Native Protein MS and Ion Mobility - American Chemical Society


Native Protein MS and Ion Mobility - American Chemical Societypubs.acs.org/doi/pdf/10.1021/ac071878cSimilarby CS Kaddis...

1 downloads 96 Views 397KB Size

Native Protein MS and Ion Mobility:

Large Flying Proteins with ESI Catherine S. Kaddis Joseph A. Loo University of California, Los Angeles

T

Gas-phase methods such as MS and ion mobility spectrometry are emerging tools for structural biology that can measure sizes of large protein assemblies.

he measurement of large biomolecules has benefited tremendously from the development of ESI coupled to gas-phase analyzers such as mass spectrometers and ion mobility spectrometers. The role of multisubunit assemblies and aggregation in normal cellular processes and diseases warrants a practical method for the study of large macromolecular complexes. X-ray crystallography and NMR spectroscopy provide unrivaled high-resolution structural information. However, protein crystallization is traditionally time-consuming; NMR is limited by the size of the protein target; and compared with MS, both methods require large quantities of purified analyte. In the absence of high-resolution structures, ESI MS and ESI ion mobility spectrometry (IMS) can provide crucial functional information about proteins and protein complexes, such as binding affinity constants, assembly states, stoichiometry, and conformational changes (1). These details are comparable to those offered by modern analytical ultracentrifugation techniques, calorimetry (differential scanning and isothermal), chromatographic methods (light scattering and fluorescence), and surface plasmon resonance (SPR). However, immobilization is not required for ESI-MS/IMS analysis, nor is protein modification required for visualization. Furthermore, ESI MS and ESI IMS may be used in isolation or coupled with other separation methods to identify and characterize protein components in heterogeneous mixtures. MS is more generally suited for these measurements because it measures an inherent fundamental property, the molecular mass, of proteins and protein complexes at an accuracy that substantially surpasses other analytical techniques. In this article, we describe the development of ESI-MS/ IMS for characterizing large protein complexes held

© 2007 AMERICAN CHEMICAL SOCIETY

together by noncovalent bonds; we will also highlight unique applications.

Opening the door for largeprotein measurement It was not too long ago when the measurement of a simple, linear peptide of mass 1000 Da was a challenge. In 1986, Frank Field remarked that “the mass region of real interest for proteins is m/z 40,000–100,000, and one can only speculate as to whether such monster gaseous ions can be produced” (2). But John Fenn’s work demonstrated that even a simple quadrupole mass analyzer of m/z 1500 could “weigh” molecules >100,000 Da because on average, a protein will pick up 1 proton per 10 amino acid residues, thereby shifting m/z of all proteins into a very similar and easily measurable range. Once this ESI-based method was developed, it became obvious how these very large molecules or “flying elephants” could be measured reliably and sensitively (3). ESI opened up new applications of MS, such as the analysis of large protein complexes and assemblies. Proteins are involved in molecular recognition, including the recognition of signals by receptors, of substrates and their transition states by enzymes, and of ligands by proteins. This is arguably the most significant molecular event in biology. The ability of ESI to ionize macromolecules without disrupting covalent bonds and to maintain weak noncovalent interactions is a key distinguishing feature of ESI for the study of biological complexes (1, 4, 5). The molecular-mass measurement provides a direct determination of the stoichiometry of the binding partners in the complex, even for multiligand heterocomplexes. Quaternary structure refers to the arrangement of subunits in multisubunit proteins. Intermolecular noncovalent interactions are responsible for agM A R C H 1 , 2 0 0 7 / A N A LY T I C A L C H E M I S T R Y

1779

13+

Ecotin (pH 3) 11+

Ecotin dimer (pH 7)

