THE

JOU&'1AL

OF BrOLOGICAL

Vol. 270, No. 31, Issue of August 4, pp. 18517-18522, 1995

Prrued in U.S.A.

CHEMISTRY

1995 by The American Society for Biochemistry and Molecular Biology, Inc.

An Abundant, trans-spliced mRNA from Toxocara canis Infective

Larvae Encodes a 26-kDa Protein with Homology to
Phosphatidylethanolamine-binding
Proteins*
(Received for publication, February 2, 1995, and in revised forrn, May 15, 1995)

David Gemst, Cheryl J. Ferguson, Brian D. Robertsonsj, Ren Nieves], Antony P. Page**,
Mark L. BlaxterH, and Rick M. MaizelsH
From the Wellcome Research Centre for Parasitic Infections, Departmeni of Biology, Imperial College of Science,
Technology and Medicine, Prince Consort Road, London SW7 2BB, United Kingdom

A full-Iength mRNAencoding a secreted 26-kDaantigen of infective larvae of the ascarid nematode parasite
Toxocara canis has been identified. This was characterized as a 1,082-basepair clone highly abundant (0.81.9%) in cDNAprepared from infective stage larvae but
absent from cDNA from adult male worms. Sequence
analysis revealed an open reading frame corresponding
to a hydrophilic 263-amino acid residue polypeptide
with a 20-residue N-terminal signal peptide, indicating
that it is secreted. The 5' end of the cDNAwas isolated
by polymerase chain reaction using a primer containing
the nematode-splced leader sequence, SL1, showing
that the mRNAis trans-spliced. The molecular mass of
the putative protein with the signal peptide removed is
26.01kDa, and antibody to the recombinant protein expressed in bacterial vectors reacts with a similarly sized
protein in T. canis excretory/secretory (TES)products.
An identical sequence was obtained from a genomic
clone isolated by expression screening with mouse antibody to TES.The 72 amino acid residues adjacent to the
signal peptide form two homologous 36-residue motifs
containing 6 cysteine residues; this motifis found also in
the T.canis-secreted glycoprotein TES-120and in genes
of Caenorhabditis elegans. Sequence data base searches
revealed significant similarity to 7 other sequences in a
newly recognized gene family of phosphatidylethanolamine-bindng proteins that includes yeast, Drosophila,
rat, bovine, simian, and human genes and a representative from the filaral nematode Onchocerca volvulus. As-

* This investigation received financial support from a Leverhulme

Trust grant, from the Medical Research Council, and from the Wellcome
Trust through the Wellcome Research Centre for Parasitic Infections.
The costs of publication of this article were defrayed in part by the
payment ofpage charges. This article must therefore be hereby marked
"adoertisement" in accordance with 18 U.S.C. Section 1734 solely to
indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to
the GenBank I EMBL Data Bank with accession number(s) U29761.
:j: Present address: Molecular Biology Program, 311 Tucker Hall,
University of Missouri, Columbia, MO 65211.
Supported by Medical Research Council postgraduate studentships.
~ Present address: Dept. ofMedical Microbiology,StoMary's Hospital
Medical School, London W2 IPG, UK.
11 Supported by Minority International Research Training Grant T37TW00046from the Fogarty International Center at the National Institutes of Health. Present address: Dept. of Biochemistry, Medical Sciences Campus, University of Puerto Rico, San Juan, PR 00936.
** Recipient of a WeJlcome Prize Studentship. Present address:
Wellcome Unit for Molecular Parasitology, University of Glasgow, 56
Dumbarton Road, Glasgow GIl 6NU, UK.
:j::j: Present Address: Inst. of Cell, Animal and Population Biology,
Ashworth Laboratories, King's Buildings, University of Edinburgh,
West Mains Rd., Edinburgh EH9 3JT, UK.
To whom correspondence should be addressed. Te!.: 44-131-6505511; Fax: 44-131-650-5450;E-mail: R.Maizels@ed.ac.uk.

says with the T. canis recombinant 26-kDa protein expressed as a fusion with maltose-bndng protein have
confirmed phosphatidylethanolamine-binding specificity for this novel producto

Toxocara canis is a parasitic ascarid nematode that infects

large numbers of dogs worldwide. Adult worms infest the gut
and lay large numbers of eggs that, on excretion, embryonate
and le dormant in the soil. Upon ingestion by non-canid
paratenic hosts, tissue-invasive infective stage larvae are released (1). In humans these cause visceral and ocular disease
syndromes (2-4). Infective larvae have been shown to persist
up to 9 years (5), demonstrating their remarkable facility for
evading host immune responses.
T. canis infective larvae are developmentally arrested, and
large numbers may easiJy be cultured in vitro for long periods
(6). For this reason this organism has been used as a laboratory
model for characterizing the nematode surface in order to understand the mechanisms of escape from immune attack (7).
Cultured larvae release five major glycoprotein excretory/secretory (TES)1 antigens (8-10), which are well characterized
biochemically C8,10-14), as well as a number ofminor components that have been lttle studied. Infective larvae al so possess a glycocalyx-Iike surface coat that overles the cuticle,
comprised largely of a 120-kDa mucin-Iike glycoprotein, TES120 (15, 16).2 Monoclonal antibodies reactive to TES-120 bind
to secretory glands within the body, from which it is thought
that coat components are ducted to the surface via the buccal
cavity and secretory pore (18).
The mRNA for TES-120 is hyperabundant,
representing approximately 10% of a cDNA library, and is trans-spliced with
the nematode 22-nucleotide spliced leader (SL) sequence.?
cDNA amplified from mRNA by PCR using SL and oligord'I')
primers reveals the TES-120 hyperabundant
band, together
with a less abundant mRNA species. We report here the isolation of cDNA corresponding to this second SL-bearing message
CTcSL-2), demonstrate its stage-specific expression in infective
larvae as the secreted antigen TES-26, and describe the primary structure of the protein. Comparisons to previously
known sequences reveal that TcSL-2 belongs to a recently
identified family of genes from a range of organisms, sorne of
which bind phosphatidylethanolamine
with high specificity.

