Edinburgh Research Explorer

Differential recognition patterns of Schistosoma haematobiumadult worm antigens by the human antibodies IgA, IgE, IgG1 andIgG4

Citation for published version:Mutapi, F, Bourke, C, Harcus, Y, Midzi, N, Mduluza, T, Turner, CM, Burchmore, R & Maizels, RM 2011,'Differential recognition patterns of Schistosoma haematobium adult worm antigens by the humanantibodies IgA, IgE, IgG1 and IgG4', Parasite Immunology, vol. 33, no. 3, pp. 181-192.https://doi.org/10.1111/j.1365-3024.2010.01270.x

Digital Object Identifier (DOI):10.1111/j.1365-3024.2010.01270.x

Link:Link to publication record in Edinburgh Research Explorer

Document Version:Publisher's PDF, also known as Version of record

Published In:Parasite Immunology

Publisher Rights Statement:Open Access.

General rightsCopyright for the publications made accessible via the Edinburgh Research Explorer is retained by the author(s)and / or other copyright owners and it is a condition of accessing these publications that users recognise andabide by the legal requirements associated with these rights.

Take down policyThe University of Edinburgh has made every reasonable effort to ensure that Edinburgh Research Explorercontent complies with UK legislation. If you believe that the public display of this file breaches copyright pleasecontact openaccess@ed.ac.uk providing details, and we will remove access to the work immediately andinvestigate your claim.

Download date: 02. Sep. 2020

Differential recognition patterns of Schistosoma haematobium adult

worm antigens by the human antibodies IgA, IgE, IgG1 and IgG4

F. MUTAPI,1 C. BOURKE,1 Y. HARCUS,1 N. MIDZI,2 T. MDULUZA,3 C. M. TURNER,4 R. BURCHMORE5 & R. M. MAIZELS1

1Ashworth Laboratories, Institute of Immunology & Infection Research, School of Biological Sciences, University of Edinburgh,Edinburgh, UK, 2National Institute of Health Research, Harare, Zimbabwe, 3Department of Biochemistry, University of Zimbabwe,Harare, Zimbabwe, 4Glasgow Biomedical Research Centre, Faculty of Biomedical and Life Sciences, University of Glasgow, Glasgow, UK,5Division of Infection and Immunity, Institute of Biomedical Life Sciences, University of Glasgow, Glasgow, UK

SUMMARY

Schistosoma haematobium antigen recognition profiles of thehuman isotypes IgA, IgE, IgG1 and IgG4 were compared byimage analysis of western blots. Adult worm antigens sepa-rated by two-dimensional gel electrophoresis were probed withpooled sera from Zimbabweans resident in a S. haematobiumendemic area, followed by the identification of individual anti-genic parasite proteins using mass spectrometry. Overall,IgG1 reacted with the largest number of antigens, followed byIgE and IgA which detected the same number, while IgG4detected the fewest antigens. IgE recognized all antigensreactive with IgG4 as well as an additional four antigens, anisoform of 28-kDa GST, phosphoglycerate kinase, actin 1 andcalreticulin. IgG1 additionally recognized fatty acid–bindingprotein, triose-phosphate isomerase and heat shock protein70, which were not recognized by IgA. Recognition patternsvaried between some isoforms, e.g. the two fructose 1-6-bis-phosphate aldolase isoforms were differentially recognized byIgA and IgG1. Although the majority of S. haematobiumadult worm antigens are recognized by all of the four isotypes,there are clear restrictions in antibody recognition for someantigens. This may partly explain differences observed in iso-type dynamics at a population level. Differential recognitionpatterns for some isoforms indicated in the study have poten-tial importance for vaccine development.

Keywords isotype, proteomics, recognition patterns, schisto-somiasis

INTRODUCTION

Schistosomiasis is a major human parasitic disease intropical and subtropical countries in Africa, the MiddleEast and South America (1). Urinary schistosomiasiscaused by Schistosoma haematobium affects over 100 mil-lion people, and a recent survey in sub-Saharan Africaindicated that 70 million individuals had experienced hae-maturia and 32 million dysuria associated with S. haemat-obium infection (http://www.who.int/vaccine_research/diseases/soa_parasitic/en/index5.html Accessed 16th Sep-tember 2010). Furthermore, it was estimated that 18 mil-lion people suffered schistosome-related bladder wallpathology and 10 million suffered hydronephrosis.

Epidemiological studies in endemic human populationshave shown that schistosome prevalence and intensity lev-els rise to peak in childhood (around ages 9–14) (2) anddecline thereafter, so that in any endemic population, chil-dren carry the heaviest infection levels while adults carrylittle or no infection. This pattern has been taken to reflectthe development of immune-meditated resistance to infec-tion ⁄ re-infection (2,3) and suggests that protective immuneresponses develop slowly as a result of cumulative expo-sure to parasite antigens (4). Early serum transfer studiesin the mouse model (5–8) as well as human immuno-epi-demiological studies have shown that schistosome-specificantibody responses are associated with protection to infec-tion and re-infection (9–13). Despite the demonstrationthat antibody-mediated responses can protect againstschistosome infection in experimental models, currenthuman schistosome vaccine research based on antibody-mediated protection has stalled with the failure of many ofthe vaccine candidate antigens to enter Phase III clinicaltrials (14). Limitations in our current understanding of the

Correspondence: Francisca Mutapi, Institute for Immunology andInfection Research, School of Biological Sciences, University ofEdinburgh, Ashworth Laboratories, King’s Buildings, West MainsRoad, Edinburgh EH93JT, UK (e-mail: f.mutapi@ed.ac.uk).Disclosures: None.Re-use of this article is permitted in accordance with the Terms andConditions set out at http://wileyonlinelibrary.com/onlineopen#OnlineOpen_Terms.Received: 2 July 2010Accepted for publication: 15 October 2010

Parasite Immunology, 2011, 33, 181–192 DOI: 10.1111/j.1365-3024.2010.01270.x

� 2011 Blackwell Publishing Ltd 181

development of protective anti-schistosome responsesagainst specific antigenic proteins may be contributingto the slow development of effective anti-schistosomevaccines. Previous studies characterizing immune responsesto schistosome proteins have relied on recombinant pro-teins expressed in bacterial systems (15–17). However,there are restrictions associated with this approach; forexample, it cannot detect immunogenic epitopes arisingfrom post-translational modifications. Recent develop-ments in two-dimensional gel electrophoresis and proteo-mics have the potential to overcome some of theserestrictions because crude parasite antigen preparationscan be separated into individual protein spots that canthen be identified by mass spectrometry analysis andmatched to newly available genomic databases.

