Dilution-induced phase separation polypeptides and applications thereof

Abstract

The application discloses dilution-induced phase separation polypeptides and application thereof. The application surprisingly finds a polypeptide TP presenting abnormal phase change dynamics, studies liquid-liquid phase separation properties and molecular mechanisms of TP dilution induction, finds that liquid-liquid phase separation can be induced by diluting a homogeneous TP solution, thereby realizing compartmentalization of biological macromolecules, which overturns the traditional understanding of liquid-liquid phase separation. The application also verifies that the dilution-induced condensation effect of the TP can realize compartmentalization of guest molecules, regulate optical output of the guest, promote secondary liquid-liquid phase separation of proteins in the condensate and optimize enzymatic reactions, thereby proving practicability of the dilution-induced condensate of the TP, and providing a brand-new regulation strategy for biological macromolecule self-assembly and function regulation, and having wide application prospects.

Description

Technical Field

[0001] This invention belongs to the field of biomedicine, specifically relating to dilution-induced phase separation peptides and their applications. Background Technology

[0002] Compartmentation is a fundamental characteristic of all living organisms. By separating active components into different compartments, living systems can perform sophisticated biological functions such as material transport, energy conversion, and information exchange. Understanding the molecular pathways by which complex compartmentalized structures are constructed in nature not only provides a unique perspective for exploring the essence of cellular life but also offers important insights into understanding the mechanisms of the origin of life on Earth. However, the efficient concentration of biomolecules within compartments faces thermodynamic challenges—the entropy loss associated with this process puts it thermodynamically disadvantageous. To achieve compartmentalization of biomolecules, cells and their precursor systems have evolved various compartmentalization strategies: forming membrane-like compartments through the self-assembly of amphiphilic lipids, polysaccharides, peptides, and even inorganic structures. Unlike traditional membrane structures or membrane-like structures, liquid-liquid phase separation (LLPS), as a novel mechanism for concentrating biomolecules into membrane-free condensates, has recently been recognized as an important way to achieve cellular compartmentalization. Peptides with phase-separation capabilities can encapsulate other molecules into the phase-separated system to exert their effects. Summary of the Invention

[0003] To overcome the shortcomings of existing technologies, this invention provides a polypeptide and its applications.

[0004] To achieve the above objectives, the present invention adopts the following technical solution:

[0005] A first aspect of the present invention provides a polypeptide comprising an amino acid sequence selected from the group consisting of:

[0006] (1) An amino acid sequence obtained by substituting, deleting, adding or inserting at least one amino acid residue into SEQ ID NO:1;

[0007] (2) An amino acid sequence that has at least 90% sequence identity with SEQ ID NO:1.

[0008] Furthermore, the substitution is performed by one or more of leucine, glutamic acid, and alanine.

[0009] Furthermore, leucine was replaced by isoleucine.

[0010] Furthermore, glutamic acid is replaced by valine and / or aspartic acid.

[0011] Furthermore, alanine was replaced by aspartic acid.

[0012] Furthermore, the sequence of the polypeptide is shown in any one of SEQ ID NO:1-5.

[0013] A second aspect of the present invention provides a polypeptide condensate comprising the polypeptide described in the first aspect of the present invention.

[0014] A third aspect of the present invention provides a polynucleotide that encodes a polypeptide as described in the first aspect of the present invention or a polypeptide condensate as described in the second aspect of the present invention.

[0015] A fourth aspect of the present invention provides a carrier comprising the polynucleotide described in the third aspect of the present invention.

[0016] In some implementations, examples of vectors include, but are not limited to, retroviruses (including lentiviruses), adenoviruses, adeno-associated viruses, herpesviruses (e.g., herpes simplex virus), poxviruses, baculoviruses, papillomaviruses, multifocal papillomaviruses (e.g., SV40), λ phages and M13 phages, and plasmids. Specific examples of vectors include, but are not limited to, pcDNA3.3, pMD18-T, pOptivec, pCMV, pEGFP, pIRES, pQD-Hyg-GSeu, pALTER, pBAD, pcDNA, pCal, pL, pET, pGEMEX, pGEX, pCI, pEGFT, pSV2, pFUSE, pVITRO, pVIVO, pMAL, pMONO, pSELECT, pUNO, pDUO, Psg5L, pBABE, pWPXL, pBI, p15TV-L, pPro18, pTD, pRS10, pLexA, pACT2.2, pCMV-SCRIPT.RTM., pCDM8, pCDNA1.1 / amp, pcDNA3.1, pRc / RSV, PCR2.1, pEF-1, pFB, pSG5, pXT1, pCDEF3, pSVSPORT, and pEF-Bos.

[0017] A fifth aspect of the present invention provides a host cell comprising the polynucleotide described in the third aspect of the present invention or the vector described in the fourth aspect of the present invention.

[0018] In some embodiments, the host cell is a cell used to receive, retain, replicate, and amplify the vector. The host cell can also be used to express the polypeptide encoded by the vector. When the host cell divides, the nucleic acids contained in the vector replicate, thereby amplifying the nucleic acids. In one embodiment, the host cell is a genetic package that can be induced to express variant polypeptides on its surface. In another embodiment, the host cell is infected with the genetic package.

[0019] In some embodiments, the host cell can be virtually any cell available for the expression vector. This includes prokaryotic cells and eukaryotic cells, with the prokaryotic cells including, but not limited to, eubacteria such as Gram-negative or Gram-positive organisms, such as Enterobacteriaceae, such as Escherichia, for example, *Escherichia coli* (DH5α, BL21DE3, BL21DE3pLysS, JM109, TOP10, HB101, SCS110, *E. coli* JM110); *Enterobacter*; *Erwinia*; *Klebsiella*; *Proteus*; *Salmonella*, for example, *Salmonella typhimurium*; and *Serratia*, for example, *Serratia marcescens*. The genera *Bacillus* include *Bacillus subtilis* and *Bacillus licheniformis*; *Pseudomonas*, such as *P. aeruginosa*; and *Streptomyces*.

[0020] Eukaryotic cells include, but are not limited to, protist cells, animal cells, or fungal cells. Animal cells include mammalian cells, avian cells, and insect cells. Mammalian cells include, but are not limited to, CHO cells, F2N cells, CSO cells, BHK cells, Bowes melanoma cells, HeLa cells, 911 cells, AT1080 cells, A549 cells, 293T cells, and 293F cells.

[0021] The sixth aspect of the present invention provides the use of the polypeptides described in the first aspect of the present invention, the polypeptide condensates described in the second aspect of the present invention, the polynucleotides described in the third aspect of the present invention, the carriers described in the fourth aspect of the present invention, and the host cells described in the fifth aspect of the present invention in liquid-liquid phase separation, enrichment, or in the preparation of products that regulate liquid-liquid phase separation and enrichment.

[0022] The seventh aspect of the present invention provides the use of the polypeptides described in the first aspect of the present invention, the polypeptide condensates described in the second aspect of the present invention, the polynucleotides described in the third aspect of the present invention, the carriers described in the fourth aspect of the present invention, and the host cells described in the fifth aspect of the present invention in regulating the compartmentalization of guest molecules or in the preparation of products that regulate the compartmentalization of guest molecules.

[0023] The eighth aspect of the present invention provides the use of the polypeptides described in the first aspect of the present invention, the polypeptide condensates described in the second aspect of the present invention, the polynucleotides described in the third aspect of the present invention, the carriers described in the fourth aspect of the present invention, and the host cells described in the fifth aspect of the present invention in aggregation-induced emission or in the preparation of aggregation-induced emission products.

[0024] Furthermore, the product includes an aggregation-induced emission probe.

[0025] The ninth aspect of the present invention provides the use of the polypeptides described in the first aspect of the present invention, the polypeptide condensates described in the second aspect of the present invention, the polynucleotides described in the third aspect of the present invention, the carriers described in the fourth aspect of the present invention, and the host cells described in the fifth aspect of the present invention in promoting enzymatic reactions or in the preparation of products that promote enzymatic reactions.

[0026] The tenth aspect of the present invention provides any of the following methods:

[0027] (1) A method for regulating liquid-liquid phase separation and enrichment, the method comprising using the polypeptide described in the first aspect of the present invention, the polypeptide condensate described in the second aspect of the present invention, the polynucleotide described in the third aspect of the present invention, the carrier described in the fourth aspect of the present invention, and the host cell described in the fifth aspect of the present invention.

[0028] (2) A method for regulating the compartmentalization of guest molecules, the method comprising using the polypeptides described in the first aspect of the present invention, the polypeptide condensates described in the second aspect of the present invention, the polynucleotides described in the third aspect of the present invention, the carriers described in the fourth aspect of the present invention, and the host cells described in the fifth aspect of the present invention;

[0029] (3) A method for promoting aggregation-induced emission, the method comprising using the polypeptide described in the first aspect of the present invention, the polypeptide condensate described in the second aspect of the present invention, the polynucleotide described in the third aspect of the present invention, the carrier described in the fourth aspect of the present invention, and the host cell described in the fifth aspect of the present invention;

[0030] (4) A method for promoting an enzymatic reaction, the method comprising using the polypeptide described in the first aspect of the present invention, the polypeptide condensate described in the second aspect of the present invention, the polynucleotide described in the third aspect of the present invention, the carrier described in the fourth aspect of the present invention, and the host cell described in the fifth aspect of the present invention.

[0031] Advantages and beneficial effects of the present invention:

[0032] This application unexpectedly discovered a polypeptide TP exhibiting anomalous phase transition kinetics. The properties and molecular mechanism of TP dilution-induced liquid-liquid phase separation were investigated. It was found that diluting a homogeneous TP solution can induce liquid-liquid phase separation, thereby achieving compartmentalization of biomolecules. This discovery overturns the traditional understanding of liquid-liquid phase separation. This application also verifies that the TP dilution-induced aggregation effect can achieve compartmentalization of guest molecules, regulate the optical output of guest molecules, promote the formation of secondary liquid-liquid phase separation of proteins within aggregates, and optimize enzymatic reactions. This not only demonstrates the practicality of TP dilution-induced aggregates but also provides a novel regulatory strategy for the self-assembly and functional regulation of biomolecules, with broad application prospects. Attached Figure Description

[0033] Figure 1 This is a graph showing the phase separation effect induced by TP solution dilution. 1A is a graph of TP solutions at different concentrations during the dilution process in a system with 20mM Tricine and pH=7.4. 1B is a bar graph of static light scattering intensity corresponding to each concentration in 1A (N=5). 1C is a graph of OD550 signal intensity corresponding to the detection of TP by the microplate reader when it travels between two concentrations of 1mM and 0.75mM.

[0034] Figure 2 This is a morphology image of TP condensate under a differential interference difference microscope;

[0035] Figure 3 This is a concentration-related phase diagram showing the separation of Tricine and TP phases;

[0036] Figure 4 This is a concentration-correlation phase diagram showing the separation of Tris and TP phases;

[0037] Figure 5 This is a concentration-related phase diagram showing the separation of TEA and TP phases;

[0038] Figure 6 It is a concentration-related phase diagram showing the separation of phosphate and TP phases;

[0039] Figure 7 These are time-series microscopic images suggesting that TP condensates readily wet glass interfaces;

[0040] Figure 8 This is a graph showing the correlation between buffer salt concentration and aggregate size revealed by dynamic light scattering;

[0041] Figure 9 These are fitting graphs of the fusion dynamics curves of TP condensates. Among them, 9A is an example image of a microscope photograph of the fusion process of TP condensates, 9B is a scatter plot of the correlation between the eccentricity and time based on morphological measurements during the fusion process, and 9C is a fitting graph of the correlation function between the relaxation time and the final radius of the TP condensate fusion.

