ARTICLE IN PRESS

0142-9612/$ - se

doi:10.1016/j.bi

�CorrespondEngineering, N

Singapore 1175

E-mail addr

Biomaterials 27 (2006) 3321–3332

www.elsevier.com/locate/biomaterials

Microparticles developed by electrohydrodynamic atomization for thelocal delivery of anticancer drug to treat C6 glioma in vitro

Jingwei Xiea, Jan C.M. Marijnissenb, Chi-Hwa Wanga,c,�

aDepartment of Chemical and Biomolecular Engineering, National University of Singapore, 4 Engineering Drive 4, Singapore 117576, SingaporebDepartment of Chemical Technology, Delft University of Technology, Particle Technology Group, Julianalaan 136, 2628 BL, Delft, The Netherlands

cMolecular Engineering of Biological and Chemical Systems, Singapore-MIT Alliance, Singapore 117576, Singapore

Received 25 October 2005; accepted 24 January 2006

Available online 21 February 2006

Abstract

This study aims to fabricate biodegradable polymeric particles by electrohydrodynamic atomization (EHDA) for applications in

sustained delivery of anticancer drug—paclitaxel to treat C6 glioma in vitro. Controllable morphologies such as spheres, donut shapes

and corrugated shapes with sizes from several tens of microns to hundred nanometers of particles were observed by scanning electron

microscopy (SEM) and field emission electron microscope (FSEM). The differential scanning calorimetry (DSC) study indicated that

paclitaxel could be either in an amorphous or disordered-crystalline phase of a molecular dispersion or a solid solution state in the

polymer matrix after fabrication. The X-ray photoelectron spectroscopy (XPS) result suggested that some amount of paclitaxel could

exist on the surface layer of the microparticles. The encapsulation efficiency was around 80% and more than 30 days in vitro sustained

release profile could be achieved. Cell cycling results suggested that paclitaxel after encapsulation by EHDA could keep its biological

function and inhibit C6 glioma cells in G2/M phase. The cytotoxicity of paclitaxel-loaded biodegradable microparticles to C6 glioma

cells could be higher than Taxols in the long-term in vitro tests evaluated by MTS assay. The drug delivery devices developed by EHDA

in this study could be promising for the local drug delivery to treat malignant glioma.

r 2006 Elsevier Ltd. All rights reserved.

Keywords: Electrohydrodynamic atomization; Biodegradable; Microparticle; Paclitaxel; Cytotoxicity

1. Introduction

Current treatments for malignant glioma sometimesremain unfavorable, and the duration of survival time ofpatients has made limited progress during the past years,despite a lot of advances in surgery, radiotherapy, adjuvantchemotherapy and imaging techniques [1]. The localizeddrug delivery device has many advantages to treatmalignant glioma. It can be implanted directly at thetumor site and the released drug can bypass the blood–brain barrier and thereby increase the concentration ofdrugs in the brain [2,3]. As a result, both systemic exposureand side effects of chemotherapy are minimized. A few

e front matter r 2006 Elsevier Ltd. All rights reserved.

omaterials.2006.01.034

ing author. Department of Chemical and Biomolecular

ational University of Singapore, 4 Engineering Drive 4,

76, Singapore. Tel.: +656874 5079; fax: +65 6779 1936.

ess: chewch@nus.edu.sg (C.-H. Wang).

types of localized drug delivery devices including wafers,discs and microparticles have been developed for the localtreatment of various tumors [4–12]. For wafers and discs,open surgery for implantation is always required fortreatment of malignant glioma. In contrast, microparticlescould be easily implanted by stereotaxy in discrete, precise,and functional areas of the brain without causing damageto the surrounding tissue [13]. Harper et al. reported thatpaclitaxel-loaded p(DAPG-EOP) microspheres are safeand potent for regional therapy of lung cancer tumornodules in mice [9]. Nsereko et al. studied paclitaxel-containing chitin and chintin-Pluronics F-108 microparti-cles and confirmed their use as biodegradable systems forlocalized administration in murine model of Lewis lungcarcinoma [10]. Attawia et al. reported that paclitaxel-loaded PMA-CPH microspheres can be used in conjunc-tion with irradiation for the treatment of Ewing’s sarcomato effectively release paclitaxel while maintaining its

ARTICLE IN PRESS

Fig. 1. (a) Diagram of electrohydrodynamic atomization experimental

setup; (b1): multiple-cone spraying (nozzle voltage: 10.0 kV; ring voltage:

0 kV); (b2): single-cone spraying (nozzle voltage: 10.0 kV; ring voltage:

9.0 kV). Operating parameters are the same as that of S2 sample. The

voltage at the grounded needle is taken as the reference point, 0V.

J. Xie et al. / Biomaterials 27 (2006) 3321–33323322

combined cytotoxic and radiosensitizing abilities [11]. In aseparate study where 9L glioblastoma cells were introducedinto the abdominal cavity of Wistar rats, Demetrick et al.found that the paclitaxel-loaded PLA microsphere for-mulation was more effective than conventional paclitaxel inpreventing tumor seeding [12].

There are many methods to fabricate microparticles suchas solvent evaporation [14], spray drying [15], spraying apolymer solution through a small orifice [16,17] andShirasu Porous Glass (SPG) membrane emulsificationtechnique [18], etc. However, most of these methods havesome disadvantages such as low encapsulation efficiency ortedious separation procedure of particles from aqueousphase.

