All human studies complied with all relevant ethical regulations. Experiments with blood samples from healthy caucasian donors of both sexes (25–49 years of age) were approved by the ethics committee of the medical association Hamburg (permission number: 2020-10067-BO). All experiments were conducted under donor anonymization and in accordance with the relevant guidelines.

C57BL/6 mice (10–12 weeks old) were used for the experiments, which were performed in accordance with German animal protection laws and reviewed by the federal health authorities of the State of Hamburg (permission numbers: N51/17; N120/2020). Mice were bred in the animal facility of the Bernhard Nocht Institute for Tropical Medicine and kept at 21–22°C (50–60% humidity) in ventilated cages under specific pathogen-free conditions. Mice were rendered unconscious using a CO₂ (20%)-filled chamber and sacrificed by cervical dislocation.

Mice (10-12 weeks old) were infected intrahepatically with 2 × 10⁵ in vitro-cultivated trophozoites of the high pathogenic E. histolytica clone B2, as described previously (35, 36). Mice were sacrificed on day 3 postinfection (pi), at the peak of disease severity (25).

Male mice (8 weeks old) were gonadectomized by testicular ligation. Testosterone substitution was performed by subcutaneous implantation of an osmotic pump (micro-osmotic pump; Model 2004, ALZET) containing 5 mg/mL testosterone diluted in 45% w/w 2-Hydroxypropyl-β-cyclodextrin (carrier solution) or carrier solution alone. Alternatively, testosterone implants (Belma Technologies: T30) were inserted subcutaneously.

Bone marrow (BM) was collected from euthanized mice by flushing it into a dish using PBS. The cell suspension was filtered and treated with an erythrocyte lysis buffer. Blood was obtained by cardiac puncture upon euthanization, collected in EDTA-coated tubes, and two erythrolysis steps were performed to isolate immune cells. Spleen and liver were mashed, flushed with PBS on ice, and filtered, followed by one erythrolysis step. Liver immune cells were isolated using 80% Percoll as described previously (37). Neutrophil granulocytes from bone marrow were isolated using a Neutrophil Isolation Kit (#130-097-658, Miltenyi Biotec), blood neutrophils were isolated by FACS-sorting.

Liver tissue from E. histolytica-infected mice was fixed in formalin (4%) and embedded in paraffin. Sections (0.2 µm) were stained with hematoxylin and eosin (H&E) or prepared for immunohistochemistry. Antibodies: rabbit anti-mouse 7/4 antibody (Neutrophils, clone 7/4; Cedarlane; 1:800 dilution); polyclonal rabbit serum raised against recombinant E. histolytica- antigens. Slices were developed using DCS SuperVision Single Species horse-radish peroxidase (HRP)-Polymere (Innovative Diagnostic-Systems) and counterstained with hemalaun.

Blood plasma was used for cytokine analysis. Briefly, collected blood was centrifuged (1000 × g, 4°C, 10 min) to obtain plasma and stored -20°C prior to cytokine measurement. Cytokine analysis was performed using multiple customized murine LEGENDplex kits (BioLegend).

