HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 277, Issue 8, 5796-5803, February 22, 2002

This Article Abstract Full Text (PDF) All Versions of this Article: 277/8/5796    most recent M106849200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sharkov, N. A. Articles by Cai, D. Search for Related Content PubMed PubMed Citation Articles by Sharkov, N. A. Articles by Cai, D. Social Bookmarking                      What's this?

Discovery of Substrate for Type I Signal Peptidase SpsB from Staphylococcus aureus^*

Nikolai A. Sharkov and Danying Cai^§

From the Affymax Research Institute, Palo Alto, California 94304

Received for publication, July 20, 2001, and in revised form, December 3, 2001

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Based on the kinetic model of substrate phage proteolysis, we have formulated a strategy for best manipulating the conditions in screening phage display libraries for protease substrates (Sharkov, N. A., Davis, R. M., Reidhaar-Olson, J. F., Navre, M., and Cai, D. (2001) J. Biol. Chem. 276, 10788-10793). This strategy is exploited in the present study with signal peptidase SpsB from Staphylococcus aureus. We demonstrate that highly active substrate phage clones can be isolated from a phage display library by systematically tuning the selection stringency in screening. Several of the selected clones exhibit superior reactivity over a control, the best clone, SIIIRIII-8, showing >100-fold improvement. Because no conserved sequence features were readily revealed that could allow delineation of the active and unreactive clones, the sequences identified in five of the active clones were tested as synthetic dodecamers, Ac-AGX₈GA-NH₂. Using electrospray ionization mass spectrometry, we show that four of these peptides can be cleaved by SpsB and that Ala is the P1 residue exclusively and Ala or Leu the P3 residue, in keeping with the (3, 1) rule for substrate recognition by signal peptidase. Our successful screening with SpsB demonstrated the general applicability of the screening strategy and allowed us to isolate the first peptide substrates for the enzyme.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Bacterial type I signal peptidase is responsible for cleaving the signal peptide from precursor proteins, and its activity is an integral part of the export and maturation of secreted proteins in vivo. The essential function of the enzyme to bacterial cell viability has been demonstrated using genetic approaches with both Gram-positive and Gram-negative organisms (1-3), supporting the notion that the signal peptidase is potentially an antibacterial target (4). Drug discovery efforts with the enzyme, however, may be hampered by the lack of an effective in vitro assay employing a nonprotein substrate such as a peptide (4).

Our current understanding is that signal peptides are highly variable in sequence (5). Based on the studies carried out over the past 2 decades, it has been established that the recognition sites for signal peptidases lie between 6 and +1 in sequences encompassing the site of cleavage (6-12). Sequence conservation analyses of a large panel of naturally occurring signal peptides in bacteria and eukaryotes reveal that the predominant residue at the P1 site is Ala and that the predominant residues at the P3 site are large aliphatic residues (Leu, Ile, Val) as well as Ala and Ser, a consensus dubbed the (3, 1) rule (9-11). The (3, 1) rule also holds for the cleavage of engineered preproteins in vivo as well as in vitro (6-8, 12). The reaction of signal peptidases with synthetic peptides, on the other hand, is not as well explored as with protein substrates. For the signal peptidase LepB from Escherichia coli, the best characterized signal peptidase, Ala was found as the only residue permitted at the P1 site through single amino acid replacements of a peptide bearing the signal peptide sequence of the E. coli maltose-binding protein (13).

As demonstrated with the E. coli LepB enzyme, the catalytic efficiency of signal peptidase toward short peptide substrates is generally several orders of magnitude lower than toward polypeptides bearing the same protease recognition sequence (14-16). Various approaches including computational designs have been attempted with limited success in search of more highly functional peptides to serve as substrates for the E. coli enzyme (17-20). For instance, peptide libraries were created by incorporating randomized sequences into the signal peptide of TEM-1 -lactamase, varying six amino acid residues between 4 and +2 positions around the signal peptidase cleavage site (19). Functional sequences were found to support the production of active TEM-1 but none better than the wild type. Reported more recently were combinatorial synthetic peptide libraries in which four positions, 4, 3, 2, and +2, were varied in the signal peptidase recognition sequence, and better than 10-fold improvements over the control were observed among the selected peptides (20).

