Anal. Chem. 2000, 72, 4266-4274
Subfemtomole MS and MS/MS Peptide Sequence Analysis Using Nano-HPLC Micro-ESI Fourier Transform Ion Cyclotron Resonance Mass Spectrometry Susan E. Martin,† Jeffrey Shabanowitz, Donald F. Hunt,*,‡ and Jarrod A. Marto
Department of Chemistry, University of Virginia, Charlottesville, Virginia 22904-4319
Subfemtomole peptide sequence analysis has been achieved using microcapillary HPLC columns, with integrated nanoelectrospray emitters, coupled directly to a Fourier transform ion cyclotron resonance mass spectrometer. Accurate mass ((0.010 Da) peptide maps are generated from a standard six-protein digest mixture, whose principle components span a concentration dynamic range of 1000:1. Iterative searches against ∼189 000 entries in the OWL database readily identify each protein, with high sequence coverage (20-60%), from as little as 10 amol loaded on-column. In addition, a simple variableflow HPLC apparatus provides for on-line tandem mass spectrometric analysis of tryptic peptides at the 400-amol level. MS/MS data are searched against ∼280 000 entries in a nonredundant protein database using SEQUEST. Accurate precursor and product ion mass information readily identifies primary amino acid sequences differing by asparagine vs aspartic acid (∆m ) 0.98 Da) and glutamine vs lysine (∆m ) 0.036 Da).
Parallel efforts in several fields have made mass spectrometry arguably the technique of choice for analysis of complex biological mixtures. The development of electrospray (ESI)1,2 and matrixassisted laser desorption/ionization (MALDI)3,4 as soft ionization sources fueled mass spectrometry’s current popularity in the biological sciences. ESI in particular has benefited from advances in microcapillary-scale high-performance liquid chromatography (HPLC),5-7 which provided on-line separation capabilities to mass spectral analysis;8,9 several groups quickly adapted the com* To whom correspondence should be addressed; (e-mail)
[email protected]; (fax) 804-982-2781. † Present address: Department of Chemistry, Furman University, 3300 Poinsett Highway; Greenville, SC 29613. ‡ Also a member of the Department of Pathology. (1) Fenn, J. B.; Yamashita, M. J. Phys. Chem. 1984, 88, 4671-4675. (2) Fenn, J. B.; Yamashita, M. J. Phys. Chem. 1984, 88, 4451-4459. (3) Tanaka, K.; Waki, H.; Ido, Y.; Akita, S.; Yoshida, Y.; Yoshida, T. Rapid Commun. Mass Spectrom. 1988, 2, 151-153. (4) Karas, M.; Bachmann, D.; Bahr, U.; Hillenkamp, F. Int. J. Mass Spectrom. Ion Processes 1987, 78, 53-68. (5) Kennedy, R. T.; Jorgenson, J. W. Anal. Chem. 1989, 61, 1128-1135. (6) Karlsson, K. E.; Novotny, M. Anal. Chem. 1988, 60, 1662-1665. (7) Flurer, C.; Borra, C.; Beale, S.; Novotny, M. Anal. Chem. 1988, 60, 18261829.
