Functional Microfabricated Sample Targets for Matrix-Assisted Laser

Quantitation of Ribonucleic Acids UsingO Labeling and Mass Spectrometry. Zhaojing Meng and Patrick A. Limbach. Analytical Chemistry 2005 77 (6), 1891-...
0 downloads 0 Views 170KB Size
Download PDF
Anal. Chem. 2003, 75, 1997-2003

Functional Microfabricated Sample Targets for Matrix-Assisted Laser Desorption/Ionization Mass Spectrometry Analysis of Ribonucleic Acids Beniam T. Berhane and Patrick A. Limbach*

Rieveschl Laboratories for Mass Spectrometry, Department of Chemistry, University of Cincinnati, P.O. Box 210172, Cincinnati, Ohio 45221

Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) is a powerful analytical tool for the structural characterization of proteins and nucleic acids. However, many proteomics or genomics methodologies that employ MALDI-MS require external sample manipulation, which limits the overall throughput of analysis. We have focused on fabricating functional MALDI sample plates that would permit the on-probe characterization of nucleic acids. Here, we present results arising from the fabrication of functional sample plates composed of poly(methyl methacrylate) (PMMA). The PMMA sample plates were fabricated by a CNC milling technique. The key structural feature of our microfabricated samples plates is the presence of individual cylindrical posts (360 µm × 360 µm), which serve as individual sample targets within the overall PMMA-based MALDI sample plate. Functionality is added to these microposts via the covalent attachment of enzymes. As an example of the applicability of these microfabricated sample plates, enzymatic digestion of ribonucleic acids was performed on probe (i.e., on the micropost) with subsequent analysis by MALDI-MS. Advantages to such an approach include a reduction in sample handling (and concomitant sample losses) and a reduction in the amount of sample required for analysis due to the small surface area of the microposts. The development of microfabricated analytical devices and components offers the benefits of increased throughput, reduced sample and reagent consumption, lower operating cost, singleuse disposable devices, and smaller sizes for portability.1,2 The typicalapplicationsofmicrofabricateddevicesincludeelectrophoresis,3-5 electrochromatography,6-8 enzyme/immunoassays,9-12 and as an * To whom correspondence should be addressed. Phone: (513) 556-1871. Fax: (513) 556-9239. E-mail: [email protected]. (1) van den Berg, A.; Lammerink, T. S. J. Top. Curr. Chem. 1998, 194, 21-49. (2) Bruin, G. J. M. Electrophoresis 2000, 21, 3931-3951. (3) Gao, J.; Xu, J. D.; Locascio, L. E.; Lee, C. S. Anal. Chem. 2001, 73, 26482655. (4) Lazar, I. M.; Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 2001, 73, 17331739. (5) Wang, J.; Chatrathi, M. P.; Mulchandani, A.; Chen, W. Anal. Chem. 2001, 73, 1804-1808. (6) Ericson, C.; Liao, J. L.; Nakazato, K.; Hjerten, S. J. Chromatogr., A 1997, 767, 33-41. (7) Ericson, C.; Holm, J.; Ericson, T.; Hjerten, S. Anal. Chem. 2000, 72, 8187. 10.1021/ac020710i CCC: $25.00 Published on Web 04/01/2003

