The Sporothrix strains used in our study are S. schenckii (ATCC MYA 4820) and S. brasiliensis (ATCC MYA 4823), both clinical isolates from the same geographical hyperendemic region, in Rio de Janeiro state, Brazil. The virulence profile for these clinical isolates has been determined (11), and their cell wall structures were well characterized (31). To obtain them in yeast form, both the strains were grown in YPD broth (pH 7.8; Difco, Detroit, MI, USA) for 4 days at 37°C under orbital agitation, as previously described (25, 31). The yeasts cells were collected, passed through a Falcon filter (40 µm) to remove any hyphal fragments and/or yeast clusters (31), washed twice in Dulbecco’s modified Eagle’s medium (DMEM) (LGC Biotecnologia, SP, Brazil), and counted in a hemocytometer. For the interaction assays, the yeast cell concentration was adjusted accordingly.

Blood samples were collected from cubital veins of eight healthy volunteers, and human monocyte-derived macrophages were generated as described previously (32, 33). All volunteers gave informed consent prior to their participating in this study. The number of the Certificate of Presentation for Ethical Consideration related to this study is 62785716.2.0000.5259. Briefly, peripheral venous blood was collected in ethylenediamine tetraacetic acid (EDTA)-coated tubes, pooled for each donor, and diluted in Hanks’ balanced salt solution (HBSS, Cultilab, SP, Brazil). Lymphoprep (Axis-Shield, Oslo, Norway) gradient fractionation was used to separate PBMCs from the whole blood. Isolated PBMCs were then washed with HBSS and resuspended in DMEM (LGC Biotecnologia, SP, Brazil). To isolate monocyte subpopulation from PBMCs, a positive selection was performed using human CD14+ microbeads (Miltenyi Biotec, Bergisch Gladbach, Germany). In parallel, sera were separated from clotted blood samples of each donor (whole serum samples); part of the sera was heat-inactivated at 56°C for 30 min (complement-inactivated serum samples). The purified CD14+ monocytes were then suspended in DMEM supplemented with 1% Pen/Strep/Glut (10,000 unit/ml penicillin; 10,000 µg/ml streptomycin; 29.2 mg/ml glutamine) solution (GIBCO^®, NT, USA) and 10% (v/v) of autologous serum samples. These cells were seeded in culture plates and maintained for 7 days at 37°C in an atmosphere of 5% CO₂ for differentiation into macrophages (hMDMs).

The hMDMs were inoculated with yeast cells of S. schenckii or S. brasiliensis. All interaction assays were performed in a 5% CO₂ incubator at 37°C. The culture medium used was DMEM supplemented with 1% (v/v) of Pen/Strep/Glut and 10% (v/v) of either autologous whole human serum (wHS) or autologous inactivated human serum (iHS). The multiplicity of infection (MOI) of Sporothrix:hMDM was 3:1 or 5:1, as indicated for each experiment.

For Fluorescence-activated cell sorting (FACS) analysis and live microscopy, live yeasts were stained for 20 min at ambient temperature in the dark with 1 mg/ml fluorescein isothiocyanate (FITC; Sigma, Dorset, UK) in 0.05 M carbonate-bicarbonate buffer pH 9.6 (BDH Chemicals, VWR International, Leicestershire, UK). Following incubation, yeasts were washed with phosphate-buffered saline (PBS; 3×) to remove excess FITC and resuspended in PBS for a final suspension concentration of 10⁸ FITC-labeled yeasts/ml.

Assays were performed using a standard protocol with modifications (33, 34). hMDMs were washed in PBS, resuspended in prewarmed supplemented serum-free CO₂-independent medium containing LysoTracker red DND-99 (1 µM; Invitrogen, Paisley, UK), and used immediately for the phagocytosis assays. Live FITC-labeled S. schenckii and S. brasiliensis yeast cells at 3:1 MOI were added to 10⁶ LysoTracker red DND-99-stained hMDMs seeded in a sterile glass-based imaging slide chamber of eight wells and cocultured in CO₂-independent medium supplemented with 10% of whole human serum. 3i-Live Cell Imaging Systems featuring Yokogawa CSU-X1 (Spinning Disk Confocal Microscope coupled with an EMCCD camera, PerkinElmer, USA) was performed at 37°C in the microscope chamber. Images were captured at 1-min intervals over up to 18 h. The time lapse to start image capture (time zero) was 5 min. The Volocity^® 6.2 imaging analysis software was used to track phagocyte migration and yeasts uptake at 1-min intervals throughout the phagocytosis duration of 30 min or overnight (18 h).

