Novel curcumin-loaded human serum albumin nanoparticles surface functionalized with folate: characterization and in vitro/vivo evaluation
Drug Design, Development and Therapy
novel curcumin-loaded human serum albumin nanoparticles surface functionalized with folate: characterization and in vitro/vivo evaluation
Zhiwang song 2
Yonglin lu 2
Xia Zhang 2
haiping Wang 1
Junyi han 0
chunyan Dong 2
0 Department of g astrointestinal s urgery, s hanghai east hospital, Tongji University , shanghai, People's republic of china
1 Department of Pharmacy
2 Breast cancer center
Folate-conjugated, curcumin-loaded human serum albumin nanoparticles (F-CM-HSANPs) were obtained by the chemical conjugation of folate to the surface of the curcumin (CM)-loaded human serum albumin nanoparticles (NPs). The NPs were characterized by various parameters, including size, polydispersity, zeta potential, morphology, encapsulation efficiency, and drug release profile. The mean particle size of F-CM-HSANPs was 165.6±15.7 nm (polydispersity index ,0.28), and the average encapsulation efficiency percentage and drug loading percentage of the F-CM-HSANPs were 88.7%±4.8% and 7.9%±0.4%, respectively. Applied in vitro, the CM NPs, after conjugation with folate, maintained sustained release, and a faster release of CM was more visibly observed than the unconjugated NPs. F-CM-HSANPs can prolong the retention time of CM significantly in vivo. However, after intravenous injection of F-CM-HSANPs, the pharmacokinetic parameters of CM were not significantly different from those of CM-loaded human serum albumin NPs. The improved antitumor activity of F-CM-HSANPs may be attributable to the protection of drug from enzymatic deactivation followed by the selective localization at the desired site. These results suggest that the intravenous injection of F-CM-HSANPs is likely to have an advantage in the current clinical CM formulation, because it does not require the use of a solubilization agent and it is better able to target the tumor tissue.
*These authors contributed equally
to this work
nanoparticles (NPs), nanoemulsions, liposomes, conjugates,
peptide carriers, cyclodextrins, and solid dispersions.8–11
In nanotechnology, human serum albumin (HSA) has
attracted a wide range of interest as a carrier system for
drug delivery, especially in the field of cancer treatment.12,13
It is a nonimmune and nontoxic material and has good
bio108 compatibility and biodegradability.14 In addition, albumin
l-J2u nanotechnology does not require surfactants or polymeric
-12 materials for preparation. Therefore, it is believed that the
8on tolerance of human serum albumin nanoparticles (HSANPs)
..73142 ibny vuivsiongis allibkuelmyitno cbaerrgieoro,di.sAthberabxeasnte®e,xwamhipclhe iasmporenpgarthede
.183 wide range of clinical applications.15 In addition, the amino
yb5 and carboxylic groups of the HSA structure can be used for
/om surface modification.16 In the chemotherapeutic treatment
.ssc of cancer, it was very important to enrich the drug to the
repe tumor cells and simultaneously reduce the drug-associated
vo adverse effects.
.//:sdwww l.yseonu used in the colloidal systems that selectively target tumor
Among the targeting ligands, folic acid has been widely
ttph lan tissues. The specific advantages of folic acid include small
rom rse size (Mw =441.4 Da), low immunogenicity, easy
dead roF tion, low cost, storage stability, solvent compatibility, and
lon high affinity (Kd =10−10 M) with the folate receptor alpha.17
odw The main objective of this study was to develop and
ypa characterize functionalized HSANPs as the potential carrier
reh for site-specific drug delivery of CM. Folate-conjugated,
dnT curcumin-loaded, human serum albumin nanoparticles
tan (F-CM-HSANPs) were prepared by the chemical
conjugaepm tion of folate to the surface of the curcumin-loaded, human
lveo serum albumin nanoparticles (CM-HSANPs). The developed
,eD formulations were also characterized by assessing
morpholisgn ogy, particle size, zeta potential, encapsulation efficiency
egD (EE), and drug release profile. The cytotoxic activity and
ruD anticancer efficacy of the developed formulations were also
evaluated in vivo.
