A total of 102 horses with necropsy-confirmed status of S. vulgaris infection were enrolled in the validation study. All necropsies were performed at either University of Kentucky in Lexington, Kentucky or East Tennessee Clinical Research (ETCR) in Rockwood, Tennessee.
All horses from University of Kentucky were naturally infected with mixed species of gastrointestinal helminth infections (n=31). They were enrolled from two main populations; a herd kept without anthelmintic intervention since 1979 , and a population of research horses maintained with four anthelmintic treatments a year. Naturally infected horses from Tennessee (n=23) participated in different anthelmintic drug trials at ETCR and all underwent necropsy. In total, samples from 54 naturally infected horses were collected.
Experimental infections were carried out for two different research projects with horses maintained at ECTR (n=48). Horses in group P (n=20) were treated once daily for five consecutive days with fenbendazole paste (10 mg/kg, Panacur, Merck, Summit, NJ, USA), housed in stables and inoculated on a single occasion with 600 embryonated P. equorum eggs obtained locally from naturally infected horses. After six months the horses were euthanatised and necropsied. Horses in group S (n=28) were acquired with unknown anthelmintic treatment history, and were therefore treated with a larvicidal regimen of moxidectin (0.4 μg/kg, Quest gel, Pfizer, Madison, NJ, USA) administered once, and fenbendazole paste (10 mg/kg, Panacur) once daily for five consecutive days during the week prior to enrolment in the study. They were subsequently infected with 5,000 cyathostomin third stage larvae (L3) five times weekly throughout the study. Five weeks into the study the horses started receiving 25 S. vulgaris L3 larvae once weekly until euthanasia and necropsy after 5.5 months .
At necropsy, the posterior aorta and cranial mesenteric artery and branches were recovered and evaluated for the presence of migrating S. vulgaris larvae in all horses.
The case-definition that served as the gold standard for classification of S. vulgaris positive horses was: horses with migratory tracts, one or more larvae, or evidence of previous infection. Migratory tracts were considered evidence of a current infection, and a circular area with raised surface, roughened or corrugated intima without evidence of active thrombosis was classified as evidence of previous infection. Horses with no migratory tracts, no larvae, and no signs of previous infection were classified as being S. vulgaris-negative.
The gastrointestinal and migrating parasites were enumerated as previously described .
A peripheral blood sample was collected from each horse (n=102) in a serum or serum-separator tube. Sera were separated by centrifugation, and duplicate aliquots were transferred to 2 mL cryovials and stored at −20°C until analysis. The serum was stored for up to 5 years prior to analysis.
The horses in this study had a mean age of 19.5 months (range 0.6 months – 22 years) with a median of 12 months of age. A total of 46 horses were female and 54 were males of which 12 were castrated. Data on age and gender was missing for two and one horse, respectively. The subset of horses 7 months and older had a mean age of 22.51 months (range 7 months – 22 years) with a median of 18 months of age with 44 female, 42 males of which 12 were castrated. Data from horses 3 months and younger were omitted; the next youngest foal was 7 months old. Breeds represented included: Tennessee Walking Horse, Paint, American Quarter Horse, Thoroughbred, Standardbred, Appaloosa, Shetland type ponies and mixed light breed.
All work involving the horses at UK and ETCR was approved by the Institutional Animal Care and Use Committee.
Migrating S. vulgaris larvae were collected by dissection of the abdominal aorta, celiac artery, cranial mesenteric artery and major branches recovered from horses at the University of Kentucky that were naturally infected with gastrointestinal parasites and where anthelmintic drugs have not been used since 1979 . The larvae were carefully lifted from the thrombus material and washed four times in 20 mL PBS (137 mM NaCl, 10 mM phosphate, 2.7 mM KCl, pH 7.4) to remove debris. The larvae were either placed in 2 mL cryotubes and snap frozen in liquid N2 for RNA extraction or used for collection of ES antigens.
Worms were collected from the caeca of horses using the necropsy technique previously described , adult S. vulgaris worms were identified by morphological criteria  and washed five times in 20 mL PBS to remove debris and used for collection of ES antigens.
