A method for isolating and purifying a recombinant collagen type III

Abstract

The application relates to the technical field of protein purification, and discloses a separation and purification method of recombinant type III collagen. The method comprises the following steps: obtaining a crude extract by centrifuging a broken wall of recombinant yeast bacteria, removing glycoprotein impurities by passing through concanavalin A affinity chromatography, gradually purifying and enriching a trimer by anion exchange chromatography and gel filtration chromatography, and finally performing ultrafiltration concentration, dialysis and desalting to obtain recombinant type III collagen with a purity not lower than 95%. The application solves the problems that in the prior art, high-mannose glycoprotein impurities from yeast are difficult to remove with high selectivity, and a trimer configuration of a target protein is easily damaged in a purification process and cannot be effectively enriched.

Description

Technical Field

[0001] This application relates to the field of protein purification technology, and in particular to a method for the isolation and purification of recombinant type III collagen. Background Technology

[0002] Recombinant type III collagen, due to its amino acid sequence and biological functions highly similar to human collagen, shows broad application prospects in fields such as medical aesthetics, tissue engineering, and wound repair. Currently, intracellular expression systems using methanol-nutritional yeasts such as Pichia pastoris as hosts have become one of the mainstream technologies for the large-scale production of recombinant collagen due to their advantages such as high secretion efficiency, low culture cost, and the ability to achieve high-density fermentation. In the downstream purification stage, existing technologies typically involve cell disruption followed by centrifugation to obtain a crude extract, which is then captured and purified using one-step or two-step chromatography (such as anion exchange chromatography, hydrophobic interaction chromatography, or metal chelate affinity chromatography). Finally, the final product is obtained through ultrafiltration concentration and buffer replacement.

[0003] However, when recombinant type III collagen is expressed using yeast as a host, the crude extract released after cell disruption is extremely complex, containing a large number of high-mannose glycoproteins derived from yeast host cells. These glycoprotein impurities partially overlap with recombinant type III collagen in terms of molecular weight and isoelectric point, making it difficult to completely separate them using traditional ion exchange or hydrophobic chromatography. Existing purification methods often use high-temperature heat treatment to denature and precipitate the host protein for coarse removal of this specific impurity. However, this heat treatment process not only causes irreversible aggregation of the target protein and a significant decrease in recovery rate, but may also disrupt the natural conformation of recombinant type III collagen, which consists of three single chains assembled into a trimer through interchain hydrogen bonds, causing the product to lose the higher-order structural integrity upon which its biological activity depends. Furthermore, existing processes generally lack fine-grained fractionation steps based on protein molecular weight, failing to effectively distinguish between different polymerization states of the target protein, such as trimers, dimers, and free single chains. This results in the final product's trimer content and uniformity failing to meet the stringent conformational purity requirements for pharmaceutical applications. Summary of the Invention

[0004] This application provides a method for separating and purifying recombinant type III collagen, which solves the problems in the prior art where yeast-derived high-mannose glycoprotein impurities are difficult to remove with high selectivity and the target protein trimer conformation is easily damaged and cannot be effectively enriched during purification. This method improves the purity, trimer content and total recovery rate of recombinant type III collagen, enabling it to meet the requirements of pharmaceutical-grade applications for product purity and conformational uniformity.

[0005] This application provides a method for isolating and purifying recombinant type III collagen, the method comprising: S1. The recombinant yeast cells containing recombinant type III collagen were suspended in a buffer solution containing protease inhibitors. The resulting cell suspension was subjected to high-pressure homogenization to disrupt the cell wall. Cell fragments were removed by centrifugation to obtain a crude extract. S2. Pass the crude extract into a balanced concanavalin A agarose affinity chromatography column and collect the unbound components in flow-through mode. This allows high-mannose glycoprotein impurities from yeast host cells to specifically bind to and be retained by concanavalin A. Recombinant type III collagen without mannose N-glycan modification flows out of the column to obtain a deglycoprotein flow-through solution. S3. The deglycoprotein flow-through solution is replaced with buffer and loaded into an anion exchange chromatography column. It is eluted with sodium chloride in a fractional gradient and the elution fraction containing recombinant type III collagen is collected to obtain the ion exchange purified solution. S4. After concentrating the ion exchange purified solution, load it into a gel filtration chromatography column for molecular weight fractionation and collect the elution fraction corresponding to the molecular weight range of recombinant type III collagen trimer to obtain the gel filtration purified solution. S5. The gel filtration purification solution is concentrated by ultrafiltration and then desalted by dialysis with phosphate buffer containing glycine and trehalose to obtain recombinant type III collagen with a purity of not less than 95%.

[0006] In the technical solution provided in this application, recombinant yeast cells are suspended in a buffer solution containing protease inhibitors. This effectively inhibits the degradation of the target protein by intracellular proteases while simultaneously homogenizing the cell walls under high pressure, thus ensuring the integrity of the recombinant type III collagen from the source. The crude extract obtained after centrifugation to remove cell fragments provides a clear injection base for subsequent chromatographic purification. A concanavalin A agarose affinity chromatography column is introduced and operated in flow-through mode. This cleverly utilizes the essential difference in glycan modification between the target protein and the host impurities. Recombinant type III collagen does not contain mannose-type N-glycan modification, and therefore does not bind to concanavalin A and flows directly out of the column. In contrast, high-mannose-type glycoprotein impurities from yeast host cells are retained because their densely packed mannose residues specifically bind to concanavalin A in a multivalent manner. This negative screening strategy based on glycan recognition, compared to traditional heat-induced denaturation precipitation or non-specific adsorption methods, achieves highly selective removal of yeast-derived glycoprotein impurities at the molecular recognition level. It avoids the risk of irreversible aggregation or conformational damage to non-denatured target proteins due to heat, while maximally preserving the natural interchain hydrogen bond assembly structure of recombinant type III collagen trimers under mild conditions, laying a high-purity foundation for subsequent purification steps. After buffer replacement, the deglycoprotein flow-through is loaded into an anion-exchange column, and residual non-glycoprotein host impurities are finely separated using a fractional sodium chloride gradient elution. This technique allows protein components with different isoelectric points and surface charge densities to elute sequentially under progressively increasing ionic strength, thereby achieving further differentiation between target and impurity proteins along the charge dimension and significantly improving the purity of the intermediate product.

[0007] The ion-exchange purified solution was concentrated and loaded onto a gel filtration chromatography column for molecular weight fractionation. The elution fractions specifically collected within the molecular weight range of recombinant type III collagen trimers are noteworthy for their design logic. This is not simply a routine desalting or polishing step, but rather a precise identification and targeted enrichment of different polymerization states of the target protein at the molecular weight level. Since recombinant type III collagen spontaneously assembles into trimers from three single chains via interchain hydrogen bonds under non-denaturing conditions, and this trimer configuration is the structural basis for its biological function, the targeted collection of the trimer fractions during molecular weight fractionation essentially elevates the purification target from simple chemical purity to conformational homogeneity. This ensures that the final product not only meets requirements for residual impurities but also achieves controllable enrichment of the functional polymeric states. Ultrafiltration concentration followed by dialysis desalting with phosphate buffer containing glycine and trehalose is employed. The introduction of glycine and trehalose as protein stabilizers maintains the conformational stability of recombinant type III collagen trimers during formulation through preferential hydration and glassy state protection mechanisms, preventing depolymerization or irreversible aggregation due to excessively high local concentrations or freeze-thaw stress during concentration and storage. In summary, the technical features of each step in this application form a multi-dimensional, progressive purification system encompassing glycan recognition, charge resolution, molecular weight fractionation, and formulation stabilization. The functional roles of each step are complementary and interconnected, collectively ensuring the overall quality of the final product in terms of purity, trimer content, and biological activity. Attached Figure Description

[0008] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0009] Figure 1 This is a schematic diagram of an embodiment of the method for isolating and purifying recombinant type III collagen in this application. Figure 2 This is a schematic diagram comparing the purification effects of Example 1 and Comparative Experiments 1 to 3 in this application. Detailed Implementation

[0010] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. The following specific embodiments will provide further detailed description of the invention. Unless otherwise specified, the pharmaceuticals and reagents used in the embodiments are commercially available products, and the methods used are conventional methods in the art.

[0011] The gene encoding recombinant type III collagen used in the various embodiments of this invention originates from the mature collagen region of the human type III procollagen α1 chain (COL3A1). Using the human COL3A1 mRNA sequence shown in GenBank accession number NM_000090.4 as a reference, a fragment encoding amino acid residues 151 to 601 (a total of 451 residues) of the mature collagen triple helix region was extracted. This fragment contains (Gly-XY). n The triplet repeat domain does not contain N-terminal or C-terminal propeptide sequences, nor does it contain a signal peptide sequence. Following codon optimization based on Pichia pastoris codon preference, the encoded sequence was chemically synthesized, digested with restriction endonucleases EcoRI and NotI, and ligated to the multiple cloning site of the Pichia pastoris intracellular expression vector pPIC9K to construct the recombinant expression plasmid pPIC9K-ColIII. This plasmid was linearized with SalI and electroporated into Pichia pastoris GS115 competent cells (His-phenotype). High-copy-count integrative transformants were screened on histidine-deficient medium and 4 mg / mL G418 resistant plates to obtain the engineered strain GS115 / pPIC9K-ColIII. The theoretical molecular weight of the single-chain polypeptide encoded by this truncated fragment is 48.7 kDa, and the apparent molecular weight, determined by SDS-PAGE after intracellular expression in Pichia pastoris, is 55 kDa. The reason why the apparent molecular weight is higher than the theoretical molecular weight is that the amino acid composition of collagen, which is rich in glycine, proline and hydroxyproline, makes its SDS binding ratio lower than that of globular proteins. This results in a lower electrophoretic mobility in SDS-PAGE, thus showing a higher apparent molecular weight reading. This phenomenon is consistent with the behavior of naturally extracted human type III collagen α1 chains in SDS-PAGE.

