Trichuris muris whey acidic protein induces type 2 protective immunity against whipworm

Human whipworm (Trichuris trichiura) infects approximately 1 in 15 people worldwide, representing the leading infectious cause of colitis and subsequent, inflammatory bowel disease (IBD). Current control measures focused on mass deworming have had limited success due to low drug efficacies. Vaccination would be an ideal, cost-effective strategy to induce protective immunity, leading to control of infection and transmission. Here we report the identification of whey acidic protein, a whipworm secretory protein, as a strong immunogen for inducing protective efficacy in a surrogate mouse T. muris infection model. The recombinant WAP protein (rTm-WAP49), as well as a single, highly conserved repeat within WAP (fragment 8) expressed as an Na-GST-1 fusion protein (rTm-WAP-F8+Na-GST-1), generate a strong T helper type 2 (Th2) immune response when delivered as subcutaneous vaccines formulated with Montanide ISA 720. Oral challenge with T. muris infective eggs following vaccination led to a significant reduction in worm burden of 48% by rTm-WAP49 and 33% by rTm-WAP-F8+Na-GST-1. The cellular immune correlates of protection included significant antigen-specific production of Th2 cytokines IL-4, IL-9, and IL-13 by cells isolated from the vaccine-draining inguinal lymph nodes, parasite-draining mesenteric lymph nodes, and spleen in mice vaccinated with either rTm-WAP49 or rTm-WAP-F8+Na-GST-1. The humoral immune correlates included a high antigen-specific ratio of IgG1 to IgG2a, without eliciting an IgE-mediated allergic response. Immunofluorescent staining of adult T. muris with WAP antisera identified the worm’s pathogenic stichosome organ as the site of secretion of native Tm-WAP protein into the colonic mucosa. Given the high sequence conservation for the WAP proteins from T. muris and T. trichiura, the results presented here support the WAP protein to be further evaluated as a potential human whipworm vaccine candidate.

Human whipworm (Trichuris trichiura) infects approximately 1 in 15 people worldwide, representing the leading infectious cause of colitis and subsequent, inflammatory bowel disease (IBD). Current control measures focused on mass deworming have had limited success due to low drug efficacies. Vaccination would be an ideal, cost-effective strategy to induce protective immunity, leading to control of infection and transmission. Here we report the identification of whey acidic protein, a whipworm secretory protein, as a strong immunogen for inducing protective efficacy in a surrogate mouse T. muris infection model. The recombinant WAP protein (rTm-WAP49), as well as a single, highly conserved repeat within WAP (fragment 8) expressed as an Na-GST-1 fusion protein (rTm-WAP-F8+Na-GST-1), generate a strong T helper type 2 (Th2) immune response when delivered as subcutaneous vaccines formulated with Montanide ISA 720. Oral challenge with T. muris infective eggs following vaccination led to a significant reduction in worm burden of 48% by rTm-WAP49 and 33% by rTm-WAP-F8+Na-GST-1. The cellular immune correlates of protection included significant antigen-specific production of Th2 cytokines IL-4, IL-9, and IL-13 by cells isolated from the vaccine-draining inguinal lymph nodes, parasite-draining mesenteric lymph nodes, and spleen in mice vaccinated with either rTm-WAP49 or rTm-WAP-F8+Na-GST-1. The humoral immune correlates included a high antigen-specific ratio of IgG1 to IgG2a, without eliciting an IgE-mediated allergic response. Immunofluorescent staining of adult T. muris with WAP antisera identified the worm's pathogenic stichosome organ as the site of secretion of native Tm-WAP protein into the colonic mucosa. Given the high sequence conservation for the WAP proteins from T. muris and T. trichiura, the results presented here support the WAP protein to be further evaluated as a potential human whipworm vaccine candidate. a1111111111 a1111111111 a1111111111 a1111111111 a1111111111

Introduction
It is estimated that 465 million people are living with whipworms (Trichuris trichiura) in their colon, predominately in the developing nations of the Americas, Asia, and Africa. In children, T. trichiura infections cause chronic colitis, resulting in nutritional deficiencies, growth stunting, and cognitive impairment [1]. In adults, Trichuris-induced colitis is being studied as a possible major environmental driver of inflammatory bowel disease [2]. The current approach to human trichuriasis control relies on mass deworming campaigns with yearly, single dose albendazole or mebendazole [3]. However, this approach has shown limited benefit against trichuriasis due to low efficacy, high rates of post-treatment reinfection, and failure to co-implement adequate sanitation and hygiene practices [4,5]. It is therefore important to design better or alternative strategies to helminth prevention, including the development of a trichuriasis vaccine [6]. While there is no direct evidence that humans develop protective immunity to trichuriasis, epidemiological studies suggest that there is an age-related reduction in worm burden and adult worm longevity in humans beginning in adolescence that may be due to acquired immunity [7,8].
The closely related Trichuris muris parasite, specific to mice, is a well-established model for disease pathophysiology and subsequent host immunity, and is widely used for evaluating the immunogenicity and efficacy of vaccine candidates [9,10]. In both mice and humans, ingested whipworm eggs hatch into larvae in the terminal ileum and caecum due to interactions with the host microbiome [11]. There, the helminths' anterior-end organ, known as a stichosome, excretes products implicated in facilitating the invasion and maintenance of the parasite into the host's gut epithelia [9,12]. Once embedded, the Trichuris worms become fixed in location, causing inflammation at the site of intracellular attachment and induce colitis-like pathology [9]. The T. muris excretions and secretions (ES), thought to predominately originate from the stichosome, can be collected from the media of in-vitro cultured helminths. We and others have reported that vaccination with T. muris ES product elicits protective immunity in murine models [13][14][15]. Protective immunity in preclinical models correlates with T helper type 2 (Th2) responses, characterized by the production of cytokines IL-4, IL-9, and IL-13 along with IgG1 antibodies and high IgG1 to IgG2a ratios [9,13,[15][16][17][18]. Conversely, susceptibility is characterized by a predominate TH1 response (IL-12 and Interferon (IFN)-γ) and induction of IgG2a antibodies [9,13,16].
Here we report the identification and analyses of immunogenicity as well as efficacy of two highly abundant secretory proteins, whey acidic protein (WAP) and cysteine-rich secretory proteins (CRISPS), Antigen 5, and Pathogenesis-related 1 (CAP-1), from among the T. muris ES antigens. Recombinant Tm-WAP (rTm-WAP49) induced significant protection against T. muris challenge, while recombinant Tm-CAP-1 was ineffective. Protection against T. muris infection by rTm-WAP49 was associated with strong Th2 cellular and humoral immune responses.

