Available online at www.sciencedirect.com

www.elsevier.com/locate/cap

www.kps.or.kr

Current Applied Physics 8 (2008) 687–691

DNA microarray scanner with a DVD pick-up head

Kyung-Ho Kim a, Seung-Yop Lee a,*, Sookyung Kim b, Seong-Gab Jeong c

a Department of Mechanical Engineering, Sogang University, Seoul 121-742, Republic of Koreab Nanostorage Co. Ltd., Seoul 120-090, Republic of Korea

c Department of Mechantronics, Cheongju College of Korea Polytechnic IV, Cheongju, Chungbuk 36-290, Republic of Korea

Received 29 July 2006; received in revised form 9 April 2007; accepted 27 April 2007Available online 10 October 2007

Abstract

A low-cost highly sensitive DNA microarray scanner for fluorescent detection is developed based on the pick-up head of a commer-cially available optical storage device, DVD. A laser beam of 650 nm, generated by a DVD laser diode, is used for dynamic auto-focusingas well as the excitation of Cy5 fluorescent dye. The fluorescence intensity emitted from Cy5 dye is measured by a photomultiplier tube(PMT). In contrast to other microarray scanners, the DVD-based scanner offers the auto-focusing function using the focus error signal(FES) and a voice coil motor (VCM), and this enables fast response, high accuracy and compact size. The fluorescence-detecting per-formance of the scanner is inspected by using a commercial BAC (bacterial artificial chromosome) oligonucleotide chip and a scannerevaluation microarray (DS01). Experiments have shown that the DVD-based scanner meets the limit of detection, ensuring the feasibilityof a low-cost, highly sensitive DNA microarray scanner.� 2007 Elsevier B.V. All rights reserved.

PACS: 42.62.Be; 42.79.Vb; 87.62.+n; 87.80.Tq

Keywords: DNA microarray; DNA chip; Biochip; DVD; Fluorescent scanner; Auto-focusing actuator

1. Introduction

Today, most DNA or protein microarrays use fluorescentdyes as a DNA marker. In general, the fluorescence-detect-ing technology has shortcomings, such as weak fluorescentlight and a highly controlled positioning mechanism. It isknown that mechanical precision to 1 lm or less and a cool-ing device are required to meet the measurement specifica-tions of commercial confocal scanners. The high cost ofthe detection mechanism prevents a wide spread in theDNA-chip market, especially when high-density diagnosticchips are used. Therefore, detection technology satisfyingboth high sensitivity and low-cost becomes one of the essen-tial factors to expand the DNA-chip market and the relatedtechnology and industry [1,2].

1567-1739/$ - see front matter � 2007 Elsevier B.V. All rights reserved.

doi:10.1016/j.cap.2007.04.047

* Corresponding author. Tel.: +82 2 705 8638; fax: +82 2 712 0977.E-mail address: sylee@sogang.ac.kr (S.-Y. Lee).

Various researches have used optical storage technologyas the detection mechanism for DNA microarrays. It wasshown that CD technology could build a detection systemwith equivalent or even better performance than conven-tional scanners [3], and a lab on a chip (LOC) system hasbeen developed for application in DNA analysis using aCD disk [4,5].

2. Auto-focusing technology

When scanning a large surface area of a DNA micro-array, it is important to keep the distance between theobjective lens and the microarray surface constant withhighest accuracy. The main advantage of the DVD pick-up-based fluorescent detection is to easily implement adynamic auto-focusing mechanism by attaching a reflectivelayer on the commercial microarray. Fig. 1 shows the func-tional difference between a conventional single pointmethod and an auto-focusing control. Fig. 1a shows that

Fig. 1. Comparison of (a) single point method and (b) auto-focusing method for misaligned or warped microarrays.

Fig. 2. (a) Schematic diagram of optical path and (b) DNA microarray with a reflective layer.

688 K.-H. Kim et al. / Current Applied Physics 8 (2008) 687–691

Fig. 4. Measurements of signal-to-noise ratio (SNR) of two cases: with/without auto-focusing.

