• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of jbacterPermissionsJournals.ASM.orgJournalJB ArticleJournal InfoAuthorsReviewers
J Bacteriol. Sep 2004; 186(17): 5919–5925.
PMCID: PMC516829

Substrate Specificity of the Escherichia coli Outer Membrane Protease OmpT


OmpT is a surface protease of gram-negative bacteria that has been shown to cleave antimicrobial peptides, activate human plasminogen, and degrade some recombinant heterologous proteins. We have analyzed the substrate specificity of OmpT by two complementary substrate filamentous phage display methods: (i) in situ cleavage of phage that display protease-susceptible peptides by Escherichia coli expressing OmpT and (ii) in vitro cleavage of phage-displayed peptides using purified enzyme. Consistent with previous reports, OmpT was found to exhibit a virtual requirement for Arg in the P1 position and a slightly less stringent preference for this residue in the P1′ position (P1 and P1′ are the residues immediately prior to and following the scissile bond). Lys, Gly, and Val were also found in the P1′ position. The most common residues in the P2′ position were Val or Ala, and the P3 and P4 positions exhibited a preference for Trp or Arg. Synthetic peptides based upon sequences selected by bacteriophage display were cleaved very efficiently, with kcat/Km values up to 7.3 × 106 M−1 s−1. In contrast, a peptide corresponding to the cleavage site of human plasminogen was hydrolyzed with a kcat/Km almost 106-fold lower. Overall, the results presented in this work indicate that in addition to the P1 and P1′ positions, additional amino acids within a six-residue window (between P4 and P2′) contribute to the binding of substrate polypeptides to the OmpT binding site.

The outer membrane protein T (OmpT) of Escherichia coli is a surface membrane serine protease and is the prototypical member of the omptin family of gram-negative bacteria (19). More than 20 years ago, Leytus and coworkers showed that OmpT catalyzes the activation of plasminogen to plasmin (17), a function that is physiologically relevant for the virulence of Yersinia pestis and for clinical E. coli isolates (18, 28). OmpT has also been found to play a role in bacterial virulence in ways that are unrelated to plasminogen activation, for example in the cleavage of protamine and other cationic peptides with antibiotic activity (11, 29).

OmpT folds into a 10-strand antiparallel β-barrel conformation with extracellular loops that extend well beyond the membrane (32). The active site is located within a deep groove formed by loops L4 and L5 on the one side and L1, L2, and L3 on the other. The structure also reveals a binding site for a single lipopolysaccharide molecule that appears to be important for the catalytic activity of the enzyme (14, 15, 16). The finding that peptide hydrolysis is weakly inhibited by certain serine protease inhibitors was the basis for the classification of OmpT and its homologues as a distinct serine protease family. However, the large distance between the putative catalytic Ser99 and the His212 seen in the crystal structure subsequently led to the hypothesis that OmpT functions through a novel mechanism involving an Asp210-His212 catalytic dyad that together with Asp83-Asp85 activates a putative nucleophilic water molecule (32).

OmpT and its homologues cleave synthetic substrates between dibasic residues with high catalytic efficiency (14, 19, 30). The cleavage of sequences containing dibasic residues has been shown to be important for the inactivation of antibiotic peptides and colicins, the proteolysis of bacterial membrane proteins in trans, and the degradation of recombinant proteins expressed in E. coli (2, 4, 11, 12, 20, 25, 26, 29, 30). Since the enzyme is a membrane protein, it fractionates with the insoluble fraction in cell lysates and copurifies with protein inclusion bodies (33). It retains activity under denaturing conditions, including boiling or in the presence of up to 4 M urea, and therefore it can be a major source of protein degradation during the solubilization and renaturation of inclusion bodies (26, 33).

Dekker et al. (5) examined the substrate specificity of OmpT using immobilized tetrapeptide libraries in which each of the P2, P1, P1′, and P2′ positions was sequentially substituted with all other amino acids. They reported cleavage of peptides with Ile, His, Ala, Phe, Pro, Leu, Met, Gln, Asn, or Val in the P1′ and that Ile is slightly preferred relative to Arg at that position. In contrast, the cleavage of protein substrates by OmpT occurs exclusively between dibasic peptides or, in the case of human plasminogen, between an Arg-Val sequence (19). The differences between the cleavage of proteins and of the peptide substrates analyzed by Dekker et al. (5) indicate that residues beyond the P2 and the P2′ position contribute to the substrate specificity of the enzyme. Consistent with this hypothesis, the cleavage of fusion proteins in the presence of 4 M urea was reported to depend on the identity of the P4 residue (23).

