Optimal formulation of CPMSD based on DOE
Taking into consideration the fact that Box-Behnken design (BBD) method has been widely applied in DOE due to the high efficiency with small numbers of tests,30 a 3-factor-3-level BBD was performed herein for optimizing the formulation of CPMSD. As shown in Table 1, the independent variables were some crucial factors concerned with complex polymeric micelles fabrication such as the types of polymer (X1), mass ratio of DOX to SAA (X2) and the feed ratio of carrier to drugs (X3), for which the corresponding levels were set according to preliminary one-factor tests for screening. More to the point, three common used amphiphilic block copolymers were applied for micellar formula optimization, namely mPEG-PLA, mPEG-PCL, and mPEG-PCL-Phe (Boc), wherein the hydrophilic segment was mPEG with an average molecular weight of about 2,000, and the hydrophobic segment was selected from polylactide (PLA), polycaprolactone (PCL), or that capped with N-t-butoxycarbonyl-phenylalanine (Boc-Phe). All these copolymers were synthesized and characterized in our lab, and the drug entrapment properties had been well demonstrated in previous studies.21, 29, 30 Resultantly, the design matrix containing a total of 17 experimental runs were obtained from BBD as shown in Table 2, which contained 12 factorial points at the midpoint of edge for each process space, and 5 replicates at the center point for estimation of pure error sum of squares. Each experimental run then could be performed in light of the design matrix in random order to avoid bias.
Table 1 Independent variables and levels involved in Box-Behnken design
Independent variable
|
Level
|
Low (-1)
|
Medium (0)
|
High (+1)
|
X1 Types of polymer
|
mPEG-PLA
|
mPEG-PCL
|
mPEG-PCL-Phe(Boc)
|
X2 Mass ratio of DOX/SAA
|
1:2
|
1:4
|
1:6
|
X3 Feed ratio
|
4:1
|
5:1
|
8:1
|
Table 2 Box-Behnken experimental design and the observed responses
Run
|
Independent variable
|
|
Dependent variable
|
X1
|
X2
|
X3
|
|
Y1 (nm)
|
Y2 (%)
|
Y3 (%)
|
1
|
0
|
0
|
0
|
|
17.11 ± 1.30
|
13.3 ± 0.4
|
85.6 ± 4.8
|
2
|
0
|
-1
|
-1
|
|
18.78 ± 0.75
|
14.5 ±1.0
|
75.0 ± 2.1
|
3
|
0
|
+1
|
+1
|
|
17.65 ± 0.97
|
4.5 ± 0.6
|
43.0 ± 2.0
|
4
|
0
|
0
|
0
|
|
16.99 ± 1.12
|
13.4 ± 1.1
|
92.9 ± 3.3
|
5
|
0
|
0
|
0
|
|
16.98 ± 0.90
|
13.9 ± 1.5
|
100.4 ± 4.2
|
6
|
0
|
-1
|
+1
|
|
17.66 ± 2.11
|
8.0 ± 0.7
|
80.6 ± 5.2
|
7
|
0
|
0
|
0
|
|
16.82 ± 2.01
|
15.1 ± 1.0
|
101.6 ± 15.0
|
8
|
+1
|
+1
|
0
|
|
17.16 ± 1.01
|
16.5± 1.2
|
101.8 ± 2.8
|
9
|
-1
|
-1
|
0
|
|
224.5 ± 22.34
|
11.7 ± 0.7
|
87.8 ± 4.6
|
10
|
-1
|
0
|
+1
|
|
15.03 ± 1.67
|
9.9 ± 0.6
|
78.9 ± 6.7
|
11
|
0
|
+1
|
-1
|
|
16.77 ± 0.52
|
19.8 ± 1.5
|
96.1 ± 8.7
|
12
|
+1
|
0
|
-1
|
|
18.44 ± 0.73
|
16.6 ± 1.7
|
87.7 ± 5.3
|
13
|
+1
|
-1
|
0
|
|
266.9 ± 24.21
|
11.7 ± 0.7
|
72.7 ± 6.5
|
14
|
+1
|
0
|
+1
|
|
18.23 ± 1.31
|
10.6 ± 0.5
|
93.9 ± 8.0
|
15
|
0
|
0
|
0
|
|
17.46 ± 0.63
|
15.7 ± 1.2
|
91.6 ± 11.3
|
16
|
-1
|
+1
|
0
|
|
16.06 ± 0.37
|
15.0 ± 1.4
|
84.9 ± 7.8
|
17
|
-1
|
0
|
-1
|
|
115.5 ± 9.78
|
14.6 ± 0.9
|
61.7 ± 9.1
|
