Polar aprotic organic solvent for dissociating streptavidin-biotin conjugate as well as screening method and application of polar aprotic organic solvent

Abstract

The invention relates to a polar aprotic organic solvent for dissociating a streptavidin-biotin conjugate as well as a screening method and application of the polar aprotic organic solvent. Aiming at the problems that a streptavidin-biotin conjugate system is difficult to elute quickly under a mild condition and a carrier is difficult to regenerate, a solvent molecule descriptor database is constructed; a polar aprotic organic solvent which meets specific molar refractive index and polarity parameters and is a non-hydrogen bond donor is screened from a large number of candidate solvents by utilizing a machine learning model. The solvent can realize rapid reversible dissociation of biotin conjugates from streptavidin at room temperature, and the carrier still keeps high activity after multiple rounds of binding-elution cycles. The solvent is further used for purification of biotinylated small molecules in a crude reaction solution, enrichment and elution of mitochondria and cell surface samples and the like, topological resolution proteomics analysis is achieved, a high-purity product can be obtained in a high-flux mode, and site level topological analysis can be completed.

Description

Technical Field

[0001] This invention belongs to the fields of biochemistry, chemical biology, and bioseparation and analysis technology, and relates to a polar aprotic organic solvent for dissociating streptavidin-biotin conjugates, its screening method, and its application. This invention also involves organic solvent property modeling, molecular descriptors, and machine learning screening methods, utilizing polar aprotic organic solvents to achieve ultra-fast, reversible dissociation of the streptavidin-biotin conjugate system and multiple rounds of carrier regeneration; and combining proximity labeling and cell surface labeling strategies to achieve topological resolution analysis of mitochondrial proteome and cell surface proteome. Background Technology

[0002] The streptavidin / avidin-biotin (SAB) system exhibits extremely high affinity, with its equilibrium dissociation constant typically around 10. -14 ~10 -15 The process is on the order of M, almost to the point of being "irreversible". Due to its high specificity and thermodynamic stability, this system has long been widely used in various biological and chemical biological operations such as enrichment, immobilization, immunoassay, imaging labeling and proximity labeling of proteins, nucleic acids and small molecule probes (Weber PC et al., Science, 1989, 243: 85–88; Dundas CM et al., Appl Microbiol Biotechnol, 2013, 97(21): 9343–9353).

[0003] In the field of proteomics, biotinylated proteins or peptides are typically enriched using streptavidin resin or magnetic beads, relying on biotinylated proximity markers (such as APEX, BioID, etc.) and surface analysis methods. In the fields of chemical biology and medicinal chemistry, biotinylated small molecule probes can achieve target protein enrichment, target identification, or small molecule purification after "click chemistry" through streptavidin solid phase (Roux KJ et al., Curr Protoc Protein Sci, 2018,91: 19.23.1–19.23.15; Rhee HW et al., Science, 2013, 339(6125): 1328–1331; Branon TC et al., Nat Biotechnol, 2018, 36(8): 880–887).

[0004] However, the ultra-high affinity of the SAB system also presents significant challenges: effective elution is almost impossible under mild conditions. Traditional methods for eluting biotin or biotinylated molecules from streptavidin mainly involve boiling at high temperatures in a buffer containing sodium dodecyl sulfate (SDS) and excess free biotin, or treating under extreme conditions such as high-temperature water, high pH, ​​or low pH. These methods often lead to irreversible conformational changes or even denaturation of the streptavidin protein, and are also highly destructive to the captured molecules, especially proteins and large molecular complexes. Therefore, in many applications, streptavidin carriers are only for single use, increasing experimental costs and limiting high-throughput applications (Holmberg A et al., Electrophoresis, 2005, 26(3): 501–510; Cheah JS et al., Biochem Biophys ResCommun, 2017, 493(4): 1522–1527).

[0005] To improve reversibility, existing technologies have proposed monomeric streptavidin, engineered low-affinity streptavidin variants, or the use of reversible ligands such as desthiobiotin to replace biotin. While these strategies improve the controllability of dissociation to some extent, they often come at the cost of reduced affinity and specificity, resulting in decreased enrichment efficiency and washability. It is difficult to simultaneously meet the requirements of "high affinity capture" and "mild and reversible elution" (Howarth M et al., NatMethods, 2006, 3(4): 267–273; Hirsch JD et al., Analytical Biochemistry, 2002, 308(2): 343–357; Dundas CM et al., Appl Microbiol Biotechnol, 2013, 97(21): 9343–9353).

[0006] On the other hand, in biotin-based proteomics and paramics analysis, the elution of streptavidin enrichment in existing methods often relies on the aforementioned strong denaturation or high temperature conditions, requiring complex buffer replacement and desalting steps downstream, which is not very user-friendly for automation and high-throughput analysis. Some methods use anti-biotin antibodies for enrichment and elution under mild conditions, but the antibody costs are high, regeneration is difficult, and the enrichment depth and reproducibility are limited (Udeshi ND et al., Nat Methods, 2017, 14(12): 1167–1170; Zhu Y et al., Biochem Biophys Rep, 2024, 38: 101711).

[0007] In the field of chemical biology, biotinylated small molecules are often prepared through organic synthesis or click chemistry. After the reaction, the crude reaction solution often contains complex components such as unreacted substrates, byproducts, catalysts, and auxiliaries, while the target biotinylated product usually constitutes only a small portion. Traditional small molecule purification methods mostly rely on column chromatography and preparative high-performance liquid chromatography, which are cumbersome and time-consuming, making them unsuitable for high-throughput screening in micro-volume, parallel, and multi-reaction systems. While using streptavidin for solid-phase purification of small molecules and biotin has theoretical advantages, it is limited by irreversible elution and the non-renewable carrier, preventing the development of a mature and widely applicable solution (Cravatt BF et al., Annu RevBiochem, 2008, 77: 383–414).

[0008] Photocrosslinking probes are a powerful tool in chemical biology and chemical proteomics research. By chemically modifying small molecules, they can be endowed with the ability to covalently crosslink. Photocrosslinking technology is often used to study the non-covalent interactions between small molecule drugs and targets. Through proper design, effectively synthesized photocrosslinking probes can be directly covalently linked to their interacting proteins, which is of great significance for identifying the targets and interaction sites of small molecule drugs.

[0009] Traditional target identification methods (such as conventional ABPP or immunoprecipitation) typically only tell researchers "which protein the drug binds to" (e.g., drug X binds to protein A), which is known as protein-level identification. This invention, however, utilizes FlashBio (Fast, low stringency, ambient, solvent-helped biotin conjugate release) for gentle elution, enabling complete recovery of the probe-bearing peptide, thus achieving peptide-level and even amino acid-level identification (Residue-level Resolution). A key drawback of conventional methods is that in conventional processes, biotinylated peptides often bind too tightly (Kd≈10). -15 However, the drug cannot be eluted from streptavidin magnetic beads, or it must be subjected to vigorous boiling, resulting in significant streptavidin contamination and masking the low-abundance target signal. This means that researchers can usually only detect "non-crosslinked peptides" to infer the protein identity, but lose the most crucial "binding site" information. By combining FlashBio technology, this invention can release the "drug-probe-peptide" complex completely without introducing streptavidin contamination. Mass spectrometry can not only identify the protein but also pinpoint which amino acid residue reacted by detecting specific mass shifts caused by probe modification.

[0010] In previous studies, some polar organic solvents have been observed to affect streptavidin-biotin binding to some extent. However, these observations are mostly sporadic and lack systematic solvent screening and mechanism analysis. Quantitative criteria for solvent properties have not yet been established, nor has a complete reversible affinity purification system been constructed around solvent selection.

[0011] At the methodological and algorithmic level, cheminformatics and machine learning have been used to predict solvent properties, reaction yields and other problems. However, there is still a lack of machine learning models that combine molecular descriptors, solvent polarity parameters and elution performance of bioaffinity systems to guide solvent screening and then use them to discover “rapid and reversible elution solvents” suitable for streptavidin-biotin systems (Götz J et al., Sci Adv, 2023, 9(43): eadj2314).

[0012] In summary, the existing technologies have the following shortcomings: (1) There is a lack of a universal method that can achieve ultra-fast and reversible dissociation of the streptavidin-biotin system under mild conditions, making it difficult to balance high affinity capture and efficient reversible elution; (2) The regeneration and multi-round use capabilities of streptavidin carriers are limited, making it difficult to meet the needs of high-throughput, small-volume and automated application scenarios; (3) The purification of biotinylated small molecules still mainly relies on traditional chromatography techniques, which are complex to operate and have limited throughput; (4) In organelles (such as mitochondria) and cell surface proteomics, the topological resolution capability is limited, making it difficult to systematically resolve the topological structure and dynamic remodeling of transmembrane proteins at the site level.

[0013] Therefore, it is necessary to provide a new technical solution that can achieve gated rapid and reversible dissociation of streptavidin-biotin conjugate systems, combined with high-throughput experimental procedures, and play a role in multiple application scenarios such as small molecule purification and topologically resolved proteomics. Summary of the Invention

[0014] To address the aforementioned technical problems, this invention provides a polar aprotic organic solvent for the dissociation of streptavidin-biotin conjugates, along with its screening method and applications. This invention addresses the difficulties in rapidly eluting the streptavidin-biotin conjugate system under mild conditions and the challenges in carrier regeneration. It constructs a solvent molecule descriptor database and utilizes a machine learning model to screen a class of polar aprotic organic solvents from a large pool of candidate solvents that meet specific molar refractive indices, polar parameters, and are non-hydrogen bond donors. These solvents enable rapid and reversible dissociation of biotin conjugates from streptavidin at room temperature, and the carrier retains high activity even after multiple binding-elution cycles. Furthermore, this invention applies these solvents to the purification of biotinylated small molecules in crude reaction solutions and the enrichment and elution of mitochondrial and cell surface samples, achieving topologically resolved proteomics analysis. This allows for the acquisition of high-purity products and site-level topological resolution in high-throughput formats.

