Functional Screening of Serine Protease Inhibitors in the Medical ...

Research

Functional Screening of Serine Protease Inhibitors in the Medical Leech Hirudo medicinalis Monitored by Intensity Fading MALDI-TOF MS* Oscar Yanes‡, Josep Villanueva§, Enrique Querol, and Francesc X. Aviles¶ The blood-feeding invertebrates are a rich biological source of drugs and lead compounds to treat cardiovascular diseases because they have evolved highly efficient mechanisms to feed on their hosts by blocking blood coagulation. In this work, we focused our attention on the leech Hirudo medicinalis. We performed, by “intensity fading” MALDI-TOF mass spectrometry, a comprehensive detection and functional analysis of pre-existent peptides and small proteins with the capability of binding to trypsin-like proteases related to blood coagulation. Combining “intensity fading MS” and off-line LC prefractionation allowed us to detect more than 75 molecules present in the leech extract that interact specifically with a trypsinlike protease over a sample profile of nearly 2,000 different peptides/proteins in the 2–20-kDa range. Moreover we resolved 232 individual components from the complex mixture, 13 of which have high sequence homology with previously described serine protease inhibitors. Our findings indicate that such extracts are much more complex than expected. Additionally, intensity fading MS, when complemented with LC separation strategies, seems to be a useful tool to investigate complex biological samples, establishing a new bridge between profiling, functional peptidomics, and subsequent drug discovery. Molecular & Cellular Proteomics 4:1602–1613, 2005.

Although the task of identifying and characterizing genes in many sequenced genomes in the last years has been very intense, it is even more challenging when applied to the much higher complex field of the protein world (1). One difficult set of proteins for such analysis are those which we could term as “small” (i.e. below 15–20 kDa) (2) because of their variety and difficulty to be detected by several established proteomic approaches such as 2D1 electrophoresis (3) or computational genomic scanning (4) among others. Therefore, the developFrom the Institut de Biotecnologia i de Biomedicina and Departament de Bioquı´mica, Universitat Auto`noma de Barcelona, 08193 Bellaterra (Barcelona), Spain Received, May 18, 2005, and in revised form, July 5, 2005 Published, MCP Papers in Press, July 18, 2005, DOI 10.1074/ mcp.M500145-MCP200 1 The abbreviations used are: 2D, two-dimensional; RP, reversedphase.

1602

Molecular & Cellular Proteomics 4.10

ment of efficient and high throughput technologies for producing functionally related data of peptides and small proteins (2–20 kDa) is increasingly attracting attention. Some of these proteins have potential use for diagnostics or therapeutics (5–7). They include families of peptide hormones, neuropeptides, cytokines, growth factors, and enzyme inhibitors acting as biochemical messengers or modulators that organize the regulatory processes in all organisms (6). In fact, most of the important biopharmaceutical products approved for therapeutic applications are molecules within a molecular mass range of 1– 40 kDa (8). The proteomic strategy focused on the analysis of peptides and small proteins from complex mixtures has been named “peptidomics” (9, 10). The primary proteomic tool, 2D gel electrophoresis with mass spectrometry, has been gradually enhanced, and sometimes replaced, by three main technologies: MALDI MS in combination with LC (off line) (9, 11, 16), ESI MS combined with nano-LC (on line) (12–16), and MALDITOF MS for in situ peptide profiling (17–21). One of the main goals of these technologies is to provide a comprehensive analysis of molecular masses, concentration levels, and molecular structures as well as interactions established by peptides and small proteins in complex biological mixtures. The detection of the interactions between biomolecules and their targets provides information on the potential biological activity of these compounds (22–24) and ultimately shortcuts for the development of new drugs for the pharmaceutical industry (7, 25–27). In this context, we recently reported a new approach to detect non-covalent interactions between molecules by direct spectral perturbation, which we called “intensity fading” MALDI-TOF mass spectrometry (28, 29). It is based on the analysis of the relative intensities derived from the MALDI ions (m/z) to study the formation of complexes between biomolecules. Complexes are detected through the decrease (fading) of the relative intensities of the m/z signal corresponding to a ligand or mixture of potential ligands (i.e. a biological extract or a peptide library) after the addition of the target molecule (i.e. a protein). One of our present goals is to apply this procedure to a series of complex biological systems to validate the procedure and expand its range of applicability to high throughput proteomic strategies. Leeches have been a useful biological source of drugs to

© 2005 by The American Society for Biochemistry and Molecular Biology, Inc. This paper is available on line at http://www.mcponline.org

Functional Screening of Trypsin Inhibitors from Leech

treat cardiovascular diseases or as lead compounds in this field because they have evolved highly specific mechanisms to feed on their hosts by blocking blood coagulation (30). A series of interesting molecules, mainly protease inhibitors, acting at different points in the coagulation cascade or in the inhibition of platelet aggregation have been purified from these organisms (31–33). In this work we describe the application of the intensity fading MS approach to screen protein inhibitors of serine proteases in the saliva of the leech Hirudo medicinalis. We modified our original version of the approach (28) by immobilization of the protein target on microbeads and incorporating prefractionation steps, which strongly increased the analytical power. This combination allowed the detection of more than 75 molecules among a complexity of nearly 2,000 molecular species of different molecular mass, in the 2–20-kDa range, that interact specifically with a trypsin-like protease. Moreover we resolved 232 individual components from the complex mixture, showing that 16 of them are new putative trypsin-like inhibitors displaying inhibitory activity against trypsin and in most cases high sequence homology with other serine protease inhibitors. EXPERIMENTAL PROCEDURES

Materials Sinapic acid was obtained from Fluka. Extract from H. medicinalis was supplied by the group of Profs. H. Fritz and C. Sommerhoff (Chirurgischen Klinik Innenstadt, Ludwig-Maximilians-Universitat, Munich, Germany). Trypsin (proteomic sequencing grade), anhydrotrypsin-agarose, bovine trypsin, and the synthetic trypsin substrate N-benzoyl-L-arginine ethyl ester hydrochloride were purchased from Sigma. Paraffin wax film (Parafilm) was purchased from Pechiney Plastic Packaging, Inc.