K.-H. Kim et al. / Current Applied Physics 8 (2008) 687–691 689

a misaligned or warped microarray can lead to unfocusedimages when the single point method is used. However,Fig. 1b shows the main advantage of the auto-focusingmethod, which compensates the mechanical tolerance ofthe DNA-chip surface. In general, the relative position ofthe objective lens and the microarray surface is determinedby 5� of freedom (three translations and two rotations).For the precise detection of the fluorescent marker, the sur-face translation motions (X and Y) must be less than 1 lm.A DVD optical pick-up, which is widely used in commer-cial optical storage devices, enables a low-cost sub-micronfocusing control [6,7].

3. Fluorescent detection mechanism

Fig. 2a shows a schematic diagram of the optical pathsof the proposed system. There are two different opticalpaths: one is for auto-focusing and the other is for fluores-cent detection. A laser diode generates a laser beam ofwavelength 650 nm. Most of the laser beam is reflectedby a dichroic filter (Chroma Optics Z660RDC), and thepart of the laser beam passing through the dichroic filteris used to regulate the laser power by a front monitor sen-sor. The laser beam transmitted through the dichroic filteris focused on the reflective layer, and the reflected beam isfinally focused onto a four-quadrant photodiode (PD) afterpassing through two dichroic filters. The focus error signalis calculated by using the intensities of the four dividedregions of the PD. The PD generates a focus error signalwhose magnitude is dependent upon the distribution ofthe beam spot across its four divided regions. The resultingservo signal is used to drive the VCM actuator in such away that the objective lens is shifted to a point where itsfocal point falls upon the spot center. The laser beamfocused on the DNA-chip surface excites Cy5 dye withDNA probes, emitting a fluorescent signal of 670 nm.The collimated emission beam is then refocused by a detec-tion lens and passes through a pinhole into the photomul-tiplier tube (PMT). The analogue electric signal from thePMT represents the intensity of the fluorescent signal.

Fig. 3. Experimental set-up of fluorescence-detec

The tracking motion is actuated and controlled by a precisepositioning stage.

In order to implement the dynamic auto-focusing func-tion, a new microarray slide structure is designed and man-ufactured by attaching a reflective layer on a commercialDNA microarray slide. Fig. 2b shows the slide structureconsisting of two different layers. Since the DNA probesurface is directly attached on the reflective layer, the laserbeam is focused on the microarray surface and the reflec-tive layer simultaneously.

4. Experimental results

Fig. 3 shows the experimental system to test the fluores-cence-detecting performance of the proposed DNA micro-array scanner. The focusing motion of the objective lens isautomatically controlled and the focal point falls upon thespot center. The tracking motion is actuated by a precisemotorized stage to scan the microarray surface. In orderto increase the fluorescence signal, the signal-to-noise ratio(SNR) is to be improved. To determine SNR, the followingformula is commonly used:

ting mechanism with a DVD pick-up head.

690 K.-H. Kim et al. / Current Applied Physics 8 (2008) 687–691

SNR ¼ Signal intensity� Background intensity

Standard deviation of background: ð1Þ

Eq. (1) implies that one may increase the signal or/and re-duce the background in order to improve SNR. Typically,the limit of detection (LOD) is defined as the minimumdetectable signal for which the SNR is 3. The sensitivityof a DNA-chip scanner is inversely related to the limit ofdetection – the detection sensitivity increases as the limitof detection decreases.

Fig. 4 shows a comparison of SNR for the two caseswith and without auto-focusing function. In the experi-ment, regular line patterns with constant dye concentration

Fig. 5. Experimental results using auto-focusing control (

Fig. 6. Detection images of evaluation slide (a) wit

are painted on a slide. Then, SNR was measured along thespot centers with constant fluorescent intensities afteradjusting the initial position and light intensity on the spotcenter. For the auto-focusing case, the magnitude and uni-formity of SNR are improved, as shown in Fig. 4. The ele-vation of SNR is due to the uniform background by thedynamic auto-focusing, the increase of fluorescent signalby the reflective layer, and the decrease of noise. The weakfluorescent level without auto-focusing is below the limit ofdetection, and it does not guarantee the minimum detect-able signals. Thus, the dynamic auto-focusing mechanismimplemented by the DVD pick-up system is required forthe scanner to meet the minimum fluorescence level. The

a) with auto-focusing and (b) without auto-focusing.

h auto-focusing and (b) without auto-focusing.

