Neuroprotective Effect of Cyclo-(L-Pro-L-Phe) Isolated from the Jellyfish-Derived Fungus Aspergillus flavus

Peroxisome proliferator-activated receptor (PPAR) expression has been implicated in pathological states such as cancer, inflammation, diabetes, and neurodegeneration. We isolated natural PPAR agonists—eight 2,5-diketopiperazines—from the jellyfish-derived fungus Aspergillus flavus. Cyclo-(L-Pro-L-Phe) was the most potent PPAR-γ activator among the eight 2,5-DKPs identified. Cyclo-(L-Pro-L-Phe) activated PPAR-γ in Ac2F rat liver cells and SH-SY5Y human neuroblastoma cells. The neuroprotective effect of this partial PPAR-γ agonist was examined using the 3-(4, 5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide assay, lactate dehydrogenase release, and the Hoechst 33342 staining assay in SH-SY5Y cells. Our findings revealed that cyclo-(L-Pro-L-Phe) reduced hydrogen peroxide-induced apoptosis as well as the generation of reactive oxygen species. Rhodamine 123 staining and western blotting revealed that cyclo-(L-Pro-L-Phe) prevented the loss of mitochondrial membrane potential and inhibited the activation of mitochondria-related apoptotic proteins, such as caspase 3 and poly (ADP-ribose) polymerase. Moreover, cyclo-(L-Pro-L-Phe) inhibited the activation and translocation of nuclear factor-kappa B. Thus, the partial PPAR-γ agonist cyclo-(L-Pro-L-Phe) demonstrated potential neuroprotective activity against oxidative stress-induced neurodegeneration in SH-SY5Y cells.


Introduction
Neurodegenerative diseases are a group of heterogeneous disorders characterized by a gradual loss of neuronal structure or function and include conditions such as Alzheimer's disease (AD), Parkinson's disease (PD), amyotrophic lateral sclerosis, and Huntington's disease (HD) [1,2]. Accumulating evidence has indicated the potential role of oxidative stress, mitochondrial dysfunction, inflammation, autophagy, and apoptotic dysfunction in neurodegenerative diseases [3][4][5][6]. Notably, oxidative stress, which induces mitochondrial DNA damage, has been implicated as the primary underlying cause of neurodegenerative diseases, such as AD and PD [7].
Peroxisome proliferator-activated receptors (PPARs) are nuclear receptors that regulate ligand-activated transcription [8] and are known to include three isotypes: PPAR-α, β/δ, and γ [8,9]. PPAR-α is expressed in the brown adipose tissue, liver, kidney, heart, and brain [9,10]. In the brain, PPAR-α upregulates the expression of the gene coding α-secretase, which is known to mediate amyloid precursor protein degradation while it downregulates the expression of the gene coding β-secretase that mainly enhances amyloid-beta (Aβ) peptide release [10]. The inhibition of PPAR-α expression may alter mitochondrial function and suppress antioxidant and anti-inflammatory activities [10]. By contrast, the activation of PPAR-α expression via PPAR-α agonists, gemfibrozil, and WY14643 reportedly ameliorates spatial learning defects and memory impairment in AD mice [11]. PPAR-β/δ is expressed in the gut, kidney, heart, and brain [9,10]. Thus, PPAR-β/δ could mediate antioxidative and anti-inflammatory processes in the damaged brain and PPAR-β/δ deletion was found to induce developmental defects in the mouse brain [12]. In addition, PPAR-β/δ activation by the selective PPAR-β/δ agonist GW0742 exhibits a neuroprotective effect by suppressing inflammation and apoptosis in a mouse model [13]. PPAR-γ is expressed in the adipose tissue, liver, colon, and brain [9,10,14]. Reportedly, PPAR-γ activation stimulates neuronal differentiation and axon polarity [15]. PPAR-γ agonists can decrease the incidence of several neurological disorders [16] and impart protective effects against apoptosis, mitochondrial dysfunction, and oxidative damage [17]. The PPAR-γ agonist pioglitazone was found to improve learning and memory impairment in mice, as well as ameliorate cognitive impairment in diabetic patients with AD [18,19]. In human neural stem cells, rosiglitazone protects cells from Aβ-induced mitochondrial dysfunction and oxidative stress [20]. In a mouse model of traumatic brain injury, rosiglitazone exhibited neuroprotective effects by mediating anti-inflammatory, anti-apoptotic, and anti-oxidative activities [21]. Although several PPAR agonists exhibit neuroprotective effects in neurodegenerative diseases, such as PD, their applications are restricted given the high-dose requirement or toxic side effects [22]. Therefore, new PPAR agonists with fewer side effects need to be developed as potential therapeutic options in neurodegenerative diseases.
Nuclear factor-kappa B (NF-κB) is a transcription factor involved in inflammatory response and apoptosis [23,24]. NF-κB is widely expressed in the central nervous system (CNS) and is associated with IκB in its inactive form [25]. However, in CNS diseases, the NF-κB inhibitor IκB is phosphorylated and degraded following stimulation by several inducers that is followed by nuclear translocation and binding to inflammatory gene response elements [25]. In CNS diseases, PPAR agonists exert benefits by inhibiting the NF-κB pathway (inhibiting the activation of NF-κB or DNA binding of the activated NF-κB) [25]. Reportedly, the PPAR-γ agonist pioglitazone decreases NF-κB activation in a 6-hydroxydopamine induced PD model [26].

