Identification and Characterization of Falcilysin, a Metallopeptidase Involved in Hemoglobin Catabolism Within the Malaria Parasite Plasmodium Falciparum*

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 274, No. 45, Issue of November 5, pp. 32411–32417, 1999 © 1999 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in U.S.A. Identification and Characterization of Falcilysin, a Metallopeptidase Involved in Hemoglobin Catabolism within the Malaria Parasite Plasmodium falciparum*

(Received for publication, May 25, 1999, and in revised form, August 24, 1999)

Kathleen Kolakovich Eggleson, Kevin L. Duffin‡, and Daniel E. Goldberg§ From the Howard Hughes Medical Institute, Washington University, Departments of Molecular Microbiology and Medicine, St. Louis, Missouri 63110 and ‡Monsanto Corporate Research, Monsanto Company, St. Louis, Missouri 63198

The malaria parasite Plasmodium falciparum de- organelle, the food vacuole. This compartment has a pH esti- grades hemoglobin in its acidic food vacuole for use as a mated at 5.0–5.4 (3, 4). major nutrient source. A novel metallopeptidase activ- Nonproteolytic acid hydrolases could not be detected in food ity, falcilysin, was purified from food vacuoles and char- vacuoles isolated from P. falciparum (5). Thus, it appears that acterized. Falcilysin appears to function downstream of the food vacuole of P. falciparum does not function in degrada- the aspartic proteases plasmepsins I and II and the cys- tion and recycling of macromolecules in general. The catabolic Downloaded from teine protease falcipain in the hemoglobin proteolytic capability of this organelle is focused on hemoglobin. Disrup- pathway. It is unable to cleave hemoglobin or denatured tion of hemoglobin catabolism causes parasite death in an globin but readily destroys peptide fragments of hemoglobin. Falcilysin cleavage sites along the ␣ and ␤ chains animal model and in culture (2, 6, 7). The vital and specialized of hemoglobin are polar in character, with charged res- process of hemoglobin degradation within the food vacuole -idues located in the P1 and/or P4؅ positions. In contrast, provides promising targets for the development of novel anti plasmepsins I and II and falcipain prefer hydrophobic malarial drugs, greatly needed in the face of increasing resist- www.jbc.org residues around the scissile bond. The gene encoding ance to existing chemotherapeutic agents (8). falcilysin has been cloned. Its coding sequence exhibits Multiple proteases within the food vacuole facilitate the deg- features characteristic of clan ME family M16 metal- radation of hemoglobin to peptide fragments. Three acidic pro- lopeptidases, including an “inverted” HXXEH active site teases have been identified, purified from food vacuoles, and at Washington University on August 9, 2006 motif. Falcilysin shares primary structural features characterized (2, 9–14). Two aspartic proteases, plasmepsin I with M16 family members such as insulysin, mitochon- and plasmepsin II, can cleave native hemoglobin. A cysteine drial processing peptidase, nardilysin, and pitrilysin as protease, falcipain, is able to cleave denatured globin but not well as with data base hypothetical proteins that are potential M16 family members. The characterization of native hemoglobin (10, 15). Exopeptidase activity capable of falcilysin increases our understanding of hemoglobin generating individual amino acids from peptide fragments is catabolism in P. falciparum and the unusual M16 family absent from the food vacuole (16). However, peptides may tra- of metallopeptidases. verse the food vacuole membrane and could be converted to amino acids by exopeptidase activity in the parasite cytoplasm. An aminopeptidase that functions at neutral pH has been Plasmodium falciparum is a protozoan parasite that causes purified from parasites and characterized (17–19). the most lethal form of human malaria. Upon invasion of a When hemoglobin was incubated with food vacuole lysate at human erythrocyte, the parasite grows and matures sur- acidic pH, a series of discrete peptide fragments was generated rounded by cytosol consisting predominantly of a single pro- (16). Cleavage sites along the hemoglobin ␣ and ␤ chains were tein, hemoglobin. Amino acids derived from the proteolysis of located an average of eight amino acids apart. Many cleavage hemoglobin are incorporated into parasite proteins and parasites corresponded to the peptide bonds previously identified as sites require supplementation with only a few amino acids that sites for the known vacuolar proteases. Twenty-four cleavage are absent or deficient in hemoglobin for normal growth in sites that could not be attributed to the known proteases of the culture (1, 2). Hemoglobin proteolysis occurs within an acidic vacuole were identified. Unlike the preference of plasmepsin I, plasmepsin II, and falcipain for hydrophobic residues, many of * This work was supported in part by National Institutes of Health the novel cleavage sites contained polar residues at the P1 Grant AI-31615. Preliminary sequence data for P. falciparum chromo- and/or P1Ј positions. These results suggested that one or more some 14 was obtained from The Institute for Genomic Research web- vacuolar endopeptidases had remained unidentified (16). In site. Sequencing of chromosome 14 was part of the International Ma- this report, we describe the discovery of a novel proteolytic laria Genome Sequencing Project and was supported by awards from the Burroughs Wellcome Fund and the U. S. Department of Defense. activity, purification of the enzyme responsible for this activity, The costs of publication of this article were defrayed in part by the characterization of the purified enzyme, and the corresponding payment of page charges. This article must therefore be hereby marked molecular sequence data. Furthermore, we provide evidence “advertisement” in accordance with 18 U.S.C. Section 1734 solely to that this enzyme, falcilysin, has a distinct, downstream role in indicate this fact. The nucleotide sequence for the falcilysin gene and the amino acid the semiordered hemoglobin degradation pathway of P. sequence for the falcilysin protein have been deposited in the GenBank falciparum. data base under GenBank accession number AF123458. § Recipient of a Burroughs Wellcome Fund Scholar Award in Molec- EXPERIMENTAL PROCEDURES ular Parasitology. To whom correspondence should be addressed: Wash- ington University School of Medicine, Box 8230, 660 S. Euclid Ave., St. Reagents—Human hemoglobin, human globin, pepstatin, E64, Louis, MO 63110. Tel.: 314-362-1514; Fax: 314-362-1232; E-mail: EDTA, dipicolinic acid, amastatin, bestatin, bacitracin, 1,10-phenan- [email protected]. throline, cobalt sulfate, magnesium sulfate, manganese sulfate, zinc

