- Email: [email protected]
New Fluorogenic Substrates forN-Arginine Dibasic Convertase
Analytical Biochemistry 269, 149 –154 (1999) Article ID abio.1999.4033, available online at http://www.idealibrary.com on
New Fluorogenic Substrates for N-Arginine Dibasic Convertase Eva Csuhai,* Maria Aparecida Juliano,† Jan St. Pyrek,‡ Amy C. Harms,‡ Luiz Juliano,† and Louis B. Hersh* *Department of Biochemistry and ‡Mass Spectrometry Facility, University of Kentucky, Lexington, Kentucky, 40536-0084; and †Department of Biophysics, Escola Paulista de Medicina, Sa˜o Paulo, Brazil
Received October 16, 1998
N-Arginine dibasic (NRD) convertase is a recently described peptidase capable of selectively cleaving peptides between paired basic residues. The characterization of this unique peptidase has been hindered by the fact that no facile assay procedure has been available. Here we report the development of a rapid and sensitive assay for NRD convertase, based on the utilization of two new internally quenched fluorogenic peptides: Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp. These peptides contain the fluorescent 2-aminobenzoyl moiety that is quenched in the intact peptide by a 2,4-dinitrophenyl moiety. Cleavage by NRD convertase at the Arg-Arg sequence results in an increase of fluorescence. NRD convertase cleaves these peptides efficiently and with high specificity as observed by both HPLC and fluorescence spectroscopy. The rate of hydrolysis of the fluorogenic substrates is proportional to enzyme concentration, and obeys Michaelis–Menten kinetics. The kinetic parameters for the fluorescent peptides (K m values of ;1.0 mM, and V max values of ;1 mM/(min z mg) are similar to those obtained with peptide hormones as substrates. © 1999 Academic Press
Peptidases have been found to play an important role in modulating the action of peptide hormones. Peptidases can hydrolytically inactivate a peptide, can convert a peptide hormone precursor to an active form, or can alter the physiological activity of the parent peptide. Recently, a novel endopeptidase, N-arginine dibasic convertase (NRD convertase), 1 was described 1 Abbreviations used: NRD convertase, N-arginine dibasic convertase; Abz, 2-aminobenzoyl; EDDnp, ethylenediamine-2,4-dinitrophenyl; TBTU, O-benzotriazol-1-yl-N,N,N9,N9-tetramethyluronium tetrafluoroborate; Hobt, 1-hydroxybenzotriazole; NMM, N-methyl
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
(1, 2). NRD convertase is a 130-kDa metallopeptidase which cleaves peptide bonds between paired basic sequences of the type Arg-Arg or Arg-Lys (1, 3). The enzyme is broadly expressed in tissues, being most abundant in testes, with high enzyme activity, as well as mRNA, also found in brain and spinal cord (4, 5). The enzyme appears to be developmentally regulated; at early stages of embryonic development it is detected exclusively in brain and spinal cord. At later developmental stages this peptidase appears to be expressed ubiquitously (6). It is interesting to note that NRD convertase is abundant in all immortalized cell lines examined to date (5, 7). NRD convertase is a soluble protein found in the cytosol as well as in secretory vesicles (8). In addition, a secreted form has been reported (3). This endopeptidase is a member of a relatively new family of metallopeptidases, the Zn-containing inverzincins. Other members of this group include insulin-degrading enzyme and protease III of Escherichia coli. The putative active site in this peptidase family is characterized by an inverted Zn-binding motif HXXEH (9). The mechanism of peptide cleavage at this inverted Zn-binding site has not yet been fully characterized. NRD convertase has been shown to cleave a relatively narrow range of substrates in vitro, including the opioid peptides dynorphin A and B, bovine adrenal medulla peptide, and a-neoendorphin (3). The in vivo substrate specificity of the enzyme, however, has yet to be determined. A distinguishing characteristic of NRD convertase is an extensive acidic domain consisting of 43–59 negatively charged residues within a 76-amino-acid stretch
morpholine; MALDI–TOF, matrix-assisted laser desorption ionization–time of flight; Fmoc, 9-fluorenylmethoxycarbonyl. 149
150
CSUHAI ET AL.
