AC Research
Anal. Chem. 1998, 70, 4879-4884
Accelerated Articles
A Microscale Electrospray Interface Incorporating a Monolithic, Poly(styrene-divinylbenzene) Support for On-Line Liquid Chromatography/ Tandem Mass Spectrometry Analysis of Peptides and Proteins Roger E. Moore, Larry Licklider, Detlef Schumann, and Terry D. Lee*
Division of Immunology, Beckman Research Institute of The City Of Hope, Duarte, California 91010
A methodology is described for creating a monolithic chromatography support within a pulled fused-silica electrospray needle. The monolith was formed from a mixture of styrene, divinylbenzene, 1-dodecanol, and toluene using 2,2′-azobis(isobutyronitrile) as the catalyst. The mixture was loaded into 150-µm-i.d. fused-silica capillary tubing with a pulled 5-10-µm needle tip at one end. Polymerization at 65 °C followed by removal of the porogen material yielded a stable, porous, monolithic support which had excellent properties for the separation and on-line, electrospray, mass spectrometry analysis of peptides and proteins. The performance of the monolithfilled electrospray needles was compared with similar needles filled with commercial C18 silica and polymeric particulate supports. Separation efficiencies for both protein and peptide mixtures were generally equal to or better than the particulate supports at comparable pressures and flow rates. The ion chromatograms derived from the on-line MS analysis were remarkably free from chemical background signals that often complicate the LC/MS analysis of femtomole amounts of sample. Good sequence coverage was obtained by LC/MS/MS analysis of the peptide mixture obtained from a protein isolated by silver-stained gel electrophoresis. The capability of the monolith to do peak parking experiments was demonstrated by the characterization of an immunoreactive HPLC fraction. The simple fabrication method, chromatographic performance, and robust nature of these microscale integrated column electrospray sources make 10.1021/ac980723p CCC: $15.00 Published on Web 10/22/1998
© 1998 American Chemical Society
them ideally suited for high-sensitivity tandem LC/MS analyses.
Sensitivity improvements in protein characterizations using online liquid chromatography/electrospray mass spectrometry (LC/ ESMS) systems have been largely due to the sensitivity advantages of lower column flow rates and on-line sample handling.1-4 Additional sensitivity improvement can be gained by using smaller ES emitters with tip dimensions in the range of 1-25 µm, which enhances stability of the ion source and allows for the lowest possible flow rates.5-7 Low-nanoliter per minute flow rates have lowered detection limits into the mid-attomole range when a minute amount of reversed-phase support was placed into the emitter tip for concentration of a dilute protein sample.8 Recently, a microscale ES interface (hereafter referred to as microspray) was developed for on-line separations of complex peptide mix* Corresponding author: (phone) 626 301-8301; (fax) 626 301-8186; (e-mail)
[email protected]. (1) Huang, E. C.; Henion, J. D. Anal. Chem. 1991, 63, 732-739. (2) Deterding, L. J.; Parker, C. E.; Perkins, J. R.; Moseley, M. A.; Jorgenson, J. W.; Tomer, K. B. J. Chromatogr. 1991, 554, 329-338. (3) Hunt, D. F.; Shabanowitz, J.; Mosely, M. A.; McCormack, A. L.; Michel, H.; Martino, P. A.; Tomer, K. B.; Jorgenson, J. W. In Methods in Protein Analysis; Jornvall, H., Hoog, J. O., Gustavsson, A. M., Eds.; Birkhauser Verlag: Basel, 1991; pp 257-266. (4) Davis, M. T.; Stahl, D. C.; Hefta, S. A.; Lee, T. D. Anal. Chem. 1995, 67, 4549-4556. (5) Wilm, M. S.; Mann, M. Int. J. Mass Spectrom. Ion Processes 1994, 136, 167-180. (6) Valaskovic, G. A.; Kelleher, N. L.; Little, D. P.; Aaserud, D. J.; McLafferty, F. W. Anal. Chem. 1995, 67, 3802-3805. (7) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1-8. (8) Emmett, M. R.; Caprioli, R. M. J. Am. Soc. Mass Spectrom. 1994, 5, 605613.
