Comparative genomic and functional insights into life-cycle plasticity and host adaptation in a wild Nippostrongylus brasiliensis strain from Guangzhou Authors: Xi Pan, Jinzhi Lu, Chuyue Wang, Yashan Song, Hongyi Li, Muchun Wan, Lihua Xiao, Yaoyu Feng, Dongjuan Yuan Published: 2026 ABSTRACT BackgroundNippostrongylus brasiliensis (rat hookworm) is a key model for studying human hookworm infection.ResultsA wild N. brasiliensis isolate from Guangzhou maintains high fecundity for 10 days and persists for 21 days in rats, which is twice as long as reported for reference strains, but survives only 10 to 11 days in mice while causing more severe pathology than that observed in rats. Comparative genomic, transcriptomic, and functional analyses were conducted to explore potential genetic factors associated with these phenotypic differences. Whole-genome sequencing revealed 92.6–93.3% identity to publicly available reference assemblies. Comparative analysis of strain-specific orthogroups identified astacin metalloproteases and four collagens, three being col-117-like genes, that were exclusively expressed in parasitic L4/L5 larvae. Two of these, col-117l12 and col-117l16, were up-regulated in mouse-derived worms compared to rat-derived worms. Lentivirus-mediated RNAi of col-117l16 reduced worm burdens by 77.2%, impaired reproduction, attenuated lesions, and disrupted adult cuticle structure and intestinal positioning.ConclusionsThese findings highlight candidate genes potentially associated with N. brasiliensis life-cycle plasticity and host specificity; however, direct causal attribution of these phenotypic traits to specific genetic factors requires further validation under controlled experimental conditions.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12864-026-13012-2. Background Human hookworm infection, caused predominantly by Ancylostoma duodenale and Necator americanus, is estimated to affect 406–480 million people in tropical and subtropical regions worldwide. The rodent-parasitic nematode Nippostrongylus brasiliensis (syn. Heligmosomum muris, Nippostrongylus muris) belongs to the superfamily Trichostrongyloidea. The life cycle and pathological features of N. brasiliensis closely resemble that of hookworms, which led to its designation as “rat hookworm”. During its life cycle, the eggs shed in feces hatch as first-stage larvae (L1). The worm develops to infective third-stage larvae (iL3) within 5–6 days after two molts. iL3 penetrates host skin and migrates to the lungs within 11–18 h, where larva undergoes molting into the fourth stage (L4). L4 larva ascends the trachea, is swallowed, and reaches the small intestine. A final molt produces adult worm that persists for egg-laying within 3–5 days; the entire cycle is completed in approximately 14–16 days, after which the worm is expelled from the rodent hosts [1, 2]. Due to its short life cycle and easy maintenance in laboratory rodents, N. brasiliensis has served as the principal laboratory model of human hookworm disease, offering a tractable, reproducible and ethically acceptable system in which to dissect skin-penetrating larval migration, intestinal blood-feeding, host immunity, tissue remodeling and anthelmintic drug development [1, 3]. The advent of lentivirus (LV)-mediated RNA interference (RNAi) has enabled functional analysis of key genes in N. brasiliensis, including β-tubulin isotype-1 (tbb-iso-1), secreted acetylcholinesterase B (ache-B), and fatty acid and retinol binding protein-1 (far-1) [4, 5], thereby establishing this parasite as a robust model for investigating nematode biology and host-parasite interactions [1, 6, 7]. N. brasiliensis strains from the USA, New Zealand, and UK have been widely reported [8–10]. However, investigating nematode biology on host adaptation of these strains was limited. In this study, an isolate from wild Rattus norvegicus in Guangzhou, China was identified with strong adaptation in rats through a range of methods, including morphological assessment, molecular identification, life-cycle profiling in rodents, fecundity assays, and whole-genome assembly. Integrated comparative genomic and transcriptomic analyses were conducted to identify candidate genes potentially associated with the prolonged life cycle and high fecundity of the isolate of N. brasiliensis from Guangzhou in rodents, and LV-mediated RNAi was subsequently employed to functionally characterize a candidate gene implicated in this life-cycle strategy. Methods Biological resources Specific pathogen-free SD rats and C57BL/6, BALB/c, KM, NIH, ICR mice were purchased from the Guangdong Experimental Animal Center. Wild Rattus norvegicus was captured on the campus of South China Agricultural University in Guangzhou, China. Worm collection and morphology observation Adult worms were collected from wild R. norvegicus captured in Guangzhou, China. Fecal eggs were cultured on filter paper moistened with 1% penicillin-streptomycin-amphotericin B at 26 °C in the dark. L1, L2 and L3 larvae were harvested at 24 h, 48 h and 7 days, respectively. L3 larvae were used for experimental infections. Five SD rats (8-week-old females) in each group received ~ 3,500 L3 larvae by subcutaneous abdominal injection; phosphate-buffered saline (PBS)-injected rats served as controls. Five mice (7-8-week-old females of the strains listed above) in each group received 500-1,000 L3 larvae subcutaneously in the neck. Eggs per gram (EPG) of feces were monitored throughout the experiments. L4 and the fifth stage (L5) larvae and adult worms were recovered from rat lungs at 3–4 days post-infection (dpi) and from rat intestines at 6 and 11 dpi, respectively. Worm morphology was examined with a stereomicroscope (Leica, Germany); length and width of the worms were measured in Image J. For scanning electron microscopy (SEM) observation, the worms were prepared as our previous study [4]. Histopathology Lungs and intestines were collected at early (dpi 5), peak (dpi 11) and late (dpi 22) patent stages, fixed in 4% paraformaldehyde, paraffin-embedded and sectioned [4]. Hematoxylin and eosin (H&E)-stained slides were digitized with an AxioScan.Z1 (Zeiss, Germany) and analyzed with ZEN software [4]. Molecular identification Single L3 larva was lysed in 10 µL distilled water plus 1 µL proteinase K (56 °C 15 min, 95 °C 10 min). Small subunit ribosomal ribonucleic acid (SSU rRNA) fragment was amplified with primers SSU-F (5′-AAACTGCGAACGGCTCAT-3′) and SSU-R (5′-CGACAAGGCAACAATACCC-3′) using Phanta Max Super-Fidelity DNA Polymerase (Vazyme, China) and sequenced (Sangon, China). A maximum-likelihood phylogeny of N. brasiliensis strains was constructed using SSU rRNAs with RAxML-NG v1.1 and Necator americanus SSU rRNA was used as the outgroup. Mitogenome and genome sequencing and assembly Total DNA was extracted from worms using the DNeasy Blood & Tissue Kit (Qiagen, Germany) and was sequenced on an Illumina NovaSeq 6000 (45.69 Gb) and PacBio Revio (34.92 Gb) platforms (Personalbio, China). Mitogenome assembly, annotation, and visualization were referred to our previous study [11]. The genetic distance of mitogenomes between the Guangzhou isolate and reference strains was calculated by GEME software. Nuclear genome assembly combined PacBio long reads (Hifiasm; parameters i = 3, n = 4, s = 0.1) and Illumina short reads. Redundant contigs were removed with Purge_dups; remaining contigs were polished with NextPolish and Pilon after BWA alignment of the Illumina sequencing reads to the contigs. Contigs were linked with SSPACE-LongRead and SSPACE [12], the resulting scaffolds were then ordered and