Protein Expression and Purification xxx (2015) xxx–xxx
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus Hao Li a,b,1, Yanli Yang a,b,1, Yan Zhang a, Songping Zhang a, Qizu Zhao c, Yuanyuan Zhu c, Xingqi Zou c, Mengran Yu a,b, Guanghui Ma a, Zhiguo Su a,⇑ a b c
National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, PR China University of Chinese Academy of Sciences, Beijing 100049, PR China China Institute of Veterinary Drug Control, Beijing 100081, PR China
a r t i c l e
i n f o
Article history: Received 8 April 2015 and in revised form 28 April 2015 Available online xxxx Keywords: Foot-and-mouth disease virus Hydrophobic interaction chromatography Purification Downstream processing
a b s t r a c t A purification scheme based on hydrophobic interaction chromatography was developed to separate inactivated foot-and-mouth disease virus (FMDV) from crude supernatant. About 92% recovery and 8.8-fold purification were achieved on Butyl Sepharose 4FF. Further purification on Superdex 200 resulted in another 29-fold purification, with 92% recovery. The columns were coupled through an intermediate ultrafiltration unit to concentrate the virus. The entire process was completed in about 3.5 h, with 75% final FMDV recovery, and 247-fold purification. The final product had purity above 98%, with over 99.5% of host cell DNA removed. High-performance size exclusion chromatography (HPSEC), Western blot, dynamic light scattering (DLS), and transmission electron microscopy (TEM) indicated that the purified virus contained the required antigen, and was structurally intact with a spherical shape and a particle size of 28 nm. Ó 2015 Elsevier Inc. All rights reserved.
Introduction Foot-and-mouth disease is an acute and contagious disease of cloven-hoofed livestock such as pig, sheep, and cattle [1]. Recently, widespread breakouts were reported in the United Kingdom, Argentina, and Uruguay, with serious consequences on agriculture and international trade [2–4]. New vaccines, for example, the self-assembling capsid protein of the foot-and-mouth disease virus (FMDV)2, have been developed [5–7]. Nevertheless, vaccines prepared from inactivated FMDV remains the most effective [2]. FMDV is a spherical particle consisting of one molecule of single-stranded positive-sense RNA and a protein capsid [8]. The capsid contains four structural proteins: VP1, VP2, and VP3 with molecular weights between 25 and 34 kDa, and VP4 of 8–10 kDa [8]. Typically, FMDV is produced in baby hamster kidney cells, separated from cell debris, and then inactivated [9,10]. It remains ⇑ Corresponding author. Tel.: +86 10 62561817. E-mail address: [email protected] (Z. Su). Both authors contributed equally to this work. 2 Abbreviations used: FMDV, foot-and-mouth disease virus; IEC, ion-exchange chromatography; HIC, hydrophobic interaction chromatography; SEC, size exclusion chromatography; HPSEC, high-performance size exclusion chromatography; TEM, Transmission electron microscopy; DLS, dynamic light scattering; HBsAg, hepatitis B surface antigen; DRT, dimensionless residence time. 1
quite common to formulate the vaccine from this relatively crude inactivated antigen [9]. However, as the public has become increasingly concerned with the safety and quality of livestock products, regulatory agencies are considering raising the quality standard for FMDV vaccines. Therefore, more stringent purification of inactivated FMDV may become necessary to remove host cell proteins and DNA, which may cause unwanted side effects [11]. FMDV has been purified by PEG precipitation [10,12], sodium sulfate precipitation [13], ultrafiltration [13,14], and aqueous two-phase partition [13]. These methods are useful, but not adequate to achieve high purity [15]. High purity has been achieved in the laboratory by ultracentrifugation in a sucrose density gradient [2,12]. However, ultracentrifugation is unsuitable for industrial-scale production, because it suffers from low capacity as well as high capital and operational cost. Chromatographic techniques may provide an alternative strategy, on the basis of high selectivity and scalability [16]. Indeed, many vaccines, including influenza virus and recombinant hepatitis B surface antigen (HBsAg), have been successfully produced through chromatography [17,18]. In addition, size exclusion chromatography (SEC) was recently used as an analytical tool to quantify inactivated FMDV [2]. Nevertheless, we are not aware of any report that describes a complete, effective program to purify inactivated FMDV by chromatography.
http://dx.doi.org/10.1016/j.pep.2015.04.011 1046-5928/Ó 2015 Elsevier Inc. All rights reserved.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
2
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
A classic chromatographic technique is ion-exchange chromatography (IEC), which is often used to purify antigens such viruses [19,20] and virus-like particles [21,22]. However, interaction with IEC media may sometimes alter the structure of purification targets, with disastrous results. For example, HBsAg disassembles during IEC, resulting in substantial loss [23,24]. For some viruses like FMDV, structural integrity is critical. The virus, with sedimentation coefficient 146S when live, is highly unstable and tends to dissociate into smaller particles with sedimentation coefficient 12S [25]. The immunogenicity of 12S particles is extremely low and unprotective [26]. What is worse, the virus was found more unstable after being inactivated, even though the inactivating agents like binary ethyleneimine were thought to only alter the viral RNA without effects on the capsid proteins [27]. Therefore, any chromatographic separation should preserve the intact FMDV structure. In this paper, we describe our efforts to purify inactivated FMDV through chromatography. We initially attempted to use IEC to purify FMDV from cell culture supernatant. However, it was difficult to achieve both satisfactory recovery and purification fold even after screening of IEC media and separation conditions. In addition, desalting by dialysis or dilution was typically required as an extra preliminary step to reduce the salt concentration in the initial crude. Thus, we evaluated hydrophobic interaction chromatography (HIC) as an alternative, which separates molecules based on hydrophobicity instead of charge. The technique has never been used to purify FMDV; indeed, HIC is far less frequently used than IEC in vaccine preparation. Nevertheless, recovery of over 98% was achieved when HBsAg was purified by HIC, more than twice the recovery from IEC [28]. In addition, HIC may be more appropriate than IEC for inactivated viruses, the structure of which may be sensitive to ionic interactions with IEC media. The final purification scheme combines HIC with ultrafiltration and SEC. The scheme was optimized by screening various chromatographic media and separation conditions. High-performance size exclusion chromatography (HPSEC), Western blotting, dynamic light scattering (DLS), and transmission electron microscopy (TEM) were used to evaluate the purified product.
Materials and methods Materials All reagents were analytical grade, and solutions were prepared using Milli-Q water (Millipore, USA). Separations were executed on an ÄKTA Purifier 100 from GE Healthcare (USA). HIC, IEC, and SEC media were obtained from GE Healthcare (Uppsala, Sweden). For HIC, we evaluated Butyl-S Sepharose 6FF, Butyl Sepharose 4FF, Octyl Sepharose 4FF and Phenyl Sepharose 6FF (high sub). DEAE-Sepharose FF and ANX Sepharose 4FF (high sub) were evaluated for ion-exchange. SEC was performed on chromatography media with different separation range, including Sepharose 4FF, Sepharose 6FF and Superdex 200. Chromatography columns (200 16 mm I.D., 1000 26 mm I.D.) were from GE Healthcare (USA).
