Serine Protease Activity in Developmental Stages of Eimeria Tenella

J. Parasitol., 93(2), 2007, pp. 333–340 ᭧ American Society of Parasitologists 2007

SERINE PROTEASE ACTIVITY IN DEVELOPMENTAL STAGES OF EIMERIA TENELLA

R. H. Fetterer, K. B. Miska, H. Lillehoj, and R. C. Barﬁeld Animal Parasitic Diseases Laboratory, Animal and Natural Resources Institute, United States Department of Agriculture, Henry A. Wallace Beltsville Agricultural Research Center, Beltsville, Maryland 20705. e-mail: [email protected]

ABSTRACT: A number of complex processes are involved in Eimeria spp. survival, including control of sporulation, intracellular invasion, evasion of host immune responses, successful reproduction, and nutrition. Proteases have been implicated in many of these processes, but the occurrence and functions of serine proteases have not been characterized. Bioinformatic analysis suggests that the Eimeria tenella genome contains several serine proteases that lack homology to trypsin. Using RT-PCR, a gene encoding a subtilisin-like and a rhomboid protease-like serine protease was shown to be developmentally regulated, both being poorly expressed in sporozoites (SZ) and merozoites (MZ). Casein substrate gel electrophoresis of oocyst extracts during sporulation demonstrated bands of proteolytic activity with relative molecular weights (Mr) of 18, 25, and 45 kDa that were eliminated by coincubation with serine protease inhibitors. A protease with Mr of 25 kDa was purified from extracts of unsporulated oocysts by a combination of affinity and anion exchange chromatography. Extracts of SZ contained only a single band of inhibitor- sensitive proteolytic activity at 25 kDa, while the pattern of proteases from extracts of MZ was similar to that of oocysts except for the occurrence of a 90 kDa protease, resistant to protease inhibitors. Excretory-secretory products (ESP) from MZ contained AEBSF (4-[2-Aminoethyl] benzenesulphonyl fluoride)–sensitive protease activity with a specific activity about 10 times greater than that observed in MZ extracts. No protease activity was observed in the ESP from SZ. Pretreatment of SZ with AEBSF significantly reduced SZ invasion and the release of the microneme protein, MIC2. The current results suggest that serine proteases are present in all the developmental stages examined.

Infection of chickens by several species of Eimeria causes rozoites and involved in cellular invasion. Results of a recent malabsorption, reduced weight gain, and decreased efficiency characterization of genes differentially expressed during spor- of feed conversion, resulting in significant economic losses to ulation of E. tenella indicated transcripts of several proteases the poultry industry, which are estimated to be over 800 million present in E. tenella (Miska et al., 2004). These include the dollars annually (Williams, 1998; Allen and Fetterer, 2002). Be- serine proteases subtilisin, rhomboid protease, prolyl endopep- cause they are intracellular parasites, poultry coccidians, like tidase and a metallo-protease, aminopeptidase N. Recently, a other apicomplexans, must have highly developed mechanisms rhomboid protease from E. tenella was cloned, expressed, and to invade and survive in host cells (Kim, 2004; Caruthers and localized to sporozoites (Li et al., 2006). In addition, an ami- Blackman, 2005). Along with intracellular invasion, other com- nopeptidase related to aminopeptidase N was purified and char- plex mechanisms are involved in Eimeria spp. survival, includ- acterized during sporulation of E. tenella, confirming a role for ing control of sporulation in the environment, evasion of host this protease in parasite development (Fetterer and Barfield, immune responses, successful reproduction, and nutrition (Al- 2003). Proteases are, therefore, potentially critical to many fac- len and Fetterer, 2002; Klemba and Goldberg, 2002). ets of the parasite’s survival including excystation, differentia- Recent research on the role of serine proteases in apicom- tion, immune invasion, and nutrition (Klemba and Goldberg, plexans has centered primarily on protein processing and other 2002). However, occurrence and function of proteases during functions related to intracellular survival. These have been rel- sporulation and in the sporozoite and merozoite stages are atively well investigated in Toxoplasma gondii and Plasmodium largely unknown. The current study presents an initial charac- species (Shaw et al., 2002; Kim, 2004; Withers-Martinez et al., terization of serine protease activity in E. tenella developmental 2004; Dowse and Soldati, 2005; O’Donnell and Blackman, stages. 2005). Recent studies have shown that serine protease inhibitors can prevent invasion of T. gondii tachyzoites (TZ) into host MATERIALS AND METHODS cells both in vitro and in vivo and that serine proteases related Bioinformatic analysis to rhomboid proteases or subtilisins are involved in protein processing of micronemes and other proteins that are essential for Sequence homologies to known serine proteases in the E. tenella invasion (Miller et al., 2001, 2003; Kim, 2004; Dowse and Sol- genome were investigated by performing a BLAST (Altschul et al., 1997) search using protein sequences representative of the serine dati, 2005). protease families from the MEROPS database (Rawlings et al., 2004) Presumably the invasion events in Eimeria spp. intracellular as a query against the sequence of the partial E. tenella genome stages are similar to those observed in other apicomplexans, located at the SANGER Institute (http://www.sanger.ac.uk/cgi-bin/ but the details of protease involvement in molecular events con- blast/submitblast/e-tenella/omni). trolling sporulation and other developmental processes are lack- ing. A study by Fuller and McDougald (1990) showed that ser- Host and parasites ine proteases related to trypsin may play a role in sporozoite Chickens (80–100 sex-sals, Moyers Hatcheries, Quakertown, Penn- invasion of host cells since classical serine protease inhibitors sylvania), 4–5 wk of age, were infected with 1.0–1.25 ϫ 105 E. tenella reduced invasion of E. tenella sporozoites into cells in vitro. In (Wampler strain) oocysts per bird, mixed in the feed. On day 7 post- inoculation (PI), birds were killed by cervical dislocation, and the cecae addition, a 20 kDa trypsin-like protease was partially purified were removed. Oocysts were recovered from infected cecae and spor- from E. tenella sporulated oocysts (Michalski et al., 1994), fur- ulated as previously described (Fetterer and Barfield, 2003). ther suggesting that a serine protease may be present in spo- To obtain oocysts in various stages of sporulation, unsporulated oocysts were suspended in PBS containing an antibiotic/antimycotic mixture (GIBCO, Gaithersburg, Maryland) and incubated under aeration at Received 28 December 2005; revised 3 August 2006; revised 21 Sep- 41 C. At the desired time interval (ranging from 0 to 72 hr), an aliquot tember 2006; accepted 21 September 2006. containing about 1 ϫ 108 oocysts was removed from the incubation

