Biosensors and Bioelectronics 67 (2015) 739–746

Contents lists available at ScienceDirect

Biosensors and Bioelectronics

http://d0956-56

n CorrE-m

journal homepage: www.elsevier.com/locate/bios

Surface-enhanced Raman spectroscopy detection of dopamine by DNATargeting amplification assay in Parkisons's model

Jeung Hee An a, Dong-Kug Choi b, Kwon-Jai Lee a,c, Jeong-Woo Choi d,n

a Division of Food Bioscience, Konkuk University, Chungju-si 380-701, South Koreab Department of Biotechnology, Konkuk University, Chungju-si 380-701, South Koreac Department of Advanced Materials Engineering, Daejeon University, Daejeon 300-716, South Koread Department of Chemical & Biomolecular Engineering, Sogang University, Seoul 121-742, South Korea

a r t i c l e i n f o

Article history:Received 5 June 2014Received in revised form20 October 2014Accepted 20 October 2014Available online 23 October 2014

Keywords:Bio-barcode assayGold nanoparticlesDopamineSurface-enhanced Raman spectroscopy

x.doi.org/10.1016/j.bios.2014.10.04963/& 2014 Elsevier B.V. All rights reserved.

esponding author. Fax: þ82 2 3273 0331.ail address: jwchoi@sogang.ac.kr (J.-W. Choi).

a b s t r a c t

Dopamine is a potent neuromodulator in the brain that influences a variety of motivated behaviors and isinvolved in several neurologic diseases. We evaluated a bio-barcode amplification assay for its ability todetect dopamine in a mouse model with and without prior administration of the neurotoxin 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine (MPTP). Our approach uses a combination of DNA barcodes andbead-based immunoassays for detecting neurotransmitters with surface-enhanced Raman spectroscopy(SERS). This method relies on a gold nanoplate with adsorbed antibodies and gold nanoparticles that areencoded with DNA and antibodies that can sandwich the target protein captured by the nanoparticle-bound antibodies. C57BL/6 mice were infused intranasally with MPTP (25 mg/kg/day) over 7 consecutivedays. At 7 and 21 days after the last administration of MPTP, dopamine was found by western blotanalysis to have decreased in the midbrain by 37.44% and 92.95%, respectively. Furthermore, the Ramanintensity of dopamine in the midbrains of MPTP-treated mice decreased by 56.77% and 61.12% on days7 and 21, respectively. Our results demonstrate that the concentration of dopamine in midbrain andstriatum of MPTP-treated mice can be easily detected using the bio-barcode assay, which is a rapid, high-throughput screening tool for detecting neurotransmitters.

& 2014 Elsevier B.V. All rights reserved.

1. Introduction

Dopamine is a potent neuromodulator in the brain that influ-ences a variety of motivated behaviors and is involved in severalneurologic disease (Schapira, 2002; Gubernator et al., 2009 ). It isnow well known that dopamine is an extrasynaptic messengerthat functions via volume transmission, escaping the synaptic cleftto bind to extrasynaptic receptors and transporters. Abnormaldopamine transmission is associated with neurologic disorderssuch as schizophrenia, Parkinson's, and Huntington's diseases(Rowe et al., 1998; Ali et al., 2007). Patients with Parkinson's dis-ease have lost more than 80% of their dopamine-producing cells inthe substantia nigra (Schapira, 2002; Kline et al., 2009; Leviel2001). Basal dopamine concentration in the extracellular fluid ofthe central nervous system is very low (0.01–1 μM) (Rowe et al.,1998). Therefore, sensitive and selective determination of dopa-mine is needed for molecular diagnostics of Parkinson's disease,design of therapeutics, and evaluation of drug efficacy (Ali et al.,

2007; Dijkstra et al., 2007). Various methods have been developedfor the detection of dopamine release by single cells, includingspectrophotometry, chemiluminescence, fluorometry, chromato-graphy, and electrochemical technology (Wei et al., 2007; Wenet al., 2009; Huang et al., 2008; Kovac et al., 2014; Ferry et al.,2014). Dopamine release is rapid; therefore, detection technolo-gies should have high spatial and temporal resolution (Cao et al.,2010; Oh et al., 2006). Recently, the novel bio-barcode amplifica-tion assay (BCAA) was reported as having ultrahigh sensitivity;(Duan et al., 2010; Chen et al., 2009) however, there have been noreports thus far describing the use of the BCAA for dopamine de-tection in an in vivo model of Parkisons's.

Surface-enhanced Raman spectroscopy (SERS) using metal na-noprobes have shown promise by increasing the sensitivity ofRaman spectroscopy (Lee et al., 2007). SERS enhances the Ramansignal intensity by a factor of 106 to 1014, thus allowing detectionof pico- to femtomolar amounts of biomolecules (Woo et al.,2009). Functionalized SERS nanoparticles, in the presence of Ra-man-active molecules, are typically used for the detection, sensing,or imaging of biological samples such as DNA, proteins, cells, andtissues (Xie et al., 2009). In particular, SERS has been used to

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746740

confirm the viability of substrates such as microlithographicallyprepared silver posts, electron-beam lithographically producedarrays of elongated gold nanoparticles, nanosphere litho-graphically produced triangular nanoparticle arrays, gold-nanos-tructured films deposited on polystyrene colloidal crystal tem-plates, and gold template particles grafted onto silanized glass(Baia et al., 2007; Fujiwara et al., 2006). SERS phenomena invol-ving Au nanoparticles have been well-documented because theirsignals can be identified at the single-molecule level by associationwith the resonance effect of the adsorbed molecule (El-Said et al.,2010).

