A cathepsin B-like protease is required for host protein degradation in Trypanosoma brucei.

Identification and analysis of Clan CA (papain) cysteine proteases in primitive protozoa and metazoa have suggested that this enzyme family is more diverse and biologically important than originally thought. The protozoan parasite Trypanosoma brucei is the etiological agent of African sleeping sickness. The cysteine protease activity of this organism is a validated drug target as first recognized by the killing of the parasite with the diazomethane inhibitor Z-Phe-Ala-CHN(2) (where Z is benzyloxycarbonyl). Whereas the presumed target of this inhibitor was rhodesain (also brucipain, trypanopain), the major cathepsin L-like cysteine protease of T. brucei, genomic analysis has now identified tbcatB, a cathepsin B-like cysteine protease as a possible inhibitor target. The mRNA of tbcatB is more abundantly expressed in the bloodstream versus the procyclic form of the parasite. Induction of RNA interference against rhodesain did not result in an abnormal phenotype in cultured T. brucei. However, induction of RNA interference against tbcatB led to enlargement of the endosome, accumulation of fluorescein isothiocyanate-transferrin, defective cytokinesis after completion of mitosis, and ultimately the death of cultured parasites. Therefore, tbcatB, but not rhodesain, is essential for T. brucei survival in culture and is the most likely target of the diazomethane protease inhibitor Z-Phe-Ala-CHN(2) in T. brucei.

Identification and analysis of Clan CA (papain) cysteine proteases in primitive protozoa and metazoa have suggested that this enzyme family is more diverse and biologically important than originally thought. The protozoan parasite Trypanosoma brucei is the etiological agent of African sleeping sickness. The cysteine protease activity of this organism is a validated drug target as first recognized by the killing of the parasite with the diazomethane inhibitor Z-Phe-Ala-CHN 2 (where Z is benzyloxycarbonyl). Whereas the presumed target of this inhibitor was rhodesain (also brucipain, trypanopain), the major cathepsin L-like cysteine protease of T. brucei, genomic analysis has now identified tbcatB, a cathepsin B-like cysteine protease as a possible inhibitor target. The mRNA of tbcatB is more abundantly expressed in the bloodstream versus the procyclic form of the parasite. Induction of RNA interference against rhodesain did not result in an abnormal phenotype in cultured T. brucei. However, induction of RNA interference against tbcatB led to enlargement of the endosome, accumulation of fluorescein isothiocyanate-transferrin, defective cytokinesis after completion of mitosis, and ultimately the death of cultured parasites. Therefore, tbcatB, but not rhodesain, is essential for T. brucei survival in culture and is the most likely target of the diazomethane protease inhibitor Z-Phe-Ala-CHN 2 in T. brucei.
The Clan CA cysteine proteases include the papain-related proteases of plants and the lysosomal cathepsins of mammalian cells. Recent genomic and biochemical studies of protozoa and primitive metazoa suggested that there was an explosion of gene diversity in this family coincident with the evolution of the first eukaryotic cell (1). In contrast to higher plants and most mammalian cells, the cysteine proteases of primitive protozoa and metazoa function in a variety of chemical environments and cellular compartments (2,3). Therefore, the biolog-ical importance and distribution of cysteine proteases is much greater than originally proposed for the cathepsins.
Because primitive protozoa include many of the major parasitic organisms of humans and domestic animals, parallel studies of the pathogenesis of parasitic diseases have underscored the importance of Clan CA proteases as virulence factors. Furthermore, considerable progress has been made in targeting these proteases for the development of new antiparasitic chemotherapy (2, 4 -7). One notable observation in this regard was that the benzyloxycarbonyl-phenylalanine-alanine diazomethane cysteine protease inhibitor (Z-Phe-Ala-CHN 2 ) 1 was lethal to Trypanosoma brucei, the causative agent of African sleeping sickness, in vitro and in vivo (8,9). The presumptive target for this inhibitor was a cathepsin L-like protease, rhodesain, isolated from T. brucei rhodesiense and the homologue of brucipain, trypanopain, or congopain in other Trypanosoma species (10,11). However, peptide diazomethanes are relatively nonselective irreversible cysteine protease inhibitors, and recent genomic analysis of trypanosomes suggested that a repertoire of Clan CA protease genes might be present.
