Development and optimization of boswellic acid-loaded proniosomal gel

Abstract

Context: Boswellic acids (BAs) are isolated from oleo gum of Boswellia serrata and are mainly used as potential anti-inflammatory, hypolipidemic, immunomodulatory, and antitumor agents. Pharmacokinetic investigations of BAs uncover its poor bioavailability through digestive system thus creates a need for improved therapeutic responses which can possibly be achieved by developing formulations through novel delivery system.

Objective: Present study was conducted to design topical BA-loaded proniosomal gel for the management of inflammatory disorders with enhanced bioavailability.

Materials and methods: Nonionic surfactant vesicles were prepared using the coacervation phase separation method. A central composite design was employed to statistically optimize formulation variables using Design-Expert software. Three independent variables were evaluated: amount of surfactant (X₁), amount of soya lecithin (X₂), and amount of cholesterol (X₃). The encapsulation efficiency percentage (Y₁) and particle size (Y₂) were selected as dependent variables.

Results and discussion: The optimum formulation (F10) displayed spherical bi-layered vesicles under transmission electron microscopy with optimum particle size of 707.9 nm and high entrapment efficiency as 98.52%. In vitro skin permeation study demonstrated the most sustained release of 84.83 ± 0.153 mg/cm² in 24 h. Anti-inflammatory activity of the gel showed a significant (p < 0.001) higher percentage inhibition as compared to the marketed gel at the same dose.

Conclusion: The present study exhibited that BA-loaded proniosomal gel was better in terms of absorption, bioavailability, and release kinetics.

Keywords:

• Boswellia serrata
• central composite design
• entrapment efficiency
• release kinetics
• vesicle size

Introduction

Boswellic acids (BAs) are chemically triterpenoids, obtained from oleo gum resin of the Boswellia serrata, also called Salai Guggul family Burseraceae is broadly distributed in India (Culioli et al., 2003). Boswellia serrata contains four important BAs: α-boswellic acid, β-boswellic acid, 11-keto-β-boswellic acid (KBA), and 3-acetyl-11-keto-β-boswellic acid (AKBA) (Singh et al., 1996). The gum resin is established to possess potent anti-inflammatory activity in traditional ayurvedic medicines (Ammon, 2006) and also used in other diseases like, joint pain (Sharma et al., 1989; Gupta et al., 1994; Kimmatkar et al., 2003), tumor (Huang et al., 2000), colitis (Gupta et al., 2001), hyperlipidemia (Pandey et al., 2005), crohn’s illness (Gerhardt et al., 2001), etc.

The anti-inflammatory action of BAs has been attributed due to the inhibition of 5-lipoxygenase in a selective, non-redox, and non-compititive way (Ammon et al., 1991) as compared to other inhibitors of the enzyme 5-lopoxygenase which posess disadvantages because of their non-selectivity, in this manner BAs have advantages of less danger and constrained symptoms when contrasted with other anti-inflammatory medications (Ammon et al., 1993; Safayhi et al., 1995; Siemoneit et al., 2009). Earlier studies have suggested that both KBA and AKBA interfere with production of leukotrienes by inhibition 5-lipoxygenase; however, more recent research has shown that other BAs specially β-boswellic acid also play an important role, targeting the microsomal prostaglandin (PG) E2 synthase-1 (mPGES-1) as well as cathepsin G (CatG) thereby complementing the inflammatory action of KBA and AKBA (Tausch et al., 2009; Abdel et al., 2011; Siemoneit et al., 2011; Skarke et al., 2012).

Preliminary pharmacokinetic investigations of BAs have demonstrated that they are lipophilic in nature therefore not solubilize into the intestinal liquid (Barthe et al., 1999). Moreover there is a extensive hepatic first-pass metabolism through which systemic absorption of BAs, particularly KBA and AKBA turns out to be low in people (Sharma et al., 2004; Abdel et al., 2011; Skarke et al., 2012) need is thus arises to overcome their low bioavailability by means of topical delivery.

Proniosomes offer a potential vesicle delivery concept and may be a promising transporter for lipophilic medications, especially in light of their simple production and trite scale up. After topical application under occlusive conditions, proniosomes tend to shape niosomes upon hydration from skin (Fang et al., 2001).