BCAAs allow highly sensitive detection of proteins and nucleicacids (Cao et al., 2010). This method involves two types of parti-cles, a magnetic nanoparticle functionalized with a group that hasaffinity for a target of interest, and nanoparticles functionalizedwith a second group that has affinity for the same target alongwith oligonucleotide barcodes that serve as reporter groups for thetarget (Oh et al., 2006; Duan et al., 2010). The detection limit of theBCAA for protein has been shown to be as low as 30 aM (atto-molar) (Oh et al., 2006). One of the key aspects of the BCAA assayis its use of gold nanoparticles, which enable simultaneous loadingof detection antibodies/probes and a large number of barcodeDNAs per particle (Chen et al., 2009). Because gold nanoparticlesare relatively easily coupled with antibodies and DNA, they over-come the complicated preparation of antibody-DNA conjugatesrequired in immuno-PCR (Chen et al., 2009). In addition, the largeratio of DNA to antibody (typically 100–300:1) on the functiona-lized gold nanoparticles further improves its sensitivity relative tothat of immuno-PCR (Chen et al., 2009).

The neurotoxin 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropy-ridine (MPTP) causes damage to dopaminergic neurons and hasbeen widely used for generation of animal models of Parkinson'sdisease (Aras et al., 2014). In our study, dopamine was detected forthe first time by BCAA combined with SERS analysis in a Parkin-son's animal model. We show that this modified BCAA can detectdopamine, which is of tremendous potential significance to thehealth care field as dopamine is recommended as the most ef-fective marker for diagnosing and monitoring patients with Par-kinson's. Moreover, this approach offers an opportunity to developa biosensor for monitoring the effect of neurotoxins on neuronalcell growth as well as viability, and could be applicable to the earlydetection of Parkinson's in MPTP-induced animal models.

2. Materials and methods

2.1. DNA oligonucleotides, bio-barcode DNA, and PCR primers

The sequences of the thio-capped oligonucleotides and the bio-barcode DNAs were 5′-CCACACTGCCG-GATGTGGATTTAACCTTT-CAGCAAATGTCTT-(A)10-(CH2)3-SH-3′ and 5′-AAG ACA TTT GCTGAA AGG TTA AAT CCA CAT CCG GCA GTG TGG-3′, respectively. Theprimers for amplification of the bio-barcode DNA were synthe-sized by Bioneer (Seoul, Korea).

2.2. Functionalization of gold substrates with immobilized dopamineantibody

The gold substrates were chemically modified by the formationof a self-assembled monolayer of 11-mercaptoundecanoic acid(MUA; Sigma-Aldrich, St. Louis, MO). The gold substrate was im-mersed in ethanol and then incubated in a 10 mM solution of MUAin ethanol for 10 min, and followed by rinsing with ethanol andthen distilled water. This procedure requires very careful handlingof the Au substrate since rinsing with or immersion in ethanolmay cause mixing of water and ethanol on the substrate surface,

which would lead to aggregation of Au nanoparticles. The thiolgroups on the MUA-coated, gold nanoparticle-deposited substratewere then derivatized to form succinimide esters by immersing inan aqueous solution containing 0.1 M 1-ethyl-3-(3-dimethylami-nopropyl) carbodiimide (EDC; Dojindo, location) and 0.1 M N-hy-droxysuccinimide (NHS; Sigma-Aldrich) for 2 h at room tempera-ture. For functionalization with anti-dopamine (abcam Cambridge,MA, USA), the NHS-activated glass chips were incubated in albu-min solution containing anti-dopamine for 24 h at 4 °C. Thefunctionalized glass chips were stored at 4 °C until further use (Fig.1a).

2.3. Preparation of functionalized gold probes

The antibody- and thiol-capped oligonucleotides were con-jugated with gold nanoparticles by a previously described one-step method (An et al., 2012). The conjugated gold nanoparticleswere re-suspended in 4 ml of PBS and bio-barcode DNA (0.8 OD atA260) was added and allowed to hybridize for 4 h at room tem-perature. The solution was centrifuged at 13,000g, 30 min (Labo-gene 1248R) at 4 °C, and the resultant dopamine-nanoparticle–dsDNA complexes were re-suspended in a solution of 0.15 M NaCland 0.01 M PBS and stored at 4 °C. The antibody- and DNA-con-jugated gold nanoparticles were detected using a UV–visiblespectrophotometer; the antibody activity on the gold nanoparticlesurface was evaluated by ELISA, and the oligonucleotides weredetected by SERS (Fig. 1a).