Virus FMDV strain O China 1999 was propagated at industrial scale in BHK-21 cell suspension cultures and inactivated by binary ethyleneimine [2]. Cell debris was removed, and the supernatant containing virus particles was collected. The FMDV supernatant was originally obtained from Lanzhou Veterinary Research Institute
(Chinese Academy of Agricultural Sciences, China) after cultivation and pretreatment as described above. Evaluation of hydrophobic interaction chromatography media To evaluate the suitability of hydrophobic interaction to separate FMDV, we first analyzed the media by static adsorption with increasing hydrophobicity, including Butyl-S Sepharose 6FF, Butyl Sepharose 4FF, Octyl Sepharose 4FF and Phenyl Sepharose 6FF (high sub). The effect of ionic strength was studied using 0–1 M (NH4)2SO4. Generally, about 0.2 g of drained media, pre-equilibrated in an appropriate buffer at pH 8.0 with a defined concentration of (NH4)2SO4, was mixed with 0.8 mL FMDV suspension that had been pre-diluted with 5.6 mL buffer. After static adsorption for 24 h at 25 °C on a shaking incubator at 170 rpm, the supernatant was analyzed for remaining FMDV and proteins. Purification by hydrophobic interaction chromatography and ionexchange chromatography For chromatographic purification by HIC, (NH4)2SO4 was added to 125 mL FMDV supernatant in amounts required to achieve a target concentration, and the resulting mixture was gently stirred until the salt dissolved. Then the supernatant was loaded onto a HIC column (50 16 mm I.D.) individually packed with the HIC media pre-equilibrated with phosphate buffer (pH 8.0) containing a desired concentration of (NH4)2SO4. FMDV was eluted stepwise with decreasing (NH4)2SO4 concentration. Fractions were collected and analyzed. To compare purification efficiency with HIC, FMDV was also purified on IEC media, including DEAE-Sepharose FF and ANX Sepharose 4FF (high sub). Thus, 125 mL FMDV was diluted with 20 mM phosphate buffer pH 8.0 until the ionic strength dropped below 4 mS/cm. Suspensions were then passed through IEC media packed in 50 16 mm I.D. columns and pre-equilibrated with 20 mM phosphate buffer pH 8.0. Samples were eluted stepwise with increasing NaCl concentration. As before, fractions were collected and analyzed. All experiments were executed on ÄKTA Purifier 100 at 2.0 mL/min. Residence time is presented as dimensionless residence time (DRT), which is defined as DRT = RT/(CV/Flow rate), where RT is residence time and CV is column volume. Concentration of HIC-purified FMDV FMDV fractions from HIC were concentrated by pump filtration (Masterflex L/S, USA) through a PES membrane with molecular weight cut-off 100 kDa (VivaFlow50, Sartorius, Germany). Briefly, about 60 mL FMDV fractions were concentrated to about 20 mL at room temperature and 130 mL/min. The membrane was flushed between samples using 0.2 M NaOH at pressure less than 1 bar until flux returned to the initial value. Size exclusion chromatography Size exclusion chromatography was used to refine purification products. SEC columns with different separation ranges, namely Sepharose 4FF, Sepharose 6FF, and Superdex 200, were evaluated. Samples of 5–10 mL concentrated, HIC-purified FMDV was loaded on a column packed with one of the three media (900 26 mm I.D.), followed by elution with pH 8.0 phosphate buffer containing 0.15 M NaCl. The flow rate was 3.0 mL/min controlled by ÄKTA purifier 100 and the detection wavelength was set to 259 nm. Residence time was converted to dimensionless residence time as described.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
3
Fig. 1. The effect of (NH4)2SO4 concentration on binding of FMDV to Butyl-S Sepharose 6FF (A), Butyl Sepharose 4FF (B), Octyl Sepharose 4FF (C) and Phenyl Sepharose 6FF (high sub) (D).
HPSEC quantification of intact FMDV Intact FMDV was quantified by high-performance size exclusion chromatography (HPSEC) as described previously [29]. The analysis was performed on a TSK G4000 SWXL (300 7.8 mm I.D.) analytical column (Tosohaas, Stuttgart, Germany) using an Agilent 1100 HPLC series system (Agilent, USA) with an inline degasser and variable-wavelength detector operating at 259 nm. Samples of 100 lL were injected and eluted at 0.6 mL/min with 50 mM phosphate buffer pH 7.2 containing 100 mM Na2SO4. A standard curve for intact FMDV was constructed using preparations purified by sucrose gradient ultracentrifugation according to published methods [2,12]. The concentration of the standard was calculated from UV absorbance, using extinction coefficient
Fig. 2. Elution profile of FMDV from a 50 16 mm I.D. column of Butyl Sepharose 4FF. A sample of 125 mL crude FMDV was loaded at 0.8 M (NH4)2SO4 and 20 mM phosphate, and then eluted in one step to 0.6 M at a flow rate of 2 mL/min.
E1% 259;1cm = 76 [30]. A dilution series of the standard was then analyzed by HPSEC. The standard eluted as a single peak at 13.4 min, as previously reported [29]. The peak area at 259 nm was linearly proportional to FMDV concentration, with R2 = 0.998 over the tested range between 0 and 60.2 lg/mL. Thus, the FMDV concentration in samples was obtained from the peak area at 259 nm. Quantification of residual host cell protein and DNA
Table 1 Purification efficiency on HIC and IEC.
a
Media
Protein recoverya (%)
FMDV recoveryb (%)
Purification foldc
Butyl Sepharose 4FF Phenyl Sepharose 6FF DEAE Sepharose FF ANX Sepharose 4FF
12
89
7.2
8.2
26
3.2
13 12
69 68
5.2 5.4
Protein recovery is the ratio between proteins eluted after chromatography and the total protein loaded, as determined by Bradford assay. b FMDV recovery is the ratio between total FMDV eluted from chromatographic columns and total FMDV loaded, as measured by HPSEC. c Purification fold is the ratio between FMDV and proteins recovered.
Protein concentration was determined according to a modified Bradford method, as described elsewhere [31]. Bovine serum albumin was used as reference standard. Host cell DNA was determined by selectively staining double stranded DNA using Quant-iT™ PicoGreenÒ reagent and kits (Invitrogen), following the manufacturer’s instructions. A microplate reader (Varioskan flash, Thermo scientific, USA) was used to analyze samples at excitation and emission wavelengths 480 nm and 520 nm, respectively. Western blot, dynamic light scattering and transmission electron microscopy Purified FMDV was analyzed by Western blot as described elsewhere [29], except that we used a secondary antibodies labeled
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
4
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
Table 2 Separation efficiency on different SEC columns.
a
Column
Separation range (kDa)
Resolution
Peak Widtha (DRT)
Sepharose 4FF Sepharose 6FF Superdex 200
60–20,000 10–4000 10–600
0.71 0.89 2.96
0.29 0.26 0.05
Peak width is dimensionless residence time, as defined in the text.
with Alexa FluorÒ 790 (Jackson ImmunoResearch, USA). Following the manufacturer’s protocol, fluorescence was measured at 800 nm using the OdysseyÒ imager (Li-Cor, USA). The size distribution of purified FMDV was determined by dynamic light scattering (DLS) in a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK). Samples were filtered through a pre-rinsed 0.22-lm filter, followed by equilibration to 25 °C. A minimum of three measurements were collected from each sample, and results are expressed in number-weighted percentage distribution. Finally, the structure of the purified virus was analyzed by transmission electron microscopy using a Philips FEI Tecnai 20 TEM (Royal Philips Electronics, Amsterdam). Samples obtained by chromatography was spotted on a 400-mesh copper grid, dried, stained with 1% uranyl acetate and visualized under TEM. Results and discussion Characterization of the FMDV crude The crude supernatant obtained from industrial-scale cell culture at Lanzhou Veterinary Research Institute was characterized after virus inactivation and separation from cell debris. The pH of the FMDV supernatant was about 8.6, with ionic strength about 17.7 mS/cm. The suspension contained 0.4 mg/mL total protein, but only 1.9 lg/mL FMDV, indicating that the initial virus pool is contaminated with many unwanted proteins. Purification of FMDV by hydrophobic interaction chromatography Recombinant hepatitis B surface antigen was successfully purified with high recovery through HIC [17,28]. To test if HIC is also suitable to purify FMDV, as well as to optimize separation parameters, several HIC media were first evaluated by static adsorption
Fig. 3. Elution of HIC-purified FMDV from a 900 26 mm I.D. Superdex 200 column. HIC-purified FMDV was concentrated by ultrafiltration to 7 mL and passed through SEC at 3 mL/min using phosphate buffer containing 0.15 M NaCl.
Fig. 4. SDS–PAGE of samples from each step of purification. Lane 1, crude FMDV; 2, HIC-purified virus; 3, retentate in ultrafiltration; 4, SEC-purified FMDV. The identity of the product from SEC was verified by Western blot against VP1, shown as lane 5.
