Collision-Induced Dissociation of Electrosprayed Protein Complexes

Subscriber access provided by UNIV OF CAMBRIDGE

Article

Collision-Induced Dissociation of Electrosprayed Protein Complexes: An All-Atom Molecular Dynamics Model with Mobile Protons Vlad Popa, Danielle A. Trecroce, Robert G. McAllister, and Lars Konermann J. Phys. Chem. B, Just Accepted Manuscript • DOI: 10.1021/acs.jpcb.6b03035 • Publication Date (Web): 24 May 2016 Downloaded from http://pubs.acs.org on June 3, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry B is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Collision-Induced Dissociation of Electrosprayed Protein Complexes: An All-Atom Molecular Dynamics Model with Mobile Protons

Vlad Popa1, Danielle A. Trecroce2, Robert G. McAllister2, and Lars Konermann1,2 *

1

Department of Biochemistry, The University of Western Ontario, London, ON, N6A 5C1, Canada 2

Department of Chemistry, The University of Western Ontario, London, ON, N6A 5B7, Canada

* corresponding author Email: [email protected] Telephone: (519) 661-2111 ext. 86313

Running Title: MD Simulations of Gaseous Protein Complexes

1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ABSTRACT: Electrospray ionization mass spectrometry (ESI-MS) has become an indispensable technique for examining noncovalent protein complexes. Collision-induced dissociation (CID) of these multiply protonated gaseous ions usually culminates in ejection of a single subunit with a disproportionately large amount of charge. Experiments suggest that this process involves subunit unfolding prior to separation from the residual complex, as well as H+ migration onto the unravelling chain. Molecular dynamics (MD) simulations are a promising avenue for gaining detailed insights into these CID events. Unfortunately, typical MD algorithms do not allow for mobile protons. Here we address this limitation by implementing a strategy that combines atomistic force fields (such as OPLS/AA and CHARMM36) with a proton hopping algorithm, focusing on the tetrameric complexes transthyretin and streptavidin. Protons are redistributed over all acidic and basic sites in 20 ps intervals, subject to an energy function that reflects electrostatic interactions and proton affinities. Our simulations predict that native-like conformers at the onset of collisional heating contain multiple salt bridges. Collisional heating initially causes subtle structural changes that lead to a gradual decline of these zwitterionic patterns. Many of the MD runs show gradual unfolding of a single subunit in conjunction with H+ migration, culminating in subunit separation from the complex. However, there are also instances where two or more chains start to unfold simultaneously, giving rise to charge competition. The scission point where the “winning” subunit separates from the complex can be attained for different degrees of unfolding, giving rise to product ions in various charge states. The simulated product ion distributions are in close agreement with experimental CID data. Proton enrichment in the departing subunit is driven by charge-charge repulsion, but the combination of salt bridge depletion, charge migration, and proton affinity causes surprising compensation effects among the various energy terms. It appears that this work provides the most detailed account to date of the mechanism whereby noncovalent protein complexes disassemble during CID. 2 ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Introduction Many proteins in solution form noncovalent multi-subunit complexes. Various methods are available for probing the structure and composition of these assemblies.1-4 One widely used approach is the transfer of protein complexes into the gas phase by electrospray ionization (ESI) for analysis by mass spectrometry (MS).5-10 Electrosprayed complexes can by further interrogated by various activation and fragmentation techniques,11-13 and by ion mobility spectrometry (IMS).14-18 Most experiments are conducted in positive ion mode, where net charge on the analyte is imparted by excess protons.19, 20 Of particular interest is the behavior of gaseous protein complexes during collision-induced dissociation (CID). CID is a slow heating process that involves thousands of collisions with background gas.21-23 When applied to protein complexes, the CID settings are usually adjusted such that only noncovalent interactions are disrupted, while covalent bonds remain intact. Under these conditions most complexes eject a single subunit with a disproportionally large amount of charge.12, 22, 24-27 As a typical example of a homo-tetrameric complex, [M4 + 14H]14+ transthyretin (TTR) releases M7+ or M8+, as well as the complementary charge-depleted trimers.28 Here, the monomeric CID products carry roughly half of the total charge but only a quarter of the total mass. The same behavior is seen when applying other slow heating methods.12 A framework has emerged that provides a phenomenological explanation for this “asymmetric charge partitioning” of collisionally activated complexes.12,

13, 22, 24, 26, 29, 30

The

concept envisions that the analyte initially retains a compact structure where excess protons are spread over the surface. Collisional heating induces unfolding of a single subunit. Subsequently, this subunit is ejected from the complex. Unfolding is accompanied by proton migration towards the unravelling subunit. This inter-subunit charge transfer takes place until the scission point is reached, i.e., until the departing subunit separates from the complex.12, 3 ACS Paragon Plus Environment

13, 22, 24, 26, 29, 30

The


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 32

described framework is consistent with the known high mobility of protons in collisionally activated biomolecules,31-33 and it is supported by the observation of semi-unfolded intermediates in IMS experiments.13, 34 Many facets of the CID process nonetheless remain poorly understood. These include the interplay between unfolding and charge migration, as well as the structures of intermediate species.13,

35

Alternative models that envision completely different mechanistic

avenues have been put forward as well.36 Molecular dynamics (MD) simulations can provide detailed insights into the behavior of biomolecular systems.37,

38

MD investigations on gaseous proteins are particularly attractive,

because the absence of solvent implies considerable savings in computer time. The gas phase protein community has made extensive use of MD techniques for exploring the structure and dynamics of desolvated proteins.39-45 A limitation of classical MD simulations is the fact they cannot account for H+ transfer events. This is a particular problem for studies on collisionally heated gaseous proteins, where protons are highly mobile.31-33 Realistic proton simulations call for QM/MM,46, 47 ab initio MD,48-50 or DFT/MD methods51 that are not practical for studying large proteins. Simplified proton transfer schemes have been proposed, but most of these are geared towards solution-phase simulations52-54 or they remain limited to relatively small systems.47 Thachuk et al.30 recently developed a simple charge hopping strategy for modeling the CID process of protein complexes. Charge was redistributed over basic sites during the MD runs. The charge patterns were governed by the tendency of the system to minimize the overall Coulomb energy, with an additional term that accounts for the proton affinity (PA) of the various sites. Focusing on TTR, those simulations30 confirmed the expected behavior12,

13, 22, 24, 26, 29

where

thermal excitation triggered unfolding of a single subunit with concomitant charge accumulation, followed by ejection from the complex.


Page 5 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Thachuk’s charge hopping work30 represents an important advance. However, the method was implemented only for the coarse-grained Martini force field. Compared to all-atom models, coarse-grained force fields are greatly simplified, as groups of atoms are mapped together into single beads.55 Hydrogens are not considered explicitly. Protonation and deprotonation are modeled by modifying charges on beads without removing or adding any atoms.30 Coarse-grained modeling is a powerful approach for studying very large systems in solution on very long time scales.55 Yet, the predictive power of coarse-grained methods tends to be lower than for all-atom models.56 The lack of solvent for gaseous proteins makes these systems readily accessible to allatom simulations. Therefore, it is highly desirable to extend a charge hopping approach of the type described by Thachuk30 to all-atom force fields. Another limitation of the method developed in ref.30 is the use a “positive-only” strategy that exclusively considers protonation changes of basic sites ( NT0/+, Arg0/+, Lys0/+, and His0/+). Acidic groups were kept permanently protonated: Asp0, Glu0, and CT0 (NT represents N-termini; CT stands for C-termini). This approach reflects the view that the high gas phase proton affinity of R-COO- will favor the neutralization of carboxylates.39, 42, 57 However, there is strong evidence that electrosprayed biomolecules can have zwitterionic properties and salt bridges.36, 58-63 Improved CID models should therefore allow for protonation changes of basic and acidic sites. The present study extends the work of Thachuk et al.30 by moving from coarse-grained to all-atom simulations of CID processes. We implement a proton hopping algorithm for redistributing charge in response to conformational changes. TTR and streptavidin serve as model systems. Both of these homo-tetrameric complexes form charge states around 14+ during native ESI, and their CID behavior has been well documented experimentally.28, 64, 65 The simulations are conducted by considering both positive-only and zwitterionic scenarios. Our data support the



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

proposed unfolding/charge migration framework,12, 13, 22, 24, 26, 29, 30 while providing novel insights into the energetics and conformational changes of protein complexes during CID.

