Cathepsin L Is Essential for Embryogenesis and Development of Caenorhabditis elegans *

Cysteine proteases play critical biological roles in both intracellular and extracellular processes. We characterized Ce-cpl-1 , a Caenorhabditis elegans cathepsin L-like cysteine protease. RNA interference with Ce-cpl-1 activity resulted in embryonic lethality and a transient delayed growth of larvae to egg producing adults, suggesting an essential role for cpl-1 during embryogenesis, and most likely during post-embryonic development. Cpl-1 gene ( Ce-cpl-1:lacZ ) is widely expressed in the intestine and hypodermal cells of transgenic worms, while the fusion protein (Ce-CPL-1::GFP) was expressed in the hypodermis, pharynx, and gonad. The CPL-1 native protein accumulates in early to late stage embryos and be-comes highly concentrated in gut cells during late embryonic development. CPL-1 is also present near the periphery of the eggshell as well as in the cuticle of larval stages suggesting that it may function not only in embryogenesis but also in further development of the worm. Although the precise role of Ce -CPL-1 during embryogenesis is not (cid:1) , to positions (cid:4) 2525 to (cid:4) 2504 relative to the ATG start codon, and OVc- plp3 (cid:1) ( Xma I), antisense, 5 (cid:1) -GACCCGGGTGAGTTTTTTTCTGTTTCTG-TTTTCC-3 (cid:1) , complementary to positions (cid:2) 1 to (cid:4) 25 relative to the ATG start codon. Following restriction enzyme digestion, the PCR fragment was cloned into lacZ reporter vector pPD90.23 containing a nuclear localization signal (kindly supplied by A. Fire). The fusion gene containing the cpl putative promoter was designated Ov-cpl:lacZ (pSL117A). Using a heterologous transformation, we created C. elegans transgenic lines expressing the O. volvulus promoter as described in generation and expression of C. elegans reporter gene constructs.

Cysteine proteases of the papain superfamily have long been recognized for their role in intracellular and extracellular protein degradation in a range of cellular processes (1). Within the papain family, the cathepsins can be subdivided into more than 10 subfamilies on the basis of their primary sequence and enzymatic activity (2). The family includes cathepsin B, C, L, and Z, all of which contain an essential cysteine residue in their active site but differ in tissue distribution and in some enzymatic properties, such as substrate specificity and pH stability. Cathepsin B-like cysteine protease genes occur as a large multigene family in a wide range of parasitic and free-living nematodes. Several cathepsin B genes were reported to be expressed in Caenorhabditis elegans, some of which were restricted to the intestines of larval and adult transgenic worms (3)(4)(5). Interestingly, the structurally similar Hemonchus contortus (3)(4)(5) and Schistosoma mansoni (6) cathepsin B homologues were also expressed in the gut and were suggested to be potentially involved in feeding (5,7), such as nutrient digestion. Heterologous transformation of C. elegans with an H. contortus cathepsin B gene promoter has demonstrated also conservation of the mechanisms controlling its spatial expression in free-living and parasitic nematodes (8), and therefore both enzymes were hypothesized to be not only structurally similar, but also functionally homologous and important for proper feeding (4,7).
Cathepsin L and Z-like proteases were shown to be present in many parasitic nematodes, where they are speculated to have diverse biological functions including invasion, feeding, molting, and immune evasion (reviewed in Refs. 6, 9, and 10). Many of these cathepsins have homologues in C. elegans suggesting that they may be involved in functions conserved across different nematode species. Yet not much is known about their precise function. In Dirofilaria immitis (11) and Brugia pahangi 1 cysteine proteases were shown to be associated with molting as well as activities that might facilitate larval migration. The potential role of cysteine proteases during molting was indirectly established in Onchocerca volvulus by showing that the peptidyl monofluoromethyl ketones, low molecular weight irreversible cysteine protease inhibitors, inhibit the molting of third stage larvae (L3) in a time-and dose-dependent manner (12). These irreversible inhibitors can block cathepsin Z and L-like, but not the B-like enzyme activities, suggesting that a cathepsin L, as well as a cathepsin Z, might be involved in the molting process. The target cysteine proteases were indirectly localized in the granules of the glandular esophagus of O. volvulus L3 using the biotin-Phe-Ala-CHN 2 inhibitor (12). In further studies, a larval O. volvulus cysteine protease named LOVCP was cloned (12). LOVCP as well as its homologue in Toxocara canis and C. elegans were recently classified as a novel monophylectic group in the papain family of cysteine proteases, named cathepsin Z (13). The O. volvulus cathepsin Z was shown to be required for molting and the development of fourth-stage larvae (L4) based on its localization using monospecific anti-LOVCP antibodies. The native enzyme was local-ized in molting L3 in the region where the separation between the cuticles of L3 and L4 takes place. The general role of the O. volvulus cathepsin Z-like enzyme as well as its C. elegans homologue (14) during molting was therefore hypothesized to be as a proteolytic enzyme involved in cuticle degradation, similar to that observed in the entomopathogenic fungi Metarhizium anisopliae with its endogenous cysteine protease Pr4 (15). This function as well as other yet unknown functions of cathepsin Z and L cysteine proteases were previously hypothesized to be present in nematodes based on the immunolocalization of onchocystatin, an endogenous cysteine protease inhibitor, in thin sections of O. volvulus. Onchocystatin was expressed in L3, molting L3, adult worms, and eggshells around developing microfilaria (16,17), suggesting that its target cysteine protease(s) is possibly involved in regulating activities such as molting, cuticle remodeling, and embryogenesis during the development of the parasite in the host. Importantly, C. elegans has two homologues of onchocystatin (14), suggesting the presence of their target enzymes in this worm as well.
