The Caenorhabditis elegans Orthologue of Mammalian Puromycin-sensitive Aminopeptidase Has Roles in Embryogenesis and Reproduction*

Mammals possess membrane-associated and cytosolic forms of the puromycin-sensitive aminopeptidase (PSA; EC 3.4.11.14). Increasing evidence suggests the membrane PSA is involved in neuromodulation within the central nervous system and in reproductive biology. The functional roles of the cytosolic PSA are less clear. The genome of the nematode Caenorhabditis elegans encodes an aminopeptidase, F49E8.3 (PAM-1), that is orthologous to PSA, and sequence analysis predicts it to be cytosolic. We have determined the spatio/temporal gene expression pattern of pam-1 by using the promoter region of F49E8.3 to control expression in the nematode of a second exon translational fusion of the aminopeptidase to green fluorescent protein. Cytosolic fluorescence was observed throughout development in the intestine and nerve cells of the head. Neuronal expression was also observed in the tail of adult males. Recombinant PAM-1, expressed and purified from Escherichia coli, hydrolyzed the N-terminal amino acid from peptide substrates. Favored substrates had positively charged or small neutral amino acids in the N-terminal position. Peptide hydrolysis was inhibited by the metal-chelating agent 1,10-phenanthroline and by the aminopeptidase inhibitors actinonin, amastatin, and leuhistin. However, the enzyme was ∼100-fold less sensitive toward puromycin (IC50, 135 μm) than other PSA homologues. Following inactivation of the enzyme, aminopeptidase activity was recovered with Zn2+, Co2+, and Ni2+. Silencing expression of pam-1 by RNA interference resulted in 30% embryonic lethality. Surviving adult hermaphrodites deposited large numbers of oocytes throughout the self-fertile period. The overall brood size was, however, unaffected. We conclude that pam-1 encodes an aminopeptidase that clusters phylogenetically with the PSAs, despite attenuated sensitivity toward puromycin, and that it functions in embryo development and reproduction of the nematode.

Metalloaminopeptidases catalyze the removal of amino acids from the N terminus of oligopeptides. Aminopeptidases are found in a wide variety of cells; they can be membrane-associated or cytosolic, and they have been implicated in a wide variety of physiological processes, including protein processing, regulation of peptide hormone action, viral infection, and cancer cell proliferation (1). The puromycin-sensitive aminopeptidase (PSA 1 ; EC 3.4.11.14) is a member of the M1 family of metallopeptidases, and, as such, it possesses the archetypal HEXXH(X) 18 E metal ion coordination site and the GAMEN motif characteristic of this family (2). A membrane-associated PSA is found in rat brain tissue (3), and studies using mouse mutants have shown that this isoform is involved in growth and behaviors associated with anxiety and pain (4). Further characterization has shown that this membrane PSA is essential for fertility in male and female mice (5,6). The cytosolic isoform of rat PSA is also predominantly found in the brain, although there is also distribution among other organs (3). Neuropeptides, such as enkephalins, have been proposed as substrates for both isoforms of PSA, although the intracellular localization of the cytosolic PSA makes such activity questionable for this isoform. Interestingly, the cytosolic rat brain PSA has been shown to be developmentally regulated, leading to the proposal that the intracellular enzyme could be required for phasic processes related to the development of the central nervous system (7). The recent isolation of a Drosophila melanogaster cytosolic PSA mutant exhibiting defects in spermatogenesis indicates likely structure-function conservation of both isoforms of this aminopeptidase across evolutionary well diverged species (8).
An aminopeptidase with increased sensitivity toward puromycin, relative to PSA, and with biochemical properties distinct from the ubiquitous PSA, has been shown to localize exclusively to neurons of the central nervous system of rats (9). The predominant intracellular localization of this enzyme and its developmental expression profile suggest a role in neuronal growth and differentiation (10). The in vivo substrates and precise physiological roles of these different aminopeptidases and their isoforms in the central nervous system remain to be elucidated.
Recently, human cytosolic PSA has been shown to hydrolyze cytotoxic T cell epitope precursors and hence contribute to the processing of major histocompatibility complex class I peptides (11,12). The cytosolic isoform has also been implicated in cell cycle regulation, because puromycin causes cell cycle arrest of murine COS cells (13). Addition of a non-competitive inhibitor of PSA to a human acute lymphoblastic leukemia cell line also caused reduced tumor cell invasion (14,15). The identification within murine PSA of two motifs, conserved among 26 S proteasome subunits and microtubule-binding sites, coupled with the subcellular localization of PSA to mitotic spindles in mammalian cell lines, may offer some insight into the mode(s) of action of the enzyme (13,16).
