Pi

Fa

b

c

d

e

a

ARAA

KNNIE

1

hil[cmtib

EgiNE[

nM

1d

Journal of Chromatography B, 877 (2009) 1561–1567

Contents lists available at ScienceDirect

Journal of Chromatography B

journa l homepage: www.e lsev ier .com/ locate /chromb

urification of histidine-tagged nucleocapsid protein of Nipah virus usingmmobilized metal affinity chromatography

ui Chin Chonga,b, Wen Siang Tanc,d, Dayang Radiah Awang Biakb, Tau Chuan Linge, Beng Ti Teyb,d,∗

Department of Chemical and Natural Resources Engineering, Faculty of Engineering, Universiti Malaysia Pahang, 25000 Kuantan, Pahang, MalaysiaDepartment of Chemical and Environmental Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor, MalaysiaDepartment of Microbiology, Faculty of Biotechnology and Biomolecular Sciences, Universiti Putra Malaysia, 43400 Serdang, Selangor, MalaysiaInstitute of Bioscience, Universiti Putra Malaysia, 43400 Serdang, Selangor, MalaysiaDepartment of Process and Food Engineering, Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia

r t i c l e i n f o

rticle history:eceived 3 February 2009ccepted 31 March 2009vailable online 7 April 2009

a b s t r a c t

Nucleocapsid (N) protein of Nipah virus (NiV) is a potential serological marker used in the diagnosis of NiVinfections. In this study, a rapid and efficient purification system, HisTrapTM 6 Fast Flow packed bed columnwas applied to purify recombinant histidine-tagged N protein of NiV from clarified feedstock. The opti-mizations of binding and elution conditions of N protein of NiV onto and from Nickel SepharoseTM 6 FastFlow were investigated. The optimal binding was achieved at pH 7.5, superficial velocity of 1.25 cm/min.

eywords:ucleocapsid proteinipah virus

mmobilized metal affinity chromatographyscherichia coli

The bound N protein was successfully recovered by a stepwise elution with different concentration ofimidazole (50, 150, 300 and 500 mM). The N protein of NiV was captured and eluted from an inlet N pro-tein concentration of 0.4 mg/ml in a scale-up immobilized metal affinity chromatography (IMAC) packedbed column of Nickel SepharoseTM 6 Fast Flow with the optimized condition obtained from the methodscouting. The purification of histidine-tagged N protein using IMAC packed bed column has resulted a

tion f

68.3% yield and a purifica

. Introduction

Nipah virus (NiV) is a deadly zoonotic virus of bat origin andas been classified as biosafety level 4 (BSL4) pathogens [1]. Since

ts presence in the 1990s, lethal infections of NiV in human and/orivestock animals are reported almost every year in Southeast Asia2–4]. In 1998 and 1999, the outbreak of NiV in Malaysia haslaimed 105 human lives and resulted in the culling of about 1.1illion pigs [5], bringing tremendous economic and social impact

o the nation. Hence, vigorous serological monitoring of the NiVnfection is necessary to prevent the occurrence of the major out-reak.

The N protein of NiV has been successfully expressed inscherichia coli (E. coli), and it is highly antigenic and immuno-enic [6]. The recombinant N protein assembles automatically

nto herringbone-like structure [6,7], which resembles the nativeiV nucleocapsid. Yu et al. [8] developed a NiV-N protein-basedLISA system as a diagnosis tool for NiV infection. Ndifuna et al.9] recommended that the N protein can be used in large-scale

∗ Corresponding author at: Department of Chemical and Environmental Engi-eering, Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor,alaysia. Tel.: +60 3 89466289; fax: +60 3 86567120.

E-mail addresses: btey@eng.upm.edu.my, btey@yahoo.com (B.T. Tey).

570-0232/$ – see front matter © 2009 Elsevier B.V. All rights reserved.oi:10.1016/j.jchromb.2009.03.048

actor of 7.94.© 2009 Elsevier B.V. All rights reserved.

epidemiological investigations and to be applied in developingcountries.

The major problem during the production and purification ofthe N protein of NiV is the low recovery yield due to the proteolyticdegradation [6]. The protein degradation can be reduced by short-ening the purification time [10]. The lab scale purification of the Nprotein using sucrose gradient ultracentrifugation is time consum-ing, and thus provides sufficient time for the protease to attack the Nprotein. Therefore, the development of a rapid and simplified purifi-cation of N protein is desired. The placement of a histidine hexamertag at the C-terminus of the N protein has enabled the purificationof the N protein by Ni2+ based immobilized metal affinity chro-matography (IMAC) [6]. IMAC can be applied in the early stage ofprotein purification before the protein precipitation and dialysissteps, hence significantly shorten the purification time.