Pretreatments of H. medicinalis Extract For direct and fractionated mass profiling analysis by MALDI-TOF MS, the extracts of H. medicinalis were either cleaned up and concentrated by a simple reversed-phase C18 resin or fractionated by reversed-phase HPLC, respectively. C18 Resin Protocol to Clean Up and Concentrate Proteins—Lyophilized H. medicinalis extract was dissolved in deionized water at a concentration of 20 mg/ml. The insoluble residues were removed by centrifugation at 8,000 ⫻ g for 10 min. The supernatant was processed by a reversed-phase C18 resin-based protocol to clean and concentrate peptides and small proteins (34). The reversed-phase resin-bound molecules were eluted with 50% isopropanol, lyophilized, and resuspended in 30 ␮l of 10 mM Tris-HCl, pH 7.5. Reversed-phase HPLC Fractionation of the Entire Extract—Lyophilized leech extract (⬃2 mg) was dissolved and subjected to reversedphase HPLC on a Vydac C18 column using a linear gradient from 10 to 50% acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 0.5 ml/min for 60 min. Fifty fractions were collected, lyophilized, and resuspended in 30 ␮l of 10 mM Tris-HCl, pH 7.5.

Enzymatic Measurements for the Detection of Trypsin Inhibitory Activity The general assay for trypsin inhibition was carried out by preincubation of 10 ␮l of sample (from each of the lyophilized and resus-

pended reversed-phase (RP)-HPLC fractions of the leech extract and single purified trypsin ligands) with 50 ␮l of bovine trypsin (33 ␮g/ml in 20 mM CaCl2, 0.1 M Tris, pH 8.0) and 790 ␮l of activity buffer (20 mM CaCl2, 0.1 M Tris, pH 8.0) for 3 min at room temperature followed by the addition of 150 ␮l of the substrate N-benzoyl-L-arginine ethyl ester hydrochloride (3 mM) and measurement of the absorbance change at 254 nm. The RP-HPLC fractionation of the extract and inhibitory activity assay of each fraction were performed in triplicate ensuring that the performance and the results reported here were reproducible.

Interaction Experiments Initial experiments involved the use of a proteomic sequencing grade trypsin, a chemically modified form resistant to autolysis. However, due to the high concentration of trypsin used in the assay (complexes between trypsin and proteic trypsin inhibitors are 1:1, w/w), residual autolytic reactions and proteolysis of the sample altered the spectra, complicating the interpretation of the results. Anhydrotrypsin, a chemically modified form of bovine trypsin with no detectable catalytic activity but with a strong affinity toward trypsin inhibitors (35, 36) was used throughout this work to avoid artifacts such as the above mentioned. Preparation of Anhydrotrypsin-Agarose—The buffered aqueous suspension (50 mM sodium acetate, pH 5.0, containing 20 mM CaCl2 and 0.02% sodium azide) of the anhydrotrypsin-agarose was eliminated by centrifugation and replaced with 10 mM Tris-HCl, pH 7.5. Interaction of Anhydrotrypsin-Agarose with the Reversed-phase HPLC Fractions of the Extract—1 ␮l of the dissolved reversed-phase HPLC fractions (see above) was mixed with 1–2 ␮l of anhydrotrypsinagarose and incubated for 3 min at room temperature on a small piece of Parafilm. In the control sample, the protease was replaced by a neutral buffer (10 mM Tris-HCl, pH 7.5). Recovery of Ligands—Given the hydrophobic nature of the Parafilm surface, the reacting drop (2–3 ␮l) exhibited an accentuated surface tension that allowed pipetting 0.5 ␮l from the top of the drop and recovering unbound molecules for further MALDI MS analysis. Next the anhydrotrypsin-agarose was washed three times with 5 ␮l of 10 mM Tris-HCl, pH 7.5, by pipetting the agarose up/down 5–10 times and subsequent elimination of the Tris-HCl solution by capillarity with a thin absorbent paper. Finally 2 ␮l of 0.1% formic acid were mixed with anhydrotrypsin-agarose, and after 3 min 0.5 ␮l of the drop was pipetted to analyze by MALDI MS those molecules initially bound to the target protein.

Sample Preparation for MALDI-TOF Mass Spectrometry Analysis Both mass profiling and intensity fading MS experiments were analyzed in the same way. The sample was mixed with a matrix solution (1:2, v/v) of sinapic acid (10 mg/ml) containing 30% (v/v) acetonitrile diluted in deionized water (pH 3). 0.5 ␮l of the mixture was deposited on the MALDI target using the dried droplet method (37).

