Immunity to pathogenic mucosal C. albicans infections mediated by oral megakaryocytes activated by IL-17 and candidalysin

Abstract

The fungus Candida albicans can cause mucosal infections including oropharyngeal candidiasis (OPC) in immunocompromised patients. In humans, an increased risk of fungal infections correlates with thrombocytopenia. However, our understanding of platelets and megakaryocytes (Mks) in mucosal fungal infections is almost entirely unknown. When megakaryocyte- and platelet-depleted mice were infected with OPC, the tongue showed higher fungal burden, due to decreased neutrophil accumulation. Protection depended on a distinct population of oral-resident Mks. Interleukin-17, important in antifungal immunity, was required since mice lacking the IL-17 receptor had decreased circulating platelets and their oral Mks did not expand during OPC. The secretion of the peptide toxin candidalysin activated human Mks to release platelets with antifungal capacity. Infection with a candidalysin-deficient strain resulted in decreased expansion of tongue Mks during OPC. This is the first time that a distinct megakaryocyte population was identified in the oral mucosa which is critical for immunity against fungal infection.

INTRODUCTION

Platelets are typically associated with maintaining hemostasis (blood clotting), including in the context of infections that induce severe vascular damage.¹ Yet, it is now better appreciated that platelets are directly involved in the response against pathogens including fungi.^2–4 While activated platelets facilitate hemostasis,^5,6 these cells are also activated by fungal cell wall components during systemic infection.⁷ Platelet microparticles released in response to Candida albicans (C. albicans) directly kill the fungus in vitro^8,9 and levels correlate with patient outcomes during invasive fungal infections (IFIs) including disseminated candidiasis.^10,11 The severity of disease caused by C. albicans is associated with a decreased level of RANTES (CCL5), a leukocyte chemokine released in abundance by platelets during IFIs.¹² Thrombocytopenic liver transplant recipients are at a higher risk of systemic fungal infections, also indicating the importance of maintaining platelet levels to prevent disease.^13,14 During systemic candidiasis, platelets adhere to circulating Candida species, including C. albicans, and these interactions are important to clear the fungi from the blood.¹⁵ Platelets also influence the function of other immune cells. In the context of bacterial infection, platelet-deficient mice have increased susceptibility to systemic Staphylococcus aureus infection due to decreased macrophage activity.^16,17 Platelets recruit neutrophils during inflammation through engagement of P-selectin on activated platelets with P-selectin glycoprotein ligand-1 (PSGL-1) on neutrophils leading to chemotaxis to the site of thrombus formation.^18–20 In addition, platelet secretion of chemokine C-X-C motif ligand 4 (CXCL4) and CXCL7 recruits and activates neutrophils during inflammation.¹⁹

Just as our understanding of platelet function is often narrowly confined to the context of hemostasis and related disorders, megakaryocytes (Mks) also contribute to host protection in ways that are not fully appreciated. As the parent cell, Mks are associated with the production of platelets, yet their function is not limited to platelet biogenesis.^21–23 Mks respond to various stimuli through expression of a wide range of immunoreceptors including Toll-like receptors and Fc-gamma receptors.²⁴ More recently, murine Mks were found to localize to other mucosal tissues such as the lung.²⁵ These distinct tissue Mks express higher levels of major histocompatibility complex class II (MHCII) and C-C chemokine receptor type 7 (CCR7) compared to bone marrow Mks.^22,25 Lung Mks also contribute to nearly 50% of all circulating platelet production in mice.²⁵ Furthermore, the lung contains both mature and immature Mk populations, as well as hematopoietic progenitor cells that can repopulate the bone marrow during thrombocytopenia or stem cell deficiency.²⁵ Mks also release platelets that subsequently contribute to tissue-specific immune responses. For instance, disruption of Mk-specific Syk-dependent signaling results in reduced phagocytic activity of responding neutrophils during murine pulmonary Aspergillus fumigatus infection.⁴ Together, this suggests Mks and platelets may be relevant to oral mucosal immune responses that rely on neutrophils for protection, such as during candidiasis.

In both humans and mice, protection against C. albicans is strongly associated with intact IL-17-mediated immunity. Interleukin-17A (IL-17) is produced by both innate lymphocytes (γδ T cells, αβ T cells, and ILC3s) and conventional Th17 cells.²⁶ During acute oropharyngeal candidiasis (OPC), IL-17 from innate sources is essential for protection against C. albicans through induction of protective neutrophil and antimicrobial peptide responses.

Even though the connection between Mks and platelets with the IL-17 family has not been explored in the context of OPC, there were indications that the Th17/IL-17 signaling axis was connected to Mk and platelet development and/or function in host defense. First, during Mk development progenitors express major histocompatibility complex (MHC) class II and act as antigen-presenting cells (APCs) that enhance Th17 responses.²⁷ Additionally, IL-17A induces expansion of Mk populations in mice, suggesting IL-17 contributes to Mk differentiation.²⁸ Platelets are also connected to Candida-directed Th17-mediated mucosal immune responses. Whereas the C. albicans secreted toxin, candidalysin, activates the GP1bα receptor on platelets, which induces the release of the immunomodulatory peptide dickkopf 1 (Dkk-1) that drives protective Th2- and Th17-mediated responses during allergic airway disease.²⁹

In this study, we further investigated the contribution of Mks and platelets during mucosal fungal infection using a model of OPC in which mice were challenged orally with C. albicans.³⁰ Mk- and platelet-depleted mice were more susceptible to OPC, which correlated with a decreased neutrophil response compared to Mk- and platelet-sufficient mice. In naïve wild-type (WT) mice, a population of Mks in tongue tissue expanded upon OPC infection. In vitro, human Mks killed C. albicans in a contact-independent manner. Mk expansion and platelet release were induced by the fungal toxin candidalysin both in vitro and in vivo. Strikingly, we uncovered that IL-17 signaling is involved in Mk function. Specifically, IL-17A stimulated platelet release by Mks and the absence of IL-17 signaling resulted in lower Mk and platelet populations in murine tongue tissue during infection. Furthermore, our data demonstrate a previously undescribed role for candidalysin in Mk development and activation in the oral cavity, resulting in the release of platelets that mediate neutrophil recruitment and subsequent protection against OPC.

