Flavopiridol Protects Against Fungal Keratitis by Alleviating Inammation Through the Promotion of Autophagy

Background: Fungal keratitis is a serious infectious keratopathy related to fungal virulence and excessive inammatory responses. Autophagy exhibits a potent ability to resolve inammation during fungal infection. This study aimed to investigate the protective function of avopiridol in fungal keratitis and explore its effects on autophagy. Methods: A mouse model of fungal keratitis was established and then treated with 5 μM avopiridol. RAW 264.7 cells were treated with 200 nM avopiridol before fungal stimulation. The severity of corneal diseases was evaluated by slit-lamp microscopy. The expression levels of cytokines were detected by RT-PCR and ELISA. The protein levels of LC3, Beclin-1 and Atg7 were determined by western blot and immunouorescence. A Cell Counting Kit-8 assay was used to test cell viability. Autolysosomes were detected by transmission electron microscopy (TEM). An inhibitor of autophagy, 3-methyladenine (3-MA), was used to pretreat RAW 264.7 cells. Phagocytosis of RAW 264.7 cells was evaluated by counting colony forming units. A. fumigatus was incubated with avopiridol, and the hyphae were stained with calcouor white. Absorbance assay, crystal violet staining and adherence assay were used to detect the antifungal activity of avopiridol. Results: Flavopiridol treatment notably reduced corneal opacity and the clinical scores of infected corneas. Compared with DMSO treatment, avopiridol treatment greatly downregulated IL-1β, IL-6 and TNF-a expression in infected corneas. In RAW 264.7 cells, avopiridol treatment inhibited IL-1β, IL-6 and TNF-a expression but promoted IL-10 expression. TEM images showed that more autolysosomes were presented in infected corneas and RAW 264.7 cells after avopiridol treatment than after DMSO treatment. Flavopiridol treatment notably upregulated the protein expression of LC3, Beclin-1 and Atg7 in infected corneas as well as in RAW 264.7 cells. 3-MA pretreatment counteracted the cytokine regulation induced by avopiridol. Moreover, avopiridol promoted the phagocytosis of RAW 264.7 cells. Flavopiridol also exhibited antifungal activity by restricting fungal growth and limiting fungal biolm formation and conidial adhesion. Conclusions: Flavopiridol signicantly alleviated the inammation of fungal keratitis by activating autophagy. In addition, avopiridol promoted the phagocytosis of RAW 264.7 cells and exhibited antifungal function, indicating the potential therapeutic role of avopiridol in fungal keratitis. protein of LC3B II, Beclin-1 and Atg-7 infected corneas and RAW 264.7 cells, avopiridol upregulated autophagy in fungal keratitis. Autophagy has been considered as a signicant player in regulating inammation during fungal infection. In our study, 3-MA, an inhibitor of autophagy, was used to pretreat RAW 264.7 cells before avopiridol treatment. 3-MA could specically block the fusion of autophagosomes and lysosomes, inhibiting the formation of autolysosomes. Our study showed that 3-MA pretreatment signicantly counteracted the downregulation of TNF-a and upregulation of IL-10 induced by avopiridol in infected RAW 264.7 cells. Thus, avopiridol exerted anti-inammatory effects by inducing autophagy in fungal keratitis. In addition to autophagy, macrophages also exhibit phagocytosis ability, which can directly phagocytose fungi and participate in the innate immune process of fungal keratitis. Our results indicated that avopiridol treatment increased macrophage phagocytosis, contributing to the elimination of fungi.