333 334 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 2, APRIL 2007

flask and centrifuged. The oocyst pellet was resuspended in 1.0 ml 40 of the protease inhibitor 4-(2-Aminoethyl) benzenesulphonyl fluoride mM Tris, pH 8.0, and stored at Ϫ70 C. (AEBSF, Sigma) dissolved in AB. The assay was incubated from 2 to Sporozoites (SZ) were prepared from fully sporulated oocysts (less 14 hr at 37 C. After incubation, 100 ␮l of cold 4.5% trichloroacetic than 30 days postharvest) as previously described (Fetterer et al., 2004). acid with 0.5 M NaCl was added to each reaction and the sample kept Merozoites (MZ) were collected from cecae at 110 hr PI from birds at 4 C for 30 min. The sample was centrifuged at 10,000 g for 5 min inoculated with 3 ϫ 105 sporulated oocysts per bird. MZ were isolated at 4 C. The supernatant was diluted with 180 ␮l of 0.5 M Tris, pH 8.0. and purified as described for SZ. Isolated SZ and MZ were resuspended The fluorescence was read in a microtiter plate fluorometer at an exci- in 40 mM Tris and frozen at Ϫ70 C. tation wavelength of 485 nm and emission wavelength of 538 nm. The amount of fluorescein released was estimated from a standard curve, Reverse transcription polymerase chain reaction (RT-PCR) and results expressed as nmol released per mg protein. All parasite material used in RNA isolation was snap frozen follow- Electrophoresis ing purification and stored at Ϫ70 C until use. Total RNA was isolated from MZ, SZ, unsporulated oocysts (0 hr), and sporulating oocysts (12– Protein samples were analyzed by polyacrylamide gel electrophoresis 48 hr), as well as fully sporulated oocysts (72 hr) using TRIzol (Invi- using 1-mm-thick gradient gels (8 ϫ 9 cm, 4–12% Bis-Tris, Invitrogen) trogen, Carlsbad, California). Each sample of oocysts or SZ was com- as described (Fetterer and Barfield, 2003). Western blot analysis was bined with approximately 3 g of diethylpyrocarbonate (DEPC) treated performed using the basic method previously described (Fetterer and Pyrex beads (3 mm diameter) (Corning, New York, New York) and 10 Barfield, 2003), except that the primary antibody consisted of a 1:500 ml of TRIzol. The samples were vortexed for 1 min, then incubated on dilution of mouse monoclonal antibody against a recombinant E. tenella ice for 1 min (4ϫ). MZ samples were treated directly with TRIzol microneme protein, MIC2 (Tomley et al., 1996; Lillehoj et al., 2005). without employing beads. The remainder of the total RNA isolation Substrate gels consisted of casein imbedded 12% Tris-glycine gels protocol was carried out using the manufacturer’s recommended instruc- (Zymogram gels, Invitrogen). Samples were prepared by dilution in 2ϫ tions. The resulting pellets containing total RNA were resuspended in Tris-glycine sample buffer, allowed to stand at room temperature for 15 DNase/RNase free water (Invitrogen) and stored until cDNA synthesis min, and applied to lanes of gel. Gels were run at 100 V for about 2 at Ϫ70 C. Prior to cDNA synthesis, the total RNA was treated with hr in a standard Tris-glycine running buffer. After electrophoresis, gels DNase I (Invitrogen) using the manufacturer’s instructions. cDNA was were incubated in 100 ml renaturing buffer (2.5% [w/v] TritonX-100 in synthesized from 0.8 ␮g of total RNA following the recommended in- water) for 30 min followed by incubation in 100 ml developing buffer structions provided with the Advantage RT for PCR kit (BD Biosci- for 30 min at room temperature and then overnight in 100 ml of de- ences Clontech, Palo Alto, California) using oligo dT or random hex- veloping buffer at 37 C. In most experiments, the developing buffer