Proniosomes are unique in nature for its stability and amphiphilicity. They can entrap both hydrophilic and hydrophobic drugs in their vesicles. The drug entrapped in noisome vesicle infiltrates the skin at a quicker rate than the free drug. Proniosomes also minimize the physical stability problems of niosomes like, aggregation, fusion, leaking, storage, dosing, etc (Hu & Rhodes, 1999).

Central composite design (CCD) is the most popular design to provide optimum conditions that improve a process. The CCD reduces the cost of expensive analytical methods and decreases the associated numeric noise. It is generally easy to decide the precise ideal condition for a solitary reaction utilizing response surface methodology (Bezerra et al., 2008). Based on this, the present study was aimed to incorporate BAs in proniosomal gel system for topical administration to increase bioavailability of drug. The prepared system was optimized and characterized for various parameters such as percentage drug entrapment, vesicle size, in vitro permeation, etc.

Materials and methods

The experiments were performed with dry Boswellia serrata extract received from Ambe Phytoextracts Pvt. Ltd., New Delhi containing not less than 65% of BAs. AKBA and KBA reference standards were obtained from Sigma Aldrich, Germany. All other chemicals and reagents used were of analytical grade and obtained from Central Drug House, New Delhi.

Experimental design

A CCD was used to determine the response pattern and afterward to build up a model utilizing three variables. Three variables, i.e. amount of surfactant (X₁), amount of soya lecithin (X₂), and amount of cholesterol (X₃) were studied at two levels (−1,+1). The dependent variables were percentage drug entrapment and vesicle size of proniosomal gel. All variables were tackled a central coded value considered as zero. CCD was built in a manner that 2k + 2k + 6 analyses were required where k speaks to the number of factors to be considered.

Percentage drug entrapment and vesicle size of proniosomal gel was assessed by the quadratic response surface model. The multivariate study encourages their interactions between variables and provides a complete exploitation of the experiment protocol with a lesser number of experiments. In the present study, three factors (n) were evaluated at the two levels (k) and accordingly, the CCD consisted of 8 (kⁿ) batches of full factorial design (1F-8F), 6 (2n) batches on axial points (1S–6S) and 6 replicates at the center points (1C–4C). A total of 20 experiments were performed (Table 1). The obtained experimental results were subjected to the statistical analysis performed by the software Design-Expert 8.0.7.1.

Table 1. Experimental design and results of central composite design.

Independent factors				Dependent factors
Formulation	Amount of surfactant (X1)	Amount of lecithin (X2)	Amount of cholesterol (X3)	Entrapment efficiency (%)	Particle size (nm)
F1	−1	−1	−1	97.54	870
F2	+1	−1	−1	98.2	720.9
F3	−1	+1	−1	97.6	863.4
F4	+1	+1	−1	98	726.6
F5	−1	−1	+1	96.6	952
F6	+1	−1	+1	97.39	890
F7	−1	+1	+1	97.13	902.9
F8	+1	+1	+1	97.47	883.4
F9	−1.682	0	0	97.7	769.9
F10	+1.682	0	0	98.52	707.9
F11	0	−1.682	0	98.36	715.8
F12	0	+1.682	0	97.69	820
F13	0	0	−1.682	97.73	752.8
F14	0	0	+1.682	97.62	852
F15	0	0	0	97.69	792
F16	0 (2 g)	0 (2 g)	0	97.73	760
F17	0 (2 g)	0 (2 g)	0	97.69	796
F18	0 (2 g)	0 (2 g)	0 (75 mg)	97.69	801
F19	0	0	0	97.69	810
F20	0	0	0	97.69	798

Drug-excipient compatibility study

The drug-excipient compatibility study was performed to check any interaction between drug and the excipients. FT-IR spectrum of drug sample and blend of drug sample and excipients was analyzed and recorded.