over a range of ionic strength (Fig. 1). In most cases, FMDV binding increased with (NH4)2SO4 concentration, until complete binding is achieved near 1.0 M (NH4)2SO4. However, less than 60% of the virus bound to Butyl-S Sepharose 4FF even at this concentration, suggesting that this media is not hydrophobic enough to adsorb FMDV efficiently. Indeed, FMDV binding increased with the hydrophobicity of the media. Thus, more than 90% of FMDV bound to Phenyl Sepharose 6FF in 0.6 M (NH4)2SO4, while only about 25% bound to Butyl Sepharose 4FF. Nevertheless, the virus bound only weakly to Octyl Sepharose 4FF, even though it is more hydrophobic than Butyl Sepharose 4FF. One possible explanation is that impurities displace FMDV from Octyl Sepharose 4FF. Conversely, interaction between impurities and Butyl Sepharose 4FF may also be too weak to interfere with FMDV binding. On the other hand, interaction between FMDV and Phenyl Sepharose 6FF may be sufficiently strong to overcome interference by contaminants. Taken together, these results suggest that HIC purification may be achieved by loading FMDV in high salt, and then eluting in low salt. To maximize recovery, we selected the concentration step over which FMDV binding changed most dramatically. This step is 0.8–0.6 M for Butyl Sepharose 4FF, and 0.6–0.4 M for Phenyl Sepharose 6FF. The chromatographic profile by Butyl Sepharose 4FF is shown in Fig. 2. Fraction A contains impurities flowing through the column, but not FMDV. The UV absorbance of fraction A was high, but the protein recovery was only 16%. DNA determination showed more than 80% of the DNA in supernatant flew through the column in fraction A, which is because of the weak hydrophobicity of nucleic acids. As would have been expected from the initial binding experiments, most FMDV was eluted in fraction B, which was collected at 0.6 M (NH4)2SO4, leading to 89% FMDV recovery with a purification fold of 7.2. Residual bound FMDV as well as contaminants were eluted in fraction C. Notably, when FMDV was eluted from Phenyl Sepharose 6FF over several step gradients, resulting in low fold-purification of 3.2 and low FMDV recovery of only 26%. Thus, FMDV is simply too strongly bound to this media to enable enrichment. As seen from Fig. 1(D), the binding of FMDV to Phenyl Sepharose 6FF was extremely weak at low salt concentration. Therefore, another strategy by passing the supernatant through Phenyl column at low concentration of (NH4)2SO4 might be possible to obtain a high FMDV recovery in breakthrough fraction while
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
5
Fig. 5. HPSEC analyses of crude (A), HIC-purified (B), concentrated (C) and SEC-purified (D) FMDV. Samples of 100 lL were analyzed at 0.6 mL/min on TSK G4000 SWxL (300 7.8 mm, I.D.) in 50 mM phosphate buffer pH 7.2. Elution was monitored by UV absorbance at 259 nm.
Table 3 Purification efficiency at each step. Sample
FMDV 100 HIC UF SEC Total
Total proteina (mg) crude 21.8 ± 1.8 20.0 ± 1.2 0.64 ± 0.07
FMDV
Residual host cell DNAc (%)
Amountb (lg)
Recovery (%)
Purification fold
208
940
100
1
864 ± 48 770 ± 62 709 ± 38
92 ± 5 89 ± 2 92 ± 2 75 ± 4
8.8 ± 0.2 1.0 ± 0.1 28.8 ± 1.8 247 ± 11
3.3 ± 0.3 2.3 ± 0.8 0.5 ± 0.1
Errors are from duplicate experiments. a Total protein was determined by Bradford assay. b Total FMDV was measured by HPSEC. c Residual host cell DNA is the ratio between DNA remaining after each step and total DNA in the initial crude. DNA concentration was determined by fluorimetry.
removing most of the non-virus protein by adsorption. However, the host cell DNA would also probably breakthrough the column together with FMDV, and thus another step is needed to remove it. Moreover, since FMDV cannot be concentrated by breakthrough strategy, additional concentration step and more time may be required for subsequent processing. Therefore, Butyl Sepharose 4FF was deemed more suitable than Phenyl Sepharose 6FF for FMDV purification.
Comparison of FMDV purification by IEC and HIC Purification efficiency on Butyl Sepharose 4FF and Phenyl Sepharose 6FF is summarized in Table 1, along with purification
efficiency on DEAE Sepharose FF and ANX Sepharose 4FF. The last two media were screened from several of frequently-used IEC media in previous work, with relative good purification efficiency. As seen from the table, the highest FMDV recovery and fold-purification are highest on Butyl Sepharose 4FF. Recovery from DEAE Sepharose FF and ANX Sepharose 4FF was acceptable, but IEC requires preliminary dilution or desalting of the crude, and therefore requires more time. Therefore, Butyl Sepharose 4FF was retained as the method of choice. Refinement by size exclusion chromatography FMDV eluted from Butyl Sepharose 4FF HIC was polished by SEC to remove residual contaminants. We evaluated SEC columns with separation range beyond, close to, and below 5500 kDa, the approximate molecular weight of FMDV [29]. Separation efficiency is listed in Table 2. SEC with Superdex 200 produced the sharpest FMDV peak, with excellent resolution of 2.96. The virus is excluded from Superdex 200, and was eluted in the exclusion volume while impurities were eluted much later (Fig. 3), indicating that impurities were of low molecular weight. In contrast, both FMDV and impurities are fractionated over Sepharose 4FF and Sepharose 6FF. Unfortunately, separation was poor on both, with resolution below 1. Therefore, Superdex 200 column was selected to polish FMDV preparations, even though preparative SEC often suffers from limited efficiency at industrial scale [32]. However, in this particular case, separation on SEC in ‘‘negative mode’’ is advantageous, and produces high-purity and high-concentration products, even in industrial scale.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
6
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
Fig. 6. DLS analysis (A) and TEM imaging (B) of the purified FMDV.
Purification efficiency of the entire scheme FMDV may thus be purified in three steps, which include HIC on Butyl Sepharose 4FF, concentration by ultrafiltration, and SEC on Superdex 200. To verify efficiency and test reproducibility, FMDV purification was repeated three times. The initial crude was also scaled up from 125 mL to 500 mL. The size of the HIC column was increased accordingly, although ultrafiltration and SEC parameters were not changed. After purification, the identity of the final FMDV product was verified by Western blot (Fig. 4), with purity above 98%, as determined by SDS–PAGE (Fig. 4) and HPSEC (Fig. 5D). The major protein impurity in the initial crude was serum albumin, which showed a molecular weight near 70 kDa as shown in Fig. 4. The serum albumin was the main culture medium which is essential for cell growth. After HIC, the serum albumin was reduced, and then was completely removed by SEC. Host cell DNA was efficiently reduced to 0.5% after purification, with most being removed at the first HIC step. The purification efficiency of each step is shown in Table 3, with recovery over 90% after HIC and SEC. The average total recovery was 75%, with 247-fold increase in purity. The standard deviation from duplicate experiments was less than 6%, suggesting that the process is reproducible. On balance, the performance of this scheme is sufficient for industrial-scale production.
Structure of purified FMDV The structure of purified FMDV was analyzed by DLS and TEM. Samples exhibit a single DLS peak with diameter 28.0 nm and polydispersity index 0.21 (Fig. 6A), indicating monodispersity. This result is in good agreement with data from Ravindra Acharya [33]. Based on TEM imaging (Fig. 6B), the majority of purified virus are spherical, with calculated mean diameter 28 nm. These results confirm that purified FMDV is structurally intact, and thus appropriate as vaccine.
Conclusions Hydrophobic interaction chromatography (HIC) is suitable to purify inactivated FMDV. Of the HIC media evaluated, Butyl Sepharose 4FF was found to perform best to achieve recovery about 92% with 8.8-fold purification. HIC is more efficient than IEC by eliminating the need for preliminary desalting and reducing processing time.