amer primers. Primer sequences for each of the genes were as follows: consisted of 50 mM Tris, 200 mM NaCl, 5 mM CaCl2, 0.02% (w/v) rhomboid protease forward primer 5Ј CCTCCAAGCATGGAGGCTT Brij 35. In some experiments, an acidic developing buffer (50 mM Ј Ј Ј ATC 3 and reverse primer 5 CTTGTTGAATTCACACGTTACTG 3 , sodium acetate, 100 mM NaCl, 5 mM CaCl2, 5 mM dithiothreitol, pH subtilisin forward primer 5Ј GCAATCATCCACGAATCCAT 3Ј and re- 5.0) was used. In other cases, protease inhibitors AEBSF, phenylmeth- verse primer 5Ј GAACGCATGGGAAGTCTGGA 3Ј. As a positive con- ylsulfonyl fluoride (PMSF), or alpha-tosyl-L-lysine chloromethyl ketone trol, primers designed to amplify a portion of the 18s rRNA were used (TLCK) (all from Sigma) were added to developing buffer. After de- (forward primer 5Ј CGGTGAAACTGCGAATGGCTCA 3Ј and reverse velopment, gels were stained with Coomassie Blue (Simply Blue, In- primer 5Ј GCCTTCCTTAGATGTGGTAGCC 3Ј). Thermocycling con- vitrogen), and the presence of clear bands against a dark background ditions were as follows: initial heat activation of polymerase, 95 C for indicated protease activity. 2 min, denaturation, 94 C for 30 sec, annealing, 60 C for 30 sec, extension, 72 C for 1 min, and a final extension of 5 min at 72 C. Cycles Protease purification 2–4 were repeated 34 times. TAP polymerase (CLP, San Diego, Cali- ␮ Soluble protein from unsporulated oocysts was diluted into loading fornia) in the presence of 400 nM primer was used, and 2.5 l of cDNA buffer (LB, 150 mM Tris, 150 mM KCl, pH 7.5) in a volume of about was used as target in each reaction. As a negative control, DNase I 7 ml (3 mg/ml). The sample was applied to an affinity column (about treated RNA was used as target in RT-PCR to determine whether con- 4 ml column volume) consisting of the serine protease inhibitor ben- taminating genomic DNA could be contributing to the presence of prod- zamidine bound to Sepharose-4 (Amersham Bioscience, Uppsala, Swe- ucts. den). The sample was applied with a flow rate of about 0.8 ml per min, and flow-through was reapplied for 4 cycles. The column was washed Protein extracts with LB, and absorbance (280 nm) was monitored from 2 ml aliquots Oocysts (1 ϫ 108) suspended in 1 ml of 40 mM Tris or PBS were (about 25 ml total) until the absorbance was less than 0.002 AUFS. The placed in a 1.5-ml capped microfuge tube containing 0.5 g of 0.5 mm column was then eluted with 0.05 M glycine, pH 3.0, and 2 ml fractions glass beads and homogenized with a mini-bead-beater (BioSpec Prod- collected into tubes containing 0.1 ml of 1.0 M Tris, pH 9.0. The frac- ucts, Bartlesville, Oklahoma). Soluble extracts were prepared as previ- tions were assayed for protease activity by monitoring the hydrolysis ously described (Fetterer and Barfield, 2003). of FITC-casein as described above; fractions containing protease activ- SZ collected as described above were incubated in 2 ml sterile Hank’s ity were pooled and dialyzed against 40 mM Tris, pH 8.0. The sample balanced salt solution (2 ϫ 108/ml) for 2 hr under an atmosphere of from the affinity column was injected onto an HPLC anion exchange ϫ 95% air, 5% CO2 at 41 C. MZ collected as described above were in- column (DEAE 5PW, Protein-Pak, 7.5 75 mm steel, Waters, Milford, cubated in 3 ml of sterile Hank’s balanced salt solution (HBSS) (3 ϫ Massachusetts). The mobile phase consisted of 0.05 M Tris, pH 8.0 (A) 8 10 /ml) for 2 hr at 41 C under an atmosphere of 95% air, 5% CO2,or and 0.05 M Tris, 1.0 M NaCl, pH 8.0 (B). The flow rate was 1.0 ml at 4 C under atmosphere of air. Following incubation, parasites were per min at 22 C. The column was eluted with solvent A for 10 min centrifuged, and the supernatant referred to as excretory-secretory prod- followed by a linear gradient from 100% A, 0% B to 50% A, 50% B uct (ESP) was removed, concentrated, and frozen at Ϫ70 C. The pellets in 35 min. Aliquots were collected at 0.5 min intervals (0.5 ml) and containing intact MZ or SZ were washed 2ϫ with PBS and frozen at absorbance monitored with a photo-diode array detector (Waters Model Ϫ70 C. Concentrations of soluble proteins were measured by BCA as- 996). The fractions were assayed for protease activity and fractions say (Pierce, Rockford, Illinois) with bovine serum albumin as the stan- containing protease activity were pooled, concentrated, and dialyzed dard. against 40 mM Tris, pH 8.0.