[0015] The objective of this invention can be achieved through the following methods:

[0016] In a first aspect, the present invention provides the application of a polar aprotic organic solvent in the dissociation of streptavidin-biotin conjugates, wherein the polar aprotic organic solvent is selected from at least one of dimethyl sulfoxide, dimethylformamide, N-methyl-2-pyrrolidone, tetramethyl sulfoxide, dimethylacetamide, and dimethylpropenylurea, or a mixture thereof with water. The polar aprotic organic solvent of the present invention is a solvent with a high dielectric constant and electronic polarization capability.

[0017] In one embodiment of the present invention, the volume fraction of the polar aprotic organic solvent in the mixture is 80% or more, preferably 90% or more.

[0018] As one embodiment of the present invention, the method for dissociating the streptavidin-biotin conjugate includes the following steps: S1. Wash the streptavidin-biotin conjugate sample with an aqueous washing solution to remove non-specifically bound components; S2. Then treat the streptavidin-biotin conjugate sample with a polar aprotic organic solvent; S3. Remove the polar aprotic organic solvent and equilibrate the streptavidin-biotin conjugate sample with aqueous buffer. Repeat steps S1 and S2.

[0019] The present invention can also add a pre-washing step with a volume fraction of about 10% to 30% of polar aprotic organic solvent before or after the aqueous washing step to remove hydrophobic non-specific adsorbed components, thereby further improving the purity of the eluted sample.

[0020] Furthermore, the linker in the biotin conjugate is selected from at least one of small molecule compounds, peptides, or oligonucleotides. The biotin conjugate exhibits higher solubility in the polar aprotic organic solvent than in the aqueous buffer solution, and the elution efficiency is positively correlated with the solubility of the linker moiety in the organic solvent.

[0021] This invention uses polar aprotic organic solvents to dissociate biotin conjugates from streptavidin into the solvent phase, and the binding efficiency and dissociation efficiency of the biotin conjugates in each round are maintained above 80%.

[0022] Further, in step S1, the aqueous washing solution includes PBS; In step S3, the aqueous buffer solution includes PBS; the number of repetitions is 10-30 times.

[0023] Furthermore, in step S2, the processing time is 1-5 minutes.

[0024] Secondly, the present invention provides a method for screening polar aprotic organic solvents in the aforementioned applications, comprising the following steps: A1. Select all organic solvents from the public database as candidate solvents and calculate molecular descriptors for the candidate solvents; A2. Select a portion of the solvent as the training set and determine its elution efficiency for biotin-Cy5 dissociation from streptavidin. A3. Using a machine learning model, the molecular descriptors from step A1 are used as input features, and elution efficiency is used as output. Cross-validation is used to evaluate the model performance and prevent overfitting. By analyzing the feature importance of the machine learning model, molecular descriptors that contribute significantly to the prediction of elution performance are selected. The molecular descriptors that contribute significantly to the prediction of elution performance include molar refractive index (AMR) and solvent polarity parameter (E). T (30); A4. Based on molecular descriptors and classical solvent parameters that significantly contribute to the prediction of elution performance, solvents with a molar refractive index (AMR) greater than 20 and solvent polarity parameter E are selected from the candidate solvents. T (30) Solvents with a value greater than 42 and no hydrogen bond donor capability are considered as the first candidate set; A5. Based on practical constraints, solvents that are liquid at room temperature, miscible with water, and chemically stable are selected from the first candidate set as polar aprotic organic solvents.

[0025] In one embodiment of the present invention, in step A1, the candidate solvent includes polar solvents, nonpolar solvents, protic solvents, and aprotic solvents. In some embodiments, the candidate solvent is composed of 431 organic solvents.

[0026] As one embodiment of the present invention, in step A1, the calculation software includes the cheminformatics software PaDELPy.

[0027] In one embodiment of the present invention, in step A3, the machine learning model includes one or more of the following: random forest model, gradient boosting tree model, and support vector machine model. Preferably, the molar refractive index (AMR) and solvent polarity parameter E are discovered. T (30) It has a positive correlation with elution performance, and the effective solvent should be a non-hydrogen bond donor solvent. Therefore, candidate polar aprotic organic solvents should meet at least: AMR>20, E T (30)>42 and does not have hydrogen bond donor capability.

[0028] Regarding solvent selection, the parameter thresholds in the screening rules can be appropriately adjusted based on the physicochemical properties of different linkers.

[0029] Thirdly, this invention provides the application of a polar aprotic organic solvent in purifying biotinylated small molecule compounds in crude reaction solutions, analyzing topologically resolved mitochondrial proteomics, analyzing topologically resolved cell surface proteomics, identifying small molecule drug targets and binding sites, and purifying biotinylated single-stranded DNA.

[0030] As one embodiment of the present invention, the method for purifying biotinylated small molecule compounds in crude reaction solution includes the following steps: B1. Add the crude reaction solution containing biotinylated small molecule compounds to the solid packing material filled with streptavidin; B2. Wash the solid packing with a washing solution to remove non-specifically bound impurities; B3. The solid-phase packing material is then eluted with a polar aprotic organic solvent to obtain the eluent of the biotinylated small molecule compound.

[0031] Further, in step B1, the crude reaction solution is a reaction solution containing biotinyne and an azide-containing substrate undergoing an azide-yne ​​cycloaddition reaction under Cu(I) catalysis. In some embodiments, the volume of the crude reaction solution is 10-600 μL, and the bed volume of the solid-phase packing is approximately 100 μL; the azide-containing substrate is derived from chemical synthesis.

[0032] Furthermore, in step B2, the washing solution includes urea.

[0033] As one embodiment of the present invention, the method for analyzing topologically resolved mitochondrial proteomics includes the following steps: C1. Express a peroxidase-labeled enzyme that targets the mitochondrial matrix in the cells to be tested, thereby localizing the labeled enzyme to the mitochondrial matrix; C2. Add biotin and hydrogen peroxide to the cells to activate the labeling reaction, triggering the labeling enzyme to catalyze the generation of active free radicals and biotinylate mitochondrial adjacent proteins in the cell survival state. C3. The labeled cells were lysed and digested with proteases to obtain a sample containing biotinylated peptides; C4. Contact the sample with streptavidin to enrich the biotinylated peptides, and then wash (strictly) with an aqueous washing solution. C5. The biotinylated peptides are then eluted with a polar aprotic organic solvent. C6. Mass spectrometry analysis was performed on the eluted peptides, and the transmembrane topology of mitochondrial proteins was assigned or modified based on biotinylation site information.

[0034] As one embodiment of the present invention, in step C6, the topological structure of mitochondrial proteins, especially mitochondrial inner membrane transmembrane proteins, is inferred based on the position of the biotinylation site in the protein sequence and in known or predicted transmembrane region annotations, and matrix-side coverage and topological verification can be achieved for the subunits of introduced complexes (such as TOM, TIM, PAM, etc.).

[0035] The number of biotinylated peptides and mitochondrial proteins obtained by the method of the present invention is higher than that of the comparative method based on enrichment and elution of anti-biotin antibodies.

[0036] As one embodiment of the present invention, the method for analyzing topologically resolved cell surface proteomics includes the following steps: D1. Add a membrane-impermeable biotinylated lysine reaction reagent to the surface of living cells to biotinylate lysine residues exposed on the outer side of the cell. D2. The cells are lysed and digested with proteases to obtain a sample containing biotinylated peptides. D3. Contact the sample with streptavidin to enrich the biotinylated peptides, and then wash it (strictly) with an aqueous washing solution. D4. The biotinylated peptides are then eluted with a polar aprotic organic solvent. D5. Perform mass spectrometry analysis on the eluted peptides and map the obtained biotinylated sites to the corresponding protein amino acid sequence and its database topological annotation or transmembrane prediction results. D6. Based on whether the biotinylated sites fall within the labeled or predicted extracellular regions, transmembrane proteins are divided into a fully matched category (all sites are located outside the cell), a partially matched category (some sites are located outside the cell), an unmatched category (no extracellular sites), and a blank category (cannot be determined due to missing topological information), thus obtaining a topologically resolved cell surface proteome map.

[0037] As one embodiment of the present invention, in step D1, the biotinylated lysine reaction reagent includes N-hydroxysuccinimide ester biotinylation reagent.

[0038] As one embodiment of the present invention, in step D1, the living cells perform steps D1 to D6 under different stimulation conditions, and compare the distribution and changes of the transmembrane proteins in each topological category under each condition to characterize the abundance and topological remodeling of cell surface proteins during stimulation.