Immunoscreening for the identification of highly abundant T. muris excreted-secreted (ES) antigens
Serum from mice immunized with T. muris ES product was used to screen the T. muris adult whole worm expressional cDNA library and identified a total of 102 positive clones. DNA sequencing revealed a whey acidic protein (Tm-WAP) as the most abundant of the candidates identified (63 clones). BLAST searching revealed that Tm-WAP shares 96.8% amino acid sequence identity with a gene product (TMUE_s0165000300) in the established T. muris genome on WormBase [19,20]. The Tm-WAP protein consists of several repeats (3-10) of a 50 amino acid fragment that share 69-96% sequence identity and possess six conserved cysteines predicted to form multiple disulfide bonds (Fig 1A) (DiANNA 1.1) [21][22][23]. The Tm-WAP gene shares high sequence alignment with the T. trichiura putative porin proteins, TT95 [24] and TT52 [25], at 54% and 47% amino acid sequence identity, respectively (Fig 1B).
The second most abundant clone identified from the cDNA library screening was T. muris cysteine-rich secretory proteins (CRISPS), Antigen 5, and Pathogenesis-related 1 (CAP-1) protein (35 clones) (Fig 1C). The CAP-domain family (also known as SCP/TAPS) was previously identified based on its abundance in the secretome of whipworm and other soil-transmitted helminths [26,27]. BLAST searching revealed that Tm-CAP-1 shares 99.3% amino acid sequence identity with a gene product (TMUE_s0030008500) in the established T. muris genome on WormBase [19,20]. Tm-CAP-1 shares the highest sequence alignment with the T. suis SCP-like protein (accession number KHJ42268.1) and T. trichiura CAP-domain containing protein (accession number CDW52210.1), at 48% and 38% amino acid sequence identity, respectively (Fig 1D).
High sequence alignment (defined as >40%) is strongly predictive of protein homology with shared structure [28]. This suggests that both WAP and CAP-1 would be translatable targets of a future T. trichiura vaccine.
The Tm-WAP protein was expressed with a 6His-tag at C-terminus in Pichia pastoris X33 as a soluble protein and purified by IMAC. After multiple unsuccessful attempts at expressing the entire protein in E. coli, ultimately yeast were found to be an efficient expression platform. The most conserved repeat of Tm-WAP, fragment 8, was expressed with a Necator americanus glutathione s-transferase-1 (Na-GST-1)-tag at the N-terminus as a soluble recombinant fusion protein in E. coli BL21 under induction of 1 mM IPTG and purified with immobilized IMAC. In addition to improving the expression and solubility of the fusion protein, Na-GST-1 is a hookworm vaccine candidate [29], and selected for the potential as a pan-anthelminthic vaccine candidate. The recombinant protein migrated at 49 kDa, therefore named rTm-WAP49. The full-length recombinant Tm-CAP-1 protein was expressed in Escherichia coli (E. coli) BL21 as an insoluble inclusion body and solubilized in 8M urea. The rTm-CAP-1 with an 8 his-tag at C-terminus was purified with immobilized metal affinity chromatography (IMAC), as previously described [30].