Protease assay In vitro cell invasion assays The proteolytic activity of extracts was determined by measuring the Madin-Darby bovine kidney (MDBK) cells, American Type Culture liberation of fluorescein from fluorescein isothiocyanate conjugated to Collection (ATCC, Fairfax, Virginia) were propagated in 75-cm2 flasks casein (FITC-casein). The assay consisted of 30 ␮l of assay buffer (AB, in culture media (CM) consisting of Earle’s modified Eagles media ␮ 0.05 M Tris, 150 mM NaCl, 5 mM CaCl2, pH 8.2), 10 l FITC-casein (EMEM, ATCC) containing 5% fetal bovine serum (FBS, ATCC); cells (Sigma, St. Louis, Missouri), in 0.1 M Tris, pH 8.0, and a 10 ␮l extract were then trypsinized, counted, diluted, and added at a concentration sample. In some cases, only 20 ␮l of AB was used, along with 10 ␮l of 3 ϫ 105/well to 24-well plates containing sterile 8 mm round glass FETTERER ET AL.—SERINE PROTEASES IN E. TENELLA DEVELOPMENTAL STAGES 335 cover slips in 1 ml CM. Plates with cells were incubated at 37 C under an atmosphere of 95% air, 5% CO2 for 48 hr to achieve an 80% con- fluent monolayer. To quantify invasion, freshly excysted SZ (4 ϫ 105) were incubated in 1 ml of EMEM (control) or EMEM containing AEBSF or monensin (Sigma) for 1 hr at 41 C. SZ were removed from treatment by centrifugation and washed in EMEM for 15 min at 22 C, then centrifuged and resuspended in CM. Prior to application of SZ, CM was removed from cell layers, and cell surfaces were washed twice with sterile PBS. SZ in CM were applied to cells on cover slips (trip- licate treatment) and incubated at 41 C for 2 hr. Following incubation, CM with SZ was removed and cover slips and cell surfaces were gently washed twice with 1ϫ PBS at 41 C, and cells were fixed with 1 ml of ice-cold methanol/well for 15 min at 4 C. To enumerate intracellular SZ, cover slips were subjected to immunofluorescent staining using a rabbit antibody to the parasite antigen SZ1, as previously described (Fetterer et al., 2004). Cell invasion was determined by counting intracellular SZ in 50 fields per cover slip at ϫ1,000 magnification. In experiments to detect release of microneme proteins during invasion, cell sheets in wells without cover slips were prepared as described above, washed with sterile 1ϫ PBS. SZ pretreated with AEBSF or monensin were added to wells (1 ϫ 107/ml/well). After incubation for 2 hr at 41 C, media with SZ was removed from cell wells, centrifuged at 3,080 g for 15 min at 22 C to remove any SZ, and the media frozen at Ϫ80 C.

RESULTS BLAST P analysis BLAST P analysis of examples from serine protease families in the MEROPS database against the E. tenella genome indicates a relatively small number of serine protease families present in the E. tenella genome. There were strong homologies to members of other serine protease families, including subtilisin (S8A, 1.5 e-19), proly-endopeptidase (S9A, 3.6 e-68), and rhomboid proteases (S54; 9.1 e-5 for Drosophila sp. and 4.4 FIGURE 1. RT-PCR analysis of the expression of subtilisin, rhomboid, and ribosomal RNA (rRNA) as a positive control during sporu- e-12 for T. gondii). lation from 0 to 72 hr and in sporozoites (SZ) or merozoites (MZ). S indicates standards in descending order from the top of 900, 700, 500, RT-PCR analysis and 250 bp. The sizes of the transcripts were (A) 176 bp for subtilisin, (B) 421 bp for rhomboid, and (C) 355 bp for rRNA. Subtilisin transcripts were observed at 0, 12, 24, and 72 hr following the initiation of sporulation; however, no transcripts were detected at 48 hr of sporulation, or in SZ and MZ (Fig. activity was obtained when 1 mM PMSF or TLCK was includ- 1A). The pattern of expression of the rhomboid gene differed ed in the developing buffer (data not shown). Analysis of ex- from that of subtilisin in that rhomboid transcripts were present tracts using FITC-casein assay indicates significant AEBSF- in all stages of sporulation (Fig. 1B). As a positive control, sensitive protease activity during sporulation, with the highest expression of the 18S rRNA was used (Fig. 1C), while a lack in the unsporulated oocysts (0 hr) and the least activity ob- of amplification when RNA treated with DNase was used as served after 72 hr of sporulation (Fig. 2B). target with all 4 primer pairs serving as a negative control (data Serine protease activity was purified from E. tenella unspo- not shown). rulated oocysts by a combination of affinity and ion exchange chromatography (Fig. 3). Initial purification using a benzamidine affinity column yielded a peak of protease activity (Fig. Proteolytic activity during sporulation 3A). Analysis of the fractions by substrate-gel electrophoresis Analysis of soluble extracts of E. tenella oocysts during spor- indicated fractions 46–50 were enriched in activity similar to ulation by substrate gel-electrophoresis using casein-impreg- that observed in the unfractionated extract (Fig. 3B). When the nated gels indicated several bands of proteolytic activity present partially purified protease was further fractionated by HPLC- during sporulation (Fig. 2). Bands with relative molecular anion exchange chromatography, a distinct peak of proteolytic weights (Mr) of about 18, 23–25, and 45 kDa and occasionally activity with a retention time of about 49 min was observed, bands of 90 kDa were observed (Fig. 2A). Activity was present which did not correlate with the protein absorbance (280 nm) from 0 hr (unsporulated) to 72 hr (fully sporulated); however, peak (Fig. 3C). When the fractions comprising this peak were activity appeared to be less after 72 hr of sporulation. When analyzed, only a single band of activity was observed on sub- acidic buffer (pH 5.0) was used to develop activity in substrate strate gels with Mr of about 25 kDa, while analysis on a re- gels, the intensity of bands was either eliminated or greatly duced, denatured gel stained with Coomassie Blue indicated 2 reduced compared to pH 8.2. When AEBSF was included in bands with Mr of about 23 and 25 kDa, a band of about 18 the developing buffer, all the bands of enzyme activity observed kDa, as well as some minor bands of higher Mr (Fig. 3D). on gels were greatly reduced. A similar inhibition of proteolytic Analysis of soluble extracts of SZ by substrate electropho- 336 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 2, APRIL 2007

