A Hypodermally Expressed Prolyl 4-Hydroxylase from the Filarial Nematode Brugia malayi Is Soluble and Active in the Absence of Protein Disulfide Isomerase*

The collagen prolyl 4-hydroxylase (P4H) class of enzymes catalyze the hydroxylation of prolines in the X -Pro-Gly repeats of collagen chains. This modification is central to the synthesis of all collagens. Most P4Hs are (cid:1) 2 (cid:2) 2 tetramers with the catalytic activity residing in the (cid:1) subunits. The (cid:2) subunits are identical to the enzyme protein disulfide isomerase. The nematode cuticle is a collagenous extracellular matrix required for maintenance of the worm body shape. Examination of the model nematode Caenorhabditis elegans has demonstrated that its unique P4Hs are essential for viability and body morphology. The filarial parasite Brugia malayi is a causative agent of lymphatic filariasis in hu-mans. We report here on the cloning and characterization of a B. malayi P4H with unusual properties. The recombinant B. malayi (cid:1) subunit, PHY-1, is a soluble and active P4H by itself, and it does not become associated with protein disulfide isomerase. The active enzyme form is a homotetramer with catalytic and inhibition properties similar to those of the C. elegans on the ExPASy proteomics B. phy-1 Promoter-Reporter Plasmid— of amplified genomic

reticulum (ER) resident enzyme is central to collagen synthesis, as collagen triple helices are thermally unstable in the absence of 4-hydroxyproline residues (2,3). P4H also acts as a chaperone in the assembly of collagen ensuring that only correctly folded collagens are released for secretion (4). In vertebrates and Drosophila melanogaster the enzyme is an ␣ 2 ␤ 2 tetramer (2,3,5,6), with hydroxylation activity residing in the catalytic ␣ subunits. Two ␣ subunit isoforms, ␣(I) and ␣(II), have been characterized in vertebrates (7,8). They become assembled into [␣(I)] 2 ␤ 2 and [␣(II)] 2 ␤ 2 tetramers with insect cell coexpression data arguing strongly against the formation of mixed ␣(I)␣(II)␤ 2 tetramers (8). The ␤ subunits of P4Hs are identical to the enzyme and chaperone protein disulfide isomerase (PDI) (EC 5.3.4.1) (9) and are required to maintain the ␣ subunits in a catalytically active nonaggregated conformation (10,11). The P4H is also maintained within its correct subcellular compartment by virtue of an ER retention signal at the C terminus of PDI (11). When expressed alone in a recombinant expression system, the ␣ subunits are insoluble and inactive, whereas coexpression with PDI results in the formation of an active, soluble P4H (7,(12)(13)(14). The PDIs from different organisms can often substitute for the authentic partner, such as the human PDI, which can function as a ␤ subunit in mouse and Drosophila P4H tetramers and in a P4H dimer with Caenorhabditis elegans PHY-1 (5,7,13).
In the model nematode C. elegans the cuticular collagen modifying function of P4H is essential for body morphology and viability (15,16). In nematodes the exoskeleton (known as the cuticle) is an extracellular matrix composed of small collagenlike molecules (17). The nematode cuticle is synthesized by the underlying hypodermal tissue and performs multiple functions including maintenance of worm body shape. Mutations in collagens forming the cuticle and in the enzymes involved in collagen biosynthesis can result in lethality and severe alterations to body shape, as illustrated by the C. elegans sqt-3 (18) and bli-4 (19) mutant phenotypes.
The P4Hs in C. elegans that are involved in the synthesis of cuticle collagens are formed from the ␣ subunits PHY-1 and PHY-2 and the ␤ subunit PDI-2 (14,15). The expression, function, and assembly of these subunits have been examined in detail showing that unique P4H forms exist in C. elegans. The most abundant form is a tetramer; this differs, however, from others described, being a mixed PHY-1/PHY-2/(PDI-2) 2 tetramer (14). PHY-1 and PHY-2 can also each individually associate with PDI-2 to form dimers (14). Such P4H forms have not been described for any other species to date. Genetic disruption of phy-1 and phy-2 simultaneously, or pdi-2 singly, results in embryonic lethality in which embryos develop normally until the first cuticle is required to maintain the elongated worm shape (15,16). The weakened cuticle is then unable to maintain this form, after which embryos collapse to a disorganized state and eventually die. The body shape defect and reduced 4-hydroxyproline levels in the cuticle collagens of the viable phy-1 null genetic mutant, dpy-18, underline the importance of collagen modification by P4H for nematode body morphology (15,16).
Applying the knowledge of P4H function in C. elegans, we examined a P4H in the filarial parasitic nematode Brugia malayi. B. malayi along with Brugia timori and Wuchereria bancrofti are the causative agents of lymphatic filariasis in humans, with over 120 million people infected and over 1 billion people at risk of infection worldwide (20). Lymphatic filariasis is a debilitating disease, with approximately onethird of those infected being incapacitated and/or disfigured by the infection (21). Commercially available inhibitors of P4H have been shown to be toxic to B. malayi adults, producing associated cuticular defects (22). These observations and the requirement in C. elegans for P4H activity highlight this enzyme class as a potential drug target in the control of human and veterinary parasitic nematode infections.
