Chapter 7. Self-Assembled hollow nanocapsule from amphiphatic
7.6 DOX Loading and Release Behavior of CHC Nanocapsules
50 nm
9 nm 9 nm
Figure 7-7 SEM images of CHC-4 nanocapsules.
7.6 DOX Loading and Release Behavior of CHC Nanocapsules
To investigate the drug-loading capacity of the CHC nanocapsules, the model drug, doxorubicin (DOX), was loaded into the CHC nanocapsules.
Figure 7-8a shows the encapsulation efficiency (EE) of DOX from CHC nanocapsules with different DH. As reported by Kitaeva et al. [117], since DOX molecule contains an amino group with a pKa of 8.6, it is expected that DOX can form polyelectrolyte complex with the carboxymethyl groups of CC and CHC. Hence, for the CC (where its DH is zero), it shows a lower encapsulation efficiency of around 26.3 %. However, as increase the level of hydrophobic substitution, the DOX encapsulation efficiency increased from 26.3 to 46.8%
as the DH increased form 0 to 0.48. The improved encapsulation efficiency with increase of DH of the CHC should be caused by the larger reservoir space and the increase of the hydrophobic interaction of inner shell, which restricts the outward diffusion of DOX from nanocapsules.
91
The release of DOX from the CHC nanocpasules was investigated using a dialysis procedure. A solution of the DOX-loaded CHC nanocapsules in a dialysis cassette was dialyzed against 100 mM buffer and the solution in the cassette was sampled at various times over a 7 day period to determine the amount of DOX remaining in the nanocapsules. Figure 7-8b shows the release profiles of DOX and DOX-loaded CHC nanocapsules with different levels of DH.
Figure 7-8 (a) DOX encapsulation efficiency and (b) DOX release profiles of CHC hollow nanocapsules
92
93
It was observed that DOX and DOX-loaded CC had a faster diffusion rate with essentially complete release after 24-36 hrs. This is consistent with the expected rate of diffusion for low molecular weight molecules across the dialysis membrane. However, DOX-loaded CHC nanocapsules show a two-step release profile. In the first step (0~12 hrs), the rapid release of DOX may caused by the diffusion of DOX on the surface of nanocapsules out of the dialysis membrane. Then, the followed a relatively slow release, which may be dominated by the DH of CHC nanocapsules. It can be found that CHC-1 showed a total release period of about 96 hrs. By contrast, with the increase of DH (i.e., CHC-2, CHC-3, and CHC-4), the entrapped DOX showed a continuous release over 7 days. It is reasonably believed that the increase of hydrophobic groups enhances encapsulation efficiency and payload of the DOX within CHC nanocapsules. These results indicate that these CHC self-assembled hollow nanocapsules can be a suitable candidate for controlled DOX delivery for anti-cancer purpose.
94
Chapter 8
Hydrophobic Effect on the Structural Evolution of Acylated-Carboxymethyl Chitosan and its Self-Assemble Forming of Doxorubicin-loading
Nanocapsule
8.1 Introduction
The construction of supra-molecular assemblies with well-defined nanostructure is of great interest owing to their potential applications in diverse fields. Since 1995, Eisenberg et al. [118] found the multiple morphologies of crew-cut micelle-like aggregates of polystyrene-b-poly (acrylic acid) diblock copolymer in solution, the preparation and assembly of amphiphatic polymers of various architectures has become a research area of interest. Polymer amphiphiles enable the preparation of nano-aggregates with shape ranging from spheres [119] to cylinders [120], vesicles [121], donuts [122], and other geometry [123] by self-assembly. Of the various aggregates, hollow vesicles represent an important class of materials because their unique structural and surface properties may lead to a wide range of applications, especially in medicine such as capsule agents for drug delivery, encapsulation of large quantities of guest molecules, and gene therapy [124].
