| Glycobiology | Pages |
High-level expression of the endo-[beta]-N-acetylglucosaminidase F2 gene in E.coli: one step purification to homogeneity
Introduction
Results and discussion
Materials and methods
Acknowledgments
Abbreviations
References
High-level expression of the endo-[beta]-N-acetylglucosaminidase F2 gene in E.coli: one step purification to homogeneity
The Endo F2 gene was overexpressed in E.coli as a fusion protein joined to the maltose-binding protein. MBP-Endo F2 was found in a highly enriched state as insoluble, inactive inclusion bodies. Extraction of the inclusion bodies with 20% acetic acid followed by exhaustive dialysis rendered the fusion protein active and soluble. MBP-Endo F2 was digested with Factor Xa and purified on Q-Sepharose. The enzyme was homogeneous by SDS-PAGE, and appeared as a single symmetrical peak on HPLC. Analysis of the amino-terminus demonstrated conclusively that recombinant Endo F2 was homogeneous and identical to the native enzyme.
Key words: Endo F2/F.meningosepticum/MBP-fusion protein/inclusion bodies
Introduction
Flavobacterium meningosepticum expresses three different endo-[beta]-N-acetylglucosaminidases designated Endo F11, Endo F2, and Endo F3 (Plummer and Tarentino, 1991). All three enzymes hydrolyze the di-N-acetylchitobiose linkage of asparagine-linked glycans but differ considerably in their substrate specificity, depending on the nature of the attached oligosaccharide moiety (Tarentino and Plummer, 1994).
X-Ray crystallographic analysis of these three endoglycosidases affords a unique opportunity to study protein structure as a function of substrate specificity in a well-defined class of proteins. Endo F1, which is secreted in large amounts by F.meningosepticum, has been purified to homogeneity and its crystal structure determined to 2.0 Å resolution (Van Roey et al., 1994). Endo F2 and Endo F3 are produced in much lower amounts in F.meningosepticum, and sufficient protein for crystallization necessitated overexpression of these genes in E.coli. We recently described the amplification of the Endo F3 gene in E.coli (Tarentino et al., 1995). Endo F3 has since been crystallized, and a comparative structural analysis with Endo F1 completed (Van Roey et al., 1997).
In the past year we attempted to overexpress Endo F2 as a fusion protein linked to the maltose-binding protein of E.coli. The gene had been properly inserted into the pMAL system (Guan et al., 1988) and the DNA sequence was verified, but little expression was obtained with clones from a variety of E.coli strains (TB-1, XL-1, DH5[alpha]) grown in a standard Tryptone broth.
In this report we describe conditions for facile overexpression of the Endo F2 gene in E.coli BL21(DE3), and present a simple isolation method for obtaining large amounts of pure protein from inclusion bodies suitable for crystallographic analysis.
Results and discussion
The Endo F2 gene was fused to the mal E gene, which codes for the maltose-binding protein (MBP), and expressed in BL21(DE3) (Novogen) as a MBP-fusion protein following induction from the strong P(tac) promotor by IPTG. A factor Xa recognition site engineered into the pMAL vector (New England Biolabs) behind the polylinker allows the MBP to be removed, and the recombinant protein to be recovered intact with the proper amino-terminus.
A high level of cytoplasmic expression of MBP-Endo F2 was achieved using the expression strain BL21 (DE3) (Studier et al., 1990), in conjunction with an enriched growth medium designated superbroth (see Materials and methods). In preliminary studies we found that cells grown in superbroth produced much larger amounts of fusion protein than those grown in tryptone broth (Ausubel, 1989; Moore et al., 1993). The MBP-Endo F2 fusion protein was obtained, however, in an inactive insoluble form as post-cell lysate inclusion bodies. As shown in Figure
Figure 1. SDS-PAGE at various stages of Endo F2 isolation. A 4-20% gradient gel was loaded as follows. (A) Lane 1, post-cell lysate inclusion bodies showing MBP-Endo F2 fusion protein; lane 2, acid-extracted and neutralized MBP-Endo F2 from inclusion bodies; lane 3, Factor Xa-digested MBP-Endo F2. (B) Lane 1, molecular mass markers; lane 2, Q-Sepharose-purified Endo F2
A modification of the procedure of Hui et al. (1993) was used successfully to render the fusion protein active and in soluble form. The inclusion bodies were treated with 20% acetic acid containing 1.5 mg/ml of the detergent CHAPS for 2.5 h at 37°C. As shown in Figure To obtain Endo F2 in an homogeneous state, the Factor Xa digest was chromatographed on Q-Sepharose at pH 8.0. In this system (Figure
Because of its known stability to low pH (Tarentino and Plummer, 1994), Endo F2 is an ideal candidate for acid extraction and renaturation from inclusion bodies. It should be noted that thiols were not needed in the reactivation protocol since Endo F2 does not contain cysteine residues. The specific activity of recombinant Endo F2, determined with a porcine fibrinogen biantennary glycopeptide substrate, was found to be the same as that for native Endo F2, demonstrating that acid extraction did not adversely affect enzyme activity.
