Novel Cuminaldehyde Self‐Emulsified Nanoemulsion for Enhanced

Research Paper

Novel cuminaldehyde self-emulsiﬁed nanoemulsion for enhanced antihepatotoxicity in carbon tetrachloride-treated mice Michael Adu-Frimponga,b , Wei Qiuyua, Caleb Kesse Firempongc, Yusif Mohammed Mukhtara, Qiuxuan Yanga, Emmanuel Omari-Siawd, Zhen Lijuna, Ximing Xua and Jiangnan Yua aDepartment of Pharmaceutics and Tissue Engineering, School of Pharmacy, Jiangsu University, Zhenjiang, China, bDepartment of Basic and Biomedical Sciences, College of Health and Well-Being, Kintampo, Bono Region, cDepartment of Biochemistry and Biotechnology, College of Science, Kwame Nkrumah University of Science and Technology, and dDepartment of Pharmaceutical Sciences, Kumasi Technical University, Kumasi, Ghana

Keywords Abstract antihepatotoxicity; bioavailability; cuminaldehyde; self-emulsified Objectives Cuminaldehyde self-emulsified nanoemulsion (CuA-SEN) was pre- nanoemulsion pared and optimised to improve its oral bioavailability and antihepatotoxicity. Methods Cuminaldehyde self-emulsified nanoemulsion was developed through Correspondence the self-nanoemulsification method using Box–Behnken Design (BBD) tool while Ximing Xu and Jiangnan Yu, Department of appropriate physicochemical indices were evaluated. The optimised CuA-SEN Pharmaceutics and Tissue Engineering, School of Pharmacy, Jiangsu University, 301 was characterised via droplet size (DS), morphology, polydispersity index (PDI), Xuefu Road, 212001 Zhenjiang, China. zeta potential (ZP), entrapment efficiency, in-vitro release, and pharmacokinetic E-mails: [email protected] (X.X.) and studies while its antihepatotoxicity was evaluated. [email protected] (J.Y.) Key findings Cuminaldehyde self-emulsified nanoemulsion with acceptable characteristics (mean DS-48.83 1.06 nm; PDI-0.232 0.140; ZP- Received November 22, 2018 29.92 1.66 mV; EE-91.51 0.44%; and drug-loading capacity (DL)- Accepted April 22, 2019 9.77 0.75%) was formulated. In-vitro drug release of CuA-SEN significantly doi: 10.1111/jphp.13112 increased with an oral relative bioavailability of 171.02%. Oral administration of CuA-SEN to CCl4-induced hepatotoxicity mice markedly increased the levels of superoxide dismutase, glutathione and catalase in serum. Also, CuA-SEN reduced the levels of tumour necrosis factor-alpha and interleukin-6 in both serum and liver tissues while aspartate aminotransferase, alanine aminotransferase and malonaldehyde levels were significantly decreased. Conclusions These findings showed that the improved bioavailability of cuminaldehyde via SEN provided an effective approach for enhancing antioxidation, anti-inflammation and antihepatotoxicity of the drug.

[5,9–11] Introduction as antitumour, anti-inflammatory and antioxidant agents. Cumin [Cuminum cyminum Linn.], a common annual aro- Natural monoterpenoid, cuminaldehyde, has an aldehyde matic herb of the Apiaceae family,[1] is indigenous to India, group and isopropyl moiety at four positions (Figure 1b) Iran and Egypt, but cultivated broadly in areas such as Asia, and is a component of cumin oil[12] and oils such as eucalyp- North Africa, southern Russia and southern America.[2,3] tus, myrrh and cassia. The health-promoting effects of Historically, cuisines of diverse cultures and food industries cumin oil are attributed to cuminaldehyde,[13] which has are prepared with spicy seeds (Figure 1a) of cumin, which other functionalities such as anticancer,[14] antidiabetic,[15] comprised of fixed oil (about 10%), volatile oil (about neuroprotection[16] and anti-inflammation.[17] 1.5%), protein, sugar[4,5] and phenolic constituents.[6] Several conditions and xenobiotics can disrupt the nor- Cumin oil gives peculiar aroma and burning taste[7] to mal functioning of liver.[18] Common xenobiotic used to foods when used as flavour. Cumin oil possesses various mimic hepatotoxicity in animal models is carbon tetrachlo- pharmacological potentials such as digestive conditions, ride (CCl4) which induces oxidative stress, lipid peroxida- [8] [19,20] toothache, epilepsy and jaundice. It can also possibly act tion and inflammation. Notably, CCl4 induces injury

Figure 1 (a) Seeds of Cuminum cyminum Linn. and (b) chemical structure of cuminaldehyde (C10H12O; Molecular weight: 148.205 g/mol). [Col- our figure can be viewed at wileyonlinelibrary.com] via its metabolites (trichloromethyl) produced by cyto- lipophilic cuminaldehyde because it is easy to prepare and chrome P450 systems.[19] Treatment options which are also has the capacity to yield smaller globule size readily available, biocompatible and effective should be (<100 nm) with large surface area, ability to inhibit P-gly- developed for long-term supplementation without any tox- coprotein (P-gp)-mediated efflux of drugs and decrease icity concerns. Plant-based seed oils such as cumin oil,[9] first-pass gut metabolism.[35–37] Other SEN benefits include pomegranate oil[21] and grape oil[22] have been reported to increased GIT membrane permeability, efficient drug-load- exhibit antihepatotoxicity. Previous work suggested that ing and enhanced stability.[38–41] The SEN is a transparent cuminaldehyde could improve the activity of liver enzymes carrier which is stabilised by interfacial film of surfactant in vitro,[23] but, its in-vivo antihepatotoxicity has not been and cosurfactant molecules with the capacity to minimise reported yet. Like other natural essential oils (EOs), the use the volatility of some EOs.[42,43] of free cuminaldehyde usually leads to decreased bioactivity Recently, quality-by-design (QbD) approach has fre- because of low bioavailability caused by volatilisation, quently been applied to design and optimise nanofor- hydrophobicity and rapid enzymatic metabolisation.[24,25] mulations development on the basis of comprehensive Cuminaldehyde could therefore be formulated via appro- understanding of the association between in-process priate novel techniques to enhance its oral bioavailability variables for better product quality.[44,45] Hence, and therapeutic efficacy as other natural bioactive com- response surface methodology (RSM) tool like Box– pounds.[20,25,26] Several reports indicated that carriers such Behnken design (BBD) was employed in this study as liposomes, cyclodextrin inclusion complexes and drug- because it can aid in optimising formulation factors and in-cyclodextrin-in-liposome systems have been used to responses with relatively minimal experimental trials.[46] encapsulate various EOs and their components, viz. clove Accordingly, as an independent, spherically rotatable or EO, cinnamaldehyde, thymol, eugenol, trans-ferulic acid, almost rotatable quadratic design (with no embedded carvacrol, p-cymene, limonene and perillaldehyde.[27–33] factorial or fractional factorial design),[47] BBD is con- Despite these enormous data on the encapsulation of such sidered an appropriate design among all the RSM tools EOs and their related components in carrier systems, little requiring fewer runs in a three-factor/three-level experi- is known about cuminaldehyde-loaded formulations. Thus, mental design.[48] available reports can only support the preparation of cumin The current work sought to entrap cuminaldehyde in oil-loaded nanoemulsion and ayurvedic polyherbal formu- SEN to successfully deliver the drug in vivo for improved lation, which contained active ingredients such as trans- oral bioavailability and enhanced antioxidation, anti-in- anethole, limonene, cuminaldehyde and thymol.[9,34] How- flammation and antihepatotoxicity. ever, the in-vitro release, pharmacokinetic and pharmacodynamics evaluations of the oral preparations of individual Materials and Methods components of cumin oil-like cuminaldehyde have not been explored. In this regard, cuminaldehyde self-emulsi- Materials fied nanoemulsion (CuA-SEN) was developed, while its in- vitro release, pharmacokinetic and pharmacodynamics Cuminaldehyde (97% purity), the active pharmaceutical studies were fully investigated. The SEN system was there- ingredient (API), was bought from Aladdin Ind. Corp., fore selected to solve the low bioavailability problem of the (Shanghai, China) and BASF (Ludwigshafen, Germany)