To identify and validate the target of Z-Phe-Ala-CHN 2 in T. brucei, as well as to clarify its biological function, a genomic scan using papain as a probe was first undertaken. The previously characterized cathepsin L-like cysteine protease, rhodesain (12), was identified as expected, but a second cathepsin B-like gene product was also discovered. Subsequent expression, biochemical analysis, and localization studies suggested that the cathepsin B-like protease in T. brucei was a plausible target for the diazomethane inhibitor. Therefore, to clarify the roles of both the T. brucei cathepsin B and cathepsin L homologues, RNA interference (RNAi) was used in conjunction with radiolabeled active site probes. Subcloning of TbcatB cDNA into Escherichia coli-Reverse transcription polymerase chain reactions were carried out using the One

Identification of Clan CA Cysteine Proteases in the T. brucei Ge
Step with Platinum Taq kit (Invitrogen). Two micrograms of total RNA from T. brucei were mixed with the splice leader primer, TbSL 5Ј-ATTATTAGAACAGTTTCTGTACTATATTG-3Ј, plus the reverse primer, TbCatBXhoIR 5Ј-GCTAATATCTCAGATACGCCGTGTTGGG-TGCA-3Ј, to amplify the full-length cDNA of tbcatB. To amplify the open reading frame of tbcatB for recombinant expression, the primers TbCatBBHIF, 5Ј-GTCTATAGGATCCATGCATCTCATGCGTGCCT-3Ј, and TbGSTHisR 5Ј-GTCGACGAGATGGAGATGGAGATGCGCCGTG-TT-3Ј, were used. All of the amplified products were subcloned into the TOPO TA cloning vector and transformed into competent E. coli for propagation (Invitrogen).
Purification of Recombinant tbcatB from E. coli-One liter of Terrific Broth (Invitrogen) containing 100 g/ml ampicillin was inoculated with 5 ml of E. coli expressing the pGSTbcatBHis plasmid. The E. coli were incubated at 37°C until the A 600 reached 1.0 and then induced with isopropyl-1-thio-␤-D-galactopyranoside to a final concentration of 1 mM for 5 h. Bacteria were centrifuged at 9,000 ϫ g for 15 min using a Beckman JLA 10.5000 rotor. After centrifugation, the pellet was resuspended in 50 ml of sonication buffer (150 mM NaCl, 1% Nonidet P-40, 50 mM Tris, pH 7.5, 10% glycerol) and sonicated on ice five times for 2-min intervals with 3-min breaks between each interval. The crude lysate was cleared by centrifugation for 15 min at 25,000 ϫ g. The cleared supernatant was decanted, and the pellet was resuspended in 50 ml of guanidine HCl buffer (6 M guanidine HCl, 20 mM Tris, pH 8.0, 100 mM NaH 2 PO 4 ) for 2 h at room temperature. After 2 h, the lysate was cleared by centrifugation, and 10 ml of equilibrated nickel-agarose beads (Qiagen) in 50% slurry were added to the cleared lysate. The beads were incubated with the lysate for 3 h at room temperature. The mixture was poured into a 10-ml gravity column and washed with guanidine HCl buffer, pH 8.0. The recombinant 6-His tagged protein was eluted from the beads according to the manufacturer's manual (Qiagen). Three microliters from each fraction were mixed with 2 l of SDS loading dye and examined by SDS-PAGE. Peak fractions were pooled and dialyzed three times in 1 liter of dialysis buffer (50 mM Tris,pH 7.5, 50 mM NaCl, 3.0 M urea), and 1.5 mg of protein was sent to AnimalPharm, Inc. for production of rabbit polyclonal antibodies.
