Ionization

Anal. Chem. 2004, 76, 1078-1082

Direct from Polyacrylamide Gel Infrared Laser Desorption/Ionization Yichuan Xu, Mark W. Little, David J. Rousell, Jorge L. Laboy,† and Kermit K. Murray*

Department of Chemistry, Louisiana State University, Baton Rouge, Louisiana 70803

The direct combination of gel electrophoresis and infrared laser desorption/ionization time-of-flight mass spectrometry has been demonstrated. We present results for infrared laser desorption and ionization mass spectrometry of peptides and proteins directly from a polyacrylamide gel without the addition of a matrix. Analyte molecules up to 6 kDa were ionized directly from a vacuum-dried sodium dodecyl sulfate-polyacrylamide gel after electrophoretic separation. Mass spectra were obtained at the wavelength of 2.94 µm, which is consistent with IR absorption by N-H and O-H stretch vibrations of water and other constituents of the gel. A 5-nmol quantity of peptide or protein was loaded per gel slot, although it was possible to obtain mass spectra from a small fraction of the gel spot. This technique shows promise for the direct identification of both parent and fragment masses of proteins contained in polyacrylamide gels. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) is a widely used technique for the separation of proteins in molecular biology, biochemistry, and medicine.1-7 A protein’s molecular mass is determined by an electrophoretic separation together with several marker proteins of known molecular masses that bracket the mass of the protein of interest. Mass determination of proteins by SDS-PAGE suffers from one main limitation: the mass determination accuracy is typically in the range of 5-10%, which is insufficient for peptide and protein identification in many cases.8 The poor mass resolution of gel separations is due to large variability in the protein quaternary and tertiary structure, posttranslational modifications, binding of proteins with the gel, and partial unfolding of the protein in the presence of detergents.9 * Corresponding author: (e-mail) [email protected]; (phone) 225-578-3417; (fax) 225-578-3458. † Current address: Department of Chemistry, University of Puerto Rico, Mayaguez, PR 00681-5000. (1) Laemmli, U. K. Nature 1970, 227, 680-685. (2) Neville, H.; Glossmann, H. Methods Enzymol. 1974, 32, 92-102. (3) O’Farrell, P. H. J. Biol. Chem. 1975, 250, 4007-4021. (4) Scha¨gger, H.; von Jalow, G. Anal. Biochem. 1978, 166, 368-379. (5) Wilkens, M. R.; Sanchez, J. C.; Williams, K. L.; Hochstrasser, D. Electrophoresis 1996, 17, 830-838. (6) Anderson, N. G.; Anderson, N. L. Electrophoresis 1996, 17, 443-453. (7) Go ¨rg, A.; Boguth, G.; Obermaier, C.; Posch, A.; Weiss, W. Electrophoresis 1995, 16, 1079-1086. (8) Dianoux, A. C.; Stasia, M. J.; Gagnon, J.; Vignais, P. V. Biochemistry 1992, 31, 5898-5905.