[0012] In some embodiments, in addition to the truncated fragment of 451 residues mentioned above, a specific (Gly-XY) fragment is extracted from the mature collagen triple helix region of the human type III procollagen α1 chain. nThe invention employs repeated structural units, tandemly repeating these units to construct recombinant type III collagen variants of varying lengths. The aim of this design is to selectively extract representative triple-helix fragments from natural collagen and tandemly copy them multiple times, thereby obtaining recombinant collagen molecules with controllable molecular weight and well-defined sequence composition while preserving the typical triple-helix folding ability of collagen. Specifically, this invention designs the following six recombinant type III collagen variants: Variant Seq 2: This variant uses a fragment containing the core motif Gly-Ala-Pro-Gly-Phe-Arg-Gly-Pro-Ala-Gly-Pro-Asn-Gly-Ile-Pro-Gly-Glu-Lys in the triple helix region of human type III collagen as the basic structural unit. The core motif is tandemly arranged 8 times, with the 1st to 5th and 7th to 8th times being the standard motifs mentioned above. A micro-variant with a proline residue deletion appears in the 6th time, namely Gly-Ala-Pro-Gly-Phe-Arg-Gly-Pro-Ala-Gly-Asn-Gly-Ile-Pro-Gly-Glu-Lys. The core motifs are connected by short peptides such as Gly-Pro-Ala-Gly-Glu-Arg and Gly-Ala-Pro-Gly-Glu-Arg, with a total length of 238 amino acid residues. The gene encoding this variant was optimized according to the codon preference of Pichia pastoris, with EcoRI and SalI restriction sites introduced at both ends, for a total length of 726 base pairs.

[0013] Variant Seq 5: Each repeating unit consists of 33 amino acid residues; the amino acid composition is Gly-Leu-Pro-Gly-Ala-Ala-Gly-Glu-Arg-Gly-Ala-Pro-Gly-Phe-Arg-Gly-Pro-Ala-Gly-Pro-Asn-Gly-Ile-Pro-Gly-Glu-Lys-Gly-Pro-Ala-Gly-Glu-Arg, with each repeating unit strictly tandemly repeated 5 times, for a total length of 165 amino acid residues. The full sequence of this variant strictly conforms to (Gly-XY). n The triplet repeats a pattern, with glycine as the first residue in each of the three residues. After codon optimization, the gene encoding this triplet is 507 base pairs long, with EcoRI and SalI restriction enzyme sites at both ends.

[0014] Variant Seq 8: This variant uses 39 amino acid residues as repeating units, with the amino acid composition Gly-Lys-Asp-Gly-Ser-Pro-Gly-Glu-Pro-Gly-Ala-Asn-Gly-Leu-Pro-Gly-Ala-Ala-Gly-Glu-Arg-Gly-Ala-Pro-Gly-Phe-Arg-Gly-Pro-Ala-Gly-Pro-Asn-Gly-Ile-Pro-Gly-Glu-Lys. This repeating unit is strictly tandemly repeated 5 times, for a total length of 195 amino acid residues. The full sequence of this variant also strictly conforms to (Gly-XY). n The triplet repeats a pattern. Its encoding gene, after codon optimization, has a total length of 597 base pairs, with EcoRI and SalI restriction enzyme sites at both ends, respectively.

[0015] Variant dSeq 2: This variant is a second tandem repeat of the complete amino acid sequence of Seq 2, that is, the 238 residues of Seq 2 are linked together and repeated once, with a total length of 476 amino acid residues, corresponding to a total coding gene length of 1452 base pairs.

[0016] Variant dSeq 5: This variant is a second tandem repeat of the complete amino acid sequence of Seq 5, that is, the 165 residues of Seq 5 are linked end to end and repeated once, which is equivalent to strictly repeating the repeating unit of 33 amino acids in tandem 10 times, with a total length of 330 amino acid residues, corresponding to a total coding gene length of 1014 base pairs.

[0017] Variant dSeq 8: This variant is a second tandem repeat of the complete amino acid sequence of Seq 8, that is, the 195 residues of Seq 8 are linked end to end and repeated once, which is equivalent to strictly repeating the repeating unit of 39 amino acids in tandem 10 times, with a total length of 390 amino acid residues, corresponding to a total coding gene length of 1194 base pairs.

[0018] The coding genes for the six variants were all codon-optimized and chemically synthesized in the same manner as the aforementioned 451-residue truncated fragments, and then ligated into a Pichia pastoris intracellular expression vector after double digestion with EcoRI and SalI. Since the amino acid sequences of these variants all originate from the (Gly-XY) triple helix region of the mature collagen α1 chain of human type III procollagen,... n The repeating domain does not contain N-terminal and C-terminal propeptide sequences or signal peptide sequences, and N-glycosylation site analysis shows that there are no N-glycosylation consensus sequences in the form of NXS / T. Therefore, it does not undergo high-mannose N-glycan modification when expressed in yeast cells.

[0019] N-glycosylation site analysis of the truncated fragment revealed the absence of an N-glycosylation consensus sequence in the form of NXS / T (N for asparagine, X for any amino acid except proline, and S / T for serine or threonine) within the 451 amino acid residues. In the yeast intracellular expression environment, this recombinant type III collagen did not undergo high-mannose N-glycan modification, which is the molecular basis of the ConA negative selection strategy in step S2. Target proteins without mannose N-glycans do not bind to ConA and exit the column directly in flow-through mode, while yeast host proteins containing high-mannose N-glycans are specifically truncated by ConA.

[0020] Under non-denaturing conditions, three truncated single-stranded fragments assemble into a trimer via interchain hydrogen bonds along the triple helix axis. The theoretical molecular weight of the trimer is 48.7 kDa × 3 = 146.1 kDa. However, due to the rigid rod-like conformation of the collagen triple helix, its hydrodynamic radius is significantly larger than that of globular proteins of the same molecular weight. Therefore, when calibrated with globular protein standards in gel filtration chromatography and SEC-HPLC, the apparent molecular weight reading of the trimer is higher, ranging from 300 to 500 kDa. The theoretical molecular weight of 165 kDa is calculated using an apparent molecular weight of 55 kDa × 3 from SDS-PAGE, a common characterization convention in the field of collagen.

[0021] Example 1 This embodiment provides a method for the isolation and purification of recombinant type III collagen, such as... Figure 1 As shown, it includes: S1. The recombinant yeast cells containing recombinant type III collagen were suspended in a buffer solution containing protease inhibitors. The resulting cell suspension was subjected to high-pressure homogenization to disrupt the cell wall. Cell fragments were removed by centrifugation to obtain a crude extract. S2. Pass the crude extract into a balanced concanavalin A agarose affinity chromatography column and collect the unbound components in flow-through mode. This allows high-mannose glycoprotein impurities from yeast host cells to specifically bind to and be retained by concanavalin A. Recombinant type III collagen without mannose N-glycan modification flows out of the column to obtain a deglycoprotein flow-through solution. S3. After replacing the deglycoprotein flow-through with buffer, load it into an anion exchange chromatography column, elute with sodium chloride in a fractional gradient, and collect the elution fraction containing recombinant type III collagen to obtain the ion exchange purified solution. S4. After concentrating the ion exchange purified solution, load it into a gel filtration chromatography column for molecular weight fractionation and collect the elution fraction corresponding to the molecular weight range of recombinant type III collagen trimer to obtain the gel filtration purified solution. S5. The gel filtration purification solution is concentrated by ultrafiltration and then desalted by dialysis with phosphate buffer containing glycine and trehalose to obtain recombinant type III collagen with a purity of not less than 95%.

[0022] The fermentation culture of *Pichia pastoris* strain GS115 / pPIC9K-ColIII, which expresses recombinant type III collagen intracellularly, was used as the raw material. Fermentation was carried out in a 30L fermenter. After inoculation, glycerol was used as the carbon source for the cell growth phase. Once the cell wet weight reached 200 g / L, the culture was switched to methanol induction. After 72 hours of induction, the fermentation broth was harvested. The fermentation broth was centrifuged at 5000g for 15 min at 4℃ to collect the cell pellet, and the supernatant was discarded. The operating temperature was controlled between 2 and 8℃ throughout the process.

[0023] The implementation process of step S1 is as follows: Take 600g of yeast cell precipitate (wet weight) and add 3000mL of lysis buffer (30mM sodium phosphate buffer, pH 7.4, containing 150mM sodium chloride, 1mM EDTA, and 1mM PMSF) to fully suspend it. Each gram of wet cell weight corresponds to 5mL of buffer. The PMSF needs to be temporarily removed from the 100mM isopropanol stock solution and added to the lysis buffer before use, and left at room temperature for no more than 30 minutes to avoid spontaneous hydrolysis in the aqueous solution, which would lead to inactivation. Transfer the cell suspension to the feed tank of a GEANiro Soavi Panda PLUS high-pressure homogenizer, set the homogenization pressure to 700bar, turn on the cooling circulating water (inlet water temperature 2℃), and perform the first homogenization cycle. Collect the effluent from the first cycle and return it to the feed tank, maintaining a pressure of 700bar for the second cycle. After the second cycle, the pressure was reduced to 500 bar for the third cycle. This pressure reduction was to prevent irreversible aggregation of the released free proteins under high shear stress during the third cycle. Between each cycle, the material temperature was controlled at 2–6°C using a heat exchanger. A 50 μL sample of the homogenized solution was observed under a 100x objective lens using an optical microscope. The proportion of intact yeast cells (including fragments and intact cells) in the field of view was less than 5%, indicating a cell disruption rate of over 95%.

