Available online at www.sciencedirect.

com

International Journal for Parasitology 38 (2008) 1579–1588

www.elsevier.com/locate/ijpara

The intestinal contortin structure in Haemonchus contortus:

An immobilised anticoagulant?
Peter Geldhof a,b, David Knox a,*
a
Moredun Research Institute, Pentlands Science Park, Bush Loan, Penicuik, Midlothian, EH26 0PZ, UK
b
Laboratory of Parasitology, Faculty of Veterinary Medicine, Ghent University, Salisburylaan 133, 9820 Merelbeke, Belgium

Received 17 March 2008; received in revised form 30 April 2008; accepted 2 May 2008

Abstract

Contortin was the first intestinal antigen of the sheep parasite Haemonchus contortus which induced significant levels of protection
when used to vaccinate lambs. This antigen is present in the intestine of L4 and adult worms as a helical polymeric structure attached
to the luminal surface of the intestinal cells. However, the nature of the protein itself and its function have never been reported. In the
present study, contortin was isolated and analysed by peptide mass fingerprint and LC/MS–MS. These analyses indicated that contortin
comprises two major proteins, Hc-PCP1 and Hc-PCP2, with homology to prolyl-carboxypeptidases. The two proteins show 64% amino
acid sequence identity to each other and both are comprised of two prolyl-carboxypeptidase S28 type domains organised in a tandem
repeat. The transcripts of both genes are present from the L4 stage onwards, coinciding with the onset of blood-feeding. Addition of
contortin to a fibrinogen solution significantly inhibited blood coagulation in a dose-dependent manner. Mass-spectrometry indicated
that the contortin-enriched fraction degraded the C-terminal end of the fibrinogen alpha-chain, which was shown previously to be essen-
tial for clot formation. The process happens within seconds after addition and can be inhibited by the dipeptidyl-peptidase IV inhibitors
Diprotin A and Bt-PEG-Glu-ProP(OPh)2. These data suggest that the prolyl-carboxypeptidases are intestinal anticoagulants used by H.
contortus to interfere with blood coagulation.
Ó 2008 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.

Keywords: Haemonchus contortus; Parasitic nematode; Blood-feeding; Anticoagulant; Prolyl-carboxypeptidase; Contortin

1. Introduction PBS extracts of adult parasites by ultracentrifugation

(Munn, 1977). The function of contortin remains unde-
Targeting proteins expressed on the gut surface has fined, but Munn (1977) suggested that it could be an
been a particularly successful approach for vaccine devel- immobilised anticoagulant. When a contortin-enriched
opment in blood-feeding nematodes (Knox et al., 2003) preparation (CEP) was used to vaccinate lambs, worm
and ectoparasites such as ticks (Willadsen et al., 1995). burdens were reduced by 78% following challenge infec-
Munn (1977) first described contortin, using electron tion (Munn et al., 1987). However, it was shown subse-
microscopy, as a helical polymeric extracellular structure quently that CEPs contained a 110 kDa major antigenic
present in large amounts in the intestine and the pharynx contaminant as judged by Western blotting despite only
from the L4 and adult stage of the sheep parasite Hae- faint staining being evident in protein gels (Smith and
monchus contortus. The polymer is associated with the Munn, 1990). This protein was purified using lectin-affin-
luminal surface of the plasma membrane of the intestinal ity chromatography and was subsequently defined as
epithelium and its helical filaments fill the spaces between H11, the most effective immunogen isolated from a par-
the microvilli. It is insoluble and can be enriched from asitic nematode to date (Knox et al., 2003). Subsequent
effort focused on H11 and other gut proteins in Haemon-
*
Corresponding author. Tel.: + 44 131 4455111; fax: +44 131 4456111. chus and the exact nature and function of contortin was
E-mail address: david.knox@moredun.ac.uk (D. Knox). not investigated further.

0020-7519/$34.00 Ó 2008 Australian Society for Parasitology Inc. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijpara.2008.05.002
1580 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588

