Method for detecting affinity of a ligand for a protein or protein having a change in energy state - Patent Application 20070122997

Abstract

The present invention discloses a method for detecting proteins with altered energy states and the affinity of ligands for proteins. Specifically, after the energy state of a protein changes, its resistance to proteolytic destruction changes. The structure of low-energy proteins can also be disrupted under non-denaturing conditions by using a large amount of enzymes, directly generating small peptide fragments with molecular weights of less than 5 KDa that can be directly used for bottom-to-top mass spectrometry analysis. This method has extremely high sensitivity. Quantitative proteomics is used to find differential peptide fragments resulting from enzymatic cleavage, and the proteins to which the differential peptide fragments belong and their locations in the proteins are analyzed. As a result, proteins with altered energy states and altered regions can be determined across the entire proteome. If the energy state of a protein changes due to the addition of a ligand, this method can determine the binding protein and the binding region of the ligand; the quantitative output of peptide fragment-level results further enables the method to determine the local affinity of the ligand binding to the protein.

Description

[Technical Field]

[0001] The present invention belongs to the research field of functional proteomics (proteomics category). Specifically, the present invention involves detecting proteins with altered energy states, identifying ligand-binding proteins, or determining local affinities between ligands and proteins in complex protein mixtures such as cell or tissue lysates. [Background technology]

[0002] Proteins are key players in biological processes. Traditional quantitative proteomics is commonly used to assess differences in protein abundance between different samples. However, many diseases, such as Parkinson's disease and Alzheimer's disease, or interactions between proteins and ligands (including drugs, endogenous metabolites, etc.), result in changes in protein folding state, post-translational modifications, or altered protein function rather than specific protein abundance [De Souza N, et al., Curr Opin Struct Biol, 2020, 60:57-65; Cappelletti V, et al., Cell, 2021, 184(2):545-559]. These changes in protein properties beyond abundance can affect the thermodynamic stability, i.e., the energy state, of a protein. Therefore, the development of methods capable of detecting changes in the energy state of proteins is of significant importance in fields such as disease diagnosis, metabolic pathway analysis, and drug target discovery.

[0003] Top-down mass spectrometry allows for the direct analysis of intact proteins to monitor changes in their energy state (Sheng Y et al., International Journal of Mass Spectrometry, 2010, 300 (2011) 118-122). However, this method has low throughput and is primarily applicable to purified proteins, making it difficult to apply to complex protein mixtures. Non-mass spectrometry methods are also available for detecting changes in protein energy state, such as fluorescence resonance energy transfer (Heyduk T, et al., Curr. Opin. Biotechnol, 2002, 13, 292-296) and nuclear magnetic resonance (Sakakibara D, et al., Nature, 2009, 458, 102-105). These techniques are suitable for purified proteins and therefore cannot be applied at the proteomics level.

[0004] Bottom-up mass spectrometry is a technique that involves enzymatically digesting large protein fragments into smaller peptides for subsequent mass spectrometric analysis. This technique allows for the identification or quantification of thousands or even tens of thousands of proteins simultaneously. In recent years, a series of high-throughput methods that combine bottom-up mass spectrometry techniques to analyze changes in protein energy state have attracted widespread attention both domestically and internationally. These methods are based on the altered resistance to heat or denaturation following a change in a protein's energy state. Therefore, proteins with lower energy states are more likely to maintain their original conformation and are less likely to precipitate or unfold when subjected to the same intensity of heat or denaturing stimulus. Proteins with altered energy states can be identified by analyzing the difference in the amount of precipitated or unfolded protein between two or more groups of samples. Such techniques include cellular thermal shift assays (CETSA, Martinez Molina et al., Science, 2013, 341(6141):84-87), thermal proteome profiling (TPP, Savitski M et al., Science, 2014, 346(6205): 1255784), and pulse proteolysis (PP, Chiwook P et al., Nature Methods, 2005, 2(3):207-212). These methods for detecting changes in protein energy states have been successfully applied to identify drug target proteins [Savitski M et al., Science, 2014, 346(6205): 1255784; Perrin J et al., Nature Biotechnology, 2020, 38(3):303-308], analyze protein-protein interactions [Tan CS et al., Science, 2018, 359(6380):1170-1177], and study functional post-translational modifications (Huang JX et al., Nature Methods, 2019, 16, 894-901).However, these methods cannot be used to analyze the changes in the energy state of proteins at the protein level and determine the specific region where the energy state of proteins changes.In addition, for proteins with multiple domains, local stability changes may have minimal impact on the overall stability of proteins, making it difficult to capture stability changes at the protein level (Mateus A, et al., Proteome Sci, 2016, 15:13).

[0005] Limited proteolysis (LiP) has long been used to study protein conformation. This method is based on the differential susceptibility of proteins to proteolysis of different conformations. By using a small amount of nonspecific enzyme, the protein is slightly cleaved, primarily targeting high-energy disordered regions to generate large protein fragments. When a protein undergoes a conformational change (e.g., disordered regions become more ordered), the susceptibility of the conformationally altered region to proteolysis changes accordingly. Conformationally altered regions of a protein are identified by detecting differences in the abundance of the resulting large protein fragments by immunoblotting. Because changes in the energy state of proteins are often accompanied by conformational changes, this method can also be used to study changes in the energy state of proteins and to identify regions of proteins where the energy state changes (Antonio Aceto, et al., The International Journal of Biochemistry & Cell Biology, 1995, 27(10):1033-1041; Polverino de Laureto, et al., J Mol Biol, 2003, 334(1):129-141). However, due to the difficulty of directly analyzing large protein fragments in bottom-up mass spectrometry for high-throughput quantification, this technique is limited to the characterization of purified proteins.

[0006] Paola et al. combined limited proteolysis (LiP) with bottom-up mass spectrometry (Feng YH, et al., Nature Biotechnology, 2014, 32(10):1036-1044). The large fragments generated from limited proteolysis (i.e., the aforementioned large protein fragments) were denatured using a denaturant and then subjected to a second complete trypsin digestion to generate semi-tryptic peptides containing non-specific cleavage sites and complete tryptic peptides without non-specific cleavage sites. Using bottom-up quantitative proteomics, proteins with conformational changes were identified by quantifying the difference in abundance of semi-tryptic or complete tryptic peptides in two or more groups of samples. In addition, conformationally altered regions were determined by analyzing the location of these peptides on the protein. To some extent, this method achieved large-scale analysis of changes in protein energy states and captured interactions between metabolites and proteins in relatively simple biological systems [Piazza I, et al., Cell, 2018, 172(1-2):358-372], as well as structural changes in proteins due to post-translational modifications, thermal stimuli, and osmotic pressure [Cappelletti V, et al., Cell, 2021, 184(2):545-559]. However, the two-step digestion in this method greatly increased sample complexity and tended to result in peptide-peptide interference. Therefore, this method is primarily applicable to relatively simple biological systems. Paola et al. (2020) further developed LiP-Quant [Piazza I, et al., Nature Communications, 2020, 11(1):4200], a method for improving the reliability of target protein identification based on LiP-MS using multiple ligand concentrations. LiP-Quant has achieved some success in identifying ligand-binding proteins and their binding regions in mammalian cell lysates. However, due to the two-step digestion, it is difficult to identify semitryptic peptides containing nonspecific cleavage sites.In addition, peptides with non-specific cleavage sites show relatively weak responses to ligand binding, resulting in very limited sensitivity in identifying ligand-bound target proteins. For example, when identifying target proteins of staurosporine, a broad-spectrum kinase inhibitor, requires that all identified candidate target proteins be kinases (>80%), only 9 kinase target proteins were identified in HeLa cell lysates by LiP-Quant, compared with 111 kinase target proteins in HeLa cell lysates according to the present invention (see Example 3) [Piazza I, et al., Nature Communications, 2020, 11(1):4200].

[0007] Previously described proteolysis-based methods for studying protein conformation involve the use of relatively small amounts of enzyme, which restricts cleavage sites to high-energy, disordered regions. Different conformations of proteins result in distinct, large protein fragments. Proteins with conformational changes are determined by detecting differences in these resulting protein fragments via immunoblotting or denaturant-assisted secondary digestion. There is no reason to use larger amounts of enzyme to induce destructive digestion of proteins (i.e., first digesting the high-energy state structure of the protein to expose the low-energy state cleavage site, and then digesting the low-energy state structure in the presence of a substantial amount of enzyme) to directly generate small peptides with molecular weights less than 5 kDa with two cleavage sites that reflect the stability of the protein region. In bottom-up mass spectrometry approaches, quantitative proteomics is used to analyze the abundance differences between these peptides to determine proteins and protein regions with altered energy states, or the local binding affinity of ligands to the protein. To address these issues, we have devised a method that can sensitively determine proteins and protein regions with altered energy states and measure the local affinity of ligands for proteins in complex protein mixtures.

[0008] This method utilizes destructive digestion to directly generate small peptides containing two cleavage sites suitable for bottom-up mass spectrometry analysis, both of which reflect the local stability of the protein. This significantly reduces sample complexity compared to the products obtained by the two-step digestion with two enzymes used in LiP-MS. Furthermore, during the generation of these peptides, the protein undergoes relatively thorough digestion (i.e., high-energy state protein structures are first destroyed, inducing further digestion of low-energy state protein structures). Within this cascade process, the cumulative difference in digestion rates at each stage results in small peptides that exhibit a significant response to changes in the protein's energy state or ligand binding, thereby substantially increasing the range of detectable peptide fragment changes. Therefore, this method exhibits high sensitivity in detecting changes in the protein's energy state. For example, compared to LiP-Quant, if we require a kinase ratio of over 80% among all identified target proteins for staurosporine, 111 kinases in HeLa cell lysates can be identified by this method (see Example 3), which is a 12.3-fold improvement over LiP-Quant (9 kinases can be identified). This method outputs quantitative results at the peptide level and can be used to calculate the local binding affinity of proteins for ligands at multiple doses. Summary of the Invention [Problem to be solved by the invention]

[0009] The objective of this invention is to develop a method capable of sensitively determining the local binding affinity of ligands to proteins or protein regions with altered energy states in complex systems, which we call the Peptide-Centered Local Stability Assay (PELSA). [Means for solving the problem]

[0010] The principle of the present invention is that when the energy state of a protein changes, its local or global stability also changes, changing its ability to resist external destruction. Enzymatic digestion is a destructive process in which changes in the energy state of a protein affect its resistance to destruction caused by enzymatic digestion, and the present invention takes advantage of this fact. Proteins are first cleaved into large subunits with a high concentration of protease. Due to the destructive nature of enzymatic digestion, subunit structures that are in a high-energy state or not bound to a ligand are destroyed and rapidly unfolded, thereby inducing further enzymatic digestion of the structure itself. This leads to the destruction of the low-energy state structure, ultimately producing small peptides (<5 kDa) that contain two cleavage sites suitable for bottom-up mass spectrometry analysis and can both reflect the stability of the protein region. On the other hand, protein regions that are in a low-energy state or bound to a ligand maintain a stable structure and are unlikely to enter the active pocket of the protease, making the generation of small peptides difficult. Proteins with altered energy states can be determined by comparing the differences in the abundance of such peptides. Protein regions with altered energy states can be determined by analyzing the location of different peptides on the protein. In this method, proteins directly generate small peptides containing two cleavage sites, both of which can reflect the local stability of the protein under non-denaturing conditions. Due to the relatively thorough cascade digestion process that generates small peptides, differences in the enzymatic digestion rates at every stage in this cascade process accumulate, resulting in small peptides that exhibit a large response to changes in the protein's energy state or binding to a ligand. Furthermore, because this method involves extensive destructive enzymatic digestion, the cleavage sites are not limited to high-energy disordered regions, resulting in a larger number of peptides that reflect changes in protein stability. In addition, using only one protease in this process significantly reduces the complexity of the peptides generated, allowing a large number of proteins to be examined. These features make this method highly sensitive.

[0011] When a ligand is exposed to a complex protein sample, binding of the ligand to its target protein results in a change in the energy state of the ligand-binding region in the target protein. Thus, proteins that are bound to a ligand and their ligand-binding regions can be determined by analyzing the change in the energy state of the intact protein using this method.

[0012] The principle of this technique for determining the local affinity of a protein for a ligand is based on the dose-dependent change in the local stability of the target protein with respect to the dose of added ligand, unless a saturating concentration is reached. By detecting the abundance of small cleaved peptides in samples treated with different ligand concentrations, the local affinity of the ligand for a protein region can be calculated.