Protective Effects of Compound 8 against H 2 O 2 -Induced Cell Injury
Activation of PPAR reportedly demonstrates beneficial effects in neurodegenerative diseases and CNS injury [53]; accordingly, ligands targeting PPARs are considered potential therapeutics in these pathologies. Thyrotropin-releasing hormone (TRH), which is a neural tripeptide amide, was first characterized in the hypothalamus and afforded neuroprotective effects in CNS trauma [54,55]. In addition to its neuroprotective effect, TRH exhibits physiological effects that may be undesirable for the treatment of neurotrauma [56]. Metabolic products of TRH, such as cyclo-(His-Pro) (CHP), protect cells against H 2 O 2 -induced injury by inhibiting oxidative stress [57]. The synthetic CHP mimetics, cyclo-[(R)-3 ,5 -di-tertbutyl-Tyr-L-Pro] and cyclo-[(S)-3 ,5 -di-tert-butyl-Tyr-L-Pro], reportedly inhibited neuronal cell death in a traumatic injury model [58]. Based on the common structural features of these neuroprotective compounds, a pharmacophore model was generated to assess neuroprotective effects of DKPs [59] ( Figure S2A). Compound 8 was found to possess common neuroprotective structural features and was mapped onto the active pharmacophore model ( Figure S2B). Therefore, compound 8 may be worth investigating for its potential neuroprotective effects.
The neuroprotective effect of 8 was evaluated using an H 2 O 2 -induced SH-SY5Y cell injury model, compared with the positive control rosiglitazone [21,60]. Before performing the neuroprotection assay, SH-SY5Y cells were exposed to various concentrations of H 2 O 2 , 8, and rosiglitazone to determine appropriate concentrations for the assay. Concentrations approximating the IC 50 values of H 2 O 2 (650 µM) and the non-cytotoxic concentrations (10, 20, and 40 µM) of 8 and rosiglitazone were selected to perform the neuroprotection assay ( Figure 5A,B). Pretreatment with 8 induced a dose-dependent increase in cell viability up to 66.4%, 74.6%, and 80.4% (at 10, 20, and 40 µM, respectively), revealing a potency higher than that of rosiglitazone ( Figure 5C). In addition, a lactate dehydrogenase (LDH) release assay was performed to demonstrate this protective effect. The pretreatment with 8 or rosiglitazone decreased the H 2 O 2 -induced cytotoxicity to 45.9% and 44.8%, respectively, at a concentration of 40 µM ( Figure 5D).
The blood-brain barrier (BBB) is a highly selective semipermeable barrier comprising endothelial cells, which prevents solutes in circulating blood from non-selectively crossing into the CNS where neurons reside [61]. As a possible neuroprotective agent, the BBB permeability of 8 was predicted using PreADMET [62] and the brain to the blood concentration ratio of 8 was determined as 0.558621 ( Figure S3A); this was higher than that of CHP (0.140492) ( Figure S3B). CHP reportedly accumulates in the CNS regardless of its low entry rate owing to its long half-life and marked resistance to enzymatic degradation [63]. Therefore, compound 8 could maintain a precise CNS concentration owing to its high stability against enzymatic hydrolysis [32].
Mitochondrial membrane potential (MMP, ∆Ψm) can regulate matrix configuration and cytochrome C release and MMP levels are reduced during apoptosis [66]. MMP loss is considered to induce cell death by damaging the mitochondria [67,68]. Rhodamine 123 (Rho 123) is employed as a probe to monitor MMP; the Rho 123 fluorescence decay rate corresponds to the MMP [69]. As shown in Figure 6C,D, the Rho 123 fluorescence intensity was significantly reduced in the H 2 O 2 -treated group; however, pretreatment with 8 (10, 20, and 40 µM) and rosiglitazone (40 µM) inhibited MMP loss. In addition, the effect of 8 was more potent than that of rosiglitazone at the same treatment concentration (40 µM). These results indicated that compound 8 suppressed H 2 O 2 -induced apoptosis in SH-SY5Y cells.