The aim of this study was to characterise contortin at Table 1

the molecular level in an effort to elucidate its potential PCR primer sequences used for reverse transcriptase-PCR experiments
function. Haemonchus contortus Primer sequence
EST cluster (gene name)
2. Materials and methods HCC00232 (Hc-pcp1) F RT-PCR 50 aag cag gtt cag cct ttt ca
R RT-PCR 50 ctt gcc cat tgt tcg ttt tt
50 RACE 50 cat aat gac gca cga tcg ac
2.1. Preparation of a contortin-enriched fraction HCC00298 (Hc-pcp2) F-RT-PCR 50 cgc tga cct gtg tca agt gt
R RT-PCR 50 aag tag ccg gct tcg tat tg
CEP was prepared as previously described by Munn 30 RACE 50 agc atc gcc gct ttc aat
(1977). In short, adult H. contortus were homogenised in
PBS (1:10, w/v) and subsequently centrifuged for 5 min at website (www.nematodes.org/nematodeESTs/nembase.
1000g. The supernatant was collected and the pellet html). Further analyses of sequences and the deduced amino
re-homogenised in PBS followed by centrifugation at acid sequences were performed using DNAstar software
1000g. The supernatants were pooled and centrifuged at (DNAstar Inc.). Database searches were performed on the
3000g for 10 min. The pellet was washed once with PBS NCBI server (www.ncbi.nlm.nih.gov/blast). Amino
and the pooled supernatants were centrifuged for 90 min at acid sequences were analysed for signal peptides and glyco-
10,000g. The pellets, which contain contortin, were resus- sylation sites using Signal P-(www.cbs.dtu.dk/services/
pended in 1 ml of PBS and stored at 80 °C prior to analysis. SignalP) and NetNGlyc-programs (www.cbs.dtu.dk/services/
NetNGlyc).
2.2. Gel electrophoresis and mass-spectrometry analysis
2.5. Reverse transcriptase PCR
An aliquot (20 ll) of CEP was fractionated in a 10%
SDS–PAGE gel under reducing conditions. The protein Reverse transcriptase PCR (RT-PCR) was used to
components were visualised by Coomassie Blue staining. determine the stage-specific transcription of Hc-pcp1 and
All visible protein bands were excised from the gel and used Hc-pcp2. Total RNA was extracted from exsheathed L3,
for mass-spectrometry analysis. Protein bands were L4 and adult parasites using Total RNA Isolation Reagent
digested in-gel using trypsin, the resultant peptides subse- (Advanced Biotechnologies Ltd.). The RT-PCRs were car-
quently isolated by HPLC using a reversed phase C18 Pep- ried out using the Superscript One-Step RT-PCR System
Map column and finally analysed by MALDI-TOF mass- (Invitrogen). The specific primers used for detection of
spectrometry and LC/MS–MS analysis. The Mascot search the Hc-pcp1 of the Hc-pcp2 transcripts are shown in Table
engine was used to statistically analyse the mass-spectrom- 1. After 35 cycles of amplification, the RT-PCR products
etry and LC/MS–MS data and to identify the proteins. were separated on 1% agarose gels. A cytoplasmic superox-
ide dismutase gene (SODc; acc. Z69621; Liddell and Knox,
2.3. Isolation and cloning of cDNAs 1998) was used to control for the efficiency of the RNA
purifications.
The LC/MS–MS analyses identified the prominent com-
ponent of CEP as prolyl-carboxypeptidases encoded by 2.6. S28 peptidase activity assays
two expressed sequence tag (EST) clusters, HCC00232
and HCC00298 (www.nema.cap.ed.ac.uk/nematodeESTs/ Dipeptidyl peptidase (DPP) activity was measured using
nembase.html). Full length cDNA sequences of the pro- the chromogenic substrates H-Ala-Pro-pNitroanilide (DPP
lyl-carboxypeptidases (PCPs; from hereon named Hc- IV, 1.0 M in methanol; Bachem) and H-Lys-Ala-pNitroani-
pcp1 and Hc-pcp2) were isolated from a cDNA library lide (DPP II, 1.0 M in methanol; Bachem). Different quanti-
made from 11-day-old H. contortus worms. Based on the ties of CEP (10–30 lg) were mixed with the substrates (final
consensus sequences of EST clusters, specific primers were conc. 1 mM) in 100 ll of buffer ranging from pH 3 to 8 (pH
designed to isolate the 50 and 30 ends of each cDNA. The 3–5: 0.1 M sodium acetate buffer, pH 6–7: 0.1 M phosphate
primer sequences are shown in Table 1. The gene-specific buffer, pH 8: 0.1 M Tris buffer). Samples were incubated for
primers were used in combination with the T3 cDNA 45 min at 37 °C when the optical density was measured at
library vector primer to amplify both full-length cDNA 405 nm. The effect of DPP IV inhibitors, Diprotin A and
sequences. PCR products were cloned in pGEM-T (Pro- Diprotin B (Bachem), on activity was determined by adding
mega) and sequenced. Sequence analyses and alignments different concentrations of the individual inhibitors (10, 100,
were performed using DNAstar software (DNAstar Inc.). 500 and 1 mM) to the reaction mixture described above. All
experiments were performed in duplicate.
2.4. Bioinformatics
2.7. Fibrin clotting/re-aggregation assay
The EST datasets of H. contortus and other nematodes
were analysed using the Partigene bioinformatics pipeline A bovine fibrin solution (concentration 15 mg/ml)
(Parkinson et al., 2004) which is available on the NEMBASE (Sigma) was clotted by the addition of 1 ll of a thrombin
 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588 1581