This paper is available on line at http://www.jbc.org 32411 32412 Falcilysin, a Vacuolar Metallopeptidase from P. falciparum sulfate, BisTris,1 Klentaq LA, and Triton X-100 were obtained from Brilliant Blue staining. Endoproteinase Lys-C was used to perform an Sigma. Endoproteinase Lys-C and PMSF were from Roche Molecular in-gel digest of the falcilysin SDS-PAGE band and the resulting pep- Biochemicals (Indianapolis, IN). N-Ethylmaleimide was from Fluka tides were extracted (24). The peptide mixture was loaded onto a mi- (Buchs, Switzerland). Coomassie Brilliant Blue, gel filtration standard crobore C18 column. The column was washed for 15 min with 0.1% vials, high range SDS-PAGE standards, and Tween 20 were from Bio- trifluoroacetic acid and the peptides were eluted by applying a linear Rad (Hercules, CA). Silver nitrate was from Eastman Kodak (Roches- gradient of acetonitrile from 0 to 50% in 0.1% trifluoroacetic acid over ter, NY). The Ultrasphere C18 HPLC column was from P. J. Cobert 1 h at a flow rate of 100 ␮l/min. 0.5-min fractions were collected and the Associates, Inc. (St. Louis, MO). The Microbore C18 column was from peptides were detected at 215 nm on a Hewlett-Packard 1090 HPLC The Separations Group (Hesperia, CA). Mono S PC 1.6/5, Superose 12 system (Hewlett-Packard, Palo Alto, CA). Peptides were sequenced on a HR 10/30, DEAE-Sepharose, and CM-Sepharose were from Amersham PROCISE 492 protein sequencer (PE Biosystems, Foster City, CA).

Pharmacia Biotech. MES and acetonitrile (CH3CN) were from Fisher Metal Requirement—The apoenzyme form of falcilysin was generated (Pittsburgh, PA). Millex HV polyvinylidine difluoride 0.45-␮m filters by incubation in dilute form at 4 °C until no activity against the ␣71–78 were from Millipore (Bedford, MA). Synthetic oligonucleotides were substrate could be detected. Sulfate salts of Co2ϩ,Mg2ϩ,Mn2ϩ,orZn2ϩ from Integrated DNA Technologies (Coralville, IA). The quenched flu- were added to the apoenzyme at various concentrations and tested for orescence substrates were from AnaSpec (San Jose, CA). their ability to restore activity in the fluorogenic assay. An amount of Culture—P. falciparum clones HB3 (gift of W. Trager, Rockefeller apoenzyme equivalent to 0.47 units of original holoenzyme activity University) and 3D7 (gift of P. Rathod, Catholic University) were grown (where 1 unit of activity equals 1 pmol of quenched fluorescence sub- by the method of Trager and Jensen (20) in serum-free medium sup- strate cleaved/min) was used in each reaction. plemented with 0.5% AlbuMAX I from Life Technologies, Inc. (Grand Falcilysin was tested for inhibition by 10 ␮M amastatin, 10 ␮M Island, NY). Synchrony was maintained by sorbitol treatment (21). bestatin, 1 ␮M pepstatin, 10 ␮M E64, 1 mM PMSF, 5 mM N-ethylmale- Food Vacuole Preparation—Food vacuoles were isolated from P. fal- imide, 1 mM EDTA, 500 ␮M dipicolinic acid, and 1 mM 1,10-phenanthro- ciparum trophozoite-stage HB3 parasites by sorbitol lysis/differential line using the fluorogenic assay for falcilysin activity. Five separate centrifugation as described previously (9) except that the low speed determinations were made for each inhibitor. 6.3 units of falcilysin Downloaded from ϫ pellet after sorbitol lysis was harvested twice with 1.5 mM MgCl2 in 1 activity was used in each reaction. Controls were used to ensure that phosphate-buffered saline and that the final Percoll gradient centrifu- inhibition was not due to solvents and to correct for the intrinsic gation step was omitted. Food vacuole pellets were stored at Ϫ70 °C fluorescence of 1,10-phenanthroline. prior to falcilysin purification. A typical falcilysin purification started Identification of Hemoglobin-derived Falcilysin Substrates—The with food vacuoles prepared from a total of 3.6 liters of cultured para- ability of falcilysin to cleave hemoglobin or globin was investigated by sites at 2% hematocrit and 10% parasitemia. incubation of 15 units of purified falcilysin with 207 pmol (3.2 ␮g) of

Fluorogenic Assay for Falcilysin Activity—Samples were incubated hemoglobin or acid-denatured globin overnight at 37 °C in 100 mM www.jbc.org with 6.25 ␮M quenched fluorescence substrate corresponding to resi- sodium acetate buffer, pH 5.2. Recombinant plasmepsin II was pre- dues 71–78 of the ␣ chain of human hemoglobin, 4-(4-dimethylamino- pared as described previously (25). 0.04 units of recombinant plasmep- phenylazo)benzoic acid (Dabcyl)-Gaba-AHVDDMPN-5-[(2-aminoethyl- sin II measured using a quenched fluorescence peptide substrate for )amino]naphthalene-1 sulfonic acid (Edans), in 100 mM sodium acetate plasmepsin II activity (25) was also incubated with 207 pmol of hemo-

