In vitro cytotoxicity, hemolysis assay, and biodegradationbehavior of biodegradable poly(3-hydroxybutyrate)–poly(ethylene glycol)–poly(3-hydroxybutyrate)nanoparticles as potential drug carriers

Cheng Chen,1,2 Yin Chung Cheng,1 Chung Him Yu,1 Shun Wan Chan,1 Man Ken Cheung,1 Peter H. F. Yu1

1State Key Laboratory of Chinese Medicine and Molecular Pharmacology, Department of Applied Biology andChemical Technology, The Hong Kong Polytechnic University, Hong Kong, People’s Republic of China2Changchun Institute of Applied Chemistry, Chinese Academy of Sciences, Changchun 130022,People’s Republic of China

Received 24 January 2007; revised 22 May 2007; accepted 31 July 2007Published online 7 January 2008 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jbm.a.31719

Abstract: Nanoparticles based on amorphous poly(3-hydroxybutyrate)–poly(ethylene glycol)–poly(3-hydroxybuty-rate) (PHB-PEG-PHB) are potential drug delivery vehicles,and so their cytotoxicity and hemolysis assay were investi-gated in vitro using two kinds of animal cells. The PHB-PEG-PHB nanoparticles showed excellent biocompatibility and hadno cytotoxicity on animal cells, even when the concentrationsof the PHB-PEG-PHB nanoparticle dispersions were increasedto 120 lg/mL. Moreover, no hemolysis was detected with thePHB-PEG-PHB nanoparticles, suggesting that the PHB-PEG-PHB nanoparticles were obviously much hemocompatible fordrug delivery applications. In the presence of intracellularenzyme esterase, the biocompatible PHB-PEG-PHB nanopar-

ticles might be hydrolyzed, and their biodegradable behaviorwas monitored by the fluorescence spectrum and the pH me-ter. The initial biodegradation rate of the PHB-PEG-PHB nano-particles was closely related to the enzymatic amount and thePHB block length. Compared with that obtained from the flu-orescence determination, the initial biodegradation rate frompH measurement was faster. The biodegraded productsmainly consisted of 3HB monomer and dimer, which were themetabolites present in the body. � 2008 Wiley Periodicals, Inc.J BiomedMater Res 87A: 290–298, 2008

Key words: nanoparticles; cytotoxicity; biocompatible; bio-degradation

INTRODUCTION

Amorphous biodegradable poly(3-hydroxybuty-rate)–poly(ethylene glycol)–poly(3-hydroxybutyrate)(PHB-PEG-PHB) nanoparticles have the potential asdrug delivery systems because of their possible supe-rior drug loading properties based on our previousstudies with pyrene as the imitative drug and theirshorter biodegradation period in vitro than microbial

PHB in the presence of the extracellular enzyme.1 Thenanoparticles can be produced from amphiphilic tri-block copolymers consisted of a PEG central block(molecular weight of 4 kDa) with two PHB blocks ofvarying molecular weight (5–70 kDa).2 It has beenshown that the PEG block is in an extended solvatedstate to form the hydrophilic outer shell of the nanopar-ticles and energetically stabilize the colloidal nanopar-ticles, while the PHB block is entrapped in the centralsolid state as hydrophobic core to minimize their inter-action with water due to their inherent hydrophobiccharacters.1,2 Taking advantage of the core-shell struc-ture, hydrophobic drugmoleculesmay be encapsulatedinto the hydrophobic region of the nanoparticle andthen released according to different mechanisms.3,4

When the nanoparticles are used as drug carriersfor medical application, the fundamental requirementis that it should display adequate biocompatibilityand biodegradability. PEG is widely used to makedrugs by pegylation which can mask certain drugssuch as interferon from the immune system and pre-vent rejection. Moreover, it can effectively reduce the

Additional Supporting Information may be found in theonline version of this article.Correspondence to: C. Chen; e-mail: chencheng510@yahoo.

comContract grant sponsor: University Grant Council of

Hong Kong; contract grant numbers: PolyU 5299/01P,PolyU 5257/02M, PolyU 5403/03MContract grant sponsor: University Grants Committee

Area of Excellence Scheme (Hong Kong); contract grantnumber: AoE/P-10/01

� 2008 Wiley Periodicals, Inc.

uptake of proteins at the reticuloendothelial sites (e.g.liver and spleen).5 Microbial PHB possesses good bio-degradability and biocompatibility. It was found thatmicrobial PHB was tolerated by animal tissues with-out the induction of inflammation or necrosis.6 How-ever, a little inflammation of the tissues adherent tothe implants was observed in the rabbit body after3 and 6 months of microbial PHB implantation.7

In addition, the in vitro toxicological investigationsof poly(3-hydroxybutyrate-co-3-hydroxyvalerate) [P(3HB-3HV)] in cell culture tests and in vivo experi-ments with animals revealed inflammatory responses,the degree of which correlated with the HV content.8

These results indicated that the biocompatibility ofPHB was determined not only by its chemical compo-sitions but also by the processing methods used, theforms of the implants, its surface properties, and thedegree of chemical purity of the materials.