[0024] The homogenized solution was transferred to a pre-chilled centrifuge bottle at 4°C for the first centrifugation: centrifuged at 8000g for 20 min (Beckman Avanti J-26S XPI centrifuge, JA-10 rotor), discarding the precipitate containing bacterial fragments and unbroken cells, and collecting the supernatant. The supernatant was then transferred back to a pre-chilled centrifuge bottle for the second centrifugation: centrifuged at 12000g for 30 min (JA-25.50 rotor) to remove cell membrane fragments, lipid particles, and subcellular organelle remnants. The supernatant from the second centrifugation was collected and vacuum filtered through a 0.45 μm polyethersulfone (Millipore Stericup) membrane to remove residual particles, yielding approximately 2800 mL of clear crude extract. The total protein concentration of the crude extract was determined by the BCA method to be 18.6 mg / mL, with a total protein content of approximately 52.1 g. 10 μL of the crude extract was analyzed by 12% reduced SDS-PAGE electrophoresis. A clear target protein band was visible at 55 kDa, while a large number of host protein bands were distributed in the 25-80 kDa range. ImageJ grayscale analysis showed that the target protein band area accounted for 12.3% of the total protein band area in the lanes, based on which the target protein content in the crude extract was estimated to be approximately 6.4 g.

[0025] The procedure for step S2 is as follows: Take 280 mL of ConA-Sepharose 4B wet gel (GE Healthcare, ConA coupling density 14 mg / mL gel) and load it into an XK 50 / 30 chromatography column (inner diameter 5.0 cm, column height approximately 14.3 cm). Use 20 mM Tris-HCl buffer (pH 7.4) containing 1 mM calcium chloride, 1 mM manganese chloride, and 0.5 M sodium chloride as the equilibration buffer, and equilibrate for 3 column volumes at a linear velocity of 30 cm / h until the UV 280 nm baseline and pH are stable. Load 2800 mL of the clarified crude extract obtained in step S1 into the ConA column at a linear velocity of 30 cm / h. During sample loading, high-mannose glycoproteins in yeast host cell proteins, including the molecular chaperone Kar2p (78 kDa, containing Man8GlcNAc2 to Man...), are detected. 12 GlcNAc2-type N-glycans, vacuolar protease carboxypeptidase Y (CPY, 61 kDa, containing 4 N-glycosylation sites), and cell wall mannose proteins, etc., have α-D-mannosyl residues on their glycan chains that are specifically recognized and captured by the sugar-binding pockets of ConA, causing these glycoprotein impurities to adsorb onto the column and be retained. The amino acid sequence of recombinant type III collagen is (Gly-XY). nThe triplet repeats form the main framework and do not contain N-glycosylated consensus sequences in the form of asparagine-X-serine / threonine (NXS / T). During intracellular expression in yeast, high-mannose N-glycosylation modification does not occur, and the target protein does not bind to ConA, exiting the column directly with the flow-through. Approximately 2900 mL of the flow-through was collected (the slight increase in volume is due to dead volume within the column and the addition of washing fluid). The total protein concentration in the flow-through, determined by the BCA method, was 7.2 mg / mL, with a total protein content of approximately 20.9 g, a 59.9% reduction from the 52.1 g in the crude extract. This indicates that the ConA column retained approximately 31.2 g of protein, meaning that about 60% of the protein in the crude extract was glycoprotein impurities containing high-mannose N-glycans. 10 μL of the flow-through buffer was analyzed by SDS-PAGE electrophoresis. Compared with the crude extract, multiple bands of contaminating proteins in the 40-80 kDa range were significantly weakened or disappeared, and the relative proportion of the target protein band at 55 kDa increased from 12.3% to 29.8%. Based on the target protein band area ratio multiplied by the total protein content, the target protein content in the flow-through buffer was estimated to be approximately 6.2 g, with a recovery rate of 96.9%.

[0026] After flow-through collection, the ConA column was eluted for 3 column volumes with equilibration buffer containing 0.2 M methyl-α-D-mannoside to dissociate the retained glycoprotein impurities from ConA. After elution, the column was washed for another 5 column volumes with equilibration buffer until baseline was restored, completing column regeneration. The regenerated ConA column underwent stability testing (after 5 reuses, the glycoprotein rejection rate remained above 95%) and was deemed reusable.

[0027] The procedure for step S3 is as follows: 2900 mL of the deglycoproteinized flow-through solution obtained in step S2 was transferred into a tangential flow ultrafiltration system equipped with a Pellicon 3 Ultracel 10 kDa regenerated cellulose ultrafiltration membrane pack (0.11 m² membrane area). Constant volume perfiltration was performed using 20 mM Tris-HCl buffer (pH 8.0) as the replacement buffer. After 6 volume-fold perfiltrations, the conductivity of the permeate was monitored using a conductivity meter until it dropped below 0.8 mS / cm (corresponding to a NaCl concentration of approximately 6 mM), meeting the requirements for anion exchange loading. After perfiltration, the solution was concentrated to 500 mL, with a total protein concentration of 38.5 mg / mL.

[0028] 200 mL of DEAE-Sepharose Fast Flow wet gel was loaded into an XK 50 / 30 column and equilibrated for 3 column volumes with 20 mM Tris-HCl buffer (pH 8.0). 500 mL of the dialysis concentrate was loaded into the DEAE column at a linear velocity of 60 cm / h. At pH 8.0, recombinant type III collagen (isoelectric point approximately 6.5) has a negative net charge and is electrostatically adsorbed onto the DEAE group, thus binding to the column. Basic host proteins with isoelectric points higher than 8.0 do not bind to DEAE during the loading phase and are discharged with the flow-through buffer. After loading, the column was washed with equilibration buffer until the UV baseline stabilized, and then a four-stage gradient elution was performed. The first stage involved eluting for 5 column volumes (1000 mL) with 20 mM Tris-HCl (pH 8.0) containing 0.05 M NaCl. A small peak was observed at 280 nm UV light; this peak corresponds to weakly bound nucleic acid fragments and small peptide impurities, which were discarded. The second eluent, containing 0.15 M NaCl, was used to elute for 5 column volumes (1000 mL). A moderate absorption peak appeared, and 10 μL was analyzed by SDS-PAGE. Multiple host protein bands were observed, but no clear target protein band was found at 55 kDa, so this was discarded. The third eluent, containing 0.25 M NaCl, was used to elute for 5 column volumes (1000 mL). A high-intensity absorption peak appeared, and approximately 600 mL of the main peak fraction was collected. 10 μL was analyzed by SDS-PAGE. A single main band appeared at 55 kDa, and ImageJ grayscale scanning quantification showed a target protein purity of 88.5%. The fourth eluent, containing 0.50 M NaCl, was used to elute for 3 column volumes (600 mL). A weak peak appeared, corresponding to a strongly bound acidic host protein, and this was discarded. The total protein concentration of the main peak fraction in the third stage was determined to be 8.5 mg / mL by the BCA method, and the total protein content was approximately 5.1 g. Based on the target protein purity of 88.5%, the target protein content was estimated to be approximately 4.5 g, and the recovery rate of step S3 was 72.6% (based on the target protein content of 6.2 g in the flow-through solution of step S2).

[0029] The procedure for step S4 is as follows: 600 mL of the ion exchange purification solution collected in step S3 is transferred into a stirred ultrafiltration cup (Millipore Amicon 8400 model, effective membrane area 10 kDa) equipped with an Ultracel 10 kDa ultrafiltration membrane. The sample was concentrated to 15 mL at 4 °C and 0.2 MPa nitrogen pressure, resulting in a protein concentration of 3.2 mg / mL. Sephacryl S-300HR packing material was loaded into an XK 26 / 60 chromatography column (2.6 cm inner diameter, 60 cm height, approximately 318 mL column volume). A 20 mM sodium phosphate buffer (pH 7.0) containing 150 mM NaCl was used as the run buffer, and two column volumes were equilibrated at a flow rate of 0.5 mL / min until baseline stability was achieved. 15 mL of the concentrated sample (4.7% of the column volume, meeting the requirement of no more than 5% loading) was injected into the top of the column using the aforementioned run buffer via a quantitative loop. The column was run at a flow rate of 0.5 mL / min, and the eluent was continuously detected using UV at 280 nm. 3 mL was collected from each tube.

[0030] Two well-separated absorption peaks appeared on the elution chromatogram. The first peak appeared at an elution volume of approximately 145–175 mL (corresponding to Kav 0.22–0.35), with a relatively high peak height; the second peak appeared at an elution volume of approximately 210–260 mL (corresponding to Kav 0.55–0.72), with a relatively low peak height. The fraction corresponding to the tube number of the first peak was collected, and the combined volume was approximately 30 mL. Samples of the first and second peak fractions were taken separately for reducing SDS-PAGE analysis. The first peak fraction showed a single main band at 55 kDa, with no visible impurities, and the purity reached 96.8% after grayscale scanning. The second peak fraction showed multiple bands of impurities (molecular weight distribution in the range of 25–65 kDa), with only a very weak trace band at 55 kDa, confirming that the main body of the second peak consisted of residual host protein impurities and a small amount of target protein degradation fragments. The first peak fraction was further analyzed by size exclusion high-performance liquid chromatography (SEC-HPLC, TSKgel G3000SWxl column, 0.5 mL / min, with the same running buffer). A sharp, symmetrical peak appeared at a retention time of approximately 12.3 min. Calibrated with globular protein molecular weight standards, the corresponding molecular weight range was 300-500 kDa (since collagen trimers are rigid rod-shaped molecules, their hydrodynamic radius is larger than that of globular proteins of the same molecular weight; the apparent molecular weight calculated by SEC-HPLC using globular protein standards is higher than the theoretical molecular weight of the trimer, 165 kDa). This peak area accounted for 92.5% of the total protein peak area. A small peak appeared at a retention time of approximately 16.8 min, corresponding to single-chain or partially dissociated dimers, with a peak area accounting for 7.5%. The total protein concentration of the first peak fraction was determined by the BCA method to be 1.35 mg / mL, with a total protein content of approximately 4.05 g. The recovery rate in step S4 was 90.0% (based on the target protein of 4.5 g in step S3).