Recently, many efforts have been performed to prepare biodegradable and non-toxic polymeric amphiphatic on the basis of natural biomaterials such as polysaccharides. Among those natural biopolymers, hydrophilically, hydrophobically, and amphiphatically modified chitosan derivatives were
95
studied and realized to form monodisperse self-aggregated nano-particles by ultrasonication in aqueous solution. However, it is more technologically desirable and critical if (a) a hollow structure can be easily achieved using chitosan derivatives without the need of extensive ultrasonication, where a nanometric carrier with higher load capacity of drugs or proteins can be expected with lower operation cost, (b) a variation of the amphiphilic nature of the resulting carrier allows an enhanced affinity to different drugs or proteins, e.g., hydrophilic drugs or hydrophobic drugs, where an improved encapsulation efficiency can be designed and (c) an adjustable biocompatibility and biodegradability is clinically manageable. In chapter 7, a simple core-template-free strategy was successfully proposed and a new type of amphiphatic chitosan hollow nano-capsules was synthesized by a self-assembly mechansim. In general, polymer amphiphiles consisting of hydrophilic and hydrophobic segments can form micelles or micelle-like self-assemblies with a hydrophobic core and a hydrophilic shell due to non-covalent association arising from intra- and/or inter-molecular interactions among hydrophobic segments in aqueous media. On the other hand, among these versatile micelles, the hollow cavity enclosed by a double layer membrane is also possibly formed [125]. However, little scientific evidences can be sufficiently employed for better understanding of such a hollow cavity development for the amphiphilic chitosan.
Hence, as one of the research objectives in present study, structural evolution of the chitosan-based nano-aggregates was systematically investigated in terms of different degrees of acyl substitution (DH) and number of carbon of the acyl ligand (Cn). In addition, the doxorubicin (DOX)-encapsulated capacity of the nano-aggregates of varying degrees of
96
hydrophobicity, in terms of DH and Cn, was also studied.
8.2 Synthesis of carboxymethyl chitosan (CC)
Following our previous report [126], 5 g of chitosan was suspended in 2-propanol (50 mL) at room temperature while being stirred for 30 min. The resulting suspension was gently mixed with 12.5 mL NaOH solution. The mixture containing NaOH of 13.3M was mixed with 25 g of chloroacetic acid to prepare carboxymethyl chitosan (CC) sample with a high degree of carboxymethyl substitution. The resulting suspensions were stirred for 30 min and heated to 60°C for 4 h, followed by filtration, washing by methanol solution, and drying.
8.3 Synthesis of acylated carboxymethyl-chitosan (ACC)
The obtained CC sample (2g) was dissolved in distilled water (50 mL) under vigorously stirring for 24 h. These resulting solutions were then mixed with methanol (50 mL), followed by addition of acyl anhydrides, namely, acetic anhydride, hexanol anhydride, decanoic anhydride, and dodecanoic anhydride, representing acyl groups of various chain lengths (or carbon numbers). The acylated chitosan derivatives were prepared as previously described [16]. Acetyl, hexanoyl, and decanoyl carboxymethyl-chitosan were prepared at ambient temperature for 20 h. Dodecanoic carboxymethyl-chitosan was prepared at 50oC for 2h, followed by standing for 18h. Different ratios of acyl anhydride to amino functionalities of CC samples were designed in order to prepare the acylated-carboxymethyl chitosan (ACC) of different degree of hydrophobic substitution (DH) until the value of DH reached up to 0.5. The DH is defined as the average number of acyl group per
97
repeat unit multiplied by 100, calculated according to the nitrogen content which determined by elemental analysis [127]. Finally, the resulting solutions were collected by dialysis membrane after dialysis with ethanol solution (25%
v/v) for 24 h. The chemically modified chitosan was expressed using the value of n and m to define the carbon number of hydrophobic side chain (Cn) and DH, respectively, of ACC derivative in the Cn-m, where n ranges from 2, 6, 10, to 12 and m ranges from 0.125, 0.25, 0.375, to 0.5. For example, C2-0.25 means the ACC having the approximate value of the degree of hydrophobic substitution of 0.25. The preparation conditions and the characteristics of ACC samples are given in Table 8-1.
8.4 Preparation of drug-loaded and non-loaded ACC nano-aggregates 100 mg/ml ACC samples were separately suspended in distill water under gentle shaking at 25oC for 24 h, followed by ultrasonication using a probe type sonifier (Automatic Ultrasonic Processor UH-500A, China) at 30W for 2 min till an optically clear solution was obtained. The sample solutions were filtered through a filter (1.0 μm, Millipore) to remove dust and impurity.
The final solutions were then stored in stock for a subsequent sample characterization.
