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

An open reading frame of gelonin (Gel), one of ribosome inactivating proteins, was inserted into the vector pBSL-C which contains 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 10 mM Tris buffer, 500 mM NaCl, pH 8.5, and about 2.4 mgGel/L culture with 96% homogeneity was directly eluted from the captured column by incubation at 25 degrees C under pH 6.5 for 48 h based on intein C-terminal self-cleavage. Western blot, ELISA, and in vitro inhibition of protein synthesis demonstrated that the bioactivity of recombinant 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.     

Refolding by High Pressure of a Toxin Containing Seven Disulfide Bonds: Bothropstoxin-1 from <Emphasis Type="Italic">Bothrops jararacussu</Emphasis>

Aggregation is a serious obstacle for recovery of biologically active heterologous proteins from inclusion bodies (IBs) produced by recombinant bacteria. E. coli transformed with a vector containing the cDNA for Bothropstoxin-1 (BthTx-1) expressed the recombinant product as IBs. In order to obtain the native toxin, insoluble and aggregated protein was refolded using high hydrostatic pressure (HHP). IBs were dissolved and refolded (2 kbar, 16 h), and the effects of protein concentration, as well as changes in ratio and concentration of oxido-shuffling reagents, guanidine hydrochloride (GdnHCl), and pH in the refolding buffer, were assayed. A 32% yield (7.6 mg per liter of bacterial culture) in refolding of the native BthTx-1 was obtained using optimal conditions of the refolding buffer (Tris–HCl buffer, pH 7.5, containing 3 mM of a 2:3 ratio of GSH/GSSG, and 1 M GdnHCl). Scanning electron microscopy (SEM) showed that that disaggregation of part of IBs particles occurred upon compression and that the morphology of the remaining IBs, spherical particles, was not substantially altered. Dose-dependent cytotoxic activity of high-pressure refolded BthTx-1 was shown in C2C12 muscle cells.     

Assembly-defective OmpC mutants of Escherichia coli K-12.

Novel ompC(Dex) alleles were utilized to isolate mutants defective in OmpC biogenesis. These ompC(Dex) alleles also conferred sensitivity to sodium dodecyl sulfate (SDS), which permitted the isolation of SDS-resistant and OmpC-specific phage-resistant mutants that remained Dex+. Many mutants acquired resistance against these lethal agents by lowering the OmpC level present in the outer membrane. In the majority of these mutants, a defect in the assembly (metastable to stable trimer formation) was responsible for lowering OmpC levels. The assembly defects in various mutant OmpC proteins were caused by single-amino-acid substitutions involving the G-39, G-42, G-223, G-224, Q-240, G-251, and G-282 residues of the mature protein. This assembly defect was correctable by an assembly suppressor allele, asmA3. In addition, we investigated one novel OmpC mutant in which an assembly defect was caused by a disulfide bond formation between two nonnative cysteine residues. The assembly defect was fully corrected in a genetic background in which the cell's ability to form disulfide bonds was compromised. The assembly defect of the two-cysteine OmpC protein was also mended by asmA3, whose suppressive effect was not achieved by preventing disulfide bond formation in the mutant OmpC protein.     

Techno-economic evaluation of an inclusion body solubilization and recombinant protein refolding process

Expression of recombinant proteins in Escherichia coli is normally accompanied by the formation of inclusion bodies (IBs). To obtain the protein product in an active (native) soluble form, the IBs must be first solubilized, and thereafter, the soluble, often denatured and reduced protein must be refolded. Several technically feasible alternatives to conduct IBs solubilization and on-column refolding have been proposed in recent years. However, rarely these on-column refolding alternatives have been evaluated from an economical point of view, questioning the feasibility of their implementation at a preparative scale. The presented study assesses the economic performance of four distinct process alternatives that include pH induced IBs solubilization and protein refolding (pH_IndSR); IBs solubilization using urea, dithiothreitol (DTT), and alkaline pH followed by batch size-exclusion protein refolding; inclusion bodies (IBs) solubilization using urea, DTT, and alkaline pH followed by simulated moving bed (SMB) size-exclusion protein refolding, and IBs solubilization using urea, DTT and alkaline pH followed by batch dilution protein refolding. The economic performance was judged on the basis of the direct fixed capital, and the production cost per unit of product (P(C)). This work shows that (1) pH_IndSR system is a relatively economical process, because of the low IBs solubilization cost; (2) substituting β-mercaptoethanol for dithiothreithol is an attractive alternative, as it significantly decreases the product cost contribution from the IBs solubilization; and (3) protein refolding by size-exclusion chromatography becomes economically attractive by changing the mode of operation of the chromatographic reactor from batch to continuous using SMB technology.     