[0039] In proteomics applications, other organelles or membrane structures besides mitochondria and the cell surface can be analyzed, as long as spatially specific biotinylated patterns can be generated through appropriate localization tags or labeling strategies, they can be used in conjunction with the reversible affinity system of this invention. In a further embodiment, the above steps can be performed on cells under different treatment conditions (e.g., before and after pharmacological stimulation or differentiation induction), and the changes in the abundance of surface proteins and the conversion between different topological classes under each condition can be compared, thereby characterizing the remodeling process of cell surface proteins in biological events at both the site and topological levels. For example, during the differentiation of THP-1 monocytes into macrophage-like cells, some transmembrane proteins can be observed to change from Full class to Partial or No class, suggesting rearrangement in the exposed region.

[0040] In terms of automation systems, affinity columns can be replaced with magnetic bead systems to achieve liquid replacement through magnetic separation; or small-volume streptavidin solid and solvent channels can be integrated in the form of microfluidic chips to achieve lower sample volume and higher integration.

[0041] As one embodiment of the present invention, the method for identifying small molecule drug targets and binding sites includes the following steps: E1. Probe assembly: Small molecule drugs are coupled with trifunctional probes through click chemistry reactions to obtain conjugates; E2. Photocrosslinking: The conjugate is incubated with biological samples and then exposed to light to covalently crosslink the probe with the target protein in the biological sample; E3. Enzymatic hydrolysis: The biological sample undergoes protein denaturation, reduction and alkylation, followed by enzymatic hydrolysis into a mixture of peptides; E4. Affinity enrichment: The peptide mixture is contacted with a solid support bonded with streptavidin to capture biotinylated cross-linked peptides. E5. Dissociation: The solid support is exposed to the Flash reagent to disrupt the non-covalent interaction between the biotin conjugate and streptavidin, and the cross-linked peptides are eluted. E6. Mass spectrometry analysis: Detects eluted peptides and identifies drug binding sites by recognizing amino acid sites carrying intact probes; screens target peptides by detecting the molecular weight of intact trifunctional probes attached to the peptide precursor ion.

[0042] As one embodiment of the present invention, in step E1, the three-functional probe has the following general structure: B—L—X; B is selected from biotin or its analogues; X is a multifunctional core structure covalently linked with: (i) an alkynyl or azide group for coupling with drug molecules; and (ii) a photoactivating group for forming a covalent bond with the target protein; L is a chemically inert linker that connects B and X.

[0043] The trifunctional probe does not contain chemical bonds (such as disulfide bonds, hydrazone bonds, and acetal bonds) that are easily broken under biological and physiological conditions or conventional proteomics enrichment conditions, thus ensuring that the probe maintains its structural integrity throughout the entire process from cell incubation to mass spectrometry analysis.

[0044] As one embodiment of the present invention, in step E5, the elution conditions are sufficient to release the biotinylated peptide, but not sufficient to cause streptavidin to detach or dissolve in large quantities from the solid support.

[0045] In one embodiment of the present invention, step E6 confirms the peptide identity by detecting the characteristic mass shift caused by the intact probe-ligand complex. The mass shift value is equal to the sum of the molecular weights of the probe and ligand minus the mass of the leaving group (e.g., nitrogen). This method is used to distinguish between the "functional binding pockets" (e.g., allosteric or orthogenic sites) of small molecule ligands on target proteins and their binding to non-functional surfaces.

[0046] As one embodiment of the present invention, the method for purifying biotinylated single-stranded DNA includes the following steps: F1. Mix streptavidin with biotinylated single-stranded DNA, vortex and incubate, then centrifuge and collect the supernatant. F2. Wash the supernatant with an aqueous washing solution; F3. Elute the supernatant with a polar aprotic organic solvent to obtain the target ssDNA.

[0047] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention introduces the concept of polar aprotic organic solvent gating into the streptavidin-biotin conjugate system. By using DMSO or other screened polar aprotic organic solvents with a volume fraction of ≥80%, preferably ≥90%, the biotin conjugate dissociation is achieved at room temperature, within minutes, or even nearly instantaneously, transforming the traditional "almost irreversible" high-affinity system into a programmable bidirectional switch.

[0048] 2. The method of the present invention achieves rapid and reversible dissociation of biotin conjugates under mild conditions, avoiding the use of high temperature, SDS and other strong denaturing conditions, and has little impact on the structure and activity of streptavidin protein. This allows the solid support to be reused for multiple rounds (e.g., at least about 20 rounds) while the binding and elution efficiency remains at a high level, significantly reducing experimental costs and increasing throughput.

[0049] 3. This invention constructs a three-step affinity purification process for biotinylated small molecules, integrating streptavidin-specific capture, vigorous washing, and elution with polar aprotic organic solvents into a single module. This allows for the rapid acquisition of high-purity biotinylated small molecule products in micro-liter volumes and complex reaction environments. It is compatible with 96-well plates and automated platforms, providing a universal and efficient purification strategy for chemical biology and drug screening.

[0050] 4. This invention, through molecular descriptor modeling and machine learning analysis such as random forests on a large number of candidate organic solvents, extracts the "AMR>20, E" formula. T (30)>42 and lacking hydrogen bond donor capability, etc., provide a set of concise screening criteria, which provide quantitative and generalizable rules for selecting and discovering polar aprotic organic solvents suitable for streptavidin-biotin conjugate systems. Based on this, a number of new high-performance solvents have been discovered, such as tetramethylenesulfoxide (TMSO), dimethylacetamide (DMAC), and dimethyl propylene urea (DMPU), expanding the solvent space of reversible affinity systems.

[0051] 5. This invention combines the above-mentioned reversible affinity system with mitochondrial proximity labeling technology, which can significantly improve the identification depth of mitochondrial biotinylated peptides and protein numbers, perform high-density, site-level analysis of the topological structure of mitochondrial inner membrane transmembrane proteins, and has the ability to cover multiple subunits of mitochondrial protein introduction complexes, providing an effective tool for mitochondrial topological proteomics.

[0052] 6. This invention applies a reversible affinity system to cell surface proteomics, which increases the number of surface proteins and biotinylated sites identified. At the same time, it systematically organizes the topological information of transmembrane proteins through a Full / Partial / No / Blank classification strategy. Furthermore, it can analyze the dynamic changes in the abundance and topological class of surface proteins under different cell types or biological treatment conditions, which helps to screen potential drug or antibody targets with actual external epitopes.

[0053] 7. This invention uses a reversible affinity system to purify biotinylated single-stranded DNA. It has good specific purification effect and practical application potential for short biotinylated ssDNA with a chain length ≤12 nt, providing the possibility for enriching biotin-modified ssDNA using this strategy in the future.

[0054] 8. The reversible affinity system of this invention is used to identify small molecule drug targets and binding sites, and has the ability to simultaneously resolve drug targets and their binding sites in complex biological systems. Attached Figure Description