1078 Analytical Chemistry, Vol. 76, No. 4, February 15, 2004

With the emergence of matrix-assisted laser desorption/ ionization (MALDI)10,11 and electrospray ionization (ESI),12,13 mass spectrometry is increasingly used to further and more specifically characterize proteins by determining their intact mass and that of chemically or enzymatically generated fragments. Due to the high accuracy, sensitivity, high throughput, and high information content, mass spectrometry has become an outstanding complement to gel electrophoresis. MALDI has an advantage over ESI in many applications because it can tolerate many of the buffers and additives used in gel separation without extensive cleanup procedures. There are now a number of reports involving various ways that MALDI and ESI can be employed to assist in the characterization of proteins isolated by gel electrophoresis. Typically, proteins are localized in spots, excised from the gel, and digested by suitable proteases in situ. The resulting peptide fragments are eluted, purified, and subjected to MALDI-MS and ESI-MS.14-17 In some procedures, proteins are electroblotted onto a membrane or other solid support for direct MALDI analysis or for subsequent in situ digestion and mass mapping of proteolytic digest components. A variety of membranes have been employed, including nitrocellulose,18 poly(vinylene difluoride), polyamide,19-21 and nylon.22 Whereas ESI-MS requires the elution of the proteins or fragments from the blot membrane, MALDI-MS offers the option of either desorbing the proteins directly from the membrane or performing (9) Schuhmacher, M.; Glocker, M.; Wunderlin, M.; Przybylski, M. Electrophoresis 1996, 17, 848-854. (10) Karas, M.; Hillenkamp, F. Anal. Chem. 1988, 60, 2299-2301. (11) Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, T. Rapid Commun. Mass Spectrom. 1988, 82, 151-153. (12) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C. M. Science 1989, 246, 64-71. (13) Smith, R. D.; Loo, J. A.; Ogorzalek Loo, R. R.; Busman, M.; Udseth, H. R. Mass Spectrom. Rev. 1991, 10, 359-451. (14) Hall, S. C.; Smith, D. M.; Masiarz, F. R.; Soo, V. M.; Tran, H. M.; Epstein, L. B.; Burlingame, A. L. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 19271931. (15) Pappin, D. J.; Hojrup, P.; Bleasby, A. J. Curr. Biol. 1993, 3, 327-332. (16) Mortz, E.; Vorm, O.; Mann, M.; Roepstorff, P. Biol. Mass Spectrom. 1994, 23, 249-261. (17) Clauser, K. R.; Hall, S. C.; Smith, D. M.; Webb, J. W.; Andrews, L. E.; Tran, H. M.; Ebstein, L. B.; Burlingame, A. L. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 5072-5076. (18) Klarskov, K.; Roepstorff, P. Biol. Mass Spectrom. 1993, 22, 433-440. (19) Eckerskorn, C.; Strupat, K.; Karas, M.; Hillenkamp, F.; Lottspeich, F. Electrophoresis 1992, 13, 664-665. (20) Strupat, K.; Karas, M.; Hillenkamp, F.; Eckerskorn, C.; Lottspeich, F. Anal. Chem. 1994, 66, 464-670. (21) Vestling, M. M.; Fenselau, C. Anal. Chem. 1994, 66, 471-477. (22) Zaluzec, E. J.; Gage, D. A.; Allison, J.; Watson, J.; Throck. J. Am. Soc. Mass Spectrom. 1994, 5, 230-237. 10.1021/ac034879n CCC: $27.50

© 2004 American Chemical Society Published on Web 01/15/2004

the analysis using standard solution preparations after elution. Several laboratories have reported MALDI-MS of proteins directly from membranes after gel separation and electroblotting.23-28 Alternatively, electroelution can be employed prior to the solutionphase analysis of gel-isolated proteins by mass spectrometry.29-32 The mass spectrometric analysis of peptides and proteins by UV-MALDI directly from polyacrylamide gels33 and from isoelectric focusing gels34 has also been reported. These methods rely on ultrathin gels, and the sample handling includes drying the gels prior to matrix addition. Recently, infrared laser desorption and ionization of peptides and proteins deposited on top of a frozen polyacrylamide gel has been presented.35 Whereas analyte signals were observed at a wavelength of 5.9 µm, no ionization was observed at 2.9 µm, which is coincident with the strong OH stretch absorption of water. In this work, we report infrared laser desorption and ionization of peptides and proteins directly from a polyacrylamide gel after electrophoretic separation without the addition of a matrix. The additional sample-processing steps of extraction of the protein from the gel, blotting the protein onto a membrane, freezing the gel, or addition of an exogenous matrix prior to mass spectrometry analysis were avoided. Ionization of analytes at the wavelength of 2.9 µm was observed at room temperature using the OH stretch vibrational absorption of the gel and the water that remains in the gel after vacuum-drying. EXPERIMENTAL SECTION Materials. Bradykinin and bovine insulin were obtained from Sigma Chemical Co. (St. Louis, MO). Acetic acid, methanol, and β-mercaptoethnol were acquired from Fisher Scientific (Pittsburgh, PA). Tris was purchased from Life Technologies (Gaithersburg, MD). Except when otherwise noted, all chemicals used for gel electrophoresis, including polypeptide standards, were supplied by Bio-Rad Laboratories (Hercules, CA). All chemicals were used without further purification. Gel Electrophoresis. Electrophoresis was carried out on a Mini-Protean 3 miniature slab gel system (Bio-Rad Laboratories). The solutions and minigels for Tris/tricine SDS-PAGE were prepared according to the procedure of Schagger and von Jagow4 and run at 150 V for 70-90 min. The concentration of acrylamide (23) Schreiner, M.; Strupat, K.; Lottspeich, F.; Eckerskorn, C. Electrophoresis 1996, 17, 954-961. (24) Blais, J. C.; Nagnan-le-Meillour, P.; Bolbach, G.; Tabet, J. C. Rapid Commun. Mass Spectrom. 1996, 10, 1-4. (25) Patterson, S. D. Electrophoresis 1995, 16, 1104-1114. (26) Sutton, C. W.; Wheeler, C. H.; Sally, U.; Corbett, J. M.; Dunn, M. J. Electrophoresis 1997, 18, 424-431. (27) Eckerskorn, C.; Strupat, K.; Schleuder, D.; Sanchez, J. C.; Hochstrasser, D.; Lottspeich, F.; Hillenkamp, F. Anal. Chem. 1997, 69, 2888-2892. (28) Schleuder D.; Hillenkamp, F.; Strupat, K. Anal. Chem. 1999, 71, 32383247. (29) Cohen, S. L.; Chait, B. T. Anal. Biochem. 1997, 247, 257-267. (30) Klarskov, K.; Roepstorff, P. Biol. Mass Spectrom. 1993, 22, 433-440. (31) Mortz, E.; Sareneva, T.; Haebel, S.; Julkunen, I.; Roespstorff, P. Electrophoresis 1996, 17, 925-931. (32) Clarke, N. J.; Li, F.; Tomlinson, A. J.; Naylor, S. J. Am. Soc. Mass Spectrom. 1998, 9, 88-91. (33) Orgorzalek Loo, R. R.; Stevenson, T. I.; Mitchel, C.; Loo, J. A.; Andrews, P. C. Anal. Chem. 1996, 68, 1910-1917. (34) Orgorzalek Loo, R. R.; Mitchel, C.; Stevenson, T. I.; Martin, S. A.; Hines, W. M.; Juhasz, P.; Patterson, D. H.; Peltier, J. M.; Loo, J. A.; Andrews, P. C. Electrophoresis 1997, 18, 382-390. (35) Baltz-Knorr, M.; Ermer, D.; Schriver, K. E.; Haglund, R. F., Jr. J. Mass Spectrom. 2002, 37, 254-258.

