Essential role of hepcidin in host resistance to disseminated candidiasis

Abstract

Candida albicans is a leading cause of life-threatening invasive infection despite antifungal therapy. Patients with chronic liver disease are at increased risk of candidemia, but the mechanisms underlying this susceptibility are incompletely defined. One consequence of chronic liver disease is an attenuated ability to produce hepcidin and maintain organismal control of iron homeostasis. To address the biology underlying this critical clinical problem, we demonstrate the mechanistic link between hepcidin insufficiency and candida infection using genetic and inducible hepcidin knockout mice. Hepcidin deficiency led to unrestrained fungal growth and increased transition to the invasive hypha morphology with exposed 1,3-β-glucan, which exacerbated kidney injury, independent of the fungal pore-forming toxin candidalysin in immunocompetent mice. Of translational relevance, the therapeutic administration of PR-73, a hepcidin mimetic, improved the outcome of infection. Thus, we identify hepcidin deficiency as a host susceptibility factor against C. albicans and hepcidin mimetics as a potential intervention.

Keywords: 1,3; CP: Metabolism; CP: Microbiology; Candida albicans; PR73; chronic liver disease; hepcidin; hepcidin mimetic; iron; β-glucan.

Figure 1.. CLD attenuates hepcidin production and is associated with renal iron accumulation: Implications for human candidiasis and disease

Compared to patients without liver disease, hepatic hepcidin gene expression and serum hepcidin levels are attenuated in patients with CLD (A and B). 8- to 10-week-old C57BLK/6 mice were injected intraperitoneally with 0.3% CCl₄ or vehicle (corn oil) at a dose of 10 μL/g of body weight twice weekly for 5 weeks and euthanized. Compared to vehicle, picrosirius red staining of the CCl₄ -injected livers revealed extensive fibrosis with multiple fibrotic septa and divided hepatic lobules (C and D). Scale bar: 50 μM. The intra-hepatic hepcidin gene expression was significantly attenuated in fibrotic livers (E). The kidneys of the vehicle- and CCl₄ -treated mice were stained for Perl’s detectable iron deposits. Perl’s detectable iron deposits were not observed in vehicle-treated kidneys (F). However, following CCl₄ treatment, blue Perl’s detectable iron deposits were observed in the renal tubules (G). Scale bar: 30 μM. Arrows denote the edge of the section. Each point represents a patient. A 2-tailed Mann-Whitney test determined statistical significance, plotted as mean ± SEM. **p < 0.005, ***p < 0.001.

Figure 2.. Hepcidin deficiency worsens outcomes of disseminated candidiasis

12-week-old littermates (WT) and Hamp−/− mice were intravenously infected with 4 × 10⁵ C. albicans (SC5314). On day 3 post infection, 10% mortality was observed in WT mice, whereas all Hamp−/− mice died. The WT mice reached a similar endpoint by day 6 (A). To reduce mortality, mice were intravenously injected with 2 × 10⁵ C. albicans, and tissues were harvested 3 days later (B). Compared to WT mice, iron-overloaded Hamp^−/− mice had a significantly higher fungal burden in the kidneys (C–E) and liver (F). In contrast, the iron-sufficient WT spleens harbored significantly more fungal colonies than iron-deficient Hamp^−/− spleens (G). Experiments were performed twice with n = 8–9 each time, and data from a representative experiment are shown. The data were significant each time. A 2-tailed Mann-Whitney test determined statistical significance. Data are presented as mean ± SEM. *p < 0.05, **p < 0.001, ***p < 0.0001. Grocott methenamine silver staining revealed that the fungus in the WT kidney was predominantly in yeast-pseudohypha form (H) but had undergone a hyphal transformation in the Hamp^−/− kidneys (I). Representative images are shown. Scale bars: 100 and 30 μM.

Figure 3.. C. albicans exacerbates AKI and inflammation in hepcidin knockout mice