[0042] Figure 10 This is a graph showing molecular exchange kinetics detected by a fluorescence bleaching recovery assay;

[0043] Figure 11 This is a full Raman spectrum of TP condensate inside and outside the body;

[0044] Figure 12 The method involves constructing a standard curve of phenylalanine concentration-signal intensity to calculate the peptide concentration map in TP aggregates;

[0045] Figure 13 This is a comparison diagram of the positions of key concentration points in the phase diagram during TP phase separation events;

[0046] Figure 14 This is a peak distribution diagram of the secondary structure of the amide I region of the TP condensate;

[0047] Figure 15 This is a comparison chart of the dilution-induced LLPS model and the classic LLPS model;

[0048] Figure 16 This is a diagram showing the AlphaFold3 prediction of TP monomer structure;

[0049] Figure 17 This is a diagram illustrating the disorder analysis of TP sequences from the IUPred3 website;

[0050] Figure 18 This is a diagram from the PLAAC website showing the prion-like characteristics of the TP sequence;

[0051] Figure 19 This is a nuclear magnetic resonance spectroscopy analysis of the monomer structure of TP in a high-concentration homogeneous state;

[0052] Figure 20 This is a circular dichroism spectrum showing the conformational changes of TP during the dilution process;

[0053] Figure 21 This is a phase diagram showing that the addition of trifluoroethanol promotes the formation of TP aggregates;

[0054] Figure 22 This is a predicted structural diagram of the TP-3A mutant in AlphaFold3;

[0055] Figure 23 This is a graph showing the dilution-induced phase separation effect of the TP-3A mutant.

[0056] Figure 24 This is a graph showing the dilution-induced phase separation effect of the TP-2E mutant;

[0057] Figure 25 This is a diagram illustrating the effect of 1,6-hexanediol on TP condensates;

[0058] Figure 26 This is a graph showing the dilution-induced phase separation effect of the TP-2A mutant.

[0059] Figure 27 This is a diagram illustrating the correlation between TP phase separation capability and pH value;

[0060] Figure 28 This is a summary of concentration-related phase diagrams for the separation of various salts from the TP phase;

[0061] Figure 29 This is a concentration-related phase diagram showing the separation of sodium chloride and TP phases under 10 mM Tirs conditions;

[0062] Figure 30 This is a concentration-related phase diagram showing the separation of magnesium chloride and TP phases under 10 mM Tirs conditions;

[0063] Figure 31 This is a concentration-related phase diagram showing the separation of sodium sulfate and TP phases under 10 mM Tirs conditions;

[0064] Figure 32 This is a diagram showing the mutation locations of TP homologous peptides;

[0065] Figure 33 This is a comparison chart of the separation capabilities of TP homologous peptide dilution induction solution and liquid phase.

[0066] Figure 34 This is a diagram illustrating the energy funnel diagram of TP dilution induction solution-liquid phase separation.

[0067] Figure 35 It is a recruitment map for guest molecules achieved by utilizing the TP dilution-induced enrichment effect;

[0068] Figure 36 It shows the structural formula of a fluorescent small molecule and its fluorescence microscopy image after enrichment.

[0069] Figure 37 This is a diagram showing the guest peptide sequence and its corresponding isoelectric point used to test the enrichment ability induced by TP dilution.

[0070] Figure 38 This is a confocal imaging display and fluorescence distribution analysis of guest polypeptides enriched in TP condensates;

[0071] Figure 39 It is a bar chart of the partition coefficients of guest polypeptides and small molecules;

[0072] Figure 40 This is a trend diagram showing that increasing RNA concentration promotes TP phase separation;

[0073] Figure 41The spatial distribution of TP and RNA condensates with different concentrations of buffer salts is shown.

[0074] Figure 42 These are a schematic diagram of the aggregation-induced emission effect and a diagram illustrating the TP dilution-induced emission effect;

[0075] Figure 43 This is a diagram illustrating the multi-compartmental distribution pattern of eGFP-Gal3 induced by dilution within TP condensates.

[0076] Figure 44 This is a concentration threshold diagram showing the phase separation formed by eGFP-Gal3 itself;

[0077] Figure 45 This is a confocal imaging map showing the multi-compartmental distribution of eGFP-Gal3 within TP condensates induced by dilution.

[0078] Figure 46 This is a diagram of eGFP-Gal3 condensate fusion events within TP condensates;

[0079] Figure 47 This is a 3D reconstruction and fluorescence distribution diagram of a multi-level phase separation system;

[0080] Figure 48 This is a visualization of the secondary compartment fluorescence bleaching of eGFP-Gal3 formed within TP condensates;

[0081] Figure 49 This is a schematic diagram of dilution-induced TP aggregates used as microreactors to accelerate enzymatic reactions;

[0082] Figure 50 This is a DIC and confocal fluorescence microscopy image showing the colocalization of His-Gal-10 and TEV within the condensate;

[0083] Figure 51 The image is a polarized light microscope image revealing that dilution-induced enrichment increases the crystal formation rate. Detailed Implementation

[0084] The following provides definitions for some of the terms used in this specification. Unless otherwise stated, all technical and scientific terms used herein generally have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0085] The invention is further illustrated below with reference to specific embodiments. It should be understood that the specific embodiments described herein are by way of example and are not intended to limit the invention. The main features of the invention can be used in various embodiments without departing from the scope of the invention.

[0086] Example

[0087] 1. Experimental Materials and Methods

[0088] 1.1 Reagents and Materials

[0089] All peptides were purchased from Anhui Guotai Pharmaceutical Co., Ltd. High-performance liquid chromatography (HPLC) and mass spectrometry (MS) analysis showed a sample purity exceeding 98%. Other experimental reagents (analytical grade) were sourced from the following suppliers: Tris, Tricine, phosphates, and TEA were purchased from Sinopharm Group; 1,6-hexanediol, guanidine hydrochloride, and hexafluoroisopropanol were purchased from Macklin; DAPI, ThT, and bovine serum albumin (BSA) were purchased from Solarbio. Total yeast RNA was purchased from Beyotime. All experimental water was ultrapure water treated with Milli-Q.

[0090] 1.2 Concentration Measurement Processing

[0091] 1.21 For polypeptides and proteins containing tryptophan and tyrosine

[0092] First, the lyophilized peptide powder used in the experiment was dissolved in ultrapure water to prepare a high-concentration stock solution. A portion of this stock solution was diluted 1:9 in 6M liquid guanidine hydrochloride. 2 μL of the solution was then used to determine the molar concentration of the peptide using a NanoDrop UV-Vis spectrophotometer (ThermoFisher). Concentration measurements were based on the NanoDrop user manual. According to the Beer-Lambert equation, the absorbance (A) of the peptide at 280 nm is related to the molar extinction coefficient (ε), optical path length (L), and peptide concentration (c), i.e.:

[0093] A=εLc (1)

[0094] The molar extinction coefficient of a polypeptide is calculated based on the sum of the molar extinction coefficients of tryptophan and tyrosine in its sequence at 280 nm:

[0095] ε=(nW*5500)+(nY*1490) (2)

[0096] The molar extinction coefficients of tryptophan and tyrosine were chosen according to the recommendations in the NanoDrop user manual.

[0097] 1.22 For peptides labeled with fluorescent probes

[0098] For TMR-labeled TP, the concentration of TMR-TP was determined by measuring the absorption of the TMR fluorescent molecule at 555 nm. The extinction coefficient of TMR was ε = 90,000 M. -1 cm -1 .

[0099] For FAM-labeled peptides, the concentration of the labeled peptide is determined by measuring the absorption of FAM fluorescent molecules at 493 nm. The extinction coefficient of FAM is ε = 83,000 M. -1 cm -1 .

[0100] 1.23 Nucleic acid sample

[0101] For the quantification of total yeast RNA used in the experiment, NanoDrop was used to measure the RNA detection module.

[0102] 1.3 Induction Liquid-Liquid Phase Separation Process

[0103] All peptides involved in liquid-liquid phase separation were stored and their concentrations determined using a high-concentration stock solution dissolved in pure water. They were then diluted in the target buffer solution as needed for the experiment. For systems where peptide dilution induced liquid-liquid phase separation, the peptide was first diluted at a high concentration in the target buffer to maintain a stable single-phase solution, and then further diluted with the target buffer to achieve phase separation. For experiments requiring fluorescence imaging, 1 mol% TMR-TP was added to the system to characterize the TP.

[0104] 1.4 Drawing of Liquid-Liquid Phase Separation Phase Diagrams

[0105] To construct a phase diagram relating peptide concentration to salt concentration, peptide solutions were added to 48-well BioJane crystal culture plates and experiments were conducted at room temperature (25°C). The plates were pretreated with 2 mg / mL BSA for 30 minutes, followed by three rinses with deionized water to reduce surface sensitivity to protein adhesion and enhance resistance to aggregate wetting, facilitating microscopic observation. To prevent evaporation, buffer solution was added to the reservoirs of the plates, and the wells were covered with a transparent sealing film. Images were captured at each time point using a vertical microscope (Nikon E200).

[0106] 1.5 Light Scattering Measurement

[0107] Static light scattering (SLS) measurements were performed at room temperature (25°C) using a Zetasizer Nano ZS90 (Malvern). The sample cell was illuminated with a 4mW, 633nm laser at a scattering angle of 90°. Measurements were performed after each sample had equilibrated for 30 seconds, with at least three measurements taken. Data were processed using Zetasizer software (Malvern).

[0108] For dynamic light scattering (DLS) measurements, sample preparation is the same as for SLS measurements. However, under certain conditions, the hydrated particle size of the sample is greater than 1 μm, and factors such as gravity affect the correlation between the intensity of the scattered light and the particle diffusion behavior, reducing the accuracy of the algorithm for fitting the particle size. Therefore, the autocorrelation function (ACFs) actually detected in the experiment is used to determine the particle size; a longer decay time indicates a larger particle size.

[0109] 1.6 Detection of Liquid-Liquid Phase Separation Agglomerates

[0110] Eight-well coverslips (Cellvis) used for observing condensates were pre-incubated with 2 mg / ml BSA for 30 min and rinsed three times with deionized water. After equilibration for 5 min, the samples were observed under a DMI8 inverted microscope (Leica) at 63× oil immersion. To facilitate the observation and collection of more large-sized condensate fusion events before the condensates wetted the coverslip due to gravity settling, 200 mM M Triricine was used as a buffer system, with a final peptide concentration of 200 μM. The focus was on the condensate closest to the bottom of the slide, and continuous imaging was performed for 10 min, with each frame lasting 0.5 s. Data processing was performed using LAS X (Leica) measurements to obtain the long axis L during the fusion of two condensates. max With short axis L min The value of eccentricity A = (L) at each time point in each fusion event. max –L min ) / (L max +L min The values ​​were normalized, and the relaxation time was fitted using the One-phase decay method. The final size of the condensate from the fusion event was plotted as a scatter plot with its relaxation time and fitted. The slope of the scatter plot is the inverse capillary rate, which reflects the ratio of the condensate viscosity to the surface tension (η / γ).