Materials and methods
Materials and animals
CM (purity .99%) was obtained from Shanghai Yuanye
Biopharma Co., Ltd. (Shanghai, People’s Republic of China)
HSA (purity 96%–99%, 65 kDa) and all other chemicals
were obtained from Shanghai Chemical Reagent Company of
Chinese Medicine. High-performance liquid chromatography
(HPLC)-grade acetonitrile and methanol were obtained from
Sigma-Aldrich Co. (St Louis, MO, USA). All other chemicals
were of analytical grade. Distilled water was used in all the
experiments. HT29 cell lines were obtained from MeiSci
Pharma (Shanghai, People’s Republic of China).
Preparation of cM-hsanPs
In this study, the albumin nanotechnology described in
the literature was used to prepare CM-HSANPs.18 Briefly,
900 mg of HSA was dissolved in 25 mL of water saturated
with chloroform and 100 mg of CM was dissolved in 1.5 mL
of chloroform saturated with water. Afterward, the CM
solution was added drop wise into the HSA solution with a
stirring rate of 300 rpm. Oil/water emulsion was obtained
by homogenizing at 20,000 psi for ten cycles. After
homogenization, the produced emulsion was rotary evaporated
for 20 minutes to allow evaporation of chloroform at 25°C
under reduced pressure. The NPs were collected after being
filtered through a 0.25 µm membrane syringe filter, and the
solvent was removed by lyophilization for 36 hours at −70°C.
The CM-HSANPs powder was collected under vacuum for
24 hours at 4°C.
Preparation of F-cM-hsanPs
N-Hydroxysuccinimide ester of folic acid (NHS-folate) was
prepared according to the previously reported method.19
In all, 300 mg of folic acid was dissolved in 6 mL of
anhydrous dimethyl sulfoxide containing 150 mL of triethylamine,
282 mg of dicyclohexylcarbodiimide, and 156 mg of NHS.
The mixture was magnetically stirred overnight at room
temperature. The insoluble by-product dicyclohexylurea
was removed by filtration. The filtrate was poured into the
solution of ice-cold anhydrous ether containing 30% acetone
to precipitate the formation of NHS-folate. The NHS-folate
was obtained by centrifugation at 10,000 rpm for 10 minutes
at 4°C and washed with anhydrous ether. Finally, the product
was dried under vacuum.
A total of 5 mg of NHS-folate was dissolved in 1 mL of
carbonate/bicarbonate buffer solution (0.2 M, pH 10) and
added dropwise to the suspension of CM-HSANPs (2 mL;
pH was adjusted to ten using 0.2 M carbonate/bicarbonate
buffer) with mild stirring. The reaction was allowed to
proceed for 1 hour at room temperature. The mixed liquid
was centrifuged at 12,000 rpm for 20 minutes and washed
with bicarbonate buffer solution. The prepared NPs were
freeze dried without other cryoprotectant.
The particle size distribution of prepared CM-HSANPs
and F-CM-HSANPs was determined by a laser diffraction
particle sizer. The measurements were performed at 25°C
after equilibration for 2 minutes. All analyses were performed
in triplicate (with three samples).
The morphology of the CM-HSANPs and
F-CMHSANPs was observed under a transmission electron
microscopy. Before observation, the NPs were prepared as
follows: the samples were diluted with distilled water and
placed on a copper grid covered with nitrocellulose before
negatively stained with phosphotungstic acid and dried at
Drug loading (DL) and EE of CM-HSANPs and
F-CMHSANPs were identified as follows. Briefly, 0.5 mL of
drug-loaded NP pulp was introduced into preweighed
Eppendorf tubes and lyophilized to a constant weight.
Then the dried deposits were dissolved and diluted with
methanol. The method of HPLC was used to determine
the CM amount in solution. Finally, DL and EE of
drugloaded NPs were calculated according to the following
DL(%) = Amount oWfCeiMghitnoffofromrmulualtaiotino−n A(pmoloyumnetro)f free CM ×100 (
EE(%) = Amount ofACmMoiunntfoorfmCuMlatiinonfo−rmAmuloatuinotnof free CM ×100 (
The concentration of CM in the prepared NPs was
determined by HPLC. The chromatographic separation was
performed on a column (C18, 4.6×250 mm, 5 µm) at the
temperature of 30°C. The mobile phase was a mixture of
methanol and water at a ratio of 70:30 (v/v) containing 1%
acetic acid. The flow rate was 1 mL/min, and the detective
wavelength was 428 nm. In all, 10 µL of each sample was
injected into the column, and blank runs were carried out
randomly between samples to allow chromatographic
carryover to be evaluated. The retention time of CM was
Prepared freeze-dried F-CM-HSANPs were stored in closed
glass containers at 25°C and 60% relative humidity and
examined on the day of production and after 3 months of
storage. NPs were analyzed for particle size, DL, EE, and
polydispersity index (PDI) at 1-month intervals.