In this article we describe the identification of a P4H phy-1 gene from B. malayi, characterize the molecular and enzymatic properties of the recombinant B. malayi P4H produced in an insect cell expression system, and examine the expression profile and putative function of the B. malayi phy-1 gene by heterologous expression in the model nematode C. elegans. Unusually, the B. malayi PHY-1 is a soluble and active P4H when expressed alone in a recombinant system in the absence of PDI, and it does not associate with PDIs from other organisms, including C. elegans. The developmental temporal expression pattern of the B. malayi phy-1 gene was analyzed by RT-PCR using stage-specific mRNA samples. Reporter gene experiments showed that the B. malayi phy-1 promoter directs tissuespecific spatial expression to the hypodermal cells of C. elegans.

MATERIALS AND METHODS
Nematode Strains and Culture Conditions-C. elegans strains were cultured as described elsewhere (23). The wild type Bristol N2, CB364 (dpy-18(e364)), and DR96 (unc-76(e911)) C. elegans strains were provided by the Caenorhabditis Genetics Centre. The B. malayi nematodes were provided by Rick Maizels (University of Edinburgh).
Isolation of cDNA and Genomic Clones-ESTs SW3D9CA480SK, MBAFCX8G05T3, and MBAFCZ7H09T3 ( Fig. 1A) were received from the Filarial Genome Project, subcloned, and sequenced. The primers X8G5F1 (5Ј-CAGTCGCTCAACACCGG-3Ј) and BMNPHYR (5Ј-CCAAT-AGTATTTAAGCAC-3Ј) were designed from the EST sequences and used to obtain a 312-bp PCR product (Fig. 1A) from B. malayi adult stage cDNA prepared as described previously (15). The purified PCR product was labeled with [␣-32 P]dCTP and used to screen a B. malayi adult male cDNA library, SAW94NLBmAm (from Steven Williams, Filarial Genome Project, Northampton, MA). Eleven positive clones, all representing the same cDNA, named B. malayi phy-1, were identified from a total of 2 ϫ 10 4 recombinants. Amplification of the 5Ј end of the B. malayi phy-1 cDNA was performed using the Invitrogen 5Ј-RACE system. Two 418-bp products from independent 5Ј-RACE PCR reactions (Fig. 1A) were sequenced to generate a consensus 5Ј sequence. The primers BMPHY-1RESF(BamHI), and BMPHY-1RESR(NotI) (for primer sequences, see "dpy-18 Rescue Experiments with B. malayi phy-1") were used to generate the B. malayi phy-1 genomic sequence from the translation start codon ATG to the stop codon TAA using B. malayi genomic DNA as a template. Three identical full-length clones were then fully sequenced to generate a consensus genomic sequence. Computer prediction for analysis of conceptual translation of B. malayi phy-1, protein alignments, signal peptide analysis, and predicted posttranslational modifications were performed on the ExPASy proteomics tools data base (www.expasy.ch).

Generation of a Recombinant Baculovirus coding for B. malayi PHY-1, and Expression and Analysis of Recombinant Proteins in Insect
Cells-The full-length coding sequence of B. malayi phy-1 cDNA was cloned from B. malayi adult stage cDNA by PCR using Pfu polymerase (Stratagene) with the primers BMPHY-1BVF(NotI) (5Ј-gagcggccgcAT-GATAGCTACCGTGGTGTTC-3Ј) and BMPHY-1BVR(XbaI) (5Ј-gctcta-gaTTAAGCACTTAGATCGCCCAC-3Ј) (artificial restriction sites are set in lowercase and underlined). The PCR product was cloned into a NotI-XbaI-digested transfer vector pVL1392 (BD Pharmingen) and sequenced. The recombinant vector was cotransfected into Spodoptera frugiperda Sf9 cells with a modified Autographa californica nuclear polyhedrosis virus DNA (BaculoGold, BD Pharmingen) by calciumphosphate precipitation.
Sf9 or High Five (Invitrogen) insect cells were cultured as monolayers in TNM-FH (modified Grace's insect cell medium) medium (Sigma) supplemented with 10% fetal bovine serum (BioClear) or in suspension in Sf900IISFM serum-free medium (Invitrogen). The cells were seeded at a density of 5 ϫ 10 6 cells/100-mm plate or 1 ϫ 10 6 cells/ml and infected at a multiplicity of 5 with the virus coding for the B. malayi PHY-1 alone or together with viruses coding for C. elegans PDI-1, PDI-2, or human PDI. In control experiments, the cells were coinfected with the viruses coding for C. elegans PHY-1, PHY-2, and PDI-2, with PHY-1 and human PDI, or with the various PDI viruses alone. The cells were harvested 72 h after infection, washed with a solution of 0.15 M NaCl and 0.02 M phosphate, pH 7.4, homogenized in a 0.1 M NaCl, 0.1 M glycine, 10 M dithiothreitol, 0.1% Triton X-100, and 0.01 M Tris buffer, pH 7.4, and centrifuged at 10,000 ϫ g for 20 min. The pellets were further solubilized in 1% SDS. Aliquots of the samples were analyzed by 8% SDS-PAGE under reducing conditions and by nondenaturing PAGE followed by Western blotting with polyclonal antibodies against B. malayi PHY-1 (see below), C. elegans PDI-1 or PDI-2 (14), or a monoclonal antibody against human PDI (5B5, Dako). P4H activity was assayed by a method based on the hydroxylation-coupled decarboxylation of 2-oxo-[1-14 C]glutarate, and K m and K i values were determined as described previously (24). The molecular weight of the recombinant B. malayi P4H was analyzed by applying the Triton X-100soluble fraction of insect cells expressing B. malayi PHY-1 to a calibrated HiPrep Sephacryl S-200 HR gel filtration column (Amersham Biosciences), equilibrated and eluted with 0.1 M NaCl, 0.1 M glycine, 10 M dithiothreitol, and 0.01 M Tris buffer, pH 7.4, and P4H activity was assayed in the eluted fractions. Purified recombinant human type I P4H (12) was used as a control in gel filtration experiments.