[0013] The small peptides detected by this technique have two cleavage sites, both of which reflect the local stability of the protein. Different stability levels of various regions of the protein in the conformation being tested result in varying degrees of susceptibility to enzymatic digestion. The cleavage site generated under such conditions is referred to as the cleavage site reflecting the local stability of the protein.

[0014] The present invention includes the following technical scheme: (a) preparing solutions of proteins in different energy states or solutions of proteins incubated with various concentrations of ligands, and then subjecting the solutions to destructive enzymatic digestion at a specific enzyme-to-protein ratio, allowing most proteins to produce small peptides with molecular weights less than 5 kDa. These peptides contain two cleavage sites, both of which can reflect the local stability of the protein; (b) isolating the cleaved peptides and quantitatively analyzing them using bottom-up mass spectrometry; and (c) analyzing the differences in the abundance of the cleaved peptides to determine proteins and protein regions with changes in energy state and the local binding affinity of the ligand to the protein.

[0015] Specifically, the technical scheme includes the following steps:

[0016] 1) Preserving the conformation of the protein to be tested in the solution described in step (a). If the tissue or cell lysate is obtained by lysis, one or more gentle lysis methods can be used, such as repeated freezing and thawing using liquid nitrogen, liquid nitrogen crushing, or homogenization. The lysis solution used also preserves the conformation of the protein to be tested.

[0017] 2) During the study of ligand-binding proteins, the biological samples containing the protein are divided into two groups. The experimental group is treated with a specific concentration of ligand, while the control group is treated with a blank solvent. However, this is not limited to only two groups, and the experimental group may also include multiple groups with various concentrations of ligand.

[0018] 3) Adding enzyme to the solution in a weight ratio of enzyme to total protein ranging from 1:1 to 1:50, and digesting the protein at a certain temperature for 0.5 to 60 minutes using a vibrating shaker at a speed of 1000 rpm to 1500 rpm.

[0019] 4) Heat the digested sample from step 3) in a metal bath at 100°C for 5 minutes to terminate the digestion.

[0020] 5) Alkylation of the digested sample by dissolving the heat-induced precipitate in guanidine hydrochloride to a final concentration of 6 M, followed by the addition of tris(2-carboxyethyl)phosphine (TCEP) to a final concentration of 10 mM and chloroacetamide (CAA) to a final concentration of 40 mM at 95°C for 5 min.

[0021] 6) Transfer the alkylated sample from step 5) to a 10 kDa ultrafiltration unit and centrifuge at 14,000 g for 20 min to 1 h to isolate small peptides. The technical scheme further includes subsequently washing the ultrafiltration membrane with 200 μL of pH 8.2 HEPES buffer, followed by further centrifugation at 14,000 g to isolate the remaining peptides, and mixing the remaining peptides with the peptides collected in the first centrifugation.

[0022] 7) Quantifying the collected peptides from step 6) using mass spectrometry by utilizing either label-based or label-free quantification methods.

[0023] 8) Analyzing the quantification results at the peptide level, where the significance or fold change of the first / second / third / fourth ranked peptides (ranked by significance or fold change) is used as an indicator of the significance or fold change of the protein. This approach is used to screen reliable target proteins.

[0024] 9) Assigning peptides to their respective protein sequences or structures to identify protein regions with changes in energy state, fitting the fold changes of peptides at different ligand concentrations and the corresponding ligand concentrations to a four-parameter logarithmic equation [Y=bottom+(top-bottom) / [1+10^(LogEC50-X)×Hill slope]], and calculating the local affinity between the ligand and the protein, represented by the EC50 value in this equation. [Effects of the Invention]

[0025] Compared to existing methods, the present invention, known as PELSA, exhibits several innovations and advantages.

[0026] (1) PELSA utilizes the destructive process of enzymatic digestion. By using a large amount of enzyme, proteins under non-denaturing conditions are first destroyed in their high-energy state protein structure by the destructive effect of digestion, which results in the exposure of cleavage sites in their low-energy state protein structure. In the presence of a large amount of enzyme, the low-energy state protein structure is also digested and destroyed, allowing most proteins to directly generate small peptides with molecular weights less than 5 kDa. These peptides are then identified and quantified using mass spectrometry. Peptides identified in PELSA are generated from a cascade-like destructive digestion process, and these peptides can be directly analyzed by mass spectrometry in a bottom-up manner. When a protein undergoes a change in energy state or binds to a ligand, the cumulative difference in enzymatic digestion rate at each stage of the cascade digestion process results in a significant response from the peptide. In addition, this method utilizes a large amount of enzyme, allowing the generation of a greater number of ligand-responsive peptides. For example, in Example 2, PELSA identified 2 to 5 times more ligand-responsive peptides than the LiP-MS method. Furthermore, the magnitude of the fold changes of these peptides measured in PELSA was 4-6 times higher than that observed in LiP-MS.

[0027] (2) Wide applicability: Compared with traditional affinity chromatography-based methods, this method does not require chemical modification of the ligand molecule when used to identify ligand-binding proteins. This means that, in theory, this method can be applied to any ligand. In addition, because the identification of ligand-binding target proteins does not depend on the strength of their binding affinity, it is suitable for identifying target proteins with low ligand affinity. For example, successful identification of target proteins of metabolites such as folate, leucine, and α-ketoglutarate (αKG) is demonstrated in Examples 4 to 6. Notably, Example 6 represents the largest number of known αKG target proteins identified in a single experiment to date.

[0028] (3) Identification of protein regions with different energy states or ligand-binding regions: Compared with the widely applied thermal proteome profiling (TPP), this method allows for the identification of regions in proteins with altered energy states or regions involved in ligand binding. This determination is achieved by analyzing the location of peptides with alterations within the protein. As demonstrated in Example 3, this method successfully identified the binding region of the kinase inhibitor staurosporine in 154 kinases.

[0029] (4) High sensitivity: Unlike digestion differences captured at the protein level, where local differences caused by protein energy state or ligand binding are averaged across the entire protein, this method directly captures local differences induced by protein energy state or ligand binding at the peptide level, making it more sensitive in detecting subtle changes. Additionally, determining ligand-binding regions using this method does not rely on identifying peptides with specific characteristics, such as those containing nonspecific cleavage sites, as required by LiP-MS. This makes it easier to identify differential peptides. For example, in Example 3, PELSA identified a 2.1-fold increase in the number of kinase target proteins for the broad-spectrum kinase inhibitor staurosporine compared to TPP, and a 12.3-fold increase compared to LiP-Quant.

[0030] (5) Assessment of local ligand-protein binding affinity: A protein sample is divided into multiple groups, and different concentrations of ligand are added to each group. Unless saturation is reached, varying ligand concentration alters the resistance of the target protein to digestion. Affinity information can be obtained by fitting a four-parameter logarithmic equation to the data, with ligand concentration on the x-axis and the fold change in abundance of cleaved peptides on the y-axis. A single protein can be identified using multiple peptides, and the affinity curve for each peptide represents the local binding affinity of the ligand to its specific region. For example, Example 11 demonstrates that the local affinity between ligand and protein measured by PELSA was very similar to that determined using a microscale thermophoresis assay.

[0031] (6) Wide applicability: Protein-ligand binding, post-translational modification of proteins, conformational changes of proteins, or protein-protein interactions can all induce changes in the energy state of proteins. Therefore, the present method can be applied to investigate where external perturbations cause changes in the energy state of proteins, providing a powerful tool for addressing various biological questions. Example 7 demonstrates the application of PELSA in analyzing changes in the energy state of proteins caused by protein-protein interactions. Example 8 shows the application of PELSA in analyzing changes in the energy state of proteins resulting from interactions between proteins and post-translationally modified peptides. Example 9 shows the application of PELSA in analyzing changes in the energy state of proteins induced by interactions between metal ions and proteins. [Brief explanation of the drawings]

[0032] [Figure 1]PELSA flowchart: Proteins with different energy states (e.g., a protein in a ligand-bound state and the same protein in an unliganded state) undergo destructive digestion at a specific enzyme-to-protein ratio. Due to the binding of the ligand to the protein in the ligand-treated group, the binding region becomes more stable, which results in a lower susceptibility to unfolding compared to the control group, and therefore the generation of fewer cleaved peptides. The digestion products are then denatured and alkylated, followed by ultrafiltration to isolate the cleaved peptides, which are then subjected to quantitative mass spectrometry analysis.

[0033] [Figure 2]Application of PELSA to analyze changes in protein energy state in cell lysates treated with the anti-breast cancer drug lapatinib. (A) Analysis of protein secondary structure at PELSA cleavage sites: PELSA cleavage sites refer to the N- and C-terminal residues of all identified peptides in the PELSA experiment. The source of protein secondary structure information was AlphaFold, and the classification of secondary structures was based on the literature [Bludau I, et al, PLoS Biol, 2022, 20(5): e3001636]. HELIX represents a low-energy helical structure, STRAND represents a folded structure, BEND represents a bent structure, TURN represents a turn structure, and unstructured represents a high-energy disordered structure. (B) Volcano plot of protein levels corresponding to BT474 cell lysates treated with 100 nM lapatinib: Peptide quantitative values ​​were obtained from BT474 cell lysates treated with 100 nM lapatinib or DMSO (four replicates). Fold changes and P values ​​for each peptide were calculated using an empirical Bayes t-test. The peptide with the smallest P value per protein was used to represent the corresponding protein. The log2 fold change and -log10 P value of the protein are plotted on the x-axis and y-axis, respectively. (C) Two-dimensional local stability profile of ERBB2: Log2 fold change (Log2FC) of quantified peptides of ERBB2 and their corresponding positions on the two-dimensional array of ERBB2. (D) Local affinity profile of ERBB2: Log2 fold change of ERBB2 peptides at different concentrations of lapatinib. (E) Volcano plot of protein levels corresponding to BT474 cell lysates treated with 1 μM lapatinib. (F) Two-dimensional local stability profile of off-target kinases identified for lapatinib using CHEK2, SLK, RIPK2, or YES1 as examples. (G) Western blot confirms stabilization of PTGES2 upon lapatinib treatment.

[0034] Unless otherwise stated, the terms "fold change," "FC" (short for fold change), or "ratio" herein refer to the ratio of peptide abundance between experimental and control groups.

[0035] [Figure 3] Comparison of PELSA and LiP-MS for identifying proteins with altered energy states in HeLa cell lysates treated with methotrexate (MTX) or SHP099. (A) (Top) Volcano plot of peptide levels obtained from LiP-MS / PELSA for HeLa cell lysates treated with 10 μM MTX. (Bottom) Volcano plot of peptide levels obtained from LiP-MS / PELSA for HeLa cell lysates treated with 10 μM SHP099. (B) In the MTX- or SHP099-treated systems, the number of ligand-responsive peptides of target proteins identified by PELSA was 2-fold and 5.25-fold higher, respectively, than that by LiP-MS. (C) The fold changes of ligand-responsive peptides of target proteins DHFR and PTPN11 identified by PELSA were 4.3-fold and 6.4-fold higher, respectively, than that by LiP-MS. (D) PDB structure of DHFR bound to MTX. The arrow indicates the peptide located in the MTX-binding site within the low-energy helical region. This peptide showed a significant fold change in PELSA but remained unchanged in LiP-MS. (E) PDB structure of PTPN11 bound to SHP099. The arrow indicates the peptide located in the SHP099-binding site and embedded within the protein structure. Similarly, this peptide showed a significant fold change in PELSA but no fold change in LiP-MS.

[0036] [Figure 4]Figure 4 demonstrates the high sensitivity of PELSA in identifying proteins with altered energy states, e.g., in screening for target proteins of staurosporine, a broad-spectrum kinase inhibitor. (A) Volcano plots of protein levels corresponding to K562 cell lysates (left) and HeLa cell lysates (right) treated with 20 μM staurosporine. (B) Comparison of the number of staurosporine target proteins identified by PELSA in K562 cell lysates, PELSA in HeLa cell lysates, LiP-Quant in HeLa cell lysates (previously reported), and TPP in K562 cell lysates (previously reported). The x-axis represents the total number of identified target proteins, and the y-axis represents the number of kinases among the identified target proteins. (C) The number of proteins and peptides identified by LiP-Quant (HeLa), TPP (K562), PELSA (K562), and PELSA (HeLa), as well as the final number of kinase targets. (D) Comparison of protein sequence coverage for all proteins identified by PELSA(HeLa) and LiP-Quant(HeLa) (left), and comparison of protein sequence coverage for kinase targets identified by PELSA(HeLa) and LiP-Quant(HeLa) (right). This graph shows that PELSA can identify target proteins with lower protein sequence coverage. (E) Split violin plot illustrating the distribution of thermal melting points for kinase targets identified by PELSA(HeLa), PELSA(K562), or TPP(K562), and for all quantified kinase proteins in each dataset. This graph demonstrates that PELSA is effective in identifying both heat-stable and heat-sensitive target proteins. (F) Overlap of kinase targets identified by PELSA(K562) and PELSA(HeLa). (G) Density plot showing the significance of stability changes of peptides located within the kinase domain relative to other peptides identified by PELSA in K562 (left) and HeLa (right).The x-axis represents the significance of the change, i.e., -log10P value, calculated using an empirical Bayes t-test, and the y-axis represents density. The pie plots show the percentage of significantly changed peptides located within the kinase domain, very close to the kinase domain (within ±10 amino acid residues away from the kinase domain), belonging to proteins without the kinase domain, and located outside the kinase domain. It can be seen that the majority of peptides with significant fold changes are located within the kinase domain.

