A

EaCpba(opooPib©

K

1

ddm

I

0d

Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79

Insight in the role of bovine serum albumin for promoting the in situsurface growth of polyhydroxybutyrate (PHB) on patterned surfaces

via enzymatic surface-initiated polymerization

Nuttawee Niamsiri a,b,e,∗, Magnus Bergkvist b, Soazig C. Delamarre a, Nathan C. Cady a,Geoffrey W. Coates c, Christopher K. Ober d, Carl A. Batt a,b

a Department of Food Science, Cornell University, Ithaca, NY 14853, United Statesb Nanobiotechnology Center, Cornell University, Ithaca, NY 14853, United States

c Department of Chemistry and Chemical Biology, Cornell University, Ithaca, NY 14853, United Statesd Department of Material Sciences, Cornell University, Ithaca, NY 14853, United States

e Department of Biotechnology, Faculty of Science, Mahidol University, Bangkok 10400, Thailand

Received 26 April 2007; received in revised form 29 May 2007; accepted 29 May 2007Available online 2 June 2007

bstract

Polyhydroxyalkanoates (PHAs) are a family of aliphatic polyesters produced by a variety of microorganisms as a reserve of carbon and energy.nzymes involved in the synthesis of PHAs can be utilized to produce polymers in vitro, both in bulk and on solid surfaces. Here, site-specificttachment of the key catalytic enzyme, PHA synthase, on lithographically patterned surfaces and subsequent addition of (R)-3-hydroxybutyryl-oA substrate allowed us to fabricate spatially ordered polyhydroxybutyrate (PHB) polymeric structures via an in situ enzymatic surface-initiatedolymerization (ESIP). By varying the reaction conditions, we enhanced the growth of PHB on solid surfaces and analyzed the resulting structuresy fluorescence microscopy, atomic force microscopy (AFM), attenuated total reflectance Fourier transform infrared (ATR-FTIR) spectroscopy,nd gel permeation chromatography (GPC). We found that stabilization of smaller PHB granule structures by an addition of bovine serum albuminBSA) was the most important factor for a successful synthesis of a PHB layer up to 1 �m in thickness, consisting mainly of larger cluster assembliesf PHB granules that cover the entire patterned area. Immunofluorescence detection and surface contact angle analysis revealed that BSA washysically bound to the PHB polymer all through the cluster, and reduced the overall hydrophobicity of the polymer surface. Based on informationbtained from AFM, kinetic measurements and various polymer characterization methods, a plausible model for roles of BSA in the enhancement

f PHB formation on surfaces is discussed. Furthermore, by using biotinylated BSA conjugates, we were able to incorporate biotin groups into theHB polymer matrix, thus generating a bioactive surface that can be used for displaying other functional biomolecules through streptavidin–biotin

nteraction on the PHB structures. Because of its versatility, our fabrication strategy is expected to be a useful surface modification tool for numerousiomedical and biotechnological applications. 2007 Elsevier B.V. All rights reserved.

ionan

cem

eywords: PHA synthase; PHB; Enzymatic surface-initiated polymerization; B

. Introduction

The desire to create more complex, functional nanoscale

evices has driven scientists to redirect their attention to theevelopment of bottom-up strategies that employ biologicalolecules. “Bionanofabrication” describes a fabrication pro-

∗ Corresponding author at: Department of Food Science, Cornell University,thaca, NY 14853, United States. Tel.: +1 607 255 7902; fax: +1 607 255 8741.

E-mail address: nn22@cornell.edu (N. Niamsiri).

ucusruth

927-7765/$ – see front matter © 2007 Elsevier B.V. All rights reserved.oi:10.1016/j.colsurfb.2007.05.023

ofabrication

ess that takes advantage of the specificity and/or catalyticfficiency of biological systems to create various types oficro/nanostructures. For example, synthetic DNA has been

sed to align inorganic nanoparticles [1–8], and as a template toreate conductive nanowires [9–14]. Certain viruses have beensed to create ordered arrays of quantum dots [15], and theelf-assembly properties of certain types of proteins such as fer-

itin [16], chaperonin [17], and bacterial S-layers [18] have beensed to construct two-dimensional periodic arrays of nanopar-icles. Recently, selected enzymatic polymerization processesave been employed to create polymeric micro/nanostructures

faces

if

scpssfiIicalSacreaotSlsekDoboya

ctpabAsostc

mIicetpttocortt[iptPm

hi[eeo(smphf[trramc

N. Niamsiri et al. / Colloids and Sur

n situ on surfaces that can be further used for novel surfaceunctionalization strategies [19–25].

Solid surfaces modified with polymeric materials have beenhown to have a wide range of applications, such as protectiveoatings, controlled adhesion, and providing improved biocom-atibility [26–34]. Among the methods for modifying solidurfaces, surface-initiated polymerization (SIP) has been inten-ively studied and shown to produce more robust and denserlms than conventional spin coating or grafting [26,30,34].n SIP, the reactive group that initiates the polymerization ismmobilized onto the surface, and the propagating polymerhain is grown directly from the surface. SIP is particularlyttractive because it is compatible with both micro- and nano-ithography processes [35]. Several studies have demonstratedIP of various monomers using techniques such as cationic,nionic, free-radical, and metal alkoxide-based initiation pro-esses [26,30,34,36–39]. However, most of these methods oftenequire rigorous reaction conditions and/or input of thermalnergy. In addition, more environmentally friendly and less haz-rdous catalysts are currently desired to broaden the applicationsf SIP, especially in biomedical areas, where the minimiza-ion of harmful species is critical. Recently, several bio-friendlyIP approaches have emerged which involve the use of bio-

ogical catalysts to synthesize biopolymer structures on solidurfaces under physiological conditions. Chow et al. [20] forxample, demonstrated a SIP strategy using a biological enzymenown as terminal deoxynucleotidyl transferase to constructNA nanostructures from oligonucleotide initiators patternedn a surface. Our laboratory has successfully developed anotheriologically based SIP process, enabling the bionanofabricationf a biodegradable and biocompatible polymer, polyhydrox-alkanoate (PHA), on various surfaces such as agarose, silicon,nd gold [22–25].

PHAs are one of the eight classes of physiologically signifi-ant organic biopolymers along with polynucleotides, polypep-ides, polysaccharides, polyisoprenoids, lignins, polyphos-hates, and polythioesters. They represent a complex class ofliphatic polyesters that are synthesized by most genera ofacteria and members of the family Halobacteriaceae of therchaea [40–44]. PHAs accumulate as non-crystalline inclu-

ions or granules (typically 0.2–0.5 �m in diameter) when

ne or more nutritional elements, such as nitrogen, phosphate,ulfur, iron, oxygen, potassium or magnesium, is limiting inhe presence of an excess of carbon sources. The polymeran constitute as much as 90% of the dry cell weight, with a

s(ut

Fig. 1. (a) PHB synthesis by PHA synthase from W. eutropha. (b) Schemat

B: Biointerfaces 60 (2007) 68–79 69

olecular weight ranging from of 2 × 105 to 3 × 106 [40,45].n the past 50 years, PHAs have been a subject of intensenterest due to their promising use as biodegradable and bio-ompatible substitutes for petroleum-based thermoplastic andlastomeric materials. Because of their properties, the poten-ial applications for PHAs span from biodegradable consumerackaging to drug delivery, surgical sutures and scaffolds inissue engineering [46–49]. The key enzymes of PHA biosyn-hesis are the PHA synthases, which catalyze the conversionf (R)-hydroxyacyl-Coenzyme A substrates to PHA with theoncomitant release of CoA [50]. There are four major classesf PHA synthases that are distinguished from one another withespect to the primary structure of the enzyme, subunit composi-ion and substrate specificity [40,41,49]. More than 100 differentypes of monomers have been reported as PHA constituents41,49], where the length and specific functional groups presentn the monomers can influence the physical properties of PHAsolymer (i.e. melting point, glass transition temperature, crys-allinity, and biodegradability). The most prevalent polymer ofHAs family is poly-(3-hydroxybutyrate), PHB, which has aethyl side chain.In the past, PHA synthases from several bacterial species