Drug-loaded ACC nano-aggregates were prepared by dissolving 20 μg/mL of doxorubicin (DOX) in 20 ml acquired suspension. Insoluble, free DOX was removed by centrifugation at 2000 rpm and 4oC for 5 min. The drug-containing nanoparticles were then separated from the aqueous solution by centrifugation at 15000 rpm and 4oC for 15 min. Drug concentration in the supernatant was analyzed by ultraviolet absorption (UV) at wavelength of 490 nm, a strong characteristic absorption band of DOX, with reference to a
98
calibration curve on a UV-VIS spectrometer (SP-8001, Metertech Inc.). The measurements were performed in triplicate. The amount of the drugs encapsulated in the nano-aggregates was then calculated by the total amount of DOX subtracting the residual DOX in the supernatant. Encapsulation efficiency (EE) were obtained as described below
EE= (A-B)/A×100 where A is the total amount of the DOX, B is the amount of DOX remaining in
the supernatant and C is the weight of the ACC nano-aggregates.
99
Table 8-1 The preparation conditions and the characteristics of ACC nano-aggregates
Samples Acyl group DHa
[(RCO)2O]/
[GlcN]
Temp (oC) /
Time (h) CACb×10-2 Wnf,maxc
(%)
CC - 0 - - 25.0 10.3
C2-0.125 Acetyl 0.11 0.2 25/20 21.0 11.8
C2-0.25 Acetyl 0.24 0.4 25/20 12.7 14.5
C2-0.325 Acetyl 0.31 0.5 25/20 7.20 15.9
C2-0.5 Acetyl 0.56 1 25/20 5.00 17.9
C6-0.125 Hexanoyl 0.13 0.2 25/20 9.20 12.4
C6-0.25 Hexanoyl 0.26 0.4 25/20 31.7 19.8
C6-0.325 Hexanoyl 0.33 0.5 25/20 0.91 24.8
C6-0.5 Hexanoyl 0.48 1 25/20 0.40 29.3
C10-0.125 Decanoyl 0.12 0.2 25/20 6.20 15
C10-0.25 Decanoyl 0.24 0.4 25/20 1.80 24.3
100
C10-0.325 Decanoyl 0.34 0.5 25/20 0.57 41.2
C10-0.5 Decanoyl 0.45 1 25/20 0.36 57.4
C12-0.125 Dodecanoyl 0.1 0.2 50/2+25/18 5.00 16.2
C12-0.25 Dodecanoyl 0.22 0.4 50/2+25/18 1.20 31.4
C12-0.325 Dodecanoyl 0.3 0.5 50/2+25/18 0.36 45.7
C12-0.5 Dodecanoyl 0.44 1 50/2+25/18 0.32 63.9
a DH determined by element analysis.
b Critical aggregation concentration determined from I372/I385 data..
c The content of bound water obtained by the DSC test.
8.5 Structural Characteristics of ACC
As shown in Scheme 8-1, ACC was synthesized using CC as the starting precursor. In previous work, it was demonstrated that a hydrogen atom on the amino group of CC can be replaced by acyl group, as confirmed by 1H NMR analysis [126]. However, in this study, the signal intensity of the characteristic acyl protons was not in proportion to the DH of the acyl groups, which suggested that acyl groups aggregated to form hydrophobic micro-domains to minimize their interaction in D2O (the resulting spectra were not shown here). This trend in the 1H NMR spectra was consistent with other polymeric amphiphiles that formed aggregates in the aqueous phase [128].
Hence, the DH of ACC samples were determined by elemental analysis [127]
in this study and the ACC samples with different DH and Cn were prepared through the control of various ratios of acyl anhydride to amino group of the CC, which is listed in Table 8-1.
R=C n-2 H 2(n-2) CH 3 ; n=2, 6, 10, 12.
Scheme 8-1 Acylation Reaction of carboxymethyl chitosan
101
Scheme 8-1 shows the molecular structure of the amphiphilically modified chitosan with different lengths of acyl side chains (with carbon number, Cn, ranging from 2, 6, 10, to 12) attached on the amino groups of chitosan. From the FT-IR analysis, shown in Fig. 8-1, the absorption peaks of
CS at ca. 1655 cm-1 can be assigned to the carbonyl stretching of secondary amides (amide I band), at 1570 cm-1 to the N–H bending vibration of nonacylated 2-aminoglucose primary amines, and at 1555 cm-1 to the N–H bending vibrations of the amide II band [127]. The presence of both 2-aminoglucose and 2-acetamidoglucose repeat units was confirmed by bands at 1655, 1570, and 1555 cm-1. After carboxymethylation, CC shows absorption peak at ca. 1730 cm-1, which was assigned to the carboxymethyl dimer (O=COH…O=COH). Compared to CC, the vibrational band of C12-0.5 corresponding to primary amino groups at 1570 cm-1 was disappeared, while prominent bands at 1555 cm-1 were observed. In addition, peaks at 2850–2950 cm-1 ascribed to –CH2 with their absorption intensity was proportional to the acyl chain length [129]. These results clearly confirmed a successful acyl substitution on the carboxymethyl chitosan.