[0055] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments with reference to the accompanying drawings: Figure 1 This is a schematic diagram illustrating the principle of the reversible affinity system of the polar aprotic organic solvent-gated streptavidin-biotin conjugate in Example 1; wherein, (A) is a schematic diagram of the dissociation of biotin-Cy5 from streptavidin-coated beads by dimethyl sulfoxide (DMSO); (B) is a schematic diagram of the dependence of biotin-Cy5 dissociation from streptavidin beads under different volume fractions of DMSO; (C) is a schematic diagram of the dissociation kinetics of biotin-Cy5 in the time range of 1–5 min under 100% DMSO conditions; (D) is a schematic diagram of the cyclic reversibility of the streptavidin-biotin-Cy5 interaction; (E) is a schematic diagram of the change in the binding efficiency of biotin-Cy5 to streptavidin beads in 20 cycles; (F) is a schematic diagram of the change in the dissociation efficiency of biotin-Cy5 from streptavidin beads in the same 20 cycles. Figure 2 The diagram below illustrates the solvent screening and machine learning modeling process based on polar aprotic organic solvents in Example 3. (A) shows the flowchart for machine learning feature importance analysis of 11 tested solvents; (B) shows the diagram of the 10 molecular descriptors with the highest contribution obtained through the random forest feature importance algorithm; (C) shows the new solvent screening criteria; and (D) shows the results of solvent elution experiments on streptavidin agarose beads. Figure 3The following is a schematic diagram of the three-step affinity purification method for biotinylated small molecules in Example 2: (A) is a schematic diagram of the three-step FlashBio purification process; (B) is a schematic diagram of the synthetic route for preparing biotin–AZT by coupling zidovudine (AZT) with biotin-acetylene via Cu(I)-catalyzed azide-acetylene cycloaddition reaction; (C) is a reversed-phase HPLC chromatogram of the crude biotin–AZT reaction solution and the FlashBio elution product; (D) is a schematic diagram of the FlashBio purification process in the “19+1” mixture system; (E) is an HPLC chromatogram of 5 μM biotin–AZT alone; (F) is an HPLC chromatogram of the “19+1” mixture in which each component is 5 μM; (G) is an HPLC chromatogram of the unbound supernatant after incubation of the “19+1” mixture with streptavidin (SA) beads; (H) is a FlashBio recovery of biotin–AZT from the “19+1” mixture. HPLC chromatogram; Figure 4 This is a schematic diagram of the mitochondrial proteome topological resolution analysis method in Example 4; wherein, (A) is a schematic diagram of the mitochondrial matrix labeling and enrichment workflow; (B) is a comparison of the number of mitochondrial annotated biotinylated peptides identified by enrichment with anti-biotin antibody and enrichment with FlashBio-streptavidin in three biological replicates; (C) is a schematic diagram of the overlap relationship between APEX2-labeled mitochondrial proteins identified by the antibody method and the FlashBio method in three replicates; (D) is a schematic diagram of the label specificity in the TOM / TIM / PAM complex, a protein importer that crosses the outer and inner mitochondrial membranes; (E) is a schematic diagram of the biotinylation site mapping of several representative mitochondrial transmembrane proteins; (F) is a schematic diagram of assigning or finely refining the transmembrane topology of examples of mitochondrial inner membrane proteins such as ATP5PB and PHB2 using high-density site information obtained by APEX2–FlashBio. Figure 5(A) Schematic diagram of the topological resolution analysis method for HeLa cell surface proteome in Example 5; (B) Schematic diagram of the experimental strategy; (C) Schematic diagram of the topological analysis process for transmembrane (TM) proteins; (D) Schematic diagram of the classification of biotinylated proteins based on UniProt subcellular localization annotation and subcellular / local localization of biotinylated sites; (E) Schematic diagram of representative transmembrane TM proteins in the Full-match category; (F) Schematic diagram of representative transmembrane TM proteins in the No-match category; (G) Schematic diagram of representative transmembrane TM proteins in the Blank category; (H) Schematic diagram of several examples of transmembrane TM proteins. Figure 6 The diagram below illustrates the flowchart of the HEK293T cell surface proteome topological resolution analysis method in Example 5. (A) is a flowchart of the plasma membrane™ protein topological analysis; (B) is a diagram illustrating the classification of biotinylated proteins based on UniProt subcellular localization annotation and subcellular / local distribution of biotinylated sites; (C) is a diagram illustrating representative plasma membrane™ proteins of the Full-match category; (D) is a diagram illustrating representative plasma membrane™ proteins of the No-match category; (E) is a diagram illustrating representative plasma membrane™ proteins of the Partial-match category; (F) is a diagram illustrating representative plasma membrane™ proteins of the Blank category; and (G) is a diagram illustrating several examples of plasma membrane™ proteins. Figure 7 This is a schematic diagram showing the changes in abundance and topological class of THP-1 cell surface proteins under different biological conditions (such as before or after stimulation or differentiation) in Example 5; where (A) is a schematic diagram of the experimental strategy; (B) is a summary bar chart of the number of biotinylated surface proteins identified by the FlashBio workflow in three independent LC-MS / MS experiments; (C) is a schematic diagram comparing the PMA-regulated surface proteins identified in this invention with the regulatory proteins reported in existing THP-1 surface genome studies; (D) is a volcano diagram showing the changes in the THP-1 cell surface genome between the PMA-treated group and the control group; (E) is a schematic diagram of representative downregulated surface proteins; (F) is a schematic diagram of representative upregulated surface proteins; (G) is a schematic diagram of representative surface proteins whose overall abundance changes are not significant but whose topological features (such as the relative exposure degree of different extracellular fragments) are reshaped during PMA-induced differentiation. Figure 8 The elution efficiency of five different lengths of ssDNA in Example 6; Figure 9Example 7 illustrates the use of a trifunctional probe for target protein and binding site discovery. (A) shows the chemical structure of the trifunctional probe; (B) shows the flowchart of photocrosslinking enrichment and LC-MS / MS analysis; (C) shows the sequence-intensity map of proteins identified by LC-MS / MS in the enrichment experiment; (D) shows the sequence coverage of human CRBN; (E) shows a representative MS / MS spectrum of the CRBN peptide (NH2-GYVHETLVYK-COOH) crosslinked with the probe; and (F) maps the identified crosslinking sites onto the crystal structure of CRBN. Detailed Implementation

[0056] The present invention will now be described in detail with reference to the accompanying drawings and specific embodiments. The following examples are implemented under the premise of the technical solution of the present invention, providing detailed implementation methods and specific operating procedures, which will help those skilled in the art to further understand the present invention. It should be noted that the scope of protection of the present invention is not limited to the following embodiments; any adjustments and improvements made under the concept of the present invention are all within the scope of protection of the present invention.

[0057] Example 1: Reversible affinity method for streptavidin-biotin conjugates gated by polar aprotic organic solvents Material: Table 1

[0058] This embodiment illustrates a general method for rapid and reversible elution of biotin conjugates using polar aprotic organic solvents, such as... Figure 1 The diagram illustrates the binding of biotin conjugates to streptavidin under aqueous conditions and their dissociation from the binding site under conditions of a high volume fraction of polar aprotic organic solvent.

[0059] (1) Preparation of solid support Commercially available streptavidin agarose beads were used. The beads were washed three times with phosphate-buffered saline (PBS, pH 7.4) according to the manufacturer's instructions to remove the stock solution. The washed streptavidin beads were then resuspended in PBS, adjusting the bead volume fraction to approximately 50%.

[0060] (2) Biotin conjugate binding Using biotin-fluorescent dye (Biotin-Cy5) as a model biotin conjugate, 1 μM of biotin-Cy5 was dissolved in PBS and added to a tube containing streptavidin beads. The beads were then reacted for 10–30 minutes under gentle inversion and mixing conditions to achieve the binding of biotin-Cy5 with streptavidin.

[0061] After binding, the beads are washed 3–5 times with PBS to remove unbound biotin-Cy5. At this point, biotin-Cy5 exists primarily in solid form on the streptavidin beads. The degree of binding can be confirmed by measuring the fluorescence or absorbance of the supernatant.

[0062] (3) Aqueous prewash To remove non-specific hydrophobic components adsorbed on the surface of the beads, a pre-washing step containing a low volume fraction of polar aprotic organic solvent (20% DMSO + 80% PBS) is added after PBS washing. The washing is repeated 1 to 2 times, and then washed again with PBS to restore the aqueous phase conditions.

[0063] (4) Elution with polar aprotic organic solvents Prepare a 0%–100% (v / v) dimethyl sulfoxide (DMSO) solution (which can be diluted with water or PBS to the desired v / v). Add sufficient DMSO solution to the tube containing the streptavidin beads, ensuring the beads are completely submerged. Gently invert to mix, ensuring the solvent and beads are in full contact, for 1–5 minutes.

[0064] Figure 1 (A) is a schematic diagram of the dissociation of biotin-Cy5 from streptavidin-coated beads by dimethyl sulfoxide (DMSO), showing that the binding was essentially eliminated within 1 min under 100% DMSO conditions.

[0065] Figure 1 (B) Schematic diagram of the dependence of biotin-Cy5 dissociation from streptavidin beads under different volume fractions of DMSO. It can be seen that when the volume fraction of DMSO is below approximately 80%, a significant decrease in elution recovery is observed; when the volume fraction is between 90% and 100%, the elution recovery approaches saturation. Under these conditions, biotin-Cy5 rapidly dissociates from the streptavidin binding site and dissolves into the DMSO phase. By centrifuging to settle the beads, the supernatant DMSO solution is transferred to a new tube to obtain the eluent containing biotin-Cy5. The elution efficiency can be confirmed by measuring the fluorescence intensity of the DMSO phase or by chromatographic analysis; a typical elution recovery rate can reach approximately 90%.

[0066] Figure 1 (C) is a schematic diagram of the biotin-Cy5 dissociation kinetics within a time range of 1–5 min under 100% DMSO conditions.

[0067] (5) Carrier regeneration and recycling After elution, PBS was added to the beads, gently mixed, and repeatedly washed to remove residual DMSO, so that the beads were completely returned to the aqueous environment. Then, steps (2) to (4) were repeated for the next binding-elution cycle.

[0068] Figure 1 (D) is a schematic diagram of the reversible cycle of streptavidin-biotin-Cy5 interaction, showing the binding-release cycle repeated 20 times after elution and removal in 100% DMSO.

[0069] After approximately 20 consecutive cycles, the binding and elution efficiency of biotin-Cy5 in each cycle was measured, and the results are as follows: Figure 1 As shown in (E)-(F), the binding efficiency and elution recovery rate were both maintained at over 80%, indicating that the structure and function of streptavidin were basically maintained under the conditions described, and the whole process can be regarded as a mild and reversible process.

[0070] (6) Applications of other polar aprotic organic solvents Following the steps described above, DMSO can be replaced with dimethylformamide (DMF), N-methyl-2-pyrrolidone (NMP), or selected novel polar aprotic organic solvents such as tetramethylene sulfoxide (TMSO), dimethylacetamide (DMAC), and dimethylpropenylurea (DMPU). Similar rapid elution effects can be obtained by adjusting the solvent volume fraction (generally not less than 80%) and the reaction time. Optimal conditions for different solvents can be determined through preliminary experiments.

[0071] Example 2: Three-step affinity purification of biotinylated small molecules based on polar aprotic organic solvents Material: Table 2

[0072] This embodiment illustrates how to apply the reversible affinity system of the present invention to the rapid purification of biotinylated small molecule crude reaction solutions, such as... Figure 3 The diagram illustrates the steps of loading the crude reaction solution, strong washing with urea in an aqueous phase, and elution with a polar aprotic organic solvent, as well as a comparison with traditional chromatographic purification methods.

[0073] (1) Biotinylated small molecule crude reaction system Using small molecule drugs and their derivatives as substrates, an azide group is first introduced into a non-functional site, followed by a Click reaction with an alkyne-containing biotinylation reagent catalyzed by Cu(I) to prepare biotinylated small molecules. The total volume of the reaction system is controlled at 100–200 μL, and the solvent is an aqueous organic solvent system. After the reaction, no further processing is performed, resulting in the crude reaction solution, which contains the target biotinylated small molecule, unreacted substrate, byproducts, and catalyst residue.

