Self-assembled drug delivery system based on low-molecular-weight bis-amide organogelator: synthesis, properties and in vivo evaluation

Abstract

Context: Orgnaogels based on amino acid derivatives have been widely used in the area of drug delivery.

Objective: An organogel system based on l-lysine derivatives was designed and prepared to induce a thermal sensitive implant with higher transition temperature, better mechanical strength, and shorter gelation time.

Materials and methods: The organogel was prepared by injectable soybean oil and methyl (S)-2,5-ditetradecanamidopentanoate (MDP), which was synthesized for the first time. Candesartan cilexetil (CC) was chosen as model drug. Different formulations were designed and optimized by response surface method. Thermal, rheology properties, and gelation kinetics of the optimized formulation had been characterized. The release behaviors in vitro, as well as in vivo were evaluated in comparison with the oily solution of drugs. Finally, the local inflammation response of in situ organogel was assessed by histological analysis.

Results and discussion: Results showed that the synthesized gelator, MDP, had a good gelation ability and the organogels obtained via the self-assembly of gelators in vegetable oils exhibited great thermal and rheology properties, which guaranteed their state in body. In vivo pharmacokinetic demonstrated that the organogel formulation could extend the drug release and maintain a therapeutically effective plasma concentration at least 10 d. In addition, this implant showed acceptable moderate inflammation.

Conclusion: The in situ forming l-lysine-derivative-based organogel could be a promising matrix for sustained drug delivery of the drugs with low solubility.

Keywords:

• Biocompatibility
• gelator
• in situ implant system
• in vivo release
• organogel

Introduction

Low-molecular-weight organogelators (LMOG), which are able to self-assembled form gels (organogels) with organic solvents, have been extensively studied in recent decades due to their potential wide applications in industrial fields, such as cosmetics, foods, medical engineering, and drug delivery (Vintiloiu & Leroux, 2008; Pandey et al., 2010; Basak et al., 2012; Jiao et al., 2012b; Balata et al., 2013; Sagiri et al., 2014). The growth of LMOG units usually result in the formation of various supramolecular polymers, such as fibers, ribbons, and sheets as a result of hydrogen bonding, van der Waals (hydrophobic) and π–π stacking interactions (Florent et al., 2010; Jiao et al., 2013; Guo et al., 2014). Among many organogelators, we have focused on the gelation behavior and their potential applications in drug delivery based on l-amino acid derivatives because l-amino acid is the component constructing a living body, thus considered as biocompatible, biodegradable, and non-toxic (Suzuki et al., 2008; Wang et al., 2010). For example, injectable in situ forming implant based on N-steroyl-l-alanine methyl ester and N-steroyl-l-alanine ethyl ester have been successfully prepared as a matrix where the therapeutic agents could be entrapped and then released from the organogels via diffusion or degradation (Plourde et al., 2005). Subcutaneous administration is frequently used as a route since it is convenient to administrate, and can avoid the first-pass metabolism, prolong the retention time, reduce the frequency of administration and improve the bioavailability of drugs (Packhaeuser et al., 2004; Madan et al., 2009; Giri et al., 2012; Ghasemi et al., 2016). The organogel, however, could be injected through syringe needle with the aid of “gelation inhibitor”, which belongs to some of the amphiphilic solvents. N-methyl-2-pyrrolodone (NMP) involved in the commercialization of a PLGA-based injectable implant (Eligard®), is found to disrupt the organogel blocks and then facilitate the injection (Oliver, 2003; Kempe & Mäder, 2012). Upon injection, NMP would diffuse into the surrounding tissues and allow the in situ formation of the organogel implant without irreversible gel damage.

Although such advances provide a bright outlook for implementation of polymeric implant, traditional l-amino acid derivatives used suffered from their characteristically weak mechanical properties due to their inherently weak nature of non-covalent interactions that comprise the self-assembled network structures (Brosse et al., 2004). The suitable mechanical strength is necessary to control the release of drug. In addition, the gelator should have good gelation ability. With the degradation of organogel, there may generate some acid compound, which was undesired as an implant. Thus, the concentration of gelator should be limited for the preparation of organogel. In order to overcome this limitation, multivalent interactions have been tried and gained some success. For example, l-lysine-based LMOG were designed and their organogelation properties were investigated (Hanabusa et al., 2000; Suzuki et al., 2008). The researchers revealed that l-lysine-based LMOG had better organogelation abilities than those with an amide bond, attributing to the fact that l-lysine group undergoes a stronger hydrogen bonding interaction (Suzuki & Hanabusa, 2009). On the other hand, the researchers also observed that the l-lysine derivatives with long alkyl chain length led to precipitation in many organic solvents. Though there are some reports on their gelation behavior in organic solvents, relative few have been investigated in the field of drug delivery (Masahiro et al., 2003; Suzuki et al., 2005; Chen et al., 2015). In addition, some gelators have a complex structure and the synthetic routes are complex, which were undesired for their application.