The indirect approaches as described above, however, do not provide conclusive information on the precise function of these proteases. Recently, a new family of cathepsin L-like sequences with similarity to previously characterized mammalian cathepsin L-like enzymes was identified in the O. volvulus as well as in other filarial L3 EST data bases, 2 and in other parasitic nematodes. 3 Their function is not as yet known. We have identified a related cathepsin L-like protease sequence (T03E6.7) within the C. elegans complete genome data base (ACeDB), and used the C. elegans powerful system for investigating its functions in vivo during development. To determine the function of the cathepsin L protease we took advantage of methods available in C. elegans that enable analysis of gene promoter activity and gene function in individual cells. Based on the obtained information we could predict its potential physiological role in filarial parasites. This study is the first to directly demonstrate the functional importance of cathepsin L in nematode development.

EXPERIMENTAL PROCEDURES
Data Base Search of Cathepsin L-like Enzyme of C. elegans-A BLAST search (18) of the C. elegans genome data base (ACeDB, www. sanger.ac.uk/projects/C_elegans/wormpep/1) and the nr data base using the Ov-CPL (accession number AF331036) and other filarial nematodes amino acid sequences, Di-CPL (accession number AF001101), Bp-CPL (accession number AF031819.1) identified a predicted C. elegans cathepsin L-like gene (T03E6.7), which was named Ce-cpl-1. 4 A Ce-cpl-1 cDNA clone containing the full-length sequence was also identified in the EST data base (accession number C12099) and the corresponding ZAP II phage was obtained from Yuji Kohara (clone yk146d10, C. elegans consortium, National Institute of Genetics, Mishima, Japan). The pBluescript phagemid was excised and the DNA sequenced in both directions to confirm the predicted amino acid sequence.
Signal sequences and putative cleavage sites were identified using the SignalP server (www.cbs.dtu.dk/services/SignalP). Prediction of the pro-region cleavage site as well as the active sites were based on alignment of the CPL protein sequences was made using Cluster W multiple sequence alignment. Analysis of the promoter region of Cecpl-1 gene was performed using Genefinder provided by BCM server (www.hgsc.bcm.tmc.edu/searchlauncher).
Stage-specific Transcript of Ce-cpl-1 Using RT-PCR-C. elegans strains used in this study were the wild-type Bristol N2 strain and unc-76 mutant strain DR96 (unc-76 (e911)V) (19), both provided by the Caenorhabditis Genetics Center. Strains were maintained on NGM agar plates as previously described (19). Semi-quantitative RT-PCR (sqRT-PCR) was carried out using first strand cDNA generated from total RNA collected from synchronous L1, L2, L3, L4, and young adult C. elegans cultures at 2-h intervals, as previously described (20). The stage-specific cDNA samples were kindly provided by Iain Johnstone (Wellcome Center for Molecular Parasitology, University of Glasgow, UK) and used according to the established protocols. Gene-specific cDNA fragments were amplified using two sets of PCR primers, one set was specific for Ce-cpl-1 (CPF1, sense, 5Ј-GTCTCCGTGCTCTGGGTC-GGTTCCGTATC-3Ј and CPR1, antisense, 5Ј-CCATGGTGTCGACACC-GAGGAGTCATAC-3Ј) and the other set was specific for an internal control, the ama-1 transcript (20). The primers were designed to span an intron to distinguish cDNA from contaminating gDNA products. The following PCR conditions, which allowed reactants to remain in excess, were used: 94°C for 3 min, followed by 30 cycles of 94°C for 30 s, 59°C for 30 s, and 72°C for 1 min, with a final extension at 72°C for 3 min. Amplified products were separated on 2% agarose gels, Southern blotted, and probed with the appropriate end-labeled oligonucleotides. After autoradiography, specific bands corresponding to amplified cpl-1 and control ama-1 products were excised from the blot for each time point and counted in scintillant. The relative content of the transcript corresponding to the Ce-cpl-1 gene is expressed as the ratio of the signal for cpl-1 for each developmental stage to that of ama-1. The sqRT-PCR was carried out on two occasions, with very similar results, one of which is presented.
The double-stranded RNA (dsRNA) used for RNAi experiments was prepared following the Fire et al. (21) protocol. Briefly, for in vitro transcription, each of the pBluescript plasmid DNA constructs was linearized with the appropriate restriction enzyme and single-stranded sense or antisense RNA was synthesized using the RiboMAX RNA Large-Scale Production System (Promega, Madison, WI) according to the manufacturers instructions. Equal amounts of each set of sense and antisense RNA strand were then annealed by incubation in injection buffer (20 mM KPO 4 , 3 mM K citrate, 2% PEG 6000, pH 7.5) for 10 min at 68°C and 30 min at 37°C. Double-stranded RNA (concentration 0.5 mg/ml) was then injected into the gonad of 25-30 young adult C. elegans hermaphrodites. Injected worms were left for 16 -24 h at 20°C to recover and to lay any eggs present in utero prior to microinjection, and then transferred to individual plates at 24-h intervals. The F1 progeny was quantified and examined for embryonic lethality or abnormal development.
RNAi using the soaking protocol was done on L3 and L4 as described by Tabara et al. (22). Briefly, 15 C. elegans worms were incubated in 15 l of 0.2 M sucrose in 0.1 ϫ PBS containing 1 mg/ml dsRNA pre-mixed with 1 l of Lipofectin (Invitrogen, Carlsbad, CA). After 24 h, soaked larvae were transferred to individual plates and their development examined for 4 -5 days. Larvae soaked in 0.2 M sucrose in 0.1 ϫ PBS without dsRNA served as the control. The length and width of each treated worm were measured under microscope using ocular and stage micrometer. Data were analyzed by One-way Analysis of Variance (ANOVA) and Student-Newman-Keuls Comparison Test (GraphPad InStat Software Inc.). Probability values of less than 0.05 were considered significant.