One unsurpassed advantage of phage display over other combinatorial approaches is its capacity to generate a vast number of possible combinations. It is experimentally feasible to randomize up to eight amino acid residues in one library. Phage display has been successfully applied to proteases for selection and optimization of peptide substrates by way of optimizing the substrate phage (21-24). A good peptide substrate in turn would aid the development of protease assays in vitro. Recently, we reported that the proteolysis of substrate phage is a single exponential process and provided the kinetic basis for how to control the rate of proteolysis to ensure the success of substrate phage selection (25). The experimental design strategy we put forward is now exploited in the present study, where we applied it to the screening of an 8-mer phage display library with the type I signal peptidase SpsB from Staphylococcus aureus. By systematically tuning the screening stringency in the selection process, we discovered several active substrate phage clones. The sequences found in the most reactive clones were subsequently evaluated as synthetic peptides and characterized for their competency to serve as substrates for SpsB with respect to the substrate recognition and digestion kinetics.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- The phage display library fTC-LIB-N8 used in this study was reported earlier (24). The recombinant signal peptidase SpsB, expressed in E. coli,¹ was kindly provided by Monica Gevi and Birger Jansson of GlaxoWellcome SpA, Italy. The purified protein was stored in enzyme storage buffer (0.5% Triton X-100, 10% glycerol in phosphate-buffered saline, pH 7.4), and the enzyme concentration was calculated from the protein concentration.¹ The biotinylated mouse monoclonal antibody mAB179 was prepared in house and kindly provided by Bruce Mortenson, and antibody mAB3-E7 was from a previous study (23). Pansorbin cells were purchased from Calbiochem. Phage clones were maintained and amplified using E. coli K91 cells (23). Synthetic peptides of >95% purity were purchased from SynPep Corp. (Dublin, CA) and were acetylated and amidated at the N and C termini, respectively.

Phage Library Screening, Phage Preparation, and Phage ELISA²-- In round one of slicing I (see Table I for designation), 30 µl of the phage library fTC-LIB-N8 at 6.7 × 10¹¹ colony-forming units/ml were mixed with 25 µl of 175 µM signal peptidase SpsB and 195 µl of assay buffer (50 mM Tris-HCl, pH 9.0, 200 mM NaCl, 5 mM MgCl₂ and 0.375% Triton X-100). The negative control was the same except for the replacement of SpsB with 25 µl of enzyme storage buffer. After 24 h of incubation at 25 °C, 30 µl of 1% bovine serum albumin, 33 µl of 3 mg/ml mAb 179, and 10 µl of 1 mg/ml mAb3-E7 were added. After 30 min on ice, 100 µl of Pansorbin cells were added, and the reaction mixture was placed on a rotating mixer at 4 °C for 1 h. The mixture was centrifuged at 4 °C for 2 min, and the supernatant was recovered. A second addition of 100 µl of Pansorbin cells was made, and the Pansorbin cell adsorption step was repeated. The final supernatant was amplified overnight in E. coli K91 cells, and the phage particles were purified (see below). A small aliquot of the final supernatant solution was also used for titering on E. coli K91 cells. Round two of slicing I was then continued for 24 h with 30 µl of the amplified phage from round one, 25 µl of 175 µM SpsB, and 195 µl of assay buffer. The selection round three was continued similarly following the completion of round two. The incubation time of other slicing experiments is indicated in Table I.

Starting at selection round two, individual colonies were picked at random from the Petri dishes from the titering step and cultured in 12-ml LB medium containing 15 µg/ml tetracycline overnight at 37 °C to produce individual phage clones. The phage particles of each clone were precipitated from the supernatant following a general procedure described previously (23). Phage preparations were typically resuspended in 50 µl of TBS (150 mM NaCl and 50 mM Tris, pH 7.5). The phage titer of these preparations was determined by phage ELISA performed following a general procedure described previously (25). These individual clones were then screened for SpsB substrate activity.