1 The abbreviations used are: TES, T. canis excretory/secretory product; SL, spliced leader; PCR, polyrnerase chain reaction; MBP, maltosebinding protein; PEBP, phosphatidylethanolamine-binding protein; PE,
phosphatidylethanolamine; kb, kilobase(s); bp, base pair(s).
2 Gems, D. H., and Maizels, R. M., submitted for publication.

18517

18518

26-kDa Antigen from Toxocara canis

MATERIALSAND METHODS
Parasites-Live adult T. canis were obtained from the intestinal
tract from post-mortem dogs. Male worms were snap frozen and stored
at -70 "C prior to use. Eggs were obtained from uteri excised from
gravid females. Hatching of eggs and in vitro culture of infective larvae
was as described previously (19).
Preparation of RNA and DNA-Approximately 1,000 infective larvae
that had been in culture for at least 3 days were suspended in a drop of
RPMI 1640 medium and snap frozen in liquid nitrogen. This was
ground to a fine powder in a pre-cooled mortal' with a pestle, and RNA
was extracted using the acid guanidinium thiocyanate method (20) and
resuspended in 10 .tI ofH"O. DNA was prepared from 10 adult males by
grmdmg to a fine powder under liquid nitrogen with a pestle and
mortar and resuspending in 5 mI oflysis buffer (100 mMNaCI 100 mM
EDTA, and 10 mMTrislHCI, pH 7.5, supplemented with 1% SOS and 1
mg/ml proteinase K). This was incubated at 50 "C for 3 h with gentle
agitation, then extraeted with an equal volume ofwarm phenol (37 OC),
and spun at 20,000 X g for 10 min at 20 "C. After two extractions with
phenol/chloroform and one with chloroforrn, RNase was added to the
aqueous phase to a concentration of 250 p,g/ml, and the sample was
incubated for 20 min at room temperature with gentle agitation. After
another chloroform extraction DNA was ethanol precipitated from the
aqueous phase.
Genomic DNA Library Construction-Aliquots of 5 fJ-gof T. canis
DNA were partially digested using RsaI, HaeIlI, andAlul. These were
pooled and protected from EcoRI restriction using EcoRI methylase.
EcoRI linkers were then Iigated on and digested withEcoRI. The inserts
were then ligated into the EcoRI site of the express ion vector Agtll.
After paekaging (Arnersham paekaging extract) a titer of 106 plaqueforming units/ml was obtained, with 98% of clones containing genomic
DNA inserts.
Antibody Screening of Expression Library-Anti- TES polyclonal sera
were raised in CBA/Ca and BALB/cmiee. These sera were found to bind
to all components ofTES in Western blots. For screening, sera were first
preabsorbed against sonicated Escherichia coli and then used at a
dilution of 1:25.
cDNA Preparation-To prepare cDNA, 1 fJ-1
oftotal RNA solution was
used with a GeneAmp RNA PCR kit (Perkin-Elmer) with Pfu polymerase (Stratagene Ltd., United Kngdom) substituting for Amplitaq in the
PCR. RNA PCR has been deseribed previously (21, 22). The cDNA was
amplified at 94 "C for 1 min, 60 "C for 3 rnin, and 72 "C for 15 min for
35 cycles. For first strand cDNA synthesis, primer DGDT containing
a 5' BamHI restriction site (5'-AATTCGGATCCCCCGGG(T)18-3')
was used. Three seeond primers were added for PCR: DGSL1 (5'GGGCGGCCGCGGTTCAATTACCCAAGTTGGAG-3'),DGSL2A (5'GGGCGGCCGCGGGTTTAATTACCCAAGTTGGAG-3'),and DGSL3
(5' -GGGCGGCCGCGGGGTTT AATTACCCAAGTTTGAG-3'). These
primers contained a Notl site positioned so that eaeh primer would
result in PCR products being in a different frame relative to the lacZ
gene after ligation into the plasmid pBluescript SK(+). Sequence
changes were introduced into the SL1 sequence ofDGSL1 and DGSL2A
to rernove termination eodons.
PCR of TcSL-2 cDNA-Total cDNA prepared as described aboye
was used as a template. Two primer pairs were used: DG9a (5'-GCGCGGCGGCCGCGTCAATTAACCTCTGCCAGAAC-3'),containing 21
bases ofthe sense genomic sequence determined for clone TcG9 and a 5'
Notl site with the oligo(dT) primer DGDT, and DG9c (5'-GCGCGGCGGCCGCGTTCTGGCAGAGGTTAATTGA-3'),containing 21 bases of
the antisense TcG9 sequenee and a 5' NotI site with DGSL-2A.Amplification was at 94 "C for 1 min, 60 "C for 2 min, and 72 "C for 3 min for
35 cycles.
Southern Blof Analysis-DNA was electrophoresed on 0.8% agarose
gels and transferred to Hybond-N membrane (Amersham Corp.). DNA
probes were labeled with digoxygenin by random priming and detected
after hybridization using a digoxygenin DNA Labeling and Detection
kit (Boehringer Mannheim). The chemiluminescent substrate AMPPD
was used as a substrate for the anti-digoxygenin hapten-conjugated
alkaline phosphatase. Hybridization and washing was carried out at
high stringency as described in the Boehringer kit protocol.
DNA Sequencing-The nucleotide sequences of the cDNAs were determined using the Sanger dideoxy chain termination method with
double-stranded DNA (23) using "'S-dATP and a T7 sequencing kit
(Pharmacia Biotech Inc.). The sequences of both strands were determined from two copies of each of the two portions of the TcSL-2 cDNA.
Where sequence differences were seen a third copy was sequenced
across the region in question.
Expression of TcSL-2 (TES26) Protein-Two expression systems