ave been investigated, with the focus primarily on the kinet-cs and mechanisms of PHA granule formation in solution51–58]. Among these, the PHA synthase from Wautersiautropha H16 (64 kDa) is the most extensively studied. Thisnzyme has been shown to catalyze an in vitro polymerizationf (R)-3-hydroxybutyryl-Coenzyme A, (3HB-CoA), to PHBFig. 1a) in an aqueous solution, resulting in a formation ofpherical polymer granules with diameters of up to 3 �m andolecular weights up to 1.3 × 107 Da [51,52,54]. The in vitro

olymerization of PHB displays an initial lag phase, whichas been speculated to be associated with the requirementor these enzymes to form dimers in order to become active51,59]. Several in vitro polymerization studies have shown thathe dimerization of PHA synthase can be promoted by eithereacting (pre-loading) the enzyme with substrate (priming theeaction) [60,61] or by adding multihydroxyl compounds (suchs glycerol or fructose) to the enzyme solution, which pro-otes the conversion of monomer to dimer due to the molecular

rowding effect [59,62].In previous work, we have demon-

trated an in vitro enzymatic surface-initiated polymerizationESIP) of PHB on agarose and homogenous silicon surfacessing immoblized N-terminal decahistidine-tagged PHA syn-hases from W. eutropha H16 (His10-tagged PhaCWe) [23].

ic illustration of PHB bionanofabrication on gold patterned surfaces.

7 urface

Bgeo[busetiidapwsommultebsmtvtsbrB

2

2

2C(tdNwa

2

(ECBc

apootbspVtf1sawrTA571w(ecBa1

2

mwwaSspDJ(pppodbEoie

0 N. Niamsiri et al. / Colloids and S

y feeding 3HB-CoA substrates to the system, distinct PHBranular structures were formed on the surfaces, with diam-ters ranging from 200 to a maximum of 600 nm dependingn the enzyme concentration used during the immobilization23,24]. The height profile of the PHB structures was found toe roughly equivalent to the diameter of individual PHB gran-les formed as a monolayer at the interface. However, completeurface coverage with PHB has remained difficult to achieve,ven when the surface has been treated with high concen-ration of enzyme [24]. In the present study, we describe anmproved ESIP method to synthesize PHB in situ on lithograph-cally defined, gold patterned surfaces that results in a higheregree of surface coverage, as well as polymer structures thatre significantly taller than previously reported. Gold micro-atterned surfaces (10 �m × 10 �m squares with a 25 �m pitch)ere used in this study for a number of reasons. It demon-

trates the capability to site-specifically grow PHB structuresn a solid surface. Also, the patterning allows for an easier andore accurate analysis of PHB structures using atomic forceicroscopy (AFM) and other imaging techniques. In addition,

sing gold surfaces allows us to create self-assembled mono-ayers (SAMs) of alkanethiols that can be chemically modifiedo present oriented reactive groups on the surface for effectivenzyme immobilization [63,64]. By adding additives such asovine serum albumin (BSA) during the immobilization andurface polymerization steps, we were able to augment the for-ation of PHB on the surfaces. The role of BSA in enhancing

he surface growth of PHB was further investigated througharious techniques. Taken together, the data strongly suggesthat direct physical binding of BSA to the PHB polymerictructure occurs, presumably via hydrophobic interactions. Thisinding of BSA is used to demonstrate a strategy for incorpo-ating functional groups into PHB structures using biotinylatedSA.

. Experimental

.1. Materials

5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB), N-methyl--pyrrolidone, dl-�-hydroxybutyryl coenzyme A (3HB-oA), 16-mercaptohexadecanoic acid, trifluoroacetic anhydride

TFA), triethylamine (TEA), N, N-dimethylformamide (DMF),etrahydrofuran (THF), sodium hydroxide (NaOH), ethylene-iaminetetraacetic acid (EDTA), nickel chloride (NiCl2), N,-bis-(carboxymethyl)-l-lysine-hydrate, and triethylene glycolere purchased from Sigma–Aldrich (St. Louis, MO) and used

s received.

.2. Bacterial growth and enzyme purification

His10-tagged PHA synthase of W. eutropha H16 (DSM428)His10-tagged PhaCWe) was purified from recombinant

scherichia coli strain BL21 (DE3) pLysS (Novagen)/pET-O1 as described elsewhere with some modifications [22,23].riefly, a 1-l culture of recombinant E. coli BL21 (DE3) pLysSells harboring His10-tagged PhaCWe was grown aerobically

2

t

s B: Biointerfaces 60 (2007) 68–79

t 37 ◦C in Luria–Bertani broth (Bacton, Dickson and Com-any) containing ampicillin (50 �g/ml) to an optical densityf 0.5 at 600 nm. The culture was then induced by additionf 200 �M isopropyl-�-d-thiogalactopyranoside (Sigma) andransferred to a 30 ◦C shaking incubator for additional incu-ation for 5 h. The cells were harvested by centrifugation andtored at −80 ◦C until purification. His10-tagged PhaCWe wasurified using a NTA-Ni2+ agarose chromatography (Qiagen,alencia, CA) under native condition according to manufac-

urer’s instructions with some modifications. Briefly, frozen cellsrom a 500 ml culture were resuspended in 50 mM Tris–HCl,0% (v/v) glycerol, 10 mM imidazole (pH 7.5) and lysed byonication on ice for 1 min. The crude extract was centrifugedt 10,000 rpm for 30 min at 4 ◦C and the clear supernatantas further purified by metal chelating column chromatog-

aphy with NTA-Ni2+ agarose pre-equilibrated with 50 mMris–HCl, 10% (v/v) glycerol, 10 mM imidazole (pH 7.5).fter loading the supernatant, the column was washed with0 mM Tris–HCl, 10% (v/v) glycerol, 20 mM imidazole (pH.5). His10-tagged PhaCWe was eluted with 50 mM Tris–HCl,0% (v/v) glycerol, 250 mM imidazole (pH 7.5). The proteinas further purified on a gel filtration chromatography column

GFC; PD-10 columns, Pharmacia, Freiburg, Germany) pre-quilibrated with 50 mM Tris–HCl (pH 7.5). The final proteinoncentration was determined by the Bradford method usingSA as a standard [65]. The purified His10-tagged PhaCWe wasliquoted and stored at −80 ◦C after addition of glycerol to0% (v/v).

.3. Fabrication of gold patterns on silicon wafers

Gold patterns on silicon wafers were produced using standardicrofabrication and lithographic techniques described else-here with some modifications [22,23,25]. Briefly, a Si waferas first cleaned with Pirhana solution (a mixture of sulfuric

cid/hydrogen peroxide) using an automatic wafer processor forC-1 & Pirhana cleans (Hamatech, Sternenfels, Germany) andubjected to molecular vapor deposition of (1H, 1H, 2H, 2H-erflorooctyl)-trichlorosilane (FOTS) using a Molecular Vaporeposition MVD-100 System (Applied MicroStructures, San

ose, CA). Subsequently, the wafer was coated with photoresistShipley S1813), and baked at 110 ◦C for 90 s on a contact hotlate. Patterns (arrays of 10 �m × 10 �m squares with a 25 �mitch) were transferred using an EVG620 Contact Aligner. Afterost-exposure baking at 110 ◦C for 60 s, patterns were devel-ped for 1 min in MIF 300 (Shipley). After a brief UV/O3escum process, a 10 nm chromium adhesion layer followedy a 30 nm gold layer were deposited using a CHA Mark 50-Beam Evaporator (CHA Industries, Fremont, CA). A gold lift-ff process was performed by submerging the wafer overnightn N-methyl-2- pyrrolidone and then rinsed thoroughly withthanol.