© 2019 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 71 (2019), pp. 1324–1338 1325 Nanoemulsification of cuminaldehyde Michael Adu-Frimpong et al. provided Kolliphor-EL. Medium-chain triglyceride (MCT) the BBD with the required quantities of cuminaldehyde as and polyethylene glycol (PEG)-200 were received from the active ingredient, pure MCT as the oil phase, Kol- Sino-pharm Chem. Co. Ltd., (Shanghai, China). Nanjing liphor-EL as surfactant and PEG-200 as cosurfactant JianCheng Bioeng. Inst. (Nanjing, China) provided the (Table S1). The CuA-SEN preconcentrate was then slowly enzyme-linked immunosorbent assay (ELISA) kits. Jiangsu mixed with double distilled water (DDW) and continu- University’s Animal Laboratory Center supplied the male ously stirred at 750 rpm with a magnetic stirrer at 25 °C. Sprague–Dawley (SD) rats (200 20 g) and Kunming The optimised CuA-SEN was also prepared by mixing mice (18–20 g). All the other materials/reagents were of cuminaldehyde and MCT (1 : 1, v/v), Kolliphor-EL pharmaceutical grade. (34.26% v/v) and PEG-200 (17.84% v/v) together, and the mixture subsequently treated as indicated earlier. Besides, the blank SEN (without cuminaldehyde) was fabricated via Methods the mixing of MCT (48.26% v/v), Kolliphor-EL (34.26% v/ v) and PEG-200 (17.84%) together. Box–Behnken design Response surface methodology–Box–Behnken design Percentage transmittance determination (RSM-BBD) tool was used for optimisation of CuA-SEN fabrication using three coded factors (Cuminaldehyde/ The %T of the CuA-SEN was determined based on earlier MCT ratio-X1, Kolliphor-EL concentration-X2 and PEG- reports.[50] The mixtures were diluted in DDW (1 : 100 v/ 200 concentration-X3) in 1; 0; +1 levels (Table 1) on dro- v) to obtain a fine nanoemulsion. Using DDW as blank, the plet size (DS, Y1) and percentage transmittance (%T, Y2). %T of the developed CuA-SEN was measured after 2 h via Design-Expert 11 software (Stat-Ease Inc., Minneapolis, UV-160A Shimadzu spectrophotometer (Japan) at MN, USA) was applied to design random ordered 17 trials 638.2 nm.[51] The entire measurements were repeated in (Table S1). The results were fitted with quadratic equa- triplicate. tion as follows: FT-IR characterisation of optimised CuA-SEN Xk Xk Xk ¼ a þ a þ a 2 þ a þ ð Þ Y 0 iXi iiXi ijXiXj e 1 The FT-IR spectra of free cuminaldehyde, blank SEN (with- ¼ ¼ i 1 i 1 1 i j out cuminaldehyde) and SEN (with cuminaldehyde) were where a0 denotes constant and ai, aii, and aij, respectively, recorded via potassium bromide (KBr) disc method with represent the coefficients of linear, quadratic and interac- Nicolet 170SX, (Thermal Fisher Scientific, Madison, WI, tions terms, while Xi, Xj and Xk denote the independent USA). The samples (4 mg) and KBr (IR grade, 200 mg) variables. The k represents the number of variables while were gently mixed in a ceramic mortar and compacted to Y is the predicted response.[49] obtain a disc. The discs were scanned within wave number range of 400–4000 cm 1 [52] at 4 scans/s with 1 cm 1 reso- lution. Preparation of CuA-SEN

The CuA-SEN was prepared based on spontaneous Morphological characterisation nanoemulsiﬁcation method described in previous reports.[42] The different CuA-SEN was also prepared using The optimised CuA-SEN was diluted with DDW (1 : 100 v/v) and mixed via mild shaking. Excess CuA-SEN was wiped out with ﬁlter paper after a drop was placed on cop- – Table 1 Coded independent variables investigated using Box Behn- per-grids. A 1% phosphotungstate-based acid solution was ken design model applied in staining the sample for 30 s. Transmission-elec- Coded levels tron microscopy (TEM; JEM_2100; JEOL, Tokyo, Japan) Low High was utilised to observe the morphology of CuA-SEN. level Intermediate level Factor Independent variables (1) level (0) (+1) DS, PDI and ZP analysis

X1 Cuminaldehyde/MCT 0.50 1.0 1.50 ratio (v/v) Laser scattering Zetasizer (Malvern Inst. Ltd., Worcester-

X2 Kolliphor-EL 25.0 30.0 35.0 shire, UK) was used to ascertain the DS, PDI and ZP as [39,50] concentration (% v/v) reported in earlier works. Diluted CuA-SEN (in X3 PEG-200 10.0 15.0 20.0 DDW, 1 : 100 v/v) was used to measure the DS, ZP and concentration (% v/v) PDI in triplicates.

Entrapment efficiency and drug loading Q ¼ kt þ Q0 zero order ð4Þ [53] Based on previous reports, the entrapment efficiency kt Q ¼ Q0e First order ð5Þ (EE) and drug loading (DL) of optimised CuA-SEN were Q ¼ ktn Korsmeyer-Peppas ð6Þ determined. Ultrafiltration method (3500 Da) was used to isolate the unincorporated cuminaldehyde via centrifugation Y ¼ a linx þ b ð7Þ (speed of 805 g) for 5 min. The cuminaldehyde content was Q ¼ 100ð1 exp½ðt TÞb=aÞ Weibull ð8Þ measured with HPLC under the following conditions: C18 column (Zorbax ODS, 5 lm, 150 9 4.6 mm I.D; Waters, = = Q1 3 ¼ kt þ Q1 3 Hixson-Crowell ð9Þ Dublin, Ireland), mobile phase-acetonitrile/DDW 0 : (50 : 50-v/v); detection wavelength-260 nm; flow rate- Q ¼ kt0 5 Higuchi ð10Þ 1.0 ml/min; retention time 9.1 min (Figure S1a). Other Q tn Q ¼ max Hill equation ð11Þ details of standard curve determination can be found in kn þ tn Appendix S1. The equations for calculating EE and DL where Q = amount (%) of drug released at the time t, were as follows: Q0 = Q start value, Qmax = maximum value of Q (100%), Mt Mf = EE ¼ 100 ð2Þ T lag time for the dissolution process and k, a, b: con- Mt stants. A nonlinear least-squares regression was performed via the KinetDS 2.0 software[59] while the R2 and Akaike’s M M [60] DL ¼ t f 100 ð3Þ Information Criterion (AIC) were estimated. A model Mc usually adjusts the data set better when the values of AIC and R2 are, respectively, smaller and higher. where Mt = the total quantity of cuminaldehyde in CuA- SEN, Mf = the quantity of free cuminaldehyde and Pharmacokinetic study Mc = the total ingredients in CuA-SEN. The Animal Ethics and Experimental Committee (SCXK (SU) 20130036 approved on March 2013) of Jiangsu Stability study University approved the protocol in accordance with the The optimised CuA-SEN was stored in airtight container at National Health Guide Institute for Laboratory Animals’ 25 2 °C/65 5% relative humidity (RH) and Care and Use. The animals were kept (1 week) at ° – 40 2 °C/75 5% RH for 3 months as indicated in 25 2 C (RH-55 5%) and a 12 : 12-h light dark cycle some reports.[50] On the first, second and third month of with unrestricted access to water and food. Before the drug storage, the physical appearance, DS, PDI and ZP were administration, the animals were fasted overnight with free assessed. access to water except food. Cuminaldehyde (suspended in DDW, 200 mg/kg) and CuA-SEN (4 mg/ml, dispersed in DDW at the same dose as cuminaldehyde) were orally In-vitro drug release given to the rats in the control and tested groups, respectively. The in-vitro drug release (IVDR) study of free cuminalde- The bioavailability of CuA-SEN was assessed in SD rats hyde and CuA-SEN was based on dialysis bag method.[54] based on earlier reports.[50,61,62] The amount of cuminalde- The dissolution media (900 ml each), viz. simulated-gastric hyde in circulation with time was ascertained using the fluid (SGF, pH 1.2) and phosphate-buffer solution (PBS, standard concentration–time curve of the plasma. The pH 7.4), was held at 37 0.5 °C and 100 r/min.[55,56] The pharmacokinetic variables were computed with PKSolver free cuminaldehyde (5 mg/ml, 1 ml) and the CuA-SEN 2.0. software (AGAH Working group PK/PD mod- (equivalent to cuminaldehyde alone, 1 ml) were loaded elling).[63] The best fit model was selected based on smaller into the membranes. At predetermined time points (5; 10; value of AIC and higher R-squared.[60,64] The oral relative 15; 30; 45; 60; 90; 120; 150 and 180 min), samples (5 ml bioavailability (RBA) of CuA-SEN to the free cuminalde- each) were withdrawn and replaced with the same volume hyde was computed as follows: of fresh media. The amount of cuminaldehyde released was determined via HPLC method described earlier. Different AUC models of release kinetics (viz. zero order, first order, Kors- RBA ¼ CuA SEN 100 ð12Þ AUC meyer-Peppas, Y = a*lnx + b, Weibull, Hixson-Crowell, Cuminaldehyde Higuchi and Hill equations) were applied to fit the IVDR Other details of the pharmacokinetic study can be found data.[57–59] in Appendix S1.

