Antibody responses to Schistosoma mansoni schistosomula antigens

Summary While antigens from Schistosoma schistosomula have been suggested as potential vaccine candidates, the association between antibody responses with schistosomula antigens and infection intensity at reinfection is not well known. Schistosoma mansoni‐infected individuals were recruited from a schistosomiasis endemic area in Uganda (n = 372), treated with 40 mg/kg praziquantel (PZQ) and followed up at five weeks and at one year post‐treatment. Pre‐treatment and five weeks post‐treatment immunoglobulin (Ig) E, IgG1 and IgG4 levels against recombinant schistosomula antigens rSmKK7, rSmLy6A, rSmLy6B and rSmTSP7 were measured using ELISA. Factors associated with detectable pre‐treatment or post‐treatment antibody response against the schistosomula antigens and the association between five‐week antibody responses and one year post‐treatment reinfection intensity among antibody responders were examined. Being male was associated with higher pre‐treatment IgG1 to rSmKK7, rSmLy6a and AWA. Five weeks post‐treatment antibody responses against schistosomula antigens were not associated with one year post‐treatment reinfection intensity among antibody responders’ antibody levels against rSmKK7, rSmLy6B and rSmTSP7 dropped, but increased against rSmLy6A, AWA and SEA at five weeks post‐treatment among antibody responders. S. mansoni‐infected individuals exhibit detectable antibody responses to schistosomula antigens that are affected by treatment. These findings indicate that schistosomula antigens induce highly varied antibody responses and could have implications for vaccine development.


Funding information
Factors associated with detectable pre-treatment or post-treatment antibody response against the schistosomula antigens and the association between five-week antibody responses and one year post-treatment reinfection intensity among antibody responders were examined. Being male was associated with higher pretreatment IgG1 to rSmKK7, rSmLy6a and AWA. Five weeks post-treatment antibody responses against schistosomula antigens were not associated with one year posttreatment reinfection intensity among antibody responders' antibody levels against rSmKK7, rSmLy6B and rSmTSP7 dropped, but increased against rSmLy6A, AWA and SEA at five weeks post-treatment among antibody responders. S. mansoni-infected individuals exhibit detectable antibody responses to schistosomula antigens that are affected by treatment. These findings indicate that schistosomula antigens induce