The sequencing of the entire genome of the nematode Caenorhabditis elegans (17), and the powerful molecular and genetic techniques applicable to this organism, make it an excellent model system for the study of biological processes. The C. elegans genome potentially encodes ϳ320 peptidases, which have been classified into five families on the basis of their catalytic mechanisms (18). The metallopeptidase family contains most members, several of which have been defined by classic genetic approaches to reveal diverse roles in development, growth, and behavior (19 -24). Genome-wide reverse genetics offers an alternative strategy for identifying physiological roles for C. elegans genes, although to date only a small number of metallopeptidases have been defined using such approaches (25,26).
Analysis of the C. elegans genome predicts a single aminopeptidase gene (F49E8.3) with high homology to the cytosolic human PSA. In the present study we show that the C. elegans PSA orthologue, termed PAM-1 (puromycin-sensitive aminopeptidase), is expressed throughout development in intestinal and nerve cells and exhibits an RNAi phenotype consistent with roles in embryo development and reproduction.
Microscopy-Transgenic C. elegans transmitting the pLS47 construct were examined using a Leica HC inverted microscope (Leica Microsystems, Wetzair, Germany) at an excitation wavelength of 475 nm. Images were captured using a charge-coupled device camera and OPENLAB 3.0.4 image capture software (Improvision Inc., Lexington, MA).
RNA Interference-A 426-bp fragment comprising exons 1 and 2 of pam-1 was excised from pLS22 (see below), using SalI-PstI, and subcloned into the corresponding sites of the plasmid L4440 (pPD129.36) (A. Fire vector kit 1999) to generate pLS31. E. coli HT115(DE3) harboring pLS31 was used as the food source for the C. elegans RNAisensitive mutant rrf-3 (28) to silence pam-1 gene expression as described previously (29,30). Briefly, expression of dsRNA was induced at 37°C in a mid-log phase culture of E. coli HT115(DE3) harboring pLS31, by addition of 0.4 mM IPTG. After a period of 4 h of growth at 37°C, a further 0.4 mM IPTG was added to the culture, and bacteria were aliquoted onto NGM agar plates supplemented with 1 mM IPTG. L4 stage larvae were allowed to feed on the bacteria for 3 days at 20°C, and then individual L4 animals (F1 generation) were transferred onto plates, prepared as above. Nematodes were transferred onto fresh plates at 24-h intervals, for a period of 3 days, and scored for numbers of oocytes, eggs, and progeny. Controls, using HT115(DE3) as the food source, were carried out in parallel using an identical protocol.
Constructs Used to Generate the Recombinant PAM-1-The complete cDNA encoding PAM-1 was PCR-amplified with Platinum Pfx polymerase (Invitrogen, Paisley, UK) and primers OL32 (5Ј-CTGCAGGCG-GCCTGTGGAAACCCAAGC-3Ј) and OL33 (5Ј-GGTACCTTAAACGTT-GCTCTGTTTAAGA-3Ј), using ZAP II yk348g1 as the template (94°C for 2 min followed by 20 cycles of 94°C for 30 s, 50°C for 30 s, and 68°C for 2.5 min). The 2.7-kb PCR product was cloned into pCR-XL-TOPO (Invitrogen) to generate pLS22. The construct was sequenced to confirm integrity and then excised and subcloned into the bacterial expression vector pBAD/His C (Invitrogen), using PstI and KpnI restriction sites, to generate pLS28. Trial expression and purification, as detailed below, revealed that recombinant PAM-1 interacted weakly with the Ni-NTA resin. To improve binding of the recombinant protein to the Ni-NTA resin the N-terminal hexahistidine tag was engineered by mutagenesis to contain 10 histidine residues. Briefly, two independent single-cycle (94°C for 30 s, 55°C for 1 min, and 68°C for 13 min and 20 s) primer extension reactions were performed, using PfuTurbo DNA polymerase (Stratagene, La Jolla, CA) and oligonucleotides OL44 (5Ј-GGGGTTCT-CATCATCATCATCATCATCATCATCATCATGG-3Ј) and OL45 (5Ј-CCATGATGATGATGATGATGATGATGATGATGAGAACCC-3Ј) (nucleotides in boldface encode the additional 4 histidine residues), with pLS28 as the template. Equal volumes of each primer extension reaction were mixed, and mutagenesis was carried out by thermocycling (18 cycles of 94°C for 30 s, 55°C for 1 min, and 68°C for 13 min and 20 s) using the QuikChange site-directed mutagenesis kit (Stratagene) as described by the manufacturer. Subsequent DNA sequence analysis confirmed the integrity of the mutated plasmid, designated pLS35.