The nature of the anti-schistosome antibody responseto crude antigen preparations has been studied in termsof the different antibody isotypes involved and their rela-tionship to each other, to host age and to schistosomeinfection level by ourselves and others (for example, see(4,11,18–21). Relatively few individual target antigenshave been analysed in the context of selective antibodyisotype recognition (22). IgA, IgE, IgG1 and IgG4 are ofparticular interest as they have been implicated in thedevelopment of protective human antischistosome immu-nity. Hagan et al. demonstrated a significant relationshipbetween antischistosome IgE and IgG4 responses inS. haematobium infections. Furthermore, they reported adecrease in IgG4 accompanied by an increase in IgE withhost age, which was associated with resistance to re-infec-tion after treatment (10). We have previously reportedthat children exposed to S. haematobium produced a pre-dominantly IgA response against egg and adult wormantigens which was gradually replaced by an IgG1response in adulthood (4). These studies were based onimmune responses to crude antigens containing a hetero-geneous mixture of proteins and thus did not provideany information on the recognition patterns of individualantigens. Therefore, the antigenic source of variation inisotype-specific responses to S. haematobium is unknown.Differences may arise from the recognition of the sameantigen by different antibody classes, from the recogni-tion of different antigens or a combination of both.Identifying the individual antigens recognized by the dif-ferent antibody isotypes in crude adult worm antigenswould help resolve the cause(s) of some of these differ-ences. In this study, we used proteomic approaches incombination with two-dimensional western blotting todetermine which adult worm antigens are recognized byspecific antibody isotypes ⁄ subclasses from schistosome-infected ⁄ exposed individuals. Thus, we identify the iso-type-specific recognition patterns of individual parasite

proteins within the crude adult S. haematobium proteomefor the first time. We identified antigens recognized byIgA, IgE, IgG1 and IgG4, focussing our attention onWHO vaccine candidate antigens (23). Characterizing theantigens recognized by each antibody class will bothimprove our understanding of naturally acquired schisto-some-specific immunity and inform vaccine developmentwhere the aim may be to stimulate an isotype-specificantibody response to a single recombinant antigen ratherthan a heterogeneous mixture of peptides. In this study,we determined the reactivity of sera from 7- to 18-year-olds, the age group where infection levels are changingfrom rising, peaking and declining in this population(24,25) to capture the range of immune reactivities as thedynamics of infection change. The analyses were con-ducted on pretreatment sera to study natural immunereactivities before enhancement of antibody reactivitywith antihelminthic treatment, which we have previouslydemonstrated in a different subgroup from the same pop-ulation (26).

MATERIALS AND METHODS

Study area

The study was conducted in two rural villages in theMashonaland East Province of Zimbabwe (31�30¢E;17�45¢S) where S. haematobium is endemic. The study areais described in detail elsewhere (26). The study receivedethical and institutional approval from the MedicalResearch Council of Zimbabwe and the University ofZimbabwe, respectively. Permission to conduct the work inthis province was obtained from the Provincial MedicalDirector. The villages were selected because health surveysconducted regularly by the Ministry of Health and ChildWelfare in the region showed little or no infection withother helminths and a low S. mansoni prevalence (<5%).The selected villages had not been included in theNational Schistosome Control Programme and thereforehad not received treatment for schistosomiasis or otherhelminth infections. The main activity in these villages issubsistence farming. Drinking water is collected from openwells while bathing and washing are conducted in twomain rivers in the villages. Most families maintain a gar-den located near the river where water is collected forwatering the crops, and the schools surveyed were all inclose proximity to rivers.

Study subjects

The study participants are described in detail elsewhere(26). Briefly, only permanent inhabitants of the villages

F. Mutapi et al. Parasite Immunology

182 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192

who had never been treated for any helminth infectionwere eligible for inclusion in the study. Informed consentwas obtained from all participants prior to enrolment intothe study. Following explanation of the project aims andprocedures to the community, school children and theirteachers, an initial parasitology (stool and urine samples)and serology (blood sample) survey was conductedamongst all compliant participants. A questionnaire sur-vey confirmed that on average, all participants frequentedwater contact sites at least four times per week and thatfrequency of water contacts at the various sites was notsignificantly different within the age range included in thisstudy (data not shown).

Parasitology samples (at least two urine and two stoolsamples collected on 3 consecutive days) and 20 mL ofvenous blood were collected from each participant. Stoolsamples were processed following the Kato-Katz procedure(27) to detect S. mansoni eggs and other intestinal helm-inths, while the urine filtration method (28) was used todetect S. haematobium eggs in urine samples. After collec-tion of the samples, all participants were offered treatmentwith the recommended dose of praziquantel, 40 mg ⁄ kgbody weight. To be included in the cohort, participantshad to meet all of the following criteria: (1) have providedat least two urine and two stool samples on consecutivedays for the detection of helminth parasites; (2) have givena blood sample for serological assays; (3) be negative forintestinal helminths and S. mansoni. In practice, all peoplemeeting criteria 1 and 2 were egg negative for intestinalhelminths and S. mansoni, and so no participants wereexcluded on this third criterion. A total of 215 people(110 men and 105 women) aged from 5 to 18 years met allcriteria. Serum samples obtained from 20 mL of venousblood from each participant were frozen and stored induplicate at )20�C in the field and transferred to a –80 �Cfreezer in the laboratory. One complete set of the sampleswas subsequently transported frozen from Zimbabwe tothe United Kingdom, stored at )80�C and defrosted forthe first time for use in this study. A serum pool for use asa positive control in the western blot assays was madeusing equal volumes (50 lL) of sera from each of the 215participants irrespective of their schistosome infection sta-tus and frozen in aliquots that were defrosted only oncefor the immunoassays.

Parasite antigens

Adult worm antigens were used in this study as previousstudies have shown that adult S. haematobium parasitessuffer immune attrition and are targeted by host antibod-ies (29). Soluble S. haematobium adult worm antigens(SWAP) were obtained freeze-dried, from the Theodor

Bilharz Institute in Egypt, and reconstituted as previouslydescribed (26). These were used in the subsequent gel elec-trophoresis and immunoassays.

