Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Okuno, K. Articles by Kinoshita, S. Search for Related Content PubMed PubMed Citation Articles by Okuno, K. Articles by Kinoshita, S. Agricola Articles by Okuno, K. Articles by Kinoshita, S.

 Previous Article  |  Next Article 

Applied and Environmental Microbiology, January 2004, p. 76-86, Vol. 70, No. 1
0099-2240/04/$08.00+0     DOI: 10.1128/AEM.70.1.76-86.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.

Utilization of Escherichia coli Outer-Membrane Endoprotease OmpT Variants as Processing Enzymes for Production of Peptides from Designer Fusion Proteins

Kazuaki Okuno,^1,2* Masayuki Yabuta,¹ Toshihiko Ooi,² and Shinichi Kinoshita²

Institute for Medicinal Research and Development, Daiichi Suntory Pharma Co., Ltd., Akaiwa, Chiyoda-machi, Ohra-gun, Gunma 370-0503,¹ Division of Molecular Chemistry, Graduate School of Engineering, Hokkaido University, Kita-ku, Sapporo, Hokkaido 060-8628, Japan²

Received 18 June 2003/ Accepted 13 October 2003

Top Abstract Introduction Materials and Methods Results Discussion References

 
Escherichia coli outer-membrane endoprotease OmpT has suitable properties for processing fusion proteins to produce peptides and proteins. However, utilization of this protease for such production has been restricted due to its generally low cleavage efficiency at Arg (or Lys)-Xaa, where Xaa is a nonbasic N-terminal amino acid of a target polypeptide. The objective of this study was to generate a specific and efficient OmpT protease and to utilize it as a processing enzyme for producing various peptides and proteins by converting its substrate specificity. Since OmpT Asp⁹⁷ is proposed to interact with the P1' amino acid of its substrates, OmpT variants with variations at Asp⁹⁷ were constructed by replacing this amino acid with 19 natural amino acids to alter the cleavage specificity at Arg (P1)-Xaa (P1'). The variant OmpT that had a methionine at this position, but not the wild-type OmpT, efficiently cleaved a fusion protein containing the amino acid sequence -Arg-Arg-Arg-Ala-Argmotilin, in which motilin is a model peptide with a phenylalanine at the N terminus. The OmpT variants with leucine and histidine at position 97 were useful in releasing human adrenocorticotropic hormone (1-24) (serine at the N terminus) and human calcitonin precursor (cysteine at the N terminus), respectively, from fusion proteins. Motilin was produced by this method and was purified up to 99.0% by two chromatographic steps; the yield was 160 mg/liter of culture. Our novel method in which the OmpT variants are used could be employed for production of various peptides and proteins.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Production of a large amount of polypeptides usually involves recombinant DNA technology. Short peptides are often expressed as fusion proteins in Escherichia coli, and somatostatin is produced as a fusion protein linked to E. coli ß-galactosidase (9). A fusion protein composed of a target peptide linked to a fusion partner (e.g., glutathione S-transferase, maltose-binding protein, or thioredoxin) via a linker peptide is expressed as a soluble protein or accumulates in the form of inclusion bodies. Generally, the fusion protein is processed to obtain the target peptide by chemical reagents, such as cyanogen bromide, or by external addition of a specific endoprotease (e.g., factor Xa [18], -thrombin [19], or Kex2 [27]). Although endoproteases can cleave fusion proteins more specifically and efficiently than chemical reagents, external addition of endoproteases increases the protein production cost. Also, use of a mammalian protease results in a risk of viral contamination. Therefore, we have tried to utilize E. coli OmpT (EC 3.4.21.87) as a processing enzyme (20, 21, 30).

OmpT, an endoprotease localized in the E. coli outer membrane (5), was originally proposed to be a serine protease, but it is now classified as an aspartyl protease (family A26). The mature enzyme consists of 297 amino acid residues and lacks cysteines (6, 25). This enzyme requires the outer-membrane lipid lipopolysaccharide for activity (14). OmpT belongs to the omptin family. Some omptin family members include the Yersinia pestis plasminogen activator (23), Salmonella enterica serovar Typhimurium E protein (7), E. coli OmpP (11), and Shigella flexneri SopA (3). A novel proteolytic mechanism that involves a His²¹²-Asp²¹⁰ dyad and an Asp⁸³-Asp⁸⁵ pair that activate a putative nucleophilic water molecule has been proposed based on the crystal structure of E. coli OmpT (28). The active site is fully conserved in the omptin family (28). The crystalline structure has shown that OmpT Asp⁹⁷ should be located close to the P1' residue of its substrate (13).

The physiological function of OmpT remains unclear. However, it has been suggested that this enzyme is involved in E. coli pathogenicity (16) and in inactivation of antimicrobial peptides (24). This enzyme is a stress protein, and its level of expression increases in response to induction of recombinant protein overexpression or heat shock (4). The protease activity in the cells is greater when they are cultured at 37°C than when they are cultured at 30°C, and expression of OmpT is higher in the late logarithmic and stationary phases than in the logarithmic phase (25).

The OmpT protease associates with inclusion bodies when a recombinant fusion protein accumulates in these bodies (10), and it is active in the presence of a high concentration of urea (29). As soon as the inclusion bodies are dissolved in a denaturation solution with urea, the inclusion body-associated OmpT can cleave the fusion protein, liberating the target peptide (21, 30). Whole cells can be also employed as the source of OmpT (5).