Simulation Methods General Aspects. Protein simulations in a vacuum environment employed GROMACS 5.1 for leap frog integration of Newton’s equations with a 1 fs time step.66, 67 Bonds were constrained using LINCS.68 All runs were conducted on quad core desktop Linux workstations with GPU acceleration. Electrostatic and van der Waals interactions were modeled without cut-offs.69 Temperature control was achieved using the Nosé-Hoover thermostat70 with a coupling constant of 0.5 ps. Test runs employing a velocity rescaling thermostat71 resulted in artifactual accumulation of angular momentum. Resorting to the Nosé-Hoover scheme eliminated this problem. Center of mass translations were eliminated throughout the runs. Unless noted otherwise, the simulations employed the OPLS all-atom (OPLS/AA) force field.72 Additional runs were performed using the CHARMM36 all-atom force field.73 TTR simulations used the X-ray structure 3GRG as starting point.74 All heteroatoms were removed from the pdb file prior to processing it for GROMACS using PDB2GMX. Due to chain disorder the X-ray data do not show the positions of the first 9 Nterminal and the last 2 C-terminal residues in the TTR chains. These residues were added in extended conformations using the Pymol editor. Subsequent relaxation was performed for 100 ps at 1 K. Conformational dynamics of the termini during the early stages of the production runs (described below) quickly erased any memory of the initial extended conformations. To allow comparisons with experimental data28 TTR simulations were run on [M4 + 14H]14+ ions. An analogous strategy was used for streptavidin simulations, based on the X-ray structure 4Y5D.75 All


Page 6 of 32

Page 7 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


runs were repeated at least five times with different starting conformations, as well as different initial atom velocities.

Proton Moving Algorithm. We implemented an algorithm for distributing H+ among possible protonation sites (NT0/+, Arg0/+, Lys0/+, His0/+, Asp0/-, Glu0/-, and CT0/-). Interactions of these charge sites give rise to a Coulomb energy ECoul that is given by ECoul =

1 4πε 0κ e

N

N

∑∑

i =1 j =i +1

qi × q j

(1)

r ij

where qi and qj are the protonation-dependent charges (-e, 0, or e), with distances rij.30 Unless noted otherwise, a dielectric constant of κe = 1 was used.30, 76 An additional term arises from the proton affinities of the various sites (PANT = 960 kJ mol-1, PAARG = 1029 kJ mol-1 ,PALYS = 937 kJ mol-1, PAHIS = 958 kJ mol-1, PAASP- = 1450 kJ mol-1, PAGLU- = 1450 kJ mol-1, PACT- = 1450 kJ mol-1).30, 77 These PA values give rise to the energy EPA, i.e., E PA = − ∑ PAk

(2)

k

where summation includes only protonated sites. These two terms result in a “total” energy30 Etot = EPA + ECoul

(3)

The premise of our model is that mobile protons will adopt a spatial pattern that minimizes Etot.29, 30, 39

A search algorithm was developed that numbers protonation sites according to sequence. The

proton from the first occupied site is tentatively moved to each of the unoccupied sites. This is repeated for all occupied sites. The pattern producing the lowest Etot is retained, provided that it is lower than the previous minimum. Once all single proton moves have been tested, the process starts over. This sequence is repeated until the search algorithm fails to find a new Etot minimum. During the early stages of this project the acceptance of new protonation patterns involved a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Monte-Carlo strategy.30 However, data obtained in this way were virtually indistinguishable from those generated by always retaining the pattern with the lowest Etot. The simulations discussed below were thus generated without invoking Monte-Carlo techniques. Similar to earlier work,76 our algorithm uses a point charge approach where protonation/deprotonation is attributed to the following sites (using OPLS/AA notation):72 N-terminal nitrogen for NT+, NH2 for Arg+, NZ for Lys+, CB for His+, C for CT-, OD2 for Asp-, and OE2 for Glu-. Neutral His was modeled in the Nε2-H form, rather than Nδ1-H, consistent with the prevalence of the former in X-ray structures.78

Mobile Proton MD Simulations. Simulations were run as a succession of brief MD segments, each of which used a static protonation pattern. These segments were interleaved with proton redistribution events at regular time intervals (∆tredist), using the algorithm outline above. Each segment started with a steepest descent energy minimization for eliminating possible atom clashes that might arise from proton hopping. Subsequently, velocities were reassigned from a Maxwell-Boltzmann distribution at the desired temperature. To mimic protein heating during CID21-23 the temperature was increased by 50 K after every 4 ns, starting at 350 K, until one of the subunits separated from the complex. For picking a suitable ∆tredist one can consider that transition state theory suggests a lower limit on the order of ∆tredist ≈ (kBT/h)-1 ≈ 0.2 ps for an ideal barrier-free process.79 Solution-phase proton transfer barriers are on the order of kBT, corresponding to hopping times on a ps time scale.48 The absence of solvent in gaseous proteins may cause barriers to be somewhat higher, with correspondingly longer hopping times. With this in mind, we conducted test runs with ∆tredist values of 5 ps, 20 ps, and 100 ps. No discernible differences were found between these three conditions (data not shown). For the data discussed below we used ∆tredist = 20 ps.


Page 8 of 32

Page 9 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


To properly account for the high H+ mobility in thermally excited protein ions12, 31-33 we did not impose a maximum proton transfer distance. The underlying idea is that a single longrange H+ transfer from an occupied site H to a vacant site A is equivalent to a series of short-range hops taking place via a relay mechanism.80 For example, H-H-H-A → A-H-H-H is equivalent to

H-H-H-A → H-H-A-H → H-A-H-H → A-H-H-H.