Introduction
Fungal keratitis is a serious infectious keratopathy with a high incidence of vision loss, which is mainly caused by Fusarium and Aspergillus (Huang et al. 2016, Mahmoudi et al. 2018). Corneal trauma after agricultural injury and contact lens usage have been recognized as the predominant risk factors (Mahmoudi et al. 2018). The severity of fungal keratitis is related to fungal virulence and excessive in ammatory responses (Niu et al. 2019). However, due to the insu ciency and severe side effects of antifungal drugs, fungal keratitis has a poor prognosis. Thus, it is urgent to explore the novel approaches to treat fungal keratitis.
Currently, CDK9 inhibitors are considered potential treatments for in ammatory diseases (Hou et al. 2007). In vivo, CDKi, such as avopiridol and R-roscovitine, exert anti-in ammatory and pro-resolution functions in pleurisy, hepatitis and pneumonia (Schmerwitz et al. 2011, Leitch et al. 2009, Rossi et al. 2006, Hoogendijk et al. 2012). In vitro, avopiridol suppresses the interaction between leukocytes and endothelial cells in the model of murine hepatitis, which is a crucial step in the initiation of in ammation (Schmerwitz et al. 2011). In addition, CDKi exerts antiangiogenic functions in corneal diseases (Liebl et al. 2011). However, the anti-in ammatory function of avopiridol has never been explored in fungal keratitis.
Although current studies have noted that avopiridol inhibits the expression of in ammatory cytokines mainly through regulating the CDK9 and NF-kB signaling pathway, the speci c anti-in ammatory mechanism of avopiridol still remains unclear (Srikumar et al. 2016). Emilia Mahoney et al. demonstrated that avopiridol could induce autophagy by increasing LC3 cytoplasmic accumulation and SQSTM1/p62 expression (Mahoney et al. 2013). Autophagy is a kind of degradation that removes longlived proteins and organelles and is mediated by lysosomes. Autophagy is induced by various stimuli including starvation, stress, hypoxia and infection (Feng et al. 2014). Autophagy exhibits a potent ability to resolve in ammation, which is associated with the limitation and degradation of in ammasomes and cytokines (Galluzzi et al. 2017, Qian et al. 2017, Renga et al. 2018). These previous results implied that autophagy may become a novel target for avopiridol to prevent in ammation. Macrophage autophagy can effectively degrade infectious microorganisms, thereby eliminating infection (Chai et al. 2016). A previous study has indicated that the promotion of autophagy could reduce the severity of fungal keratitis by regulating cytokines production ). Thus, we hypothesized that avopiridol may exhibit a protective role in fungal keratitis by affecting autophagy.
Moreover, avopiridol has been considered as an inhibitor of UDP-galactopyranose mutase (UGM) in A. fumigatus, suggesting its potential antifungal effect (Martín et al. 2017, Hostetter et al. 1994). UGM is a crucial enzyme for fungal cell wall biogenesis, which promotes the transformation from UDPgalactofuranose (Galf) to UDP-galactopyranose (Galp) (Latgé 2010). The fungal cell wall is involved in the interaction between pathogens and hosts, which determines the pathogenicity of fungal keratitis (Hostetter 1994). As an inhibitor of UGM, avopiridol may inhibit the growth and virulence of fungi by restricting fungal cell wall formation. In addition, avopiridol has been recognized as a candidate in infectious diseases (Ou et al. 2013, Nelson et al. 2001). Various studies have indicated that avopiridol inhibits viral transcription without cytotoxicity and alleviates virus-associated diseases, including adenoviral epidemic keratoconjunctivitis (Ou et al. 2013, Nelson et al. 2001).
Our study demonstrated that avopiridol exerted anti-in ammatory activity by promoting autophagy in fungal keratitis. Additionally, avopiridol could improve the phagocytosis ability of macrophages and restrict the growth of fungi. These data suggest that avopiridol may exert therapeutic effects in fungal keratitis.

Materials And Methods
Mice corneal infection 8-week-old C57BL/6 female mice were provided by Jinan Pengyue corporation (Jinan, China). After anesthetization, the 2 mm-diameter central corneal epithelium of mice left eyes was scraped. 10 8 colony forming units (CFU) /mL A. fumigatus (5 μL) was added on the surface of mouse corneas and then covered with a contact lens. All animal experiments abided by the ARVO Statement for the Use of Cell Counting Kit-8 assay (MCE) was used to detect the cell viability after incubation with avopiridol.

