Anal. Chem. 2003, 75, 3250-3255
Technical Notes
Automated Bead Alignment Apparatus Using a Single Bead Capturing Technique for Fabrication of a Miniaturized Bead-Based DNA Probe Array Hideyuki Noda,* Yoshinobu Kohara, Kazunori Okano, and Hideki Kambara
Hitachi, Ltd., Central Research Laboratory, 1-280 Higashi-Koigakubo, Kokubunji, Tokyo 185-8601, Japan
We have developed an automated bead alignment apparatus for fabricating a bead-based DNA probe array inside a capillary. The apparatus uses 16 micro vacuum tweezers to extract single beads from among a large amount of beads in bead stock wells. It then manipulates single beads into the probe array capillaries. Single 100µm-diameter beads were successfully extracted from the water-contained bead-stock well by the vacuum tweezers, which have inner and outer diameters of 50 and 150 µm. An interesting aspect is that unexpected extra beads adsorbed on the outer wall of the vacuum tweezers can be removed using the surface tension force between the water and the atmosphere. In testing the total performance of this apparatus, the DNA probe arrays with 10 sets of probe-conjugated beads and 2 plain beads were produced in the intended order in the capillaries. The time needed to align the 12 beads was 10 min, and the 16 bead arrays were fabricated simultaneously. After hybridization experiments using these fabricated DNA probe arrays, fluorescence from each bead was clearly observed. DNA probe array has become one of the most important tools for understanding the functions and variations of multiplex genes.1-3 Among the tools, DNA microarrays, manufactured by spotting DNA probe solutions on slide glasses or by photolithographic oligonucleotide synthesis on glass substrates, are already in commercial use.3-9 However, some serious challenges still remain, such as the cost-effective production of the devices and a reduction in the reaction time. Bead-based probe arrays have been * Corresponding author: Phone: +81-423-23-1111. Fax: +81-423-27-7833. E-mail:
[email protected]. (1) Khan, J.; Bittner, M. L.; Chen, Y.; Meltzer, P. S.; Trent, J. M. Biochim. Biophys. Acta 1999, 1423, 17-28. (2) Diamandis, E. P. Clin. Chem. 2000, 46, 1523-1525. (3) Schena, M.; Shalon, D.; Davis, R. W.; Brown, P. O. Science 1995, 270, 467-470. (4) DeRisi, J.; Penland, L.; Brown, P. O.; Bittner, M. L.; Meltzer, P. S.; Ray, M.; Chen, Y.; Su, Y. A.; Trent, J. M. Nature Genet. 1996, 14, 457-460. (5) Okamoto, T.; Suzuki, T.; Yamamoto N. Nat. Biotechnol. 2000, 18, 438441. (6) Foder, S. P.; Read, J. L.; Pirrung, M. C.; Stryer, L.; Lu, A. T.; Solas, D. Science 1991, 251, 767-773. (7) Lockhart, D. J.; Dong, H.; Byrne, M. C.; Follettie, M. T.; Gallo, M. V.; Chee, M. S.; Mittmann, M.; Wang, C.; Kobayashi, M.; Horton, H.; Brown, E. L. Nat. Biotechnol. 1996, 14, 1675-1680.