stock solution (10 mg/ml) (Sigma) and incubated at 37 °C 3. Results

for 15 min. The fibrin clot was collected and resuspended
in 8 ml of 1 M NaBr/0.05 Na acetate buffer pH 5.3. The 3.1. Characterisation of the contortin-enriched fraction
resultant fibrin monomer solution was subsequently con- (CEP)
centrated to 750 ll and stored at 4 °C prior to use.
Re-aggregation of the fibrin solution was carried out by The protein profile of the CEP is shown in Fig. 1. The
mixing 5 ll of the fibrin monomer solution with 95 ll of prominent band (arrowed in Fig. 1) and the remaining vis-
PBS. The re-aggregation was monitored in a fluorimeter ible bands were excised from a Coomassie-stained gel, ana-
by measuring the scattered light at a 90° angle at 350 nm. lysed by LC/MS–MS after a tryptic digest and then used to
Results are shown as the change in scattered light over a generate a peptide mass fingerprint which was screened
10 min period. The effect of CEP on the fibrin re-aggrega- against all the H. contortus EST datasets available on the
tion was assessed by incubating 5 ll of the fibrin monomer Nembase website. The prominent band yielded three
solution with different quantities of CEP for 30 min at peptide sequences which showed a 100% match with clus-
37 °C prior to the fluorimetric analysis. The effect of the ters HCC00232 and HCC00298 (Table 2). Clusters
DPP IV inhibitor Diprotin A was determined by adding HCC00232 and HCC00298, respectively, contain 421 and
0.4 ll of a 0.5 M inhibitor stock solution to the reaction 37 individual EST sequences. The consensus sequence of
mixture described above. All incubations were done in cluster HCC00232 was 2579 bp long. An additional
the presence of the inhibitors (all from Sigma), Amin- 884 bp at the 50 end was isolated from a cDNA library
oethylbenzenesulfonyl fluoride hydrochloride (AEBSF, by a PCR approach using gene-specific primers in combi-
1 mM), Pepstatin (0.1 mM), trans-epoxysucciny-L-leucyl- nation with the T3 vector primer. This resulted in a full-
amido (4-guanidino) butane (E64, 0.1 mM) and 1.10 Phe- length coding sequence which encodes a 122 kDa protein.
nanthroline (1, 10 Phe, 50 mM) to block the activity of ser-
ine, aspartyl, cysteine and metallo-proteases, respectively.
1 2
2.8. Fibrin degradation assay

Twenty-five micrograms of fibrin monomers, prepared

as described above, were mixed with different amounts of 250 1
CEP and incubated at 37 °C for different lengths of time,
148 2
ranging from 15 s to 30 min. The protease inhibitors
AEBSF, Pepstatin, E64 and 1.10 Phe were included in
the reaction mixtures at final concentrations described
above. Samples were subsequently analysed on 10% 98
SDS–PAGE gels under reducing conditions and visualised
by Coomassie Blue staining. The effects of the DPP IV-spe-
cific inhibitors Diprotin A and Bt-PEG-Glu-ProP(OPh)2
(Gilmore et al., 2006) were investigated by adding the 60
3
inhibitors to the reaction mixture described above at a final
concentration of 0.1 M and 50 lM, respectively. The inhib-
itors were preincubated with contortin for 30 min prior to
50 4
addition of fibrin monomers.

2.9. Labelling of CEP with Bt-PEG-Glu-ProP(OPh)2

5
Ten micrograms of CEP was mixed with 50 lM of the 36 6
DPP IV-specific biotinylated inhibitor Bt-PEG-Glu-Pro-
P 7
(OPh)2 (Gilmore et al., 2006) and incubated for 15 min
at 37 °C. Samples were subsequently run onto a 10%
SDS–PAGE under reducing conditions and then blot
transferred onto a polyvinylidene fluoride (PVDF) mem-
brane. Membranes were blocked in 10% horse serum in
PBS containing 0.05% Tween 20 (PBST) for 1 h and then
incubated with conjugated streptavidin (Sigma) diluted
Fig. 1. Ten percent SDS–PAGE profile of the contortin-enriched protein
1:5000 in PBST for 1 h. Bands were visualised by adding
fraction (CEP) under reducing conditions. Lane 1: marker, lane 2: CEP.
0.05% 3,3 diaminobenzidine tetrachloride in PBS contain- The numbered zones and the prominent band around 55 kDa (arrow)
ing 0.01% H2O2. CEP incubated without the addition of were excised and analysed by tryptic digest, peptide mass fingerprint
the inhibitor served as the negative control. analysis and LC–MS/MS analysis.
1582 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588