In this work we employed a novel version of substrate phage display (21) to analyze the extended substrate specificity of OmpT. Substrate phage display is based on the selective cleavage of specific peptide sequences sandwiched between the gene III minor coat protein of M13 bacteriophage and an affinity tag. The phage is immobilized on a solid support via the affinity tag and, following treatment with a purified protease, clones containing susceptible peptide sequences are cleaved, released from the support, and amplified. This process is repeated several times until clones containing a consensus sequence that corresponds to the preferred cleavage site of the enzyme are isolated. We find that for proteases displayed on the cell surface, such as OmpT, cleavage can occur in situ during the growth of the phage, circumventing the need for treatment of the phage library with purified protein. Our analysis shows very strong preference for basic residues in the P1 and P1′ positions and also significant preference for amino acids occupying P4, P3, and P2′. Further, OmpT was found to exhibit very high catalytic activity with optimal peptide substrates.



Electrocompetent XL1-Blue E. coli [F′ proAB lacIqZΔM15 Tn10 (Tetr)] was purchased from Stratagene. E. coli strains NK5507 [F′ lacIq lacZp-4008(L8) lacI4500::Tn10 (Tetr)] and AD202 {F [araD139] Δ(argF-lac)169 ompT::kan flhD5301 fruA25 relA1 rpsL150(Strr) rbsR22 deoC1} were obtained from the E. coli Genetic Stock Center (Yale University). An F′ Tetr episome was transferred to the latter strain by conjugation. Pansorbin (protein A-bearing Staphylococcus aureus cells) was purchased from Calbiochem. Immobilized protein A on Biomag beads was obtained from Perseptive Biosystems. Peptides were synthesized by Chiron Corp. (Emeryville, Calif.). The monoclonal antibodies (MAb) anti-Glu (Chiron Corp., Emeryville, Calif.) and MAb anti-T7 (Novagen) recognize the peptide epitopes EYMPME and MASMTGGQQMG, respectively. Bacteriophage lp140 is a derivative of M13mp19 (7, 34).

Construction of bacteriophage libraries.

Sequences encoding the Glu and T7 epitopes were fused to gene III by standard PCR techniques with bacteriophage lp140 DNA as template, a 5′ primer containing KpnI and NcoI restriction sites (5′-CTTTAGTGGTACCTTTCTATTCTCACTCCGCTGAATACATGCCAATGGAAGGAATGGCTAGCATGACTGGTGGACAGCAAATGGGTCCATGGGTTACAATTGAAAGTTGTTTAG), and a 3′ primer containing an AlwNI site (5′-GGAAAGCGCAGTCTCTG), followed by ligation of the KpnI/AlwNI-digested PCR product into lp140. DNA for the tagged 6-mer library was synthesized by PCR using 140T DNA (100 pg) as template, a 5′ primer (10 pmol) 5′-AAATGGGTCCATGGGGCGGTNNKNNKNNKNNKNNKNNKGGTACAATTGAAAGTTG (where N represents equimolar A, C, G, or T, and K is equimolar G or T), and a 3′ primer (5 pmol) 5′-GGAAAGCGCAGTCTCTG, followed by digestion with NcoI/AlwNI and ligation into 140T. Library DNA was ligated at a ratio of 3:1 insert to vector DNA. The ligated DNA (~1 μg) was used to transform electrocompetent E. coli XL1-Blue (~80 μl) via electroporation. Immediately following electroporation, the cells were added to 1 ml of 10× SOC medium, allowed to grow 1 h at 37°C, and then added to 1 liter of SB medium containing 100 μg of ampicillin/ml. The cells were grown for 24 h (with further addition of ampicillin to 100 μg/ml at 8 and 18 h) and pelleted at 4,500 × g for 15 min. The supernatant was filtered, and the phage was precipitated by the addition of 0.1 volume of 25% polyethylene glycol 8000 containing 2.5 M NaCl and centrifugation at 14,000 × g for 30 min. The resulting pellet was resuspended in 1 ml of 10% glycerol-Tris-buffered saline. This primary library was aliquoted and stored at −80°C. Bacteriophage libraries in the ompT-deficient strain AD202 F′ were also constructed by infecting the cells with 1 × 1,010 to ~1 × 1,011 PFU, followed by growth in 1 liter of SB medium containing 100 μg of ampicillin/ml for 24 h.

Bacteriophage selection.

