A new tetrapeptide biomimetic chromatographic resin for antibody separation with high adsorption capacity and selectivity

Journal of Chromatography A 1604 (2019) 460474

Contents lists available at ScienceDirect

Journal of Chromatography A journal homepage: www.elsevier.com/locate/chroma

A new tetrapeptide biomimetic chromatographic resin for antibody separation with high adsorption capacity and selectivity Yu-Ming Fang, Sheng-Gang Chen, Dong-Qiang Lin, Shan-Jing Yao∗ Key Laboratory of Biomass Chemical Engineering of Ministry of Education, College of Chemical and Biological Engineering, Zhejiang University, Hangzhou 310027, China

a r t i c l e

i n f o

Article history: Received 11 June 2019 Revised 16 August 2019 Accepted 21 August 2019 Available online 22 August 2019 Keywords: Biomimetic chromatography Tetrapeptide ligands Antibody Separation and purification

a b s t r a c t Biomimetic affinity chromatography with short peptide ligands is a developing technology, which has great potential for antibody separation and purification. In this study, a tetrapeptide library with critical residues of natural ligands to hIgG was constructed and a novel tetrapeptide ligand (Ac-FYHE) with high LibDock scores was selected by molecular docking. Then, Ac-FYHE ligand was linked to agarose bead to prepare a new chromatography resin. The properties of antibody adsorption were measured and evaluated by static/dynamic adsorption. It was found that the resin with ligand Ac-FYHE has high binding capacity and selectivity for hIgG. The results showed the Qm-hIgG of Ac-FYHE-4FF resin was 87.9 mg/g resin while the Qm-BSA of this resin was only 16.5 mg/g resin at pH 7.0. Moreover, at pH 7.0, Q10% of AcFYHE-4FF resin was 24.1 mg/mL for hIgG but just 2.1 mg/mL for BSA, which presented high selectivity of the screened resin at pH 7.0. Subsequently, the adsorption and separation properties of the Ac-FYHE-4FF resin were further investigated. As a result, with the addition of 0.5 M NaCl, Qm decreased by less than 20% but Qm decreased by 70% with the addition of 50% (v/v) ethylene glycol, which indicated that hydrophobic interaction would be the driving force for the binding between resin and hIgG. Besides, pH 7.5 and pH 4.5 could be the optimal loading and elution condition for hIgG, respectively. Finally, the AcFYHE-4FF resin was applied to separate mAb or/and hIgG from BSA containing feedstock, CHO cell culture supernatant and human serum, and the purity and recovery were both more than 90% with only one-step separation. The results indicate that the Ac-FYHE-4FF resin developed in this work would be promising for antibody separation and purification. © 2019 Elsevier B.V. All rights reserved.

1. Introduction Antibodies as large Y-shaped proteins have been found in the blood or other body fluids of vertebrates, which could recognize a unique part of the foreign target called an antigen [1]. Due to antibodies’ high specificity, they have been widely used in the preparation and development of therapeutic drugs and diagnostic reagents. Monoclonal antibodies (mAbs) as the most important of the antibodies have been well used in the treatment of cancer, cardiovascular and cerebrovascular diseases, inflammation and so on with the advantages of high specificity, strong targeting and small toxicity [2–5]. However, with the increasing market demand for mAbs and the rapid development of upstream processing, it is important to develop downstream processing by finding new resins for antibody purification [5–7]. Nowadays, Protein A affinity chromatography is commonly applied for antibody purification. However, the use of Protein A affin∗

Corresponding author. E-mail address: [email protected] (S.-J. Yao).

https://doi.org/10.1016/j.chroma.2019.460474 0021-9673/© 2019 Elsevier B.V. All rights reserved.

ity chromatography is challenged because of toxic ligand leakage, high cost, harsh elution condition and so on [8–11]. As a potential alternative to protein A affinity chromatography, peptide biomimetic chromatography has the advantages of cheaper to produce, milder elution and resistance to enzymatic degradation [12,13]. Peptide biomimetic chromatography using peptide ligands with critical residues from natural ligands has raised increasing interest in recent years [14,15]. For example, Wang et al. [11] discovered a new peptide resin (Ac-YFRH-4FF resin) by flexible docking and molecular dynamics simulation. The results showed this resin could purify hIgG with the purity of 98.4% and recovery of 89.4% from the BSA-containing feedstock. Moreover, this resin could separate mAbs from CHO cell culture supernatant with the purity of 98.0% and recovery of 79.5%. In addition, Yang et al. [16–18] screened HWRGWV ligand from a synthetic solid phase library. The results indicated HWRGWV resin could bind all human IgG subclasses and IgGs from bovine, mouse, goat and rabbit. Moreover, this resin could separate hIgG from complete mammalian cell culture medium with both recovery and purity as high

2

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

as 95%. Therefore, this research was carried out to screen new peptide ligands with high selectivity and adsorption for antibody purification. In order to select peptide ligands with high affinity, a peptide library with potential candidates is needed. In our previous study, according to the literature, Gln129, Phe132, Tyr133, Leu136, Glu143, Ile150, and Lys154 from ProA have relatively great contributions to the Fc-ligand binding, while Trp4, His5, Leu6, Glu8, Leu9, Val10 and Trp11 from peptide Fc-III are critical residues to Fc-ligand binding [19]. Therefore, a tetrapeptide library were constituted by choosing four residues from discovered critical residues and randomly combining them. Then, LibDock as a high-through docking module was applied to screen tetrapeptide ligand with high LibDock Score from the library. In this work, a novel tetrapeptide ligand (Ac-FYHE) with high LibDock Score would be screened by molecular docking. Then, AcFYHE ligand would be coupled to agarose bead for preparing a new resin and the static/dynamic adsorptions of this resin for hIgG and BSA would be investigated. Finally, the Ac-FYHE-4FF resin would be used to separate hIgG and mAb from human serum and CHO cell culture supernatant for promotion and application of peptide biomimetic chromatography.