Loss of Ce-cpl-1 Transcript Following RNAi-Loss of cpl-1 transcript following dsRNA injection was examined by RT-PCR using the same cpl-1 and ama-1 internal primers described above. Approximately 20 RNAi injected adult hermaphrodites or 300 RNAi mutant embryos were collected and washed twice in 1 ml of PBS. Wild-type adults and embryos were used as controls for normal gene expression levels. Adult and embryo pellets were then resuspended in 200 l of lysis buffer (0.5% SDS, 5% ␤-mercaptoethanol, 10 mM EDTA, 10 mM Tris-HCl, pH 7.5, and 0.5 mg/ml proteinase K), quick-frozen at Ϫ80°C for 10 min, followed by incubation at 55°C for 1 h. The RNA was extracted using Total RNA Isolation Reagent (Advanced Biotechnologies Ltd.) and the RT-PCR was carried out using SuperScript One-Step RT-PCR System (Invitrogen) according to the manufacturers instructions. Each 50-l reaction mixture was split into two tubes into which either cpl-1 primers or ama-1 control primers were added. After 35 cycles of amplification, the RT-PCR products were separated on 2% agarose gels.
Generation and Expression of C. elegans Reporter Gene Constructs-Both transcriptional and translational Ce-cpl-1 reporter gene fusion constructs were generated and their expression patterns examined. For the transcriptional fusion construct, a promoter region of 1.76-kb of cpl-1 upstream sequence, was generated by PCR on T03E6 cosmid DNA using Vent DNA Polymerase (New England Biolabs) and the following PCR primers: CPpromF1 (SphI), sense, 5Ј-ACAGCATGCTCCCGAAA-AAAACTTCAATATTCTG-3Ј, corresponding to positions Ϫ1761 to Ϫ1736 relative to the ATG start codon, and CPpromR1 (XbaI), antisense, 5Ј-CGGTCTAGACTGGAATTTTATAACATTTAAAAT-3Ј, complementary to positions Ϫ2 to Ϫ25 relative to the ATG start codon (Fig.  1A). Following restriction enzyme digestion, the PCR fragment was cloned into lacZ reporter vector pPD96.04 containing a nuclear localization signal (kindly supplied by A. Fire).
The translational fusion construct (pSL104) contained a 3.6-kb genomic fragment including 1.03 kb of the potential promoter region and all the four exons and three introns. This fragment was amplified by PCR on T03E6 cosmid DNA using the following primers: F1 (SphI), sense, 5Ј-CATGCATGCATCTCACCGTCTTCACCAGG-3Ј corresponding to position Ϫ1030 to Ϫ1010 relative to the ATG start codon and R1 (AgeI), antisense, 5Ј-GCTACCGGTGCGACTCCGCAGTGATTGTT-3Ј, complementary to position 2550 to 2570 within the cpl-1 gene relative to the ATG start codon. The PCR amplified gene fragment was first cloned into the PCR2.1 cloning vector using the TOPO cloning kit (Invitrogen) and then subcloned into the gfp reporter vector pPD95.75 (kindly supplied by A. Fire). Plasmid DNA of the construct was prepared using the Concert TM Rapid Plasmid Miniprep System (Invitrogen) and sequenced to confirm that the last exon of cpl-1 is in-frame with gfp.
Transformation of C. elegans was performed by microinjection of plasmid DNA into the distal arm of the hermaphrodite gonad as described previously (23,24). Reporter plasmid DNA corresponding to the transcriptional construct (25 g/ml) or translational construct (60 g/ ml) was co-injected with a marker plasmid DNA (pRF4 at 100 g/ml) containing a dominant mutant allele of the rol-6 gene (su1006) (24). Transformants were identified by their right roller phenotype (25). To allow examination of expression from transcriptional fusion constructs in the absence of any enhancers from the rol-6 gene, transformation rescue of the unc-76 mutant strain DR96 was carried out with plasmid p76 -16B (a gift of Laird Bloom, MIT Center for Cancer Research), which contains the wild-type neuronally expressed unc-76 gene. Lines in which F2 and subsequent generations showed the roller or uncrescue phenotype were stained for ␤-galactosidase expression, as previously described (3) using 4Ј6-diamidino-2-phenylindole (final concentration 0.1%) as a co-stain to aid in the identification of cell types. GFP was visualized by mounting live transgenic worms on a 2% agarose pad in 0.01% sodium azide, that inhibits their movement, and viewing under a fluorescence filter. At least three independent lines were examined for each construct.
Production of Recombinant C. elegans CPL-1 Fusion Polypeptide-A fragment of Ce-cpl-1 cDNA encoding the mature enzyme (Fig. 1A, amino acids 121-337) was amplified and cloned into the BamHI and XhoI sites of the pGEX4T-3 expression vector (Amersham Bioscience Inc., Piscataway, NJ). The recombinant GST-Ce-mCPL-1 protein was expressed in the form of inclusion bodies and was therefore purified from the 50 mM Tris-HCl, pH 8.0, insoluble pellet. The pellet was solubilized in 6 M urea in 50 mM Tris-HCl buffer, pH 8.0, followed by preparative separation using the Prep Cell (Bio-Rad) according to the manufacturers instructions. Fractions containing the purified recombinant GST-Ce-mCPL-1 polypeptide were identified using antibodies against GST. A mouse antiserum was raised using the repetitive multiple site immunization strategy (26). Each mouse received a total of 10 g of GST-mCPL-1 in RIBIs adjuvant on days 0, 3, 6, 8, and 10.