To select for bacteriophage displaying sequences that were cleaved in situ, ~1 × 1,010 (10 μl) PFU isolated from E. coli XL1-Blue cells were suspended in a solution containing 50 mM HEPES, pH 7.6, 3 mM MgCl2, 1 mM dithiothreitol, 0.1% bovine serum albumin, and 50% glycerol (250 μl) and were incubated 30 min at 25°C. Aliquots of MAb anti-Glu (10 μg) and MAb anti-T7 (5 μg) were added. After 30 min on ice, Pansorbin cells (100 μl) were added, the mixture was rocked gently for 30 min at 4°C, and the mixture was centrifuged for 2 min. The supernatant was recovered, and the process was repeated. An aliquot (5 μl) of the final supernatant was used to determine output bacteriophage titer, while the remaining bacteriophage-containing solution (~350 μl) was used to infect 1 ml of log-phase XL1-Blue cells, by incubation for 15 min at 25°C, followed by addition of 20 ml of 2YT medium containing 100 μg of ampicillin/ml and growth overnight at 37°C. Amplified bacteriophage was used to infect fresh cells, and the selection and subsequent amplification procedures were repeated for several rounds. Individual clones were selected from the titer plates for sequencing or for growth in 2-ml cultures for isolation of individual bacteriophage clones.

To select for bacteriophage displaying sequences cleaved by exogenous OmpT, 1 × 1,010 library phage isolated from E. coli AN1 (ompT) cells was suspended in a solution containing 50 mM HEPES, pH 7.6, 3 mM MgCl2, 1 mM dithiothreitol, 0.1% bovine serum albumin, and 50% glycerol (250 μl). Aliquots of MAb anti-Glu (10 μg) and MAb anti-T7 (5 μg) were added, and the mixture was incubated for 30 min on ice. Pansorbin cells (100 μl) or Biomag beads bearing immobilized protein A (50 μl) were added, and the resulting suspension was rocked gently for 60 min at 4°C followed by microcentrifugation for 2 min and washing of the cells (or beads) six times with 300 μl of 50 mM Tris-HCl, pH 7.5, and 50 mM NaCl. The cells (or beads) were resuspended in 100 μl of this buffer and incubated 90 min at 37°C with 160 to 800 nM purified OmpT. After microcentrifugation, an aliquot (5 μl) of the supernatant was used to determine the output bacteriophage titer, while the remaining bacteriophage-containing solution (~100 μl) was amplified by addition to mid-exponential-phase E. coli AN1 cells (1 ml). After incubation for 15 min at 25°C, 20 ml of 2YT medium containing 100 μg of ampicillin/ml was added, and the cells were grown overnight at 37°C. Amplified bacteriophage was isolated and processed as described above.

Cells from isolated plaques were transferred with a sterile toothpick into 2YT medium (2 ml) with 100 μl of log-phase cells, and the DNA region encoding the randomized peptide insert was sequenced.

OmpT protease expression and purification.

Cells transformed with plasmid pML19, which consists of a 2.0-kbp EcoRI-PstI chromosomal fragment from E. coli K-12 containing the ompT gene cloned into pUC19 (10, 24), were grown in Luria-Bertani media supplemented with 100 μg of ampicillin/ml at 37°C. OmpT was purified from stationary-phase cells according to the procedure of Mangel et al. (19). Protein concentrations were determined according to the method of Bradford (3). Extracted protein preparations routinely exceeded 90% purity as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis.

Peptide proteolysis assays.

Appropriate concentrations of OmpT and peptide were incubated at 37°C in 100 μl of phosphate-buffered saline, pH 7.2, for various times. In each reaction, enzyme concentrations (typically, 0.1 to 10 nM) and reaction times were chosen such that a maximum of 20% of the substrate was consumed. Forty microliters of the reaction was then immediately injected into a Waters 626 liquid chromatography pump using a C18 reverse-phase column (Supelco) and analyzed by a Waters 996 photodiode array detector. The separation was performed on a gradient that consisted of the following steps: (i) 5% acetonitrile (AcCN)-95% H2O-0.07% trifluoroacetic acid (TFA) for 5 min; (ii) 5% AcCN-95% H2O-0.07% TFA to 95% AcCN-5% H2O-0.07% TFA over 30 min; (iii) 95% AcCN-5% H2O-0.07% TFA for 5 min; (iv) 95% AcCN-5% H2O-0.07% TFA to 5% AcCN-95% H2O-0.07% TFA over 5 min; and (v) 5% AcCN-95% H2O-0.07% TFA for 5 min. The initial rates calculated were fitted to the Michaelis-Menten equation, and values of kcat and Km were determined. The amino acid sequence of the reaction products was determined by first pooling the appropriate fractions. The collected fractions were lyophilized in a centrifugal vacuum concentration system and dissolved in 50% AcCN-0.1% TFA to a concentration of 1 μg/μl and subjected to electrospray ionization mass spectrometry (ESI/MS) (Finnegan LCQ mass spectrometer, 4.5 kV laser voltage).


Library construction and screening.

