METHOD FOR PREPARING GLY-TB4

Abstract

The present invention relates to a method for preparing glycine-thymosin β4 (Gly-Tβ4). The present invention has an excellent effect in that, according to the present invention, a large amount of high-purity Gly-Tβ4 can be obtained with high productivity through chromatography, enzyme treatment, and filtration of a sample containing GST fusion thymosin β4 (GST-Tβ4).

Claims

1. A method of preparing glycine-thymosin β4 (Gly-thymosin β4, Gly-Tβ4), comprising: (S1) loading a sample in a primary anion exchange chromatography column equilibrated with an equilibration buffer, washing the column with a wash buffer, and eluting a fraction attached to the column with the elution buffer; (S2) performing affinity chromatography on an eluate; (S3) enzymatically cleaving the eluate by thrombin treatment; (S4) loading the cleaved product in a secondary anion exchange chromatography column equilibrated with an equilibration buffer, washing the column with a washing buffer, and eluting a fraction attached to the column with an elution buffer at a dynamic binding capacity (DBC) of 3 mg/mL or less; (S5) loading the eluate in a cation exchange chromatography column equilibrated with an equilibration buffer, washing the column with a wash buffer, and eluting a fraction attached to the column with an elution buffer at a DBC of 3 to 8 mg/mL; and (S6) filtrating the eluate.

2. The method of claim 1, wherein, in Step (S4), the cleaved product is loaded in the column at pH 8 to 9 and a conductivity of 4 mS/cm or less.

3. The method of claim 1, wherein the fraction attached to the column in the anion exchange chromatography in Step (S4) is eluted with a buffer of pH 7 to 9, which contains 10 to 30 mM Tris-HCl and 30 to 130 mM NaCl, at a flow rate of 100 to 300 cm/hr.

4. The method of claim 1, wherein, in Step (S5), the eluted product is loaded in the column at pH 4.5±0.1 and a conductivity of 2±0.2 mS/cm or less.

5. The method of claim 1, wherein, in the cation exchange chromatography in Step (S5), the cleaved product is loaded in the column at pH 4 to 7 and a conductivity of 4 mS/cm or less.

6. The method of claim 1, wherein, in the cation exchange chromatography in Step (S5), the fraction attached to the column is eluted with a buffer of pH 4 to 5, which contains 10 to 30 mM acetate and 150 to 210 mM NaCl, at a flow rate of 100 to 300 cm/hr.

7. The method of claim 1, wherein, in step (S3), the ratio of GST-Tβ4 to thrombin is 5 to 0.5:1 (unit:mg).

8. The method of claim 1, wherein the reaction temperature in Step (S3) is 13 to 27° C.

9. The method of claim 1, wherein the purity of the final purified extract according to the method is 97% or more.

10. The method of claim 1, wherein the sample in step (S1) is obtained by the following steps: culturing host cells expressing a target protein; lysing the host cells; and washing and filtering the lysate.

11. The method of claim 10, wherein main culture of the host cells is fed-batch culture at an OD.sub.600 of 40 to 65 and an induction temperature of 25 to 35° C. in the initiation of induction.

12. The method of claim 10, wherein cell productivity according to the method is 180 g pellet/L or more, and the productivity of the target protein is 54 mg/g pellet or more.

Description

DESCRIPTION OF DRAWINGS

[0106] FIG. 1 shows a 5 L verification run feeding strategy.

[0107] FIG. 2 shows a 50 L verification run feeding strategy.

[0108] FIG. 3 shows cell growth and glucose & acetate concentrations (verification run 5 L culture).

[0109] FIG. 4 shows a result of SDS-PAGE assay (5 L verification run).

[0110] FIG. 5 shows cell growth and glucose & acetate concentrations (verification run 50 L culture).

[0111] FIG. 6 shows a result of SDS-PAGE (verification run 50 L culture).

[0112] FIG. 7 shows flask culture cell growth (flask 1: w/YE, flask 2: w/o YE).

[0113] FIG. 8 shows cell growth and glucose & acetate concentrations (5 L fed-batch growth test).

[0114] FIG. 9 (FIGS. 9A to 9M) shows cell growth (DoE 13 runs) and glucose & acetate residual amounts in a culture broth.

[0115] FIG. 10 shows SDS-PAGE & protein quantification (Run data: DoE 12, 7, 8, 5, 10, 2, 13).

[0116] FIG. 11 shows SDS-PAGE & protein quantification (Run data: DoE 11, 3, 9, 1, 6, 4).

[0117] FIG. 12 shows a result of the goodness of fit for DoE variance analysis (residual plot).

[0118] FIG. 13 shows the optimization of Minitab response conditions.

[0119] FIG. 14 shows the contour plot for responses (GST-Tβ4 and cell productivity).

[0120] FIG. 15 shows cell growth and glucose & acetate concentrations (confirming DoE optimal condition reproducibility).

[0121] FIG. 16 shows SDS-PAGE & protein quantification analysis 1 (optimal condition reproducibility batch).

[0122] FIG. 17 shows SDS-PAGE & protein quantification analysis 2 (optimal condition reproducibility batch).

[0123] FIG. 18 shows cell growth and glucose & acetate concentrations (50 L optimal condition scale-up process).

[0124] FIG. 19 shows a 50 L verification run downstream process flow chart.

[0125] FIG. 20 shows the SDS-PAGE analysis for a cell lysate by cell lysis buffer conditions.

[0126] FIG. 21 shows the SDS-PAGE analysis for an ammonium sulfate study in a cell lysate.

[0127] FIG. 22 shows a regression curve for GST-Tβ4 using an affinity column.

[0128] FIG. 23 shows an AEXI process chromatogram for a linear flow rate (61 cm/hr).

[0129] FIG. 24 shows an AEXI process chromatogram for a linear flow rate (200 cm/hr).

[0130] FIG. 25 (FIGS. 25A to 25B) show the SDS-PAGE analysis for a linear flow rate study (a: 61 cm/hr, b: 200 cm/hr).

[0131] FIG. 26 shows an AEXI process chromatogram for a 500 mM NaCl elution condition.

[0132] FIG. 27 shows an AEXI process chromatogram for a 200 mM NaCl elution condition.

[0133] FIG. 28 (FIGS. 28A and 28B) shows the SDS-PAGE analysis for an elution buffer study (a: 500 mM NaCl, b: 200 mM NaCl).

[0134] FIG. 29 shows an AEXI process chromatogram for a loading sample condition (Control).

[0135] FIG. 30 shows an AEXI process chromatogram for a loading sample condition (50 mM NaCl).

[0136] FIG. 31 shows an AEXI process chromatogram for a loading sample condition (80 mM NaCl).

[0137] FIG. 32 (FIGS. 32A and 32B) shows the SDS-PAGE analysis for a loading sample & an EQ buffer study (a: 80 mM NaCl, b: 50 mM NaCl).

[0138] FIG. 33 (FIGS. 33A, 33B and 33C) shows the SDS-PAGE analysis for an AEX DBC test (a: 10 mg GST-Tβ4/mL of resin, b: 8 mg GST-Tβ4/mL of resin, and c: 75 mg GST-Tβ4/mL of resin).

[0139] FIG. 34 (FIG. 34A and FIG. 34B) shows a hold time study with an AEXI eluted pool in two different conditions (a: storage temperature: 2 to 8° C., b: storage temperature: 18 to 22° C.).

[0140] FIG. 35 shows an affinity chromatogram for a linear flow rate (90 cm/hr).

[0141] FIG. 36 shows an affinity chromatogram for a linear flow rate (180 cm/hr).

[0142] FIG. 37 (FIGS. 37A and 37B) show AEXI chromatograms for a scale up test from XK16(a) to XK50(b), respectively.

[0143] FIG. 38 (FIGS. 38A and 38B) shows affinity chromatograms for a feasibility test in LRC10 scale (a) and XK50 scale (b), respectively.

[0144] FIG. 39 (FIGS. 39A and 39B) shows a hold time study with an affinity eluted pool in two different conditions (A: storage temperature: 2 to 8° C., B: storage temperature: 18 to 22° C.).

[0145] FIG. 40 shows an SEC-HPLC chromatogram of reaction mixtures after digestion at three different temperatures.

[0146] FIG. 41 shows a CEX process chromatogram in a pH 4.5 condition.

[0147] FIG. 42 shows an RPC-HPLC chromatogram for a CEX process in a pH 4.5 condition.

[0148] FIG. 43 (FIGS. 43A and 43B) shows a set of HIC chromatograms for butyl (a) and Phenyl (b).

[0149] FIG. 44 shows a Capto MMC process chromatogram.

[0150] FIG. 45 shows the RP-HPLC analysis for MMC process eluted fractions.

[0151] FIG. 46 shows an MMC chromatogram by step elution after washing.

[0152] FIG. 47 shows an MMC chromatogram for a 100 mM NaCl washing step.

[0153] FIG. 48 shows an MMC chromatogram for a 120 mM NaCl washing step.

[0154] FIG. 49 shows an MMC chromatogram for a 140 mM NaCl washing step.

[0155] FIG. 50 shows a DSP flow chart.

[0156] FIG. 51 shows the SEC-HPLC analysis for an AEXII DBC elution sample.

[0157] FIG. 52 shows an AEXII process chromatogram for a load amount (1 mg Gly-Tβ4 per resin mL).

[0158] FIG. 53 shows an AEXII process chromatogram for a load amount (15 mg Gly-Tβ4 per resin mL).

[0159] FIG. 54 shows the RP-HPLC analysis for an AEXII process elution sample.

[0160] FIG. 55 shows an AEXII process chromatogram with increased conductivity of a loading sample.

[0161] FIG. 56 shows an RP-HPLC chromatogram for an AEXII hold time study.

[0162] FIG. 57 shows a CEX process chromatogram (at Huons).

[0163] FIG. 58 shows a CEX process chromatogram (at BINEX).

[0164] FIG. 59 shows a CEX process chromatogram for a prepack column (height: 10 cm).

[0165] FIG. 60 shows a CEX process chromatogram for a XK16 column (height: 20 cm).

[0166] FIG. 61 shows a CEX process chromatogram with a column having a bed height of 10 cm.

[0167] FIG. 62 shows a RP-HPLC chromatogram for a CEX hold time study.

[0168] FIG. 63 shows an AEXI process chromatogram for a 5 L scale process.

[0169] FIG. 64 shows an affinity process chromatogram for a 5 L scale process.

[0170] FIG. 65 shows an AEXII process chromatogram for a 5 L scale process.

[0171] FIG. 66 shows a CEX process chromatogram for a 5 L scale process.

[0172] FIG. 67 (FIGS. 67A, 67B and 67C) shows AEXI process chromatograms (a: Cycle 1, b: Cycle 2, c: Cycle 3).

[0173] FIG. 68 (FIGS. 68A, 68B and 68C) shows affinity process chromatograms (a: Cycle 1, b: Cycle 2, c: Cycle 3).

[0174] FIG. 69 (FIGS. 69A and 69B) shows AEXII process chromatograms (a: Cycle 1, b: Cycle 2).

[0175] FIG. 70 (FIGS. 70A and 70B) show CEX process Chromatograms (a: Cycle 1, b: Cycle 2).

MODES OF THE INVENTION

[0176] Hereinafter, to help in understanding the present invention, exemplary examples will be suggested. However, the following examples are merely provided to more easily understand the present invention, and not to limit the present invention.

[0177] [1. Fermentation Process]

Example (1) Strain Information

[0178] A GST-fusion human thymosinβ4 (GST-Tβ4)-producing recombinant strain was prepared using an Escherichia coli BL21 strain, and an RCB was prepared to be used in the present invention (Table 1).

TABLE-US-00001 TABLE 1 Items Detail Host Escherichia coli BL21 Product GST-fusion Human Thymosin β 4 Promoter Tac promoter Inducer IPTG or Lactose Expression Soluble expression in cytoplasm

Example (2) Confirmation of Reproducibility of China Northland Process

[0179] (2-1) 5 L Scale Culture

[0180] Referring to the 30 L culture batch record (hereinafter, “NR process”) of Northland Biotech, Beijing, China, conditions for a 5 L scale-down culture process were deduced (Tables 2, 3 and 4), and 1.sup.st to 4.sup.th cultures were carried out.

TABLE-US-00002 TABLE 2 Process Item 5L scale N/A Strain HU024 RCB Seed culture Inoculation 1 mL (RCB) (Baffled Flask) Medium LB media Culture volume 50 mL Flask volume 500 mL Kanamycin 50 μg/mL Shaking speed 200-230 rpm Temperature 37° C. Culture time 3~4 h Culture end OD.sub.600 = 2~6 Main culture Incubation volume 50 mL (5L fermenter) Culture start volume 3 L Incubator 5 L fermenter Process type Fed-batch Kanamycin Not added Culture temperature 37° C. Incubation Start OD.sub.600 80 ± 10 condition Temperature 32° C. Inducer IPTG 1 mM Added 3 mL (1M stock) amount Time 4 h Agitation speed 300~1,000 rpm DO control 30% pH 7.0 ± 0.2 Aeration 1 vvm (3 L/min) Culture time ≥17 h Culture conditions for 5 L verification run

TABLE-US-00003 TABLE 3 Preparation Preparation Item Composition criterion (L.sup.−1) Seed culture LB powder 25 g Initial medium K.sub.2HPO.sub.4 13.3 g (3 L) (NH.sub.4).sub.2HPO.sub.4 4.0 g Citric acid 1.7 g Soy peptone 1.0 g Yeast extract (Biospringer 0202) 0.5 g Yeast extract (Biospringer 0251) 0.5 g Sodium hydroxide 2.9 g Antifoam 204 0.2 mL MgSO.sub.4•7H.sub.2O 1.1 g Glucose 5.0 g Microelement solution 5.0 mL Feed A Glucose 500.0 g (1 L) MgSO.sub.4•7H.sub.2O 1.8 g Feed B Tryptone 33.3 g (0.5 L) Yeast extract (Biospringer 0202) 50.0 g Yeast extract (Biospringer 0251) 50.0 g Medium composition for 5 L verification run

TABLE-US-00004 TABLE 4 Preparation Composition criterion (L.sup.−1) Ferric chloride 16.0 g Cobalt dichloride 2.0 g Copper chloride 1.0 g Zinc chloride 2.0 g Sodium molybdate 2.0 g Boric acid 0.5 g Concentrated hydrochloric acid 100.0 mL Calcium chloride 15.0 g Vitamin B1 1.0 g Composition of microelement solution (verification run)

[0181] (2-2) Production of 50 L Scale Sample

[0182] Based on the results of 5 L-scale culture and the results of partial condition changes, 50 L scale-up conditions for sample production were deduced (Tables 5 and 6). In addition, after establishing a 50 L feeding strategy, a process for 2-batch non-clinical sample production was carried out (FIG. 1).

TABLE-US-00005 TABLE 5 Process Item Condition Seed culture 1 Inoculation volume RCB 1 mL (500 mL Flask) Medium LB medium Culture volume 100 mL × 1 ea Flask volume 500 mL Kanamycin stock (50 mg/mL) 100 μL Shaking speed 200 rpm Temperature 37° C. Culture time 3~4 h Culture end 2~5 OD.sub.600 Seed culture 2 Inoculation volume 15 ML (1 L flask) Medium LB medium Culture volume 300 mL × 3 ea Flask volume 1 L Kanamycin stock (50 mg/mL) 300 μL × 3 ea Shaking speed 200 rpm Temperature 37° C. Culture time 2.5 h Culture end 2~5 OD.sub.600 Main culture Incubation 0.9 L (75 L fermenter) Culture start volume 30 L Incubator 75 L fermenter Culture type Fed-batch Culture temperature 37 ± 0.5° C. pH 7.0 ± 0.1 Incubation Start OD.sub.600 80 ± 10 Temperature 32° C. Inducer IPTG 1 mM Added 35 mL (1M stock) amount Time 4 h Agitation speed 300~700 rpm Aeration 30 L./min DO control 30% Culture time ≥17 h Culture conditions for 50 L verification run

TABLE-US-00006 TABLE 6 Preparation criterion Preparation Item Volume Composition (L.sup.−1) Batch medium   30 L KH.sub.2PO.sub.4 13.3 g (NH.sub.4).sub.2HPO.sub.4 4.0 g Citric acid 1.7 g Soytone 1.0 g Y. extract (springer-0202) 0.5 g Y. extract (springer-0251) 0.5 g Sodium hyroxide 2.9 g Antifoam204 (Sigma) 0.3 mL MgSO.sub.4•7H.sub.2O 1.1 g Glucose 5.0 g Microelement solution 5.0 mL Feed I   8 L Glucose 500.0 g (Carbon source) MgSO.sub.4•7H.sub.2O 1.8 g Feed II   5 L Soytone 33.3 g (Nitrogen source) Y. extract (Biospringer-0202) 50.0 g Y. extract (Biospringer-0251) 50.0 g IPTG stock 0.04 L IPTG (238.3 g/mol) 237.5 g (1M, 1000 x ) Medium composition for 50 L verification run

Example (3) Development of Fed-Batch Fermentation Process for Improving GST-Tβ4 Productivity

[0183] It was confirmed that the above-described NR process has several problems such as low productivity, the formation of a medium precipitate and acetate accumulation. Accordingly, it was intended to develop a basic medium composition and process conditions that are able to solve the above-described problems.

[0184] (3-1) Flask Culture (Batch-Culture)

[0185] Flask culture experiments (flask 1, 2) were performed according to the presence or absence of a yeast extract (hereinafter, “YE”) in a medium composition of a BNS-platform (Tables 7 and 8). A strain-specific growth parameter for the fed-batch process development was identified, and seed culture conditions for performing the main culture in a fermenter were confirmed.

TABLE-US-00007 TABLE 7 Preparation Item Composition Criterion Flask 1 Flask 2 Tri-sodium citrate dihydrate 2.0 g (w/ YE) (w/o YE) KH.sub.2PO.sub.4 4.0 g Na.sub.2HPO.sub.4 10.0 g (NH.sub.4).sub.2SO.sub.4 4.0 g Glucose 5.0 g MgSO.sub.4•7H.sub.2O 0.4 g Microelement solution 6.0 mL Yeast extract 1.0 g

[0186] Flask Test Medium Composition

TABLE-US-00008 TABLE 8 Item Condition Inoculation volume RCB 1 mL Medium Main culture initial medium Culture volume 100 mL Flask volume 1 L (Baffled flask) Kanamycin 50 μg/mL Shaking speed 200 rpm Temperature 37° C. Culture end Stationary phase Flask test culture conditions

[0187] (3-2) 5 L-Scale Fed-Batch Culture Conditions

[0188] (3-2-1) Feeding Strategy

[0189] Fed-batch culture parameters (Table 9) were set to calculate the optimal feeding rate of a Feed-A medium. Feed-B was applied by calculating a rate of adding all preparation quantities by a constant feeding method during an induction section (the initiation of induction to the end of culture) according to a DoE experiment condition (Table 10).

TABLE-US-00009 TABLE 9 Variables Definition V.sub.0 Initial fermentation volume (L) X.sub.0 Initial strain concentration (OD) Y.sub.x/s Yield of strain based on substrate (OD/g/L) dt Feeding rate adjustment interval (h) Si Substrate concentration in feed medium (g/L) S.sub.0 Initial substrate concentration (g/L) μ Cell specific growth rate (h.sup.−1) Gravity Specific gravity of feed medium (g/mL) F.sub.0 Flow rate of feed medium (L/h) Exponential feeding parameters

TABLE-US-00010 TABLE 10 Feeding strategy Feed-A Feed-B Step (C-source) (N-source) Batch growth N/A N/A Fed-batch growth Exponential Induction feeding Constant feeding DoE fed-batch culture feeding strategy

[0190] (3-2-2) Medium Preparation and Process Conditions

[0191] A DoE fed-batch culture process starting from a 2.5 L initial medium and terminated at a 4.2 L (0.7 L of feed-A, 1.0 L of feed-B) scale was established (Tables 11, 12 and 13), and the same conditions were applied to the DoE culture experiment.

TABLE-US-00011 TABLE 11 Preparation criterion Preparation item Volume Composition (L.sup.−1) Initial medium 2.5 L Tri-sodium citrate dihydrate 2.0 g KH.sub.2PO.sub.4 4.0 g Na.sub.2HPO.sub.4 10.0 g (NH.sub.4).sub.2SO.sub.4 4.0 g Y. extract (Biospringer-0202) 1.0 g Antifoam204 (sigma) 100.0 μL Glucose 5.0 g MgSO.sub.4•7H.sub.2O 0.4 g Microelement solution 6.0 mL Feed-A 0.7 L Glucose 500.0 g (Carbon source) MgSO.sub.4•7H.sub.2O 10.0 g Feed-B   1 L Y. extract (Biospringer-0202) 150.0 g (Nitrogen source) Antifoam (sigma 204) 200.0 μL Composition of DoE fed-batch culture medium (BNS-Platform)

TABLE-US-00012 TABLE 12 Preparation Composition criterion (L.sup.−1) Ferrous sulfate heptahydrate 10.0 g Zinc sulfate heptahydrate 2.25 g Cupric sulfate pentahydrate 1.00 g Manganese sulfate monohydrate 0.35 g Sodium borate decahydrate 0.23 g Calcium chloride dehydrate 2.00 g Ammonium heptamolybdate tetrahydrate 0.10 g HCl 37% 44.40 mL Composition of microelement solution (BNS-Platform)

TABLE-US-00013 TABLE 13 Process Item Condition Seed culture Inoculation 1 mL (1 L baffled flask) Culture volume 125 mL Flask volume 1 L Kanamycin 50 μg/mL Shaking speed 200 rpm Temperature 37° C. Culture time ≥4 h Culture end ≥2 OD.sub.600 Main culture Inoculation volume 125 mL (5 L fermenter) Culture volume 2.5 L .fwdarw. 4.2 L Incubator Sartorius B-plus Culture type Fed-batch Culture temperature    37 ± 0.5° C. pH 6.8 ± 0.1 Induction Start condition DoE condition Temperature DoE condition Inducer IPTG 1 mM Added amount 3 mL (1M stock) Time Until the end of culture Agitation speed 300~1000 rpm Aeration 3 L/min Conditions for DoE fed-batch culture process

[0192] (3-3) Deduction of Optimal Fed-Batch Culture Conditions

[0193] To optimize an induction strategy, a response surface method (hereinafter, “RSM”) test was designed. A 2-level full central composite experiment for two factors (induction starting point and induction temperature) was designed using Minitab software to carry out 13 runs of fed-batch culture experiments (Tables 14 to 16).