Analyses of immunogenicity and protective efficacy of the T. muris proteins selected from immunoscreening
While the expression and purification of rTm-WAP49 was in progress, we evaluated the immunogenicity of rTm-WAP-F8+Na-GST-1 and rTm-CAP-1 in the AKR mouse model. We selected the AKR mouse model because these mice develop a type 1 immune response to T. muris challenge alone, leading to a high-burden of disease, yet retain the ability to develop a protective type 2 immune response if vaccinated with Tm-ES prior to challenge [13].
Separate groups of mice were immunized three times at two-week intervals with Tm-ES, rTm-WAP-F8+Na-GST-1 or rTm-CAP-1 using the Montanide ISA 720 adjuvant (Fig 2A). Montanide ISA 720 is a squalene-based water-in-oil emulsion used in multiple efficacious helminth vaccine animal studies, including the closely related Trichinella spiralis, and has demonstrated safety in humans [31][32][33]. Control groups included mice injected with PBS or adjuvant alone. Two weeks after the final immunization, serum levels of antigen-specific IgG1 and IgG2a were determined by ELISA to calculate end-point titers. Mice immunized with Tm-ES, rTm-WAP-F8+Na-GST-1, or rTm-CAP-1 exhibited significant levels of respective antigenspecific IgG1 as well as IgG2a titers relative to those in control groups of mice immunized with PBS or adjuvant ISA-720 (Fig 2A and 2B). However, antigen-specific IgG2a levels were significantly higher in mice immunized with rTm-CAP-1 relative to those in mice immunized with Tm-ES (Fig 2C). Consequently, mice immunized with rTm-CAP-1 had a significantly lower The predicted signal peptide of each protein is highlighted in red (SignalP 4.1). The Tm-WAP protein contains ten repeats of a whey acidic protein (WAP)-type four-disulfide core domain that were aligned using CLUSTAL W and prepared for display using BOXSHADE. Identical amino acids are shaded in black and similar amino acids in gray. This 50 amino acid domain contains 6 conserved cysteine residues. The 49 kDa N-terminus sequence was expressed in yeast as rTm-WAP49 (in red box). The repeat fragment #8 was expressed in E. coli as Tm-WAP-F8 (in yellow box) using an Na-GST-1 expression tag. Phylogenetic tree depicting the sequence identify of (B) Tm-WAP and corresponding gene TMUE_s0165000300 and (D) rTm-CAP-1 and corresponding gene TMUE (s0030008500), to their respective homologues in different Trichuris species. https://doi.org/10.1371/journal.ppat.1007273.g001 Trichuris muris whey acidic protein induces type 2 protective immunity against whipworm IgG1:IgG2a ratio relative to those immunized with either Tm-ES or rTm-WAP-F8+Na-GST-1 (Fig 2D). Next, an analysis of antigen-specific T cell response of mice immunized with the different antigens was conducted. Spleen cells were isolated and restimulated with the cognate proteins for 72 hours in vitro. In comparison to the mice in the two negative control groups (PBS or adjuvant control) cognate antigen-specific T cell responses, in terms of type 2 cytokine production, were significantly higher in mice immunized with Tm-ES, rTm-WAP-F8+Na-GST-1, and rTm-CAP-1 (Fig 2E). However, the production of type 1 cytokines IFN-γ and IL-12 in splenocytes was low and comparable between mice in the different vaccine groups, as well as the adjuvant control group. Importantly, production of type 2 cytokines IL-4 and IL-13 was significantly higher in mice immunized with either rTm-WAP-F8+Na-GST-1 or Tm-ES compared to that in mice immunized with rTm-CAP-1 (S1 Fig).
In addition to a strong humoral and cellular type 2 response generated by vaccination, a continued type 2 immune response is necessary during the "critical period" of worm expulsion (day 0-21 p.i.) [13]. We evaluated humoral and cellular immune responses post-challenge (anamnestic responses) by measuring the antigen-specific IgG1 and IgG2a levels in the serum along with T cell responses in splenocytes after restimulation in vitro with the cognate antigens. In mice vaccinated with Tm-ES and rTm-WAP-F8+Na-GST-1, relative to adjuvant control group, we observed a significantly higher type 2 immune response, characterized by high IgG1:IgG2a ratios and T cells producing significantly higher levels of IL-4 and IL-13 in the spleen (Fig 3C-3F). These antibody and T cell responses were significantly lower in mice vaccinated with rTm-CAP-1 when compared to those in mice vaccinated with Tm-ES or rTm-WAP-F8+Na-GST-1. Furthermore, post-challenge, we observed T cells producing significant levels of IFN-γ in the spleens from mice immunized with rTm-CAP-1 (S2 Fig).

Recombinant WAP protein(rTm-WAP49) induces strong type 2 protective immunity
The Tm-WAP protein, named rTm-WAP49 expressed in Pichia pastoris X33 was used to evaluate immunogenicity and protective efficacy in comparison to that by the WAP fragment (rTm-WAP-F8+Na-GST-1) following the immunization and challenge schedule shown in Fig  2A. To determine the contribution of the Na-GST-1 tag within the rTm-WAP-F8+Na-GST-1,

