Characterization of a cysteine protease from Tritrichomonas foetus that induces host-cell apoptosis

Archives of Biochemistry and Biophysics 477 (2008) 239–243

Contents lists available at ScienceDirect

Archives of Biochemistry and Biophysics journal homepage: www.elsevier.com/locate/yabbi

Characterization of a cysteine protease from Tritrichomonas foetus that induces host-cell apoptosis John J. Lucas a, Gary R. Hayes a, Hardip K. Kalsi a, Robert O. Gilbert b, Yongchool Choe c, Charles S. Craik c, Bibhuti N. Singh a,* a

Department of Biochemistry and Molecular Biology, SUNY Upstate Medical University, 4257 Weiskotten Hall, 766 Irving Avenue, Syracuse, NY 13210, USA Department of Clinical Sciences, College of Veterinary School, Cornell University, Ithaca, NY 14853, USA c Department of Pharmaceutical Chemistry, University of California at San Francisco, CA 94143, USA b

a r t i c l e

i n f o

Article history: Received 31 March 2008 and in revised form 14 May 2008 Available online 11 June 2008 Keywords: Bovine trichomoniasis Cysteine protease Apoptosis DNA sequencing

a b s t r a c t Tritrichomonas foetus is a serious veterinary pathogen, causing bovine trichomoniasis and affecting cattle herds world-wide, resulting in inflammation of the genital tract, infertility and huge economic losses. The parasite secretes a cysteine protease (CP8), which induces cytotoxicity and apoptosis in bovine vaginal and uterine epithelial cells. Mallinson et al. [D.J. Mallinson, J. Livingstone, K.M. Appleton, S.J. Lees, G.H. Coombs, M.J. North, Microbiology 1995, 141 (12) 3077–3085.] originally reported a partial DNA sequence of T. foetus CP8 based on PCR cloning of T. foetus genomic DNA. Here we report the biochemical properties of the CP8 enzyme. Kinetic properties and the substrate specificity profile of T. foetus CP8 were studied using positional scanning synthetic combinatorial libraries and Michaelis–Menten kinetic analysis of three synthetic fluorogenic substrates. The preferred substrate Z-Leu-Arg-MCA prevented host-cell death/apoptosis induced by CP8. In addition, the DNA sequence was completed by 30 and 50 rapid amplification of cDNA ends (RACE) and the full-length amino acid sequence was obtained. Ó 2008 Elsevier Inc. All rights reserved.

Tritrichomonas foetus is a parasite of particular veterinary importance, causing the sexually transmitted disease, bovine trichomoniasis. Bovine trichomoniasis creates a serious global economic burden in areas where free-ranging herds are maintained using natural service for insemination [1–3]. The infection normally persists for years in bulls without any clinical signs. In cows, the parasites initially adhere to and infect vaginal epithelial cells and later, upon migration to the uterus, infect epithelial cells of this organ and the placenta. Infection in cows results in vaginitis, endometritis, and, rarely pyometra. These consequences eventually lead to infertility and sometimes permanent sterility. In previous studies we showed that a cysteine protease (CP)1 obtained from T. foetus conditioned medium induced apoptosis of cultured host cells derived from the bovine reproductive tract (vaginal and uterine epithelial cells, BVECs and BUECs) [4,5]. The induction of apoptosis is correlated with protease activity since the specific CP inhibitor E-64 inhibited cytotoxicity and apoptosis in BVECs and BUECs treated with the CP [4,5]. Peptide sequencing iden-

* Corresponding author. Fax: +1 315 464 8750. E-mail address: [email protected] (B.N. Singh). 1 Abbreviations used: CP, cysteine protease; RACE, rapid amplification of cDNA ends; PS-SCL, positional scanning of synthetic combinatorial libraries; TYM, trypticaseyeast extract-maltose; TIB, trichomonad incubation buffer; MWCO, molecular weight cut-off; BVECs, Bovine vaginal epithelial cells; PLG, Phase lock gel; TFECP, Tritrichomonas foetus extracellular protease; SF, soluble fraction. 0003-9861/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.abb.2008.05.018