[0074] Figure 3(B) is a schematic diagram of the synthetic route for preparing biotin–AZT by coupling zidovudine (AZT) with biotin-acetylene via Cu(I)-catalyzed azido-acetylene cycloaddition reaction.

[0075] (2) Sample loading and binding Take about 100 μL of streptavidin agarose beads and put them into a small column or centrifuge tube. After washing the beads three times with ddH2O, add the crude reaction solution directly to the beads and incubate them at room temperature with shaking for 10 to 20 minutes to allow the target biotinylated small molecule to fully bind with streptavidin.

[0076] (3) Urea aqueous phase washing After removing the effluent by centrifugation or negative pressure, add a washing buffer containing 6 mol / L urea to the column and wash for 3–5 column volumes to remove non-specifically adsorbed small molecules and other impurities. Then wash with ddH2O for 1–2 column volumes to remove residual urea.

[0077] (4) Elution with polar aprotic organic solvents Add 90%–100% (v / v) DMSO solution (or other screening solvent) to the column, let stand for 1–2 minutes, or pass 1–2 column volumes of solvent at a moderate flow rate and collect the eluent. After elution, the purity and recovery rate of the target biotinylated small molecule in the eluent can be detected by reversed-phase high-performance liquid chromatography.

[0078] Figure 3 (A) is a schematic diagram of the three-step FlashBio purification process, showing that the crude reaction solution is directly loaded into streptavidin agarose, strictly washed with 6 M urea, and then eluted with 100% DMSO to recover the biotin conjugated product.

[0079] Figure 3 (C) is the reversed-phase HPLC chromatogram of the crude biotin–AZT reaction solution and the FlashBio elution product. The comparison shows that the product purity increased from 54.23% to 99.16%, and the separation yield was approximately 55.35%.

[0080] The results show that, in reaction systems based on representative substrates (such as zidovudine, lenalidomide, and sitagliptin), the purity of the biotinylated target product can reach approximately 95%–99% using the above three-step method, with yields generally between 50% and 70%. Even when the target biotinylated small molecule is incorporated into a complex mixture containing approximately 19 representative drugs, the elution recovery rate of the target product can still be maintained at over 65% after processing using the same procedure.

[0081] Figure 3(D) is a schematic diagram of the FlashBio purification process in the “19+1” mixture system, in which biotin–AZT is incorporated into a mixture containing 19 representative drug / drug class small molecules.

[0082] Figure 3 (E) is the HPLC chromatogram of 5 μM biotin–AZT alone.

[0083] Figure 3 (F) is the HPLC chromatogram of the “19+1” mixture in which all components are 5 μM.

[0084] Figure 3 (G) is the HPLC chromatogram of the unbound supernatant after incubation of the “19+1” mixture with streptavidin (SA) beads.

[0085] Figure 3 (H) is an HPLC chromatogram of the recovery of biotin-AZT from the “19+1” mixture by FlashBio, showing that the product is a nearly single peak and the recovery is not less than about 65%.

[0086] (5) Perforated plates and automated formats The above process can be extended to a 96-well plate format: streptavidin beads are pre-loaded into a 96-well filter plate, and 10–200 μL of crude reaction solution is processed per well. The crude reaction solution is loaded, washed with urea, and eluted with DMSO sequentially through an automated pipetting workstation, so as to achieve rapid purification of products for parallel multiple reactions.

[0087] Example 3: Screening and Machine Learning Modeling of Polar Aprotic Organic Solvents This embodiment illustrates how to use molecular descriptors and machine learning models to screen polar aprotic organic solvents suitable for the method of this invention, such as... Figure 2 The diagram illustrates the steps involved in setting up a candidate solvent set, calculating molecular descriptors, training a machine learning model, screening key descriptors, and applying screening rules to obtain effective solvents.

[0088] (1) Candidate solvent set A candidate set of 431 organic solvents was selected from publicly available solvent databases and literature, including commonly used polar and nonpolar solvents, protic and non-protic solvents, etc.

[0089] (2) Calculation of molecular descriptors Using the cheminformatics software PaDELPy, multiple molecular descriptors were calculated for each candidate solvent, including molecular weight, molecular volume, molar refractive index (AMR), topological polar surface area, and hydrogen bond donor / acceptor counts. A total of 1,875 molecular descriptors were obtained, exceeding the thousand-level. Simultaneously, the empirical parameter E, widely used to characterize solvent polarity, was introduced. T (30) thus forming an initial feature set containing 1,876 descriptors.

[0090] Subsequently, the descriptor data was preprocessed to remove descriptors with missing values ​​and those with constant values ​​across solvents. The Pearson correlation coefficients among the remaining descriptors were then calculated. For highly correlated descriptors with an absolute correlation coefficient of not less than 0.80, only one was retained to reduce feature redundancy and improve model stability. After these processing steps, a feature subset containing 58 molecular descriptors was finally obtained.

[0091] (3) Determination of experimental elution efficiency A subset of solvents, including dimethyl sulfoxide (DMSO), dimethylacetamide (DMF), N-methyl-2-pyrrolidone (NMP), ethylene glycol (MEG), formamide (MF), ethanol (Et), acetonitrile (ACN), n-butanol (NBA), isopropanol (IPA), methanol (Mt), and acetone (DMK), were selected as the training set. Under uniform conditions (100% solvent volume fraction, room temperature, and reaction time of 1 min), their elution efficiency for biotin-Cy5 from streptavidin beads was determined, and the elution efficiency was used as the tag data.

[0092] (4) Machine learning model training A random forest regression model was employed, using the aforementioned molecular descriptors as input features and elution efficiency as output. Cross-validation was used to evaluate model performance and prevent overfitting. Analysis of the feature importance of the random forest model revealed that some descriptors (such as molar refractive index AMR, solvent polarity parameter E)... T (30) etc.) contribute significantly to the prediction of elution performance, such as Figure 2 As shown in (B).

[0093] Figure 2 (A) is a flowchart illustrating the process of performing machine learning feature importance analysis on 11 tested solvents.

[0094] Figure 2 (C) is a schematic diagram of the new solvent screening criteria. Each dot represents an organic solvent. Black dots indicate the absence of hydrogen bond donors, and blue boxes indicate the screening threshold region (molar refractive index greater than 20 and E). T(30) Greater than 42), of which 12 solvents marked in red that do not contain hydrogen bond donors and meet the threshold conditions were selected for further testing.

[0095] (5) Extraction of screening rules Further analysis of the physicochemical properties of validated polar aprotic solvents reveals that typical solvents such as DMSO, DMF, and NMP all have molar refractive indices greater than 20. Simultaneously, the ET(30) values ​​of these solvents are all greater than 42, consistent with the characteristics of polar aprotic solvents possessing high polarity and strong electronic polarization. Based on machine learning results and classical solvent parameter knowledge, the following screening rules are proposed: effective solvents should be non-hydrogen bond donors with high electronic polarization and polarity, such as AMR > 20 and E30 > 42. T (30)>42. Apply these rules to all candidate solvents to obtain the first round of screening results.

[0096] (6) Practicality constraints and experimental verification Based on the results of the first round of testing, and considering practical conditions such as room temperature liquid properties, partial miscibility with aqueous buffer solutions, chemical stability, and controllable toxicity, several commercially available or easily prepared candidate solvents were screened, such as TMSO, DMAC, and DMPU. Subsequently, under experimental conditions similar to or the same as DMSO, the elution efficiency of these solvents on the biotin-streptavidin system was measured, confirming that their elution efficiency could reach or approach the level of DMSO.

[0097] Figure 2(D) is a schematic diagram of the solvent elution experiment performed on streptavidin agarose beads. The solvents tested included tetramethylene sulfoxide (TMSO), N,N-dimethylacetamide (DMAC), dimethyl propylene urea (DMPU), methyl(methylthio)methyl sulfoxide (MMMSO), trimethyl phosphate (TMP), 3-methyl sulfolane (3-MeSF), 1-methylpyrrolidin-2-thione (MTP), N,N-dimethylthioformamide (DMTF), propylene carbonate (PC), N-cyanopyrrolidine (NCP), and N-cyanomorpholine. NCM and N,N-diethylcyanamide (DECAM).

[0098] Through the above process, a closed-loop screening strategy was established from "solvent structure-molecular descriptor-machine learning model-physical property and safety constraints" to "candidate polar aprotic organic solvents", forming a general method that can be used to guide the discovery or design of new solvents.

[0099] Example 4: Topological resolution analysis of mitochondrial proteome Material: Table 3

[0100] This embodiment illustrates how to use the reversible affinity system of the present invention for topological resolution analysis of mitochondrial proteomes, such as... Figure 4 As shown, the process of mitochondrial matrix-targeted labeling enzyme expression, biotinylated neighbor labeling, streptavidin enrichment and elution with polar aprotic organic solvents, as well as mass spectrometry analysis and topological inference is illustrated.

[0101] (1) Cellular and labeling enzyme expression In mammalian cells such as HEK293T, peroxidase-labeled enzymes (e.g., APEX2) carrying mitochondrial matrix-targeting sequences were constructed and stably expressed. The location of this labeling enzyme on the mitochondrial matrix side was confirmed by immunofluorescence or subcellular fractionation.