in vitro drug release
The free CM, CM-HSANPs, and F-CM-HSANPs
(containing 20 mg CM) were put into three dialysis tubes
(molecular weight cut-off 8,000–12,000), respectively,
and subjected to dialysis against 50 mL of normal saline
with 1% Tween 80 in the dark environment at 37°C.20 In
all, 1 mL of the sample was taken out from the released
medium periodically at 0.25, 0.5, 1, 2, 4, 6, 8, 10, 12, 24,
and 36 hours. After each sampling, 1 mL of fresh normal
saline with 1% Tween 80 was added into the medium to keep
the volume constant. HPLC was used for the determination
of drug concentration.
Twelve rats weighing between 200 g and 250 g were used
in this study. The rats were kept under fasting conditions
for 24 hours before the experiment started. In this
experiment, three groups of rats (six per group) were
intravenously administered by tail vein with either CM solution
(10 mg/kg, as a control) or CM-HSANPs (10 mg/kg) and
F-CM-HSANPs formulation (10 mg/kg). In all, 2 mL of
serum samples were collected in pressure equalization tubes
at different time intervals before and after drug
administration (0.5, 1, 2, 3, 4, 6, 8, 10, 12, and 24 hours). The obtained
serum samples were frozen at −20°C until the analysis of
HPLC was performed. All animal study protocols and care
guidelines were approved by the Institutional Animal Care
and Use Committee at the Hospital of Shanghai.
in vivo antitumor activity
BALB/c male mice (6–7 weeks old) were used as a tumor
xenograft model. Tumors were established by inoculating HT
29 cells (5×108 cells/mice, 50 µL injection) subcutaneously
into a dorsal flank of each mouse. Five days after tumor cell
inoculation, the mice were administered phosphate-buffered
saline, free CM (10 mg/kg, intravenous [IV], solubilized in
Cremophor EL/ethanol), CM-HSANPs, and F-CM-HSANPs
(10 mg/kg, IV) dissolved in normal saline every other day for
10 days (eight mice in each). A digital caliper was used to
measure the tumor diameters, and tumor volumes (in mm3)
were calculated using the following formula:
Tumor volume = Length × Width2 ×0.5
The body weights were measured every 2 days in order to
monitor the potential toxicities. The in vivo antitumor study
was terminated 20 days after injection.
Data were shown as mean ± SD (n=5). Statistical data were
analyzed by the Student’s t-test at the level of P,0.05.
Results and discussion
The physicochemical characteristics of F-CM-HSANPs
formulations were compared to the corresponding CM-HSANPs
(Figure 1). The mean particle size of F-CM-HSANPs
(165.6±15.7 nm, PDI ,0.28) was slightly higher, yet deviated
by no more than ±30 nm from that of the corresponding
CM-HSANPs (132.2±9.8 nm, PDI ,0.26). The increase
in particle size was attributed to the bulkiness provided
by folate moiety to the NPs. In the experiment, HSA has
a negative charge, so the prepared NPs also exhibited
negative zeta potential. Since the folate moiety contains a
negative residue in its structure, the zeta potential of
folateconjugated NPs (−27.3±4.2 mV) was slightly more negative
./dwww l.yeon than the unconjugated ones (−21.3±2.5 mV).21 The average
/:ttsp lsua EE% and DL% of the CM-HSANPs and F-CM-HSANPs
h no were 86.3%±6.8% and 88.7%±4.8%, and 7.5%±0.8% and
from rsep 7.9%±0.4%, respectively.
The stabilization of our preparations was assessed according
to the International Conference on Harmonisation guidelines.