Polyclonal antiserum was raised in two rabbits against a synthetic peptide corresponding to the C-terminal region of B. malayi PHY-1. The peptide CRRPCGLSRSVEEQFVGDLSA was conjugated to keyhole limpet hemocyanin (Sigma GenoSys) via an added cysteine residue (underlined) and used for immunization.
The B. malayi phy-1 cDNA (Ϯ a synthetic intron) and genomic rescue constructs were microinjected into the syncytial gonad of C. elegans phy-1 null, dpy-18(e364), nematodes at concentrations of 10 and 100 g/ml. A marker plasmid with a dpy-7 cuticle collagen promoter in the green fluorescent protein (GFP) fusion vector pPD95-67 (from Iain Johnstone, University of Glasgow) was coinjected at 5 g/ml, and the injection mixes were made up to a final concentration of 150 g/ml with pBluescript SKM. Transformants were selected by GFP fluorescence, and more than five semi-stable transmitting lines were examined for each concentration.
Cloning of the B. malayi phy-1 Promoter-To identify the putative promoter region, a genomic B. malayi BAC library was screened with a 1.7-kb biotin-labeled probe generated by PCR using the primers T7PL (5Ј-CTCACTATAGGGCGAATTGG-3Ј) (New England Biolabs) and BMPHY1.1IS3R(B) (5Ј-GCGTGGATGATTTGGATC-3Ј) and a plasmid containing a T7 site and the 5Ј genomic coding sequence from B. malayi phy-1 as a template. The gridded BAC filters were hybridized using NEBlot Phototope and Phototope-star (New England Biolabs) detection kits. BLASTX analysis was performed on the ExPASy proteomics tools data base.
Construction of a B. malayi phy-1 Promoter-Reporter Plasmid-A 2.2-kb putative promoter fragment of the B. malayi phy-1 gene was amplified from B. malayi genomic DNA using Pfu polymerase and the primers BMPHY-1PF(SphI) (5Ј-ggcgcatgcGAATGAGACAATTGCACA-AG-3Ј and BMPHY-1PR(BamHI) (5Ј-ggcggatccGCTATCATCACTGGC-TCTGGA-3Ј). This fragment, extending from Ϫ2189 to ϩ8 relative to the translation start site, was cloned into SphI-BamHI-digested C. elegans reporter gene vector, pPD96-04. This was microinjected into the syncytial gonad of the C. elegans strain DR96(unc-76) together with the unc-76 rescue plasmid (p7616B), both at 100 g/ml. Six semi-stable transgenic lines were identified and examined for reporter gene expression by viewing GFP expression in live worms and by sensitive staining of fixed worms for ␤-galactosidase activity (26). Live nematodes were transferred to 2% agarose, 0.065% sodium azide pads, and images were taken with an Axioskop 2 microscope using a Hamamatsu digital camera and Improvision Openlab processing software.

Cloning of B. malayi phy-1-Three
ESTs from the B. malayi data base were identified that encode amino acid sequences homologous to C. elegans PHY-1 and PHY-2. MBAFCX8G05T3 (AA509222) and MBAFCZ7H09T3 (AA406985) were both derived from a B. malayi adult female cDNA library and SW3D9CA480SK (AA585698) from a B. malayi L3 cDNA library (GenBank TM accession numbers in parentheses). Sequencing of the ESTs established that they were all derived from a single gene, termed B. malayi phy-1 (Fig. 1A). A 312-bp PCR probe was generated from B. malayi adult stage cDNA, based on the EST sequences, and used to screen a B. malayi adult male cDNA library (Fig. 1A). Eleven positive clones were identified; full-length sequencing was performed on two 1.6-kb clones, and both were found to represent the B. malayi phy-1 cDNA. The nine other identified clones contained inserts with sizes ranging from 0.6 to 1.6 kb, and all corresponded to the B. malayi phy-1 gene. The 1669-bp sequence generated from the library screening contained a 1515-bp open reading frame and 154-bp of 3Ј-UTR sequence (Fig. 1A). The 3Ј untranslated region did not contain a consensus polyadenylation signal (AATAAA). A divergent poly(A) signal sequence (GATAAA) was located, however, 11-bp upstream of the poly(A) tail, representing a variant poly(A) signal sequence also found in ϳ5% of the C. elegans genes examined (27).