© 2003 American Chemical Society

integrated part in mass spectrometry.13,14 Silica-based substrates have been commonly used for producing microfabricated systems due to their advantageous features such as high mechanical strength and reproducibility of structures.15 More recently, polymeric materials have been used in microfabrication due to their potential uses as disposable, low-cost devices.16-18 Mass spectrometry (MS) has emerged as one of the most widely used analytical techniques for the analysis of biological materials.19 The development of the “soft ionization” methods, electrospray ionization (ESI)20 and matrix-assisted laser desorption/ionization (MALDI),21 has increased the applicability of mass spectrometry to biological materials. Microfabricated devices have been interfaced with mass spectrometry for the analysis of biological materials.13,14,22,23 Most applications have utilized ESIMS due to the ease of coupling microfluidics with this ionization source. However, several applications with MALDI-MS have been reported.24 Marko-Varga et al. have developed a “microfabricated toolbox” out of silicon substrate that has a reactor, microdispenser for sample spotting, and nanovial MALDI target.25-30 Those researchers have recently used polymer substrates as nanovial (8) Ceriotti, L.; de Rooij, N. F.; Verpoorte, E. Anal. Chem. 2002, 74, 639-647. (9) Mao, H.; Yang, T.; Cremer, P. T. Anal. Chem. 2002, 74, 379-385. (10) Wang, J.; Chatrathi, M. P.; Mulchandani, A.; Chen, W. Electrophoresis 2002, 23, 713-718. (11) Dodge, A.; Fluri, K.; Verpoorte, E.; de Rooij, N. F. Anal. Chem. 2001, 73, 3400-3409. (12) Sato, K.; Tokeshi, M.; Kimura, H.; Kitamori, T. Anal. Chem. 2001, 73, 12131218. (13) Limbach, P. A.; Meng, Z. Analyst 2002, 127, 693-700. (14) Figeys, D.; Pinto, D. Electrophoresis 2001, 22, 208-216. (15) Kop, M. U.; de Mello, A. J.; Manz, A. Science 1998, 280, 1046-1048. (16) McCormick, R. M.; Nelson, R. J.; Alonso-Amigo, M. G.; Benvegnu, D. J.; Hooper, H. H. Anal. Chem. 1997, 69, 2626-2630. (17) Becker, H.; Gartner, C. Electrophoresis 2000, 21, 12-26. (18) Soper, S. A.; Ford, S. M.; Qi, S.; McCarley, R. L.; Kelly, K.; Murphy, M. C. Anal. Chem. 2000, 72, 642A-651A. (19) Yates, J. R. J. Mass Spectrom. 1998, 33, 1-19. (20) Yamashita, M.; Fenn, J. B. J. Phys. Chem. 1984, 88, 4451. (21) Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. Int. J. Mass Spectrom. Ion Processes 1987, 78, 53. (22) Oleschuk, R. D.; Harrison, D. J. Trends Anal. Chem. 2000, 19, 379-388. (23) de Mello, A. J. Lab Chip 2001, 1, 7N-12N. (24) Foret, F.; Preisler, J. Proteomics 2002, 2, 360-372. (25) Ekstrom, S.; Onnerfjord, P.; Nilsson, J.; Bengtsson, M.; Laurell, T.; MarkoVarga, G. Anal. Chem. 2000, 72, 286-293. (26) Laurell, T.; Nilsson, J.; Marko-Varga, G. Trends Anal. Chem. 2001a, 20, 225-231. (27) Laurell, T.; Nilsson, J.; Marko-Varga, G. J. Chromatogr., B 2001, 752, 217232.