The electrospray technique was widely employed tofabricate inorganic nanoparticles, thin films, fibers, hollowfibers, deposition of nanoparticle clusters, micro/nanoen-capsulation and production of pharmaceutical particles[19–25]. Ding et al. [26] reported an electrohydrodynamicatomization (EHDA) technique to fabricate uniformmicroparticles in the range of 1–15 mm. However, the yieldfor particles obtained at the collection station was notsatisfactory and the particle morphology was not con-trollable. Extending from the work by Ding et al. [26], themain objective of this study was to examine the morphol-ogy and characterization of EHDA particles obtainedunder various operating factors such as polymer solutionflow rate, polymer concentration, type of polymers, andtype of organic solvents. In performing these experiments,the present study aimed to fine tune the particle size rangeby systematic variation of these operating factors. Deliveryand cytotoxicity of paclitaxel to C6 glioma were evaluatedin vitro by downstream characterization through cellculture experiments. Cell cycling analysis and in vitrocytotoxicity were compared with the case of systemicdelivery where paclitaxel has considerable side effects and ashort lifetime in serum [27].

2. Materials and methods

2.1. Materials

Paclitaxel samples used in the present study were kindly offered by

Bristol-Myers Squibb (New Brunswick, NJ) at no cost. Taxols was

purchased from Bristol-Myers Squibb (Princeton, NJ). DCM (Cat. No.

DR-0440) and acetonitrile (Cat. No. AS-1122) were purchased from Tedia

Company Inc. (Fairfield, OH). Polymers such as poly(D,L-lactide-co-

glycolic acid) (PLGA) with L:G molar ratio of 50:50

(MW ¼ 90,000–120,000), poly(DL-lactide) (PDLA) (MW ¼ 106,000),

poly(L-lactide) (PLLA) (MW ¼ 85,000–160,000), polycaprolactone

(PCL) (MW ¼ 65,000) and polycaprolactone (PCL) (MW ¼ 14,000) were

purchased from Sigma Aldrich (St. Louis, MO USA). Ethylene/vinyl

acetate copolymer (EVAC) (Cat# 787) was purchased from SP Scientific

Polymer Products Inc. (NY, USA). Phosphate buffer saline (PBS) for in

vitro release study was bought from Sigma Aldrich, containing 0.1M

sodium phosphate, 0.15M sodium chloride, pH 7.4. Propidium iodide (PI,

Cat. No. P1304MP) was obtained from Molecular Probes Inc. (Eugene,

OR). CellTiter 96s AQueous One Solution Cell Proliferation Assay was

bought from Promega Corporation (Madison, WI). All other materials

and reagents used were of analytical grade.

2.2. Microparticle fabrication by electrospray

The experimental setup is shown in Fig. 1a. A well-defined potential

difference is created between a nozzle and a ring, by applying a high

voltage on the nozzle and a lower high voltage on the ring. Through the

nozzle a solution, containing dissolved polymer and drug, is pumped at a

low predetermined rate. At the nozzle tip a liquid cone is formed, with a

thin jet emerging from the apex. This jet breaks up in mono-dispersed

droplets [28,29]. The charged droplets need to be neutralized to make them

manageable and to avoid coulomb fission, due to solvent evaporation.

Neutralization is accomplished by a corona discharge system, which is

created at the tip of a sharp, grounded needle, positioned opposite the

high-voltage ring and needle. Fig. 1 shows both spraying in a single-jet

mode (Fig. 1b2) and in a multiple-jet mode (Fig. 1b1). Going from the

cone-jet mode to the multiple-jet mode is accomplished by increasing the

potential difference between the nozzle and the ring. This is done here by

keeping the potential of the nozzle constant and decreasing the potential

of the ring.

2.3. Scanning electron microscopy (SEM)

The morphology of microparticles observed by SEM (Jeol JSM

5600LV) required an ion coating with platinum by a sputter coater

(JFC-1300, Jeol, Tokyo) for 40 s in a vacuum at a current intensity of

40mA after mounting the sample on metallic studs with double-sided

conductive tape. The accelerating voltage ranged from 10 to 15 kV during

scanning.

2.4. X-ray photoelectron spectroscopy (XPS)

The surface chemistry of the microspheres was determined by AXIS

His X-ray photoelectron spectroscopy (Kratos Analytical Inc., NY) using

the curve fitting software provided by the manufacture. For all samples, a

survey spectrum was recorded over a binding energy range of 0–1100 eV

using a pass energy of 80 eV. In all cases, the survey spectra recorded the

presence of oxygen (O1s 533 eV), carbon (C1s 285 eV) and nitrogen (N1s

399 eV) at the surface.

ARTICLE IN PRESSJ. Xie et al. / Biomaterials 27 (2006) 3321–3332 3323

2.5. Encapsulation efficiency and in vitro release

Encapsulation efficiency (EE) is defined as

EE;% ¼Drug loaded in particles

initial amount of drug loaded� 100%. (1)

To determine encapsulation efficiency, 3mg of paclitaxel-loaded

microparticles were dissolved in 1ml DCM. Subsequently, 9ml mobile

phase (acetonitrile:water ¼ 1:1) was added and DCM was made to

evaporate completely. The resulting solution was analyzed by HPLC on

the condition mentioned below. 10mg paclitaxel-loaded microparticles

were dispersed in 50ml PBS, pH ¼ 7.4, which contained 1% of Tween 80

to enhance the paclitaxel solubility. The buffer solution was kept constant

at 37.2 1C. At given time intervals, three tubes of each formulation of

microparticles were withdrawn and centrifuged at 10 000 rpm (equivalent

to 1250g) for 20min. The precipitated microparticles were taken and

resuspended in 50ml of fresh release medium and placed back in the

shaker. The supernatant solution was kept for high-performance liquid

chromatography (HPLC) analysis. Paclitaxel in the release medium was

firstly extracted with 1ml of DCM. Subsequently, 5ml mobile phase

(acetonitrile:water ¼ 1:1) was added and DCM was made to evaporate

completely. The resulting solution was analyzed by HPLC, in which C-18

Column was used and the mobile phase was delivered at a rate of 1ml/min.