Northern Blot Analysis of T. brucei RNA-T. brucei total RNA was purified from 10 8 parasites using TRIzol reagent (Invitrogen). Thirty micrograms of total RNA were loaded on a 1.2% agarose-formaldehyde gel and resolved by electrophoresis at 50 V for 2.5 h. After electrophoresis, the RNA was transferred to polyvinylidene diflouride membrane and cross-linked to the membrane with a Stratalinker (Stratagene). 32 P-Labeled cDNA probes were generated using the random primed labeling kit (Amersham Biosciences). The probes were denatured at 100°C and hybridized to the blot overnight at 42°C in buffer containing 50% formamide, 5ϫ SSC, 4ϫ Denhardt's solution, 0.1% SDS, and 0.1% sodium pyrophosphate.
Culturing of T. brucei-Procyclic T. brucei strain 427 was incubated at 27°C without carbon dioxide and maintained in complete Cunningham's media containing 10% fetal bovine serum and 1ϫ penicillin/ streptomycin. The transgenic bloodstream form T. brucei clone 90-13 was a gift from the laboratory of George A. M. Cross (15). Bloodstream parasites were incubated in 5% CO 2 at 37°C in HMI-9 medium containing 10% fetal bovine serum, 10% Serum Plus (Omega Scientific), 1ϫ penicillin/streptomycin with 5.0 g/ml hygromycin B, and 2.5 g/ml G418.
Introduction of RNAi Transgenes into Bloodstream T. brucei by Electroporation-For electroporation, 10 8 parasites were pelleted by centrifugation and washed twice with 10 ml of cytomix (120 mM KCl, 150 M CaCl 2 , 10 mM K 2 HPO 4 /KH 2 PO 4 , pH 7.6, 25 mM HEPES, pH 7.6, 2 mM EGTA, pH 7.6, 5 mM MgCl 2 , pH 7.6, adjusted with KOH) (16). After the second wash, the parasites were resuspended in 0.5 ml of cytomix. The plasmid pZJM, which allows for stable tetracycline-inducible expression of double-stranded RNA in T. brucei, was a gift from the laboratory of Paul T. Englund (17). One hundred micrograms of pZJM vector containing the first 597 nucleotides of the rhodesain cDNA (pZJMT-bRho) or nucleotides 300 -1020 of the tbcatB cDNA (pZJMTbCB) were linearized with NotI restriction endonuclease and precipitated with ethanol. The DNA was resuspended in 100 l of cytomix buffer and mixed with the 0.5-ml suspension of T. brucei in a 4-mm electroporator cuvette. The parasites were pulsed with 1.7 kV and 25 microfarads. After pulsing, the parasites were transferred to 24 ml of complete medium and incubated overnight at 37°C with 5% CO 2 . To select for parasites stably expressing the transgenes, phleomycin was added to the parasites at a concentration of 2.5 g/ml, and the parasites were aliquoted in a 24-well plate for 7-12 days. Stably transfected parasites were maintained in complete medium containing 2.5 g/ml phleomycin. Induction of the double-stranded RNA in bloodstream parasites was carried out by adding tetracycline to a final concentration of 100 ng/ml.
After transferring and blocking, the polyvinylidene difluoride membranes were incubated with rabbit anti-tbcatB (1:2,000 dilution) or anti-rhodesain antiserum (1:5,000 dilution) (12) for 1 h and washed three times for 5 min with TBST (10 mM Tris, pH 7.4, 150 mM NaCl, 0.4% Tween 20). After the third wash, horseradish peroxidase-conjugated donkey anti-rabbit IgG (1:1,000 dilution) was added to the blots for 1 h. The blots were then washed again in the same buffer three times for 5 min and then examined by ECL (Amersham Biosciences).
