Name: combining chemical cross-linking and mass spectrometry of intact protein complexes to study the architecture of multi-subunit protein assemblies
Uploaded: 2017-11-28T15:00:00
Duration: 10 min 1 s
Description: Watch this Scientific Journal Video about Combining Chemical Cross-linking and Mass Spectrometry of Intact Protein Complexes to Study the Architecture of Multi-subunit Protein Assemblies at JoVE.com

We thank our colleagues for helpful discussions. We also thank Ilme Schlichting and Karl-Peter Hopfner for providing protein complexes. We acknowledge funding from the Federal Ministry for Education and Research (BMBF, ZIK programme, 03Z22HN22), the European Regional Development Funds (EFRE, ZS/2016/04/78115) and the MLU Halle-Wittenberg to C.S. and funding from the Wellcome Trust (109854/658 Z/15/Z) to A.P.

The authors have nothing to disclose.

Protocols are provided for mass spectrometry-based structural analysis of multi-subunit protein complexes. The two techniques, described in the protocol, mostly deliver complementary results and are well-suited to gain insights into the structural arrangements within protein (-ligand) complexes which are difficult to study by conventional structural techniques. Native mass spectrometry delivers insights into protein stoichiometries as well as protein interactions by analyzing subcomplexes and stable interaction modules. Cross-linking, on the other hand, yields information on direct contact sites. Depending on the cross-linker used, a certain flexibility can or should be included in the analysis.
The provided protocols are in general easy to perform and not time-consuming. The entire protocol can be executed within a week and can be applied to almost all protein complexes, although, a certain amount of the protein complex is required for successful analysis. Sample preparation is simple and does not require specifically purified protein complexes. However, one common pitfall is contamination of the sample during sample preparation for mass spectrometry-based protein identification. These contaminations in most cases include keratins which originate from dust, skin, or hair. Therefore, extra care such as wearing gloves and lab coats, filtrating aqueous buffers, and using high-purity solvents should be taken during sample preparation for mass spectrometry-based protein identification. Other contaminating proteins such as chaperones are usually introduced during protein purification, e.g., when using affinity tags. In these cases, protein purification should be improved, for instance by increasing washing steps. In any case, protein contaminations in the sample are easily identified during the database search by omitting the taxonomy filter (i.e., searching against proteins from all species). If only a few peptides are observed (i.e., a low protein coverage could be obtained), even though sufficient sample is available, it might be necessary to use a different proteinase during digestion. In general Trypsin yields a sufficient number of peptides; however, in some cases such as membrane proteins or membrane domains of proteins, the number of tryptic cleavage sites is reduced and other enzymes targeting hydrophobic amino acids are a better choice.
In terms of instrumentation, a particularly modified instrument is required for native mass spectrometry which maintains non-covalent interactions during transfer into the gas-phase. Several instrument types have been introduced, including Q-ToF and orbitrap&#160;instruments. While modified Q-ToF mass spectrometers are commercially available for native mass spectrometry since several years, the latter were only recently introduced and in most cases require specialized modification45. However, the application of high-resolution instruments allowed studying binding of multiple ligands and their quantification46,47 and is promising for future applications.
To identify cross-linked di-peptides by liquid chromatography-coupled mass spectrometry, standard procedures with few modifications can be applied. However, the database search is the limiting factor as specialized software can rarely deal with large databases, and reduced databases containing the protein subunits of the complexes are required. Recent studies used mass spectrometry-cleavable cross-linkers targeting protein interactions in entire cell lysates48,49. The use of chemical cross-linkers that fragment in tandem mass spectrometry experiments mostly yields linear peptides (modified by the cross-linker), which can be identified by further fragmentation and database searching of linear peptides, and this reduces search time and computational search space. However, to perform these experiments, an ion trap mass analyzer or a hybrid mass spectrometer with an ion trap is required. In general, as false-positives are an important issue, mass spectra of cross-linked peptides are often validated manually by the quality of their fragment spectra which extends data analysis time enormously. Developing robust scoring systems that can be applied without further validation steps are therefore potential future applications. One way to improve data analysis and to reduce the number of false-positives was the introduction of false discovery rate calculations and their application to cross-linking data sets50.
In general, the techniques described here can be complemented with further mass spectrometry techniques (e.g., covalent labeling) to increase the output from the analysis. Other modifications and improvements of the protocols can be easily implemented. As such comparative cross-linking34 unravels conformational changes in the protein assembly. Further developments in native mass spectrometry nowadays allow the analysis of membrane proteins51,52 and their interactions with lipids28,52,53,54. New developments of high-resolution mass spectrometers for native mass spectrometry have extended the application and ligand binding, e.g., binding of lipids to membrane proteins, can now be included in the analysis45,46. In combination with computational modeling approaches, these techniques can deliver structural models of varying resolution55. If no crystal structures are available for the intact complex or single subunits, mass spectrometry can deliver first insights into protein interactions and the topology of the unknown complex. Depending on the techniques used and the results obtained, low-resolution models of the unknown complex can be obtained56,57,58. If crystal structures or homology models are available, the structural information received from mass spectrometry can yield even near-native models59.
In comparison with other structural techniques, mass spectrometry has the advantage that it requires low sample amounts, it can deal with heterogeneous samples and is applicable to protein complexes of unlimited size. In addition, mass spectrometry allows the investigation of dynamic protein systems. Different populations of the protein or protein complex that exist in solution are usually analyzed together and therefore, unlike with other structural techniques which require selection of certain populations, all conformations are maintained during analysis and are assessable in one experiment. Quantitative cross-linking approaches were recently introduced34,60,61 and are promising for future applications describing conformational changes under different conditions.

