Honokiol Reduces Fungal Load, Toll-Like Receptor-2, and Inflammatory Cytokines in Aspergillus fumigatus Keratitis

Abstract

Purpose: We characterized the effects of Honokiol (HNK) on Aspergillus fumigatus–caused keratomycosis and the underlying mechanisms. HNK is known to have anti-inflammatory and antifungal properties, but the influence on fungal keratitis (FK) remains unknown.

Methods: In ex vivo, minimum inhibitory concentration and Cell Count Kit-8 assay were carried out spectrophotometrically to provide preferred concentration applied in vivo. Time kill assay pointed that HNK was fungicidal and fungistatic chronologically. Adherence assay, crystal violet staining, and membrane permeability assay tested HNK effects on different fungal stages. In vivo, clinical scores reflected the improvement degree of keratitis outcome. Myeloperoxidase (MPO) assay, flow cytometry (FCM), and immunohistofluorescence staining (IFS) were done to evaluate neutrophil infiltration. Plate count detected HNK fungicidal potentiality. RT-PCR, Western blot, and enzyme-linked immunosorbent assay (ELISA) verified the anti-inflammatory activity of HNK collaboratively.

Results: In vitro, MIC₉₀ HNK was 8 µg/mL (no cytotoxicity), and Minimal Fungicidal Concentration (MFC) was 12 µg/mL for A. fumigatus. HNK played the fungistatic and fungicidal roles at 6 and 24 hours, respectively, inhibiting adherence at the beginning, diminishing biofilms formation, and increasing membrane permeability all the time. In vivo, HNK improved C57BL/6 mice outcome by reducing disease severity (clinical scores), neutrophil infiltration (MPO, FCM, and IFS), and fungal loading (plate count). RT-PCR, Western blot, and ELISA revealed that HNK downregulated mRNA and protein expression levels of Toll-like receptor-2 (TLR-2), high mobility group box 1 (HMGB1), IL-1β, and TNF-α.

Conclusions: Our study suggested HNK played antifungal and anti-inflammatory roles on keratomycosis by reducing survival of fungi, infiltration of leucocytes, and expression of HMGB1, TLR-2, and proinflammatory cytokines, providing a potential treatment for FK.

Fungal keratitis (FK) is a serious infectious eye disease in developing countries, associated with agriculture-related ocular trauma, overuse of contact lenses, and postoperative corneal infection, leading to vision loss or blindness.^1,2 Aspergillus fumigatus keratitis is one of the most common FK, which tends to result in poor prognosis because of the lack of effective antifungal agents and excessive innate immune response.^3,4 Therefore it is of significance to develop effective antifungal drugs. 

Honokiol (HNK; 5,5′-diallyl-2,4′-dihydroxybiphenyl), a natural biphenolic compound extracted from the cortex of Magnolia officinalis, is a traditional Chinese herbal material with multiple biological functions, such as anti-tumorigenesis,⁵ anti-inflammation,^6,7 antibacterial,⁸ and neuroprotective properties.⁹ Previous studies suggested that HNK might have an antifungal property.¹⁰ 

Fungal infection mainly involves three steps, including adhesion to host cells, hyphal growth, and biofilm formation.^11–16 It has been demonstrated that HNK can inhibit the adhesion and hyphal growth of Candida albicans in vitro.¹⁷ Biofilm consists of polysaccharides, polypeptides, and extracellular DNA, which are crucial to protect fungi from being recognized and phagocytosed by the host immune response and increase multidrug-resistant isolates.¹⁸ A previous study showed that HNK could resist the invasion of pathogenic microbials by reducing biofilm formation.¹⁷ 

However, the uncontrollable innate immune response is one of the main causes of the poor prognosis of FK.^19,20 Inflammation-induced recruitment of inflammatory cells, like neutrophils in the cornea, trigger the release of cytokines and chemokines. This release further attracts more immune cells and results in precipitation, causing corneal opacity and vision loss.¹⁹ Thus to control the excessive inflammatory response is the key of FK treatment.^21,22 Interestingly, recent studies showed HNK could attenuate the inflammatory response through inhibiting high mobility group box 1 (HMGB1), Toll-like receptor-2 (TLR-2), and proinflammatory molecules in acute pancreatitis²³ and acute kidney injury²⁴ in rat models. Here we hypothesized that HNK could provide an alternative to alleviate A. fumigatus keratitis through its anti-inflammatory activities. 