We constructed a phage library displaying the peptide extension NH2-AEYMPMEGMASMTGGQQMGPWGGXXXXXXGTIES (where X can be any amino acid) at the N terminus of the M13 protein III. The invariant sequence comprised the peptide epitopes for the MAb anti-Glu and MAb anti-T7 (as the affinity domains) and was followed by a randomized region of six amino acids flanked by Gly residues. The randomized sequence was encoded by an NN(G/T) scheme, where N is A, G, C, or T. Following three rounds of electroporation, a total of 108 independent transformants were obtained. Sequencing of 30 randomly selected clones from this library revealed that 100% (30/30) contained inserts. As expected, the sequences of the 30 randomly selected phage clones showed a random distribution of amino acids (data not shown).

The phage library was subjected to several rounds of selection following treatment with purified OmpT (21). Briefly, the phage library was amplified in an (ompT-negative) E. coli strain and immobilized via the affinity domain, and unbound bacteriophage was removed by extensive washing. Clones displaying protease-susceptible sequences were specifically cleaved and released from the solid support by treatment with purified recombinant OmpT protease. After five rounds (OmpT concentration, 800 nM in rounds one to three and 160 nM in rounds four to six), the enrichment of recovered bacteriophage was 6% of the input titer. For comparison, when the library was cycled through the same number of rounds in the absence of added OmpT, only 0.1% of input titer was recovered. As is frequently the case following many rounds of phage library screening, a slight de-enrichment was observed following the fifth round. Thus, clones from the fifth round were selected at random, and the sequence around the hexapeptide region was determined by DNA sequencing (Table (Table11).

Amino acid sequences of the randomized region in clones selected in the presence of exogenous OmpT protease by the bound cleavage methoda

In a separate experiment, the bacteriophage library was amplified in E. coli (ompT+) cells without incubation with added purified protease. During the course of bacteriophage amplification, susceptible sequences in the randomer may be cleaved either as the bacteriophage particle emerges from the cell through the pIV pore or, most likely, in trans. Following phage amplification, uncleaved phages were captured and removed via their affinity domains (27). Protease-cleaved phages that remained in the supernatant were recovered, amplified in XL1-Blue cells, and subjected to further rounds of selection. After four rounds, the output phage represented 12% of the input. The sequences of clones selected after the fourth round are shown in Table Table22.

Amino acid sequence of the randomized region in clones selected by the solution cleavage methoda

Determination of a common consensus sequence.

The randomized hexamer sequences selected in the presence of exogenous OmpT were aligned such that the nearly invariant consecutive basic residues occupy the P1 and P1′ positions of the substrate, in accordance with the known specificity of OmpT for cleavage between paired basic residues. The frequency distribution of residues occupying different positions indicated that the enzyme has an overwhelming preference for Arg in the P1 or in P1′, with Lys as the only observed alternative (in 6 out of 21 clones; Table Table1).1). Val is preferred in the P2′ position (15 out of 21 clones), and we also observed an elevated frequency of Ala in P1′. Further, Arg and Trp are preferred in P3 and in P4 (Arg, 15 of 21 clones in P4; Arg or Trp in P3, 18 of 21 clones). Thus, although interactions with the P1 and P1′ amino acids of the substrate are critical for recognition by OmpT, preferences in the P3 to P2′ positions of the substrate are also likely to be important in determining specificity. The data in Table Table11 are consistent with an apparent preference for Gly in the P3′ and Gly or Thr in P4′. We note, however, that these amino acids are derived from the invariant region of the bacteriophage and therefore are not likely to have been specifically selected.

Sequences selected from libraries incubated with cells expressing OmpT were aligned, and the P1 and P1′ residues were identified as above (Table (Table2).2). Clone 4-13 contained only one Lys within the randomized sequence, and this residue was assigned to the P1 position. Clones 4-6 and 4-9 did not contain consecutive basic residues. In clone 4-6 there are two Arg-Gly dipeptides, and the identity of the peptide bond that serves as the primary cleavage site could not be ascertained. In clone 4-9, the Arg-Val sequence was assigned to the P1 and P1′ positions. Clone 4-9 also contains an Arg-Trp sequence, and cleavage could potentially occur between these two amino acids. However, OmpT does not cleave before a Trp residue (5; also data not shown). Furthermore, as discussed above, Arg and Trp are preferred in the P3 and P4 positions, consistent with cleavage occurring between the Arg-Val dipeptide of 4-9.

