Dynamic Article LinksC<Lab on a Chip

Cite this: Lab Chip, 2011, 11, 1059

www.rsc.org/loc PAPER

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online / Journal Homepage / Table of Contents for this issue

DNA hybridization enhancement using piezoelectric microagitation througha liquid coupling medium

Kiattimant Rodaree,†a Thitima Maturos,†a Sastra Chaotheing,†*b Tawee Pogfay,†a Nattida Suwanakitti,b

Chayapat Wongsombat,b Kata Jaruwongrungsee,a Anurat Wisitsoraat,a Sumalee Kamchonwongpaisan,b

Tanom Lomasa and Adisorn Tuantranont*a

Received 17th September 2010, Accepted 6th January 2011

DOI: 10.1039/c0lc00419g

In conventional DNA microarray hybridization, delivery of target cDNAs to surface-bounded probes

depends solely on diffusion, which is notoriously slow, and thus typically requires 6–20 h to complete. In

this study, piezoelectric microagitation through a liquid coupling medium is employed to enhance DNA

hybridization efficiency and the results are compared with the standard static hybridization method.

DNA hybridization was performed in a sealed aluminium chamber containing DNA microarray glass

chip, coupling medium and piezoelectric transducers. 3 � SSC (Saline Sodium Citrate) was used as

a coupling medium to prevent overheating of the piezoelectric transducers and to effectively transmit

ultrasonic wave to the glass chip. Flow visualization using fluidic dye and velocimetry (PTV) technique

was applied to observe fluid transport in the hybridization chamber. It was revealed that the dye solution

was homogeneously distributed within 10 min under dynamic agitation while it took over 1 h to reach the

same level of homogeneity in static condition. Plasmodium falciparum DNA microarrays and total RNA

extracted from parasite cells were used as a model for DNA microarray experiments. It was found that

the required hybridization time may be substantially reduced from 16 h to 4 h by the use of dynamic

hybridization scheme. With the same hybridization time of 16 h, dynamic hybridization resulted in higher

fluorescent signals of �33% and �24% compared to static hybridization in Cy3 and Cy5 channels,

respectively. Additionally, good/effective spots, some of which were not formed by static method, were

enhanced and distributed more uniformly over the microarray. Therefore, the developed dynamic

hybridization with integrated piezoelectric microagitation platform is highly promising for DNA

analysis in molecular biology and medical applications.

Introduction

DNA microarrays have been recognized as a prominent tool in

molecular biology and medicine. They have been applied to

a wide range of research including gene discovery and mapping,

gene regulation studies, disease diagnosis, drug discovery and

toxicology.1 In conventional DNA microarray hybridization, the

reaction takes place after diffusion process, in which fluo-

rescently labeled targets and probes come into contact by natural

concentration gradient. Diffusion of DNAs in water is very slow

due to their low diffusion constant (D). Statistically, DNAs take

aNanoelectronics and MEMS laboratory, National Electronics andComputer Technology Center (NECTEC), 112 Paholyothin Rd., Klong1, Klong Luang, Pathumthani, 12120, Thailand. E-mail: adisorn.tuantranont@nectec.or.th; Fax: +66-2-564-6756; Tel: +66-2-564-6900ext. 2111; sastra.cha@biotec.or.thbNational Center for Genetic Engineering and Biotechnology (BIOTEC),113 Paholyothin Rd., Klong 1, Klong Luang, Pathumthani, 12120,Thailand; Fax: +66-2-564-6707; Tel: +66-2-564-6700

† These authors contributed equally to the work.