In this study, we first demonstrated the antifungal and anti-inflammatory roles of HNK in FK mouse models and investigated the underlying mechanisms. Our study may provide a possible therapeutic approach for FK. 

Materials and Methods

Preparation of HNK Solution

HNK powder, purchased from MCE (Shanghai, China), was dissolved in PBS (Solarbio, Beijing, China) or other culture mediums at a concentration of 16 µg/mL, and then was diluted with the corresponding mediums as requested. 

Cell Viability (CCK-8)

Human corneal epithelial cells (HCECs; provided by Laboratory, University of Xiamen, Fujian, China) (3 × 10⁴/mL) were suspended and seeded in the 96-well plate and treated with HNK (0, 2, 4, 8, and 16 µg/mL) for 12, 24, and 48 hours. The cells were incubated for 2 hours with Cell Counting Kit-8 (CCK-8; MCE), and the absorbance was measured at 450 nm. Each sample had five replicates. 

Cell Scratch Test

HCECs (3 × 10⁵/mL) suspension was plated in the 6-well plate and incubated overnight at 37°C. Three parallel lines were scraped on the cell layer using sterile 200 µL pipette tips (Corning, NewYork, USA). The cells were then incubated with HNK (0, 4, 8, and 12 µg/mL) for 24 hours. The width of the scratches observed using an optical microscopy (Axio Vert; Zeiss, Jena, Germany, 100×) were measured before and after HNK treatment. 

HNK Minimum Inhibitory Concentration (MIC)

Conidia were harvested by rinsing the A. fumigatus (species number 3.0772; General Microbiological Culture Collection Center, Beijing, China) malt agar slants with PBS containing 0.1% Tween 20 (Sigma-Aldrich Corp., St. Louis, MO, USA). Conidia suspension was prepared by repeated resuspending, centrifuging (12,000g for 5 minutes), and washing using PBS. MIC HNK for A. fumigatus was assayed by a standardized microdilution method in the 96-well plate described as before.²⁵ Briefly, 100 µL of Sabouraud liquid culture medium was transferred into second to sixth vertical rows. Then removing half of HNK (16 µg/mL, 200 µL) in the seventh column to the left adjacent one realized serial dilutions. Finally, 5 µL of prepared conidia suspension (4 × 10⁶ cfu/mL) was added into the third to seventh columns. The second column was the blank control. The plates were incubated at 37°C without shaking for 36 hours. The HNK MIC₉₀ was determined spectrophotometrically, recognized as the lowest concentration that could inhibit 90% growth of A. fumigatus. 

Time Kill Assay

Based on the MIC, time kill assays were used to evaluate fungicidal/fungistatic activity conveniently and intuitively. As illustrated before,²⁶ conidia suspension was incubated at 37°C, 120 rpm for 9 hours, and was exposed to HNK (0.5 × MIC, 1.0 × MIC, 1.5 × MIC) for 24 hours. Then the conidial aliquots were plated on the Sabouraud agar plates at different time points (0, 3, 6, 12, 18, and 24 hours), and incubated at 37°C for 18 hours. Fungicidal activity was defined as a reduction in colony count >2log10 cfu/mL, and fungistatic activity was defined as a decrease in colony count <2log10 cfu/mL, compared with cfu on the mediums at the beginning. 