Flow cytometry was performed from organ single-cell suspensions or purified neutrophils, live cells were identified using zombie UV dye (#423108 BioLegend) or live/dead blue (#L34961 Invitrogen), intracellular cytokine staining was performed following 4 h restimulation with phorbol myristate acetate (PMA; 50 ng/mL) and ionomycin (500 ng/mL), followed by a 1h Brefeldin A (5 µg/ml) incubation. Antibodies: (all from BioLegend unless stated otherwise): CD11b FITC (1:200, M1/70), CD11b PerCP (1:100 M1/70), CD11b Bv510 (1:100, M1/70), Ly6C PerCP/Cy5.5 (1:400, HK 1-4), Ly6C PE (1:800, HK 1-4), Ly6G APC/Cy7 (1:400, 1A8), Ly6G PE (1:400, 1A8), Ly6G Bv785 (1:100, 1 A8), PD-L1 PE (1:100, 10F.9G2), CD54 AF647 (1:200, YN1/1.74), CD117 BUV395 (1:100, 2B8, BD), TNF-α BV421 (1:200, MP6-XT22), CCL2 PE (1:50, 2H5), CXCL1 AF647 (1:50, 1174A, R&D). Human neutrophils were analyzed in peripheral blood or after isolation using the MACSxpress^® kit (Miltenyi Biotech). Isolated cells were stimulated with PMA (10 ng/mL), LPS (0.1 µl/mL) and CL097 (1µg/mL) for 4 hours. Antibodies: CD16-APC-Cy7 (1:200, 3G8) and CD66B-APC (1:400, G10F5, Biolegend), viperin (RSAD2) PE (1:200, MaP,VIP; Biosience). Flow cytometry analysis was performed on a Cytek Aurora (Cytek)- or BD LSRII cytometer and data analysis were performed using FlowJo V10.4.2 software.

RNA was isolated from BM-derived neutrophils or blood leukocytes using TRIzol reagent (Life Technology) or a RNeasy MinElute kit (Qiagen). RNA was transcribed into cDNA using the Maxima First Strand cDNA Kit (Thermo Scientific). Androgen receptor (AR) and Ly6G mRNA levels were calculated using the 2^-ΔΔCt method, with the ribosomal protein S9 (RPS9) as housekeeping gene. The following primers were used: fwd-AR, 5´-TGAGTACCGCATGCA-CAAGT-3´; rev-AR, 5´-GCCCATCCACTGGAATAATGC-3´. Ly6G SG QuantiTect primers (Qiagen) were used for quantification of Ly6G. QPCR was performed using the Maxima SYBR Green qPCR Master Mix kit (Thermo Scientific) and a Roche LightCycler^®.

Sequencing library was prepared using the QIAseq Stranded mRNA Library Kit (Qiagen, Hilden, Germany) at the NGS Core Facility at the Bernhard-Nocht-Institute for Tropical Medicine and sequenced on a NextSeq 2000 Illumina Platform. Constructed libraries were sequenced as barcoded pooled samples on a NextSeq 550, resulting in 14–17.5 million reads for blood and 15.1–22.4 million reads for BM. The read length of the nucleotides was 100 base pairs. Library preparation and sequencing were performed at the Bernhard Nocht Institute for Tropical Medicine.

As previously shown, depletion of inflammatory Ly6C^hi monocytes and neutrophils reduces liver damage in the murine model for hepatic amebiasis (23, 25). The same experimental design was used to study neutrophil granulocytes: Male and female mice were infected intrahepatically with E. histolytica trophozoites and examined on day 0 and day 3 post infection. ( Figure 1A ). Immunohistological examination of infected livers demonstrated greater tissue damage, massive immune cell infiltration, stronger abscess formation and a distinct margin area of Ly6G⁺ neutrophils in males than in females ( Figure 1B ). Flow cytometry analysis identified neutrophils as Ly6G⁺ cells out of CD11b⁺ cells ( Figure 1C ) and revealed higher percentages of neutrophils in the blood of naive male animals, but not in the BM and liver; after infection, a significant increase in the amounts of neutrophils in the blood of female animals and in the liver of both sexes was observed ( Figures 1D, E ).

The more severe pathology in males was reflected by higher pro- and anti-inflammatory cytokine serum levels in male than in female mice ( Figure 1F ). Infection triggered an increase in neutrophil accumulation in the liver; these cells produced TNF-α (significantly higher in females after infection), CCL2, and CXCL1, and the percentage of neutrophils producing CXCL1 was significant higher in males and females compared to uninfected mice ( Figure 1G ). In addition, phagocytic capacity ( Supplementary Figures 1A, B ), production of reactive oxygen species (ROS) ( Supplementary Figures 1C, D ) as well as myeloperoxidase (MPO) production ( Supplementary Figures 1E, F ) as assayed on BM-derived neutrophils, was not significant higher in female neutrophils than in male counterparts ( Supplementary Figure 1A ). In summary, amoebic infection of the liver results in a higher prevalence of neutrophilic granulocytes in male mice.