4266 Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
bination of LC-ESI-MS for peptide sequence analysis, building upon the original work of Hunt10-13 and Biemann.14-16 Overall performance (separation efficiency, sample recovery, ionization efficiency, detection limits, etc.) continues to improve as the mechanical scale of separation and electrospray devices reaches the micrometer and submicrometer regime.17-24 Concurrent developments in various computer algorithms now allow one to query the ever-expanding protein and nucleotide databases with mass spectrometric (MS)25-31 or tandem mass spectrometric (MS/MS)32-34 data. Finally, recent advances in instrument control (8) Hall, M.; Lewis, S.; Jardine, I.; Liu, J.; Novotny, M. J. Microcolumn Sep. 1990, 2, 285-292. (9) Griffin, P. R.; Coffman, J. A.; Hood, L. E.; Yates, J. R. Int. J. Mass Spectrom. Ion Processes 1991, 111, 131-149. (10) Hunt, D. F.; Bone, W. M.; Shabanowitz, J.; Rhodes, J.; Ballard, J. M. Anal. Chem. 1981, 53, 1704-1706. (11) Hunt, D. F.; Zhu, N. Z.; Shabanowitz, J. Rapid Commun. Mass Spectrom. 1989, 3, 122-124. (12) Hunt, D. F.; Yates, J. R.; Shabanowitz, J.; Winston, S.; Hauer, C. R. Proc. Natl. Acad. Sci. U.S.A. 1986, 83, 6233-6237. (13) Hunt, D. F.; Shabanowitz, J.; Yates, J. R. J. Chem. Soc., Chem. Commun. 1987, 548-550. (14) Gibson, B. W.; Biemann, K. Proc. Natl. Acad. Sci. U.S.A. 1984, 81, 19561960. (15) Biemann, K. Anal. Chem. 1986, 58, 1288A-1300A. (16) Biemann, K. Methods Enzymol. 1990, 193, 351-360. (17) Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1-8. (18) Wilm, M.; Shevchenko, A.; Houthaeve, T.; Breit, S.; Schweigerer, L.; Fotsis, T.; Mann, M. Nature 1996, 379, 466-469. (19) Valaskovic, G. A.; Kelleher, N. L.; Little, D. P.; Aaserud, D. J.; McLafferty, F. W. Anal. Chem. 1995, 67, 3802-3805. (20) Davis, M. T.; Stahl, D. C.; Hefta, S. A.; Lee, T. D. Anal. Chem. 1995, 67, 4549-4556. (21) Emmett, M. R.; White, F. M.; Hendrickson, C. L.; Shi, S. D. H.; Marshall, A. G. J. Am. Soc. Mass Spectrom. 1998, 9, 333-340. (22) Figeys, D.; Gygi, S. P.; McKinnon, G.; Aebersold, R. Anal. Chem. 1998, 70, 3728-3734. (23) Figeys, D.; Ning, Y.; Aebersold, R. Anal. Chem. 1997, 69, 3153-3160. (24) Figeys, D.; Aebersold, R. J. Biomech. Eng. 1999, 121, 7-12. (25) Mann, M.; Hojrup, P.; Roepstorff, P. Biol. Mass Spectrom. 1993, 22, 338345. (26) Pappin, D. J. C.; Hojrup, P.; Bleasby, A. J. Curr. Biol. 1993, 3, 327-332. (27) James, P.; Quadroni, M.; Carafoli, E.; Gonnet, G. Biochem. Biophys. Res. Commun. 1993, 195, 58-64. (28) Henzel, W. J.; Billeci, T. M.; Stults, J. T.; Wong, S. C.; Grimley, C.; Watanabe, C. Proc. Natl. Acad. Sci. U.S.A. 1993, 90, 5011-5015. (29) Yates, J. R.; Speicher, S.; Griffin, P. R.; Hunkapiller, T. Anal. Biochem. 1993, 214, 397-408. (30) Mortz, E.; O’Connor, P. B.; Roepstorff, P.; Kelleher, N. L.; Wood, T. D.; McLafferty, F. W.; Mann, M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 82648267. (31) Mann, M.; Wilm, M. Anal. Chem. 1994, 66, 4390-4399. 10.1021/ac000497v CCC: $19.00
© 2000 American Chemical Society Published on Web 08/12/2000
software have provided an unprecedented degree of automation;20,35,36 in fact the new field of proteomics is based, in large part, on integrating protein isolation, proteolytic degradation, mass analysis, and database searches into a high-throughput, automated process.37 To date most researchers have employed triple-quadrupole (TQ), time-of-flight (TOF), and quadrupole ion trap (QIT) mass spectrometers for peptide analysis. Generally speaking, MALDITOF instruments excel at accurate mass peptide mass fingerprinting, while both TQ and QIT instruments have demonstrated superior MS/MS capabilities. For example, Mann38 demonstrated mass accuracy better than 20 ppm for peptide tryptic maps while Takach et al. achieved values of 99%). Emmett et al.21 recently demonstrated low-femtomole MS detection limits for LC-FTMS operated in external ion accumulation mode. In this report, we present data from our FTMS instrument, which has been configured with electrospray ionization and external ion accumulation. Significant features include