Figure 2. Q-Sepharose chromatography of Factor Xa-digested MBP-Endo F2. MBP-Endo F2 (125 A280 units) digested with Factor Xa was purified on a 1.5 × 12 cm Q-Sepharose column equilibrated in 10 mM Tris-chloride, pH 8.0. The flow rate was 3 ml/min and 4.5 ml fractions were collected. Endo F2 activity eluted in the nonretarded fraction (tubes 3-19), and nowhere else. The gradient (see Materials and methods) was begun at the arrow.
Figure 3. Analytical chromatography of Q-Sepharose purified Endo F2 on a Waters Protein Pak SP (HR15) column. An aliquot of Q-Sepharose-purified Endo F2 was applied to a 0.5 × 10.5 cm Protein-Pak SP (HR15) column equilibrated in 10 mM sodium acetate, pH 6.0. The flow rate was 33 ml per h and 0.55 ml fractions were collected. A linear gradient was developed to 0.20 M NaCl over 2.5 h. By changing growth conditions, such as growing the cells at 25°C instead of 37°C, and induction with 0.1 mM IPTG instead of 1 mM IPTG, the synthesis of Endo F2 could be directed to the cytoplasm in a soluble form. However, the expression of Endo F2 was lower, and purification of the enzyme was more complicated from the soluble cell lysate than from the relatively purer inclusion body fraction. Typically 1 l of superbroth yielded 49 A280 units of pure Endo F2, corresponding to approximately 24 mg of enzyme protein. The BL21(DE3) strain, which produces T7 RNA polymerase upon induction with IPTG, is intended for use with pET vectors containing T7 promoters. It is interesting that T7 polymerase recognized the hybrid P(tac) promoter in pMAL efficiently. Our results suggest that BL21(DE3) could be a useful expression strain for other plasmids containing the P(tac) promoter.
Recombinant F2:
AVNLSNLIAYKN....
Native F2:
AVNLSNLIAYKN....
Materials and methods
Construction of Endo F2 plasmid vector
The Endo F2 gene was amplified from a pBluescript clone (Tarentino et al., 1993) using standard PCR methodology. The overall strategy used the pMAL expression system (New England Biolabs) and was developed as follows: oligonucleotide primers were designed to position an FspI site on the 5[prime]-end at the junction of the signal peptide and the amino-terminus of Endo F2, and an XbaI site on the 3[prime]-end just beyond the first termination codon.
Primer 1: 5[prime]-GTTACAGTATGCGCAGTTAATCTAAGCAAT-3[prime]
Primer 2: 5[prime]-AAAAGTATTGTTCTAGAGTTGGTATAAAAT-3[prime]
The optimized conditions for PCR were identical to those reported previously for the corresponding Endo F3 gene (Tarentino et al., , 1995). The PCR product was purified from an 0.5% agarose gel using a Centricon 100 spin filter (Amicon) to remove Vent polymerase and the primer. The newly introduced linkers were double-digested with 30 U each of FspI and XbaI for 6 h at 37°C.
The Endo F2 gene was fused to the mal E gene in the vector as follows. The expression vector pMal-C2 (20 µg) was double-digested with XmnI and XbaI for 4 h at 37°C. The restricted Endo F2 PCR product and the vector were purified by agarose gel electrophoresis using an Amicon Centriluter, and concentrated with a Centricon-100 spin filter. Ligation and transformation of E.coli BL21(DE3) (Novogen), were accomplished from standard protocols.