S. schenckii and S. brasiliensis were analyzed for their interaction with hMDM in the presence of wHS or heat-inactivated human serum (iHS) as well as to evaluate phagocytic index upon blocking CR3, TLRs and Dectin-1. To block CR3, TLR2, TLR4, and Dectin-1 receptors, monoclonal anti-CD11b (Thermo Fisher, MA5-16528), anti-TLR2 (Abcam, ab16894), anti-TLR4 (Abcam, ab22048), and anti-Dectin-1 antibodies (35) were used, respectively. Briefly, hMDMs were preincubated with fresh DMEM medium supplemented with 10% of wHS for 30 min at 37°C in an atmosphere of 5% of CO₂. Monoclonal antibodies were then added in a final concentration of 10 µg/ml and incubated further at 8°C for 1 h. Then, hMDMs were infected with FITC-labeled yeasts of S. schenckii or S. brasiliensis, and the interaction was allowed for a duration of 1 h at 37°C in a CO₂ incubator. The hMDMs were washed one time with PBS (GIBCO^®, NY, USA) pH 7.4 and incubated for 15 min at ambient temperature with Trypan blue (GIBCO^®, NY, USA) at a final concentration of 1.2 mg/ml. The cells were washed twice with PBS and fixed with 1% p-formaldehyde for 15 min at ambient temperature. Finally, hMDMs were washed once with PBS and, with the aid of a cell scraper, were gently released from the wells, suspended in PBS supplemented with 3% fetal bovine serum, and analyzed in a BD FACS-Canto™ II.

To perform kinetics assay of phagocytosis with image analysis, hMDMs were cultivated on circular coverslips in a 24-well culture plate (2.5 × 10⁵ hMDMs per well) and infected with S. schenckii or S. brasiliensis yeasts at MOI 3:1. The interaction assay was performed for 1, 4, and 18 h, and the culture medium was supplemented with wHS or iHS. After each interaction time, the number of yeasts inside the macrophages was counted (defined when a clear halo was observed around the yeast). At least 10 microscopy fields were imaged, and a minimum of 50 hMDMs were counted. Photomicrographs were done in the ×100 optical magnitude. This experiment was performed two times, each time in duplicate.

To determine cytokine and long pentraxin 3 (PTX3) secretions, hMDMs were cultivated in a 96-well culture plate (0.5 × 10⁵ hMDMs per well), infected with S. schenckii or S. brasiliensis (MOI, 5:1). The interaction assay was performed for 4 or 18 h. At these time points, the culture plates were centrifuged (200 g, 15 min, 4°C) to remove cell debris. The culture supernatants were collected, used immediately, or stored at -80°C until use. PTX3 secretion was determined by Quantikine ELISA Pentraxin kit (R&D Systems). Tumor necrosis factor (TNF)-α, interleukin (IL)-1β, and IL-10 concentrations in the culture supernatants were determined by ELISA using kits (R&D Systems^®, USA) according to the manufacturer’s instructions. Non-infected hMDM was used as a negative control, and hMDM incubated with lipopolysaccharide (LPS; 10 ng/µl; Sigma-Aldrich^®, Brazil) was used as a positive control. CR3-, TLR2-, or TLR4-blocked hMDMs were also infected with S. schenckii or S. brasiliensis for 18 h. TNF-α and IL-1β concentrations in the culture supernatants were determined by ELISA. Non-infected hMDMs were used as non-infected control; hMDMs infected with Sporothrix species in the presence of wHS were used as the infected controls. Experiments were performed two/three times in duplicate (n = 4 or 6).

The cytotoxicity assay was performed in 96-well culture plates seeded with 0.5 × 10⁵ hMDMs/well, infected with Sporothrix yeasts at MOI 5:1; the interaction was arrested after 4–18 h. After each time point, the culture supernatant was collected, used immediately, or stored at -80°C until use. The lactate dehydrogenase (LDH) activity was measured using CytoTox^® Non-Radioactive Cytotoxicity Assay (Promega) according to the manufacturer’s instructions. This assay was performed two times in duplicate/triplicate (n = 5).

