Microsoft Word - 6. 2016-0038_new_-RZFeb1-F.docdoi:
10.3724/SP.J.1009.2017.00142
Chinese Journal of Natural Medicines
Pharmacokinetics and correlation between in vitro release and in
vivo absorption of bio-adhesive pellets of panax notoginseng
saponins
Institute of Medicinal Plant Development, Chinese Academy of
Medical Sciences & Peking Union Medical College, Beijing
100193, China
Available online 20 Feb. 2017
[ABSTRACT] The present study was designed to prepare and compare
bio-adhesive pellets of panax notoginseng saponins (PNS)
with hydroxy propyl methyl cellulose (HPMC), chitosan, and chitosan
: carbomer, explore the influence of different bio-adhesive
materials on pharmacokinetics behaviors of PNSbio-adhesive pellets,
and evaluate the correlation between in vivo absorption and in
vitro release (IVIVC). In order to predict the in vivo
concentration-time profile by the in vitro release data of
bio-adhesive pellets, the release experiment was performed using
the rotating basket method in pH 6.8 phosphate buffer. The PNS
concentrations in rat plasma were analyzed by HPLC-MS-MS method and
the relative bioavailability and other pharmacokinetic parameters
were estimated using Kinetica4.4 pharmacokinetic software.
Numerical deconvolution method was used to evaluate IVIVC. Our
results indicated that, compared with ordinary pellets, PNS
bio-adhesive pellets showed increased oral bioavailability by 1.45
to 3.20 times, increased Cmax, and extended MRT. What’s more, the
release behavior of drug in HPMC pellets was shown to follow a
Fickian diffusion mechanism, a synergetic function of diffusion and
skeleton corrosion. The in vitro release and the in vivo biological
activity had a good correlation, demonstrating that the PNS
bio-adhesive pellets had a better sustained release. Numerical
deconvolution technique showed the advantage in evaluation of IVIVC
for self-designed bio-adhesive pellets with HPMC. In conclusion,
the in vitro release data of bio-adhesive pellets with HPMC can
predict its concentration-time profile in vivo.
[KEY WORDS] Panax notoginseng saponins; Bio-adhesive pellets;
Pharmacokinetics; In vivo and in vitro correlation
[CLC Number] R969.1 [Document code] A [Article ID]
2095-6975(2017)02-0142-10
Introduction
Panax notoginseng saponins (PNS) are the effective active
substances of root of notoginseng [Panax notoginseng (Burk.) F.H.C
hen], a kind of perennial herbaceous plants affiliated to
Araliacede, with the main active ingredients being notoginseng
saponin and ginseng saponin. Studies have reported that PNS have
various activities such as effects against cerebrovascular
ischemia, anti-arrhythmic effects, diastolic relaxing blood
vessels, improving blood rheology and microcirculation, inhibiting
platelet aggregation and
[Received on] 17- Feb.-2016 [Research funding] This work was
supported by the National
Natural Science Foundation of China (No. 81274094).
[*Corresponding author] Tel: 86-10-57833276, Fax: 86-10- 57833276,
E-mail:
[email protected] These authors have no conflict of
interest to declare.
thrombosis, and reducing hematic fat and resistance to
atherosclerosis [1-5]. PNS have good water solubility, and the
solubility and dissolution rate are not the main factors affecting
drug absorption [6-7]. However, poor stability under stomach
condition, low membrane permeability, and high molecular weight are
the main factors resulting in poor bioavailability.
Bio-adhesive preparation as a new drug delivery system has become
more and more popularly in recent years; it could extend the time
of pharmaceutical preparations’ effects on target sites, increase
the contact with the absorption membrane, change membrane fluidity,
and increase drug penetration to the intestinal epithelial cells,
thus promoting the absorption of drugs, and improving drug oral
bioavailability [8-9]. Therefore, studying bio-adhesive
preparations is highly significant.
Several researchers have prepared the controlled-release
formulations using enteric technology [10-11], micro-porous osmotic
pump tablets [12] or pulsatile controlled-release tablets
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 143 –
[13], in order to improve the oral bioavailability by avoiding PNS
degradation in gastric acid. Our team has prepared bio-adhesive
tablets to improve the oral bioavailability [14-16] by increasing
the drug intestinal absorption time via adhesion to the
gastrointestinal tract. What’s more, the bio-adhesive formulations
have sustained-release effect to a certain extent to avoid PNS
degradation in gastric fluid to a certain extent. The present study
was designed to prepare skeleton-type bio-adhesive pellets using
bio-adhesive materials that could promote intestinal absorption of
PNS. PNS bio-adhesive pellets had larger superficial area than
bio-adhesive tablets, increasing the adhesion area, prolonging the
contact time of drugs with mucous membrane, promoting the
absorption of drug, and effectively improving drug bioavailability
[17].
In the present study, the in vitro release behaviors of the
bio-adhesive pellets were evaluated, and the drug concentrations in
blood were analyzed following adminis- tration of different PNS
bio-adhesive pellets with different bio-adhesive materials. The
pharmacokinetic parameters were estimated, so that the effects of
different bio-adhesive materials on in vivo absorption of PNS could
be determined. Additionally, the correlation between in vitro
release and in vivo absorption was analyzed in order to provide
reference for future researches on dosage forms and clinical
applications.
Materials and Methods
Animals Male Sprague–Dawley (SD) rats weighing 250–300 g
were obtained from Vital River Laboratories, Beijing, China. The
animals were housed at temperature of (25 ± 1) °C and relative
humidity of 45%–55% under 12 : 12 light : dark cycle. Before study,
the rats were allowed to acclimate to the environment for 7 days.
