A quenched fluorescent dipeptide for assaying dispase- and thermolysin-like proteases

Abstract Metalloproteases such as dispase and thermolysin play a crucial role in the life cycle of bacteria. Commonly, they prefer hydrophobic amino acids at P1� of substrate proteins, thereby cleaving the peptide bond at the alpha amino group. Activity of such proteases has been measured by the use of tailor-made oligopeptides provided with Xuorescence resonance energy transfer dyes. We can now show that the short dipeptide Dabcyl-Ser-Phe-EDANS is an appropriate substrate of dispase and thermolysin. It was cleaved by both enzymes at the single peptide bond accompanied by a steep increase in Xuorescence. Substantial quenching eVects of the formed products were observed only when more than 80�M substrate was hydrolyzed. High aYnity of the proteases for the dipeptide resulted in low Km values of 91 § 9 and 104 § 18 �M, which are comparable to those measured for longer peptides. Dabcyl-Ser-Phe-EDANS was also used to determine the pH and optimal temperature of dispase, which were found at pH 7.0 and 50 °C. BuVer substances such as acetate, citrate, and tris(hydroxymethyl)aminomethane had no signiWcant eVect on enzyme activity. Measurements up to 100 °C revealed that hydrolysis of the quenched Xuorescent dipeptide took place only in the presence of active dispase. © 2006 Elsevier Inc. All rights reserved. Keywords: P1� metalloproteases; Quenched Xuorescent dipeptide; Dabcyl-Ser-Phe-EDANS; P1� endoprotease assay P1� proteases such as dispase from Bacillus polymyxa are hydrolases cleaving peptide bonds at the N side of spe- ciWc amino acids [1]. Until recently, continuous assays for P1� proteases could not be readily established because a chromophore at the alpha amino group of a synthetic peptide hardly changes the absorption by proteolytic release. Only short furylacryloyl peptides, such as N-(3- furylacryloyl)glycyl-L-leucylamide and N-(3-furylacry- loyl)glycyl-L-phenylalanylamide, were known to decrease UV absorption after cleavage of the sole peptide bond [2]. However, the small eVect induced by P1� proteases limits a broader application of the compounds. During the past resonance energy transfer (FRET)1 assays for the deter- mination of speciWc proteases. The combination of the Xuorescent donor N-(2-aminoethyl)aminonaphthalene-5- sulfonic acid (EDANS) and the quenching acceptor 4-(4�- N,N-dimethylaminophenyl)azobenzoic acid (Dabcyl) is one of the most frequently used FRET systems due to a well-suited overlap of the donor–acceptor wavelengths and a large increase in Xuorescence when Dabcyl– EDANS peptides are cleaved proteolytically [3]. 1 Abbreviations used: FRET, Xuorescence resonance energy transfer; Analytical Biochemistry 35 A quenched Xuorescent dispase- and thermo Stefanie Weimer a, Kai Oertel b a Department of Chemical Engineering and Biotechnology, Universit b N-Zyme BioTec GmbH, D- Received 21 N Available online 0003-2697/$ - see front matter © 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.ab.2006.02.029 decade, many synthetic peptides have been coupled with Xuorescent dyes to develop highly sensitive Xuorescence * Corresponding author. Fax: +49 6151 168404. E-mail address: [email protected] (H.-L. Fuchsbauer). ANALYTICAL BIOCHEMISTRY 2 (2006) 110–119 www.elsevier.com/locate/yabio dipeptide for assaying lysin-like proteases , Hans-Lothar Fuchsbauer a,¤ y of Applied Sciences of Darmstadt, D-64287 Darmstadt, Germany 64293 Darmstadt, Germany ovember 2005 15 March 2006 EDANS, N-(2-aminoethyl)aminonaphthalene-5-sulfonic acid; Dabcyl, 4- (4�-N,N-dimethylaminophenyl)azobenzoic acid; Boc, tert-butoxycarbonyl; OSu, oxysuccinimide; FT, Fourier transform; ESI, electrospray ionization; OPFP, oxypentaXuorophenylester; DMF, N,N-dimethylformamide; DCC, N,N�-dicyclohexylcarbodiimide; EDTA, ethylenediaminetetraacetic acid; AEBSF, 4-(2-aminoethyl)benzenesulfonylXuoride; rfu, relative Xuores- cence units; TLC, thin-layer chromatography. Assaying dispase- and thermolysin-like proteases / S. W Quenched Xuorogenic peptides were constructed to deter- mine matrix metalloproteases [4,5], the TNF�-cleaving enzyme [6], �-secretase [7], caspase [8], cathepsin D [9], virus proteases [10–12], and several other proteases such as renin [13], papain [14], and trypsin [15]. In all cases, extended peptides, exhibiting more than 6 (typically 10– 12) amino acids, were synthesized to address the speciWc binding proWle of a protease. Spacers such as �-aminobu- tyric acid were inserted as well as a precaution to prevent a potential steric hindrance of substrate binding by the bulky dyes [3]. For standard P1� proteases, it would be desirable to have larger amounts of a quenched Xuores- cent peptide and to reduce synthesis costs. This implicates the preparation of shorter peptides, at most without any spacer. However, no attempt has been made to comply with such demands. We have now synthesized without any spacer a Dab- cyl–EDANS dipeptide exhibiting only the two amino acids serine and phenylalanine. The Ser-Phe motif has been chosen because it is the preferred binding site of the P1� proteases TAMEP and dispase during activation of bacterial transglutaminase [16–18]. The quenched Xuores- cent dipeptide was proved to be an appropriate substrate of the neutral metalloproteases