microcapillary HPLC columns with integrated nanospray emitters and a simple apparatus for controlling chromatographic peak widths, thus enabling real-time MS/MS experiments for peptide sequence analysis. Our initial results improve dramatically on previously reported LC-FTMS detection limits;21 more importantly, we demonstrate on-line MS/MS analysis (via infrared multiphoton photodissociation, IRMPD) of peptides at the subfemtomole level. High mass resolution (m/∆m >5000) and mass accuracy (∼510 ppm) are maintained in both MS and MS/MS modes. Data from these experiments are readily used in conjunction with existing database search algorithms to identify candidate source proteins. EXPERIMENTAL SECTION Fourier Transform Mass Spectrometer. Figure 1 shows a schematic representation of our current Fourier transform mass spectrometer, which has been significantly modified from earlier versions.51-54 The electrospray source is built around a standard Finnigan APCI/ESI module (Finnigan MAT, San Jose, CA). The heated metal capillary (150 °C) and tube lens assemblies were used in their commercial form. The mounting surface of the standard skimmer lens (SL) was faced off by 0.025 in. and secured to the ESI block with Teflon shim stock such that its dc potential could be floated relative to ground. Ions are transported to the Penning trap through a series of rf-only ion guides, consisting of two octupoles (O1 and O2: 2-mm-diameter rods, inscribed circle radius (r0) ) 2.7 mm, rod length (L)) 21.3 and 11.8 cm, respectively) followed by two quadrupoles (Q1 and Q2: 6.4-mmdiameter rods, r0 ) 2.7 mm, L ) 25.4 and 86.4 cm, respectively). 4268
Analytical Chemistry, Vol. 72, No. 18, September 15, 2000
Flat lenses, [interoctupole lens (IOL), octupole-quadrupole lens (OQL), and interquadrupole lens (IQL)] each with ∼2-mm i.d., provide conductance limits between the various pumping stages (P2-P5). A quadrupole “cap” lens (QCL, 5.6-mm i.d.) is located just after Q2 in order to minimize perturbations in the Penning trap electric fields due to rf and/or dc potentials on Q2. During operation, P1 is evacuated to 750 mTorr by a dual-stage rotary vane mechanical pump (Alcatel, Hingham, MS; model 2033, 27 cfm); the pressure is reduced to 1.5 mTorr in P2 by a turbomolecular drag pump (Balzers, Hudson, NH; model TMH 260, 230 L/s). The remaining regions, P3, P4, P5, are evacuated by three cryopumps (APD, Allentown, PA; model AP-6, 680 L/s, driven by APD HC-4 compressors) to 1 × 10-6, 5 × 10-9, and 2 × 10-9 Torr, respectively; pressures are monitored by magnetically shielded Bayard-Alpert-type ion gauges, located above each cryopump (∼70 cm from the central axis of the vacuum manifold). An open, cylindrical Penning trap was constructed from OFHC copper stock (2.25-in. i.d.) and assembled with ceramic spacers, aluminum support rods, and titanium machine screws. The cell assembly was secured on a 4.625-in.-diameter conflat-type flange at the far end of the high vacuum region (P5). Note that the vacuum manifold is pumped from only one end; in addition, the second quadrupole (Q2) terminates ∼20 cm from the cell flange. As a result, we chose somewhat atypical dimensions for the cell electrodes (trap plates, 1.125-in. length; excite/detect plates, 3.5in. length), in an attempt to obtain an acceptable compromise between trapping field quadrupolarity, excitation field linearity, and overall ion storage capacity. A barium fluoride window (Bicron, Solon, OH) was centered on the cell flange to allow transmission of an infrared (10.6 µm) laser beam (Synrad, Mukiteo, WA, model 48-2 60-W CW) for photodissociation of peptide ions. An Odyssey data station (Finnigan FTMS, Madison, WI) running software version 4.0 was used for instrument control and data acquisition. DAC outputs controlled dc potentials on both octupoles (O1 and O2), the IOL, and the OQL. Finnigan electronic modules, modified in-house, were used to control RF voltages on all multipole ion guides. External ion accumulation66 was performed by dropping the dc offset on O1 to -5 V while simultaneously raising the interoctupole lens to +9.75 V for a period of 1-2 s. Ions were injected into the Penning trap by raising both octupole dc offsets to 0 V, while dropping the voltage on IOL to -9.75 V. Ion injection time was optimized at 2.1 ms; gated trapping67,68 was employed at the cell, with the front and rear trap
