The Protein Encoded at the 3′ End of the Serine Protease Gene of Aeromonas sobria Functions as a Chaperone in the Production of the Protease

Abstract

For the successful production of Aeromonas sobria serine protease (ASP), open reading frame 2 (ORF2) protein, encoded at the 3′ end of the protease operon, is required. In this study, we examined the action of ORF2 protein. The results showed that the protein associated with ASP in the periplasm and helped ASP to form an active structure.

Aeromonas spp. are involved in a number of human illnesses, including not only intestinal but also nonintestinal diseases (12). In the intestine, the pathogens induce diarrhea with pain, but the mortality due to infection is low (2). In nonintestinal diseases, symptoms such as septicemia, skin and soft-tissue lesions with bulla and necrotizing fasciitis, and meningitis have been reported (9, 10). As the Aeromonas serine protease (ASP) has activity to induce edema in dorsal skin, the protease is thought to be related to the soft-tissue lesions induced by Aeromonas spp. (17).

Two proteins, open reading frame 1 (ORF1) protein and ORF2 protein, are synthesized by translation of the gene that codes for ASP. They are encoded at the 5′ and 3′ ends of the gene, respectively (16). ORF1 protein is ASP itself. ORF2 protein is composed of 152 amino acid residues. Sequences homologous with the operon have not been found in other organisms. Our previous studies showed that ORF2 protein was indispensable for the successful production of ASP by the cell. Subsequent analysis revealed that the transcription of the ASP gene was not affected by ORF2 protein and that ORF2 protein synthesized in the cytoplasm moved into the periplasmic space (16). Therefore, we speculated that ORF2 protein functions in the periplasmic space to help ASP form a proper structure. In this paper, we attempted to confirm this hypothesis.

Association of ASP with ORF2 protein in the periplasm.

We previously demonstrated that the 24-amino-acid peptide at the amino terminus of ASP was a signal peptide that functions in translocation across the inner membrane (16). We also showed that ORF2 protein synthesized in the cytoplasm moved into the periplasmic space and that the protein in the periplasm was smaller than that in the cytoplasm (16). To determine the exact size of ORF2 protein in the periplasm, periplasmic proteins of Escherichia coli HB101 harboring pET11-ORF2, which contains the ORF2 gene, were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and the bands on the gel were transferred to a nitrocellulose filter. The band for ORF2 protein on the filter was extracted, and its amino-terminal sequence was determined by the Edman degradation method. The comparison of the sequence determined with that deduced from the DNA sequence revealed that ORF2 protein is synthesized as a precursor and that a 22-amino-acid peptide at the amino terminus of the precursor is removed in the ORF2 protein on the gel. The region might be a signal peptide functioning in translocation across the inner membrane. The theoretical molecular weight of the ORF2 protein in the periplasmic space is 13,735.

Subsequently, a cross-linking experiment was performed to examine the association of ORF2 protein with ASP. Plasmid pUC119-5528, which contains the ASP operon (16), was introduced into E. coli HB101. A periplasmic fraction of cells which were grown to early logarithmic phase was prepared by treatment with polymyxin B (16). Then, the periplasmic fraction was treated with 5 mM dimethyl suberimidate (DMS), a cross-linker (15). The samples treated were separated by SDS-PAGE. The bands composing ASP and ORF2 protein were detected by immunoblotting using, respectively, anti-ASP antiserum and anti-ORF2 protein antiserum, which were prepared by the injection into rabbits of purified ASP and the 12-amino-acid peptide (CIELSGAEQQPK) whose sequence is identical with that of the carboxy-terminal end of ORF2 protein, respectively.

The result is shown in Fig. 1. A protein reacting with anti-ASP antiserum with a size of 64 kDa was found in the samples from cells producing ASP and ORF2 protein (Fig. 1A, lanes 2 and 4). As the theoretical molecular size of ASP is 64,314, the band detected must have been ASP itself. In addition to this band, a band of 78 kDa appeared in the lane to which the sample treated with DMS was applied (lane 4). The band did not appear in the sample without DMS treatment (lane 2). As the molecular weight of ORF2 protein in the periplasm is 13,735, the 78-kDa band was thought to be an association of ASP (64 kDa) and ORF2 protein (14 kDa). To verify the association, we examined the reactivity of the protein with anti-ORF2 protein serum. As shown in Fig. 1B, lane 5, the band reacted with the antiserum, showing that the 78-kDa protein was an association of ASP and ORF2 protein. The band was not detected in the sample that was not treated with a cross-linker (lane 7), meaning that the associated proteins were dissociated by heating them in SDS-dye solution before they were applied to the gel. To confirm the dissociation, we analyzed the amount of free ORF2 protein in the sample. To detect the free ORF2 protein, a gel containing a large amount of polyacrylamide (18%) was used, because ORF2 is small. As shown in Fig. 1C, the amount of free ORF2 protein in the sample that was not treated with a cross-linker was higher than that in the treated sample (lane 11 versus lane 9). This strongly indicated that ORF2 protein associated with ASP in the periplasmic space.