All studies in mice were performed in accordance with guidelines
approved by the Ethics Committee of the Chinese Academy of Medical
Sciences (CAMS) and Peking Union Medical College (PUMC). Drugs and
chemicals
PNS extract (PNS extract contained 2.30% of notoginsenoside R1,
15.02% of ginsenoside Rg1, and 26.80% of Rb1, respectively, the
batch number is 201304) was purchased from Wenshan Kangzhou
bio-technique Co. Ltd. (Wenshan, Yunnan, China). Standards of
Notoginsenoside R1 (NGR1), Ginsenoside Rg1 (GRg1), and Ginsenoside
Rb1 (GRb1) were purchased from the National Institute of the
Control of Pharmaceutical and Biological Products (Beijing, China).
In addition, Digoxin (internal standard) was purchased from the
National Institute of the Control of Pharmaceutical and Biological
Products (Beijing, China). Carbomer was purchased from BF Goodrich
(Cleveland, USA). Chitosan was purchased from Jinqiao Biochemistry
Company (Taizhou, Zhejiang, China). HPMC (K4M) was purchased from
Colorcon Company (Shanghai, China). Microcrystalline cellulose
(MCC) (PH101) was purchased from BF Goodrich. Chitosan was
purchased from Asahi Kasei Corporation (Tokyo, Japan). HPLC-grade
acetonitrile and
methanol were purchased from Fisher Scientific (New Jersey, USA).
Heparin sodium was purchased from Beijing Yaobei Biological and
Chemical Reagents Company (Beijing, China). Preparation of normal
and bio-adhesive pellets of PNS
We prepared normal and bio-adhesive PNS pellets, according to the
methods of our previous research [18], using three types of
bio-adhesive materials: chitosan, HPMC, and chitosan: carbomer. We
homogeneously mixed PNS with MCC and other bio-adhesive materials
after sifting and used 30% NaCl to get soft materials with modest
viscoelasticity and plasticity. The soft materials were passed
through screw extruder to get the same diameter, irregular length,
smooth, and dense strings, at the rotate speed of 35 r·min−1; the
strings were passed through spheronizator, and the extrudates were
rolled into balls under the effect of friction force and
centrifugal force. The pellets were dried at 60 °C. Normal pellets
were prepared by the similar ratio and method, but without
bio-adhesive materials. Release testing of PNS pellets Quantitative
analysis by high-performance liquid chromatography (HPLC)
A Shimadzu (Japan) Class VP HPLC system with a
Kromasil C18 (10 mm × 4.6 mm2.4 μm) (Akzo Nobel,
Goteborg, Sweden), and UV detector was used for drug content
analysis. Mobile phase was composed of aqueous (A) and water (B),
and the gradient elution program was as follows: 0−10 min: 20%−40%
A; and 10−25 min: 40%−20% A. The flow rate was set at 1 mL·min−1,
the measurement wavelength was set at 203 nm, the column
temperature was set at 30 oC, and injection volume was 20 µL. The
regression equation and correlation coefficients of standard curves
were as follows: NGR1: A = 2 649.2C − 241.7 (r = 0.999 7); GRg1: A
= 13 686.3C − 2 043.2 (r = 0.999 9); GRb1: A = 2 087.9C −538.4 (r =
0.999 9). A represents peak area, C represents drug concentration.
The linearity ranges was 0.2−50 μg·mL−1. In vitro release
testing
In vitro release testing was carried out in release medium (pH 7.4
phosphate- buffered saline (PBS) solution) at (37 ± 0.5) oC. 0.5 g
of pellets were suspended in the rotative baskets with 500 mL of
release medium. At pre-determined time intervals, aliquots of 2-mL
solutions were withdrawn and filtered through 0.45-μm filters. The
sample volumes were replaced with equal volume of the fresh medium.
NGR1, GRg1, and GRb1 released from pellets were quantified by HPLC,
three tests were performed for each sample and the mean values were
used as the final results. Analysis of elease kinetics
The release data of optimized formulation were fitted to different
mathematical models to reveal the release mechanism from the
pellets: Zero order (% cumulative drug release vs time), first
order (log% drug release vs time), Higuchi model (% cumulative drug
release vs square root of
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 144 –
time), and Peppas exponential equation (log% drug release vs log
time). All the curve fitting, simulation, and plotting were
performed using commercially available Microsoft excel solver, and
regression coefficient (r2) values were calculated. In vivo
pharmacokinetic study Instrumentation and analytical
conditions
Method validation proved that the current assay can be successfully
applied to the determination of NGR1, GRg1, and GRb1 after a single
oral administration of 90 mg·kg−1 of PNS in three different kinds
bio-adhesive pellets of PNS. The analytical column C18: Thermo C18
(10 mm × 4.6 mm; 2.4 μm) was used at 25 oC, with water containing
10 mmol·L−1 of ammonium acetate, 10% methol and 5% N-butylamine (A)
and methanol (B) gradient as mobile phase at a flow rate of 0.5
mL·min−1. The elution was carried out as follows: 50%−95% B at
0−2.0 min; 95% B at 2.0−4.0 min; 95%−50% B at 4.0−6.0 min; and 50%
B at 6.0−8.0 min. The injection volume was 20 μL and the partial
loop with needle overfill mode was used for sample injection.
