- Email: [email protected]
Polysorbate 20 vesicles as oral delivery system: In vitro characterization
Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
Contents lists available at SciVerse ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Polysorbate 20 vesicles as oral delivery system: In vitro characterization Luisa Di Marzio a,∗ , Sara Esposito a , Federica Rinaldi b , Carlotta Marianecci b , Maria Carafa b a b
Dip. di Farmacia, University “G. d’Annunzio”, Chieti-Pescara, Italy Dip. di Chimica e Tecnologie del Farmaco, University “Sapienza”, Rome, Italy
a r t i c l e
i n f o
Article history: Received 2 August 2012 Received in revised form 19 October 2012 Accepted 23 October 2012 Available online 20 December 2012 Keywords: Pegylated surfactant vesicles Mucoadhesion Gastrointestinal stability
a b s t r a c t Non-phospholipid vesicles made with non-ionic surfactants represent a promising alternative to the more widely studied liposomes. The main aim of the present work is to evaluate if vesicles of polysorbate 20 may be used as delivery systems for oral administration of drugs. Then in vitro stability and mucoadhesion studies in simulated gastrointestinal fluids were carried out. The colloidal stability of the surfactant vesicles was determined by size and fluorescence-dequenching assay, while their mucoadhesive properties were evaluated by light-scattering and protein assay. The results of in vitro stability demonstrated that the pHs and enzymes (pepsin and/or pancreatin) of the gastrointestinal fluids had not influence on surfactant vesicle stability. However, in presence of bile salts the nanosize vesicles showed a release of fluorescent marker (about 11% at 2 h and 28% at 4 h), whereas they were stable in size as confirmed by the light scattering experiments. Finally, the in vitro mucoadhesive experiments showed that the capacity of nanovesicles to adsorb mucin was higher at neutral pH than at acidic pH. As a conclusion of these preliminary studies, the surfactant vesicles could be considered a versatile tool for the oral delivery of drugs with poor stability in gastrointestinal tract and low permeability. Nevertheless, further work is required in order to examine the interaction with and/or the transport route through the epithelial cells of the gastrointestinal wall. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Most front-line therapeutic agents currently in use are untargeted, toxic compounds that act in a nonspecific way, often eliciting unwanted, dose-limiting, and often debilitating side effect. The development of an efficient, versatile and targeted delivery vehicle of bioactive molecules can bring a potential and complete solution to disease outcomes. The growing work in the field of drug delivery over the past 30 years has been aimed to the development of various safe nanosize delivery systems. Among different colloidal delivery carriers, vesicular systems such as liposomes have been investigated more than the other systems. Liposomes are mainly designed for parenteral use in order to protect labile drugs against environmental aggression and to achieve high concentration of pharmaceutical compounds in a specific area of body [1,2]. However, several drawbacks exist with the use of liposomes such as in vitro [3] and in vivo [4] instability, particularly in the gastrointestinal tract (GIT). After oral administration the lipid vesicles are easily degraded not only by bile salts, but also by pancreatic enzymes and by the low pH present in the gut [5,6]. This
∗ Corresponding author at: Università “G. D’Annunzio”, Chieti-Pescara, Dipartimento di Farmacia, Via dei Vestini, 66100 Chieti, Italy. Tel.: +39 0871 3554705/3554718; fax: +39 0871 3554911. E-mail address: [email protected] (L. Di Marzio). 0927-7765/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2012.10.036
gastrointestinal degradation leads both to a reduction of liposome structures reaching the bowel and to release the encapsulated bioactive compounds into the GIT. To improve liposomal membrane integrity in the GIT a possible approach is coating them with polymers such as chitosan, pectin and polyethylene glycol (PEG) [7]. Polyethylene glycols are hydrophilic, non-ionic and nontoxic polymers, and they have been extensively investigated for pharmaceutical and medical purpose. Indeed it has been shown that in parenteral administration, the PEG reduces the interaction of different nanoscale systems with opsonins and slows down their capture by the RES, leading to longer blood circulation halflives for the nanovectors in vivo [8]. Moreover, in recent years it has been reported that the PEG coating of nanosystems promotes and enhances the specific bioadhesive interactions with mucosal tissues due to the ability of PEG chains to diffuse across mucus network enhancing interpenetration [9,10]. Thus pegylated nanovectors can be considered as suitable systems to improve the bioavailability of drugs by prolonging the residence time of the bioactive molecules on the absorptive epithelium and controlling drug release properties. In general, it is established that the interactions between a polymer and the mucous layer can be due to physical or mechanical bonds, secondary chemical bonds and covalent chemical bonds. Secondary chemical interactions include ionic bonds, van der Waals interactions and hydrogen bonding. The functional groups that form hydrogen bonds are hydroxyls, carboxyls, sulfate and amino groups. Even these types of forces are
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
weaks, numerous interaction sites lead to strong mucoadhesion. Indeed one of the characteristics that seems to increase mucoadhesion phenomenon is the presence of hydrogen bonding [11]. A possible approach to obtain a specific bioadhesive drug delivery system could be the hydration of a mixture of cholesterol and non-ionic surfactants such as alkyl esters or ether surfactant, in which the hydrophilic region consists of repeated oxyethylene units. The vesicles formed from the non-ionic amphiphiles are known as niosomes or non-ionic surfactant vesicles (NSVs). They are analogous in terms of physical properties to phospholipid vesicles, but they possess several advantages over them. The main advantages lie in their chemical stability, lower toxicity, improved therapeutic performance and easier handling and storage. Another advantage of surfactant amphiphiles is their relatively low cost, which makes them attractive for industrial production both in pharmaceutical and cosmetic applications. Moreover niosome preparation for the routine and large-scale production can be performed without using pharmaceutically unacceptable solvents [12,13]. The NSV structural properties allow the entrapment of drugs with different lipophilicities: strongly lipophilic drugs are entrapped almost completely in the bilayer, strongly hydrophilic drugs are located exclusively in the aqueous compartment, and drugs with intermediate log P easily distribute between the lipid and aqueous phases, both in the bilayer and in the aqueous core [14]. Although pharmaceutical niosome formulations have yet to be commercially exploited, a number of studies demonstrated the potential of surfactant vesicles in drug delivery. Examination of the literature reveals that surfactant drug delivery systems play a significant role in formulation of bioactive molecules to improve safety and efficacy to administration of several classes of molecules like anti-neoplastic, anti-viral, anti-inflammatory, anti-microbial and proteins [15–21]. Based on these considerations, the aim of the present work was to investigate the stability of NSVs, composed by polisorbate 20, in the artificial gastrointestinal fluid. The polisorbate 20 (Tween® 20 or Tw20) is a non-ionic surfactant, the stability and relative non-toxicity of which, allows it to be used as excipient for oral administration. It is polyoxyethylene sorbitan monolaurate, in which the hydrophilic region consist of 20 oxyethylene units. The stability of Tw20 vesicles was evaluated in simulated digestive fluids in terms of release of the encapsulated fluorescent probes and size. As a second step, the interaction between non-ionic surfactant vesicles and mucin was evaluated to asses the mucoadhesive properties of polisorbate 20, particularly, in the surfactant vesicles. 2. Materials and methods 2.1. Materials Tween 20® (Tw20), cholesteryl hemisuccinate (CHEMS), Sephadex G-75, Hepes salt {N-(2-idroxyethyl), piperazine-N-(2ethanesulfonic acid)}, p-Xylene-bis (N-pyridinium bromide) (DPX), pancreatin (from porcine pancreas, 4XUSP), pepsin (from porcine gastric mucosa, 800–2500 UI/mg protein), and bile salts were Sigma–Aldrich products (Sigma–Aldrich SRL, Milan, Italy). Cholesterol (CHOL) and hydroxypyrene-1,3,6-trisulfonic acid (HPTS) were obtained from Acros Organics (Acros Organics BVBA, Geel, Belgium). Bovine submaxillary gland mucin was obtained from Merck Chemicals (United Kingdom). Bio-Rad protein assay was BioRad Laboratories (Bio-Rad Laboratories S.r.l., Milan, Italy). All other products and reagents were of analytical grade. 2.2. Preparation and characterization of non-ionic surfactant vesicles Non-ionic surfactant vesicles (NSVs) were prepared using different amounts of Tw20, CHOL and CHEMS (Table 1). Tw20