A comparison of the amino acid sequences encoded by the 1515-bp B. malayi phy-1 cDNA sequence with known P4H ␣ subunits signified that the 5Ј coding sequence was incomplete. Additionally, according to the signal sequence prediction program, SignalP, the N terminus did not contain a characteristic signal peptide. The 5Ј-UTR, the trans-spliced leader sequence, and the coding sequence for 37 additional N-terminal amino acids was obtained using the 5Ј-RACE system for rapid amplification of cDNA ends (Fig. 1A). These data were assembled with the 1669-bp B. malayi phy-1 sequence obtained from the cDNA libraries to give the full-length cDNA sequence, which was confirmed by the sequencing of a full-length Pfu-generated PCR product (Fig. 1A). The complete B. malayi phy-1 cDNA sequence contains a consensus 22-bp SL1 trans-spliced leader sequence, an 8-bp 5Ј-UTR, a single open-reading frame of 1626-bp that encodes a 541 amino acid polypeptide (GenBank TM accession No. AJ297845), and a 154-bp 3Ј-UTR (Fig. 1A).
The 4596-bp consensus genomic sequence of the B. malayi phy-1, determined from three individual full-length PCR products, contains 12 exons and 11 introns ( Comparison of the B. malayi PHY-1 Amino Acid Sequence with Those of Other P4H ␣ Subunits-A signal peptide cleavage site between Ala-17 and Asp-18 was predicted by the SignalP program, and thus the processed B. malayi PHY-1 consists of 524 amino acids. The highest amino acid sequence homology was found between the processed B. malayi PHY-1 and C. elegans PHY-1 sequences (13), the identity being 59% and similarity 76%, whereas the identity and similarity between the B. malayi PHY-1 and C. elegans PHY-2 (15,16) are 53 and 71%, respectively (Fig. 2). The amino acid sequence homology between the B. malayi and C. elegans PHY polypeptides is slightly higher than that between the B. malayi PHY-1 and the PHY-1 from a closely related filarial nematode, Onchocerca volvulus (22), with the B. malayi and O. volvulus PHY-1 polypeptides being 49% identical and 70% similar. The amino acid sequence identities between the B. malayi PHY-1 and the human ␣(I) and ␣(II) subunits are 45 and 44%, and the similarities are 62 and 63%, respectively. The cysteine residues essential for intrachain disulfide bonding (28,29) and the ac- tive site histidine, aspartic acid, and lysine residues (29,30) are all conserved in B. malayi PHY-1 (Fig. 2). The extended Cterminal regions present in the C. elegans and O. volvulus PHY-1 polypeptides are not found in B. malayi PHY-1 (Fig. 2).
B. malayi PHY-1 is a Soluble and Active P4H When Expressed in the Absence of PDI in a Baculovirus System-The B. malayi phy-1 cDNA was cloned into the baculovirus expression vector pVL1392, a recombinant baculovirus was generated and used to infect insect cells. The cells were harvested 72 h after infection, homogenized in a 0.1% Triton X-100-containing buffer, and centrifuged. The remaining pellet was solubilized in 1% SDS, and the samples were analyzed by reducing SDS-PAGE followed by Western blotting (Fig. 3A). The C. elegans PHY polypeptides (13,14) and vertebrate P4H ␣ subunits (7,12) require association with PDI to form soluble and active P4Hs. When expressed alone in a recombinant system, these polypeptides form inactive aggregates, and 1% SDS is required for their efficient solubilization (7,(12)(13)(14). In contrast to these P4H ␣ subunits, Western blotting showed that the majority of the recombinant B. malayi PHY-1 was soluble in the Triton X-100-containing buffer (Fig. 3A, lane 1), and only a minor amount formed insoluble aggregates that required 1% SDS for solubilization (Fig. 3A, lane 2). The Triton X-100 extracts from insect cells expressing the B. malayi PHY-1 were analyzed for P4H activity with an assay based on the hydroxylation-coupled decarboxylation of 2-oxo-[1-14 C]glutarate (24), and a significant amount of P4H activity was observed ( Table I). The B. malayi PHY-1 thus differs from the C. elegans PHY polypeptides and the vertebrate P4H ␣ subunits, which require association with PDI to form soluble and active P4Hs (7,(12)(13)(14).