The mass spectrometer was operated in positive ionization mode
using MRM to assess the three ginsenosides and Digoxin: m/z
955.7→775.4 for NGR1, m/z 823.6→643.6 for GRg1, m/z 1 132.1→365.7
for GRb1 and m/z 803.5→283.4 for Digoxin (IS), respectively. The
optimized cone voltage and collision energy were 100 V and 22 eV
for NGR1, 90 V and 24 eV for GRg1, 100 V and 30 eV for GRb1, and 50
V and 20 eV for IS, respectively. A spray voltage of 5 500 V was
used, and the capillary temperature was set at 550 ºC. The scanning
range was selected as m/z 100→1 200: no interference was observed
around target compound peaks. Data acquisition and processing were
accomplished on a 3200 Q TRAP® mass spectrometer (Applied
Biosystems, Waltham, Massachusetts, USA). Sample preparation
20 μL of Digoxin (1 000 ng·mL−1 in methanol, IS) was added to 150
μL of plasma. The sample was mixed with 0.5 mL of acetone for 5 min
and then centrifuged at 12 000 r·min−1 for 10 min. The supernatant
was collected, transferred to a clean centrifuge tube, and
evaporated to dryness at 60 ºC. The resultant residue was dissolved
in 150 μL of 80% methanol, vortexed for 3 min, and centrifuged at
12 000 r·min−1 for 5 min. The supernatant was collected, filtered
through 0.22-μm microfiltration membrane, and injected onto
HPLC-MS/MS system for analysis. Method validation Specificity
The specificity of the method was investigated by analyzing blood
samples from healthy rats, spiked blood samples, and blood samples
after intranasal administration of PNS pellets, to exclude any
endogenous co-eluting interference. Linearity and LLOQ
The ginsenoside concentrations were plotted against the ratio of
ginsenoside peak area over that of Digoxin. A weighted (1/[nominal
concentration]) least-square linear
regression (y = bx + a) was used to fit the curves. The least
concentration of the calibration curve was considered LLOQ.
Precision and accuracy
The intra- and inter-day precision and accuracy of the method were
determined by analyzing five replicates at low, medium, and high
concentrations with the same analytical run on three consecutive
days. Stability
The stability of NGR1, GRg1 and GRb1 in blood samples was assessed
at three quality control (QC) levels with five replicates under
different storage and processing conditions, including 4 h storage
at ambient temperature, freeze-thaw for three cycles, to determine
long-term and post-preparative stability. The calibration curves
were freshly generated during the stability assays. Administration
of PNS and sample collection
According to our double cycle and crossover experi- mental design,
normal and bio-adhesive pellets were administered orally at a dose
of 200 mg·kg−1 to six rats each group. All the rats were deprived
of food but given free access to water for 12 h before the
experiments. The blood samples (0.5 mL each time) were collected
via orbit at 0.25, 0.5, 0.75, 1, 1.5, 2, 4, 6, 8, 12, and 24 h
after administration. The plasma was separated after centrifugation
for 10 min at 7 000 r·min−1 and stored at −20 °C until analysis.
Correlation study between in vitro release and in vivo
pharmacokinetic characteristics
Deconvolution method was used to investigate the in vitro release
and in vivo absorption correlation. The correlation coefficient was
calculated with WinNonlin5.2 pharmacokinetic software [19]. The
cumulative release degree F represented in vitro release
characteristics, the AUC at all time points represented the in vivo
absorption, and the ratio of AUC at each point with AUClast
represented the Fraction Input (R) [20].
Results
In vitro drug release The cumulative release of PNS from the
bio-adhesive
pellets prepared with different bio-adhesive materials are shown in
Fig. 1. The drug release rates were more than 80% for ordinary
pellets and bio-adhesive pellets with chitosan within 20 min. The
bio-adhesive pellets with HPMC had a better sustained release
effect. Release kinetics
The bio-adhesive pellets with HPMC and chitosan- carbomer were
simulated with various mathematical models, such as zero order,
first order, Higuchi, and Peppas kinetic models. The best fitted
regression results were obtained with the Peppas model (Table 1).
The release mechanism was then studied with the Peppas equation,
indicatig a coupling of diffusion and erosion mechanism, suggesting
that the drug release was controlled by more than one
process.
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 145 –
Fig. 1 In vitro release profiles of bio-adhesive pellets with
different bio-adhesive materials. A: NGR1, B: GRg1, and C:
GRb1
Pharmacokinetic applications and plasma drug concentration- time
curves of rats Method validation Specificity
In the present study, chromatogram examination ruled out any
endogenous peaks at the retention times of NGR1, GRg1, GRb1, and
Digoxin under optimized analytical conditions (Fig. 2). Linearity
and LLOQ
The regression equation (correlation coefficients) and linearity
ranges of the analytes were as follows: y = 0.090 3x + 3.804 6 (r =
0.997 7) and 7.18−1 435.00 ng·mL−1 for NGR1; y = 0.081 9x + 23.544
5 (r = 0.997 7) and 6.75−1 350 ng·mL−1 for GRg1; y = 0.002 4x +
1.615 2 (r = 0.995 3) and 15.45−1 545.00 ng·mL−1 for GRb1. The LLOQ
values were 7.18 ng·mL–1 for NGR1, 6.75 ng·mL−1 GRg1, and 15.45
ng·mL−1 for GRb1. respectively. The results indicated a good
linearity within the selected range. In addition, LLOQ with a ratio
of signal to noise of more than 10 (S/N > 10) for the three
analytes indicated acceptable assay accuracy and precision in blood
sample anlysis. Stability
Stability values (< 15% of change from the baseline) indicated
that all analytes were stable when stored at ambient temperature
for 4 oC and subjected to three freeze-thaw cycles (−80 oC) (data
not shown).