| INTRODUC TI ON
Approximately 206 million people worldwide required treatment for schistosomiasis in 2016. 1 Control programmes in affected countries have reduced the morbidity associated with schistosomiasis, 2 yet despite this, treatment is unable to prevent reinfection with Schistosoma species, and consequently, transmission of schistosomiasis remains a global problem. Mass drug administration with praziquantel (PZQ) alone may not ultimately control schistosomiasis, 3 and therefore, a solution such as a prophylactic vaccine is needed to provide long-term immunity and preferably complement the PZQbased schistosomiasis control strategy. 4 There is currently no vaccine against schistosomes, and only a few vaccine candidates are in clinical trials. Despite the limited success, defined antigens continue to be investigated as vaccine targets. Among these are those from the larvae of the schistosome, known as schistosomula. The schistosomula develop when free-living cercariae penetrate host skin, lose their bifurcated tails and shed their cercarial glycocalyx coat in the skin in a process called transformation. Early work on schistosome parasite biology showed that transformation makes the skin schistosomula vulnerable to killing by host immune responses. 5 This killing is mediated through complement fixation 6,7 and antibody (IgE, IgA and IgG)-dependent cell-mediated cytotoxicity (ADCC). [8][9][10] The antibodies coat the schistosomula and allow eosinophils to kill the parasite in vitro. 11,12 This suggests that the schistosomula are a potential source of vaccine candidates.
By examining the S. mansoni transcriptome, genes that are highly expressed or upregulated during the schistosomula stage compared to the infective cercariae can be identified. 13 Indeed, DNA microarray-based analysis of the S. mansoni life cycle has revealed schistosomula-enriched gene products, such as those coding for SmKK7 (smp_194830), S. mansoni lymphocyte antigen 6 isoforms A and B (SmLy6A; smp_019350, SmLy6B; smp_105220) and S. mansoni tetraspanin 7 (SmTSP7; smp_099770). 14 SmLy6A and SmLy6B are probably glycophosphatidylinositol (GPI)-anchored antigens, found in schistosomula, 13,15 but also in adult tegumental and mesenchymal tissues. 16,17 Although SmLy6A and SmLy6B are homologues of human CD59, which inhibits the formation of the complement membrane attack complex, they do not inhibit complement fixation and their function remains unknown. 17 SmKK7 is a putative immunomodulatory protein found in the peripheral nervous system of S. mansoni adults 18 and upregulated in 7-and 14day schistosomula. 15 SmTSP7 is a membrane-spanning tetraspanin with unknown function but has been identified in the tegument of the adult worm. 19 Epidemiological studies have shown that specific host antibodies to Schistosoma antigens contribute to immunity to schistosomiasis. For instance, antibody responses targeting S. mansoni adult worm antigens are associated with resistance or susceptibility to reinfection of people at risk in endemic areas. 20,21 Although adult worm antigen-specific IgE is associated with resistance against reinfection, it is important to note that vaccine candidates that induce IgE production may not be safe for use, particularly in previously helminth-infected individuals. This was accentuated during a trial involving a hookworm vaccine candidate that was discontinued when previously hookworm-infected participants developed an IgE-mediated allergic reaction. 22,23 The same hookworm vaccine had earlier been shown to be safe with healthy hookworm-naive individuals. 24 This suggests that helminth (including Schistosoma) antigens should be pre-clinically screened using sera from infected people to determine whether antigen-specific IgE is present.
On the other hand, IgG4 blocks IgE-mediated protective immunity by competing with IgE to bind shared epitopes 25 and engaging the inhibitory IgG receptor, FcγIIb, that downregulates signalling from the IgE receptor, FcεRI, on effector cells. 26,27 As a result, activation of effector cells is inhibited 28 and protective responses may be less effective. Indeed, Schistosoma-infected people, especially children (who are the most susceptible to reinfection), produce high levels of IgG4. 29,30 Although antibody responses (IgE and IgG) to crude purified extracts of schistosomula have been shown to contribute to human resistance, 31,32 few studies have looked at antibody responses to recombinant schistosomula antigens. 33 The aim of this study was to determine how infection intensity, age and sex affect antibody responses to the recombinant S. mansoni schistosomula antigens rSmLy6A, rSmLy6B, rSmKK7 and rSmTSP7 in an endemic population. We also analysed the effect of treatment on the antibody responses against these recombinant antigens and the correlation between the antibody and cytokine responses (in the companion paper). Finally, we examined whether pre-or five weeks post-treatment antibody responses were associated with reinfection one year later.

| Ethical statement
Informed consent was obtained from adults in Namoni to participate in this study. Children gave written assent to participate in the study.

| Recruitment of study participants
Individuals were recruited from Namoni village, a Ugandan fishing community that is endemic with schistosomiasis at the shore of Lake Victoria as part of TheSchistoVac study (http://www.theschistovac. eu) to develop antigens for a prophylactic schistosomiasis vaccine. A total of 372 individuals aged between 6 and 40 years were recruited from Namoni (Figure 1). TheSchistoVac cohort in Namoni has been described elsewhere. 34 In September 2011, the participants were treated with two doses of praziquantel (40 mg/kg) one week apart and followed up for 5 weeks and a year. Infection intensity was determined from stool samples collected pre-treatment, at 5 weeks for efficacy of the PZQ treatment and one year for reinfection.
Participants were asked to donate a venous blood sample immediately before the first praziquantel treatment and a further blood sample five weeks later. Plasma samples were separated, stored and later tested for antibody responses.
Of those recruited and followed up to the end of the study, 244 (65.6%) completed the study. However, 240 (64.5%) of those recruited had complete data on infection intensity, age, sex and levels of IgG1, IgG4 and IgE to AWA, SEA and the antigens at baseline, five weeks and one year after PZQ treatment (Figure 1).
The water contact behaviour of the study participants was obtained and recorded before treatment and included the duration of contact (time spent in the lake) and whether or not the participants bathed, swam, played, fished or processed fish, washed utensils or clothes in the lake, or engaged in transport across the lake or engaged in farm irrigation using lake water. The water contact behaviour was self-reported during questionnaire interviews.