The data in Table Table22 confirm the virtual requirement for an Arg residue in P1, with Lys (in 2 out of 30 clones) as the only observed alternative. There is a slightly less stringent preference for Arg (23 of 30 clones) or Lys (4 of 30 clones) residues in the P1′ position, with Gly or Val residues also found at that position (3 of 30 clones). There is a strong propensity for Val or Ala residues in P2′ (21 out of 30 clones). Acidic residues appear to be disfavored in P2, as none of the 51 individual clones selected in the two experiments contained Asp or Glu at that position. In addition to the occurrence of Ala in P2, an elevated frequency of Gly, Tyr, or Phe is also observed in that position. A basic residue or Trp is strongly preferred at the P3′ position, whereas in P4′ there is a high frequency of tryptophan and, to a lesser extent, Arg (6 of 30 clones). We note, however, that many of the Trp residues in P4′ are derived from the invariant portion of the linker sequence in the bacteriophage.

Hydrolysis rates of consensus-derived peptide substrates.

Six peptides, corresponding to two sequences selected by the bound cleavage method and four sequences selected by the solution cleavage technique, as well as a peptide that included the known OmpT cleavage site in plasminogen, were synthesized by solid-phase synthesis. Each peptide consisted of 12 amino acids (the 6-mer randomer region and flanking constant region amino acids from the phage pIII) and featured a C-terminal carboxamide and an N-terminal N-acetyl group (with the exception of the plasminogen cleavage site peptide, where the N-acetyl peptide was only sparingly soluble and the uncapped peptide was substituted). The rates of hydrolysis were determined by quantitating the amount of product produced in a given incubation time by high-pressure liquid chromatography, and the kinetic parameters kcat and Km were determined by fitting the observed rate data to the Michaelis-Menten equation (Table (Table33).

Kinetic parameters for the OmpT-catalyzed hydrolysis of peptides corresponding to sequences selected by bacteriophage display methodsa

Peptides 1, 2, and 3, corresponding to sequences of the most frequently selected phage clones 5-3, 5-4, and 4-12, respectively, were indeed the most readily cleaved substrates of OmpT. kcat/Km values of the selected peptides ranged from 7.3 × 106 M−1 s−1 (for peptide 1, clone 5-3) to 4.2 × 104 M−1 s−1 (for peptide 6, clone 4-28). kcat/Km values of the selected substrate sequences spanned a 200-fold range, while determined Km values differed by at most sixfold. The increases in catalytic efficiencies of the selected peptide substrates (compared to control plasminogen peptide 7) evidently were driven primarily by lowering the energy of the transition states, manifested in greater kcat, rather than lowering the energy of the ground states, which would be manifested in lower Km. OmpT-catalyzed hydrolysis rates for peptides 1, 2, 3, 4, and 5 varied by only an order of magnitude and were sufficiently rapid (~106 to 107 M−1 s−1) so that these rates would be expected to be at least partially diffusion controlled. In that sense, these peptides approach the most highly optimized substrates possible. By comparison, the rate of hydrolysis of the plasminogen-derived peptide 7 was about 5 orders of magnitude lower.

ESI/MS analysis identified the position of the OmpT cleavage sites in each peptide. In all cases, cleavage occurred between the two basic residues of each peptide. In the plasminogen-derived peptide, cleavage occurred at the Arg-Val bond, as expected. It is of note that two of the peptides, 1 and 2, corresponding to the sequences displayed in the 5-3 and 5-4 clones, have two dibasic amino acid motifs, and thus two potential OmpT cleavage sites. ESI/MS analysis confirmed that peptide 1 was cleaved at the Arg-Lys bond (i.e., WGGRWAR↓KKGTI), while peptide 2 was attacked at the C-terminal Arg-Arg bond (i.e., WGGRRSR↓RVGTI). Cleavage between these particular consecutive basic residues in 1 and 2 may be directed by the occupation of the S4 subsite by an arginyl side chain. This is consistent with the consensus sequence derived from alignment of the sequences selected by either bacteriophage display method.


In this work, the optimal subsite occupancy of OmpT was analyzed by substrate phage display (21). This method has been employed to define the subsite specificities of a number of proteases, including factor Xa (21), mutant subtilisins (1), matrix metalloproteases (27), herpes simplex virus protease (22), human tissue plasminogen activator (h-tPA) (6), and human urokinase-type plasminogen activator (13). Elucidation of protease specificity via this technique is based upon the selective cleavage of specific peptide sequences from within a large peptide library (displayed as fusions to the gene III minor coat protein of M13 bacteriophage) that are susceptible to proteolysis upon treatment with the protease of interest. In the original version of the method (21), bacteriophage from a library displaying an appropriate randomer region was captured via an engineered affinity domain at the N terminus of pIII that bound a ligand immobilized on a solid support. Clones displaying protease-susceptible sequences in the randomer were cleaved, leading to their release from the solid support into the supernatant. In a subsequent variation of this method (27), library bacteriophage was cleaved in solution with the protease of interest, followed by the capture and removal of uncleaved bacteriophage via the binding of their intact affinity domains to immobilized ligand.

