weighting factor of 1/x² for drugs to IS concentration to produce the best fit for the concentration-detector response relationship. The LLOQ was determined as the minimum concentration of analyte which gives a signal-to-noise ratio of 10:1. The extraction efficiency (recovery) of CUR, DMC and BDMC were determined (at LLOQ, LQC, MQC and HQC levels) by comparing the mean area response of six replicates from spiked samples to that of extracted drug free samples (plasma, lungs and brain homogenate). For intra-day and inter-day (accuracy and precision) analyses, six replicate analyses of spiked samples (consisting of LLOQC, LQC, MQC and HQC) were performed on the same day and three consecutive days, respectively. The precision of the assay was calculated as percent coefficient of variation over the given concentration range, while the accuracy was calculated as the ratio of the calculated mean values of the samples to their respective nominal values. Robustness of the method was determined by making slight changes in operating conditions (mobile phase composition, flow rate and pH) while for ruggedness, six replicates were runs of one batch of precision and accuracy was performed using a different column (same type) and by a different analyst employing the same instrument. Stability of spiked samples were evaluated by analysing six replicates of three different QC levels (LQC, MQC and HQC) each exposed to different conditions (time and temperature). The long-term stability was assessed after 1 month storage at deep freeze (−80 °C), short-term stability was determined after the exposure (of processed samples) at 10 °C for 24 h in auto-sampler, freeze-thaw stability was evaluated for three consecutive freeze-thaw cycles from −80 °C to room temperature while bench-top stability was determined for 24 h storage in optimized conditions. The QC samples were quantified against the freshly spiked calibration curve standards. Percentage stability was determined by the formula: $\text{Stability}\left(\%\right)=\left(\begin{array}{l}\text{mean corrected response of stability stock/}\\ \text{mean response of comparison stock}\end{array}\right)\times 100$

Pharmacokinetics, biodistribution, brain-targeting efficiency and brain drug-targeting potential

For pharmacokinetic and biodistribution study, rats were divided into 12 groups (n = 3) and CUR solution, CUR-PNIPAM, DMC solution DMC-PNIPAM, BDMC solution and BDMC-PNIPAM equivalent to dose of 100 µg/kg body weight were administered through intravenous (i.v.) and i.n. route, separately. Blood samples were collected at 0.5, 1, 2, 4, 8, 12 and 24 h of pre-treatment, processed and analysed as per the method described in the section Bioanalytical method development and validation by UPLC/quadrupole time-of-flight MS/MS. C_max, AUC_0–t and t_1/2 were acquired from the obtained plasma data. For the biodistribution study, 21 animals (3 × 7 time points) were taken for each of 12 groups and dosing was performed as mentioned for pharmacokinetic study. At each time point, animals (n = 3) was sacrificed for plasma, lungs and brain homogenates analysed as per the method in section dealing with bioanalytical method development and validation. From the data obtained from biodistribution study, brain-targeting efficiency (%DTE) and brain drug-targeting potential (%DTP) were calculated as per the formulae (Babbar et al., 2000; Kumar et al., 2008): $\%\text{DTE}={\left({\text{AUC}}_{\text{brain}}/{\text{AUC}}_{\text{blood}}\right)}_{\text{in}}/{\left({\text{AUC}}_{\text{brain}}/{\text{AUC}}_{\text{blood}}\right)}_{\text{iv}}\times 100$ $\%\text{DTP}=\left(B_{in}-B_x\right)/B_{in}\times 100$ where, B_x = (B_iv/P_iv) × P_in, B_x is the brain area under the curve (AUC) fraction contributed by systemic circulation through the BBB following an i.n. administration; B_iv and P_iv are the AUC_0–24 of brain and blood, respectively following i.v. administration; B_in and P_in are the AUC_0–24 of brain and blood, respectively following i.n. administration.

Pharmacodynamic study in cerebral ischemia

Rats were divided into 15 groups consisting of six rats in each group: (Group I) Sham operated (control), (Group II) Sham + PNIPAM placebo operated (substantial control), (Group III) MCAO induced, (Group IV–XV) MCAO induced + 100 µg/kg body weight equivalent dose of CUR solution, CUR-PNIPAM, DMC solution, DMC-PNIPAM, BDMC solution and BDMC-PNIPAM through both i.v. and i.n. route, separately. The right MCAO was done by intraluminal filament model (Longa et al., 1989). In brief, the rats were anesthetized with chloral hydrate (400 mg/kg, i.p.) and a silicone rubber (4 0-3033REPK10; DOCCOL, MA)-coated monofilament was inserted into the external carotid artery to reach middle cerebral artery (MCA) through the internal carotid artery until a slight resistance was observed which is indicative of occlusion. The filament was slowly withdrawn after 2 h of induction of ischemia. Sham group rats, undergone all surgical procedures except the MCAO. After the reperfusion period (22 h), the animals were assessed for neurobehavioural activity (locomotor and grip strength) and then sacrificed to remove brains for estimation inflammatory cytokines levels (TNF-α and IL-1β), antioxidant activity (data not shown) and histopathological studies (data not shown). Neurological function was evaluated using a 0–5-point scale neurological score: 0 = no neurological dysfunction; 1 = failure to extend left forelimb fully when lifted by tail; 2 = circling to the contralateral side; 3 = falling to the left; 4 = no spontaneous walk or in a comatose state and 5 = death. The scores were assessed in a blinded fashion (Masuo et al., 1997). For the measurement of cytokine level, brain tissues were homogenized in a pH 7.6 buffer consisting of 20 mM Tris–HCl, 100 mM KCl, 5 mM NaCl, 2 mM EDTA, 1 mM EGTA, 2 mM dithiothreitol and 2 mM PMSF. Homogenates were centrifuged at 4 °C and 12 000g for 15 min. Supernatants were removed and assayed in duplicate using TNF-α and IL-1β assay kits (eBioscience, USA) according to the manufacturer's guidelines. Tissue cytokine concentrations were expressed as picograms of antigen per millilitre of protein. All data are reported as mean ± SEM and the differences between the groups were tested using Student's t-test at the level of p < 0.05. More than two groups were compared using analysis of variance and the difference greater at p < 0.05 was considered significant.

Toxicity study

In vivo toxicity evaluation of PNIPAM-NPs (placebo) (Group A), CUR-PNIPAM (Group B), DMC-PNIPAM (Group C) and BDMC-PNIPAM (Group D) was carried out to assess the mortality (if any) followed by nasal and brain histology (after 14 days treatment) at an equivalent dose of 100 µg/kg of curcuminoids. The rats were dosed once daily in morning (between 9.00 and 10.00 am) with 50 μL suspension form, equivalent to 100 µg/kg of CUR, DMC and BDMC by i.n. route for 14 days. Prior to treatment, every day the animals were examined for any abnormal behaviour, mortality and morbidity. Group E was untreated group and served as normal control. After 14 days, the rats were sacrificed to dissect out brain and nasal mucosa, fixed in 10% neutral-buffered formalin solution and transverse sections of the tissues were stained with hematoxylin and eosin and were examined microscopically for any histological changes as compared to normal control group.