TABLE-US-00014 TABLE 14 Experimental Center point Factor condition condition Induction cell High OD.sub.600 = 80 ± 10 OD.sub.600 = 55 ± 10 density Low OD.sub.600 = 30 ± 10 Induction High 34° C. 31° C. temperature Low 28° C. DoE factors and condition ranges

TABLE-US-00015 TABLE 15 Item Condition Details DoE design Central present the minimum experimental points to find the method synthesis optimal response surface point for each factor method forming experimental conditions with “vertex + axis point + central point” or the like. Factor 2 Refer to Table 14 Level 2 Response 2 GST-Tβ34 productivity (mg/g), cell productivity (g/L) Resolution complete As the resolution of the DoE design is higher, it is recognized to be more statistically reliable. “complete” is the highest level of resolution, confirming the interaction between a factor and a main effect, and deducing the optimal condition Repeated 1 Number of repeated experiments: 1 cycle experiment Center point 5 Experiment conditions for the intermediate value of each factor OD.sub.600 = 55, temperature: 31° C. Designed so that the central point is close to the expected point in the optimal condition. When k (factor) = 2, statistically reliable no. of central points is 5 or more. Axis point 4 Experimental point at the α value position from the center. Composed of 2k experimental points. Vertex 4 Complete factor batch experiment condition: composed of 2k experimental points α value 1.41 Distance from the center to the axis point, different according to the number of factors (k) Run 13 No. of combinations of experiment conditions Conditions for DoE experiment design

TABLE-US-00016 TABLE 16 Factor Center Induction cell Induction temp. Batch No. DoE No. point density (OD.sub.600) (° C.) 160928-B1 12 0 55.0 31.0 161005-B1 7 −1 55.0 26.8 161005-B2 8 −1 55.0 35.2 161027-B1 5 −1 19.6 31.0 161011-B2 10 0 55.0 31.0 161101-B1 2 1 80.0 28.0 161018-B2 13 0 55.0 31.0 161018-B1 11 0 55.0 31.0 161103-B1 3 1 30.0 34.0 161011-B1 9 0 55.0 31.0 161025-B1 1 1 30.0 28.0 161103-B2 6 −1 90.4 31.0 161101-B2 4 1 80.0 34.0 Design of 2-level full central composite factor experiment

[0194] (3-4) Synthesis of Reproducibility of Optimal Culture Process

[0195] Results of a RSM DoE culture experiment were statistically analyzed to deduce optimal conditions and the target standard of response values, and culture was carried out under optimal conditions to confirm the reproducibility of the cultures (Table 17).

TABLE-US-00017 TABLE 17 Induction YE (Yeast OD & extract) Antifoam Batch No. Temp. manufacturer manufacturer Aeration 161122-B1 DoE experiment Biospringer Sigma 1 vvm optimal condition 0202 161115-B1 BD Sigma 1 vvm 161122-B2 BD Sigma 1 vvm 161124-B1 BD Dow corning 0.5 vvm (“DC”) Confirmation of reproducibility of DoE optimal conditions

Example (4) Optimal Condition Scale-Up for 50 L-Scale Process (Sample Production)

[0196] (4-1) Conditions for Culture Process

[0197] A 50 L culture process for sample production was performed by scale-up of optimal conditions for developing a 5 L-scale process (Tables 18 and 19).

TABLE-US-00018 TABLE 18 Process Item Condition Seed culture 1 Inoculation volume RCB 1 mL (500 mL Flask) Culture volume 100 mL × 1 ea Flask volume 500 mL baffled flask Shaking speed 200 rpm Temperature 37° C. Culture time 5~6 h Culture end ≥2 OD.sub.600 Seed culture 2 Inoculation volume 15 mL (1 L Flask) Culture volume 315 mL × 4 ea Flask volume 1 L baffled flask Shaking speed 200 rpm Temperature 37° C. Culture time 3~4 h Culture end ≥2 OD.sub.600 Main culture Inoculation 1.25 L (5%, v/v) (75 L fermenter) Culture start volume 25 L Incubator 75 L fermenter Culture type Fed-batch Culture temperature* 36.5 ± 0.5° C. pH* 6.8 ± 0.1  Induction Start OD.sub.600 = 54 ± 5 Temperature 30.0 ± 1.0° C. Inducer IPTG 1 mM Added amount 40 mL (1M stock) Time ≥9 h Agitation speed** 300~700 rpm Aeration 15 L/min DO control 30% Culture time ≥2 h Conditions for optimal 50 L scale-up culture *A culture temperature and pH are the same as BNS-platform conditions. **The maximum agitation speed of 75 L equipment (700 rpm) is reflected.

TABLE-US-00019 TABLE 19 Preparation Preparation item Volume Composition criterion (L.sup.−1) Batch medium Tri-sodium citrate 2.0 g dihydrate KH.sub.2PO.sub.4 4.0 g Na.sub.2HPO.sub.4 10.0 g g (NH.sub.4).sub.2SO.sub.4 4.0 g 25 L Y. extract (BD) 1.0 g Antifoam (Dow corning) 2.0 mL Glucose 5.0 g MgSO.sub.4•7H.sub.2O 0.4 g Microelement solution 6.0 mL Feed-A 7 L Glucose 500.0 g (Carbon source) MgSO.sub.4•7H.sub.2O 10.0 g Feed-B 10 L Yeast extract (BD) 150.0 g (Nitrogen source) Antifoam (Dow corning) 5.0 mL IPTG stock 0.04 L IPTG 238.3 g (1M, 1000 x) Seed medium 1.4 Tri-sodium citrate 2.0 g dihydrate KH.sub.2PO.sub.4 4.0 g Na.sub.2HPO.sub.4 10.0 g (NH.sub.4).sub.2SO.sub.4 4.0 g Y. extract (BD) 1.0 g Glucose 5.0 g MgSO.sub.4•7H.sub.2O 0.4 g Microelement solution 6.0 mL Medium composition for 50 L optimal condition scale-up culture

[0198] (4-2) Recovery Conditions

[0199] (4-2-1) Cell Slurry Recovery

[0200] (A) Recovery of Culture Broth [0201] Recovery time: overnight after cooling of culture broth (≤15.0° C.) [0202] Recovery volume: 40 to 42 L

[0203] (B) Recovery of Cell Slurry (Continuous Centrifugation)

[0204] {circle around (1)} 1.sup.st recovery [0205] Solid content before recovery: ≥20% [0206] Bowl rotation speed: 9,500 rpm (11,119×g) [0207] Inlet flow rate: 100 L/M [0208] Outlet pressure: 100 kPa [0209] Cell washing: 40 L of deionized water (DIW)

[0210] {circle around (2)} 2.sup.nd recovery [0211] Solid content before recovery: ≥10% [0212] Bowl rotation speed: 9,500 rpm [0213] Inlet flow rate: 100 L/m [0214] Outlet pressure: 100 kPa [0215] Cell washing: 40 L of DIW

[0216] {circle around (3)} 3.sup.rd recovery [0217] Solid content before recovery: ≥10% [0218] Bowl rotation speed: 9,500 rpm [0219] Inlet flow rate: 100 L/m [0220] Outlet pressure: 100 kPa

[0221] (4-2-2) Cell Lysis and Supernatant Recovery

[0222] (A) Preparation of lysis buffer (20×) [0223] 400 mM Tris buffer, pH 8.0±0.2 (suitable reagent: HCl), Conductivity 17±2 mS/cm [0224] 400 mM Tris buffer composition: Tris 21.8 g/L, Tris-HCl 34.7 g/L

[0225] (B) Cell Slurry Preparation (15±1.5%, w/w)

[0226] {circle around (1)} The volume was adjusted so that the solid content was 15±1.5% by adding 20× lysis buffer and DIW.

[0227] {circle around (2)} Calculation formula

Total weight=Cell cake weight/0.15

20× Lysis Buffer weight=Total weight/20

Amount of added DIW=Total weight−slurry weight−20× Lysis buffer weight

[0228] (C) CIP (Microfluidizer) [0229] 0.5 N NaOH: 11,000 psi condition, 300 mL once, 500 mL once [0230] DIW: 11,000 psi condition, 300 mL once, 500 mL once

[0231] (D) Cell Lysis (Microfluidizer)

[0232] {circle around (1)} EQ: 11,000 psi condition, * 300 mL 1× lysis buffer

[0233] {circle around (2)} Cell lysis: 11,000 psi condition, 15±1.5% cell slurry lysis twice in a row [0234] Cell lysis criteria: OD ratio before and after lysis ≤30% [0235] Calculation of cell lysis rate: (OD.sub.600 after lysis/OD.sub.600 before lysis)×100

[0236] {circle around (3)} Flushing: 11,000 psi condition, 1× lysis buffer 50 mL

[0237] (E) Recovery of Supernatant of Cell Lysate (Continuous Centrifugation) [0238] Bowl rotation speed: 9,500 rpm [0239] Inlet flow rate: 100 L/M [0240] Outlet pressure: 250 kPa

[0241] (4-2-3) Depth Filtration

[0242] (A) Filter information

[0243] {circle around (1)} Depth filter: [0244] Millistak+POD C0HC: 0.11 m.sup.2×2 ea [0245] Millistak+POD F0HC: 0.11 m.sup.2×2 ea

[0246] {circle around (2)} Membrane filter [0247] Opticap XL10 (0.45 μm) 0.69 m.sup.2×1 ea

[0248] (B) Filter Washing [0249] Depth filter: 100 L/m.sup.2 of DIW, flow rate 300 LMH (1,100 mL/min) [0250] Sterile filter: 10 L/m.sup.2 of DIW, flow rate 1,000 mL/min

[0251] (C) Depth Filtration

TABLE-US-00020 Step Input Volume Feed Flux Clarification Cell lysate N/A 75 LMH (275 mL/min) Flushing DIW 11.6 L 75 LMH (275 mL/min) Air Flushing N/A N/A 300 LM (1,100 mL/min)

Example (5) Culture Broth Analysis

[0252] (5-1) Measurement of Cell Density

[0253] The cell concentration in the culture broth was obtained by measuring optical density at 600 nm. 1 mL of the culture broth was diluted with distilled water to have OD.sub.600=0.2 to 0.5, and an optical density (OD.sub.600) was measured, and then multiplied by a dilution factor, thereby obtaining a cell density. Sample dilution was performed by standardization in the manner as shown in Table 20.

TABLE-US-00021 TABLE 20 Dilution Dilution OD600 rate volume 1~3  10 x 1 mL 4~7  20 x 1 mL  8~10  30 x 6 mL 10~20  50 x 10 mL 20~40 100 x 10 mL 40~60 200 x 20 mL  60~100 300 x 30 mL 100≤ 500 x 50 mL Dilution conditions for OD value measurement

[0254] (5-2) Cell Metabolite Measurement

[0255] Glucose and acetate concentrations in the culture broth were measured from a supernatant obtained by centrifuging 1 mL of the culture broth at 13,000 rpm for 5 minutes using a Cedex Bio (Roche, Switzerland) analyzer (SOP No. FE02-921).

[0256] (5-3) Measurement of Cell Productivity

[0257] (5-3-1) Measurement of Wet Cell Weight (g/L)

[0258] A wet cell weight was measured by the following method.

[0259] {circle around (1)} Measure 500 mL tube+SUPPORT CUSHION weight (prepare two)

[0260] {circle around (2)} Add aliquots of culture broth by 500 mL (use 500 mL measuring cylinder)

[0261] {circle around (3)} Centrifuge under conditions of 4500 rpm and 4° C. for 1 hour (CE-6140 equipment)

[0262] {circle around (4)} Discard supernatant and measure weight of pellet containing container (tube+cushion)

[0263] {circle around (5)} Calculate wet cell weight (g/L)

Wet cell weight=(pellet contained weight−container weight)g/0.5 L (volume of culture broth)

[0264] {circle around (6)} Determine cell productivity by the average of two measured wet cell weights.

[0265] (5-3-2) Measurement of Solid Content (%)

[0266] {circle around (1)} Prepare two micro tubes

[0267] {circle around (2)} Measure and record tube weights using precision balance

[0268] {circle around (3)} Measure and record weight of tube containing 1 mL culture broth

[0269] {circle around (4)} Centrifuge under conditions of 13,000 rpm and 4° C. for 5 minutes

[0270] {circle around (5)} Discard supernatant and measure weight of tube containing pellet

[0271] {circle around (6)} Calculate solid content (%):

[00001]$\frac{\mathrm{pellet}\mathrm{contained}\mathrm{weight}-\mathrm{tube}\mathrm{weight}}{\mathrm{weight}\mathrm{including}1\mathrm{mL}\mathrm{medium}-\mathrm{tube}\mathrm{weight}}\times 100$

[0272] {circle around (7)} Take mean value of two measured solid contents

[0273] (5-4) Measurement of GST-Tβ4 Productivity (SDS-PAGE & Protein Quantification)

[0274] {circle around (1)} Add 1,000 μL of lysis buffer (50 mM Tris, 100 mM NaCl, 0.1% 2ME, 5 mM EDTA, pH 7.0) to cell pellet sample of OD.sub.600=5 and perform vortexing

[0275] {circle around (2)} Perform sonication for 10 seconds and leave on ice for 1 minute (repeated 7 times)

[0276] {circle around (3)} After centrifugation at 10,000×g for 20 minutes, take 15 μL of supernatant

[0277] {circle around (4)} Prepare 15 μL of bovine serum albumin (BSA) protein standard reagent to be contained at 0.5, 1, 2 or 4 μg

[0278] {circle around (5)} Mix 15 μL sample (centrifugation supernatant and BSA standard) with 5 μL of the sample buffer mixture prepared in {circle around (4)} (total volume: 20 μL)

[0279] {circle around (6)} React at 70° C. for 10 minutes

[0280] {circle around (7)} Load entire sample (20 μL) to each well of gel

[0281] {circle around (8)} Run under conditions of 200 V and 400 mA for 35 minutes

[0282] {circle around (9)} Perform Coomassie blue staining for 20 minutes or more

[0283] {circle around (10)} Destain for 2 hours or more

[0284] {circle around (11)} Analyze image of gel band density using densitometer

Experimental Example (1) Confirmation of Reproducibility of Chinese Northland Process

[0285] (1-1) 5 L Scale Culture

[0286] L scale verification run was carried out by scaling down an NR process, and the normal expression of GST-Tβ4 was confirmed through SDS-PAGE (FIG. 4). The final cell density was OD.sub.600=85, 81, 61, 41, 86 (Table 21), and except two tertiary batches (Batch No. 160428-B1, B2), growth similar to the NR process was shown (FIG. 3). However, in the entire process, glucose in a culture broth was accumulated to control growth by interrupting and resuming feed-A supply, and while proceeding to the 4.sup.th process, continuous improvement in C-source (feed-A medium) supply was attempted. However, except the 4.sup.th process, acetate accumulation was observed. Accordingly, it was necessary to more precisely improve a feeding rate of a feed-A medium. In addition, after the medium was prepared, a large amount of precipitates were formed, and the pH of the medium composition was not optimized, and therefore, the disadvantage of a conventional NR process, such as the requirement of a separate titration process before inoculation, was confirmed.

TABLE-US-00022 TABLE 21 Induction Induction temper- Cell Batch Batch start ature Final productivity No. name (OD.sub.600) (° C.) OD.sub.600 (g/L) 160407-B1 5 L 1.sup.st 77 32 85 104 160407-B2 5 L 2.sup.nd 75 32 81 106 160428-B1 5 L 3.sup.rd-1 60 32 61 86 160428-B2 5 L 3.sup.rd-2 40 32 41 58 160512-B2 5 L 4.sup.th 75 32 86 125 Results for 5 L verification run culture

[0287] (1-2) 50 L Scale Culture (Sample Production)

[0288] Based on results for 5 L scale verification culture performed by scale down of an NR process, 50 L scale process conditions were deduced (Tables 5 and 6, FIG. 2), 2 batches of 50 L sample production were conducted. Compared to the NR process (results for 30 L scale culture), the lag phase of the initial cell growth was long in both batches (1.sup.st and 2.sup.nd), but the final cell density reached OD.sub.600=84 or, 98, which was similar to that of the NR process (FIG. 5, Table 22). An acetate concentration was 1 g/L or less in both batches, which was well controlled until the end of the culture, but in the case of the 2.sup.nd batch (Batch No. 160729), glucose was accumulated up to 2 g/L in the second half (after 17 hours) of the culture. GST-Tβ4 expression was confirmed through SDS-PAGE (FIG. 6), and after continuous centrifugation of the culture broth, finally, 3.0 kg or 3.3 kg of cells were recovered (Table 22).

TABLE-US-00023 TABLE 22 Cell Recovered Batch Batch Final Volume of Productivity Recovery cell weight No. name OD.sub.600 culture broth (g/L) type (kg) 160629 50 L 1.sup.st 84.3 36.7 128 Cell Slurry 3.3 160729 50 L 2.sup.nd 98.3 36.9 122 Cell cake 3.0 Culture results of 50 L verification run

Experimental Example (2) Development of Fed-Batch Fermentation Process for Improvement in GST-Tβ4 Productivity

[0289] (2-1) Flask Culture

[0290] To confirm cell growth parameters and seed culture process time, required for calculating a feeding rate of a DoE fed-batch culture process, a flask culture experiment was performed. Flask 1 was carried out with the same medium composition as the initial medium of a fed-batch, and flask 2 was carried out with a composition excluding only YE from the initial medium. Flask 1 (containing YE) has a shorter lag phase than flask 2, and the maximum cell density (OD.sub.600) was grown until 8.3 (flask 1) or 6.9 (flask 2) (FIG. 7). Based on the OD value and glucose concentration of an exponential phase, a cell yield (Y.sub.X/S) with respect to a glucose substrate and the maximum specific growth rate (μm) were calculated (Table 23). In the 5 L fed-batch culture experiment using a fermenter, a YE-containing medium composition was selected, and therefore, a feeding rate was calculated using the parameter value of flask 1. For a 5 L fed-batch seed culture process, OD.sub.600 was ≥2, and culture time was determined as 4 hours or more.

TABLE-US-00024 TABLE 23 Parameter Flask 1 Flask 2 Exponential Time (h) 4 7 5.6 8.6 phase Glucose 4.8 0 4.2 0 (g/L) OD.sub.600 2.1 7.5 0.7 5.4 Y.sub.X/S 1.13 1.11 μ.sub.m (h.sup.−1) 0.43 0.66 X.sub.m (OD.sub.600) 8.3 6.9 HU024 cell growth parameters

[0291] (2-2) Feeding Strategy

[0292] (2-2-1) Exponential Fed-Batch Parameter

[0293] Referring to results of the flask 1 culture test that has been previously performed, each parameter for calculating the feeding rate of the C-source was set as shown in Table 24. In the flask batch culture of the YE-containing initial medium composition, the maximum specific growth rate (μm) was 0.43 h.sup.−1 (Table 23), and since the rapid growth of cells in E. coli culture is disadvantageous for the production of a soluble protein-type product, a quasi-steady state was induced through precise control of a feeding rate to control a specific growth rate to be 0.3 to 0.14 h.sup.−1.

TABLE-US-00025 TABLE 24 Parameter Set value V.sub.0 2.5 X.sub.0 8 Y.sub.X/S 1.13 dt 1 Si 500 μ 0.3~0.14 gravity 1.1 F.sub.0 0.012 Exponential fed-batch parameters

[0294] (2-2-2) Calculation of Feeding Rate (Feed-A)

[0295] A feeding rate of the feed-A medium was calculated using the following formulas. The first feeding rate at the time point of ending batch-culture was calculated by substituting each parameter value (Table 24) into Equation {circle around (4)}, and subsequently, a VX value (Equation {circle around (1)}), a V value (Equation {circle around (2)}), a X value (Equation {circle around (3)}) and a F value (Equation {circle around (4)}) were sequentially calculated every hour (dt). Since the calculated flow rate (F) is expressed as L/h, after conversion to mL/min, finally, the flow rate was converted to the units of g/min, which is the unit of speed of pumping equipment (Table 25).