[0031] The procedure for step S5 is as follows: Transfer the 30 mL gel filtration purified solution collected in step S4 into an Amicon 8010 stirred ultrafiltration cup (equipped with an Ultracel 30 kDa regenerated cellulose ultrafiltration membrane, effective membrane area...). Ultrafiltration concentration was initiated at 4℃ and 0.15MPa nitrogen pressure. When the volume of the feed solution in the ultrafiltration vessel was concentrated from 30mL to half of its original volume, i.e., 15mL (determined by reading the cumulative permeate volume as 15mL on the collection tube scale), pressurization was paused, the top of the ultrafiltration vessel was opened, and 15mL of pre-cooled conformationally stable dialysis buffer (5mM glycine, 0.5% (w / v) trehalose, 10mM sodium phosphate, pH 7.0) was added. After mixing with magnetic stirring, the concentration was re-pressurized. When the feed solution volume was concentrated from 30mL to 15mL again, the first round of concentration-dilution was completed. The second and third rounds of concentration-dilution were performed in the same manner, for a total of 3 rounds. The essence of the 3 rounds of concentration-dilution is to gradually replace the high concentration of NaCl in the gel filtration running buffer with conformationally stable dialysis buffer. Theoretically, each round reduces the NaCl concentration to half of the previous round, and the theoretical residual ratio after 3 rounds is (1 / 2). 3 =12.5%. The conductivity of the concentrate after the third round of concentration was measured to be 1.2 mS / cm, corresponding to a NaCl residual concentration of approximately 8-10 mM (the NaCl concentration in the running buffer in step S4 was 150 mM, and the actual residual rate was approximately 5.3-6.7%, slightly higher than half of the theoretical value of 12.5%, because the ultrafiltration membrane's NaCl rejection rate is not zero, and the actual NaCl permeation rate is approximately 85-90%).

[0032] The concentrated solution after crude desalting was transferred into a dialysis bag (Spectra / Por 7, 25mm wide) with a molecular weight cutoff of 10kDa. The bag was clamped at both ends with metal clamps and placed in a 1L beaker containing 500mL of conformationally stable dialysis buffer. The bag was then slowly stirred with a magnetic stirrer at 80-100rpm in a 4°C freezer (without vortex generation). The dialysis buffer was changed every 4 hours (each time, the dialysis bag was removed, the old dialysis buffer was discarded, and 500mL of fresh, pre-cooled conformationally stable dialysis buffer was added before reinserting the bag). This process was repeated 3 times, for a total dialysis time of 12 hours. During dialysis, NaCl diffused from the inside of the dialysis bag through the semi-permeable membrane into the dialysis buffer outside the bag. The ratio of dialysis buffer volume to the volume of liquid inside the bag was approximately 500mL:15mL ≈ 33:1. The theoretical equilibrium NaCl concentration after each dialysis buffer change was 1 / 34 of the concentration inside the bag. After 3 changes, the NaCl concentration inside the bag theoretically dropped to an extremely low level.

[0033] The role of glycine in conformationally stable dialysis buffer is as follows: Glycine is the most frequently occurring amino acid in the GXY repeating sequence of collagen (one in every three residues). Free glycine molecules competitively interact with the carbonyl oxygen of glycine residues on the triple helix surface through hydrogen bonds, mimicking the hydrogen bond microenvironment at the collagen trimer interface. When the NaCl concentration in the buffer decreases significantly, weakening the ion shielding effect, this competitive protection by glycine maintains the integrity of the hydrogen bond network between triple helix chains. The role of trehalose is as follows: As a non-reducing disaccharide, trehalose forms a dense hydration layer on the protein surface through a preferential repulsion mechanism, increasing the free energy barrier for the transition of the protein from the folded to unfolded state. In the concentration polarization region of ultrafiltration (where the local protein concentration on the membrane surface is much higher than the bulk concentration), it effectively inhibits protein aggregation.

[0034] After dialysis, the fluid in the dialysis bag was removed, and the conductivity was measured to be 0.35 mS / cm, corresponding to a residual NaCl concentration of approximately 2.5 mM, meeting the requirement of being reduced to below 5 mM. After aseptic filtration through a 0.22 μm PES membrane, approximately 8 mL of the final product was obtained. The protein concentration was determined by the BCA method to be 4.8 mg / mL, with a total protein content of approximately 3.84 g. The recovery rate in step S5 was 94.8% (based on the target protein of 4.05 g in step S4).

[0035] The overall recovery rate of the complete process in this embodiment is: Step S2 (96.9%) × Step S3 (72.6%) × Step S4 (90.0%) × Step S5 (94.8%) = 60.0%.

[0036] The final product was analyzed by 12% reduced SDS-PAGE electrophoresis, showing a single main band at 55 kDa. After Coomassie Brilliant Blue G-250 staining and ImageJ grayscale analysis, the purity was 96.8%. SEC-HPLC analysis under non-denaturing conditions showed that the trimer peak area accounted for 92.5% of the total protein peak area. Western blotting (primary antibody: rabbit anti-human type III collagen polyclonal antibody, dilution 1:5000; secondary antibody: HRP-labeled goat anti-rabbit IgG, dilution 1:10000; DAB staining) showed a specific positive band at 55 kDa. A negative control, the product obtained by purifying unconverted GS115 yeast lysate using the same procedure, showed no band.

[0037] The final product was analyzed by 8% non-reducing SDS-PAGE electrophoresis (without β-mercaptoethanol or DTT; the sample was treated with sample buffer containing 1% SDS for 10 min at room temperature without heating or boiling). Under non-reducing and non-heating conditions, the interchain hydrogen bonds and interchain hydrophobic interactions in the trimer were partially maintained, and the trimer migrated in a complete or partially dissociated form. The electrophoresis results showed a main band at approximately 180 kDa (corresponding to the partially dissociated intermediate state of the trimer; due to the denaturing effect of SDS, the electrophoretic mobility of the trimer under non-reducing conditions was between that of the complete trimer and the single chain), and a weak band at 55 kDa (corresponding to the completely dissociated single chain). The grayscale ratio of the 180 kDa main band was 84.3%, and that of the 55 kDa weak band was 15.7%. After boiling the same product at 95°C for 5 minutes, non-reducing SDS-PAGE was performed. The 180 kDa band completely disappeared, and a single main band (accounting for 98.2%) appeared at 55 kDa, confirming that the 180 kDa band was indeed a trimer rather than a covalently cross-linked protein aggregate (covalently cross-linked aggregates do not dissociate due to boiling). This result is consistent with the 92.5% trimer content detected by SEC-HPLC (the 84.3% trimer content in non-reducing SDS-PAGE is lower than the 92.5% in SEC-HPLC because the denaturation effect of 1% SDS during SDS-PAGE sample preparation caused some trimers to dissociate before electrophoresis, while SEC-HPLC, running in non-denaturing buffer, kept the trimers intact).

[0038] The final product was diluted to 0.5 mg / mL with conformationally stable dialysis buffer and subjected to far-UV CD spectroscopy scanning using a Jasco J-815 circular dichroism chromatograph with a 0.1 cm path length quartz cuvette. The scanning wavelength range was 190-260 nm, the scan speed was 50 nm / min, the bandwidth was 1.0 nm, the response time was 1 s, and four scans were accumulated and averaged. The temperature was controlled at 20 °C, and the same batch of dialysis buffer was used as the blank baseline. A positive peak (ellipticity) appeared at 221 nm in the CD spectrum. = +3850 deg· A negative peak appears at 197nm. = -28600 deg· The 221nm positive peak is a characteristic signal of the triple helix conformation of collagen, generated by the polyproline type II helix (PPII) conformation where every three residues form a repeating unit. Its positive peak position (221±1nm) and the ratio of positive to negative peaks Rpn = |[θ] 221 / [θ] 197The value of | =0.135 falls within the reported Rpn value range (0.10-0.15) for both natural type III collagen and recombinant collagen. The globular proteins and random coiled peptides do not exhibit a positive peak at 221 nm. This CD spectroscopy result confirms that the trimers in the purified product are indeed assembled using a triple-helix hydrogen bond network, rather than being non-specific aggregates of proteins.

[0039] Thermal denaturation experiments were conducted using the 221 nm signal from the CD spectrum as a probe. The temperature was increased from 20 °C to 80 °C at a rate of 1 °C / min, and continuous monitoring was performed. The change in [θ] occurs within the range of 20℃ to 42℃. 221 Remain stable; within the range of 42℃ to 58℃, from +3850 deg· It monotonically decreases to near 0; above 58℃, [θ] 221 The curve tends to flatten. The thermal transition temperature Tm is defined as the midpoint of the thermal denaturation curve, and the temperature corresponding to the minimum value of the first derivative dθ / dT curve is taken as Tm = 48.5℃. This Tm value is consistent with the reported triple-helix unwinding temperature range (45-52℃) of recombinant human type III collagen in the literature, further confirming that the target protein exists in a triple-helix conformation. This Tm value also explains, conversely, why the trimer content decreased to 68.4% after heat treatment at 65℃ in Comparative Experiment 2. 65℃ is much higher than the triple-helix Tm value (48.5℃), and the heat treatment led to irreversible triple-helix unwinding and interchain dissociation.

[0040] Comparative Experiment 1: Using the same yeast fermentation cells and the same cell disruption conditions in step S1, but omitting the ConA negative selection step in step S2, the crude extract was directly replaced with buffer and loaded into a DEAE column for step S3. The remaining parameters were the same as in Example 1. In the 0.25M NaCl elution section of the DEAE column, the target protein peak severely overlapped with a large number of glycoprotein impurity peaks. After collecting the main peak fraction, the SDS-PAGE purity was only 62.3% (compared to 88.5% in Example 1). Subsequent gel filtration also resulted in partial overlap between the target protein peak and impurity peaks due to excessive impurity load, ultimately leading to a product purity of only 81.7%, failing to meet the 95% quality standard. The overall recovery rate was 42.3% (lower than 60.0% in Example 1). This was because impurity co-purification forced a narrowing of the collection window at each fraction collection step, sacrificing recovery for purity.

