www.elsevier.com/locate/yprep

Protein Expression and Purification 37 (2004) 361–367

Intein-mediated fusion expression, high efficient refolding,and one-step purification of gelonin toxin

Chenyun Guoa, Zhuoyu Lia, Yawei Shia, Mingqun Xub, John G. Wisec,Wolfgong E. Trommerc, Jingming Yuana,*

a Key Laboratory of Chemical Biology and Molecular Engineering of Ministry of Education, Institute of Biotechnology,

Shanxi University, Taiyuan 030006, PR Chinab New England Biolabs, Berverly, MA 01915, USA

c Department of Chemistry, University of Kaiserslautern, Kaiserslautern 67653, Germany

Received 5 April 2004, and in revised form 14 June 2004

Available online 6 August 2004

Abstract

An open reading frame of gelonin (Gel), one of ribosome inactivating proteins, was inserted into the vector pBSL-C which con-

tains the coding region of chitin binding domain (CBD)-intein, resulting in the fusion expression of CBD-intein–Gel in Escherichia

coli BL21 (DE3) by the induction of IPTG. The fusion product formed an aggregate of the misfolded protein, commonly referred to

as inclusion bodies (IBs). The IBs were denatured and then refolded by step-wise dialysis. About 69% fusion protein was in vitro

refolded to native state in the presence of GSSG and GSH as monitored by size-exclusion HPLC. The refolded CBD-intein–Gel

was loaded onto chitin beads column equilibrated with 10mM Tris buffer, 500mM NaCl, pH 8.5, and about 2.4mgGel/L culture

with 96% homogeneity was directly eluted from the captured column by incubation at 25 �C under pH 6.5 for 48h based on intein C-

terminal self-cleavage. Western blot, ELISA, and in vitro inhibition of protein synthesis demonstrated that the bioactivity of recom-

binant Gel was comparable to that of native Gel purified from seeds. This implied that the purified Gel by this method is biologically

active and suitable for further studies.

� 2004 Elsevier Inc. All rights reserved.

Keywords: Gelonin; Intein; In vitro Refolding; Self-cleavage; One-step purification; Bioactivity

Protein self-splicing is a post-translational processing

event in which an internal protein segment, the intein,

can catalyze its own excision from a precursor protein

and concomitantly ligate the flanking regions, the ex-

teins, to form a mature protein [1,2]. Since the mecha-

nism of protein splicing was elucidated, the research

and application of splicing element, intein, have been de-veloped in the field of protein engineering, purification

of recombinant proteins in particular [3,4]. A conven-

tional method for recombinant protein expression and

purification is to make the target protein to be a fusion

1046-5928/$ - see front matter � 2004 Elsevier Inc. All rights reserved.

doi:10.1016/j.pep.2004.06.037

* Corresponding author. Fax: +86 351 7018268.

E-mail address: jmyuan@sxu.edu.cn (J. Yuan).

product harboring an affinity tag, such as polyhistidine

(His-tag) [5], Escherichia coli maltose-binding protein

(MBP) [6], Schistosoma glutathione S-transferase

(GST) [7], Staphylococcus protein A [8], and so on.

However, all of these methods suffer a drawback that

a site-specific protease is necessary to cleave the target

protein from its affinity tag, the high cost and uncom-pleted cleavage of these proteases have limited their ap-

plication. In recent years, a rapid, simple protein

expression, and purification system has been performed

by using intein self-cleavage. The intein reactivity can be

controlled to reach the cleavage reaction at either its C-

terminus or its N-terminus. If C-terminal Asn of intein is

substituted to Ala, the cleavage of fusion protein will

362 C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367

only occur at N-terminus of the intein. For example, hu-

man neurotrophin-3 (hNT-3) has been fused to N-termi-

nus of intein from Mycobacterium xenopi gyrA (Mxe

GyrA intein) and successfully purified on a chitin affinity

column by one-step manipulation, based on DTT induc-

ible peptide bond cleavage [9]. On the other hand, a tar-get protein can also be fused to the C-terminus of an

intein whose N-terminal CySH (Ser, Thr) was substitut-

ed to Ala, then the target protein can be purified by cy-

clization of the Asn residue at the C-terminus of intein

with a pH or temperature shift [10].