Although OmpT primarily cleaves the peptide bond between consecutive basic amino acid residues (25), the specificity for the P1' residue is not absolute (2). It has also been shown that this enzyme can cleave the peptide bond between Arg (or Lys) and Xaa residues of the PRX fusion protein, where Xaa is any amino acid except aspartic acid, glutamic acid, or proline (20). Like trypsin, OmpT requires an arginine or lysine residue at the P1 position for cleavage. Nevertheless, unlike trypsin, OmpT does not cleave its substrate at all dibasic and monobasic sites. On the contrary, its substrate specificity is different from and narrower than that of trypsin. It seems that OmpT interacts with the amino acid residues adjacent to a putative dibasic target and a monobasic target in the substrate, which exert a significant influence on cleavage specificity and enzyme efficiency. In a previous paper (21), it was demonstrated that basic amino acid residues at P4 and P6 enhance OmpT cleavage efficiency, while acidic amino acid residues at the same positions have the opposite effect. OmpT cleaves protamine, a highly basic antimicrobial peptide (24), and the OmpT extracellular domain, where the enzyme exerts its proteolytic activity, contains a large negatively charged groove (28). Hence, the electrostatic properties of amino acid residues proximal to the cleavage sites of substrates are likely to affect OmpT cleavage. Furthermore, it has been demonstrated that mature human atrial natriuretic peptide, which has a serine residue at the N terminus at P1', was released from the fusion protein with endogenous OmpT (21). However, the types of peptides that can be liberated from fusion proteins by OmpT are restricted due to low cleavage efficiency at Arg-Xaa, if Xaa is not a basic amino acid.

Thus, OmpT has suitable properties for processing the fusion protein since it has high substrate specificity and since it is active under the denaturing conditions in the inclusion bodies where the fusion protein is expressed. The purpose of this study was to change the substrate specificity at the P1' site and enhance the OmpT cleavage efficiency by mutating the enzyme and designing the amino acid sequences of substrate fusion proteins proximal to the cleavage site.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Bacterial strains.
E. coli strain JM109 (Takara Shuzo, Kyoto, Japan) was used for plasmid construction. E. coli strain W3110 M25, an OmpT protease-deficient mutant derived from strain W3110 (26), was used for fusion protein expression. This strain and E. coli strain BL21 (Novagen, Madison, Wis.) were employed for OmpT variant expression. The general procedures for DNA manipulation were carried out as described previously (20), and site-directed mutagenesis was performed by PCR with KOD Plus DNA polymerase (Toyobo, Osaka, Japan) used according to the manufacturer's instructions.

Expression plasmid constructs for fusion proteins.
Plasmid pG117S4HompPA, which was used to express a PA fusion protein (Fig. 1), was constructed by a PCR, and oligonucleotides were replaced by synthetic oligonucleotides from plasmid pG117S4HompPRR (20). The latter plasmid, a pBR322-derived high-copy-number plasmid used for production of glucagon-like peptide-1 (7-36 amide) [GLP-1 (7-36)] under control of the E. coli lac promoter, expressed a PRR fusion protein (Fig. 2). The PA fusion protein expressed, containing GLP-1 (7-37), has an OmpT cleavage site between Arg¹⁴⁰ (P1) and Arg¹⁴¹ (P1') in a 26-amino-acid linker peptide, P13-P13' (Fig. 1). Plasmid pG117S4HompPA3', expressing PA3', was prepared from pG117S4HompPA, from which plasmids expressing fusion proteins PA23' and PA2'3' (pG117S4HompPA23' and pG117S4HompPA2'3', respectively) were constructed. The plasmid expressing fusion protein PA323' (pG117S4HompPA323') was constructed by using plasmid pG117S4HompPA23'. In order to obtain the plasmid expressing the PRMT fusion protein (Fig. 3A), the 3' region of pG117S4HompPRR encoding the amino acid residues following Arg¹⁴¹ (Fig. 2) was replaced by the synthesized human motilin gene. The plasmid expressing fusion protein PMT (pG117S4HompPMT) (Fig. 4A) was prepared from the plasmid expressing fusion protein PRMT and pG117S4HompPA23'. The plasmids expressing fusion proteins PAC and PCT (Fig. 5A) were constructed from pG117S4HompPMT. All site-directed mutations were verified by DNA sequencing.

View larger version (13K): [in this window] [in a new window]

FIG. 1. Structures of fusion proteins containing GLP-1 (7-37) and their OmpT cleavage sites. The amino acid sequence around the cleavage site of fusion protein PA is shown. The arrowhead indicates an OmpT cleavage site between Arg140 and Arg141. The cleavage efficiencies at Arg140Arg141 (•) and Arg143Ala144 () of wild-type OmpT for the PA variants are indicated on the right. Arginine residues are indicated by boldface type. The arrows indicate the OmpT cleavage sites. The OmpT cleavage efficiency of the original fusion protein PA was defined as 100%. The cleavage efficiency at Arg140Arg141 for PA23' (39%) includes the cleavage efficiency at Arg139Arg140. The cleavage efficiencies at Arg140Arg141 and Arg143Ala144 for PA2'3' (60 and 63%, respectively) include parts of the cleavage efficiency at Arg142Arg143. The concentrations of fusion proteins PA, PA3', PA23', PA323', and PA2'3' were between 6 and 8 mg/ml.

View larger version (10K): [in this window] [in a new window]

FIG. 2. Structures of fusion proteins PRR and PRX. The arrowhead indicates the OmpT cleavage site of fusion protein PRR between Arg140 and Arg141. The PRR P1' residue was changed to 20 different amino acids, resulting in 20 types of PRX fusion proteins (20).

View larger version (15K): [in this window] [in a new window]

FIG. 3. (A) Structure of fusion protein PRMT. (B) HPLC elution profiles of reaction mixtures after cleavage of PRMT by insoluble cellular fractions of E. coli W3110 M25 cells expressing wild-type OmpT and OmpT variant D97M. The concentrations of wild-type OmpT and the D97M variant were 0.40 and 0.52 mg/ml, respectively, and the concentration of fusion protein PRMT was 4.4 mg/ml.

View larger version (21K): [in this window] [in a new window]

FIG. 4. (A) Structure of the PMT fusion protein containing human motilin. (B) HPLC elution profiles of reaction mixtures after cleavage of PMT by insoluble cellular fractions of E. coli cells expressing wild-type OmpT and OmpT variant D97M. The numbers in parentheses indicate the relative peak areas (expressed as percentages); the peak area of RRAR-motilin generated by wild-type OmpT was defined as 100%. The concentrations of wild-type OmpT and the D97M variant were 0.40 and 0.52 mg/ml, respectively, and the concentration of fusion protein PMT was 4.6 mg/ml.