Antibody enzyme-linked immunosorbent assays

Circulating levels of IgA, IgE, IgG1 and IgG4 directedagainst the S. haematobium adult worm antigens weredetected by indirect enzyme-linked immunosorbent assays(ELISA) as previously described (4) using 5 lg ⁄ mL of anti-gen to coat plates, sera diluted at 1 : 50 and the secondaryhorseradish peroxidase–conjugated secondary antibodiesIgE (Sigma A9667, Sigma Aldrich, Gillingham, UK),IgG1(The Binding Site, AP006) and IgG4 (Ap009), all at1 : 1000 dilution, and IgA (P0216; Dako, Cambridgeshire,UK) which was diluted at 1 : 2000. Negative control serafrom five European volunteers who have never travelled totropical countries and therefore presumed not previouslyexposed to schistosomiasis were also included in the assaysto generate reactivity cut-off points.

Gel electrophoresis

The soluble adult worm antigen preparation was separatedinto constituent proteins by two-dimensional gel electro-phoresis on 13-cm gels as previously described (26) to gen-erate a reference gel used for identifying proteins (200 lgof SWAP used) and the second (100 lg of SWAP used)for western blotting to determine which antigens were rec-ognized by the sera following a previously established pro-tocol (4).

Determining sera and secondary antibody dilutions forwestern blotting

To minimize differences in sensitivity and specificity of thedifferent assays arising from technical differences, assayswere optimized by titration assays. A pool of sera from allof the Zimbabwean participants was used as a positivecontrol, while a pool of sera from five European volun-teers (same as the ones used in the ELISA assays above)was used as a negative control. The dilutions used for thesera and secondary antibodies (IgA, IgE, IgG1 and IgG4)were determined by a combination of enzyme-linkedimmunosorbent assay (ELISA) titration assays and trialwestern blots. First, ELISAs were conducted (varying bothsecondary antibody (1 : 500–1 : 2000) and sera dilutions(1 : 10–1 : 11280; see example in Figure S1) with 5 lg ⁄ mLof adult worm antigen. This informed the subsequent trialwestern blots conducted using sera diluted at 1 : 200 or1 : 400 to allow enough sample for all replicate assays andto test a range of potential secondary antibody dilutions

Volume 33, Number 3, March 2011 Isotype-specific reactivity of helminth antigens

� 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192 183

(1 : 1000, 1 : 2000, 1 : 4000) on a blot of 100 lg of two-dimensional electrophoresis-separated SWAP. From thesedata, we determined dilutions that gave the most consis-tent replicates and optimal sensitivity and specificity foreach antibody isotype. Higher dilutions of secondary anti-bodies resulted in the detection of fewer antigenic spots(see Figure S2), while reducing the secondary antibodydilution factor from 1 : 1000 (IgE, IgG1, IgG4) or1 : 2000 (IgA) resulted in higher background rather thanan increase in the number of spots detected. Furthermore,sera and secondary antibody dilutions used were consis-tent with those we have previously used in ELISA (24,30)and western blot assays (16), enabling our results to berelated to previous studies. The finalized protocol isdescribed elsewhere.

Western blotting

Western blotting was used to determine the proteins reac-tive with the isotypes IgE, IgG4, IgA and IgG1 in serafrom the participants. To achieve this, proteins were trans-ferred from the gel onto nitrocellulose membrane as previ-ously described (26). Blots were run in pairs (onegel ⁄ membrane) to probe for two antibody isotypes at atime in parallel assays. To verify efficient transfer ofSWAP proteins, membranes were stained with Ponceau Ssolution (Sigma) and then blocked for 1 h at room tem-perature in TBS blocking buffer [Pierce, (Thermo Scienti-fic), Surrey, UK] with 0.05% Tween 20. This was followedby 2 · 10 min washes with TBS ⁄ 0.05% Tween 20 ⁄ 0.5%Triton-X 100 (TBS ⁄ TT) (used for all washes). The pool ofpositive control sera (diluted at 1 : 200 in TBS blockingbuffer with 0.02% Tween 20) was added to each of themembranes and incubated overnight at 4�C followed by3 · 10 min washes. Horseradish peroxidise–conjugated sec-ondary antibodies diluted in TBS blocking buffer ⁄ 0.05%Tween 20 were then added to respective membranesdiluted at 1 : 1000 for IgE and 1 : 2000 for IgA. Mem-branes were subsequently incubated at room temperaturefor 1 h, followed by 4 · 10 min washes in TBS ⁄ TT and1 · 10 min washes in TBS alone. The proteins were visual-ized using the chemiluminescence product ECL Plus (GEHealthcare, Buckinghamshire, UK) according to the man-ufacturer’s instructions. Films exposed to the blots for 5 swere developed and spots matched to the Coomassie blue–stained reference gel using the image analysis softwareImage master. Following visualization, the membraneswere stripped of the ECL reagent, secondary antibody andsera following the manufacturer’s protocol. The samemembranes were then re-probed using the same pool ofsera. The IgE membrane was re-probed with IgG4 dilutedat 1 : 1000, while the IgA membrane was re-probed with

IgG1 diluted at 1 : 1000. A previous assay showed thatthe stripping procedure removed all proteins not directlybound to the nitrocellulose membrane as indicated by thelack of ECL reactivity with a stripped membrane. Thisprocedure did not remove any of the parasite proteins asevidenced by probing the same membrane with threeserum samples successively, i.e. sera from the endemicpopulation, followed by negative control sera (a pool fromfive Europeans who had never travelled to tropicalregions) and then by the same sera from the endemic pop-ulation. An example of a membrane blotted with negativecontrol sera and the IgG4-specific secondary antibody isshown in Figure S3. Gel electrophoresis and western blot-ting were repeated twice using the secondary antibodies indifferent order on the same membrane, i.e. IgG1 followedby IgA and IgG4 followed by IgE to confirm the recogni-tion patterns obtained.

Image analysis and mass spectrometry

Images from the western blots were electronically scannedfor image analysis. Spots on these blots were detected bypixel analysis and matched across the different blots foreach isotype using the 2-D gel image analysis softwareImageMaster from GE Healthcare. All electronic spotidentifications and predicted matches were manually veri-fied via independent visualization by two researchers.Spots on the Coomassie blue–stained gel, which matchedto those on the western blots, were excised, digested withtrypsin and analysed by mass spectrometry analysis aspreviously described (26). Data thus obtained were submit-ted to an MSMS ion search via the Mascot search engine(Matrix Science), searching both locally established data-bases for S. mansoni EST sequence and the current nonre-dundant NCBI database.