RESULTS

Preparation and characterization of PNIPAM NPs

Twelve placebo formulations with varying concentrations of NIPAM, VP and AA are presented in Table 1. The results revealed that minimum (82.4 ± 5.3 nm) and maximum (989.8 ± 115.3 nm) size of the developed submicron particles were obtained with formulation A2 and D3, respectively. Lowest PDI was obtained in the formulation D3 (0.106 ± 0.039%) while maximum PY was obtained in the case of formulation D1 (82.90 ± 3.78%), followed by D2 (80.24 ± 1.17%) and D3 (72.24 ± 2.13%). Form the data obtained from placebo, formulation D3 was selected for the drug loading because of smallest size, lowest PDI and optimum PY.

Table 1. Optimization of placebo NPs (PNIPAM-NPs) on the basis of NIPAM, VP, AA and MBA ratio.

Formulation code	NIPAM (%)	VP (%)	AA (%)	Concentration of MBA (mg/mL)	Mean particle size (nm ± SD)	Mean PDI ± SD	PY (%) ± SD
A1	50	25	25	20	815.9 ± 212.1	0.906 ± 0.048	21.78 ± 1.75
A2	50	15	35	20	989.8 ± 115.3	0.989 ± 0.032	17.34 ± 2.35
A3	50	35	15	20	727.4 ± 106.6	0.856 ± 0.028	23.73 ± 1021
B1	60	20	20	20	415.1 ± 66.8	0.325 ± 0.045	27.34 ± 2.45
B2	60	15	25	20	519.2 ± 45.7	0.336 ± 0.033	35.54 ± 1.76
B3	60	25	15	20	489.2 ± 59.9	0.346 ± 0.029	38.46 ± 2.12
C1	70	15	15	20	268.3 ± 27.7	0.291 ± 0.022	42.56 ± 2.33
C2	70	10	20	20	288.9 ± 21.8	0.286 ± 0.018	41.61 ± 3.46
C3	70	20	10	20	232.6 ± 22.3	0.281 ± 0.015	48.63 ± 5.16
D1	80	10	10	20	88.6 ± 45	0.272 ± 0.028	82.90 ± 3.78
D2	80	5	15	20	108.3 ± 3.8	0.189 ± 0.029	80.24 ± 1.17
D3	80	15	5	20	82.4 ± 5.3	0.106 ± 0.039	72.24 ± 2.13

Particle morphology, particle size, PDI, zeta potential, PY, LC and EE after drug loading

Three different drug polymer ratios (1:10, 2:10 and 3:10) were tried to evaluate the particle size, PDI, zeta potential, LC, EE and PY for each drug. The data showed that drug polymer ratio 1:10 produced best results in terms of the selected parameters. For CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM, particle size was found to be 92.46 ± 2.8, 91.23 ± 4.2 and 94.28 ± 1.91 nm; PDI = 0.291 ± 0.052, 0.283 ± 0.063 and 0.242 ± 0.046; zeta potential =−16.2 ± 1.42, −15.6 ± 1.33 and −16.6 ± 1.21 mV; PY =81.83 ± 4.69, 82.06 ± 4.29 and 82.29 ± 4.71%; LC = 39.31 ± 3.7, 38.91 ± 3.6 and 40.61 ± 3.6% and EE = 84.63 ± 4.2, 84.71 ± 3.99 and 85.73 ± 4.31%, respectively (Table 2). TEM and SEM microphotographs of CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM are presented in Figure 1 which revealed spherical or nearly spherical structures of NPs. The size analysis conforms to the results obtained by Zetasizer (Malvern Instruments Ltd).

Figure 1. Comparative SEM (left) and TEM (right) images of (A and D) PNIPAM-CUR, (B and E) PNIPAM-DMC and (C and F) PNIPAM-BDMC.

Table 2. Optimization of drug polymer ratio of CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM on the basis of particle size, PDI, zeta potential, PY, LC and EE (%).

Formulation code	Drug:polymer ratio	Mean particle size (nm ± SD)	Mean PDI ± SD	PY ± SD	Mean zeta potential (mV ± SD)	LC (%)	EE (%)
CUR-PNIPAM-1	1:10	92.46 ± 2.8	0.191 ± 0.052	81.83 ± 4.69	−16.2 ± 1.42	39.31 ± 3.7	84.63 ± 4.2
CUR-PNIPAM-2	2:10	198.74 ± 6.6	0.423 ± 0.042	83.29 ± 5.24	−15.02 ± 2.90	51.73 ± 2.4	81.24 ± 4.9
CUR-PNIPAM-3	3:10	328.24 ± 8.9	0.462 ± 0.064	87.23 ± 4.78	−12.1 ± 3.16	59.29 ± 2.8	74.13 ± 3.9
DMC-PNIPAM-1	1:10	91.23 ± 4.2	0.183 ± 0.063	82.06 ± 4.29	−15.6 ± 1.33	38.91 ± 3.6	84.71 ± 3.99
DMC-PNIPAM-2	2:10	196.78 ± 6.2	0.429 ± 0.049	84.91 ± 6.11	−14.9 ± 2.78	52.16 ± 2.9	80.33 ± 4.2
DMC-PNIPAM-3	3:10	337.63 ± 7.2	0.469 ± 0.061	86.73 ± 4.62	−13.6 ± 3.76	62.39 ± 3.4	71.23 ± 2.6
BDMC-PNIPAM-1	1:10	94.28 ± 1.91	0.142 ± 0.046	82.29 ± 4.71	−16.6 ± 1.21	40.61 ± 3.6	85.73 ± 4.31
BDMC-PNIPAM-2	2:10	205.53 ± 3.16	0.424 ± 0.046	86.74 ± 2.12	−14.3 ± 2.69	53.23 ± 2.9	82.16 ± 4.62
BDMC-PNIPAM-3	3:10	346.23 ± 6.9	0.472 ± 0.068	88.24 ± 4.18	−12.8 ± 3.27	65.26 ± 4.2	72.93 ± 3.9

SD, standard deviation. All the above nine formulation PNIPAM -NPs, PNIPAM/MBA ratio fixed at 99.4/0.6 in which NIPAM/VP/AA::85/10/05.

NMR (¹H-NMR)

¹H-NMR spectroscopy of NIPAAM-VP-AA copolymer revealed the characteristic peaks of NIPAAM and VP while those of AA are lost in the noise due low percentage (Figure 2). The spectrum showed the absence of resonance due to the vinyl end group protons of the monomers which is indicative of polymerization. Instead, these protons resonated at upfield region indicating the presence of saturated protons. The ratio of the peak intensities of the methine proton (−CH−) of NIPAM group to half of the intensity of β-methylene protons (−CH₂−) of pyrrolidone group was found to be 8.7:1 which means that for each VP unit there are about nine NIPAAM units present which is nearly similar to the initial composition of the monomers taken.