In-vivo evaluation of antioxidation, anti-inflam- indicated the adequacy of BBD for this study. The suitabil- mation and antihepatotoxicity ity is further supported by the observed lack of fit (Table S2), which suggested that the chosen quadratic Kunming mice (48) were divided randomly into six groups regression model was suitable for response surface analysis (n = 8), viz. normal control (NC), model control (MC, (P > 0.05). Based on the ANOVA result, it was observed CCl -treated group), positive control (PC), free cuminalde- 4 that Kolliphor-EL concentration was significant (P < 0.05) hyde (FC), CuA-SEN (CS) and blank SEN (BS). The NC in linear (X ), quadratic (X ) and interaction (X X ) terms and MC groups were given water through intragastric 2 2 2 3 for both DS (Y ) and %T (Y ) determinations (Table S2). route, while the PC group received silibinin (100 mg/kg 1 2 Likewise, PEG-200 concentration was also statistically sig- per day suspended in 0.5% CMC-Na). The mice in the FC, nificant in linear (X ), quadratic (X ) and interaction CS and BS groups were each treated with 200 mg/kg per 3 3 (X X ) for DS (Y ) and %T (Y ). Thus, interaction between day (based on previous report[12] and preliminary study 2 3 1 2 Kolliphor-EL and PEG-200 could influence both DS and % data not shown) of free cuminaldehyde, CuA-SEN and T of CuA-SEN (Figure 2). The CuA-SEN formed from blank SEN with all samples dispersed in DDW, respectively, MCT, Kolliphor-EL and PEG-200 had smaller nanosized for 7 consecutive days on the account of earlier works.[65,66] droplets, which could improve permeability and uptake of ELISA kits were used to determine the biochemical mark- drugs sensitive to P-gp mediated efflux,[51] as well as ers, viz. ALT, AST, CAT GSH, MDA, SOD TNF-a and IL-6 enhance absorption of aquaphobic drugs at the intestinal (for other details refer to Appendix S1). lymph and hepatic portal blood.[50] Cosurfactants usually serve as second surface active agent for optimum formation Statistical analysis of nanoemulsion.[68] The choice of the excipients for CuA- SEN was supported by previous investigations of perillalde- Data are presented as mean SD. The significant differ- hyde (PAH)-loaded nanoemulsion.[50] Therefore, the two ences within the groups were analysed using ANOVA. A P non-ionic compounds were chosen for incorporating cumi- value of <0.05 was considered as an acceptable level for sta- naldehyde in SEN. tistical significance. The Graph-pad prism 6.00 (Graph-pad The cuminaldehyde/MCT ratio was significant for linear software Inc., San Diego, CA, USA) and Origin Pro version (X ) and quadratic (X ) terms (Table S2). Three-dimen- 2018 (OriginLab Corporation, Northampton, MA 01060, 1 1 sional plots for responses Y and Y were constructed (Fig- USA) were used to plot the graphs. 1 2 ure 2a–2d) to investigate the interactive effects of factors on responses. An optimal formulation was prepared with Results and Discussion optimised conditions, viz. cuminaldehyde/MCT ratio (1.00), Kolliphor-EL concentration (34.26%) and PEG-200 Box–Behnken design of experiment concentration (17.84%) with a computed composite desir- ability index of 0.9360. Through one-factor design experiment, drug/oil ratio, surfactant and cosurfactant concentrations were found to influence DS (Y1) and %T (Y2) of CuA-SEN. BBD design Characterisation of CuA-SEN was therefore utilised to select optimal conditions.[67] As shown in Table S1, the response values for Y1 and Y2 FT-IR characterisation of optimal CuA-SEN obtained from experimental results ranged from 25.93 to The possible interaction between pure drugs and excipients 96.41 nm and 31.59–94.95%, respectively. The predicted could normally be checked with FT-IR.[69] The FT-IR spec- responses values were fitted to the following quadratic tra of free cuminaldehyde, blank SEN and CuA-SEN are equations using the coded variables; depicted in Figure 3a–3c. The FT-IR spectrum of free cuminaldehyde (Figure 3a) was characterised by primary Y1 ¼34:91 þ 25:13X1 11:16X2 þ 10:17X3þ absorption peaks at 2964.70 (–C–H– stretching vibration of 13:69X2 þ 10:74X2 þ 16:70X2 1:83X X ð13Þ 1 2 3 1 2 isopropyl group), 1701.85 (aromatic C=C stretching vibra- : þ : 2 18X1X3 4 88X2X3 tion), 1607.71 (–C=O– stretching frequency) and 833.65 (aromatic C–H bending) cm 1.[70–72] The blank SEN (Fig- ¼ : : þ : : Y2 88 49 19 92X1 14 65X2 5 70X3 ure 3b) showed broad absorption at 3439.98, 2927.15, : 2 : 2 : 2 þ : ð Þ 1 11 56X1 11 43X2 13 35X3 2 23X1X2 14 1740.70 and 1109.66 cm . Upon incorporation of cumi- þ 3:38X1X3 4:09X2X3 naldehyde into SEN (Figure 3c), the characteristic absorption peaks of cuminaldehyde at 1701.85, 1607.71 and The high R2 values (Table S2) obtained for the responses 833.65 cm 1 were retained with broad bands of the excipi- 2 (R = 0.9889 and 0.9899 for Y1 and Y2, respectively) ents while other peaks of the pure drug disappeared. The

