Mcpt8Cre mice on C57BL/6 background were bred and maintained in the Franz-Penzoldt Center in Erlangen and kept under specific pathogen free conditions. In Mcpt8Cre mice basophils are specifically and constitutively deleted as a result of Cre toxicity (15). C57BL/6 mice were obtained from Charles River Laboratories.

Animal experiments were approved by the Local Government of Lower Franconia and performed in accordance with German animal protection law and European Union guidelines.

For N. brasiliensis (Nb) infection mice were subcutaneously (s.c.) injected with 500 L3 stage larvae as previously described (15). To assess parasite fecundity, fecal egg counts were determined on day 7 post infection (p.i.) using a modified MacMaster counting chamber. Worm burden in the lung was analyzed on day 2 p.i. by enumeration of larvae that migrated out of the harvested lung tissue.

For preparation of Nb excretory/secretory proteins (NES) from L3 stage larvae or adult worms, the larvae were collected from the culture plates (L3) or intestine of infected mice (adults) and washed extensively (PBS/PenStrep). For collection of NES-containing supernatants, 10,000 larvae/mL for L3 or 100 worms/mL for adults were cultured in 1% glucose in PBS for 48-72h in 24-well plates at 37°C and 5% CO₂. NES was passed through a 0.2 µm filter and stored at -80°C until used. N. brasiliensis somatic extract (NEX) was prepared by homogenization of L3 larvae on ice with stainless steel beads in PBS (TissueLyser, Qiagen, Hilden, Germany) followed by centrifugation and recovery of supernatant.

For expression of His-tagged Nb-LSA1a protein, fullength cDNA was cloned in pcDNA3.1 (+) C-HA vector ( Supplementary Figure S1 ). Transient transfection of HEK293T cells was performed at a cell confluency of 70-90% using 20 µg plasmid and standard calcium phosphate transfection technique (250 mM CaCl₂ and HEPES-buffered saline). Supernatant containing His-tagged Nb-LSA1a protein was stored at -20°C until Ni-NTA purification (HisPur™ Ni-NTA Spin Columns, Thermo Fisher Scientific, Waltham, MA). For immunization experiments purified Nb-LSA1a protein or collected supernatant was used as indicated.

Female C57BL/6 or Mcpt8Cre mice were immunized with Nb-LSA1a (purified or supernatant), NES or control (buffer used for Ni-NTA purification of protein or supernatant from empty vector transfected HEK293T cells) by intraperitoneal injection (i.p.) with 200 µL Imject Alum (Thermo Fisher Scientific). Nb-LSA1a or NES protein was used at 5-10 µg/mouse for prime and 1 µg/mouse for boost immunizations. Immunizations were performed on day 0, then boosted on day 7, before infection with Nb on day 14.

NES samples (1-10 µg) were prepared in 5x Laemmli buffer without β-mercaptoethanol (non-reducing condition), heated (95°C, 5 min) and analyzed by SDS-PAGE. Following visualization of proteins using colloidal Coomassie Brilliant Blue, gel lanes were cut into 6 slices covering approx. the mass range between 10 to 95 kDa. Slices were washed and destained by alternatingly incubating them with 10 mM NH₄HCO₃ and 50% (v/v) acetonitrile (ACN)/10 mM NH₄HCO₃ (10 min at room temperature (RT) each). Cysteine residues were reduced (5 mM TCEP/10 mM NH₄HCO₃, 30 min at RT) and alkylated (50 mM 2-chloroacetamid/10 mM NH₄HCO₃; 30 min at RT) followed by proteolytic digestion of proteins using trypsin (60 ng per slice; overnight at 37°C). Peptides were eluted with 0.5% (v/v) trifluoroacetic acid (TFA)/50% (v/v) ACN, dried in vacuo, resuspended in 30 μl 0.1% TFA and desalted with in-house prepared STAGE tips prior to LC-MS analysis.

