Mass spectrometric methods for protein structure determination

Mass spectrometry has proven in the last decade to be a powerful technique for structural characterisation of proteins and their interactions. Therefore, the application of mass spectrometry as an important tool in biochemical and biomedical science has rapidly increased over the last years. Proteins have been analysed by mass spectrometry for identification of amino acid sequences and post-translational modifications, as well for characterisation of their interactions with different ligands.

The capability to study extremely small quantities of molecules and mixtures of proteins with high sensitivity is a major advantage of mass spectrometric analyses over other analytical techniques.

In 2002, the Nobel Prize for Chemistry was awarded to John Fenn and Koichi Tanaka for the development of the two major “gentle ionisation” techniques, electrospray ionisation (ESI) and matrix-assisted laser desorption/ionisation (MALDI) mass spectrometry. Both ionisation processes have been shown to be able of the analysis of protein-ligand complexes.

A number of mass spectrometric methods [57, 59, 142] and combinations of mass spectrometry methods with H/D exchange [49, 50], Protein Ligand Interaction by Mass Spectrometry, Titration and H/D Exchange approach (PLIMSTEX) [143-145], Stability of Unpurified Proteins from Rates of H/D Exchange (SUPREX) [146, 147] and bioaffinity [98, 103, 104] have been established and developed for detection and quantification of protein-ligand interactions. HDX takes advantage of the drastically slower kinetics of HD exchange rates in regions of a polypeptide antigen shielded by binding to an antibody [148, 149]. SUPREX makes use of the extent of protein HDX in various concentrations of denaturants to determine the affinities of

protein-ligand complexes. PLIMSTEX, first developed by Zhu et. al. [143], tracks changes in the extent of protein HDX for various ligand/protein ratios to determine affinity, stoichiometry and conformational changes that occur upon ligand binding.

Thus far, there are no reports of extending these MS methods to antibody-antigen binding. Nevertheless, HDX is effective for epitope mapping to probe antibody-antigen interactions [148, 150].

HDX-MS is a powerful technique for characterisation of conformation and dynamics of proteins and protein complexes. The method can be applied in cases when conventional structural approaches such as NMR and X-ray crystallography cannot be applied due to low concentrations and impure samples. The application of mass spectrometry has therefore found high interest, particularly due to its high sensitivity (picomole-femtomole detection limits) [49, 50].

In H/D exchange, amide hydrogen atoms exchange for deuterium atoms when a protein is incubated in D₂O. The kinetics of exchange is sensitive to the hydrogen bonding of the amide backbone-amide hydrogens. In dynamic regions, the hydrogens will exchange quickly while tightly hydrogen-bonded amides exchange much more slowly. After subjecting a deuterated protein to rapid proteolysis under conditions that preserve the deuterium label, MS analysis reveals the extent of deuteration at a resolution of 5-20 residues (Figure 4). In protein-ligand complexes the binding domains are shielded from HDX and thus can be identified by mass spectrometry.

H→DExchange D₂O incubation

H+, 0°C

Proteolysis

MS HPLC

Protein Amide hydrogens in

disordered regions exchange fast

H/DExchange is measured as m/z

shift by MS

Figure 4: Schematic representation of the HDX-MS approach. Protein is incubated with D₂O solution, reaction which is essentially quenched by shifting the pH to 2 at 0 °C. The HDX-exchanged protein is then proteolyzed with pepsin. The peptic fragments are then chromatographically separated and their masses determined by mass spectrometry. The experiment is repeated in the absence of deuterium and the weight gain of each fragment attributed to deuteration.

Despite the fact that mass spectrometry measures biopolymer ions in the gas phase, it has been established and widely applied to protein-ligand interactions, since non-covalent interactions of protein structures are generally preserved in electrospray mass spectra [59, 90, 142, 151-154]. Electrospray ionisation is a soft ionisation method for producing intact molecular ions of biopolymers which has found high general application in bioanalytical chemistry. ESI-MS produces gaseous ionised molecules from solution [90] by creating a fine spray of charged droplets in an electric field [154-156]. The basic steps for ion formation involve the production of charged droplets at the electrospray tip in the presence of an external electric field, successive droplet disintegration, solvent evaporation leading to highly charged smaller droplets, and finally the release of multiply charged gas phase ions from the smaller droplets.

An electric field and gas flow force ions created near the atmosphere-vacuum interface

of a mass spectrometer to accelerate toward the inlet and enter the first vacuum pumping stage. An illustration of the ESI process is shown in Figure 5.

+ +

Spray needle tip Tailor cone Aerosol of fine charged droplets

Figure 5: Schematic representation of electrospray ionisation. (a) A high positive potential is applied to the capillary, causing positive ions in solution to drift towards the meniscus. Destabilization of the meniscus occurs, leading to the formation of a “Taylor cone” and a fine jet emitting droplets with excess positive charge. (b) Gas phase ions are formed from charged droplets in a series of solvent evaporation-Coulomb fission [59].