(4, 10). An acidic domain of this length is unprecedented in the literature. This domain is postulated to play a role in regulation of enzymatic activity by charge– charge interactions with other cellular components (11). In addition to the regulation of enzymatic activity, similar charge– charge interactions through this unique domain might also lead to protein–protein interactions with implications on cellular signaling. To date the assay used to measure NRD convertase activity is based on following the conversion of in vitro peptide substrates (e.g., somatostatin analogs or bovine adrenal medulla peptide) into cleavage products by HPLC using reverse-phase chromatography. This assay is tedious, time-consuming, and, since it relies on the detection of products in a stopped reaction mixture, is discontinuous. In our hands the HPLC assay requires the generation of at least 0.1 nmol of product. A more sensitive, rapid, and continuous method of assaying NRD convertase would be preferable and would facilitate studies on this unique enzyme. Fluorogenic peptides have recently been used to characterize various peptidases, including, most recently, the E. coli leader peptidase (12) and a human cytomegalovirus protease (13). In these peptide derivatives, based on a concept originally developed by Yaron et al. (14), a fluorescent acceptor group is attached to one of the amino acid residues and a quenching donor is attached to another residue in the sequence. Cleavage that takes place at any of the peptide bonds between the donor/acceptor pair leads to an increase in fluorescence that can be quantified fluorometrically. This technology has the advantage of high sensitivity and the ability to monitor the reaction on a continuous basis. Thus any deviations from linearity are readily apparent. To this end we have synthesized two internally quenched fluorogenic peptide substrates for NRD convertase which are based on the sequences found in the peptides dynorphin and somatostatin. These substrates contain an N-terminal fluorescent 2-aminobenzyl (Abz) group and a C-terminal fluorescent-quenching ethylenediamine-2,4-dinitrophenyl (EDDnp) group (15, 16). Cleavage at any peptide bond leads to the separation of the quenching EDDnp group from the fluorescent Abz group and an increase in fluorescence. The Abz/EDDnp pair produces one of the best quencher effects; consequently, a large increase of fluorescence is observed after substrate hydrolysis. MATERIALS AND METHODS
Buffers and chemicals were obtained from Sigma Chemical Co. Matrix-assisted laser desorption ionization mass spectrometry was performed at the University of Kentucky Mass Spectrometry Facility. An automated benchtop simultaneous multiple solid-phase peptide synthesizer (PSSM 8 system from Shimadzu,
Tokyo) was used for solid-phase peptide synthesis with Fmoc methodology (17). Nova Syn Tentagel resin (0.3 meq/g) from NovaBiochem (San Diego, CA) was employed, and all couplings were performed with TBTU/ HOBt using NMM as base.
Synthesis of Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp Fmoc-Glu-(CO a EDDnp) (5 eq) was coupled with TBTU (5 eq), HOBT (5 eq), and NMM (7.5 eq) in DMF to the free amino group of the linker p-[(R,S)a - [1-(9H-fluoren-9-yl)methoxyformamido] - 2,4-dimethoxybenzyl]-phenoxyacetic acid already attached to Nova Syn TGR resin (0.3 meq/g) from NovaBiochem. The Fmoc group was removed by exposure to 30% piperidine in DMF for 10 min and each amino acid was coupled and deprotected by the same procedure. The guanido side chain of Arg was protected by the 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmc) group. After the coupling of Boc-Abz, obtained as previously described (18), the resin was treated with a solution containing TFA:anisol:EDT:H 2O (85:5:3:7) for 8 h to cleave the side chain protections and to release the peptides from the resin. This long treatment period was required to remove arginine side chain protection.
Synthesis of N-Fmoc-Glu-(CO aEDDnp) EDDnp (N-[2,4-dinitrophenyl]-ethylenediamine) as its hydrochloride was obtained as follows: 1.6 mol of ethylenediamine was dissolved in 360 ml of dry dioxane, cooled in an ice bath under nitrogen atmosphere and 0.08 mol of 1-fluor-2,4-dinitrobenzene, dissolved in 120 ml of dry dioxane, was added dropwise under stirring during 1 h. The mixture was then kept under stirring for 3 h at room temperature. The solvent was removed by evaporation under reduced pressure and water was added. The precipitate was collected by filtration, dissolved in EtOH and the pH was adjusted to 2. The precipitate was collected and washed with chilled EtOH. Yield was 94%, mp 5 268 –270°C. N-Fmoc-Glu-(CO aEDDnp) was obtained essentially as previously described (15). Briefly, N a-Boc-Glu-(COBzl)-(CO-EDDnp) was obtained by the mixed anhydride procedure using isobutyl chlorocarbonate; mp 5 157–159°C. This compound was treated with anhydrous HF containing anisol (2% v/v) for 1 h at 0°C. After evaporating the HF, the residue was dissolved in water, washed with ethyl ether, and lyophilized. The resulting compound [H 3N-Glu-(CO-EDDnp)] was reacted with Fmoc-succinimide and the product was crystallized from ethyl ether, mp 5 148 –152°C.
FLUOROGENIC ASSAY WITH N-ARGININE DIBASIC CONVERTASE
151
netic parameters were calculated using the computer programs of Cleland (20).