To study whether B. malayi PHY-1 has the potential to associate with PDI, the recombinant protein was expressed in insect cells either alone, with the C. elegans ␤ subunit, PDI-2 (31), with the C. elegans PDI isoform, PDI-1 (31), or with the human PDI (12). In control experiments the insect cells were infected with viruses coding for the various PDIs alone or coinfected with viruses coding for the C. elegans PHY-1/PHY-2/(PDI-2) 2 tetramer or the hybrid C. elegans PHY-1/human PDI dimer. Triton X-100-soluble extracts of the cell lysates were analyzed by nondenaturing PAGE followed by Coomassie Blue staining (Fig. 3B). In the control experiments, bands corresponding to the C. elegans PHY-1/PHY-2/(PDI-2) 2 tetramer (Fig. 3B, lane 5) and the C. elegans PHY-1/human PDI dimer . Cells were harvested and homogenized as described above, and the Triton X-100-soluble fractions were analyzed by nondenaturing PAGE followed by Coomassie Blue staining. The C. elegans PHY-1/PHY-2/(PDI-2) 2 tetramer (T) and the C. elegans PHY-1/ human PDI dimer (D) are indicated by arrows. The band indicated by an asterisk is also found in extracts from cells infected with a wild type baculovirus (not shown), and the faint bands below it represents the nonassociated PDI subunits. (Fig. 3B, lane 6) were detected. In contrast, neither a tetramer nor a dimer was detected by Coomassie Blue staining in the extracts from cells expressing the B. malayi PHY-1 alone or in combination with different PDIs (Fig. 3B, lanes 1-4). However, three B. malayi PHY-1 immunoreactive bands were detected by nondenaturing Western analysis in extracts from cells expressing B. malayi PHY-1 alone (Fig. 4A, lane 1). Three bands with approximately the same mobilities were also detected in the extracts from cells coexpressing the B. malayi PHY-1 and the different PDIs (Fig. 4A, lanes 2-4). The mobility of the B. malayi PHY-1 upper band was similar to those of the C. elegans PHY-1/PHY-2/(PDI-2) 2 tetramer (Fig. 4A, lane 8) and the human P4H tetramer (data not shown), and the mobility of the middle band was likewise similar to that of the PHY-1/PDI-2 dimer (Fig. 4A, lane 8). No B. malayi PHY-1 immunoreactive bands were detected in the extracts from cells expressing the different PDIs alone (Fig. 4A, lanes 5-7). Coexpression of the recombinant B. malayi PHY-1 with the different PDIs did not significantly increase the amount of P4H activity obtained (Table I), and Western blotting of the nondenaturing PAGE with antibodies against the different PDIs showed that the B. malayi PHY-1 does not associate with the human PDI (Fig.  4B), C. elegans PDI-1 (Fig. 4C), or C. elegans PDI-2 (Fig. 4D). This is in contrast to the vertebrate, C. elegans, and Drosophila ␣ subunits, which can each associate with orthologous P4H ␤ subunits (5,7,13).
Gel filtration experiments in a calibrated HiPrep Sephacryl S-200 HR column of the Triton X-100 extracts from cells expressing the recombinant B. malayi PHY-1 alone showed that P4H activity was eluted in fractions that corresponded to a molecular weight of ϳ350,000 (Fig. 5A). Control experiments with purified recombinant human type I P4H tetramer showed that it eluted in exactly the same position as the recombinant B. malayi P4H (Fig. 5A). Previous gel filtration studies have also shown that the elution position of a purified chick embryo P4H tetramer corresponds to a molecular weight of 350,000 (32). The calculated molecular weights of a B. malayi PHY-1 tetramer and the human type I P4H tetramer are 241,468 and 228,808, respectively. The fractions containing B. malayi P4H activity were pooled and analyzed by reducing SDS-PAGE and nondenaturing PAGE followed by Western blotting (Fig. 5, B and C). Two B. malayi PHY-1 immunoreactive bands corresponding to the nonglycosylated and glycosylated forms (see Fig. 6B) were detected in SDS-PAGE (Fig. 5B), and two bands with mobilities similar to the human or C. elegans P4H tetramers and a C. elegans P4H dimer, respectively, were detected in the nondenaturing PAGE (Fig. 5C). No detectable B. malayi P4H activity was eluted from the gel filtration column in a position that corresponds to that of the C. elegans P4H dimer ( Fig. 5A) (13). Therefore, our results indicate that the B. malayi PHY-1 self-associates into active P4H tetramers (Figs. 4A and 5C), but during the nondenaturing PAGE partial dissociation of the B. malayi P4H tetramers into dimers (Figs. 4A and 5C) and monomers (Fig. 4A) occurs.
Analysis of Tissue-extracted B. malayi PHY-1 and Native Glycosylation-Three immunoreactive recombinant B. malayi PHY-1 bands were observed when Triton X-100-soluble extracts from insect cells were analyzed by nondenaturing PAGE and Western blotting (Fig. 4A, lane 1). Native extracts were prepared from the nematodes and were likewise analyzed to determine whether similar PHY-1 immunoreactive bands are found in B. malayi in vivo. Two major immunoreactive bands were observed in freshly prepared parasite extracts (Fig. 6A, lane 1) that have a native gel migration pattern similar to the characterized C. elegans PHY-1/PHY-2/(PDI-2) 2 tetramer and PHY-1/PDI-2 dimer from native C. elegans extracts (Fig. 6A,  lane 2).