| Stool examination by microscopy
Before treatment, three stool samples were collected on three consecutive days in the morning and two thick smears made from each sample. The six slides were examined using the Kato Katz method for the number of S. mansoni eggs as previously described. 35 The egg count was the average of the six slides from the three samples.

| Schistosoma mansoni antigens
The antigens used in this study were S. mansoni adult worm (AWA) and egg antigens (SEA) and recombinant schistosomula antigens rSmKK7, rSmLy6A, rSmLy6B and rSmTSP7. These antigens were identified as highly expressed products in the schistosomula life cycle stage after screening the Schistosoma transcriptome using DNA microarrays, as previously described. 14,16 Specifically, a >5-fold increase in expression when comparing normalized expression averages from snail (egg, miracidia, mother sporocyst and daughter sporocyst) to schistosomula (3-h, 24-h, 3-day and 6-day schistosomula) life stages was observed. 16 Recombinant antigens were expressed in Escherichia coli and purified using methods as previously described for rSmKK7 (smp_194830), 14 rSmLy6A and rSmLy6B. 15 For SmTSP7, 78 amino acids representing the extracellular loop 2 of the protein (108-185AA; defined by TMHMM2.0 software 39 were expressed using the same vector, expression parameters and purification methods as SmLy6A and SmLy6B. The selected proteins used in this study were screened for in silico similarity to known allergens, and any with a similarity were discarded from the pipeline. The selected proteins were further screened for IgE reactivity using sera from a schistosomiasis endemic cohort, Musoli. It was found that Musoli residents had minimal IgE reactivity to these proteins, and the antigens were therefore selected for further study.

| Blood collection and estimation of antibody responses
Plasma samples were separated from whole blood before and five weeks after PZQ treatment by centrifugation and stored at -20°C The study profile describing the screening, recruitment and follow-up of the study participants until ready for use. The concentrations of IgG1, IgG4 and IgE specific for the schistosomula antigens (rSmKK7, rSLy6A, rSmLy6B, rSmTSP7) and the crude parasite antigens (AWA and SEA) were quantified using ELISA as previously described. 40

| Cytokine responses to schistosomula antigens
Stored peripheral blood mononuclear cells were thawed, stimulated with rSmKK7, rSLy6A, rSmLy6B and rSmTSP7 and the supernatant collected after 24 h and tested for cytokines, chemokines and growth factors as described in the companion paper (Egesa unpublished).

| Statistical methods
Antibody (IgG1, IgG4 and IgE) levels against the schistosomula antigens before treatment were categorized as antibody responders and nonresponders. Antibody responders were participants with antibody levels greater than the mean plus 3 standard deviations against the schistosomula antigens detected in the plasma of 26 nonendemic European individuals. Factors associated with detectable pretreatment or post-treatment antibody (IgG1, IgG4 and IgE) responses against crude schistosome and schistosomula antigens were determined using multiple logistic regression. The correlation between the antibody levels and cytokine responses to schistosomula antigens among antibody responders was investigated using Spearman's rank correlation. Multivariable logistic regression was used to investigate whether detectable five weeks post-treatment antibody responses were associated with one year post-treatment reinfection intensity.
Because antibody levels were not normally distributed, the levels were log-transformed and the transformed antibody levels compared before and 5 weeks after treatment among responders, using the paired t-test. Statistical analysis was performed using Stata version 13 (Stata Corp, College Station, TX, USA) and graphs drawn using GraphPad Prism version 6.0 g (GraphPad Software, Inc., , San Diego, CA, USA). Table 1 shows the demographics of the Namoni study participants at baseline. The age range of the study participants was 6-40 years. There were slightly more females than males. The study participants mainly had a heavy infection intensity (>400 epg) before treatment.

| Demographics of the study participants
The majority of the study participants spent no longer than one hour a day in the lake (Table S1). However, the male study participants spent more time in the lake than their female counterparts (X 2 (5, N = 226) = 20.24, P = 0.001). There was no age difference in water contact behaviour (X 2 (10, N = 226) = 16.37, P = 0.089).

