Mass spectrometry (MS) has long been used to study proteins mainly via sequence identification and quantitation of expression abundance. In recent years, MS has emerged as a tool for structural biology. Intact protein structural analysis has been enabled by the development of methods such as native MS, top-down proteomics, and ion mobility MS. Other MS-based structural methods include affinity purification MS, chemical cross-linking, and protein footprinting. These methods have enabled the study of protein–protein and protein–ligand interactions and regions of conformational change. The coupling of MS with liquid chromatography has permitted the analysis of complex samples. This bottom-up proteomics workflow enables the study of protein structure in the native cellular environment and provides structural information across the proteome. It has been demonstrated that the crowded environment of the cell affects protein binding interactions and affinities. Performing studies in this complex environment is essential for understanding the functional roles of proteins. MS-based structural methods permit analysis of samples such as cell lysates, intact cells, and tissue to provide a more physiological view of protein structure. This mini-review discusses the various MS-based methods that can be used for proteome-wide structural studies and highlights some of their application.

Introduction

The development of new methods has paved the way for mass spectrometry (MS) to be used as a tool for structural biology [1]. MS has evolved from the identification and quantitation of proteins to study their interactions and conformations. The emergence of methods such as native MS, ion mobility, chemical cross-linking, and protein footprinting has allowed for the analysis of oligomerization states, conformations, and interaction networks [2–9]. Requiring only microgram quantities of proteins, the sensitivity of MS makes it advantageous for structural measurements. Protein size is also not a limiting factor with native MS studies on intact protein aided by developments in MS instrumentation that expands the m/z analysis range allowing for the study of purified proteins and protein complexes [10]. In some cases, such as chemical cross-linking and protein footprinting, MS analysis is performed on proteolyzed proteins. These methods involve a chemical reaction that modifies a protein in some manner prior to MS analysis. The structural information that is encoded in the modification differences is retained after proteolysis and detected by liquid chromatography coupled to tandem MS (LC–MS/MS) analysis. This bottom-up workflow enables the analysis of thousands of proteins in a single experiment. The coupling of chromatographic separation with MS of proteolyzed proteins allows for the analysis of complex systems such as cell lysates, intact cells, and tissue. These methods enable proteome-wide structural studies, a feat that is not possible with other structural methods.

The crowded cellular environment

The complexity of the cellular environment plays a role in the stability and ligand affinities of proteins [11]. The high concentration of macromolecules in the cell, 80–400 mg/ml [12–14] with proteins occupying 15–35% of the cell volume [15], leads to transient interactions amongst neighboring proteins. This quinary structure is essential for cellular organization [16]. Owing to the significant effect the crowded cellular environment has on protein structure, it is imperative to study proteins directly in cells. The rise of cellular-based methods such as fluorescence imaging and in-cell NMR has enabled the study of protein interactions and conformations in the cell [17–19]. However, the MS-based bottom-up proteomics workflow is advantageous due to its ability to provide structural information on hundreds to thousands of proteins across the proteome. These proteome-wide structural methods have been reviewed in detail elsewhere [20]. Here, we will briefly discuss these methods and highlight some applications. Although several of these methods provide similar structural information, each method has some unique advantages (Table 1).

Table 1.
Summary of proteome-wide structural methods
MethodApplicationsAdvantagesLimitations
TPP Drug target discovery; Protein interaction network analysis High proteome coverage provides abundant structural information Thermal stability of some proteins does not change in response to ligand binding 
LiP Protein conformation differences; Drug target discovery Specifically detects peptides involved in protein interactions and conformational change Requires separation of limited proteolyzed fragments from tryptic peptides. 
SPROX Drug target discovery; Biological state analysis Quantification of biophysical properties of protein interactions (binding free energies, Kd values) Requires the identification of methionine-containing proteins 
XL-MS Protein interaction network analysis Analysis of cell lysates, intact cells, and tissue Difficulty in the identification of cross-linked peptides 
IV-FPOP Protein interactions; protein conformational change Structural studies in live animals Lack of enrichment strategy for oxidatively modified peptides reduces proteome coverage 
MethodApplicationsAdvantagesLimitations
TPP Drug target discovery; Protein interaction network analysis High proteome coverage provides abundant structural information Thermal stability of some proteins does not change in response to ligand binding 
LiP Protein conformation differences; Drug target discovery Specifically detects peptides involved in protein interactions and conformational change Requires separation of limited proteolyzed fragments from tryptic peptides. 
SPROX Drug target discovery; Biological state analysis Quantification of biophysical properties of protein interactions (binding free energies, Kd values) Requires the identification of methionine-containing proteins 
XL-MS Protein interaction network analysis Analysis of cell lysates, intact cells, and tissue Difficulty in the identification of cross-linked peptides 
IV-FPOP Protein interactions; protein conformational change Structural studies in live animals Lack of enrichment strategy for oxidatively modified peptides reduces proteome coverage 