oriented into chromosome-level assemblies with RagTag (v2.1.0) against the reference genome GCA_030553155.1 [13], and manually curated in Mauve (v2.4.0) [14]. Completeness was assessed with BUSCO (v6) and QUAST (v5.0) [15, 16]. Repetitive elements were identified using RepeatMasker (v4.1.6) against the RepBase (v24.07), complemented by de novo predication with RepeatModeler [17, 18]. Genome-wide gene structure annotation was performed with BRAKER3 [19]. Stage-specific RNA-seq data were integrated with the orthologs from N. brasiliensis and four closely related species of N. americanus, Ancylostoma caninum, Haemonchus contortus, and Caenorhabditis elegans to train and validate genome-specific gene prediction models. Within this framework, AUGUSTUS (v3.4.0) and GeneMark-ETP (v1.0) accurately defined translation start and stop sites, exon–intron boundaries, and complete coding sequences [20, 21]. The structural annotation information of protein-coding genes (PCGs) was integrated with EVidenceModeler [22]. Functional annotation used BLASTP against Swiss-Prot database (released February 2024), NR (released February 2024), KEGG (released May 2024) and GO (released July 2024) [23]. Synteny alignment of N. brasiliensis from Guangzhou and reference strain (GCA_030553155.1) was generated with JCVI and visualized with CIRCOS [24, 25]. A maximum-likelihood phylogeny of N. brasiliensis strains was constructed by single-copy orthologs with RAxML-NG v1.1 using N. americanus as the outgroup [26]. Orthologous gene families were analyzed in the genomes between the worm and reference strain (GCA_030553155.1) using OrthoFinder software and the expansion/contraction of genes were analyzed using CAFE v2.1 to obtain unique and expanded orthogroups [27, 28]. These genes in the orthogroups were functionally annotated with EggNOG-mapper [29]. Transcriptomics Eggs, L1-L5 larvae and adult worms (11 dpi from rats; 8 dpi from C57BL/6 and BALB/c mice) were collected for RNA-seq. Total RNA was extracted with TRIzol (Invitrogen, USA) and sequenced on NovaSeq 6000 (Illumina, USA). After quality control using FastQC (v0.11.9), low quality reads were filtered using Fastp (v0.23.2) [30]. Reads were analyzed with STAR (v2.7.10b), RSEM (v1.3.1), featureCounts and DESeq2 (v1.20.0) (|log2FC| ≥ 1, p < 0.05) [31–34]. Heatmaps and GO enrichment were produced with TBtools (v1.108) [35]. Genome-wide annotation of astacin metallopeptidases and collagens The domains of astacins and collagens in N. brasiliensis genome were identified by BLASTP using C. elegans corresponding domain sequences for queries. Sequences were extracted with TBtools after domain prediction by Pfam profiles. Phylogenetic analyses used the LG + I+G4 model with 1,000 bootstrap replicates using MEGA 11 and were visualized with iTOL v6 [36, 37]. Chromosomal locations were plotted with MG2C-v2.1 [38]. Cloning and identification of small interfering RNA (siRNA) Total RNA was extracted from adults and reverse transcribed to cDNA using the PrimeScript RT Reagent Kit (TaKaRa, Japan) according to the manufacturer’s protocols after removing contaminating genomic DNA with DNase I (Sigma, USA). The cDNA encoding NbCOL-117L16 protein was amplified using the primers of 5’- ATGGAGTCCGGATCTAGAGTTGCG − 3’ and 5’- GCTGTCAACGATACGCTACGTAACGGCATGACAGTG(T)12 − 3’ and Phanta Max Super-Fidelity DNA Polymerase (Vazyme, China). The Nbcol-117l16 gene was cloned into the pEASY-T1 vector (TransGen Biotech, China) and verified by sequencing. Five siRNAs were biosynthesized (GenePharma, China) and added into culture medium of adult worms to evaluate the effects of siRNAs on Nbcol-117l16 gene expression. Adults were divided into six groups: (i) wild-type (WT), (ii) 0.2 mM siRNA negative control (NC), (iii) 0.2 mM siRNA-col-117l16-139, (iv) 0.2 mM siRNA-col-117l16-663, (v) 0.2 mM siRNA-col-117l16-796, and (vi) 0.2 mM siRNA-col-117l16-891, and (vii) 0.2 mM siRNA-col-117l16-930. After 24 h of incubation with siRNAs, fifty adult worms with mixed sex in each group were washed three times with PBS buffer and harvested for quantitative real-time PCR (qRT-PCR). LV col-117l16-891 and col-117l16-930 administration Target gene-specific hairpin stem sequences were composed of siRNA sense sequences, loop sequences, siRNA antisense sequences, and polyA. The sense and antisense strands of shRNA-col-117l16-891 and shRNA-col-117l16-930 with BamH I and Xho I were biosynthesized (Sangon, China) and annealed using Annealing buffer for DNA Oligos, 5× (Solarbio, China). The sense and antisense strand sequences of shRNA-col-117l16-891 were 5’-TCGAGGAGAATTTCTCGTAAATCTTTCAAGAGAAGATTTACGAGAAATTCTCTTTTTTG-3’ and 5’-GATCCAAAAAAGTAGTGAAGTCGCTTACAATCTCTTGAATTGTAAGCGACTTCACTACC − 3’. The sense and antisense strand sequences of shRNA-col-117l16-930 were 5’- TCGAGGTAGTGAAGTCGCTTACAATTCAAGAGATTGTAAGCGACTTCACTACTTTTTTG-3’ and 5’-GATCCAAAAAAGAGAATTTCTCGTAAATCTTCTCTTGAAAGATTTACGAGAAATTCTCC − 3’. These sequences were cloned into the virus-encoding plasmid pGIPZ under the transcript of human miRNA30 and were designated as miRNA-adapted shRNA (shRNAmir). The primers of shRNAmir (5’-ACAGAATCGTTGCCTGCACA-3’ and 5’-GGTTGATTGTTCCAGACGCG-3’) were used to identify the shRNAmir sequences in the vectors of pGIPZ-col-117l16-891 and pGIPZ-col-117l16-930. VSV-G-pseudotyped, replication-incompetent LV particles in the HEK293T cell supernatant were prepared as previous study [4]. The virus was concentrated and stored at -80 °C in 1 mL aliquots. Functional virus titers were estimated as TCID50 (50% tissue culture infectious dose) using the Reed-Muench method. Parasite infection, recovery and transduction with LV N. brasiliensis was maintained in female SD rats, and iL3s were isolated and washed extensively in wash buffer (PBS buffer, 0.45% glucose, 100 U/mL penicillin, 100 µg/mL streptomycin, 100 µg/mL gentamicin) by centrifugation at 150 g for 3 min. iL3s were immersed in sodium hypochlorite with a final concentration of 0.1% for 2–3 min, after which exsheathment was assessed under a microscope. Approximately 2,000 activated L3s were exposed to LV, followed by subcutaneous injection into 7-week-old female SD rats for infection, as described in our previous study [4]. Rat lung and intestine tissues were fixed in 4% paraformaldehyde for H&E staining to analyze pathological changes. Reverse transcription PCR (RT-PCR) and qPCR Total RNA was extracted from recombinant LV using Simply P Total RNA Extraction Kit (BioFlux, USA) and total RNA was extracted from adult worms using Trizol (Invitrogen, USA). Total RNA was reverse-transcribed to cDNA using PrimeScript RT Master Mix (Takara, Japan) according to the manufacturer’s protocols after removing contaminating genomic DNA with DNase I (Sigma, USA). Total DNA of eggs from adult worms with LV treatment was extracted using DNeasy Blood & Tissue Kit (Qiagen, Germany). The PCR amplification of shRNAmir sequence from the cDNA of recombinant LV and total DNA of eggs was performed using One Taq DNA polymerase (New England Biolabs, USA) and the primers of shRNAmir. The qPCR amplification of Nbcol-117l16 gene in adult worms was performed using SYBR Green Real-Time PCR Master Mix (Toyobo, Japan) on a Light Cycler 480 Instrument II (Roche, Switzerland). The primers for qPCR of Nbcol-117l16 gene were 5’-GAGAATGCTCCTTCGACTCCC-3’ and 5’-TCCGTCCTCGAAGAATACACC-3’. The primers for the gapdh internal control gene were 5’-GCAGCAGACGGACCAATGAAGG-3’ and 5’-CACGAAGTTAGGGTTGAGCGAGATG-3’. The qPCR reactions were performed at 95 °C for 3 min denaturation; 50 cycles at 95 °C for 5 s, 60 °C for 10 s, and 72 °C for 40 s. PCR amplification was performed