Figure 2 3D response surface plot depicting (a) the effect of Kolliphor-EL concentration and cuminaldehyde/medium chain triglyceride (CuA/ MCT) ratio on droplet size (DS), (b) the effect of polyethylene glycol 200 (PEG-200) concentration and CuA/MCT ratio on DS, (c) the effect of PEG-200 concentration and Kolliphor-EL concentration on DS, (d) the impact of CuA/MCT ratio and Kolliphor-EL concentration on percentage transmittance (%T), (e) the impact of cuminaldehyde/MCT ratio and PEG-200 concentration on %T, and the impact of Kolliphor-EL concentration and PEG-200 concentration on %T. [Colour figure can be viewed at wileyonlinelibrary.com] disappearance of characteristic peaks of cuminaldehyde spontaneous production of emulsions with smaller DS and might be ascribed to the overlapping of drug peaks with stability.[76] Other indices such as EE and DL which were that of the excipients.[61] also significant parameters of SEN may be used to evaluate the efficiency of drug loading into nanocarriers.[77] The higher EE (91.51 0.44%) and DL (9.77 0.75%) values Physical characteristics of optimal CuA-SEN of cuminaldehyde entrapment in nanoemulsion indicated a The optimal CuA-SEN was morphologically homogenous better loading capability. These results might be due to the and spherical or elliptical (Figure 3d) in shape[36,73] with extreme lipophilicity of cuminaldehyde, which may result the absence of aggregation when observed under TEM. The in the internalisation of cuminaldehyde in the aquaphobic average DS, PDI and ZP produced were 48.83 1.06 nm, core of the formulation. 0.232 0.14 and 29.92 1.66 mV (Table 2), respectively. The smaller DS of CuA-SEN could be ascribed to the Stability study initial dispersion of the surfactant in the oil phase before the addition of the aqueous phase.[74] Besides, the low PDI The physical appearance, DS, PDI value and ZP of the value suggested possible narrow and homogenous DS dis- blank SEN and optimised CuA-SEN were determined dur- tribution (Figure 2b), while the high ZP indicated stable ing the stability study. At 25 °C/65% RH, the DS of CuA- nanosized globules with low possibility of conglutination SEN showed a slight insignificant increase from 48.83 to attributable to sufficient repulsion.[75] The negative value 51.05 nm (P > 0.05), which indicated a physically stable of ZP might be ascribed to the unbound fatty acids in the nanoemulsion for the 3-month storage period (Table 2). excipients of SEN preconcentrate. The varying hydrocarbon Likewise, the DS of CuA-SEN fairly increased from 49.60 to chain lengths of fatty acids in oil, surface-active agent and 52.32 nm upon storage at 40 °C/75% RH (P > 0.05). The the degree of unsaturation play crucial roles in the PDI and ZP of CuA-SEN also varied minimally after

Figure 3 Fourier transform infrared (FT-IR) spectra of (a) free cuminaldehyde (b) blank SEN, (c) CuA-SEN and characterization of optimized CuA- SEN (d) TEM image of CuA-SEN (e) size distribution of CuA-SEN. [Colour ﬁgure can be viewed at wileyonlinelibrary.com]

Table 2 Storage stability of blank and optimised cuminaldehyde self-emulsiﬁed nanoemulsion dispersions at different temperatures over 3-month study period

Measurement Physical Storage temperature Droplet size Formulations time (month) appearance (°C) (nm) PDI ZP (mV)

CuA-SEN 1st Clear, homogenous 25 2 °C/65 5% RH 48.83 1.06 0.232 0.14 29.92 1.66 Clear, homogenous 40 2 °C/75 5% RH 49.60 1.20 0.196 0.01 28.52 1.72 2nd Clear, homogenous 25 2 °C/65 5% RH 50.99 1.11 0.222 0.021 27.26 1.20 Clear, homogenous 40 2 °C/75 5% RH 51.05 1.25 0.201 0.15 27.90 2.94 3rd Clear, homogenous 25 2 °C/65 5% RH 51.44 0.59 0.227 0.07 29.47 2.34 Clear, homogenous 40 2 °C/75 5% RH 52.32 1.28 0.262 0.05 26.36 5.62 Blank SEN 1st Clear, homogenous 25 2 °C/65 5% RH 35.92 4.39 0.115 0.021 48.12 5.15 Clear, homogenous 40 2 °C/75 5% RH 36.27 2.36 0.154 0.05 47.80 2.87 2nd Clear, homogenous 25 2 °C/65 5% RH 36.51 1.92 0.160 0.07 47.28 4.45 Clear, homogenous 40 2 °C/75 5% RH 37.55 2.13 0.205 0.08 46.25 0.97 3rd Clear, homogenous 25 2 °C/65 5% RH 38.07 1.41 0.249 0.06 46.09 0.66 Clear, homogenous 40 2 °C/75 5% RH 40.04 2.36 0.192 0.04 45.68 0.98

storage at the same climatic conditions (Table 2). Gener- the stable optimal CuA-SEN which could effectively deliver ally, CuA-SEN is classiﬁed as thermodynamically stable cuminaldehyde to the target site. with no signs of phase-separation, haziness or precipitation.[61] Thus, the stability of cuminaldehyde can be In-vitro drug release studies enhanced with SEN nanocarrier. Similarly, the DS of the blank SEN demonstrated statistically non-signiﬁcant In general, the highest drug release was obtained from changes from 35.92 to 38.07 nm (25 °C/65% RH) and CuA-SEN in comparison with the free cuminaldehyde in 36.27 to 40.04 nm (40 °C/75% RH). The high ZP (45.68 suspension over the 180 mins irrespective of the media (pH to 48.12 mV) of the blank SEN showed that the excipient 1.2 and pH 7.4 buffers; Figure 4a). The release phe- mixture was stable. These results might have contributed to nomenon might be ascribed to the small DS of the

1330 © 2019 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 71 (2019), pp. 1324–1338 Michael Adu-Frimpong et al. Nanoemulsification of cuminaldehyde nanoemulsion which is significant for improving in-vivo equations. The release data of the kinetics of CuA-SEN fit- absorption.[50] The release of cuminaldehyde from the drug ted well with the Korsmeyer-Peppas equation (PBS, suspension was markedly lower (P < 0.0001) with only R2 = 0.8284 and AIC = 19.70; SGF, R2 = 0.8317 and 33.27% (SGF) and 36.93% (PBS) drug release in 180 min. AIC = 18.97), which might indicate non-Fickian diffu- The CuA-SEN demonstrated initial quick release of about sional release mechanism. These findings highly agreed 42.34% (SGF) and 44.02% (PBS) within 30 min which with other related reports.[81,83] might be owing to the release of unincorporated cuminaldehyde.[78] This phenomenon supported the slow and In-vivo pharmacokinetic study continuous release of the encapsulated drug reaching a maximum of 75.51% (SGF) and 81.98% (PBS) at the end The concentration–time curve of CuA-SEN in plasma was of 180 min (Figure 4a). The CuA-SEN release in PBS was compared with free cuminaldehyde (Figure 4b). The slightly higher than SGF but it was statistically insignificant. absorption behaviour of cuminaldehyde from CuA-SEN This result implied that the pH of the dissolution medium was determined using pharmacokinetic parameters, and the might not have affected the release pattern of cuminalde- data were fitted to one-compartmental model (Table 3). hyde. Substituted benzaldehydes like cuminaldehyde usu- The CuA-SEN demonstrated significantly higher plasma ally undergo hydrolysis to stably soluble benzoic acid or drug curve than the free cuminaldehyde. The Cmax of CuA- conjugate base in acidic or basic conditions.[79] Ideally, SEN (7.13 lg/ml) was 1.5-fold greater than the free drug every formulation seeks to achieve 100% dissolution in (4.63 lg/ml), suggesting significant improvement of the release media; however, approximately 82% (PBS) release bioavailability of cuminaldehyde. Similarly, the Tmax of suggested a significantly improved solubility of cuminalde- CuA-SEN (2.11 h) was obviously longer than the free hyde compared with <30% release of the free drug. The cuminaldehyde (1.85 h). The AUC(0∞) for CuA-SEN data corroborated the proposed 75% accumulative release (41.49 lg h/ml) was also greater than the suspended free for immediate dissolution formulations.[56] drug (24.26 lg h/ml). Additionally, the higher MRT of In order to decrease the deviations of theoretical values CuA-SEN (4.44 h) compared with the free drug (3.85 h) from that of the practical values,[80] the data obtained from suggested a sustained release of the cuminaldehyde from the in-vitro release of CuA-SEN in both media were fitted the formulation. These results implied that the encapsu- to eight mathematical models, viz. zero order, first order, lated drug might be rapidly absorbed but slowly eliminated Korsmeyer-Peppas, Weibull’s equations Y = a*lnx + b, from systemic circulation than the free cuminaldehyde. Hixson-Crowell, Higuchi and Hill equations (Table S3). This phenomenon contributed to the high RBA of CuA- The most appropriate drug release model was selected SEN (171.02%) relative to the free cuminaldehyde. The based on the best-fitted model with smaller AIC value and CuA-SEN favoured rapid absorption probably owing to the higher R2.[81] Previous report[82] has indicated that experi- polar vehicles (MCT, Kolliphor-EL and PEG-200) utilised mental data fitting can simply be done with empirical in the nanoformulation. Previous report has also posited