Results and Discussion Performance of the Proton Redistribution Algorithm. The simulations discussed below require an efficient algorithm for identifying energetically favorable protonation patterns within protein complexes of arbitrary structure. The conceptually simplest approach for determining the lowest possible Etot (equation 3) would be to test all possible patterns. Unfortunately, the astronomical number of combinations60 N!/((N-n)! × n!) in which n protons can be distributed over N possible protonation sites precludes such an analysis. The search algorithm used here typically converges after testing ~2 × 105 patterns that can be computed within a few seconds. We first illustrate the algorithm for native-like TTR14+, without MD simulations. Figure 1a represents a highly improbable positive-only distribution, with charges 14+/0/0/0 on subunits A/B/C/D. Symmetry considerations suggest that after optimization each chain should carry roughly the same charge. The output of the algorithm confirms this expectation, generating a 3+/3+/4+/4+ pattern that lowers the energy by |∆Etot| = 4412 kJ mol-1 (Figure 1b). Similarly, a zwitterionic 17+/14+/7-/10- distribution is turned into 4+/3+/3+/4+, with |∆Etot| = 27500 kJ mol-1 (Figure 1c,d). Considering the vast parameter space, it is unlikely the algorithm will find the global



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 10 of 32

Etot minimum. Nonetheless, Figure 1 illustrates that our steepest descent approach quickly converges on approximate solutions that represent physically reasonable protonation patterns. It is instructive to consider the zwitterionic scenario in more detail (Figure 1c,d). Charge redistribution causes neutralization of many carboxylates, consistent with their high proton affinity.77 However, the algorithm retains a considerable number of R-COO- moieties that are involved in salt-bridges such as +/- or +/-/+ (Figure 1d). The presence of such motifs is consistent with studies on various gaseous ions.36, 58-61 For a closely spaced acceptor/donor pair a +/- saltbridged structure is more favorable than a neutralized 0/0 arrangement. Similarly, a +/-/+ scenario has a lower Etot than +/0/0, provided that the titratable sites are in close proximity to one another (see Supporting Figure S1 for a detailed analysis). Clearly, equation 3 represents a relatively crude tool for predicting zwitterionic vs. neutralized contacts. In-depth analyses would require DFT methods,61, 63 but the computational cost of such endeavors precludes their use in the present case. The salt bridge patterns used here thus represent semi-quantitative approximations. Despite these limitations, the possibility to conduct simulations on zwitterionic proteins represents a step beyond earlier positive-only models.30, 39, 42

Collision-Induced Dissociation of TTR. Figure 2a-d illustrates a typical CID run in positive-only mode. After 12 ns subunit B (blue) starts to protrude with its positively charged N-terminus. The subunit gradually unravels, but remains associated with the complex until separation occurs at t ≈ 18 ns. Unraveling is accompanied by charge accumulation on subunit B, culminating in 8+ at the scission point (see also Supporting Movie 1). Similar events take place in zwitterionic mode, illustrated in Figure 2e-h which shows ejection of a highly extended subunit C as an 8+ ion (Supporting Movie 2). We found no evidence of preferential ejection for any particular subunit, consistent with the symmetric structure of native TTR. This is different from the simulations of 10 ACS Paragon Plus Environment

Page 11 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ref.30 where subunit A was primed for ejection by making it the only chain with a chargeable Nterminus. Figure 3 takes a closer look at the CID events in positive-only and in zwitterionic mode. The two top panels of Figure 3 show the charge on each subunit, while the second row displays radii of gyration (Rg). Comparison of these profiles highlights the close correlation between subunit unfolding and charge accumulation. For the positive-only case ECoul decreases significantly as protons spread into the unraveling subunit (Figure 3d). The process does not cause major changes in EPA, because the PA values of the various basic sites are similar to one another.30 Hence, the Etot profile is dominated by changes in ECoul (Figure 3e). The energetics of this positive-only scenario confirm the expectation that H+ migration during CID is governed primarily by electrostatic repulsion.29, 30, 76 Surprisingly, under zwitterionic conditions Etot remains almost constant throughout the CID process. This effect arises from the combination of an ECoul increase and an EPA decrease (Figure 3j). The origin of this behavior requires closer examination. As noted, the proximity of acidic and basic sites in native-like conformations favors the formation of salt bridges (Figure 1c,d, see also Supporting Figure S1). Heating induces structural perturbations that render salt bridge formation less favorable. These events result in a gradual loss of zwitterionic character during heating, i.e., a steadily decreasing number of R-COO- sites. The decline of negative carboxylates is accompanied by a drop in the number of positive sites, keeping in mind that the overall TTR charge is fixed at 14+ (Figure 3h). Elimination of salt bridges (DH+ + A- → D + AH) lowers EPA, owing to the high PA of the acceptor sites Glu-, Asp-, and CT- (Figure 3j). At the same time, the decline of electrostatic attraction between opposite charges causes ECoul to increase. This interplay of EPA and ECoul results in Etot ≈ const (Figure 3i, j). The loss of native structure and zwitterionic character is highlighted in Figure 4; panel A shows subunit B at t = 0 with its native β sandwich 11 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 12 of 32

and numerous salt bridges. After 15.1 ns the structure is greatly perturbed, with retention of only one salt bridge (Figure 4b). The events seen in our MD simulations are consistent with IMS data, where structural changes are detectable long before subunit ejection takes place.64 For understanding why compact and extended structures can share the same Etot (Figure 3j) it is helpful to consider a minimalist model with 4 possible protonation sites and z = 2+. Figure 5a shows a compact “native state” where two positive charges (termed 1 and 2) interact with a salt bridge (Etot = -3295 kJ mol-1). In Figure 5b a structural perturbation has repositioned the salt bridge and increased the donor/acceptor distance, causing Etot to increase. The perturbed system can lower Etot by salt bridge neutralization (Figure 5c); however, this new Etot remains less favorable than for the native state of Figure 5a. In Figure 5d charges 1 and 2 are given the opportunity to separate from each other, e.g., as the result of subunit unfolding. The reduced electrostatic repulsion lowers Etot to -3295 kJ mol-1, which is identical to the value of Figure 5a. Overall, Figures 5a,d illustrate how a compact native state with salt bridges can have the same Etot as an extended state without salt bridges, reminiscent of the zwitterionic TTR scenario of Figure 3j. It is concluded that for zwitterionic TTR charge enrichment is driven by proton-proton repulsion, similar to the positive-only scenario. However, the former case is complicated by the interplay of attractive/repulsive interactions and proton affinity effects. The overall electrostatics of the zwitterionic model are dominated by the loss of salt bridges during heating, masking the energetically favorable effects of charge accumulation in the unraveling subunit.

Different Modes of Subunit Ejection. Not all simulation runs produced exactly the same CID behavior. This section illustrates the various types of events that were encountered in ten different OPLS/AA simulations on zwitterionic TTR14+. For most of these runs the initial step towards dissociation was the protrusion of a positively charged N-terminus. In addition, there were


Page 13 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


instances where a charged hairpin stuck out first. Roughly half of the runs followed a sequence similar to that depicted in Figure 2e-h, with extensive unfolding of the departing chain prior to separation from the residual trimer. The other runs produced a range of scission point structures where the departing subunits retained more compact conformations. In one case subunit B started to unravel, but then separated from the complex with very little further unfolding (Figure 6a). In another instance subunit C showed extensive unfolding at the expense of contacts with chain A, such that the latter was the one to separate (Figure 6b). Some of the runs displayed competition behavior, with two protruding chains that underwent unfolding in parallel (Figure 6c,d). The extent of charge accumulation on the ejected subunit is governed by its degree of unfolding DU relative to the residual trimer at the scission point. We define DU as Rg(departing chain)/ΣRg(all chains). A case where all four subunits are native (or equally unfolded) corresponds to DU = 0.25. Conversely, DU > 0.25 represents the case where the departing subunit is more unfolded than the three other chains. Figure 6e reveals a linear relationship between DU and the charge of the departing subunit. This correlation reflects the fact that extended chains have a larger capacity to accommodate excess protons. The observation of different DU values and various product charge states in our simulation is consistent with experiments, where CID always produces a range of z values rather than a single charge state.12, 13, 22, 24, 26, 29, 30

Versatility of the Mobile Proton Algorithm. Minor changes of the code allowed CID simulations to be conducted under a variety of conditions. Except for a single instance (one out of ten OPLS/AA runs on zwitterionic TTR), all mobile proton simulations showed the expected monomer/trimer separation. Figure 7 summarizes MD data for different scenarios by displaying the average charge on the ejected subunit, as well as the average temperature TCID at which separation took place. 13 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