This journal is ª The Royal Society of Chemistry 2011

more than 24 hours to travel over a short distance of a few

millimetres (this can be estimated from the equation of diffusion

length (L): L ¼ (Dt)0.5 where t is the diffusion time and D is the

DNA diffusion constant, which is in the order of 10�6 to 10�7 cm2

s�1).1 Typically, conventional static hybridization requires long

hybridization times (usually 6–20 h) to achieve sufficient

hybridization signals and uniformity of signals across an array.2

The awareness of this limitation has led various research

groups to develop ways to improve DNA hybridization effi-

ciency. Some of the developed methods include chaotic mixing

through coordination of pump and valve,3 surface acoustic wave

(SAW) generated by LiNbO3 agitation chip,4 recirculation flow,5

rotation of magnetic stirring bar,6 cavitation microstreaming

induced by vibration of bubble membrane,7 and rotating

microchamber.8,9 However, most of the aforementioned tech-

niques require specialized components, which are difficult to

implement in standard DNA microarray hybridization systems.

Another effective mixing technique applicable to microarrays is

the acoustic streaming method.10,11 In addition, it is suitable for

solution containing cDNA. It is considered to be one of the most

Lab Chip, 2011, 11, 1059–1064 | 1059

Fig. 1 Experimental setup of the DNA microarray with acoustic

streaming hybridization. (a) Schematic fabrication diagram of a micro-

chamber. (b) Overall view of the assembled hybridization chambers. Each

glass slide consists of two microarrays. (c) Acoustic streaming inside the

microchamber is generated by absorption of mechanical energy from

ultrasonic wave propagating in upward direction (red arrow).

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online

practical agitation techniques for DNA hybridization process

since there is no need of an external pumping unit or complicated

mechanical parts for moving the microarray slide and coverslip.

However, fabrication and integration of the acoustic streaming

source to the microarray is still a challenging task. For example,

the previously developed microstreamed DNA hybridization

systems require a modification to the microarrays and cover-

slip,6,7,12–14 or need some material coating, which could cause

contamination. It is desirable to integrate a piezoelectric trans-

ducer into DNA hybridization systems with no modification to

microarray and coverslip, so that the hybridization device can be

used without any hybridization procedure modifications.

During acoustic streaming, a steady flow pattern is formed as

a result of mechanical energy transfer from acoustic wave to the

fluid. Flow occurs in the direction of acoustic beam, away from

the transducer. An ultrasonic wave generated by a piezoelectric

transducer (PZT) is used to create acoustic streaming. An

ultrasonic wave propagating across an interface results in

transmission and reflection. Since reflection could result in less

effective induction of acoustic streaming in the fluid, a coupling

medium should be properly selected to minimize the loss of

energy at the interface. In this case, the medium material whose

acoustic impedance matches to that of glass slide is preferred.

Metals such as aluminium15 or steel16 are widely used while

liquids such as water,17 glycerin,17 or gel18 are among attractive

candidates. Since temperature control is crucial for biochemical

reaction including DNA hybridization, a medium with high heat

capacitance (Cp) such as water is needed to prevent the heat flux

generated by piezoelectric transducer from interfering with the

hybridization temperature within the microchamber.

In this work, a simple and low cost device for enhancing the

efficiency of DNA hybridization is developed based on an

acoustic streaming method. The configuration of the device

described herein is designed to be compatible with the standard

microarray setup. The fabrication process and the hybridization

enhancement of the ultrasonic micromixer using piezoelectric

actuation through a coupling medium are reported.