15+

gregation of folded polypeptide chains into multimers; this determines a protein system’s quaternary structure. Hydrophobicity, van der Waals, ionic and dipole interactions, hydrogen bonding, and bond constraints are many of the forces that contribute to the formation of secondary, tertiary, and quaternary structures and govern how proteins fold and interact with other molecules to form complex structures. ESI MS can measure proteins and their complexes in aqueous solution at near-neutral pH. The protein interactions are often sufficiently retained upon transition to the gas phase so that the size and binding stoichiometry can be measured (6 ). In some circumstances, the secondary and tertiary structures can be probed by gas-phase methods, such as hydrogen–deuterium exchange and ion mobility (7 ). Consider a researcher who has isolated a protein of unknown function or has overexpressed a protein in E. coli and wishes to characterize it. A common practice today is to submit a small amount of the protein to a core MS laboratory for a molecular weight (MW) measurement. By ESI or MALDI, a MW with a precision and accuracy of ±0.05% or better can be measured. However, this depends heavily on the purity of the protein sample, the relative size of the protein, the presence or absence of posttranslational modifications (e.g., phosphorylation or glycosylation), and the resolution of the mass analyzer. The usual course of action is to measure the MW of the denatured protein by unfolding its regular, ordered structure using solvents that are compatible with ESI and MALDI, such as 50:50 water:acetonitrile (v/v) mixtures with organic acids (e.g., acetic acid, trifluoroacetic acid, or formic acid) to lower the pH. Although these conditions can yield intense signals for the protein, information is lost regarding the possible presence of interacting ligands and other proteins and the high-order structure of the protein. This occurs because weak interactions can be disrupted by varying environmental conditions, such as pH, temperature, solvent composition, and the addition of other denaturants (e.g., detergents).

Noncovalent protein interactions In the early days of developing ESI, it became clear that this technique could measure small proteins bound to small molecular ligands. The first report of specific associations probed by ESI MS was written by the Ganem and Henion groups. The intact ~13-kDa receptor–ligand complex between FK binding protein and the macrolides rapamycin and FK506 (8) and the enzyme– substrate pairing between 14-kDa lysozyme and N-acetylglycosamine and its cleavage products were reported (9). Several reports of other biochemical, noncovalently bound systems measured by ESI MS followed shortly, including a ternary complex between the HIV protease dimer bound to a substrate-based inhibitor (10). The noncovalent binding of heme (protoporphyrin IX) to myoglobin was reported by Katta and Chait (11). Dramatic differences are observed in the mass spectra of aqueous solutions with pH 3.35–3.90, as myoglobin is fully denatured at pH 3.35. At pH 3.35, the spectrum showed only ions for the apo- or nonbinding form of the protein. Raising the pH to 3.90 allows the protein to fold properly to the “native” configuration, and non1780

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 7

Streptavidin tetramer

20+

GroES heptamer

21+

Enolase dimer

27+ 1000

3000

5000

Yeast ADH tetramer

7000

6000

8000 10,000 m/z

32.2 kDa 52.2 kDa 72.7 kDa 93.3 kDa 147.6 kDa 50,000 100,000 150,000

20S proteasome 28 -mer

63+

16.1 kDa

12,000

690 kDa 600,000 650,000 700,000 Mass

FIGURE 1. ESI mass spectra of noncovalent protein complexes. The corresponding mass spectra converted to the mass domain are shown on the right. As protein MW increases, the average charge and m/z range increase for the multiply charged molecules. The accuracy of the MW determinations ranges from ±0.02% for the smaller complexes to ±0.1% for the larger complexes.

covalent binding of heme occurs. The effect of solution thermodynamics and the relative stability of the gas-phase complex was studied by the Smith group (12). Ribonuclease S (RNase S) is composed of the 11.5-kDa S-protein hydrophobically bound to the 2-kDa S-peptide. In solution, S-peptide binds to S-protein with a solution dissociation constant Kd ~1–10 nM. The solution temperature dependence on the gas-phase stability correlates with that predicted from thermodynamic parameters (13). The development of low-flow electrospray, or nanoelectrospray (14, 15), down to 10–200 nL/min, played a significant role in the ESI MS of biomolecules in general and in the study of noncovalent complexes. The advantages of nanoliter-perminute analyte flow include not only reduction of the overall consumption of precious sample without compromising signal intensity but also the generation of smaller droplets that result in increased signal levels. This is especially advantageous for solutions with a large amount of water. These early examples established the feasibility of ESI MS and the potential link to the solution-phase world. These papers also helped to move biology, biochemistry, and medicinal chemistry to the forefront of applications for ESI MS technology.