WT and Hamp^−/− mice were infected with 2e⁵ C. albicans yeast, and tissues were harvested 3 days later. 10% formalin-fixed kidney slices were used for histopathology. Hematoxylin and eosin (H&E) staining revealed that, compared to WT kidneys (A and B), the Hamp^−/− kidneys had more tubular epithelial cell necrosis (dark pink fragmented cytoplasm with no nuclei), with casts, luminal debris, and multiple foci of inflammation (C and D) Scale bars: 100 μM (A and C) and 50 μM (B and D). Markers of kidney injury and inflammation were comparable in naive WT and Hamp^−/− mice. Post infection, AKI, measured by plasma creatinine and the proximal tubular injury markers Ngal and Kim1, was exacerbated in Hamp^−/− mice (E–G). Grocott’s methenamine silver staining of the infected Hamp^−/− kidneys revealed widespread distribution of fungal hyphae, most of which were in the corticomedullar region (the region between the two yellow circles) and the deep cortex (the region outside the larger yellow circle). In contrast, the renal medulla (the region between the smaller yellow circle and black circle) and the pelvis/papilla (the region inside the black circle) had limited fungal growth (H). The region of fungal growth correlated well with the iron-loaded regions of the Hamp^−/− kidneys: the corticomedullary region (the region between the two yellow circles) and the deep cortex (the region outside the larger yellow circle) (I). Scale bar: 30 μM. Naive WT and Hamp^−/− mice had comparable intra-renal gene expression of chemoattractants such as Csf3 (J) Ccl2 (K), and Cxcl11 (L). C. albicans significantly increased the expression of all three chemoattractants in the WT kidneys, which were further significantly elevated in Hamp^−/− kidneys (J–L). The immune infiltrates composed of neutrophils (M) and monocytes (N) showed the same trend. Experiments were performed twice with n = 8–9 each time, and data from a representative experiment are shown. The data were significant each time. Data were analyzed using two-way ANOVA with HolmŠídák’s multiple comparisons test and represented as mean ± SEM. *p < 0.05, **p < 0.005, ***p < 0.001. ****p < 0.0001.

Figure 4.. Expression of Ece1 correlates with cell death, inflammation, and loss of renal parenchyma

3 day post C. albicans infection, the kidneys of WT and Hamp^−/− mice were analyzed for expression of the fungal gene Ece1. Ece1 gene expression was observed only in the Hamp^−/− kidneys (A). Cell death markers were comparable in naive WT and Hamp^−/− kidneys (B and C). GsdmD (pyroptosis) and Mlkl (necrosis) gene expression were not significantly induced in the infected WT kidneys (B and C). However, the infected Hamp^−/− kidneys significantly upregulated the expression of GsdmD and Mlkl (B and C). Experiments were performed twice with n = 8–9 each time, and data from a representative experiment are shown. The data were significant each time. Data were analyzed using 2-way ANOVA with Holm-Šídák’s multiple-comparisons test and represented as mean ± SEM. *p < 0.05, **p < 0.005, ***p < 0.001, ****p < 0.0001. Proximal tubular epithelial cells (PTECs) are the most abundant epithelial cells in the kidneys. Lotus tetragonolobus lectin (LTL) staining, a marker for PTECs, revealed a uniform distribution throughout the infected WT kidneys (D) (the red arrow indicates the edge of the section). However, infected Hamp^−/− kidneys presented with multiple sections devoid of LTL-positive PTECs, indicating loss of kidney parenchyma (E). Scale bars:100 μM (D and E).

Figure 5.. Hepcidin deficiency drives fungal burden, but the severity of AKI depends on candidalysin

WT and Hamp^−/− mice were infected with 2e⁵ Ece1-deficient C. albicans yeast (ece1ΔΔ), and the kidneys were harvested on day 3. Compared to WT kidneys, the fungal burden was significantly higher in the Hamp^−/− kidneys (A and B) and was comparable to Hamp^−/− kidneys infected with an isogenic strain sufficient for candidalysin (ece1ΔΔ + ECE1) (B). The ece1ΔΔ C. albicans had undergone a hyphal transformation in Hamp^−/− kidneys, and histology showed multiple foci of infection, which were not observed in WT kidneys (C and D). Scale bars: 10×, 100 μM (C); 40×, 40 μM (D). Compared to WT mice, ece1ΔΔ strain-infected Hamp^−/− mice developed significantly more AKI, as measured by plasma creatinine and intra-renal Ngal gene expression (E and F). AKI was exacerbated by the ece1ΔΔ + ECE1 strain (E and F). Compared to WT kidneys, ece1ΔΔ infected Hamp^−/− kidneys had significantly greater neutrophil and monocyte infiltrates (G and H). However, both immune cell types were significantly lower than ece1ΔΔ + ECE1 infected Hamp^−/− kidneys (G and H). Data from two independent experiments were pooled and analyzed (n = 9–10) using 2-way ANOVA with Holm-Šídák’s multiple-comparisons test and are represented as mean ± SEM. *p < 0.05, **p < 0.005, ***p < 0.001, ****p < 0.0001.