Activity Screening and Sequencing of Individual Phage Clones-- For the activity screening of individual clones, each clone was diluted in 40 µl of assay buffer containing 17.5 µM SpsB alongside 40 µl of diluted phage with no enzyme as a control. The dilution was determined according to the titer such that following a further 50× dilution, the starting phage titer would produce an ELISA signal corresponding to the uppermost part of the "useful range" (25). The reaction was incubated at 25 °C, and 5-µl aliquots were typically removed at 10 min, 30 min, 1 h, and 2 h and diluted in 250 µl of 0.1% bovine serum albumin in phosphate-buffered saline to stop the reaction. 200 µl of the stopped reaction were transferred into ELISA plates and developed to check for the extent of cleavage. Clones showing 40% digestion in this time period were considered active, and the most active clones showed significant if not complete digestion within 10 min. Active clones were cultured overnight in 1 liter of LB medium containing 15 µg/ml tetracycline at 37 °C, and the purified phage particles were typically resuspended in 1 ml of phage storage buffer (25 mM Tris, pH 7.4, 20% glycerol, and 75 mM NaCl) and then titered.

The top half of the clones resulting from all slicing experiments were further screened and ranked according to the digestion rate constant k_obs, obtained by following the complete digestion time course in the presence of 15 or 17.5 µM SpsB, because the single-exponential rate constant is a more precise measure of the substrate reactivity. The amount of substrate remaining at various time points was determined using the ELISA assay, and the data were fitted to Equation 1 as described previously using nonlinear regression (25) to determine k_obs,

(Eq. 1)

where [S]₀ represents the starting substrate concentration.

Phage DNA for sequencing was isolated from a 12-ml overnight culture using a commercially available M13 single-stranded DNA isolation kit, the Wizard M13 DNA Purification System, from Promega (Madison, WI). The sequencing primer was 5'-CGATCTAAAGTTTTGTCGTCT-3', which generated the sequence of the complementary antisense strand. The automated sequencing service was provided by Biotech Core (Palo Alto, CA).

Full Kinetic Characterization of Selective Substrate Phage Clones-- For the careful determination of proteolytic activity of the active phage clones, the digestion and ELISA were performed by automated means with a robotic system from Scitec Inc. (Ashland, MA) as previously described (25). The optimal dilution of phage in the digestion reaction was determined individually for each clone. The starting phage concentration was chosen such that following a 33.3-fold dilution in stop solution, it would produce an ELISA signal corresponding to the uppermost part of the "useful range" of the phage standard curve (25). The k_obs was determined for five of the active clones at various SpsB concentrations, and the data were fitted to Equation 2, derived for a pseudo-first-order reaction model (25), to calculate the substrate and enzyme binding constant k₁.

(Eq. 2)

All enzymatic reactions were carried out at 25 °C.

Mass Spectrometry Analysis of the Proteolysis of Synthetic Peptide-- The synthetic peptides were typically dissolved in Me₂SO: peptides 1 and 6 to 10 mM, peptide 2 to 50 mM, peptide 3 to 25 mM, and peptides 4 and 5 to 100 mM (all based on solubility in Me₂SO). To determine the cleavage site, 250 µM peptide was incubated with 17.5 µM SpsB in 200 µl of assay buffer at 25 °C for 16.5 h or otherwise as indicated, all containing 1% Me₂SO in the final digestion reaction except for peptides 1 and 6, which had 2.5% Me₂SO. The digestion reaction (digest) was stopped by adding an equal volume of stop solution (20% acetic acid plus 80% ACN), and the mixture was spun in a microcentrifuge at maximum speed for 10 min. In parallel, two control reactions were prepared similarly, one containing the enzyme only (SpsB only; see Fig. 3) and one the substrate only (peptide only). 5 µl of the supernatant were removed and analyzed using an LC/MS instrument from Agilent (Palo Alto, CA) consisting of a quadrupole ion analyzer (Agilent 1100 MSD), a binary pump, an autosampler, a temperature-controlled column compartment, and a diode array detector. The sample was injected onto a YMC J'sphere column and eluted with an ACN gradient from 10% ACN, 0.2% formic acid to 95% ACN, 0.2% formic acid in 5 min at a flow rate of 1 ml/min. The LC/MS instrument was operated in electrospray positive ionization mode, and the spectra of 200-2000 mass range were recorded. The acquired mass spectra were analyzed using the Agilent ChemStation software to search for the appearance of cleaved products. For each peptide, the expected molecular weight of products from all possible cleavage sites was calculated first, and the ions of the exact molecular weight were extracted from the mass spectra. The products resulting from the actual digestion were those present in the digest spectrum and absent in the SpsB-only and peptide-only control spectra.