It is well known that the dimensional characteristics of nanomicelles may contribute to passive targeting to tumor through the enhanced permeability and retention (EPR) effect, since it preferred to avoid macrophages uptake and clearance by mononuclear phagocyte system (MPS) or the reticuloendothelial system (RES), and achieved long circulation and fairly high chance of reaching tumor site. Furthermore, the nanomicelles with high DLC and EE but small particle size are usually expected to avoid immune clearance and maintain desired systematic bioavailability.34 Herein, several key performance indices (KPI) of micellar system thus were used as the dependent variables for statistical analysis and evaluation of CPMSD formulation, namely particle size (Y1), drug loading capacity (DLC, Y2), and entrapment efficiency (EE, Y3). The observed responses of KPI (Table 2) clearly revealed that there was significant differentiation among these experimental runs. Specifically, the particle size ranged from 15 nm to 267 nm, while DLC and EE changed between 4.5% and 19.8%, and 43.0% and 101.8%, respectively. The influence of all these independent variables (X1, X2, X3) on KPI were further investigated by using multiple regression analysis. The cubic polynomial regression model for each KPI could be obtained with determination coefficient (R2) more than 0.9 and the p-value less than 0.05, and the 3-D response surface plots illustrated overall influence of the independent variables (Fig. 2 A-C). Among all the three formulation factors involved in the present study, X3, namely the feed ratio of copolymer carrier to both drugs, was found to be the most important one for fabricating CPMSD micelles, especially had a great impact on DLC and EE. According to these results, all the three factors X1, X2 and X3 thus were determined at the medium level for fabricating CPMSD micelles. That is to say, the multifunctional nanomicellles of CPMSD could be achieved by using the amphiphilic block copolymer of mPEG-PCL as a carrier for encapsulating both drugs under a carrier-drug mass ratio of 5:1 and DOX-SAA mass ratio of 1:4.
On the basis of optimal formulation obtained from BBD study, several batches of CPMSD nanomicelles were prepared by conventional thin film hydration technique, then characterized for verification.21 DLS assay clearly demonstrated that the CPMSD micelles were usually the uniform and small particles with mean particle size ranging within 15 to 25 nm and the PDI values less than 0.2. Meanwhile, HPLC quantitation indicated that the DLC values for DOX and SAA together were (15.7 ± 0.8) %, and EE values were generally more than 95%. All these KPI values determined were in close agreement with the predicted values of optimal micellar system of CPMSD from DOE model that showed particle size of 17.1 nm, DLC of 14.3%, and EE of 94.4%, suggesting that the DOE-based optimal formulation would be reliable for fabricating the aimed dual drug-loaded nanomicelles of CPMSD for further evaluation.
Molecular mechanism of drugs entrapment into micelles
The possible mechanism of amphiphilic block copolymer encapsulating both drugs to form novel nanomicelles of CPMSD was further investigated in the present study. The approaches of molecular mechanics (MM) and molecular dynamics (MD) were applied to examine molecular interactions among the polymeric carrier (mPEG-PCL) and small-molecule drugs (DOX and SAA) based on simulating their theoretical structures.31, 32 As illustrated in Fig. 3 (a-d), the copolymer initially displayed a curvilinear conformation, then gradually bended and changed with the heating process, finally formed into a spherical shape after 100ps MD simulation, which consisted of hydrophilic and hydrophobic parts and provided suitable binding sites for small molecules. Meanwhile, both small-molecule drugs constantly adjusted the conformation and distance from copolymer to obtain favorable interaction modes (Fig. 3 e-h). Resultantly the copolymer encapsulated SAA into the hydrophobic cavity and DOX to the hydrophilic surface, respectively (Fig. 3 i).
