All animal experiments were conducted in strict accordance with China regulations on the protection of experimental animals (Regulations for the Administration of Affairs Concerning Experimental Animals, version from July 18, 2013) and specifically approved by the Institutional Animal Care and Use Committee of Xiamen University, China (Permit Number: 2013–0053).

Tumor cell line HeLa was obtained from the Han’s Lab, Xiamen University (Xiamen, China). HEK-293 T cell was conserved by the State Key Laboratory of Cellular Stress Biology, Xiamen University, China. They were incubated in DMEM (Dulbecco’ s modified Eagle’ s medium; HyClone, United States) with 10% fetal bovine serum (HyClone, United States) at 37°C, 5% CO₂. Hydrogen peroxide was obtained from Sinopharm (China). NAC was obtained from Beyotime (China). Cobalt chloride was obtained from Sigma (United States). U0126 was obtained from Selleck Chemicals (United States). YC-1 and Trolox were obtained from MedChemExpress (United States).

The parasite isolate used in this study was obtained from Hulunbeier Pasture of Inner Mongolia of China and maintained by in vivo propagation of the parasite material in mice [supplied by Xiamen University Laboratory Animals Center, (XMULAC), China] (Tang et al., 2004). In vitro cultivation of metacestode vesicles was performed using HeLa conditioned medium according to a previously established protocol (Spiliotis and Brehm, 2009). Normoxia or hypoxia conditions were maintained at 37°C in the incubator with 20% O₂ and 5% CO₂ or 94% N₂, 5% CO₂, and 1% O₂. Unless otherwise described in the text, all experiments of vesicle were performed after 48 h incubation with HeLa medium (HM). For the growth assay, vesicles (diameter ≤ 1 mm) were manually picked up and cultured with HM including beta mercaptoethanol (β-Me) in 6-well cell plates supplemented with different reagent. Parasite growth was determined by the measurement of vesicle’s diameter every 7 days. Each group contains at least 2 replicates and more than 80 vesicles in total for each group were analyzed. Two-three independent experiments were performed.

Unless otherwise described in the text, metacestode vesicles and protoscoleces were incubated with 50 μM of EdU for 4 h and wholemount prepared according to Cheng and colleagues (Cheng et al., 2015). Click-iT-EdU Alexa Fluor 555 Imaging Kit (Life Technologies, Shanghai, China) or BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 488 (Beyotime, Shanghai, China) was used for the detection of EdU. DNA was counterstained with 4′, 6-diamidino-2-phenylindole (DAPI) (Sigma, United States) for all labeling experiments. For the quantification of EdU⁺ cells in metacestode vesicles, 3–5 random microscopic fields per vesicle from 6 to 10 vesicles were captured and the positive cells were manually counted. For the quantification of EdU⁺ cells in protoscoleces, the protoscoleces were photographed and the image with the largest number of EdU⁺ cells was taken for counting. At least 2 labeling experiments were performed and analyzed for each control and treatment group.

For the frozen sections, liver tissue with lesions were excised form infected mice and enbedded in OCT (Optimal Cutting Temperature, Sakara, United States). Serial sections were performed with Leica cryostat (Leica Biosystems, Germany) and mounted onto slides. ROS were determined by the BBoxiProbe™ Frozen Section ROS Detection Kit (BestBio, China). EdU were determined by the BeyoClick™ EdU Cell Proliferation Kit with Alexa Fluor 594 (Beyotime, China). After treatment, total parasite lesions were excised and weighed.

There were five mice per group in lesion weight experiments, and one animal was excluded with no detectable lesion. For EdU experiment, there were five mice in the NAC group and three mice in the saline group. For the quantification of EdU⁺ cells in mice, 3–5 microscopic fields per mouse were captured and the positive cells were counted. Kunming female mice aged 8 ~ 10 weeks were used for all animal experiments, and animal procedures were approved in advance by the Institutional Animal Care and Use Committee of Xiamen University.

After cultivation with HeLa cells for 2 months, metacestode vesicles (2 mm < diameter < 4 mm) were picked up and cultured in HM for 2 days. Metacestode vesicles were then subjected to various treatments mentioned in the text. For primary cell isolation, we referred to Spiliotis et al. (2008) with some modifications. In brief, metacestode vesicles were sheared by pipetting with a 5 mL syringe. After centrifugation (2,000 g, 5 min, room temperature) and three washing steps with PBS (15 mM NaH₂PO₄, 100 mM NaCl, 85 mM Na₂HPO₄, pH 7.4), 8 vol. of pre-warmed (37°C) trypsin (GIBCO, United States) was added to the tube. After incubation (37°C, 10 min) and adding the same volume of HM to terminate the dissociation. Then the cells were passed through a 30 μm sieve (Miltenyi Biotec, Germany) and centrifuged for 10 min at 1,000 g. The sediment was resuspended in HM. Primary cells were directly used for flow cytometry. Before the instrument analysis, primary cells were incubated with dyes 2,7-Dichlorodihydrofluorescein diacetate (DCFH-DA), DOJINDO, Japan; Hoechst 33342, Beyotime, China. The analytical and sorting instruments were Fortessa and FACSAria III (BD, United States). The data was analyzed using FlowJo X10.0 software.