A volume of 50 ml of sample was injected with an auto-sampler and the

column effluent was detected at 227 nm with an ultraviolet (UV) detector.

2.6. Cell culture

C6 glioma cells (courtesy, Singapore General Hospital, passages 6–10)

were grown and routinely maintained at 37 1C in 75 cm2 culture flasks, in

Dulbecco’s Modified Eagles Medium supplemented with 10% fetal calf

serum, 1% penicillin–streptomycin in atmosphere of 5% CO2 and 90%

relative humidity. The cells were harvested with trypsin-EDTA. Medium

was changed every other day.

2.7. Microparticle sterilization

Gamma irradiation of the samples for sterilization was done using 60Co

with a total dose of 15 kGy.

2.8. Cell cycling analysis

C6 glioma cells were cultured in T25 culture flasks. When the cells

reached 80% confluence, medium contained different concentrations of

Taxols and paclitaxel-loaded PLGA microparticles was changed. At

different intervals, medium was taken out and culture flasks were washed

with ice cold PBS three times to remove microparticles. The cells were

harvested by trypsin and fixed by ethanol. Then, the cells were stained with

PI. Finally, the cell solutions were measured by flow cytometry for

excitation with blue light and detection of PI emission at red wavelengths.

2.9. Microparticle cytotoxicity

The cell viability was assessed by the MTS-CellTiter 96s aqueous non-

radioactive cell proliferation assay (Promega, Madison, WI). Cells were

transferred to 96-well plate first to ensure 5� 103 cells per well. Medium

was changed until 80% confluence was reached. Then the medium was

changed with 100ml medium with Taxols and paclitaxel-loaded micro-

particles with different concentrations. The plate was incubated with

different intervals. One row of 96-well plates was taken as control without

adding microparticles. At different intervals, suspension was removed and

the wells were washed 3 times by PBS. Then 200ml of medium and 20 ml ofMTS assay were added to the wells. After incubation for around 3–4 h, the

plate was observed by microplate reader (GENios, Groedig, Austria) at

492nm. Cell viability was determined by

Cell viability ð%Þ ¼ ðAbstest cells=Abscontrol cellsÞ � 100, (2)

where Abstest cells and Abscontrol cells represent the amount of formazan

determined for cells treated with different formulations and for control

cells (non-treated), respectively.

3. Results and discussion

3.1. Particle fabrication

3.1.1. Polymer concentration

Fig. 2 illustrates that different polymer concentrationscould affect the particle size and morphology. It isobserved that the particle size could decrease withdecreasing polymer concentration. The surface of PCLmicroparticles may change from smooth surface to roughcorrugated surface when polymer solution concentrationvaries from 6% to 0.5%. The PCL microparticles couldchange from sphere to biconcave to irregular shapes withdecreasing polymer solution concentration. In addition,when the polymer solution flow rate was low enough,nanoparticles of around 300 and 600 nm could be obtainedwhen the polymer concentrations were 2% and 8%,respectively (shown in Fig. 2(e) and (f)).

3.1.2. Polymer solution flow rate

It is found that the particle size can be fine-tuned withvariation of polymer solution flow rate. The particle sizecould decrease from 11.170.8, 6.570.8 to 4.970.8 mmwith decreasing polymer solution flow rate from 3, 1 to0.5ml/h when the polymer concentration was maintainedat 3%. The particle size may increase from about 17 to30 mm with increasing polymer flow rate from 3 to 15ml/h[shown in Figs. 3(a)–(c)]. In addition, the particlemorphology seemed to be smoother when the polymersolution flow rate was high. Fig. 3(b) and (c) suggest that partof organic solvents could not evaporate completely when thepolymer solution flow rate was high, which may be due to theshort residence time of the particles caused by the shortvertical dimension of our setup, limiting the evaporation ofthe organic solvent. From the SEM micrograph, in additionto discrete particles, a corrugated polymer film was observedon the surface of the collector. This is because some of theparticles were not dry, and as they hit the collector, wetparticles deformed and a film is formed.In the cone-jet model, Hartman et al. described the

relationship between droplet size and various variables bythe following scaling law [28,29]. Although for the lowconductivity liquids used here, a more complicated currentscaling law was applicable [30], we used here the simplerversion to obtain an indication of the variation of dropletsize with changing parameters:

I / ðgKQÞ1=2, (3)

d ¼ cre0Q4

I2

� �1=6

/ Q1=2, (4)

ARTICLE IN PRESS

Fig. 3. SEM images of microparticles fabricated under different polymer solution flow rates: (a) 3ml/h particle size 17 mm; (b) 10ml/h particle size 26mm;

(c) 15ml/h particle size 32mm. Operating parameters: PCL, polymer air flow rate 20 l/min, needle size 0.91mm, polymer concentration 6%, voltages of

nozzle and ring 8.8 and 7.1 kV. The voltage at the grounded needle is taken as the reference point, 0V.

Fig. 2. SEM images of microparticles fabricated under different polymer concentrations: (a) 6%; (b) 3%; (c) 1%; (d) 0.5%. Operating parameters: PCL

polymer, air flow rate 20 l/min, polymer solution flow rate 3ml/h, nozzle size 0.91mm, voltages of nozzle and ring 8.8 and 7.1 kV. (e) 8%; f: 2%. Operating

parameters: PLGA polymer, air flow rate 25 l/min, polymer solution flow rate 0.1ml/h, nozzle size 0.41mm, voltages of nozzle and ring 14.0 and 12.0 kV.