As our series of studies on the PHB-PEG-PHB nano-particles as potential drug carriers, this article pre-sented the cytotoxicity and hemolysis activity of thePHB-PEG-PHB nanoparticles on different animal cellsin vitro, especially in the PHB block was obtained bychemical synthesis. The effect of the molecular weight(Mn) of the PHB block on the biocompatibility wasalso investigated. In our previous studies,1 it wasfound that the PHB-PEG-PHB nanoparticles might bebiodegraded by the extracellular PHB depolymerasewhich was widely used to investigate the biodegrada-tion of microbial PHB. In fact, little exists the PHBdepolymerase in animal body. To overcome the short-coming, in the present works, the intracellularenzyme, esterase obtained from porcine liver, wasused to further investigate the biodegradation behav-ior of the PHB-PEG-PHB nanoparticles in vitro.

MATERIALS AND METHODS

Materials

All chemicals, unless otherwise stated, were purchasedfrom Aldrich Chemical. Two triblock PHB-PEG-PHB

copolymers with different PHB block lengths were pre-pared following the procedures described elsewhere.2 TheMn of the PHB block varied from 10 kDa (35 repeatedunits) to 20 kDa (100 repeated units). The PEG blockremained constant throughout the series at 4 kDa. Thecharacteristics of both PHB-PEG-PHB triblock copolymersused in this study are listed in Table I. Esterase fromporcine liver was purchased from Fluka Chemical, andused at a constant concentration of 10 lg/mL duringbiodegradation.

Preparation of nanoparticles

The nanoparticles were prepared via a surfactant-freeapproach according to our previously described method.1,2

Typically, a triblock copolymer solution in acetone wasadded dropwise to distilled and deionized water (DDH2O) under ultrasonic situation. The acetone was removedunder a low pressure and ultrasonic situation. The initial con-centration of the PHB-PEG-PHB dispersions was 3 mg/mL.This method was very easy and convenient because itavoided the use of aggressive conditions (emulsion adju-vants, complicated procedures etc.) that compromised thestability of the encapsulated molecules. Scheme 1 showsthe preparation procedures of the PHB-PEG-PHB nanopar-ticles, together with the corresponding morphology. Theimage exhibited that the PHB-PEG-PHB copolymers formspherical, discrete particles in aqueous medium.

The effect of different concentrations of the nanoparticledispersions on cell cytotoxicity was investigated by dilut-ing the initial dispersions to the desired concentrations(120, 60, 40, 20, 10, and 5 lg/mL). All dispersions werefiltered using disposable 0.45-lm Millipore filters, with-out any significant effect on the particle yield or sizedistribution.

Cell adhesion assay and cell proliferation studies

The effects of the nanoparticles on (a) cell adhesionsand (b) cell proliferation were investigated using two typesof animal cells (Mouse Fibroblasts L929 and ChineseHamster Ovary cell CHO-K1), respectively. The cells wereincubated at 378C in a 5% CO2 atmosphere. (a) For cell adhe-sion, L929 cells (3 3 105 per mL) and CHO-K1 cells (2.5 3105 per mL) were seeded in the nanoparticle dispersions

TABLE ICharacteristics of the PHB-PEG-PHB Triblock Copolymers

SamplesStructurea

(HBx-EGy-HBx) Mn Dnb (nm) Polydispersityb

Zeta Potentials (mV)

in DD H2O in PBS

1 35-91-35 10,020 72.6 6 7.7 1.2 0 9.8 6 0.12 100-91-100 21,200 88.9 6 5.6 1.1 1.8 6 0.1 10.0 6 0.1

HB and EG represented the PHB and PEG blocks, respectively; x and y represented the number-average degree of poly-merization of the PHB and PEG blocks.

aCalculated for the integration of NMR resonances belonging to the PEG block at 3.64 ppm and to the PHB block at 5.26ppm.

bDv and Dn were the volume and number average particle diameters, respectively. The polydispersity was determinedby Dv/Dn. Data presented the mean and the standard deviation of 10 independent experiments.

BIODEGRADABLE PHB-PEG-PHB NANOPARTICLES AS POTENTIAL DRUG CARRIERS 291

Journal of Biomedical Materials Research Part A

with different concentrations, and then incubated for 4 h.(b) For cell proliferation, L929 cells (2.5 3 105 per mL) andCHO-K1 cells (2.5 3 105 per mL) were first seeded in 96-well tissue-culture plates. After 24 h, the culture mediumin the well was replaced by the medium containing nano-particle dispersions with different concentrations, and thecells were then further incubated for 1 day and 3 days,respectively.

In both systems, 4 lL of nanoparticle dispersions withdifferent concentrations (120–5 lg/mL) was added to100 lL cells with the culture medium. For control experi-ments, sterilized DD H2O and 10% ethanol were used asreferences of normal cell viability and negative control. Cellnumbers were determined using (3-(4,5-dimethylthiazol-2-yl)-5-(3-carboxy methoxyphenyl)-2-(4-sulfophenyl)-2H- tet-razolium (MTS, Promega) assay at 378C for 2 h after the re-moval at 490 nm. Viable cell numbers were then determinedfrom the standard curve based on their MTS absorbency.