RT-PCR
Total RNA was extracted from mouse corneas and RAW 264.7 cells. The PCR method was described previously [25] . The primers used are exhibited in Table 1. Transmission electron microscopy RAW 264.7 cells were treated with DMSO or avopiridol for 2 hours, and then stimulated by fungi for 8 hours. Then cells were scraped off and placed in a microcentrifuge tube. After centrifuged at the speed of 3000 rpm for 10 minutes, cell precipitate was obtained and xed in 2.5% glutaraldehyde. Mouse corneas were removed and xed in 2.5% glutaraldehyde. The preparation steps of samples before transmission electron microscopy (TEM) observation were described in the previous study (Liu et  A. fumigatus growth analysis A. fumigatus strain 3.0772 was provided by The China General Microbiological Culture Collection Center (China). A. fumigatus was incubated with DMSO and avopiridol (200, 400 and 800 nM; MCE) in 96-well plate for 1, 2, 3 ,4 and 5 days. The absorbance of fungi was measured at 540 nm. Fungi was also stained with Calco uor white (Sigma, Santa Clara, CA, USA) for 10 minutes. The images of stained-hyphae were captured and the uorescence intensity was measured.
Fungal bio lm formation assay The preparation of fungal bio lm formation has been described in previous publications (Wiederhold et al. 2018). The fungal bio lms were incubated with DMSO and avopiridol (200, 400 and 800 nM) for 48 hours. Bio lms were xed with methanol and stained by 0.1% crystal violet. After washed for three times, ethanol was used to release crystal violet bounded to bio lms. The absorbance was detected at 570 nm three times.

Fungal adherence assay
HCECs (2×10 4 /mL) were treated with the mixture of conidia suspension (at an MOI of 10) and 200 nM avopiridol or DMSO, and plated on chambered slides (Cameron et al. 1988). After incubation for 3 hours at 37°C, hematoxylin and eosin (HE) were used to stain the conidia and cells. The images of adherent conidia to HCECs were captured by microscopy (Thermo Fisher Scienti c, 600×).

Statistical analysis
Student's t-test was used to analyze differences between two groups. One-way ANOVA was used to evaluate differences among three or more groups. Differences were considered signi cant at P ≤ 0.05. All data are shown as the mean ± SEM.