.4. Modification of gold surfaces with NTA-Ni2+

Gold surfaces were functionalized with NTA-Ni2+ forhe immobilization of His-tagged proteins as described pre-

faces

vs1touafaoaNgTbwttdas

2a

toiip0wfeostat2Tdgs

i1s(actrpt5

ap1pdc[otdc

2p

vssit4srjtIba1tatPad(Frw2

mspfpSWSttb

N. Niamsiri et al. / Colloids and Sur

iously with some modifications [24,66]. Briefly, a goldurface was immersed in an ethanol solution containing 2 mM6-mercaptohexadecanoic acid overnight. The resulting func-ionalized gold surface, containing a self-assembled monolayerf the carboxylic acid, was rinsed with ethanol, blown drynder a stream of nitrogen, and, subsequently immersed insolution of 0.1 M TFA and 0.2 M TEA in anhydrous DMF

or 30 min at room temperature to generate reactive interchainnhydride groups. After reaction, the surface was rinsed thor-ughly with DMF, blown dry under a stream of nitrogen,nd immediately immersed in a solution containing 10 mM, N-bis-(carboxymethyl)-l-lysine-hydrate, 10 mM triethylenelycol, and 7.5 mM NaOH at room temperature for 30 min.he gold surface was then rinsed with deionized water andlown dry under a stream of nitrogen. NTA-Ni2+ activationas performed by immersing the gold surface in a solu-

ion of 40 mM NiCl2 at room temperature for 2 h. Finally,he functionalized gold surface was rinsed thoroughly witheionized water and blown dry under a stream of nitrogen,nd cut into small pieces (1 cm2) using a diamond-tippedtylus.

.5. Bionanofabrication of PHB on gold patterned surfacesnd activity assay of immobilized PHA synthase

The fundamental in situ bionanofabrication of PHB on pat-erned surfaces involves two main steps; an immobilizationf PHA synthase and an enzymatic polymerization as shownn Fig. 1b. The immobilization was carried out by incubat-ng the NTA-Ni2+ functionalized gold surface in PBS buffer,H 7.4, (0.14 M NaCl, 0.01 M Na2HPO4, 0.003 M KCl and.0018 M KH2PO4) containing 200 �g/ml His10-tagged PhaCWe

ith slight agitation for 2 h at room temperature. The sur-ace was then washed thoroughly with PBS buffer to removexcess enzymes, and used immediately. Surface polymerizationf 3HB-CoA was carried out by immersing the enzyme-coatedurface into a monomer solution (PBS buffer, pH 7.4, con-aining 10 �g/ml Nile blue dye (Sigma) and 5 mM 3HB-CoA)t room temperature for 12 h or more with continuous agita-ion. For optimization experiments, various additives includingmg/ml BSA (Sigma), 0.03 mM Tween-20 (Sigma), 0.1 mMriton X-100 (Sigma), and 4 mM CHAPS (Sigma) were addeduring the immobilization and polymerization steps to investi-ate their influence on the polymer formation on the patternedurfaces.

For studying enzymatic activity, His10-tagged PhaCWe wasmmobilized onto functionalized planar gold substrates (cut tocm2 in size). The amount of synthases immobilized to the gold

urface was quantified using Micro BCATM Protein Assay KitPierce, Rockford, IL) according to the manufacturer’s protocolfter being removed from the surface using a 3% sodium dode-yl sulfate (Sigma) solution in PBS (pH 4.5). The activity ofhe immobilized PHA synthases was assayed spectrophotomet-

ically by measuring the amount of free thiol released duringolymerization as follows: A 10-�l aliquot of the reaction mix-ure was taken at various time points and mixed with 390 �l ofmM sodium acetate buffer (pH 4.7) containing 50 mM NaCl

wRfia

B: Biointerfaces 60 (2007) 68–79 71

nd 0.5 mM EDTA. A total of 590 �l of 40 mM sodium phos-hate buffer (pH 7.6) containing 2 mM EDTA and 10 �l of00 mM DTNB was subsequently added and mixed at room tem-erature for 2 min. Absorbance was measured at 412 nm usingistilled water as a blank. The concentration of free thiol was cal-ulated using a molar extinction coefficient of 13,600 cm−1 M−1

67,68]. One unit of enzyme activity is defined as the amountf enzyme (mg) required to convert 1 �mol of substrateo product in 1 min. The specific activity (unit/mg) wasetermined by analyzing the linear part of the conversionurve.

.6. Characterization of PHB synthesized on goldatterned surfaces

Nile blue was added into the reaction mixture for specificisualization of in situ synthesized PHB on the gold patternedurfaces as previously described. After the polymerization, theurface synthesized PHB was washed thoroughly with deion-zed water and observed by fluorescence microscopy, wherehe excitation and emission wavelength of Nile blue are at50–630 nm and 625–725 nm, respectively [69]. All surfaceynthesized PHB samples were kept in deionized water atoom temperature and blown dry under a stream of nitrogenust prior to conducting analysis. The surface topography ofhe PHB surface structures was examined by AFM (Digitalnstrument Dimension 3100, Veeco Instruments Inc., Wood-ury, NY) using cantilevers with spring constants of 20–100 N/mnd nominal tip radii of 5–15 nm. Areas of 15 �m × 15 �m,0 �m × 10 �m, or 5 �m × 5 �m squares were scanned inapping modeTM, with the cantilevers operating with a freemplitude between 3 and 4 V at 330 ± 50 kHz. Images wereaken with a 0.5–1.0 Hz scan rate. The diameter of an individualHB granule was measured by AFM horizontal cross-sectionnalysis. At least 400 granules were analyzed for each con-ition. Attenuated total reflectance Fourier transform infraredATR-FTIR) spectra were recorded using a Bruker Vertex80vTIR spectrometer (Bruker Optics, Billerica, MA) at a spectralesolution of 4 cm−1. All spectra were recorded with 256 scans,here the sample chamber was held at a vacuum better than.3 hPa.

The weight-average molar mass (Mw), number-averageolar mass (Mn), and polydispersity index (PDI) of PHB

ynthesized on the gold surfaces were determined by gelermeation chromatography (GPC) after extraction with chloro-orm. Calibration was performed with narrow-molecular weightolystyrene standards (PSS Polymer Standards Service, Silverpring, MD). GPC data were collected and analyzed with PSSINGPC scientific V6.20 (PSS Polymer Standards Service,

ilver Spring, MD). GPC measurements were performed inriplicate and averaged. Relative hydrophobicity of PHB syn-hesized on flat, non-patterned, gold surfaces was evaluatedy measuring the surface contact angle of a drop of distilled

ater (3-�l) using a NRC contact goniometer (model 100-00,ame-Hart Inc., Mountain Lakes, NJ). For each angle reported,ve measurements from different surface locations wereveraged.