The structural investigation of protein assemblies has become particularly important for the understanding of cellular processes. Consequently, many, techniques have been developed and improved in structural biology1. However, these techniques are sometimes limited in their application due to size, flexibility, or heterogeneity of the protein complexes under investigation. Mass spectrometry can deal with these challenges and, therefore, emerged as a powerful tool in structural biology2,3,4,5. The greatest advantage of mass spectrometry, however, is the ability to unambiguously identify proteins even in complex and heterogeneous mixtures6. To this end, proteins are usually digested with endoproteinases and the resulting peptide mixture is separated by liquid chromatography and directly eluted into the mass spectrometer. Peptide masses are subsequently determined and precursor ions are selected for further fragmentation of the peptides. Proteins are then identified by searching peptide and corresponding fragment masses against a known database. This procedure not only allows the identification of peptides/proteins but also their post-translational modifications which cause a mass shift of the precursor peptides and the fragment ions carrying the modification7. Some of the many techniques in structural mass spectrometry are based on this principle4,8. For instance, labeling techniques such as hydrogen deuterium exchange9,10, chemical labeling strategies11,12, or hydroxyl radical foot printing13,14, give insights into the surface accessibility of the proteins under certain conditions.
Another technique is (chemical) cross-linking involving the covalent linkage of amino acids in close proximity through their functional groups. For this, chemical reagents, UV-activatable reagents, or amino acids are employed15,16. After cross-linking, the proteins are usually hydrolyzed with proteases and cross-linked di-peptides are analyzed by liquid chromatography-coupled mass spectrometry. The identification of cross-linking products by database searching, however, requires the use of specialized software which concatenate peptide sequences of various proteins and disparate regions17,18,19. The use of chemical cross-linkers has the advantage that it can be employed to almost every protein complex of interest and does not require incorporation of UV-activatable amino acids, which can only be achieved when expressing the protein of interest in host cells. As such, cross-linking is a versatile tool and was successfully employed in many structural studies of even large protein assemblies20.
Mass spectrometry of intact protein complexes (sometimes called &#39;native&#39; mass spectrometry), on the other hand, involves the analysis of intact proteins and protein complexes without hydrolysis into peptides. It reveals composition, heterogeneity, stoichiometry, topology, and subunit interactions of protein complexes21,22. In combination with ion mobility, native mass spectrometry further allows determination of their conformation23,24. This makes it a powerful tool for the structural investigation of protein complexes that are difficult to assess by conventional structural techniques. However, native mass spectrometry requires analysis buffers which maintain non-covalent protein interactions during electrospray ionization. This is usually achieved by using aqueous, volatile buffers such as ammonium acetate25. In addition, instrument modifications are necessary to prevent dissociation during transmission into the gas phase of the mass spectrometer26. Applied in this way, many (large) protein complexes were analyzed. Impressive examples are the studies of intact ribosomes27, ATP synthases28, or viruses29.
The combination of native mass spectrometry and cross-linking proved particularly successful in previous studies. For instance, the stoichiometries of chaperone complexes, including an unexpected Hsp70 dimer, could be obtained from mass spectrometry experiments of the intact protein complexes, while chemical cross-linking revealed the arrangements of the proteins in the assemblies30,31. In a different study, the effects of post-translational modifications on an intact ATP synthase complex were studied. Native mass spectrometry provided insights into the protein complex stability in the presence or absence of phosphorylation or acetylation32,33. A comparative cross-linking strategy34 then revealed conformational changes of the protein complex under the different conditions.
Here, we provide the protocols for protein identification by mass spectrometry, (chemical) cross-linking, and native mass spectrometry, including data analysis and interpretation (Figure 1). The combination of complementary results obtained from these methods with the help of two well-characterized protein complexes is demonstrated35. Our protocol can be applied to any protein assembly which can be purified at certain purity and concentration. The approach is in some cases limited by data analysis of cross-linking data, i.e., the size of the database employed, which determines the required search space and time. In addition, manual validation of identified cross-links is often required and further reduces the output. Native mass spectrometry is mostly limited by the sample quality, e.g., buffers and adducts used during purification and the possibility to exchange them by aqueous and volatile buffers. However, protein complexes purified for structural analysis usually have the quality required for successful analysis with our protocols.