Published sequences of HIF1α of the human, mouse, Drosophila, Xenopus, zebrafish and C. elegans (Supplementary Table S1) were used as queries to BLAST the E. multilocularis genome database (Tsai et al., 2013) available at Wormbase database.¹ EmuJ_000599400 was identified as the homologs HIF1α and its full coding sequences were amplified from the cDNA preparations as described previously (Brehm et al., 2000). RACE was performed using the SMART RACE cDNA Amplification Kit (Clontech, United States) according to the manufacturer’s instructions. Specific primers were used as shown in Supplementary Table S2. The domains of PAS and PAC were determined using the SMART.² The phylogenetic tree was generated by the maximum likelihood method (bootstrap = 1,000) using the MEGA 7.0.26. The analysis of three-dimensional structure was generated using the SWISS-MODEL.³ Primers for amplification of the full coding sequences of EmHIF1α were used as shown in Supplementary Table S2.

EmHIF1α, EmHIF1β and HsHIF1β, tagged at their N-terminus with FLAG-tag, MYC-tag or HA-tag, respectively, were sub-cloned into pcDNA3.3 plasmid (gifts from Prof. Lin Shengcai, Xiamen University, China). The expression plasmids were co-transfected into the HEK-293 T cells with the aid of Lipofectamine™ 3,000 and Opti-MEM™ I Reduced Serum Medium (Thermo Scientific, United States). Cell lysates were harvested at 36 h post-transfection using RIPA lysis buffer (Beyotime, China). Co-Immunoprecipitation experiments were performed using anti-FLAG (Sigma, United States, RRID: AB_262044), anti-MYC (Cell Signaling Technology, United States, RRID: AB_490778) or anti-HA antibodies (Cell Signaling Technology, United States, RRID: AB_1549585) conjugated Sepharose Beads (Cell Signaling Technology, United States). Lysates of experiments were electrophoresed on SDS-polyacrylamide gels and transferred onto the PVDF membranes in a humid environment. Membranes were blocked with 5% BSA in TBST and incubated with primary antibodies at 4°C overnight. Primary antibodies: β-Tubulin (Cell Signaling Technology, United States, RRID: AB_2210545), GAPDH (Proteintech, United States, RRID: AB_2107436), Phospho-ERK (Thr185, Tyr187) (Thermo Scientific, United States, RRID: AB_2533719). Then the membranes were washed three times by TBST and incubated with the horseradish peroxidase (Invitrogen, United States) which was conjugated with the anti-rabbit or anti-mouse IgG (Thermo Scientific, United States). Blots were developed using Bio-Rad ChemiDoc Touch (Bio-Rad, United States) and analyzed by Image Lab Software.

The EmHIF1α polyclonal antibody was prepared by immunizing New Zealand Rabbit with the synthetic peptide “CDVKQFQVDSIETSN” of EmHIF1α (Genscript, United States). Purification of the anti-EmHIF1α antibody from the antiserum was performed further by protein A and peptide affinity chromatography. Western blot was performed using the EmHIF1α polyclonal antibody with a dilution of 1:1,000.

The EmHIF1α target sequences were determined using the BLOCK-iT RNAi Designer software.⁴ The three selected siRNAs (7,013, 8,053 and 9,063) were synthesized by Biotechnology Co., Ltd. (Ribo, China). To determine the transfection efficiency, another negative control (siNC) and the FAM fluorescence label was purchased from Biotechnology Co., Ltd. (Ribo, China). The sequences of the three siRNAs targeting EmHIF1α were listed in Supplementary Table S2.