Superdex 200 was used to polish FMDV preparations, because it was the most efficient among the SEC media evaluated. On this column, the virus was excluded from the matrix, and thus separated from contaminants, resulting in resolution near 3, 92% recovery, and 28-fold purification. However, optimum and efficient SEC purification requires an intermediate concentration step by ultrafiltration. On balance, 75% of FMDV could be reliably recovered from crude, and purified 247-fold to purities above 98%. More than 99% of host cell DNA was removed. Western blot, HPSEC, DLS and TEM indicate that the final product contained the required antigen, and was structurally intact with spherical shape and particle size 28 nm. Thus, our work demonstrates that HIC could be useful in purifying inactivated viruses whose structural integrity may be sensitive to industrial processing. Acknowledgments Financial supports from the Natural Sciences Foundation of China (Nos. 21336010, 21306207), National High Technology Research and Development Program of China (863 Program, Nos. 2012AA02A406 and 2014AA021006), National Key Scientific Instrument and Equipment Development Project (No. 2013YQ14040508), and Special Fund for Agro-scientific Research in the Public Interest (No. 201303046) are acknowledged. References [1] Z.Y. Li, J.X. Liu, The current state of vaccines used in the field for foot and mouth disease virus in China, Expert Rev. Vaccines 10 (2011) 13–15. [2] M.A. Spitteler, I. Fernandez, E. Schabes, A. Krimer, E.G. Regulier, M. Guinzburg, E. Smitsaart, M.S. Levy, Foot and mouth disease (FMD) virus: quantification of whole virus particles during the vaccine manufacturing process by size exclusion chromatography, Vaccine 29 (2011) 7182–7187. [3] S. Goodwin, T.J. Tuthill, A. Arias, R.A. Killington, D.J. Rowlands, Foot-and-mouth disease virus assembly: processing of recombinant capsid precursor by exogenous protease induces self-assembly of pentamers in vitro in a myristoylation-dependent manner, J. Virol. 83 (2009) 11275–11282. [4] P. Kitching, J. Hammond, M. Jeggo, B. Charleston, D. Paton, L. Rodriguez, R. Heckert, Global FMD control—is it an option?, Vaccine 25 (2007) 5660–5664 [5] C. Porta, X.D. Xu, S. Loureiro, S. Paramasivam, J.Y. Ren, T. Al-Khalil, A. Burman, T. Jackson, G.J. Belsham, S. Curry, G.P. Lomonossoff, S. Parida, D. Paton, Y.M. Li, G. Wilsden, N. Ferris, R. Owens, A. Kotecha, E. Fry, D.I. Stuart, B. Charleston, I.M. Jones, Efficient production of foot-and-mouth disease virus empty capsids in insect cells following down regulation of 3C protease activity, J. Virol. Methods 187 (2013) 406–412. [6] B.M. Subramanian, M. Madhanmohan, R. Sriraman, R.V.C. Reddy, S. Yuvaraj, K. Manikumar, S. Rajalakshmi, S.B. Nagendrakumar, S.K. Rana, V.A. Srinivasan, Development of foot-and-mouth disease virus (FMDV) serotype O virus-likeparticles (VLPs) vaccine and evaluation of its potency, Antivir. Res. 96 (2012) 288–295.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx [7] Z.Y. Li, Y.Z. Yi, X.P. Yin, Y. Zhang, M. Liu, H. Liu, X.R. Li, Y.N. Li, Z.F. Zhang, J.X. Liu, Development of a foot-and-mouth disease virus serotype A empty capsid subunit vaccine using silkworm (Bombyx mori) pupae, PLoS ONE 7 (2012) e43849. [8] H.W. van Vlijmen, S. Curry, M. Schaefer, M. Karplus, Titration calculations of foot-and-mouth disease virus capsids and their stabilities as a function of pH, J. Mol. Biol. 275 (1998) 295–308. [9] T.R. Doel, FMD vaccines, Virus Res. 91 (2003) 81–99. [10] S.J. Barteling, Development and performance of inactivated vaccines against foot and mouth disease, Rev. Sci. Tech. 21 (2002) 577–588. [11] L.L. Rodriguez, M.J. Grubman, Foot and mouth disease virus vaccines, Vaccine 27 (2009) D90–D94. [12] S.J. Barteling, R.H. Meloen, A simple method for the quantification of 140S particles of foot-and-mouth disease virus (FMDV), Arch. Ges. Virusforsch. 45 (1974) 362–364. [13] A.W. Morrow, C.J. Whittle, W.A. Eales, A comparison of methods for the concentration of foot-and-mouth disease virus for vaccine preparation, Bull. Off. Int. Epiz. 81 (1974) 1155–1167. [14] I.A. Galloway, M. Schlesinger, Purification and concentration of the virus of foot-and-mouth disease by combined centrifugation and ultrafiltration methods, J. Hyg. 37 (1937) 463–470. [15] A.A. Navarro del Canizo, M. Mazza, R. Bellinzoni, O. Cascone, Foot and mouth disease virus concentration and purification by affinity chromatography, Appl. Biochem. Biotechnol. 61 (1996) 399–409. [16] G. Iyer, S. Ramaswamy, D. Asher, U. Mehta, A. Leahy, F. Chung, K.S. Cheng, Reduced surface area chromatography for flow-through purification of viruses and virus like particles, J. Chromatogr. A 1218 (2011) 3973–3981. [17] Y. Huang, J. Bi, Y. Zhang, W. Zhou, Y. Li, L. Zhao, Z. Su, A highly efficient integrated chromatographic procedure for the purification of recombinant hepatitis B surface antigen from Hansenula polymorpha, Protein Expr. Purif. 56 (2007) 301–310. [18] B. Kalbfuss, M. Wolff, R. Morenweiser, U. Reichl, Purification of cell culturederived human influenza A virus by size-exclusion and anion-exchange chromatography, Biotechnol. Bioeng. 96 (2007) 932–944. [19] T. Weigel, T. Solomaier, A. Peuker, T. Pathapati, M.W. Wolff, U. Reichl, A flowthrough chromatography process for influenza A and B virus purification, J. Virol. Methods 207 (2014) 45–53. [20] J. Jimenez, H.K. Hammond, A. Miyanohara, T. Takahashi, D.M. Roth, N.C. Lai, Ion exchange chromatography purification of adeno-associated virus type 5, FASEB J. 19 (2005). A1540–A1540. [21] C.F. Yap, W.S. Tan, C.C. Sieo, B.T. Tey, Purification of long helical capsid of newcastle disease virus from Escherichia coli using anion exchange chromatography, Biotechnol. Prog. 29 (2013) 564–567.
7
[22] J.O. Baek, J.W. Seo, I.H. Kim, C.H. Kim, Production and purification of human papillomavirus type 33 L1 virus-like particles from Spodoptera frugiperda 9 cells using two-step column chromatography, Protein Expr. Purif. 75 (2011) 211–217. [23] M.R. Yu, Y. Li, S.P. Zhang, X.N. Li, Y.L. Yang, Y. Chen, G.H. Ma, Z.G. Su, Improving stability of virus-like particles by ion-exchange chromatographic supports with large pore size: advantages of gigaporous media beyond enhanced binding capacity, J. Chromatogr. A 1331 (2014) 69–79. [24] Y. Huang, J. Bi, L. Zhao, G. Ma, Z. Su, Regulation of protein multipoint adsorption on ion-exchange adsorbent and its application to the purification of macromolecules, Protein Expr. Purif. 74 (2010) 257–263. [25] T. Liang, D. Yang, M. Liu, C. Sun, F. Wang, J. Wang, H. Wang, S. Song, G. Zhou, L. Yu, Selection and characterization of an acid-resistant mutant of serotype O foot-and-mouth disease virus, Arch. Virol. 159 (2014) 657–667. [26] M.G. Rao, G. Butchaiah, A.K. Sen, Antibody-response to 146S particle, 12S protein subunit and isolated Vp1 polypeptide of foot-and-mouth-disease virus type Asia-1, Vet. Microbiol. 39 (1994) 135–143. [27] N.P. Ferris, A.I. Donaldson, I.T.R. Barnett, R.W. Osborne, Inactivation, purification and stability of 146S antigens of foot and mouth disease virus for use as reagents in the complement fixation test, Rev. Sci. Tech. Off. Int. Epiz. 3 (1984) 339–350. [28] W.B. Zhou, J.X. Bi, L. Zhao, Y.M. Wang, Y. Li, Y.D. Huang, G.H. Ma, Z.G. Su, A highly efficient hydrophobic interaction chromatographic absorbent improved the purification of hepatitis B surface antigen (HBsAg) derived from Hansenula polymorpha cell, Process Biochem. 42 (2007) 751–756. [29] Y. Yang, H. Li, Z. Li, Y. Zhang, S. Zhang, Y. Chen, M. Yu, G. Ma, Z. Su, Sizeexclusion HPLC provides a simple, rapid, and versatile alternative method for quality control of vaccines by characterizing the assembly of antigens, Vaccine 33 (2015) 1143–1150. [30] H.L. Bachrach, R. Trautman, S.S. Breese Jr., Chemical physical properties of virtually pure foot-and-mouth disease virus, Am. J. Vet. Res. 25 (1964) 333– 342. [31] W. Zhou, J. Bi, J.C. Janson, A. Dong, Y. Li, Y. Zhang, Y. Huang, Z. Su, Ion-exchange chromatography of hepatitis B virus surface antigen from a recombinant Chinese hamster ovary cell line, J. Chromatogr. A 1095 (2005) 119–125. [32] B. Kalbfuss, D. Flockerzi, A. Seidel-Morgenstern, U. Reichl, Size-exclusion chromatography as a linear transfer system: purification of human influenza virus as an example, J. Chromatogr. B 873 (2008) 102–112. [33] R. Acharya, E. Fry, D. Stuart, G. Fox, D. Rowlands, F. Brown, The threedimensional structure of foot-and-mouth disease virus at 2.9 Å resolution, Nature 337 (1989) 709–716.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
Contents lists available at ScienceDirect
Protein Expression and Purification journal homepage: www.elsevier.com/locate/yprep
A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus Hao Li a,b,1, Yanli Yang a,b,1, Yan Zhang a, Songping Zhang a, Qizu Zhao c, Yuanyuan Zhu c, Xingqi Zou c, Mengran Yu a,b, Guanghui Ma a, Zhiguo Su a,⇑ a b c
National Key Laboratory of Biochemical Engineering, Institute of Process Engineering, Chinese Academy of Sciences, Beijing 100190, PR China University of Chinese Academy of Sciences, Beijing 100049, PR China China Institute of Veterinary Drug Control, Beijing 100081, PR China
a r t i c l e
i n f o
Article history: Received 8 April 2015 and in revised form 28 April 2015 Available online xxxx Keywords: Foot-and-mouth disease virus Hydrophobic interaction chromatography Purification Downstream processing
a b s t r a c t A purification scheme based on hydrophobic interaction chromatography was developed to separate inactivated foot-and-mouth disease virus (FMDV) from crude supernatant. About 92% recovery and 8.8-fold purification were achieved on Butyl Sepharose 4FF. Further purification on Superdex 200 resulted in another 29-fold purification, with 92% recovery. The columns were coupled through an intermediate ultrafiltration unit to concentrate the virus. The entire process was completed in about 3.5 h, with 75% final FMDV recovery, and 247-fold purification. The final product had purity above 98%, with over 99.5% of host cell DNA removed. High-performance size exclusion chromatography (HPSEC), Western blot, dynamic light scattering (DLS), and transmission electron microscopy (TEM) indicated that the purified virus contained the required antigen, and was structurally intact with a spherical shape and a particle size of 28 nm. Ó 2015 Elsevier Inc. All rights reserved.