Detection of the Native Ce-CPL-1 Protein in C. elegans by Immunofluorescence-For detection of the native Ce-CPL-1 enzyme in C. elegans embryos, gravid hermaphrodites were washed off culture plates in PBS and cut open to release the embryos. The embryos were then collected and fixed in methanol/acetone using the freeze-cracking protocol (27). For whole mounted immunostaining, mixed-stage population of larvae and adults were collected and washed in PBS and Ruvkun Fixation buffer before being fixed and permeabilized using 1% paraformaldehyde in Ruvkun Fixation buffer for 30 min and two freezethaw cycles in a dry ice/ethanol bath (28). The fixed and permeabilized embryos or the whole worms were treated with a blocking solution (60 mg/ml normal goat serum in PBS, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA) for 1 h before reaction with antibodies. Primary anti-GST-mCPL-1 antibodies were used at 1:20 and 1:40 dilution. Fluorescein isothiocyanate-conjugated rabbit anti-mouse secondary antibodies were used at a 1:50 dilution. Immunostained specimens were viewed under fluorescence microscope using appropriate filter sets in the presence of mounting medium, Vectashield containing 4Ј6-diamidino-2-phenylindole (Vector Laboratories, Inc., Burlingame, CA).
Ultrastructural Localization of the Native Cathepsin L Protein in C. elegans and O. volvulus-C. elegans mixed stage larvae, hermaphrodites, and embryos collected from 5% NaOCl-treated hermaphrodites were fixed for 60 min in 4% paraformaldehyde, 0.1% glutaraldehyde in 0.1 M phosphate buffer, pH 7.4, containing 1% sucrose. The fixed worms were then processed for immunoelectron microscopy as previously described (17). Thin sections of C. elegans embedded worms were probed with mouse antisera raised against the recombinant C. elegans GST-Ce-mCPL-1 fusion polypeptide before incubation with 15-nm gold particles coated with anti-mouse IgG (Amersham Bioscience, Inc.). Mouse preimmune serum or antibodies to GST were used as controls. The localization of the native CPL in O. volvulus was done as previously described (17) using thin sections of O. volvulus L3 and adult-embedded worms and rabbit antisera raised against the recombinant O. volvulus mature region or the pro-region. Rabbit preimmune serum or antibodies to GST were used as controls. The antisera raised against the CPL enzymes recognized by Western blot their corresponding recombinant proteins and did not cross-react with recombinant proteins expressing the cathepsin Z-like proteins of C. elegans or O. volvulus (data not shown).
Generation and Expression of O. volvulus Reporter Gene Construct for Heterologous Transformation in C. elegans-The potential promoter region of Ov-cpl was amplified by sequential PCR reactions using O. volvulus genomic library, Fix II (kindly provided by Dr. R. D. Walter), the vector primer and gene-specific primers designed based on known sequences. A 1.3-kb fragment containing the ATG start codon was amplified first followed by another 1.8-kb fragment after using the vector primer and an antisense primer corresponding to a region within the 1.3-kb fragment. The final region identified as the putative promoter of the gene was ϳ2.5 kb upstream of the ATG of Ov-cpl encoding region (accession number AF442768). For the transcriptional fusion construct, this fragment was amplified from O. volvulus genomic DNA using the following pair of PCR primers: OVcplp5Ј (XbaI), sense, 5Ј-CGTCTAGACTTAGATTTCCATCCCGACGAG-3Ј, corresponding to positions Ϫ2525 to Ϫ2504 relative to the ATG start codon, and OVc-plp3Ј (XmaI), antisense, 5Ј-GACCCGGGTGAGTTTTTTTCTGTTTCTG-TTTTCC-3Ј, complementary to positions ϩ1 to Ϫ25 relative to the ATG start codon. Following restriction enzyme digestion, the PCR fragment was cloned into lacZ reporter vector pPD90.23 containing a nuclear localization signal (kindly supplied by A. Fire). The fusion gene containing the cpl putative promoter was designated Ov-cpl:lacZ (pSL117A). Using a heterologous transformation, we created C. elegans transgenic lines expressing the O. volvulus promoter as described in generation and expression of C. elegans reporter gene constructs.

RESULTS
The Cathepsin L-like Cysteine Proteases of C. elegans-Based on BLASTP search of ACeDB using the mature region of the Ov-CPL amino acid and the other filarial nematode sequences, a related sequence was identified in cosmid T03E6.7 (accession number Z92812) encoding a predicted C. elegans cathepsin L-like enzyme designated Ce-CPL-1. The mature region of Ce-CPL-1 was 45-49.8% identical to the mature regions of the filarial and ϳ30% to human CPL enzyme sequences. The initiator methionine in the predicted Ce-CPL-1 protein sequence was confirmed by matching it with the open reading frame encoded by the full-length yk146d10 cDNA clone (accession number C12099). The C. elegans cathepsin L-like enzyme sequence contains 337 amino acids with a potential signal pep- ) and a predicted pro-region of 100 amino acids and mature region of 216 amino acids, with a predicted molecular weight of 24,000. The Ce-CPL-1 pro-region is shorter than the 160 -167 amino acids region present in the filarial CPL enzymes. Nonetheless, all the pro-regions contain five of the six residues that make up the ERFNIN motif (29) that is characteristic of the noncathepsin B papain-like cysteine proteases. The motif in the nematode sequences differs in Ile for Thr in the Ov-CPL and the other filarial sequences (data not shown) and in Arg for Tyr in the Ce-CPL-1 sequence.