In vitro hemolysis assay of the nanoparticles

Hemolysis assay was performed using fresh pig blood.The erythrocytes were collected by centrifugation at1500 rpm for 15 min, and then washed three times withPhosphate buffered saline (PBS) buffer (Dulbecco’s PBS,Gibco) at pH 7.4. The stock dispersion was prepared bymixing 3 mL of centrifuged erythrocytes into 11 mL of PBS.The PHB-PEG-PHB nanoparticle dispersions were preparedin PBS buffer with the above-mentioned concentrations(120–5 lg/mL). One hundred microliter of stock dispersionwas added to 1 mL of the nanoparticle dispersions. The sol-utions were mixed and incubated for 4 h at 378C in an Incu-bator Shaker. The percentage of hemolysis was measuredby UV–vis analysis of the supernatant at 394 nm absorbanceafter centrifugation at 13,000 rpm for 15 min. One milliliterof PBS was used as the negative control with 0% hemolysis,and 1 mL of DD H2O was used as the positive control with100% hemolysis. All hemolysis data points were presentedas the percentage of the complete hemolysis.

Statistical analysis

In the cell adhesion assay, cell proliferation studies andin vitro hemolysis assay, the results are expressed as mean6 standard deviation from five independent experiments

for cell cytotoxicity and three independent experiments forhemolysis assay. Statistical analysis was performed usinganalysis of variance (ANOVA) with post hoc Bonferroni’scorrection or Student’s t-test, where appropriate. Differen-ces were considered significant when p < 0.05.

Biodegradation of the nanoparticles in vitro

In a typical enzymatic biodegradation experiment, dust-free nanoparticle dispersions with a concentration of1 mg/mL were prepared in the presence of pyrene (6 31027M) aqueous solution. To remove excess pyrene, thenanoparticle dispersion was purified by placing it into a5 Da molecular weight cut-off (Spectra/Por1CE) and dia-lyzed against DD H2O for 2 days at room temperatureunder shielded light. Dust-free enzyme solution (0.01 and0.05 mL) was added into 1 mL nanoparticle dispersions atroom temperature to start the biodegradation experiments.The fluorescence spectra of pyrene and the pH values inthe nanoparticle dispersions were recorded during biode-gradation. Afterwards, the biodegraded products were fro-zen and lyophilized, and then redissolved in chloroform-d(CDCl3) for qualitative analysis.

Analytic methods

The NMR analysis of the specimens was carried out ona Varian Inova 500-MHz NMR spectrometer. Chemicalshifts were given in ppm using tetramethylsilane (TMS) inCDCl3 as the internal reference and sodium 3-trimethylsi-lylpropionate-d4 (TSP) in deuterated water (D2O) as theexternal reference. The sizes and zeta potentials of thenanoparticles were measured using a Malvern Zetasizer3000HSA (Malvern, UK) equipped with a 10 mW He-Nelaser (633 nm) and operating at 208C and an angle of 908.For the measurement of zeta potentials, DD H2O and PBSbuffers were used as the suspension fluids, respectively.Since depended on the environmental pH values, zetapotentials in the present studies were carried out at thecorresponding pH values of DD H2O (pH 5 5.8) and a rel-evant physiologic pH (7.4). The scanning number was 10for particle sizes and 5 for zeta potentials. The UV–visanalysis during the hemolysis process was carried outwith a U-2800 Spectrophotometer (Hitachi, Japan). Fluores-cence measurements were carried out with a Perkin-Elmer

Scheme 1. Schematic formation of the PHB-PEG-PHB nanoparticles. (A) Copolymerization of the PEG block with the b-butyrolactone (BL) monomer. (B) Dispersion of the amphiphilic PHB-PEG-PHB triblock copolymers in aqueous mediumunder ultrasonic situation. (C) Self-assembly of the amphiphilic PHB-PEG-PHB triblock copolymers in aqueous medium.(D) SEM micrograph of the PHB-PEG-PHB nanoparticles for Sample 1. [Color figure can be viewed in the online issue,which is available at www.interscience.wiley.com.]

292 CHEN ET AL.

Journal of Biomedical Materials Research Part A

LS50B Luminescence spectrometer. The excitation wave-length was 339 nm, and the emission wavelength was394 nm. Both excitation and emission bandwidths were2.5 nm. The pH values were measured with a pH/ISEmeter mold 710A.

RESULTS AND DISCUSSION

Characterization of the PHB-PEG-PHBnanoparticles

Table I lists the mean diameters of Sample 1 andSample 2 at the initial concentration of 3 mg/mL. Itwas found that their sizes were 72.6 and 88.9 nm,respectively. The PHB-PEG-PHB nanoparticles witha longer PHB block length exhibited a great size. Inaddition, both PHB-PEG-PHB nanoparticle disper-sions showed high monopolydispersity since theirDv/Dn values were between 1.1 and 1.2.