[0041] Comparative Experiment 2: Using the same yeast fermentation cells and the same cell disruption conditions in step S1, the ConA negative selection step in step S2 was omitted. The crude extract was heat-treated at 65℃ for 30 min (referring to the method in patent CN107033238B) to denature the host protein. After centrifugation to remove the denatured protein precipitate, steps S3-S5 were performed. The total protein content in the supernatant after heat treatment was 15.3 g (29.4% of the 52.1 g crude extract, indicating that 70.6% of the protein was removed by heat precipitation). SDS-PAGE analysis showed that the target protein band intensity at 55 kDa was significantly weaker than that in the crude extract. The target protein content was estimated to be 3.1 g by grayscale scanning, indicating that heat treatment caused approximately 51.6% of the target protein to be lost through co-precipitation with the denatured host protein. After subsequent steps S3-S5, the final product purity was 93.2%, and the total recovery rate was only 29.8%. The trimer content was 68.4% by SEC-HPLC analysis (lower than 92.5% in Example 1), indicating that the 65°C heat treatment caused a considerable proportion of the trimer to unwind into single chains, consistent with the Tm=48.5°C obtained from the CD thermal denaturation experiment above.

[0042] Comparative Experiment 3: Using the same yeast fermentation cells, the ion exchange purified solution (600 mL, target protein purity 88.5%, target protein content approximately 4.5 g) was obtained according to steps S1-S3 of Example 1. The gel filtration molecular weight fractionation step in step S4 was omitted, and the ion exchange purified solution was directly subjected to the ultrafiltration concentration and dialysis desalting operation in step S5. The 600 mL ion exchange purified solution was concentrated to 15 mL using a regenerated cellulose ultrafiltration membrane with a molecular weight cutoff of 30 kDa (Millipore Ultracel 30 kDa, Amicon 8400 stirred ultrafiltration cup) at 4°C and 0.2 MPa nitrogen pressure. Then, desalting was performed according to the three rounds of concentration-dilution operation and 12 h dialysis operation in step S5 of Example 1.

[0043] The final product was analyzed by 12% reduced SDS-PAGE electrophoresis. The grayscale purity of the main band at 55 kDa was 88.2%, which was basically consistent with the 88.5% purity in the ion exchange stage. However, several weak bands with molecular weights distributed in the 25-50 kDa range appeared below the 55 kDa main band, corresponding to host protein impurities and target protein degradation fragments that were not completely removed in step S3. Since the molecular weight cutoff of the 30 kDa ultrafiltration membrane is higher than that of most impurity proteins in the 25-50 kDa range, these impurity proteins were retained by the ultrafiltration membrane and co-concentrated with the target protein. The product purity of 88.2% did not meet the 95% quality standard. SEC-HPLC analysis of the product showed that the trimer peak area accounted for 72.3% of the total protein peak area, significantly lower than 92.5% in Example 1. Two distinct shoulder peaks appeared after the trimer peak, corresponding to the dimer / partially dissociated intermediate and the single chain, respectively. The results indicate that residual host proteases (protease impurities not completely removed in the ion-exchange purification step) in the ion-exchange purification solution continue to degrade the target protein during the prolonged concentration and dialysis process in step S5, leading to partial dissociation of the trimer into single chains. In contrast, the molecular weight fractionation of gel filtration chromatography in step S4 of Example 1 separates and removes host proteases and degradation fragments with molecular weights below 100 kDa from the trimer fraction, significantly reducing the risk of target protein degradation in subsequent step S5. The total protein content of the final product, determined by the BCA method, was 3.72 g, and the recovery rate in step S5 was 82.7% (based on 4.5 g of target protein in the ion-exchange purification solution), lower than the 94.8% in step S5 of Example 1. The main reasons for the reduced recovery rate are twofold: when concentrating protein solutions with high impurity loads using a 30 kDa ultrafiltration membrane, impurity proteins form a gel polarization layer on the membrane surface, increasing filtration resistance and irreversible protein adsorption loss; and degradation fragments smaller than 30 kDa generated by protease degradation permeate through the ultrafiltration membrane.

[0044] The three comparative experiments above demonstrate that the ConA negative selection step of this invention selectively removes approximately 60% of yeast glycoprotein impurities without damaging the target protein, significantly improving the resolution of subsequent ion exchange chromatography. Step S4, gel filtration chromatography, plays two irreplaceable roles in the overall purification process: molecular weight-based physical fractionation (increasing purity from 88.5% to 96.8%) and removal of residual protease impurities (protecting the trimer from degradation in subsequent steps). Compared to traditional direct chromatography methods that omit the ConA step, heat treatment methods, and methods that omit gel filtration, the complete five-step process of this invention has significant advantages in purity, recovery rate, and trimer retention rate.

[0045] Figure 2 This is a schematic diagram comparing the purification effects of Example 1 and Comparative Experiments 1 to 3 in this application. Figure 2The results of comparing Example 1 with Comparative Experiments 1, 2, and 3 in terms of three key indicators: purity, total recovery rate, and trimer content are shown. Figure 2 It is evident that the complete process of Example 1 is significantly superior to the comparative experiments in terms of purity (96.8%) and trimer content (92.5%), indicating that the concanavalin A affinity chromatography step plays an irreplaceable role in removing glycoprotein impurities from yeast host, while the gel filtration chromatography step is crucial for enriching trimer configurations and improving product uniformity. Omitting or replacing any key step will lead to a significant decrease in purity and / or trimer content, verifying the synergistic necessity among the steps of the complete separation and purification process described in this application.

[0046] Example 2 This embodiment provides a method for isolating and purifying recombinant type III collagen, using the same Pichia pastoris engineered fermentation culture as raw material. The operating temperature is controlled between 2 and 8°C throughout the process.

[0047] In step S1, 400g of yeast cell precipitate (wet weight) was taken and resuspended in 2400mL of lysis buffer (20mM sodium phosphate buffer, pH 7.0, containing 100mM sodium chloride, 0.5mM EDTA, and 0.5mM PMSF) (6mL buffer per gram of wet cell weight). The hydrolysis half-life of PMSF in aqueous solution varies with pH and temperature, and is approximately 55min at pH 7.0 and 4℃ (longer than the approximately 35min half-life at pH 7.4 in Example 1). In this example, the relatively low initial concentration of 0.5mM maintained an effective concentration above 0.25mM during homogenization (4 cycles, total time approximately 40min), which was sufficient to inhibit the serine protease activity in the Pichia pastoris lysate. The PMSF stock solution (100mM isopropanol solution) was removed from the -20℃ freezer and added directly to the lysis buffer within 5min before use to ensure an effective concentration. Four cycles were performed at a homogenizing pressure of 500 bar, with the material temperature controlled at 2–4°C after each cycle via a heat exchanger. After two stages of centrifugation (8000g for 20 min and 12000g for 30 min) and filtration through a 0.45 μm polyethersulfone membrane, approximately 2200 mL of clear crude extract was obtained, with a total protein concentration of 14.2 mg / mL (total protein weight 31.2 g) and a target protein purity of 11.8% (approximately 3.7 g) based on grayscale scanning. The cell disruption rate, as measured by microscopy, was 96.2%. Compared to Example 1 (700 bar, 3 cycles), adding one cycle at a lower pressure of 500 bar also achieved a cell disruption rate of over 95%.

[0048] In step S2, 220 mL of ConA-Sepharose 4B wet gel was loaded into an XK 50 / 20 column, using a 15 mM Tris-HCl buffer (pH 7.0) containing 0.5 mM calcium chloride, 0.5 mM manganese chloride, and 0.3 M sodium chloride as the equilibration buffer. The crude extract was loaded at a linear velocity of 20 cm / h. Approximately 2350 mL of flow-through was collected, with a total protein concentration of 5.8 mg / mL (total protein weight 13.6 g), a reduction of 56.4% compared to the crude extract. SDS-PAGE analysis showed that the target protein grayscale purity was 25.3% (approximately 3.4 g), with a recovery rate of 91.9%. The recovery rate was slightly lower than that in Example 1 (96.9%) because the NaCl concentration in the equilibration buffer was lower (0.3 M vs 0.5 M), resulting in a slight increase in non-specific ion adsorption and causing a small amount of target protein to be non-specifically retained. However, the 91.9% recovery rate is still within the acceptable range of the process. After flow-through collection, the ConA column was regenerated with equilibration buffer containing 0.1 M methyl-α-D-mannoside.

[0049] In step S3, the flow-through buffer system was replaced with 15 mM Tris-HCl (pH 7.8) using an ultrafiltration membrane with a molecular weight cutoff of 10 kDa. After five volume-fold permeation, the conductivity decreased to 0.9 mS / cm (NaCl concentration approximately 7 mM). A DEAE-Sepharose Fast Flow column (150 mL) was loaded, and a four-stage gradient elution was performed (0.03 M / 0.10 M / 0.20 M / 0.40 M NaCl, each elution stage using 15 mM Tris-HCl pH 7.8 as the base buffer). Under pH 7.8 conditions (lower than pH 8.0 in Example 1), the binding strength of the target protein to DEAE was slightly weaker, causing the target protein to begin eluting in the third stage (0.20 M NaCl elution). Approximately 500 mL of the third peak fraction was collected, with an SDS-PAGE purity of 85.2%, a target protein content of approximately 2.4 g, and a recovery rate of 70.6%.

[0050] In step S4, 500 mL of ion-exchange purification buffer was concentrated to 12 mL (protein concentration 2.1 mg / mL), loaded onto a Sephacryl S-300 HR column (XK 26 / 60, running buffer containing 100 mM NaCl and 10 mM sodium phosphate, pH 6.8), and approximately 28 mL of the fraction with a Kav value in the range of 0.18–0.38 was collected. The SDS-PAGE purity was 95.3%, and the SEC-HPLC trimer content was 91.2%. The protein content was approximately 2.1 g, with a recovery rate of 87.5%.