Refolding of human beta-1-2 GlcNAc transferase (GnT1) and the role of its unpaired Cys 121

Human beta1-2N-acetylglucosaminyltransferase (hGnT1) lacking the first 103 amino acids was expressed as a maltose binding protein (MBP) fusion protein in inclusion bodies (IBs) in Escherichia coli and refolded using an oxido-shuffling method. GnT1 mutants were prepared by replacing a predicted unpaired cysteine (C121) with alanine (C121A), serine (C121S), threonine (C121T) or aspartic acid (C121D). A double mutant R120A/C121H, was generated to mimic Gly14, the Caenorhabditis elegans GnT1 counterpart to hGNT1. Each mutant hGnT1 was constructed as an MBP fusion protein and resultant IBs were isolated and refolded. Wild type hGnT1 and mutants C121A, C121S and R120A/C121H transferred UDP-GlcNAc to the glycoprotein acceptor Man(5)-RNAse B, whereas mutants C121T and C121D were inactive. These findings indicated that cysteine 121 has a structural role in maintaining active site geometry of hGnT1, rather than a catalytic role, and illustrates for the first time the potential utility of E. coli as an expression system for hGnT1.     

Protease-deficient DegP suppresses lethal effects of a mutant OmpC protein by its capture

The expression of assembly-defective outer membrane proteins can confer lethality if they are not degraded by envelope proteases. We report here that the expression of a mutant OmpC protein, OmpC(2Cys), which forms disulfide bonds in the periplasm due to the presence of two non-native cysteine residues, is lethal in cells lacking the major periplasmic protease, DegP. This lethality is not observed in dsbA strains that have diminished ability to form periplasmic disulfide bonds. Our data show that this OmpC(2Cys)-mediated lethality in a degP::Km(r) dsbA(+) background can be reversed by a DegP variant, DegP(S210A), that is devoid of its proteolytic activity but retains its reported chaperone activity. However, DegP(S210A) does not reverse the lethal effect of OmpC(2Cys) by correcting its assembly but rather by capturing misfolded mutant OmpC polypeptides and thus removing them from the assembly pathway. Displacement of OmpC(2Cys) by DegP(S210A) also alleviates the negative effect that the mutant OmpC protein has on wild-type OmpF.     

Molecular modelling of epitope presentation using membrane protein OmpC

Three-dimensional models of the chimeric S. typhi OmpC protein carrying an epitope from rotavirus VP4 capsid protein on either of two exposed loops (fourth and sixth) were constructed separately, using computer-aided homology modelling. The theoretical model of S. typhi OmpC was used as a template. The monomers were initially energy minimized. The trimers were generated for both the chimeric S. typhi OmpC proteins and the structures were optimized after several cycles of minimization. The surface accessibility calculations for the resulting models show that epitope recognition should be more effective in the fourth loop than in the sixth loop, in accordance with the experimental results on the immunogenic nature of the rotaviral epitope inserted into the two putative loops of S. typhi OmpC.     

Roles of lipopolysaccharide and outer membrane protein OmpC of Escherichia coli K-12 in the receptor function for bacteriophage T4

The roles of lipopolysaccharide and OmpC, a major outer membrane protein, in the receptor function for bacteriophage T4 were studied by using Escherichia coli K-12 strains having mutations in the ompC gene or in genes controlling different stages of lipopolysaccharide synthesis. The receptor activity for T4 was monitored by (i) T4 sensitivity of intact cells, (ii) phage inactivation activity of cell envelopes, and (iii) phage inactivation activity of specimens reconstituted from purified OmpC and lipopolysaccharide. It was found that (i) in the presence of the OmpC protein, the essential region of the lipopolysaccharide for the receptor activity was the core-lipid A region that includes the heptose region, whereas the glucose region was not necessarily required for the receptor function; (ii) the OmpC protein was not required at all when the distal end of the lipopolysaccharide was removed to expose a glucose residue at the distal end; and (iii) when cells lacked both the OmpC protein and the glucose region, they became extremely resistant to T4. Based on these findings, the roles of the OmpC protein and lipopolysaccharide in T4 infection are discussed.     

Role of lipopolysaccharide in assembly of Escherichia coli outer membrane proteins OmpA, OmpC, and OmpF.