Materials and methods

Design and fabrication

The hybridization chamber consisted of two units (Fig. 1(a and

b)) including ultrasonic piezoelectric transducers and the corre-

sponding coupling liquid. The overall dimension of the device

was 5 � 12 � 2 cm3. In order to optimize the transmission of

acoustic energy, three coupling liquids were tested. They were

ultrasonic gel, 3 � SSC (Saline Sodium Citrate) solution, and

glycerin having impedance values of Zgel ¼ 1.52 � 106 N s m�3,

Zsolution ¼ 1.48 � 106 N s m�3, Zglycerin ¼ 2.42 � 106 N s m�3,

respectively. To perform hybridization, a thin layer of coupling

liquid was first placed in the 0.8 mm thick circular well such that

the liquid covered the entire area of the piezoelectric transducer

(Fig. 1(a and b)). A DNA microarray glass slide was then placed

on top of the coupling liquid-layered piezoelectric transducers

and a 22 � 25 mm2 glass LifterSlip (Erie Scientific, Portsmouth,

USA) was used to cover the spotted DNA microarray. 35 mL of

hybridization mixture (containing fluorescently labeled cDNA

targets) was gently applied to the microarray, allowing the

1060 | Lab Chip, 2011, 11, 1059–1064

mixture to seep under the coverslip that covered the whole area

of microarray spots. In order to provide humidity and thus

prevent overdrying, two rectangular wells (22 � 15 � 1 mm3)

were created at two opposite ends of the cartridge to accom-

modate 3 � SSC solution. Another piece of aluminium acting as

a lid was screwed on the hybridization chamber to provide a tight

seal. O-Ring was implemented to achieve sufficient sealing as

seen in Fig. 2. The transducers were controlled by an ac signal

generated by a driving circuit.

Our hybridization design is based on the concept of acoustic

streaming. Travelling ultrasonic wave generates flow in the

direction away from the surface of the glass slide as shown in

Fig. 1(c). Since flow cannot pass through the coverslip, it diverges

in left and right directions resulting in global convective flow

inside the microchamber, allowing transportation of target DNA

to their corresponding probes.

Fluid mixing and visualization

Fluid mixing experiments were performed at room temperature

(25 �C). 20 mL of deionized water, loaded by a micropipette, was

used as a liquid sample. 10 mL of dye was added for flow

This journal is ª The Royal Society of Chemistry 2011

Fig. 2 View of the dynamic hybridization device. DNA microarrays are

placed inside the hybridization cartridge above the piezoelectric ultra-

sonic transducers and coupling liquid. An O-ring rubber is used to

provide a tight seal. Each spotted DNA microarray area is covered with

a glass coverslip (not shown in picture).

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online

visualization after sample loading. The fluidic dye, which has

comparable size to DNA, was used to directly represent DNA in

the hybridization solution.

Particle tracking velocimetry (PTV) method was then used to

further study the local behavior of fluid and particles influenced

by acoustic streaming. The flow activity of colloidal polystyrene

(PS) microspheres with 4.5 mm diameter (Polysciences, Inc.,

Warrington, PA, USA) was observed under a CCD camera-fitted

optical microscope (Olympus CX41, Olympus, Inc., PA, USA) at

200� magnification. Nine locations throughout the chamber

were monitored. It should be noted that micron-sized particles

were used as opposed to nanoparticles because the precision of

PTV is limited by the optical resolution of a few microns. Despite

the size discrepancy, PTV should be applicable to envisage the

flow of DNA in the hybridization solution under acoustic stream

since transportation in this case is dominated by fluid convection.

The transducers oscillated at the resonance frequency of 1.67

MHz and 24 Vrms sinusoidal amplitude. Due to the fact that the

dramatic heat flux was generated during piezoelectric trans-

ducers activation, it could not operate in a continuous manner.

For the control experiment, the transducers were turned on and

off using a square wave signal. The transducers operated in

different modes by varying duty cycles (%) and periods (T) of the

square wave signal. The instances of signal variation included

20% (T ¼ 5 s), 25% (T ¼ 4 s), 33% (T ¼ 3 s), 40% (T ¼ 5 s), 50%

(T ¼ 2 s) and 50% (T ¼ 4 s).

Parasite culture and total RNA extraction

Asynchronous cultures of Plasmodium falciparum strain K1 were

grown in vitro by standard methods.19 Parasites were harvested

and freed from the erythrocytes by saponin lysis. Total RNA was

purified from parasite cells using Trizol reagent (Invitrogen)

according to manufacturer’s instructions. The quality and

quantity of the total RNA were assessed to ensure purity using

a NanoDrop spectrophotometer (NanoDrop Technologies Inc.,

Wilmington, USA).

