All procedures with mice were approved by the Boston University Institutional Animal Care and Use Committee and performed in compliance with the guidelines of the National Institutes of Health. C57BL/6J mice were obtained from The Jackson Laboratory (Bar Harbor, ME). C5aR1^-/- (B6.129S4-C5ar1tm1Cge) were originally provided by Prof. J. Köhl, Lübeck University. C5aR2^-/- (C5ar2tm1b(EUCOMM)Wtsi) from targeted embryonic stem cells and C5aR1/2^-/- (C57BL/6J-Del(7C5ar2-C5ar1)1Bosm) mice were generated in our laboratory and the Transgenic Facility of the University of Mainz, Germany (submitted for publication elsewhere). The mice were bred and genotyped at the animal facilities of Boston University under specific pathogen-free conditions and a controlled light/dark cycle. Male and female mice at 8-12 weeks of age were used for experiments.

N. brasiliensis infective L3 larvae (3-6 mm length) were visualized through an Amscope binocular (4 x magnification) and recorded using a camera with 10 x magnification ( Supplementary Movie 1 ). Larvae were maintained and separated from coprocultures, and ultimately prepared for inoculations according to an established protocol (7). Mice received subcutaneous (s.c.) injection in the flank with n=500 L3 in 0.1 mL of sterile PBS (Thermo Fisher Scientific, Waltham, MA). Sham mice treated with sterile PBS s.c. were included as controls.

Alveolar and lung viable larvae burdens were assessed in bronchoalveolar lavage fluids (BALFs) and lung tissues respectively, isolated from mice 48 h post-infection following a previously established protocol (30).

Alveolar macrophages were isolated from naïve C57BL/6J mice by performing BAL using 6x1 mL aliquots of cold sterile phosphate buffered saline (PBS) supplemented with 0.1% BSA (Sigma) and 2 mM ethylenediaminetetraacetic acid (EDTA) (Invitrogen, San Diego, USA) by inserting an 18G angiocatheter (Exel International, Redondo Beach, USA) into the trachea and kept on ice. The lavages were combined and centrifuged at 650 g for 10 min at 4°C. BALF cells were resuspended in sterile Roswell Park Memorial Institute medium (RPMI)-1640 (Gibco, Life Technologies, Grand Island, USA) medium containing 0.1% bovine serum albumin (BSA) and 100 U/mL Penicillin-Streptomycin (Thermo Fisher Scientific), and counted with a cell counter (CytoSMART, Skillman, USA). Equal numbers of cells were seeded in 48-well plates (Corning, Corning, USA) and cells were cultivated for 2 h in 5% CO₂ humidified incubator at 37°C. Thereafter, non-adherent cells were removed by washing the wells once with sterile PBS and the attached cells were further stimulated with 100 mM recombinant mouse C5a (rmC5a, R&D systems, Minneapolis, USA).

BALF total protein was measured with the Pierce BCA Protein Assay Kit (Thermo Fisher Scientific). Alveolar hemorrhage was assessed as described elsewhere (30). Briefly, a serial dilution (8000-250 mg/mL) of human hemoglobin (Sigma-Aldrich, St. Louis, MO) was prepared, and 50 μL of each standard and BALF samples were added to duplicate wells of a 96-well plate. A volume of 100 μL of 6% SDS (Sigma-Aldrich) was added to all wells and resuspended several times. The absorbance at 560 nm was measured with a Tecan Infinite M Nano plate reader (Tecan, Mannedorf, Switzerland), and alveolar hemorrhage (i.e., total amount of hemoglobin recovered in BALF) was determined from the standard curve generated by Magellan version 7.2 software (Tecan).

The ELISA kits for IL-27p28, IL-6, CXCL10, TNF-α were purchased from R&D System (Minneapolis, MN). BALF samples were diluted in PBS with 0.1% BSA to fit in standard range of the ELISA kit and measured according to the ELISA kit protocol. The absorbance at 450 nm was measured with a Tecan Infinite M Nano plate reader, and concentration was determined from the standard curve generated by Magellan software. The manufacturer’s protocol was followed for analyses. For detection of citrullinated histones, an ELISA kit for total histone detection was purchased from Roche (Mannheim, Germany). Anti-citrullinated Histone H3 antibody (Abcam) was used as coating antibody instead of anti-histone biotin. Optical density was read at 450 nm or 405 nm. This modified citH3 ELISA produced minimal background signal when sham mouse BALF were used as negative control (31, 32).