[0037] [Figure 5]Figure 5 shows the application of PELSA in the analysis of proteins with local energy state changes in HeLa cell lysates upon metabolite treatment. (A) Volcano plot of protein levels corresponding to K562 cell lysates treated with 50 μM folic acid. (B) Three-dimensional structure plot demonstrating the stability change of DHFR upon folic acid binding as measured by PELSA. The arrow indicates the protein region with the greatest stability change. (C) Three-dimensional structure plot demonstrating the stability change of ATIC upon the addition of 50 μM folic acid as measured by PELSA. The arrow indicates the protein fragment with the greatest stability change. (D) Two-dimensional plot demonstrating the stability change of MTHFR upon the addition of 50 μM folic acid as measured by PELSA. (E) Three-dimensional structure plot demonstrating the stability change of GART upon the addition of 50 μM folic acid as measured by PELSA. The arrow indicates the protein fragment with the greatest stability change. (F) Two-dimensional plot demonstrating the change in stability of P3H1 upon the addition of 50 μM folic acid as measured by PELSA. (G) Volcano plot of protein levels corresponding to K562 cell lysates treated with 5 mM leucine. (H) Two-dimensional plot demonstrating the change in stability of LARS1 upon the addition of 5 mM leucine as measured by PELSA. (I) Topological structure diagram of the membrane protein SLC1A5 with the start and end positions of three quantified peptides (190–212, 493–502, and 523–541) shown with their corresponding |log2FC| values ​​labeled along each individual peptide. (J) Two-dimensional plot demonstrating the change in stability of PPIP5K1 and PPIP5K2 upon the addition of 5 mM leucine as measured by PELSA. (K) Volcano plot of protein levels corresponding to HeLa cell lysates treated with 2 mM α-ketoglutarate (αKG). Dashed lines represent -log10P value = 3.4 and log2FC = -0.5, respectively. Of the 40 proteins, 30 that meet the criteria of -log10P value > 3.4 and log2FC < -0.5 are known target proteins of αKG. (L) Two-dimensional plot demonstrating the change in stability of EGLN1, RSBN1L, or KDM3B as determined by PELSA in HeLa cell lysates treated with 2 mM αKG.The protein fragments detected by PELSA as having altered energy states are known αKG binding regions.

[0038] [Figure 6] Figure 6 shows the application of PELSA in analyzing changes in the local energy state of proteins due to protein-protein interactions (as indicated by antibody-antigen binding) in HeLa cell lysates. (A) Schematic representation of PELSA identifying antibody-binding epitopes in cell lysates. (B) (Left) Protein-level volcano plot corresponding to DHFR antibody-treated HeLa cell lysates. (Right) Protein-level volcano plot corresponding to CDK9 antibody-treated HeLa cell lysates. (C) Three-dimensional structure plot demonstrating the change in stability of DHFR upon the addition of DHFR antibody as measured by PELSA (PDB:1BOZ). Shaded circles indicate antibody-binding epitopes, and arrows indicate the protein fragments with the greatest stability change. (D) (Top) Two-dimensional plot demonstrating the change in stability of DHFR upon the addition of DHFR antibody as measured by PELSA. (Bottom) Two-dimensional plot demonstrating the change in stability of CDK9 upon the addition of CDK9 antibody as measured by PELSA. (E) Changes in abundance values ​​of two peptides of CDK9, NPATTNQTEFERVF and NPATTNQTEFER, upon addition of CDK9 antibody.

[0039] [Figure 7]Figure 7 shows the application of PELSA to analyze changes in the local energy state of proteins in BT474 cell lysates resulting from post-translational modifications and protein interactions (as indicated by identifying recognition domains for phosphotyrosines). (A) Schematic diagram of PELSA for identifying recognition domains for post-translational modifications in cell lysates. (B) Scatter plot of protein levels in a PELSA experiment. Here, pSEEI and YEEI were used as controls, and the -log10P values ​​(pYEEI / pSEEI) and -log10P values ​​(pYEEI / YEEI) were calculated for each peptide. The peptide with the largest sum of the -log10P values ​​(pYEEI / pSEEI) and -log10P values ​​(pYEEI / YEEI) was used to represent the corresponding protein. (C) Scatter plot of protein levels in a pull-down experiment. Here, pSEEI and YEEI are used as controls, and the -log10P values ​​(pYEEI / pSEEI) and -log10P values ​​(pYEEI / YEEI) are calculated for each protein. (D) Violin plot showing the log2FC distribution of peptides inside and outside the SH2 domains of nine SH2 domain-containing target proteins identified by PELSA. (E) Two-dimensional plot demonstrating the stability changes of SH2 domain-containing proteins identified by PELSA, showing examples such as YESI, TNS3, PLCG1, and GRB10. (F) Two-dimensional plot demonstrating the stability changes of calcium-binding proteins identified by PELSA, showing examples such as EFHD1 and CALM1.

[0040] [Figure 8]Figure 8 shows the application of PELSA in the analysis of proteins with altered energy states in HeLa cell lysates treated with metal ions (exemplified by zinc ions). (A) Volcano plot of protein levels corresponding to HeLa cell lysates treated with 30 μM ZnCl2. (B) Percentage of metal ion-binding proteins among all identified proteins (left) and among 280 significantly stabilized proteins (right). The pie chart on the right shows the ratio of zinc ion-binding proteins among stabilized metal ion-binding proteins. (C) Bar graph showing the number of various metal ion-binding proteins among significantly stabilized metal ion-binding proteins. (D) Distribution of log2 fold changes for peptides derived from 60 PELSA-identified proteins containing zinc finger motifs, grouped based on peptides located within and outside the zinc finger motif. (E) Two-dimensional plot demonstrating the stability changes of PELSA-identified proteins containing zinc finger motifs, showing examples such as SQSTM1, TRAD1, CHAMP1, and YY1. (F) Two-dimensional plot demonstrating the stability change of the zinc ion-binding protein LIMA1, which lacks a zinc finger motif, after treatment with 30 μM ZnCl2. Three peptides within the LIM domain are labeled 1, 2, and 3, respectively. (G) Three-dimensional structure of the LIM domain of LIMA1. Circles represent zinc ions. The three identified peptides in the LIM domain are labeled in the same order as in (F). (H) Distribution of log2 fold changes for peptides derived from 20 PELSA-identified proteins containing EF-hand / EH motifs, grouped based on peptides located within and outside the EF-hand / EH motifs. (I) Distribution of log2 fold changes for peptides derived from nine PELSA-identified target proteins containing Fe2+-binding domains, grouped based on peptides located within and outside the Fe2+-binding domains.(J) Zoomed-in protein-level volcano plot corresponding to HeLa cell lysate treated with 30 μM ZnCl2, retaining only proteins with increased energy states (-log10P value > 6, log2 fold change > 0). (K) Two-dimensional plot demonstrating the stability changes of IQ motif-containing proteins (UBE3C, MYO1B, MYO1C) and HNRNPU containing the B30.2 SPRY domain after the addition of 30 μM ZnCl2. (L) Distribution of log2 fold changes for peptides derived from PSMC1-6, grouped based on peptides located within and outside the P-loop-NTPase domain after the addition of 30 μM ZnCl2.

[0041] [Figure 9] Comparison of protein coverage and identification of proteins with altered energy states upon digestion with different enzymes by PELSA. (A) Comparison of the number of peptides identified by trypsin-, chymotrypsin-, and proteinase K-based PELSA (hereafter abbreviated as trypsin-PELSA, chymotrypsin-PELSA, and proteinase K-PELSA, respectively). (B) Venn diagram showing the overlap of identified proteins between trypsin-PELSA, chymotrypsin-PELSA, and proteinase K-PELSA. (C) Volcano plot of protein levels for trypsin-PELSA (left), chymotrypsin-PELSA (center), and proteinase K-PELSA (right) in HeLa cell lysates treated with 20 μM methotrexate (MTX).

[0042] [Figure 10]Dimethyl-labeling-based PELSA was used to identify proteins with altered energy states in HeLa cell lysates treated with three heat shock protein (Hsp) inhibitors. (A) Structures of three heat shock protein inhibitors: geldanamycin (left), tanespimycin (center), and ganetespib (right). (B) Experimental groups were treated with 100 μM geldanamycin, 100 μM tanespimycin, or 100 μM ganetespib, while the control group was treated with an equal volume of blank solvent. A scatter plot was generated, where the peptide with the second largest change represents the protein. The x- and y-axes of the plot represent the log2 fold change obtained from two replicates for mass spectrometry analysis. (C) Fold changes in abundance of peptides from different domains of HSP90AA1, HSP90AB1, HSP90B1, and HSP90AB2P after treatment with geldanamycin. (D) Verification of the interactions between MAT2A and ganetespib, AKR1C2 and ganetespib, and AKR1C2 and geldanamycin using thermal shift assays.

[0043] [Figure 11] Figure 11 shows the local affinity between ligands and proteins assessed using PELSA. (A) Fold changes in peptide abundance for heat shock proteins HSP90AA1, HSP90AB1, HSP90B1, and TRAP1 at different concentrations of ganetespib (ganetespib treatment vs. vehicle treatment). (B) Affinity measurements between proteins from the heat shock protein family and three inhibitors calculated by PELSA. The x-axis represents different inhibitor concentrations increasing from left to right, and the y-axis represents the fold change in peptide abundance upon drug treatment. A fold change of 1 indicates no abundance change. The annotated values ​​indicate the half-maximal inhibitory concentrations of the three inhibitors for HSP90AA1 obtained from PELSA calculations. (C) Affinity characterization of the three heat shock protein inhibitors with purified HSP90AA1 using microscale thermophoresis (MST). The annotated values ​​in the figure represent the dissociation constants between HSP90AA1 and the three heat shock protein inhibitors determined by MST. DETAILED DESCRIPTION OF THE INVENTION

[0044] In order to provide a clearer description of the technical solutions of the present invention and highlight the advantages thereof, the following detailed description is presented in conjunction with specific examples. It should be noted that these examples are not intended to limit the scope of the present invention. [Example]

[0045] In the following examples, Examples 1 and 4-9 demonstrate that this method can be used to identify proteins and protein regions that change energy state in cell lysates upon treatment with anticancer drugs, metabolites, antibodies, peptides with post-translational modifications, or metal ions. Example 2 demonstrates that PELSA generates more peptides in response to changes in energy state or in response to ligand binding, and that the magnitude of these peptide changes in PELSA is significantly greater than that of competing methods such as LiP-MS. Example 3 demonstrates that PELSA is currently the most sensitive method available for identifying proteins and protein regions with altered energy states. Example 10 demonstrates the use of other nonspecific or specific enzymes for digestion by detecting peptides containing two cleavage sites, both of which can reflect the local stability of the protein and can also identify proteins with changes in energy state. Example 11 demonstrates that peptide quantification by PELSA can also be performed using dimethyl-labeling-based quantification. Example 12 demonstrates that PELSA can be used to determine the affinity between ligands and their binding regions. In addition, Examples 1-9 demonstrate that PELSA can be used to identify proteins and their corresponding binding regions that bind to drugs, broad-spectrum kinase inhibitors, metabolites, antibodies, peptides with post-translational modifications, and metal ions. Furthermore, the binding of ligands to target proteins demonstrated in Examples 1-9 can induce conformational changes in the target proteins, thereby highlighting the applicability of this method in studying conformational changes in proteins.

[0046] In Example 1, treatment of BT474 cell lysates with 100 nM lapatinib induced a significant energy state change (-log 10This energy state change occurred precisely in the region corresponding to the kinase domain, confirming that PELSA can identify proteins and regions with energy state changes, or proteins bound to ligands or their ligand-binding regions. Furthermore, based on the PELSA results, we observed that a greater number of off-target kinase proteins were identified when the lapatinib concentration was increased up to 1 μM.