Gelonin (Gel) is one of the single chain plant ribo-

some inactivating proteins (RIPs). Due to lacking lec-

tin subunit, single chain RIPs are generally not ashighly toxic to intact cells as the double chain RIPs

such as ricin. Therefore, the single chain RIPs have

mostly been selected to construct the potent and spe-

cific immunoconjugates. Gel has been conjugated with

some monoclonal antibodies by gene fusion or chemi-

Fig. 1. A diagrammatic sketch for one-step purification of Gel by using inte

pBSL-C was blocked due to the mutation of CySH1–Ala1. The expression

proteins are washed off the beads directly with washing buffer. When the captu

peptide bond between Asn at C-terminus of intein and CySH at N-terminus o

succinic amide, and Gel was cleaved from the fusion product and eluted fro

cal cross to effectively kill the target cell, attempting to

cure tumors and autoimmune diseases [11,12]. Herein,

Gel gene was inserted into plasmid pBSL-C harboring

Ssp DnaB mini-intein attached CBD as a chitin affinity

tag and transformed into E. coli BL21(DE3). Then

pure Gel can be directly eluted from chitin beads col-umn by a pH shift and purified by single-step manip-

ulation without chemical cleavage or protease

digestion (Fig. 1). Thus, the intein purification system

provides a facile method for preparing recombinant

proteins.

Materials and methods

Bacterial strains, plasmids, reagents, and media

Escherichia coli BL21(DE3) was stored in this labora-

tory at �70 �C. Vector pBSL-C was generously provided

in C-terminal self-cleavage The N-terminal cleavage of intein in vector

product, CBD-intein–Gel, can bind to chitin beads and the unbound

red beads are incubated at pH 6.5, nucleophilic attack will occur at the

f target protein. Therefore, the Asn was transformed to an intermediate,

m the beads directly.

C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367 363

by Dr. Xu in New England Biolabs (NEB, Beverly,

USA). PUC-Gel, native Gel, and mouse anti-Gel serum

are kindly provided by Prof. Dr. Trommer in Chemistry

Department of Kaiserslautern University, Germany. All

enzymes used, DNA and protein markers as well as chi-

tin beads are from NEB. Horseradish peroxidase-conju-gated goat anti-mouse immunoglobulin G was

purchased from Sigma (Missouri, USA). Agarose and

all other chemicals are of analytical grade. Cells were

grown on LB liquid or on LB solid medium as described

by Sambrook and Russell [13].

Construction of plasmid pBSL-Gel and expression of

recombinant in E. coli

To fuse the target gene to the C-terminus of CBD-in-

tein reading frame, the Gel gene (753bp) was amplified

from plasmid pUC-Gel by polymerase chain reaction

(PCR) with the following primers which, respectively,

contained unique restriction sites for AgeI (forward)

and PstI (reverse) to facilitate cloning: forward primer:

5 0-GGT GGT ACC GGT GGC CTG GAT ACCGTG AGC-3 0, reverse primer: 5 0-GGT GGT CTG

CAG TTA TTT CGG ATC TTT ATC G AC-3 0. The

conditions used were: 95 �C for 5min, 30 cycles of

(94 �C 30s, 52 �C 30s, and 72 �C 1min), and a final ex-

tension of 72 �C for 10min. The PCR product was puri-

fied and digested with corresponding enzymes,

subsequently inserted into the vector pBSL-C at AgeI

and PstI sites of MCS to form the recombinant plasmidpBSL-Gel [14]. After restriction enzyme digestion and

DNA sequencing verification, the pBSL-Gel was trans-

formed into competent E. coli strain BL21(DE3) by cal-

cium chloride method [13]. The resulting engineered

strain E. coli BL21(DE3)/pBSL-Gel was grown in LB

medium supplemented with 100g/mL ampicillin at

37 �C until the optical density (OD) at 600nm reached

0.5–0.6, and then induced by 0.5mM IPTG at 12 �Cfor 16h. The fusion expression product, CBD-intein–

Gel was confirmed by SDS–PAGE and Western blot

analysis.

