International Journal of Biological Macromolecules 75 (2015) 409–417

Contents lists available at ScienceDirect

International Journal of Biological Macromolecules journal homepage: www.elsevier.com/locate/ijbiomac

Calcium alginate-carboxymethyl cellulose beads for colon-targeted drug delivery Tarun Agarwal a,1 , S.N. Gautham Hari Narayana a,2 , Kunal Pal a,3 , Krishna Pramanik a,4 , Supratim Giri b,5 , Indranil Banerjee a,∗ a b

Department of Biotechnology and Medical Engineering, National Institute of Technology, Rourkela Odisha 769008, India Department of Chemistry, National Institute of Technology, Rourkela Odisha, 769008, India

a r t i c l e

i n f o

Article history: Received 18 October 2014 Received in revised form 12 December 2014 Accepted 15 December 2014 Available online 10 February 2015 Keywords: Calcium alginate Carboxymethyl cellulose Bead Colon-specific delivery pH sensitivity Mucoadhesivity

a b s t r a c t The present study delineates preparation, characterization and application of calcium alginate (CA)carboxymethyl cellulose (CMC) beads for colon-specific oral drug delivery. Here, we exploited pH responsive swelling, mucoadhesivity and colonic microflora-catered biodegradability of the formulations for colon-specific drug delivery. The CA-CMC beads were prepared by ionic gelation method and its physicochemical characterization was done by SEM, XRD, EDAX, DSC and texture analyzer. The swelling and mucoadhesivity of the beads was found higher at the simulated colonic environment. Variation was more prominent in compositions with lower CMC concentrations. CA-CMC formulations degraded slowly in simulated colonic fluid, however the degradation rate increased drastically in the presence of colonic microflora. In vitro release study of anticancer drug 5-fluorouracil (5-FU) showed a release (>90%) in the presence of colonic enzymes. A critical analysis of drug release profile along with FRAP (fluorescence recovery after photobleaching) study revealed that the presence of CMC in the formulation retarded the release rate of 5-FU. 5-FU-loaded formulations were tested against colon adenocarcinoma cells (HT29). Cytotoxicity data, nuclear condensation-fragmentation and apoptosis analysis (by flow cytometry) together confirmed the therapeutic potential of the CA-CMC formulations. In conclusion, CA-CMC beads can be used for colon-specific drug delivery. © 2015 Elsevier B.V. All rights reserved.

1. Introduction The colon-targeted oral drug delivery is desirable in order to treat a variety of colon diseases, such as ulcerative colitis, Crohn’s disease, amebiosis, colonic cancer, etc. [1–6]. In recent years, there have been a number of developments for the improvement of target specificity of colon-targeted delivery systems [7]. The primary approaches pertaining to the colon-specific delivery include: (i) covalent linkage of a drug with polymers as a prodrug, (ii) coating of the delivery system with the pH-sensitive

∗ Corresponding author. Tel.: +91 9438507035. E-mail addresses: [email protected] (T. Agarwal), [email protected] (S.N.G.H. Narayana), [email protected] (K. Pal), [email protected] (K. Pramanik), [email protected] (S. Giri), [email protected] (I. Banerjee). 1 Tel.: +91 9933968910. 2 Tel.: +91 7750826049. 3 Tel.: +91 6612462289. 4 Tel.: +91 6612462283. 5 Tel.: +91 6612462666. http://dx.doi.org/10.1016/j.ijbiomac.2014.12.052 0141-8130/© 2015 Elsevier B.V. All rights reserved.

polymers (e.g. Eudragit polymers), bioadhesive polymers (e.g. polycarbophil-based polymers) or biodegradable polymers and (iii) microbially triggered release of the drug. In addition, some of the novel drug-delivery approaches have also been introduced such as: (i) pressure-controlled drug delivery, (ii) CODESTM (combined approach of pH-dependent and microbially triggered drug delivery), (iii) osmotic pressure-controlled drug delivery through a semipermeable membrane and (iv) multiparticulate systems like microspheres and nanoparticles [1]. However, in a recent review, Talaei et al. [8] highlighted that although the novel drug-delivery systems have shown good potential, yet further improvements are needed before their full translation into clinical use. Calcium alginate (CA) and carboxymethyl cellulose (CMC) are two biopolymers that can be used for developing oral drug-delivery systems. Alginate (salts of alginic acid) is a linear polysaccharide composed of alternating blocks of ␤ (1→4) linked d-mannuronic acid and ␣ (1→4) linked l-guluronic acid residues [9–14], whereas CMC consists of linear chains containing ␤ (1→4)-linked glucopyranose residues [15]. These biopolymers have been reported to show a pH-dependent swelling behavior [13,16–18]. Both the polymers are anionic in nature due to the presence of negatively charged

410

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

Table 1 Composition of the CA-CMC beads. S. No.