Analytical Chemistry, Vol. 75, No. 9, May 1, 2003 1997

MALDI targets31,32 on which tryptic products are deposited for MALDI-MS analysis. A strong motivation for coupling microfabricated devices with mass spectrometry is the advantage of performing sample pretreatment at the microscale. Sample pretreatment at this scale reduces the amount of sample required and offers the potential for reduced overall analysis time.33 Performing sample pretreatment on microfabricated devices requires adding functionality to these devices. To date, the most popular approach has been to immobilize enzymes of interest onto surfaces (e.g., beads, chromatographic supports) prior to use with a microfabricated device34 or MALDI sample target.25,35 Although much of the previous integrations of microfabricated devices with MALDI-MS have focused on the off-line processing of samples, Reinhoudt et al.36 recently demonstrated the on-line MALDI-MS analysis of various (bio)chemical processes occurring within a microfluidic device. Unique to that work was the development of a microreactor system that allows for the (bio)chemical processing to occur while the MALDI sample plate is contained in the vacuum chamber of the instrument. Reagents are mixed due to the vacuum-induced flow, and the analytes were monitored directly by MALDI-MS. In contrast to the previously mentioned approaches, here we report the development of a functional microfabricated MALDI sample target that permits the on-probe analysis of nucleic acids without requiring microfluidics. The MALDI sample target was fabricated from poly(methyl methacrylate) (PMMA) by a CNC machining process. After fabrication, RNases were covalently immobilized onto the surface of the MALDI target and utilized for the structural analysis of ribonucleic acids. Functional microfabricated MALDI targets reduce sample requirements, add flexibility to the experimental methodology, and can be utilized in a general scheme for the structural characterization of nucleic acids. MATERIALS AND METHODS Materials. PMMA bead (6.5-µm diameter) were purchased from Bangs Laboratories (Fishers, IN) and 0.5-mm-thick PMMA sheets were purchased from GoodFellow (Berwyn, PA). N-Butyllithium, 1,3-diaminopropane, glutaraldehyde, diammonium hydrogen citrate (DAHC), ethylenediamine, magnesium chloride, sodium chloride, 2,4,6-trihydroxyacetophenone (THAP), and Escherichia coli tRNAVal were obtained from Sigma-Aldrich (St. Louis, MO). Tris-HCl was purchased from Gibco BRL (Grand Island, (28) Laurell, T.; Wallman, L.; Nilsson, J. J. Micromech. Microeng. 1999, 9, 369376. (29) Onnerfjord, P.; Nilsson, J.; Wallman, L.; Laurell, T.; Marko-Varga, G. Anal. Chem. 1998, 70, 4755-4760. (30) Miliotis, T.; Kjellstrom, S.; Onnerfjord, P.; Nilsson, J.; Laurell, T.; L-E.;, E.; Marko-Varga, G. J. Chromatogr., A 2000, 886, 99-110. (31) Marko-Varga, G.; Ekstrom, S.; Helldin, G.; Nilsson, J.; Laurell, T. Electrophoresis 2001, 22, 3978-3983. (32) Ekstrom, S.; Nilsson, J.; Helldin, G.; Laurell, T.; Marko-Varga, G. Electrophoresis 2001, 22, 3984-3992. (33) Nelson, R. W.; Nedelkov, D.; Tubbs, K. A. Electrophoresis 2000, 21, 11551163. (34) Wang, C.; Oleschuk, R.; Ouchen, F.; Li, J. J.; Thibault, P.; Harrison, D. J. Rapid Commun. Mass Spectrom. 2000, 14, 1377-1383. (35) Brivio, M.; Fokkens, R. H.; Verboom, W.; Tas, N. R.; Goedbloed, M.; van den Berg, A.; Reinhoudt, D. N. Anal. Chem. 2002, 2002, 3972-3976. (36) Peterson, D. S.; Rohr, T.; Svec, F.; Frechet, J. M. J. Anal. Chem. 2002, 74, 4081-4088.

1998

Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

NY). Sodium phosphate, Borax, dimethyl sulfoxide, and disodium ethylendiaminetetracetic acid were obtained from Fisher Scientific (Hampton, NH). Fluorescien-5-isothiocyanate (FITC), 5-(aminoacetamido)fluorescien and 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride were obtained from Molecular Probes (Eugene, OR). RNase A was purchased from ICN Biomedical (Costa Mesa, CA), RNase T1 was purchased from Roche Biomolecular Chemical (Indianapolis, IN), and RNase U2 was purchased from Gracefield Research Center (Glycobiology and Enzyme Technology Industrial Research Ltd., Lower Hutt, New Zealand). All phosphoroamidites and reagents used for synthesizing oligonucleotides on an Expedite 8909 nucleic acid synthesizer (Applied Biosystems, Foster City, CA) via standard phosphoramidite chemistry were obtained from Glen Research (Sterling, VA). Oligonucleotide purification cartridges were purchased from Waters (Plymouth, MN), and C18 ZipTips were obtained from Millipore (Bedford, MA). Nanopure water (18 MΩ) was filtered using a Barnstead Nanopure System (Dubuque, IA) followed by sterilization to remove adventitious RNases. Methods. Microfabrication of MALDI Target. Three PMMA sheets were bonded together by isopropyl alcohol treatment and heating to 80 °C. Microposts (360 µm diameter × 360 µm height) were milled out of the surface using an RBI bed-type vertical spindle milling machine (Taichung Hsien, Taiwan) with Trak CNC2 controller (Southwest Industries, Gardena, CA) and a drill bead cutter from Antares Instruments (Horsham, PA). The Trak CNC2 controller was manipulated to work in a closed loop so that the tool offset will be less than that of the radius of the drill bead cutter, which was 0.030 in. A polymer mask was also machined to allow for enzyme immobilization to occur only on the top surface of the microposts. This mask was made from 500-µm-thick PMMA. Holes (365-µm diameter) were drilled into the mask to allow its placement over the microfabricated target during the enzyme immobilization process. To monitor and confirm the integrity of the machining process, optical images of the MALDI target were taken using a Nikon Objective optical camera from Optotek (Cottonwood, CA). In addition, SEM images of the microposts were obtained using an XL30 E-SEM (FEI Co., Peabody, MA). The total time to fabricate the PMMA sample target is ∼4 h. Enzyme Immobilization. RNase immobilization was first performed on PMMA beads.37 After verifying that the RNase activity was intact when immobilized on the PMMA beads, the RNase was then immobilized on the microposts of the MALDI target fabricated from PMMA. The immobilization of the RNase on the PMMA beads and the MALDI target was performed by following a procedure similar to that described previously.38 Briefly, three steps are necessary to covalently attach the RNase to the PMMA surface. The first step is a derivatization of PMMA with a suitable diamine. The second, a functionalization step, converts the amine to an aldehyde via reaction with glutaraldehyde. The final step is the reaction of the aldehyde with free amines from the enzyme to yield an immobilized enzyme. Confocal Microscopy. Previously, Henry et al. demonstrated the covalent immobilization of a restriction enzyme to PMMA.38 Thus, in this work, the immobilization of RNases to the surface of the (37) Dominick, W. D.; Berhane, B. T.; Mccomber, J. S.; Limbach, P. A. Anal. Bioanal. Chem., in press. (38) Henry, A. C.; Tutt, T. J.; Galloway, M.; Davidson, Y. Y.; McWhorter, C. S.; Soper, S. A.; McCarley, R. L. Anal. Chem. 2000, 72, 5331-5337.