Reversed-phase liquid chromatography-mass spectrometry was performed using the UltiMateTM 3000 RSLCnano system (Dionex LC Packings/Thermo Fisher Scientific, Dreieich, Germany) coupled online to a Q Exactive Plus (Thermo Fisher Scientific, Bremen, Germany) instrument. The UHPLC system was equipped with two C18 μ-precolumns (Ø 0.3 mm × 5 mm; PepMap, Thermo Fisher Scientific) and an Acclaim PepMap™ analytical column (ID: 75 μm x 500 mm, 2 μm, 100 Å, Dionex LC Packings/Thermo Fisher Scientific). Peptides eluting from the LC column were transferred to a fused silica emitter for electrospray ionization using a Nanospray Flex ion source with DirectJunctionTM adaptor (Thermo Fisher Scientific) and applying a spray voltage of 1.5 kV and a capillary temperature of 200°C. The MS instrument was externally calibrated using standard compounds and equipped with a nanoelectrospray ion source and a stainless steel emitter (Thermo Fischer Scientific). MS parameters were as follows: MS scan range, m/z 375–1,700; resolution, 70,000 (at m/z 200); target value, 3 x 10⁶ ions; max injection time, 60 ms; TOP12-higher-energy collisional dissociation of multiply charged peptides; NCE of 28%; target value of 1 x 10⁵, maximum injection time of 120 ms; dynamic exclusion time of 45 s.

For this study, the MaxQuant 1.6.10.43 was used with the UniProt database for Nippostrongylus brasiliensis, Taxonomy ID 27835, (release 2020_05; 22636 protein entries). The precursor mass tolerance was set to 20 ppm for the first search and to 4.5 ppm for the main search. Trypsin was set as proteolytic enzyme (≤2 missed cleavages). Oxidation of methionine and acetylation of the protein N-terminus was allowed as variable modifications and cysteine carbamidomethylation as fixed modification. A false discovery rate (FDR) of 1% was applied on both peptide (on modified peptides separately) and protein lists. The mass spectrometry proteomics data have been deposited to the ProteomeXchange Consortium via the PRIDE (16) partner repository with the dataset identifier PXD035568.

The three-dimensonal structure of Nb-LSA1a was predicted using AlphaFold v2.0 (without homologous structure templates and using a selected portion of the BFD database) (17, 18). The prediction is colored by model confidence band and the accuracy of the AlphaFold model was scored as highly accurate with a predicted local distance difference (pLDDT>90) on a scale from 0 to 100.

NES and NEX samples (1-10 µg), purified Nb-LSA1a protein or supernatant of transfected cells was subjected to reducing and non-reducing SDS-PAGE using precast gels (10-12% Mini-PROTEAN TGX, Biorad, Hercules, CA) and blotted onto a PVDF membrane according to manufacturer´s instructions (Trans-Blot Turbo System, Biorad). Therefore, samples were prepared in Laemmli buffer containing either 5% (reducing) or no β-mercaptoethanol (non-reducing). Membranes were blocked in 5% milk powder in Tris-buffered saline (TBS) with 0.1% Tween-20 (TBST) overnight at 4°C, before being incubated with indicated mouse serum samples (1:10 dilution in 3% bovine serum albumin (BSA)/PBS) overnight at 4°C. After extensive washing in TBST, bound immunoglobulin was detected by incubation with HRP-conjugated anti-mouse IgG (Fcγ fragment specific, Jackson ImmunoResearch, Ely, UK), 1:5000 diluted in 5% milk powder/TBST for 1 h at RT. Alternatively, blots were incubated with rat anti-mouse IgE or rat anti-mouse IgG1 (SouthernBiotech, Birmingham, AL), 1:200 in 5% milk powder/TBST) for 2 h at RT, followed by HRP-conjugated goat anti-rat (Jackson ImmunoResearch), 1:5000 in 5% milk powder/TBST for 1 h at RT. For detection of the His-tag, the blot was incubated with polyclonal rabbit anti-His antibody (Cell Signaling, Danvers, MA), 1:1000 3% BSA/PBS for 2 h at RT. Detection followed by HRP-conjugated donkey anti-rabbit (Jackson ImmunoResearch, 1:5000 in 5% milk powder/TBST) and membrane was developed as above. For 2D SDS-PAGE, proteins are separated by isoelectric focusing (IEF) using precast gels (SERVAGel) prior to standard separation by size (SDS-PAGE). In contrast to standard SDS-PAGE, the used NES samples were desalted (Zeba spin columns, Thermo Fisher Scientific) and directly eluted in IEF sample buffer and loaded onto the gel. Subsequent western blotting was carried out as described above.