<table>
	<tbody>
		<tr>
			<td>Chemicals, Reagents, Consumables</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>2-[4-(2-hydroxyethyl)piperazin-1-yl]ethanesulfonic acid (HEPES)</td>
			<td>Sigma Aldrich</td>
			<td>H4034</td>
			<td>buffer</td>
		</tr>
		<tr>
			<td>Acetic acid</td>
			<td>Sigma Aldrich</td>
			<td>695092</td>
			<td>pH</td>
		</tr>
		<tr>
			<td>Acetonitrile Optima LC/MS</td>
			<td>Fisher Chemicals</td>
			<td>A955-1</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Amicon Ultra centrifugal devices (different MWCO)</td>
			<td>Millipore</td>
			<td>i.e. UFC500396</td>
			<td>buffer exchange, concentration</td>
		</tr>
		<tr>
			<td>Ammonium acetate solution</td>
			<td>Sigma Aldrich</td>
			<td>A2706</td>
			<td>native MS</td>
		</tr>
		<tr>
			<td>Ammonium bicarbonate</td>
			<td>Sigma Aldrich</td>
			<td>9830</td>
			<td>in-gel digestion</td>
		</tr>
		<tr>
			<td>Ammonium solution</td>
			<td>Sigma Aldrich</td>
			<td>9859</td>
			<td>pH</td>
		</tr>
		<tr>
			<td>Bis(Sulfosuccinimidyl)suberate-d0 (BS3-d0)</td>
			<td>Thermo Scientific</td>
			<td>21590</td>
			<td>cross-linker</td>
		</tr>
		<tr>
			<td>Bis(Sulfosuccinimidyl)suberate-d4 (BS3-d4)</td>
			<td>Thermo Scientific</td>
			<td>21595</td>
			<td>cross-linker</td>
		</tr>
		<tr>
			<td>Caesium iodide</td>
			<td>Sigma Aldrich</td>
			<td>203033</td>
			<td>calibration</td>
		</tr>
		<tr>
			<td>Calcium chloride</td>
			<td>Sigma Aldrich</td>
			<td>C5670</td>
			<td>in-gel digestion</td>
		</tr>
		<tr>
			<td>Capillaries (1.0 OD &#215; 0.78 ID &#215; 100 L mm)</td>
			<td>Harvard Apparatus</td>
			<td>30-0038</td>
			<td>native MS</td>
		</tr>
		<tr>
			<td>Disuccinimidyl suberate (DSS)</td>
			<td>Thermo Scientific</td>
			<td>21655</td>
			<td>cross-linker</td>
		</tr>
		<tr>
			<td>DL-Dithiothreitol</td>
			<td>Sigma Aldrich</td>
			<td>D5545</td>
			<td>in-gel digestion</td>
		</tr>
		<tr>
			<td>DMSO (dimethyl sulfoxide)</td>
			<td>Sigma Aldrich</td>
			<td>D8418</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Ethanol</td>
			<td>Fisher Chemicals</td>
			<td>BP2818</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Formic acid Optima LC/MS</td>
			<td>Fisher Chemicals</td>
			<td>A117-50</td>
			<td>solvent supplement</td>
		</tr>
		<tr>
			<td>Instant Blue Coomassie staining solution</td>
			<td>expedeon</td>
			<td>ISB1L</td>
			<td>staining solution for gel electrophoresis</td>
		</tr>
		<tr>
			<td>Invitrogen NuPAGE 4-12% Bis-Tris gels (1 mm, 10 well)</td>
			<td>life technologies</td>
			<td>NP0323BOX</td>
			<td>gel electrophoresis</td>
		</tr>
		<tr>
			<td>Iodoacetamide</td>
			<td>Sigma Aldrich</td>
			<td>I1149</td>
			<td>In-gel digestion</td>
		</tr>
		<tr>
			<td>Isopropyl alcohol, HPLC grade</td>
			<td>Fisher Chemicals</td>
			<td>P750717</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Methanol</td>
			<td>Fisher Chemicals</td>
			<td>A456-212</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Micro BioSpin 6 columns</td>
			<td>BioRad</td>
			<td>732-6222</td>
			<td>buffer exchange</td>
		</tr>
		<tr>
			<td>NuPAGE Antioxidant</td>
			<td>invitrogen</td>
			<td>NP0005</td>
			<td>running buffer, gel electrophoresis</td>
		</tr>
		<tr>
			<td>NuPAGE LDS Sample Buffer (4 &#215;)</td>
			<td>invitrogen</td>
			<td>NP0007</td>
			<td>sample loading buffer (non-reducing) for gel electrophoresis</td>
		</tr>
		<tr>
			<td>NuPAGE MES SDS Running buffer</td>
			<td>life technologies</td>
			<td>NP0001</td>
			<td>gel electrophoresis</td>
		</tr>
		<tr>
			<td>NuPAGE MOPS SDS Running buffer</td>
			<td>life technologies</td>
			<td>NP0001</td>
			<td>gel electrophoresis</td>
		</tr>
		<tr>
			<td>NuPAGE Sample Reducing Agent (10 &#215;)</td>
			<td>invitrogen</td>
			<td>NP0004</td>
			<td>reducing Agent for gel electrophoresis</td>
		</tr>
		<tr>
			<td>PBS - phosphate buffered saline</td>
			<td>Sigma Aldrich</td>
			<td>P4417</td>
			<td>buffer</td>
		</tr>
		<tr>
			<td>SeeBlue Plus2 Prestained Standard Protein Marker</td>
			<td>invitrogen</td>
			<td>LC5925</td>
			<td>prestained Protein Marker for gel electrophoresis</td>
		</tr>
		<tr>
			<td>Sodium acetate</td>
			<td>Sigma Aldrich</td>
			<td>S2889</td>