[0051] In step S5, 28 mL of gel filtration purification solution was concentrated using a 30 kDa regenerated cellulose ultrafiltration membrane. When the solution volume was reduced from 28 mL to half of its original volume (14 mL), 14 mL of pre-cooled conformational stabilization dialysis buffer (3 mM glycine, 0.3% (w / v) trehalose, 8 mM sodium phosphate, pH 6.8) was added. After mixing, the solution was concentrated to 14 mL, completing the first round of concentration-dilution. The second round was performed in the same manner, for a total of two rounds of concentration-dilution and crude desalting. After crude desalting, the solution was transferred to a dialysis bag with a molecular weight cutoff of 10 kDa and dialyzed against 500 mL of conformational stabilization dialysis buffer for 12 h (the buffer was changed every 4 h, for a total of 3 times). After dialysis, the conductivity was 0.42 mS / cm, and the NaCl concentration was approximately 3.0 mM. The final product protein content was 1.91 g, and the recovery rate in step S5 was 91.0%. The total recovery rate was 91.9% × 70.6% × 87.5% × 91.0% = 51.6%, the purity was 95.3%, and the trimer content was 91.2%.

[0052] This embodiment demonstrates that, under the lower limit of the parameter range (500 bar homogenization, 4 cycles, 0.5 mM PMSF, 0.3 M NaCl equilibration solution, pH 7.8 ion exchange, 3 mM glycine / 0.3% trehalose dialysis solution, 2 rounds of concentration-dilution), a qualified product with a purity of not less than 95% and a trimer content of not less than 90% can still be obtained.

[0053] Example 3 This embodiment provides a method for isolating and purifying recombinant type III collagen, using the same Pichia pastoris engineered fermentation culture as raw material. The operating temperature is controlled at 2–8℃ throughout the process.

[0054] In step S1, 500g of wet bacterial cells were suspended in 2000mL of lysis buffer (50mM sodium phosphate buffer, pH 7.8, containing 200mM sodium chloride, 2mM EDTA, and 2mM PMSF) (4mL buffer per gram of wet bacterial cells). The mixture was cycled twice at 800 bar, with the material temperature controlled at 6–8°C after each cycle via a heat exchanger. After two stages of centrifugation (6000g for 25 min and 10000g for 35 min) and filtration through a 0.22μm polyethersulfone membrane, approximately 1900mL of crude extract was obtained. The cell disruption rate was 95.8%. Compared to the low-pressure, high-cycle combination in Example 2, the high-pressure, low-cycle combination achieved a comparable cell disruption rate but with a shorter processing time (2 cycles vs. 4 cycles), verifying the equivalence of the two parameter combinations. The total protein concentration was 22.4 mg / mL (total protein amount 42.6 g), and the purity of the target protein in grayscale scanning was 13.1% (approximately 5.6 g).

[0055] In step S2, 190 mL of ConA-Sepharose 4B was packed into a column and equilibrated with 25 mM Tris-HCl buffer (pH 7.8) containing 1.0 mM calcium chloride, 1.0 mM manganese chloride, and 0.5 M sodium chloride. The crude extract was loaded at a linear velocity of 40 cm / h. The total protein content of the flow-through was 16.8 g (a reduction of 60.6%), the target protein ash purity was 31.5% (approximately 5.3 g), and the recovery rate was 94.6%. The column was regenerated with equilibration buffer containing 0.3 M methyl-α-D-mannoside.

[0056] In step S3, the flow-through was permeated eight times by volume to a 25 mM Tris-HCl (pH 8.5) buffer system using an ultrafiltration membrane with a molecular weight cutoff of 12 kDa, reducing the conductivity to 0.4 mS / cm (NaCl concentration approximately 3 mM). The solution was then loaded onto a DEAE-Sepharose 6Fast Flow column (180 mL) and subjected to a four-stage gradient elution (0.08 M / 0.20 M / 0.35 M / 0.60 M NaCl, each elution using 25 mM Tris-HCl pH 8.5 as the base buffer). At the higher pH of 8.5, the binding strength of the target protein to DEAE was enhanced. The SDS-PAGE purity of the main peak fraction eluted in the third stage (0.35 M NaCl) reached 91.2%, with a target protein content of approximately 4.2 g and a recovery rate of 79.2%, higher than that of Example 2 (70.6%).

[0057] In step S4, the ion-exchange purification buffer was concentrated to 18 mL (protein concentration 4.8 mg / mL), loaded onto a Superdex 200 prep-grade column (XK 26 / 60, running buffer containing 200 mM NaCl and 25 mM sodium phosphate, pH 7.4), and approximately 25 mL of the fraction with a Kav range of 0.15-0.35 was collected. The SDS-PAGE purity was 97.5%, and the SEC-HPLC trimer content was 93.8%. The protein content was approximately 3.8 g, with a recovery rate of 90.5%. The Superdex 200 prep-grade packing material has a higher resolution than the Sephacryl S-300 HR used in Example 1, which is reflected in higher purity and trimer content in this example.

[0058] In step S5, 25 mL of the gel filtration purified solution was concentrated using a 30 kDa regenerated cellulose ultrafiltration membrane. Four rounds of concentration-dilution desalting were performed, adding an equal volume of conformation-stabilized dialysis buffer (8 mM glycine, 0.8% (w / v) trehalose, 15 mM sodium phosphate, pH 7.4) each time the solution was concentrated to half its original volume. The theoretical NaCl residual ratio after four rounds of operation was (1 / 2). 4=6.25%, measured conductivity 0.85 mS / cm. Transferred to a dialysis bag with a molecular weight cutoff of 12 kDa and dialyzed for 16 h (medium changed every 5 h, for a total of 3 times). After dialysis, the conductivity was 0.28 mS / cm, and the NaCl concentration was approximately 2.0 mM. The final product protein content was 3.54 g, and the recovery rate in step S5 was 93.2%. The total recovery rate was 94.6% × 79.2% × 90.5% × 93.2% = 63.2%, the purity was 97.5%, and the trimer content was 93.8%.

[0059] This embodiment demonstrates that the process is equally effective even under the upper limit of the parameter range (800 bar homogenization, 2 cycles, 2 mM PMSF, pH 8.5 ion exchange, DEAE-Sepharose 6 Fast Flow packing material, Superdex 200 prep grade packing material, 8 mM glycine / 0.8% trehalose dialysis solution, 4 rounds of concentration-dilution, 12 kDa dialysis bag, 16 h dialysis). Furthermore, with the higher ion exchange pH of pH 8.5 and the combination of Superdex 200 high-resolution packing material, higher purity (97.5%) and trimer content (93.8%) can be obtained.

[0060] Example 4 This example is used to verify the necessity of glycine and trehalose components in conformationally stable dialysis buffer.

[0061] Following the same steps S1-S4 as in Example 1, approximately 30 mL of gel filtration purification buffer (protein concentration 1.35 mg / mL) was obtained. The purification buffer was divided into four equal portions (approximately 7.5 mL each), and each portion was subjected to step S5 using one of four different dialysis buffers.

[0062] Group A (complete formulation): 10mM sodium phosphate buffer (pH 7.0) containing 5mM glycine and 0.5% (w / v) trehalose was used as the dialysate. The operation method was the same as step S5 in Example 1 (dialysis for 12 hours after 3 rounds of concentration-dilution and crude desalting, with the solution changed every 4 hours).

[0063] Group B (without glycine): Use 10mM sodium phosphate buffer (pH 7.0) containing 0.5% (w / v) trehalose as the dialysate, and perform the other procedures as in Group A.

[0064] Group C (without trehalose): 10mM sodium phosphate buffer (pH 7.0) containing 5mM glycine was used as the dialysate, and the rest of the operation was the same as Group A.

[0065] Group D (without glycine and trehalose): Use 10mM sodium phosphate buffer (pH 7.0) as the dialysate, and perform the same procedures as Group A.

[0066] The test results of the processed products in each group are as follows.

[0067] Trimer content was analyzed by SEC-HPLC: Group A was 92.5%, Group B was 84.3%, Group C was 89.6%, and Group D was 76.8%. Group D showed the largest decrease in trimer content (from 92.5% in the gel filtration stage to 76.8%), indicating that in the absence of any stabilizers, the dramatic change in ionic strength during ultrafiltration concentration and dialysis desalination led to approximately 15.7% of the trimers unwinding into single chains. Group B (without glycine) had a trimer content of 84.3%, a decrease of 8.2 percentage points compared to Group A (92.5%), indicating that glycine significantly contributes to maintaining the triple-helix hydrogen bond network. Group C (without trehalose) had a trimer content of 89.6%, a decrease of 2.9 percentage points compared to Group A, indicating that the anti-aggregation protective effect of trehalose is also significant.

[0068] Visual inspection of product clarity revealed that Group A and Group B were clear and transparent; Group C showed a slight opalescence; and Group D exhibited a distinct opalescence and a small amount of white flocculent precipitate. After filtration through a 0.22 μm membrane, the membrane weight gain was measured: Group A was 0 mg (no precipitate), Group B was 0 mg, Group C was 0.3 mg, and Group D was 1.2 mg. This indicates that trehalose is a key component in preventing protein aggregation and precipitation. Groups C and D, which do not contain trehalose, showed varying degrees of protein aggregation, while Group B, although lacking glycine, contained trehalose and did not exhibit aggregation.

[0069] The recovery rates of step S5 were calculated based on protein recovery rates: Group A was 94.8%, Group B was 93.5%, Group C was 88.7%, and Group D was 81.4%. The lower recovery rates of Groups C and D are due to losses caused by protein aggregation and precipitation.

[0070] This example demonstrates that glycine and trehalose each play distinct protective roles in conformationally stable dialysis buffer, and neither can be dispensed with. Glycine primarily maintains the integrity of the trimer triple helix structure (preventing unwinding), while trehalose primarily prevents protein aggregation and precipitation during concentration (preventing aggregation). Their synergistic effect ensures that the final product simultaneously meets the quality standards of a trimer content of not less than 90% and a purity of not less than 95%.

[0071] Example 5 This embodiment is used to verify the stability of the ConA column in step S2 for repeated use.