Detection of IgE and IgG1 levels in the serum of naïve and infected mice was determined as follows: Purified mouse anti-IgE (clone R35–72, BD Biosciences, Franklin Lakes, NJ) or a commercial IgG1 ELISA kit (SouthernBiotech) was used for coating. As secondary reagents anti-mouse IgE-AP or IgG1-AP (SouthernBiotech), followed by development with pNPP substrate (SouthernBiotech) was applied. For detection of parasite-specific IgE or IgG1, a 10-20 μg/mL NES protein suspension ( Supplementary Figure S2 ) was coated on 96-well polystyrene plates overnight (4°C), blocked with 3% BSA/PBS for 2 h and then incubated for 2 h with serum dilutions. Parasite-specific antibodies were determined using the secondary reagents described above. For Nb-LSA1a-specific ELISA, 96-well polystyrene plates were coated with a 10-20 µg/mL Nb-LSA1a suspension. Absorption was measured at 405 nm on a Multiskan FC photometer (Thermo Fisher) and blank wells were used for background subtraction.

Statistical analysis was performed with Sigmaplot (Version 12.3, Systat Software) using Mann-Whitney U-test. Data is always indicated as mean + standard error (SEM). Levels of significance: *p < 0.05, **p < 0.01. n.s. = not significant.

Infective L3 stage larvae of Nb secrete a large variety of proteins and other molecules (collectively termed Nippostrongylus brasiliensis excretory/secretory products, NES) some of which may play a critical role for entry of L3 larvae into the host organism via the skin barrier and for successful establishment of parasitism within their hosts. We therefore reasoned that identification of immunogenic proteins in NES could help to develop a vaccination strategy and dissect the mechanisms of protective immunity against the early stage of infection in the skin.

As a first step we determined total and NES-specific IgE and IgG1 levels in the serum after primary and secondary Nb infection of mice on C57BL/6 background. While total IgE and IgG1 levels increased after primary infection, we could not detect NES-specific IgE or IgG1 in the serum by ELISA ( Figure 1A ). This could be due to bystander activation of unspecific B cells or production of low-affinity antibodies. However, after secondary infection NES-specific IgE and IgG1 antibodies were clearly detectable by ELISA ( Figure 1A ). Next, we performed Western blot analysis. There was no antigen-specific IgG1 or IgE response to NES detectable in the serum of naïve mice while the serum after primary Nb infection showed a faint staining for secretions of adult worms (L5) and preparations of whole worm homogenates at ~100 kDa and above 180 kDa ( Figure 1B and Supplementary Figure S3 ). This may indicate that only some antigen multimers are detected by low-affinity antibodies after primary infection. However, a discrete and overlapping set of NES antigens was recognized by both IgG1 and IgE antibodies from secondary Nb-infected mice with a broad signal between 45 and 55 kDa and additional signals at 70 kDa and above 180 kDa ( Figure 1B ). Importantly, this discrete band pattern was only detectable in non-reducing conditions which preserves inter- and intramolecular disulfide bonds of the proteins. We then further separated NES proteins by size and charge using two-dimensional gel electrophoresis (2D SDS-PAGE) followed by Western blotting to reveal the complexity of the detected NES antigens. We identified several spots at about 55 or 70 kDa separated by the pH gradient suggesting that the detected antigens consist of numerous proteins with similar size but different charge ( Figure 1C ). Interestingly, the 70 kDa spots basically mirrored the charge-based distribution of the 55 kDa spots. This may indicate differences in glycosylation although further analysis would be required to confirm this assumption.