7 urface

2c

faA(swT7ab4fwCs

pwwcnwTaw

2P

w(tdtnDoQuisobirtbastPfl

3

3Pv

PaaPowftwwrt2(ebHttaa

0citee(rCstwieia

aat((sb

2 N. Niamsiri et al. / Colloids and S

.7. Immunofluorescence detection of BSA proteins andonfocal analysis of fluorescent BSA

To investigate if BSA became incorporated into the sur-ace synthesized PHB structures, a rabbit anti-BSA primaryntibody (Molecular ProbesTM, Invitrogen, Carlsbad, CA) andlexa Fluor® 488-conjugated anti-rabbit secondary antibody

Molecular ProbesTM) were used. In this experiment, PHB wasynthesized on gold patterned surfaces as described earlier, butithout the presence of Nile blue dye in the reaction mixture.he fabricated PHB surface was immersed in PBS buffer (pH.4) containing a 1:1000 dilution of rabbit anti-BSA primaryntibody. After 1 h, the surface was rinsed thoroughly with PBSuffer (pH 7.4) and immediately incubated with Alexa Fluor®

88-conjugated anti-rabbit secondary antibody (1:1000 dilution)or 1 h. The surface was then subsequently washed extensivelyith deionized water and visualized by fluorescence microscopy.ontrols were performed, where BSA, primary antibody, or

econdary antibody was absent.In order to determine whether the BSA molecules became

hysically bound through the entire PHB polymer matrix, PHBas synthesized on gold patterned surfaces as described earlierithout the presence of Nile blue, and Alexa Fluor® 488-

onjugated BSA (Molecular ProbesTM) was used instead ofative BSA during the polymerization. The resulting surfaceas visualized by confocal laser scanning microscopy (LeicaCS SP2 system, Leica Microsystems Inc., Bannockburn, IL),nd series of cross-section images with 200 nm-sectional viewsere taken.

.8. Biotinylation of BSA and creation of biotin-derivatizedHB surfaces

For the preparation of biotinylated BSA (bBSA), BSAas reacted with sulfosuccinimidyl-6-(biotin-amido) hexanoate

Sulfo-NHS-LC-Biotin) (Pierce) according to the manufac-urer’s instructions. Briefly, the indicated amount of biotin esterissolved in ultrapure water was incubated with BSA solu-ion (2 mg/ml in PBS) for 1 h at room temperature. Excesson-reacted and hydrolyzed biotin reagent was removed using-saltTM Dextran Desalting columns (Pierce) and the levelf biotin incorporation was determined using EZTM Biotinuantification Kit:HABA assay (Pierce) according to the man-facturer’s method. The number of biotin units incorporatednto each BSA molecule was calculated to be approximatelyix biotin molecules per BSA molecule. Bionanofabricationf PHB–bBSA on the gold patterned surfaces was performedy immobilization of His10-tagged PhaCWe onto the surfacen the presence of native BSA (2 mg/ml in PBS buffer) atoom temperature for 2 h, and subsequent surface polymeriza-ion of 3HB-CoA in the presence of bBSA (2 mg/ml in PBSuffer containing 5 mM 3HB-CoA) overnight at room temper-ture. In order to verify the presence of biotin groups on the

urface synthesized PHB, Alexa Fluor® 488-conjugated strep-avidin (Molecular ProbesTM) was incubated with the fabricatedHB–bBSA surface. The resulting surface was visualized usinguorescence microscopy.

Iht1

s B: Biointerfaces 60 (2007) 68–79

. Results and discussion

.1. Enzymatic surface-initiated polymerization (ESIP) ofHB on gold patterned arrays under the influence ofarious additives

Previously, Kim et al. [23] demonstrated that His10-taggedhaCWe immobilized onto Ni2+-NTA derivatized surfaces suchs silicon and agarose beads without additives was still activend able to catalyze the polymerization of (R)-3HB-CoA toHB. In this study, we show that PHB could also be synthesizedn a gold micro-patterned surface (an array of 10 �m × 10 �mith a 25 �m pitch). The PHB synthesized without any additives

ormed very distinct and granular structures, and appeared as ahin film covering the gold patterned surfaces when observedith fluorescence microscopy (Fig. 2c), similar to previousork [23]. Tapping mode AFM micrographs of this surface

evealed that the film was about 100 ± 15 nm thick (Fig. 3b), andhat the average size of the PHB granules was approximately00–300 nm in diameter (Fig. 4a–c). Larger PHB granules>300 nm in diameter) were also observed, but to a lesserxtent.The height of the bare gold patterns was measured toe 40 ± 3 nm from the surface of the silicon wafer (Fig. 3a).owever, for growth conditions without additives, we noticed

hat there were a number of PHB granules synthesized outsidehe gold patterned surfaces, which might due to non-specificdsorption of active PHA synthases onto those surroundingreas.

When mild non-ionic detergents (0.03 mM Tween-20,.1 mM Triton X-100, and 4 mM CHAPS) were added at aoncentration below their critical micelle concentrations dur-ng both the immobilization and polymerization steps, in ordero minimize non-specific adsorption and also evaluate theirffect on the ESIP PHB, no PHB granules was formed onither the patterned gold surface or the surrounding areasSee Supporting Information, Fig. S1). Roo et al. [70] haveeported that detergents such as Tween-20, Triton X-100, andHAPS all appeared to be strong inhibitors of purified PHA

ynthase from Pseudomonas oleovorans GPo1 (PhaCPo) forhe in vitro synthesis of PHA in aqueous solution, and itas suggested that they might interfere with protein–protein

nteractions, disrupting the dimerization of the PHA synthasenzyme subunits. Our results show that these detergents stronglynhibit the formation of PHB by surface anchored enzymess well.

In contrast to the above results, we found that when BSA wasdded into the reaction system during both the immobilizationnd polymerization steps, more PHB polymers were formed onhe gold patterned surfaces and less on the surrounding areasFig. 2d) as compared to when BSA was absent from the systemFig. 2c). This result shows that BSA, not only reduces non-pecific adsorption of the synthases to the background areas,ut also enhances the PHB synthesis on the patterned surfaces.

n the presence of BSA, the gold patterns were completely andomogenously covered with larger PHB granular cluster struc-ures. The height of the PHB structures was measured to be± 0.2 �m above the gold surface, which was almost 10-times

N. Niamsiri et al. / Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79 73

F poly3 d arraw e abs

ho1esitHtlcdriBf

3S

sBatpods

Fa

ig. 2. Fluorescence analysis of synthesized PHB on gold arrays. PHB surfaceHB-CoA, and 100 �g/ml Nile blue for 12 h. (a) An optical micrograph of a golith no added 3HB-CoA substrate (negative control). (c) PHB synthesized in th

igher than the height of the PHB film synthesized in the absencef BSA (Fig. 3c). The diameter of these PHB clusters was about–1.5 �m (Fig. 4a, d, and e). AFM analyses revealed an inter-sting polymer surface topography, wherein large PHB clusterseemed to be comprised of distinct granules with 200–300 nmn diameter that were similar in size to individual granules syn-hesized on the surface in the absence of BSA (Fig. 4c and e).iraishi et al. [71] reported that in vitro PHB synthesis in solu-

ion could result in cluster formation where the formation ofarger PHB structures (200–700 nm in diameter) appeared to beonsisting of smaller individual PHB granules (100–250 nm iniameter). Because we found that the addition of BSA into the

eaction mixture had the greatest stimulatory effect (resultingn more PHB formation at the interface), the possible roles ofSA in the surface polymerization of PHB were investigated in

urther detail.

bb

i

ig. 3. AFM analysis measuring the height profile of PHB structures synthesized obsence of BSA, and (c) PHB synthesized in the presence of 2 mg/ml BSA.

merization on the patterned surfaces was carried out in the presence of 5 mMy consisting of 10 �m × 10 �m squares. (b) Immobilized His10-tagged PhaCWe

ence of BSA, and (d) PHB synthesized in the presence of 2 mg/ml BSA.

.2. Investigating the Effect of BSA on ESIP of PHBynthesis on Gold Surfaces

In order to investigate in more detail how BSA enhances theurface synthesis of PHB, experiments were carried out, whereSA was added, either to the enzyme immobilization reactionlone, to the PHB polymerization reaction alone, or to both reac-ions successively. We found that when BSA was present in theolymerization step, PHB growth was maximized. This is thenly condition under which large PHB clusters of 1–1.5 �m iniameter with a height profile of ∼1 �m formed on the patternedurface. However, we found that in order to reduce non-specific

ackground adsorption, BSA had to be present during the immo-ilization step.