Table 1 Release kinetics simulated with the Peppas model
Groups NGR1 GRg1 GRb1
HPMC y = 0.476 6x + 2.038 2 y = 0.503 9x + 1.830 7 y = 0.591 9x +
1.368
R² = 0.984 R² = 0.9844 R² = 0.9566
Chitosan : Carbomer Y = 0.593 6x + 0.572 Y = 0.572 5x + 0.678 2 Y =
0.539 6x + 0.721 6
R² = 0.996 3 R² = 0.995 1 R² = 0.984 4
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 146 –
Fig. 2 Ion Chromatorgraphy of blank blood samples of rats (on the
left) and the rat blood samples spiked with corresponding standard
sample (on the right). A: NGR1; B: GRg1; C: Digoxin; D: GRb1).
Noting: The y-axis are peak area ratio against internal
standard
Pharmacokinetics We investigated the pharmacokinetic properties of
PNS in
rats, the concentration-time curves of three components of six rats
are shown in Fig. 3. The AUC values of NGR1, GRg1 and GRb1 with
bio-adhesive pellets were increased remarkably, compared with that
of the normal pellets (P < 0.05), and the relative
bioavailability data after oral administration of bio-adhesive
pellets are shown in Table 2 for NGR1, Table 3 for GRg1, Table 4
for GRb1. These results indicated that bio-adhesive formulation
considerably improved the oral bioavailability of PNS. Compared
with the normal pellets, the parameters of Tmax for NGR1, GRg1 and
GRb1 were significantly increased (P < 0.05) and the parameters
of Cmax for NGR1 and GRb1 were increased remarkably for the
bio-adhesive pellets (P < 0.05), which revealed that there was a
process of accelerating release or improving assimilation.
Correlations between in vitro release and in vivo pharmacokinetic
characteristics
The correlations between in vitro release and in vivo
pharmacokinetic characteristics were evaluated using the
deconvolution method. The Fraction input was calculated with
WinNonlin5.2 software, the results are shown in Tables 5−7.
The cumulative rate of in vitro release (Q) is as the dependent
variable Y, Fraction Input (R) is as the independent
Variable Xand the regression equations are as followsY =
75.272X + 7.531 5, r = 0.983 7 (NGR1); Y = 73.485X + 8.583,
r = 0.980 5 (GRg1); and Y = 65.929X + 10.402r = 0.960 3
(GRb1). The results are shown in Fig.4. When the degree of freedom
(df) was 6 – 2 = 4 and critical value (r4,
Fig. 3 Mean concentration-time profiles of Notoginsenoside R1 (A),
Ginsenoside Rg1 (B) and Ginsenoside Rb1 (C) in rat plasma after
oral administration of normal and bio-adhesive pellets. The dose of
oral administration of PNS was 90 mg·kg−1 (n = 6)
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 147 –
Table 2 Pharmacokinetic parameters of Notoginsenoside R1 after oral
administration of PNS normal and bio-adhesive pellets (200 mg·kg−1)
to rats (n = 6)
Parameter Control HPMC Chitosan Chitosan : Carbomer
Cmax/(ng·mL−1) 112.01 391.47** 272.1** 456.9**
Tmax/h 0.75 1.5** 1.0** 1.0**
MRT/min 24.38 26.9 13.34 6.43
AUC0-t/(ng·mL−1·h) 31 982.5 83 419.0 58 282.9 70 968.4
AUC0-∞/(ng·mL−1·h) 51 038.6 139 949.1 70 601.7 75 564.8
Fr% 100.00 260.83** 182.23** 221.90**
Fr: relative bioavailability = (AUC0-24(test) ×
100)/(AUC0-24(control)). Comparative pharmacokinetics profile of
NGR1 in blood (**P < 0.01 between normal
and bio-adhesive pellets)
Table 3 Pharmacokinetic parameters of Ginsenoside Rg1 after oral
administration of PNS normal and bio-adhesive pellets (200 mg·kg−1)
to rats (n = 6)
Parameter Control HPMC Chitosan Chitosan : Carbomer
Cmax/(ng·mL−1) 513.8 1 099.5** 1 458.5** 1 556.1**
Tmax/h 0.75 1.5** 1.0** 1.0**
MRT/min 24.49 47.23 14.95 6.91
AUC0-t/(ng·mL−1·h) 128 818.5 311 731.8 261 009.1 293 905.2
AUC0-∞/(ng·mL−1·h) 208 919.0 669 067.2 330 085.5 315 191.4
Fr% 100 320.25** 228.15** 202.62**
Fr: relative bioavailability = (AUC0-24(test) ×
100)/(AUC0-24(control)). Comparative Pharmacokinetics profile of
GRG1 in blood (**P < 0.01 between normal
and bio-adhesive pellets)
Table 4 Pharmacokinetic parameters of Ginsenoside Rb1 after oral
administration of PNS normal and bio-adhesive pellets (200 mg·kg−1)
to rats (n = 6)
Parameter Control HPMC Chitosan Chitosan : Carbomer
Cmax/(ng·mL−1) 4 977.8 11 883.9** 9549.5** 14 884.1**
Tmax/h 1.0 1.0 1.5** 1.0
MRT/min 14.94 11.87 48.76 11.04
AUC0-t/(ng·mL−1·h) 2 693 861 5 284 163.6 2 740 140.0 4 129
113.1
AUC0-∞/(ng·mL−1·h) 3 359 531 6 350 225.3 6 376 695.8 4 871
884.3
Fr% 100.0 189.02** 145.01* 189.8**
Fr: relative bioavailability = (AUC0-24(test) ×
100)/(AUC0-24(control)). Comparative Pharmacokinetics profile of
GRb1 in blood (**P < 0.01 between normal and bio-adhesive
pellets).