Fungal Adherence Assay

Conidia suspension (2 × 10⁵ /mL) containing HNK (0 and 8 µg/mL) was mixed with HCECs and plated on the chambered slides (4/slide) as described previously.²⁷ Each slide was incubated at 37°C for 3 hours, then washed and stained by hematoxylin and eosin (HE) staining.²⁸ The spores adhering to HCECs were photographed by an optical microscopy (Axio Vert; Zeiss, Jena, Germany, 400×). 

Determination of Biofilm Formation Capacity

The preparation of the microbial inoculum before the biofilm formation was carried out by the standardized microdilution method as described previously.²⁵ After 48 hours incubation, biofilm was fixed by methanol and then stained with 0.1% crystal violet (Sigma-Aldrich). Rinsing unbound dye repeatedly with PBS, we released the bound dye of the dry biofilm with 95% ethanol. OD was measured at 570 nm three times. 

Membrane Permeability

Membrane permeability of HNK was detected using Live/Dead FungaLight Yeast Viability L34952 solutions (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) by a flow cytometry (FCM) (CytoFLEX FCM, Beckman Coulter, Indianapolis, IN, USA). Cell samples were stained followed by the experimental protocols to examine cell membrane integrity. All recorded data were analyzed and processed with FlowJo_V10 (Becton, Dickinson, and Company, Franklin Lakes, NJ, USA). 

Animal Models of FK

Healthy C57BL/6 mice (female, 8 weeks old) were purchased from Jinan Pengyue Laboratory Animal Co. Ltd. (Jinan, China) and were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Visual Research. Mice were abdominally anesthetized with 8% chloral hydrate, and intrastromal injections were given using a sterile microliter syringe (10 µL; Hamilton Corp., Bonaduz, GR, Switzerland). After loading 2.5 µL of A. fumigatus conidia suspension (2.5 × 10⁷ cfu/mL) into the syringe, it was inserted obliquely into the midstromal level in the center of the right cornea. The left eyes were blank control. Experimental eyes were treated with 5 µL of HNK (8 µg/mL) topically, whereas conditional control eyes were treated with PBS topically. HNK topical treatment began at 4 hours post infection (p.i.) and then three times per day (dosing every 4 hours in the daytime) at 1 to 5 days p.i. Subconjunctival injection was given at 16 and 40 hours p.i. Based on the observation under a slit lamp at 1, 3, and 5 days p.i., the severity of keratitis was evaluated by clinical score that was the sum of the three aspects of cornea, including opacity density, opacity area, and surface regularity, each of which has a grade of 0 to 4. Meanwhile, ranging from 0 to 12, the severity of keratitis was divided into normal (0), mild (1–5), moderate (6–9), and severe (10–12). Taking a normal cornea as an example, the unsacrificed cornea was given a score of 0 in each aspect, and thus tallied to yield a score of 0.²⁹ Mice corneas removed by a scalpel and microscissor at the indicated times after treatments were prepared for RT-PCR, Western blot, myeloperoxidase (MPO), plate count, FCM, and enzyme-linked immunosorbent assay (ELISA), respectively. Then whole eyes were harvested for immunohistofluorescence staining (IFS). 

MPO Assay

To determine the activity of polymorphonuclear neutrophils (PMNs) quantitatively,³⁰ the corneas were harvested at 3 days p.i. (n = 5/group/time) and treated following the protocol of MPO kit (Nanjing Jiancheng Bioengineering Institute, Nanjing, Jiangsu, China). MPO was measured spectrophotometrically at 460 nm at 37°C; the slope of the line was related to the MPO units/g cornea. 

Plate Count

The advantage of modified counting was that we could count fungi surviving on the cornea after the treatment. Corneal homogenate after treatment was diluted by PBS, then plated on Sabouraud agar mediums, and incubated at 37°C for 24 hours. Colonies were photographed and counted to reflect fungi surviving on the treated cornea (n = 3/group/time). 