View larger version (23K): [in this window] [in a new window]

FIG. 5. (A) Structures of fusion proteins PAC and PCT containing human ACTH (1-24) and CT (1-33), respectively. (B) HPLC elution profiles of reaction mixtures after cleavage of PAC and PCT by insoluble cellular fractions of wild-type and variant OmpT-expressing E. coli cells. The concentrations of PAC and PCT were 3.4 and 4.2 mg/ml, respectively, and the concentrations of wild-type OmpT, D97H, and D97L were 0.40, 0.76, and 0.48 mg/ml, respectively.

Preparation of inclusion bodies containing fusion proteins.
Inclusion bodies containing fusion proteins were prepared by using strain W3110 M25 as described previously (20). The relationships between protein concentrations and optical densities at 660 nm (OD₆₆₀) of the inclusion body suspensions were assessed as follows. An inclusion body suspension was appropriately diluted with 1 M acetic acid and 2 M urea. After centrifugation, the supernatant was subjected to a high-performance liquid chromatography (HPLC) analysis by using bovine serum albumin (Pierce, Rockford, Ill.) as a standard. HPLC was performed with a YMC Protein RP column (YMC, Kyoto, Japan). Elution was carried out with a linear gradient of 20 to 60% acetonitrile in a 0.1% trifluoroacetic acid solution at a flow rate of 1 ml/min for 40 min, and the effluent was monitored by measuring the absorbance at 220 nm.

Expression plasmids for OmpT variants.
The codon corresponding to OmpT Asp⁹⁷ was mutated by site-directed mutagenesis by using PCR to generate 19 different D97X variants (where X is a substituted amino acid residue at position 97). The variant pOmpTD97X plasmids were constructed by using pOmpTTcE (20), a pBR322-derived high-copy-number plasmid used for production of wild-type OmpT under control of the E.coli lac promoter. All site-directed mutations were verified by DNA sequencing.

OmpT protease activity assay.
For measurement of OmpT protease activity, the substrate dynorphin A (Peptide Institute, Osaka, Japan) was utilized as described previously (20). One unit of OmpT protease activity was defined as the activity necessary to cleave 1 µmol of dynorphin A per min at 25°C.

Fusion protein cleavage by OmpT.
OmpT was purified as described previously (20). A 10-µl portion of an inclusion body suspension (OD₆₆₀, 100) of each fusion protein containing GLP-1 (7-37) was solubilized in 20 µl of 10 M urea. Sodium phosphate (1 M, pH 7.0), EDTA (50 mM), and purified OmpT were added to final concentrations of 50 mM, 2 mM, and 0.14 U/ml, respectively. Cleavage reactions were performed in the presence of 4 M urea to dissolve the inclusion bodies; under these conditions the OmpT proteolytic activity is retained (29). A 150-µl aliquot of 1 M acetic acid-2 M urea was added to the reaction mixture (50 µl) after incubation for 30 min at 25°C. After centrifugation, the supernatant was analyzed by reversed-phase HPLC by using a YMC Protein RP column. Elution was performed with a linear gradient of 30 to 50% acetonitrile in a 0.1% trifluoroacetic acid solution for 20 min at a flow rate of 1 ml/min. The OmpT cleavage efficiency was estimated by the peak area of the cleavage product on HPLC and was compared to that of the fusion protein PA-derived product, which was defined as 100%. In order to identify the cleavage products containing GLP-1 (7-37) based on their molecular masses, the products were isolated by reversed-phase HPLC and analyzed with an SSQ710 mass spectrometer (Thermo Finnigan, San Jose, Calif.).

Cleavage of fusion proteins by OmpT variants.
OmpT protease-deficient E. coli BL21 cells harboring each OmpT variant pOmpTD97X expression plasmid were grown in 2 ml of Luria-Bertani medium containing 10 µg of tetracycline per ml at 37°C until the OD₆₆₀ reached 1. The cells were harvested by centrifugation and then washed with TE (10 mM Tris-HCl, 1 mM EDTA; pH 8.0) twice. The cells were resuspended in TE (OD₆₆₀, 2) and used as an OmpT protease D97X variant (where D97 is Asp⁹⁷ and X is another amino acid at position 97) for cleavage of PRX fusion proteins (Table 1 and Fig. 2). Western blot analysis showed that the levels of expression of all OmpT variants were similar (data not shown). A 5-µl portion of an inclusion body suspension (OD₆₆₀, 100) of each PRX fusion protein was solubilized in 20 µl of 10 M urea. Sodium phosphate and EDTA were added as described above. The reaction was started by addition of a 10-µl portion of a whole-cell suspension expressing the OmpT protease D97X variant. The reaction mixtures (50 µl) were incubated for 60 min at 25°C and analyzed by reversed-phase HPLC by using a YMC Protein RP column as described above.

View this table: [in this window] [in a new window]

TABLE 1. Cleavage efficiencies of OmpT D97X variants with PRX fusion proteinsa

Quantification of human motilin by HPLC.
The mature human motilin concentrations in the cleavage reaction mixtures were assessed by HPLC by using a YMC Protein RP column. Each reaction mixture was appropriately diluted with 1 M acetic acid and 2 M urea. After centrifugation, the supernatant was injected onto the column. Elution was carried out with a linear gradient of 20 to 27.5% acetonitrile in a 0.1% trifluoroacetic acid solution at a flow rate of 1 ml/min for 15 min, and the effluent was monitored by determining the absorbance at 214 nm. Human motilin concentrations were determined from the ratio of the peak area to the area of a human motilin standard (Peptide Institute). For analysis of purity, elution was carried out with a linear gradient of 0 to 50% acetonitrile in a 0.1% trifluoroacetic acid solution at a flow rate of 1 ml/min for 50 min.

Preparation of insoluble cellular fractions from W3110 M25 cells expressing an OmpT variant.
W3110 M25 cells harboring the expression plasmid for wild-type OmpT or for OmpT variant D97L, D97M, or D97H were grown at 37°C in 400 ml of Luria-Bertani medium containing 10 µg of tetracycline per ml. The cells were harvested by centrifugation. After resuspension in TE, the cells were disrupted by ultrasonication. The insoluble fractions containing membrane fractions were washed with TE twice, suspended in TE (OD₆₆₀, 10), and stored at -20°C until they were needed. The insoluble cellular fractions were analyzed by sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) to determine the concentrations of OmpT. The band intensities for wild-type OmpT and OmpT variants were measured by scanning with a computing densitometer (model 300A; Molecular Dynamics, Sunnyvale, Calif.) after Coomassie brilliant blue R250 staining; purified OmpT was used as the standard.

