765 Braz J Med Biol Res 33(7) 2000 Operational substrate for ZapABrazilian Journal of Medical and Biological Research (2000) 33: 765-770 ISSN 0100-879X Development of an operational substrate for ZapA, a metalloprotease secreted by the bacterium Proteus mirabilis 1Departamento de Microbiologia, Instituto de Ciências Biomédicas, Universidade de São Paulo, São Paulo, SP, Brasil 2Departamento de Biofísica e Biologia Molecular, Universidade Federal de São Paulo, São Paulo, SP, Brasil 3Departamento de Biologia, Instituto de Ciências Biológicas, Universidade Estadual de São Paulo, Rio Claro, SP, Brasil 4Departamento de Bioquímica e Biofísica, Instituto Butantan, São Paulo, SP, Brasil B.L. Fernandes1, M.A.F. Anéas1, L. Juliano2, M.S. Palma3, I. Lebrun4 and F.C.V. Portaro4 Abstract The protease ZapA, secreted by Proteus mirabilis, has been consid- ered to be a virulence factor of this opportunistic bacterium. The control of its expression requires the use of an appropriate methodol- ogy, which until now has not been developed. The present study focused on the replacement of azocasein with fluorogenic substrates, and on the definition of enzyme specificity. Eight fluorogenic sub- strates were tested, and the peptide Abz-Ala-Phe-Arg-Ser-Ala-Ala- Gln-EDDnp was found to be the most convenient for use as an operational substrate for ZapA. A single peptide bond (Arg-Ser) was cleaved with a Km of 4.6 µM, a kcat of 1.73 s-1, and a catalytic efficiency of 376 (mM s)-1. Another good substrate for ZapA was peptide 6 (Abz- Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-Gln-EDDnp) which was cleaved at a single bond (Phe-Ser) with a Km of 13.6 µM, a kcat of 3.96 s-1 and a catalytic efficiency of 291 (mM s)-1. The properties of the amino acids flanking the scissile bonds were also evaluated, and no clear requirement for the amino acid residue at P1 was found, although the enzyme seems to have a preference for a hydrophobic residue at P2. Correspondence B.L. Fernandes Departamento de Microbiologia ICB II, USP Av. Prof. Lineu Prestes, 1374 05508-900 São Paulo, SP Brasil Fax: +55-11-818-7240 E-mail: blfernan@usp.br Research supported by FAPESP and PIBIC/CNPq/USP. Received November 4, 1999 Accepted March 10, 2000 Key words · Metalloprotease · Substrate specificity · Quenched fluorescence peptides · Proteus mirabilis Introduction The proteins albumin, azocasein and de- natured hemoglobin were extensively used in the past to detect and partially characterize new proteolytic enzymes (see Ref. 1 for a review). However, difficulties in monitoring enzymatic activity, the low sensitivity of these methods, and the comparison of the data obtained with the properties of other well- characterized enzymes were some of the prob- lems when these proteins were used as sub- strates to define novel proteases. More im- portantly, these substrates were unsuitable for the determination of enzyme specificity, both concerning the most susceptible scissile bond cleaved by the enzyme and the require- ments for specific amino acid side chains in the primary sequence of the substrate. Thus, the development of synthetic peptide sub- strates containing chromogenic or fluoro- genic groups (1,2) greatly accelerated the 766 Braz J Med Biol Res 33(7) 2000 B.L. Fernandes et al. progress of the methodologies used to iso- late and characterize new proteolytic en- zymes. We have been studying the genetic insta- bility of ZapA expression (3), a metallopro- tease which is considered to be a virulence factor of Proteus mirabilis (4). In order to pursue the project concerning the genetic control of ZapA expression by P. mirabilis, an appropriate