Figure 2. NMR spectra of the developed co-poly-(NIPAM–VP–AA) revealing the formation of PNIPAM.

DSC study

DSC thermograms (Figure 3) of pure CUR, DMC and BDMC showed sharp endothermic peaks (equivalent to the melting point) at ∼186.6, 176.0 and 231.8 °C, respectively, which were retained in the physical mixtures of PNIPAM and curcuminoids (individually). Freeze-dried NPs showed diminished peaks which is indicative of molecular dispersion of curcuminoids within the PNIPAM-NPs.

Figure 3. DSC thermograms of pure CUR, DMC and BDMC, physical mixtures of monomers and curcuminoids and freeze-dried NPs.

In vitro release study

The cumulative percentage release of CUR, DMC and BDMC from CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM (drug:polymer ratio 10:1) was 78.64 ± 2.77, 76.79 ± 2.56 and 73.99 ± 1.95%, while for CUR, DMC and BDMC solution it was 36.26 ± 2.31, 30.98 ± 1.97 and 25.66 ± 2.15%, respectively over a period of 24 h (Figure 4).

Figure 4. The cumulative percentage release of CUR, DMC, BDMC from CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM (drug:polymer ratio 10:1) as compared to pure CUR, DMC and BDMC.

Ex vivo permeability study

Ex vivo permeability study revealed remarkably higher maximum percentage drug permeation with NPs (78.02 ± 3.21, 75.18 ± 2.78 and 79.99 ± 3.11% for CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM, respectively) as compared to pure drug suspension (25.56 ± 1.98, 18.73 ± 1.03 and 17.29 ± 1.11% for CUR, DMC and BDMC, respectively) through goat nasal mucosa in 24 h (Figure 5).

Figure 5. Ex vivo permeation profiles of developed NPs as compared to pure curcuminoids through goat nasal mucosa.

Bioanalytical method development and validation by UPLC/Q-ToF-MS/MS

MS and MS/MS scans of analytes and IS are shown in Figure 6 while typical chromatograms obtained for blank, rat plasma (PL), lungs (LH) and brain homogenates (BH) sample (extracted) have been shown in Figure 7. The mean recovery (n = 6) of CUR, DMC and BDMC from plasma, lungs and brain homogenates were >80%. Method was found to be linear (r² > 0.999) in PL, LH and BH over the range of 1–1000 ng mL⁻¹ for all curcuminoids. Representative chromatograms of extracted blank plasma, lungs and brain homogenates fortified with CUR, DMC and BDMC demonstrate the selectivity of the method. Inter-day and intra-day precision and accuracy of the method are listed in Tables 3–5. Summarily, the intra-batch and inter-batch %CV for all the QC levels of CUR, DMC and BDMC were obtained between 0.09–3.07, 0.44–2.72 and 0.27–3.86 in LH, BH and PL. Intra-batch and inter-batch accuracy were within the range of 93.00–99.19, 94.22–99.08 and 91.5–98.97% for LH, BH and PL, respectively for all QC levels of CUR, DMC and BDMC. Tables 6–8 summarizes the data obtained for stability experiments which revealed stable life of CUR, DMC and BDMC in all storage conditions (long-term, freeze-thaw, bench-top and post-processing stability).

Figure 6. MS and MSMS scans of CUR, DMC, BDMC and IS (nimesulide).

Figure 7. Typical chromatograms of brain homogenate extracted (A) CUR, (B) DMC, (C) BDMC, (D) Blank and (E) Nimesulide (IS).

Table 3. Validation: precision and accuracy data for CUR in different biomatrixes.

			Intra-batch precision			Inter-batch precision
Biomatrix	Quality controls samples	Theoretical concentration (ng mL^−1 or ng g^−1)	Observed concentration (ng mL^−1 or ng^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Observed concentration (ng mL^−1 or ng g^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Recovery (%)
Brain  homogenate	LLOQQC	1.00	0.97 ± 0.011	96.29	1.17	0.97 ± 0.012	96.33	1.21	86.79
	LQC	2.90	2.83 ± 0.062	97.41	2.20	2.84 ± 0.050	98.02	1.74	87.96
	MQC	400.00	393.50 ± 3.820	98.37	0.97	393.05 ± 3.925	98.26	1.00	89.10
	HQC	800.00	792.60 ± 3.741	99.08	0.47	792.53 ± 4.132	99.07	0.52	89.16
Lungs  homogenate	LLOQQC	1.00	0.95 ± 0.014	95.00	1.49	0.94 ± 0.02	93.5	2.27	87.67
	LQC	2.90	2.79 ± 0.035	96.03	1.27	2.78 ± 0.014	95.86	0.51	88.19
	MQC	400.00	390.5 ± 3.56	97.63	0.91	390.25 ± 0.354	97.56	0.09	89.27
	HQC	800.00	793.5 ± 2.12	99.19	0.27	791.75 ± 2.475	98.97	0.31	89.79
Plasma	LLOQQC	1.00	0.925 ± 0.035	92.5	3.82	0.93 ± 0.007	93.00	0.76	81.42
	LQC	2.90	2.73 ± 0.057	94.14	2.07	2.75 ± 0.028	94.83	1.03	95.30
	MQC	400.00	379.50 ± 2.121	94.88	0.56	382.75 ± 4.596	95.69	1.20	94.12
	HQC	800.00	783.5 ± 3.54	97.94	0.45	791.75 ± 2.475	98.97	0.31	92.32

Table 4. Validation: precision and accuracy data for DMC in different biomatrixes.

			Intra-batch precision			Inter-batch precision
Biomatrix	Quality controls samples	Theoretical concentration (ng mL^−1 or ng g^−1)	Observed concentration (ng mL^−1 or ng g^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Observed concentration (ng mL^−1 or ng g^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Recovery (%)
Brain  homogenate	LOQQC	1.00	0.96 ± 0.013	95.38	1.36	0.97 ± 0.013	95.87	1.30	87.67
	LQC	2.90	2.82 ± 0.047	97.18	1.65	2.84 ± 0.042	97.90	1.47	87.54
	MQC	400.00	393.46 ± 3.807	98.36	0.97	392.53 ± 3.663	98.13	0.93	89.33
	HQC	800.00	792.60 ± 3.690	99.03	0.47	792.24 ± 4.063	99.03	0.51	88.86
Lungs  homogenate	LOQQC	1.00	0.95 ± 0.021	94.5	2.24	0.94 ± 0.014	94.00	1.50	83.23
	LQC	2.90	2.80 ± 0.035	96.38	1.26	2.79 ± 0.021	96.03	0.76	86.28
	MQC	400.00	390.79 ± 2.12	97.63	0.54	390.25 ± 0.354	97.56	0.09	89.44
	HQC	800.00	792.25 ± 3.54	99.06	0.45	791.25 ± 1.77	98.91	0.22	89.85
Plasma	LOQQC	1.00	0.93 ± 0.028	93.00	3.04	0.93 ± 0.007	93.00	0.76	88.76
	LQC	2.90	2.74 ± 0.078	94.31	2.84	2.75 ± 0.028	94.83	1.03	88.92
	MQC	400.00	378.50 ± 2.121	94.63	0.56	382.75 ± 4.59	95.69	1.20	89.14
	HQC	800.00	781.00 ± 2.83	97.63	0.36	791.75 ± 2.475	98.97	0.31	89.93

Table 5. Validation: precision and accuracy data for BDMC in different biomatrixes.