There are a lot of active compounds which cannot be developed as a drug and used in clinical, because the low solubility directly leads a low bioavailability. As a prodrug of candesartan, candesartan cilexetil (CC) is a non-peptide angiotensin II receptor antagonist and usually used in the treatment of hypertension (Meineke et al., 1997; Surampalli et al., 2014). However, its therapeutic application was limited as an oral dosage form by its very low solubility and oral bioavailability (McClellan & Kl, 1998; Detroja et al., 2011; Dudhipala & Veerabrahma, 2016). This simple prepared drug delivery system is an ideal vehicle for the administration of lipophilic drug such as canderstan celiexetil.

In this study, a novel amino acid derivative based on l-lysine was designed and synthesized. Attempts were made to characterize and develop the organogel formulation with CC as a model drug for its probable application in drug delivery system. Thermal, rheology properties, and gelation kinetics of the drug-loaded organogel were investigated and the release profile was also determined in vitro. The in vivo pharmacokinetic study was conducted in comparison with the oily solution of drugs. In addition, the local inflammation response was assessed by histological analysis for the biocompatibility of this organogel matrix after injection.

Materials and methods

Materials

l-Lysine methyl ester hydrochloride was purchased from Yangzhou Baosheng Biochemical Co., Ltd., Yangzhou, China. Myristic acid and oxalyl chloride were obtained from Chengdu Gracia Chemical Technology Co., Ltd., Chengdu, China. Injectable soybean oil and injectable medium chain triglycerides (MTC) were purchased from Zhonghang Tieling Pharmaceuticals Co., Ltd., Tieling, China. Salad oil, corn oil, sunflower oil, and olive oil were purchased from China Oil & Foodstuffs Corporation, Beijing, China. Candesartan ciexetil (CC) and candesartan (CDS) were purchased from Wuhan Combined Biochemical Co., Ltd., Wuhan, China. CDS standard was purchased from Takeda Pharmaceutical Co., Ltd., Tianjin, China. Dimethylformamide was purchased from Shanghai Jinjinle Industrial Co., Ltd., Shanghai, China. Methanol of HPLC grade was obtained from Shandong Yuwang Industrial Co., Ltd., Shandong, China. The other solvents were obtained from Tianjin Bodi Chemical Co., Ltd., Tianjin, China.

Synthesis of the gelator

Methyl (S)-2,5-ditetradecanamidopentanoate (MDP) was synthesized by acylation of l-lysine methyl easer with acyl chlorides through a one-pot synthetic procedure as reported previously (Motulsky et al., 2005). Briefly, oxalyl chloride was added drop-wise to an anhydrous chloroform solution (50 mL) of myristic acid (0.042 mol) containing a small amount of dimethylformamide as catalyst with stirring at 0 °C. The mixture was stirred for about 3 h until no bubbles existed in the system, after which myristoyl (C₁₄) chloride precipitated upon the solvent evaporated.

Triethylamine (0.085 mol) was introduced to an anhydrous chloroform solution (100 mL) of l-lysine methyl ester hydrochloride (0.02 mol) and the achieved solution was stirred for 1 h at room temperature. Then myristoyl (C₁₄) chloride (0.05 mol) dissolved in chloroform (2 mL) was added drop-wise to above mixture with stirring at 0 °C until the end of the reaction (about 7 h). Progression of the reaction was monitored by thin layer chromatography (TLC). The solution was successively washed with distilled water, saturated aqueous solution of sodium chloride, saturated aqueous solution of sodium bicarbonate, hydrochloric acid (1M) and again with water. The organic solvent was removed under reduced pressure and the crude obtained was recrystallized with tetrahydrofuran. Chemical structures were identified by FTIR (IFS 55, Bruker, Germany) (Figure 11 in supplementary), NMR spectra (Switzerland Burker, ARX-400) spectrometers in chloroform-D1 with tetramethylsilane (TMS) as internal standard (Figure 12 in supplementary), and MS (solarix 7.0T, Bruker Daltonics, Germany).