[00002]$\begin{array}{cc}\mathrm{VX}=V_0.Math.X_0+F_0*\mathrm{dt}*\mathrm{Si}*Y_{X/S}& 1\\ V=V_0+\mathrm{dt}*F_0*\mathrm{gravity}& 2\\ X=\mathrm{VX}/V& 3\\ F=\frac{X_0.Math.V_0\left(e^{\mu t}-1\right)}{S_0.Math.Y_{X/S}.Math.t}& 4\end{array}$

TABLE-US-00026 TABLE 25 F F t X V VX F (L/hr) (mL/min) (g/min) μ 0~4 Batch culture 4 8 2.50 20.0 0.012 0.21 0.2 0.3 5 11 2.51 27.0 0.017 0.28 0.3 0.3 6 14 2.53 36.4 0.023 0.38 0.4 0.3 7 19 2.55 49.2 0.019 0.32 0.4 0.2 8 23 2.57 60.1 0.024 0.39 0.4 0.2 9 28 2.59 73.4 0.026 0.43 0.5 0.18 10 34 2.62 87.9 0.031 0.51 0.6 0.18 11 40 2.65 105.2 0.037 0.61 0.7 0.18 12 47 2.69 125.9 0.044 0.73 0.8 0.18 13 55 2.73 150.8 0.046 0.77 0.8 0.16 14 64 2.78 176.9 0.054 0.91 1.0 0.16 15 73 2.83 207.6 0.064 1.06 1.2 0.16 16 84 2.90 243.6 0.065 1.08 1.2 0.14 17 95 2.96 280.3 0.075 1.24 1.4 0.14 18 106 3.04 322.4 0.086 1.43 1.6 0.14 19 119 3.13 370.8 0.099 1.64 1.8 0.14 20 132 3.23 426.6 0.113 1.89 2.1 0.14 Calculation of exponential feeding rate

[0296] (2-3) Confirmation of Cell Growth in Fed-Batch Process

[0297] To confirm basic cell growth in a fed-batch process using a 5 L fermenter, fed-batch culture was conducted by supplying feed-A at the calculated feed-A feeding rate (Table 23) (FIG. 8). A total of 788.8 mL of the feed-A medium was supplied until the end of the culture, and the maximum cell density (OD.sub.600) predicted when the substrate yield (Y.sub.X/S) was applied by converting a total glucose amount of 397 g (containing initial glucose) added into an incubator into the final culture volume (3.3 L) was 136, almost similar to the maximum cell density (OD.sub.600=142) as the actual culture result. A acetate concentration was well controlled to less than 1 g/L from the beginning to the end of the culture, and as the initial glucose in the initial media was rapidly consumed after 3 hours of culture, the concentration was well controlled close to 0 g/L until the end of culture even after the beginning of glucose feeding (4 h), and since the quasi-steady state of the actual cell growth was well induced close to a set μ value, it was considered that the previously calculated feeding rate (Table 25) was appropriate. In addition, the cell growth result for the fed-batch culture experiment was utilized as a material for predicting the starting point of a DoE culture experiment.

[0298] (2-4) Deduction of Optimal Culture Conditions (DoE)

[0299] (2-4-1) Culture Result

[0300] Thirteen runs of culture were performed according to DoE RSM design conditions. At the starting point of induction under each DoE condition, IPTG addition and a temperature shift as well as feed-B medium feeding were initiated, and a feed-B medium consisting of high concentration YE was only provided in the induction phase, so that culture was conducted with a strategy of maximizing cell growth and GST-Tβ4 productivity. Except the DoE 5.sup.th batch (Batch No. 161027-B1), the maximum cell density (OD.sub.600) grew to 140 to 196, and was overall well controlled without excessive glucose and acetate accumulation (FIGS. 9A to 9M). In the DoE 5.sup.th batch (161027-B1), according to the RSM DoE experiment design, as the fastest induction starting condition (OD.sub.600=19.6), glucose and acetate were accumulated the most in the culture broth, considered to be due to induction and temperature shift before sufficient cell growth. The mean cell productivity of the entire DoE batch runs was 185±31 g/L (wet cell weight), and the total mean GST-Tβ4 productivity confirmed through SDS-PAGE and gel band image analysis (FIGS. 10 and 11) was 56.4±10.0 mg/g (productivity per wet cell) and 10.4±2.2 g/L (productivity per culture volume), and cell productivity with respect to the mean value of 2 batches (primary, secondary) of a 50 L verification run was increased approximately 1.5-fold, and GST-Tβ4 productivity was increased approximately 2.5-fold (Table 26).

TABLE-US-00027 TABLE 26 Initiation Temperature Cell GST-Tβ of Induction of induction Final productivity productivity Batch No. DoE No. (OD600) (° C.) OD600 (g/L) (mg/g) 160928-   1 DoE 12 52 31 196.1 202 60.56 161005-   1 DoE 7 56 26.8 189.5 203 61.36 161005-   2 DoE 8 56 35.2 148.6 140 59.68 161027-   1 DoE 5 17 31 128.3 108 66.51 161011-   2 DoE 10 54 31 183.9 183 67.37 161101-   1 DoE 2 78 28 161.3 221 33.86 161018-   2 DoE 13 55 31 156.4 183 61.92 161018-   1 DoE 11 53 31 164.4 182 52.11 161103-   1 DoE 3 29 34 157.0 198 63.51 161011-   1 DoE 9 54 31 185.3 194 56.95 161025-   1 DoE 1 28 28 171.0 172 60.43 161103-   2 DoE  93 31 158.0 212 40.78 161101-   2 DoE 4 77 34 173.8 202 48.17 160629 32 25.2 160729 70 32 122 Summary of DoE 13 runs culture process results indicates data missing or illegible when filed

[0301] (2-4-2) DoE Statistical Analysis

[0302] (2-4-2-1) Analysis of Variance (ANOVA)

[0303] ANOVA was conducted on cell productivity and GST-Tβ4 productivity values of all DoE-based batches (13 runs) using Minitab software (Table 27). R.sup.2 values representing model explanatory power were 79.78% (GST-Tβ4 productivity) and 58.57% (cell productivity), determined as common levels, considering an experiment design without repetition (considering development period) and characteristics of a culture field in which several factors were involved in a response value. Since the p-values of induction OD were 0.002 (GST-Tβ4 productivity) and 0.029 (cell productivity), they were confirmed as significant factors with respect to both of the response values (GST-Tβ4 and cell productivity) at a confidence level less than 0.05. As the p-values of an induction temperature were 0.399 (GST-Tβ4 productivity) and 0.299 (cell productivity), it was determined that there is no effect on GSTTB4 and cell productivity results. Since p-values in the “induction OD*temperature” term were 0.374 and 0.413, it was confirmed that there was no “reciprocal” effect, referred to as the inter-condition interference between two factors, on the response values.

TABLE-US-00028 TABLE 27 Term Coefficients P-value GST-Tβ4 productivity Constant 59.780 0.000 Induction OD −9.788 0.002 Temperature 1.878 0.399 Induction OD × −4.464 0.087 Induction OD Temperature × −1.028 0.661 Temperature Induction OD × 2.809 0.374 Temperature Curvature effect — 0.205 R.sup.2 — 79.78% R.sup.2-adjusted — 65.34% Cell productivity Constant 188.800 0.000 Induction OD 25.010 0.029 Temperature −10.262 0.299 Induction OD × −6.275 0.543 Induction OD Temperature × −0.525 0.959 Temperature Induction OD × −11.250 0.413 Temperature Curvature effect — 0.819 R.sup.2 — 58.57% R.sup.2-adjusted — 28.98% Regression coefficient of RSM DoE experiment and results for ANOVA

[0304] (2-4-2-2) Residual Plot

[0305] Referring to a result of a residual plot of the conducted DoE model results, generally, a constant variance was shown without specific tendency or singularity. In addition, the experiment was conducted independently and randomly according to time and an experimental order, and according to normal distribution, it was confirmed that a lack of fit was not shown (FIG. 12).

[0306] (2-4-2-3) Deduction of Optimal Conditions and Determination of Design Space

[0307] According to the ANOVA result of the culture experiment through DoE RSM design, optimal conditions were obtained using a response optimization tool of Minitab software. The optimal conditions presented at a level satisfying a synthesis satisfaction degree of 0.67 with respect to cell productivity and GST-Tβ4 productivity were an induction OD.sub.600 of 53.9 and an induction temperature of 28.9° C., and target response values were predicted to be 58.5 mg/g and 194 g/L, respectively (Table 28, FIG. 13). Since the suggested optimal conditions are close to center point conditions, the setting of factor levels for the initial DoE design was considered appropriate, and the lower limit of target productivity in the optimal process was set as the lowest point of the average of center points (Table 29).

TABLE-US-00029 TABLE 28 Induction Optimal Target (estimated) value parameter condition of response value Induction 53.9 *GST-TB4 productivity: OD.sub.600 58.5 mg/g pellet *Cell productivity: 194 g pellet/L Induction Temperature 28.9 (° C.) Optimization of Minitab response conditions

[0308] A contour plot plotted using the lowest point of the average center points (54.1 mg/g & 180 g/L) as the lower limit of the target standard, and the mean value of top five batches (64.1 mg/g & 208 g/L) as the upper limit of the target standard was confirmed (Table 29, FIG. 14). The white area of the contour plot indicates a design space satisfying both of GST-Tβ4 productivity and the target standard of cell productivity (≥54.1 mg/g, ≥180 g/L), and the final optimized process conditions deduced again based on the stable range of this space were determined to be an induction OD.sub.600 of 54±5 and an induction temperature of 30±1° C. (Table 30).

TABLE-US-00030 TABLE 29 Top 5 Mean of Response Total batch central Target value mean mean points criterion Cell 185 ± 31  208 ± 8  189 ± 9  ≥180 productivity (g/L) GST-TB4 56.4 ± 10.0 64.1 ± 2.7 59.8 ± 5.7 ≥54.1 productivity (mg/g) Target standard of response values of optimal condition process for HU024 culture

TABLE-US-00031 TABLE 30 Process factor Optimal condition Induction OD.sub.600 54 ± 5 Induction temperature 30 ± 1° C. Optimal process conditions for HU024 culture

[0309] (2-5) Confirmation of Reproducibility of Optimal Culture Conditions

[0310] Culture for confirming response value reproducibility was performed with the final optimal conditions (OD.sub.600=54±5, temperature: 30±1° C.) deduced from statistical analysis for the DoE RSM culture experiment results. A total of 4 batch culture processes were conducted (Table 17): 1 batch (Batch No. 161122-B1) using the same product as the YE product (Springer 0202) used in a DoE experiment; 2 batches (Batch No. 161115-B1, 161122-B2) using YE manufactured by BD; and 1 batch (161124-B1) with an antifoam product (Dow Corning) and a changed aeration condition (0.5 vvm). In the case of an antifoam manufactured by Dow Corning, the amount of antifoam used has increased more than 10 times compared to an existing antifoam (Sigma 204). However, all 4 batches showed similar growth, but the maximum cell density (OD.sub.600) was 160 or more, and glucose and acetate concentrations were well controlled to 1 g/L or less (FIG. 15). Cell productivity was 182, 202, 208 or, 205 g/L (Table 32), the average of the two repeated analyses of GST-Tβ4 productivity was 66.5±0.3, 57.5±2.4, 65.1±9.5 or 66.6±2.6 mg/g (Table 31, FIGS. 16 and 17), and all of the 4 batches achieved the result of the target standard of optimal condition response values (Table 29) or more. As a result, the reproducibility of the 5 L scale optimal culture process condition established in this study was confirmed, and compared to the verification 50 L culture result that reproduced the NR process (Batch No. 160629, 160729), cell productivity was increased approximately 1.6-fold, and GST-Tβ4 productivity was increased approximately 2.8-fold (Table 32).

TABLE-US-00032 TABLE 31 GST-TB4 productivity GST-TB4 productivity per (mg/g pellet) culture volume (g/L) Analysis Analysis Analysis Analysis Batch name. <1> <2> Mean <1> <2> Mean 161122-B1 66.3 66.7 66.5 ± 0.3 11.7 11.8 11.8 ± 0.0 (Springer YE*) 161122-B1 71.8 58.3 65.1 ± 9.5 12.1 9.8 11.0 ± 1.6 (using BD YE) 161122-B2 59.2 55.8 57.5 ± 2.4 11.0 10.3 10.6 ± 0.4 (using BD YE) 161124-B1 68.5 64.8 66.6 ± 2.6 12.3 11.6 11.9 ± 0.5 (BD YE/DC** antifoam/Air0.5) 160629 25.2 19.9 22.5 ± 3.7 2.7 2.2  2.4 ± 0.4 (50 L 1.sup.st, 2.sup.nd) *YE: Yeast Extract **DC: Dow Corning Analysis result for GST-Tβ4 productivity under optimal conditions

TABLE-US-00033 TABLE 32 Optimal process condition Confirmation Change in Verification deduced of optimal antifoam, run 50 L from DoE condition Aeration 1.sup.st, Item experiment reproducibility confirmation 2.sup.nd Batch No. N/A 161122-B1 161115-B1 161124-B1 160629 161122-B2 160729 YE Springer Springer BD BD Biospringer manufacturer.sup.(1) (0202) (0202) (0202) Biospringer (0251) Soy peptone Antifoam N/A Antifoam 204 Dow Antifoam manufacturer.sup.(2) (amount used: 0.5~0.6 Corning 204 (6 mL) mL) (amount used: 10 mL) Aeration N/A 1 vvm 0.5 vvm 1 vvm condition.sup.(3) Induction OD.sub.600 54 ± 5 53.8 52.4 54.4 69 51.3 70 Induction 30 ± 1 30 30 30 32 Temperature (C.) Cell ≥180 182 205 207 125 ± 4 productivity 208 (g pellet/L) GST-TB4 ≥54.1 66.5 ± 0.3 65.1 ± 9.5 66.6 ± 2.6 22.5 ± 3.7 productivity 57.5 ± 2.4 (mg/g pellet) Kanamycin ◯ X X X ◯ added in total culture broth.sup.(4) .sup.(1)Change of YE manufacturer: Springer product is food grade, BD product is bioproduct grade. .sup.(2)Change of antifoam manufacturer: Sigma product is reagent grade, Dow Corning product is medical grade. .sup.(3)Change of aeration condition: Reduced aeration volume to suppress excessive foam generation due to antifoam change. .sup.(4)Elimination of kanamycin antibiotic-related regulatory issue. Summary of results of confirming reproducibility of 5 L-scale optimal culture conditions

Experimental Example (3) 50 L-Scale Optimal Condition Scale-Up Process (Sample Production)

[0311] A 50 L culture process for sample production was conducted by scale-up of the optimal conditions for developing a 5 L-scale process. A feeding rate of a 50 L scale fed-batch culture was calculated by applying the same parameters as in the 5 L scale feeding strategy (Table 33). As a result of 2 batches (Batch No. 170104, 170114) of 50 L scale sample production culture, cell productivity was 207 and 208 g/L, GST-Tβ4 productivity was 76.9 and 69.7 mg/g (productivity per pellet), respectively (Table 34), the target standards of the optimal condition response values (Table 29, cell productivity ≥180 g/L & GST-Tβ4 productivity ≥54.1 mg/g) were satisfied, and growth and metabolism (glucose & acetate) trends also showed the same patterns as a L-scale optimal condition-reproducible batch (FIG. 18). In the case of the batch (batch No. 170114) conducted as a sample production process, from the end of culture to the recovery of a culture broth, the temperature was cooled down to 10° C. in an incubator, followed by conducting an overnight (hereinafter, “O/N”) holding test. The OD.sub.600 values before and after O/N were 174.6 (before) and 171.5 (after), and the pellet productivity was measured to be 221 g/L (before) and 208 g/L (after), respectively (Table 34). Results for the sample production (batch No. 170114) are summarized in Table 35.

TABLE-US-00034 TABLE 33 t X V F(g/min) μ 0~4 Batch culture 4 8 2.50 2.3 0.3 5 11 2.51 3.1 0.3 6 14 2.53 3.8 0.3 7 19 2.55 4.1 0.2 8 23 2.57 4.3 0.2 9 28 2.59 4.7 0.18 10 34 2.62 5.6 0.18 11 40 2.65 6.7 0.18 12 47 2.69 8.1 0.18 13 55 2.73 8.5 0.16 14 64 2.78 10.0 0.16 15 73 2.83 11.7 0.16 16 84 2.90 11.9 0.14 17 95 2.96 13.7 0.14 18 106 3.04 15.7 0.14 19 119 3.13 18.1 0.14 20 132 3.23 20.8 0.14 Exponential feeding rate in 50 L scale-up process

TABLE-US-00035 TABLE 34 Induction Induction Cell GST-TB4 Batch Start temperature Final productivity productivity No. (OD.sub.600) (° C.) OD.sub.600 (g/L) (mg/g) 170104 55.4 30 169.2 207 76.9 170114 50.4 30 174.6 (End of culture) 221 (End of culture) 69.7 171.5 (10° C. O/N) 208 (10° C. O/N) Summary of results for 50 L optimal condition scale-up culture

TABLE-US-00036 TABLE 35 Process step Result Seed culture 1 Culture time 5 h Final cell OD.sub.600 = 2.1 concentration Seed culture 2 Culture time 3 h Final cell OD.sub.600 = 2.4, 2.5, 2.6, 2.6 concentration Main culture Feed-A start OD.sub.600 = 6.2 (4 h) Induction start OD.sub.600 = 50.4 (11.5 h) Culture end OD.sub.600 = 174.6 (20 h) 15° C. overnight OD.sub.600 = 171.5 (39 h) Volume of final 36.0 kg (20.8% of solid content) culture broth Continuous 1.sup.st recovery and 48.7 kg (13.8% of solid content) centrifugation washing with DIW (recovery of 2.sup.nd recovery and 52.7 kg (11.6% of solid content) culture broth) washing with DIW 3 .sup.rd recovery 13.6 kg (36.2% of solid content) (final cell slurry) Cell lysis Lysis buffer Cell slurry: 13.6 kg suspension 20X lysis buffer: 1.63 kg DIW: 17.4 kg Total weight: 32.6 kg Measurement of 1.sup.st 107.0 (before)/13.2 (after) lysis OD.sub.600 Measurement of 2.sup.nd 13.2 (before)/8.1 (after) lysis OD.sub.600 Lysis rate 7.5% Cell lysate 32.7 kg Continuous centrifugation 24.8 kg (Recovery of supernatant of cell lysate) Filter washing C0HC washing 22.0 L (DIW) F0HC washing 22.0 L Opticap XL10 washing 6.9 L Cell lysate supernatant 24.8 kg (input) Depth Flushing (DIW) 11.6 kg filtration Final recovery after 34.2 kg membrane filter Summary of results for 50 L sample production (batch No. 170114)

[0312] [2. Purification Process]

Example (1) Background and Preliminary Experiment

[0313] The Chinese process conducted with Q FF-GST FF-enzyme digestion-UF/DF-Q XL-Source 30RPC-Q HP-UF/DF was compared with a more optimized Huons process conducted with Q FF-GST FF-enzyme digestion-Q FF-UF/DF. Accordingly, based on the Huons process, experiments were conducted with the aim of improving purity, optimizing a unit process and minimizing a high cost of GST FF resins.

[0314] (1-1) Resin Screening of Primary Column

[0315] To maximize a DBC for expensive GST columns, instead of a Q FF resin that had been used as a primary column in the Chinese process, Nuvia Q, Factrogel TMAE(M) was applied to confirm an effect of improving a DBC for a GST resin by the improvement in purity of an eluted fraction. When an elution pool purified using two types of resins was applied to a GST column, the DBC was approximately 42 to 44 mg/mL regardless of resin type (Tables 36 and 37). Therefore, it was decided to use Fractogel TMAE(M), which has a relatively high yield, in process development.

TABLE-US-00037 TABLE 36 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) % AEX (Nuvia Q)* 0.98 50 48.9 0.53 77 41.0 83.8 Affinity** 0.53 28.3 15 0.54 7.7 4.2 *CV of Nuvia Q: 6.8 mL, Loading amount: 7.2 mg/mL of resin **Affinity CV: 1 mL columns were used Step yield for Nuvia Q resin and DBC for Affinity resin using Nuvia Q eluted sample

TABLE-US-00038 TABLE 37 Loading Sample Eluted Pool Conc. Conc. Vol- (mg/ Volume Amount (mg/ ume Amount Yield mL) (mL) (mg) mL) (mL) (mg) (%) AEX- 1.17 50 58.6 0.66 87.3 57.5 98.1 (TMAE(M))* Affinity** 0.66 23 15 0.44 10.0 4.4 *CV of TMAE (M): 8.6 mL, loading amount: 6.8 mg/mL of resin **Affinity CV: 1 mL columns were used Step yield for TMAE(M) resin and DBC for Affinity resin using TMAE(M) eluted sample

[0316] (1-2) 50 L Verification Run Test

[0317] Based on the Huons process, the reproducibility of a small scale process was confirmed, and both of a 1.sup.st verification run and a 2.sup.nd verification run were carried out with a 50 L scale culture broth (FIG. 19, Tables 38 and 39).

TABLE-US-00039 TABLE 38 Items Process conditions Cell Lysis & Recovery- Cell pellet 1366 g, 1419 g (Batch No. 160629-50L) Depth filtration Microfluidizer: 15,000 psi, 3 pass centrifugation (8500 rpm, 50 min, 10° C.) C0HC (0.11 m.sup.2) × 2ea + F0HC (0.11 m.sup.2) + Opticap XL10 (0.45 μm) AEXI Resin Fractogel EMD TMAE(M) Column scale BPG 200 (i.d 20 cm, H 13 cm), CV 4145 mL Equilibration buffer 20 mM Tris-HCl, pH 8.0 Elution buffer 20 mM Tris-HCl, 500 mM NaCl, pH 8.0 Linear flow rate 61 cm/hr (Flow rate 319 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity <2.8 mS/cm Elution pooling Start point: UV.sub.280 nm Value ≥100 mAU criteria End point: UV.sub.280 nm Value ≤100 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) Affinity Resin Glutathione Sepharose 4 Fast Flow Column scale BPG 200 (i.d 20 cm, h 11 cm), CV 3454 mL Equilibration buffer 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 Elution buffer 50 mM Tris-HCl, 10 mM Glutathione (Reduced), pH 8.0 Acidic buffer 100 mM Sodium Acetate, 500 mM NaCl, pH 4.5 Alkaline buffer 100 mM Tris-HCl, 500 mM NaCl, pH 8.0 Linear flow rate 63 cm/hr (Flow rate 329 mL/min) Loading sample Directly Loading condition Elution pooling Start point: UV.sub.280 nm Value ≥100 mAU criteria End point: UV.sub.280 nm Value ≤100 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) Acidic buffer (5 CV) Alkaline buffer (5 CV) 6M Guanidine hydrochloride (2 CV) DW (2 CV) Equilibration buffer (3 CV) Thrombin Enzyme Bovine Thrombin Reaction & Enzyme ratio 1 mg GST-Tβ4 per 2 unit thrombin AEXII Reaction time 18 ± 2 H Reaction temp 22 ± 2° C. Agitation 60 rpm Resin Fractogel EMA TMAE(M) Column scale BPG 100 (i.d 10 cm, H 14.2 cm), CV 1115 mL Equilibration buffer 20 mM Tris-HCl, pH 8.0 Elution buffer 20 mM Tris-HCl, 100 mM NaCl, pH 8.0 Linear flow rate 170 cm/hr (Flow rate 223 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity <2.8 mS/cm Elution pooling Start point: UV.sub.214 nm Value ≥50 mAU criteria End point: UV.sub.214 nm Value ≤50 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) Conditions for 1.sup.st 50 L verification run process

[0318] The 1.sup.st verification run was conducted in the same manner as used in the small scale process, and the results are shown as follows (Tables 39 and 40).