To further examine the effect of BSA on the activity of themmobilized His10-tagged PhaCWe, an in vitro activity assay of

n gold patterned arrays. (a) Patterned gold alone, (b) PHB synthesized in the

74 N. Niamsiri et al. / Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79

F or prs sencea m sc

ttrsFsstWtcectioart(

miiaiittaol(sobt

ig. 4. (a) Size distribution of PHB granules synthesized either in the absenceurfaces (a gold patch of 10 �m × 10 �m squares). PHB synthesized in the abbove image. PHB synthesized the presence of 2 mg/ml BSA: (d) 15 �m × 15 �

he immobilized enzyme was performed based on the detec-ion of released CoA using DTNB, a compound that specificallyeacts with thiol groups [67,68]. The formed adduct was mea-ured spectrophotometrically at a wavelength of 412 nm [70].or this study, we immobilized PHA synthase on a planar goldurface (1 cm2). The amount of synthase immobilized on thisurface either in the absence or presence of BSA was measuredo be 1.49 ± 0.1 �g/cm2 and 1.51 ± 0.2 �g/cm2, respectively.

e are aware that obtaining a precise direct quantification ofhe amount of synthase immobilized onto the surface is quitehallenging due to the following issues. First, all immobilizednzymes might not be completely removed under the employedondition (3% SDS solution in PBS at pH 4.5 with harsh agita-ion for 24 h). Second, when BSA is present during the enzymemmobilization, there is a possibility for BSA to co-adsorbnto the surface as well. As a result, we analyzed the surface

fter the protein desorption experiment by using immunofluo-escence detection against PHA synthase and/or BSA, and foundhat no significant amount of proteins was left on the surfacedata not shown). This result shows that our protein desorption

bgbb

esence of BSA. AFM analysis of surface synthesized PHB on gold patternedof BSA: (b) 15 �m × 15 �m scan size, and (c) 5 �m × 5 �m scan size of the

an size and (e) 5 �m × 5 �m scan size of the above image.

ethod is effective, and thus the amount of proteins measuredn the desorption solution represents the amount of proteinsmmobilized on the surface. To investigate whether BSA co-dsorbs with PHA synthase onto the surface, we employed themmunofluorescence detection against BSA directly after themmobilization step, and found that BSA indeed was present onhe surface (Fig. 5a). Therefore, the amount of adsorbed pro-eins quantified when BSA is present represents the combinedmount of both proteins bound to the surface. Since the amountf removed proteins from the surface was found to be simi-ar whether or not BSA is present during the immobilizationapproximately 1.5 �g/cm2 for both conditions), these findingsuggest that when BSA is present in the system, the true amountf His10-tagged PhaCWe immobilized onto the surface shoulde somewhat less than 1.5 �g/cm2. The presence of BSA duringhe immobilization step might physically interfere the binding

etween His10-tagged PhaCWe and NTA-Ni2+ functionalizedold surfaces, lowering the amount of PHA synthases immo-ilized onto the surface. However, its presence might still beeneficial to the system. BSA has been used as a stabilizer of

N. Niamsiri et al. / Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79 75

F ) Funco a Fluoa d subs

viihnaan

tssthisfPtpspaawi

Fbetrd

btEltfbf

smiapffpcross-section analyses of surface synthesized PHB in the pres-ence of Alexa Fluor® 488-conjugated BSA confirms the abilityof BSA to bind within the polymer structure, and also reveals that

ig. 5. Immunofluorescence detection of BSA incorporated into the surfaces. (af BSA, then subjected to anti-BSA primary antibody and subsequently to Alexrrays in the presence of BSA, then subjected to anti-BSA primary antibody an

arious proteins in a variety of biological buffers [72–74]. Thus,t might improve the stability of the PHA synthase during themmobilization as well. However, most likely the main effect ofaving BSA present during the immobilization step is to reduceon-specific adsorption of active synthases to the backgroundreas of the patterned surfaces (i.e. a passivation effect). This iswell-known phenomena and BSA is frequently used to reduceon-specific binding [75].

Fig. 6 shows that whether or not BSA is added to the sys-em, there is no longer an apparent lag phase present in theurface synthesis reaction of PHB, compared to the in vitroolution phase synthesis of PHB by free, non-immobilized syn-hases, where a lag time of 4 min is typically seen [23,51]. Itas been suggested that the disappearance of the lag phase withmmobilized PHA synthase can be due to the possibility that theynthase is already immobilized in a dimeric form (an activeorm), and thus can immediately initiate the polymerization ofHB polymer upon the addition of substrates [23]. In Fig. 6,

he polymerization times required for the reaction to reach thelateau are quite similar whether or not BSA is present in theystem (both are approximately 5 h). However, when BSA isresent, the polymerization reaches the plateau at a much higherbsorbance value, indicating more polymer growth. The specific

ctivity of the immobilized PHA synthase in the absence of BSAas calculated to be 1.47 units/mg. The specific activity of the

mmobilized PHA synthase in the presence of BSA could not

ig. 6. Time course of CoA release from in vitro catalyzed 3HB-CoA by immo-ilized His10-PhaCWe on planar gold surfaces (1 cm2) for comparative kineticsither in the absence of BSA (control) or presence of 2 mg/ml BSA. The reac-ion mixtures contain 5 mM 3HB-CoA. Slope (Δ) of the linear part of the curveeflecting the velocities of the reactions. The error bars represent the standardeviation of triplicate experiments for each condition.

FsF

tionalized planar gold surface immobilized with PHA synthase in the presencer® 488-conjugated secondary antibody. (b) PHB synthesized on gold patternedequently to Alexa Fluor® 488-conjugated secondary antibody.

e obtained, since the exact amount of immobilized PHA syn-hase when BSA is present could not be directly determined.ven so, the activity of immobilized PHA synthase should be at

east three-fold higher as compared to when BSA is absent fromhe system (Fig. 6). These results, therefore, suggest that BSAacilitates the utilization of 3HB-CoA substrate by the immo-ilized PHA synthases, thus increasing the overall rate of PHBormation on the surface.

In particular, it is possible that BSA might interact with andtabilizes the nascent PHB polymer chains in order to allowore substrates to reach the immobilized enzymes below. To

nvestigate this, we performed immunofluorescence detectionnd found that BSA was indeed physically bound to the PHBolymeric structures synthesized on the gold patterned sur-aces (Fig. 5b). We also noted some background fluorescencerom BSA bound to the surrounding areas, consistent with itsassivation effect described earlier. The results from confocal

ig. 7. Confocal laser scanning microscopy sectional images of surface synthe-ized PHB on a gold pattern (10 �m × 10 �m) in the presence of 2 mg/ml Alexaluor® 488-conjugated BSA. It shows 200-nm sectional views.

76 N. Niamsiri et al. / Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79

Table 1Average molar mass and contact angle of PHB synthesized via ESIP

Type of sample Mn (10−5 g/mol) Mw (10−5 g/mol) PDI Contact angle (H2O) advancing Contact angle (H2O) receding

PHB alone 0.59 ± 0.04 1.06 ± 0.09 1.80 ± 0.05 62◦ ± 2 54◦ ± 3PHB–BSA 0.62 ± 0.08 1.13 ± 0.15 1.81 ± 0.09 36◦ ± 4 22◦ ± 3

N ersitys

B(odbtipiBmhliosos

3

opfipI2ap

ppsasaBpnomostapm

batf1swfhcTdiwsssfets

3

dspstTrapacaodaltpb

umber-average molar mass (Mn), weight-average molar mass (Mw), polydispurfaces (1 cm2) either in the absence or presence of 2 mg/ml BSA.

