**The E**ff**ect of Nano-Epigallocatechin-Gallate on Oxidative Stress and Matrix Metalloproteinases in Experimental Diabetes Mellitus**

**Adriana Elena Bulboaca 1, Paul-Mihai Boarescu 1,\*, Alina Silvia Porfire 2,\*, Gabriela Dogaru 3,\*, Cristina Barbalata 2, Madalina Valeanu 4, Constantin Munteanu 5, Ruxandra Mioara Râjnoveanu 6, Cristina Ariadna Nicula <sup>7</sup> and Ioana Cristina Stanescu <sup>8</sup>**


Received: 11 January 2020; Accepted: 18 February 2020; Published: 20 February 2020

**Abstract:** Background: The antioxidant properties of epigallocatechin-gallate (EGCG), a green tea compound, have been already studied in various diseases. Improving the bioavailability of EGCG by nanoformulation may contribute to a more effective treatment of diabetes mellitus (DM) metabolic consequences and vascular complications. The aim of this study was to test the comparative effect of liposomal EGCG with EGCG solution in experimental DM induced by streptozotocin (STZ) in rats. Method: 28 Wistar-Bratislava rats were randomly divided into four groups (7 animals/group): group 1—control group, with intraperitoneal (i.p.) administration of 1 mL saline solution (C); group 2—STZ administration by i.p. route (60 mg/100 g body weight, bw) (STZ); group 3—STZ administration as before + i.p. administration of EGCG solution (EGCG), 2.5 mg/100 g b.w. as pretreatment; group 4—STZ administration as before + i.p. administration of liposomal EGCG, 2.5 mg/100 g b.w. (L-EGCG). The comparative effects of EGCG and L-EGCG were studied on: (i) oxidative stress parameters such as malondialdehyde (MDA), indirect nitric oxide (NOx) synthesis, and total oxidative status (TOS); (ii) antioxidant status assessed by total antioxidant capacity of plasma (TAC), thiols, and catalase; (iii) matrix-metalloproteinase-2 (MMP-2) and -9 (MMP-9). Results: L-EGCG has a better efficiency regarding the improvement of oxidative stress parameters (highly statistically significant with *p*-values < 0.001 for MDA, NOx, and TOS) and for antioxidant capacity of plasma (highly significant *p* < 0.001 for thiols and significant for catalase and TAC with *p* < 0.05). MMP-2 and -9 were also significantly reduced in the L-EGCG-treated group compared with the EGCG group (*p* < 0.001). Conclusions: the liposomal nanoformulation of EGCG may serve as an adjuvant therapy in DM due to its unique modulatory effect on oxidative stress/antioxidant biomarkers and MMP-2 and -9.

**Keywords:** epigallocatechin-gallate; liposomes; diabetes mellitus; oxidative stress

#### **1. Introduction**

Consuming green tea has been linked to human health and longevity for centuries. In particular, green tea catechins are involved in many biological processes such as antioxidant activity and modulation of various cellular lipid and protein metabolisms [1]. Green tea contains a great amount of polyphenols (flavonols, flavones, and flavanols) with similar structure, possessing lots of therapeutic active components including catechin, epicatechin, epicatechin-3-gallate, and epigallocatechin-3-gallate (EGCG) [2]. EGCG is the most active and abundant compound (65% of total catechin content) [3,4].

Green tea therapeutic effects have been studied intensively, proving beneficial in various diseases such as cancer [5], hyperlipidemia [6,7], cardiovascular diseases [8,9], neurodegenerative diseases [10,11], and infectious diseases [12,13]. Some reports also suggest that daily consumption of tea catechins may help in controlling type 1 [14] and type 2 diabetes mellitus [1]. It has been demonstrated that green tea consumption reduces fasting glucose levels, an effect mediated by EGCG [15]. Lipophilic EGCG has been shown to reduce glycemia and serum lipids in experimental diabetes mellitus induced by streptozotocin (STZ) in rats [16].

Type 1 diabetes mellitus (DM) is associated with an autoimmune-mediated destruction of pancreatic beta cells, leading to absolute insulin deficiency [17]. One of the most used experimental models for testing various therapies addressing type 1 DM is based on STZ administration. STZ induces type 1 DM, with destruction of pancreatic beta cells and associated insulin deficiency, as a result of its cytotoxic effect, mediated by increased synthesis of reactive oxygen species (ROS) and subsequent inflammation [18–20]. A protective effect of EGCG on pancreatic beta cells has been already demonstrated in experimental studies [21]; meanwhile, oral chronic administration of EGCG proved to have hypoglycemic and hypolipidemic effects and to reduce oxidative stress in streptozotocin-diabetic rats [22]. EGCG can exert antioxidant, anti-inflammatory, antiangiogenetic, and antifibrotic effects [2]. The catechol or galloyl groups from catechins act as scavengers for metal ions, reducing further production of free radicals [23]. Another essential effect is represented by the scavenging activity for free radicals, through phenoxyl compounds [24]. EGCG treatment can also reduce oxidative stress by increasing the level of antioxidant enzymes, such as superoxide dismutase (SOD), glutathione peroxidase (GP), and catalase (CAT), emerging in an antiapoptotic consequence [25].

Matrix metalloproteinases (MMPs) are a family of enzymes (peptidases) involved in degradation and remodeling of extracellular matrix (ECM) [26]. Recent studies reveal that MMPs can regulate chemokines and cytokines synthesis, thus participating in innate immunity processes, inflammation, and angiogenesis [27]. MMPs can be generated by various cell types, such as endothelial cells and mononuclear cells of the immune system [28]. Pathological induction of MMP synthesis is associated with an imbalance between synthesis and degradation of ECM proteins leading to ECM degradation [29]. High glucose ambience influences the MMPs' increased synthesis and low tissue inhibitors of MMPs (TIMP) activity [30]. Increased levels of MMP-2 and MMP-9 are observed in type 1 diabetic patients and animal models, such as STZ-induced diabetes mellitus in rats [31,32], and are associated with microvascular complications of DM [28].

Analyzing the EGCG therapeutic properties and pharmacokinetic parameters, considerable individual differences and variations between results were noted [33]. EGCG is highly lipophilic, which explains its low bioavailability (0.2% to 2% of the total load ingested by healthy people), mainly because a large amount of the ingested EGCG is degraded by local microbiota and does not enter into the blood circulation [34]. Improvement of bioavailability and stability of EGCG can be obtained by encapsulation in nanoparticles [35]. Catechin nanoemulsions proved to be stable for long periods of time (120 days at 4 ◦C) [36]. Liposomes, assembled from phospholipid bilayers similar to cell membranes, are one of the nanoparticles frequently used for drug delivery [23]. Their biphasic character makes them suitable for being carriers for both hydrophilic (in the central aqueous compartment) and hydrophobic (in lipid bilayers) compounds [37,38]. Nanoformulation by encapsulation in liposomes could also facilitate the solubility for hydrophobic particles [4]. Through all of these properties, liposomes can offer an enhanced bioavailability, stability, and shelf life for sensitive ingredients [39].

The aim of this study was to investigate the effect of two forms of EGCG (EGCG solution and liposomal EGCG) on oxidative stress parameters, antioxidant capacity, serum MMP-2 and -9, and pancreatic and liver function in STZ-induced diabetes mellitus in rats.

#### **2. Materials and Methods**

#### *2.1. Materials*

The substances used for liposomal preparation were: Epigallocatechin-gallate (EGCG) derived from green tea (Sigma-Aldrich, Steinheim, Germany); 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC): *N*-(carbonyl-methoxypolyethylenglycol-2000)-1,2-distearoylsn-glycero-3-phosphoethanolamine Na-salt (MPEG-2000-DSPE) (Lipoid GmbH, Ludwigshafen am Rhein, Germany); and cholesterol (CHO) obtained from sheep wool (Sigma-Aldrich, Steinheim, Germany). All other solvents and reactive substances were obtained from Sigma-Aldrich, Steinheim, Germany, and had an analytical degree of purity.