Effects of Compound 8 on H 2 O 2 -Induced Oxidative Stress
Oxidative stress generates reactive oxygen species (ROS) and disrupts mitochondrial membrane permeability and mitochondrial defense systems; theses are known features that underlie the development of neurodegenerative diseases [70]. Apart from the generation of endogenous ROS, the mitochondria also act as a ROS target via feedback [71]. Oxidative stress directly targets mitochondria to induce apoptotic cell death [71]. As shown in Figure 7A,B, pretreatment with 8 (10, 20, and 40 µM) or rosiglitazone (40 µM) decreased H 2 O 2 -induced [72] ROS generation in SH-SY5Y cells.
Superoxide dismutase (SOD) is a metalloenzyme that plays a vital role against oxidative stress in the body [73]. SOD scavenges ROS to attenuate cell death [72]. Catalase (CAT) is the second most abundant enzymatic antioxidant that decomposes ROS [74,75]. Both SOD and CAT are the first lines of defense against free radical-induced tissue damage [76]. The treatment of SH-SY5Y cells with 8 (10, 20, and 40 µM) or rosiglitazone (40 µM) increased SOD and CAT enzyme levels but the activity was not significant (Figure 7C,D). These results suggested that compound 8 could suppress H 2 O 2 -induced oxidative stress by attenuating ROS generation in SH-SY5Y cells.

Effects of Compound 8 on H 2 O 2 -Induced Apoptosis-Related Proteins
Under oxidative stress, ROS-induced cell death is reportedly associated with caspaseactivated apoptosis [72]. Activation of caspases is related to mitochondria-dependent apoptosis [71]. Morphological changes in mitochondria and ROS generation are mediated via caspase 9 [77]. Caspase 9 activates caspase 3, which is essential for brain development and contributes to apoptosis [77,78]. In addition, caspase 3 is responsible for the cleavage and activation of poly (ADP-ribose) polymerase (PARP), which activates DNA strand breakage [79]. In addition to caspase 3, caspase 9 also activates caspase 7 [77,80], which is a pivotal mediator of MMP loss (∆Ψm). Procaspases are inactive zymogens that need to be activated through cleavage [81]. We measured protein levels of cleaved-caspase 3, 7, and 9 and cleaved-PARP. As shown in Figure 8A-E, the ratio of cleaved-caspase 3 and cleaved-PARP to their inactive zymogens was significantly decreased when treated with 8 or rosiglitazone. Even though the error range of the data of caspases 7 and 9 ( Figure 8C,D) was large, the trend of caspase inhibition by compound 8 can be observed. These results indicated that compound 8 could reduce the protein levels of cleaved-caspase 3 and cleaved-PARP in H 2 O 2 -induced damage in SH-SY5Y cells.

Effects of Compound 8 on NF-κB Activation and Nuclear Translocation
In patients with PD, elevated nuclear translocation of NF-κB has been observed in dopaminergic neurons [82]. In Aβ 25-35 -exposed rats (experimental AD model), IκB-α degradation was found to be enhanced; however, the neuroprotective agent sodium hydrosulfide, which enhances protein levels of PPAR-α and PPAR-γ, can block IκB-α degradation (i.e., NF-κB activation) [83]. NF-κB, which is a crucial mediator of host defense against pathogens, is activated by various stimuli, such as inflammatory factors or oxidants [84]. After activation of latent NF-κB in the cytoplasm, the NF-κB complex is translocated into the nucleus, thereby promoting NF-κB-regulated gene expression [84,85]. PPAR-γ can block tissue injury by suppressing the NF-κB pathway to decrease inflammation while promoting the nuclear factor erythroid 2-related factor 2 (Nrf2)/antioxidant response element (ARE) axis to reduce oxidative stress [86]. As an oxidant, H 2 O 2 can promote NF-κB p65 activation and nuclear translocation in SH-SY5Y cells [87]. In the present study, western blot and immunofluorescence assays were performed to measure the activation and endonuclear translocation of NF-κB p65 after treatment with 8. As shown in Figure 9C, the immunofluorescent staining assay revealed that 8 suppressed H 2 O 2 -induced NF-κB activation and endonuclear translocation. For further confirmation, the Western blot assay was performed and H 2 O 2 induced the nuclear translocation of NF-κB p65, but 8 (10, 20, and 40 µM) and rosiglitazone (40 µM) decreased the nuclear protein level of NF-κB ( Figure 9A,B). These results suggested that compound 8 inhibited NF-κB activation and its translocation through the activation of PPAR-γ.