RESULTS

Platelets drive immunity to oropharyngeal candidiasis

Previously, we compared the transcriptional profile of tongue tissue from OPC-infected Il17ra^−/− mice to infected control mice (WT) using RNA-Seq.³¹ Further analyses of these data sets showed differentially expressed platelet-related genes. These included Selp (P-selectin, involved in platelet activation and leukocyte tethering) and Prg4 (Proteoglycan 4, a Mk stimulating factor), which were upregulated in Il17ra^−/−mice. Meanwhile, G6b (Mk and platelet inhibitory receptor, involved in Mk differentiation, function, and platelet production) was downregulated in Il17ra^−/− mice (Fig. 1A), indicating that mice deficient in IL-17RA signaling have a defect related to platelet levels or function. Complete blood count analysis (collected by cardiac puncture) showed that while Il17ra^−/− mice were not thrombocytopenic (levels below 150,000 platelets/μL blood), platelets numbers were lower compared to WT control mice (Fig. 1B). The Il17ra^−/− mice had a defect specifically in platelets since there was not a deficit of other blood cells (Supplementary Fig. 1). We next determined if platelets contributed to susceptibility during OPC. These studies employed a Cre-lox model of platelet depletion via ablation of Mks. When using this system, mice expressing Cre recombinase (Cre) driven by the platelet factor 4 (PF4) promoter were crossed with mice ubiquitously expressing simian diphtheria toxin receptor (iDTR) preceded by a floxed stop codon.^17,32 The stop codon was removed in PF4⁺ Cre-expressing cells, resulting in expression of diphtheria toxin receptor on Mks, allowing for ablation by administering low doses (200–400 ng) of diphtheria toxin. Using a well-established model of OPC,³³ WT (PF4Cre⁻-iDTR^fl/fl referred to as WT) and platelet-depleted (PF4Cre⁺-iDTR^fl/fl referred to as PF4-iDTR) mice were challenged with C. albicans, followed by tongue tissue harvest and processing for fungal colony forming unit (CFU) enumeration 3 days after infection (Day 3). Typically, WT mice clear C. albicans from the oral cavity by Day 3–5 after infection, resulting in negligible tissue fungal burden, whereas immunocompromised mice have increased susceptibility correlating with higher fungal loads.²⁶ PF4-iDTR mice had significantly higher fungal burden (404 CFU/g tongue tissue) compared to immunocompetent WT mice (10 CFU/g tongue tissue) (Fig. 1C). Even though OPC typically is not fatal in WT mice, some PF4-iDTR mice succumbed during the infection. Mortality was not related to dissemination of Candida systemically, but rather to uncontrolled bleeding events in the absence of platelets (data not shown). There was also lethality in PF4-iDTR sham-infected mice, although the rate of death was not significantly different from platelet-sufficient WT mice (Fig. 1D). Nonetheless, the higher rate of mortality during platelet deficiency necessitated that fungal susceptibility be assessed on Day 3 post-infection. Taken together, these results show that IL-17 signaling promotes platelet homeostasis in the blood and that platelet depletion is detrimental during OPC.

Fig. 1.

Platelets contribute to the immune response during oropharyngeal candidiasis. (A) Differentially expressed genes in WT versus Il17ra^−/− mice on Day 1 of OPC. Fold change of genes induced in WT tongues compared to Il17ra^−/− tongues. (B) Circulating level of platelets in naïve WT and Il17ra^−/− mice (dotted line represents clinical level of thrombocytopenia). Results pooled from four separate experiments, n = minimum three mice per cohort per experiment, analyzed by Student’s t test. Mean ± SEM. (C) Fungal burden (CFU/g) of WT and PF4-iDTR murine tongue tissue during OPC. Results pooled from three experiments, n = minimum four mice per cohort per experiment, analyzed by Mann-Whitney t test. (D) Survival curve of WT and PF4-iDTR mice during OPC. N.s. represents comparison between WT Sham and PF4-iDTR Sham as well as comparison between PF4-iDTR Sham and PF4-iDTR OPC. PF4-iDTR OPC group was also compared to WT OPC and the p-value was less than 0.05. Results pooled from three experiments, n = minimum three mice per cohort per experiment, significance analyzed by log-rank test. *p < 0.05, **p < 0.01, ****p < 0.001, n.s. is not significant. CFU = colony forming unit; OPC = oropharyngeal candidiasis; SEM = standard error of the mean; WT = wild-type.

Platelets facilitate neutrophil influx to the tongue during oropharyngeal candidiasis

To further understand susceptibility to OPC during platelet insufficiency, we examined the influx of neutrophils into the tongue tissue by flow cytometric analysis in WT and PF4-iDTR mice. Neutrophils are involved in protection against OPC, and we have characterized the neutrophil populations involved in the response to Candida in the oral cavity by flow cytometric analysis.^26,31,34,35 On Day 1, the neutrophil population (GR-1⁺) was lower in PF4-iDTR tongue tissue compared to control mice (9.3% vs. 40.1% in WT mice) (Figs. 2A and 2B, Supplementary Fig. 2). This decrease in population size correlated with decreased expression of CD11b, a marker of neutrophil activity, in the PF4-iDTR mice compared to WT mice (Figs. 2C and 2D). These results showed that platelet depletion led to a decreased influx of neutrophils into the tongue tissue during OPC and that these cells displayed an altered activation status compared to neutrophils in oral tissue with platelets present.

Fig. 2.

Platelet depletion disrupts neutrophil influx to tongue tissue during oropharyngeal candidiasis. (A) Representative flow cytometry images of neutrophil (GR-1⁺) populations from the CD11b parent gate in WT and PF4-iDTR tongue tissue on Day 1 of OPC. Gating strategy in Supplementary Figure 2. (B) Quantification of GR-1⁺ flow cytometry data. Results pooled from two experiments, n = minimum three mice per cohort per experiment, analyzed by one-way ANOVA. Mean ± SEM. (C+D) Quantification of CD11b⁺ MFI (of GR-1⁺ cells) and GR-1⁺ cells positive for CD11b marker. Results pooled from two experiments, n = minimum three mice per cohort per experiment, analyzed by one-way ANOVA. Mean ± SEM. *p < 0.05, ***p < 0.001, ****p < 0.0001. ANOVA = analysis of variance; CD = cluster of differentiation; OPC = oropharyngeal candidiasis; SEM = standard error of the mean; WT = wild-type.

Platelets promote neutrophil trafficking in the oral mucosa

A productive immune response during OPC results in both an influx of neutrophils to the tongue tissue from circulation as well as chemotaxis of neutrophils to Candida-infected areas in the superficial epithelial layers.³⁶ Histological evaluation of tongue tissue on Day 1 of OPC (Fig. 3A) showed reduced neutrophil chemotaxis to the supra-basal layer of PF4-iDTR mice (~10 neutrophils per field of view) compared to WT mice (~1150 neutrophils per field of view) (Fig. 3B). Analysis of transcript levels of neutrophil chemokine and growth factor genes in tongue tissue that are important during OPC showed expression of Cxcl2, Ccl20, and Csf3 were lower in PF4-iDTR mice compared to WT mice, while Cxcl1 and 5 showed no difference (Fig. 3C). Even though a few neutrophil recruiting chemokines were lower in platelet-depleted mice, and could at least partially explain the attenuated neutrophil response, this would not account for the requirement of any direct interactions between platelets and neutrophils. These interactions could include platelets acting as a tether for neutrophils at the endothelium through expression of adhesion molecules to facilitate neutrophil chemotaxis to the highly vascularized oral mucosa. Similar to how platelets recruit neutrophils during thrombosis,³⁷ the most likely mechanism mediating platelet-neutrophil interaction is PSGL-1, which adheres to selectins including P-selectin (CD62P) expressed on the surface of activated platelets.³⁷ To test this, we blocked this interaction by administering WT mice with neutralizing PSGL-1 monoclonal antibody or immunoglobulin G (IgG) isotype control followed by OPC infection. Mice that were treated with α-PSGL-1 showed a defect in neutrophil recruitment compared to mice treated with IgG control (Figs. 4A and 4B). Since endothelial cells also express P-selectin we next tested if reconstitution of platelets rescued the neutrophil defect in PF4-iDTR mice. We intravenously transfused platelets from healthy, naïve WT mice into PF4-iDTR mice after OPC infection and analyzed neutrophil levels in the tongue. Neutrophil numbers in the tongue increased substantially in PF4-iDTR mice transfused with platelets compared to both WT and non-transfused PF4-iDTR mice (Figs. 4C and 4D). The replenishment of neutrophils correlated with a reduction of fungal susceptibility in the platelet-transfused PF4-iDTR mice (0 CFU/g tongue tissue) (Fig. 4E). Our results show that platelet depletion resulted in reduced neutrophil influx and chemotaxis into C. albicans-infected tongue tissue during OPC and that platelet reconstitution rescued the neutrophil defect in PF4-iDTR mice.