The voltage at the grounded needle is taken as the reference point, 0V.

J. Xie et al. / Biomaterials 27 (2006) 3321–33323324

where d is the droplet diameter (m), c is a constant, Q is theliquid flow rate (m3 s�1), r is the liquid density, g is theliquid surface tension, K is the liquid conductivity, I is thecurrent and e0 is the permittivity in vacuum.

This equation suggested that the droplet size decreaseswith deceasing flow rate of polymer solution. In the presentstudy, the size of microparticles was found to decrease withdecreasing polymer solution flow rate. Since the particlesize has a direct relationship with the droplet size at a fixed

polymer concentration, our experimental results seem to bequantitatively consistent with the prediction by Eq. (4).

3.1.3. Voltages difference between nozzle and ring

Ding et al. [26] reported that with the same liquid flowrate and system geometry, applied voltage differencesbetween spray nozzle and ring play key roles in determin-ing the spraying pattern. This has also been well explainedin available literature reports and not reproduced here [31].

ARTICLE IN PRESSJ. Xie et al. / Biomaterials 27 (2006) 3321–3332 3325

For all data of voltages applied to the nozzle and ringshown in the captions of various figures throughout thiswork, the voltage at the grounded needle is taken as thereference point, 0V.

3.1.4. Polymers of different molecular weights

Under the same operating conditions, similar particlesize of about 10 mm microparticles could be obtained usingdifferent molecular weights of PCL (data not shown). Forlower molecular weight PCL, there was a hole on thesurface of the microparticle. The diameter of the hole wasaround 1 mm which may be due to the difference of polymermechanical properties and the reason for such formation isto be further investigated.

3.1.5. Different organic solvents

Under the same operating conditions, it is seen from Fig.4(a)–(c) that the particle size could be smaller whenacetonitrile was used as the organic solvent. This couldbe due to the difference in dielectric constants (orequivalently conductivities) between DCM (9.1 at 20 1C)and acetonitrile (37.5 at 20 1C). Due to the similar values of

Fig. 4. SEM images microparticles fabricated under different organic solvents.

air flow rate 20 l/min, polymer concentration 3%, polymer solution flow rate

Fabrication under different polymers. ((d) EVAC; (e) PLGA; (f) PDLA). O

polymer solution flow rate 3ml/h, needle size 0.34mm, voltages of nozzle and

reference point, 0V.

dielectric constants (or conductivities) of DCM and THF(7.6 at 20 1C), similar particle size could be obtained whenusing THF. The literature reported that organic solventswith high dielectric constants usually also have higherconductivity [32]. In this study, DCM (28.1 dyn/cm),acetonitrile (29.1 dyn/cm) and THF (28 dyn/cm) have thesimilar value of surface tension at 20 1C [33]. In ourexperiments, the measured conductivities of DCM(0.000275 mS/cm), THF (0.000745 mS/cm) and acetonitrile(0.0710 mS/cm) were collected using YSI 3200 conductivityinstrument (YSI Inc., Yellow Springs, OH, USA). Thedifferences of particle size may be caused by the differenceof their conductivities, which could be explained by thescaling laws shown in Eqs. (3) and (4) [29], which can berewritten for d as

d /1

K

� �1=6

. (5)

In this study, the size of microparticles decreases withincreasing conductivity of polymer solution. Therefore, ourresults are qualitatively consistent with the prediction bythe scaling laws.

((a) DCM; (b) acetonitrile; (c) THF). Operating parameters: polymer PCL,

3ml/h, needle size 0.34mm, voltages of nozzle and ring 8.8 and 7.1 kV.

perating parameters: air flow rate 20 l/min, polymer concentration 3%,

ring 8.8 and 7.1 kV. The voltage at the grounded needle is taken as the

ARTICLE IN PRESSJ. Xie et al. / Biomaterials 27 (2006) 3321–33323326

3.1.6. Different biodegradable polymers

In this study, different biodegradable polymers such asPCL, PLGA, PDLA and PLLA and non-biodegradablepolymers such as EVAC were employed to fabricatemicroparticles by EHDA and the results were compared.Figs. 4(d)–(f) show the SEM images of the microparticlesfabricated by the different polymers. Under the sameoperating conditions, various polymeric microparticleswith different morphologies and similar particle size couldbe obtained.

Fig. 5 shows SEM images of a number of samplesmanufactured in this study. This figure suggests that for allthe samples shown, the neutralization is sufficient to avoidRayleigh disintegration during evaporation, because inthe case of insufficient neutralization we would expect toalso see smaller particles. So after electro-spraying thesolution, the droplets are (partly) neutralized and dried toform the particles shown in the figure. For differentconcentrations, different morphologies of particles couldbe obtained, which may be due to different evaporationrates of the organic solvents. The optimized operatingconditions are summarized in Table 1. Therefore, con-

Fig. 5. SEM images of microparticles for sample

Table 1

Optimized EHDA fabrication parameters and their effects on particle size and

C Qa Qp Vn

Size Large High — High Low

Small Low — Low High

Morphology Smooth High Low High High

Rough Low High Low Low

This table provides empirical operating parameters in controlling particle size.

values to obtain larger size of particles.