Fig 2. Immunogenicity of rTm-WAP-F8+Na-GST-1 and rTm-CAP-1 proteins in a murine model. (A)
Vaccination schedule to evaluate immunogenicity in male AKR/J mice. Humoral immune response was measured by end-point serum titer for antigenspecific (B) IgG1 and (C) IgG2a by ELISA to calculate the (D) IgG1 to IgG2a ratio for the mice in each group. Mice in groups injected with PBS or Montanide ISA 720 adjuvant were used as negative controls for each coating antigen. Statistical significance: Ã p<0.05, ÃÃ p<0.01 ÃÃÃ p<0.001, ÃÃÃÃ p<0.0001. (E) Cellular immune responses in terms of cytokines in the supernatants of splenocytes stimulated with media or cognate antigens, rTm-WAP-F8+Na-GST-1 (10 μg/mL), rTm-CAP-1 (10 μg/mL), or Tm-ES (50 μg/mL), for 72 hours. Values for cells cultured in media only were subtracted as background from those of cells stimulated with cognate antigens in the corresponding groups of mice. Values from duplicate wells for each treatment/stimulation were averaged for individual cytokines for the 5 mice in each group to calculate statistical significance. Individual cytokines and statistics are shown in S1 Fig  the recombinant Na-GST-1 protein was included as an additional control immunogen. Montanide ISA 720 adjuvant alone and Tm-ES were used as negative and positive control reagents, respectively. Endpoint titers of antigen-specific IgG1 (Fig 4A) and IgG2a (Fig 4B) in the serum samples rose significantly for all vaccine groups compared to adjuvant alone. While the cognate antigen-specific IgG1:IgG2a ratios in mice vaccinated with rTm-WAP-F8+Na-GST-1 or rTm-WAP49 were significantly higher compared to the adjuvant alone, these levels were lower relative to those in mice vaccinated with Tm-ES (Fig 4C).
Next, we compared the efficacy and immune correlates for rTm-WAP-F8+Na-GST-1 and rTm-WAP49 immunogens using the same vaccination and challenge schedule as described above. We also included a group of mice vaccinated with rNa-GST-1 to account for its potential contribution to protection observed for rTm-WAP-F8+Na-GST-1. In comparison to the Montanide ISA 720 adjuvant control, a significant reduction in worm burden was observed in mice vaccinated with rTm-WAP49 (48%, p = 0.000103) or rTm-WAP-F8+Na-GST-1 (38%, p = 0.00278), with no significant difference among these two groups. Importantly, vaccination with Tm-ES resulted in significant protection, while rNa-GST-1 was ineffective yielding no significant reduction in worm burden after challenge. These results suggest that the protective efficacy observed with the rTm-WAP-F8+Na-GST-1 against challenge with the T. muris eggs was related to the WAP fragment 8, and not the Na-GST-1 tag (Fig 5A).
Immune correlates for the protective efficacy by rTm-WAP49 and rTm-WAP-F8+Na-GST-1 included significantly higher antigen-specific IgG1 endpoint titers post-infection compared to that in mice immunized with the Montanide ISA 720 adjuvant alone (Fig 5B). However, rTm-WAP-F8+Na-GST-1 generated a higher antigen-specific IgG2a response (Fig 5C), leading to a significantly lower IgG1:IgG2a ratio relative to that observed in mice vaccinated with rTm-WAP49 (Fig 5D).
We also analyzed for post-challenge cellular immune correlates of protection in the spleen, the parasite-draining mesenteric lymph nodes (MLNs), and the vaccine-draining inguinal lymph nodes (ILNs). In mice vaccinated with rTm-WAP-F8+Na-GST-1 or rTm-WAP49, restimulation of spleen cells with cognate antigen showed high levels of IL-4, IL-5, and IL-6, indicative of type 2 immune responses (Fig 5E). Interestingly, we observed that the IL-13 response was significantly higher for rTm-WAP-F8+Na-GST-1 whereas the IL-9 response was significantly higher for rTm-WAP49, relative to each other (S3A and S3B Fig). Strong type 2 responses were also observed in the MLNs of mice vaccinated with rTm-WAP-F8+Na-GST-1 or rTm-WAP49 (Fig 5F), but IL-6 and IL-13 responses were significantly higher in rTm-WAP49 vaccinated mice compared to those vaccinated with rTm-WAP-F8+Na-GST-1 (S3C  and S3D Fig). Finally, in the inguinal lymph nodes (ILNs) of mice vaccinated with rTm-WAP49, we observed significantly higher type 2 immune responses in terms of higher production of cytokines IL-4, IL-5, IL-6, IL-9, and IL-13, compared to that in mice vaccinated with rTm-WAP-F8+Na-GST-1, but both these groups of mice exhibited significantly higher type 2 immune responses compared to control group of mice immunized with adjuvant alone (Fig  5G, S3E and S3F Fig). specific (C) IgG1 and (D) IgG2a by ELISA to calculate the (E) IgG1 to IgG2a ratio for the mice in each group. Mice in groups injected with Montanide ISA 720 were used as negative controls for each coating antigen. Statistical significance: Ã p<0.05, ÃÃ p<0.01 ÃÃÃ p<0.001, ÃÃÃÃ p<0.0001. (F) Cellular immune responses in terms of cytokines in the supernatants of splenocytes stimulated with media or cognate antigen, rTm-WAP-F8+Na-GST-1 (10 μg/mL), rTm-CAP-1 (10 μg/mL), or Tm-ES (

Tm-WAP is the immunodominant component of Tm-ES
Serum samples from mice immunized with the various T. muris-derived immunogens (rTm-WAP49, rTm-WAP-F8+Na-GST-1, and rTm-CAP-1) were tested for their antibody recognition of Tm-ES (Fig 6A). ELISA endpoint titers revealed that both rTm-WAP49 and rTm-WAP-F8+Na-GST-1 have significantly greater recognition of Tm-ES than Tm-CAP-1. Furthermore, antibody recognition of Tm-ES by rTm-WAP49 and rTm-WAP-F8+Na-GST-1 were the same as Tm-ES antisera. We performed western blot analyses of multiple T. muris products using serum samples from mice vaccinated with rTm-WAP49, rTm-WAP-F8+Na-GST-1, or Tm-ES (Fig 6B). We observed that mouse anti-Tm-ES serum originally used for the immunoscreening of the T. muris cDNA library clearly identified 49kd and 31kd protein bands corresponding to rTm-WAP49 and rTm-WAP-F8+Na-GST-1, respectively, and confirming Tm-WAP as an immunodominant antigen in the ES products of the T. muris adult worm. The mouse anti-rTm-WAP49 sera recognized rTm-WAP49 and the WAP fragment within rTm-WAP-F8+Na-GST-1, but not the rNa-GST-1 tag. Sera from mice immunized with rTm-WAP-F8+Na-GST-1 recognized rTm-WAP49 and rNa-GST-1. Importantly, sera from mice immunized with either Tm-WAP49 or rTm-WAP-F8+Na-GST-1 exhibited a similar recognition pattern of Tm-ES products, with several bands from 28-148 kDa range, corroborating findings that highly-conserved WAPs may be expressed as several secretory proteins of different molecular weights by both T. muris [27] and T. trichiura [24].