MALDI-TOF Mass Spectrometry Mass profiles and intensity fading MS interaction experiments were analyzed with an Ultraflex MALDI-TOF mass spectrometer (Bruker, Bremen, Germany) equipped with a 337 nm laser, a gridless ion source, delayed extraction electronics, a high resolution timed ion selector, and a 2-GHz digitizer. For direct and fractionated mass profiling, separate spectra were obtained for two restricted m/z ranges, corresponding to peptides and small proteins with molecular mass of 2– 8 and 8 –20 kDa, under specifically optimized instruments settings. Each spectrum was the result of 500 laser shots per m/z segment per sample delivered in 10 sets of 50 shots distributed in


1603


three different locations on the surface of the matrix spot. Spectra were acquired in linear mode geometry under 20 kV and with timed ion selector deflection of mass ions ⱕ1.500 m/z (2– 8-kDa segment) or ⱕ6.000 m/z (8 –20-kDa segment). Delayed extraction was maintained for 100 (2– 8 kDa) or 330 ns (8 –20 kDa) to give appropriate time lag focusing after each laser shot. Some peptides were consecutively eluted in two or more adjacent RP-HPLC fractions. A script was designed to eliminate those repeated m/z ions of adjacent fractions. External calibration was implemented to obtain more accurate m/z values of all the components analyzed.

Multidimensional Liquid Chromatography (2D HPLC) Approximately 0.1 g of lyophilized leech extract was dissolved in 20 mM Tris acetate buffer, pH 8.0, and insoluble residues were removed by centrifugation at 13,000 ⫻ g for 10 min. After pH equilibration, the supernatant was loaded onto a preparative anion-exchange column (TSK-DEAE 5PW, 2.5 ⫻ 15 cm; Toyo-Soda) connected to an automated liquid chromatography system (⌬KTA purifier 10/100, Amersham Biosciences). Elution was performed using a linear gradient from 2 to 100% 0.8 M ammonium acetate in 20 mM Tris acetate at a flow rate of 2 ml/min for 80 min. Thirty fractions of 5 ml were collected, lyophilized, and subjected to reversed-phase HPLC on a Vydac C18 column using a linear gradient from 15 to 50% acetonitrile in 0.1% trifluoroacetic acid at a flow rate of 0.5 ml/min for 60 min. Peaks were manually collected, and the determination of the molecular mass (m/z) and purity of the isolated molecules from each reversed-phase chromatographic peak were checked by MALDI-TOF MS.

Preparation of Samples for Edman Degradation Liquid samples containing acetonitrile and 0.1% trifluoroacetic acid were directly absorbed onto micro TFA filters (Applied Biosystems, Foster City, CA). The isolated small protein inhibitors were analyzed by automated amino acid sequencing on a Procise protein sequencing system (Applied Biosystems). RESULTS

Direct and Fractionated Mass Profiling of H. medicinalis by MALDI-TOF MS—The analytical potential of the intensity fading MS approach to screen complex biological mixtures to identify new ligands for a specific target protein depends on the complexity of the molecular ion profile detected in the control mass spectrum. The more ionized molecules displayed, the better the peptide and small protein representation of the biological sample for further detection of biomolecular interactions. When a direct sample mass profiling of a salivary gland extract of the leech H. medicinalis was performed by MALDITOF mass spectrometry, only 146 ion signals appeared in the spectrum (Fig. 1A) probably due to suppression effects (38 – 40). Signal suppression may be caused by abundant or dominant species in a mixture that suppress the ionization and detection of the less abundant ones. To determine whether such negative effects take place and to try to mitigate them, a previous fractionation of the extract was carried out. The entire extract of H. medicinalis was fractionated by means of RP-HPLC. This is a favorable technique for resolving low molecular weight proteins and peptides. Most of the serine

1604


protease inhibitors described until now are peptides and small extracellular proteins with compact structures, rich in disulfide bonds (i.e. aprotinin and hirudin), and resistant to the harsh conditions (high amount of acetonitrile and low pH) of RPHPLC. It is worth remembering that we cannot ensure that inhibitory activity of all serine protease inhibitors present in nature are fully maintained after RP-HPLC and lyophilization (especially those not stabilized by disulfide bridges). In our case, only 50 HPLC fractions were collected and subsequently subjected to off-line MALDI-TOF MS analysis. Such analysis revealed a high complexity of ion signals (m/z) for many of the fractions that increased from the previous 146 ionized species to 1,953 in the range of 2–20 kDa. To illustrate this complexity, Fig. 1B shows the 50 MALDI-TOF mass spectra derived from the 50 RP-HPLC fractions in the 2– 8- and 8 –20-kDa ranges. To represent the differences in complexity between direct and fractionated mass profiling, the molecular weight range distribution and the quantitation of all these peptides and small proteins (after computational discrimination to delete co-eluted compounds) have been plotted in Fig. 1C. Most of the ionized molecules are peptides in the range of 2– 8 kDa. On the other hand, RP-HPLC fractionation allowed the detection of 164 small proteins in the 10 –20-kDa upper range not previously displayed with direct mass profiling by MALDI-TOF MS. Functional Screening of Serine Protease Inhibitors by Intensity Fading MALDI-TOF MS—Given the large complexity of the sample from H. medicinalis, its analysis was simplified by focusing on those species with the capability to inhibit (and therefore bind) serine proteases. All 50 RP-HPLC chromatographic fractions were submitted to an inhibitory activity assay against trypsin using a classical but sensitive spectrophotometric assay (41) (results not shown). Nineteen fractions exhibited inhibitory activity and were subsequent subjected to intensity fading MS. With this step, the potential trypsin-ligand population was reduced from nearly 2,000 to 750 species approximately. Fig. 2 shows this analytical process, indicating the level of inhibition detected for each chromatographic fraction and the subsequent mass spectra of three selected fractions of the extract (used as examples) before and after the addition of anhydrotrypsin (bound to microbeads). Some ion signals are greatly diminished in the mass spectra after the addition of the target enzyme. Given that the catalytic (but not binding) capability of the protease was eliminated by previous conversion in its anhydro derivative, the few ion signals (m/z) affected by the addition of the enzyme should correspond to molecules that specifically bind to the target protein. Once it had been shown that some of the ion signals of the spectra were faded and given that the protease was immobilized on agarose (see “Experimental Procedures”), an additional step of the analytical assay was performed to assess the specificity of the assay. Anhydrotrypsin-agarose was washed with neutral buffer after the binding reaction to discard unbound molecules followed by lowering the pH to re-