2. Materials and methods 2.1. Materials Bestarose 4FF (4% highly cross-linked agarose beads) was purchased from Bestchrom Bio-Technology Co., Ltd. (Shanghai, China). The ligand of Ac- Phenylalanine-Tyrosine-Histidine-Glutamic acid (Ac-FYHE, purity: 98.0%) was synthesized by Chinese Peptide Company (Hangzhou, China). Human immunoglobulin G (hIgG) for intravenous injection was obtained from Boya Biopharmaceutical Ltd. (Jiangxi, China) and Bovine albumin (BSA) was purchased from BBI Life Science (Shanghai, China). allyl bromide (AB), N-Bromosuccinimide (NBS), 2-(7-Aza-1H-benzotriazole-1-yl)1,1,3,3- tetramethyluronium hexafluorophos-phate (HATU), N,NDiisopropylethylamine (DIPEA) and anhydrous dimethylformamide (DMF) were purchased from Aladdin Industrial Corporation (Shanghai, China). Chinese hamster ovary (CHO) cell culture supernatant containing monoclonal antibody was obtained from a local biotechnology company (Hangzhou, China). Human serum was obtained from Zhejiang university hospital. Other reagents were of analytical grade and used as received.

2.2. Molecular docking The molecular docking of tetrapeptide ligand onto the Fc-A molecule was carried out by the “LibDock” modules provided by Discovery Studio 2.5. LibDock is a high-through docking module that aligns the ligand conformations to the polar and apolar receptor interaction sites (hotspots or potential binding sites) [20]. Normally, the higher the LibDock score, the more stable of the binding structure. For LibDock, tetrapeptide and Fc-A corresponded to the ligand and the receptor, respectively. And the procedure could be carried out by the following steps: Firstly, Fc-A fragment was applied with CHARMm force field and covered by 18 interaction range spheres with the radii of 20 A˚ in a way that whole Fc-A could be fully overlapped. Then tetrapeptide was aligned to polar and apolar receptor interaction site (hotspot) for each interaction range sphere, which was deemed as the potential binding site [20]. Finally, Ac-FYHE ligand was screened with high LibDock score, which could be seen in Fig. 1, and it was selected for the further usage.

2.3. Preparation of the screened resins The preparation process of the Ac-FYHE-4FF resin is shown in Fig. 2. 10.0 g Bestarose 4FF agarose bead was drained and mixed with 10 mL allyl bromide and 5 g sodium hydroxide in 10 mL 20% (v/v) dimethyl sulfoxide solution. The reaction was performed under 180 rpm and 30 °C for 24 h. The beads were then rinsed with ethanol and deionized water, and added into 50% acetone solution with 5 g NBS and reacted at 30 °C and 180 rpm for 3 h. Subsequently, the brominated beads were rinsed with deionized water and added into 1 M carbonate buffer (pH 12.0) to react with 3 mL hexamethylene diamine under 180 rpm and 30 °C for 24 h. The prepared resin was rinsed with deionized water, 0.1 M HCl, 0.1 M NaOH and deionized water in sequence. Finally, the resin was drained and stored in 20% ethanol at 4 °C for further usage [11]. Then the tetrapeptide ligand (Ac-FYHE) was grafted to the resin by the following steps. Firstly, 3.0 g resin prepared before was sequentially rinsed with deionized water, ethanol and DMF before peptide coupling. Then, HATU and DIPEA were used to link peptide to the aminated resin. The molar ratio of peptide: amino groups: HATU: DIPEA was set as 1:1:2:4 and the reaction was performed under 180 rpm and 25 °C for 8 h. Then, the resin was washed extensively with DMF, ethanol and deionized water. Before being stored, the resin was added into the mixture of acetic anhydride and sodium acetate at 25 °C and 180 rpm for 1 h to block the untreated free primary amine on the resin to avoid unspecific adsorption. Finally, the Ac-FYHE-4FF resin was prepared and stored in 20% ethanol at 4 °C for further usage [11]. 2.4. Determination of amino density and ligand density The amino density of the amino-activated matrix was determined by the HCl titration. Firstly, the gel was successively washed by deionized water, 0.1 M HCl solution and 0.1 M NaOH solution for three times. Secondly, 1 g drained resin was equilibrated in 5 mL 0.1 M NaCl solution. Finally, 0.1 M HCl was used to titrate the solution to pH 4.5. The experiment was conducted in triplicate to get the ionic capacity of the resin calculated by mass balance. And the amino density was described by equation (Eq. (1)) as,

ρ=

c ×V × 10 0 0 2 ×m

(1)

where ρ is the amino density of the resin (μmol/g resin), c is the concentration of the HCl solution (mol/L), V is the volume of the HCl solution (mL) and m is the mass of the resin (g), respectively. Subsequently, the ligand density of the tetrapeptide resin can be calculated by the decrease of the peptide in the reaction solution. The peptide concentration was measured with reversedphase high performance liquid chromatography (RP-HPLC) on Agilent 1100 series (Agilent Technologies, Santa Clara, CA). A Hypersil ODS-2 C18 column (Thermo Fisher Scientific Inc., Waltham, MA, USA) was used. And solvent A was 0.01% Trifluoroacetic acid (TFA) in H2 O, solvent B was 0.09% TFA in the acetonitrile and 20% H2 O. And all the ligands were analyzed using stepwise gradient from 22% to 32% B in 20 min. Ultraviolet (UV) detection of the tetrapeptide was performed at 220 nm. 2.5. Adsorption isotherms The hIgG and BSA adsorption isotherms of the prepared AcFYHE-4FF resin were measured in 20 mM acetate buffer (pH 5.0), 20 mM phosphate buffer (pH 6.0∼8.0), 20 mM Glycine-NaOH (pH 9.0) at 25 °C. 30 mg drained resins were mixed with 0.8 mL hIgG solution or BSA solution with different concentration, and the mixture was incubated in a thermomixer under 1500 rpm and 25 °C for 3 h. The supernatant was separated by centrifugation at 70 0 0 rpm

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

Fig. 1. (a) The structures of Ac-FYHE and (b) the typical binding conformation of Ac-FYHE ligand on the Fc-A.

