A Novel Cyclophilin from Parasitic and Free-living Nematodes with a Unique Substrate- and Drug-binding Domain*

A highly diversified member of the cyclophilin family of peptidyl-prolyl cis-trans isomerases has been isolated from the human parasite Onchocerca volvulus(OvCYP-16). This 25-kDa cyclophilin shares 43–46% similarity to other filarial cyclophilins but does not belong to any of the groups previously defined in invertebrates or vertebrates. A homolog was also isolated from Caenorhabditis elegans(CeCYP-16). Both recombinant O. volvulus and C. elegans cyclophilins were found to possess an enzyme activity with similar substrate preference and insensitivity to cyclosporin A. They represent novel cyclophilins with important differences in the composition of the drug-binding site in particular, namely, a Glu124 (C. elegans) or Asp123 (O. volvulus) residue present in a critical position. Site-directed mutagenesis studies and kinetic characterization demonstrated that the single residue dictates the degree of binding to substrate and cyclosporin A.CeCYP-16::GFP-expressing lines were generated with expression in the anterior and posterior distal portions of the intestine, in all larval stages and adults. An exception was found in the dauer stage, where fluorescence was observed in both the cell bodies and processes of the ventral chord motor neurons but was absent from the intestine. These studies highlight the extensive diversification of cyclophilins in an important human parasite and a closely related model organism.

Cyclophilins belong to a large family of proteins that have been found in most organisms including parasites. It is thought that cyclophilins play an important role in protein folding because of their peptidyl-prolyl cis-trans isomerase (PPIase) 1 activity, which can be measured in vivo (1,2) and in vitro (3,4). Most cyclophilins bind the immunosuppressive drug cyclosporin A (CsA), resulting in specific inhibition of their PPIase activity (5,6). Therefore, CsA may interfere with the correct folding of proline-containing proteins that are the natural substrates for cyclophilins. It remains to be determined whether this is the mechanism by which CsA and its nonimmunosup-pressive derivatives exert lethal structural damage on a number of important parasites. For example, subimmunosuppressive levels of CsA cause gross herniation in the gut and blistering of the tegumental surface of Schistosoma mansoni (7). In the case of Litomosoides carinii microfilariae, the drug causes shrinkage of the parasite and stiffening of the surrounding sheath (8).
Most parasite cyclophilins published to date possess a high degree of similarity to human cyclophilin A (CypA) (5), an 18-kDa cytoplasmic protein that is abundantly expressed in all mammalian tissues (9). Like human CypA, the PPIases described from S. mansoni (10 -12), Toxoplasma gondii (13,14), and Plasmodium falciparum (15) increase the rate of isomerization of a standard proline-containing peptide substrate (Nsuccinyl-Ala-Ala-Pro-Phe-p-nitroanilide) in vitro, and their PPIase activity is easily inhibited by nanomolar concentrations of CsA (9, 16 -18). Thus far, only CypA homologs have been found in parasites, with the exception of the filarial worms. In addition to this highly conserved form, designated CYP-2 in filarial parasites, Brugia malayi, Onchocerca volvulus, and Dirofilaria immitis express two divergent cyclophilins that are more related to human nuclear-specific cyclophilin (CYP-3/4) and natural killer cell cyclophilin (CYP-1). These cyclophilins are considerably larger than human CypA, prefer other synthetic substrates, and display a reduced sensitivity to CsA (19 -21).
We report here the cloning, expression, and characterization of a new class of cyclophilin from the important human parasite O. volvulus  and the model organism Caenorhabditis elegans (CeCYP-16). These cyclophilins are distinct from the other cyclophilins present in the data base from C. elegans (designated CeCYP-1 through CeCYP- 15, or O. volvulus (designated OvCYP-1, OvCYP-2, OvCYP-4, OvCYP-5, and OvCYP-10). The CYP-16 cyclophilins represent novel, highly diversified cyclophilins with respect to the composition of the drug-binding site and are particularly interesting because, unlike other parasite cyclophilins described thus far, they are apparently not found in mammals. We present molecular and biochemical studies on these new enzymes and use transgenic methodologies in C. elegans to analyze developmental and spatial expression of CeCYP-16 to gain insight into the potential natural substrate(s) for these enzymes.

