Abstract
Secondary fungal infections represent a major complication following thermal injuries. However, the mechanisms of fungal colonization of burn tissue and how the host subsequently responds to fungi within this niche remain unclear. We have previously reported a zebrafish model of thermal injury that recapitulates many of the features of human burn wounds. Here, we characterize host-fungal interaction dynamics within the burn wound niche using two of the most common fungal pathogens isolated from burn injuries, Aspergillus fumigatus and Candida albicans. Both A. fumigatus and C. albicans colonize burned tissue in zebrafish larvae and induce a largely conserved innate immune response following colonization. Using drug-induced cell depletion strategies and transgenic zebrafish lines with impaired innate immune function, we found that macrophages control fungal burden while neutrophils primarily control invasive hyphal growth at the early stages of infection. However, we also found that loss of either immune cell can be compensated by the other at the later stages of infection, and that fish with both macrophage and neutrophil deficiencies show more invasive hyphal growth that is sustained throughout the infection process, suggesting redundancy in their antifungal activities. Finally, we demonstrate that C. albicans strains with increased β(1,3)-glucan exposure are cleared faster from the burn wound, demonstrating a need for shielding this immunogenic cell wall epitope for successful fungal colonization of burn tissue. Together, our findings support the use of zebrafish larvae as a model to study host-fungal interaction dynamics within burn wounds.
Importance Secondary fungal infections within burn wound injuries are a significant problem that delay wound healing and increase the risk of patient mortality. Currently, little is known about how fungi colonize and infect burn tissue or how the host responds to pathogen presence. In this report, we expand upon an existing thermal injury model using zebrafish larvae to begin to elucidate both the host immune response to fungal burn colonization and fungal mechanisms for persistence within burn tissue. We found that both Aspergillus fumigatus and Candida albicans, common fungal burn wound isolates, successfully colonize burn tissue and are effectively cleared in immunocompetent zebrafish by both macrophages and neutrophils. We also find that C. albicans mutants harboring mutations that impact their ability to evade host immune system recognition are cleared more readily from burn tissue. Collectively, our work highlights the efficacy of using zebrafish to study host-fungal interaction dynamics within burn wounds.
Introduction
The World Health Organization (WHO) estimates that around 11 million burn injuries per year occur worldwide, with almost 200,000 resulting in patient death (1). During burn injury, the skin barrier and microbiome are disrupted, and subsequent antimicrobial treatment and immune system dysregulation make these wounds susceptible to microbial infections (2). Burn wound infections (BWIs) are the leading cause of mortality among burn wound patients and add an additional layer of complexity that can delay the healing process and further alter the immune response (3–5). BWIs are primarily caused by bacterial species like Pseudomonas aeruginosa and Staphylococcus aureus (3, 6). However, fungal pathogens are also a significant cause of infection at the burn wound interface. Indeed, global reports have shown that up to 44% of severe BWIs develop secondary infections from fungal species (5). Once colonized, the risk for the development of systemic infections increases substantially, with mortality rates reported to be as high as 76% once dissemination occurs (7–9). Despite this high prevalence, little is known about how fungi successfully colonize the burn wound tissue to cause disease, nor how the host subsequently responds to invading fungi at this niche. Indeed, most of the available data is based on clinical studies, highlighting the need for a robust infection model to further elucidate these dynamics.
Although fungal BWIs are understudied, in vitro and in vivo models do exist that have provided initial insights into the host response and therapeutic strategies following disease onset. For example, BWIs in human tissue explants have been recently established as an ex vivo model to study the host response to Candida albicans and have elegantly shown a role for neutrophils in responding to both burn damage and fungi (8). Animal burn models using mice, rats, pigs and galleria also have been utilized to study drug efficacy following fungal burn wound colonization (10–12). However, these models have yet to be leveraged to extensively study the mechanisms of fungal colonization and host immune response, and many do not easily permit in vivo imaging of host-fungal interactions to better characterize disease dynamics. Zebrafish larvae, which are naturally transparent, provide an advantage for imaging infection kinetics in vivo (13–15). These vertebrates have a largely conserved innate immune response with humans, and zebrafish larvae have been established as a reliable model to study the inflammatory profile and tissue remodeling processes following burn injury (16, 17). In addition, we have shown that the inflammatory profile of burn wounds can be altered by P. aeruginosa colonization and have recently expanded upon the zebrafish burn model to show that thermal injuries can be colonized by the fungal pathogen C. albicans (18, 19). However, the mechanisms of fungal colonization and host mediated clearance remain unknown.