Very large protein systems Heck and van den Heuvel coined the term “native protein MS”, which parallels the technique native protein gel electrophoresis. Native protein MS seeks to characterize proteins in their native, functional form (16). Solvent is evaporated from the ESI-generated microdroplets as they enter the opening of the atmosphericpressure/vacuum interface. Remaining solvent from the droplet and from the protein is removed in this interface region and throughout its journey within the mass spectrometer prior to detection of the protein molecule. The desolvation process requires an input of energy, in the form of thermal energy and collisional heating of the droplet itself. However, this can be detrimental for the measurement of the intact protein complex because of its fragility in the gas phase. Fortunately, for many proteins report-

210

Po Aerosol to DMA

Capillary

+ – 2 to 3 kV DC

Aerosol to detector

Air

0 to – 10 kV DC

Normalized particle counts

Air CO2

[M]1+

100 75 50

[M]2+

25

3+

[M]

[2M]1+

ESI and ion mobility analyzers

ESI IMS methods complement high-resolution ESI MS ESI source DMA techniques and have been used to probe details about proteins and protein assemFIGURE 2. Schematic of ESI GEMMA system. blies. For example, highlight210 Multiply charged molecules generated by ESI are neutralized to singly charged species within the Po ing the potential of IMS as chamber. Proteins in the aerosol are separated and sized within the DMA and detected by the CPC. The GEMMA spectrum on the right is from 66-kDa human serum albumin (7-nm EMD). (Images of the albua structure-elucidation tool, a min structures were generated by WebLab ViewerPro 3.7 from the Protein Data Bank coordinates 1BM0.) number of groups have demonstrated that protein and protein-complex dimensions measured by ESI IMS are ed in the literature, this process is sufficiently gentle to accom- consistent with known crystal structure size and topology modate the MW measurement of a number of large protein (18–22). Demonstrating the practical application of IMS in biocomplexes. (More advanced studies using tandem MS methods logical chemistry research, ESI-MS/IMS data on amyloid -proto selectively dissociate the weaker noncovalent bonds can yield teins supports proposed mechanisms of protein aggregation ininformation on the structure and topology of the protein com- volved in Alzheimer’s disease (23). The application of a unique gas-phase electrophoretic mobility molecular analyzer (GEMMA) plex; 17 .) As smaller protein–ligand complexes are found to be ame- has extended measurements of biological macromolecules to the nable to ESI MS, the ability to access larger MW complexes is megadalton range, thereby expanding the utility of ion mobility tested. As MW increases, multiple charging increases, but the to the study of very large protein complexes (18, 20, 22, 24 –26). Figure 2 shows a GEMMA instrument composed of an ESI multiple charge distribution is shifted to increasing m/z (Figure 1). For the small 32-kDa protease dimer ecotin, the protein is source and a differential mobility analyzer (DMA) to measure denatured in the presence of acetonitrile and by lowering the gas-phase electrophoretic mobility diameter (EMD); a condenpH. As a result, the protein is dissociated in solution into its sation particle counter (CPC) serves as the detector (24, 26). monomer components, and its multiply charged molecules are The commercially available