FIG. 1.

Cross-linking experiment to detect the association of ASP with ORF2 protein. The periplasmic fraction of E. coli HB101 cells harboring plasmid pUC119-5528, which contains the ASP operon, was incubated with 5 mM DMS (+), a cross-linker (15), or a distilled water (−) and then separated by SDS-PAGE. The reactivities of bands on the gel with anti-ASP antiserum and anti-ORF2 protein antiserum were examined by immunoblotting. A 10% acrylamide gel (10% gel) (A and B) and 18% acrylamide gel (18% gel) (C) were used to detect the proteins reactive to anti-ASP antiserum and anti-ORF2 protein antiserum, respectively. Numbers along the sides of the gels indicate molecular masses in kilodaltons. Arrows and asterisks indicate ASP and ASP-associated ORF2 protein, respectively.

A minor band appeared at a higher molecular weight than that of the main band in lane 5 of Fig. 1B. The sample was treated with a cross-linker, and the proteins which were linked with ORF2 protein were detected with the antiserum in this lane. The appearance of a minor band may indicate that there is another protein with which ORF2 protein associates. However, the properties of the protein are unclear at this time.

To confirm the association of ORF2 protein with ASP, we recovered the periplasmic proteins which were reactive with anti-ASP antiserum. A periplasmic solution of the cells (HB101/pUC119-5528) was prepared by treatment with lysozyme-EDTA (14). Then, the samples obtained were incubated with protein A-Sepharose resin to remove materials which nonspecifically react with the resin. After incubation, the resins were removed by centrifugation. Anti-ASP antiserum was added to the supernatant, and the mixture was incubated at 37°C for 1 h. Then, the protein A-Sepharose resin was added to the solution. After incubation for 1 h at 0°C, the resin was collected by centrifugation. The precipitates were washed five times with sodium phosphate buffer (pH 7.2). Next, they were mixed with SDS-dye solution and heated for 5 min in boiling water. After the heat treatment, the resin was removed by centrifugation. The supernatants obtained were separated by SDS-PAGE and the reactivities of the bands on the gel with antiserum were examined (Fig. 2).

FIG. 2.

Analysis of periplasmic proteins which are reactive to anti-ASP antiserum. The periplasmic fraction of HB101/pUC119-5528 was prepared by treatment with lysozyme-EDTA. The materials in the fraction which reacted with protein A-Sepharose resin were adsorbed to the resin by incubation. The resin was removed from the solution by centrifugation. Then the supernatant was incubated with anti-ASP antiserum. Subsequently, protein A-Sepharose resin was added to the mixture to recover the antibodies in the solution. The resin was collected by centrifugation and suspended in SDS-dye solution. After being heated in boiled water, the solutions were separated by SDS-PAGE. The reactivities of the bands with the antiserum were examined by immunoblotting. To examine the reactivity of the band with anti-ASP antiserum, 10% polyacrylamide gel was used (A), and to examine the reactivity of the band with anti-ORF2 protein antiserum, 18% polyacrylamide gel was used (B).

To examine reactivity with anti-ASP antiserum, a 10% gel was used. A protein band reacting with anti-ASP antiserum was detected (Fig. 2A). To detect the protein which was reactive with anti-ORF2 protein antiserum, an 18% gel was used. A band with a size of 14 kDa reacted with anti-ORF2 protein antiserum (Fig. 2B). This demonstrated that ORF2 protein associated with ASP in the periplasm, and the associated proteins were recovered with anti-ASP antiserum.

ORF2 protein is needed for construction of the active structure by ASP.