The coding region of the C. elegans cpl-1 gene resides within 2.6 kb of DNA, and spans 4 exons and 3 introns (Fig. 1B). Ce-cpl-1 has larger introns (470, 542, and 570 bp) compared with most C. elegans genes having small introns (40 -60 bp) (30). The Ce-cpl-1 gene lies within cosmid T03E6 on position 12.45 on chromosome V to the left of the dpy-21 gene and to the right of the rrs-1 gene (C. elegans genome project).
Ce-cpl-1 Is Expressed in All C. elegans Developmental Stages and Its Expression Increases Prior to Each Molt-Semiquantitative RT-PCR was used to examine the temporal pattern of Ce-cpl-1 expression throughout development. We used the ama-1 gene as an internal control transcript to allow the relative quantification of cpl-1 expression in each stage. It has previously been shown that the levels of ama-1, which encodes the large subunit of RNA polymerase II, are relatively constant during development and it is therefore a suitable control gene (20,31). As shown in Fig. 2, cpl-1 is expressed throughout development, in mixed stage embryos and in all larval and adult stages. The highest level of expression was detected in adult stages. Interestingly, in each larval stage, the level of cpl-1 expression increases significantly in the intermolt period, approximately 4 h prior to each molt, then it gradually decreases, returning to basal levels immediately after molting. At the molt from L4 to adult, cpl-1 level decreases slightly, but rapidly increase and remain at a constant level as the adults mature.
Ce-CPL-1 Is Essential for Embryogenesis-To determine the possible function(s) of the Ce-CPL-1 enzyme during development of C. elegans, RNAi was carried out to selectively interfere with cpl-1 gene expression. DsRNA corresponding to the entire cDNA or to three different regions of the cpl-1 gene (Fig.  1B) was injected into adult hermaphrodites and the effect on the adults and F1 progeny was examined. In repeated experiments, an early embryonic lethal phenotype was observed, with 95-100% of F1 embryos arresting with ϳ100 -200 cells ( Table  I). The number of F1 progeny was slightly reduced in some injected adults compared with noninjected controls, although this could have been due to gonadal tissue damage during the injection procedure. A time course of development of RNAi affected embryos compared with wild-type embryos is shown in Fig. 3. Embryos from Ce-cpl-1 dsRNA-injected hermaphrodites initially underwent normal cell divisions, but ϳ50 min after the first cleavage (approximately 12-cell stage) the cell divisions occurred at a slower rate than that of wild-type embryos. Some development of muscle and neuronal tissue did occur, as indicated by twitching of the dying embryos, but embryos from the injected worms failed to undergo any morphogenesis and arrested as a ball of cells. The few larvae, which did occasionally develop, appeared abnormal, often with incomplete gut development and usually died at the L1 stage. Soaking of L4 in cpl-1 dsRNA resulted in an identical embryonic lethal effect in ϳ92% of the F1 progeny of soaked worms (Table I).
As identical RNAi phenotypes were obtained on L4 and adult hermaphrodites using dsRNA prepared from three separate regions as well as the full-length cDNA of the Ce-cpl-1 gene, we have concluded that the observed effect was due to specific interference with cpl-1 expression. However, to confirm the absence of cpl-1 transcript following RNAi treatment, RT-PCR using cpl-1 and ama-l specific primers was carried out on treated as well as wild-type worms. Fig. 4 shows the absence of cpl-1 cDNA from RNAi-treated adults and progeny embryos, while the control gene ama-l is present at the same level as in N2 noninjected adults and embryos.
Ce-CPL-1 Is Important for Development of L3 to Adult Stages-The sqRT-PCR demonstrated that cpl-1 levels increase about 4 h prior to molting (Fig. 2). To examine whether RNAi would also have any effect on molting and/or the development of adult stages the RNAi was performed on L3 using the soaking technique (22). Larvae do not survive the microinjection procedure well. L3 were soaked overnight in dsRNA corresponding to the full-length Ce-cpl-1 cDNA and the phenotype obtained was observed for 5 days. Although all the treated larvae developed by day 5 into normal size adults, the growth of 53% of them was significantly delayed. The affected adult worms on day 3 were significantly shorter (458 Ϯ 25.7 versus 481 Ϯ 13.4 m) and thinner (33 Ϯ 5.0 versus 47 Ϯ 6.0 m) in comparison to untreated wild-type worms (Table I). Moreover, the affected worms also laid fewer eggs than the control worms: 0 -21 eggs versus 21-47 eggs on day 2 and 16 -119 versus 87-195 eggs on day 3. Although by day 5 all the affected worms reached normal size, their egg production was still diminished. The laid eggs, nonetheless, developed normally to adult stages. Thus, it appears that interference with cpl-1 gene expression also significantly affected the normal growth of L3 to egglaying adults and their subsequent egg production.
Temporal and Spatial Expression of cpl-1 in Transgenic Worms-The spatial and temporal expression of Ce-cpl-1 was examined by transformation of C. elegans worms with 2 reporter gene constructs. Following germ line microinjection, injected DNA normally recombines to form large, extrachromosomal arrays that are transmitted to between 10 and 90% of The active site cysteine (C), histidine (H), and asparagine (N) residues are indicated by an asterisk. Blank and dark arrowheads show the putative cleavage sites of the signal peptide and pro-region, respectively. The position and size of the three introns are indicated between the exons by thin lines. Thick lines mark the sequence regions corresponding to PCR products used to generate dsRNA for RNAi analysis ("Experimental Procedures").

FIG. 2. Temporal pattern of Ce-cpl-1 gene expression as determined by semi-quantitative RT-PCR.