Living washed adult worms were incubated in 5 mL RPMI-1640 (Life Technologies, Grand Island, NY, USA) with penicillin (100 IU/mL), streptomycin (100 IU/mL), amphotericin B (0.25 μg/mL) and a protease inhibitor cocktail (Protease inhibitor cocktail, P2714, Sigma-Aldrich, St. Louis, MO, USA) in a 5% CO2 incubator at 37°C. The medium was collected after 12 and 24 h and fresh medium was added. The protein concentration was analysed using the Bradford Protein quantitation assay (Pierce, Rockford, IL, USA) as per the manufacturer’s protocol. The ES antigen-rich medium was then dialysed against PBS at 4°C using a 3 mL 3.5 kDa molecular cut-off Slide-A-Lyzer® (Pierce, Rockford, IL, USA) according to the manufacturer’s protocol. Dialysed ES antigen was frozen and shipped to Cocalico Biologicals, Inc. (Reamstown, PA, USA) for production of hyperimmune serum. One rat was immunised with 250 μg adult S. vulgaris ES antigen mixed with Titermax® as adjuvant and given as six inoculations over a 61-day period. Due to background reactivity against E. coli proteins, the hyperimmune rat serum was pre-absorbed extensively against E. coli XL1-BLUE lysates .
Larval ES antigens were obtained from S. vulgaris larvae using 2 mL of RPMI-1640 (Life Technologies, Grand Island, NY, USA) with penicillin (100 IU/mL), streptomycin (100 IU/mL), amphotericin B (0.25 μg/mL) and a protease inhibitor cocktail (Protease inhibitor cocktail, P2714, Sigma-Aldrich, St. Louis, MO, USA) and incubating the larvae in a 5% CO2 incubator at 37°C. The medium was collected after 14 days during which time the dead larvae were removed daily with a sterile needle.
A total of 300 ng of S. vulgaris larval ES antigens was adjusted to 15 μL with water and mixed with 3 μL 5X sodium dodecyl sulphate-polyacrylamide gel electrophoresis (SDS-PAGE) sample buffer containing protease inhibitor cocktail. The same amount of S. vulgaris adult ES was prepared similarly. The ES mixtures were denatured at 95°C for 5 min, placed on ice for 5 min and resolved in individual wells under reducing conditions in a 12% polyacrylamide gel, where after the gel was stained by silver staining.
Construction and immunoscreening of a larval S. vulgaris cDNA library
RNA was extracted from N2 frozen S. vulgaris larvae using the Trizol reagent (Life Technologies, Grand Island, NY, USA) and poly(A)+ RNA (mRNA) was purified from total RNA using the NucleoTrap® mRNA-kit (Clontech, Mountain View, CA, USA) according to the manufacturer's protocol. A total of 400 ng purified mRNA was used as template to synthesise cDNA using the SMART cDNA library construction kit (Clontech, Mountain View, CA, USA), cloned into bacteriophage TriplEx2 lambda vector digested with NdeI and packaged using the Giga Pack III gold packaging extract (Agilent Technologies, Stratagene products division, La Jolla, CA, USA) as previously described for a larval cyathostomin cDNA library . The titre of the cDNA library was evaluated and the quality was assessed by analysing cDNA inserts in 30 randomly-picked plaques by PCR analysis. The PCR products were cleaned using the Wizard® SV Gel and PCR clean-up system (Promega, Madison, WI, USA) and sequenced at the University of Kentucky’s Advanced Genetic Technology Center.