201
Table 1 Surfactant vesicle composition. Sample
Tw20 (mM)
CHOL (mM)
CHEMS (mM)
1 2
15 15
15 2
– 13
concentration was always remarkably above CMC (0.048 mM in water, at 20 ◦ C). The vesicles were obtained by the “film” method, as previously reported [22]. The dried films were hydrated by addition of HEPES buffer (10 mM, pH 7.4) alone or Hepes buffer (10 mM, pH 7.4) containing equimolar ratio HPTS/DPX (30 mM/30 mM). The surfactant dispersion was mechanically stirred for about 5 min and then sonicated for 10 min at 60 ◦ C (UP200H, Hielscher, Teltow, Germany) equipped with an exponential microprobe operating at 24 kHz and an amplitude of 60%. Vesicle dispersions were purified by size exclusion chromatography on Sephadex G75 glass columns. To determine size (nm) and -potential (mV) of the NSVs, vesicle dispersions were diluted 1:100 and 1:10 in HEPES buffer for size and -potential measurements, respectively. Vesicle size distribution and -potential were measured by dynamic light scattering (DLS) on a Malvern Zetasizer Nano ZS90 (Malvern Instruments Ltd., Worcestershire, United Kingdom) at 20 ◦ C. It should be pointed out that size distribution results were determined as % of intensity of the colloidal dispersion. The polydispersity index (p.i.) value was determined as a measurement of the width of the size distribution: a p.i. value lower than 0.3 indicates a homogenous and monodisperse population [23]. The surfactant structured in the vesicles was determined by the spectrophotometric method as previously reported [24]. 2.3. SGF and SIF Simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were made according to USP XXIV [25] with some modifications. The Table 2 summarize the compositions of the simulated gastric and intestinal media used in this study. 2.4. pH-stability studies In order to evaluate the stability of niosomes in presence of digestive fluids, 20 l of purified vesicles containing entrapped dies (HPTS/DPX) were incubated in media prepared according to the recommendations given in USP XXIV at 37 ◦ C. Samples were collected at times 0, 1, 2, 4 h. The SIF1 and SIF2 samples were centrifuged for 10 min at 13,600 × g and 4 ◦ C to eliminate pancreatin aggregates in medium. The stability of surfactant vesicles in simulated gastric and intestinal fluids was estimated by DLS measurements and evaluating the leakage of water-soluble fluorescent dyes, able to change their fluorescent behaviour upon release. In this study, the leakage of aqueous markers was assayed according to Smolarsky et al. [26] with the fluorescent probe HPTS and its quencher DPX entrapped in the surfactant vesicles. When the HPTS-DPX were released from vesicles, they dissociated into free HPTS and DPX, producing an increase in HPTS fluorescence revealed at 510 nm. The release of entrapped HPTS from the surfactant vesicles was determined by means of a fluorescence-dequenching assay using a LS55 spectrofluorometer (PerkinElmer, Massachusetts, USA) at excitation and emission wavelengths of 403 and 510 nm, respectively. The extent of HPTS release was expressed as a percentage and calculated according to the following equation: % HPTS release =
Fx − F0 × 100 Ft − F0
(1)
202
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
Table 2 Composition of the media to simulate gastric (SGF) and intestinal (SIF) fluids according to USP XXIV. Media composition
SGF
SGF1
SIF
SIF1
SIF2
Sodium chloride Sodium monobasic phosphate anhydrous Chloridic acid Sodium hydroxide Pepsin Pancreatin Bile salts pH
34.2 mM – 84 mM – – – – 1.2
34.2 mM – 84 mM – 2560 U/ml – – 1.2
– 50 mM – 25 mM – – – 6.8
– 50 mM – 25 mM – 250 U/ml – 6.8
– 50 mM – 25 mM – 250 U/ml 5 mM 6.8
where Fx is the fluorescence intensity of HPTS in the sample incubated in different media and/or different time, F0 is the fluorescence intensity of HPTS loaded niosomes in different media at room temperature at time = 0, Ft is the total fluorescence intensity of HPTS loaded surfactant vesicles in SIF after addition of isopropanol to induce vesicular structure rupture. The samples had different surfactant concentration of 1.2 M and 11.4 M for sample 1 and 2 respectively. The samples were maintained at room temperature during the measurements. The in vitro stability of NSVs in gastrointestinal fluid was also determined evaluating size variations and colloidal stability of vesicles by means of DLS.
where M0 is the glycoprotein total concentration incubated with nanosystems and ML is the free mucin amount after incubation and centrifugation. 2.7. Statistical analysis Results are expressed as the mean of three experiments ± S.E.M. (standard error of the mean). Statistical data analysis was performed using the t-test. To evaluate if obtained differences in experimental results were statistically significant, p ≤ 0.05 was used as the significance criterion. 3. Results
2.5. Turbidity measurement
3.1. Characterization of surfactant vesicles
Aggregation of nanosize vesicular samples was monitored as an increase in turbidity by right angle light scattering (excitation and emission = 600 nm) using a LS55 fluorimeter (PerkinElmer, USA). The samples were maintained at 25 ◦ C during the measurements using surfactant concentrations about 11 M for both samples.
The surfactant vesicles were prepared, as previously reported [22,27] by the film method associated with sonication in order to reduce the size down the sub-micron range. As reported in Table 3, the percentage of structured surfactant is 40 ± 4% and 38 ± 2% for sample 1 and 2, respectively. In addition vesicles are characterized
2.6. Mucoadhesive studies 2.6.1. Preparation of mucin solution The commercially available bovine submaxillary gland mucin was hydrated in 10 mM Hepes pH 7.4 at 4 ◦ C overnight. The mucin solution was centrifuged at 15,557 × g for 30 min at 4 ◦ C and then filtered through a 0.8 m cellulose nitrate filters prior to use. The final mucin concentration in solution was determined through the BioRad protein assay (Hercules, CA) using the bovine serum albumin standards.
2.6.2. In vitro interaction between vesicles and mucin The adsorption of mucin on vesicle surface was used as a method to assess mucoadhesive properties of the prepared surfactant vesicle. The interaction was studied by incubating mucin (0.5 mg/ml) and vesicles (surfactant concentration in the dispersion was fixed at 0.98 mg/ml) at pH 1.2 (SGF) and 6.8 (SIF). The incubation was carried out under magnetic stirring (700 rpm) at temperature of 37 ± 1 ◦ C for 4 h. Then the dispersion were centrifuged at 21,000 × g and 4 ◦ C for 30 min. After centrifugation the pellet were resuspended in H2 O and used to determine the vesicle size as described in Section 2.2, whereas the supernatant was collected and used for the determination of free mucin using Bio-Rad protein assay. In the protein assay the amount of mucin adsorbed on the surfactant vesicle surface was determined as a difference between its initial concentration and the concentration in the dispersion after incubation and centrifugation, according to Eq. (2): % bounded mucin =
M0 − ML × 100 M0
(2)
Fig. 1. In vitro stability of surfactant vesicles at the pH range of the gastrointestinal tract. The stability of surfactant vesicles (sample 1: Tw20:CHOL = 15 mM:15 mM; sample 2: Tw20:CHOL:CHEMS = 15 mM:2 mM:13 mM), incubated at pH 1.2 (SGF) and pH 6.8 (SIF) as described in Section 2, was evaluated by the release of encapsulated HPTS (A) and turbidity increase (B). The data, reported as percentage of HPTS release, was determined by Eq. (1). AU = arbitrary unit. The showed data represented the mean obtained from three independent experiments ± S.E.M. *p < 0.05 with respect to vesicles incubated in SIF.
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
203
Table 3 Characterization of surfactant vesicles. Percentage of non-ionic surfactant structured (%), vesicle dimensions (diameter), polydispersity index (p.i.) and -potential values. Reported data were means of three experiments ± S.E.M. Sample
%
Size (nm)
-Potential (mV)
p.i.
Sample 1 Sample 2
40 ± 4 38 ± 2
159.63 ± 8.19 156.43 ± 2.52
−32.89 ± 2.34 −29.83 ± 0.33
0.15 ± 0.03 0.29 ± 0.01
by dimensions in the sub-200 nm size irrespective of the composition, confirming our previous data [28]. Furthermore as shown in Table 3, both samples had a polydispersity index less than 0.3, value that indicated an homogeneity and monodispersity of the obtained vesicular population. The measurements carried out by DLS also showed that the entrapment of fluorescent probe HPTS and its quencher DPX did not cause a significant increase in surfactant vesicle size with respect to the corresponding empty ones, showing that the fluorescent probe and quencher were trapped within the surfactant samples (data not shown). The nanovesicles were negatively charged as reflected in the -potential values (≈−30 mV), which indicated the stability of the prepared samples [29]. The absolute negative value of -potential is likely attributed to the presence of the oxyethylene units on the vesicular surface [30].