Table 5 Parameter of devolution of NGR1, GRg1, and GRb1after oral
administration of bio-adhesive pellet with HPMC to rats (n =
6)
C/(ng·mL−1) AUC/(ng·min·mL−1) Fraction Input (R) Q/% t/min
NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1
0 0 0 0 0 0 0 0 0 0 0 0 0
30 57.65 215.17 5 521.76 591.103 2 2 355.017 4 127 458 0.000 499 3
0.028 05 0.028 63 13.02 13.88 13.82
60 213.40 1 047.16 11 883.96 2 347.674 6 8 912.332 9 250 068.9
0.073 60 0.106 1 0.113 7 19.21 18.07 17.77
90 391.47 1 099.57 11 534.13 4 137.341 0 15 637.732 6 336 417.9
0.144 39 0.186 2 0.200 4 23.56 24.75 24.05
120 247.26 1 032.80 11 857.18 5 556.846 9 21 506.175 3 663 347.1
0.194 25 0.256 1 0.269 1 29.20 30.08 25.01
240 60.09 237.48 9 330.85 10 548.798 1 41 400.191 8 950 573.1 0.383
02 0.493 1 0.510 9 48.05 46.26 38.54
360 53.52 112.25 3 054.29 13 499.047 1 52 830.494 8 1 209 729.9
0.548 9 0.629 2 0.653 7 53.85 52.73 44.39
480 34.10 134.98 2 632.78 15 991.441 5 63 612.743 1 864.75 0.698 5
0.757 6 0.774 4 64.62 64.12 52.86
720 30.20 125.16 1 875.22 20 649.123 2 83 961.242 1 1 731 881.1 1 1
1 83.55 81.36 77.81
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 148 –
Fig. 4 The in vivo and in vitro correlation curves for NGR1, GRg1
and GRb1 in bio-adhesive pellets with HPMC. Q(%) represents the
cumulative rates of in vitro release, R represents Fraction
Input
0.05) was 0.811, the correlation coefficient of regression equation
(r) was > r4, 0.05, which indicated that NGR1, GRg1 and GRb1 in
bio-adhesive pellets with HPMC had a good correlation between in
vitro release and in vivo absorption, demonstrating that we could
predict the concentration-time data in vivo using the in vitro
release data. of bio-adhesive pellets with HPMC.
The cumulative rates of in vitro release (Q) is as the dependent
variable Y, Fraction Input (R) is as the
independent Variable Xand the regression equations are as
followsY = 86.252X + 25.094, r = 0.811 8 (NGR1), Y =
79.724X + 23.587, r = 0.802 5 (GRg1), and Y = 70.11X + 24.387, r =
0.772 3 (GRb1). The results are shown in Fig. 5. When the degree of
freedom (df) was 6 − 2 = 4 and critical value (r4, 0.05) was 0.811,
the correlation coefficient of regression equation (r) was < r4,
0.05, indicating that NGR1, GRg1 and GRb1 in bio-adhesive pellets
with chitosan : carbomer had a bad correlation between in vitro
release and in vivo absorption, which demonstrated that we couldn’t
predict the concentration- time data in vivo using the in vitro
release data of bio-adhesive pellets with chitosan :
carbomer.
Table 6 Parameters of devolution of NGR1, GRg1, and GRb1 after oral
administration of bio-adhesive pellet with chitosan- carbomer (n =
6)
C/(ng·mL−1) AUC/(ng·min·mL−1) Fraction Input (R) Q/% t/min
NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1
0 0 0 0 0 0 0 0.0 0.0 0.0 0 0 0
30 44.125 2 169.189 0 2 741.836 081 591.103 2 2 355.017 4 26 395.35
0.056 0 0.056 88 0.039 79 37.02 34.45 33.69
60 456.988 2 1 556.140 9 14 884.168 7 2 347.67 46 8 912.332 9 127
458 0.222 6 0.215 3 0.192 1 51.35 47.64 47.73
90 185.868 7 786.238 6 8 609.312 637 4 137.341 0 15 637.732 6 250
068.9 0.392 2 0.377 7 0.377 0 71.84 63.00 60.14
120 310.294 3 1 252.914 1 12 339.668 71 5 556.846 9 21 506.175 3
336 417.9 0.526 9 0.519 5 0.507 2 81.22 78.37 68.77
240 56.597 2 252.284 5 3 804.050 877 1 0548.798 41 400.191 8 663
347.1 1 1 1 98.68 91.01 84.35
Fig. 5 The in vivo and in vitro correlation curves for NGR1, GRg1
and GRb1 in bio-adhesive pellets with chitosan: carbomer. Q(%)
represents the cumulative rates of in vitro release, R represents
Fraction Input
Similarly, the cumulative rates of in vitro release (Q) is as
the dependent variable Y, Fraction Input (R) is as the
independent Variable X, and the regression equations are as
follows: Y = 71.351X + 45.98, r = 0.398 8 (NGR1); Y =
68.176X + 42.198, r = 0.413 9 (GRg1); and Y = 64.417X +
45.18, r = 0.772 3 (GRb1). The results are shown in Fig. 6.
When the degree of freedom (df) was 6 − 2 = 4 and
critical value (r4, 0.05) was 0.811, the correlation
coefficient
of regression equation (r) was < r4, 0.05, indicating that
NGR1, GRg1, and GRb1 in bio-adhesive pellets chitosan had
a bad correlation between in vitro release and in vivo
absorption, which demonstrated that we couldn’t predict the
concentration-time data in vivo using the in vitro release
data
of bio-adhesive pellets with chitosan.