			<td>ethanol precipitation</td>
		</tr>
		<tr>
			<td>Tris(hydroxymethyl)aminomethane (Tris)</td>
			<td>Sigma Aldrich</td>
			<td>252859</td>
			<td>buffer</td>
		</tr>
		<tr>
			<td>Trypsin</td>
			<td>Sigma Aldrich</td>
			<td>TRYPSEQ-RO (11418475001)</td>
			<td>in-gel digestion</td>
		</tr>
		<tr>
			<td>Vivaspin centrifugal devices (different MWCO)</td>
			<td>Sartorius</td>
			<td>i.e. VS0101</td>
			<td>buffer exchange, concentration</td>
		</tr>
		<tr>
			<td>Water for HPLC</td>
			<td>Sigma Aldrich</td>
			<td>270733-2.5L-M</td>
			<td>solvent</td>
		</tr>
		<tr>
			<td>Name</td>
			<td>Company</td>
			<td>Catalog Number</td>
			<td>Comments</td>
		</tr>
		<tr>
			<td>Instruments</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Centrifuge Heraeus Fresco 21, bench-top centrifuge</td>
			<td>Thermo Scientific</td>
			<td>75002425</td>
			<td></td>
		</tr>
		<tr>
			<td>Flaming / Brown Micropipette puller Model P-1000</td>
			<td>Sutter Instruments</td>
			<td>P-1000</td>
			<td></td>
		</tr>
		<tr>
			<td>Gelelectrophoresis chamber Xcell SureLock MiniCell</td>
			<td>Invitrogen</td>
			<td>EI0001</td>
			<td></td>
		</tr>
		<tr>
			<td>Gold Coater Quorum Q150R S</td>
			<td>Quorum Technologies Ltd.</td>
			<td>Q150RS</td>
			<td></td>
		</tr>
		<tr>
			<td>Horizontal gel shaker Rotamax 120 T</td>
			<td>Heidolph Instruments</td>
			<td>544-41200-00</td>
			<td></td>
		</tr>
		<tr>
			<td>Q-TOF Ultima mass spectrometer, MS Vision high mass upgrade</td>
			<td>Waters (Micromass)</td>
			<td>-</td>
			<td></td>
		</tr>
		<tr>
			<td>reversed-phase C18 analytical column Acclaim PepMap (C18, 75 mm I.D., 50 cm, 3 mm pore size)</td>
			<td>Thermo Scientific</td>
			<td>164570</td>
			<td></td>
		</tr>
		<tr>
			<td>reversed-phase C18 pre-column (C18, 150 mm I.D., 2 cm, 5 mm pore size)</td>
			<td>Thermo Scientific</td>
			<td>164213</td>
			<td></td>
		</tr>
		<tr>
			<td>scalpel</td>
			<td>Fisher Scientifc</td>
			<td>10463989</td>
			<td></td>
		</tr>
		<tr>
			<td>SpeedVac SPD121P vacuum centrifuge</td>
			<td>Thermo Scientific</td>
			<td>SPD121DP-230</td>
			<td></td>
		</tr>
		<tr>
			<td>Thermomixer C</td>
			<td>Eppendorf</td>
			<td>5382000015</td>
			<td></td>
		</tr>
		<tr>
			<td>Tweezers AA</td>
			<td>Sigma Aldrich</td>
			<td>Z680184-1EA</td>
			<td></td>
		</tr>
		<tr>
			<td>Ultrasonic cleaner USC-TH, sonication bath</td>
			<td>VWR</td>
			<td>142-0084</td>
			<td></td>
		</tr>
		<tr>
			<td>UltiMate Dionex 3000 nano-LC system, coupled to Q-Exactive plus hybrid mass spectrometer (nano-ESI source)</td>
			<td>Thermo Scientific</td>
			<td>0726030+</td>
			<td></td>
		</tr>
		<tr>
			<td>Name</td>
			<td>Company</td>
			<td>Catalog Number</td>
			<td>Comments</td>
		</tr>
		<tr>
			<td>Software, Software Tools, Database search</td>
			<td></td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Mascot</td>
			<td>www.matrixscience.com</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Massign</td>
			<td>http://massign.chem.ox.ac.uk</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>MassLynx</td>
			<td>Micromass</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>MassMatrix</td>
			<td>Xu, H. J Proteome Res. 9 (2010)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>MaxQuant</td>
			<td>Cox, J. Nat. Biotechn. 26 (2008)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>pLink</td>
			<td>Yang, B. Nat Methods 9 (2012)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>pXtract conversion tool</td>
			<td>http://pfind.ict.ac.cn/downloads.html</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>UniDec</td>
			<td>Marty, M.T. Anal. Chem. 87 (2015)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>XCalibur</td>
			<td>Thermo Scientific</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>XiNET</td>
			<td>http://crosslinkviewer.org</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>XLinkX</td>
			<td>Liu, F. Curr Op Struct Biol 35 (2015)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>xQuest</td>
			<td>Rinner, O. Nat Methods 5 (2008)</td>
			<td></td>
			<td></td>
		</tr>
		<tr>
			<td>Xvis</td>
			<td>https://xvis.genzentrum.lmu.de</td>
			<td></td>
			<td></td>
		</tr>
	</tbody>
</table>