The Proscan program (ExPASy) predicted that B. malayi PHY-1 has two N-linked glycosylation sites at positions 49 -52 (NRSL) and 140 -143 (NASG) (amino acid positions given relate to the mature processed protein) (Fig. 2). The extracts from B. malayi nematodes and from insect cells expressing recombinant B. malayi PHY-1 were treated with N-glycosidase F and analyzed by reducing SDS-PAGE followed by Western blotting (Fig. 6B) to compare the glycosylation of native and recombinant B. malayi PHY-1 polypeptides. Following glycosidase treatment of insect cell and worm extracts (Fig. 6B, lanes 1 and  3), a single band of ϳ60 kDa was observed in both samples, the size of the immunoreactive band being consistent with the predicted size of B. malayi PHY-1. In addition to the 60-kDa band, an additional band migrating slightly higher was detected in the untreated insect cell extract (Fig. 6B, lane 2). The 60-kDa band was not detected in the untreated worm extract (Fig 6B, lane 4), but instead two bands with higher mobilities were observed, representing B. malayi PHY-1 polypeptides in which one or both glycosylation sites have been modified (Fig.  6B, lane 4). Thus, approximately half of the recombinant B. malayi PHY-1 polypeptides expressed in insect cells remain unglycosylated, whereas in the remaining polypeptides, only one of the glycosylation sites is modified. In B. malayi in vivo no unglycosylated forms are present, and a minority of PHY-1 polypeptides are glycosylated at both sites.
Examination of the Functional Conservation between the B. malayi and C. elegans PHY-1 Polypeptides-Phenotypic rescue of the C. elegans phy-1 null mutant (dpy-18(e364)) was attempted to assess interspecies conservation of the phy-1 gene function. Attempts were made to rescue the C. elegans mutant strain CB364(dpy18(e364)) using the B. malayi phy-1 coding sequence expressed under the control of the C. elegans phy-1 promoter (Fig. 7). The B. malayi PHY-1 should therefore be expressed in the relevant tissues with appropriate developmental timing and at levels comparable with that of the C. elegans PHY-1. A vector was constructed that contained the previously defined C. elegans phy-1 promoter (15) and a splice site containing 3Ј-UTR with the polyadenylation signal and poly(A) transfer sequences (13) (vector pAW1) (Fig. 7B). The B. malayi phy-1 coding sequence, with and without introns, was inserted between the two C. elegans sequences (Fig. 7B) and transformed into the CB364 strain, and the ability to rescue the medium dumpy phenotype in the transformed animals carrying the B. malayi PHY-1-encoding transgene was analyzed. Multiple semi-stable lines of nematodes transformed with 10 and 100 g/ml of the rescue construct containing the B. malayi phy-1 cDNA sequence were examined and found not to rescue the dpy-18 phenotype (data not shown). The lack of introns may result in extremely low levels or absence of expression of the heterologous protein (33,34), and therefore an artificial intron was synthesized based on standard C. elegans introns and inserted into the B. malayi phy-1 cDNA coding sequence (Fig. 7B). Microinjection of this cDNA construct containing a synthetic intron at 10 and 100 g/ml concentrations likewise failed to rescue the dumpy phenotype (data not shown). The ability of the genomic B. malayi phy-1 sequence to rescue the dpy-18 phenotype was also studied. High concentrations (100 g/ml) of this construct were toxic, however, as all of the transformed nematodes died during embryogenesis, and no transformed lines could be generated. Injections at lower concentrations (10 g/ml) yielded transformed lines, but examination of multiple lines again revealed a failure to rescue the dumpy phenotype (data not shown).

Temporal Expression Analysis of B. malayi phy-1 mRNA-
Data from ESTs and library screenings indicated that the B. malayi phy-1 mRNA was expressed in adult females, males, and L3 larvae. A more detailed analysis of the temporal expression pattern of B. malayi phy-1 in developmental stages from L3 to adults was performed by RT-PCR using mRNA samples prepared from extracts isolated from infected jirds; daily extracts were taken up to day 14 post-infection, after which extracts were taken every 2-4 days (25). The 655-bp PCR product from B. malayi phy-1 cDNA can be distinguished from that produced by any genomic contamination, as the primers applied in this study were designed to span introns. A second set of primers, applied simultaneously in the same PCR reactions, was used to amplify a fragment of the tubulin gene as an internal control. The B. malayi phy-1 transcript was detectable in all L3, L4, and adult samples analyzed (Fig. 8). Expression was likewise found in an L1 (microfilaria) cDNA sample (data not shown). Taken together with the EST and library screening data, the RT-PCR results from stage-specific samples show that the B. malayi phy-1 is expressed in L1 (microfilaria), at all time points examined throughout L3 and L4 development, and in both adult sexes. Visual examination of this expression profile reveals peaks of abundance corresponding to the L3-L4 and L4 to adult molts (Fig. 8, lanes 8 and 17).  , lane 8). The cells were harvested and homogenized as described in the legend to Fig. 3. Triton X-100-soluble fractions were analyzed by nondenaturing 8% PAGE followed by Western blotting using antibodies against B. malayi PHY-1 (A, lanes 1-7) Analysis of the Function of the B. malayi phy-1 Promoter Region-A 2.2-kb region of the putative B. malayi phy-1 promoter sequence (GenBank TM accession No. AJ421994) was identified by screening a genomic BAC Brugia library. This 2.2-kb region does not contain any other coding sequences as determined by BLASTX analysis. A 3Ј trans-splicing acceptor site (27) is located at position Ϫ15 to Ϫ8 with respect to the translation start codon. The 2.2-kb promoter fragment followed by the codons for the first three amino acids of B. malayi PHY-1 was generated by PCR and fused in-frame to lacZ/GFP in the multi-intron reporter gene vector, pPD96-04. The reporter gene construct was transformed into C. elegans at 100 g/ml, and multiple lines were examined, with expression found consistently in hypodermal cells of all lines examined (Fig. 9, A-D). GFP expression was particularly prominent in the hypodermal cell nuclei hyp5, hyp6, and hyp7 of the head (Fig. 9A) and in the pair of hyp7 cell nuclei in the tail of adult worms (Fig. 9B). Using sensitive staining techniques for the detection of ␤-galactosidase activity, expression was also detected in larval stages, again strongly in the head and tail regions (Fig. 9, C  and D). The lacZ expression was less pronounced in the hypodermal cells of the body (Fig. 9C), with expression being detected in the hypodermal cell nuclei hyp4, hyp5, hyp6, and hyp7 of the head. With prolonged staining (overnight at room temperature) lacZ expression was also observed in the vulval cells of the adult, which are of hypodermal origin, in the adult body wall muscle cells, and in embryonic stages (data not shown).