Here, we characterize a fungal BWI model in zebrafish using two of the most common fungal pathogens isolated from burn wounds, C. albicans and the filamentous fungus Aspergillus fumigatus (2, 6, 7, 9, 20). We demonstrate that both colonize the burn wound and are successfully cleared in an immunocompetent host. We found that macrophages and neutrophils play synergistic and partially redundant roles in controlling fungal burden and hyphal formation, and that infection in zebrafish with both deficient macrophages and neutrophils results in extensive colonization and invasive hyphal growth from the site of injury. We also show that fungal mutants with increased β(1,3)-glucan exposure are cleared faster from the burn site, demonstrating a need for shielding immunogenic epitopes in the fungal cell wall for persistent colonization to occur. Collectively, our data demonstrate conserved immune responses against two distinct fungal pathogens in the burn wound and highlight the potential of zebrafish larvae as a model to study fungal BWIs.
Results
Pathogenic fungi colonize burn wounds in immunocompetent zebrafish larvae
We have recently expanded on a preexisting burn wound injury model using 3 days post-fertilization (dpf) larvae to show that C. albicans can successfully colonize burn tissue in this model (19). However, characterization of the infection process and the subsequent host response has yet to be assessed. Furthermore, it is currently unknown if additional pathogenic fungi commonly isolated from BWIs are capable of infecting damaged tissue in a larval zebrafish burn model. To address this, we induced burn wound injury in 3 dpf zebrafish larvae via cauterization and incubated fish with either fluorescent labelled spores of Aspergillus fumigatus or Candida albicans yeast to permit injury colonization (Fig. 1A). Both confocal microscopy images and plating for colony forming units (CFUs) showed that A. fumigatus was able to successfully colonize the burn wound, with peak fungal burden at the 24-hour time point followed by dramatic fungal clearance by 72 hours post burn (hpb) (Fig. 1B-E). Importantly, the infection was restricted to the burn wound area by 24 hpb, highlighting the need for tissue disruption for colonization to occur (Supplemental Fig 1D). In addition to colonization, germinated spores and/or hyphae were seen across all time points but was limited to small hyphal extensions localized to the tail fin (Fig. 1D). It should be noted that at time points prior to 24 hpb there is an abundance of fungi superficially adhered to necrotic tissue of the burn wound, but by 24 hpb fungal cells are interstitial within viable tissue. Therefore, we considered the 24 hpb time point as the start of stable colonization of the burn wound for further analysis.
3-dpf wild-type larvae were injured and infected with either A. fumigatus CEA10 (RFP) or C. albicans (RFP). Control fish were uninfected. (A) Representative brightfield images of larvae infected with A. fumigatus CEA10 or uninfected at 48hpb. (B) 2D-tail fin area at 24-72 hpb of larvae uninfected or infected with A. fumigatus CEA10. (C) 2D-tail fin area at 24-72 hpb of larvae uninfected or infected with C. albicans. (D) Representative tile image of entire wild-type larvae infected with A. fumigatus CEA10 (RFP) (scale bar 200 µm). Bars on graph presented as mean+s.d. Each dot represents data from an individual fish. Results represent data pooled from 3 independent experiments. n= 24-30 larvae per condition. p values calculated by ANOVA with Tukey’s multiple comparisons
(A) Fungal burn wound infection schematic. The tail fins of 3 days post-fertilization (dpf) larvae are injured using a cauterizer and immediately transferred to E3 -MB media with 2 x 107 fungal cells/ml. (B) Representative images of wild-type larvae infected with fluorescent (RFP) A. fumigatus CEA10 and imaged at 2, 6, 24, 48 and 72-hours post-burn (hpb). Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail fin edge. (C) 2D-fungal (RFP) area from 24-72 hpb. (D) Percent of larvae with germinated spores from 24-72 hpb. Each dot represents percent of an individual experiment. (E) Larvae infected with A. fumigatus CEA10 were homogenized and plated for CFUs at 24-72 hpb. (F) Representative images of wild-type larvae infected with fluorescent (RFP) C. albicans and imaged at 2, 6, 24, 48 and 72-hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail-fin edge. (G) 2D-fungal (RFP) area and (H) CFUs/fish from 24-72 hpb. (I-J) 2D-fungal area at 24 hpb from wild-type larvae innoculated at 3dpf with either live or paraformaldehyde (PFA)--fixed (I) A. fumigatus CEA10 spores or C. albicans yeast cells (RFP). Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by ANOVA with Tukey’s multiple comparisons. *p<0.05, ***p<0.001, ****p<0001.
The ability to colonize the burn wound was conserved by C. albicans (Fig. 1F-H). Strikingly, C. albicans seemed to persist longer, with higher fungal burden compared to A. fumigatus, especially at 72 hpb (Fig 1G-H). Similar to A. fumigatus, morphological transition to hyphae was seen during C. albicans infection, but the kinetics of these transitions differed. A. fumigatus germination was seen starting at 24 hpb, but C. albicans hyphae was observed as early as 2-6 hpb and was largely controlled by 24 hpb (Fig. 1F). This is consistent with previously published work in which C. albicans germination starts within 6 hours-post infection (hpi) in a hindbrain ventricle zebrafish infection model, but by 24 hpi the host immune system effectively controls hyphal growth (21).