The first three scenarios in Figure 7 refer to mobile proton TTR simulations using the OPLS/AA force field.72 The results obtained in positive-only mode and in zwitterionic mode are quite similar, with monomer charge states around 7+, and TCID between 500 and 550 K. A detailed discussion of the trajectories seen under these two conditions has been provided in the preceding sections. We also conducted positive-only runs that did not consider changes in EPA, such that proton redistribution was governed exclusively by electrostatic repulsion (Figure 7, first pair of bars). These conditions resulted in slightly higher z values, but the overall reaction sequence was similar to that of Figures 2a-d. The robustness of the TTR results for these different scenarios is reassuring. Also, the MD-generated monomer charge states are consistent with experiments, which show 7+ and 8+ monomers as the dominant CID products of TTR14+.28 To demonstrate the compatibility of our approach with different atomistic force fields we conducted additional TTR simulations in zwitterionic mode using CHARMM3673 (Figure 7, fourth set of bars). Compared to the OPLS/AA runs, the CHARMM36 data exhibited scission point structures that were more heterogeneous. Ejection of compact and extended chains took place with comparable probability (3/7 and 4/7 runs, respectively). The resulting average monomer charge is 5.5+ with a considerable standard deviation of ± 2.5. Different structural propensities for CHARMM and OPLS/AA have noted previously.81 While CHARMM has been highly successful for modeling proteins in solution,37 the results of the current work suggest that OPLS/AA yields more consistent data for gas phase CID simulations. After extensively characterizing the CID behavior of TTR with mobile protons, we also conducted OPLS/AA runs with a static 4+/3+/4+/3+ pattern, i.e., without charge hopping. This static pattern was generated in zwitterionic mode, analogously to the data of Figure 1d. Three out of five runs culminated in dimerization, with highly extended structures for both fragment ions. In two runs we observed separation of a compact monomer, with charge states 3+ and 4+ (Supporting 14 ACS Paragon Plus Environment

Page 15 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure S2). Hence, ejection of a highly extended monomer is the dominant dissociation pathway only for mobile proton models. Also, simulations with static charges resulted in the highest TCID (averaged over all five runs: 690 K, Figure 7), suggesting that the presence of mobile protons lowers the activation barriers for dissociation. Coulombic interactions in a bulk medium are modulated by dielectric effects, reflecting the effects of orientational and electronic polarization. Charge interactions are much more pronounced in the gas phase (κe = 1) than in water (κe ≈ 80). There is some debate regarding the appropriate κe value for gas phase proteins. As in previous work,30 we used κe = 1 for all preceding calculations, consistent with experimental data on gaseous peptides.82 Importantly, even with κe = 1 dielectric effects are explicitly considered in our model (albeit in a relatively crude way) due to the presence of zwitterionic/dipolar arrangements that interact with charge sites. Some earlier simulations42, 83 used κe = 2, an estimate that originates from gas phase basicity measurements on cytochrome c.57 For completeness, we conducted mobile proton OPLS/AA simulations on TTR14+ in zwitterionic mode with κe = 2. The reduction in electrostatic interactions destabilizes salt bridges (equation 1), causing the proton redistribution algorithm to eliminate all negative charges. Moreover, the diminished electrostatic repulsion lowers the propensity for subunit unfolding. As a result, all runs with κe = 2 culminated in ejection of compact monomers with an average charge of 4+ (Figure 7). The lack of unfolding and asymmetric charge partitioning under these conditions is not consistent with experimental observations.12, 13, 22, 24, 26 In other words, meaningful CID simulations for the model developed here require κe = 1. As a final step, we conducted mobile proton CID simulations on a noncovalent assembly other than TTR. Streptavidin represents one of the first complexes for which asymmetric charge partitioning was reported. Similar to TTR, experiments on streptavidin reveal the formation of 7+



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 16 of 32

and 8+ monomers as main CID products from 14+ tetramer ions.65 All five OPLS/AA runs of streptavidin in zwitterionic mode (with κe = 1) produced highly extended monomers. Subunit unraveling is favored by the presence of an N-terminal region that tends to carry two positive charges, owing to its high basicity with two Lys and the N-terminal amino group (Supporting Figure S3, Supporting Movie 3). The simulated monomer charge states of 8 ± 1 agree well with experimental spectra65 (last data set Figure 7).

Conclusions This study marks the first time that all-atom MD simulations were combined with a mobile proton model for examining the CID behavior of gaseous protein complexes. Our results are consistent with the view that asymmetric charge partitioning is caused by the interplay of subunit unfolding and charge migration, as previously suggested on the basis of experimental studies.12, 13, 22, 24, 26 Similar to earlier work30 gas collisions were not explicitly modeled in our study; instead, temperature control was achieved via a thermostat. This approach greatly reduces the computational cost of the simulations.84 Experiments confirmed that asymmetric charge partitioning takes place for different slow heating methods (collisional excitation and blackbody radiation).12 This implies that the absence of an explicit collision gas is not a major shortcoming of the MD strategy used here. The product charge states generated in our simulations are in good agreement with experimental data. The results are also consistent with the coarse-grained MD simulations of Thachuk et al.30 We observed the ejection of unfolded monomeric chains in high charge states under various conditions, i.e., positive-only with and without accounting for proton affinity, as


Page 17 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


well as zwitterionic scenarios with two different force fields. It is reassuring that the observed behavior is largely model-independent. Nonetheless, zwitterionic OPLS/AA simulations appear most suitable for describing the CID process, because they produced the most robust results while accounting for the known presence of salt bridges. 36, 58-63 Our simulations reveal that monomers can separate from the complex in various conformations, from relatively compact to highly extended. The corresponding DU values translate into different product charge states, consistent with experimental CID spectra that show a range of z values. A contributing factor to this charge state heterogeneity is the possibility that two or more chains can start to unfold simultaneously, causing competition for mobile protons. We are not aware of previous reports that have discussed such charge competition scenarios. For most MD runs the charge-depleted trimeric products adopt compact structures after monomer ejection. However, all subunits undergo extensive conformational changes during collisional heating. The interrogation of residual sub-complexes is therefore unlikely to provide information that is relevant to the solution structure of the protein. This last conclusion highlights the need to use methods other than CID for generating biologically relevant sub-complexes, such as surface-induced dissociation (SID)13, 85 or mild solution-phase destabilization. 86 In future work it will be interesting to extend simulations of the type described here to larger protein complexes that can undergo both collapse and unfolding.18, 42 SID simulations will be another intriguing application,13 as well as studies on the mechanisms by which bound ligands and metal ions affect the dissociation behavior.28

Acknowledgements. Funding for this work was provided by the Natural Sciences and Engineering Research Council of Canada (Discovery Grant 217080-2013).



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 18 of 32

Supporting Information Available. Supporting Figures and Movies as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.


