Papain Entrapment in Alginate Beads for Stability Improvement and Site Specific Delivery : Physicochemical Characterization and Factorial Optimization using Neural Network Modeling

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.





Papain Entrapment in Alginate Beads for Stability Improvement
and Site Specific Delivery: Physicochemical Characterization and
Factorial Optimization using Neural Network Modeling



Mayur G. Sankalia, Rajshree C. Mashru,* Jolly M. Sankalia, and Vijay B. Sutariya


Center of Relevance and Excellence in NDDS,
Pharmacy Department,
G. H. Patel Building,
The M. S. University of Baroda,
Vadodara, Gujarat, India – 390 002
Telephone: +91-265-2434187 / 2794051.
Fax: +91-265-2418927.
E-mail: sankalia_mayur@hotmail.com










*Corresponding Author: Dr. (Ms.) Rajshree C. Mashru, Center of Relevance and
Excellence in NDDS, Pharmacy Department, The M. S. University of Baroda, G. H. Patel
Building, Donor’s Plaza, Vadodara, Gujarat, India – 390 002; Telephone: +91-265-2434187 /
2794051; Fax: +91-265-2418927; E-mail: sankalia_mayur@hotmail.com


1

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

ABSTRACT

This work examines the influence of various process parameters (like sodium alginate
concentration, calcium chloride concentration and hardening time) on papain entrapped in
ionotropically crosslinked alginate beads for stability improvement and site specific delivery
to the small intestine using neural network modeling. A 33 full factorial design and feed
forward neural network with multilayer perceptron was employed to investigate the effect of
process variables on % entrapment, time required for 50% and 90% of the enzyme release,
particle size and angle of repose. Topographical characterization was carried out by SEM and
entrapment was confirmed by FTIR and DSC. T50 and T90 were increased with increase in all
three process variables. Percentage entrapment and particle size were found to be directly
proportional to sodium alginate concentration and inversely proportional to calcium chloride
concentration and hardening time, while angle of repose and degree of cresslinking showed
exactly opposite proportionality. Beads with more than 90% entrapment and T50 of less than
10 min could be obtained at the low levels of all three process variables. Inability of bead to
dissolve in acidic environment, while complete dissolution in buffer of pH 6.8 or higher,
showed suitability of beads to release papain into small intestine. Shelf-life of the capsules
prepared using the papain loaded alginate beads was found to be 3.60 years compared to 1.01
year of the marketed formulation. It can be inferred from the above results that the proposed
methodology can be used to prepare papain loaded alginate beads for stability improvement
and site specific delivery.

Keywords: alginate beads; neural network; multilayer perceptron, optimization; entrapment;
factorial design; papain


2

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

INTRODUCTION

Papain (EC 3.4.22.2) is one of the thiol proteases, and its active site consists of Cys-25, His-
159 and Asp-158. Papain shows extensive proteolytic activity towards proteins, short-chain
peptides, amino acid esters and amide links, and is applied extensively in the fields of food
and medicine. The reverse reaction of 'papain hydrolysis', can be used in the synthesis of
peptides and oligomers. Most marketed formulations containing papain and other digestive
enzymes need to be stored at cold (2 – 8º C) or cool (8 – 25 º C) temperatures conditions and
still have the self-life of not more than one year. Entrapment of the papain in ionotropically
crosslinked biodegradable hydrogels may improve the stability of the parent enzymes and
make it less prone to interference of various formulation excipients. Immobilized enzymes
are stable at higher temperature and might be stored at room temperature with extended self-
life.1 Optimum pH for activity of papain is in the range of 3-9 which varies with different
substrate.2 However, papain is almost inactive at gastric pH of 1.2 so the ideal place for
papain delivery is the small intestine. Multiple unit dosage forms are particularly useful: (i)
for delivering highly-irritant drugs such as nonsteroidal anti-inflammatory drugs (NSAIDs);3
(ii) for site-specific targeting of acid labile drugs within the gastrointestinal tract;4 and (iii) for
delivery of enzymes, peptides/proteins and vaccines.5 Above advantages are of great
commercial interest for the pharmaceutical industries hence it was the objective of the
research to develop an extended shelf-life formulation for site specific delivery of papain by
immobilization in ionotropically crosslinked biodegradable alginate beads which results in
better and efficient utilization of enzyme. This paper will also deal with ‘in vitro’ dissolution
studies of beads and physicochemical characterization for evaluating the bead formation and
its release behavior.

