PC66CH20-Robinson

ARI

V I E W

9:53

Review in Advance first posted online on January 14, 2015. (Changes may still occur before final publication online and in print.)

A

N

I N

C E

S

R

E

7 January 2015

D V A

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

Mass Spectrometry of Protein Complexes: From Origins to Applications Shahid Mehmood,∗ Timothy M. Allison,∗ and Carol V. Robinson Department of Chemistry, University of Oxford, Oxford OX1 3QZ, United Kingdom; email: [email protected]

Annu. Rev. Phys. Chem. 2015. 66:453–74

Keywords

The Annual Review of Physical Chemistry is online at physchem.annualreviews.org

noncovalent complexes, membrane proteins, ionization mechanisms, protein-lipid interactions, ion-mobility mass spectrometry

This article’s doi: 10.1146/annurev-physchem-040214-121732 c 2015 by Annual Reviews. Copyright  All rights reserved ∗

These authors contributed equally to this work.

Abstract Now routine is the ability to investigate soluble and membrane protein complexes in the gas phase of a mass spectrometer while preserving folded structure and ligand-binding properties. Several recent transformative developments have occurred to arrive at this point. These include advances in mass spectrometry instrumentation, particularly with respect to resolution; the ability to study intact membrane protein complexes released from detergent micelles; and the use of protein unfolding in the gas phase to obtain stability parameters. Together, these discoveries are providing unprecedented information on the compositional heterogeneity of biomacromolecules, the unfolding trajectories of multidomain proteins, and the stability imparted by ligand binding to both soluble and membrane-embedded protein complexes. We review these recent breakthroughs, highlighting the challenges that had to be overcome and the physicochemical insight that can now be gained from studying proteins and their assemblies in the gas phase.

453

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

STATE-OF-THE-ART MASS SPECTROMETRY OF BIOMACROMOLECULES

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

The recent landmark publication of a draft map of the human proteome is the result of several developments in mass spectrometry applied to proteomics, in which peptides are sequenced to identify proteins in databases (1). The widespread uptake of these developments for proteomics, together with continued improvement in instrumentation, made this milestone possible. With this draft map of the human proteome, many questions arise concerning interaction partners, cofactors, complexes, and substrates. Developments in the mass spectrometry of noncovalent complexes, which have occurred in parallel with those in proteomics, although much less widespread, can be applied to define these interaction partners. Although not yet at the level of high-throughput implementation, or of the required sensitivity to obtain information on low-abundance protein complexes, the progress of these so-called top-down methods, in which intact protein assemblies are directly interrogated, is advancing rapidly. It is now possible to use mass spectrometry to measure the stoichiometry, assembly, and topology of protein complexes isolated directly from cells. This is achieved by first determining the masses of the component subunits, then forming many subcomplexes by disruption of the intact assembly in the solution and gas phases, and subsequently using network algorithms to link these subunits and subcomplexes into multiprotein assemblies (2, 3). In such studies, it is also often possible to link changes in the stability of the protein complexes with the extent of post-translational modification. For example, the unstructured C-terminal regions of U1-snRNP were removed prior to crystallization of the recombinant complex but are phosphorylated in vivo and were found to affect subunit interactions within the endogenous complex (4). For real-time assembly reactions, the ability to define a complete readout of all components in an equilibrating mixture enables the capture of various intermediates. For homomeric assemblies, differences in mass can be introduced through the use of labeled subunits, either isotopically labeled or tagged. With regard to polydisperse assemblies, the mass-to-charge (m/z) ratios of the multimers do not coincide, as might be anticipated for multiple copies of the same protein, because the subunit interfaces bury charge and surface area upon complex formation, and thus multimers can be readily separated on an m/z scale. Investigating the topology of protein assemblies has become possible only relatively recently through the introduction of ion-mobility mass spectrometry. This development is adding a new dimension to the study of protein assemblies, particularly with respect to modeling novel topological arrangements of subunits (5) and to observing conformational changes in response to various stimuli (6). As far as instrumentation is concerned, the mass spectrometers used for the characterization of intact proteins consist of three parts: the ion source, analyzer, and detector. Optimal instrument development and design were initially focused on the analysis of small molecules. Applications to larger biomolecules, such as proteins, were dependent on several advances, including the development of soft-ionization methods, modifications of time-of-flight (TOF) mass spectrometers for high-mass transmission, and high-mass quadrupoles (7–9). These developments are ongoing, with the very recent modification of an Orbitrap instrument for the noncovalent mass spectrometry of large complexes, such as whole virus capsids, and for the characterization of antibodies (10–12). In addition to modifications to mass spectrometer hardware, improvements in sample isolation and preparation have played an important role, opening up unheralded opportunities for the characterization of membrane proteins (13, 14). Mass spectrometry has also been successfully coupled with ion mobility, which allows the separation of ions on the basis of charge and shape inside the mass spectrometer and additionally enables the measurement of the rotationally averaged collision cross section (CCS) (15). Ion-mobility mass

454

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

spectrometry has therefore been applied to examine the functional mechanism of complex biological systems, for example, the conformational flexibility of rotary ATPases (16), the intermolecular processes that underlie amyloid formation (17, 18), the gas-phase structure of intrinsically disordered proteins (19, 20), and the effect of ligand binding on protein stability (14, 21, 22).

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

DIFFERENT MECHANISMS OF ION FORMATION As all mass spectrometry experiments depend critically on the ionization of the molecule of interest, it is important to consider the different ionization techniques used for biomacromolecules. Several high-energy methods are available to generate ions for small organic samples, which are naturally volatile. In the case of intact proteins, however, it is common to generate gas-phase ions directly from solution-phase molecules using so-called soft-ionization methods, which avoid undesired fragmentation and untoward alteration of the protein structure. Intact gas-phase molecular ions, ranging from a few daltons to megadaltons, can be produced using these soft-ionization techniques, which can be divided into two categories, energy transfer or high electric field, based on the strategy employed. Methods in the first category use the activation of an analyte by energy transfer, for which several methods exist, such as fast atom bombardment, secondary ionization mass spectrometry, and plasma desorption (23–25). Although early examples showed that these energy-transfer techniques are efficient in ionizing bioorganic molecules, ionization and transfer of proteins were not achieved. These energy-transfer techniques have most commonly been supplanted by the use of matrixassisted laser desorption ionization (MALDI) for biomacromolecule analysis.

Matrix-Assisted Laser Desorption Ionization MALDI mass spectrometry uses a rapid photovolatilization of biomolecules embedded in an ultraviolet-absorbing matrix followed by mass analysis. The mechanism of ion formation in MALDI mass spectrometry is not well understood, but it is generally considered a two-step process. First, activation of the matrix gives rise to a hot plume carrying highly ionized matrix ions, and then the analyte in the hot plume is ionized by the highly activated matrix ions (26). Notably, MALDI mass spectrometry tends to produce predominantly singly charged molecular ions, although in some cases doubly and triply charged ions are also observed. This low charging is an advantage when interpreting mass spectra containing a complex mixture of components. MALDI is also more tolerant to the presence of contaminants compared to electrospray ionization (ESI). There are a few examples in which MALDI, after coupling with liquid chromatography (27), has been used to study mixtures of peptides for protein identification; however, for liquid chromatography mass spectrometry and tandem mass spectrometry analysis, ESI is the dominant choice.

High Electric Field Ionization: Electrospray Ionization The alternative ionization method to energy transfer is the use of a high electric field, commonly implemented as ESI. In a standard ESI setup, a high electrical potential is applied on a capillary containing the solution for study, causing an accumulation of positive ions (in the positive-ion mode) at the tip of the capillary, which in turn promotes the evolution of charged droplets carrying the precursors of the charged gas-phase ions (Figure 1a). The confined nature of the ions inside the limited space of the droplet causes electrostatic repulsion leading to droplet fission, ultimately yielding the gas-phase ions. Solvent evaporation is also essential as this promotes the Coulombic repulsion between ions as they become increasingly confined in the shrinking droplet (28). www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

455

PC66CH20-Robinson

ARI

7 January 2015

9:53

Instrument entry

a Capillary +++

+++

Sample +++

+++

+++

++++++ +++

Droplet fission



High voltage

+ +

+

+ +

Droplet

+ + +

+

Protein

+

+

+

+

+

+ +

+ + + +

+

Chain ejection model

+ + + + + +

++

+ + +

+ + +

+ + + +

+

+ + +

+

+

+

+

+

+

+

+

+

Disordered protein in droplet

+ + +

+

+ + +

++

+

+

+ + +

+ + ++ + + + + + + + ++ +

+ + +

+

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

+ + + + + + + + + ++ + + + ++ + ++ + ++ + + + + ++ + ++ + +++ ++ + + + + + ++ + + + + ++ + +

Charge residue model

+

c

+ ++ + + + + +

Droplet formation

+

b

+++

+++

Protein chain ejecting from droplet

Figure 1 (a) In the electrospray ionization process, charged droplets are emitted from a highly charged capillary. Through fission and evaporation, charged ions of the macromolecule are produced. (b) The charge residue model suggests that the charge of an ion (orange sphere) is inherited from the charge on the original droplet (blue shell ). (c) Disordered proteins tend to gain high charge after ionization, accounted for by the chain ejection model in which the protein (orange) is emitted from the droplet (blue).