combining chemical cross-linking and mass spectrometry of intact protein complexes to study the architecture of multi-subunit protein assemblies

Proteins interact with their ligands to form active and dynamic assemblies which carry out various cellular functions. Elucidating these interactions is therefore fundamental for the understanding of cellular processes. However, many protein complexes are dynamic assemblies and are not accessible by conventional structural techniques. Mass spectrometry contributes to the structural investigation of these assemblies, and particularly the combination of various mass spectrometric techniques delivers valuable insights into their structural arrangement.
In this article, we describe the application and combination of two complementary mass spectrometric techniques, namely chemical cross-linking coupled with mass spectrometry and native mass spectrometry. Chemical cross-linking involves the covalent linkage of amino acids in close proximity by using chemical reagents. After digestion with proteases, cross-linked di-peptides are identified by mass spectrometry and protein interactions sites are uncovered. Native mass spectrometry on the other hand is the analysis of intact protein assemblies in the gas phase of a mass spectrometer. It reveals protein stoichiometries as well as protein and ligand interactions. Both techniques therefore deliver complementary information on the structure of protein-ligand assemblies and their combination proved powerful in previous studies.

The structural analysis of proteins and the complexes they form is fundamental for understanding their function. Mass spectrometry considerably contributes to the structural investigation in that it can be applied to almost every complex of interest irrespective of size or sample heterogeneity. We exemplify the protocol by using two well-characterized protein complexes; first, the hetero-dodecamer RvB1/B2 from C. thermophilum and second, the hetero-octameric CPS complex from E. coli.
First, we identified the protein components of the two complexes. For this, the proteins were separated by SDS-PAGE (Figure 2) and gel bands were cut from the gel. After in-gel digestion of the proteins, the peptide mixture was analyzed by liquid chromatography-coupled mass spectrometry and peptide and fragment masses were subjected to database searching. Following this workflow, we identified all protein subunits of the two complexes with high confidence, i.e., a high number of peptides with reasonable peptide scores was observed yielding high sequence coverage for all protein subunits (Table 1).
We then analyzed the intact RvB1/B2 complex by native mass spectrometry (Figure 3A). The mass spectrum revealed two species, one at approximately 8,000 m/z and another species at approximately 11,000-12,000 m/z. The calculated masses for these species correspond to the hexameric (RvB1)3(RvB2)3 ring (approximately 310 kDa) and the dodecameric double-ring (RvB1)6(RvB2)6 (approximately 620 kDa). Both peak series show two populations; these originate from a mixture of His-tagged and untagged RvB2 subunits in the complexes. A crystal structure for the RvB1/B2 complex was previously obtained36 and shows the arrangement of the double-ring (Figure 3B). The native mass spectrum therefore confirms the stoichiometry of the intact dodecamer and furthermore reveals a stable sub-complex. In addition, co-existing populations are identified.
To identify protein interaction sites in the RvB1/B2 complex, we chemically cross-linked the purified complex with BS3 cross-linker. We first titrated the amount of BS3 during the cross-linking reaction to determine the optimal concentration. BS3 is amine-specific and covalently links lysine side chains as well as the N-termini of proteins. The cross-linking reaction was followed by SDS-PAGE (Figure 2A). The non-cross-linked complex showed both RvB1 and RvB2 subunits. Adding BS3 to the reaction mixture caused covalent linkage of the proteins resulting in protein bands at higher molecular weight. The SDS gel shows that increasing amounts of BS3 yield higher amounts of cross-linked species while non-cross-linked protein subunits are reduced. We then cut the protein bands from the gel and followed the protocol provided above to identify protein interaction sites. An example spectrum of a cross-linked di-peptide is shown (Figure 3C). The spectrum shows y-ion series of both peptides confirming this protein interaction. In total, we obtained 14 protein interactions, including four cross-links between subunits RvB1 and RvB2 and two cross-links between two copies of RvB2 (Table 2). The results from BS3 cross-linking are visualized in an interaction network (Figure 3D) showing intra-molecular interactions as well as interactions between different subunits. Intra-molecular cross-links suggest that both RvB1 and RvB2 subunits fold in a way that N- and C-terminal domains are in close proximity. Note, that