[0102] (2) Neighbor labeling reaction Cells expressing the labeled enzyme were cultured to an appropriate density, biotinylate was added to the culture medium, and a small amount of hydrogen peroxide was added within a short time (e.g., within 1 minute) to activate the labeling reaction. The reaction was then quickly terminated with a quenching solution. This process was carried out while the cells were alive, and biotinylation modification mainly occurred around the labeled enzyme, with a preference for the mitochondrial matrix and the inner membrane region.

[0103] (3) Protein extraction and enzymatic hydrolysis Labeled cells were collected and lysed using a lysis buffer containing detergent and protease inhibitors to obtain a total protein extract. This extract was then subjected to reduction and alkylation, followed by trypsin digestion to yield a peptide mixture.

[0104] (4) Streptomycin enrichment The peptide mixture was incubated with pre-equilibrated streptavidin beads for 1–2 hours under appropriate salt concentration and pH conditions to allow the biotinylated peptides to specifically bind to the beads. This was followed by multiple washes with a medium-strength or high-salt aqueous wash to remove non-specifically bound peptides.

[0105] Figure 4 (A) is a schematic diagram of the mitochondrial matrix labeling and enrichment workflow. HEK293T cells express APEX2, which targets the mitochondrial matrix. After being labeled with biotinylated phenol / H2O2 pulse, the cells are cleaved and enzymatically digested. The resulting biotinylated peptides are enriched by anti-biotin antibody (upper workflow) or enriched by streptavidin beads and then eluted by FlashBio (lower workflow).

[0106] (5) Elution with polar aprotic organic solvents The beads were treated using the elution conditions described in Example 1, such as DMSO, a polar aprotic organic solvent. For example, 90%–100% DMSO solution was added, reacted for 1 minute, and the eluent was collected. Compared to the conventional method using anti-biotin antibodies for elution, this step is simpler and more reproducible.

[0107] (6) Mass spectrometry analysis and topological inference The eluted peptides were desalted and then analyzed by liquid chromatography-tandem mass spectrometry to identify biotinylated peptides and their corresponding proteins. Based on the location of each biotinylation site in the protein sequence, combined with topological annotations of the inner and outer mitochondrial membranes in the database and transmembrane prediction results, the topological structure of each transmembrane protein was inferred, such as the location of the N-terminus / C-terminus, the number of transmembrane segments, and their distribution on the matrix side and intermembrane side.

[0108] By comparing data obtained using anti-biotin antibody enrichment methods, it was found that the method of the present invention can identify more mitochondrial proteins and a higher number of biotinylation sites, especially with significantly improved coverage of transmembrane proteins and importation complex subunits.

[0109] Figure 4 (B) Comparison of the number of mitochondrial annotated biotinylated peptides identified by enrichment with anti-biotin antibody and enrichment with FlashBio-streptavidin in three biological replicates.

[0110] Figure 4 (C) is a schematic diagram showing the overlap of APEX2-labeled mitochondrial proteins identified in three replicates using the antibody method and the FlashBio method. It shows that APEX2–FlashBio identified a total of 3,791 biotinylated peptides, corresponding to 1,738 proteins, with a significantly improved depth.

[0111] Figure 4 (D) Schematic diagram of label specificity in the TOM / TIM / PAM complex, a protein delivery machine that crosses the outer and inner mitochondrial membranes. APEX2–FlashBio detected all known matrix-side exposed subunits, while the antibody enrichment process missed one of these components, demonstrating improved coverage of this key pathway.

[0112] Figure 4 (E) is a schematic diagram of biotinylation sites of several representative mitochondrial transmembrane proteins, marking both sites reported by previous antibody strategies and newly detected sites in key proteins such as LETM1, IMMT, ATAD3A and NNT.

[0113] Figure 4 (F) is a schematic diagram showing the assignment or fine-tuning of the transmembrane topology of mitochondrial inner membrane proteins such as ATP5PB and PHB2 using high-density site information obtained by APEX2-FlashBio.

[0114] Example 5: Topological resolution and dynamic remodeling analysis of cell surface proteome This embodiment illustrates how to use the method of the present invention for topological resolution and dynamic analysis of cell surface proteomes, such as... Figure 5 As shown, the steps of biotinylated lysine labeling on the HeLa cell surface, streptavidin enrichment and elution with polar aprotic organic solvents, mass spectrometry identification, and Full / Partial / No / Blank topological classification are illustrated; among them, Figure 5(A) is a schematic diagram of the experimental strategy, showing the labeling of live HeLa cells with a membrane-impermeable EZ-Link™ Sulfo-NHS-LC-biotin probe, followed by streptavidin capture, FlashBio elution, and LC–MS / MS analysis to define the cell surface proteome. Figure 6 The diagram illustrates the steps of biotinylated lysine labeling on the surface of HEK293K cells, streptavidin enrichment and elution with polar aprotic organic solvents, mass spectrometry identification, and Full / Partial / No / Blank topological classification.

[0115] (1) Cell surface markers Taking adherent cell lines such as HeLa or HEK293T as examples, cells are washed with PBS on ice or at 4°C to remove amine-containing substances from the culture medium. Then, a chilled solution containing the Sulfo-NHS-LC-biotin probe is added, and the reaction is carried out under mild conditions for a certain time (e.g., 30 minutes). This reagent is membrane-impermeable and can only label primary amine residues exposed on the outer side of the cell. After the reaction, the reaction is terminated with a quenching solution containing an amino compound, and residual reagents are removed.

[0116] (2) Protein extraction and enzymatic hydrolysis The labeled cells were lysed to prepare a total protein extract, which was then reduced, alkylated, and digested with trypsin to obtain a peptide mixture.

[0117] (3) Streptavidin enrichment and elution with polar aprotic organic solvents The peptide mixture was incubated with streptavidin beads to capture biotinylated peptides, and then washed multiple times with PBS containing detergent and high salt. Subsequently, the surface biotinylated peptides were rapidly eluted and collected using the polar aprotic organic solvent elution conditions described in Example 1 (e.g., 90%-100% DMSO).

[0118] (4) Mass spectrometry identification and site mapping The eluted peptides were desalted and then identified by mass spectrometry to determine the sequence of each peptide and the biotinylated lysine residues it contained. These residues were then mapped onto the amino acid sequences of the corresponding proteins to obtain or predict the transmembrane structure information of the proteins.

[0119] Figure 5 (B) is a schematic diagram of the topological analysis process for transmembrane (TM) proteins, which integrates the topological annotations in UniProt with the location information of biotinylated lysine sites; Figure 5(C) is a schematic diagram of the classification of biotinylated proteins based on UniProt subcellular localization annotation and biotinylation site subcellular / local localization, highlighting plasma membrane™ proteins with extracellular exposure modifications.

[0120] Figure 6 (A) is a schematic diagram of the plasma membrane™ protein topology analysis workflow, which integrates the topological annotations in UniProt with the biotinylated lysine sites identified by FlashBio. Figure 6 (B) is a schematic diagram of the classification of biotinylated proteins based on UniProt subcellular localization annotation and subcellular / local distribution of biotinylated sites, highlighting plasma membrane™ proteins with extracellular exposure modifications.

[0121] (5) Topological classification Based on the location of the biotinylation site in the transmembrane topology model, each transmembrane protein is classified into four categories: Full, Partial, No, or Blank, thereby obtaining the overall topological distribution of transmembrane proteins on the cell surface.

[0122] Figure 5 (D) is a schematic diagram of a representative membrane™ protein of the Full Match category, with all its biotinylated residues mapped to the annotated extracellular region; Figure 6 (C) is a schematic diagram of a representative membrane™ protein of the Full Match category, with all its biotinylated residues mapped to the annotated extracellular region; Figure 5 (E) is a schematic diagram of a representative membrane™ protein of the No matching category, where no biotinylated residues are mapped to the annotated extracellular region; Figure 6 (D) is a schematic diagram of a representative membrane™ protein of the No matching category, where no biotinylated residues are mapped to the annotated extracellular region; Figure 5 (F) is a schematic diagram of a representative membrane™ protein of the Partial Matching category, with some biotinylated residues located in the extracellular region and some in the non-extracellular region; Figure 6 (E) is a schematic diagram of a representative membrane™ protein of the Partial Matching Class, with some biotinylated residues located in the extracellular region and some in the non-extracellular region; Figure 5 (G) is a schematic diagram of a representative membrane TM protein of the Blank class. The existing topological annotation is insufficient to determine the intracellular and extracellular orientation of its modified residues. Figure 6(F) is a schematic diagram of a representative membrane TM protein of the Blank class. The existing topological annotation is insufficient to determine the orientation of its modified residues.

[0123] Figure 5 (H) is a schematic diagram of several examples of plasma membrane TM proteins. By combining the predicted TM fragments with the biotinylation site map based on FlashBio, their transmembrane topology can be inferred or refined. Figure 6 (G) is a schematic diagram of several plasma membrane TM proteins. By combining the predicted TM fragments with the biotinylation site map based on FlashBio, their transmembrane topology can be inferred or refined.

[0124] (6) Dynamic reshaping analysis Taking the THP-1 mononuclear cell line as an example, surface labeling and analysis can be performed in the untreated state, followed by the addition of phorbol ester to induce differentiation into macrophage-like cells, and then surface labeling and analysis can be performed again. By comparing the abundance changes of various transmembrane proteins in the two states and the switching between Full, Partial, No, and Blank categories, changes in the degree of membrane protein surface exposure during differentiation can be revealed, helping to identify surface targets related to differentiation or with immunomodulatory significance.