Introduction Foot-and-mouth disease is an acute and contagious disease of cloven-hoofed livestock such as pig, sheep, and cattle [1]. Recently, widespread breakouts were reported in the United Kingdom, Argentina, and Uruguay, with serious consequences on agriculture and international trade [2–4]. New vaccines, for example, the self-assembling capsid protein of the foot-and-mouth disease virus (FMDV)2, have been developed [5–7]. Nevertheless, vaccines prepared from inactivated FMDV remains the most effective [2]. FMDV is a spherical particle consisting of one molecule of single-stranded positive-sense RNA and a protein capsid [8]. The capsid contains four structural proteins: VP1, VP2, and VP3 with molecular weights between 25 and 34 kDa, and VP4 of 8–10 kDa [8]. Typically, FMDV is produced in baby hamster kidney cells, separated from cell debris, and then inactivated [9,10]. It remains ⇑ Corresponding author. Tel.: +86 10 62561817. E-mail address: [email protected] (Z. Su). Both authors contributed equally to this work. 2 Abbreviations used: FMDV, foot-and-mouth disease virus; IEC, ion-exchange chromatography; HIC, hydrophobic interaction chromatography; SEC, size exclusion chromatography; HPSEC, high-performance size exclusion chromatography; TEM, Transmission electron microscopy; DLS, dynamic light scattering; HBsAg, hepatitis B surface antigen; DRT, dimensionless residence time. 1
quite common to formulate the vaccine from this relatively crude inactivated antigen [9]. However, as the public has become increasingly concerned with the safety and quality of livestock products, regulatory agencies are considering raising the quality standard for FMDV vaccines. Therefore, more stringent purification of inactivated FMDV may become necessary to remove host cell proteins and DNA, which may cause unwanted side effects [11]. FMDV has been purified by PEG precipitation [10,12], sodium sulfate precipitation [13], ultrafiltration [13,14], and aqueous two-phase partition [13]. These methods are useful, but not adequate to achieve high purity [15]. High purity has been achieved in the laboratory by ultracentrifugation in a sucrose density gradient [2,12]. However, ultracentrifugation is unsuitable for industrial-scale production, because it suffers from low capacity as well as high capital and operational cost. Chromatographic techniques may provide an alternative strategy, on the basis of high selectivity and scalability [16]. Indeed, many vaccines, including influenza virus and recombinant hepatitis B surface antigen (HBsAg), have been successfully produced through chromatography [17,18]. In addition, size exclusion chromatography (SEC) was recently used as an analytical tool to quantify inactivated FMDV [2]. Nevertheless, we are not aware of any report that describes a complete, effective program to purify inactivated FMDV by chromatography.
http://dx.doi.org/10.1016/j.pep.2015.04.011 1046-5928/Ó 2015 Elsevier Inc. All rights reserved.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
2
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
A classic chromatographic technique is ion-exchange chromatography (IEC), which is often used to purify antigens such viruses [19,20] and virus-like particles [21,22]. However, interaction with IEC media may sometimes alter the structure of purification targets, with disastrous results. For example, HBsAg disassembles during IEC, resulting in substantial loss [23,24]. For some viruses like FMDV, structural integrity is critical. The virus, with sedimentation coefficient 146S when live, is highly unstable and tends to dissociate into smaller particles with sedimentation coefficient 12S [25]. The immunogenicity of 12S particles is extremely low and unprotective [26]. What is worse, the virus was found more unstable after being inactivated, even though the inactivating agents like binary ethyleneimine were thought to only alter the viral RNA without effects on the capsid proteins [27]. Therefore, any chromatographic separation should preserve the intact FMDV structure. In this paper, we describe our efforts to purify inactivated FMDV through chromatography. We initially attempted to use IEC to purify FMDV from cell culture supernatant. However, it was difficult to achieve both satisfactory recovery and purification fold even after screening of IEC media and separation conditions. In addition, desalting by dialysis or dilution was typically required as an extra preliminary step to reduce the salt concentration in the initial crude. Thus, we evaluated hydrophobic interaction chromatography (HIC) as an alternative, which separates molecules based on hydrophobicity instead of charge. The technique has never been used to purify FMDV; indeed, HIC is far less frequently used than IEC in vaccine preparation. Nevertheless, recovery of over 98% was achieved when HBsAg was purified by HIC, more than twice the recovery from IEC [28]. In addition, HIC may be more appropriate than IEC for inactivated viruses, the structure of which may be sensitive to ionic interactions with IEC media. The final purification scheme combines HIC with ultrafiltration and SEC. The scheme was optimized by screening various chromatographic media and separation conditions. High-performance size exclusion chromatography (HPSEC), Western blotting, dynamic light scattering (DLS), and transmission electron microscopy (TEM) were used to evaluate the purified product.
Materials and methods Materials All reagents were analytical grade, and solutions were prepared using Milli-Q water (Millipore, USA). Separations were executed on an ÄKTA Purifier 100 from GE Healthcare (USA). HIC, IEC, and SEC media were obtained from GE Healthcare (Uppsala, Sweden). For HIC, we evaluated Butyl-S Sepharose 6FF, Butyl Sepharose 4FF, Octyl Sepharose 4FF and Phenyl Sepharose 6FF (high sub). DEAE-Sepharose FF and ANX Sepharose 4FF (high sub) were evaluated for ion-exchange. SEC was performed on chromatography media with different separation range, including Sepharose 4FF, Sepharose 6FF and Superdex 200. Chromatography columns (200 16 mm I.D., 1000 26 mm I.D.) were from GE Healthcare (USA).