Fig. 3. Effect of pancreatic lipase and bile salts on the stability of surfactant systems. The niosomes were incubated in simulated intestinal fluids in absence (SIF) and in presence of pancreatin (SIF1) or bile salts (SIF2) at 37 ± 1 ◦ C for different time. At specified times the fluorescence intensity and size of sample were recorded as described in Section 2.4. (A) HPTS release: the data of probe leakage from vesicles were expressed as percentage and were calculated using Eq. (1). The results were reported as mean of three independent experiments and S.E.M. values were always lower than 10% of the mean value. *p ≤ 0.05 with respect to vesicles incubated in SIF. (B) Size: the particle dimensions (nm) were evaluated by dynamic light scattering. The data of size were shown as mean ± S.E.M. obtained from three independent experiments.
3.2. Stability of surfactant vesicles in model of gastrointestinal fluids
Fig. 2. Effect of pepsin on the stability of non-ionic nanovesicles. The vesicles were incubated in gastric media in presence (SGF1) and in absence (SGF) of pepsin at 37 ± 1 ◦ C for different time. At specified times the fluorescence intensity and size of sample 1 were determined as described in Section 2.4. (A) HPTS release: the probe release from vesicles was expressed as percentage and was calculated using Eq. (1). The data were expressed as mean of three independent experiments and S.E.M. values were always lower than 10% of the mean value. (B) Size: the particle dimensions (nm) were evaluated by dynamic light scattering. The results were the mean of three independent experiments ± S.E.M.
Keeping in mind that the main purpose of our work was to design a drug delivery vesicular system recommend for oral administration, its stability in simulated digestive fluids should be evaluated. It is known that the gastric pH (1.2) is a key factor in the destabilization of vesicular systems, therefore their stability at pH 1.2 (SGF) and 6.8 (SIF) was evaluated measuring the release of entrapped HPTS and turbidity after 1 h of incubation. Fig. 1A reported the percentages of HPTS release from both surfactant nanovesicles at two different pH after 1 h incubation. As shown by the data reported in Fig. 1A, the percentage of HPTS release was quite similar at both different pH for samples 1, while it was significantly (p < 0.05) different at two pH for sample 2. In fact the percentage of released probe for sample 2 raised from 3 ± 2% to 15 ± 1% after the pH reduction from 6.8 to 1.2. Moreover no significant change in percentage of HPTS release from both sample was observed when HPTS release was compared at pH 6.8 and at the neutral pH 7.4 (data not shown). Furthermore the stability of both non-ionic surfactant vesicles was followed by turbidity measurements. As expected no changes in turbidity were observed in the case of sample 1 whereas the sample 2 showed a 1.5-fold increase in turbidity at pH 1.2 compared to 6.8 (Fig. 1B). In the present work
204
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
Fig. 4. Size profiles of surfactant vesicles incubated in intestinal fluid containing bile salts. The dimension of sample were determined by DLS analysis after incubation in intestinal fluid in absence and in presence of bile salts at different times. (A) Niosomes incubated in absence of bile salts (SIF), (B) vesicles incubated in presence of bile salts (SIF2) for 2 h (T2), and (C) vesicles incubated in presence of bile salts (SIF2) for 4 h (T4). The reported DLS data were triplicate analyses from a representative experiments of three experiment giving similar results. The S.E.M were ever lower than 10% of the mean value.
sample 2 was chosen as reference pH-sensitive system because its stability was dependent on the environmental pH [31]. From the collected data, sample 1 seemed to be stable at gastric pH in contrast with previously reported data on different liposomal formulations, showing a nearly 100% release of carboxyfluorescein after just 10 min at gastric pH [32]. According to the evidence that sample 1 was stable in the pH range of the gastrointestinal tract, the effect of enzymes (pepsin or pancreatin) and or bile salts was also evaluated. The stability
of sample 1 incubated in gastrointestinal fluids containing digestive compounds was compared with that of sample incubated in absence of gastrointestinal fluid constituents. In Fig. 2A the percentages of HPTS released by surfactant vesicles (sample 1) incubated in gastric fluid in presence (SGF1) and absence of pepsin (SGF) for 4 h were reported. The results clearly showed that nanovesicles were not subjected to enzymatic degradation. In addition the DLS data (Fig. 2B) were in close agreement with those obtained by HPTS release. The collected results indicated that neither the acidic
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
pH nor the gastric enzyme had effect on the vesicular integrity in simulated gastric fluid. Similar results were obtained when the surfactant vesicles were incubated in intestinal media in presence (SIF1) or absence (SIF) of the pancreatic lipase (Fig. 3A and B). Consequently, it seems reasonable to conclude that the oxyethylene units of hydrophilic region of Tw20 on the vesicle surface prevent breakage of the ester bond between the carboxyl group of the fatty acid and the hydroxyl group of the alcohol of Tw20 moiety by the pancreatic lipase. These data are in agreement with those previously reported on the stability of PLA or PA nanosystems coated with PEG in gastrointestinal media [33,34]. However, different behaviour was observed when stability studies were performed in intestinal fluid containing bile salts (SIF2). As showed in Fig. 3A in presence of bile salts the HPTS released from nanosize samples was unchanged during the first 60 min of incubation, while a relatively high amount of dye released from sample 1 was detected after 1 h (Fig. 3A). The percentage of dye released was approximately the 11% at 2 h and 28% at 4 h, suggesting that the stability was dependent on the incubation time (p ≤ 0.05). The bile salts concentration is above the critical micelle concentration and the time dependent stability could be related to an interaction leading to mixed vesicles [35] that might cause a leakage of HPTS. On the contrary, the mean size of samples remained unchanged (Fig. 3B), indeed the DLS data did not show any evidence of the micelle formation over the time. Fig. 4 illustrates the distribution profiles obtained from a representative experiment of vesicle stability in SIF2 at different times. Incubation of sample 1 in presence of bile salts, after incubation at 2 h and 4 h (Fig. 4B and C) produced vesicle size distribution and p.i. (<0.2) comparable to those of niosomes incubated in the absence of bile salts (Fig. 4A). Considering the size stability, indicating that the vesicles were not destroyed, the HPTS leakage from surfactant sample in SIF2 could be attributed to small membrane defects. This behaviour was seen previously for liposomes which stayed intact at low pH and leakage of small molecules was pronounced [36,37].
3.3. Mucoadhesive properties of surfactant vesicles It was reported that grafting of PEG into polymeric hydrogels enhanced the mucoadhesion [9,10]. Mucoadhesive behaviour of sample 1 was assessed in vitro by the suspension of sample 1 in a mucin aqueous solution at pH 1.2 (SGF) and 6.8 (SIF), and the amount of glycoprotein adsorbed onto vesicles and size was determined as reported in Section 2.6.2. The results of mucoadhesion are presented in Fig. 5. The DLS analyses showed a significant increase of size with respect to initial value of nanovesicles for the sample incubated with mucin at both pHs investigated. The increase in particle dimension was about 10% at pH 1.2 and 26% at pH 6.8. Indeed, sample size was 159 ± 2 nm in absence of glycoprotein and was significantly (p ≤ 0.05) increased to 175 ± 4 nm and 200 ± 3 nm after incubation in mucin solution at pH 1.2 and pH 6.8, respectively. To obtain further insight into the effect of pH on glycoprotein adhesion onto surfactant systems, the determination of mucin adsorbed to nanovesicles was also investigated using protein assay. The data obtained by Bio-Rad assay indicated that the percentage of mucin bound to the surface of nanosystem varied in a significant way (p ≤ 0.05) from 9 ± 1.5% at pH 1.2 to 20 ± 2.4% at pH 6.8, suggesting that a simulated medium at pH 6.8 respect to pH 1.2 had a more marked effect (p ≤ 0.05) on mucin interaction with nanosize vesicles. This is in agreement with previous data, which demonstrated that pegylated particles possessed higher affinity to adhere to intestinal than to the gut mucosa [38]. All together these results suggested that mucin amount bounded to the surface of surfactant vesicles could be due to the combination of two factor:
205
Fig. 5. Mucoadhesive properties of non-ionic surfactant vesicles. The interaction between vesicles and glycoprotein was studied by incubating sample 1 and mucin (0.5 mg/ml) at pH 1.2 (SGF) and 6.8 (SIF) for 4 h at 37 ± 1◦ C. The data, obtained as reported in material and methods, were expressed as bounded mucin and as the % of dimension increase respect to initial size of nanovesicles. In DLS determination the initial dimension of vesicles was set as 100%, whereas in the protein assay the adsorbed mucin was calculated by Eq. (2). The results were shown as black columns (initial size vesicle), white columns (sample incubated in SGF) and grey columns (sample incubated in SIF). The reported data were the mean ± S.E.M of triplicate analyses from a representative experiment of three experiments giving similar results. *p ≤ 0.05 with respect to initial size of nanovesicles and § p ≤ 0.05 with respect to nanovesicles incubated at pH 1.2.