Activity Assay of Peptide Substrates-- To quantitate the extent of digestion, 50 µl of the same supernatant used above for the LC/MS experiment were also analyzed using an HPLC instrument by Rainin (Emeryville, CA) equipped with a Dynamax UV detector. The separation was achieved on a Macherey-Nagel ET 250/3 Nucleosil 100-5C18 column using an ACN gradient from 10% ACN, 0.1% trifluoroacetic acid to 100% ACN, 0.1% trifluoroacetic acid in 30 min at a flow rate of 0.5 ml/min. The percentage of digestion was calculated by taking the ratio of the area of the eluted substrate peak from the digest sample and that from the peptide-only control.

The activity of peptides 2 and 4 was further assayed under the steady-state conditions using an end point method. The substrate was diluted in Me₂SO, and the final substrate concentration in the assay was 5-100 and 2-500 µM for peptides 2 and 4, respectively. The Me₂SO concentration was kept constant at 1%. The diluted substrate was incubated with 17.5 µM SpsB at 25 °C for 16 h for peptide 2 and with 1 µM SpsB for 90 or 290 min for peptide 4. Following the addition of an equal volume of stop solution and centrifugation, 50 µl of the supernatant were removed and analyzed using an HPLC instrument. Blanks containing the substrate only were also set up alongside for each test concentration and subjected to the same manipulation and HPLC analysis as follows. The HPLC analysis was performed using an instrument by Rainin as described above or using an Alliance 2690XE HPLC system by Waters (Milford, MA) equipped with a Waters 2487 UV detector. The separation on the Waters instrument was achieved on a Nova-Pak C18 column (3.9 × 150 mm) with an ACN gradient from 0.1% trifluoroacetic acid to 100% ACN, 0.1% trifluoroacetic acid in 12 min at a flow rate of 0.5 ml/min. The area of the eluted substrate peak was used to quantitate the amount of substrate remaining in each sample. The amount of substrate consumed due to digestion was thus calculated from the difference of the amount of substrate remaining in the digestion sample and in its corresponding blank. The enzyme activity was estimated as the amount of substrate consumed over time. The observed substrate consumption was 16-37% and 2-30% for peptides 2 and 4, respectively.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Positive Control Substrate-- Signal peptidase SpsB from S. aureus is known to complement LepB-deficient E. coli mutants (1), indicating that it may recognize and correctly process substrates bearing a cleavage sequence of the E. coli enzyme. A control clone SigPep1 was thus created, containing the cleavage sequence of the E. coli maltose-binding protein (see Table II). Using the phage ELISA that we developed to determine the concentration of intact substrate phage (25), we found that 17.5 µM SpsB was able to cleave SigPep1 at 25 °C with a t_1/2 of about 10 h (Fig. 1). The course of digestion was single-exponential, consistent with our previous observation with other proteases (25). The experiment demonstrated that the activity of SpsB could be detected and assayed properly even for a substrate with a poor reactivity.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Digestion time course of control substrate SigPep1. The amount of substrate remaining was determined at different time points in the presence () and absence () of 17.5 µM SpsB. The data collected in the presence of SpsB were fitted to Equation 1, yielding kobs = (1.17 ± 0.98) × 103 min1, and the line represents the nonlinear curve fit to Equation 1.

Phage Library Screening-- The phage display library fTC-LIB-N8 was screened according to the procedure described under "Experimental Procedures." Five independent screening experiments, slicings I-V, were performed, and the results are summarized in Table I. The screening experiments are arranged in Table I in the order of increased screening stringency, which was controlled by varying the incubation time in each round of the selection. The incubation was kept constant in all selection rounds in every experiment except for slicing III, where it was shortened from 24 to 1 h after the first two rounds. As expected (25), the number of active clones declined as the selection became more stringent. The activity of each clone picked was scored according to the extent of substrate cleavage upon incubation with 17.5 µM SpsB at 25 °C for up to 2 h, and the clones showing 40% cleavage were considered active. According to this criterion, only a small fraction of the total colonies screened was deemed active, and the ratio became even smaller without a 24-h incubation in all rounds of selection, implying that the library contained only a small population of potential substrates for SpsB.