tified the enzyme as CP8, which was one of several CPs originally cloned and sequenced by Mallinson et al.[6]. The predicted amino acid sequence reported by Mallinson et al. [6] is incomplete and based on the molecular weight determined using MALDI-TOF mass spectrometry we postulated that a significant portion of the sequence is missing [4]. In the present study we report completion of the CP8 DNA sequence using 30 , 50 RACE and extend the predicted amino acid sequence of CP8. Furthermore, we report the substrate preference and kinetic properties of the CP8 enzyme using a combination of positional scanning of synthetic combinatorial libraries (PS-SCL) [7] and kinetic measurements using synthetic substrates. Methods and materials Parasites and soluble fraction (SF) Tritrichomonas foetus (strain D1, obtained from Dr. R. BonDurant, University of California, Davis) was cultured in Diamond’s trypticase-yeast extract-maltose (TYM) media, pH 7.2 as described by Singh et al. [8]. SF was obtained and purified as previously described [4]. Briefly, parasites were grown in screw-cap 500 mL serum bottles (4-L batches) and harvested in late log phase (24 h). Parasite densities were 107 to 1.4  107 parasites/mL. Parasites were harvested by centrifugation (6300g for 15 min). The pellet was washed sequentially with ice-cold PBS (10 mM sodium phosphate, 150 mM sodium chloride, pH 7.2), and trichomonad incubation buffer (TIB, PBS supplemented with 10 mM HEPES and 0.05% ascorbic acid, pH 7.2). The final washed pellet was resuspended in TIB buffer (9  108 parasites/mL), and incubated at 37 °C for 2.5 h. Preparations were used only when >95% of the