Fig. 3.

Platelet depletion disrupts neutrophil chemotaxis during oropharyngeal candidiasis. (A) Hematoxylin-eosin staining of WT and PF4-iDTR tongue tissue on Day 1 of OPC, enlarged images of selected regions. (B) Neutrophil quantification of the suprabasal layer on Day 1 and 2 of OPC. Results are pooled from three separate mice per group analyzed by Student’s t test between days. Mean ± SEM. (C) Relative expression of relevant neutrophil genes of WT sham-infected, WT OPC-infected, and PF4-iDTR OPC-infected mice on Day 1 of OPC. Results pooled from two experiments, n = minimum three mice per cohort per experiment, qPCR analysis performed in triplicate, analyzed by one-way ANOVA. Mean ± SEM. **p > 0.01, ***p > 0.001, ****p < 0.0001, n.s. is not significant. ANOVA = analysis of variance; OPC = oropharyngeal candidiasis; qPCR = quantitative polymerase chain reaction; SEM = standard error of the mean; WT = wild-type.

Fig. 4.

Blockade of PSGL-1 signaling results in decreased neutrophil response during oropharyngeal candidiasis. (A) Representative flow cytometry plots of neutrophil populations (GR-1⁺) in tongue tissue on Day 1 of OPC between WT Sham, WT OPC, Isotype control OPC, and α-PSGL-1 OPC mice. (B) Quantification of neutrophil flow cytometry data from (A). N = minimum three mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. (C) Representative flow cytometry plots of neutrophil populations (GR-1⁺) in tongue tissue on Day 1 of OPC between WT OPC, PF4-iDTR OPC, and platelet-transfused PF4-iDTR mice. (D) Quantification of neutrophil flow cytometry data from (C). N = minimum four mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. (E) Fungal burden (CFU/g) of platelet-transfused PF4-iDTR murine tongue tissue on Day 1 of OPC. Results representative of two independent experiments. **p > 0.01, ***p > 0.001, ****p < 0.0001, n.s. is not significant. ANOVA = analysis of variance; CD = cluster of differentiation; OPC = oropharyngeal candidiasis; PSGL-1 = P-selectin glycoprotein ligand-1; SEM = standard error of the mean; WT = wild-type.

Oral megakaryocytes expand during oropharyngeal candidiasis

Since the neutrophil response in the oral mucosa occurs within 24 hours during OPC, we next investigated if there was a localized Mk population in the tongue tissue contributing quickly to protection against Candida. For the data presented in this study, we refined the protocol to detect oral immune cells by flow cytometry using one processed murine tongue to represent each data point, versus having to combine multiple tongues for each sample.³⁴ This allowed the reproducible detection of relatively rare oral immune cell populations such as Mks. Similar to what has been shown in the lung, tongue tissue from perfused mice contained a distinct population of CD41⁺ Mks (0.113% of total tongue cells vs. 0.0003% isotype control) that was not present in PF4-iDTR mice (0.033%) (Figs. 5A and 5B).²² This population of Mks was also positive for the nuclear marker DRAQ5 eliminating the likelihood that the cells were platelet aggregates (Supplementary Fig. 3). In the bone marrow, Mks were detected at a similar percent of total cells (0.11%) compared to tongue Mks, and in both oral and bone marrow tissue the population was rare (Fig. 5B). Next, OPC was induced in WT mice and the Mk populations were assessed on Days 1 through 4 to examine the effects of infection on this population. Compared to WT sham-infected mice, the Mk population in infected mice was present on Day 1, expanded ~2-fold on Day 2, and was maintained on Day 4 post-infection (Figs. 5C and 5D). In comparison, bone marrow Mks did not expand upon infection in WT mice (Fig. 5E). In the tongue tissue, only a small percentage of the CD41⁺ population expressed macrophage or neutrophil markers (Supplementary Figs 4–6). These CD41⁺CD11B⁺F4/80⁺ macrophages or CD41⁺CD11B⁺Ly6G⁺ neutrophils represented cells likely coated with platelets. In addition, there were relatively few macrophages (80 cells per tongue) within the tissue compared to Mks (1000 cells per tongue). Moreover, only the Mks, not the macrophages, expanded upon OPC infection (Supplementary Fig. 4). For subsequent Mk analyses we gated out the CD41⁺CD11B⁺ population and focused on the CD41⁺CD11B⁻ cells.³⁸ This CD41⁺ CD11B⁻ Mk population in the tongue also coexpressed CD61, which along with CD41 comprises the GPIIb/IIIa integrin found on Mks and platelets. While the oral Mks were also largely positive for P-selectin (CD62P), this population expressed platelet glycoprotein V (CD42d), one of the subunits of the receptor for von Willebrand factor, to a lesser extent (Supplementary Fig. 7).³⁹ Along with cell surface markers, we also considered cell size in the flow cytometry gating strategy of oral Mks to eliminate platelets (Supplementary Fig. 3). To confirm the localization of Mks, we performed confocal microscopy to visualize Mks in the layers of tongue tissue. Mks were observed outside of the blood vessels of the tongue, near the basal stem cell layer, but predominantly in the sub-basal regions in WT sham and OPC mice (Fig. 6A). WT OPC mice on Day 2 of infection had more Mks (~0.56 per 1 mm² field of view) compared to WT Sham mice (~0.22 per 1 mm²) and PF4-iDTR OPC mice had no detectable Mks (0.00 per 1 mm²) (Fig. 6B), and these numbers were comparable to the number of cells detected by flow cytometry (Supplementary Fig. 8). Additionally, CD41⁺Draq5⁺ Mks from OPC-infected WT mice proliferated within the tongue tissue, shown by higher Ki67 expression compared to both sham tongue and bone marrow tissue (Figs. 6C and 6D). This suggested that tongue Mks proliferate within the tissue and do not accumulate from the bone marrow via circulation. Together, these data showed that Mks were resident within uninfected tongue tissue and expanded independently of the bone marrow during OPC, pointing to a role for Mks in fungal clearance in the oral mucosa.

Fig. 5.

The tongue contains megakaryocytes. (A) Representative flow cytometry plots of megakaryocyte populations (CD41⁺DRAQ5⁺). (B) Percentage of live cells that are megakaryocytes (CD41⁺DRAQ5⁺) in WT tongue tissue, PF4-iDTR tongue tissue, or WT bone marrow by flow cytometry. N = minimum four mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. (C) Representative flow cytometry images of megakaryocytes populations (CD41⁺DRAQ5⁺) in WT Sham and WT OPC mice on Day 2. (D) Quantification of megakaryocyte populations (CD41⁺DRAQ5⁺) in WT Sham and WT OPC mice on Days 1–4 (percentage of live cells). Results pooled from two experiments, n = minimum four mice per cohort per experiment, analyzed by Student’s t test between days. Mean ± SEM. (E) Percentage of bone marrow cells that are megakaryocytes (CD41⁺DRAQ5⁺) in sham and OPC mice. N = 3 mice per cohort, analyzed by Student’s t test. Mean ± SEM. *p > 0.05, **p > 0.01, ***p < 0.001, n.s. is not significant. ANOVA = analysis of variance; CD = cluster of differentiation; OPC = oropharyngeal candidiasis; SEM = standard error of the mean; WT = wild-type.