MALDI target was monitored using confocal microscopy. Two fluorescent dyes were attached to available amine or carboxylic acid groups separately. Fluorescence was then monitored using a Zeiss LSM 510 laser scanning confocal microscope (Zeiss Instruments, Thornwood, NY). The dyes used were the aminereactive probe FITC and the carboxylic acid reactive probe 5-(aminoacetamido)fluorescein. In the case of the amine-reactive probe, 1 µM FITC was prepared in a borate buffer, adjusted to pH 9.2, containing 1% isopropyl alcohol. In the case of the carboxylic-reactive probe, 0.5 mM 5-(aminoacetamido)fluorescein was made in a solution containing 14 mM 1-ethyl-3-(3-dimethylaminopropyl)carbodiimide hydrochloride, and the pH was adjusted to 9.0. Microfabricated MALDI targets were kept overnight in the dye solutions and then washed three times with Nanopure water prior to imaging. RNase Digestion. In-solution digestions of tRNAVal were performed as follows. (1) RNase A: 100 µg of tRNAVal was digested by incubating for 20 min at 37 °C with 0.2 µg of RNase A in 10 µL of 30 mM Tris-HCl buffer, pH 7.8, containing 70% (v/v) DMSO. (2) RNase T1: 100 µg of tRNAVal was digested by incubating for 30 min at 37 °C with 10 000 units of RNase T1 in 50 µL of 50 mM of Tris-HCl, pH 7, containing 1 mM EDTA. (3) RNase U2: 100 µg of tRNAVal was digested by incubating for 15 min at 50 °C with 50 units of RNase U2 in 10 µL of 20 mM sodium citrate buffer, pH 5, containing 1 mM EDTA. Digestion of tRNAVal using RNases immobilized to PMMA beads was performed as described above for the in-solution digestions with the exception that the total amount of beadimmobilized RNase was unknown. In the case of on-target digestions, the tRNAVal was dissolved in the digestion buffer first and then deposited on the surface of the MALDI target on which the RNases were immobilized. Then, the MALDI target with immobilized RNase and the deposited tRNAVal was incubated for the same amount of time as in the case of in-solution digestions. To limit the addition of sample to the top of the microposts, less than 1 µL of solution was applied by repeated pipetting. MALDI Analysis. A Bruker Reflex IV MALDI time-of-flight (TOF) instrument with a 3-m effective flight path and a two-stage gridless ion reflector equipped with pulsed ion extraction was used in reflectron mode (Bruker Daltonics, Billerica, MA). This MALDI instrument is equipped with a nitrogen laser (λ ) 337 nm), collision-induced dissociation capability, and fragmentation analysis and structural TOF MS capability. Flex control software and Xmass software were used for data acquisition and processing, respectively. The in-solution and on-bead immobilized digestion products were purified using ZipTip C18 micropipet tips following the protocol for oligonucleotide purification provided by the manufacturer. After desalting, 1 µL of 300 mM THAP and 1 µL of 250 mM DAHC were added to 2 µL of the digestion products and vortexed. Then, 0.5 µL of the mixture was spotted on top of a layer of previously spotted THAP. After the mixture was dried in a custom-built vacuum drier, MALDI analysis was performed. For on-target digestion, 0.5 µL of the THAP/DHAC matrix solution was added directly to the micropost(s) of interest. After drying, MALDI analysis was performed as described above.