buffer, pH 5.2, at 37 °C. Incubations in the presence of 10 ␮M E64, 1 ␮M globin or acid-denatured globin under the same conditions. As controls, at Washington University on August 9, 2006 pepstatin, and 1 mM PMSF were used to detect falcilysin during the similar incubations without enzyme were performed. 10 ␮l of each 40-␮l purification process. Fluorescence was measured with an LS50B Lumi- sample was subjected to 15% SDS-PAGE. nescence Spectrometer from Perkin-Elmer (Buckinghamshire, United In order to determine whether falcilysin could cleave fragments of Kingdom) at excitation wavelength 336 nm, slit width 5 nm, and emis- hemoglobin, 162 ␮g of human hemoglobin was incubated overnight at sion wavelength 490 nm, slit width 12 nm. The assay for plasmepsin II 37 °C with 41 units of recombinant plasmepsin II in 150 mM sodium activity was performed as above, except that the quenched fluorescence acetate buffer, pH 5.2. More than 90% of the 162 ␮g of hemoglobin was substrate corresponded to residues 30–37 of the ␣ chain of human degraded as monitored by the loss of absorbance at 410 nm relative to hemoglobin (Dabcyl-Gaba-ERMFLSFP-Edans). hemoglobin incubated separately. The sample was treated with pepsta- Falcilsysin Purification—Food vacuole pellets were resuspended in tin to inhibit plasmepsin II. This substrate material was split into two 20 mM BisTris, pH 7.5, 1% Triton X-100, 1 mM PMSF, 10 ␮M E64, 1 ␮M equal portions and incubated overnight at 37 °C with or without 26 pepstatin. Resuspended vacuoles were lysed with 10 strokes by pestle A units of purified falcilysin in 150 mM sodium acetate buffer, pH 5.2. of a Dounce homogenizer and incubated on ice for 1 h. Samples were Each sample was applied to an Ultrasphere C18 column for reverse centrifuged at 20,000 ϫ g for1hat4°Candsupernatant was applied phase chromatography with a linear gradient of 0–60% acetonitrile at to a column of DEAE-Sepharose equilibrated with 20 mM BisTris, pH 1 ml/min over 50 min and peptide elution was monitored at 214 nm. 7.5. Bound proteins were eluted stepwise with 0.1, 0.2, 0.3, and 0.5 M Determination of Falcilysin Cleavage Sites—Hemoglobin fragments NaCl in 20 mM BisTris, pH 7.5. DEAE fractions were tested for activity were generated by incubation of 560 ␮g of human hemoglobin with 176 against the ␣71–78 quenched fluorescence substrate. The fractions with units of recombinant plasmepsin II as above. Pepstatin was added to the greatest activity were pooled, diluted with an equal volume of 100 block further plasmepsin II action. This material served as substrate for mM MES, pH 5.5, and subjected to cation exchange chromatography. the next reaction. Half was incubated overnight at 37 °C as above with Either CM-Sepharose or Mono S HPLC chromatography could be used 123 units of purified falcilysin and half was incubated separately from but Mono S proved more reproducible. Cation exchange resin was the falcilysin only control. The overnight reaction was stopped by add- equilibrated with 20 mM MES, pH 5.5. Bound proteins were eluted with ing glacial acetic acid to 20% final concentration. The substrate only NaCl in a stepwise gradient from 50 to 500 mM (CM) or a linear gradient and falcilysin only controls were combined. Both the falcilysin digest from0to2M over 100 min (Mono S). Fractions were tested for activity and combined control samples were desiccated and resuspended in 50 against the ␣71–78 quenched fluorescence substrate. In some prepara- ␮l of aqueous 0.1% trifluoroacetic acid. Five microliters was injected tions, a further gel filtration step was performed. The fractions with the onto a 0.32-mm inner diameter ϫ 15-cm capillary column packed with greatest activity were pooled, filtered with a .45 ␮M polyvinylidine Vydac C-18 (300 Å) stationary phase. Peptide fragments were eluted difluoride membrane, and applied to a Superose 12 HPLC column run from the column with 0–80% acetonitrile in 0.1% trifluoroacetic acid isocratically in PBS. Collected fractions were assayed and those with over 45 min at a flow rate of 5 ␮l/min. Column eluate passed through a the greatest activity represent purified falcilysin. Purity and molecular 100-␮m inner diameter fused silica transfer line to the electrospray weight were assessed by SDS-PAGE (22) and silver staining (23). interface of a Micromass Q-TOF mass spectrometer (Manchester, Internal Peptide Sequence Determination—SDS-PAGE was per- United Kingdom). Mass scans from m/z 400–4000 were summed over formed on purified falcilysin and the band was visualized by Coomassie 1 s to generate each mass spectrum in the LC/MS chromatogram. Retention times of eluted peptides were determined by an increase in the total ion current of the mass chromatogram. Molecular weights of 1 The abbreviations used are: BisTris, 2-[bis(2-hydroxyethyl)amino]- eluted peptides were determined after averaging the mass spectral scans over which the peptide eluted. Monoisotopic molecular weights 2-(hydroxymethyl)-propane-1,3-diol; E64, L-trans-epoxysuccinyl-leucyl- 12 amide-(4-guanidino)-butane; LC, liquid chromatography; MES, 2-(4- were determined by multiplying the m/z value of the peptide C isotope morpholino)ethanesulfonic acid; MS, mass spectrometry; PMSF, by the charge state of the peptide. The abundance values were deter- ϩ phenylmethanesulphonyl fluoride; PP, processing peptidase; Dabcyl, mined by summing the selected ion currents of the (M ϩ H) ,(Mϩ ϩ ϩ 4-(4-dimethylaminophenylazo)benzoic acid; Edans, 5-[(2-aminoethyl- 2H)2 ,(Mϩ 3H)3 , etc. from the LC/MS chromatograms. The units are )amino]naphthalene-1 sulfonic acid; PAGE, polyacrylamide gel electro- counts/s, and represent the ion current that the mass spectrometer phoresis; HPLC, high performance liquid chromatography. detects for each ion. This value is consistent with the concentration of Falcilysin, a Vacuolar Metallopeptidase from P. falciparum 32413 peptide in the LC eluate. Peptides of interest with respect to their abundance in the falcilysin digest versus control sample were selected for LC/MS/MS characterization after making a second LC injection. In this experiment, the qua- drupole mass analyzer sequentially passed peptides of selected m/z values into the collision cell of the mass spectrometer, where the peptides underwent 50 eV collisions with nitrogen gas. Collision-induced dissociation fragments (product ions) of each of the peptides were then mass analyzed by the time-of-flight segment of the mass spectrometer over a 1-s period to generate a MS/MS spectrum of the peptide. Product ion masses present in the MS/MS spectra were programmed into MS- Tag software (Protein Prospector, University of California, San Fran- cisco) (26, 27) to determine the identities of selected peptides. Falcilysin Gene Identification and Sequencing—Preliminary sequence data for P. falciparum chromosome 14 was obtained from The Institute for Genomic Research website. Sequencing of chromosome 14 was part of the International Malaria Genome Sequencing Project. The entire coding region was amplified from strain 3D7 genomic DNA with synthetic oligonucleotides annealing 56 base pairs 5Ј of the start codon (5Ј-TTATTTATTTATTCATTTTTATATTACACC-3Ј) and 63 base pairs 3Ј of the stop codon (5Ј-ACAAGGTGTATAATGTAAGTTTGC-3Ј) using Klentaq LA and was directly cloned into the pCR 2.1 vector using the FIG.1. Falcilysin migrates as a 125-kDa monomeric enzyme.