FIGURE 2. Analysis of protease activity in soluble extracts of E. tenella oocysts during sporulation. (A) Casein substrate electrophoresis of soluble extract from 0 to 72 hr after sporulation performed either at pH 8.2, pH 5.0, or in the presence of 1 mM AEBSF. Clear bands against dark background indicate areas of proteolytic activity. (B) Proteolytic activity of soluble extracts of E. tenella oocysts after 0 to 72 hr of sporulation assayed by hydrolysis of FITC-casein. Second bar for each time point represents protease activity determined with 1mM AEBSF in the assay mixture. Values are means from duplicate measurements. resis indicates a single band of proteolytic activity with Mr of to the 1 hr pretreatment. Pretreatment of SZ with 1 ␮g/ml mo- about 25 kDa, while MZ extracts had multiple bands at 23–26 nensin had only a minimal effect on MIC2 release. kDa and 18 kDa as well as multiple bands at 90–116 kDa (Fig. 4). Addition of 1 mM AEBSF to developing buffer completely DISCUSSION abolished proteolytic activity observed in SZ, but the bands of The results of the current study demonstrate that alkaline 90 kDa were still apparent in MZ extract treated with AEBSF. proteases are present in extracts of all developmental stages of Proteolytic activity could not be detected in ESP from sporo- E. tenella examined. Proteolytic activities in the substrate gels zoites (data not shown). and in a FITC-casein assay were significantly inhibited by spe- Analysis of ESP collected from MZ incubated at 41 C indi- cific serine protease inhibitors, indicating proteolytic activity is cated prominent proteolytic bands at 45 and 23–25 kDa. MZ likely due to one or more proteases of the serine class. In ad- incubated at 4 C had a similar pattern of proteolytic activity, dition, acidic conditions that should suppress activity of most but the intensity was much less than that for MZ incubated at serine proteases and be more conducive to detection of cysteine 41 C (Fig. 5, left). Inclusion of 1 mM AEBSF in the developing protease activity eliminated or greatly reduced proteolytic ac- buffer greatly reduced the intensity of bands observed. Analysis tivity. A limited number of studies of serine proteases in E. of proteolytic activity indicated that the total proteolytic activity tenella have suggested that serine protease activity was due to of ESP from MZ collected at 41 C was about 10 times that a trypsin or trypsin-like protein (Michalski et al., 1990; Fuller observed for the extract (Fig. 5, right). and McDougald, 1994). Since there was little homology to the trypsin/chymotrypsin subfamily in the E. tenella genome, while In vitro cell invasion strong homologies were observed for other serine protease fam- Treatment of E. tenella SZ for 1 hr with AEBSF prior to ilies including subtilisin, oligo endopeptidase, and rhomboid application to MBDK cells resulted in a dose-related decrease protease, it is unlikely that the serine protease activity observed in cellular invasion (Fig. 6A). A concentration of 300 ␮M in the current study is due to a trypsin-like protein. In addition, AEBSF caused about an 88% decrease in numbers of SZ in- our current demonstration of developmental expression of a vading cells, while at a concentration of 1 mM, a greater than subtilisin and a rhomboid protease, as well as the report of the 95% decrease in cell invasion was observed (data not shown). expression of other classes of serine proteases (Miska et al., The anticoccidial monensin, at a concentration of 1 ␮g/ml, com- 2004; Fetterer and Barfield, 2003), provides at least circum- pletely inhibited invasion. stantial evidence of the presence of several serine protease fam- When SZ were allowed to invade MBDK cells in vitro for 2 ilies in E. tenella. Recently a gene for a 24 kDa rhomboid hr, the microneme protein MIC2 was detected in the cell culture protease from E. tenella has been cloned and expressed (Li et supernatant by Western blot analysis (Fig. 6B). The level of al., 2006). The sequence of a gene deduced from the rhomboid MIC2 released was greatly reduced relative to untreated con- protease transcript used in the present study (K. Miska, unpubl. trols when SZ were preincubated in AEBSF (1 mM) for 1 hr, obs.) has a 99% identity to the recently cloned rhomboid pro- washed free of the inhibitor, and then allowed to invade cells tease, suggesting they represent the same gene. Interestingly, for 2 hr. A similar result was observed if 1 mM AEBSF was the antibody generated to the recombinant rhomboid protease added to cell culture media during the 2 hr invasion in addition localized a protein in the SZ (Li et al., 2006), but the current FETTERER ET AL.—SERINE PROTEASES IN E. TENELLA DEVELOPMENTAL STAGES 337