1 2 3 a

Samplesa

T1G T3G T5G

Polymer concentrations (wt%)

Glutaraldehyde concentration (%)

Sodium Alginate

Carboxymethyl Cellulose

1.8 1.8 1.8

— 0.5 1

0.25 0.25 0.25

G stands for glutaraldehyde crosslinked beads.

carboxyl groups at pH > 5. These negative charges allow the polymer to shrink in the acidic pH and to swell when they are exposed to neutral or basic pH. These properties make these polymers suitable for applications in the design of oral drug-delivery systems. Apart from the pH sensitivity, SA and CMC have also been known to possess an excellent mucoadhesive property [19]. Keeping the aforesaid perspective in mind, here we have explored the potential of calcium alginate (CA)-carboxymethyl cellulose (CMC) bead as a colon-specific drug-delivery system. We hypothesize that an appropriate composition of CA-CMC will ensure: (1) least amount of drug release at non-specific sites (stomach and small intestine) during its transit through the GI tract, (2) higher adhesion to the colonic mucosa in comparison to other parts of the GI tract and (3) controlled degradation of the formulations by the colonic microflora. These factors are expected to promote sustained release of the drug in the colon. The rationale behind such ventures is: (i) to ensure the appropriate therapeutic dose at colon for effective treatment, (ii) to avoid dose and activity loss of the therapeutics during the GI transit and (iii) to minimize the adverse side effects of the therapeutics caused from absorption at non-specific tissue locations. 5-Fluorouracil (5-FU) was taken as the reference drug in this study. 2. Materials and Methods 2.1. Materials Sodium alginate (SA) (molecular weight: 7.72 × 104 g/mol, degree of polymerization: 476, M/G ratio 1.08) was bought from SDFCL, Mumbai, India. Calcium chloride (CaCl2 , fused) was purchased from MERCK, Mumbai, India. Glutaraldehyde (25% aqueous solution) was procured from LOBA Chemie, Mumbai, India. Carboxymethyl cellulose sodium (CMC) salt (molecular weight: 6.62 × 105 g/mol, degree of polymerization: 3062, degree of substitution: 0.68), MaCoy’s 5A media, Dulbecco’s phosphatebuffered saline (DPBS), trypsin-EDTA solution, fetal bovine serum, antibiotic-antimycotic solution, MTT assay kit and nutrient broth were purchased from Himedia, Mumbai, India. HT29 adenocarcinoma cell line was procured from NCCS, Pune. 5-FU and FITC-Dextran (molecular weight: 10 kDa) were obtained from Sigma-Aldrich, Mumbai. 2.2. Methods 2.2.1. Preparation of CA-CMC beads Preparation of the CA-CMC beads was done by ionic gelation method as described by Girhepunje et al. [20]. Both SA and CMC were dissolved in deionized water at a specific concentration (Table 1). Thereafter, the prepared polymeric solution was extruded as droplets using a syringe (28G) and poured into 2% calcium chloride (w/v) solution under constant stirring at 80 rpm and 37 ◦ C and cured for 10 min. Then, 1.1 ml of glutaraldehyde reagent [glutaraldehyde (25%, 0.5 ml) + ethanol (0.5 ml) + HCl (0.1 N, 100 ␮l)] was added to 50 ml of calcium chloride solution and bead curing was done for another 10 min. The beads were washed with deionized water, neutralized with glycine and dried overnight at 40 ◦ C. To

determine the average size of the swollen and dried beads, images were taken using the camera (Canon A2400 IS) and the images were analyzed by NIH ImageJ software.

2.2.2. Physicochemical characterization of the beads Morphological characterization of the dried beads was carried out using scanning electron microscopy (JOEL India JSM-6480Lv) at 15 kV after platinum sputter coating. The calcium content of the beads was analyzed by energy-dispersive X-ray spectroscopy. Variation in percentage crystallinity of CA-CMC beads were recorded using X-ray diffraction (Philips XRD-PW1700 diffractometer). Scanning was done in the range of 5–60◦ 2 with a step size of 0.02◦ /s using monochromatic Cu K␣ radiation of wavelength ˚ Differential scanning calorimetry analysis was carried ( = 1.514 A). out by heating 20 mg of CA-CMC beads from 35 to 250 ◦ C, at a rate of 5 ◦ C/min using DSC-200-F3 MAIA instrument (Netzsch, Germany). Bulk compressive strength of the CA-CMC beads was analyzed by TA.XT2i Texture analyzer (Stable Micro Systems Ltd, Surrey, UK). The analysis was done using 30 mm probe, 1 mm/s test speed and auto (force) mode (5 g, 5 mm).

2.2.3. Swelling analysis Swelling of the beads in simulated GI fluids was studied following the protocol described by Pasparakis et al. [21]. For this, accurately weighed, dried CA-CMC beads were immersed in phosphate-buffered saline (PBS; pH 7.4 and 6.8) and in 0.1 N HCl (pH 1.2) at 37 ◦ C. At defined time intervals, the beads were withdrawn from the solution and increase in the weight of the beads was measured as a function of time. Swelling ratio (SR) was expressed as SR =

W2 − W1 W1

where W1 and W2 represent the dry and wet weight of the beads, respectively.

2.2.4. Mucoadhesivity Mucoadhesivity testing was carried out following in vitro washoff protocol as reported by Lehr et al. [22]. In brief, fresh tissue portions from the goat stomach and colon were obtained from a commercial slaughter house and cleaned with cold normal saline. The tissues (1.5 cm × 1.5 cm) were fixed on a glass slide using adhesive glue keeping the mucosal surface upward. Fifty milligrams of the beads were placed on the mucosa and a 5 g load was applied onto it for 15 min to ensure uniform adhesion of the bead on the mucosa. Thereafter, bead-loaded stomach and colon mucosa were placed in 0.1 N HCl (pH adjusted to 1.2, specific to stomach) and PBS (pH 6.8, specific to colon), respectively, onto the groves of USP24 tablet disintegration apparatus. The disintegration apparatus was then operated in a way that ensured up and down movement of tissue specimen in 1 l of buffer at 37 ◦ C. The experiment was run for 24 h and the time corresponding to the complete wash off of the beads was noted.