intra-molecular interactions cannot be distinguished from inter-molecular interactions of the same subunit in this case. Inter-molecular cross-links between the two subunits were also observed. Of these, two inter-molecular cross-links between RvB1 and RvB2 could be visualized in the structure validating the cross-linking approach. The other inter-molecular cross-links are located in flexible loops which are not included in the crystal structure. We also identified two cross-links in RvB2 containing the same peptide sequences. These cross-links can unambiguously be classified as inter-molecular as they must originate from two copies of the same protein (Figure 3D). Our cross-linking experiments reveal protein interaction sites within the complex but also within the protein subunits providing insights into their structural arrangement that could also be confirmed by the existing crystal structure (Figure 3B).
The second protein complex that we studied was CPS. The native mass spectrum (Figure 4A) revealed three protein complexes between 6,000 and 12,000 m/z. The largest complex of 640 kDa corresponds to the intact hetero-octamer. The smaller complexes represent two sub-complexes; the dimer of the small and large CPS subunits (160 kDa) and a hetero-tetramer containing two copies of each subunit (320 kDa). These sub-complexes deliver first insights into the protein assembly; i.e., the large and small subunits are in direct contact (as revealed by the hetero-dimer) and the tetramer might be a product of two dimers. To gain more information on the structural arrangement in the intact CPS complex, we performed tandem mass spectrometry (MS/MS) of the hetero-octamer and the hetero-tetramer. In both cases, the small subunit dissociated from the precursor suggesting that the small subunit is located in the periphery of the assembly (Figure 4B). Indeed, the small subunit is peripheral in the available crystal structure (Figure 4C)44.
Chemical cross-linking using the BS3 cross-linker was also performed. Using increasing amounts, the covalent linkage of the CPS subunits was enhanced. After digestion of the proteins and analysis of the peptides as described above, many protein interactions within the large subunit and one cross-link in the small subunit were obtained (Table 2). In addition, similar to the RvB1/B2 complex, we found two inter-molecular cross-links between two copies of the large CPS subunit. These cross-links place the two large subunits facing each other at their C-terminal sides. In a previous study, combining structural mass spectrometry and computational modeling35, we identified three additional interactions in the large subunit which most likely originate from the interface of two copies of the large subunit validated by the crystal structure and the obtained model (Figure 4C and Table 2). These interactions allow the arrangement of the CPS core complex consisting of four large subunits. However, no inter-subunit cross-links between the small and the large subunits were observed. By inspecting the available crystal structure (Figure 4C), it becomes apparent that the interaction surface between the tetrameric core of the complex, consisting of the large subunit, and the peripheral small subunits is very small, which might explain the absence of inter-subunit interactions. This is confirmed by native mass spectrometry which showed that the small subunit readily dissociates from the intact complexes most likely due to a small binding interface. Nonetheless, protein interactions in the CPS complex combined from chemical cross-linking and native mass spectrometry allow deducing their structural arrangement (Figure 4D).
Taken together, the combination of native mass spectrometry and chemical cross-linking coupled with mass spectrometric identification of cross-linked peptides, allows reconstitution of the structural arrangement of both example complexes. While chemical cross-linking revealed arrangement of the protein subunits, for instance the interactions between RvB1 and RvB2 or within the tetrameric core of CPS, native mass spectrometry delivered protein stoichiometries of the intact complexes and prevalent subcomplexes. In the case of CPS, for which no inter-molecular interactions between the two subunits could be observed by chemical cross-linking, native mass spectrometry suggests that each large subunit interacts with one small subunit (Figure 4D). Tandem mass spectrometry proposed the peripheral location of the small subunit in the complex and a small interface between both subunits.
<img alt="Figure 1" class="xfigimg" src="/files/ftp_upload/56747/56747fig1.jpg" /> 