Refolding of CBD-intein–Gel aggregate

The cultured cells were harvested by centrifugation at

6000rpm for 10min and the cell pellets from 1L IPTGinduced culture were suspended and sonicated in lysis

buffer A (10mM Tris–HCl, 500mM NaCl, 1mM

PMSF, pH 8.5, and 0.1% Triton X-100). The lysate

was then centrifuged at 12,000rpm for 30min. Because

the fusion protein almost completely existed in the form

of IBs by 10% SDS–PAGE analysis, the precipitate was

washed with buffer A containing 4M urea and centri-

fuged at 12,000rpm for 30min. This step was repeatedtwice to remove the cell debris and other impurities.

Then the IBs were dissolved in buffer B (10mM Tris–

HCl, 500mM NaCl, pH 8.5) containing 8M urea to

reach a protein concentration of about 0.3mg/mL. The

dissolved precipitate was kept for 2–3h at room temper-

ature to completely solubilize the aggregate. After cen-

trifugation at 12,000rpm for 30min, about 10mL

supernatant was step-wisely dialyzed against 1L bufferB containing 4, 2, and 1M urea sequentially at 4 �Cfor about 17–18h each. Finally, the protein solution

was dialyzed against 500mL buffer B containing

0.2mM GSH, 0.02mM GSSG, and 0.5M LL-Arg for in

vitro refolding at 4 �C for 24h. After centrifugation at

12,000rpm for 30min, the supernatant was further puri-

fied in the next step. The in vitro refolding process was

simultaneously monitored by size-exclusion HPLCequipped with a SW 300 gel column (7.8 · 300mm) with

10mM Tris buffer, pH 8.5, as the mobile phase at a flow

rate of 1.0mL/min, and absorbance at 280nm was mon-

itored with a UV detector.

One-step purification of Gel by intein C-terminal self-

cleavage

The refolded protein supernatant was applied to a

5mL chitin beads column pre-equilibrated with buffer

B at 4 �C. The captured column was washed with 20 col-

umn volumes of buffer B to remove unbound proteins

and flushed with 15mL cleavage buffer (10mM Tris–

HCl, 500mM NaCl, pH 6.5) before the outlet was

closed. After incubating the captured column at 25 �Cfor 48h, 5–10mL cleavage buffer was flowed throughthe column, the fractionation of Gel was collected ac-

cording to the absorbance at 280nm in LKB protein pu-

rification chromatography system and then examined by

12% SDS–PAGE. The chitin resin could be regenerated

with buffer C (10mM Tris–HCl, 500mM NaCl, and 1–

2% SDS, pH 8.5).

Determination of protein concentration and purity

Protein concentration was determined by Bradford

method using bovine serum albumin as the standard

[15]. Protein expression level and protein purity were es-

timated by comparing the intensity of Coomassie bril-

liant blue staining of samples run on SDS–PAGE. The

stained gel was quantified by gel document scanning

(GDS) with BIO-PROFIL Bio-ID V99.01 software.

Western blot and ELISA

Fusion protein CBD-intein–Gel and purified Gel

were run on 12% SDS–PAGE and transferred electro-

phoretically to a nitrocellulose membrane using a mini

trans-blot electrophoretic transfer cell (Bio-Rad) at

100V for 1h in transfer buffer (25mM Tris, 192mM gly-cine, and 0.025% SDS, pH 8.3). After blocking with 1%

BSA, the membrane was incubated with mouse anti-Gel

Fig. 2. SDS–PAGE analysis of the expression products from engi-

neered strain E. coli BL 21 (DE3)/pBSL-Gel. Lane 1, culture from host

strain; lane 2, uninduced culture from recombinant strain; lane 3,

induced culture from recombinant strain; lane 4, supernatant of the

cell lysate after centrifugation; and lane 5, precipitate of the cell lysate

after centrifugation.