Fig. 2. Preparation of the screened tetrapeptide resin.

3

4

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

and unbound protein was measured with spectrophotometer (One Drop OD-10 0 0+, Wins Technology, Nanjing, China) at 280 nm. The adsorption isotherm was described by Langmuir equation (Eq. (2)) as,

Q∗ =

Qm × C ∗ Kd + C ∗

(2)

where Q∗ and C∗ are the equilibrium protein concentration on the resin (mg/g resin) and the equilibrium protein concentration in the solution (mg/mL), respectively. Qm stands for the saturation adsorption capacity (mg/g resin) and Kd is the dissociation constant (mg/mL). Then, the adsorption isotherms of hIgG with the peptide resin prepared were measured at different NaCl concentrations (0–1 M) at pH 7.0 or at different ethylene glycol concentrations (0, 25%, 50%) a fixed pH value of 7.0 to see the effect of the NaCl and ethylene glycol to the adsorption behavior of hIgG on the Ac-FYHE-4FF resin for getting the interactions between the Ac-FYHE ligand and the hIgG. 2.6. Measurement of dynamic binding capacity at 10% breakthrough The measurement of dynamic binding capacity at 10% breakthrough (Q10% ) was performed with a Tricorn 5/50 column (GE Healthcare, Uppsala, Sweden). Ac-FYHE-4FF resin was packed into column and equilibrated with the different buffer (pH 5.0–9.0) at linear velocities of 100 cm/h or with the linear velocities from 100 cm/h to 300 cm/h at pH 7.0 to get the influence of the pH and the linear velocities to the 10% breakthrough of the selected resin. The processes were carried out by the following steps. Ac-FYHE4FF resin was packed into column and equilibrated with the buffer (pH 5.0–9.0) until the absorbance baseline at 280 nm was stable. Then, 2.0 mg/mL hIgG or BSA dissolved in the equilibrium buffer was loaded at 100 cm/h (when the buffer was pH 7.0, the linear velocities were from 100 cm/h to 300 cm/h). The protein concentration in the effluent was detected by a UV detector at 280 nm. Subsequently, the column was eluted with 20 mM acetate buffer (pH 4.0) and regenerated and re-equilibrated with 0.1 M NaOH and the equilibrium buffer, respectively. Q10% (mg/mL resin) was calculated as follow,

resin at different pH conditions were evaluated to get the best loading pH value by the following steps. Ac-FYHE-4FF resin was packed into column and equilibrated with the different buffer system (pH 5.0, pH 6.0, pH 6.5, pH 7.0, pH 7.5, pH 8.0, pH 8.5) until the absorbance baseline at 280 nm was stable. Then, hIgG (about 2 mg/mL) or BSA (about 8 mg/mL) was loaded at 0.5 mL/min. Subsequently, buffer system was used to washout the non-bound proteins at 1.0 mL/min. Finally, the elution and a re-equilibrium were carried out to regenerate the resin. The percentage of the adsorbed protein was defined as follow,



PA % =

AA 1− A0



× 100

(4)

where PA % is the percentage of the adsorbed protein. AA is the breakthrough peak area of the proteins and A0 is the total peak area of the proteins (mL). 2.9. Elution of hIgG by the Ac-FYHE-4FF resin Ac-FYHE-4FF resin was packed into column and equilibrated at pH 7.5 until the absorbance baseline at 280 nm was stable. Then, 2 mg/mL hIgG was loaded at 0.5 mL/min. After buffer system was used to washout the non-bound proteins at 1.0 mL/min, elution was carried out at different pH conditions (pH 3.5, pH 4.0, pH 4.5, pH 5.0, pH 5.5, pH 6.0). Finally, a clean in place (CIP) and a reequilibrium were carried out to regenerate the resin. The elution efficiency of the adsorbed protein was defined as follow,

PE (% ) =

AE × 100 A0

(5)

where PE % is the elution efficiency of the adsorbed protein. AE is the elution peak area of the hIgG and A0 is the total peak area of the hIgG. 2.10. Separation of hIgG from the BSA containing feedstock

where C and C0 are the protein concentration of outlet and initial fluid, respectively (mg/mL). V10% is the loading volume at 10% breakthrough and Vresin is the volume of packed resin (mL).

The ÄKTA start system was used and the same as the following sections. the protein mixture of 2 mg/mL hIgG and 8 mg/mL BSA was prepared with the equilibrium buffer (20 mM phosphate buffer at pH 7.5, named buffer A). 5 mL of protein mixture was loaded and elution was performed at pH 4.5 (6 CV of 20 mM acetate buffer, pH 4.5, named buffer B). The elution peaks were collected with 2 mL fraction for further analysis. Finally, a CIP and a re-equilibrium were carried out to regenerate the resin. The flow rate was 0.5 mL/min and the chromatographic run was monitored on-line at 280 nm.

2.7. Adsorption kinetics

2.11. Purification of mAb from CHO cell culture supernatant

Adsorption kinetics of hIgG on Ac-FYHE-4FF resin were measured at different pH conditions (pH 5.0–9.0) by the following steps. 2 mg/mL hIgG was prepared with different pH conditions (pH 5.0–9.0). Then, 0.03 g drained resins were added into 1 mL protein solution in a 2 mL tube. Subsequently, the mixture was incubated in a thermomixer under 1500 rpm and 25 °C. The supernatant was taken to measure the protein concentration periodically (0–10 min interval was 1 min, 10–20 min interval was 2 min, 20– 30 min interval was 5 min, finally, 40 min and 60 min). Then the adsorption kinetics curves were obtained and fitted by the pore diffusion model (PDM) [21–23]. Finally, the effective pore diffusivity (De ) was calculated.