SA is incorporated throughout the entire PHB polymer matrixFig. 7). BSA is a well-known protein capable of binding to vari-us types of hydrophobic compounds, particularly hydrophobicrugs in blood plasma [72], and PHB is generally regarded aseing a hydrophobic polymer [49,61]. In fact, the ability of BSAo bind to PHB has been evidenced by Reusch et al. [76], whodentified “albumin” as the major PHB-binding protein in humanlasma. As a result, a possible explanation as to why the bind-ng of BSA might improve the PHB surface synthesis is thatSA binding increases the hydrophilic nature of the PHB poly-er chain, and thus keeps it from collapsing, presumably due to

ydration forces [77]. This could then, in turn, give the immobi-ized enzymes a better access to substrate and subsequently resultn an increased rate of polymer synthesis. Combining the abovebservations, we believe that BSA improves the overall surfaceynthesis of patterned PHB by reducing non-specific adsorptionf the synthases to the background areas, and concomitantlytabilizing the hydrophobic PHB polymer chains.

.3. Characterization of PHB synthesized on gold surfaces

In order to chemically verify that the polymer synthesizedn the surface either in the absence or presence of BSA is PHBolymer, ATR-FTIR analyses were performed on the polymerlms synthesized on planar gold surfaces either in the absence orresence of BSA. The spectra results were shown in Supportingnformation, Fig. S2. The C H stretching peaks at 2991 and932 cm−1, the C O carbonyl stretching peak at 1725 cm−1,nd the C O ester peak at 1280 cm−1 are all indicative of theresence of PHB on the surface [78].

Interestingly, we found that the Mw, Mn and PDI of the PHBolymer synthesized on the gold surface either in the absence orresence of BSA are similar (Table 1). The Mw and PDI of PHBynthesized in the absence of BSA are 1.06 × 105 ± 0.09 g/molnd 1.8 ± 0.05, respectively, while the Mw and PDI of PHBynthesized in the presence of BSA are 1.13 × 105 ± 0.15 g/molnd 1.81 ± 0.09, respectively. From these results, it appears thatSA has no influence on the molecular weight or the polydis-ersity of PHB synthesized on the surface. Therefore, BSA doesot affect the processivity of the immobilized synthase, but maynly affect the turnover rate of the enzyme (i.e. more PHB poly-er chains synthesized per active enzyme molecule). Since little

r no change was observed in the molecular weight of the surfaceynthesized PHB, this indicates that BSA does not act as a chain

ransfer agent in the ESIP process. However, there seems to belimitation in the ultimate molecular mass of the growing PHBolymer via ESIP. Gerngross and Smith [51] reported that theolecular weight of PHB synthesized in vitro was not affected

t

th

index (PDI), and contact angle for water of PHB synthesized on planar gold

y either reaction time or initial substrate concentrations. Theylso reported that excess amounts of substrate did not increasehe molecular mass of the PHB synthesized, and resulted in aormation of PHB polymer with a molecular weight of about.2 × 107 g/mol, which may be the size limit of PHB synthe-ized in vitro in solution [51]. In this study, the lower moleculareight of PHB synthesized on the surface might be due to the

act that the synthases are immobilized to the surface, thus mayave their conformation and/or orientation more restricted asompared to the free, non-immobilized enzymes in solution.he conformational constraint might impose the lower limit ofegree of polymerization occurred at the catalytic site of themmobilized synthase. According to goniometry measurements,e found that the surface contact angle of the PHB film synthe-

ized in the presence of BSA was much lower than the PHB filmynthesized in the absence of BSA (Table 1). This result alsoupports the fact that BSA, which contains various hydrophilicunctional groups (such as carbonyl, amine, carboxylic acid,tc.), is incorporated into the PHB structures, thus decreasinghe overall hydrophobicity of the PHB film synthesized on theurface.

.4. Mechanism of PHB formation on gold surfaces

On the basis of our AFM observations, and combined withata from the kinetic measurements, the immunofluorescencetudies, and the molecular weight determinations, we propose alausible mechanism of how BSA enhances the ESIP of PHB onolid surfaces as shown in Fig. 8. In the absence of BSA, His10-agged PhaCWe dimers become immobilized to the surface.he immobilized synthases react with the substrate, 3HB-CoA,

esulting in the polymerization of PHB polymer chains, whichre extruded out from the immobilized enzymes. Since the PHBolymer chain is hydrophobic in nature, the polymerization inn aqueous solution causes the hydrophobic polymer chains tooil up and instantaneously form PHB granular structures withdiameter of 200–300 nm on top of the immobilized enzymesn the surface. These individual PHB granules can also ran-omly aggregate to generate larger granular structures, if theyre located close enough to one another. Since there is mostikely a lower degree of access for new substrates to diffuse tohe immobilized enzymes located underneath the synthesizedolymeric structures, only a limited amount of substrate woulde able to reach the catalytic sites over a prolonged period of

ime.

When BSA proteins are present in the reaction mixture,hey can stabilize the PHB structures by binding to the nascentydrophobic polymer chain, perhaps as soon as it is synthe-

N. Niamsiri et al. / Colloids and Surfaces B: Biointerfaces 60 (2007) 68–79 77

synt

stfTesmwseiBitdmeisl

PthWia

3p

muwf(

FFt

Fig. 8. Schematic illustration of the plausible mechanisms of PHB

ized and protrudes out of the enzyme. By directly binding tohe polymer, BSA may prevent the hydrophobic polymer chainsrom collapsing immediately above the immobilized enzymes.his would allow more substrates to diffuse to the immobilizednzymes below, and thus increase the overall rate of polymerynthesis. Since there seems to be a limitation in the ultimateolecular mass of each polymer chain synthesized via ESIPhich is not affected by the duration of polymerization, initial

ubstrate concentrations or BSA, the PHB polymers synthesizedither in the absence or presence of BSA do not differ signif-cantly in their average molecular weights. In the presence ofSA, as more and more PHB granules with size ∼200–300 nm

n diameter are formed, they begin to aggregate with one anothero form larger PHB clusters having sizes of ∼1–1.5 �m iniameter. Over time, as more clusters are formed, the PHB poly-eric structures eventually cover the underlying gold surface

ntirely. With the simultaneous consumption of substrate thats provided in excess (5 mM 3HB-CoA), PHB is continuouslyynthesized, pushing the large, previously formed PHB granu-ar clusters upwards from the surface. The maximum height of

aAop

ig. 9. (a) Schematic illustration of biotin functionalized PHB patterned surfaces tluorescence analysis of biotin-derivatized PHB patterned surfaces after the binding

he presence of 2 mg/ml biotinylated-BSA subjected to Alexa Fluor® 488-conjugated

hesis on solid surfaces either in the absence or presence of BSA.

HB polymeric structures formed on the surface thus far appearso be around 1 �m, which is almost 10-times higher than theeight of PHB structures synthesized in the absence of BSA.e also tested other proteins (lysozyme, myoglobin, protein A,

mmunoglobulin G (IgG), and fibrinogen) but none of them gaven increased PHB growth similar to BSA (data not shown).

.5. Bionanofabrication of biotin-derivatized PHBolymeric structures

Since PHB does not contain any functional groups on its poly-er backbone for further chemical modifications, we decided to

se biotinylated BSA during the surface polymerization of PHB,hich allow us to create a novel biotin-functionalized PHB sur-

ace through the physical binding of bBSA to the PHB polymerFig. 9a). Fluorescence analysis of the resulting surface (Fig. 9b

nd c) shows a successful binding of streptavidin-conjugatedlexa Fluor® 488 to biotin groups displayed only on the surfacef PHB structures. This result, thus, indicates the in situ incor-oration of bBSA into the PHB polymer matrix. In addition, the

hrough the incorporation of biotinylated BSA into the PHB polymer matrix.of Alexa Fluor® 488-conjugated streptavidin. (b) Surface synthesized PHB instreptavidin. (c) Zoom in of the left image.