[0047] In Example 2, a comparison between LiP-MS and PELSA was performed to identify proteins with energy state changes induced by treatment with MTX and SHP099 in HeLa cell lysates. The results demonstrated that PELSA was able to identify a greater number of peptides responding to changes in protein energy state or ligand binding compared to LiP-MS (2-5 times more than those identified by LiP-MS). In addition, the amplitude of these peptides responding to ligand binding or energy state changes was 4-6 times greater than that in LiP-MS.

[0048] In Example 3, 121 and 111 proteins were identified by PELSA as kinase target proteins of staurosporine in K562 and HeLa cell lysates, respectively. These numbers were 2.1-fold higher than those reported using the TPP method (53 kinase target proteins) and 12.3-fold higher than those reported using the LiP-Quant method (9 kinase target proteins), when the same proportions of kinase target proteins were achieved. This result emphasizes that PELSA is the most sensitive method for identifying proteins with altered energy states.

[0049] Examples 4-6 demonstrated the high sensitivity of PELSA in detecting weak protein-ligand interactions and precisely locating the ligand-binding region.

[0050] In Example 7, PELSA was successfully used to identify epitopes of antigens that bind to antibodies in HeLa cell lysates, demonstrating the ability of PELSA to identify interfaces of other protein-protein interactions.

[0051] In Example 8, PELSA successfully identified the recognition domain for tyrosine phosphorylation (pYEEI) in BT474 cell lysates, providing a powerful tool for identifying the recognition domains of other post-translational modifications. In Example 9, PELSA successfully identified the recognition domain for 112 Zn 2+ We successfully identified the binding protein and detected Zn in cell lysates. 2+ The binding region was precisely localized. This finding demonstrates that PELSA can identify ligand-protein interactions regardless of the size of the ligand. Furthermore, based on the PELSA results, the addition of zinc ions stabilized the calcium ion-binding motif (EF-hand / EH motif) but destabilized the IQ motif, which interacts with the EF-hand motif. This suggests that zinc ion binding to the EF-hand motif leads to its dissociation from the IQ motif, resulting in its destabilization. These results also highlight the potential of PELSA in studying the binding interface of protein complexes.

[0052] Example 10 investigated the use of various enzymes other than trypsin in PELSA. The results showed that the number of peptides or proteins identified using chymotrypsin and proteinase K was significantly lower than that using trypsin, but the identification of MTX-binding proteins could be achieved. This suggests that the use of enzymes other than trypsin in PELSA also allows the identification of high-abundance target proteins.

[0053] In Example 11, the combination of PELSA and dimethyl labeling-based quantification was explored to identify target proteins for heat shock protein inhibitors. The results showed that the combination of PELSA with dimethyl labeling-based quantification and the second significantly altered peptide screening strategy enabled the identification of target and off-target proteins of heat shock protein inhibitors with high specificity.

[0054] Example 12 demonstrates that PELSA can accurately determine the binding affinity between a ligand and a protein. [Example 1]

[0055] Application of PELSA to identify proteins and protein regions with altered energy state after treatment of BT474 cell lysates with the anticancer drug lapatinib

[0056] (1) One dish of BT474 cells (approximately 5 × 10 7 The cells (1000 cells) were resuspended in 1 mL of lysis buffer (PBS supplemented with 1% (v / v) protease inhibitor (Sigma, Cat. No. P8340-5mL)). The mixture was subjected to three freeze-thaw cycles (freezing in liquid nitrogen for 2 minutes, thawing in a 37°C water bath for 2 minutes, repeated three times) to obtain a mixture of cell debris and cell lysate. The mixture was centrifuged at 500 g and 4°C for 10 minutes, and the supernatant was collected to obtain the cell lysate. Protein concentration was determined using a Pierce™ 660 nm protein assay (Thermo, USA).

[0057] (2) The protein concentration of the extracted cell lysate was adjusted to 1 mg / mL using cell lysis buffer [PBS buffer containing 1% (v / v) protease inhibitors, Sigma, catalog number P8340-5mL]. Four aliquots of the cell lysate were transferred to EP tubes. For the experimental group, 0.5 μL of lapatinib (Selleck, catalog number S2111) stock solution at different concentrations was added to each 50 μL aliquot to achieve final concentrations of 100 nM, 1 μM, 10 μM, and 100 μM (corresponding to lapatinib stock concentrations of 10 μM, 100 μM, 1 mM, and 10 mM, respectively; lapatinib was dissolved in DMSO). For the control group, 0.5 μL of DMSO was added to 50 μL of the cell lysate. Both the experimental and control groups were performed in quadruplicate and incubated at room temperature for 30 minutes.

[0058] (3) Trypsin (Sigma, Catalog No. T1426) was added to the experimental and control groups at an enzyme-to-protein ratio of 1:2 (by weight). Samples were digested on a shaker at 37°C and 1000 rpm for 1 minute. Digestion was terminated by adding 165 μL of pH 8.2 HEPES buffer (Sigma, Catalog No. H3375) containing 8 M guanidine hydrochloride (Sigma, Catalog No. G3272) to obtain digested samples.

[0059] (4) TCEP (Sigma, Cat. No. C4706) and CAA (Sigma, Cat. No. 22790) were added to the digested sample to a final concentration of 10 mM and 40 mM, respectively. The digested sample was heated at 95°C for 5 minutes and then cooled to room temperature. The sample was then transferred to a 10 kDa ultrafiltration tube (Sartorius, Cat. No. VN01H02) and centrifuged at 14,000 g for 50 minutes. The digested peptides were collected by filtration through the membrane, and the ultrafiltration membrane was washed with 200 μL of pH 8.2 HEPES buffer. The peptides from both ultrafiltration steps were combined.

[0060] (5) The peptide solution was desalted using a 200 μL tip column filled with 2 mg of Oasis HLB filter (Waters, USA). The desalted peptides were lyophilized to obtain a peptide mixture.

[0061] (6) The peptide mixture described above was reconstituted in 20 μL of 0.1% (v / v) formic acid (Sigma, catalog number V900803) and subjected to LC-MS / MS analysis. Each sample was analyzed once using data-independent acquisition (DIA) mass spectrometry. Spectronaut software (Biognosys, Switzerland) was used to identify and quantify peptides, providing information about the corresponding protein, the location of the peptide within the protein, and its abundance in each sample. In this example, we first analyzed the secondary structure at the cleavage sites in PELSA. The results showed that 59% of the cleavage sites in the PELSA experiment were located within the helical regions of the protein (Figure 2A), suggesting that the protein structure was disrupted during PELSA, resulting in the digestion of protein regions in a stable, low-energy state.

[0062] (7) The abundance of peptides in the lapatinib-treated and untreated groups was statistically evaluated using an empirical Bayes t-test to determine the P value and fold change for each peptide, which represents the significance and magnitude of the change in peptide abundance upon lapatinib treatment. For each protein, the peptide with the lowest P value (highest significance) among all its peptides was selected to represent the protein for screening ligand-binding proteins.

[0063] In this study, -log 10 Proteins with a P value >5 were defined as proteins with a change in energy state. Figure 2B shows the -log on the y-axis for all quantified proteins. 10Volcano plots with p-values ​​and log2 fold change (log2FC) on the x-axis are shown. In Figure 2B, among the 5,774 quantified proteins, only ERBB2, a known target protein of lapatinib, showed a significant fold change under treatment with 100 nM lapatinib.

[0064] ERBB2 is a receptor tyrosine kinase located on the cell membrane. It consists of extracellular, transmembrane, and intracellular regions, including the kinase domain of ERBB2 on which lapatinib acts. In Figure 2C, the x-axis represents the protein sequence from the N-terminus to the C-terminus, the boxes represent the kinase domain on which lapatinib acts, and the y-axis represents the fold change in abundance of ERBB2 peptides under 100 nM lapatinib treatment. Analysis of the identified ERBB2 peptides (Figure 2C) showed that only peptides located in or adjacent to the kinase domain showed significant changes (|log2FC|>0.3 and -log 10 The results demonstrated the ability of PELSA to determine drug-binding regions.

[0065] Furthermore, at higher concentrations of lapatinib (100 μM, 10 μM, and 1 μM), peptides within the kinase domain of ERBB2 were observed to be more resistant to digestion (Figure 2D). Of note, only ERBB2 exhibited changes in its energy state at 100 nM lapatinib, and increasing lapatinib concentrations led to changes in the energy state of several other kinases, such as CHEK2, SLK, RIPK2, and YES1 (Figure 2E). Furthermore, most of the peptides that showed changes in these kinases were located within the individual kinase domains of the proteins (Figure 2F). These findings indicate that lapatinib exhibits off-target interactions with other kinases at higher doses, suggesting that PELSA can be used to assess drug promiscuity and guide drug design and synthesis.

[0066] Furthermore, we also observed that the non-kinase protein PTGES2 exhibited a change in its energy state at high concentrations of lapatinib (Figure 2E). To verify the binding of PTGES2 and lapatinib, we performed Western blots using procedures and conditions similar to those described above, with the following modifications: BT474 cells (cell count approximately 5 × 10) in a large dish were cultured. 7The cells (800 cells) were suspended in PBS, and cell lysis and protein concentration determination were performed as described. The resulting cell lysate was divided into eight portions and treated with equal volumes of different final concentrations of lapatinib (lapatinib dissolved in DMSO) or DMSO alone, resulting in final lapatinib concentrations of 100 μM, 50 μM, 10 μM, 1 μM, 0.1 μM, 0.01 μM, 0.001 μM, and 0 μM. The mixture was incubated at room temperature (25°C) for 30 minutes. After incubation, trypsin was added at a protease-to-protein weight ratio of 1:40, and the mixture was digested at 37°C for 1 minute. The sample was heated at 95°C, followed by the addition of 5x loading buffer (1 / 4 volume of the protein solution) and heating at 95°C for an additional 5 minutes to terminate the digestion. Western blotting was used to detect undigested protein. Samples were run in a stacking gel at 80 V for 20 minutes, then in a resolving gel at 120 V for 60 minutes. Transfer blots were performed at a constant current of 250 mA for 40 minutes. After blocking, a primary antibody against PTGES2 (Proteintech, USA) was added at a dilution of 1:800 and incubated at room temperature for 1 hour, followed by incubation with a goat anti-rabbit HRP-IgG secondary antibody (Abcam, UK) for 1 hour at room temperature. Chemiluminescence detection was performed using ECL reagents (Thermo Fisher Scientific, USA), and a Fusion FX5 chemiluminescence system (Vilber Infinit, France) was used for imaging of undigested proteins. Undigested protein bands were quantified based on their intensity. GAPDH was used as a reference protein and was incubated with a primary antibody against GAPDH overnight at 4°C, followed by incubation with a rabbit anti-mouse HRP-IgG secondary antibody (Abcam, UK) for 1 hour at room temperature for chemiluminescence detection. As shown in Figure 2G, with increasing concentrations of lapatinib, the bands became progressively darker, indicating that an increasing amount of PTGES2 protein was protected from digestion, reaching a maximum at a lapatinib concentration of 50 μM.These results suggest that the resistance of PTGES2 to digestion is dependent on lapatinib concentration, thus providing further evidence supporting the view that PTGES2 may be an off-target protein of lapatinib.

[0067] The above analysis results demonstrate that PELSA can identify ligand-binding proteins with high specificity and determine the ligand-binding region based on peptides with altered abundance. This unbiased omics-level screening method also facilitates the identification of off-target proteins, such as other kinases and PTGES2, as identified in this example. This demonstrates that this method can assess drug promiscuity and provide guidance for drug design and development. Furthermore, only peptides within the ligand-binding region showed significant abundance fold changes, demonstrating that this method can reveal the drug's active region. The dose-dependent changes in target protein abundance upon treatment with different concentrations of ligand suggest the ability of PELSA to determine the binding affinity between the ligand and the target protein. [Example 2]

[0068] Comparison between LiP-MS and PELSA for identifying proteins with altered energy states upon treatment with methotrexate (MTX) and SHP099 in HeLa cell lysates

[0069] To demonstrate the advantages of PELSA in identifying proteins with altered energy states, in this example, the existing method LiP-MS was compared with the present invention (PELSA) for identifying proteins with altered energy states based on digestion after treatment with MTX and SHP099 in HeLa cell lysates. DHFR is the target protein of MTX, and PTPN11 is the target protein of SHP099. In this example, the PELSA procedure and experimental conditions were the same as those in Example 1, with the following differences: HeLa cell samples were used, and the ligands were MTX (Selleck, Cat. No. S1210) at a final concentration of 10 μM or SHP099 (Selleck, Cat. No. S8278) at a final concentration of 10 μM. The LiP-MS procedure adheres to the protocol described in the literature by Piazza et al. [Nature Communication, 2020, 11(1):4200], with the following steps (the following describes the procedures and experimental conditions that differ from those in Example 1):

[0070] (1) The cell lysis buffer used consisted of 60 mM HEPES pH 7.5, 150 mM KCl, and 1 mM MgCl2 [Piazza et al., Nature Communication, 2020, 11(1):4200].