3250 Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
developed as a technique to meet these challenges.10-14 The production costs decrease by distributing probe-conjugated beads among the devices or tubes because the expensive probes can be used effectively to immobilize many of the beads in a batch reaction. Furthermore, the hybridization reaction on beads tends to be faster than on the typical DNA microarray. Combing of colorcoded microbeads with a flow cytometer, as well as fixed microbeads mounted on the terminal wells of optical fibers, have been reported.10-12 We have recently developed a new miniaturized bead-based DNA probe array called “bead array”.14 This device consists of an arrangement of different probe-conjugated beads of 100-µm diameter in a predetermined order inside a 150-µm-diameter capillary. The ordinal number of the beads identifies which probes are on the beads. In our previous work,14 very efficient hybridization reactions of this device were observed. First, the time needed for the reaction to reach a plateau was within 10 min, independent of the target concentration, and second, 1-amol fluorescencelabeled target DNAs were selectively detected on the basis of their sequences. These attractive features of the bead array are superior, not only to that of the typical DNA microarrays but also to that of the previous bead-based probe arrays.3-12 Therefore, bead arrays might be used for powerful multityping devices particularly for diagnosis tests. However, fabrication by manually introducing the beads one-by-one into the capillary is labor-intensive and timeconsuming. To manufacture the bead array more efficiently, an apparatus for aligning beads in a predetermined order must be developed. (8) Fan, J. B.; Chen, X.; Halushka, M. K.; Berno, A.; Huang, X.; Ryder, T.; Lipshutz, R. J.; Lockhart, D. J.; Chakravarti, A. Genome Res. 2000, 10, 853860. (9) Sapolsky, R. J.; Hsie. L.; Berno, A.; Ghandour, G.; Mittmann, M.; Fan, J. B. Genet. Anal. 1999, 14, 187-192. (10) Fulton, R. J.; McDade, R. L.; Smith, P. L.; Kienker, L. J.; Kettman, J. R., Jr. Clin. Chem. 1997, 43, 1749-1756. (11) Walt, R. D. Science 2000, 287, 451-452. (12) Ferguson, J. A.; Steemers, F. J.; Walt, R. D. Anal. Chem. 2000, 72, 56185624. (13) Brenner, S.; Johnson, M.; Bridgham, J.; Golda, G.; Lloyd, D. H.; Johnson, D.; Luo, S.; McCurdy, S.; Foy, M.; Ewan, M.; Roth, R.; George, D.; Eletr, S.; Albrecht, G.; Vermaas, E.; Williams, S. R.; Moon, K.; Burcham, T.; Pallas, M.; DuBridge, R. B.; Kirchner, J.; Fearon, K.; Mao, J.; Corcoran, K. Nat. Biotechnol. 2000, 18, 630-634. (14) Kohara, Y.; Noda, H.; Okano, K.; Kambara, H. Nucleic Acids Res. 2002, 30, e87. 10.1021/ac020674n CCC: $25.00
© 2003 American Chemical Society Published on Web 05/22/2003
Figure 1. Schematic diagram of the bead alignment apparatus. The micro vacuum tweezers unit is moved to the position above the holes for bead-dropping (dotted lines).
In this article, we present an automated bead alignment apparatus for fabricating the bead array, and a new method for the single-bead capturing technique using micro vacuum tweezers. How to extract only single 100-µm-diameter beads from bead stock wells is described. We also demonstrate multiplex hybridization using DNA probe arrays fabricated by this apparatus. EXPERIMENTAL SECTION Configuration of Bead-Alignment Apparatus. Figure 1 shows a schematic diagram of the bead alignment apparatus. It consisted of four main parts: microplate stages, micro vacuum tweezers unit, bead detection unit, and bead arraying unit. Two 384-well microplates with DNA conjugated beads were able to set on the microplate stages. Probe-conjugated beads (768 sets) identified by the position of the microwells on the plate were used. This represents the maximum for bead array fabrication in one attempt. The micro vacuum tweezers unit was designed to hold 16 capillary tubes (GL Science, Japan) with an innner diameter (i.d.) of 50 µm. The capillaries were mounted onto the unit through inner seal connectors (GL Science, Japan) and aligned in the pitch of the wells of a 384-well microplate (well pitch, 4.5 mm). The inner seal connector can fix the capillaries of both 150-µm- and 375-µm-outer-diameters (o.d.). The 16 vacuum tweezers were operated by an aspirator (A-3S, Tokyo Rikakikai, Japan) and compressed air through 16 tubes (Tygon Tube, Norton, U.S.A.). The aspirator evacuated the inside of the vacuum tweezers which was used to capture a single bead. The compressed air produced high pressure in the inside of the vacuum tweezers. This was used to release the bead. By changing the port of the air valves connected between the aspirator and the compressed air, the capturing and releasing of beads on the vacuum tweezers were controlled. Three actuators controlled the micro vacuum tweezers unit. A vision sensor (CV-700, Keyence, Japan) was used to detect whether a single bead existed on the vacuum tweezers. A ringshaped light from white LEDs (CA-DRW3, Keyence, Japan) was