			Intra-batch precision			Inter-batch precision
Biomatrix	Quality controls samples	Theoretical concentration (ng mL^−1 or ng g^−1)	Observed concentration (ng mL^−1 or ng g^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Observed concentration (ng mL^−1 or ng g^−1) ± SD	Accuracy (%)	Precision (%C.V.)	Recovery (%)
Brain homogenate	LOQQC	1.00	0.95 ± 0.026	94.22	2.72	0.96 ± 0.019	94.72	1.97	87.34
	LQC	2.90	2.84 ± 0.036	97.82	1.27	2.84 ± 0.033	98.05	1.17	87.97
	MQC	400.00	392.58 ± 3.419	98.15	0.87	392.33 ± 3.560	98.08	0.91	89.09
	HQC	800.00	792.25 ± 3.455	99.03	0.44	792.51 ± 3.641	99.06	0.46	89.59
Lungs homogenate	LOQQC	1.00	0.94 ± 0.028	94.00	3.01	0.93 ± 0.014	93.00	1.52	82.97
	LQC	2.90	2.80 ± 0.014	96.55	0.51	2.79 ± 0.021	96.03	0.76	88.32
	MQC	400.00	389.5 ± 2.121	97.38	0.54	389.57 ± 0.354	97.44	0.09	87.95
	HQC	800.00	792.50 ± 3.54	99.06	0.45	791.25 ± 1.768	98.91	0.22	88.67
Plasma	LOQQC	1.00	0.915 ± 0.035	91.50	3.86	0.94 ± 0.014	92.50	1.53	80.26
	LQC	2.90	2.77 ± 0.035	95.34	1.28	2.78 ± 0.064	95.69	2.29	86.42
	MQC	400.00	384.5 ± 3.456	96.125	0.92	385.25 ± 1.061	96.31	0.28	81.82
	HQC	800.00	782.5 ± 2.12	97.81	0.27	785.25 ± 3.889	98.16	0.50	87.06

Table 6. Validation: stability data for CUR in different biomatrixes.

	LQC (2.90 ng/mL or ng g^−1)			MQC (400 ng/mL or ng g^−1)			HQC (800 ng/mL or ng g^-1)
Exposure condition	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma
Long-term stability; recovery (ng) after storage (−80 °C)
Previous day	2.82 ± 0.03	2.79 ± 0.035	2.73 ± 0.057	393.50 ± 3.820	390.5 ± 3.56	379.50 ± 2.121	797.93 ± 1.40	793.53 ± 2.12	783.55 ± 3.541
Thirtieth day	2.80 ± 0.03 (99.29%)	2.77 ± 0.043 (99.28%)	2.72 ± 1.412 (99.63%)	389.23 ± 3.191 (98.91%)	385.42 ± 2.89 (98.70%)	376.14 ± 2.045 (99.11%)	788.37 ± 2.36 (98.80%)	783.91 ± 2.191 (98.79%)	780.94 ± 2.141 (99.67%)
Freeze-thaw stress; recovery (ng) after freeze-thaw cycles (−80 to 25 °C)
Pre-cycle	2.82 ± 0.03	2.79 ± 0.035	2.73 ± 0.057	393.50 ± 3.820	390.5 ± 3.56	379.50 ± 2.121	797.93 ± 1.40	793.53 ± 2.122	783.55 ± 3.541
First cycle	2.81 ± 0.03 (99.65%)	2.78 ± 0.025 (99.64%)	2.73 ± 1.151 (100%)	392.94 ± 2.791 (99.86%)	389.90 ± 2.46 (99.85%)	378.12 ± 1.210 (99.64%)	795.67 ± 1.59 (99.72%)	790.12 ± 1.493 (99.57%)	781.92 ± 1.491 (99.79%)
Second cycle	2.79 ± 0.02 (98.94%)	2.78 ± 0.012 (99.64%)	2.72 ± 1.211 (99.63%)	391.16 ± 1.141 (99.41)	388.79 ± 1.46 (99.56%)	377.49 ± 2.018 (99.47%)	794.80 ± 1.15 (99.61%)	788.79 ± 1.791 (99.40%)	778.91 ± 2.091 (99.41%)
Third cycle	2.76 ± 0.01 (97.87%)	2.77 ± 0.051 (99.28%)	2.71 ± 2.001 (99.27%)	390.02 ± 1.213 (99.12%)	388.19 ± 1.49 (99.41%)	376.91 ± 1.791 (99.32%)	791.32 ± 2.29 (99.17%)	783.146 ± 2.492 (98.69%)	775.11 ± 1.791 (98.92%)
Bench-top stability; recovery (ng) at room temperature (25 °C)
0 h	2.82 ± 0.03	2.79 ± 0.035	2.73 ± 0.057	393.50 ± 3.820	390.5 ± 3.56	379.50 ± 2.121	797.93 ± 1.40	793.53 ± 2.12	783.55 ± 3.541
24 h	2.81 ± 0.02 (99.65%)	2.77 ± 0.049 (99.28%)	2.72 ± 1.718 (99.63%)	391.15 ± 2.746 (99.40%)	389.04 ± 2.94 (99.63%)	378.09 ± 1.952 (99.63%)	795.29 ± 0.38 (99.67%)	786.491 ± 2.791 (99.11%)	777.12 ± 2.491 (98.92%)
Post-processing stability; recovery (ng) after storage in auto-sampler (4 °C)
0 h	2.82 ± 0.03	2.79 ± 0.035	2.73 ± 0.057	393.50 ± 3.820	390.5 ± 3.56	379.50 ± 2.121	797.93 ± 1.40	793.53 ± 2.12	783.55 ± 3.541
4 h	2.81 ± 0.02 (99.65%)	2.78 ± 0.028 (99.64%)	2.72 ± 0.091 (99.63%)	391.59 ± 2.491 (99.51%)	388.94 ± 1.29 (99.60%)	378.91 ± 1.107 (99.84%)	796.11 ± 1.06 (99.77%)	787.194 ± 1.927 (99.20%)	780.16 ± 2.121 (99.57%)

Values (mean ± SD) are derived from six replicates. Figures in parenthesis represent analyte concentration (%) relative to time zero. Theoretical contents: LQC: 2.9, MQC: 400 and HQC: 800 ng mL⁻¹.