dispase and thermolysin. Moreover, moderate activity of proteinase K revealed that the novel compound may also be truncated at the carboxyl side of phenylalanine, thereby releasing EDANS. Materials and methods Materials EDANS, N-tert-butoxycarbonyl-L-phenylalanine (Boc- Phe), N-(4-[4�-N�,N�-(dimethylamino)phenylazo]benzoyl- oxy)succinimide (Dabcyl-OSu), thermolysin from Bacillus thermoproteolyticus rokko, collagenase from Clostridium hist- olyticum, proteinase K from Tritirachium album, subtilisin from Bacillus globigii, bovine chymotrypsin, papain from Papaya carica, bromelain from pineapple stem, and pepsin from porcine gastric mucosa were obtained from Sigma– Aldrich (St. Louis, MO, USA). N-tert-butoxycarbonyl-O- tert-butyl-L-serine (Boc-Ser(tBu)-OH) was obtained from Bachem (Bubendorf, Switzerland). Dispase was obtained from Roche Diagnostics (Mannheim, Germany). All other chemicals were purchased in analytical grade or for synthesis by Merck or AppliChem (Darmstadt, Germany). Apparatus NMR and MS were carried out on a Bruker AC 300 pulsed Fourier transform (FT) NMR and a Bruker Esquire-LC (electrospray ionization (ESI)–MS) by the Technical University of Darmstadt. Absorption and emission spectra of Dabcyl-Ser-Phe- EDANS were measured using a PerkinElmer Lambda 2 UV/Vis spectrometer and a Hitachi F4000 Xuorometer. eimer et al. / Anal. Biochem. 352 (2006) 110–119 111 Peptide synthesis N-tert-Butoxycarbonyl-L-phenylalanyl pentaXuorophenylester (Boc-Phe-OPFP) A mixture of N-tert-butoxycarbonyl-L-phenylalanine (400 mg, 1.50 mmol) and pentaXuorophenol (331 mg, 1.80 mmol) in 3.4 ml dioxane and 1.1 ml N,N-dimethylform- amide (DMF) was cooled to ¡7 °C. Within 30 min, N,N�- dicyclohexylcarbodiimide (DCC, 340 mg, 1.65 mmol) was added to the reaction mixture. The stirred solution was allowed to warm to room temperature, and the reaction was completed by stirring overnight. The precipitated white solid was separated by Wltration, and the Wltrate was evapo- rated to dryness in vacuo. The residue was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, resulting in 700 mg (90%) of white solid. TLC Rf D 0.93 (silica, 65:25:4 CHCl3:MeOH:H2O). 2-N-(N�-tert-Butoxycarbonyl-L-phenylalanylamido)-N�- ethylaminonaphthalene-5-sulfonic acid (Boc-Phe-EDANS) N-(2-Aminoethyl)aminonaphthalene-5-sulfonic acid (128 mg, 0.500 mmol) and 0.5 ml of 1 M NaOH in ethanol were evaporated to dryness in vacuo. The residue and 69.4�l (0.500 mmol) triethylamine were dissolved in 105 ml DMF. Boc-Phe-OPFP (225 mg, 0.500 mmol) was added to the stirred mixture in small portions at room temperature, and the reaction was completed by stirring overnight. The reaction mixture was concentrated to approximately 3 ml in vacuo and was separated by silica 60 chromatography using CHCl3/MeOH/H2O (70:25:4). After removal of sol- vent from combined product fractions by evaporation in vacuo, residue was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, resulting in 233 mg (87%) of a brownish-yellow Xuorescent solid. TLC Rf D 0.76 (silica, 70:25:4 CHCl3:MeOH:H2O). 1H NMR (400 MHz, [D6]-DMSO): � [ppm] D 1.30 (s, 9 H, t- Bu), 2.75–2.79 (m, 1 H, �-phenylalanyl-CH2), 2.95–3.00 (m, 1 H, �-phenylalanyl-CH2), 3.20 (t, 2 H, EDANS-CH2), 3.29–3.67 (m, 2 H, EDANS-CH2, covered by HOD), 4.13– 4.23 (m, 1 H, �-phenylalanyl-CH), 6.58 (d, 1 H, naphthyl-H, J D 7.6 Hz), 6.93 (d, 1 H, NH), 7.12–7.18 (m, 1 H, phenyl-H), 7.19–7.25 (m, 4 H, phenyl-H), 7.28 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 8.4 Hz), 7.33 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 8.4 Hz), 7.94 (d/d, 1 H, naphthyl-H, J D 7.0 Hz, 1.0 Hz), 7.95 (t, 1 H, NH), 8.08 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.16 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.20 (t, 1 H, NH). MS (LC- spray): m/z D 512 (100%, M¡H), 456 (11%, M¡H¡i- butene), 438 (9%, M¡2 H¡OtBu), 412 (80%, M¡H¡Boc), 183 (10%, N-(2-aminoethyl)naphthylamine¡3 H). 2-N-(L-Phenylalanylamido)-N�-ethylaminonaphthalene-5- sulfonic acid (Phe-EDANS) Boc-Phe-EDANS (53.6 mg, 0.100 mmol) was dissolved in 0.5 ml dichloromethane and 0.5 ml triXuoroacetic acid. The reaction mixture was stirred for 1 h and evaporated to dry- ness in vacuo. The residue was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, 112 Assaying dispase- and thermolysin-like proteases / S. W resulting in 40 mg (92%) of a light brown Xuorescent solid. TLC Rf D 0.34 (silica, 65:25:4 CHCl3:MeOH:H2O). 1H NMR (400 MHz, [D6]-DMSO): � [ppm] D 2.99–3.08 (m, 1 H, �-phenylalanyl-CH2), 3.10–3.22 (m, 1 H, �-phenylalanyl- CH2), 3.28–3.36 (m, 2 H, EDANS-CH2), 3.43–3.51 (m, 2 H, EDANS-CH2), 3.60–4.10 (m, 1 H, �-phenylalanyl-CH cov- ered by HOD), 6.55 (d, 1 H, naphthyl-H, J D 7.6 Hz), 7.19– 7.26 (m, 6 H, NH, phenyl-H), 7.28 (t, 1 H, naphthyl-H, J D 8.0 Hz), 7.32 (t, 1 H, naphthyl-H, J D 7.7 Hz), 7.95 (d, 1 H, naphthyl-H, J D 7.0 Hz), 8.07 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.15 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.26 (broad s, 2 H, NH2), 8.60 (t, 1 H, NH). MS (LC-spray): m/z D 436 (72%, M+Na), 414 (100%, M+H), 396 (4%, M¡NH2¡H). N-tert-Butoxycarbonyl-O-tert-butyl-L-serine pentaXuorophenylester (Boc-Ser(tBu)-OPFP) A mixture of 391 mg (1.50 mmol) N-tert-butoxycar- bonyl-O-tert-butyl-L-serine and 332 mg (1.80 mmol) penta- Xuorophenol in 3 ml dioxane and 6 ml DMF was cooled to ¡7 °C. Within 30 min, 340 mg (1.60 mmol) DCC was added in portions to the reaction mixture. The stirred solution was allowed to warm to room temperature, and the reaction was completed by stirring overnight. The precipitated white solid was separated by Wltration, and the remaining solu- tion was evaporated to dryness in vacuo. The residue was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, resulting in 550 mg (86%) of a white solid. TLC Rf D 0.92 (silica, 70:25:4 CHCl3:MeOH:H2O). 