Fig. 6.

Megakaryocytes localize to the infected oral epithelium. (A) Visualization of megakaryocytes in the dorsal region of infected WT tongue tissue on Day 2 by confocal microscopy (top image), enlarged image of selected region from red box (middle image), and images of individual cells (bottom images), DAPI (blue) merged with CD41 (red). Papillae (P), area of infection (Inf), red blood cells (RBC), and basal stem cell layer (red line) labeled for reference. (B) Megakaryocyte quantification using confocal microscopy, shown as number per field of view between WT sham, WT OPC, and PF4-iDTR OPC mice on Day 2. N = 3 mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. (C) Histogram of Ki67 stained megakaryocytes (CD41⁺DRAQ5⁺) in Sham (dark gray) and OPC (light gray) tongue tissue and Sham bone marrow (black), n = four mice per cohort. (D) Quantification of mean fluorescent intensity of Ki67 expression of tongue and bone marrow megakaryocytes from (H), analyzed by one-way ANOVA. Mean ± SEM. **p < 0.01, ****p < 0.0001. ANOVA = analysis of variance; CD = cluster of differentiation; DAPI = 4’,6’-diamidino-2-phenylindole; Inf = infection; OPC = oropharyngeal candidiasis; P = papillae; RBC = red blood cell; SEM = standard error of the mean; WT = wild-type.

Oral megakaryocytes are distinct from bone marrow-derived megakaryocytes

Fully differentiated Mks have undergone a multistage developmental program that includes Mk progenitors identified by expression of the hematopoietic stem cell marker CD34 and Mk-specific molecules including CD41.⁴⁰ To further characterize the function of tongue Mks, we first sought to understand expansion of Mk progenitors during OPC. CD34⁺CD41⁺ progenitor cells were detected in the tongue tissue of sham-infected WT mice, but this population remained unchanged on Day 2 after infection, unlike Mks that expanded in OPC-infected WT mice (Figs. 7A and 5C, D). Next, we compared several immune surface markers differentially expressed between lung and bone marrow Mks. Through high expression of CCR7, lung Mks traffic to secondary lymphoid tissue where they express MHCII and the costimulatory molecule B7-2 (CD86) and act as antigen-presenting cells (APC).²² Tongue Mks showed higher levels of surface CCR7 (2082 mean fluorescence intensity (MFI)) and MHCII (3176 MFI) compared to bone marrow Mks (1617 and 2174, respectively) (Fig. 7B and Supplementary Fig. 2) indicating tongue Mks may be capable of acting in the same capacity as lung Mks to activate adaptive immune responses when necessary. While this potential APC function of Mks is likely not relevant during the acute form of OPC studied here, this showed that tongue Mks have increased expression of immune surface markers compared to bone marrow Mks, supporting the role of these cells in oral immunity.

Fig. 7.

Tongue megakaryocytes differ from bone marrow megakaryocytes. (A) Quantification of flow cytometry data showing percentage of megakaryocytes (CD41⁺ cells) that are also positive for CD34 in sham or infected tongue and bone marrow on Day 2 using gating strategy in Supplementary Figure 3. N = 3 mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. (B) Quantification of mean fluorescent intensity of CCR7, MHCII, and CD86 expression on tongue and bone marrow megakaryocytes. N = 3 mice per cohort, analyzed by Student’s t test. Mean ± SEM. **p < 0.01, ***p < 0.001, n.s. is not significant. ANOVA = analysis of variance; CCR7 = C-C chemokine receptor type 7; CD = cluster of differentiation; MHCII = major histocompatibility complex class II; SEM = standard error of the mean.

Megakaryocyte expansion and activation are interleukin-17-dependent

Since IL-17 plays a role in megakaryopoiesis and is also critical for effective host responses to C. albicans, we sought to establish if the cytokine was involved in the expansion or function of oral Mks during OPC.^26,28 First, analysis of the human Meg-01 cell line revealed the presence of both the IL-17RA and IL-17RC subunits of the receptor on the surface (Figs. 8A and 8B). To establish the functional outcome of receptor engagement, Meg-01 cells were cultured with IL-17A or the established Mk stimulator, thrombopoietin (TPO),⁴¹ followed by analysis of the supernatant for platelets. IL-17A stimulated release of platelets from Mks (8651 counts) similarly to TPO stimulation (8452 counts), while addition of α-IL-17RA antibody blocked platelet release (Fig. 8C). Next, we sought to establish if IL-17 deficiency affected the Mk populations as it did for circulating platelet levels (Fig. 1B). Bone marrow was harvested from naïve WT and Il17ra^−/− mice and the Mk population assessed. Il17ra^−/− mice had less bone marrow Mks (0.052%) compared to WT mice (0.098%) (Figs. 8D and 8E). The same trend was found in the tongue where Mk levels were higher in the WT compared Il17ra^−/− sham mice (0.11% and 0.061%, respectively) (Figs. 8E, 8F, and 8G). In addition, upon infection, the tongue Mk population failed to expand in Il17ra^−/− mice (0.056%) as it did in WT mice (0.2%) (Figs. 8F and 8G). IL-23 is required for expansion and function of innate and adaptive Type-17 lymphocytes during disease, including OPC.⁴² Mice deficient in the IL-23p19 subunit (Il23a^−/−), also had lower levels of tongue Mks during infection compared to WT mice (Supplementary Fig. 9). Taken together, these results showed a convergence of the Type-17/IL-17 signaling pathway with control of the Mk population in the oral mucosa in response to Candida.

Fig. 8.

Interleukin-17 signaling can drive megakaryocyte activity. (A) Histogram of IL-17RA and (B) IL-17RC receptor expression on human megakaryocytes (Meg-01 cell line). (C) Flow cytometry analysis of platelet (CD41⁺) counts in supernatant after a 4hr incubation of Meg-01’s with rIL-17A cytokine (100 ng/mL) compared to control supernatant or incubation with rTPO (10 ng/mL), rIL-17A+αIL-17RA antibody or IL-17A-isotype control antibody. Triplicate samples run over the course of 120 seconds, analyzed by one-way ANOVA. Mean ± SEM. Representative of three independent experiments. (D) Representative flow cytometry images of megakaryocyte populations (CD41⁺DRAQ5⁺) in naïve WT and Il17ra^−/− bone marrow. (E) Megakaryocyte populations (CD41⁺DRAQ5⁺) in bone marrow of naïve WT and Il17ra^−/− mice. N = 3 mice per cohort, analyzed by Student’s t test. Mean ± SEM. (F) Representative flow cytometry images of megakaryocyte populations (CD41⁺DRAQ5⁺) in WT and Il17ra^−/− OPC tongue tissue on Day 2. (G) Quantification of megakaryocyte populations (CD41⁺DRAQ5⁺) in tongue tissue of sham or infected WT or Il17ra^−/− mice on Day 2 (percentage of live cells). N = 3 mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. Representative of two independent experiments. *p < 0.05, **p < 0.01, ***p < 0.001, n.s. is not significant. ANOVA = analysis of variance; CD = cluster of differentiation; IL = interleukin; OPC = oropharyngeal candidiasis; rTPO = recombinant thrombopoietin; SEM = standard error of the mean; WT = wild-type.