Table 7. Validation: stability data for DMC in different biomatrixes.

	LQC (2.90 ng/mL or ng g^−1)			MQC (400 ng/mL or ng g^−1)			HQC (800 ng/mL or ng g^−1)
Exposure condition	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma
Long-term stability; recovery (ng) after storage (−80 °C)
Previous day	2.85 ± 0.02	2.80 ± 0.035	2.74 ± 0.078	393.46 ± 3.807	390.79 ± 2.12	378.50 ± 2.121	795.68 ± 3.12	792.25 ± 3.54	781.00 ± 2.83
Thirtieth day	2.79 ± 0.03 (97.89%)	2.75 ± 0.041 (98.21%)	2.70 ± 1.121 (98.54%)	387.73 ± 2.791 (98.54%)	388.11 ± 1.711 (99.31)	375.13 ± 2.046 (99.11%)	786.46 ± 1.57 (98.84%)	785.91 ± 3.419 (99.20%)	778.06 ± 2.491 (99.62%)
Freeze-thaw stress; recovery (ng) after freeze-thaw cycles (−80 to 25 °C)
Pre-cycle	2.85 ± 0.02	2.80 ± 0.035	2.74 ± 0.078	393.46 ± 3.807	390.79 ± 2.12	378.50 ± 2.121	795.68 ± 3.12	792.25 ± 3.54	781.00 ± 2.83
First cycle	2.82 ± 0.02 (98.95%)	2.79 ± 0.0431 (99.64%)	2.73 ± 0.071 (99.64%)	391.13 ± 2.194 (99.41%)	389.17 ± 1.913 (99.59%)	377.03 ± 1.034 (99.61%)	791.55 ± 3.08 (99.48%)	790.37 ± 1.416 (99.76%)	779.93 ± 1.613 (99.86%)
Second cycle	2.79 ± 0.02 (97.89%)	2.77 ± 1.011 (98.93%)	2.71 ± 0.070 (98.91%)	389.19 ± 1.791 (98.91%)	388.13 ± 1.712 (99.32%)	376.49 ± 1.034 (99.47%)	788.74 ± 2.32 (99.13%)	788.13 ± 2.016 (99.48)	778.03 ± 1.037 (99.62%)
Third cycle	2.76 ± 0.01 (96.84%)	2.76 ± 1.104 (98.57%)	2.70 ± 0.072 (98.54%)	387.19 ± 1.919 (98.41%)	387.18 ± 1.818 (99.08%)	374.16 ± 1.161 (98.85%)	786.97 ± 2.85 (98.91%)	786.14 ± 1.037 (99.23%)	776.02 ± 2.709 (99.36%)
Bench-top stability; recovery (ng) at room temperature (25 °C)
0 h	2.85 ± 0.02	2.80 ± 0.035	2.74 ± 0.078	393.46 ± 3.807	390.79 ± 2.12	378.50 ± 2.121	795.68 ± 3.12	792.25 ± 3.54	781.00 ± 2.83
24 h	2.77 ± 0.02 (97.19%)	2.74 ± 0.094 (97.86%)	2.70 ± 0.079 (98.54%)	388.57 ± 2.716 (98.76%)	386.99 ± 1.146 (99.03%)	375.66 ± 1.312 (99.25%)	786.81 ± 2.29 (98.89%)	785.94 ± 2.161 (99.20%)	778.12 ± 1.373 (99.63%)
Post-processing stability; recovery (ng) after storage in auto-sampler (4 °C)
0 h	2.85 ± 0.02	2.80 ± 0.035	2.74 ± 0.078	393.46 ± 3.807	390.79 ± 2.12	378.50 ± 2.121	795.68 ± 3.12	792.25 ± 3.54	781.00 ± 2.83
24 h	2.81 ± 0.04 (98.60%)	2.77 ± 0.471 (98.93%)	2.72 ± 0.074 (99.27%)	390.16 ± 1.161 (99.16%)	388.93 ± 1.819 (99.52%)	376.31 ± 1.493 (99.42%)	794.73 ± 4.23 (99.88%)	789.96 ± 1.037 (99.71%)	779.19 ± 1.033 (99.77%)

Values (mean ± SD) are derived from six replicates. Figures in parenthesis represent analyte concentration (%) relative to time zero. Theoretical contents: LQC: 2.9; MQC: 400 and HQC: 800 ng mL⁻¹.

Table 8. Validation: stability data for BDMC in different biomatrixes.

	LQC (2.90 ng/mL or ng g^−1)			MQC (400 ng/mL or ng g^−1)			HQC (800 ng/mL or ng g^−1)
Exposure condition	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma	Brain homogenate	Lungs homogenate	Plasma
Long-term stability; recovery (ng) after storage (−80 °C)
Previous day	2.86 ± 0.01	2.80 ± 0.014	2.77 ± 0.035	392.58 ± 3.419	389.5 ± 2.121	384.5 ± 3.456	795.69 ± 2.90	792.50 ± 3.54	782.5 ± 2.12
Thirtieth day	2.78 ± 0.03 (97.20%)	2.74 ± 0.046 (97.86%)	2.71 ± 0.049 (97.83%)	379.55 ± 2.488 (96.68%)	377.12 ± 2.790 (96.82%)	372.14 ± 2.791 (96.79%)	785.73 ± 1.48 (98.75%)	780.15 ± 2.761 (98.44%)	771.23 ± 3.106 (98.56%)
Freeze-thaw stress; recovery (ng) after freeze-thaw cycles (−80 to 25 °C)
Pre-cycle	2.86 ± 0.01	2.80 ± 0.014	2.77 ± 0.035	392.58 ± 3.419	389.5 ± 2.121	384.5 ± 3.456	795.69 ± 2.90	792.50 ± 3.54	782.5 ± 2.12
First cycle	2.84 ± 0.01 (99.30%)	2.78 ± 0.046 (99.29%)	2.75 ± 0.041 (99.28%)	388.14 ± 1.021 (98.87%)	387.19 ± 1.731 (99.41%)	381.76 ± 1.216 (99.29%)	792.99 ± 3.05 (99.66%)	788.19 ± 1.379 (99.46%)	777.35 ± 1.067 (99.34%)
Second cycle	2.82 ± 0.02 (98.60%)	2.75 ± 0.079 (98.21%)	2.72 ± 0.071 (98.19%)	384.19 ± 1.079 (97.86%)	385.08 ± 1.911 (98.87%)	378.91 ± 1.659 (98.55%)	789.85 ± 3.33 (99.27%)	782.13 ± 1.761 (98.69%)	774.03 ± 2.309 (98.92%)
Third cycle	2.79 ± 0.03 (97.55%)	2.73 ± 1.033 (97.50%)	2.70 ± 0.016 (97.47%)	379.13 ± 1.491 (96.57%)	378.81 ± 1.611 (97.26%)	373.192 ± 2.10 (97.06%)	786.24 ± 2.48 (98.81%)	779.46 ± 2.731 (98.35%)	771.34 ± 2.673 (98.57%)
Bench-top stability; recovery (ng) at room temperature (25 °C)
0 h	2.86 ± 0.01	2.80 ± 0.014	2.77 ± 0.035	392.58 ± 3.419	389.5 ± 2.121	384.5 ± 3.456	795.69 ± 2.90	792.50 ± 3.54	782.5 ± 2.12
24 h	2.79 ± 0.04 (97.55%)	2.72 ± 0.079 (97.14%)	2.71 ± 0.036 (97.83%)	378.99 ± 1.094 (96.54%)	377.81 ± 1.412 (97.00%)	372.91 ± 1.093 (96.99%)	787.12 ± 2.24 (98.92%)	779.91 ± 1.316 (98.41%)	772.93 ± 2.641 (98.78%)
Post-processing stability; recovery (ng) after storage in auto-sampler (4 °C)
0 h	2.86 ± 0.01	2.80 ± 0.014	2.77 ± 0.035	392.58 ± 3.419	389.5 ± 2.121	384.5 ± 3.456	795.69 ± 2.90	792.50 ± 3.54	782.5 ± 2.12
24 h	2.85 ± 0.01 (99.65%)	2.77 ± 0.46 (98.93%)	2.72 ± 0.070 (98.19%)	382.16 ± 2.716 (97.35%)	383.19 ± 1.614 (98.38%)	378.03 ± 2.011 (98.32%)	794.32 ± 1.90 (99.83%)	781.06 ± 1.694 (98.56%)	772.16 ± 2.046 (98.68%)