in triplicate for each sample. Expression was normalized to gapdh using 2−△△Ct. Hoechst staining Twenty adult worms were collected and washed extensively with PBS buffer. Specimens were incubated with Hoechst 33,258 (Solarbio, China) at a final concentration of 10 µg/mL for 30 min at room temperature. Unbound dye was washed by PBS buffer, and worms were subsequently visualized under Zeiss microscope equipped with a DAPI filter (Zeiss, Germany). Masson staining Three female worms were fixed in 4% paraformaldehyde, paraffin-embedded and sectioned [4]. Sections were stained using the Masson’s Trichrome Staining Kit (Beyotime, China) according to the manufacturer’s protocol. The stained slides were digitized with an AxioScan.Z1 (Zeiss, Germany) and the collagen fibers were analyzed using Image J. Detection of total collagen content Adult worms were weighed, snap-frozen, and homogenized in 6 M HCl (1:10, w/v, mg:µL) on ice for 5 min. Homogenates were neutralized with PBS buffer, resuspended in 200 µL, and sonicated. Samples were then mixed with 12 M HCl to a final concentration of 6 M HCl and hydrolyzed at 95 °C for 20 h. After cooling, hydrolysates were centrifuged (13,000 × g for 10 min) and diluted to 4 M HCl. Standards (S1-S7) and a blank control (S8) were processed identically. Total protein was quantified using a BCA Protein Assay Kit (Beyotime, China). For collagen measurement, 35 µL of each sample or standard was incubated with 75 µL assay buffer in a 96-well plate for 20 min with shaking, then 75 µL chromogen was added and incubated at 60 °C for 1 h. Absorbance was measured at 570 nm. Collagen concentrations were determined from a standard curve (0–300 µg/mL). SEM and transmission electron microscopy (TEM) For SEM and TEM observation, the worms were collected from rat intestine and processed as described in our previous study [4]. The samples for SEM observation were coated with platinum-palladium in an ACE 600 ion sputter (Leica, Germany) and observed in an EVO MA 15 (ZEISS, Germany). The samples for TEM observation were sectioned after drying and observed in model Talos F200S (Thermo Fisher Scientific, USA). Statistics Data were presented as mean ± SEM (worm burdens, EPG) or mean ± SD (all others). Two-group comparisons used unpaired t-tests and multiple comparisons used one-way ANOVA using GraphPad Prism 9. *p < 0.05; **p < 0.01; ***p < 0.001; n.s., not significant. Results Morphological characterization of the worm isolated from wild rodent in Guangzhou The fecal eggs (61–68 × 34–41 μm) of the worm developed into L1 larvae (236–413 × 16–23 μm) within 24 h, L2 larvae (436–631 × 25–32 μm) after 48 h, and infective L3 larvae (651–790 × 26–34 μm) by day 7 (Fig. 1Aa, Ab, Ah). Infective L3 larvae were then inoculated subcutaneously into Sprague-Dawley (SD) rats. L4 larvae (869–1318 × 34–46 μm) were obtained from the lungs, whereas adult worms were obtained from the small intestine: females measured 4243–4943 × 95–113 μm and males measured 3102–3782 × 79–103 μm (Fig. 1Ac, Af-Ah). Fig. 1The morphology and life cycle of N. brasiliensis from Guangzhou. A The morphology of egg, larvae and adult worms. Aa The elliptical egg from the infected rat feces. Ab Infective L3 larva stained with iodine. Ac L4 larva from infected rat lung. Ad L5 male larva with primitive copulatory bursa (pcb) from infected rat intestine. Ae L5 female larva with primitive vulva (pv) from infected rat intestine. Af Adult male with copulatory spicules (cs) and copulatory bursa (cb) from infected rat intestine. Ag Adult female with intrauterine eggs (e) from infected rat intestine. Ah Body length of worms across developmental stages. Black arrows indicated the esophageal-intestinal junction. Data was mean ± SD of 22 worms. B SEM observation of the morphology of L4 larva and adult worms from infected rat. Ba The whole body of early L4 larva. Bb The anterior end of L4 larva. Bc The whole body of adult male. Bd The anterior end of adult worm showing oral opening with three cuticular plates, and two amphid (a) and six labial papillae (lp) on the lateral side. Be The posterior end of adult female. Bf The copulatory bursa and copulatory spicule of adult male. C-D The curves of body weight and egg shedding of SD rats infected with 3500 L3 larvae by subcutaneous injection, n = 5 / group. E H&E staining observation of the pathological changes of infected rat lung and intestine at dpi 5 (the prepatent phase), dpi 11 (the patent phase), and dpi 22 (the post-patent phase). Black arrows indicated the worms in the villous surface and lumen of rat intestine. F-H The changes of villi and crypt of infected rats; the height of villi (F), the crypt depth (G), and the ratio of villus height to crypt depth (H) at dpi 11. I Comparison of life cycle of the worm from Guangzhou to reported N. brasiliensis strains The morphology and life cycle of N. brasiliensis from Guangzhou. A The morphology of egg, larvae and adult worms. Aa The elliptical egg from the infected rat feces. Ab Infective L3 larva stained with iodine. Ac L4 larva from infected rat lung. Ad L5 male larva with primitive copulatory bursa (pcb) from infected rat intestine. Ae L5 female larva with primitive vulva (pv) from infected rat intestine. Af Adult male with copulatory spicules (cs) and copulatory bursa (cb) from infected rat intestine. Ag Adult female with intrauterine eggs (e) from infected rat intestine. Ah Body length of worms across developmental stages. Black arrows indicated the esophageal-intestinal junction. Data was mean ± SD of 22 worms. B SEM observation of the morphology of L4 larva and adult worms from infected rat. Ba The whole body of early L4 larva. Bb The anterior end of L4 larva. Bc The whole body of adult male. Bd The anterior end of adult worm showing oral opening with three cuticular plates, and two amphid (a) and six labial papillae (lp) on the lateral side. Be The posterior end of adult female. Bf The copulatory bursa and copulatory spicule of adult male. C-D The curves of body weight and egg shedding of SD rats infected with 3500 L3 larvae by subcutaneous injection, n = 5 / group. E H&E staining observation of the pathological changes of infected rat lung and intestine at dpi 5 (the prepatent phase), dpi 11 (the patent phase), and dpi 22 (the post-patent phase). Black arrows indicated the worms in the villous surface and lumen of rat intestine. F-H The changes of villi and crypt of infected rats; the height of villi (F), the crypt depth (G), and the ratio of villus height to crypt depth (H) at dpi 11. I Comparison of life cycle of the worm from Guangzhou to reported N. brasiliensis strains L4 larvae are transparent, tapered at both ends, gently curved, and contain a conspicuous dark-brown intestine (Fig. 1Ac, 1Ba-Bb). L5 larvae exhibit incipient sexual differentiation. Males display a rudimentary copulatory bursa, while females possess a sharply curved tail; no eggs were found in the ovaries of these young adult females (Fig. 1Ad-Ae). Adult worms have three prominent lips around the mouth opening; the head region has two amphids and six labial papillae; the body surface has 14 longitudinal ridges. The cylindrical, translucent esophagus connects to a red, hemoglobin-filled intestine (Fig. 1Af, Ag, 1Bd). In males, the posterior end has a copulatory spine and a bursa comprising two large lateral lobes and a smaller dorsal lobe (Fig. 1Bc, Bf). In females, the posterior end has a pointed tail tip (Fig. 1Be), and the posterior ovary contains numerous eggs (Fig. 1Ag). These morphological features and the migratory route showed that the worm may be N. brasiliensis [1]. The