Cleavage of fusion proteins by insoluble cellular fractions of W3110 M25 cells expressing OmpT variants.
A 10-µl portion of an inclusion body suspension (OD₆₆₀, 100) containing the fusion protein PRMT, PMT, PAC, or PCT was solubilized in 20 µl of 10 M urea. Sodium phosphate and EDTA were added as described above. The reaction was started by adding 5 µl of the insoluble cellular fraction of E. coli W3110 M25 cells expressing wild-type OmpT or an OmpT variant. A portion of each reaction mixture (50 µl) was analyzed by HPLC and SDS-PAGE after incubation for 10 min (PAC) or 120 min (PRMT, PMT, and PCT) at 25°C. HPLC analysis of human adrenocorticotropic hormone (1-24) or human calcitonin precursor was carried out by using a YMC Protein RP column with a linear gradient of 10 to 50% acetonitrile for 50 min.

Conditions for high-cell-density cultivation.
E. coli W3110 M25 cells harboring the OmpT variant D97M expression plasmid pOmpTD97M were grown to obtain a high-cell-density culture in a 3-liter fermentor at 32°C with 2 liters of medium containing (per liter) 4 g of K₂HPO₄, 4 g of KH₂PO₄, 2.7 g of Na₂HPO₄, 0.2 g of NH₄Cl, 1.2 g of (NH₄)₂SO₄, 4 g of yeast extract, 2 g of MgSO₄ · 7H₂O, 40 mg of CaCl₂ · 2H₂O, 40 mg of FeSO₄ · 7H₂O, 10 mg of MnSO₄ · 5H₂O, 10 mg of AlCl₃ · 6H₂O, 4 mg of CoCl₂ · 6H₂O, 2 mg of ZnSO₄ · 7H₂O, 2 mg of Na₂MoO₄ · 2H₂O, 1 mg of CuCl₂ · 2H₂O, 0.5 mg of H₃BO₄, and 10 mg of tetracycline, to which glucose was added at a final concentration of 1.5% as a carbon source. After glucose starvation, glycerol was added to a final concentration of 2.0%, and the temperature was shifted to 37°C to enhance expression of the OmpT variant. Cells were harvested by centrifugation and then washed with TE twice. The cells obtained were resuspended in TE (OD₆₆₀, 320) and then stored at -20°C until they were needed. E. coli W3110 M25 cells harboring the fusion protein PMT expression plasmid pG117S4HompPMT were grown similarly. The cells were disrupted with a Manton-Gaullin homogenizer. The expressed fusion proteins were present as insoluble inclusion bodies. This insoluble fraction was harvested by centrifugation and then washed with deionized water. The fraction was washed with 50 mM Tris-HCl (pH 7.5) containing 5 mM EDTA and 1% Triton X-100 and then washed with deionized water. The pellet was resuspended in deionized water (OD₆₆₀, 250) and stored at -20°C until it was used for human motilin production.

Production and purification of human motilin.
A 4-ml portion of the inclusion body suspension of the PMT fusion protein produced by W3110 M25(pG117S4HompPMT) cells was solubilized in 8 ml of 10 M urea. Sodium phosphate and EDTA were added to final concentrations of 50 and 2 mM, respectively. The reaction was started by addition of 1 ml of a W3110 M25(pOmpTD97M) cell suspension. The reaction mixture (20 ml) was incubated at 25°C. A 13.5-ml portion of the reaction mixture was added to 40.5 ml of 20 mM sodium acetate (pH 4.0) for acid precipitation. After centrifugation, the supernatant was diluted with 20 mM sodium acetate (pH 4.0) to obtain a final volume of 200 ml. The diluted supernatant was loaded onto an SP Sepharose fast-flow column (26 by 50 mm; Amersham Pharmacia Biotech, Uppsala, Sweden) that had been equilibrated with 20 mM sodium acetate (pH 4.0), washed with 100 ml of the same buffer, and then washed with 100 ml of the buffer containing 0.1 M NaCl. Human motilin was eluted with a linear gradient of 0.1 to 0.5 M NaCl in the acetate buffer at a rate of 5 ml/min. The eluate was loaded onto a Vydac 214TPB1520 column (10 by 300 mm; Grace Vydac, Hesperia, Calif.) that had been equilibrated with 0.1% trifluoroacetic acid, washed with 100 ml of the same solution, and then eluted with a linear gradient of 0 to 30% acetonitrile in 0.1% trifluoroacetic acid at a rate of 1.6 ml/min.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Substrate specificity of the OmpT variants.
The OmpT crystal structure suggested that the P1' residues of its substrates would be located close to Asp⁹⁷ (13). In order to study the substrate specificity of the D97X OmpT variants (where X is a substituted amino acid residue at position 97) at the P1' site, the PRX fusion proteins (Fig. 2), which were prepared by replacement of Arg¹⁴¹ at the P1' position in the PRR fusion protein (20), were treated with the variant proteases (Table 1). Some variants had different P1' specificities than the wild-type OmpT. Among the variants, D97A, D97L, D97F, D97M, D97S, D97T, D97C, D97N, D97Q, and D97H preferred fusion proteins with monobasic cleavage sites instead of the PRK and PRR fusion proteins, which have dibasic cleavage sites. The OmpT variant D97M preferred phenylalanine at the P1' position the most among these variants.

Release of human motilin by OmpT variant D97M, but not by wild-type OmpT.
Based on the results described above, we expected that the OmpT variant D97M might liberate human motilin, which has a phenylalanine residue at its N terminus, from the PRMT fusion protein, which consisted of 140 N-terminal amino acids of fusion protein PRR and human motilin (Fig. 3A). The cleavage reaction was performed by using the insoluble cellular fraction of W3110 M25 cells harboring the wild-type OmpT or OmpT D97M expression plasmid, and the products were analyzed by HPLC. The OmpT variant D97M could release human motilin from the fusion protein PRMT (Fig. 3B), and the motilin concentration was 0.10 mg/ml after the cleavage reaction. The wild-type OmpT did not cleave the fusion protein at all (Fig. 3B). Thus, it was found that replacing OmpT Asp⁹⁷ with methionine was necessary in order to cleave PRMT between Arg¹⁴⁰ and Phe¹⁴¹.