Cell lysates

Several MS-based structural methods use a denaturant or protease to report information on protein structure. These methods, initially used on purified protein systems, have been adapted for studies on cell lysates. Examples of these include limited proteolysis (LiP) [21] and pulse proteolysis [22] which both utilize proteolytic digestion to report on protein structure and stability of proteins from rates of oxidation (SPROX) which combines chemical denaturation with oxidative labeling of methionine [23]. The LiP workflow utilizes a double-digestion protocol where after a native protein extraction, samples are first subjected to digestion with a broad-specificity protease and then a complete digestion with trypsin (Figure 1). A comparison of the LiP sample peptides with samples only subjected to trypsin digest reveals which regions of proteins are accessible to digestion. When two conditions are compared, such as comparison of the apo and holo state of a protein, the differences in digestion detail the structural differences between states. Initial LiP studies on mammalian cell lysates identified ∼3500 LiP cleavage sites and ∼1200 proteins. A large-scale LiP study performed on yeast grown in glucose- compared with ethanol-based media yielded 4267 LiP sites across 1001 proteins [24]. Of the 283 proteins that displayed at least a two-fold abundance difference in LiP peptides between the two metabolic states, 8% were at known or predicted protein–protein interactions sites. Most of these 283 proteins were in the glycolytic and carbon metabolism pathways demonstrating the ability of the method to identify proteins that undergo structural changes upon the metabolic transition from glucose- to ethanol-based growth.

LiP-SRM workflow.

Figure 1.
LiP-SRM workflow.

In the first step of the LiP-SRM workflow, a proteome is extracted from cells under nondenaturing, native conditions and limited proteolysis is conducted with a broad-specificity protease, such as proteinase K, for a short time. Red arrows indicate limited proteolysis sites. Protease activity is quenched by shifting the proteome to denaturing conditions. A complete trypsin digestion is then performed. A fraction of the same sample is only subjected to the trypsin step under denaturing conditions. The samples are quantitatively analyzed by MS and levels of the resulting and peptides are compared. A fully tryptic peptide containing a LiP cleavage site will be detected in the trypsin control and replaced by two half-tryptic halves in the sample subjected to LiP. Each proteome (biological triplicates) undergoes the double-digestion step to detect protein structural differences in differently treated proteome extracts and the resulting digests are compared. Discovery-based shotgun LC–MS/MS analyses, followed by SRM validation or hypothesis-based SRM alone, are applied to the analysis, in case of discovery or targeted applications, respectively. Reproduced with permission from Feng et al. [24].

Figure 1.
LiP-SRM workflow.

In the first step of the LiP-SRM workflow, a proteome is extracted from cells under nondenaturing, native conditions and limited proteolysis is conducted with a broad-specificity protease, such as proteinase K, for a short time. Red arrows indicate limited proteolysis sites. Protease activity is quenched by shifting the proteome to denaturing conditions. A complete trypsin digestion is then performed. A fraction of the same sample is only subjected to the trypsin step under denaturing conditions. The samples are quantitatively analyzed by MS and levels of the resulting and peptides are compared. A fully tryptic peptide containing a LiP cleavage site will be detected in the trypsin control and replaced by two half-tryptic halves in the sample subjected to LiP. Each proteome (biological triplicates) undergoes the double-digestion step to detect protein structural differences in differently treated proteome extracts and the resulting digests are compared. Discovery-based shotgun LC–MS/MS analyses, followed by SRM validation or hypothesis-based SRM alone, are applied to the analysis, in case of discovery or targeted applications, respectively. Reproduced with permission from Feng et al. [24].

In SPROX, the oxidative labeling of methionine residues provides information on solvent accessibility differences that occur due to ligand binding or conformational change. Samples are exposed to hydrogen peroxide in the presence of increasing concentrations of a chemical denaturant with both urea and guanidium chloride used depending on the application [25]. The stability of the protein is measured based on the concentration of denaturant at which the protein unfolds. This concentration will shift upon ligand binding or conformational change. Unlike LiP, which utilizes a more targeted selected reaction monitoring MS to identify digested peptides, SPROX utilizes a shotgun proteomics approach for analysis [25]. However, to increase the structural information, methionine-containing peptides are enriched prior to MS analysis. Enrichment of unoxidized methionine-containing peptides led to a five-fold increase in the frequency of their identification improving the utility of the method for proteome-wide studies [25]. SPROX has been used to determine stability changes in proteins in MCF-7 cells, breast cancer cells, after treatment with tamoxifen (TAM) and its metabolite n-desmethyl tamoxifen (NDT) [26]. Two quantitative proteomics methods, stable isotope labeling with amino acids in cell culture (SILAC) and isobaric tags for relative and absolute quantitation (iTRAQ), were coupled with SPROX for this study. In the NDT study, the high-confidence protein hit TUBB3 demonstrates stability changes in peptide LATPTYGDLNHLVSATMSGVTTSLR (Figure 2). Altogether, 11 and 20 proteins were identified as high-confidence hits in the TAM and NDT experiments, respectively. A subset of these, 27 protein hits, were reproducibly identified with both SILAC-SPROX and iTRAQ-SPROX. Approximately 85% of these 27 protein hits are linked to breast cancer demonstrating the specificity of the method. A difficulty in the analysis of SPROX data is distinguishing whether observed differences are due to alterations in direct binding or in allosteric effects. To distinguish these for the Y-box binding protein (YBX1), which displayed differences in unfolding upon TAM binding, pulse proteolysis, and phenacyl bromide SPROX (PAB-SPROX) were performed on purified YBX1 in the presence and absence of TAM. Both methods showed differences between the two conditions verifying the original SPROX data and confirming the difference in unfolding is due to direct binding of TM to YBX1 [26].