The core and shell compositions of the PHB-PEG-PHB nanoparticle were quantitatively analyzed byusing the 1H NMR spectrum according to previousreported method,1,9 and the results are listed in Ta-ble II. Twelve- and ten-percentage methyl protons(CH3) of the PHB block were detected in Sample 1and Sample 2, respectively. The results implied thatmost of the PHB block formed the hydrophobic core.The shorter the PHB block, the higher the relativelyhigh percentage of the PHB methyl protons (CH3%)detected by 1H NMR. This indicated that in the caseof the PHB-PEG-PHB nanoparticles with the shortPHB block there were some movements of the PHBcore. The result was reasonable considering theweak hydrophobic interaction for the short PHBblock. In addition, both the methine and the methyl-ene signals of the PHB block were not detected,which implied that both proton groups were in astate which could not be resolved by the NMRexperiment and hence they were in a solid-like envi-ronment. However, since some of the methyl groupsof the PHB block were seen, it was expected that asmall methine and methylene signals of the PHBblock would also be present. The lack of the methineand methylene signals might be due to the weaknessof the signal because the corresponding methyl reso-nance was also very weak. To consider the effect ofthe relaxation decay time on the NMR experiment, itwas changed from 0.1 to 20 s. However, both themethine and the methylene groups of the PHB blockwere still not detected, which further supported thatboth kinds of the protons would be at a more re-stricted environment. Similar results were alsoreported in the PLA-PEG micelle system.9 For thePEG block, it was found that 24 and 19% of themethylene protons in Sample 1 and Sample 2 wereundetectable (Table II). The results indicated that

some portions in the PEG block might be trappedinside the hydrophobic core, especially for the por-tions adjacent to the PHB block.

Table I also lists the zeta potentials of the PHB-PEG-PHB nanoparticles dispersed in DD H2O andPBS solution. For the nanoparticles consisted of ali-phatic polyesters, for example, PCL-dextran copoly-mers10 and PLLA-PEG copolymers,11 they usuallypresented negative zeta potentials because of thepresence of ionized carboxyl groups on the surface.For two PHB-PEG-PHB nanoparticles dispersed inDD H2O, their zeta potentials were almost zero,implying that little charges appeared on the surfaceof the PHB-PEG-PHB nanoparticles. It might be theresult of the PEG block on the surface covering thenegative surface charges of the PHB carboxylgroups. The result was also supported by the factthat the methylene signals of the PHB block adjacentto the carboxyl groups were undetectable by 1HNMR. In PBS buffer, the zeta potentials of two PHB-PEG-PHB nanoparticle dispersions were þ9.8 andþ10.0 mV, respectively. Although the PHB blocklength was different, the zeta potentials of bothnanoparticle dispersions were almost the same evenwhen they were dispersed in different media. It waswell known that low zeta potential was correlatedwith low cytotoxicity,12,13 since a high absolute valueof the surface charge would trigger the immune sys-tem of the body against the nanoparticles or induceblood coagulation. Hence, the results of low zetapotentials may be considered as an indirect evidencethat the PHB-PEG-PHB nanoparticles should bemuch biocompatible, especially in PBS bufferbecause it mimicked the surrounding environmentof the nanoparticles in body.

In vitro cytotoxicity of thePHB-PEG-PHB nanoparticles

The cytotoxicity of the PHB-PEG-PHB nanopar-ticles was investigated in two aspects: cell adhesionand cell proliferation. To compare conveniently, thecell number cultured in DD H2O was considered as100% and acted as its cell viability. As a negative

TABLE IIPercentage of PHB Methyl (CH3) and PEG Methylene

(CH2) Resonances Detected by 1H NMR for thePHB-PEG-PHB Nanoparticles at 208C

Samples CH3% (PHB block) CH2% (PEG block)

1 12 762 10 81

The percentage was determined as the ratio of theresonances obtained in D2O compared to that in CDCl3assuming that 100% resonances were seen in the room-temperature CDCl3 spectra.

BIODEGRADABLE PHB-PEG-PHB NANOPARTICLES AS POTENTIAL DRUG CARRIERS 293

Journal of Biomedical Materials Research Part A

control, 10% ethanol showed considerable inhibitioneffect on cell growth during cell adhesion. Its cell vi-ability decreased to 26% (p < 0.001) for CHO-K1 and37% (p < 0.001) for L929, respectively (Fig. 1). Gener-ally, when the cell numbers decreased significantlymore than 10% in the media, cell inhibition in themedia was suggested.14 For Sample 1, its cell viabil-ity slightly reduced in both kinds of the cells. Thegreatest reduction was 6.47% (p > 0.05) in CHO-K1(40 lg/mL) and 9.6% (p > 0.05) in L929 (120 lg/mL)in the whole range (Fig. 1). Meanwhile, Sample 2with long PHB block showed higher cell viabilitythan Sample 1, indicating that long PHB segmentsmight improve the biocompatibility to a certainextent. When the inhibition of cell growth was in therange of 0–10%, the cytotoxicity index was zero.14 Itwas known that cell adhesion was mediated by theinteraction of surface proteins such as integrins withproteins in the extracellular matrix or on the surfaceof other cells or particles.15,16 The phenomenon ofcell adhesion was of crucial importance in governinga range of cellular functions including cell growth,migration, differentiation, survival, and tissue orga-nization.15,16 Thus the delivery system based on thePHB-PEG-PHB nanoparticles should not elicit ageneric and chronic inflammatory response thatcould ultimately result in a failure to achieve normalcell growth and function on the cell-particle surface.