prolonged life cycle of the worm in rats compared to reference strains Approximately 3500 L3 larvae were inoculated subcutaneously into the abdominal wall of rats. Body weight increased modestly compared to uninfected control (Fig. 1C). Fecal egg excretion began at 7 dpi, peaked at 3.6 × 105 eggs on 11 dpi, and ceased entirely by 21 dpi (Fig. 1D). Accordingly, the infection was staged as the prepatent phase (0–6 dpi), patent phase (7–15 dpi), and post-patent phase (15–21 dpi). During the prepatent phase, larvae were recovered from the lungs, which appeared dark-red, congested, and diffusely hemorrhagic. Bronchioles were densely infiltrated by lymphocytes and monocytes; alveoli were dilated, with their septa thickened by extravasated erythrocytes (Fig. 1E). In the small intestine, a few larvae encircled the villi, and the intestine was diffusely dark-red with mild oedema (Fig. 1E). During the patent phase, the lungs remained dark-red with scattered hematomas and retracted margins, but lesions showed signs of regression. In the intestine, large numbers of adult worms were observed occupying the villous surface and lumen from the distal duodenum to the proximal jejunum. The intestine was intensely hemorrhagic, congested, and edematous, with dense inflammatory infiltrates (Fig. 1E). In the post-patent phase, the lungs became light-red with smooth surfaces, and lesions showed significant improvement. Intestinal worm burdens declined, accompanied by amelioration of mucosal hemorrhage and edema improved (Fig. 1E). Villus height decreased, crypt depth increased, and the villus-to-crypt ratio declined progressively (Fig. 1F-H). The worm from Guangzhou exhibited the same life cycle and migration route as other N. brasiliensis strains. The 14-day ovulation period observed in rats exceeds the 3 ~ 5 days previously reported for reference strains in published studies (Fig. 1I and Supplementary Table 1). Infection with the worm from Guangzhou also caused predominant lesions in the rat lungs and intestines. The divergent host adaptation of the worm in mice Groups of NIH, ICR, KM, C57BL/6 and BALB/c mice were inoculated subcutaneously with 500 or 1,000 L3 larvae. In each strain, the higher dose of worms caused much more weight loss. The lowest body weight appeared at 3–5 dpi in C57BL/6, BALB/c and ICR mice, at 3 dpi in NIH mice, and 1 dpi in KM mice (Fig. 2A-B and Supplementary Fig. 1). Larvae and adults were recovered from the lungs and small intestine of all five strains, respectively. By 8 dpi, the lungs were hemorrhagic and congested, and the duodenum and jejunum were edematous in all five strains; lesions were most severe in C57BL/6 mice infected with 1,000 L3 larvae. As for fecal eggs, C57BL/6 and BALB/c mice shed eggs for 6 days, achieving peak counts of 23,000 EPG (C57BL/6, 9 dpi) and 18,000 EPG (BALB/c, 8 dpi) when infected with 1,000 L3 larvae, respectively (Fig. 2C-D). Other strains produced few or no eggs. Thus, C57BL/6 and BALB/c mice were the definitive hosts of the worm. However, the worms sustained a three-day period of relatively high fecundity in mice, which was shorter than that observed in rats and comparable to values previously reported for reference strains in mice in published studies (Supplementary Table 1). Fig. 2The development of N. brasiliensis Guangzhou strain in mice. A-D The curves of body weight and egg shedding of C57BL/6 (A, C) and BALB/c (B, D) mice infected with 1000 L3 larvae. E-F H&E staining observation of the pathological changes of lung and intestine in C57BL/6 (E) and BALB/c (F) mice infected with 1000 L3 larvae at dpi 8. G The worm burden in both infected mice. H-J The changes of villi and crypt of infected mice at dpi 8; the height of villi (H), the crypt depth (I), and the ratio of villus height to crypt depth (J). ns, p ≥ 0.05; ***p < 0.001 The development of N. brasiliensis Guangzhou strain in mice. A-D The curves of body weight and egg shedding of C57BL/6 (A, C) and BALB/c (B, D) mice infected with 1000 L3 larvae. E-F H&E staining observation of the pathological changes of lung and intestine in C57BL/6 (E) and BALB/c (F) mice infected with 1000 L3 larvae at dpi 8. G The worm burden in both infected mice. H-J The changes of villi and crypt of infected mice at dpi 8; the height of villi (H), the crypt depth (I), and the ratio of villus height to crypt depth (J). ns, p ≥ 0.05; ***p < 0.001 C57BL/6 and BALB/c mice inoculated with 1,000 L3 larvae showed severe weight loss, dyspnea and rough and dishevelled coat. As worms migrated from lungs to intestines, respiratory signs waned and body weights began to recover. At the ovulatory peak (8 dpi), lungs were dark-red with scattered ecchymoses and collapsed surfaces (Fig. 2E-F). Worms were concentrated in the distal duodenum and proximal jejunum, with higher worm burdens in C57BL/6 than in BALB/c mice, which was consistent with the EPGs in both mice (Fig. 2E-G). The intestine was hemorrhagic, congested and edematous; villi were blunted, desquamated and infiltrated by inflammatory cells (Fig. 2E-F). Crypt depth of infected mice increased while the ratio of villus height to crypt depth decreased (Fig. 2H-J). The worm infection caused more severe weight loss and damages to the intestines in mice than in rats, while the worm exhibited a longer life cycle and higher fecundity in rats. These observations appear distinct from phenotypes reported for reference strains in published studies, though direct comparison under matched conditions was not performed. Taxonomic identification of the worm by SSU rRNA, cox-1 and mitogenome sequences A 970 bp fragment of the worm SSU rRNA sequence was identical (100%) to N. brasiliensis (GenBank accession: KY462825.1) and grouped unambiguously with reference N. brasiliensis sequences in phylogenetic analyses (Fig. 3A and Supplementary Fig. 2). The complete mitogenome of the worm (GenBank accession: OR897706), assembled from Illumina sequencing of total DNA, is 13,351 bp long and contains 12 PCGs, 22 transfer ribosomal RNA (tRNA), 2 ribosomal RNA (rRNA), 2 non-coding regions (NCR), and an AT-rich region, but lacks atp8 (Fig. 3B and Supplementary Table 2). The worm mitogenome shared 94.35–94.72% nucleotide identity with reference strains. The genetic distance of the worm mitogenome from the three reference strains was 0.0557, however, it was 0.00 between reference strains. It shared 94.6% nucleotide identity with the known N. brasiliensis mitogenome (GenBank accession: AP017690.1), the 774 single nucleotide polymorphisms (SNPs) were concentrated in the AT-rich region (88.0% identity) (Supplementary Table 2). Among PCGs, nad2 and nad6 were the most divergent (93.9% and 94.4% amino-acid identity, respectively), whereas the remaining genes exceeded 98.7% identity (Supplementary Table 2). The worm cox-1 sequence extracted from mitogenome had 92.88–93.90% identity to the sequences from 6 reference strains, and sequence identity of cox-1 was 99.83–100% among 6 reference strains. The worm cox-1 sequence did not cluster with those of the reference strains into one group (Fig. 3C). Fig. 3The molecular identification and comparative genomic analyses of N. brasiliensis. A Phylogenetic tree was constructed by the worm SSU rRNA with known N. brasiliensis strains and Necator americanus as the outgroup. B Spatial distribution of mitochondrial genes and SNP sites of the worm in comparison with reference strain of N. brasiliensis. Outer circle denoted the distribution of mitochondrial genes, and inner circle indicated the distribution of variable sites compared to reference strain. C Phylogenetic tree was constructed by the worm cox-1 from mitogenome with known geographical strains of N. brasiliensis. D BUSCO-based completeness assessment of the genomes of geographical strains of N. brasiliensis.