The protein adsorption on IMAC is governed by various factorssuch as the type of chelating ligand, metal ion, the surround-ing chemical environment and experimental parameters on thedynamic binding capacity [11–14]. Furthermore, Sharma et al.[12] demonstrated that the role of ionic strength and pH of the

chromatographic medium is significant in governing the bindinginteraction between the protein and the adsorbent. The perfor-mance of a chromatography process such as dynamic bindingcapacity and recovery throughput can be assessed by frontal break-through analysis. Finette et al. [14] examined frontal breakthrough

1 atogr.

mdh

pllhapcafob

2

2

aaXltprp

2

tcT�b(a8

piircfdwcff

2

2

dtwtta(h

562 F.C. Chong et al. / J. Chrom

easurements in packed bed IMAC with different flow rates andifferent inlet concentrations of hen egg white lysozyme anduman serum albumin.

The benefits of the uses of IMAC as a ligand are its stability, highrotein loading, mild elution conditions, simple regeneration and

ow in cost. These factors are important to be considered whenarge-scale purification procedures are involved. In this respect, weave selected the Nickel SepharoseTM 6 Fast Flow with iminodi-cetic acid (IDA) chelating group, since it has low Ni2+ leakage, highrotein binding capacity, high flow properties and reproduciblehromatographic performance. The detailed investigations on thedsorption and elution behaviors of proteins in IMAC were per-ormed. The yield and efficiency of purification procedure of variousperating parameter for Nickel SepharoseTM 6 Fast Flow in packeded column were evaluated.

. Experimental

.1. Materials and equipment

Precharged Nickel SepharoseTM 6 Fast Flow affinity adsorbentnd prepacked HisTrapTM FF 1 ml column containing the samedsorbents were purchased from GE Healthcare (Uppsala, Sweden).K 16/20 column (GE Healthcare) with a 16 mm diameter and 20 cm

ength and a cross-sectional column area of 201 mm2 was usedo perform a lab scale IMAC purification of N protein of NiV. Theacked bed column was connected to the Äkta FPLC chromatog-aphy system (GE Healthcare) throughout the protein purificationrocess.

.2. Feedstock preparation

E. coli strain BL21 (DE3) harbouring plasmid pTrcHis2 expressinghe N protein of NiV [6] was cultured in Luria–Bertani (LB) mediumontaining ampicillin at 25 ◦C with vigorous shaking at 200 rpm.he expression of N protein was induced by adding isopropylthio--d-galactoside (IPTG) to a final concentration of 1 mM when theiomass concentration of the culture reached absorbance at 600 nmA600) about 0.6–0.8. The induced culture was further incubated fornother 5 h at 25 ◦C. The cells were harvested by centrifugation at,000 × g (JLA 16.25 rotor, Beckman, USA) for 20 min at 4 ◦C.

The pelleted cells were resuspended in buffer solution sup-lemented with 0.2 �g/ml lysozyme and 4 mM MgCl2. Protease

nhibitor (1 mM phenylmethylsulphonyl fluoride (PMSF)) wasncluded in the buffer to inhibit the activity of host proteases thateleased together with the N protein during cell disruption pro-ess. The cell suspension was subjected to ultrasonication at 200 Wor 30 s with 60 s intervals in an ice bath for duration of 50 min asescribed previously by Ho et al. [15]. The cell lysate was treatedith DNase (5 �g/ml) and incubated for 1 h at 4 ◦C. The lysate was

larified by centrifugation at 18,000 × g (JA 20 rotor, Beckman, USA)or 20 min at 4 ◦C. The clarified supernatant was used as feedstockor the subsequent purification process.

.3. Method scouting and optimization

.3.1. Optimization of binding buffer conditionAn experiment to determine the optimal binding buffer con-

ition for purifying the N protein of NiV was performed by usinghe HisTrapTM FF 1 ml column. The HisTrapTM FF 1 ml columnas loaded with the same buffer solution in each equilibra-

ion step and binding step. The compounds chosen for preparinghe binding buffer solutions were 2-(N-morpholino)ethanesulfoniccid (MES) for pH 6, piperazine-N,N-bis(2-ethanesulfonic acid)PIPES) for pH 6.5, sodium phosphate for pH 7 and 7.5, 4-2-ydroxyethyl-1-piperazineethanesulfonic acid (HEPES) for pH 8

B 877 (2009) 1561–1567

and tris-(hydroxymethyl)-methylamine (Tris) for pH 8.5. The con-centration of imidazole and NaCl added into all the buffer solutionswas 20 mM and 500 mM, respectively, before the pH was adjusted.The pH range of the binding buffer solutions was prepared from 6.0to 8.5. The HisTrapTM FF 1 ml column was equilibrated with imida-zole prior to chromatographic purification to avoid drastic pH dropdue to the effect of imidazole protonTM pump [16].

The HisTrapTM FF 1 ml column was rinsed with 5 column vol-ume (CV) of distilled water and equilibrated with 5 CV of bindingbuffer. The column was then loaded with clarified lysate at 0.8 mg Nprotein of NiV/ml adsorbent. Fractions of unbound protein sampleat 0.5 ml volume throughout the binding stage were collected andanalysed to determine the amount of total protein and N proteinof NiV. The amount of bound total protein and N protein was cal-culated by applying mass balance established from the initial andunbound amount of protein [17].