GEMMA system (TSI Inc.) was infound at m/z 1000–2000. Raising the solution pH to near phys- troduced by Kaufman and co-workers to analyze proteins and iological values (and with deletion of acetonitrile) results in a other biological macromolecules (24). Within the electrospray shift in the charge distribution to m/z 2000–3000 and an in- unit, the protein solution is sprayed into a mist of highly charged crease in MW to 32 kDa, which is consistent for the native dimer droplets, which dry and are then charge-reduced by a bipolar complex. In general, the charge distribution for most denatured neutralizer (24, 27 ). A 210Po -particle source ionizes the air proteins measured is in the m/z 800–2500 range, regardless of molecules to produce a high concentration of bipolar (positively their MW. However, ions for the native complexes are found at and negatively charged) ions. Neutralization of the multiply increasing m/z as the size of the ion increases. Thus, the m/z charged ESI ions occurs through the collisions between the localrange of the MS analyzer is an important consideration for such ized bipolar ions in the air and highly charged ESI droplets (27 ). measurements, but several analyzers, such as TOF, can accomMost of the particles will be neutral, a size-dependent permodate this requirement. centage of the particles will possess one charge, and a lesser fracEven the 690-kDa 20S proteasome complex, composed of an tion of particles will display multiple charges. The DMA sepaassembly of 14 -subunits and 14 -subunits, can be measured rates the charged particles on the basis of their electrophoretic by ESI quadrupole-TOFMS (Figure 1). The proteasome is a mobility in air. Particles enter the DMA through a slit in the multicatalytic protease complex expressed in the nucleus and cy- outer cylindrical electrode, which is electrically grounded (24, toplasm of all eukaryotic cells. As a recycler of damaged, mis- 26). The central electrode rod of a coaxial electrode pair is confolded, and short-lived regulatory proteins, the proteasome is the nected to the negative power supply. The particles are subjected central enzyme of nonlysosomal protein degradation (i.e., the to a laminar, coaxial flow of filtered sheath air perpendicular to ubiquitin–proteasome pathway). Our lab has measured the 28- the electric field created by the potential difference between component 20S proteasome complex and the 42-component these two electrodes. Carried by the sheath flow of air, positive20S proteasome–inhibitor complex; a peptide-based inhibitor of ly charged particles are attracted by the inner electrode. Negathe proteasome’s protease activity binds to each of the 14 inner tively charged particles are repelled by the inner electrode, and -subunits noncovalently (18). Very impressive ESI MS results neutral particles exit the DMA with the excess air. Only positivehave been demonstrated for even larger protein complexes, such ly charged particles within a narrow range of electrophoretic moas the 2.3-MDa 70S ribosome from Thermus thermophilus, a bilities are selected for detection by the CPC. Within the CPC, a thermophilic bacterium. Peaks for its multiply charged molecules supersaturated vapor of butanol condenses onto the particles, are found in the m/z 25,000 –27,000 range measured by a which in turn are detected by laser light scattering. If a range of quadrupole-TOF analyzer (5). voltage differences between the two electrodes is scanned, the Sample