One potential drawback of both approaches is that bacteriophage clones displaying certain sequences may be cleaved by unrelated enzyme contaminants present in the purified protease preparation. This problem may be circumvented if the protease of interest is expressed in E. coli in a manner that allows access to the pIII fusion on the phage. Proteins that are either natively or heterologously expressed on the cell surface can cleave assembled phage particles that have been released from the cell. Cleavage by a protease expressed on the cell surface in situ is thus technically simpler and avoids artifacts that may arise due to the contamination of the target enzyme with unrelated proteases. While OmpT is a native E. coli outer membrane enzyme, this strategy may also be employed for normally soluble proteases, provided that they are expressed as fusions to surface anchoring sequences (8, 9).

We observed slight differences (cf. Tables Tables11 and and2)2) between the frequency distribution of amino acids surrounding the cleavage site in phage obtained from phage treated with purified OmpT compared to phage obtained from cells that express the protease endogenously. These differences may reflect biases in the two methods of library screening. First, possible contamination of the purified OmpT preparation (>90% purity as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis) with other proteases may account for a low level of cleavage by unrelated proteases. Second, clones that weakly bind antibody or clones precleaved by other endogenous proteases would be expected to be depleted in a phage population selected following treatment of immobilized phage by incubation with purified OmpT, while these clones would be enriched in a phage population selected by solution cleavage methods (27).

A number of proteins and peptides that are cleaved by OmpT are known (19, 30, 33). The amino acid sequences adjacent to the scissile bond of several known OmpT substrates are shown in Table Table4.4. Inspection reveals several features in common with the consensus obtained in this study: the invariant basic residue in the P1 position, the nearly invariant Arg or Lys residues in P1′ (with the notable exception of a single valyl residue in human plasminogen), and the occurrence in several of these previously known substrates of the consensus amino acid residues most often found by substrate bacteriophage display in the P4, P3, and P2′ positions. Indeed, amino acid sequences near the cleavage sites of two of the known protein substrates, parathyroid hormone and Torpedo californica creatine kinase (WLR KKL and IYK KLR, respectively), each differ from selected bacteriophage sequences by only two (clone 5-03, WAR KKG) or three (clone 4-18, GYR KMR) conservative substitutions in the region spanning the presumed P3-P3′ positions (31, 33). On the other hand, a few differences between the sequences identified by the phage method and the cleavage of known substrates by OmpT were also apparent. For example, whereas none of the clones isolated by bacteriophage display contained an acidic amino acid in P2, two of the known protein substrates, creatine kinase from Torpedo californica and rabbit, feature a Glu at that position. A potential limitation of bacteriophage display methods is that many factors, in addition to the selection method employed, influence the sequences selected, including codon usage, expression levels, and proper export to the bacteriophage surface.

Amino acid residues in previously identified substrate peptides and proteinsa

It has been previously noted (30, 31, 33) that in peptides or proteins that feature various consecutive basic residues, OmpT-catalyzed cleavage occurs most rapidly between Arg-Arg dipeptides. Such a propensity is consistent with our finding of predominantly Arg-Arg-containing peptide sequences by both bacteriophage screening methods and contrasts with a previous report which found cleavage predominantly within lysine-lysine-containing peptides (5). We also noted a number of additional differences in the OmpT substrate specificity profile determined by bacteriophage selection and the earlier analysis of its specificity using synthetic tetrapeptide SPOT (5). First, unlike the data of Dekker et al. (5), we find that other than basic residues, only Gly and Val can be accepted to a small degree in P1′. As reported in Table Table3,3, the rate of cleavage of a peptide containing an Arg-Val sequence is 5 orders of magnitude lower than that of optimal substrates. In contrast, Dekker et al. (5) reported an appreciable cleavage of peptides containing many other amino acids in P1′, including His, Ala, Ile, Gln, Met, and others. However, when we examined the cleavage of longer peptides with His or Ala in P1′, we could not detect any cleavage with purified OmpT even after 24 h of incubation (data not shown). Second, our substrate phage data show that Ser and Thr, although not preferred, can nonetheless be accepted in P2, whereas in the SPOT libraries these residues were not allowed at all at that position. A possible explanation for these and other more minor discrepancies between the results of the SPOT libraries and the substrate phage data shown here is that the former utilized tetramer peptides and therefore interactions between substrates and the S4 and S3 subsites of OmpT could not occur. In fact, we find that interactions of the substrate with the S4 and S3 subsites are likely to be responsible for determining the site where cleavage occurs in peptides that contain two Arg-Arg dipeptides close to each other, as is the case with substrate 5-4.