Designing the amino acid sequence of the fusion protein proximal to the monobasic cleavage site.
Although the OmpT variant D97M cleaved the PRMT fusion protein at the monobasic site between the arginine and phenylalanine residues, the cleavage was partial (data not shown), and the cleavage efficiency was low. In order to increase the cleavage efficiency of OmpT at monobasic sites, the effect of a basic amino acid residue(s) proximal to the cleavage site was investigated by using GLP-1 (7-37) fusion proteins. It was previously reported that basic amino acid residues at the P4 and P6 sites enhanced OmpT cleavage efficiency at a dibasic site (21) and used the wild-type OmpT protease and the monobasic cleavage site between arginine and alanine.

By investigating the effect of the arginine residue proximal to the cleavage site on OmpT using the PA substrate fusion protein, in which amino acid residues P10 to P2 and P2' to P5' of the PRR fusion protein PRR are replaced by nine and four consecutive alanine residues, respectively, we found that the PA3' fusion protein was partially cleaved between Arg¹⁴³ and Ala¹⁴⁴ (Fig. 1). Next, an arginine residue(s) was placed around the monobasic cleavage site in order to find a substrate in which OmpT prefers the monobasic cleavage site to the dibasic site (PA23', PA323', and PA2'3') (Fig. 1). PA23', which contained the sequence -Arg¹³⁹-Arg¹⁴⁰-Arg¹⁴¹-Ala¹⁴²-Arg¹⁴³ -Ala¹⁴⁴-, was found to be the best substrate that OmpT efficiently and specifically cleaved between Arg¹⁴³ and Ala¹⁴⁴.

Utilization of fusion proteins PMT and OmpT variant D97M for release of human motilin.
We examined whether the OmpT variant D97M can liberate human motilin more efficiently from the fusion protein PMT (Fig. 4A) than from the fusion protein PRMT (Fig. 3A). Insoluble cellular fractions of W3110 M25 cell lines harboring the OmpT and OmpT D97M expression plasmids were utilized. HPLC elution profiles are shown in Fig. 3B and 4B. Both enzymes cleaved PMT almost completely (data not shown). Human motilin was found to be released from the PMT fusion protein when it was treated with the OmpT variant D97M but not when it was treated with the wild-type OmpT (Fig. 4B). From PMT, D97M released motilin (0.36 mg/ml) and RRAR-motilin at a ratio of approximately 3:1 (Fig. 4B). The wild-type OmpT cleaved PMT at Arg¹³⁹-Arg¹⁴⁰ and Arg¹⁴⁰-Arg¹⁴¹ to liberate RRAR-motilin and RAR-motilin, respectively.

When the substrates PMT and PRMT were compared, the amount of motilin released by the OmpT variant D97M from PMT was 3.6-fold higher than the amount released from PRMT (0.36 mg/ml versus 0.10 mg/ml). These results indicated that motilin release required utilization of OmpT variant D97M and that the cleavage efficiency was enhanced by the primary structure -Arg-Arg-Arg-Ala-Arg (P1)- in front of the monobasic cleavage site.

Utilization of OmpT D97X variants for release of human peptides from fusion proteins.
Like the use of D97M for release of human motilin from PMT, we investigated whether other OmpT variants could cleave fusion proteins with the amino acid sequence -Arg-Arg-Arg-Ala-Arg (P1)- and release target peptides. The fusion proteins PAC and PCT, containing human adrenocorticotropic hormone (1-24) [ACTH (1-24)] with an N-terminal serine and human calcitonin precursor (1-33) [CT (1-33)] with a cysteine residue at the N terminus, were prepared (Fig. 5A). The OmpT variants D97L and D97H were chosen for cleavage of PAC and PCT, respectively (Table 1).

The fusion protein PAC was incubated with the insoluble cellular fraction of W3110 M25 expressing wild-type OmpT or OmpT variant D97L for 10 min and was cleaved almost completely (Fig. 5B). The OmpT variant D97L successfully released ACTH (1-24) without any by-products. Wild-type OmpT also released ACTH (1-24), but the by-product RAR-ACTH was generated from cleavage between Arg¹⁴⁰ and Arg¹⁴¹. Additionally, ACTH (1-15) and ACTH (16-24) were generated by cleavage at Arg¹⁴³-Ser¹⁴⁴ and Lys¹⁵⁸-Lys¹⁵⁹ (data not shown). When the variant OmpT and wild-type OmpT were compared, the amount of ACTH (1-24) released by OmpT variant D97L was 2.9-fold higher than the amount released by the wild-type OmpT.

The fusion protein PCT was similarly incubated with wild-type OmpT or OmpT variant D97H for 120 min and was cleaved almost completely (Fig. 5B). The OmpT variant D97H released CT (1-33), and the by-products RRAR-CT and AR-CT were generated from cleavage at Arg¹³⁹-Arg¹⁴⁰ and Arg¹⁴¹-Ala¹⁴², respectively. Mass spectroscopy analysis suggested that the CT (1-33) peak overlapped the AR-CT peak. Wild-type OmpT could not cleave PCT at the monobasic site between Arg¹⁴³ and Cys¹⁴⁴ to release CT (1-33) but instead cleaved it at Arg¹³⁹-Arg¹⁴⁰ and Arg¹⁴⁰-Arg¹⁴¹ to liberate RRAR-CT and RAR-CT, respectively.

Production of human motilin.
Our findings were applied to production of human motilin on a larger scale. High-cell-density cultures of recombinant E. coli W3110 M25 cell lines harboring the expression plasmids for the fusion protein PMT [W3110 M25(pG117S4HompPMT)] and the OmpT variant D97M [W3110 M25(pOmpTD97M)] were grown, and this was followed by a cleavage reaction and purification of human motilin. High expression of PMT and D97M did not affect cell growth, and the OD₆₆₀ of both cell cultures kept increasing, even at the end of the culture period. The OD₆₆₀ of the W3110 M25(pG117S4HompPMT) cell culture reached 67 in 24 h, and 15 g (wet weight) of inclusion bodies per liter was obtained. The OD₆₆₀ of the W3110 M25(pOmpTD97M) cell culture was 65 after 20 h, and 150 g (wet weight) of cells per liter of culture was harvested.