The feedstock of CHO cell culture supernatant was prepared and adjusted to pH 7.5. 5 mL CHO cell culture supernatant (about 0.9 mg hIgG/mL) was loaded at 0.5 mL/min, and non-bound proteins were washed out at 1.0 mL/min. Then, the elution was performed at 1.0 mL/min at pH 4.5. The elution peaks were collected with 2 mL fraction for further analysis. Finally, a CIP and a reequilibrium step were carried out to regenerate the resin. The chromatographic run was monitored on-line at 280 nm.

Q10% =

C0 · ∫V010% (1 − C/C0 )dV Vresin

(3)

2.8. Adsorptions of hIgG and BSA by the Ac-FYHE-4FF resin With the ÄKTA start system (GE Healthcare, Uppsala, Sweden), the adsorption behaviors of hIgG and BSA in the Ac-FYHE-4FF

2.12. Purification of hIgG from human serum The feedstock of human serum (about 1 mg/mL hIgG) was prepared and adjusted to pH 7.5. The column was pre-equilibrated with buffer A until the stable baselines of UV absorption and conductivity were reached. Then 10 mL feed stock was loaded at 0.5 mL/min, and 10 CV of buffer A was used to wash out the nonbound proteins at 1.0 mL/min. Then, the elution was performed at

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

5

Table 1 Docking results between peptide ligand and Fc-A fragment. Ligands

Ac-FYHE DAAG HWRGWV FYWHCLDE

Total poses (dock score > 90)

142 102 104 92

Binding site 1

Binding site 2

Poses (score > 90)

Maximum dock score

Poses (score > 90)

Maximum dock score

39 22 16 14

166.3 149.3 154.1 139.4

90 65 67 62

175.4 151.4 162.4 148.9

1.0 mL/min at pH 4.5. The elution peaks were collected with 2 mL fraction for further analysis. Finally, a CIP and a re-equilibrium were carried out to regenerate the resin. The chromatographic run was monitored on-line at 280 nm. 2.13. SEC-HPLC analysis The concentration and purity of hIgG were analyzed by size exclusion chromatography-high performance liquid chromatography (SEC-HPLC) with LC-30 0 0 HPLC system (Beijing Chuangxintongheng Science & Technology Co., Ltd, Beijing, China). The analytical column TSKgel G30 0 0 SWXL (ID 7.8 mm × 30.0 cm,TOSOH BIOSCIENCE, Tokyo, Japan) and a guarding column (TSK-guardcolumn SWXL , ID 6.0 mm × 4.0 cm, TOSOH BIOSCIENCE, Tokyo, Japan) were used. The injection volume was 20 μL, and the flow rate was 0.6 mL/min. The mobile phase was 0.1 M Na2 SO4 in 0.1 M phosphate solution, pH 6.7. The detection wave length was 280 nm. All samples should be filtered with a 0.22 μm microporous membrane. The concentration of hIgG was determined by referring to the standard curve, while the purity of hIgG was defined as the ratio of the peak area of hIgG monomer to that of total protein peak:

PhIgG (% ) =

Peak area o f IgG × 100 T otal protein peak area

(6)

The recovery of hIgG was calculated as the ratio of the amount of hIgG in the elution fraction to that in the loading feedstock as follows,

Elution amount RhIgG (% ) = Loading amount

(IgG peak area × VE ) × 100 (IgG peak area × VL )

(7)

where VE is equal to the elution volume of hIgG peak and VL presents the loading volume of feedstock. 2.14. SDS-PAGE analysis The feedstock and the elution fractions were analyzed by 10% SDS-PAGE under non-reducing conditions. The protein solution was diluted or concentrated to the appropriate concentration (0.5– 1 mg/mL). 5 μL of protein marker and 10 μL of protein sample were applied. The protein migration was run at a constant voltage of 120 V for 90 min. The gel was stained with Coomassie Blue R-250, and then destained. The stained protein gel was analyzed by the Gel Doc 20 0 0 imaging system (Bio-Rad, Hercules, CA, USA). 3. Results and discussion 3.1. Screening for tetrapeptide ligands with high affinity According to the literature, ten residues (Gln, Phe, Tyr, Leu, Glu, Ile, Lys, Trp, His and Val) were discovered as critical residues of natural ligands to bind hIgG [19]. Therefore, a tetrapeptide library was constituted by choosing four residues from these residues, and randomly combining these four amino acids. Then, Fc-A fragment was covered by 18 interaction range spheres with the radii of 20 A˚ in a way that whole Fc-A could be fully overlapped (Fig. S1(a)), and two potential binding sites on Fc-A were selected (Fig. S1(b)).