364 C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367

serum diluted 1:500 and the horseradish peroxidase-con-

jugated goat anti-mouse immunoglobulin G (1:10000).

Finally, the sample membrane was subjected to 3,3 0-di-

aminobenzidine tetrahydrochloride (DAB) solution to

visualize the antigen–antibody complex. For ELISA,

the color in the microtiter wells was developed withDAB as substrate for 40min and measured at 405nm us-

ing a microtiter plate reader.

Reticulocyte lysate activity

The bioactivity of Gel on inhibition protein synthesis

was assessed using a cell-free rabbit reticulocyte lysate

protein translation system purchased from GIBCO(Grand Island, NY, USA). After diluted in PBS, 5llGel sample and 40ll cell-free rabbit reticulocyte lysate

containing 10mM creatine phosphokinase, and 0.5mM

KClwere added into 96wells on themicrotiter plate, incu-

bating at 37 �C for 5min, then 10ll master mixture con-

taining 10mM creatine phosphate, 0.5mM MgCl2,

79mM KCl, 5lCi/mL [14C]valine, and amino acid mix-

tures were added and continuously incubated at 37 �Cfor 10min. Five microliter reaction mixture was taken

from the wells and added into 1mL ice chilled water con-

taining 500ll valine (1mg/mL), incubating 10min at

37 �C. The assay was repeated four times at each protein

concentration. The reaction was stopped by adding 25%

trichloroacetic acid. The protein obtained by the glassmi-

crofiber filter was dried and counted by liquid scintillator.

Results and discussion

Cloning and expression of Gel gene

It was shown from the results of double-enzymatic di-

gestion and DNA sequence analysis for plasmid pBSL-

Gel that Gel gene fragment was correctly inserted intovector pBSL-C and suitable for fusion expression in E.

coli (data not shown). The engineered strain E. coli

BL21(DE3)/pBSL-Gel can express the target product,

CBD-intein–Gel in the form of IBs only as estimated

by SDS–PAGE (Fig. 2) or Western blot analysis (data

not shown). To explore its soluble expression, various

factors were examined, including the optical density of

culture at induction (OD600nm 0.4–0.8), the concentra-tion of inducer IPTG (0.1–1.0 mM), induction tempera-

ture (37, 25, 16, and 12 �C) [16–18] as well as

co-expression with molecular chaperone GroESL [19].

Unfortunately, no significant amount of the soluble tar-

get product was observed under any of the above condi-

tions. Over-expression of a recombinant harboring the

gene of eukaryotic cells in host strain E. coli often results

in the formation of biologically inactive aggregates. For-tunately, it is possible to make the aggregates soluble by

some refolding methods [20].

Refolding of CBD-intein–Gel aggregate

It is well known that the IBs occurred during recom-

binant expression in bacteria as a random protein aggre-gate in an unfolded, partially folded or inactive

conformational state, which can be in vitro refolded to

partially recover its active and native state under the de-

fined conditions [20]. Due to the retention time of linear

and globular macromolecules being quite different on

size-exclusion chromatography, the whole process of in

vitro refolding was accurately monitored by size-exclu-

sion HPLC while the urea concentration was reducedin step-wise dialysis (Fig. 3). The IBs were initially

washed with 4M urea to remove the cell debris and

other impurities before completely solubilized in 8M

urea. As can be seen in Fig. 3, the protein was in linear

or random state under strong chaotropic solvent condi-

tion (purple trace, 2). There was little difference for the

chromatograms (purple and green traces, 2 and 3) in

the presence of 8 or 4M urea, which indicated that thecomplete unfolded polypeptide was still dominative with

