The gene encoding FIP-nha (GeneBank: XM3043608) and mutations (for which the asparagine on either position 5 (N5A), position 39 (N39A) or both (N5+39A) was replaced by an alanine) were synthesized in the laboratories of BaseClear B.V. (Leiden, the Netherlands). For protein expression, the plasmid pPICZα A (Invitrogen, California, United States) was transformed into Pichia pastoris strain X-33 (Biolab, Beijing, China). Mutagenesis construction of FIP-nha, the induction of protein expression, and protein purification are described in a previous publication (Liu et al., 2023).

FITC (ThermoFisher, California, United States) was dissolved in DMSO with the same concentration as rFIP-nha. Protein in PBS mixed with FITC in DMSO with 10:1 ratio, and keep in 4°C overnight. FITC-labelled rFIP-nha (FITC-rFIP-nha) were separated via NAP-5 column with DPBS buffer according to its manufacturer’s protocol (GE Healthcare, Chicago, United States).

Protein concentrations (in 10 mM PBS solution; pH 7.4) were quantified by BCA Protein Assay Kit (ThermoFisher, California, United States), and all samples were snap frozen by liquid nitrogen and stored at −80°C. LPS content of rFIP-nha was tested via Endonext Kit according to the manufacturer’s protocol (Biomerieux, Marcy-l'Étoile, France). Protein information regarding concentration and LPS content is listed in Supplementary Table S1; The purified proteins visualised on SDS-PAGE gel is shown in Supplementary Figure S1.

THP-1 monocytes (ATCC (#TIB-202), Massachusetts, United States) were cultured as described previously (Hoppenbrouwers et al., 2022). For differentiation into THP-1 macrophages, cells were plated at 1 × 10⁶ cells/mL in 12-well tissue culture plates (Greiner Bio-one; 665,180) in RPMI-1640 containing 25 mM HEPES and 2 mM Glutamax (Lonza, Basel, Switzerland), supplemented with 10% FBS and 1% penicillin and streptomycin (Pen/Strep; Sigma-Aldrich, St. Louis, MO, United States) with 100 ng/mL PMA (Sigma-Aldrich, Zwijndrecht, Netherlands). After 48 h, cells were washed twice with complete RPMI medium and left for 5 days without touching the plates.

Macrophages were treated with 1 μM FITC-rFIP-nha for 5 min, 10 min, 30 min, and 60 min with RPMI without phenol red. After treatment, macrophages were washed with DPBS buffer (Sigma, Zwijndrecht, Netherlands) twice and stained with 10 mM Hoechst in DPBS buffer for 10 min. Subsequently, macrophages were washed with DPBS buffer twice, and any residual FITC-dye was quenched using 0.004% trypan blue, to eliminate any signal from FITC-rFIP-nha possibly still bound to the outside of the cell. Fluorescence images were taken using the Evos FL Auto 2 (Invitrogen, California, United States) at GFP, DAPI and Trans channels with 60 times magnification separately.

Macrophages were treated with 10 μM rFIP-nha for 16 h, with the LPS amount of rFIP-nha (Supplementary Table S1; 3.15 EU/mL) as LPS-negative control (LPS-CN), and 1 μg/mL LPS as LPS-positive control (LPS + CN). Medium-exposed cells were included as non-treated control (CN). The supernatant was collected for Enzyme-Linked Immunosorbent Assay (ELISA), while the cells were further tested via 7AAD cell viability assay and phagocytosis assay. Cytotoxicity of rFIP-nha exposure to cells was analysed using the 7AAD staining (BD Biosciences, Drachten, The Nethrlands) according to the manufacturer’s protocol.

After supernatant removal, 1 mL medium +4 μg/mL AlexaFluor-conjugated Escherichia coli (K-12 strain) BioParticles® (Molecular Probes, Life Technologies, Eugene, OR, United States) was added to each well. Incubation occurred for 1 h in a 37 °C incubator with 5% CO₂, followed by two washes using 1 mL of DPBS (Sigma, Zwijndrecht, Netherlands). Macrophages were harvested using 0.3 mL 0.25% Trypsin/EDTA per well, washed, resuspended with DPBS, and quantified using a CytoFlex flow cytometer (Beckman Coulter, Brea, United States) in the FITC-A channel, with data collected for up to 5000 events.