The reactivity of 11 active clones resulting from the screening was analyzed further. We followed their complete digestion time course in the presence of 15 or 17.5 µM SpsB using the phage ELISA and found that the digestion process was single-exponential for all of the clones (data not shown). The first-order rate constants k_obs are reported in Table II.

Pseudo-first-order Kinetics of Substrate Phage Digestion-- The four most active clones SIIIRIII-8, SIRIII-5, SIIBRIII-11, and SIRIII-30 were subjected to full kinetic characterization along with clone SIRIII-35, which was selected because the sequence in its variable region was closely related to that in clone SIIBRIII-11. The proteolysis kinetics were investigated as a function of the SpsB enzyme concentration. Shown in Fig. 2A is an example with SIRIII-5 to illustrate the single-exponential disappearance of the substrate and the dependence of the first-order rate constant on enzyme concentration. The data shown in Fig. 2B indicate a linear function of the enzyme concentration dependence for all five clones, which conforms with our previous conclusion that the reaction between substrate phage and enzyme obeys pseudo-first-order kinetics (25). Only one data point was obtained for the control SigPep1 with 15 µM SpsB, and, for the purpose of comparison, it was included in Fig. 2B. It is clear that the five clones we characterized demonstrate a superior reactivity over the control. Based on the values of k_obs with 15 µM SpsB as reported in Fig. 2B, we found a 54-, 23-, 16-, and 9-fold improvement over SigPep1 for clones SIRIII-5, SIIBRIII-11, SIRIII-30, and SIRIII-35, respectively. SIIIRIII-8 shows the most improvement, estimated at 112-fold over SigPep1 (for SIIIRIII-8 and SIIBRIII-11, k_obs values with 15 µM SpsB were extrapolated from the linear fit in Fig. 2B). The relative reactivity of these clones was also manifested in the slope of the linear curve fit (Table III), i.e. the binding rate constant k₁ between the substrate and enzyme (25).

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effect of SpsB enzyme concentration on substrate phage digestion. A, digestion of substrate phage clone SIRIII-5 in the presence of 2 (), 5 (×), and 15 µM () SpsB. The lines represent the nonlinear curve fit to Equation 1. B, enzyme concentration dependence of five substrate phage clones: SIIIRIII-8 (×), SIRIII-5 (), SIIBRIII-11 (+), SIRIII-30 (), and SIRIII-35 (). The lines represent the linear curve fit of data to Equation 2, and the slopes are shown in Table III. A single data point determined for SigPep1 was also included ().

Sequence Analysis-- The phage DNA was isolated from the control SigPep1 and the 11 active phage clones shown in Table II. The portion of DNA encompassing the variable region was sequenced, and the amino acid sequence was translated from the nucleotide sequence. 10 additional clones were also picked at random, and their phage DNA was purified and sequenced after it was shown that they did not exhibit any appreciable digestion by 17.5 µM SpsB following a 2-h incubation. The sequencing results for both active and unreactive clones are shown in Table II. We found, examining the sequences, that most clones contained at least one Ala residue in the variable region regardless of their reactivity. The occurrence of charged residues and the distribution of other residues seemed indistinguishable between the two groups of clones and might appear rather to reflect the abundance of available codons encoding the residues. For instance, the frequent occurrence of Arg in almost every sequence might result from the fact that it can be encoded by six codons, as can Leu and Ser, and might have no real significance in rendering certain reactivity to the sequence. Therefore, no sequence conservation was evident that would allow for deduction of a consensus sequence or for the sequence of active clones to be aligned in a manner that would be reflective of a plausible cleavage pattern.