[0111] 1.7 Raman Spectroscopy Acquisition

[0112] Raman spectra were acquired using an Alpha 300 R confocal Raman microscope (WITec), with excitation parameters of 532 nm and 60 mW. Raman data were processed using Suite SIX (WITec) and Origin software.

[0113] The method for measuring the concentration of peptides within condensates using Raman spectroscopy was referenced from previous literature. In short, the concentration of phenylalanine side chains at 1005 cm⁻¹ was determined. -1 The characteristic absorption peak at 1005 cm⁻¹ is used as a reference. First, the absorption peak of free phenylalanine at different concentrations is measured at 1005 cm⁻¹. -1 The intensity of the characteristic absorption peak at a specific point was used to construct a standard curve. The intensity of the sample condensate at 1005 cm⁻¹ was measured. -1The Raman peak intensity at a specific point, combined with a standard curve, allows for the calculation of phenylalanine concentration in the condensate, which can then be converted to the peptide concentration. The effectiveness of this method was validated with BSA protein solutions of different concentrations.

[0114] Raman spectroscopy was used to analyze the secondary structure within the condensate, with a reference location at 1600 cm⁻¹. -1 -1700cm -1 Amide I region. Baseline subtraction was performed using Suite SIX software, and secondary structure peaks were determined and fitted using Origin, with the peak position fixed at 1615 cm⁻¹. -1 (Phenylalanine and tyrosine signals), 1657cm -1 (α-helix), 1670cm -1 (β-sheet), 1681cm -1 (disorder).

[0115] 1.8 Circular Dichroism Spectrum

[0116] The secondary structure of the peptide in solution was characterized using a circular dichroism spectroscopy system (Jasco J-1500) at room temperature (25°C). Data acquisition was performed using quartz cuvettes with a path length of 0.1 cm. The peptide was dissolved in 20 mM Tricine solution at pH 7.4. The spectral scan rate was 100 nm / min, with a digital integration time of 1 second and a bandwidth of 2 nm. At least two scans were performed in the 260 nm–190 nm range with a step size of 1 nm. Data processing was performed using the instrument's built-in Spectra Manager software, and graphs were plotted using Origin. The ellipticity (θ, deg. cm) was used as the plotting parameter. 2 .dmol -1 The sample signal is normalized based on concentration, with the unit being )

[0117] 1.9 pH-dependent detection

[0118] Britton–Robinson buffer, with a pH range from 2 to 12, was used as the buffer solvent. This buffer was prepared by mixing 0.04 M acetic acid, 0.04 M phosphate, and 0.04 M boric acid, and the pH was adjusted with 0.2 M sodium hydroxide to obtain buffer solutions at different pH values. The buffer solution was mixed with TP at a volume ratio of 19:1 to bring the final TP concentration to 200 μM, and 20 μL was placed in each well of a 384 flat-bottomed transparent glass plate (Cellvis). After incubation at room temperature for 30 minutes, images of each group were taken under a DMI8 inverted microscope, and the absorbance of each group at 550 nm was measured using a SynergyH1 microplate reader (Biotek) to assess the degree of phase separation.

[0119] 1.10 Fluorescence bleaching recovery experiment

[0120] Fluorescence recovery after bleaching (FRAP) assays were performed using a Leica Stellaris 5 Confocal Microscope (Leica) with a 63× oil immersion lens and the FRAP module in the Las X (Leica) software. 5 μL of a 200 μM TP phase-separated sample containing 1 mol% TMR-TP was placed on a BSA-pretreated 8-well coverslip in 200 mM Tricine buffer and equilibrated for 5 min. The confocal excitation wavelength was 561 nm, and the emission and reception range was 570–620 nm. Peptide aggregates of approximately 5 μm in size were located near the slide surface for FRAP. Three images were taken as controls before quenching. Depending on the requirements of full-, partial-, and half-FRAP, the quenching regions were set as the entire aggregate, a 1 μm diameter circular region at the center of the aggregate, and a semicircular region halfway between the aggregates, respectively. The quenching regions were quenched at 100% power for 560 ms, followed by 1 min of imaging every 280 ms. In addition, a non-fluorescent region was set as the background, and a non-quenched agglomerate adjacent to the quenched agglomerate was set as a reference to ensure that the laser used for imaging did not produce additional fluorescence quenching. For full- and partial-FRAP, the region of origin (ROI) where the fluorescence signal was observed was the quenched region. For half-FRAP, the change in fluorescence intensity of the other half of the agglomerate that was not quenched was also recorded. The processing of the fluorescence recovery curve was as previously reported. In short, it included background subtraction, normalization of the pre-quenching fluorescence of the reference ROI, image plotting, and fitting using the One-phase decay method to obtain the half-recovery time. At least four sets of repeated controls of the agglomerate were used.

[0121] 1.11 Guest molecule recruitment experiment

[0122] The recruitment of fluorescent small molecules, peptides, and proteins by TP condensates is achieved by utilizing the dilution-induced phase separation effect. Simply put, the guest molecules are first mixed with a high concentration of TP in a buffer solution, which is still a homogeneous solution at this stage. Then, the system is diluted with the buffer solution to reduce the TP concentration and achieve phase separation.

[0123] DAPI final concentration 0.5 μM, excitation wavelength 360 nm. ThT final concentration 0.5 μM, excitation wavelength 450 nm. Yeast RNA (Solarbio) final concentration 200 ng / ml, specific tracking fluorescence imaging using SYBR Green II (Thermo Fisher), excitation wavelength 488 nm. FAM-labeled peptide final concentration 1 μM, excitation wavelength 488 nm. EGFP fusion expression protein, excitation wavelength 488 nm.

[0124] 1.12 Protein Expression and Labeling

[0125] Recombinant protein expression was performed according to previously reported literature. The process involved plasmid transformation, antibiotic screening, scale-up culture, sonication, nickel column purification, processing with an AKTA Pure 25 protein purification system (Superdex 75 column), and protein concentration. Since this study only used the recruited macromolecule, the specific steps are not detailed here.

[0126] Fluorescent labeling of the recombinant protein was performed using a protein labeling kit (Alexa Fluor™ 633 and Alexa Fluor™ 488, Thermo Fisher).

[0127] 1.13 Nuclear Magnetic Resonance Spectroscopy

[0128] 2 mM TP solution was dissolved in 20 mM Tricine buffer (pH 7.4, 10% D2O), with 2,2-dimethyl-2-silylpentanesulfonic acid (DSS) used as an internal standard chemical shift reference. NMR experiments were performed at 25 °C using a Bruker Avance 700 MHz NMR spectrometer equipped with four RF channels and a triple resonance cryogenic probe with pulsed field gradient. Standard curves were plotted using myoglobin (17.1 kDa), ovalbumin (43.1 kDa), cytochrome C (12.3 kDa), and carbonic anhydrase (29.0 kDa), with experimental measurements referenced from previously published articles. DOSY experiments employed the pulsed sequence stebpgp1s19, with the diffusion coefficient (Dt) determined by the stimulus-echo method. Water peak suppression was achieved using a 3-9-19 pulsed sequence. The experimental parameters (including diffusion time and pulse gradient field duration) were optimized, and the results were analyzed and fitted using the T1 / T2 module of BrukerTopSpin 3.2 software.

[0129] Chemical shifts of main-chain and side-chain atoms were obtained through two-dimensional HCCH-TOCSY (80 ms mixing time) and 1H-1HNOESY (250 ms mixing time) experiments. 1H-1H NOESY spectra were used to confirm the NOE effect. NMR spectra were processed using NMRPipe and analyzed using NMRView. Parameters calculated using the LBDB structure included the proton spacing confinement from the nuclear NOE and the dihedral angles calculated from chemical shifts using TALOS. The initial structures were generated using the CANDID module of CYANA, and the 20 lowest-energy structures were selected as models. NOE assignment was extended using the SANE procedure. Further calculations were performed on 200 structures using CYANA, and the 100 lowest-energy structures were further refined using AMBER. Finally, the 20 lowest-energy conformations were selected as representative structures and analyzed using MOLMOL and PROCHECK_NMR.

[0130] 2. Experimental Results

[0131] 2.1 Study on the properties of dilution-induced liquid-liquid phase separation

[0132] 2.11TP exhibited anomalous dilution-induced phase separation.

[0133] In preliminary experimental studies, this application unexpectedly discovered a polypeptide TP (SYYRPREEEAIPHPLALTHKMGWLQLLGRMF, SEQ ID NO:1) exhibiting anomalous phase transition kinetics. Figure 1 As shown in Figure A, in a 20 mM tris(hydroxymethyl)glycine (Tricine) buffer system at pH 7.4, the solution remains clear when the TP concentration is above 1 mM; however, when diluted to below the critical concentration (<1 mM), the solution immediately undergoes phase separation, manifested as a significant increase in turbidity.

[0134] Static light scattering experimental data ( Figure 1 B) This phenomenon was further verified: the scattering signal intensity of TP at a concentration of 2 mM was only about 42 kcps, indicating that no large molecular assemblies existed in the solution. Unlike common dispersion systems, the scattering signal of TP gradually increased with dilution, reaching ~86 kcps at 1.25 mM; reaching ~318 kcps at the critical 1 mM; when the concentration dropped to 0.75 mM, the solution became turbid, and the signal intensity surged to ~2470 kcps, indicating the formation of a large number of large assemblies; the maximum signal value of ~6076 kcps was reached at 0.1 mM, after which the signal decreased with dilution.

[0135] This anomalous phase transition effect is reversible. In absorbance detection based on concentration gradient changes, the solution was cyclically switched between 1 mM and 0.75 mM using methods such as "buffer dilution" and "addition of high-concentration TP," and this was confirmed by absorbance detection at 550 nm. Figure 1 C): When the concentration of the phase-separated system rises back to above the critical value, it can be completely reconstituted and restored to a clear state, exhibiting instantaneous response characteristics and complete reversibility.

[0136] Differential interferometry (DIC) microscopy observations provided direct evidence for this anomaly. Figure 2At a TP concentration of 0.5 mM, numerous spherical droplet-like aggregates were visible in the field of view; however, when the concentration increased above a critical value, the aggregates completely disappeared. This "low-concentration aggregation-high-concentration dissolution" phase transition behavior contrasts sharply with the concentration-dependent phase separation (dispersion at low concentrations, aggregation at high concentrations) exhibited by conventional biomacromolecule systems. A systematic literature review revealed that this type of inverse concentration phase transition phenomenon in a single component has not been previously reported in biomacromolecule solution systems. This discovery provides a new perspective for understanding the regulatory mechanisms of phase transitions in biomacromolecules.