Virus FMDV strain O China 1999 was propagated at industrial scale in BHK-21 cell suspension cultures and inactivated by binary ethyleneimine [2]. Cell debris was removed, and the supernatant containing virus particles was collected. The FMDV supernatant was originally obtained from Lanzhou Veterinary Research Institute
(Chinese Academy of Agricultural Sciences, China) after cultivation and pretreatment as described above. Evaluation of hydrophobic interaction chromatography media To evaluate the suitability of hydrophobic interaction to separate FMDV, we first analyzed the media by static adsorption with increasing hydrophobicity, including Butyl-S Sepharose 6FF, Butyl Sepharose 4FF, Octyl Sepharose 4FF and Phenyl Sepharose 6FF (high sub). The effect of ionic strength was studied using 0–1 M (NH4)2SO4. Generally, about 0.2 g of drained media, pre-equilibrated in an appropriate buffer at pH 8.0 with a defined concentration of (NH4)2SO4, was mixed with 0.8 mL FMDV suspension that had been pre-diluted with 5.6 mL buffer. After static adsorption for 24 h at 25 °C on a shaking incubator at 170 rpm, the supernatant was analyzed for remaining FMDV and proteins. Purification by hydrophobic interaction chromatography and ionexchange chromatography For chromatographic purification by HIC, (NH4)2SO4 was added to 125 mL FMDV supernatant in amounts required to achieve a target concentration, and the resulting mixture was gently stirred until the salt dissolved. Then the supernatant was loaded onto a HIC column (50 16 mm I.D.) individually packed with the HIC media pre-equilibrated with phosphate buffer (pH 8.0) containing a desired concentration of (NH4)2SO4. FMDV was eluted stepwise with decreasing (NH4)2SO4 concentration. Fractions were collected and analyzed. To compare purification efficiency with HIC, FMDV was also purified on IEC media, including DEAE-Sepharose FF and ANX Sepharose 4FF (high sub). Thus, 125 mL FMDV was diluted with 20 mM phosphate buffer pH 8.0 until the ionic strength dropped below 4 mS/cm. Suspensions were then passed through IEC media packed in 50 16 mm I.D. columns and pre-equilibrated with 20 mM phosphate buffer pH 8.0. Samples were eluted stepwise with increasing NaCl concentration. As before, fractions were collected and analyzed. All experiments were executed on ÄKTA Purifier 100 at 2.0 mL/min. Residence time is presented as dimensionless residence time (DRT), which is defined as DRT = RT/(CV/Flow rate), where RT is residence time and CV is column volume. Concentration of HIC-purified FMDV FMDV fractions from HIC were concentrated by pump filtration (Masterflex L/S, USA) through a PES membrane with molecular weight cut-off 100 kDa (VivaFlow50, Sartorius, Germany). Briefly, about 60 mL FMDV fractions were concentrated to about 20 mL at room temperature and 130 mL/min. The membrane was flushed between samples using 0.2 M NaOH at pressure less than 1 bar until flux returned to the initial value. Size exclusion chromatography Size exclusion chromatography was used to refine purification products. SEC columns with different separation ranges, namely Sepharose 4FF, Sepharose 6FF, and Superdex 200, were evaluated. Samples of 5–10 mL concentrated, HIC-purified FMDV was loaded on a column packed with one of the three media (900 26 mm I.D.), followed by elution with pH 8.0 phosphate buffer containing 0.15 M NaCl. The flow rate was 3.0 mL/min controlled by ÄKTA purifier 100 and the detection wavelength was set to 259 nm. Residence time was converted to dimensionless residence time as described.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
3
Fig. 1. The effect of (NH4)2SO4 concentration on binding of FMDV to Butyl-S Sepharose 6FF (A), Butyl Sepharose 4FF (B), Octyl Sepharose 4FF (C) and Phenyl Sepharose 6FF (high sub) (D).
HPSEC quantification of intact FMDV Intact FMDV was quantified by high-performance size exclusion chromatography (HPSEC) as described previously [29]. The analysis was performed on a TSK G4000 SWXL (300 7.8 mm I.D.) analytical column (Tosohaas, Stuttgart, Germany) using an Agilent 1100 HPLC series system (Agilent, USA) with an inline degasser and variable-wavelength detector operating at 259 nm. Samples of 100 lL were injected and eluted at 0.6 mL/min with 50 mM phosphate buffer pH 7.2 containing 100 mM Na2SO4. A standard curve for intact FMDV was constructed using preparations purified by sucrose gradient ultracentrifugation according to published methods [2,12]. The concentration of the standard was calculated from UV absorbance, using extinction coefficient
Fig. 2. Elution profile of FMDV from a 50 16 mm I.D. column of Butyl Sepharose 4FF. A sample of 125 mL crude FMDV was loaded at 0.8 M (NH4)2SO4 and 20 mM phosphate, and then eluted in one step to 0.6 M at a flow rate of 2 mL/min.
E1% 259;1cm = 76 [30]. A dilution series of the standard was then analyzed by HPSEC. The standard eluted as a single peak at 13.4 min, as previously reported [29]. The peak area at 259 nm was linearly proportional to FMDV concentration, with R2 = 0.998 over the tested range between 0 and 60.2 lg/mL. Thus, the FMDV concentration in samples was obtained from the peak area at 259 nm. Quantification of residual host cell protein and DNA
Table 1 Purification efficiency on HIC and IEC.
a
Media
Protein recoverya (%)
FMDV recoveryb (%)
Purification foldc
Butyl Sepharose 4FF Phenyl Sepharose 6FF DEAE Sepharose FF ANX Sepharose 4FF
12
89
7.2
8.2
26
3.2
13 12
69 68
5.2 5.4
Protein recovery is the ratio between proteins eluted after chromatography and the total protein loaded, as determined by Bradford assay. b FMDV recovery is the ratio between total FMDV eluted from chromatographic columns and total FMDV loaded, as measured by HPSEC. c Purification fold is the ratio between FMDV and proteins recovered.
Protein concentration was determined according to a modified Bradford method, as described elsewhere [31]. Bovine serum albumin was used as reference standard. Host cell DNA was determined by selectively staining double stranded DNA using Quant-iT™ PicoGreenÒ reagent and kits (Invitrogen), following the manufacturer’s instructions. A microplate reader (Varioskan flash, Thermo scientific, USA) was used to analyze samples at excitation and emission wavelengths 480 nm and 520 nm, respectively. Western blot, dynamic light scattering and transmission electron microscopy Purified FMDV was analyzed by Western blot as described elsewhere [29], except that we used a secondary antibodies labeled
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
4
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
Table 2 Separation efficiency on different SEC columns.
a
Column
Separation range (kDa)
Resolution
Peak Widtha (DRT)
Sepharose 4FF Sepharose 6FF Superdex 200
60–20,000 10–4000 10–600
0.71 0.89 2.96
0.29 0.26 0.05
Peak width is dimensionless residence time, as defined in the text.
with Alexa FluorÒ 790 (Jackson ImmunoResearch, USA). Following the manufacturer’s protocol, fluorescence was measured at 800 nm using the OdysseyÒ imager (Li-Cor, USA). The size distribution of purified FMDV was determined by dynamic light scattering (DLS) in a Zetasizer Nano ZS (Malvern Instruments, Malvern, UK). Samples were filtered through a pre-rinsed 0.22-lm filter, followed by equilibration to 25 °C. A minimum of three measurements were collected from each sample, and results are expressed in number-weighted percentage distribution. Finally, the structure of the purified virus was analyzed by transmission electron microscopy using a Philips FEI Tecnai 20 TEM (Royal Philips Electronics, Amsterdam). Samples obtained by chromatography was spotted on a 400-mesh copper grid, dried, stained with 1% uranyl acetate and visualized under TEM. Results and discussion Characterization of the FMDV crude The crude supernatant obtained from industrial-scale cell culture at Lanzhou Veterinary Research Institute was characterized after virus inactivation and separation from cell debris. The pH of the FMDV supernatant was about 8.6, with ionic strength about 17.7 mS/cm. The suspension contained 0.4 mg/mL total protein, but only 1.9 lg/mL FMDV, indicating that the initial virus pool is contaminated with many unwanted proteins. Purification of FMDV by hydrophobic interaction chromatography Recombinant hepatitis B surface antigen was successfully purified with high recovery through HIC [17,28]. To test if HIC is also suitable to purify FMDV, as well as to optimize separation parameters, several HIC media were first evaluated by static adsorption
Fig. 3. Elution of HIC-purified FMDV from a 900 26 mm I.D. Superdex 200 column. HIC-purified FMDV was concentrated by ultrafiltration to 7 mL and passed through SEC at 3 mL/min using phosphate buffer containing 0.15 M NaCl.
Fig. 4. SDS–PAGE of samples from each step of purification. Lane 1, crude FMDV; 2, HIC-purified virus; 3, retentate in ultrafiltration; 4, SEC-purified FMDV. The identity of the product from SEC was verified by Western blot against VP1, shown as lane 5.