TABLE 1
Characterization of Fluorogenic Peptides Mol wt
RESULTS AND DISCUSSION
Peptide
HPLC retention time (min)
Calcd
Obsd [M1H] 1
Yield (%)
Abz-GGFLRRVGQ-EDDnp Abz-GGFLRRIQ-EDDnp
10.4 9.1
1316 1273
1317 1274
40 45
The characteristics of the two synthetic peptides are given in Table 1. Separation of Peptides and Cleavage Products Peptides and Abz/EDDnp-containing peptide analogs were separated by reverse-phase chromatography on a Vydac C 4 analytical column using a Waters highpressure liquid chromatography system with detection of the peptides by absorbance at 214 nm in a Waters 484 UV detector. The mobile phase was 0.1% trifluoroacetic acid in water/acetonitrile. The separations were carried out in a gradient of acetonitrile from 5 to 50%. Peptides were purified for enzymatic assays and for mass spectrometry by HPLC as described above. Starting materials and products were determined by injecting commercially available standards where applicable. Enzyme Purification N-Arginine dibasic convertase was purified from rat testes as previously described (11, 19). Fluorometric Assay of NRD Convertase The fluorescence of reaction mixtures was directly monitored with a Hitachi F-2000 spectrofluorometer. Excitation was set at 319 nm, while emission was measured at 419 nm. In a stirred cell, a 400-ml reaction containing 0.25–20 mM fluorogenic peptide substrate was preincubated at 34°C for 5 min in 20 mM potassium phosphate, pH 7.0. No fluorescence change was observed during this time. Peptide cleavage was initiated by the addition of purified NRD convertase. Emission was monitored continuously for 1–5 min by displaying the output of the spectrofluorometer on a strip chart recorder. For concentration measurements, a standard curve was established by cleaving known amounts of the fluorogenic peptides with trypsin and recording the total fluorescence change. Trypsin cleaves these substrates (1 mM at pH 7.5) with a specific activity of ;4500 mmol/min z mg. The contribution of the intrinsic fluorescence of NRD convertase was negligible under these experimental conditions. Ki-
The need for a simple, rapid, and sensitive assay for measuring N-arginine dibasic convertase activity has been evident since the discovery of this enzyme. We previously reported that the fluorogenic peptide analog Abz-FRRV-EDDnp, which contains the R–R putative NRD convertase cleavage site, was not cleaved by the enzyme (3). Two new, longer fluorogenic peptides, AbzGGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp, were synthesized as potential substrates. In these peptides the fluorescence of the N-terminal aminobenzoyl group is masked by the through-space-quenching effect of the dinitrophenyl moiety. Upon cleavage of any of the peptide bonds, fluorescence increases and thus can be used to measure peptidase activity in a continuously monitored, sensitive assay. The peptide sequence of Abz-GGFLRRVGQ-EDDnp is based on that of bovine adrenal medulla peptide, an excellent substrate of NRD convertase (3, 11), while the sequence in AbzGGFLRRIQ-EDDnp is based on dynorphin A, the substrate with the highest affinity for the enzyme (11). The maximal fluorescence change upon cleavage of the two substrates was obtained by complete cleavage with either trypsin or NRD convertase in 20 mM potassium phosphate buffer, pH 7.0. A total relative fluorescence increase of 7- to 10-fold over the starting value was observed upon the complete cleavage of either of these substrates (Fig. 1). Digestion of Abz-GGFLRRVGQ-EDDnp or Abz-GGFLRRIQ-EDDnp with NRD convertase yielded only two cleavage products from each of the peptide analogs. The cleavage products were separated and purified by reverse-phase HPLC. Mass spectrometry was used to identify Abz-GGFLR as a common product from
FIG. 1. Fluorescence spectra of Abz-GGFLRRVGQ-EDDnp and its cleavage products. Abz-GGFLRRVGQEDDnp (1 mM) in 400 ml of 20 mM potassium phosphate buffer, pH 7.0, was monitored by fluorescence before (b) and after (a) cleavage by NRD convertase. For reference, the absorbance spectrum of the peptide is included (c).
152
CSUHAI ET AL.
FIG. 2. MALDI-FT-ICR-MS spectrum of the two cleavage products of Abz-GGFLRRVGQ-EDDnp. (A) Peptide fragment Abz-GGFLR showing a protonated ion at m/z 668.350 (calcd 668.352) and sodiated ion at m/z 690.334 (calcd 690.334). Ions detected at m/z 678 – 681 represent an unidentified impurity. (B) Peptide fragment RVGQ-EDDnp showing a protonated ion at m/z 667.332 (calcd 667.328), protonated ion minus O at m/z 651.337 (calcd 651.333), protonated ion minus 2O and 2H at m/z 637.358 (calcd 637.353), and protonated ion minus 3O and 2H at m/z 621.363 (calcd 621.359). Peptides were measured in a 2,5-dihydroxy-benzoic acid matrix on a 4.7 T FTMS from IONspec with a N 2 laser.
FLUOROGENIC ASSAY WITH N-ARGININE DIBASIC CONVERTASE
FIG. 3. Dependence of the rate of cleavage on NRD convertase concentration. Abz-GGFLRRVGQ-EDDnp (1 mM) was incubated at 34°C in 20 mM potassium phosphate buffer, pH 7.0, in a 400 ml final volume. Emission was monitored at 419 nm with excitation set at 319 nm. Output was read from a strip chart recorder and rates were determined from the initial velocity of each reaction. NRD convertase was varied from 0.01 to 0.28 mg of protein.
both peptides and RVGQ-EDDnp and RIQ-EDDnp (not shown) were identified as the C-terminal products of Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQEDDnp, respectively (Fig. 2). Both of the peptide analogs were found to be excellent substrates for NRD convertase. The initial velocity of peptide cleavage, shown for Abz-GGFKRRVGQEDDnp, is linearly proportional to enzyme concentration (Fig. 3). We could measure a rate of cleavage of as low as 0.5 pmol/min and detect as little as 1 to 2 ng of
153
NRDc. The cleavage reaction obeyed classical Michaelis–Menten kinetics (Fig. 4A), yielding a K m value of 1.0 mM and a V max value of 1.3 mmol/min z mg for AbzGGFLRRIQ-EDDnp. Although not shown, a K m of 0.70 mM and a V max of 0.86 mmol/min z mg was obtained for Abz-GGFLRRVGQ-EDDnp. These values can be compared to K m of 6 mM and a V max of 3.6 mmol/min z mg for bovine adrenal medulla peptide (11) and K m values for a-neoendorphin and dynorphin B, two similar opioid peptides, of 3.6 and 0.6 mM, respectively. The observed K m values for the two new fluorogenic substrates thus fall within the range observed for peptides of similar size and sequence. FLRKI, the minimal peptide necessary for recognition and selective cleavage by NRD convertase (3), competitively inhibits the cleavage of Abz-GGFLRRVGQ-EDDnp (Fig. 4B). During the purification of NRD convertase through several chromatographic steps, peptidase activity, measured with Abz-GGFLRRVGQ-EDDnp, was directly proportional to enzyme activity measured by the more traditional assay of following the hydrolysis of bovine adrenal medulla peptide by HPLC. Thus, the fluorogenic peptide derivatives provide a convenient substrate to follow the enzyme during its purification. However, these substrates are not specific for NRDc. They will be cleaved by trypsin-like enzymes, and potentially by a variety of other peptidases. However, in the present case, NRDc was the predominant enzyme cleaving these substrates in rat testes homogenates. The new, rapid, and sensitive fluorogenic assay reported here should facilitate studies of the physiological functions of this novel peptidase.