One of the advantages of ESI for the mass spectrometry of proteins is that it can generally be applied to a solution of the protein of interest and can therefore be directly coupled to liquid chromatography (29, 30). In addition, electrospray generates ions with multiple charge states; thus, only low-mass-range analyzers are required for their study (see the sidebar Charge States and Electrospray Ionization). Although simple in concept and in implementation, the mechanisms proposed for ion formation in ESI are often controversial and constantly evolving. Of the many mechanisms proposed, the 456

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

CHARGE STATES AND ELECTROSPRAY IONIZATION

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

The ESI process gives rise to multiply protonated protein ions (in the positive-ion mode). These ions exist in a charge-state distribution, which is largely influenced by the folded state of the protein in the solution phase. Unfolded proteins give rise to higher and broader charge-state distributions, whereas folded proteins give relatively narrower distributions with lower average charge (104). This observation led to the use of mass spectrometry as an analytical tool to probe the solution states of proteins. In the case of partially unfolded proteins, a bimodal charge-state distribution consisting of both high and low average charge distributions is observed (104).

ion evaporation model (31–33) and charge residue model (CRM) (34, 35) are the most widely employed, and recently a chain ejection model for unfolded proteins has been put forward (30). The ion evaporation model is thought to be appropriate for small ions. For larger ions, such as proteins, the CRM and the chain ejection model are considered more accurate than the ion evaporation model (36).

The Charge Residue Model According to the CRM, a droplet carrying a single protein shrinks by evaporation and fission until the complete removal of solvent, with the droplet charge transferred concomitantly to the protein (Figure 1b) (36). An important metric for the CRM is the Rayleigh-limit hypothesis (36), which suggests that the charge state of a protein is limited to the maximum charge on the precursor droplet. A derivative of the CRM model, referred to as the charge carrier field-emission model (37), incorporates the ion evaporation model to allow for the expulsion of charge-carrying ions from the droplet through field emission. This assists in explaining deviation from the charge predicted from the Rayleigh limit.

The Chain Ejection Model Although the CRM describes the charge states for folded proteins from ESI, it may not account for how intrinsically disordered and partially hydrophobic proteins acquire higher charge states. A chain ejection model has been proposed to account for this phenomenon (30): These proteins are suggested to migrate to the surface of the droplet owing to the exposure of hydrophobic residues, which make it unfavorable to reside in the interior of the droplet. The end of the protein chain is expelled into the vapor phase, followed by the ejection of the rest of the chain until it completely separates from the droplet (Figure 1c). Charge equilibration during chain ejection accounts for the high charge states.

The Process of Nano-Electrospray Ionization Nano-electrospray ionization (nano-ESI) is a variant of ESI that requires microliter volumes of a sample as opposed to the milliliters required for standard electrospray. The method is also more tolerant to the presence of salts in the protein solution and can be used with aqueous solutions without the addition of volatile organic solvents, normally necessary for standard electrospray. The size of the electrospray droplet for ESI is in the micrometer range, whereas it is believed to be at least 10 orders of magnitude lower for nano-ESI (38). Given that ESI relies on the evaporation of solvent during the ion formation process, nano-ESI by design requires less harsh conditions to www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

457

PC66CH20-Robinson

ARI

7 January 2015

9:53

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PROTEIN INTERACTIONS IN THE GAS PHASE The interactions of proteins and their complexes in aqueous buffer solutions are stabilized by hydrogen bonding, van der Waals forces, and ionic interactions. These combine to maintain the folded state of a protein and are particularly important interactions for multimeric protein complexes. What happens to protein interactions as they are dehydrated in transition to the gas phase has been addressed by many researchers. Early investigations examined the complex formed between acyl-CoA binding protein and a series of substrate derivatives and protein variants. These studies showed that the interactions between a protein and its ligand are preserved in the gas phase and by the mutation of active-site residues, which resulted in changes to the complex formation, confirming that the interactions were specific and ligand dependent (105). Work by Klassen and colleagues (106) showed the hydrophobic interactions between a protein and its ligand were retained in transition to the gas phase and, in addition, using ion-mobility measurements, that the protein maintains a folded conformation despite the requisite loss of the aqueous environment. These studies provide a rationale for the use of native mass spectrometry for various applications, including determining the ligand-binding affinities of both soluble and membrane proteins (107–109).

evaporate the solvent and can be applied directly to aqueous solutions. This makes it ideally suited for the mass spectrometry of proteins and their complexes in buffered aqueous solutions in which noncovalent interactions are maintained. The smaller droplet size formed in nano-ESI also helps prevent nonspecific interactions from forming between proteins and between proteins and small molecules. Because the smaller droplets generated in nano-ESI require fewer and lower energy collisions to desolvate, the internal energy of the resulting protein ions is therefore reduced. This is of particular relevance for so-called native mass spectrometry, in which the noncovalent interactions of proteins need to be preserved (29) (see the sidebar Protein Interactions in the Gas Phase). Nano-ESI also produces higher-quality mass spectra compared to conventional ESI, which has lower resolution peaks because of the incomplete removal of adducts. Spectra of the GroEL complex clearly show the differences between the two techniques: A bimodal Gaussian charge-state distribution and nonspecific higher-order oligomer were observed in a conventional ESI mass spectrum but not in its nano-ESI counterpart, in which only one stoichiometry was observed: the 14-mer with high resolution and a single Gaussian charge-state distribution (29).

MASS ANALYZERS FOR LARGE PROTEINS AND THEIR ASSEMBLIES Time-of-Flight Mass Analyzers The ions generated by either ESI or MALDI are separated, based on their m/z ratios, by magnetic or electric fields within mass analyzers, before their detection. TOF mass analyzers are commonly used for the characterization of protein complexes owing to their effectively unlimited mass range (39, 40). After emission from the ion source, ions are focused within electrostatic plates, which control their movement and extraction. The function of these plates is important for ESI because the continuous flow of ions formed can be released in packets, necessary for the function of TOF analyzers. TOF analyzers are based on the principle that the time taken by ions to travel a known field-free distance, under high vacuum, is directly related to their m/z ratios (41). The addition of a reflector at the end of the flight tube instead of the detector helps correct the difference in the kinetic-energy distribution of the same m/z ions that would otherwise lead 458

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

ARI

7 January 2015

9:53

to peak broadening and decreased resolution (42). TOF analyzers are generally hybridized with a quadrupole mass analyzer, which acts as a mass filter, and a collision cell for tandem mass spectrometry purposes. In native mass spectrometry, the collision cell is used for the dissociation of protein complexes and, for example, in the case of membrane proteins for the removal of bound detergent (see the section Mass Spectrometry of Membrane Proteins). One problem with the mass spectrometry of noncovalent protein complexes is their tendency to dissociate owing to their high internal energy when transferred from the solution to gas phase (9). This obstacle is overcome by lowering the internal energy of ions by collision with an inert gas such as argon; the transmission of high-mass protein complexes is enhanced by increasing the pressure in the ion source and increasing the vacuum at various stages during the flight path (43). The success of this approach was demonstrated by the first spectrum of an intact GroEL complex with a molecular mass of 800 kDa (44) and very recently by a study of an 18-MDa virus assembly (45). Further development involved a custom-built radiofrequency (RF) generator with reduced frequency of the quadrupole to allow mass selection of greater than 32,000 m/z (46). One of the advantages of this quadrupole time-of-flight (Q-TOF) type of instrument is its ability to couple with ion mobility to obtain information on the shape of proteins, which can be used to build functional and structural models (5, 47–49) (see the section Ion-Mobility Mass Spectrometry). Following the realization that the modified Q-TOF technology could be applied to noncovalent protein complexes, the technique began to be used in the study of protein-protein and proteinligand interactions (17, 50), particularly in cases in which conventional biophysical approaches are less successful. Examples include polydispersity in α-crystallins (51), the assembly pathway of the small heat-shock proteins (52), and the relationship between protein quaternary structure evolution and function (53).