[0137] 2.12 Constructing the TP dilution-induced phase separation phase diagram

[0138] Phase diagram analysis is a key method for analyzing phase separation behavior. To systematically study the influence of buffer systems on the phase transition of total phosphorus (TP), four buffer salts with buffering capacities meeting the pH 7.4 condition were selected: N-(Tris(hydroxymethyl)methyl)glycine (Tricine), tris(hydroxymethyl)aminomethane (Tris), triethanolamine (TEA), and phosphate. A concentration-phase diagram showing their correlation with TP was constructed. Figure 3-6 Experimental results show that: First, the dilution-induced phase separation phenomenon is universal under different buffer systems and different buffer salt concentrations, confirming the widespread existence of this anomalous phase transition behavior—when the buffer salt concentration is fixed, gradually increasing the TP concentration shows that the phase states of "homogeneous-phase separation-homogeneous" occur sequentially; Second, when the TP concentration is fixed, increasing the buffer salt concentration (e.g., increasing Tricine from 20 mM to 50 mM) shifts the critical concentration on the right side of phase separation to the right (from 1 mM to 2 mM), meaning that the higher the buffer salt concentration, the more favorable it is for promoting phase separation of TP at high concentrations.

[0139] It is noteworthy that metastable conditions were generally detected in the phase diagram boundary regions of all buffer systems. Figure 3-6(Blue data points). Under these conditions, the condensates formed by sample dilution spontaneously disappeared within 10 minutes of standing, and slight further dilution could induce phase separation again. This kinetic behavior is the opposite of classical liquid-liquid phase separation (LLPS) theory and previous related experimental reports: metastable systems exist between the binodal and spinodal lines, characterized by an initial homogeneous phase that spontaneously separates to reach the lowest free energy upon activation or after a period of standing. However, the observations in this study show that in the TP system, the metastable state, characterized by condensates, actually relaxes back to a homogeneous phase, suggesting that the system may have a special energy barrier distribution—under specific boundary conditions, the homogeneous phase is the more thermodynamically stable final state.

[0140] 2.13 Characterization of the liquid-like properties of TP condensates

[0141] Microscopic observation shows that TP condensates generally exhibit a spherical droplet morphology. Figure 2 It is also prone to wetting with the cover glass substrate. Figure 7 This suggests that it may possess liquid-like characteristics. To systematically analyze its physical properties, this study reveals the following key characteristics through multi-scale property characterization:

[0142] (1) Size regulation mechanism. Based on the positive correlation between buffer salt concentration and phase separation trend in phase diagram studies, ( Figure 3-6 Dynamic light scattering (DLS) was used to quantitatively study the particle size distribution. Given the limitations of the traditional Stokes-Einstein model for particles >1 μm (gravitational settling interfering with Brownian motion), the relaxation time τ of the autocorrelation function (ACF) was used to indirectly compare the relative particle sizes. Figure 8 As shown, the equivalent particle size corresponding to the τ value in the 20 mM Tricine system is <1 μm, while it increases to ~3 μm at 200 mM, confirming that the buffer salt concentration promotes the formation of larger aggregates. Furthermore, when the buffer salt concentration is fixed (Tricine and PBS as examples), changes in TP concentration (0.2-0.5 mM) have no significant effect on the τ value, indicating that the aggregate size is mainly regulated by the buffer salt concentration. Since large aggregates are beneficial for the accuracy and convenience of microscopic observation experiments, a 200 mM Tricine buffer system was chosen for characterizing this part of the physical properties to obtain suitable aggregates for observation.

[0143] (2) Fusion Dynamics and Rheological Properties. The fusion event of condensates is the gold standard for determining whether they possess liquid properties. By tracing the dynamic process from contact to spheroidization of two adjacent condensates of similar size (…), Figure 9 AB), establish the quantitative relationship between fusion time (t) and final radius (R). Figure 9 C). The slope of the fitted data reflects the reverse capillary rate η / γ (viscosity / surface tension) of the system. The η / γ of TP condensates is 68.88±1.84 s / μm, which is 1-2 orders of magnitude higher than that of classic liquid condensates such as Pgranules (2.00 s / μm) and GFP-HP1α (0.35 s / μm), revealing its extremely high viscoelasticity.

[0144] (3) Molecular exchange kinetics. The fluorescence recovery after photobleaching (FRAP) experiment further revealed the intramolecular mobility of the condensates, using a 4 mol% TMR-TP labeling system. Interestingly, the partial bleaching experiment (Part-FRAP) showed that the fluorescence relaxation time τ of the TP condensates was 2.45 ± 0.06 s, and the fluorescence intensity recovered to 40% of its pre-bleaching level during the plateau phase. Figure 10 A). This rapid recovery kinetics significantly contradicts the high viscoelasticity characterized in the fusion experiment. To resolve this phenomenon, full-FRAP and half-FRAP experiments were designed for further verification. Figure 10 As shown in B and C, both the full bleaching and semi-bleaching experiments exhibited kinetic curves similar to those of the partial bleaching experiment, with relaxation times of 2.75±0.1s and 2.48±0.06s, respectively, and recovery rates of approximately 40%. The full bleaching experiment results indicate that the fluorescence recovery is entirely due to molecular exchange with free TP in the external solution environment, and this exchangeable component accounts for approximately 40% of the aggregate. In the semi-bleaching experiment, no significant decrease in fluorescence intensity was detected in the unbleached area (<5%), indicating that the flow within the aggregate is restricted, and the fluorescence recovery in the quenched area also mainly comes from molecular exchange with the external environment.

[0145] Combining fusion events and fluorescence bleaching recovery experiments, TP condensates exhibit unique kinetic heterogeneity—a coexistence of surface tension-driven liquid-like macroscopic behaviors (such as spheroidization fusion) and restricted molecular mobility. This characteristic may stem from a continuous density gradient from the outside to the inside: a loose outer layer (contributing ~40% of the fluorescence signal) enables rapid exchange (τ ~2.5 s), while a high-density core restricts molecular motion.

[0146] 2.14 Raman spectroscopy characterizes the properties of TP condensates

[0147] This study employed high-resolution Raman confocal microscopy for in-situ detection of TP condensates. First, the TP condensates and their surrounding solution were analyzed within a range of 400–3800 cm⁻¹. -1 Raman broad-spectrum scanning. For example... Figure 11As shown, there are significant differences in the spectral signal distribution inside and outside the condensate. These differences are particularly pronounced in the following regions:

[0148] 1002cm -1 The peak corresponds to the respiratory peak of phenylalanine (Phe) and is related to the C-S stretching vibration of the benzene ring. This signal is not easily affected by protein structure and is suitable for quantifying protein concentration.

[0149] 1220cm -1 -1400cm -1 The region is the amide III region, which is mainly generated by the combined action of NH bending vibration and CN stretching vibration, and can be used to identify secondary structures;

[0150] 1450cm -1 The peaks are usually associated with the bending vibration modes of CH2 and CH3;

[0151] 1600cm -1 Up to 1700cm -1 The region is the amide I region, which is mainly related to C=O stretching vibrations and is suitable for the analysis of secondary structures.

[0152] 2800–3100cm -1 The region is typically associated with CH stretching vibrations in fatty acid chains and alkyl chains;

[0153] 3100–3700cm -1 The region is associated with NH and OH stretching vibrations (hydrogen bonds in amino acids and proteins). Because the OH stretching vibration signal contributed by water molecules is very strong, the majority of the signal in aqueous solution samples is usually contributed by water molecules.

[0154] By comparing the heights of characteristic protein signal peaks, it can be roughly inferred that the interior of TP condensates is an extremely concentrated and crowded environment. Furthermore, 3100–3700 cm⁻¹ -1 The difference in signal area within the region also supports this conclusion: the signal area in this region within the condensate is less than 50% of that in the external solution environment. Considering that the peptides within the condensate also contribute a portion of the signal in this region, it means that the water molecule content inside the condensate is significantly lower than that outside. This phenomenon is significantly different from previous Raman spectroscopy studies of protein condensates, where the signal difference between the inside and outside of the condensate is usually small, revealing that TP condensates have unique dehydration and densification characteristics.

[0155] Furthermore, by constructing a standard curve between phenylalanine concentration and respiratory peak signal intensity, the concentration of peptides within the condensate was quantified. First, standard curves (R0, R0, R0) were plotted by measuring phenylalanine solutions at 10 mM, 20 mM, and 36 mM. 2=0.999), and verified using 0.3mM, 0.6mM, and 1.2mM BSA solutions, confirming that the phenylalanine signal accurately reflects protein concentration ( ). Figure 12 A–C). Then, the TP condensate (N=4) was heated at 1002 cm⁻¹. -1 Substituting the signal intensity at that location into the standard curve, and based on the number of phenylalanine residues in TP (n=1), the peptide concentration inside the aggregate was found to be as high as 230±30 mM. Figure 12 D) is more than 20 times the concentration at the right boundary of phase separation under the detection conditions (200 mM Tricine, CT concentration of 10 mM). Figure 13 This result indicates that the concentration of the TP rephase (CD) is significantly higher than that of the CT, and this is corroborated by the extremely high viscoelasticity exhibited in the fusion event (η / γ = 68.88 ± 1.84 s / μm), suggesting that the phase separation behavior of TP is significantly different from the classical LLPS model.

[0156] Furthermore, by analyzing the area at 1600cm -1 Up to 1700cm -1 The amide I region peak diagram allows for further exploration of the secondary structure of peptides within condensates. This can be achieved by analyzing the peaks at 1615 cm⁻¹. -1 (Phenylalanine and tyrosine), 1657cm -1 (α-helix), 1670cm -1 (β-fold) and 1681cm -1 Gaussian fitting of the (random structure) peak positions showed that the α-helical conformation dominated within the condensate, accounting for approximately 76.5% of the conformations, followed by the random structure, accounting for 18.7%. Figure 14 Unlike most proteins in condensates that tend to conform to random patterns, polypeptides in TP condensates tend to exist in α-helical form. This phase separation pattern, dominated by rigid secondary structures, may originate from the directional stacking effect of amphiphilic α-helical conformations in the TP sequence.

[0157] 2.15 Establishment of a dilution-induced liquid-liquid phase separation model

[0158] Based on the above experimental results, although the phase separation behavior of TP conforms to the macroscopic characteristics of liquid-liquid phase separation (LLPS) (such as spherical shape and surface tension-dominated deformation), its concentration effect differs fundamentally from the classical LLPS model. Therefore, a phase transition kinetic model of TP based on the classical LLPS model is constructed and called "dilution-induced liquid-liquid phase separation (DILLPS)".

[0159] like Figure 15As shown, unlike the classic LLPS model, the right boundary concentration CT of phase separation in the DILLPS model is much lower than the heavy phase concentration CD. Beyond this concentration, the two phases will revert to a homogeneous state. Because the peptide concentration between CT and CD is too high for experimental observation (currently, the highest observed concentration is 30 mM TP as a clear homogeneous phase), there is no sufficient evidence to fully demonstrate that the TP concentration increases linearly in the 3→4 interval, or that there exists a secondary phase separation state with a predominantly heavy phase and a secondary light phase. Therefore, this is represented by the dashed square region, and the system eventually enters a homogeneous state of heavy phase.

[0160] 2.2 Study on the molecular mechanism of dilution-induced liquid-liquid phase separation

[0161] 2.21 Prediction of Phase Separation Trend in TP

[0162] This application attempts to assess the phase separation trend and key sites of TP using existing computational prediction tools. First, AlphaFold3 was used to predict the TP monomer structure, and the results showed that TP exhibits N-terminal random coil and C-terminal α-helix characteristics, with an overall confidence level pIDDT > 70%. Figure 16 Since the conformation of biomacromolecules in solution often deviates from the predicted lowest energy state, this prediction can only be used as a reference to reveal the differences in the tendency of different regions on the sequence to form stable secondary structures and disordered conformations.