K.-H. Kim et al. / Current Applied Physics 8 (2008) 687–691 691

use of the dynamic auto-focusing and reflective layer canremarkably improve the detection performance. In con-trast, the non-uniform profile of SNR without auto-focus-ing is mainly caused by the background noise, surfacevariability, and misaligned slide surface.

Fig. 5 shows the fluorescent intensities of the auto-focus-ing and defocused cases. The DNA microarray used in thisexperiment is a BAC (bacterial artificial chromosome) oli-gonucleotide chip, which is permitted as a commercial diag-nostic DNA chip by Korea Food and Drug Administration(KFDA). As shown in Fig. 5, the fluorescent intensity andthe signal uniformity of the auto-focusing case are muchhigher than those without auto-focusing. For the case with-out auto-focusing, there is a remarkable difference betweenthe signal levels measured at two different spot centers, andthe non-uniform signal pattern decreases SNR.

Fig. 6 shows scanning images of an evaluation micro-array slide, which is designed for quantitatively analyzingthe dynamic range, detection limit, and uniformity ofmicroarray scanners. The experiment uses the scanner cal-ibration slide (DS01) by Full Moon Biosystems Inc. Thisslide contains an array block (32 columns · 12 rows) ofdilution series of Cy5 fluorescent dyes. The slide can alsodetermine the stability of the laser and the alignment ofthe scanner because each column contains 12 repeats ofeach dye dilution. The detection performance of the auto-focusing case is shown to be about two times greater thanthat without auto-focusing in Fig. 6. The non-uniformitycaused mainly by the surface variability and misalignmentcan be compensated by the dynamic auto-focusing func-tion. These experimental results show that the pick-up-based auto-focusing mechanism improves the detectionperformance, compared to conventional diagnostic scan-ners without auto-focusing.

5. Conclusions

This study has demonstrated a new DNA microarrayscanner using a commercially available DVD optical

pick-up. A laser beam of 650 nm, generated by a DVDlaser diode, is used for the dynamic auto-focusing as wellas the excitation of Cy5 dye. The use of the dynamicauto-focusing and the reflective layer attached on DNA-chip slides improves the sensitivity of fluorescent detection.Experiments using a typical evaluation slide haveshown that the new DVD-based scanner meets the limitof detection (SNR = 3) which is the minimum detectablesignal. The new DNA microarray scanner based on opticalpick-up technology enables high sensitivity as well as low-cost.

Acknowledgements

This study was supported by a grant from the Techno-logical Development Program by the Ministry of Com-merce, Industry and Energy (10006623-2006-22) and agrant from the Seoul Research and Business DevelopmentProgram (10816).

References

[1] M. Schena, DNA Chip Analysis, John Wiley & Sons, Inc., New Jersey,2003.

[2] M. Schena, DNA Chips, Oxford University Press, New York, 1999.[3] F. Perraut, A. Lagrange, P. Pouteau, O. Peyssonneaux, P. Puget, G.

McGall, L. Menou, R. Gonzalez, P. Labeye, F. Ginot, Biosens.Bioelectron. 7 (2002) 803–813.

[4] T. Matsui, T. Shimonura, in: Optical Data Storage Topical Meeting,2006, pp. 103–105.

[5] H. Kido, A. Maquieira, B. Hammock, Anal. Chim. Acta 411 (2000) 1–11.

[6] M. Bohmer, F.R. Pampaloni, M. Wahl, H.-J. Rahn, R. Erdmann, J.Enderlein, Rev. Sci. Instrum. 72 (2001) 4145–4152.

[7] K. Kim, S.-Y. Lee, S.H. Lee, S. Kim, S.G. Jeong, Microsyst. Technol.13 (2007) 1359–1369.