Native Tm-WAP protein localizes to the stichosome organ of T. muris
Over the past 20 years, the Trichuris spp. stichosome has been of particular interest due to its pathogenic role in facilitating its intercellular existence within the intestinal epithelial layer  [6,9,24,35]. To determine whether Tm-WAP is stichosome-derived, we used anti-rTm-WAP49 sera to perform fluorescent immunohistochemistry (IHC) of adult T. muris embedded Trichuris muris whey acidic protein induces type 2 protective immunity against whipworm in the caecal tissue (visualized by DAPI nuclear stain). Tissue cross-sections revealed an anti-Tm-WAP staining in the tissue-embedded stichosome of T. muris (Fig 7) with no discernable staining in the posterior, non-stichosome end of the worm. A predominance of the staining occurred in a ring-like, granular pattern (orange arrow) beneath the cuticle of the helminth. Directly adjacent to the outer layer of the helminth cuticle, we observed strong anti-Tm-WAP binding (white arrow), suggesting secretion of the WAP protein into the mouse caecal intraluminal space. Of note, in certain cross sections, there were delineated structures with granular anti-Tm-WAP staining (red arrow) within the stichosome. A similar pattern of variable distribution of antigen-specific secretory granules was previously described in the stichosome of T. spiralis [36].

Recombinant Tm-WAP proteins do not elicit antigen-specific IgE response
The generation of antigen-specific IgE antibodies secondary to vaccination or pre-existing in the population would likely impair the protective efficacy of immunization with that specific candidate vaccine [37]. In response to a systemic vaccination these circulating parasite-specific IgE antibodies can lead to a deleterious allergic response, as seen with the hookworm Na-ASP-2 vaccine [38]. Serum-specific IgE to total IgE ratio is a clinical indicator of allergen-specific response [39,40]. We observed that vaccination with either rTm-WAP49 or rTm-WAP-F8+Na-GST-1 did not lead to any significant serum-specific IgE to total IgE ratio compared to Montanide ISA 720 (S4A-S4C Fig). Ratios fell below a cutoff defined by the average values observed analyzing serum samples from control group of mice immunized with PBS. Furthermore, we evaluated mice for serum-specific IgE endpoint titers post-infection to the two recombinant WAPs by ELISA ( S4D  Fig). While IgE antibodies did react to Tm-Lysate, titers against the recombinant WAP proteins remained comparable to the negative control, Na-GST-1. Additionally, Tm-ES, which contains the immunodominant native Tm-WAP, did not induce an IgE response after vaccination nor was Tm-ES recognized by IgE antibodies in the sera of infected mice.

Discussion
With nearly a half-billion people afflicted, trichuriasis is one the most prevalent infectious diseases in the world, and yet there are limited prophylactic and therapeutic options. Using the T. muris infection of AKR mouse as a surrogate model of T. trichiura infection in humans, here we describe the identification of a whey acidic protein (WAP) as a highly expressed secretory protein from the stichosome of the worm that is effective in inducing a strong Th2 immunity and significant protection against oral challenge with T. muris eggs. The recombinant WAP protein (rTm-WAP49) as well as the highly conserved fragment 8 of WAP expressed as a fusion protein with Na-GST-1 tag (rTm-WAP-F8+Na-GST-1) were equally effective in inducing protective Th2 immunity.
Vaccination with either Tm-WAP (rTm-WAP49) or Tm-WAP fragment fusion protein (rTm-WAP-F8+Na-GST-1) led to a significantly reduced worm burden after T. muris infection. Interestingly, relative protection after vaccination with rTm-WAP49 was greater but not significantly different from that observed after vaccination with rTm-WAP-F8+Na-GST-1.
Antibody recognition of both rTm-WAP49 protein and the fragment rTm-WAP-F8 by Tm-ES antisera and the reciprocal recognition of native Tm-WAP in Tm-ES by anti-rTm-WAP49 and anti-rTm-WAP-F8+Na-GST-1 sera, suggest that both candidates, as well as Tm-WAP in the Tm-ES, contain epitopes with conserved antigenicity. Furthermore, using rTm-WAP49 serum, we localized the native protein to the stichosome organ of T. muris. Protection generated in mice vaccinated against Tm-WAP further supports previous studies suggesting that the stichosome plays a critical role in helminth pathogenicity [41,42]. The homologous stichosome-derived secretory Trichuris trichiura WAPs, Tt52 and Tt95, are thought to be part of a multigene WAP family, due to gene duplication and elongation events. Consistent with this finding in T. trichiura, multiple proteins from the T. muris ES are cross-reactive to anti-Tm-WAP49 and anti-Tm-WAP-F8+Na-GST-1 sera by western blot. Furthermore, Tt52 and Tt95 have been described for their ability to generate pores in a planar lipid bilayer [12,24,25]. Future studies testing the ability of Tm-WAP antisera to inhibit pore-formation, preventing stichosome penetration and maintenance in the colonic mucosa would be important to fully realize the vaccine potential of WAP.
Critical cellular mediators of protection against T. muris identified in our studies are the Th2 cytokines IL-4, IL-9, and IL-13. In this regard, it is well-established that the Th2 cytokine IL-4 is a critical mediator to enhance antibody responses [43], while IL-13 is recognized for its role in promoting worm expulsion through goblet cell expansion, mucin production, and smooth muscle hypercontractility [43,44]. Similarly, IL-9 mediates smooth muscle hypercontractility, as well as tissue mast cell maturation [44,45]. Given the important roles for these Th2 cytokines in the infection process of the worm, our results demonstrating the protective immunity elicited by the recombinant WAP protein, or the highly conserved fragment 8 of the WAP, strongly support their potential as vaccine candidates.
In contrast to the protective efficacy of the WAP, Tm-CAP-1, the second highly expressed protein identified in our studies from the excreted-secreted products of T. muris, was ineffective in inducing strong type 2 immune response as well as protection. The lack of protective efficacy for Tm-CAP-1 is in line with our results showing that antisera against this protein do not exhibit ELISA recognition of Tm-ES. On the other hand, the antisera from mice immunized with rTm-WAP49 and rTm-WAP-F8+Na-GST-1, suggesting that the native Tm-WAP protein is the immunodominant and major protective component of Tm-ES. This is in contrast to literature reports identifying the CAP-domain family (also known as SCP/TAPS) as a promising vaccine target due to its abundance in the secretome of whipworm and other soiltransmitted helminths [26,27]. Additional studies testing different immunization regimens and/or adjuvant formulations should inform whether or not the CAP-1 protein can induce Th2 immunity and protective efficacy.
Pre-formed IgE antibodies to recombinant Na-ASP-2 in vaccine recipients from hookworm endemic regions are thought to have induced generalized urticaria, halting the vaccine's development [46]. In this regard, after three immunizations with either of the two recombinant WAP proteins we did not observe the parasite-specific IgE to total IgE ratio to be above background levels. Furthermore, T. muris infection did not induce IgE antibodies against either recombinant WAP protein. Future studies evaluating the potential of the recombinant WAP proteins as vaccine candidates should determine whether natural infection with T. trichiura generates pre-formed IgE antibodies that recognize the recombinant Tm-WAP proteins.
The high degree of amino acid sequence identity for the WAP proteins between T. muris and T. trichiura also bodes well for WAP as a strong candidate for vaccine development strategies against T. trichiura. Further evaluation of the immunogenicity of these vaccine candidates with more potent Th2-promoting adjuvants, such as the alum-based adjuvants, along with optimizing the dose and the route of delivery (e.g. mucosal vs systemic), may augment protection. Additional Tm-ES derived antigens, such as those recently reported by Shears et al. [34], may be required in conjunction to Tm-WAP to achieve the same near-sterile immunity generated by Tm-ES vaccination.
In summary, here we report the identification of a novel immunogenic whipworm protein, Tm-WAP, a secretory, stichosome-derived protein. We have developed two recombinant derivatives of Tm-WAP, rTm-WAP-F8+Na-GST-1 and rTm-WAP49, which are both protective in the AKR animal model. Furthermore, we demonstrate that both candidates generate a strong type 2 immune response, similar to the highly protective Tm-ES. Future evaluations will require verification that Na-GST-1 fused with Tm-WAP-F8 retains protective efficacy as an individual candidate against hookworm, supporting the potential to developing pan-anthelminthic vaccine [29,47,48].