Page 19 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


References 1. Alber, F.; Dokudovskaya, S.; Veenhoff, L. M.; Zhang, W. Z.; Kipper, J.; Devos, D.; Suprapto, A.; KarniSchmidt, O.; Williams, R.; Chait, B. T.; Rout, M. P.; Sali, A., Determining the architectures of macromolecular assemblies. Nature 2007, 450, 683-694. 2. Kristensen, A. R.; Gsponer, J.; Foster, L. J., A high-throughput approach for measuring temporal changes in the interactome. Nat. Methods 2012, 9, 907-+. 3. Robinson, C. V.; Sali, A.; Baumeister, W., The molecular sociology of the cell. Nature 2007, 450, 973-982. 4. Bai, X.-C.; McMullan, G.; Scheres, S. H. W., How cryo-EM is revolutionizing structural biology. Trends Biochem. Sci. 2015, 40, 49-57. 5. Dyachenko, A.; Gruber, R.; Shimon, L.; Horovitz, A.; Sharon, M., Allosteric mechanisms can be distinguished using structural mass spectrometry. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 7235-7239. 6. Uetrecht, C.; Heck, A. J. R., Modern Biomolecular Mass Spectrometry and its Role in Studying Virus Structure, Dynamics, and Assembly. Angew. Chem. Int. Ed. 2011, 50, 8248-8262. 7. Mehmood, S.; Allison, T. M.; Robinson, C. V., Mass Spectrometry of Protein Complexes: From Origins to Applications. Annu. Rev. Phys. Chem. 2015, 66, 453-474. 8. Kaddis, C. S.; Loo, J. A., Native Protein MS and Ion Mobility: Large Flying Proteins with ESI. Anal. Chem. 2007, 79, 1779-1784. 9. Benesch, J. L. P.; Ruotolo, B. T., Mass spectrometry: come of age for structural and dynamical biology. Curr. Op. Struct. Biol. 2011, 21, 641-649. 10. Chen, F.; Gulbakan, B.; Weidmann, S.; Fagerer, S. R.; Ibanez, A. J.; Zenobi, R., Applying mass spectrometry to study non-covalent biomolecule complexes. Mass Spectrom. Rev. 2016, 35, 48-70. 11. Lermyte, F.; Konijnenberg, A.; Williams, J. P.; Brown, J. M.; Valkenborg, D.; Sobott, F., ETD Allows for Native Surface Mapping of a 150 kDa Noncovalent Complex on a Commercial Q-TWIMS-TOF Instrument. J. Am. Soc. Mass Spectrom. 2014, 25, 343-350. 12. Felitsyn, N.; Kitova, E. N.; Klassen, J. S., Thermal Decomposition of a Gaseous Multiprotein Complex Studied by Blackbody Infrared Radiative Dissociation. Investigating the Origin of the Asymmetric Dissociation Behavior. Anal. Chem. 2001, 73, 4647-4661. 13. Quintyn, R. S.; Zhou, M.; Yan, J.; Wysocki, V. H., Surface-Induced Dissociation Mass Spectra as a Tool for Distinguishing Different Structural Forms of Gas-Phase Multimeric Protein Complexes. Anal. Chem. 2015, 87, 1187911886. 14. Bernstein, S. L.; Dupuis, N. F.; Lazo, N. D.; Wyttenbach, T.; Condron, M. M.; Bitan, G.; Teplow, D. B.; Shea, J.-E.; Ruotolo, B. T.; Robinson, C. V.; Bowers, M. T., Amyloid-β protein oligomerization and the importance of tetramers and dodecamers in the aetiology of Alzheimer's disease. Nat. Chem. 2009, 1, 326-331. 15. Ashcroft, A. E., Mass Spectrometry and the Amyloid Problem - How Far Can We Go in the Gas Phase? J. Am. Soc. Mass Spectrom. 2010, 21, 1087-1096. 16. Servage, K. A.; Silveira, J. A.; Fort, K. L.; Clemmer, D. E.; Russell, D. H., Water-Mediated Dimerization of Ubiquitin Ions Captured by Cryogenic Ion Mobility-Mass Spectrometry. J. Phys. Chem. Lett. 2015, 6, 4947-4951. 17. Bush, M. F.; Hall, Z.; Giles, K.; Hoyes, J.; Robinson, C. V.; Ruotolo, B. T., Collision Cross Sections of Proteins and Their Complexes: A Calibration Framework and Database for Gas-Phase Structural Biology. Anal. Chem. 2010, 82, 9667-9565. 18. Hogan, C. J.; Ruotolo, B. T.; Robinson, C. V.; de la Mora, J. F., Tandem Differential Mobility Analysis-Mass Spectrometry Reveals Partial Gas-Phase Collapse of the GroEL Complex. J. Phys. Chem. B 2011, 115, 3614-3621. 19. Heck, A. J. R.; Van den Heuvel, R. H. H., Investigation of intact protein complexes by mass spectrometry. Mass Spectrom. Rev. 2004, 23, 368-389. 20. Allen, S. J.; Schwartz, A. M.; Bush, M. F., Effects of Polarity on the Structures and Charge States of NativeLike Proteins and Protein Complexes in the Gas Phase. Anal. Chem. 2013, 85, 12055−12061. 21. McLuckey, S. A.; Goeringer, D. E., Slow heating methods in tandem mass spectrometry. J. Mass Spectrom. 1997, 32, 461-474. 22. Benesch, J. L. P.; Ruotolo, B. T.; Simmons, D. A.; Robinson, C. V., Protein Complexes in the Gas Phase: Technology for Structural Genomics and Proteomics. Chem. Rev. 2007, 107, 3544-3567. 23. Mayer, P. M.; Poon, C., The Mechanism of Collisional Activation of Ions in Mass Spectrometry. Mass Spectrom. Rev. 2009, 28, 608-639. 24. Light-Wahl, K. J.; Schwartz, B. L.; Smith, R. D., Observation of the Noncovalent Quaternary Association of Proteins by Electrospray Ionization Mass Spectrometry. J. Am. Chem. Soc. 1994, 116, 5271-5278.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 20 of 32