Analytical Chemistry, Vol. 70, No. 23, December 1, 1998 4879
tures.4,9 A computer-controlled microgradient HPLC system was coupled with the microspray interface in which a microspray needle was packed with reversed-phase media.9 This interface has been used routinely to perform variable flow rate (low nLµL/min) LC/MS/MS analyses of femtomole amounts of peptide mixtures. One drawback of the packed needle interface is the level of technical expertise needed to assemble the needle structure and pack it with particulate supports. In this report, a more convenient means of placing a reversed-phase support into the needles is shown which is based on demonstrations by others of macroporous polymer monoliths formed in situ.10,11 Poly(styrene-divinylbenzene) (PS/DVB) monoliths were formed in the ES needle by simply filling it with the monomer mixture and controlling temperature during the polymerization reaction. PS/DVB monoliths synthesized elsewhere under similar conditions have been well characterized in regard to control of their pore properties.11 Careful attention to the preparation of the reaction mixture was necessary to allow reproducible separations on the filled needles. The effectiveness of the monolith-filled needles was examined by performing on-line LC/MS/MS analyses of femtomole amounts of protein and peptide mixtures. EXPERIMENTAL METHODS Materials. The styrene (99%), divinylbenzene (80% divinyl monomer, technical grade), 1-dodecanol (98%), toluene (HPLC grade), and 2,2′-azobis(isobutyronitrile) (AIBN, 98%) were purchased from Aldrich and used without further purification. The Sequenal grade trifluoroacetic acid (TFA) was purchased from Pierce (Rockford, IL). Fused-silica tubing was purchased from Polymicro Technologies (Phoenix, AZ). The Vydac 218TP C18 bonded silica packing was purchased from The Separations Group (Hesperia, CA), the Zorbax 300SB C18 bonded silica was from Rockland Technologies (Newport, DE) and the Poros 10 R2 polymeric support was from Perseptive Biosystems (Framingham, MA). The protein standard containing ribonuclease A, cytochrome c, holotransferrin, and apomyoglobin was purchased from Sigma (St. Louis, MO) and was quantitated by amino acid analysis by the City of Hope Mass Spectrometry core facility, which also provided the Lys-C cytochrome c digest standard. The immunoreactive fraction used for the peak parking experiment was provided by Dr. Michael Adams of the Department of Entomology, University of California, Riverside. All other chemicals were reagent grade or higher. Microspray Needles. The microspray needles were pulled from 350-µm-o.d., 150-µm-i.d. fused-silica tubing using a laser-based micropipet puller (Sutter Instruments, model P-2000, Novato CA). The program used a two-stage pulling process to produce a tip with a 1-2-mm-long narrowed region 30-50 µm in diameter with a 5-10-µm orifice. The instrument parameters for the first stage were as follows: 550 heat, 0 filament, 30 velocity, 200 delay, and 0 pull. For the second stage of the program, the heat value was reduced to 500 and the filament was increased to 5. The second stage was repeated until the needles separated, which typically required four cycles. Needles were initially 5-6 cm in length and then trimmed to 3-4 cm after filling. The initial needle length (9) Davis, M. T.; Lee, T. D. J. Am. Soc. Mass Spectrom. 1998, 9, 194-201. (10) Wang, Q. C.; Svec, F.; Frechet, J. M. J. Anal. Chem. 1993, 65, 2243-2248. (11) Viklund, C.; Svec, F.; Frechet, J. M. J. Chem. Mater. 1996, 8, 744-750.
4880 Analytical Chemistry, Vol. 70, No. 23, December 1, 1998
(5-6 cm) is conveniently made using the laser puller. The final length was convenient for mounting in our microspray interface. After a number of injections of dirty samples, it is sometimes necessary to trim the top part of the column to restore normal flow. This has no noticeable effect on performance. Monolith-Filled Microspray Needles. AIBN (20 mg) was dissolved in a mixture of styrene (750 mg), divinylbenzene (1.250 g), 1-dodecanol (2.250 g), and toluene (750 mg). This mixture was then stirred magnetically for 10 min, sparged with Ar for 1 min, and then sonicated for 10 min. Microspray needles were filled by placing their butt ends into the premonolith mixture and allowing capillary action to pull the mixture to the tip. The filled needles were then placed in a HPLC column oven thermostated to 65 °C. The needles were supported by a ridged plastic block to minimize contact with hot oven surfaces and promote even heating by the surrounding air. The mixture was then allowed to polymerize overnight. After polymerization, the needles were visually inspected under a light-field dissecting microscope to detect any abnormalities. The needles were then flushed with a 50% acetonitrile/0.1% TFA solution to remove the dodecanol, toluene, and any other soluble compounds. Flushing was carried out at 65 °C using a pressure gradient (10-500 psi over 1 min, 500 psi for 15 min, 500-10 psi over 3 min) to avoid excessively rapid pressurization and depressurization of the monolith. At some point during the polymerization to form the monolith, the mixture volume decreases leaving a gap of a few millimeters at the back end the needle and a smaller void (a few nanoliters) at the needle tip. Particle-Packed Needles. Fritted microspray needles were packed with C18 bonded silica as previously described9 except that Whatman GF/A glass fiber filter was substituted for Durapore membrane for the frit when the Vydac C18 and Poros 10R2 structures were made. A lower pressure (1000 vs 4000 psi) was used to pack the Poros 10 R2 filled needles. LC/MS Analysis. Mass spectra were generated using a Finnigan LCQ ion trap mass spectrometer. The micro-LC system and microspray interface were used as previously described.4,9 Approximate flow rates were determined by collecting column effluent in a graduated glass capillary and were proportional to applied pressure. Except where indicated, all samples were loaded at 500 psi for 90 s followed by reduction in pressure to 50 psi for gradient elution. For the analysis of the unknown protein digest, a sample (10 µL) was injected (500 psi for the monolith, 1500 psi for the Zorbax C18) and then rinsed at the loading pressure for 10 min to remove contaminants. Gradient elution was then performed at 25 psi for the monolith and 50 psi for the C18 support. The gradient used in all analyses was from 2% to 92% buffer B over 10 µL (buffer A, 0.02% TFA; buffer B, 90% acetonitrile/0.014% TFA). Preparation of an Unknown Protein Digest. The unknown protein was extracted from 108 transfected murine colon carcinoma MC38 cells which were lysed by vortex stirring. Immunoprecipitation was performed by a procedure based on a commercial protocol that uses monoclonal antibody (mAb) coupled beads (Boehringer Mannheim, kit 1719394). Precipitated proteins were recovered, and 2D gel electrophoresis was performed as described elsewhere.12 Silver staining and in-gel digestion of an abundant (12) O’Farrell, P. H. J. Biol. Chem. 1975, 250, 4007-4021.