1. high concentration of hydroxyl group at the end of PEG chains on the vesicular surface. In general, it is established that the interactions between a polymer and the mucous layer can be due to physical or mechanical bonds, secondary chemical bonds and covalent chemical bonds, in particular one of the characteristics that seems to increase mucoadhesion phenomenon is the presence of hydrogen bonding [11]. 2. Molecular weight of PEG (PM ≈ 900). Indeed, Yoncheva et al. reported that the ability of nanosystems to adsorb mucin at neutral pH was higher for PEG 1000 than for PEG 2000; the lower interaction between mucosal glycoprotein and PEG 2000nanosystems was due to the steric hindrance of PEG “brush” layer on the surface of the nanoparticles [38].
4. Conclusion Challenges for oral mucoadhesive systems include the harsh environment of the stomach, which, due to its low pH, results in an activation of a wide range of drugs as well as the degradation of protein drugs administrated for therapeutic purposes. On the other hand, the wash out effect in the GI tract due to the intestinal motility results in a low residence time of the drug at the site of absorption. For these reasons, designing a stable and mucoadhesive delivery system could be of great interest. In the present, from in vitro simulations, it could be evidenced that Tween® 20 vesicles show good stability in SIF and SGF, also in presence of digestive enzymes and bile salts, in fact the release of fluorescent probe is most likely related to small membrane damages but not to a disintegration of the vesicles indicated by size stability. Furthermore, pegylated vesicles possessed the ability to adhere to mucin in intestinal medium rather than in gastric one. Irrespective of the limited value of the in vitro results, it could be underlined that the reported data could be helpful in designing innovative drug delivery systems for oral administration. However, the mechanism of interaction and transport of non ionic surfactant vesicles through the gastrointestinal epithelium have to be clarified, together with toxicity studies.
206
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
References [1] T.A. Elbayoumi, V.P. Torchilin, Methods Mol. Biol. 605 (2010) 1. [2] M.S. Webb, P. Rebstein, W. Lamson, M.B. Bally, Recent Pat. Drug Deliv. Formul. 1 (2007) 185. [3] M. Ausborn, P. Nuhn, H. Schreier, Eur. J. Pharm. Biopharm. 38 (1992) 133. [4] G. Poste, Biol. Cell 47 (1983) 19. [5] K. Iwanaga, S. Ono, K. Narioka, M. Kakemi, K. Morimoto, S. Yamashita, Y. Namba, N. Oku, J. Pharm. Sci. 88 (1999) 248. [6] K. Iwanaga, S. Ono, K. Narioka, K. Morimoto, M. Kakemi, S. Yamashita, M. Nango, N. Oku, Int. J. Pharm. 157 (1997) 73. [7] H. Takeuchi, Y. Matsui, H. Sugihara, H. Yamamoto, Y. Kawashima, Int. J. Pharm. 303 (2005) 160. [8] P. Aggarwal, J.B. Hall, C.B. McLeland, M.A. Dobrovolskaia, S.E. McNeil, Adv. Drug Deliv. Rev. 61 (2009) 428. [9] N.V. Efremova, Y. Huang, N.A. Peppas, D.E. Leckband, Langmuir 18 (2002) 836. [10] L. Serra, J. Doménech, N.A. Peppas, Eur. J. Pharm. Biopharm. 63 (2006) 11. [11] L. Serra, J. Doménech, N.A. Peppas, Eur. J. Pharm. Biopharm. 71 (2009) 519. [12] A. Manosroi, R. Chutoprapat, M. Abe, J. Manosroi, Int. J. Pharm. 352 (2008) 248. [13] T.T. Pham, C. Jaafar-Maalej, C. Charcosset, H. Fessi, Colloids Surf. B: Biointerfaces 94 (2012) 15. [14] I.F. Uchegbu, S.P. Vyas, Int. J. Pharm. 172 (1998) 33. [15] R. Muzzalupo, F.P. Nicoletta, S. Trombino, R. Cassano, F. Iemma, N. Picci, Colloids Surf. B: Biointerfaces 58 (2007) 197. [16] C. Marianecci, D. Paolino, C. Celia, M. Fresta, M. Carafa, F. Alhaique, J. Control. Release 147 (2010) 127. [17] S.K. Mehta, N. Jindal, G. Kaur, Colloids Surf. B: Biointerfaces 87 (2011) 173. [18] S. Akhter, S. Kushwaha, M.H. Warsi, M. Anwar, M.Z. Ahmad, I. Ahmad, S. Talegaonkar, Z.I. Khan, R.K. Khar, F.J. Ahmad, Drug Dev. Ind. Pharm. 38 (2012) 84. [19] M. Bragagni, N. Mennini, C. Ghelardini, P. Mura, J. Pharm. Pharm. Sci. 15 (2012) 184. [20] L. Di Marzio, C. Marianecci, F. Rinaldi, S. Esposito, M. Carafa, Lett. Drug Des. Discov. 9 (2012) 494.
[21] C. Marianecci, F. Rinaldi, M. Mastriota, S. Pieretti, E. Trapasso, D. Paolino, M. Carafa, J. Control. Release 164 (2012) 17. [22] M. Carafa, C. Marianecci, G. Lucania, E. Marchei, E. Santucci, J. Control. Release 95 (2004) 67. [23] V. Centis, P. Vermette, Colloids Surf. B: Biointerfaces 65 (2008) 239. [24] H. Kato, Y. Nagai, K. Yamamoto, Y. Sakabe, J. Assoc. Off. Anal. Chem. 72 (1989) 27. [25] US Pharmacopeia XXIV, 2006. [26] M. Smolarsky, D. Teitelbaum, M. Sela, C. Gitler, J. Immunol. Methods 15 (1977) 255. [27] E. Santucci, M. Carafa, T. Coviello, E. Murtas, F.M. Riccieri, F. Alhaique, A. Modesti, A. Modica, STP Pharm. Sci. 6 (1996) 29. [28] M. Carafa, L. Di Marzio, C. Marianecci, B. Cinque, G. Lucania, K. Kajiwara, M.G. Cifone, E. Santucci, Eur. J. Pharm. Sci. 28 (2006) 385. [29] L. Di Marzio, C. Marianecci, M. Petrone, F. Rinaldi, M. Carafa, Colloids Surf. B: Biointerfaces 82 (2011) 18. [30] R. Gref, A. Domb, P. Quellec, T. Blunk, R.H. Muller, J.M. Verbavatz, R. Langer, Adv. Drug Deliv. Rev 16 (1995) 215. [31] L. Di Marzio, C. Marianecci, B. Cinque, M. Nazzarri, A.M. Cimini, L. Cristiano, M.G. Cifone, F. Alhaique, M. Carafa, Biochim. Biophys. Acta 1778 (2008) 2749. [32] J. Parmentier, M.M. Becker, U. Heintz, G. Fricker, Int. J. Pharm. 405 (2011) 210. [33] M. Tobío, A. Sánchez, A. Vila, I.I. Soriano, C. Evora, J.L. Vila-Jato, M.J. Alonso, Colloids Surf. B: Biointerfaces 18 (2000) 315. [34] A. Sahu, U. Bora, N. Kasoju, P. Goswami, Acta Biomater. 4 (2008) 1752. [35] P. Walde, J. Sunamoto, C.J. O’Connor, Biochim. Biophys. Acta 905 (1987) 30. [36] G.B. Patel, B.J. Agnew, L. Deschatelets, L.P. Fleming, G.D. Sprott, Int. J. Pharm. 194 (2000) 39. [37] M. Taira, N. Chiaramoni, K. Pecuch, S. Alonso-Romanowski, Drug Deliv. 11 (2004) 123. [38] K. Yoncheva, E. Lizarraga, J.M. Irache, Eur. J. Pharm. Sci. 24 (2005) 411.