SPROX study in breast cancer cells.

Figure 2.
SPROX study in breast cancer cells.

(A,B) iTRAQ- and SILAC-SPROX data (respectively) collected on the methionine-containing peptide LATPTYGDLNHL-VSATMSGVTTSLR from TUBB3 in the NDT binding study. In (A), the dark- and light-shaded data points represent the data collected in the presence and in the absence of NDT, respectively, and the solid lines represent the best fit of the data. The data points indicated with an ‘X” were not included in the fit. In (B), the date points were fit and the solid and dotted curves represent the theoretical SPROX curves generated from the SILAC-SPROX data. In (A) and (B), the dotted vertical lines indicate the C1/2 values of the SPROX curves. Reproduced with permission from Ogburn et al. [26].

Figure 2.
SPROX study in breast cancer cells.

(A,B) iTRAQ- and SILAC-SPROX data (respectively) collected on the methionine-containing peptide LATPTYGDLNHL-VSATMSGVTTSLR from TUBB3 in the NDT binding study. In (A), the dark- and light-shaded data points represent the data collected in the presence and in the absence of NDT, respectively, and the solid lines represent the best fit of the data. The data points indicated with an ‘X” were not included in the fit. In (B), the date points were fit and the solid and dotted curves represent the theoretical SPROX curves generated from the SILAC-SPROX data. In (A) and (B), the dotted vertical lines indicate the C1/2 values of the SPROX curves. Reproduced with permission from Ogburn et al. [26].

Studies in intact cells

Studies in intact cells are necessary to truly study quinary structure because cell disruption methods may perturb these weak interactions [16]. An example of a proteome-wide method that can be used to study intact cells is thermal proteome profiling (TPP). TPP is based on the fact that proteins in the cell denature and aggregate near their melting temperature (Tm). After exposure of cells to a range of temperatures and cell lysis, MS is used to detect the loss of protein in the tryptic digest of the soluble cell fraction. The Tm of proteins can then be extracted from these data. The method has been used on both cell lysates and intact cells [27]. A comparison study showed proteins had 2.7° higher Tm values in cell extracts than in intact cells suggesting cell disruption alters protein stability [28].

Owing to ligand binding stabilizing proteins, thus increasing their melting temperature, TPP is especially suited to study this process including for studying drug-target engagement and regulatory roles of small molecules [28,29]. A study of the treatment of K562 cells with panobinostat, a histone deacetylase (HDAC) inhibitor, demonstrated shifts in thermal stability for five HDAC proteins (HDAC 1, 2, 6, 8, and 10) and other proteins including TTC38, syntaxin-4 (STX4), and zinc-finger and the H2A histone family member V or Z protein (H2AFV or H2AFZ) (Figure 3) [30]. A variant of TPP called compound concentration range (TPP-CCR), where the temperature is kept constant and the concentration of ligand is varied, was also used as verification of the TPP results. The results showed that STX4 did not display dose-dependent behavior suggesting the result from the TPP experiment with varied temperature was a false-positive. All other proteins that demonstrated changes in TPP also showed differences in TPP-CCR. In this study, a total of 6004 proteins were identified establishing the utility of TPP as a method for proteome-wide structural studies.

TPP for drug target engagement.

Figure 3.
TPP for drug target engagement.

(A) Scatter plot of Tm shifts calculated from the two biological replicates of the panobinostat versus vehicle treatment experiment. Panobinostat-induced Tm shifts that passed the significance criteria are shown in red. The points corresponding to the HDAC1 and HDAC2 proteins are colored orange; both proteins passed the significance requirement in one biological replicate and were just outside the threshold for significance in the other. (B–I) Melting curves of proteins for which strong and significant or borderline significant thermal stabilization was observed when treating cells with panobinostat; HDAC 1, 2, 6, 8 and 10 (B–F); TTC38 (G); STX4 (H); and H2AFV or H2AFZ (I). Reproduced with permission from Franken et al. [30].

Figure 3.
TPP for drug target engagement.

(A) Scatter plot of Tm shifts calculated from the two biological replicates of the panobinostat versus vehicle treatment experiment. Panobinostat-induced Tm shifts that passed the significance criteria are shown in red. The points corresponding to the HDAC1 and HDAC2 proteins are colored orange; both proteins passed the significance requirement in one biological replicate and were just outside the threshold for significance in the other. (B–I) Melting curves of proteins for which strong and significant or borderline significant thermal stabilization was observed when treating cells with panobinostat; HDAC 1, 2, 6, 8 and 10 (B–F); TTC38 (G); STX4 (H); and H2AFV or H2AFZ (I). Reproduced with permission from Franken et al. [30].

Chemical cross-linking provides structural information in heart tissue

Cross-linking mass spectrometry (XL-MS) entails the initiation of covalent linkages, both intra- and intermolecular, between amino acid side chains. Cross-linkers of various chemistries make it possible to study a wide variety of protein systems [31]. Structurally, the method provides distance restraints between cross-linked peptides providing insights on protein interaction networks. The development of MS-cleavable cross-linkers has increased the utility of the XL-MS for complex systems [32]. XL-MS has been used on purified proteins [33], isolated organelles [34], cell extracts [35], and intact cells [36].