EXPERIMENTAL PROCEDURES
Isolation of OvCYP-16 and CeCYP-16 -All reagents, kits, and bacterial strains used in cloning, expression, and sequencing (described below) were obtained from New England Biolabs (Beverly, MA) and used as described by the manufacturer, unless otherwise specified.
A partial cDNA clone (552 bp) encoding a putative cyclophilin was isolated from an O. volvulus L3 stage Lambda Uni-ZAP XR cDNA library kindly provided by Dr. Steven Williams. The library was screened by hybridization (22) (23) were hybridized in 6ϫ SSC at 60°C overnight with the 32 P-labeled probe (1 ϫ 10 6 cpm/ml hybridization solution). After hybridization, the plaque lifts were washed twice at room temperature and once at 60°C for 30 min each in 2ϫ SSC containing 0.1% SDS. After plaque purification of nine clones, the Bluescript phagemids were excised from Lambda ZAP (Stratagene, La Jolla, CA). After sequencing, the phagemid clone with the smallest insert (552 bp; fragment B) was identified as a fragment of a putative O. volvulus cyclophilin (OvCYP-16) by using the National Center for Biotechnology Information BLAST program. This clone was isolated due to significant homology in a short stretch of sequence (39 nucleotides) encoding part of the catalytic domain of each enzyme (data not shown).
To obtain a full-length cDNA of OvCYP-16, two primers, 5Ј-gcgagtggacattcctttgacc-3Ј (antisense) and 5Ј-ccattcaatgatattgttcc-3Ј (sense), were designed using the sequence derived from the 5Ј and 3Ј ends of the partial cDNA. Two PCR products, designated fragment A (241 bp) and fragment C (322 bp), were obtained by performing thermal cycling on the O. volvulus L3 cDNA library with the following primer pairs: antisense primer/T3 primer, 5Ј-aattaaccctcactaaaggg-3Ј; and sense primer/T7 primer, 5Ј-gtaatacgctcactatagggc-3Ј. Typical PCR reactions contained 2 l of O. volvulus L3 cDNA library stock boiled for 10 min before use, 2 units of Vent polymerase, 1ϫ thermal polymerase buffer, 6 mM MgSO 4 , 0.2 mM deoxynucleotide triphosphate, and 250 nM of each primer. The reactions were heated at 95°C for 5 min, followed by 20 cycles of 95°C for 1 min, 55°C for 30 s, and 72°C for 1 min. PCR products were purified with the QIAquick PCR purification kit (Qiagen) and then subcloned into the pGEM-T vector (Promega, Madison, WI) for DNA sequence analysis. Fragments A and C were digested separately with EcoRI/ClaI and BtgI/Xhol, respectively. The 459-bp partial OvCYP-16 cDNA (fragment B) was released from the Bluescript phagemid with ClaI and BtgI. These three fragments (fragments A, B, and C) were ligated into pUC19 digested with EcoRI and SalI. The ligation was performed at 16°C overnight. Deduced amino acid sequences were aligned and compared using the Clustal method (24), and searches for homologies to other cyclophilins were performed using the National Center for Biotechnology Information BLAST program.