FIGURE 3. Purification of proteolytic activity from soluble extracts of E. tenella unsporulated (0 hr) oocysts. (A) Isolation of proteolytic activity from a benzamidine affinity column. Enzyme activity was eluted from column with 0.05 M glycine, pH 2.0, as indicated by the arrow. Protein was monitored by absorbance at 280 nm and the enzyme activity determined by fluorescence intensity (FI) released by hydrolysis of FITC-casein. (B) Fractions 46–50 eluted from the affinity column and a sample of the unfractionated extract (UF) were assayed by casein substrate gel electrophoresis. (C) Anion exchange (HPLC-AX) fractionation of proteolytic activity isolated by affinity column. The peak of proteolytic activity had retention time of 49 min. Fractions 48–51 were pooled and concentrated. (D) Casein substrate gel electrophoresis (right panel) of proteolytic activity purified by HPLC-AX (AX) and aliquot of sample from benzamidine affinity (BZ) prior to HPLC-AX. Analysis by gel electrophoresis and CB staining of fraction purified by HPLC-AX is shown in left panel. study does not indicate any transcription of the rhomboid pro- founding the measurement of proteolytic activity (Michalski et tease in the SZ, suggesting significant developmental regula- al., 1990). Protease inhibitors have not been previously inves- tion. tigated in E. tenella, but serpins and other types of serine pro- Proteolytic activity was demonstrated and examined in oo- tease inhibitors have been characterized in other apicomplexans cyst extracts from all time points collected, and Mr of bands (Lindh et al., 2001; Morris et al., 2002; Bruno et al., 2004) and observed on substrate gels were similar. However, the amount a gene for a serpin has been identified in E. tenella oocysts of activity was reduced in fully sporulated oocysts, suggesting (Miska et al., 2004). decreased activity associated with complete formation of spo- At least 3 bands of proteolytic activities were observed on rozoites. Alternatively, the decreased activity may not reflect a substrate gels of extracts of unsporulated oocysts, and these lack of proteases, but may be due to endogenous inhibitors. A bands were also enriched after affinity purification. However, previous investigation of serine protease activity in sporulated only 1 band of activity was observed on substrate gels after oocysts also reported difficulty in measuring activity in soluble further purification with anion exchange. The reason for the extracts, suggesting the presence of endogenous inhibitors con- inability to purify multiple proteases is unclear, but it is possible 338 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 2, APRIL 2007

Goldberg, 2002). The release of proteases could be due to ac- tive secretion or just the result of release from structurally com- promised parasites. The amount of ESP released is a temperature-sensitive process, suggesting that the protease secretion is a metabolically dependent process. The specific activity of protease is about 10 times greater in ESP compared to the MZ extract, as may be expected if the protease was being actively secreted and not just the result of proteins being released from degenerating parasites. Differing from the other developmental stages examined, SZ extracts contained only a single band of AEBSF-sensitive proteolytic activity. In contrast to a previous report (Fuller and McDougald, 1994), proteolytic activity was not detected in ESP or in culture media during in vitro cell invasion. This distinction may be due to differences in assay systems used in the 2 studies or represent degradation of SZ and the subsequent release of endogenous proteases into culture media. Although it is possible that SZ release proteases that function in survival within the host, the absence of protease activity in both the ESP and in the culture media during cell invasion suggests a functional role for serine protease within the SZ rather than a direct effect on the host. Our observation that the specific water-soluble, nontoxic serine protease inhibitor AEBSF prevents invasion of SZ into cultured cells, while reducing the release of the microneme protein MIC2 suggests a distinct function for proteases within the parasite. This observation is also consistent with recent evidence that implicates serine proteases in protein-processing events essential for cell invasion by apicomplexans (Kim, 2004; Caruthers and Blackman, 2005). The effect of a serine protease inhibitor on invasion is consistent with previous observations that serine protease inhibitors prevent invasion of E. tenella SZ (Fuller and McDougald, 1994) and T. gondii TZ in vitro (Conseil et al., 1999; Buitrago-Rey et al., 2002). How- ever, caution is required when interpreting the results of invasion studies utilizing inhibitors because the inhibitor alone may FIGURE 4. Casein substrate electrophoresis of soluble extracts from have detrimental effects on parasite survival. Even specific in- E. tenella sporozoites (SZ) or MZ in the absence (Ϫ) or presence of 1 mM AEBSF. hibitors such as AEBSF have the potential to inhibit other en- zymes such as esterases (Dentan et al., 1996). In the current study, AEBSF did not appear to have any great effect on SZ that some of the protease activities were not stable during prep- viability or motility that could be detected visually, although aration. Supporting this idea is the recent purification of a 45 the SZ appear somewhat rounder than normal after treatment kDa protease from unsporulated oocysts using a different pu- with AEBSF. Of interest is our observation that monensin, al- rification scheme (R. Fetterer, unpubl. obs.). The relationship though it completely inhibited SZ invasion, did not affect re- between the 25 kDa protease purified from unsporulated oo- lease of MIC2 into culture media. The release of MIC2 con- cysts and a 20 kDa protease that was previously purified from current with monensin treatment may be due to loss of integrity sporulated oocysts (Michalski et al., 1990) is unclear. It is pos- by parasites that were distorted and rendered immotile, sug- sible that the 20 kDa protein is related to a band of about 18 gesting that MIC2 release alone is not predictive of SZ’s ability kDa observed on many of the substrate gels. to invade cells. Both extracts and ESP from E. tenella MZ had AEBSF-sen- It is clear from the current results that alkaline proteases of sitive bands of proteolytic activity of a pattern similar to that the serine class are widely distributed in both the free-living observed in oocyst extracts. This differs from observations and intracellular stages of E. tenella. Although the present re- made in S. neurona, which had serine proteases with Mr be- sults support the previous work that indicates a role for serine tween 65 and 75 kDa in MZ extracts (Barr and Warner, 2003). protease in microneme processing and cell invasion, the func- However, in MZ extracts, an AEBSF-insensitive band with Mr tions and identities of the serine proteases in oocysts remain of about 90 kDa was often observed, suggesting some proteo- quite speculative. Since mechanisms of invasion are likely to lytic activity unrelated to serine protease is present in E. tenella be conserved, serine protease may be involved in the invasion MZ extracts. The function of protease in MZ ESP is unknown, process of MZ, but the presence of protease in the ESP suggests but it is possible that it interacts with the host tissues to play a a role for these proteases in host-parasite interactions. Studies role in immune invasion or some other host-parasite interaction, to express and characterize protease genes during development as has been suggested for other parasite proteases (Klema and and to further purify and identify proteases within developmen- FETTERER ET AL.—SERINE PROTEASES IN E. TENELLA DEVELOPMENTAL STAGES 339