[0072] Using the same ConA-Sepharose 4B column (280 mL, XK 50 / 30) as in Example 1, five independent batches of crude yeast extract (approximately 2800 mL per batch, from different fermenter batches under the same fermentation conditions) were processed in step S2 with the equilibration buffer and regeneration conditions of Example 1. After each batch, elution and regeneration were performed with equilibration buffer containing 0.2 M methyl-α-D-mannoside and reequilibration was performed with equilibration buffer. The glycoprotein rejection rate (characterized as a percentage reduction in total protein) and the target protein recovery rate were recorded for each batch.

[0073] Batch 1: Glycoprotein rejection rate 59.9%, target protein recovery rate 96.9%. Batch 2: Glycoprotein rejection rate 58.7%, target protein recovery rate 96.2%. Batch 3: Glycoprotein rejection rate 57.3%, target protein recovery rate 95.8%. Batch 4: Glycoprotein rejection rate 56.1%, target protein recovery rate 95.4%. Batch 5: Glycoprotein rejection rate 55.6%, target protein recovery rate 95.1%.

[0074] In five batches, the glycoprotein rejection rate slowly decreased from 59.9% to 55.6% (a decrease of 4.3 percentage points), and the target protein recovery rate slowly decreased from 96.9% to 95.1% (a decrease of 1.8 percentage points). The slow decrease in rejection rate is related to the partial inactivation of ConA tetramer after multiple uses, but the rejection rate remained above 55% after five uses. Quality testing was performed on the final products from the five batches after complete steps S1-S5. The purities were 96.8%, 96.5%, 96.1%, 95.8%, and 95.3%, respectively, and the trimer contents were 92.5%, 92.1%, 91.8%, 91.5%, and 91.2%, respectively, all meeting the quality standards of a purity of not less than 95% and a trimer content of not less than 90%. This example demonstrates that ConA columns can be reused at least five times on the same batch of packing material without affecting the quality of the final product.

[0075] Example 6 This example is used to verify the biological activity of the final product.

[0076] The purified product obtained in step S5 of Example 1 was subjected to the following in vitro biological activity tests.

[0077] Anti-inflammatory activity was assessed using human immortalized keratinocytes (HaCaT) as a model. HaCaT cells were cultured at a density of 5 × 10⁶ cells / year. 4 Seeds were planted at a density of 10 cells / well in 96-well plates and incubated in DMEM medium containing 10% fetal bovine serum at 37°C. Cells were cultured for 24 hours under controlled conditions to allow them to adhere to the culture medium. After discarding the medium, an inflammation model was established by treatment with serum-free DMEM medium containing 10 μg / mL lipopolysaccharide (LPS, derived from *E. coli* O111:B4). Simultaneously with LPS stimulation, purified recombinant type III collagen was added at concentrations of 1 μg / mL, 5 μg / mL, and 10 μg / mL, respectively. A blank control group (without LPS and collagen) and a model group (with LPS only) were established. Each group had three replicates. After 24 hours of culture, the culture supernatant was collected, and the concentrations of IL-6, IL-8, and TNF-α were detected by enzyme-linked immunosorbent assay (ELISA).

[0078] IL-6 detection results: The blank control group was 52.3±8.7 pg / mL, the model group was 826.4±42.3 pg / mL, the 1 μg / mL treatment group was 638.5±35.6 pg / mL (inhibition rate 22.7%), the 5 μg / mL treatment group was 456.2±28.4 pg / mL (inhibition rate 44.8%), and the 10 μg / mL treatment group was 284.7±31.5 pg / mL (inhibition rate 65.5%).

[0079] IL-8 detection results: The blank control group was 98.6±12.3 pg / mL, the model group was 1542.8±68.7 pg / mL, the 1 μg / mL treatment group was 1205.3±55.2 pg / mL (inhibition rate 21.9%), the 5 μg / mL treatment group was 836.7±42.8 pg / mL (inhibition rate 45.8%), and the 10 μg / mL treatment group was 523.6±45.2 pg / mL (inhibition rate 66.1%).

[0080] TNF-α detection results: The blank control group was 18.5±4.2 pg / mL, the model group was 368.5±22.1 pg / mL, the 1 μg / mL treatment group was 292.3±18.6 pg / mL (inhibition rate 20.7%), the 5 μg / mL treatment group was 198.4±15.3 pg / mL (inhibition rate 46.1%), and the 10 μg / mL treatment group was 142.3±18.7 pg / mL (inhibition rate 61.4%).

[0081] The inhibition rates of the three inflammatory factors were all significantly concentration-dependent, with the inhibition rates of the 10 μg / mL treatment group all exceeding 60%, demonstrating that the purified product maintained its anti-inflammatory biological activity.

[0082] The proliferative activity was assessed using human skin fibroblasts (HSF) as a model. HSF cells were cultured at a density of 3 × 10⁶ cells / year. 3Cells were seeded at a density of [number] cells / well in 96-well plates and cultured adherently for 24 h. Then, culture medium containing 0 μg / mL (blank control), 1 μg / mL, 5 μg / mL, and 10 μg / mL purified recombinant type III collagen were added, respectively. A positive control group containing 10 μg / mL bovine type I collagen (Sigma, C9791) was also included. Each group had 6 replicates. After 48 h of culture, cell proliferation activity was assessed using the CCK-8 assay (10 μL of CCK-8 reagent was added, and the cells were incubated at 37°C for 2 h; OD values ​​were read at 450 nm using a microplate reader).

[0083] The results were as follows: the OD450 value was 0.82±0.05 in the blank control group, 0.98±0.06 in the 1 μg / mL group (proliferation promotion rate 19.5%), 1.24±0.07 in the 5 μg / mL group (proliferation promotion rate 51.2%), 1.47±0.08 in the 10 μg / mL group (proliferation promotion rate 79.3%), and 1.18±0.06 in the bovine type I collagen control group (proliferation promotion rate 43.9%). The 10 μg / mL purified product showed significantly better proliferation-promoting effects than bovine type I collagen at the same concentration.

[0084] The migration-promoting activity assay also used HSF cells as a model. HSF cells were injected at a concentration of 1 × 10⁻⁶ cells. 5 Cells were seeded at a density of cells / well in 12-well plates and cultured until monolayer confluence. Using a 200 μL pipette tip, vertical straight scratches approximately 500 μm wide were made on the cell monolayer. After washing three times with PBS to remove detached cells, serum-free medium containing 10 μg / mL purified recombinant type III collagen, 10 μg / mL bovine type I collagen, and blank serum-free medium were added, respectively. Scratch widths were recorded at 0 h and 24 h using an inverted microscope, and the scratch area was measured and closure rate calculated using ImageJ software.

[0085] The results showed that the scratch closure rate at 24 h was 78.2%±4.3% in the 10μg / mL purified product group, 52.6%±5.1% in the bovine type I collagen group, and 31.4%±3.8% in the blank control group. The migration-promoting effect of the purified product was also superior to that of bovine type I collagen.

[0086] The above in vitro experimental results show that the recombinant type III collagen obtained by the separation and purification method of this invention maintains complete biological activity in terms of anti-inflammatory, proliferative and migration-promoting aspects, and its activity is superior to that of commercial bovine type I collagen.

[0087] Example 7 This example is used to verify the residual amount of yeast host cell protein (HCP) in the final product.

[0088] The final product obtained in step S5 of Example 1 was used to quantify residual host protein using the Pichia pastoris HCP ELISA kit (Cygnus Technologies, Cat. No. F600). Following the kit instructions, the final product was diluted with sample diluent to two concentration gradients: 100 μg / mL and 10 μg / mL, with three replicates for each concentration. A standard curve (0-100 ng / mL) was plotted using the Pichia pastoris HCP standard provided in the kit. OD values ​​were read at 450 nm and converted to HCP concentration.

[0089] The test results were as follows: At a dilution of 100 μg / mL, the HCP concentration was 12.8 ± 1.5 ng / mL, which translates to an HCP content of 128 ± 15 ng / mg in the original product (i.e., 128 ng of host protein per milligram of target protein); at a dilution of 10 μg / mL, the HCP concentration was 1.4 ± 0.3 ng / mL, which translates to an HCP content of 140 ± 30 ng / mg. The values ​​at the two dilutions showed good consistency (relative deviation 9.4%), and the average value was taken as 134 ng / mg.

[0090] Samples were taken from the crude extract (S1, target protein purity 12.3%), flow-through buffer (S2, purity 29.8%), ion-exchange purification buffer (S3, purity 88.5%), and gel filtration purification buffer (S4, purity 96.8%) in step S1 and analyzed using the same HCP ELISA. The HCP content in each step was as follows: (in crude extract...) (That is, approximately 7.13 mg HCP per milligram of total protein, consistent with the target protein purity of 12.3%), in the flow-through solution In ion exchange purification solution The HCP content in the gel filtration purification solution was 2150 ng / mg. From the crude extract to the final product, the HCP content decreased from... The concentration was reduced to 134 ng / mg, with a total removal factor of approximately 53,200 times (corresponding to an HCP removal rate of 99.998%). The HCP removal contributions of each chromatographic step were as follows: the ConA flow-through step removed 3.0 times (…). to The anion exchange step removed 18.1 times ( to ), gel filtration removes 60.5 times ( The removal rate was 16.0 times (from 2150 to 134) by ultrafiltration and dialysis steps. The single-step HCP removal by gel filtration contributed the most, further confirming the conclusion in comparative experiment 3 that omitting this step led to substandard purity.

[0091] The final product had a residual HCP content of 134 ng / mg of the target protein, which translates to 134 ppm. This level is lower than the general control requirements for host cell protein residues in recombinant protein drugs in the International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH) Q6B guidelines, indicating that the HCP removal capability of the purification process of this invention has reached the purification level of pharmaceutical-grade recombinant proteins.

[0092] Example 8 This example is used to verify the endotoxin and nucleic acid residue levels in the final product.