FT-IR (KBr, v, cm⁻¹): v 3323 (v^sN–H), 3299 (v^sN–H), 2964 (v CH₃), 2915 (v^asCH₂), 1641 (v C = O), 1556 (β N–H). 1H NMR (300 MHz,CDCl3) δ (ppm): 6.19 (d, J = 6.0 Hz, 1H), 5.73 (s, 1H), 4.61–4.54 (m, 1H), 3.74 (s, 3H), 3.24 (dd, J₁= 9.0 Hz, J₂= 13.5 Hz 2H), 2.35–2.14 (m, 4H), 1.25 (s, 50H), 0.88 (t, J = 6.0 Hz, 6H). MS m/z: 581.5255 [M + H]+, 603.5075 [M + Na]+.

Gelation ability

The gelation ability of MDP was measured by inverse flow method (Luboradzki et al., 2005). Briefly, 50 mg of the gelator was weighted and placed in a vial. 0.5 mL of the organic solvent was added and the mixture was heated at 60 °C until the gelator dissolved absolutely. After balanced in water bath (25 °C) for 15 min, the vial was inversed vertically to assess the formation of organogel. If the gel formed, a small amount of corresponding solvent was added. Otherwise, certain amount of the gelator was added and repeated the procedures above until the minimal gelator concentration (MGC) of MDP was obtained. All determinations were conducted in triplicate.

Preparation of organogels

The organogel was simply prepared by dissolving amount of gelator (MDP) in vegetable oils under magnetic stirring at 60 °C. The solution was cooled to room temperature and turned upside down to check whether the organogel formed. When preparing the drug-loaded organogel, certain amount of drug dissolved in the pre-heated oil solution was used to replace the oil phase. The gel samples were stored at 4 °C.

Thermal stability of organogels

Inverse flow method was employed to determine the sol–gel (T_sg) and gel–sol (T_gs) phase transition temperatures of organogels. To elucidate the thermal stability, a set of organogels was prepared with varying concentrations of MDP. The organogels were allowed to balance at respective temperature for 15 min and then observed by inversing the vial vertically (Dan et al., 2014). The temperature at which the organogel could not form was considered to be T_gs and the determination of sol–gel transition temperature was carried out in an opposite procedure with a cooling procedure until the organogel formed. Temperature range was taken from 25 °C to 80 °C with 2 °C/time change.

Scanning electron microscopy (SEM) observation

The organogel at certain concentration (10% or 15%) built on MDP/toluene was prepared and dried in vacuum for 24 h to obtain the xerogel. After coated with gold, SURA 35 field emission scanning electron microscope (ZEISS, Oberkochen, Germany) operating at 15 kV was used to observe the morphology of the xerogel.

Optimization of candesartan cilexetil organogel

In order to obtain the optimal formulation with appropriate release profile of CC quickly, a computer optimization technique based on response surface methodology (RSM) was used to evaluate the effects of different amounts of MDP (X₁), NMP (X₂), and CC (X₃) on drug release (Changa et al., 2007; Yang et al., 2012). A three-factor, three-level Box–Behnken design (BBD) was used to optimize the formulation of CC-loaded organogel. Cumulative release amounts of CC from the organogels in 24 h (Y_24h) and 15 d (Y_15d) were chosen as the response values. The response and formation variables of all model formulations were analyzed by Design-Expert software (Design Expert® v 8.0.1, Stat-Ease, Minneapolis, MN). Details of studied levels and the response variables are shown in Table 1.

Table 1. The response values for the different levels of experimental design.