Expression and Purification of Recombinant PAM-1-E. coli TOP10 harboring pLS35 was grown at 37°C to mid-log phase, and expression of PAM-1 was induced by addition of 0.2% (w/v) arabinose. Four hours post-induction, cells were harvested and recombinant PAM-1 was purified from the soluble fraction by Ni-NTA affinity chromatography as described by the manufacturer (Qiagen, Crawley, West Sussex, UK). The recombinant enzyme was eluted from the Ni-NTA column with 50 mM imidazole, and 4 ϫ 1-ml fractions were collected and analyzed by SDS-PAGE. The eluate containing purified PAM-1 was passed through a 30K microsep concentrator (Flowgen, Lichfield, Staffordshire, UK) to allow buffer exchange with 100 mM Tris-HCl, pH 7.6. The recombinant enzyme was stored at Ϫ20°C in 50% (v/v) glycerol. Protein quantification was carried out using the bicinchoninic acid assay (Pierce, Rockford, IL) and bovine serum albumin as the standard.
Biochemical Assays-Amino acid methyl coumarin (AMC) substrates were purchased from Sigma-Aldrich and Bachem (Bubendorf, Switzerland). Assays were performed at ambient temperature in a Wallac Victor 2 (PerkinElmer Life Sciences, Beaconsfield, Buckinghamshire, UK) microplate reader adapted with emission and excitation filters of 380 and 460 nm, respectively. Each assay was performed in 50 mM Tris-HCl, pH 7.6, and comprised 10 ng of recombinant PAM-1 and a range of substrate concentrations between 1 M and 0.8 mM in the presence of 1 mg ml Ϫ1 bovine serum albumin and 1 mM glutathione. Inhibitors were purchased from Sigma-Aldrich and preincubated for 15 min at varying concentrations with enzyme in the assay buffer. Inhibitor assays were initiated by addition of 10 M Arg-AMC. Enzyme kinetics was determined using WorkOut (PerkinElmer Life Sciences) and EnzPack (Biosoft) software. The IC 50 values were calculated from semi-log plots with a minimum of nine different concentrations of inhibitor.

Sequence
Analysis of the C. elegans Gene Encoding PAM-1-A full-length expressed sequence tag (yk348g1), derived from a him-8 (high incidence of males) mixed-stage cDNA library, was completely sequenced by primer walking, confirming the 884-amino acid PAM-1 sequence prediction (CE10790) in the C. elegans data base (www.wormbase.org). Examination using SignalP (version 1.1, www.cbs.dtu.dk/services/SignalP) of the genomic sequence upstream of the designated start methionine codon did not reveal the presence of a potential signal peptide sequence, and hence PAM-1 is predicted to be an intracellular aminopeptidase. The amino acid sequence contained the active site residues HEXXH(X) 18 E and GAMEN, as expected for a member of the M1 family of aminopeptidases (Fig. 1A). Residues 227-251 of PAM-1 show 46% identity, 71% similarity, to the putative ␣-type proteasome subunit motif identified in the murine PSA (13). The microtubule-binding site motif identified in mouse PSA (13) is poorly conserved in PAM-1 (data not shown). Overall, PAM-1 is 36% identical to human membrane PSA, 37% identical to human cytosolic PSA, and 38% identical to the Drosophila melanogaster PSA. Cluster analysis confirmed PAM-1 to be the C. elegans PSA orthologue (Fig. 1B).
Expression Pattern Analysis of PAM-1-The spatio/temporal expression pattern of PAM-1 was studied by generation of transgenic C. elegans expressing extrachromosomal arrays of the pPD95.69 plasmid engineered to comprise 4.6 kb of sequence immediately 5Ј of pam-1 and an additional 0.5 kb of pam-1 sequence, generating a second exon translational fusion to GFP (pLS47). Independent lines of transgenic C. elegans expressing the reporter transgene showed fluorescence in embryo, larval, and adult stages (Fig. 2). Embryonic reporter gene activity was observed around the time of gastrulation and continued throughout development (Fig. 2, A-D). Strong fluo-rescence was observed in the intestinal cells of larvae and adults, and this was particularly enhanced in the cells of the posterior gut (Fig. 2E). Larval and adult fluorescence was also observed in neuronal cell bodies around the nerve ring and in processes extending toward the tip of the nose, indicative of them being the amphid sensory neurons (Fig. 2F). The expression pattern in males was determined by mating males (N2, Bristol strain) with hermaphrodites expressing the pam-1::gfp transgene. In addition to the components described above, adult males showed transgene neuronal expression in the developing and mature male tail (Fig. 2, G and H).