Megakaryocytes kill Candida albicans through platelet release

While our data implicated Mks in the response to OPC, we next aimed to better understand how Mks interact with Candida. Since platelets have direct activity against C. albicans,^8,9 we determined whether Mks function similarly in vitro. Increasing numbers of human Mks (MEG-01) were incubated with both yeast and hyphal forms of C. albicans followed by colony enumeration on YPD agar plates. Mks killed Candida yeast at a concentration as low as 1:3 (MEG-01:Candida cells) but Mks did not kill Candida hyphae (Fig. 9A). MEG-01 cells exposed to C. albicans expressed a panel of immune mediators including antimicrobial peptides that typically are associated with killing of Candida (Supplementary Fig. 10).⁴³ We next determined whether Mk anti-Candida activity requires direct cell interaction with the fungus. MEG-01 and Candida yeast were incubated together using a transwell system. The average human platelet is 3–4 μm in diameter, while Mks are between 50–100 μm.⁴⁴ Therefore, we used an 8 μm filter for transwell assays to prevent Mks from interacting directly with Candida, but allowing released products, including platelets, to pass. A 0.4 μm transwell filter was used to prevent platelets from passing but allowed platelet microparticles and granules to interact directly with C. albicans cells. With the 0.4 μm filter, there was minimal killing of Candida. It was only when the 8 μm filter was used, and platelets could pass, that Candida was killed optimally (Fig. 9B). We then verified that Mks responded to the presence of C. albicans by releasing platelets. For this, MEG-01s were incubated with C. albicans for 1hr. Flow cytometry of the resulting supernatant showed a higher number of CD41⁺ DRAQ5^neg anucleate platelets (15816 counts) when C. albicans was present compared to the MEG-01-only control (7602 counts) (Figs. 9C and 9D). Next, Mks were incubated with Cytochalasin D (CytoD) to prevent platelet release.⁴⁵ CytoD-treated MEG-01 cells killed Candida less effectively (~16111 CFU/mL) compared to MEG-01 cells that released platelets (~8056 CFU/mL) (Fig. 9E). These data indicated that direct contact between Mks and C. albicans was not required for fungal killing, but platelet release from Mks was necessary.

Fig. 9.

Direct contact is not required for megakaryocyte activity against Candida albicans. (A) Killing potential (CFU/mL reduction) of varying numbers of megakaryocytes against the yeast and hyphal form of 1×10⁵/mL C. albicans after 1hr incubation at 37 °C. Performed in triplicate analyzed by one-way ANOVA. Mean ± SEM. Representative of at least three independent experiments. (B) Killing potential (CFU/mL reduction) of 1×10⁵/mL megakaryocytes against 1×10⁵/mL C. albicans after 1hr incubation at 37 °C incubation with or without 0.4mm or 8mm transwell filters. Performed in triplicate, analyzed by one-way ANOVA. Mean ± SEM. Representative of at least two independent experiments. (C) Representative flow cytometry plots of platelets (CD41⁺) in megakaryocyte-only or megakaryocyte + C. albicans supernatants. (D) Flow cytometry analysis of platelet (CD41⁺) counts in megakaryocyte-only control or megakaryocytes + C. albicans supernatants after 1hr incubation at 37 °C. Triplicate samples run over the course of 120 seconds, analyzed by Student’s t test. Mean ± SEM. Representative of at least two independent experiments. (E) Killing potential (CFU/mL reduction) of 1×10⁵/mL untreated megakaryocytes or cytochalasin D (50 μM)-treated megakaryocytes against C. albicans after 1hr incubation at 37 °C in. Performed in triplicate analyzed by one-way ANOVA. Mean ± SEM. Representative of two independent experiments *p < 0.05, **p < 0.01, n.s. is not significant. ANOVA = analysis of variance; CD = cluster of differentiation; CFU = colony forming unit; SEM = standard error of the mean.

Candidalysin stimulates platelet release from megakaryocytes

Since direct contact between Mks and C. albicans was not required for subsequent platelet-mediated fungal killing, we sought to identify a secreted component of C. albicans that stimulated platelet release. Notably, C. albicans produces a peptide toxin, candidalysin, that is essential for inducing protective immune responses during OPC.^46–48 Since candidalysin activates platelets²⁹ we stimulated Mks with either synthetic candidalysin or lipopolysaccharide (LPS) as a non-C. albicans-derived control that stimulates Mks in other inflammatory conditions.²² Mks treated with candidalysin released the same number of activated platelets (13792 counts) as those incubated with C. albicans (14538 counts), and both treatments released higher numbers of CD41⁺ DRAQ5^neg platelets compared to untreated MEG-01-only controls and LPS-only treated MEG-01’s (7969 and 7410 counts, respectively) Also, LPS and candidalysin treatment together did not have an additive effect (14532 counts) (Figs. 10A and 10B). The supernatants were also collected to assess the Candida-killing potential of platelets and other Mk releasates. C. albicans incubated with supernatants from candidalysin-stimulated MEG-01 cells or from C. albicans-stimulated MEG-01 cells killed Candida better (~52000 CFU/mL and ~32000 CFU/mL, respectively) than control supernatant from unstimulated MEG-01 cells (~105600 CFU/mL) (Fig. 10C). To confirm that candidalysin induced Mk activity, we used a C. albicans mutant strain that does not secrete candidalysin (ece1ΔΔ) and an ece1ΔΔ-ECE1 revertant control strain, which has one copy of the ECE1 gene recomplemented, thus restoring candidalysin secretion.⁴⁹ When Mks were incubated with ece1ΔΔ in a transwell assay, little reduction in C. albicans CFU/mL was observed (~53800 CFU/mL) compared to the ece1ΔΔ-only control (~67000 CFU/mL), whereas incubating MEG-01’s with the ece1ΔΔ-ECE1 revertant strain resulted in 50% killing of C. albicans (~31400 CFU/mL) (Fig. 10D). When synthetic candidalysin was added to the ece1ΔΔ:MEG-01 transwell system, C. albicans killing was restored to levels similar (~34900 CFU/mL) to the ece1ΔΔ-ECE1 revertant strain:MEG-01 (Fig. 9D). We also verified that when Mks were incubated with the ece1ΔΔ-ECE1 revertant strain, similar C. albicans killing was observed as MEG-01 cells incubated with WT (SC5314) C. albicans (Supplementary Fig. 11). Next, when Mks were incubated with the ece1ΔΔ strain, a smaller population of activated platelets were released (7348 counts) compared to MEG-01 cells incubated with ece1ΔΔ + candidalysin and with the ece1ΔΔ-ECE1 revertant strain (17102 & 16838 counts respectively) (Fig. 10E). These data confirm that candidalysin secretion was required for platelet release. In addition, there appeared to be no additive effect of IL-17 and candidalysin on platelet release since MEG-01 cells incubated with both released equivalent numbers of platelets compared to independent stimulation by either (Supplementary Fig. 12). Since candidalysin interacts with CD42b (GP1bα) one of the receptor subunits for von Willebrand factor on platelets, we sought to determine if this was the receptor candidalysin interacted with on Mks.²⁹ To test this, we targeted the receptor on Mks with antibodies specific for CD42b before candidalysin incubation. Compared to Mk-only control (6746 counts), blocking CD42b resulted in similar platelet release by Mks (6874 counts), whereas Mks without antibody treatment and Mks treated with isotype control had increased platelet release (13983 and 13750 counts, respectively) (Fig. 10F). Taken together, these results showed that candidalysin induced Mks through activation of the CD42b receptor to release platelets capable of killing C. albicans.

Fig. 10.