FCM Analysis

Neutrophils were shown by monitoring singular cell suspensions stained with fluorescently labeled antibodies. The degree of recruitment was illustrated by the amount and proportion of neutrophils in total inflammatory cells, which were from infected corneas after 3 days treatment. Among the CD45⁺ cells, CD11b⁺Gr-1⁺ cells were neutrophils assayed by FCM. The infected corneas of C57BL/6 mice (n = 5/group/time) were treated with PBS or HNK for 3 days t.i.d. topically, harvested and separated into individual cells, and then stained with CD45-PEvio (1:200; Miltenyi Biotec, Bergisch Gladbach, Germany), CD11b-FITC (1:200; BioLegend, San Diego, CA, USA), and Gr-1-PE (1:80; BioLegend) for 20 minutes. FCM was performed according to the manufacturer's instruction, and PMNs were distinguished from leukocytes by staining with CD11b-FITC and Gr-1-PE. 

Real-Time RT-PCR

Total RNA of corneas were extracted by the RNAiso plus reagent (TaKaRa, Dalian, China), and the RNA was quantified by a NanoDrop ND-1000 Spectrophotometry (Thermo Fisher Scientific) according to the manufacturer's instructions as described earlier.³⁰ The extracted mRNA (n = 5/group/time) was used as the template, and cDNA productions (1 µg) were synthesized by a two-step method using PrimeScript RT reagent Kit with gDNA Eraser (TaKaRa). PCR was assayed in 20 µL reaction system (2 µL of diluted cDNA (1:12.5), 10 µL of SYBR Premix Ex TaqTM (TaKaRa), 1 µL of diluted primers (1:9), and 7 µL DEPC-treated water). The PCR program for the reactions was described as before.³¹ Analysis of target genes was quantified by the 2^ΔΔCT method.³² The corresponding primers are listed in the Table. 

Table.

 

Nucleotide Sequence of Mouse Primers Used for RT-PCR

 

Western Blot

The protein separation procedure has been interpreted previously.³³ Six corneas as one sample (n = 6/group/time) were lysed in 196 µL RIPA buffer (Solarbio), 2 µL phenylmethanesulfonyl fluoride (Solarbio), and 2 µL phosphatase inhibitor (MCE) for 2 hours. Protein concentration was determined by BCA assay (Solarbio). Then the proteins were separated by SDS-PAGE electrophoresis and transferred onto polyvinylidene difluoride membrane (Solarbio). After being blocked with blocking buffer (Solarbio), membranes were incubated with primary antibodies against β-actin (1:3000; Elabscience, Wuhan, China), tubulin (1:3000; Elabscience), TLR-2 (1:1000; ABclonal, Wuhan, China), and TNF-α (1:1000; Abcam, Cambridge, MA, USA) at 4°C overnight. The membranes were washed in PBST three times, subsequent to incubation with a secondary antibody at 37°C for 1 hour. Then the blots were visualized using chemiluminescence (Thermo Fisher Scientific). 

Immunohistofluorescence Staining

IFS was performed according to the methods described previously.³¹ The euthanized mouse eyeballs (n = 3/group/time) were embedded and frozen. A total of 10 µm of corneal lesions were acquired and then fixed in acetone. After being blocked with goat serum (1:100), sections were incubated with monoclonal rat anti-mouse neutrophil marker (1:100; Santa Cruz Biotechnology, Santa Cruz, CA, USA), followed by FITC-conjugated goat anti-rat secondary antibody (1:200; Elabscience). Cell nuclei were stained with DAPI. Images were photographed by a fluorescence microscope (Zeiss Axio Vert, 400×). 

Enzyme-Linked Immunosorbent Assay

Following the manufacturer's instructions (Elabscience), individual cornea (n = 5/group/time) was homogenized in 500 µL of PBS containing 1% protease inhibitors (MCE). Undiluted supernatant was quantified using TNF-α, IL-1β, and HMGB1 ELISA kits. The supernatant, diluted two-fold, was used to evaluate the protein level of HMGB1. Sensitivities were as follows: 9.38 pg/mL (HMGB1), 4.69 pg/mL (TNF-α), and 4.69 pg/mL (IL-1β). 