To study the impact of testosterone on neutrophil dynamics during E. histolytica infection, male mice underwent gonadectomy and then received either testosterone or a carrier solution for two more weeks before being infected intrahepatic with E. histolytica ( Figure 2A ). Gonadectomy of male mice significantly reduced the number of neutrophils, regardless of infection ( Figure 2B ). In contrast, testosterone substitution in gonadectomized naïve male mice resulted in a marked increase in neutrophil abundance in the blood and tended to increase in the liver as well. ( Figure 2C ). A similar effect on neutrophils was observed in female individuals substituted with testosterone ( Figure 2D ). qPCR analysis of blood leukocytes revealed increased AR and reduced Ly6G mRNA expression in males following gonadectomy, whereas testosterone treatment reduced expression of AR mRNA and increased expression of Ly6G mRNA in females ( Figures 2E, F ). Consequently, there was a correlation between expression of both, AR and Ly6G (R^2 = 0,1077; p < 0.05) ( Figure 2G ). Further, sex-specific correlation analyses are provided in Supplementary Figures 2C, D . Next, we analyzed the expression of CD54⁺ (inflammatory N1) and PD-L1⁺ (anti-inflammatory N2) on neutrophils (CD11b⁺Ly6C⁺Ly6G⁺) (scheme and representative FACS Plots are shown in Supplementary Figures 2A, B ). The amount of N1 neutrophils increased after infection in the BM, blood and liver in female individuals, as well as in the BM and the liver in males ( Figure 2H ). N2 neutrophils showed a similar pattern as N1, with the greatest quantities in the liver in both sexes ( Figure 2I ). No significant differences were found in N1 neutrophils following gonadectomy and testosterone substitution. Only a slight decrease in the BM following infection as well as in the blood and the liver following testosterone substitution and infection was observed ( Figure 2J ). Although not significant, N2 neutrophils are increased in the blood in naive gonadectomized and infected male mice upon testosterone treatment, and increased in gonadectomized and testosterone substituted mice in the liver upon infection ( Figure 2K ).

Ex vivo exposure of isolated BM-derived neutrophils increased expression of CD54 (N1) and PD-L1 (N2) following LPS stimulation in both sexes and increased production of TNF-α in males following stimulation with amoebic antigens ( Supplementary Figures 3A–C ).

Based on the data published by Kim et al. (38) and the nomenclature of Xie et al. (17), deconvolution analysis was performed to investigate further maturation stages. Pre-neutrophils (G2) were present in the BM but not in the blood. Immature neutrophils (G3) increased after infection. Mature neutrophils (G4) and ISG-expressing neutrophils (G5b (17, 38);) significantly increased in infected females’ BM and blood. Late-mature neutrophils (G5a, G5c; migration & inflammatory responses (17, 38);) decreased in the BM and in the blood in females during infection but increased in males after infection ( Figures 2L, M ).

In summary, not all neutrophil granulocytes meet the N1- and N2-like definitions and testosterone further suppresses both of them in the blood during infection. Females tend to have a higher proportion of N1 and N2 neutrophils in the blood and liver, and deconvolution analysis revealed that a higher number of mature neutrophils (G4) is found in females and G5a neutrophils are higher in males during E. histolytica infection in the blood.