#### *2.2. Experimental Model*

The study was approved by the Ethic Committee of the University and by the National Sanitary Veterinary Authority number 137/13.11.2018. Twenty-eight male Wistar-Bratislava rats were procured from the Centre of Experimental Medicine, University of Medicine and Pharmacy, Cluj-Napoca, Romania. The rats weighed 200–250 g, were kept in polypropylene cages, with day–night regimen, at constant temperature (24 ± 2 ◦C) and humidity (60 ± 5%). Free access to food (standardized pellets from Cantacuzino Institute, Bucharest, Romania) and water was provided to all animals. The animals were randomly divided into 4 groups (7 rats/group). The groups were organized as follows:

group 1—control group (C)—with intraperitoneal (i.p.) administration of 1 mL saline solution, group 2—STZ administration by i.p. route (STZ),

group 3—STZ administration as before + i.p. administration of EGCG solution (EGCG),

group 4—STZ administration as before + i.p. administration of liposomal EGCG (L-EGCG).

Each medication was dissolved in saline solution (0.9% sodium chloride) and the volume administrated i.p. was 1 mL [19]. The following doses were used: STZ—60 mg/100 g body weight (b.w.) [40]; EGCG in saline solution or in liposomal form were freshly prepared and were administrated i.p. in a dose of 2.5 mg/100 g b.w./day as pretreatment, two consecutive days before STZ administration [41]. Intraperitoneal administration was preferred as a method that improves EGCG bioavailability, compared to low bioavailability with oral administration [42].

Blood samples were taken at 48 h after STZ administration, under ketamine anesthesia (5 mg/100 g bw, i.p. route) from retro-orbital sinus, followed by rat euthanasia by cervical dislocation [43]. Rats with glucose higher or equal to 200 mg/dL were considered to have diabetes mellitus [20].

#### *2.3. Preparation and Physicochemical Characterization of EGCG-Loaded Liposomes*

For the preparation of liposomes, we used a modified film hydration method [44,45]. The lipid double-layer components, having a 70 mM concentration (DPPC:MPEG-2000-DSPE:CHO = 4.75:0.25:1 molar ratio), were dissolved in ethanol in a round-bottomed glass flask. Ethanol was evaporated at 45 ◦C under low pressure; the lipid film product was hydrated with a solution of EGCG diluted in highly purified water, pH = 5.00, at the same temperature. The resulted liposomal dispersion was then extruded through polycarbonate membranes with 200 nm final pore dimension, with LiposoFastLF-50 equipment (Avestin Europe GmbH, Mannheim, Germany). Unencapsulated

EGCG particles were removed by dialysis method, using Slide-A-Lyzer filters (cassettes) with 10 kDa molecular weight cut-off.

To assess the amount of liposomal-loaded EGCG, we used a spectrophotometric method—the reaction with Folin–Ciocâlteu reagent (Merck, Darmstadt, Germany) [46]. During this procedure, a dilution of liposomal dispersion with methanol 1:10 (*v*/*v*) was made, and a UV-VIS spectrophotometer (Specord 200 Plus, Analytik Jena, Überlingen, Germany) measured the absorbance value.

The size and polydispersity index of liposomes were assessed by dynamic light scattering method (with a 90◦ scattering angle), and the zeta potential was measured by laser Doppler electrophoresis; a Zetasizer Nano ZS analyzer was used for both assessments (Malvern Instruments Co., Malvern, UK).

The mean liposomal concentration of the L-EGCG solution was about 900 μg/mL, and encapsulation efficiency was over 80%. Liposomal vesicles' mean size was 170 nm, and polydispersity index was less than 0.2, meaning that the vesicles' size and uniformity were appropriate to ensure a prolonged circulation in the blood. Aggregative stability was ensured by values of 51.83 mV of the zeta potential.

#### *2.4. Oxidative Stress and Antioxidant Parameters Assessment*

Parameters associated with oxidative stress and antioxidant status were determined from collected blood samples. The parameters used to assess oxidative stress were: malondialdehyde (MDA) [47], indirect nitric oxide (NOx) synthesis assessment [48], and total oxidative status (TOS) [49]. Antioxidant status parameters were represented by total antioxidant capacity of plasma (TAC) [50], thiols [51], and catalase [52]. All measurements were performed using a Jasco V-350 UV-VIS spectrophotometer (Jasco International Co, Ltd., Tokyo, Japan). Matrix metalloproteinases (MMPs) were appraised from serum using a rat ELISA kit (Boster Biological technology, Pleasanton, CA, USA) and a Stat Fax 303 ELISA reader (Quantikine, McKinley Place NE, MN, USA).

#### *2.5. Assessment of Beta Pancreatic Cells and Hepatic Cells Function*

Glycemia was measured at 48 h after DM induction, as it was previously observed that STZ induces significant beta cell death at 48 after administration [53]. Glycemia was also used as a parameter for pancreatic function changes induced by experimental diabetes mellitus. Hepatic cytolysis was assessed by serum levels of aspartate aminotransferase (AST) and alanine aminotransferase (ALT) measured by a standardized technique (Vita Lab Flexor E, Spankeren, The Netherlands) [40].

#### *2.6. Data Analysis*

The SPSS software package version 21.0 (SPSS Inc., Chicago, IL, USA) was used for statistical analysis and graphic representations. The acceptable error threshold was *p* = 0.05. In order to describe the continuous quantitative data, we used the arithmetic mean and the standard deviation (SD). The distribution of investigated markers in groups was plotted as individual values (circles) and median (line), as recommended by Weissgerber and coauthors [54]. The Kruskal–Wallis ANOVA was used to test the differences in the investigated markers. The Mann–Whitney test was used in post hoc analysis when significant differences were identified by the Kruskal–Wallis ANOVA test.

#### **3. Results**

No rat died during the experiment, so the analysis was conducted on all seven rats in each group. All P values for comparison between groups are presented in Supplementary Table S1.

In our experimental model, diabetes mellitus was successfully induced by STZ: all rats that received STZ were definitely diabetic, proven by glycemia >200 mg/dL and values significantly higher in diabetic rats compared to control group: 401.81(11.31) mg/dL versus 84.27 (2.87) mg/dL, respectively (expressed as mean and standard deviation), with a *p*-value < 0.001. Also, hepatic damage was detected in the STZ group, quantified by significant elevation of transaminases AST and ALT (Table 1).


**Table 1.** Values of oxidative stress parameters, antioxidants levels, glycemia, hepatic enzymes, and matrix metalloproteinases in the four groups, expressed as mean and standard deviation.

MDA = malondialdehyde; NOx = indirect nitric oxide; TOS = total oxidative status; AST= aspartate aminotransferase; ALT = alanine aminotransferase; MMP-2 = matrix metalloproteinase 2; MMP-9 = matrix metalloproteinase 9; STZ = streptozotocin control; STZ + EGCG = STZ and EGCG solution i.p. as pretreatment; STZ + L-EGCG = STZ and liposomal EGCG i.p. as pretreatment.

Oxidative stress parameters (MDA, NOx, and TOS) significantly increased after induction of DM (*p*-values <0.001 in all items, Figure 1a–c, Table 1). MMP-2 and MMP-9 levels were significantly higher in the STZ-induced DM group compared with control group (*p*-values <0.001, Figure 4a,b, Table 1). Serum antioxidant capacity, measured by thiol, catalase, and TAC levels, was significantly reduced in diabetic rats compared to control animals (*p*-values < 0.001 in all items, Figure 2a–c, Table 1).

In the diabetic group pretreated with EGCG, oxidative stress parameters NOx and TOS were significantly reduced compared to the untreated STZ group (with *p*-values of 0.017 and <0.001, respectively, Figure 1b,c).