0

3

4 5 6 7 8 910 EMD (nm)

M A R C H 1 , 2 0 0 7 / A N A LY T I C A L C H E M I S T R Y

1781

9.2–12 MDa

1200 4.6 MDa

1000

MW (kDa)

800 1.7 MDa

600 400

480 kDa

the MW divided by VEMD and Avogadro’s number N0. (Unlike MW, the values for EMD and VEMD are based on individual particles rather than a molar quantity.)

MW  = MW = VEMD N0  (EMD3) 6

()

For a series of spherical particles that display similar compactness, there will be a linear relationship between MW and volume (24, 30). For the pro0 teins and protein complexes studied by our and 10 20 40 30 EMD (nm) Allmaier’s groups, a correlation exists between EMD and MW (Figure 3; 26). Displaying an average  = 0.6 g/mL, this data set includes proteins as FIGURE 3. EMDs of proteins measured by GEMMA. small as 3 kDa and protein complexes >10 MDa. (Left to right) The protein structures are myoglobin (1WLA), apoferritin (1IER), The analytical performance of the ESI GEMGroEL:GroES complex (1SX4), cowpea chlorotic mottle virus (1CWP), and the ribonucleoprotein vault. Accuracy of MW measurements for large protein complexes by MA system is sufficient to address most issues reGEMMA has been estimated at 1–5% (26). (Image of the vault protein complex is lated to the measurement of protein–protein comreprinted with permission from Ref. 31.) plexes. Nanomolar concentrations of protein species can be detected (32). The detection efficiency is best at the higher end of the sampling electrophoretic mobility distribution of particles from the pro- range, i.e., larger EMD (24, 25). For nonglobular large protein complexes, such as viruses and the  ring of the 20S proteasome tein solution can be determined (24). The GEMMA instrument differs from traditional ion mobility (18, 20, 22), the EMD is in general agreement with crystal strucspectrometers in a number of ways. In conventional IMS instru- ture dimensions and is reproducible. The resolution of GEMMA measurements has been comments, the electric field is parallel rather than perpendicular to the direction of flow (28). Given the higher mobilities of proteins ob- pared with that of size-exclusion chromatography (SEC; 24). Reserved using traditional IMS, a pulsed operation and “relatively solving power for proteins and other biological macromolecules fast timing” are utilized. Particles are separated by GEMMA in studied by GEMMA is ~7 fwhm (18, 22, 24, 26, 29), but imspace rather than in time (24, 28, 29). In addition, ion reactions proved resolution may be realized through use of other DMA within the analyzer are utilized in traditional IMS instruments, models, which potentially offer resolution >100 (29). Although but they have not been observed through use of the DMA (24). the resolving power of GEMMA is lower than that of MS, Particular features of ESI GEMMA assist in the study of large GEMMA’s sensitivity is superior for analysis of very large protein (kilo- to megadalton range) protein assemblies relevant to dis- complexes (i.e., >500 kDa). ease and normal cellular function. Generally, ESI of proteins and protein complexes results in multiply charged particles and a Applications complex MS spectrum. Through charge neutralization, GEMMA GEMMA has been exploited to determine the stoichiometry of allows for the resolution of peaks representing a mixture of pro- subunits within large protein complexes and to estimate the teins and protein complexes (24). Molecular masses of most pro- number of proteins encapsulated within larger macromolecular teins can then be calculated by using empirically determined assemblies. For example, Kaufman et al. determined the stoiequations or estimated values for protein density (25, 26). The chiometry of subunits within the heteromultimeric, hexagonal examples described highlight how large complexes of similar bilayer hemoglobin from the earthworm Lumbricus terrestris; composition can be separated, identified, and quantified by IMS the 3:1 globin–linker ratio was determined from the spectrum of to answer typical biological research questions and to find utility the assembly after it was denatured by SDS (25). beyond the research laboratory. This lab has applied ESI GEMMA to gain insights about the The estimation of protein size is obtained by considering the function of a number of protein complexes, including vaults EMD, which is defined as the diameter of a singly charged sphere (32); the 20S proteasome (18); and ErbB3, which is a member with the same electrophoretic mobility as the particle. The vol- of an epidermal growth factor receptor family of human receptor ume VEMD can be calculated with the formula for the volume of proteins associated with cancer cell proliferation (33). The vault a sphere, assuming that the particles are approximately spherical is the largest known ribonucleoprotein particle found in a variety of eukaryotic species, yet its cellular function remains unknown. (24, 25), so The hollow, ellipsoidal 13-MDa native complex is composed of VEMD =  (EMD3) a short sequence of untranslated RNA and 3 different protein 6 components: MVP (major vault protein), TEP1, and VPARP. The effective density of the particles  can then be calculated as Given the potential involvement of vaults in the transport of ma200

1782

18 kDa

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 7

100 [hHb]1+

0 100 [bHb]1+ 50

0 100

50

0

2

4 6 8 10 20 30 50 EMD (nm)

FIGURE 4. GEMMA spectra of human Hb (hHb), bovine Hb (bHb), and Hemopure (bottom). The average MW of hHb and bHb is 65 kDa; the average MW of Hemopure is 250 kDa (dotted line).