TA cloning kit from Invitrogen (San Diego, CA). Plasmid DNA template Activity of Superose 12 fractions in the fluorogenic assay for falcilysin Downloaded from was extracted from Escherichia coli and prepared for sequencing using activity. Retention times of marker proteins are shown. Inset, purified the Nucleobond Plasmid Midi Kit from CLONTECH (Palo Alto, CA) or falcilysin was subjected to reducing SDS-PAGE and silver staining. Migration is compared with high range SDS-PAGE standards. the Qiagen (Valencia, CA) plasmid mini kit. Sequencing samples were generated using an ABI PRISM BigDye Terminator Cycle Sequencing Ready Reaction Kit from the Applied Biosystems division of Perkin- The apoenzyme form of purified falcilysin was generated. No Elmer (Foster City, CA). Automated sequencing was performed at the activity against the ␣71–78 substrate could be detected using

DNA Sequencing Facility of the Protein and Nucleic Acid Chemistry the apoenzyme. Divalent metal cations were added to the www.jbc.org Laboratory at Washington University School of Medicine. Both strands apoenzyme at varying final concentrations and tested for their of the falcilysin gene were sequenced once, new polymerase chain re- ability to reconstitute activity against the ␣71–78 substrate action products were generated and cloned into pCR 2.1, and then both strands of the falcilysin gene were sequenced a second time. Compari- (Fig. 2). Magnesium restored very low levels of activity in a son of pairs of sequences was performed using the BESTFIT program range of concentrations spanning 5 orders of magnitude. Man- at Washington University on August 9, 2006 from the GCG package (28). Multiple sequence alignments were per- ganese was ineffective at low concentrations but reconstituted formed with Clustal W (29). a small amount of activity at 10 mM. Cobalt at 100 ␮M generated the highest level of activity observed. Zinc restored high RESULTS levels of activity at 1 ␮M and greater concentrations were Purification of Falcilysin from Food Vacuoles—In our previ- inhibitory, typical of zinc metalloenzymes (30). ous work the peptide products of human hemoglobin digestion A variety of inhibitors were tested for their abilities to block by food vacuole lysate were identified as discrete fragments of cleavage of the ␣71–78 substrate by falcilysin. As expected for the ␣ and ␤ chains (16). Analysis of food vacuole enzyme cleav- a metalloendopeptidase, high levels of activity remained in the age sites indicated the existence of one or more new vacuolar presence of the aminopeptidase inhibitors amastatin and endopeptidases. A prominent cleavage site between two adja- bestatin, the aspartic protease inhibitor pepstatin, the cysteine cent aspartic acid residues at positions 74 and 75 on the human protease inhibitor E64, and the serine protease inhibitor ␣ chain was found (16); therefore, a quenched fluorescence PMSF. Potent inhibition was observed with the chelators octapeptide substrate corresponding to ␣71–78, Dabcyl-Gaba- EDTA, dipicolinic acid, and 1,10-phenanthroline. The alkylat- AHVDDMPN-Edans, was synthesized. This substrate was ing agent N-ethylmaleimide also inhibited falcilysin activity by cleaved by food vacuole lysate. Cleavage was not blocked by 80%. inhibitors of aspartic, cysteine, or serine proteases but potent The fluorogenic assay was performed with sodium acetate inhibition was observed with metal chelators (data not shown). buffer at different pH values using saturating substrate. The Prior to this experiment, aspartic and cysteine proteases were pH optimum of falcilysin was determined to be approximately the only proteolytic enzymes known to function in the food 5.2 (data not shown). Optimal function at acidic pH is expected vacuole. The ␣71–78 substrate was used to detect a novel of vacuolar enzymes, since the pH of the food vacuole is be- metallopeptidase, which we call falcilysin, during purification tween 5.0 and 5.4 (3, 4). from food vacuoles. Falcilysin Cleaves Fragments of Hemoglobin—Plasmepsin I, Falcilysin was purified from the P. falciparum food vacuoles plasmepsin II, and falcipain readily cleave hemoglobin and/or by sequential anion and cation exchange chromatography. acid-denatured globin at vacuolar pH (10). We wished to deter- Yields of up to 11% were obtained. Purified falcilysin could be mine whether falcilysin also participates in the destruction of stabilized for storage at Ϫ70 °C using final concentrations of 1 hemoglobin or globin in the food vacuole. Hemoglobin or acid- mg/ml bovine serum albumin, 0.01% Tween 20, or 50 mM so- denatured globin were incubated overnight with 0.04 units of dium acetate, pH 5.2. plasmepsin II or 15 units of falcilysin (where 1 unit of activity Properties of Falcilysin—When cation exchange fractions equals 1 pmol of quenched fluorescence substrate cleaved/min). were applied to a Superose 12 column, a peak of activity No cleavage of either hemoglobin or acid-denatured globin by against the ␣71–78 quenched fluorescence substrate eluted falcilysin could be detected by silver stained SDS-PAGE. In with an estimated mass of 125 kDa (Fig. 1). The activity peak contrast, recombinant plasmepsin II was able to cleave both has a slight asymmetry. The reason for this is not known. substrates. The falcilysin activity used was sufficient to cleave However, it could be due to a minor proteolytic fragment of a quantity of its quenched fluorescence substrate 68-fold falcilysin or an interaction of the enzyme with the column. A greater than the molar amount of hemoglobin/globin present in silver-stained SDS-PAGE gel showed that falcilysin migrates the reaction. as a single band at 125 kDa (Fig. 1, inset). Since hemoglobin and globin were not cleaved by falcilysin, 32414 Falcilysin, a Vacuolar Metallopeptidase from P. falciparum