**Figure 1.** *Cont.*

**Figure 1.** Distribution of oxidative stress intensity by groups: (**a**) MDA (malondialdehyde), (**b**) NOx (indirect nitric oxide), (**c**) TOS (total oxidative status) on all study groups (7 rats/group). STZ = streptozotocin control; STZ + EGCG = STZ and EGCG solution i.p. as pretreatment; STZ + L-EGCG = STZ and liposomal EGCG i.p. as pretreatment. The symbol–number codes correspond to the *p*-values < 0.05 as follows: α—STZ compared to control; β—STZ + EGCG compared to control; ε—STZ + EGCG compared to STZ; γ—STZ + L-EGCG compared to control; λ—STZ + L-EGCG compared to STZ; μ—STZ + L-EGCG compared to STZ + EGCG.

**Figure 2.** Distribution of plasmatic antioxidant capacity by groups: (**a**) Thiols, (**b**) Catalase, (**c**) TAC (total antioxidant capacity) on all study groups (7 rats/group). STZ = streptozotocin control; STZ + EGCG = STZ and EGCG solution i.p. as pretreatment; STZ + L-EGCG = STZ and liposomal EGCG i.p. as pretreatment. The symbol–number codes correspond to the *p*-values < 0.05 as follows: α—STZ compared to control; β—STZ + EGCG compared to control; ε—STZ + EGCG compared to STZ; γ—STZ + L-EGCG compared to control; λ—STZ + L-EGCG compared to STZ; μ—STZ + L-EGCG compared to STZ + EGCG.

All antioxidant parameters (thiols, catalase, and TAC) were significantly higher in the STZ-treated group (*p*-values of < 0.001, 0.026, and 0.017 respectively, Figure 2a–c).

No significant differences were noted in MDA and MMP values between the pretreated group with EGCG compared to the untreated STZ group (Figure 1a, Figure 4a,b). Also, glycemia and liver parameters were not significantly different in the EGCG pretreated group, with the exception of a decrease in ALT (*p*-value = 0.038, Figure 3c).

**Figure 3.** Distribution of (**a**) Glycemia, (**b**) AST (aspartate aminotransferase), (**c**) ALT (alanine aminotransferase) on all study groups (7 rats/group). STZ = streptozotocin control; STZ + EGCG = STZ and EGCG solution i.p. as pretreatment; STZ + L-EGCG = STZ and liposomal EGCG i.p. as pretreatment. The symbol–number codes correspond to the *p*-values < 0.05 as follows: α—STZ compared to control; β—STZ + EGCG compared to control; ε—STZ + EGCG compared to STZ; γ—STZ + L-EGCG compared to control; λ—STZ + L-EGCG compared to STZ; μ—STZ + L-EGCG compared to STZ + EGCG.

In the STZ group pretreated with L-EGCG, all oxidative stress parameters were significantly decreased and serum antioxidant capacity parameters were all increased, with better results compared to the STZ group pretreated with EGCG (*p* < 0.017, Figures 1 and 2). Also, the L-EGCG solution improved glycemic values and decreased transaminases levels better than EGCG (*p* < 0.001, Figure 3). The MMP levels were significantly lower in the L-EGCG-treated group compared to the diabetic untreated group or compared to the STZ group pretreated with EGCG (<0.001, Figure 4).

**Figure 4.** Distribution of matrix metalloproteinase (MMP): (**a**) MMP-2 and (**b**) MMP-9 on all study groups (7 rats/group). STZ = streptozotocin control; STZ + EGCG = STZ and EGCG solution i.p. as pretreatment; STZ + L-EGCG = STZ and liposomal EGCG i.p. as pretreatment. The symbol–number codes correspond to the *p*-values < 0.05 as follows: α—STZ compared to control; β—STZ + EGCG compared to control; γ—STZ + L-EGCG compared to control; λ—STZ + L-EGCG compared to STZ; μ—STZ + L-EGCG compared to STZ + EGCG.

The Kruskal–Wallis ANOVA test identified significant differences between the groups with diabetes and EGCG pretreatment for all evaluated parameters (*p*-values < 0.0001). The post hoc analysis identified significant differences in most of the cases with better protection for the EGCG-treated group, and significantly higher protection when liposomal EGCG solution was used (Figures 1–4).

#### **4. Discussion**

#### *4.1. Protective E*ff*ects of EGCG on Pancreatic and Hepatic Cell Function in Diabetic Rats*

In our study, EGCG reduced blood glucose levels in pretreated animals but the reduction was not statically significant (Table 1, Figure 3). Some of the antidiabetic effects of EGCG are suggested to be the suppression of appetite, adjustment of dietary fat emulsification in the gastrointestinal tract, inhibition of gastrointestinal lipolysis, and reduction of nutrient absorption enzymes [55]. The most significant hypoglycemia was obtained in liposomal EGCG-pretreated groups. This indicates a protective effect of EGCG on pancreatic cell function. Meng et al. showed that EGCG can inhibit inflammation by reducing reactive oxygen species and downregulating the production of inducible nitric oxide synthetase (iNOS) [56]. Furthermore, EGCG increases glucose tolerance [57] and decrease HbA1c levels in STZ-induced diabetes in rats, contributing to further prevention of diabetic complications [58]. Another suggested mechanism of EGCG's protective effect is the increased glucose uptake due to promoting the glucose transporter-4 (GLUT4) translocation in skeletal muscle, through activation of both phosphoinositol 3-kinase and AMP-activated protein kinase pathways [58]. EGCG also increases tyrosine phosphorylation of insulin receptors, having an insulin-like effect on H4IIE hepatoma cell lines [59].

The liver is extremely adversely affected in type 1 diabetes mellitus. In our study, we found elevated AST and ALT levels, showing liver damage, in STZ diabetic rats (Table 1, Figure 3). In STZ-induced diabetes, transaminases elevation is the consequence of the toxic effect of STZ on hepatocytes, which induces lipid peroxidation, oxidative stress enhancement, peroxisome proliferation, and mitochondrial dysfunction [60–62]. Rodriguez et al. identified increased NO levels and hepatic oxidative stress in STZ-induced diabetic rats [63]. In our study, pretreatment with EGCG decreased ALT levels, preventing hepatic damage induced by STZ. Furthermore, liposomal EGCG administration significantly reduced AST and ALT values, confirming the enhanced protective effect of L-EGCG on

hepatic cells. Other studies also demonstrated the hepatic-protective effect of green tea extracts in hepatic injury reflected by decreased serum transaminase levels, and improved structural changes in histopathological examination [64]. Moreover, long-time consumption of EGCG (in healthy Wistar rats) decreases age-induced hepatic damage by lowering the ALT and AST serum levels and improving microscopic changes of the liver tissue due to the aging process [65].

#### *4.2. E*ff*ect of EGCG on Oxidative Stress Parameters and Plasmatic Antioxidant Capacity*

In this study, increased levels of MDA, NO, and TOS were observed in diabetic rats (Table 1 and Figure 1), together with low levels of antioxidant biomarkers such as thiols, catalase, and TOS (Table 1 and Figure 2). Pretreatment with EGCG and L-EGCG induced protection against STZ toxic effects, as demonstrated by reduction of oxidative stress parameters (Table 1, Figure 1) and by enhancement of antioxidant defense (Table 1, Figure 2), with best results for the liposomal form. STZ-induced diabetes in experimental models is followed by an enhanced production of reactive oxygen species (ROS) and consumption of cell antioxidant systems, as a consequence of necrotic and apoptotic degeneration of pancreatic β cells [66,67]. Hyperglycemia itself is another factor generating intracellular ROS [68]. Oxidative stress (by excessive ROS production, auto-oxidation of glycated proteins, and increased lipid peroxidation) and decreased antioxidant capacity (free radical scavengers and enzymatic systems) are also involved in the pathogenesis of diabetic complications [69–72].