4M urea [21]. However, when the characteristics of sol-

vent surrounded denatured protein were changed with

decreasing urea concentration, nucleation of protein

conformation seemed to be appearing after dialyzed

against 2M urea, as evidenced by the appearance of a

new peak at the retention time of 7.48min (turquoise

trace, 4). The unsymmetrical chromatogram profile indi-cated that the protein could partially be folded to an in-

termediate, and there existed a competition between the

first-order (correct) folding reaction and the higher-or-

der aggregation reaction. After dialyzed against 1M

urea, the majority of the target protein was refolded as

indicated by the main peak around 11min (pink trace,

5), and also a significant amount of protein seem to be

in a misfolded or aggregate state as indicated by thepeak around 5min (pink trace, 5). The misfolding or ag-

gregation of the protein could be from the association of

Fig. 3. Size-exclusion HPLC chromatograms of in vitro refolding process for CBD-intein–Gel. Conditions: Equipment, HPLC 1525 (Waters, USA);

Column, SW300 (7.8 · 300mm); Mobile phase, 10mM Tris–HCl, 0.5mM NaCl, pH 8.5; Flow rate, 1mL/min; Injected volume, 25lL; Detection:

A280nm; Black trace (1): buffer containing 8M urea; Purple trace (2): inclusion bodies solubilized in buffer with 8M urea; Green trace (3): fusion

protein after dialyzed against buffer with 4M urea; Turquoise trace (4): fusion protein after dialyzed against buffer with 2M urea; Pink trace (5):

fusion protein after dialyzed against buffer with 1M urea; and Brown trace (6): completely refolded fusion protein after dialyzed against buffer

without urea (0M urea).

Fig. 4. Effect of pH on the cleavage of CBD-intein–Gel. Lane 1,

expressed products (IBs); lane 2, in vitro refolded CBD-intein–Gel; and

lanes 3–7, cleavage products of lane 2 at pH 6.0, 6.5, 7.0, 7.5, and 8.0,

respectively. (CBD-intein–Gel, 55kDa; Gel, 28kDa).

C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367 365

hydrophobic surfaces that were exposed in folding inter-

mediate or improper disulfide bridge formation [22]. The

competition of misfolding may kinetically limit the pro-

tein to be folded into its native state [23]. In our study,

the majority of the protein was shifted to the correct

conformation after urea was completely removed by di-alysis in the refolding buffer (10mM Tris–HCl, 500mM

NaCl, pH 8.5, 0.2mM GSH, 0.02mM GSSG, and 0.5M

LL-Arg) (brown trace, 6). Reduced and oxidized glutathi-

one (GSH and GSSG) are commonly used as oxido-

shuffling reagents, because thiol-disulfide exchange

reactions are rapidly reversible. The oxido-shuffling re-

agents can increase both the rate and the yield of correct

disulfide bond formation by rapid reshuffling of improp-er disulfide bonds [23]. LL-Arg contains a guanidino

group and it may play a role in suppressing aggregation

of the protein during refolding [24]. In conclusion, the

results from size-exclusion HPLC analysis figuratively

demonstrated in vitro refolding process during the

step-wise dialysis, which was also confirmed by follow-

ing bioactivity assay.

In vitro pH inducible cleavage of fusion protein

The target protein fused with intein can be cleaved at

the C-terminus of intein by pH shift based on the back-

ground of the recombinant plasmid. To investigate the

optimal pH for the cleavage reaction, the refolded

CBD-intein–Gel was incubated in vitro at five different

pH values, ranging from pH 6.0 to 8.0 at 25 �C. It wasshown from the results of SDS–PAGE in Fig. 4 that

the cleavage reaction was completely inhibited at

pHP7.5 and was gradually increased at either pH 6.0

or 7.0, while the optimal value for the yield and purity

of Gel seemed to be at pH 6.5. During the cleavage pro-

cess, three bands corresponding to fusion protein

(55kDa), CBD-intein (27kDa), and Gel (28kDa) should

occurred by SDS–PAGE analysis. However, the molec-

ular weights of Gel and its fusion partner are too close

to be separated on the gel plate. For confirming the tar-get protein Gel, a positive product at 28kDa band was

occurred by Western blot analysis (data not shown). It

has been speculated that pH sensitivity of the intein

aroused from protonation of the highly conserved pen-

ultimate histidine residue (pKa, approximately 6.5) of

the intein C-terminus [25]. Mutation of the penultimate

histidine inhibited intein C-terminal cleavage, which also

indicated the importance of the conserved histidine res-idue for intein C-terminal cleavage [26–28].