TABLE-US-00040 TABLE 39 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) (%) Qunatitation Cell breakage 1144 (160602_50) Depth Filtration 1,144 1326 Densitometer Microfiltration 1030 966.5 AEX I Preparation of LS 2.45 200.0 489.8 1.662 309.0 513.6 104.9% Densitometer Process 1.66 257.0 427.1 1.24 385.0 475.5 111.3% Densitometer Gluathione S4FF 1.24 175.6 216.9 1.99 81.9 163.1  75.2% Densitometer Thromblin 1.99 79.0 157.4 0.28 91.0 25.5  16.2% UV/SEC AEX II Preparation of LS 0.28 89.0 24.9 0.08 254.0 20.8  83.6% SEC Process 0.08 251.0 20.6 0.32 21.0 7  32.6% SEC UF/DF 0.32 18.0 5.8 9.90 0.5 5  85.9% SEC/RP Total Yield (%)   3.3% Step Yield of HU024 Small Scale verification run

TABLE-US-00041 TABLE 40 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) (%) Qunatitation Cell breakage* 11,065 Depth Filtration&MF* 11,605 43,968 27,795 37,815 Densitometer AEXI* 1.36 27,358 37,308 2.14 16,981 36,322 96.1% Densitometer Gluathione S4FF** 2.14 16,981 36,322 3.21  6,845 21,972 60.5% Densitometer Thrombin 3.21  6,845 21,972 0.48  7,850  3,768 17.1% UV/SEC AEX II Flow through 0.09 41,060  3,695 0.06 33,494  1,909 51.7% RP (1.1L CV) Elution 1.19    955  1,140 30.9% RP UF/DF*** 0.06 33,494  1,909 9.95    188  1,866 97.7% RP Total Yield (%)  4.3% *Recovery & AEXI were the results obtained by 2 cycles **Affinity Process was the result obtained by 3 cycles ***Resulted from flow through fraction only Step yield of 1.sup.st HU024 verification run

[0319] Referring to the small scale purification results and the 1.sup.st verification run results, many target proteins were confirmed from flow through (F/T) fractions in an AEXII process (AEX process after thrombin digestion) (Table 40). Accordingly, in the 2.sup.nd verification, the process was conducted in a flow through mode (Table 41).

TABLE-US-00042 TABLE 41 Items Process conditions Cell Lysis & Recovery Cell pellet 1500 g, 1530 g (Batch No. 160729.-50L) Depth filtration Microfluidizer: 15,000 psi, 3 pass Tubular continuous centrifugation (15000 rpm, feed rate 700 mL/min) C0HC (0.11 m.sup.2) × 2ea + F0HC (0.11 m.sup.2) × 2ea + Opticap XL10 (0.45 μm) AEXI Column scale BPG 200 (i.d 20 cm, H 13 cm), CV 4145 mL Equilibration buffer 20 mM Tris-HCl, pH 8.0 Elution buffer 20 mM Tris-HCl, 500 mM NaCl, pH 8.0 Linear flow rate 61 cm/hr (Flow rate 319 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity <2.8 mS/cm Elution pooling Start point: UV.sub.280 nm Value ≥100 mAU criteria End point: UV.sub.280 nm Value ≤100 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) Affinity Column Scale BPG 200 (i.d 20 cm, h 11 cm), CV 3454 mL Equilibration buffer 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 Elution buffer 50 mM Tris-HCl, 10 mM Glutathione (Reduced), pH 8.0 Acidic buffer 100 mM Sodium Acetate, 500 mM NaCl, pH 4.5 Alkaline buffer 100 mM Tris-HCl, 500 mM NaCl, pH 8.0 Linear flow rate 63 cm/hr (Flow rate 329 mL/min) Loading sample Direct Loading condition Elution pooling Start point: UV.sub.230 nm Value ≥100 mAU criteria End point: UV.sub.280 nm Value ≤100 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) Acidic buffer (5 CV) Alkaline buffer (5 CV) 6M Guanidine hydrochloride (2 CV) DW (2CV) Equilibration buffer (3CV) Thrombin Enzyme Bovine Thrombin Reaction & Enzyme ratio 1 mg GST-Tβ4 per 2 unit thrombin AEXII Reaction time 18 ± 2 H Reaction temp 22 ± 2° C. Agitation 60 rpm Column scale BPG 100 (i.d 10 cm, H 14.2 cm), CV 1115 mL Equilibration buffer 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 Linear flow rate 170 cm/hr (Flow rate 223 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity <2.8 mS/cm Elution pooling Start point: UV.sub.214 nm Value ≥50 mAU criteria End point: UV.sub.214 nm Value ≤50 mAU Process Equilibration buffer (2 CV) Sample Loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) *Each process resin was the same as used in 1st verification run Conditions for 2.sup.nd 50 L verification run process

TABLE-US-00043 TABLE 42 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) (%) Qunatitation Cell breakage* 14,036 Depth Filtration&MF* 14,036 52,159 42,720 45,196 Densitometer AEXI* 1.06 42,202 44,644 1.50 27,330 40,995 91.8% Densitometer Glutathione S4FF** 1.50 25,866 39,699 3.83  6,919 26,500 66.8% Densitorneter Thrombin 3.83  6,901 26,431 0.55  7,907  4,349 16.5% UV/SEC AEX II 0.37 12,049  4,349 0.24 16,672  4,018 92.4% SEC UF/DF*** 0.24 16,672  4,018 9.87    329  3,250 80.9% RP Total Yield (%)  7.6% *Recovery & AEXI were the results obtained by 2 cycles **Affinity Process was the result obtained by 3 cycles ***UF/DF were the UF/DF value obtained by reprocessing twice due to endotoxin Step yield of 2.sup.nd HU024 verification run

[0320] In the 1st verification run, the purities of the F/T fractions in the AEXII process were confirmed to be 99.4% (RP-HPLC) and 97.9% (SE-HPLC), whereas the purities of eluted fractions were confirmed to be non-standard (Table 43). In the case of the 2.sup.nd Verification run, when the AEXII process was conducted in a flow through mode, a purification yield was improved (4.3-≥7.6%), but the purity was slightly reduced (97.9-≥97.0%), compared to the 1st verification (Table 44).

TABLE-US-00044 TABLE 43 HCP Conc. HCP Conc. HCP Endotoxin Purity (%) Purity (%) Sample (mg/mL) (ng/mL) (ppm) (EU/mg) (SEC-HPLC) (RP-HPLC) AEXI 2.14 10,193 4,793 34,963 Gluathione S4FF* 3.71 1,041 324 3,121 Thrombin 0.48 982 2,045 29,042 AEX II Flow through 0.06 — — 0.08 Elution 1.19 143 120 0.21 UF/DF Flow through 9.97 217 22 0.76 97.9 99.4 Elution 7.93 1178 149 1.49 93.4 93.4 Quality profile for 1.sup.st verification run

TABLE-US-00045 TABLE 44 Purity Purity HCP (%) (%) Conc. HCP Conc. HCP Endotoxin (SEC- (RP- Sample (mg/mL) (ng/mL) (ppm) (EU/mg) HPLC) HPLC) AEXI 1.50 21,690 14,460 98,973 Gluathione 3.83 1,252 327 1,120 S4FF** Thrombin 0.55 1,250 2,273 6,411 AEX II 0.24 14 58 0.02 UF/DF 11.5 560 49 2.58 96.9 97.1 UF/DF 9.87 — — 0.35 97.0 98.7 (Re- processing) *reprocessing was the sample to which 2 cycles of AEX(F/T) ---> AEX(B/E) was applied Quality profile for 2.sup.nd verification run

[0321] However, in the two verification runs, the SE-HPLC purity of the final purified extract was less than 98%. Therefore, the process development was planned to improve purity to be 98% or more, in addition to the optimization of the unit process.

[0322] (1-3) Optimization of Cell Disruption Conditions

[0323] For the unit process optimization, it was checked whether to improve cell disruption efficiency.

[0324] (1-3-1) Experiment of Improving Lysis Buffer

[0325] The possibility of improving cell disruption efficiency was confirmed. When EDTA and Triton X-100 as a detergent were added, the increase in recovered target protein content was insignificant (approximately 4 to 6% increase), so that it was determined to stick to the previous conditions (FIG. 20 and Table 45).

TABLE-US-00046 TABLE 45 Conditions Control 3 mM DTT 10 mM EDTA 20 mM EDTA Band Vol (OD*mm.sup.2) 21.7 21.1 23.3 22.3 Percentage (%) 100.0 97.0 106.7 103.5 0.1% 0.5% 0.1M 0.2M Conditions TX-100 TX-100 Guanidine-HCl Guanidine-HCl Band Vol (OD*mm.sup.2) 22.1 22.3 16.7 15.9 Percentage (%) 102.7 104.1 78.4 74.8 *Basic buffer composition: 20 mM Tris-HCl, pH 8.0 Conditions for comparison GST-Tβ4 band density by cell disruption buffer

[0326] (1-3-2) Impurity Removal Experiment

[0327] According to NR (Chinese process), the possibility of removing impurities by utilizing salting-out caused by the addition of ammonium sulfate was checked. In the NR process, a process of adding 5.6 g (33%) of ammonium sulfate per 100 mL of a lysate solution was applied, whereas in this experiment, a GST-Tβ4 amount in a lysate solution treated with 16.5%, 33% or 66% of ammonium sulfate was confirmed. As the ammonium sulfate content increased, the GST-Tβ4 content was decreased by 20%, 37% or 43%, compared to a sample without ammonium sulfate (FIG. 21 and Table 46). When the ammonium sulfate was used, the GST-Tβ4 contents were also reduced, so that it was determined not to add ammonium sulfate.

TABLE-US-00047 TABLE 46 GST-Tβ4 Relative rate of Condition Amount (ug)* target Loss 0% Ammonium sulfate 2.4  0% 16.5% Ammonium sulfate 1.9 20% 33.0% Ammonium sulfate 1.5 37% 66.0% Ammonium sulfate 1.4 43% *Quantitation by Densitometer Relative reduction of GST-Tβ4 amount by adding ammonium sulfate

[0328] (1-3-3) Changes of Disruption Conditions

[0329] Microfluidizer conditions were changed. The cell disruption conditions used in two verification runs (30% cell slurry, a pressure of 15,000 psi, and disruption repeated three times) were changed into the cell disruption conditions currently applied in a Binex GMP facility (15% cell slurry, a pressure of 11,000 psi, and disruption repeated two times). When the changed disruption conditions were applied, a degree of reducing optical density (OD.sub.600) of a sample after disruption was equal to or higher than that of the conventional method (Table 47). The changed conditions were applied from the development of a purification process using a 5 L scale culture broth.

TABLE-US-00048 TABLE 47 Cell disruption condition 15,000 psi, 3pass, 11,000 psi, 2pass. 30% slurry 15% slurry 50 L 50 L 2.sup.nd Depth 50 L Step 1.sup.st Run Run test 3.sup.rd Run Cell slurry (OD.sub.600) 148 141.5 95.6 107 1.sup.st Disruption (OD.sub.600) 62.7 47.6 17.2 13.2 2.sup.nd Disruption (OD.sub.600) 35.2 24.6 5.6 8.1 3.sup.rd Disruption (OD.sub.600) 22.9 13.5 N/A N/A End OD/initial OD(%) 15.4 9.5 5.8 7.5 Cell optical density according to different conditions of disruption

[0330] (1-3-4) Depth Filtration Sizing Test

[0331] According to the first result of a filter test provided by Merck Millipore, C0HC and F0HC were selected, and applied to verification runs. Filter clogging occurred during a process, and the cause thereof was not analyzed yet. Afterward, the same filter was used in 5 L and 50 L processes, but clogging was not found. Therefore, it was confirmed that clogging occurred regardless of filter type, and clogging was expected to be resolved by changing disruption conditions.

[0332] After changing the disruption conditions, a second filter test was requested at Merck Millipore, and X0HC was suggested. A filter test was requested at Satorius, and thus Sartoclear PC2 was selected (Table 48). As a result, a depth filter manufactured by Merck Millipore was used, and a 5 L process and a 3.sup.rd 50 L verification run were conducted with C0HC and F0HC (Merck report Revision #2 <2016.06.17>, Merck report revision #1 <2016.10.20>, Sartorius report <Doc No. SKB201604>).

TABLE-US-00049 TABLE 48 Depth filter Nominal Sterile Pore Company type retention μm filter size(μm) Merck Millipore X0HC 0.1 Durapore 0.45 Sartorius PC2 0.3 Sartopore2 0.45, 02 (Double layer) Information on depth filter and sterile filter recommended by each vendor

Example (2) Analysis Conditions for Each Process Sample

[0333] (2-1) UV Quantification Method

[0334] Protein quantification was performed by the following equation (Equation 1), and used for protein quantification of a process sample for an affinity process.

[00003]$\begin{array}{cc}\mathrm{Protien}\mathrm{Concnetration}\left(\mathrm{mg}/\mathrm{mL}\right)=\frac{Abs\left(280\mathrm{nm}\right)}{1.385}& \left[\mathrm{Equation}1\right]\end{array}$

[0335] (2-2) SEC-HPLC Analysis

[0336] Analysis was conducted on a process sample after thrombin digestion, and SEC-HPLC analysis for the process sample was conducted according to a sample size-exclusion chromatography (SEC-HPLC) quantification test and a Gly-TB4 SEC-HPLC purity test (according to Standard Operating Procedure (SOP) for pharmaceuticals) revised on Nov. 16, 2016) enabling purity measurement in a Gly-TB4 process enabling quantitative measurement of Gly-TB4 using conditions enabling GST and Gly-TB4 separation. The conditions are as follows.

TABLE-US-00050 TABLE 49 Items Conditions Column TSKgel G2000swxl (7.8 × 300 mm, 5 μm particle) TSKgel Guard column (6.0 × 40 mm) Flow rate 0.6 mL/min Injection volume 20 μL Mobile phase 50 mM Sodium phosphate, 0.3M NaCl, 10% ACN, pH 7.2 ± 0.1 Column temperature 30 ± 5° C. Autosampler  5 ± 5° C. temperature Wavelength 214 nm Running 30 time min Sample Dilution with formulation buffer preparation (Formulation Buffer: 20 mM Sodium phosphate, 150 mM NaCl, pH 8.0) SEC-HPLC Analysis Method and Conditions

[0337] (2-3) RP-HPLC Analysis

[0338] Analysis was conducted on a process sample after thrombin digestion, and RP-HPLC analysis for the process sample was conducted according to a Gly-TB4 RP-HPLC quantification test enabling quantification measurement and a Gly-TB4 RP-HPLC purity test enabling purity measurement (according to SOP for pharmaceuticals revised on Nov. 16, 2016) using conditions enabling Gly-TB4 and impurity separation. The conditions are as follows. In addition, the purities of purified extracts obtained from 1.sup.st and 2.sup.nd verification runs that had been conducted before SOP revision were considered to be unreliable measurement values.

TABLE-US-00051 TABLE 50 Items Conditions SOP ver. 1.0 2.0 Column Agilent zorbax 300SB C18 (4.6 × 250 mm, 5 μm particle) Agilent 300SB C18 guard column (4.6 × 12.5 mm) Flow rate 0.5 mL/min 0.8 mL/min Injection  20 μL volume Mobile A: Water (w/0.1% TFA) A: Water (w/0.1% TEA) phase B: ACN B: 80% ACN (w/0.1% TFA) (w/0.085% TEA) Column 30 ± 5° C. 45 ± 2° C. temperature Autosampler  5 ± 2° C. temperature Wavelength 220 nm Running 45 min 45 min time Gradient Time (min) A (%) B (%) Time (min) A (%) B (%) 0 85 15 0 91 9 method 5 85 15 5 91 9 25 50 50 30 64 36 25.1 10 90 31 10 90 30 10 90 35 10 90 30.1 85 15 36 91 9 45 85 15 45 91 9 Sample Dilution with formulation buffer preparation Method and Conditions for RP-HPLC analysis

[0339] (2-4) Endotoxin Analysis

[0340] Endotoxin quantification was carried out according to the endotoxin test method using PTS of Gly-TB4 developed by Binex, thereby identifying endotoxin in the process sample.

[0341] (2-5) Host Cell Protein (HCP) Analysis

[0342] HCP quantification was carried out according to the HCP analysis of Gly-TB4 developed by Binex, thereby identifying the tendency of removing HCPs from the process sample.

[0343] (2-6) Host Cell DNA (HCD) Analysis

[0344] HCD quantification was carried out according to residual E. coli HCD analysis of Gly-TB4 developed by Binex, thereby confirming a residual HCD amount of the final product.

[0345] (2-7) SDS-PAGE Analysis

[0346] SDS-PAGE analysis was carried out according to electrophoresis of Gly-TB4 using conditions for separating Gly-TB4 and impurities by size, thereby qualitatively confirming impurity profiles, and cell recovery and an AEXI eluted pool were quantitatively analyzed with BSA as a reference or an affinity pool using a densitometer, as needed.

Example (3) Anion Exchange Chromatography (AEXI)

[0347] (3-1) Preliminary Experiment

[0348] A preliminary experiment was conducted for improving purity by lowering an NaCl concentration in an elution buffer as much as possible.

[0349] (3-1-1) Confirmation of Target Protein Elution Site

[0350] A NaCl concentration at which GST-Tβ4 was eluted was screened under a linear gradient condition in an NaCl range of 0-500 mM using 2 mg of purified GST-Tβ4 (160709, 2.sup.nd column eluted fraction IPC sample acquired from 1.sup.st verification run). As a result, the location of the apex of the eluted main peak was identified at 20 mS/cm, and it was confirmed that an elution was conducted in a range from 200 to 300 mM NaCl.

[0351] (3-1-2) GST-T114 Quantitative Analysis (Simple Purification)

[0352] In the case of an AEXI process, it was impossible to measure a step yield through UV quantification due to a large quantity of impurities in a loading sample and an elution sample. An analysis method for plotting the standard curve of the peak area of a standard material using an affinity column and confirming the amount of a target protein was intended to be developed (Table 51).

TABLE-US-00052 TABLE 51 Items Process conditions Resin Glutathione Sepharose 4 Fast Flow Column LRC 10 (ID 1.0 cm, Height 1.2 cm), Column Volume (0.94 mL) Flow rate 1 mL/min Process Equilibration 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 (5 CV) Sample Injection 500 μL (Dilution with 20 mM Tris-HCl, 500 mM NaCl, pH 8.0) Unbound wash 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 (5 CV) Elution 50 mM Tris-HCl, 10 mM Glutathione, pH 8.0 (5 CV) Conditions for GST-Tβ4 Quantity Analysis

[0353] 0.3, 0.6, 0.9 and 1.2 mg/mL of standard materials were prepared through UV quantification (absorbance coefficient: 1.385) by buffer exchange of the 1.sup.st verification run affinity pool with a dilution buffer (20 mM Tris-HCl, 500 mM NaCl, pH8), and analyzed by the methods listed in Table 51, and then an integrated area was calculated using an AKTA operation program (Unicon 5.11). A linearity curve was plotted using the calculated values, resulting in obtaining a curve having a linearity of 0.99 or more, and AEXI process samples (loading sample and elution pool) were analyzed.

[0354] (3-2) Flow Rate Verification Test for Shortening Process Time

[0355] (3-2-1) Purpose of Experiment

[0356] The process was conducted while a linear velocity was maintained at 61 cm/hr, and the linear velocity was increased up to 200 cm/hr to shorten the time required for a unit process. The maintenance of a step yield and a purity was confirmed.

[0357] (3-2-2) Experimental Method

[0358] The AEXI process was performed as shown in the following table (Table 52). Experiments were carried out by applying flow rates of 61 cm/hr and 200 cm/hr, respectively. Samples obtained before and after the process were analyzed by the method described in (3-1-2), and impurity profiles were confirmed through SDS-PAGE.