C: polymer concentration; Qa: air flow rate; Qp: polymer solution flow rate;

crystallinity of polymers; K: conductivity; —: not much effect; high: high valu

trollable biodegradable particle morphology and sizecould be realized in this study. The operating conditionsfor the particle fabrication are summarized in Table 1. Byadjusting the parameters listed in the table, controllablesize and morphology of particles could be obtained. Fig. 5shows the SEM images of the typical samples fabricated inthis study. From Fig. 5 it is seen that PCL microspheres,PLGA microspheres and biconcave-shaped PLGA micro-particles of the similar size (shown in Table 2) could beobtained.

3.2. Differential scanning calorimetry (DSC) analysis

Fig. 6 shows the DSC thermogram analysis, whichprovided qualitative and quantitative information aboutthe physical status of the drug in the microparticles. Thedrug loadings of the samples in DSC experiments are about8% and 16%. The pure paclitaxel showed an endothermicpeak of melting at 223.0 1C. There was no peak observed atthe temperature range of 150–250 1C for the samples. TheDSC study did not detect any crystalline drug material inthe microparticle samples. It could thus be concluded that

s S1–S5. (a) S1; (b) S2; (c) S3; (d) S4; (e) S5.

morphology

MW Cr K S

— — Low Low dielectric constant

— — High High dielectric constant

High High crystallinity — High dielectric constant

Low Low crystallinity — Low dielectric constant

For example, one needs to use high C and Op values and low Vn, K and S

MW: polymers of different molecular weights; S: organic solvents; Cr:

e; low: low value.

ARTICLE IN PRESS

Table 4

XPS analysis for characteristic samples

Sample Element Atomic conc % Atomic mass conc %

Pure Paclitaxel C 79.3 74.3

N 1.4 1.6

O 19.3 24.1

J. Xie et al. / Biomaterials 27 (2006) 3321–3332 3327

the paclitaxel formulated in the samples was in anamorphous or disordered-crystalline phase of a moleculardispersion or a solid solution state in the polymer matrixafter fabrication. Moreover, the glass transition tempera-ture of the polymers employed in paclitaxel-loaded PLGAand PCL microparticles was not influenced obviously bythe EHDA process (Table 3).

-40

-30

-20

-10

0

10

20

0 50 100 150 200 250Temperature (°C)

mW

Paclitaxel

PCL microparticles

S1

S5

S2

S3

S4

Fig. 6. DSC diagram of different characteristic samples.

Table 3

Operating parameters for the fabrication of samples S1–S5

Samples Operating parameters

Polymer Polymer

concentration (%)

Polymer solution

flow rate (ml/h)

Air flow rate

(l/min)

Voltage at nozzle

(kV)

Voltage at ring

(kV)

Nozzle size

(mm)

S1 PCL 6 3 25 9.8 7.1 0.91

S2 PLGA 8 5 25 10.0 9.0 0.91

S3 PLGA 6 3 25 9.8 7.1 0.91

S4 PLGA 8 5 25 10.0 9.0 0.91

S5 PLGA 8 5 25 10.0 9.0 0.91

The voltage at the grounded needle is taken as the reference point, 0V.

Table 2

Characterization for samples S1–S5

Samples Drug loading

(%)

Encapsulation

efficiency (%)

Particle size (mm)7SD

S1 8.1 81.3 11.470.9

S2 7.9 82.3 15.271.7

S3 8.4 84.1 14.272.2

S4 15.8 78.1 15.170.7

S5 0 — 12.670.8

S1—Paclitaxel-loaded PCL microspheres; S2—Paclitaxel-loaded PLGA

microspheres; S3—Paclitaxel-loaded PLGA particles of irregular shape;

S4—Paclitaxel-loaded PLGA microspheres; S5—Blank PLGA micro-

spheres.

PCL microparticles C 23.9 29.5

N 0.0 0.0

O 76.1 70.5

S1 C 22.4 27.7

N 0.7 0.8

O 76.9 71.5

S2 C 31.9 38.4

N 0.6 0.6

O 67.5 61.0

S4 C 29.6 35.8

N 0.6 0.7

O 69.8 63.5

S5 C 34.2 40.9

N 0.0 0.0

O 65.8 59.1

ARTICLE IN PRESS

0

10

20

30

40

50

60

70

80

0 5 10 15 20 25 30 35

Time (day)

Cu

mu

lati

ve r

ele

ase (

%)

S1

S2

S3

S4

Fig. 7. In vitro release curves for samples S1–S4.

Fig. 8. Cell cycling analysis treated by Taxols after 24 h. The forward scatter (F

charts) indicate cells.

J. Xie et al. / Biomaterials 27 (2006) 3321–33323328

3.3. Surface chemistry XPS analysis

In order to analyze the distribution of paclitaxel inthe microparticles, XPS analysis was carried out toexamine C, N and O element compositions. FromTable 4, N was detected in S1, S2 and S4 samples andthis finding suggests that some amount of paclitaxel existson the surface layer of the microparticles because XPScould detect the depth of several nanometers of polymermaterials, which could explain the initial burst for the invitro release.

3.4. Encapsulation efficiency and in vitro release

The encapsulation efficiency and drug loading of eachsample are listed in Table 2. It is seen from the table that

SC) measurement is related to cell size; events (y-axis of the DNA contents

ARTICLE IN PRESS

Fig. 9. Cell cycling analysis treated by microparticles after 24 h. The forward scatter (FSC) measurement is related to cell size; events (y-axis of the DNA

contents charts) indicate cells.