To further analyze NES components and identify individual antigens, we performed Liquid Chromatography Mass Spectrometry (LC-MS/MS) of eluted gel slices in the area of interest based on the Western blot analysis. We identified a total of 76 proteins, of which the top 25 most abundant proteins are listed in Figure 2A . Only three proteins showed a match with already described proteins exhibiting peptidase activity (legumain, aminopeptidase), as well as a protein disulfide-isomerase, while all others were uncharacterized proteins. Some of the uncharacterized proteins contained domains found in serine proteases (PF05577), histidine phosphatases (PF00328), copper-binding tyrosinase (PF01549), or macroglobulin (PF01835). ID68, a macroglobulin-related protein, may confer endopeptidase inhibitor activity. ID52 contains a Serpin domain (PF00079), characteristic for serine protease inhibitors whose role in nematodes is still poorly defined. ID70 contains a CUB domain (PF00431) often found in peptidases. The most frequently represented group of proteins in the NES products belonged to the SCP/TAPS superfamily (19). This superfamily also contains members of the Venom Allergen-Like (VAL) or Ancylostoma Secreted Protein (ASP) families, which are very abundant in helminth secretions (20, 21). However, molecular targets and functions remain largely elusive. The core of helminth VALs consists of CAP domains with characteristic Cysteine-rich regions (PF00188). Blast annotation and domain analysis showed that at least five proteins contained single or double CAP-domains. Interestingly, one of the proteins (ID78) with a single CAP domain is closely related to C. elegans Venom-Allergen-like protein 1 (vap-1).

ID72 was the most abundant protein in all NES preparations. BLAST search against the Nb genome on the WormBase ParaSite database revealed that this 16,696 kDa protein with 144 amino acids (aa) is encoded by the gene NBR_0001642601 with seven exons. Additionally, this gene has 5 uncharacterized paralogues in the Nb genome, two of which were also detected in our LC-MS/MS analysis ( Figure 2A ). We termed this protein family Nb-LSA1 for “Nippostrongylus brasiliensis larval secreted protein 1”, and assigned Nb-LSA1a to ID72. Then, Clustal Omega (22) was used to align the Nb-LSA1a protein sequence with the other 5 family members ( Figure 2B ). Nb-LSA1a is most closely related to Nb-LSA1b (97% protein sequence identity and same size, encoded by the gene NBR_0001642701). The protein sequence identity of Nb-LSA1a to the other family members is only 19-29%. The sizes of these proteins are: 27,6 kDa (Nb-LSA1c, encoded by NBR_0001642501), 35,8 kDa (Nb-LSA1d, encoded by NBR_0001642801), 23,1 kDa (Nb-LSA1e, encoded by NBR_0002055701) and 17,2 kDa (Nb-LSA1f, encoded by NBR_0000291501). A signal peptide motif (first 17-aa) is only present in four family members and missing in Nb-LSA1c and Nb-LSA1f.

The sequence of the initially identified protein Nb-LSA1a was then used to search for homologues in other nematode species using the SWISS-MODEL database (23). A sequence similarity of 21,15% was found for the dog hookworm protein Ac-ASP-7, and 14,29% similarity for the human hookworm protein Na-ASP-1. Although the algorithm used for conserved domain search in Figure 2A did not identify a CAP motif for Nb-LSA1a, the result of the homology analysis and the known sequence diversity of the CAP domains supported the idea that Nb-LSA1a might indeed contain a CAP domain. To investigate this more closely, Nb-LSA1a was subjected to comparative analysis with ASPs of known structure, namely Ac-ASP-7 (PDB entry 3s6s), Na-ASP-1 (PDB entry 3nt8) and Na-ASP-2 (PDB entry 1u53). By comparing the sequence and structural features, conserved CAP sequence motifs could be identified in Nb-LSA1a ( Figure 2C ).