240

J.J. Lucas et al. / Archives of Biochemistry and Biophysics 477 (2008) 239–243

parasites were motile after incubation. Following incubation, parasites were centrifuged at 17,400g for 15 min, and the supernatant was filtered through a 0.45 lm filter (Corning, Corning, NY), followed by a 0.22 lm (Millipore, MA) filter, and centrifuged again at 48,400g for 2.5 h to remove insoluble material(s). Isolation of CP8 from soluble fraction As previously reported , SF was concentrated by centrifugation using a Jumbosep (PALL Corp., MI) centrifugal device, 10,000 molecular weight cut-off (MWCO). CP8 was further purified by bacitracin affinity chromatography according to the methods described earlier [9], except that Affi-Gel 10 (Bio-Rad, CA) was used as the support matrix [4]. The concentrated SF in TIB was diluted (1:3) with sodium acetate buffer (20 mM, pH 4.0) and applied to a column (1 by 30 cm) of the affinity matrix equilibrated with sodium acetate buffer at a flow rate of 0.3 mL/min. The column was rinsed (0.8 mL/min) with 20 mM sodium acetate buffer until the A280 read zero. Bound material was eluted (0.6 mL/min) with 0.1 M Tris–HCl (pH 7.0), 1.0 M NaCl-25% 2-propanol as reported by Thomford et al. [9]. The peak (A280) was collected and dialyzed for 3 h at 4 °C against water and concentrated by Jumbosep centrifugal device as described above. Finally, it was concentrated to 1 mL by centrifugal filtration using Amicon Ultra 10,000 MWCO (Millipore, MA), and stored at 40 °C until pooled with other identical preparations. The combined affinity column eluates were applied to a BioGel P60 column (1  60 cm) equilibrated with 0.1 M ammonium acetate (pH 6.0). Elution (0.4 mL/ min) was monitored at A280 and 2 mL fractions were collected as described[10]. The protein-containing peak was concentrated by centrifugation using Amicon Ultra as described above. This preparation contains a single Coomassie blue staining band upon SDS–PAGE analysis and only peptides from CP8 were observed when it was subjected to extensive mass spectral sequencing analysis [4]. PS-SCL analysis of CP8 PS-SCL assays were conducted using a library of fluorescent-labeled tetrapeptides as previously described [11]. The library was constructed such that for every substrate, one of the four positions was a fixed amino acid, while the other three positions were randomized, containing a total of 8000 compounds. Briefly, aliquots of 25 nmol in 1 lL from each of 20 sublibraries of the P1, P2, P3, and P4 were added to each well of a 96-well Microfluor-1 U-bottom plate (Dynex Technologies, Chantilly, VA) to a final concentration of 31.25 nM for each of the 8000 compounds in a 100 lL volume per well. CP8, preincubated with 5 mM DTT for 5 min, was added to the reaction mixture and fluorescence was measured with a SpectraMax Gemini fluorescence spectrophotometer (Molecular Devices, Sunnyvale, CA), with an excitation at 380 nm and emission at 460 nm, with a 435 nm cut-off. Assays were performed in triplicate in buffer containing 100 mM sodium acetate (pH 7.0), 100 mM sodium chloride, 1 mM EDTA, 10 mM DTT, and 0.01% Brij35. Data analysis was conducted using SOFTMAX software (Sunnyvale, CA) and Microsoft Excel (Microsoft Corp., Redmond, WA). Enzyme assays CP8 was assayed using fluorogenically labeled synthetic peptides as described by Singh et al. [4]. Generally, CP8 is added to PBS (pH 7.6) containing 5 mM DTT, and incubated for 5 min at 25 °C to reduce the enzyme prior to addition of 10 lL of substrate (e.g., Z-RR-MCA, benzyloxycarbonyl-L-arginyl-L-arginyl-4-methylcoumaryl-7-amide) dissolved in DMSO, to give 1 mL total volume in PBS. In experiments outlined below three different substrates were employed; Z-RR-MCA, Z-LRMCA, and Z-FR-MCA (Peptide International, Inc., KY). For routine screening Z-RRMCA was used at a concentration of 15 lM. The production of the fluorescent product (MCA) was monitored continuously using a Fluorolog 3; in conjunction with Datamax version 2.20 software (both from Instruments SA, Inc.), and Grams/32 version 4.14 Level II software (Galactic Industries Inc.); at an excitation wavelength of 380 nm (slit width of 1 nm) and emission wavelength of 460 nm (slit width of 1 nm). The CP8 concentration was based on Bradford protein assays and the previously published molecular weight of 23.7 kDa [4]. The concentration of active enzyme in CP8 preparations was determined by E-64 titration [12]. Assay conditions were as follows: CP8 (2 lg/mL; 84 nM), 5 mM DTT, and varying concentrations of E-64 in PBS (pH 7.6) were incubated for 30 min at 25 °C prior to the addition of 3.3 lM Z-RR-MCA. These titrations showed the concentration of active enzyme to be 58% of the total enzyme concentration. Data analysis was conducted in Sigmaplot version 8.0 (Systat Software, Inc., San Jose, CA), and Excel 2003 (Microsoft Corp.) using the Michaelis–Menten equation. Culture of host cells and assay for cytotoxicity/apoptosis Bovine vaginal epithelial cells (BVECs) were cultured as reported earlier by Singh et al. [4,5]. BVECs were maintained at 37 °C in an atmosphere of 5% CO2 in William’s complete medium [4,5]. Once cells were approaching confluence, any contaminating fibroblasts were removed by differential trypsinization and epithelial cells were subcultured in 96-well plates [4,5]. Experiments were performed