in the resolving gel was 16.5% T and the acrylamide-to-biscrylamide ratio was 30:1. Here % T denotes the weight percentage of total monomer (acrylamide and bisacrylamide) whereas % C denotes the percentage of the cross-linker relative to the total monomer mass. The small-pore resolving gel was overlaid by a 7.5% T, 3% C stacking gel (2 cm). The thickness of the gel was ∼750 µm. The protein samples were incubated for 30 min at 40 °C in 4% SDS, 12% glycerol, 50 mM Tris-HCl (pH 6.8), and 0.02% Coomassie Brilliant Blue G-250. A mixture of polypeptide standards (triosephosphate isomerase, 26.6 kDa; myoglobin, 16.95 kDa; R-lactalbumin, 14.4 kDa; aprotinin, 6.5 kDa; insulin B chain, 3.5 kDa; bacitracin, 1.4 kDa) was loaded onto the gel with amounts of 0.5 µg/lane. Aliquots of 5 nmol of bradykinin and bovine insulin were loaded per gel lane. The effect of reducing agents during electrophoresis was studied by adding 5% β-mercaptoethnol. After electrophoresis, the gel was stained with 0.25% Coomassie Blue R250 in 50% methanol in water containing 10% acetic acid for 30 min and destained overnight in 10% methanol in 1% aqueous acetic acid. Gel pieces 5 mm in size containing the analyte were excised and mounted on the sample target using double-sided conductive tape (Electron Microscopy Sciences, Ft. Washington, PA). Preparation of Polypeptides Spotted onto Gel Surfaces. The bradykinin and bovine insulin were dissolved in doubly distilled water containing 0.1% trifluoroaetic acid to concentrations of 5 and 3 mM, respectively. A 5-mm-diameter piece of SDScontaining polyacrylamide gel was cut out and mounted on the sample target using double-sided conductive tape; 5-µL aliquots of analyte solutions were applied to the gel sections. After 10 min, the gel sections were rinsed with dd-H2O. Fourier Transform Infrared-Attenuated Total Reflection (FT-IR-ATR) Spectra of Gels. Transmission spectra of gels were acquired in a Perkin-Elmer spectrometer with a liquid nitrogencooled mercury cadmium telluride detector, continuously purged with dry nitrogen, at 4-cm-1 resolution. A 45° single pass trapezoidal ATR silicon plate (dimensions 50 × 20 × 2 mm, Harrick Scientific Corp., New York) internal reflection element was mounted in a single-beam multiple internal reflection attachment (model 9, Foxboro, MA). Background and sample spectra were recorded by averaging 100 scans. Mass Spectrometry. The MALDI-TOF mass spectrometer used in this work has been described in details previously.36-38 Briefly, the instrument is a 1-m linear time-of-flight instrument with delayed ion extraction. The IR source is a Nd:YAG pumped optical parametric oscillator (OPO) that is tunable from 1.45 to 4.0 µm. Experiments were run at 2.94-µm wavelength with the sample at room temperature. The OPO output was focused onto the sample to a spot size of approximately 200 by 300 µm as determined using laser burn paper and a measuring magnifier. At this spot size, IR fluences ranged from 1000 to 5000 J/m2. Because a large fraction of the laser shots did not provide sufficient signal, mass spectra were acquired by averaging single mass spectra selected by the computer. Single laser shot spectra in which peaks in the mass range of the analyte were larger than 40 mV were summed, while the spectra with signals lower than this (36) Sheffer, J. D.; Murray, K. K. J. Mass Spectrom. 2000, 35, 95-97. (37) Caldwell, K. L.; Murray, K. K. Appl. Surf. Sci. 1998, 127-129, 242-247. (38) Bhattacharya, S. H.; Raiford, T. J.; Murray, K. K. Anal. Chem. 2002, 74, 2228-2231.