Chemically DOX is a kind of antibiotics with the amino sugar linked to anthracycline via a glycosidic bond (Fig. 1), which contributes to its hydrophilicity and the preference for binding to the hydrophilic surface of the copolymer mPEG-PCL. In contrast, SAA is a kind of salvianolic acid with fairly high hydrophobicity, thus could tight bind with the hydrophobic cavity of the copolymer. More to the point, the phenolic hydroxyl and carboxyl groups in SAA (Fig. 1) would significantly enhance the molecular interaction with the other drug molecule DOX via its basic amino sugar. From this point of view, SAA plays an important role as a bridge between DOX and the copolymer, which greatly promotes copolymer-drug interaction, and leads to significant increase in drug encapsulation efficiency of the complex micellar system CPMSD, especially for the relatively hydrophilic drug molecule DOX. The findings from MD simulation clearly demonstrated that CPMSD could act as a new and efficient dual drug-loaded micellar DDS by a unique mechanism involved in drug entrapment, which might also result in specific drug release profiles of CPMSD.
Colloidal properties and stability of CPMSD
Polymeric nanomicelles are a kind of self-assembled colloidal particles for drug delivery, and the colloidal stability has a great impact on pharmaceutical performance such as in vitro and in vivo drug release behaviors.35 Therefore the colloidal properties and stability of CPMSD were investigated through observation of appearance and particle size. As illustrated in Fig. 4A, the nanomicelles of CPMSD fabricated under the optimal formulation presented as a stable colloidal suspension clearly displaying a Tyndall effect, and lyophilization could provide a yellowish red powder with good disparity without collapse or atrophy, as well as good re-dispersibility in saline, PBS or double distilled water. The TEM image for morphology further revealed that CPMSD micelles were uniform nanoparticles with spherical or nearly spherical shape and a typical particle size of about 20 nm (Fig. 4B).
In order to know the colloidal stability of CPMSD micelles, how the particle size changed with temperature was investigated at first. Six batches of fresh prepared micelles (2 ml/branch) were divided into two groups (n=3) and placed at 4 °C and room temperature (25°C), respectively, then the particle size was monitored periodically. Although a significant increase in particle size was observed after storage at 4 °C for 12 hrs, the fresh prepared CPMSD could maintain a stable micellar particle size for at least 24 hrs at room temperature (Fig. 4C), suggesting a reliable condition to produce lyophillized powder of CPMSD for storage.
The particle size change in simulated serum was further inspected to evaluate the colloidal stability of CPMSD in systemic circulation. The micelles were incubated with fetal bovine serum (FBS, pH 7.4) at 37 °C under gentle stirring, and an aliquot of the sample was withdrawn for the measurement according to the time schedule. As shown in Fig. 4D, there was no significant change with time in micellar particle size within 24 hrs incubation in the simulated body fluid containing 1% or 10% FBS (p > 0.05), indicating a good colloidal stability of CPMSD in vivo. Meanwhile, the CPMSD micellar particles in PBS displayed a slightly negative zeta potential, i.e. – (1.77±0.49) mV, and this value became more negative with increasing FBS content, namely – (3.20±0.53) mV and – (3.62±0.27) mV in the incubation system containing 1% FBS, and 10% FBS, respectively. These findings together demonstrated the coating effect of endogenous proteins such as serum albumin might be chiefly responsible for in vivo colloidal stability of CPMSD.
In vitro drug release from CPMSD
The in vitro drug release characteristics of CPMSD were investigated via dialysis method, and PBS solution containing 1% Tween 80 (pH 7.4) was used as the release medium to maintain the chemical stability and sink condition for both DOX and SAA. Resultantly, there was significant difference in the in vitro release profile between DOX and SAA (Fig. 5). Although the encapsulated drugs were both sustained released from CPMSD, DOX had a much higher release rate at early stage than SAA did, and an initial release burst with a percent cumulative release of 45% within 0.5 h could be observed. The drug release of DOX increased steadily, and the cumulative drug release amount reached a peak of about 90% within 3 hrs. In contrast, the nanomicelles of CPMSD constantly released SAA at a rather slow rate, which had a percent cumulative release up to 70% within 72 hrs.