Statistical Analyses

Disease score was analyzed by the Mann-Whitney U test (SPSS Statistics 19; IBM Corporation, Armonk, NY, USA), MIC, Cell Scratch Test, RT-PCR, Western blot, and so on by an unpaired, two-tailed Student's t-test (GraphPad Prism; GraphPad, San Diego, CA, USA), and three or more groups by Bonferroni multiple comparison test (GraphPad Prism). These data were median or mean + SEM, P < 0.05 (*P < 0.05, **P < 0.01, ***P < 0.001) as significance. All experiments were performed at least twice to ensure practicability. 

Results

HNK Cytotoxicity Test

Cell viability was tested using CCK-8 for different concentrations of HNK in HCECs (Fig. 1A). HNK started to inhibit HCECs proliferation at 8 µg/mL for 48 hours incubation, and no significant cytotoxic effect was observed at 0 to 8 µg/mL for 12 and 24 hours (Fig. 1A). In addition, scratch-wound assay demonstrated that HNK at 8 µg/mL does not affect cell migration compared with control group (Figs. 1B, C). Thus HNK with the concentration no more than 8 µg/mL and incubation time less than 24 hours were considered nontoxic and used in the following in vitro and in vivo experiments. 

Figure 1.

 

Effects of HNK on cell viability and migration. CCK-8 cell viability assay was performed on HCECs at 12, 24, and 48 hours using different concentrations of HNK (0–16 µg/mL) (A). Scratch assay (B) and quantitative analysis (C) of HNK on re-epithelization potentiality. HNK less than 12 µg/mL hardly affected the migration of HCECs (*P < 0.05, **P < 0.01, ***P < 0.001). All data were mean + SEM and analyzed by an unpaired, two-tailed Student’s t-test.

HNK Inhibits the Growth, Adhesion Ability, and Biofilm Formation of A. fumigatus

MIC showed HNK started to inhibit the growth of A. fumigatus at 4 µg/mL and prevented 90% A. fumigatus growth (MIC₉₀) at 8 µg/mL (Fig. 2A). In the time kill assay, A. fumigatus were exposed to HNK at 0.5, 1.0, and 1.5 MIC (MIC = 8 µg/mL), which illustrated HNK killed A. fumigatus in a time- and concentration-dependent manner. The colony counts were significantly decreased at 1.0 and 1.5 MIC compared with 0.5 MIC, suggesting HNK is effective to kill A. fumigatus at 8 µg/mL. Within the first 6 hours, the declined trend of killing fungi was similar between the yellow line (1.0 × MIC) and the blue one (1.5 × MIC). As time went by, the difference between the two lines was obvious. For 0.5 × MIC, we only recorded the formation of colonies within 6 hours. Because fungal colony agglomerated after 6 hours, we failed to count the single units (Fig. 2B). HE staining showed a significant decreased number of adherent conidia on the surface of HNK-treated (8 µg/mL) HCECs compared with PBS-treated cells, indicating HNK inhibits adhesion ability of A. fumigatus in HCECs (Figs. 2C–E). Moreover, biofilm formations were tested in HCECs, which suggested HNK at 8 µg/mL remarkably inhibited biofilm formation (Fig. 2F). 

Figure 2.

 

Effects of HNK on MIC, time kill, adherence, biofilm formation in vitro and in vivo. For A. fumigatus, MIC₉₀ of HNK was 8 µg/mL, MIC₉₉ was 12 or 16 µg/mL (A). Time kill curves for A. fumigatus exposed to 0.5, 1, and 1.5 × MIC of HNK were performed over a period of 24 hours (B; *<2log10 cfu/mL, **>2log10 cfu/mL). The red line (0.5 × MIC) was incomplete because of the high fungi load. HE staining of A. fumigatus infected HCECs treated with PBS (C) or HNK (D). The cell framed by a black box in each panel is magnified in the lower right corner (magnification = ×40 µm). Transparent and rice-like conidia on the cell surface are indicated by black arrows, and quantitative diagram is shown at bottom (E). Biofilm formation was inhibited significantly by HNK at more than 8 µg/mL in vitro compared with PBS treatment, which was demonstrated by the absorbance values of crystal violet released from the biofilm (F). All data were mean + SEM and analyzed by an unpaired, two-tailed Student’s t-test.