One-step purification of gelonin

The refolded CBD-intein–Gel was loaded onto a chi-

tin beads column at pH 8.5. After washing the captured

column with the equilibrated buffer, about 90% of

fusion protein was bound to the chitin beads (Fig. 5,lane 3). Pure Gel with 96% homogeneity was directly

eluted from the column after incubation at pH 6.5,

Fig. 5. SDS–PAGE for one-step purification of recombinant Gel on

chitin beads column. Lane 1, pellets after cell lysate; lane 2:

supernatant after renaturation and centrifugation; lane 3: flow-through

from chitin beads column; lane 4, recombinant Gel eluted from the

column after cleavage incubation at pH 6.5.

Table 2

ELISA analysis of recombinant Gel

Sample Recombinant Gel Native Gela BSAb

OD405nm 0.24 ± 0.01 0.30 ± 0.01 0.01 ± 0.01

Note. The data were the average values obtained by multiply tests.a Positive control.b Negative control.

Fig. 6. Comparison of inhibitory activity of native and recombinant

Gel in the cell-free protein synthesis assay. IC50 of native Gel is at

15pM; IC50 of recombinant Gel is at 20pM.

366 C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367

25 �C for 48h (Fig. 5, lane 4). It is indicated from Table 1

that the refolding percentage of CBD-intein–Gel

reached about 69%, but Gel yield is only 5%, compared

with the quantity of inclusion bodies. The cleavage reac-

tion as described above was so incomplete that about

20% fusion product still stuck in the affinity column as

occurred by SDS–PAGE analysis with a small portion

of affinity beads. There was not much improvement inthe yield of Gel even if the cleavage reaction on-column

was extended to longer time. However, this system re-

quires no protease or chemical cleavage to obtain the

target protein from its fusion product, compared with

other affinity tag expression systems [29,30].

Bioactivity of recombinant Gel

It was demonstrated from ELISA (Table 2) and Wes-

tern blot analysis (data not shown) that the recombinant

fusion product or Gel only reveals the positive immuno-

reactivity. The bioactivity of recombinant Gel was fur-

ther confirmed with the functional analysis by using

inhibition assay of cell-free protein synthesis in rabbit

reticulocyte lysates. As shown in Fig. 6, native Gel in-

hibits cell protein synthesis by 50% at 15pM, whilstthe recombinant Gel is at 20pM with 50% inhibition.

The result indicated that the bioactivity of the recombi-

nant Gel was comparable to that of native Gel.

Table 1

Purification of recombinant Gel from 1L E. coli cell lysatea

Purification steps Total proteins (mg)

Inclusion bodies (in 8mol urea) 76

Refolding supernatant 51

Chitin column and intein-mediated cleavage 2.4

a Step and overall yields were calculated starting from the pure inclusion bo

method.b Purity is defined as the percentage of target protein in the purified prot

Conclusions

In this report, a fusion protein, CBD-intein–Gel was

over-expressed in the form of inclusion bodies in E. coli

and was successfully in vitro refolded through 8M urea

denaturation and step-wise dialysis. The refolding pro-

cess along the renaturation was concomitantly moni-

tored by size-exclusion HPLC, and the refoldingrecovery of the fusion product reaches about 69%.

About 2.4mgGel/L culture with 96% homogeneity was

obtained by single-step purification on chitin beads col-

umn, without any chemicals or protease treatment.

Moreover, immunoreactivity and functional assay also

demonstrated that the recombinant Gel possessed the

same properties as the natural Gel purified from seeds.