* the exact origins of these two genomes are unclear, but they were described by researchers from New Zealand and UK universities. E Collinearity analysis of the genomes between the worm versus reference genome (GCA_030553155.1). F Phylogenetic tree was constructed by 687 single-copy orthologs of the worm with known N. brasiliensis, Necator americanus, Ancylostoma duodenale, Ancylostoma caninum and Heligmosomoides polygyrus. G GO enrichment analysis of genes in unique and expanded orthogroups in the Guangzhou strain compared to reference genome (GCA_030553155.1). Red box indicated the GO terms of metallopeptidase activity, molting cycle, collagen and cuticulin-based cuticle, structural constituent of cuticle in unique and expanded orthogroups The molecular identification and comparative genomic analyses of N. brasiliensis. A Phylogenetic tree was constructed by the worm SSU rRNA with known N. brasiliensis strains and Necator americanus as the outgroup. B Spatial distribution of mitochondrial genes and SNP sites of the worm in comparison with reference strain of N. brasiliensis. Outer circle denoted the distribution of mitochondrial genes, and inner circle indicated the distribution of variable sites compared to reference strain. C Phylogenetic tree was constructed by the worm cox-1 from mitogenome with known geographical strains of N. brasiliensis. D BUSCO-based completeness assessment of the genomes of geographical strains of N. brasiliensis.* the exact origins of these two genomes are unclear, but they were described by researchers from New Zealand and UK universities. E Collinearity analysis of the genomes between the worm versus reference genome (GCA_030553155.1). F Phylogenetic tree was constructed by 687 single-copy orthologs of the worm with known N. brasiliensis, Necator americanus, Ancylostoma duodenale, Ancylostoma caninum and Heligmosomoides polygyrus. G GO enrichment analysis of genes in unique and expanded orthogroups in the Guangzhou strain compared to reference genome (GCA_030553155.1). Red box indicated the GO terms of metallopeptidase activity, molting cycle, collagen and cuticulin-based cuticle, structural constituent of cuticle in unique and expanded orthogroups Draft genome sequence of N. brasiliensis from Guangzhou The nuclear genome was sequenced to high coverage depth with Illumina HiSeq 2000 (118 ×) and PacBio Revio (30 ×). The assembled genome had 259.6 Mb arranged in 143 contigs across 6 chromosomes, with contig N50 value of 4.9 Mb and scaffold N50 value of 47.2 Mb, which substantially exceeded the contiguity of the reference genome (contig N50 = 1.6 Mb; GCA_030553155.1) (Table 1). The genome completeness (including single-copy and multi-copy genes) was comparable to the reference genome (GCA_030553155.1) using BUSCO assessment (Fig. 3D). Whole-genome alignment showed that six chromosomes had excellent collinearity with reference genome (Fig. 3E and Supplementary Table 3). High-quality transcriptomic data of the worms were obtained across developmental stages. Principal-component analysis showed tight clustering of the three biological replicates within each stage and clear separation between stages (Supplementary Fig. 3). Additionally, the mapped RNA-seq reads across developmental stages to the worm genome exceeded those obtained with reference strains (Supplementary Table 4). Table 1Comparative genomic analysis of the two N. brasiliensis strainsN. brasiliensis Guangzhou strainN. brasiliensis (GCA_030553155.1)Total length (Mb)259.6257.4No. of chromosomes66No. of contigs143379Contig N50 (Mb)4.91.6Scaffold N50 (Mb)47.245.6Repeats (%)39.9339.32GC content (%)43.2843.17Nucleotide sequence identity (%)93.28-Number of genes15,64015,654Number of coding proteins22,56522,090Total length of CDS (Mb)34.3031.74Avg_length of CDS (bp)1,5201,437GC content in CDS (%)49.9549.81Gene density (gene/Mb)60.2460.81Percent coding (%)13.2112.33No. of genes with introns14,36814,215Genes with introns (%)91.8791.04No. of tRNA10,8827,853Avg_length of tRNA (bp)7474Protein with putative signal peptide2,7592,663Protein with putative transmembrane domain4,2624,308 Table 1Comparative genomic analysis of the two N. brasiliensis strainsN. brasiliensis Guangzhou strainN. brasiliensis (GCA_030553155.1)Total length (Mb)259.6257.4No. of chromosomes66No. of contigs143379Contig N50 (Mb)4.91.6Scaffold N50 (Mb)47.245.6Repeats (%)39.9339.32GC content (%)43.2843.17Nucleotide sequence identity (%)93.28-Number of genes15,64015,654Number of coding proteins22,56522,090Total length of CDS (Mb)34.3031.74Avg_length of CDS (bp)1,5201,437GC content in CDS (%)49.9549.81Gene density (gene/Mb)60.2460.81Percent coding (%)13.2112.33No. of genes with introns14,36814,215Genes with introns (%)91.8791.04No. of tRNA10,8827,853Avg_length of tRNA (bp)7474Protein with putative signal peptide2,7592,663Protein with putative transmembrane domain4,2624,308 Comparative genomic analysis of the two N. brasiliensis strains The genome encoded 15,640 PCGs and 22,565 proteins (34.30 Mb; 13.21% of the assembly), of which 17,985 (79.70%) were annotated in Swiss-Prot, KEGG, Pfam or NCBI-NR (Supplementary Table 5). KEGG pathway profiling showed signal transduction, transport and catabolism, endocrine system and immune system as major functional categories (Supplementary Fig. 4). Phylogenetic analysis of 687 single-copy orthologs showed the three reference genomes were more closely related to each other than to the worm, while simultaneously resolving their underlying geographic substructure (Fig. 3F). Repetitive sequences of the assembled genome accounted for 39.93%, including retroelements 21.64 Mb (8.28%), DNA transposons 16.92 Mb (6.47%), which was slightly smaller than in the reference genome (GCA_900200055.1, 42.37%) and was comparable to the reference genome (GCA_030553155.1, 39.32%) (Supplementary Table 6). Thus, the worm was identified as N. brasiliensis Guangzhou strain. Comparative genomic analysis of the Guangzhou strain and reference strain To explore the genetic basis of the extended life cycle, we compared gene-family composition between the assembled genome and the reference genome (GCA_030553155.1). OrthoFinder recovered 12,094 orthogroups in the assembled genome. The genome of Guangzhou strain had 26 unique orthogroups (141 genes), including hookworm platelet inhibitor 1, zinc-dependent metalloproteases, myosin, and mitochondrial NAD-dependent malic enzyme compared to reference genome (Supplementary Table 7). Hookworm platelet inhibitor 1 had low expression level across developmental stages; several myosin domain proteins had relatively high expression in L1 and L2 larvae; one malic enzyme was relatively highly expressed in L3 larvae (Supplementary Table 8). Astacin metalloproteases showed a relatively high expression in the parasitic stages of L4 and L5 larvae (Supplementary Table 8). GO enrichment showed that the genes in the unique orthogroups were dominated by astacin metallopeptidases for molting cycle, collagen and cuticulin-based cuticle (Fig. 3G and Supplementary Table 9). The genome of Guangzhou strain had 37 expanded orthogroups (256 genes), mainly including cuticle collagen, and adenylyl/guanylyl cyclase, and transposases compared to reference genome (Supplementary Table 10). The adenylyl/guanylyl cyclase and transposases had low expression level across developmental stages (Supplementary Table 11), but cuticle collagens had high expression in free-living L1/L2 