Contents lists available at SciVerse ScienceDirect
Colloids and Surfaces B: Biointerfaces journal homepage: www.elsevier.com/locate/colsurfb
Polysorbate 20 vesicles as oral delivery system: In vitro characterization Luisa Di Marzio a,∗ , Sara Esposito a , Federica Rinaldi b , Carlotta Marianecci b , Maria Carafa b a b
Dip. di Farmacia, University “G. d’Annunzio”, Chieti-Pescara, Italy Dip. di Chimica e Tecnologie del Farmaco, University “Sapienza”, Rome, Italy
a r t i c l e
i n f o
Article history: Received 2 August 2012 Received in revised form 19 October 2012 Accepted 23 October 2012 Available online 20 December 2012 Keywords: Pegylated surfactant vesicles Mucoadhesion Gastrointestinal stability
a b s t r a c t Non-phospholipid vesicles made with non-ionic surfactants represent a promising alternative to the more widely studied liposomes. The main aim of the present work is to evaluate if vesicles of polysorbate 20 may be used as delivery systems for oral administration of drugs. Then in vitro stability and mucoadhesion studies in simulated gastrointestinal fluids were carried out. The colloidal stability of the surfactant vesicles was determined by size and fluorescence-dequenching assay, while their mucoadhesive properties were evaluated by light-scattering and protein assay. The results of in vitro stability demonstrated that the pHs and enzymes (pepsin and/or pancreatin) of the gastrointestinal fluids had not influence on surfactant vesicle stability. However, in presence of bile salts the nanosize vesicles showed a release of fluorescent marker (about 11% at 2 h and 28% at 4 h), whereas they were stable in size as confirmed by the light scattering experiments. Finally, the in vitro mucoadhesive experiments showed that the capacity of nanovesicles to adsorb mucin was higher at neutral pH than at acidic pH. As a conclusion of these preliminary studies, the surfactant vesicles could be considered a versatile tool for the oral delivery of drugs with poor stability in gastrointestinal tract and low permeability. Nevertheless, further work is required in order to examine the interaction with and/or the transport route through the epithelial cells of the gastrointestinal wall. © 2012 Elsevier B.V. All rights reserved.
1. Introduction Most front-line therapeutic agents currently in use are untargeted, toxic compounds that act in a nonspecific way, often eliciting unwanted, dose-limiting, and often debilitating side effect. The development of an efficient, versatile and targeted delivery vehicle of bioactive molecules can bring a potential and complete solution to disease outcomes. The growing work in the field of drug delivery over the past 30 years has been aimed to the development of various safe nanosize delivery systems. Among different colloidal delivery carriers, vesicular systems such as liposomes have been investigated more than the other systems. Liposomes are mainly designed for parenteral use in order to protect labile drugs against environmental aggression and to achieve high concentration of pharmaceutical compounds in a specific area of body [1,2]. However, several drawbacks exist with the use of liposomes such as in vitro [3] and in vivo [4] instability, particularly in the gastrointestinal tract (GIT). After oral administration the lipid vesicles are easily degraded not only by bile salts, but also by pancreatic enzymes and by the low pH present in the gut [5,6]. This
∗ Corresponding author at: Università “G. D’Annunzio”, Chieti-Pescara, Dipartimento di Farmacia, Via dei Vestini, 66100 Chieti, Italy. Tel.: +39 0871 3554705/3554718; fax: +39 0871 3554911. E-mail address: [email protected] (L. Di Marzio). 0927-7765/$ – see front matter © 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.colsurfb.2012.10.036
gastrointestinal degradation leads both to a reduction of liposome structures reaching the bowel and to release the encapsulated bioactive compounds into the GIT. To improve liposomal membrane integrity in the GIT a possible approach is coating them with polymers such as chitosan, pectin and polyethylene glycol (PEG) [7]. Polyethylene glycols are hydrophilic, non-ionic and nontoxic polymers, and they have been extensively investigated for pharmaceutical and medical purpose. Indeed it has been shown that in parenteral administration, the PEG reduces the interaction of different nanoscale systems with opsonins and slows down their capture by the RES, leading to longer blood circulation halflives for the nanovectors in vivo [8]. Moreover, in recent years it has been reported that the PEG coating of nanosystems promotes and enhances the specific bioadhesive interactions with mucosal tissues due to the ability of PEG chains to diffuse across mucus network enhancing interpenetration [9,10]. Thus pegylated nanovectors can be considered as suitable systems to improve the bioavailability of drugs by prolonging the residence time of the bioactive molecules on the absorptive epithelium and controlling drug release properties. In general, it is established that the interactions between a polymer and the mucous layer can be due to physical or mechanical bonds, secondary chemical bonds and covalent chemical bonds. Secondary chemical interactions include ionic bonds, van der Waals interactions and hydrogen bonding. The functional groups that form hydrogen bonds are hydroxyls, carboxyls, sulfate and amino groups. Even these types of forces are
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
weaks, numerous interaction sites lead to strong mucoadhesion. Indeed one of the characteristics that seems to increase mucoadhesion phenomenon is the presence of hydrogen bonding [11]. A possible approach to obtain a specific bioadhesive drug delivery system could be the hydration of a mixture of cholesterol and non-ionic surfactants such as alkyl esters or ether surfactant, in which the hydrophilic region consists of repeated oxyethylene units. The vesicles formed from the non-ionic amphiphiles are known as niosomes or non-ionic surfactant vesicles (NSVs). They are analogous in terms of physical properties to phospholipid vesicles, but they possess several advantages over them. The main advantages lie in their chemical stability, lower toxicity, improved therapeutic performance and easier handling and storage. Another advantage of surfactant amphiphiles is their relatively low cost, which makes them attractive for industrial production both in pharmaceutical and cosmetic applications. Moreover niosome preparation for the routine and large-scale production can be performed without using pharmaceutically unacceptable solvents [12,13]. The NSV structural properties allow the entrapment of drugs with different lipophilicities: strongly lipophilic drugs are entrapped almost completely in the bilayer, strongly hydrophilic drugs are located exclusively in the aqueous compartment, and drugs with intermediate log P easily distribute between the lipid and aqueous phases, both in the bilayer and in the aqueous core [14]. Although pharmaceutical niosome formulations have yet to be commercially exploited, a number of studies demonstrated the potential of surfactant vesicles in drug delivery. Examination of the literature reveals that surfactant drug delivery systems play a significant role in formulation of bioactive molecules to improve safety and efficacy to administration of several classes of molecules like anti-neoplastic, anti-viral, anti-inflammatory, anti-microbial and proteins [15–21]. Based on these considerations, the aim of the present work was to investigate the stability of NSVs, composed by polisorbate 20, in the artificial gastrointestinal fluid. The polisorbate 20 (Tween® 20 or Tw20) is a non-ionic surfactant, the stability and relative non-toxicity of which, allows it to be used as excipient for oral administration. It is polyoxyethylene sorbitan monolaurate, in which the hydrophilic region consist of 20 oxyethylene units. The stability of Tw20 vesicles was evaluated in simulated digestive fluids in terms of release of the encapsulated fluorescent probes and size. As a second step, the interaction between non-ionic surfactant vesicles and mucin was evaluated to asses the mucoadhesive properties of polisorbate 20, particularly, in the surfactant vesicles. 2. Materials and methods 2.1. Materials Tween 20® (Tw20), cholesteryl hemisuccinate (CHEMS), Sephadex G-75, Hepes salt {N-(2-idroxyethyl), piperazine-N-(2ethanesulfonic acid)}, p-Xylene-bis (N-pyridinium bromide) (DPX), pancreatin (from porcine pancreas, 4XUSP), pepsin (from porcine gastric mucosa, 800–2500 UI/mg protein), and bile salts were Sigma–Aldrich products (Sigma–Aldrich SRL, Milan, Italy). Cholesterol (CHOL) and hydroxypyrene-1,3,6-trisulfonic acid (HPTS) were obtained from Acros Organics (Acros Organics BVBA, Geel, Belgium). Bovine submaxillary gland mucin was obtained from Merck Chemicals (United Kingdom). Bio-Rad protein assay was BioRad Laboratories (Bio-Rad Laboratories S.r.l., Milan, Italy). All other products and reagents were of analytical grade. 2.2. Preparation and characterization of non-ionic surfactant vesicles Non-ionic surfactant vesicles (NSVs) were prepared using different amounts of Tw20, CHOL and CHEMS (Table 1). Tw20