After 3 months of storage stabilization at 25°C and 60% relative
humidity, lyophilized F-CM-HSANPs appeared to be stable as
dried cakes, showing no collapse or contraction. In addition,
particle size and PDI measurements showed no change in
stability during the storage procedure (Table 1). Long-term stability
is a major requirement that must be fully met, and lyophilizing
is considered as a viable long-term stability test. Therefore,
since HSA acts as a cryoprotectant and aids reconstitution,
F-CM-HSANPs were lyophilized with no excipients.
in vitro drug release
The in vitro release profile of CM from the free CM,
CMHSANPs, and F-CM-HSANPs was studied in normal saline
Table 1 storage stability of F-cM-hsanPs at 25
rh (n =3)
°c and 60%
0 day 1 month 2 months 3 months
Particle size (nm) 165.6±15.7 165.9±16.2 167.3±15.5 168.4±14.9
Dl% 5.5±0.7 5.6±0.8 5.5±0.5 5.4±0.6
ee% 83.2±6.9 81.8±7.2 82.1±6.4 80.3±5.6
PDi ,0.28 ,0.30 ,0.31 ,0.33
Abbreviations: F-cM-hsanPs, folate-conjugated, curcumin-loaded human serum
albumin nanoparticles; rh, relative humidity; Dl, drug loading; ee, encapsulation
efficiency; PDI, polydispersity index.
(Figure 2). Over time, CM in HSANPs was released much
more slowly than free CM. It was clearly observed that the
release curve of CM-NPs exhibited a biphasic pattern: an
initial burst release during the first 4 hours followed by a slower
and sustained release over a long period of time. Owing to the
poor water solubility of CM, it was difficult for it to penetrate
into the matrix of HSANPs. The diffusion rates of the NPs
prepared with HSA in the release medium determine whether
it has a good sustained-release performance.
Therefore, the results indicated that the NPs have a good
sustained-release function. After conjugated with folate, the
CM NPs maintained the properties of sustained release, and
a faster release of CM was more visibly observed than the
unconjugated NPs. However, there was no obvious difference
between the two NPs in the release curve.
To investigate the kinetic modeling of drug release from
NPs, the dissolution profiles were fitted to zero-order, Q = k0⋅t;
first-order, ln (100−Q) = ln Q0−k1⋅t; Higuchi, Q = kH⋅t1/2; and
Korsmeyer–Peppas models.22 The best-fitted model of rhein
released from the NPs was the Higuchi kinetic model; the
correlation coefficient (r) was 0.9991, revealing that the form
of NPs could control the CM release.
The pharmacokinetics studies of CM solution, CM-HSANPs,
and F-CM-HSANPs were conducted in rats (Figure 3).
Based on the above results, in the group of CM solution,
high concentrations of free CM in the plasma were recorded
5 minutes after dosing; however, its concentrations decreased
rapidly and became nondetectable 6 hours later. In contrast,
the concentrations of CM encapsulated in HSANPs could
still be determined at 24 hours following injection. The
pharmacokinetic results indicated that the clearance of CM
encapsulated in NPs decreased, and preparation of CM into
NPs could significantly change the pharmacokinetic behavior
of CM. The elimination half-life (t1/2) values of CM-HSANPs
and F-CM-HSANPs were 0.36 hours and 0.43 hours,
respectively (Table 2). NPs provided higher AUC0–∞ (2.92-fold) and
half-life (t1/2; 5.3-fold) compared to the free CM. Thus, it was
reasonable to conclude that NPs could significantly extend the
role of CM in vivo (provided higher bioavailability). This may
be due to the fact that a lot of HSA blocks were conjugated
to the surface of the NPs, which could substantially increase
the circulating time of NPs in the blood. Moreover, there was
no significant difference in the pharmacokinetic parameters
between F-CM-HSANPs and CM-HSANPs after IV
administration, just because there was no obvious difference in the
contents of HSA blocks of the two nanosystems.
in vivo antitumor activity
To evaluate the antitumor activity of F-CM-HSANPs in
human colon cancer (HT 29) xenograft models in vivo, we
Parameter IV administration
Solution CM-HSANPs F-CM-HSANPs
t1/2 (h) 0.08±0.01 0.36±0.05* 0.43±0.07*
aUc 0–t (ng⋅h/ml) 2,095.2±165.3 5,427.3±487.9* 6,450.9±632.2*
aUc 0–∞ (ng⋅h/ml) 2,298.3±187.6 5,761.4±518.5* 6,721.6±643.1*
MrT (h) 2.9±0.8 13.7±1.4* 14.7±1.5*
cl (l/h) 16.7±1.8 3.2±0.6* 3.6±0.4*
Note: *P,0.05 vs cM solution.