Isolation of 2,5-DKPs
The ethyl acetate extract of the jellyfish-derived fungus A. flavus was subjected to ODS column and HPLC separation to obtain eight 2,5-DKPs (1-8) (Figure 1). It was not requisite to acquire the approval of ethical commission for the isolation of the fungus from the jellyfish Aurelia aurita.

Molecular Docking Study
The crystal structures of PPAR-α, β/δ, and PPAR-γ with PDB codes 4BCR, 5U46, and 2PRG were downloaded from the Protein Data Bank [88]. Proteins were prepared using the Chimera 1.10.2 software package (National Institutes of Health, Bethesda, MD, USA) [89]. Ligand preparation and the addition of polar hydrogen, Kollman charges, setting grid box parameters for proteins, and docking calculations were performed using AutoDockTools 1.5.6 (The Scripps Research Institute, La Jolla, CA, USA) and AutoDock Vina 1.1.2 (The Scripps Research Institute) [90]. Discovery Studio 4.5 (NeoTrident Technology Ltd., Beijing, China) [91] was used to analyze the protein-ligand interactions.

LDH Release
The SH-SY5Y cells were seeded into 96-well plates with the background control (contained the assay medium), low control (spontaneous LDH release), and high control

Luciferase Assay
The SH-SY5Y and Ac2F cells were seeded in 48-well plates. When cell density reached 90% confluence, the plasmids pcDNA3, TK-PPRE, PPAR-α, β/δ, and PPAR-γ were transfected into cells using the free medium for 4 h (Ac2F) or 24 h (SH-SY5Y) (this experiment was performed as described in our previous report) [92]. After treatment, the free medium was removed and the cells were incubated with complete medium overnight. Transfected cells were treated with 1-8, WY-14643, GW501516, or rosiglitazone for 6 h (Ac2F) or 24 h (SH-SY5Y). The cells were lysed and data values were measured using the ONE-Glo™ Luciferase Assay System regent with GloMax ® -Multi Microplate Multimode Reader (Promega Co., Madison, WA, USA).

Hoechst 33342 Staining
The SH-SY5Y cells were seeded in confocal dishes and incubated overnight. The cells were pretreated with 8 and rosiglitazone for 10 h and then exposed to H 2 O 2 for 14 h. After treatment, the cells were fixed with 10% formalin solution for 15 min and stained with Hoechst 33342 reagent (10 µg/mL) for 20 min in the dark. The cells were washed three times with phosphate-buffered saline (PBS) and then visualized using a ZEISS LSM 800 confocal microscope (Oberkochen, Baden-Württemberg, German).

MMP Assay
The SH-SY5Y cells were seeded in confocal dishes and treated with 8 (10, 20, and 40 µM) and rosiglitazone (40 µM) for 10 h and then exposed to H 2 O 2 for another 14 h. Next, the cells were stained with Rho 123 (10 µg/mL) for 20 min in the dark and analyzed using a ZEISS LSM 800 confocal microscope (Oberkochen, Baden-Württemberg, German) at 529 nm. The mean fluorescence intensity was quantified using ImageJ (National Institutes of Health, Bethesda, MD).

ROS Generation
The SH-SY5Y cells were seeded in confocal dishes and treated as described above. After treatment, the cells were washed with PBS and stained with DCFDA (5 µM) in free medium for 30 min in the dark. Finally, the cells were washed with PBS thrice and then analyzed using a ZEISS LSM 800 confocal microscope (Oberkochen, Baden-Württemberg, German) at 525 nm.