larvae, parasitic L4/L5 larvae, or all these four stages (Supplementary Table 11). GO enrichment also showed that these highly expressed cuticle collagens in the expanded orthogroups were structural constituent of cuticle (Fig. 3G and Supplementary Table 12). Stage-specific expression of astacins and cuticle collagens in the Guangzhou strain A genome-wide screen identified 88 astacin metallopeptidase genes harboring the conserved catalytic domain (CD; Pfam PF01400, ~ 200 residues) and four additional domains, CUB (complement C1r/C1s, Uegf, Bmp1), TSP (thrombospondin domain), ShK (Stichodactyla helianthus K channel toxin) and EGF_CA (calcium-binding EGF-like domain), positioned downstream of the CD (Fig. 4A). Phylogeny analysis showed that astacins could be separated into three groups. Genes with additional domain were mainly distributed in groups 1 and 2. Genes in these two groups had high expression across developmental stages, especially in L4/L5 larvae (Fig. 4A). In group 3, these genes had distinct expression patterns across developmental stages, including relatively high expression in L1/L2 larvae (subgroup 4) or in L4/L5 larvae (subgroups 1–3), and low expression across developmental stages (Fig. 4A). Notably, strain-specific astacins were concentrated in subgroups 1 and 3 of group 3, but had high expression in L4/L5 larvae (Fig. 4A). Fig. 4Phylogenetic trees of astacin metallopeptidases and collagens in the genome of N. brasiliensis from Guangzhou. A Phylogenetic tree of astacin metallopeptidases in the genome of N. brasiliensis from Guangzhou. B Phylogenetic tree of collagens in the genome of N. brasiliensis from Guangzhou. Heatmap indicated that expression pattern of astacin metallopeptidases and collagens across development stages. Red gene ID indicated astacin genes in unique orthogroup. Green gene ID indicated collagen genes in expanded group. * indicated four collagen proteins that cannot be identified as protein orthologs to C. elegans COL-117 Phylogenetic trees of astacin metallopeptidases and collagens in the genome of N. brasiliensis from Guangzhou. A Phylogenetic tree of astacin metallopeptidases in the genome of N. brasiliensis from Guangzhou. B Phylogenetic tree of collagens in the genome of N. brasiliensis from Guangzhou. Heatmap indicated that expression pattern of astacin metallopeptidases and collagens across development stages. Red gene ID indicated astacin genes in unique orthogroup. Green gene ID indicated collagen genes in expanded group. * indicated four collagen proteins that cannot be identified as protein orthologs to C. elegans COL-117 A genome-survey collected 215 collagen genes, containing a typical nematode-specific collagen domain at the N-terminus or one or more tandem collagen domains in Guangzhou strain (Fig. 4B). Chromosomal mapping showed that these collagen genes were mainly located on chromosomes I, II, and IV among 6 chromosomes. Phylogenetic analysis of 215 collagen proteins showed that a col-117-like branch comprised 38 proteins with the most reliable sequence identity with C. elegans col-117 and 4 proteins with the most reliable sequence identity with collagens of dpy-13, col-43 and col-119 among 188 collagens by TBtools. Then, these 38 COL-117-like proteins in col-117-like branch were named according to protein sequence identity with C. elegans COL-117 (Supplementary Table 13). Transcriptomic profiling showed stage-specific expression of these collagens in Guangzhou strain. Some genes were highly expressed in the free-living L1/L2 larvae, some had high expression in the parasitic L4/L5 larvae, whereas other genes were highly expressed across developmental stages (Fig. 4B). Furthermore, collagens within the expanded orthogroups were mainly found in branch 1 and the col-117-like branch. In these two branches, the genes from the expanded orthogroups had mainly high expression in L1/L2 larvae, but genes of g3899.t1, g9846.t1 (col-117-like-12, col-117l12), g10736.t2 (col-117-like-15, col-117l15) and g10382.t1 (col-117-like-16, col-117l16) were highly expressed in L4/L5 larvae. Transcriptomic analysis of developmental divergence of adult worms in rodent hosts Principal component analysis of the transcriptomes clearly separated mouse-derived worms (BALB/c and C57BL/6) from those recovered from rats (Fig. 5A). Consistently, differential expression analysis showed much more differentially expressed genes (DEGs) between mouse- and rat-derived worms than between the two mouse strains (Fig. 5B). Mouse-derived worms shared 219 down-regulated genes and 609 up-regulated genes compared with rat-derived worms (Fig. 5C). The down-regulated genes were enriched for GO terms related to ATP, nucleic acid and metal ion binding, as well as to embryonic or larval development (Fig. 5D). Most of these genes showed low expression across all developmental stages (Supplementary Table 14). Fig. 5The transcriptomic comparison of N. brasiliensis Guangzhou strain from mice and rats. A Principal components analysis of transcriptomic data in adult worms from rats and mice. B Number of DEGs in adult worms from rats, BALB/c and C57BL/6 mice. C Venn diagram of DEGs in adult worms from rats, BALB/c and C57BL/6 mice. D GO enrichment of up- and down-regulated DEGs comparing worms derived from mice and rats. E Heatmaps of up-regulated expression of collagens comparing worms derived from mice and rats. Red gene ID indicated col-117l12 and col-117l16 in expanded orthogroup The transcriptomic comparison of N. brasiliensis Guangzhou strain from mice and rats. A Principal components analysis of transcriptomic data in adult worms from rats and mice. B Number of DEGs in adult worms from rats, BALB/c and C57BL/6 mice. C Venn diagram of DEGs in adult worms from rats, BALB/c and C57BL/6 mice. D GO enrichment of up- and down-regulated DEGs comparing worms derived from mice and rats. E Heatmaps of up-regulated expression of collagens comparing worms derived from mice and rats. Red gene ID indicated col-117l12 and col-117l16 in expanded orthogroup Conversely, the up-regulated genes were enriched for GO terms related to mitochondrial oxidation-reduction processes, nitrogen compound metabolic process, structural constituent of ribosome, galactoside binding, transferase and heme binding activities, and cuticular collagens. Notably, highly expressed genes in adult worms included actin proteins, ATP carrier protein 1, elongation factors, globins, myoglobins, small ribosomal subunit proteins, and ten cuticular collagens (Fig. 5E and Supplementary Table 15). Of the 609 up-regulated genes, 15 were identified within unique and expanded orthogroups by comparative genomics. Three of 15 genes were highly expressed in L4/L5 larvae. Two encoded cuticular collagens g9846.t1 (col-117l12) and g10382.t1 (col-117l16) which were found among these three genes belonging to an expanded orthogroup (Fig. 5E). Thus, these data suggest that cuticular collagens are critical for adult worm development within the murine host. Nbcol-117l16 orchestrates cuticle assembly and determines parasite fitness NbCOL-117L16 contains a canonical N-terminal collagen domain followed by two tandem collagen domains (Supplementary Fig. 5). Nbcol-117l16 had stage-specific expression in L4 and L5 larvae (Fig. 6A). Two interference fragments of siRNA-930 and siRNA-891 could effectively reduce the mRNA level of Nbcol-117l16 gene by 44.3% and 55.9%, respectively (Supplementary Fig. 6A). A recombinant pGIPZ-shRNAmir LV was obtained to treat L3 larvae, and then L3 larvae were inoculated into SD rats subcutaneously (Supplementary Fig. 6B-D). Nbcol-117l16 expression