Values (mean ± SD) are derived from six replicates. Figures in parenthesis represent analyte concentration (%) relative to time zero. Theoretical contents: LQC: 2.9, MQC: 400 and HQC: 800 ng mL⁻¹.

Pharmacokinetics, biodistribution, %DTE and %DTP

Pharmacokinetic parameters (C_max, T_max, t_1/2, K_e and AUC_0–t) obtained after i.n. and i.v. administration of CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM compared with pure CUR, DMC and BDMC are presented in Tables 9–11, respectively. The data revealed a significantly higher AUC_0–t after administration of the developed NPs as compared to pure curcuminoids in all three areas (brain, lungs and plasma). Brain/plasma ratio (Tables 9–11) showed that through i.n. administration, brain concentration was found to be higher with respect to i.v. administration. %DTE, %DTP and AUC_i.n/AUC_i.v. (Table 12, Figure 8) clearly showed that when administered through i.n. route, all the developed PNIPAM NPs were found to be very effective in enhancing brain bioavailability.

Figure 8. Pharmacokinetic profiles of CUR, DMC and BDMC concentration in brain at different time intervals after administration of developed NPs compared with pure curcuminoids.

Table 9. Pharmacokinetic parameters of CUR-PNIPAM after i.n. and i.v. administration to rats at the dose of 0.100 mg kg⁻¹ in brain, lungs and plasma (n = 6, mean ± SD).

Formulation administration	Samples	Cmax (ng/mL g)	Tmax	t1/2 (h)	Ke (h^−1)	AUC0–t (ng min/mL g)
CUR (i.n.)	Brain	200.67 ± 14.90	2.00	21.04 ± 4.10	0.03294 ± 0.01	2476.17 ± 311.66
	Lungs	23.33 ± 3.20	2.00	23.36 ± 3.58	0.02967 ± 0.00472	276.33 ± 38.72
	Plasma	60.50 ± 4.32	0.5	6.43 ± 0.67	0.10784 ± 0.01141	450.63 ± 44.65
CUR (i.v.)	Brain	108.50 ± 13.02	2.00	7.48 ± 0.65	0.09268 ± 0.0082	1032.75 ± 67.16
	Lungs	177.67 ± 14.02	2.00	6.48 ± 1.00	0.10693 ± 0.0218	1570.88 ± 57.62
	Plasma	1314.17 ± 35.41	1.00	8.92 ± 0.81	0.07769 ± 0.0067	17894.96 ± 377.06
CUR-PNIPAM (i.n.)	Brain	989.33 ± 14.29	2.00	219.44 ± 33.64	0.00316 ± 0.00046	15597.13 ± 373.31
	Lungs	112.33 ± 8.78	2.00	82.51 ± 58.70	0.00840 ± 0.00419	1796.54 ± 148.20
	Plasma	419.50 ± 7.79	2.00	28.50 ± 0.77	0.02432 ± 0.00066	6424.79 ± 151.74
CUR-PNIPAM (i.v.)	Brain	410.83 ± 28.74	2.00	38.96 ± 14.50	0.01779 ± 0.01	6355.88 ± 262.95
	Lungs	215.33 ± 7.50	2.00	147.10 ± 1089.82	0.00471 ± 0.0044	3366.67 ± 142.22
	Plasma	1219.17 ± 23.21	1.00	9.27 ± 1.11	0.07480 ± 0.00893	17501.46 ± 297.96
CUR (i.n.)	Brain/plasma	3.32	4.00	3.27	0.31	5.49
CUR (i.v.)	Brain/plasma	0.08	2.00	0.84	1.19	0.06
CUR-PNIPAM (i.n.)	Brain/plasma	2.36	1.00	7.70	0.13	2.43
CUR-PNIPAM (i.v.)	Brain/plasma	0.34	2.00	4.20	0.14	0.36

Table 10. Pharmacokinetic parameters of DMC-PNIPAM after i.n. and i.v. administration to rats at the dose of 0.100 mg kg⁻¹ in brain, lungs and plasma (n = 6, mean ± SD).

Formulation administration	Samples	Cmax (ng/mL g)	Tmax	t1/2 (h)	Ke (h^−1)	AUC0–t (ng min/mL g)
DMC (i.n.)	Brain	205.50 ± 13.14	2.00	21.57 ± 3.44	0.03213 ± 0.00470	2560.92 ± 236.86
	Lungs	21.67 ± 4.41	2.00	19.31 ± 7.30	0.03590 ± 0.02245	240.13 ± 74.70
	Plasma	58.5 ± 6.53	0.50	5.85 ± 1.10	0.11855 ± 0.02614	411.29 ± 93.51
DMC (i.v.)	Brain	109.5 ± 22.37	2.00	7.37 ± 1.44	0.09405 ± 0.0093	1045.83 ± 72.00
	Lungs	176.83 ± 12.14	2.00	6.37 ± 2.94	0.10883 ± 0.0210	1551.96 ± 46.34
	Plasma	1297.50 ± 34.49	1.00	8.55 ± 2.16	0.08109 ± 0.0114	17503.29 ± 868.06
DMC-NIPAM (i.n.)	Brain	985.33 ± 18.42	2.00	195.11 ± 50.13	0.00355 ± 0.00091	15649.63 ± 782.57
	Lungs	113.83 ± 7.31	2.00	72.39 ± 61.76	0.00958 ± 0.00431	1814.79 ± 149.37
	Plasma	416.17 ± 9.13	2.00	27.89 ± 2.07	0.02485 ± 0.00206	6343.13 ± 119.57
DMC-NIPAM (i.v.)	Brain	411.17 ± 30.15	2.00	39.17 ± 12.22	0.01770 ± 0.00538	6377.17 ± 300.64
	Lungs	211.83 ± 4.96	2.00	74.55 ± 16.48	0.00930 ± 0.00233	3214.29 ± 99.68
	Plasma	1214.17 ± 11.60	1.00	9.45 ± 0.90	0.07337 ± 0.00671	17494.96 ± 299.53
DMC (i.n.)	Brain/plasma	3.51	4	3.69	0.27	6.23
DMC (i.v.)	Brain/plasma	0.08	2.00	0.86	1.16	0.06
DMC-NIPAM (i.n.)	Brain/plasma	2.37	1.00	7.00	0.14	2.47
DMC-NIPAM (i.v.)	Brain/plasma	0.34	2.00	4.14	0.24	0.36