Analytical Chemistry, Vol. 76, No. 4, February 15, 2004

1079

Figure 1. FT-IR-ATR spectra of dry (black solid line) and wet (dashed line) polyacrylamide gel.

were rejected. The sample target was continuously moved to provide a fresh area of sample for each laser shot. Here, signal loss may be due to ablation of the gel, chemical modification of the gel, desorption of the sample, or desorption of the solvent. When operated in this manner, ∼10% of the laser shots were summed. Mass spectra shown below are the summed average of five selected single laser shot mass spectra unless otherwise indicated. RESULTS AND DISCUSSION The significance of a sufficient optical absorption of the matrix at the laser wavelength has been recognized since the early days of MALDI.39,40 As a general rule, spectrum quality increases with absorption,41,42 although, once the absorption exceeds a certain value, further improvements cannot be obtained.43 Qualitatively, the best mass spectrometric performance has been obtained at wavelengths near the absorption maximum of the matrix. In this study, the material acting as the matrix is either the gel or water held within the gel that is not completely removed by vacuum desiccation. To determine the relative contributions of gel and water absorption, the IR absorption spectrum of the polyacrylamide gel was investigated. Figure 1 shows the FT-IR-ATR spectrum of a dry 16.5% sodium dodecyl sulfate polyacrylamide gel (black solid line) taken at room temperature. As with the mass spectrometric analysis, the gel was dried under vacuum for 30 min before the IR spectrum was obtained. The IR spectrum of a wet gel (dashed line) is shown for comparison. The FT-IR-ATR spectrum of the wet gel is similar to the FT-IR transmission spectrum of wet polyacrylamide gel previously reported.35 The broad peak centered around 3 µm contains the N-H and O-H stretching modes of both water and acrylamide gel. The peak centered on 6 µm contains the CdO and CsN stretching modes, the NsH deformation mode of the gel, and the OsH bending mode of the gel and water. The ratio of the 6- to 3-µm peaks in (39) Karas, M.; Bachmann, D.; Hillenkamp, F. Anal. Chem. 1985, 57, 29352539. (40) Beavis, R. C.; Chait, B. T. Rapid Commun. Mass Spectrom. 1989, 3, 432435. (41) Dreisewerd, K. Chem. Rev. 2003, 103, 395-425. (42) Horneffer, V.; Dreisewerd, K.; Luedemann, H. C.; Hillenkamp, F.; Laege, M.; Strupat, K. Int. J. Mass Spectrom. 1999, 185/186/187, 859-870. (43) Chen, X.; Caroll, J. A.; Beavis, R. C. J. Am. Soc. Mass Spectrom. 1998, 9, 885.