25. Jones, C. M.; Beardsley, R. L.; Galhena, A. S.; Dagan, S.; Cheng, G.; Wysocki, V. H., Symmetrical GasPhase Dissociation of Noncovalent Protein Complexes via Surface Collisions. J. Am. Chem. Soc. 2006, 128, 1504415045. 26. Jurchen, J. C.; Williams, E. R., Origin of Asymmetric Charge Partitioning in the Dissociation of Gas-Phase Protein Homodimers. J. Am. Chem. Soc. 2003, 125, 2817-2826. 27. Abzalimov, R. R.; Frimpong, A. K.; Kaltashov, I. A., Gas-phase processes and measurements of macromolecular properties in solution: On the possibility of false positive and false negative signals of protein unfolding. Int. J. Mass Spectrom. 2006, 253, 207-216. 28. Han, L.; Hyung, S.-J.; Ruotolo, B. T., Bound Cations Significantly Stabilize the Structure of Multiprotein Complexes in the Gas Phase. Angew. Chem. Int. Ed. 2012, 51, 5692-5695. 29. Sciuto, S. V.; Liu, J.; Konermann, L., An Electrostatic Charge Partitioning Model for the Dissociation of Protein Complexes in the Gas Phase. J. Am. Soc. Mass Spectrom. 2011, 22, 1679-1689. 30. Fegan, S. K.; Thachuk, M., A Charge Moving Algorithm for Molecular Dynamics Simulations of Gas-Phase Proteins. J. Chem. Theory Comput. 2013, 9, 2531-2539. 31. Jørgensen, T. J. D.; Gårdsvoll, H.; Ploug, M.; Roepstorff, P., Intramolecular Migration of Amide Hydrogens in Protonated Peptides upon Collisional Activation. J. Am. Chem. Soc. 2005, 127, 2785-2793. 32. Dongré, A. R.; Jones, J. L.; Somogyi, Á.; Wysocki, V. H., Influence of Peptide Composition, Gas-Phase Basicity, and Chemical Modification on Fragmentation Efficiency: Evidence for the Mobile Proton Model. J. Am. Chem. Soc. 1996, 118, 8365-8374. 33. Boyd, R. K.; Somogyi, Á., The Mobile Proton Hypothesis in Fragmentation of Protonated Peptides: A Perspective. J. Am. Soc. Mass Spectrom. 2010, 21, 1275-1278. 34. Ruotolo, B. T.; Hyung, S.-J.; Robinson, P. M.; Giles, K.; Bateman, R. H.; Robinson, C. V., Ion Mobility– Mass Spectrometry Reveals Long-Lived, Unfolded Intermediates in the Dissociation of Protein Complexes. Angew. Chem. Int. Ed. 2007, 46, 8001-8004. 35. Benesch, J. L. P., Collisional Activation of Protein Complexes: Picking Up the Pieces. J. Am. Soc. Mass Spectrom. 2009, 20, 341-348. 36. Ogorzalek Loo, R. R.; Loo, J. A., Salt Bridge Rearrangement (SaBRe) Explains the Dissociation Behavior of Noncovalent Complexes. J. Am. Soc. Mass Spectrom. 2016, in press, DOI: 10.1007/s13361-016-1375-3. 37. Piana, S.; Lindorff-Larsen, K.; Shaw, D. E., Atomic-level description of ubiquitin folding. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, 5915-5920. 38. Karplus, M., Development of Multiscale Models for Complex Chemical Systems: From H+H-2 to Biomolecules (Nobel Lecture). Angew. Chem.-Int. Edit. 2014, 53, 9992-10005. 39. Mao, Y.; Woenckhaus, J.; Kolafa, J.; Ratner, M. A.; Jarrold, M. F., Thermal unfolding of unsolvated cytochrome c: Experiment and molecular dynamics simulations. J. Am. Chem. Soc. 1999, 121, 2712-2721. 40. Chen, S. H.; Chen, L. X.; Russell, D. H., Metal-Induced Conformational Changes of Human Metallothionein2A: A Combined Theoretical and Experimental Study of Metal-Free and Partially Metalated Intermediates. J. Am. Chem. Soc. 2014, 136, 9499-9508. 41. Khakinejad, M.; Kondalaji, S. G.; Maleki, H.; Arndt, J. R.; Donohoe, G. C.; Valentine, S. J., Combining Ion Mobility Spectrometry with Hydrogen-Deuterium Exchange and Top-Down MS for Peptide Ion Structure Analysis. J. Am. Soc. Mass Spectrom. 2014, 25, 2103-2115. 42. Hall, Z.; Politis, A.; Bush, M. F.; Smith, L. J.; Robinson, C. V., Charge-State Dependent Compaction and Dissociation of Protein Complexes: Insights from Ion Mobility and Molecular Dynamics. J. Am. Chem. Soc. 2012, 134, 3429-3438. 43. Wyttenbach, T.; Pierson, N. A.; Clemmer, D. E.; Bowers, M. T., Ion Mobility Analysis of Molecular Dynamics. Annu. Rev. Phys. Chem. 2014, 65, 175–196. 44. Patriksson, A.; Adams, C. M.; Kjeldsen, F.; Zubarev, R. A.; van der Spoel, D., A direct comparison of protein structure in the gas and solution phase: The TRP-cage. J. Phys. Chem. B 2007, 111, 13147-13150. 45. Steinberg, M. Z.; Elber, R.; McLafferty, F. W.; Gerber, R. B.; Breuker, K., Early Structural Evolution of Native Cytochrome c After Solvent Removal. ChemBioChem 2008, 9, 2417-2423. 46. Goyal, P.; Qian, H. J.; Irle, S.; Lu, X. Y.; Roston, D.; Mori, T.; Elstner, M.; Cui, Q., Molecular Simulation of Water and Hydration Effects in Different Environments: Challenges and Developments for DFTB Based Models. J. Phys. Chem. B 2014, 118, 11007-11027. 47. Barnes, G. L.; Hase, W. L., Energy Transfer, Unfolding, and Fragmentation Dynamics in Collisions of NProtonated Octaglycine with an H-SAM Surface. J. Am. Chem. Soc. 2009, 131, 17185-17193. 48. Marx, D.; Chandra, A.; Tuckerman, M. E., Aqueous Basic Solutions: Hydroxide Solvation, Structural Diffusion, and Comparison to the Hydrated Proton. Chem. Rev. 2010, 110, 2174-2216.


Page 21 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


49. Iyengar, S. S.; Day, T. J. F.; Voth, G. A., On the amphiphilic behavior of the hydrated proton: an ab initio molecular dynamics study. Int. J. Mass Spectrom. 2005, 241, 197-204. 50. Burlet, O.; Orkiszewski, R. S.; Ballard, K. D.; Gaskell, S. J., Charge Promotion of Low-Energy Fragmentations of Peptide Ions. Rapid Commun. Mass Spectrom. 1992, 6, 658-662. 51. Nakai, H.; Sakti, A. W.; Nishimura, Y., Divide-and-Conquer-Type Density-Functional Tight-Binding Molecular Dynamics Simulations of Proton Diffusion in a Bulk Water System. J. Phys. Chem. B 2016, 120, 217-221. 52. Lill, M. A.; Helms, V., Molecular dynamics simulation of proton transport with quantum mechanically derived proton hopping rates (Q-HOP MD). J. Chem. Phys. 2001, 115, 7993-8005. 53. Donnini, S.; Tegeler, F.; Groenhof, G.; Grubmuller, H., Constant pH Molecular Dynamics in Explicit Solvent with lambda-Dynamics. J. Chem. Theory Comput. 2011, 7, 1962-1978. 54. Peng, Y. X.; Swanson, J. M. J.; Kang, S. G.; Zhou, R. H.; Voth, G. A., Hydrated Excess Protons Can Create Their Own Water Wires. J. Phys. Chem. B 2015, 119, 9212-9218. 55. Marrink, S. J.; Tieleman, D. P., Perspective on the Martini model. Chem. Soc. Rev. 2013, 42, 6801-6822. 56. Tozzini, V., Coarse-grained models for proteins. Curr. Op. Struct. Biol. 2005, 15, 144-150. 57. Schnier, P. D.; Gross, D. S.; Williams, E. R., Electrostatic Forces and Dielectric Polarizability of Multiply Protonated Gas-Phase Cytochrome c Ions Probed by Ion/Molecule Chemistry. J. Am. Chem. Soc. 1995, 117, 67476757. 58. Julian, R. R.; Hodyss, R.; Kinnear, B.; Jarrold, M. F.; Beauchamp, J. L., Nanocrystalline aggregation of serine detected by electrospray ionization mass spectrometry: Origin of the stable homochiral gas-phase serine octamer. J. Phys. Chem. B 2002, 106, 1219-1228. 59. Yoo, H. J.; Wang, N.; Zhuang, S. Y.; Song, H. T.; Hakansson, K., Negative-Ion Electron Capture Dissociation: Radical-Driven Fragmentation of Charge-Increased Gaseous Peptide Anions. J. Am. Chem. Soc. 2011, 133, 16790-16793. 60. Li, J.; Santambrogio, C.; Brocca, S.; Rossetti, G.; Carloni, P.; Grandori, R., Conformational Effects in Protein Electrospray Ionization Mass Spectrometry. Mass Spectrom. Rev. 2016, 35, 111-122. 61. Forbes, M. W.; Bush, M. F.; Polfer, N. C.; Oomens, J.; Dunbar, R. C.; Williams, E. R.; Jockusch, R. A., Infrared spectroscopy of arginine cation complexes: Direct observation of gas-phase zwitterions. J. Phys. Chem. A 2007, 111, 11759-11770. 62. Zhang, Z.; Browne, S. J.; Vachet, R. W., Exploring Salt Bridge Structures of Gas-Phase Protein Ions using Multiple Stages of Electron Transfer and Collision Induced Dissociation. J. Am. Soc. Mass Spectrom. 2014, 25, 604613. 63. Jockusch, R. A.; Price, W. D.; Williams, E. R., Structure of cationized arginine (Arg center dot M+, M = H, Li, Na, K, Rb, and Cs) in the gas phase: Further evidence for zwitterionic arginine. J. Phys. Chem. A 1999, 103, 92669274. 64. Pacholarz, K. J.; Barran, P. E., Distinguishing Loss of Structure from Subunit Dissociation for Protein Complexes with Variable Temperature Ion Mobility Mass Spectrometry. Anal. Chem. 2015, 87, 6271-6279. 65. Schwartz, B. L.; Bruce, J. E.; Anderson, G. A.; Hofstadler, S. A.; Rockwood, A. L.; Smith, R. D.; Chilkoti, A.; Stayton, P. S., Dissociation of Tetrameric Ions of Noncovalent Streptavidin Complexes Formed by Electrospray Ionization. J. Am. Soc. Mass Spectrom. 1995, 6, 459-465. 66. Hess, B.; Kutzner, C.; van der Spoel, D.; Lindahl, E., GROMACS 4: Algorithms for Highly Efficient, LoadBalanced, and Scalable Molecular Simulation. J. Chem. Theory Comput. 2008, 4, 435-447. 67. Van der Spoel, D.; Lindahl, E.; Hess, B.; Groenhof, G.; Mark, A. E.; Berendsen, H. J. C., GROMACS: Fast, flexible, and free. J. Comput. Chem. 2005, 26, 1701-1718. 68. Hess, B.; Henk, B.; Berendsen, H. J. C.; Fraaije, J. G. E. M., LINCS: A linear constraint solver for molecular simulations. J. Comput. Chem. 1997, 18, 1463-1472. 69. McAllister, R. G.; Metwally, H.; Sun, Y.; Konermann, L., Release of Native-Like Gaseous Proteins from Electrospray Droplets via The Charged Residue Mechanism: Insights from Molecular Dynamics Simulations. J. Am. Chem. Soc. 2015, 137, 12667-12676. 70. Hoover, W. G., Canonical dynamics: Equilibrium phase-space distributions. Phys. Rev. A 1985, 31, 16951697. 71. Bussi, G.; Donadio, D.; Parrinello, M., Canonical sampling through velocity rescaling. J. Chem. Phys. 2007, 126, 0141011-0141017. 72. Kaminski, G.; Duffy, E. M.; Matsui, T.; Jorgensen, W. L., Free Energies of Hydration and Pure Liquid Properties of Hydrocarbons from the OPLS All-Atom Model. J. Phys. Chem. 1994, 98, 13077-13082. 73. Huang, J.; MacKerell, A. D., CHARMM36 all-atom additive protein force field: Validation based on comparison to NMR data. J. Comput. Chem. 2013, 34, 2135-2145.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 22 of 32