Run	X1 (mg)	X2 (mL)	X3 (mg)	Y24h (%)	Y15d (%)
1	80.00	0.10	10.00	21.13	62.83
2	50.00	0.20	10.00	24.57	75.04
3	50.00	0.20	10.00	24.38	72.61
4	50.00	0.20	10.00	26.33	75.54
5	50.00	0.30	15.00	25.95	68.10
6	80.00	0.20	15.00	21.28	60.28
7	20.00	0.20	15.00	29.12	77.21
8	80.00	0.20	5.00	30.58	89.20
9	50.00	0.30	5.00	30.27	101.45
10	50.00	0.20	10.00	24.21	75.04
11	20.00	0.20	5.00	36.44	100.12
12	50.00	0.10	15.00	20.09	69.10
13	20.00	0.30	10.00	28.72	97.48
14	80.00	0.30	10.00	22.83	66.98
15	50.00	0.20	10.00	24.62	77.86
16	20.00	0.10	10.00	23.04	75.73
17	50.00	0.30	5.00	28.60	81.68

X₁, X₂, X₃ – different amounts of MDP, NMP, and CC.

Y_24h and Y_15d represent the drug release percent at 24 h and 15 d.

In vitro release evaluation

1 mL of the MDP/injectable soybean oil/NMP formulation solution was transferred to the dialysis bag with a mass cut-off of 8000–14 000 Da, and then incubated in 100 mL release medium (10 mM phosphate buffer containing 0.5% SDS and 0.1% sodium azide, pH 7.4). The systems were placed in a water bath oscillator (100 rpm) at 37 °C. At predetermined time intervals, 6 mL release medium was withdrawn for analysis and replaced by fresh buffer to maintain sink condition. The samples were then filtered with 0.45 μm filter and analyzed by ultraviolet-visible spectrophotometer (UV) at 254 nm. Drug release profile of its oily solution was also determined as control (n = 3).

Evaluation of the optimized CC formulation

DSC determination

The thermal properties of drug, blank organogel, and the optimized formulation were measured by differential scanning calorimeter (DSC, TA-60WS, Shimidazu Co., Kyoto, Japan). Samples (about 5 mg) were weighted and added in an aluminum pan. The scanning range of temperature was 30 °C to 200 °C at a heating or cooling rate of 10 °C/min under nitrogen atmosphere. Both of heating and cooling curves were recorded except the drug powder.

Rheological measurements

Rheology analyses of the developed formulation and its blank organogel were performed using AR 2000ex rheometer (TA, Co Ltd.) with a circulating environment system for temperature control. A parallel-plate geometry (20 mm diameter) was used and the gap between the plates was fixed at 1000 μm. The measurements of elasticity modulus (G′) and the viscosity modulus (G″) were carried out over a frequency sweep between 0.1 and 10 Hz with a particular stain (0.5%) at 37 ± 1 °C.

Gelation kinetics determination

The gelation process of injectable MDP/injectable soybean oil/NMP solution was investigated by electrical conductivity method at 37 °C (Wang et al., 2011b). Briefly, the optimized formulation was introduced into the dialysis bag (8000–14 000 Da), followed by the immersion of the system into 20 mL phosphate-buffered saline (PBS) solution (pH 7.4). The values of electrical conductivity of dialysis solvent were recorded and the organogel formed completely when the electrical conductivity became constant.

In vivo pharmacokinetic study of CC formulation

Pharmacokinetic study was carried out in male Sprague-Dawley rats weighted 200–250 g, which were obtained from the Experimental Animal Center of Shenyang Pharmaceutical University. All the experiments were approved by the Animal Ethics Committee of Shenyang Pharmaceutical University. Before in vivo experiment, animals were acclimatized under laboratory condition for a week. Ten rats were randomly divided into two groups. A single dose of formulation (0.2 mL, 6.4 mg/kg CC) was administered subcutaneously in the dorsal area with 22-G syringe needle. The optimal formulation was shown as 4.9% MDP with 30% NMP and the control formulation was a corresponding amount of CC dissolved in the oil solution. After administration, the blood samples (0.5 mL) were collected from retro-orbital plexus at the time of 0.5, 1, 2, 6, 12, and 24 h. After day 1, blood samples were collected once a day until the end of 10 d. The plasma was separated by centrifugation (5000 rpm × 10 min) and stored at −20 °C for further analysis.

CDS was extracted from the plasma according to the method reported by Nekkanti et al. (2009) with a minor modification. Briefly, 20 μL of valsartan (1.25 mg/mL), as an internal standard, and 10 μL HCl (1M) were added to the plastic centrifugal tube containing 200 μL of plasma sample and the solution was vortexed for 1 min. After 3 mL of methanol added to the tube, this mixture was vortexed for 3 min and centrifuged at 10 000 rpm for 10 min. 2 mL of the supernatant was separated and evaporated to dryness under a gentle steam of nitrogen. The residue was then reconstituted in 200 μL of the mobile phase and centrifuged at 10 000 rpm for 10 min. An aliquot of the 20 μL was then withdrawn for the assay.