2.4. Capture of dopamine in SH-SY5Y cells and midbrain and stratumof MPTP-treated mice by bio-barcode DNA assay

Cells of SH-SY5Y neuroblastoma and tissues of midbrain andstratum were lysed in ice-cold lysis buffer (RIPA, 20 mM Tris–HClpH7.5, 150 mM NaCl, 1 mM Na2EDTA, 1 mM EGTA, 1% NP-40, 1%sodium deoxycholate, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, 1 mM Na3VO4, 1 mg/ml leupeptin, 1 mM PMSF).Homogenates were centrifuged once at 500g for 5 min to removeunbroken cells and tissues and the supernatant was collected andcentrifuged at 13,000g for 10 min at 4 C. The supernatants wereused to antigen solution. In a typical experiment, an immunoassayconsisting of an immunoprobe/dopamine/gold immunoprobesandwich was conducted, and then the DNA bio-barcodes werecollected after enzyme treatment and eluted with chloroform:isopropanol:phenol (25:24:1). The extracted DNAwas treated withprotease K to hydrolyze any contaminating proteins by incubationfor 1 h at 50 °C. The supernatant of the reaction mixture was usedas a template for PCR or incubated with the nanogold plates forbinding of complementary DNAs. DNA was also detected by SERS.Complementary DNA was detected by its Raman signal; however,non-complementary DNA did not exhibit a Raman signal.

2.5. Cell culture

The neuroblast-like SH-SY5Y cell line (Korea Cell Line Bank,Seoul, Korea) were cultured in the presence of CO2 (5%) at 37 °C inDulbecco's Modified Eagle's Medium (Gibco) supplemented withfetal bovine serum (10%, Hyclone, Thermo Fisher Scientific, Rock-ford, IL, USA) and 0.5% antibiotics (Gibco).

2.6. MPTP exposure mice

Eight- to ten-week-old male C57BL/6J mice (20–25 g) wereused for the MPTP treatment experiments, with a total of 5 mice ineach group. All animal experiments were conducted according toprotocols approved by the Institutional Animal Care and UseCommittee, Konkuk University. For the MPTP lesioning phase,

Fig. 1. (a) Schematic diagram of the surface-immobilized bio-barcode assay protocol, and (b) scanning electron microscope (SEM) image of gold nanopatterns fabricated onITO glass substrate and anti-dopamine-conjugated gold nanoparticles.

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746 741

mice received daily subcutaneous injections at the nape of theneck of either phosphate-buffered saline (PBS) or 25 mg/kg of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP-HCl Sigma-Al-drich) in PBS for 5 days. Mice were allowed to recover for 7 and 21days after the last MPTP dose and then sacrificed by decapitationfor tissue harvesting.

2.7. Preparation of complementary DNA in nanogold plates for SERSdetection

Well-ordered nanoporous alumina masks were prepared fromaluminum foil (99.99%, 100 μm in thickness) via a two-step ano-dization process (Jung et al., 2011). The nanoporous alumina maskswere placed on indium tin oxide (ITO) glass substrates, and goldwas deposited on the ITO substrate through the pores of the maskusing a thermal evaporator (ULVACVPC-260) with a vacuumpressure of 3�10�6 Torr and an evaporation rate of approxi-mately 0.1 Å s�1. After gold deposition, the alumina masks weredissolved for 5 min in 10% (w/v) NaOH, and then the substrate wasrinsed three times with distilled water. After the pre-treatment, awell-ordered complementary DNA (5′-AAG ACA TTT GCT GAA AGGTTA AAT CCA CAT CCG GCA GTG TGG-3′) was fabricated on thefreshly cleaned gold electrodes.

2.8. Raman spectroscopy

The biocomposition of living cells and the effect of neurotoxicagents on SH-SY5Y cells were investigated by SERS using a RamanNTEGRA spectra (NT-MDT, city, Russia). The maximum scan range(XYZ) was 100 μm�100 μm�6 μm; the resolution of the spec-trometer was 200 nm in the XY plane and 500 nm along the Z axis.Raman spectra were recorded using an NIR 785 nm laser. Tenscans of 1 s each from 1 cm�1 to 2000 cm�1 were recorded, andthe means of these scans were used in the analysis. The spectrawere typically acquired with a 1 s exposure time at laser powers of10 mW and at 785 nm. Raman peak measurements are controlledby NT-MDT's Nova software. Images obtained were processedusing the program, Image Analysis 3.5 (NT-MDT, Russia).

3. Results and discussion

3.1. Synthesis and characterization of anti-dopamine-conjugatedbarcode DNA

The barcode assay utilizes 60 nm Au nanoparticle probes, eachco-functionalized with a polyclonal antibody for recognizing thetarget antigen (Fig. 1a). One-half of the barcode for each target is

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746742

the target-reporting oligonucleotide probe, while the other half isidentical for all the barcodes and is a universal sequence. Fur-thermore, the target antigens are captured in anti-dopamineconjugated gold substrate, each conjugated with a monoclonalantibody specific for an epitope of the target antigen that is dif-ferent from the one recognized by the gold probe (Oh et al., 2006).The barcode DNA –antibody and gold nanoparticle complexes aresandwiched between Au nanoparticle probes with an antibodythat can bind to the target in a region different from that of theanti-dopamine conjugated gold substrate (Fig. 1a). These com-plexes were lysed using protease enzyme and washed, and thebarcode strands are released by a ligand exchange process inducedby the addition of dithiothreitol. The bio-barcode strands are thenidentified by SERS. The Au nanoparticle probes are modified for

Fig. 2. (a) Raman spectra of the DNA barcodes for different dopamine concentrations idbands at 670, 780, 1179, and 1420 cm�1 at various dopamine concentrations.

detection of antibody and barcode DNA strands (43 bp/particle).The Au nanoparticle probes are stored at a concentration of 0.8 O.D., with the excess barcode DNA removed in 0.15 M NaCl, 0.025%Tween 20, 0.1% BSA, and 10 mM PBS (pH 7.4).