This journal is ª The Royal Society of Chemistry 2011

Fluorescent cDNA preparation and hybridization

Amino-allyl cDNA labeling and hybridization were performed

as previously described.20 DNA microarrays used in the study

contain 8088 70-mer oligonucleotide probes representing the

majority of annotated open reading frames (ORF) as listed in

PlasmoDB 4.4.21 The DNA microarrays were fabricated on

polylysine coated glass slides using a new generation ultrafast,

linear servo driven DeRisi microarrayer controlled by an

ArrayMaker software (http://derisilab.ucsf.edu/microarray/

software.html).

For each cDNA synthesis reaction, 10–15 mg of total RNA was

mixed with 5 mg oligodT(21) primer and reverse-transcribed to

produce amino-allyl-dUTP(Sigma)-labeled cDNA using

Imprompt II reverse transcriptase (Promega). The labeled cDNA

was coupled with either monoreactive Cy3 or Cy5 (Amersham

Biosciences) as described previously.22 Purified Cy3- and Cy5-

labeled cDNA of equal pmole amount (26 pmole each) were

resuspended in 70 mL Pronto hybridization solution (Corning),

mixed thoroughly, and split into two equal volumes (35 mL each)

for hybridization of DNA microarrays by either static or piezo-

mediated methods. This was to ensure that the labeled cDNA used

for both hybridization methods came from the same pool of

labeled cDNA so that the results from both hybridization methods

could be compared. In the end, each hybridization experiment

contained 13 pmole of each of the fluorescent dye-labeled cDNA.

The static hybridizations were carried out for 16 hours while

dynamic hybridization time was varied from 2 to 16 hours to

determine a minimum time that yields better result than 16 h static

hybridization. All hybridizations were conducted at 42 �C.

Image analysis

After washing, arrays were scanned using an Axon GenePix

4000B microarray scanner and the intensity of spots was quan-

tified using GenePix Pro 6.0 Software (Axon Instrument, Inc.).

The scanning was done at 100% laser power, with the same PMT

gain settings for all arrays. Further analysis of the spot intensity

and other parameters such as signal-to-noise ratio (SNR), % >

B532 + 2SD, % > B653 + 2SD, and Rgn R2, which were used to

distinguish good/effective spots, was carried out in Microsoft

Excel. We used the criteria that were previously reported by

Yatskou et al.23 for spot quality analysis. These criteria were

SNR532 $ 3, SNR653 $ 3, ‘% > B532 + 2SD’ $ 55 or ‘% > B653

+ 2SD’ $ 55, and Rgn R2 $ 0.5.

Results and discussion

Choice of coupling medium

Three coupling media, namely ultrasonic gel, glycerin and 3 �SSC, were tested to determine which one would be the most

appropriate medium for the piezoelectric microagitation of our

hybridization device. It was found that the coupling media per-

formed differently under the same duty cycle and period

parameters. The ultrasonic gel effectively transmitted the ultra-

sound from the piezoelectric transducers. However, it possessed

many disadvantages. Firstly, heat generated by the transducers

over an hour of operation caused the temperature of the gel to

increase dramatically, resulting in bubble formation in the gel.

Lab Chip, 2011, 11, 1059–1064 | 1061

Fig. 3 Photographs of dye solution mixing in the hybridization micro-

chamber after 10 min (a) without acoustic streaming (PZT off) and (b)

with the PZT agitation at 1.67 MHz and 24 Vrms.

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online

Secondly, undesired heat transported from the PZT disk to the

microarray through the gel affected the optimal hybridization

temperature, which should be maintained constant at 42 �C.