[0071] (2) After incubation of the cell lysate with the drug, proteinase K (Sigma, catalog number P2308) was added at a protease to protein ratio of 1:100 (wt / wt). The mixture was incubated at 25°C on a shaker at 1000 rpm for 4 minutes, heated at 98°C for 1 minute, and then an equal volume of 10% by volume sodium deoxycholate (Sigma, catalog number D6750) was added. The mixture was heated at 98°C for an additional 4 minutes to denature the protein fragments and obtain a protein solution.

[0072] (3) After denaturation, TCEP (Sigma, Cat. No. C4706) and CAA (Sigma, Cat. No. 22790) were added to the above protein solution to final concentrations of 10 mM and 40 mM, respectively. The sample was heated at 98°C for 5 minutes, cooled to room temperature, and then diluted with 4 volumes of pH 8.2 60 mM HEPES buffer to obtain a final concentration of 1% sodium deoxycholate.

[0073] (4) Lys-C (Wako Chemicals) was added at a protease to protein ratio of 1:100 (wt / wt) and incubated for 4 hours, followed by trypsin (Promega) at a protease to protein ratio of 1:50 (wt / wt) and incubation for 16 hours.

[0074] (5) After digestion, formic acid (FA) was added to achieve a final volume of 1.5%. The sample was allowed to precipitate for 10 min. Once the sodium deoxycholate precipitate reached equilibrium, the solution was centrifuged (twice) at 20,000 g for 10 min at room temperature to remove the sodium deoxycholate precipitate.

[0075] (6) The processes of peptide desalting, mass spectrometry quantification, software searching, and data analysis were the same as in Example 1, except that when searching using Spectronaut, the enzyme was set as “trypsin” and the digestion type was set as “semi-trypsin.”

[0076] In Figures 3A and 3B, -log 10Peptides meeting the criteria of a P value >2 (above the horizontal dashed line in the figure) and a |log2 fold change| >0.3 (outside the vertical dashed line in the figure) were defined as ligand-responsive peptides. In the MTX-DHFR system, PELSA identified 2-fold more MTX-responsive DHFR peptides than those identified by LiP-MS. In the SHP099-PTPN11 system, PELSA identified 5.25-fold more SHP099-responsive PTPN11 peptides than those identified by LiP-MS. Furthermore, the magnitude of the response of DHFR peptides to MTX in PELSA was 4.3-fold higher than that in LiP-MS (Figure 3C), and the magnitude of the response of PTPN11 peptides to SHP099 in PELSA was 6.4-fold higher than that in LiP-MS (Figure 3C). For example, in the MTX-DHFR system, a peptide located in the MTX-binding site showed a significant fold change in the PELSA experiment, but no fold change was detected for the same peptide in the LiP-MS experiment (Figure 3D). This peptide in DHFR is in a low-energy state helical structure, which is difficult to access by gentle digestion in LiP-MS. However, destructive digestion in PELSA allows for cleavage of regions of the protein with low energy states, enabling detection of changes in this peptide. Similarly, in the SHP099-PTPN11 interaction system, changes in the peptide located in the SHP099-binding site and within the internal structure of the protein were only detected in the PELSA experiment (Figure 3E). This finding further demonstrates that PELSA can disrupt stable internal structures of proteins, thereby capturing changes in energy states occurring within these regions. These findings suggest that PELSA outperforms LiP-MS in identifying a larger number of ligand-responsive peptides in target proteins. Furthermore, ligand-responsive peptides identified by PELSA exhibit significantly higher response magnitudes compared to those identified by LiP-MS, again making PELSA more sensitive in identifying proteins with altered energy states.For example, in this example, GART, which is involved in the pharmacokinetics of MTX [Mikkelsen TS et al., Pharmacogenet Genomics, 2011, 21(10): 679-686], was identified as a protein with a change in energy state only in PELSA but not in LiP-MS (Figure 3A).

[0077] This example demonstrates that the present invention (PELSA) can generate more ligand-responsive peptides than existing LiP-MS methods, and these ligand-responsive peptides exhibit larger fold changes in response to ligand binding. Therefore, PELSA allows for the identification of a greater number of ligand-binding proteins. This partly explains the extremely high sensitivity of the present invention in identifying ligand-binding proteins. [Example 3]

[0078] PELSA identification of proteins and protein regions in HeLa and K562 cell lysates with altered energy states upon treatment with staurosporine, a broad-spectrum kinase inhibitor.

[0079] This example demonstrated the high sensitivity of PELSA in identifying staurosporine-binding proteins and proteins with altered energy states by comparison with results reported in the literature.

[0080] The experimental procedures and conditions were similar to those in Example 1, except that K562 and HeLa cell samples were used. The ligand molecule used in the experimental group was 20 μM staurosporine (final concentration, Selleck, catalog number S1421). As shown in Figure 4A, PELSA revealed that a large number of kinases were stabilized in both HeLa and K562 lysates upon staurosporine treatment. By setting the cutoff for the proportion of kinases among identified target proteins at 80%, PELSA identified 121 kinase targets stabilized in K562 cell lysates (out of a total of 143 target proteins) and 111 kinase targets stabilized in HeLa cell lysates (out of a total of 135 target proteins). In contrast, according to published studies, LiP-Quant identified only nine kinase targets that responded to staurosporine under the same kinase percentage cutoff (i.e., ≥80% of target proteins that are kinases), while TPP identified 53 kinase targets (out of a total of 60 target proteins) that responded to staurosporine (Figure 4B). These findings clearly demonstrate the vastly superior sensitivity of PELSA compared to existing methods.

[0081] We further compared the protein and peptide coverage depths of LiP-Quant, TPP, and PELSA. We observed that although a greater number of peptides were identified in the LiP-Quant experiment (Figure 4C), and LiP-Quant achieved higher overall protein sequence coverage compared to PELSA (Figure 4D), PELSA identified a significantly higher number of kinases (i.e., kinases identified as target proteins) that responded to staurosporine than LiP-Quant (111 kinases vs. 9 kinases). Further analysis revealed that the sequence coverage of kinase targets identified by PELSA was lower than that identified by LiP-Quant in the LiP-Quant experiment (Figure 4D). This suggests that PELSA can identify binding proteins as target proteins even at lower sequence coverage. In contrast, LiP-Quant requires higher protein sequence coverage to accurately identify target proteins due to the presence of many unrelated peptides resulting from complete digestion with trypsin.

[0082] Although TPP identified more proteins (7673) than PELSA (6310) in -K562 (Figure 4C), PELSA ultimately identified 2.28-fold more kinase targets than TPP, further highlighting the high sensitivity of PELSA in target protein identification. Analysis of the melting points of kinase targets identified by TPP and PELSA revealed that TPP has limited ability to identify kinase targets with very high or very low melting temperatures. In contrast, PELSA efficiently identifies kinase targets with a wide range of melting temperature points (Figure 4E).

[0083] PELSA identified a total of 192 target proteins for staurosporine in the two cell lines, of which 154 were kinases, accounting for 80% of all target proteins (Figure 4F). In addition, PELSA could identify the region that binds to staurosporine, as shown in Figure 4G. Among the significantly altered peptides identified by PELSA, over 92% of the peptides were located within or close to (within 10 amino acid residues of) the kinase domain in both K562 and HeLa lysates.

[0084] The experimental results show that PELSA not only exhibits extremely high sensitivity in identifying ligand-binding proteins, but also accurately identifies the regions of the protein that bind to the ligand. [Example 4]

[0085] PELSA identification of proteins and protein regions with altered energy status upon treatment with the metabolite folate in K562 cell lysates.

[0086] Unlike the strong interaction between lapatinib and ERBB2 (with an affinity of approximately 9 nM), folic acid exhibits a weaker affinity for its binding protein. For example, previous studies have shown that folic acid binds to its target protein, DHFR, with a dissociation constant ranging from 3 to 60 μM [Ozaki Y et al., Biochemistry, 1981, 20(11): 3219-3225]. Therefore, we used folic acid to examine whether PELSA can analyze changes in the energy state of proteins induced by ligand binding to low-affinity proteins.

[0087] Most experimental procedures and conditions were the same as in Example 1, except that K562 cell samples were used. After the cell samples underwent three rounds of freeze-thawing, the supernatant was obtained by centrifugation at 500 g and 4° C. for 10 minutes. To remove endogenous folic acid, a protein desalting step was performed using a Zeba Spin desalting column (Thermo Fisher Scientific). The protein concentration of the desalted lysate was determined using a Pierce™ 660 nm protein assay (Thermo, USA). The protein concentration was adjusted to 1 mg / mL using cell lysis buffer, i.e., PBS buffer containing 1% (v / v) protease inhibitor (i.e., cocktail; Sigma, catalog number P8340-5mL). The ligand used in this experiment was folic acid (Sigma, catalog number F7879) at a final concentration of 50 μM. Subsequent steps were performed as described in Example 1.

[0088] As shown in Figure 5A, the stability of DHFR, a known protein targeted by folate, showed the most significant change. By mapping DHFR peptides with changes in their protein structure (Figure 5B), we observed that the region experiencing the most significant stability change corresponded to the folate binding site. Furthermore, we identified stability changes in three proteins (ATIC, MTHFR, and GART) that interact with folate analogs. The stabilized regions of these proteins coincided with the binding sites of folate analogs (Figures 5C-5E). P3H1 is a prolyl 3-hydroxylase involved in the prolyl hydroxylation of collagen, and literature has suggested that folate may participate in proline hydroxylation in collagen [Haustvast J, et al., British Journal of Nutrition, 1974, 32(2): 457-469]. Our experimental results revealed that the hydroxylase domain of P3H1 was stabilized upon addition of folate ( Figure 5F ), providing evidence for the involvement of folate in collagen prolyl hydroxylation. [Example 5]

[0089] PELSA identification of proteins and protein regions with altered energy status upon treatment with the metabolite leucine in K562 cell lysates.

[0090] The dissociation constants of leucine and its target proteins, LARS1 and SESN2, are 95 μM [Kim S, et al., Cell Reports, 2021, 35(4): 109301] and 20 μM [Wolfson RL, et al., Science, 2016, 351(6268): 43-48], respectively. Therefore, we also use leucine to evaluate the applicability of PELSA in analyzing changes in protein energy states induced by weak-affinity ligand-protein interactions.

[0091] The procedure and conditions were the same as those outlined in Example 4, except that leucine (Sigma, Cat. No. 61819) was used as the ligand at a final concentration of 5 mM. As shown in Figure 5G, we observed significant changes in the energy states of known leucine target proteins, including LARS1, LARS2, GLUD1, and SENS2. LARS1 has two leucine-binding sites located in the CD and CP domains, corresponding to the synthesis and editing sites, respectively. We noted that peptides from the synthesis site of LARS1 (labeled CP in Figure 5H) showed a more pronounced magnitude of change compared to those from the editing site (labeled CD in Figure 5H). Furthermore, peptides from the C-terminal domain, which does not contribute to leucine binding, showed no response (Figure 5H). These findings demonstrate the ability of PELSA to identify distinct binding sites in a single protein.

[0092] We also observed a significant change in the stability of SLC1A5, another known leucine-binding protein (Figure 5G). SLC1A5 is a membrane protein composed of intracellular, transmembrane, and extracellular regions. The Na-dicarboxylate symporter domain (residues 54–483) located in the extracellular region plays a role in amino acid transport. Among the identified peptides of SLC1A5, only the peptide from the Na-dicarboxylate symporter domain (residues 190–212) showed significant changes (Figure 5I), demonstrating that our method can be used to identify and determine the binding region of membrane proteins. Interestingly, we also observed that leucine stabilized PPIP5K1 and PPIP5K2 (Figure 5G), and this was specific to their shared histidine phosphatase domain (Figure 5J).

[0093] This example shows that PELSA is very effective in identifying weak interactions between metabolites and proteins, and also in accurately identifying the binding region of metabolites.In addition, the successful identification of SLC1A5 as a leucine target protein demonstrates that ELSA works well in identifying target membrane proteins.Several proteins that can interact with leucine were also identified by PELSA, shedding light on the future investigation of the function of leucine. [Example 6]

[0094] PELSA identification of proteins and protein regions with altered energy states upon treatment with the metabolite alpha-ketoglutarate (αKG) in HeLa cell lysates.