Ethics statement
All animal procedures were conducted in accordance with Baylor College of Medicine Institutional Animal Care and Use Committee (IACUC) approved protocol AN-6297 in compliance with the Animal Welfare Act, PHS Policy, and other Federal statutes and regulations relating to animals and experiments involving animals.

Generation of T. muris excretory and secretory (ES) products and their anti-sera
The T. muris ES products were obtained from the overnight culture of T. muris adult worms isolated from laboratory maintained STAT6/KO mice based on the established protocol [10,13,17]. The concentrated T. muris ES products containing the stichosome-secreted proteins (100 μg/mouse) formulated with Montanide ISA 720 were used to immunize AKR mice subcutaneously three times at 2 weeks interval. The antisera were obtained from immunized mice 10 days after the last immunization.

Immunoscreening of the T. muris adult cDNA library
Mouse anti-ES sera were used to immunoscreen T. muris adult cDNA library to identify immunodominant antigens based on the methods described previously [49]. Briefly, 5x10 4 plaques of the cDNA library made from T. muris adult worm were plated on each LB agar plate. The expression of T. muris adult proteins was induced and transferred by covering each plate with a 10 mM IPTG soaked nitrocellulose membrane. The blotted membranes were incubated with 1:2,000 dilutions of mouse anti-ES sera and then reacted with HRP-conjugated goat anti-mouse IgG (1:5000) (Invitrogen, USA). The putative positive clones were scored and picked for secondary screening with the same reagents until a single positive clone was obtained.
The DNA sequences of positive clones were obtained by double strand DNA sequencing using vector flanking primers, T3 and T7 promoter. Nucleotide and deduced amino acid sequences were compared to existing sequences by BLAST searching in the GenBank (http:// www.ncbi.nlm.nih.gov) and in the T. muris reference genome on WormBase ParaSite (http:// parasite.wormbase.org/Trichuris_muris_prjeb126/). MUSCLE 3.8.31 was used for multiple sequence alignment while PhyML 3.1 for phylogeny and TreeDyn 198.3 was used for phylogenetic tree rendering using one click analysis mode online at http://www.phylogeny.fr [50,51].

Expression and purification of Tm-WAPs and Tm-CAP-1
DNA encoding for the Tm-WAP49, comprising 70% of the immunodominant Tm-WAP (amino acids 24-393), was amplified without signal peptide from the total T. muris adult cDNA using forward (5'-GCGAATTCGCTAAAATAGGTTCATGTCC-3') and reverse (5'-ATGCGGCCGCTCAATGATGATGATGATGATGCTCGGCAGTGCTGTCGTCTCGGT-3') primers and subcloned into yeast expression vector pPICZαA (Invitrogen, USA) using EcoRI and XbaI restriction enzyme sites. The correct open reading frame (ORF) was confirmed by sequencing using the vector flanking primers corresponding to the regions encoding the α-factor and 3'AOX1. The recombinant plasmids were linearized following digestion with SacI and transformed into P. pastoris X33 strain by electroporation according to the manufacturer's instructions (Invitrogen, USA). Transformants were selected on zeocin-resistant YPD plates. The expression of recombinant Tm-WAP49 with 6His-tag at C-terminus were induced in medium with methanol in 10 L fermentation for 96 hours. The expressed recombinant Tm-WAP49 was purified with immobilized metal affinity chromatography (IMAC), as previously described [30].
For cloning Tm-WAP-F8+Na-GST-1, DNA coding for a Tm-WAP repeat fragment 8, the most conserved (69-96%) of the 10 repeats of Tm-WAP, was synthesized by GenScript. Tm-WAP-F8 was then fused with DNA coding for Na-GST-1 at N-terminus, a hookworm vaccine antigen with GST function that helps the solubilization of fusion protein [29,52] and cloned into pET41a by using NdeI/NdeI/XhoI sites. The Tm-WAP-F8+Na-GST-1 was expressed as soluble recombinant fusion protein in E. coli BL21 under induction of 1 mM IPTG and purified with immobilized metal affinity chromatography.