Whereas with optimal substrates the catalytic efficiency of OmpT approaches the diffusion limit, the kcat/Km value obtained with peptide 7, containing the plasminogen cleavage site, is only 17 M−1 s−1. Nonetheless, it is of note that the catalytic efficiency of OmpT with the plasminogen peptide is significantly greater than the value reported for h-tPA (kcat/Km = 0.29 M−1 s−1 [6]) or for human plasminogen-type urokinase (kcat/Km = 0.88 M−1 s−1 [13]). Much like OmpT, both h-tPA and urokinase cleave intact plasminogen with a far greater efficiency than peptide 7 (6, 13, 19). All three proteins cleave optimal peptide substrates with rate constants that are many orders of magnitude greater than those observed with the plasminogen target sequence. Urokinase and h-tPA cleave peptides with sequences deduced from substrate phage analyses at rates about 5,300 times faster than the plasminogen target sequence (6, 13).

An earlier substrate phage analysis of h-tPA (6) revealed that this enzyme's preference for amino acid side chains in the P1, P2′, and P3′ positions is strikingly similar to that of OmpT. Specifically, it was reported that the P1-P2′ consensus sequence cleaved by h-tPA is R-X-(G/A), where X is most commonly (75%) Arg and gaps are represented by hyphens. Proteases belonging to the omptin family exhibit no sequence homology with h-tPA or for that matter any other members of the serine protease family. Indeed, the crystal structure of OmpT (32) confirmed that OmpT, like other outer membrane proteins, folds into a β-barrel. Yet, despite their drastically different overall fold, lack of sequence homology, and phylogenetic origin, OmpT and h-tPA exhibit many similarities in terms of their substrate preference. The physiological role of human plasminogen activator is solely fibrin-dependent activation of plasminogen to plasmin. While h-tPA is able to catalyze the cleavage of short peptides up to 5,000-fold more efficiently than a similarly sized peptide containing the plasminogen cleavage site (6), there is no evidence that cleavage of such peptides is of physiological significance. In contrast, both OmpT's ability to cleave plasminogen and to hydrolyze cationic antibiotic peptides appear to be exploited by bacteria in the course of pathogenesis.


We thank Jill Winter (Chiron Corp.) for helpful discussions and Philip Stewart for assistance with kinetics experiments.

This work was supported by the Director, Office of Energy Research, Office of Basic Energy Sciences, Divisions of Material Sciences and of Energy Biosciences of the U.S. Department of Energy under contract no. DE-AC03-76SF00098 to Lawrence Berkeley National Laboratory (J.F.K.) and by a grant from NSF-BES (G.G.). Financial support in the form of a postdoctoral fellowship (to J.D.M.) from the Medical Research Council of Canada is gratefully acknowledged.