BALF samples were collected from infected mice as described above and kept on ice. Cell viability from the single cell suspension (>98%) was confirmed by staining with fixable viability dye eFluor 780. BALF cells were centrifuged at 650 g, 4°C for 5 min, and incubated at 4°C in the dark for 10 min with Fc receptor block (anti CD16/CD32 antibody) followed by fluorophore-labelled mouse antibodies, listed in the Supplementary Table . All antibodies were diluted in fluorescence-activated cell sorting (FACS) buffer containing sterile PBS supplemented with 1% (w/v) BSA, 0.01% (w/v) sodium azide, and 2.5 mM EDTA. The stained cells were washed twice in FACS buffer and fixed in 2% paraformaldehyde (Santa Cruz Biotechnology, Dallas, TX) at 4°C for 20 min in the dark.

For intranuclear staining of DEDD2 and intracytoplasmic staining of THBS1, cells were fixed and permeabilized after surface marker staining using the eBioscience FoxP3/TF staining buffer set (Thermo Fisher Scientific) for DEDD2, or the Cytofix/Cytoperm kit (BD Biosciences) for THBS1. Cells were further incubated with unconjugated anti-mouse DEDD2 polyclonal antibody (or isotype control) for 45 min at 4°C, or with unconjugated anti-mouse THBS1 monoclonal antibody (or isotype control) for 20 min at 4°C. After two washing steps, cells were stained for another 30 min at 4°C with the corresponding fluorophore-labelled secondary antibodies.

Stained/fixed cells were washed twice and centrifuged as described before, resuspended in FACS buffer, and transferred to a FACS tube containing 20 μL of CountBright Absolute Counting Beads (Thermo Fisher Scientific). For all single-stained compensation controls, we used a pool of BALF cells or UltraComp eBeads Compensation Beads (Invitrogen, Waltham, MA) according to the manufacturer’s instructions. Multiplets were identified as FSC-W^hi (i.e., FSC-A^hi) cells and excluded from further analysis (33). For Figure 1F , isotype controls for all antibodies were pooled and total live alveolar cells are shown. Flow cytometric acquisition of samples was performed on a BD LSR II flow cytometer with BD FACSDiva software or an Astrios MoFlo instrument (Beckman and Coulter). Preparation and analyses of final plots were performed in FlowJo version 10.

C5a protein detection was performed via tandem immunoprecipitation followed by western blotting of lung and BALF samples. This previously described protocol was intended to optimize sensitivity and specificity (34, 35). The faint signal in the BALF negative control without sample lane was interpreted as spillover from the recombinant mouse C5a (rmC5a) positive control lane ( Figure 1A ).

Total RNA was isolated from macrophages by Qiazol/chloroform/isopropanol/EtOH purification or the innuPREP RNA Mini kit (AnalytikJena, Jena, Germany), and quantified using a NanoDrop 2000c spectrophotometer (Thermo Fisher Scientific). The cDNA was generated using the Maxima™ H Minus cDNA Synthesis Master Mix kit (Thermo Fisher Scientific). 3-5 ng cDNA complemented with PowerUp™ SYBR™ Green Master Mix (Applied Biosystems, Waltham, USA) and specific primers at a concentration of 500 nM each were used for RT-PCR in a QuantStudio 3 Real-Time PCR instrument (Applied Biosystems). Dedd2 gene expression was compared between samples by normalizing to Gapdh expression and applying the 2^-ΔΔCt formula. Primer sequences are provided in the Supplementary Table .

RNA was extracted using the QIAGEN RNAeasy Plus Micro kit. Ultra-low input RNA-seq was performed by Novogene (USA). Analysis methods are further detailed in the Supplementary Material . C5ar1 was not included in the volcano plot for better axis scaling, but absence of expression was confirmed in all C5aR1^-/- mice. Pathways with no apparent biological relevance to pulmonary hookworm infection (e.g. pancreas beta cells, estrogen early response, allograft rejection, adipogenesis, xenobiotic metabolism and heme metabolism) are not shown.