2-N-(N�-tert-Butoxycarbonyl-O-tert-butyl-L-serinyl-L- phenylalanylamido)-N�-ethylaminonaphthalene-5-sulfonic acid (Boc-Ser(tBu)-Phe-EDANS) Phe-EDANS (326 mg, 0.750 mmol) and 209�l (1.50 mmol) triethylamine were dissolved in 7 ml DMF. Boc-Ser(tBu)-OPFP (321 mg, 0.750 mmol) was added in small portions to the stirred mixture at room temperature, and the reaction was completed by stirring overnight. After concentration to approximately 3 ml, the residue was sepa- rated by silica 60 chromatography using CHCl3/MeOH/ H2O (70:25:4). Product fractions displayed by a single TLC spot with Rf of 0.47 were combined and evaporated to dry- ness in vacuo. The residue was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, resulting in 432 mg (84%) of a brownish-yellow Xuorescent solid. TLC Rf D 0.47 (silica, 70:25:4 CHCl3:MeOH:H2O). 1H NMR (400 MHz, [D6]-DMSO): � [ppm] D 1.08 (s, 9 H, Ot-Bu), 1.39 (s, 9 H, Boc-tBu), 2.82–2.92 (m, 1 H, �-phenyl- alanyl-CH2), 2.93–3.04 (m, 1 H, �-phenylalanyl-CH2), 3.13–3.18 (m, 2 H, EDANS-CH2), 3.20–3.60 (m, 4 H, �-seri- nyl-CH2, EDANS-CH2, covered by HOD), 3.89–4.03 (m, 1 H, �-serinyl-CH), 4.48–4.59 (m, 1 H, �-phenylalanyl-CH), 6.03 (t, 1 H, NH), 6.54 (d, 1 H, naphthyl-H, J D 7.6 Hz), 6.65 (d, 1 H, NH), 7.10–7.16 (m, H, phenyl-H), 7.16–7.19 (m, 4 H, phenyl-H), 7.26 (t, 1 H, naphthyl-H, J D 8.0 Hz), 7.32 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 8.0 Hz), 7.91 (d, 1 H, NH), 7.93 eimer et al. / Anal. Biochem. 352 (2006) 110–119 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 1 Hz), 8.06 (d, 1 H, naph- thyl-H, J D 8.4 Hz), 8.13 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.19 (t, 1 H, NH). MS (LC-spray): m/z D 655 (100%, M¡H), 555 (6%, M¡Boc), 412 (9%, M¡Boc-Ser), 183 (11%, N-(2- aminoethyl)naphthylamine¡3 H). 2-N-(L-Serinyl-L-phenylalanylamido)-N�- ethylaminonaphthalene-5-sulfonic acid (Ser-Phe-EDANS) Boc-Ser(tBu)-Phe-EDANS (67.8 mg, 0.100 mmol) was dissolved in 0.6 ml dichloromethane and 0.6 ml triXuoroace- tic acid. The reaction mixture was stirred for 1 h at room temperature and evaporated to dryness in vacuo. The resi- due was digerated with diethyl ether, Wltered, washed with diethyl ether, and dried over P4O10, resulting in 43 mg (82%) of a light brown Xuorescent solid. TLC Rf D 0.22 (silica, 65:25:4 CHCl3:MeOH:H2O). 1H NMR (400 MHz, [D6]- DMSO): � [ppm] D 2.84–2.89 (m, 1 H, �-phenylalanyl-CH2), 3.01–3.05 (m, 1 H, �-phenylalanyl-CH2), 3.15–3.20 (m, 2 H, EDANS-CH2), 3.30–3.36 (m, 1 H, �-serinyl-CH2), 3.41–3.48 (m, 1 H, �-serinyl-CH2), 3.60–4.10 (HOD, EDANS-CH2 visible at 3.65, �-serinyl-CH visible at 3.97–4.03), 4.52–4.58 (m, 1 H, �-phenylalanyl-CH), 6.55 (d, 1 H, naphthyl-H, J D 7.6 Hz), 7.14–7.19 (m, H, phenyl-H), 7.20–7.25 (m, 4 H, phenyl-H), 7.28 (t, 1 H, naphthyl-H, J D 8.0 Hz), 7.33 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 7.8 Hz), 7.95 (d, 1 H, naphthyl- H, J D 8.0 Hz), 8.05 (d, 1 H, naphthyl-H), 8.07 (t, 1 H, NH), 8.14 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.32 (t, 1 H, NH), 8.71 (d, 1 H, NH). MS (LC-spray): m/z D 523 (100%, M+H+Na), 501 (44%, M+H), 472 (8%, M¡CO). 2-N-(4-[4�-N�,N�-(Dimethylamino)phenylazo]benzoyl-L- serinyl-L-phenylalanylamido)-N�-ethylaminonaphthalene-5- sulfonic acid (Dabcyl-Ser-Phe-EDANS)2 N-(4-[4�-N�,N�-(Dimethylamino)phenylazo]benzoyl- oxy)succinimide (16.5 mg, 4.00�mol) was added in small portions to a stirred solution of 20.9 mg (4.00�mol) Ser- Phe-EDANS and 11.2�l (8.00�mol) triethylamine in 4 ml DMF. After the reaction was completed by stirring over- night, the mixture was concentrated in vacuo to approxi- mately 1 ml and separated by silica 60 chromatography using CHCl3/MeOH/H2O (80:18:2). The combined product fractions were evaporated to dryness in vacuo and dried over P4O10, resulting in 25 mg (80%) of a brick red solid. TLC Rf D 0.5 (silica, 70:25:4 CHCl3:MeOH:H2O). 1H NMR (400 MHz, [D6]-DMSO): � [ppm] D 2.86–2.91 (m, 1 H, �- phenylalanyl-CH2), 3.04–3.09 (m, 1 H, �-phenylalanyl- CH2), 3.08 (s, 6 H, 2 CH3), 3.17–3.22 (m, 3 H, EDANS-CH2, �-serinyl-CH2), 3.67–3.74 (m, 2 H, EDANS-CH2), 3.25–3.60 (HOD covering �-serinyl-CH2), 4.15 (broad s, 1 H, OH), 4.49–4.56 (m, 2 H, �-phenylalanyl-CH, �-serinyl-CH), 5.15 (t, 1 H, NH), 6.04 (t, 1 H, NH), 6.55 (d, 1 H, naphthyl-H, J D 7.6 Hz), 6.85 (d, 2 H, phenylazo-H, J D 9.6 Hz), 7.10–7.23 2 The compound is available from N-Zyme BioTec (Darmstadt, Germany). Assaying dispase- and thermolysin-like proteases / S. W (m, 5 H, phenyl-H), 7.27 (t, 1 H, naphthyl-H, J D 8.0 Hz), 7.32 (d/d, 1 H, naphthyl-H, J D 7.6 Hz, 7.8 Hz), 7.84 (d, 2 H, phenylazo-H, J D 9.6 Hz), 7.85 (d, 2 H, azobenzoyl-H, J D 9.6 Hz), 7.93 (d, 1 H, naphthyl-H, J D 7.6 Hz), 8.05 (d, 1 H, naphthyl-H, J D 8.0 Hz), 8.07 (d, 2 H, azobenzoyl-H), 8.13 (d, 1 H, naphthyl-H, J D 8.4 Hz), 8.17 (d, 1 H, NH), 8.47 (d, 1 H, NH). MS (LC-spray): m/z