Furthermore, the kinetic mechanism of drug release from CPMSD was investigated by fitting several common kinetic models to the cumulative release profiles, including the zero-order, first-order and Higuchi models. As shown in Table 3, the first-order model was determined as the optimal one with the highest goodness of fit, and the values of determination coefficient (R2) reached up to 0.97 for both drugs. Thus, it could be concluded that non-constant diffusion was the chief mechanism involved in drug release from CPMSD micelles for both DOX and SAA, while matrix swelling and dissolution could be negligible in the present dual drug-loaded polymeric micellar system. All these findings clearly demonstrated the special drug releasing characteristics of CPMSD, and also provided a strong evidence for the above-mentioned distinctive mechanism of drug entrapment based on MD studies, which thought that CPMSD micelles would be prone to a relatively fast drug release for DOX from the hydrophilic surface, but an extended release for SAA from the inner core.
Table 3 Model fitting for in vitro drug release from nanomicelles of CPMSD micelles
Model
|
DOX*
|
SAA*
|
Zero-order equation
|
Y=0.78t+71.07 (0.2103)
|
Y=0.86t+21.50 (0.7811)
|
First-order equation
|
Y=87.66[1-exp(-1.26t)] (0.9819)
|
Y=65.18[1-exp(-0.10t)] (0.9765)
|
Higuchi equation
|
Y=6.36t1/2+61.67 (0.4924)
|
Y=8.55t1/2+7.05 (0.9402)
|
Notes: * The data in parenthesis were the determination coefficient for model fitting.
In vitro cytotoxicity and cellular uptake
The in vitro cytotoxicity activity of CPMSD against human breast cancer MCF-7 cells was evaluated by using CCK-8 assay of cell viability. Several relative preparations were used for comparison, including free solution of DOX, SAA and the cocktail formulation of DOX and SAA. Free DOX alone could significantly inhibit in vitro proliferation of MCF-7 cells in a concentration-dependent manner with the final concentration ranging within 0.01 ~ 5 μM (Fig. 6A), and the half inhibitory concentration (IC50) was determined as 0.27 μM, which was in good consistence with the data previously reported and demonstrated the high potency of DOX as a chemotherapy agent for human breast cancer.36
Fig. 6B clearly illustrated the difference in cytotoxicity against MCF-7 cells among various treatments involved. More to the point, free DOX alone at a final concentration of 230 ng/ml close to the IC50 value caused an inhibition rate of ~ 60% against in vitro growth of MCF-7 cells. The treatment with CPMSD preparation or DOX-SAA cocktail displayed a similar inhibition rate to free DOX at the same concentration, and no significance difference was observed among the three groups (p > 0.05). In contrast, the treatment with free SAA alone only yielded ~ 40% inhibition at a final concentration of 930 ng/ml, which was set four times as much as DOX according to their mass ratio in CPMSD (1:4). Furthermore, the inhibition rate of SAA treatment was found to be quite lower than that of free DOX alone or DOX-SAA combination via the CPMSD preparation (p < 0.05). These findings demonstrated that neither copolymer entrapment nor the combination with SAA would have significant effect on the inhibitory potency of DOX against MCF-7 cells growth, and DOX thus could be regarded as the major active pharmaceutical ingredient responsible for anticancer efficacy of the present CPMSD nano-formulation used for cancer chemotherapy.
DOX is an anthracycline drug widely used in breast cancer chemotherapy. So far, the best-known and widely accepted mechanisms are based on the inhibition of DNA replication, transcription and repair processes, which is mediated by the drug intercalation into DNA and occur in the nucleus.37 The final target location of DOX is the nucleus, which is usually regarded as the main target responsible for the anticancer potency of DOX.38 The cellular uptake of DOX and the effect of CPMSD formulation thus were investigated following the evaluation of anticancer potency in vitro. Cell nuclei were stained by DAPI, and CLSM was employed for observation through detecting its blue fluorescence and the obvious specific red fluorescence of DOX. Continuous increase in the fluorescence intensity of DOX could be clearly observed with extension of the incubation time. After a 2-h incubation, all the three groups treated with DOX at an equivalent concentration displayed similarly strong red fluorescence when compared with the control group without any drug treatment. According to the Merge column, the regions with red fluorescence of DOX were well overlapped with those with blue fluorescence of DAPI and there was no significant difference among these DOX containing treatments (Fig. 6C), indicating a rapid uptake of DOX almost entirely into the nuclei of MCF-7 cells, no matter what the preparation of DOX was. The results from in vitro evaluation together revealed that the CPMSD preparation might be an efficient nano-formulation of DOX with full maintenance of the anticancer potency and the final target of this chemotherapy drug.