7 urface

bsbdbitcnivmsiop

4

i(pisppfhetPcsf

ifpacAcfttmtmtos

A

N

9UiS

A

i

R

[

[[

[

[

[[

[

[

[[

[

[

[

[

[

[

8 N. Niamsiri et al. / Colloids and S

ackground fluorescence from non-specific adsorption of thetreptavidin onto the surrounding areas was found to be negligi-le, presumably, due to the passivation effect of native BSA useduring the immobilization step. The specificity of streptavidininding to the biotinylated PHB surface was also confirmed byncubating the streptavidin-conjugated Alexa Fluor® 488 withhe PHB surface synthesized in the presence of native BSA as aontrol. We found that the fluorescent streptavidin conjugate didot bind to the PHB–BSA surface, thereby indicating the bind-ng of the streptavidin to the biotinylated PHB–bBSA surface isery specific (data not shown). Furthermore, we found that thisolecular assembly of PHB–bBSA/streptavidin was stable over

everal days and unaffected by vigorous washing. Incorporat-ng biotin into the PHB matrix allows continued incorporationf other biological ligands and/or proteins, thus increasing itsotential utility as a biocompatible platform.

. Conclusions

In summary, we have demonstrated and studied an enhancedn situ bionanofabrication method to create polyhydroxybutyratePHB) polymeric micro/nanostructures on solid surfaces underhysiological conditions. BSA is added to both the enzymemmobilization reaction (to prevent non-specific binding of theynthases) and the polymerization reaction (to stabilize the PHBolymer). In this study, the effect of BSA on stabilizing the PHBolymeric structures was found to be the most important factoror the successful synthesis of PHB structures up to 1 �m ineight with homogenous surface coverage. An additional ben-fit of having BSA physically bound to the PHB polymer ishat it allows incorporation of functional groups throughout theHB polymer matrix. We demonstrate this by using a biotinonjugated BSA which results in in situ reactive PHB surfacetructures, and can thus potentially be useful for constructingunctional biocompatible ESIP structures on solid surfaces.

We believe that the use of these BSA-enhanced PHB surfacesn general offers many practical applications in different areas,or example, the modification of solid surfaces with biocom-atible and biodegradable PHB polymer could potentially yieldnovel type of coated implants for tissue engineering appli-

ations that promote cell attachment and growth [23,48,79].s a result, one of our future goals is to study how different

ells interact and behave on this type of fabricated PHB sur-ace. In addition to biocompatible surface coatings, we envisionhat these PHB polymeric structures can be built into spaceshat cannot be accessed by conventional lithographic tools or by

ost other standard fabrication processes. For example, func-ional PHB microstructures can be formed in situ inside sealed

icrofluidic channels at designated areas. Also, we believe thathe knowledge obtained through this study may help extendur understanding of the mechanism of ESIP of PHB on solidurfaces.

cknowledgements

This research is supported by the STC program of theational Science Foundation under agreement no. ECS-

[

[

s B: Biointerfaces 60 (2007) 68–79

876771 through the Nanobiotechnology Center at Cornellniversity. We also thank Dr. Young-Rok Kim for construct-

ng the His10-tagged PHA synthase from W. eutropha and Dr.onny Mark for helpful discussions.

ppendix A. Supplementary data

Supplementary data associated with this article can be found,n the online version, at doi:10.1016/j.colsurfb.2007.05.023.

eferences

[1] A.P. Alivisatos, K.P. Johnsson, X. Peng, T.E. Wilson, C.J. Loweth, M.P.J.Bruchez, P. Schultz, Nature 382 (1996) 609.

[2] R. Elghanian, J.J. Storhoff, R.C. Mucic, R.L. Letsinger, C.A. Mirkin, Sci-ence 277 (1997) 1078.

[3] C.A. Mirkin, R.L. Letsinger, R.C. Mucic, J.J. Storhoff, Nature 382 (1996)607.

[4] W.F. Patolsky, K.T. Ranjit, A. Lichtenstein, I. Willner, Chem. Commun.(2000) 1025.

[5] W.F. Patolsky, J. Wasserman, Angew. Chem. Int. Ed. 40 (2001) 1861.[6] W.F. Patolsky, Y. Weizmann, O. Lioubashevski, I. Willner, Angew. Chem.

Int. Ed. 41 (2002) 2323.[7] J. Storhoff, J.R. Elghanian, R.C. Mucic, C.A. Mirkin, R.L. Letsinger, J.

Am. Chem. Soc. 120 (1998) 1959.[8] T.A. Taton, R.C. Mucic, C.A. Mirkin, R.L. Letsinger, J. Am. Chem. Soc.

122 (2000) 6305.[9] E. Braun, Y. Eichen, U. Sivan, G. Ben-Yoseph, Nature 391 (1998) 775.10] W.E. Ford, O. Harnack, A.J. Yasuda, M. Wessels, Adv. Mater. 13 (2001)

1793.11] C.F. Monson, A.T. Woolley, Nano Lett. 3 (2003) 359.12] S.H. Park, R. Barish, H. Li, J.H. Reif, G. Finkelstein, H. Yan, T.H. LaBean,

Nano Lett. 5 (2005) 693.13] J. Richter, R. Seidel, R. Kirsch, M. Mertig, W. Pompe, J.H. Planschke, K.

Schackert, Adv. Mater. 12 (2000) 507.14] T. Torimoto, M. Yamashita, S. Kuwabata, T. Sakata, H. Mori, H. Yoneyama,

J. Phys. Chem. B 103 (1999) 8799.15] S.W. Lee, M. C., C.E. Flynn, A.M. Belcher, Science 296 (2002) 892.16] E. Mayes, A. Bewick, D. Gleeson, J. Hoinville, R. Jones, O. Kasyutich, A.

Nartowski, B. Warne, J. Wiggins, K.K.W. Wong, IEEE Trans. Magn. 39(2003) 624.

17] W. Baumeister, M. Barth, R. Hegerl, R. Guckenberger, M. Hahn, W.O.Saxton, J. Mol. Biol. 187 (1986) 241.

18] M. Bergkvist, S.S. Mark, X. Yang, E.R. Angert, C.A. Batt, J. Phys. Chem.B 108 (2004) 8241.

19] Y.S. Chi, Y.H. Jung, I.S. Choi, Y.-G. Kim, Langmuir 21 (2005) 4669.20] D.C. Chow, W.-K. Lee, S. Zauscher, A. Chilkoti, J. Am. Chem. Soc. 127

(2005) 14122.21] J. Hyun, J. Kim, S.L. Craig, A. Chilkoti, J. Am. Chem. Soc. 126 (2004)

4770.22] Y.R. Kim, Nanoengineering of Solid Surfaces Using an In Vitro Synthesized

Biological Polymer, Cornell University, Ithaca, 2003.23] Y.R. Kim, H.J. Paik, C.K. Ober, G.W. Coates, C.A. Batt, Biomacro-

molecules 5 (2004) 889.24] Y.R. Kim, H.J. Paik, C.K. Ober, G.W. Coates, C.A. Batt, Macromol. Biosci.

6 (2005) 145.25] S.C. Delamarre, Engineering of Biological Systems for the Production of

Polyhydroxyalkanoate Copolymers and the Biofabrication of PolymericMicrostructures, Cornell University, Ithaca, 2004.

26] R. Jordan, A. Ulman, J. Am. Chem. Soc. 120 (1998) 243.

27] F.E. Black, M. Hartshorne, M.C. Davies, C.J. Roberts, S.J.B. Tendler, P.M.

Williams, K.M. Shakesheff, S.M. Cannizzaro, I. Kim, R. Langer, Langmuir15 (1999) 3157.