were used to prepare TcSL-2 (TES-26) protein by inserting a PCRamplified fragment extending from Gln-21 to the C-terminal Ala-263
thereby omitting the putative signal peptide. For pET15b (Novagen):
the 5' pnmer contained a restriction site for Ndel (AGTGGACATATGCAACAGTGTATGGACAGT),and the 3' primer contained a BamHI
site (TACGGATCCTTAGGCCTGCGATCGATAGAA).Cloning of this
insert into a site under control of the T7 promoter was accomplished in
E. coli TB-1 cells, which lack the T7 RNA polymerase gene; expression
was achieved in BL21 cells carrying the DE3 bacteriophage, which
carries T7. The recombinant TcSL-2 was found in the soluble fraction of
a bacterial Iysate and was purified on a NiSO.-charged chelating
Sepharose column by elution at pH 3.8. A second expression system,
pMAL (New England BioLabs Inc.) produced a form of TcSL-2, fused
with maltose-binding protein (MBP), which performed better in enzyme-Iinked mmunosorbent assays (see below). For this vector, the
same 3' PCR primer contaning a BamHI site was combined with a 5'
primer utilizing a BglI restriction sequence.
Immunological Assays-Western blot analysis was performed on
TES antigens as described previously (18). Briefly, 5 fJ-gof TES/track
was electrophoresed on a 5-25% polyacrylamide gradient gel, transferred onto nitrocellulose paper, blocked in 5% fetal calf serum and
incubated for 2 h with 1/400 dilution of sera from normal or TES- or
TcSL-2-immunized mice. Blots were then probed with 1/2500 alkaline
phosphatase-conjugated anti-mouse Ig and finally incubated in substrate, nitro blue tetrazolium/5-bromo-4-chloro-3-indolyl phosphate
(Sigma). For enzyme-linked immunosorbent assays, MBP-TES-26 or
control MBP from unrecombined pMAL were coated onto plates at 1
ug/ml overnight, and sera were tested at 1/200, incubating for 2 h at
37 "C. Wells subsequently received peroxidase-conjugated anti-human
IgG, followed by 2,2'-di-(3-ethylbenzthiazoline sulfonic acid) substrate
(Krkegaard and Perry Laboratories, Inc.). Absorbances for each serum
on MBP and MBP-TES-26 were read at 405 nm, and the net reactivity
to TES-26 was calculated by subtraction.
Lipid Binding Assay-A binding assay for phospholipids was
adapted from that described for other members ofthe PEBP family (24,
25). Lipidex-1000 beads (Packard catalog number 6008301) were first
washed in 10 volumes of 10 mM potassium phosphate buffer, pH 7.4,
and then poured into 2-ml columns in siliconized glass. Radiolabeled
[14C]phosphatidylethanolamine (DuPont NEN catalog number NEC783) was mixed with either recombinant 26-kDa MBP fusion protein or
MBP alone; all mixtures contained 500 pmol of PE and 0-10 fJ-gof
protein in a total volume of300 fJ-lof10 mMpotassium phosphate buffer,
pH 7. Samples were incubated in siliconized glass for 15 min at 37 "C
and then cooled on ice before loading onto parallel Lipidex columns.
Columns were washed with 5 ml of phosphate buffer, and the flowthrough fraction was collected; this fraction contains lipids sequestered
by proteins. Free lipids bound to the column were then eluted with 5 mi
of methanol at room temperature. Aliquots from each fraction were
assayed by liquid scintillation counting, and the proportion of proteinbound PE was calculated as a percentage of total counts recovered.
Additional mixtures were set up in which proteins were preincubated
with 5 nmol ofunlabeled PE or phosphatidylcholine for 15 min at 37 "C.
Computer Analysis of Sequences-Nucleic acid sequences were anaIyzed using MacMolly (Softgene) and the University of Wisconsin Sequence Analysis (UWGCG) package at the SEQNET facility (Daresbury, UK). Data base searches for amino acid sequence similarities
were performed using the Lipman and Pearson algorithm (26) FASTA
on the SwissProt data bank, TFASTAon a translated data base ofDNA
sequences (GenEMBL), and by the BLAST E-mail server (27) to search
a collection of data bases including GenBank TM, SwissProt, and dbest.
RESULTS

Preparation of cDNA from T. canis Adult Males and Infectiue

Laruae-cDNA was synthesized from unfractionated RNAprepared from T. canis infective larvae and from adult males. For
first strand cDNAsynthesis a poly.A)"-complementary primer
carrying a 5' EcoRI restriction enzyme site (DGDT)was used.
A second primer containing the conserved nematode 5' spliced
leader (SL1) sequence and a 5' NotI site was added, and PCR
was carried out. It has been demonstrated in the closelyrelated
nematode Ascaris lumbricoides that most mRNAspecies carry
a 5' SLl sequence (28); the occurrence of spliced leaders in T.
canis was previously unknown. The cDNA preparations were
separated by agarose gel electrophoresis (Fig. lA). The cDNA
profilesfrom adult male and infectivelarvae were similar apart

18519

26-kDa Antigen from Toxocara canis

2

Size

(kb)

1.0

0.85
0.51
0.73 kb

0.23

FIG. l. RNA PCR reveals hyperabundant

stage-specific transcripts among total SL + cDNA. A, ethidium brornide-stained agarose
gel with cDNAprepared by RNAPCR using as template unfractionated
RNA from infective larvae (lane 2) and adult males (lane 3). Prirners
used were complementary to the 5' splieed leader (SLl) and the 3'
poly(A)+tail. Lane 1 eontains l-kb ladder size standards (Life Teehnologies Ine.). B, Southern blot of the same gel probed with digoxygeninlabeled DNA from a genomie exeretory/seeretory antigen clone, TcG9.

from two distinct bands visible only in infective larval cDNA.