J. Xie et al. / Biomaterials 27 (2006) 3321–3332 3329

the encapsulation efficiency of all samples was around 80%and drug loadings are about 8% and 16% for theoreticaldrug loadings of 10% and 20%, respectively. The releaseprofiles of different morphologies of paclitaxel-loadedPLLA microparticles were investigated [34]. Fig. 7 showsthe in vitro release profile of the samples in this study. Theinitial burst could be attributed to some amount ofpaclitaxel on the surface layer of the microparticle whichwas proved by XPS results. A slower, zero-order phase ofrelease was followed. From Fig. 7, it is seen that paclitaxel-loaded PLGA microparticles could release faster than

paclitaxel-loaded PCL microparticles and biconcave-shaped PLGA microparticles may release slightly fasterthan PLGA microspheres. In addition, 16% drugloading paclitaxel-loaded PLGA microparticles couldrelease slightly faster than 8% drug loading samples. It isobserved that the total amount of paclitaxel releasedseemed to be less than 60% of the total amount of drug inthe microparticles in the defined time. Therefore, theremained paclitaxel within the polymer matrix would bereleased slowly until the polymer matrix was completelyeroded.

ARTICLE IN PRESSJ. Xie et al. / Biomaterials 27 (2006) 3321–33323330

3.5. Cell cycling analysis

Paclitaxel binds to the microtubules and suppresses thespindle-microtubule dynamics and prevents the dividingcancer cells from metaphase into anaphase and the cells

Table 5

Cell cycling analysis

G0/G1 (%) S (%) G2/M (%)

Control 69.5 15.8 14.7

S5 250mg/ml 56.2 17.4 26.4

S5 500mg/ml 60.1 14.9 25.0

Taxols 2.5 mg/ml 76.1 10.1 13.8

Taxols 5.0 mg/ml 26.6 12.8 61.6

Taxols 10.0mg/ml 18.2 2.3 79.5

Taxols 20.0mg/ml 38.0 1.9 60.1

S2 125mg/ml 81.0 12.5 6.5

S2 250mg/ml 77.5 14.1 8.4

S2 500mg/ml 5.7 3.3 91.0

S2 500mg/ml (48 h)a 13.4 9.1 78.2

S2 1000mg/ml 7.8 2.9 89.3

aAll samples were tested after 24 h exposure except for this sample

which was tested after 48 h.

Fig. 10. Representative optical images of C6 glioma cells after being treated by

5 days; (d–f) 1250mg/ml—1, 2, 5days; (g–i) 2000mg/ml—1, 2, 5 days.

eventually die by apoptosis [35,36]. From Fig. 8 it wasobserved that Taxols-treated C6 glioma cells wereaccumulated in G2/M (61.6%) phase beginning withconcentrations as low as 5 mg/ml. In addition, flowcytometry showed that the sharp peak of G0/G1 cells wasmarkedly attenuated at the paclitaxel concentration of5 mg/ml (26.6%) compared to 2.5 mg/ml (76.1%). However,at higher concentrations, e.g. 20 mg/ml, two discrete cellpeaks could be discerned. This may have been either due tolysis and release of nuclei from ruptured multinucleatedcells or additional cell cycle restrictions, such as late G1 orG1/S boundary [37]. Fig. 9 shows the cell cycling analysisafter the microparticle treatment, which indicated thatblank PLGA microparticle had almost no interaction withthe cells. When the concentration of paclitaxel-loadedPLGA microparticles was below 250 mg/ml, the cells couldnot be accumulated in G2/M phase after 24 h, which maybe due to very little amount of drug released out from themicroparticles. However, when the concentrations ofpaclitaxel-loaded PLGA microparticles were 500 and1000 mg/ml, most of the living cells were accumulated inG2/M phases. The percentage of the living cells in G1/G0, Sand G2/M phases after treatment of different formulations

paclitaxel-loaded PLGA microparticles ( 50 mm). (a–c) 250mg/ml—1, 2,

ARTICLE IN PRESS

0%

20%

40%

60%

80%

100%

120%

0 20 40 60 80 100 120 140 160

Taxol concentration (µg/ml)

Cell v

iab

ilit

y (

%)

1day

2days

3days

4days

5days

0%

20%

40%

60%

80%

100%

120%

0 250 500 750 1000 1250 1500 1750 2000 2250 2500 2750

Paclitaxel-loaded PLGAmicroparticle concentration (µg/ml)

Cell v

iab

ilit

y (

%)

1day

2days

3days

4days

5days

(a)

(b)

Fig. 11. Cell viability (a) Taxols; (b) paclitaxel-loaded PLGA micro-

particles (sample S2).

J. Xie et al. / Biomaterials 27 (2006) 3321–3332 3331

was analyzed (shown in Table 5). Flow cytometry analysisshowed that cells were mostly in G2/M phase but not in Sphase, which was accordance to increased DNA turnoverdue to cell cycle restriction imposed by paclitaxel. Fromthis study, the drug sensitivity could be obtained, whichmay provide useful information for the cell viabilityexperiment. The other important information we couldget was that the bioactivity of anticancer drug paclitaxelcould not change after encapsulation in biodegradablepolymers in the electrospray process.

3.6. Microparticle in vitro cytotoxicity

Under optical microscopy, it is clearly seen that thenumber of cells treated by Taxols and paclitaxel-loadedmicroparticles decreased drastically with increasing incu-bation time and increasing paclitaxel concentration (shownin Fig. 10). The morphology of Taxols-treated andpaclitaxel-loaded microparticles treated cells becamerounder at 24 h and subsequently appeared to be flatter.Marked hydropic changes and vacuolated appearanceswere observed clearly from Taxols-treated cells (data notshown). These effects were less pronounced at lower dosesof the drug and cells treated with 20 mg/ml paclitaxel werenot morphologically different from untreated control cellsafter 24 h.