To elucidate the action of ORF2 protein, the ASP gene was expressed with or without the expression of ORF2 protein in an in vitro transcription-translation system. The part of the ASP gene which encodes the downstream region from the 25th amino acid residue from the amino terminus of ASP (the region appearing in the periplasm) was inserted at the NdeI site of plasmid pET11 (16). The prepared plasmid was designated pET11-ASP-1. Similarly, the part of the ORF2 gene which encodes the downstream region from the 23rd amino acid residue from the amino terminus of ORF2 protein (the region appearing in the periplasm) was inserted into the same position of another pET11 plasmid. This plasmid was designated pET11-ORF2-1. The transcription of these genes was dependent on the exogenously added T7 RNA polymerase.

We added the two plasmids, pET11-ASP-1 and pET11-ORF2-1, into one tube, and pET11-ASP-1 and pET11 (as the control for pET11-ORF2-1) were placed into a second tube. The genes were expressed with a Rapid Translation System RTS 100 E. coli HY kit (Roche Diagnostics, Mannheim, Germany). The reaction was carried out at 30°C for 24 h. After the reaction, the reaction mixture was separated by SDS-PAGE (18% gel), and the protein bands reactive with anti-ASP antiserum and with anti-ORF2 protein antiserum were detected by immunoblotting. To detect these protein bands, a mixture containing anti-ASP antiserum and anti-ORF2 protein antiserum was used. The result is shown in Fig. 3A. ORF2 protein was produced only in the first tube, as expected (lane 2), because the second tube did not contain the ORF2 gene. ASP was produced in both samples (lanes 2 and 3), showing that ASP was synthesized independently of ORF2.

FIG. 3.

Analysis of proteins synthesized by the in vitro transcription-translation system (A) and their proteolytic activities (B). (A) The structural genes of ASP and ORF2 protein were inserted into pET11 as described in the text. The genes inserted were expressed by in vitro transcription-translation. After expression, the reactant was separated by SDS-PAGE (18% polyacrylamide gel) and the reactivity of the band to a mixture containing anti-ASP antiserum and anti-ORF2 protein antiserum was examined by immunoblotting. The proteolytic activity of the reactant was determined by an assay using azocasein (17). (B) The amount of azocasein digested by the proteases was measured as absorbance at 450 nm.

Subsequently, the proteolytic activities of these reactants were examined. The reactants were diluted 10-fold with 10 mM Tris-HCl buffer (pH 7.4). Azocasein (Sigma, St. Louis, Mo.), a substrate of the protease, was added to the diluted sample at a ratio of 1% (wt/vol), and the sample was incubated at 37°C for 5 h. After the incubation, the absorbance at 450 nm of the sample was measured. The absorbance reflects the intensity of the proteolytic activity of the sample. The result is shown in Fig. 3B. The sample obtained from the tube containing both genes showed proteolytic activity (column 2), but the sample from the tube which did not contain the ORF2 gene did not. This means that ASP produced without the concomitant production of ORF2 protein cannot take an active form. From these results, we conclude that the ORF2 protein is a chaperone helping ASP to form an active structure in the periplasm.

It has been shown that several prokaryotic proteases are produced as a precursor containing the propeptide region which functions as an intramolecular chaperone in the folding of the mature region (1). Some intramolecular chaperones of proteases directly affect the folding of the mature region by lowering a high-energy barrier to the process (1, 3). A similar function was reported in the lipase-specific foldase (Lif), whose gene is located at the 3′ end of the lipase (LipA) gene of Burkholderia glumae (6, 7). As Lif contributes to the production of active LipA by helping to construct the proper structure, it has been classified as a steric chaperone (4, 5). Our ORF2 protein may also be a steric chaperone helping to form the correct structure of ASP, but the precise role of the protein has remained unclear.

In other chaperone proteins, functions such as prevention of the aggregation of host protein and the refolding of denatured host proteins have been reported (8, 11, 13). In order to examine whether ASP aggregates in the absence of ORF2 protein, we gathered the insoluble materials in the reaction mixture of the second tube of the experiment whose results are shown in Fig. 3 by ultracentrifugation and analyzed whether the insoluble materials contained ASP. ASP was not found in the materials (data not shown), indicating that ASP did not aggregate in the absence of ORF2 protein. In addition, we showed in the previous report that ASPs produced by living cells without the concomitant production of ORF2 protein were easily digested into pieces (16). This suggests that ASP forms a structure sensitive to proteases in the absence of ORF2 protein without aggregating. Therefore, it may not be necessary for ORF2 protein to possess the ability to prevent the aggregation of ASP and to refold denatured ASP, because ASP does not take such a formation even in the absence of ORF2 protein.