We established four siRNA groups: a negative control siRNA group (siNC) and three EmHIF1α siRNA-treated groups (7,013, 8,053 and 9,063). Electroporation was used to deliver siRNA into protoscoleces in vitro according to a previously established protocol (Spiliotis et al., 2010). In brief, 2000 protoscoleces were washed three times with RNAi electroporation buffer (120 mM trehalose, 20 mM HEPES, 1 mM myo-inositol, 1 mM KCl, 1 mM MgCl₂, 1 mM K₂HPO₄, 0.4 mM KH₂PO₄ and 1 mM gluthatione, pH 6.9) and then resuspended in 100 μL electroporation buffer containing FAM-labelled control siRNA to a final concentration of 3 μM in a 1-mm electroporation cuvette. Electroporation was performed pulses once with 125 V for 20 ms by Gene Pulser II (Bio-Rad, United States). After incubation at 37°C for 10 min, 2 mL culture medium was added, and the protoscoleces were transferred to 12-well plates for an additional 60 h of incubation at 37°C in 5% CO₂ in the dark. Then, some protoscoleces were used to collect protein samples. Remaining protoscoleces were subjected to EdU staining after incubating with 50 μM of EdU for 8 h and observed under a fluorescence microscope (SONY, Japan).

Total vesicle RNA was collected by RNeasy Mini Kit (QIAGEN, Germany). RNA was converted to cDNA by PrimeScript RT reagent Kit with gDNA Eraser (Takara, Japan). Real-time PCR was carried out using Hieff qPCR SYBR Green Master Mix (YEASEN, China) in LightCycler 96 (Roche, Germany). Elp was internal control (Brehm et al., 2003). QPCR was performed with using the oligonucleotides as listed in Supplementary Table S2.

Data are shown as mean ± SD as indicated in the respective figure legend unless otherwise indicated. The mean values of the data from the experimental groups were compared by performing a two-tailed Student’s t-test. SPSS 18.0 was used for statistical analysis and significance was set at p < 0.05. p-values were defined as follows: * p < 0.05, ** p < 0.01, *** p < 0.001, NS not significant. Asterisks without horizontal lines represent significant differences compared to the control group. Horizontal lines with asterisk on top indicate significant differences between groups.

Infection with parasites induces host immune responses to produce ROS. We first examined the in vivo ROS pattern of AE in the mouse liver using the ROS-sensitive dye DCFH-DA (2,7-Dichlorodihydrofluorescein diacetate) and found an obvious ROS accumulation around the lesion (Figure 1A). To investigate the effect of ROS on the growth of metacestode larvae, we then injected ROS scavenger NAC (N-Acetyl-L-Cysteine) into the mice and found that the accumulation of ROS was significantly reduced (Figure 1B). Interestingly, the parasite weight was decreased after NAC treatment (Figure 1C). To confirm this effect of ROS on parasite growth in vitro, we added NAC to the culture medium and found that the growth rate of the metacestode vesicles greatly decreased. By contrast, addition of H₂O₂ (hydrogen peroxide) increased the growth rate (Figure 1D). These results suggest that the growth of E. multilocularis larvae requires the participation of ROS.

The proliferation of germinative cells on the germinal layer is the basis of E. multilocularis larval growth (Koziol et al., 2014). So we examined the effect of ROS on the proliferation of germinative cells in vivo. AE-infected mice were administered to NAC treatment first, followed by intraperitoneal injection of EdU for labeling the proliferating cells. The results showed that the number of EdU⁺ cells in metacestode decreased greatly in the NAC treatment group (Figures 2A,B). Then we treated in vitro-cultured metacestode vesicles with two ROS scavengers NAC and Trolox (6-hydroxy-2,5,7,8-tetramethylchroman-2-carboxylic acid) respectively, both of which could efficiently down-regulate the ROS levels in the primary cells of metacestode vesicles (Figure 2C). We found that the number of EdU⁺ cells decreased after NAC/Trolox treatment in a dose-dependent manner (Figures 2D,E), accompanied with a significant down-regulation of the mRNA levels of cell cycle-related genes (Supplementary Figures S1A,B). Furthermore, we treated the metacestode vesicles with H₂O₂. The results showed that the number of EdU⁺ cells and the mRNA levels of cell cycle-related genes were increased greatly after 60 μM H₂O₂ treatment, whereas 500 μM H₂O₂ reduced the number of EdU⁺ cells (Figure 2F and Supplementary Figure S1C). These results suggest that the moderate ROS can promote the proliferation of germinative cells and too low or too high ROS levels are not conducive to proliferation.