2.3.2. Optimization of elution conditionsPrior to the elution, the prepacked HisTrapTM FF 1 ml column

was loaded with the N protein as described in Section 2.3.1. Thecolumn was then washed with 5 CV of binding buffer to removeloosely bound protein. The N protein was eluted from the columnin stepwise elution using binding buffer added with different con-centrations of imidazole of 50, 150, 300 and 500 mM (5 ml for eachconcentration). Eluted protein fractions were collected and anal-ysed for the amount of N protein and total protein. The purity andyield of the amount N protein obtained in the purified fractionswere calculated as described by Ng et al. [17] and Tan et al. [18].

2.4. Operation of packed bed column

2.4.1. Packing of affinity adsorbentNickel SepharoseTM 6 Fast Flow affinity adsorbent was packed

in an XK 16/20 column according to the procedure recommendedby GE Healthcare [19]. Briefly, slurry of 70% settled medium to30% distilled water was prepared and poured into the column.The remainder column was filled up with buffer and the columnadapter was mounted. By connecting to a pump, a packing flowrate of 2.5 ml/min was run in first step and 8.7 ml/min in secondstep. The packing flow rate was maintained for 3 column volumeafter a constant bed height was reached. The pump was stopped andthe adapter was locked in position. The column was washed with5 CV of distilled water and equilibrated with 5 CV binding buffer(20 mM sodium phosphate, 500 mM NaCl and 20 mM imidazole,pH 7.5) [19].

2.4.2. Breakthrough curve and dynamic binding capacityThe breakthrough curve was performed to determine the

dynamic capacity and the productivity of the Nickel SepharoseTM

6 Fast Flow affinity adsorbent in the XK 16/20 column. NickelSepharoseTM 6 Fast Flow affinity adsorbent was packed in an XK16/20 column at a packed bed height of 10 cm which was installedonto the Äkta FPLC. The breakthrough curve for the packed bedsystem was obtained by loading clarified feedstock of 0.4 mg/ml Nprotein concentration at a superficial velocity of 1.25 cm/min. Pro-tein fractions (10 ml) collected throughout the operation were thenanalysed to determine the concentration of N protein. The dynamicbinding capacity, QB, is the total amount of N protein adsorbedin the packed bed column per unit adsorbent volume, when theoutlet concentration of N protein is 10% of the inlet concentration.Recovery throughput is another parameter in affinity purification.

This refers to the amount of adsorbed protein at 10% breakthroughdivided by the volume of adsorbent and the processing time; bothwere calculated according to Chang and Chase [20]. The time forprotein adsorption was recorded once the 10% breakthrough wasreached. The processing time in one purification cycle was defined

atogr.

aai5abb

2

pifpidczai

2

2

Scitt1wi2bhpipaebcpr

mbtT0fss

TE

Ba

222222

F.C. Chong et al. / J. Chrom

s the sum of the time for equilibration, loading, washing, elution,nd cleaning-in-place (CIP). The time for each equilibration, wash-ng and elution step for all experiments running at 2.5 ml/min overCV was 0.67 h. The CIP time for XK 16/20 column using 1 M NaOH

nd distilled water to remove any precipitated proteins, hydropho-ically bound proteins and lipoproteins was 2 h, as recommendedy the manufacturer.

.5. Purification of the N protein of NiV using IMAC

IMAC was performed with Nickel SepharoseTM 6 Fast Flowacked in XK 16/20 column. The clarified feedstock was loaded

nto the equilibrated column followed by washing. Equilibration,eed application and washing were performed with 20 mM sodiumhosphate, 500 mM NaCl, 20 mM imidazole, pH 7.5. After the wash-

ng step, the elution buffer was run at 2.5 ml/min in an upwardirection by applying two steps elution. Firstly, 3 CV of elution bufferontaining 20 mM sodium phosphate, 500 mM NaCl, 50 mM imida-ole, pH 7.5 was used to remove unbound contaminant proteinsnd then followed by 3 CV of elution buffer containing 300 mMmidazole to elute the bound N protein from the adsorbent.

.6. Analytical procedure

.6.1. Protein analysis and quantitationElectrophoresis separation of proteins in the cell lysate by

DS-PAGE was performed with 12% polyacrylamide gel [21]. Nitro-ellulose membrane was cut to the size of the gel and equilibratedn Tris–glycine–methanol transfer buffer. Proteins were transferredo the nitrocellulose membrane using the Transblot SD semidryransfer cell (Biorad, USA). The current was kept at 0.8 mA percm2 of gel at 15 V for 1 h. All antigen and antibody dilutionsere done in TBS (50 mM Tris, 150 mM NaCl, pH 7.5). All wash-

ng steps were carried out three times with TTBS (0.1% Tween0 in TBS). After removal from the blotting apparatus, the mem-rane was placed in blocking buffer (5% non-fat dried milk thatad been dissolved in TTBS) and incubated for 1 h at room tem-erature. Subsequently, the membrane was rinsed 3 times and

ncubated for 10 min at room temperature. The protein bands wererobed by adding alkaline phosphatase-conjugated antihistidinentibody (1:5000 dilution) (Invitrogen, USA) (or mentioned oth-rwise) and colour was developed in 10 ml of alkaline phosphataseuffer (100 mM Tris–HCl (pH 9.5), 100 mM NaCl and 5 mM MgCl2)ontaining 33 �l of 50 mg/ml 1-bromo-3-chloro-3-indolyl phos-hate (BCIP) and 66 �l of 50 mg/ml nitro blue tetrazolium (NBT)eagent (Fermentas, USA).