D 772 (2%, M¡H+Na), 750 (100%, M¡H), 603 (3%, M¡C8H10N3). Fluorometric assay For the used proteases and Dabcyl-Ser-Phe-EDANS, stock solutions of 1 mg/ml were made up in 100 mM Tris– HCl (pH 7.5) containing 2 mM CaCl2 (dispase, thermoly- sin, collagenase, proteinase K, subtilisin, and chymotryp- sin), 100 mM citrate (pH 6.5, papain and bromelain), or 10 mM HCl (pepsin). Stock solutions of papain and bro- melain had concentrations of 4 mg/ml. In typical experi- ments, 15 �l (20.0 nmol) of the quenched dipeptide was incubated in 418–470 �l of 100 mM Tris–HCl/2 mM CaCl2 and 100 mM citrate (pH 6.5) or 10 mM HCl at 37 °C for 10 min before hydrolysis was initiated by adding 15–67 �l of the protease stock solutions (Wnal concentrations of 1.2 �M for dispase, thermolysin, collagenase, proteinase K, subtilisin, chymotrypsin, and pepsin and 6.0 �M for papain and bromelain). The reaction was continued for an additional 10 min and stopped by 500 �l of 20 mM ethyl- enediaminetetraacetic acid (EDTA) (metalloproteases), 2 mM 4-(2-aminoethyl)benzenesulfonylXuoride (AEBSF) (serine and cysteine proteases), or 100 mM Tris–HCl (pH 7.5) (pepsin). Blanks were obtained by adding the inhibi- tor prior to protease. Emission (�ex 336 nm, �em 490 nm [3]) was monitored immediately after the addition of the inhibitory solution using 1-cm cuvettes. All data given in the Wgures represent the means of at least three measurements. Final point measurements were performed at various substrate concentrations to correlate the turnover rate with the increase of emission intensity, where 1 nmole of cleaved Dabcyl-Ser-Phe-EDANS enhances Xuorescence intensity by 3.7 relative Xuorescence units (rfu). Kinetic parameters of dispase and thermolysin were determined using double-reciprocal Lineweaver–Burk plots. Results and discussion Synthesis of Dabcyl-Ser-Phe-EDANS The quenched Xuorogenic dipeptide was synthesized using the well-established active ester chemistry combined with Boc protection techniques (Fig. 1). In brief, all reac- tion steps ran with yields of more than 80%. 1H NMR spec- tra and the obtained data from MS corresponded with the expected molecular structures. The established procedure enables preparation of 25–100 mg of the novel substance within a few days. eimer et al. / Anal. Biochem. 352 (2006) 110–119 113 Hydrolysis of Dabcyl-Ser-Phe-EDANS by various proteases Because it was unclear whether the dipeptide was too short to Wt into the active core of a protease or whether the directly coupled dyes could prevent an attachment, several proteases were chosen to study primarily the proteolytic cleavage of Dabcyl-Ser-Phe-EDANS, among them metallo- proteases with P1� speciWcity as well as serine, cysteine, and aspartyl proteases commonly exhibiting P1 speciWcity. As can be seen in Fig. 2, the P1� proteases dispase from B. poly- Fig. 1. (A) Scheme of Dabcyl-Ser-Phe-EDANS synthesis. PFP, pentaXuor- ophenylester; TFA, triXuoroacetic acid. (B) Cleavage sites for proteases. BOC-Phe PFP DCC (1) DCH BOC-Phe-OPFP PFP EDANS BOC-Phe-EDANS 90 % 87 % 92 % TFA i-butene CO2 Phe-EDANS Ser-Phe-EDANS 86 % BOC-Ser(tBu)-Phe-EDANS BOC-Ser(tBu)-OPFP DCH (2) PFP DCC BOC-Ser(tBu) Ser-Phe-EDANS Dabcyl-OSu (3) SuOH Dabcyl-Ser-Phe-EDANS 80 % 84 % Phe-EDANS PFPi-butene CO2 TFA 82 % N NO OH O N O N N SO3H N N H3C CH3 P1-proteases H H H H P1'-proteases A B myxa and thermolysin from B. thermoproteolyticus rokko displayed by far the highest proteolytic activity. Nearly the 114 Assaying dispase- and thermolysin-like proteases / S. W entire substrate was hydrolyzed within 10 min. The third used metalloprotease, collagenase from C. histolyticum, that usually opens Leu-Gly and Gly-Gly peptide bonds could also cleave Dabcyl-Ser-Phe-EDANS, even with con- siderably reduced activity. Among the P1 proteases, only proteinase K revealed a signiWcant turnover rate. The emis- sion maximum of the Xuorescent product formed by pro- Fig. 2. Hydrolysis of Dabcyl-Ser-Phe-EDANS by various endoproteases. He (pH 7.5) was incubated with 1.2 �M dispase (�), 1.2 �M thermolysin (�), 1 1.2 �M chymotrypsin (¤) at 37 °C for 10 min (Wnal volume of 500 �l). For p enzyme concentration was enhanced to 6 �M. The 1.2-�M pepsin reaction ( EDTA (thermolysin, dispase, and collagenase), 2 mM AEBSF (proteinase K Xuorescence intensity was monitored at 490 nm (�ex D 336 nm). The control (+ 0 50 100 150 200 400 450 500 wave re la tiv e flu or es ce nc e exclusively, revealing the typical P1� speciWcity (Fig. 3, lines 4 and 5). After an overnight incubation, the starting eimer et al. / Anal. Biochem. 352 (2006) 110–119 re, 41 �M of the quenched Xuorophore in 2 mM CaCl2 and 0.1 M Tris–HCl .2 �M collagenase (�), 1.2 �M proteinase K (�), 1.2 �M subtilisin (�), or apain (�) and bromelain (�), a 0.1 M citrate buVer (pH 6.5) was used, and ) was carried out in 0.01 M HCl. Reactions were stopped by 500 �l of 20 mM , chymotrypsin, papain, and bromelain), or Tris–HCl (pH 7.5) (pepsin), and ) contained no protease. 