Fluorescence-activated Cell Sorter (FACS) Analysis and 4,6-Diamidino-2-phenylindole (DAPI) Staining-For FACS analysis, parasites were fixed overnight in 70% methanol and 30% phosphate-buffered saline at 4°C. Afterward, the parasites were washed with cold phosphate-buffered saline and resuspended to 10 6 cells/ml. RNase A and propidium iodide were added to the cells to a final concentration of 10 g/ml each and incubated at 37°C for 60 min. FACS analysis was performed with a Becton Dickinson FACScalibur using FL2-H. Data analysis was carried out with either FlowJo version 4.52 or CellQuest version 3.3.
Analyzing Degradation Fluorescein Isothiocyanate (FITC)-transferrin in T. brucei-FITC-conjugated bovine transferrin (Sigma) was resuspended at 1 mg/ml in 10 mM HEPES buffer, pH 7.4. Parasites were incubated in HMI-9 medium containing 15 g/ml FITC-conjugated bovine transferrin at 37°C and 5% carbon dioxide for 3 h. The parasites were then washed twice in phosphate-buffered saline, pH 7.4, and fixed in 0.1 M cacodylate buffer containing 2.5% glutaraldehyde and 1.6% paraformaldehyde and applied to poly-L-lysine-coated slides. Parasites were visualized using an Axioskop2 microscope (Zeiss).
Identification of Clan CA Cysteine Protease Active Sites with 125 I-Labeled Inhibitors-Ten million T. brucei cultured in 10 ml of complete HMI-9 medium at 37°C and 5% carbon dioxide were incubated with 125 I-JPM565 for 2 h (18). After 2 h, the parasites were pelleted by centrifugation at 900 ϫ g for 20 min with a Beckman GH 3.8A rotor and lysed in 100 l of lysis buffer (50 mM sodium acetate, 1 mM EDTA, 1% Triton X-100, pH 5.5) for 1 h on ice. Alternatively, 10 million unlabeled parasites were pelleted and lysed using identical conditions. Lysates were cleared by centrifugation at 16,000 ϫ g for 15 min at 4°C in Eppendorf tubes. The protein concentration of the cleared lysates was determined by Bradford assay, and equal amounts of parasite lysates were labeled with 125 I-MB-074 for 1 h at 25°C (19). Quantification of labeled enzymes was determined by PhosphorImager analysis (Molecular Dynamics).
T. brucei Viability Assay-Uninduced control or tetracycline-induced T. brucei were grown in a T-25 flask under standard conditions for 48 h. After 48 h, 100 l of the cultures were aliquoted into white opaque 96-well plates. To assay for viability, an equal volume of CellTiter-Glo TM (Promega) was added to the corresponding wells. The mixture was placed on an orbital shaker at room temperature for 5 min before reading the plates with a SpectraFluor Plus multidetection plate reader (Tecan). Alternatively, 48 h after tetracycline induction, the parasites were counted by hemocytometer.

Identification of TbcatB, a Cathepsin B-like Homologue in
T. brucei-Using the highly conserved catalytic region of papain as a probe, we searched T. brucei data bases for homologues to papain. From this search, cathepsin L-like and cathepsin B-like protease sequences were identified. The cDNAs encoding these two gene products were amplified from T. brucei RNA by reverse transcriptase PCR, subcloned, and sequenced. The 1.3-kb cDNA encoded a 450-amino acid polypeptide with a predicted M r of 48,431 identical to rhodesain (or brucipain or trypanopain), the previously characterized cathepsin L-like protease identified by biochemical assays in T. brucei lysates (20). The 1.0-kb cDNA encoded a cathepsin B-like protease, tbcatB, having 341 amino acids with a predicted M r of 37,223. The tbcatB open reading frame contained each of the conserved motifs identified in the active site of lysosomal cathepsins including the residues that form the catalytic triad, Cys-42, His-282, and Asn-302 (tbcatB numbering). The tbcatB open reading frame also encodes two conserved motifs identified in cathepsin B family proteases: Gly-Cys-Xaa-Gly-Gly, which is identical to human cathepsin B (residues 70 -74), and an occluding loop motif that is thought to be responsible for the exopeptidase activity of cathepsin B-like proteases (21) (Fig. 1).