Alternative Mass Analyzers Since the development of these mass spectrometry methods for noncovalent protein complexes, QTOF mass analyzers have been commonly used because these instruments are readily customized, such as for modifications to RF generators to promote the selection and transmission of higher m/z ratios (39, 46). A few reports describe the mass spectrometry of protein complexes performed with a hybrid mass spectrometer, which combines a quadrupole and collision cell for mass selection and fragmentation with Fourier transform ion cyclotron resonance. This was used in a top-down electron-capture dissociation study for improved sequence coverage of protein assemblies (54, 55) and to map the location of a ligand-binding site (56, 57). The quest for high-resolution mass analyzers to differentiate the diversity between polydisperse higher-order oligomers such as antibodies is a long-term objective. Although it is not clear if the issue is one of desolvation or mass resolution (40), it is apparent that the Q-TOF configuration of mass analyzers provides limited resolution for noncovalent complexes, which likely derives from incomplete desolvation. One answer to the lack of resolving power involves the implementation and customization of Orbitraps for the analysis of protein complexes. This technology employs a novel mass analyzer based on the orbital trapping of ions in an electrostatic field (58). The image current produced by the movement of ions around a spindle-like central electrode is monitored and transformed into an m/z ratio using a Fourier transform (58). Introduced as a possible substitute for Fourier transform ion cyclotron resonance with relatively low cost and maintenance, Orbitraps have provided good accuracy and high resolving power for low-m/z ions and are widely employed in proteomics (59, 60). Recently, it was shown that with modifications to the instrument, an Orbitrap with a quadrupole mass filter is capable of analyzing protein complexes in their native www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

459

PC66CH20-Robinson

ARI

7 January 2015

9:53

states (61). Results demonstrated that these instruments have the potential to preserve noncovalent interactions in some instances, with better resolving power than alternative instrument designs. Customization of the Orbitrap RF generator was required, together with efficient desolvation of droplets in the source region, to achieve this enhanced resolution. Trapping of ions for longer time intervals within the higher-energy collisional dissociation cell also improved the removal of buffer molecules. In contrast to Q-TOF mass spectrometers, increased pressure within the source region showed no significant effect on ion transmission (61, 62). Although still early in development, and by no means applicable to a wide range of systems, the potential of Orbitraps to contribute to the mass spectrometry of high-mass complexes is immense.

OBTAINING STRUCTURAL INFORMATION FROM GAS-PHASE IONS Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

Collision-Induced Dissociation An important aspect of the mass spectrometry of protein complexes is the need to carefully balance the energy used to activate the ions, the sufficient energy to remove residual bound molecules (e.g., solvent or salt), and that at which the protein begins to undergo unwanted events, such as unfolding, dissociation, or fragmentation. The most common method of activation is collisioninduced dissociation (CID) (63), in which ions are accelerated by a voltage potential through a cell containing neutral gas atoms or molecules (64). The collisions with the gas convert some of the ion’s kinetic energy into internal modes (65, 66). Depending on the magnitude of the energy increase, the outcome can be any step along an energy dissipation pathway (63). Briefly, from lowest to highest energy, the events are so-called ion cleaning, conformational rearrangement, dissociation, and fragmentation. For noncovalent mass spectrometry, in which noncovalent interactions are to be retained, it is then only necessary to ensure that the energy is not too great to cause the dissociation of a protein complex. However, for nondenaturing native mass spectrometry, in which the folded conformations of proteins need to be preserved, a much finer balance is required to sufficiently clean the ions but not induce a conformational rearrangement, such as unfolding or collapse. Gas-phase activation can, however, be exploited to obtain information about the quaternary structure of protein complexes. Ions can be carefully activated to give sequential dissociation of subunits in which the dissociation pathway provides information about the subunit arrangement in the protein complexes or the connectivity of a subcomplex within a supercomplex. For example, for the eukaryotic initiation factor 3, gas-phase analyses revealed the exact stoichiometry of the complex (67). CID in combination with in-solution disruption led to an architecture of the complex that defined the intersubunit connections.

Surface-Induced Dissociation Surface-induced dissociation (SID) is an alternative activation method emerging as a useful approach to probe the quaternary structure of protein complexes (68). Instead of multiple collisions with inert gas molecules, as is the case for CID, SID is a single-collision event, in which collision with a surface results in disruption of the complex. Consequently, whereas CID occurs on a timescale of microseconds, SID occurs on a much shorter picosecond timescale. One application of SID is to pursue dissociation pathways, alternate to those observed for CID, for the dissociation of protein complexes in the gas phase. Dissociated monomers from protein complexes disrupted by SID are less charged than those ejected by CID and also tend to be more compact and folded-like, which is rarely observed for CID (69, 70). The foundation 460

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

for this difference is the rapid timescale and single-collision nature of SID, which causes ejection of subunits with their folded state preserved (69). In a recent example, SID was applied to a large-molecular-weight protein complex, GroEL, resulting in an atypical protein dissociation pathway in which tetradecamer complexes dissociated into a heptamer, as well as a dimer and monomer. This illustrates the potential of SID for investigating the quaternary structure of large protein complexes. The technology to carry out SID-coupled mass spectrometry is still under development, but results so far are promising (71).

Ion mobility is an established technique in its own right (72–74), but it has also been coupled with mass spectrometers to produce instruments in which protein ions can be separated based on their charge and shape, adding an extra dimension to mass spectra (48, 75–77). This is useful not only for helping to deconvolute more complex spectra, but also for interrogating the composition and topology of protein complexes (48, 77–79). There are two types of drift tubes commonly used for protein ion-mobility mass spectrometry. The first type separates ions using a potential along the drift tube, and the second type uses a traveling wave. In both types, a bath of neutral gas provides resistance to the ions as they travel through the mobility cell; the larger the ion is, the longer it will take to traverse the tube and therefore the greater the arrival time (Figure 2).

a

Traveling wave/voltage potential

Inert counter gas flow Drift region

Ion packet Ions begin to separate by shape and charge

Mobility-separated ions arrive ready for time-of-flight measurement

b

Intensity

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

Ion-Mobility Mass Spectrometry

Arrival time distribution Figure 2 (a) The ion-mobility technique can separate different ions based on their shape and charge. Ions are transmitted by the influence of an electric field through an inert gas–filled drift tube. Small and compact ions travel faster than larger, more elongated ions. (b) The time taken by different ions to move through the drift tube is a function of their mobility; the profile is called an arrival time distribution. www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

461

ARI

7 January 2015

9:53

Ion-mobility mass spectrometry can not only separate ions based on size and charge, but can also measure the size of the ions themselves, as the rotationally averaged CCS. For drifttube instruments, this is a direct measurement using the Mason-Schamp equation (76, 80). For traveling-wave instruments, it is not possible to directly measure the CCS because the travelingwave transport of ions is not well understood. Instead, a calibration can be formed that relates the CCS values measured on a drift-tube instrument to the arrival times measured under the same traveling-wave instrument conditions as the unknown (81). Theoretical CCS values can be calculated from protein structure models (82, 83), for example, from crystal structures, which can then be compared to experimental values. These theoretical calculations are based on a variety of methods, the most common of which is the projection approximation. Although these theoretical techniques are generally accurate for relatively small molecules, a correction factor is usually applied for larger structures, such as proteins (84). However, care must be taken when comparing experimental to theoretical values. An inherent assumption is that the experimental setup is carefully designed, such that the ion is not activated beyond cleanup. That a mass spectrometry experiment is performed in such a way to preserve noncovalent interactions is no guarantee that an excess of energy has not been supplied, causing the protein, or a population of ions, to collapse or unfold. Another consideration is that the theoretical value is appropriately calculated. For example, many protein models may be incomplete and also may not account for the flexible or dynamic behavior of a protein. In these cases, computer simulations can play an important role in better defining likely models of the protein. Furthermore, as the technique is a single-number measurement, in scenarios in which the experimental and theoretical CCS values match, it is difficult to distinguish between coincidence and the fact that the ion persists in a desired conformation. This is particularly a problem when trying to distinguish different protein conformations; given the predisposition of ions to be very sensitive to structural rearrangements in the gas phase, it is tenuous at best to ascribe observed size differences to biologically relevant conformational change above the possibility of having observed a different form due to gas-phase collapse, unfolding, or both. There is a distinct limitation of interpretation that must be rigorously applied by those practicing the technique. The term native mass spectrometry is generally interpreted as meaning mass spectrometry in which noncovalent interactions (e.g., between protein subunits) are maintained. However, it is now known that for a number of protein complexes, their subunit interactions are maintained but their folded structure is perturbed, giving rise to CCSs much larger than the native state. The possibility of retaining not only noncovalent interactions but also conformations very similar to the solution phase prompts the need for new terminology to adequately distinguish between native, in which interactions are maintained, and folded, in which both interactions and compact/folded structures are retained.