To assess sex-specific differences in neutrophils at the transcriptomic level, we conducted a differential expression analysis of neutrophils isolated from BM and blood samples from both male and female mice. Due to the extremely low number of neutrophils in the female liver ( Figure 1E ), we did not sufficient amounts of RNA required for high throughput sequencing. Our analysis revealed major differences in gene expression profiles during infection, with a focus on differentially expressed genes (DEGs) exhibiting a log-fold change (FC) > 1.5 and a p-adjusted < 0.01. The results are depicted in Figures 3A–D . In BM-derived neutrophils, we identified 413 DEGs in males and 472 DEGs in females during infection, when compared to uninfected controls. In blood-derived neutrophils, 40 DEGs were observed in males, and 254 DEGs were identified in females during infection ( Figures 3A-D ). A detailed list of all DEGs is provided in Supplementary Table 4 . Infection-dependent differences in the expression of genes were more prominent in blood compared to BM-derived neutrophils from male and female mice. The differential changes of male ALA versus naïve and female ALA versus naïve show similar expression changes ( Figure 3A ). We visualized the normalized expression values of the DEGs in two heatmaps ( Figures 3A, B , right side) for bone marrow to compare expression levels across conditions and sexes. Although the differences were not strongly pronounced, they were discernible in individual samples and between the sexes. For instance, a subcluster was observed in the female ALA versus naive comparison that was distinct from the male cluster, indicated in Figure 3B , that is not visible in the male comparison of BM ( Figure 3A ). In blood neutrophils, infection led to a more pronounced upregulation of genes, with females showing higher activation in the infected phase ( Figure 3D ) compared to expression changes in males ( Figure 3C ). Female blood neutrophils displayed upregulated expression of Arg1 (FC = 4.11; padj < 0.01), a gene associated with N2-like neutrophils ( Figure 3B ) (30, 31). The heatmaps depict a rather homogenous distribution of gene expression levels throughout the samples. Further, Volcano plots stratified by sex illustrated lower expression in male ALA BM-derived neutrophils than in uninfected BM-derived neutrophils. Notably, sex differences were more pronounced in blood compared to BM-derived neutrophils from both male and female mice ( Supplementary Figures 4 A, B ). In Venn diagrams, we observed that a lower number of DEGs were down-regulated in male BM-derived neutrophils (249/413) compared to females (322/472) during infection. Conversely, more DEGs were upregulated in male neutrophils (164/413). In blood neutrophils, we found fewer downregulated DEGs in males (17/40) than in females (131/254), while more upregulated DEGs were observed in females (123/254) than in males (23/40) ( Figure 3E ).

We also investigated the infection-related enrichment of general signal transduction pathways and depicted the top 20 pathways of GO biological processes (GOBP). A detailed list of all GOBP is provided in Supplementary Table 1 . These were separated into females ( Figure 3F ) and males ( Figure 3G ). In BM-derived neutrophils, infection was associated with the upregulation of pathways related to biosynthetic processes and increased cell proliferation. In contrast, blood neutrophils exhibited strong activation of pathways related to type I and type II interferon signaling. In summary, during infection, more genes, especially ISGs, were upregulated in blood neutrophils compared with BM-derived neutrophils, with greater activation of gene expression in female neutrophils ( Figures 3F, G ).

Since interferon signaling is important for pathogen elimination, we next focused on sex- and infection-dependent expression of type I and type II ISGs in BM and Blood neutrophils (16, 31) (included ISGs are shown in Supplementary Table 5 ). Overall, female BM-derived neutrophils showed higher expression of type I ISGs under steady state conditions (BM and blood, Oas2, Ifih1, Rsad2), whereas a subset of ISGs was also highly regulated in their male counterparts (BM, Ddit4, Slc25a28, Trim30b, Irf1, Jade 2; Blood, Ssbp3) ( Figure 4A ). During infection, more ISGs were upregulated in female (comparison ALA vs naive) blood neutrophils than in male samples (comparison ALA vs. naive), whereas a downregulation of ISGs was observed in male neutrophils and no changes were found in female BM neutrophils ( Figure 4A ). Type II ISGs in BM neutrophils showed less pronounced sex differences under steady state conditions, however, expression of chemokines involved in adhesion and recruitment of neutrophils was higher in male-derived neutrophils. In blood neutrophils, in the steady state (e.g. Myc, H2-DMb2, H2-M3, H2-Q7…) but also during infection e.g. Cxcl10, H2-D1, C4a, Psme1, Stat1, Psme2, more genes were upregulated in neutrophils from female mice. However, one set of type II ISGs was also upregulated in male-derived neutrophils following infection e.g. Casp1, CtsI, Ccl2 ( Figure 4B ). An additional illustration, showing organ-and sex-specific comparative analysis of ISGs, is provided in Supplementary Figures 4C, D .