methodology for enzyme de- tection was required. Surprisingly, azocasein had been the only substrate employed (5). In this study we analyzed several quenched fluorogenic substrates to identify the peptide with the best kinetic properties to be used as an operational substrate in the further char- acterization of this enzyme and its role in P. mirabilis pathogenicity. Material and Methods Bacterial strain and growth condition Escherichia coli strain DH5a [F-, f80dlac ZDM15 D(lacZYA-argF) U169 endA1 recA1 hsdR17 (rK - mK +) deoR thi-1 supE44 l-gyrA 96 relA1] carries plasmid pCW101, which contains the whole zap gene cluster (5). LB medium (1% tryptone, 0.5% yeast extract, 1% NaCl) supplemented with ampicillin (50 µg/ml) was used for E. coli growth. Peptide substrates The intramolecularly quenched fluores- cent peptide substrates were synthesized, purified, and characterized by described pro- cedures (2), using the multiple automated peptide synthesizer PSSM-8 from Shimadzu Corporation (Tokyo, Japan). These peptides contain an Abz-group (ortho-aminobenzoyl) at the N-terminus, and EDDnp (2,4-dinitro- phenyl-ethylenediamine) at the C-terminus, as fluorescent and quencher pair, respec- tively. All peptides used in this study are listed in Table 1. Protease purification by phenyl-Sepharose affinity chromatography The recombinant ZapA protease was pu- rified as described (5), with little modifica- tion. Briefly, the culture supernatant of over- night grown bacteria was centrifuged (7000 g, 30 min, 4oC) and filtered through 0.45-µm pore-size Millipore filters. The filtrates were loaded at a flow rate of 1 ml/min at 4oC onto columns (2 x 60 cm) of phenyl-Sepharose (Pharmacia, Uppsala, Sweden) equilibrated in 50 mM Tris-HCl, pH 8.0. Columns were then washed with 10 volumes of the same buffer. Bound protease was eluted with 50 mM Tris-HCl, pH 11.0, and the collected Table 1 - Scissile bonds and kinetic parameters of fluorogenic peptide substrates (Abz-...-EDDnp) hydrolyzed by ZapA. The intramolecularly quenched fluorogenic peptide substrates were hydrolyzed in 50 mM Tris-HCl, pH 8.0, and 2 mM CaCl2 at 37oC, using 2.8 pM ZapA. Peptide Sequence Km (µM) kcat (s-1) kcat/Km (mM s)-1 P4 P3 P2 P1 P’1 P’2 P’3 P’4 1 A F R ¯ S A A Q 4.6 ± 0.2 1.73 ± 0.09 376 2 A F R ¯ S D A Q 16.5 ± 1.1 0.63 ± 0.03 38 3 A K F R ¯ S A Q 8.5 ± 1.1 0.36 ± 0.03 42 4 A D F R ¯ S A Q 14.8 ± 1.9 0.082 ± 0.007 6 5 W A F R ¯ S A Q 0.37 ± 0.038 0.032 ± 0.001 86 6 R P P G F ¯ S P F R Q 13.6 ± 1.9 3.96 ± 0.40 291 7 G G F ¯ L R R Q 2.3 ± 0.1 0.031 ± 0.001 13 8 E A S Q F E ¯ T S A Q 3.9 ± 0.2 0.020 ± 0.001 5 767 Braz J Med Biol Res 33(7) 2000 Operational substrate for ZapA fractions (3.8 ml) were monitored for pro- tein content with a spectrophotometer at A280 nm, after the pH had been adjusted to pH 8.0. The peak fractions were pooled, and concen- trated by differential centrifugation using Millipore Ultrafree filters (size cut-off, 30 kDa). Material retained by the filter was aliquoted at appropriate concentrations and kept at -20oC in 15% glycerol. The homoge- neity of the preparation was analyzed by SDS-PAGE (6). Enzymatic assays Hydrolysis of the intramolecularly fluo- rogenic quenched peptide substrates at 37oC in 50 mM Tris-HCl, 2 mM CaCl2, pH 8.0, was monitored by measuring fluorescence at lem = 420 nm and lex = 320 nm with a Hitachi F-2000 spectrofluorometer, as pre- viously described (7). Standard hydrolysis conditions were strictly maintained for dif- ferent substrates. The enzyme concentrations varied from 2.8 pM for the best substrates to 8.4 pM for the less susceptible ones. In most cases, substrate concentration ranged