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

RECENT DEVELOPMENTS AND CHALLENGES Mass Spectrometry of Membrane Proteins A major challenge, ever since the first mass spectra of soluble protein complexes were recorded, has been the study of membrane proteins and their complexes in the gas phase. Integral membrane proteins span the lipid bilayer in cellular membranes, a richly hydrophobic environment with which these proteins are tailored with matching hydrophobic surfaces. Consequently, these proteins are not soluble in aqueous environments and are unfavorably and undesirably prone to unfolding and aggregation. To study membrane proteins in vitro, investigators can reconstitute them in a variety of systems, such as detergent micelles (85), bicelles (86), liposomes (87), amphipols (88), 462

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

Bicelle ll

Relative intensity (%)

High collisional activation detergent release results in dissociation 100

High collisional activation of bicellesolublized protein results in dissociation 6+

100

23+

11+ 12+ 5+

2,000

6,000

10,000

2,000

14,000

Low collisional activation detergent release

Amphipol

4,000

6,000

8,000

10,000

Unresolved charge states using a sticky detergent 100

Relative intensity (%)

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

100 Nanodisc

5+

0

0

Micelle

6+ 7+

29+

0 2,000

0 6,000

10,000

m/z

14,000

2,500

7,500

12,500 17,500 22,500

m/z

Figure 3 Membrane proteins can be solubilized in a variety of different systems, which are all compatible with mass spectrometry. Different solubilization systems and the choice of components in the system (e.g., the detergent for detergent-solubilized membrane proteins) affect the amount of energy required to remove the solubilizing components and thus resolve the charge states. A desire to maintain the folded conformations of a membrane protein predicates the use of the lowest collisional activation possible. Mass spectra are reprinted with permission from References 13 and 99.

and nanodiscs (89). These systems solubilize membrane proteins in aqueous environments by encasing the hydrophobic patches of the protein with complementary hydrophobic components of the respective systems. Although the mass spectrometry of soluble proteins needs only to remove solutes bound to the protein to achieve resolvable spectra, reconstituted membrane proteins pose a problem because the detergent and/or some lipids must be removed, as the membrane protein is taken from solution into the gas phase and through the mass spectrometer. If these components cannot be removed, then mass measurements suffer from severe heterogeneity, such that charge states often cannot be resolved, making masses unmeasurable (Figure 3). However, these unwanted bound adducts from the reconstitution system can often be removed by delipidation protocols in solution or collisional activation inside the mass spectrometer. Membrane proteins have been analyzed using a variety of ionization techniques. For example, in laser-induced liquid bead ion desorption, a protein is solubilized in detergent, and an infrared laser desorbs the protein from microdroplets (90). This technique is sensitive to laser intensity: At low intensity/soft conditions, intact complexes are generated with up to 15% of the detergent molecules bound. At higher energy, subunits dissociate. Despite the low charge-state distributions, this specialized technique is difficult to implement because the balance of activation needs to be carefully selected. A more robust ionization technique, MALDI, has also been used for membrane www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

463

PC66CH20-Robinson

ARI

7 January 2015

9:53

protein mass spectrometry (91, 92). The low charge states that arise from this ionization in many cases necessitate a high-mass detector. It is also challenging to ionize proteins efficiently in the presence of detergents. MALDI, however, has been used to study amphipol-trapped membrane proteins (92), but the complexes did not remain intact. The majority of the mass spectrometry studies of membrane proteins have used the softer conditions of ESI, and it is this we focus on next.

Mass Spectrometry of Detergent-Solubilized Membrane Proteins

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

The seminal work for the noncovalent mass spectrometry of a membrane protein was described in 2008–2009 (93, 94), when a membrane protein complex, containing both soluble and transmembrane subunits, BtuC2 D2 , a vitamin B12 importer, was successfully maintained intact in the gas phase of a mass spectrometer for the first time. This was important as it highlighted that the best approach for the mass spectrometry of membrane proteins was to introduce the protein into the mass spectrometer as a protein-detergent complex, which helps prevent dissociation or structural perturbation. Subsequently, the mechanism of how detergents permit the noncovalent mass spectrometry of membrane protein complexes has been investigated (95). When the detergent molecules that provide a solubilizing protective jacket around the membrane protein are removed by collisional activation, the removal promotes ion cooling, which reduces the internal energy of the ion, thereby preserving the folded protein structure—a model consistent with evaporative cooling. Results with other membrane proteins and detergents are consistent with this proposal, in which sticky detergents, ones that are harder to remove by collisional activation, offer less protection of the protein structure, as the internal energy of the ion is not so easily dissipated by the release of bound detergent molecules (13, 14). Presumably at the point at which enough energy is provided to remove the bound detergent, the internal energy of the ion is already high enough that the protein has begun to unfold, or this is concomitant with the energy sufficient for detergent release. There was some initial success with the detergent n-dodecyl-β-D-maltopyranoside (DDM), which is a relatively common detergent for the solubilization of membrane proteins (85), predominantly because many membrane proteins are stable and active in this detergent. Yet, for many membrane proteins, DDM is not a good choice for mass spectrometry analysis as it can be difficult to remove by collisional activation (13). Consequently, in the mass spectrum, a broad unresolved hump is often observed, representing severe heterogeneity caused by bound detergent and lipids, analogous to a very salty soluble protein. High collisional activation can produce resolved charge states in these cases, but at the expense of retaining noncovalent interactions and preserving folded conformations. To address this problem, researchers developed protocols for screening a range of different detergents for mass spectrometry (13). It is clear that the use of many detergents will yield resolved charge states, many of which require less energy for release from the protein than DDM. This makes these detergents appealing for nondenaturing mass spectrometry in which compact states are retained, although care must be taken that detergent micelles in the spectrum do not interfere with the m/z regions occupied by the protein. One of the most useful detergents that meets this criterion is C8E4 from the polyoxyethyleneglycol class of detergents. Screening a range of detergents by mass spectrometry also provides information on the differing abilities of detergents to remove bound lipid from membrane proteins. It is interesting to note, given that many membrane proteins appear to need bound lipid for stability (85), that the lipid-bound state may alter the observed stabilities in different detergents. For many types of experiments, not exclusive to mass spectrometry, bound lipid is a potentially undesirable source 464

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

ARI

7 January 2015

9:53

of sample heterogeneity. Critically, few other techniques can so readily provide information on the lipid-bound state of a membrane protein as can mass spectrometry. A fascinating property of the different detergents is the resulting average charge state of the protein. It is known for soluble proteins that lower charge states tend to favor the retention of the protein structure more than do higher charge states (96, 97). The same is also true for membrane proteins (98), and fortuitously, the best detergents in terms of ease of removal also yield average charge states that are substantially lower than in other detergents. This breakthrough has provided the opportunity for many more detergent-solubilized membrane proteins to be analyzed by mass spectrometry in their folded conformations. The reason why different detergents give different charge states, and the effect of the charge state on membrane protein structure, is currently under investigation. At this stage, it is clear that a combination of hydrogen-bonding capacity, surface tension, size, and other physico-chemical properties of the detergent micelle contributes to the varying influences of detergents on preserving the folded conformations of membrane proteins. The choice of detergent for nondenaturing mass spectrometry of a membrane protein complex should therefore balance the energy required to remove the bound detergent molecules with the input of energy that will cause disruption of the noncovalent interactions that are crucial for the retention of protein structure. Additionally, a detergent that meets the aforementioned criteria and yet yields the lowest charge states is desirable to retain the folded state. Interestingly, the major advancements for membrane proteins have come not only from many instrument modifications previously utilized for soluble protein complexes, but also from developments in the isolation and solubilization of membrane proteins in solution.