pam-1 Has Roles in Embryogenesis and Reproduction-RNA interference was carried out by allowing the RNAi-sensitive mutant rrf-3 to feed on bacteria expressing dsRNA targeting exons 1 and 2 of the pam-1 gene. The F1 larvae appeared to develop normally to adulthood. During the initial 24-h time period of self-fertility, a small number of oocytes were laid by RNAi-treated nematodes, whereas oocytes were never observed on control plates during this time period (data not shown). During the second and third 24-h time periods of self-fertility, the frequency of oocytes laid by RNAi-treated animals was significantly greater (Student's t test, p Ͻ 0.0001) than by nematodes fed on E. coli not expressing dsRNA (Fig. 3A). The penetrance of the phenotype was ϳ60%, and overall, ϳ10-fold more oocytes were laid by animals subjected to dsRNA (243 Ϯ 7, n ϭ 20) than controls (24 Ϯ 3, n ϭ 7). The brood size for the group of nematodes exposed to dsRNA (79 Ϯ 3, mean Ϯ S.E.; n ϭ 20) was not significantly different when compared to that of the control group (79 Ϯ 6, mean Ϯ S.E.; n ϭ 7) (Fig. 3B). In addition to the high numbers of oocytes laid by RNAi-treated hermaphrodites, it was also noted that embryonic lethality occurred (penetrance of ϳ30%, n ϭ 40). On closer inspection, embryogenesis appeared to arrest during gastrulation, generating a mass of undifferentiated cells (Fig. 3C).
Expression and Purification of Recombinant PAM-1-The cDNA encoding PAM-1 was subcloned into pBAD/His C to allow induced expression in E. coli TOP10 of a 6ϫ His tag/ Xpress leader/F49E8.3 fusion protein. The recombinant protein was overexpressed into the bacterial insoluble fraction and to a lesser extent the soluble fraction, but it could not be purified by affinity chromatography due to weak interaction with the Ni-NTA resin. Manipulation of the construct by mutagenesis generated a plasmid (pLS35) for the inducible expression of a 105-kDa 10ϫ His tag/Xpress leader/PAM-1 fusion protein. The solubility characteristics were not altered by the additional His residues (Fig. 4A), and Ni-NTA affinity chromatography utilizing native purification conditions allowed ϳ0.6 mg of recombinant protein to be purified from 1 liter of bacterial culture (Fig.  4B). A single species of 105 kDa, corresponding to the predicted size of the fusion protein, was detected by Western blotting with the anti-6ϫ His tag antibody (Fig. 4C).
Biochemical Characterization of PAM-1-The substrate specificity of the recombinant PAM-1 was studied using amino acid methyl coumarins. Of the substrates tested, those with the highest hydrolytic efficiency (k cat /K m ) were Arg-AMC and Lys-AMC, followed by Leu-AMC, Ala-AMC, Met-AMC, Tyr-AMC, Phe-AMC, Gly-AMC, and Ser-AMC (Table  I). Hydrolysis of Pro-, Glu-, and Asp-AMC substrates was not detected. The optimal pH for hydrolysis of Leu-AMC was between 7.2 and 7.6.
Effect of Inhibitors-The inhibitor profile of PAM-1 was determined by measuring the rate of hydrolysis of Arg-AMC in the presence of different inhibitors (Table II). The recombinant PAM-1 was insensitive to inhibition by the serine protease inhibitor, phenylmethylsulfonyl fluoride, and the cysteine protease inhibitor, E-64. In contrast, the enzyme was sensitive to the chelating agent, 1,10-phenanthroline, as would be expected for a metallopeptidase. The aminopeptidase inhibitors actinonin, amastatin, and leuhistin were the most potent inhibitors of PAM-1 (IC 50 values 0.1-0.6 M). Puromycin inhibited the enzyme with an IC 50 value of 135 M (Fig. 5).