In vivo anticancer efficacy
Female BALB/c nude mice bearing human breast cancer MCF-7 cells were employed to evaluate in vivo therapeutic responses of the CPMSD preparation such as anticancer efficacy, as well as the protective effect against DOX-induced cardiotoxicity. All the model mice were randomly divided into several groups (n=8) administering different formulations of DOX under the same regime, and the anti-tumor therapeutic efficacy was assessed through characterizing the tumor with time after administration. More specifically, the aim drug DOX, which was delivered by the free drug alone, CPMSD or the cocktail formulation both with a DOX/SAA mass ratio of 1:4, was intravenously administered at a single DOX dose of 2.5 mg/kg for five times every each other day.21 The mice only given the same volume of vehicle (saline) was used as the negative control (NC) for comparison.
In contrast to the mice in NC group, the animals administered DOX all had drastic inhibition of tumor growth, no matter what the drug formulation was. As illustrated in Fig. 7A, tumor grew gradually in the NC group and the relative tumor volume reached 150% after the last injection of saline, whereas the other three groups (DOX, Cocktail and CPMSD) similarly showed an opposite temporal profile and all displayed a significant difference from the NC group in the relative tumor volume (p < 0.01). After the accomplishment of all the five DOX dosages, the relative tumor volume was observed ranging from less than 40% to about 50%, which varied with the formulations of DOX. Moreover, significant difference among these DOX formulations could be found after the second dosage, and some tumors were even completely eradicated through treatment by the CPMSD micelles. Since it would take time for the micelles to penetrate into tumors, concentrate at the tumor site and release drug,36 these findings thus indicated that CPMSD might be a kind of sustained release preparation of DOX with tumor-targeting efficiency, as well as the most potent DOX formulation against breast cancer growth in mice.
In order to confirm the observation of tumor volume changes, all the mice were sacrificed one day after the last dosage and tumors were resected and weighed. As shown in Fig. 7B, the tumor weight averaged 70.1±11.5 mg for the mice in NC group, 25.9±5.6 mg for DOX group, 30.5±7.2 mg for Cocktail group, and 27.4±8.7 mg for CPMSD group, respectively. The mice only treated by saline showed the maximum mean tumor weight, while all the DOX-containing treatments, no matter what the drug formulation was, led to dramatic shrinkage of the tumor weight in comparison with that in the NC group (p < 0.01). Then the tumor growth inhibition (TGI) was calculated to quantify treatment effects. The TGI value for antitumor activity rating reached up to 63.0% by free DOX alone, 56.5% by the cocktail formulation of DOX, and 60.9% by CPMSD preparation, respectively, and there was no significant difference among these DOX formulations (p > 0.05).
Meanwhile, the body weight was monitored for each animal to investigate the potential difference in systemic toxicity among various treatments. Although the mice in each group had a similar mean body weight till the second dosage, significant difference could be observed among the three DOX formulations involved after finishing all the five dosages (p < 0.01), and the final body weight of the mice in CPMSD group was the closest to that in NC group (Fig. 7C), suggesting excellent biosafety of this DOX formulation. Therefore, it was concluded that co-treatment with SAA by CPMSD micelles or the simple cocktail would not attenuate in vivo antitumor potency of DOX, and its systemic toxicity could be greatly alleviated by the CPMSD formulation. Indeed, distinctive differences could be observed in bio-distribution characteristics (Fig. 8) between the dual drug-loaded polymeric micellar system of CPMSD and the simple cocktail formulation of free DOX and SAA, which revealed the site-specific drug releasing properties of CPMSD and therefore the possible mechanisms responsible for its high antitumor potency and low toxicity in vivo.