Detergent-Free Membrane Protein Mass Spectrometry Although many of the developments of membrane protein mass spectrometry have centered on detergent-solubilized proteins, mass spectrometry is also compatible with other solubilization methods. It is possible to use membrane proteins solubilized in nanodiscs, bicelles, and amphipols (92, 99, 100). The advantage of these approaches is that the membrane proteins are considered to be in more native-like environments, as these systems may better mimic the lipidic environment, lateral forces, and curvature of the membrane. Although the nanodisc approach benefits from the absence of the detergent micelles in the mass spectrum, a disadvantage is the energy cost to disrupt the nanodisc structure, necessary to resolve the protein under study, which can cause dissociation of the protein complex. In addition, the technique introduces heterogeneity from the presence of lipids in that the resulting protein after the disruption of the protein belt and some lipids remains associated with variable numbers of lipids. It is clear from recent work showing the stabilization by lipids of membrane proteins by mass spectrometry (14) that, in the general case, it is difficult to remove bound lipids by collisional activation at even the highest energies attainable in the mass spectrometers commonly used: well above the energies at which a protein may remain folded in its native state and above energies at which monomeric units of a protein complex begin to dissociate (14). Similar to nanodiscs, bicelles also require higher collisional activation to release the protein they contain but also tend to give lower charge states and smaller charge-state distributions than do detergent-solubilized membrane proteins. Lower charge states are also observed for amphipols, which are amphipathic polymers, thought to support increased membrane protein stability. Amphipols are designed to be effectively permanently attached to membrane proteins, which appears undesirable. Equivalent to a difficult-to-remove detergent, in which high energies are required to resolve the protein only, they are also likely to promote complex dissociation. Despite this, it has been shown to be possible to retain folded conformations of membrane proteins introduced to the mass spectrometer in amphipols (100). www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

465

PC66CH20-Robinson

ARI

7 January 2015

9:53

If methods can be found to more readily remove the disc, bicelle, or amphipol at low collisional activation, which would both help preserve the native/folded structure and negate the tendency for subunit dissociation, then the lower charge states could be advantageous for structural studies of membrane proteins in the gas phase. In addition, regardless of the method of membrane protein solubilization, instrument conditions need to be carefully optimized to prevent spectral degradation from, for example, detergent micelle clusters and free amphipols.

Collision-Induced Unfolding for Analyzing Protein Stability

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

A mass spectrometry method that is gaining popularity for interrogating proteins is collisioninduced unfolding (CIU). The protein ions are introduced under soft-ionization conditions that preserve their folded conformations, collisional activation is then applied, and the coupling of ion mobility means that the size of the ion can be measured as it undergoes conformational changes (generally collapse or unfolding) owing to the collisional activation. The collision voltage at which the transitions between conformations occur, and the transitions/intermediates themselves, can be compared, enabling the assessment of gas-phase stability. For example, this facilitates investigations of the effect of ligand binding on protein stability, informative for protein crystallization, drug discovery, and the identification of ligands that may be important for function. This application of collisional activation is akin to solution-based thermal denaturation experiments, in which temperature is replaced by the slow-heating collisional activation. There are many solution techniques to measure similar properties, including differential scanning fluorimetry, circular dichroism, differential scanning calorimetry, any activity assay under temperature control, and any other technique in which intactness is measured, either directly or indirectly, as a function of a denaturant. The benefit of the mass spectrometry approach, however, is that the so-called naked protein is studied in the absence of artifacts that can interfere with other assays. Additionally, and more importantly, because the mass is measured simultaneously with unfolding, it is possible to distinguish and identify the effects of individual binding events—the compound of interest is bound and the stabilizing effects of different numbers of bound species can be probed with certainty. Early work used CIU to assess the stabilization of the wild-type transthyretin complex and a disease variant by the substrate thyroxine (22). For these data, the ligand-bound ions survive longer in the gas phase before unfolding to larger, more unfolded, intermediates. In addition, thyroxine preserves the folded form of the disease variant more than it does for the wild type. Similar analyses showed that concanavalin A is stabilized by different sugars, with the rank order of stabilization correlating with binding strengths (101). Another implementation of CIU used a simple analysis to assess the effects of ligands on protein stability; the intensity of the native species for a range of monomeric proteins was monitored as a function of Elab (the product of the ion charge state and the collision voltage used for activation) (21). The Elab at which folded species had depleted by 50% was termed Ec50(unfold) , and these values were compared for ligand-free and ligand-bound forms. This analysis clearly showed that for the range of interactions of monomeric proteins investigated, all had an unambiguously positive effect on protein stability in the gas phase. These results demonstrate that CIU is an effective method of probing ligand effects on the stability of soluble protein complexes. In general, CIU data are visualized using contour plots (Figure 4a), which provide a view of the size of a particular ion as it changes with increasing collisional activation (22, 101–103). Although useful for visualizing and understanding the conformational changes of the ion, it is difficult to quantify differences in the unfolding plots. This difficultly has hampered more robust analyses and has restricted the method to highlighting qualitative differences in observed unfolding 466

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

1

a

N

I1

I2

I3

1

b

I3

N

150

0.5

Molar fraction

Trap voltage (V)

200

0.5

I1

100

I2

0 50

0

50 5,000

7,000

6,000

100

150

200

250

Trap voltage (V)

c

d

Most compact form Intermediates

Medium

CCS

High

CCS

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

CCS (Å2)

Low

Trap voltage (V)

Charge state

Figure 4 (a) The collision cross section (CCS) of an ion can be monitored as a function of the trap voltage, and distinct gas-phase forms can be identified (N, I1 , I2 , and I3 ). (b) The intensity of these forms can be extracted and visualized as a function of trap voltage. Panels a and b reprinted with permission from Reference 14. (c) The conformation of a protein in the gas phase changes as a function of the collisional activation. At low (trap) voltages, more compact (folded) forms can be observed, whereas the application of a greater amount of energy creates larger, unfolded species. (d ) When proteins are unfolded in solution, the numerous charge states observed in the gas phase have different conformations. The number of regions in which the CCS of the different charge states changes most rapidly (highlighted) has been correlated to the number of domains in a protein. The number of domains has also been correlated to the number of conformational forms observed during collision-induced unfolding. Panels c and d adapted from Reference 102 with permission from Wiley-VCH.

trajectories. Recognizing that an ion at different sizes can be thought of as an ensemble of different gas-phase intermediates, one can extract the intensity of these intermediates and can visualize the disappearance and appearance of different species as a function of collisional activation (Figure 4b). Based on observing the differences in the transitions between intermediates, a method to distinguish kinase inhibitor types was developed by combining CIU data for ligand-bound, ligand-dissociated, and charge-stripped ions (103). This method used differences in the observed intermediates at particular collision voltage ranges to distinguish between the inhibitor types. www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

467

ARI

7 January 2015

9:53

An extension to the analysis of the features of an unfolding plot investigated the collisional unfolding of proteins in the gas phase (102). During CIU, multiple gas-phase intermediates are normally observed, with a greater number observed the higher the charge state. Here, for particular charge states, the number of transitions observed between folded and different unfolding intermediates was shown to correlate to the number of domains in the protein structures (Figure 4c). This correlation was shown to also exist for Coulombic unfolding, in which proteins were unfolded in solution, resulting in a wide charge-state distribution. The number of charge-state ranges in which the CCS was most rapidly changing also correlated to the number of protein-structure domains (Figure 4d ). In this study, monomeric proteins were used; for multimeric proteins, it would be difficult to apply the same type of analysis, as there would have to be an assumption that only one subunit unfolds at any one time. These data comfortably support the notion that the unfolding process is one in which subunits progressively unfold, and it is sensible to suggest that this would happen sequentially through the discrete components of the protein tertiary structure. One difficulty in using CIU, however, has been the lack of methods to quantify stability. Although preliminary steps to do this have been made, recognizing that a protein unfolds in the gas phase through distinct intermediates was important. Based on this, a new method has recently been developed that is capable of rigorously quantifying gas-phase stability. In this approach (14), a model is generated for the unfolding data, with which the original unfolding contour plot can be back-calculated. The model supposes that the unfolding proceeds from an initial species (taken to be the native/folded one), through a series of intermediates, which are of different sizes. The basis of the model is a protein unfolding model, in which the relative fractions of each observed species are calculated at each collision voltage. Using Gaussian functions to model the intensity of each intermediate at each collision voltage, one can generate a theoretical unfolding plot, allowing the model to be minimized against the experimental data. With this model, it is now possible to quantify gas-phase stability, as the transition points at which a particular species diminishes, analogous to IC50 values, can be compared. This grants the ability to rank stability, for example, between the stabilizing effects of different ligands binding to a protein. The rank order then provides key information to interrogate the interaction further by other biochemical means. Up until very recently, CIU experiments had been limited to soluble proteins, for which native and folded forms of these proteins in the gas phase are generally easily accessible. However, with the discovery of a new range of detergents compatible with the mass spectrometry of membrane proteins (13), select members of which are amenable to preserving folded conformations in the gas phase (14), this has allowed native and folded-like conformations to be attained, opening the door to the CIU analysis of membrane proteins. Experiments exploiting these developments have already been used to show, directly, that particular individual lipids stabilize membrane proteins differentially, which led to new insights on the role lipids play in membrane protein structure and function (14).