Electrophoresis and immunoblotting of recombinant Tm-WAP proteins
Samples were separated by 4-20% NuPAGE/MES gel (Thermo Fisher Scientific, USA) and either immediately stained with Coomaisse Blue for imaging analysis or transferred onto PVDF membrane (Thermo Fisher Scientific, USA). PVDF membranes were blocked with 5% (w/v) skim milk powder in PBST (PBS +0.05% Tween-20), and then incubated with sera (1:3,000) from immunized mice (rTm-WAP49, rTm-WAP-F8+Na-GST-1 or Tm-ES. HRPconjugated goat anti-mouse IgG (Invitrogen, USA) was used as a secondary antibody, followed by ECL substrate (GE Healthcare, USA) to develop antibody bound bands. The nitrocellulose membranes of anti-rTm-WAP-F8+Na-GST-1 and anti-Tm-ES were cut just prior to ECL substrate reaction between lanes 4-5 and 3-4, respectively, to allow for different length of reaction. Both T. muris adult homogenate and Tm-ES were used at 10 μg for SDS-PAGE and 2.0 μg for western blot. rTm-WAP49 was used at 2 μg for SDS-PAGE and 200 ng for western blot. Both rTm-WAP-F8+Na-GST-1 and rNa-GST-1 were used at 2 μg for SDS-PAGE and 100 ng for western blot.

Immunofluorescence staining of mouse cecum
Immunofluorescent staining for Tm-WAP in adult Trichuris muris was done following the M. O.M. Immunodetection Kit (Vector Labs, USA). In brief, SCID mice were sacrificed on day 42 post T. muris infection. Intestinal tissue with worms embedded were fixed in paraffin and sectioned. Primary staining was performed with PBS, naïve sera (1:100), or anti-rTm-WAP49 (1:100). Secondary staining was with FITC-conjugated goat anti-mouse IgG (1:500). DAPI was embedded in the mounting medium.

T. muris parasite and mouse model
T. muris E (Edinburgh) isolate, kindly provided by Dr. Joe Urban from the USDA, were maintained in susceptible STAT6K/O mice. At day 15 post-infection (p.i.) (42 days after initial vaccination), adult worms were gently extracted by forceps from the caecum of sacrificed mice. Day 15 p.i., which corresponds to the L2/L3 larvae [18], was selected due our observed peak intensity of infection in AKR mice. T. muris eggs collected from cultured T. muris adult worms were washed in sterile water and centrifuged at 500xg for 5 minutes and supernatant debris was removed from pelleted eggs. Collected eggs were stored in distilled water in culture flasks (VWR, USA) and allowed to embryonate at room temperature in the dark for 8 weeks. 300 embryonated eggs were separated into 100 μL aliquots. Each aliquot was counted by light microscopy to verify 300 ± 10 embryonated eggs. 6-7-week-old male AKR mice were purchased from Jackson Labs (USA) with 5 mice per group for immunogenicity and 13-15 mice per group for efficacy studies. Mice undergoing challenge were infected by oral gavage of 300 embryonated eggs in 100 ul of diH2O. Worm burden and anamnestic immunological response were assessed at day 15 p.i. Blood collection at various time points was collected by tail bleeding and the terminal bleeds were collected by cardiac puncture at the time of sacrifice. Mouse spleen, mesenteric lymph nodes, and inguinal lymph nodes were collected at the time of sacrifice, and immediately resuspended in a singlecell suspension for immunological analysis.