Figure 4 (a) In-vitro release profiles of free cuminaldehyde and CuA-SEN (mean SD, n = 6) and (b) plasma concentration–time profiles of free cuminaldehyde and CuA-SEN (mean SD, n = 3) at a dose of 200 mg/kg. [Colour figure can be viewed at wileyonlinelibrary.com]

© 2019 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 71 (2019), pp. 1324–1338 1331 Nanoemulsification of cuminaldehyde Michael Adu-Frimpong et al. that carvone, a dietary monoterpene-like cuminaldehyde, In-vivo evaluation of antioxidation, anti-inflam- was quickly absorbed in vivo with limited detection in mation and antihepatotoxicity human plasma because of rapid biotransformation.[84] The outcomes of the current report showed that the admin- However, there is lack of data on the bioavailability of istration of CCl via oral route induced elevated oxidative cuminaldehyde. Therefore, the present study provides pre- 4 stress, inflammation and concomitant hepatotoxicity. CCl liminary result on cuminaldehyde bioavailability enhance- 4 induces hepatotoxicity via metabolites such as trichloro- ment. ˙ ˙ methyl ( CCl3) and trichloromethyl peroxy ( OOCCl3), which can covalently bind to macromolecules such as pro- teins, lipids and nucleic acids. Disruption of structural Table 3 Plasma pharmacokinetic parameters of free cuminaldehyde integrity of liver cells results in an increased release of hep- and cuminaldehyde self-emulsified nanoemulsion in SD rats atic enzymes (AST and ALT) from the cytoplasm into cir- (mean SD, n =3)aftertheoraladministrationatdoseof200mg/kg culation.[85,86] Silibinin and optimised CuA-SEN Pharmacokinetic Free significantly (P < 0.05) increased the levels of SOD, GSH parameters cuminaldehyde CuA-SEN and CAT and reduced MDA in rat plasma compared with the free cuminaldehyde (Figure 5a–5d). The outcome indi- Cmax (lg/ml) 4.63 0.09 7.13 0.17*** cated that the cuminaldehyde possesses antioxidant activity Tmax (h) 1.85 0.07 2.11 0.01* [87] AUC0∞ (lg h/ml) 24.26 0.04 41.49 0.13*** which corroborated other studies. The result also t1/2 (h) 1.45 0.06 1.89 0.08* implied that the nanoformulation of cuminaldehyde could t1/2 Ka (h) 1.23 0.07 1.18 0.04 potentially protect vital organs like the liver from oxidative l Cl ((mg)/( g/ml)/h)) 8.26 0.09*** 4.82 0.09 stress-related injury. Hence, enhanced antioxidation of MRT (h) 3.85 0.05 4.44 0.14* CuA-SEN might be due to the improved bioavailability of RBA (%) 171.02 cuminaldehyde.[88] AUC, area under the curve; Cmax, peak drug concentration; Cl, clear- ance; T , time to reach the peak concentration; MRT, mean resi- The aromatic ring of cuminaldehyde provides its ability max [89] dence time; t1/2, half-life; abs, absorption half-life; RBA(%), to scavenge ROS, viz. hydroxyl radical, hydrogen perox- [90] [91] percentage relative bioavailability. *P < 0.05, **P < 0.001 and ide, superoxide anion and metal ion chelants. The ***P < 0.0001, compared with free cuminaldehyde. antioxidation of cuminaldehyde is attributable to the

Figure 5 Effect of CuA-SEN on hepatic antioxidant markers in CCl4-induced hepatotoxicity. Normal control (NC), model control (MC, CCl4-treated group), positive control (PC, CCl4 + silibinin), free cuminaldehyde group (FC, CCl4 + free cuminaldehyde), CuA-SEN (CS, CCl4 + CuA-SEN), and blank SEN group (BS, CCl4 + Blank SEN). (a) Superoxide dismutase (SOD), (b) glutathione (GSH), (c) catalase (CAT), and (d) malondialdehyde (MDA), *P ˂ 0.05, **P ˂ 0.001 and ***P ˂ 0.0001 ANOVA, compared with the MC group.

Figure 6 Effect of CuA-SEN on hepatic and serum pro-inﬂammatory mediators in CCl4-hepatotoxic mice. Normal control (NC), model control (MC, CCl4-treated group), positive control (PC, CCl4 + silibinin), free cuminaldehyde group (FC, CCl4 + free cuminaldehyde), CuA-SEN (CS, CCl4 + CuA-SEN) and blank SEN group (BS, CCl4 + Blank SEN). IL-6, interleukin-6; TNF-a, tumour necrosis factor-alpha, *P ˂ 0.05 and ***P ˂ 0.0001 ANOVA, compared with the model group.

Figure 7 Effect of CuA-SEN on serum ALT and AST levels in CCl4 intoxicated mice. Normal control (NC), model control (MC, CCl4-treated group), positive control (PC, CCl4 + silibinin), free cuminaldehyde group (FC, CCl4 + free cuminaldehyde), CuA-SEN (CS, CCl4 + CuA-SEN) and blank SEN group (BS, CCl4 + Blank SEN). (a) Alanine aminotransferase (ALT), (b) aspartate aminotransferase (AST) and (c) effect of CuA-SEN on liver weight-to-body weight ratio. *P ˂ 0.05 and ***P ˂ 0.0001 ANOVA, compared with the MC group.

© 2019 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 71 (2019), pp. 1324–1338 1333 Nanoemulsification of cuminaldehyde Michael Adu-Frimpong et al. aldehyde functional group (C=O–H), which can donate a and 7b) might be ascribed to the small DS of the nanofor- ˙ ˙ hydrogen atom to the CCl3 and OOCCl3 radicals. Cumi- mulation and improved bioavailability. Thus, the formula- naldehyde then exists in a radical form with delocalised tion improved the pharmacological potentials of the drug benzene ring to produce a stable resonance. This stabilises via prolonged antioxidation, anti-inflammation and anti- free radicals via electron conjugation culminating in pre- hepatotoxicity. The results of the biochemical analysis were vention of the radicals from participating in destruc- also supported by the normalised liver weight-to-body tive biochemical reactions.[92] weight ratio (Figure 7c) and the histopathological investi- ˙ Pro-inflammatory mediators’ (TNF-a and IL-6 levels) gations (Figure 8d and 8e) of the rats. The CCl3 and ˙ levels in serum and liver tissues were significantly reduced OOCCl3 radicals generated via the activation of CCl4 by (P < 0.0001, hepatic IL-6 and TNF-a as well as serum IL-6 the cytochromes (CYP2E1, CYP2B1 or CYP2B2) can and TNF-a) in the treated groups compared with the impair essential cellular processes like lipid metabolism model group (Figure 6a–6d). Likewise, the cytokines levels with fatty degeneration (steatosis) as the possible out- in serum and liver tissues were significantly ameliorated come.[93] Usually, treatment with antioxidants ameliorates (P < 0.0001) in CuA-SEN and silibinin groups than the the toxification process of the radicals and concomitant free cuminaldehyde (Figure 6a–6d). This observation sug- fatty degeneration. From Figure 8e, it can be observed that gested that formulation of cuminaldehyde enhanced its treatment with cuminaldehyde self-emulsified nanoemul- anti-inflammatory activity which might play key role in sion (CuA-SEN) significantly improved histopathological attenuating CCl4-induced hepatotoxicity in mice. findings albeit some fatty degeneration occurring which The significantly reduced ALT and AST contents of could possibly be attributed to the cytosolic lipid droplets CuA-SEN compared with free cuminaldehyde (Figure 7a usually present in healthy liver tissues.[94] Besides, the fatty