Figure 1. (a) MALDI mass spectrum arising from the analysis of tRNAVal on-bead endonuclease digestion products. RNase U2 was immobilized to PMMA beads. The digestion products were cleaned with ZipTips and the matrix/comatrix solution was added to the digestion products before depositing on a conventional MALDI sample target. The peaks identified by asterisks, *, are those corresponding to expected endonuclease digestion products. (b) Same as in (a) excepting that the RNase U2 was not immobilized on PMMA beads.

RESULTS AND DISCUSSION Development of a functional, microfabricated MALDI sample target requires first confirming that the covalent immobilization of the enzyme(s) of interest does not result in a loss of enzymatic activity.37 The first enzyme immobilizations were performed on PMMA beads. As the surface of the PMMA bead is presumed to be identical to the surface of a PMMA sheet, the successful digestion of a nucleic acid by an RNase immobilized to a PMMA bead would serve as evidence that enzyme activity is not lost upon the immobilization process. Figure 1a is the mass spectrum resulting from the digestion of tRNAVal using RNase U2 that had been first immobilized onto PMMA beads. For comparison, the in-solution digest of tRNAVal with this same RNase is shown in Figure 1b. As can be seen by a visual inspection of the mass spectral data, a number of lower molecular weight ions are detected in both cases. The peaks identified by asterisks, *, are those corresponding to expected endonuclease digestion products. While the quality of the mass spectral data obtained from the in-solution digestion was higher than that found for the bead-based digestion, it is clear from these results that immobilization of this particular RNase at available amine groups does not lead to the complete loss of enzymatic activity. A direct quantitative comparison of enzyme activity between these two cases cannot be made as the amount of enzyme immobilized onto the PMMA beads is unknown. However, the Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

1999

Figure 2. MALDI mass spectrum arising from the analysis of the synthetic oligonucleotides dT5 and dT15. The oligonucleotides were mixed with the matrix/comatrix solution before depositing on a MALDI sample target made from a sheet of PMMA. No interferences from the use of the PMMA as the sample target are noted.

starting amount of enzyme used for the bead immobilization process was the same as that used for the in-solution digests. Once we had confirmed that enzyme immobilization does not lead to the complete loss of enzyme activity, the next step was to determine whether the PMMA substrate would be compatible with MALDI-based analysis of oligonucleotides. One concern was whether the polymer would lead to a high background signal within the mass range of interest for these analyses. Another concern was whether the covalently immobilized enzyme (or any nonspecifically adsorbed enzyme) would interfere with the MALDI analysis. To examine these issues, MALDI sample targets made of PMMA sheets were fabricated and analyzed directly by laser desorption/ionization and MALDI under instrumental conditions similar to those expected for conventional MALDI experiments of oligonucleotides. Pristine PMMA, PMMA that had been derivatized with a diamine, PMMA that had been derivatized with a diamine and then functionalized with glutaraldehyde, and PMMA that had RNases covalently immobilized onto the surface were all examined. In no case was any background signal attributable to the polymer or the surface modifications detected under laser powers typically utilized for MALDI experiments (data not shown). We did notice that, at higher laser powers (>4× of normal operating power), background signal attributed to PMMA could be detected. However, as those conditions are not typically encountered during the MALDI-MS analysis of oligonucleotides, the use of PMMA as a MALDI target substrate did not appear to pose problems for typical analyses. Further, the lack of signal arising from the enzymes of interest in this study provided evidence that the nonspecific adsorption of those enzymes did not occur on the modified PMMA surface. After confirming that the PMMA-based MALDI target would not generate background ions that would interfere with analysis of the analytes of interest in this study, the next investigations focused on determining whether the PMMA-based MALDI target was compatible with a standard MALDI analysis of oligonucleotides. Figure 2 is the mass spectrum resulting from the analysis of a 5-mer and a 15-mer of poly(thymidilic acid) (dT5 and dT15) spotted in matrix directly onto a PMMA-based MALDI target. As is evident in this figure, both oligonucleotides are readily detected along with the usual matrix-related ions at lower mass. No interference from the PMMA substrate is noted, and the analytical 2000 Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