201
Table 1 Surfactant vesicle composition. Sample
Tw20 (mM)
CHOL (mM)
CHEMS (mM)
1 2
15 15
15 2
– 13
concentration was always remarkably above CMC (0.048 mM in water, at 20 ◦ C). The vesicles were obtained by the “film” method, as previously reported [22]. The dried films were hydrated by addition of HEPES buffer (10 mM, pH 7.4) alone or Hepes buffer (10 mM, pH 7.4) containing equimolar ratio HPTS/DPX (30 mM/30 mM). The surfactant dispersion was mechanically stirred for about 5 min and then sonicated for 10 min at 60 ◦ C (UP200H, Hielscher, Teltow, Germany) equipped with an exponential microprobe operating at 24 kHz and an amplitude of 60%. Vesicle dispersions were purified by size exclusion chromatography on Sephadex G75 glass columns. To determine size (nm) and -potential (mV) of the NSVs, vesicle dispersions were diluted 1:100 and 1:10 in HEPES buffer for size and -potential measurements, respectively. Vesicle size distribution and -potential were measured by dynamic light scattering (DLS) on a Malvern Zetasizer Nano ZS90 (Malvern Instruments Ltd., Worcestershire, United Kingdom) at 20 ◦ C. It should be pointed out that size distribution results were determined as % of intensity of the colloidal dispersion. The polydispersity index (p.i.) value was determined as a measurement of the width of the size distribution: a p.i. value lower than 0.3 indicates a homogenous and monodisperse population [23]. The surfactant structured in the vesicles was determined by the spectrophotometric method as previously reported [24]. 2.3. SGF and SIF Simulated gastric fluid (SGF) and simulated intestinal fluid (SIF) were made according to USP XXIV [25] with some modifications. The Table 2 summarize the compositions of the simulated gastric and intestinal media used in this study. 2.4. pH-stability studies In order to evaluate the stability of niosomes in presence of digestive fluids, 20 l of purified vesicles containing entrapped dies (HPTS/DPX) were incubated in media prepared according to the recommendations given in USP XXIV at 37 ◦ C. Samples were collected at times 0, 1, 2, 4 h. The SIF1 and SIF2 samples were centrifuged for 10 min at 13,600 × g and 4 ◦ C to eliminate pancreatin aggregates in medium. The stability of surfactant vesicles in simulated gastric and intestinal fluids was estimated by DLS measurements and evaluating the leakage of water-soluble fluorescent dyes, able to change their fluorescent behaviour upon release. In this study, the leakage of aqueous markers was assayed according to Smolarsky et al. [26] with the fluorescent probe HPTS and its quencher DPX entrapped in the surfactant vesicles. When the HPTS-DPX were released from vesicles, they dissociated into free HPTS and DPX, producing an increase in HPTS fluorescence revealed at 510 nm. The release of entrapped HPTS from the surfactant vesicles was determined by means of a fluorescence-dequenching assay using a LS55 spectrofluorometer (PerkinElmer, Massachusetts, USA) at excitation and emission wavelengths of 403 and 510 nm, respectively. The extent of HPTS release was expressed as a percentage and calculated according to the following equation: % HPTS release =
Fx − F0 × 100 Ft − F0
(1)
202
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
Table 2 Composition of the media to simulate gastric (SGF) and intestinal (SIF) fluids according to USP XXIV. Media composition
SGF
SGF1
SIF
SIF1
SIF2
Sodium chloride Sodium monobasic phosphate anhydrous Chloridic acid Sodium hydroxide Pepsin Pancreatin Bile salts pH
34.2 mM – 84 mM – – – – 1.2
34.2 mM – 84 mM – 2560 U/ml – – 1.2
– 50 mM – 25 mM – – – 6.8
– 50 mM – 25 mM – 250 U/ml – 6.8
– 50 mM – 25 mM – 250 U/ml 5 mM 6.8
where Fx is the fluorescence intensity of HPTS in the sample incubated in different media and/or different time, F0 is the fluorescence intensity of HPTS loaded niosomes in different media at room temperature at time = 0, Ft is the total fluorescence intensity of HPTS loaded surfactant vesicles in SIF after addition of isopropanol to induce vesicular structure rupture. The samples had different surfactant concentration of 1.2 M and 11.4 M for sample 1 and 2 respectively. The samples were maintained at room temperature during the measurements. The in vitro stability of NSVs in gastrointestinal fluid was also determined evaluating size variations and colloidal stability of vesicles by means of DLS.
where M0 is the glycoprotein total concentration incubated with nanosystems and ML is the free mucin amount after incubation and centrifugation. 2.7. Statistical analysis Results are expressed as the mean of three experiments ± S.E.M. (standard error of the mean). Statistical data analysis was performed using the t-test. To evaluate if obtained differences in experimental results were statistically significant, p ≤ 0.05 was used as the significance criterion. 3. Results
2.5. Turbidity measurement
3.1. Characterization of surfactant vesicles
Aggregation of nanosize vesicular samples was monitored as an increase in turbidity by right angle light scattering (excitation and emission = 600 nm) using a LS55 fluorimeter (PerkinElmer, USA). The samples were maintained at 25 ◦ C during the measurements using surfactant concentrations about 11 M for both samples.
The surfactant vesicles were prepared, as previously reported [22,27] by the film method associated with sonication in order to reduce the size down the sub-micron range. As reported in Table 3, the percentage of structured surfactant is 40 ± 4% and 38 ± 2% for sample 1 and 2, respectively. In addition vesicles are characterized
2.6. Mucoadhesive studies 2.6.1. Preparation of mucin solution The commercially available bovine submaxillary gland mucin was hydrated in 10 mM Hepes pH 7.4 at 4 ◦ C overnight. The mucin solution was centrifuged at 15,557 × g for 30 min at 4 ◦ C and then filtered through a 0.8 m cellulose nitrate filters prior to use. The final mucin concentration in solution was determined through the BioRad protein assay (Hercules, CA) using the bovine serum albumin standards.
2.6.2. In vitro interaction between vesicles and mucin The adsorption of mucin on vesicle surface was used as a method to assess mucoadhesive properties of the prepared surfactant vesicle. The interaction was studied by incubating mucin (0.5 mg/ml) and vesicles (surfactant concentration in the dispersion was fixed at 0.98 mg/ml) at pH 1.2 (SGF) and 6.8 (SIF). The incubation was carried out under magnetic stirring (700 rpm) at temperature of 37 ± 1 ◦ C for 4 h. Then the dispersion were centrifuged at 21,000 × g and 4 ◦ C for 30 min. After centrifugation the pellet were resuspended in H2 O and used to determine the vesicle size as described in Section 2.2, whereas the supernatant was collected and used for the determination of free mucin using Bio-Rad protein assay. In the protein assay the amount of mucin adsorbed on the surfactant vesicle surface was determined as a difference between its initial concentration and the concentration in the dispersion after incubation and centrifugation, according to Eq. (2): % bounded mucin =
M0 − ML × 100 M0
(2)
Fig. 1. In vitro stability of surfactant vesicles at the pH range of the gastrointestinal tract. The stability of surfactant vesicles (sample 1: Tw20:CHOL = 15 mM:15 mM; sample 2: Tw20:CHOL:CHEMS = 15 mM:2 mM:13 mM), incubated at pH 1.2 (SGF) and pH 6.8 (SIF) as described in Section 2, was evaluated by the release of encapsulated HPTS (A) and turbidity increase (B). The data, reported as percentage of HPTS release, was determined by Eq. (1). AU = arbitrary unit. The showed data represented the mean obtained from three independent experiments ± S.E.M. *p < 0.05 with respect to vesicles incubated in SIF.
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
203
Table 3 Characterization of surfactant vesicles. Percentage of non-ionic surfactant structured (%), vesicle dimensions (diameter), polydispersity index (p.i.) and -potential values. Reported data were means of three experiments ± S.E.M. Sample
%
Size (nm)
-Potential (mV)
p.i.
Sample 1 Sample 2
40 ± 4 38 ± 2
159.63 ± 8.19 156.43 ± 2.52
−32.89 ± 2.34 −29.83 ± 0.33
0.15 ± 0.03 0.29 ± 0.01
by dimensions in the sub-200 nm size irrespective of the composition, confirming our previous data [28]. Furthermore as shown in Table 3, both samples had a polydispersity index less than 0.3, value that indicated an homogeneity and monodispersity of the obtained vesicular population. The measurements carried out by DLS also showed that the entrapment of fluorescent probe HPTS and its quencher DPX did not cause a significant increase in surfactant vesicle size with respect to the corresponding empty ones, showing that the fluorescent probe and quencher were trapped within the surfactant samples (data not shown). The nanovesicles were negatively charged as reflected in the -potential values (≈−30 mV), which indicated the stability of the prepared samples [29]. The absolute negative value of -potential is likely attributed to the presence of the oxyethylene units on the vesicular surface [30].