DISCUSSION
In this study we have identified and characterized a P4H with unusual properties from the filarial parasite B. malayi. The recombinant B. malayi PHY-1 assembles into an active soluble P4H tetramer when expressed in insect cells in the absence of PDI. High levels of B. malayi phy-1 transcript expression were detected in all parasite stages examined. Analysis of B. malayi phy-1 promoter-reporter gene fusions showed that the construct is expressed in hypodermal cells in a heterologous C. elegans system, thus indicating a conservation of control elements between these species. Although B. malayi PHY-1 appears to function as a P4H independently, without a PDI subunit, the B. malayi phy-1 was not able to replace P4H function in a C. elegans phy-1 mutant.
The essential nature of the cuticle collagen modifying function of P4H has previously been demonstrated in the nematode C. elegans, where animals lacking or deficient in this function   7. Representation of phy-1 null dpy-18(e364) C. elegans rescue constructs. A, depiction of the C. elegans phy-1 gene construct (promoter, genomic sequence, and 3Ј-UTR) successfully used to rescue dpy-18(e364) mutants (ϩ) (15). B, B. malayi PHY-1 rescue constructs contained the C. elegans phy-1 promoter and 3Ј-UTR sequences and the full-length B. malayi phy-1 coding cDNA, the full-length B. malayi phy-1 coding cDNA with a single C. elegans synthetic intron, or the B. malayi phy-1 genomic sequence from the translation initiation codon to the stop codon. Sizes of the fragments are not drawn to scale, and the full number of introns in each of the genomic fragments is not depicted. The constructs were injected into the C. elegans phy-1 null strain CB364(dpy-18) for the examination of interspecies rescue. None of the B. malayi constructs were able to rescue the dumpy phenotype of C. elegans phy-1 null mutants (Ϫ).

FIG. 8. Developmental expression analysis of B. malayi phy-1.
Expression of the B. malayi phy-1 transcript was compared with the tubulin transcript through L3 to L4 and L4 to adult molts. PCR reactions were performed on cDNA samples using primer pairs from both genes simultaneously. B. malayi extracts were prepared from daily samples of infected jirds through L3 and L4 and then at longer intervals after day 14. The numbers at the bottom of the figure refer to days post-infection. are nonviable or malformed (15,16,35). Chemical inhibitors of P4H have also been shown to have similar cuticle-specific effects in both C. elegans (14,16) and B. malayi (22). As P4H is involved in the synthesis of the nematode cuticle, it is therefore an excellent target for parasite control by chemical inhibition.
The general structure and composition of the cuticle or exoskeleton are well conserved throughout the nematode phyla (17). The cuticle is synthesized five times during development and is critical for many functions such as movement and protection from the environment. In C. elegans the structure of the cuticle has been studied in detail, demonstrating that despite overall similarity, cuticles from each developmental stage are chemically and structurally distinct (36). Following the completion of embryonic elongation the cuticle of the unhatched first larval stage maintains the elongated form of the worm (18). A weakened or malformed cuticle may cause altered body morphology in viable worms, or in some cases, as in the combined disruption of the P4H ␣ subunit encoding genes phy-1 and phy-2 in C. elegans, the resulting cuticle is too weak for survival (15,16). The cuticle is composed predominantly of collagen, and more than 150 cuticular collagen genes are present in the C. elegans genome (37). Similar large gene families are likewise present in parasitic nematodes (17), and a representative cuticular collagen, COL-2 (38), has been cloned from B. malayi. The size of B. malayi COL-2 (33 kDa) is similar to the small C. elegans collagens (26 -35 kDa) and shows overall amino acid sequence similarity to the C. elegans cuticle collagens. A comparison at the amino acid level reveals 46% identity to the C. elegans group 2 (37) collagen, T07H6.3. Precise positioning of the cysteine residues, which are required for registering and stabilizing the collagen triple helix, is likewise conserved between the B. malayi COL-2 and the C. elegans group 2 collagens. We propose that B. malayi PHY-1 may function to modify cuticular collagens such as B. malayi COL-2, a process that in C. elegans is an essential feature of development.