HNK Affected Membrane Permeability of A. fumigatus

Dead A. fumigatus conidia were stained and counted by FCM, showing there were more dead conidia in HNK-treated HCECs (Fig. 3C) than PBS-treated cells (Fig. 3B), which indicates HNK can increase the membrane permeability of A. fumigatus. Moreover, a significant increase in the percentage of dead fungi treated with HNK was observed compared with the control (Fig. 3D). 

Figure 3.

 

Effect of HNK on membrane permeability. Conidia were tested by a live/dead kit. Unstained live conidia (A). The ratio of live and dead cell populations was analyzed and possessed by flow cytometer. Signals of double-stained conidia (FITC-SYTO9 and ECD-PI) are shown in the graphs. Representative PBS treated HCECs (B); representative HNK (8 µg/mL) treated HCECs (C). The graph shows the difference of ratio of dead conidia between PBS and HNK treatment (D; P < 0.05). All data were mean + SEM and analyzed by an unpaired, two-tailed Student’s t-test.

Therapeutic Treatment with HNK

The therapeutic efficacy of HNK was tested with treatment beginning at 1 day p.i. Clinical score in the HNK treatment group was significantly lower than the PBS group at 3 and 5 days p.i., with no difference between groups at 1 day p.i. (Fig. 4A), indicating that HNK could improve corneal transparency and may have corneal protective effects from 3 day p.i. Representative photographs taken with a slit lamp camera illustrated that corneas were more transparent in HNK-treated groups versus PBS control at 1, 3, and 5 days p.i. (Fig. 4B). HNK significantly reduced the neutrophil infiltration at 3 days p.i. (Fig. 4C), and viable fungal load in the infected cornea at 5 days p.i. (Figs. 4D–F). 

Figure 4.

 

Disease severity after A. fumigatus infection followed by PBS or HNK treatment. Clinical sores of PBS or HNK treated A. fumigatus keratitis mice at day 1, 3, and 5 (A). Representative slit lamp photographs of PBS- or HNK-treated mouse cornea (B). Quantitative analysis of MPO at 3 days p.i. (C). Representative plates of PBS (E) or HNK (F) treated A. fumigatus infected mouse cornea, and quantitative diagram (D). Data were analyzed using a nonparametric Mann-Whitney U test. Horizontal lines indicate the median values. Magnification ×8.

HNK Reduces the Number and Vitality of Neutrophils in A. fumigatus Keratitis Mouse Model

Neutrophils in PBS-treated (Fig. 5A) or HNK-treated (Fig. 5B) mouse cornea were counted using FCM, showing that HNK treatment dramatically reduced the number of neutrophils compared with PBS treatment (Fig. 5D), which is also consistent with immunofluorescence result (Fig. 5E). The neutrophil proportion between the HNK-treated group and control was statistically significant (Fig. 5C). Moreover, MPO results suggested that HNK treatment inhibited neutrophil vitality in A. fumigatus keratitis mouse model (Fig. 4C). 

Figure 5.

 

Effects of HNK on neutrophil infiltration. The representative images of recruited and accumulated neutrophils FCM analyzed are shown in A–D. Scatter plots gated on CD45+ cells (A,B), FCM quantified the mouse cornea infiltrating neutrophils (CD11b+Gr-1+cells) for PBS-treated (B) and HNK-treated (D) mice at 3 days after A. fumigatus infection in the form of representative dot plots, respectively. The outcomes were monitored by the proportion and the number of PMNs (C,D). Splenocytes were stained as positive control. (E) Immunofluorescent staining demonstrated that HNK treatment reduced neutrophilic infiltrate in the mouse corneal stroma after the fungi infection. NIMP-R14-FITC, neutrophil marker, could specifically label neutrophils. The labeled neutrophils were green with a fluorescence microscope. The results showed that the PBS-treated group showed greater staining intensity than the HNK-treated group, which means the larger number of PMN could be observed in the former group. Green: NIMP-R14-FITC staining; blue: nuclear staining (DAPI) (400× ); merge: neutrophil localization. All data were mean + SEM and analyzed by an unpaired, two-tailed Student’s t-test.