Table 11. Pharmacokinetic parameters of BDMC-PNIPAM after i.n. and i.v. administration to rats at the dose of 0.100 mg kg⁻¹ in brain, lungs and plasma (n = 6, mean ± SD).

Formulation administration	Samples	Cmax (ng/mL g)	Tmax	t1/2 (h)	Ke (h^−1)	AUC0–t (ng min/mL g)
BDMC (i.n.)	Brain	202.17 ± 12.06	2.00	20.90 ± 3.84	0.03317 ± 0.00548	2493.38 ± 243.51
	Lungs	23.67 ± 5.09	2.00	20.94 ± 6.78	0.03310 ± 0.02244	269.17 ± 76.08
	Plasma	55.67 ± 6.31	0.50	5.94 ± 1.13	0.11661 ± 0.02664	367.00 ± 93.10
BDMC (i.v.)	Brain	105.83 ± 17.41	2.00	7.61 ± 1.69	0.09106 ± 0.0087	1002.13 ± 104.71
	Lungs	176.50 ± 12.88	2.00	6.32 ± 1.93	0.10964 ± 0.0207	1547.13 ± 50.36
	Plasma	1315.83 ± 54.52	1.00	8.89 ± 1.83	0.09106 ± 0.0069	17918.13 ± 375.04
BDMC-PNIPAM (i.n.)	Brain	989.83 ± 20.38	2.00	189.37 ± 53.32	0.00366 ± 0.00104	15668.92 ± 813.51
	Lungs	112.50 ± 7.56	2.00	70.37 ± 62.56	0.00985 ± 0.00432	1787.92 ± 145.87
	Plasma	414.67 ± 10.05	2.00	27.51 ± 2.14	0.02520 ± 0.00210	6309.33 ± 164.92
BDMC-PNIPAM (i.v.)	Brain	408.33 ± 31.61	2.00	41.25 ± 10.58	0.01680 ± 0.00390	6338.79 ± 312.59
	Lungs	209.50 ± 6.50	2.00	77.44 ± 17.84	0.00895 ± 0.00248	3192.96 ± 122.03
	Plasma	1218.33 ± 11.71	1.00	9.54 ± 0.87	0.07262 ± 0.00671	17569.75 ± 284.71
BDMC (i.n.)	Brain/plasma	3.63	4.00	3.52	0.28	6.79
BDMC (i.v.)	Brain/plasma	0.08	2.00	0.86	1.00	0.06
BDMC-PNIPAM (i.n.)	Brain/plasma	2.39	1.00	6.88	0.15	2.48
BDMC-PNIPAM (i.v.)	Brain/plasma	0.34	2.00	4.32	0.23	0.36

Table 12. Drug targeting efficiency and direct nose-to-brain transport following i.n. administration of different formulations.

			Comparative bioavailability^a (AUCi.n./AUCi.v) (%)
Formulations	Drug targeting efficiency (%DTE)^a	Direct nose-to-brain transport (%DTP)^a	Blood	Brain
CUR	203.97 ± 56.43	51.01 ± 8.09	12.53 ± 2.16	38.96 ± 5.36
CUR-PNIPAM	659.23 ± 83.59	84.03 ± 1.81	36.71 ± 5.61	245.40 ± 11.92
DMC	201.26 ± 53.79	52.62 ± 7.01	12.35 ± 2.63	40.16 ± 4.41
DMC-PNIPAM	667.84 ± 85.12	85.23 ± 2.19	36.26 ± 5.49	246.70 ± 12.65
BDMC	205.40 ± 54.76	53.64 ± 7.46	13.09 ± 2.77	39.34 ± 4.69
BDMC-PNIPAM	677.12 ± 289.99	85.47 ± 2.49	35.91 ± 5.31	247.19 ± 12.81

^aParameters are derived using mean ± standard error of mean values of six different estimations.

Pharmacodynamic study in cerebral ischemia

Summarized results of pharmacodynamic studies based on behavioural test and quantitation of cytokine levels is given in Figure 9. MCAO rats showed poor neurological function as compared with the sham group. Data showed that treatment with NPs (CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM) significantly improved the neurological deficits as compared with curcuminoids treatment. The activity was found to be greatest in the case of CUR-PNIPAM followed by DMC-PNIPAM and BDMC-PNIPAM. Levels of TNF-α and IL-1β in the infracted and surrounding areas by enzyme-linked immunosorbent assay (ELISA) showed a significant increase in the level of TNF-α and IL-1β in the MCAO group as compared to the sham group (p < 0.001). Significant decrease in the levels of TNF-α and IL-1β was found after treatment with CUR-PNIPAM (p < 0.001), DMC-PNIPAM and BDMC-PNIPAM (p < 0.01) as compared with MCAO group. However, treatment with CUR, DMC and BDMC showed no significant alteration (Figure 9).

Figure 9. CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM pre-treatment for 21 days improves performances in the neurological deficits after stroke: (A) neurological scores and (B and C) quantification of TNF-α and IL-1β activity by ELISA in the sham, MCAO and CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM + MCAO groups. [***p < 0.001 (Sham vs MCAO); ##p < 0.01 and ###p < 0.001 (Treatment vs MCAO group)].

Toxicity study

There were no mortalities observed in any of the groups during the 14-day treatment period. Macroscopic examination of the brain sections as well as nasal mucosa exposed to PNIPAM-NPs, CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM did not show any change in the morphology or tissue microstructure as compared to normal control group rats. No visible sign of inflammation or necrosis revealed the safety of the developed formulation (Figure 10).

Figure 10. Representative photomicrographs showing the TS of rats' brain (A, B, C and D) and nasal mucosa (E, F, G and H) for normal control, PNIPAM-NPs (placebo), CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM-treated groups, respectively after 14 days.