(a) (b)

(e)

Figure 8 Histopathological changes in mice livers stained with H & E. (a) Normal control depicting normal hepatic architecture, (b) mice supplemented with 2% CCl4 (10 ml/kg body weight) and showing the presence of fatty degeneration, cell necrosis and inﬂammations, (c) animals pre- treated with silibinin (100 mg/kg per day) and then later treated with 2% CCl4 (10 ml/kg of body weight), (d) mice supplemented with free cuminaldehyde (200 mg/kg per daily) and 2% CCl4 (10 ml/kg of body weight) and (e) mice supplemented with CuA-SEN (200 mg/kg per daily) and 2% CCl4 (10 ml/kg of body weight). CV, central vein, HCN, hepatic cell necrosis, H, hepatocytes, S, sinusoid. [Colour ﬁgure can be viewed at wileyonlinelibrary.com]

1334 © 2019 Royal Pharmaceutical Society, Journal of Pharmacy and Pharmacology, 71 (2019), pp. 1324–1338 Michael Adu-Frimpong et al. Nanoemulsification of cuminaldehyde degeneration could be due to potential mild signs of toxic- Declarations ity on liver by higher doses of SEN nanocarriers as reported by others.[95] Therefore, future research will comprehen- Conflict of interest sively investigate the possible acute and subchronic toxicity of the CuA-SEN in experimental animals. Overall, these There is no conflict of interest. findings suggested that the optimised CuA-SEN exhibited remarkable antihepatotoxicity against CCl4-induced liver Acknowledgement damage. This work was supported by the National Natural Science Foundation of China [Grant 81373371], and China Post- Conclusion doctoral Science Foundation [2014M560410 and Cuminaldehyde self-emulsified nanoemulsion with improved 2014M560409]. We also express our sincere thanks to the bioavailability and antihepatotoxicity was successfully Jiangsu University Ethics Committee for their kind guid- designed and optimised. The optimised CuA-SEN composed ance in the mice experiments. of cuminaldehyde/MCT ratio as oil phase (cuminaldehyde/ MCT ratio 1.00 v/v), Kolliphor-EL concentration 34.26% v/v Author contributions and PEG-200 17.84% v/v. The CuA-SEN showed elliptical or spherical and homogenous droplets with the absence of Michael Adu-Frimpong, Zhen Lijun and Ximing Xu aggregation and acceptable physical characteristics. CuA-SEN designed the experiments and acquired the data. Qiuxuan showed a marked increase in the rate of in-vitro release and Yang, Yusif Mohammed Mukhtar and Qiuyu Wei analysed enhanced relative oral bioavailability. The optimised CuA- data while Michael Adu-Frimpong, Qiuyu Wei and Jiang- SEN exhibited substantial higher serum levels of GSH, SOD, nan Yu drafted the manuscript. Caleb Kesse Firempong CAT and TNF-a plus IL-6 (serum and liver tissues) along and Emmanuel Omari-Siaw revised the manuscript for sig- with reduced AST, ALT and MDA contents were observed. nificant intellectual content, while Jiangnan Yu and Ximing Thus, improving bioavailability of cuminaldehyde via SEN Xu approved the final version of the manuscript. All the could be a promising tool for enhancing the antioxidation, authors reviewed the manuscript. anti-inflammation and antihepatotoxicity of the drug.

References oil: a high effectiveness against Vibrio cuminaldehyde, cuminol and an inhi- spp. strains. Food Chem Toxicol 2010; bitor isolated from Cuminum cymi- 1. Al Juhaimi FY, Ghafoor K. Extraction 48: 2186–2192. num L. in streptozotocin-induced optimization and in vitro antioxidant 6. Rebey IB et al. Effect of drought on diabetic rats. Br J Nutr 2013; 110: properties of phenolic compounds the biochemical composition and 1434–1443. from cumin (Cuminum cyminum L.) antioxidant activities of cumin (Cumi- 11. Einafshar S et al. Antioxidant activity seed. Int Food Res J 2013; 20: 1669– num cyminum L.) seeds. Ind Crop of the essential oil and methanolic 1675. Prod 2012; 36: 238–245. extract of cumin seed (Cuminum cym- 2. Motamedifar M et al. The effect of 7. Mandal M, Mandal S. Cumin (Cumi- inum L.). Eur J Lipid Sci Technol cumin seed extracts against herpes num cyminum L.) oils. In: Preedy VR, 2012; 114: 168–174. simplex virus type 1 in Vero cell cul- ed. Essential Oils in Food Preservation, 12. Jafari S et al. Evaluation the effect of ture. Iran J Med Sci 2010; 35: 304– Flavor and Safety. Cambridge, MA: 50 and 100 mg doses of Cuminum 309. Academic Press, 2016: 377–383. cyminum L. essential oil on glycemic 3. Tuncturk R, Tuncturk M. Effects of 8. Muthamma MKS et al. Enhancement indices, insulin resistance and serum different phosphorus levels on the of digestive enzymatic activity by inﬂammatory factors on patients with yield and quality components of cumin (Cuminum cyminum L.) and diabetes type II: a double-blind ran- cumin (Cuminum cyminum L.). Res J role of spent cumin as a bionutrient. domized placebo-controlled clinical Agric Biol Sci 2006; 2: 336–340. Food Chem 2008; 110: 678–683. trial. J Tradit Complement Med 2017; 4. Li R, Jiang Z. Chemical composition 9. Mostafa DM et al. Transdermal cumin 7: 332–338. of the essential oil of Cuminum cymi- essential oil nanoemulsions with 13. Moghaddam M et al. Variation in num L. from China. Flavour Fragr J potent antioxidant and hepatoprotec- essential oil composition and antioxi- 2004; 19: 311–313. tive activities: in-vitro and in-vivo dant activity of cumin (Cuminum 5. Hajlaoui H et al. Chemical composi- evaluation. J Mol Liq 2015; 212: 6–15. cyminum L.) fruits during stages of tion and biological activities of Tuni- 10. Patil SB et al. Insulinotropic and b- maturity. Ind Crops Prod 2015; 70: sian Cuminum cyminum L. essential cell protective action of 163–169.