Figure 1: Workflow of native mass spectrometry and cross-linking. Both techniques deliver complementary results. While native mass spectrometry reveals stoichiometries and interaction modules, cross-linking gives insights into the protein interaction sites within the complexes. Note that chemical cross-linking only reveals binary interactions. (A) The first step in native mass spectrometry is buffer exchange to a volatile and aqueous buffer using filter units or gel filtration columns. Mass spectrometry of the intact protein complexes then reveals their stoichiometry. In tandem mass spectrometry experiments, peripheral subunits are dissociated. (B) For chemical cross-linking, the protein complex is incubated with a cross-linking reagent. The cross-linked proteins are then digested into peptides which are subsequently analyzed by liquid chromatography-coupled mass spectrometry. <a href="https://cloudflare.jove.com/files/ftp_upload/56747/56747fig1large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 2" class="xfigimg" src="/files/ftp_upload/56747/56747fig2.jpg" /> 
Figure 2: SDS-PAGE of cross-linked RvB1/B2 (A) and CPS (B) complexes. (A) 2.5 &#181;M RvB1/B2 were loaded per gel lane. The concentration of BS3 was varied. Non-cross-linked RvB1/B2 shows the two protein subunits at approximately 50 kDa. Addition of BS3 caused covalent linkage of the protein subunits resulting in protein bands at higher molecular weight. The amount of cross-linked species is increased with higher BS3 concentrations. Optimal cross-linking conditions are highlighted (red). (B) 10 &#181;M CPS were loaded per gel lane. The large (90 kDa) and small (40 kDa) CPS subunits are obtained. Addition of BS3 caused covalent linkage of the protein subunits resulting in protein bands at higher molecular weight. Optimal cross-linking conditions are highlighted (red). <a href="https://cloudflare.jove.com/files/ftp_upload/56747/56747fig2large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 3" class="xfigimg" src="/files/ftp_upload/56747/56747fig3.jpg" /> 
Figure 3: Native mass spectrometry and chemical cross-linking of RvB1/B2 complex. (A) The native mass spectrum reveals two species of RvB1/B2; the intact dodecamer (i.e., (RvB1)6(RvB2)6) at approximately 11,000 to 12,000 m/z and the hexameric ring (RvB1)3(RvB2)3 at approximately 8,000 m/z. Both species show two populations resulting from His-tagged and untagged RvB2. The spectrum has been modified from35. (B) The crystal structure of RvB1/B2 is shown (PDB ID 4WVY). Alternating RvB1 and RvB2 subunits form two hexameric rings. (C) Fragmentation spectrum of a cross-linked di-peptide. The N-terminus of RvB1 was cross-linked with K23 of RvB1. y-ion series were obtained for both peptides (red and cyan). (D) Intra- and inter-protein interactions obtained in the RvB1/B2 complex. Intra-cross-links are shown in red, inter-cross-links are shown in blue. The insert shows two inter-molecular cross-links between RvB1 and RvB2 subunits which could be visualized in the crystal structure (green, insert). Interactions that originate from two RvB2 copies are shown as blue dotted lines. <a href="https://cloudflare.jove.com/files/ftp_upload/56747/56747fig3large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
<img alt="Figure 4" class="xfigimg" src="/files/ftp_upload/56747/56747fig4.jpg" /> 
Figure 4: Native mass spectrometry and chemical cross-linking of CPS. (A) The native mass spectrum of CPS shows three complexes. The hetero-dimer (160 kDa), hetero-tetramer (320 kDa) and the hetero-octamer (640 kDa). The spectrum has been modified from35. (B) Tandem mass spectrometry of the tetrameric and octameric CPS complex revealed dissociation of the small CPS subunit. (C) The crystal structure of CPS is shown (PDB ID 1BXR). The large subunits form a tetrameric core and the small subunits are located in the periphery of the complex. Inter-molecular cross-links between two copies of the large subunit are shown (green). (D) Interactions of the large and small CPS subunits. Native mass spectrometry revealed subcomplexes and suggests a peripheral location of the small subunit. Chemical cross-links indicate arrangements in the tetrameric core of CPS. <a href="https://cloudflare.jove.com/files/ftp_upload/56747/56747fig4large.jpg" target="_blank">Please click here to view a larger version of this figure.</a>
Table 1: Database search results. The proteins were identified by liquid chromatography-coupled mass spectrometry and database searching. The protein names, accession number, and description as well as protein mass are given. The protein score, number of observed spectra per protein, and the number of observed peptide sequences are listed. The five peptides with highest Mascot peptides scores are listed for each protein subunit. <a href="http://cloudflare.jove.com/files/ftp_upload/56747/Table1.xlsx">Please click here to download this file.</a>
Table 2: Cross-links observed in RvB1/B2 and CPS. The subunits of the complexes and the cross-linked residues are given. The type of cross-link (intra- or inter-molecular) was revealed from overlapping peptide sequences or a previous study35. <a href="http://cloudflare.jove.com/files/ftp_upload/56747/Table2.xlsx">Please click here to download this file.</a>