Moreover, prior to microarray slide washing and scanning,

cleaning was required to remove all traces of the gel that had

stuck to the back of the microarray slide and this task proved

difficult and problematic. Even with thorough cleaning, a stain

which would fluoresce and thus interfere with the fluorescence

signal emanating from the microarray was still visible. Similarly,

glycerin provided comparable ultrasonic transmission to that of

the gel. However, issues of heat generation, bubble formation,

and difficult stain removal still remained. 3 � SSC, an aqueous-

based solution, provided the best result. Firstly, neither

temperature rise of the coupling liquid and the microarray nor

formation of bubbles could be observed since the specific heat

capacity of water was relatively high. Additionally, aqueous 3 �SSC solution could be easily removed from the back of the

microarray, without leaving any stain. As a result, the fluores-

cence signal was not interfered. Moreover, aqueous solution

yielded better adhesion to the glass slide.

Selection of optimal duty cycle

To obtain a suitable mode of operation for the ultrasonic

transducer, total operation time, amount of heat generation,

formation of bubble and the overall fluidic movement were

optimized by varying duty cycle and period of transducer on–off

signal. For 50% (T¼ 4 s) and 40% (T¼ 5 s) duty cycles, the active

time of the transducer was too long (2 s) because large amount of

heat generated from the piezoelectric transducers and bubbles

formation was observed inside the coupling media. Hence, these

two modes of operation were not desirable. Next, the activation

time was reduced by decreasing period, leading to the conditions:

50% (T¼ 2 s) and 33% (T¼ 3 s). In these cases, a number of small

bubbles generated by heat were still observed inside the cou-

plants because the off time for heat release became too short.

Lastly, the activation time was decreased by reducing the duty

cycle to 20% (T ¼ 5 s) and 25% (T ¼ 4 s). These conditions were

proven to be suitable in terms of heat reduction and overall fluid

movement. By taking the overall operation time into account, the

most appropriate condition was 25% (T ¼ 4 s).

Fluid mixing visualization under dynamic hybridization

Fluidic dye was added into the microchambers to observe lateral

fluidic movement with or without piezoelectric agitation. There

are two important effects resulting from the presence and

absence of piezoelectric agitation: (1) acoustic streaming, which

plays an important role when the piezoelectric disk was activated

and (2) the diffusion, which took place when there was no acti-

vation. Each piezoelectric disk was alternately turned on for 1 s

with a 3 s pause before activating the other disk. The 25% duty

cycle (T ¼ 4 s) was found to be the most optimal condition since

it provided significant liquid circulation and reduction of heat

generated from the piezoelectric transducers.

It was found that fluidic dye distributed throughout the

microchamber within 10 min under acoustic agitation while it

took more than 1 h for the dye to diffuse half way across the

chamber when acoustic streaming was absent (Fig. 3). With our

1062 | Lab Chip, 2011, 11, 1059–1064

piezo-mediated microagitation system, the dye could quickly

flow under the influence of both diffusion (PZT off) and acoustic

streaming (PZT on), which was much more effective than the dye

movement that relied solely on diffusion in the absence of

agitation.

Flow visualization by PVT experiments showed that poly-

styrene (PS) microspheres moved only slightly, but on average

they remained static without acoustic field agitation. In contrast,

significant movement of PS beads was observed when an acoustic

field was present. This clearly illustrated that acoustic streaming

in the hybridization microchamber generated by ultrasonic

waves was directly responsible for the increased movement of the

beads. Moreover, the bead experiment also revealed the presence

of vertical movement of beads in the chamber. The finding that

acoustic streaming caused the beads to move in both horizontal

and vertical directions suggested that acoustic streaming could

enhance hybridization efficacy by causing DNA targets to move

in similar manners in the hybridization chamber. The PVT

results agree well with those from microfluidic dye visualization.

This confirms the applicability of PVT for flow analysis of the

acoustic streaming.