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

2.2.5. Colonic microflora-specific biodegradation Colonic microflora was obtained from human stool culture of a healthy volunteer. The culture was inoculated in nutrient broth at 37 ◦ C and 50 mg of the dried CA-CMC beads were added to the culture. A similar set of experiment was also performed in phosphate-buffered saline (pH 6.8) and nutrient broth (pH 6.8, without bacteria). The study was monitored for 1 month and time required for the complete degradation was recorded. Also, the interaction of the colonic microflora with the bead material was examined through field emission scanning electron microscope (FESEM) (Nova NanoSEM 450) after gold sputter coating (Quorum Technologies, Q150R ES) at 3 kV. 2.2.6. In vitro drug release study Loading of 5-FU in the CA-CMC bead was done by swelling [23]. In brief, 10 mg of the dried beads were incubated in to 50 ␮l of aqueous drug solution (10 mg/ml) at pH 7.4 for 9 h. Volume of the drug solution and duration of loading was fixed on the basis of swelling data. After 9 h, beads were taken out from the drug solution washed gently with PBS and then subjected for drug release study. To analyze the extent of loading, each set was dried separately, then crushed using mortar pestle and total drug was extracted using 10 ml of PBS (pH 7.4). Extraction was done in a step-wise manner using 2 ml of extraction buffer at a time. Concentration of drug in the extract was measured using UV–vis spectrophotometry at 266 nm. In vitro drug release from the beads was carried out in simulated gastrointestinal environment as described by Ahmad et al. [24]. In brief, 5-FU was loaded into the polymeric solution at a concentration of 5% (w/w) of the total polymer concentration. Thereafter, accurately weighed 100 mg of the drug-loaded dried beads were initially incubated in 50 ml 0.1 N HCl (simulated gastric fluid, pH adjusted to 1.2) for 2 h, then incubated in 50 ml of PBS (simulated small intestinal fluid, pH 7.4) for 3 h and finally transferred to 50 ml of PBS (simulated colonic fluid, pH 6.8) and kept for another 115 h [14,20]. In addition, a similar analysis was also performed wherein the enzyme cocktail released by the colonic microflora was added to PBS (pH 6.8) in 1:10 ratio. The enzyme cocktail was prepared from the human stool culture of a healthy volunteer. For this, the stool culture was centrifuged at 5000 rpm for 10 min. The supernatant thus obtained was filtered using 0.22 ␮m filter and then used for the analysis. The beads in the release medium were kept under shaking conditions in a shaker incubator (WADEGATI Labequip Pvt. Ltd.) at 60 rpm and 37 ◦ C. At definite time intervals, 3 ml of the sample was taken out of the flasks and was replaced by 3 ml of fresh PBS and analyzed spectrophotometrically at 266 nm. Furthermore, mobility of large molecules entrapped in the CA-CMC beads at simulated intestinal fluid (PBS, pH 6.8) with or without enzyme cocktail was analyzed by fluorescence recovery after photobleaching (FRAP) using confocal scanning laser microscope (Olympus IX 81 confocal microscope using Fluoview1000). For this purpose, FITC Dextran (mol. wt. 10 kDa) entrapped beads were subjected to bleaching in a specific region of interest (ROI) using 95% intensity of a multiargon laser (40mW, 488 nm). Fluorescence recovery in the region of interest was recorded at 0.2% intensity using the same laser source. 2.2.7. In vitro evaluation of therapeutic potential of drug loaded beads Efficacy of the drug-loaded bead was tested against HT-29 colon adenocarcinoma cell line by MTT assay, flow cytometry and immunocytochemistry. In brief, the cell line was maintained in MaCoy’s 5A supplemented with 10% FBS in a humidified (95%), CO2 (5%) incubator at 37 ◦ C. Cells were harvested using 0.25% trypsinEDTA solution. Thereafter, the cells were seeded in a 12-well plate at a viable cell concentration of 1 × 105 cells/ml. Dried beads (blank and drug loaded) of each formulation were UV sterilized and 10 beads were placed in each well. The culture plate was incubated