1080 Analytical Chemistry, Vol. 76, No. 4, February 15, 2004

Figure 2. Infrared (2.94 µm) laser desorption/ionization mass spectra of (a) bradykinin and (b) bovine insulin deposited on the top of a polyacrylamide gel with no added matrix.

the spectra suggests that the vacuum-dried gel still contains a significant fraction of water but with relatively high molar ratio of the gel to the water. Initial mass spectrometry experiments were carried out with desorption and ionization of analytes deposited on polyacrylamide gels without the addition of matrix. The IR-LDI mass spectra of bradykinin and bovine insulin deposited on top of the gels are shown in Figure 2. The base peaks in the mass spectra are protonated bradykinin (Figure 2a) and protonated bovine insulin (Figure 2b). The bradykinin mass spectrum contains intense lowmass peaks associated with Na+ and K+ as well as additional peaks below m/z 200 that are most likely associated with gel components. The peak marked with an asterisk is 220 Da above the base peak in Figure 2a and could not be assigned to an adduct of any intact component of the gel with the bradykinin. In Figure 2b, there is an unresolved low-mass background near m/z 2000 that may be due to gel components or water clusters. Mass spectra obtained at shorter extraction delays suggest that the molecules giving rise to these peaks contain carbon atoms. To evaluate the ability of the direct combination of gel electrophoresis and laser desorption/ionization mass spectrometry, bradykinin and bovine insulin were separated in a 16.5% T polyacrylamide gel. An image of a gel run under the conditions used to obtain the mass spectra is shown in Figure 3. Tricine SDS-PAGE allows a resolution of small proteins at lower acrylamide concentrations than glycine SDS-PAGE.4 Using the Tris/tricine system with 16.5% T/3% C narrow pore size separation gels, bradykinin, bovine insulin, insulin A chain, and insulin B

Figure 3. Separation on a 16.5% T polyacrylamide gel of molecular mass standards (lane 1), bradykinin (lane 2), bovine insulin (lane 3), and reduced bovine insulin (lane 4).

chain migrate as resolved bands. The polypeptides were subsequently visualized using a Coomassie Brilliant Blue stain. The first lane contains the molecular mass size standards at 1.4, 3.5, 6.5, 14.4, 17.0, and 26.6 kDa (vide supra). Note that the molecules in lane 1 of Figure 3 were used for gel calibration only and were not analyzed by mass spectrometry. The second lane contains bradykinin and the third lane bovine insulin. The fourth lane contains bovine insulin with β-mercaptoethnol reducing agent. Note that when the reducing agent is added, both the insulin A chain and B chain are observed in the gel but the intact insulin molecule is not. Mass spectra were obtained by direct desorption and ionization of the analyte from the gels with no added matrix. Figure 4a shows the IR-LDI mass spectrum of bradykinin in the gel, corresponding to lane 2 in Figure 3. The largest mass spectral peak associated with the analyte is the sodium cation adduct, denoted [M + Na]+ in Figure 4a. Intense Na+ and K+ peaks are observed in the lowmass region. Between m/z 100 and the bradykinin [M + Na]+ peak, additional interference peaks are observed, which may result from fragmentation of the Coomassie dye. The largest analyte ion observed in this work was bovine insulin (Mr ) 5733.6); a mass spectrum of bovine insulin using IR-LDI from gels is shown in Figure 4b. The spectrum is characterized by a large protonated molecule peak. The peaks marked with an asterisk are separated from the protonated insulin by ∼800 Da and are tentatively assigned as adducts of insulin with the Coomassie dye. Interfering peaks below m/z 1800 may also be associated with the dye, possibly clustered with other gel components. To denature the native protein conformation and improve the performance of the gel separation, a reducing agent is often used. A standard procedure involving the use of 5% β-mercaptoethanol during electrophoresis was used with bovine insulin. The results of this procedure are shown in lane 4 of Figure 3. Here, the insulin A and B chains are clearly separated (MR ) 2239 and 3495, respectively). The 2.94-µm IR-LDI mass spectrum obtained from this separation is shown in Figure 5, which is the average of 10 single laser shot mass spectra. The mass spectrum of the insulin B chain shows an intense [M + H]+ peak with a small adduct with acrylamide. (MR ) 71). The spot corresponding to the insulin A chain was not tested due to the low ionization efficiency of that peptide.44 The result shown in Figure 5 demonstrates the advantage of laser desorption mass spectrometry directly from the gel for identification of chemically modified components.

Figure 4. Infrared laser desorption/ionization mass spectra of (a) bradykinin and (b) bovine insulin following gel elution.

Figure 5. Infrared laser desorption/ionization mass spectrum of bovine insulin B chain following reduction and gel elution of insulin.