Statistical analyses were performed with Prism v8 software (GraphPad). Sample sizes and number of technical and biological replicates are included in the figure legends. Data in the bar graphs represent mean ± standard error of the mean (SEM). Comparisons of two groups were performed based on the two-sided Student’s t-test while multiple group testing was done through one-way analysis of variance (ANOVA) with Bonferroni correction. The statistical significance threshold was set at P<0.05.

To determine if pulmonary hookworm infection triggers local C5a generation as an indicator of early complement activation, we used N. brasiliensis, a well-characterized murine nematode closely related to human-pathogenic hookworms. This infection model is well established for studying the immunological response during the pre-intestinal stage of primary infection (3, 7).

C57BL/6J wild type (WT) mice were inoculated sub-cutaneously (s.c.) with viable stage 3 N. brasiliensis larvae (L3) from coprocultures ( Supplementary Movie 1 ). Bronchoalveolar lavage (BALF) and lung tissue were collected 48 h post-infection and processed to detect C5a by immunoprecipitation followed by western blot. Infected mice displayed a considerable presence of C5a in BALF and dissociated lungs, whereas C5a was not detected in uninfected sham mice ( Figures 1A, B ). Next, we proceeded with immunophenotyping of BALF immune cells isolated from WT and C5aR1^-/- mice using flow cytometry (gating strategy provided in Supplementary Figure 1A ). N. brasiliensis caused a marked accumulation of PMNs (neutrophils), monocytes, and eosinophils into the airspaces in the lungs in both strains ( Figures 1C, D ), which are not present in the BALF of uninfected sham mice ( Supplementary Figure 1B ) (35, 36). Importantly, the influx of PMNs, monocytes, and eosinophils was significantly reduced in C5aR1^-/- mice compared to WT mice, suggesting that the recruitment of inflammatory cells to the alveoli during the early phase of hookworm infection is dependent on the C5a/C5aR1 axis.

In contrast, the numbers of C5aR1⁺ resident AMs (alveolar macrophages) and invading NK cells remained unchanged in the absence of C5aR1 signaling, since these cells do not require C5a/C5aR1 to appear in the small airways of the lung ( Figures 1D–F ). All studied subsets of BALF leukocytes except NK cells expressed C5aR1 ( Figures 1E, F ). PMNs, not eosinophils, were the most abundant cell type, and PMNs expressed highest amounts of C5aR1 ( Figures 1D–F ). These data suggest that C5a mediates innate immune responses, particularly PMN influx, during acute pulmonary hookworm infection.

PMNs are a first line of defense and have the potential to kill nematode larvae (37, 38). Consequently, BALF and lung lysates from WT and C5aR1^-/- mice were used to evaluate the impact of C5aR1 deficiency on alveolar and lung parasite burden. Surprisingly, WT mice compared to C5aR1^-/- mice showed a significant ~4-fold and ~2-fold higher viable parasite load in BALF and lung homogenates, respectively ( Figures 2A, B ). To assess a potential role of the C5aR2 receptor, we compared WT mice with C5aR1^-/-, C5aR2^-/-, and C5aR1/2 dual knockout mice. No relevant changes in the alveolar live parasite counts in infected C5aR2^-/- and C5aR1/2^-/- mice were present ( Supplementary Figure 2A ). These results indicated a more efficient pathogen clearance in alveolar spaces and lung tissue when C5a signaling through C5aR1 was absent, despite a reduced PMN count. Since hookworms cause vascular damage in lung tissue, we assessed the alveolar hemorrhage by measuring hemoglobin concentration in the BALF of infected WT and C5aR1^-/- mice. In line with greater parasite burden, we observed greater alveolar hemorrhage in WT mice compared to C5aR1^-/- mice ( Figure 2C ). On the other hand, the disruption of C5aR2 signaling alone or in combination with C5aR1 did not significantly alter pulmonary hemorrhage ( Supplementary Figure 2B ). Consistently, lung injury was more severe in WT mice than in the C5aR1^-/- mice, as shown by the exacerbated protein leakage detected in BALF ( Figure 2D ), a parameter used as a marker of capillary/alveoli barrier disruption (39). Furthermore, PMN reduction in C5aR1^-/- mice was associated with a lower concentration of soluble citrullinated histone 3 (CitH3), suggesting a dampened release of neutrophil extracellular traps (NET)s, consistent with the reduction of tissue damage observed in C5aR1^-/- mice ( Figure 2E ). CitH3 is regarded as a reliable marker of NETosis (31, 40). For this reason, we opted not to analyze additional parameters of neutrophil activation. To evaluate the intensity of parasite-induced inflammation, we determined a range of inflammatory markers in BALF of WT and C5aR1^-/- mice early after N. brasiliensis infection. Significantly lower levels of TNF-α, IL-6, IL-27p28 and CXCL10 were found in the BALF of C5aR1^-/- mice, showing that abrogation of C5aR1 signaling attenuated inflammatory cytokines and chemokines ( Figure 2F ). Of note, we determined that the surge of BALF larvae burden and lung injury/inflammation occurs 2 days after infection compared to uninfected sham controls ( Supplementary Figures 3A-D ).