For cell proliferation, the growth of both cells wasstrongly suppressed by 10% ethanol as a negativecontrol (Figs. 2 and 3). However, when the L929 cellswere added into two PHB-PEG-PHB nanoparticledispersions, the cell growth percentages in bothnanoparticle dispersions were almost the same as

that in DD H2O in 1 day, and slightly lower thanthat in 3 days (Fig. 2). For CHO-K1 cells for 3 days,the cell growth percentages in both PHB-PEG-PHBnanoparticle dispersions were significantly higherthan those in DD H2O (p < 0.001) (Fig. 3). Moreover,their values generally decreased with the nanopar-ticle concentrations. The results showed that thegrowth of both kinds of cells was not affected by theintroduction of the PHB-PEG-PHB nanoparticle dis-persions. It was noted for the CHO-K1 cells that thePHB-PEG-PHB nanoparticles might slightly improve

Figure 1. Cell viability of CHO-K1 (black bars) and L929(red bars) cells as a function of nanoparticle concentrationsat 4 h. Data presented the mean and standard deviation offive independent experiments. Sample 1: hollow column;Sample 2: cross-patterned column. [Color figure can beviewed in the online issue, which is available at www.interscience.wiley.com.]

Figure 2. Cell growth percentage of L929 cells as a func-tion of nanoparticle concentrations for 1 day (black bars)and 3 days (red bars). Data presented the mean and stand-ard deviation of five independent experiments. Sample 1:hollow column; Sample 2: cross-patterned column. [Colorfigure can be viewed in the online issue, which is availableat www.interscience.wiley.com.]

Figure 3. Cell growth percentage of CHO-K1 cells as afunction of nanoparticle concentrations for 1 day (blackbars) and 3 days (red bars). Data presented the mean andstandard deviation of five independent experiments. Sam-ple 1: hollow column; Sample 2: cross-patterned column.[Color figure can be viewed in the online issue, which isavailable at www.interscience.wiley.com.]

294 CHEN ET AL.

Journal of Biomedical Materials Research Part A

its cell proliferation. With the increase in the PHBblock length, the cell growth percentage did not ex-hibit remarkable differences.

When the nanoparticles were used to deliverdrugs, they could be accumulated in cells with ahigh concentration and then lead to toxicity, forexample, carbon-based nanoparticles.17 In the case ofPluronic micelles, significant effects on prostate car-cinoma cells were not observed until after 144 h oftreatment. It was because of a greater accumulationof the micelles in the cells, which was the key forproducing cytotoxicity in cells with a low mitotic ac-tivity.18 In the present studies, the PHB-PEG-PHBnanoparticle dispersions with a high concentrationof 120 lg/mL were used to investigate the cytotoxic-ity resulting from their accumulation. The value wasmuch higher than the one reported (9 lg/mL).8 Theexperimental results showed that both kinds of cellsstill exhibited excellent cell growth percentage inboth PHB-PEG-PHB nanoparticle dispersions (Figs. 2and 3), indicating that the biocompatibility of thePHB-PEG-PHB nanoparticles was very reliable.The fact also indicated that it was feasible to developthe PHB-PEG-PHB nanoparticles as a drug deliverysystem. Some low molecular weight forms of PHBwere detected in human tissues and blood serum,and up to 20–30% of PHB was correlated with lipo-protein.19 Hence, it was not difficult to understandthat the PHB-PEG-PHB nanoparticle exhibited excel-lent biocompatibility and had no cytotoxicity.

Hemolytic properties of thePHB-PEG-PHB nanoparticles

When the nanoparticles are injected into the bloodfor drug delivery or drug detoxification, detrimentalinteraction of these particles with blood constituentsmust be avoided. Although the concentrations of thenanoparticle dispersions were very high, both PHB-PEG-PHB nanoparticle dispersions did not show anyobservational hemolytic activities in the red bloodcell in the experimental range. Figure 4 shows the he-molytic activities of two PHB-PEG-PHB nanoparticledispersions with different PHB block lengths. It wasobserved that the hemolytic percentage of the nano-particle dispersions depended on its concentrations.The hemolytic percentage of Sample 1 was alwayslower than 0.1% in the whole tested concentrationrange. For Sample 2, hemolysis was not detected atconcentrations below 40 lg/mL, while 0.48% hemoly-sis was detected at the concentration of 120 lg/mL.According to the ISO/TR 7405-1984(f), the sampleswere considered as hemolytic if the hemolyticpercentage was above 5%. Both PHB-PEG-PHB nano-particles induced hemolytic percentages which weresignificantly lower than 5% (p < 0.001). Consequently,

the PHB-PEG-PHB nanoparticles had no hemolyticeffect on the red cell suspension. The results also sug-gested that the PHB-PEG-PHB nanoparticles weresuitable for a wide safety margin in blood-contactingapplications and suitability for intravenous adminis-tration.20 However, it was also reported that PHBwould activate the hemostasis system. A probablereason was that the microbial PHB had some impur-ities or endotoxin produced by the gram-negativebacteria.21 In the present studies, atactic PHB wasused to replace microbial PHB, thus these disadvanta-geous factors were avoided. In the PLGA nanopar-ticles system,22 surfactant stabilized PLGA nanopar-ticles could cause 80% hemolysis because there wasan interaction between these particles and red bloodcells, damaging most of red blood cells. For PEGy-lated PLGA nanoparticles, although their hemolysiswas much less than surfactant stabilized PLGA, thehemolytic percentage also increased up to 10%. Com-pared to PLGA nanoparticles, the PHB-PEG-PHBnanoparticles were obviously more hemocompatiblefor drug delivery applications.