[0163] Furthermore, the TP sequence was analyzed using protein phase separation prediction tools such as IUPred3 and PLAAC. The prediction results of each evaluation method are as follows ( Figure 17 ,18):

[0164] IUPred3 long disorder

[0165] Predicting longer disordered regions (>30 amino acids) is suitable for identifying proteins with overall disorder or long fragments of disorder, which are often associated with LLPS. TP prediction results did not show fragments with scores >0.5, suggesting a low LLPS tendency.

[0166] IUPred3 short disorder

[0167] Predicting short disordered regions (<30 amino acids) is applicable to identifying locally disordered regions within an overall ordered protein, such as binding sites or flexible domains. Results show that TP exhibits some flexible regions at its N-terminus and C-terminus, while the central region is relatively rigid.

[0168] ANCHOR2

[0169] Potential binding sites in disordered regions are predicted. A score >0.5 indicates that the region is disordered in its free state, but may undergo conformational changes (such as folding into an α-helix or β-sheet) upon target binding. TP prediction results show that the whole sequence score is <0.5, indicating no obvious binding tendency and a low probability of LLPS.

[0170] IUPred3 structural domains

[0171] Identify ordered structural domains (i.e., regions that fold into stable structures) in the protein. A TP full-sequence score <0.5 indicates that it may be relatively stable in its monomeric state and is unlikely to directly drive LLPS.

[0172] PrD-like (Prion Domain-like Score)

[0173] The protein was assessed for prion-like domain (PrLD) characteristics; a higher value (closer to 1) indicated a greater tendency to form reversible aggregates, thus promoting LLPS. TP did not exhibit prion-like features.

[0174] PLAAC Score

[0175] By combining the results of PrD-like and Hidden Markov Model (HMM) calculations, we assessed whether the protein might rely on PrLD for phase separation. A low PLAAC score (<0.5) suggests that TP may not rely on a prion-like mechanism, but may involve other interactions (such as electrostatics, hydrophobicity, and π-π stacking) to drive phase separation.

[0176] 4*PAPA Score

[0177] The ability of a protein to form reversible aggregates was assessed using the PAPA (Prion Aggregation Prediction Algorithm). Higher values ​​indicate a stronger aggregation tendency and a higher correlation with LLPS proteins. TP showed a high aggregation tendency at its N-terminus, but its overall score was still lower than that of typical LLPS proteins.

[0178] Fold Index

[0179] This predicts whether a protein is more likely to fold into a stable structure or remain in a disordered state. A Fold index > 0 indicates that the protein is more likely to fold into a stable three-dimensional conformation, while a Fold index < 0 indicates stronger disorder and a greater tendency to participate in LLPS. The TP prediction results are similar to those of AlphaFold3, with the C-terminus tending to form a stable structure, while the N-terminus is more flexible.

[0180] In summary, TP exhibits both ordered and disordered characteristics, but computational predictions do not support a strong LLPS trend, whether based on IDR predictions or prion-like features. The C-terminus of TP shows higher rigidity, while the N-terminus is more flexible, potentially harboring binding sites. Therefore, its dilution-induced phase separation phenomenon may not be entirely dependent on IDR or PrLD, failing to conform to common protein phase separation driving mechanisms, and may involve more complex intermolecular interactions, environmental factors, and dynamic conformational changes.

[0181] 2.22 Nuclear magnetic resonance spectroscopy reveals the homogeneous conformation of TP at high concentrations.

[0182] This application selects NMR technology to perform structural analysis on the high-concentration homogeneous state of TP in order to obtain detailed conformational information of TP. In this study, 20 mM Tricine buffer was used, and the TP concentration was adjusted to 2 mM to ensure that the sample was in a high-concentration homogeneous state, which is conducive to the accurate acquisition of NMR signals.

[0183] This application employs diffusion-ordered spectroscopy (DOSY), a technique that infers the apparent molecular weight and conformational state of molecules by measuring their diffusion coefficients, thus providing reliable data support for resolving the TP monomer structure. As shown in Table 1, the apparent molecular weight corresponding to the TP diffusion coefficient obtained through DOSY measurement is approximately 5.2 kDa, indicating that under the selected conditions, TP mainly exists in monomeric form. Furthermore, analysis using one-dimensional 1H NMR spectroscopy (Table 2) revealed that TP in a homogeneous state lacks secondary structure and exhibits a compact, coiled, spherical conformation. Figure 19 A, PDB ID: 9KVR). More detailed spectral analysis revealed two pairs of electrostatically interacting arginine (Arg)-glutamic acid (Glu) residues on the surface of the TP molecule. Figure 19 B), while most hydrophobic side chains tend to concentrate inside the molecule. Figure 19 C (highlighted in blue). This characteristic of internal hydrophobicity and surface electrostatic interaction is likely a key factor in maintaining the compact conformation of the molecule, preventing the aggregation of TP molecules, and thus inhibiting or reversing LLPS.

[0184] In summary, by utilizing NMR and DOSY techniques, we not only clarified that TP exists primarily in monomeric form under high-concentration homogeneous conditions, but also revealed its unique conformational characteristics and intramolecular interaction patterns. This provides solid experimental evidence for further exploring the molecular mechanism of TP dilution-induced LLPS and lays the foundation for resolving the details of key interactions and driving forces in subsequent studies.

[0185] Table 1. Diffusion ordering spectral data of TP and standard proteins of different molecular weights.

[0186]

[0187]

[0188] Table 2. Statistical analysis of NMR structure calculation data for TP samples.

[0189]

[0190] 2.23TP conformation changes towards α-helix with concentration dilution.

[0191] To investigate the conformational changes of TP at different concentrations and their relationship with phase separation behavior, circular dichroism (CD) spectroscopy was used to measure the conformational changes of TP during the stepwise dilution from a high concentration homogeneous phase to the occurrence of phase separation. Four concentration gradients were set up, with the solution remaining homogeneous at the first three concentrations, while at the lowest concentration (0.5 mM), it was in the LLPS state. Figure 20 The results showed that at 2.0 mM, TP exhibited a weak α-helical trend, manifested as characteristic negative peaks at 208 nm and 222 nm. Simultaneously, no typical random coil signal (the approximately 195 nm negative peak formed due to exposure of the main chain solvent and the absence of a fixed hydrogen bonding mode) was observed, which is consistent with NMR observations—while TP lacks a clear secondary structure, it is not entirely in a loose random coil state. As the concentration decreased, the α-helical conformation signal in the solution gradually increased, reaching a peak at 0.5 mM under phase separation conditions (note that during LLPS, due to solution turbidity, coexistence of two phases, laser scattering, and signal obstruction, some signal loss may occur). Combined with the NMR structural data, it can be concluded that under high-concentration homogeneous conditions, TP mainly exists in a relatively compact, spherical conformation, while with dilution, the TP conformation gradually transforms into an extended α-helical state.

[0192] To further verify the correlation between the α-helical conformation and TP phase separation, 2,2,2-trifluoroethanol (TFE), a reagent that promotes α-helix formation, was added to the solution. Experimental results showed that the addition of TFE significantly enhanced the phase separation tendency of TP: under conditions that were originally high-concentration homogeneous, obvious aggregates appeared under the induction of 5% TFE. Figure 21 This further demonstrates the crucial role of α-helical conformational transformation in the TP phase separation process.

[0193] 2.24 Study on key sequence characteristics for maintaining the compact and coiled state of monomers under high TP concentration

[0194] Based on the aforementioned structural studies, it was found that TP tends to form compact coiled monomers that are more conducive to intramolecular interactions under high concentration conditions, while it gradually transforms into an α-helix conformation at low concentrations, enhancing the potential for intermolecular interactions. This concentration-dependent allosteric response effect is undoubtedly achieved through a special sequence arrangement, causing the α-helix and compact coiled conformations to exhibit an "antagonistic tug-of-war" with changes in concentration, resulting in this anomalous phase transition kinetics effect. This leads to three questions: (1) Which sequence features are conducive to maintaining the compact coiled monomer state at high TP concentrations? (2) What are the driving forces for condensate formation? What dynamic changes occur during the allosteric process? (3) Why does a dilution-induced allosteric effect exist? The following sections will analyze these three questions in turn.

[0195] Previous studies have shown that proline (Pro) typically disrupts the formation of the α-helix, and the TP sequence contains three Pro residues. Nuclear magnetic resonance (NMR) structural analysis reveals that the positions of these three Pro residues coincide with a significant torsion of the main chain. Figure 19 C (marked in magenta) suggests that they may play a key role in maintaining a compact conformation and intramolecular interactions at high TP concentrations. To verify this hypothesis, a mutant TP-3A was designed and synthesized, in which all three Pros were replaced with alanine (Ala), which is conducive to α-helix formation. The structural prediction results of TP-3A in AlphaFold3 also support the possibility of a more α-helix conformation. Figure 22 ).

[0196] In the DILLPS effect assay, TP-3A exhibited phase transition kinetics significantly different from the prototype TP. Specifically, in 20 mM Tricine buffer, the prototype TP remained clear and homogeneous at concentrations exceeding 1 mM; while TP-3A solutions remained turbid even at concentrations up to 6 mM, and the formation of large assemblies was immediately observed upon mixing with the solvent. Figure 23 A). Furthermore, in pure water, TP-3A initially appears as a clear, homogeneous solution, but after standing at room temperature for 30 minutes, the solution becomes cloudy, and large assemblies are visible under an optical microscope. Figure 23 B) This indicates that it is thermodynamically more inclined towards intermolecular assembly. In contrast, the prototype TP does not undergo phase separation in pure water; a certain concentration of buffer salt is a necessary condition for its phase separation. Overall, TP-3A loses the DILLPS effect and exhibits a strong tendency for intermolecular aggregation.

[0197] Furthermore, based on NMR structural analysis ( Figure 19(B) The two pairs of salt bridges on the surface of the TP molecule may be related to the main chain conformational orientation, providing enhanced stability for maintaining its compact coiled conformation. To verify this hypothesis, a mutant TP-2E was designed and synthesized, replacing the two Arg residues with Glu. The purpose was twofold: first, to disrupt the original electrostatic effect, and second, to minimize changes to the hydrophilicity / hydrophobicity and secondary structure tendency of the original residues. In the DILLPS effect test, although TP-2E still retained the ability to induce phase separation through dilution, its critical concentration for achieving a high-concentration homogeneous phase was significantly increased. Figure 24 In 20 mM Tricine buffer, the critical concentration increased from 1 mM for the prototype TP to 4 mM. Notably, the TP-2E aggregates exhibited a large and irregular assembly, with only a few spherical droplets visible in the field of view. This further reflects the crucial role of electrostatic interactions in regulating the physical properties of TP aggregates.

[0198] In summary, these results indicate that the compact coil conformation of TP monomers at high concentrations depends on a specific sequence arrangement. Proline plays a crucial role in maintaining this compact conformation by blocking α-helix formation, while electrostatic interaction residues on the molecular surface contribute to stabilizing this conformation. Mutations in Pro→Ala and Arg→Glu significantly alter the phase behavior of TP in solution, providing solid experimental evidence for a deeper understanding of the molecular mechanism of TP dilution-induced LLPS.