We constructed nanogold substrates containing approximatelyequal numbers of Au nanostructures that contained SERS-activespots. The average diameter of the nanoporous alumina wasmeasured by scanning electron microscopy (SEM) and found to be6075 nm. The nanoporous alumina mask was approximately200 nm thick and its thickness was previously shown to dependon the second anodization time (Jung et al., 2011). A long-range-ordered ultrathin alumina mask with through-holes was placed onthe ITO glass substrate. Previously, this was shown to bind to thesubstrate via van der Waals interactions (Mei et al., 2003). A

entified by surface-enhanced Raman spectroscopy, and (b) linear graph of Raman

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746 743

representative SEM image of the gold nanosubstrate arrays formedon the ITO glass after removal of the nanoporous alumina mask isshown in Fig. 2b. The average diameter of the Au nanoislands ar-rays was approximately 7075 nm, and their mean height was4575 nm (Jung et al., 2009). The shape of the gold nanoislandarrays depend on the pore diameter and thickness of the aluminamask which is known to play a crucial role in the ordering of thearray (Jung et al., 2011). Here, it was removed after the secondetching process. Thus, gold nanoisland arrays with a regular sizeand uniform spatial distribution, for use as a nanobioplatform forSERS detection, were obtained using a nanoporous alumina mask.Similar to earlier reports, most gold nanoparticles self-assemblednear each other by forming randomly distributed one-dimensionalpatterns (Kafi et al., 2010).

3.2. Detection of dopamine using barcode probes

Many factors can limit the performance of DNA barcode-basedimmunoassays. These factors include the types of particles or an-tibodies, the attachment chemistry, and the effectiveness of theremoval and washing steps. Among them, the affinity of the im-munoprobe for its analytical target arguably plays the most im-portant role. The dopamine bound by DNA-conjugated nano-particles, which formed a self-assembled gold substrate, was ex-amined by Raman spectroscopy. We detected levels correspondingto five different dopamine concentrations ranging from 0 pM to1 nM, and the intensity of the dopamine barcode DNA increased ina dose-dependent manner (Fig. 2).

We obtained SERS spectra for a range of wave numbers from500 to 1600 cm�1 (Fig. 2a). Raman spectra of DNA barcode ofdopamine showed broad bands at 670, 780, 1179, 1383 and1420 cm�1. Raman peaks of DNA are found at 670 (guanine), 780(thymine), 1179 (thymine, cytosine), 1289 (adenine, thymine) 1420(deoxyribose) and 1478 cm�1 (adenine) (Barhoumi et al., 2008;Kneipp et al., 2002; Lamba et al., 1989; Ruiz-Chica et al., 2004; Xieet al., 2009). A detection limit of 0.1 pM dopamine could be esti-mated using a signal/noise (S/N) ratio of 3. The box in Fig. 2, theraman signal of 0.1 pM showed at 1179 and 1420 cm�1 peaks. Thisdetection limit is 4 orders of magnitude lower than that of theaptamer conformational change-based SERS method (Chen et al.,2008), one order of magnitude lower than that of the strand dis-placement amplification-based fluorescence detection (He et al.,2010), and lower than that of the other optical or electrochemicaltechniques (Li, et al., 2007, 2012; Sharma and Heemstra, 2012).The Raman peaks that appeared in our experiments are those mostaffected by adsorption, along with the conventional Raman spec-trum of the test molecule excited by the 785 nm laser (note thatthe conventional Raman spectra were all recorded using excitationlines, but only one was selected because of their similarity).Moreover, the SERS spectra recorded from different spots on thesurface are identical, which demonstrates the reproducibility ofthe SERS substrate. Interesting results were observed upon com-parison of samples; a gold substrate, a surface with single DNAwith non-complementary DNA, and a DNA base with com-plementary DNA. The Raman spectrum of the gold substrate wasobserved at 1383 cm�1, which was similar to the pattern of thesurface with non-complementary DNA. However, the com-plementary DNA displayed various peaks at 670, 780, 1179, 1289,1420 and 1478 cm�1.

Fig. 2b shows that the intensity of the spectral patterns of SERSincreased with increasing dopamine concentrations; moreover,the antigen-concentration-dependent SERS spectra changed. De-creasing the concentration of dopamine resulted in a decrease inthe relative intensities of the Raman peaks at 670 (R2¼0.9831),780 (R2¼0.9804), 1179 (R2¼0.9950) and 1420 cm�1 (R2¼0.9792),which correspond to the DNA bases.

Especially, there was a good linear relationship between theRaman intensity of the 670 cm�1 peak and the dopamine con-centration from 1�10�13 M to 1�10�9 M (Fig. 2C). The regressionequationwas Y¼10436.79þ736.60X (Y is for the Raman intensity, Xis for the concentration of dopamine). Also, the regression equa-tion of 780, 1179 and 1420 cm�1 were Y¼9739.10þ690.50X,Y¼11700.30þ844.29X and Y¼11392.60þ784.79X, respectively.Thus, our result shows that the SERS response increased linearlywith increasing concentrations of dopamine. A series of 10 re-spective measurements of 1�10�13 M dopamine was used for es-timating the assay precision, and the relative standard deviationwas 1.81%, thus showing good reproducibility of the proposedmethod. These results suggest that a dopamine-conjugated bio-barcode DNA on nanopatterned Au is an effective tool for neuro-transmitter detection by SERS-based measurements. Importantly,similar patterns were found between the SERS spectra in terms ofthe dose-dependent expression of dopamine. Therefore, our resultssuggest that SERS may be useful for monitoring neurotransmitterssuch as dopamine, epinephrine, and norepinephrine.