By using these inclusion bodies and cells, the releasing reaction of human motilin was performed (Fig. 6A). After 60 min of incubation, the concentration of human motilin released from PMT was about 2 mg/ml, and then the concentration reached a plateau (Fig. 6A). Almost all fusion protein had been cleaved by that time (Fig. 6B, lane 2). The fact that the motilin concentration did not increase after 60 min (Fig. 6A) suggested that the native peptide was not made from the by-product RRAR-motilin. Although the amount of RRAR-motilin seemed to be more than the amount of motilin, as determined by SDS-PAGE (Fig. 6B), HPLC quantitatively showed that the PMT fusion protein generated more motilin than the by-product (Fig. 6C). Replacement of Ala¹³⁷ or Ala¹³⁸ of PMT by aspartic acid was effective in preventing generation of RRAR-motilin (data not shown).

View larger version (20K): [in this window] [in a new window]

FIG. 6. Production of human motilin from fusion protein PMT by E. coli cells expressing OmpT variant D97M. (A) Time course of the release of human motilin from PMT. (B) SDS-16% PAGE analysis. Lane 1, PMT; lane 2, PMT incubated for 60 min with OmpT variant D97M-expressing E. coli W3110 M25 whole cells; lane 3, authentic human motilin. (C) HPLC elution profile of PMT after 60 min of incubation with OmpT variant D97M-expressing E. coli.

Human motilin was purified by employing the releasing reaction mixture incubated for 60 min. Table 2 summarizes the purification steps and shows the amount, yield, and purity of motilin. Most of the undigested fusion protein, the partner peptide, and a large part of the impurities derived from host cells could be effectively removed by acid precipitation. Almost all RRAR-motilin was efficiently separated by cation-exchange chromatography with a linear NaCl gradient because the theoretical isoelectric points of motilin and RRAR-motilin are 8.50 and 10.88, respectively. Finally, human motilin was purified by reversed-phase chromatography, which produced the pure peptide in an HPLC elution profile (data not shown). The molecular mass and N-terminal amino acid sequence of the purified product corresponded with those of standard human motilin (data not shown). The motilin yield and purity were satisfactory.

View this table: [in this window] [in a new window]

TABLE 2. Summary of human motilin purificationa

Top Abstract Introduction Materials and Methods Results Discussion References

 
Several systems in which OmpT is used as the processing enzyme to release the target peptide from a fusion protein have been constructed (8, 12, 17, 20, 21, 30). However, these systems can only be employed to liberate a limited number of target polypeptides that have basic amino acids at the C terminus or a few types of amino acid sequences at the N terminus preferred by OmpT. In this study, we developed a novel system which has the potential to effectively produce a variety of short target peptides from fusion proteins by using OmpT variants with converted P1' specificities and fusion proteins with a particular primary structure proximal to the monobasic cleavage sites. We employed two strategies: (i) replacing OmpT Asp⁹⁷ with various amino acids to obtain OmpT variants that efficiently release peptides with various N-terminal amino acid residues from fusion proteins and (ii) optimizing the fusion protein amino acid sequence in front of the monobasic cleavage site.

First, OmpT variants with changed P1' specificity were obtained. OmpT D97X variants (e.g., D97H, D97M, and D97L) that hydrolyze peptide bonds preferentially at monobasic sites rather than dibasic sites were constructed (Table 1). This result supports the hypothesis that the OmpT substrate P1' residue is located close to Asp⁹⁷ (13). The P1' specificity of D97E approximately coincided with that of wild-type OmpT, suggesting that the interaction between the P1' residue and OmpT Asp⁹⁷ is primarily electrostatic. However, we observed no other remarkable relationship between the characteristics of the substituted amino acid residue of an OmpT variant and its P1' specificity.

There are some instances of designer enzymes that have been described previously. For example, Ballinger et al. changed subtilisin Asn⁶² and Gly¹⁶⁶ to aspartic acid, which resulted in a shift in specificity toward dibasic substrates (1). They also identified a substrate (Asn-Leu-Met-Arg-Lys) that was selectively cleaved by the mutated protease. Kurth et al. replaced Lys⁶⁰ with glutamic acid and aspartic acid to introduce high specificity for basic residues into the trypsin S1' site (15). To our knowledge, our study is the first application of molecular modeling in which OmpT substrate specificity was changed.

Olsen et al. (22) took another approach to obtain OmpT proteases with novel substrate specificities. In their study, enzymatic cleavage of substrates resulted in release of a fluorescent resonance energy transfer quenching partner, while the fluorescent product was retained on the cell surface. By screening a library of random OmpT variants by fluorescence-activated cell sorting using a peptide substrate with an Arg-Val cleavage sequence, they isolated a variant protease in which the catalytic activities were enhanced 60-fold. The variant OmpT had amino acid substitutions at some positions, but interestingly Asp⁹⁷ was conserved. On the other hand, we showed that not only the wild-type OmpT but also the OmpT variant D97M cleaved at the dibasic site between Arg¹³⁹ and Arg¹⁴⁰ of the PMT fusion protein (Fig. 4). Hence, OmpT amino acid residues other than that at position 97 may also be important for P1' specificity.

Our second strategy was to optimize the amino acid sequences of fusion proteins near the OmpT monobasic cleavage site. A good substrate was found with the amino acid sequence -Arg-Arg-Arg-Ala-Arg (P1)-Ala (P1')-, which the wild-type OmpT cleaved effectively and preferentially. Replacement of OmpT Asp⁹⁷ did not change the substrate specificities at any of the other sites except P1', as the particular amino acid sequence of the wild-type OmpT was applicable to at least some OmpT variants. Since OmpT favors small hydrophobic amino acids at P2 (2), a similar increase in cleavage efficiency might be obtained by using -Arg-Arg-Arg-Phe-Arg (P1)-Ala (P1')- and -Arg-Arg-Arg-Ile-Arg (P1)-Ala (P1')-. Fusion protein PRA was cleaved similarly to fusion protein PRF by OmpT variant D97M, whose cleavage efficiencies were 6.0% with PRA and 7.7% with PRF (Table 1). However, D97M mainly cleaved the fusion protein PMT (Fig. 4) at Arg¹⁴³-Phe¹⁴⁴ but not at Arg¹⁴¹-Ala¹⁴². Fusion proteins PMT and PRA/PRF have different amino acid sequences near the P1-P1' site. Therefore, the results suggest that amino acid residues proximal to the cleavage site of the substrate are also important for cleavage by the OmpT enzyme.