411

for the next 48 h. Cytotoxicity of the formulation was first analyzed using MTT Assay. For flow cytometry-based apoptosis analysis, 1 × 105 cells were incubated for 48 h in a 6-well plate in the presence of drug-loaded beads (10 mg). After 48 h of incubation, percentage of apoptosis was analyzed by flow cytometer (BD Accuri) using PE Annexin V apoptosis detection kit (BD pharmingen). Furthermore, HT-29 cells exposed to the drug (leachant from the 5-FUloaded beads) was checked for nuclear condensation and fragmentation using DAPI staining (1:300 dilution) by confocal microscopy (Olympus IX 81 confocal microscope using Fluoview1000) [25]. 2.2.8. Statistical analysis All the data were reported as mean ± S.D. (standard deviation). For evaluating statistical significance of the data, one-way ANOVA was performed. 3. Results and Discussion 3.1. Bead preparation Calcium alginate-carboxymethyl cellulose (CA-CMC) beads were prepared by the ionic gelation method. In this process, anionic carboxylic groups present in alginate and carboxymethyl cellulose interacted with bivalent calcium ion to form the gel. It was observed that an increase in the CMC concentration resulted in an overall decrease in the percentage yield. The percentage yield of T1G beads was found 95.93 ± 2.3% and that of T3G and T5G was 92.04 ± 1.9 and 88.56 ± 2.7%, respectively. In an earlier study, Arica et al. [26] reported similar trend while working with alginate beads. In practice, with an increase in the biopolymer concentration (here, CMC), the viscosity of the solution increases which contributes toward higher retention of liquid volume at the tip of the nozzles and thus increased bead size [14]. The average size of dried T1G, T3G and T5G beads was found to be 738.51 ± 19.41, 875.81 ± 25.37 and 918.92 ± 37.45 ␮m, respectively. Preliminary stability study showed that beads crosslinked by calcium chloride were degraded completely over a period of 12 h in PBS (pH 7.4) [27]. This happened because of the release of Ca2+ from the beads to the solution which leads to the disruption of the polymeric network. The stability of the beads was increased by crosslinking with 0.25% glutaraldehyde. In this case, glutaraldehyde crosslinked the polymers via acetal bond formation [28,29]. The swollen bead formulations appeared white, translucent and spherical in shape. The average diameter of swollen T1G, T3G and T5G were found 1938.16 ± 37.15, 2156.45 ± 57.56 and 2270.11 ± 73.78 ␮m, respectively (Fig. 1G). 3.2. Physicochemical characterization Analysis of surface morphology of the dried beads by scanning electron microscopy revealed that pure calcium alginate beads were spherical with a smooth surface topography. Beads appeared solid (devoid of any core) and without any micropores on their surface. A critical examination showed that an increase in the CMC concentrations resulted in an increase in the bead size and formation of wrinkles on the bead surface (Fig. 1A–H). Similar morphological changes were reported earlier by Kim et al. [30]. It is important to mention that there were cracks on the surface of the beads. However, such cracks were not seen under light microscope and might have developed during the sample processing for scanning electron microscopy. The elemental analysis of the beads demonstrated that with an increase in the CMC concentration, the calcium content increased significantly from 12.89 ± 0.37% (%w) in T1G to 31.74 ± 0.93% (%w) in T5G (p < 0.05) (Fig. 1I). Calcium is involved in the crosslinking of both alginate and CMC [31,32].

412

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

Fig. 1. Scanning electron micrograph of dried CA-CMC bead. (A, B) T1G; (C, D) T3G; (E, F) T5G. Average diameter of the swollen (G) and dried (H) beads. (I) EDAX-based analysis of calcium content (wt%) in the CA-CMC beads. Sodium alginate (SA) and carboxymethyl cellulose (CMC) were taken as reference.

The XRD analysis of the pure sodium alginate demonstrated characteristic peaks at 13.6◦ 2 and 22.1◦ 2, while the diffractogram of CMC consisted of a broad characteristic peak at 20◦ 2. Aforementioned three characteristic peaks corresponding to SA and CMC were all present in the XRD profile of CA-CMC beads. However, the peaks were broadened, suggesting an increase in the amorphous nature of the formulations (Fig. 2). The DSC thermogram of the CA-CMC beads showed a broad endothermic peak near 100 ◦ C which may be due to the evaporation of water molecules. The peak was found to be more intense in T1G followed by T3G and T5G respectively, which suggest the presence of higher proportion of water in T1G. It is also important to mention that a secondary endothermic peak was also observed which may be associated with the bound water molecules. The secondary peak in T1G, T3G and T5G were found to be present at ∼186, ∼186 and ∼193 ◦ C, respectively. The presence of secondary peak at elevated temperature indicates thermal degradation of the polymer matrix (Fig. 3A). The bulk compressive modulus of T3G and T5G was found 0.4605 ± 0.004 and 0.3726 ± 0.517 MPa, respectively, which are significantly lower in comparison to that of T1G (1.331 ± 0.181 MPa). This could possibly due to inherent property (compressive strength) of the polymeric formulation or the variation in the packing of the CACMC beads. The packing of the beads is dependent on their size. As already mentioned, T1G have smaller and regular bead size, owing to its proper packing and higher bulk compressive strength (Fig. 3B).

3.3. Swelling analysis Calcium alginate and carboxymethyl cellulose, being polyelectrolytes, exhibit pH-dependent swelling. Such pH-responsive swelling property is attributed to the presence of negatively charged carboxyl groups present in the polymer backbone. In acidic pH, carboxylic acid groups remain undissociated and therefore no net charge is developed in the polymeric network. Once exposed to neutral or alkaline medium, carboxylic acid group is converted into negatively charged carboxylate ions, resulting in an electrostatic repulsion amongst the different polymer chains. This in turn compels the polymer network to swell. Such pH-dependent swelling often modulates release of a drug molecule from a carrier system in oral drug delivery. The swelling study of the beads was carried out in three different pH conditions: 1.2, 7.4 and 6.8 corresponding to the pH of the stomach, small intestine and colon, respectively. The highest swelling for all formulations took place at pH 7.4, while lowest swelling happened at pH 1.2. Difference in swelling at pH 6.8 and 7.4 was insignificant for all the formulations. Interestingly, at higher pH, i.e. at pH 6.8 and 7.4, the swelling was reduced with an increase in the CMC concentration while a reverse trend was observed at pH 1.2. When compared to T1G (without CMC), there was a 1.81- and 3.26-fold decrease in the swelling index in T3G (with 0.5% CMC) and T5G (with 1.0% CMC), respectively, at pH 7.4 (Fig. 3C). On the other hand, the swelling ratio of the T3G and T5G beads in pH 1.2 increased by 1.3- and 1.6-fold in comparison to T1G, respectively.