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

OUTLOOK Above we consider the generation of gas-phase protein ions from different ionization sources and the effects of these processes on the folded structure and interactions of proteins. A particular focus is on instrumental advances that have enabled experiments to progress from individual proteins to large heterogeneous complexes. Developments in applications to membrane protein complexes are also considered, and a qualitative description is given of the energetics required for releasing membrane proteins from detergent micelles, and alternative nondetergent matrices. 468

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

PC66CH20-Robinson

ARI

7 January 2015

9:53

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

We place emphasis on recent applications to probe gas-phase unfolding and stability, particularly of membrane proteins, and consider the problems of extracting quantitative data from these processes. This new approach for parameterizing these unfolding data, together with the range of detergents compatible with maintaining the folded state of membrane protein complexes, has opened up new horizons for mass spectrometry in terms of providing unique information about individual ligand-binding events. For all these processes, from initial ionization to transmission and detection in the gas phase, the removal of the detergent micelle, and the analysis of gas-phase unfolding trajectories, it is clear that a deeper understanding of the physicochemical principles that underlie these processes will aid the optimization and rationalization of the experimental procedures. It is our aspiration that this understanding, together with further instrument development, automation, and rigorous interpretation of data, will inspire greater acceptance of the gas phase as a powerful and versatile medium for garnering information from recalcitrant protein assemblies.

DISCLOSURE STATEMENT The authors are not aware of any affiliations, memberships, funding, or financial holdings that might be perceived as affecting the objectivity of this review.

ACKNOWLEDGMENTS We thank members of the Robinson group, past and present, for helpful contributions and discussions. In particular, we acknowledge contributions to the figures by Jonathon Hopper and Eamonn Reading. Funding for this research is gratefully acknowledged from the Royal Society, a Medical Research Council Program grant (98101), and an Advanced Investigator Award (26851) from the ERC (IMPRESS).

LITERATURE CITED 1. Kim MS, Pinto SM, Getnet D, Nirujogi RS, Manda SS, et al. 2014. A draft map of the human proteome. Nature 509:575–81 2. Hern´andez H, Dziembowski A, Taverner T, Seraphin B, Robinson CV. 2006. Subunit architecture of multimeric complexes isolated directly from cells. EMBO Rep. 7:605–10 3. Taverner T, Hern´andez H, Sharon M, Ruotolo BT, Matak-Vinkovic D, et al. 2008. Subunit architecture of intact protein complexes from mass spectrometry and homology modeling. Acc. Chem. Res. 41:617–27 4. Hern´andez H, Makarova OV, Makarov EM, Morgner N, Muto Y, et al. 2009. Isoforms of U1-70k control subunit dynamics in the human spliceosomal U1 snRNP. PLoS ONE 4:e7202 5. Politis A, Stengel F, Hall Z, Hern´andez H, Leitner A, et al. 2014. A mass spectrometry–based hybrid method for structural modeling of protein complexes. Nat. Methods 11:403–6 6. Marcoux J, Wang SC, Politis A, Reading E, Ma J, et al. 2013. Mass spectrometry reveals synergistic effects of nucleotides, lipids, and drugs binding to a multidrug resistance efflux pump. Proc. Natl. Acad. Sci. USA 110:9704–9 7. Sobott F, Robinson CV. 2002. Protein complexes gain momentum. Curr. Opin. Struct. Biol. 12:729–34 8. Sobott F, McCammon MG, Hern´andez H, Robinson CV. 2005. The flight of macromolecular complexes in a mass spectrometer. Philos. Trans. R. Soc. A 363:379–89. Discussion. 2005. Philos. Trans. R. Soc. A 363:389–91 9. Robinson CV. 2002. Protein complexes take flight. Nat. Struct. Biol. 9:505–6 10. Rose RJ, Damoc E, Denisov E, Makarov A, Heck AJR. 2012. High-sensitivity Orbitrap mass analysis of intact macromolecular assemblies. Nat. Methods 9:1084–86 www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

469

ARI

7 January 2015

9:53

11. Thompson NJ, Rosati S, Heck AJ. 2014. Performing native mass spectrometry analysis on therapeutic antibodies. Methods 65:11–17 12. Snijder J, van de Waterbeemd M, Damoc E, Denisov E, Grinfeld D, et al. 2014. Defining the stoichiometry and cargo load of viral and bacterial nanoparticles by Orbitrap mass spectrometry. J. Am. Chem. Soc. 136:7295–99 13. Laganowsky A, Reading E, Hopper JT, Robinson CV. 2013. Mass spectrometry of intact membrane protein complexes. Nat. Protoc. 8:639–51 14. Laganowsky A, Reading E, Allison TM, Ulmschneider MB, Degiacomi MT, et al. 2014. Membrane proteins bind lipids selectively to modulate their structure and function. Nature 510:172–75 15. Uetrecht C, Rose RJ, van Duijn E, Lorenzen K, Heck AJR. 2010. Ion mobility mass spectrometry of proteins and protein assemblies. Chem. Soc. Rev. 39:1633–55 16. Zhou M, Politis A, Davies RB, Liko I, Wu K-J, et al. 2014. Ion mobility–mass spectrometry of a rotary ATPase reveals ATP-induced reduction in conformational flexibility. Nat. Chem. 6:208–15 17. Bernstein SL, Dupuis NF, Lazo ND, Wyttenbach T, Condron MM, et al. 2009. Amyloid-β protein oligomerization and the importance of tetramers and dodecamers in the aetiology of Alzheimer’s disease. Nat. Chem. 1:326–31 18. Bleiholder C, Dupuis NF, Wyttenbach T, Bowers MT. 2011. Ion mobility–mass spectrometry reveals a conformational conversion from random assembly to β-sheet in amyloid fibril formation. Nat. Chem. 3:172–77 19. Knapman TW, Valette NM, Warriner SL, Ashcroft AE. 2013. Ion mobility spectrometry–mass spectrometry of intrinsically unfolded proteins: trying to put order into disorder. Curr. Anal. Chem. 9:181–91 20. Pagel K, Natan E, Hall Z, Fersht AR, Robinson CV. 2013. Intrinsically disordered p53 and its complexes populate compact conformations in the gas phase. Angew. Chem. Int. Ed. Engl. 52:361–65 21. Hopper JT, Oldham NJ. 2009. Collision induced unfolding of protein ions in the gas phase studied by ion mobility–mass spectrometry: the effect of ligand binding on conformational stability. J. Am. Soc. Mass. Spectrom. 20:1851–58 22. Hyung SJ, Robinson CV, Ruotolo BT. 2009. Gas-phase unfolding and disassembly reveals stability differences in ligand-bound multiprotein complexes. Chem. Biol. 16:382–90 23. Barber M, Bordoli RS, Sedgwick RD, Tyler AN. 1981. Fast atom bombardment of solids (F.A.B.): a new ion source for mass spectrometry. J. Chem. Soc. Chem. Commun. 1981:325–27 24. Devienne FM, Roustan JC. 1982. “Fast atom bombardment”: a rediscovered method for mass spectrometry. Org. Mass Spectrom. 17:173–81 25. Macfarlane RD, Torgerson DF. 1976. Californium-252 plasma desorption mass spectroscopy. Science 191:920–25 26. Zenobi R, Knochenmuss R. 1998. Ion formation in MALDI mass spectrometry. Mass Spectrom. Rev. 17:337–66 27. Miliotis T, Kjellstrom S, Nilsson J, Laurell T, Edholm LE, Marko-Varga G. 2000. Capillary liquid chromatography interfaced to matrix-assisted laser desorption/ionization time-of-flight mass spectrometry using an on-line coupled piezoelectric flow-through microdispenser. J. Mass Spectrom. 35:369–77 28. Kebarle P, Verkerk UH. 2009. Electrospray: from ions in solution to ions in the gas phase, what we know now. Mass Spectrom. Rev. 28:898–917 29. Benesch JL, Ruotolo BT, Simmons DA, Robinson CV. 2007. Protein complexes in the gas phase: technology for structural genomics and proteomics. Chem. Rev. 107:3544–67 30. Konermann L, Ahadi E, Rodriguez AD, Vahidi S. 2013. Unraveling the mechanism of electrospray ionization. Anal. Chem. 85:2–9 31. Thomson BA, Iribarne JV. 1979. Field induced ion evaporation from liquid surfaces at atmospheric pressure. J. Chem. Phys. 71:4451–63 32. Iribarne JV, Thomson BA. 1976. On the evaporation of small ions from charged droplets. J. Chem. Phys. 64:2287–94 33. Nguyen S, Fenn JB. 2007. Gas-phase ions of solute species from charged droplets of solutions. Proc. Natl. Acad. Sci. USA 104:1111–17 34. Schmelzeisen-Redeker G, Butfering L, Rollgen FW. 1989. Desolvation of ions and molecules in ther¨ ¨ mospray mass spectrometry. Int. J. Mass Spectrom. Ion Process. 90:139–50