One particular type of cross-linker based on protein interaction reporter (PIR) technologies has further increased the complexity of samples that XL-MS can be used to study [37]. The PIR technology utilizes a MS-cleavable cross-linker coupled with a mass-encoded reporter and affinity group. This allows for enrichment of the cross-linked peptides which are present in low abundance. PIR-based cross-linkers have been used to study protein interactions in heart tissue, the most complex system studied to date by XL-MS [38]. The heart was excised from anesthetized mice and the tissue was minced into 1 mm cubes for the cross-linking reaction. In total, 2026 cross-linked lysine pairs in 316 proteins lead to the identification of 419 non-redundant protein–protein interactions. These cross-linked peptide pairs were used as constraints on a structural model of the sarcomeric proteins, myosin, actin, troponin, and tropomyosin (Figure 4). Myosin 6 in the sarcomere is a highly flexible molecule, whose multiple ensembles were detected by XL-MS with 10 links clustered in the three flexible regions of this protein (Figure 4A, numbered in yellow). This study also identified interactions amongst the complexes in the oxidative phosphorylation pathway.

XL-MS in heart tissue.

Figure 4.
XL-MS in heart tissue.

(A) Structural model of sarcomere protein interactions including the thick filament proteins: myosin motor (MYH6, red), myosin essential light chain (MYL3, teal blue), myosin regulatory light chain (MLRV, blue-violet), and the thin filament proteins: actin (ACTA green and red-violet), tropomyosin (yellow), and the three troponin subunits (TNNC1, light blue; TNNT2, blue; TNNI3, dark blue). The model utilizes structural information from structures PDB: 5H53, 5CJ1, 4XA4, 5CHX, 5CJ0, 1J1E, and 5JLH. The second MYH6 molecule is represented by a gray semi-transparent structure. Cross-linked sites are indicated by green space-filled residues, and links between residues are displayed as gray bars (Cα–Cα distance < 42 Å) or yellow bars (Cα–Cα distance > 42 Å). A total of 10 cross-links (2 MYH6-MLRV and 8 MYH6-MYL3) exceed 42 Å. These links are clustered around three flexible hinge regions indicated by numbered yellow circles in the MYH6 light chain binding domain. It is likely that these links are representative of the conformational flexibility of the myosin motor and are formed from a different conformation than the rigor state model shown here. Forty-six homodimer cross-linked peptide pairs were also identified in the coiled-coil domain of MYH6. (B) Zoomed inset of a 90° rotation of the structure shown in (A). Calcium ions bound to TNNC1 are shown as magenta spheres. The yellow links between K165 of TNNI3 and K330 of ACTA and K43 of TNNC1 and K330 of ACTA exceed the possible cross-linkable distance and are not compatible with the calcium-saturated structure of troponin (PDB: 1J1E), and are instead indicative of the calcium-depleted state of troponin in which TNNI3 changes conformation to interact with ACTA, effectively blocking the myosin interaction site. Reproduced with permission from Chavez et al. [38].

Figure 4.
XL-MS in heart tissue.

(A) Structural model of sarcomere protein interactions including the thick filament proteins: myosin motor (MYH6, red), myosin essential light chain (MYL3, teal blue), myosin regulatory light chain (MLRV, blue-violet), and the thin filament proteins: actin (ACTA green and red-violet), tropomyosin (yellow), and the three troponin subunits (TNNC1, light blue; TNNT2, blue; TNNI3, dark blue). The model utilizes structural information from structures PDB: 5H53, 5CJ1, 4XA4, 5CHX, 5CJ0, 1J1E, and 5JLH. The second MYH6 molecule is represented by a gray semi-transparent structure. Cross-linked sites are indicated by green space-filled residues, and links between residues are displayed as gray bars (Cα–Cα distance < 42 Å) or yellow bars (Cα–Cα distance > 42 Å). A total of 10 cross-links (2 MYH6-MLRV and 8 MYH6-MYL3) exceed 42 Å. These links are clustered around three flexible hinge regions indicated by numbered yellow circles in the MYH6 light chain binding domain. It is likely that these links are representative of the conformational flexibility of the myosin motor and are formed from a different conformation than the rigor state model shown here. Forty-six homodimer cross-linked peptide pairs were also identified in the coiled-coil domain of MYH6. (B) Zoomed inset of a 90° rotation of the structure shown in (A). Calcium ions bound to TNNC1 are shown as magenta spheres. The yellow links between K165 of TNNI3 and K330 of ACTA and K43 of TNNC1 and K330 of ACTA exceed the possible cross-linkable distance and are not compatible with the calcium-saturated structure of troponin (PDB: 1J1E), and are instead indicative of the calcium-depleted state of troponin in which TNNI3 changes conformation to interact with ACTA, effectively blocking the myosin interaction site. Reproduced with permission from Chavez et al. [38].