The S. vulgaris larval cDNA library was immunoscreened as described by the manufacturer (Clontech, Mountain View, CA, USA). The primary immunoscreening was conducted on 100,000 cDNA clones. Plaque lifts were made onto nitrocellulose membranes (Fisher Scientific, Pittsburg, PA, USA) and the membranes were washed five times for 5 min in 25 mL TBST (100 mM Tris, 0.15 M NaCl and 0.05% Tween-20) and kept in TBST overnight at 4°C. The membranes were blocked with TBST + 1% gelatine for 1 h at room temperature (RT) and probed with preabsorbed hyperimmune rat serum at 1:100 in TBST after washing. As secondary antibodies, horseradish peroxidase (HRP)-conjugated goat anti-rat IgG (H+L) (Jackson ImmunoResearch, Inc. West Grove, PA, USA) were used at 1:10,000 in TBST after washing. The signal was developed using a chromogenic substrate (TMB stabilised substrate for HRP, Promega, Madison, WI, USA) after washing. The membranes were aligned with the agar plate, and the positive clones were picked using p200 pipette tips. The selected plaque was placed in 500 μL Lambda dilution buffer (0.1 M NaCl, 10 mM MgSO4, 3.5 mM Tris and 0.01% gelatin), vortexed for 30 s and kept at 4°C. The positive clones underwent secondary immunoscreening with the hyperimmune rat serum and rat-pre-immunisation serum as negative control to exclude false positive clones.
Clones isolated from the cDNA library were amplified by PCR using vector-specific primers and sequenced at the University of Kentucky’s Advanced Genetic Technology Center. The resulting sequences were used as queries in BLASTN searches against non-human, non-mouse nucleotide sequences and BLASTX searches against the non-redundant protein database from NCBI. The presence of a signal peptide was predicted using SignalP 4.0 , the presence of glycosylation sites were analysed using the ExPASy Prosite  and the presence of transmembrane domains and protein localisation were predicted using TMHMM 2.0 server . Pairwise alignment and phylogenetic comparison was performed using the ClustalOmega software on the EBI-server . Homologues from Cylicostephanus goldi (courtesy of Dr. Jane Hodgkinson, University of Liverpool) and P. equorum (courtesy of Dr. Georg von Samson-Himmelstjerna, University of Berlin) were obtained and the partial sequences compared by pairwise alignment using the ClustalOmega software.
Expression of recombinant protein
Primers were designed to amplify the coding sequence from an immune-reactive cDNA clone for subcloning into a pET22b(+) vector. The primer sequences were as follows: Forward: 5’-GATCCATATG CAAAATGGACCTCCACC-3' and reverse: 5'-GATCCTCGAG TCCCTTCATAGCGTCC-3' which incorporated the NdeI and XhoI restriction sites (underlined) to allow for unidirectional cloning. The PCR was performed using Verbatim High Fidelity Polymerase (Thermo Fisher Scientific, Pittsburg, PA, USA) and the amplified fragment was digested with NdeI and XhoI and ligated with the pET22b(+) plasmid. The resulting plasmid was transformed into E. coli BL21 cells and expression of the recombinant protein was induced by 1 mM isopropyl-β-D-thiogalactopyranoside (IPTG) at an OD600nm of 0.6 and the cells were incubated under agitation for 9 h at 30°C. The recombinant protein was purified on immobilised cobalt by affinity chromatography using BD TALON resin (Clontech, Mountain View, CA, USA) as per the manufacturer’s protocol for soluble proteins. Purified recombinant protein was stored in aliquots at −20°C.
Hyperimmune serum against recombinant protein
A total of 300 μg of recombinant protein was resolved in 12% polyacrylamide gels under reducing conditions, a strip of the gel was stained with GelCode Blue stain (Thermo Scientific, Pittsburg, PA, USA) to verify where the recombinant antigen travelled in the gel and the corresponding part of the gel containing the desired molecular size was cut out and shipped to Cocalico Biologicals, Inc. (Reamstown, PA, USA) and used to immunise a guinea pig with Freund’s complete adjuvant for initial immunisation and Freund’s incomplete adjuvant for subsequent boosters. The guinea pig was immunised by 6 inoculations of recombinant antigen over a period of 17 weeks.