FIG. 4. Kinetic studies with Abz-GGFLRRVGQ-EDDnp. (A) Michaelis–Menten kinetics of the cleavage of Abz-GGFLRRVGQ-EDDnp: Reaction mixtures contained 20 mM potassium phosphate buffer, pH 7.0, 0.1 mg of enzyme, and 0.25–20 mM Abz-GGFLRRVGQ-EDDnp. Activity was determined as described in the legend to Fig. 3. (B) Dixon plot showing inhibition of Abz-GGFLRRVGQ-EDDnp cleavage by a shorter peptide substrate of NRD convertase: Assay conditions were identical to those described above with Abz-GGFLRRVGQ-EDDnp as the fixed variable substrate at 0.5, 0.75, or 1.0 mM. The peptide FLRKI was varied as indicated.
154
CSUHAI ET AL.
ACKNOWLEDGMENTS This work was supported in Brazil by Fundac¸a˜o de Amparo Pesquisa do Estado de Sa˜o Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Cientıf´ico e Tecnolo´gico (CNPq). This work was also supported in part by NIDA/NIH Grant DA02243. E.Cs. is a recipient of a NIDA/NIH Postdoctoral Fellowship DA05671.
REFERENCES 1. Chesneau, V., Pierotti, A. R., Barre´, N., Cre´minon, C., Tougard, C., and Cohen, P. (1994) J. Biol. Chem. 269, 2056 –2061. 2. Foulon, T., Cadel, S., Chesneau, V., Draoui, M., Prat, A., and Cohen, P. (1996) Ann. NY Acad. Sci. 780, 106 –120. 3. Csuhai, E., Safavi, A., and Hersh, L. B. (1995) Biochemistry 34, 12411–12419. 4. Pierotti, A. R., Prat, A., Chesneau, V., Gaudoux, F., Leseney, A.-M., Foulon, T., and Cohen, P. (1994) Proc. Natl. Acad. Sci. USA 91, 6078 – 6082. 5. Hospital, V., Prat, A., Joulie, C., Che´rif, D., Day, R., and Cohen, P. (1997) Biochem. J. 327, 773–779. 6. Fumagalli, P., Accarino, M., Egeo, A., Scartezzini, P., Rappazzo, G., Pizzuti, A., Avvantaggiato, V., Simeone, A., Arrigo, G., Zuffardi, O., Ottolenghi, S., and Taramelli, R. (1998) Genomics 47, 238 –245. 7. Draoui, M., Bellincampi, L., Hospital, V., Cadel, S., Foulon, T., Prat, A., Barre´, N., Reichert, U., Melino, G., and Cohen, P. (1997) J. NeuroOncol. 31, 99 –106.
8. Chesneau, V., Prat, A., Segretain, D., Hospital, V., Dupaix, A., Foulon, T., Je´gou, B., and Cohen, P. (1997) J. Cell Sci. 109, 2737–2745. 9. Hooper, N. M. (1994) FEBS Lett. 345, 1– 6. 10. Chesneau, V., Pierotti, A. R., Prat, A., Gaudoux, F., Foulon, T., and Cohen, P. (1994) Biochimie 76, 234 –240. 11. Csuhai, E., Chen, G., and Hersh, L. B. (1998) Biochemistry 37, 3787–3794. 12. Zhong, W., and Benkovic, S. J. (1998) Anal. Biochem. 255, 66 – 73. 13. Bonneau, P. R., Plouffe, C., Pelletier, A., Wernic, D., and Poupart, M.-A. (1998) Anal. Biochem. 255, 59 – 65. 14. Yaron, A., Carme, A., and Katchalsi-Katzir, E. (1979) Anal. Biochem. 95, 228 –235. 15. Chagas, J. R., Juliano, L., and Prado, E. (1991) Anal. Biochem. 192, 419 – 425. 16. Gershkovich, A. A., and Kholodovych, V. V. (1996) J. Biochem. Biophys. Methods 33, 135–162. 17. Y. Hirata, Cezari, M. H. S., Nakaie, C. R., Boschcov, P., Ito, A. S., Juliano, A. M., and Juliano, L. (1994) Lett. Peptide Sci. 1, 299 – 308. 18. Meldal, M., and Breddam, K. (1991) Anal. Biochem. 195, 141– 147. 19. Cohen, P., Pierotti, A. R., Chesneau, V., Foulon, T., and Prat, A. (1995) Methods Enzymol. 248, 703–716. 20. Cleland, W. W. (1979) Methods Enzymol. 63, 103–137.