| Number of antibody responders to the schistosomula antigens
The number and percentage of responders are shown in Table 2.
The number of IgG1, IgG4 and IgE responders to the crude S. mansoni AWA and SEA was relatively high compared to the schistosomula antigens.   (17) Heavy (>400) 155 (65) Before treatment, being male was associated with higher pretreatment IgG1 to AWA (P = 0.013) and IgE to AWA (P = 0.005) and SEA (P = 0.029) after adjusting for gender. Being 10 years and over was associated with higher IgE levels against AWA (P = 0.031).

| Being male was associated with pretreatment and post-treatment antibody responses against the recombinant schistosomula antigens among antibody responders
Tables 4 and S3 show factors that were associated with pretreatment and five weeks post-treatment antibody responses against the recombinant schistosomula antigens among responders, respectively. Before treatment, being male was associated with higher IgG1 to rSmKK7 (P = 0.009) and rSmLy6a (P = 0.006).
Five weeks following treatment, being male was associated with higher IgG4 levels against SmKK7 (P = 0.0001) and SmLy6B (P = 0.045).

| Those reinfected after treatment had high pretreatment intensity
Predisposition describes the phenomenon where those that are re- infection intensity (and subsequently resistance to reinfection) was statistically significant after adjusting for age and gender (P = 0.011).

| Five weeks post-treatment antibody responses against schistosomula antigens were not associated with one year post-treatment reinfection intensity among antibody responders
To assess whether antibody responses to schistosomula antigens predicted reinfection, the association between detectable five   (Table 5).

| PZQ treatment differentially affects antibody responses against the recombinant schistosomula antigens among antibody responders
The antibody levels against the recombinant schistosomula antigens pre-and post-treatment among antibody responders are shown in The effect of PZQ treatment on crude antigens AWA and SEA among antibody responders is shown in Figure S1. IgG1 to the crude antigens AWA (P < 0.0001) and SEA (P < 0.0001), IgG4 to AWA (P < 0.0001) and IgE to AWA (P = 0.003) increased five weeks posttreatment. IgG4 (P = 0.526) and IgE (P = 0.438) levels to SEA were not affected by treatment.