A genomic DNA sequence encoding the putative homolog of OvCYP-16 was identified in the C. elegans cosmid Y17G7B using the C. elegans BLAST program (Sanger Institute, Cambridge, United Kingdom). This genomic DNA sequence was used to search the C. elegans expressed sequence tag data base, and a clone (yk648d4) was found that likely represented a full-length cDNA (CeCYP-16) homolog of OvCYP- 16. Preparation and Purification of Recombinant O. volvulus CYP-16 and C. elegans CYP-16 -Thermal cycling primers were designed to enable cloning of OvCYP-16 into plasmid pMAL-c2X to generate a fusion with maltose-binding protein (MBP). The forward primer corresponded to the open reading frame of OvCYP-16 and had the sequence 5Ј-atgtcaaacgttattatcgaattcggc-3Ј, generating a 5Ј blunt end. The reverse primer, 5Ј-cccaagcttctattcaactttattgaagaccgc-3Ј, corresponded to the 3Ј end of the gene including a downstream termination codon and a HindIII recognition site. 50-l PCR reactions were carried out using 0.1 g of the pUC19-OvCYP-16 construct as template, 2 units of Vent DNA polymerase, 5 l of 10ϫ thermal polymerase buffer, 5 mM MgSO 4, 0.2 mM deoxynucleotide triphosphate, and 250 nM of each primer. The thermal cycling conditions used were 95°C for 5 min, followed by 25 cycles of 95°C for 1 min, 60°C for 30 s, and 72°C for 1 min. The reaction product was purified and digested with HindIII before ligation into pMAL-c2X digested with XmnI and HindIII. Plasmid DNA was isolated, and the insert was sequenced in both directions using the CircumVent thermal cycle dideoxy DNA sequencing kit. Production and purification of the MBP fusion protein were as described by the manufacturer.
The cDNA clone yk648d4 (GenBank TM accession number AV195981) was obtained from Dr. Yuji Kohara, and two thermal cycling primers were designed to subclone CeCYP-16 into pMAL-c2X for protein expression. The forward primer (5Ј-atgagtaatcaatatatcaacgagccg-3Ј) corresponded to the open reading frame of CeCYP-16 preceded by an ATG codon. The reverse primer (5Ј-cccaagcttctaaaccttattaaaaacggcc-3Ј) corresponded to the 3Ј end of the gene and included a downstream termination codon and a HindIII recognition site. PCR reactions (50 l) were performed as described above using 3 l of yk648d4 Lambda DNA stock as template. PCR products were purified and digested with HindIII before ligation into pMAL-c2X digested with XmnI and HindIII. The recombinant plasmid DNA was isolated, and the insert was sequenced in both directions to ensure authenticity. Production and purification of the MBP fusion protein were as described by the manufacturer.
Production of Active Site Mutants of Filarial Cyclophilin-Site-directed mutagenesis of the PPIase domain of the previously described B. malayi BmCYP-1 (19) was accomplished by the method of Kunkel (25). The histidine residue (132) of BmCYP-1 was substituted with aspartic acid using the following mutagenic primer: 5Ј-attactacaacacctgcgccagatctcaatatatccatgtggtatttgg-3Ј. The bases encoding the mutated amino acid are underlined. Mutagenesis of BmCYP-1 was verified using the CircumVent thermal cycle dideoxy DNA sequencing kit. The protocol used for the production and purification of BmCYP-1 (H132D) was as described previously (26).
Sequencing Analysis-DNA sequences were analyzed using the Genetics Computer Group (Madison, WI) software. Pairwise identity comparisons of OvCYP-16 and CeCYP-16 to other cyclophilins were performed using the program GAP. Alignment of the derived amino acid sequence of the enzyme domains of OvCYP-16, CeCYP-16, and other  Table I. Reactions were performed at 10°C and monitored at 0.3-s intervals at 400 nm using a Beckman DU 640 spectrophotometer. Pseudo-first-order rate kinetics were calculated using the following formula: k obs ϭ (k cat /K m ) [E].
To determine inhibition of enzyme activity by CsA (Sigma), recombinant enzyme (30 nM to 10 M) was preincubated for 1 h at 4°C with CsA (1 nM to 5 M), and the assay was performed as described above. Data were fitted into the following equation: k obs ϭ k obs */(1 ϩ [CsA]/ IC 50 ), where k obs * is k obs in the absence of CsA (19).
Nematode Culture-Wild-type C. elegans were obtained from the Caenorhabditis Genetics Center (St. Paul, MN). Worms were main-tained on nematode growth medium agar plates with Escherichia coli (OP50) as a food source (28).
Amplification products were subcloned using the TOPO TA Dual Promotor Cloning Kit (Invitrogen) according to the manufacturer's instructions. Restriction sites near the 5Ј ends of the primers (XbaI and HindIII) were then used to excise the insert and clone it into a multicloning site upstream of GFP in the pPD95.75 vector (29).