To investigate whether the effects of rFIP-nha were initiated by the protein itself, the rFIP-nha WT was dissolved in medium and heat-inactivated at 100°C for 10 min, and subsequently used alongside the non-heated rFIP-nha WT to analyse mRNA expression of inflammasome genes using qPCR. 2 μg/mL poly (dA:dT) (InvivoGene, California, United States) was transferred to THP-1 macrophages via Lipofectamine™ 2000 (Thermo Fisher Scientific, California, United States) following the manufacturer’s protocol as AIM2 positive control (AIM2+). 1 μg/mL Ultrapure flagellin from S. typhimurium (InvivoGen, California, United States) was transferred to THP-1 macrophage via Lipofectamine™ 2000 following the manufacturer’s protocol as NLRC4 positive control (NLRC4+). 1 μg/mL LPS with 5 mM ATP was considered as NLRP3 positive control (NLRP3+). Medium-treated cells were considered as non-treated control (CN). THP-1 macrophages were treated with samples for 6 h in a 37°C cell culture stove with 5% CO₂, then media were removed, and cells were collected for RNA extraction.

ASC (apoptosis-associated speck-like protein containing a CARD domain) is an essential protein adaptor implicated in canonical inflammasome responses involving NLRP3 and AIM2 (Mathur et al., 2018). To verify the involvement of ASC, ASC knockout human monocytes (THP1-KO-ASC cells) were obtained from InvivoGen (InvivoGen, California, United States), and cultured according to the manufacturer’s protocol. For differentiation into THP1-KO-ASC macrophages, cells were plated at 1 × 10⁶ cells/mL with 1mL/well in 12-well tissue culture plates (Greiner Bio-one; 665,180) in RPMI-1640 with HEPES and Glutamax (Lonza, Basel, Switzerland) supplemented with 10% FBS, 100 ng/mL PMA (Sigma-Aldrich, Zwijndrecht, Netherlands). 1% Normocin (InvivoGen, California, United States), a formulation of three antibiotics to prevent contamination from mycoplasmas, bacteria, and fungi was used instead of 1% Pen/Strep, as advised by the HP1-KO-ASC culture instruction manual. After 48 h, cells were washed twice with complete RPMI medium and rested for 5 days without touching the plates.

The antagonists ODN TTAGGG (A151; InvivoGen, California, United States) and NLRP3/AIM2-IN-3 (J114; MedChemExpress, New York, United States) were added to THP-1 derived macrophages 1.5 h before stimulation. Both THP-1 derived and the THP1-KO-ASC macrophages were treated with 10 μM rFIP-nha (WT) for 16 h, with the LPS amount of rFIP-nha (WT) as LPS-CN, and 1 μg/mL LPS as LPS + CN. 2 μg/mL poly (dA:dT) (InvivoGen, California, United States) was transferred to macrophages via Lipofectamine™ 2000 following the manufacturer’s protocol as AIM2 positive control (AIM2+). Cells exposed to medium were included as non-treated control (CN). Medium was collected for further analysis.

In the macrophage supernatant, IL-1β, IL-12/IL-23 (p40) (shown as IL-12 in the following content) and IL-10 were measured according to the manufacturer’s protocol (BioLegend, Koblenz, Germany). All samples were measured using a Tecan Infinite 200PRO (Tecan, Männedorf, Switzerland).

Macrophages were treated with 10 μM rFIP-nha for 3 h and 6 h, with the LPS amount of rFIP-nha as LPS-CN (Supplementary Table S1; 3.15 EU/mL), and 1 μg/mL LPS as LPS + CN. RNA was extracted by lysing cells with 400 μL TRIzol (Invitrogen, Bleiswijk, Netherlands) for each well. This was followed by RNeasy (Qiagen, Venlo, Netherlands) on column clean-up according to manufacturer’s protocol. The integrity of the ribosomal RNA was verified through 1% agarose (Eurogentec, Liège, Belgium) gel electrophoresis. RNA concentration and purity were checked using the Nanodrop spectrophotometer system (Nanodrop Technologies, Wilmington, DE, United States) and only samples with a ratio (Abs 260/280 nm) between 1.8 and 2.1 were used for qPCR. Subsequently, cDNA was synthesized using iScript (Bio-Rad, Veenendaal, Netherlands) and 200 ng RNA per reaction, according to manufacturer’s protocol.