Normalized concentration (%)

terial within the cell, the ability of vaults to encapsulate one of its subunit components, VPARP, after assembly of the 96-subunit MVP shell was demonstrated (32). Differences in the GEMMAdetermined mass of recombinant MVP-only vaults and recombinant VPARP/MVP vaults were used to calculate the number of VPARP subunits per vault complex and confirmed the ability of the complex to incorporate proteins into its interior after shell formation (32). ESI GEMMA was also used to detect differences in complex sizes between the 20S proteasome and the 20S proteasome encapsulating protein substrates. To trap substrates within the proteasome, the proteolytic activity of the complex can be irreversibly inhibited prior to incubation of the complex with the protein substrates. Based on the change in GEMMA-determined MW, an average of 4.5 substrate proteins (myoglobin) was estimated to be sequestered within the complex (18). High-resolution MS data have since shown that the 20S proteasome from Thermoplasma acidophilum can sequester a maximum of 3– 4 substrate proteins of similar size (34). Hemopure (Biopure Corp.) is a hemoglobin (Hb)-based oxygen carrier that has been investigated for its potential as a red blood cell transfusion substitute (35). With an average MW of 250 kDa, Hemopure is a glutaraldehyde-linked polymer of bovine Hb. In South Africa, it has been approved for the treatment of adults who are facing surgery and are acutely anemic, to decrease or delay the need for allogenic blood transfusions. However, Hb-based blood substitutes such as Hemopure have also been included on the World Anti-Doping Agency Prohibited List in response to use of the substances to enhance oxygen transfer and bolster endurance among athletes. A number of methods, including LC/MS, gel electrophoresis, and SEC, to detect Hemopure in blood have appeared in the literature (36–39). Potentially, ESI GEMMA could be used to differentiate cross-linked Hb from the natural form. Figure 4 shows GEMMA differentiation of human Hb, bovine Hb, and Hemopure. An advantage of GEMMA over gel electrophoresis and SEC/HPLC is the short (2– 4-min) analysis time. However, additional experimentation with immunodepletion and other methods to treat plasma samples prior to analysis is required before the potential of IMS/GEMMA to detect and quantify Hemopure and other Hb-based substitutes can be explored. ESI GEMMA has also been used to assess the risk of cardiovascular disease by profiling lipoproteins. A patient’s risk for cardiovascular disease can be determined by a number of factors, including relative plasma concentrations and particle size distributions of high-density (HDL) and low-density (LDL) lipoproteins (40, 41). LDL particles, classified as proatherogenic, participate in the delivery of cholesterol from the liver to peripheral cells; HDL particles are considered antiatherogenic because they participate in the reverse cholesterol-transport system from peripheral cells to the liver. Within each subspecies, HDL or LDL, particles can be further categorized by their size, density, charge, and chemical composition. Because ESI GEMMA can quantify intact noncovalent biological complexes, it has been applied to plasma lipoproteins for medical diagnostics (42). Figure 5 shows GEMMA spectra of

Normalized particle counts (%)

50

HDL

100

LDL IDL VLDL

75 50 25 0 2

4

6 8 10

20

30 40 50

EMD (nm)

FIGURE 5. GEMMA spectra of plasma lipoprotein fractions from patients who are normal (solid line), have been diagnosed with coronary heart disease (CHD; dotted line), and are at risk for CHD (dashed line). HDL particles (density range 1.063–1.210 g/mL) exhibit a diameter range of 4–10 nm; LDL particles (density range 1.019–1.063 g/mL) exhibit a diameter range of 19–23 nm; intermediate density lipoproteins (IDL; density range 1.006–1.019 g/mL) exhibit a diameter range of 22–24 nm; and very LDL (VLDL) particles (density range 0.95–1.006 g/mL) exhibit a diameter range of 25–70 nm. Density and diameter ranges are based on techniques other than IMS, such as analytical ultracentrifugation and gel electrophoresis, and vary among laboratories (43 ).