550 600 length [nm] 100 88 25 19 1 2 9 4 1 dispase thermolysin collagenase proteinase K subtilisin chymotrypsin papain bromelain pepsin Rel. Fl. (%)Protease Dabcyl-Ser-Phe-EDANS Phe-EDANS teinase K seems to be slightly moved to lower wavelengths that coincide more with the emission spectrum of EDANS than with that of Phe-EDANS. All other used proteases could not produce an adequate increase in Xuorescence to establish sensitive protease assays. It should be noted that papain and bromelain were used in Wvefold concentrations. The metalloproteases were assumed to cleave the peptide bond between serine and phenylalanine, whereas proteinase K preferably should remove EDANS. Thin-layer chroma- tography (TLC) was used to determine the degradation products (Fig. 3). The orange-red protease substrate Dab- cyl-Ser-Phe-EDANS can be seen under UV light as a dark dot or band exhibiting an Rf value of 0.72 (Fig. 3, line 1). Correspondingly, dark bands at an Rf value of 0.53 most likely represent Dabcyl-serine. The references of the poten- tial products EDANS and Phe-EDANS are shown in Fig. 3 (lines 2 and 3) as white spots. EDANS remained under the used conditions near the starting line (Rf D 0–0.1), whereas Phe-EDANS had an Rf value of 0.34. Ser-Phe-EDANS (Rf D 0.22) would appear between the spots of EDANS and Phe-EDANS (not shown). Lack of a corresponding band indicated that none of the used proteases could release the Dabcyl residue. Dispase and thermolysin cleaved the Ser-Phe bond material was completely dissipated and only Dabcyl-Ser and Phe-EDANS could be detected. Similarly, collagenase preferably hydrolyzed the Xuorogenic peptide such as dis- pase and thermolysin. The weaker Phe-EDANS formation and a faint EDANS band (Rf D 0.1) suggested an additional disposition to open the amide bond between the phenylala- nyl residue and EDANS. Removal of EDANS was the main preference of protein- ase K, as was expected for a P1 protease. Only traces of a Xuorescent Phe-EDANS band are recognizable on the TLC (Fig. 3, line 7). Compared with it, EDANS showed a sub- stantial spot. However, Dabcyl-Ser-Phe was absent, and besides the remaining Dabcyl-Ser-Phe-EDANS, only the Fig. 3. Products of Dabcyl-Ser-Phe-EDANS after proteolytic cleavage indicated by TLC. The reaction mixtures of Fig. 2 were incubated over- night, applied onto silica 60 aluminum sheets, separated by trichlorome- thane/methanol/water (65:25:4), and irradiated at 365 nm by a Biometra transilluminator. Lane 1, Dabcyl-Ser-Phe-EDANS; lane 2, EDANS; lane 3, Phe-EDANS; lanes 4–8, Dabcyl-Ser-Phe-EDANS treated with dispase, thermolysin, collagenase, proteinase K, and papain, respectively. 1 2 3 4 5 6 7 8 EDANS distinct band of the assumed Dabcyl-Ser product was detectable after the overnight incubation. Additional Assaying dispase- and thermolysin-like proteases / S. W truncation of serine by proteinase K must have occurred in a second step. Attempts to identify the intermediate Ser- Phe-EDANS at a reduced reaction time gave no other result (not shown). Papain, in turn, seemed to exhibit the same speciWcity as the metalloproteases (Fig. 3, line 8). However, it should be taken into consideration that only a small portion of Dab- cyl-Ser-Phe-EDANS was transformed by the protease after a long incubation over several hours. Quenching eVects of the hydrolysis products Fig. 4. EVect of product formation on Xuorescence intensity after complete cle hydrolyzed up to 3 h by 181 nM dispase in 0.1 M Tris–HCl (pH 7.5) containi surements between 90 and 180 min when Xuorescence remained unchanged. In turnover rates, where 1 nm of hydrolyzed dipeptide per milliliter enhances Xu 0 50 100 150 200 250 300 350 0 50 100 substrate c re la tiv e flu or es ce nc e in te ns ity 0 20 40 60 80 100 120 140 160 180 re la tiv e flu or es ce nc e in te ns ity Accordingly, 1 nm of hydrolyzed dipeptide per milliliter increases the emission intensity by 3.7 rfu. eimer et al. / Anal. Biochem. 352 (2006) 110–119 115 Kinetic behaviour of Dabcyl-Ser-Phe-EDANS The Xuorogenic P1� protease substrate was studied in concentrations below 80�M to avoid quenching eVects by the proteolytically formed products. In most experiments, 40�M Dabcyl-Ser-Phe-EDANS were used to attain more than 50% of the highest possible Xuorescence (cf. Fig. 4). Initial linear rates of Xuorescence enhancement were observed at least within the Wrst 15 min when 60 nM dispase hydrolyzed no more than 20% of the protease substrate (Fig. 5). An emission plateau after 2.5 h at approximately avage of Dabcyl-Ser-Phe-EDANS by dispase. The Xuorescent dipeptide was ng 2 mM CaCl2 at 37 °C. The data represent the means of the last four mea- set: concentration range with linear increasing Xuorescence used to calculate orescence by 3.7 U (correlation factor of 0.9847). 150 200 250 oncentration [μM] 0 10 20 30 40 substrate concentration [μM] Fluorescence of the naphthalene residue of Dabcyl-Ser- Phe-EDANS is eVectively quenched by energy transfer to the phenylazobenzene group. No more than 26–28 rfu was moni- tored at 200–250�M. Increasing amounts of the quenched Xuorescent compound correlated approximately in a linear manner with increasing Xuorescence, at least in the concen- trations of 0–80�M used below (not shown). When Dabcyl- Ser-Phe-EDANS was incubated with dispase, so long as emission intensity could not be more elevated, Xuorescence was enhanced up to 300 rfu (Fig. 4). The maximum was already reached above 150�M. Cleavage of more than 240�M substrate resulted in a decrease in Xuorescence (not shown). It seems obvious that there is a concentration limit where the average distance between the released donor and acceptor molecules becomes too short to allow further enhancement of Xuorescence intensity. The linear increasing emission range was clearly ascertained to be below 40�M of the Xuorescent product (Fig. 4, inset). Because Dabcyl-Ser- Phe-EDANS was completely cleaved in these measurements, Fig. 4 can be used as a calibration curve to convert relative Xuorescence units into enzyme units or turnover rates. 