Preparation of proniosomal gel

Proniosomal gel was prepared by the coacervation phase separation method. The drug (BAs) with surfactant, lecithin and cholesterol were mixed with absolute ethanol in a wide mouth glass container. Then the open end of the glass container was warmed in a water bath at 65 ± 3 °C for 5 min. Phosphate buffer of pH 7.4 was then added and again warmed on the water bath for about 2 min until a clear solution was obtained. The mixture was then allowed to cool down at room temperature till the dispersion was converted to proniosomal gel (Ammar et al., 2011).

Entrapment efficiency determination

The proniosomal gel was mixed with 10 ml phosphate buffer of pH 7.4 in the glass tube and the resultant fluid suspension was then sonicated in a sonicator bath. The mixture was then centrifuged at 25 000 rpm at 20 °C for 30 min to isolate BA-containing niosomes from the unentrapped drug. The supernatant was then diluted with methanol. The concentration of BAs in the subsequent solution was tested by HPTLC and percentage of drug encapsulation was calculated by the following mathematical statement:

where C_t is the concentration of total BAs and C_f is the concentration of free drug. The entrapment efficiency was assessed by CAMAG Linomat V, HPTLC system (Lalit et al., 2011).

Vesicle size examination

Light scattering measurements were performed by using zetasizer (Malvern-Zetasizer Nano ZS 90, Worcestershire, UK). The proniosomal gel in the glass tube was diluted with 10 ml phosphate buffer of pH 7.4 to determine the vesicle size. The instrument was set at temperature 200 °C, viscosity 0.01 poise, refractive index 1.333, run time 200s and range 0–3000 nm (Gupta et al., 2007).

Surface morphology

The morphology of the vesicles was studied by taking one drop of niosomal scattering and diluted 10 times to load further into a carbon covered grid for 1 min. The grid was then observed by transmission electron microscopy utilizing imaging viewer programing. The pictures were captured for further investigation.

In vitro permeation study

To study the permeation of BAs from optimized proniosomal formulation, a semi-permeable membrane was placed between the donor and receptor compartment of Franz (vertical) diffusion cell. Paraffin paper was used to cover the top of the diffusion cell. The donor compartment was filled with the proniosomal formulation. A 20 ml aliquot of phosphate buffer pH 7.4 was used as a receptor medium which was maintained at 37 °C and stirred by a magnetic bar at 600 rpm. The available diffusion area of the cell was 1.25 cm². 2 ml portion from the receptor medium was withdrawn and simultaneously replaced by an equal volume of fresh receptor solution at regular intervals (1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 12, and 24 h). The samples were filtered through 0.45 μm membrane filter before estimation of the drug content by HPTLC at 254 nm. The cumulative amount of the drug permeated through the skin was then calculated to determine average percentage release and flux values (Uchegbu & Vyas, 1998).

Release kinetics

In vitro permeation studies data were fitted in different kinetic models: zero order as a cumulative percentage of drug permeated versus time, first order as log cumulative percentage of drug remaining versus time, Higuchi’s model as cumulative percentage drug permeated versus Square root of time and Korsmeyer and Peppas equation as log cumulative percentage of drug released versus log time. The exponent “n” was calculated from the straight’s incline line (Korsmeyer et al., 1983). The value was then used to characterize different release mechanisms, as shown in Table 2.

Table 2. Interpretation of diffusional release mechanisms from polymeric films.

Diffusion exponent (n)	Overall solute diffusion mechanism
0.45	Fickian diffusion
0.45 < n < 0.89	Anomalous (non-Fickian) diffusion
0.89	Case-II transport
n > 0.89	Super case-II transport

Anti-inflammatory activity–carrageenan-induced rat paw edema

The anti-inflammatory activity of the optimized formulations was assessed by the carrageenan-induced hind paw edema method (Winter et al., 1962). Young Wistar rats (180–220 g) of either sex were taken and housed in polypropylene cages. Six animals were kept in each cage with free access to standard lab diet and water ad libitum under standard research laboratory conditions (temperature: 25 ± 2 °C; relative humidity 55 ± 5%). The experimental protocol was approved by Institutional Animal Ethics Committe (IAEC) of Maharshi Dayanand University, Rohtak, India [Reg. no./date: 1767/RE/S/14/CPSCEA, 18/07/2014] as per the guidance of CPCSEA, Ministry of Social Justice and Experiment, Government of India with approved protocol no.: MDU/CAH/151-67, dated 30 March 2015. Proper guidelines were followed while conducting the anti-inflammatory studies. The anti-inflammatory effect of proniosomal gel was compared with standard anti-inflammatory marketed gel (Voveran). Thirty animals were divided into five groups, each consisting of six rats and treated as follows:

Group 1 (Control Group I): with plain proniosomal gel.