This intein-mediated purification scheme would provide

Target protein (mg) Purityb (%) Yield (%)

46 (CBD–intein–Gel) 60 100

32 (CBD–intein–Gel) 63 69

2.3 (Gel) 96 5

dy preparation and protein concentration was determined by Bradford

ein preparation.

C. Guo et al. / Protein Expression and Purification 37 (2004) 361–367 367

a convenient and economic method to prepare other

recombinant proteins.

Acknowledgments

This work was supported by the Natural Science

Foundation of China (NSFC: 30270292). We also thank

Dr. Tao Yuan in Aventis Pasteur, Canada, for his help-

ful suggestion.

References

[1] M.Q. Xu, M.W. Southworth, F.B. Mersha, L.J. Hornstra, F.B.

Perler, In vitro protein splicing of purified precursor and the

identification of a branched intermediate, Cell 75 (1993) 1371–

1377.

[2] F.B. Perler, E.O. Davis, G.E. Dean, F.S. Gimble, W.E. Jack, N.

Neff, C.J. Noren, J. Thorner, M. Belfort, Protein splicing

elements: inteins and exteins a definition of terms and the

recommended nomenclature, Nucleic Acids Res. 22 (1994) 1125–

1127.

[3] F.B. Perler, E. Adam, Protein splicing and its applications, Curr.

Opin. Biotechnol. 11 (2000) 377–383.

[4] S.F. Singleton, R.A. Simonette, N.C. Sharma, A.I. Roca, Intein-

mediated affinity-fusion purification of the Escherichia coli RecA

protein, Protein Express. Purif. 26 (2002) 476–488.

[5] M.W. Van Dyke, M. Sirito, M. Sawadogo, Single-step purifica-

tion of bacterially expressed polypeptides containing an oligo-

histidine domain, Gene 111 (1992) 99–104.

[6] C. Guan, P. Li, P.D. Riggs, H. Inouye, Vectors that facilitate the

expression and purification of foreign peptides in Escherichia coli

by fusion to maltose-binding protein, Gene 67 (1998) 21–30.

[7] D.B. Smith, K.S. Johnson, Single-step purification of polypeptides

expressed in Escherichia coli as fusions with glutathione S-

transferase, Gene 67 (1988) 31–40.

[8] B. Nilsson, L. Abrahmsen, Fusions to Staphylococcal protein A,

Mehtods Enzymol. 185 (1990) 144–161.

[9] Z.Y. Li, J.H. Fan, J.M. Yuan, Single-column purification of

recombinant human neurotrophin-3 (hNT3) by using the intein

mediated self-cleavage system, Biotechnol. Lett. 24 (2002) 1723–

1727.

[10] H.E. Humphries, M. Christodoulides, J.E. Heckels, Expression of

the class 1 outer-membrane protein of Neisseria meningitidis in

Escherichia coli and purification using a self-cleavable affinity tag,

Protein Express. Purif. 26 (2002) 243–248.

[11] M.G. Rosenblum, L.H. Cheung, Y. Liu, H.W. Marks, Design,

expression, purification, and characterization, in vitro and in vivo,

of an antimelanoma single-chain Fv antibody fused to the toxin

Gel, Cancer Res. 63 (2003) 3995–4002.

[12] I.L. Urbatsch, P.K. Sterz, K. Peper, W.E. Tommer, Antigen-

specific therapy of experimental myasthenia gravis with acetyl-

choline receptor–gelonin conjugates in vivo, Eur. J. Immunol. 23

(1993) 776–779.

[13] J. Sambrook, D.W. Russell, Molecular Cloning: A Laboratory

Manual, third ed., Cold Spring Harbor Laboratory Press, NY,

2001.

[14] S. Mathys, T.C. Evans, I.C. Chute, H. Wu, S. Chong, J. Benner,

X.Q. Liu, M.Q. Xu, Characterization of a self-splicing mini-intein

and its conversion into autocatalytic N- and C-terminal cleavage

elements: facile production of protein building blocks for protein

ligation, Gene 231 (1999) 1–13.