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

413

Table 2 Differential mucoadhesion of CA-CMC beads. S. No.

Sample

1 2 3

T1G T3G T5G

Mucoadhesivity (h)a pH 1.2

pH 6.8

1.9 2.5 2.6

>24 >24 >24

a Mucoadhesivity is expressed in terms of retention time of beads on mucosal surface.

tend to get shielded; however, the charges present in the core of beads may contribute to a slight increase in the swelling index due to inadequate shielding [30]. A critical analysis of the swelling profile of all the formulation at pH 7.4 and 6.8 revealed that the extent of swelling for all the three formulations at initial phase (up to 1 h) was significantly different irrespective of the formulation studied. 3.4. Mucoadhesivity of the beads at different parts of the GI tract

Fig. 2. XRD profile of dried CA-CMC beads. Sodium alginate (SA) and carboxymethyl cellulose (CMC) were taken as reference.

Swelling of T1G, T3G and T5G at pH 6.8 followed a similar pattern as in the case of pH 7.4 with swelling index of 30.05 ± 1.277, 17.278 ± 0.955 and 8.914 ± 0.97, respectively. With an increase in the CMC concentration, relative charge density of the beads tends to increase. At pH 1.2, the negative charges present on the surface

Mucoadhesivity is a property of the polymeric formulation which allows it to adhere onto the mucus membrane [33]. Mucoadhesivity needs special consideration while designing an oral drug-delivery system as it ensures prolonged retention of the formulation at a specific location in the GI tract and thus helps in achieving a sustained drug release for a longer period [33,34]. It is important to mention that the entire gastrointestinal tract in humans is lined by a mucus membrane but the characteristics of this mucosal lining tend to vary from one region of the GI tract to the other [35]. The major component of this mucus membrane is a branched glycoprotein named mucin which shows a pH-dependent variation in its configuration [36]. Such pH dependence alters its affinity for mucoadhesive materials. This implies that the formulation having the maximum mucoadhesivity for colon mucosa is desirable for an effective colon-targeted delivery. Keeping this fact in mind, the mucoadhesiveness of the beads was analyzed

Fig. 3. (A) DSC thermogram of the CA-CMC beads, (B) bulk compressive strength of CA-CMC beads and (C) swelling of CA-CMC beads in different GIT simulated fluids.

414

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

Fig. 4. (A) Degradation of CA-CMC beads at colonic pH. Degradation was studied in the presence and absence of colonic microflora and expressed in terms of time (hours) required for complete degradation of 100 mg beads of different formulations. Data were expressed as mean ± SD (*, **p < 0.005). (B) Differential adhesions of colonic microflora on CA-CMC beads at colonic pH. (A) T1G, (B) T3G, (C) T5G.

for colon and stomach mucosa at pH 6.8 and 1.2, respectively (Table 2). Data showed that at pH 6.8, a major fraction of the beads remained adhered to the colon mucosa surface after 24 h, while significant decrease in mucoadhesiveness was observed in the case of stomach mucosa at pH 1.2. At acidic pH, all three formulations showed weak mucoadhesivity and beads were washed off from the mucosal surface within 2–3 h. In the acidic pH, the intrinsic negative charges of the beads get shielded off, which may play a critical role for the adhesiveness of the beads with the mucus membrane. Also, at neutral pH, the charges of the beads tend to get expose, allowing it to interact with the mucus membrane through strong electrostatic interactions and thus showing a greater mucoadhesivity.

3.5. Colonic microflora-specific degradation of the beads The colonic region of the gastrointestinal tract is a known habitat of over 400 distinct bacterial species, including Bacteroides, Bifidobacterium, Eubacterium, Peptococcus, Lactobacillus, Clostridium and Escherichia coli [1,6,9,24,35,37]. These bacterial species produce a number of reductive and hydrolytic enzymes such as ␤-glucuronidase, ␤-xylosidase, ␤-galactosidase, ␣-arabinosidase, nitroreductase, azoreductase, deaminase and urea hydroxylase. This enzymatic cocktail produced by the colonic microflora help in the drug release by degrading the biopolymeric matrix of the delivery system [6,9,24]. It is usually observed that before reaching the colon region of the GI tract, a significant amount of the entrapped