Direct C3b binding to Sporothrix yeast surfaces was checked by immunofluorescence labeling. Briefly, Sporothrix yeasts were fixed with 2.5% paraformaldehyde (PFA) at ambient temperature for 1 h and then overnight at 4°C. After quenching with 0.1 M NH₄Cl, fixed Sporothrix yeasts were blocked with PBS-BSA (1%) at ambient temperature for 1 h, incubated with recombinant C3b [5 μg/ml in GVB+ (preparation that is explained below); Merck Millipore], added with anti-C3b antibodies (5 μg/ml in PBS-BSA; mouse monoclonal, MA1-40155, Invitrogen), incubated for 1 h, washed with PBS (3×), incubated with tetramethylrhodamine (TRITC)-conjugated mouse immunoglobulin G (IgG; Sigma) in the dark for 1 h, washed with PBS (3×), mounted on slides, and observed for fluorescent labeling under confocal microscope (LSM700, ZEISS).

PRMs obtained from both Sporothrix species as described previously (36) were dissolved at 50 µg/ml in sodium bicarbonate buffer (50 mM, pH 9.6). The 96-well microtiter plate was coated with PRM solution (100 μl/well) overnight at ambient temperature. The supernatants were discarded, and the wells were washed once in wash buffer (PBS supplemented with 0.05% Tween-20). Wells were then blocked with PBS containing 1% BSA (PBS-BSA) at ambient temperature for 1 h and washed once with wash buffer. Gelatin veronal buffer (GVB-) was prepared with 5 mM barbital, 145 mM NaCl, 0.1% gelatin, pH 7.4. By adding MgCl₂ and CaCl₂ to the final concentrations of 0.5 mM and 0.15 mM, respectively, GVB+ was prepared; EGTA was added to the final concentrations of 5 mM to ensure Ca⁺²/Mg⁺² depletion. wHS or complement component Factor-B-depleted serum (CompTech, TX, USA) was diluted in GVB+/- (8 µl serum + 92 µl GVB) and was added to the PRM-coated wells, incubated at ambient temperature for 1 h, washed twice with wash buffer, added with peroxidase-conjugated anti-human complement C3 antibody (MP Biomedicals; Cat-No. 55237) in PBS-BSA (1:5,000 dilution, 100 µl/well), and incubated at ambient temperature for 1 h. The wells were washed thrice with wash buffer and added with the substrate (TMBW, Tebu-Bio; 100 µl/well). After 5 min, the color development was arrested by adding 4% H₂SO₄ (50 µl/well). The optical density was measured at 450 nm using microplate reader (Infinite m200 pro, TECAN). C3-PRM interaction was also studied using C1q and MBL-depleted sera following the protocol described for wHS.

For PRM-CR3 interaction study, PRM-coated wells (50 µg/ml in sodium bicarbonate buffer) in a 96-well microtiter plate were added with recombinant human CD11b (50–100 ng/well; Abcam ab126010) in PBS-BSA (1%), incubated at ambient temperature for 1 h, washed with PBS-Tween-20 (0.05%), added sequentially with monoclonal anti-human CD11b antibodies (Sony, clone M1/70) and secondary peroxidase-conjugated human IgG, incubated at ambient temperature for each antibody added, and washed with PBS-Tween-20 between each step. Furthermore, chromogenic peroxidase substrate (TMBW, Tebu-Bio; 100 µl/well) was added, and the color developed was optically read at 450 nm.

PRMs from S. brasiliensis (16.0 mg/ml) and S. schenckii (14.5 mg/ml) were obtained as described previously (36), dissolved in D₂O (99.99%, Eurisotop, France) to be analyzed by NMR in natural ¹³C abundance. NMR experiments were performed at 25°C on an 18.8 T Avance Neo spectrometer (Bruker, Billerica, USA) with a ¹H resonating frequency of 800.62 MHz equipped with a triple resonance (¹H, ¹⁵N, ¹³C) cryogenically cooled probe. Spectra were recorded and processed with Topspin 4.08 (Bruker) and analyzed with CCPNMR Analysis 2.5 (37).