TABLE-US-00053 TABLE 52 Items Process conditions Resin Fractogel EMD TMAE(M) Column XK16 (ID 1.6 cm, Height 20 cm), Column Volume (40 mL) Flow rate 61 cm/hr, 200 cm/hr Sample preparation Cell cake (Batch No. 160629-50L) .fwdarw. Prepare 15% (w/w) slurry 11,000 psi, 2pass disruption Centrifuge (8,500 rpm, 50 min, 10° C.) Microfiltration (0.45 + 0.22 μm Double Layer) .fwdarw. Adjust pH and conductivity with 1M Tris (pH 8.0, ≥2.8 mS/cm) Process Equilibration 20 mM Tris-HCl, pH 8.0 (2 CV) Sample loading Cell Lysate adjusted pH and conductivity (4.2~4.9 mg GST-Tβ4/mL) Unbound wash 20 mM Tris-HCl, pH 8.0 (5 CV) Elution 20 mM Tris-HCl, 500 mM NaCl, pH 8.0 (5 CV) (100 mAU~100 mAU) Strip 2M NaCl (3 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 20 mM Tris-HCl, pH 8.0 (3 CV) Conditions for AEXI Process

[0359] (3-2-3) Result of Experiment

[0360] The chromatograms according to a flow rate after the AEXI process are shown in FIGS. 23 and 24, and the area indicated by a red dotted circle corresponds to an elution peak.

[0361] As a result of a step yield (Table 53) and SDS-PAGE analysis (FIG. 25) for an elution pool under each condition, it was confirmed that, even when the flow rate in the AEXI process was increased to 200 cm/hr, there was no effect on step yield and purity. However, since the load amount was approximately 5 mg/mL resin, reconfirmation was required when the load amount increased.

TABLE-US-00054 TABLE 53 Start material Elution pool Volume Amounts Concentration Volume Amounts Yield Sample (mL) (mg) (mg/mL) (mL) (mg) (%)  61 80 167.5 0.93 160 149 89% cm/hr 200 80 167.5 0.93 155 144 86% cm/hr Yield for AEXI Linear flow rate study

[0362] (3-3) AEXI Elution Condition Test for Purity Improvement

[0363] (3-3-1) Purpose of Experiment

[0364] Considering the economic feasibility of an expensive affinity column (GST FF), a DBC was intended to be maximized. To this end, the purity of a loading sample was raised as high as possible. Therefore, the experiment was conducted by changing a 500 mM NaCl condition into a 200 mM NaCl condition, and then a step yield and purity were comparatively analyzed.

[0365] (3-3-2) Experimental Method

[0366] An experimental method was basically the same as shown in Table 51 of (3-2-2) (Cell cake batch No. 160603 used in this experiment), and the experiment was conducted with an elution buffer having an NaCl concentration of 500 mM or 200 mM NaCl. HU024 was identified by analyzing samples before and after the process according to the method described in (2-1-2), and impurity profiles were confirmed through SDS-PAGE.

[0367] (3-3-3) Result of Experiment

[0368] The chromatograms according to an elution buffer composition after the AEXI process are shown in FIGS. 26 and 27, and the area indicated by a red dotted circle corresponds to an elution peak.

[0369] According to the step yield result (Table 54) and SDS-PAGE analysis result (FIG. 28) for the elution pools by condition, there was no difference in AEXI yield, and as a result of the SDS-PAGE analysis, the major impurity peak below the target band was observed in the strip pool, rather than the elution pool, thus, a purity improving effect was expected. However, it was confirmed that an elution volume increased with a decreased NaCl concentration. In addition, when the load amount is increased by 5 mg/mL resin or more, additional experiments may be required.

TABLE-US-00055 TABLE 54 Elution pool Start material Con- Volume Amounts centration Volume Amounts Yield Sample (mL) (mg) (mg/mL) (mL) (mg) (%) Control 80 142 0.79 146 115 81% (500 mM NaCl) 200 mM 80 142 0.69 179 123 87% NaCl Yield for AEXI Elution buffer study

[0370] (3-4) AEXI Loading Sample Condition Study

[0371] (3-4-1) Purpose of Experiment

[0372] The conductivity of a loading sample in the AEXI column is known as an important process variable by which a target protein is able to bind to the column. This experiment was conducted to confirm a conductivity range of an applicable loading sample, and the change in impurity content and step yield according to the variation.

[0373] (3-4-2) Experimental Method

[0374] An experimental method was basically the same as shown in Table 51 of (3-2-2) (Cell cake batch No. 160603 used in this experiment), and the loading samples were prepared to have a conductivity of 10 mS/cm (80 mM NaCl), 8 mS/cm (50 mM NaCl), and <2.8 mS/cm as a control, and an equilibrium buffer was prepared to have an NaCl concentration of 80 mM, 50 mM or 0 mM. The process was conducted with an elution buffer containing 200 mM NaCl. HU024 in the samples before and after the process was analyzed in the same manner as described in (2-1-2), and impurity profiles were confirmed through SDS-PAGE.

[0375] (3-4-3) Result of Experiment

[0376] The chromatograms according to the change in conductivity of the loading samples and equilibration buffer after AEXI process are shown in FIGS. 29 to 31, and the area indicated by a red dotted circle corresponds to an elution peak.

[0377] Referring to the yields (Table 55) and SDS-PAGE result (FIG. 32) for the elution pool by condition, in the cases of the control (0 mM NaCl) and 50 mM NaCl, the yields were 81% and 97%, respectively, but in the case of 80 mM NaCl, the yield was 19.1%, and then F/T fractions were analyzed. According to the analysis result, it was confirmed that approximately 57% of the target proteins were observed in the F/T fractions. According to SDS-PAGE, it was confirmed that there was no difference in purity between the control and the 50 mM NaCl condition.

TABLE-US-00056 TABLE 55 Start material Elution pool Salt conc. Vol- Conc. Vol- of load ume Amounts (mg/ ume Amounts Yield Re- sample (mL) (mg) mL) (mL) (mg) (%) mark Control 80 198 0.87 184 161 81% N/A (0 mM) 50 mM 80 171 0.82 201 165 97% N/A NaCl 80 mM 80 158 0.17 180 30.1 19% FIT NaCl 0.34 268 89.9 57% Elution Step yield with different conductivities of loading sample for AEXI

[0378] According to the result of the test for the conductivities of a loading sample and an equilibrium buffer, it can be confirmed that, when the conductivities of loading samples were ≤2.8 mS/cm and 8.0 mS/cm (50 mM NaCl), a step yield and purity were the same. However, when the conductivity was 10 mS/cm (80 mM NaCl), the target protein was lost by F/T, so that it was confirmed to be inappropriate to apply, and it was confirmed that the conductivity of a loading sample in the AEXI process up to 8.0 mS/cm does not significantly affect AEXI binding. However, when the load amount was increased by 5 mg/mL resin or more, additional experiments were needed.

[0379] (3-5) AEXI Dynamic Binding capacity (DBC) Test

[0380] (3-5-1) Purpose of experiment

[0381] A DBC for GST-Tβ4 with respect to AEXI was to be confirmed.

[0382] (3-5-2) Experimental method

[0383] An experimental method was basically the same as shown in Table 51 of (3-2-2) (Cell cake batch No. 160603 used in this experiment), but a LRC10 column was packed with 2 mL of Fractogel TMAE(M), a working linear flow rate was 150 cm/hr (Flow rate 2 mL/min), and compositions for an elution buffer and an EQ buffer were the same as the conditions established in process development (EQ: 20 mM Tris-HCl, 50 mM NaCl, pH 8.0, Elution: 20 mM Tris-HCl, 200 mM NaCl, pH 8.0). The process was conducted with a load amount of 10 mg, 8 mg or 7.5 mg GST-Tβ4/mL of resin (the conventional experiment condition was <5 mg/mL of resin). F/T and elution pools were analyzed by SDS-PAGE, and the location of the target protein was confirmed.

[0384] (3-5-3) Result of Experiment

[0385] The result of analyzing the F/T and elution pools through SDS-PAGE is shown in FIG. 33. When 7.5 mg of GST-Tβ4 was injected, the target protein was not found in the F/T pool, and when 8.0 or 10.0 mg of GST-Tβ4 was injected, it was confirmed that the larger the load amount, the more target bands in the F/T pool. Accordingly, the DBC of AEXI was expected to be 7.5 mg/mL or less.

[0386] (3-6) Hold Time Study for AEXI Process Samples

[0387] (3-6-1) Purpose of Experiment

[0388] To secure the storage condition and period that can be applied to GMP production, in an AEXI process, eluted fraction samples were stored in a cold room (2 to 8° C.) and at room temperature (18 to 22° C.) for 7 days, followed by confirming the change in purity.

[0389] (3-6-2) Experimental Method

[0390] An experimental method was basically the same as shown in Table 52 of (3-2-2) (Cell cake batch No. 160906 used in this experiment), and the experiment was conducted under the conductivity of a loading sample of 8 mS/cm (50 mM NaCl), and an elution condition of 200 mM NaCl. An eluted pool was subjected to sterile filtration, divided into microtubes by 500 uL, and stored for 7 days in refrigeration and room temperature conditions. The sampled samples were stored at −70° C., completely frozen, and simultaneously thawed one day after the storage period, followed by SDS-PAGE.

[0391] (3-6-3) Result of Experiment

[0392] The SDS-PAGE result is shown in FIG. 34. There was no significant difference between the samples stored in the refrigerator. However, from day 2 of the room temperature storage, it was confirmed that the intensity of the band corresponding to approximately 55.4 kDa increased, and the intensity of the band corresponding to approximately 40 kDa decreased. Accordingly, it was confirmed that the AEXI process sample was stable for 7 days in a cold room condition (2 to 8° C.), but unstable in a room temperature (18 to 22° C.) condition.

[0393] (3-7) Summary of AEXI Process

TABLE-US-00057 TABLE 56 Items Process conditions Resin Fractogel EMD TMAE(M) Dynamic binding capacity Not more than 7.5 mg/mL Equilibration buffer 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 Elution buffer* 20 mM Tris-HCl, 200 mM NaCl, pH 8.0 Process temperature 15~25° C. Linear flow 200 rate cm/hr Loading sample pH 8.0 ± 0.1 (adjustment with 1M Tris) condition Conductivity max 8.0 mS/cm (50 mM NaCl) Process Equilibration 2 CV Sample loading Sample Unbound wash (with Equilibration buffer) 5 CV Elution 5 CV Strip (2M NaCl) 5 CV CIP (0.5M NaOH) 4 CV Strip (2M NaCl) 2 CV DW 3 CV Storage (20% Ethanol, 150 mM NaCl) 3 CV Elution pooling criteria Start point: UV.sub.280 nm Value ≥100 mAU End point: UV.sub.280 nm Value ≤100 mAU Hold time for Stable at 2 ~ 8° C. for up to 7 days AEXI Eluted pool *According to client's request, finally determined as 300 mM NaCl Summary of AEXI process

Example (4) Glutathione Sepharose 4 Fast Flow (Affinity Chromatography)

[0394] (4-1) Flow Rate Condition Test for Shortening Process Time

[0395] (4-1-1) Purpose of Experiment

[0396] In a conventional process, the linear velocity was maintained at 90 cm/hr, and the linear velocity was increased up to 180 cm/hr to shorten the time required for a unit process. The maintenance of a step yield was confirmed.

[0397] (4-1-2) Experimental Method

[0398] The affinity process was performed as shown in the following table (Table 57). Experiments were carried out by applying flow rates of 90 cm/hr and 180 cm/hr. HU024 in the samples before and after the process was analyzed in the manner described in (2-1), thereby confirming a step yield.

TABLE-US-00058 TABLE 57 Resin Glutathione Sepharose 4 Fast Flow Column LRC10 (ID 1.0 cm, Height 15 cm), Column Volume (11.8 mL) Flow rate 90 cm/hr, 180 cm/hr Sample preparation AEXI Eluted pool (EQ 50 mM NaCl, Elution 200 mM NaCl) Process Equilibration 20 mM Tirs-HCl, 150 mM NaCl, pH 8.0 (2 CV) Sample loading AEXI Eluted pool, directly loading (4.4 mg GST-Tβ4/mL resin) Unbound wash 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 (5 CV) Elution 50 mM Tris-HCl, 10 mM Glutathione (Reduced), pH 8.0 (5 CV) (50 mAU~50 mAU) Acidic buffer 100 mM Sodium acetate, 500 mM NaCl, pH 4.5 (5 CV) Alkaline buffer 100 mM Tris-HCl, 500 mM NaCl, pH 8.0 (5 CV) CIP 6M Guanidine hydrochloride (2 CV) DW DW (2 CV) Equilibration 20 mM Tirs-HCl, 150 mM NaCl, pH 8.0 (3 CV) Conditions for affinity process

[0399] (4-1-3) Result of Experiment

[0400] Chromatograms by condition according to the affinity process are shown in FIGS. 35 and 36.

[0401] Referring to the yields for the elution pool by condition (Table 58), it was confirmed that there was no effect on a step yield even when the flow rate in the affinity process was increased to 180 cm/hr. However, since the load amount was approximately 4.4 mg/mL resin, reconfirmation was required when the load amount increased.

TABLE-US-00059 TABLE 58 Start material Elution pool Sam- Volume Amounts Concentration Volume Amounts Yield ple (mL) (mg) (mg/mL) (mL) (mg) (%)  90 50 0.82 2.91 8 . . . 8 25.5 62% cm/hr 180 50 0.82 2.33 11.2 26.1 64% cm/hr Yield for affinity linear flow rate study

[0402] (4-2) Dynamic Binding Capacity (DBC) Test

[0403] (4-2-1) Purpose of Experiment

[0404] A DBC of GST-Tβ4 with respect to an affinity column was to be confirmed.

[0405] (4-2-2) Experimental Method

[0406] An affinity process was conducted under the condition shown in Table 57 of (4-2), and a LRC10 column was packed with 1 mL of Glutathione Sepharose 4 Fast Flow resin, and a linear flow rate of 76.4 cm/hr (flow rate: 1 mL/min, however, 150 cm/hr during elution) was applied. A loading sample was overloaded by injecting an AEXI eluted pool (GST-Tβ4 41 mg) obtained through an AEXI optimized process, compared with a DBC (≈11 mg GST-tagged protein/mL medium) in a resin manual, and after the affinity process, the process sample was analyzed through the UV quantification method described in (2-1) UV quantification method.

[0407] (4-2-3) Result of Experiment

[0408] The content of an eluted fraction obtained by elution after overloading in 1 mL of resin was measured (Table 59), and the DBC of the glutathione Sepharose 4 fast flow with respect to GST-Tβ4 was confirmed to be 9.3 mg/mL.

TABLE-US-00060 TABLE 59 Condition & Results AEXI eluted pool (GST-Tβ4 41 mg, volume 50 mL) Vol- Resin Items Conc. ume Amounts volume DBC Loading sample (mg/mL) (mL) (mg) (mL) (mg/mL) Eluted pool 1.36 6.5 8.75 0.94 9.31 DBC for Glutathione Sepharose 4 Fast Flow

[0409] (4-3) AEXI Process Scale-Up and Affinity Process Reproducibility Experiments

[0410] (4-3-1) Purpose of Experiment

[0411] An AEXI process scale-up experiment (XK16-≥ XK50) was conducted.

[0412] Reproducibility and scale-up experiments were performed by applying the affinity process conditions (linear flow rate in sample loading: 80 cm/hr, linear flow rate in elution: 150 cm/hr, LRC10-≥ XK50 condition).

[0413] (4-3-2) Experimental Method

[0414] The same conditions specified in (3-7) were applied to the AEXI process, and XK16 (h: 20 cm) or XK50 (h: 20 cm) was applied to each column. The affinity process was conducted in the same manner as shown in the table below (Table 60), and LRC10 (h: 15 cm) or XK50 (h: 15 cm) was applied to each column.

TABLE-US-00061 TABLE 60 Items Process conditions Resin Glutathione Sepharose 4 Fast Flow Flow rate Sample Loading & Unbound Wash 80 cm/hr Elution step: 150 cm/hr Loading sample AEXI Eluted pool (w/o adjustment of pH or conductivity) Process Equilibration 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 (2 CV) Sample loading AEXI eluted pool, directly loading Unbound wash 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 (5 CV) Elution 50 mM Tris-HCl, 10 mM Glutathione, pH 8.0 (5 CV) Acidic buffer 100 mM Sodium Acetate, 500 mM NaCl, pH 4.5 (5 CV) Alkaline buffer 100 mM Tris-HCl, 500 mM NaCl, pH 8.0 (5 CV) CIP* 6M Guanidine hydrochloride (2 CV) DW* DW (2 CV) Storage 20% Ethanol Elution pooling criteria Start point: UV.sub.280 nm Value ≥100 mAU End point: UV.sub.280 nm Value ≤100 mAU *When CIP flows, the linear flow rate was set to be 80 cm/hr, and when DW flows, the same flow rate needs to be applied. (Max pressure of resin: 1 bar) Conditions for affinity chromatography

[0415] (4-3-3) Result of Experiment

[0416] As a result of the AEXI column scale-up, a chromatogram is shown in FIG. 37, and step yields are shown in Table 61.

TABLE-US-00062 TABLE 61 Start material Elution pool Volume Amounts Concentration Volume Amounts Yield Sample (mL) (mg) (mg/mL) (mL) (mg) (%) XK16 50   269 4.02 71   287 107% scale* XK50 450 2,601 4.25 595 2,529  97% scale* *Loaded at a loading amount of 6.7 mg/mLof resin AEXI step yield for scale up study from XK16 to XK50

[0417] Scale-up was confirmed by measuring similar step yields in the two types of columns. The eluted fractions were then applied to affinity columns (FIG. 38 and Table 62).

TABLE-US-00063 TABLE 62 Start material Elution pool Volume Amounts Concentration Volume Amounts Yield Sample (mL) (mg) (mg/mL) (mL) (mg) (%) LRC10 25   101 6.43 9    61 60% scale* XK50 595 2,529 8.32 221 1,842 73% scale* *Loaded at a loading amount of 8.6 mg/mL of resin Affinity step yield for feasibility test at LRC10 and XK50 scale

[0418] The experiments were conducted by setting a loading amount (8.6 mg/mL of resin) of a sample to be 90% of the DBC (9.3 mg/mL of resin) determined in the previous experiment and applying a loading rate of 80 cm/hr and an elution rate of 150 cm/hr, confirming that step yields were 60% and 73%.

[0419] (4-4) Affinity Column Cleaning

[0420] (4-4-1) Background

[0421] In the 1.sup.st batch of a 50 L verification run, an endotoxin content measured after 6 M GHCl-treated CIP was measured was higher than 0.1 EU/mL, so that additional CIP methods were conducted (Table 63).

TABLE-US-00064 TABLE 63 Endotoxin removal in affinity column by various CIP methods Run Endotoxin No. Process condition (EU/mL) CIP 6M Guanidine hydrochloride(Gdn-HCl) (2CV) - >1 EU/mL Run1 contact time 34 min CIP 6M Gdn-HCl(2CV) - contact time 34 min −> 0.58 EU/mL Run2 DW −> 2M NaCl (2CV) CIP 400 mM Potassium phosphate(KPi) (1CV) - 0.27 EU/mL Run3 contact time 17 min CIP 0.1N NaOH (contact time 10 min) −> 400 0.24 EU/mL Run4 mM KPi CIP 0.1N NaOH (contact time 34 min) −> 2M NaCl 0.085 EU/mL Run5 (2CV) −> DW −> 20% EtOH (3CV, contact 69 min)

[0422] A CIP method (CIP Run5) to which 0.1N NaOH and ethanol recommended by GE were applied was applied to a subsequent process (Table 63).

[0423] (4-5) Hold Time Study for Affinity Process Samples

[0424] (4-5-1) Purpose of Experiment

[0425] Affinity process samples were stored for 7 days in cold room (2 to 8° C.) and room temperature (18 to 22° C.) conditions, and then the change in purity was confirmed to set storage conditions and period for the process samples.

[0426] (4-5-2) Experimental Method

[0427] An experimental method was basically the same as shown in Table 58 of (4-2), and as loading samples, process samples except the AEXI hold time study samples in (3-6) were used. The eluted pool was subjected to sterile filtration, divided into microtubes by 500 uL and stored for 7 days in the cold room and room temperature conditions. The sampled samples were stored at −70° C., completely frozen, simultaneously thawed one day after the end of the storage period, followed by SDS-PAGE.

[0428] (4-5-3) Result of Experiment

[0429] The result of SDS-PAGE after the experiments for the storage conditions (room temperature and cold room) is shown in FIG. 39. No significant difference was found in the SDS-PAGE result of the samples stored in refrigeration and room temperatures. Accordingly, it was confirmed that the affinity process samples were stable for 7 days in refrigeration (2 to 8° C.) or room temperature (18 to 22° C.) storage.

[0430] (4-6) Summary of Affinity Process

TABLE-US-00065 TABLE 64 Summary of affinity process Items Process conditions Resin Glutathione Sepharose 4 Fast Flow Dynamic binding Not more than 9.3 mg/mL capacity Equilibration buffer 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 Elution buffer 50 mM Tris-HCl, 10 mM Glutathione (reduced), pH 8.0 Process temperature 15~25° C. Linear flow rate Sample Loading, unbound wash 80 cm/hr Elution 150 cm/hr Loading sample AEXI eluted pool, directly loading condition Process Equilibration 2 CV Sample loading Sample Unbound wash (with 5 CV Equilibration buffer) Elution 5 CV Acidic buffer 5 CV Alkaline buffer 5 CV CIP (6M Guanidine 2 CV hydrochloride)* DW* 2 CV Storage (20% Ethanol) 3 CV Elution pooling Start point: UV.sub.280 nm Value ≥ 100 mAU criteria End point: UV.sub.280 nm Value ≤ 100 mAU Hold time for Stable at 2~8° C. and 18~22° C. for up to 7 days Affinity eluted pool *When CIP flows, the linear flow rate is set to be 80 cm/hr, and when DW flows, the same flow rate is applied. (Max pressure of resin: 1 bar)

Example (5) Thrombin Digestion

[0431] (5-1) Background

[0432] When the thrombin digestion condition (2.6 mg GST-Tβ4) was applied according to the Chinese process, it was confirmed that, when a target protein concentration in an affinity pool is low, enzyme cleavage efficiency was low (when 2 mg/unit of thrombin was added at the concentration of the affinity pool of 0.3 mg/mL, only approximately 50% reacted in SDS-PAGE). Accordingly, in this experiment, 1 mg GST-Tβ4/2 units of thrombin was applied, and the reaction time was fixed at 16 hours, and then the similarity of enzyme cleavage efficiencies in the range of 13 to 27° C. was confirmed to set a reaction temperature range.