Data analysis

The ratios of IgG1 ⁄ IgA and IgE ⁄ IgG4 were calculated foreach individual from ELISA data for each individual. Par-tial correlations were conducted to determine the associa-tion between the ratios and schistosome infection intensityallowing for residential village, age and sex. Significancelevel for the statistics tests was set at P < 0.05. For allanalyses of antigenic spots, if the recognition intensity ofthe spot after subtraction of the background intensity was0 pixels, then the spot was designated ‘not recognized’ bythe isotype, and if it was greater than 0, then the antigenicspot was designated as ‘recognized’. To allow comparisonsof the intensity of antigen recognition by the different iso-types, the detected spots were categorized by their recogni-tion intensity into five groups; group 1 comprised all

F. Mutapi et al. Parasite Immunology

184 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192

antigens not recognized by the pooled sera and groups2–5 were based on the interquartile ranges for each isotypeusing the mean intensity from the three replicate blots perisotype. The interquartile ranges were calculated from theantigens using a spot detected from pixel analysis andmanual verification. These ranges were calculated for eachisotype separately so that the spot intensities werestandardized within the gel. The intensity categories ratherthan absolute values could then be compared acrossthe different gels. This approach was used as it does notrequire any spots to be recognized equally by the fourisotypes for calibrating recognition intensity across allisotypes.

RESULTS

Levels of IgA, IgE, IgG1 and IgG4 detected in individuals

The prevalence of S. haematobium infection in the partic-ipants was 57%, and the corresponding mean infectionintensity was 36 eggs ⁄ 10 mL urine (SE of the mean =5.7), with individual egg counts ranging from 0 to 676eggs ⁄ 10 mL urine. Individual infection intensity followedthe typical age infection profile for schistosome infec-tions, rising with age to peak in childhood and thendeclining thereafter as shown in Figure 1a. The levels ofthe four antibody isotypes IgA, IgE, IgG1 and IgG4directed against adult worm antigen were measured foreach individual who was included in the serum pool. Aparticipant was considered reactive for an isotype if theirtitre was above the cut-off point of mean +2 standarddeviations of the negative control sera. Based on thiscut-off point, the least prevalent schistosome-specific iso-type was IgA being detected in 31% of the participantsand then IgE detected in 50% of the participants. IgG1was detected in 58% of the participants, while IgG4 wasthe most prevalent response, detected in 74% of the par-ticipants. Levels of the isotypes were plotted by agegroups in Figure 1b.

Earlier studies have shown that higher ratios ofIgG1 ⁄ IgA and IgE ⁄ IgG4 are associated with lower infec-tion levels (10,31). Therefore, to determine whether therelationship between S. haematobium infection and anti-body responses in this study was consistent with thosereported from other earlier studies, the ratios of IgG1 ⁄ IgAand IgE ⁄ IgG4 were calculated and plotted against schisto-some infection intensities. The ratio of IgE ⁄ IgG4 showeda significant inverse relationship with schistosome infec-tion intensity (r = )0.21, P = 0.002, df = 210), so thatpeople with high values for these ratios carried little or noschistosome infection (Figure 1c). The IgA ⁄ IgG1 ratio didnot show a significant association with infection intensity

(r = 0.11, P = 0.10, df = 210). The subsequent westernblotting analyses were performed to determine whetherthese different isotypes recognized the same individualnative antigens.

Antigens recognized by the different isotypes

The potential repertoire of schistosome antigens recog-nized on two-dimensional blots of adult parasite extractwas first defined with a highly reactive positive post-treat-ment serum pool made from a subgroup of this popula-tion previously used to investigate the effects ofpraziquantel treatment on schistosome-specific responses(26), which was probed with a secondary antibody combi-nation reactive to IgA ⁄ IgG ⁄ IgM. This serum pool reactedwith a total of 71 spots compared to negative nonendemicsera that reacted with none of the protein spots. The 71proteins were identified and characterized in our previousstudy (26) which gives their predicted molecular weights,isoelectric point (pI), Mascot output statistic and acces-sion numbers from NCBI.

Compared to the IgA ⁄ IgG ⁄ IgM positive control, no sin-gle isotype in the serum pool recognized all of the 71 anti-genic spots. The differences in recognition of the antigenicspots between the isotypes were both qualitative andquantitative. IgG1 recognized the largest number of spots(n = 45) followed by IgA and IgE which both recognized39 spots, with IgG4 recognizing the fewest antigenicspots (n = 35) as shown in Figure 2 and Table 1. The moststrongly recognized antigens by all subclasses were the twoisoforms of glyceraldehyde-3-phosphate, some isoforms ofheat shock protein (HSP) 70 as well as some proteins stillto be identified (spots 60 and 61). The antigenic spots 2(protein yet to be indentified), 6 (triose-phosphate isomer-ase) and 15 (yet to be identified) were strongly recognizedby IgG1 but were not detected by any of the other iso-types. There were two isoforms of fructose-1-6-bis-phos-phate aldolase, one reactive with IgA (antigenic spot 25 inFigure 2a) and one reactive with IgG1 (antigenic spot 23in Figure 2a). Some antigens were not recognized at allbyany of the 4 isotypes tested; these included the muscleproteins myosin (heavy and light chain), tropomyosinand paramyosin, HSP 60 and some isoforms of enolaseand immunophilin. Some antigenic spots such as thosebetween spots 22 and 52 in Figure 2 were strongly recog-nized by the sera but were in such low abundance in theCoomassie reference gel that they could not be isolatedand identified by mass spectrometry as reported in previ-ous studies (26,32).

To quantify the relationships between protein spot recog-nition by different isotypes, the intensity of recognition ofeach antigenic spot by each isotype was categorized by

Volume 33, Number 3, March 2011 Isotype-specific reactivity of helminth antigens

� 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192 185

interquartile range (to reduce the effects of differencesarising because of intra-isotype variation in detectionthresholds) as shown in Table 1. The recognition intensityof the spots was positively correlated so that for most spots,those intensely recognized by IgA were also intensity recog-

nized by IgG1 (Figure 3a). Thirty-six of the 39 antigenicspots recognized by IgA were also recognized by IgG1(Table 1). However, IgG1 recognized an additional ninespots indicated in Figure 3a (numbered spots on the y axis).These resolved to the antigens fatty acid–binding protein

(a) (c)

(b)

Figure 1 Description of the individual study participants partitioned by age group. (a) Infection intensity for each individual calculated asa mean of eggs ⁄ 10 mL of urine from at least two urine samples collected on consecutive days. (b) Antibody levels measured as optical den-sities for each individual (c) The relationship between the ratios of IgG1 ⁄ IgA and IgE ⁄ IgG4 calculated from the values in (1B) above andschistosome infection intensity for each individual participant. These show a significant negative association between the ratio IgE ⁄ IgG4and infection intensity and no significant association between the IgG1 ⁄ IgA ratio and infection intensity.