Group 2 (Standard Group): with marketed gel 1% w/w (Voveran).

Group 3 (Test Group I): with BA-loaded proniosomal formulation; dose 1% w/w.

Group 4 (Test Group II): with BA-loaded proniosomal formulation; dose 2% w/w.

Group 5 (Test Group III): with BA-loaded proniosomal formulation; dose 3% w/w.

Paw edema was induced in all the rats by injecting 0.1 ml of 1% w/w homogeneous suspension of carrageenan into the plantar surface of the right hind paw. The gel was applied to the right hind paw of rats after 1 h of carrageenan injection. The bandages were applied over area of application and kept for 2 h. The dressing was then removed and paw volume was measured with Plethysmometer up to 5 h at 1 h interval.

The rate of edema and percentage inhibition of each group was calculated as follows:

Where V_o was the mean paw volume before carrageenan injection, the V_t was the mean paw volume after the carrageenan injection at time t, E_c was the edema rate of the control group and E_t was the edema rate of the treated group at time t.

Results and discussion

Drug-excipient compatibility study

The FTIR spectra of pure BAs and mixture of drug with excipients are shown in Figure 5. The spectrum of pure BAs showed characteristic peaks at 3437 cm⁻¹ (OH stretching), 2932 cm⁻¹ (C-H stretching), 1697 cm⁻¹ (C = O stretching of aryl acid), 1453 cm⁻¹ (C-H bend), 1375 cm⁻¹ (COO⁻ symmetric stretching of carboxylates), 1240 cm⁻¹ (C-CO-C stretching of aryl ketone), 1025 cm⁻¹ and 988 cm⁻¹ (ring structures of cyclohexane). From Figure 1, it was observed that there were no significant changes in the position of the characteristic peaks of drug when mixed with a surfactant, cholesterol and soya lecithin which indicated compatibility of excipients with the drug.

Figure 1. FTIR overlay spectra of boswellic acids and mixture of boswellic acids with excipients.

Experimental design

A central composite experimental design using three independent variables at their three unique levels was utilized to study their impacts on the dependent variables. This design offers an advantage of fewer trial runs (20 runs) as compared with that of 3³ full factorial design which requires 27 runs. All the batches of proniosomal gel were prepared as per experimental design. Transformed values of all the batches along with their responses are shown in Table 2.

Statistical analysis

The relation of dependent and independent variables was established using mathematical relationships obtained with the statistical package.

The polynomial equations obtained were:

Where Y₁ represents the response, i.e. entrapment efficiency and Y₂ represent particle size. X₁, X₂, X₃, X₁ X₂, X₁ X₃, X₂ X₃, X₁², X₂², and X₃² are the variables.

The polynomial equations can be utilized to make inferences subsequent to considering the numerical sign and extent of the coefficient. The value of correlation coefficient (r²) for entrapment efficiency and particle size was found to be 0.91 and 0.94 indicating a good fit. A positive sign of the coefficient of the term X₁ and X₂ for entrapment efficiency represents a favorable effect. This may be due to more availability of surfactant and soya lecithin resulting in high value of entrapment efficiency. A negative sign for the coefficient X₃ indicates the antagonistic effect of this variable on entrapment efficiency. In case of particle size a negative sign for the coefficient X₁ indicates the antagonistic effect of this variable on particle size while the other two, i.e. X₂ and X₃ indicate favorable effect.

ANOVA analysis report

Model signification and suitability was tested by analysis of variance (ANOVA). Various statistical data such as standard error, the sum of squares, F-ratio, or p value are given in Table 3.

Table 3. Summary of results of analysis of variance (ANOVA).