Candidalysin drives megakaryocyte activation. (A) Representative flow cytometry plots of platelets (CD41⁺) collected from the releasates of megakaryocytes incubated with or without candidalysin (Clys). (B) Flow cytometry analysis of platelet (CD41⁺) counts in releasates from megakaryocyte-only control, megakaryocytes + C. albicans, megakaryocytes + Clys peptide (15 μM), megakaryocytes + LPS (5 μg/mL), or megakaryocytes + Clys peptide + LPS incubated for 1hr at 37 °C. Triplicate samples run over the course of 120 seconds, analyzed by one-way ANOVA. Mean ± SEM. (C) Killing potential (CFU/mL reduction) against C. albicans of releasates collected from megakaryocyte-only control, megakaryocyte + C. albicans, or megakaryocytes + Clys peptide (15 μM) incubated for 1hr at 37 °C. Performed in triplicate, analyzed by one-way ANOVA. Mean ± SEM. (D) Killing potential of megakaryocytes against Clys-deficient C. albicans (ece1ΔΔ) +/− the addition of Clys peptide (15 μM) or the ece1ΔΔ-ECE1 strain after incubation for 1hr at 37 °C. Performed in triplicate, analyzed by one-way ANOVA. Mean ± SEM. Representative of at least two independent experiments. (E) Flow cytometry analysis of platelets (CD41⁺) counts in releasates from megakaryocyte-only control, megakaryocytes + ece1ΔΔ, megakaryocytes + ece1ΔΔ-ECE1, or megakaryocytes + ece1ΔΔ + Clys peptide (15 μM) after 1hr incubation at 37 °C. Triplicate samples run over the course of 120 seconds, analyzed by one-way ANOVA. Mean ± SEM. (F) Flow cytometry analysis of platelets (CD41⁺) count in releasates from megakaryocyte-only control, megakaryocytes + Clys peptide (15μM), megakaryocytes + IgG Isotype control + Clys, or megakaryocytes + αCD42b (1μg/mL) + Clys after 1hr incubation at 37 °C. Triplicate samples run over the course of 120 seconds, analyzed by one-way ANOVA. Mean ± SEM. For A-D 1×10⁵/mL megakaryocytes were incubated with 1×10⁵/mL C. albicans where applicable. *p < 0.01, **p < 0.01, ***p < 0.001, ****p < 0.0001, n.s. is not significant. ANOVA = analysis of variance; CD = cluster of differentiation; CFU = colony forming unit; Clys = candidalysin; OPC = oropharyngeal candidiasis; SEM = standard error of the mean.

Megakaryocyte expansion in the tongue tissue requires candidalysin expression during oropharyngeal candidiasis

Since candidalysin was necessary for the Mk response to Candida in vitro, we next considered if candidalysin was responsible for Mk expansion during OPC. WT mice were infected with either ece1ΔΔ or ece1ΔΔ-ECE1 revertant C. albicans, followed by tongue tissue harvest 2 days post-infection. First, we confirmed that mice infected with either strain presented with similar fungal burden on Day 2, which allowed us to compare the Mk populations between groups since the levels of Candida present on the tongue were not significantly different across mouse cohorts as shown by others in immunocompetent WT mice⁴⁷ (Fig. 11A). In mice infected with the ece1ΔΔ-ECE1 revertant strain or SC5314 the Mk population expanded (0.19% and 0.18% of total tongue cells, respectively) more compared to sham mice (0.092%) and ece1ΔΔ infected mice (0.086%) (Figs. 11B and C). Lastly, we examined if the clinical commensal C. albicans strain, 529L, which produces very low levels of candidalysin in vivo⁵⁰ would also fail to stimulate Mk expansion during OPC. Like sham-infected mice (0.076%), mice infected with C. albicans strain 529L had minimal Mk population expansion (0.068%), compared to mice infected with C. albicans strain SC5314 (0.16%) even though fungal burden with 529L was higher (Figs. 11D, 11E, and 11F). Together, these data demonstrated that candidalysin secretion induced the expansion of Mks in tongue tissue during OPC.

Fig. 11.

Candidalysin drives megakaryocyte expansion in murine tongue tissue during oropharyngeal candidiasis. (A) Fungal burden (CFU/g) of ece1ΔΔ or ece1ΔΔ-ECE1 revertant strain OPC murine tongue tissue on Day 2 of OPC. (B) Representative flow cytometry plots of megakaryocytes (CD41⁺DRAQ5⁺) in murine tongue tissue on Day 2 of OPC using ece1ΔΔ or ece1ΔΔ-ECE1 revertant strain. (C) Quantification of megakaryocyte populations (CD41⁺DRAQ5⁺) in sham-, SC5314 OPC-, ece1ΔΔ OPC-, or ece1ΔΔ-ECE1 OPC-infected tongue tissue on Day 2 (percentage of live cells). Results pooled from two experiments, n = minimum of three mice per cohort per experiment, analyzed by Student’s t test. Mean ± SEM. (D) Fungal burden (CFU/g) of SC5314 or 529L infected OPC murine tongue tissue on Day 2 of OPC. (E) Representative flow cytometry plots of megakaryocytes (CD41⁺DRAQ5⁺) in murine tongue tissue on Day 2 of OPC using C. albicans strains SC5314 and 529L. (F) Quantification of megakaryocyte populations (CD41⁺DRAQ5⁺) in sham, SC5314 OPC, or 529L OPC tongue tissue on Day 2 (percentage of live cells). n = 3 mice per cohort, analyzed by one-way ANOVA. Mean ± SEM. **p < 0.01, n.s. is not significant. ANOVA = analysis of variance; CFU = colony forming unit; OPC = oropharyngeal candidiasis; SEM = standard error of the mean.

DISCUSSION AND CONCLUSION

Our understanding of how Mks and platelets contribute to immunity is expanding. The discovery of Mk populations outside of the bone marrow that function in host defense mechanisms contrasts with the notion that the singular role of Mks is to release platelets for blood clotting. For example, the lungs are a major site of platelet biogenesis derived from mature Mks found within extravascular spaces.²⁵ During thrombocytopenic conditions, immature Mks and Mk progenitor cells in the lung can migrate and replenish bone marrow Mk populations.²⁵ In the context of immune function, lung Mks are distinct from bone marrow Mks.²² Lung Mks not only express higher levels of APC-related molecules, specifically MHCII and CD11c, but also produce higher levels of inflammatory molecules, including IL-1α, CXCL2, CCL5.²² This suggests that the immune phenotypes of Mks, regardless of their location in the bone marrow or lung, are plastic and their function is dictated by the inflammatory status of the tissue environment.^22,51,52

While the role of Mks in systemic bacterial infection is better understood,²¹ the cells have been linked to the host response against fungi including Candida and Aspergillus species.^4,23 With regard to fungal infections, Mk progenitor cells enhance anti-Candida activity of peripheral blood mononuclear cells and mediate the increased expression of IL-17 by CD4⁺ cell populations when C. albicans is present.²⁷ Related to platelets in fungal immunity, these cells control the overgrowth of the pathogenic fungus A. fumigatus in the lungs.⁴ Prior to this study, there was also evidence that platelets were involved in host responses to Candida. Platelets, isolated from human and murine blood, incubated with C. albicans led to reduced fungal growth.⁸ Platelet activity against C. albicans was dependent on platelet stimulation, as blocking granular release disrupted C. albicans killing.⁸ The antifungal capacity and tissue-specific role of both Mks and platelets predicted there was a contribution from these cells during OPC.