HNK Attenuates Inflammatory Response in A. fumigatus Keratitis Mouse Model

To detect the effects of HNK in the inflammatory response, we evaluated the mRNA and protein expressions of different inflammatory cytokines at 1, 3, and 5 days after A. fumigatus infection. The A. fumigatus elevated mRNA expression levels of TLR-2 (Fig. 6A), TNF-α (Fig. 6B), IL-1β (Fig. 6E), and HMGB1 (Fig. 6F), which were significantly repressed by HNK at 3 and 5 days p.i. Similarly, the increased protein expression levels of TLR-2 (Fig. 6C), TNF-α (Fig. 6D), IL-1β (Fig. 6G), and HMGB1 (Fig. 6H) were also inhibited by HNK at 3 and 5 days p.i. For the uninfected corneas (N), no difference in mRNA or protein expression for any cytokines was detected between the HNK and PBS groups. 

Figure 6.

 

Effects of HNK on the inflammatory response. Real-time PCR results for TLR-2 (A), TNF-α (B), IL-1β (E), and HMGB1 (F) at 1, 3, and 5 days after infection in A. fumigatus infected mouse cornea and treated with PBS or HNK. In PBS- or HNK-treated A. fumigatus–infected mouse cornea, protein expressions of TLR-2, TNF-α, IL-1β, and HMGB1 were detected using Western blot or ELISA. Representative examples of TLR-2 Western blot (C) and TNF-α Western blot (D) at 3 and 5 days after infection. ELISA results of IL-1β (G) and HMGB1 (H) at 3 and 5 days after infection. Data were mean + SEM analyzed using an unpaired, two-tailed Student’s t-test.

Discussion

A. fumigatus is one of the most common Aspergillus species causing FK in agricultural districts.^34,35 Compared with bacterial ocular infections, presentations of A. fumigatus keratitis tend to be more severe because of strong fungal invasiveness and virulence,³⁶ lack of effective antifungal agents, and long-term use of antibiotic-induced drug resistance. These issues suggest the need to develop more effective antifungal treatments.^35,37,38 HNK is extracted from magnolia bark that is a traditional Chinese herbal material and has been used for thousands of years for its antioxidant, anti-inflammatory, and antibiotic effects.^8,39 Recently, both in vitro and in vivo studies demonstrated that magnolia bark extract has no mutagenic nor genotoxic potential, and no adverse effect was observed for concentrated magnolia bark extract >240 mg/kg body weight/d (mg/kg b.w/d).⁴⁰ HNK is one of the main substances of magnolia bark.⁴¹ In this study, we identified that HNK had no cytotoxic effects at 8 µg/mL within 24 hours in HCECs, lending support that HNK could be applied at an appropriate concentration and considered nontoxic. Interestingly, we found that HNK treatment could significantly improve the transparency of mouse cornea with A. fumigatus keratitis, relieving the corneal ulcer and protecting cornea integrity. To understand how HNK exhibits the corneal protective effects, we sought to explore its antifungal and anti-inflammatory mechanisms. 