The upper band was approximately 1.1 kb in size; analysis of
the lower 0.73-kb band is described elsewhere.f
Hybridization of TcG9 to cDNA and genomic DNA-In a
parallel experiment, a 1.3-kb genomic Agt11 clone, TcG9, was
isolated by immunoreactivity to an anti-TES polyclonal serum
raised in mice. To determine if mRNA from this antigen was
represented in the SLcDNA, TcG9 was used to probe a Southern blot of adult and infective larva-derived cDNA (Fig. lB). It
was found that hybridization occurred to a single band approximately 1.1 kb in size in the infective larval cDNA. No hybridization to the adult male cDNA was observed. Digoxygeninlabeled TcG9 was hybridized to a Southern blot of T. canis
genomic DNA digested with EcoRI at high stringency. A single
band of approximately 3 kb was seen (data not shown), indicating the presence of a single copy of the gene.
PCR Amplification of the 1.1-kb cDNA Using TcG9-complementary Primers-273 bp ofthe 5' end ofTcG9 were sequenced,
and upstream (DG9a) and downstream (DG9c) primers were
designed from the 5' 21 bp. PCR was carried out on template
cDNA from adults or infective larvae. Using infective larval
cDNA and DG9a with DGDT, an 850-bp product was obtained
(Fig. 2). This was not obtained when adult cDNA was used as
a template. The up tream primer (DG9c) was used in PCRs
with three SL1-complementary primers. When infective larval
cDNA was used as a template, a 235-bp product was obtained
in each case (Fig. 2). Where adult cDNAtemplate was used, no
such product was seen. Both fragments were ligated into the
plasmid pBluescript SK(+). A plasmid containing the 850-bp
fragment was digoxygenin-labeled and used to probe TcG9 on a
Southern blot, and hybridization occurred (data not shown).
These results confirm the correspondence of the cloned 1.1-kb
cDNA with the TcG9 genomic sequen ce and indicate that expression ofthe gene from which the 1.1-kb mRNA is derived is
stage-specific.
DNA Sequence Analysis-Using
double-stranded sequencing
methods two copies of each of the 5' and 3' PCR products were
sequenced along the fulllength ofboth strands (Fig. 3). A group
of overlapping clones was assembled of 1,056 bp, including SLl
and excluding the 3' polytA)" tail. The sequence contained a
single open reading frame encoding a protein of 263 amino acid
residues. The start codon was located 9 bases downstream of

FIG.2. Reverse transcriptase PCR amplification of cDNA using primers designed from TcG9 sequence. Lane 1, size standard
(l-kb ladder, Life Technologies, Inc.); lanes 2 and 3, reverse transcriptase PCR using as template cDNAprepared from infective larvae
and as primers DGDT and DGK9a (lane 2) and DGSLl and DGKge
(lane 3); lanes 4 and 5, reverse transcriptase PCR using adult cD A
with the primers DGDT and DGK9a (lane 4) and DGSLl and DGK9c
(lane 5).

the end of the spliced leader sequence. Analysi based on a

weight matrix consensus (29) indicated that the first 20 residues comprised a signal peptide for targeting secretion. The
predicted protein molecular mass was 28,034, but it was 26,007
after removal ofthe signal peptide with a pl of 8.5 in each case.
The most abundant residues in TcSL-2 are alanine (11.0%),
asparagine (8.8%), valine (8.8%), threonine (7.6%), and serine
(7.2%).
A hydropathy plot shows the protein to be largely hydrophilic
apart from the N-terminal signal peptide and a 20-amino acid
region at residues 94-113 (data not shown). The presence of
charged residues on either side ofthis region suggests that this
may be a transmembrane domain, although no direct evidence
for membrane association exists for this protein. The N-terminal domain also contains an NPT triplet at amino acids 78-80,
but the presence of proline generally prohibits N-glycosylation
because the asparagine and hydroxyamino acid side chains
cannot be juxtaposed.
Antibody Reactivity of Recombinant and Native ProteinsRecombinant TcSL-2 was expressed as a full-length protein
from Gln-21 to the C-terminal Ala-263 in two vectors, pET15b
and pMAL. The former was used to prepare recombinant protein for immunization of mice, and sera collected after secondary challenge were used to probe a Western blot of TES antigens. The TcSL-2-specific antiserum reacted with a 26-kDa
band distinct from the major TES-32 glycoprotein, which was
designated TES-26 (Fig. 4). The 26-kDa antigen showed no
shift in mobility when anti-TcSL2ffES-26 was used to probe
N-glycanase-treated TES (data not shown). In testing for naturally elicited antibodies in human infection with T. canis, it
was found that the pMAL-expressed protein fused to maltosebinding protein performed better than the pET producto Sera
from a total of 118 patients with toxocariasis (as confirmed by
A >0.5 by enzyme-linked immunosorbent assay on TES) were
screened, together with 38 low titer Toxocara-reactive sera and
8 normal controls. Only 18 individuals (11.5%)were seropositive against TES-26, but one normal control was also reactive
(data not shown). The relatively low frequency and nonspecific
reactivity to TES-26 excludes it from consideration as a diagnostic antigen. The same antibodies have been tested in immunofluorescence assays in which strong staining of the anterior
surface of the larvae was observed (data not shown). Because
the nematode cuticle and surface coats are extracellular structures, this pattern supports the contention that TES-26 follows
a secretory pathway.

26-kDa Antigen from. Toxocara canis

18520

Protein:
M
S
V
V
H
K
A
e L I A L L F
DNA, GGTTTCATCATGTCAGTTGTACACAAAGCTTGCTTAATAGCTTTACTGTTT

Ant.
NMS TcSL2

Ant.
TES

v s S G V A Q Q le M o S A S D e A A N
GTGTCGAGTGGAGTGGCACAACAGTGTATGGACAGTGCGTCAGACTGTGCGGCAAAC
A

R e

GCTGGGTCCTGTTTCACACGACCAGTCAGCCAAGTTCTGCAGAACAGGTGTCAAAGG
T C N T C D)~
R D E A N N C A A S I N
ACATGCAATACATGCGACTGCCGGGACGAGGCTAACAATTGTGCAGCTTCAATTAAC
L