Green tea component EGCG is a flavonoid with antioxidant and anti-inflammatory properties conferred by its particular structure, a flavanol core and two gallocatechol rings, which are able to bind metal ions and scavenge free oxygen radicals. As a consequence, EGCG exerts direct antioxidant effects (scavenger of ROS and cheater of metal ions), but also indirect antioxidant effects (inductor of antioxidant enzymes, such as catalase, and inhibitor of oxydases, such as NADPH—nicotinamide adenine dinucleotide phosphate, lipoxygenase, or xantin-oxydase) [73]. Anti-inflammatory effects of EGCG were also related to the increase of circulating levels of interleukin-10 (an anti-inflammatory cytokine) in nonobese diabetic mice [14]. EGCG can decrease lipid peroxidation in the liver, kidney, and brain, and reduce lymphocyte DNA damage in diabetic mice [74].

EGCG has low bioavailability which can be modified by incorporation in special drug delivery systems. Because of its highly lipophilic nature, EGCG is suitable for incorporation in liposome nanoparticles, composed of phospholipid bilayers. Minnelli et al. showed that pretreatment of adult retinal pigmented epithelium (ARPE) cells with EGCG encapsulated in magnesium liposomes increases the survival of cells exposed to hydrogen peroxide (H2O2), with better preserved mitochondria structure on electron microscopy examination, showing the superior antioxidant activity of L-EGCG compared with free EGCG [75]. In this regard, natural antioxidant products could be a promising therapeutic option for prevention of diabetes mellitus and its complications, conferring protection against oxidative damage by liposomal nanostructure encapsulation [69].

#### *4.3. EGCG E*ff*ect on Matrix Metalloproteinases*

In the present study, serum levels of MMP-2 and -9 increased after DM induction and were better modulated by L-EGCG (Table 1 and Figure 4). In experimental models of DM, increased MMP-2 expression and activity were linked to elevated ROS levels and oxidative stress, with consecutive pancreatic beta cell apoptosis, showing MMP-2- s important role in DM pathogenesis [76]. Thus, inhibition of intracellular MMP-2 expression is an essential target for beta cell protection and DM prevention. There is also a postulated connection between MMP production and inflammatory process and proinflammatory cytokine production associated with DM. Chemokines such as MCP-1 and NF-kB can induce MMP overproduction in DM [77]. After their secretion as inactive forms, proinflammatory molecules contribute to further transformation of MMPs in active forms by different proteases that are implicated in their cleavage [38]. MMPs are also involved in regulation and duration of immune response, endothelial cell function, vascular smooth muscle migration and

proliferation, Ca2<sup>+</sup> signaling pathways, and vessel contraction, all of these consistently influencing vascular remodeling in DM [78,79].

Activated inflammatory cells such as leucocytes can contribute to endothelial cell dysfunction and vascular damage by direct and indirect pathways. Indirect loops comprise augmentation of MMP production by proinflammatory cytokines synthesized in activated leucocytes [70].

Activation of MMP-2 and MMP-9 is important in pathogenesis of diabetic microangiopathic complications such as diabetic retinopathy, nephropathy, and neuropathy [39]. Diabetic retinopathy, by inducing apoptosis of retinal endothelial cells and by degrading the junction proteins, is followed by increased vascular permeability [80,81]. In experimental models of DM, increased oxidative stress activates MMP-2, and antioxidant therapies inhibit the development of diabetic retinopathy by modulating retinal MMP-2 levels [32,82]. Diabetic nephropathy, one of the most severe microangiopathy in diabetes mellitus, is also characterized by MMP overexpression and accelerated ECM degradation, both being a hallmark of associated histopathologic changes [30]. MMPs' increased synthesis can also lead to neuronal injury through blood–nerve barrier (BNB) disruption, contributing to the neuropathic pain associated with diabetic neuropathy [83,84].

The multiple and complex roles exhibited by MMPs are explained by their multiple localizations. MMP-2 and MMP-9 are colocalized in vessel walls and atherosclerotic plaque, being involved in endothelial dysfunction and DM macrovascular complication and vascular remodeling [85,86]. Wang et al. reported a protective effect of EGCG after i.p. administration, by reducing the plasma levels of TNF-α, IL-6, and monocyte chemoattractant protein-1 (MCP-1) [38]. There is also evidence that EGCG can inhibit MMP-2 activation [87]. Multiple compounds of green tea can inhibit MMP-2 and -9, but the most efficient ones proved to be EGCG and epigallocatechin (EGC) [88]. Therefore, we chose the EGCG compound for our experimental study. Moreover, liposomal encapsulation brings an increased bioavailability with better results in reducing oxidative stress biomarkers and MMP plasma level. EGCG reduces MMP-2 activity by targeting the fibronectin type II repeated regions 1 and 3 of MMP-2, binds the amino acids that constitute the exosite of this enzyme, and hinders proper positioning of the substrate [89]. Due to its antioxidants effects and inhibitory action on the protein tyrosine kinases, EGCG reduces MMP-9 activity by reducing its release from the activated neutrophils [90].

From our knowledge, this is the first experimental study addressing liposomal EGCG effects in experimental DM induced by STZ in rats. Decreasing the hepatic and pancreatic damage due to STZ administration is a valuable effect of liposomal EGCG.

#### *4.4. Potential Limitations of the Study*

No measurements of EGCG and L-EGCG in the blood or pancreatic and hepatic tissue were done in this study since such quantifications were outside of our aim. Future studies could be conducted to measure the concentration of EGCG and L-EGCG in the blood and tissues. Moreover, oxidative stress parameters and MMPs could be measured in liver and pancreas tissue. Another limitation of our study is that the evaluation of endogenous insulin levels and measurement of HOMA-IR for endogenous pancreatic function were not performed.

Future studies should also investigate the effects of long-term administration of EGCG and L-EGCG on DM and its complications, as this study was focused on assessing their effects 48 h after DM induction.

#### **5. Conclusions**

L-EGCG pretreatment reduces oxidative stress biomarkers and MMP plasma levels 48 h after DM induction. Further studies are needed to detect other particularities regarding the EGCG protective mechanisms in order to improve their therapeutic efficiency. Due to the beneficial effects of EGCG nanoformulation proven by this study on oxidative stress, antioxidative defense, and MMP-2 and -9, we propose that L-EGCG could be considered as a novel adjuvant therapy in DM management.

**Supplementary Materials:** The following is available online at http://www.mdpi.com/2076-3921/9/2/172/s1, Table S1: *p*-values for comparisons between the study groups for all studied parameters.

**Author Contributions:** Conceptualization, G.D., C.M., and C.A.N.; Data curation, A.E.B., C.M., and I.C.S.; Formal analysis, M.V. and R.M.R.; Funding acquisition, G.D. and I.C.S.; Investigation, A.E.B., A.S.P., C.B., and M.V.; Methodology, P.-M.B., C.B., and C.A.N.; Project administration, A.E.B. and P.-M.B.; Resources, A.S.P., C.B., and I.C.S.; Software, P.-M.B. and M.V.; Supervision, R.M.R.; Validation, P.-M.B., A.S.P., C.M., and R.M.R.; Visualization, G.D. and C.M.; Writing—original draft, A.E.B. and I.C.S.; Writing—review & editing, A.S.P., G.D., and C.A.N. All authors have read and agreed to the published version of the manuscript.

**Acknowledgments:** The authors would like to thank Olivia Veris, ezan-Ros,u for professional English language editing of the manuscript.