The cytotoxic activity of Taxols and paclitaxel-loadedPLGA microparticles fabricated in this study (shown inFig. 11) were evaluated by assessing cell viability using theMTS assay against C6 glioma cells. Cells were incubatedwith microparticles which contained the same amount ofdrug as that of Taxols sample in which paclitaxelconcentrations were in the ranged from 20 to 160 mg/ml.The concentrations selected here correspond to plasmalevels of the drug achievable in humans [38,39]. Fig. 11(a)shows the cytotoxicity of Taxols of different concentra-tions, which suggests that the cell viability could decreasewith increasing Taxols concentration and incubation time.It is seen that the cell viability could verify little when theTaxols concentration ranges from 80 to 160 mg/ml. Also,IC50 values of 70, 50, 30, 18 and15 mg/ml for Taxols aftertreatment of 1, 2, 3, 4 and 5 days could be obtained. Cellviability was also evaluated with different concentrationsof particles and it was found that higher concentration inthe range from 250 to 2000 mg/ml could reduce much morecell viability (shown in Fig. 11(b)). Cell viability could notreach 50% after 3-day treatment by the microparticles ofthe concentrations in this study. However, IC50 values of300 and 200 mg/ml for microparticles (S2) could beachieved after treatment of 4 days and 5 days, respectively(shown in Fig. 11(b)). According to the actual amount ofdrug (around 22%) released from the microparticles(shown in Fig. 7), the microparticles seem to be morecytotoxicity than Taxols after 4–5 days. The antitumoreffect of paclitaxel is dependent upon whether sustainedconcentrations of the drug are higher than therapeutic levelrather than the maximal plasma concentrations can be

reached because paclitaxel toxicity requires entry of cellsinto the M phase [40]. It was reported that increasing thetime of paclitaxel exposure results in increasing paclitaxelcytotoxicity [41]. Therefore, the local drug delivery systemdeveloped in this study may have its advantage oncytotoxicity as compared to Taxols because of thesustained release property shown in Fig. 7.

4. Conclusions

In this study, polymeric particles with controllablemorphology and uniform size were successfully developedfor drug delivery applications by a modified EHDA setupunder different operating conditions. The encapsulationefficiency was around 80% and more than 30-day in vitrorelease was achieved. Cell cycling results suggested thatpaclitaxel after encapsulation by EHDA could keep itsbiological function and inhibit most of C6 glioma cells inthe G2/M phase. The cytotoxicity of paclitaxel-loadedbiodegradable microparticles to C6 glioma cells is higherthan Taxols after 4–5 days treatment because of theirsustained release properties. The drug delivery devices

ARTICLE IN PRESSJ. Xie et al. / Biomaterials 27 (2006) 3321–33323332

developed by EHDA in this study could be promising forthe local drug delivery to treat malignant glioma and willbe tested in vivo in future work.

Acknowledgments

This work is supported by the National University ofSingapore under Grant No. R279-000-095-112. Theauthors thank Lim Liang Kuang for many helpfuldiscussions on the EHDA particle fabrication.

References

[1] Kornblith PL, Walker M. Chemotherapy for malignant gliomas.

J Neurosurg 1998;68:1–17.

[2] Fung LK, Saltzman WM. Polymeric implants for cancer chemother-

apy. Adv Drug Deliv Rev 1997;26:209–30.

[3] Brem H. Polymers to treat brain tumors. Biomaterials 1990;11:

699–701.

[4] Langer R. Polymer implants for drug delivery in the brain. J Control

Rel 1991;16:53–60.

[5] Walter KA, Cahan MA, Gur A, Tyler B, Hilton J, Colvin OM, et al.

Interstitial Taxol delivered from a biodegradable polymer implant

against experimental malignant glioma. Cancer Res 1994;54:2207–12.

[6] Brem H, Garbikian P. Biodegradable polymer implants to treat brain

tumors. J Control Rel 2001;74:63–7.

[7] Seong H, An TK, Khang G, Choi SU, Lee CO, Lee HB. BCNU-

loaded poly (D, L-lactide-co-glycolide) wafer and antitumor against

XF-498 human CNS tumor cells in vitro. Int J Pharm 2003;251:1–12.

[8] Li Y, Owusu A, Lehnert S. Treatment of intracranial rat glioma

model with implant of radiosensitizer and biomodulator drug

combined with external beam radiotherapy. Int J Radiat Oncol Biol

Phys 2004;58(2):519–27.

[9] Harper E, Dang W, Lapidus RG, Garver RI. Enhanced efficacy of a

novel controlled release paclitaxel formulation (PACLIMER delivery

system) for local-regional therapy of lung cancer tumor nodules in

mice. Clin Cancer Res 1999;5:4242–8.

[10] Nsereko S, Amiji M. Localized delivery of paclitaxel in solid tumors

from biodegradable chitin microparticle formulations. Biomaterials

2002;23:2723–31.

[11] Attawia MA, Borden MD, Herbert KM, Katti DS, Asrari F, Uhrich

KE, et al. Regional drug delivery with radiation for the treatment of

Ewing’s sarcoma. In vitro release development of a taxol release

system. J Control Rel 2001;71:193–202.

[12] Demetrick JS, Liggins RT, Machan L, Davis NL, Burt HM, Hunter

WL. The development of a novel intraperitoneal tumor-seeding

prophylactic. Am J Surg 1997;173:403–6.

[13] Menei P, Benoit JP, Boisdron-Celle M, Fournier D, Mercier P, Guy

G. Drug targeting into the central nervous system by stereotactic

implantation of biodegradable microspheres. Neurosurgery 1994;

34(6):1058–64.