Fungal colonization increases immune cell recruitment to burns
In both humans and fish, macrophages and neutrophils are recruited to the burn wound and function to clear cellular debris from damaged and necrotic tissue (16, 18). To assess if fungal BWI further impacted immune cell recruitment we burned and infected larvae with fluorescently labeled macrophages (mpeg-GFP) and neutrophils (lyz-BFP) and quantified leukocyte recruitment. By 24 hpb, macrophages and neutrophils are recruited to the burn wound during both A. fumigatus (Fig. 2A-C) and C. albicans (Fig. 2D-F) infection, and their numbers drop over a 72-hour period in accordance with wound healing and fungal clearance. Like the inflammatory profile of the burn wound at basal levels (16), macrophages were the predominant cell type present during both fungal infections (Fig. 2B&E). Interestingly, at 24 hpb fish infected with C. albicans showed increased levels of both macrophages and neutrophils at the tail fin when compared to uninfected burned fish (Fig. 2E-F), but only neutrophil levels were increased at 24 hpb during A. fumigatus infection (Fig. 2C). However, neither C. albicans nor A. fumigatus significantly impaired tail fin regeneration compared to the uninfected control following burn (Supplemental Fig. 1B-C).
(A-C) 3-days post fertilization wild-type larvae with fluorescent macrophages (mpeg-GFP) and neutrophils (lyz-BFP) were burned and infected with fluorescent A. fumigatus CEA10 (RFP). Control larvae were injured but uninfected. Larvae were imaged at 24-, 48- and 72-hpb. (A) Representative images of larvae infected with A. fumigatus CEA10 at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail-fin edge. (B) 2D-macrophage area from 24-72 hpb. (C) 2D-neutrophil area from 24-72 hpb. (D-F) 3-days post fertilization wild-type larvae with fluorescent macrophages (mpeg-GFP) and neutrophils (lyz-BFP) were burned and infected with fluorescent C. albicans (RFP). (D) Representative images of larvae infected with C. albicans at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail-fin edge. (E) 2D-macrophage area from 24-72 hpb. (F) 2D-neutrophil area from 24-72 hpb. Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test. *p<0.05, ***p<0.001.
To elucidate which phagocyte may be predominantly driving fungal clearance, we next quantified the number of neutrophils and macrophages that were in contact with fungal cells. Throughout imaging both A. fumigatus and C. albicans infections, macrophages consistently harbored numerous phagocytosed fungi, while direct neutrophil-fungal interactions were less common (Fig. 3A-B). Quantification of the number of immune cells interacting with fungi per fish showed that macrophages were engaged with fungal cells around twice as often as neutrophils, possibly suggesting a more direct role for macrophages in controlling fungal infection in burn wounds (Fig. 3C-D).
(A-B) 24 hpb representative images of wild-type larvae with fluorescent macrophages (mpeg-GFP) and neutrophils (lyz-BFP) inoculated with fluorescent (A) A. fumigatus CEA10 spores (RFP) or (B) C. albicans yeast cells (RFP). Images represent maximum intensity projections of z-stacks (scale bar is 100µm) (C-D) Number of macrophages and neutrophils contacting (C) A. fumigatus cells and (D) C. albicans cells. Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24 larvae per condition. p values calculated by a student’s t-test. ****p<0001.
Macrophage depletion leads to increased fungal burden in the burn wound
Our data suggest that macrophages may be the predominant leukocyte driving fungal BWI clearance (Fig. 2). However, previous work has shown that fungal phagocytosis by macrophages often does not result in fungal clearance, as both A. fumigatus and C. albicans can germinate within and escape from macrophages (21–24). To better elucidate the contribution that macrophages play in BWI resolution we transiently depleted macrophages by injecting clodronate liposomes into the caudal vein of 2 dpf larvae expressing fluorescently labeled macrophages (mpeg-GFP) and proceeded to burn and infect larvae at 3 dpf (25). For A. fumigatus infection, we found that clodronate effectively depleted macrophages and resulted in a significant increase in fungal burden at both 24 and 48 hpb when compared to PBS liposome injected larvae (Fig. 4A-C). Interestingly, we did not observe significant differences in germination, as both germ tubes and hyphae were observed throughout the experiment (Fig. 4D). However, at 72 hpb there was a slight, but insignificant, increase in germination in clodronate treated fish. Similar to A. fumigatus, clodronate injected fish infected with C. albicans showed a significant increase in fungal burden at 24 hpb (Fig. 4E-G), although this difference was lost at later time points throughout the infection process. Altogether, our data support an important role for macrophages in controlling fungal burden, especially during the first 24 hpb.