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

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **A Novel Pectic Polysaccharide of Jujube Pomace: Structural Analysis and Intracellular Antioxidant Activities**

#### **Ximeng Lin 1, Keshan Liu 1, Sheng Yin 2, Yimin Qin 3, Peili Shen 3,\* and Qiang Peng 1,\***


Received: 29 December 2019; Accepted: 30 January 2020; Published: 2 February 2020

**Abstract:** After extraction from jujube pomace and purification by two columns (DEAE-Sepharose Fast Flow and Sepharcyl S-300), the structure of SAZMP4 was investigated by HPGPC, GC, FI-IR, GC-MS, NMR, SEM, and AFM. Analysis determined that SAZMP4 (Mw = 28.94 kDa) was a pectic polysaccharide mainly containing 1,4-linked GalA (93.48%) with side chains of 1,2,4-linked Rha and 1,3,5-linked Ara and terminals of 1-linked Rha and 1-linked Ara, which might be the homogalacturonan (HG) type with side chains of the RG-I type, corresponding to the results of NMR. In AFM and SEM images, self-assembly and aggregation of SAZMP4 were respectively observed indicating its structural features. The antioxidant activity of SAZMP4 against H2O2-induced oxidative stress in Caco-2 cells was determined by activity of superoxide dismutase (SOD) and glutathione peroxidase (GSH-Px) as well as malondialdehyde (MDA) and reactive oxygen species (ROS) levels, indicating SAZMP4 can be a natural antioxidant. Also, a better water retention capacity and thermal stability of SAZMP4 was observed based on DSC analysis, which could be applied in food industry as an additive.

**Keywords:** polysaccharide; jujube pomace; structural analysis; antioxidant activity

#### **1. Introduction**

Pectin is a natural macromolecular compound, generally considered as a complex polysaccharide containing α-1,4-linked galacturonic acid, which might be partly methyl esterified and have side chains of various neutral sugars, such as rhamnose, arabinose, galactose, and so on [1]. It widely exists in the cell wall and the middle lamella structure of all higher plants [2]. Because of pectin's gelatification, thickening, and stabilization, it is widely applied in food, medical, chemical, and other industries [3]. China, the original country of jujube, has cultivated jujube since around 7000 years ago. Since jujube fruits are rich in sugar, and the abundant intracellular and cell wall polysaccharides are more soluble in alkaline solution, it is better to use alkaline solution for extraction to take full advantage of the pomace [4]. According to previous research, jujube polysaccharides have different biological activities, such as antioxidant activity [5], immunoregulatory activity [6], hepatoprotective effects [7], anti-hyperlipidemia effects [8], and antitumor activity [9]. Obviously, biological activities of polysaccharides are associated with their structural characteristics. Many researchers have determined the composition, the average molecular weight, and the type of linkages of polysaccharides can affect the biological activities of polysaccharides [10–12]. Thus, there exists an importance to determine the structural characterizations of pectic polysaccharide.

It is widely acknowledged that free radicals are indispensable in metabolic processes. However, oxidative stress is an important factor in these diseases. Reactive oxygen species (ROS), chemically active substances mainly including peroxide, superoxide, hydroxyl radicals, and singlet oxygen [13], plays a necessary role in physiological regulation and message passing in moderate amount. Normally, there exists an antioxidant system containing antioxidants and antioxidant enzymes that controls the metabolic balance of free radicals. However, as a result of some unnormal factors and conditions like ischemia, hypoxia, chemicals, ionizing radiation, chemotherapy drugs, and ultraviolet radiation, abundant free radicals will be produced and break the metabolic balance. Under this circumstance, oxidative stress will occur. Also, excessive free radical accumulation can injure the components of cells, such as DNA and proteins, leading to the development and progression of diseases such as diabetes, cancer, and cardiovascular diseases [14,15]. Thus, it is imperative to improve antioxidant activity in order to prevent and control these diseases.

In this study, the structure of a pectic polysaccharide extracted from jujube pomace was characterized by chemical and instrumental methods, and the antioxidant activity was investigated by the Caco-2 cells model.

#### **2. Materials and Methods**

#### *2.1. Materials and Reagents*

The dry fruits (*Ziziphus jujuba cv.* Muzao) were provided by the Loess Plateau Experimental Orchard from Yulin in Shaanxi province, China. The chromatographic columns (DEAE-Sepharose Fast Flow and Sepharcyl S-300) were purchased from GE Healthcare Life Sciences (Piscataway, NJ, USA). The standards arabinose, fucose, galactose, glucose, mannose, rhamnose, and xylose were from Solarbio Life Sciences Co. (Beijing, China). The standards glucuronic acid and galacturonic acid were purchased from Aladdin Biochemical Technology Co., (Shanghai, China). All other chemicals were analytical grade.

#### *2.2. Extraction and Purification of SAZMP4*

Jujube pomace was obtained by removing the water-insoluble polysaccharide jujube powder. Alkaline extraction (0.1 M NaOH, 25 ◦C, 1 h) was applied to jujube pomace for obtaining crude polysaccharide. Then, an anion-exchange column of DEAE-Sepharose Fast Flow (2.6 × 100 cm), using 0.3 M NaCl as mobile phase at room temperature with a flow rate of 1.2 mL/min, and a gel-permeation chromatography column of Sepharcyl S-300 (2.6 × 100 cm), using ultrapure water as mobile phase at room temperature with a flow rate of 0.8 mL/min, were used to purify the crude polysaccharide in order to derive the purified polysaccharide, SAZMP4 [16].

#### *2.3. Structural analysis of SAZMP4*

#### 2.3.1. Physical and Chemical Analysis

The phenol-sulfuric acid method [17], the Bradford method [18], and the Folin–Ciocalteu reagent method [19] were used to measure the content of total sugar, protein, and total phenolics.

The UV-Vis spectrum was recorded by a UV7 spectrophotometer (METTLER, TOLEDO, Zurich, Switzerland) in the 200–400 nm region to detect the protein and nuclear acids [20], and the FI-IR spectrum was recorded with a Fourier transform infrared spectrometer (FI-IR, Vetex70, Bruker Co., Ettlingen, Germany) in the 4000–400 cm−<sup>1</sup> region by KBr pellets to determine the primary functional groups [21].

High-performance gel-permeation chromatography (HPGPC) equipped with an Agilent 1200 series high-performance liquid chromatography system, a Waters 2414 refractive index detector, and a TSK gel G5000PWXL column (300 × 7.8 mm, Tosoh, Japan) were used to determine the homogeneity and average molecular weight of the purified polysaccharide [22]. The retention time was used to calculate the average molecular weight of SAZMP4.

Gas chromatography (GC, GC-2014, Shimadzu Co., Kyoto, Japan) with a capillary column of DB-17 (30 m × 0.25 mm × 0.25 μm, Agilent, Santa Clara, CA, US) was used to indicate the monosaccharide composition of SAZMP4 [23], and the mixed standard monosaccharides were used for the monosaccharide identification and quantification.

A differential scanning calorimeter (DSC, Q2000, Waters, Milford, MA, USA) was used to analyze the thermal properties of SAZMP4 [24]. The dried and powdered polysaccharide (3 mg) was put into a standard aluminum crucible and sealed immediately. The program raised the temperature from 40 ◦C to 300 ◦C at a rate of 10 ◦C/min in a dynamic inert nitrogen atmosphere (50 mL/min). Simultaneously, an empty standard aluminum crucible was used as a reference.

#### 2.3.2. Methylation Analysis

After the uronic acid reduction by Taylor and Conrad [25], SAZMP4 was methylated by a method reported previously [26]. The disappearance of the absorption band of O-H around 3400 cm−<sup>1</sup> in the FI-IR spectrum indicated complete methylation of the sample. Then, the sample was hydrolyzed by trifluoroacetic acid, restored with sodium borohydride, acetylated by acetic anhydride, and dissolved in chloroform. A GCMS-QP2010A instrument (Shimadzu Co., Kyoto, Japan) equipped with a Rtx-50 capillary column (30 m × 0.25 mm × 0.25 μm) and an ion trap MS detector was used to determine the derivatives.