FIG.2.Activity of falcilysin apoenzyme is restored by metals. The apoenzyme form of falcilysin exhibited no activity in the fluorogenic assay for falcilysin. Cobalt sulfate (open circle), magnesium sulfate FIG.3. Falcilysin cleaves hemoglobin fragments. Hemoglobin (filled circle), manganese sulfate (open square), or zinc sulfate (filled was digested with recombinant plasmepsin II, the resulting material square) were added to the apoenzyme at various concentrations and was split into two equal portions, and each portion was incubated with tested for their ability to restore activity against the ␣71–78 quenched (lower panel) or without (upper panel) purified falcilysin. Each sample Downloaded from fluorescence substrate. was applied to an Ultrasphere C18 column for reverse phase chromatography with a linear gradient of 0–60% CH CN. we wished to investigate whether falcilysin had a role in the 3 hemoglobin catabolism process subsequent to the fragmenta- dase because the ␤ subunit has been shown to be responsible tion of globin. Fragments of hemoglobin generated with recom- for catalytic activity (32, 33) while the ␣ subunit is thought to binant plasmepsin II were incubated with or without purified function in substrate recognition (34). All of these enzymes, falcilysin, and subjected to reverse phase HPLC (Fig. 3). In the including falcilysin, are classified in the family M16 of metal- www.jbc.org falcilysin-treated sample, the plasmepsin II-generated frag- lopeptidase clan ME (35). Clan ME metallopeptidases are de- ments were efficiently destroyed and several peaks represent- fined by an active site motif, HXXEH. This sequence is inverted ing peptide products appeared at lower retention times. compared with the HEXXH motif found in the MA and MB ␣ ␤ Falcilysin Cleavage Sites Along the and Chains of Hemo- clans. The inversion observed in metallopeptidase clan ME in at Washington University on August 9, 2006 globin—Hemoglobin fragments were generated by digestion fact extends beyond the HXXEH motif, as a hydrophobic resi- with recombinant plasmepsin II and incubated in the presence due found two residues COOH-terminal to the HEXXH motif of and absence of falcilysin. Products were separated by reverse clans MA and MB is located two residues NH2-terminal to the phase HPLC and analyzed by tandem MS-MS (Table I). Some HXXEH motif of clan ME metallopeptidases (36). The M16 peptides were found in greater abundance in the control than family has some common sequence features. Three residues in the falcilysin digest and were therefore judged to be sub- function as zinc ligands, the two histidines of the HXXEH motif strates for falcilysin cleavage. Peptides unique to the falcilysin- and a glutamic acid residue located 75–84 residues COOH- treated sample represent the result of cleavage by falcilysin terminal to the HXXEH motif (37). Two other glutamic acid (Fig. 4). Analysis of these seven cleavage sites reveals that residues are part of the active site and are important for activ- falcilysin prefers sites with charged residues at the P1 and/or ity. The glutamic acid within the HXXEX motif is essential for Ј P4 positions. A glutamine or positively charged residue was activity but is not required for zinc binding (36, 38, 39). When preferred at the P2 position. the glutamic acid residue seven residues NH2-terminal to the Nucleotide Sequence and Deduced Amino Acid Sequence of zinc ligand glutamic acid was mutated to glutamine in pitrily- the Falcilysin Gene—Purified falcilysin was digested with en- sin, the resulting enzyme retained only one-third of wild type doproteinase Lys-C and amino acid sequences were determined activity and one-half the wild type amount of zinc (37). An for the resulting peptides. Eight sequences were obtained, and asparagine located 23–28 residues COOH-terminal to the these were used to search for the falcilysin gene in the data HXXEH motif is strictly conserved among all M16 family mem- bases of The International Malaria Genome Sequencing Con- bers known to be proteolytically active (35). A tyrosine located sortium for the P. falciparum genome project. A match was 129–163 residues COOH-terminal from the HXXEH motif is obtained with preliminary shotgun data for chromosome 14 conserved among all members known to be active except AXL1 from The Institute for Genomic Research. This information (35). allowed the falcilysin gene to be amplified from strain 3D7 genomic DNA by polymerase chain reaction, cloned, and DISCUSSION sequenced. Our previous work suggested that there may be an addi- Falcilysin is an 1193-amino acid protein encoded by a single tional unknown enzyme(s) in the food vacuole that participates continuous open reading frame 3579 base pairs in length. The in hemoglobin degradation (16). This paper describes the iden- predicted mass of the protein is 138.8 kDa, somewhat larger tification of a novel metallopeptidase activity in P. falciparum. than the 125 kDa determined for the native enzyme. The rea- Falcilysin was purified from food vacuoles and characterized. It son for this discrepancy is unclear at this time. The sequences is an oligoendopeptidase of the M16 family which contains two of all eight peptides derived from falcilysin could be found in subfamilies. Subfamily B contains consists of the mitochondrial the deduced amino acid sequence (Fig. 5). processing peptidases and chloroplast stromal processing pep-