74. Palmieri, L. D.; Lima, L.; Freire, J. B. B.; Bleicher, L.; Polikarpov, I.; Almeida, F. C. L.; Foguel, D., Novel Zn2+-binding Sites in Human Transthyretin: Implications for Amyloidogenesis and Retinol-Binding Protein Recognition. J. Biol. Chem. 2010, 285, 31731-31741. 75. Terai, T.; Kohno, M.; Boncompain, G.; Sugiyama, S.; Saito, N.; Fujikake, R.; Ueno, T.; Komatsu, T.; Hanaoka, K.; Okabe, T.; Urano, Y.; Perez, F.; Nagano, T., Artificial Ligands of Streptavidin (ALiS): Discovery, Characterization, and Application for Reversible Control of Intracellular Protein Transport. J. Am. Chem. Soc. 2015, 137, 10464-10467. 76. Sinelnikov, I.; Kitova, E. N.; Klassen, J. S., Influence of Coulombic Repulsion on the Dissociation Pathways and Energetics of Multiprotein Complexes in the Gas Phase. J. Am. Soc. Mass Spectrom. 2007, 18, 617-631. 77. Bienkowski, T.; Swider, P.; Blaziak, K.; Danikiewicz, W., Proton affinities of the anions of aromatic carboxylic acids measured by kinetic method. Int. J. Mass Spectrom. 2014, 357, 29-33. 78. Sudmeier, J. L.; Bradshaw, E. M.; Haddad, K. E. C.; Day, R. M.; Thalhauser, C. J.; Bullock, P. A.; Bachovchin, W. W., Identification of histidine tautomers in proteins by 2D H-1/C-13(delta 2) one-bond correlated NMR. J. Am. Chem. Soc. 2003, 125, 8430-8431. 79. Hanggi, P.; Talkner, P.; Borkovec, M., Reaction-rate theory: fifty years after Kramers. Rev. Mod. Pys. 1990, 62, 251-342. 80. Cukierman, S., Et tu, Grotthuss! and other unfinished stories. Biochim. Biophys. Acta 2006, 1757, 876-885. 81. Cino, E. A.; Choy, W. Y.; Karttunen, M., Comparison of Secondary Structure Formation Using 10 Different Force Fields in Microsecond Molecular Dynamics Simulations. J. Chem. Theo. Comp. 2012, 8, 2725-2740. 82. Gross, D. S.; Williams, E. R., Experimental Measurement of Coulomb Energy and Intrinsic Dielectric Polarizibility of a Multiply Protonated Peptide Ion Using Electrospray Ionization Fourier-Transform Mas Spectrometry. J. Am. Chem. Soc. 1995, 117, 883-890. 83. Shelimov, K. B.; Clemmer, D. E.; Hudgins, R. R.; Jarrold, M. F., Protein Structure in Vacuo: The Gas-Phase Conformation of BPTI and Cytochrome c. J. Am. Chem. Soc. 1997, 119, 2240-2248. 84. Wedekind, J.; Reguera, D.; Strey, R., Influence of thermostats and carrier gas on simulations of nucleation. J. Chem. Phys. 2007, 127, 0645011-06450112. 85. Snijder, J.; Heck, A. J. R., Analytical Approaches for Size and Mass Analysis of Large Protein Assemblies. Annu. Rev. Anal. Chem. 2014, 7, 43-64. 86. Hall, Z.; Politis, A.; Robinson, C. V., Structural Modeling of Heteromeric Protein Complexes from Disassembly Pathways and Ion Mobility-Mass Spectrometry. Structure 2012, 20, 1596-1609.


Page 23 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure Captions Figure 1. Illustration of the proton redistribution algorithm for gaseous TTR14+. Subunits are shown in different colors, with positive and negative charges as white and red spheres, respectively. (a) Manually selected, highly improbable protonation pattern in positive-only mode; (b) pattern after energy minimization. (c) Highly improbable pattern in zwitterionic mode; (d) optimized pattern after energy minimization. Tables display the charges on individual sites; the chemical nature of each site is denoted along the top, in order of sequence.

Figure 2. MD snapshots, representing different time points during the collision-induced dissociation of TTR14+ in positive-only mode (a-d), and in zwitterionic mode (e-h).