#### 2.3.3. NMR Analysis

The dried SAZMP4 was dissolved (D2O) and lyophilized three times. Fifty milligrams of deuterium-exchanged SAZMP4 was dissolved in 0.5 mL D2O. NMR spectra of 1H and 13C were recorded with a Brucker AVANCE III 500 MHz nuclear magnetic resonance spectrometer (NMR) using standard pulse sequences at 25 ◦C [27].

#### *2.4. Molecular Morphological Analysis*

A field emission scanning electron microscope (SEM, S-4800, Hitachi, Tokyo, Japan) was used to record the surface morphological properties of SAZMP4 [28]. Before observation, SAZMP4 was covered with a gold layer.

An atomic force microscope (AFM, Multimode-8, Bruker Co., Billerica, MA, USA) was used to document the properties of the molecular morphology of the polysaccharide. Ten microliters of the polysaccharide solution (1 μg/mL) was dropped onto a mica carrier and then dried at room temperature, using tapping mode on the AFM for record [29].

#### *2.5. Antioxidant Activity of SAZMP4*

#### 2.5.1. Cell Culture

Human colorectal adenocarcinoma cells (Caco-2) were obtained from Shanghai Institute of Cell Biology (Shanghai, China). The cells were cultured in high-glucose Dulbecco's modified Eagle's medium (H-DEME, Hyclone, Logan, UT, USA) with 10% fetal bovine serum (FBS, Biological Industries Beit Haemek, Kibbutz, Israel), 100 units/mL penicillin, and 100 μg/mL streptomycin and in a humidified atmosphere of 5% CO2 at 37 °C. Between 3 and 15 passages of the cells were used in this study.

#### 2.5.2. Cell Viability Analysis

Cell Counting Kit-8 (CCK-8, EnoGene Co., Shanghai, China) was used to evaluate the cell viability. In brief, Caco-2 cells were cultured in 96-well plates with a density of 5 <sup>×</sup> 103 cells/mL and incubated for 24 h in a 37 ◦C incubator with a humidified 5% CO2 atmosphere. After that, the cells were treated with different concentrations of SAZMP4 (50, 100, 200, 400, and 800 μg/mL) for 24 h. Then, 10 μL of

CCK-8 solution was added, and the cells were incubated in the same environment for 1 h. The cell viability was determined by a multifunctional enzyme marker (victorX3, PerkinElmer Co., Waltham, Massachusetts, US) at a wavelength of 450 nm and was expressed as a relative percentage to the blank control group.

#### 2.5.3. Treatment Procedure

For treatment, the cells (5 <sup>×</sup> <sup>10</sup><sup>3</sup> cells/mL) were cultured in 96-well plates and incubated at 37 ◦C for 24 h. Then, the cells were treated with different concentrates of SAZMP4 (25, 50, 100, and 200 μg/mL) for 24 h. After removing the medium, the cells were exposed to 200 μM of H2O2 for 2 h. The cell viability, superoxide dismutase (SOD), glutathione peroxidase (GSH-Px), ROS, and malondialdehyde (MDA) levels were determined by relevant commercial kits.

#### 2.5.4. Measurement of SOD and GSH-Px

Cell lysates treated without or with different concentrations of SAZMP4 were collected for antioxidant enzymes (SOD and GSH-Px) analysis. The activities of SOD were measured by the relevant commercial kits (Beyotime, Biotechnology, Shanghai, China) using the xanthine oxidase method for determination. The activity of SOD was defined as the corresponding SOD content when the SOD inhabitation rate in each milliliter of reaction liquid reached 50%.

The activities of GSH-Px were measured by the relative commercial kits (Beyotime, Biotechnology, Shanghai, China). The activities of GSH-Px were determined by the consumption of GSH in enzymatic reactions.

#### 2.5.5. Intracellular ROS and MDA Levels

The intracellular ROS was determined by a Reactive Oxygen Species Assay Kit (Beyotime, Biotechnology, Shanghai, China), investigated by fluorophore 2,7-dichlorofluorescein diacetate (DCFH-DA). After incubating the cells in a black 96-well plate for 24 h and removing the medium, the cells were washed with phosphate-buffered saline (PBS, 100 μL). After that, the cells were cultured with DCFH-DA at a concentration of 10 μM at 37 ◦C for 30 min. The results were determined by a multimode microplate reader (PerkinElmer, Waltham, MA, USA) and expressed as fold changes in fluorescence intensity versus control.

The MDA levels were indicated by corresponding detection kits (Jiancheng Bioengineering Institute, Nanjing, China) following the manufacturer's instructions.

#### *2.6. Data Analysis*

All test data were expressed as mean ± SD from no fewer than three determinations and analyzed with variance (ANOVA) following multiple tests. SPSS version 22.0 was used for all statistical analyses, and *p* < 0.05 was considered to be significant.

#### **3. Results**

#### *3.1. Separation and Purification of SAZMP4*

SAZMP4 was extracted from jujube pomace by alkaline solution and purified by the column of DEAE-Sepharose Fast Flow with mobile phase of 0.3 M NaCl and the column of Sepharcyl S-300 with ultrapure water. The yield of crude polysaccharide was 5.3% relative to jujube pomace, and the yield of SAZMP4 was 5.10% relative to crude polysaccharide. Other jujube researchers [5,7,9] obtained similar results to this study.

#### *3.2. Preliminary Characterizations of SAZMP4*

According to the phenol-sulfuric acid assay, SAZMP4 was 96.52% sugar. It had a low protein content of 0.78%, coinciding with UV–Vis analysis that the polysaccharide contained no protein (<3%) based on the no absorption peaks at 280 nm. Also, no absorption peaks at 260 nm in the spectrum indicated no nucleic acid in SAZMP4. The total phenol content was not detected in SAZMP4. These results were similar to other acidic jujube polysaccharides from *Z. Jujuba* [30].

In the FI-IR spectrum (Figure 1), the peaks of the intramolecular or intermolecular stretching vibration of O-H was around 3400 cm−<sup>1</sup> and the stretching vibration of C–H was around 2940 cm<sup>−</sup>1, indicating the SAZMP4 was a polysaccharide. The peaks at approximately 1620 cm−<sup>1</sup> were attributed to the stretching vibration of carboxyl, which implied that SAZMP4 might contain uronic acid [31]. The absorptions at 1420 and 1325 cm−<sup>1</sup> belonged to the bending vibration of C–H. Also, the signals at 1200–800 cm−<sup>1</sup> of the fingerprint area of carbohydrates indicated that the bands at 1093 and 1012 cm−<sup>1</sup> were the bending vibration of C–O in the pyranose form. In addition, the weak absorption bands at 941 and 838 cm−<sup>1</sup> were probably attributed to α-glycosidic bonds, further supported by the out-of-plane bending vibration of C–H at around 630 cm−1, which implied the presence of α-glycosidic bonds in SAZMP4 [32].

**Figure 1.** FI-IR spectrum of SAZMP4.

According to the equation, the average molecular weight of SAZMP4 was calculated to be 28.94 kDa with the retention time of 19.91 min. Also, the single and symmetric elution peak from the HPGPC indicated SAZMP4 was a homogeneous fraction.

GC analysis, as shown in Table 1, determined that the monosaccharide composition of SAZMP4 mainly contained galacturonic acid at a molar rate of 93.48%, which coincided with the feature of pectin. This result was similar with other pectic polysaccharides [30,33,34], but the galacturonic acid content of SAZMP4 was higher than them.