The concentration of N protein from various samples was esti-ated by comparing the intensity of their bands from Western

lots with the help of an internal standard of corresponding pro-ein purified using sucrose gradient ultracentrifugation [17,18,22].

he concentration of the internal standards was from 0.01 to.09 mg/ml. The bands were quantified by using the Volume Toolsrom the Bio-Rad imaging devices supported by the Quantity One®

oftware (Bio-Rad, Hercules, California, USA). A volume is the inten-ity data inside a defined boundary drawn on the images. The

able 1ffect of buffer and pH on the binding conditions of histidine-tagged N protein of NiV from

inding buffer (all buffersdded with 0.5 M NaCl)

pH Amount of bound Nprotein (mg)

Perceprote

0 mM MES 6 1.3 15.90 mM PIPES 6.5 1.0 12.30 mM phosphate 7 5.6 70.50 mM phosphate 7.5 5.9 73.80 mM HEPES 8 4.8 59.50 mM Tris 8.5 3.0 37.6

B 877 (2009) 1561–1567 1563

intensity of the data inside the boundary and that of other objectscan be compared using the Volume Analysis report. The yield of Nprotein was expressed as the amount of N protein obtained dividedby the initial amount of N protein. The amount of total proteinwas determined using the Bradford assay [23]. Bovine serum albu-min (BSA) was used as standard. All the samples were analysed induplicate.

2.6.2. Enzyme-linked immunosorbent assay (ELISA)The antigenicity of the purified N protein of NiV by IMAC was

determined by ELISA as described by Tan et al. [18]. The purifiedproduct was used as the capturing antigen in ELISA to determinethe presence of the anti-NiV antibodies in serum samples collectedfrom infected rabbit [6]. A U-bottom polystyrene microtiter platewas used as solid-phase adsorbents (TPP Immunomax high bindingELISA plate, USA). The amount of purified histidine tagged N pro-tein (10–1000 �g; 100 �l) pooled elution fractions were preparedin TBS. The plate was first coated with the purified N protein forovernight at 4 ◦C. The coated plate was blocked with 200 �l 10%milk diluent (KPL, USA) for 2 h at 4 ◦C. For the entire washing step,the solution was decanted and the plate was washed and incubatedfor 5 min at room temperature (RT) with TTBS (0.05% Tween-20 inTBS) for three times. The remaining solution was removed by pat-ting the plate on a paper towel. The plate was then incubated for 2 hat RT with the anti-N protein rabbit serum (1:100 dilution, 100 �l)[6] from the infected and uninfected animals. The uninfected rab-bit serum was used as a negative control in every assay. Afterthe washing step, the anti-rabbit antibody conjugated to alkalinephosphatase (1:5000 dilution, 100 �l) (Invitrogen) was added andincubated for 2 h at RT. Subsequently the plate was washed againand follow by the addition of 200 �l of the enzyme–substrate solu-tion, containing 1 mg/ml PNPP (p-Nitrophenyl phosphate, Sigma,USA) in 0.1 M diethanolamine (Sigma, USA), pH 10.3. The reactionwas stopped after 15 min by the addition of 50 �l NaOH (1N) to eachwell. The absorbance value at 405 nm (A405) was determined in amicroplate reader (Bio Tek Instruments Inc. Model Elx 800, USA).The assays were performed in triplicates.

3. Results and discussion

3.1. Optimization of binding buffer condition

In IMAC, the strength of binding between a protein and a metalion depends on the pH of buffer [12,24,25]. In this study, the effectof pH for adsorption of N protein was investigated in HisTrap FF1 ml column. A pH range from 6.0 to 8.5 and sodium chloride at500 mM added in IMAC buffer was being considered in this study.Under weakly alkaline pH and in the presence of high ionic strengthbuffers, adsorption of histidine tagged protein in IMAC is favourableinduced [18,26]. The electrolytes of high ionic strength buffer will

weaken the interaction between of the metal ion and the solvatedwater molecules, and thus favours the specific ionic interactionsbetween histidine tagged protein and IMAC matrix [26]. While atmore alkaline range, the pH condition favours coordination withamino functional groups, thus selectivity is decreased [27].