CTCTGCCAGAACCCCACATTTGAACCTTTAGTTCGTGATCGATGCCAAAAGACTTGT

G L c) A G C G F I S S G I V P L V V T
GGTTTGTGTGCGGGGTGCGGATTCATCAGTAGCGGAATCGTTCCATTAGTTGTGACG

'.-*100

v s v T F A N N V Q v N e
AGTGCACCTTCAAGACGCGTGAGTGTGACCTTCGCTAACAACGTTCAGGTGAACTGC
G N T L T T A Q v A N Q P T V T W E A
GGTAATACTTTGACGACGGCGCAAGTAGCCAACCAACCGACAGTAACGTGGGAAGCG
Q

CAACCGAATGACAGGTACACCCTCATCATGGTCGATCCTGACTTCCCAAGCGCGGCG
N G Q Q G Q R L H W W V I N I P G N N
AACGGCCAGCAGGGTCAGCGACTTCACTGGTGGGTAATAAACATTCCAGGAAATAAT
lAG
G T T L A A F Q P S T P A A N T
ATCGCCGGTGGTACGACGCTCGCAGCGTTCCAGCCATCAACACCCGCCGCTAATACT
G

Rly

ylR

GGTGTGCACCGATACGTGTTTCTGGTATACAGACAGCCTGCCGCAATAAACTCTCCG
L L N N L V v Q D S E R P G F G T T A
CTACTTAATAATTTGGTAGTGCAGGATTCGGAGCGGCCAGGTTTTGGTACGACTGCC
F

TTTGCCACTCAGTTTAACCTTGGATCACCCTATGCTGGCAATTTCTATCGATCGCAG
A

GCCTAATAGTCGGTACTTCCGATGTGCAAATAACTCCGCGAGAAACTAGTACAACCA
CAACTGCTGTGAATACTGGACTCCATAATGATTTAGACCTGATTTCCAGAATTCCAG
TTTCCTCGAAATTCTATTGTTTTCCCGAAGTCCGTAGATTGCACAACCGTTTCTGTT
TTCTTTGCATTGCAGATGCCTCCCAGACTTGGAGTTCTTATCAATTTGCTAfl~TAAT
AAACAAGTTGTTGAk~~AAAAAAAAAAAAAAA
FIG. 3. DNA and protein sequence of abundant l.I-kb infective
larval cDNA- The amino acid residues of the putative N-terminal
hydrophobic secretory signal peptide are shown in italics. The 36residue NC6 motif repeats are highlighted by open boxes. The potential
transmembrane sequence is marked with salid boxes, and the flanking
pairs ofcharged residues are circled. The putative phosphatidylethanolamine-binding site identified in rat and bovine homologuesis shown in
the shaded box.

Sequence Data Base Searches-Seven

protein sequences
showing a high degree of similarity to the 1.1-kb cDNA were
identified by data base searching: the filarial nematode Onehoeerca uoluulus Ag16 antigen (30, 31), the 21-kDa bovine
brain cytosol PEBP (24, 32, 33), the 23-kDa rat sperm plasma
membrane PEBP (25), identical to the 23-kDa morphine binding protein found in brain and liver cytosol (34), a macaque
epididymal homologue (35), a homologous sequence from human HepG3 cDNA (36), an antennal cDNA from Drosophila
(37), and the Saceharomyees eereuisiae 24-kDa TFSl protein
(38). Similarities within this gene family are summarized in
Fig. 5. The Ag16 cDNAsequence from O. uoluulus (30) shows a
frame shift due to a l-bp deletion relative to the other members
of the family at nueleotide 454 (residue 136, Fig. 5). A strong
residual similarity can be found between the -1 frame and the
rest of the family almost to the end of the protein (37), indicating that this deletion is likely to be a recent evento
The sequence alignment reveals several blocks of highly
conserved sequence, especially between residues 240 and 253.

FIG. 4. Western blot reactivity of TES antigens with normal

mouse serum (NMS), antibody to pET-expressed TcSL-2 protein
(Anti-TcSL2),
and antibody to TES (Anti-TES)_ The major TES
glycoproteins, TES-32 and TES-120, are indicated, together with the
TES-26 band reactive with anti-TcSL-2 antibodies.

The run ofhydrophobic residues YV(WfF)LVYat 245-250 (Fig.

3) was previously noted as a putative PE-binding sequence (33)
and has a similar composition to a known binding site for
phosphatidylcholine, VFMYYF (39). In al! seven homologues
this hydrophobic stretch is bracketed by pairs of charged residues (H,R and EIKIR,Q),possibly reflecting hydrophilic boundaries of a short hydrophobic functional region.
Both the 2l-kDa bovine protein (33) and the 23-kDa rat
protein (25) have been shown to bind PE. For this reason the
1.1-kb cDNA putative protein was compared with three other
families of lipid-associated proteins: (a) the 14-16-kDa fatty
acid-binding proteins, which inelude surface proteins of Sehistosoma manso ni (40) and Faseiola hepatiea (41); (b) the mammalian lipid-binding proteins, e.g. human adipocyte lipid-binding protein (42) and rat nonspecific lipid transfer protein (43);
and (e) the plant phospholipid transfer proteins, e.g. barley
phospholipid transfer protein (44). No significant similarity
was detected between the PEBP family and the other three
lipid associated protein families (data not shown). A short
stretch of similarity (7 out of 8 residues identical) was also
found between residues 54-61 ofTcSL-2 and residues 107-114
of von Willebrand's factor (45), a mammalian serum protein
involved in platelet adhesion and blood coagulation (Fig. 5).
TES-26 ls a Funetional Phosphatidylethanolamine-binding
Protein-To test whether the sequence homology reflected a
functional conservation, recombinant TES-26, expressed as a
fusion protein with MBP, was tested in a Lipidex bead assay for
binding to radioactive PE. These beads bind free lipid but not
lipid sequestered by proteins, and this assay has demonstrated
PE binding by other TES-26 homologues (24, 25). Fig. 6 shows
that TES-26 but not MBP binds directly to labeled PE. This
binding can be largely inhibited by a 10-fold molar excess of
cold PE but not by phosphatidylcholine.
TES-26 Contains Two 36-Residue 6-Cysteine Partial Repeats-Amino acid residues 23-94 comprise partial repeats of a
cysteine-rich 36-residue consensus sequence, designated the
NC6 motif (Fig. 7). This contains the consensus sequence
XCXDXAX2CX6CXllRCX2TCX2C previously identified at the
C terminus of another major secreted glycoprotein of T. eanis
infective larvae, TES-120, and this motif is also found in several predicted genes from Caenorhabditis elegons? Although
the 5' and 3' NC6 motifs ofTES-26 are homologous,there is a
greater similarity between these sequences and the corresponding motifs in TES-120, indicating that the duplication to
create a 72-amino acid domain was an ancient event (Fig, 7).
Abundanee of 1.1-kb cDNA-A cDNAlibrary prepared from