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

drug is retained inside the beads. So, it is very essential that the drug-delivery systems must be degraded completely by the microbial population residing in the colon and release the remaining drug entrapped within them. In this regard, we tried to analyze in vitro, the potential of the intrinsic colon bacteria to degrade beads of the CA-CMC formulation. In vitro degradation studies carried out in the presence of native colon microflora obtained from the human stool culture revealed that the prepared beads were completely degradable in such environment (Fig. 4A). The time required for complete degradation of the beads varied largely from one formulation to another. It was observed that the pure alginate beads took approximately 42 h to get degraded, while T5G alginate beads containing 1% CMC get degraded in around 98 h. An increase in the CMC concentration in the beads tends to slow down the rate of degradation. A comparative kinetic study of bead degradation revealed that the beads of all the formulations failed to get degraded in the absence of bacteria in phosphate buffer saline (pH 7.4) and nutrient broth (pH 7.4) within a considerable time frame and took a time span of four weeks. In PBS and media, T1G took approximately 12–13 days for degradation while T5G took approximately 600–640 h (4 weeks). The analysis demonstrated that the addition of CMC had a profound effect on the degradation of the beads irrespective of the environmental conditions. In PBS and nutrient broth, addition of CMC in the formulation significantly reduced the degradation rate of the formulated beads (p < 6.5E−6). Furthermore, with the addition of the microbial culture to the degradation media to simulate colonic condition, the degradation rate of all the formulations increased significantly (p < 5.00E−6), however degradation profile followed the same trend as in the case of PBS and nutrient broth. Here, it is important to mention that the transit time for the colon is reported to be 8–72 h [38]. Our previous data already suggest that beads of all formulations are highly mucoadhesive to colon mucosa at pH 6.8. The high mucoadhesivity of the beads may increase their retention time, which may provide a chance for the complete degradation of these beads and thus resulting in the complete drug release. A microscopic analysis pertaining to the interaction between beads and microbes (Fig. 4B–D) revealed a preferential bacterial adhesion on a pure alginate composition with formation of bacterial biofilm on T1G bead surface. However, the bacterial adhesion was found to decrease with the addition of CMC. This explains about a higher degradation rate of T1G in comparison to T3G and T5G in the presence of the bacterial microflora. 3.6. In vitro drug release study The colon-specific drug-delivery systems must release a negligible or low amount of the loaded drug in the stomach or small intestine. The residence time of any solid dose in human stomach and small intestine during GI transit is 2–3 and 3–4 h, respectively [38]. Since it is difficult to control the transit time, efforts have been made to tailor the release of the drugs from the carrier system during transit. Thus, it becomes essential to analyze the percentage of the drug released from the formulations before they reach the colon. Drug loading is an important parameter in this regard because it determines the theoretical limit of maximum drug release. Here drug loading was done by swelling method. Analysis of drug loading showed that percentage loading of 5-FU in dried alginate bead (T1G) was 66% of the initial drug take. The same for T3G and T5G were 72 and 82%, respectively. The loading with respect to the dry weight of the bead was 30 mg/g for T1G, 33 mg/g for T3G and 41 mg/g for T5G. It is important to mention that extent of drug loading in the beads did not follow the swelling profile rather a reverse trend was observed. There could be a drug–CMC interaction that caused higher retention of the drug molecules by the polymer. Analysis of the drug-loaded beads by XRD revealed that there was

415

Fig. 5. (A) Analysis of release of 5-fluorouracil from CA-CMC matrix in the presence and absence of colonic enzymes in simulated fluids corresponds to different compartment of the GI tract. The residence time of the bead at each compartment was fixed as per the actual in vivo value. Data are expressed as mean ± SD. All data are statistically significant with respect to groups (p < 0.005). (B) Analysis of solute mobility inside CA-CMC matrix through fluorescence recovery after photobleaching. Percentage recovery corresponds to the free mobility of solute molecules (FITC-Dextran, 10 kDa). E stands for the experimental conditions in which colonic enzymes are present and active.

a variation in crystallite structure of CMC in the presence of 5-FU (data not shown) which indicates about such possibility. The drug release analysis showed that the T3G and T5G beads released 6 and 4.7% of the drug, respectively, in comparison to 8.6% drug release in T1G after 2 h of analysis at pH 1.2. A sudden release of the drug was observed in case of beads of all formulations when they were transferred from simulated gastric fluid (pH 1.2) to simulated intestinal fluid (pH 7.4), although this burst release significantly reduced with an increase in the CMC concentration (p < 0.05) [39,40]. A cumulative release of 34 and 27% was observed in T3G and T5G, respectively, after 5 h of analysis at pH 7.4. T1G beads showed significant difference with respect to T3G (p < 0.005) and T5G (p < 0.005) under same conditions with a cumulative release of 41%. Further, when the beads were transferred to the simulated colonic fluid (pH 6.8), a cumulative release of 75.5, 60.6 and 51.5% was observed in the case of T1G, T3G and T5G, respectively, at the end of analysis. This data showed that the typical drug release in the colon compartment is 34.2% (T1G), 26.5% (T3G) and 24.36% (T5G). It is important to mention that a higher drug release occurred in the presence of enzymes with respect to control (without enzymes). A critical analysis of the drug release in simulated colonic fluid (pH 6.8) demonstrated that the drug released from the beads of all three formulations in presence/absence of enzymes was not significant during the initial 24 h of the

416

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

Fig. 6. Study of the efficacy of 5-FU-loaded CA-CMC beads on HT-29 cells. (A) Cell viability was measured by MTT after 48 h of drug exposure. (B) Flow cytometry analysis of apoptosis in HT-29 cells using PE–annexin V. The cells were exposed to the drug loaded formulation for 48 h. (C) Immunocytochemistry-based study of nuclear condensation and fragmentation using DAPI (Blue): (I) Control, (II) T1G, (III) T3G, (IV) T5G. Arrow designates the nuclear condensation and fragmentation.