We first examined the antifungal effects of HNK. We showed that HNK could inhibit the growth of A. fumigatus, which is in agreement with recent studies that HNK inhibited the fungal growth of Alternaria alternata and Fusarium oxysporum.^42,43 Then we further identified that HNK could repress the adhesion ability and biofilm formation of A. fumigatus, which are the critical steps for fungi to infect host cells and resist the host immune system.⁴⁴ Furthermore, our data indicated that HNK increased the membrane permeability of A. fumigatus, which may cause leakage and imbalance of osmotic pressure between intra- and extracellular membranes, resulting in the irreversible damage to the fungal membrane.⁴⁵ Further experimentation could be performed to show which cell membrane component is affected by HNK. In summary, these results demonstrated that HNK has antifungal properties on HCECs through inhibiting the growth, adhesion, biofilm formation, and increasing membrane permeability of A. fumigatus, which, to our knowledge, is the first time that these antifungal effects have been confirmed in FK mice and HCECs. 

In addition, we tested whether HNK would affect inflammatory response. Innate immune response acts as the first line against fungal infection,¹⁹ however, excessive inflammatory response may contribute to the poor prognosis of FK because of the cornea damage caused by the higher production of proinflammatory factors, cytokines, and oxidative stress.^23,46,47 Neutrophils are professional phagocytes playing critical roles in fungal infections through various neutrophil pathogen recognition receptors (PRRs), signaling transductions, and cytotoxicity.⁴⁸ Still, excessive neutrophil presence can lead to excess monocyte and macrophage recruitment and infiltration, in turn releasing numerous proinflammatory cytokines and chemokines to cause corneal damage.⁴⁹ Studies have shown HNK has anti-inflammatory activity,^50,51 and it is reported that HNK not only protects rat brain from focal cerebral ischemia-reperfusion injury through inhibiting neutrophil infiltration and reactive oxygen species (ROS) production,⁵² but also relieves acute pancreatitis via inhibiting expression of TLRs (2/4) and inflammatory mediators, such as TNF-α, IL-1β, and IL-6.^24,53 Thus we investigated whether HNK would influence neutrophil infiltration, as well as the expression of PRRs and proinflammatory mediators in A. fumigatus keratitis. Our data suggested that HNK attenuated neutrophil vitality and infiltration to the infected area in vitro and in vivo. Moreover, HNK inhibited the expression of TLR-2 and inflammatory cytokines, including TNF-α, IL-1β, and HMGB1, in A. fumigatus HCECs. TLR-2 and TLR-4 activation results in neutrophil activation and the production of ROS, promoting the continuation of the inflammation cycle.^54,55 HNK represses the expression of TLR-2 so that it in turn inhibits neutrophil activation,⁵⁶ which is consistent with our result that showed HNK suppressed neutrophil infiltration. This result supports our hypothesis that HNK can inhibit inflammatory response in A. fumigatus keratitis. However, the underlying mechanism of the inhibitory effects is still unclear. A few studies suggested HNK could bind to and enhance the activity of deacetylase SIRT3, a class III histone deacetylase predominantly located in mitochondria,⁵⁰ and SIRT3 could diminish inflammation and mitigate endotoxin-mediated acute lung injury,⁵⁷ which indicates HNK may exhibit its anti-inflammatory effects via mitochondrial function. Further study will be focusing on the exact anti-inflammatory mechanisms of HNK. 

Conclusions

Our study demonstrated that HNK has antifungal and anti-inflammatory activities in A. fumigatus keratitis mice and HECEs through inhibiting the growth/adhesion/biofilm formation of A. fumigatus, increasing fungal membrane permeability, and repressing the expression of TLR-2 and inflammatory cytokines, including TNF-α, IL-1β, and HMGB1. Thus HNK, as a potent antifungal and anti-inflammatory molecule for A. fumigatus keratitis, could be a novel therapeutic agent used for FK treatment. 

Acknowledgments

Supported by the National Natural Science Foundation of China (No. 81470609; No. 81870632; No. 81500695), and the Natural Science Foundation of Shandong Province (No. ZR2019BH004). 

Disclosure: L. Zhan, None; X. Peng, None; J. Lin, None; Y. Zhang, None; H. Gao, None; Y. Zhu, None; Y. Huan, None; G. Zhao, None