[0095] The procedures and conditions were followed as outlined in Example 4, except that the cell lysate was derived from HeLa cells and αKG (Sigma, Cat. No. 75890-25g) was used as the ligand under study at a final concentration of 2 mM. 10 Proteins meeting the criteria of P value > 3.4 and log2FC < -0.5 were considered as proteins stabilized by 2 mM αKG.

[0096] As shown in Figure 5K, PELSA identified 40 proteins stabilized by 2 mM αKG in 2 mM αKG-treated HeLa cell lysates, of which 30 were previously known αKG target proteins. This represents the largest number of known αKG target proteins identified in a single experiment to date. Although literature reports the use of LiP-MS to identify αKG target proteins in Escherichia coli [Piazza I et al., Cell, 2018, 172(1-15)], only two of the 34 identified target proteins were previously known αKG binding proteins. In addition, the two-dimensional profiles showing the stability changes determined by PELSA (Figure 5L) demonstrated that the regions of the αKG target proteins with energy changes accurately matched the αKG binding regions.

[0097] In summary, this example demonstrates that PELSA exhibits extremely high sensitivity in analyzing weak interactions between metabolites and proteins and can identify metabolite-binding regions. [Example 7]

[0098] PELSA identification of proteins and protein regions with changes in energy state upon treatment with antibodies in HeLa cell lysates.

[0099] This example demonstrates the application of PELSA in identifying antibody-binding epitopes by identifying protein regions with altered energy states in HeLa cell lysates upon treatment with an antibody.

[0100] Most experimental procedures and conditions were the same as in Example 1, except that HeLa cells were used, and in the experimental group, two commercially available antibodies, DHFR antibody (Wabways, China, RRID:AB_2877179) and CDK9 antibody (Wabways, China, RRID:AB_2877178), were used as the investigated ligands at final concentrations of 2% (v / v) and 1% (v / v), respectively, while in the control group, an equal volume of vehicle was used. 10 Proteins with a P value >5 were considered to have a change in energy state. Figure 6A shows a schematic diagram of the identification of antibody-binding epitopes in cell lysates. Figure 6B (left) shows a volcano plot of proteins obtained from HeLa cell lysates treated with DHFR antibody, while Figure 6B (right) shows a volcano plot of proteins obtained from HeLa cell lysates treated with CDK9 antibody. Addition of DHFR and CDK9 antibodies caused significant changes in the energy states of DHFR and CDK9, respectively, demonstrating that PELSA can directly identify antigen proteins in cell lysates. Other proteins showing changes in energy state may be due to nonspecific binding to the antibody.

[0101] Further analysis of all DHFR peptide changes identified by PELSA revealed that peptides with altered energy states corresponded precisely to known epitopes recognized by the antibody (Figures 6C and 6D). For CDK9, two peptides exhibited opposite abundance changes. The CDK9 antibody recognized the epitope sequence PATTNQTEFERVF located at the C-terminus of CDK9. The sequence NPATTNQTEFER contains a site within the epitope, resulting in a reduction in peptide yield upon antibody addition (Figure 6D). Conversely, the sequence NPATTNQTEFERVF (NPxxVF), which completely encompasses the epitope sequence but lacks the trypsin cleavage site located in the epitope, showed an increase in peptide yield upon antibody addition (Figure 6D). Further analysis revealed that the combined intensities of these two peptides were minimally affected by the addition of the antibody (Figure 6E). Therefore, the opposing changes observed in these two peptides may be due to the protective effect of the antibody, which makes the C-terminal R site of the NPATTNQTEFER sequence less susceptible to cleavage, resulting in a decrease in the yield of the corresponding peptide. As a result, the NPATTNQTEFERVF (NPxxVF) sequence, which should remain unchanged upon CDK9 antibody treatment, remained more abundant. These results demonstrate the high resolution of PELSA in identifying antibody-binding epitopes.

[0102] This example demonstrates that PELSA can efficiently identify interactions between antibodies and proteins and reveal antibody-protein binding sites, demonstrating the potential of PELSA for application to various protein-protein interaction systems. [Example 8]

[0103] PELSA identification of proteins and protein regions with altered energy state upon treatment with post-translationally modified peptides in BT474 cell lysates.

[0104] Protein post-translational modifications (PTMs) play crucial roles in various biological activities. Regulation of these activities often requires the recognition and recruitment of effector proteins by downstream proteins. Therefore, identifying proteins that recognize PTMs is essential for understanding the function of PTMs and studying disease mechanisms. Phosphorylated tyrosine-glutamic acid-glutamic acid-isoleucine (referred to as pYEEI) is known to be recognized by proteins containing SH2 domains. In this analysis, we demonstrated that PELSA can be used to identify domains that recognize PTMs by studying proteins and protein regions that have changes in energy state upon treatment with pYEEI in BT474 cell lysates.

[0105] Most experimental procedures and conditions were the same as those in Example 1, with the following exception: an additional 2 mM phosphatase inhibitor was added to the cell lysis buffer to prevent the phosphorylated peptide from being dephosphorylated by active phosphatases present in the cell lysate. For parallel comparison with the results of the pull-down experiments, we used 100 μM N-terminal biotin-conjugated pYEEI (abbreviated as biotin-pYEEI, synthesized by Qiangyao Biotechnology) as a ligand molecule. Two control groups were included to eliminate any potential effects arising from the phosphate group and the YEEI motif. One control group was treated with 100 μM N-terminally biotinylated phosphorylated serine-glutamic acid-glutamic acid-isoleucine (abbreviated as biotin-pSEEI, synthesized by Qiangyao Biotechnology), and the other control group was treated with 100 μM N-terminally biotinylated tyrosine-glutamic acid-glutamic acid-isoleucine (abbreviated as biotin-YEEI, synthesized by Qiangyao Biotechnology).

[0106] The pull-down experiment was performed as follows: BT474 cells were used, and cell lysis, lysate preparation, and protein concentration measurement and adjustment were performed as described in Example 1. After adjusting the protein concentration of the cell lysate to 1 mg / mL, 100 μL of the cell lysate was treated with biotin-pYEEI at a final concentration of 100 μM as the experimental group. Two additional 100 μL of the cell lysate were treated with biotin-pSEEI at a final concentration of 100 μM and biotin-YEEI at a final concentration of 100 μM, respectively, to serve as control groups. This process was repeated three times, and the cell lysate was incubated with the added ligand molecule at room temperature for 30 minutes. After incubation, 200 μL of avidin beads (Thermo, USA) was added to each group and incubated overnight at 4 °C. The beads were then washed four times with 400 μL of wash buffer (1% phosphatase inhibitors in PBS (i.e., 1% cocktail) and 0.5% NP40) followed by four washes with 400 μL of cell lysis buffer (1% phosphatase inhibitors in PBS (i.e., 1% cocktail)). The protein was eluted by adding 100 μL of HEPES buffer (pH 8.2) containing 8 M guanidine hydrochloride, and the elution step was repeated twice. The eluted protein was treated with 10 mM TCEP and 40 mM CAA and heated at 95°C for 5 minutes to ensure complete alkalinization. The eluted protein solution was transferred to an ultrafiltration tube and centrifuged at 14,000 g for 30 minutes, and the filtrate was discarded. The ultrafiltration membrane was washed twice with 200 μL of 10 mM ammonium bicarbonate (NH4HCO3) buffer, and the filtrate was discarded. Then, 100 μL of 10 mM NH4HCO3 buffer was added to resuspend the proteins retained on the membrane, and 2 μg of trypsin (Promega) was added overnight for digestion. The next day, the ultrafiltration tube was centrifuged at 14,000 g for 50 minutes to collect the peptides obtained from the digestion. The ultrafiltration membrane was washed with 100 μL of NH4HCO3 buffer, and the filtrate was collected to reuse the remaining peptides. The peptides from both ultrafiltration steps were pooled and lyophilized.The procedures for mass spectrometry, software analysis, and data validation were the same as those described in Example 1.

[0107] Figure 7A shows a schematic diagram illustrating the identification of tyrosine phosphorylation recognition domains in cell lysates. Using YEEI as a control, an empirical Bayes t-test was performed to obtain a P value (pYEEI / YEEI). Using pSEEI as a control, an empirical Bayes t-test was performed to obtain a P value (pYEEI / pSEEI). The logarithms of both P values ​​were plotted, and log 10 P value(pYEEI / YEEI)>-3.1, log2FC(pYEEI / YEEI)<0, -log 10 Proteins meeting the criteria of P value (pYEEI / pSEEI) > 3.1 and log2FC(pYEEI / pSEEI) < 0 were defined as proteins with a reduced energy state, i.e., proteins stabilized by pYEEI treatment. This resulted in the identification of 28 proteins stabilized by pYEEI. Of these 28 proteins, 9 contained SH2 domains and 8 were found to be calcium ion-associated proteins (Figure 7B). However, no SH2 domain-containing proteins were identified in the pull-down experiment (Figure 7C), likely due to the high wash intensity. PELSA also successfully identified the recognition domain for pYEEI. As shown in Figures 7D and 7E, only peptides within the SH2 domain were stabilized upon addition of pYEEI, while the abundance of other peptides in other regions remained unchanged. Furthermore, the identified calcium ion-associated proteins were only stabilized in specific EF-hand domains (Figure 7F).

[0108] These results demonstrate that PELSA can identify proteins that recognize PTMs and recognition regions with altered energy states upon treatment of cell lysates with post-translationally modified peptides. [Example 9]

[0109] PELSA identification of proteins and protein regions with altered energy states upon treatment with metal ions in HeLa cell lysates.

[0110] In this example, we investigated whether PELSA can be applied to detect changes in the energy state of proteins induced by their binding to small metal ions.

[0111] Most experimental procedures and conditions were the same as those in Example 1, except that HeLa cells were used, and the cell lysis buffer was that of Example 1, supplemented with 2 mM ethylenediaminetetraacetic acid sodium salt (EDTA, purchased from Sigma). After cell lysis according to the procedure described in Example 1, the added EDTA was removed by two rounds of protein desalting using a Zeba Spin Desalting Column (Thermo Fisher Scientific). Zinc chloride (Sigma, Cat. No. 450111-10G) was added to the cell lysate at a final concentration of 30 μM, which served as the experimental group. -log 10 Proteins meeting the criteria of P value > 3 and log2FC < -0.5 were selected as Zn 2+ The protein was considered to be stabilized by

[0112] Figure 8A shows that PELSA 2+ This indicates that a significant number of metal ion-binding proteins were identified that were stabilized by treatment with Zn. 2+ Of all 280 proteins stabilized by the treatment, over 66% (185 proteins) were already known metal ion-binding proteins. In contrast, within the entire identified proteome, the proportion of metal ion-binding proteins was only 19%, indicating that PELSA was unable to identify Zn 2+ We demonstrated that metal ion-binding proteins can be successfully identified by detecting changes in their energy state during processing. 2+ But, Ca 2+It can bind to the EF-hand motif of binding proteins (Tsvetkov PO et al., Front Mol Neurosci, 2018, 11: 459) and Mg 2+ Binds to binding proteins and Mg 2+ It has been reported that divalent metal ions can occupy binding sites [Dudev T et al., Chemical Reviews, 2003, 103(3): 773-787], suggesting the promiscuous nature of these divalent metal ions in metal ion-binding proteins. Consistent with the literature, we found that of 185 metal ion-binding proteins stabilized by zinc ions, 112 proteins were zinc ion-binding proteins and 73 proteins were other metal ion-binding proteins (Figure 8B). Of these 73 proteins, 25 proteins were Ca ion-binding proteins. 2+ 20 proteins are known to be Mg binding proteins. 2+ 15 proteins are known iron ion binding proteins, and 9 proteins are known Mn 2+ Zn-binding proteins, and six proteins were proteins that bound additional metal ions (Figure 8C). 2+ Within the group of binding proteins, 60 proteins contained zinc finger motifs (Figure 8C). Analysis of proteins with zinc finger motifs revealed that the median log2FC values ​​for peptides within zinc finger motifs were significantly lower compared to peptides outside the zinc finger motif (Figure 8D). Furthermore, analysis of individual zinc finger-containing proteins showed that only peptides located within the zinc finger structure were significantly stabilized (Figure 8E). These results support the conclusion that Zn 2+ We demonstrate the ability of PELSA to pinpoint binding sites. Furthermore, we demonstrate the ability of PELSA to pinpoint binding sites in proteins lacking zinc finger motifs, e.g., LIM Zn 2+ In the case of LIMA1, which has a binding domain, three peptides from this domain were quantified using PELSA (Figure 8F). 2+ The one directly involved in binding showed the largest fold change, but Zn2+ The other two peptides not directly involved in binding showed minimal changes (|log2FC|<0.3, Figure 8G). These results support the high accuracy of PELSA in precisely localizing the binding site.