New Fluorogenic Substrates for N-Arginine Dibasic Convertase Eva Csuhai,* Maria Aparecida Juliano,† Jan St. Pyrek,‡ Amy C. Harms,‡ Luiz Juliano,† and Louis B. Hersh* *Department of Biochemistry and ‡Mass Spectrometry Facility, University of Kentucky, Lexington, Kentucky, 40536-0084; and †Department of Biophysics, Escola Paulista de Medicina, Sa˜o Paulo, Brazil
Received October 16, 1998
N-Arginine dibasic (NRD) convertase is a recently described peptidase capable of selectively cleaving peptides between paired basic residues. The characterization of this unique peptidase has been hindered by the fact that no facile assay procedure has been available. Here we report the development of a rapid and sensitive assay for NRD convertase, based on the utilization of two new internally quenched fluorogenic peptides: Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp. These peptides contain the fluorescent 2-aminobenzoyl moiety that is quenched in the intact peptide by a 2,4-dinitrophenyl moiety. Cleavage by NRD convertase at the Arg-Arg sequence results in an increase of fluorescence. NRD convertase cleaves these peptides efficiently and with high specificity as observed by both HPLC and fluorescence spectroscopy. The rate of hydrolysis of the fluorogenic substrates is proportional to enzyme concentration, and obeys Michaelis–Menten kinetics. The kinetic parameters for the fluorescent peptides (K m values of ;1.0 mM, and V max values of ;1 mM/(min z mg) are similar to those obtained with peptide hormones as substrates. © 1999 Academic Press
Peptidases have been found to play an important role in modulating the action of peptide hormones. Peptidases can hydrolytically inactivate a peptide, can convert a peptide hormone precursor to an active form, or can alter the physiological activity of the parent peptide. Recently, a novel endopeptidase, N-arginine dibasic convertase (NRD convertase), 1 was described 1 Abbreviations used: NRD convertase, N-arginine dibasic convertase; Abz, 2-aminobenzoyl; EDDnp, ethylenediamine-2,4-dinitrophenyl; TBTU, O-benzotriazol-1-yl-N,N,N9,N9-tetramethyluronium tetrafluoroborate; Hobt, 1-hydroxybenzotriazole; NMM, N-methyl
0003-2697/99 $30.00 Copyright © 1999 by Academic Press All rights of reproduction in any form reserved.
(1, 2). NRD convertase is a 130-kDa metallopeptidase which cleaves peptide bonds between paired basic sequences of the type Arg-Arg or Arg-Lys (1, 3). The enzyme is broadly expressed in tissues, being most abundant in testes, with high enzyme activity, as well as mRNA, also found in brain and spinal cord (4, 5). The enzyme appears to be developmentally regulated; at early stages of embryonic development it is detected exclusively in brain and spinal cord. At later developmental stages this peptidase appears to be expressed ubiquitously (6). It is interesting to note that NRD convertase is abundant in all immortalized cell lines examined to date (5, 7). NRD convertase is a soluble protein found in the cytosol as well as in secretory vesicles (8). In addition, a secreted form has been reported (3). This endopeptidase is a member of a relatively new family of metallopeptidases, the Zn-containing inverzincins. Other members of this group include insulin-degrading enzyme and protease III of Escherichia coli. The putative active site in this peptidase family is characterized by an inverted Zn-binding motif HXXEH (9). The mechanism of peptide cleavage at this inverted Zn-binding site has not yet been fully characterized. NRD convertase has been shown to cleave a relatively narrow range of substrates in vitro, including the opioid peptides dynorphin A and B, bovine adrenal medulla peptide, and a-neoendorphin (3). The in vivo substrate specificity of the enzyme, however, has yet to be determined. A distinguishing characteristic of NRD convertase is an extensive acidic domain consisting of 43–59 negatively charged residues within a 76-amino-acid stretch
morpholine; MALDI–TOF, matrix-assisted laser desorption ionization–time of flight; Fmoc, 9-fluorenylmethoxycarbonyl. 149
150
CSUHAI ET AL.
(4, 10). An acidic domain of this length is unprecedented in the literature. This domain is postulated to play a role in regulation of enzymatic activity by charge– charge interactions with other cellular components (11). In addition to the regulation of enzymatic activity, similar charge– charge interactions through this unique domain might also lead to protein–protein interactions with implications on cellular signaling. To date the assay used to measure NRD convertase activity is based on following the conversion of in vitro peptide substrates (e.g., somatostatin analogs or bovine adrenal medulla peptide) into cleavage products by HPLC using reverse-phase chromatography. This assay is tedious, time-consuming, and, since it relies on the detection of products in a stopped reaction mixture, is discontinuous. In our hands the HPLC assay requires the generation of at least 0.1 nmol of product. A more sensitive, rapid, and continuous method of assaying NRD convertase would be preferable and would facilitate studies on this unique enzyme. Fluorogenic peptides have recently been used to characterize various peptidases, including, most recently, the E. coli leader peptidase (12) and a human cytomegalovirus protease (13). In these peptide derivatives, based on a concept originally developed by Yaron et al. (14), a fluorescent acceptor group is attached to one of the amino acid residues and a quenching donor is attached to another residue in the sequence. Cleavage that takes place at any of the peptide bonds between the donor/acceptor pair leads to an increase in fluorescence that can be quantified fluorometrically. This technology has the advantage of high sensitivity and the ability to monitor the reaction on a continuous basis. Thus any deviations from linearity are readily apparent. To this end we have synthesized two internally quenched fluorogenic peptide substrates for NRD convertase which are based on the sequences found in the peptides dynorphin and somatostatin. These substrates contain an N-terminal fluorescent 2-aminobenzyl (Abz) group and a C-terminal fluorescent-quenching ethylenediamine-2,4-dinitrophenyl (EDDnp) group (15, 16). Cleavage at any peptide bond leads to the separation of the quenching EDDnp group from the fluorescent Abz group and an increase in fluorescence. The Abz/EDDnp pair produces one of the best quencher effects; consequently, a large increase of fluorescence is observed after substrate hydrolysis. MATERIALS AND METHODS
Buffers and chemicals were obtained from Sigma Chemical Co. Matrix-assisted laser desorption ionization mass spectrometry was performed at the University of Kentucky Mass Spectrometry Facility. An automated benchtop simultaneous multiple solid-phase peptide synthesizer (PSSM 8 system from Shimadzu,
Tokyo) was used for solid-phase peptide synthesis with Fmoc methodology (17). Nova Syn Tentagel resin (0.3 meq/g) from NovaBiochem (San Diego, CA) was employed, and all couplings were performed with TBTU/ HOBt using NMM as base.