To assign the carbohydrate moiety resonances of PRMs, we analyzed ¹H homonuclear 1Ds and 2D DQFCOSY [double-quantum-filtered correlation spectroscopy (38)], TOCSY [total correlation spectroscopy, 60 ms mixing time (39)], ROESY [rotating frame Overhauser enhancement spectroscopy, 200 ms mixing time (40)], and NOESY [nuclear Overhauser enhancement spectroscopy, 200 ms mixing time (41)] spectra and 2D ¹H–¹³C heteronuclear experiments: HSQC [heteronuclear single quantum coherence (42)] and its edited version to differentiate CH₂ groups from CH and CH₃ groups, HMBC [heteronuclear multiple bond coherence (43), HSQCTOCSY (30, 60 and 100 ms mixing times] and H2BC [heteronuclear two bond correlation (44)]. Typically, the spectral width for ¹H was 5.5 ppm (centered at 3.1 ppm) and 56 ppm for carbon (centered at 84 ppm). ¹H–¹³C spectra were folded and recorded with non-uniform sampling (50%) in the indirect ¹³C dimension to increase resolution.

Statistical analyses were performed using GraphPad Prism 9 (San Diego, CA, USA). Unpaired two-tailed t-tests and one-way analysis of variance (ANOVA) with Tukey’s multiple comparison posttest or Dunnett’s multiple comparison posttest were performed. The statistical details of each result can be found in the figure legends; “n” represents the total number of replicates used in the assay and statistical analysis. Differences between data were considered significant when p ≤ 0.05.

First, the ability and dynamics of hMDMs to recognize and phagocytose S. schenckii and S. brasiliensis yeasts were monitored. These human immune cells in the presence of autologous serum were able to actively phagocytose the yeasts of both Sporothrix species within 30 min of interaction ( Supplementary Videos S1 and S2 ). After 18 h of interaction of Sporothrix yeasts with hMDMs, some germinating cells could be observed ( Figure 1A ). Moreover, an increase in the number of S. brasiliensis yeasts inside the hMDMs after 18 h of interaction was noticed. The germination of S. brasiliensis inside hMDM is illustrated by a microscopic observation after an overnight interaction (18 h) in the presence of wHS ( Supplementary Video S3 ), which demonstrates the uptake of S. brasiliensis yeasts by hMDMs and their germination within these cells.

We then performed phagocytic studies in media supplemented with wHS or heat-inactivated human serum (iHS) ( Figure 1B ). After 1 h interaction with hMDMs, flow cytometry analysis showed a significant decrease in the uptake of the two Sporothrix species in the medium containing iHS compared to that containing wHS, suggesting that thermolabile serum factors facilitate the uptake of Sporothrix yeasts by hMDMs. On the other hand, the number of Sporothrix yeasts taken up by hMDMs was higher for S. brasiliensis compared to S. schenckii ( Figure 2 ). Moreover, there was an increase in the number of S. brasiliensis yeast in hMDMs after 18 h ( Figure 2B ), consistent with the microscopic observation ( Figure 1 ), suggesting their replication inside hMDMs. Together, these results confirm that hMDMs recognize S. schenckii and S. brasiliensis yeasts distinctively and that thermolabile serum factors mediate the uptake of both species, albeit at different degrees.

Cytotoxicity of Sporothrix species was determined by the LDH assay upon 4–18 h of interaction in hMDMs challenged with the Sporothrix yeasts. The results are expressed by the relative percentage to the uninfected control in Figure 3 . In the medium containing iHS, both Sporothrix species showed less than 20% of cytotoxicity, which could be related to their low phagocytic levels. On the other hand, after 18 h of interaction in the presence of medium containing wHS, the cytotoxicity of S. brasiliensis was significantly higher compared to that of S. schenckii (p < 0.05), which could be attributed to the higher level of phagocytosis of S. brasiliensis and its capacity to germinate/multiply.