The graph shows the ratio of the signal (counts per minute) from Ce-cpl-1 amplified cDNA to that of a control gene ama-1 (y axis). cDNA was synthesized from mRNA from mixed stage embryos and synchronized larval and adult populations collected at 2-h intervals, as indicated on the x axis. Note that cpl-1 transcripts increased in the intermolt period of each larval stage. the progeny (24). Because individual transformants show mosaic patterns of expression, the staining pattern of many transformants derived from at least three independent lines was examined. Expression of the lacZ transcriptional fusion construct was detected in all developmental stages, from early embryos through to the adult stages. This was in accordance with the cpl-1 transcript pattern of expression as detected by sqRT-PCR, suggesting that the gene is regulated at the level of transcription. In embryos, ␤-galactosidase expression was confined exclusively to gut cells (Fig. 5a). In larvae and adults, both hermaphrodites and males, strong expression was also detected in all gut cells (Fig. 5, b-d). In addition, ␤-galactosidase staining was also observed in hypodermal cells of many larval and adult transgenic worms, although the number of hypodermal cells showing expression and the level of expression was more mosaic than gut cell expression. Examination of transformants with high levels of hypodermal cell expression showed that all types of hypodermal cells, dorsal, lateral, and ventral cells, expressed the cpl-1:lacZ transgene. Identical staining patterns were observed using rol-6 or unc-76 rescue as a marker of transformation, indicating that any enhancer elements present in the marker gene did not influence the expression pattern.
A distinct expression pattern in the hypodermal and cuticular regions of all stages was obtained from the Ce-cpl-1 translational fusion construct (Fig. 6, A-D). Expression of the CPL-1::GFP fusion protein was also observed along the length of the pharyngeal lining in L1-L4 stages, but was restricted to the posterior bulb of the pharynx in adult worms (Fig. 6, A and  B). Furthermore, robust expression was observed in the eggshells surrounding the embryos inside the hermaphrodite and in the regions of the vulva, uterus, and spermatheca (Fig. 6C). GFP was also detected in the eggshells of laid eggs in different stages of development (Fig. 6E). In contrast to the transcriptional reporter construct there was no CPL-1::GFP expression in the gut.
Localization of the Native CPL-1 Protein in C. elegans by Immunofluorescence and Immunogold Labeling-The distribution of the native CPL-1 in C. elegans embryos, larvae, and adult hermaphrodites was determined by immunofluorescence staining with antibodies raised against the recombinant C. elegans GST-Ce-mCPL-1 fusion polypeptide. CPL-1 was localized in various cells in early to late embryonic stages (Fig. 7, A-C) as well as in the eggshell surrounding the embryos (Fig.  7D). CPL-1 accumulated in numerous cells during early embryogenesis (Fig. 7A); however, as embryos progressed through middle and late developmental stages, the CPL-1 protein clustered in gut cells of the embryos at the early L1 stage (Fig. 7C). In addition, Ce-CPL-1 was present in the hypodermal and cuticular regions in all larval and adult stages (Fig. 7E), in the pharyngeal lining of all stages, and in the pharyngeal gland of the adult worms only (Fig. 7F). The subcellular localization of  3 and 7). An asterisk indicates PCR fragments amplified from contaminating genomic DNA, present in some RNA preparations. M shows 1-kb DNA size marker. The figure is a negative of an ethidium bromide-stained agarose gel. a Twenty-five young hermaphrodite worms were microinjected with cpl-1 dsRNA and the effect of dsRNA was observed on the injected as well as on the progenies of the injected worms as described under "Experimental Procedures." The same experiment was done with 4 different preparations of dsRNA and each experiment was repeated at least once. Similar results were obtained each time.
b Fifteen L4 worms were soaked in cpl-1 dsRNA for 24 h and the effect was observed on the soaked as well as on the progenies of the soaked worms as described under "Experimental Procedures." Four different dsRNA preparations were used as above with similar results. The experiment was repeated once.
c Fifteen synchronized C. elegans L3 worms were soaked in cpl-1 dsRNA for 24 h and the effect of dsRNA was observed on the soaked as well as on the progenies. The experiment was repeated once. As the effect was on growth of the worms the length and width of the affected worms were measured. On day 3 after treatment the affected worms were significantly shorter (458 Ϯ 25.7 versus 481 Ϯ 13.4 m, p Ͻ 0.01) and thinner (33 Ϯ 5.0 versus 47 Ϯ 6.0 m, p Ͻ 0.001) than the control worms. In addition, the affected worms laid fewer eggs on day 3 than the controls, 16 -119 versus 87-195, respectively. the native CPL protein in C. elegans was determined by immunoelectron microscopy staining. In larval and adult stages the native protein was localized in the cuticular region of the worms and on the outer cuticle of the ridge of the annuli (Fig.  8, A-C). In molting larvae, the protein was localized in both the old and new cuticle (Fig. 8B). In addition, the protein was localized on the surface of late-stage embryos (Fig. 8D).
Localization of the Native Cathepsin L Protein in O. volvulus by Immunoelectron Microscopy-Interestingly, antibodies raised against the recombinant O. volvulus Ov-CPL mature protein recognized the corresponding native O. volvulus enzyme in regions similar to those identified in C. elegans. The antibodies reacted specifically with the O. volvulus protein in the basal layer of the cuticle and the hypodermis of male and female worms (Fig. 9, A and B). In sections of molting L3, the protein was localized in the area between the basal layer of the cuticle and the hypodermis of the larvae (Fig. 9C). In addition, antibodies raised against the pro-region of Ov-CPL reacted specifically with the protein in the granules of the glandular esophagus of molting L3 (Fig. 9D). Specific labeling was also obtained in the eggshells surrounding the embryos and the developing microfilariae (L1) inside the uterus of the female worms with antibodies against the mature region (Fig. 9, E-G). Serum against GST or preimmune sera did not cross-react with any proteins in the larvae and the adult worms (data not shown).