Figure 3. (a) Optical image of a typical microfabricated MALDI sample target. The sample target was generated using PMMA as the polymer substrate. Free-standing microposts are equally spaced near the perimeter of the sample target. (b) SEM image of a single micropost on the microfabricated MALDI sample target. Micropost dimensions are 360 µm × 360 µm. All surface modifications leading to enzyme immobilization occur on the top surface of each micropost.

performance of the PMMA-based target was similar to that found with conventional stainless steel targets. Thus, the use of the PMMA-based target is compatible with the MALDI-MS analysis of oligonucleotides and presumably other typical MALDI analytes. The final step in developing functional microfabricated MALDI sample targets was the fabrication of the PMMA-based sample targets and the immobilization of the enzymes of interest onto these targets. As mentioned by Ekstrom et al., the microfabrication of MALDI sample targets allows for the partitioning of minimal sample volumes on the target surface. However, rather than fabricate nanovials, which could lead to interferences from the spatial distribution of the matrix-analyte crystals,25 we chose to fabricate microstructures in the form of cylindrical posts. Additionally, microposts were fabricated to provide a planar surface suitable for subsequent surface modification. Optical and SEM images of a typical microfabricated MALDI sample target are seen in Figure 3. Although the goal of this initial work was not to generate the densest possible array of posts, it is clear from these images that free-standing microposts can be fabricated on the surface of the PMMA sheet. Further, although in the current work these posts were fabricated by a milling technique, a master metal mold insert could just as easily be fabricated and used for the hot embossing or injection molding generation of multiple MALDI sample targets in a rapid, costeffective manner. To selectively derivatize the top surface(s) of the microfabricated MALDI sample target, a polymer mask was also fabricated to limit any surface modifications to the top surfaces of the microposts on the MALDI sample target. As a means of monitor-

Figure 4. Confocal fluorescence microscopy images of a micropost top surface after incubation with an amine-reactive fluorescent dye at various stages in the immobilization process. The predicted surface functional groups at each step38 are shown below their respective images. (a) Pristine PMMA surface. (b) Surface after derivatization with diamine. (c) Surface after functionalization with glutaraldehyde. (d) Surface after immobilization of RNase.

ing the effectiveness of the various steps necessary for the ultimate immobilization of an RNase to the micropost, fluorescent probes specific for amine or carboxylic acid functional groups were

utilized at each step in the immobilization process. By incubating the microfabricated MALDI sample target with the appropriate fluorescent probe, changes in the surface chemistry of the Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

2001

micropost could be monitored using confocal microscopy. Based upon the known surface chemistries expected to result from this process, fluorescence from the amine-based dye should be seen after the derivatization and immobilization steps. Images of a representative micropost sample surface at various stages in the immobilization process are seen in Figure 4. The images in Figure 4 arise from the use of an amine-reactive dye. As predicted, fluorescence is only detected after the derivatization and immobilization steps, confirming that the PMMA surface now contains amine groups. Importantly, no fluorescence is found after the functionalization step, confirming that the derivatized (i.e., amine) surface has been converted to the aldehyde-based surface that allows for the ultimate immobilization of the RNase through available free amines (such as the N-terminus of the protein). In addition, microscopy performed in the focal plane of the MALDI sample probe surface did not yield any fluorescence signals, confirming that the immobilization process was limited to the planar surface of the microposts. The reduction in fluorescence seen between the amineterminated surface (Figure 4b) and the enzyme-immobilized surface (Figure 4d) is presumed to occur due to the reduction in the number of free amines available for conjugation with the fluorescent dye. In support of this rationale, examination of the enzyme-immobilized surface with a carboxylic acid-reactive dye was also done. The C-terminus of the enzyme should be available for conjugation with the dye (unlike the N-terminus, which can serve as the site of attachment to the modified surface). As expected, increased fluorescence was found using the carboxylic acid-reactive dye (data not shown), suggesting the planar surface of the micropost contains primarily immobilized RNase. To confirm that a functional microfabricated MALDI sample target had been generated, various RNases were immobilized onto the sample target microposts and then used for the on-probe digestion of nucleic acids. A representative example of the mass spectra resulting from the analysis of on-probe digested nucleic acids is shown in Figure 5. Figure 5a is the mass spectrum resulting from the on-probe digestion of tRNAVal with RNase A, Figure 5b is the mass spectrum resulting upon digestion with RNase T1, and Figure 5c is the mass spectrum resulting upon digestion with RNase U2. For each of the examples shown here, 10 pmol of tRNA was spotted onto the micropost. In all cases, enzymatic digestion of the intact tRNA occurred yielding m/z values that correspond to predicted endonuclease digestion products. As with the in-solution and bead-based digestions, no evidence was found that intact tRNA remained after digestion. The functional microfabricated sample plates could be reused after washing in phosphate buffer. In addition, after use and washing, these samples plates can be stored in the freezer (0 °C) and reused for subsequent characterizations of nucleic acids. Taken together, these data illustrate that the covalent immobilization of RNases to the microfabricated sample plate provides a robust and stable platform for the structural analysis of nucleic acids. Advantages of the developed approach for RNA endonuclease mapping include the reduced sample requirements, the reduced sample preparation time, and the added functionality arising from having separate microreactors (i.e., microposts) that could allow for the parallel processing of multiple analytes by a single 2002 Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