Subsequently, the software of Discovery Studio was used for the peptide screening and a tetrapeptide ligand (Ac-FYHE, Fig. 1(a)) with high LibDock Score was selected. Compared to the reported short peptide ligands (DAAG [24], HWRGWV [25], FYWHCLDE [26]) shown in Table 1, the screened ligand showed comparable binding ability with the highest LibDock Score between 160 and 180. Moreover, the typical binding conformation of Ac-FYHE on the Fc-A could be seen in Fig. 1(b), which showed that Binding site 1 could bind ligand by inserting aromatic rings to hydrophobic cavity and Binding site 2 could bind ligand by inserting two aromatic rings to two hydrophobic cavities to form a “pincer” structure. Therefore, Ac-FYHE was selected as potential tetrapeptide ligand by forming stable structure to target protein. 3.2. Preparation of screened resins Bestarose 4FF (4% highly cross-linked agarose beads) was used to prepare the peptide resin. The amination and peptide coupling procedures are shown in Fig. 2. Here diaminehexane acted as the spacer arm and provided a free amine for the ligand coupling. Then, tetrapeptide ligand (Ac-FYHE) was coupled by HATU chemistry. The ligand density of the drained Ac-FYHE-4FF resin prepared was 89 μmol/g resin. 3.3. Adsorption of BSA and hIgG with screened resins With hIgG and BSA as the model proteins, the adsorption isotherms of the screened tetrapeptide resins were determined under different pH conditions (pH 5.0–9.0). From the Fig. 3, the adsorption of the Ac-FYHE-4FF resin showed strong pH-dependency, and the neutral environment (pH 7.0) around pI of hIgG (pH = 6.5 [5]) was considered as the adequate adsorption condition of the resin to adsorb hIgG with the advantages of avoiding denaturation or aggregation of antibody [11]. When the adsorption condition closes to the pI of the hIgG, the surface hydration layer of hIgG becomes thinner and the binding effort of hydrophobic interaction between the resins and the hIgG becomes stronger, which helps increase the adsorption ability of the resin. Moreover, the saturation adsorption capacity (Qm ) and the dissociation constant (Kd ) of the resin to hIgG at pH 7.0 were 87.9 mg/g resin and 0.31 mg/mL, respectively, while the Qm and the Kd of the resin to BSA at pH 7.0 were 16.5 mg/g resin and 3.21 mg/mL, respectively. The results showed to the Ac-FYHE-4FF resin at pH 7.0, Qm of the BSA is 82% less than Qm of the hIgG, and the Kd of the BSA is more than 10 times to the Kd of the hIgG (The higher the Kd , the lower the binding ability). The results indicated that the binding ability of the Ac-FYHE-4FF resin to the hIgG is much higher than the BSA, which meant the high binding ability and selectivity of the screened resin. On the one hand, by the increase of the pH (8–9) away from the pI of hIgG, the decrease of the hydrophobic interaction between the ligand and the hIgG leads to the lower of the Qm . On the other hand, when the pH condition between pH 5.0 to pH 6.0, hIgG is positively charged, and the Histidine of tetrapeptide ligands has the same charge, which creates the electrostatic repulsion between the ligand and the hIgG and leads the decrease of the Qm .

6

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

Fig. 4. Effect of (a) NaCl and (b) ethylene glycol on the adsorption behavior of hIgG on Ac-FYHE-4FF resin. Fig. 3. Adsorption isotherms of (a) hIgG and (b) BSA on the Ac-FYHE resin at pH 5.0–9.0.

Moreover, pH 5.0 is the best adsorption condition of the resin to the BSA, which closes to the pI of the BSA (pI = 4.8 [5]). Therefore, the increase of the pH decreased the binding ability between the resins and the BSA. What’s more, Ac-FYHE-4FF resin almost could not adsorb BSA at pH 7.0–9.0 shown in Fig. 3(b), which was caused by the existence of the Glutamate. At pH 7.0–9.0, Glutamate is negatively charged and BSA is also negatively charged, which presents the strong electrostatic repulsion between the ligand and the BSA. Then, to figure out the interactions between the resin and the hIgG, following researches were carried out. Firstly, NaCl as a neutral salt was used to calculate the influence of the electrostatic interaction to the resin, which could effectively shield electrostatic interaction without changing hydrophobic effect [27,28]. Moreover, ethylene glycol was also used to observe the interactions between the ligand and the resin by the reason that the addition of ethylene glycol, which influenced the binding of the ligands and the resin, could see the contribution of the hydrophobic interaction to the binding [29,30]. Fig. 4(a) shows Qm of the hIgG decreased by 6.78–13.16% when the NaCl concentration increased from 0 M to 0.1 or 0.25 M. While NaCl concentration was further increased to 0.5 or 1 M, Qm remained basically stable. Especially, when the salt concentration increased to 1 M, Qm only decreased by 18.54%, compared to the

FYWHCLDE reported before (87% hIgG was lost in the 0.5 M NaCl concentration [31]). Therefore, Ac-FYHE-4FF resin showed high salt tolerance to bind hIgG. Salt tolerance of the resins could improve efficiency and save operating costs when it is not necessary to dilute or desalinate the feedstock in order to reduce its conductivity in the real separation process. Fig. 4(b) shows Qm of the hIgG decreased by 24.7% when the ethylene glycol concentration increased from 0 to 25% (v/v). Moreover, when the ethylene glycol concentration increased to 50% (v/v), the resin almost couldn’t adsorb hIgG, which meant that hydrophobic interaction was the dominant interaction in the binding of the Ac-FYHE-4FF resin and hIgG at pH 7.0. 3.4. Measurement of dynamic binding capacity at 10% breakthrough Ac-FYHE-4FF resin was used to see the influence of the different pH conditions and different flow rates to the DBC of hIgG. Firstly, the breakthrough curves of hIgG in Ac-FYHE-4FF resin at 100 cm/h with different pH conditions (pH 5.0–9.0) were shown in Fig. S2, which showed the same adsorption rules as the static binding ability of the Ac-FYHE-4FF resin. pH 7.0 is the best adsorption condition and the Q10% is 24.1 mg/mL resin. Therefore, compared to the DBC of several commercial Protein A resins, such as 12–19 mg/mL resin of Pierce Protein A Agarose to hIgG,

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

7

Fig. 5. Adsorption kinetics curves of hIgG on Ac-FYHE-4FF at varying pHs.