150 rfu indicated a complete turnover of the starting mate- rial and the absence of any quenching eVect. The same reaction was studied using various dispase concentrations for an incubation time of 10 min (Fig. 6). Linear enhancement in Xuorescence was interrupted by the lack of substrate at 0.3–1.8 �M dispase, as could be expected. The maximum corresponded with values already obtained for 40 �M of completely hydrolyzed substrate (cf. Figs. 4 and 5). However, beyond the maxi- mum, a nearly linear decline in Xuorescence occurred with increasing amounts of the protease. Usually emis- sion activity of a Xuorescent compound, in this case of Phe-EDANS, is intensiWed and not reduced by replace- ment of a watery environment for the hydrophobic sur- face of a protein. An inhibitory eVect of the products should be excluded because hydrolysis of Dabcyl-Ser- Phe-EDANS was already completed at a lower dispase concentration. It seems more likely that the zinc ion in the catalytic core of dispase diminished Xuorescence intensity of Phe-EDANS as it was known for other Xuo- rophores such as Zinquin and Zinpyr compounds [19,20]. Therefore, zinc metalloproteases such as dispase should be used in concentrations of less than 2 �m/L to accom- plish full sensitivity. 116 Assaying dispase- and thermolysin-like proteases / S. W Fig. 5. Hydrolysis of Dabcyl-Ser-Phe-EDANS by dispase. Here, 41.2 �M of t incubated with 60 nM dispase at 37 °C (Wnal volume of 500 �l). The reaction w was monitored at 490 nm (�ex D 336 nm). Inset: initial phase of dispase hydrol re la tiv e flu or es ce nc e 0 30 60 90 120 150 180 0 30 60 90 time re la tiv e flu or es ce nc e in te ns ity 100 120 140 160 e in te ns ity 3 4 V (nm tide bond may contribute to the rapid formation of distorted enzyme–substrate complexes. A necessary reorientation that eimer et al. / Anal. Biochem. 352 (2006) 110–119 improves the Wt within the active site may then be due to comparably low hydrolysis rates, resulting as a whole in moderate catalytic eYciencies. Nevertheless, the short pep- tide appears to be an inexpensive alternative to specially tai- lored oligopeptides that do not force themselves consistently he quenched Xuorophore in 2 mM CaCl2 and 0.1 M Tris–HCl (pH 7.5) was as stopped by adding of 500 �l of 20 mM EDTA, and Xuorescence intensity ysis used for rate determinations. 0 5 10 15 20 25 30 35 0 300 600 900 time [s] 120 150 180 210 [min] 0 10 20 30 40 n m ol hydrolysed Dabcyl-Ser -Phe-EDA NS0 5 10 15 20 25 30 35 0 300 600 900 time [s] re la tiv e flu or es ce nc e Final measurements of turnover rates at various substrate concentrations revealed that the dispase-mediated hydrolysis of Dabcyl-Ser-Phe-EDANS obeys a simple Michaelis–Men- ten relationship (Fig. 7). The high aYnity of dispase and thermolysin for the short dipeptide was expressed by low Km values of 91§9 and 104§18�M, respectively, which are comparable to those determined with other proteases and longer peptides (Table 1). Contingently, the close proximity of the hydrophobic dyes to the amino acids at the single pep- by more favorable kinetic parameters. In any case, it is a much better option than furylacryloylpeptides for measuring activity of P1� proteases. Properties of dispase assayed with Dabcyl-Ser-Phe-EDANS Three buVers were used to study the inXuence of buVer substances and pH on the interaction of the Xuorogenic substrate and dispase. As can be seen in Fig. 8, the eVect of buVer compounds on enzyme activity was low. The nearly ideal bell-shaped curve of the neutral metalloprotease revealed a tight maximum at pH 7.0. Below pH 6.5 and above pH 7.5, no more than 25% of maximum activity could be measured. In addition, activity of dispase was examined using Dab- cyl-Ser-Phe-EDANS at various temperatures to get infor- mation about the stability of the novel substrate (Fig. 9). The determined optimal temperature of 50 °C was consis- tent with similar data of neutral metalloproteases from other mesophilic Bacillus species [21–23]. As can be judged from Xuorescence at 100 °C, hydrolysis of the Xuorogenic dipeptide did not take place without dispase. Conclusions Fig. 6. EVect of dispase concentration on hydrolysis of Dabcyl-Ser-Phe- EDANS. Here, 41.2 �M of the quenched Xuorophore in 2 mM CaCl2 and 0.1 M Tris–HCl (pH 7.5) was incubated with various amounts of dispase at 37 °C for 10 min. The assay was continued as described in Fig. 5. 0 20 40 60 80 0 1 2 3 dispase [μM] re la tiv e flu or es ce nc 0 1 2 ol min -1 m l -1) Although metalloproteases such as thermolysin have been studied for many years, no convenient substrate to measure Assaying dispase- and thermolysin-like proteases / S. W 80 100 120 140 sc en ce in te ns ity in a steep increase in Xuorescence following Wrst-order kinet- ics. The problem with the quenched Xuorogenic substrates is eimer et al. / Anal. Biochem. 