Fig. 3. Effect of pancreatic lipase and bile salts on the stability of surfactant systems. The niosomes were incubated in simulated intestinal fluids in absence (SIF) and in presence of pancreatin (SIF1) or bile salts (SIF2) at 37 ± 1 ◦ C for different time. At specified times the fluorescence intensity and size of sample were recorded as described in Section 2.4. (A) HPTS release: the data of probe leakage from vesicles were expressed as percentage and were calculated using Eq. (1). The results were reported as mean of three independent experiments and S.E.M. values were always lower than 10% of the mean value. *p ≤ 0.05 with respect to vesicles incubated in SIF. (B) Size: the particle dimensions (nm) were evaluated by dynamic light scattering. The data of size were shown as mean ± S.E.M. obtained from three independent experiments.
3.2. Stability of surfactant vesicles in model of gastrointestinal fluids
Fig. 2. Effect of pepsin on the stability of non-ionic nanovesicles. The vesicles were incubated in gastric media in presence (SGF1) and in absence (SGF) of pepsin at 37 ± 1 ◦ C for different time. At specified times the fluorescence intensity and size of sample 1 were determined as described in Section 2.4. (A) HPTS release: the probe release from vesicles was expressed as percentage and was calculated using Eq. (1). The data were expressed as mean of three independent experiments and S.E.M. values were always lower than 10% of the mean value. (B) Size: the particle dimensions (nm) were evaluated by dynamic light scattering. The results were the mean of three independent experiments ± S.E.M.
Keeping in mind that the main purpose of our work was to design a drug delivery vesicular system recommend for oral administration, its stability in simulated digestive fluids should be evaluated. It is known that the gastric pH (1.2) is a key factor in the destabilization of vesicular systems, therefore their stability at pH 1.2 (SGF) and 6.8 (SIF) was evaluated measuring the release of entrapped HPTS and turbidity after 1 h of incubation. Fig. 1A reported the percentages of HPTS release from both surfactant nanovesicles at two different pH after 1 h incubation. As shown by the data reported in Fig. 1A, the percentage of HPTS release was quite similar at both different pH for samples 1, while it was significantly (p < 0.05) different at two pH for sample 2. In fact the percentage of released probe for sample 2 raised from 3 ± 2% to 15 ± 1% after the pH reduction from 6.8 to 1.2. Moreover no significant change in percentage of HPTS release from both sample was observed when HPTS release was compared at pH 6.8 and at the neutral pH 7.4 (data not shown). Furthermore the stability of both non-ionic surfactant vesicles was followed by turbidity measurements. As expected no changes in turbidity were observed in the case of sample 1 whereas the sample 2 showed a 1.5-fold increase in turbidity at pH 1.2 compared to 6.8 (Fig. 1B). In the present work
204
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
Fig. 4. Size profiles of surfactant vesicles incubated in intestinal fluid containing bile salts. The dimension of sample were determined by DLS analysis after incubation in intestinal fluid in absence and in presence of bile salts at different times. (A) Niosomes incubated in absence of bile salts (SIF), (B) vesicles incubated in presence of bile salts (SIF2) for 2 h (T2), and (C) vesicles incubated in presence of bile salts (SIF2) for 4 h (T4). The reported DLS data were triplicate analyses from a representative experiments of three experiment giving similar results. The S.E.M were ever lower than 10% of the mean value.
sample 2 was chosen as reference pH-sensitive system because its stability was dependent on the environmental pH [31]. From the collected data, sample 1 seemed to be stable at gastric pH in contrast with previously reported data on different liposomal formulations, showing a nearly 100% release of carboxyfluorescein after just 10 min at gastric pH [32]. According to the evidence that sample 1 was stable in the pH range of the gastrointestinal tract, the effect of enzymes (pepsin or pancreatin) and or bile salts was also evaluated. The stability
of sample 1 incubated in gastrointestinal fluids containing digestive compounds was compared with that of sample incubated in absence of gastrointestinal fluid constituents. In Fig. 2A the percentages of HPTS released by surfactant vesicles (sample 1) incubated in gastric fluid in presence (SGF1) and absence of pepsin (SGF) for 4 h were reported. The results clearly showed that nanovesicles were not subjected to enzymatic degradation. In addition the DLS data (Fig. 2B) were in close agreement with those obtained by HPTS release. The collected results indicated that neither the acidic
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
pH nor the gastric enzyme had effect on the vesicular integrity in simulated gastric fluid. Similar results were obtained when the surfactant vesicles were incubated in intestinal media in presence (SIF1) or absence (SIF) of the pancreatic lipase (Fig. 3A and B). Consequently, it seems reasonable to conclude that the oxyethylene units of hydrophilic region of Tw20 on the vesicle surface prevent breakage of the ester bond between the carboxyl group of the fatty acid and the hydroxyl group of the alcohol of Tw20 moiety by the pancreatic lipase. These data are in agreement with those previously reported on the stability of PLA or PA nanosystems coated with PEG in gastrointestinal media [33,34]. However, different behaviour was observed when stability studies were performed in intestinal fluid containing bile salts (SIF2). As showed in Fig. 3A in presence of bile salts the HPTS released from nanosize samples was unchanged during the first 60 min of incubation, while a relatively high amount of dye released from sample 1 was detected after 1 h (Fig. 3A). The percentage of dye released was approximately the 11% at 2 h and 28% at 4 h, suggesting that the stability was dependent on the incubation time (p ≤ 0.05). The bile salts concentration is above the critical micelle concentration and the time dependent stability could be related to an interaction leading to mixed vesicles [35] that might cause a leakage of HPTS. On the contrary, the mean size of samples remained unchanged (Fig. 3B), indeed the DLS data did not show any evidence of the micelle formation over the time. Fig. 4 illustrates the distribution profiles obtained from a representative experiment of vesicle stability in SIF2 at different times. Incubation of sample 1 in presence of bile salts, after incubation at 2 h and 4 h (Fig. 4B and C) produced vesicle size distribution and p.i. (<0.2) comparable to those of niosomes incubated in the absence of bile salts (Fig. 4A). Considering the size stability, indicating that the vesicles were not destroyed, the HPTS leakage from surfactant sample in SIF2 could be attributed to small membrane defects. This behaviour was seen previously for liposomes which stayed intact at low pH and leakage of small molecules was pronounced [36,37].
3.3. Mucoadhesive properties of surfactant vesicles It was reported that grafting of PEG into polymeric hydrogels enhanced the mucoadhesion [9,10]. Mucoadhesive behaviour of sample 1 was assessed in vitro by the suspension of sample 1 in a mucin aqueous solution at pH 1.2 (SGF) and 6.8 (SIF), and the amount of glycoprotein adsorbed onto vesicles and size was determined as reported in Section 2.6.2. The results of mucoadhesion are presented in Fig. 5. The DLS analyses showed a significant increase of size with respect to initial value of nanovesicles for the sample incubated with mucin at both pHs investigated. The increase in particle dimension was about 10% at pH 1.2 and 26% at pH 6.8. Indeed, sample size was 159 ± 2 nm in absence of glycoprotein and was significantly (p ≤ 0.05) increased to 175 ± 4 nm and 200 ± 3 nm after incubation in mucin solution at pH 1.2 and pH 6.8, respectively. To obtain further insight into the effect of pH on glycoprotein adhesion onto surfactant systems, the determination of mucin adsorbed to nanovesicles was also investigated using protein assay. The data obtained by Bio-Rad assay indicated that the percentage of mucin bound to the surface of nanosystem varied in a significant way (p ≤ 0.05) from 9 ± 1.5% at pH 1.2 to 20 ± 2.4% at pH 6.8, suggesting that a simulated medium at pH 6.8 respect to pH 1.2 had a more marked effect (p ≤ 0.05) on mucin interaction with nanosize vesicles. This is in agreement with previous data, which demonstrated that pegylated particles possessed higher affinity to adhere to intestinal than to the gut mucosa [38]. All together these results suggested that mucin amount bounded to the surface of surfactant vesicles could be due to the combination of two factor:
205
Fig. 5. Mucoadhesive properties of non-ionic surfactant vesicles. The interaction between vesicles and glycoprotein was studied by incubating sample 1 and mucin (0.5 mg/ml) at pH 1.2 (SGF) and 6.8 (SIF) for 4 h at 37 ± 1◦ C. The data, obtained as reported in material and methods, were expressed as bounded mucin and as the % of dimension increase respect to initial size of nanovesicles. In DLS determination the initial dimension of vesicles was set as 100%, whereas in the protein assay the adsorbed mucin was calculated by Eq. (2). The results were shown as black columns (initial size vesicle), white columns (sample incubated in SGF) and grey columns (sample incubated in SIF). The reported data were the mean ± S.E.M of triplicate analyses from a representative experiment of three experiments giving similar results. *p ≤ 0.05 with respect to initial size of nanovesicles and § p ≤ 0.05 with respect to nanovesicles incubated at pH 1.2.