analysis (p > 0.05). Thereafter, considerable differences were observed in comparison to the control set (without enzymes). A cumulative drug release of 94.7, 75.9 and 60.8% occurred from T1G, T3G and T5G beads, respectively, at the end of the analysis (Fig. 5A). In the absence of enzymes, the amount of the drug released in the colon, followed a ratio of 1.0:0.78:0.71. This ratio changes significantly in the presence of colonic enzyme cocktail and becomes 1.0:0.74:0.62. It is important to mention that all though the bead formulations released the drug mostly in simulated colonic environment, however, a small portion of the loaded drug also get released at other part of GI tract (simulated stomach and small intestine environment). This could further be prevented by finer tuning of the polymeric formulations or using an enteric coating on the beads with polymers such as Eudragit® . The enteric coating polymers offer various advantages including pH-dependent drug release, protection of actives sensitive to gastric fluid, protection of gastric mucosa from aggressive actives. In addition, these polymers offers GI and colon targeting depending on their grades, for example; Eudragit® S100-55, Eudragit® L100 and Eudragit® S100

dissolves above pH 5.5, 6 and 7 respectively, thus offering selectivity in targeted site for delivery. However, the coating must not interfere with the mucoadhesive property of the CA-CMC beads formulations. Our previous biodegradation study has suggested that in all CA-CMC beads formulations, T1G showed the highest rate of biodegradation followed by T3G and T5G, respectively. However, our mucoadhesivity study suggests that beads of all formulations exhibited higher adhesion with colonic mucosa at pH 6.8. This clearly indicates the higher retention of the CA-CMC beads in the colon which will allow the degradation of the beads by colonic microflora, aiding in a faster and complete drug release. Such variations in the rate of degradation along with higher mucoadhesivity will allow the usage of CA-CMC formulations for colon-specific drug delivery applications. Further, the FRAP analysis of CA-CMC beads in PBS (pH 6.8) demonstrated that the recovery of the FITC-Dextran at the bleached ROI attained 84, 79 and 77% of its initial intensity in T1G, T3G and T5G, respectively (Fig. 5B). Such variations in the FRAP profile are due to their compositions which aids in the variation in the local viscosity of the formulations and microarchitecture

T. Agarwal et al. / International Journal of Biological Macromolecules 75 (2015) 409–417

of the beads. These factors contribute to the variations in the internal solute transport. However, the rate of recovery of fluorophore increased when the beads were exposed to the microbial enzyme cocktail. A complete recovery (approx. 100%) was obtained in the case of T1G when exposed to the enzyme cocktail while in case of T3G and T5G; nearly 90 and 81%, respectively, of the recovery could be achieved. 3.7. In vitro evaluation of therapeutic potential of anticancer drug (5-FU)-loaded beads 5-FU-loaded beads were evaluated for their potential as a drug carrier system against HT-29 adenocarcinoma cell lines. The viability of the cells was estimated by MTT assay [41]. The results demonstrated that the cell viability decreased significantly after 48 h of exposure of cells to the drug-loaded beads of all formulations. The percentage cell death in case of T1G was found to be 1.29 folds higher than T3G and 1.42 folds higher than T5G (Fig. 6A). The percent cell death observed in all the three formulations was significant with respect to the control (without drug) and amongst themselves (p < 0.005). It is also important to mention that the blank beads of all the formulations were found to be biocompatible (data not shown). Furthermore, in response to the bead treatment, apoptosis in HT-29 cells was analyzed by flow cytometry (Fig. 6B) and by checking nuclear condensation and fragmentation (Fig. 6C). Cell-cycle analysis showed that the percentage apoptosis was highest for T1G (22.5%; late apoptotic 6.97, early apoptotic 15.5), followed by T3G (15.7%; late apoptotic 5.61%, early apoptotic 10.08%) and T5G (15.4% late apoptotic 5.88%, early apoptotic 9.55% respectively). The same for control was 7% (late apoptotic 1.72%, early apoptotic 5.32%). Further, immunocytochemistry data showed that the cells, which were exposed to leachant from 5-FU drug-loaded beads, underwent nuclear condensation and fragmentation. In comparison, the leachants from the blank beads did not show any such apoptotic response. This data clearly suggest that the CA-CMC beads can work as an effective colon targeted drugdelivery system and the efficacy of the formulation can be modified be by changing the composition of the bead. 4. Conclusion Conventional strategies of colon-targeted drug delivery rely upon the exploitation of either pH-sensitive property of the polymer or biodegradability (especially microbial degradation of the carrier matrix). Recently, combined approach of pH-dependent and microbially triggered drug-delivery (CODESTM) system is employed for the aforesaid purpose with reasonable success. However, none of the existing system has exploited the combined effect of pH sensitivity, colonic microbial degradation and colon mucosa-specific preferential mucoadhesivity. In this regard, the present study is novel in the field of colon-targeted delivery system. In this study, we have successfully proved that calcium alginate-carboxymethyl cellulose beads are potential candidates for colon-specific oral delivery of therapeutics. We showed that these beads can effectively protect the therapeutics from the harshlytic environment of the stomach during its transit through the GI tract and can preferentially deliver the drug at colon under the influence of colonic pH and microflora. It was evident from the study that such formulations are working well in in vitro colon cancer model. The study gives a clear indication that such formulation

417

can be used for delivery of other therapeutics. The system reported here is a prototype only and improvement in colon-specific delivery of therapeutics from the formulation is highly possible by changing the composition. However, the study needs to be substantiated in in vivo system for further progress. Acknowledgement The authors would like to thank Dr. T.K. Maiti and Dr. Bibhas Roy, Department of Biotechnology, Indian Institute of Technology, Kharagpur, for providing confocal imaging facility. References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41]