when cells were 70–80% confluent and proliferating. BVECs were exposed to CP8 (30 lg/well) for 4 h at 37 °C. Cytotoxicity required the presence of a reducing agent [4–6]. Therefore, prior to the start of experiments fresh medium containing cysteine (10 mM) was added to each well. Cytotoxicity of host cells depended on the CP8 concentration [4–6]. Preliminary experimental conditions were established to show that it took more than 15–20 h to kill BVECs when they were exposed to 15–20 lg of CP8. In control experiments, (a) CP8 was omitted, and (b) only the peptide substrate (Z-Leu-Arg-MCA) was added. In other experiments, BVECs were treated with CP8 (30 lg) plus the peptide substrate (Z-Leu-Arg-MCA, 0.83 mM) together. The WST-1 (BioVision, Inc. CA) assay was used to measure cytotoxicity and viability of BVECs, according to manufacturer’s instructions. Microscopic examination was usually performed prior to the use of WST-1 to asses the damage of cells. CP8 DNA sequence completion Extension of the known CP8 sequence was performed using 30 and 50 RACE as described below. Total RNA was isolated from T. foetus (1  107) grown in suspension to a density of 5–10  106 cells/mL. Parasites were harvested by centrifugation, suspended in 1 mL of TRIzol reagent (Invitrogen Corp., Carlsbad, CA) and processed as described by the manufacturer, including the optional isolation step described in the product manual. Samples were transferred to Phase Lock Gel (PLG) Heavy tubes (Eppendorf, Westbury, NY) for better recovery of total RNA. All subsequent isolation steps were conducted as described in the TRIzol reagent product manual. Total RNA was used to create 50 and 30 cDNAs using the SMART RACE cDNA Amplification Kit (Clonetech, Mountain View, CA). Gene specific primers for RACE PCR were chosen using Primer3 (Rozen and Skaletsky, 2000), based on the known partial sequence of CP8[6], such that the internal sequence was partially covered, as well as allowing for appropriate primer extension, and such that melting temperatures were above 70 °C, as per manufacturer’s recommendations. The gene specific primer sequences for 50 and 30 RACE PCR were CGTTCGAGCTTTCCGGATTTGATGC, and TGGCTGCGTCGGTTTCGGTGCT, respectively. RACE products were purified using the Montage PCR Cleanup Kit (Millipore, Billerica, MA). The products were cloned into the pcR2.1Topo vector for maintenance and sequencing (Invitrogen Corp., Carlsbad, CA). Plasmid preparations were digested using EcoRI to confirm the presence of inserts and clones were sequenced at the DNA sequencing core facility at SUNY Upstate Medical University, Syracuse, NY. Sequence analysis and comparison were conducted using Seqweb Version 3 (Accelrys, Burlington, MA).

Results T. foetus CP8 was previously shown to have cytotoxic properties when incubated with host cells (BVECs and BUECs) in culture and the cytotoxicity is dependent upon enzyme activity [4,5]. Thus, we initiated studies of the enzymatic properties of CP8. An early step in that process included analysis of substrate specificity using PSSCL. As shown in Fig. 1, the enzyme displays a strong preference for Arg and Lys residues in the P1 position, whereas, the P2 position is strongly selective for the hydrophobic residues leucine and valine, and less for phenylalanine, suggesting an aliphatic over aromatic preference. Other residues in the P2 position show very low activity. The P3 and P4 positions show little discriminatory power, although there is a subtle preference for Arg and Pro residues in the P3 position. Overall, these data suggest that the P1 and P2 positions are the most selective, which is consistent with cathepsin L-like subfamily proteases, but the slight preferences shown in the P3 position suggest similarities to cathepsin S [11]. Based on the results of the PS-SCL analysis three synthetic substrates were tested. The kinetic parameters shown in Table 1 reveal that although the Km for Z-RR-MCA is substantially lower than that of the other two substrates, the enzyme turnover (kcat) for Z-LRMCA is 200 times greater and the turnover for Z-FR-MCA is more than 40 times greater than that of Z-RR-MCA. The kinetic data measured directly are consistent with the results obtained by PSSCL assays. That is, CP8 prefers aliphatic hydrophobic residues at the P2 position, while arginine was less favorable. Based on these observations we predicted that the peptides produced by digestion with CP8 should inhibit apoptosis. Z-Leu-ArgMCA would be the best substrate to investigate this. Preliminary experiments using 15 lg/well of CP8 in the presence of 5 mM ZLeu-Arg-MCA (22 h) showed a significant reduction in damage to BVECs, as determined by microscopic examination. To quantify this

J.J. Lucas et al. / Archives of Biochemistry and Biophysics 477 (2008) 239–243

241

Fig. 1. Substrate specificity profile of CP8. The y axis represents picomolar per second. The x axis denotes 20 amino acids, where n = norleucine. Amino acids are arranged based on biochemical properties of their side chains. The height of the bars and error bars indicate means ± SD.