(A) Representative images of clodronate liposome or PBS liposome treated mpeg-GFP larvae infected with A. fumigatus CEA10 at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail fin edge. (B) 2D-macrophage area from 24-72 hpb. (C) 2D-fungal (RFP) area from 24-72 hpb. (D) Percent of larvae with germinated spores from 24-72 hpb. Each dot represents percent of an individual experiment. (E) Representative images of clodronate liposome or PBS liposome treated mpeg-GFP larvae infected with C. albicans (RFP) at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail-fin area. (F) 2D-macrophage area from 24-72 hpb. (G) 2D-fungal (RFP) area from 24-72 hpb. Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test. **p<0.01, ***p<0.001, ****p<0001.
Neutrophil deficiency leads to increased hyphal growth within burn tissue
While our data place macrophages as the primary leukocyte directly interacting with fungi at the burn wound interface, there is ample evidence showing neutrophils as key players of the innate immune response against fungal infections (21, 22, 26–28). Indeed, neutrophil deficiency in human patients has been associated with invasive fungal growth and increased risk for burn wound infections, and ex vivo burn wound models using human tissue have shown an important role for neutrophils in the epidermal antifungal immune response (8, 29, 30). To assess the role of neutrophils during zebrafish fungal BWI, we infected larvae with impaired neutrophil function that express a dominant-negative Rac2 (RacD57N) protein under the regulatory control of a neutrophil specific promoter (mpx:rac2D57N). These neutrophils are not able to leave the circulation and are therefore unable to migrate to burn tissue at the infection site (31). Previous work using mpx:rac2D57N fish demonstrated that host-survival is decreased during A. fumigatus hindbrain infection, as well as for C. albicans swim-bladder infection (28, 32). Interestingly, we found no differences in fungal burden for A. fumigatus burn infection in these fish compared to WT (Fig. 5A&C). However, a significant increase in germination and hyphal development was observed in infected mpx:rac2D57N fish at 72 hpb (Fig. 5D). Surprisingly, a small number (3.3%) of mpx:rac2D57N fish presented invasive hyphal growth that extended from the wound towards the head by penetrating muscle tissue of the dorsal fin (Fig. 5B&E). Although rare, this is indicative of severe infection, in which the fungus was not only able to colonize viable tissue but to also invade beyond the infection site.
(A) Representative images at 24 hpb of 3-dpf mpx:Rac2D57N (mCherry) or mpx:Rac2 (mCherry) wild-type larvae infected with A. fumigatus CEA10 (RFP). Images represent maximum intensity projections of z-stacks (scale bar 50µm). Dashed white lines outline tail-fin area. (B) 2D-fungal (RFP) area from 24-72 hpb. (C) Percent of larvae with germinated spores from 24-72 hpb. Each dot represents percent of an individual experiment. (D) Representative image of entire mpx:Rac2D57N larvae infected with A. fumigatus CEA10 (RFP) at 72 hpb (scale bar is 200µm). (E) Percent of fish with invasive hyphal growth extending beyond the notochord and growing toward the head at 72 hpb. Bar represents the sum of fish showing invasive growth from 4 experiments. (F) Representative images at 24 hpb of 3-dpf mpx:Rac2D57N (mCherry) or mpx:Rac2 (mCherry) wild-type larvae infected with C. albicans (mNeon) at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail fin edge. (G) 2D-fungal area from 24-72 hpb. (H) Representative tile image of entire mpx:Rac2D57N (mCherry) larvae infected with C. albicans (mNeon) (scale bar is 100µm). (I) Percent of fish with hyphal growth from 24-72 hpb. Each dot represents percent of an individual experiment. (Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test. *p<0.05,**p<0.01, ***p<0.001)
In contrast to A. fumigatus, mutant fish infected with C. albicans had a significant increase in fungal burden at 24 hpb compared to WT that resolved by 72 hpb (Fig. 5F and 5H). The increase in fungal burden largely correlated with extensive hyphal networks within the tail fins of mpx:rac2D57N infected fish by 24 hpb (Fig. 5G and 5I); ∼55% of mpx:rac2D57N fish showed branched hyphae extending throughout the tail fin by 24 hpb that was not observed in immunocompetent fish. Surprisingly, by 72 hpb invasive hyphal growth was absent in many of these fish, suggesting a redundant role for other immune cells in controlling fungal hyphal formation. Yet, hyphae extending beyond the notochord and growing towards the head of the zebrafish was still seen in ∼15% of the Rac2D57N infected fish (Fig. 5I), making it a more common occurrence as compared to A. fumigatus infection (Fig. 5E). Together, these findings highlight an essential role for neutrophils in controlling invasive hyphal growth during infection with both A. fumigatus and C. albicans.