tbcatB Essential for T. brucei Survival in Culture
The tbcatB sequence has been deposited in the GenBank TM data base under accession number AY508515.
Examination of Rhodesain and TbcatB mRNA Expression in T. brucei-Rhodesain has been characterized extensively with synthetic peptides demonstrating that it was the major cysteine protease activity in T. brucei (22)(23)(24). Consistent with it being the most abundantly expressed cysteine protease gene product in T. brucei, genomic analysis has indicated that there were between 10 and 20 copies of the rhodesain gene repeated in tandem (25). We examined the mRNA expression of rhodesain and tbcatB to elucidate their relative abundance in the two major developmental stages of T. brucei. Northern blot analysis demonstrated that the amounts of rhodesain mRNA expressed in either the bloodstream (mammalian host stage) or the procyclic (insect stage) forms of the parasite were equivalent ( Fig. 2A). In shorter exposures of the autoradiography, the rhodesain mRNA could be visualized as a doublet in the RNA from both forms of the parasite. These mRNA results were surprising because studies from previous protein expression assays concluded that rhodesain protein was up-regulated in the bloodstream-form parasites (12). It is therefore reasonable to conclude that rhodesain, like its homologue in Trypanosoma cruzi, may be translationally regulated in procyclic parasites. We used gene-specific probes to hybridize to tbcatB mRNA for Northern blot analysis. Our results demonstrated that tbcatB mRNA was less abundantly expressed than rhodesain mRNA in both developmental stages of T. brucei; however, the sizes of the two transcripts were very similar. To avoid ambiguous results, we carried out the tbcatB and rhodesain blots in parallel rather than using blots that were stripped and reprobed. Northern blot analysis also demonstrated that the message for tbcatB was up-regulated in the bloodstream-form parasites (Fig. 2B). The differential expression of tbcatB observed in the bloodstream-form versus procyclic parasites suggests that in contrast to rhodesain, tbcatB is regulated at the transcriptional level. Up-regulation of tbcatB mRNA in the bloodstream-form parasites suggests that it may primarily function in the bloodstream parasites.
Silencing of Rhodesain mRNA Expression in T. brucei-To examine the consequences of inhibiting the activity of each protease in T. brucei, RNAi was used to silence the mRNA expression of rhodesain or tbcatB. We used tetracycline to induce RNAi in T. brucei, stably expressing the rhodesain, pZJMTbRho (or tbcatB), or pZJMTbCB transgenes. The tetracycline-induced pZJMTbRho parasites did not exhibit visible phenotypic abnormalities, signs of cell cycle defects, or changes in growth rate compared with the uninduced parasites. Total RNA from induced and control parasites was extracted and analyzed by Northern blot analysis 24 h postinduction. The rhodesain probe hybridized with two distinct messages on the Northern blot with about equal intensity. Because the Northern blot analysis was carried out on total RNA, the two transcripts recognized by the probe could have arisen if the rhodesain gene lay within a cluster of related genes as described for the cathepsin L-like genes of Leishmania species (26). Alternatively, the two transcripts may represent a mixture of mature rhodesain mRNAs and either immature or alternatively processed transcripts. In either case, both of these transcripts were reduced to almost undetectable levels after tetracyclineinduced RNAi was done in these parasites (Fig. 3A). Immunoblot analysis was used to compare the amount of rhodesain in control parasites with the amount in RNAi-silenced parasites. Affinity-purified rabbit anti-rhodesain antibodies recognized two bands, 45 and 47 kDa, in the Western blots of control parasites. These two bands could reflect that rhodesain has become glycosylated because a cryptic glycosylation site has been identified (12) and these antibodies do not distinguish between glycosylated and non-glycosylated forms of rhodesain.