[0093] Endotoxin detection: The final product obtained in step S5 of Example 1 was used for bacterial endotoxin quantification using the LonzaPyroGene Recombinant Factor C Assay (Cat. No. 50-658U). The final product was diluted with pyrogen-free water to 1 mg / mL before detection. The negative control was pyrogen-free water, and the positive control was the standard endotoxin (0.005-50 EU / mL) provided in the kit. Two replicates were set for each sample, and a product inhibition / enhancement assay was also performed to exclude matrix interference.

[0094] The test results showed that the endotoxin content of the final product was 0.18 ± 0.05 EU / mg protein. Pichia pastoris is a eukaryotic microorganism whose cell wall does not contain lipopolysaccharide (LPS). The background endotoxin level in the crude extract derived from yeast was significantly lower than that of the Escherichia coli expression system. The multi-step chromatographic treatment and ultrafiltration dialysis in the purification process of this invention further reduced the endotoxin level. This detection value is lower than the endotoxin limit for injectable biological products in General Chapter 1143 of the 2025 edition of the Chinese Pharmacopoeia (generally not exceeding 10 EU / mg protein), indicating that the product meets the endotoxin control requirements for injectable grade products.

[0095] Nucleic acid residue detection: The final product was used to quantify double-stranded DNA residue using the Quant-iT PicoGreen dsDNA Detection Kit (Invitrogen, Cat. No. P7589). The final product was diluted to 0.5 mg / mL, and an equal volume of PicoGreen working solution was added. After incubation in the dark for 5 min, the fluorescence intensity was read using a fluorescence microplate reader (excitation 480 nm / emission 520 nm). A standard curve was plotted using λ DNA standards (0-1000 pg / mL).

[0096] The test results showed that the residual dsDNA in the final product was 8.5 ± 1.2 pg / mg protein. The nucleic acid removal pathway from the crude extract to the final product was as follows: Step S3, the first stage of anion exchange chromatography, involved low-salt elution (0.05M NaCl), which selectively removed weakly bound nucleic acid fragments; this was the main step in nucleic acid removal. Step S4, gel filtration chromatography, effectively separated small molecular weight oligonucleotide fragments (molecular weight below 30 kDa) from the trimer target protein (apparent molecular weight 300-500 kDa). Step S5, ultrafiltration, had a molecular weight cutoff of 30 kDa, and small molecular weight nucleic acid fragments were lost with the permeate during ultrafiltration. The residual dsDNA in the final product, 8.5 pg / mg, is lower than the World Health Organization (WHO) recommended limit for residual host cell DNA in biological products (10 ng / dose; calculated based on a dosage of 1 mg protein per dose, the residual DNA in this product is 0.0085 ng, far below the limit).

[0097] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for isolating and purifying recombinant type III collagen, characterized in that, Includes the following steps: S1. The recombinant yeast cells containing recombinant type III collagen were suspended in a buffer solution containing protease inhibitors. The resulting cell suspension was subjected to high-pressure homogenization to disrupt the cell wall. Cell fragments were removed by centrifugation to obtain a crude extract. S2. Pass the crude extract into a balanced concanavalin A agarose affinity chromatography column and collect the unbound components in flow-through mode. This allows high-mannose glycoprotein impurities from yeast host cells to specifically bind to and be retained by concanavalin A. Recombinant type III collagen without mannose N-glycan modification flows out of the column to obtain a deglycoprotein flow-through solution. S3. The deglycoprotein flow-through solution is replaced with buffer and loaded into an anion exchange chromatography column. It is eluted with sodium chloride in a fractional gradient and the elution fraction containing recombinant type III collagen is collected to obtain the ion exchange purified solution. S4. After concentrating the ion exchange purified solution, load it into a gel filtration chromatography column for molecular weight fractionation and collect the elution fraction corresponding to the molecular weight range of recombinant type III collagen trimer to obtain the gel filtration purified solution. S5. The gel filtration purification solution is concentrated by ultrafiltration and then desalted by dialysis with phosphate buffer containing glycine and trehalose to obtain recombinant type III collagen with a purity of not less than 95%.

2. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In step S1, the protease inhibitor is benzyl sulfonyl fluoride, and its concentration in the buffer solution is 0.5–2 mM; the buffer solution is a sodium phosphate buffer solution with a pH of 7.0–7.8, a concentration of 20–50 mM, and also contains 100–200 mM sodium chloride and 0.5–2 mM disodium ethylenediaminetetraacetate.

3. The method for isolating and purifying recombinant type III collagen as described in claim 1 or 2, characterized in that: In step S1, the specific operation of the high-pressure homogenization cell disruption treatment is as follows: the recombinant yeast cells are suspended at a ratio of 4-6 mL buffer per gram of wet cell weight, and then subjected to 2-4 cycles at a homogenization pressure of 500-800 bar. Between each cycle, the material temperature is controlled at 2-8°C using a heat exchanger. The centrifugation is a two-stage centrifugation. The first stage centrifugation is carried out at 6000-10000 g for 15-25 min to remove cell fragments and undisrupted cells. The supernatant of the first stage centrifugation is collected, and then the second stage centrifugation is carried out at 10000-15000 g for 25-35 min to remove cell membrane fragments and lipid particles. The supernatant of the second stage centrifugation is collected and filtered through a 0.22-0.45 μm polyethersulfone filter membrane to obtain the crude extract.

4. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In step S2, the concanavalin A agarose affinity chromatography column is equilibrated with a 15-25 mM Tris-HCl buffer containing 0.5-1.0 mM calcium chloride, 0.5-1.0 mM manganese chloride, and 0.3-0.5 M sodium chloride, at a pH of 7.0-7.

8. The crude extract is loaded into the column at a linear velocity of 20-40 cm / h. After flow-through collection, the concanavalin A agarose affinity chromatography column is eluted with an equilibration buffer containing 0.1-0.3 M methyl-α-D-mannoside to remove trapped glycoprotein impurities, allowing the column to be regenerated and reused.

5. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In step S3, the specific operation of the buffer replacement is as follows: the buffer system of the deglycoprotein flow-through solution is replaced with 15-25 mM Tris-HCl buffer solution, pH 7.8-8.5, using an ultrafiltration membrane with a molecular weight cutoff of 8-12 kDa, at a replacement factor of 5-8 times the original volume, so that the sodium chloride concentration in the flow-through solution is reduced to below 10 mM; the packing material of the anion exchange chromatography column is DEAE-Sepharose Fast Flow or DEAE-Sepharose 6 Fast Flow.

6. The method for isolating and purifying recombinant type III collagen as described in claim 5, characterized in that: In step S3, the sodium chloride fractional gradient elution includes the following four stages: the first stage elutes 3-6 column volumes with 0.03-0.08M sodium chloride to remove weakly bound nucleic acid fragments and low molecular weight peptides; the second stage elutes 3-6 column volumes with 0.10-0.20M sodium chloride to elute non-glycoprotein host proteins with moderate binding strength to the DEAE group; the third stage elutes 3-6 column volumes with 0.20-0.35M sodium chloride, collecting the fraction corresponding to the main absorption peak at 280nm UV as the eluted fraction containing recombinant type III collagen; and the fourth stage elutes 2-4 column volumes with 0.40-0.60M sodium chloride to remove strongly bound acidic host proteins.

7. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In step S4, the gel filtration chromatography column is packed with Sephacryl S-300 HR or Superdex 200 prep grade, with a separation range covering 10 kDa to 1500 kDa; the running buffer is a 10-25 mM sodium phosphate buffer containing 100-200 mM sodium chloride, pH 6.8-7.4; before being loaded into the gel filtration chromatography column, the ion exchange purification solution is concentrated to a protein concentration of 2-5 mg / mL using an ultrafiltration membrane with a molecular weight cutoff of 10 kDa; the concentrated sample is loaded into the gel filtration chromatography column using the running buffer as the carrier liquid; the sample loading volume does not exceed 5% of the gel filtration chromatography column volume.

8. The method for isolating and purifying recombinant type III collagen as described in claim 7, characterized in that: In step S4, the recombinant type III collagen is assembled into a trimer by three single chains with a molecular weight of 55 kDa through interchain hydrogen bonds under non-denaturing conditions. The theoretical molecular weight of the trimer is 165 kDa. The elution fraction with a partition coefficient Kav in the range of 0.15 to 0.40 in the gel filtration chromatography elution spectrum is collected as the gel filtration purification solution.

9. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In step S5, the ultrafiltration concentration is carried out using a regenerated cellulose ultrafiltration membrane with a molecular weight cutoff of 30 kDa, under nitrogen pressure conditions of 2–6°C and 0.1–0.3 MPa. The combined operation of ultrafiltration concentration and dialysis desalting is as follows: 2–4 rounds of concentration-dilution are performed by adding an equal volume of dialysate every time the concentration reaches 1 / 2 of the original volume, to perform coarse desalting of the gel filtration purified solution. Then, the concentrated solution is transferred to a dialysis bag with a molecular weight cutoff of 8–12 kDa, and dialyzed against the phosphate buffer containing glycine and trehalose for 8–16 hours, with the dialysate being replaced every 3–5 hours. The combined operation of ultrafiltration concentration and dialysis desalting reduces the sodium chloride concentration to below 5 mM. In the phosphate buffer containing glycine and trehalose, the glycine concentration is 3–8 mM, the trehalose concentration is 0.3%–0.8% (w / v), the phosphate concentration is 8–15 mM, and the pH is 6.8–7.

4.

10. The method for isolating and purifying recombinant type III collagen as described in claim 1, characterized in that: In the recombinant type III collagen obtained in step S5, the proportion of the trimer peak area to the total protein peak area is not less than 90% as detected by size exclusion high performance liquid chromatography under non-denaturing conditions. The recombinant yeast strain was detected by SDS-PAGE electrophoresis and showed a single main band at 55 kDa with a purity of not less than 95%. Western blotting with anti-human type III collagen antibody confirmed that a specific positive band was observed at 55 kDa. The recombinant yeast strain was a Pichia pastoris engineered strain that expressed recombinant type III collagen intracellularly.

Images