BLAST searches (31) conducted with the falcilysin sequence tidases that cleave NH2-terminal signal peptides from proteins revealed homology with other metallopeptidases (Fig. 6). While (35). Subfamily A of family M16 consists of large proteins only one sequence is shown for each type of enzyme in this (approximately 100 kDa or more) that function as oligoen- analysis, it should be noted that many of these enzymes have dopeptidases. Substrates for these enzymes include peptides been identified in a wide variety of species. Fig. 6 includes only such as a-factor (40, 41), insulin (42), glucagon (43), atrial the ␤ subunit of heterodimeric mitochondrial processing pepti- natriuretic peptide (44), transforming growth factor ␣ (45), Falcilysin, a Vacuolar Metallopeptidase from P. falciparum 32415

TABLE I Peptides identified from samples of hemoglobin fragments incubated with falcilysin and in an untreated control Fragments of human hemoglobin were generated by digestion with recombinant plasmepsin II and incubated with purified falcilysin. Falcilysin only and substrate only controls were also incubated separately and combined after enzymatic inactivation. Components of the falcilysin digest and combined control samples were separated by reverse phase HPLC. Peptides in each sample were definitively identified as fragments of the ␣ and ␤ chains of human hemoglobin by MS/MS: mass spectrometry of peptide, fragmentation of peptide into subfragments, and mass spectrometry of subfragments. The relative abundances of a given peptide in the control and falcilysin digest samples indicated whether it represents a product of falcilysin digestion, a falcilysin substrate, or a peptide not cleaved by falcilysin. The abundance values were determined by summing the selected ion currents from the LC/MS chromatograms.

ϩFalcilysin Control Peptide mass Peptide identity abundance abundance Role in reaction 1877.9 ␤33–47 31 612 Preferred substrate 1471.9 ␤33–43 49 412 Preferred substrate 1706.0 ␤33–45 55 332 Preferred substrate 1701.0 ␣34–47 94 928 Preferred substrate 2096.4 ␤13–32 100 210 Substrate 2102.4 ␣83–100 208 345 Substrate 1200.9 ␣91–100 2270 Background Abundant product 435.3 Unknown 711 Background Abundant product 894.6 ␣34–41 596 Background Abundant product 1048.7 Unknown 554 Background Abundant product 544.4 Unknown 549 Background Abundant product

588.3 Unknown 543 Background Abundant product Downloaded from 509.3 ␤129–132 505 Background Abundant product 825.4 ␣42–47 489 Background Abundant product 465.3 Unknown 447 Background Abundant product 421.2 Unknown 323 Background Product 547.2 ␣75–79a 291 Background Product 848.4 ␤41–47 272 Background Product 558.4 Unknown 248 Background Product 510.8 ␤74–78 120 Background Product www.jbc.org a Indicates assignment by LC/MS only. Some masses could have been derived from more than one possible fragment and their identity is therefore marked as unknown.

food vacuole. In this study falcilysin cleaved peptides up to 20 at Washington University on August 9, 2006 amino acids and preferred those 11–15 residues in length (Ta- ble I). It thus functions downstream of other vacuolar proteases, providing strong evidence for order in the hemoglobin degradation process. Our current understanding of the sequential nature of the proteolytic events is that the initial cleavages of the native hemoglobin molecule are made by the plasmepsins. Denatured or fragmented globin is susceptible to cleavage by falcipain as well as the plasmepsins (15). Falcilysin cannot cleave hemoglobin or globin, but is able to cleave hemo- FIG.4. Falcilysin cleavage sites along the ␣ and ␤ chains of globin fragments. It is likely that many peptide intermediates human hemoglobin. Peptide products of falcilysin digestion could be are susceptible to cleavage by multiple enzymes. The distinct ␣ ␤ assigned to specific segments of the and chains of human hemoglo- but overlapping roles of the proteolytic enzymes of the food bin by LC/MS or MS/MS (Table I). Deduced falcilysin cleavage sites are listed. A relative consensus sequence is displayed below. Ϯ indicates vacuole and the location of exopeptidase activity across the food that a charged residue is preferred. vacuole membrane in the parasite cytoplasm suggest that hemoglobin catabolism within P. falciparum is a semiordered pathway. ␤-amyloid protein (46), somatostatin 28 (47), ␤-galactosidase The importance of zinc to intraerythrocytic P. falciparum has (48), and now, peptide fragments of hemoglobin (this work). been demonstrated. Levels of zinc increase in parallel with Inhibition by alkylating agents has been observed in insulysin parasite maturation within the red blood cell (50). Treatment of (42), nardilysin (47), mitochondrial processing peptidase of Neurospora crassa (49), and falcilysin (this work). Nardilysin is cultures with dipicolinic acid does not prevent schizont rupture inhibited by the aminopeptidase inhibitors amastatin and or reinvasion of host erythrocytes but blocks parasite matura- bestatin (47) but this was not observed in falcilysin. tion from the ring to the trophozoite stage, coincident with the The M16 family of metallopeptidases is growing. Besides the onset of hemoglobin degradation (50, 51). Inhibition of hemo- addition of falcilysin presented in this work, BLAST searches globin catabolism is lethal to intraerythrocytic P. falciparum revealed sequence characteristics common to M16 family mem- (2, 7). The specific role of falcilysin in the essential process of bers in several uncharacterized protein sequences from differ- hemoglobin degradation may have encouraging implications ent organisms (Fig. 7). In fact, many potential M16 family for the development of new chemotherapeutic agents. Plas- members share greater sequence identity with falcilysin than mepsins I and II as well as falcipain prefer to cleave the ␣ and the known members. Future investigations of these sequences ␤ chains of hemoglobin at sites with hydrophobic residues at will reveal which of these proteins function as metallopepti- the P1 and/or P1Ј positions (10). In contrast, falcilysin sites are dases and whether one or more new types of enzymes with polar in character, with charged residues at the P1 and/or P4Ј distinct biological functions are among them. positions. Internal cleavage of the large peptides generated by As expected for a metallopeptidase of this class, falcilysin is the plasmepsins and falcipain at distinct sites is likely to be unable to cleave polypeptides such as the ␣ and ␤ chains of critical for production of peptides small enough to cross the food hemoglobin (141 and 146 amino acids in length) but is able to vacuole membrane for cytosolic amino acid production. Falci- cleave within peptides generated by the other proteases of the lysin appears to play an integral role in this process. 32416 Falcilysin, a Vacuolar Metallopeptidase from P. falciparum