MS-based studies in an animal model for human disease

In some cases, pathogenesis of human diseases involves the dynamic interplay between molecular and cellular systems. The complexity of these interacting systems cannot be recapitulated in cell culture where only single cell types are present. This underscores the need to study protein conformations in animal models to further understand complex disease states. The protein footprinting method fast photochemical oxidation of proteins (FPOP) has recently been extended to study protein interactions and conformations in Caenorhabditis elegans, an animal model for human disease [39]. FPOP utilizes hydroxyl radicals to oxidatively modify the side chains of solvent accessible amino acids [40]. Hydroxyl radicals are generated via photolysis of hydrogen peroxide using an excimer laser at 248 nm. FPOP is a general labeling strategy with the ability to label 19 of the 20 amino acids. The method was first extended for analysis of proteins in intact cells and was shown to modify over a thousand proteins in various cellular organelles [41,42]. The transparency of C. elegans permits the laser irradiation to penetrate the animal to generate hydroxyl radicals in situ. The worms can both ingest hydrogen peroxide and absorb it through their skin.

In initial in vivo FPOP (IV-FPOP) studies, over 500 proteins from various body systems throughout the worm were oxidatively modified [39]. Residue-level quantification of the extent of oxidation demonstrates the ability of the method to probe solvent accessibility in the worms similar to in vitro FPOP. Four residues in the UNC-45 binding peptide on Hsp90 were oxidatively modified by IV-FPOP. Although each residue displayed a +16 modification, chromatographic separation was able to distinguish the isomeric peptides and MS/MS spectra were obtained for each individual modification (Figure 5). The extent of modification for each residue was normalized to residue reactivity with hydroxyl radicals. This protection factor ln(PF) correlates hydroxyl radical labeling directly to structural information [43]. The normalized extent of modification data correlates very well to the solvent accessible surface area calculated from the crystal structure of the C. elegans Hsp90 peptide bound to UNC-45 (Figure 5C). M698 has the highest level of protection (less solvent accessible) in comparison with the other residues. In the crystal structure, the sidechain of M698 is directed inward into the UNC-45 protein limiting its solvent accessibility (Figure 5A, inset). In some cases, residue-level quantification cannot be achieved. This is evident for two peptides in the UNC-45 protein where the multiple modified residues are not separated chromatographically (Figure 5B). For these, peptide-level quantification can be performed to provide structural information.

Correlating IV-FPOP modification to solvent accessibility.

Figure 5.
Correlating IV-FPOP modification to solvent accessibility.

(A) Myosin chaperon protein UNC-45 (gray) (PDB ID 4I2Z) highlighting two modified peptides identified by LC/MS/MS analysis, 669−680 and 698−706 (green, left inset). UNC-45 is bound to the Hsp90 peptide fragment (blue). Oxidatively modified residues within this fragment are shown in sticks (red), and UNC-45 is rendered as a surface (right inset). (B) Tandem MS spectra of UNC-45 peptide 669−680 (top) and 698−706 (bottom) showing b- and y-ions for the loss of CO2, an FPOP modification. (C) The calculated ln(PF) for the Hsp90 oxidatively modified residues, R697, M698, E699, and E700. (D) Tandem MS spectra for R697, M698, E699, and E700 showing a +16 FPOP modification. Reproduced with permission from Espino et al. [39].

Figure 5.
Correlating IV-FPOP modification to solvent accessibility.

(A) Myosin chaperon protein UNC-45 (gray) (PDB ID 4I2Z) highlighting two modified peptides identified by LC/MS/MS analysis, 669−680 and 698−706 (green, left inset). UNC-45 is bound to the Hsp90 peptide fragment (blue). Oxidatively modified residues within this fragment are shown in sticks (red), and UNC-45 is rendered as a surface (right inset). (B) Tandem MS spectra of UNC-45 peptide 669−680 (top) and 698−706 (bottom) showing b- and y-ions for the loss of CO2, an FPOP modification. (C) The calculated ln(PF) for the Hsp90 oxidatively modified residues, R697, M698, E699, and E700. (D) Tandem MS spectra for R697, M698, E699, and E700 showing a +16 FPOP modification. Reproduced with permission from Espino et al. [39].

A new platform for in-cell and IV-FPOP was recently developed that further increases the utility of each method. Whereas these methods were originally used using a flow system for the cells and worms, this new system, entitled platform incubator with XY movement (PIXY), is a static platform (Figure 6). PIXY consists of a stage-top incubator with a movable stage. The temperature-controlled incubator enables cell culture and worm growth directly at the optical bench in six-well plates. The stage moves the incubator so that each well can be individually irradiated by the laser. Peristaltic pumps are utilized to infuse hydrogen peroxide and quench solutions. IV-FPOP using PIXY increased the number of oxidatively modified proteins by 1.5-fold in comparison with the flow system [44]. Furthermore, 71% of the proteins modified with both PIXY and the flow system had a higher number of modified residues when labeled using PIXY. This higher modification coverage provides more structural information on individual proteins.

Schematic of the PIXY system.

Figure 6.
Schematic of the PIXY system.

The system consists of a stage-top incubator (A), XY movable stage (B), and four peristaltic pumps (C) equipped with perfusion (D) lines for chemical transfer. Cell culture media is withdrawn from each well. DPBS, hydrogen peroxide, and quench solutions are infused into each well as indicated by the colored arrows. White lines indicate the laser path for irradiation. Reproduced with permission from Johnson et al. [44].