Table 2 Peptide mass fingerprint and LC/MS–MS analysis of

Peptide sequences from the LC–MS/MS analysis of the 55 kDa band additional protein bands (numbered in Fig. 1) resulted in
Peptides EST cluster the identification of a further eight proteins present in
LDYHEYYQVVEASIR DFDEEGWASVDR HCC00232 CEP (Table 3). These include a myosin, the aminopepti-
FDFWEGTQFAEDIYR DFDEEGWASVDR HCC00298 dase H11, glutamate dehydrogenase, apical gut membrane
protein, metallo-peptidase 3, galectin and cysteine
Cluster HCC00298 has a consensus sequence of 3037 bp. proteinases.
Together with an additional 562 bp at the 30 end, which The full-length H. contortus PCP1 and PCP2 sequences
was isolated in a similar approach as described above, it were also used to search EST databases of other parasitic
codes for a 129 kDa protein. Both proteins show homology nematodes. EST clusters encoding PCPs were identified
to PCP-like proteins in Caenorhabditis elegans (PCP-2 acc. in Ancylostoma caninum (cluster ACC04826, two ESTs),
NP_501599 and PCP-3 acc. NP_501598.1). The shared Ascaris suum (cluster ASC18950, one EST), Onchocerca
identity between the H. contortus and C. elegans proteins volvulus (cluster OVC01220, one EST), Strongyloides ratti
was approximately 38%. The two H. contortus proteins (cluster SRC00324, six ESTs) and Parastrongyloides tricho-
are herein named Hc-PCP1 (cluster HCC00232) and Hc- suri (cluster PTC00674, one EST). Sequence homology var-
PCP2 (cluster HCC00298). An alignment of Hc-PCP1 ied between 32% and 42% over a maximum of 205 amino
and Hc-PCP2 is shown in Fig. 2A. The two proteins show acids.
64% identity and 77% similarity to each other in amino
acid sequence. Both contain a predicted signal peptide, 3.2. Transcription profile of the Hc-pcp’s
albeit at a different location. Both proteins also contain
multiple putative glycosylation sites, marked in Fig. 2A. Hc-pcp1 and Hc-pcp2-specific primers were used in
Both Hc-PCP1 and Hc-PCP2 are comprised of two serine RT-PCR on total RNA samples from exsheathed L3,
peptidase S28 type domains (pfam05577) organised in a L4, 11- and 22-day-old parasites (Fig. 3). Transcripts
tandem repeat. This feature seems to be nematode-specific; of the two genes were present from the L4 stage
only the C. elegans PCPs show a similar structure. Each of onwards whilst SODc was transcribed by all the parasite
the S28 domains encode a 51 kDa peptidase, which stages.
approximately matches the size of the prominent band on
SDS–PAGE gels (arrowed in Fig. 1). Therefore, it is likely
that the two tandem structured PCPs are post-translation- 3.3. S28 peptidase activity assays
ally processed into the four individual prolyl-carboxypep-
tidases, from hereon termed PCP1A, PCP1B, PCP2A and At pH 7.5, CEP hydrolysed the DPP IV-specific sub-
PCP2B. Interestingly, Hc-PCP2 has a unique C-terminal strate in a dose-dependent manner (Fig. 4A). The opti-
extension which is proline-rich and hydrophobic. The exact mal activity was at pH 6 (Fig. 4B) and this was
function of this part of the protein is unknown, but it might partially inhibited by Diprotin A (H-Ile-Pro-Leu; Fig.
be involved in anchoring the protein to a cell membrane. 4C) in a dose-dependent manner but unaffected by
Fig. 2B shows a partial sequence alignment of the region Diprotin B. (H-Val-Pro-Leu-OH). No activity could be
around the active sites of the different H. contortus PCPs detected against the DPP II substrate (results not
and human PCP. Catalysis by this type of peptidase is shown).
dependent on three active site residues, namely serine, his-
tidine and aspartic acid which, although being dispersed 3.4. Coagulation assays
throughout the primary protein sequence, come together
in close proximity to form the active site when the enzyme Mentlein and Heymann (1982) reported previously
is folded correctly. The catalytic residues in the Haemon- that DPP IV inhibited the polymerisation of fibrin
chus PCP sequences here are ordered serine, aspartic acid monomers. Therefore, a coagulation assay with purified
and histidine and this feature, alongside the homology to fibrinogen was used to analyse if CEP showed anti-clot-
PCPs place these enzymes in the clan SC, family S28 ting activity and if it was targeted against fibrinogen.
(Underwood et al., 1999). The region around the putative The results of this fibrin re-aggregation assay are shown
active site serine is highly conserved with 6/10 residues in Fig. 5A. The addition of CEP to a fibrin monomer
identical in all the comparators, this number diminishing solution significantly inhibited the re-aggregation in a
to 3/10 and 1/8 around the putative active site aspartic acid dose-dependent manner, a 55% and 75% reduction with
and histidine, respectively. Of the PCP domains, PCP 1A 2.5 and 5 ll, respectively, as measured by the change in
showed closest homology to PCP2B – 69.4% identity with scattered light over a 10 min period. This effect was sub-
81.8% similarity. Comparisons between other domain com- stantially overcome by the addition of the DPP IV inhib-
binations all resulted in 30% identity, 50% similarity. All itor Diprotin A (Fig. 5B). Addition of specific inhibitors
domains showed around 19% identity and 30% similarity for all four classes of proteinases (serine, cysteine, metal-
to DPP IV of human origin and 25% identity, 40% similar- lo and aspartyl proteases) did not inhibit the anti-clotting
ity to the PCP of human origin. activity.
 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588 1583

Fig. 2. The figure shows a pairwise alignment of the derived amino acid sequences from the two Haemonchus prolyl-carboxypeptidase (PCP)-encoding
transcripts (A) and a comparison of the active site regions of the individual PCP domains they encode (B). Predicted cleavage sites of the N-terminal signal
sequences in both proteins are marked with an arrow. Putative glycosylation sites are marked with an * (A). Both proteins contain two prolyl-
carboxypeptidase S28 type domains which are marked with []. A partial sequence alignment of the region around the active sites of the different
H. contortus PCPs and human PCP is shown in (B). The residues around the active site serine are shown in bold, the residues around the aspartic acid site
by single underline and the histidine site in double underline. The order of these active site residues places the PCPs in the peptidase clan SC, family S28.
1584 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588

Table 3
Additional proteins identified in the contortin-enriched protein fraction by A 0.9
pH 7.5 DPP IV
LC/MS–MS 0.8
0.7
Zone on SDS–PAGE Protein ID

OD 405 nm
gel (Fig. 1) 0.6

1 Myosin (DQ310759) 0.5

2 Integral membrane protein H11 (X94187) 0.4
3 Glutamate dehydrogenase (Hc-gldh1, 0.3
AF000967)
4 Apical gut membrane protein (GA1, U55864) 0.2
and metallo-peptidase (MEP III, AF080172) 0.1
5 Cysteine proteinase (HaeCytPro2 – M80386) 0
6 Galectin (Hco-gal-1, AF077944) 0 µg 10 µg 20 µg 30 µg
7 Cysteine proteinase (GCP7, AF046229) CEP concentration