qPCR was performed as previously described by Tang et al. (2017). Primers were derived from the Harvard Primerbank (http://pga.mgh.harvard.edu/primerbank/) and synthesized by Biolegio (Nijmegen, Netherlands). The cycling conditions used for amplifying the target sequences using the CFX96 Touch Real-Time PCR Detection System (Bio-Rad) were as follows: 90 s at 95°C, followed by 40 cycles at 95°C for 10 s, 58°C for 10 s and 72°C for 15 s, with a final elongation step at 72°C for 2 min (Tang et al., 2017). qPCR was performed in technical duplicate, and all samples were normalized to the geometric means of RPLP0 (ribosomal protein lateral stalk subunit P0; 60S acidic ribosomal protein P0) and ActinB expression (ΔCT normalisation towards two reference genes) and then normalised against medium-stimulated macrophages (ΔΔCT normalisation, setting NT to 1) via the ΔΔCT method using the qbase + software (Biogazelle, Gent, Belgium). The primers’ information is listed in Supplementary Table S2.

Phagocytosis assays and cytokine secretion assays were independently repeated five times (n = 5 independent biological replicates); other experiments were repeated three times independently (n = 3). Statistical analyses were carried out using Graphpad Prism 8. All parameters are presented as means ± SEM. A One-way ANOVA with a Dunnett post-hoc test to correct for multiple comparisons was used to assess the parameters for significance (p < 0.05; p < 0.1 is considered a trend). All groups were only compared to the non-treated control (NT). Outliers were identified using the ROUT method with Q = 1%. Graphs were plotted using Graphpad Prism 8.

To investigate the immunomodulatory activity of rFIP-nha on macrophages, we first determined whether the rFIP-nha was taken up by the macrophages using FITC-rFIP-nha (no cytotoxicity was observed on macrophages in the tested concentration range up to 1 µM FIP-nha; data not shown). After application, the macrophages started gradually to take up FITC-rFIP-nha (Figure 1). Within 1 h, FITC-rFIP-nha was present in the cytoplasm of almost all macrophages (see merged enlargements in Figures 1A–D). These results suggest that rFIP-nha could exert its activity inside the cell.

Macrophages are crucial in the elimination of diseased and damaged cells through phagocytosis. To assess the impact of rFIP-nha on macrophage phagocytic activity, we considered both the percentage of macrophages engaged in the phagocytosis of fluorescent Escherichia coli particles and the quantity of E. coli particles consumed in each macrophage, measured as mean fluorescence intensity (MFI). As shown in Figure 2, rFIP-nha dose-dependently inhibited macrophage phagocytosis. Initially, the MFI of macrophages was around 1,000,000, with a phagocytic rate of approximately 40% in the control media treatment. Exposure to rFIP-nha (0.01 μM–10 μM) led to a gradual decrease in macrophage MFI to approximately 200,000 and a reduction in phagocytosis to 10% (Figures 2A,B; Supplementary Figures S3–S5). As pro-inflammatory macrophages (M1) have been described to be less phagocytotic compared with anti-inflammatory (M2) macrophages (Hoppenbrouwers et al., 2022), this could potentially indicate that rFIP-nha might skew macrophages towards a pro-inflammatory phenotype. Meanwhile, rFIP-nha glycosylation did not seem to affect phagocytosis (Figures 2C,D; Supplementary Figures S6–S8). A degree of cell aggregation was observed for cells exposed to rFIP-nha and its mutants (Supplementary Figure S9). This is likely caused by cross-linking of cell surface carbohydrates, often observed by lectin-like proteins (Chen et al., 2021).