Flavopiridol induced autophagy in fungal keratitis
To detect the effect of avopiridol on autophagy in corneal fungal infection, autolysosomes were observed by TEM. No autolysosomes were observed in images of normal mouse cornea (Fig. 3a) and normal mitochondria was observed at higher magni cation (Fig. 3d). Few autolysosomes could be seen in infected cornea after DMSO treatment at 3 days p.i. (Fig. 3b). While, the cristae of mitochondria decreased or disappeared in infected cornea (Fig. 3e). As shown in Figs. 3c and 3f, more autolysosomes were formed and the number of mitochondria cristae increased in infected corneas after avopiridol treatment.
To further explore the effects of avopiridol on autophagy, the expression of LC3B , LC3B , Beclin-1 and Atg-7 in infected corneas was detected. DMSO-treated corneas were set as the control group. Compared with control corneas, LC3B , Beclin-1 and Atg-7 protein expression was increased in infected corneas. Flavopiridol treatment further elevated LC3B , Beclin-1 and Atg-7 protein expression in infected corneas, compared with DMSO treatment (Figs. 3g-3l).
Flavopiridol induced autophagy in RAW 264.7 cells Autolysosomes in RAW 264.7 cells were detected by TEM. Fig. 4a displays the TEM image of a normal RAW 264.7 cell. After fungal stimulation for 8 hours, phagocytosed fungus and autolysosomes could be seen in the cytoplasm of RAW 264.7 cell (Fig. 4b). In the avopiridol group, RAW 264.7 cells were incubated with 200 nM avopiridol and then infected with A. fumigatus for 8 hours. Phagocytosed fungus (Fig. 4c) and an increased number of autolysosomes (Fig. 4d) were observed in RAW 264.7 cell in the avopiridol group. To investigate the effects of avopiridol on autophagy, the protein expression levels of LC3, Beclin-1 and Atg-7 were detected. LC3 expression in RAW 264.7 cells was detected by immuno uorescence staining (Fig. 4e). DMSO-treated cells were set as the control group. The images showed that minimal LC3 was expressed in the control group. The immuno uorescence of LC3 was increased in infected RAW 264.7 cells after DMSO treatment and was mainly distributed in the cytoplasm. Compared with DMSO treatment, avopiridol treatment further improved LC3 expression in infected RAW 264.7 cells. Next, Beclin-1 and Atg-7 protein expression levels in RAW 264.7 cells were measured. Western blot results showed that, fungal stimulation upregulated the protein levels of Beclin-1 (Figs. 4f, 4g; P 0.05) and Atg-7 (Figs. 4h, 4i; P 0.001) in RAW 264.7 cells. In addition, the protein levels of Beclin-1 (Figs. 4f, 4g; P 0.05) and Atg-7 (Figs. 4h, 4i; P 0.05) were both further elevated after avopiridol treatment in infected cells. Immunostaining results demonstrated that both Beclin-1 (Fig. 4j) and Atg-7 (Fig. 4k) expression increased after A. fumigatus stimulation. Flavopiridol treatment further notably enhanced the uorescence of Beclin-1 and Atg-7, compared with DMSO treatment.

Autophagy inhibition restricted the anti-in ammatory activity of avopiridol
To demonstrate the role of autophagy during the anti-in ammatory process of avopiridol, 3-MA, an inhibitor of autophagy, was used to pretreat RAW 264.7 cells before avopiridol treatment and A. fumigatus stimulation. The mRNA expression and protein levels of cytokines were examined by RT-PCR and ELISA. As shown in Figs. 8a and 8b, 3-MA treatment dampened the avopiridol-induced downregulation of TNF-a at both the mRNA ( Fig. 5a; P 0.01) and protein ( Fig. 5b; P 0.05) levels in infected RAW 264.7 cells. In addition, compared with avopiridol treatment, the mRNA ( Fig. 5c; P 0.01) and protein ( Fig. 5d; P 0.001) expression levels of IL-10 were reduced in infected RAW 264.7 cells in the 3-

MA pretreatment group.
Flavopiridol enhanced the phagocytosis of RAW 264.7 cells To assess the effect of avopiridol on the phagocytosis of RAW 264.7 cells, cells were treated with 200 nM avopiridol and an equivalent number of conidia. The ectocytic conidia remained in the supernatant and the number of CFUs was counted (Fig. 6a). Phagocytosis was measured by the formula described in the methods. Compared with DMSO treatment, avopiridol treatment notably increased the P (120 min) of RAW 264.7 cells ( Fig. 6b; P 0.01).

Flavopiridol suppresses the growth, bio lm formation, and adhesion ability of A. fumigatus
To test the antifungal effects of avopiridol, A. fumigatus was incubated with 0, 200, 400 and 800 nM avopiridol for 1, 2, 3, 4 and 5 days (Fig. 7a). After incubation with avopiridol for 2 days, the absorbance of the medium and the fungal mass were measured. Compared with DMSO treatment, avopiridol signi cantly decreased the absorbance (Fig. 7b; P<0.05; P<0.001; P<0.001). The hyphae were stained with calco uor white, and images of stained hyphae were photographed. The images showed that fewer hyphae were present after 200, 400 and 800 nM avopiridol treatment for 2 days compared to DMSO treatment (Fig. 7e). The fungal mass was quanti ed by assessing the uorescence intensity of the stained hyphae. The fungal mass was reduced after avopiridol treatment ( Fig. 7c; P<0.001; P<0.001; P<0.001). Flavopiridol restricted the bio lm formation of A. fumigatus at 200, 400 and 800 nM (Fig. 7d). In addition, HE staining was used to show the conidia that were adherent to HCECs. Because 200 nM avopiridol exhibited no signi cant cytotoxic effect on HCECs, cells were treated with 200 nM avopiridol. HE staining demonstrated that fewer conidia adhered to cells after avopiridol treatment compared to DMSO treatment (Figs. 7f, 7g). In addition, avopiridol has been considered an inhibitor of UGM, which has antifungal potential (Martín et al. 2017). Thus, we hypothesized that avopiridol might protect against fungal keratitis.