102
1800 1700 1600 1500 14003200 3100 3000 2900 2800 1570 cm-1
1555 cm-1 1723 cm-1
C12
C2
In te n sity ( a.u .)
CCWavelength (cm
-1)
-CH2
C12
C2 CC
Figure 8-1 FTIR spectra of CC, C2-0.5, and C12-0.5.
103
8.6 Self-assemble behavior of ACC
After dispersing in water, the ACC molecules can form self-assembled nanometric aggregates with an aid of ultrasonication for better dispersion as has been demonstrated in previous work [126]. The fluorescence probe technique can be used to monitor the self-aggregation behavior of ACC at a molecular level, and pyrene was chosen as a fluorescence probe. The intensity ratio of the first peak (372 nm) and the third peak (385 nm) I372/I385 in its fluorescent spectrum is quite sensitive to a subtle change in the environment around the pyrene molecules. Hence, the critical aggregation concentration (CAC), which is defined as the threshold concentration of self-aggregate formation by intra- and/or inter-molecular interactions, can be determined from the variation of the I372/I385 value of pyrene in the presence of polymeric amphiphiles. Fig. 8-2a shows the change of CAC of ACC nano-aggregates with different levels of DH and hydrophobic side chain length (Cn). It was found that the CAC of ACC with DH and Cn varying from 0.125 to 0.5 and from 2 to 12, respectively, has a smaller value than that of the CC (which was 0.25mg/ml as previously reported [126]), indicating that the ACC possesses stronger self-aggregation behavior than that of CC in aqueous solutions. This stronger tendency of the self-aggregation can be deduced from the increased hydrophobicity caused by the introduction of greater DH and Cn. In addition, it is also observed that the CAC values of ACC nano-aggregates were decreased with the increase of DH and/or Cn. It has realized that the increase of hydrophobicity in amphiphatic polymer causes the difficulty to dissolve in aqueous solution because the energy is essential to dissociate the H-bonding in water molecules (need to do work). Therefore, the dissolution of hydrophobic groups in water will give rise to an increase of the surface free
energy. In order to reduce the energy, the hydrophobic groups will aggregate to form micelle-like structure with hydrophobic core and hydrophilic shell, or a bi-layer structure of hydrophilic-hydrophobic-hydrophilic configuration. Hence, with larger DH and/or Cn, the ACC is able to form nano-aggregates easily at very low concentrations because it tends to reach the largest surface energy.
This demonstrates that the self-aggregation ability can be effectively enhanced with increasing hydrophobic nature, e.g., DH and/or Cn, of the ACC in aqueous environment. However, it seems to have the lowest value of CAC even if the DH or/and Cn is infinitely increased. This means that the CAC of the ACC nano-aggregates will reach a lowest value over a critical level of hydrophobicity.
In addition, a linear correlation between ln(CAC) and DH was observed, except the case when Cn is 12 and DH is 0.5, as shown in Fig. 8-2b. The same linear correlation between ln(CAC) and Cn can be also obtained. Such a linear correlation proves that the CAC of the ACC nano-aggregates decreased exponentially with the increase of DH and Cn. Hence, a mathematical relationship among ln(CAC), DH and Cn can be derived as follow,
1 1
1
) 1
ln(CAC =a XCn ⋅XDH +b XDH +c XCn +d at XDH < 0.5, XCn < 12 (8-1)
or
[
1 1 1 1expa X X b X c X d
CAC = cn⋅ DH + DH + cn +
]
at XDH < 0.5, XCn < 12 (8-2)104
where the coefficients, i.e., a1, b1, c1, and d1 having the value of 0.24, -6.87, -0.21, and -0.38, respectively, were obtained by regression analysis. From eq.
(8-2), it suggests that the hydrophobic effect resulting from both the DH and
side chain length, Cn dominates the self-aggregation ability of the ACC nano-aggregates. Compared to the case of C2, the slopes were apparently increased for the cases of C6 and C12 in the semi-log plot, Fig. 8-2b, indicating that a potentially stronger hydrophobic effect can be developed through a longer-chain acyl substitution, where a much lower value of CAC, by 2-3 orders of magnitude, can be obtained.
a
0.000-7 0.125 0.250 0.375 0.500 -6 the critical aggregation concentration, CAC and (b) ln(CAC), respectively.