FIG.5.The falcilysin gene encodes an 1193 amino acid protein containing internal peptide sequences obtained from the native enzyme. Residues are numbered at the end of each row. Underlined sequences represent all eight internal peptide sequences obtained by endoproteinase Lys-C digestion of purified falcilysin. Downloaded from www.jbc.org at Washington University on August 9, 2006

FIG.6.Sequence alignment of known members of metallopeptidase clan ME family M16. An alignment of amino acid residues 233–416 of nardilysin from Rattus norvegicus (Swiss-Prot P47245, 1161 amino acids, 17.9% identical to falcilysin), 87–280 of insulysin from Homo sapiens (Swiss-Prot P14735, 1019 amino acids, 18.6% identical to falcilysin), 56–245 of sporozoite developmental protein from Eimeria bovis (Swiss-Prot P42789, 596 amino acids, 19.1% identical to falcilysin), 67–257 of pitrilysin from E. coli (Swiss-Prot P05458, 962 amino acids, 20.6% identical to falcilysin), 47–247 of AXL1 from Saccharomyces cerevisiae (Swiss-Prot P40851, 1208 amino acids, 18.8% identical to falcilysin), 217–403 of chloroplast stromal processing peptidase from Pisum sativum (GenBank U25111, 1259 amino acids, 14.7% identical to falcilysin), 62–248 of mitochondrial processing peptidase ␤ subunit from N. crassa (Swiss-Prot P11913, 476 amino acids, 21.2% identical to falcilysin), and 112–327 of falcilysin from P. falciparum (GenBank AF123458) is shown. Zinc ligands (*), active site residues (∧), and conserved residues (ϩ) are labeled above the aligned sequences.

FIG.7.Sequence alignment of potential members of metallopeptidase clan ME family M16 with falcilysin. An alignment of amino acid residues 42–239 of Swiss-Prot P48053 from Caenorhabditis elegans (995 amino acids, 21.1% identical to falcilysin), 42–238 of Swiss-Prot Q12496 from S. cerevisiae (1037 amino acids, 21.0% identical to falcilysin), 45–244 of GenBank AE001187 from Treponema pallidum (1023 amino acids, 24.0% identical to falcilysin), 42–238 of Swiss-Prot Q10068 from Schizosaccharomyces pombe (1036 amino acids, 20.4% identical to falcilysin), 49–228 of GenBank AF082738 from Streptococcus pyogenes (429 amino acids, 21.8% identical to falcilysin), 16–148 of Swiss-Prot Q46205 from Clostridium perfringens (148 amino acids, 49.3% identical to falcilysin), 112–327 of falcilysin from P. falciparum (GenBank AF123458), 44–240 of Swiss-Prot O51246 from Borrelia burgdorferi (971 amino acids, 28.0% identical to falcilysin), 66–266 of Swiss-Prot P32898 from S. cerevisiae (989 amino acids, 23.1% identical to falcilysin), and 73–276 of Swiss-Prot O42908 from S. pombe (882 amino acids, 25.5% identical to falcilysin) is shown. Zinc ligands (*), active site residues (∧), and conserved residues (ϩ) are labeled above the aligned sequences. Falcilysin, a Vacuolar Metallopeptidase from P. falciparum 32417