Impaired macrophage and neutrophil function increases invasive fungal infection
Our data have shown a role for both macrophages and neutrophils in controlling both A. fumigatus and C. albicans infections. Therefore, we hypothesized that infection in fish with deficiencies in both cell types would further exacerbate the disease phenotype. To test this, we used a zebrafish line that expresses a dominant-negative Rac2D57N in leukocytes (coro1a:GFP-rac2D57N) (33). Thus, both macrophages and neutrophils expressing this mutation have migration deficiencies that will impede their movement to the BWI site. Surprisingly, infection of coro1a:GFP-rac2D57N fish with A. fumigatus did not appear to result in increased fungal burden or germination rates (Fig. 6A-C), but did result in an increase in the number of fish showing invasive hyphal growth extending beyond the notochord as early as 24 hpb. Indeed, 10% and 5% of all fish imaged showed hyphal growth extending past the notochord and moving toward the head at 24 and 48 hpb, respectively (Fig. 6D). Therefore, loss of both macrophages and neutrophils further impacts invasive hyphal growth of A. fumigatus, although it appears to be modest.
(A) Representative images of 3-dpf coro1a:GFP-rac2D57N (GFP) and wild-type larvae infected with A. fumigatus CEA10 (RFP) at 24 hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). (B) 2D-fungal area (RFP) at 24-72 hpb. (C) Percent of larvae with germinated spores at 24-72 hpb. Each dot represents percent of an individual experiment. (B) Percent of fish with invasive hyphal growth at 72 hpb. Bar represents the sum of fish with invasive hyphal growth from 3 experiments. Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test.
Unlike A. fumigatus, infection in coro1a:GFP-rac2D57N fish with C. albicans showed a significant increase in fungal burden at all the times analyzed (Fig. 7A-C). This was accompanied by robust hyphal development as early as 24 hpb that persisted throughout infection. Overall, between 50-75% of C. albicans infected coro1a:GFP-rac2D57N zebrafish showed hyphal growth of varying degrees throughout the infection (Fig. 7D). We then further categorized hyphal presence based on the severity of invasive growth (Fig. 7E). At 24 hpb, the hyphal growth was either small hyphae localized to the burn wound edge, denoted as level 1, or extensive and branched hyphal growth that was localized within the tail fin, denoted as level 2 (Fig. 7F). However, as the infection progressed the percentage of fish displaying hyphae extending past the notochord and towards the head of the fish, denoted as very invasive level 3 hyphal growth, increased. Indeed, ∼55% and ∼80% of hyphal growth at 48 hpb and 72 hpb, respectively, were classified as severely invasive (Fig. 7F). Therefore, it appears that both macrophages and neutrophils are essential to control C. albicans during infection. However, it is important to note that coro1a:GFP-rac2D57N fish also showed an innate sensitivity to burn wound injury, and excessive tissue necrosis and death were observed even in the absence of infection (Supplementary Fig. 2A-B). Fungal presence did not appear to further increase death, and it seems likely that the excessive tissue damage observed in coro1a:GFP-rac2D57N fish allowed C. albicans to more effectively colonize and invade the burn wound niche.
3-dpf coro1a:GFP-rac2D57N (GFP) or wild-type larvae were injured and infected with A. fumigatus CEA10 or C. albicans and imaged at 24 hpb. (A) Representative brightfield tile image of entire uninfected wild-type or coro1a:Rac2D57N larvae infected (scale bar 200µm). (B) Fish were monitored for survival at 24-72hpb. Survival curve of infected versus uninfected wild-type and coro1a:GFP-rac2D57N larvae.
(A) Representative images of 3-dpf coro1a:GFP-rac2D57N (GFP) or wild-type larvae infected with C. albicans (dTomato) and imaged at 24, 48 and 72-hpb. Images represent maximum intensity projections of z-stacks (scale bar is 50µm). Dashed white lines outline tail fin edge. (B) Representative tile images of entire coro1a:GFP-rac2D57N (GFP) larvae infected with C. albicans (dTomato) at 72 hpb. (scale bar is 100µm). (C) Log10 2D-fungal area from 24-72 hpb. (D) Percent of fish showing hyphal growth from 24-72 hpb. Each dot represents percent of an individual experiment. (D) Schematic of hyphal growth levels classification. Level 1 indicates small and controlled hyphal growth present at the burn wound edge. Level 2 is denoted as extensive hyphal growth within the tail fin. Level 3 is classified as extensive and invasive hyphal growth that penetrates beyond the notochord and extends towards the head of the fish. (F) Percent of fish showing hyphal growth within each classification level (1-3). Bar colors represent the different levels and are derived from the mean of 3 individual experiments. (Bars on graphs presented as mean + standard deviation. Each dot represents data from an individual fish and results represent data pooled from 3 independent experiments. n = 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test. ***p<0.001, ****p<0001)
Increased β(1,3)-glucan exposure attenuates C. albicans burn wound colonization in zebrafish
Both macrophages and neutrophils are needed to control burn wound infection by C. albicans (Fig. 7) and we hypothesized that infection with fungal mutants that have an attenuated ability to evade host immune system recognition would result in enhanced fungal clearance within this niche. C. albicans mutants with a hyperactive Cek1 mitogen activated protein kinase pathway (STE11/Ptet-offSTE11ΔN467) or that are deficient in the cell wall protein Fgr41 (fgr41 Δ/Δ) display increased exposure of the immunogenic cell wall epitope β(1,3)-glucan (Fig. 8A), induce increased TNFα production by macrophages in vitro and display virulence attenuations that are host immune system dependent during systemic infection in mice (34–36). Therefore, we predicted that these mutants would also be attenuated in their ability to colonize burn tissue within an immunocompetent zebrafish burn model. Indeed, WT fish infected with either STE11/Ptet-offSTE11ΔN467 or fgr41Δ/Δ showed significantly less fungal burden than WT C. albicans infected fish at both 24 hpb and 48 hpb (Fig. 8B). Thus, masking immunogenic β(1,3)-glucan is an important mechanism utilized by C. albicans to successfully colonize burn tissue.