Discussion
In a previous study, we have explored in vitro adhesion of pellets
with different bio-adhesive materials using the tissue retention
method, the results demonstrated that in vitro adhesion of HPMC,
chitosan and chitosan : carbomer
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 149 –
Table 7 Parameters of devolution of NGR1, GRg1 and GRb1after oral
administration of bio-adhesive pellet with chitosan (n = 6)
C/(ng·mL−1) AUC/(ng·min·mL−1) Fraction Input (R) Q/% t/min
NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1 NGR1 GRg1 GRb1
0 0 0 0 0 0 0 0.0 0.0 0.0 0 0 0
30 20.986 5 138.260 8 2 750.225 0 591.103 2 2 355.0174 0 26 395.35
0.142 9 0.150 6 0.105 6 1.021 5 0.948 5 0.921 0
60 272.115 5 1 458.586 0 7 305.430 7 2 347.674 6 8 912.332 9 127
458 0.567 4 0.569 9 0.509 7 1.016 6 0.954 6 0.969 7
90 73.463 3 359.755 0 9 549.556 1 4 137.341 0 15 637.732 6 250
068.9 1 1 1 1.021 41 0.957 8 0.957 0
Fig. 6 The in vivo and in vitro correlation curves for NGR1, GRg1
and GRb1 in bio-adhesive pellets with chitosan. Noting: Q(%)
represents the cumulative rates of in vitro release, R represents
Fraction Input
combination were both 3−4 times larger than the control group
(pellets without bio-adhesive materials) [21]. The results
demonstrated our prepared pellets with bio-adhesive materials had
mucous membrane adhesion characteristics.
In the present study, we prepared three types of bio-adhesive
pellets, i.e., HPMC, chitosan and chitosan : carbomer (1 : 1),
selecting the best preparation for future studies. Because PNS are
not stable under acid conditions and degraded easily, choosing the
artificial intestinal fluid as release medium to explore drug
release behavior is proper. Our results demonstrated that
bio-adhesive pellets with HPMC released in sustained fashion for 12
h. Bio-adhesive pellets with chitosan: carbomer (1 : 1) could
release completely within 3−4 h. Bio-adhesive pellets with chitosan
disintegrated rapidly and released completely, which showed no
difference from that normal pellets.
HPMC is a kind of hydrophilic gel material. The bio-adhesive with
HPMC has hydrophilic gel skeleton type release mechanism. When HPMC
is placed in aqueous
medium, due to the hydration and a thick gel barrier formation,
internal medicine would slowly spread to the surface and dissolve
in medium, which would prolong the time of drug release and block
PNS release in the stomach, avoiding PNS degradation in the
stomach. The impetus of drugs diffusion comes from drug
concentration gradient in the skeleton, and the release behavior is
slow after the first fast release. The release mechanism
approximated Higuchi equation or first order kinetic
equation.
Under high pH conditions, the chitosan expands slowly, gel layer is
formed slowly, the solution could seep into the inside of skeleton
and dissolve drugs, and eventually the skeleton is disintegrated,
leading to the rapid release. Carbomer is different from chitosan;
due to its low pKa value, carbomer expands rapidly under the high
pH conditions, delaying the release of drugs.
In establishing and validating the analytical methods for
determination of in vivo samples, we showed that HPLC-MS/MS could
determinate three major components NGR1, GRg1, and GRb1 in PNS
preparations. We found the NGR1, GRg1, GRb1 of [M − H]− and
internal standard of [M + HCOO]− peaks in the negative ion mode
with ESI ion source, but the abundance values were not high. On the
contrary, the abundance values of [M + Na]+ in the positive ion
detection mode with ESI were higher, and the ion pair was stable,
so MRM detection of drugs was chosen in positive ion mode
[22-23].
During establishing the biological sample pretreatment methods, we
investigated the influence of extraction solvents (methylene
chloride, methanol and acetone, acetone and methanol), the volume
of solvent (2, 3, and 10 times), extraction times (once or twice),
and mobile phases (methanol, acetonitrile, formic acid aqueous
solution, and ammonium acetate solution) as well as the internal
standard. Using methanol or acetonitrile alone as solvent
extraction, the extraction efficiency was incomplete, and the
matrix effect exceeded the requirement range. After methanol was
added into acetone, the ingredients in plasma that were bound to
plasma proteins were extracted, further improving the extraction
efficiency [24-25]. Therefore, according to the literature [24],
the final extraction methods were established as follows: solvent
was acetone: methanol (3 : 1), the extraction solvent volume was
three times of the amount of samples, and extracting 1 times. The
impurities were removed relatively completely, the process was
simple and easy to handle, and
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 150 –
the extraction rate and the matrix effect could meet the
requirements for the determination of biological samples.
To determine three main ingredients of PNS, NGR1, GRg1, and GRb1,
simultaneously, the linear gradient elution method was used in the
present study. In HPLC-UV detection, if mobile phase was methanol,
the baseline fluctuated widely, and methanol had end absorption
which could affect the PNS determination results. But methanol
would not affect the function of the mass spectrometry detector
with HPLC- MS/MS. So we chose methanol as organic mobile phase. We
found that water with ammonium acetate and n-butyl amine solution
could significantly improve the efficiency of ionization of three
ingredients [26], and therefore, we selected 10 mmol·L−1 of
ammonium acetate solution (containing 0.5% n-butyl amine) as the
water phase.
Internal standard must have similar polarity to NGR1, GRb1 and GRg1
and no obvious difference from their molecular weight, and do not
interfere with their determination. According to the relevant
literatures [27-30], we chose Digoxin as the internal standard,
considering that its polarity, molecular weight (779.5) and
retention time were similar to that of the three compounds, which
met the requirements for sample analysis.
Our results indicated that the bio-adhesive pellets with HPMC and
chitosan : carbomer (1 : 1) had good sustained release effect, the
Tmax values of each ingredient showed different degrees of
extension, and the Cmax values were greater than that of the normal
pellets. Bio-adhesive pellets has a quick release or accelerated
absorption in vivo process, which was effective absorption
enhancer. According to the dissolution of the pellets in the
artificial intestinal fluid, the release of bio-adhesive pellets
with HPMC and chitosan: carbomer (1: 1) was slow, and therefore,
the Cmax increases of the bio-adhesive pellets with HPMC and
chitosan: carbomer (1: 1) were due to the acceleration PNS
absorption and the extension of Tmax was the result of sustained
release.