used to illuminate the bead. By setting a suitable brightness threshold value for the typical bead image, beads and background, including the vacuum tweezers, could be distinguished, and the existence of single beads on the vacuum tweezers was detectable. The bead arraying unit consisted of a head with 16 holes for released beads and a holder for setting 16 bead array capillaries. The holes on the head were aligned in the pitch of the microwells of a 384-well microplate. Sixteen bead array capillaries with i.d. of 150 and o.d. of 375 µm were set on the holder. By attaching the holder under the head with the bead-relased holes, all capillaries could link to the released-bead holes. To supply water to the holes, a syringe pump (sp-230iw, WPI, U.S.A.) was connected to the head. Furthermore, an aspirator was connected to the holder through a valve and was used to evacuate the inside of the bead array capillaries. Conjugation of DNA Probes on Glass Beads. DNA probe conjugation was performed on the 384-well microplate.14 Amino and maleimide groups were sequentially introduced into 103-µm glass beads (Matsunami Glass Ind. Ltd., Japan). Next, 5 mg of these maleimide-group-induced glass beads were distributed among the microwells of the 384-well microplates. The DNA probes were then conjugated on the bead surfaces of these plates. In other words, 5′-thiol-modified synthesized-oligo DNAs (Genset Japan, Japan) were put into the microwells and conjugated on the bead surfaces. After the beads in the microplate were dried in a vacuum, the bead-stocked 384-well microplates were set on the stage on the apparatus. Dionized water was introduced to the beadstocked wells before the beads were arrayed. Procedure of Bead Array Fabrication. The standard procedure of the bead alignment by the apparatus consisted of four steps. The first step was the single-bead capturing. The 16 capillary vacuum tweezers were moved above and in line with the 16 microwells which store the required beads. Then, the end of the capillary vacuum tweezers was immersed in the water and plunged into the beads at the bottom of the well. After being held at the bottom of the bead-stocked well for 2 s, the end of the tweezers Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
3251
Figure 2. Real-time images of micro vacuum tweezers capturing a single bead from among many beads (glass beads with an average diameter of 103 µm) in a microwell. Inner (i.d.) and outer diameters (o.d.) of the capillary vacuum tweezers were 50 and 150 µm, respectively.
was pulled out from the water with a single bead attached to it. Thus, 16 capillary vacuum tweezers picked up 16 single beads from the 16 microwells. The second was the detection of the single bead on the vacuum tweezers. The vision sensor scanned the end of the 16 vacuum tweezers one-by-one. The scanning velocity for our experiment was optimized to 0.9 s per bead. Thus, the total scanning time over 16 single beads on all vacuum tweezers’ was 14.4 s. In the case of a failure to capture or detect a single bead, the vacuum tweezers moved back to the previous position above the microwells and retried to capture the bead. The third step was to introduce the single beads into the bead array capillaries. At first, the valve, connected between the bead array capillary and the aspirator, was closed,and the holes were filled with 30 µL of water using the syringe pump. When the tweezers approached the released-bead holes, compressed air at 3 Kg/cm3 was supplied to the vacuum tweezers for 5s. The beads were released by the compressed air into the holes, and then the valve was opened. By the assistance of the water flow, the beads were introduced into the bead array capillaries. The final step was to check that all the tweezers released the single beads. As with the second step, the vision sensor scanned the ends of the 16 vacuum tweezers one-by-one. If the single beads were detected on the vacuum tweezers and were not released, the tweezers moved back to the released-bead holes and retried to release the bead. By repeating these four steps, the intended beads prepared on the microplate were introduced into the bead array capillaries, and bead arrays in a predetermined order were fabricated. RESULTS AND DISCUSSION Characteristics of a Single Bead Capturing Using Micro Vacuum Tweezers. Figure 2 shows typical real time images of the micro vacuum tweezers extracting a single bead from the beads in a well on a 384-well microplate. The i.d. and o.d. of the micro vacuum tweezers were 50 and 150 µm, respectively, and the glass beads had an average diameter of 103 µm. Interesting aspects of the process are that all adsorbed beads on the outer wall of the tweezers were removed and only a single 3252
Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
Figure 3. Theoretical schematic drawing of forces on a bead, caused by the water/atmosphere interface.