20 ml clarified feedstock (5.2 mg/ml) using HisTrapTM FF 1 ml column.

ntage of bound Nin (%)

Amount of bound totalprotein (mg)

Percentage of boundtotal protein (%)

11.8 11.313.1 12.615.6 1516.6 16.014.2 13.710.4 10

1564 F.C. Chong et al. / J. Chromatogr. B 877 (2009) 1561–1567

F 1 mla in (A)l lanes

esoar27f1fSpbBact

3

5ieubfsTtoz(b5ht

TEs

E

EEEE

ig. 1. Graphical illustration showing elution profile of N protein from HisTrapTM FFnd (B) graphical illustration showing elution profile of IMAC. The N protein bandsanes 2–6: elution 1 (50 mM imidazole); lanes 7–11: elution 2 (150 mM imidazole);

In the present study, six binding buffer conditions were consid-red for optimization (Table 1). The results showed that pH bufferystem ranging from 6.0 to 8.5 affected differently the adsorptionf N protein. More than 60% of the total N protein loaded wasdsorbed to the Nickel SepharoseTM 6 Fast Flow adsorbent at pHange of 7–8. Among these eight binding conditions, the buffer with0 mM sodium phosphate, 500 mM NaCl and 20 mM imidazole, pH.5 yielded the highest percentage of bound N protein, about 74%rom the initial amount of N protein loaded into the HisTrap FFml column. Similarly, Tan et al. [18] reported that the optimal pH

or adsorption of N protein of Newcastle disease virus (NDV) ontotreamline chelating adsorbent was at pH 8. Small amount of totalrotein and degraded proteins were binding onto the adsorbent,ut it was relatively small (<16%) compared to that of the N protein.inding between the contaminating protein and the adsorbent waslso occurred at pH ranging from 6.0 to 8.5. The optimized bindingondition selected would be utilized to scale up the purification ofhe N protein of NiV using the Äkta FPLC.

.2. Optimization of imidazole concentration in elution buffer

The protein loaded column was eluted in stepwise elution usingml of each binding buffer containing various concentrations of

midazole, 50, 150, 300 and 500 mM. The purity and yield of theluted N protein fractions were confirmed by Western blottingsing antihistidine antibody. The elution profile of the N proteinands detected by Western blot of 15 �l of 1 ml fractions elutedrom the HisTrapTM FF 1 ml column is shown in Fig. 1A. Table 2ummarises the purity and yield values for the eluted N protein.he highest purity of 0.95 of eluted N protein was achieved whenhe 500 mM imidazole buffer was applied. However, higher amountf N protein was eluted with the buffer containing 300 mM imida-ole (lanes 12–16), and the thickest band was observed on lane 13

Fig. 1). As tabulated in Table 2, the yield of N protein for elutionuffer containing 300 mM imidazole was 3-fold higher than that of00 mM imidazole. The elution buffer containing 50 mM imidazolead washed out most of the unwanted host cell proteins bound tohe adsorbent.

able 2ffect of increasing imidazole concentration in 20 mM phosphate buffer pH 7.5 in elutiontepwise increments.

lution step Total elutionvolume (ml)

Imidazoleconcentration (mM)

Total proteelution (�

1 5 50 80502 5 150 31183 5 300 34204 5 500 993

column. The (A) Western blot analysis of the N protein fractions eluted by stepwisewere detected by anti-histidine antibody. Lane 1: clarified feedstock (0.4 mg/ml);

12–16: elution 3 (300 mM imidazole); lanes 17–21: elution 4 (500 mM imidazole).

Tan et al. [18] successfully designed a two stages elution whichwas particularly useful in washing out unwanted proteins and atthe same time to recover the recombinant N protein of Newcastledisease virus from a Streamline chelating adsorbent. Based on theabove observation and the work of Tan et al. [18], a two-stage elu-tion was applied in this study to scale-up the packed bed column,50 mM imidazole buffer was used in the first stage elution to elutecontaminating proteins, followed by the elution buffer containing300 mM imidazole buffer to elute the N protein from the adsorbent.

3.3. Breakthrough curves

3.3.1. Effect of superficial velocity on the breakthrough curveThe effects of superficial velocity on the adsorption of N pro-

tein of NiV onto the Nickel SepharoseTM 6 Fast Flow adsorbentin the XK 16/20 packed bed column experiments were performedwith 0.4 mg/ml N protein loading, at pH 7.5. The superficial veloc-ities used were 0.5, 1.25, 2.5 and 5 cm/min and the breakthroughcurves are depicted in Fig. 2. An early breakthrough was recorded asthe linear superficial velocity was increased from 0.5 to 5 cm/min.Chang and Chase [20], and Chase [28] also reported that the shape ofbreakthrough curves varied with superficial velocity when adsorp-tion kinetics were slow relative to the mass transfer rate. A slowersuperficial velocity during feedstock loading in a packed bed opera-tion may have assisted in giving more interaction time for proteinsand adsorbents, thus led to a higher binding of N protein onto theadsorbent as shown by the results tabulated in Table 3.