Enhancement of DNA hybridization via piezoelectric transducer-

mediated agitation

Hybridization performance of our piezoelectric transducer-

mediated agitation was compared to that of standard hybrid-

ization method that uses coverslips and incubation in a water

bath. We chose to exploit the availability of P. falciparum DNA

microarrays and the well-established parasite culture facilities in

our laboratory to carry out such a head-to-head performance

comparison. Our experimental design was aimed to evaluate the

performance of our agitation system under ‘real’ microarray

experiments using RNA extracted from biological samples and

8088-oligo-containing DNA microarrays that have been used

previously in many transcriptomic profiling studies.20,21,24,25 In

this regard, our design is different from that of most previous

studies investigating hybridization enhancement, which used

chemically synthesized oligo primers hybridizing to a ‘test’ array

that contains only a few dozens or hundreds of spots.5,8,14,26

We hybridized equal pmole amounts of fluorescent dye-

labeled cDNA (that came from the same pooled sample) to our

P. falciparum DNA microarrays. The time for static hybridiza-

tion was fixed at 16 hours while that of dynamic hybridization

was varied from 2 to 16 hours. Scanned pictures of microarrays

This journal is ª The Royal Society of Chemistry 2011

Fig. 4 Microarray scan pictures of P. falciparum DNA microarrays with (a) 16 h static hybridization, (b) 4 h dynamic hybridization and (c) 16 h

dynamic hybridization. Equal amounts (13 pmole) of Cy3- and Cy5-labeled cDNA targets (from a pooled sample) were used in the hybridization assays

at 42 �C.

Fig. 5 Comparison of (a) the average hybridization signal intensity

(background-corrected) obtained from 16 h piezoelectric transducer-

mediated agitation and standard 16 h static hybridization experiments.

The signal intensities of Cy3 and Cy5 channels were shown. (b) Effective

spots analysis. Effective spots were defined as having SNR532 or

SNR635 $ 3, ‘% > B532 + 2SD’ $ 55 or ‘% > B635 + 2SD’ $ 55, and

Rgn R2 $ 0.5.

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online

after 16 h static, 4 h dynamic and 16 h dynamic hybridizations

are shown in Fig. 4(a–c), respectively. It is seen that the 4 h

dynamic hybridization produces comparable hybridization

signals to 16 h static hybridization and considerably improved

overall hybridization signals are obtained from dynamic

hybridization with the same hybridization time of 16 hours. The

fluorescent signals were then quantified and the results are shown

in Fig. 5. Fig. 5(a) shows that the background-corrected mean

signal intensities obtained from our agitation system were higher

in both fluorescent channels than those from static hybridization.

In comparison with the 16 h static experiment, 16 h dynamic

hybridization produced approximately 33% and 24% increase in

the mean signal intensities in the Cy3 and Cy5 channels,

respectively. The result suggests that our piezoelectric-mediated

microagitation could reduce the required hybridization time or

improve the overall hybridization qualities.

Effective spot analysis was then used to statistically analyze

fluorescent intensity data from DNA microarrays. In our study,

effective spots are defined based on the parameters suggested by

the GenePix Pro user’s manual and a previous study on advanced

spot quality analysis by Yatskou et al.23 This analysis has taken

into account of appropriate statistical parameters so that the

analyzed results are globally valid and meaningful. These include

(1) signal-to-noise ratio (SNR) $ 3, (2) features with more than

55% of pixels that have intensity values greater than two stan-

dard deviation (2SD) above the median background intensity

value in either Cy3 or Cy5 channel (‘% > B532 + 2SD’ parameter

$ 55% or ‘% > B635 + 2SD’ parameter $ 55%, respectively), and

(3) Rgn R2 $ 0.5, which is the coefficient of determination for

the least-squares of regression fit for a given feature. SNR, % >

B532 + 2SD, and % > B635 + 2SD parameters are used to

identify those features that have pixel intensity value significantly

above that of the background, and therefore have decent signal

quality and are probably reliable values, while the Rgn R2

determines the uniformity of a given spot, and is thus a good

parameter for filtering those non-uniform spots with low Rgn R2

values that may have resulted from inefficient hybridization.