Western blot analyses
S. vulgaris ES
For Western blot (WB) analysis of anti-adult S. vulgaris ES rat serum against both S. vulgaris larval and adult ES antigens, the antigens were mixed individually with SDS-PAGE sample buffer containing protein inhibitor cocktail under reducing conditions and resolved in two 1-well 12% polyacrylamide gels. The proteins were transferred to 0.45 μm nitrocellulose membranes by semidry electrophoretic transfer in Tris-glycine buffer. Membranes were blocked for 1 h in PBST (PBS + 0.05% Tween-20), and probed with pre-immunisation rat serum at 1:100 in PBST or hyperimmune rat serum raised against S. vulgaris adult ES at 1:100 in PBST. The signal was developed using TMB stabilised substrate for HRP (Promega, Madison, WI, USA).
For WB analysis of anti-recombinant protein guinea pig serum against larval ES, 4.4 μg S. vulgaris larval ES was resolved in a 1-well 12% polyacrylamide gels, transferred to a nitrocellulose membrane and blocked. The nitrocellulose membrane was cut and strips were placed in individual trays. The strips were probed with either 500 μL of guinea pig pre-immunisation serum at 1:500 in PBST or guinea pig hyperimmune serum anti-recombinant protein at 1:2,500 in PBST. HRP-conjugated goat anti-horse IgG(T) (Bethyl Laboratories, Inc., Montgomery, TX, USA), were used as secondary antibodies at 1:10,000 in PBST. The signal was developed using Supersignal WestPico chemiluminescent substrate (Pierce, Rockford, IL, USA).
Recombinant protein (233 ng) was resolved in each of two 1-well 12% polyacrylamide gels by SDS-PAGE under reducing conditions and transferred to nitrocellulose membranes. The membranes were blocked and one was placed in a multi-slot apparatus (Bio Rad, Hercules, CA, USA). The blot was probed with 500 μL horse serum samples (1:50 in PBST) in each slot. As secondary antibodies, HRP-conjugated goat anti-horse IgG(T) (Bethyl Laboratories, Inc., Montgomery, TX, USA), were used at 1:10,000 in PBST. The signal was developed using Supersignal WestPico chemiluminescent substrate.
Strips were cut from the other blocked blot and placed in individual trays. The strips were probed with either pre-immunisation guinea pig serum at 1:500 in PBST or anti-recombinant protein guinea pig serum at 1:2500 in PBST. As secondary antibody, HRP-conjugated donkey anti-guinea pig IgG (H+L) (Jackson ImmunoResearch, Inc., West Grove, PA, USA) were used at 1:10,000 in PBST. The signal was developed using Supersignal WestPico chemiluminescent substrate.
The indirect ELISA was optimised by sequential checkerboard titration to the following setup. Individual wells of a 96 well EIA/RIA plate (Costar® easy-wash, Corning Inc., Corning, NY, USA) were coated with 75 μL of recombinant protein diluted to 0.1 μg/mL in PBS and incubated overnight at 4°C. The wells were washed three times for 1 min with PBST and blocked for 1½ h at RT with 200 μL block solution (PBS containing 5% normal goat serum, 1% dry milk powder and 1% Tween-20). The wells were washed once and 75 μL of horse serum diluted 1:50 in diluent solution (block solution in PBST, 1:10) was added to individual wells in duplicates and incubated for 1 h at 37°C. Positive, negative and blank controls were included on each plate in duplicates. The wells were washed five times with PBST, and 75 μL of HRP-conjugated goat anti-horse IgG (H+L) (Jackson ImmunoResearch, Inc. West Grove, PA, USA) diluted 1:10,000 was added to each well and incubated for 1 h at 37°C. The wells were washed five times with PBST and incubated for 10 min at RT in the dark with 75 μL of RT 1-step Ultra TMB ELISA substrate (Thermo Scientific, Rockford, IL, USA) per well. The reactions were stopped with 75 μL 2 M H2SO4 per well, and the OD450nm determined using an E-max Precision Microplate Reader (Molecular Devices, Sunnyvale, CA, USA) with a photometric range of 0.000 to 4.000 OD and a resolution of 0.001 OD.