In vitro biodegradation of thePHB-PEG-PHB nanoparticles

To investigate the biodegradation behavior of thePHB-PEG-PHB nanoparticles, pyrene was used asmodel drug to encapsulate into the PHB core becauseof its unique fluorescence. It was noted that the addi-tion of the enzyme had no any influence on the fluo-rescence spectrum of pyrene. Consequently, takingadvantage of the differences of pyrene fluorescencepeak in pure water (at 333.1 nm) and in the hydropho-bic core (at 336.6 nm), the biodegradation behavior ofthe PHB-PEG-PHB nanoparticles might be revealed

Figure 4. Hemolytic activities of the PHB-PEG-PHB nano-particles with different PHB block lengths as a function ofnanoparticle concentrations. Data presented the mean andstandard deviation of three independent experiments.

BIODEGRADABLE PHB-PEG-PHB NANOPARTICLES AS POTENTIAL DRUG CARRIERS 295

Journal of Biomedical Materials Research Part A

by monitoring the change of the corresponding inten-sity at 336.6 nm [I(336.6)]. It was because the release ofpyrene in the PHB core during the biodegradationwould induce the decrease of I(336.6) (Fig. 5).

Figure 6 shows the relative ratio of I(336.6)t/I(336.6)0as a function of biodegradation time (tbiodegradation) ata constant initial concentration of the dispersions(1 mg/mL). It was found that esterase was also ableto biodegrade the PHB-PEG-PHB nanoparticles,although the PHB block was amorphous. This wascompletely different from common PHB biodeg-radation which needed crystalline phase to induceenzymatic hydrolysis.23 Considering the core-shellstructure of the PHB-PEG-PHB nanoparticles, theouter PEG shell should have a steric hindrancetoward esterase. However, the molecules of esterasestill passed through the shell and biodegraded thePHB core, implying that the core-shell structure ofthe nanoparticles was loose and the PEG steric hin-drance was not significant especially compared to achemical barrier for the bioreaction of esterase itself.As the biodegradation time increased, the values ofI(336.6)t/I(336.6)0 gradually reduced. When the timewas above 20 min, it almost kept constant. It meansthat during the biodegradation the enzyme mole-cules could lose their activity. It was well-knownthat the activity of enzymes depended on the tem-perature and the pH values in the environment.PHB is an aliphatic polyester, and its biodegradedproducts included some small acids. The productionof small acids would result in the pH decrease inthe dispersions, which was supported by the follow-ing results. For esterase, its optimal pH condition inuse was 8.0 according to supplier. Consequently, thepH decrease of the dispersions was disadvantageousto enzymatic degradation. In addition, the remained

PEG chains after PHB biodegradation might act asthe surfactant and surround the enzymatic mole-cules.24 Because the devitalization of the enzymeresulted in that the PHB-PEG-PHB nanoparticleswere not completely biodegraded. Therefore, thefinal values of I(336.6)t/I(336.6)0 were not zero. It wasalso shown in Figure 6 that the biodegradation ofthe PHB-PEG-PHB nanoparticles was also relatedwith the enzymatic amount. The greater the enzy-matic amount, the higher the biodegradation rate,since the values of I(336.6)t/I(336.6)0 with 5% enzymaticamount were clearly lower than that with 1% enzy-matic amount. The result also indicated that the bio-degradation extent of the nanoparticle dispersionsincreased with the increase in enzymatic amount.

Biodegradation behavior monitored by the changeof pH showed different trends compared with thatby the change of I(336.6). When the enzymatic amountwas 1%, the relative pH values almost decreased lin-early with biodegradation time during the total bio-degradation. As the enzymatic amount reached 5%,the relative pH values remarkably decreased in theinitial 20 min, and then hardly changed with theincrease in biodegradation time, indicating thatthe biodegradation of the PHB-PEG-PHB nanopar-ticles mainly finished in the initial 20 min (Fig. 7).

By [dY/dt]t?0 [Y was I(336.6) or pH value], the ini-tial biodegradation rate (V0) can be determined, andthe results are listed in Table III. The V0 valuesobtained by the change of pH were much faster thanthat by the change of I(336.6) under the same experi-mental conditions. The discrepancy mainly resultedfrom the different natures of the two methods.1 Itcould be imaged that under the influence of theenzyme the PHB chains were first broken. The pHvalues directly reflected the amount of small acids

Figure 5. Fluorescence excitation spectra of the PHB-PEG-PHB nanoparticles (Sample 2) loading pyrene duringthe biodegradation.

Figure 6. The relative ratio of I(336.6)t/I(336.6)0 for Sample 2as a function of biodegradation time at a constant initialconcentration, where the subscripts ‘‘t’’ and ‘‘0’’ representedbiodegradation time t 5 0 and t 5 t, respectively.