bead at the end of the tweezers remained. In water, as shown in Figure 2b and c, the many beads were adsorbed on the outer wall of the vacuum tweezers, as well as at the end of the vacuum tweezers, and moved upward together. Since this is also observed in the bead-capturing process without water, it is considered that static electricity caused the adsorption of beads onto the surface of the outer wall. At the water/atmosphere interface, however, we can see that the extra beads adsorbed on the outer wall were removed by a tension force (Figure 2d-f) and that only the single bead was successfully captured on the vacuum tweezers after passing through the interface (Figure 2f). Figure 3 shows a theoretical schematic drawing of the forces on a bead, originating at the water/atmosphere interface. The surface tension force of the liquid/gas interface acts on beads adsorbed on the wall when the bead passes through the interface (Figure 3). The surface tension force Fσ can be expressed as15
Fσ ) Pσlg cos θ1
(1)
where P is the perimeter length contacted between the bead and the water surface, σlg is the surface tension, and cos θ1 is the contact angle of the liquid/gas interface. As can be seen in eq 1 (15) Defay, R.; Prigogine, I.; Bellemans, A.; Everett, D. H. Surface Tens. Adsorpt. 1966, 1-20.
them.15 As shown in Figure 3, on the other hand, the friction force Ff drives between the bead and the wall by static electricity. The force Ff enables the beads to remain on the wall. Therefore, when the F is greater than the friction force Ff, as written by
F > Ff
Figure 4. Relationship between inner (i.d.) and outer diameters (o.d.) of micro vacuum tweezers and bead size for single bead capture. (a) Micro vacuum tweezers; i.d., 50 µm; o.d., 150 µm; bead size, 103 µm. (b) Micro vacuum tweezers; i.d., 50 µm; o.d., 150 µm; bead size, 75 µm. (c) Micro vacuum tweezers; i.d., 50 µm; o.d., 375 µm; bead size, 103 µm.
and Figure 3, a downward force, F, which is the cosine component of Fσ is given by
F ) Fσ cos θ2
(2)
acts on the beads absorbed on the outer wall, and tries to remove
(3)
extra beads absorbed on the wall can be removed by the downward force. This dependence of the outer diameter of the vacuum tweezers on the capturing performance is shown in Figure 4. The ends of the tweezers with capillaries with o.d. of 150 and i.d. of 50 µm were able to pick out a single bead with a diameter of 103 or 75 µm from the stocked beads (Figures 2, 4a,b). However, when the capillaries with an o.d. of 375 µm (i.d. 50 µm) were used, extra beads with a diameter of 103 µm remained absorbed not only at the end of the tweezers but also on the outer wall, even after the tweezers were pulled out of the water into the air (Figure 4c). These results indicate that to avoid bead adsorption on the outer wall, it is very important to choose the outer diameter of the tweezers according to the bead size. It can be explained that, as the curvature of the wall on the vacuum tweezers increases, the friction force, Ff, between the outer wall and the bead becomes strong. In the case of using the 375-o.d. vacuum tweezers,
Figure 5. (a) Sequences of 10 pairs of 18-mer complementary DNA derived from 10 exons of p53 gene, used as probes and targets for multiplex hybridization experiments. (b) Arrangement of 10 sets of probe-conjugated beads and two sets of plain glass beads without a probe on a 384-microwell plate. The same kind of DNA-conjugated beads were stocked in 16 microwells in each row.
Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
3253
Figure 6. Schematic diagram (a) and transparent microscope image (b) of bead array with 12 sets of beads, fabricated by bead alignment apparatus.