3.3. Detection of dopamine secreted by SH-SY5Y cells by barcodeanalysis

As a step towards establishing the biological relevance of thismethod, we determined whether the immunoprobes capture do-pamine secreted by dopaminergic cells such as the SH-SY5Y cellline. SERS analysis was exploited for this purpose as a straight-forward approach to validate the utility of the immunoprobe.Potential concerns regarding the detection of dopamine by thismethod include: (1) failure to capture dopamine due to theblocking effect of the microbeads, (2) non-specific interactionsbetween the beads and the dopamine leading to false-positives,and (3) non-specific interactions of the immunoprobe with othercatecholamines. To address these potential issues and evaluate thespecificity and clinical applicability of the proposed method, re-covery experiments were performed.

We obtained SERS spectra for dopamine secreted by the SH-SY5Y cells for a range of wave numbers from 500 to 1600 cm�1. Aswith the commercially supplied dopamine, the peaks that ap-peared in these spectra were mostly those affected by adsorption,along with the conventional Raman spectrum of the test moleculeexcited by a 785 nm laser (Fig. 3a). The Raman spectrum of SH-SY5Y cells showed Raman bands at 670, 780, 1179, 1420 and1478 cm�1. Interestingly, the results show a change in intensity at1478 cm�1 (a change that is associated with catechol ring vibra-tion) with a change in the concentration of dopamine (Lee et al.,1988). But the 1478 cm�1 observed as adenine in DNA. Also, the1179 cm–1 was previously demonstrated to represent thymine,whereas the 670 cm–1 band was identified as guanine (Barhoumiet al., 2008; Ruiz-Chica et al., 2004).

Three neurotoxins (polychlorinated biphenyl (PCB), rotenone,and bisphenol A) were selected as representative oxidative stress-inducing drugs to study their effects on dopamine secretion in SH-SY5Y cells using the SERS-based DNA bio-barcode analysis. Theconcentration of dopamine secreted by untreated SH-SY5Y cellswas determined to be 2.45�10�11 M using our bio-barcode ana-lysis (Fig. 3b). The cellular concentration of dopamine was calcu-lated at Raman peaks of 670 cm�1 and shown in Fig. 2b. The Ra-man spectra indicated that differences in biochemical compositionwere evident after treatment with the neurotoxins, particularly atthe Raman peaks associated with DNA. Overall, treatment with theneurotoxins decreased the relative intensities of the Raman peaks.The concentration of dopamine was found to be 1.12�10�12 Mafter PCB treatment, 1.19�10�13 M in rotenone-treated cells, and1.76�10�13 M in bisphenol A-treated cells. The dopamine extracts

Fig. 3. SERS spectra (a) and barcode analysis (b) of dopamine secreted by untreatedSH-SY5Y cells (Con) and those exposed to 500 nM of bisphenol A, polychlorinatedbiphenyl (PCB), or rotenone for 24 h.

Fig. 4. (a) The expression of dopamine and thyrosine hydroxylase was determinedby western-blot analysis. Equal loading was confirmed by β-actin quantification.(b) For protein, the relative density was measured normalized to β-actin bands.

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746744

were from 1�106 cells/ml samples prepared from SH-SY5Y cells.The dopamine recovery ranged from 95.6% to 102% (Fig. 3b).

PCB exposure is known to pose numerous health risks; indeed,several studies have shown PCBs to be detrimental to the centralnervous system, resulting in motor and cognitive deficits (Simonet al., 2007). One common result that has been reported is a sig-nificant reduction in brain dopamine levels following exposure toPCBs (Lyng and Seegal, 2008). Rotenone is a common pesticide anda well-characterized, specific inhibitor of complex I of the mi-tochondrial respiratory chain. Rats chronically exposed to rote-none develop neuropathological and behavioral symptoms ofParkinson's disease resulting from the induction of apoptosis andacceleration of α-synuclein formation as has been shown inpharmacodynamic models in vivo and in vitro (Betarbet et al.,2000). Bisphenol A at environmentally relevant levels (either 25 or250 ng per kg of body mass) has been shown to cause altered brainsexual differentiation in rat offspring (Rubin et al., 2006). Fur-thermore, bisphenol A has been reported to traverse the blood–brain barrier (Sun et al., 2002) and to induce changes in the be-havioral activities of offspring after maternal exposure.