The dynamic binding capacity (QB) and the adsorption through-put of the adsorbents under various superficial velocities wereestimated from the breakthrough curve (Fig. 2). An increment ofsuperficial velocity from 0.5 to 5 cm/min resulted in the decreaseof QB of 56% (Table 3). Similarly, other researchers have reportedthat the value of QB obtained reduced as the superficial velocity

increased [14]. The highest QB at 10% breakthrough of 2.5 mg/mlwas achieved at the lowest linear flow velocity, 0.5 cm/min.The long residential time provided under this operating condi-tion may contribute to the higher protein adsorption. However,the longer adsorption time has affected negatively its recovery

profile using HisTrapTM FF 1 ml column by loading in 5 ml of each elution buffer in

in ing)

N protein inelution (�g)

Purity of elutedN protein (%)

Yield of elutedN protein (%)

440.5 5.5 4.891455.4 46.7 16.172953.2 86.4 32.81

940.6 94.7 10.45

F.C. Chong et al. / J. Chromatogr. B 877 (2009) 1561–1567 1565

Fat(

ttthcofr

3

cichIp0

a

Fig. 3. Effect of inlet N protein concentration on the breakthrough curve obtained

TP

S(

0125

TE

IN

000

ig. 2. Packed bed N protein breakthrough curves as a function of superficial velocityt 25 ◦C on Nickel SepharoseTM 6 FF (particle size 90 �m). C is the N protein concen-ration in the effluent. C0 is the N protein concentration in the clarified feedstock0.4 mg/ml) and V is the effluent volume.

hroughput. Hence, the throughput calculated is almost similar tohat obtained in 1.25 cm/min. Nevertheless, the volume of break-hrough and the adsorption time for 0.5 cm/min is 1.3× and 3.3×igher than that of 1.25 cm/min. Therefore, the linear superfi-ial velocity of 1.25 cm/min was selected for the later purificationperation. The QB at 10% breakthrough and recovery throughputor 1.25 cm/min was 1.7 mg/ml adsorbent and 5.97 �g ml−1 min−1,espectively.

.3.2. Effect of protein concentration on dynamic binding capacityThe breakthrough curves of feedstock with various protein con-

entrations are shown in Fig. 3. As the inlet N protein concentrationncreased from 0.2 to 0.6 mg/ml, the steepness of the breakthroughurves increased and the curve positions were shifted to the left,ence, a shorter time was required to achieve 10% breakthrough.

ndeed, the 10% breakthrough time required by the feedstock withrotein concentration of 0.2 mg/ml is 4.2× higher than that of.6 mg/ml.

The increased in inlet protein concentration has resulted inhigher concentration gradient in the adsorbent bead. Higher

able 3acked bed dynamic binding capacity at 10% breakthrough as a function of superficial vel

uperficial velocitycm/min)

Volume at 10%breakthrough

Dynamic binding capacity at10% breakthrough, QB

(mg/ml)

.5 124 2.5

.25 95 1.9

.5 75 1.555 1.1

able 4ffect of protein concentration on dynamic binding capacity.

nlet concentration ofprotein, C0 (mg/ml)

Volume at 10%breakthrough (ml)

Dynamic binding capacityat 10% breakthrough, QB

(mg/ml)

.6 36.0 1.1

.4 95.0 1.9

.2 144.0 1.4

for the adsorption of N protein to Nickel SepharoseTM 6 FF at a constant superficialvelocity of 2.5 ml/min. C is the N protein concentration in the effluent. C0 is the Nprotein concentration in the clarified feedstock and V is the effluent volume.

concentration gradient has enhanced the diffusion of protein tothe surface of the adsorbent bead. Hence, a faster saturation ofthe adsorbent beads is achieved and a shorter time is requiredto achieve the breakthrough [13]. Chase et al. [28] reported thatthe change in shape and position of the breakthrough curvewas noticed when the parameter of inlet concentration (C0) iscomparable or smaller than the desorption equilibrium constant(Kd). However, the highest recovery throughput was obtainedat the operation using feedstock with protein concentration of0.4 mg/ml due to its higher QB that had been achieved. The QBand recovery throughput for feedstock with protein concentrationof 0.4 mg/ml were 1.9 mg/ml and 6.82 �g ml−1 min−1, respectively(Table 4).

3.4. Purification of the N protein by packed bed column

The optimized binding conditions and elution strategy from the

optimization study was used to develop an optimized purificationoperation for the N protein of NiV from clarified E. coli feedstock. Thepurification operation was carried out at 25 ◦C, inlet N protein con-centration of 0.4 mg/ml and pH 7.5. The feedstock was applied into

ocity.