[0113] Identified Ca 2+ Analysis of binding proteins revealed that 20 of the 27 proteins contain EF-hand / EH motifs (Ca 2+ Similar to proteins containing zinc finger motifs, the median log2FC values ​​for peptides within EF-hand / EH motifs were significantly lower than those for peptides outside these motifs (Figure 8H), revealing that Zn 2+ These Ca 2+ Furthermore, among the nine proteins containing iron ion-binding domains, the median log2FC values ​​for peptides within the iron ion-binding domains were significantly lower than those for peptides outside these domains (Figure 8I). 2+ showed that these iron ion-binding domains can act on iron ion-binding proteins.

[0114] When a ligand binds to a protein, it can dissociate the protein from its original complex, resulting in destabilization of the partner protein in the complex with the protein. We have observed destabilization in several proteins containing IQ motifs (-log 10 P value > 6 and log2FC > 0) (Fig. 8J). The destabilized peptides were located exactly at or close to the IQ motifs (Fig. 8K). The IQ motif is a known binding site for the EF-hand motif, which binds Zn 2+ (Figure 8H). Thus, the destabilization of the IQ motif is due to the interaction between the EF-hand motif and Zn. 2+This may be due to the dissociation between the IQ motif and the EF-hand motif caused by the interaction between IQ and EF-hand motifs. The IQ and EF-hand motifs are important interfaces for protein-protein interactions, suggesting that PELSA can be used to analyze the changes in energy state resulting from the association and dissociation of protein complexes in cells and to reveal the binding interfaces of proteins within these complexes.

[0115] We also observed destabilization of components of the 26S proteasome regulatory subunits, specifically PSMC1-6 (Figure 8J). Each of the PSMC1-6 proteins contains a P-loop-NTPase domain located at the binding interface of the complex formed with PSMC1-6. PELSA results demonstrated that only peptides within this domain showed significant destabilization (Figure 8L), indicating that Zn 2+ We showed that induced dissociation of the 26S proteasome regulatory subunit, thereby destabilizing the protein-protein interaction interface of PSMC1-6. These findings further suggest that this method can be used to study the dynamic changes resulting from the association and dissociation of protein complexes. [Example 10]

[0116] Analysis of proteins with altered energy state upon treatment with methotrexate (MTX) by other enzyme-based PELSA in HeLa cell lysates.

[0117] To evaluate the applicability of PELSA using proteases other than trypsin in identifying proteins with altered energy states, we performed a parallel comparison using trypsin and two other proteases with different cleavage specificities: chymotrypsin (purchased from Sigma, catalog number C3142) and proteinase K (abbreviated as PK, purchased from Sigma, catalog number P2308). These enzymes were used to analyze proteins with altered energy states in HeLa cell lysates upon treatment with MTX. Chymotrypsin primarily cleaves at the N-terminus of aromatic amino acids, whereas proteinase K exhibits broad cleavage specificity.

[0118] The procedures and conditions for trypsin-PELSA were the same as those in Example 1, except that HeLa cells were used; the experimental group was treated with MTX (dissolved in DMSO, 1 mM stock concentration; purchased from Selleck, catalog number S1210) at a final concentration of 10 μM as the ligand molecule.

[0119] The procedures and conditions for chymotrypsin-PELSA were the same as those in trypsin-PELSA, except for the following: trypsin was replaced with chymotrypsin; digestion was performed at 25°C, 1000 rpm for 1 minute; and the cleavage sites for digestion were set as F, W, Y, L, and M when searching with Spectronaut.

[0120] The procedure and conditions for proteinase K-PELSA were the same as those in trypsin-PELSA, except for the following: trypsin was replaced with proteinase K; digestion was performed at 25°C, 1000 rpm for 1 min; and the cleavage site for digestion was set as trypsin, nonspecific when searching with Spectronaut.

[0121] Figure 9A shows that trypsin-PELSA identified a total of 69,245 peptides, while chymotrypsin-PELSA and proteinase K-PELSA identified significantly fewer peptides compared to trypsin-PELSA (18,027 and 28,702 peptides, respectively). Correspondingly, trypsin-PELSA identified a greater number of proteins (5,487) compared to chymotrypsin-PELSA (2,710) and proteinase K-PELSA (1,937) (Figure 9B). Furthermore, proteins identified by trypsin-PELSA covered the majority of proteins identified by proteinase K-PELSA and chymotrypsin-PELSA (Figure 9B). This observation may be due to the fact that tryptic peptides generated by trypsin are more favorable for mass spectrometry identification compared to peptides generated by other enzymes. In terms of identifying proteins with altered energy states, DHFR, a known MTX-binding protein, is relatively abundant, so all three proteases used in PELSA (trypsin, chymotrypsin, and proteinase K) were able to detect this protein and distinguish it from background proteins (Figure 9C). This example demonstrates that PELSA is not limited to the use of trypsin and can also utilize other specific and nonspecific proteases to identify proteins with altered energy states. However, the problem of fewer proteins and peptides for identification may be encountered. [Example 11]

[0122] PELSA combined with dimethyl labeling quantification to identify proteins and protein regions with altered energy states in HeLa cell lysates upon treatment with three heat shock protein inhibitors.

[0123] The procedure was as follows.

[0124] (1) HeLa cells were used. The processes of cell lysis, protein extraction, and protein concentration determination followed the protocol described in Example 1. Six 50 μL aliquots of cell lysate were prepared in Eppendorf tubes. Three aliquots were treated with 100 μM geldanamycin, 100 μM tanespimycin, and 100 μM ganetespib (stock concentration: 10 mM; all dissolved in DMSO, purchased from Selleck). An equal volume of DMSO was added to the remaining three aliquots. The samples were then incubated at room temperature (25°C) for 30 minutes. After incubation, trypsin was added to each sample at a trypsin-to-protein ratio of 1:2 (wt / wt), and the samples were incubated at 37°C for 1 minute. Digestion was terminated by heating the samples at 100°C for 5 minutes.

[0125] (2) To each of the above protein solutions, a three-fold volume of sodium dihydrogen phosphate buffer (pH 6.5) containing 8 M guanidine hydrochloride was added. TCEP and CAA were added to final concentrations of 10 mM and 40 mM, respectively. The sample was heated at 95°C for an additional 5 minutes and then cooled to room temperature. The sample was then transferred to a 10 kDa ultrafiltration tube and centrifuged at 14,000 g for 50 minutes to collect the peptides obtained from the digestion. The ultrafiltration membrane was washed twice with 200 μL of sodium dihydrogen phosphate buffer (pH 6.5), and the filtrates from both ultrafiltration steps were pooled.

[0126] (3) Dimethyl labeling: Peptides in the drug-treated group were labeled with medium reagents, i.e., 16 μL of 4% by volume denatured formaldehyde (Sigma, Cat. No. 596388) and 16 μL of 0.6 M cyanoborohydride (Sigma, Cat. No. 156159). For the control group, 16 μL of 4% by volume formaldehyde (Sigma, Cat. No. 252549) and 16 μL of 0.6 M cyanoborohydride were added for photolabeling. The labeling reaction was carried out at 30 °C for 1 hour. After the reaction, 10 μL of 10% by volume ammonium hydroxide (Sigma, Cat. No. 338818) was added and incubated for another 30 minutes. Finally, peptides from the drug-treated and control groups were mixed in equal mass for further analysis.

[0127] (4) The dimethyl-labeled sample was acidified by adding 4.5 μL of trifluoroacetic acid (Sigma, Cat. No. T6508). The solution was desalted using a tip column (maximum volume 200 μL) packed with 2 mg of HLB resin (Waters, USA), and then lyophilized to obtain a peptide mixture.

[0128] (5) The above peptide mixture was reconstituted in 30 μL of 0.1% by volume formic acid. 1 μg of peptide was then injected and subjected to LC-MS / MS analysis using data-dependent acquisition (DDA). Each sample was analyzed twice by the mass spectrometer, and duplicate quantitative results were obtained for each peptide.

[0129] (6) The spectra obtained by the DDA method were analyzed using MaxQuant software (Cox, Germany) to identify the protein to which the peptide belonged, determine the position of the peptide on said protein, and calculate the fold change (FC) in peptide abundance between the experimental and control groups.

[0130] In this example, three representative heat shock protein inhibitors were used: geldanamycin and tanespimycin (inhibitors of the ansamycin class containing a benzoquinone group) and ganetespib (a second-generation inhibitor with a novel structure) (Figure 10A). Geldanamycin and tanespimycin differ from each other only in the circled structural region. The second most significantly altered peptide (i.e., the peptide with the second largest |log2FC| value among all quantified peptides for a given protein) represents each protein, and the quantitative results for duplicate replicates are shown in Figure 10B. By considering indicators of proteins with altered energy states with a quantitative result of |log2FC| > 1.4 for duplicate replicates, we successfully identified heat shock proteins, including HSP90AB1, HSP90AA1, HSP90A1, and HSP90AB2P, demonstrating that PELSA based on dimethyl-labeling quantification can efficiently identify proteins with altered energy states. Furthermore, it was observed that TRAP1, a mitochondrial heat shock protein, was specifically identified as a protein with altered energy status only in the ganetespib-treated group.

[0131] The HSP90 protein consists of three domains: the N-terminal ATP-binding domain, the central domain, and the C-terminal domain. Geldanamycin, tanespimycin, and ganetespib all target the N-terminal ATP-binding domain of HSP90. As shown in Figure 10C, taking geldanamycin as an example, analysis of the fold change in peptide abundance across different domains revealed that only peptides from the N-terminal domain (drug-binding region) showed significant changes in abundance. This finding further supports the ability of PELSA to accurately identify protein regions with altered energy states.

[0132] In addition to known target proteins, PELSA also identified several previously unknown off-target proteins. For geldanamycin, PELSA identified the destabilization of several PRDX family proteins with log2FC > 1.4 (Figure 10B). It has been reported in the literature that geldanamycin can induce the generation of reactive oxygen species (ROS), resulting in severe liver toxicity (Clark et al., Free Radical Biology & Medicine, 2009, 1440-1449). PRDX family proteins play a protective role in cells by scavenging ROS under oxidative stress conditions. The destabilization of PRDX family proteins suggests that their structure is disrupted at high concentrations of geldanamycin, which provides a possible explanation for the mechanism of geldanamycin hepatotoxicity. For tanespimycin, a molecule structurally similar to geldanamycin, PELSA identified PRDX5 and NQO1 as common off-target proteins with geldanamycin. NQO1 has been reported to be involved in the metabolism of ansamycin-class heat shock protein inhibitors [Reigan PD, et al., Molecular Pharmacology, 2011, 79(5): 823-832]. Additionally, PELSA identified two previously unreported off-target proteins of ganetespib, namely, MAT2A and AKR1C2 (Figure 10B). To further verify the binding between these two proteins and ganetespib, both proteins were purified and thermal shift assays were performed to evaluate their interaction with ganetespib. As shown in Figure 10D, the thermal melting temperatures of both MAT2A and AKR1C2 significantly increased upon the addition of ganetespib, confirming the binding of ganetespib to AKR1C2 and MAT2A. Furthermore, the thermal shift assay results demonstrate that AKR1C2 binds to geldanamycin, despite the lower stabilizing effect of geldanamycin compared to ganetespib (Figure 10D).Consistently, the PELSA results also showed a smaller stabilizing effect of geldanamycin on AKR1C2 compared to ganetespib (FIG. 10B).

[0133] These findings clearly demonstrate that PELSA, combined with dimethyl labeling, can accurately determine proteins with altered energy states and efficiently distinguish target proteins among structurally similar distinct inhibitors. In vitro thermal shift assays using purified proteins verified the high reliability of off-target proteins identified by PELSA. The identification of these target proteins provides valuable insights into the hepatotoxicity of ansamycin HSP90 inhibitors and the exploration of novel applications of ganetespib. [Example 12]

[0134] PELSA assessment of local affinity between heat shock protein inhibitors and their target proteins.