(A) Representative microscopy images of β(1,3)-glucan exposure of wild-type, STE11/Ptet-OFFSTE11ΔN467 and fgr41Δ/Δ C. albicans strains. The scale bar indicates 10μm. (B) CFUs per fish from wild-type zebrafish larvae that were infected with either the C. albicans wild-type or mutants that show increased β(1,3)-glucan exposure (STE11/Ptet-OFFSTE11ΔN467 and fgr41Δ/Δ). Error bars on graph presented as mean + standard deviation. Each dot represents data from an individual fish. Results represent data pooled from 3 independent experiments. n= 24-30 larvae per condition. p values calculated by a two-way ANOVA with Šidák’s multiple comparisons test. **p<0.01, ****p<0001.
Discussion
Fungal burn wound infections are an understudied, but significant complication that increases the risk of disease and mortality in patients. The challenge of finding suitable animal models to study the biology behind fungal infections during burn injury represents a primary obstacle to better understanding disease kinetics. Because of this, most of the available data is limited to clinical studies, ex vivo models and sparse in vivo studies that have primarily focused on reactive treatment strategies after infection (8, 10, 12, 37, 38). Therefore, there remain significant gaps in our understanding of the basic biology of the fungal-host burn wound interface. In this study we exploited an already established burn injury model in zebrafish larvae and developed a fungal burn wound infection model using two of the most common species isolated from burn wounds, A. fumigatus and C. albicans.
By combining live imaging and quantification of fungal burden, we demonstrate that both A. fumigatus and C. albicans successfully colonize and are cleared from immunocompetent zebrafish larvae (Fig. 1). Both macrophages and neutrophils are recruited to the site of trauma (Fig. 2). Fungal colonization results in an increase in recruitment of neutrophils in response to both pathogens, but only C. albicans induces increased macrophage infiltration as well. Thus, it appears that each fungal isolate has distinct effects on the host immune response following burn injury. It is possible that the divergent metabolic nature between the pathogens may account for these differences. C. albicans yeast cells are active and appear to readily undergo morphological transitioning to hyphae at the early stages (2-6 hpb) of burn wound colonization (Fig. 1). Hyphae are both more invasive than yeast cells, but also have increased exposure of β-glucan moieties within their cell wall that may further exacerbate the host immune response (39–41). By contrast, A. fumigatus conidia are generally considered metabolically dormant, and germination does not occur until 24-48 hpb (Fig. 1) (42). Structurally, the spore surface is covered by hydrophobins that mask underlying immunogenic cell wall epitopes and aids in immune cell evasion (43, 44). Therefore, the delayed germination rate of A. fumigatus spores likely contributes to the reduced immune cell infiltration as compared to C. albicans.
Our findings suggest distinct roles for neutrophils and macrophages at different stages of infection. Depletion of macrophages results in increased fungal burden at early stages of infection (Fig. 4), while infection in neutrophil defective zebrafish results in invasive hyphal growth by both fungi (Fig. 5). However, both fungal burden and hyphal presence are largely restored to WT levels at the later stages of infection in both macrophage or neutrophil deficient zebrafish, suggesting redundant functions. Infection in coro1a:GFP-rac2D57N zebrafish larvae, which have defective neutrophils and macrophages, increases both fungal burden and invasive hyphal growth with C. albicans (Fig. 7), supporting the idea that macrophages and neutrophils have synergistic and partially redundant roles in antifungal immunity in burns. The observation that infection in fish with defective neutrophils results in increased invasive hyphal development (Fig. 5) recapitulates findings from both clinical reports and ex vivo human skin infection models (8, 30). However, our findings that loss of macrophages impacts fungal colonization (Fig. 4) raises interesting questions about the role of macrophages in fungal control during thermal injury. C. albicans yeast cells can remain viable, divide and be transported within macrophages to distant sites from the initial infection (45). Studies with A. fumigatus using the zebrafish hindbrain infection model have similarly shown that macrophages phagocytose but often do not kill conidia, and actually delay clearance by inhibiting germination and subsequent hyphal killing by neutrophils (22). Yet, it was recently reported that macrophage specific Rac2 rescues control of invasive hyphal growth by A. fumigatus in pan-Rac2 deficient fish during hindbrain infection (46). Our data further indicates that macrophages can control hyphal growth in the absence of neutrophils, demonstrating that macrophages do play an important role in antifungal immunity. A caveat is that there may be off target effects of clodronate treatment that contribute to some of the observed phenotypes, since studies in mice suggest that clodronate liposomes may also have effects on neutrophil function (47).