Carbomer could open tight junction and promote absorption by
paracellular route. Besides, it also has effect of tissue adhesion
and enzymatic degradation of gastrointestinal tract. For pellets
with chitosan alone, cationic chitosan can interact with anionic
membranes, which could open tight junction and promote absorption
by paracellular route [31], so as to improve oral bioavailability
of the drug.
The three ingredients of PNS analyzed in the present study are
water-soluble, and the poor membrane permeability and bile drainage
are the two main reasons for low bioavailability. In addition GRg1
undergoes a rapid kidney metabolism [32]. Therefore, delaying the
drug release and slowing kidney metabolism could increase oral
bioavailability greatly. Indeed, the extent of oral bioavailability
increase in GRg1 for all kinds of bio-adhesive materials was the
largest among the three ingredients of PNS.
The deconvolution method does not rely on the simulation of
compartmental model and is especially suitable
for drugs that do not comply with compartmental model and may be
suitable for all kinds of in vivo and in vitro correlation
research, with the characteristics of simple concept and intuitive
mathematical operations, which has been recorded by the United
States Pharmacopoeia.
The in vitro release pattern of bio-adhesive pellets with HPMC is
through multi-mechanism of diffusion and skeleton dissolution,
releasing slowly. In the present study, we got a good correlation
between in vivo and in vitro results using the deconvolution
method. However, the three ingredients of bio-adhesive pellets with
chitosan : carbomer and chitosan showed no significant in vivo and
in vitro correlations. Chitosan has different release
characteristics under artificial gastric fluid and intestinal fluid
conditions. The main factors affecting dissolution of soluble drugs
from expansible skeleton are associated with the speed of gel layer
formation on the preparation surface. If the gel layer forms
slowly, the solution could seep into the inside of skeleton,
dissolving drugs, disintegrating skeleton eventually, and causing
rapid drug release. Chitosan has good solubility in the artificial
gastric fluid and poor solubility in artificial intestinal fluid.
Under low pH conditions, chitosan expands rapidly, and the gel
layer forms fast, which could delay the release of drugs. Under
high pH conditions, drugs release fast, and the drug release
behavior reflects drug absorption in vivo.
Conclusion
oral bioavailability of bio-adhesive pellets with different
bio-adhesive materials were as follows: HPMC pellets >
chitosan: carbomer pellets > chitosan pellets >
ordinary
pellets, which provided further evidence that the preparation
of PNS bio-adhesive pellets could improve PNS oral
bioavailability. According to previous research [18], HPMC
K4M skeleton materials could promote PNS absorption in
intestinal perfusion model, and bio-adhesive pellets with
HPMC had the characteristics of sustained release and higher
bio-adhesive property. Compared with other preparations, the
increase in oral bioavailability of PNS bio-adhesive pellets
with HPMC was the largest, with a good correlation between
in vitro release and in vivo bioavailability, demonstrating
that the in vitro release behaviors could forecast the in
vivo
absorption properties, so as to guide preparation of
prescription and technology screening.
References
[1] Liu L, Ning BB, Cui JG, et al. miR-29c is implicated in the
cardioprotective activity of Panax notoginseng saponins against
isoproterenol-induced myocardial fibrogenesis [J]. J
Ethnopharmacol, 2017, 198: 1-4.
[2] Yang BR, Cheung KK, Zhou X, et al. Amelioration of acute
myocardial infarction by saponins from flower buds of Panax
notoginseng via pro-angiogenesis and anti-apoptosis [J]. J
Ethnopharmacol, 2016, 181: 50-58.
LI Ying, et al. / Chin J Nat Med, 2017, 15(2): 142151
– 151 –
[3] Wu JH, Leung GPH, Kwan YW, et al. Suppression of diet-induced
hypercholesterolaemia by saponins from Panax notoginseng in rats
[J]. J Funct Foods, 2013, 5(3): 1159-1169.
[4] Chen SX, Liu JL, Liu XY, et al. Panax notoginseng saponins
inhibit ischemia-induced apoptosis by activating PI3K/Akt pathway
in cardiomyocytes [J]. J Ethnopharmacol, 2011, 137(1):
263-270.
[5] Yuan ZM, Liao Y, Tian G, et al. Panax notoginseng saponins
inhibit Zymosan A induced atherosclerosis by suppressing integrin
expression, FAK activation and NF-κB translocation [J]. J
Ethnopharmacol, 2011, 138(1): 150-155.
[6] Gil EC, Colarte AI, Bataille B, et al. Development and
optimization of a novel sustained-release dextran tablet
formulation for propranolol hydrochloride [J]. Int J Pharm, 2006,
317(1): 32-39.
[7] Feng L, Jiang X. Absorption profiles of sanchinoside Rl and
ginsenoside Rgl in the rat intestine [J]. Eur J Drug Metab PH,
2005, 30(4): 261-268.
[8] Akiyama Y, Nagahara N, Nara E, et al. Evaluation of oral
mucoadhesive microspheres in man on the basis of the
pharmacokinetics of furose mide and riboflavin, compounds with
limited gastrointestinal absorption sites [J]. J Pharm Pharncol,
1998, 50(2): 159-166.
[9] Honsy EA, Angary AA. Bioavailability of sustained release
indomethacin suppositories containing polycarbophil [J]. Int J
Pharm, 1995, 113: 209-213.