The contribution of platelets to neutrophil recruitment in other infections and inflammatory conditions also suggested there could be a convergence of platelet-mediated and IL-17-mediated immunity in OPC to potentiate a robust neutrophil response. In healthy humans and mice, IL-17/IL-17RA signaling leads to upregulation of chemokines essential for neutrophil recruitment.²⁶ However, the neutrophil response is not entirely IL-17-dependent, and, as such, other mediators are involved.³⁶ When using the PF4-iDTR model, platelet-depleted mice had higher tongue tissue fungal burden compared to WT mice. This aligned with both the lower influx of neutrophils to tongue tissue and poor neutrophil activation when platelets were absent. Additionally, chemotaxis of the responding neutrophils from the sub-basal to the supra-basal epithelial layer was impeded during platelet depletion presumably due, at least in part, to the decreased expression of neutrophil chemokines and growth factors. Importantly, direct interaction between platelets and neutrophils was necessary through PSGL-1 and P-selectin. The compromised neutrophil response at early infection time points in platelet-depleted mice likely contributed to increased fungal burden and susceptibility to OPC since replenishment of platelets resulted in neutrophils trafficking to the oral mucosa and clearance of C. albicans. This time frame aligns with the neutrophil response in healthy mice and the impaired neutrophil response in Il17ra^−/− mice within 24 hours after infection. Since Il17ra^−/− mice lacked a robust Mk population, this further aligns IL-17 with the neutrophil response via Mk/platelet function.

Mks did not require direct contact with C. albicans to become activated (Fig. 8B). Candidalysin, which is secreted by C. albicans hyphae, was responsible for Mk proliferation and function. Candidalysin is a cytolytic toxin associated with virulence, epithelial damage, and immune activation during oral C. albicans infection.⁵³ Candidalysin induces the release of several epithelial antimicrobial peptides and alarmins,⁵⁴ along with chemokine release leading to neutrophil recruitment.⁵³ While candidalysin has been implicated in platelet function,²⁹ this is the first time a role has been ascribed to this toxin in Mk biology. We speculate that hyphal formation and tissue invasion in proximity to the location of oral Mks in the sub-basal regions of the tongue is first required and then Mks expand and are activated to release platelets. These activated platelets then participate in killing the yeast form before the transition to more hyphae occurs. As such, commensal strains of C. albicans including 529L, which form hyphae but secrete low levels of candidalysin, may not induce localized expansion of tongue Mks and platelet production. This would allow commensal C. albicans strains to colonize for extended periods of time with minimal immunopathology.

It is intriguing that candidalysin and IL-17 activated Mks and platelets in the absence of additional thrombotic stimuli in cell culture. Mk differentiation and activation are traditionally associated with the availability of TPO, soluble agonists such as adenosine diphosphate (ADP), and platelet-secreted products released during hemostasis.^41,55 Our data adds another mechanism for Mk/platelet activation during mucosal infection and lends to the notion that Mks and platelets are diverse immune cells that can differentially respond to stimuli. While relatively rare, Mk numbers in the tongue tissue were higher than tissue-localized macrophages, supporting the importance of Mk in oral immune responses. Further study will elucidate how Mk intermediate cell types also contribute to fungal immunity.

The continuous circulation of platelets allows these cells to react broadly to both the pathogen and the vascular damage related to pathogenesis.¹ We now identify Mks in the oral mucosa as another local source of platelets during infection. A more in-depth study of the contribution of oral Mks, like those in the lung, to systemic replenishment of platelets will increase our knowledge of tissue-specific immune responses. Platelet activation in other tissue compartments contributes to the recruitment and function of key immune cells such as neutrophils and macrophages.^17,56,57 As such, a better understanding of how platelets and fungi interact to enhance the function of other immune cells is critical, especially since thrombocytopenia is a risk factor for serious fungal infections.¹³ Although, most patients with platelet insufficiency are susceptible to disseminated forms of candidiasis, not necessarily mucosal OPC. Intriguingly, our findings suggest oral Mks could be a local source of platelets that protect against localized mucosal infection even when there is a systemic platelet insufficiency. While this is speculative, our findings do indicate that further studies of oral Mks in humans are critical and may elucidate the importance of these cells in other oral infections and inflammatory conditions.

MATERIALS AND METHODS

Mice

To generate PF4-iDTR mice, C57BL/6 mice transgenic for simian iDTR containing a LoxP-flanked stop codon (Jackson Laboratories, Bar Harbor, ME strain #007900) were crossed with C57BL/6 expressing Cre recombinase under control of the PF4 promoter (Jackson Laboratories, strain #008535) as previously described.¹⁷ Polymerase chain reaction (PCR) based genotyping of ear punches was used to select Cre-expressing mice homozygous or heterozygous for iDTR. Il23a−/− mice were obtained from Genentech (San Francisco, CA) through a materials transfer agreement and Il17ra^−/− mice from Amgen (Thousand Oaks, CA).⁵⁸ Male and female mice aged 6–12 weeks were used for all experiments. For platelet depletion, an initial dosage of 400 ng/mouse DT (diphtheria toxin) (Sigma Aldrich, St. Louis, MO) was given followed by 200 ng/mouse DT toxin every 48 h until completion of experiment. Platelet depletion was verified after animal sacrifice via cardiac puncture to collect blood, followed by staining with CD41-APC (MWReg30) (BioLegend, San Diego, CA) and analysis on an LSRFortessa flow cytometer (BD Biosciences, Franklin Lakes, NJ). All mice were housed with food and water ad libitum under a 12 h dark/light cycle in a specific pathogen-free facility at the University of Toledo. All animal studies were approved by the University of Toledo IACUC.

Murine model of oropharyngeal candidiasis

Mice were challenged with either Candida albicans SC5314, 529L, ece1ΔΔ, or ece1ΔΔ-ECE1 revertant using a previously described model of murine OPC.³³ Briefly, Candida was cultured overnight in YPD broth from which a working solution of 1×10⁷-yeast/mL was prepared. Pre-weighed cotton balls soaked in either a phosphate-buffered solution (PBS) control or Candida solution were placed under the tongues of anesthetized mice for 75 min. Fungal burden analysis was performed by sacrificing mice and then harvesting tongues for mechanical homogenization on a GentleMACS Dissociator (Miltenyi Biotec, Auburn, CA) before plating on YPD agar for colony enumeration. For platelet-depletion studies, OPC infection began 7 days after initial DT treatments. Depletion was verified by post-mortem collection of blood via cardiac puncture and flow cytometry analysis of platelet levels. For PSGL-1 neutralization studies, mice received α-PSGL-1 neutralizing antibody or isotype control (4 mg/kg; BioXCell, Lebanon, NH) on Day −2,−1, and Day 0 of OPC infection.