Sequence
The CeCYP-16 cDNA is 669 bp in length (GenBank TM accession number AF393636) and codes for a 223-amino acid protein with a predicted molecular mass of 25.2 kDa with a pI of 7.44 and contains a PPIase domain (Figs. 1 and 3).
Expression Pattern of CYP-16 in C. elegans-Transgenic lines IP102 and IP103 were obtained after coinjection of pIP10 and pRF-4 into the gonad of adult worms. Consistent GFP expression patterns were observed in these lines throughout the intestine, with particularly strong fluorescence in the anterior and posterior ends, in all the larval stages and adults (Fig. 6). An exception was found in the dauer stage, where fluorescence was observed in both the cell bodies and processes of the ventral chord motor neurons but was absent from the intestine (Fig. 6).  -1, BmCYP-2, DiCYP-3,  OvCYP-16, CeCYP-16, and mutant Bm-CYP-1 (H132D) with cyclosporin A. Appropriate synthetic substrates were used in the assays, namely, N-succinyl-Ala-Leu-Pro-Phe-p-nitroanilide for OvCYP-16 and CeCYP-16 and N-succinyl-Ala-Ala-Pro-Phe-p-nitroanilide for BmCYP-1, Bm-CYP-1 (H132D), BmCYP-2, and DiCYP-3, respectively.

DISCUSSION
Cyclophilins appear to have undergone proliferation and extensive diversification in filarial worms. In contrast to other parasites that appear to only express homologs of human CypA, the filariae possess, in addition, highly distinctive cyclophilins. The three filarial PPIases (CYP-1, CYP-2, and CYP-3) analyzed in detail thus far differ in size and display unique substrate preferences and a range of sensitivity to inhibition with CsA (21,27,32). The other cyclophilins (designated CYP-4, CYP-5, and CYP-10) identified from O. volvulus as potential vaccine and drug target candidates after immunoscreening of cDNA libraries and expressed sequence tag anal- ysis (35) belong to previously defined groups (19,21,27,32). Structure/function studies in the human and filarial parasite systems have given some insight into the reasons for the apparent differences in the enzymes. There are 13 residues that form the CsA-binding site of human CypA (33), and 1 of these residues (tryptophan 121) is essential for drug binding (34). These amino acids are highly conserved in all parasite cyclophilins described, with the exception of the filarial CYP-1 (19) and CYP-3/4 (21, 32) enzymes that have a histidine or tyrosine residue instead of tryptophan, respectively. X-ray crystallography analysis (36 -38) and site-directed mutagenesis studies (26) on B. malayi CYP-1 have demonstrated that the single H132W difference correlates with increased sensitivity to CsA.
Diversification of cyclophilins may be a feature of nematodes because the free-living nematode C. elegans possesses many different cyclophilins, including homologs of the filarial cyclophilins described above (31,32,39). This is also the case with the novel OvCYP-16 cyclophilin from O. volvulus and Ce-CYP-16 from C. elegans. The CYP-16 proteins possess a PPIase domain and an additional C-terminal domain that is rich in lysine and arginine. However, a distinctive feature of these cyclophilins is that they belong to an independent group and, based on a data base search, do not appear to be present in mammals.
In the OvCYP-16 and CeCYP-16 cyclophilins, 9 and 11 of the 13 residues are conserved, and, unlike any other cyclophilin published to date, a Glu 124 (C. elegans) or Asp 123 (O. volvulus) residue is present at the critical position (tryptophan) in the drug-binding site. Because previous studies have shown that filarial parasites possess CsA-insensitive (CYP-1 and CYP-3) (19,21) and CsA-sensitive (CYP-2) cyclophilins (27), similar experiments were performed using OvCYP-16 and CeCYP-16 fusion proteins. At the highest concentrations of CsA tested, we were unable to detect any inhibition of enzyme activity. These enzymes are therefore considerably more resistant to CsA inhibition than CYP-1 or CYP-2 and are more similar to the more recently described filarial CYP-3 enzymes (21).