Selection was performed for resistance to a phage, Ox2, specific for the Escherichia coli outer membrane protein OmpA, under conditions which excluded recovery of ompA mutants. All mutants analyzed produced normal quantities of OmpA, which was also normally assembled in the outer membrane. They had become essentially resistant to OmpC and OmpF-specific phages and synthesized these outer membrane porins at much reduced rates. The inhibition of synthesis acted at the level of translation. This was due to the presence of lipopolysaccharides (LPS) with defective core oligosaccharides. Cerulenin blocks fatty acid synthesis and therefore that of LPS. It also inhibits synthesis of OmpC and OmpF but not of OmpA (C. Bocquet-Pagès, C. Lazdunski, and A. Lazdunski, Eur. J. Biochem. 118:105-111, 1981). In the presence of the antibiotic, OmpA synthesis and membrane incorporation remained unaffected at a time when OmpC and OmpF synthesis had almost ceased. The similarity of these results with those obtained with the mutants suggests that normal porin synthesis is not only interfered with by production of mutant LPS but also requires de novo synthesis of LPS. Since synthesis and assembly of OmpA into the outer membrane was not affected in the mutants or in the presence of cerulenin, association of this protein with LPS appears to occur with outer membrane-located LPS.     

Purification effect of artificial chaperone in the refolding of recombinant ribonuclease A from inclusion bodies

Artificial chaperone (AC) containing cetyltrimethylammonium bromide (CTAB) and β-cyclodextrin (β-CD) has been used to refold recombinant ribonuclease A (RNase A) from inclusion bodies (IBs). At low urea concentration (0.8 M), the AC could enhance the refolding yield of RNase A by effectively suppressing its intermolecular interaction-induced aggregation. As a result, 0.9 mg/mL RNase A could be 77% refolded, which was a 57% increase as compared to that without the AC. At high protein concentration range (0.9–2.3 mg/mL in total protein concentrations) and 1.6 M urea, CTAB selectively precipitated contaminant proteins distinctly, so a purification effect was achieved. For example, 1.5 mg/mL RNase A could be 62% refolded and recovered at a purity of 87%, which was a 34% increase in purity as compared to that in IBs (65%). The precipitation selectivity was considered due to the differences in the hydrophobicity of the proteins. The work indicates that by using the AC, RNase A could be efficiently refolded at low urea concentration and purified at high urea concentration from IBs at high protein concentrations.     

Translational control of exported proteins that results from OmpC porin overexpression.

The regulation of synthesis and export of outer membrane proteins of Escherichia coli was examined by overexpressing ompC in multicopy either from its own promoter or from an inducible promoter in an expression vector. Overexpression of OmpC protein resulted in a nearly complete inhibition of synthesis of the OmpA and LamB outer membrane proteins but had no effect on synthesis of the periplasmic maltose-binding protein. Immunoprecipitation of labeled proteins showed no evidence of accumulation of uncleaved precursor forms of OmpA or maltose-binding protein following induction of OmpC overexpression. The inhibition of OmpA and LamB was tightly coupled to OmpC overexpression and occurred very rapidly, reaching a high level within 2 min after induction. OmpC overexpression did not cause a significant decrease in expression of a LamB-LacZ hybrid protein produced from a lamB-lacZ fusion in which the fusion joint was at the second amino acid of the LamB signal sequence. There was no significant decrease in rate of synthesis of ompA mRNA as measured by filter hybridization of pulse-labeled RNA. These results indicate that the inhibition is at the level of translation. We propose that cells are able to monitor expression of exported proteins by sensing occupancy of some limiting component in the export machinery and use this to regulate translation of these proteins.     

Isolation and preliminary characterization of wild-type OmpC porin dimers from Escherichia coli K-12

Escherichia coli K-12 strain PLB3255 contains a mutation in the ompF gene that results in a 15 amino acid deletion in the porin protein. The mutation (dex) appears to increase the OmpF channel size, allowing the PLB3255 cells to grow on maltodextrins in the absence of a functional maltoporin. Porin isolated from strain PLB3255, which contains a wild-type ompC gene, was separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and shown to contain 50-60% trimer aggregates and 35-40% of a 50-kDa "dimer". When the porin isolate was heated to 100 degrees C and separated on SDS gels containing 8 M urea, both the trimer and the "dimer" were recovered in a single band at 36 kDa that corresponded in mobility to wild-type OmpC porin. An analysis of the temperature stability of the isolate showed that the OmpC "dimer" was less stable and denatured at 66 degrees C compared to 81 degrees C for the trimer. To separate these aggregates, the unheated porin was suspended in 30% SDS, applied to a Sephadex G-200 gel filtration column, and eluted with 0.5% sodium deoxycholate. Two peaks were recovered containing separated trimers and "dimers". Circular dichroism spectra of isolated dimers and trimers indicated similar amounts of beta-structure. The isolated dimers and trimers were reconstituted into artificial membranes. Electrical conductance across planar bilayer lipid membranes and liposome swelling assays showed that the two isolates had similar channel-forming activity. Four other ompF deletion mutants of the same phenotype were also shown to produce 50-kDa OmpC porin "dimers".(ABSTRACT TRUNCATED AT 250 WORDS)     