M A R C H 1 , 2 0 0 7 / A N A LY T I C A L C H E M I S T R Y

1783

plasma-derived lipoprotein fractions from patients who are normal, diagnosed with coronary heart disease (CHD), and at risk for CHD. HDL particles from normal patients display a bimodal distribution, whereas HDL from patients diagnosed with CHD and at risk for CHD display a unimodal distribution (42). In addition, the modal diameter of the LDL peak decreases with increasing patient CHD risk. The modal diameter is highest for the normal patient (EMDLDL = 22.5 nm), lowest for the patient with CHD (EMDLDL = 20.9 nm), and intermediate for the patient at risk for CHD (EMDLDL = 21.7 nm). This latter observation is consistent with other lipoprotein quantification techniques. Small, dense LDL particles are more likely to contribute to arterial lipid deposition than large buoyant LDL particles and are associated with a high risk for myocardial infarction (40, 43).

clude protein MS, proteomics, and disease biomarker discovery. Address correspondence about this article to Loo at 405 Hilgard Ave., Paul D. Boyer Hall, Molecular Biology Institute, University of California at Los Angeles, Los Angeles, CA 90095 ([email protected]).

References (1) (2) (3) (4) (5) (6) (7)

Size matters Methods based on the combination of ESI with MS and GEMMA will continue to be refined and developed further for the characterization of large protein macroassemblies. Ion mobility techniques for measuring collision cross sections of protein complexes can be used to provide information on shape and architecture (21). Tandem MS will continue to play an important role in defining the sizes of large gas-phase protein oligomers, especially from heterogeneous mixtures, perhaps to sizes >2 MDa (17, 44). New ion activation and dissociation methods, such as electron capture dissociation, may be used as a probe of the gas-phase noncovalent complex or to determine the site of ligand binding to a protein target (6, 45). Although lower in resolving power and mass accuracy compared with ESI MS, ESI GEMMA’s simple design (no vacuum required), ease of use, and MW range (>50 MDa) will continue to complement MS measurements. GEMMA could in principle replace other biophysical techniques that are used to measure large protein assemblies and that require much greater amounts of sample, such as light scattering and ultracentrifugation. Applications resulting from the use of ESI will continue for some time to come, and an important biomedical application will be to propel large protein complexes into mass spectrometers and ion mobility devices for measuring their sizes. Field’s “monster gaseous ions” and Fenn’s “flying elephants” generated by ESI will find longevity within the scientific community.

(8) (9) (10) (11) (12) (13) (14) (15) (16)

We are grateful for the support for the UCLA Functional Proteomics Center provided by the W. M. Keck Foundation, and the funding of our research by the National Institutes of Health (RR 20004 to JAL). CSK was supported by the UCLA-NIH Chemistry–Biology Interface Training grant. We acknowledge research contributions from Shirley Lomeli, Sheng Yin, Marcin Apostol, and Beniam Berhane (UCLA) and our collaboration with L. H. Rome (UCLA) and Mario Thevis (German Sport University, Cologne). Appreciation goes to Gary Hewett (Berkeley HeartLab) for information about lipoprotein diameters, for providing the lipoprotein fractions, and for reviewing the manuscript, and to Stanley L. Kaufman (TSI) for additional advice.

(17) (18) (19) (20) (21) (22) (23) (24) (25) (26) (27) (28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39) (40) (41) (42)

Catherine S. Kaddis is a graduate student at UCLA. Joseph A. Loo is a professor in the department of chemistry and biochemistry and in the David Geffen School of Medicine at UCLA. His research interests in-

(43) (44) (45)