It has been suggested that the various isoforms of filarial cyclophilins may be involved in the folding of different proteins in vivo (40). Therefore, we compared the ability of OvCYP-16 and CeCYP-16 to catalyze the isomerization to the trans form of 13 different synthetic peptides of the general structure Nsuccinyl-Ala-Xaa-cis-Pro-Phe-p-nitroanilide. The previously characterized filarial enzymes CYP-1, CYP-2, and CYP-3 were also included. The catalytic efficiency (k cat /K m ) of the substrates varied, and a distinct profile emerged for each cyclophilin. CYP-1, CYP-2, and CYP-3 favored peptides containing the short chain residues alanine (found in the standard substrate) and glycine. In contrast, OvCYP-16 and CeCYP-16 did not demonstrate any activity with these particular substrates but instead were found to prefer hydrophobic, acidic, or amide amino acids. These results support the notion that the various members of the cyclophilin family could have distinct natural substrates. Amazingly, a single amino acid substitution in the mutated BmCYP-1 (H132D) resulted in an enzyme with the characteristics of OvCYP-16 and CeCYP-16 with regard to substrate preference, enzyme activity (decreased ϳ40-fold compared with wild-type), and drug sensitivity. This emphasizes how a single amino acid in this position dictates the behavior of the enzyme. We have also found that mutation of CeCYP-16 (E124W) significantly increased the activity of the enzyme fused to MBP (data not shown). This result indicates that the low activity associated with the CYP-16 enzymes is probably not due to the presence of MBP. Consistent with this is the fact that a mixture containing MBP cleaved from the fusion protein and CeCYP-16 also has a low enzyme activity (data not shown).
It is known from the extensive work done on mammalian cyclophilins that the various isoforms can vary in their expression pattern (41). To gain information on the function of filarial CYP-16, properties of its homolog were studied in the genetic model C. elegans. Developmental and spatial expression patterns of CeCYP-16 were determined using a GFP reporter system. It was found that GFP expression under the control of the CeCYP-16 promoter was strong in the distal portions of the intestine. This pattern was seen throughout development, with the exception of the dauer larva, in which expression was seen only in the ventral chord motor neurons. Therefore, CeCYP-16 would seem to be functioning in distinct cell types during regular development as compared with the dauer stage. During the dauer stage, the intestine is closed off to the environment, and the animal does not feed (42). Therefore, perhaps a protein functioning in the intestine would no longer be needed during this stage, resulting in a loss of expression in the gut. Of greater interest is the apparent recruitment of this protein in completely unrelated cells during the dauer stage. Because dauer animals undergo less movement compared with other developmental stages, one might speculate that a dauer-specific function for CeCYP-16 in the motor neurons might be of an inhibitory nature. It would be interesting to determine whether a parallel expression pattern is seen in filarial parasites, where the dauer stage is represented by the infective stage larva that remains in the insect vector until transmitted to the mammalian host.
The expression pattern of CeCYP-16 resembles that of Ce-CYP-8, another cyclophilin with a homolog in filarial parasites (BmCYP-1). CeCYP-8 was found specifically in gut cells in various stages (39). However, this study did not include the dauer stage, so it remains to be seen whether there is complete overlap in the expression patterns of these two proteins. A completely different pattern was observed for CeCYP-4, a muscle-specific cyclophilin that has been shown to be essential for normal muscle development in early larvae. RNA interference experiments of the CeCYP-4 resulted in progeny with a lumpy appearance (31). Similar experiments performed using Ce-CYP-8 (39) or CeCYP-16 (data not shown) did not show any obvious phenotype, perhaps indicating that they perform the same function in the gut, although this is unlikely due to the distinct substrate preference profile observed for CeCYP-16 and BmCYP-1 (a homolog of CeCYP-8). It is also noteworthy that RNA interference is ineffective against neuronal expressed products (43), so other approaches would be necessary to address the importance of CeCYP-16 in this tissue.
Additional studies on both conserved and divergent cyclophilins in nematodes may give insight into the function of the various PPIases and reveal potential drug targets for antifilarial chemotherapy.