B 1.4
20 µg
1.2
XL3 L4 d11 d22
1

OD 405 nm
Hc-pcp1 0.8

0.6

0.4
Hc-pcp2
0.2

0
pH3 pH4 pH5 pH6 pH7 pH8
SODc
0.8
C
Fig. 3. Presence of Hc-pcp1 and Hc-pcp2 transcripts as revealed by reverse 0.7
transcriptase (RT)-PCR. The lanes are independent RT-PCR reactions DiprotinB
using target RNA from exsheathed L3, L4, 11- and 22-day-old adult 0.6
parasites. RT-PCR for a cytoplasmic superoxide dismutase (SODc) was
OD 405 nm

0.5
used as a control to check the uniformity of the RNA purifications.
0.4

3.5. Fibrin degradation assay 0.3

0.2 DiprotinA
SDS–PAGE gel analysis of the fibrin monomer solution
0.1
under reducing conditions (Fig. 6A lane 1) revealed three
protein bands of approximately 63, 56 and 47 kDa, the 0
0 uM 10 uM 100 uM 500 uM 1000 uM
a-, b- and c-chains, respectively. The addition of CEP Inhibitor concentration
resulted in the proteolytic degradation of both the a- and
Fig. 4. Peptidase activity assays. (A) Dose-dependent degradation of the
b-chain after 15 s of incubation at 37 °C (lane 2). The addi-
dipeptidyl peptidase IV-specific substrate by the CEP at pH 7.5. (B)
tion of specific inhibitors for all four classes of proteinases Activity against the DPP IV-specific substrate by 20 lg of the CEP
(serine, cysteine, metallo and aspartyl proteases) stopped incubated at different pH. (C) Testing the effect of two DPP IV-specific
the degradation of the b-chain, but not the a-chain (lane inhibitors on the DPP IV type activity of the CEP.
3). However, preincubating CEP with the DPP IV-specific
inhibitors Diprotin A (lane 4) and Bt-PEG-Glu-Pro-
P tein, which suggested that the C-terminal end of the a-
(OPh)2 (lane 5) completely abolished the degradation of
chain was degraded. As a control, the a-chain of an
the fibrinogen a-chain.
untreated sample was analysed and this analysis confirmed
The specific degradation of the a-chain coincided with
that C-terminal peptides covering the full a-chain sequence
an intensification of the c-chain, as shown in lane 3
could normally be generated.
(arrow), suggesting the presence of proteolytic degradation
products of the a-chain. Mass-spectrometric fingerprint
and LC–MS/MS analysis on this protein band revealed 3.6. Bt-PEG-Glu-ProP(OPh)2 affinity labelling
the presence of both the c- and a-chain, whereas in the c-
band of an untreated sample no peptides of the a-chain The labelling of CEP with the DPP IV-specific inhibitor
were present. Fig. 6B shows the location of the identified Bt-PEG-Glu-ProP(OPh)2 is shown in Fig. 7A. A strong
peptides in the complete a-chain protein sequence. No pep- band is visible around 55 kDa, (lane 3), coinciding with
tides could be identified in the C-terminal part of the pro- the size of the Hc-PCP1 and Hc-PCP2 on a Coomassie-
 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588 1585

in the total Haemonchus EST dataset (Geldhof et al., 2005)

A
Change in scattered light
2.5
and by far the most abundant EST in the intestinal EST
2 dataset. This is consistent with the observations by Munn
1.5 (1977) who showed that contortin was a major protein in
1
the intestine of H. contortus, present from the L4 stage
onwards. The Hc-pcp2 transcript on the other hand seems
0.5 to be present at a much lower level. Only 37 ESTs are pres-
0 ent in the EST dataset of which two were identified in the
Control 2 µg CEP 5 µg CEP intestinal cDNA library. This suggests that the contortin
structure is likely to be mainly composed of the two PCPs
B 1.6
coded by Hc-PCP1. Nevertheless, a peptide specific for
Change in scattered light