18521

26-kDa Antigen (rom Toxocara canis

50
TES-26
Oncho
Drome
Human
Macaca

MSVVHKACLI

ALLFVSSGVA

QQCMDSASDC

AANAGSCFTR

45

PVSQVLQNRC
.<3

"

40

35

Bovine

"O

Rat
Yeast

30

'""".5.

25

51

100

TES-26
Oncho

NLCQNPTFEP
.MHCLQVVIA IVLYSFGKIS
.MK LPALHLLFLG

Dr-cme

LVRDRCQKTC
AENANCKKCT
YICLARSQDN

GLCAGCGFIS
Pl1LVDSAFKE
DENVRRIMKE
.MPVDLSKWS
.MPVDLSKWS
.PVDLSKWS
.MAADISQWA
.MNQAIDF AQASIDSYKK

Human
Macaca
Bovine
Rat
Yeast

::

~o
'"

r
t

Q..

20
15
10
5

Drome
Human

Macaca
Bovne
Rat
Yeast

101
SGIVPLVrS
HGIVPDVIsT
MEVIPEILDE

... KLVN IrFANNVQVN

~YNNLTV.N
..IRRVS
...p. ELLR IKYDNTIDIE

L
IIAm
E T

GPLSLQEIE
GPLSLQE DE
GPLSLQE
E
GPLSLQE
E
HGILEDVIHD

...
...
...
...P
TSFQ

L
L
L
L

ug Protein Added to Lipid

HPLH IYAGAAVDE
HPLH
YAGAALDE
HPLQ
YGGAEVDE
HALR
YGGVTVDE
GILA
YSSSAPVA

TSI ANIQ
KNQ .KVS
.~J50
TEL KFQ ..RL
TI
T'
PT
T
PTEKA

KNR .SIS
KN
.SIS
KNR .SI
MN
.SIS
RSK FQFTF

151
TES-26

Rat
Yeast

E.........
...A.QPND
D.........
...A.EPGA
N
A.DPES
D. .... .... .
GLDSGK
D. ....
..GLDSGK
D. ....
..GLDPGK
D..
....GLDPGK
NKQMQKSVPQ ANAYVPQDDD

TES-26
Oncho
Drome
Human
Macaca

201
P
S
P
K
K

Oncho
Drome
Human
Macaca

Bovine

GQQG

~UI

FIG.6. Phosphatidylethanolamine
binding by recombinant
TES26. Lipidex-l.O beads were used to seprate free r14C1PE from
lipid bound to TES-26MBP fusion protein (e) or control MBP (L.).
Positive binding of [l4CIPE to TES-26-MBP is inhibited by 10-fold
excess of free PE (O) but not by free phosphatidylcholine (.). For each
combination of protein, Iipid, and competitor, Lipidex column separation was performed to yield flow-through (protein-bound) and methanol-eluted (free lipid bound to column) fractions. The percentage of
protein-bound radioactivity is plotted.

O
TES 2 fi

FL
TF~ 120

EFC LVECDL

~~;~;I~~I

.
.
_..
.
.

Bovine K

.
.