Our findings highlight a distinction between colonization and infection. In humans, skin infection is characterized by fungal penetration to variable depths within viable tissue, along with angioinvasion (48–50). Colonization is limited to the wound surface and proliferation at the interface between viable and non-viable tissue (9, 49, 51). In immunocompetent and macrophage depleted fish, fungi were highly localized and displayed minimal hyphal development, which suggests colonization of the tissue. However, neutrophil deficiency was associated with fungal invasion of neighboring tissues. Although zebrafish larvae do not fully develop vasculature until later stages of development, we propose that the invasive hyphal growth observed in larvae with impaired neutrophil function represents infected tissue and supports a key role of neutrophils in preventing invasive growth within burns (52).
To test the utility of this fungal burn wound model for identifying fungal mechanisms of tissue colonization, we infected fish with C. albicans mutants that had increased β(1,3)-glucan exposure within their cell walls (Fig. 8). β(1,3)-glucan is highly immunogenic, and these mutants have been previously shown to induce robust pro-inflammatory cytokine production by macrophages in vitro and show immune system-dependent virulence attenuations during systemic infections in mice (34–36). Our results show limited tail fin colonization with unmasked C. albicans mutants (Fig. 8), similar to the phenotypes observed in mice. These findings suggest that masking of β(1,3)-glucan is necessary for persistent tissue colonization in larval burn wounds, and highlights the efficacy of this model for studying fungal mechanisms of tissue colonization during burn injury. However, further analysis should be performed to better characterize the host response to infection with cell wall mutants within the context of BWIs, with the potential for these findings to have broader implications across different infection models and disease manifestations.
Materials and Methods
Ethics Statement
The use of zebrafish in this research was approved by the Institutional Animal Care and Use Committee (IACUC) at University of Wisconsin-Madison College of Agricultural and Life Sciences (CALS). The animal care and use protocol M005405-A02 adheres to the guidelines established by the federal Health Research Extension Act and the Public Health Service Policy on the Humane Care and Use of Laboratory Animals, led by the National Institutes of Health (NIH) Office of Laboratory Animal Welfare (OLAW).
Zebrafish Husbandry and Maintenance
Adult zebrafish were maintained under a light/dark cycle of 14/10 hours, as previously described (32). For experiments, embryos were collected, screened and maintained at 28.5°C. For A. fumigatus infection in wild-type Rac2, zebrafish larvae that weren’t expressing mCherry were chosen as the fungus was RFP positive. All zebrafish lines utilized in this study are listed in Table 1.
Fungal strains growth and conditions
Aspergillus fumigatus and C. albicans strains were maintained as glycerol stocks at −80°C. For A. fumigatus, the strain was activated before every experiment by streaking on a plate of Glucose Minimal Media (GMM) and incubating for 3 days in darkness at 37°C. Conidia were harvested in 0.01% Tween-water by scraping with an L-shaped spreader, and immediately filtered through sterile Miracloth. To prepare for burn wound infection, conidia solution was centrifuged and washed three times in sterile 1x DPBS (-CaCl2, -MgCl2). After counting conidia, solution was resuspended in PBS. The appropriate volume to achieve 2 x 107 spores/mL was then added to a conical tube (per dish of larvae) with 5 mL of E3 without methylene blue (-MB).
For C. albicans, strains were activated on YPD agar plates (1% Yeast Extract, 2% peptone, 2% dextrose, 2% agar) from glycerol stocks. One day prior to infection, overnight cultures in liquid YPD media were started and left to incubate for ∼16 hours while shaking at 225rpm at 30°C. Following incubation, cultures were transferred to a 15ml conical tube and centrifuged at 3500rpm for 5 minutes. Following centrifugation, the supernatant was aspirated, and the fungal cell pellet was washed 2 times with 10ml of 1x DPBS. After washing, the fungal cell pellet was resuspended in 5ml of E3 -MB media, counted on a hemocytometer, and diluted to a concentration of 2×107 cells/ml in 6ml of E3 -MB media. All fungal strains used in this study are listed in Table 2.
For fixation of fungal cells, A. fumigatus and C. albicans cells were incubated in 4% Paraformaldehyde (PFA) for 30 minutes at room temperature and washed three times in 1x DPBS.