Since PMNs were the most abundant cell type with the highest C5aR1 expression, and C5aR1^-/- mice showed significantly lower neutrophil recruitment to the alveolar compartment than WT mice, we investigated C5aR1-dependent alterations in PMNs using bulk RNA sequencing. RNA was purified from fluorescence activated cell (FAC)Sorted CD45⁺Ly6C⁺Ly6G⁺ PMNs isolated from the BALF of C5aR1^-/- and WT mice infected with N. brasiliensis for 48 h ( Figure 3A ). The principal component analysis plot showed 2 distinct clusters of samples respective of each experimental group ( Figure 3B ). N. brasiliensis infection induced significant changes in the abundance of 63 genes in C5aR1^-/- versus WT mice, with a large majority of differentially expressed genes (DEGs) being downregulated (n=55) and only a few DEGs being upregulated (n=8) ( Figure 3C ). Among DEGs with the highest expression values, we identified Dedd2, Thbs1 (Thrombospondin1), Ccrl2, and Cd274 (encoding for PD-L1), that were downregulated in C5aR1^-/- PMNs ( Figure 3D ). Enrichment analysis of the DEGs showed that apoptosis pathway/P53 signaling, inflammatory response represented by PI3K/AKT/MTOR signaling, IL2/STAT5 signaling, IL6/JAK/STAT3 signaling, MTOR signaling, peroxisome, TNF-α signaling and hypoxia, plus interferon α/β pathways, were significantly inhibited in PMNs during N. brasiliensis infection ( Figure 3E ). The number of genes involved in the regulation of the innate immune response, defense, and metabolic processes were homogenously distributed ( Supplementary Figure 4 ). In conclusion, these results suggest that C5aR1^-/- PMNs harbor a less inflammatory phenotype than WT PMNs.

We manually curated a short list of significant DEGs of interest based on perceived biological relevance, novelty, and abundance. The expression values of Dedd2, Thbs1, Ccrl2 and Cd274 were significantly reduced in C5aR1^-/- alveolar PMNs after N. brasiliensis infection ( Figure 4A ). Since Dedd2, Thbs1, and Ccrl2 genes (but not Cd274) showed high expression values, we proceeded with protein validation by flow cytometry. No significant difference was observed in THBS1 or CCRL2 protein expression in Ly6G⁺Ly6C⁺ PMNs and CD11c⁺SiglecF⁺ AMs between C5aR1^-/- mice and WT controls ( Figure 4B ). However, protein staining for DEDD2 was reduced by 2-fold in PMNs and by 1.5-fold in AMs of C5aR1^-/- mice compared with WT mice following N. brasiliensis infection ( Figures 4C, D ). These findings confirmed the downregulation of DEDD2 at the protein level in neutrophils lacking C5aR1 during hookworm infection, while suggesting that potential post-transcriptional regulation mechanisms may occur for CCRL2 and THBS1. DEDD2 is a recently identified protein with potential functions as a cell death regulator and possibly apoptotic processes (28) ( Figure 2E ). To further evaluate the effect of C5a signaling through C5aR1 on Dedd2 expression, we exposed primary AMs isolated from naïve WT mice to mouse recombinant C5a and assessed the gene expression of Dedd2. Consistent with the in vivo RNA sequencing data ( Figure 4A ), C5a stimulation resulted in upregulation of Dedd2 in AMs ( Figure 4E ), demonstrating the C5a-dependent regulation of DEDD2 through C5aR1.