**Table 1.** Monosaccharide analysis data of SAZMP4.

The thermodynamic properties of SAZMP4 were examined by DSC from 40 to 300 ◦C. In the DSC thermogram of SAZMP4 (Figure 2), an endothermic peak and an exothermic peak were observed, and the parameters of them were labeled, such as melting temperature (Tm), melting enthalpy (ΔHm), degradation temperature (Td), and degradation enthalpy (ΔHd). Tm and ΔHm were determined by the composition and structural characterizations of polysaccharides. A polysaccharide with lower molecular weight and less uronic acid content has worse capacity to sustain water, so the Tm and ΔHm were lower [35]. The high Tm and ΔHm indicated the better capacity of SAZMP4 to sustain water, coinciding with results of HPGPC and GC analyses. The second peak was caused by the degradation of the polysaccharide in the process [36]. Apparently, Td was primarily impacted by the composition of the polysaccharides, while the ΔHd of polysaccharides was mainly affected by its galacturonic acid content. The Td implied that SAZMP4 was stable below 240 ◦C, related to the better thermal stability, which might be applied in the food industry.

**Figure 2.** DSC thermogram of SAZMP4.

#### *3.3. Methylation Analysis*

For the determination of linkage types between sugar residues with GC-MS, SAZMP4 was firstly subjected to uronic acid reduction in order to avoid β-condensation reaction during methylation, which might cause structural changes to sugar chains [37]. According to the retention time of sugar residues and the standard data of CCRC Spectral Database, the results of the methylation analysis are exhibited in Table 2. SAZMP4 mainly included five types of glycosidic linkages: 1-linked Rha*p*, 1-linked Ara*f*, 1,2,4-liked Rha*p*, 1,3,5-linked Ara*f,* and 1,4-linked Gal*p* with a molar ratio of 0.4: 0.38: 0.62: 0.58: 28.7. This conformed to the GC analysis, which indicated that the structures of sugar chains were not destroyed in the methylation process. Besides, GC analysis showed SAZMP4 contained only galacturonic acid, and no galactose exited in it. Thus, the linkage type of galacturonic acid was 1,4-linked GalA*p*. Obviously, SAZMP4 was a pectic polysaccharide containing 1,4-linked galacturonic acid with side chains of 1,2,4-liked Rha*p* and 1,3,5-linked Ara*f* as well as terminals of 1-linked Rha*p* and 1-linked Ara*f,* which indicated that SAZMP4 might be homogalacturonan (HG) with side chains of rhamnogalacturonan (RG) type I [38].


**Table 2.** Methylation analysis data of SAZMP4.

#### *3.4. NMR Analysis*

The 1H (Figure 3A) and 13C NMR (Figure 3B) spectra of SAZMP4 were displayed, and the chemical shifts in the spectra were classified based on previous research [39]. In the 1H spectrum, the weak peak of δ 1.84 belonged to H-6 of 1,2,4-linked Rhap or 1-linked Rhap, and the other strong signs of δ 4.34, 4.06, 3.92, and 3.60 were respectively attributed to H-5, H-4, H-3, and H-2 of 1,4-linked GalAp. The strong signs of δ 160.45, 99.07, 71.44, 69.67, 68.28, and 62.70 belonged to the C-6, C-1, C-4, C-5, C-3, and C-2 of 1,4-linked GalAp in the 13C spectrum, respectively.

**Figure 3.** *Cont.*

**Figure 3.** The NMR spectra of SAZMP4 in D2O: 1H spectrum (**A**); 13C spectrum (**B**).

#### *3.5. Molecular Morphological Properties*

It is widely acknowledged that SEM can be used to observe the surface morphology of polysaccharides, and the SEM images can exhibit the molecular morphological properties of polysaccharide. At low magnification (400-fold and 3000-fold, Figure 4A,B), SAZMP4 was observed to have a smooth surface and debris shape, while it showed a smooth surface and a thin but large lamellar shape in the image of high magnification (8000-fold and 20,000-fold, Figure 4C,D). These results were different to those of previous research [40] possibly because of the different preparation and purification methods and the structural differences of polysaccharides.

**Figure 4.** The SEM images of SAZMP4 (**A**: × 400, **B**: × 3000, **C**: × 8000, **D**: × 20,000).

AFM is another tool that can not only provide two-dimensional images but also observe three-dimensional surface images of polysaccharides directly in the natural conditions. Generally speaking, sugar chains with different compositions always have the tendency to form into a conformation with the lowest free energy. According to the AFM images (Figure 5B), SAZMP4 was observed to have an irregular bulk structure, indicating the molecular aggregation caused by intermolecular and intramolecular interactions of hydroxyl groups on polysaccharide chains, and these results coincided with the results of SEM analysis. The height analysis in the planar image (Figure 5E,F) revealed the aggregation extent of SAZMP4. A branched, ring-like, helical or an interconnected network structure in the AFM image (Figure 5A,B) implied molecular self-assembly of SAZMP4, which undergoes a spontaneous process from disorder to order based on the weak, noncovalent interaction of hydrogen bonds and van der Waals forces as well as the hydrophobic effect [41]. These results indicated that SAZMP4 might aggregate at first and then self-assemble to form a long chain; this is why the molecular morphology of SAZMP4 appeared the way it did in SEM and AFM images. In addition, SAZMP4 was also observed to have a core structure with branches probably related to the α-1,4 glycosidic bonds, which might be galacturonic acid, coinciding with the methylation analysis [42].

**Figure 5.** The AFM planar image (**A**) and (**B**); three-dimensional image of AFM (**C**) and (**D**); the height analysis of the planar image at the line (**E**) and (**F**).

#### *3.6. Antioxidant Activity of SAZMP4*

#### 3.6.1. The Cytotoxicity of SAZMP4 to Caco-2 Cells

The cytotoxic effects of different concentrations of SAMZP4 on Caco-2 cells were evaluated by the CCK-8 commercial kit. As shown in Figure 6A, SAZMP4 of 50 and 100 μg/mL presented no significant effects on cell viability compared to the control group, while SAZMP4 at high concentrations (200, 400, and 800 μg/mL) caused a significant dose-dependent decrease of cell viability. Hence, the following experiments were carried out with 25, 50, and 100 μg/mL of SAZMP4 for treatment in order to reduce the interference with the polysaccharide.

**Figure 6.** Cell viability of Caco-2 with different concentrations of SAZMP4 (**A**). Cell viability of Caco-2 with 200 μM H2O2 and different concentrations of SAZMP4 (**B**). Intracellular level of ROS in Caco-2 cells (**C**). Intracellular level of MDA in Caco-2 cells (**D**). Activity of SOD in Caco-2 cells (**E**). Activity of GSH-Px in Caco-2 cells (**F**). The data are expressed as the mean ± SD (*n* = 5 wells per group). (\*) *p* < 0.05 and (\*\*) *p* < 0.01 versus the control group.

#### 3.6.2. Protective Effect against H2O2-induced Toxicity

After treating with different concentrations of SAZMP4 for 24 h, the cells were exposed to 200 μM H2O2 for 2 h. The results are exhibited in Figure 6B. The cell viability of the group only exposed to H2O2 declined to 68.6%, indicating the cells were in the state of oxidative stress. Pretreatment of SAZMP4 for 24 h enabled the cells to resist the toxic effects of H2O2, causing the viability to be greater than 81.6%. However, these results were not precise to indicate SAZMP4 had antioxidant activity. Thus, the determinations of ROS and MDA levels and SOD and GSH-Px activities were inevitable.

#### 3.6.3. ROS and MDA Levels in Caco-2 Cells

The ROS and MDA levels in Caco-2 cells after pretreatment of SAZMP4 and treatment of 200 μM H2O2 are exhibited in Figure 6C. There was a significant increase in the ROS level in the group without pretreatment of SAZMP4 compared to the control group. Pretreated groups had significant decreases of ROS levels and exhibited a dose-dependent response. Also, similar results were exhibited in the changes of MDA levels (Figure 6D). At 100 μg/mL, SAZMP4 presented significant decreases compared to the group only exposed to 200 μM H2O2. Apparently, these results suggested that SAZMP4 could help the cells resist the toxicity of H2O2.