Recently, it has been demonstrated that a species of mammalian cathepsin L is able to initiate translation from a second tbcatB Essential for T. brucei Survival in Culture downstream methionine codon and become translocated into the nucleus (27). Rhodesain is a cathepsin L-like protease that also contains several downstream methionine codons where translation can initiate. Initiation at a downstream methionine codon may also account for the second species of rhodesain detected by the antibodies. In either case, these two bands were nearly undetectable by the same antibodies in extracts where rhodesain expression was silenced in parasites by RNAi (Fig. 3B).
To confirm the effects of mRNA silencing on protease activity in vivo, we used JPM565, a derivative of E-64, the irreversible inhibitor of papain and other cysteine proteases. E-64 and its derivatives form a thioether bond with the sulfhydryl group in the active center of cysteine proteases, and they are therefore ideal for active site titration because 1 mol of the inhibitor inhibits 1 mol of protease (28). Furthermore, JPM565 has a preference for cathepsin L-like proteases and can be iodinated allowing quantitative analysis to be carried out on labeled proteases (18,29). When control parasites were incubated with 125 I-JPM565, both species of rhodesain were detected by the label (Fig. 3C). The tetracycline-induced parasites showed a 65% reduction in labeling of rhodesain by 125 I-JPM565 (Fig.  3D). Together these observations suggested that rhodesain was not essential for parasite replication in culture.
Knock Down of TbcatB Leads to Dysmorphism and Death of T. brucei-When the pZJMTbCB clones were induced with tetracycline, distension of the posterior endosome/lysosome compartment of the parasites was easily visualized within 12 h postinduction (Fig. 4). Distension of these RNAi-induced parasites in the region of the endosome reached a maximum by 24 h postinduction. In addition to posterior swelling, some RNAiinduced parasites displayed multiple flagella that looked similar to uninduced control parasites undergoing cytokinesis (Fig.  4A, Phase Contrast). To examine the consequences of endosome/lysosome dysfunction in the parasites, we used DAPI to stain nuclear DNA and kinetoplast DNA (which is the mitochondrial DNA equivalent in T. brucei) of control and induced parasites. In uninduced controls, the population was composed of non-dividing parasites containing one kinetoplast and one nucleus, parasites undergoing kinetoplast division and segregation containing two kinetoplasts and one nucleus, and parasites undergoing nuclear mitosis containing two kinetoplasts and two nuclei. These observations were consistent with the control parasites existing as an asynchronous population (30). The nuclei of RNAi-induced pZJMTbCB clones were stained with DAPI 24 h postinduction. These clones demonstrated a dramatic increase in the number of parasites containing two kinetoplasts and two nuclei, but no parasites containing one kinetoplast and one nucleus were detected 48 h postinduction (Fig. 4A, DAPI).
The DNA content of control or induced parasites was then analyzed by flow cytometry to verify the karyotype results observed by microscopy. In the uninduced control population, a similar number of parasites contained either non-replicating 2C or postmitotic 4C DNA. A similar profile was observed in the parental 90-13 and wild-type 221 bloodstream-form parasites consistent with uninduced parasites existing as an asynchronous population (data not shown). After 24 h of tetracycline induction, the majority of parasites observed contained 4C DNA or greater, suggesting that the tetracycline-induced population of parasites was arrested in cytokinesis after completing several rounds of DNA synthesis and mitosis (Fig. 4A,  FACS). No proliferation in the pZJMTbCB clones was detected by either direct counting with a hemocytomer or by a luciferase-based proliferation assay that detects adenosine triphosphate generation by 48 h postinduction (Fig. 4B) (31). The tetracycline-induced parasites died by 72 h postinduction. These abnormalities were not observed in uninduced control parasites where normal proliferation occurred. We carried out dosage-dependent induction of the parental 90-13 strain versus pZJMTbCB clones. After 48 h of induction, the proliferation of the 90-13 strain was not affected at any of the concentrations of tetracycline, indicating that this drug does not produce any toxic effects on T. brucei. Proliferation of the pZJMTbCB clones was inhibited at tetracycline concentrations as low as 1 ng/ml, showing dosage-dependent inhibition of proliferation by RNAi. The parasites that remain at 1 ng/ml of tetracycline may represent those that inefficiently produce RNAi having a more modest reduction in mRNA for tbcatB (Fig. 4C).