The concentration of CDS was monitored by the HPLC system, which equipped with a Hitachi 10-AT pump, and a UV detector set as 254 nm. Analyses were performed by Thermo ODS-2 Hypersil (250 mm × 4.6 mm, 5 μm, Waltham, MA) column with guard column JanuSep C18 (Benxi, China) and the optimized mobile phase consisted of methanol–water–phosphoric acid (70:30:0.1, v/v/v) at a flow rate of 1 mL/min. The CDS content in plasma was calculated according to the peak area ratios of drug to the internal standard.

Pharmacokinetic analysis

Pharmacokinetic parameters were calculated with the pharmacokinetic program Drug And Statistics (DAS 2.0). The differences in pharmacokinetic parameters were evaluated statically by using the 2-sample t-test and it was considered significant when p values lower than 0.05. The percentage of drug dose absorbed profiles was calculated using the Wagner–Nelson model and the equation was performed as follows: where F_at is the fraction of drug absorbed at time t; C_t is concentration of drug in the plasma at time t. The in vivo absorption values are directly related to the in vitro release data to complete the in vivo–in vitro correlations.

In addition, the kinetics of drug release from the organogel system was determined by Ritger–Peppas model. The equation was expressed as follows: n represents the diffusion index, which could be used to characterize the mechanism of release. When n ≤ 0.43, it reflects the release behavior is dominated by diffusion. While if n ≥ 0.89, the release mechanism of drug is erosion. When the value is between 0.43 and 0.89, it means the release mechanism is controlled by a combined action of both diffusion and erosion.

Histological analysis

Histological analysis of the organogel was carried out on mice after subcutaneous injection. After 10 d, the rats were sacrificed and the tissues adjacent to the injection site were collected, fixed with 10% formalin solution, and embedded in paraffin. The tissue section was stained by hematoxylin–eosin and used to evaluate the local inflammatory response.

Results and discussion

The MGC of MDP in varying solvent was determined, which demonstrated the gelation ability of the corresponding organogels (Samai et al., 2011). As shown in Table 2, MDP can form gel in all of the vegetable oils, but lost the ability in some of aromatic solvents and glycols. Moreover, the MGC values varied in different solvents, suggesting the organogelation properties depend on the organic solvent (Wang et al., 2011a; Jiao et al., 2012a; Sawalha et al., 2013). Compared with our previous study, MDP showed extremely better gelation ability. For example, in the vegetable oils such as soybean oil, sunflower oil and olive oil, the MGC (g/L) values of MDP were 3.3, 7.2, and 4.4, which were significantly lower than N-stearoyl-l-alanine methyl ester (SAM) with values of 20.0, 20.83, and 25.42, respectively (Wang et al., 2010). This attributed to the fact that the increased number of amide unites in the structure of MDP improved the weak hydrogen bonding compared with SAM. In addition, the two alkyl chains in both ends of the gelator further enhanced the van der Waals interactions between gelator–gelator and gelator–oil molecules.

Table 2. Minimal gelator concentration of MDP in different solvent at 25 °C.

Solution	MGC (g/L)	Solution	MGC (g/L)	Solution	MGC (g/L)
Injectable soybean oil	3.3	Cyclohexane	96.5	Ethylene glycol	–
Corn oil	4.5	Tetrachloromethane	125.0	Acetonitrile	20.8
MCT	6.3	Toluene	166.7	Ethanol	51.7
Sunflower oil	7.2	Dimethylbenzene	–	Methanol	50.0
Olive oil	4.4	Chloroform	–	DMSO	13.2
Blend oil	3.6	p-Dioxane	–	Heptane	50.0

“–”means value higher than 200.