FIG. 1. Mass profiling by MALDI-TOF MS of the salivary glands of the leech H. medicinalis. A, MALDI mass spectra derived from the analysis of a concentrated and cleaned-up sample of the leech extract. B, MALDI mass spectra derived from the reversed-phase HPLC fractionation of the leech extract. C, plot showing the molecular mass distribution of all peptides and small proteins displayed with both direct and fractionated mass profiling. Co-eluted peptides and small proteins were deleted to quantify the sample complexity.


1605


FIG. 2. A, reversed-phase HPLC chromatogram of the entire extract of the leech H. medicinalis. Shaded areas indicate regions displaying inhibitory activity against trypsin. B, plot representing the slope of the spectrophotometric inhibitory assay for each one of the 50 reversedphase HPLC fractions. C, MALDI-TOF mass spectra of three different reversed-phase HPLC fractions from the leech extract showing inhibitory activity before (upper) and after (lower) the addition of the trypsin-like protease. Faded m/z ions are indicated by dotted lines. Abs, absorbance.

1606



FIG. 3. MALDI-TOF mass spectra of four different reversed-phase HPLC fractions from the leech extract showing inhibitory activity before (upper) and after (middle) the addition of the trypsin-like protease. The recovery of the m/z signal corresponding to trypsin inhibitors, after acidification of the sample, is shown (lower).

lease species bound to the protease (see “Experimental Procedures”). Fig. 3 shows the results of the intensity fading MS assay including this additional step for four of the RP-HPLC fractions. By breaking the interactions of the protease with the putative binding molecules, the faded signals after the addition of the protease appear again in the MS spectra. Therefore, these signals are attributed to molecules that specifically bind to the protease. A triplicate screening of all the chromatographic fractions showing inhibitory activity against trypsin was performed fol-

lowing a two-step strategy. First, immobilized protease was added to check which molecules of the biological extract were bound to the protease followed by an acidification of the mixture. Only the ion signals presenting a clear mass spectroscopy difference were considered for further experiments, that is, those that showed a fading or disappearance of its relative intensity signal after the addition of the enzyme and a reappearance after the acid treatment of the mixture in each triplicate assay. This functional screening gave rise to a list of more than 75 m/z ions of putative serine protease inhibitors


1607


TABLE I Molecular weights (m/z) of peptides and small proteins that specifically interacts with the trypsin-like protease by intensity fading MS

present in the H. medicinalis extract (Table I). Further analysis allowed characterization of some them (see below). Multidimensional Liquid Chromatography for the Separation and Identification of Novel Serine Protease Inhibitors—A strategy of combining 2D HPLC and MALDI-TOF MS off line was followed using the crude H. medicinalis extract to isolate (in sufficient quantity), identify, and characterize the molecules with molecular masses present in Table I. For the 2D HPLC approach we selected two modes characterized by different separation selectivities: coupling anion-exchange and reversed-phase chromatography as the first and second dimension, respectively. Fraction collection of the anion-exchange chromatography followed by the injection of the individual fractions onto the second dimension column resulted in about 690 reversed-phase chromatographic peaks. Fig. 4, A and B, shows this preparative process, displaying four examples of fractions from reversed-phase chromatography from which we isolated five new serine protease inhibitors. Analyzing by MALDI-TOF MS each of the RP-HPLC peaks allowed us to visualize 232 isolated single components in the range of 2–20 kDa. The remainder of the peaks showed two or more coeluted components. The resolved single molecules with m/z shown in Table I were subjected to automated Edman N-terminal sequencing analysis (without fragmentation), giving rise to an array of partial sequences (see Table II). After a database search using such partial sequences, clear homology relationships with other serine protease inhibitors from leeches or other parasites for 12 of them were obtained. No homology was found with any other sequences in the databases for four of them (see Table II). We also were able to isolate and identify some previously described serine protease inhibitors from H. medicinalis such as hirustasin (5879 m/z) or bdellin A (6333 m/z) (data not shown). Validating the Approach—If the small proteins identified by intensity fading MS, purified, and partially sequenced are trypsin inhibitors, they should display inhibitory activity. To ensure that the molecules isolated and identified in this work are actually trypsin inhibitors, classical enzymology assays were performed with the purified molecules. Fig. 4C shows the inhibition data of the RP-HPLC-purified forms, which were later used for identification by Edman degradation. Within the 17 partial sequences shown in Table II, we detected trypsin inhibitory activity in 16 of them. The one remaining, with no detectable inhibitory activity, showed a molecular mass of 7,765 Da and 100% sequence similarity with hemoglobin ␣-1