SOD and CAT Activities
The SH-SY5Y cells were seeded into 6-well plates and pretreated with 8 at 10, 20, and 40 µM and rosiglitazone (40 µM) for 10 h, followed by treatment with 650 µM H 2 O 2 for an additional 14 h. After treatment, the cells were collected and lysed in the lysis buffer for 30 min on ice. The SOD and CAT activities were measured according to the manufacturer's instructions of EZ-Catalase assay kit (DoGenBio Co., Ltd., Seoul, Korea) and the Superoxide dismutase assay kit was used (DoGenBio Co., Ltd., Seoul, Korea). SOD inhibition activity was determined at 450 nm using a microplate reader (Elx 800, Bio-Tek, Winooski, VT, USA). CAT reacted with H 2 O 2 to produce water and oxygen and the unconverted H 2 O 2 reacted with OxiRed TM to generate a product measured at 570 nm using a microplate reader (Elx 800, Bio-Tek, Winooski, VT, USA).

Western Blotting
The SH-SY5Y cells were seeded on cell culture dishes and pretreated with 8 at 10, 20, and 40 µM and rosiglitazone (40 µM) for 10 h, followed by treatment with 650 µM H 2 O 2 for an additional 14 h. After treatment, cells were collected and washed with PBS. Cell lysis buffer was added to the cell pellet to lyse the cells for 30 min on ice. The lysed cells were centrifuged at 13,000 rpm for 15 min and the supernatant protein concentration was measured using a BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA). Proteins were loaded and separated using sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Then, proteins were transferred to PVDF membranes and blocked with 5% non-fat milk for 1 h at room temperature (25 • C). Next, the membranes were incubated with primary antibodies (cleaved-caspase 3, 7, and 9; cleaved-PARP; caspase 3, 7, and 9; and PARP) overnight. The membranes were washed three times with TBST and then incubated with secondary antibodies for 1 h at room temperature (25 • C). Finally, the membranes were washed three times with TBST and visualized using an ECL kit using the ChemiDoc™ Touch Imaging System (Bio-Rad Laboratories, Hercules, CA, USA).

Immunofluorescence Assay
The SH-SY5Y cells were seeded on confocal dishes and pretreated with 8 at 10, 20, and 40 µM and rosiglitazone (40 µM) for 10 h, followed by treatment with 650 µM H 2 O 2 for another an additional 14 h. Then, the cells were fixed with 10% formalin solution and treated with 0.3% Triton X-100 for 15 min. The fixed cells were blocked with 10% BSA for 30 min at room temperature and incubated with the primary antibody anti-NF-κB overnight. The cells were washed three times with PBS and incubated with Alexa 488 secondary antibodies for 30 min at room temperature. Finally, 10 µg/mL of propidium iodide and 10 µg/mL RNase were added to confocal dishes and cultured for 30 min individually. The fluorescence of the SH-SY5Y cells was analyzed using a ZEISS LSM 800 confocal microscope (Oberkochen, Baden-Württemberg, German).

Statistical Analysis
Data analyses were performed using GraphPad Prism 5 (San Diego, CA, USA). Data values are presented as means ± standard error of the mean. One-way analysis of variance and Tukey's HSD-post hoc test were used to analyze significant differences. * p < 0.05, ** p < 0.01, and *** p < 0.001 were used to determine statistical significance.

Conclusions
In the course of our search for natural PPAR agonists, eight 2,5-DKPs (1-8) were isolated from the jellyfish-derived fungus A. flavus. Compound 8 was selected as a partial PPAR-γ agonist and evaluated for neuroprotective effect using SH-SY5Y neuroblastoma cells. Compound 8 showed inhibition of H 2 O 2 -induced cell injury and ROS generation in SH-SY5Y cells, together with inhibition of H 2 O 2 -induced apoptosis and the loss of mitochondrial membrane potential. The activation of the apoptosis-related proteinscaspase 3 and PARP-was inhibited by 8. In addition, compound 8 inhibited H 2 O 2 -induced activation and endonuclear translocation of NF-κB, which is a key physiological marker in patients with PD and experimental AD models. Therefore, compound 8, which is a partial PPAR-γ agonist, was proposed to exert neuroprotective effects by modulating the NF-κB pathway. According to the in vitro results, compound 8 may be utilized as a partial PPAR-γ agonist for in vivo study in neurodegenerative diseases models.

Conflicts of Interest:
The authors declare no conflict of interest.