Thermal stability of organogel

The prepared MDP organogels were thermal reversible and their thermal stabilities were concentration dependent. With the increasing concentration of MDP, a monotonous increase in their both phase transition temperatures was observed (Figure 1), indicating the density of the gel assembly increase to ensure participation of a greater number of gelator molecules per unit volume in the thermally induced transition. For the organogels prepared by MDP and injectable soybean oil, the difference values between T_sg and T_gs were 5.3 °C and 14.3 °C for the concentration of 0.5% and 2%, respectively, suggesting the organogel can stay stable at a larger temperature range. To be a suitable drug depot in body, T_sg and T_gs should be both higher than 37 °C. The organogels formed by MDP can meet this standard at extremely low concentration of about 1%. As the gelator had a relatively good gelation ability in injectable soybean oil, which was safe when preparing a subcutaneous implant, it was chosen to prepare the organogel formulation (Poullain-Termeau et al., 2008).

Figure 1. Phase transition temperature of the organogel based on MDP. (A) Temperature of sol to gel (T_sg) and (B) temperature of gel to sol (T_gs).

Scheme 1. The synthetic scheme of methyl (S)-2,5-ditetradecanamidopentanoate.

Scanning electron microscopy (SEM)

To obtain an insight into the aggregation mode of gelators, SEM studies were performed on xerogel. As shown in Figure 2, a lamellar morphology was observed and the lamellar structure inclined to link to each other. At the same time, the lamellar structure became bigger and tightness with the increase of MDP concentration, indicating the structure increment of mechanical strength, which was consistent with the increase of thermal stability along with gelator concentration.

Figure 2. SEM images of xerogels formed by MDP in toluene at the concentration of 10% (A) and 15% (B).

Optimization of the CC organogels

Based on the design and results of Box–Behnken tests, the Design-Expert software was used to fit the regression model. The values of response variables Y_24h (percent drug release in 24 h) and Y_15d (percent drug release in 15 d) obtained from batches prepared according to response surface design were shown in Table 1 and the obtained equations were expressed as follows:

The significance level of the model was less than 0.001 and the model determination coefficient R² were 0.9285 and 0.9546, respectively, which suggested the model was reliable and could fit the experimental results well. The optimal formulation was chosen based on two principles: the drug released slowly in the first 24 h and as more as possible at the end of 15 d. Results showed that the initial burst release (24 h) was a combined action of MDP, NMP and CC, however, the drug release in 15 d was mostly affected by the amount of drug. On the basis of these, the predicted formulation contained MDP: 49.80 mg, NMP: 0.3 mL, CC: 7.11 mg and injectable soybean oil: 1 mL.

Evaluation of the optimized CC formulation

DSC measurement

The DSC curve of drug was shown in Figure 3(A) with a sharp endothermic peak of 172.3 °C. While the melting point of the drug was disappeared when loaded into organogel. As shown in Figure 3(B), the melting temperatures (T_gs) were found to be higher than the solidification temperature (T_sg). This type of behavior has been commonly observed with other LMOG-based gels (Dan et al., 2014). Both of the phase transition temperatures were higher than 37 °C, but they changed a little after the loading of drug. The T_gs of blank organogel and drug-loaded organogel were 86.0 °C and 84.3 °C respectively, T_sg of which was 58.3 °C and 56.6 °C. It may result from the breakup of interactions between gelator molecules after the adding of drug. These results indicated that the drug would influence the thermal properties of organogels, however, the impact could be ignored. In all, the CC-loaded organogel had a suitable thermal behavior as an in situ implant and could form stable organogel in body.

Figure 3. DSC thermograms of the drug (A) and organogel (B) with (a), (d) CC organogel and (b), (c) blank organogel.

Rheological measurement

Rheological property is an important factor in the developing of in situ forming organogels to measure their physical and structure stability, which is strongly influenced by the aggregation of gelator molecular (Samai et al., 2011). When preparing the formulation, the vehicle should have certain mechanical strength as many native tissues have moduli in the kPa range (Carter et al., 2001; Aamer et al., 2004). Otherwise, the gelling agent is likely to diffuse out of the depot and into the surrounding tissue since the fluid flow in subcutaneous tissue usually ranges from 7 to 53 mL/100 g per minute (Schoenhammer et al., 2009; Supper et al., 2014). As Figure 4(A) shows, the elastic modulus (G′) increased with the viscous modulus (G″) decreased at the beginning, and then both of them increased gradually, indicating a behavior of rheopexy. This phenomena was common in the prepared organogel like Gui-Chao et al., who had synthesized a series of amino acids-based dendrons with various focal moieties or peripheral groups, and the organogels exhibited the same rheological trends (Kuang et al., 2011). Frequency sweep rheometry measurement of the blank organogel and CC organogel showed the same trend that the values of their G′ were always greater than viscous modulus G″, over the entire frequency range, suggesting the behavior organogel was more elastic-like than fluid-like.