The CAP motifs CAP1, CAP2 and CAP3, which are relatively well conserved between the so far known CAP domain-containing proteins, are also present in Nb-LSA1a. Sequence alignment furthermore showed that Nb-LSA1a does not contain a CAP4 motif and is also missing the cysteine-rich region. This region is only weakly conserved and is not a central component for the 3D-structure of the CAP domain. Furthermore, the most likely structure for Nb-LSA1a was generated using AlphaFold prediction algorithm (24) and compared to the known crystal structure of Na-ASP-1 which is composed of two CAP domains ( Figure 2D ). The accuracy of our AlphaFold model was scored as highly accurate with a predicted local distance difference test (pLDDT) >90% ( Supplementary Figure S4 ). The arrangement of α-helices and β-strands of Nb-LSA1a clearly resembles one CAP domain of Na-ASP-1. Therefore, it appears that Nb-LSA1a is a CAP domain protein.

To further characterize the immunogenicity of Nb-LSA1a, we expressed a C-terminally His-tagged version in HEK293T cells and first performed Western blot analysis of supernatants with anti-His antibodies. Under reducing conditions (+β-ME) Nb-LSA1a appeared as a dominant band of approximately 17 kDa ( Figure 3A ). However, under non-reducing conditions (-β-ME), the 17 kDa band was almost gone and three other bands at around 30-40 kDa appeared ( Figure 3A ). This suggests that Nb-LSA1a is actually expressed as dimer/trimer.

Next, we addressed the question whether Nb-LSA1a is indeed recognized by immune serum from Nb-infected mice. As expected, no bands appeared when blots were hybridized with serum from naïve mice. In contrast, serum isolated from mice after secondary Nb infection showed basically the same staining pattern as the anti-His antibodies ( Figure 3B ). ELISA analysis further revealed that Nb-LSA1a-specific IgG1 and IgE is generated in Nb-infected mice and both antibody levels increased after secondary as compared to primary infection ( Figure 3C ). Based on the strong humoral immune response against Nb-LSA1a we further investigated whether immunization of mice with Nb-LSA1a could protect against Nb infection.

To determine whether immunization of mice with Nb-LSA1a is sufficient to protect against Nb infection we performed experiments using a standard intraperitoneal immunization protocol with alum adjuvant ( Figure 4A ). In brief, mice were immunized with Nb-LSA1a or NES in alum on day 0 and 7, infected with Nb on day 14 and analyzed 7 days after infection. While Nb infection elicited similar levels of total IgG1 serum antibodies in all groups of mice, Nb-LSA1a-specific IgG1 was only present in the serum of Nb-LSA1a-immunized mice ( Figure 4B, C ). IgG1 in the serum of Nb-LSA1a-immunized mice also bound to NES-coated plates, which confirms that Nb-LSA1a is a prominent antigen in the whole secreted protein mixture ( Figure 4D ). Unexpectedly, we did not detect a significant increase of anti-Nb-LSA1a IgE ( Figures 4E–G ).

Importantly, Nb-LSA1a-immunized mice showed strongly reduced egg burden in fecal pellets, similar to NES-immunized mice ( Figure 4H ). Previous studies have shown that basophils contribute to protection against secondary Nb infection. Hence, we decided to compare the protective effect of Nb-LSA1a immunization in wild-type and basophil-deficient Mcpt8Cre mice (15). Egg counts in fecal pellets of Nb-LSA1a-immunized Mcpt8Cre mice were similar to egg counts from non-immunized wild-type or Mcpt8Cre mice ( Figure 4I ). This was not due to an impaired anti-Nb-LSA1a IgG1 response in Mcpt8Cre mice (data not shown). To further analyze whether this protective basophil-mediated effect occurs already in the skin as the first anatomical site of infection, we determined the number of larvae that reached the lung on day 2 after infection. While non-immunized wild-type or Mcpt8Cre mice contained about 200 larvae, this number was reduced to about 50 larvae only in Nb-LSA1a-immunized wild-type mice ( Figure 4J ).

Overall, these data indicate that immunization with Nb-LSA1a, a newly identified secreted protein of Nb L3 larvae, elicits a strong IgG1 response and provides basophil-mediated protective immunity against Nb infection mainly in the skin or before they reach the lung. This finding illustrates that secreted proteins of the free-living larval stage can have important and yet to be determined functions for migration and survival within the infected host.