3.4. Expression levels of dopamine and tyrosine hydroxylase in MPTPmice

Progressive degeneration of dopaminergic neurons and dopa-mine depletion in the nigrostriatal system is a key component inthe pathogenesis of Parkinson's disease. (Schapira, 2002). It isnoteworthy that the initial clinical symptoms first appear in hu-mans many years after the onset of the neurodegenerative processafter a threshold loss of �50% of dopaminergic neurons (cell

bodies) in the substantia nigra and 70–80% of axonal terminals inthe striatum. (Ehringer anf Hornykiexicz, 1998) It has been shownthat all the crucial characteristics of Parkinsonism are replicated inMPTP-treated mice, including motor behavior, dopamine content,and the number of dopaminergic neurons (cell bodies and axonalterminals) in the nigrostriatal system (Kozina et al., 2014). Takinginto account that tyrosine hydroxylase (TH) is a rate-limiting en-zyme of dopamine synthesis, TH activity is generally considered tobe an indication of the rate of dopamine synthesis (Kozina et al.,2014). Our data show that the expression level of TH decreased by23.4% and 74.64% at 7 and 21 days, respectively, after MPTP wasinjected in the midbrain (Fig. 4a and b); the expression of dopa-mine was found to have decreased by 37.44% and 91.95%, respec-tively (Fig. 4). Our results are consistent with expression patternsreported by others. Pain et al. (2013) reported that TH im-munoreactivity was significantly reduced in the striatum of MPTP-intoxicated mice compared to saline-treated mice (40%, 66% and79% for acute, sub-acute and chronic, respectively). A positivecorrelation has been shown between the severity of Parkinsonismand the decrease of dopamine uptake by MPTP (Perez et al., 2008).In addition, a decrease in aromatic L-amino acid decarboxylaseactivity has been demonstrated in nigrostriatal dopaminergicneurons in humans with Parkinson's (Hefti et al., 1981), whichcould result in a decrease in dopamine synthesis despite therebeing relatively high TH activity.

3.5. Detection of dopamine in midbrain and striatum of control andMPTP-treated mice by barcode analysis

SERS spectra of dopamine from midbrain and striatum werecollected for wave numbers ranging from 530 to 1250 cm�1.Consistent with the SERS analysis of purified dopamine and thatsecreted by SH-S5Y5 cells, the peaks that appeared in thesespectra were mostly those affected by adsorption, along with theconventional Raman spectrum of the test molecule excited by a785 nm laser (Fig. 5). The Raman spectrum of midbrain dopamineshowed bands at 670, 725, 780, 857, 1078, 1383 and 1482 cm�1.The peak at 670 and 780 cm�1, which was the most intense of allthe peaks, was similar to that in SERS spectra of the DNA base(Fig. 5a). Interestingly, the results show a change in intensity at670 and 780 cm�1, which is known to be associated with guanine

Fig. 5. (a) SERS spectra of dopamine from midbrains and striata in untreated andMPTP-treated mice. The C57BL/6 mice were infused intranasally with MPTP(25 mg/kg/day) over 7 consecutive days. (B) Relative concentration of dopaminefrom midbrain and striatum compare to untreated and MPTP-treated mice.

Fig. 6. (a) SERS spectra of dopamine in plasma from control and MPTP-treatedmice. (b) Relative concentration of dopamine from plasma compare to normal andMPTP mice.

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746 745

and Thymine (Barhoumi et al., 2008). Thymine and cytosine havebeen shown to be associated with the bands at 1078 cm�1 and1482 cm�1, cytosine with the bands at 1078 cm�1 and 1482 cm�1,and adenine with the bands at 725 cm�1, and 1482 cm�1 (Bar-houmi et al., 2008). Also, the Raman spectrum of the gold sub-strate was observed at 1383 cm�1. The Raman spectra of dopa-mine were consistent with the results obtained by western blotanalysis. The intensity of dopamine the midbrains of untreatedmice was found to have decreased by 7 day, and further decreasedby day 21 (Fig. 5b). The dopamine concentrations in the midbrainsof control mice were 2.63�10�13 M using our bio-barcode ana-lysis (Fig. 5b) by the SERS analysis. The intensity of dopamine inmidbrains of the MPTP-treated mice decreased at 7 and 21 days by56.77% and 61.12%, respectively, compared to control. Thus, theconcentration of dopamine was found to be 1.137�10�13 M and1.023�10�13 at 7 and 21 day after MPTP treatment, respectively.

Interestingly, the Intensity of Raman spectra of dopamine fromstriatum exhibited the same pattern by day 7; however, by day 21was found to have increased (Fig. 5a). The Raman spectra ofstriatum dopamine also showed bands at 670, 725, 780, 857, 1078,1179, 1383, and 1482 cm�1. Similarly to midbrain dopamine, theintensity of the 780 cm�1 peak was higher than the others peaks,and the band at 780 cm–1 was similar to those in SERS spectra of

the DNA base (Fig. 5b). The same bands that were identified asbelonging to nucleotides in the midbrain dopamine spectra (670,725, 1078, 1383 and 1482 cm�1) were seen in those from striatumdopamine. The intensity of dopamine decreased by 68.94% and47.48% by days 7 and 21, respectively, in the striatum of MPTP-treated mice versus control mice (Fig. 5b). The dopamine con-centrations in the midbrains of control mice were 3.07�10�13 Musing our bio-barcode analysis by the SERS analysis. Also, theconcentration of dopamine in striatum was found to be9.54�10�14 M and 1.613�10�13 at 7 and 21 day after MPTPtreatment, respectively.