Discussion
In our study, photographs taken by a slit-lamp camera demonstrated infected corneas with reduced edema and opacity following avopiridol treatment. Additionally, the clinical scores were notably decreased after avopiridol treatment. These results indicate that avopiridol exhibits a protective role in fungal keratitis. Compared with DMSO treatment, avopiridol treatment notably downregulated cytokines expression, including IL-1β, IL-6 and TNF-a, in infected corneas. Our results are consistent with ndings of prior studies showing that avopiridol suppressed IL-6 and IL-1β production in postinjury cartilage explants and prevented injury after knee trauma (Hu et al. 2016). Flavopiridol has been veri ed to exert an anti-in ammatory function by reducing the induction and transactivation of cytokines (Hu et al. 2018). In vitro, we further tested the anti-in ammatory function of avopiridol in RAW 264.7 cells. In prior studies, avopiridol was demonstrated to inhibit LPS-induced cytokines production via a MyD88-dependent pathway in RAW 264.7 cells (Haque et al. 2011). Our results showed that avopiridol signi cantly downregulated the levels of pro-in ammatory cytokines such as IL-1β, IL-6 and TNF-a but enhanced the level of the anti-in ammatory cytokine IL-10 in A. fumigatus-stimulated RAW 264.7 cells. Based on these results, avopiridol may represent a novel approach for alleviating in ammation in fungal keratitis.
Previous studies have highlighted the crucial role of autophagy in innate immunity during fungal infection (Li et al. 2020, Kanayama et al. 2015). During infection, autophagy helps to eliminate intracellular pathogens and present antigens, and protects normal cells from being infected (Xian et al. 2016, Lu et al. 2017, Ferreira et al. 2013. At the beginning, phagophores are formed inside the cell. After phagocytizing the damaged organelles and denatured proteins, autophagosomes are formed. Then, autophagosomes bind with lysosomes to form autolysosomes, which help to degrade organelles, proteins and invading microorganisms (Galluzzi et al. 2019). In our study, TEM images exhibited no autolysosome in normal mice cornea or RAW 264.7 cells, while increased autolysosomes were present in infected corneas or RAW 264.7 cells. In addition, the cristae of mitochondria decreased or disappeared in infected corneas. Previous studies showed that avopiridol induced autophagy by upregulating LC3Band downregulating p62 expression, which exerted a cytoprotective role (Wang et al. 2017, Okada et al. 2017, Jeong et al. 2018. Our results showed that avopiridol treatment further increased autolysosomes in infected mouse corneas or RAW 264.7 cells, compared with DMSO treatment. The number of mitochondria cristae were back to normal after avopiridol treatment. During autophagy, LC3-I is modi ed and processed by a ubiquitin-like system including Atg7 and Atg3 to form LC3-II, which locates on the membrane of autophagosome. The expression of LC3 has been considered as a crucial criterion for detecting autophagy. In addition, Beclin-1 and Atg7 are crucial for autophagic responses ). Thus, the expressions of LC3, Beclin-1 and Atg7 were examined in fungal keratitis. Compared with uninfected corneas and RAW 264.7 cells, the protein levels of LC3B II, Beclin-1 and Atg-7 were increased after fungal stimulation. Our results indicated that autophagy is promoted in fungal keratitis, which is consistent to the previous study . We found that avopiridol treatment could further elevate protein levels of LC3B II, Beclin-1 and Atg-7 in infected corneas and RAW 264.7 cells, implying that avopiridol upregulated autophagy in fungal keratitis. Autophagy has been considered as a signi cant player in regulating in ammation during fungal infection. In our study, 3-MA, an inhibitor of autophagy, was used to pretreat RAW 264.7 cells before avopiridol treatment. 3-MA could speci cally block the fusion of autophagosomes and lysosomes, inhibiting the formation of autolysosomes. Our study showed that 3-MA pretreatment signi cantly counteracted the downregulation of TNF-a and upregulation of IL-10 induced by avopiridol in infected RAW 264.7 cells. Thus, avopiridol exerted anti-in ammatory effects by inducing autophagy in fungal keratitis. In addition to autophagy, macrophages also exhibit phagocytosis ability, which can directly phagocytose fungi and participate in the innate immune process of fungal keratitis. Our results indicated that avopiridol treatment increased macrophage phagocytosis, contributing to the elimination of fungi.
In addition, avopiridol exhibits potential antifungal effects. Galactomannan, the production of which is initiated by UGM, is a ubiquitous component in fungal cell walls. The absence of UGM leads to thinner fungal cell wall and reduced fungal virulence (Martín et al. 2017, Schmalhorst et al. 2008). In our study, A. fumigatus was cultivated with different concentrations of avopiridol for 1, 2, 3, 4 and 5 days. The absorbance of the medium was reduced after avopiridol treatment. Fungal mass results and images of immunostained fungi indicated that, avopiridol treatment reduced the amount of A. fumigatus in a concentration-dependent manner. Our results further demonstrated that avopiridol treatment effectively decreased fungal bio lm formation and restricted the adhesion of conidia to HCECs. Bio lm formation and adhesion are crucial for fungi to initiate infection and resist the immune system (Taylor et al. 2014, Beauvais et al. 2015). These results provide evidence that avopiridol exerts antifungal effects in fungal keratitis.
In conclusion, avopiridol protects against fungi-induced corneal damage by suppressing excessive in ammatory responses and limiting fungal growth. Our study demonstrated that avopiridol upregulated autophagy activity and improved the phagocytosis ability of macrophages, contributing to the anti-in ammatory function and protective role of avopiridol in fungal keratitis.