[14] Rosca ID, Watari F, Uo M. Microparticle formation and its

mechanism in single and double emulsion solvent evaporation. J

Control Rel 2004;99:271–80.

[15] Mu L, Feng SS. Fabrication, characterization and in vitro release of

paclitaxel (Taxols) loaded poly (lactic-co-glycolic acid) microspheres

prepared by spray drying technique with lipid/cholesterol emulsifiers.

J Control Rel 2001;76:239–54.

[16] Berkland C, Kim C, Pack DW. Fabrication of PLG microspheres

with precisely controlled and monodisperse size distributions. J

Control Rel 2001;73:59–74.

[17] Berkland C, King M, Cox A, Kim K, Pack DW. Precise control of

PLG microsphere size provides enhanced control of drug release rate.

J Control Rel 2002;82:137–47.

[18] Ma G, Nagai M, Omi S. Preparation of uniform poly(lacide)

microspheres by employing the Shirasu Porous Glass (SPG)

emulsifying technique. Colloids Surf A: Physiochem Eng Aspects

1999;153:383–94.

[19] Okuyama K, Lenggoro IW. Preparation of nanoparticles via spray

route. Chem Eng Sci 2003;58:537–47.

[20] Buchko GJ, Chen LC, Shen Y, Martin DC. Processing and

microstructured characterization of porous biocompatible protein

polymer thin films. Polymer 1999;40:7397–407.

[21] Berkland C, Pack DW, Kim K. Controlling surface nano-structure

using flow-limited field-injection electrostatic spraying (FFESS) of

poly (D, L-lactide-co-glycolide). Biomaterials 2004;25:5649–58.

[22] Loscertales IG, Barrero A, Marquez M, Spretz R, Velarde-Ortiz R,

Larsen G. Electrically forced coaxial nanojets for one-step hollow

nanofiber design. J Am Chem Soc 2004;126:5376–7.

[23] Jayasinghe SN, Edirisinghe MJ, Zhang DZ. Controlled deposition of

nanoparticle clusters by electrohydrodynamic atomization. Nano-

technology 2004;15:1519–23.

[24] Loscertales IG, Barrero A, Guerrero I, Cortijo R, Marquez M,

Ganan-Calvo AM. Micro/nano encapsulation via electrified coaxial

liquid jets. Science 2002;295:1695–8.

[25] Ijsebaert JC, Geerse KB, Marijnissen JCM, Lammers JJ, Zanen P.

Electro-hydrodynamic atomization of drug solutions for inhalation

purposes. J Appl Physiol 2001;91:2735–41.

[26] Ding LN, Lee T, Wang CH. Fabrication of monodispersed Taxol-

loaded particles using electrohydrodynamic atomization. J Control

Rel 2005;102(2):395–413.

[27] Singla AK, Garg A, Aggarwal D. Paclitaxel and its formulations. Int

J Pharm 2002;35:179–92.

[28] Hartman RPA, Brunner DJ, Camelot DMA, Marijnissen JCM,

Scarlett B. Jet break-up electro dynamic atomization in the cone jet

mode. J Aerosol Sci 2000;31:65–95.

[29] Hartman RPA, Brunner DJ, Camelot DMA, Marijnissen JCM,

Scarlett B. Electrohydrodynamic atomization in the cone-jet mode

physical modeling of the liquid cone and jet. J Aerosol Sci

1999;30(7):823–49.

[30] Hartman RPA, Electrohydrodynamic atomization in the cone-jet

mode, 1998, Ph.D. dissertation, Department of Chemical Technol-

ogy, Delft University of Technology.

[31] Duft D, Lebius H, Huber BA, Guet C, Leisner T. Shape oscillations

and stability of charged microdroplets. Phys Rev Lett 2002;89(8)

084503-1–2.

[32] Castellanos A. Electrohydrodynamics. New. Berlin: Springer; 1998.

[33] Smallwood IM. Handbook of organic solvent properties. London:

Arnold; 1996.

[34] Liggins RT, Burt HM. Paclitaxel loaded poly (L-lactide acid) (PLLA)

microspheres II. The effect of processing parameters on microsphere

morphology and drug release kinetics. Int J Pharm 2004;281:103–6.

[35] Jordan MA, Wilson L. Microtubules as a target for anticancer drugs.

Nat Rev Cancer 2004;4:253–65.

[36] Orr GA, Verdier-Pinard P, Mcdaid H, Horwitz SB. Mechanisms of

Taxol resistance related to microtubules. Oncogene 2003;22:7280–95.

[37] Gagandeep S, Novikoff PM, Ott M, Gupta S. Paclitaxel shows

cytotoxic activity in human hepatocellular carcinoma cell lines.

Cancer Lett 1999;136:109–18.

[38] Fonseca C, Simoes S, Gaspar R. Paclitaxel-loaded PLGA nanopar-

ticles: preparation, physicochemical characterization and in vitro

anti-tumoral activity. J Control Rel 2002;83:273–86.

[39] Raymond E, Hanauske A, Faivre S. Effects of prolonged versus

short-term exposure paclitaxel (Taxols) on human tumor colony-

forming units. Anti-Cancer Drugs 1997;8(4):379–85.

[40] Gagandeep S, Novikoff PM, Ott M, Gupta S. Paclitaxel shows

cytotoxic activity in human hepatocellular carcinoma cell lines.

Cancer Lett 1999;136:109–18.

[41] Liebmann JE, Cook JA, Lipschultz C, Teague D, Fisher J, Mitchell

JB. Cytotoxic studies of paclitaxel (Taxols) in human tumour cell

lines. Br J Cancer 1993;68:1104–9.