A.K. Philip, B. Philip, Oman Med. J. 25 (2010) 79–87. B.P. Kolte, Asian J. Biomed. Pharm. Sci. 2 (2012) 21–28. L. Yang, J.S. Chu, J.A. Fix, Int. J. Pharm. 235 (2002) 1–15. A. Jain, Y. Gupta, S.K. Jain, J. Pharm. Pharm. Sci. 10 (2007) 86–128. V. Sinha, R. Kumria, Int. J. Pharm. 224 (2001) 19–38. M. Chourasia, S. Jain, J. Pharm. Pharm. Sci. 6 (2003) 33–66. S.V. Sastry, J.R. Nyshadham, J.A. Fix, Pharm. Sci. Technol. 3 (2000) 138–145. F. Talaei, F. Atyabi, M. Azhdarzadeh, R. Dinarvand, A. Saadatzadeh, Eur. J. Pharm. Sci. 49 (2013) 712–722. S.K. Bajpai, S. Sharma, React. Funct. Polym. 59 (2004) 129–140. R.E. Forster, F. Thürmer, C. Wallrapp, A.W. Lloyd, W. Macfarlane, G.J. Phillips, J.-P. Boutrand, A.L. Lewis, J. Mater. Sci. Mater. Med. 21 (2010) 2243–2251. D. Tahtat, M. Mahlous, S. Benamer, A.N. Khodja, H. Oussedik-Oumehdi, F. Laraba-Djebari, Int. J. Biol. Macromol. 58 (2013) 160–168. M. George, T.E. Abraham, J. Control. Release 114 (2006) 1–14. S.C. Kumar, A.K. Kalekar, Int. J. Pharm. Sci. Drug Res. 4 (2012) 115–122. K. Manjanna, K.T. Pramod, B. Shivakumar, Drug Discov. Ther. 4 (2010) 109–122. J.-F. Su, Z. Huang, X.-Y. Yuan, X.-Y. Wang, M. Li, Carbohydr. Polym. 79 (2010) 145–153. M.A. Abd El-Ghaffar, M.S. Hashem, M.K. El-Awady, A.M. Rabie, Carbohydr. Polym. 89 (2012) 667–675. J.-P. Zhang, Q. Wang, X.-L. Xie, X. Li, A.-Q. Wang, J. Biomed. Mater. Res. B 92B (2010) 205–214. A. Sannino, C. Demitri, M. Madaghiele, Materials 2 (2009) 353–373. V. Grabovac, D. Guggi, A. Bernkop-Schnürch, Adv. Drug Deliv. Rev. 57 (2005) 1713–1723. K. Girhepunje, V. Krishnapiillai, R. Pal, H. Gevariya, N. Thirumoorthy, Int. J. Curr. Pharm. Res. 2 (2010) 46–55. G. Pasparakis, N. Bouropoulos, Int. J. Pharm. 323 (2006) 34–42. C.-M. Lehr, J.A. Bouwstra, J.J. Tukker, H.E. Junginger, J. Control. Release 13 (1990) 51–62. P. Sriamornsak, J. Nunthanid, K. Cheewatanakornkool, S. Manchun, AAPS Pharm. Sci. Tech. 11 (2010) 1315–1319. M.Z. Ahmad, S. Akhter, M. Anwar, F.J. Ahmad, Mol. Pharm. 9 (2012) 2986–2994. R.G. Deschesnes, J. Huot, K. Valerie, J. Landry, Mol. Biol. Cell 12 (2001) 1569–1582. B. Arica, S. Calis, P. Atilla, N. Durlu, N. Cakar, H. Kas, A. Hincal, J. Microencapsulation 22 (2005) 153–165. A.K. Anal, D. Bhopatkar, S. Tokura, H. Tamura, W.F. Stevens, Drug Dev. Ind. Pharm. 29 (2003) 713–724. S. Distantina, R. Rochmadi, M. Fahrurrozi, W. Wiratni, Eng. J. 17 (2013) 57–66. A.W. Chan, R.A. Whitney, R.J. Neufeld, Biomacromolecules 10 (2009) 609–616. M.S. Kim, S.J. Park, B.K. Gu, C.-H. Kim, Appl. Surf. Sci. 262 (2012) 28–33. M. Saeedi, J. Akbari, R. Enayatifard, K. Morteza-Semnani, M. Tahernia, H. Valizadeh, Iran. J. Pharm. R. 8 (2010) 241–249. J. Wang, P. Somasundaran, J. Colloids Interface Sci. 291 (2005) 75–83. T.R.R.S. Rahamatullah Shaikh, M.J. Garland, A.D. Woolfson, R.F. Donnelly, J. Pharm. Bioallied Sci. 3 (2011) 89–100. A. Ahuja, R.K. Khar, J. Ali, Drug Dev. Ind. Pharm. 23 (1997) 489–515. Y.K. Yadav, A.B. Gupta, R. Kumar, J.S. Yadav, B. Kumar, J. Chem. Pharm. Res. 2 (2010) 418–432. X. Cao, R. Bansil, K.R. Bhaskar, B.S. Turner, J.T. LaMont, N. Niu, N.H. Afdhal, Biophys. J. 76 (1999) 1250–1258. I. Sekirov, S.L. Russell, L.C.M. Antunes, B.B. Finlay, Physiol. Rev. 90 (2010) 859–904. T.T. Kararli, Biopharm. Drug Dispos. 16 (1995) 351–380. K. Kesavan, G. Nath, J.K. Pandit, Sci. Pharm. 78 (2010) 941–957. K. Nishida, Y. Ando, M. Enomoto, Colloid Polym. Sci. 260 (1982) 511–513. P. Twentyman, M. Luscombe, Br. J. Cancer 56 (1987) 279–285.