3

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

Alginate, a high-molecular-mass polysaccharide, is a naturally occurring biodegradable
copolymer of 1,4-linked ?-D-mannuronic acid (M) and ?-L-guluronic acid (G) and extracted
from brown seaweeds (Phaeophyceae, mainly Laminaria). It has been shown that the G and
M units are joined together in blocks and as such, three types of blocks may be found: homo-
polymeric G blocks (GG), homopolymeric M blocks (MM) and heteropolymeric sequentially
alternating blocks (MG). The reactivity with calcium and the subsequent gel formation
capacity is a direct function of the average chain length of the G blocks. Hence, alginates
containing the highest GG fractions possess the strongest ability to form gels. This initially
arises from the ability of the divalent calcium cation to fit into the guluronate structures like
eggs in an “egg box junctions”. Consequently, this binds the alginate chains together by
forming junction zones, and sequentially leading to gelling of the solution mixture and bead
formation. When an aqueous solution of sodium alginate is added dropwise to an aqueous
solution of calcium chloride, a spherical gel with regular shape and size, also known as
‘‘alginate bead’’. Alginate beads have the advantages of being nontoxic orally and having
high biocompatibility.6 Another advantageous property is their inability to reswell in acidic
environment while easily reswells in alkaline environment, so acid-sensitive drugs
incorporated into the beads would be protected from gastric juice.7 Therefore, alginate is used
as an entrapment matrix for cells and enzymes as well as for pharmaceutical and food
adjuvants. In the past, conventional crosslinked calcium–alginate beads have been
investigated for the development of a multiple unit drug delivery system.8-11 However, not
even a single reference could be cited in literature till date for entrapment of papain in
alginate beads for improvisation of shelf-life.
Neural network (NN) models might generalize better than regression models since regression
analyses are dependent on predetermined statistical significance levels (i.e. less significant
terms are not included in the model).12, 13 With the NN method all data are used potentially

4

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

making the models more accurate. Hence NN was selected for modeling and evaluating tool
in this paper. The use of at least one hidden layer enables the NNs to describe nonlinear
systems.12 One layer is usually sufficient to provide adequate prediction even if continuous
variables are adopted as the units in the output layer and also there is a little evidence to
suggest that a larger number of hidden layers improves performance.14
The multilayer perceptron (MLP) with backpropagation algorithm is one of the most widely
implemented neural network topologies and is important in the study of nonlinear dynamics.
Two important characteristics of the MLP are: its smooth nonlinear neurons (sigmoidal
function); and their massive interconnectivity.
NN has been successfully applied to many pharmaceutical areas in recent years15 e.g.: QSAR
analysis and drug modeling,16 pharmacokinetic17–pharmacodynamic studies,18 optimization
and pharmaceutical formulation development,19 powder flow,20 compound determination
using HPLC,21 analysis of NMR spectra,22 prediction of drug release profile,23 prediction of
physicochemical properties,24 prediction of octanol-water partition coefficient,25 prediction
of solubility,26 etc.


5

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

MATERIALS AND METHODS

Materials

Hammersten type casein USP (Himedia Laboratories Pvt. Ltd., Mumbai, India) and
trichloroacetic acid (98.0%, Qualigens Fine Chemicals, Mumbai, India) were used as
received. Purified papain IP, sodium alginate IP, calcium chloride dihydrate (98.0%), dibasic
sodium phosphate (99.5%), disodium ethylenediaminetetraacetate (99.5%), cystein
hydrochloride (99.0%), and citric acid (98.0%) were purchased from S. D. Fine-Chem Ltd.,
Mumbai, India. All the other chemicals and solvents were of analytical grade and were used
without further purification. Deionized double-distilled water was used through out the study.

Preparation of Beads

Concentrated sodium alginate solution in distilled water was prepared well before required.
Required quantity of enzyme (200 mg papain in 50 ml of final sodium alginate solution) was
dissolved in small quantity of water and mixed with concentrated sodium alginate solution.
Final concentration of sodium alginate was adjusted in the range of 1-2% w/v and was used
after being degassed under a vacuum. The beads were prepared by dropping the sodium
alginate solution (10 ml) containing papain from the dropping device such as syringe with
26G×½” flat-tip hypodermic needle to a magnetically stirred calcium chloride solution (40
ml) at a rate of 5 ml/min and were allowed to harden for specific time. Different levels (Table
1) of sodium alginate, calcium chloride and hardening time were selected. The beads were
collected by decanting calcium chloride solution, washed with deionized water and dried to a

6

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

constant weight in vacuum desiccator (Tarsons Products Pvt. Ltd., Kolkata, India) at room
temperature for 36 hours.

Factorial Design

In this study a 33 full factorial design was used to determine the effect of the sodium alginate
concentration, the calcium chloride concentration, and the hardening time. Before the
application of the design, a number of preliminary trials were conducted to determine the
conditions at which the process resulted to beads. The matrix of the experiments and the
results of the responses are listed in Table 2. To determine the experimental error, the
experiment at the centre point was repeated five times at different days. The mean %
entrapment, T50, T90, particle size and angle of repose of these experiments were
85.88±1.02%, 15.57±0.28 min, 82.43±0.33 min, 261.1±0.9 µm and 20.61±0.29º respectively.
The above-mentioned values showed good reproducibility of the process.

Neural Network Software and Network Topology

The Microsoft®-Windows® based neural network software; NeuroSolutions® Version 4.24
(NeuroDimension, Inc.) was used. The MLP with single hidden layer architecture was
chosen. The experimental matrix of 32 input:desired output data sets (Table 2) was
introduced in to the model, with three input neuron (process variables), one hidden layer and
five output neuron (response variables). Various adjustable parameters like number of
neurons in hidden layer, step size and momentum of hidden layer and output layer, etc. were
optimized. Training was repeated for 3 times for optimization of all parameters. At the start
of the training run, weights were initialized with random values. During training, 5 additional

7

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

data sets of input:desired output were used for the cross-validation and was back-propagated
through the network to evaluate the trained network. The training termination criterion was
the rise in minimum standard error (MSE) of the cross valibation set than that of the training
set for continuous 100 epochs. The network trained under optimum conditions was used to
predict responses at different factor values and response surface were generated for
interpretation.