Synthesis of Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp Fmoc-Glu-(CO a EDDnp) (5 eq) was coupled with TBTU (5 eq), HOBT (5 eq), and NMM (7.5 eq) in DMF to the free amino group of the linker p-[(R,S)a - [1-(9H-fluoren-9-yl)methoxyformamido] - 2,4-dimethoxybenzyl]-phenoxyacetic acid already attached to Nova Syn TGR resin (0.3 meq/g) from NovaBiochem. The Fmoc group was removed by exposure to 30% piperidine in DMF for 10 min and each amino acid was coupled and deprotected by the same procedure. The guanido side chain of Arg was protected by the 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmc) group. After the coupling of Boc-Abz, obtained as previously described (18), the resin was treated with a solution containing TFA:anisol:EDT:H 2O (85:5:3:7) for 8 h to cleave the side chain protections and to release the peptides from the resin. This long treatment period was required to remove arginine side chain protection.
Synthesis of N-Fmoc-Glu-(CO aEDDnp) EDDnp (N-[2,4-dinitrophenyl]-ethylenediamine) as its hydrochloride was obtained as follows: 1.6 mol of ethylenediamine was dissolved in 360 ml of dry dioxane, cooled in an ice bath under nitrogen atmosphere and 0.08 mol of 1-fluor-2,4-dinitrobenzene, dissolved in 120 ml of dry dioxane, was added dropwise under stirring during 1 h. The mixture was then kept under stirring for 3 h at room temperature. The solvent was removed by evaporation under reduced pressure and water was added. The precipitate was collected by filtration, dissolved in EtOH and the pH was adjusted to 2. The precipitate was collected and washed with chilled EtOH. Yield was 94%, mp 5 268 –270°C. N-Fmoc-Glu-(CO aEDDnp) was obtained essentially as previously described (15). Briefly, N a-Boc-Glu-(COBzl)-(CO-EDDnp) was obtained by the mixed anhydride procedure using isobutyl chlorocarbonate; mp 5 157–159°C. This compound was treated with anhydrous HF containing anisol (2% v/v) for 1 h at 0°C. After evaporating the HF, the residue was dissolved in water, washed with ethyl ether, and lyophilized. The resulting compound [H 3N-Glu-(CO-EDDnp)] was reacted with Fmoc-succinimide and the product was crystallized from ethyl ether, mp 5 148 –152°C.
FLUOROGENIC ASSAY WITH N-ARGININE DIBASIC CONVERTASE
151
netic parameters were calculated using the computer programs of Cleland (20).
TABLE 1
Characterization of Fluorogenic Peptides Mol wt
RESULTS AND DISCUSSION
Peptide
HPLC retention time (min)
Calcd
Obsd [M1H] 1
Yield (%)
Abz-GGFLRRVGQ-EDDnp Abz-GGFLRRIQ-EDDnp
10.4 9.1
1316 1273
1317 1274
40 45
The characteristics of the two synthetic peptides are given in Table 1. Separation of Peptides and Cleavage Products Peptides and Abz/EDDnp-containing peptide analogs were separated by reverse-phase chromatography on a Vydac C 4 analytical column using a Waters highpressure liquid chromatography system with detection of the peptides by absorbance at 214 nm in a Waters 484 UV detector. The mobile phase was 0.1% trifluoroacetic acid in water/acetonitrile. The separations were carried out in a gradient of acetonitrile from 5 to 50%. Peptides were purified for enzymatic assays and for mass spectrometry by HPLC as described above. Starting materials and products were determined by injecting commercially available standards where applicable. Enzyme Purification N-Arginine dibasic convertase was purified from rat testes as previously described (11, 19). Fluorometric Assay of NRD Convertase The fluorescence of reaction mixtures was directly monitored with a Hitachi F-2000 spectrofluorometer. Excitation was set at 319 nm, while emission was measured at 419 nm. In a stirred cell, a 400-ml reaction containing 0.25–20 mM fluorogenic peptide substrate was preincubated at 34°C for 5 min in 20 mM potassium phosphate, pH 7.0. No fluorescence change was observed during this time. Peptide cleavage was initiated by the addition of purified NRD convertase. Emission was monitored continuously for 1–5 min by displaying the output of the spectrofluorometer on a strip chart recorder. For concentration measurements, a standard curve was established by cleaving known amounts of the fluorogenic peptides with trypsin and recording the total fluorescence change. Trypsin cleaves these substrates (1 mM at pH 7.5) with a specific activity of ;4500 mmol/min z mg. The contribution of the intrinsic fluorescence of NRD convertase was negligible under these experimental conditions. Ki-