1.4 PCP2A was also identified in the mass spectrometric anal-

1.2 ysis of the contortin structure.
1 The PCPs described in this paper show homology to the
0.8 peptidase S28 family. This is a family of serine peptidases
0.6 which include both amino- and carboxy-exopeptidases,
0.4 examples of which are lysosomal Pro-Xaa carboxypepti-
0.2 dase, dipeptidyl peptidase II and DPP IV. Enzyme activity
0
analyses indicated that DPP IV activity predominated with
Control 5 µg CEP 5 µg CEP a pH optimum at pH 6, activity being substantially inhib-
+DiprotinA ited by the competitive DPP IV inhibitor Diprotin A.
Fig. 5. Inhibition of fibrin re-aggregation by the addition of CEP. (A) The Munn (1977) previously proposed that contortin was an
re-aggregation of control fibrin monomers compared with the re-aggre- immobilised anticoagulant. To investigate this hypothesis
gation of fibrin monomers incubated with different concentrations (2 and we have analysed the anti-clotting activity of CEP against
5 lg, respectively) of CEP. (B) Analysis of the effect of Diprotin A on the purified fibrin monomers. CEP significantly interfered with
inhibition of fibrin re-aggregation. All results are shown as changes in
fibrin clot formation within seconds after addition. The
scattered light measured at 350 nm, which is a measure of the amount of
re-aggregated fibrin monomers. high rate of this enzymatic reaction might be important
for the parasite to immediately prevent blockage of the
intestine by the blood meal. The reduced clotting is a result
stained gel (lane 1). Other fainter bands were visible, partic- of the degradation of the C-terminal end of the fibrinogen
ularly around 38 kDa. No bands were recognised in the alpha-chain (aC-domain). Cysteine proteinases have previ-
negative control sample, which consisted of CEP without ously been implicated in fibrinogen degradation by extracts
the addition of Bt-PEG-Glu-ProP(OPh)2 (lane 2). of adult H. contortus (Boisvenue et al., 1992). Here, addi-
tion of specific inhibitors for all four classes of proteinases
4. Discussion (serine, cysteine, metallo and aspartyl proteases) to CEP
did not inhibit the anti-clotting activity, nor did it have
The intestinal surface of H. contortus is lined by a helical an effect on the carboxypeptidase type degradation of the
polymeric structure, termed contortin, which fills the fibrinogen alpha-chain. Degradation of the alpha-chain
spaces between the microvilli (Munn, 1977). In the present was only inhibited when the DPP IV inhibitors Diprotin
study, we have identified the major protein component of A and Bt-PEG-Glu-ProP(OPh)2 were included in the reac-
the structure as prolyl-carboxypeptidase-like enzymes tion mixture, results which suggest that the PCPs are
encoded by two mRNA transcripts, each of which, in turn, responsible for the degradation of the fibrinogen alpha-
encodes two related but distinct PCPs. Consistent with chain. However, apart from the PCPs, the use of the biotin-
these findings, we show that a protein fraction highly ylated DPP IV inhibitor also resulted in the labelling of
enriched for these enzymes has DPP IV activity, impairs some fainter bands around 38 kDa. At this stage it is
fibrin clot formation and we provide compelling evidence unclear if this is background recognition or if it is actually
that this property is mediated by DPP IV-dependent cleav- caused by the presence of additional DPP IV like enzymes
age of the alpha-fibrinogen chain. The latter was unaffected in the CEP fraction. If this is the case, we cannot exclude
by the addition of a cocktail of inhibitors against serine, their involvement in the fibrinolytic activity.
cysteine, metallo and aspartyl proteases but completely Blood-feeding parasites have developed an array of spe-
abolished by the addition of DPP IV-specific inhibitors. cific inhibitors to interfere with mammalian blood coagula-
The genes (hc-pcp1 and hc-pcp2) encoding these proteins tion (Ledizet et al., 2005). Most of these inhibitors target
are both transcribed from the L4 stage onwards. The level the serine proteases involved in the early steps of the clot-
of transcript, based on EST levels, is, for Hc-pcp1, extre- ting cascade. To date, a number of anticoagulants have
mely high. The Haemonchus EST dataset contains 421 been identified in both hookworms and ticks which inhibit
ESTs for Hc-pcp1 of which 362 were identified in an intes- the coagulation factor Xa, the prothrombinase complex
tinal cDNA library. This makes it the third biggest cluster and the fVIIa/tissue factor complex (Mans and Neitz,
1586 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588

A 1 2 3 4 5

-chain
-chain
-chain
-chain
+ -chain

MFSVRDLCLVLSLVGAIKTEDGSDPPSGDFLTEGGGVRGPRLVERQQSACKETGWPFCS
DEDWNTKCPSGCRMKGLIDEVDQDFTSRINKLRDSLFNYQKNSKDSNTLTKNIVELMRG
DFAKANNNDNTFKQISEDLRSRIEILRRKVIEQVQRIKVLQKNVRDQLVDMKRLEVDIDIKIR
SCKGSCSRALEHKVDLEDYKNQQKQLEQVIAINLLPSRDIQYLPLIKMSTITGPVPREFKSQ
LQEAPLEWKALLEMQQTKMVLETFGGDGHARGDSVSQGTGLAPGSPRKPGTSSIGNVNP
GSYGPGSSGTWNPGRPEPGSAGTWNPGRPEPGSAGTWNPGRPEPGSAGTWNPGRPE
PGSAGTWNPGRPEPGSAGTWNTGSSGSSSFRPDSSGHGNIRPSSPDWGTFREEGSVSS
GTKQEFHTGKLVTTKGDKELLIDNEKVTSGHTTTTRRSCSKVITKTVTNADGRTETTKEVV
KSEDGSDCGDADFDWHHTFPSRGNLDDFFHRDKDDFFTRSSHEFDGRTGLAPEFAALGE
SGSSSSKTSTHSKQFVSSSTTVNRGGSAIESKHFKMEDEAESLEDLGFKGAHGTQKGHTK
ARPARGIHTSPLGEPSLTP

Control
…..…………TVTNADGRTETTKEVVKSEDGSDCGDADFDWHHTFPSRGNLDDFFHRDKD
DFFTRSSHEFDGRTGLAPEFAALGESGSSSSKTSTHSKQFVSSSTTVNRGGSAIESKHFK
MEDEAESLEDLGFKGAHGTQKGHTKARPA RGIHTSPLGE PSLTP