[0135] The procedures and conditions were the same as those in Example 11, with the following modifications: 14 aliquots each containing 50 μL of HeLa cell lysate were divided into an experimental group and a control group, with 7 aliquots in each group. In the experimental group, different concentrations of geldanamycin were added to the lysate to achieve final concentrations of 100 μM, 10 μM, 1 μM, 100 nM, 10 nM, 1 nM, or 0.1 nM (geldanamycin dissolved in DMSO), while samples in the control group received an equal volume of DMSO. The subsequent experimental steps, mass spectrometry quantification, and software analysis procedures were the same as those in Example 11. The processing procedures for tanespimycin and ganetespib followed the same protocol as for geldanamycin.

[0136] Figure 11A shows, using ganetespib as an example, that only peptides within the N-terminal domain of HSP90 protein exhibited increasing changes in peptide abundance with increasing drug concentration, eventually reaching a plateau. The affinity between heat shock proteins and three HSP90 protein inhibitors was then calculated based on the fold changes of these peptides with various drug concentrations. Peptides of heat shock proteins with at least 12 quantification values ​​(14 quantification values ​​in total) were fitted to a four-parameter logarithmic equation using Prism software (GraphPad): Y = bottom + (top - bottom) / [1 + 10^(LogEC50 - X) × Hill slope)]. The fitting quality between the raw data and this equation was assessed using the Pearson correlation coefficient (R2), with a high fitting quality (R2) being considered. 2 Peptides with a K > 0.9 were selected as candidate peptides. In addition, the magnitude of the abundance change of the candidate peptide at the highest ligand concentration must be at least 30% or more. The median fold change of all quantified peptides from the same protein was considered as the fold change of the protein at a given ligand concentration, corresponding to the Y value in the four-parameter logarithmic equation. X still represents the ligand concentration. The four-parameter logarithmic equation was refitted using Prism software to determine the EC50 (Figure 11B). As shown in Figure 11C, the protein was purified and the affinity (K) between the three inhibitors and HSP90AA1 was measured. d ) was measured using microscale thermophoresis (MST). The affinity values ​​K obtained by MST were d is in close agreement with the EC50 calculated using PELSA.

[0137] These findings demonstrate that affinity values ​​determined by PELSA are consistent with conventional methods for determining affinity, such as microscale thermophoresis (MST), which was used in this example. Furthermore, this approach allows for the acquisition of affinity data at the peptide level, which will contribute to our understanding of the interaction between ligands and proteins.

Claims

1. 1. A method for detecting proteins having an altered energy state, comprising: A. Mixing and incubating two or more groups of protein solutions, each representing a different energy state (one group serving as a control, being a protein solution in its original energy state, and the other group serving as an experimental group, being a protein solution with an altered energy state), with a protease at a weight ratio of protease to total protein ranging from 1:1 to 1:50 for 0.5 to 60 minutes to allow digestion of low-energy protein structures (i.e., digestion of proteins in a destructive manner), resulting in the direct generation of small peptides (with molecular weights <5 kDa) suitable for bottom-up (protein identification by peptide) mass spectrometry analysis containing two cleavage sites reflecting the local stability of the protein; B. Isolating small peptides (having a molecular weight of <5-10 kDa) from larger protein fragments (having a molecular weight of ≥ 10 kDa); C. Determining the abundance of small peptides (having a molecular weight of <5 kDa), the protein to which they belong, or their location on said protein by quantitative proteomic techniques and software analysis (i.e., mass spectrometric quantification and analysis); D. Determining proteins with altered energy states by analyzing differences in abundance of small peptides (having a molecular weight of <5 kDa) between experimental and control samples; and E. Determining protein regions with altered energy states by analyzing the locations of small peptides (with molecular weights <5 kDa) with altered abundance in the protein to which they belong. A method comprising:

2. The different energy states of proteins are: A. A change in the energy state of a protein resulting from the interaction of the protein with a ligand (including one or more of a drug molecule, a human or animal metabolite, a plant extract, a nucleic acid, a metal ion, a peptide, an antibody, and a protein); B. A change in the energy state of a protein resulting from an altered post-translational modification of the protein, or C. Changes in the energy state of a protein resulting from one or more of the following: thermal stress, osmotic changes, denaturant stress, oxidative stress, or disease.

10. The method of claim 1, wherein the energy state change is caused by one or more of the following:

3. 1. A method for detecting affinity between a ligand and a protein, comprising: A. Mixing and incubating a protein solution with a ligand, each at a series (or multiple) of different final concentrations (one of the concentrations can serve as a control group as a blank control without ligand, and the others as experimental groups); B. After incubation, subject the sample to digestion under non-denaturing conditions with a protease to protein mass ratio of 1:1 to 1:50 for 0.5 to 60 minutes to ensure that low-energy protein structures are disrupted by enzymatic cleavage (i.e., induce destructive enzymatic cleavage of the protein) and directly generate a large number of small peptides (with molecular weights of <5 kDa) suitable for bottom-up mass spectrometric analysis containing two cleavage sites that can reflect the local stability of the protein; C. Isolating small peptides suitable for mass spectrometry (having a molecular weight of <5-10 kDa) from large protein fragments (having a molecular weight of ≥ 10 kDa); D. Determining the abundance of small peptides (having a molecular weight of <5 kDa), the protein to which they belong, or the location of the small peptides on said protein by quantitative proteomic techniques and software analysis (i.e., mass spectrometry quantification and analysis); and E. Calculating the local affinity between the protein and the ligand, expressed as the half-maximal effective concentration (EC50), according to the abundance change of the small peptide in the sample treated with different ligand concentrations, where the calculation is performed based on a four-parameter logarithmic equation: Y = bottom + (top - bottom) / [1 + 10^(LogEC50 - X) x Hill slope], where Y (vertical axis) represents the fold change in peptide abundance between the experimental group and the control group, X (horizontal axis) represents the different ligand concentrations, bottom and top represent the lower and upper plateau values ​​of the four-parameter logarithmic equation curve, respectively (same units as Y), Hill slope represents the absolute value of the maximum slope of the curve at the midpoint, and EC50 is the calculated half-maximal effective concentration; a smaller EC50 indicates a stronger affinity between the ligand and the protein region to which the peptide belongs. and The method wherein the "series of different final concentrations" represents at least three or more final concentrations, preferably five or more final concentrations.

4. 4. The method of claim 3, wherein the ligand may be one or more of a drug molecule, a metabolite in an animal or plant, a plant extract, a nucleic acid, a metal ion, a peptide, an antibody, and a protein, or any other substance capable of interacting with a protein.

5. 4. The method of claim 1 or 3, wherein the protein solution comprises a single protein solution containing one protein or a mixed protein solution containing two or more proteins; the mixed protein solution comprises a cell or tissue extract derived from a human, an animal, a plant, or a bacterium; a mild extraction method is used for protein extraction, including, but not limited to, liquid nitrogen freeze-thaw extraction, liquid nitrogen crushing extraction, or homogenization extraction; when an extraction buffer is used during the extraction process, detergents or denaturants at concentrations that disrupt the conformation of the protein are avoided to ensure that the protein retains the conformation to be tested; and the protein to be tested may exist in the form of a free protein and / or an immobilized protein.

6. The method comprises subjecting solutions of proteins of different energy states in the conformations to be tested to digestion with proteases added according to a mass ratio of protease to protein under the same digestion conditions, such that the high-energy state protein structure is exposed to the protease, destroyed, and unfolded by digestion, thereby inducing exposure of the cleavage sites of the low-energy state of the protein; and in the presence of a relatively large amount of protease, the low-energy state structure is destroyed by digestion, directly generating small peptides (with molecular weights of <5 kDa) suitable for identification by bottom-up mass spectrometry (using peptides to identify the protein), resulting in various degrees of digestion of the proteins of different energy states; 2. The method of claim 1, wherein the optional protease comprises one or more selected from trypsin, glutamyl endopeptidase, thermolysin, proteinase K, or other site-specific or non-site-specific proteases, the mass ratio of protease to protein is 1:1 to 1:50, preferably 1:1 to 1:10, in the presence of 1% to 2% by volume of a protease inhibitor, the digestion time is within the range of 0.5 to 60 minutes, and the digestion temperature is determined based on the optimum temperature required to maintain the activity of the protease.

7. The method includes subjecting a ligand-treated group treated with various ligand concentrations and a control group treated with a blank solvent to digestion with protease added according to a mass ratio of protease to protein, such that the high-energy state protein structure is exposed to the protease, destroyed, and unfolded by digestion, thereby inducing exposure of a low-energy state cleavage site of the protein; in the presence of a relatively large amount of protease, the low-energy state structure is destroyed by digestion, directly producing small peptides (having a molecular weight of <5 kDa) suitable for identification by bottom-up mass spectrometry (using peptides to identify proteins), resulting in a different degree of digestion of the ligand-bound protein from that of the unligand-bound protein; 4. The method of claim 3, wherein the optional protease comprises one or more selected from trypsin, glutamyl endopeptidase, thermolysin, proteinase K, or other site-specific or non-site-specific proteases, the mass ratio of protease to protein is 1:1 to 1:50, preferably 1:1 to 1:10, in the presence of 1% to 2% by volume of a protease inhibitor, the digestion time is within the range of 0.5 to 60 minutes, and the digestion temperature is determined based on the optimum temperature required to maintain the activity of the protease.

8. 4. The method of claim 1 or 3, wherein the separation of small peptides from larger protein fragments includes, but is not limited to, utilizing differences in one or more of molecular weight, hydrophobicity, or thermal stability.

9. Quantitative proteomics techniques refer to quantitative analysis of cleaved peptides in a bottom-up mass spectrometry approach (using peptides to identify proteins); mass spectrometry data acquisition includes data-dependent acquisition (DDA) and data-independent acquisition (DIA); quantification methods include label-free quantification and / or label-based quantification (e.g., dimethyl-label quantification, iTRAQ-label quantification, TMT-label quantification, TMTpro-label quantification, or a combination thereof); The method of claim 1 or 3, wherein software analysis refers to importing raw spectral files generated by mass spectrometry into database search software to obtain quantitative information of small peptides, information about the protein to which the small peptides belong, and information about the position of the small peptides on the protein; and the software includes Maxquant (Cox, Germany), MS Fragger (Alexay, USA), Spectronaut (Biognosys, Switzerland), or a combination thereof.

10. The method for determining proteins with altered energy state is based on the following factor: -log of peptide １０ The significance of the difference between the differential peptides represented by the P value, the |log ２ based on one or any combination of the magnitude of the fold change of the peptide, represented by fold change |, the number of differential peptides (N), whether the fold change of the peptide exhibits a dose-dependent change with respect to the ligand concentration, or a combination thereof; Determining the significance of the difference in differential peptides involves performing a significance test, such as an empirical Bayes t-test or a Student's t-test, on the quantification values ​​of the peptides in the experimental and control groups, and the resulting P-value is used as an indicator of the significance of the difference in peptide abundance between the control and experimental groups; -log １０ Proteins containing peptides with P values ​​>2-5 were considered to represent proteins with altered energy state changes; Determining the fold change of differential peptides includes determining it directly from database search software or via data validation methods such as empirical Bayes t-test or Student's t-test; ２ Proteins containing peptides with fold change |>0.3-3 are considered as proteins with altered energy state; The process for counting the number of differential peptides involves ranking all peptides within each protein based on the significance of their difference or fold change; the peptide with the highest significance of difference or fold change is defined as the top-ranked altered peptide for that protein; the second-ranked peptide represents the peptide with the second-highest significance of difference or fold change, followed by the third- and fourth-ranked peptides; and the −log １０ P-value and log ２ The fold change is defined as the quantitative value of the protein; １０ P value > 2 to 5 or |log ２ Proteins with a fold change |>0.3 to 3 are defined as having a change in energy state; proteins with a change in energy state specified in these ways are defined as having a change in energy state -log １０ P value > 2 to 5 or |log ２ containing one, two, three, or four peptides that meet the criteria of fold change |>0.3-3; Determining whether the fold change in peptide abundance exhibits a dose-dependent response to ligand concentration involves fitting the fold change of peptide at different ligand concentrations and the ligand concentration as variables to a four-parameter logarithmic equation, and the quality of the fit between this equation and the raw data is measured using the Pearson correlation coefficient (R ２ ) and evaluated using R ２ If the ratio is >0.75-0.95, the peptide is considered to exhibit a dose-dependent response to the ligand concentration, and the protein to which the peptide belongs is defined as the target protein of the ligand; Any combination of factors may be calculated using the -log α to generate a global score that determines proteins that satisfy any two or more of the aforementioned factors or have a change in energy state. １０ P value, |log ２ Fold change|, N (number of differential peptides), and R ２ 4. The method of claim 1 or 3, comprising combining in any proportion:

Images