Acknowledgments—We thank Anna Oksman for assistance with cul- 224, 451–455 turing parasites, Ilya Gluzman for providing recombinant plasmepsin 25. Luker, K. E., Francis, S. E., Gluzman, I. Y., and Goldberg, D. E. (1996) Mol. II, Choukri Ben Mamoun for genomic DNA and assistance with com- Biochem. Parasitol. 79, 71–78 puter analyses, and Christoph Turck for determining internal peptide 26. Clauser, K. R., Hall, S. C., Smith, D. M., Webb, J. W., Andrews, L. E., Tran, H. M., Epstein, L. B., and Burlingame, A. L. (1995) Proc. Natl. Acad. Sci. sequences. U. S. A. 92, 5072–5076 REFERENCES 27. Yates, J. R., Eng, J. K., Clauser, K. R., and Burlingame, A. L. (1996) J. Am. Soc. Mass Spectrom. 7, 1089–1098 1. Divo, A. A., Geary, G., T., Davis, N. L., and Jensen, J. B. (1985) J. Protozool. 28. Devereux, J., Haeberli, P., and Smithies, O. (1984) Nucleic Acids Res. 12, 32, 59–64 387–395 2. Francis, S. E., Gluzman, I. Y., Oksman, A., Knickerbocker, A., Mueller, R., 29. Thompson, J. D., Higgins, D. G., and Gibson, T. J. (1994) Nucleic Acids Res. 22, Bryant, M. L., Sherman, D. R., Russell, D. G., and Goldberg, D. E. (1994) 4673–4680 EMBO J. 13, 306–317 30. Auld, D. S. (1995) Methods Enzymol. 248, 228–242 3. Yayon, A., Cabantchik, Z. I., and Ginsburg, H. (1984) EMBO J. 3, 2695–2700 31. Altschul, S. F., Gish, W., Miller, W., Myers, E. W., and Lipman, D. J. (1990) J. 4. Krogstad, D. J., Schlesinger, P. H., and Gluzman, I. Y. (1985) J. Cell Biol. 101, Mol. Biol. 215, 403–410 2302–2309 32. Kitada, S., Shimokata, K., Niidome, T., Ogishima, T., and Ito, A. (1995) 5. Goldberg, D. E., Slater, A. F. G., Cerami, A., and Henderson, G. B. (1990) Proc. J. Biochem. (Tokyo) 117, 1148–1150 Natl. Acad. Sci. U. S. A. 87, 2931–2935 33. Striebel, H. M., Rysavy, P., Spizek, J., and Kalousek, F. (1996) Arch. Biochem. 6. Rosenthal, P. J., Lee, G. K., and Smith, R. E. (1993) J. Clin. Invest. 91, Biophys. 335, 211–218 1052–1056 34. Shimokata, K., Kitada, S., Ogishima, T., and Ito, A. (1998) J. Biol. Chem. 273, 7. Rosenthal, P. J., Wollish, W. S., Palmer, J. T., and Rasnick, D. (1991) J. Clin. 25158–25163 Invest. 88, 1467–1472 35. Rawlings, N. D. (1998) in Handbook of Proteolytic Enzymes (Barret, A. J., 8. Olliaro, P., Cattani, J., and Wirth, D. (1996) J. Am. Med. Assoc. 275, 230–233 Rawlings, N. D., and Woessner, J. F., eds) pp. 1360–1362, Academic Press, 9. Goldberg, D. E., Slater, A. F. G., Beavis, R., Cerami, A., Chait, B., and Bath, United Kingdom Henderson, G. B. (1991) J. Exp. Med. 173, 961–969 36. Becker, A. B., and Roth, R. A. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 10. Gluzman, I. Y., Francis, S. E., Oksman, A., Smith, C. E., Duffin, K. L., and 3835–3839 Goldberg, D. E. (1994) J. Clin. Invest. 93, 1602–1608 37. Becker, A. B., and Roth, R. A. (1993) Biochem. J. 292, 137–142 Downloaded from 11. Dame, J. B., Reddy, G. R., Yowell, C. A., Dunn, B. M., Kay, J., and Berry, C. 38. Perlman, R. K., Gehm, B. D., Kuo, W. L., and Rosner, M. R. (1993) J. Biol. (1994) Mol. Biol. Parasitol. 64, 177–190 Chem. 268, 21538–21544 12. Rosenthal, P. J., McKerrow, J. H., Aikawa, M., Nagasawa, H., and Leech, J. H. (1988) J. Clin. Invest. 82, 1560–1566 39. Kitada, S., Kojima, K., Shimokata, K., Ogishima, T., and Ito, A. (1998) J. Biol. 13. Rosenthal, P. J., and Nelson, R. G. (1992) Mol. Biochem. Parasitol. 51, 143–152 Chem. 273, 32547–32553 14. Vander Jagt, D. L., Hunsaker, L. A., Campos, N. M., and Scaletti, J. V. (1992) 40. Fujita, A., Oka, C., Arikawa, Y., Katagai, T., Tonouchi, A., Kuhara, S., and Biochim. Biophys. Acta 1122, 256–264 Misumi, Y. (1994) Nature 372, 567–570 15. Francis, S. E., Gluzman, I. Y., Oksman, A., Banerjee, D., and Goldberg, D. E. 41. Adames, N., Blundell, K., Ashby, M. N., and Boone, C. (1995) Science 270, 464–466

(1996) Mol. Biol. Parasitol. 83, 189–200 www.jbc.org 16. Kolakovich, K. A., Gluzman, I. Y., Duffin, K. L., and Goldberg, D. E. (1997) 42. Shii, K., Yokono, K., Baba, S., and Roth, R. A. (1986) Diabetes 35, 675–683 Mol. Biol. Parasitol. 87, 123–135 43. Kirschner, R. J., and Goldberg, A. L. (1983) J. Biol. Chem. 258, 967–976 17. Florent, I., Derhy, Z., Allary, M., Monsigny, M., Mayer, R., and Schrevel, J. 44. Muller, D., Baumeister, H., Buck, F., and Richter, D. (1991) Eur. J. Biochem. (1998) Mol. Biol. Parasitol. 97, 149–160 202, 285–292 18. Curley, G. P., O’Donovan, S. M., McNally, J., Mullally, M., O’Hara, H., Troy, 45. Gehm, B. D., and Rosner, M. R. (1991) Endocrinology 128, 1603–1610

A., O’Callaghan, S. A., and Dalton, J. P. (1994) J. Eukaryot. Microbiol. 41, 46. Kurochkin, I., and Goto, S. (1994) FEBS Lett. 345, 33–37 at Washington University on August 9, 2006 119–123 47. Chesneau, V., Pierotti, A. R., Barre, N., Creminon, C., Tougard, C., and Cohen, 19. Vander Jagt, D. L., Baack, B. R., and Hunsaker, L. A. (1984) Mol. Biol. P. (1994) J. Biol. Chem. 269, 2056–2061 Parasitol. 10, 45–54 48. Cheng, Y. E., and Zisper, D. (1979) J. Biol. Chem. 254, 4698–4706 20. Trager, W., and Jensen, J. B. (1976) Science 193, 673–675 49. Schneider, A., Arretz, M., Wachter, E., and Neupert, W. (1990) J. Biol. Chem. 21. Lambros, C., and Vanderberg, J. P. (1979) J. Parasitol. 65, 418–420 265, 9881–9887 22. Laemmli, U. K. (1970) Nature 227, 680–685 50. Ginsburg, H., Gorodetsky, R., and Krugliak, M. (1986) Biochim. Biophys. Acta 23. Wray, W., Boulikas, T., Wray, V. P., and Hancock, R. (1981) Anal. Biochem. 886, 337–344 118, 197–203 51. Yayon, A., Vande Waa, J. A., Yayon, M., Geary, T. G., and Jensen, J. B. (1983) 24. Hellman, U., Wernstedt, C., Gonez, J., and Heldin, C. H. (1995) Anal. Biochem. J. Protozool. 30, 642–647