106
8.7 Molecular configuration of water in ACC
The hydrophobic substitution is also found to affect profoundly water uptake behavior of the resulting ACC. It was generally accepted that the hydrophobic modification of amphiphatic chitosan derivatives with various chain lengths may cause a change in the molecular configuration of water that being trapped in the matrix. In order to better understand the physical state of water molecules that being trapped in the ACC network, DSC analysis was conducted. Accordingly, the configuration of the water molecules, corresponding to the resulting DSC spectrum; can be classified into three different physical states [130]: (1) free water which can freeze at the usual freezing point (corresponding to an endothermic peak close to 4.8 oC), (2) freezable bound water which freezes at a temperature lower than the usually freezing point (corresponding to an endothermic peak much lower than 4.8 oC) and (3) non-freezable bound water which cannot freeze at the usual freezing point (which is hardly detected by the DSC).
In the initial swelling process, water molecules first disrupt the intermolecular hydrogen bonds and then bind to the hydrophilic sites, such as carboxymethyl groups in this case. These water molecules, which are isolated and uniformly distributed throughout the polymer, have greatly restricted mobility and are referred to as non-freezable bound water. Above a certain level of the bound water, the additional water is preferentially oriented around the bound water and the polymer network structure as a secondary or tertiary hydration shell, which is in a form generally called “clusters”. These cage-like structures result from the tendency of water molecules to form the maximum amount of hydrogen bonds among them in the available space. This type of water is called freezable bound water. As the water uptake further increases,
the splitting of the melting peak will becomes more apparent in the DSC curves, suggesting the existence of two states of freezing water, i.e., freezable bound water and free water, in the polymers. Figure 8-3 shows the DSC spectra of the samples with different chain length of acyl groups measured at water content of about 200%.
-10 -5 0 5 10
CC
C2-0.5 I
Temperature (
oC)
C6-0.5
I I
II II
II
Endot her m ic
C12-0.5
Figure 8-3 DSC curves of CC and ACC derivates (Cn-0.5) measured at water content=200%. Dash lines represent the curve fitting by Lorentzian curve-fitting procedure.
107
A heavily overlapped band (solid line) is observed and can be clearly differentiated as an overlap of several peaks (depicted as dash lines) using the Lorentzian curve-fitting procedure. For the CC sample, the endothermic peak of free water was not observed under fully swollen state. Instead, the endothermic peaks at 1.3 oC and below assigned to freezable bound water were detected. However, after acyl modification, the samples show different DSC spectra, suggesting the state of water in the ACC network was
108
significantly different from that in the CC. It can be found that the peak Ι which is close to 4.8 oC suggesting that the ACC contained free water when the water content > 200%. On the other hand, the peak II was assigned to freezable bound water. With increase of the side chain length from Cn = 2 to 12, peaks I and II shifted toward lower temperature and the peak intensity was decreased, which may be caused by the decrease of the intermolecular volume for free water and freezable bound water. This implies that the chain length of acyl group plays an important role in affecting the amount of free water and freezable bound water within the network structure.
Hence, in order to understand the interaction between water molecules and hydrophobic side chain of the ACC, DSC is also used to quantitatively determine the amount of water, i.e., both freezable and non-freezable bound water in the ACC matrix. Figure 8-4 shows the Wnf,max in ACC with different hydrophobic side chain lengths (carbon number, Cn). It was found that the CC displays the lowest Wnf, max of 10.3% than other samples. This is due to the fact that the samples with high carboxymethyl substitution favored the formation of intermolecular hydrogen bonding (polymer-polymer interaction), resulting in a structure with fewer tight water-binding sites (water-polymer interactions) [24].
However, with the increase of the hydrophobic side chain length, the Wnf,max of ACC is proportionally increased with the Cn until Cn = 12. This Cn-dependent increment can be attributed to a so-called “hydrophobic effect” under which water becomes more structured and less mobile in the vicinity of the hydrophobic group [131]. It has been known from surface-thermodynamic analyses that the attractive interactions between apolar (hydrophobic) molecules immersed in water are driven by the strong hydrogen-bonding free energy of cohesion between the surrounding water molecules [132]. Hence,
increasing Cn of side chain will cause an increase of the hydrophobizing capacity of water since it tends to increase the tendency of
increasing Cn of side chain will cause an increase of the hydrophobizing capacity of water since it tends to increase the tendency of