The repulsion between the charges on the surface causes intact ions to leave the droplet by a process known as a “Taylor cone” [157, 158]. The low flow rates of this technique provide significant advantages, including the use of buffers, detergents and other co-solvents required for the solubilisation of biopolymers. Multiple charging allows ions to be analysed based on a mass-to-charge (m/z) ratio, which greatly extends the mass range of the mass analyzer. Generally, ESI mass spectra show a steady increase of the charge state of ions with increasing molecular weight. The number of charges varies, depending on several parameters such as analyte, solvent, pH and temperature. For positive ion analysis of peptides and proteins in acidic solutions (pH < 4), the charges are normally associated with the most basic amino acids (Lys, Arg) of the molecule and the amino terminus.

During the last years, several types of non-covalent complexes have been analysed by ESI-MS. Example include protein-peptide, polypeptide-metal ion and protein-nucleic

ESI-MS are illustrated in Figure 6. The first example (Figure 6a), the complex formation between a synthetic peptide (R16L) and calmodulin was used as a model to analyse the interactions between calmodulin (CaM) and the CaM-binding sequences of smooth-muscle myosin light chain kinase in the presence and absence of Ca²⁺. The results indicate the importance of electrostatic forces in interactions between CaM and targets, particularly in the presence of Ca²⁺, and the role of hydrophobic forces in contributing additional stability to the complexes regardless if Ca²⁺ is present or absent. The stoichiometry of peptide binding was observed to be one peptide per one CaM. [159]. A second example illustrating a non-covalent protein complex analysis by ESI-MS is the complex between a nucleotide and the protein binding regions in the EF-Tu from T. thermophilus [57]. The transition of EF-Tu from an ”inactive” GDP to the “active” GTP binding form upon interaction with the nucleotide exchange factor EF-Ts was characterised by selective chemical modification and direct mass spectrometric analysis of the EF-Tu/-Ts and GDP complexes, and the Lys residues have been identified for the distinct nucleotide binding region. As an example, ESI spectra of free and GDP-bound forms of EF-Tu are compared in Figure 6c vs. Figure 6d which revealed a homogeneous ion series of a ratio 1:2 EFTu/GDP complex under the conditions employed [57].

For the first time simultaneous structure identification/characterisation by mass spectrometry, and kinetic data determinations for CaM-peptides complexes and nucleotide-peptide library have been possible due to the online combination of an SAW biosensor with electrospray mass spectrometry.

a

b

c

d

Figure 6: (a) ESI-FTICR mass spectrum of CaM with R16L peptide (concentration ratio 1:1.5) in 5 mM ammonium acetate buffer, pH 5.9. Insets show the expansions of the 7+ and 8+ charge-states. C represents CaM and P represents peptide, respectively [159]. (b) ESI mass spectra and structure model of (c) free EF-Tu; (d) EF-Tu complexed with GDP. The residue numbers denote lysine residues found shielded in the EF-Tu/GDP complex

Compared to the continuous ionisation mode of ESI from solution, matrix-assisted laser desorption ionisation (MALDI) utilizes a pulsed laser for the desorption of analyte molecules from a target surface on which they are co-crystallized with an excess of specific wavelength-absorbing matrix molecules, such as α-cyano-4-hydroxy cinnamic acid [160]. Ions are produced by short laser pulses at 337 nm wave length and accelerated into the mass analyser by applying a high potential electric field between the sample and the orifice. Although the details of energy conversion for ion formation are not fully understood, a general scheme of the MALDI process can be formulated as shown in Figure 7.

c ++

matrix/analyt could hν

laser pulse matrix/analyt

cluster

+

macro-ion in cluster

macro-ion

Figure 7: Schematic representation of the ion formation in MALDI mass spectrometry. A matrix/analyte cloud is desorbed from the microcrystalline matrix/sample preparation by a laser pulse. Gas-phase proton-transfer with matrix ions is considered to be primarily responsible for the subsequent generation of analyte ions.

The laser causes a rapid heating of the matrix crystals leading to the sublimation of matrix and analyte molecules. Ions are formed through gas-phase proton-transfer reactions with photoionized matrix molecules, generally producing low charge state ions [91, 161]. MALDI offers the advantages of intact protein ionisation, low sample requirements and high throughput analysis. Moreover, it has a higher tolerance towards salts compared to ESI analysis. MALDI-MS has also been applied to the study of protein complexes, DNA duplexes and anionic compounds binding to polybasic peptides [152, 162, 163], although with the limitation of uncertain structural states. Figure 8 illustrates the mass spectrum of heparin disaccharides complexed with a synthetic peptide.

a

b

[P+2H]²⁺

[P+D1+2H]²⁺

[P+H]⁺ 1442.8

[P+D1+H]⁺ 1981.8

[P+2D1+H]⁺

[P+2H]²⁺

[P+H]⁺ 1442.8

[P+D2+H]⁺ 2019.5

-SO3 1940.4

Figure 8: IR-MALDI mass spectra of heparin disaccharides D1 and D2 mixed with the synthetic peptide SP1. The lability of N-sulfate group(s) is obvious from spectrum (b) [163].

Even though the MALDI technique is a soft ionisation process, solvent evaporation during sample preparation may affect the complex stability and represents a limiting factor for the observation of non-covalent complexes. Furthermore, during evaporation, pH and ionic strength of the solution may be altered, leading to altered structure and interactions, thus presenting limitations for the use of MALDI-MS as a general tool for studying protein-ligand interactions.