Fig. 5(b) indicates that when piezo-mediated agitation was

applied, the number of effective spots dramatically increased for

all of the filtering parameters used. Our analysis shows that

This journal is ª The Royal Society of Chemistry 2011 Lab Chip, 2011, 11, 1059–1064 | 1063

Publ

ishe

d on

02

Febr

uary

201

1. D

ownl

oade

d by

CA

SE W

EST

ER

N R

ESE

RV

E U

NIV

ER

SIT

Y o

n 29

/10/

2014

20:

26:2

5.

View Article Online

additional 1186 and 1882 spots can be recognized as effective

spots under agitated hybridization when filtering by SNR532 and

SNR635, respectively. This is consistent with the number of

additional effective spots when the % $ B532 + 2SD, B635 +

2SD, and Rgn R2 parameters were used (1102 for % $ B532 +

2SD, 1882 for % $ B635 + 2SD, and 2002 for Rgn R2).

Spot quality control is a pivotal part of DNA microarray data

analysis as it is paramount to distinguish between good/effective

spots from bad spots that may represent false positives or noise

and thus compromise the subsequent biological interpretation.

We have shown here that our piezo-mediated agitation system

can increase the number of effective spots, which would other-

wise be excluded from downstream microarray analysis due to

inefficient static hybridization. This suggests that our agitation

system is a powerful tool for increasing sensitivity of detecting

targets that may have produced low hybridization signals (due to

their low expression level) but were biologically relevant. The

power of discovering more effective spots would be extremely

beneficial in any kind of transcriptomic profiling studies because

it would help better microarray data interpretation and allow

identification of, for example, disease-related biomarkers that

could not be detected by conventional hybridization method.

Conclusions

In conclusion, we have developed a simple and low-cost dynamic

DNA hybridization device, which employs a novel hybridiza-

tion-enhancing approach in which piezoelectric transducer and 3

� SSC as the coupling medium were used to generate efficient

agitated mixing within the microarray chamber. Another

attractive feature of our hybridization device is that it can be used

without the need to make any modifications to the standard

hybridization procedures. Our static versus dynamic hybridiza-

tion comparison under ‘real’ microarray experimental parame-

ters demonstrated that the required hybridization time may be

substantially reduced from 16 h to 4 h by the use of dynamic

hybridization scheme. Moreover, dynamic hybridization

increased the average fluorescent signal intensities and the signal-

to-noise ratios in both Cy3 and Cy5 channels over a static

hybridization, allowing more effective spots to be included in the

subsequent microarray data analysis. This provides an immense

advantage in transcriptomic profiling studies as the additional

good spots detected under dynamic hybridization, but not static

hybridization could have potential biological or medical signifi-

cance in, for example, drug target and disease biomarker

discovery studies.

Acknowledgements

We would like to acknowledge Drs Joseph Derisi and Jennifer

Shock at University of San Francisco for providing the oligo-set,

training and facilities for the production of DNA microarray

1064 | Lab Chip, 2011, 11, 1059–1064

slides through generous support from the Howard Hughes

Medical Institute (HHMI, USA). S.K. is an international

research scholar of HHMI, USA. We would also like to thank

Drs Philip James Shaw and Chairat Uthaipibull for useful

discussion on DNA microarray. A.T. expresses his great grati-

tude to the Thailand Research Fund (RSA5380005) for a

research career development grant.