		Entrapment efficiency				Particle size
Source	DF	SS	MS	F	p	SS	MS	F	p
X1	1	7.11	7.11	16.22	0.00	30 134.79	30 134.79	12.58	0.0053
X2	1	1.14	1.14	2.60	0.13	1407.49	1407.49	0.59	0.4610
X3	1	0.43	0.43	0.99	0.34	29 454.26	29 454.26	12.30	0.0057
X1 X2	1	0.005	0.005	0.013	0.91	151.38	151.38	0.063	0.8066
X1 X3	1	0.041	0.041	0.093	0.76	4250.42	4250.42	1.77	0.2124
X2 X3	1	0.20	0.20	0.45	0.51	151.38	151.38	0.063	0.8066
X1^2	1	1.86	1.86	4.24	0.06	1860.70	1860.70	0.78	0.3988
X2^2	1	0.047	0.047	0.11	0.75	495.73	495.73	0.21	0.6589
X3^2	1	0.13	0.13	0.30	0.59	4701.82	4701.82	1.96	0.1914
Lack of fit	5	4.38	0.876	14.33	0.0043	22 472.68	4494.54	15.22	0.0048
Pure error	5	0.001	0.002			1476.83
Cor. total	19	15.43				95 684.81

SS, sum of square; DF, degree of freedom; MS, mean of square; F, Fisher’s ratio.

As shown in Table 3, a p value of ≤0.05 for the independent variables and their interaction in ANOVA indicates significant effect of the corresponding factors on the entrapment efficiency and particle size. Three-dimensional (3D) reaction surface plots give a varieties’ representation in every reaction when the two components are all the while changed from lower to more elevated amount. It gives a 3D bend of the adjustment accordingly at diverse component levels. It additionally gives the variety in outline focuses from the anticipated reaction esteem. The 3D response surface plots and the corresponding contour plots for drug entrapment efficiency and particle size are shown in Figures 2–5, respectively.

Figure 2. Contour plot for entrapment efficiency.

Figure 3. 3D response surface plot for entrapment efficiency.

Figure 4. Contour plot for particle size.

Figure 5. 3D surface plot for particle size.

Optimum formula

After studying the effects of the independent variables, the optimum responses were determined. The optimum formulation of BA-loaded proniosomal gel was selected on the criteria of attaining a low particle size and maximum entrapment efficiency. Hence formulation F₁₀ was considered the optimized formulation which has maximum entrapment efficiency, i.e. 98.52% and minimum particle size, i.e. 707.9 nm.

Effect of surfactant

The effect of concentration of nonionic surfactant was studied on the entrapment efficiency and vesicle size. Different formulations were prepared using varying concentrations of Span 40. There are reports that noisome formed in span 40 proniosomal gel demonstrates higher percentage drug entrapment because of the fact that span 40 is solid at room temperature, has highest phase transition temperature, low permeability and act as gelator (Ibrahim et al., 2008; Mokhtar et al., 2008; Alam et al., 2010). It was observed that decrease in concentration of surfactant results in reduction in the entrapment of the drug and also the overall size of the noisome which may happen due to insufficient material to form a bilayer and encapsulate the drug efficiently.

Effect of soya lecithin

Soya lecithin has a property to enhance penetration due to the presence of unsaturated fatty acids, oleic acid, and linoleic acid (Mokhtar et al., 2008), so it was chosen as independent variables for influencing entrapment efficiency. Incorporation of lecithin resulted in improved drug entrapment to 98.52% and reduced the vesicle size due to increase in hydrophobicity. Lecithin forms more compact and well organized bilayers which prevent the leakage of the drug.

Effect of cholesterol

The main use of surfactant is formation of bilayer formation as the core material however nonionic surfactant alone is not strong enough to serve as host for the drug bilayer. So cholesterol having a steroidal rigid structure gives the required strength to the bilayer, but cholesterol itself is not able to form any layer (Korchowiec et al., 2006). Entrapment efficiency increases with increase in cholesterol content. It was also observed that high cholesterol content had a lowering effect on drug entrapment (98.5% to 96.6%) because cholesterol beyond a certain level disrupts the bilayered structure leading to loss of drug entrapment.