Fig. 6. Proteolytic degradation of the fibrin a-chain by CEP and the inhibition of this activity by dipeptidyl IV-specific inhibitors Diprotin A and Bt-PEG-
Glu-ProP(OPh)2. (A) 10% reducing SDS–PAGE gel of a control fibrin solution (10 lg) (lane 1), fibrin (10 lg) incubated with CEP for 15 s (lane 2), fibrin
(10 lg) incubated for 15 s with CEP in the presence of E64, 1,10 Phe, AEBSF and Pepstatin (lane 3), fibrin (10 lg) incubated for 15 s with CEP
preincubated with the DPP IV-specific inhibitors Diprotin A (lane 4) and Bt-PEG-Glu-ProP(OPh)2 (lane 5). (B) Complete peptide sequence of the bovine
fibrinogen a-chain (P02672). The peptides highlighted were identified by mass-spectrometry analysis or LC/MS–MS in the fibrinogen c-band after
incubation with CEP. The peptides identified in the C-terminal end of a control fibrinogen c-band (without incubation with the PCP-enriched protein
fraction) are highlighted separately. The results suggest that the C-terminal end of the a-chain was degraded by the PCP activity.

2004; Mieszczanek et al., 2004). The anticoagulant activity far as we know, it is the first time this type of anti-throm-
in CEP, most likely due to the PCPs, targets the key struc- botic activity has been identified in any organism. Some
tural component of the blood clot itself, fibrinogen. The serine and metallo-proteases have been identified in snake
structure of fibrinogen and fibrin has recently been venom with a substrate preference for the fibrinogen a-
reviewed by Mosesson (2005). In short, fibrinogen has a chain (Matsui et al., 2000) and a dipeptidyl peptidase IV
dimeric structure of with each monomer composed of a-, has been identified in human placenta which cleaved the
b- and c-chain. The N-terminal domains of the six chains glycylproline residues from the N-terminal end of the fibrin
in a fibrinogen molecule are linked together by disulphide a-chain (Mentlein and Heymann, 1982). However, the
bonds to form the E-domain. When clotting occurs, throm- specificity for the aC-domain seems to be unique. This part
bin cleaves off the fibrinopeptides at the N-terminal ends of of the a-domain has shown to be important for both intra-
the a- and b-chain. The newly exposed N-terminal ends of and intermolecular interactions in fibrin and fibrils (Moses-
these chains will respectively bind to the terminal cC-and son, 2005). Fibrinogen molecules lacking this domain dis-
bC-domain of an adjacent fibrinogen molecule and form play a slower rate of assembly, a reduced turbidity and
fibrin. Multiple fibrin molecules will align to form fibrils generate thinner fibres (Mosesson, 2005). The aC-domains
and these will undergo lateral associations with other fibrils are normally non-covalently associated with the central E-
to create multi-stranded fibres and eventually the clot. domain. When thrombin cleaves the fibrinopeptide B’s, the
Here, degradation specifically targets the aC-domain. As aC-domains dissociates from this E-domain and become
 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588 1587

Given their apparently novel mode of action, PCPs

A 1 2 3 might have a potential as drug targets. In recent years,
much attention has been given in the medical research field
98
to the development of inhibitors for this type of enzyme, in
62 particular for DPP IV (Rosenblum and Kozarich, 2003).
49 Such inhibitors, when available, could easily be tested in
the coagulation assays for their inhibitory activity of the
38 H. contortus PCPs. Moreover, the H. contortus PCPs might
also have medical applications. A number of anti-throm-
28 botics from haematophagous invertebrates have been eval-
uated in vivo for the treatment of a variety of conditions
17 associated with blood clotting. Some of these are already
in varying stages of preclinical and clinical development,
14
for example the nematode anticoagulant protein-2 (Ledizet
et al., 2005).

Fig. 7. Affinity labelling of contortin-enriched preparation (CEP) with the Acknowledgements

dipeptidyl peptidase IV (DPP IV)-specific inhibitor Bt-PEG-Glu-Pro-
P
(OPh)2. Coomassie and Western blot analysis of CEP. Lane 1: 10 lg CEP
on Coomassie-stained SDS–PAGE. Lanes 2 and 3: 10 lg CEP on Western
P.G. is a Postdoctoral Fellow of the Fund for Scientific
blot developed with streptavidin conjugate, respectively, with and without Research – Flanders (Belgium) (F.W.O. – Vlaanderen). We
the addition of Bt-PEG-Glu-. The PCP band is marked with an arrow. thank Dr. Brendan Gilmore for the biotinylated protease
inhibitor.