352 (2006) 110–119 117 60 80 V (nmol mi80 100 m g) proteolytic activity continuously was available. Symptomatic of this is that furylacryloyl peptides of low sensitivity are still in use. For medical purposes, highly speciWc oligopeptides provided with FRET systems have been constructed. Fre- quently, Xuorescent EDANS, contributing to a better solubil- ity of the peptide substrate, has been combined with Dabcyl, serving as a quenching acceptor. Hydrolysis of such Dabcyl– EDANS peptides by speciWc endoproteases usually resulted that they exhibit, without exception, a long peptide chain normally requiring an automated solid-phase peptide synthe- sis apparatus. Such a procedure limits the availability of a peptide and is cost-intensive. Moreover, in most cases, the sequences of the FRET peptides have been derived from nat- ural substrates to meet the speciWcity of the concerned endo- proteases. Therefore, the general usefulness of the peptides needs to be questioned. Note. Measurements were carried out as described in Materials and methods. DTT, dithiothreitol; BSA, bovine serum albumin; DMSO, dimethyl sulfoxide. a X, �-aminobutyric acid. b Calculated from published data. protease �-Secretase RE(ƒ)EVNLDAEFK(ƒ)R 0.1 M acetate (pH 4.5), 10% DMSO 5.4 0.24 0.74b [7] Trypsin ƒ GPARLAIG ƒ 50 mM Hepes (pH 8.0), 10 mM CaCl2, 0.1 M NaCl 34 2400b 1170 [15] Cathepsin D Ac-EE(ƒ)KPILFF RLGK(ƒ)E-NH2 50 mM glycine (pH 3.5), 2% DMSO 5.7 2376b 7000 [9] MMP-1 ƒ XaPQGLE(ƒ)AK-NH2 50 mM Tris (pH 7.6), 0.15 M NaCl, 5 mM CaCl2, 1 �M ZnCl2, 0.01% Brij35 21 [5] MMP-2 619 Fig. 7. Hydrolysis rates of Dabcyl-Ser-Phe-EDANS at various concentrations. The quenched Xuorophore in 2 mM CaCl2 and 0.1 M Tris–HCl (pH 7.5) was incubated with 1.1 �M dispase at 37 °C for 10 min. The assay was continued as described in Fig. 5. Inset: double-reciprocal plot of velocity (means of three measurements) versus substrate concentration used to determine the kinetic parameters of dispase. 0 20 40 60 0 20 40 60 80 substrate [μM] re la tiv e flu or e 0 20 40 n -1 m g -1) 0 20 40 60 -50 0 50 100 150 200 1/S (mM-1) 1/ v (µ mo l-1 m in Table 1 Kinetic parameters of dispase and thermolysin using Dabcyl-Ser-Phe-EDANS compared with data published for various proteases and longer peptides Protease Substrate (Dabcyl ƒ EDANS) BuVer Km (�M) kcat (min ¡1) kcat/Km (mM ¡1 s¡1) Reference Dispase ƒ SF ƒ 0.1 M Tris (pH 7.5), 2 mM CaCl2 91 (§ 9) 15.4 (§ 1.0) 2.8 (§ 0.01) Thermolysin ƒ SF ƒ 0.1 M Tris (pH 7.5), 2 mM CaCl2 104 (§ 18) 25.8 (§ 3.6) 4.1 (§ 0.1) HIV-1 PR ƒ XaSQNYPIVQ ƒ 0.1 M acetate (pH 4.7), 1 M NaCl, 1 mM DTT, 1 mM EDTA, 1mg/ml BSA, 10% DMSO 103 294b 48b [3] ADAM33 protease K(ƒ)YRVAFQKLAE(ƒ)K 20 mM Hepes (pH 7.0), 0.5 M NaCl, 0.2 mg/ml BSA 32 72 36 [4] SARS protease ƒ VNSTLQSGLRK(ƒ)M 20 mM phosphate (pH 7.5), 0.1 M NaCl, 5 mM DTT, 1 mM EDTA 404 1.08 0.045b [21] SARS protease ƒ KTSAVLQSGFRKME ƒ 20 mM Bis–Tris (pH 7.0) 17 114b 112 [10] HTLV-1 ƒ XaPQVLNphVMH ƒ 10 mM acetate (pH 5.3) 58 12b 4 [11] We have shown in this article that a short dipeptide, Dabcyl-Ser-Phe-EDANS, is an appropriate substrate for 118 Assaying dispase- and thermolysin-like proteases / S. W frequently studied metalloproteases of dispase and thermoly- sin. Both endoproteases prefer a hydrophobic amino acid in the P1� position, and it could be expected that other prote- ases exhibiting the same speciWcity will hydrolyze the peptide with a similar turnover rate. Actually, C. histolyticum collage- nase with a quite diVerent speciWcity and the serine protease proteinase K from T. album could generate a Xuorescence signal, even with considerably reduced activity. The simple synthesis procedure of Dabcyl-Ser-Phe-EDANS, along with its high aYnity for dispase and thermolysin, should allow a Fig. 8. Determination of the optimal pH of dispase using Dabcyl-Ser-Phe- EDANS. Here, 41.2 �M substrate was hydrolyzed by 0.3 �M dispase in 0.2 M acetate/2 mM CaCl2 (- - -), 0.2 M citrate/2 mM CaCl2 (—), and 0.2 M Tris–HCl/2 mM CaCl2 (-·-) at 37 °C for 10 min as described in Fig. 5. 0 200 400 600 800 1000 3.5 5.5 7.5 9.5 pH V (n mo l m in- 1 m g- 1 ) Fig. 9. Determination of the optimal temperature of dispase using Dabcyl- Ser-Phe-EDANS. Here, 41.2 �M substrate was hydrolyzed by 0.12 �M dis- pase for 10 min as described in Fig. 5. 0 200 400 600 800 1000 1200 1400 1600 1800 20 40 60 80 100 temperature [˚C] V (n mo l m in- 1 m g- 1 ) broader application of the novel protease substrate than could be expected of the tailor-made oligopeptides. eimer et al. / Anal. Biochem. 352 (2006) 110–119 Acknowledgments This work was supported by the R&D Centre of the University of Applied Sciences of Darmstadt, Germany. S. Weimer gratefully acknowledges N-Zyme BioTec (Darmstadt) for Wnancial donations. References [1] I. Schechter, A. Berger, On the size of the active site in proteases: I. Papain, Biochem. Biophys. Res. Commun. 27 (1967) 157–162. [2] J. Feder, L.R. Garrett, A rapid method for removal of zinc from the metallo neutral proteases, Biochem. Biophys. Res. Commun. 