from 10 times lower than the Km to 10 times higher than the Km. The kinetic parameters were calculated according to Wilkinson (8). The fluorogenic peptide 1 was used to determine the activities of the recombinant enzyme. One unit of ZapA activity is the amount of enzyme which hydrolyzes 1 µmol of peptide 1 in 1 min. All enzyme assays were per- formed in triplicate. HPLC analysis of synthetic fluorogenic substrates and of their enzymatic hydrolysis products The peptide solutions (20-50 µM) in 50 mM Tris-HCl, pH 8.0, and 2 mM CaCl2 were incubated with the native and the recombi- nant proteases at 37oC. Samples (500 µl) from the substrate and reactions were peri- odically removed for HPLC analysis until 100% hydrolysis was achieved. The hydroly- sis products were separated by HPLC, col- lected manually, and submitted to mass spec- trometry. The scissile bonds were deduced from the sequences of the substrate frag- ments. The HPLC conditions used for the analytical procedure were 0.1% trifluoro- acetic acid in water (solvent A) and acetoni- trile solvent A (9:1) as solvent B. The sepa- rations were performed at a flow rate of 1 ml/ min using a J.T. Baker C-18 column (4.6 x 300 mm). Analytical HPLC was performed using an SPD-10AV Shimadzu UV/visible detector, and an RF-10 AX fluorescence detector. For peptide purification, a semi- prep Ultrapac TSK C-18-120T column (7.8 x 300 mm, 10 µm particles; LKB, Bromma, Sweden) was used with the same solvents as above at a flow rate of 2 ml/min. In all cases, elution was followed by ultraviolet absorp- tion (214 nm) and by fluorescence (lem = 420 nm and lex = 320 nm). Mass spectrometry The mass spectrometry experiments were performed using a QUATTRO II triple qua- drupole mass spectrometer equipped with a standard ES probe (Micromass, Manchester, UK) adjusted to ca. 40 µl/min with an ODS- HG-5 microcolumn (0.3 mm ID x 150 mm, 5 µm particles), using a stream splitter at a ratio of approximately 1:15. The mass spec- trometer data acquisition and treatment sys- tem was equipped with the MassLynx and MaxEnt software for handling spectra. Dur- ing all experiments the source temperature was maintained at 80oC and the needle volt- age at 3.6 kV, applying a drying gas flow (nitrogen) of 200 l/h and a nebulizer gas flow of 20 l/h. The mass spectrometer was cali- brated with intact horse heart myoglobin and its typical cone voltage-induced fragments. For reliable mass determination of the intact ß-chain of insulin and its proteolytic frag- ments, reduction and carboxymethylation were performed as previously described (9). Mass spectrometric detection was achieved 768 Braz J Med Biol Res 33(7) 2000 B.L. Fernandes et al. with different parameters for each type of experiment. The LC-separated peptides were detected by scanning from m/z 50 to 2000 at 6 s/scan, with a 31-V cone. Product ions from MS/MS experiments were detected in several runs by scanning the appropriate mass for each situation, using high energy (25 eV) for single charged precursor ions and low collision energy (15 eV) for multiple charged precursor ions. No tandem MS was recorded for peptides smaller than four amino acid residues. Results and Discussion A small series of fluorescent substrates used to characterize cathepsin B (10) proved to be adequate to determine the specificity of the metalloprotease secreted by Proteus mirabilis, ZapA. Peptide 1 (Abz-A-F-R-S- A-A-Q-EDDnp) showed the best catalytic efficiency (Table 1) and therefore was se- lected for use as an operational peptide sub- strate. It should replace the azocaseine method, introducing a series of obvious ad- vantages in the studies