[0125] Figure 7 This diagram illustrates the changes in the abundance and topological class of THP-1 cell surface proteins under different biological conditions (such as before or after stimulation or differentiation), showing the switching between different topological classes of some representative transmembrane proteins. (A) is a schematic diagram of the experimental strategy, showing that, under conditions with or without PMA stimulation, the surface of live THP-1 cells was labeled with the EZ-Link™ Sulfo-NHS-LC-biotin probe on ice, followed by lysis and streptavidin capture, FlashBio elution, and LC-MS / MS analysis to define the THP-1 cell surface proteome; (B) is a bar chart summarizing the number of biotinylated surface proteins identified by the FlashBio workflow in three independent LC-MS / MS experiments; (C) is a schematic diagram comparing the PMA-regulated surface proteins identified in this invention with the regulatory proteins reported in existing THP-1 surface proteome studies, showing that the FlashBio method detected more regulated surface proteins during the differentiation of monocytes into macrophage-like cells; (D) is a volcano plot comparing the changes in the THP-1 cell surface proteome between the PMA-treated group and the control group, with the horizontal axis representing log2 fold changes and the vertical axis representing -log 10p-values ​​highlight surface proteins that are significantly upregulated and downregulated by PMA stimulation; (E) is a schematic diagram of representative downregulated surface proteins, showing their decreased abundance and accompanying topological changes after PMA stimulation through peptide-level site mapping; (F) is a schematic diagram of representative upregulated surface proteins, showing their increased surface abundance and topological changes after PMA treatment; (G) is a schematic diagram of representative surface proteins whose overall abundance changes are not significant but whose topological features (such as the relative exposure of different extracellular fragments) are reshaped during PMA-induced differentiation.

[0126] Example 6: Purification method of biotinylated single-stranded DNA (ssDNA) Material: Table 4

[0127] This embodiment expands upon the original application scenario of the present invention, focusing on verifying the specific purification performance of the present invention for biotin-modified single-stranded DNA (ssDNA). The specific operation steps and result analysis are as follows: (1) Specific capture of biotinylated ssDNA Take 100 μL of streptavidin agarose beads and place them in a 1.5 mL centrifuge tube. Wash the beads with ddH2O, adding 1 mL of ddH2O each time. Centrifuge and discard the supernatant. Repeat the washing process three times to remove residual preservatives and impurities from the bead surface. After washing, add 200 μL of a 40.0 nmol / mL biotinylated ssDNA solution (sequences shown in Table 5, as shown in SEQ ID NO. 1-5) to the centrifuge tube. Seal the centrifuge tube and incubate it in a shaker at 4°C for 10–20 min to allow the biotinylated ssDNA to specifically bind to streptavidin on the surface of the streptavidin agarose beads. After incubation, centrifuge and collect all the supernatant (i.e., the unbound liquid) for later use.

[0128] Table 5

[0129] (2) Urea aqueous phase washing to remove non-specific adsorbates Add washing solution (ddH2O solution containing 6 mol / L urea) pre-cooled to 4°C to the centrifuge tubes that have undergone the capture treatment. After standing at room temperature for 2 min, centrifuge and discard the supernatant. Repeat the above washing operation 3 to 4 times to fully remove non-specific impurities on the surface of the beads. Then, rinse the beads in the centrifuge tubes with ddH2O, adding 1 mL of ddH2O each time, centrifuging and discarding the supernatant, rinsing 1 to 2 times to thoroughly remove residual urea on the surface of the beads and in the centrifuge tubes, so as to avoid interference with subsequent elution steps.

[0130] (3) Elution of target ssDNA with polar aprotic organic solvents Add 600 μL of 90%–100% dimethyl sulfoxide (DMSO) eluent to the washed centrifuge tube, seal the centrifuge tube, and place it on a vortex mixer at room temperature for 1–2 min. After vortexing, centrifuge and collect all the supernatant (i.e., eluent) into the centrifuge tube and label it as the eluted sample. In this step, other screened polar aprotic organic solvents (such as N,N-dimethylformamide, acetonitrile, etc.) can also be used as eluents. The elution conditions can be adjusted according to the type of solvent.

[0131] (4) Elution effect detection and result analysis The purity and recovery rate of target biotinylated ssDNA in the eluted samples were quantitatively determined by reversed-phase high-performance liquid chromatography (RP-HPLC). The detection conditions were as follows: the chromatographic column was a C18 reversed-phase column (4.6 mm × 250 mm, 5 μm); mobile phase A was a ddH2O solution containing 0.1% triethylamine acetate (TEAA), and mobile phase B was an acetonitrile solution; the elution gradient was as follows: the volume fraction of mobile phase B increased from 2% to 25% within 0–20 min, and decreased from 25% to 2% within 20–20.1 min, then held for 28 min; the flow rate was 1.0 mL / min; the detection wavelength was 260 nm; the column temperature was 40℃; and the injection volume was 10 μL. Unpurified biotinylated ssDNA standards were used as controls to calculate the recovery rate of the target ssDNA.

[0132] The test results showed that for biotinylated ssDNA with a chain length of 12 nt, the purification and recovery efficiency of this invention reached 17%. However, as the ssDNA chain length increased, the recovery efficiency showed a significant decreasing trend. When the ssDNA chain length increased to 24 nt, the recovery efficiency was less than 0.5%, which can be considered as essentially impossible to achieve effective recovery. These results indicate that this invention has certain chain length-dependent limitations in the purification application of biotinylated ssDNA, but it still has good specific purification effects and practical application potential for short-chain biotinylated ssDNA with a chain length ≤12 nt.

[0133] Figure 8To assess the elution efficiency of five different ssDNA lengths and expand the application scope of Flashbio technology, this invention attempted to enrich ssDNA of different lengths (12 nt, 16 nt, 20 nt, 24 nt, and 48 nt) using this technology. It was found that the recovery efficiency decreased with increasing chain length. Nevertheless, within the 12 nt-16 nt length range, DMSO could still recover 5%-17% of the biotin-modified ssDNA bound to streptavidin beads, providing a possibility for future enrichment of biotin-modified ssDNA using this strategy.

[0134] Example 7: Trifunctional probes for the discovery of target proteins and binding sites (1) Probe preparation and incubation with cell lysis buffer First, a trifunctional probe, Biotin-Diazirine-Lenalidomide (hereinafter referred to as "probe L"), was synthesized. This probe was obtained through total chemical synthesis and contains a biotin affinity tag, a diaziridine photocrosslinking group, and a lenalidomide pharmacophore. Figure 9 (A) is a schematic diagram of the chemical structure of the trifunctional probe, which consists of a lenalidomide ligand (blue), a photoactivated diazirine crosslinking group (red), and a biotin affinity tag (green). HeLa cells were collected, and cell lysates were prepared using RIPA lysis buffer containing a protease inhibitor, with the protein concentration adjusted to 2 mg / mL. Probe L was added to the lysis buffer to achieve a final concentration of 20 µM. The mixture was incubated at room temperature by rotation for 10 minutes to establish non-covalent binding between the drug and the target protein.

[0135] (2) In-situ photocrosslinking After incubation, the sample was placed under a 365 nm wavelength UV lamp and irradiated for 2 minutes at a distance of 5 cm from the light source to induce the biacrididin group to generate a carbene intermediate, thereby covalently crosslinking the probe L with its binding protein (target).

[0136] (3) Protein denaturation, reduction and alkylation Solid urea was added directly to the light-exposed sample to a final concentration of 8 M, and the mixture was shaken at room temperature for 30 minutes to completely denature the protein. Dithiothreitol (DTT) was then added to a final concentration of 10 mM, and the mixture was reacted at 37°C for 30 minutes to reduce disulfide bonds. After cooling to room temperature, iodoacetamide (IAA) was added to a final concentration of 20 mM, and the mixture was reacted at room temperature for 30 minutes under light-protected conditions for alkylation.

[0137] (4) Enzymatic hydrolysis Dilute the sample with 50 mM ammonium bicarbonate (NH4HCO3) buffer to reduce the urea concentration to below 1.5 M. Add sequencing-grade trypsin at an enzyme-to-substrate ratio of 1:50 and incubate at 37°C for 12-16 hours to digest the protein into a peptide mixture.

[0138] (5) Affinity enrichment and rigorous washing Add 100 µL of streptavidin agarose beads to the enzymatic digest and incubate at room temperature for 1 hour by rotation to capture biotin-tagged cross-linked peptides. Centrifuge to separate the agarose beads and discard the supernatant. Then, wash the agarose beads three times with PBS buffer containing 6 M urea for 5 minutes each time to thoroughly remove non-specifically adsorbed impurity peptides and retain the covalently cross-linked target peptides.

[0139] (6) FlashBio-specific dissociation and sample recovery FlashBio's dedicated dissociation buffer (DMSO) was added to the washed agarose beads to directly disrupt the non-covalent interaction between biotin and streptavidin under mild conditions without cleaving the probe's chemical bonds. The supernatant, containing the intact "drug-probe-peptide" ternary complex, was collected. The eluent was dried in a vacuum centrifuge (SpeedVac) to remove organic solvents, then redissolved in 10 µL of an aqueous solution containing 0.1% formic acid, and analyzed by LC-MS / MS mass spectrometry.