[0199] 2.25 TP Low Concentration: Key Driving Factors for Aggregate Formation

[0200] In high-concentration homogeneous environments, TP maintains a compact coiled conformation due to a specific sequence arrangement. However, with concentration dilution, the conformation gradually shifts towards an α-helix. It is foreseeable that this conformational transition will be accompanied by the gradual exposure of key interaction sites driving phase separation, leading to a reorganization of the original intramolecular interaction patterns. This, in turn, promotes intermolecular interactions and induces LLPS formation. Therefore, identifying which interactions provide the key driving forces for TP condensate formation and how these interactions work synergistically is fundamental to further understanding the TP dilution-induced phase separation effect.

[0201] Existing literature shows that electrostatic interactions, hydrophobic interactions, and π–π and cation–π interactions play crucial roles in the self-assembly of peptides and proteins. Electrostatic attraction and repulsion can regulate the initial aggregation between molecules; while hydrophobic interactions promote the formation of aggregates by encouraging hydrophobic side chains to approach each other and lowering the system's free energy; simultaneously, π–π stacking and cation–π interactions provide additional stability to multivalent interaction networks, further strengthening intermolecular binding. Based on this, this study will systematically analyze hydrophobic interactions, π–π / cation–π interactions, and electrostatic interactions in turn.

[0202] First, 1,6-Hexanediol (1,6-HD) is a commonly used reagent for studying the effects of hydrophobic interactions in LLPS. Its hydrophobic segments can insert into the hydrophobic regions of proteins or peptides, thereby interfering with and weakening these hydrophobic interactions, leading to the dissolution or disappearance of droplets or aggregates. However, it has a smaller impact on LLPS systems that do not depend on hydrophobic interactions. Therefore, it is suitable for identifying whether hydrophobic interactions are a key driver of aggregate formation. Figure 25 As shown, the addition of 5% 1,6-HD dissolved most TP condensates, and after increasing the concentration to 10%, all condensates disappeared from the field of view, suggesting that hydrophobic interactions play a significant role in TP condensate formation. It is worth noting that, based on nuclear magnetic resonance structural analysis (… Figure 19 C) In the high-concentration monomer state, most hydrophobic residues are located at the core of the spheroidal conformation; therefore, the transition to the α-helical conformation will gradually expose these hydrophobic residues, thereby transforming the hydrophobic effect, which was originally dominated by intramolecular interactions, into a mode that is more conducive to intermolecular interactions.

[0203] Secondly, regarding π–π stacking and cation–π interactions, TP contains four aromatic amino acids with potential interaction capabilities. Based on the NMR conformation in the high-concentration monomeric state ( Figure 19 C) The two tyrosine (Tyr) side chains are partially exposed to the solvent and should be less affected by conformational changes; the tryptophan (Trp) and phenylalanine side chains are located in the hydrophobic core region and may be more affected by conformational changes. To focus on the impact of conformational changes on potential interaction modes, a mutant TP-2A was designed, replacing the tryptophan and tyrosine in the hydrophobic core with alanine, a non-aromatic amino acid with similar hydrophobicity. In the DILLPS test, TP-2A completely lost its phase separation ability, and even in 20 mM Tricine buffer (… Figure 26 Under various buffer conditions, including Tris and phosphate, and even after the addition of the α-helical inducer TFE, no phase separation response was observed, suggesting that π–π stacking or cation–π interaction plays a key driving role in TP phase separation. In addition, the solvent exposure caused by the conformational change of the Trp and Phe side chains located in the hydrophobic core is of great significance for the formation of aggregates.

[0204] Finally, regarding the contribution of electrostatic interaction, the phase separation of TP was tested for pH dependence using Britton–Robinson buffer with a buffer range of 3–12 as background. The results showed that the phase separation trend of TP was most obvious in the pH range of 7–10. Figure 27However, at pH 3-5 and 12, the aggregates completely disappear. Since TP contains both acidic and basic amino acids, its theoretical isoelectric point (pI) is approximately 8.2. This phenomenon, where phase separation tends to increase as pH approaches pI, points to two opposing yet unified explanations:

[0205] (1) The reduction of long-range electrostatic repulsion. When the pH of the solution is close to pI, the net charge of the protein approaches zero, and the long-range electrostatic repulsion between molecules is significantly reduced. This reduction in repulsion makes it easier for molecules to approach each other, and thus short-range interactions such as hydrophobicity and π–π / cation–π between molecules are more likely to form, suggesting that electrostatic interaction is unfavorable to the occurrence of phase separation.

[0206] (2) Enhancement of local short-range electrostatic attraction. Near pI, the full ionization of local charges (deprotonation of carboxyl groups of acidic residues and protonation of amino groups of basic residues) maximizes the dipole strength, which can contribute to the formation of condensates by providing intermolecular salt bridges, suggesting that electrostatic interaction is conducive to the occurrence of phase separation.

[0207] These two explanations are essentially two sides of the same coin regarding electrostatic interactions; they are not contradictory and reflect the complexity of electrostatic interactions in driving phase separation. Which effect is dominant requires further investigation in conjunction with ionic effects. It is noteworthy that in the pH-phase separation correlation diagram, pH 8 serves as the dividing line. Below pH 8, aggregates are prone to wetting, while above pH 8, aggregates tend to maintain a droplet-like morphology. This may be related to the pre-coating of the quartz glass slide with BSA (pI≈4.7) to form a protein film. When the solution pH is greater than 4.7, the protein film carries a negative charge. If the solution pH is also lower than the isoelectric point of TP (8.2), TP carries a positive charge, resulting in an electropositive aggregate surface that readily forms electrostatic adsorption with the protein film. Conversely, TP carries a negative charge, resulting in an electronegative aggregate surface that electrostatically repels the underlying protein film, preventing wetting.

[0208] In summary, during dilution, the compact coil conformation of TP gradually transforms into a relatively extended α-helix, accompanied by the exposure of key hydrophobic residues and π–π stacking / cation–π interaction sites, thereby enhancing intermolecular interactions and driving the occurrence of LLPS. The electrostatic interaction mechanism is more complex and will be discussed in detail in the next subsection in conjunction with ionic effects.

[0209] 2.26 Discussion on the mechanism of dilution-induced conformational change

[0210] According to literature review, reports on protein / peptide conformational changes with concentration are scarce, and systematic theoretical and experimental verification is lacking. Existing research mostly focuses on the transformation from monomers to aggregates as protein / peptide concentration increases. This process is often accompanied by intermolecular conformational induction, leading to a shift from random conformations to ordered secondary structures. For example, as peptide concentration increases, melittin transforms from a monomer to a tetramer with an α-helical conformation; after reaching a certain concentration threshold, amyloid-β and α-synuclein spontaneously assemble into β-sheet structures, forming amyloid aggregates. Most proteins exhibiting the LLPS effect are exceptions; when their concentration exceeds the phase separation threshold, they lose their original stable conformation and transform into a random conformation, thereby exposing multivalent binding sites and promoting intermolecular assembly to form aggregates. The common logic behind these phenomena is that increased protein / peptide concentration shortens the intermolecular distance, increasing the frequency of intermolecular collisions, making intermolecular interactions dominant, and leading to the synergistic occurrence of conformational changes and molecular aggregation. However, for TP, the decrease in peptide concentration and increase in intermolecular distance during dilution seem intuitively unfavorable for intermolecular interactions, which contradicts common sense. Currently, there is no mature theory that can fully explain this dilution-induced allosteric or aggregation effect. Therefore, this section will discuss this effect comprehensively by combining experimental phenomena with existing theories.

[0211] First, from a macroscopic perspective, the only change before and after dilution is the concentration of TP itself, while the overall solvent environment (including water structure, ion types and concentrations, etc.) remains unchanged. Therefore, it is necessary to consider whether long-range intermolecular interactions in the solution, especially the electrostatic interactions between peptide molecules, play a key role in the conformational change. Since TP has an isoelectric point of 8.2, in a solution with pH 7.4, the molecules are positively charged as a whole, thus exhibiting intermolecular electrostatic repulsion. One possible hypothesis is that at high concentrations, the intermolecular electrostatic repulsion network restricts the conformational adjustment of TP, contributing to the compact coiled monomer conformation; however, after dilution, the increased intermolecular distance exceeds the electrostatic shielding range, the electrostatic repulsion disappears, and thus the original compact coiled state cannot be maintained, leading to the conformational change.

[0212] If the hypothesis is correct, then the phase separation tendency observed in the experiment should be negatively correlated with the electrostatic shielding distance; that is, weakening the Debye length can enhance the phase separation tendency. This should be analyzed in conjunction with phase diagrams of different buffer systems previously studied. Figure 28A) It was found that monovalent organic buffer salts (Tris, Tricine, TEA) showed a generally consistent trend in their influence on phase separation, while divalent inorganic salts (phosphates) were significantly better than the former three. The Debye distance for each of the four buffer salts at various concentrations was calculated as follows: The ion dissociation forms of each buffer salt at pH 7.4 were calculated using the Henderson–Hasselbalch formula:

[0213]

[0214] The results showed that at pH 7.4, protonated Tris and TEA accounted for 83% and 71%, respectively; phosphate existed as 40% NaH2PO4 and 60% Na2HPO4; Tricine was unique, as it had two pKa values, ~2.3 (corresponding to the carboxyl group) and ~8.15 (corresponding to the amino group), and at pH 7.4, 85% of it existed as zwitterions (overall neutral), with only 15% being unprotonated and negatively charged.

[0215] The ionic strength of each solvent is calculated based on its ionization form, according to the ionic strength calculation formula:

[0216]

[0217] Where c i is the molar concentration (M) of ion i, and z is the charge number of the ion.

[0218] The effective ionic strength provided by the four buffer salts as a function of their concentrations (C) was calculated to be: Tris: 0.915C, Tricine: 0.075C, TEA: 0.85C, and phosphate: 1.7C. The electrostatic shielding distance in each buffer solution was then calculated using the Debye formula.

[0219]

[0220] Where λ D ε is the Debye distance; I is the ionic strength; ε r ε0 is the relative permittivity of the solvent; ε0 is the vacuum permittivity; k B is Boltzmann constant; T is absolute temperature; e is elementary charge.

[0221] The electrostatic shielding distances of each buffer salt at 10, 20, 50, and 200 mM were calculated, and the results are shown in Table 3. Phosphate showed the highest ionic strength and the most significant electrostatic shielding effect, consistent with the phase diagram trend. Tris was similar to TEA; Tricine was the weakest, with its Debye distance (2.48 nm) at 200 mM only approaching that of Tris (2.25 nm) at 20 mM. Since Tricine and Tris are similar in structure and buffering capacity, but offer significantly different ionic strengths, they are suitable for comparison. If the conformational change during dilution is primarily driven by the disappearance of long-range electrostatic repulsion, then Tricine's ability to promote phase separation should be significantly lower than that of Tris. For example, the TP phase separation trend at 200 mM Tricine should be close to that at 20 mM Tris. However, considering the phase diagram... Figure 28 A) Tricine and Tris have similar overall effects on the TP phase separation trend, even at the same 20 mM, when Tricine (λ) D The phase separation trend in (≈7.84nm) is slightly stronger than that in Tris(λ) D The result (≈2.25 nm) clearly does not support a simple explanation of electrostatic repulsion. It also suggests that the key factor promoting phase separation by increasing buffer salt concentration is not simply an increase in ionic strength, but may be related to other properties of the buffer salt itself. For example, Tricine is mainly in zwitterionic form (85%) at pH 7.4. These neutral molecules contribute nothing to ionic strength, but may affect peptide conformation through hydrogen bonding, electrostatics, hydrophobic interactions, or van der Waals forces. This contribution may be much stronger than the effect of ionic strength.