was suppressed by 57.0%, 69.8%, and 33.6% in the adult worms of the F1, F2, and F3 generations in the siRNA-891 group, respectively, compared to empty-vector (EV) group (Fig. 6C and Supplementary Fig. 6E). No significant RNAi effect was observed in the siRNA-930 group (Supplementary Fig. 6F). Fig. 6Effects of LV-mediated siRNA of col-117l16 gene on the development of N. brasiliensis and pathological changes of rats after treatment. A Expression pattern of col-117l16 across developmental stages. B Location of five siRNAs on col-117l16 gene. C col-117l16 gene transcription in F1 of adult worms after LV treatment by qPCR analysis. D Effects of col-117l16 RNAi on the egg-shedding curve of adult females (n = 3). E Effects of col-117l16 RNAi on the changes of fecal eggs. F-H Effects of col-117l16 RNAi on the changes of egg hatchability (F), developmental rates of eggs to L3s (G) and developmental rates of L3s to adults (H). I Hoechst permeability assays of cuticle and Masson staining observation of collagen fibrils after col-117l16 RNAi. J Effects of col-117l16 RNAi on the total collagen abundance in adult worms. K Effects of col-117l16 RNAi on the area of collagen fibers in adult worms. L SEM and TEM observations of the morphological changes of adult worms (F2 generation). M H&E staining observation of the pathological changes of the lungs and intestines of infected rats after col-117l16 RNAi. N-O The villi height and the crypt depth of infected rats after col-117l16 RNAi. P The numbers of adult worms in each intestinal segment after col-117l16 RNAi. r: ridge, cz: cortical zone, cs: collagen strut, aes: amorphous electron-dense substances, mz: medial zone, bz: basal zone, e: epidermis. ns, p ≥ 0.05; *p < 0.05; ***p < 0.001 Effects of LV-mediated siRNA of col-117l16 gene on the development of N. brasiliensis and pathological changes of rats after treatment. A Expression pattern of col-117l16 across developmental stages. B Location of five siRNAs on col-117l16 gene. C col-117l16 gene transcription in F1 of adult worms after LV treatment by qPCR analysis. D Effects of col-117l16 RNAi on the egg-shedding curve of adult females (n = 3). E Effects of col-117l16 RNAi on the changes of fecal eggs. F-H Effects of col-117l16 RNAi on the changes of egg hatchability (F), developmental rates of eggs to L3s (G) and developmental rates of L3s to adults (H). I Hoechst permeability assays of cuticle and Masson staining observation of collagen fibrils after col-117l16 RNAi. J Effects of col-117l16 RNAi on the total collagen abundance in adult worms. K Effects of col-117l16 RNAi on the area of collagen fibers in adult worms. L SEM and TEM observations of the morphological changes of adult worms (F2 generation). M H&E staining observation of the pathological changes of the lungs and intestines of infected rats after col-117l16 RNAi. N-O The villi height and the crypt depth of infected rats after col-117l16 RNAi. P The numbers of adult worms in each intestinal segment after col-117l16 RNAi. r: ridge, cz: cortical zone, cs: collagen strut, aes: amorphous electron-dense substances, mz: medial zone, bz: basal zone, e: epidermis. ns, p ≥ 0.05; *p < 0.05; ***p < 0.001 The fecal EPGs decreased in the siRNA-891 group compared to the WT and EV groups (Fig. 6D). Eggs were round and shorter in length and exhibited a reduced aspect ratio (Fig. 6E and Supplementary Fig. 7A). Both egg hatchability and development to iL3 larvae were severely reduced in the col-117l16-891 group (Fig. 6F-G). Furthermore, the total worm burdens declined by 77.2% (Fig. 6H) in the col-117l16-891 group compared to the WT and EV groups. Adult worms were significantly shorter in col-117l16-891 group (Supplementary Fig. 7B). Hoechst permeability assays indicated increased cuticular permeability, while Masson staining revealed a drastic reduction in collagen fibrils (Fig. 6I-K). SEM showed increased cephalic wrinkling and disrupted annular and longitudinal ridges; TEM revealed longitudinal ridges with an irregular, thickened basal zone laden with amorphous electron-dense material (Fig. 6L). The body weight of infected rats remained unchanged (Supplementary Fig. 8A), but the lesions in the lungs and intestines of infected rats were alleviated (Fig. 6M and Supplementary Fig. 8B) and the villus height increased and the crypt depth decreased to the normal level (Fig. 6N-O) in col-117l16-891 group compared with the WT and EV groups. Adult worms were displaced anteriorly, localizing predominantly in the duodenum (Fig. 6P), which may be consistent with impaired migratory capacity. Thus, interfering with Nbcol-117l16 expression compromises collagen contents, producing an aberrant, fragile cuticle that curtails parasite survival, reproduction and migration within the host. Discussion The survival of parasitic nematodes within hosts depends on maintaining a dynamic equilibrium, wherein parasites must simultaneously evade immune attack, acquire nutrients, detoxify hostile environments, and synchronize their developmental program with host physiology. In the developmental stages, collagens confer tensile strength and elasticity to the cuticle, preserving worm integrity as parasites traverse hostile tissues [39]. Digestive proteases, cathepsins, metallopeptidases, peroxiredoxins and superoxide dismutases neutralize host immune responses and participate in modulating host-pathogen interactions. Up-regulated glycolysis, malate dismutation and rhodoquinone-dependent fumarate reductase enable strict anaerobic respiration in hypoxic niches, while specialized haem- and lipid-binding proteins facilitate nutrient acquisition. The differential equilibrium between parasites and hosts determines distinct parasitic adaptations, ultimately influencing nematode development and fecundity, as well as host pathological changes. Host-adapted biology of the Guangzhou strain The extended life cycle and enhanced fecundity in rats are consistent with a Guangzhou strain-specific developmental program that appears distinct from reference strains. It should be noted that these comparisons rely on life-cycle data reported in published studies for reference strains, rather than parallel experimental analyses conducted under matched conditions in the present study. Subcutaneous inoculation of L3 larvae initiates a hookworm-like migratory route. In rats, patency extends from day 7 to 21 (15 days) with a peak egg output of 360,000, substantially longer than the 6-day patent period reported for reference strains [1, 3] (Supplementary Table 1). In C57BL/6 and BALB/c mice, patency extends from day 6 to day 11 (6 days) with peak egg counts of 18,000 ~ 23,000, comparable to the 4-day patent period of the reference strain in mice [1] (Supplementary Table 1). By contrast, KM, NIH and ICR mice exhibit minimal or no patency. Susceptible C57BL/6 and BALB/c mice are thus suitable for mechanistic studies of helminth expulsion. Body-weight curves also showed the divergence of Guangzhou strain in both rodent hosts. Rats infected with Guangzhou strain had modest growth retardation, whereas all mouse strains infected with Guangzhou strain showed continuous weight loss, most precipitously during the pulmonary phase (days 3–5). C57BL/6 mice infected with reference strains reached their nadir at 2 dpi and returned to baseline body weight by 7 dpi [1]. Histopathological examination revealed pulmonary hemorrhage and emphysema, followed by intestinal hemorrhage and edema, in both rodent hosts infected with the Guangzhou strain. These pathological changes, granuloma formation and fibrotic nodules along the lung–intestine axis, have