Discussion

Addition of the initiator lead to the start of the polymerization of NIPAM monomer until reaches a certain critical length and finally it collapses giving rise to precursor particles at the temperature higher than LCST. The precursors particles then grow by either aggregation with the other precursor particles or are trapped by existing colloid stable particles Moreover, the average particle sizes of PNIPAM NPs is measured as a function of increasing temperature i.e. when the temperature is raised above the LCST, the polymer undergoes a phase transition and hydrophilic random coil state collapses to form a globular hydrophobic state (Xu et al., 2006). Increasing trend in the size of particles with increasing ratio of VP and AA suggests that copolymerization of NIPAM with the hydrophilic monomers (VP and AA) would have increased the LCST of the resultant copolymer leading to minimized rearrangement (Naha et al., 2010). Polymer shows amphiphilic character with a hydrophobic core inside the micelles and hydrophilic outer shell consisting of hydrated amides, pyrrolidone and carboxylic groups (Xu et al., 2006). Various concentrations of monomers were tried out of which formulation D3 with NIPAM:VP:AA:MBA ratio of 80:15:5:20 was selected for drug loading. The core of co-poly-(NIPAM–VP–AA) micelles serves as a non-aqueous reservoir for drugs where the drugs are stabilized against chemical modifications. Different drug polymer ratio (1:10, 2:10 and 3:10) produced NPs of different sizes. Particle size was found to be in direct variation with drug: polymer ratio. From some of the previous studies it can be concluded that effective brain targeting can be achieved when particle size is kept below 200 nm (Fazil et al., 2012; Haque et al., 2012; Md et al., 2012). So, we selected the formulations with ratio of 1:10, having particle sizes of <100 nm (CUR-NIPAM = 92.46 ± 2.8, DMC-PNIPAM = 91.23 ± 4.2 and BDMC-PNIPAM =94.28 ± 1.91 nm) for further studies. Moreover, these formulations were having PDI < 0.2 suggesting fairly uniform particle size distribution. NIPAM exhibits a very sensitive phase transition at ∼31–32 °C (LCST). Below LCST, the polymer is soluble in aqueous media, while it becomes opaque and forms precipitate above LCST (Ramírez-Fuentes et al., 2008). However, the addition of AA results in an increase in LCST (Ng et al., 2005) and the VP rendered a hydrogel and mucoadhesive properties of the polymer (Alsarra et al., 2011). Thus, addition of AA and VP increased the stability of the formulation as well as its retention at nasal mucosa. DSC findings shows diminished peaks as compared to pure curcuminoids which suggests that they are molecularly dispersed and present in amorphous form within the NPs (Joshi et al., 2010). Drug release profile obtained revealed that the developed NPs produced a sustained and enhanced release. Interestingly the difference in release profile with CUR, DMC and BDMC due to different aqueous solubility (CUR = 0.61, DMC = 0.42 and BDMC = 0.33 mg/mL) was also seemed to be minimized when incorporated in NPs. All the three formulation (CUR-PNIPAM, DMC-PNIPAM and BDMC-PNIPAM) showed a similar release pattern. This confirms the existence of a molecularly dispersed and amorphous nature of curcuminoids within the NPs. Significantly enhanced permeation was obtained with NPs as compared with curcuminoids. The hydrophobicity of curcuminoids would have led enhanced partitioning through the biological membrane and smaller size (<100 nm) of NPs could have been contributing factors in transport by via different trans- and para-cellular transport (Richter & Keipert, 2004). The biodistribution results showed that the drug concentration was highest in the brain via i.n. administration of NPs which is direct correlation with the enhanced permeation and existence of direct nose-to-brain transport bypassing the BBB. Moreover, the special mucoadhesive property of the developed NPs would have decreased mucociliary clearance (Wang et al., 2006). Interestingly, higher concentration in the brain was achieved even with i.v. administration of NPs as compared to pure curcuminoids (i.v). This showed the capability of NPs to cross BBB up to some extent (Tosi et al., 2007). Summarily, highest bioavailability in the brain with i.n. route might be the consequences three proposed pathways: (i) systemic pathway, by which some of the drug is absorbed into the systemic circulation and subsequently reaches the brain by crossing the BBB; (ii) olfactory and (iii) trigeminal neural pathway, by which drug travels directly to the CSF and brain tissue from the nasal cavity (Thorne et al., 2004). The determined pharmacokinetic parameters also supported superiority of nose-to-brain delivery of CUR, DMC and BDMC through NPs. Higher T_max value in brain (2 h) compared to blood (1 h) suggests preferential nose-to-brain transport following i.n. administration. Long T_max gives an added benefit in cerebral ischemia since concentration of ROS is gradually increased in the brain and by the use of NPs, curcuminoids concentration in brain can be maintained for at least 6 h. Similarly, C_max and AUC was significantly higher after i.n. administration of NPs as compared to i.v. route as well as pure curcuminoids (i.v. or i.n.) (Alsarra et al., 2011). The %DTE and %DTP represent the percentage of drug directly transported to the brain via the olfactory pathway. Higher %DTE and %DTP suggest that NPs have better brain-targeting efficiency through i.n. route (p < 0.001). Similar results have been reported by Zhang et al. (2004). Pharmacodynamically, MCAO rats have poor neurological function as compared with the sham group. Treatment with NPs improved the condition significantly. However, no significant alteration was observed in the curcuminoids treated group. This is in direct corroboration with enhanced bioavailability of curcuminoids via administration NPs through i.n. route. Pro-inflammatory cytokines TNF-α and IL-1β are reported be over expressed under ischemic conditions which is considered to be a detrimental in cerebral ischemia (Berger et al., 2002; Rezaie & Dean, 2002). Our results showed that curcuminoids NPs pre-treatment decreased the activity of these proteins upto significant levels. The short-term (14 days) toxicity studies, after repeated i.n. administration of the CUR, DMC and BDMC-loaded nanoformulations in rats caused no significant inflammation, or tissue toxicity proving the safety of developed brain-targeted formulations.

Conclusion

Enhancement of brain bioavailability of curcuminoids through PNIPAM-NPs for the effective treatment of cerebral ischemia through via direct nose-to-brain has been successfully established in this study. UPLC/ESI-Q-ToF-MS/MS-based bioanalytical method was developed, validated and successfully applied for the pharmacokinetic and biodistribution studies. Pharmacodynamic performance of the developed NPs was assessed on MCAO-induced cerebral ischemia model in rats based on neurological deficit, quantification of proinflammatory cytokines and histopathological studies. Moreover, toxicity studies revealed the safety of the developed nanoformulations. Encouraging results were obtained to conclude that CUR, DMC and BDMC-loaded NPs are novel, effective, non-invasive and safe brain-targeted drug delivery system for the treatment of cerebral ischemia.

Acknowledgements

The authors are thankful to Council for Scientific and Industrial Research (CSIR), Pusa Road, New Delhi, India, for the granting CSIR-SRF which supported this study.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.