43 (1971) 943–948. [3] E.D. Matayoshi, G.T. Wang, G.A. KraVt, J. Erickson, Novel Xuoro- genic substrates for assaying retroviral proteases by resonance energy transfer, Science 247 (1990) 954–958. [4] J. Zou, R. Zhang, F. Zhu, J. Liu, V. Madison, S.P. Umland, ADAM33 enzyme properties and substrate speciWcity, Biochemistry 44 (2005) 4247–4256. [5] B. Beekman, B. van Ei, J.W. Drijfhout, H.K. Ronday, J.M. teKoppele, Highly increased levels of active stromelysin in rheumatoid synovial Xuid determined by a selective Xuorogenic assay, FEBS Lett. 418 (1997) 305–309. [6] B.F. Becker, S. Gilles, C.P. SommerhoV, S. Zahler, Application of pep- tides containing the cleavage sequence of pro-TNF� in assessing TACE activity of whole cells, Biol. Chem. 383 (2002) 1821–1826. [7] J. ErmolieV, J.A. Loy, G. Koelsch, J. Tang, Proteolytic activation of recombinant pro-memapsin 2 (pro-�-secretase) studied with new Xuorogenic substrates, Biochemistry 39 (2000) 12450–12456. [8] E. Schmitt, A. Steyaert, G. Cimoli, R. Bertrand, Bax-� promotes apoptosis induced by cancer chemotherapy and accelerates the activation of caspase 3-like cysteine proteases in p53 double mutant B lymphoma Namalwa cells, Cell Death DiVer. 5 (1998) 506–516. [9] S.V. Gulnik, L.I. Suvorov, P. Majer, J. Collins, B.P. Kane, D.G. John- son, J.W. Erickson, Design of sensitive Xuorogenic substrates for human cathepsin D, FEBS Lett. 413 (1997) 379–384. [10] C.J. Kuo, Y.H. Chi, J.T. Hsu, P.H. Liang, Characterization of SARS main protease and inhibitor assay using a Xuorogenic substrate, Bio- chem. Biophys. Res. Commun. 318 (2004) 862–867. [11] J.J. Ha, D.A. Gaul, V.L. Mariani, Y.S. Ding, R.A. Ikeda, S.B. Shuker, HTLV-I protease cleavage of P19/24 substrates is not dependent on NaCl concentration, Bioorg. Chem. 30 (2002) 138–144. [12] M. Taliani, E. Bianchi, F. Narjes, M. Fossatelli, A. Urbani, C. Stein- kühler, R. De Francesco, A. Pessi, A continuous assay of hepatitis C virus protease based on resonance energy transfer depsipeptide sub- strates, Anal. Biochem. 240 (1996) 60–67. [13] G.T. Wang, C.C. Chung, T.F. Holzman, G.A. KraVt, A continuous Xuorescence assay of renin activity, Anal. Biochem. 210 (1993) 351–359. [14] C. Garcia-Echeverria, D.H. Rich, New intramolecularly quenched Xuorogenic peptide substrates for the study of the kinetic speciWcity of papain, FEBS Lett. 297 (1992) 100–102. [15] S. Grahn, D. Ullmann, H. Jakubke, Design and synthesis of Xuoro- genic trypsin peptide substrates based on resonance energy transfer, Anal. Biochem. 265 (1998) 225–231. [16] J. Zotzel, P. Keller, H.-L. Fuchsbauer, Transglutaminase from Strep- tomyces mobaraensis is activated by an endogenous metalloprotease, Eur. J. Biochem. 270 (2003) 3214–3222. [17] J. Zotzel, R. Pasternack, C. Pelzer, M. Mainusch, H.-L. Fuchsbauer, Activated transglutaminase from Streptomyces mobaraensis is pro- cessed by a tripeptidyl aminopeptidase in the Wnal step, Eur. J. Bio- chem. 270 (2003) 4149–4155. [18] R. Pasternack, S. Dorsch, J.T. Otterbach, I.R. Robenek, S. Wolf, H.-L. Fuchsbauer, Bacterial pro-transglutaminase from Streptoverticillium Assaying dispase- and thermolysin-like proteases / S. Weimer et al. / Anal. Biochem. 352 (2006) 110–119 119 mobaraense: PuriWcation, characterisation, and sequence of the zymo- gen, Eur. J. Biochem. 257 (1998) 570–576. [19] P.D. Zalewski, I.J. Forbes, W.H. Betts, Correlation of apoptosis with change in intracellular labile Zn(II) using zinquin [(2-methyl-8-p-tolu- enesulphonamido-6-quinolyloxy)acetic acid], a new speciWc Xuores- cent probe for Zn(II), Biochem. J. 296 (1993) 403–408. [20] S.C. Burdette, G.K. Walkup, B. Spingler, R.Y. Tsien, S.J. Lippard, Fluorescent sensors for Zn2+ based on a Xuorescein platform: synthe- sis, properties, and intracellular distribution, J. Am. Chem. Soc. 123 (2001) 7831–7841. [21] J.K. Sierecka, PuriWcation and partial characterization of a neutral protease from a virulent strain of Bacillus cereus, Int. J. Biochem. Cell Biol. 30 (1998) 579–595. [22] G.A. Rufo Jr., B.J. Sullivan, A. Sloma, J. Pero, Isolation and charac- terization of a novel extracellular metalloprotease from Bacillus sub- tilis, J. Bacteriol. 172 (1990) 1019–1023. [23] M.K. Pangburn, P.L. Levy, K.A. Walsh, H. Neurath, Thermal stability of homologous neutral metalloendopeptidases in thermophilic and mesophilic bacteria: structural considerations, Experientia Suppl. 26 (1976) 19–30. A quenched fluorescent dipeptide for assaying dispase- and thermolysin-like proteases Materials and methods Materials Apparatus Peptide synthesis N-tert-Butoxycarbonyl-l-phenylalanyl pentafluorophenylester (Boc-Phe-OPFP) 2-N-(Nprime-tert-Butoxycarbonyl-l-phenylalanylamido)-NPrime-ethylaminonaphthalene-5-sulfonic acid (Boc-Phe-EDANS) 2-N-(l-Phenylalanylamido)-Nprime-ethylaminonaphthalene-5-sulfonic acid (Phe-EDANS) N-tert-Butoxycarbonyl-O-tert-butyl-l-serine pentafluorophenylester (Boc-Ser(tBu)-OPFP) 2-N-(Nprime-tert-Butoxycarbonyl-O-tert-butyl-l-serinyl-l-phenylalanylamido)-NPrime-ethylaminonaphthalene-5-sulfonic acid (Boc-Ser(tBu)-Phe-EDANS) 2-N-(l-Serinyl-l-phenylalanylamido)-Nprime-ethylaminonaphthalene-5-sulfonic acid (Ser-Phe-EDANS) 2-N-(4-[4prime-Nprime,Nprime-(Dimethylamino)phenylazo]benzoyl-l-serinyl-l-phenylalanylamido)-NPrime-ethylaminonaphthalene-5-sulfonic acid (Dabcyl-Ser-Phe-EDANS)2The compound is available from N-Zyme BioTec (Darmstadt, Germany).2 Fluorometric assay Results and discussion Synthesis of Dabcyl-Ser-Phe-EDANS Hydrolysis of Dabcyl-Ser-Phe-EDANS by various proteases Quenching effects of the hydrolysis products Kinetic behaviour of Dabcyl-Ser-Phe-EDANS Properties of dispase assayed with Dabcyl-Ser-Phe-EDANS Conclusions Acknowledgments References