TABLE-US-00066 TABLE 65 Conditions for thrombin digestion Items Process conditions Enzyme Chromatopur Bovine Thrombin (MP bio) (Cat # 02199907) Enzyme/Protein Ratio 1 mg GST-Tβ4/2 unit Thrombin Mix Agitation 60 rpm Reaction Time 16 hrs

[0433] (5-2) Preparation of Thrombin Digestion Samples

[0434] To perform a thrombin digestion study, AEXI and affinity processes were conducted, and the process conditions were the same as those specified in process development. The process-completed affinity pool was divided into conical tubes by 40 mL, stored in a cold room (−20° C.), and thawed as needed.

[0435] (5-3) Thrombin Digestion Study by Reaction Temperature

[0436] (5-3-1) Purpose of Experiment

[0437] Under a thrombin digestion reaction condition, a temperature range was to be additionally set.

[0438] (5-3-2) Experimental Method

[0439] The affinity elution pool (209.1 mg GST-Tβ4) prepared in (5-2) was used, and Thrombin was dissolved in 20 mM Tris-HCl (pH 8.0) of 15% (v/v) of the volume of the affinity pool to be reacted, and then applied to treat the elution pool. Afterward, reactions were performed with the method specified in Table 65, but the reactions were simultaneously conducted at 15±2° C. (cold room), 20±2° C. (room temperature) and 25±2° C. (incubator). The reaction-completed samples were subjected to SEC-HPLC specified in (2-2) to quantify Gly-Tβ4, thereby confirming yields (2016-037, p 13, 18).

[0440] (5-3-3) Result of Experiment

[0441] The SEC-HPLC results for the samples that had been subjected to thrombin digestion by reaction temperature are shown in FIG. 40. When the chromatogram and yields were examined (Table 66), no significant difference according to temperature was confirmed. Therefore, it was confirmed that the thrombin reaction temperature condition ranging from 13 to 27° C. does not affect a yield and purity.

TABLE-US-00067 TABLE 66 Start material Elution pool Volume Amounts Conc. Volume Amounts Yield Sample (mL) (mg) (mg/mL) (mL) (mg) (%) 15 ± 2° C. 34.5 209 1.07 34.5 36.9 17.7% 20 ± 2° C. 34.5 209 1.09 34.5 37.5 17.9% 25 ± 2° C. 34.5 209 1.07 34.5 37.1 17.7% Step yield for cleavage reaction mixtures

Example (6) 4.SUP.th .Process Study

[0442] (6-1) Background

[0443] According to the result of the 1.sup.st 50 L verification run, the loss of a target protein due to an F/T fraction in the AEXII process was observed, and it was confirmed that the purities of RP-HPLC and SEC-HPLC of an elution pool are lower than the F/T Fraction. Based on the result of the experiment, the condition of the 2.sup.nd verification run was changed to an F/T mode, so that a purity and a yield were improved. However, in SE-HPLC of the final purified extract, it was necessary to further improve purity to be 97.0%, and in non-optimized RP-HPLC, the purity was confirmed at 98.7%, and therefore, 4.sup.th column introduction was considered.

[0444] (6-2) Sample Production for Developing 4.sup.th Column Process

[0445] To ensure a sample required for a 4.sup.th column process experiment, a sample was prepared by performing an AEX II process in an F/T mode, which is the 2.sup.nd verification run condition. For sample production, a cell lysate was prepared for a depth sizing test of two 5 L DoE batches (Batch No. 161019-B1, B2). After the sizing test, 7099 mg of Gly-Tβ4 was secured by performing AEXI, affinity, thrombin digestion and AEXII processes using the remaining cell lysate (approximately 1200 mL). While the AEXI and affinity processes were conducted in the same manner as described in (3-7) and (4-1), a column scale was applied to the AEXI (XK50, h: 20 cm) and affinity (XK50, h: 15 cm) processes, and the AEXII process was conducted by the method specified in Table 67. The obtained sample was subjected to microfiltration (0.2 μm), divided by 100 mL, and stored at −20° C., followed by the development of the 4th column process using this.

TABLE-US-00068 TABLE 67 AEXII process conditions Items Process conditions Resin Fractogel EMD TMAE(M) Column XK50 (ID 5.0 cm, Height 20 cm), Column Volume (392.5 mL) Flow rate 200 cm/hr (65.41 mL/min) Sample preparation After Thrombin dioestion, Adjust pH conductivity (pH 8.0, conductivity 6 mS/cm) with DW Process Equilibration 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 (2 CV) Sample Thrombin digestion sample loading Unbound 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 (5 CV) wash Strip 2M NaCl (3 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 (3 CV) Elution criteria Start point: UV.sub.(214 nm) Value ≥ 50 mAU (Flow through pooling) End point: UV.sub.(214 mn) Value ≤ 50 mAU

[0446] The step yields and purities of the secured 3.sup.rd column eluted fractions were analyzed (Tables 68 and 69). While the SE-HPLC purity was improved up to 98.9%, the RP-HPLC purity was 97.6%. However, when analyzed by the optimized RP-HPLC analysis method, purity fell to 92.2%, and thus it was intended to improve purity 99.0% or more.

TABLE-US-00069 TABLE 68 Start material Elution pool Vol- Conc. Vol- ume Amounts (mg/ ume Amounts Yield HCP Sample (mL) (mg) mL) (mL) (mg) (%) (ppm) Cell 5.07 1,212 6,145 — — Lysate AEXI 1,212 6,145 3.45 1,927 6,648 108% 1,916 Affinity 1,835 6,329 7.79   625 4,871  77%   137 Thrombin   625 4,871 1.09   707   771 16%   821 Reaction AEXII  1069   770 0.33  2125   710  92%    17 Step yield of AEX II for preparing load sample on 4.sup.th column

TABLE-US-00070 TABLE 69 SEC-HPLC RPC-HPLC (ver. 1.0) Sample Test 1 Test 2 Average Test 1 Test 2 Average UF/DF 98.9 99.0 98.9 97.4 97.8 97.6 Purity analysis after UF/DF for preparing load sample on 4.sup.th column

[0447] (6-3) Resin Scouting

[0448] (6-3-1) Cation Exchange Chromatogram (CEX) Process

[0449] To improve purity, SP FF, which is a basic CEX resin, was applied (Table 70).

TABLE-US-00071 TABLE 70 CEX process conditions for 4.sup.th column Items Process conditions Resin SP Sepharose Fast Flow Column LRC10 (ID 1.0 cm, Height 16.5 cm), Column volume (13.0 mL) Flow rate 200 cm/hr (2.63 mL/min) Sample preparation AEXII adjusted pH and conductivity (pH 4.5, 5.0, conductivity ≤ 2.8 mS/cm) Process Equilibration 20 mM Sodium Acetate, pH 4.5, 5.0 (2 CV) Sample AEXII sample adjusted pH and conductivity loading (11.8 mg Gly-Tβ4) Unbound 20 mM Sodium Acetate, pH 4.5, 5.0 (5 CV) wash Elution 20 mM Sodium Acetate, 1.0M NaCl, pH 4.5, 5.0 (10 CV, 0~100% B Linear gradient) Strip 2M NaCl (3 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 20 mM Sodium Acetate, pH 4.5, 5.0 (3 CV)

[0450] The chromatogram for the CEX process is shown in FIG. 41. Two separated peaks were shown, and step yields of the processes proceeded with pH 4.5 and 5.0 were confirmed (Table 71).

TABLE-US-00072 TABLE 71 Elution pool Start material Conc. Volume Amounts (mg/ Volume Amounts Yield Sample (mL) (mg) mL) (mL) (mg) (%) pH 4.5 130 11.8 1.15 10 11.5 97.5% pH 5.0 62.5 8.8 0.24 25 6.1 69.3% Yield for CEX process by pH

[0451] In the case of pH 5.0, no separated peak was found, and the yield was confirmed to be lower than that in the case of the pH 4.5 condition. When the F/T fraction was analyzed, a target protein was identified, and thus it was expected that the above result was caused by the target protein. Accordingly, only the pH 4.5 condition was subjected to RP-HPLC analysis, and the analysis result is shown in FIG. 42.

[0452] It was confirmed that many impurity peaks before and after the main peak were removed, and the RP-HPLC purities of a loading sample and an elution sample were confirmed to be 92.2% and 97.6%, respectively. However, the impurity peak present before the main peak was not removed, and referring to a UF/DF result (Table 69), there was no significant difference from the RP-HPLC purity (97.6%), so that a hydrophobic resin, instead of a CEX-type resin, was to be confirmed to ensure a purity of 99% or more.

(6-3-2) Hydrophobic Interaction Chromatography (HIC) Process

[0453] Considering the separation between the main and impurity peaks and a target peak in RP-HPLC, hydrophobic columns were tested, and process conditions are shown in Table 72. Two resins (butyl and phenyl) were tested according to the degree of hydrophobicity, and as salt conditions, both NaCl and Na.sub.2SO.sub.4 were used.

TABLE-US-00073 TABLE 72 HIC process conditions for 4.sup.th column Items Process conditions Resin Butyl Sepharose 4 Fast Flow Phenyl Sepharose 6 Fast Flow (High sub) Column LRC10 (ID 1.0 cm, Height 15.0 cm), Column volume (11.8 ml) Flow rate Sample Loading 80 cm/hr (1.04 mL/min) Elution 150 cm/hr (1.96 mL/min) Sample preparation AEXII adjusted pH and conductivity with 4M NaCl and 2M Na.sub.2SO.sub.4 (pH 8.0, Conductivity ≥ 90 mS/cm) Process Equili- 20 mM Tris-HCl, 2M NaCl, 1.0M Na.sub.2SO.sub.4, bration pH 8.0 (2 CV) Sample AEXII sample adjusted pH and conductivity loading (16.2 mg Gly-Tβ4) Unbound 20 mM Tris-HCl, 2M NaCl, 1.0M Na.sub.2SO.sub.4, wash pH 8.0 (5 CV) Elution 20 mM Tris-HCl, pH 8.0 (10 CV, 0~100% B Linear gradient) Strip DW (3 CV) CIP 0.5M NaOH (3 CV) Strip 100 mM Sodium Acetate, pH 4.5 (2 CV) Equili- 20 mM Tris-HCl, 2M NaCl, 1.0M Na.sub.2SO.sub.4, bration pH 8.0 (3 CV)

[0454] The process chromatogram is shown in FIG. 43. A high peak was formed in the front of the chromatogram, and there was no peak observed in the elution step.

[0455] As a result, the target protein did not bind to both of the resins, and thus there was no additional progression.

[0456] (6-3-3) Mixed-Mode Chromatography (MMC) Process

[0457] As a result of the CEX process and the HIC resin study, the removal of main impurities (17.5 min peak when RP-HPLC analysis method ver. 1.0 was applied) was not confirmed. Therefore, Capto MMC was applied.

[0458] (6-3-3-1) Experimental Method

[0459] In the case of the MMC resin, for the binding condition, a process was conducted at low pH to induce a CEX-mode operation, and usually, an elution method is used by changing the pH to neutral and giving pH and salt gradients (Reference Document 6.2). According to the experiment, the pH of an EQ buffer was changed to 4.5 to 5.0, and the pH of an elution buffer was changed to 4.5 to 7.5, confirming that, as the elution pH increased, a peak area became smaller, and the resolution of a main impurity and a target protein decreased. Accordingly, it was determined that the Capto MMC condition is suitable to be performed at low pH. Therefore, in the low pH (EQ, elution buffer, pH 4.5) condition, the process was conducted under the conditions in Table 73.

TABLE-US-00074 TABLE 73 MMC process conditions for 4.sup.th column Items Process conditions Resin Capto MMC Column LRC10 (ID 1.0 cm, Height 15.0 cm), Column volume (11.8 mL) Flow rate 200 cm/hr (2.61 mL/min) Sample preparation Dilution with Equilibration by equivalent volume of AEXII sample Process Equilibration 50 mM Sodium Acetate, pH 4.5 (2 CV) Sample AEXII sample adjusted pH and conductivity loading (6.1 mg Gly-Tβ4) Unbound 50 mM Sodium Acetate, pH 4.5 (5 CV) wash Elution 50 mM Sodium Acetate, 1M NaCl, pH 4.5 (15 CV, 0~100% B Linear gradient) Strip 2M NaCl (3 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 50 mM Sodium Acetate, pH 4.5 (3 CV)

[0460] The fractionation of an elution pool was performed to carry out RP-HPLC analysis, and purities and yields were confirmed by collecting the pools having a purity of 98% or more.

[0461] (6-3-3-2) Result of Experiment

[0462] The chromatograms for Capto MMC are shown in FIG. 37. In the elution step, the formation of a single peak was confirmed, and the RP-HPLC (ver 1.0) result for each fraction is shown in FIG. 44. As a result, in the fractions in the front of the peak, the main impurity peak was highly confirmed, whereas almost no main impurity peaks were found in the fractions in the middle to back of the peak.

[0463] Referring to the RP-HPLC purity analysis result by eluted fraction, it can be confirmed that the main impurity peaks are concentrated in the front fractions. In pooling of the B8, B7 and B6 fractions, 99.6% purity was able to be ensured at 70% step yield (FIG. 74).

TABLE-US-00075 TABLE 74 Yield for Capto MMC process Conc. Volume Amounts Purity Yield Sample (mg/mL) (mL) (mg) (%) (%) Capto MMC Loading 0.15 40 6.1 — — Sample B10 Fraction 0.05 5 0.27 84.3 — B9 Fraction 0.14 5 0.69 92.6 — B8 Fraction 0.35 5 1.74 98.7 — B7 Fraction 0.35 5 1.74 99.8 — B6 Fraction 0.16 5 0.77 99.4 — B8~B6 Fraction 0.29 15 4.3 99.6 70.5 pooling

[0464] (6-4) 4.sup.th Process Optimization Study 1

[0465] (6-4-1) Purpose of Experiment

[0466] A step elution condition optimization experiment for improving a step yield was conducted.

[0467] (6-4-2) Experimental Method

[0468] Target elution conditions by step were confirmed, and it can be confirmed that the eluted condition is elution with 200 mM NaCl or more. Accordingly, a washing step condition was to be confirmed by applying 185 mM NaCl, which is lower than the salt concentration at which a target can be eluted, and the basic purification conditions are shown in Table 73 of (6-3-3-1), but the NaCl concentration of the elution buffer was changed to 400 mM. In addition, process reproducibility when loading amounts were applied under the confirmed conditions was to be confirmed.

[0469] (6-4-3) Result of Experiment

[0470] The chromatogram obtained by applying 185 mM NaCl (conductivity 20.7 to 21.2 mS/cm) in the washing step is shown in FIG. 46. After loading at 0.5 mg/mL of resin, the elution step fractions (C6-C9) were pooled into two types to perform RP-HPLC. As a result, it was confirmed that the purities are 98.6% (C6-C9 pool) and 99.0% (C7-C9 pool), which are 98% or more, and the yield is 78.7% in the C6-C9 fraction pool (Table 75).

TABLE-US-00076 TABLE 75 Step yield of MMC step elution with load amounts of 0.5 mg/mL of resin) Conc. Volume Amounts Purity Yield Sample (mg/mL) (mL) (mg) (%) (%) Capto MMC 0.15 40 6.1 — — Loading sample C6~C9 Fraction 0.24 20 4.8 98.6 78.6 pool C7~C9 Fraction 0.27 15 4.1 99.0 67.3 pool

TABLE-US-00077 TABLE 76 Step yield of MMC step elution with load amounts of 4.0 mg/mL of resin Conc. Volume Amounts Purity Yield Sample (mg/mL) (mL) (mg) (%) (%) Capto MMC 0.3 159 47.8 — — Loading sample Wash 0.35 80.0 28.3 99.6 59.2 Elution 0.43 35 15.3 99.7 32.0

[0471] However, when 47.8 mg of Gly-Tβ4 was loaded (4.0 mg/mL of resin), most of the target protein flowed out in the washing step, and the yield was confirmed to be 32% (Table 76), so that it was impossible to be applied to the process.

[0472] (6-5) 4.sup.th Process Optimization Study 2

[0473] (6-5-1) Purpose of Experiment

[0474] Under the condition of a loading amount of 3 mg Gly-Tβ4/mL of resin, which is higher than 0.5 mg (conventional amount) and lower than 4.0 mg, purities and step yields according to the NaCl concentration conditions (100, 120 and 140 mM) of the washing step were confirmed.

[0475] (6-5-2) Experimental Method

[0476] An experimental method is the same as specified in Table 77. After the process, samples were analyzed by RP-HPLC (ver 2.0), and according to the analysis result, the purities and step yields of the process samples were confirmed.

TABLE-US-00078 TABLE 77 MMC process conditions with loading of 3 mg/mL resin Items Process conditions Resin Capto MMC Column LRC10 (ID 1.0 cm, Height 15.0 cm), Column volume (11.8 mL) Flow rate 200 cm/hr (2.61 mL/min) Sample preparation Adjust pH with 1M Acetic acid (do not dilution) Process Equilibration 20 mM Sodium Acetate, pH 4.5 (2 CV) Sample AEXII sample adjusted pH (34 mg Gly-Tβ4) loading Unbound 20 mM Sodium Acetate, pH 4.5 (5 CV) wash Washing 20 mM Sodium Acetate, 100 mM, 120 mM 140 mM NaCl, pH 4.5 (20 CV) Elution 50 mM Sodium Acetate, 500 mM NaCl, pH 4.5 (5 CV) Strip 2M NaCl (3 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 20 mM Sodium Acetate, pH 4.5 (3 CV)

[0477] (6-5-3) Result of Experiment

[0478] The chromatograms according to NaCl concentrations in a washing step are shown in FIGS. 47, 48 and 49. In the elution step, it was confirmed that the peak is split into two. It was assumed that elution is delayed by the decrease in pH due to a salt in an elution buffer. The two split elution peaks were fractioned and analyzed, followed by pooling and then reanalysis (Table 78).

TABLE-US-00079 TABLE 78 Conc. Volume Amounts Sample (mg/mL) (mL) (mg) Purity (%) Yield (%) 100 mM NaCl LS 0.3 115 34.6 — Wash — — — — Elution 1.32 25 32.9 93.7 95.0 pool 120 mM NaCl LS 0.3 116 34.7 — Wash 0.05 186  9.7 — 27.9 Elution 0.85 30 25.6 99.3 73.7 pool 140 mM NaCl LS 0.3 113 33.9 — — Wash 0.09 165 15.3 — 45.0 Elution 0.54 30 16.1 99.4 47.4 pool Purity and Step yield of MMC with increasing NaCl for washing

[0479] When washing was conducted with 100 mM NaCl, it was confirmed that the yield was high at 95%, but the mean purity in an elution pool was approximately 93.7%, which was insufficient for removing impurities. When washing was conducted with 140 mM NaCl, it was confirmed that the purity was 99% or more, but the yield was very low at 47%. Accordingly, the NaCl concentration in the washing step, which is the most suitable in terms of a purity and an yield, was confirmed to be 120 mM, and here, it was confirmed that the yield was 74% and the purity was 99% or more.

[0480] (6-6) 4.sup.th Process Dynamic Binding Capacity (DBC) Test

[0481] (6-6-1) Purpose of Experiment

[0482] A DBC of MMC with respect to Gly-Tβ4 was to be confirmed.

[0483] (6-6-2) Experimental Method

[0484] An experimental method was conducted as specified in Table 79, and all binding proteins without a washing step were eluted at one time, and quantified by RP-HPLC.

TABLE-US-00080 TABLE 79 Capto MMC process conditions for DBC test Items Process conditions Resin Capto MMC Column LRC10 (ID 1.0 cm, Height 1.2 cm), Column volume (1 mL) Flow rate 76.4 cm/hr (1.0 mL/min) Sample preparation Adjust pH with Capto MMC Equilibration buffer Process Equilibration 20 mM Sodium Acetate, pH 4.5 (10 CV) Sample AEXH sample adjusted pH (26.4 mg Gly-Tβ4) loading Unbound 20 mM Sodium Acetate, pH 4.5 (50 CV) wash Elution 50 mM Sodium Acetate, 500 mM NaCl, pH 4.5 (20 CV) Strip 2M NaCl (10 CV) CIP 0.5M NaOH (10 CV) Strip 2M NaCl (10 CV) Equilibration 20 mM Sodium Acetate, pH 4.5 (10 CV)

[0485] (6-6-3) Result of Experiment

[0486] The RP-HPLC result for the MMC process samples obtained by overloading in 1 mL of resin is shown in Table 80, and the DBC of Capto MMC with respect to Gly-Tβ4 was confirmed to be 9.6 mg/mL. This result has a difference from the previous experiment result, and that is, in the 4.0 mg/mL of resin condition, most of the protein was present in wash fractions. Accordingly, a condition difference between the two experiments (flow rate 200 vs 76 cm/hr) was suspected.