Table 1 Kinetic Parameters of CP8 for chosen substrates Substrate Z-RR-MCA Z-FR-MCA Z-LR-MCA

kcat (s1) 0.0538 ± 0.0027 2.34 ± 0.59 11.3 ± 2.2

kcat/Km (M1 s1)

Km (M) 7

2.93  10 1.74  106 6.72  106

1.83  105 1.32  106 1.69  106

Each substrate was tested at least in triplicate. For determination of kinetic constants using the three synthetic substrates 0.2 lg of CP8 was added to the assay mixture and data were collected for 5 min. The concentration of active CP8 (E0 = 4.9 nM, 58% active) was determined by titration with the CP-specific inhibitor E-64, and constants were calculated based on the active CP concentration. The substrates are ordered from least to most favorable based on their turnover numbers.

effect, higher doses of CP8 were used to determine the cytotoxicity induced by CP8 in the presence and absence of Z-Leu-Arg-MCA. Results are shown in Fig. 2. BVECs are completely destroyed (>98%) by CP8 and this destruction was prevented by the presence of preferred substrate, Z-Leu-Arg-MCA. The substrate alone had no affect. These results are totally consistent with the kinetic observations above and lend strong support to the conclusion that CP8 induces host-cell apoptosis through its CP enzymatic activity. Mallinson et al. [6] published a partial sequence for CP8. Based on the mass spectral sequence analysis and the MALDI-TOF MS molecular weight determinations published by Singh et al. [4] it was apparent that a large portion of the enzyme sequence was missing; perhaps between 25% and 40%. Therefore, as part of the characterization of CP8 we used RACE to complete the sequence. Using primers that overlapped the known DNA sequence available from GenBank (Accession No. X87781) at the 50 and 30 ends, an additional 489 bases were added to the known sequence (GenBank Accession No. EF610628). That extended the predicted amino acid sequence by 125 amino acid residues at the N-terminus and by 38

Fig. 2. Cytotoxicity of CP8. BVECs in 96-well plate were incubated for 4 h at 37 °C with CP8 (30 lg/well); in presence of peptide substrate Z-Leu-Arg-MCA (0.83 mM) plus CP8, and Z-Leu-Arg-MCA (0.83 mM). Cytotoxicity was measured by the WST-1 assay as described in Methods and materials. Experiments were performed in quintuplicate two times. The mean absorbance at 450 nm ± SD is shown.

residues at the C-terminal end of CP8. The complete amino acid sequence of CP8 is shown in Fig. 3. The sequence determined in the present study, is shown in red. The pro-form of the enzyme has a calculated molecular weight of 34.7 kDa. Discussion CP8 was originally described by Mallinson et al. [6], based on PCR cloning and DNA sequencing of T. foetus genomic DNA. In that study they identified seven T. foetus CPs. The cloning was

242

J.J. Lucas et al. / Archives of Biochemistry and Biophysics 477 (2008) 239–243

Fig. 3. Complete amino acid sequence of CP8 sequence added during this study is shown in red. The ‘‘ERFNIN” motif is underlined and appears in the pro region of the molecule. The most N-terminal sequence previously found by mass spectrometry [4] is from residues 101–107.* Represents the stop code.