The need for a simple, rapid, and sensitive assay for measuring N-arginine dibasic convertase activity has been evident since the discovery of this enzyme. We previously reported that the fluorogenic peptide analog Abz-FRRV-EDDnp, which contains the R–R putative NRD convertase cleavage site, was not cleaved by the enzyme (3). Two new, longer fluorogenic peptides, AbzGGFLRRVGQ-EDDnp and Abz-GGFLRRIQ-EDDnp, were synthesized as potential substrates. In these peptides the fluorescence of the N-terminal aminobenzoyl group is masked by the through-space-quenching effect of the dinitrophenyl moiety. Upon cleavage of any of the peptide bonds, fluorescence increases and thus can be used to measure peptidase activity in a continuously monitored, sensitive assay. The peptide sequence of Abz-GGFLRRVGQ-EDDnp is based on that of bovine adrenal medulla peptide, an excellent substrate of NRD convertase (3, 11), while the sequence in AbzGGFLRRIQ-EDDnp is based on dynorphin A, the substrate with the highest affinity for the enzyme (11). The maximal fluorescence change upon cleavage of the two substrates was obtained by complete cleavage with either trypsin or NRD convertase in 20 mM potassium phosphate buffer, pH 7.0. A total relative fluorescence increase of 7- to 10-fold over the starting value was observed upon the complete cleavage of either of these substrates (Fig. 1). Digestion of Abz-GGFLRRVGQ-EDDnp or Abz-GGFLRRIQ-EDDnp with NRD convertase yielded only two cleavage products from each of the peptide analogs. The cleavage products were separated and purified by reverse-phase HPLC. Mass spectrometry was used to identify Abz-GGFLR as a common product from
FIG. 1. Fluorescence spectra of Abz-GGFLRRVGQ-EDDnp and its cleavage products. Abz-GGFLRRVGQEDDnp (1 mM) in 400 ml of 20 mM potassium phosphate buffer, pH 7.0, was monitored by fluorescence before (b) and after (a) cleavage by NRD convertase. For reference, the absorbance spectrum of the peptide is included (c).
152
CSUHAI ET AL.
FIG. 2. MALDI-FT-ICR-MS spectrum of the two cleavage products of Abz-GGFLRRVGQ-EDDnp. (A) Peptide fragment Abz-GGFLR showing a protonated ion at m/z 668.350 (calcd 668.352) and sodiated ion at m/z 690.334 (calcd 690.334). Ions detected at m/z 678 – 681 represent an unidentified impurity. (B) Peptide fragment RVGQ-EDDnp showing a protonated ion at m/z 667.332 (calcd 667.328), protonated ion minus O at m/z 651.337 (calcd 651.333), protonated ion minus 2O and 2H at m/z 637.358 (calcd 637.353), and protonated ion minus 3O and 2H at m/z 621.363 (calcd 621.359). Peptides were measured in a 2,5-dihydroxy-benzoic acid matrix on a 4.7 T FTMS from IONspec with a N 2 laser.
FLUOROGENIC ASSAY WITH N-ARGININE DIBASIC CONVERTASE
FIG. 3. Dependence of the rate of cleavage on NRD convertase concentration. Abz-GGFLRRVGQ-EDDnp (1 mM) was incubated at 34°C in 20 mM potassium phosphate buffer, pH 7.0, in a 400 ml final volume. Emission was monitored at 419 nm with excitation set at 319 nm. Output was read from a strip chart recorder and rates were determined from the initial velocity of each reaction. NRD convertase was varied from 0.01 to 0.28 mg of protein.
both peptides and RVGQ-EDDnp and RIQ-EDDnp (not shown) were identified as the C-terminal products of Abz-GGFLRRVGQ-EDDnp and Abz-GGFLRRIQEDDnp, respectively (Fig. 2). Both of the peptide analogs were found to be excellent substrates for NRD convertase. The initial velocity of peptide cleavage, shown for Abz-GGFKRRVGQEDDnp, is linearly proportional to enzyme concentration (Fig. 3). We could measure a rate of cleavage of as low as 0.5 pmol/min and detect as little as 1 to 2 ng of
153
NRDc. The cleavage reaction obeyed classical Michaelis–Menten kinetics (Fig. 4A), yielding a K m value of 1.0 mM and a V max value of 1.3 mmol/min z mg for AbzGGFLRRIQ-EDDnp. Although not shown, a K m of 0.70 mM and a V max of 0.86 mmol/min z mg was obtained for Abz-GGFLRRVGQ-EDDnp. These values can be compared to K m of 6 mM and a V max of 3.6 mmol/min z mg for bovine adrenal medulla peptide (11) and K m values for a-neoendorphin and dynorphin B, two similar opioid peptides, of 3.6 and 0.6 mM, respectively. The observed K m values for the two new fluorogenic substrates thus fall within the range observed for peptides of similar size and sequence. FLRKI, the minimal peptide necessary for recognition and selective cleavage by NRD convertase (3), competitively inhibits the cleavage of Abz-GGFLRRVGQ-EDDnp (Fig. 4B). During the purification of NRD convertase through several chromatographic steps, peptidase activity, measured with Abz-GGFLRRVGQ-EDDnp, was directly proportional to enzyme activity measured by the more traditional assay of following the hydrolysis of bovine adrenal medulla peptide by HPLC. Thus, the fluorogenic peptide derivatives provide a convenient substrate to follow the enzyme during its purification. However, these substrates are not specific for NRDc. They will be cleaved by trypsin-like enzymes, and potentially by a variety of other peptidases. However, in the present case, NRDc was the predominant enzyme cleaving these substrates in rat testes homogenates. The new, rapid, and sensitive fluorogenic assay reported here should facilitate studies of the physiological functions of this novel peptidase.