TABLE-US-00081 TABLE 80 DBC for Capto MMC process Condition & Results AEXII Eluted pool (Gly-Tβ4 26.4 mg, volume 160 mL) Items Resin Loading Conc. Volume Amounts volume DBC sample (mg/mL) (mL) (mg) (mL) (mg/mL) Eluted pool 0.64 15 9.6 1.0 9.6

[0487] (6-7) Summary of Capto MMC Process

TABLE-US-00082 TABLE 81 Summary for Capto MMC process Items Process conditions Resin Capto MMC Dynamic binding 9.6 mg/mL (Max DBC) capacity Equilibration buffer 20 mM Sodium Acetate, pH 4.5 Wash buffer 20 mM Sodium Acetate, 120 mM NaCl, pH 4.5 Elution buffer 20 mM Sodium Acetate, 500 mM NaCl, pH 4.5 Process temperature 15~25° C. Linear flow rate* 76 or 200 cm/hr (RT 6 min) Loading sample pH 4.5 ± 0.1 (adjustment with Acetic condition acid, but do not dilution) Process Equilibration 2CV Sample Loading Sample Unbound Wash (with 5CV Equilibration buffer) Wash 20CV Elution 5CV Strip (2M NaCl) 5CV CIP (0.5M NaOH) 4CV Strip (2M NaCl) 2CV DW 3CV Storage (20% Ethanol) 3CV Elution pooling Start point: from elution step criteria End point: UV.sub.214 nm Value ≤ 100 mAU *to be more studied later

Example (7) Process Study

[0488] (7-1) Background

[0489] The conductivity condition of a load sample and the NaCl concentration in an elution buffer in an AEXI process were changed, and the AEXII process was changed to a bind-elute mode, and a process using the Nuvia HR-S resin as a 4.sup.th column was determined as the final process. In addition, the purity criterion of the final purified extract was confirmed to be 97% or more (in RP-HPLC). Accordingly, a task of reproducing the process in small-scale experiments and scaling up the process was conducted.

TABLE-US-00083 TABLE 82 Conditions for comparison DSP Process conditions Items Binex Huons AEXI Resin Fractogel EMD TMAE(M) Equilibration buffer 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 20 mM Tris-HCl, pH 8.0 Elution buffer* 20 mM Tris-HCl, 200 mM NaCl, pH 20 mM Tris-HCl, 300 mM NaCl, pH 8.0 8.0 Linear flow rate 200 cm/hr (RT 6 min) Loading sample pH 8.0 ± 0.1 pH 8.0 ± 0.1 condition* Conductivity max 8.0 mS/cm Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: UV.sub.280 nm Value ≥ 100 mAU criteria End point: UV.sub.280 nm Value ≤ 100 mAU Affinity Same as above conditions Thrombin Same as above conditions Reaction AEXII Resin Fractogei EMD TMAE(M) Mode* F/T B/E Equilibration Buffer 20 mM Tris-HCl, 50 mM NaCl, pH 8.0 20 mM Tris-HCl, pH 8.0 Elution Buffer* N/A 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 Linear Flow Rate 200 cm/hr (RT 6 min) Loading Sample pH 8.0 ± 0.1 pH 8.0 ± 0.1 Condition* Conductivity < 6.5 mS/cm conductivity < 2 mS/cm <10 Dilution with DW> Elution pooling Start point: UV.sub.214 nm Value ≥ 50 mAU N/A criteria End point: UV.sub.214 nm Value ≤ 50 mAU 4th Resin* Capto MMC Nuvia HR-S Column Mode B/E B/E Equilibration Buffer 20 mM Acetate, pH 4.5 Wash 20 mM Acetate, 120 mM NaCl, pH 20 mM Acetate, 100 mM NaCl, pH 4.5 4.5 Elution Buffer 20 mM Acetate, 500 mM NaCl, pH 20 mM Acetate, 180 mM NaCl, pH 4.5 4.5 Linear Flow rate To be more studied(76 or 200 cm/hr) 200 cm/hr (R.T. 6 min) Loading Sample pH 4.5 ± 0.1 pH 4.5 ± 0.1 Condition* (adjust only pH) Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: from elution step N/A criteria End point: UV.sub.214 nm Value ≤ 100 mAU

[0490] (7-2) Sample Preparation for 3.sup.rd and 4.sup.th Column Process Studies

[0491] As a sample used in this experiment, the affinity pool specified in (5-2) was used (GST-Tβ4 concentration 6.97 mg/mL), and prepared by initiating thrombin digestion in the manner specified in Table 65 one day before the 3.sup.rd process.

[0492] (7-3) 3.sup.rd Column Process Study (AEXII-B/E Mode)

[0493] (7-3-1) Confirmation of 3.sup.rd Column Process Dynamic Binding Capacity (DBC)

[0494] (7-3-1-1) Purpose of Experiment

[0495] A DBC for AEXII was to be confirmed. However, the experiment was performed by changing a flow rate from 200 cm/hr (conventional) to 153 cm/hr.

[0496] (7-3-1-2) Experimental method

[0497] An experimental method was conducted as in Table 83, and after overloading a sample (5 mg/mL of resin) in a column, an eluted fraction was quantified by SEC-HPLC to confirm the amount of eluted Gly-Tβ4.

TABLE-US-00084 TABLE 83 AEXII process conditions for DBC test Items Process conditions Resin Fractogel EMD TMAE(M) Column LRC10 (ID 1.0 cm, Height 2.5 cm), Column volume (2 mL) Flow rate 153 cm/hr (2 mL/min) Sample preparation After thrombin digestion, Adjust pH and conductivity (pH 8.0, conductivity < 2 mS/cm - with DW) Process Equilibration 20 mM Tris-HCl, pH 8.0 (10 CV) Sample Thrombin digestion sample adjusted pH and loading conductivity <Loading amounts 10.5 mg Gly-Tβ4> Unbound 20 mM Tris-HCl, pH 8.0 (50 CV) wash Elution 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 (20 CV) Strip 2M NaCl (10 CV) CIP 0.5M NaOH (10 CV) Strip 2M NaCl (10 CV) Equilibration 20 mM Tris-HCl, pH 8.0 (10 CV)

[0498] (7-3-1-3) Result of Experiment

[0499] The SEC-HPLC result for the elution pool obtained by being overloaded in 2 mL of resin is shown in FIG. 51. When the elution pool was overloaded in AEXII, GST was also found, and when AEXII had the following conditions, the Gly-Tβ4 DBC attached to AEXII was confirmed to be 1.7 mg Gly-Tβ4/mL (Table 84). However, when the flow rate was changed from 153 cm/hr to 200 cm/hr, additional experiments were required.

TABLE-US-00085 TABLE 84 DBC for AEXII process Condition & Results Items Conc. Volume Amounts Resin volume DBC Loading sample (mg/mL) (mL) (mg) (mL) (mg/mL) Thrombin digestion Sample (Gly-Tβ4 10.5 mg, volume 100 mL) Eluted pool 0.048 70 3.4 2 1.7 Thrombin digestion Sample (Gly-Tβ4 8.96 mg, volume 85 mL) Eluted pool 0.049 70 3.4 2 1.7

[0500] (7-3-2) 3.sup.rd Column Process Reproducibility Experiment

[0501] (7-3-2-1) Purpose of Experiment

[0502] Reproducibility (step yield: within 80%, purity: RP-HPLC 90 to 96% level) was to be confirmed by applying only approximately 90% to the set DBC and increasing a flow rate to 200 cm/hr.

[0503] (7-3-2-2) Experimental Method

[0504] An experimental method for the 3.sup.rd process was conducted as shown in Table 85 below, and after the process was completed, quantification was performed through SEC-HPLC, followed by purity analysis through RP-HPLC.

TABLE-US-00086 TABLE 85 AEXII process conditions Items Process conditions Resin Fractogel EMD TMAE(M) Column XK16 (ID 1.6 cm, Height 20 cm), Column volume (40 mL) Flow rate 200 cm/hr (6.67 mL/min) Sample preparation After thrombin digestion, Adjust pH and conductivity (pH 8.0, conductivity < 1 mS/cm - with DW) Process Equilibration 20 mM Tris-HCl, pH 8.0 (2 CV) Sample AEXII sample adjusted pH and conductivity loading Unbound 20 mM Tris-HCl, pH 8.0 (5 CV) wash Elution 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 (5 CV) Strip 2M NaCl (4 CV) CIP 0.5M NaOH (3 CV) Strip 2M NaCl (2 CV) Equilibration 20 mM Tris-HCl, pH 8.0 (3 CV) Elution criteria Start point: UV.sub.(214 nm) ≥ 50 mAU End point: UV.sub.(214 nm) ≤ 50 mAU

[0505] (7-3-2-3) Result of Experiment

[0506] The chromatograms obtained by loading 1.0 mg and 1.5 mg Gly-Tβ4 per mL of resin are shown in FIGS. 52 and 53, respectively.

[0507] Before elution, a high UV peak was found, and as a result of analyzing the peak, it was confirmed that there was no target protein, and the peak was caused by impurities (GSH). The yields and purities for the processes were confirmed as shown in Table 86.

TABLE-US-00087 TABLE 86 Conc. Volume Amounts Sample (mg/mL) (mL) (mg) Purity (%) Yield (%) 1 mg Gly-Tβ4/ Loading 0.11 395 41.7 — — Resin mL sample Elution 0.29 113 33.0 93.6 79.1 1.5 mg Gly-Tβ4/ Loading 0.11 579 61.4 — Resin mL sample Elution 0.34 146 49.1 92.2 80.0 Quality and quantity profile for AEXII process

[0508] No significant differences in yield and purity with respect to loading amounts were observed. Therefore, reproducibility for the DBC and conditions for AEXII was able to be confirmed.

[0509] (7-3-3) Experiment for Reducing Volume of 3.sup.rd Process Loading Sample

[0510] (7-3-3-1) Purpose of Experiment

[0511] It was intended to confirm the effect of 4-fold dilution for reducing the volume of a loading sample that increased when the conventional 10-fold dilution factor was applied according to a change to a B/E mode.

[0512] (7-3-3-2) Experimental Method

[0513] An experimental method was conducted in the same manner as specified in Table 85, and a loading sample was diluted 4-fold with DW to maintain a conductivity at 2 to 3 mS/cm. The sample after the process was quantified by SEC-HPLC, and a purity was analyzed by RP-HPLC.

[0514] (7-3-3-3) Result of Experiment

[0515] The chromatogram obtained by performing the process with a change in conductivity of a loading sample is shown in FIG. 55.

TABLE-US-00088 TABLE 87 Conc. Volume Amounts Sample (mg/mL) (mL) (mg) Purity (%) Yield (%) 1 mg Gly-Tβ4/ Loading 0.27  94 24.9 — — Resin mL sample Elution 0.26  76 19.5 93.1 78.5 1.5 mg Gly-Tβ34/ Loading 0.28 234 65.4 — Resin mL sample Elution 0.40 125 49.8 90.6 76.1 Purity and yield of AEXII with increased conductivity of loading sample

[0516] Due to the decrease in dilution factor of a loading sample, the conductivity was increased (1 mg Gly-Tβ4/resin mL Loading sample conductivity 1.16-≥2.76 mS/cm, 1.5 mg Gly-Tβ4/resin mL Loading sample conductivity 1.13-≥2.80 mS/cm), but there was no significant difference in yield and purity (Table 87).

[0517] (7-3-4) 3.sup.rd Process Hold Time Study

[0518] (7-3-4-1) Purpose of Experiment

[0519] To ensure the storage conditions and period that can be applied in GMP production, in an AEXII process, eluted fraction samples were stored for 7 days in cold room (2 to 8° C.) and room temperature (18 to 22° C.) conditions, and then the change in purity was confirmed.

[0520] (7-3-4-2) Experimental Method

[0521] In an experimental method, in a 5 L scale process, a 3.sup.rd process eluted pool was subjected to sterile filtration, divided into microtubes by 300 uL, and stored for 7 days in cold room and room temperature conditions. Here, at the end of the storage period, samples were stored at −70° C. to completely freeze, and one day after the end of the storage period, all of the samples were simultaneously thawed, followed by RP-HPLC.

[0522] (7-3-4-3) Result of Experiment

[0523] The chromatogram obtained by the test in each storage condition (room temperature or cold room) and RP-HPLC is shown in FIG. 56, and the table exhibiting a purity and an area for the main peak is shown in Table 88.

[0524] In terms of the stability criteria of a process sample, % CV of the area of the main peak with the lapse of time is confirmed to be 2.0 or less, compared with the system suitability criteria (% CV≤2.0%) according to the RP-HPLC of (2-3), determining that there is no change. Therefore, it was confirmed that the 3.sup.rd process samples are stable for 7 days in the cold room (2 to 8° C.) or room temperature (18 to 22° C.) storage condition.

TABLE-US-00089 TABLE 88 RP-HPLC analysis for AEXII hold time study Hold time at 2~8° C. Hold time at 18~22° C. Sample Main peak Area Purity (%) Main peak Area Purity (%) 0 day 8,623 96.3 8,861 96.3 1 day 9,014 95.9 9,077 95.8 2 day 8,846 95.9 9,047 96.4 4 day 8,930 96.1 8,992 96.0 7 day 9,117 96.0 9,048 95.5 % CV 1.9 0.2 0.9 0.3

[0525] (7-3-5) AEXII Process Summary

TABLE-US-00090 TABLE 89 Summary of AEXII process Items Process conditions Resin Fractogel EMD TMAE(M) Dynamic binding Not more than 1.8 mg/mL capacity Equilibration buffer 20 mM Tris-HCl, , pH 8.0 Elution buffer 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 Process temperature 15~25° C. Linear flow rate 200 cm/hr Loading sample Thrombin digestion sample adjust pH and condition conductivity (pH 8.0, conductivity, 2~3 mS/cm with DW) Process Equilibration 2 CV Sample loading Sample Unbound wash (with 5 CV Equilibration buffer) Elution 5 CV Strip (2M NaCl) 4 CV CIP (0.5M NaOH) 3 CV Strip (2M NaCl) 2 CV Equilibration 3 CV DW 2 CV Storage (20% Ethanol, 3 CV 150 mM NaCl) Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU Hold time for Stable at 2~8° C and 18~22° C. for up to 7 days AEXII eluted pool

[0526] (7-4) 4.sup.th Column Process Study

[0527] (7-4-1) Experiment of Confirming Condition Reproducibility

[0528] Loading samples to be used in a 4.sup.th column reproducibility experiment were prepared in a B/E mode by applying 2.sup.nd 50 L verification run criteria (50 mAU˜50 mAU) as elution criteria for AEXII, and other than that, all 4.sup.th column conditions are described in Table 90. The experiment was repeated twice (FIGS. 57 and 58).

TABLE-US-00091 TABLE 90 CEX process conditions Items Process conditions Resin Nuvia HR-S Column XK16 (ID 1.6 cm, Height 20 cm), Column volume (40 mL) Flow rate 200 cm/hr (6.67 mL/min) Sample preparation AEXII elution pool adjust pH and conductivity (pH 4.5, conductivity, 1~2 mS/cm - with DW) Process Equili- 20 mM Sodium Acetate anhydrous. pH 4.5 (2 CV) bration Sample AEXII eluted pool adjusted pH and loading conductivity <Loading amounts 45 mg Gly-Tβ4> Unbound 20 mM Sodium Acetate anhydrous. pH 4.5 (5 CV) wash Wash 20 mM Sodium Acetate anhydrous. 80 mM NaCl, pH 4.5 (10 CV) Elution 20 mM Sodium Acetate anhydrous, 180 mM NaCl, pH 8.0 (10 CV) Strip 2M NaCl (4 CV) CIP 0.5M NaOH (4 CV) Strip 2M NaCl (2 CV) Equili- 20 mM Sodium Acetate anhydrous. pH 4.5 (3 CV) bration

TABLE-US-00092 TABLE 91 Load amount Con- Conc. Volume (mg/mL ductivity Sample (mg/mL) (mL) of resin) (mS/cm) pH CEX Loading 0.067 680 1.1 1.74 4.53 Run1 sample (17.0 CV) CEX Loading 0.069 580 1.0 1.68 4.50 Run2 Sample (14.5 CV) Loading Sample preparation for CEX feasibility test

[0529] As a result of the experiment, a main peak elution position was slightly different from the conventional result (FIGS. 57 and 58). The difference in experiment conditions was examined, and it was confirmed that the conditions were the same. except a resin bed height, and the protein concentration and volume of a loading sample. In the conventional example, the Gly-Tβ4 concentration in the AEXII elution pool was 1 to 1.5 mg/mL, but in the in-house experiment performed with the elution criteria of 50 mAU˜50 mAU, the concentration of Gly-Tβ4 in the elution pool was confirmed as 0.4 mg/mL. Accordingly, a loading sample volume increased: for example, the conventional condition was 7.5 CV, and the in-house experiment repeated twice was conducted with 14.5 and 17.0 CV of the samples, respectively. To identify the exact cause, first, the effect of the difference in bed height was to be examined.

[0530] (7-4-2) Effect Caused by Bed Height Difference

[0531] (7-4-2-1) Purpose of Experiment

[0532] It was to be confirmed whether there is a difference in elution position according to the difference in bed height.

[0533] (7-4-2-2) Loading Sample Preparation

[0534] To prepare a sample as similar as possible to the conventionally used loading sample, an elution fraction was obtained by changing the AEXII elution criteria to 100 mAU˜172 mAU. The purity and step yield of the elution fraction prepared as described above were confirmed (Table 92). However, the protein concentration was 0.9 mg/mL, which was slightly lower than the conventional case (1 to 1.5 mg/mL).

TABLE-US-00093 TABLE 92 Purity Yield (%) (%) Conc. Volume Amounts RP- SEC- Sample (mg/mL) (mL) (mg) HPLC HPLC AEXII Loading 0.28 230 64.8 — — sample Elution 0.90  57 51.1 94.8 78.3 Quality and quantity profile for AEXII process

[0535] (7-4-2-3) Experimental Method

[0536] An experimental method was conducted in the same manner as specified in Table 90, experiments were independently conducted using a manually packed XK16 (Bed height 20 cm, CV 40 mL) column and a prepacked column (Bed height 10 cm, CV 5 mL), and the loading for each process was performed according to the conditions shown in Table 93.

[0537] The same volume (5.5 CV) of loading sample was applied in both experiments. However, compared to the previous experiment, the volume was approximately 3-fold smaller. While a flow rate was constantly maintained at 200 cm/h, there was a two fold difference in resistance time between the experiments due to the difference in bed height. Compared with the previous experiment, the conductivity of the loading sample slightly increased from 1.7 mS/cm to 2.0 mS/cm. After the process was completed, the purity of the sample was analyzed by RP-HPLC.

TABLE-US-00094 TABLE 93 Load Loading amount Sample Conc. Volume (mg/mL of Volume Conductivity Bed height (mg/mL) (mL) resin) (CV) (mS/cm) pH Prepack column 0.18 27.5 1.1 5.5 2.01 4.58 (10 cm) XK 16 columnn 0.18 220 1.0 5.5 2.02 4.45 (20 cm) Loading Sample Volume for bed height study

[0538] (7-4-2-4) Result of Experiment

[0539] The chromatograms for the experiment processes are shown in FIGS. 59 and 60. From the result of the experiment conducted with a bed height of 10 cm, it was confirmed that a peak was found in the elution step, not the washing step. It was confirmed that this result is similar to the conventional chromatogram (FIG. 56).

[0540] The result of the experiment conducted with a bed height of 20 cm shows that elution was initiated at a time point of 8CV in the washing step, and completed in the elution step. However, according to the result of the previous experiment conducted with the same bed height condition, the main peak was fully eluted in the washing step. Such a difference is estimated to be associated with the difference in loading sample volume (14.5 vs 5.5 CV).

TABLE-US-00095 TABLE 94 Conc. Volume Amounts Sample (mg/mL) (mL) (mg) Purity (%) Yield (%) H 10 cm Loading sample 0.18 27.5 5.4 — — Elution 1.47 1.9 2.8 99.7 N/A* Loading sample 0.18 220 40.9 — — H 20 cm Elution 1 0.25 93.1 23.1 99.1 56.4 Elution 2 0.69 16.2 11.2 99.3 27.3 Sum 83.6 *not measured due to the loss of elution sample during process Quality and quantity profile for CEX column length study

[0541] Consequently, to constantly maintain the bed height causing the difference in residence time, a 10 cm condition was selected, and it was confirmed that the same elution position as the conventional result was reproduced under conditions of a flow rate of 200 cm/hr and a loading sample volume of 5.5 CV.

[0542] (7-5) 4.sup.th Column Process Reproducibility Experiment

[0543] (7-5-1) Preparation of 4.sup.th Column Process Sample

[0544] To confirm the reproducibility of a 4.sup.th column process, an AEXII process was conducted, and the 4.sup.th column process was conducted in the same manner as shown in Table 90, except that the pooling criteria was changed to 200 mAU˜200 mAU to reduce the volume of a 4.sup.th column eluted fraction sample. The 4.sup.th process-completed sample was subjected to quantification by SEC-HPLC and purity measurement by RP-HPLC, confirming that there was no difference from the previous result, and the result is shown in Table 95.

TABLE-US-00096 TABLE 95 Conc. Volume Amounts Purity Yield Sample (mg/mL) (mL) (mg) (%) (%) AEXII Loading 0.28 230 64.8 — — sample Elution 1.19  39 46.5 94.1 71.8 Purity and step yield for preparation of loading sample for the study of CEX bed height

[0545] (7-5-2) Purpose of Experiment

[0546] It was to be confirmed whether the experiment result is reproduced using a manually packed column, rather than a prepacked column, under the 10-cm bed height (XK16 packing) condition.

[0547] (7-5-3) Experimental Method

[0548] While the experimental method was conducted as shown in Table 96, the volume of a loading sample was 7.0 CV, similar to the previous experiment result of 5.5 CV, and a loading amount was 1.65 mg Gly-Tβ4/mL of resin, which is slightly larger than the previous experiment result of 1.1 mg. The process-completed sample was quantified through SEC-HPLC, and its purity was confirmed through RP-HPLC.