F. Mutapi et al. Parasite Immunology

186 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192

(Sm14), phosphoglycerate kinase, triose-phosphate isomer-ase, an isoform of fructose 1-6-bis-phosphate aldolase, HSP70, calreticulin, and two proteins yet to be identified (spots2 and 15). Conversely, IgA recognized three antigenic spotsnot recognized by IgG1 (numbered spots on the x-axis ofFigure 3a). These were an isoform of enolase, an isoform offructose 1-6-bis-phosphate aldolase, and a protein (spot 50)yet to be identified. These were also not recognized byIgG4 (Figure 3d) or IgE (Figure 3f).

Similar patterns of correlations between isotypes, butwith some antigens being exclusively or largely recognizedby only one isotype, were seen in all other pairwise com-parisons (Figure 3). IgE recognized all 34 antigenic spotsrecognized by IgG4 as shown in Table 1 and Figure 3b. Inaddition, it also recognized an isoform of 28-kDa glutathi-one-S-transferase, phosphoglycerate kinase, actin 1, calreti-culin and one protein (spot 16) still to be identified.

Of all recognized antigens, only triose-phosphateisomerase (spot 6 in Figure 2 and Table 1), isoforms offructose 1-6-bis-phosphate aldolase (spots 23 and 25), eno-lase (spot 26), HSP70 (spot 29) and spots 2 and 15 whichare proteins still to be identified were recognized by a sin-gle isotype. The isoforms of fructose-1-6-bis-phosphatealdolase were reactive with either IgA (antigenic spot 25in Figure 2a) or IgG1 (antigenic spot 23 in Figure 2a) butnot with both. IgG1 recognized all of the antigenic spotsrecognized by IgG4 (Table 1, Figure 3c) and IgE (Table 1,Figure 3e).

Of the current World Health Organisation’s 10 schisto-some vaccine candidates, only three were recognized by allof the isotypes assayed. These were an isoform of 28-kDaglutathione-S-transferase (spot 8), both isoforms of glyce-ladeyde-3-phosphate (spots 21 and 22) and several iso-forms of actin-binding ⁄ filamin-like protein (spots 66–69).

MW kDa

62 63 64 65 50 49 48

200

150

70

(a)

56 53 54

31 32 33 34

41 42 43 44 45 47 46

26 25 22

52

21

52

78

18

50

37

20

15

10

IgA

MW kDa

69 68 67 66

38 39 4854

65646362616059

495356

41

200

150

70

30

31 32 33 34

2329

18 17

52 2122

464745444342 50

37

89 720

25 25

15

10

IgG1MW kDa

200

(b)

66676869 150

70

50

37

31

5960 62

38

39

61 63

6465 49 48

47

52 21

30 32 33 34

22

464543

41

42

5356

44

54

25

20

161718

8 7

IgEpH 3pH 10

15

10

MW kDa

200

150

70

50

37

39

31 32 33

41 4342 44 47 46

4845

6463

54

62

5356

61

60

59

25

208

52 2122

9

18

IgG4pH 3pH 10

15

101

Figure 2 Antigen recognition patterns of the serum pool compared by two-dimensional western blotting. (a) IgA vs. IgG1. (b) IgE vs.IgG4.

Volume 33, Number 3, March 2011 Isotype-specific reactivity of helminth antigens

� 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192 187

Table 1 The table gives identities of the antigenic spots in Figure 3, relating the spot number to its protein identity from the peptidesearches and recognition intensity by each of the isotypes

F. Mutapi et al. Parasite Immunology

188 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192

The vaccine candidate paramyosin was not recognized byany of the antibodies assayed.

DISCUSSION

Passive transfer studies in animal models show that anti-bodies can protect against schistosomiasis (5–8,33). In par-allel, human immuno-epidemiology studies have shownthat some antibody classes are associated with low levelsof infection in endemic populations as well as low re-infec-tion rates following treatment with antihelminthic drugs(10,12,13,34–36). However, all previous studies to datehave focused on describing antibody responses either toindividual recombinant antigens or to uncharacterized‘whole worm’ antigen preparations. We have previouslydefined native schistosome adult proteins recognized bythe total IgG fraction of human serum from samples col-lected from this same population (26). In this study, wecompare the reactivity of different isotypes to the solublecomponent of adult worm crude antigen, separated intoits constituent defined antigens, to answer two questions.First, which antigens are recognized by the different anti-body iostypes IgA, IgE, IgG1 and IgG4? Second, do dif-ferent isotypes recognize the same antigens? It is possiblethat observed differences in recognition intensity reflectnot only the biological (functional) differences between theisotypes but also differences in the sensitivity and specific-

ity of the isotypes, so that some antigens would be belowthe detection limit of the assays. However, we havecarefully calibrated the dilutions of sera and secondaryantibodies to enable optimal comparison and minimizethis possibility. We have used three replicates of each blotto verify the patterns and ratios of recognition intensity ofeach protein spot that we observed for each of the fourisotypes. Thus, we have shown that some antigens fromwithin the adult S. haematobium proteome are preferen-tially recognized by certain isotypes. For example, oneisoform of the antigen fructose-1-bis-phosphate aldolasewas intensely recognized by IgA but not detected by IgE,IgG1 or IgG4. Furthermore, comparing recognition inten-sity between isotypes using quartiles calculated within eachsubclass (Table 1), rather than absolute values (Figure 3),corrected for some intra-isotype variation in sensitivity.The study was designed to extend our understanding ofthe individual antigens detected in immunoassays routinelyconducted by ELISA using crude antigens. In these formerassays, protocols are designed to optimize sensitivity andspecificity as carried out here to reduce the effects of dif-ferences in detection thresholds of the different isotypes.