therefore, the downward force was not sufficient to remove the extra beads absorbed on the wall. It was concluded, from Figure 4, that the use of micro vacuum tweezers with an outer diameter of less than twice the bead diameter enables a single-bead to be picked up. The effect of the speed at which the 150-µm-o.d. vacuum tweezers were removed from the water on the single-bead-capture was also investigated. The speed was changed from 50 µm/sec to 3.2 mm/sec, but no significant effect on the single-beadcapturing phenomenon was observed. That is, the vacuum tweezers captured a single bead independently of the removal speed. It was also found that the performance of the single-bead capture did not depend on the material of the vacuum tweezers. For instance, stainless steel pipe with i.d. of 50 µm and o.d. of 100 µm (KS Sangyo, Japan) could also capture a single 103-µm bead. Fabrication of a Bead Array and Detection of Multiplex DNA Hybridization. Bead arrays consisting of 10 DNA-probeconjugated beads with a plain bead on each side in a capillary and in a determined order were produced. The plain beads were used to prevent the two stainless wires, introduced on both sides of the bead array, from damaging the surface of the DNAconjugated beads. Ten pairs of 18-mer complementary DNAs derived from 10 exons of the p53 gene were used as probes and targets in the multiplex hybridization experiments. The sequences of the 10 pairs of probes and targets are shown in Figure 5a. We arranged the 10 sets of DNA-probe-conjugated 103-µm beads and two sets of 130-µm glass beads without probe on the 384-microwell plate, as shown in Figure 5b. The same kind of DNA-conjugated beads were stocked in each row, consisting of 16 microwells, making 12 rows in total. This corresponds to the number of DNA-probe-conjugated- and plain-bead sets. Each of the 16 micro vacuum tweezers extracted 16 single beads from the 16 microwells, one-by-one from the first to the 12th rows of the plate, and thus, 16 bead arrays with 12 beads in each capillary in a determined order were produced automatically. Figure 6a shows a schematic diagram of a bead array with 12 beads aligned in this process. The transparent microscope image shown in Figure 6b indicates that the 12 beads were successfully aligned in the capillary. The time needed to array 12 sets of beads 3254 Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
Figure 7. Transparent microscopic image (a) and fluorescent microscope images (b-k) of the fabricated bead arrays after hybridization reactions. All sets of target DNAs were labeled with Texas Red. Only the beads with probes complementary to the target were observed in the exact position on the fluorescence image, and other beads with noncomplementary probes showed no fluorescence.
into 16 bead arrays was 10 min on average. In other words, the time for manipulating 16 beads using 16 vacuum tweezers was less than 1 min on average. Ten sets of bead arrays were used in the multiplex DNA hybridization experiments. All sets of target DNAs (Figure 5a) were labeled with Texas Red. Only one target was introduced to each bead array. Hybridization was performed at 45 °C, and the reaction time was 10 min. The details of the hybridization procedure were the same as in our previous work.14 Figure 7 shows a transparent microscopic image (Figure 7a) and fluorescent microscope images (Figure 7b-k) of the fabricated bead arrays after hybridization. As shown in Figure 7b-k, the beads with the probes complementary to the target were observed in the exact position in the fluorescence image, and the other beads with noncomplementary probes showed no fluorescence. The fluorescence intensity from hybridization-reacted beads was ∼1 × 103 times larger than that from the other beads. The discrimination ratio between complementary and noncomplementary hybridization was almost the same as that for a bead array fabricated manually.14 This indicates that the procedure did not damage the probe on the bead surface; therefore, it did not hinder the efficient hybridization reaction. It is thus considered that the bead alignment apparatus based on single-bead capturing
by the micro vacuum tweezers can significantly improve bead array fabrication. CONCLUSIONS AND SUMMARY An automated bead alignment apparatus based on a technique for single-bead capture by micro vacuum tweezers was developed. The vacuum tweezers were used to extract only a single 100-µm diameter bead in bead-stock wells and to manipulate it into the probe array’s capillaries. The idea is very simple, to capture single beads on the end of an evacuated capillary, but no similar technique existed. The biggest problem was the absorption effect of extra beads in the capillary. The conditions for avoiding the absorption of extra beads to the vacuum tweezers were determined, and the effect of the diameter of the vacuum tweezers on the capture performance was evaluated. The apparatus could manipulate a single bead without damaging the probe on the beads. It is, thus, suitable for inexpensive multiplex gene analysis on the miniaturized bead-based DNA probe array, that is, beadarray. Furthermore, the apparatus uses many other bead-based
applications, for example, distributing beads with probes to microwells or as a microdevice for controlling the number of beads. It is believed that soon, single-bead capture by capillary vacuum tweezers will be widely applied in bead-based DNA and immunoassay chip fabrications. ACKNOWLEDGMENT This work was performed as part of the research and development project of the Industrial, Science and Technology Program supported by the New Energy and Industrial Technology Development Organization, Japan. The authors thank Hayami Toba and Shinichiro Watase at Eastern Japan Semiconductors, Inc., for their help in constructing the apparatus.
Received for review November 4, 2002. Accepted April 5, 2003. AC020674N
Analytical Chemistry, Vol. 75, No. 13, July 1, 2003
3255