A novel protein refolding method using a zeolite

We have succeeded in developing a simple and effective protein refolding method using the inorganic catalyst, beta-zeolite. The method involves the adsorption of proteins solubilized with 6M guanidine hydrochloride from inclusion body (IB) preparations onto the zeolite. The denaturant is then removed, and the proteins in the IBs are released from the zeolite with polyoxyethylene detergent and salt. All of the IBs tested (11 different species) were successfully refolded under these conditions. The refolded proteins are biochemically active, and NMR analysis of one of the proteins (replication protein A 8) supports the conclusion that correct refolding does occur. Based on these results, we discuss the refolding mechanism.     

Overexpression and purification of recombinant human interferon alpha2b in Escherichia coli

Overexpression of rhIFN-alpha2b was obtained by synthesizing a codon optimized gene for IFN-alpha2b and expressing it in the form of inclusion bodies (IBs) in Escherichia coli. The recombinant plasmid pRSET-IFNalpha, which had the IFN-alpha2b gene under the T7 promoter, was coexpressed with plasmid pGP1-2, which carried the gene for T7 RNA polymerase under the heat inducible lambdaP(L) promoter. This two plasmid expression system was optimized with respect to heat shock time, media, and time of induction in shake flask cultures. This was then scaled up into a bioreactor to get a maximum volumetric product yield of 5.2g/L at a final OD(600) of 67. At this point, the IBs represented approximately 40% of the total cellular protein. This high specific product yields eased the further downstream processing steps and improved product recoveries. The IBs were isolated and purified through ion exchange followed by step refolding to give a final product yield of approximately 3g/L, which is maximum reported in the literature. The bioassay of the refolded protein gave a specific activity of approximately 3 x 10(9)IU/mg protein.     

Isolation and characterization of OmpC porin mutants with altered pore properties.

The LamB protein is normally required for the uptake of maltodextrins. Starting with a LamB- OmpF- strain, we have isolated mutants that will grow on maltodextrins. The mutation conferring the Dex+ phenotype in the majority of these mutants has been mapped to the ompC locus. These mutants, unlike LamB- OmpF- strains, grew on maltotriose and maltotetraose, but not on maltopentaose, and showed a significantly higher rate of [14C]maltose uptake than the parent strain did. In addition, these mutants showed increased sensitivity to certain beta-lactam antibiotics and sodium dodecyl sulfate, but did not exhibit an increase in sensitivity to other antibiotics and detergents. The nucleotide sequence of these mutants has been determined. In all cases, residue 74 (arginine) of the mature OmpC protein was affected. The results suggest that this region of the OmpC protein is involved in the pore domain and that the alterations lead to an increased pore size.     

Purification of integral outer-membrane protein OmpC, a surface antigen from Salmonella typhi for structure-function studies: a method applicable to enterobacterial major outer-membrane protein

Extraction of the outer-membrane porin, OmpC, from Salmonella typhi Ty21a was done by using a modified salt-extraction procedure. It was possible to extract only the major outer-membrane protein (OMP) from the crude membrane using this method. Aberrant lipopolysaccharide (LPS) production in the galE mutant Ty21a has resulted in more isoforms of OmpC and subsequently led to anomalous mobility in SDS-PAGE. The purity of the preparation was confirmed by denaturing urea SDS-PAGE and N-terminal sequencing. The major OMP extracts had LPS of both bound and free forms. The free form of LPS could be removed by gel filtration and the bound form, largely, was removed using ion-exchange chromatography and by passing through ultrafiltration devices. This method has been used to extract the native trimer of OmpC, the major OMP, in a large scale, for structure-function studies. S. typhi Ty21a OmpC preparation yielded reproducible diffraction-quality crystals. Extracts of porin from wild-type Escherichia coli HB101, grown under high osmolarity conditions, showed a single species of OMP on SDS-PAGE. This suggests the possible application of the method to other gram-negative bacterial porins.     

The porin OmpC of Salmonella typhimurium mediates adherence to macrophages.