Ac-FYHE-4FF resin showed the comparable binding ability [32]. Then the different flow rates (10 0–30 0 cm/h) were carried to see the influence of the binding ability, which was shown in Fig. S3. Q10% decreased from 24.1 mg/mL to 16.7 mg/mL with flow rate increasing from 100 cm/h to 300 cm/h. The Q10% was decreased by the increase of the linear flow rate, which was also reported before in other peptide resins [33–35]. The results showed mass transfer resistance was more obvious at high flow rates [20]. 3.5. Adsorption kinetics Since Ac-FYHE-4FF resin was screened for the further research, the hIgG adsorption kinetic curves of the resin were shown in Fig. 5, which indicated that the calculated effective pore diffusivity De was greatly dependent on the pH conditions. For Ac-FYHE4FF resin, De values increased firstly and then decreased with the increase of pH numbers (5.0–9.0). This resin has the highest De value (2.51 × 10−11 m2 /s) at pH 7.0. When the pH number increases to 8.0–9.0, weaker hydrophobic interactions would cause stronger surface binding resistance for pore diffusion. Moreover, higher electrostatic repulsion (both have positive charge) blocked the binding of the hIgG and the ligand, which decreased the De value to 1.24 − 1.75 × 10−11 m2 /s at pH 8.0–9.0. However, when the pH decreases to 5.0–6.0, higher electrostatic repulsion (both have negative charge) also blocked the binding of the hIgG and the ligand, which decreased the De value to 0.3 − 1.05 × 10−11 m2 /s. 3.6. Adsorptions of hIgG and BSA by the Ac-FYHE-4FF resin To get the best adsorption pH value of the Ac-FYHE-4FF resin, further research was carried out by adsorbing hIgG and BSA at different pH values shown in Fig. S4. The first curve of the Fig. S4 is breakthrough curve, the second curve is elution curve and the final curve is regeneration curve. The results showed the area of the breakthrough curve decreased firstly and then increased with the increase of pH numbers (5.0–9.0). Moreover, the minimum areas of breakthrough curves were gained at pH 6.5–7.5. Referring to BSA, the Ac-FYHE-4FF resin almost couldn’t adsorb BSA especially at pH 7.0–8.5. Subsequently, the effect of loading pH on BSA and hIgG adsorption in the column of Ac-FYHE-4FF resin was gotten by calculating the PA shown in Fig. S5. The results showed the highest PA (98.1%) was gained at pH 7.5, which also exhibited high value of PA -hIgG/PA -BSA (about 12.6) indicating high selectivity of the resin to hIgG. Therefore, pH 7.5 was selected as the adsorption condition for the further research.

Fig. 6. Separation of hIgG from artificial protein mixture by Ac-FYHE-4FF resin. (a) chromatographic curve; (b) SEC-HPLC analysis.

3.7. Elution of hIgG by the Ac-FYHE-4FF resin Since pH 7.5 has been discovered as the best condition for hIgG adsorption and BSA almost couldn’t be bound at this situation, further research was carried out by adsorbing hIgG at pH 7.5 and eluting it at different pH values to get the best elution condition. Fig. S6 shows hIgG could be eluted absolutely except at pH 5.5–6.0 by the electrostatic repulsion. Fig. S7 shows the elution efficiency of hIgG (PE ) is about 50% at pH 6.0, and increases to 88.4% at

8

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

Fig. 7. Purification of mAb from CHO supernatant by Ac-FYHE-4FF resin. (a) chromatographic curve; (b) SEC-HPLC analysis; (c) Reducing SDS-PAGE analysis.

pH 5.5. Moreover, PE is about 94.7%−99.86% at pH 5.0–3.5, which means hIgG could be well eluted from the resin. Therefore, pH 4.5 as a relatively mild elution condition [36–38] has been selected in the further research for the elution of the target protein.

3.8. Separation performance of the Ac-FYHE-4FF resin While Ac-FYHE-4FF resin could well adsorb hIgG at pH 7.5 and has high adsorption selectivity compared to the BSA, it is selected

to separate target proteins from the BSA containing feedstock, CHO supernatant and human serum. As shown in Fig. 6, Ac-FYHE-4FF resin was used to separate hIgG from BSA containing feedstock (2 mg/mL hIgG and 8 mg/mL BSA). According to Fig. 6(b), BSA was mainly present in the breakthrough curve, while hIgG was mainly present in the elution curve, and only a small amount of hIgG was detected in the breakthrough curve. The results indicated that the Ac-FYHE-4FF resin adsorbed hIgG selectively and BSA directly flowed out during the loading of the protein mixture. Moreover, there almost couldn’t detect any

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

9

Fig. 8. Purification of hIgG from human serum by Ac-FYHE-4FF resin. (a) chromatographic curve; (b) SEC-HPLC analysis; (c) Reducing SDS-PAGE analysis.

hIgG in the CIP which meant hIgG could be totally eluted at pH 4.5. The results of SEC-HPLC analysis showed the hIgG purity was more than 99.0% and the hIgG yield was about 94.4%, which was much higher than the reported peptide resins [14]. The Ac-FYHE-4FF resin was further challenged to purify mAbs from CHO cell culture supernatant and separate hIgG from human serum. Fig. 7 showed that almost all impurities flowed through the column and there was almost no mAb detected in breakthrough fraction. The purity of the mAb was as high as 94.4% and the recovery was about 93.2%. Fig. 8 also showed that almost all impurities flowed through the column, and the purity of hIgG was 94.2% with the recovery about 91.1%. The results indicated that AcFYHE-4FF resin could be well used to separate and purify mAbs compared to the peptide resins reported before [14]. For exam-

ple, Sugita et al. [39] discovered NARKFYKG and NKFRGKYK resins, which could separate mAb from cell culture medium with both purity and recovery less than 85%. Zhao et al. [26] found FWHCLDE resin which could separate hIgG from human serum with the recovery of 87% and purity of 90%. Therefore, Ac-FYHE-4FF resin with high purification and separation of mAbs has a potential application for antibody purification. 4. Conclusions The purpose of this work is to discovery a novel peptide ligand as a biomimetic affinity ligand with high adsorption capacity and selectivity for antibody purification. Owing to the critical residues of natural ligands to hIgG, a tetrapeptide library was designed and