1. Ballinger, M. D., J. Tom, and J. A. Wells. 1995. Designing subtilisin BPN′ to cleave substrates containing dibasic residues. Biochemistry 34:13312-13319. [PubMed]
2. Baneyx, F., and G. Georgiou. 1990. In vivo degradation of secreted fusion proteins by the Escherichia coli outer membrane protease OmpT. J. Bacteriol. 172:491-494. [PMC free article] [PubMed]
3. Bradford, M. M. 1976. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72:248-254. [PubMed]
4. Cavard, D., and C. Lazdunski. 1990. Colicin cleavage by OmpT protease during both entry into and release from Escherichia coli cells. J. Bacteriol. 172:648-652. [PMC free article] [PubMed]
5. Dekker, N., R. Cox, A. Kramer, and M. Egmond. 2001. Substrate specificity of the integral membrane protease OmpT determined by spatially addressed peptide libraries. Biochemistry 40:1694-1701. [PubMed]
6. Ding, L., G. S. Coombs, L. Strandberg, M. Navre, D. Corey, and E. L. Madison. 1995. Origins of the specificity of tissue-type plasminogen activator. Proc. Natl. Acad. Sci. USA 92:7627-7631. [PMC free article] [PubMed]
7. Ebright, R., Q. Dong, and J. Messing. 1992. Corrected nucleotide sequence of M13mp18 gene III. Gene 114:81-83. [PubMed]
8. Francisco, J. A., C. F. Earhart, and G. Georgiou. 1992. Transport and anchoring of beta-lactamase to the external surface of Escherichia coli. Proc. Natl. Acad. Sci. USA 89:2713-2717. [PMC free article] [PubMed]
9. Georgiou, G., D. Stathopoulos, P. S. Daugherty, A. R. Nayak, B. L. Iverson, and R. Curtiss III. 1997. Display of heterologous proteins on the surface of microorganisms: from the screening of combinatorial libraries to live recombinant vaccines. Nat. Biotechnol. 15:29-34. [PubMed]
10. Grodberg, J., and J. Dunn. 1988. OmpT encodes the Escherichia coli outer membrane protease that cleaves T7 RNA polymerase during purification. J. Bacteriol. 170:1245-1253. [PMC free article] [PubMed]
11. Guina, T., E. C. Yi, H. Wang, M. Hackett, and S. I. Miller. 2000. A PhoP-regulated outer membrane protease of Salmonella enterica serovar Typhimurium promotes resistance to alpha-helical antimicrobial peptides. J. Bacteriol. 182:4077-4086. [PMC free article] [PubMed]
12. Hollifield, W. C., Jr., E. H. Fiss, and J. B. Neilands. 1978. Modification of a ferric enterobactin receptor protein from the outer membrane of Escherichia coli. Biochem. Biophys. Res. Commun. 83:739-746. [PubMed]
13. Ke, S.-H., G. Coombs, K. Tachias, D. Corey, and E. Madison. 1997. Optimal subsite occupancy and design of a selective inhibitor of urokinase. J. Biol. Chem. 272:20456-20462. [PubMed]
14. Kramer, R. A., D. Zandwijken, M. R. Egmond, and N. Dekker. 2000. In vitro folding, purification and characterization of Escherichia coli outer membrane protease OmpT. Eur. J. Biochem. 267:885-893. [PubMed]
15. 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. [PubMed]
16. 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. [PubMed]
17. Leytus, S. P., L. K. Bowles, K. Konisky, and W. F. Mangel. 1981. Activation of plasminogen to plasmin by a protease associated with the outer membrane of Escherichia coli. Proc. Natl. Acad. Sci. USA 78:1485-1489. [PMC free article] [PubMed]
18. Lundrigan, M. D., and R.M. Webb. 1992. Prevalence of ompT among Escherichia coli isolates of human origin. FEMS Microbiol. Lett. 76:51-56. [PubMed]
19. Mangel, W., D. L. Toledo, M. T. Brown, K. Worzalla, M. Lee, and J. J. Dunn. 1994. Omptin: an Escherichia coli outer membrane proteinase that activates plasminogen. Methods Enzymol. 244:384-399. [PubMed]
20. Matsuo, E., G. Sampei, K. Mizobuchi, and K. Ito. 1999. The plasmid F OmpP protease, a homologue of OmpT, as a potential obstacle to E. coli-based protein production. FEBS Lett. 461:6-8. [PubMed]
21. Matthews, D. J., and J. A. Wells. 1993. Substrate phage: selection of protease substrates by monovalent phage display. Science 260:1113-1117. [PubMed]
22. O'Boyle, D. R., K. A. Pokornowski, P. J. McCann, and S. P. Weinheimer. 1997. Identification of a novel peptide substrate of HSV-1 protease using substrate phage display. Virology 236:338-347. [PubMed]
23. 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. [PubMed]
24. 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. [PubMed]
25. Quick, M., and E. M. Wright. 2002. Employing Escherichia coli to functionally express, purify and characterize a human transporter. Proc. Natl. Acad. Sci. USA 99:8597-8601. [PMC free article] [PubMed]
26. Schmidt, M., E. Viaplana, F. Hoffman, S. Marten, A. Vellaverde, and U. Rinas. 1999. Secretion-dependent proteolysis of heterologous protein by recombinant Escherichia coli is connected to an increased activity of the energy-generating dissimilatory pathway. Biotechnol. Bioeng. 66:61-67. [PubMed]
27. Smith, M. M., L. Shi, and M. Navre. 1995. Rapid identification of highly active and selective substrates for stromelysin and matrilysin using bacteriophage peptide display libraries. J. Biol. Chem. 270:6440-6449. [PubMed]
28. Sodeinde, O., Y. Subrahmanyam, K. Stark, T. Quan, Y. Bao, and J. Goguen. 1992. A surface protease and the invasive character of plague. Science 258:1004-1007. [PubMed]
29. Stumpe, S., R. Schmid, D. L. Stephens, G. Georgiou, and E. 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. [PMC free article] [PubMed]
30. Sugimura, K., and N. Higashi. 1988. A novel outer-membrane-associated protease in Escherichia coli. J. Bacteriol. 170:3650-3654. [PMC free article] [PubMed]
31. 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. [PMC free article] [PubMed]
32. Vandeputte-Rutten, L., R. A. Kramer, J. Kroon, N. Dekker, M. R. Egmond, and P. Gross. 2001. Crystal structure of the outer membrane protease OmpT from Escherichia coli suggests a novel catalytic site. EMBO J. 20:5033-5039. [PMC free article] [PubMed]
33. 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. [PubMed]
34. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. Improved M13 phage cloning vectors and host strains: nucleotide sequences of the M13mp18 and pUC19 vectors. Gene 33:103-119. [PubMed]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...


Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...