Macrophages recognize, adhere to, and phagocytose Salmonella typhimurium. The major outer membrane protein OmpC is a candidate ligand for macrophage recognition. To confirm this we used transposon mutagenesis to develop an ompC-deficient mutant in a known virulent strain of S. typhimurium; mutant and wild type were compared in macrophage adherence and association assays. Radiolabeled wild type S. typhimurium bound to macrophages at five-fold higher levels than did the ompC mutant. In association assays, macrophages in monolayers bound and internalized three-fold more wild type than mutant, while macrophages in suspension bound and internalized 40-fold more wild type than mutant. The ompC gene of our test strain of S. typhimurium contains several discrete differences compared with the ompC genes of Salmonella typhi and Escherichia coli. The deduced OmpC amino acid sequence of S. typhimurium shares 77 and 98% identity with OmpC amino acid sequence of E. coli and S. typhi, respectively. Evidence from this study supports a role for the OmpC protein in initial recognition by macrophages and distinguishes regions of this protein that potentially participate in host-cell recognition of bacteria by phagocytic cells.     

Crystallization of the immunodominant outer membrane protein OmpC; the first protein crystals from Salmonella typhi, a human pathogen.

OmpC, a surface antigen of Salmonella typhi was crystallized after several attempts, using PEG 3350. Well shaped hexagonal crystals were grown from vapor diffusion method using octyl glucoside and C12E9 as detergents. Crystals are sensitive to X-ray and diffract weakly up to 7 A. Porin isoforms, due to the bound lipopolysaccharides, could be the cause for poor diffraction. Crystal quality depends largely on the purification method, and in case of LPS contamination, the genetic background of the bacteria. Crystallization and initial data collection suggest optimum conditions and the method of choice for OmpC crystallization.     

Construction and deconstruction of bacterial inclusion bodies

Bacterial inclusion bodies (IBs) are refractile aggregates of protease-resistant misfolded protein that often occur in recombinant bacteria upon gratuitous overexpression of cloned genes. In biotechnology, the formation of IBs represents a main obstacle for protein production since even favouring high protein yields, the in vitro recovery of functional protein from insoluble deposits depends on technically diverse and often complex re-folding procedures. On the other hand, IBs represent an exciting model to approach the in vivo analysis of protein folding and to explore aggregation dynamics. Recent findings on the molecular organisation of embodied polypeptides and on the kinetics of inclusion body formation have revealed an unexpected dynamism of these protein aggregates, from which polypeptides are steadily released in living cells to be further refolded or degraded. The close connection between in vivo protein folding, aggregation, solubilisation and proteolytic digestion offers an integrated view of the bacterial protein quality control system of which IBs might be an important component especially in recombinant bacteria.     

Demonstration of a bacteriophage receptor site on the Escherichia coli K12 outer-membrane protein OmpC by the use of a protease

The Escherichia coli K12 outer-membrane proteins OmpA, OmpC, OmpF, PhoE, and LamB (all of transmembrane nature) can serve as phage receptors. We have shown previously that one OmpA-specific phage, Ox2, can give rise to the host range mutants Ox2h10 and Ox2h12, with the latter being derived from the former [Morona, R. & Henning, U. (1984) J. Bacteriol. 159, 579-582]. Unlike Ox2, both host range phages can use the OmpA and OmpC proteins as receptors and Ox2h12 is better adapted to the OmpC protein than Ox2h10. In a search for the site(s) of OmpC protein involved in phage recognition, it was found that proteinase K is able to cleave all of the proteins mentioned above. OmpC protein (Mr = 38306) could be cleaved from outside the cell by proteinase K resulting in two fragments of Mr approximately equal to 21000 and Mr approximately equal to 17500. The use of OmpC-PhoE hybrid proteins allowed us to assign the approximately equal to 21000-Mr fragment to the CO2H-terminal moiety of the protein. Proteinase K treatment of intact cells abolished their activity to neutralize the OmpC-specific phage Tulb and reduced this ability towards phage Ox2h12. The OmpA, OmpF, PhoE and LamB proteins were cleaved by the protease not in intact cells but only when acting on cell envelopes. The sizes of the OmpC protein fragments and the results obtained with the hybrid proteins very strongly suggest that the protein is cleaved from outside the cell at a region involving amino acid residues 150-178 of the 346-residue protein, which shows homology to two regions of the OmpA protein which are involved in its phage receptor site (loc. cit.). These areas also exhibit some homology to a region of the LamB protein which is thought to be part of this protein's receptor site [Charbit et al. (1984) J. Mol. Biol. 175, 395-401]. This suggests that there is a common denominator for proteinaceous phage receptor site because the LamB-specific phage lambda and phage Tulb are of completely different nature. We conclude that the region of the OmpC protein in question is cell-surface-exposed and acts as a phage receptor site.