available for interaction with other aC-domains (Moses- References

son, 2005). This increased sterical accessibility of the aC-
domain might be the reason why it is targeted by the PCPs. Boisvenue, R.J., Stiff, M.I., Tonkinson, L.V., Cox, G.N., Hageman, R.,
The high intestinal transcription levels of the PCPs seem 1992. Fibrinogen-degrading proteins from Haemonchus contortus used
to be very specific for H. contortus. Using the Haemonchus to vaccinate sheep. Am. J. Vet. Res. 53, 1263–1265.
Geldhof, P., Whitton, C., Gregory, W.F., Blaxter, M., Knox, D.P.,
PCP sequences as search queries, we identified EST clusters
2005. Characterisation of the two most abundant genes in the
from A. caninum, A. suum, O. volvulus, S. ratti and P. Haemonchus contortus expressed sequence tag dataset. Int. J.
trichosuri that encode for putative PCPs. However, each Parasitol. 35, 513–522.
of these clusters contained a small number of ESTs, rang- Gilmore, B.F., Carson, L., McShane, L.L., Quinn, D., Coulter, W.A.,
ing from one to six ESTs, indicating an overall low tran- Walker, B., 2006. Synthesis, kinetic evaluation, and utilization of a
biotinylated dipeptide proline diphenyl phosphonate for the disclosure
scription level. In addition, no PCPs have been identified
of dipeptidyl peptidase IV-like serine proteases. Biochem. Biophys.
in the intestinal EST libraries from the two hookworm spe- Res. Commun. 347, 373–379.
cies Necator americanus and A. caninum (Ranjit et al., Knox, D.P., Redmond, D.L., Newlands, G.F., Skuce, P.J., Pettit, D.,
2006) and no new sequence homologues were identified in Smith, W.D., 2003. The nature and prospects for gut membrane
a recent search of the databases. Munn (1977) previously proteins as vaccine candidates for Haemonchus contortus and other
ruminant trichostrongyloids. Int. J. Parasitol. 33, 1129–1137.
suggested, based on microscopic observations, the presence
Ledizet, M., Harrison, L.M., Koskia, R.A., Cappello, M., 2005. Discovery
of a contortin-like structure in the intestine of the sheep and pre-clinical development of antithrombotics from hematophagous
parasite Teladorsagia circumcincta. However, analysis of invertebrates. Curr. Med. Chem. Cardiovasc. Hematol. Agents 3, 1–
a T. circumcincta protein extract, purified according to 10.
the methods described by Munn (1977), showed a totally Liddell, S., Knox, D.P., 1998. Extracellular and cytoplasmic Cu/Zn
superoxide dismutases from Haemonchus contortus. Parasitology 116,
different protein profile and did not contain any detectable
383–394.
levels of DPP IV activity (P. Geldhof and D. Knox, unpub- Mans, B.J., Neitz, A.W., 2004. Adaptation of ticks to a blood-feeding
lished data). environment: evolution from a functional perspective. Insect Biochem.
The data here indicate that PCPs potentially play a cru- Mol. Biol. 34, 1–17.
cial role in the survival of H. contortus. Despite a successful Matsui, T., Fujimura, Y., Titani, K., 2000. Snake venom proteases
affecting hemostasis and thrombosis. Biochim. Biophys. Acta 1477,
vaccination trial with the CEP fraction (Munn et al., 1987),
146–156.
it is unclear what the protective capacity is of the PCPs Mentlein, R., Heymann, E., 1982. Dipeptidyl peptidase IV inhibits the
themselves. The analysis of CEP here showed the presence polymerization of fibrin monomers. Arch. Biochem. Biophys. 217,
of several other protective antigens, such as H11, an apical 748–750.
gut membrane protein, cysteine proteinases and metallo- Mieszczanek, J., Harrison, L.M., Cappello, M., 2004. Ancylostoma
ceylanicum anticoagulant peptide-1: role of the predicted reactive site
protease III, all of which could have been implicated in
amino acid in mediating inhibition of coagulation factors Xa and VIIa.
vaccine-induced protection against Haemonchus challenge. Mol. Biochem. Parasitol. 137, 151–159.
Further purification is required to fully analyse the protec- Mosesson, M.W., 2005. Fibrinogen and fibrin structure and functions. J.
tive capacity of the PCPs. Thromb. Haemost. 3, 1894–1904.
1588 P. Geldhof, D. Knox / International Journal for Parasitology 38 (2008) 1579–1588

Munn, E.A., 1977. A helical, polymeric extracellular protein associated Rosenblum, J.S., Kozarich, J.W., 2003. Prolyl peptidases: a serine
with the luminal surface of Haemonchus contortus intestinal cells. protease subfamily with high potential for drug discovery. Curr. Opin.
Tissue Cell 9, 23–34. Chem. Biol. 7, 496–504.
Munn, E.A., Greenwood, C.A., Coadwell, W.J., 1987. Vaccination of Smith, T.S., Munn, E.A., 1990. Strategies for vaccination against gastro-
young lambs by means of a protein fraction extracted from adult intestinal nematodes. Rev. Sci. Tech. 9, 577–595.
Haemonchus contortus. Parasitology 94, 385–397. Underwood, R., Chiravuri, M., Lee, H., Schmitz, T., Kabcenell, A.K.,
Parkinson, J., Anthony, A., Wasmuth, J., Schmid, R., Hedley, A., Blaxter, M., Yardley, K., Huber, B.T., 1999. Sequence, purification, and cloning of
2004.PartiGene – constructing partial genomes.Bioinformatics 20, 1398–1404. an intracellular serine protease, quiescent cell proline dipeptidase. J.
Ranjit, N., Jones, M.K., Stenzel, D.J., Gasser, R.B., Loukas, A., 2006. A Biol. Chem. 274, 34053–34058.
survey of the intestinal transcriptomes of the hookworms, Necator Willadsen, P., Bird, P., Cobon, G.S., Hungerford, J., 1995. Commerciali-
americanus and Ancylostoma caninum, using tissues isolated by laser sation of a recombinant vaccine against Boophilus microplus. Parasi-
microdissection microscopy. Int. J. Parasitol. 36, 701–710. tology 110, S43–S50.