Figure 6.
Schematic of the PIXY system.

The system consists of a stage-top incubator (A), XY movable stage (B), and four peristaltic pumps (C) equipped with perfusion (D) lines for chemical transfer. Cell culture media is withdrawn from each well. DPBS, hydrogen peroxide, and quench solutions are infused into each well as indicated by the colored arrows. White lines indicate the laser path for irradiation. Reproduced with permission from Johnson et al. [44].

Conclusions

MS-based methods have advanced the field of structural biology via the analysis of complex samples that better reflect physiological conditions. By utilizing the well-established bottom-up MS workflow, these methods can provide structural information across the proteome. Currently, these methods have demonstrated the ability to acquire structural information on a few hundred to several thousand proteins in a single experiment. To date, no other structural method can report on such a large number of proteins.

Perspectives

  • Importance of the field: The ability to obtain structural information across the proteome allows for a better understanding of the role of quinary structure and cellular crowding on protein interactions and binding affinities. This will aid in our understanding of their biological role and their link to disease pathogenesis.

  • Current thinking: It is now recognized that these MS-based structural methods can be used to study complex systems such as lysates and intact cells. These methods can be used to identify protein interaction networks, determine protein stability, and detect conformational changes.

  • Future directions: Currently, these methods are only used by a few laboratories. A more general adoption of these methods requires further development. A major bottleneck of most of these methods is in data analysis. Development of new software platforms that streamline the data analysis would increase the accessibility of these methods. In addition, automation, where possible, would also be beneficial for the adoption of these methods to a wider community.

Competing Interests

The authors declare that there are no competing interests associated with this manuscript.

Abbreviations

     
  • CCR

    compound concentration range

  •  
  • FPOP

    fast photochemical oxidation of proteins

  •  
  • IV-FPOP

    in vivo fast photochemical oxidation of proteins

  •  
  • LC–MS/MS

    liquid chromatography coupled to tandem mass spectrometry

  •  
  • LiP

    limited proteolysis

  •  
  • ln(PF)