[15] M.M. Bradford, A rapid and sensitive method for the quantita-

tion of microgram quantities of protein utilizing the principle of

protein-dye binding, Anal. Biochem. 72 (1976) 248–254.

[16] C.H. Schein, H.M. Noteborn, Formation of soluble recombinant

proteins in Escherichia coli is favored by lower growth tempera-

ture, Bio/Technology 6 (1988) 291–294.

[17] S. Cabilly, Growth at sub-optimal temperature allows the

production of functional antigen-binding Fab fragments in

Escherichia coli, Gene 85 (1989) 553–557.

[18] E. Kopetzki, G. Schumacher, P. Buckel, Control of formation of

active soluble or inactive insoluble baker�s yeast alpha-glucosidasePI in Escherichia coli by induction and growth conditions, Mol.

Gen. Genet. 216 (1989) 149–155.

[19] Z. Zhang, L.P. Song, M. Fang, F. Wang, D. He, R. Zhao, J. Liu,

Z.Y. Zhou, C.C. Yin, Q. Lin, H.L. Huang, Production of soluble

and functional engineered antibodies in Escherichia coli improved

by FkpA, Biotechniques 35 (2003) 1041–1042.

[20] R. Rudolph, H. Lilie, In vitro folding of inclusion body proteins,

FASEB J. 10 (1996) 49–57.

[21] H. Yoshii, T. Furuta, T. Yonehara, D. Ito, Y.Y. Linko, P.

Linko, Refolding of denatured/reduced lysozyme at high concen-

tration with diafiltration, Biosci. Biotechnol. Biochem. 64 (2000)

1159–1165.

[22] J. Buchner, U. Brinkmann, I. Pastan, Renaturation of a single-

chain immunotoxin facilitated by chaperonins and protein disul-

fide isomerase, Bio/Technology 10 (1992) 682–685.

[23] A. Mitraki, J. King, Protein folding intermediates and inclusion

body formation, Bio/Technology 7 (1989) 690–697.

[24] T. Arakawa, K. Tsumoto, The effects of arginine on refolding of

aggregated proteins: not facilitate refolding, but suppress aggre-

gation, Biochem. Biophys. Res. Commu. 304 (2003) 148–152.

[25] D.W. Wood, W. Wu, G. Belfort, V. Derbyshire, M. Belfort, A

genetic system yields self-cleaving inteins for bioseparations, Nat.

Biotech. 17 (1999) 889–892.

[26] M.Q. Xu, F.B. Perler, The mechanism of protein splicing and its

modulation by mutation, EMBO J. 15 (1996) 5153–5164.

[27] S. Chong, S. Yang, H. Paulus, J. Benner, F.B. Perler, M.Q. Xu,

Protein splicing involving the Saccharomyces cerevisiae VMA

intein: the steps in the splicing pathway, side reactions leading to

protein cleavage and establishment of an in vitro splicing system,

J. Biol. Chem. 271 (1996) 22159–22168.

[28] S. Chong, K.S. Williams, C. Wotkowies, M.Q. Xu, Modulation

of protein splicing of the Saccharomyces cerevisiae vacuolar

membrane ATPase intein, J. Biol. Chem. 273 (1998) 10567–

10577.

[29] S. Chong, F.B. Mersha, D.G. Comb, M.E. Scott, D. Landry,

L.M. Vence, F.B. Perler, J. Benner, R.B. Kucera, C.A.

Hirvonen, J.J. Pelletier, H. Paulus, M.Q. Xu, Single-column

purification of free recombinant proteins using a self-cleavable

affinity tag derived from a protein splicing element, Gene 192

(1997) 271–281.

[30] S. Chong, G.E. Monotello, A. Zhang, E.L. Cantor, W. Liao,

M.Q. Xu, J. Benner, Utilizing the C-terminal cleavage activity

of a protein splicing element to purify recombinant proteins in

a single chromatographic step, Nucleic Acids Res. 26 (1998)

5109–5115.