over a range of ionic strength (Fig. 1). In most cases, FMDV binding increased with (NH4)2SO4 concentration, until complete binding is achieved near 1.0 M (NH4)2SO4. However, less than 60% of the virus bound to Butyl-S Sepharose 4FF even at this concentration, suggesting that this media is not hydrophobic enough to adsorb FMDV efficiently. Indeed, FMDV binding increased with the hydrophobicity of the media. Thus, more than 90% of FMDV bound to Phenyl Sepharose 6FF in 0.6 M (NH4)2SO4, while only about 25% bound to Butyl Sepharose 4FF. Nevertheless, the virus bound only weakly to Octyl Sepharose 4FF, even though it is more hydrophobic than Butyl Sepharose 4FF. One possible explanation is that impurities displace FMDV from Octyl Sepharose 4FF. Conversely, interaction between impurities and Butyl Sepharose 4FF may also be too weak to interfere with FMDV binding. On the other hand, interaction between FMDV and Phenyl Sepharose 6FF may be sufficiently strong to overcome interference by contaminants. Taken together, these results suggest that HIC purification may be achieved by loading FMDV in high salt, and then eluting in low salt. To maximize recovery, we selected the concentration step over which FMDV binding changed most dramatically. This step is 0.8–0.6 M for Butyl Sepharose 4FF, and 0.6–0.4 M for Phenyl Sepharose 6FF. The chromatographic profile by Butyl Sepharose 4FF is shown in Fig. 2. Fraction A contains impurities flowing through the column, but not FMDV. The UV absorbance of fraction A was high, but the protein recovery was only 16%. DNA determination showed more than 80% of the DNA in supernatant flew through the column in fraction A, which is because of the weak hydrophobicity of nucleic acids. As would have been expected from the initial binding experiments, most FMDV was eluted in fraction B, which was collected at 0.6 M (NH4)2SO4, leading to 89% FMDV recovery with a purification fold of 7.2. Residual bound FMDV as well as contaminants were eluted in fraction C. Notably, when FMDV was eluted from Phenyl Sepharose 6FF over several step gradients, resulting in low fold-purification of 3.2 and low FMDV recovery of only 26%. Thus, FMDV is simply too strongly bound to this media to enable enrichment. As seen from Fig. 1(D), the binding of FMDV to Phenyl Sepharose 6FF was extremely weak at low salt concentration. Therefore, another strategy by passing the supernatant through Phenyl column at low concentration of (NH4)2SO4 might be possible to obtain a high FMDV recovery in breakthrough fraction while
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
5
Fig. 5. HPSEC analyses of crude (A), HIC-purified (B), concentrated (C) and SEC-purified (D) FMDV. Samples of 100 lL were analyzed at 0.6 mL/min on TSK G4000 SWxL (300 7.8 mm, I.D.) in 50 mM phosphate buffer pH 7.2. Elution was monitored by UV absorbance at 259 nm.
Table 3 Purification efficiency at each step. Sample
FMDV 100 HIC UF SEC Total
Total proteina (mg) crude 21.8 ± 1.8 20.0 ± 1.2 0.64 ± 0.07
FMDV
Residual host cell DNAc (%)
Amountb (lg)
Recovery (%)
Purification fold
208
940
100
1
864 ± 48 770 ± 62 709 ± 38
92 ± 5 89 ± 2 92 ± 2 75 ± 4
8.8 ± 0.2 1.0 ± 0.1 28.8 ± 1.8 247 ± 11
3.3 ± 0.3 2.3 ± 0.8 0.5 ± 0.1
Errors are from duplicate experiments. a Total protein was determined by Bradford assay. b Total FMDV was measured by HPSEC. c Residual host cell DNA is the ratio between DNA remaining after each step and total DNA in the initial crude. DNA concentration was determined by fluorimetry.
removing most of the non-virus protein by adsorption. However, the host cell DNA would also probably breakthrough the column together with FMDV, and thus another step is needed to remove it. Moreover, since FMDV cannot be concentrated by breakthrough strategy, additional concentration step and more time may be required for subsequent processing. Therefore, Butyl Sepharose 4FF was deemed more suitable than Phenyl Sepharose 6FF for FMDV purification.
Comparison of FMDV purification by IEC and HIC Purification efficiency on Butyl Sepharose 4FF and Phenyl Sepharose 6FF is summarized in Table 1, along with purification
efficiency on DEAE Sepharose FF and ANX Sepharose 4FF. The last two media were screened from several of frequently-used IEC media in previous work, with relative good purification efficiency. As seen from the table, the highest FMDV recovery and fold-purification are highest on Butyl Sepharose 4FF. Recovery from DEAE Sepharose FF and ANX Sepharose 4FF was acceptable, but IEC requires preliminary dilution or desalting of the crude, and therefore requires more time. Therefore, Butyl Sepharose 4FF was retained as the method of choice. Refinement by size exclusion chromatography FMDV eluted from Butyl Sepharose 4FF HIC was polished by SEC to remove residual contaminants. We evaluated SEC columns with separation range beyond, close to, and below 5500 kDa, the approximate molecular weight of FMDV [29]. Separation efficiency is listed in Table 2. SEC with Superdex 200 produced the sharpest FMDV peak, with excellent resolution of 2.96. The virus is excluded from Superdex 200, and was eluted in the exclusion volume while impurities were eluted much later (Fig. 3), indicating that impurities were of low molecular weight. In contrast, both FMDV and impurities are fractionated over Sepharose 4FF and Sepharose 6FF. Unfortunately, separation was poor on both, with resolution below 1. Therefore, Superdex 200 column was selected to polish FMDV preparations, even though preparative SEC often suffers from limited efficiency at industrial scale [32]. However, in this particular case, separation on SEC in ‘‘negative mode’’ is advantageous, and produces high-purity and high-concentration products, even in industrial scale.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
6
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx
Fig. 6. DLS analysis (A) and TEM imaging (B) of the purified FMDV.
Purification efficiency of the entire scheme FMDV may thus be purified in three steps, which include HIC on Butyl Sepharose 4FF, concentration by ultrafiltration, and SEC on Superdex 200. To verify efficiency and test reproducibility, FMDV purification was repeated three times. The initial crude was also scaled up from 125 mL to 500 mL. The size of the HIC column was increased accordingly, although ultrafiltration and SEC parameters were not changed. After purification, the identity of the final FMDV product was verified by Western blot (Fig. 4), with purity above 98%, as determined by SDS–PAGE (Fig. 4) and HPSEC (Fig. 5D). The major protein impurity in the initial crude was serum albumin, which showed a molecular weight near 70 kDa as shown in Fig. 4. The serum albumin was the main culture medium which is essential for cell growth. After HIC, the serum albumin was reduced, and then was completely removed by SEC. Host cell DNA was efficiently reduced to 0.5% after purification, with most being removed at the first HIC step. The purification efficiency of each step is shown in Table 3, with recovery over 90% after HIC and SEC. The average total recovery was 75%, with 247-fold increase in purity. The standard deviation from duplicate experiments was less than 6%, suggesting that the process is reproducible. On balance, the performance of this scheme is sufficient for industrial-scale production.
Structure of purified FMDV The structure of purified FMDV was analyzed by DLS and TEM. Samples exhibit a single DLS peak with diameter 28.0 nm and polydispersity index 0.21 (Fig. 6A), indicating monodispersity. This result is in good agreement with data from Ravindra Acharya [33]. Based on TEM imaging (Fig. 6B), the majority of purified virus are spherical, with calculated mean diameter 28 nm. These results confirm that purified FMDV is structurally intact, and thus appropriate as vaccine.
Conclusions Hydrophobic interaction chromatography (HIC) is suitable to purify inactivated FMDV. Of the HIC media evaluated, Butyl Sepharose 4FF was found to perform best to achieve recovery about 92% with 8.8-fold purification. HIC is more efficient than IEC by eliminating the need for preliminary desalting and reducing processing time.