10

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474

a novel ligand (Ac-FYHE) was screened by the molecular docking. The results of static and dynamic adsorption of the screened ligand (Ac-FYHE) have showed excellent adsorption properties to antibody, and indicated the high selectivity of antibody to BSA. It is found in further research that the hydrophobic interaction was considered as the main force of the screened resin. The high adsorption and selectivity of this resin was proven in the followed separation experiments. Ac-FYHE-4FF resin could be successfully used to separate hIgG from BSA containing feedstock and human serum and purify mAbs from CHO cell culture supernatant. With the sample loading at pH 7.5 and eluting at pH 4.5, hIgG or mAbs could be well separated with the purity of more than 99.0% and recovery of 94.4% from BSA containing feedstock, purity of 94.4% and recovery of 93.2% from CHO cell culture supernatant, purity of 94.2% and recovery of 91.1% from human serum. In general, Ac-FYHE-4FF resin exhibited several advantages for the purification of antibodies. Firstly, compared to the reported peptide resins (such as DAAG resin, HWRGWV resin and FYWHCLDE resin), Ac-FYHE-4FF resin showed high adsorption capacity and selectivity for antibody purification. Moreover, effective elution condition (pH 4.5) could avoid denaturation or aggregation of antibody under serious acidic conditions. In addition, the short peptide ligands with small molecular could enhance the chemical stability of ligands and reduce preparation cost compared to the ligands with long chains. Therefore, this novel tetrapeptide ligand (Ac-FYHE), which has been obtained and successfully applied to the antibody purification, would contribute to the promotion and application of peptide biomimetic chromatography. Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Acknowledgements This work was supported by the National Natural Science Foundation of China (No. 21576233, and No. 21878263). Supplementary material Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.chroma.2019.460474. References [1] G.W. Litman, J.P. Rast, M.J. Shamblott, R.N. Haire, M. Hulst, W. Roess, R.T. Litman, K.R. Hindsfrey, A. Zilch, C.T. Amemiya, Phylogenetic diversification of immunoglobulin genes and the antibody repertoire, Mol. Biol. Evol. 10 (1993) 60–72. [2] H. Borghaei, L. Paz-Ares, L. Horn, D.R. Spigel, M. Steins, N.E. Ready, L.Q. Chow, E.E. Vokes, E. Felip, E. Holgado, F. Barlesi, M. Kohlhaeufl, O. Arrieta, M.A. Burgio, J. Fayette, H. Lena, E. Poddubskaya, D.E. Gerber, S.N. Gettinger, C.M. Rudin, N. Rizvi, L. Crino, G.R. Blumenschein, S.J. Antonia, C. Dorange, C.T. Harbison, F.G. Finckenstein, J.R. Brahmer, Nivolumab versus docetaxel in advanced nonsquamous non-small-cell lung cancer, New Engl. J. Med. 373 (2015) 1627–1639. [3] A.C. Chan, P.J. Carter, Therapeutic antibodies for autoimmunity and inflammation, Nat. Rev. Immunol. 10 (2010) 301–316. [4] T.H.O. Munnink, M.J. Henstra, L.I. Segerink, K.L.L. Movig, P. Brummelhuis-Visser, Therapeutic drug monitoring of monoclonal antibodies in inflammatory and malignant disease: translating TNF-alpha experience to oncology, Clin. Pharmacol. Ther. 99 (2016) 419–431. [5] Y.-M. Fang, D.-Q. Lin, S.-J. Yao, Adsorption of IGG and BSA on two chromatographic resins-poly(ethylenimine)-4ff resin and tetrapeptide-poly(ethylenimine)-4ff resin, J. Chem. Eng. Data 63 (2018) 4418–4424. [6] J.F. Buyel, R.M. Twyman, R. Fischer, Very-large-scale production of antibodies in plants: the biologization of manufacturing, Biotechol. Adv. 35 (2017) 458–465. [7] S.-G. Chen, T. Liu, R.-Q. Yang, D.-Q. Lin, S.-J. Yao, Preparation of copolymer– grafted mixed-mode resins for immunoglobulin G adsorption, Front. Chem. Sci. Eng. 13 (2019) 70–79.