28] M.D.K. Ingall, S.J. Joray, D.J. Duffy, D.P. Long, P.A. Bianconi, J. Am.Chem. Soc. 122 (2000) 7845.

faces

[

[

[[

[

[[

[[

[

[[

[[[[

[

[[[[[

[[

[

[[[[

[

[

[

[

[

[[

[[

[[[[

[

[

[[[[

[

N. Niamsiri et al. / Colloids and Sur

29] S.L. Ishaug-Riley, L.E. Okun, G. Prado, M.A. Applegate, A. Ratcliffe,Biomaterials 20 (1999) 2245.

30] R. Jordan, A. Ulman, J.F. Kang, M.H. Rafailovich, J. Sokolov, J. Am.Chem. Soc. 121 (1999) 1016.

31] B.S. Kim, J.S. Hrkach, R. Langer, Biomaterials 21 (1999) 259.32] N. Luo, J.B. Hutchison, K.S. Anseth, C.N. Bowman, J. Polym. Sci., Part

A: Polym. Chem. 40 (2002) 1885.33] M. Moller, F. Nederberg, L.S. Lim, R. Kange, C.J. Hawker, J.L. Hedrick,

Y.D. Gu, R. Shah, N.L. Abbott, J. Polym. Sci., Part A: Polym. Chem. 39(2001) 3529.

34] O. Prucker, J. Ruehe, Langmuir 14 (1998) 6893.35] M. Husemann, M. Morrison, D. Benoit, J. Frommer, C.M. Mate, W.D.

Hinsberg, J.L. Hedrick, C.J. Hawker, J. Am. Chem. Soc. 122 (2000) 1844.36] D.J. Ehrlich, J.Y. Tsao, Appl. Phys. Lett. 46 (1985) 198.37] K. Matyjaszewski, P.J. Miller, N. Shukla, B. Immaraporn, A. Gelman, B.B.

Luokala, T.M. Siclovan, G. Kickelbick, T. Vallant, H. Hoffmann, T. Pakula,Macromolecules 32 (1999) 8716.

38] L.R. Rieth, D.R. Moore, E.B. Lobkovsky, G.W. Coates, J. Am. Chem. Soc.124 (2002) 15239.

39] B. Zhao, W.J. Brittain, J. Am. Chem. Soc. 121 (1999) 3557.40] A. Steinbuchel, in: D. Byrom (Ed.), Biomaterials, Basingstoke, Hants,

1991, p. 123.41] A. Steinbuchel, H.E. Valentin, FEMS Microbiol. Lett. 128 (1995) 219.42] B.H.A. Rehm, A. Steinbuchel, Int. J. Biol. Macromol. 25 (1999) 3.43] D. Jendrossek, Adv. Biochem. Eng. Biotechnol. 71 (2001) 294.44] F.F. Hezayen, B.J. Tindall, A. Steinbuchel, B.H.A. Rehm, Int. J. Syst. Evol.

Microbiol. 52 (2002) 2271.45] A. Steinbuchel, K. Aerts, W. Babel, C. Follner, M. Liebergesell, M.H.

Madkour, F. Mayer, U. Pieper-Furst, A. Pries, H.E. Valentin, R. Wieczorek,Can. J. Microbiol. 41 (1995) 94–105.

46] A.J. Anderson, E.A. Dawes, Microbiol. Rev. 54 (1990) 450.47] C.W. Pouton, S. Akhtar, Adv. Drug. Deliv. Rev. 18 (1996) 133.48] M. Zinn, B. Witholt, T. Egli, Adv. Drug. Deliv. Rev. 53 (2001) 5.49] B.H.A. Rehm, Biochem. J. 376 (2003) 15.50] B.H.A. Rehm, A. Steinbuchel, in: A. Steinbuchel, Y. Doi (Eds.), Biopoly-

mers, 3, Heidelberg, Germany, 2001.

51] T.U. Gerngross, D.P. Martin, Proc. Natl. Acad. Sci. U.S.A. 92 (1995) 6279.52] R.W. Lenz, C. Farcet, P.J. Dijkstra, S. Goodwin, S. Zhang, Int. J. Biol.

Macromol. 25 (1999) 55.53] R. Jossek, R. Reichelt, A. Steinbuchel, Appl. Microbiol. Biotechnol. 49

(1998) 258.

[

[

B: Biointerfaces 60 (2007) 68–79 79

54] R. Jossek, A. Steinbucel, FEMS Microbiol. Lett. 168 (1998) 319.55] A. Steinbuchel, S. Hein, Adv. Biochem. Eng. Biotechnol. 71 (2001) 81.56] S.-J. Liu, A. Steinbuchel, Appl. Microbiol. Biotechnol. 53 (2000) 545.57] Q. Qi, A. Steinbuchel, B.H.A. Rehm, Appl. Microbiol. Biotechnol. 54

(2000) 37.58] B.H.A. Rehm, Q. Qi, B.B. Beermann, H.-J. Hinz, A. Steinbuchel, Biochem.

J. 358 (2001) 263.59] S. Zhang, T. Yasuo, R.W. Lenz, S. Goodwin, Biomacromolecules 1 (2000)

244.60] T.U. Gerngross, K.D. Snell, O.P. Peoples, A.J. Sinskey, E. Csuhai, S.

Masamune, J. Stubbe, Biochemistry 33 (1994) 9311.61] J. Wodzinska, K.D. Snell, A. Rhomberg, A.J. Sinskey, K. Biemann, J.

Stubbe, J. Am. Chem. Soc. 118 (1996) 6319.62] S.M. Zhang, S. Kolvek, R.W. Lenz, S. Goodwin, Biomacromolecules 4

(2003) 504.63] A.G. Frutos, J.M. Brockman, R.M. Corn, Langmuir 16 (2000) 2192.64] E.A. Smith, M.J. Wanat, Y. Cheng, S.V.P. Barreira, A.G. Frutos, R.M. Corn,

Langmuir 17 (2001) 2502.65] M. Bradford, Anal. Biochem. 72 (1976) 248.66] L. Yan, C. Marzolin, A. Terfort, G.M. Whitesides, Langmuir 13 (1997)

6704.67] R. Singh, W.A. Blattler, A.R.A. Collinson, Anal. Biochem. 213 (1993).68] L. Ellman, Arch. Biochem. 82 (1959) 70.69] A.G. Ostle, J.G. Holt, Appl. Environ. Microbiol. 44 (1982) 238–241.70] G. Roo, Q. Qun Ren, B. Witholt, B. Kessler, J. Microbiol. Methods 41

(2000) 1.71] T. Hiraishi, Y. Kikkawa, M. Fujita, Y.M. Normi, M. Kanesato, T. Tsuge,

K. Sudesh, M. Maeda, Y. Doi, Biomacromolecules 6 (2005) 2671.72] T. Peters, in: Biochemistry, Genetics and Medical Applications, Academic

Press, San Diego, CA, 1996.73] A.C. Purvis, C.R. Tischler, R.C. Fites, Plant Physiol. 58 (1976) 95.74] B. Yang, C.E. Wyman, Biotechnol. Bioeng. 94 (2006) 611.75] W. Senaratne, L. Andruzzi, C.K. Ober, Biomacromolecules 6 (2005) 2427.76] R.N. Reusch, A.W. Sparrow, J. Gardiner, Biochim. Biophys. Acta 1123

(1992) 33.77] J.A. Molina-Bolıvar, J.L. Ortega-Vinuesa, Langmuir 15 (1999) 2635.

78] S. Bloembergen, D.A. Holden, G.K. Hamer, T.L. Bluhm, R.H. Marches-

sault, Macromolecules 19 (1986) 2865.79] S.C. Delamarre, N. Niamsiri, C.A. Batt, in: B.H.A. Rehm (Ed.), Microbial

Bionanotechnology: Biological Self-assembly Systems and Biopolymer-based Nanostructures, Norwich, U.K., 2006.