FIG. 4. Kinetic studies with Abz-GGFLRRVGQ-EDDnp. (A) Michaelis–Menten kinetics of the cleavage of Abz-GGFLRRVGQ-EDDnp: Reaction mixtures contained 20 mM potassium phosphate buffer, pH 7.0, 0.1 mg of enzyme, and 0.25–20 mM Abz-GGFLRRVGQ-EDDnp. Activity was determined as described in the legend to Fig. 3. (B) Dixon plot showing inhibition of Abz-GGFLRRVGQ-EDDnp cleavage by a shorter peptide substrate of NRD convertase: Assay conditions were identical to those described above with Abz-GGFLRRVGQ-EDDnp as the fixed variable substrate at 0.5, 0.75, or 1.0 mM. The peptide FLRKI was varied as indicated.
154
CSUHAI ET AL.
ACKNOWLEDGMENTS This work was supported in Brazil by Fundac¸a˜o de Amparo Pesquisa do Estado de Sa˜o Paulo (FAPESP) and Conselho Nacional de Desenvolvimento Cientıf´ico e Tecnolo´gico (CNPq). This work was also supported in part by NIDA/NIH Grant DA02243. E.Cs. is a recipient of a NIDA/NIH Postdoctoral Fellowship DA05671.
REFERENCES 1. Chesneau, V., Pierotti, A. R., Barre´, N., Cre´minon, C., Tougard, C., and Cohen, P. (1994) J. Biol. Chem. 269, 2056 –2061. 2. Foulon, T., Cadel, S., Chesneau, V., Draoui, M., Prat, A., and Cohen, P. (1996) Ann. NY Acad. Sci. 780, 106 –120. 3. Csuhai, E., Safavi, A., and Hersh, L. B. (1995) Biochemistry 34, 12411–12419. 4. Pierotti, A. R., Prat, A., Chesneau, V., Gaudoux, F., Leseney, A.-M., Foulon, T., and Cohen, P. (1994) Proc. Natl. Acad. Sci. USA 91, 6078 – 6082. 5. Hospital, V., Prat, A., Joulie, C., Che´rif, D., Day, R., and Cohen, P. (1997) Biochem. J. 327, 773–779. 6. Fumagalli, P., Accarino, M., Egeo, A., Scartezzini, P., Rappazzo, G., Pizzuti, A., Avvantaggiato, V., Simeone, A., Arrigo, G., Zuffardi, O., Ottolenghi, S., and Taramelli, R. (1998) Genomics 47, 238 –245. 7. Draoui, M., Bellincampi, L., Hospital, V., Cadel, S., Foulon, T., Prat, A., Barre´, N., Reichert, U., Melino, G., and Cohen, P. (1997) J. NeuroOncol. 31, 99 –106.
8. Chesneau, V., Prat, A., Segretain, D., Hospital, V., Dupaix, A., Foulon, T., Je´gou, B., and Cohen, P. (1997) J. Cell Sci. 109, 2737–2745. 9. Hooper, N. M. (1994) FEBS Lett. 345, 1– 6. 10. Chesneau, V., Pierotti, A. R., Prat, A., Gaudoux, F., Foulon, T., and Cohen, P. (1994) Biochimie 76, 234 –240. 11. Csuhai, E., Chen, G., and Hersh, L. B. (1998) Biochemistry 37, 3787–3794. 12. Zhong, W., and Benkovic, S. J. (1998) Anal. Biochem. 255, 66 – 73. 13. Bonneau, P. R., Plouffe, C., Pelletier, A., Wernic, D., and Poupart, M.-A. (1998) Anal. Biochem. 255, 59 – 65. 14. Yaron, A., Carme, A., and Katchalsi-Katzir, E. (1979) Anal. Biochem. 95, 228 –235. 15. Chagas, J. R., Juliano, L., and Prado, E. (1991) Anal. Biochem. 192, 419 – 425. 16. Gershkovich, A. A., and Kholodovych, V. V. (1996) J. Biochem. Biophys. Methods 33, 135–162. 17. Y. Hirata, Cezari, M. H. S., Nakaie, C. R., Boschcov, P., Ito, A. S., Juliano, A. M., and Juliano, L. (1994) Lett. Peptide Sci. 1, 299 – 308. 18. Meldal, M., and Breddam, K. (1991) Anal. Biochem. 195, 141– 147. 19. Cohen, P., Pierotti, A. R., Chesneau, V., Foulon, T., and Prat, A. (1995) Methods Enzymol. 248, 703–716. 20. Cleland, W. W. (1979) Methods Enzymol. 63, 103–137.