[0222] Since adjusting the pH of the buffer salt inevitably introduces additional neutral salts in specific experiments, causing the ionic strength to deviate from the theoretical value, an experiment was designed to further verify the above conclusions and specifically investigate the influence of ionic strength and ion type on the phase separation trend of TP. A 10mM Tris solution at pH 7.4 was selected as the background buffer solvent because TP at any concentration does not undergo phase separation under this condition. Based on this, monovalent neutral salt sodium chloride, divalent cationic neutral salt magnesium chloride, and divalent anionic neutral salt sodium sulfate were added at different concentration gradients, and their effects on the TP DILLPS effect were tested. The results were then compiled into a phase diagram. Figure 28 A, Figures 29-31The ionic strength and Debye distance for each neutral salt are shown in Table 4. The results show that the promoting effect of the three salts on TP phase separation is ranked as follows: sodium sulfate > magnesium chloride > sodium chloride. Notably, under all sodium chloride concentrations, TP failed to form stable aggregates (even when metastable phase separation occurred), indicating that TP in solution thermodynamically still tends towards a homogeneous form, suggesting that the phase separation ability of TP is not sensitive to simple changes in ionic strength. The difference in the effects of magnesium chloride and sodium sulfate further illustrates that the type of salt ion has a more significant impact on TP phase separation than ionic strength. Therefore, the macroscopic intermolecular long-range electrostatic repulsion network is not the key factor driving conformational changes; further exploration of potential driving forces from more interaction perspectives, combined with other theoretical models, is needed.

[0223] Furthermore, the above work clarifies that long-range intermolecular electrostatic interactions do not significantly affect TP phase separation. Therefore, the phenomenon in the previous section that "the closer to the pI value, the stronger the phase separation trend" is more likely to be explained by "enhanced local short-range electrostatic attraction," that is, the participation of electrostatic interactions is conducive to the formation and stability of condensates.

[0224] On the other hand, from a microscopic perspective, the ratio of peptide molecules to salt ions in the solution changes before and after dilution: as the concentration decreases, each peptide molecule has the potential to "acquire" more salt ions. Referring to the Counterion Atmosphere Model (CAM), ions are not uniformly distributed in the solution, but are concentrated around charged molecules, thus playing a greater role in shielding local charges. The core mechanisms of this model include: (1) Electrostatic shielding: Counterions reduce the strength of electrostatic repulsion or attraction by neutralizing the surface charge of peptides. (2) Ion concentration gradient: Counterions form a non-uniform distribution (following the Boltzmann distribution) around charged molecules, affecting the local dielectric environment and intermolecular forces. The Counterion Atmosphere Model has wide applications in the study of nucleic acids, especially in describing how nucleic acid molecules interact with ions in solution and how ions affect the structure and stability of nucleic acids. However, its research methods are relatively limited, generally small-angle X-ray scattering (SAXS) and molecular dynamics (MD) simulations. For proteins / peptides, verification is difficult due to their relatively low charge density, but some reports have shown that counterions can indeed affect the secondary structure of peptides.

[0225] In summary, based on the counterion atmosphere model, it is deduced that during the dilution-induced phase separation process, the increased accumulation of counterions around TP molecules (e.g., under 20 mM Tricine conditions, the molar ratio of TP to Tricine increases from 1:10 to 1:20 when diluted from 2 mM to 1 mM) may lead to sufficient occupancy of key interaction sites, resulting in the following changes:

[0226] (1) Dynamic balance of local charge environment: The number of counterions around each TP molecule increases, which neutralizes the surface positive charge by directly binding and weakens the intermolecular repulsion of like charges.

[0227] (2) Entropy effect of hydration competition: The hydration capacity of counterions allows them to compete with polar groups of TP (such as hydroxyl and amino groups) for water molecules, disrupting the hydration layer structure on the TP surface and affecting the conformation of the polypeptide. The released bound water molecules lead to an increase in the entropy of the system, providing a thermodynamic driving force for phase separation.

[0228] (3) Conformational flexibility and aggregation synergy: Local electrical neutrality and disruption of the hydration layer cause the TP conformation to change from a compact coiled state to an extended state, exposing hidden hydrophobic or polar interaction sites (such as α-helical hydrogen bond networks), thereby promoting molecular aggregation and phase separation.

[0229] Considering that TP exhibits insensitivity to ionic strength (e.g., under 200mM sodium chloride conditions, the Debye distance is about 0.66nm, yet phase separation still cannot occur stably), it suggests that the weakening of intermolecular electrostatic repulsion due to local charge neutralization is not the main reason for conformational change. However, this counterion salt bridge can assist salt ions in residing on the surface of peptides, thereby achieving conformational regulation at the hydration level.

[0230] Based on experimental data, the phase diagram shows that the promoting ability of various salts for TP LLPS is ranked as follows: phosphate ≈ sulfate > Tris ≈ Tricine ≈ TEA > magnesium chloride > sodium chloride.

[0231] Based on the anti-ion atmosphere model, by integrating ion retention capacity, hydration disturbance effect and conformational co-transformation, a self-consistent theoretical framework can be provided to explain the "compact coil-α-helix" conformational transformation and phase separation phenomenon of TP at low concentrations.

[0232] Table 3. Ionic strength and Debye distance of four buffer solutions at different concentrations

[0233]

[0234]

[0235] Table 4. Ionic strength and Debye distance of three neutral salts at different concentrations

[0236]

[0237] 2.27 Detection of the DILLPS effect of TP homologous peptides

[0238] Homology comparison revealed a unique L11I mutation in the TP sequence of hominids (Homo sapiens) compared to other mammals. Following comprehensive comparison, representative species with homologous sequences, including rhesus monkeys (Macaca mulatta), house mice (Mus musculus), wild boars (Sus scrofa), and Egyptian fruit bats (Rousettus aegyptiacus), were selected to synthesize TP homologous peptides, and their respective dillps-producing abilities were tested. Figure 32 Based on their characteristics, the homologous peptides were named TP-L (L11I reversion mutation, sequence: SYYRPREEEALPHPLALTHKMGWLQLLGRMF, SEQ ID NO:2), TP-VL (L11I, E8V, G22S, sequence: SYYRPREVEALPHPLALTHKMGWLQLLGRMF, SEQ ID NO:3), TP-KL (L11I, E8K, sequence: SYYRPREKEALPHPLALTHKMGWLQLLGRMF, SEQ ID NO:4), and TP-M (rat-derived, sequence: SYYRPREDEDLPHPLALTHKMSWLQLLGRMF, SEQ ID NO:5).

[0239] DILLPS effect detection indicates that ( Figure 33 ):

[0240] TP-L: Retained dilution-induced phase separation ability comparable to wild-type TP, confirming that the L11I mutation has no significant effect on function;

[0241] TP-VL: The mutation of glutamic acid to hydrophobic valine (E8V) resulted in a small number of aggregates, but retained the DILLPS effect, suggesting that acidic residues have a promoting effect on phase separation.

[0242] TP-KL: The mutation of glutamic acid to lysine (E8K) triggers charge reversal and completely loses the ability to separate phases. Combined with the results of TP-L and TP-VL, it is suggested that Glu8 may provide a key electrostatic adsorption effect on the ability of TP to form aggregates.

[0243] TP-M: The mutation of glutamic acid to aspartic acid (E8D) is a similar amino acid mutation, and the mutation of alanine to aspartic acid (A10D) provides additional acidic residues. The observed condensate size was larger than that of TP-L, indicating that the additional acidic residues can enhance the phase separation tendency, which may also be related to the pI value being closer to 7.4.

[0244] The universality of the DILLPS effect was confirmed through homology comparison and comparison of homologous peptide phase separation capabilities. Furthermore, the electrostatic adsorption effect contributed by Glu8 was found to be one of the key factors determining the liquid-liquid phase separation capability of TP.

[0245] Summary of the molecular mechanism of TP dilution-induced liquid-liquid phase separation 2.28

[0246] In summary, the molecular mechanism of dilution-induced phase separation is similar to that of an energy funnel. Figure 34 As shown, peptide molecules (TPs) exist as dynamically conformational aggregates, comprising two main conformational subtypes: compact coiled and α-helical. At high concentrations, the compact coiled monomer form is energy-dominant, thanks to its specific sequence arrangement, making intramolecular interactions dominant. With dilution, changes in the solution microenvironment reshape the energy landscape of peptide folding—the local enrichment of counterions (mainly buffer salts that tend to bind to peptides) weakens intermolecular repulsion by residing on the peptide surface through charge neutralization and hydrogen bonding. Simultaneously, counterions disrupt the hydration layer structure on the TP surface through their hydration capacity, releasing bound water molecules (driven by entropy increase). This process tilts the energy balance towards the α-helical state. The conformational change exposes hydrophobic residues and π–π stacking / cation–π interaction sites, which are favorable for intermolecular interactions, to the solvent. Simultaneously, electrostatic rearrangement makes intermolecular interactions dominant, promoting peptide assembly and aggregation through multivalent binding sites, leading to phase separation events. This mechanism reveals the crucial role of solvation effects and ion-specific binding in regulating the phase separation of biomolecules.

[0247] 2.3 Compartmentalization Study of Dilution-Induced Liquid-Liquid Phase Separation

[0248] 2.31TP exhibited a dilution-induced compartmentalization enrichment effect.

[0249] Phase separation-mediated biological compartmentalization is one of the core mechanisms by which cells achieve precise spatiotemporal regulation. Membrane-free organelles (such as nucleoli, stress granules, and P-body molecule) formed through liquid-liquid phase separation provide cells with highly dynamic functional microcompartments. These bioconcentrates can locally concentrate specific biomolecules, significantly increasing the local concentration of target substances. This allows biochemical reactions that are difficult to occur at low concentrations to proceed rapidly in the microenvironment, thereby improving reaction efficiency. Furthermore, bioconcentrates provide a controlled microenvironment for molecules, facilitating the regulation of signal transduction, gene expression, and stress responses. Through isolation and enrichment, bioconcentrates not only ensure the spatiotemporal ordered distribution of intracellular substances but also rapidly adjust local reaction dynamics in response to environmental changes, safeguarding cellular homeostasis. Therefore, in-depth research into phase separation mechanisms is of significant scientific value for revealing the fundamental principles of life activities and developing novel therapeutic strategies.

[0250] The dilution-induced phase separation effect of TP provides a novel paradigm for the enrichment of guest molecules in aggregates: by adjusting the concentration of TP, the recruitment and release of guest molecules can be effectively controlled. Specifically, when TP is in a high-concentration homogeneous state, guest molecules are uniformly distributed in the solution; when the solution is diluted, although the global concentration of guest molecules decreases, due to the occurrence of TP phase separation, guest molecules are recruited by aggregates, and their local concentration can be significantly increased, thereby achieving a dilution-induced compartmentalized enrichment effect. Figure 35 Correspondingly, increasing the TP concentration can effectively release guest molecules.