Cytokine analysis
Spleens, mesenteric lymph nodes (MLNs), and inguinal lymph nodes (ILNs) were obtained from mice two weeks after the third immunization (immunogenicity groups) or at additional 15 days after a subsequent challenge infection (efficacy groups). Tissues were disassociated into a single-cell suspension using a 100 μm cell strainer and resuspended in complete RMPI medium (RPMI 1640 with L-glutamine [VWR, USA], 10% heat inactivated FBS [VWR, Radnor, PA], and 1x pen/strep solution [Thermo Fisher Scientific, USA]). After a 5-minute centrifugation (300 x g), splenocytes were resuspended in 2 mL of ACK lysis buffer (Thermo Fisher Scientific, USA) for 5 minutes to remove red blood cells. After centrifugation cells were resuspended in complete RPMI media and incubated for 2 hours in a 37˚C + 5% CO 2 incubator before counting the number of live cells using a Cellometer Auto 2000 Cell Viability Counter (Nexelcom, USA). To have sufficient cells for multiple cytokine assays, MLNs and ILNs, were separately pooled into 3 sets of 5 mice per tissue.
Cells were seeded in a 96-well U-bottom culture plate (Corning, USA) at 1x10 6 cells per well in 250 μl medium and restimulated with either rTm-WAP49 (10 μg/mL), rTm-WAP-F8 +Na-GST-1 (10 μg/mL), rTm-CAP-1 (10 μg/mL), rNa-GST-1 (10 μg/mL), or Tm-ES (50 μg/ mL) at 37˚C, 5% CO2 for 72 hours. An unstimulated, media only negative control and a positive control of 20 ng/mL PMA and 1 μg/mL Ionomycin were performed concurrently. After 72 hours the cells were pelleted by centrifugation at 300 x g for 5 min and the supernatants were collected for measuring cytokine production. Concentrations of different stimulants and duration of stimulation were selected based on optimal pretested antigen to negative control ratio. Initially supernatant samples were tested using a custom Bio-Plex Pro Mouse Cytokine 10-plex kit (Bio-Rad, USA) for levels of IL-4, IL-5, Il-6, IL-9, IL-10, IL-12(p70), IL-13, IL-17A, IFN-γ and TNF-α. To evaluate accessory cytokines/chemokines associated with protection, subsequent experiments used the expanded Bio-Plex Pro Mouse Cytokine 23-plex kit, which including the original 10 cytokines, added Eotaxin, G-CSF, GM-CSF, IL-1α, IL-1β, IL-2, IL-3, IL-12 (p40), KC, MCP-1 (MCAF), MIP-1α, MIP-1β, and Rantes. To increase the sensitivity of the experiment, the kit was used in combination with DA-Bead plates (Curiox Biosystems, Singapore), as previously described [54]. Samples were run on a Luminex Magpix multiplex reader according to manufacturer's recommendations (Luminex, USA). Raw Luminex data were analyzed using the Bio-Plex Manager 6.0 software and plotted in GraphPad Prism 7.04 or R programming language. Sensitivity cutoff values for the detection of different cytokines were dependent on the included standards for each lot number of the Bio-Rad Bio-Plex kit. Cytokine values from media-only stimulated cells were deemed background and thus were subtracted from the values of antigen restimulated samples. For S1, S2 and S3 Figs, data was loaded into R programming language environment version 3.4.3 using Rstudio version 1.1.423 (http://www.rstudio.com/). Readxl package was used to import the data from the spreadsheet (https://CRAN.R-project.org/package=readxl). Tidyverse package (version 1.2.1) was used to transform the imported data for the analysis (https:// CRAN.R-project.org/package=tidyverse). Mean for each cytokine in a group was calculated by bootstrapping using smean.cl.boot function with default setting of Hmisc package (version 4.1-1) (https://CRAN.R-project.org/package=Hmisc) Wilcox test for each cytokine between groups was performed using coin package (version 1.2-2) and effect size was calculated using Cliff delta using effsize package [55,56]. Plot of the calculated results was generated using ggplot2 package (version ggplot2 2.2.1) [57].

Graphical presentation and statistical analysis
For all graphs, p-values were represented as Ã , ÃÃ , ÃÃÃ , and ÃÃÃÃ , corresponding to p-values of <0.05, <0.01, <0.001, and <0.0001, respectively. +Na-GST-1, or (C) rTm-CAP-1 for 72 hours and the cytokines were detected using a 10-plex assay kit. Mice in groups injected with Montanide ISA 720 were used as negative controls for each coating antigen. Cells receiving no stimulus (media only) were subtracted as background from stimulated cells from corresponding mice. Values from duplicate wells for each treatment/stimulation were averaged for individual cytokines for the 5 mice in each group to calculate statistical significance. Data are presented as means ± 95% confidence interval. Statistical significance: Ã p<0.05, ÃÃ p<0.01 ÃÃÃ p<0.001, ÃÃÃÃ p<0.0001. Effect size: d<0.2, d<0.5, d<0.8, and d>0.8, correspond to N (negligible), S (small), M (medium), and L (large), respectively. for 72 hours and the cytokines were detected using a 10-plex assay kit. Mice in groups injected with Montanide ISA 720 were used as negative controls for each coating antigen. Cells receiving no stimulus (media only) were subtracted as background from stimulated cells from corresponding mice. Values from duplicate wells for each treatment/stimulation were averaged for individual cytokines for the 13-15 mice in each group to calculate statistical significance. Data are presented as means ± 95% confidence interval. Statistical significance: Ã p<0.05, ÃÃ p<0.01 ÃÃÃ p<0.001, ÃÃÃÃ p<0.0001. Effect size: d<0.2, d<0.5, d<0.8, and d>0.8, correspond to N (negligible), S (small), M (medium), and L (large), respectively. (TIF)

S3 Fig. Post-challenge analyses of cytokine profile in the supernatants of cells from AKR mice immunized with rTm-WAP-F8+Na-GST-1 or rTm-WAP49 and challenged with T. muris.
Cytokines detected in the supernatants of (A, B) splenocytes, (C, D) mesenteric lymph nodes (MLNs), or (E, F) inguinal lymph nodes (ILNs) after stimulation with media only or cognate antigen, rTm-WAP-F8+Na-GST-1 (10 μg/mL) or rTm-WAP49 (10 μg/mL) for 72 hours were detected using a 23-plex assay kit. Mice in groups injected with Montanide ISA 720 were used as negative controls for each coating antigen. Cells receiving no stimulus (media only) were subtracted as background from stimulated cells from corresponding mice. Values from duplicate wells for each treatment/stimulation were averaged for individual cytokines for the 15 mice in each group to calculate statistical significance. Data are presented as means ± 95% confidence interval. Statistical significance: Ã p<0.05, ÃÃ p<0.01 ÃÃÃ p<0.001, ÃÃÃÃ p<0.0001. Effect size: d<0.2, d<0.5, d<0.8, and d>0.