Watch this Scientific Journal Video about Combining Chemical Cross-linking and Mass Spectrometry of Intact Protein Complexes to Study the Architecture of Multi-subunit Protein Assemblies at JoVE.com

Combining Chemical Cross-linking and Mass Spectrometry of Intact Protein Complexes to Study the Architecture of Multi-subunit Protein Assemblies

The architecture of protein complexes is essential for their function. Combining various mass spectrometric techniques proved powerful to study their assembly. We provide protocols for chemical cross-linking and native mass spectrometry and show how these complementary techniques help to elucidate the architecture of multi-subunit protein assemblies.

Proteins interact with their ligands to form active and dynamic assemblies which carry out various cellular functions. Elucidating these interactions is ...

Research

Education

JoVE Journal

JoVE Core

Molecular Biology

Biology

Unit 1

Protein Structure

SCIENTISTS IN ACTION

MALDI Sample Preparation: the Ultra Thin Layer Method

This video demonstrates the preparation of an ultra-thin matrix/analyte layer for analyzing peptides and proteins by Matrix-Assisted Laser Desorption ...

Staining of Proteins in Gels with Coomassie G-250 without Organic Solvent and Acetic Acid

In classical protein staining protocols using Coomassie Brilliant Blue (CBB), solutions with high contents of toxic and flammable organic solvents ...

Using the GELFREE 8100 Fractionation System for Molecular Weight-Based Fractionation with Liquid Phase Recovery

The GELFREE 8100 Fractionation System is a novel protein fractionation system designed to maximize protein recovery during molecular weight based ...

Isolation of Labile Multi-protein Complexes by in vivo Controlled Cellular Cross-Linking and Immuno-magnetic Affinity Chromatography

Isolation of Labile Multi-protein Complexes by in vivo Controlled Cellular Cross-Linking and Immuno-magnetic Affinity Chromatography

The dynamic nature of cellular machineries is frequently built on transient and/or weak protein associations. These low affinity interactions preclude ...

Analyzing Large Protein Complexes by Structural Mass Spectrometry

Living cells control and regulate their biological processes through the coordinated action of a large number of proteins that assemble themselves into an ...

Isolation and Biophysical Study of Fruit Cuticles

The cuticle, a hydrophobic protective layer on the aerial parts of terrestrial plants, functions as a versatile defensive barrier to various biotic and ...

Identification of Protein Complexes in Escherichia coli using Sequential Peptide Affinity Purification in Combination with Tandem Mass Spectrometry

Identification of Protein Complexes in Escherichia coli using Sequential Peptide Affinity Purification in Combination with Tandem Mass Spectrometry

Since most cellular processes are mediated by macromolecular assemblies, the systematic identification of protein-protein interactions (PPI) and the ...

A Protocol for the Identification of Protein-protein Interactions Based on 15N Metabolic Labeling, Immunoprecipitation, Quantitative Mass Spectrometry and Affinity Modulation

A Protocol for the Identification of Protein-protein Interactions Based on 15N Metabolic Labeling, Immunoprecipitation, Quantitative Mass Spectrometry and Affinity Modulation

Protein-protein interactions are fundamental for many biological processes in the cell. Therefore, their characterization plays an important role in ...