The effect of RNAi on tbcatB protease in the parasites was examined by Western blot analysis. A modest but significant decrease in the amount of tbcatB protein was detected in the extracts of the induced parasites versus the control parasites (Fig. 4D). We used the cathepsin B-specific inhibitor MB-074 to confirm reduction of cathepsin B activity in the tetracyclineinduced clones. MB-074 is a derivative of CA-074 that binds specifically to the occluding loop found in cathepsin B-like FIG. 2. Northern blot analysis of bloodstream and procyclic forms of T. brucei. Total RNA from bloodstream (BF) or procyclic (PF) forms of T. brucei were cross-linked to polyvinylidene difluoride membrane and then hybridized with rhodesain-specific (A) or tbcatB-specific (B) probes. Ribosomal RNA loading controls are shown beneath each blot. Note that while the message for tbcatB is less abundant than rhodesain, it is differentially up-regulated in the mammalian bloodstream form.
tbcatB Essential for T. brucei Survival in Culture proteases but not other cysteine proteases (32). MB-074 has a tyramine group that allows it to be labeled by 125 I. The labeled inhibitor was added to crude lysates of control or tetracyclineinduced pZJMTbCB clones and resolved by SDS-PAGE for quantitative analysis. A 34-kDa band consistent in size with the band recognized by tbcatB antibodies was labeled by 125 I-MB-074 in both extracts, but a slight decrease in the 34 kDa band was visualized in the extracts of the induced parasites (Fig. 4E). When the intensities of the labeled bands from the extracts were compared by PhosphorImager analysis, a 32% reduction of the 34 kDa protein was observed in the extracts of the RNAi-induced parasites (Fig. 4F).

TbcatB but Not Rhodesain Is Required for Degradation of
Transferrin-Because they lack cytochromes, the bloodstream form of T. brucei acquires iron from the host by internalizing transferrin through receptor-mediated endocytosis. The host transferrin is then rapidly degraded in the "endosome/lysosome" located between the nucleus and kinetoplast of the parasite. Because of the observation that tbcatB RNAi led to swelling of this compartment, FITC-transferrin was used to assay the effect that knocking down tbcatB had on the ability of the parasites to degrade this key host protein. No accumulation of FITC-transferrin was detected in uninduced control pZJMT-bCB clones indicating that the ability of the parasites to de- tbcatB Essential for T. brucei Survival in Culture grade transferrin was not hindered. However, when the clones were induced by tetracycline, FITC-transferrin began to accumulate throughout the parasite and within the endosome/lysosome, suggesting that their ability to efficiently and effectively degrade transferrin was hindered (Fig. 5A). A similar accumulation of FITC-transferrin was also previously reported when T. brucei were treated with the cysteine protease inhibitor Z-Phe-Ala-CHN 2 (8). No accumulation of FITC-transferrin was detected in control pZJMTbRho parasites or in tetracyclineinduced pZJMTbRho clones, indicating that knocking down rhodesain activity in the parasites does not interfere with the ability of the parasites to degrade transferrin (Fig. 5B). DISCUSSION Scory et al. (8) showed that a peptide diazomethane inhibitor was lethal to T. brucei in culture and produced swelling of the endosome/lysosome compartment of T. brucei. The presumptive target of this inhibitor was rhodesain, the major proteolytic activity and cathepsin L-like protease of the parasite. We have identified a cathepsin B-like homologue, tbcatB, in the T. brucei genome that confirms the Clan CA cysteine protease repertoire of T. brucei is more complex than previously thought. The tbcatB transcript is similar in size to the rhodesain transcript, but the amount of the message and protease is relatively low in comparison with the amount of rhodesain message and protein. This may explain why tbcatB was not identified in previous studies where synthetic peptides were used to characterize protease activities in T. brucei.