TABLE-US-00097 TABLE 96 CEX process conditions Items Process conditions Resin Nuvia HR-S Column XK16 (ID 1.6 cm, Height 10 cm), Column volume (20 mL) Flow rate 200 cm/hr (6.67 mL/min) Sample preparation AEXII elution pool adjust pH and conductivity (pH 4.5, conductivity, 2 mS/cm - with DW) Process Equili- 20 mM Sodium Acetate anhydrous. pH 4.5 (2 CV) bration Sample AEXII eluted pool adjusted pH and loading conductivity <Loading amounts 32.9 mg Gly-Tβ4> Unbound 20 mM Sodium Acetate anhydrous. pH 4.5 (5 CV) wash Wash 20 mM Sodium Acetate anhydrous. 80 mM NaCl, pH 4.5 (5 CV) Elution 20 mM Sodium Acetate anhydrous, 180 mM NaCl, pH 4.5 (5 CV) Strip 2M NaCl (4 CV) CIP 0.5M NaOH (4 CV) Strip 2M NaCl (2 CV) Equili- 20 mM Sodium Acetate anhydrous. pH 4.5 (3 CV) bration

[0549] (7-5-4) Result of Experiment

[0550] The chromatogram for the process is shown in FIG. 61, and unexpectedly, the main peak began to be eluted from the end of washing step before changing to an elution buffer. It is assumed that the reason that the elution start time was slightly faster than the previous experiment result is due to an increase in loading amount (1.1-≥1.7 mg/mL of resin) and an increase in loading volume (5.5-≥7.0 CV).

TABLE-US-00098 TABLE 97 Loading Sample preparation for CEX feasibility test Load amount Conc. Volume (mg/mL of Conductivity Sample (mg/mL) (mL) resin) (mS/cm) pH Loading 0.24 139 1.65 1.91 4.53 sample (7.0 CV)

TABLE-US-00099 TABLE 98 Conc. Volume Amounts Purity Yield Sample (mg/mL) (mL) (mg) (%) (%) H Loading 0.24 139 32.9 — — 10 cm sample Elution 2.21 13.5 29.9 98.8 90.9 Purity and yield of purified sample for CEX bed height study

[0551] As a result of a CEX process, it was confirmed that a step yield was 91%, and a purity was 98.8% (Table 98). The purity was slightly lower than the previous experiment result (99.7%), and the step yield could not be compared due to the absence of a previous result.

[0552] (7-6) 4.sup.th Column Process Hold Time Study

[0553] (7-6-1) Purpose of Experiment

[0554] To ensure the storage condition and period that can be applied in GMP production, 4.sup.th column eluted fraction samples were stored for 7 days in cold room (2 to 8° C.) and room temperature (18 to 22° C.) conditions, and then a change in purity was examined.

[0555] (7-6-2) Experimental Method

[0556] For the experiment, a 4.sup.th process eluted pool was subjected to sterile filtration in a 5 L scale process, divided into microtubes by 300 uL, and stored for 7 days in cold room and room temperature conditions. Here, the sample was stored at −70° C. at the end of the storage period to completely freeze, and one day after the end of the storage period, all samples were simultaneously thawed and analyzed through RP-HPLC.

[0557] (7-6-3) Result of Experiment

[0558] The chromatogram obtained by performing the experiment under each storage condition (room temperature or cold room) and then, RP-HPLC, is shown in FIG. 62, and purities and main peak areas are shown in Table 99. As stability criteria for process samples, the system suitability criteria (% CV≤2.0%) according to RPC-HPLC (HU024-SOP-004) in (2-3) was applied, and even with the passage of time under the cold room and room temperature conditions, the % CV of the main peak area was 2.0 or less, confirming that there was no change. However, when the impurity peak of the part (retention time 27.4 min) indicated by a red circle in the sample stored in a cold room was compared with a day 0 sample, with the passage of time, it was confirmed that the peak area increased (Table 100). Accordingly, it was determined that the 4.sup.th process sample is stable in a cold room condition (2 to 8° C.) for 7 days, but is not appropriate for room temperature storage.

TABLE-US-00100 TABLE 99 Purity in RP-HPLC analysis for CEX hold time study Hold time at 2~8° C. Hold time at 18~22° C. Sample Main peak Area Purity (%) Main peak Area Purity (%) 0 day 8,982 98.4 8,707 98.4 1 day 8,720 98.4 8,704 98.1 2 day 9,078 98.4 8,987 98.0 4 day 9,028 98.3 8,677 97.9 7 day 8,745 98.3 8,703 97.6 % CV 1.7 0.05 1.3 0.3

TABLE-US-00101 TABLE 100 Impurity analysis in RP-HPLC analysis for CEX eluted pool hold time at room temperature (18 to 22° C.) Area (mAU) Main peak Sample 24.1 min 25.1 min (25.9 min) 27.1 min 27.4 min 0 day 104.5 N/D 8,707.0 35.9 N.D 1 day 110.5 19.9 8,704.5 34.0 14.8 2 day 109.6 20.0 8,986.8 40.4 19.0 4 day 112.4 18.1 8,676.9 35.9 30.7 7 day 114.9 16.0 8,703.1 43.4 42.9

[0559] (7-7) CEX Process Summary

TABLE-US-00102 TABLE 101 Summary of CEX process Items Process conditions Resin Nuvia HR-S Dynamic binding 1.65 mg Gly-Tβ4/mL capacity Equilibration 20 mM Sodium acetate anhydrate, pH 4.5 buffer Wash buffer 20 mM Sodium acetate anhydrate, 80 mM NaCl, pH 4.5 Elution buffer 20 mM Sodium acetate anhydrate, 180 mM NaCl, pH 4.5 Process 15~25° C. temperature Linear flow rate 200 cm/hr (Bed height 10 cm) Loading sample AEXII eluted sample adjust pH and condition conductivity (pH 4.5 with acetic acid, conductivity, 2 mS/cm with DW) Process Equilibration 2 CV Sample loading Sample Unbound wash (with 5 CV Equilibration buffer) Wash 5 CV Elution 5 CV Strip (2M NaCl) 4 CV CIP (0.5M NaOH) 3 CV Strip (2M NaCl) 2 CV Equilibration 3 CV DW 2 CV Storage (20% Ethanol) 3 CV Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU Hold time for Stable at 2~8° C. and 18~22° C. for up to 7 days AEXII eluted pool

Example (8) Purification of SL Scale Culture Broth

[0560] (8-1) Purpose of Experiment

[0561] It was to be confirmed whether the result performed at the XK16 size is reproduced with a BPG100 column size or more.

[0562] (8-2) Experimental Method

[0563] Fourth column purification was conducted by the method specified in Table 101, and the other process conditions are shown in Table 102. For depth filtration and an AEXI process, process step yields were calculated using a GST-Tβ4 standard material (prepared in 1st 50 L verification run) by the method described in (2-6), and for thrombin, AEXII and CEX processes, step yields were calculated through SEC-HPLC and RP-HPLC. In addition, each process sample was subjected to endotoxin and HCP analyses according to the methods described in (2-4) and (2-5).

TABLE-US-00103 TABLE 102 5 L scale process buffer and sample preparation condition Items Process conditions AEXI Resin Fractogel EMD TMAE(M) Equilibration 20 mM Tris-HCl, pH 8.0 buffer Elution buffer 20 mM Tris-HCl, 300 mM NaCl, pH 8.0 Linear flow rate 200 cm/hr (RT 6 min) Loading sample pH 8.0 ± 0.1 condition Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: UV.sub.280 nm Value ≥ 100 mAU criteria End point: UV.sub.280 nm Value ≤ 100 mAU Affinity Same of above conditions Thrombin Same of above conditions Reaction AEXII Resin Fractogel EMD TMAE(M) Mode Bind/Elute mode Equilibration 20 mM Tris-HCl, pH 8.0 buffer Elution buffer 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 Linear flow rate 200 cm/hr (RT 6 min) Loading sample pH 8.0 ± 0.1 condition conductivity 2 ± 0.2 mS/cm <5 Dilution with DW> Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU CEX Resin Nuvia HR-S Mode Cation-Exchange chromatography Equilibration 20 mM Acetate, pH 4.5 buffer Wash 20 mM Acetate, 80 mM NaCl, pH 4.5 Elution buffer 20 mM Acetate, 180 mM NaCl, pH 4.5 Linear flow rate 200 cm/hr (RT 3 min) Loading sample pH 4.5 ± 0.1, Conductivity 2 ± 0.2 condition mS/cm (adjust with Acetic acid and DW) <5 Dilution with DW> Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU

TABLE-US-00104 TABLE 103 5 L scale process conditions Items Process conditions Cell Lysis & Recovery Cell pellet 1080 g (batch No. 161115-B1, 161122-B2) Microfluidizer: 11,000 psi, 2 pass Continuous centrifugation (9,500 rpm, Pressure 250 kPa, Fee flow 100 L/hr) Depth filtration C0HC (0.054 m.sup.2) + F0HC (0.054 m.sup.2) + Opticap XL10 (0.45 μm) Column controller AKTA Pilot AEXI Resin Fractogel EMD TMAE(M) (2 cycle) Column BPG 100 (ID 10 cm, Height 20 ± 1 cm, column volume(CV): 1570 mL) Flow rate 262 mL/min Linear flow rate 200 cm/hr (RT 6 min) Affinity Resin Glutathione Sepharose 4 Fast Flow (1 cycle) Column BPG140 (ID 14 cm, Height 15 ± 1 cm, column volume(CV): 2308 mL) Flow rate 205 mL/min (Sample loading, Unboundwash) 385 mL/min (Elution) Linear flow rate 80 cm/hr (Sample loading, Unboundwash) 150 cm/hr (Elution) Thrombin Enzyme/Protein 1 mg GST-Tβ4/2 unit Thrombin Digestion ratio Incubation time 14~18 hrs Incubation temp 13~27° C. AEXII Resin Fractogel EMD TMAE(M) (1 cycle) Column BPG 100 (ID 10 cm, Height 20 ± 1 cm, column volume(CV): 1570 mL) Flow rate 262 mL/min Linear flow rate 200 cm/hr (RT 6 min) CEX Resin Nuvia HR-S (1 cycle) Column BPG100 (ID 10 cm, Height 10 ± 1 cm, column volume(CV): 785 mL) Flow rate 262 mL/min Linear flow rate 200 cm/hr (RT 3 min)

TABLE-US-00105 TABLE 104 Loading Sample preparation for 5 L scale process Load amount Conc. Volume (mg/mL of Conductivity Step (mg/mL) (mL) resin) (mS/cm) pH 1.sup.st column 4.34 3,650 8.15 2.02 8.09 (AEXI) 2.sup.nd column N/A (Affinity) 3.sup.rd column 0.21 11,536  1.48 1.79 8.05 (AEXII) 4.sup.th column 0.33 6,925 2.79 2.08 4.48 (CEX) (8.8 CV)

[0564] (8-3) Result of Experiment

[0565] The chromatogram for each process was shown as follows, and it was confirmed that the process was performed without a significant difference from the process development (FIGS. 63, 64 and 65). However, in the previous CEX experiment performed under the XK16 column condition, the main peak eluted when washing progressed at approximately 9 CV or more began to elute from the 6 CV time point (FIG. 66). As described above, it may be reconfirmed that the reason that the elution start time of the main peak became faster is due to an increase in loading amount (1.7-≥2.8 mg/mL of resin) and an increase in loading volume (7.0-≥8.8 CV). Therefore, to keep a constant elution position, it was determined that 4.sup.th column loading sample data should be considered in future GMP production.

TABLE-US-00106 TABLE 105 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield Purity (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) (%) (%) Cell breakage* 7,208 7,180 Recovery 7,180 5.34 6,480 34,603 Depth 5.34 6,480 34,603 4.14 8,795 36,411 105.2 Filtration/MF AEX I Cyc;e1 4.14 3,650 15,111 4.57 2,386 10,904 Cycle2 4.14 3,650 15,111 3.79 2,426 9,195 Thrombin 7.69 1,979 15,221 1.11 2,240 2,475 16.3 AEX II 0.21 11,536 2,434 1.66 1,405 2,334 96.7 96.3 CEX 0.33 6,925 2,299 2.61 875 2,286 99.4 98.4 UF/DF 2.61 872 2,278 10.29 221 2,276 99.9. 98.0 Total Yield (%) 7.8 *Cell Batch No. 161115-B1, 161122-B2 Yield and purity in 5 L scale process

TABLE-US-00107 TABLE 106 HCP Eluted Pool HCP Endotoxin Conc. Volume Amount Conc. Conc. HCP Conc. Conc. (mg/mL) (mL) (mg) (mg/mL) (ng/mL) (ppm) (EU/mL) (EU/mg) AEX I 4.35 4,840 21,054 4.35 14,339 3,296 >500000 Affinity 7.69 1,956 15,042 7.69 1,065 138 24,102 3,134 AEX II 1.66 1,416 2,354 1.66 16 10 0.065 0.04 CEX 2.61 880 2,299 2.61 4 1.5 0.005 0.002 UF/DF 10.29 221 2,273 10.29 21 2.0 4.55 0.44 Impurity analysis in 5 L scale process

[0566] It was confirmed that final yield obtained by a 5 L scale process was 7.8%, and the RP-HPLC purity of the 3.sup.rd column process eluted solution was 96.3%, which was equal to a small scale experiment result, and the RP-HPLC purity of the 4.sup.th column process eluted solution was 98.0%, which also satisfied the standard, 97% or more. However, compared to the previous small scale CEX purification results (99.6% and 98.8%), the purity was relatively low (98.4%), but the step yield was 99.4%, an increase of approximately 8% or more. It is determined that this is caused by the main peak elution time being faster than that of the previous experiment in the 4th column process. Purity was expected to decrease according to an increase in step yield (Table 105). The removal rates of process-derived impurities (HCP and endotoxin) were HCP 97.0% for HCP and 98.0% for endotoxin in the affinity process, compared to the AEXI process sample. After the affinity process, 99.9% or more of endotoxin was removed in the AEXII process, and the endotoxin removal rate was confirmed to be most excellent in the AEXII process. The endotoxin (≤0.5 EU/mg) and HCP (≤100 ppm) results in the final purification extract also met the criterion (Table 106). Based on the above results, a GMP technical transfer protocol was written.

[0567] (9-1) Purpose of Experiment

[0568] Through the purification of a 50 L scale culture broth, a sample for non-clinical testing was produced, and 100 vials of a standard product was to be prepared.

[0569] (9-2) Experimental Method

[0570] The conditions for a buffer for a 50 L verification run are shown in Table 107, and in 4.sup.th column process, unlike the washing condition used in the 5 L scale process, and NaCl concentration was changed from 80 mM to 60 mM, and the other processes were the same as used in the conventional process. In addition, in the 1.sup.st, 2.sup.nd and 3.sup.rd column processes, the flow rate applied in the conventional 50 L verification run was applied, and the 4.sup.th column process used a BPG200 column, so that the maximum flow rate that can be driven in an AKTA pilot was applied, but the conventional 200 cm/hr condition was not applied. Accordingly, the 150 cm/hr condition was applied (manufacturer's instructions and records, batch No HU024-RD1603).

TABLE-US-00108 TABLE 107 3.sup.rd 50 L verification run process conditions Items Process conditions AEXI Column scale BPG 200 (i.d 20 cm, H 15.8 cm), CV 4961 mL (3 cycle) Equilibration buffer 20 mM Tris-HCl, pH 8.0 Elution buffer 20 mM Tris-HCl, 300 mM NaCl, pH 8.0 Linear flow rate 61 cm/hr (Flow rate 319 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: UV.sub.280 nm Value ≥ 100 mAU criteria End point: UV.sub.280 nm Value ≤ 100 mAU Process Equilibration buffer (2 CV) Sample loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) Affinity Column scale BPG 200 (i.d 20 cm, h 10.5 cm), CV 3297 mL (3 cycle) Equilibration buffer 20 mM Tris-HCl, 150 mM NaCl, pH 8.0 Elution buffer 50 mM Tris-HCl, 10 mM Glutathione (Reduced), pH 8.0 Acidic buffer 100 mM Sodium Acetate, 500 mM NaCl, pH 4.5 Alkaline buffer 100 mM Tris-HCl, 500 mM NaCl, pH 8.0 Linear flow rate 61 cm/hr (Flow rate 319 mL/min) Loading sample Direct Loading condition Elution pooling Start point: UV.sub.280 nm Value ≥ 100 mAU criteria End point: UV.sub.280 nm Value ≤ 100 mAU Process Equilibration buffer (2 CV) Sample loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) Acidic buffer (5 CV) Alkaline buffer (5 CV) 6M Guanidine hydrochloride (2 CV) DW (2 CV) Equilibration buffer (3 CV) Thrombin Enzyme Bovine Thrombin Reaction Enzyme ratio 1 mg GST-Tβ4 per 2 unit thrombin & AEXII Reaction time 16 + 2 H (2 cycle) Reaction temp 22 ± 2° C. Agitation 100 ± 10 ppm Column scale BPG 200 (i.d 2 0 cm, H 15.8 cm), CV 4961 mL Equilibration buffer 20 mM Tris-HCl, pH 8.0 Elution buffer 20 mM Tris-HCl, 80 mM NaCl, pH 8.0 Linear flow rate 61 cm/hr (Flow rate 319 mL/min) Loading sample pH 8.0 ± 0.1 condition Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU Process Equilibration buffer (2 CV) Sample loading Unbound wash (with EQ buffer) (5 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV) CEX Column scale BPG 140 (i.d 14 cm, H 9.0 cm), CV 1385 mL (2 cycle) Equilibration buffer 20 mM Sodium Acetate, pH 4.5 Wash buffer* 20 mM Sodium Acetate, 60 mM NaCl, pH 4.5 Elution buffer 20 mM Sodium Acetate, 180 mM NaCl, pH 4.5 Linear flow rate* 150 cm/hr (Flow rate 385 mL/min) Loading sample pH 4.5 ± 0.1 condition Conductivity 2.0 ± 0.2 mS/cm Elution pooling Start point: UV.sub.214 nm Value ≥ 200 mAU criteria End point: UV.sub.214 nm Value ≤ 200 mAU Process Equilibration buffer (2 CV) Sample loading Wash buffer (10 CV) Elution buffer (5 CV) 2M NaCl (4 CV) 0.5M NaOH (5 CV) 2M NaCl (3 CV) Equilibration buffer (3 CV)

[0571] (9-3) Result of Experiment

[0572] The chromatograms for the processes are shown in FIGS. 67 to 70, and it was confirmed that reproducible results are derived. The cumulative purification process yield was 12.4% (Table 108).

TABLE-US-00109 TABLE 108 Loading Sample Eluted Pool Conc. Volume Amount Conc. Volume Amount Yield Purity* (mg/mL) (mL) (mg) (mg/mL) (mL) (mg) (%) (%) Cell breakage 32,640 5.62 24,805 139,404 (170114.) Depth 5.62 24,805 139,404 4.25 34,225 145,456 104.3 Filtration&MF AEXI 4.25 34,180 145,265 6.03 23,910 144,106 99.2 Glutathione 6.03 17,930 108,064 10.47 8,370 87,634 81.1 S4FF Thrombin 10.47 8,330 87,215 2.04 9,555 19,586 22.5 AEX II 2.04 9,,545 19,565 1.97 7,215 14,206 72.6 96.3 CEX 1.97 7,175 14,128 4.61 2,901 13,371 94.6 98.1 UF/DF 4.61 2,891 13,323 10.23 1,299 13,283 99.7 98.1 Total Yield (%) 12.4 *Purity was confirmed through RP-HPLC Purity and Yield for 3.sup.rd 50 L Verification run

[0573] The content of process-derived impurities for each process measured (Table 109).

TABLE-US-00110 TABLE 109 Eluted Pool HCP Endotoxin Conc. Volume Amount Conc. HCP Conc. HCP Conc. Conc. (mg/mL) (mL) (mg) (mg/mL) (ng/mL) (ppm) (EU/mL) (EU/mg) AEX I 6.03 23,910 144,106 6.03 11,554 1,916 173,276 28,735 Affinity 10.47 8,370 87,634 10.47 745.5 71.2 2,476 236 AEX II 1.97 7,215 14,206 1.97 13.3 6.7 0.065 0.033 CEX I 4.61 2,901 13,371 4.61 3.7 0.8 <0.005 0.001 UF/DF 10.23 1,299 13,283 10.23 17.0 1.7 0.866 0.085 HCP and endotoxin analysis for 3.sup.rd 50 L verification run

[0574] Approximately 13 g of the final purified extract was obtained, and the RP-HPLC purity was confirmed to be 98.1%, which meets the criterion (97% or more) (Table 108). HCP and endotoxin contents also met the standards (Table 109). It was confirmed that the unit process removal rates of the process-derived impurities were HCP 97.7% for HCP and 99.5% for endotoxin in the affinity process. After the affinity process, in the AEXII process, the endotoxin unit process removal rate was confirmed to be 99.9%. It can be seen that this result was equal to the result of the 5 L scale culture broth purification, which is the previous experiment.

[0575] It should be understood by those of ordinary skill in the art that the above description of the present invention is exemplary, and the exemplary embodiments disclosed herein can be easily modified into other specific forms without departing from the technical spirit or essential features of the present invention. Therefore, the exemplary embodiments described above should be interpreted as illustrative and not limited in any aspect.

INDUSTRIAL APPLICABILITY

[0576] According to the present invention, high-purity Gly-Tβ4 can be stably prepared on a large scale, so that it is considered that the present invention has a great industrial value in that Gly-Tβ4 having an excellent industrial value can be obtained economically and with high efficiency.