Isolation of active MBP-Endo F2 fusion protein
One liter of superbroth (32 g tryptone, 20 g yeast extract, 5 g NaCl, and 5 ml 1 N NaOH per liter), supplemented with ampicillin (100 µg/ml), was inoculated with a 10 ml overnight culture grown from a single BL21(DE3) colony containing pMAL C-2/Endo F2 plasmid. The cells were grown at 37°C until the absorbance reached 0.2-0.30, and then induced with 1.0 mM IPTG at 37°C for 6 h. The cells were harvested by centrifugation and the cell pellet was resuspended in 80 ml of cold 20 mM Tris·chloride pH 8.0, 0.1 M NaCl, 10 mM EDTA (lysis buffer). The cells were lysed by sonication as described previously (Tarentino et al., 1993), and centrifuged at 14,000 × g × 20 min. The soluble cell lysate showed little or no Endo F2 activity, and gave no evidence for the presence of an amplified fusion protein (MBP-Endo F2) migrating at about 70 kDa on SDS-PAGE. This fraction was discarded. The insoluble material, which was designated as the inclusion-body fraction, consisted mainly of inactive MBP-Endo F2 as determined by SDS-PAGE. The inclusion-body fraction was washed once with 40 ml lysis buffer and recentrifuged as described above. The insoluble material was then resuspended in 20% acetic acid containing 1.5 mg CHAPS/ml and incubated at a slow speed on a rotary shaker at 37°C for 2.5 h. The extract was centrifuged at 14,000 × g × 20 min, and the supernatant fraction containing the solubilized MBP-Endo F2 fusion protein was dialyzed exhaustively as follows: three changes in 2 days of 6 l each of distilled water, followed by one 6 l change of 10 mM Tris-chloride, pH 8.0.
Isolation of homogeneous Endo F2 from MBP-Endo F2
The denatured/renatured MBP-Endo F2 fraction in 10 mM Tris-chloride, pH 8.0 was adjusted to 2 mM CaCl2, and to 0.1 M NaCl. The fusion protein (125 absorbance 280 units in 54 ml) was then treated with 70 µL of a solution of 1 mg Factor Xa/ml (New England Biolabs), and digested at room temperature for 3 days. After centrifugation at 14,000 × g × 20 min to remove a small amount of insoluble material that formed during the digestion, the solution was dialyzed against 4 l of 10 mM Tris-chloride, pH 8.0.
The dialyzed Factor Xa-digested MBP-Endo F2 extract was applied to a 1.5 ×12 cm of Q-Sepharose equilibrated in 10 mM Tris-chloride, pH 8.0. Endo F2 appeared in the nonretarded fraction in an homogeneous state. Fractions 3-19 which were enzymatically active (Figure
Miscellaneous techniques
Endo F2 activity was assayed quantitatively on HPLC with a di-dansyl porcine fibrinogen glycopeptide, dnsVEN(CHO)-dns K, in 0.1 M sodium acetate, pH 4.75 as described previously (Plummer et al., 1996). Column fractions were assayed qualitatively by paper chromatography with the same substrate. The intensity of the fluorescent product was scored by comparison to a standard. Automated protein microsequencing was performed with a model 477A Applied Biosystems pulsed liquid sequenator equipped with a model 120A amino-acid analyzer. Edman protein microsequencing from PVDF membranes following SDS-PAGE (Laemmli, 1970) was described previously (Tarentino et al., 1993).
Acknowledgments
We express our appreciation to Dr. Li-Ming Changchien for performing amino-terminal sequencing at the Wadsworth Center's Biochemistry Core Facility. This work was supported in part by Grant 30471 from the National Institute of General Medical Sciences, USPHS/DHH.
Abbreviations
Endo, endo-[beta]-N-acetylglucosaminidase (EC 3.2.1.96); IPTG, isopropyl-[beta]-thioglucoside; MBP, maltose-binding protein; SDS-PAGE, sodium dodecyl sulfate-polyacrylamide gel electrophoresis; PCR, polymerase chain reaction; CHAPS, (3-[(cholamidopropyl) dimethyl-ammonio]1-propane sulfonate.
References
This page is run by Oxford University Press, Great Clarendon Street, Oxford OX2 6DP, as part of the OUP Journals
Comments and feedback: www-admin{at}oup.co.uk
Last modification: 18 May 1998
Copyright©Oxford University Press, 1998.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||