1. high concentration of hydroxyl group at the end of PEG chains on the vesicular surface. In general, it is established that the interactions between a polymer and the mucous layer can be due to physical or mechanical bonds, secondary chemical bonds and covalent chemical bonds, in particular one of the characteristics that seems to increase mucoadhesion phenomenon is the presence of hydrogen bonding [11]. 2. Molecular weight of PEG (PM ≈ 900). Indeed, Yoncheva et al. reported that the ability of nanosystems to adsorb mucin at neutral pH was higher for PEG 1000 than for PEG 2000; the lower interaction between mucosal glycoprotein and PEG 2000nanosystems was due to the steric hindrance of PEG “brush” layer on the surface of the nanoparticles [38].
4. Conclusion Challenges for oral mucoadhesive systems include the harsh environment of the stomach, which, due to its low pH, results in an activation of a wide range of drugs as well as the degradation of protein drugs administrated for therapeutic purposes. On the other hand, the wash out effect in the GI tract due to the intestinal motility results in a low residence time of the drug at the site of absorption. For these reasons, designing a stable and mucoadhesive delivery system could be of great interest. In the present, from in vitro simulations, it could be evidenced that Tween® 20 vesicles show good stability in SIF and SGF, also in presence of digestive enzymes and bile salts, in fact the release of fluorescent probe is most likely related to small membrane damages but not to a disintegration of the vesicles indicated by size stability. Furthermore, pegylated vesicles possessed the ability to adhere to mucin in intestinal medium rather than in gastric one. Irrespective of the limited value of the in vitro results, it could be underlined that the reported data could be helpful in designing innovative drug delivery systems for oral administration. However, the mechanism of interaction and transport of non ionic surfactant vesicles through the gastrointestinal epithelium have to be clarified, together with toxicity studies.
206
L. Di Marzio et al. / Colloids and Surfaces B: Biointerfaces 104 (2013) 200–206
References [1] T.A. Elbayoumi, V.P. Torchilin, Methods Mol. Biol. 605 (2010) 1. [2] M.S. Webb, P. Rebstein, W. Lamson, M.B. Bally, Recent Pat. Drug Deliv. Formul. 1 (2007) 185. [3] M. Ausborn, P. Nuhn, H. Schreier, Eur. J. Pharm. Biopharm. 38 (1992) 133. [4] G. Poste, Biol. Cell 47 (1983) 19. [5] K. Iwanaga, S. Ono, K. Narioka, M. Kakemi, K. Morimoto, S. Yamashita, Y. Namba, N. Oku, J. Pharm. Sci. 88 (1999) 248. [6] K. Iwanaga, S. Ono, K. Narioka, K. Morimoto, M. Kakemi, S. Yamashita, M. Nango, N. Oku, Int. J. Pharm. 157 (1997) 73. [7] H. Takeuchi, Y. Matsui, H. Sugihara, H. Yamamoto, Y. Kawashima, Int. J. Pharm. 303 (2005) 160. [8] P. Aggarwal, J.B. Hall, C.B. McLeland, M.A. Dobrovolskaia, S.E. McNeil, Adv. Drug Deliv. Rev. 61 (2009) 428. [9] N.V. Efremova, Y. Huang, N.A. Peppas, D.E. Leckband, Langmuir 18 (2002) 836. [10] L. Serra, J. Doménech, N.A. Peppas, Eur. J. Pharm. Biopharm. 63 (2006) 11. [11] L. Serra, J. Doménech, N.A. Peppas, Eur. J. Pharm. Biopharm. 71 (2009) 519. [12] A. Manosroi, R. Chutoprapat, M. Abe, J. Manosroi, Int. J. Pharm. 352 (2008) 248. [13] T.T. Pham, C. Jaafar-Maalej, C. Charcosset, H. Fessi, Colloids Surf. B: Biointerfaces 94 (2012) 15. [14] I.F. Uchegbu, S.P. Vyas, Int. J. Pharm. 172 (1998) 33. [15] R. Muzzalupo, F.P. Nicoletta, S. Trombino, R. Cassano, F. Iemma, N. Picci, Colloids Surf. B: Biointerfaces 58 (2007) 197. [16] C. Marianecci, D. Paolino, C. Celia, M. Fresta, M. Carafa, F. Alhaique, J. Control. Release 147 (2010) 127. [17] S.K. Mehta, N. Jindal, G. Kaur, Colloids Surf. B: Biointerfaces 87 (2011) 173. [18] S. Akhter, S. Kushwaha, M.H. Warsi, M. Anwar, M.Z. Ahmad, I. Ahmad, S. Talegaonkar, Z.I. Khan, R.K. Khar, F.J. Ahmad, Drug Dev. Ind. Pharm. 38 (2012) 84. [19] M. Bragagni, N. Mennini, C. Ghelardini, P. Mura, J. Pharm. Pharm. Sci. 15 (2012) 184. [20] L. Di Marzio, C. Marianecci, F. Rinaldi, S. Esposito, M. Carafa, Lett. Drug Des. Discov. 9 (2012) 494.
[21] C. Marianecci, F. Rinaldi, M. Mastriota, S. Pieretti, E. Trapasso, D. Paolino, M. Carafa, J. Control. Release 164 (2012) 17. [22] M. Carafa, C. Marianecci, G. Lucania, E. Marchei, E. Santucci, J. Control. Release 95 (2004) 67. [23] V. Centis, P. Vermette, Colloids Surf. B: Biointerfaces 65 (2008) 239. [24] H. Kato, Y. Nagai, K. Yamamoto, Y. Sakabe, J. Assoc. Off. Anal. Chem. 72 (1989) 27. [25] US Pharmacopeia XXIV, 2006. [26] M. Smolarsky, D. Teitelbaum, M. Sela, C. Gitler, J. Immunol. Methods 15 (1977) 255. [27] E. Santucci, M. Carafa, T. Coviello, E. Murtas, F.M. Riccieri, F. Alhaique, A. Modesti, A. Modica, STP Pharm. Sci. 6 (1996) 29. [28] M. Carafa, L. Di Marzio, C. Marianecci, B. Cinque, G. Lucania, K. Kajiwara, M.G. Cifone, E. Santucci, Eur. J. Pharm. Sci. 28 (2006) 385. [29] L. Di Marzio, C. Marianecci, M. Petrone, F. Rinaldi, M. Carafa, Colloids Surf. B: Biointerfaces 82 (2011) 18. [30] R. Gref, A. Domb, P. Quellec, T. Blunk, R.H. Muller, J.M. Verbavatz, R. Langer, Adv. Drug Deliv. Rev 16 (1995) 215. [31] L. Di Marzio, C. Marianecci, B. Cinque, M. Nazzarri, A.M. Cimini, L. Cristiano, M.G. Cifone, F. Alhaique, M. Carafa, Biochim. Biophys. Acta 1778 (2008) 2749. [32] J. Parmentier, M.M. Becker, U. Heintz, G. Fricker, Int. J. Pharm. 405 (2011) 210. [33] M. Tobío, A. Sánchez, A. Vila, I.I. Soriano, C. Evora, J.L. Vila-Jato, M.J. Alonso, Colloids Surf. B: Biointerfaces 18 (2000) 315. [34] A. Sahu, U. Bora, N. Kasoju, P. Goswami, Acta Biomater. 4 (2008) 1752. [35] P. Walde, J. Sunamoto, C.J. O’Connor, Biochim. Biophys. Acta 905 (1987) 30. [36] G.B. Patel, B.J. Agnew, L. Deschatelets, L.P. Fleming, G.D. Sprott, Int. J. Pharm. 194 (2000) 39. [37] M. Taira, N. Chiaramoni, K. Pecuch, S. Alonso-Romanowski, Drug Deliv. 11 (2004) 123. [38] K. Yoncheva, E. Lizarraga, J.M. Irache, Eur. J. Pharm. Sci. 24 (2005) 411.