The four isotypes we investigated detected a total of 48antigenic spots of a possible 71 spots detected with totalIgG antibodies previously described by screening adultworm antigens with post-treatment sera (26). The sera usedin this current study were from 7- to 18-year-olds before

Table 1 (Continued)

The WHO-listed schistosome vaccine candidates are highlighted in bold. The intensity of recognition is coded from no colour (antigenic spotnot detected), very light (1st quartile), (excluding the undetected spots in former category, i.e. recognition intensity >0), light (2nd quartile),medium (3rd quartile) and darkest (4th quartile). The interquartile ranges were calculated separately for each isotype.

Volume 33, Number 3, March 2011 Isotype-specific reactivity of helminth antigens

� 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192 189

enhancement of antibody reactivity with antihelminthictreatment (26). Our previous study involved older people,i.e. up to 42 years old, and included the reactivities ofthe additional IgG subclasses IgG2, IgG3 which are reac-tive against some of the schistosome vaccine candidates(23). Therefore, it is not surprising that the number ofspots recognized by the IgG subclasses (IgG1 and IgG4)in this study would be fewer than in previous studies.Comparing the number of antigenic spots detected andthe prevalence of the response in the study population, itis interesting that while IgG4 was the most prevalentresponse in the study population (present in 74% of thepopulation), this isotype detected the least number ofantigenic spots, suggesting that there are few dominantantigens stimulating the IgG4 response. Conversely, IgG1detected in 58% of the participants, recognized the largestnumber of spots, suggesting that the IgG1 response isdirected against a larger repertoire of antigens than theother isotypes.

We have previously shown a dichotomous associationbetween IgA and IgG1 responses directed against adultworm antigens in a different Zimbabwean population (13),

and the spot identification in Figure 3a shows that thereare some antigens recognized either by IgG1 or by IgAbut not by both antibodies. An inverse relationship hasalso been reported in a case of idiopathic arthritis wherepatients deficient in IgG1 had high titres of IgA (37), sug-gesting that the expression of these two subclasses is dif-ferentially regulated. IgG1 recognized the largest numberof antigenic spots: 45 including 8 of the 10 of the WHO’sschistosome vaccine candidates. Furthermore, in thisstudy, IgG1 levels rose with age to peak in the oldest agegroup with the least infection, suggesting that the antibodyis associated with protection against infection. This isconsistent with our previous studies showing an associa-tion between antiworm IgG1 responses and protection toS. haematobium infection (13). IgA recognized fewer anti-genic spots than IgG1, reacting with only 3 of the 10schistosome vaccine candidates including the leading vac-cine candidate 28-kDa glutathione-S-transferase.

The ratio of IgE to IgG4 levels has previously been asso-ciated with protection to re-infection by S. haematobiuminfection (10), and the association of a high IgE ⁄ IgG4ratio with little or no infection in our present study is con-

0

0·2

0·4

0·6

0·8

1

1·2

1·4

1·6

0 0·2 0·4 0·6 0·8

IgE

reco

gnit

ion

inte

nsit

yIgA recognition intensity

P < 0·001, r = 0·816

2550260

0·2

0·4

0·6

0·8

1

1·2

1·4

1·6

1·8

0 0·2 0·4 0·6 0·8

IgG

4 re

cogn

itio

n in

tesi

ty

IgA recognition intensity

P < 0·001, r = 0·758

2516, 26, 50 30

0

0·2

0·4

0·6

0·8

1

1·2

1·4

1·6

1·8

0 0·5 1 1·5 2 2·5

IgG

4 re

cogn

itio

n in

tens

ity

IgG1 recognition intensity

P < 0·001, r = 0·702

615217,7,3816,23,29

300

0·2

0·4

0·6

0·8

1

1·2

1·4

1·6

1·8

0 0·5 1 1·5 2

IgG

4 re

cogn

itio

n in

tens

ity

IgE recognition intensity

P < 0·001, r = 0·819

17,7381630

0

0·2

0·4

0·6

0·8

1

1·2

1·4

1·6

0 1 2 3

IgE

reco

gni�

on in

tens

ity

IgG1 recognition intensity

P < 0·001, r = 0·757

2 615923,29

0

0·5

1

1·5

2

2·5

0 0·2 0·4 0·6 0·8

IgG

1 re

cogn

itio

n in

tens

ity

IgA recognition intensity

P < 0·001, r = 0·604

255026

6

152

1, 389, 17

2923

(a)

(d) (e) (f)

(b) (c)

Figure 3 Comparison of recognition intensity for each antigenic spot by the different isotypes showing a significant positive correlationbetween all 4 isotypes. Italicized numbers refer to the spot identity of antigenic spots preferentially recognized by one of the two isotypes,and their protein identities are given in Table 1. (a) IgG1 vs. IgA. (b) IgG4 vs. IgE. (c) IgG4 vs. IgG1. (d) IgG4 vs. IgA. (e) IgE vs. IgG1.(f) IgE vs. IgA.

F. Mutapi et al. Parasite Immunology

190 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192

sistent with these findings. The inverse relationship betweenIgG4 and IgE has been used to suggest that IgG4 anti-bodies might block IgE effector functions by competitivelybinding to epitopes recognized by both isotypes and therebyinhibit IgE-dependent cell cytotoxicity mediated by mono-cytes, eosinophils or platelets (11,38). The IgG4 subclass isupregulated in association with anti-inflammatory factors[for example, IL-10 (39)] and its own anti-inflammatorycharacteristics are thought to help the immune system todampen inappropriate inflammatory reactions (40). Thiswould only be possible if the two isotypes recognized thesame antigens. In this study, IgG4 bound to the fewest anti-genic spots, while IgE recognized all those recognized byIgG4. IgE reacted with seven of the 10 vaccine candidates,five of which were also recognized by IgG4. The recognitionintensities (for each antigenic spot representing both affinityand avidity) of these two antibodies were the most stronglycorrelated in the study, suggesting a tightly regulated associ-ation between them. Earlier studies in filariasis using wes-tern blotting of one-dimensional gels indicated that the levelof cross-reactivity between IgE and IgG4 differed with thelevel of pathology (41), which suggests that during thecourse of infection, there is dissociation in antigen recogni-tion patterns between the two antibodies. How this dissocia-tion occurs is unclear but may be related to the order ofantibody class switching where IgG4-secreting B cells canswitch to IgE but not the converse (42) and ⁄ or a thresholdbeing reached in antigen levels (32).