Our system should efficiently produce more peptides than any previously described system in which OmpT is used, as exemplified by human motilin, ACTH (1-24), and CT (1-33). Previously, we demonstrated that mature human atrial natriuretic peptide, which has a serine residue at the N terminus at P1', is released from the fusion protein by endogenous (wild-type) OmpT (21). The wild-type OmpT might have some preference for a serine residue at the P1' site, as it partially cleaved PAC between Arg¹⁴³ and Ser¹⁴⁴ (Fig. 5B). However, the wild-type enzyme was not specific for the monobasic cleavage site on PAC and generated not only ACTH (1-24) but also the by-products RAR-ACTH, ACTH (1-15), and ACTH (16-24). On the other hand, the OmpT variant D97L released ACTH (1-24) specifically and efficiently, which coincides well with its substrate specificity for the fusion protein PRS instead of PRR (Table 1). The substrate specificity at P1' of the OmpT variant D97H is broad (Table 1). Therefore, the variant should be able to release peptides having many different amino acids at the N terminus, but it is liable to generate by-products. As an example of reducing the generation of by-products, the by-product AR-CT shown in Fig. 5B might be reduced without decreasing the release of CT (1-33) by substituting the linker amino acid sequence -Arg-Arg-Arg-Ile-Arg- for -Arg-Arg-Arg-Ala-Arg-, since D97H preferred PRA to PRI (Table 1).

In some experiments, we used whole E. coli cells that express the OmpT variant as the processing enzyme without adding the protease externally, which eliminated the cost of OmpT purification. In order to improve this production system, we are now moving towards coexpression of the fusion protein and the OmpT variant with compatible expression plasmids.

 
We are grateful to Takuroh Yoshida for his help with the N-terminal amino acid sequence analysis and to Fuyuki Iwasa for critically reading the manuscript and for helpful discussions.

 
* Corresponding author. Mailing address: Institute for Medicinal Research and Development, Daiichi Suntory Pharma Co., Ltd., 2716-1 Kurakake, Akaiwa, Chiyoda-machi, Ohra-gun, Gunma 370-0503, Japan. Phone: 81-276-86-5784. Fax: 81-276-86-5760. E-mail: Kazuaki_Okuno{at}dsup.co.jp.

Top Abstract Introduction Materials and Methods Results Discussion References

 