Differential exon usage (DEU) is a type of alternative splicing that involves differential inclusion or exclusion of one or more exons from the mature mRNA transcript of a gene. The differential exon usage is calculated comparing infected versus naïve samples for males and females separately ( Figures 4C, E ). The figures show the fitted expression values (model effect estimates) of the comparisons for each of the exons of the genes for male (blue) and female (red).

We obtained 90 (female) and 1 (male) DEU for blood female and male, respectively, while female and male BM had 42 and 3 DEU, respectively ( Supplementary Table 2 ). The three ISGs (Oas2, Ifih1, and Samhd1) showed a higher fitted expression of exons in females than in males BM neutrophils ( Figure 4C ). Interestingly, the number of DEUs was higher in BM- and blood neutrophils from females than males, suggesting a more versatile rearrangement and reaction to infection in females.

Furthermore, we aimed to gain insight into the transcription factor changes that might explain differential gene regulation in sex-specific transcriptional profiles in blood and BM neutrophils. We employed transcription factor interference to analyze the activity profiles of ISG-related transcription factors (TFs) and identified differentially activated TFs ( Supplementary Table 3 ). We found that Irf2, Irf1, and Stat2 exhibited female-specific upregulation in infected BM, while their expression was upregulated in blood of both, male and female, under disease conditions. Creb3 was downregulated in uninfected female blood samples and downregulated for uninfected male samples of BM ( Figure 4D ). Additionally, a higher activity of Stat1 and Stat4 in female BM-derived neutrophils at steady state could be detected. This was verified by a higher Stat1/Stat4 exon expression in female BM neutrophils under steady state conditions (only the part that changes significantly are shown) ( Figure 4E ).

In conclusion, type I ISGs are more highly expressed in female neutrophils in the blood, independent of infection. The expression of type II ISGs of neutrophils in blood is higher in female than in those of male individuals both in the naive state and after infection. Furthermore, the high number of DEUs in BM- and blood neutrophils from females suggest a more versatile rearrangement and reaction to infection in females.

We have observed increased expression of type I ISGs at the transcriptional level, we were now interested in studying the expression and testosterone dependence of a prototypical ISG. For this purpose, we chose one of the best-studied ISGs, the viperin/RSAD2 ( Figure 4A ; Figure 5A ) (32–34) and analyzed it by flow cytometry ( Figure 5B ). Upon infection, RSAD2-expressing neutrophils significantly increased in both sexes in the blood (***p < 0.001) and the liver (*p< 0.05 (males); ***p < 0.001 (females)) ( Figure 5C ). Following testosterone substitution of female mice, a noticeable trend towards decreased RSAD2-expressing neutrophils from blood, but not the liver, was observed after infection ( Figure 5D ). Unlike in mice, we found no sex difference in the percentage of peripheral neutrophils (FACS Plot of human neutrophils expressing RSAD2, Figure 5E ) between men and women ( Figure 5F ), however, RSAD2 expression by neutrophils from women was significantly higher after stimulation with PMA compared to unstimulated neutrophils (*p<0.05, Figure 5G ). After stimulation with LPS or CL097, a tendency for higher RSAD2 production was observed in female neutrophils compared with male neutrophils ( Figure 5G ).

In conclusion, male neutrophil granulocytes exhibit lower type I ISG RSAD2 expression compared to females, which is likely influenced by testosterone.