Temporal and Spatial Expression of Ov-cpl-1 in C. elegans Transgenic Worms-Having similar expression patterns of Ce-CPL-1 and Ov-CPL native proteins in their respective nematodes, we questioned whether the O. volvulus cpl promoter would be expressed in a similar manner as its C. elegans counterpart when used in a heterologous transformation model. Lines of transgenic C. elegans that carry a heterologous Ov-cpl gene promoter linked to a lacZ reporter gene were generated, thus allowing the study of the Ov-cpl gene promoter activity in individual cells of C. elegans, and comparing their expression patterns in vivo. As indicated by the ␤-galactosidase staining pattern, expression of the Ov-cpl gene was similar in all developmental stages of the worm, and was restricted to intestinal cells (Fig. 10). This pattern of staining was consistent in all independent lines and was seen in the large nuclei of 12 cells comprising the anterior and posterior regions of the intestine (Fig. 10). However, no staining was observed in the hypodermal cells as seen with C. elegans cpl transcriptional construct.

DISCUSSION
In these studies we show that cathepsin L is important for embryogenesis and development in nematodes. Disruption of C. elegans CPL-1 function in hermaphrodites or L4 worms by RNAi results in arrested, nonviable early stage embryos, indicating that this enzyme is essential for the normal progression of early development and plays a crucial role in embryogenesis. The premature arrest of C. elegans developing embryos suggests that the cathepsin L is important during the early stages of embryogenesis when cell division and proliferation are occurring most rapidly. The C. elegans embryo undergoes inten- sive protein synthesis, mitosis, and cycles of re-organization in the first few hours after fertilization. Many of the early developmental activities of the embryo depend on the uptake and utilization of maternally derived molecules (32). Of these, the yolk proteins, or vitellogenins, are the major sources of nutrients for the developing embryo. These are taken up by all cells of the developing embryo and become localized in gut cells as the embryo develops (32). Immunostaining with antibodies raised against the recombinant Ce-CPL-1 protein established that the C. elegans native protease was also expressed in early to late stages of embryos and became highly concentrated in gut cells during late embryonic development. Based on the similarity in the staining patterns of these two proteins, it is tempting to speculate that Ce-CPL-1 might play a role in the proteolysis of essential nutrients, such as the yolk protein, during embryonic development. Support for this hypothesis was found in previous reports where cysteine proteases have been shown to be involved in yolk degradation during invertebrate embryonic development (33)(34)(35). In vertebrate models the expression of active cathepsin L was significantly higher in visceral yolk sac than in placenta (36). This was consistent with a higher proteolytic activity needed in the yolk sac for the production of amino acids for protein synthesis. Moreover, in vitro perturbation of yolk sac with the specific cysteine protease FIG. 8. Ultrastructural localization of the native CPL-1 protein in C. elegans at different stages of development. Antibodies raised against the mature region of the C. elegans enzyme were used to study the subcellular localization of CPL-1 as described under "Experimental Procedures." A, the protein is localized in the cuticles (cu) of larvae (A and B) and adult stages (C). Note that when the larvae molt the protein is present in both, the old and new cuticles (B, arrowheads). D, the protein is also localized on the surface (arrowheads) of the late-stage embryos (emb) (bar, 500 nm). inhibitors E-64 or leupeptin, resulted in decreased protein processing and embryo growth retardation in rat (37), confirming the postulated role of cathepsin L for normal breakdown of proteins during embryogenesis in this system (38). In C. elegans this potential function is supported by preliminary studies that have shown defects in yolk processing in Ce-cpl-1 RNAi mutant embryos. 3 However, whether or not CPL-1 plays any direct role in yolk processing needs further investigation.
As well as being detected within cells of the C. elegans embryo, Ce-CPL-1 also localized in eggshells surrounding the embryos and in the uterus, spermatheca, and vulva of hermaphrodites. Immunoelectron staining of thin sections from the uterine regions of C. elegans indicated that the native protein is mostly present in the periphery of the developing embryos. This may also reflect uptake of the CPL enzyme in the nematode into developing embryos from the surrounding adult reproductive tissues in a manner similar to the uptake of the yolk proteins via specific receptors (39). Alternatively, this could indicate a role for CPL-1 in the synthesis or maturation of the eggshell.
Differences were observed in the spatial expression of the cpl-1 transcriptional and translational reporter gene constructs in transgenic C. elegans. The transcriptional construct directed expression mostly in the gut and in hypodermal cells, indicating activation of the promoter in these tissues. In contrast, the translational construct was expressed in the hypodermis, pharynx, reproductive tissue, and around the embryo. As would be expected, expression from the translational construct was similar to that observed with antibodies to the native Ce-CPL-1 protein. Similar differential spatial pattern of expression and localization has also been described for the C. elegans yolk proteins. Although the yolk genes were expressed exclusively in the adult hermaphrodite gut (40), the GFP translational reporter constructs and immunostaining localized the yolk proteins to the gonad and developing oocytes and embryos (32,39). Our data suggest that Ce-CPL-1 that functions during embryogenesis is expressed in a similar manner to the yolk proteins and is transported to the gonad after synthesis in the gut cells, which could be consistent with the potential role of Ce-CPL-1 during embryogenesis. Whether this proposed transport of the CPL-1 protein is linked to transport of yolk proteins will also require further examination. The differential expression patterns of the constructs, however, could also be due to a larger size of the promoter region used in the transcriptional construct (1753 versus 1030 bp). Although, in both constructs GATA-like motifs and several AP-1, TATAA box, and CAAT sequences are present, the additional 723-bp upstream sequence contains an extra 6 GATA-like and 3 AP-1 transcriptional factors. GATA elements were shown to be involved in intestinal cell-specific expression of many C. elegans genes (3,41). Only a detailed analysis of the cpl-1 promoter would resolve the contrasted expression in the gut cells.