predicated on PCR using primers designed to mimic consensus active site sequences of CPs. The CP8 sequence presented by Mallinson et al. [6] was clearly incomplete. Since our overall efforts are aimed at understanding the biology of CP8 and how it affects host cells we cloned a full-length form of the enzyme using 30 , 50 RACE. Both the start and termination codons were revealed by DNA sequencing. The predicted amino acid sequence shown in Fig. 3 nearly doubles the known sequence. The MW of the full-length pro-form of CP8 is consistent with that of other CPs and suggests that a nearly one hundred amino acid segment is removed during conversion to the active (mature) form of the enzyme—leaving an enzyme with a MW of approximately 23.4 kDa. In our previous study [4] over ninety amino acids of the one hundred fifty-two reported by Mallinson et al. [6] were sequenced by mass spectrometry. In addition, we observed four peptides that could not be aligned with the GenBank sequence, which can now be found in the sequence extension reported in Fig. 3. The most N-terminal peptide we sequenced was VESLDWR [4], which begins at position 101 (Fig. 3). This observation suggests cleavage between residues 100 and 101 during activation of CP8 and predicts a mature enzyme with a MW of 23.4 kDa, in close agreement with our previous determination of 23.7 kDa [4]. An additional consequence of cleavage between residues 100 and 101 is that a Cys residue would be in position 24 in the active enzyme, which is consistent with the location of active site Cys residues in other CPs of this clan. As shown by Singh et al. [4] CP8 is secreted into SF. Interestingly, the completed CP8 sequence allowed comparison to several short peptides sequenced by Thomford et al. [9], obtained from a protein they referred to as TFECP (T. foetus extracellular protease) and also obtained from SF. Only one of their peptides aligns with the CP8 sequence published by Mallinson et al. [6]. The remaining peptides are C-terminal to the then known CP8 sequence, but they are exact matches for the sequence we present in Fig. 3 (positions 278–307). Thus, it seems clear that TFECP and CP8 are the same enzyme. The sequence analysis of CP8 indicates that the enzyme is a member of the papain family of CPs, clan CA1, family C1. Furthermore, the sequence suggests that it is a member of the cathepsin L family. The pro-form of CP8 contains an ERFNIN sequence between residues 37 and 56. The ERFNIN motif has been proposed to assist correct folding of cathepsin L-like proteases, by acting as a structural support [13]. The PS-SCL studies showed a substrate specificity preferring a cationic amino acid at P1 and a hydrophobic amino acid at P2. This observation was confirmed by direct kinetic analysis of three synthetic substrates. Thomford et al. [9] surveyed T. foetus conditioned medium (the equivalent of SF) with a panel of synthetic substrates, including Z-RR-MCA and Z-FR-MCA. At the substrate concentration they used, their data suggested that CP8 is more active with Z-RRMCA. Clearly, the extensive kinetic analysis summarized in Table 1 show that Z-FR-MCA and Z-LR-MCA are more efficient substrates. Indeed, the kcats are nearly 50- and 200-fold greater for the preferred substrates, respectively. The PS-SCL analysis is consistent with CP8 being a member of the cathepsin L family and suggests it is closely related to cathepsin S, because of the slight preference