IgG subclass antibodies
For evaluation of the optimal diagnostic antibody target in serum from 15 horses with known S. vulgaris larval infection status, the level of antigen specific IgG subclasses IgGa, IgGb, IgGc and IgG(T) were evaluated using HRP-conjugated goat anti-horse IgGa, IgGb, IgGc and IgG(T) antibodies (Bethyl Laboratories, Inc., Montgomery, TX, USA) at 1:40,000 as secondary antibodies as per the manufacturer’s recommendation alongside the HRP-conjugated goat anti-horse IgG (H+L) secondary antibody as described in the ELISA setup.
After identifying the best antibody target, the secondary antibody dilution for the ELISA was optimised by checkerboard titration.
Diagnostic accuracy of ELISA
These ELISAs were performed in June and July, 2012. Serum samples from all horses (n=102) and horses seven months and older (n=86) were evaluated for the level of antigen specific IgG and IgG(T) antibodies using the optimised ELISA in separate assays. The intra-assay variability of the ELISA was calculated from duplicate measurements from all horses. The inter-assay variability was calculated from the positive and negative controls included in each assay as well as specifically for three horses that were selected on the basis of their rSvSXP-specific IgG(T) OD450nm (OD value given in parentheses). These animals served as a high positive (2.708), an intermediate positive (1.278) and a negative (0.022). From each sample, a volume of 1,000 μL of serum diluted 1:50 in diluent solution was prepared to test triplicate samples on each of four sequential days to evaluate inter-assay variability as a normalised value, percentage of a positive control (PP), as previously described .
The statistical program R, version 2.12  was used to generate graphs and perform statistical analyses. For all statistical analyses, a P-value less than or equal to 0.05 was considered significant.
Evaluation of IgG subclasses
The non-parametric Wilcoxon rank sum test was used to compare IgG levels within each IgG subclass between S. vulgaris infected and uninfected horses.
Intra- and inter-assay variability
The intra- and inter-assay % coefficient of variability (% CV) was calculated for each series of assays for each of the secondary antibodies. An intra-assay % CV below 10% and an inter-assay % CV below 20% were considered acceptable .
Receiver operator characteristics (ROC) curve analysis
ROC curve analyses were performed using the software package Epi for R  for two sets of horses: all horses (n=102) and horses seven months or older (n=86). The reason for excluding foals in the second analysis was due to observations of high larval counts and corresponding low PP-values in some of the younger foals. After excluding horses younger than three months of age the remaining horses were all seven months or older. The optimal cut-off was determined on the basis of the ROC curve analysis.
For the two sets of horses: all horses and horses 7 months and older, the software package EpiR for R  was used to calculate the diagnostic accuracy values; sensitivity, specificity, odds ratio, positive likelihood ratio (LR) and negative LR with corresponding 95% confidence intervals.
Correlation of arterial S. vulgaris larvae and rSvSXP-specific antibodies
The correlation between the number of S. vulgaris larvae in the cranial mesenteric artery and branches and the level of rSvSXP specific IgG(T)-antibodies expressed as the normalised PP-value was evaluated by the Spearman correlation test from the software package fBasics for R .
All horses (n=102) were categorised by the number of larvae present in the CMA and branches in the following five groups: Group 0: No larvae, migratory tracts or evidence of previous infection (n=42); group 1: No larvae, but migratory tracts or evidence of previous infection (n=16); group 2: 1–5 larvae (n=16); group 3: 6–25 larvae (n=17); and group 4: above 25 larvae (range: 25–292) (n=11).
A subset of the horses, horses 7 months or older (n=86), was categorised by the number of S. vulgaris larvae in the CMA and branches in the following five groups: Group 0: No larvae, migratory tracts or evidence of previous infection (n=28); group 1: Migratory tracts or evidence of previous infection (n=16); group 2: 1–5 larvae (n=16); group 3: 6–25 larvae (n=17); and group 4: above 25 larvae (range: 25–200) (n=9).
The relationship between the different groups and the PP values was evaluated graphically for both sets of horses. A Kruskal-Wallis test was performed to evaluate if there was a significant difference between groups and a multiple comparison test after Kruskal-Wallis performed to identify the significantly different groups using the statistical software package pgirmess for R . This was performed on both sets of horses.