296 CHEN ET AL.

Journal of Biomedical Materials Research Part A

produced from the PHB chain scission. As the biode-gradation extent of the PHB chains inside each nano-particles increased, the encapsulated pyrene wasreleased into water, while the nanoparticles still keptthe micelle state. Only when the biodegradation ofthe PHB chains reached a certain extent, the nano-particles started to disintegrate. This was the reasonthe V0 values obtained by the change of pH wasfaster than that by the change of I(336.6). In addition,the V0 values measured by two methods alsoincreased with the length of PHB block (Table III). Itwas expected because esterase can only interact withthe PHB chains, and the hydrophilic PEG shellshould have no effect on the biodegradation.

It was worthy to be noted in Figure 7 that in thecase of 1% enzymatic amount the biodegradation pe-riod monitored by pH meter was about 60 min,while that monitored by the fluorescence was about20 min. The results suggested that when the biode-gradation time was above initial 20 min the encapsu-lated pyrene was difficult to be released from the

PHB core, although the hydrolysis of the PHB seg-ments was still in process. When the enzymaticamount was 5%, the biodegradation periods meas-ured by two methods were almost the same, due tothe high initial biodegradation rate.

Figure 8 shows the 1H NMR spectrum of the bio-degraded products from Sample 2. Compared withthe nonbiodegraded 1H NMR spectrum,1,2 it wasshown that the peaks of the methyl and the methinegroups of 3HB monomer were observed at about1.21 and 4.03 ppm, respectively. The methyl and themethylene groups attributed to 3HB dimer were alsoclearly seen at 1.26 and 2.55 ppm.1,23 However, anyresonances corresponding to the 3HB trimer werenot detected.1 Therefore, in the presence of esterase,the biodegraded products of the PHB-PEG-PHBnanoparticles mainly consisted of 3HB monomer anddimer, which are widely found in the body.

Further detailed works about the encapsulationand release of proteins using the PHB-PEG-PHBnanoparticles have been investigated in vitro.

CONCLUSIONS

In this article, the cytotoxicity and hemolysis assayof the PHB-PEG-PHB nanoparticles was investigatedin vitro using two kinds of animal cells, and its bio-degradation profiles was also studied in the presenceof the intracellular enzyme esterase. Cell adhesionassay and cell proliferation studies showed that theaddition of the PHB-PEG-PHB nanoparticles hardlyaffect the growth of animal cells, even when the con-centrations of the nanoparticle dispersions addedup to 120 lg/mL. Moreover, for the PHB-PEG-PHB

TABLE IIIInitial Biodegradation Rate (V0) of the PHB-PEG-PHB

Nanoparticles Determined With Fluorescenceand pH Meter

Samples

V0 byFluorescence(mg/mL/min)

V0 bypH Meter

(mg/mL/min)

1% 5% 1% 5%

1 0.02 0.11 0.33 1.272 0.04 0.20 0.80 1.87

V0 was determined by [dY/dt]t?0 [Y was I(336.6) or pHvalue].

Figure 8. 1H NMR spectrum of the biodegraded PHB-PEG-PHB nanoparticles for Sample 2.

Figure 7. The relative ratio of pH values as a function ofbiodegradation time for Sample 2, where Y was pH values,and the subscripts ‘‘0,’’ ‘‘t,’’ and ‘‘1’’ represented biodegra-dation time t 5 0, t 5 t, and t 5 end, respectively.

BIODEGRADABLE PHB-PEG-PHB NANOPARTICLES AS POTENTIAL DRUG CARRIERS 297

Journal of Biomedical Materials Research Part A

nanoparticle system, no hemolysis was found in thewhole experimental concentration range. Althoughthe nanoparticles were amorphous, they could stillbe hydrolyzed by the esterase. The biodegrada-tion behavior of the PHB-PEG-PHB nanoparticlesdepended on the monitoring methods. The enzymaticamount has significant effect on the initial biodegra-dation rate of the PHB-PEG-PHB nanoparticles. Thebiodegraded products of the PHB-PEG-PHB nanopar-ticles mainly consisted of 3HB monomer and dimer.These features showed that the PHB-PEG-PHB nano-particles were very suitable as potential drug carriers.

References

1. Chen C, Yu CH, Cheng YC, Yu Peter HF, Cheung MK. Bio-degradable nanoparticles of amphiphilic triblock copolymersbased on poly(3-hydroxybutyrate) and poly(ethylene glycol)as drug carriers. Biomaterials 2006;27:4804–4814.

2. Chen C, Yu CH, Cheng YC, Yu Peter HF, CheungMK. Prepara-tion and characterization of biodegradable nanoparticlesbased on amphiphilic poly(3-hydroxybutyrate)-poly(ethyleneglycol)-poly(3-hydroxybutyrate) triblock copolymer. Eur PolymJ. 2006;42:2211–2220.

3. Hans ML, Lowman AM. Biodegradable nanoparticles fordrug delivery and targeting. Curr Opin Solid State Mater Sci2002;6:319–327.