1
1. Ballinger, M. D., J. Tom, and J. A. Wells. 1995. Designing subtilisin BPN' to cleave substrates containing dibasic residues. Biochemistry 34:13312-13319.[CrossRef][Medline]
2
2. Dekker, N., R. C. Cox, R. A. Kramer, and M. R. Egmond. 2001. Substrate specificity of the integral membrane protease OmpT determined by spatially addressed peptide libraries. Biochemistry 40:1694-1701.[CrossRef][Medline]
3
3. Egile, C., H. d'Hauteville, C. Parsot, and P. J. Sansonetti. 1997. SopA, the outer membrane protease responsible for polar localization of IcsA in Shigella flexneri. Mol. Microbiol. 23:1063-1073.[CrossRef][Medline]
4
4. Gill, R. T., M. P. DeLisa, M. Shiloach, T. R. Holoman, and W. E. Bentley. 2000. OmpT expression and activity increase in response to recombinant chloramphenicol acetyltransferase overexpression and heat shock in E. coli. J. Mol. Microbiol. Biotechnol. 2:283-289.[CrossRef][Medline]
5
5. Grodberg, J., and J. J. Dunn. 1988. ompT encodes the Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J. Bacterial. 170:1245-1253.[Abstract/Free Full Text]
6
6. Grodberg, J., M. D. Lundrigan, D. L. Toledo, W. F. Mangel, and J. J. Dunn. 1988. Complete nucleotide sequence and deduced amino acid sequence of the ompT gene of Escherichia coli K-12. Nucleic Acids Res. 16:1209.[Free Full Text]
7
7. Grodberg, J., and J. J. Dunn. 1989. Comparison of Escherichia coli K-12 outer membrane protease OmpT and Salmonella typhimurium E protein. J. Bacteriol. 171:2903-2905.[Abstract/Free Full Text]
8
8. Hanke, C., J. Hess, G. Schumacher, and W. Goebel. 1992. Processing by OmpT of fusion proteins carrying the HlyA transport signal during secretion by Escherichia coli hemolysin transport system. Mol. Gen. Genet. 233:42-48.[CrossRef][Medline]
9
9. Itakura, K., T. Hirose, R. Crea, A. D. Riggs, H. L. Heyneker, F. Bolivar, and H. W. Boyer. 1977. Expression in Escherichia coli of a chemically synthesized gene for the hormone somatostatin. Science 198:1056-1063.[Abstract/Free Full Text]
10
10. Jürgen, B., H. Y. Lin, S. Riemschneider, C. Scharf, P. Neubauer, R. Schmid, M. Hecker, and T. Schweder. 2000. Monitoring of genes that respond to overproduction of an insoluble recombinant protein in Escherichia coli glucose-limited fed-batch fermentations. Biotechnol. Bioeng. 70:217-224.[CrossRef][Medline]
11
11. Kaufmann, A., Y.-D. Stierhof, and U. Henning. 1994. New outer membrane-associated protease of Escherichia coli K-12. J. Bacteriol. 176:359-367.[Abstract/Free Full Text]
12
12. Klauser, T., J. Pohlner, and T. F. Meyer. 1992. Selective extracellular release of cholera toxin B subunit by Escherichia coli: dissection of Neisseria Iga_ß-mediated outer membrane transport. EMBO J. 11:2327-2335.[Medline]
13
13. Kramer, R. A., L. Vandeputte-Rutten, G. J. de Roon, P. Gros, N. Dekker, and M. R. Egmond. 2001. Identification of essential acidic residues of outer membrane protease OmpT supports a novel active site. FEBS Lett. 505:426-430.[CrossRef][Medline]
14
14. Kramer, R. A., K. Brandenburg, L. Vandeputte-Rutten, M. Werkhoven, P. Gros, N. Dekker, and M. R. Egmond. 2002. Lipopolysaccharide regions involved in the activation of Escherichia coli outer membrane protease OmpT. Eur. J. Biochem. 269:1746-1752.[Medline]
15
15. Kurth, T., S. Grahn, M. Tormann, D. Ullmann, H.-J. Hofmann, H.-D. Jakubke, and L. Hedstrom. 1998. Engineering the S1' subsite of trypsin: design of a protease which cleaves between dibasic residues. Biochemistry 37:11434-11440.[CrossRef][Medline]
16
16. Lundrigan, M. D., and R. M. Webb. 1992. Prevalence of ompT among Escherichia coli isolates of human origin. FEMS Microbiol. Lett. 97:51-56.[CrossRef]
17
17. Maurer, J., J. Jose, and T. F. Meyer. 1997. Autodisplay: one-component system for efficient surface display and release of soluble recombinant proteins from Escherichia coli. J. Bacteriol. 179:794-804.[Abstract/Free Full Text]
18
18. Nagai, K., and H. C. Thogersen. 1984. Generation of beta-globin by sequence-specific proteolysis of a hybrid protein produced in Escherichia coli. Nature 309:810-812.[CrossRef][Medline]
19
19. Nishikawa, S., K. Yanase, T. Tokunaga-Doi, K. Kodama, H. Gomi, S. Uesugi, E. Ohtsuka, Y. Kato, F. Suzuki, and M. Ikehara. 1987. Efficient cleavage by alpha-thrombin of a recombinant fused protein which contains insulin-like growth factor I. Protein Eng. 1:487-492.[Abstract/Free Full Text]
20
20. Okuno, K., M. Yabuta, K. Kawanishi, K. Ohsuye, T. Ooi, and S. Kinoshita. 2002. Substrate specificity at the P1' site of Escherichia coli OmpT under denaturing conditions. Biosci. Biotechnol. Biochem. 66:127-134.[CrossRef][Medline]
21
21. Okuno, K., M. Yabuta, K. Ohsuye, T. Ooi, and S. Kinoshita. 2002. An analysis of target preferences of Escherichia coli outer-membrane endoprotease OmpT for use in therapeutic peptide production: efficient cleavage of substrates with basic amino acids at the P4 and P6 positions. Biotechnol. Appl. Biochem. 36:77-84.[CrossRef][Medline]
22
22. Olsen, M. J., D. Stephens, D. Griffiths, P. Daugherty, G. Georgiou, and B. L. Iverson. 2000. Function-based isolation of novel enzymes from a large library. Nat. Biotechnol. 18:1071-1074.[CrossRef][Medline]
23
23. Sodeinde, O. A., and J. D. Goguen. 1989. Nucleotide sequence of the plasminogen activator gene of Yersinia pestis: relationship to ompT of Escherichia coli and gene E of Salmonella typhimurium. Infect. Immun. 57:1517-1523.[Abstract/Free Full Text]
24
24. Stumpe, S., R. Schmid, D. L. Stephens, G. Georgiou, and E. P. Bakker. 1998. Identification of OmpT as the protease that hydrolyzes the antimicrobial peptide protamine before it enters growing cells of Escherichia coli. J. Bacteriol. 180:4002-4006.[Abstract/Free Full Text]
25
25. Sugimura, K., and T. Nishihara. 1988. Purification, characterization, and primary structure of Escherichia coli protease VII with specificity for paired basic residues: identity of protease VII and OmpT. J. Bacteriol. 170:5625-5632.[Abstract/Free Full Text]
26
26. Sugimura, K. 1988. Mutant isolation and cloning of the gene encoding protease VII from Escherichia coli. Biochem. Biophys. Res. Commun. 153:753-759.[CrossRef][Medline]
27
27. Suzuki, Y., M. Yabuta, and K. Ohsuye. 1998. High-level production of recombinant human parathyroid hormone 1-34. Appl. Environ. Microbiol. 64:526-529.[Abstract/Free Full Text]
28
28. Vandeputte-Rutten, L., R. A. Kramer, J. Kroon, N. Dekker, M. R. Egmond, and P. Gros. 2001. Crystal structure of the outer membrane protease OmpT from Escherichia coli suggests a novel catalytic site. EMBO J. 20:5033-5039.[CrossRef][Medline]
29
29. White, C. B., Q. Chen, G. L. Kenyon, and P. C. Babbitt. 1995. A novel activity of OmpT. Proteolysis under extreme denaturing conditions. J. Biol. Chem. 270:12990-12994.[Abstract/Free Full Text]
30
30. Yabuta, M., N. Ochi, and K. Ohsuye. 1995. Hyperproduction of a recombinant fusion protein of Staphylococcus aureus V8 protease in Escherichia coli and its processing by OmpT protease to release an active V8 protease derivative. Appl. Microbiol. Biotechnol. 44:118-125.[Medline]

---

Applied and Environmental Microbiology, January 2004, p. 76-86, Vol. 70, No. 1
0099-2240/04/$08.00+0     DOI: 10.1128/AEM.70.1.76-86.2004
Copyright © 2004, American Society for Microbiology. All Rights Reserved.

This article has been cited by other articles:

• Varadarajan, N., Gam, J., Olsen, M. J., Georgiou, G., Iverson, B. L. (2005). Engineering of protease variants exhibiting high catalytic activity and exquisite substrate selectivity. Proc. Natl. Acad. Sci. USA 102: 6855-6860 [Abstract] [Full Text]  

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Okuno, K. Articles by Kinoshita, S. Search for Related Content PubMed PubMed Citation Articles by Okuno, K. Articles by Kinoshita, S. Agricola Articles by Okuno, K. Articles by Kinoshita, S.