Abbreviations: cM, curcumin; iV, intravenous; cM-hsanPs, curcumin-loaded,
human serum albumin nanoparticles; F-cM-hsanPs, folate-conjugated,
curcuminloaded human serum albumin nanoparticles; t1/2, half-life; h, hours; aUc, area under
the curve; MrT, mean residence time; cl, clearance.
examined tumor growths and body weight changes in nude
mice treated with saline, free CM, CM-HSANPs, and
F-CMHSANPs (10 mg/kg). Tumors cells were formed at all sites
after 5 days of tumor cell injection (Figure 4). The highest
growth rate of tumor was found in the saline-treated mice,
and free CM was found to have a slight antitumor effect (18%
tumor growth inhibitions at the end of the treatment).
However, CM-HSANPs and F-CM-HSANPs significantly
inhibited the growth of HT 29-derived tumors (45% and 64%) at
the tenth day (Figure 5). The nanocarriers can be transported
through the lymphatic system.23,24 The antitumor activity of
the F-CM-HSANPs was improved, which may be due to the
protection of the drug from enzymatic deactivation followed
ttp la by the selective localization at the desired site. The changes
o in body weights could reflect the toxicities after the
f p ments. Animals treated with free CM showed a decrease in
body weight vs the control group (phosphate-buffered saline
treated), whereas mice given NPs showed no significant
reduction in body weight (Figure 6). All results indicated
that the IV injection of F-CM-HSANPs is likely to have an
advantage in the current clinical CM formulation, because
it does not require the use of a solubilization agent and is
better able to target the tumor tissue.
In this study, F-CM-HSANPs were obtained by the chemical
conjugation of folate to the surface of the CM-HSANPs. The
mean particle size of F-CM-HSANPs was 165.6±15.7 nm
(PDI ,0.28), and the average EE% and DL% of the
F-CMHSANPs were 83.2%±6.9% and 5.5%±0.7%, respectively.
Applied in vitro, the CM NPs, after conjugated with folate,
maintained sustained release, and a faster release of CM was
more visibly observed than the unconjugated NPs. In vivo,
F-CM-HSANPs can significantly prolong the retention time
of CM. However, the pharmacokinetic parameters of CM
after IV administration of F-CM-HSANPs had no obvious
difference compared with CM-HSANPs. The improved
antitumor activity of the F-CM-HSANPs may be attributable
to the protection of the drug from enzymatic deactivation
followed by the selective localization at the desired site.
These results suggest that the IV injection of F-CM-HSANPs
is likely to have an advantage in the current clinical CM
formulation, because it does not require the use of a
solubilization agent and is better able to target the tumor tissue.
This research was supported in part by the National Nature
Science Foundation of China (81573008), the Fund of
Pudong Health Bureau of Shanghai (PWRd2014-01), and
the Key Disciplines Group Construction Project of Pudong
Health Bureau of Shanghai (PWZxq2014-04).
The authors report no conflicts of interest in this work.
Publish your work in this journal
Drug Design, Development and Therapy is an international,
peerreviewed open-access journal that spans the spectrum of drug design
and development through to clinical applications. Clinical outcomes,
patient safety, and programs for the development and effective, safe,
and sustained use of medicines are the features of the journal, which
1. Jurenka JS . Anti-inflammatory properties of curcumin, a major constituent of Curcuma longa: a review of preclinical and clinical research . Altern Med Rev . 2009 ; 14 ( 2 ): 141 - 153 .
2. Moghadamtousi SZ , Kadir HA , Hassandarvish P , Tajik H , Abubakar S , Zandi K. A review on antibacterial, antiviral, and antifungal activity of curcumin . Biomed Res Int . 2014 ; 2014 : 186864 .
3. Bar-Sela G , Epelbaum R , Schaffer M . Curcumin as an anti-cancer agent: review of the gap between basic and clinical applications . Curr Med Chem . 2010 ; 17 ( 3 ): 190 - 197 .
4. Kumar G , Mittal S , Sak K , Tuli HS . Molecular mechanisms underlying chemopreventive potential of curcumin: current challenges and future perspectives . Life Sci . 2016 ; 148 : 313 - 328 .
5. Wegiel LA , Zhao Y , Mauer LJ , Edgar KJ , Taylor LS . Curcumin amorphous solid dispersions: the influence of intra and intermolecular bonding on physical stability . Pharm Dev Technol . 2014 ; 19 ( 8 ): 976 - 986 .
6. Anand P , Kunnumakkara AB , Newman RA , Aggarwal BB . Bioavailability of curcumin: problems and promises . Mol Pharm . 2007 ; 4 ( 6 ): 807 - 818 .