Adsorption time (min) Processing time (min) Recovery throughput(�g ml−1 min−1)

124 364.6 6.8038 278.6 6.8215 255.6 5.875.5 246.1 4.47

Adsorption time (min) Processing time (min) Recovery throughput(�g ml−1 min−1)

14 255 4.2438 279 6.8258 298 4.83

1566 F.C. Chong et al. / J. Chromatogr. B 877 (2009) 1561–1567

Fig. 4. SDS-PAGE analysis (A) and Western blot analysis (B) show the protein fractions (15 �l samples in each well) collected in the IMAC process. Lane 1: molecular massmarkers; lane 2: clarified Escherichia coli lysate; lanes 3–12: flowthrough; lanes 13–22: washing; lanes 23–28: elution with 50 mM imidazole; lanes 29–34: elution with300 mM imidazole.

Table 5The purification of histidine-tagged N protein of NiV from clarified feedstock using IMAC packed bed column.

Purification stage Total volume (ml) Total protein (mg) N protein purity (%) Amount of N protein (mg) Recovery yield (%) Purification factor

Clarified feedstock 100 522.0 8.29 43.26Flowthrough 100 304.9 3.76 11.46W 1.84E 8.9 20.57 7.41E 20.66 47.76 7.94

ttkttSi

flottitot

2TtwbhortrfNb4fiopF

ashing 100 170.6 1.08lution 1 (50 mM imidazole) 60 14.5 61.38lution 2 (300 mM imidazole) 60 31.4 65.80

he IMAC packed column at a flow rate similar to that used duringhe adsorbent bed equilibration (2.5 ml/min) and this flow rate wasept constant throughout the entire purification process (equilibra-ion, flowthrough, washing and elution stages). The protein profilehroughout the entire IMAC packed bed operation analysed withDS-PAGE is shown in Fig. 4, and the quantitative data are presentedn Table 5.

In Fig. 4A, the results showed that the protein loss in theowthrough and washing stages was about 31% of the total amountf N protein in the initial feedstock. The binding capacity of N pro-ein calculated was 1.5 mg/ml, which is still lower than the QBhat was achieved during the method scouting using the samenlet N protein concentration. Hence, the loss of N protein inhe flowthrough was probably due to the competitive adsorptionccurred between the target protein and other contaminant pro-eins at the binding sites of the adsorbent [18].

As shown in Table 5, the two-stage elution process has yielded9.6 mg of N protein, which was collected in IMAC fractions 23–34.he total recovery yield of the histidine-tagged N protein from bothhe two-stage elution was 68.3%, almost all the adsorbed N proteinas successfully eluted out from the adsorbent. The IMAC adsor-ent used in the present study has interacted very specifically to theistidine-tagged N protein. The purity of intact N protein (63 kDa)btained in the first and second stage of elution was 61.4 and 65.8%,espectively. The lower purity of intact N protein eluted is due tohe present of the degraded portion of the N protein (45 kDa), whichepresented about 35.6% of the total protein present in the elutionractions. In the elution profile obtained from the IMAC process, the

protein was detected with the anti-N protein rabbit serum anti-ody in Western blot analysis (Fig. 4B) which confirmed that the

5 kDa protein band was indeed the degraded N protein. A similarnding was reported by Tan et al. [6]. As a summary, the purificationf his-tagged N protein of NiV using IMAC packed bed purificationrocess has achieved a 68.3% yield and a purification factor of 7.94.urther improve of the histidine-tagged N protein purity can be

Fig. 5. The antigenicity of histidine-tagged N protein of NiV purified with IMAC.Error bars: standard deviation of the triplicate analysis.

achieved with size exclusion chromatography as been described byHo et al. [29]. The purified N protein of NiV from the IMAC systemwas evaluated by ELISA using the anti-N protein rabbit serum. TheELISA result depicted in Fig. 5 revealed that the antigenicity of thepurified N protein is still preserved.

4. Conclusion

In the current study, we have developed a scale-able IMAC for thepurification of the recombinant histidine-tagged N protein of NiVin E. coli feedstock. Optimization of buffers in binding and elutionstages is easily performed in HisTrapTM FF 1 ml column. The binding

atogr.

bdNdTpSbt2mtcT

A

SPst

R

[

[[[[

[

[[

[[[

[[

F.C. Chong et al. / J. Chrom

uffer of 20 mM sodium phosphate, 500 mM NaCl and 20 mM imi-azole at pH 7.5 has the highest affinity for the N protein. The boundprotein was successfully recovered by a stepwise elution with

ifferent concentrations of imidazole of 50, 150, 300 and 500 mM.hese binding and elution conditions were used for the scale-upurification process in a 20 ml packed bed column volume of NickelepharoseTM 6 Fast Flow. Based on the frontal analysis, the dynamicinding capacity of Nickel SepharoseTM 6 Fast Flow at 10% break-hrough for 0.4 mg N protein of NiV per ml of clarified feedstock at.5 ml/min constant velocity operation was 1.9 ml/min and the opti-um adsorption throughout was 6.82 �g ml−1 min−1. The purifica-

ion of the histidine-tagged N protein using the IMAC packed bedolumn has resulted a 68.3% yield and a purification factor of 7.94.he antigenicity of the purified N protein is still well preserved.