[0140] (7) Mass spectrometry detection and data analysis results Mass spectrometry data were retrieved using proteomics analysis software. The fixed modification was set to cysteine ​​carbamoyl methylation, and the variable modification was set to "probe L full mass modification" (mass offset = probe L molecular weight - 28 Da, corresponding to the removed nitrogen molecules). Analysis results showed: (1) Target identification: The known positive target protein of lenalidomide, Cereblon (CRBN), was identified with high confidence directly from the complex background of cell lysates. (2) Site resolution: Characteristic peptides carrying the complete probe modification were identified on the CRBN protein sequence. Secondary mass spectrometry (MS / MS) fragment ion identification of crosslinking sites was located in the tryptophan residue (Trp380, Trp386, Trp400) region within the thalidomide binding domain (TBD) of CRBN. The results highly overlap / correlate with the hydrophobic cage binding mode of lenalidomide and CRBN in X-ray crystal structures reported in the literature, confirming the ability of this method to simultaneously resolve drug targets and their binding sites in complex biological systems without the need for recombinant proteins.

[0141] Figure 9 This is a trifunctional probe used for target protein and binding site discovery. It also identifies the lenalidomide target CRBN and uses photoaffinity probes to map its binding pocket.

[0142] Figure 9 (B) is a schematic diagram of the process of photocrosslinking enrichment and LC-MS / MS analysis. This method can identify target proteins and their specific crosslinking sites in complex lysis buffers in parallel.

[0143] Figure 9 (C) is a ranking-intensity plot of proteins identified by LC-MS / MS in the enrichment experiment. Proteins are ranked by intensity, with the highest-ranking protein, CRBN, highlighted in red.

[0144] Figure 9 (D) is a sequence coverage diagram of human CRBN, with identified peptides marked in blue; specific modification sites are marked, including pLen (+620.25 Da) which is cross-linked with the probe, and is indicated by blue squares.

[0145] Figure 9 (E) is a representative MS / MS spectrum of the CRBN peptide (NH2-GYVHETLVYK-COOH) crosslinked with the probe. The presence of the probe residue (+620.25 Da) confirmed histidine (H) as the crosslinking site. The annotated b and y ion series validated the peptide sequence and its modification site.

[0146] Figure 9 (F) shows a magnified view of the lenalidomide binding pocket on the right, mapping the identified crosslinking sites onto the crystal structure of CRBN. Known key residues (W380, W386, W400) are marked in green. Identified crosslinking sites are colored according to frequency of occurrence: high-frequency sites are red, and low-frequency sites are orange. The identified high-frequency sites highly overlap with known key binding residues.

[0147] The specific embodiments of the present invention have been described above. It should be understood that the present invention is not limited to the specific embodiments described above, and those skilled in the art can make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. The application of a polar aprotic organic solvent in the dissociation of streptavidin-biotin conjugates, characterized in that, The polar aprotic organic solvent is selected from at least one of dimethyl sulfoxide, dimethylformamide, N-methyl-2-pyrrolidone, tetramethylene sulfoxide, dimethylacetamide, dimethylpropenylurea, or a mixture thereof with water.

2. The application according to claim 1, characterized in that, The volume fraction of the polar aprotic organic solvent in the mixture is above 80%.

3. The application according to claim 1, characterized in that, The method for dissociating the streptavidin-biotin conjugate includes the following steps: S1. Wash the streptavidin-biotin conjugate sample with an aqueous washing solution to remove non-specifically bound components; S2. Then treat the streptavidin-biotin conjugate sample with a polar aprotic organic solvent; S3. Remove the polar aprotic organic solvent and equilibrate the streptavidin-biotin conjugate sample with aqueous buffer. Repeat steps S1 and S2.

4. The application according to claim 3, characterized in that, The linker in the biotin conjugate is selected from at least one of small molecule compounds, peptides, or oligonucleotides.

5. A method for screening polar aprotic organic solvents in any one of claims 1-4, characterized in that, Includes the following steps: A1. Select all organic solvents from the public database as candidate solvents and calculate molecular descriptors for the candidate solvents; A2. Select a portion of the solvent as the training set and determine its elution efficiency for biotin-Cy5 dissociation from streptavidin. A3. A machine learning model is used, with the molecular descriptors in step A1 as input features and the elution efficiency in step A2 as output. Cross-validation is used to evaluate the model performance and prevent overfitting. By analyzing the feature importance of the machine learning model, molecular descriptors that contribute significantly to the prediction of elution performance are selected. The molecular descriptors that contribute significantly to the prediction of elution performance include molar refractive index (AMR) and solvent polarity parameter (E). T (30); A4. Based on molecular descriptors and classical solvent parameters that significantly contribute to the prediction of elution performance, solvents with a molar refractive index (AMR) greater than 20 and solvent polarity parameter E are selected from the candidate solvents. T (30) Solvents with a value greater than 42 and no hydrogen bond donor capability are considered as the first candidate set; A5. Based on practical constraints, solvents that are liquid at room temperature, miscible with water, and chemically stable are selected from the first candidate set as polar aprotic organic solvents.

6. The screening method according to claim 5, characterized in that, In step A1, the candidate solvents include polar solvents, nonpolar solvents, protic solvents, and aprotic solvents; And / or, in step A3, the machine learning model includes one or more of the following: random forest model, gradient boosting tree model, and support vector machine model.

7. The application of a polar aprotic organic solvent in purifying biotinylated small molecule compounds in crude reaction solutions, analyzing topologically resolved mitochondrial proteomics, analyzing topologically resolved cell surface proteomics, identifying small molecule drug targets and binding sites, and purifying biotinylated single-stranded DNA.

8. The application according to claim 7, characterized in that, The method for purifying biotinylated small molecule compounds in the crude reaction solution includes the following steps: B1. Add the crude reaction solution containing biotinylated small molecule compounds to the solid packing material filled with streptavidin; B2. Wash the solid packing with a washing solution to remove non-specifically bound impurities; B3. The solid-phase packing material is then eluted with a polar aprotic organic solvent to obtain the eluent of the biotinylated small molecule compound.

9. The application according to claim 7, characterized in that, The method for analyzing topologically resolved mitochondrial proteomics includes the following steps: C1. Express a peroxidase-labeled enzyme that targets the mitochondrial matrix in the cells to be tested, thereby localizing the labeled enzyme to the mitochondrial matrix; C2. Add biotin and hydrogen peroxide to the cells to activate the labeling reaction, triggering the labeling enzyme to catalyze the generation of active free radicals and biotinylate mitochondrial adjacent proteins in the cell survival state. C3. The labeled cells were lysed and digested with proteases to obtain a sample containing biotinylated peptides; C4. Contact the sample with streptavidin to enrich the biotinylated peptides, and wash with an aqueous washing solution. C5. The biotinylated peptides are then eluted with a polar aprotic organic solvent. C6. Mass spectrometry analysis was performed on the eluted peptides, and the transmembrane topology of mitochondrial proteins was assigned or modified based on biotinylation site information.

10. The application according to claim 7, characterized in that, The method for analyzing topologically resolved cell surface proteomics includes the following steps: D1. Add a membrane-impermeable biotinylated lysine reaction reagent to the surface of living cells to biotinylate lysine residues exposed on the outer side of the cell. D2. The cells are lysed and digested with proteases to obtain a sample containing biotinylated peptides. D3. Contact the sample with streptavidin to enrich the biotinylated peptides, and wash with an aqueous washing solution. D4. The biotinylated peptides are then eluted with a polar aprotic organic solvent. D5. Perform mass spectrometry analysis on the eluted peptides and map the obtained biotinylated sites to the corresponding protein amino acid sequence and its database topological annotation or transmembrane prediction results. D6. Based on whether the biotinylated sites fall within the labeled or predicted extracellular regions, transmembrane proteins are divided into a fully matched category (all sites are located outside the cell), a partially matched category (some sites are located outside the cell), an unmatched category (no extracellular sites), and a blank category (cannot be determined due to missing topological information), thus obtaining a topologically resolved cell surface proteome map.

11. The application according to claim 7, characterized in that, The method for identifying small molecule drug targets and binding sites includes the following steps: E1. Probe assembly: Small molecule drugs are coupled with trifunctional probes through click chemistry reactions to obtain conjugates; E2. Photocrosslinking: The conjugate is incubated with biological samples and then exposed to light to covalently crosslink the probe with the target protein in the biological sample; E3. Enzymatic hydrolysis: The biological sample undergoes protein denaturation, reduction and alkylation, followed by enzymatic hydrolysis into a mixture of peptides; E4. Affinity enrichment: The peptide mixture is contacted with a solid support bonded with streptavidin to capture biotinylated cross-linked peptides. E5. Dissociation: The solid support is exposed to a polar aprotic organic solvent to elute the cross-linked peptides. E6. Mass spectrometry analysis: Detects eluted peptides and identifies drug binding sites by recognizing amino acid sites carrying intact probes; screens target peptides by detecting the molecular weight of intact trifunctional probes attached to the peptide precursor ion.

12. The application according to claim 7, characterized in that, The method for purifying biotinylated single-stranded DNA includes the following steps: F1. Mix streptavidin with biotinylated single-stranded DNA, vortex and incubate, then centrifuge and collect the supernatant. F2. Wash the supernatant with an aqueous washing solution; F3. Elute the supernatant with a polar aprotic organic solvent to obtain the target ssDNA.

Images