Figure 4. Variation of storage modulus (G′) and loss modulus (G″) of blank organogel (A) and CC organogel (B).

In this study, we aimed to synthesize a gelator, which could form an organogel had suitable mechanical strength at low concentration. l-alanine derivatives were developed as drug vehicles for lipid drugs, but their mechanical strength was not enough to some extent. Bastiat et al. had reported a series of gelators based on tyrosine, tryptophan, and phenylalanine, and all of them had a better gelation ability than l-alanine. The researchers also reported that tyrosine derivatives produced the strongest gelation ability and the organogel obtained via N-stearoyl-l-tyrosine methyl ester (StyrOCH₃) in sanflower oil showed a G′ value of 100 kPa with 20 kPa for G″ at 25 °C, which was almost the same as MDP at 37 °C, indicating a strong mechanical strength of MDP organogel (Bastiat & Leroux, 2009). Comparing to other in situ forming depot for drug delivery, the system formed by the l-lysine derivative showed a typically gel-like character at low concentration (Yang et al., 2012; Chen et al., 2014).

After the addition of drug into the organogel (Figure 4B), G′ value and G″ value were both decreased, indicating a decrease of mechanical strength. These results were in accordance with the decrease of transition temperature, which was related to the interactions between gelator and drug molecules. Nevertheless, the drug-loaded organogels remained G′ as 70 kPa with 16 kPa of G″, demonstrating the dominant elastic behavior of the system with a strong mechanical profile facilitated for the particular applications as in situ implant.

Gelation kinetics determination

The gelation process was traced by the method based on determination of electrical conductivity. As organogel solution should take sol/gel transition in situ under body temperature, its gelation time was determined in PBS solution at 37 °C. Figure 5 shows that the electrical conductivity was in line with the amount of NMP (R² = 0.9987) over the range of 0–1.0 mL. Indeed, the electrical conductivity was almost constant in the blank PBS solution, which was presented in Figure 6. Conversely, when the organogel solution was incubated in the medium, the electrical conductivity decreased faster in the initial 20 min and slightly decreased afterward until 40 min. After 40 min, the values of electrical conductivity almost reached a constant value (5.70 ms/cm) equaled to the electrical conductivity of 0.3 mL NMP in PBS, indicating the organogel formed absolutely. In view of our previous studies, the new synthesized MDP organogel showed a shorter gelation time than N-lauroyl-l-alanine methyl ester (LAM) at a certainly low concentration (40 min at 5% versus 110 min at 30%) (Wang et al., 2011b).

Figure 5. Calibration curve of electrical conductivity with NMP in PBS.

Figure 6. Electrical conductivity curves of organogel without NMP (▪) and organogel contained NMP (•).

In vitro release evaluation

The release behaviors of oily solution and organogel formation were determined in vitro at the drug-loading amount of 7.11 mg/mL. Results (Figure 7) indicated that the organogel formulation behaved like a semi-solid drug depot, which could extend the release of drug from 36 h to almost 15 d. Specially, the release of CC from the control oily solution reached almost 87.0% in the first 24 h, which was only 28.2% for CC-loaded organogel. However, there still existed initial burst of drug to some extent, which was a ubiquitous phenomenon for this in situ drug delivery system due to the lag time between the injection of formulation solution and the formation of semi-solid implant (Bastiat et al., 2010). The percent prediction values of Y_24h and Y_15d were 27.6% and 92.9%, respectively, which were 28.2% and 92.4% as actual, suggesting a good prediction ability of the model equation.

Figure 7. In vitro release profiles of oil solution (•) and organogel formulation (▪).

In vivo pharmacokinetic study of CC formulation

The optimal organogel formulation and the oily drug solution were subjected for further pharmacokinetic evaluation in male rats. CC was hydrolyzed to active candesartan CDS in body, which can be detected in plasma. The pharmacokinetic parameters determined from the plasma concentration-time date were listed in Table 3. As shown in Figure 8, CC was completely released from the control oily solution within 2 d, however, the organogel formulation exhibited a 10-d plasma profile. According to the results, the injection of organogel resulted in a significant (p < 0.05) decrease of the plasma drug level and a significant (p < 0.05) prolongation of the release time, where showed C_max of 16.086 μg/mL at T_max of 5.2 h versus 20.48 μg/mL at 2.0 h, respectively. The AUC_(0−∞) values of CC organogel formulation exceeded by about 3-fold than that of the oily solution and the MRT prolonged from 6.5 h to 45.62 h. The in vitro–in vivo correlation (IVIVC) result was as follows:

Figure 8. In vivo plasma concentration–time curves of drug oily solution (▪) and organogel formulation (•).