The antigens recognized by IgE but not IgG4, and byIgG1 but not IgA, are of interest as they may be associatedwith the development of resistance to schistosomiasis; thisis supported by the fact that most of these antigens arealready known vaccine candidates (23). Studies are nowunderway to identify the proteins that were not serologi-cally reactive, particularly those that are abundant in thesoluble proteome which might play an important role inimmune evasion and ⁄ or modulation of the host immuneresponse (43). Similarly, there is a need to identify theintensely recognized proteins that occurred at insufficient

concentrations in the adult worm proteome to be identifiedby mass spectrometry, as these may indicate highly immu-nogenic vaccine candidates as yet undiscovered.

Overall, our study has indicated that although themajority of the antigens are recognized by all of the fourantibody isotypes tested, there may be restriction in anti-gen recognition for some antigens, for example prevalentisotypes such as IgG4 appear to react to relatively few ofthe major adult worm antigens. This may partly explaindifferences observed in population level isotype dynamicsin human schistosomiasis. Furthermore, there are differ-ences in the recognition of isoforms of some antigens suchas the leading vaccine candidate, 28-kDa glutathione-S-transferase and glyceraldehyde-3-phosphate dehydroge-nase. These isoform-specific differences are an importantconsideration for vaccine development, where recombinantvaccines lack post-translational modifications and there-fore epitopes crucial for induction of the relevant isotype-specific antibody responses. Studies on antigen recognitionpatterns by different antibodies present in sera from indi-viduals or partitioning the study population by age andinfection status as we previously reported for total IgGwill be valuable for identifying further vaccine candidates(32).

ACKNOWLEDGEMENTS

The investigation received financial support from theMedical Research Council, UK, the Wellcome Trust, UK(Grant no WT082028MA), and the Carnegie Trust forthe Universities of Scotland. FM acknowledges supportfrom the MRC, UK (Grant no G81 ⁄ 538). We are grate-ful for the co-operation of the Ministry of Health andChild Welfare in Zimbabwe, the Provincial MedicalDirector of Mashonaland East, the Environmental HealthWorkers, residents, teachers and school children inMutoko and Rusike. We also thank Members of theNational Institutes for Health Research (Zimbabwe) fortechnical support.

REFERENCES

1 Utzinger J, Raso G, Brooker S, et al. Schis-tosomiasis and neglected tropical diseases:towards integrated and sustainable controland a word of caution. Parasitology 2009;136: 1859–1874.

2 Anderson RM & May RM. Infectious Dis-eases of Humans Dynamics and Control.Oxford, Oxford Science, 1992.

3 Woolhouse MEJ, Taylor P, Matanhire D &Chandiwana SK. Acquired immunity andepidemiology of Schistosoma haematobium.Nature 1991; 351: 757–759.

4 Mutapi F, Hagan P, Ndhlovu P & WoolhouseMEJ. Comparison of humoral responses toSchistosoma haematobium in areas with highand low levels of infection. Parasite Immunol1997; 19: 255–263.

5 Dunne DW, Jones FM, Cook L & MoloneyNA. Passively transferable protection againstSchistosoma japonicum induced in the mouseby multiple vaccination with attenuated lar-vae: the development of immunity, antibodyisotype responses and antigen recognition.Parasite Immunol 1994; 16: 655–668.

6 Maloney N, Hinchcliffe P & Webbe G. Passivetransfere of resistance to mice with serafrom rabbits, rats or mice vaccinated withultraviolet-attenuated cercariae of schistosomajaponicum. Parasitology 1987; 94: 497–508.

7 Sher A, Smithers SR & Mackenzie P. Passivetransfer of acquired resistance to Schistoso-ma mansoni in laboratory mice. Parasitology1975; 70 Part 3: 347–357.

8 Mangold BL & Dean DA. Passive transferwith serum and IgG antibodies of irradiatedcercaria-induced resistance against Schistoso-

Volume 33, Number 3, March 2011 Isotype-specific reactivity of helminth antigens

� 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192 191

ma mansoni in mice. J Immunol 1986; 136:2644–2648.

9 Dunne DW, Butterworth AE, Fulford AJ,Ouma JH & Sturrock RF. Human IgEresponses to Schistosoma mansoni andresistance to reinfection. Mem Inst OswaldoCruz 1992; 87: 99–103.

10 Hagan P, Blumenthal UJ, Dunne D, Simp-son AJG & Wilkins AH. Human IgE, IgG4and resistance to reinfection with Schistoso-ma haematobium. Nature 1991; 349: 234–245.

11 Demeure CE, Rihet P, Abel L, Ouattara M,Bourgois A & Dessein AJ. Resistance toSchistosoma mansoni in humans: influence ofthe IgE ⁄ IgG4 balance and IgG2 in immunityto reinfection after chemotherapy. J InfectDis 1993; 168: 1000–1008.

12 Grogan JL, Kremsner PG, van Dam GJ,Deelder AM & Yazdanbakhsh M. Anti-schis-tosome IgG4 and IgE at 2 years afterchemotherapy: infected versus uninfectedindividuals. J Infect Dis 1997; 176: 1344–1350.

13 Mutapi F, Ndhlovu PD, Hagan P, et al. Che-motherapy accelerates the development ofacquired immune responses to Schistosomahaematobium infection. J Infect Dis 1998;178: 289–293.

14 Hagan P & Sharaf O. Schistosomiasis vaccines.Expert Opin Biol Ther 2003; 3: 1271–1278.

15 Capron A & Dessaint JP. Molecular basis ofhost-parasite relationship: towards the defini-tion of protective antigens. Immunol Rev1989; 112: 27–48.

16 Correa-Oliveira R, Pearce EJ, Oliveira GC,et al. The human immune response todefined immunogens of Schistosoma mansoni:elevated antibody levels to paramyosin instool-negative individuals from two endemicareas in Brazil. Trans R Soc Trop Med Hyg1989; 83: 798–804.

17 Zhang Y, Taylor MG, McCrossan MV &Bickle QD. Molecular cloning and character-ization of a novel Schistosoma japonicum‘‘irradiated vaccine-specific’’ antigen, Sj14-3-3. Mol Biochem Parasitol 1999; 103: 25–34.

18 Butterworth AE, Bebsted-Smith R & CapronA. Immunity in human schistosomiasis:prevention by blocking antibodies of theexpression of immunity in young children.Parasitology 1987; 94: 281–300.

19 Gryseels B, Stelma F, Talla I, et al. Immuno-epidemiology of Schistosoma mansoni infec-tions in a recently exposed community inSenegal. Mem Inst Oswaldo Cruz 1995; 90:271–276.