Figure 3. Time-dependent summary for typical TTR14+ CID simulations in positive-only mode (ae), and in zwitterionic mode (f-j), referring to the data of Figure 2. Vertical dotted lines indicate the scission point, where a highly charged and extended subunit separates from the complex. Panels a, f: Net charge on individual subunits. The charge of the departing monomer at the scission point is indicated. Panels b, g: Radius of gyration (Rg) of individual subunits. Panels c, h: Total number of positive and negative charges in the complex. Panels d, i: Contributions to ECoul, separated according to the type of charge-charge interaction. Panels e, j: EPA and Etot.

Figure 4. Structures of subunit B in TTR14+ at 0 ns (a), and at 15.10 ns (b). Secondary structure (β strand) detection was performed using Pymol. Positive and negative charges are shown as white and red spheres, respectively.

Figure 5. Energetics of charge arrangements in a two dimensional toy model. (a) Charges 1 and 2 23 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 24 of 32

are in contact with a stable salt bridge. (b) Charges 1 and 2 interact with an unstable salt bridge. (c) Same geometry as in panel B, but the salt bridge has been eliminated by proton transfer. (d) Same as panel C, except that the distance between charges 1 and 2 has increased. Scenarios (a) and (d) share the same Etot. Energies were calculated using equations 1-3, with PA values for Lys and Glu-.

Figure 6. (a-c) TTR structures just after subunit ejection from the 14+ tetramer. Each panel represents a different MD run conducted in zwitterionic mode. (d) Rg profiles for the competition scenario of panel C. (e) Charge of the ejected subunit vs. degree of subunit unfolding DU for 9 TTR runs in zwitterionic mode. The solid line represents a linear regression (r2 = 0.93).

Figure 7. TTR14+ and streptavidin14+ CID simulation results obtained under different conditions. Black bars represent the average charge state of the unfolded monomer at the scission point. Gray bars display the corresponding temperature. Only two of the five “TTR static charges” runs resulted in monomer ejection. Except for the “TTR zwitterionic CHARMM” data all results were generated with the OPLS/AA force field. Error bars represent standard deviations of five to nine MD runs for each condition. Additional information is provided in the text.


Page 25 of 32

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1

(a) (+14)

A B C D

(4)

A B C D

+14

0

0 +3

NT E K K D R H R K D D E K E E H E E E E K E D K K H E H E D R R K E CT 1 0 0 0

0 0 0 0

1 0 0 0

1 0 0 0

0 0 0 0

1 0 0 0

1 0 0 0

1 0 0 0

1 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

1 0 0 0

0 0 0 0

0 0 0 0

1 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

1 0 0 0

0 0 0 0

0 0 0 0

1 0 0 0

1 0 0 0

1 0 0 0

0 0 0 0

1 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

+4

(b) (3)

0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

1 0 1 1

0 0 0 0

1 1 1 1

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 1

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 0 0 0

0 1 1 0

1 1 1 0

0 0 0 1

0 0 0 0

0 0 0 0

+4 (c) A B C D

1 0 1 0 1-1 0-1

1 1 1 0

1 0 1 0 1-1 1-1

1 1 0 0

1 1 1 0

1 1 1 1

1 0 0 0 1 0 0 0 0-1-1-1 1-1-1-1

1 0 0 1 0 0 1-1-1 1-1-1

1 0 0 0 0 1 0 0 0 0 0-1-1-1-1 0-1-1-1-1

1 0 0 1 0 0 1-1-1 0-1-1

1 1 1 1

1 1 1 1

1 0 1 0 0-1 0-1

1 0 0 1 0 0 0-1-1 0-1-1

1 1 1 1

1 1 1 1

1 0 1-1 1 0 1-1

0 1 0 0

0-1 0 0 1 0 1 0

1 0 0 0

0 0 1 0

1 1 1 1

0 0 0 1

0 0 0 0 0 0 0-1-1 0 0 0

1 1 1 1

0 0 0 0 0-1 0-1

0 0 0 0-1 0 0 0 0 0 0-1

0 0 0 0

0 0 0 0

0 0-1 1-1-1 1-1-1 1 0-1

1 1 1 1

1 0 1 1

0-1 0 0 0-1 0-1

0 0 0 0

0-1 0 0 0-1 0-1

1 1 1 1


1 1 1 1

-10

+17

-7 +4

+14 +4

+3

+3

1 0 0 0-1-1 0-1-1 0-1-1

(d) A B C D

+3

0 0 0-1 0 0 0 0

0 0 0 0


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 26 of 32

Figure 2 Positive-Only

Zwitterionic

3+

(a)

(e)

4+

3+

3+

12.00 ns

4+

(b)

3+

4+

4+

(f)

3+

12.00 ns

3+

3+

16.00 ns

3+

13.00 ns 2+

5+

(c)

(g)

2+

7+

2+ 2+

2+

2+

18.00 ns

15.12 ns

2+

2+ 8+

8+

(d)

(h) 1+

3+

2+

2+ 2+

18.12 ns

2+

8+


15.14 ns

8+

Page 27 of 32

Figure 3 Positive-Only 500

550

600 T (K) 400

(a)

9

450

500

550

600

(f)

9

8+

8+

6

6

3

3

0

0

(b)

(g) 9

Subunit A Subunit B Subunit C Subunit D

6

6

3

Rg (nm)

Rg (nm)

9

3

0 40 Total Charges

Net Charge on Subunit

450

Zwitterionic

0 40

(h)

(c)

positive

30

30

20

20

10

10

Total Charges

Net Charge on Subunit

T (K) 400

negative 4e+4

ECoul (+/+)

4e+3

ECoul (+/+)

3e+3

-1

0

ECoul (-/-)

2e+3

-4e+4

ECoul (+/-)

-8e+4

1e+3

(j)

-1

Etot

0.0

-1

(e)

-1.14e+5

ECoul (all contributions) -4.0e+4 -1.17e+5

-8.0e+4

EPA

EPA

-1.2e+5

Etot -1.20e+5 0

4

8

12

16

20 0

4

8

Time (ns)


12

Time (ns)

Energy (kJ mol )

-1

Energy (kJ mol )

0

(i)

(d)

Energy (kJ mol )

0

Energy (kJ mol )

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


16

20


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 28 of 32

Figure 4

(a)

(b)


Page 29 of 32

Figure 5

(a)

(b)

(c)

(d)

+1

1.0 5.0

0.0

-0.5

+

1

+1

+

1

+ +

2

+2

+

2

stable salt bridge

Position (nm)

0.5

+

unstable salt bridge

0

+ 0

neutralized charges

0

2

0

neutralized charges

1000

0

-1000

kJ mol-1

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ECoul Series1 Series2 E PA

Series3 E

-2000

tot

-3000

-4000



Figure 6 (a)

(b)

(c)

(d)

Rg (nm)

4

Subunit A Subunit B Subunit C Subunit D

3

2

1 0

4

8

12

16

20

Time (ns) Charge of Ejected Subunit

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 30 of 32

10 9

(e)

8 7 6 5 4 3 0.2

0.3

0.4

0.5

0.6

Degree of Unfolding, DU


Page 31 of 32

Figure 7

Charge TCID

10

900 8

6

600

4 300 2

0

0

c e M es ni c PA = 2 ioni tiv g M i o t r i c s r r i u a o AR tte tte ctr ch i i -p H tho l i e l l c C i a w zw zw ic die tat R n s. R i n s T o R id T TT terio -p TR TT pt av all t T i R zw tre TT S R TT


TCID (K)

Ejected Monomer Charge

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 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 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

For TOC Only


Page 32 of 32

Collision-Induced Dissociation of Electrosprayed Protein Complexes

Recommend Documents