The secretion of the pro-inflammatory cytokines, TNF-α and IL-1β, was evaluated after 4 and 18 h of interaction of Sporothrix yeasts with hMDMs ( Figure 4 ). In the medium containing wHS, both S. schenckii and S. brasiliensis stimulated the secretion of TNF-α by hMDMs at comparable levels, which remained at the same level for S. schenckii even after 18 h of interaction. However, from 4 to 18 h, there was a significant increase in the TNF-α secretion by hMDMs stimulated with S. brasiliensis. IL-1β was detectable in the Sporothrix–hMDM culture supernatant only after 18-h interaction. For the medium supplemented with iHS, lower levels of TNF-α and IL-1β secretion by hMDMs were observed for both Sporothrix species. Furthermore, only S. brasiliensis-stimulated hMDMs showed a significant increase in the TNF-α secretion during the course of incubation from 4 to 18 h. Neither of the Sporothrix species could stimulate a significant release of IL-10 in either wHS- or iHS-supplemented media. To rule out any artifact in the hMDM cytokine secretion profile due to any indirect effect of iHS compared to wHS, similar experiments were performed with LPS as a positive control. The hMDM cytokine secretion profile in response to LPS stimulation was similar in both wHS- and iHS-supplemented media ( Figure 4 ). These results suggest that the hMDM cytokine response was not directly influenced by human serum supplement (wHS or iHS) in the culture medium, but through their effect on the phagocytosis of Sporothrix yeasts ( Figure 2 ).

As heat inactivation of the human serum directly impacts the phagocytosis and inflammatory response of hMDMs toward Sporothrix species, we investigated the role of the heat-labile complement system in the serum against these species. We first targeted CR3 (the major complement receptor expressed on macrophages) (45). Upon blocking CD11b and CD18, the CR3 components, using monoclonal antibodies, there was a significant reduction in the secretion of IL-1β by Sporothrix-infected hMDMs ( Figure 5A ). Nevertheless, the blockage of CD11b (the unique component of CR3) did not inhibit the uptake of S. schenckii and S. brasiliensis yeasts, suggesting the involvement or the cooperation of other pattern recognition receptors (PRRs) in their recognition and phagocytosis by hMDMs ( Figure 5B ).

As the TNF-α secretion by hMDMs upon Sporothrix species infection was not inhibited by blocking CR3, we further investigated the possible participation of other PRRs ( Figure 6 ) already described to mediate TNF-α release by human PBMCs infected with Sporothrix species morphotypes (25). TLR4 blockage on hMDMs resulted in an increase in the production of TNF-α upon infection of hMDMs with Sporothrix species ( Figure 6A ). In contrast, a significant reduction in TNF-α secretion was observed upon TLR2 blockage of hMDMs challenged with Sporothrix species, although this reduction was not statistically significant for S. schenckii-infected hMDMs blocked for TLR2 ( Figure 6A ). Blockage of Dectin-1 on hMDMs also resulted in significant differences in the TNF-α secretion by hMDMs compared to unblocked hMDMs, suggesting that TNF-α secretion by hMDMs upon interaction with Sporothrix is also Dectin-1 recognition dependent ( Figure 6B ).

We first checked the opsonization of Sporothrix yeasts with C3b. Immunolabeling indicated that C3b gets deposited on the surface of the yeasts of both S. schenckii and S. brasiliensis ( Figure 7A ). As the fungal cell wall components are the major source of PAMPs and the first to interact with the host immune system (22) and considering that the PRM is a well-characterized component located on the outermost cell wall layer of S. schenckii and S. brasiliensis yeast parasitic phase (31), we examined the complement activation capacity of the PRMs extracted from these two Sporothrix species. Total carbohydrate content in the PRMs of the two Sporothrix species was determined with phenol-sulfuric acid method to use them at equal concentrations for the investigation of their complement activation capacities. In GVB+ (containing Ca⁺²/Mg⁺²) medium, which facilitates the activation of all three complement (classical, lectin, and alternative) pathways, PRMs from both S. schenckii and S. brasiliensis could activate the complement system (readout was deposited C3b).