7. Liu W , Zhai Y , Heng X , et al. Oral bioavailability of curcumin: problems and advancements . J Drug Target . 2016 : 1 - 9 .
8. Hani U , Shivakumar HG . Solubility enhancement and delivery systems of curcumin a herbal medicine: a review . Curr Drug Deliv . 2014 ; 11 ( 6 ): 792 - 804 .
9. Naksuriya O , Okonogi S , Schiffelers RM , Hennink WE . Curcumin nanoformulations: a review of pharmaceutical properties and preclinical studies and clinical data related to cancer treatment . Biomaterials . 2014 ; 35 ( 10 ): 3365 - 3383 .
10. Siddiqui IA , Sanna V . Impact of nanotechnology on delivery of natural products for cancer prevention and therapy . Mol Nutr Food Res . 2016 ; 60 ( 6 ): 1330 - 1341 .
11. Mahmood K , Zia KM , Zuber M , Salman M , Anjum MN . Recent developments in curcumin and curcumin based polymeric materials for biomedical applications: a review . Int J Biol Macromol . 2015 ; 81 : 877 - 890 .
12. Dreis S , Rothweiler F , Michaelis M , Cinatl J Jr, Kreuter J , Langer K . Preparation, characterisation and maintenance of drug efficacy of doxorubicin-loaded human serum albumin (HSA) nanoparticles . Int J Pharm . 2007 ; 341 ( 1-2 ): 207 - 214 .
13. Quan Q , Xie J , Gao H , et al. HSA coated iron oxide nanoparticles as drug delivery vehicles for cancer therapy . Mol Pharm . 2011 ; 8 ( 5 ): 1669 - 1676 .
14. Stollenwerk MM , Pashkunova-Martic I , Kremser C , et al. Albuminbased nanoparticles as magnetic resonance contrast agents: I. Concept, first syntheses and characterisation . Histochem Cell Biol . 2010 ; 133 ( 4 ): 375 - 404 .
15. Fabi A , Giannarelli D , Malaguti P , et al. Prospective study on nanoparticle albumin-bound paclitaxel in advanced breast cancer: clinical results and biological observations in taxane-pretreated patients . Drug Des Devel Ther . 2015 ; 9 : 6177 - 6183 .
16. Weber C , Kreuter J , Langer K. Desolvation process and surface characteristics of HSA-nanoparticles . Int J Pharm . 2000 ; 196 ( 2 ): 197 - 200 .
17. Ulbrich K , Michaelis M , Rothweiler F , et al. Interaction of folateconjugated human serum albumin (HSA) nanoparticles with tumour cells . Int J Pharm . 2011 ; 406 ( 1-2 ): 128 - 134 .
18. Li C , Li Y , Gao Y , et al. Direct comparison of two albumin-based paclitaxel-loaded nanoparticle formulations: is the crosslinked version more advantageous? Int J Pharm . 2014 ; 468 ( 1-2 ): 15 - 25 .
19. Zhang L , Hou S , Mao S , Wei D , Song X , Lu Y. Uptake of folateconjugated albumin nanoparticles to the SKOV3 cells . Int J Pharm . 2004 ; 287 ( 1-2 ): 155 - 162 .
20. Zhao L , Du J , Duan Y , et al. Curcumin loaded mixed micelles composed of Pluronic P123 and F68: preparation, optimization and in vitro characterization . Colloids Surf B Biointerfaces . 2012 ; 97 : 101 - 108 .
21. Licciardi M , Paolino D , Celia C , Giammona G , Cavallaro G , Fresta M . Folate-targeted supramolecular vesicular aggregates based on polyaspartyl-hydrazide copolymers for the selective delivery of antitumoral drugs . Biomaterials . 2010 ; 31 ( 28 ): 7340 - 7354 .
22. Mehta AK , Yadav KS , Sawant KK . Nimodipine loaded PLGA nanoparticles: formulation optimization using factorial design, characterization and in vitro evaluation . Curr Drug Deliv . 2007 ; 4 ( 3 ): 185 - 193 .
23. Lu H , Li B , Kang Y , et al. Paclitaxel nanoparticle inhibits growth of ovarian cancer xenografts and enhances lymphatic targeting . Cancer Chemother Pharmacol . 2007 ; 59 ( 2 ): 175 - 181 .
24. Zhang XY , Lu WY . Recent advances in lymphatic targeted drug delivery system for tumor metastasis . Cancer Biol Med . 2014 ; 11 ( 4 ): 247 - 254 .