1784

A N A LY T I C A L C H E M I S T R Y / M A R C H 1 , 2 0 0 7

Loo, J. A. Mass Spectrom. Rev. 1997, 16, 1–23. Field, F. In Mass Spectrometry in the Analysis of Large Molecules ; McNeal, C. J., Ed.; John Wiley & Sons: Chichester, U.K., 1986, pp 213–214. Fenn, J. B. Angew. Chem., Int. Ed. 2003, 42, 3871–3894. Heck, A. J. R.; van den Heuvel, R. H. H. Mass Spectrom. Rev. 2004, 23, 368–389. McKay, A. R.; et al. J. Am. Chem. Soc. 2006, 128, 11,433–11,442. Xie, Y.; et al. J. Am. Chem. Soc. 2006, 128, 14,432–14,433. Valentine, S. J.; Clemmer, D. E. J. Am. Soc. Mass Spectrom. 2002, 13, 506–517. Ganem, B.; Li, Y. T.; Henion, J. D. J. Am. Chem. Soc. 1991, 113, 6294–6296. Ganem, B.; Li, Y. T.; Henion, J. D. J. Am. Chem. Soc. 1991, 113, 7818–7819. Baca, M.; Kent, S. B. H. J. Am. Chem. Soc. 1992, 114, 3992–3993. Katta, V.; Chait, B. T. J. Am. Chem. Soc. 1991, 113, 8534–8535. Ogorzalek Loo, R. R.; et al. J. Am. Chem. Soc. 1993, 115, 4391–4392. Goodlett, D. R.; et al. J. Am. Soc. Mass Spectrom. 1994, 5, 614–622. Emmett, M. R.; Caprioli, R. M. J. Am. Soc. Mass Spectrom. 1994, 5, 605–613. Wilm, M.; Mann, M. Anal. Chem. 1996, 68, 1–8. van den Heuvel, R. H. H.; Heck, A. J. R. Curr. Opin. Chem. Biol. 2004, 8, 519–526. Benesch, J. L. P.; Robinson, C. V. Curr. Opin. Struct. Biol. 2006, 16, 245–251. Loo, J. A.; et al. J. Am. Soc. Mass Spectrom. 2005, 16, 998–1008. Clemmer, D. E.; et al. J. Am. Chem. Soc. 1995, 117, 10,141–10,142. Thomas, J. J.; et al. Spectroscopy 2004, 18, 31–36. Ruotolo, B. T.; et al. Science 2005, 310, 1658–1661. Hogan, C. J.; et al. Anal. Chem. 2006, 78, 844–852. Bernstein, S. L.; et al. J. Am. Chem. Soc. 2005, 127, 2075–2084. Kaufman, S. L.; et al. Anal. Chem. 1996, 68, 1895–1904. Kaufman, S. L.; et al. Anal. Biochem. 1998, 259, 195–202. Bacher, G.; et al. J. Mass Spectrom. 2001, 36, 1038–1052. Scalf, M.; et al. Science 1999, 283, 194–197. Jarrold, M. F. Annu. Rev. Phys. Chem. 2000, 51, 179–207. de la Mora, J. F.; Ude, S.; Thomson, B. A. Biotechnol. J. 2006, 1, 988–997. Ude, S.; et al. J. Am. Chem. Soc. 2004, 126, 12,184–12,190. Mikyas, Y.; et al. J. Mol. Biol. 2004, 344, 91–105. Poderycki, M. J.; et al. Biochemistry 2006, 45, 12,184–12,193. Kani, K.; et al. J. Biol. Chem. 2005, 280, 8238–8247. Sharon, M.; et al. J. Biol. Chem. 2006, 281, 9569–9575. Anbari, K. K.; Garino, J. P.; Mackenzie, C. F. Eur. Spine J. 2004, 13, S76–82. Thevis, M.; et al. W. Anal. Chem. 2003, 75, 3287–3293. Lasne, F.; et al. Clin. Chem. 2004, 50, 410–415. Guan, F.; et al. Anal. Chem. 2004, 76, 5127–5135. Varlet-Marie, E.; et al. Clin. Chem. 2004, 50, 723–731. Rizzo, M.; Berneis, K. QJM 2006, 99, 1–14. Meyers, C. D.; Kashyap, M. L. Curr. Opin. Cardiol. 2004, 19, 366–373. Benner, H. W.; Krauss, R. M.; Blanche, P. J. World Intellectual Property Organization WO 03/042704 A1, 2003. Tulenko, T. N.; Sumner, A. E. J. Nucl. Cardiol. 2002, 9, 638–649. van den Heuvel, R. H. H.; et al. Anal. Chem. 2006, 78, 7473–7483. Geels, R. B. J.; et al. Anal. Chem. 2006, 78, 7191–7196.