Metal Ion Reactivation of PAM-1-Following complete inactivation of PAM-1 in 4 mM 1,10-phenanthroline, enzyme activity was recovered by addition of Zn 2ϩ , Ni 2ϩ , and Co 2ϩ (Fig. 6). Zn 2ϩ was the most effective ion, restoring the greatest enzyme activity at all concentrations between 1 and 180 M. No restoration of enzyme activity was obtained using Cu 2ϩ , Mn 2ϩ , Mg 2ϩ , Fe 2ϩ , and Hg 2ϩ at concentrations between 0.001 and 2 mM. DISCUSSION The present study describes the role of the puromycin-sensitive aminopeptidase, PAM-1, in the biology of the free-living nematode C. elegans. The expression pattern analysis shows that PAM-1 is expressed throughout development, and the neuronal component is consistent with PSA localization reported for other organisms (4,31). This may reflect an evolutionary conserved role for PSA in the processing of neuropeptides. The precise nature of such a role remains controversial, although studies with mammalian brain thimet oligopeptidase, and neurolysin, demonstrate that cytosolic metallopeptidases are able to function in neuropeptide metabolism (32). One possibility is that PAM-1, which appears to lack a signal sequence, may enter the secretory pathway by a non-classic mechanism, as described for thimet oligopeptidase (33) and neurolysin (34). The alternative possibility is that neuropeptides are internalized into neuronal cells by surface receptors, as demonstrated for neurotensin (35) and opioids (36), and following release, they are processed by cytosolic peptidases. The strong intestinal reporter gene expression indicates that PAM-1 may also function in the processing of peptides, absorbed by the intestinal cells following breakdown of proteins in the gut lumen. Such a function has been proposed for the C. elegans homologues of aminopeptidase P (37) and leucine aminopeptidase (23). Indeed, it is not surprising that a plethora of proteases, including cathepsin B-like (38,39) and cathepsin L-like (40) cysteine proteases and aspartic proteases (41), is expressed in the intestine, where they are able to function in digestion.
The family of Phe-Met-Arg-Phe amide-related neuropeptides (FaRPs) are extensively expressed in neurons and gland cells of C. elegans (42,43), and members are therefore potential PAM-1 substrates. Disruption of one FaRP encoding gene, flp-1, results in multiple behavioral defects, including nose-touch insensitivity (44,45). Recent work using a combination of fluorescence-activated cell sorting and microarray analyses indicates that PAM-1 might be present within C. elegans touch receptors (46), which is consistent with the pam-1 expression pattern reported herein, and hence the possibility exists that it may metabolize FaRP-1. However, examination of the C. elegans RNAi-sensitive mutant rrf-3 (28) following pam-1 (RNAi) did not reveal an observable phenotype related to touch recep-tor deficiency. 2 Nevertheless, the refractory nature of neuronally expressed genes to RNAi is well documented, and it remains a formal possibility that PAM-1 may be involved in sensory perception. RNAi did, however, result in two observ-2 D. R. Brooks, unpublished data.

FIG. 3. Silencing pam-1 by RNAi shows that it has roles in reproduction and embryogenesis.
Numbers of oocytes, embryos, and progeny were recorded throughout the adult hermaphrodite self-fertile period, showing that the RNAi-treated animals laid a high proportion of oocytes (A), whereas brood sizes were not significantly different (B). RNAi-treated, n ϭ 20; control, n ϭ 7. The experiment was repeated in triplicate. C, brightfield images of a typical embryo following arrest during the early stages of gastrulation due to pam-1 (RNAi) (i) and a wild-type embryo undergoing gastrulation (ii).  able phenotypes: (i) the deposition of excessive numbers of oocytes and (ii) embryonic lethality. Many C. elegans mutants with defects in fertilization have been isolated, and genes have been characterized with roles in both the male and female reproductive tracts (47). Metalloproteases from the ADAMs family are involved in aspects of C. elegans germ line development and fertilization (22,48). Proteases have also been shown to activate the amoeboid-like sperm of nematodes in vitro (49), although the relevance of this activation to in vivo events remains unknown. In the presence of pam-1 dsRNA, adult hermaphrodites laid oocytes early and in increasing numbers during the self-fertile period. No significant differences were observed between the brood sizes of dsRNA-treated and control animals. The reporter gene expression pattern confirms that pam-1 has a male-specific component, and hence a role in the male reproductive tract is likely. The brood size data, in conjunction with the observed embryonic lethality (see below), suggests that the number of male germ line nuclei have increased in the RNAi-treated animals. The recent isolation of a D. melanogaster mutant of cytosolic PSA during a screen for animals defective in spermatogenesis (8) may reflect an evolutionary conserved role for this aminopeptidase in the male germ line of invertebrates. The large numbers of oocytes deposited by RNAi-treated nematodes is also indicative of a function for pam-1 in the female germ line. A subtle sperm defect may thus not affect the overall brood size of the RNAi-treated animal provided excess numbers of oocytes are available for fertilization. Disruption of the C. elegans inositol triphosphate signaling pathway results in ovulation of increased numbers of unfertilized oocytes (50). Interestingly, the archetypal M1 metallopeptidase, aminopeptidase N/CD13, has been shown to elevate intracellular Ca 2ϩ levels in monocytes through signal transduction pathways involving inositol triphosphate and mitogen-activated protein kinases (51). In summary, the RNAi data suggest a role for pam-1 in both germ lines of the hermaphrodite, and the lack of pam-1::gfp reporter gene expression in germ line cells was not unexpected given that C. elegans reporter genes rarely express in these cells (52). Microarray analyses of C. elegans mutants defective in development of either the male or female germ line indeed suggest pam-1 as a candidate for expression in both germ lines (53).