Plasma samples for 20 min at 100g, followed removed theparticulates and then assayed immediately by bio-barcode assay(Fig. 6a). In plasma, the intensity of the Raman peaks in the MPTP-treated mice decreased by 62.9% on 7 day plasma compared tonormal mice (Fig. 6b). Our plasma dopamine concentration iscalculated by the standard dopamine concentration. Thus theconcentration of dopamine in the normal and MPTP mice wasfound to be 6.30�10�12 and 2.34�10�12 M, respectively. Zig-mond and Hastings (1998) showed that a loss of high-affinitydopamine uptake sites was correlated with the degeneration ofdopamine terminals. Indeed, in our study we detected a markeddecrease in dopamine transporter levels in striatum of MPTP-treated animals. Furthermore, we showed that the mesocorticalpathway was also sensitive to MPTP treatment. Indeed, chronicMPTP intoxication was associated with frontal cortex dopamineloss. These changes are in agreement with the reported alterationof the frontal cortex in mice and monkeys intoxicated with MPTP(JacKson-Lewis et al. 1995) and in Parkinson's patients (Javoy-Agid

J.H. An et al. / Biosensors and Bioelectronics 67 (2015) 739–746746

and Agid, 1980). Therefore, Raman spectra are a simple way toinvestigate the effects of toxins on dopaminergic mice by mon-itoring changes in cellular physiology and viability. Thus, ourtechnique can be used to fabricate a highly sensitive, low-costneurotransmitter sensor for detecting dopamine.

4. Conclusions

Gold nanoparticles, modified with an antibody and dsDNA,were used in an immunosorbent bio-barcode assay to detect traceamounts of dopamine. The dopamine concentrations in controlcells were detected at 2.45�10�11 M using our bio-barcode ana-lysis. The results of our study argue for a prospective evaluation ofanimals after MPTP treatment, such that the time interval ofplasma sampling can be controlled postoperatively for rapididentification of those animals and humans who may benefit mostfrom early detection. Furthermore, dopamine secreted by a do-paminergic cell line and dopaminergic neurons from mice can bedetected easily and rapidly using this bio-barcode assay. Takentogether, our data show that Raman spectroscopy is a relativelysimple and accurate technique for investigating the effects oftoxins on dopaminergic neurons in mice and for monitoringchanges in cellular physiology and viability. Thus, our method canbe used to fabricate a highly sensitive, high-throughput and rela-tively low-cost dopamine sensor for detecting the effects of en-vironmental toxins on dopaminergic neurons.

Acknowledgments

This work was supported by Konkuk University, by the BasicScience Research Program through the National Research Foun-dation of Korea (NRF) funded by the Ministry of Education, Scienceand Technology (2010-0011876), and by the Leading ForeignResearch Institute Recruitment Program through the NationalResearch Foundation of Korea (NRF) funded by the Ministry ofScience, ICT & Future Planning (MSIP) (2013K1A4A3055268).

References

Ali, S.R., Ma, Y., Parajuli, R.R., Balogun, Y., Lai, W.Y., He, H., 2007. Anal. Chem. 79 (6),2583–2587.

An, J.H., El-Said, W.A., Choi, J.W., 2012. J. Nnanosci. Nanotechnol. 12 (1), 764–768.Aras, S., Tanriover, G., Aslan, M., Yargicoglu, P., Agar, A.A., 2014. Neurochem. Int. 72,

48–57.Baia, M., Toderas, F., Baia, L., Popp, J., Astilean, S., 2007. Chem. Phys. Lett. 422, 127.Barhoumi, A., Zhang, D., Tam, F., Halas, N.J., 2008. J. Am. Chem. Soc. 130 (16),

5523–5529.Betarbet, R.R., Sherer, T.B., MacKenzie, G., Garcia-Osuna, M., Panov, A.V., Green-

amyre, J.T., 2000. Nat. Neurosci. 3, 1301–1306.Cao, C., Dhumpa, R., Bang, D.D., Ghavifekr, Z., Høgberg, J., Wolff, A., 2010. Analyst

135 (2), 337–342.Chen, J., Jiang, X., Gao, G., Liu, G., Shen, R., Yu, 2008. Chemistry 14, 8374–8382.Chen, L., Guo, Y. Cui, Zhang, Z.Z., Zhang, X.E., 2009. J. Immunol. Methods 346 (1–2),

64–70.

Dijkstra, R.J., Scheenen, W.J., Dam, N., Roubos, E.W., ter Meulen, J.J., 2007. J. Neu-rosci. Methods 159, 43–50.

Duan, R., Zhou, X., Xing, D., 2010. Anal. Chem. 82, 3099–3103.Ehringer, H., Hornykiexicz, O., 1998. Parkinsonism Relat. Disord. 4 (2), 53–57.El-Said, W.A., Kim, T.H., Kim, H., Choi, J.W., 2010. Biosens. Bioelectron. 26 (4),

1486–1492.Ferry, B., Gifu, E.P., Sandu, I., Denoroy, L., Parrot, S., 2014. J. Chromatogr. B: Anal.

Technol. Biomed. Life Sci. 2014 (951-952), 52–57.Fujiwara, K., Watarai, H., Itoh, H., Nakahama, E., Ogawa, N., 2006. Anal. Bionanl.

Chem. 386, 639.Gubernator, N.G., Zhang, H., Staal, R.G., Mosharov, E.V., Pereira, D.B., Yue, M., Bal-

sanek, V., Vadola, P.A., Mukherjee, B., Edwards, R.H., Sulzer, D., Sames, D., 2009.Science 324, 1441–1444.