HIF1α (hypoxia inducible factor 1α) is reported to act as a major effector of cellular redox levels (Lee et al., 2017). We excavated the genome information of E. multilocularis by BLAST analyses using human, D. melanogaster, and C. elegans HIF1α as the queries and procured as best hit protein encoded by locus EmuJ_000599400 (Supplementary Figures S2, S3). E. multilocularis HIF1α homolog (EmHIF1α) contains conserved HIF-characterized PAS and PAC domains (Supplementary Figure S2A), has a close evolutionary relationship with C. elegans HIF1α and exhibits a three-dimensional structure similar to human HIF1α (Supplementary Figures S2B,C). In mammals, HIF1α is stabilized, accumulated, and forms heterodimers with HIF1β to transcriptionally activate various downstream genes (Nandal et al., 2011). We also identified a HIF1β homologue of E. multilocularis by genome mining (EmuJ_000805200) (Supplementary Figures S2D–F). Co-IP from HEK-293 T cell line indicated that EmHIF1α immunoprecipitated with human HIF1β and EmHIF1β (Supplementary Figures S2G,H), suggesting a conserved complexing mechanism of HIF1 subunits in E. multilocularis. We then generated a polyclonal antibody against EmHIF1α, which effectively detected the recombinant His-tagged EmHIF1α protein as well as the endogenous EmHIF1α in the in vitro-cultivated metacestode vesicles (Supplementary Figures S2I,J). Further, we examined the expression of EmHIF1α in metacestode vesicles and found that EmHIF1α expression was increased under hypoxia (Supplementary Figure S2K).

To investigate the role of EmHIF1α in germinative cells proliferation, we treated metacestode vesicles with CoCl₂ (cobaltous chloride), an eminent hypoxia mimetic chemical and inducer of HIF1α (Munoz-Sanchez and Chanez-Cardenas, 2019). The results showed that CoCl₂ treatment resulted in a great induction of EmHIF1α (Figure 3A). Along with this, the growth rate of metacestode vesicles and the number of EdU⁺ cells significantly increased (Figures 3B,C) and the mRNA levels of cell cycle-related genes were up-regulated (Supplementary Figure S4A). We found that YC-1 [3-(5′-hydroxymethyl-2′-furyl)-1-benzyl indazole], an inhibitor of HIF1α (Kim et al., 2006), could reduce the expression of EmHIF1α (Figure 3D). After the treatment with YC-1, the growth rate of metacestode vesicles, the number of EdU⁺ cells and the mRNA levels of proliferation marker genes were decreased significantly (Figures 3E,F and Supplementary Figure S4B). These results suggest that EmHIF1α is involved in the proliferation of germinative cells.

We found that increasing the ROS levels in metacestode vesicles significantly enhanced EmHIF1α expression and this phenotype could be largely restored by NAC or Trolox (Figure 4A). To determine whether ROS promote the proliferation of germinative cells by regulating EmHIF1α, the metacestode vesicles were treated with YC-1 and/or H₂O₂. The results showed that YC-1 significantly compromised the H₂O₂-induced proliferation of germinative cells (Figure 4B). We further alternatively performed siRNA to suppress EmHIF1α expression in protoscoleces. The results showed that the number of EdU⁺ cells was significantly decreased by EmHIF1α knockdown in comparison with control siRNA (siNC). In addition, the increased number of EdU⁺ cells induced by H₂O₂ was clearly inhibited by EmHIF1α knockdown (Figures 4C,D). These results suggest that the ROS-EmHIF1α axis may be important in the growth of metacestode larvae and the proliferation of germinative cells.

ROS have been reported to induce HIF1α protein synthesis through the ERK signaling transduction pathway (Mottet et al., 2002). The previous studies have shown that the activation of EmERK signaling promotes the proliferation of germinative cells in E. multilocularis (Spiliotis et al., 2005, 2006; Gelmedin et al., 2010; Cheng et al., 2017). To know whether EmERK regulates the proliferation of germinative cells by involving ROS-EmHIF1α axis, we first isolated primary cells from the in vitro-cultivated metacestode vesicles (Supplementary Figures S5A,B) and analyzed the expression of phosphorylated EmERK (p-EmERK) and EmHIF1α in the sorted S/G2/M and G0/G1 cells. The results showed that p-EmERK and EmHIF1α were highly expressed in the S/G2/M cells (Figure 5A), suggesting that they may function in regulating germinative cell proliferation. Then we found that, similar to EmHIF1α, EmERK phosphorylation was also regulated by ROS (Figure 5B). We next treated the metacestode vesicles with the MEK/ERK inhibitor U0126 (Cheng et al., 2020) and found that U0126 significantly compromised the H₂O₂-induced proliferation of germinative cells (Figure 5C). In order to explore the effect of p-EmERK on EmHIF1α expression, we inhibited EmERK phosphorylation and EmHIF1α expression respectively, and found that the protein level of p-EmERK did not change significantly upon YC-1 treatment, while the treatment with U0126 effectively blocked the expression of EmHIF1α (Figure 5D). Furthermore, we activated EmHIF1α while inhibiting p-EmERK and found that the proliferation of germinative cells was restored (Figure 5E). Taken together, these results suggest that ROS stabilize the expression of EmHIF1α by phosphorylating EmERK, which ultimately affects the proliferation of germinative cells.