14. Nitoda T et al. Effects of cuminalde- Lipid Sci Technol 2018; 120: 1800177. nanoemulsifying drug delivery system hyde on melanoma cells. Phytother Early view: 1–11. (SNEDDS) for enhanced chemothera- Res 2008; 22: 809–813. 25. Lin L et al. Improving the stability of peutic effect. Int J Pharm 2013; 452: 15. Lee H-S. Cuminaldehyde: aldose thyme essential oil solid liposome by 412–420. reductase and a-glucosidase inhibitor using b-cyclodextrin as a cryoprotec- 36. Zhang Z et al. A high-drug-loading derived from Cuminum cyminum L. tant. Carbohydr Polym 2018; 188: self-assembled nanoemulsion enhances seeds. J Agric Food Chem 2005; 53: 243–251. the oral absorption of probucol in 2446–2450. 26. Manjunath K, Venkateswarlu V. Phar- rats. J Pharm Sci 2013; 102: 1301– 16. Morshedi D et al. Cuminaldehyde as macokinetics, tissue distribution and 1306. the major component of Cuminum bioavailability of clozapine solid lipid 37. Sood S et al. Optimization of cur- cyminum L., a natural aldehyde with nanoparticles after intravenous and cumin nanoemulsion for intranasal inhibitory effect on alpha-synuclein intraduodenal administration. J Con- delivery using design of experiment fibrillation and cytotoxicity. J Food Sci trol Release 2005; 107: 215–228. and its toxicity assessment. Colloids 2015; 80: H2336–H2344. 27. Ponce Cevallos PA et al. Encapsula- Surf B Biointerfaces 2014; 113: 330– 17. Tomy MJ et al. Cuminaldehyde as a tion of cinnamon and thyme essential 337. lipoxygenase inhibitor: in vitro and in oils components (cinnamaldehyde and 38. McEvoy CL et al. In vitro–in vivo silico validation. Appl Biochem thymol) in b-cyclodextrin: effect of evaluation of lipid based formulations Biotechnol 2014; 174: 388–397. interactions with water on complex of the CETP inhibitors CP-529,414 18. Farghali H et al. Hepatoprotective stability. J Food Eng 2010; 99: 70–75. (torcetrapib) and CP-532,623. Eur J properties of extensively studied 28. Petrovic GM et al. Encapsulation of Pharm Biopharm 2014; 88: 973–985. medicinal plant active constituents: cinnamon oil in b-cyclodextrin. J Med 39. Wang H et al. Self-nanoemulsifying possible common mechanisms. Pharm Plant Res 2010; 4: 1382–1390. drug delivery system of trans-cin- Biol 2015; 53: 781–791. 29. Ayala-Zavala JF et al. Microencapsula- namic acid: formulation development 19. Makni M et al. Evaluation of the tion of cinnamon leaf (Cinnamomum and pharmacodynamic evaluation in antioxidant, anti-inflammatory and zeylanicum) and garlic (Allium sati- alloxan-induced type 2 901 diabetic hepatoprotective properties of vanillin vum) oils in beta-cyclodextrin. J Incl rat model. Drug Dev Res 2015; 76: 82– in carbon tetrachloride-treated rats. Phenom Macrocycl Chem 2008; 60: 93. Eur J Pharmacol 2011; 668: 133–139. 359–368. 40. Larsen AT et al. Oral bioavailability of 20. Adu-Frimpong M et al. Preparation, 30. Sun X et al. Antimicrobial and cinnarizine in dogs: relation to optimization, and pharmacokinetic mechanical properties of b-cyclodex- SNEDDS droplet size, drug solubility study of nanoliposomes loaded with trin inclusion with essential oils in and in vitro precipitation. Eur J triacylglycerol-bound punicic acid for chitosan films. J Agric Food Chem Pharm Sci 2013; 48: 339–350. increased antihepatotoxic activity. 2014; 62: 8914–8918. 41. Hunter AC et al. Polymeric particu- Drug Dev Res 2018; 80: 230–245. 31. Liolios CC et al. Liposomal incorpo- late technologies for oral drug delivery 21. Gram DY et al. Effect of pomegranate ration of carvacrol and thymol iso- and targeting: a pathophysiological (Punica granatum) seed oil on carbon lated from the essential oil of perspective. Maturitas 2012; 73: 5–18. tetrachloride-induced acute and Origanum dictamnus L. and in vitro 42. Yildirim ST et al. Cinnamon oil chronic hepatotoxicity in rats. Phcog antimicrobial activity. Food Chem nanoemulsions by spontaneous emul- Res 2018; 10: 124–129. 2009; 112: 77–83. sification: formulation, characteriza- 22. Ismail AFM et al. Hepatoprotective 32. Sebaaly C et al. Preparation and char- tion and antimicrobial activity. LWT effect of grape seed oil against carbon acterization of clove essential oil – Food Sci Technol 2017; 84: 122–128. tetrachloride induced oxidative stress loaded liposomes. Food Chem 2015; 43. Nirmal NP et al. Formulation, charac- in liver of c-irradiated rat. J Pho- 178: 52–62. terisation and antibacterial activity of tochem Photobiol B Biol 2016; 160: 1– 33. Sebaaly C et al. Clove essential oil-in lemon myrtle and Nanoemulsion, 10. cyclodextrin- in-liposomes in the anise myrtle essential oil in water. 23. Roozi H, Boojar M. A comparative aqueous and lyophilized states: from Food Chem 2018; 254: 1–7. study of cumin aldehyde and para- laboratory to large scale using a mem- 44. Kaithwas V et al. Nanostructured cymene effects on mice liver tissue’s brane contactor. Carbohydr Polym lipid carriers of olmesartan medox- antioxidant activity after 24 hours of 2016; 138: 75–85. omil with enhanced oral bioavailabil- incubation in vitro. Indian J Sci Res 34. Saini N, Singh G. Gas chromatographic ity. Colloids Surf B Biointerfaces 2017; 2014; 5: 369–374. validated method for quantification of 154: 10–20. 24. Adu-Frimpong M et al. Formulation ayurvedic polyherbal formulation. 45. Shah B et al. Application of quality by of pomegranate seed oil: a promising Asian J Pharm 2015; 9: 200–205. design approach for intranasal deliv- approach of improving stability and 35. Seo YG et al. Development of doc- ery of rivastigmine loaded solid lipid health-promoting properties. Eur J etaxel-loaded solid self nanoparticles: effect on formulation