Characterization of Beads

Estimation of Papain
Papain was estimated by modified casein digestion method of USP XXVI in presence of
cysteine hydrochloride. Different aliquots of standard papain solution in phosphate-cysteine
disodium ethylenediaminetetraacetate buffer were added to 5 ml of buffered substrate
(hammersten-type casein 10 mg/ml, pH 6.0±0.1) and incubated for 60 min at 40°C. Digestion
process of casein was stopped by adding 3 ml of 30% w/v trichloroacetic acid solution and
allowed to stand for 30 – 40 min at 40°C. Digested amino acids were filtered through
whatman® filter paper no. 42 by discarding first 3 ml of filtrate and absorbance was measured
at 280 nm against their respective blanks. The method was found to be linear over an
analytical range of 3 – 100 µg/ml with correlation coefficient (r) of 0.9996. Limit of
detection, limit of quantitation, and regression equation were found to be 0.77 µg/ml, 2.57
µg/ml, and y = 0.0042x – 0.0033 respectively.

Determination of Entrapment Efficiency
Entrapment efficiency was determined by dissolving the enzyme loaded beads in a
magnetically stirred simulated intestinal fluid without enzyme (USP XXVI) for about 45 min.

8

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.

The resulting solution was centrifuged at 2500 rpm for 10 min (Remi Instruments Ltd,
Mumbai, India) and supernatant was assayed (n=3) for enzyme content by modified casein
digestion method of USP XXVI. Entrapment efficiency was calculated as:
loaded

Enzyme
efficiency

Entrapment
=
×100
(1)
loading

enzyme
l
Theoratica

Effect of pH on Release Profile
To study the effect of pH on papain release profile, ‘in vitro’ dissolution study was carried
out using the USP XXVI dissolution apparatus 2 (TDT-60T, Electrolab, Mumbai, India) in
500 ml of different pH media (simulated gastric juce pH 1.2 (USP), phosphate buffer pH 4.0
(IP), neutralized phthalate buffer pH 5.4 (IP), simulated intestinal fluid without enzyme pH
6.8 (USP), phosphate buffer pH 7.4 (IP), and phosphate buffer pH 8.0 (IP)) on the optimized
batch at 37±0.5?C with paddle speed of 75 rpm. Accurately weighed samples (n=3)
equivalent to about 40 mg of papain were introduced to dissolution media and samples of 2
ml were collected at 0, 0.25, 0.50, 0.75, 1.0, 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5 and 6.0 hr.
Samples were filtered through 0.4 µm whatman® membrane filter and assayed for enzyme
content as before.

Determination of T50 and T90
Time required for 50 (T50) and 90 (T90) percent of enzyme release are important parameters
for enzyme release study and were used to evaluate the onset of action and duration of action
respectively. For optimization purpose, dissolution study of all batches was carried out in 500
ml of simulated intestinal fluid without enzyme as before. Accurately weighed samples (n=3)
equivalent to about 40 mg of papain were subjected to dissolution and aliquots of 2 ml were
assayed at 0, 5, 10, 15, 20, 30, 45, 60, 90 and 120 min. T50 and T90 were found by
extrapolating the % enzyme released versus time plot.

9

---

© AAPS PharmSciTech. Accepted: March 10, 2005.
Author’s final version.


Particle Size Measurements
Particle size is an important parameter for the formulation development. Optimized batch of
the beads was filled in the capsules during which the particle size was the evolutionary
parameter. Larger particles show higher weight variation during capsule filling, hence the
experimental conditions result in smaller particles is preferable. Particle size was determined
with the laser diffraction particle size analyzer (MAN 0244/ HYDRO 2000 SM, Malvern
Instruments Ltd., UK) using isopropyl alcohol as a vehicle.

Angle of Repose Measurements
Angle of repose was measured for estimating flowability of the beads. If the angle exceeds
50?, the material will not flow satisfactorily while materials having values near the
minimum, flow easily and well. The rougher and more irregular the surface of the particles,
the higher will be the angle of repose. The angle also increases with decrease in particle size.
Angle of repose was measured by passing beads through a funnel on the horizontal surface.
The height (h) of the heap formed was measured with a cathetometer and the radius (r) of the
cone base was also determined. The angle of repose (?) was calculated from:
h
tan ? =










(2)
r

Fourier Transform Infra-red Spectroscopy (FTIR )
IR transmission spectra were obtained using a FTIR spectrophotometer (FTIR-8300,
Shimadzu, Japan). A total of 5% (w/w) of sample, with respect to the potassium bromide
(KBr) disc, was mixed with dry KBr (S. D. Fine Chem Ltd., Mumbai, India). The mixture
was ground into a fine powder using an agate mortar before compressing into KBr disc under

10

---