    protection factor

  •  
  • MS

    mass spectrometry

  •  
  • NDT

    n-desmethyl tamoxifen

  •  
  • PIR

    protein interaction reporter

  •  
  • PIXY

    platform incubator with XY movement

  •  
  • SPROX

    stability of proteins from rates of oxidation

  •  
  • TAM

    tamoxifen

  •  
  • Tm

    melting temperature

  •  
  • TPP

    thermal proteome profiling

  •  
  • XL-MS

    cross-linking mass spectrometry

References

References
1
Kaur
,
U.
,
Johnson
,
D.T.
,
Chea
,
E.E.
,
Deredge
,
D.J.
,
Espino
,
J.A.
and
Jones
,
L.M.
(
2019
)
Evolution of structural biology through the lens of mass spectrometry
.
Anal. Chem.
91
,
142
155
2
van de Waterbeemd
,
M.
,
Fort
,
K.L.
,
Boll
,
D.
,
Reinhardt-Szyba
,
M.
,
Routh
,
A.
,
Makarov
,
A.
et al (
2017
)
High-fidelity mass analysis unveils heterogeneity in intact ribosomal particles
.
Nat. Methods
14
,
283
286
3
Walker
,
L.R.
,
Marzluff
,
E.M.
,
Townsend
,
J.A.
,
Resager
,
W.C.
and
Marty
,
M.T.
(
2019
)
Native mass spectrometry of antimicrobial peptides in lipid nanodiscs elucidates complex assembly
.
Anal. Chem.
91
,
9284
9291
4
Upton
,
R.
,
Migas
,
L.G.
,
Pacholarz
,
K.J.
,
Beniston
,
R.G.
,
Estdale
,
S.
,
Firth
,
D.
et al (
2019
)
Hybrid mass spectrometry methods reveal lot-to-lot differences and delineate the effects of glycosylation on the tertiary structure of Herceptin(R)
.
Chem. Sci.
10
,
2811
2820
5
Tian
,
Y.
,
Lippens
,
J.L.
,
Netirojjanakul
,
C.
,
Campuzano
,
I.D.G.
and
Ruotolo
,
B.T.
(
2019
)
Quantitative collision-induced unfolding differentiates model antibody-drug conjugates
.
Protein Sci.
28
,
598
608
6
Kastritis
,
P.L.
,
O'Reilly
,
F.J.
,
Bock
,
T.
,
Li
,
Y.
,
Rogon
,
M.Z.
,
Buczak
,
K.
et al (
2017
)
Capturing protein communities by structural proteomics in a thermophilic eukaryote
.
Mol. Syst. Biol.
13
,
936
7
Haupl
,
B.
,
Ihling
,
C.H.
and
Sinz
,
A.
(
2017
)
Combining affinity enrichment, cross-linking with photo-amino acids, and mass spectrometry for probing protein kinase D2 interactions
.
Proteomics
17
,
e1600459
8
Cornwell
,
O.
,
Radford
,
S.E.
,
Ashcroft
,
A.E.
and
Ault
,
J.R.
(
2018
)
Comparing hydrogen deuterium exchange and fast photochemical oxidation of proteins: a structural characterisation of wild-type and ΔN6 β2-microglobulin
.
J. Am. Soc. Mass Spectrom.
29
,
2413
2426
9
Gustavsson
,
M.
,
Wang
,
L.
,
van Gils
,
N.
,
Stephens
,
B.S.
,
Zhang
,
P.
,
Schall
,
T.J.
et al (
2017
)
Structural basis of ligand interaction with atypical chemokine receptor 3
.
Nat. Commun.
8
,
14135
10
Fort
,
K.L.
,
van de Waterbeemd
,
M.
,
Boll
,
D.
,
Reinhardt-Szyba
,
M.
,
Belov
,
M.E.
,
Sasaki
,
E.
et al (
2017
)
Expanding the structural analysis capabilities on an Orbitrap-based mass spectrometer for large macromolecular complexes
.
Analyst
143
,
100
105
11
Gierasch
,
L.M.
and
Gershenson
,
A.
(
2009
)
Post-reductionist protein science, or putting Humpty Dumpty back together again
.
Nat. Chem. Biol.
5
,
774
777
12
Zimmerman
,
S.B.
and
Trach
,
S.O.
(
1991
)
Estimation of macromolecule concentrations and excluded volume effects for the cytoplasm of Escherichia coli
.
J. Mol. Biol.
222
,
599
620
13
van den Berg
,
B.
,
Ellis
,
R.J.
and
Dobson
,
C.M.
(
1999
)
Effects of macromolecular crowding on protein folding and aggregation
.
EMBO J.
18
,
6927
6933
14
Rivas
,
G.
,
Ferrone
,
F.
and
Herzfeld
,
J.
(
2004
)
Life in a crowded world
.
EMBO Rep.
5
,
23
27
15
Brown
,
G.C.
(
1991
)
Total cell protein concentration as an evolutionary constraint on the metabolic control distribution in cells
.
J. Theor. Biol.
153
,
195
203
16
McConkey
,
E.H.
(
1982
)
Molecular evolution, intracellular organization, and the quinary structure of proteins.
Proc. Natl. Acad. Sci. U.S.A.
79
,
3236
3240
17
Freedberg
,
D.I.
and
Selenko
,
P.
(
2014
)
Live cell NMR
.
Annu. Rev. Biophys.
43
,
171
192
18
Crivat
,
G.
and
Taraska
,
J.W.
(
2012
)
Imaging proteins inside cells with fluorescent tags
.
Trends Biotechnol.
30
,
8
16
19
Sun
,
Y.
,
Hays
,
N.M.
,
Periasamy
,
A.
,
Davidson
,
M.W.
and
Day
,
R.N.
(
2012
)
Monitoring protein interactions in living cells with fluorescence lifetime imaging microscopy
.
Methods Enzymol.
504
,
371
391
20
Kaur
,
U.
,
Meng
,
H.
,
Lui
,
F.
,
Ma
,
R.
,
Ogburn
,
R.N.
,
Johnson
,
J.H.R.
et al (
2018
)
Proteome-wide structural biology: an emerging field for the structural analysis of proteins on the proteomic scale
.
J. Proteome Res.
17
,
3614
3627
21
Fontana
,
A.
,
de Laureto
,
P.P.
,
Spolaore
,
B.
,
Frare
,
E.
,
Picotti
,
P.
and
Zambonin
,
M.
(
2004
)
Probing protein structure by limited proteolysis
.
Acta Biochim. Pol.
51
,
299
321
.
22
Park
,
C.
and
Marqusee
,
S.
(
2005
)
Pulse proteolysis: a simple method for quantitative determination of protein stability and ligand binding
.
Nat. Methods
2
,
207
212
23
West
,
G.M.
,
Tang