In GVB- medium (with EGTA, a Ca⁺²/Mg⁺² chelator) that allows the activation of only the alternative pathway, PRM from S. schenckii showed C3b deposition similar to that in GVB+ medium, but there was a significant decrease in the complement activation by S. brasiliensis PRM. This suggests that the PRM of S. schenckii activates the complement system mainly via the alternative pathway, whereas that of S. brasiliensis partially activates complement system through alternative pathway. In accordance, complement Factor-B (FB)-depleted serum in GVB- medium completely abolished the complement activation by PRMs extracted from both S. schenckii and S. brasiliensis ( Figure 7B ), as FB is involved in the amplification loop for the activation of C3 into C3b via alternative pathway. On the other hand, FB-depleted serum in GVB+ medium allowed partial regaining of the complement activation by PRMs from both Sporothrix species ( Figure 7B ), suggesting that S. brasiliensis PRM is also capable of activating complement system through classical and lectin pathways. Furthermore, these observations are supported by mannose-binding lectin (MBL)- and C1q-depleted sera (the key components of lectin and classical pathways, respectively), wherein S. schenckii PRM was mostly independent of MBL/C1q for C3 activation and S. brasiliensis PRM showed partial dependence ( Figure 7C ).

We also investigated the direct interaction between PRMs and CD11b, the specific component of CR3. ELISA upon coating PRMs and incubating them with CD11b indicated that there is direct and specific interaction between PRMs from both S. schenckii and S. brasiliensis, as the binding between PRM and CD11b was concentration dependent ( Figure 7D ). Altogether, the results suggest that PRMs could be the PAMPs of Sporothrix species directly recognized by the CR3 expressed on hMDMs.

At this point, it was important to have structural data to test whether S. brasiliensis and S. schenckii PRMs were different, which could be at the origin of the differential complement activation pathways. Thus, to correlate the PRM structures with the different immune response behavior to Sporothrix species, we characterized S. schenckii and S. brasiliensis PRMs by ¹H and ¹H-¹³C NMR. Analysis of the 2D homonuclear and ¹H-¹³C spectra of PRMs led to the assignment of many monosaccharide units and their linkages ( Figure 8 ). As described earlier for S schenckii (46), S. brasiliensis PRM contains chains of linear Rhap-α (1,4) GlcAp-α (1,2) Manp α1. α-Rhamnose residues were additionally linked to disubstituted GlcAp through α (1,2) linkages ( Figure 8 ). These chains have previously been ascribed to O-linked rhamnomannans of S. schenckii, linked to Ser/Thr residues of the peptide moiety by an additional mannose residue (46, 47). Here, however, we could not assign the later peptide O-linked mannose. These Rhap-α (1,4) GlcAp-α (1,2) Manp α1 and Rhap-α (1,4) [Rhap-α (1,2)] GlcAp-α (1,2) Manp α1 structure is here described, for the first time, in the S. brasiliensis PRM. NMR data further indicated that S. brasiliensis displayed Man α (1,6) chains with α-Rha residues branched at position 3 (31), a type of chain that was previously ascribed to N-linked PRMs of S. schenckii (47, 48). Thus, PRMs of both the Sporothrix species share common O- and N-linked chains. Nevertheless, it is important to underline that S. brasiliensis contains at least nine additional sugar residues with respect to S. schenckii ( Figure 8 ). Five of these were assigned to rhamnose, in agreement with the higher rhamnose content observed of S. brasiliensis (31). Despite common elements in both species PRMs, the S. brasiliensis PRMs contain additional rhamnose-rich structure(s).

The long PTX3 is a soluble pattern recognition molecule produced by somatic and immune cells at the site of infection. It is reported to crosstalk with other factors of the innate immune system (48). Our data showed an increase in PTX3 secretion by hMDMs challenged with Sporothrix species ( Figure 9A ). Interestingly, PTX3 secretion by hMDMs stimulated with S. schenckii and S. brasiliensis was evident only in the medium supplemented with wHS but not in the presence of iHS, suggesting the requirement of complement-mediated phagocytosis of Sporothrix species for the secretion of PTX3 by the immune cells. On the other hand, blockage of CR3 did not have any effect on PTX3 secretion by hMDMs upon interaction with S. schenckii and S. brasiliensis yeasts, whereas blockage of both CR3 and TLR4 (but not TLR4 alone) resulted in a significant decrease in the PTX3 secretion ( Figure 9B ). This crosstalk between the complement system and TLR4 in stimulating PTX3 production by hMDMs upon interaction with S. schenckii or S. brasiliensis and the role of PTX3 against Sporothrix species are subjects to be investigated further.