Surface morphology

The vesicles were found uniform and spherical in shape in transmission electron microscopy photomicrographs. Moreover, the drug to be packed in the core of the vesicles with drug-free bilayer was showed during positive staining (Figure 6).

Figure 6. TEM photomicrographs of boswellic acid-loaded proniosomes.

In vitro drug release study

In vitro drug release of optimized formulation was studied to predict how proniosomal gel may function in a perfect circumstance and additionally give a few indications of its in vivo execution since drug release deals with the measure of drug accessible for absorption (Gupta et al., 2007). The cumulative amount of drug released was observed to be 84.83 ± 0.153 mg/cm² in 24 h; Figure 7. In vitro release profile demonstrated a greatest flux estimation of 16.39 ± 0.014 mg/cm²/h and least estimation of 5.88 ± 0.15 mg/cm²/h.

Figure 7. In vitro release profile of boswellic acid-containing proniosomes.

Kinetics of drug release

The optimized formulation F10 was subjected to the graphical treatment to evaluate the energy of drug release. The data acquired from the best formulation was fitted to different kinetc mathematical statements to decide the system of drug release and release rate as showed by a higher correlation coefficient (r²). For the characterization of the release kinetics, the in vitro drug release data were fitted to zero-order plots, Higuchi diffusion plots and Peppas log log plots. The data were best fitted to Korsmeyer–Peppas model for niosomal drug with good linearity; r² value 0.9132 compares to r² value of 0.7083 and 0.8754 for zero order and Higuchi diffusion plot, respectively. Slope values of the Peppas log–log plot were also calculated. In almost all the cases slope values of the Peppas plots were in between 0.45 and 0.89 suggesting that the drug was released by non-Fickian (anomalous) release mechanism, i.e. the drug was released by the combination of both diffusion and erosion controlled drug release.

Anti-inflammatory activity

Carrageenan paw is a standout amongst the most commonly used models for the examination of novel anti-inflammatory agents (Villar et al., 1987). Development of paw edema of rats induces via carrageenan is usually related with right on time exudative period of inflammation which is a standout amongst the most vital system of inflammation pathology (Liu et al., 2002). In our study, the carrageenan actuated inflammatory pattern of the rodent paw was in close accordance with the past reports (Kochi et al., 2006). The intraplanter injection of carrageenan to hind paw in rats induced an increase in paw thickness. Treatment with BAs containing proniosomal gel (PG 1%) and standard Voveran gel (1%) resulted in 42.5 ± 0.024% and 50 ± 0.02% inhibition in paw edema respectively at 5 h compared to control. The increase in paw size in first hour could be because of the activity of histamine and serotonin on vascular permeability (Garcia et al., 1973). Inflammation increased gradually and attained peak at 3 h which could be due to release of PGs and kinins in paw tissue (Vane, 1971). As shown in Figure 8, BAs containing proniosomal gel treated animals with the doses of 1%, 2%, and 3% (single application) at distinctive time intervals demonstrated dose-dependent significant difference (p < 0.001) in paw edema when compared with the control treated animals at the same time point. Dose-dependent inhibition after topical application recommends that BAs may act in both early and late periods of inflammation. The increase in anti-inflammatory efect of proniosomal gel may be because of improved saturation of BAs through skin and availability of large concentration of BAs to the target tissue.

Figure 8. The time-dependent inhibitory effect of topical application of boswellic acid-loaded proniosomal gel against carrageenan induced rat paw edema in rats.

Conclusion

The present study showed that proniosomal gel is a suitable carrier for the delivery of BAs with enhanced transdermal delivery due to its small particle size and better encapsulation of drug within the vesicles. The results also concluded that the studied variables have a significant impact on the entrapment and vesicle size of the drug. Furthermore, optimized proniosomal gel showed significant anti-inflammatory effect as compared to the standard Voveran gel. From these data, it is presumed that BAs, which are considered as a leukotriene inhibitor and in clinical use as anti-inflammatory specialists has indicated good results with topical application as potential anti-inflammatory agent.

Declaration of interest

The authors have no conflicts of interest.

Financial support from University grants commission, Ministry of Human Resources and Development, Government of India under special assistance program is acknowledged.