Thermal injury and infection
Before every experiment, larvae were anesthetized in E3 -MB containing 0.2 mg/mL Tricaine (ethyl 3-aminobenzoate, Sigma). For thermal injury and infection, larvae were placed in 10 mL of E3 -MB + Tricaine in a 60 mm tissue culture treated dish (8-10 larvae per dish). A fine tip cautery pen (Bovie, Symmetry Surgical, Antioch, TN) was used to burn the caudal fin without injuring the notochord (16). Immediately after burn, 5 mL of liquid was removed from the dish and replaced with 5 mL of fungal solution (containing 2×107 cells/ml of fungi) to achieve a final concentration of 1 x 107 cells/mL. For control larvae, this was done with 5 mL of E3 -MB + 100uL of PBS. Larvae were slowly shaken and kept in solution for 1 hour. Afterwards, larvae were rinsed five times with E3 -MB and incubated at 28.5°C for further analysis.
Clodronate liposome injection
Transgenic larvae with fluorescent macrophages (mpeg: GFP) were microinjected at 2dpf into the caudal vein plexus with 2nL of clodronate or PBS liposomes (Liposoma) with 0.1% phenol red for easier visualization (25).
Fungal Burden Enumeration by colony forming units
For A. fumigatus and C. albicans, individual larvae were collected at 24, 48 and 72 hpb and placed in 1.5 ml microcentrifuge tubes with 90uL of 1x DPBS with 50ug/mL gentamycin and 50ug/mL kanamycin. Larvae were homogenized for 15 seconds in a mini bead beater at maximum speed. The entire volume was plated on solid GMM (A. fumigatus) or YPD (C. albicans) and plates were incubated at 30-37°C for 2 days and CFUs were counted. Three biological replicates were run for all cell lines tested and the total number of fish used is listed in the appropriate figure legends. Statistical analysis was performed using a one-way ANOVA with a Tukey’s multiple comparison analysis (GraphPad Prism, v7.0c software).
Live imaging acquisition
For live imaging acquisition, larvae were anesthetized and embedded flat on their side in a 35mm glass bottom dish (CellVis) by pouring 1-2% low melting point agarose (E3 -MB) on top. Z-series images (5µm slices) of the tail fin were acquired on a spinning disk confocal microscope (CSU-X; Yokogawa) with a confocal scan head on a Zeiss Observer Z.1 inverted microscope, Plan-Apochromat NA 0.8/20x objective, and a Photometrics Evolve EMCCD camera. All images were acquired using ZEN 2.6 software. Tiles and Z-series were acquired and stitched using ZEN software for whole larvae images.
Imaging analyses and processing
Images were processed and analyzed using FIJI ImageJ (56). Maximum intensity projections of Z-series images were used for representative images and for further analysis. Fungal burden and leukocyte recruitment were analyzed by manual thresholding and measuring 2D area of the corresponding fluorescent signal. Germination for A. fumigatus was scored as the presence or absence of germinated conidia (germ tube or hyphae). Invasive hyphal growth was quantified by the presence of branched hyphal networks extending beyond the tail fin edge toward the caudal vein of the fish. For immune cells contacting fungus analysis, each image in the whole Z-series was analyzed and the number of macrophages or neutrophils physically contacting fungal cells within the same image plane were counted. Counts were then pooled for all planes to generate contact numbers per fish. In all representative images, brightness and contrast histograms were adjusted for visual purposes only, and no alterations were made to images prior to analysis. For measuring tissue regrowth, the total fin tissue area distal to the notochord was outlined using the polygon tool (16). Three biological replicates were run for all experiments and the total number of fish used is listed in the appropriate figure legends. Statistical analysis was performed using a two-way ANOVA with a Šidák’s multiple comparisons test (GraphPad Prism, v7.0c software).
Β(1,3)-glucan staining
Fungal Β(1,3)-glucan was stained as previously described, with modification to the secondary antibody used in this study (57). Here, a secondary rat-anti-mouse FITC conjugated antibody (Biolegend; 406001) was used at a 1:250 dilution for staining. Following staining, cells were imaged on a Zeiss Observer Z.1 inverted microscope and images processed via Fiji ImageJ.
Funding Sources
This work was supported by the National Science Foundation (NSF) Graduate Research Fellowship Program under Grant No. (DGE-2137424) awarded to N.M.S., F32AI183696-01 awarded to A.S.W. from the National Institute of Allergy and Infectious Diseases (NIAID) of the National Institutes of Health (NIH) and by NIH GMS K99GM147303 awarded to A.Horn and R35 GM118027 to AH. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the NSF nor the NIH.
Acknowledgements
We would like to thank Dr. Robert Wheeler (University of Maine) for providing the fluorescent C. albicans strains used in this study. Burn wound infection schematic (Figure 1) created in BioRender. Huttenlocher, A. (2024) https://BioRender.com/i26e377