Site-specific Cleavage of Synthetic Peptides-- Because the amino acid sequence analysis offered no plausible clues to deciphering conserved features among the active phage clones, we decided to test the sequences found in the five fully characterized clones in synthetic peptides. In designing the peptides, we included the eight residues present in the variable region and two additional residues on each side so that the peptides were 12 residues long (Table II). Three reaction mixtures were set up for each peptide containing 250 µM peptide only (peptide only; see Fig. 3), 17.5 µM SpsB only (SpsB only), or 250 µM peptide and 17.5 µM SpsB (digest). Electrospray ionization mass spectrometry was employed to detect products produced from the proteolysis of peptide by 17.5 µM SpsB following a 16.5-h incubation at 25 °C. Shown in Fig. 3 are the results from an LC/MS experiment in which the digestion of SIRIII-35pep (peptide 5) was analyzed. The incubation resulted in a decrease in the mass signals due to the starting peptide (S) and the concomitant appearance of two new signals (N and C; Fig. 3A), each of which appeared to contain one major parent ion. The molecular weight of the parent ion was 684 and 432 for products N and C, respectively, which matched precisely those of the products resulting from a cleavage between Ala and Val in peptide 5 (Table III). Signal N was the N-terminal portion of the peptide, hence designated product N; signal C was the C-terminal portion, hence product C. The proteolytic products could be identified alternatively as follows. The expected product pairs from the cleavage at each peptide bond were generated first, and their molecular weights were calculated. Following the LC/MS experiment, positive ion species (+H and/or +Na) corresponding to the expected molecular weight were then extracted from total ion chromatograms. The ions absent from the chromatograms of peptide 5-only (Fig. 3, 5 only) and SpsB-only (SpsB only) control samples but unique to that of digest (digest) were considered the products from actual proteolytic cleavage. This approach proved to be very useful, especially when the products co-eluted with certain component(s) in the reaction mixture.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   LC/MS analysis of the proteolysis of peptide 5 (SIRIII-35pep) by SpsB. A, total ion chromatograms of three reaction mixtures containing the enzyme only (SpsB only), substrate (5 only), and both enzyme and substrate (digest). S indicates the retention time at which the substrate was eluted, and N and C represent the retention times at which products N and C were found, respectively. B, positive ion mass spectra of eluents at the indicated retention time. The mass of each ion species is indicated.

Using a combination of the methods described above, the proteolytic products were identified unambiguously for peptide 2 (SIRIII-5pep), peptide 3 (SIIBRIII-11pep), and the control peptide 6 (SigPep1pep). Only one product of m/z 1140 could be positively identified from the digest of peptide 4 (SIRIII-30pep), indicating that peptide 4 was cleaved at the C terminus of the acetylated Ala residue, resulting in an N product too small to be detected using our LC/MS method. No digestion products were detected with peptide 1 (SIIIRIII-8pep) even after prolonged incubations extended over several days, during which the substrate remained intact, as suggested by a constant LC/MS profile (data not shown).

The results of the synthetic peptide proteolysis are summarized in Table III, where the sequences of cleavable peptides are aligned according to the cleavage pattern. Without exception, the cleavage by SpsB occurred at the C terminus of an Ala residue, consistent with the notion that Ala is the preferred P1 residue for signal peptidase. The P3 residue was either an Ala or Leu in peptides 2, 3, and 5, again in keeping with the preference of signal peptidase for an aliphatic residue at P3. Taken together, our results conformed to the (3, 1) rule for substrate recognition by signal peptidases (9-11) except for peptide 4, which lacks residues beyond 1.

Reactivity of Synthetic Peptides-- The same reaction mixtures used in the LC/MS experiments were also analyzed using HPLC. Upon 16-h incubation of 250 µM peptide with 17.5 µM SpsB, we found that peptides 2, 5, and 6 were cleaved 20, 87, and 62%, respectively (Table III). The cleavage of peptide 3 was complete within 6 h, whereas it took only 4 h for peptide 4, and peptide 1 was completely resistant to cleavage. Based on the extent of cleavage, we determined that peptide 4 was the most active, followed by peptides 3, 5, and 2. The activity of peptides 2 and 4 was further assayed under steady-state conditions using the HPLC method described under "Experimental Procedures." We found that for both peptides the activity remained linear to substrate concentration in the range tested (Fig. 4). The highest concentration tested for peptide 4 was 500 µM, indicating a K_m of >500 µM. k_cat/K_m was 0.339 and 23.2 M¹ s¹ for peptides 2 and 4, respectively, calculated from the slope of the linear curve fit to data shown in Fig. 4.