1608


chain after partial sequencing using Edman degradation. We checked the sequence of this protein and verified the presence of an Arg in the C terminus, suggesting that the detected binding was not a consequence of its inhibitory activity but as a consequence of the binding characteristics of anhydrotrypsin with the Arg residue in the C terminus. So far, this is the only one false positive detected using this approach. Homology Relationships among the Novel Protease Inhibitors Isolated—Most of the new serine protease inhibitors found in this work can be located in different families of the leech trypsin inhibitors. However, others show homology relationships with other families. For instance, the inhibitor with a molecular mass of 3,128 Da displays a well conserved region (7DENTPCP13) showing 100% sequence identity with chymotrypsin/elastase isoinhibitor 1 (42) from Ascaris lumbricoides. In addition, two couples of protein inhibitors with molecular masses of 6,218/6,227 and 5,463/5,325 Da show identical sequences within the first 49 and 19 amino acids from the N terminus, respectively, suggesting degradation of the proteins in the extract or the presence of isoforms. Also the protein with a molecular mass of 7,900 Da showed 100% sequence identity with Eglin C (43), but a difference of 199 Da in its molecular mass suggests that it could be a degraded form of Eglin C or just a new isoform not described previously. Some of the isolated putative serine protease inhibitors of this work display a clear similarity with other inhibitors of the antistasin family previously found in leeches, such as guamerin (44, 45), piguamerin (46), bdellin A (47), and hirustasin (48) (see Table II). In Fig. 5, the amino acid sequences of five inhibitors displaying homology with these four inhibitors of the antistasin family are compared, and amino acids that are conserved in the nine proteins are boxed. The region corresponding to the “reactive center” in this type of protease inhibitors is located in an exposed binding loop (49, 50) clearly sequenced in this work. The homology there is reduced quite significantly (except for a cysteine residue) and breaks down abruptly at the boundaries. Nowhere else in the molecule can a stretch of comparable length approaching this degree of divergence be found. Finally the N-terminal sequences of molecules with 3,038, 4,014, 5,448, and 4,083 Da revealed no similarities with any other sequences in protein sequence databases, although they showed trypsin inhibitory activity. DISCUSSION

Increasing attention is paid nowadays to peptidomic research because of its potential use diagnostically or thera-


FIG. 4. A, anion-exchange chromatogram of the entire extract of the leech H. medicinalis indicating the 30 fractions collected at the preparative level. B, reversed-phase HPLC chromatograms derived from four selected anion-exchange chromatographic fractions. Shaded peaks correspond to four of the isolated single inhibitors. C, representation of the slope derived from the spectrophotometric inhibitory assay for each one of the single isolated forms of the trypsin inhibitors assayed prior to the Edman degradation. Abs, absorbance.


1609


TABLE II Some of the new small proteins and peptides identified as trypsin ligands in the leech H. medicinalis by intensity fading MS

A indicates no sequence homology found with any other protease inhibitor.

FIG. 5. Comparison of the protein sequences for five fragments of trypsin inhibitors identified in this work (denoted by numbers referring to their masses) and four previously described leech inhibitors. Identical amino acids for five of the nine proteins are boxed. A solid line indicates the reactive center region. The P1 residue of guamerin, piguamerin, hirustasin, and bdellin A is marked with an arrow.

peutically. In this sense, the leech is an organism with biomedical applications that generate outstanding interest. Several anticoagulants are present in the salivary glands of leeches because these organisms depend on a diet of fresh blood and have evolved mechanisms that interfere with the coagulation of the blood “donor.” These bioactive molecules

1610


are mainly small protease inhibitors of serine proteases involved in the coagulation cascade (32). Peptidomics, like proteomics, has three subareas that are currently attracting very active research: profiling, functional, and structural analysis. In this work, we tried to join and complement the former two subareas, that is, profiling and


functional aspects, through the use of the intensity fading MS approach. We combined LC-MS off line by means of (onedimensional) RP or (2D) ion-exchange RP-HPLC and MALDITOF MS to display the H. medicinalis complexity. In addition we isolated some of its components and semicharacterized them by Edman sequencing. Complementarily the application of the intensity fading MS methodology for screening trypsinlike inhibitors among the displayed population of peptides and small proteins detected more than 75 different proteins showing binding properties for a trypsin-like protease with molecular masses ranging from 2,000 to 20,000 Da. This is the most interesting mass range for our present purposes and also because this is the spectral range in which the present average MALDI-TOF spectrometer performs its best for peptides and proteins. The (one-dimensional) RP-HPLC and MALDI-TOF combined strategy substantially mitigated sample suppression effects in mass spectrometry and allowed the visualization of more than 10-fold ionized species from the H. medicinalis peptidome with respect to direct mass profiling strategy. However, separation based on one single parameter like reversed-phase HPLC does not provide the required resolution and capacity to resolve individual components of an extremely complex biological mixture like the H. medicinalis extract. For this reason, in a second step the strategy was improved by using off-line 2D LC based on a strong anionexchange column as a first dimension and reversed-phase liquid chromatography as the second separation. Despite the 2D HPLC strategy, it was not feasible to resolve the nearly 2,000 individual components detected by mass spectrometry, and some components were co-eluted in the second dimension. The use of immobilized anhydrotrypsin to generate spectral perturbations following the intensity fading MS assay allowed us to overcome the problems derived from the autolytic and catalytic activity toward possible substrates of trypsin and minimized potential contaminations from the added target; it also permitted us to have additional control over the putative inhibitors through their isolation and by monitoring their complex dissociation. One of the limitations of the approach is the fact that anhydrotrypsin can selectively bind peptides with Arg, Lys, or S-aminoethylcysteine residues at the C terminus, resulting in some possible false positives. To get around this limitation, a complementary spectrophotometric assay on the purified species (see “Experimental Procedures”) ensured their trypsin inhibitory activity. Some of the isolated serine protease inhibitors of this work display a clear similarity among them and with several trypsin inhibitors previously reported from leeches, such as guamerin, piguamerin, bdellin A, and hirustasin (see Table II) (44 – 48). These four serine protease inhibitors belong to the antistasin family, and they were isolated from H. medicinalis and Hirudo nipponia. Despite having a high similarity, guamerin, piguamerin, bdellin A, and hirustasin have different protease specificities (see Table II). Based on these data, it is quite