Whereas rhodesain mRNA was expressed at equivalent levels in both digenic stages of T. brucei, the mRNA of tbcatB was differentially up-regulated in the bloodstream-form parasites. Previous studies examining the expression of cathepsin-B and L-like cysteine proteases in related kinetoplastids showed that the mRNA for the cathepsin L-like protease in Leishmania species was up-regulated in the intracellular amastigote stage, but the mRNA for the cathepsin B-like cysteine protease was equivalently expressed in the promastigote and amastigote stages (33)(34)(35). In T. cruzi, the mRNAs for cathepsin L and cathepsin B-like proteases are both up-regulated in the bloodstream trypomastigote stage (36,37). However, as shown here for T. brucei, the cathepsin L-like protease of T. cruzi represents the most abundant cysteine protease activity and is posttranscriptionally up-regulated in the intracellular amastigote stage.
Because both proteases are expressed in the bloodstream form of T. brucei, it is conceivable that they represent a redundant system of host-protein degradation. The observation that T. brucei bloodstream forms degrade host proteins in an unusual endosome/lysosome compartment was thought to differ from the intracellular protein degradation classically attributed to the Clan CA family proteases of mammalian and plant cells. Recently, similar phenomena have been observed in mammalian systems where extracellular proteins were degraded by cathepsins either exocytosed from lysosomal compartments (38) or released through a regulated secretory pathway (13). Taken together, these results suggest that the Clan CA cysteine protease family evolved very early to carry out a broad array of biologic functions in primitive eukaryotic cells, some of which have been retained by higher eukaryotes. tbcatB Essential for T. brucei Survival in Culture The differential regulation of tbcatB and rhodesain suggests that these two proteases may have distinct biological roles. To address this issue in T. brucei, we utilized RNA interference to selectively knock down rhodesain or tbcatB activity in T. brucei. In cultured bloodstream trypanosomes, no observable phenotype was found after nearly complete knockdown of rhodesain message, protein, and proteolytic activity. In contrast, induction of RNAi in the pZJMTbCB clones produced a modest 32% reduction in protein and protease activity but led to a distinctive dysmorphic phenotype. Dilation of the anterior end and a block in cytokinesis of the tbcatB knockdown parasites produced a tadpole-like morphology with an enlarged lysosome/ endosome compartment. These parasites were able to complete multiple rounds of genomic replication and mitosis as evidenced by the appearance of multiple kinetoplasts and nuclei but were not able to complete cytokinesis. This defect in cytokinesis is reminiscent of that seen and reported for ␣-tubulindepleted parasites (14). It is possible that tbcatB could also be involved in a microtubule-related event alternatively; lack of cytokinesis may be an indirect consequence of iron depletion because of the inability of the parasite to degrade transferrin. The fact that a dysmorphic and later lethal phenotype was observed with modest reduction in the total protease suggests that this enzyme may be slowly replenished or be at a critical steady-state concentration in the parasites.
Finally, the lethal tbcatB RNAi effect associated with inhibition of FITC-transferrin degradation recapitulated the phenotype seen with parasites exposed to a peptide diazomethane cysteine protease inhibitor. This suggests that tbcatB was the most probable target of that inhibitor and plays a major role in host serum protein degradation by the parasite.
T. brucei is the causative agent of African sleeping sickness, a major health problem in sub-Saharan Africa and one of the great neglected diseases. Identification of a specific enzyme required for the viability of bloodstream trypanosomes represents an exploitable target for the development of new chemotherapy.