4. Soppimatha KS, Aminabhavia TM, Kulkarnia AR, RudzinskiWE. Biodegradable polymeric nanoparticles as drug deliverydevices. J Controlled Release 2001;70:1–20.

5. Gref R, Minamitake Y, Peracchia MT, Trubetskoy V, TorchilinV, Langer R. Biodegradable long-circulating polymeric nano-spheres. Science 1994;263:1600–1603.

6. Doi Y. Microbial Polyester. New York: VCH; 1990.7. Qu XH, Wu Q, Zhang KY, Chen GQ. In vivo studies of

poly(3-hydroxybutyrate-co-3-hydroxyhexanoate) based poly-mers: Biodegradation and tissue reactions. Biomaterials2006;27:3540–3548.

8. Gogolewski S, Jovanovic M, Perren SM, Dillon JG, Hughes MK.Tissue response and in vivo degradation of selected poly-hydroxyacids: Polylactides (PLA), poly(3-hydroxybutyrate)(PHB), and poly(3-hydroxybutyrate-co-3-hydroxyvalerate)(PHB/VA). J Biomed Mater Res 1993;27:1135–1148.

9. Heald CR, Stolnik S, Kujawinski KS, Matteis CD, GarnettMC, Illum L, Davis SS, Purkiss SC, Barlow RJ, Geller PR. Poly(lactic acid)-poly(ethylene oxide) (PLA-PEG) nanoparticles:NMR studies of the central solidlike PLA core and the liquidPEG corona. Langmuir 2002;18:3669–3675.

10. Rodrigues JS, Santos-Magalhaes NS, Coelho LCBB, CouvreurP, Ponchel G, Gref R. Novel core(polyester)-shell(polysacchar-ide) nanoparticles: Protein loading and surface modificationwith lectins. J Controlled Release 2003;92:103–112.

11. Zhang Y, Zhang QZ, Zha LS, Yang WL, Wang CC, Jiang XG,Fu SK. Preparation, characterization and application of py-rene-loaded methoxy poly(ethylene glucol)-poly(lactic acid)copolymer nanoparticles. Colloid Polym Sci 2004;282:1323–1328.

12. Shuai XT, Merdan T, Unger F, Wittmar M, Kissel T. Novelbiodegradable ternary copolymers hy-PEI-g-PCL-b-PEG: Syn-thesis, characterization, and potential as efficient nonviralgene delivery vectors. Macromolecules 2003;36:5751–5759.

13. Cai KY, Frant M, Bossert J, Hildebrand G, Liefeith K, JandtKD. Surface functionalized titanium thin films: Zeta-potential,protein adsorption and cell proliferation. Colloids Surf B Bio-interfaces 2006;50:1–8.

14. Dijkhuizen-Radersma R, Hesseling SC, Kaim PE, Groot K,Bezemer JM. Biocompatibility and degradation of poly(ether-ester) microspheres: In vitro and in vivo evaluation. Biomate-rials 2002;23:4719–4729.

15. Haas TA, Plow EF. Integrin-ligarid interactions: A year inreview. Curr Opin Cell Biol 1994;6:656–662.

16. Lamazi C, Fugimoto LM, Yin HL, Schmid SL. The actin cyto-skeleton is required for receptor-mediated endocytosis inmammalian cells. J Biol Chem 1997;33:20332–20335.

17. Monteiro-Riviere NA, Inman AO. Challenges for assessingcarbon nanomaterial toxicity to the skin. Carbon 2006;44:1070–1078.

18. McNealy TL, Trojan L, Knoll T, Alken P, Michel1 MS. Micelledelivery of doxorubicin increases cytotoxicity to prostate car-cinoma cells. Urol Res 2004;32:255–260.

19. Piddubnyak V, Kurcok P, Matuszowicz A, Glowala M, Kierz-kowska AF, Jedlinski Z, Juzwa M, Krawczyk Z. Oligo-3-hydroxybutyrates as potential carriers for drug delivery. Bio-materials 2004;25:5271–5279.

20. Lee DW, Powersb K, Baney R. Physicochemical propertiesand blood compatibility of acylated chitosan nanoparticles.Carbohydr Polym 2004;58:371–377.

21. Volova T. Polyhydroxyalkanoates—Plastics Material of the21st Century. New York: Nova Science Publishers; 2004.

22. Kim D, El-Shall H, Dennis D, Morey T. Interaction of PLGAnanoparticles with human blood constituents. Colloids Surf BBiointerfaces 2005;40:83–91.

23. Wang Y, Inagawa Y, Osanai Y, Kasuya KI, Saito T, Matsu-mura S, Doi Y, Inoue Y. Enzymatic hydrolysis of bacterialpoly(3-hydroxybutyrate-co-3-hydroxypropionate)s by poly(3-hydroxyalkanoate) depolymerase from Acidovorax Sp. TP4.Biomacromolecules 2002;3:828–834.

24. Nie T, Zhao Y, Xie ZW, Wu C. Nanoparticlear formation ofpoly(caprolactone-block-ethylene oxide-block-caprolactone) andits enzymatic biodegradation in aqueous dispersion. Macromo-lecules 2003;36:8825–8829.

298 CHEN ET AL.

Journal of Biomedical Materials Research Part A