View larger version (22K): [in this window] [in a new window]   Fig. 4.   Activity assay of peptides 2 and 4 under steady-state conditions. The activity of peptides 2 () and 4 () was measured using the end point assay and expressed as the amount of substrate consumed over time. The lines represent the linear curve fit with a zero intercept, and the slopes are shown in Table III. Inset, the same graph shown on an expanded scale.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Having an efficient in vitro assay not only can facilitate the biochemical study of an enzyme but is also imperative to inhibitor screens if the enzyme is a potential drug target. For an enzyme such as a protease, which catalyzes transformations to protein substrates under physiologic conditions, finding a peptide substrate is usually the first step toward achieving this goal, and phage display technology has been instrumental to the discovery of substrates for proteases.

Substrate Phage Reactivity-- We recently described the detailed kinetics for the proteolysis of substrate phage and provided a quantitative basis for the design of a phage library screening experiment (25). We further formulated a screening strategy for meeting the desired outputs through the control of screening stringency. In the present study, we methodically applied the strategy to the signal peptidase SpsB from S. aureus and succeeded in finding reactive substrate phage for the enzyme, validating our screening strategy. In the five independent screening experiments performed, the screening stringency was controlled by varying the length of time the phage library was incubated with the enzyme. The incubation was as short as 30 min to as long as 24 h, but only when it was extended to 24 h did we start to uncover a significant number of active clones (Table I). The overall number of active clones was quite small, but it increased as we relaxed the screening stringency, confirming our prediction that a relaxed screening condition would facilitate the enrichment of reactive substrates (25). We recognize that adopting such a condition is especially important for an enzyme like signal peptidase, which appears to have a low catalytic efficiency. The single-exponential kinetic process, which we described previously for the reaction of stromelysin with its substrate phage (25), was shown here for the signal peptidase. Furthermore, the reaction of the signal peptidase with substrate phage obeyed pseudo-first-order kinetics (Fig. 2), suggesting that the kinetic model developed using stromelysin (25) should be generally applicable to other enzymatic systems.

Although the criterion we used to define the reactivity of a phage clone was somewhat arbitrary (Table I), the fraction of active clones identified out of the total number of colonies screened was apparently small, implying that only a small population of potential substrates for SpsB was present in the phage display library. The activity of the best clones was not especially high (Table II), and it proved difficult to achieve further enrichment probably due to the low abundance of potential substrates in the library. It is conceivable that the segment displayed on phage needs to be over eight residues long for it to be an efficient substrate for SpsB. In such a case, a phage display library of eight randomized residues would not be a rich source of substrate for SpsB.

Peptide Substrate-- While we were successful in finding active substrate phage for SpsB, the sequences of the phage clones did not readily allow the differentiation of the active from the unreactive. We found no plausible hints in the sequences that would allow prediction of the site of cleavage and therefore turned to synthetic peptides. The synthetic peptides were dodecamers, containing eight residues found in the variable region in addition to two flanking residues on each side from the constant region of the library. Using an LC/MS method to identify the proteolytic products, we found that all substrates were cleaved at the peptide bond at the C terminus of an Ala residue, which is consistent with Ala being the preferred P1 residue for signal peptidase, and that the cleavage pattern generally obeyed the (3, 1) rule for substrate recognition by signal peptidase (9-11). It is interesting that peptide 4 is cleaved at an Ala that is common to all phage clones and not in the variable region as seen with other peptides. It is very curious that peptide 1, derived from the most reactive phage clone, is completely resistant to cleavage by SpsB. One Ala residue found in the variable region of peptide 1 is N-terminally linked to a Pro. If peptide 1 were to be cleaved at this Ala by SpsB, it would place the Pro at the 2 position. The fact that we did not detect the digestion of peptide 1 at such a site is consistent with the current understanding that Pro is not allowed in the 3 to +1 region in the substrate for bacterial signal peptidase (5, 11). In light of how peptide 4 (and hence substrate phage clone SIRIII-30) was cleaved, it is conceivable that the digestion of substrate phage clone SIIIRIII-8 could also occur outside of the variable region.