FIGURE 5. Analysis of proteolytic activity from excretory-secretory products (ESP) from E. tenella merozoites (MZ). Casein substrate electrophoresis of ESP from MZ maintained at either 4 or 41 C in the absence (Ϫ) or presence (ϩ) of 1 mM AEBSF (left panel). Total protease activity determined by hydrolysis of FITC-casein of extract or ESP from E. tenella MZ (right panel).

FIGURE 6. The effect of AEBSF on cellular invasion and microneme release by E. tenella sporozoites (SZ). (A) Effect of AEBSF SZ invasion. One ␮g/ml monensin (Mon) served as a positive control. Values are means of 3 determinations, and error bars represent 1 SEM. (B) Western blot analysis demonstrating the effect of AEBSF on release of the microneme protein MIC2 by E. tenella SZ during in vitro cell invasion. Mon or AEBSF was applied to SZ as a pretreatment prior to cell invasion. AEBSFϩ indicates AEBSF was applied as a pretreatment and was also present in culture media during invasion. 340 THE JOURNAL OF PARASITOLOGY, VOL. 93, NO. 2, APRIL 2007 tal stages are required to delineate protease function in E. te- of Eimeria tenella (Coccidia) sporozoites by protease inhibitors and nella. partial characterization of proteolytic activity associated with intact sporozoites and merozoites. Journal of Parasitology 76: 464–467. KIM, K. 2004. Role of host protease in host cell invasion by Toxoplasma ACKNOWLEDGMENTS gondii and other Apicomplexa. Acta Tropica 91: 69–81. The authors are grateful for the expert technical assistance of Gary KLEMA,M.,AND D. E. GOLDBERG. 2002. Biological roles of proteases Wilkins, Jenifer Herrmann, Mark Paape, and Eli Miramontes. We are in parasitic protozoa. Annual Review of Biochemistry 71: 275– also indebted to M. Shirley for granting permission to use the E. tenella 305. genome sequence available at the Sanger Institute website. This work LI, J., X. ZHANG,Q.LUI,J.YIN, AND J. YANG. 2006. Eimeria tenella: was supported by USDA/ARS CRIS project 1265-31320-070-00D. cloning of novel Eimeria tenella cDNA encoding a protein related to rhomboid family from F2 hybrid strain. Experimental Parasitol- ogy 113: 215–220. LITERATURE CITED LILLEHOJ, H., S. R. A. DALLOUL,T.SATO,A.YASUDA, AND E. P. LIL- ALLEN,P.C.,AND R. H. FETTERER. 2002. Recent advances in biology LEHOJ. 2005. Embryo vaccination against Eimeria tenella and E. and immunobiology of Eimeria species and the diagnosis and con- acervulina infections using recombinant proteins and cytokine ad- trol of infection with these coccidian parasites of poultry. Clinical juvants. Journal of Parasitology 91: 666–673. Microbiological Reviews 15: 58–65. LINDH, J. G., S. BOTERO-KLEIVEN,J.I.ARBOLEDA, AND M. WAHLGREN. ALTSCUL, S. F., T. L. MADDEN,A.A.SHAFFER,J.ZHANG,Z.ZHANG,W. 2001. A protease inhibitor associated with the surface of Toxo- MILLER, AND D. J. LIPMAN. 1997. Gapped BLAST and PSI-BLAST: plasma gondii. Molecular and Biochemical Parasitology 116: 137– A new generation of protein database search programs. Nucleic 145. Acids Research 25: 3389–3402. MICHALSKI, W. P., J. K. CROOKS, AND S. L. PROWSE. 1994. Purification BARR,S.C.,AND K. WARNER. 2003. Characterization of a serine pro- and characterization of a serine-type protease from Eimeria tenella tease activity in Sarcocystis neurona merozoites. Journal of Para- oocysts. International Journal for Parasitology 24: 189–195. sitology 89: 385–388. MILLER, S. A., E. M. BINDER,M.J.BLACKMAN,V.B.CARRUTHERS, AND BRUNO, S., V. G. DUSCHAK,B.LEDESMA,M.FERELLA,B.ADERSSON,E. K. KIM. 2001. A conserved subtilisin-like protein TgSub1 in mi- A. GUARNERA, AND S. O. ANGEL. 