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

470

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

ARI

7 January 2015

9:53

35. Dole M, Mack LL, Hines RL, Mobley RC, Ferguson LD, Alice MB. 1968. Molecular beams of macroions. J. Chem. Phys. 49:2240–49 36. de la Mora JF. 2000. Electrospray ionization of large multiply charged species proceeds via Dole’s charged residue mechanism. Anal. Chim. Acta 406:93–104 37. Hogan CJ Jr, Carroll JA, Rohrs HW, Biswas P, Gross ML. 2009. Combined charged residue-field emission model of macromolecular electrospray ionization. Anal. Chem. 81:369–77 38. Juraschek R, Dulcks T, Karas M. 1999. Nanoelectrospray: more than just a minimized-flow electrospray ionization source. J. Am. Soc. Mass. Spectrom. 10:300–8 39. Konijnenberg A, Butterer A, Sobott F. 2013. Native ion mobility–mass spectrometry and related methods in structural biology. Biochim. Biophys. Acta 1834:1239–56 40. Lossl P, Snijder J, Heck AJ. 2014. Boundaries of mass resolution in native mass spectrometry. J. Am. Soc. Mass. Spectrom. 25:906–17 41. Chernushevich IV, Loboda AV, Thomson BA. 2001. An introduction to quadrupole-time-of-flight mass spectrometry. J. Mass Spectrom. 36:849–65 42. Wollnik H, Przewloka M. 1990. Time-of-flight mass spectrometers with multiply reflected ion trajectories. Int. J. Mass Spectrom. Ion Process. 96:267–74 43. Tahallah N, Pinkse M, Maier CS, Heck AJR. 2001. The effect of the source pressure on the abundance of ions of noncovalent protein assemblies in an electrospray ionization orthogonal time-of-flight instrument. Rapid Commun. Mass Spectrom. 15:596–601 44. Rostom AA, Robinson CV. 1999. Detection of the intact GroEL chaperonin assembly by mass spectrometry. J. Am. Chem. Soc. 121:4718–19 45. Snijder J, Rose RJ, Veesler D, Johnson JE, Heck AJR. 2013. Studying 18 MDa virus assemblies with native mass spectrometry. Angew. Chem. Int. Ed. Engl. 52:4020–23 46. Sobott F, Hern´andez H, McCammon MG, Tito MA, Robinson CV. 2002. A tandem mass spectrometer for improved transmission and analysis of large macromolecular assemblies. Anal. Chem. 74:1402–7 47. Politis A, Park AY, Hall Z, Ruotolo BT, Robinson CV. 2013. Integrative modelling coupled with ion mobility mass spectrometry reveals structural features of the clamp loader in complex with single-stranded DNA binding protein. J. Mol. Biol. 425:4790–801 48. Ruotolo BT, Giles K, Campuzano I, Sandercock AM, Bateman RH, Robinson CV. 2005. Evidence for macromolecular protein rings in the absence of bulk water. Science 310:1658–61 49. Marcoux J, Politis A, Rinehart D, Marshall DP, Wallace MI, et al. 2014. Mass spectrometry defines the C-terminal dimerization domain and enables modeling of the structure of full-length OmpA. Structure 22:781–90 50. Sharon M. 2013. Biochemistry: Structural MS pulls its weight. Science 340:1059–60 51. Baldwin AJ, Lioe H, Robinson CV, Kay LE, Benesch JL. 2011. αB-crystallin polydispersity is a consequence of unbiased quaternary dynamics. J. Mol. Biol. 413:297–309 52. Ebong IO, Morgner N, Zhou M, Saraiva MA, Daturpalli S, et al. 2011. Heterogeneity and dynamics in the assembly of the heat shock protein 90 chaperone complexes. Proc. Natl. Acad. Sci. USA 108:17939–44 53. Levy ED, Boeri Erba E, Robinson CV, Teichmann SA. 2008. Assembly reflects evolution of protein complexes. Nature 453:1262–65 54. Li H, Wolff JJ, Van Orden SL, Loo JA. 2013. Native top-down electrospray ionization–mass spectrometry of 158 kDa protein complex by high-resolution Fourier transform ion cyclotron resonance mass spectrometry. Anal. Chem. 86:317–20 55. Zhang H, Cui W, Wen J, Blankenship RE, Gross ML. 2011. Native electrospray and electron-capture dissociation FTICR mass spectrometry for top-down studies of protein assemblies. Anal. Chem. 83:5598– 606 56. Yin S, Loo JA. 2011. Top-down mass spectrometry of supercharged native protein–ligand complexes. Int. J. Mass Spectrom. 300:118–22 57. Clarke D, Murray E, Hupp T, Mackay CL, Langridge-Smith PR. 2011. Mapping a noncovalent protein– peptide interface by top-down FTICR mass spectrometry using electron capture dissociation. J. Am. Soc. Mass Spectrom. 22:1432–40 58. Makarov A. 2000. Electrostatic axially harmonic orbital trapping: a high-performance technique of mass analysis. Anal. Chem. 72:1156–62 www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

471

ARI

7 January 2015

9:53

59. Hardman M, Makarov AA. 2003. Interfacing the orbitrap mass analyzer to an electrospray ion source. Anal. Chem. 75:1699–705 60. Perry RH, Cooks RG, Noll RJ. 2008. Orbitrap mass spectrometry: instrumentation, ion motion and applications. Mass Spectrom. Rev. 27:661–99 61. Rose RJ, Damoc E, Denisov E, Makarov A, Heck AJ. 2012. High-sensitivity Orbitrap mass analysis of intact macromolecular assemblies. Nat. Methods 9:1084–86 62. Snijder J, Heck AJ. 2014. Analytical approaches for size and mass analysis of large protein assemblies. Annu. Rev. Anal. Chem. 7:43–64 63. Benesch JLP. 2009. Collisional activation of protein complexes: picking up the pieces. J. Am. Soc. Mass Spectrom. 20:341–48 64. Jennings KR. 2000. The changing impact of the collision-induced decomposition of ions on mass spectrometry. Int. J. Mass Spectrom. 200:479–93 65. Shukla AK, Futrell JH. 2000. Tandem mass spectrometry: dissociation of ions by collisional activation. J. Mass Spectrom. 35:1069–90 66. Sleno L, Volmer DA. 2004. Ion activation methods for tandem mass spectrometry. J. Mass Spectrom. 39:1091–112 67. Zhou M, Sandercock AM, Fraser CS, Ridlova G, Stephens E, et al. 2008. Mass spectrometry reveals modularity and a complete subunit interaction map of the eukaryotic translation factor eIF3. Proc. Natl. Acad. Sci. USA 105:18139–44 68. Zhou M, Wysocki VH. 2014. Surface induced dissociation: dissecting noncovalent protein complexes in the gas phase. Acc. Chem. Res. 47:1010–18 69. Zhou M, Dagan S, Wysocki VH. 2012. Protein subunits released by surface collisions of noncovalent complexes: nativelike compact structures revealed by ion mobility mass spectrometry. Angew. Chem. Int. Ed. Engl. 51:4336–39 70. van den Heuvel RH, van Duijn E, Mazon H, Synowsky SA, Lorenzen K, et al. 2006. Improving the performance of a quadrupole time-of-flight instrument for macromolecular mass spectrometry. Anal. Chem. 78:7473–83 71. Zhou M, Jones CM, Wysocki VH. 2013. Dissecting the large noncovalent protein complex GroEL with surface-induced dissociation and ion mobility–mass spectrometry. Anal. Chem. 85:8262–67 72. von Helden G, Wyttenbach T, Bowers MT. 1995. Conformation of macromolecules in the gas phase: use of matrix-assisted laser desorption methods in ion chromatography. Science 267:1483–85 73. Jarrold MF. 2000. Peptides and proteins in the vapor phase. Annu. Rev. Phys. Chem. 51:179–207 74. Hoaglund-Hyzer CS, Counterman AE, Clemmer DE. 1999. Anhydrous protein ions. Chem. Rev. 99:3037–80 75. Loo J, Berhane B, Kaddis C, Wooding K, Xie Y, et al. 2005. Electrospray ionization mass spectrometry and ion mobility analysis of the 20S proteasome complex. J. Am. Soc. Mass. Spectrom. 16:998–1008 76. Ruotolo BT, Benesch JL, Sandercock AM, Hyung SJ, Robinson CV. 2008. Ion mobility–mass spectrometry analysis of large protein complexes. Nat. Protoc. 3:1139–52 77. Ruotolo BT, Robinson CV. 2006. Aspects of native proteins are retained in vacuum. Curr. Opin. Chem. Biol. 10:402–8 78. Ruotolo BT, Hyung SJ, Robinson PM, Giles K, Bateman RH, Robinson CV. 2007. Ion mobility–mass spectrometry reveals long-lived, unfolded intermediates in the dissociation of protein complexes. Angew. Chem. Int. Ed. Engl. 46:8001–4 79. Snijder J, Uetrecht C, Rose RJ, Sanchez-Eugenia R, Marti GA, et al. 2013. Probing the biophysical interplay between a viral genome and its capsid. Nat. Chem. 5:502–9 80. Mason EA, Schamp HW. 1958. Mobility of gaseous ions in weak electric fields. Ann. Phys. 4:233–70 81. Bush MF, Hall Z, Giles K, Hoyes J, Robinson CV, Ruotolo BT. 2010. Collision cross sections of proteins and their complexes: a calibration framework and database for gas-phase structural biology. Anal. Chem. 82:9557–65 82. Shvartsburg AA, Jarrold MF. 1996. An exact hard-spheres scattering model for the mobilities of polyatomic ions. Chem. Phys. Lett. 261:86–91 83. Mesleh MF, Hunter JM, Shvartsburg AA, Schatz GC, Jarrold MF. 1996. Structural information from ion mobility measurements: effects of the long-range potential. J. Phys. Chem. 100:16082–86