cknowledgements

This study was supported by the Research University Grantcheme (RUGS) 2007 (Grant No: 05/01/07/0225RU) from Universitiutra Malaysia. Fui Chin Chong is supported financially by Univer-iti Malaysia Pahang. We are grateful to Swee Tin Ong for providinghe E. coli pTrcHis2 clone expressing the N protein of NiV.

eferences

[1] S.K. Lam, Antiviral Res. 57 (2003) 113.[2] M.S. Chadha, J.A. Comer, L. Lowe, P.A. Rota, P.E. Rollin, W.J. Bellini, T.G. Ksiazek,

A. Mishra, Emerg. Infect. Dis. 12 (2006) 235.[3] J.G. Olson, C. Rupprecht, P.E. Rollin, U.S. An, M. Niezgoda, T. Clemins, J. Walston,

T.G. Ksiazek, Emerg. Infect. Dis. 8 (2002) 987.

[[[[[

B 877 (2009) 1561–1567 1567

[4] K.B. Chua, W.J. Bellini, P.A. Rota, B.H. Harcourt, A. Tamin, S.K. Lam, T.G. Ksiazek,P.E. Rollin, S.R. Zaki, W. Shieh, C.S. Goldsmith, D.J. Gubler, J.T. Roehrig, B. Eaton,A.R. Gould, J. Olson, H. Field, P. Daniels, A.E. Ling, C.J. Peters, L.J. Anderson, B.W.Mahy, Science 288 (2000) 1432.

[5] K.B. Chua, J. Clin. Virol. 26 (2003) 265.[6] W.S. Tan, S.T. Ong, M. Eshaghi, S.S. Foo, K. Yusoff, J. Med. Virol. 73 (2004)

105.[7] S.T. Ong, K. Yusoff, C.L. Kho, J.O. Abdullah, W.S. Tan, J. Gen. Virol. 90 (2009)

392.[8] F. Yu, N.S. Khairullah, S. Inoue, V. Balasubramaniam, S.J. Berendam, L.K. Teh,

N.S.W. Ibrahim, S.A. Rahman, S.S. Hassan, F. Hasebe, M. Sinniah, K. Morita, J.Clin. Microb. (2006) 3134.

[9] A. Ndifuna, A.K. Waters, M. Zhou, E.W. Collisson, J. Virol. Methods 70 (1998) 37.10] Y.P. Tan, T.C. Ling, K. Yusoff, W.S. Tan, B.T. Tey, J. Microbiol. 43 (2005) 295.

[11] S. Vancan, E.A. Miranda, S.M.A. Bueno, Proc. Biochem. 37 (2002) 573.12] S. Sharma, G.P. Agarwal, Anal. Biochem. 288 (2001) 126.13] H. Kempe, A. Axelsson, B. Nilsson, G. Zacchi, J. Chromatogr. A 846 (1999) 1.14] M.S.G. Finette, Q.M. Mao, M.T.W. Hearn, Biotechnol. Bioeng. 58 (1) (1998) 35.15] C.W. Ho, T.K. Chew, T.C. Ling, S. Kamaruddin, W.S. Tan, B.T. Tey, Proc. Biochem.

41 (2006) 1829.16] E. Sulkowski, J. Mol. Recognit. 9 (1996) 389.

[17] Y.T.M. Ng, W.S. Tan, N. Abdullah, T.C. Ling, B.T. Tey, J. Chromatogr. A 1172 (2007)47.

18] Y.P. Tan, T.C. Ling, W.S. Tan, K. Yusoff, B.T. Tey, Proc. Biochem. 41 (2006) 874.19] P. Vincent, W. Dieryck, L. Maneta-Peyret, P. Moreau, C. Cassagne, X. Santarelli, J.

Chromatogr. B 808 (2004) 83.20] Y.K. Chang, H.A. Chase, Biotechnol. Bioeng. 49 (1996) 512.21] U.K. Laemmli, Nature 227 (1970) 680.22] B.T. Tey, K.H. Yong, H.P. Ong, T.C. Ling, S.T. Ong, Y.P. Tan, A. Ariff, W.S. Tan,

Biotechnol. Bioproc. Eng. 9 (2004) 374.23] M.M. Bradford, Anal. Biochem. 72 (1976) 248.24] R.H. Clemmitt, H.A. Chase, Biotechnol. Bioeng. 67 (2000) 206.

25] R.D. Johnson, R.J. Todd, F.H. Arnold, J. Chromatogr. A 725 (1996) 225.26] E.K.M. Ueda, P.W. Gout, L. Morganti, J. Chromatogr. A 988 (1) (2003) 1.27] J.W. Wong, R.L. Albright, N.-H. Wang, Sep. Purif. Methods 20 (1991) 49.28] H.A. Chase, J. Chromatogr. 297 (1984) 179.29] C.W. Ho, W.S. Tan, F.C. Chong, T.C. Ling, B.T. Tey, J. Microbiol. Biotechnol. (2008),

doi:10.4014/jmb.0804.254.