involving this en- zyme. Figure 1 (panel A) shows the Michaelis- Menten and the Lineweaver-Burk plots (1/V x 1/[S] (11) obtained with peptide 1 hydro- lyzed by ZapA. The other fluorogenic pep- tide substrates tested presented the same Michaelis-Menten profiles (data not shown). Panel B illustrates the HPLC elution profile of the intact peptide substrate 1, detected by fluorescence and absorbance at 214 nm. Panel C shows the HPLC elution profile of the two fragments generated by the hydrolytic activ- ity of the ZapA metalloprotease on peptide 1, detected by absorbance at 214 nm. Only the peak corresponding to the N-terminal fragment showed fluorescence, while the other fragment was only detected by UV absorbance. The peptide bond cleaved was Arg-Ser (see Table 1), as determined by mass spectrometry. In the series of related substrates (pep- Figure 1 - Hydrolysis of Abz-A-F-R-S-A-A-Q-EDDnp (peptide 1) by ZapA. Panel A, Michae- lis-Menten profile obtained with 0.5-8.0 µM of peptide 1, hydrolyzed by ZapA (2.8 pM) and monitored by fluorescence. The insert is the 1/V x 1/[S] plot. Panel B shows the HPLC profile of the intact quenched fluorogenic peptide 1, and panel C is the HPLC elution profile of the two fragments generated by the incubation of peptide 1 with 2.8 pM purified native ZapA in 50 mM Tris-HCl and 2 mM CaCl2, pH 8.0, at 37oC for 3 h. The HPLC analysis conditions were as described in Material and Methods. Dashed lines indicate fluorescence and continuous lines show the absorbance at A214nm. 800 600 400 0 200 V el oc ity (F U /m in ) 400 200 0 400 200 0 0.012 0.008 0.0041/ V (m in /F U ) 0.8 1.2 1.6 2.00.4 1/[S] (1 µM) 8 12 16 204 [S] (µM) 0 Abz-A-F-R-S-A-A-Q-EDDnp 10 15 2050 Abz-A-F-R- S-A-A-Q-EDDnp 10 15 2050 Time (min) A B C 1000 FU ( ) FU ( ) 500 1000 500 A bs 2 14 x 1 0- 3 ( ) A bs 2 14 x 1 0- 3 ( ) 769 Braz J Med Biol Res 33(7) 2000 Operational substrate for ZapA tides 1-5, Table 1) we found that, on the P� site (according to the nomenclature of Schechter and Berger, Ref. 12), the replace- ment of one alanine with aspartic acid (pep- tide 2), or the removal of one alanine (pep- tide 3) reduced the catalytic efficiency of ZapA by one order of magnitude, and the replacement of lysine at position P3 with aspartic acid further reduced the catalytic efficiency by one order of magnitude (pep- tide 4). In contrast, the introduction of a bulky hydrophobic residue, tryptophan, at position P4, significantly improved the bind- ing, resulting in the lowest Km value in the series of peptides used in this study (peptide 5). For this whole series of substrates, a single bond, only the Arg-Ser bond, was cleaved. For the other quenched fluorescent sub- strates used, displaying rather different amino acid sequences, we found that the bradyki- nin-derived peptide (peptide 6) represented the best substrate, displaying a kcat/Km value only about 23% lower than that found for peptide 1. The only peptide bond cleaved was the Phe-Ser bond. Another quenched fluorescent substrate, which has a hydropho- bic residue at position P1, was the peptide whose sequence was derived from dynorphin (peptide 7), and was hydrolyzed at the Leu- Arg bond. Finally, in peptide 8, whose amino acid sequence was derived from synaptobrevin (the natural substrate for the tetanus toxin metalloprotease (13)), the only peptide bond cleaved was the Glu-Thr bond, this turning out to be the worst substrate in the series of peptides used in this study. The limited number of substrates studied does not permit a more detailed description of the enzyme specificity. However, it seems likely that a hydrophobic residue in P2 favors the