Based on the results obtained from the temporal and spatial pattern of Ce-cpl-1 expression as well as localization of native CPL-1 proteins in C. elegans, we suggest that Ce-CPL-1 is also involved in processes associated with cuticle remodeling in larvae and adult worms during molting and growth after the final molt. All nematodes molt four times. During each molt, a new cuticle is synthesized and the old cuticle is shed. Shedding of the old cuticle begins at the mouth of the animal where secretory vesicles from pharyngeal gland cells, combined with contraction of the pharynx, weaken and break the old cuticle, and then the worm sheds the old cuticle from head to tail as its moves (42). During the molting process and growth, profound changes in nematode structure and extensive tissue remodeling occur; C. elegans increases in size by about one-third in each larval stage and again in adult stages after the final molt (43). In addition, the basic development cues controlling postembryonic cell lineages, and even developmental plasticity in cell morphology and function, are activated by the molting cycles (44).
The following results clearly support the hypothesis that the cathepsin L enzyme of C. elegans has an additional role during molting and in cuticle remodeling of larval and adult stages: 1) in transgenic worms high levels of CPL-1::GFP expression were seen in hypodermal and cuticular regions of all stages of the worms. Furthermore, the native protein was localized to the cuticular regions of larvae and adult worms, including the outer cuticle, and in molting larvae, the protein was present in the old and the new cuticles around the region where the cuticles separate during molting, as well as in the pharynx; 2) a strict correlation exists between cpl-1 mRNA production and molting based on the sqRT-PCR data. In each larval stage, the level of cpl-1 expression increased significantly in the intermolt period, approximately 4 h prior to each molt, and then gradually decreased, returning to basal levels immediately after molting. CPL-1 may be one of the enzymes required for a specific and tightly regulated process of a limited duration which is consistent with the expression pattern of cpl-1 during molting of each larval stage; and 3) soaking of L3 with cpl-1 dsRNA affected the normal growth of the larvae to adult stages. Over a period of 3 days, the worms were significantly thinner and shorter. The transient effect might be due to the presence of other proteolytic enzymes that compensate for the interference with CPL-1 activity.
Cathepsin L may have a direct role on the nematode cuticle and may be involved in degradation of the old cuticle, in processing of the new cuticle, and/or in digesting cuticular anchoring proteins. Cathepsin L enzymes have been shown to degrade collagen, laminin, and elastin, all components of basement membrane, in cancer cells (45,46), and act as a proteolytic enzyme causing cuticle degradation in the entomopathogenic fungi M. anisopliae (15). Alternatively, CPL-1 may act indirectly to process and activate other enzymes or hormones involved in molting (47). It would be of interest to identify the target proteins for its enzymatic activity.
Notably, similar staining pattern of the native enzyme was also obtained in O. volvulus; antibodies directed against Ov-CPL reacted with the cuticle and hypodermis of molting L3 and adult worms, in the granules of the glandular esophagus of molting L3, as well as in eggshells surrounding the embryos and the developing microfilariae (L1) inside the uterus of the female worms, and thus indirectly suggesting that the CPL enzyme may have similar functions in both nematodes. In B. pahangi the cathepsin L-like enzyme Bp-CPL was also suggested to have a role in molting of B. pahangi L3 based on the observations that Z-Phe-Ala-FMK inhibited their molting in vitro, 1 and that antibody to Bm-CPL localized the protein to hypodermal cells, cuticle and eggshells. 5 Z-Phe-Ala-FMK also inhibited L3 molting in O. volvulus (12) and D. immitis (45). Taking advantage of successful heterologous transformations in C. elegans with other nematode promoter regions (14), we also examined the cellular expression of the Ov-cpl promoter in C. elegans by microinjection a transcriptional reporter construct carrying ␤-galactosidase gene. Similar to C. elegans cpl promoter expression, the heterologous O. volvulus cpl gene promoter activity was restricted to intestinal cells during embryonic and post-embryonic development of C. elegans. Although not all the 20 intestinal cells were stained, it still points to possible conservation of the regulatory regions controlling gut expression from both genes. The variable expression observed in hypodermal cells with the Ce-cpl-1 promoter was not observed with the Ov-cpl promoter. This may reflect some differences in the control elements of the O. volvulus and C. elegans promoters. A thorough study of the 2.5-kb region of Ov-cpl promoter surrounding the cpl initiation codon using a series of deletions may identify the control elements that influence the differential expression. Such studies are underway in our laboratory. Differential pattern of heterologous expression of three other parasitic nematode gene promoters (pep-1, ac-2, and colost-1) in C. elegans has been reported (8). Although those parasite promoters showed a tissue-specific expression in C. elegans transgenic which correlated with the localization of the corresponding native proteins in the parasite, their temporal pattern of expression was different suggesting that the regulatory mechanism influencing the timing of expression might have evolved more rapidly than those controlling the spatial expression of genes (8).
In conclusion, we have demonstrated a critical role for the Ce-CPL-1 enzyme during C. elegans embryogenesis and shown that it is also involved in larval development and molting. Although there are parallels between localization of the O. volvulus and C. elegans CPL enzymes, as well as the temporal and spatial expression of cpl genes in both nematodes, the precise function of Ov-CPL in O. volvulus will need further confirmation. However, creation of C. elegans mutants disrupted in the cpl-1 gene will enable us to further elucidate its function in embryogenesis, molting, and development in both nematodes, as the mutants will enable us to test functional similarity between the parasite and C. elegans genes by complementation rescue experiments using heterologous translational constructs.