KLLNEATHET

T~~IS~C~~IL
I
I

DYVGS

KG

DYVGSG
DYVGSG
EYVGSG
EYMG P

KG
l
KG TI'
KD

KG S

251

300

Br.ifAAINSP LLNNLVVQDS

...ERPGmPT

T~TQFN!I

S~FYRS

~~~~~~~~
:::~~~~I~
~~IQ~I~
~..:..~~Q~
1.~~~~~
EPILSNRSGD
EPILSNRSGD
EPILSNRSGD
EPILSNKSGD
KGVDSS KFSKIKDRPN
RPLKCD
RPLKCD
GPLKCD
QPLNCD

301
TES-26
Oncho
Drome
Human
Macaca
Bovine
Rat
Yeast

QA

SIN FRnVSQVl.Qll1
NPTFEPL,R
QLF'PP FVQPYSR~Or,' R
NFVS
SYNQN

NP1'YQf?VLtl.
SVKEYSSLt'1HR

P
P

92
s'
JC
1/6

250

P.at
Yeast

',Q
N
22IS_NAGIS
111';
.
141

c. e l oq an s

TES-26
Oncho
Drome
Human
Macaca
Bovine
Rat
Yeast

10

0.5

O
TES-26
Oncho

HRGK V
HRGK V
HRGK V
NRGK V
WGYGTPATGV

AS KKY
AS KKY
AS KKY
ES KKY
GKWAKE

,.
CYQA
'.
CYQA

CYQA

CFQA
L.V NFFYA

318
.

RWDEYVPELM
EWDDYVPKLY
EWDDYVPKL,Y
EWDDYVPKLY
EWDDSVPKLH
ETK
.

KTLYGVSE
EQLSGK ..
EQLSGK ..
EQLSGK ..
DQLAGK ..

FIG.5. Sequence alignment of TcSL2 and seven homologues.

Residues identical to TcSL2 in at least 5 of the 7 homologues are
marked with solid boxes. The region in the TcSL2 sequence with
similarity to von WiJlebrand's factor (amino acids 107-114) is denoted
in the shaded box; identical residues within this region are indicated
with asterisks. Oncho, Ag16, a cuticular and secreted antigen from the
filarial nematode parasite O. volvulus (30. 31); Drome, an antennal
protein fromDrosophila melanogaster (37);Human: a protein expressed
by HepG3 hepatoma cells (36); Macaca, an epididyrnal protein from
Macaca fasciculatis (35); Bocine, Bov21, a bovine cytosolic phosphatidylethanolamine-binding protein (33);Rat, a 23-kDa epididymal, brain,
and liver protein also identified as a morphine-binding protein (25, 34);
Yeast, S. cerevisiae TFSl, 24-kDa suppressor of CDC25 mutation, also
tenned DKAl and NSPI (38).
T. canis infective larvae (kindly provided by C. Tripp and R. B.
Grieve, Paravax Inc., Fort Collins, COl was screened with the
850-bp fragment of the 1.1-kb cDNA at high stringency. Hybridization occurred to 0.8-1.9% of plaques (sample size,
1,387), showing that this transcript is highly abundant in infective larvae.

FrG.7. Amino acid sequence alignment of NC6 motifs from

TcSL-2 and TES-120 sequences and in the C. elegans genomic
sequence zk643.6.2 Residues identical in all 5 sequences are shown in
salid boxes.
DISCUSSION

We describe here a cDNA clone that represents an abundantly expressed and stage-specific mRNA and a corresponding
excretory/secretory protein in infective Iarvae of the parasitic
nematode T. canis. Sequence similarity analyses show that
TcSL-2 (TES-26) belongs to a recently defined gene family of
PEBPs. Moreover, TES-26 retains a hydrophobic motifthought
in mammalian homologues to mediate lipid binding. Consistent
with these indications ofits function, we have also shown direct
PE binding by TES-26, and we have preliminary evidence for
an extracellular (surface) association of this molecule. It is
interesting to note that a homologous protein is al so associated
with the surface (cuticle) of o. uoluulus adult worms (30),
which, like larval T. canis, are tissue-dwelling parasites.
Despite the similarities within the PEBP gene families,
TES-26 has numerous distinct features. The most extensive is
that TES-26 alone possesses paired NC6 motifs, suggesting
that it is a hybrid gene composed of domains of diverse ancestry. The presence of similar paired NC6 sequences on two
proteins secreted by infective larvae suggests that this motif
may be associated with sorne surface-related function, possibly
involving rnodulation ofhost immunity. However, the presence
of related sequences in the free-living nematode C. elegans
indicates a more fundamental
role in the biology of the
organismo
TES-26 differs in other important respects. It possesses a
20-residue hydrophobic domain that separates the NC6 motifs
from the PEBP homology regin and that could permit mernbrane association. In other family members, the corresponding
sequences contain numerous charged residues and are unlikely
to fufill this function. AH non-nematode homologues lack signal
peptides and are cellular proteins, although the murine PEBP
has been shown to be secreted, being highly abundant in epididymal fluid (25). Because nematode cuticles are extracellular

18522

26-kDa Antigen from Toxocaracanis

structures (46),the acquisition of signal peptide sequences may

serve both to direct the PEBP to a surface location and to allow
its secretion into the environment of the parasite.
A role for this protein in the uptake and transport of host
lipids is possible because parasitic helminths generally have
limited ability to synthesize long chain fatty acids and cholesterols de novo. For example, the trematodes F. hepatica and S.
mansoni express surface fatty acid-binding proteins (40, 41),
which sequester fatty acids from host serum and incorporate
them into the tegumento However, we found no sequence similarity between TES-26 and proteins known to be involved in
lipid transporto A second possibility, arising from the preferential location of PE on the cytoplasme leaflet of plasma membranes, is that a PEBP could asymmetrically associate with
mernbrane-bound proteins (25). Finally, PEBPs may be involved in the segregation of lipids into biologically distinct
compartments (35). This last possibility is intriguing in the
light of strong experimental evidence that the surface of T.
canis larvae is made up of nondiffusing, discontinuous lipid
domains (17, 47).
We are now studying each of these possibilities to establish
the precise function ofthe nematode PEBPs and to understand
why T. canis should produce this protein in such quantity. The
extraordinarily high expression of TES-120, the surface coat