catalytic efficiency of ZapA. The excep- tion is peptide 6, which has Gly at position P2. This substrate has a good catalytic effi- ciency (kcat/Km = 291 (mM s)-1), but mostly due to the high kcat value. Thus, it seems likely that a hydrophobic residue in P2 is important to increase the affinity of the sub- strate for the enzyme. These findings agree with the substrate requirements of other bac- terial metalloproteases, which also belong to the serralysin family of bacterial alkaline proteases (14). The S1 site of ZapA showed broad speci- ficity, efficiently accommodating substrates presenting an arginine, a phenylalanine, a leucine or a glutamic acid residue in P1 (see Table 1, peptides 1-8). The latter peptide offered an additional reason for hydrolysis at the Leu-Arg bond, which is the presence of Phe at position P2. In contrast, the results obtained with peptide 8 (Table 1) yielded the lowest value for kcat, indicating that an acidic residue at position P1 does not favor hy- drolysis, even though a hydrophobic residue is located at position P2. Investigations of the natural substrates of ZapA as well as the genetic control of its expression will be greatly facilitated by the method described here. References 1. Barrett AJ (1977). Proteinases in Mam- malian Cells and Tissues. Elsevier/North- Holland Biomedical Press, Amsterdam, The Netherlands. 2. Hirata IY, Cezari MHS, Nakaie CR, Boschcov P, Ito AS, Juliano MA & Juliano L (1994). Internally quenched fluorogenic protease substrates: solid-phase synthe- sis and fluorescence spectroscopy of pep- tides containing ortho-aminobenzoyl/dini- trophenyl groups as donor-acceptor pairs. Letters in Peptide Science, 1: 299-308. 3. Costa SOP (1990). A review of protease secretion instability in Proteus mirabilis. Brazilian Journal of Genetics, 13: 165-172. 4. Rozalski A, Sidorczyk Z & Kotelko K (1997). Potential virulence factors of Proteus ba- cilli. Microbiology and Molecular Biology Reviews, 61: 65-89. 5. Wassif C, Cheek D & Belas R (1995). Mo- lecular analysis of a metalloprotease from Proteus mirabilis. Journal of Bacteriology, 177: 5790-5798. 6. Laemmli UK (1970). Cleavage of struc- tural proteins during the assembly of the head of bacteriophage T4. Nature, 227: 680-685. 7. Chagas JR, Juliano L & Prado ES (1991). Intramolecularly quenched fluorogenic tetrapeptide substrates for tissue and 770 Braz J Med Biol Res 33(7) 2000 B.L. Fernandes et al. plasma kallikreins. Analytical Biochemis- try, 192: 419-425. 8. Wilkinson GN (1961). Statistical estima- tions in enzyme kinetics. Biochemical Journal, 80: 324-332. 9. Van Baar BLM, Hulst AG & Wils ER (1999). Characterization of cholera toxin by liquid chromatography - electrospray mass spectrometry. Toxicon, 37: 85-98. 10. Portaro FCV, Santos ABF, Juliano MA, Cezari MHA, Juliano L & Carmona E (2000). Probing the specificity of cysteine proteinases at subsites remote from the active site: analysis of P4, P3, P2’ and P3’ variations in extended substrates. Bio- chemical Journal, 347: 123-129. 11. Segel IM (1975). Enzyme Kinetics. John Wiley and Sons, New York. 12. Schechter I & Berger A (1967). On size of the active site in proteases. Biochemical and Biophysical Research Communica- tions, 27: 157-162. 13. Yamazaki S, Baumeister A, Binz T, Blasi J, Link E, Cornille F, Roques B, Fykse EM, Sudhof TC, Jahn R & Niemann H (1994). Cleavage of members of the synaptobre- vin VAMP family by type E and F botulinical neurotoxin and tetanus toxin. Journal of Biological Chemistry, 269: 12764-12772. 14. Maeda H & Morihara K (1995). Serralysin and related bacterial proteinases. Meth- ods in Enzymology, 248: 395-413.