20 Ndhlovu P, Cadman H, Vennervald BJ, Chris-tensen NO, Chidimu N & Chandiwana SK.Age-related antibody profiles in Schistosomahaematobium in a rural community inZimbabwe. Parasite Immunol 1996; 18: 181–191.

21 Webster M, LibrandaRamirez BDL, AliguiGD, et al. The influence of sex and age onantibody isotype responses to Schistosomamansoni and Schistosoma japonicum inhuman populations in Kenya and the Philip-pines. Parasitology 1997; 114: 383–393.

22 Fitzsimmons CM, Stewart TJ, HoffmannKF, Grogan JL, Yazdanbakhsh M & Dunne

DW. Human IgE response to the Schistoso-ma haematobium 22.6 kDa antigen. ParasiteImmunol 2004; 26: 371–376.

23 Al-Sherbiny M, Osman A, Barakat R, ElMorshedy H, Bergquist R & Olds R. In vitrocellular and humoral responses to Schistosomamansoni vaccine candidate antigens. Acta Trop2003; 88: 117–130.

24 Mutapi F, Mduluza T, Gomez-Escobar N,et al. Immuno-epidemiology of human Schis-tosoma haematobium infection: preferentialIgG3 antibody responsiveness to a recombi-nant antigen dependent on age and parasiteburden. BMC Infect Dis 2006; 6: 96.

25 Mutapi F, Winborn G, Midzi N, Taylor M,Mduluza T & Maizels RM. Cytokineresponses to Schistosoma haematobium in aZimbabwean population: contrasting profilesfor IFN-gamma, IL-4, IL-5 and IL-10 withage. BMC Infect Dis 2007; 7: 139.

26 Mutapi F, Burchmore R, Foucher A, et al.Praziquantel treatment of people exposed toSchistosoma haematobium enhances serologi-cal recognition of defined parasite antigens.J Infect Dis 2005; 192: 1108–1118.

27 Katz N, Chaves A & Pellegrino J. A simpledevice for quantitative stool thick smeartechnique in schistosomiasis mansoni. RevInst Med Trop Sao Paulo 1972; 14: 397–400.

28 Mott KE. A reusable polyamide filter fordiagnosis of S. haematobium infection byurine filtration. Bull Soc Pathol Exot 1983;76: 101–104.

29 Agnew AM, Murare HM & Doenhoff MJ.Immune attrition of adult schistosomes. Par-asite Immunol 1993; 15: 261–271.

30 Reilly LJ, Magkrioti C, Cavanagh DR,Mduluza T & Mutapi F. Effect of treatingSchistosoma haematobium infection onPlasmodium falciparum-specific antibodyresponses. BMC Infect Dis 2008; 8: 158.

31 Woolhouse MEJ & Hagan P. Seeking theghost of worms past. Nat Med 1999; 5:1225–1227.

32 Mutapi F, Burchmore R, Mduluza T, MidziN, Turner CM & Maizels RM. Age-relatedand infection intensity-related shifts in anti-body recognition of defined protein antigensin a schistosome-exposed population. J InfectDis 2008; 198: 167–175.

33 Capron A, Dessaint JP, Capron M, JosephM & Pestel J. Role of anaphylactic antibod-ies in immunity to schistosomes. Am J TropMed Hyg 1980; 29: 849–857.

34 Dunne DW, Butterworth AE, Fulford AJC,et al. Immunity after treatment of humanschistosomiasis: association between IgEantibodies to adult worm antigens and resis-tance to reinfection. Eur J Immunol 1992; 22:1483–1494.

35 Sturrock RF, Kimani R, Cottrell BJ, et al.Observations on possible immunity to rein-fection among Kenyan school children aftertreatment for Schistosoma mansoni. Trans RSoc Trop Med Hyg 1983; 77: 363–371.

36 Webster M, Fulford AJC, Braun G, et al.Human Immunoglobulin E responses to arecombinant antigen from Schistosoma man-soni adult worms are associated with low

intensities of reinfection after treatment.Infect Immun 1996; 64: 4042–4046.

37 Yildiz B & Kural N. IgG1 deficiency andhigh IgA level with juvenile idiopathic arthri-tis. Eur J Pediatr 2007; 166: 1179–1180.

38 Rihet P, Demeure CE, Dessein AJ & Bour-gois A. Strong serun inhibition of specificIgE correlated to competing IgG4, revealedby a new methodology in subjects from aS. mansoni endemic area. Eur J Immunol1992; 22: 2063–2070.

39 Satoguina JS, Weyand E, Larbi J & HoeraufA. T regulatory-1 cells induce IgG4 produc-tion by B cells: role of IL-10. J Immunol2005; 174: 4718–4726.

40 Aalberse RC, Stapel SO, Schuurman J &Rispens T. Immunoglobulin G4: an odd anti-body. Clin Exp Allergy 2009; 39: 469–477.

41 Hussain R & Ottesen EA. IgE responses inhuman filariasis. III. Specificities of IgE andIgG antibodies compared by immunoblotanalysis. J Immunol 1985; 135: 1415–1420.

42 Stavnezer J & Amemiya CT. Evolution ofisotype switching. Semin Immunol 2004; 16:257–275.

43 Hewitson JP, Grainger JR & Maizels RM.Helminth immunoregulation: the role of para-site secreted proteins in modulating host immu-nity. Mol Biochem Parasitol 2009; 167: 1–11.

SUPPORTING INFORMATION

Additional Supporting Informationmay be found in the online version ofthis article:

Figure S1. Enzyme-linked immuno-sorbent titration assays to optimizethe sensitivity and specificity of the is-otypes used in the assay using adultworm antigen at 5 lg ⁄ mL.

Figure S2. Western blot assays usedto optimize the sensitivity and speci-ficity of the isotypes used.

Figure S3. Results of the westernblot assays using the negative controlsera (pooled from five Europeandonors who have never travelled totropical countries) diluted at 1 : 200and IgG4 secondary antibody dilutedat 1 : 1000 showing no reactivity ofthe negative control sera with schisto-some adult women antigens.

Please note: Wiley-Blackwell arenot responsible for the content orfunctionality of any supporting mate-rials supplied by the authors. Anyqueries (other than missing material)should be directed to the correspond-ing author for the article.

F. Mutapi et al. Parasite Immunology

192 � 2011 Blackwell Publishing Ltd, Parasite Immunology, 33, 181–192