Previous researchers have reported a lack of success at IRLDI and IR-MALDI directly from a gel under conditions that are similar, but not identical to those reported above. Hillenkamp and co-workers were unable to desorb proteins directly from a gel using IR-MALDI.20 However, proteins could be ionized efficiently by IR-MALDI after electroblotting. All of the analytes tested in the IR-MALDI electroblotting work were larger than 10 kDa in mass, and the quantities used for electrophoresis were between 1 and 10 pmol per lane. These previously reported results are consistent with our observation that nanomole quantities of analyte per electrophoresis lane were required to obtain mass spectra. (44) Zhu, Y. F.; Lee, K. L.; Tang, K.; Allman, S. L.; Taranenko, N. I.; Chen, C. H. Rapid Commun Mass Spectrom 1995, 9, 1315-1320.

Analytical Chemistry, Vol. 76, No. 4, February 15, 2004

1081

Haglund and co-workers were unsuccessful at obtaining matrixfree IR-LDI mass spectra from gels at 2.9 µm but were successful at 6 µm.35 In these experiments, the gels were cooled to 105 K rather than dried as in the data reported above. The quantity of analyte loaded on the gel was in the range of 10 nmol/lane. The fact that direct from gel ionization was successful for dried gels and not for wet ones suggests that the removal of water is essential for obtaining ionization at 2.9 µm. This is consistent with previous reports of IR-MALDI using the waters of hydration of lyophilized proteins as the matrix; it was found that water-ice does not function well as a matrix due to its propensity for forming large protonated clusters.45 CONCLUSIONS We have demonstrated the direct coupling of matrix-free 2.94µm infrared laser desorption/ionization mass spectrometry with polyacrylamide gel electrophoresis. Because this method does not require the addition of matrix, many of the intermediate processing steps are avoided, and the method provides potential advantages in speed and reduced complexity. Although mass spectra of polypeptides and protein have been obtained directly from gels, the practical application of this technique is restricted by the limit of detection. In this work, it was necessary to load nanomole quantities of peptide and protein analytes into each band for a successful analysis. In a typical gel electrophoresis analysis, the amount of proteins loaded per gel band is a factor of 10 or more lower. On the other hand, the amount of gel required for IR-LDI (45) Berkenkamp, S.; Karas, M.; Hillenkamp, F. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 7003-7007. (46) Yao, S.; Anex, D. S.; Caldwell, W. B.; Arnold, D. W.; Smith, K. B.; Schultz, P. G. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 5372-5377. (47) Hofmann, O.; Che, D.; Cruickshank, K. A.; Muller, U. R. Anal. Chem 1999, 71, 678-686. (48) Ullom, J. N.; Frank, M.; Gard, E. E.; Horn, J. M.; Labov, S. E.; Langry, K.; Magnotta, F.; Stanion, K. A.; Hack, C. A.; Benner, W. H. Anal. Chem. 2001, 73, 2331-2337.

1082

Analytical Chemistry, Vol. 76, No. 4, February 15, 2004

analysis is relatively small since the laser irradiates a small fraction of a gel band. The key to minimizing sample consumption and optimizing the limit of detection may be moving the gel separation to a microfluidic chip format.46,47 IR-LDI mass spectrometry applied to a capillary gel microchip will provide the best detection limit while at the same time allowing multiple lanes to be run in a small format device. A typical channel on a microfluidic device is 100 µm wide and 50 µm deep. A focused laser will irradiate a 100-µm length of this channel, a volume of 5 × 105 µm3. This is more than 104 times less volume than in the analyte containing spots analyzed in this work, suggesting that substantial improvement in detection limit might be obtained from the microfluidic chip format. Direct from gel ionization of molecules up to 6 kDa was observed in this study. This size range is comparable to what has been reported previously for matrix-free IR laser desorption.35,38,48 Although the observed mass range is suitable for peptide mass mapping of protein digests, an improvement in the mass range is desirable for the direct analysis of intact gel-separated proteins. Future studies will focus on applying IR-LDI to gel-separated proteins of higher mass. Adduct formation with the Coomassie dye and other gel components can cause additional spectral complexity under some conditions, but it may be possible to further optimize the gel and staining procedures to minimize these interferences. Further optimization may be possible through the selection of the IR laser wavelength. ACKNOWLEDGMENT This work is supported by the National Science Foundation Grant CHE-0196568, by the National Institutes of Health Grant R42RR15134, and by Louisiana State University. Received for review July 30, 2003. Accepted December 3, 2003. AC034879N