[10] Lai L, Liu HG, Lu SH, et al. Pharmacokinetics and correlation
between in vivo absorption and in vitro release of Panax
notoginseng saponins enteric microcapsule [J]. Chin J New Drug,
2012, 21(6): 693-696.
[11] Lai L, Liu HG, Wen L, et al. Correlation between in vivo
absorption and in vitro release of Panax notoginseng saponins
enteric pellets [J]. Chin J Exper Tradit Med Form, 2011, 17(24):
97-100.
[12] Chen QY, Hu PY, Wang WP, et al. Drug release in vitro
mathematical model of Panax notoginoside micro-porous osmotic pump
tablet [J]. Chin Tradit Patent Med, 2009, 31(4): 538-540.
[13] Wu Y, Zhang GZ, Lin N. Study on preparation and release
characteristics of Panax notoginseng saponins controlled- release
preparations [J]. J Hubei Univer Chin Med, 2013, 15(2):
33-36.
[14] Feng H, Chen W, Zhu C. Pharmacokinetics study of bio-adhesive
tablet of Panax notoginseng saponins [J]. Inter Arch Med, 2011, 4:
18.
[15] Chen W, Zhu C. Evaluation of adhesion and release kinetics of
Panax notoginseng saponines bioadhesive tablets in vitro [J]. Chin
Pharm J, 2006, 41(12): 917-920.
[16] Zhu C, Chen W. Preparation and in vitro evaluation of Panax
notoginseng saponines (PNS) bioadhesive tablet swithcar-bopo [J].
Chin Pharm J, 2007, 42(24): 1877-1880.
[17] Gao C, Jiang X, Zhang C, et al. Determination of in vitro
adhesion force and trial on the biological adhesion formulation
screening [J]. Chin J Pharm, 2006, 37(9): 606-609
[18] Zhang Y, Zhu CY. Effects of bioadhesive materials on small
intesstinal absorption of Panax notoginseng saponins [J]. Chin
Pharm J, 2015, 50(14): 1215-1220.
[19] Bai J, Lu Y, Du SY, et al. Preparation and in vitro in
vivo
evaluation of Chuanxiong Cataplasm [J]. CJTCMP, 2013, 28(5):
1259-1266.
[20] Langenbucher F. Handling of computational in vitro/in vivo
correlation problems by Microsoft excel III convolution and
deconvolution [J]. Eur J Pharm Biopharm, 2003, 56(3):
429-437.
[21] Li Y, Zhang Y, Zhu CY. Study on in vitro adhesion and
retention in artificial gastric acid fluid of Panax notoginseng
saponins bio-adhesive pellets [J]. Chin Pharm J, impressing.
[22] Dai X. Non-parametric efficiency estimation using Richardson-
Lucy blind deconvolution [J]. Eur J Oper Res, 2016, 248(2):
731-739.
[23] Song M, Zhang S, Xu X, et al. Simultaneous determination of
three Panax notoginseng saponins at sub-nanograms by LC-MS/MS in
dog plasma for pharmacokinetics of compound Danshen tablets [J]. J
Chromatogr B, 2010, 878(32): 3331-3337.
[24] Xia C, Wang G, Sun J, et al. Simultaneous determination of
ginsenoside Rg1, Re, Rd, Rb1 and ophiopogonin D in rat plasma by
liquid chromatography/electrospray ionization mass spectrometric
method and its application to pharmacokinetic study of ‘SHENMAI’
injection [J]. J Chromatogr B, 2008, 862(1-2): 72-78.
[25] Guo Q, Li P, Wang Z, et al. Brain distribution pharma-
cokinetics and integrated pharmacokinetics of Panax Notoginsenoside
R1, Ginsenosides Rg1, Rb1, Re and Rd inrats after intranasal
administration of Panax notoginseng saponins assessed by UPLC/MS/MS
[J]. J Chromatogr B, 2014, 969: 264-271.
[26] Yu K, Ma Y, Shao Q, et al. Simultaneously determination of
five ginsenosides in rabbit plasma using solid-phase extraction and
HPLC/MS technique after intravenous administration of ‘SHENMAI’
injection [J]. J Pharmaceut Biomed, 2007, 44(2): 532-539.
[27] Yang S, Zhang K, Lin X, et al. Pharmacokinetic comparisons of
single herb extract of Fufang Danshen preparation with different
combinations of its constituent herbs in rats [J]. J Pharmat
Biomed, 2012, 67-68(3): 77-85.
[28] Kong LT, Wang Q, Xiao BX, et al. Different pharmacokinetics of
the two structurally similar dammarane sapogenins, protopanaxatriol
and protopanaxadiol in rats [J]. Fitoterapia, 2013, 86(1):
48-53.
[29] Wu H, Liu H, Bai J, et al. Simultaneous determination of
notoginsenoside R1, ginsenoside Rg1, ginsenoside Re and 20(S)
protopanaxatriol in beagle dog plasma by ultra high performance
liquid mass spectrometry after oraladministration of a Panax
notoginseng saponin preparation [J]. J Chromatogr B, 2015, 974:
42-47.
[30] Kong L. Ginseng saponin pharmacokinetic study [D]. Peking
Union Medical College, 2013.
[31] Alexei M, Charles R, Brian R, et al. Effect of chitosan
glutamate, carbomer 974P, and EDTA on the in vitro Caco-2
permeability and oral pharmacokinetic profile of acyclovir in rats
[J]. Drug Dev Ind Pharm, 2009, 35(9): 1082-1091.
[32] Liu O, Yang J, Du F, et al. Absorption and disposition of
ginsenosides after oral administration of Panax notoginseng extract
to rats [J]. Drug Metab Dispos, 2009, 37(12): 2290-2298.