2004. Identification and charac- croneme organelles of Toxoplasma gondii. Journal of Biological terization of serine protease inhibitors from Neospora caninum. Chemistry 276: 45341–45348. Molecular and Biochemical Parasitology 136: 101–107. ———, V. THATY,J.W.AJIOKA,M.J.BLACKMAN, AND K. KIM. 2003. BUITRAGO-REY, R., J. OLARTE, AND J. E. GOMEZ-MARTIN. 2002. Evalu- TgSUB2 is a Toxoplasma gondii rhoptry organelle processing pro- ation of two inhibitors of invasion: LY311727{3-(3-acetamied-1- teinase. Molecular Microbiology 49: 883–894. benzyl-e-ethyl-inolyl-5-oxy) propane phosphonic acid} and AEBSF MISKA, K. B., R. H. FETTERER, AND R. C. BARFIELD. 2004. Analysis of [4-(2-aminoethyl)-benzesulphonyl fluoride] in acute murine toxo- transcripts expressed by Eimeria tenella oocysts using subtractive plasmosis. Journal of Antimicrobial Chemotherapy 49: 871–874. hybridization. Journal of Parasitology 90: 1245–1252. CARRUTHERS,V.B.,AND M. J. BLACKMAN. 2005. A new release on life: MORRIS,M.T.,A.COPPIN,S.TOMAVO, AND V. B . C ARRUTHERS. 2002. Emerging concepts in proteolysis and parasite invasion. Molecular Functional analysis of Toxoplasma gondii protease inhibitor 1. Microbiology 55: 1617–1630. Journal of Biological Chemistry 277: 45259–45266. CONSEIL, V. M., M. SOETE, AND J. F. DUBREMETZ. 1999. Serine protease O’DONNELL,R.A.,AND M. J. BLACKMAN. 2005. The role of malaria inhibitors block invasion of host cells by Toxoplasma gondii. An- merozoite proteases in red blood cell invasion. Current Opinion in timicrobial Agents and Chemotherapeutics 43: 358–361. Microbiology 8: 422–427. DENTAN, C., A. O. TSELEPIS,M.J.CHAPMAN, AND E. NIRO. 1996. Pe- RAWLINGS,N.D.,D.P.TOLLE, AND A. J. BARRETT. 2004. MEROPS: fabloc, 4-[2-aminoethyl]benzesulfonyl fluoride, is a new potent The peptidase database. Nucleic Acids Research 32(Database is- nontoxic and irreversible inhibitor of PAF-degrading acetylhydro- sue): D160–D164. lases. Biochimica et Biophysica Acta 1299: 353–357. SHAW, M. K., D. S. ROOS, AND L. G. TILNEY. 2002. Cysteine and serine DOWSE,T.J.,AND D. SOLDATI. 2005. Rhomboid-like proteins in Api- protease inhibitors block intracellular development and disrupt the complexa: Phylogeny and nomenclature. Trends in Parasitology 21: secretory pathway of Toxoplasma gondii. Microbes and Infections 254–258. 4: 119–132. FETTERER,R.H.,AND R. C. BARFIELD. 2003. Characterization of a de- TOMLEY, F., J. M. BUMSTEAD,K.J.BILLINGTON, AND P. P. J . D UNN. 1996. velopmentally regulated oocyst protein from Eimeria tenella. Jour- Molecular cloning and characterization of a novel microneme pro- nal of Parasitology 89: 553–564. tein (Etmic-2) from the apicomplexan protozoan parasite, E. tenel- ———, K. B. MISKA, AND R. C. BARFIELD. 2005. Partial purification la. Molecular Biochemistry and Parasitology 79: 195–206. and characterization of an aminopeptidase from Eimeria tenella. WILLIAMS, R. B. 1998. Epidemiological aspects of the use of live an- Journal of Parasitology 91: 1280–1286. ticoccidial vaccines for chickens. International Journal for Parasi- ———, ———, M. C. JENKINS, AND R. C. BARFIELD. 2004. A con- tology 28: 1089–1098. served 19 kDa Eimeria tenella antigen is a profilin-like antigen. WITHERS-MARTINEZ, C., L. JEAN, AND M. J. BLACKMAN. 2004. Subtilisin- Journal of Parasitology 90: 1321–1328. like proteases from the malaria parasite. Molecular Microbiology FULLER,A.L.,AND L. R. MCDOUGALD. 1990. Reduction in cell entry 53: 55–63.