B. malayi PHY-1 is unusual among the metazoan P4Hs in that it is a soluble and active P4H in the absence of a PDI subunit. The function of PDI subunits in the vertebrate and D. melanogaster P4Hs, and in the C. elegans P4Hs that are involved in the synthesis of cuticle collagens, is to keep the ␣, or PHY, subunits in a catalytically active, nonaggregated conformation and to maintain the enzyme within the ER (10,11). Active soluble monomeric P4Hs have been cloned and characterized from the Paramecium bursaria Chlorella virus-1 (39), Arabidopsis thaliana (40), and partially purified from algae (41). It is also possible that the C. elegans PHY-3, which is involved in the hydroxylation of collagens in early embryos, acts as a monomeric P4H (42). Furthermore, the recently identified P4Hs that hydroxylate the hypoxia-inducible transcription factor HIF␣ are monomeric cytoplasmic enzymes (43,44). Interestingly, PHY-1 from the related parasite O. volvulus may require a PDI subunit to form an active P4H, but neither subunit contains a recognizable ER retention signal (22,45). These unusual parasite P4Hs and the unique C. elegans P4H forms suggest that nematodes have evolved specialized forms of these enzymes, possibly in order to process the large number of collagens required to form the cuticular extracellular matrix.
When recombinant B. malayi PHY-1 was coexpressed in insect cells with PDI from other organisms, no detectable association of the polypeptides was observed. In contrast, the C. elegans PHY-1 and the Drosophila and vertebrate ␣ subunits have been shown to assemble into an active P4H with PDIs from other species (5,7,13). The assembly of the recombinant C. elegans PHY-2 with PDI-2 is very inefficient, and only when it is coexpressed with C. elegans PHY-1 and PDI-2 does it assemble into a fully active, mixed tetramer (14). Thus, although the recombinant B. malayi PHY-1 can function as an active P4H by itself, it may assemble with additional, as yet unidentified assembly partners.
The B. malayi phy-1 expression constructs under the control of the C. elegans phy-1 promoter were not able to rescue the C. elegans phy-1 null (dpy-18(e364)) mutant phenotype. A similar approach, however, has been successfully applied in the rescue of the same mutant strain with a wild type copy of the endogenous gene (15) (Fig. 7A). The inability of the B. malayi PHY-1 to functionally compensate for the lack of C. elegans PHY-1 could be attributable to a number of causes such as low expression levels, differences in codon usage, missplicing of introns, or low activity of the transgenic protein. Alternatively, although the recombinant B. malayi PHY-1 was shown to be an active P4H by itself, it may have higher activity in vivo when assembled with an as yet unidentified partner. In support of this contention, the C. elegans PDI-2 does not associate with the B. malayi PHY-1 in a recombinant insect cell expression system (Fig. 4D). Thus, if additional subunits are needed for B. malayi PHY-1 to rescue dpy-18, these could not be provided by C. elegans. We plan to address this possibility following the completion of sequencing the B. malayi genome, when all possible partners can be assessed for associations with B. malayi PHY-1. Heterologous rescue of C. elegans mutants by B. malayi genes has not been described to date, but C. elegans rescue has been obtained with genes from the more closely related parasitic nematode Haemonchus contortus (46). Using a system similar to that employed in our study, the embryonic lethal phenotype resulting from the lack of a cathepsin L protease in C. elegans was rescued by transformation with a homologous gene from H. contortus (46), demonstrating the conservation of the essential developmental function of these two genes in different species. Amino acid sequence identity of mature ca- thepsin L proteases between C. elegans and H. contortus is 87%, which is significantly higher than the 59% identity found between PHY-1 polypeptides from C. elegans and B. malayi. The lower identity reflects the more distant evolutionary relationship between the latter pair of nematodes.
The B. malayi phy-1 promoter is capable of directing tissuespecific expression in C. elegans to the cuticle-synthesizing hypodermis, a location that is consistent with the proposed function of its gene product. Appropriate tissue-specific expression of reporter genes in C. elegans has been described for promoters from a number of genes from related parasitic nematodes (47)(48)(49), although to date, none have been described from B. malayi. Our present study therefore confirms that C. elegans represents an excellent heterologous system for the analysis of gene expression in the filarial nematodes. Although the native localization of PHY-1 in B. malayi is not known, its function as a P4H, taken together with data from reporter gene studies of the homologous genes from C. elegans (and the reporter gene studies reported here), would predict a hypodermal localization for B. malayi phy-1 expression, which is consistent with a role in cuticle collagen modification. The secondary staining of muscle cells and vulval cells is also identical to that found for the C. elegans homologues phy-1 and phy-2, which are both expressed in vulval cells with additional muscle cell staining found for phy-2 (15). Like the C. elegans homologues, the B. malayi phy-1 transcript was strongly expressed throughout larval development, with the greatest peaks of abundance consistent with a role in the molting cycle and cuticle synthesis.
The essential nature of P4H in development and body morphology in C. elegans has been demonstrated both by genetic and RNA interference analyses and by chemical inhibition studies. The extrapolation of the knowledge gained in C. elegans and other species along with the emerging use of C. elegans as a surrogate system for expression of foreign proteins provide powerful approaches to the further characterization of parasite P4Hs and the assessment of their potential as targets for chemical control.