Figure 6.. Iron exposes fungal β-glucan to potentiate chemokine production in human proximal renal tubular cells independent of candidalysin

Ece1 knockout C. albicans yeast (ece1ΔΔ) and its isogenic candidalysin revertant (ece1ΔΔ + ECE1) were grown in standard or iron-rich medium (100 μM FeCl₃, equivalent to serum iron content of Hamp^−/− mice) for 24 h and stained for exposed β-glucan and concanavalin A. Both fungal strains remained as yeast in the standard medium and did not expose β-glucan (A and B). However, adding iron to the growth medium transformed and sustained both ece1ΔΔ and ece1ΔΔ + ECE1 in their hyphal form with exposed β-glucans (C and D). Scale bar: 100 μM. Experiments were performed 3 times each with 3 technical repeats. Representative images from one of the experiments are shown. For in vivo detection and quantification of exposed 1,3-β-glucan, WT and Hamp^−/− mice were intravenously infected with 200,000 red fluorescent C. albicans (CAF2–1-dTomato). After 3 days, whole-kidney homogenates were analyzed for fungal cells with exposed 1,3-β-glucan. Infected WT kidneys revealed red yeast and occasional hyphae with exposed 1,3-β-glucan (E and G). In contrast, the infected kidney digests of Hamp^−/− mice had more red-fluorescent hyphae with exposed 1,3-β-glucan (green) (F and H). Scale bars: 20×, 100 μM (E and F) and 100×, 30 μM (F and H). For quantifying fungi with exposed 1,3-β-glucan, five random 20× images from each slide (n = 5 from each strain) were captured. Each dot represents the number of yellow fungi within an image. Individual values are plotted with standard deviation (I). A 2-tailed Mann-Whitney test was used to determine statistical significance. ***p < 0.0001. ece1ΔΔ C. albicans was grown overnight in standard or iron-rich medium (100 μM FeCl₃) at 33°C and then co-cultured with HK-2 cells at an MOI of 1:3 (cells:fungus) at 37°C. Compared to ece1ΔΔ C. albicans grown in standard medium, that grown in iron-rich medium induced significantly more IL-8 secretion (J). Data were normalized to IL-8 secreted by naive cells at each time point. The growth of ece1ΔΔ C. albicans originally grown in standard and iron-rich medium and subsequently in HK-2 medium was followed at 37°C for 4 h. There was no significant difference between the growth rate of the fungus under the two conditions (K). Experiments were performed twice each with 3 technical replicates each time, and the data were pooled for analysis. A 2-tailed Mann-Whitney test determined statistical significance at each time point, plotted as mean ± SEM. *p < 0.05, **p < 0.001.

Figure 7.. Acute hepcidin deficiency exacerbates C. albicans-induced renal pathology and is attenuated by PR-73, a synthetic hepcidin mimetic

Inducible hepcidin knockout mice (iHamp^−/−) and their WT littermates were injected with tamoxifen (A). Tamoxifen injection did not attenuate serum hepcidin levels in WT mice, which increased significantly post C. albicans infection (B). In contrast, tamoxifen significantly attenuated serum hepcidin levels in iHamp^−/− mice, which remained low after C. albicans infection (B). Statistical significance was determined by a 2-tailed Wilcoxon matched-pair sign rank test, plotted as mean ± SEM. ***p < 0.001, ****p < 0.0001. WT littermates and iHamp^−/− mice were injected with tamoxifen and C. albicans. A cohort of iHamp^−/− mice received vehicle or PR-73 (50 nmol, intraperitoneal) starting 4 h after Candida infection (C). Compared to WT littermates, the kidneys of vehicle-treated iHamp^−/− mice had a significantly higher fungal burden, which was significantly reduced by PR-73 therapy (D). The fungus was still in the yeast form in the WT littermates (E) but had undergone a hyphal transformation in the vehicle-treated iHamp^−/− mice (F). PR-73 treatment was associated with a reduction in hyphal morphogenesis (G). Reduced fungal burden and hyphal transformation in PR-73 treated iHamp^−/− mice mitigated PTEC injury, as measured by renal Ngal gene expression (H). Data from two independent experiments were pooled and analyzed (n = 10) using 2-way ANOVA with Holm-Šídák’s multiple-comparisons test and represented as mean ± SEM. *p < 0.05, **p < 0.005, ****p < 0.0001. Scale bars: 100 and 30 μM (E–G).