To further investigate the rFIP-nha’s effect on macrophage cytokine secretion, IL-1β, IL-12 and IL-10 were measured. As shown in Figure 3, rFIP-nha increased the production of all cytokines significantly compared to the medium control and LPS negative control. Interestingly, the secretion of IL-12 and IL-10 induced by rFIP-nha was significantly higher compared to the LPS positive control. These results indicate that rFIP-nha induces a pro-inflammatory THP-1 macrophage phenotype, which is similar, but not exactly like M1 macrophages induced by LPS. There was no obvious difference between the WT rFIP-nha and the glycosylation mutants in IL-1β secretion. Noteworthy, N39A induced less IL-12 and IL-10 secretion compared with other variants (Supplementary Figure S1). From microscopic photo analysis (Supplementary Figure S2), it was concluded that no obvious morphological changes were observed after exposure to rFIP-nha.

To further investigate the mechanism behind the secretion of IL-1β by rFIP-nha treated macrophages, the main classes of inflammasomes were evaluated at transcriptional level. As no significant difference of IL-1β secretion was observed between rFIP-nha and its mutants, we used rFIP-nha WT as a representative in the next experiments. As shown in Figure 4, the AIM2 inflammasome and IL-1β were significantly up-regulated after rFIP-nha treatment for 6 h. IL-1β and AIM2 expression increased around 10-fold, indicating that rFIP-nha induced IL-1β expression and secretion via AIM2 inflammasome regulation.

As rFIP-nha was expressed by P. pastoris, we investigated whether (a non-protein) yeast contamination could induce the false appearance. We inactivated rFIP-nha with heat treatment. The heated rFIP-nha failed to modulate inflammasome activation (Figure 5), suggesting that the activation of the AIM2 inflammasome was induced by rFIP-nha rather than by (for instance) DNA contamination originating from yeast. To validate inflammasome modulation, positive controls for activation of AIM2, NLRP3, and NLRC4 were included. Although not statistically significant, the AIM2 positive control (p = 0.06) and WT (p = 0.08) upregulated IL-1β at transcriptional level (Figure 5A), but not in IL-18 (Figure 5B). The AIM2 and NLRP3 (not significant) positive controls also upregulated their respective inflammasomes at the transcriptional level (Figures 5C,E). RFIP-nha WT showed a trend in AIM2 upregulation (p = 0.06). However, the NLRC4 positive control did not upregulate NLRC4 expression. It may be necessary to optimize the timing of gene expression assessment for this particular inflammasome complex.

To investigate whether IL-1β release is indeed AIM2 inflammasome-dependent, a THP1-KO-ASC macrophage cell line and two inhibitors associated with AIM2 inflammasome regulation were tested. The apoptosis-associated speck-like protein (ASC) is an essential adaptor of the inflammasomes AIM2 and NLRP3 to process proinflammatory cytokines such as IL-1β (Li Y. et al., 2021). The IL-1β release induced by rFIP-nha WT from ASC-KO-THP-1 macrophages was significantly lower than from native THP-1 macrophages, indicating that the observed IL-1β release is at least partially ASC pathway dependent (Figure 6). A151, a synthetic oligodeoxynucleotide, functions as a competitive inhibitor of the dsDNA–AIM2 interaction (Kaminski et al., 2013). It did not downregulate the IL-1β release induced by rFIP-nha WT, but partially blocked the inflammasome assembly in both macrophages. In the THP1 macrophages the A151 contributes to a more than 500 pg/mL downregulation (P value around 0.01), while it also diminishes IL-1β release in the ASC-KO-THP-1 macrophage (P value around 0.05). J114, another inhibitor, disrupts the interaction of NLRP3 or AIM2 with the adaptor protein ASC and inhibits ASC oligomerization to diminish IL-1β release (Jiao et al., 2022). This inhibitor worked less effectively in inhibiting the effects of the AIM2 positive control, while exhibiting limited efficacy in IL-1β release induced by rFIP-nha (Supplementary Figure S10). This discrepancy may arise from the possibility that rFIP-nha interacts with the AIM2 inflammasome in a manner distinct from dsDNA. Also, there might be other ASC-related pathways involved, as the IL-1β release induced by rFIP-nha in the THP1-KO-ASC macrophages was still 2-fold higher than induced by the AIM2 positive control.