and characterization parameters. Eur J studies. Int J Pharm 2010; 383: 170– 68. David A. Dispersed system. In: Aulton Pharm Sci 2015; 78: 54–66. 177. M, ed. Pharmaceutics: the Science of 46. Baig MS et al. Application of Box- 56. Pharmacopoeia U. The United States Dosage form Design, vol. 2. New York, Behnken design for preparation of Pharmacopoeia Convention, Inc., NY: Churchill Livingstone, 2002: 70– levofloxacin-loaded stearic acid solid Rockville, MD. USP30-NF25, 2007. 100. lipid nanoparticles for ocular delivery: 57. Costa P, Sousa J. Modeling and com- 69. Joe JH et al. Effect of the solid-disper- optimization, in vitro release, ocular parison of dissolution profiles. Eur J sion method on the solubility and tolerance, and antibacterial activity. Pharm Sci 2001; 13: 123–133. crystalline property of tacrolimus. Int Int J Biol Macromol 2016; 85: 258– 58. Jain A, Jain S. In-vitro release kinetics J Pharm 2010; 395: 161–166. 270. model fitting of liposomes: an insight. 70. Sribharathy V, Rajendran S. Cuminum 47. Box G, Behnken D. Some new three Chem Phys Lipids 2016; 201: 28–40. cyminum L. extracts as eco-friendly level designs for the study of quantita- 59. Mendyk A et al. KinetDS: an open corrosion inhibitor for mild steel in tive variables. Technometrics 1960; 2: source software for dissolution test seawater. ISRN Corros 2013; 1: 1–7. 455–475. data analysis. Dissolut Technol 2012; 1: 71. Bennie RB et al. Synthesis, spectral 48. Nazzal S et al. Optimization of a self- 6–10. characterization and antimicrobial nanoemulsified tablet dosage form of 60. Yamaoka K et al. Application of Akai- studies of Schiff base transition metal Ubiquinone using response surface ke’s in formation criterion (AIC) in complexes derived from cuminalde- methodology: effect of formulation the evaluation of linear pharmacoki- hyde and 4-aminoantipyrine. Chem ingredients. Int J Pharm 2002; 240: netic equations. J Pharmacokinet Bio- Sci Trans 2014; 3: 937–944. 103–114. pharm 1978; 6: 165–175. 72. Mojarrab M et al. Antimalarial evalu- 49. Montgomery D. Design and Analysis 61. Inugala S et al. Solid self-nanoemulsi- ation of cuminaldehyde, an aromatic of Experiments, 5th edn. New York, fying drug delivery system (S- monoterpenoid, using cell free b-he- NY: John Wiley and Sons, 2001. SNEDDS) of darunavir for improved matin formation assay. J Reports 50. Omari-Siaw E et al. Hypolipidemic dissolution and oral bioavailability: Pharm Sci 2014; 3: 36–41. potential of perillaldehyde-loaded self- in vitro and in vivo evaluation. Eur J 73. Shi F et al. Realgar nanoparticle-based nanoemulsifying delivery system in Pharm Sci 2015; 74: 1–10. microcapsules: preparation and in- high-fat diet induced hyperlipidemic 62. Balakumar K et al. Self nanoemulsify- vitro/in-vivo characterizations. J mice: formulation, in vitro and ing drug delivery system (SNEDDS) Pharm Pharmacol 2014; 67: 35–42. in vivo evaluation. Eur J Pharm Sci of rosuvastatin calcium: design, for- 74. Komaiko J, McClements D. Low-en- 2016; 85: 112–122. mulation, bioavailability and pharma- ergy formation of edible nanoemul- 51. Elnaggar YSR et al. Self-nanoemulsify- cokinetic evaluation. Colloids Surf B sions by spontaneous emulsification: ing drug delivery systems of tamoxifen Biointerfaces 2013; 112: 337–343. factors influencing particle size. J Food citrate: design and optimization. Int J 63. Ambhore NS et al. Pharmacokinetic Eng 2015; 146: 122–128. Pharm 2009; 380: 133–141. and tissue distribution studies of1,9- 75. El-Laithy HM et al. Novel self-na- 52. Karthik P, Anandharamakrishnan C. pyrazoloanthrone, a c-Jun-N-terminal noemulsifying self-nanosuspension Microencapsulation of docosahex- kinase inhibitor in Wistar rats by a (SNESNS) for enhancing oral aenoic acid by spray-freeze-drying simple and sensitive HPLC method. J bioavailability of diacerein: simultane- Method and comparison of its stabil- Pharm Biomed Anal 2016; 120: 57– ous portal blood absorption and lym- ity with spray-drying and freeze-dry- 64. phatic delivery. Int J Pharm 2015; 490: ing methods. Food Bioprocess Technol 64. Zhang H et al. Preparation, in-vitro 146–154. 2013; 6: 2780–2790. release, and pharmacokinetics in rab- 76. Shah N et al. Self-emulsifying drug 53. Bu H et al. A TPGS-incorporating bits of lyophilized injection of sorafe- delivery systems (SEDDS) with polyg- nanoemulsion of paclitaxel circum- nib solid lipid nanoparticles. Int J lycolized glycerides for improving vents drug resistance in breast cancer. Nanomed 2012; 7: 2901–2910. in vitro dissolution and oral absorp- Int J Pharm 2014; 471: 206–213. 65. Abed MA, Azeez O. The effect of tion of lipophilic drugs. Int J Pharm 54. Kang BK et al. Development of self- cumin on induced diabetes in rats. 1994; 106: 15–23. microemulsifying drug delivery sys- Basra J Vet Res 2013; 12: 69–80. 77. Ke ZC et al. Design and optimization tems (SMEDDS) for oral bioavailabil- 66. Xu Y et al. Antioxidant and hepato- of self-nanoemulsifying drug delivery ity enhancement of simvastatin in protective effects of purified Rhodiola systems for improved bioavailability beagle dogs. Int J Pharm 2004; 274: rosea polysaccharides. Int J Biol of cyclovirobuxine D. Drug Des Devel 65–73. Macromol 2018; 117: 167–178. Ther 2016; 10: 2049–2060. 55. Zhao Y et al. Self nanoemulsifying 67. Ferreira SLC et al. Box-Behnken 78. Wang MM et al. Reduced burst drug delivery system (SNEDDS) for design: an alternative for the opti- release and enhanced oral bioavailabil- oral delivery of zedoary essential mization of analytical methods. Anal ity in shikimic acid-loaded polylactic oil: formulation and bioavailability Chim Acta 2007; 597: 179–186. acid submicron particles by coaxial

electrospray. J Pharm Sci 2016; 105: 87. Al-Okbi SY et al. Phytochemical con- ing Drug Delivery Systems (SNEDDS) 2427–2436. stituents, antioxidant and anticancer from chloroform bay leaf extract 79. Bover WJ, Zuman P. Lewis acid prop- activity of Mentha citrata and Mentha (Eugenia Polyantha W.) with palm erties of benzaldehydes and sub- longifolia. Res J Pharm Biol Chem Sci kernel oil as a carrier. IOP Conf Ser stituent effects. J Chem Soc Perkin 2015; 6: 739–751. Mater Sci Eng 2018; 333: 012066. Trans 1973; 24: 786–790. 88. Feng Y et al. Enhanced oral bioavail- 80. Kalam M. Development of chitosan ability and in vivo antioxidant activity Supporting Information nanoparticles coated with hyaluronic of chlorogenic acid via liposomal for- acid for topical ocular delivery of dex- mulation. Int J Pharm 2016; 501: 342– Additional Supporting Information amethasone. Int J Biol Macromol 2016; 349. may be found in the online version of 89: 127–136. 89. Bors W, Michel C. Chemistry of the this article: 81. Villar AMS et al. Design and opti- antioxidant effect of polyphenols. Ann mization of self-nanoemulsifying drug N Y Acad Sci 2002; 957: 57–69. Figure S1. Chromatograms of (a) delivery systems (SNEDDS) for 90. Zhou K et al. ESR determination of cuminaldehyde in-vitro in acetonitrile/ enhanced dissolution of gemfibrozil. the reactions between selected pheno- double distilled water (50 : 50 v/v) at Int J Pharm 2012; 431: 161–175. lic acids and free radicals or transition 30 °C with detecting wavelength of 82. Siepmann J, Peppas N. Modelling of metals. Food Chem 2006; 95: 446–457. 260 nm, retention time of 9.1 min drug release from delivery systems 91. Dua A et al. A study of antioxidant and flow rate of 1.0 ml/min and (b) based on hydroxypropyl methylcellu- properties and antioxidant com- internal standard (IS, propyl-p-hy- lose (HPMC). Adv Drug Deliv Rev pounds of cumin (Cuminum cyminum droxy benzoate) and cuminaldehyde 2001; 48: 139–157. L.). Int J Pharm Biol Arch 2012; 3: in rat plasma detected with acetoni- 83. Gupta S et al. Self-nanoemulsifying 1110–1116. trile/double distilled water (50 : 50 v/ drug delivery system for adefovir dip- 92. Nguyen DT et al. Antioxidant activi- v) at 1.0 ml/min flow rate and tem- ivoxil: design, characterization, ties of thiosemicarbazones from sub- perature of 30 °C as well as wave- in vitro and ex vivo evaluation. Col- stituted benzaldehydes and N-(tetra- length of 260 nm with retention b loids Surf A Physicochem Eng Asp O-acetyl- -d-galactopyranosyl) times of 5.0 min for and 9.5 min for – 2011; 392: 145 155. thiosemicarbazide. Eur J Med Chem IS (propyl-p-hydroxy benzoate) and 84. Jager W et al. Percutaneous absorp- 2013; 60: 199–207. 9.5 min for cuminaldehyde. tion of the montoterpene carvone: 93. Weber LW et al. Hepatotoxicity and Table S1. Box-Behnken design for implication of stereoselective metabo- mechanism of action of haloalkanes: the development of CuA-SEN with lism on blood levels. J Pharm Phar- carbon tetrachloride as a toxicological acceptable characteristics. macol 2001; 53: 637–642. model. Crit Rev Toxicol 2003; 2: 105– 85. Ahmed MB, Khater M. Evaluation of 136. Table S2. Summarized ANOVA the protective potential of Ambrosia 94. Arisqueta L et al. Involvement of lipid results of response surface methodol- maritima extract on acetaminophen- droplets in hepatic responses to ogy quadratic regression model. induced liver damage. J Ethnopharma- lipopolysaccharide treatment in mice. Table S3. Parameters obtained after col 2001; 75: 169–174. Biochim Biophys Acta 2013; 1831: fitting the release data of CuA-SEN to 86. Sallie R et al. Drugs and the liver. 1357–1367. different kinetic model equations. Part 1: testing liver function. Bio- 95. Prihapsara F et al. Acute and sub- Appendix S1. Supplementary meth- pharm Drug Dispos 1991; 12: 251–259. chronic toxicity of Self Nanoemulsify- ods.