for Arg and Pro in the P3 position. It is clearly distinct from the cathepsin B family. A number of CPs have been identified in T. foetus, beginning with the first report of a purified trichomonad CP from T. foetus by McLaughlin and Müller in 1977 [14]. North, Coombs and their colleagues studied a number of secreted proteases from both T. vaginalis and T. foetus [15–17] as did Thomford et al. [9]. In addition to the seven CPs identified and cloned by Mallinson et al. two others were identified and cloned by Ikeda et al. (GenBank Accession Nos. U13153 and U13154). The recently published T. vaginalis genome [18] suggests that over forty papain family CPs are encoded by that closely related organism. Various biological functions have been proposed for parasite CPs, including immune evasion and nutrient acquisition. Trichomonads are obligate anaerobes and obtain many of their nutrients from the host, including lipids, nucleotides, and iron. We have shown that the levels of similar proteases from the human pathogen, T. vaginalis, are modulated by changes in iron concentration in culture [19]. Similarly, CP8 may enable T. foetus to obtain vital nutrients from the host. CPs may also contribute directly to pathogenesis when released into the host mucosal surface [9]. In addition to other mechanisms they may contribute to inflammation via induction of host-cell apoptosis, as described by Singh et al. [4]. The kinetic analysis, substrate preferences and inhibition of apoptosis by the preferred synthetic substrate provide important background information on CP8. For example, the study of peptide substrate specificity may help identify the host-cell target molecules or at least help confirm their identity when they are identified by other means. The substrate specificity data may also help in the design of CP8 inhibitors, which could be used as therapeutic agents for the treatment or prevention of bovine trichomoniasis. Acknowledgments The authors gratefully acknowledge the assistance of Dr. James McKerrow and Dr. K. Land, UCSF for making arrangements with Dr. C. Craik’s laboratory (UCSF) to conduct PS-SCL studies. This work was supported by NRI-USDA Grant (2003-032517 to B. N. Singh). References [1] R.H. BonDurant, Vet. Clin. North Am. Food Anim. Pract. 13 (1997) 345–361. [2] R.S. Felleisen, Microbes Infect. 1 (1999) 807–816. [3] D.O. Rae, J.E. Crews, E.C. Greiner, G.A. Donovan, Theriogenology 61 (2004) 605– 618. [4] B.N. Singh, J.J. Lucas, G.R. Hayes, I. Kumar, D.H. Beach, M. Frajblat, R.O. Gilbert, U. Sommer, C.E. Costello, Infect. Immun. 72 (2004) 4151–4158. [5] B.N. Singh, G.R. Hayes, J.J. Lucas, D.H. Beach, R.O. Gilbert, Am. J. Vet. Res. 66 (2005) 1181–1186. [6] D.J. Mallinson, J. Livingstone, K.M. Appleton, S.J. Lees, G.H. Coombs, M.J. North, Microbiology 141 (Pt 12) (1995) 3077–3085. [7] J.L. Harris, B.J. Backes, F. Leonetti, S. Mahrus, J.A. Ellman, C.S. Craik, Proc. Natl. Acad. Sci. USA 97 (2000) 7754–7759. [8] B.N. Singh, J.J. Lucas, D.H. Beach, S.T. Shin, R.O. Gilbert, Infect. Immun. 67 (1999) 3847–3854. [9] J.W. Thomford, J.A. Talbot, J.S. Ikeda, L.B. Corbeil, J. Parasitol. 82 (1996) 112– 117. [10] U. Sommer, C.E. Costello, G.R. Hayes, D.H. Beach, R.O. Gilbert, J.J. Lucas, B.N. Singh, J. Biol. Chem. 280 (2005) 23853–23860. [11] Y. Choe, F. Leonetti, D.C. Greenbaum, F. Lecaille, M. Bogyo, D. Bromme, J.A. Ellman, C.S. Craik, J. Biol. Chem. 281 (2006) 12824–12832. [12] A.J. Barrett, H. Kirschke, Methods Enzymol. 80 (Pt C) (1981) 535–561. [13] K. Schilling, S. Pietschmann, M. Fehn, I. Wenz, B. Wiederanders, Biol. Chem. 382 (2001) 859–865. [14] J. McLaughlin, M. Muller, J. Biol. Chem. 254 (1979) 1526–1533. [15] B.C. Lockwood, M.J. North, K.I. Scott, A.F. Bremner, G.H. Coombs, Mol. Biochem. Parasitol. 24 (1987) 89–95. [16] B.C. Lockwood, M.J. North, G.H. Coombs, Exp. Parasitol. 58 (1984) 245–253. [17] B.C. Lockwood, M.J. North, G.H. Coombs, Mol. Biochem. Parasitol. 30 (1988) 135–142. [18] J.M. Carlton, R.P. Hirt, J.C. Silva, A.L. Delcher, M. Schatz, Q. Zhao, J.R. Wortman, S.L. Bidwell, U.C. Alsmark, S. Besteiro, T. Sicheritz-Ponten, C.J. Noel, J.B. Dacks, P.G. Foster, C. Simillion, Y. Van de Peer, D. Miranda-Saavedra, G.J. Barton, G.D.

J.J. Lucas et al. / Archives of Biochemistry and Biophysics 477 (2008) 239–243 Westrop, S. Muller, D. Dessi, P.L. Fiori, Q. Ren, I. Paulsen, H. Zhang, F.D. BastidaCorcuera, A. Simoes-Barbosa, M.T. Brown, R.D. Hayes, M. Mukherjee, C.Y. Okumura, R. Schneider, A.J. Smith, S. Vanacova, M. Villalvazo, B.J. Haas, M. Pertea, T.V. Feldblyum, T.R. Utterback, C.L. Shu, K. Osoegawa, P.J. de Jong, I. Hrdy, L. Horvathova, Z. Zubacova, P. Dolezal, S.B. Malik, J.M. Logsdon Jr., K.

243

Henze, A. Gupta, C.C. Wang, R.L. Dunne, J.A. Upcroft, P. Upcroft, O. White, S.L. Salzberg, P. Tang, C.H. Chiu, Y.S. Lee, T.M. Embley, G.H. Coombs, J.C. Mottram, J. Tachezy, C.M. Fraser-Liggett, P.J. Johnson, Science 315 (2007) 207–212. [19] S. Kummer, G.R. Hayes, R.O. Gilbert, D.H. Beach, J.J. Lucas, B.N. Singh, Microb. Pathog. 44 (2008) 197–203.