been reported for reference strains [3]. However, the severity of hemorrhage and edema in lungs and intestines was greater in mice than in rats. Thus, these observations suggest that infection with the Guangzhou strain produces sharply divergent fitness outcomes across rodent hosts, a pattern that differs from reports of comparable fitness of reference strains in both rodents. However, because these cross-study comparisons may be confounded by differences in host background, experimental protocols, environmental conditions, or long-term laboratory adaptation, direct attribution of these phenotypic differences to strain-specific genetic factors is not possible within the current comparative framework. Astacins and collagens in strain-specific orthogroups: the molecular engine of nematode cuticle formation A chromosome-level genome assembly of the Guangzhou strain was generated. Orthogroup analyses of this genome with the reference strain revealed Guangzhou strain-specific variants in astacin metalloproteases and cuticular collagens—key enzymes and substrates that orchestrate cuticle remodeling and molting. Humans possess six astacin protein orthologs that regulate fertilization, embryogenesis and inflammation [40]. C. elegans encodes 40 astacins that are essential for embryogenesis, cuticle synthesis, and molting, with their transcripts being most abundant in the hypodermis, digestive tract, and sensory support cells [41, 42]. Loss-of-function studies show that nas-35/dpy-31 deletion is embryonic-lethal due to exoskeletal defects [43], hch-1/nas-34 knockdown blocks hatching by preventing eggshell digestion [44], and nas-9 silencing causes partial embryonic lethality [45]. The nematode cuticle is an extracellular matrix composed primarily of collagens, cuticlins, lipids and chitin [46]. Soluble procollagens are proteolytically processed by astacins and other metalloproteases to generate highly cross-linked fibrils [47–49]. The Guangzhou strain encodes 215 collagen genes in its genome, which is more than the 188 in C. elegans and far exceeding the 28 in humans [50, 51]. Protein sequence identity analysis revealed that the Guangzhou strain harbors an expanded col-117-like branch compared to C. elegans collagens. Transcriptomic profiling revealed stage-specific expression programs: some collagens peak in free-living L1/L2 larvae, some in parasitic L4/L5 larvae, while a subset remained constitutively low. These patterns presumably fine-tune tissue remodeling across the life cycle [42, 52]. Surface armor and exoskeleton dynamics therefore rely on stage-specific collagens, cuticlins, and glycoproteins that confer mechanical strength and elasticity while enabling rapid molting. Moreover, these stage-specific collagen expression patterns may coordinate extracellular matrix turnover and immune evasion [53–55]. In this study, strain-specific astacins in the Guangzhou strain were highly expressed in parasitic L4/L5 larvae, underscoring their potential importance for parasitism in hosts. Twenty-nine collagens in expanded orthogroups were clustered into branch 1 and the col-117-like branch. Most of these 29 collagens were highly expressed in L1/L2 larvae; only four (g3899.t1 from branch 1, col-117l12, col-117l15, and col-117l16 from col-117-like branch) were highly expressed in L4/L5 stages. Thus, these astacins and three collagens in the col-117-like branch with high expression in parasitic L4/L5 stages may be associated with maintaining cuticle tensile strength and elasticity, potentially contributing to the parasitism of the Guangzhou strain in rodent hosts. Transcriptomic signatures of host adaptation Transcriptomic data of the worms derived from both rodent species showed that genes associated with embryonic or larval development were down-regulated, whereas genes regulating redox homeostasis, nitrogen metabolism, endopeptidase inhibition, adult lifespan and cuticle collagen synthesis were transcriptionally activated in mouse-derived worms compared with rat-derived worms. This concerted program potentially mitigates oxidative and immunological stresses from the host but forces parasites to reallocate resources from reproduction to cuticle maintenance and redox homeostasis. This finding is consistent with the severe pathological changes observed in infected mice, suggesting aggressive host immune responses to the worms. Compared to their counterparts in rats, worms in mice might undergo a trade-off: they gained enhanced cuticle maintenance and redox homeostasis at the cost of diminished fecundity and shortened patency. Among strain-associated genes identified by orthogroup analyses, 15 genes exhibited upregulated expression in mouse-derived worms, and only three genes (col-117l12, col-117l16, and g12781.t1) were highly expressed during parasitic stages. Col-117l12 and col-117l16 may be important for the survival of N. brasiliensis in mice. Functional interrogation of a key collagen LV-mediated RNAi targeting Nbcol-117l16 stably suppresses collagen expression across three generations and provided an experimentally tractable system to dissect collagen-dependent cuticle formation in parasitic nematodes. As in C. elegans, single-collagen mutations elicit readily scored phenotypes (Lon, Dpy, Rol, Bli, Ram, Sqt) that reflect discrete cuticular defects [50, 56]. Nbcol-117l16 was significantly downregulated upon silencing of the developmental regulator Nb-far-1 [4] and was upregulated in mouse-derived worms compared with rat-derived worms. These data implicate Nbcol-117l16 as a candidate contributor to cuticle formation in the parasitic stage of the Guangzhou strain—a function that may be relevant to prolonged life cycle and high fecundity in rats. Consequently, Nbcol-117l16 may serve as a molecular tool for dissecting the developmental mechanisms of parasitic nematodes. Limitations of the current comparative framework The phenotypic comparisons between the Guangzhou and reference strains rely on published literature rather than parallel experiments under matched conditions; consequently, observed differences in life-cycle duration, fecundity, and pathology cannot be definitively attributed to strain-specific genetic factors, as they may also reflect variation in host background, protocols, environment, or laboratory adaptation. Similarly, genomic comparisons are based on publicly available assemblies that may derive from strains with distinct passage histories. The identified orthogroups and strain-specific genes therefore represent associative candidates rather than validated causal determinants. Future studies employing controlled, side-by-side experimental comparisons of the Guangzhou strain and a reference strain under matched conditions will be essential to establish a robust causal link between the genomic differences identified here and the enhanced parasitism and fecundity reported for the Guangzhou strain. Conclusions The Guangzhou strain of N. brasiliensis exhibits superior parasitic adaptation in rats compared to mice. This observed adaptability is potentially associated with strain-specific expansions of astacin metalloprotease and certain collagen genes, particularly the col-117 family, which may contribute to worm adaptability within rodent hosts. In the absence of controlled parallel comparison with a reference strain under matched conditions, our findings highlight candidate genetic factors associated with nematode life-cycle plasticity and host specificity, pending further experimental validation to establish causal relationships. Supplementary Information Supplementary Material 1. Supplementary Material 1.