possible that the trypsin inhibitors isolated in this work show more than one specificity for different serine proteases like thrombin, plasmin, kallikrein, chymotrypsin, acrosin, or cathepsin G. The difference in specificity between these four proteins has been attributed to the high degree of non-homology within the short stretch of amino acids of the reactive center region and particularly to the resulting change in the P1 amino acid (32). As initially postulated by Hill and Hastie (51) for the serpin family, we suggest that all these small serine protease inhibitors pertaining to the antistasin family are an example of accelerated evolution in the reactive center region of serine protease inhibitors (see Fig. 5) (52). The protein sequence of the reactive center region has undergone much more rapid evolution than the rest of the molecule as a consequence of the selective forces proceeding from extrinsic proteases of the host. Because leeches are exclusively fed by blood of poikilothermic and homeothermic animals, fast adaptation to this kind of nutrition has resulted in accelerated evolutionary selection and fixation of very specific interactions between proteinase inhibitors of the leech and proteases of the host to prevent clotting of the sucked blood. Given the specific recognition by proteases of defined amino acid sequences, it may be possible to inhibit enzymes involved in pathological processes. Potent inhibitors have the potential to be developed as new therapeutic agents. In the last decade, it became obvious that invertebrates have been shown to be truly useful models in drug discovery for many cardiovascular, tumoral, and inflammatory diseases. Leeches provide a source of new candidate molecules for drug discovery, especially serine protease inhibitors, that can be exploited by the medical industry, i.e. to treat emphysema, coagulation, inflammation, dermatitis, and cancer (30 –32, 53). In conclusion, the array of new serine protease inhibitors identified in this work together with those previously described from the leech H. medicinalis confirms its potentiality as a biological source of lead compounds for the medical industry, especially for those diseases in which serine protease are involved. Our good results confirm the intensity fading MS approach as a new robust strategy for the functional screening of complex biological samples, particularly the ones containing proteases and inhibitors of interest. The approach takes advantage of the low cost and rapid performance of MALDI MS technology together with the capability to check affinity properties of the analyzed compound. Acknowledgments—We are indebted to Profs. H. Fritz and C. Sommerhoff (Chirurgical Clinic, Munich, Germany) for providing us the leech extract. We thank F. Canals for helpful discussions. * This work was supported in part by Ministerio de Ciencia y Tecnologı´a, Spain (MCYT) Grants BIO2004-05879 and GEN2003-20642C09-05 and by the Centre de Refere`ncia en Biotecnologia (CERBA) de la Generalitat de Catalunya. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.