Table 3. Pharmacokinetic parameters obtained after subcutaneous injection of oily solution and organogel formulation in rats (mean ± SD; n = 5).

Formulations	Cmax (μg/mL)	Tmax (h)	t1/2 (h)	AUC0-∞ (μg/mL/h)
Oily solution	20.48 ± 0.52	2 ± 0.50	6.50 ± 1.69	192.05 ± 25.56
CC organogel	16.07 ± 1.45	5.2 ± 1.79	45.627 ± 21.06	561.67 ± 119.38

F_a is the drug absorbed percentage in vivo and F_r represents the fraction release in vitro. The results showed that the in vivo fraction absorbed could be predicted from the in vitro release data. As reported by Shimizu et al. (1999), the plasma concentration of CDS needed to be higher than 0.05 μg/mL for an adequate therapeutic effect. In our study, the in vivo pharmacokinetic results suggested that at least 10 d was set a dosage interval for one subcutaneously injection of CC organogel. Thus, the in situ implant organogel could provide an opportunity for the delivery of these drugs with low solubility.

In addition, the release kinetics of organogel formulation in vivo was described by the Ritger–Peppas model. Correlation coefficient (R²) of 10-d drug release profile (Figure 9) was 0.9690, which indicated that the model fitted the release behavior of organogel system well. The diffusion index was 0.5092, suggesting the drug release behavior was controlled by both drug diffusion and frame erosion. As shown in the section “Gelation kinetics determination”, it took about 40 min to form a stable organogel in body. Upon injection, the drug inclined to diffuse from the oily solution and this effect would be weakened with the formation of organogel, which may be responsible for the initial release in early time. At the same time, there would be some pores formed along with the diffusion of NMP to the surrounding medium and the drugs could easily pass through these pores. On the other hand, drugs would certainly release along with organogel, which degraded in body under the action of enzyme. Thus, the release behavior of this system was a combined action of diffusion and erosion.

Figure 9. In vivo cumulative release-time curve of CC organogel formulation.

Histological analysis

The biocompatibility of the organogel formulation was evaluated by the histological analysis. Figure 10 shows the transverse section of normal subcutaneous tissue and tissues surrounding the injection site, which was harvested after 10 d. As expected, there was a moderate inflammation reaction due to the implantation of the organogel. The existence of fibroblast proliferation surrounding the injection site indicated that the inflammatory reaction began to heal. This phenomenon was commonly observed around the well-known foreign body reaction as a normal wound healing response to biocompatible and biodegradable materials (Anderson & Shive, 1997; Motulsky et al., 2005). Overall, this suggested a reasonably good compatibility of organogel as an in situ forming implant.

Figure 10. Histological evaluation of blank group (A) and adjacent tissues injected with the formulation after 10 d (B). Original magnification: 400×.

Conclusion

In this study, we first synthesized a MDP, possessing two of the same relatively long alky chains, which showed a better gelation ability due to its well hydrophobic–hydrophilic balance of the molecule. It could form thermoreversible organogels in oils at critical gelator concentration with higher transition temperatures and better mechanical properties. In vivo pharmacokinetic study revealed that organogel formulation can keep an effective therapeutic concentration at least 10 d and increase the bioavailability greatly. Also, the study of release mechanism of the Ritger–Peppas model showed that the release behavior was a combined action of drug diffusion and frame erosion. In all, the organogel based on this new synthesized gelator was well tolerated and could be used as a carrier for lipophilic drug for parenteral application.

However, to be a suitable in situ implant system, further investigation is needed to reduce the initial burst release of drug and investigate the relationship between degradation and release, in order to control the release of drug in step with organogel degradation.

Declaration of interest

The authors report no conflicts of interest.

This research work is financially supported by the National Natural Science Foundation of China (No. 81273445).

Supplementary material available online