Superdex 200 was used to polish FMDV preparations, because it was the most efficient among the SEC media evaluated. On this column, the virus was excluded from the matrix, and thus separated from contaminants, resulting in resolution near 3, 92% recovery, and 28-fold purification. However, optimum and efficient SEC purification requires an intermediate concentration step by ultrafiltration. On balance, 75% of FMDV could be reliably recovered from crude, and purified 247-fold to purities above 98%. More than 99% of host cell DNA was removed. Western blot, HPSEC, DLS and TEM indicate that the final product contained the required antigen, and was structurally intact with spherical shape and particle size 28 nm. Thus, our work demonstrates that HIC could be useful in purifying inactivated viruses whose structural integrity may be sensitive to industrial processing. Acknowledgments Financial supports from the Natural Sciences Foundation of China (Nos. 21336010, 21306207), National High Technology Research and Development Program of China (863 Program, Nos. 2012AA02A406 and 2014AA021006), National Key Scientific Instrument and Equipment Development Project (No. 2013YQ14040508), and Special Fund for Agro-scientific Research in the Public Interest (No. 201303046) are acknowledged. References [1] Z.Y. Li, J.X. Liu, The current state of vaccines used in the field for foot and mouth disease virus in China, Expert Rev. Vaccines 10 (2011) 13–15. [2] M.A. Spitteler, I. Fernandez, E. Schabes, A. Krimer, E.G. Regulier, M. Guinzburg, E. Smitsaart, M.S. Levy, Foot and mouth disease (FMD) virus: quantification of whole virus particles during the vaccine manufacturing process by size exclusion chromatography, Vaccine 29 (2011) 7182–7187. [3] S. Goodwin, T.J. Tuthill, A. Arias, R.A. Killington, D.J. Rowlands, Foot-and-mouth disease virus assembly: processing of recombinant capsid precursor by exogenous protease induces self-assembly of pentamers in vitro in a myristoylation-dependent manner, J. Virol. 83 (2009) 11275–11282. [4] P. Kitching, J. Hammond, M. Jeggo, B. Charleston, D. Paton, L. Rodriguez, R. Heckert, Global FMD control—is it an option?, Vaccine 25 (2007) 5660–5664 [5] C. Porta, X.D. Xu, S. Loureiro, S. Paramasivam, J.Y. Ren, T. Al-Khalil, A. Burman, T. Jackson, G.J. Belsham, S. Curry, G.P. Lomonossoff, S. Parida, D. Paton, Y.M. Li, G. Wilsden, N. Ferris, R. Owens, A. Kotecha, E. Fry, D.I. Stuart, B. Charleston, I.M. Jones, Efficient production of foot-and-mouth disease virus empty capsids in insect cells following down regulation of 3C protease activity, J. Virol. Methods 187 (2013) 406–412. [6] B.M. Subramanian, M. Madhanmohan, R. Sriraman, R.V.C. Reddy, S. Yuvaraj, K. Manikumar, S. Rajalakshmi, S.B. Nagendrakumar, S.K. Rana, V.A. Srinivasan, Development of foot-and-mouth disease virus (FMDV) serotype O virus-likeparticles (VLPs) vaccine and evaluation of its potency, Antivir. Res. 96 (2012) 288–295.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
H. Li et al. / Protein Expression and Purification xxx (2015) xxx–xxx [7] Z.Y. Li, Y.Z. Yi, X.P. Yin, Y. Zhang, M. Liu, H. Liu, X.R. Li, Y.N. Li, Z.F. Zhang, J.X. Liu, Development of a foot-and-mouth disease virus serotype A empty capsid subunit vaccine using silkworm (Bombyx mori) pupae, PLoS ONE 7 (2012) e43849. [8] H.W. van Vlijmen, S. Curry, M. Schaefer, M. Karplus, Titration calculations of foot-and-mouth disease virus capsids and their stabilities as a function of pH, J. Mol. Biol. 275 (1998) 295–308. [9] T.R. Doel, FMD vaccines, Virus Res. 91 (2003) 81–99. [10] S.J. Barteling, Development and performance of inactivated vaccines against foot and mouth disease, Rev. Sci. Tech. 21 (2002) 577–588. [11] L.L. Rodriguez, M.J. Grubman, Foot and mouth disease virus vaccines, Vaccine 27 (2009) D90–D94. [12] S.J. Barteling, R.H. Meloen, A simple method for the quantification of 140S particles of foot-and-mouth disease virus (FMDV), Arch. Ges. Virusforsch. 45 (1974) 362–364. [13] A.W. Morrow, C.J. Whittle, W.A. Eales, A comparison of methods for the concentration of foot-and-mouth disease virus for vaccine preparation, Bull. Off. Int. Epiz. 81 (1974) 1155–1167. [14] I.A. Galloway, M. Schlesinger, Purification and concentration of the virus of foot-and-mouth disease by combined centrifugation and ultrafiltration methods, J. Hyg. 37 (1937) 463–470. [15] A.A. Navarro del Canizo, M. Mazza, R. Bellinzoni, O. Cascone, Foot and mouth disease virus concentration and purification by affinity chromatography, Appl. Biochem. Biotechnol. 61 (1996) 399–409. [16] G. Iyer, S. Ramaswamy, D. Asher, U. Mehta, A. Leahy, F. Chung, K.S. Cheng, Reduced surface area chromatography for flow-through purification of viruses and virus like particles, J. Chromatogr. A 1218 (2011) 3973–3981. [17] Y. Huang, J. Bi, Y. Zhang, W. Zhou, Y. Li, L. Zhao, Z. Su, A highly efficient integrated chromatographic procedure for the purification of recombinant hepatitis B surface antigen from Hansenula polymorpha, Protein Expr. Purif. 56 (2007) 301–310. [18] B. Kalbfuss, M. Wolff, R. Morenweiser, U. Reichl, Purification of cell culturederived human influenza A virus by size-exclusion and anion-exchange chromatography, Biotechnol. Bioeng. 96 (2007) 932–944. [19] T. Weigel, T. Solomaier, A. Peuker, T. Pathapati, M.W. Wolff, U. Reichl, A flowthrough chromatography process for influenza A and B virus purification, J. Virol. Methods 207 (2014) 45–53. [20] J. Jimenez, H.K. Hammond, A. Miyanohara, T. Takahashi, D.M. Roth, N.C. Lai, Ion exchange chromatography purification of adeno-associated virus type 5, FASEB J. 19 (2005). A1540–A1540. [21] C.F. Yap, W.S. Tan, C.C. Sieo, B.T. Tey, Purification of long helical capsid of newcastle disease virus from Escherichia coli using anion exchange chromatography, Biotechnol. Prog. 29 (2013) 564–567.
7
[22] J.O. Baek, J.W. Seo, I.H. Kim, C.H. Kim, Production and purification of human papillomavirus type 33 L1 virus-like particles from Spodoptera frugiperda 9 cells using two-step column chromatography, Protein Expr. Purif. 75 (2011) 211–217. [23] M.R. Yu, Y. Li, S.P. Zhang, X.N. Li, Y.L. Yang, Y. Chen, G.H. Ma, Z.G. Su, Improving stability of virus-like particles by ion-exchange chromatographic supports with large pore size: advantages of gigaporous media beyond enhanced binding capacity, J. Chromatogr. A 1331 (2014) 69–79. [24] Y. Huang, J. Bi, L. Zhao, G. Ma, Z. Su, Regulation of protein multipoint adsorption on ion-exchange adsorbent and its application to the purification of macromolecules, Protein Expr. Purif. 74 (2010) 257–263. [25] T. Liang, D. Yang, M. Liu, C. Sun, F. Wang, J. Wang, H. Wang, S. Song, G. Zhou, L. Yu, Selection and characterization of an acid-resistant mutant of serotype O foot-and-mouth disease virus, Arch. Virol. 159 (2014) 657–667. [26] M.G. Rao, G. Butchaiah, A.K. Sen, Antibody-response to 146S particle, 12S protein subunit and isolated Vp1 polypeptide of foot-and-mouth-disease virus type Asia-1, Vet. Microbiol. 39 (1994) 135–143. [27] N.P. Ferris, A.I. Donaldson, I.T.R. Barnett, R.W. Osborne, Inactivation, purification and stability of 146S antigens of foot and mouth disease virus for use as reagents in the complement fixation test, Rev. Sci. Tech. Off. Int. Epiz. 3 (1984) 339–350. [28] W.B. Zhou, J.X. Bi, L. Zhao, Y.M. Wang, Y. Li, Y.D. Huang, G.H. Ma, Z.G. Su, A highly efficient hydrophobic interaction chromatographic absorbent improved the purification of hepatitis B surface antigen (HBsAg) derived from Hansenula polymorpha cell, Process Biochem. 42 (2007) 751–756. [29] Y. Yang, H. Li, Z. Li, Y. Zhang, S. Zhang, Y. Chen, M. Yu, G. Ma, Z. Su, Sizeexclusion HPLC provides a simple, rapid, and versatile alternative method for quality control of vaccines by characterizing the assembly of antigens, Vaccine 33 (2015) 1143–1150. [30] H.L. Bachrach, R. Trautman, S.S. Breese Jr., Chemical physical properties of virtually pure foot-and-mouth disease virus, Am. J. Vet. Res. 25 (1964) 333– 342. [31] W. Zhou, J. Bi, J.C. Janson, A. Dong, Y. Li, Y. Zhang, Y. Huang, Z. Su, Ion-exchange chromatography of hepatitis B virus surface antigen from a recombinant Chinese hamster ovary cell line, J. Chromatogr. A 1095 (2005) 119–125. [32] B. Kalbfuss, D. Flockerzi, A. Seidel-Morgenstern, U. Reichl, Size-exclusion chromatography as a linear transfer system: purification of human influenza virus as an example, J. Chromatogr. B 873 (2008) 102–112. [33] R. Acharya, E. Fry, D. Stuart, G. Fox, D. Rowlands, F. Brown, The threedimensional structure of foot-and-mouth disease virus at 2.9 Å resolution, Nature 337 (1989) 709–716.
Please cite this article in press as: H. Li et al., A hydrophobic interaction chromatography strategy for purification of inactivated foot-and-mouth disease virus, Protein Expr. Purif. (2015), http://dx.doi.org/10.1016/j.pep.2015.04.011