[8] J.W. Bloom, M.F. Wong, G. Mitra, Detection and reduction of Protein-A contamination in immobilized Protein-A purified monoclonal-antibody preparations, J. Immunol. Methods 117 (1989) 83–89. [9] P. Fuglistaller, Comparison of immunoglobulin binding-capacities and ligand leakage using 8 different Protein-A affinity-chromatography matrices, J. Immunol. Methods 124 (1989) 171–177. [10] G. Hale, A. Drumm, P. Harrison, J. Phillips, Repeated cleaning of Protein-A affinity column with sodium-hydroxied, J. Immunol. Methods 171 (1994) 15–21. [11] R.-Z. Wang, D.-Q. Lin, W.-N. Chu, Q.-L. Zhang, S.-J. Yao, New tetrapeptide ligands designed for antibody purification with biomimetic chromatography: molecular simulation and experimental validation, Biochem. Eng. J. 114 (2016) 194–204. [12] Q. Tan, D. Dong, L. Ye, C. Huo, F. Huang, R. Li, Pre-fractionation of rat liver cytosol proteins prior to mass spectrometry-based proteomic analysis using tandem biomimetic affinity chromatography, J. Mol. Recognit. 23 (2010) 93– 100. [13] C.R. Lowe, S.J. Burton, N.P. Burton, W.K. Alderton, J.M. Pitts, J.A. Thomas, Designer dyes - biomimetic ligands for the purification of pharmaceutical proteins by affinity-chromatography, Trends Biotechnol. 10 (1992) 442–448. [14] Y.-M. Fang, D.-Q. Lin, S.-J. Yao, Review on biomimetic affinity chromatography with short peptide ligands and its application to protein purification, J. Chromatogr A. 1571 (2018) 1–15. [15] H.-L. Lu, D.-Q. Lin, S.-J. Yao, Affinity biomimetic chromatography and its applications for antibody purification, CIESC J. 67 (2016) 3523–3535. [16] H. Yang, P.V. Gurgel, R.G. Carbonell, Hexamer peptide affinity resins that bind the Fc region of human immunoglobulin G, J. Peptide Res. 661 (2005) 120– 137. [17] H. Yang, P.V. Gurgel, R.G. Carbonell, Purification of human immunoglobulin G via Fc-specific small peptide ligand affinity chromatography, J. Chromatogr. A 1216 (2009) 910–918. [18] H. Yang, P.V. Gurgel, D.K. Williams Jr, B.G. Bobay, J. Cavanagh, D.C. Muddiman, R.G. Carbonell, Binding site on human immunoglobulin G for the affinity ligand HWRGWV, J. Mol. Recognit. 23 (2010) 271–282. [19] H.-F. Tong, D.-Q. Lin, Q.-L. Zhang, R.-Z. Wang, S.-J. Yao, Molecular recognition of Fc-specific ligands binding onto the consensus binding site of IgG: insights from molecular simulation, J. Mol. Recognit. 27 (2014) 501–509. [20] R.-Z. Wang, D.-Q. Lin, H.-F. Tong, S.-J. Yao, Molecular insights into the binding selectivity of a synthetic ligand DAAG to Fc fragment of IgG, J. Mol. Recognit. 27 (5) (2014) 250–259. [21] W.D. Chen, X.Y. Dong, S. Bai, Y. Sun, Dependence of pore diffusivity of protein on adsorption density in anion-exchange adsorbent, Biochem. Eng. J. 14 (2003) 45–50. [22] W.D. Chen, X.Y. Dong, Y. Sun, Analysis of diffusion models for protein adsorption to porous anion-exchange adsorbent, J. Chromatogr. A 962 (2002) 29–40. [23] H. Yoshida, M. Yoshikawa, T. Kataoka, Parallel transport of BSA by surface and pore diffusion in strongly basic chitosan, Aiche J. 40 (1994) 2034–2044. [24] L.N. Lund, P. Gustavsson, R. Michael, J. Lindgren, L. Norskov-Lauritsen, M. Lund, G. Houen, A. Staby, P.M. St. Hilaire, Novel peptide ligand with high binding capacity for antibody purification, J Chromatogr. A 1225 (2012) 158–167. [25] Y. Liu, S. Teng, L. Yang, K. Zhou, B. Zhang, C. Li. Purifying a human intravenous immunoglobulin by dissolving precipitated component in water for injection, filtering, regulating pH of filtrate using acetic acid solution, and carrying out strong anion exchange gel chromatography (2018). [26] W. Zhao, Q. Shi, Y. Sun, FYWHCLDE-based affinity chromatography of IgG: effect of ligand density and purifications of human IgG and monoclonal antibody, J. Chromatogr. A 1355 (2014) 107–114. [27] M.E. Lienqueo, A. Mahn, J.C. Salgado, J.A. Asenjo, Current insights on protein behaviour in hydrophobic interaction chromatography, J. Chromatogr. B 849 (2007) 53–68. [28] F. Xia, D. Nagrath, S. Garde, S.M. Cramer, Evaluation of selectivity changes in HIC systems using a preferential interaction based analysis, Biotechnol. Bioeng. 87 (2004) 354–363. [29] H.-F. Tong, D.-Q. Lin, X.-M. Yuan, S.-J. Yao, Enhancing IgG purification from serum albumin containing feedstock with hydrophobic charge-induction chromatography, J. Chromatogr. A 1244 (2012) 116–122. [30] J. Gu, H. Tong, D. Lin, Evaluation of magnetic particles modified with a hydrophobic charge-induction ligand for antibody capture, J. Chromatogr. A 1460 (2016) 61–67. [31] W. Zhao, F. Liu, Q. Shi, X. Dong, Y. Sun, Biomimetic design of affinity peptide ligands for human IgG based on protein A-IgG complex, Biochem. Eng. J. 88 (2014) 1–11. [32] H.-L. Lu, D.-Q. Lin, S.-J. Yao, Chromatographic technology in antibody purification and its progress, CIESC J. 69 (2018) 341–351. [33] S. Menegatti, M. Hussain, A.D. Naik, R.G. Carbonell, B.M. Rao, MRNA display selection and solid-phase synthesis of Fc-binding cyclic peptide affinity ligands, Biotechnol. Bioeng. 110 (2013) 857–870. [34] T. Sugita, M. Katayama, M. Okochi, R. Kato, T. Ichihara, H. Honda, Screening of peptide ligands that bind to the Fc region of IgG using peptide array and its application to affinity purification of antibody, Biochem. Eng. J. 79 (2013) 33–40. [35] W.S. Kish, A.D. Naik, S. Menegatti, R.G. Carbonell, Peptide-Based affinity adsorbents with high binding capacity for the purification of monoclonal antibodies, Ind. Eng. Chem. Res. 52 (2013) 8800–8811. [36] H.-F. Tong, D.-Q. Lin, W.-N. Chu, Q.-L. Zhang, D. Gao, R.-Z. Wang, S.-J. Yao, Multimodal charge-induction chromatography for antibody purification, J. Chromatogr. A 1429 (2016) 258–264.

Y.-M. Fang, S.-G. Chen and D.-Q. Lin et al. / Journal of Chromatography A 1604 (2019) 460474 [37] J. Qian, G. El Khoury, H. Issa, K. Al-Qaoud, P. Shihab, C.R. Lowe, A synthetic Protein G adsorbent based on the multi-component Ugi reaction for the purification of mammalian immunoglobulins, J. Chromatogr. B 898 (2012) 15–23. [38] L.N. Lund, P. Gustavsson, R. Michael, J. Lindgren, L. Norskov-Lauritsen, M. Lund, G. Houen, A. Staby, P.M. St. Hilaire, Novel peptide ligand with high binding capacity for antibody purification, J. Chromatogr. A 1225 (2012) 158–167.

11

[39] T. Sugita, M. Katayama, M. Okochi, R. Kato, T. Ichihara, H. Honda, Screening of peptide ligands that bind to the Fc region of IgG using peptide array and its application to affinity purification of antibody, Biochem. Eng. J. 79 (2013) 33–40.