[0251] Firstly, regarding the recruitment of small molecules, TP exhibited a dilution-induced enrichment effect on the aromatic fluorescent molecular probes 4',6-diamino-2-phenylindole (DAPI) and thioflavin-T (ThT): after diluting the homogeneous solution containing TP and the fluorescent probe, the fluorescent probe was enriched in the TP aggregates, with partition coefficients (K = C heavy phase / C light phase) reaching 7 and 12, respectively, indicating a high enrichment effect. Figure 36 , Figure 39 ).

[0252] Secondly, to detect whether TP condensates exhibit selectivity in terms of the charged properties of guest molecules, a series of 5-carboxyfluorescein (FAM)-labeled peptides with pI ranges of 3–13 were selected as guest molecules for verification. Figure 37 Through dilution, it was found that these peptides were all enriched in TP condensates, with partition coefficients ranging from 10 to over 255 (confocal resolution limit), all falling within the high enrichment range. Figures 38-39 This study reveals that TP condensates can recruit guest molecules with different charged properties in a broad spectrum.

[0253] 2.32 The spatial distribution of RNA in TP condensates is regulated by buffer salt concentration.

[0254] Because RNA itself has strong electronegativity, it can strongly bind to arginine and lysine in TP. Therefore, an experiment was designed to investigate whether RNA has a regulatory effect on the dilution-induced phase separation of TP. Total yeast RNA was used as the guest molecule, and 10 mM Tris (pH 7.4) was used as the buffer salt. When TP was at a high concentration (2 mM), unlike the complete clarity without RNA, the addition of RNA resulted in the presence of sparse aggregates in the solution. Figure 40 A), and the phase separation trend increases with increasing RNA concentration. Figure 40 (B) This may be due to the direct interaction between RNA and TP; and after dilution (0.5 mM), the phase separation ability of TP is enhanced, and RNA is thus further enriched in the newly formed TP condensates. Figure 41A). Furthermore, by altering the buffer salt concentration, the spatial distribution of RNA within TP condensates can be regulated: under 10 mM Tris buffer conditions, RNA is primarily enriched in the center of the TP condensate, with less distribution at the periphery. Figure 41 A); Increasing the buffer salt concentration to 50mM can cause RNA localization to gradually shift from the inside to the outside layer, with the signal intensity ratio at the edge to the center being approximately 4:1. Figure 41 (B) This phenomenon may stem from the competition between RNA and buffer salts for different phase separation driving mechanisms of TP: Under low-concentration buffer salt conditions, the α-helicalization of TP is insufficient, and the interaction between polypeptides is weak. At this time, the scaffolding effect of RNA becomes dominant, using electrostatic interactions to mediate TP phase separation with itself as the core, and the size of the resulting aggregates is also small (~1 nm). Under high-concentration buffer salt conditions, the intermolecular interactions of TP molecules are strong, and they tend to rely on their own aggregation to form aggregates (aggregates visible in the field of view up to 5 nm), and recruit RNA on this basis. Since the total size of yeast RNA is relatively large, it is not easy to penetrate into the high-density TP aggregates, thus exhibiting a "shell-like" distribution.

[0255] 2.33TP-mediated dilution-induced emission effect

[0256] The dilution-induced enrichment effect of TP provides a novel and economical approach to achieving aggregation-induced emission (AIE). AIE is a phenomenon where the luminescence intensity of organic compounds is enhanced in the aggregated state. Unlike the concentration-caused quenching (ACQ) effect that traditional fluorescent materials easily exhibit at high concentrations, AIE materials show almost no luminescence in dilute solutions, but their luminescence intensity is significantly enhanced in the aggregated state. Triphenylethylene (TVP) is used as an example... Figure 42 In the solution state, the benzene rings can rotate freely, resulting in energy dissipation through non-radiative means and low luminescence efficiency. However, in the aggregated state, these rotations are restricted, non-radiative dissipation is reduced, and luminescence efficiency is significantly improved. Figure 42 The schematic diagram and confocal microscopy imaging show that in a homogeneous mixed solution containing 2 mM TP and 500 nM TVP, the solution is almost undetectable due to the free and unaggregated state of TVP; however, when the solution is diluted 10 times, the TVP molecules emit bright red fluorescence at 620 nm, with an intensity more than 20 times higher than the initial state.

[0257] This phenomenon provides a theoretical basis and experimental support for developing a low-cost, high-efficiency detection platform based on dilution-induced phase separation to achieve the AIE effect. This strategy of regulating luminescence using molecular aggregation states has broad application prospects in fields such as bioimaging, sensing, and drug delivery.

[0258] 2.34TP forms a multi-stage phase separation mode through dilution-induced enrichment.

[0259] Previous work has demonstrated that TP (transferase-induced enrichment) can extensively enrich guest molecules through dilution-induced enrichment, achieving effective compartmentalization regulation. Here, this study uses gallectin-3 (Gal-3), a protein that undergoes LLPS (low-level phase separation) at high concentrations, as an example. It participates in the TP dilution process as a guest protein, simulating the multilevel phase separation and stratification mechanism of the nucleolus. Through dilution-induced enrichment, the concentration of the guest protein locally increases in the opposite direction, forming a secondary phase separation event. This provides a new reference model for understanding the formation of complex biological compartments. Figure 43 ).

[0260] Gal-3 is a multifunctional protein that has been reported to participate in various cellular processes, such as cell adhesion, apoptosis, and immune responses, through LLPS. The eGFP-Gal3 obtained through recombinant expression has a high phase separation concentration threshold; even at 18 mg / ml without the presence of the congestion agent PEG, phase separation events were not observed. Figure 44 In the experiment, 2 mg / ml of eGFP-Gal3 and a high concentration of TP were used to form a homogeneous solution, at which point no signal was detected in any of the fluorescence channels. However, with halving using buffer, TP acted as a scaffold to form primary aggregates, accompanied by the enrichment of eGFP-Gal3. This triggered phase separation within the TP aggregates, forming multi-compartmental secondary aggregates. Figure 45 As a control, the eGFP protein, which lacks phase separation ability, is uniformly distributed within the TP condensate instead of forming secondary compartments. Figure 46 A). Merging events also occur between these secondary compartments, but kinetically, they need to overcome the TP barrier and eliminate the TP to form their own compartments. Figure 46 B). Through 3D reconstruction and fluorescence distribution analysis of the multi-phase separation system, it was found that eGFP-Gal3 exists in three levels: its signal in the external solution is almost undetectable, while inside the TP aggregate, the signal intensity ratio between the external and external regions in the secondary zone is approximately 10:1; for TP, the signal is almost undetectable in the secondary zone, suggesting that it has been sufficiently excluded to form a "cavity" ( Figure 47 Furthermore, Part-FRAP experiments revealed that eGFP-Gal3 exhibits extremely high mobility within the secondary compartments, completing mass exchange during quenching and leading to a decrease in the global signal. Figure 48).

[0261] In summary, TP condensates, through local enrichment effects, can significantly reduce the phase separation threshold of guest proteins (from high to low concentration in terms of global concentration), demonstrating the core role of microenvironment remodeling in regulating biomolecular behavior. This "concentration amplifier" mechanism provides new insights into signal transduction and stress response, and also offers a theoretical basis and application inspiration for designing novel biomaterials and drug delivery systems.

[0262] 2.35TP achieves enzymatic catalysis through dilution-induced enrichment.

[0263] TP's excellent guest molecule enrichment capability makes it a potential candidate for concentration-responsive microreactors. By locally enriching substrates and enzymes, the rate of enzyme-catalyzed reactions can be increased, making dilution-induced phase separation not only a physical separation but also a switch for functional activity.

[0264] This study uses gallectin-10 (Gal-10), a member of the Gal-3 family, as a guest molecule in an enzymatic catalytic reaction. Gal-10 is a protein that readily self-assembles into crystals both in vivo and in vitro and is associated with immune and inflammatory diseases. Previous studies have shown that fusion expression of a His-tag at the N-terminus of Gal-10 effectively inhibits its aggregation tendency, facilitating protein expression. Furthermore, enzymatic cleavage of the His-tag linker site by TEV protease restores Gal-10's ability to self-assemble into crystals. Based on the properties of this enzymatic reaction system, experiments were designed to compare the crystal formation rate before and after dilution-induced phase separation, evaluating the effectiveness of TP as a microreactor. Figure 49 His-Gal-10 and TEV protease were labeled with Alexa 488 and Alexa 633, respectively, to construct a three-component mixed solution of His-Gal-10 / TEV protease / TP. After dilution, His-Gal-10 and TEV protease were enriched in TP condensates, exhibiting spatial co-localization. Figure 50 The reaction systems under each condition were incubated on a shaker at 4°C. Observation under a polarizing microscope showed that for the diluted low-concentration group, Gal-10 crystals began to appear at 40 minutes and became clearly visible at 80 minutes. Figure 51 A), while in the high-concentration group, no observable crystals were found in the field of view at 80 minutes. Figure 51 B). For reference, without any TP involved, the catalytic reaction requires 180 minutes to observe crystal formation. Figure 51 C).

[0265] In summary, the experimental results demonstrate that the compartmentalized microreactors formed by TP possess highly efficient biocatalytic capabilities and a unique concentration regulation mechanism. This system not only fully simulates the dynamic regulation principle of "molecular enrichment-functional emergence" in membrane-free organelles but also expands the application boundaries of phase separation in synthetic biology—by designing responsive scaffolds (such as TP) coupled with modular enzyme systems (such as His-TEV protease-Gal-10), programmable in-situ catalytic enhancement and product self-assembly can be achieved, providing innovative tools for intelligent drug factories, dynamic biomaterials, and in vitro metabolic pathway reconstruction.

[0266] The above description of the embodiments is only for understanding the method and core ideas of the present invention. It should be noted that those skilled in the art can make various improvements and modifications to the present invention without departing from the principles of the invention, and these improvements and modifications will also fall within the protection scope of the claims of the present invention.

Claims

1. A polypeptide, characterized in that, The sequence of the polypeptide is set forth in any one of SEQ ID NO: 1, SEQ ID NO: 2, SEQ ID NO:

5.

2. A polypeptide condensate, characterized in that, The polypeptide aggregate comprises the polypeptide of claim 1.

3. A polynucleotide, comprising, The polynucleotide encodes the polypeptide of claim 1 or the polypeptide aggregate of claim 2.

4. A vector, characterized by, The vector comprises the polynucleotide of claim 3.

5. A host cell, characterized in that, The host cell comprises the polynucleotide of claim 3 or the vector of claim 4.

6. Use of the polypeptide of claim 1, the polypeptide aggregate of claim 2, the polynucleotide of claim 3, the vector of claim 4, the host cell of claim 5 in aggregation-induced emission or in the preparation of a product of aggregation-induced emission.

7. Use according to claim 6, characterized in that, The product comprises an aggregation-induced emission probe.

8. A method of facilitating aggregation-induced emission, the method comprising: The method comprises using the polypeptide of claim 1, the polypeptide aggregate of claim 2, the polynucleotide of claim 3, the vector of claim 4, the host cell of claim 5.

Images