| D ISCUSS I ON
Schistosomula are vulnerable to both complement and antibodydependent cell-mediated cytotoxicity in in vitro experiments. [9][10][11]41,42 This makes the newly transformed schistosomulum a plausible source of candidate antigens for a vaccine against S. mansoni. 43 There is information on antibody responses to various Schistosoma antigens, but protective antibody responses against schistosomula have not been identified yet and factors that affect these antibody responses are unknown. In this study, we have presented an analysis of antibody responses to schistosomula-enriched antigens in a cohort from an endemic area. Not surprisingly, we found that being male was associated with high pre-and five weeks post-treatment antibody responses against the schistosomula antigens SmKK7 and  47 The host is briefly exposed to the schistosomula in a natural situation, and therefore, the required antigen threshold is probably not reached to build immunity. 48,49 This may suggest that cumulative exposure that is closely related to age of the host may have no impact on immune responses to schistosomula antigens. The implication this has on vaccine design is that vaccines based on schistosomula antigens could provide prolonged exposure to a sufficient amount of schistosomula antigen and generate the same level of protection irrespective of age of the host. On the other hand, the present study observed that sex and age were associated with specific antibody responses to schistosome crude antigens AWA and SEA consistent with previous data. 45 Where participants had both antibody and cytokine data described in the companion paper (n = 54), very few weak correlations between the antibody and cytokine responses were significant.
Antibody responses to Schistosoma antigens are affected by treatment with PZQ. 20,50 Praziquantel (PZQ) is the most effective drug for treatment of schistosomiasis 51 and is thought to disrupt the regulation of calcium (Ca 2+ ) ion permeability through surface membranes. 52 The resulting Ca 2+ -induced contractions are sustained, which simultaneously paralyse the parasite and dislodge them from venules. PZQ also damages the worm's body surface by forming fragile blebs and vacuoles in the schistosome tegument. 53 The blebs and vacuoles rupture, damaging the surface and facilitating host immune responses against schistosomes. 54,55 In addition, treatment exposes sequestered antigens and antibody responses towards these exposed antigens are elevated. 54,55 Although the effect of PZQ on the schistosomula is thought to be limited, 56 treating schistosomula in vitro with drugs (PZQ and oxamniquine) exposes a magnitude of sequestered antigens that are otherwise not accessible on living intact schistosomula. 15 However, not all Schistosoma antigens induce an increase in antibodies following PZQ treatment. For example, IgG1 levels against rSmLy6A and rSmLy6B have been shown to drop following PZQ treatment among antibody responders. 16 Similarly, we also found a drop in IgG1, IgG4 and IgE to rSmKK7, rSmLy6B and rSmTSP7, respectively, after PZQ treatment among antibody responders. On the other hand, we did observe an increase to SmLy6A. These data are in apparent contradiction to a previously As Schistosoma has more than one life cycle stage in the definitive host, some Schistosoma antigens are shared across these stages of Schistosoma including the schistosomulum, the adult and the egg. 57 As a result, there is potential for cross-reactive antibody responses to the adult worm, the egg and the schistosomula. The schistosomula antigens tested in the present study SmKK7, SmLy6a, SmLy6b and Smtsp7 are also expressed by the adults. 14,17,57 These studies have shown that there is a low expression of the mRNA transcripts for these antigens in miracidia, sporocysts and cercariae, increasing in schistosomula and adult worms. Most importantly, the egg that induces the immunopathology associated with chronic schistosomiasis had minimal levels of transcripts. Because treatment with PZQ alters immune responses to adult Schistosoma antigens, 58 it could indirectly modify host immune responses to the schistosomula. 59 This implies that cross-reactive antibodies against larvae could be boosted by killing the adults and releasing antigens. Therefore, shared epitopes could be good antigens to study further for vaccines that target both the schistosomula and the adults.
In longitudinal studies, antibody responses to specific schistosome antigens have been shown to correlate with resistance to reinfection. 60,61 Generally, the antibody responses associated with human resistance to reinfection are directed against the schistosome adult worm and not egg antigens. 62 This finding may not be surprising as the immune responses to crude adult worm and egg antigens have been widely studied. However, limited work has shown that IgG 32 and IgE 31 antibody responses to larval antigens are associated with resistance to reinfection. By looking at those infected individuals who were cured five weeks following treatment, we related their reinfection intensity to antibody responses to schistosomula antigens pre-treatment and five weeks post-treatment.
Determinants of reinfection (reviewed in 38 ) we considered in this analysis were age and sex. In the present study, pre-treatment or five weeks post-treatment antibody responses against schistosomula antigens were not associated with one year post-treatment reinfection intensity. These findings imply that although the schistosomula antigens tested in the present study were targeted by IgG1 and IgE, the responses were not protective. With the schistosomula as the target of immune attrition, using the radiation-attenuated cercariae the only experimental vaccine to date to consistently provide high levels of protective immunity (60%-80%) against challenge, [63][64][65] there is a need to identify other schistosomula antigens that are targets of protective immunity in humans. It could be that other, perhaps nonprotein, schistosomula molecules such as glycans may be targets of protective immunity in humans. A glycomic analysis of the life cycle of S. mansoni shows stage-specific expression of glycans, including during the schistosomula stage. 66 Glycans are highly antigenic, 67 and schistosomula glycans are the dominant targets of antibody  79 This was, in fact, performed for the schistosomula antigens used in this study, by TheSchistoVac (http://www.theschistovac.eu/). Our present study cohort (the Namoni cohort) is a similar cohort to the Musoli cohort (also in Uganda) but showed that there were individuals with high IgE responses to some of the schistosomula antigens. There appears to be heterogeneity in the response to schistosomula, even between geographically close populations (Musoli vs Namoni). It may be that across a population, there will be people that respond to any novel Schistosoma antigen with IgE, and it may not be possible to screen this reactivity out, meaning that there may be some people who will end up with an adverse reaction to any novel antigen. Of course, if the risk is small and can be managed, then the benefit of administering an antischistosome vaccine will far outweigh the risks.

ACK N OWLED G EM ENTS
We are grateful to the participants and authorities of Namoni for their participation and cooperation in this study. We

CO N FLI C T O F I NTE R E S T
The authors declare that there is no conflict of interest.

D I SCLOS U R E S
None.