mucin, and the major antigen targetted by the host response
can be related to the high turnover of the coat as a mechanism
ofimmune evasion by this parasite. Unlike TES-120, for which
the mRNAis hyperabundant, the level of expression of TES-26
appears low relative to its specificmessage. Moreover, the lack
of antibodies in human infection sera was unexpected, especially because the gene was originally selected for reactivity
with antisera from TES-immunized mice. It is conceivablethat
TES-26 is invisible to the human immune system in the context
of an active Toxocara infection and that a lack of response to
this antigen could even be a prerequisite for establishment.
Future studies will therefore aim to assess TES-26 both as a
physiologicalrequirement for parasite survival and as a potential target to induce immunological protection against infection. These questions are made especially important by the
remarkable ability of T. canis to live for many years in a wide
range of mammalian species.
REFERENCES
Sprent, J. F. A. (1958) Parasitology 48, 184-209
Glickman, L. T., and Schantz, P. M. (1981) Epidemiol. ReD. 3, 230-250
GiIlespie, S. H. (1988) Parasitol. Today 4, 180-182
Schantz, P. M. (1989)Am. J. Trop. Med. Hyg. 41, (suppl.) 21-34
Beaver, P. C. (1962) Bull. SocoPathol. Exot. 55,555-576
de Savigny, D. H. (1975) J. Parasitol. 61, 781-782
Maizels, R M., Gems, D. H., and Page, A. P. (1993) in Toxocara and Toxocariasis: Epidemiological, Clinical and Molecular Perspectives (Lewis, J. W.,
and Maizels, R M., eds) pp. 141-150, Institute of Biology, London
8. Maizels, R M., de Savigny, D., and Ogilvie, B. M. (1984) Parasite lmmunol.
(Oxn 6, 23-37

1.
2.
3.
4.
5.
6.
7.

9. Glickman, L. T., Schantz, P. M., and Grieve, R B. (1986) in Immunodiagnosis

of Parasitic Diseases:Part 1, Helminthiasis (Walls, K, and Schantz, P. M.,
eds) pp. 201-232, Academic Presa, Orlando, FL
10. Badley, J. E., Grieve, R B., Bowman, D. D., Glickman, L. T., and Rockey, J. H.
(1987) J. Parasitol. 73, 593-600
11. Sugane, K, and Oshima, T. (1983) Immunology 50, 113-120
12. Meghji, M., and Maizels, R M. (1986) Mol. Biochem. Parasitol. 18, 155-170
13. Page, A. P., and Maizels, R M. (1992) Parasitology 105,297-308
14. Maizels, R M., and Page, A. P. (1990)Acta Trop. 47,355-364
15. Page, A. P., Rudin, W., Fluri, E., Blaxtor, M. L., and Maizels, R M. (1992)Exp.
Parasitol. 75, 72-86
16. Blaxter, M. L., Page, A_ P., Rudin, W., and Maizels, R M. (1992) Parasitol.
Today 8, 243-247
17. Kennedy, M. W., and Proudfoot, L. (1993) in Toxocara canis: Clinical,
Epidemiological and Molecular Perspectives (Lewis, J., and Maizels, R M.,
eds) pp. 151-161, Institute of Biology, London
18. Page, A. P., Hamilton, A. J., and Maizels, R. M. (1992) Exp. Parasitol. 75,
56-71
19. Page, A. P., Richards, D. T., Lewis, J. W., Ornar, H. M., and Maizels, R M.
(1991) Parasitology 103,451-464
20. Chomczynski, P., and Sacchi, N. (1987)Anal. Biochem. 162,156-159
21. Belyavsky, A., Vinogradova, T., and Rajewsky, K (1989)Nucleic Acids Res. 17,
2919-2932
22. Domec, C., Garbay, B., Fournier, M., and Bonnet, J. (1990)Anal. Bioehem. 188,
422-426
23. Sambrook, J., Fritsch, E. F., and Maniatis, T. (1989) Molecular Cloning:
A Laboratory Manual, pp. 13.01-13.102, Cold Spring Harbor Laboratory,
Cold 8pring Harbor, NY
24. Bernier, l., Tresca, J. P., and Jolles, P. (1986) Biochim. Biophye. Acta 871,
19-23
25. Jones, R, and Hall, L. (1991) Biochim. Biophys. Acta 1080, 78-82
26. Lipman, D. J., and Pearson, W. R (1985)Scienee 227, 1435-1441
27. Altschul, S. F., Gish, W., Mmer, W., Myers, E. W., and Lipman, D. J. (1990)
J. Mol. Biol. 215, 403-410
28. Nilsen, T. W. (1993)Annu. Rev. Microbiol. 47,413-440
29. von Heijne, G. (1986)Nucleic Acids Res. 14,4683-4690
30. Lobos, E., Altmann, M., Mengod, G., Weiss, N., Rudin, W., and Karam, M.
(1990) Mol. Biochem. Parasitol. 39, 135-146
31. Lobos, E., Weiss, N., Karam, M., Taylor, H. R., Ottesen, E. A., and Nutman,
T. B. (1991) Science 251, 1603-1605
32. Bernier, t., and JoUs, P. (1984) Biochim. Biophys. Acta 790, 174-181
33. Schoentgen, F., Saccoccio, F., JoBes, J., Bernier, l., and JoBes, P. (1987)
Eur. J. Bioehem. 166, 333-338
34. Grandy, D. K, Hanneman, E., Bunzow, J., Shih, M., Machida, C. A., Bidlack,
J. B., and Civelli, O. (1990) Mol. Endoerinol. 4, 1370-1376
35. Perry, A. C. F., Hall, L., Bell, A. E., and Jones, R (1994) Biochem. J. 301,
235-242
36. Hori, N.,Chae, K-S., Murakawa, K, Matoba, R, Fukushima, A., Okubo, K,
and Matsubara, K (1994) Gene (Amst.) 140, 293-294
37. Pikielny, C. W., Hasan, G., Rouyer, F., and Rosbash, M. (1994) Neuron 12,
35-49
38. Robinson, L. C., and Tatchell, K (1991)Mol. & Gen. Genet. 230,241-250
39. Akeroyd, R, Moonen, P., Westerman, J., Puyk, W. C., and Wirtz, K W. A.
(1981) Eur. J. Biochem. 114,385-391
40. Moser, D., Tendler, M., Griffiths, G., and Klinkert, M.-Q. (1991) J. Biol. Chem.
266,8447-8454
41. Rodriguez-Perez, J., Rodriguez-Medina, J. R, Garcia-Blanco, M. A., and
Hillyer, G. V. (1992) Exp. Parasitol. 74,400-407
42. Baxa, C. A., Sha, R S., Buelt, M. K, Smith, A. J., Matarese, V., Chinander,
L. L., Boundy, K L., and Bernlohr, D. A. (1989)Biochemistry 28, 8683-8690
43. Mori, T., Tsukamoto, T., Mori, H., Tashiro, Y., and Fujiki, Y. (1991)Proc. Natl.
Acad. Sci. U. S. A. 88, 4338-4342
44. Linnestad, e., Loenneborg, A., Kalla, R, and Olsen, O. A. (1991) Plant
Physiology 97, 841-843
45. Sadler, J. E., Shelton-Inloes, B. B., Sorace, J. M., Harlan, J. M., Titani, K, and
Davie, E. W. (1985) Proc. Natl. Acad. Sei. U. S. A. 82,6394-6398
46. Maizels, R M., Blaxter, M. L., and Selkirk, M. E. (1993) Exp. Parasitol. 77,
380-384
47. Kennedy, M. W., Foley, M., Kuo, Y-M., Kusel, J. R, and Garland, P. B. (1987)
Mol. Biochem. Parasitol. 22, 233-240