protein spot (pH 5-6, MW 40 000-50 000) using modified trypsin (Promega) were performed as described previously.13 Database searches were performed using the SEQUEST database search program included in the Finnigan LCQ data system software.14 RESULTS AND DISCUSSION Packed Electrospray Needles. Packing microscale electrospray needles with LC supports is problematic because of the small dimensions of the needle orifices (typically 5 µL/min at 500 psi) and good retentive and separation properties for both protein and peptide mixtures. (13) Shevchenko, A.; Wilm, M.; Vorm, O.; Mann, M. Anal. Chem. 1996, 68, 850-858. (14) Eng, J. K.; McCormack, A. L.; Yates, J. R. J. Am. Soc. Mass Spectrom. 1994, 5, 976-989.
Figure 1. Base peak chromatograms for the LC/MS analyses of 20 ng (in 5 µL) of a protein standard containing (1) ribonuclease A, (2) cytochrome c, (3) holotransferrin, and (4) apomyoglobin using (A) a PS/DVB monolith-filled needle, (B) a Vydac C18-packed needle, and (C) a Poros R2-packed needle. NL, normalization level; the number below is the absolute intensity corresponding to 100% on the relative-abundance scale.
Peptide and Protein Separations. When compared to either the Vydac C18 (Figure 1B) or Poros R2 (Figure 1C) supports, the PS/DVB monolith (Figure 1A) gave generally higher resolution separations of a standard protein mixture containing equal portions by weight (20 ng of total protein) ribonuclease A (1), cytochrome c (2), holotransferrin (3), and apomyoglobin (4). For these analyses, the protein test mixture was analyzed on all three supports with flow rates of ∼700 nL/min at 50 psi and chromatograms were plotted using the intensity of the most intense ion for each mass spectrum. The performance of the Poros support was particularly sensitive to flow rate. At twice the flow rate (Figure 1C inset), the observed separation was similar to that of the monolith. Fluid flow in Poros supports above a minimum linear velocity facilitates mass transfer of proteins into smaller diffusive pores within the particles.15,16 The Poros R2-filled needle also had a broad background peak which eluted in the same range (15) Afeyan, N. B.; Gordon, N. F.; Mazsaroff, I.; Varady, L.; Fulton, S. P.; Yang, Y. B.; Regnier, F. E. J. Chromatogr. 1990, 519, 1-29. (16) Kassel, D. B.; Shushan, B.; Sakuma, T.; Salzmann, J. P. Anal. Chem. 1994, 66, 236-243.
Analytical Chemistry, Vol. 70, No. 23, December 1, 1998
4881
Figure 2. LC/MS analyses at a higher dilution (25 µL, 1 ng of total protein) of the same protein standard used for Figure 1. Base peak and selected ion chromatograms are shown for the (A) PS/DVB monolith-filled needle, (B)Vydac C18-packed needle, and (C)Poros R2-packed needle.
of gradient compositions as the protein peaks but was not sample related. It should be noted that the Vydac C18 support partially separated a contaminating protein that in the case of the monolith or the Poros had coeluted in the peak labeled 3. The generally high separation efficiency of the monolith is a result of the unique selectivity of a PS/DVB stationary phase and the physical nature of the pore structure. The volume of the pores in the monolith was assumed to have a bimodal distribution and occupy >60% of the volume of an unfilled needle based on well-characterized PS/ DVB monoliths of similar origin.11 Large canallike pores (∼700nm diameter) accounted for the high flow rates of the monolithfilled needles, while the much smaller