It is clear that the reactivity of the synthetic peptides correlated poorly with that of the corresponding phage (Table III), even if the results with peptide 1 were not taken into account. This finding reinforces our previous argument that such a correlation is not expected, since synthetic peptide and substrate phage are two different substrate entities, subjected to different kinetic rules (25). Substrate phage is selected under single-turnover conditions, and any improvement in its reactivity would strictly affect the rate of substrate and enzyme binding. On the other hand, the activity of the peptide is usually assayed under steady-state conditions, and the substrate binding rate constant may be a small contributing factor to its overall catalytic efficiency.

In general, peptides are known to be inefficient substrates for a signal peptidase such as the E. coli enzyme LepB (14, 15). The catalytic efficiencies of the peptide substrates for SpsB characterized in this study are not fundamentally different from those found for LepB (15). For instance, K_m for the most active peptide 4 (and most likely for peptide 2 as well) is still on the order of millimolar. Previous studies have shown that remarkable improvements in k_cat/K_m can be achieved when the peptide is modified so that it forms micelles, which enhance the enzyme and substrate interaction (26); both k_cat and K_m values were found to be improved to a level approaching that of a preprotein substrate (16). It would be interesting to incorporate the various SpsB substrate sequences from this study into a micelle-forming peptide or a preprotein construct and to test what effect it would have on the peptide reactivity. One interesting observation with peptide 4 is that only one residue, an Ala at P1, is present on the P side of the substrate sequence. It is not known what is the minimal requirement for the sequence on the P' side of the substrate molecule. In a previous attempt to define the minimal substrate, it was found that the E. coli enzyme could cleave peptides with a sequence encompassing 2 and +5 or 7 and +2 residues (15). The crystal structure of -lactam-acylated LepB has been solved (27); however, it is not clear from the structure what kind of interactions would exist in the active site for P' residues on the substrate sequence. Evidently, more experiments are needed to probe the contribution of P' residues to support the activity of a substrate like peptide 4.

Conclusions-- This study shows that through systematic tuning of the selection stringency in screening phage display libraries, the isolation of highly reactive substrate phage clones for an enzyme like signal peptidase, which appears inefficient catalytically toward simple peptide substrates, can be achieved. It demonstrates the general applicability of the kinetic model of substrate phage proteolysis and the phage display library screening strategy that we described earlier (25). These experiments resulted in the isolation of the first peptide substrates for signal peptidase SpsB from S. aureus and provided the first insight into the recognition of substrate cleavage site by the enzyme.

	ACKNOWLEDGEMENTS

We thank Chianne Chen and Jamie Fong for technical assistance; Matt Smith for construction of the phage display library and control clone SigPep1; Nikhil Shah, Mark Gao, and Bill Fitch for assistance with the LC/MS instrumentation; and Xueying Lin and Jeff Northrop for assistance with the automation.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

Present address: Signature BioScience, Inc., 21124 Cabot Blvd., Hayward, CA 94545.

^§ Present address: Telik, Inc., 750 Gateway Blvd., South San Francisco, CA 94080. To whom correspondence should be addressed. Tel.: 650-238-3377; Fax: 650-244-9388; E-mail: dcai@telik.com.

Published, JBC Papers in Press, December 10, 2001, DOI 10.1074/jbc.M106849200

¹ Monica Gevi and Birger Jansson, unpublished results.

	ABBREVIATIONS

The abbreviations used are: ELISA, enzyme-linked immunosorbent assay; ACN, acetonitrile; HPLC, high-pressure liquid chromatography; LC/MS, liquid chromatography/mass spectrometry.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				S. Stephenson, C. Mueller, M. Jiang, and M. Perego Molecular Analysis of Phr Peptide Processing in Bacillus subtilis J. Bacteriol., August 15, 2003; 185(16): 4861 - 4871. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) All Versions of this Article: 277/8/5796    most recent M106849200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Sharkov, N. A. Articles by Cai, D. Search for Related Content PubMed PubMed Citation Articles by Sharkov, N. A. Articles by Cai, D. Social Bookmarking                      What's this?