plates held at 0 and +4 V respectively during ion injection. A 200ms ion cooling period followed in which both trap plates were held at +4 V (100 ms) and then +2 V (100 ms) prior to broadband excitation (54 kHz-1 MHz at -100 Hz/µs, ∼60 Vp-p) and detection (Nyquist bandwidth, 331 kHz with 64k data points). For MS/MS experiments, stored-waveform inverse Fourier transform (SWIFT) waveforms were used to select precursor ions. Power output from a CO2 laser was then adjusted to effect infrared multiphoton dissociation (IRMPD) of selected ions. All timedomain data sets were subjected to baseline correction and Hanning apodization prior to fast Fourier transform and magnitude mode display. To minimize spurious electronic noise, the quadrupole rf voltage (870 kHz, ∼1300 Vp-p) was gated “on” only during ion injection. Note that dc potentials critical for external ion accumulation were held in the “accumulation” state at all times (except during ion injection); similarly, rf voltage (2.5 MHz, ∼450 Vp-p) was applied continuously to both octupoles. Under these conditions, ion sampling efficiency >99% was obtained over the course of on-line chromatographic acquisitions. Voltage values on all other ion guide optics were held constant during data acquisition: heated metal capillary, +3 V; SL, +3 V; OQL, -9.75 V; IQL, 0 V; quadrupole dc offset (Q1 and Q2), -12 V; QCL, 0 V. HPLC Column/Spray Tip Assembly. In an effort to minimize sample loss and undesired peak broadening, particularly during MS-only acquisitions, we sought to construct microcapillary HPLC columns with integrated nanospray emitters (e.g., from a single piece of fused silica (FS)). To this end, we have pursued strategies in which the emitter tip is pulled as the last step in the assembly process. One such approach is briefly outlined below: A 30-cm length of 360 µm o.d. × 50 µm i.d. FS is cut, and ∼1 cm of polyimide coating is removed ∼4 cm from one end. A “bottleneck” restriction (∼15-µm i.d.; see Figure 2) is created at the exposed section using a laser puller (Sutter Instruments, Novato, CA, model P-2000); the program (Lines 1-3: heat 400; filament 11; velocity 5; delay 127; pull 100. Lines 4 and 5: heat 300; filament 11; velocity 5; delay 127; pull 100) is allowed to cycle once and is then manually terminated. Next, a 2-propanol slurry containing Poros 10 R2 (Perseptive Biosystems, Cambridge, MA, ∼10-µm-diameter particles) packing material is flushed through the FS with a helium bomb (∼700 psi). After 1-2 cm of Poros has packed at the bottleneck, the bomb is depressurized and the slurry replaced with a 0.1% acetic acid. Upon washing with aqueous solution, the Poros plug tends to shrink somewhat and pack tightly into the bottleneck (but not past the “waist” of the restriction). Next, 6-8 cm of 5-µm-diameter C18 reversed-phase beads (YMC, Wilmington, NC) are added from a 60:40 acetonitrile/propanol slurry; total packing time is typically 20 min. Column conditioning is performed by first loading 100 fmol of standard peptide and then running two HPLC gradients (0-70% B in 12 min; A ) aqueous with 0.1% acetic acid, B ) acetonitrile with 0.1% acetic acid). Finally, the column is remounted in the laser puller, and the emitter tip is pulled ∼5 mm from the bottleneck region. Two different tip sizes were used for the work reported herein: MS-only screening was performed with emitter tips of ∼5 µm in diameter (heat 290; filament 0; velocity 5; delay 138; pull 0), while MS/MS data were (67) Beu, S. C.; Laude, D. A. Anal. Chem. 1992, 64, 177-180. (68) Beu, S. C.; Laude, D. A. Int. J. Mass Spectrom. Ion Processes 1992, 112, 215-230.
Figure 2. Assembly used for variable-flow HPLC. Inset shows a detailed view of a microcapillary HPLC column with integrated electrospray emitter tip (see text for description).
collected using tips of