In vivo protective effects against cardiotoxicity
The clinical use of DOX is limited by severe cardiotoxic side effects, although it is a potent anticancer drug. Oxidative stress is generally recognized as one of the main mechanisms responsible for DOX induced cardiotoxicity and much research has recently been devoted to this challenge, mainly including some doxorubicin-antioxidant co-drugs,39 and dual drug-loaded nano-platform for targeted cancer therapy toward clinical therapeutic efficacy of multifunctionality.40 Our present study aimed at the CPMSD formulation for breast cancer chemotherapy, which was a high-performance multifunctional polymeric micellar delivery system co-loading the anticancer drug DOX and the highly potent natural antioxidant SAA. Following evaluation of antitumor efficacy, the detoxication effect of SAA delivered by CPMSD was further investigated in tumor-bearing nude mice through assay of several principal physiological and biochemical markers related to myocardial function, as well as inspection of myocardial histologic changes concerned with cardiotoxicity.
Along with monitoring body weight changes, the heart weight of each animal was measured for calculating the cardiac weight index (CWI) by its ratio to body weight. As shown in Fig. 9A, the NC group had a maximum CWI value of (0.48±0.01) %, while the minimum value was found as (0.34±0.03) % for the mice administered free DOX alone. Significant difference in CWI could be observed between the NC group and any of the other two groups treated with DOX-containing formulation but CPMSD at an equivalent dosage (p < 0.01). There was no significant difference between the DOX and cocktail group, even though it was co-administered with the antioxidant agent SAA through the cocktail formulation. However, the CWI value for the mice administered CPMSD micelles was found as (0.46±0.02) %, which was significantly higher than the other two DOX-containing groups (p < 0.01), and even close to that in the NC group (p > 0.05). Taking into consideration of the changes in both CWI and body weight (Fig. 7C), these results thus indicated the differentiated effect of co-treatment formulation on DOX induced cardiotoxicity and the effectiveness of CPMDC that could nearly bring the two major physiological indexes back to normal.
In order to quantitatively evaluate the protective efficacy of CPMSD against cardiotoxic side effects caused by DOX dosing, biochemical analysis was further performed on the crucial markers involved in cardiac toxicity, including SOD and MDA in heart tissues, and LDH, CK and cTnT in plasma.5, 21 When compared with the NC group, the mice administered free DOX alone showed significantly decreased SOD level along with elevated MDA content (p < 0.01), suggesting the DOX-induced oxidative injury and remarkable amelioration of co-administration of SAA through the CPMDC formulation (Fig. 9B). Meanwhile, significant increases in plasma level of LDH, CK and cTnT could correspondingly be observed in the mice administered free DOX alone or the simple cocktail of DOX and SAA (Fig. 9C), which confirmed the occurrence of cardiac injury in both DOX and cocktail group. Comparison among these DOX-containing formulations further revealed the reducing effect on all these markers of co-treatment with SAA by CPMDC, but not the cocktail of free DOX and SAA (p < 0.01). Due to the effective protection against DOX induced cardiotoxicity, it therefore was concluded that the tumor-bearing mice would greatly benefit from the present CPMSD formulation of DOX and SAA into dual drug-loaded polymeric micelles.
Finally, the histopathological examination of cardiac tissue specimens was performed to manifest DOX-induced toxic injuries on the main target organ. As illustrated in Fig. 9D, the tumor-bearing nude mice in NC group showed basically normal morphology of cardiac myocyte in the left ventricle, while the DOX-containing treatments could affect cardiomyocytes in different ways and to different degrees depending on the formulation, among which the severest myocardial damage was observed in the mice only treated with free DOX. Accompanied by infiltration of inflammatory cells, the DOX alone group exhibited obvious myocardial injuries such as cross-striations, myocardial endochylema puffing and sarcoplasmic matrix partly resorbed, as well as myocardial fiber disarrangement, cellular swelling and degeneration, hinting toward toxin-mediated necrosis of cardiomyocytes. More to the point, these DOX-induced cardiomyocytes injuries could be alleviated by co-administering SAA, especially through the CPMSD micelles, which displayed less histopathological changes than the cocktail formulation at an equivalent dosage. These findings altogether demonstrated the high potency of SAA against cardiotoxic effect of DOX by co-administering both drugs through the CPMSD formulation.