The RNAi data reported in this study also demonstrate that PAM-1 is required for embryogenesis, hence extending the physiological roles attributed to PSA. A role in embryo development is consistent with the embryonic reporter gene expression pattern reported herein, and it may reflect an essential role in the cell cycle for PAM-1, despite the lack of a recognizable microtubule-binding site motif (13). It seems reasonable to speculate that PAM-1 processes bioactive peptides involved in signaling events within the germ lines and subsequently during embryogenesis. The presence of a potential ␣-type proteasome subunit motif within PAM-1 may offer an insight into its function. As previously proposed (13) and recently demonstrated with mammalian cells (11,12), PSA may function downstream of the proteasome, with the resulting peptides targeted for release at the cell surface. During C. elegans embryogenesis, the yolk proteins, or vitellogenins, are the major sources of nourishment (54), and hence an alternative explanation for the embryonic lethal phenotype is that PAM-1 may be required to complete digestion of these vital nutrients.
The expression and purification of recombinant PAM-1 allowed analysis of the biochemical properties of the peptidase. With respect to substrate specificity, the enzyme exhibited similarity to cytosolic PSAs of mammals (13,31,55,56), preferring peptides with either a positive charge or small neutral amino acid at the N terminus. The determined K m and catalytic efficiency (k cat /K m ) values of PAM-1 and human liver PSA (56) show good agreement, whereas PAM-1 substrate binding affinity is approximately an order of magnitude less than reported for rat brain PSA (9). This discrepancy in catalytic efficiencies might reflect limitations of the recombinant PAM-1 enzyme preparation, such as incomplete folding, relative to the rodent PSA, which was purified to homogeneity from rat brain (9).
The inhibitor profile of PAM-1 is very similar to other PSAs (9, 56) with the notable exception of inhibition by puromycin. The PAM-1 IC 50 constant for puromycin (135 M) indicates that the nematode enzyme has significantly reduced sensitivity toward this drug. Indeed, human PSA is completely inhibited by 100 M puromycin (56), and IC 50 values are reported to be of the order of 1 M (57). In this respect, PAM-1 is more like aminopeptidase N (K i ϳ 100 M) (58). However, we cannot exclude the possibility that the N-terminal tag present on the recombinant nematode enzyme is influencing binding of puromycin, even though it appears not to affect interactions with the other inhibitors tested.
Divalent cations are essential for metallopeptidase activity, and they bind to the two histidine residues and the distal glutamate present within the active site motif, HEXXH(X) 18 E, of the M1 class of metallopeptidases. Different metal ions can be exchanged within the active site of metallopeptidases, resulting in modulation of catalytic efficiency (59). Following inactivation by metal ion chelation, the activity of PAM-1 could be recovered using Zn 2ϩ , Ni 2ϩ , and Co 2ϩ . Both Zn 2ϩ and Co 2ϩ were able to restore activity to metal ion-depleted mammalian PSA (57). Physiochemical analysis would be required to determine which of these metal ions reside within the active site, although the profile of recovery observed with PAM-1 would suggest Zn 2ϩ as the most likely candidate.
In summary, we have demonstrated using RNAi that the C. elegans PSA orthologue, PAM-1, is involved in reproduction and embryogenesis. The gene expression pattern analysis shows that pam-1 is expressed throughout development, and spatio components indicate that it is also likely to function in the nematode nervous system and in digestion. The biochemical characterization of recombinant PAM-1 shows that many properties are similar to mammalian PSA, with the exception that inhibition of the nematode peptidase by puromycin is attenuated.