1611


‡ Supported by a fellowship from MCYT. § To whom correspondence may be addressed: Memorial SloanKettering Cancer Center, 1275 York Ave., New York, NY 10021. Tel.: 212-639-6676; Fax: 212-717-3604; E-mail: [email protected]. ¶ To whom correspondence may be addressed: Institut de Biotecnologia i de Biomedicina, Universitat Auto`noma de Barcelona, 08193 Bellaterra (Barcelona), Spain. Tel.: 34-93-581-1315; Fax: 34-93-5812011; E-mail: [email protected]. REFERENCES 1. de Hoog, C. L., and Mann, M. (2004) Proteomics. Annu. Rev. Genomics Hum. Genet. 5, 267–293 2. Baggerman, G., Verleyen, P., Clynen, E., Huybrechts, J., De Loof, A., and Schoofs, L. (2004) Peptidomics. J. Chromatogr. B Anal. Technol. Biomed. Life Sci. 15, 3–16 3. Rill, R. L., and Al-Sayah, M. A. (2004) Peptide separations by slab gel electrophoresis in pluronic F127 polymer liquid crystals. Electrophoresis 25, 1249 –1254 4. Southan, C. (2004) Has the yo-yo stopped?—an assessment of human protein-coding gene number. Proteomics 4, 1712–1726 5. Clynen, E., De Loof, A., and Schoofs, L. (2003) The use of peptidomics in endocrine research. Gen. Comp. Endocrinol. 132, 1–9 6. Schrader, M., and Schulz-Knappe, P. (2001) Peptidomics technologies for human body fluids. Trends Biotechnol. 19, Suppl. 10, S55–S60 7. Adermann, K., John, H., Standker, L., and Forssmann, W. G. (2004) Exploiting natural peptide diversity: novel research tools and drug leads. Curr. Opin. Biotechnol. 15, 599 – 606 8. Walsh, G. (1998) Biopharmaceuticals: Biochemistry and Biotechnology, John Wiley & Sons, New York 9. Schulz-Knappe, P., Zucht, H. D., Heine, G., Jurgens, M., Hess, R., and Schrader, M. (2001) Peptidomics: the comprehensive analysis of peptides in complex biological mixtures. Comb. Chem. High Throughput Screen. 4, 207–217 10. Jurgens, M., and Schrader, M. (2002) Peptidomic approaches in proteomic research. Curr. Opin. Mol. Ther. 4, 236 –241 11. Audsley, N., and Weaver, R. J. (2003) A comparison of the neuropeptides from the retrocerebral complex of adult male and female Manduca sexta using MALDI-TOF mass spectrometry. Regul. Pept. 116, 127–137 12. Baggerman, G., Cerstiaens, A., De Loof, A., and Schoofs, L. (2002) Peptidomics of the larval Drosophila melanogaster central nervous system. J. Biol. Chem. 277, 40368 – 40374 13. Skold, K., Svensson, M., Kaplan, A., Bjorkesten, L., Astrom, J., and Andren, P. E. (2002) A neuroproteomic approach to targeting neuropeptides in the brain. Proteomics 2, 447– 454 14. Heine, G., Raida, M., and Forssmann, W. G. (1997) Mapping of peptides and protein fragments in human urine using liquid chromatography mass spectrometry. J. Chromatogr. A 776, 117–124 15. Baggerman, G., Clynen, E., Huybrechts, J., Verleyen, P., Clerens, S., De Loof, A., and Schoofs, L. (2003) Peptide profiling of a single Locusta migratoria corpus cardiacum by nano-LC tandem mass spectrometry. Peptides 24, 1475–1485 16. Huybrechts, J., De Loof, A., and Schoofs, L. (2004) Diapausing Colorado potato beetles are devoid of short neuropeptide F I and II. Biochem. Biophys. Res. Commun. 317, 909 –916 17. Rubakhin, S. S., Garden, R. W., Fuller, R. R., and Sweedler, J. V. (2000) Measuring the peptides in individual organelles with mass spectrometry. Nat. Biotechnol. 18, 172–175 18. Fenselau, C., and Demirev, P. A. (2001) Characterization of intact microorganisms by MALDI mass spectrometry. Mass Spectrom. Rev. 20, 157–171 19. Jimenez, C. R., Li, K. W., Dreisewerd, K., Spijker, S., Kingston, R., Bateman, R. H., Burlingame, A. L., Smit, A. B., van Minnen, J., and Geraerts, W. P. (1998) Direct mass spectrometric peptide profiling and sequencing of single neurons reveals differential peptide patterns in a small neuronal network. Biochemistry 37, 2070 –2076 20. Jimenez, C. R., Li, K. W., Dreisewerd, K., Mansvelder, H. D., Brussaard, A. B., Reinhold, B. B., Van der Schors, R. C., Karas, M., Hillenkamp, F., Burbach, J. P., Costello, C. E., and Geraerts, W. P. (1997) Pattern changes of pituitary peptides in rat after salt-loading as detected by means of direct, semiquantitative mass spectrometric profiling. Proc.