He, J.L., Wu, Z.S., Zhou, H., Wang, H.Q., Jiang, J.H., Shen, G.L., Yu, R.Q., 2010. Anal.Chem. 82 (4), 1358–1364.

Hefti, F., Melamed, E., Wurtman, R., 1981. J. Pharmacol. Exp. Ther. 217, 189–197.Huang, J., Liu, Y., Hou, H., You, T., 2008. Biosens. Bioelectron. 24 (4), 632–637.Jackson-Lewis, V., Jakowec, M., Burke, R.E., Przedborski, S., 1995. Neurodegenera-

tion 1995 (4), 257–269.Javoy-Agid, F., Agid, Y., 1980. Neurology 30, 1326–1330.Jung, M., El-Said, W.A., Choi, J.W., 2011. Nanotechnology 22, 235304.Jung, M., Oh, B.Y., Choi, J.W., 2009. Ultramiscopy 109, 1006–1010.Kafi, M.A., Kim, T.H., Yea, C.H., Kim, H., Choi, J.W., 2010. Biosens. Bioelectron. 26,

1359–1365.Kline, D.D., Hendricks, G., Hermann, G., Rogers, R.C., Kunze, D.L., 2009. J. Neuro-

physiol. 101, 2270–2278.Kneipp, K., Kneipp, H., Itzkan, I., Dasari, R.R., Feld, M.S., 2002. J. Phys.: Condens.

Matter 14, R597.Kovac, A., Somikova, Z., Zilka, N., Novak, M., 2014. Talanta 119, 284–290.Kozina, E.A., Khakimova, G.R., Khaindrava, V.G., Kucheryanu, V.G., Vorobyeva, N.E.,

Krasnov, A.N., Georgieva, S.G., Kerkerian-Le Goff, L., Ugrumov, M.V., 2014. J.Neurol. Sci. 340 (1–2), 198–207.

Lee, N.S., Hsieh, Y.Z., Paisley, R.F., Morris, M.D., 1988. Anal. Chem 60, 442.Lee, S., Kim, S., Choo, J., Shin, S.Y., Lee, Y.H., Choi, H.Y., Ha, S., Kang, K., Oh, C.H., 2007.

Anal. Chem. 79, 916.Leviel, V., 2001. Neurochem. Int. 38, 83–106.Li, D., Shlyahovsky, B., Elbaz, J., Willner, I., 2007. J. Am. Chem. Soc. 129 (18),

5804–5805.Li, M., Zhang, J., Suri, S., Sooter, L.J., Ma, D., Wu, N., 2012. Anal. Chem. 84,

2837–2842.Lamba, O.P., Wang, A.H., Thomas, G.J., 1989. Biopolymers 28, 667.Lyng, G.D., Seegal., R.F., 2008. Neurotoxicology 29, 301.Mei, X., Blumin, M., Kim, D., Wu, Z.H., Guo, Q.X., Ruda, H.E., 2003. J. Cryst. Growth

251, 253.Oh, B.K., Nam, J.M., Lee, S.W., Mirkin, C.A., 2006. Small 2 (1), 103–108.Pain, S., Gochard, A., Bodard, S., Gulhan, Z., Prunier-Aesch, C., Chalon, S., 2013. Exp.

Toxicol. Pathol. 65 (5), 689–694.Perez, X., Parameswaran, N., Huang, L., O'Leary, K., Quik, M., 2008. J. Neurochem.

105, 1861–1872.Rowe, D.B., Le, W., Smith, R.G., Appel, S.H., 1998. J. Neurosci. Res. 53 (5), 551–558.Rubin, B.S., Lenkowski, J.R., Schaeberle, C.M., Vandenberg, L.N., Ronsheim, P.M.,

Soto, A.M., 2006. Endocrinology 147, 3681–3691.Ruiz-Chica, A.J., Medina, M.A., Sanchez-Jimenze, F., Ramirez, F.J., 2004. Nucleic Acid

Res. 32, 579–589.Schapira, A.H., 2002. Eur. J. Neurol. 9, 7–14.Sharma, AK, Heemstra, JM., 2012. Anal. Chem. 84, 6104–6109.Simon, T., Britt, J.K., James, R.C., 2007. Regul. Toxicol. Pharmacol. 48 (148), 2007.Sun, Y., Nakashima, M.N., Takahashi, M., Kuroda, N., Nakashima, K., 2002. Biomed.

Chromatogr. 16, 319–326.Wei, W., Wang, H.J., Jiang, C.Q., 2007. Luminescence 22, 581–587.Wen, J., Zhou, L., Jin, L., Cao, X., Ye, B.C., 2009. J. Chromatogr. B: Anal. Technol.

Biomed. Life Sci. 877, 1793–1798.Woo, M.A., Lee, S.M., Kim, G., Baek, J., Noh, M.S., Kim, J.E., Park, S.J., Minai-Tehrani,

A., Park, S.C., Seo, Y.T., Kim, Y.K., Lee, Y.S., Jeong, D.H., Cho, M.H., 2009. Anal.Chem. 81, 1008–1015.

Xie, W., Wang, L., Zhang, Y., Su, L., Shen, A., Tan, J., Hu, J., 2009. Bioconjug. Chem. 20,768.

Zigmond, M.J., Hastings, T.G., 1998. Adv. Pharmacol. 42, 788–792.