Declarations
Autolysosome was shown in A. fumigatus infected cornea. The white arrow indicates the autolysosome (b). Increased autolysosomes were observed in avopiridol-treated infected cornea. The white arrows denote autolysosomes (c). Normal mitochondria could be observed in the cytoplasm of normal corneal epithelial cell at higher magni cation of Fig. 3a. The black arrows denote the normal mitochondria (d).
The cristae of mitochondria decreased or disappeared in A. fumigatus infected cornea. The black arrows denote the mitochondria (e). Mitochondria cristae were dense in avopiridol-treated infected corneal epithelium. The black arrows denote the mitochondria (f). Flavopiridol induced LC3B , LC3B (g, h), Beclin-1 (i, j) and Atg-7 (k, l) protein expression in infected mouse corneas.    Effects of avopiridol on fungal growth, bio lm formation and adherence. Treatment with different concentrations of avopiridol for 1, 2, 3, 4 and 5 days decreased the absorbance of the medium (a). 200, 400 and 800 nM avopiridol treatment for 2 days notably reduced the absorbance of A. fumigatus (b). The fungal mass was also decreased after 200, 400 and 800 nM avopiridol treatment for 2 days (c).
Flavopiridol signi cantly inhibited fungal bio lm formation at 200, 400 and 800 nM, as determined by analyzing the amount of crystal violet from bio lms (d). Few hyphae were present after high concentrations of avopiridol treatment, as shown in images of stained hyphae. Magni cation: × 100.