1612


Natl. Acad. Sci. U. S. A. 94, 9481–9486 21. Li, L. J., Garden, R. W., and Sweedler, J. V. (2000) Single-cell MALDI: a new tool for direct peptide profiling. Trends Biotechnol. 18, 151–160 22. Figeys, D. (2002) Functional proteomics: mapping protein-protein interactions and pathways. Curr. Opin. Mol. Ther. 4, 210 –215 23. Whitelegge, J. P., and le Coutre, J. (2001) Proteomics: making sense of genomic information for drug discovery. Am. J. Pharmacogenomics 1, 29 –35 24. Pillutla, R. C., Goldstein, N. I., Blume, A. J., and Fisher, P. B. (2002) Target validation and drug discovery using genomic and protein-protein interaction technologies. Expert Opin. Ther. Targets 6, 517–531 25. Chanda, S. K., and Caldwell, J. S. (2003) Fulfilling the promise: drug discovery in the post-genomic era. Drug Discov. Today 8, 168 –174 26. Serebriiskii, I. G., Mitina, O., Pugacheva, E. N., Benevolenskaya, E., Kotova, E., Toby, G. G., Khazak, V., Kaelin, W. G., Chernoff, J., and Golemis, E. A. (2002) Detection of peptides, proteins, and drugs that selectively interact with protein targets. Genome Res. 12, 1785–1791 27. Figeys, D. (2002) Proteomics approaches in drug discovery. Anal. Chem. 74, 413A– 419A 28. Villanueva, J., Yanes, O., Querol, E., Serrano, L., and Aviles, F. X. (2003) Identification of protein ligands in complex biological samples using intensity-fading MALDI-TOF mass spectrometry. Anal. Chem. 75, 3385–3395 29. Yanes, O., Villanueva, J., Querol, E., and Aviles, F. X. (2004) Intensity-fading MALDI-TOF-MS: novel screening for ligand binding and drug discovery. Drug Discov. Today: TARGETS 3, Suppl. 1, 23–30 30. de Eguileor, M., Tettamanti, G., Grimaldi, A., Congiu, T., Ferrarese, R., Perletti, G., Valvassori, R., and Cooper, E. L. (2003) Leeches: immune response, angiogenesis and biomedical applications. Curr. Pharm. Des. 9, 133–147 31. Salzet, M. (2002) Leech thrombin inhibitors. Curr. Pharm. Des. 8, 493–503 32. Schoofs, L., Clynen, E., and Salzet, M. (2002) Trypsin and chymotrypsin inhibitors in insects and gut leeches. Curr. Pharm. Des. 8, 483– 491 33. Salzet, M. (2001) Anticoagulants and inhibitors of platelet aggregation derived from leeches. FEBS Lett. 492, 187–192 34. Villanueva, J., Canals, F., Querol, E., and Aviles, F. X. (2001) Monitoring the expression and purification of recombinant proteins by MALDI-TOF mass spectrometry. Enzyme Microb. Technol. 5, 99 –103 35. Ako, H., Foster, R. J., and Ryan, C. A. (1974) Mechanism of action of naturally occurring proteinase-inhibitors—studies with anhydrotrypsin and anhydrochymotrypsin purified by affinity chromatography. Biochemistry. 13, 132–139 36. Feinstein, G., and Feeney, R. E. (1966) Interaction of inactive derivatives of chymotrypsin and trypsin with protein inhibitors. J. Biol. Chem. 241, 5183–5189 37. Hillenkamp, F., and Karas, M. (1988) Laser desorption ionization of proteins with molecular masses exceeding 10000 daltons. Anal. Chem. 60, 2299 –2301 38. Karas, M., and Kruger, R. (2003) Ion formation in MALDI: the cluster ionization mechanism. Chem. Rev. 103, 427– 440 39. Knochenmuss, R., Stortelder, A., Breuker, K., and Zenobi, R. (2000) Secondary ion-molecule reactions in matrix-assisted laser desorption/ionization. J. Mass Spectrom. 35, 1237–1245 40. Knochenmuss, R., and Zenobi, R. (2003) MALDI ionization: the role of in-plume processes. Chem. Rev. 103, 441– 452 41. Reverter, D., Vendrell, J., Canals, F., Horstmann, J., Avile´s, F. X., Fritz, H., and Sommerhoff, C. P. (1998) A carboxypeptidase inhibitor from the medical leech Hirudo medicinalis. Isolation, sequence analysis, cDNA cloning, recombinant expression, and characterization. J. Biol. Chem. 273, 32927–32933 42. Babin, D. R., Peanasky, R. J., and Goos, S. M. (1984) The isoinhibitors of chymotrypsin elastase from ascaris-lumbricoides—the primary structure. Arch. Biochem. Biophys. 232, 143–161 43. Knecht, R., Seemu¨ller, U., Liersch, M., Fritz, H., Braun, D. G., and Chang, J. Y. (1983) Sequence determination of eglin C using combined microtechniques of amino acid analysis, peptide isolation, and automatic Edman degradation. Anal. Biochem. 130, 65–71 44. Kim, K. Y., Lim, H. K., Lee, K. J., Park, D. H., Kang, K. W., Chung, S. I., and Jung, K. H. (2000) Production and characterization of recombinant guamerin, an elastase-specific inhibitor, in the methylotrophic yeast Pichia pastoris. Protein Expr. Purif. 20, 1–9


45. Jung, H. I., Kim, S. I., Ha, K. S., Joe, C. O., and Kang, K. W. (1995) Isolation and characterization of guamerin, a new human leukocyte elastase inhibitor from Hirudo nipponia. J. Biol. Chem. 270, 13879 –13884 46. Kim, D. R., and Kang, K. W. (1998) Amino acid sequence of piguamerin, an antistasin-type protease inhibitor from the blood sucking leech Hirudo nipponia. Eur. J. Biochem. 254, 692– 697 47. Moser, M., Auerswald, E., Mentele, R., Eckerskorn, C., Fritz, H., and Fink, E. (1998) Bdellastasin, a serine protease inhibitor of the antistasin family from the medical leech (Hirudo medicinalis)—primary structure, expression in yeast, and characterisation of native and recombinant inhibitor. Eur. J. Biochem. 253, 212–220 48. Sollner, C., Mentele, R., Eckerskorn, C., Fritz, H., and Sommerhoff, C. P. (1994) Isolation and characterization of hirustasin, an antistasin-type serine-proteinase inhibitor from the medical leech Hirudo medicinalis. Eur. J. Biochem. 219, 937–943

49. Ascenzi, P., Amiconi, G., Bode, W., Bolognesi, M., Coletta, M., and Menegatti, E. (1995) Proteinase-inhibitors from the European medicinal leech Hirudo medicinalis structural, functional and biomedical aspects. Mol. Asp. Med. 16, 215–313 50. Rester, U., Bode, W., Moser, M., Parry, M. A., Huber, R., and Auerswald. E. (1999) Structure of the complex of the antistasin-type inhibitor bdellastasin with trypsin and modelling of the bdellastasin-microplasmin system. J. Mol. Biol. 293, 93–106 51. Hill, R. E., and Hastie, N. D. (1987) Accelerated evolution in the reactive center regions of serine protease inhibitors. Nature. 326, 96 –99 52. Zang, X., and Maizels, R. M. (2001) Serine proteinase inhibitors from nematodes and the arms race between host and pathogen. Trends Biochem. Sci. 26, 191–197 53. Baskova, I. P., and Zavalova, L. L. (2001) Proteinase inhibitors from the medicinal leech Hirudo medicinalis. Biochemistry (Mosc.) 66, 703–714


1613