References

1 J. M. Hertzsch, R. Sturman and S. Wiggins, Small, 2007, 3, 202–218.2 M. Schena, Microarray Biochip Technology, Eaton Publishing,

Natick, MA, 2000.3 M. K. McQuain, K. Seale, J. Peek, T. S. Fisher, S. Levy,

M. A. Stremler and F. R. Haselton, Anal. Biochem., 2004, 325, 215–226.

4 A. Toegl, R. Kirchner, C. Gauer and A. Wixforth, J. Biomol. Tech.,2003, 14, 8.

5 H. H. Lee, J. Smoot, Z. McMurray, D. A. Stahl and P. Yager, LabChip, 2006, 6, 1163–1170.

6 R. H. Liu, J. Yang, M. Z. Pindera, M. Athavale and P. Grodzinski,Lab Chip, 2002, 2, 7.

7 R. H. Liu, R. Lenigk, R. L. Druyor-Sanchez, J. N. Yang andP. Grodzinski, Anal. Chem., 2003, 75, 1911–1917.

8 J. Vanderhoeven, K. Pappaert, B. Dutta, P. Van Hummelen andG. Desmet, Anal. Chem., 2005, 77, 4474–4480.

9 M. A. Bynum and G. B. Gordon, Anal. Chem., 2004, 76, 7039–7044.10 L. A. Kuznetsova and W. T. Coakley, Biosens. Bioelectron., 2007, 22,

1567–1577.11 D. Li, Encyclopedia of Microfluidics and Nanofluidics, Springer, New

York, 2008, pp. 28–33.12 Z. Guttenberg, H. Muller, H. Habermuller, A. Geisbauer, J. Pipper,

J. Felbel, M. Kielpinski, J. Scriba and A. Wixforth, Lab Chip, 2005,5, 308–317.

13 A. Wixforth, C. Strobl, C. Gauer, A. Toegl, J. Scriba and Z. vonGuttenberg, Anal. Bioanal. Chem., 2004, 379, 982–991.

14 Z. Q. Pan, Y. Q. Li, D. R. Zhou, J. Tang, M. H. Zhang, P. F. Mao andZ. H. Lu, Anal. Biochem., 2008, 376, 280–282.

15 O. Manneberg, J. Svennebring, H. M. Hertz and M. Wiklund, J.Micromech. Microeng., 2008, 18, 095025.

16 G. Rus, S. C. Wooh and R. Gallego, Ultrasonics, 2005, 43, 391–395.17 J. E. Griffin, Phys. Ther., 1980, 60, 1010–1016.18 M. Berson, J. M. Gregoire, F. Gens, J. Rateau, F. Jamet, L. Vaillant,

F. Tranquart and L. Pourcelot, European Journal of Ultrasonic, 1999,10, 53–63.

19 W. Trager and J. Jensen, Science, 1976, 193, 673–675.20 Z. Bozdech, M. Llinas, B. L. Pulliam, E. D. Wong, J. C. Zhu and

J. L. DeRisi, PLoS Biol., 2003, 1, 85–100.21 J. L. Shock, K. F. Fischer and J. L. DeRisi, GenomeBiology, 2007, 8,

R134.22 Z. Bozdech, J. C. Zhu, M. P. Joachimiak, F. E. Cohen, B. Pulliam and

J. L. DeRisi, GenomeBiology, 2003, 4, R9.23 M. Yatskou, E. Novikov, G. Vetter, A. Muller, E. Barillot, L. Vallar

and E. Friederich, BMC Res. Notes, 2008, 1, 80.24 K. Ganesan, N. Ponmee, L. Jiang, J. W. Fowble, J. White,

S. Kamchonwongpaisan, Y. Yuthavong, P. Wilairat andP. K. Rathod, PLoS Pathog., 2008, 4, e1000214.

25 G. A. Hu, A. Cabrera, M. Kono, S. Mok, B. K. Chaal, S. Haase,K. Engelberg, S. Cheemadan, T. Spielmann, P. R. Preiser,T. W. Gilberger and Z. Bozdech, Nat. Biotechnol., 2010, 28, 91–U126.

26 P. K. Yuen, G. S. Li, Y. J. Bao and U. R. Muller, Lab Chip, 2003, 3,46–50.

This journal is ª The Royal Society of Chemistry 2011