Figure 5. MALDI mass spectra resulting from the on-probe digestion of tRNAVal by immobilized (a) RNase A, (b) RNase T1, and (c) RNase U2. The peaks identified by asterisks, *, are those corresponding to expected endonuclease digestion products. In all cases, enzymatic activity is retained upon immobilization and structurally informative endonuclease digestion products are detected. The percent sequence coverage for each endonuclease digestion is also presented.

endonuclease or, alternatively, could allow for the parallel processing of a single analyte by multiple endonucleases. As an example of the latter, the sequence coverage obtained by the three endonucleases investigated here are reported in Figure 5. While no single endonuclease resulted in complete sequence coverage, the combination of all three enzymes allows for 97% of the total tRNA sequence to be characterized. The noticeable difference between the mass spectral results obtained from an in-solution versus an on-probe digestion of a nucleic acid are the lower signal-to-noise level and presence of salt adducts in the on-probe case. As demonstrated earlier, the lower signal-to-noise level cannot be attributed to an increase in the background or noise level due to either the polymer substrate or the presence of surface modifications. Most likely, this reduction in S/N is due to a reduced analyte signal. Whether the signal has decreased due to smaller amounts of enzyme-generated

oligonucleotides or to sample contamination has not been established conclusively. However, the increased level of salt adducts in the case of on-probe digestion suggests that sample contamination is the most likely reason for the reduced S/N seen in these cases. The increased level of salt adducts likely arises from the inability to effectively desalt the analyte prior to analysissfor the in-solution analyses, samples were desalted using C18-packed micropipet tips prior to spotting. Modifications to the on-probe enzymatic digestion conditions are underway to reduce sample contamination concerns. CONCLUSIONS Here we have demonstrated that functional microfabricated MALDI sample targets can be produced and utilized for the structural analysis of nucleic acids. Immobilization of RNases to a polymer support does not lead to the complete loss of enzymatic activity of the nuclease. The use of PMMA as the MALDI substrate does not interfere with the analysis nor does the presence of surface modifications necessary for the on-probe digestion step. Although the MALDI sample target in this work was fabricated by a CNC milling approach, a metal mold insert could also be used for the rapid and economical hot embossing of multiple sample targets from the PMMA substrate. The additional of functional enzymes directly to the MALDI sample target will allow for a reduction in sample consumption and

reduces the analysis time. An additional benefit of such devices that is currently being explored in our laboratory is the integration of the on-probe digestion/MALDI-MS analysis for endonuclease mapping with subsequent postsource decay sequencing of oligonucleotides to locate modified nucleosides within the overall sequence of the RNA. Although demonstrated with RNases and nucleic acids, these functional microfabricated MALDI sample targets could be used for potentially any applications requiring the enzymatic processing of biological compounds provided the immobilization step does not lead to the complete reduction of enzymatic activity. Additional work in our laboratory demonstrating such approaches for proteomics applications is currently in progress. ACKNOWLEDGMENT Financial support of this work was provided by the National Institutes of Health (GM58843) and the University of Cincinnati. The authors also thank J. S. Mecomber, D. Hurd, S. Conklin, and C. Bennett-Stamper for their assistance at various stages of this project.

Received for review November 15, 2002. Accepted February 20, 2003. AC020710I

Analytical Chemistry, Vol. 75, No. 9, May 1, 2003

2003