,
L.
and
Fitzgerald
,
M.C.
(
2008
)
Thermodynamic analysis of protein stability and ligand binding using a chemical modification- and mass spectrometry-based strategy
.
Anal. Chem.
80
,
4175
4185
24
Feng
,
Y.
,
De Franceschi
,
G.
,
Kahraman
,
A.
,
Soste
,
M.
,
Melnik
,
A.
,
Boersema
,
P.J.
et al (
2014
)
Global analysis of protein structural changes in complex proteomes
.
Nat. Biotechnol.
32
,
1036
1044
25
Dearmond
,
P.D.
,
Xu
,
Y.
,
Strickland
,
E.C.
,
Daniels
,
K.G.
and
Fitzgerald
,
M.C.
(
2011
)
Thermodynamic analysis of protein-ligand interactions in complex biological mixtures using a shotgun proteomics approach
.
J. Proteome Res.
10
,
4948
4958
26
Ogburn
,
R.N.
,
Jin
,
L.
,
Meng
,
H.
and
Fitzgerald
,
M.C.
(
2017
)
Discovery of tamoxifen and N-desmethyl tamoxifen protein targets in MCF-7 cells using large-scale protein folding and stability measurements
.
J. Proteome Res.
16
,
4073
4085
27
Reinhard
,
F.B.
,
Eberhard
,
D.
,
Werner
,
T.
,
Franken
,
H.
,
Childs
,
D.
,
Doce
,
C.
et al (
2015
)
Thermal proteome profiling monitors ligand interactions with cellular membrane proteins
.
Nat. Methods
12
,
1129
1131
28
Savitski
,
M.M.
,
Reinhard
,
F.B.
,
Franken
,
H.
,
Werner
,
T.
,
Savitski
,
M.F.
,
Eberhard
,
D.
et al (
2014
)
Tracking cancer drugs in living cells by thermal profiling of the proteome
.
Science
346
,
1255784
29
Sridharan
,
S.
,
Kurzawa
,
N.
,
Werner
,
T.
,
Gunthner
,
I.
,
Helm
,
D.
,
Huber
,
W.
et al (
2019
)
Proteome-wide solubility and thermal stability profiling reveals distinct regulatory roles for ATP
.
Nat. Commun.
10
,
1155
30
Franken
,
H.
,
Mathieson
,
T.
,
Childs
,
D.
,
Sweetman
,
G.M.
,
Werner
,
T.
,
Togel
,
I.
et al (
2015
)
Thermal proteome profiling for unbiased identification of direct and indirect drug targets using multiplexed quantitative mass spectrometry
.
Nat. Protoc.
10
,
1567
1593
31
Sinz
,
A.
(
2006
)
Chemical cross-linking and mass spectrometry to map three-dimensional protein structures and protein-protein interactions
.
Mass Spectrom. Rev.
25
,
663
682
32
Hage
,
C.
,
Falvo
,
F.
,
Schafer
,
M.
and
Sinz
,
A.
(
2017
)
Novel concepts of MS-cleavable cross-linkers for improved peptide structure analysis
.
J. Am. Soc. Mass Spectrom.
28
,
2022
2038
33
Pimenova
,
T.
,
Nazabal
,
A.
,
Roschitzki
,
B.
,
Seebacher
,
J.
,
Rinner
,
O.
and
Zenobi
,
R.
(
2008
)
Epitope mapping on bovine prion protein using chemical cross-linking and mass spectrometry
.
J. Mass Spectrom.
43
,
185
195
34
Schweppe
,
D.K.
,
Chavez
,
J.D.
,
Lee
,
C.F.
,
Caudal
,
A.
,
Kruse
,
S.E.
,
Stuppard
,
R.
et al (
2017
)
Mitochondrial protein interactome elucidated by chemical cross-linking mass spectrometry.
.
Proc. Natl. Acad. Sci. U.S.A.
114
,
1732
1737
35
Gotze
,
M.
,
Iacobucci
,
C.
,
Ihling
,
C.H.
and
Sinz
,
A.
(
2019
)
A simple cross-linking/mass spectrometry workflow for studying system-wide protein interactions
.
Anal. Chem.
91
,
10236
10244
36
Schweppe
,
D.K.
,
Harding
,
C.
,
Chavez
,
J.D.
,
Wu
,
X.
,
Ramage
,
E.
,
Singh
,
P.K.
et al (
2015
)
Host-microbe protein interactions during bacterial infection
.
Chem. Biol.
22
,
1521
1530
37
Weisbrod
,
C.R.
,
Chavez
,
J.D.
,
Eng
,
J.K.
,
Yang
,
L.
,
Zheng
,
C.
and
Bruce
,
J.E.
(
2013
)
In vivo protein interaction network identified with a novel real-time cross-linked peptide identification strategy
.
J. Proteome Res.
12
,
1569
1579
38
Chavez
,
J.D.
,
Lee
,
C.F.
,
Caudal
,
A.
,
Keller
,
A.
,
Tian
,
R.
and
Bruce
,
J.E.
(
2018
)
Chemical crosslinking mass spectrometry analysis of protein conformations and supercomplexes in heart tissue
.
Cell Syst.
6
,
136
141.e5
39
Espino
,
J.A.
and
Jones
,
L.M.
(
2019
)
Illuminating biological interactions with in vivo protein footprinting
.
Anal. Chem.
91
,
6577
6584
40
Hambly
,
D.M.
and
Gross
,
M.L.
(
2005
)
Laser flash photolysis of hydrogen peroxide to oxidize protein solvent-accessible residues on the microsecond timescale
.
J. Am. Soc. Mass Spectrom.
16
,
2057
2063
41
Espino
,
J.A.
,
Mali
,
V.S.
and
Jones
,
L.M.
(
2015
)
In cell footprinting coupled with mass spectrometry for the structural analysis of proteins in live cells
.
Anal. Chem.
87
,
7971
7978
42
Rinas
,
A.
,
Mali
,
V.S.
,
Espino
,
J.A.
and
Jones
,
L.M.
(
2016
)
Development of a microflow system for in-cell footprinting coupled with mass spectrometry
.
Anal. Chem.
88
,
10052
10058
43
Huang
,
W.
,
Ravikumar
,
K.M.
,
Chance
,
M.R.
and
Yang
,
S.
(
2015
)
Quantitative mapping of protein structure by hydroxyl radical footprinting-mediated structural mass spectrometry: a protection factor analysis
.
Biophys. J.
108
,
107
115
44
Johnson
,
D.T.
,
Punshon-Smith
,
B.
,
Espino
,
J.A.
,
Gershenson
,
A.
and
Jones
,
L.M.
(
2020
)
Implementing in-cell fast photochemical oxidation of proteins in a platform incubator with a movable XY stage
.
Anal. Chem.
92
,
1691
1696