Candida in vitro assays

C. albicans strain SC5314, ece1ΔΔ, and ece1ΔΔ-ECE1 revertant were grown overnight at 30 °C in YPD broth, followed by subculturing in fresh YPD broth (10% FBS was added and temperature increased to 37 °C for experiments requiring hyphal Candida) the next morning until reaching an optical density between 0.8 and 0.9. After centrifugation, cells were reconstituted in PBS and enumerated using a hemocytometer then diluted to proper working dilution. MEG-01 cells were centrifuged, reconstituted in PBS, enumerated, and diluted to proper working dilutions. 1 × 10⁵/mL of C. albicans and 1 × 10⁵/mL to 1 × 10²/mL of Mks were incubated together for 1 h at 37 °C in PBS, followed by serial dilution and plating on YPD agar for colony enumeration or flow cytometric analysis for Mk platelet release. For non-contact experiments, transwell assays with a pore size of 0.4 μm and 8 μm were utilized. Candidalysin peptide (15 μM) (Vivitide LLC, Gardner, MA),⁵³ IL-17A (100 ng/mL) (R&D Systems, Minneapolis, MN), Thrombopoetin (TPO) (10 ng/mL) (R&D Systems), LPS (5 μg/mL) (Sigma Aldrich), Cytochalasin D (50 μM) (Sigma Aldrich), and αCD42b antibody (5 μg/mL) (BioLegend) were also used when necessary to assess platelet release by Mks.

Real-time quantitative polymerase chain reaction

Human megakaryocyte MEG-01 (ATCC, Manassas, VA) cell line was cultured in RPMI 1640 supplemented with 10% FBS, 100 IU/ml penicillin, and 100 mg/ml streptomycin at 37 °C in a 5% CO₂ enriched environment. MEG-01 cells were normalized to 1 × 10⁶ and incubated with C. albicans at MOI 1 or TPO at 100 ng/ml for 3 hours at 37 °C. Cells were then centrifuged at 200 g to remove Mks, supernatant collected, and centrifuged at 450 g to collect platelets. Total RNA was extracted using TRI reagent (Sigma Aldrich, St. Louis, MO, USA) and RNA was reverse-transcribed by High-Capacity cDNA RT kit (Thermo Fisher Scientific, Waltham, MA, USA) at 25 °C for 10 min, 37 °C for 120 min, followed by 85 °C for 5 min. Quantitative PCR was performed using Luna Universal qPCR Mastermix (New England Biolabs, Ipswich, MA) and a Quant Studio 3 detection system (Applied Biosystems, Waltham, MA, USA), as specified by the manufacturer. The crossing point was defined as the maximum of the second derivative from the fluorescence curve. For quantification, relative mRNA expression of specific genes using the 2–ΔCT method and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) housekeeping gene for normalization was performed. Primers for genes DEFB4A, CCL20, PF4, S100A8, S100A9, CCL5, RANTES, and DEFB103A were purchased from an existing Prime-Time qPCR database (Integrated DNA Technologies, Coralville, IA). Assays were performed in biological triplicate in technical triplicate.

Flow cytometry

For tongue analysis, mice were perfused with sterile saline post-sacrifice and tissue was harvested and mechanically homogenized in RPMI 1640 media and then incubated at 37 °C for 42 min on a GentleMACS Dissociator (Miltenyi Biotec) using Tissue Dissociation Kits (Miltentyi Biotec) then passed through a 40 mm cell strainer (for neutrophil analysis) or a 70 mm strainer (for megakaryocyte (Mk) analysis) to form single-cell suspensions. After brief centrifugation, cells were reconstituted with PBS supplemented with 2% FBS and 2 mM EDTA. 1 × 10⁶/mL viable cells were obtained by staining with Trypan blue and counting on a hemocytometer. An initial incubation of CD16/CD32 Fc Block (BD Biosciences) was followed by staining with the following antibodies, all from BioLegend: CD45-PE/Cyanine7 (I3/2.3), CD11b-PerCP/Cyanine5.5 (M1/70), and GR-1-APC (RB6-8C5) for the neutrophil staining panel and CD41-Brillant Violet 421 (MWReg30), MHCII-FITC (M5/114.15.2), CCR7-PE (4B12), and CD86-Brillant Violet 510 (GL-1), and DRAQ5 for the Mk staining panel.

For bone marrow analysis, femurs were harvested, cleared of remaining tissue, both ends cut off, and then flushed with PBS. The resulting solution was passed through a 70 mm cell strainer to form a single-cell suspension. After a brief centrifugation and reconstitution, cells were stained with the following antibodies, all from BioLegend; CD41-Brillant Violet 421 (MWReg30), MHCII-FITC (M5/114.15.2), CCR7-PE (4B12), CD86-Brillant Violet 510 (GL-1), and DRAQ5. For in vitro platelet analysis of MEG-01 cells, supernatant was collected, passed through a 40 mm cell strainer, briefly centrifuged, and then stained with CD41-FITC (HIP8) from BioLegend. Additional flow cytometry analysis (supplemental data) included the following antibodies, all from BioLegend: CD11b-Brillant Violet 510, CD42D-PerCP/Cyanine5.5, CD62P-APC, GR-1-AlexaFluor700, CD61-PE/Cyanine7, and F4/80-PE. All flow cytometry was performed on an LSRFortessa (BD Biosciences).

Flow cytometry analysis

For in vitro platelet-megakaryocyte experiments, an initial gate was made on the SSC-A versus FSC-A plot of total cells to eliminate cellular debris and platelet microparticles from platelets. Location of platelets on the plot was based on gating of purified platelets as a control. Secondary gating was performed using DRAQ5 (to differentiate anucelate platelets from nucleated Mks), CD41. To detect Mks in murine tongue tissue, doublets were first removed using a gate on the FSC-A versus FSC-H plot, followed by a general live cell gate to remove cellular debris. Within the live cell gate, Mks were defined as nucleated cells (DRAQ5⁺) positive for CD41. This population was also used as the parent gate for subsequent Mk phenotyping. For Mks in murine bone marrow, nucleated cells (DRAQ5⁺) positive for CD41 were gated on from the total cell population. To detect neutrophil populations in murine tongue tissue during OPC, doublets were removed, followed by a general leukocyte gate using SSC-A versus FSC-A to eliminate cellular debris. Neutrophils were classified into two populations using the GR-1 marker: GR-1^high and GR-1^{intermediate(Int)}, followed by analysis of CD11b expression (percentage or MFI).

Analyses of flow cytometry results were performed using FlowJo (BD Bioscience). Gating strategy can be found in Supplemental Figs. 4, 5, and 6.

Histology

Tissues were fixed in formalin, paraffin-embedded, and sectioned at 5 μm by the University of Toledo Integrated Core Facilities. Neutrophils were observed and enumerated in H&E-stained tissue using an EVOS FLc Microscope (Life Technologies, Carlsbad, CA). For immunofluorescence images, Mks were observed in tongue tissue sections stained overnight at 4 °C with rat anti-CD41 (R&D Systems, Minneapolis, MN) then 1 hour with anti-rat AlexaFluor555 antibody (BioLegend) followed by 4’,6-dia midino-2-phenylindole (DAPI) (Sigma Aldrich) counterstaining. Mounted slides were analyzed on a Stellaris 5 laser scanning confocal equipped with an HC PL APO 63x/1.40 OIL CS2 objective, Power HyD Detectors, and LAS X software (Leica Microsystems, Wetzlar, Germany). Tiling was carried out within LAS X and TauGating was used to separate unwanted background fluorescence.

Cell line

Human megakaryocyte MEG-01 (ATCC, Manassas, VA) cell line was cultured in RPMI 1640 supplemented with 10% FBS, 100 IU/ml penicillin, and 100 mg/ml streptomycin at 37 °C in a 5% CO₂ enriched environment.

Data analysis

All experiments included at least three biological replicates per group and each experiment was repeated at least twice. Normally distributed data was analyzed via analysis of variance with Tukey’s post hoc analysis or Student’s t test. Nonparametric data was analyzed by Mann-Whitney using GraphPad Prism (V8.4.3). All p values < 0.05 were considered significant.