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

472

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

PC66CH20-Robinson

ARI

7 January 2015

9:53

84. Benesch JLP, Ruotolo BT. 2011. Mass spectrometry: come of age for structural and dynamical biology. Curr. Opin. Struct. Biol. 21:641–49 85. Seddon AM, Curnow P, Booth PJ. 2004. Membrane proteins, lipids and detergents: not just a soap opera. Biochim. Biophys. Acta 1666:105–17 86. Morrison EA, Henzler-Wildman KA. 2012. Reconstitution of integral membrane proteins into isotropic bicelles with improved sample stability and expanded lipid composition profile. Biochim. Biophys. Acta 1818:814–20 87. Rigaud J-L, L´evy D. 2003. Reconstitution of membrane proteins into liposomes. Methods Enzymol. 372:65–86 88. Popot JL, Berry EA, Charvolin D, Creuzenet C, Ebel C, et al. 2003. Amphipols: polymeric surfactants for membrane biology research. Cell. Mol. Life Sci. 60:1559–74 89. Ritchie TK, Grinkova YV, Bayburt TH, Denisov IG, Zolnerciks JK, et al. 2009. Reconstitution of membrane proteins in phospholipid bilayer nanodiscs. Methods Enzymol. 464:211–31 90. Morgner N, Kleinschroth T, Barth H-D, Ludwig B, Brutschy B. 2007. A novel approach to analyze membrane proteins by laser mass spectrometry: from protein subunits to the integral complex. J. Am. Soc. Mass. Spectrom. 18:1429–38 91. Chen F, Gerber S, Heuser K, Korkhov VM, Lizak C, et al. 2013. High-mass matrix-assisted laser desorption ionization–mass spectrometry of integral membrane proteins and their complexes. Anal. Chem. 85:3483–88 92. Bechara C, Bolbach G, Bazzaco P, Sharma KS, Durand G, et al. 2012. MALDI-TOF mass spectrometry analysis of amphipol-trapped membrane proteins. Anal. Chem. 84:6128–35 93. Barrera NP, Di Bartolo N, Booth PJ, Robinson CV. 2008. Micelles protect membrane complexes from solution to vacuum. Science 321:243–46 94. Barrera NP, Isaacson SC, Zhou M, Bavro VN, Welch A, et al. 2009. Mass spectrometry of membrane transporters reveals subunit stoichiometry and interactions. Nat. Methods 6:585–87 95. Borysik AJ, Hewitt DJ, Robinson CV. 2013. Detergent release prolongs the lifetime of native-like membrane protein conformations in the gas phase. J. Am. Chem. Soc. 135:6078–83 96. Hall Z, Politis A, Bush MF, Smith LJ, Robinson CV. 2012. Charge-state dependent compaction and dissociation of protein complexes: insights from ion mobility and molecular dynamics. J. Am. Chem. Soc. 134:3429–38 97. Hall Z, Robinson CV. 2012. Do charge state signatures guarantee protein conformations? J. Am. Soc. Mass. Spectrom. 23:1161–68 98. Mehmood S, Marcoux J, Hopper JTS, Allison TM, Liko I, et al. 2014. Charge reduction stabilizes intact membrane protein complexes for mass spectrometry. J. Am. Chem. Soc. 136:17010–12 99. Hopper JT, Yu YT, Li D, Raymond A, Bostock M, et al. 2013. Detergent-free mass spectrometry of membrane protein complexes. Nat. Methods 10:1206–8 100. Leney AC, McMorran LM, Radford SE, Ashcroft AE. 2012. Amphipathic polymers enable the study of functional membrane proteins in the gas phase. Anal. Chem. 84:9841–47 101. Niu S, Rabuck JN, Ruotolo BT. 2013. Ion mobility–mass spectrometry of intact protein–ligand complexes for pharmaceutical drug discovery and development. Curr. Opin. Chem. Biol. 17:809–17 102. Zhong Y, Han L, Ruotolo BT. 2014. Collisional and coulombic unfolding of gas-phase proteins: high correlation to their domain structures in solution. Angew. Chem. Int. Ed. Engl. 53:9209–12 103. Rabuck JN, Hyung S-J, Ko KS, Fox CC, Soellner MB, Ruotolo BT. 2013. Activation state-selective kinase inhibitor assay based on ion mobility–mass spectrometry. Anal. Chem. 85:6995–7002 104. Vahidi S, Stocks BB, Konermann L. 2013. Partially disordered proteins studied by ion mobility–mass spectrometry: implications for the preservation of solution phase structure in the gas phase. Anal. Chem. 85:10471–78 105. Robinson CV, Chung EW, Kragelund BB, Knudsen J, Aplin RT, et al. 1996. Probing the nature of noncovalent interactions by mass spectrometry: a study of protein–CoA ligand binding and assembly. J. Am. Chem. Soc. 118:8646–53 106. Liu L, Bagal D, Kitova EN, Schnier PD, Klassen JS. 2009. Hydrophobic protein–ligand interactions preserved in the gas phase. J. Am. Chem. Soc. 131:15980–81 www.annualreviews.org • Mass Spectrometry of Proteins

Changes may still occur before final publication online and in print

473

PC66CH20-Robinson

ARI

7 January 2015

9:53

Annu. Rev. Phys. Chem. 2015.66. Downloaded from www.annualreviews.org Access provided by Tulane University on 01/19/15. For personal use only.

107. Cubrilovic D, Haap W, Barylyuk K, Ruf A, Badertscher M, et al. 2014. Determination of protein–ligand binding constants of a cooperatively regulated tetrameric enzyme using electrospray mass spectrometry. ACS Chem. Biol. 9:218–26 108. Housden NG, Hopper JT, Lukoyanova N, Rodriguez-Larrea D, Wojdyla JA, et al. 2013. Intrinsically disordered protein threads through the bacterial outer-membrane porin OmpF. Science 340:1570–74 109. Mathavan I, Zirah S, Mehmood S, Choudhury HG, Goulard C, et al. 2014. Structural basis for hijacking siderophore receptors by antimicrobial lasso peptides. Nat. Chem. Biol. 10:340–42

474

Mehmood

·

Allison

·

Robinson

Changes may still occur before final publication online and in print

Mass spectrometry of protein complexes: from origins to applications.

Now routine is the ability to investigate soluble and membrane protein complexes in the gas phase of a mass spectrometer while preserving folded struc...
1MB Sizes 4 Downloads 13 Views