#### 3.6.4. The Activity of SOD and GSH-Px in Caco-2 Cells

Figure 6E,F, respectively, showed the activity of SOD and GSH-Px in Caco-2 cells. Compared to the control group, the activity of SOD and GSH-Px significantly declined in the no pretreatment group. With pretreatment of SAZMP4, the activity of SOD and GSH-Px had a significant increase comparing to the 200 μM H2O2-treated group and showed a dose-dependent relationship. Obviously, these results indicated that SAZMP4 might protect the cells from oxidant injury of H2O2 by activating the antioxidant enzymes (SOD and GSH-Px).

#### **4. Discussion**

Mitochondria, bearing the responsibilities of the generation of cell energy (adenosine triphosphate, ATP), the main source of ROS, and the apoptosis of cells, is the core and hub of the entire cell and its vital activities. Also, it can participate in cell signaling. All these physiological functions of mitochondria are mainly needed to regulate energy metabolism and ROS production. In normal cells, there exists a balance between oxidation and antioxidation. However, metabolic disorders of ROS in mitochondria cause oxidative stress, leading to cell apoptosis and some diseases such as cancer, cardiovascular diseases (hypertension, diabetes), and neurodegenerative diseases (Parkinson's disease). One of the ways to cause oxidative stress is lack of antioxidants. When cells are in the state of oxidative stress, they cannot scavenge the free radicals generated by mitochondria. As previous research reported [43–47], the antioxidant activity of polysaccharides is associated with their composition of sugar chains, branched chains, molecular weight, substituents, and conformation. The higher the molecular weight and the higher the content of uronic acid the polysaccharide has, the stronger its antioxidant ability. In this study, the structural analysis of SAZMP4 determined that it mainly contained galacturonic acid with higher molecular weight and branched chains. Hence, SAZMP4 could become a natural antioxidant because it could be the electronic acceptor and scavenge free radicals.

As polysaccharides can only be digested in the intestinal tract, this study used Caco-2 as the cell model and H2O2 as the irritant to induce cellular oxidative stress [47]. The antioxidant system in the human body has the ability to recover and regulate itself. When the body is in a state of oxidative stress, the relevant antioxidant system will produce corresponding antioxidants to control the injury of oxidative stress. SOD is an important kind of enzyme in the antioxidant system that can catalyze superoxide anions to translate into H2O2. GSH-Px is another kind of enzyme in the human body. It can scavenge H2O2 and block the lipid peroxidation radical chain reaction to protect cell membranes and other biological tissues from oxidant injury. Compared to the blank group, the cell viability of the model group had an obvious reduction indicating the success of this oxidative stress model. The stronger activity of antioxidant enzymes (SOD and GSH-Px) and the reduction of MDA and ROS levels compared to the model group indicated SAZMP4 had antioxidant effects and could improve the antioxidant ability of the cells. However, the antioxidant mechanism of polysaccharides has not clearly determined yet. The signaling pathway of Nrf2-keap2-ARE is the primary antioxidant signaling pathway in the body [48]. Nrf2 is the most important transcription factor in this signaling pathway. Active Nrf2 dissociates with keap1 and enters the nucleus to interact with antioxidant response elements (AREs) to start the transcription of antioxidant enzyme genes. At present, many

antioxidants from natural plants have been determined to play an antioxidant role by promoting the dissociation of keap1-Nrf2 and activation of Nrf2, such as curcumin and phenyl ethyl caffeic acid [49]. Since polysaccharides are macromolecules, they cannot get into the membrane. Thus, it is possible that polysaccharides are decomposed into small molecules by intestinal flora at first and then enter the cells to active Nrf2. Moreover, antioxidant polysaccharides also can serve as electronic acceptors by their hydroxyl and carboxyl groups and scavenge free radicals by their special structure to play an antioxidant role in the body. Hence, polysaccharides with more uronic acid and a higher molecular weight can have a stronger antioxidant activity.

In recent years, because of the toxicity and carcinogenesis of synthetic antioxidants, people are more interested in natural antioxidants. Thus, SAZMP4 could be used as a natural antioxidant for food and medicine industries to produce some products for the prevention and control of diseases. In this study, we only investigated the structure and intercellular antioxidant activity of SAZMP4. There is no denying that further research in mice or humans is needed in order to determine the structure–activity relationship based on this study.

#### **5. Conclusions**

In conclusion, SAZMP4 (Mw = 28.94 kDa) is a novel pectic polysaccharide, mainly containing 1,4-linked GalA with side chains of 1,2,4-linked Rha and 1,3,5-linked Ara and terminals of 1-linked Rha and 1-linked Ara, and it has a tendency to aggregate and self-assemble. SAZMP4 can be a natural antioxidant and can be applied in the medicine industry. In addition, SAZMP4 has a better water retention capacity and thermal stability, indicating the potential capacity to be used as an additive in the food industry. This is a systematic work investigating the structure and antioxidant activity of a novel pectic polysaccharide, and it can provide a theoretical basis in further research.

**Author Contributions:** Conceptualization, X.L. and K.L.; methodology, X.L.; software, X.L.; validation, S.Y., P.S. and Q.P.; formal analysis, X.L.; investigation, K.L.; resources, Q.P.; data curation, X.L.; writing—original draft preparation, X.L.; writing—review and editing, X.L. and Y.Q.; visualization, K.L.; supervision, Q.P.; project administration, P.S.; funding acquisition, P.S. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was financially supported by the Beijing Engineering and Technology Research Center of Food Additives, Beijing Technology & Business University (BTBU) and Open Foundation of Ministry of Agriculture Key Laboratory of Seaweed Fertilizers (MAKLSF1807).

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

#### **Abbreviations**

SAZMP4: purified polysaccharide from crude polysaccharide with sodium hydroxide solution (0.3 M); HPGPC, high-performance gel permeation chromatography; GC, gas chromatography; FI-IR, Fourier transform infrared spectroscopy; NMR, nuclear magnetic resonance; SEM, scanning electron microscopy; AFM, atomic force microscopy; DSC, differential scanning calorimeter; RG, rhamnogalacturonan; SOD, superoxide dismutase; GSH-Px, glutathione peroxidase; MDA, malondialdehyde; ROS, reactive oxygen species.

#### **References**


© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

### *Article* **Induction of Antioxidant Protein HO-1 Through Nrf2-ARE Signaling Due to Pteryxin in** *Peucedanum Japonicum* **Thunb in RAW264.7 Macrophage Cells**

**Junsei Taira 1,\* and Takayuki Ogi <sup>2</sup>**


Received: 11 November 2019; Accepted: 1 December 2019; Published: 5 December 2019

**Abstract:** This study focused on exploring the nuclear factor-erythroid-2-related factor (Nrf2) active compound to avoid oxidative stress related to various diseases, such as obesity and diabetes mellitus. The activity of the Nrf2-ARE (antioxidant response element) signaling was evaluated by a reporter assay involving over five hundred various edible medicinal herbs, and the highest Nrf2 activity was found in the ethanol extract of *Peucedanum japonicum* leaves. The active compound in the extract was isolated by high performance liquid chromatography (HPLC), and the chemical structure was identical to pteryxin based on 1H, 13C-NMR spectra and liquid chromatography/time-of-fright mass spectrometer (LC/TOF/MS). From the pteryxin, the transcription factor Nrf2 was accumulated in the nucleus and resulted in the expression of the antioxidant protein, heme oxygenase-1 (HO-1). In addition, the Nrf2 activity involving HO-1 expression due to coumarin derivatives was evaluated together with pteryxin. This suggested that the electrophilicity, due to the α,β-carbonyl and/or substituted acyl groups in the molecule, modulates the cysteine residue in Keap1 via the Michel reaction, at which point the Nrf2 is dissociated from the Keap1. These results suggest that pteryxin will be a useful agent for developing functional foods.

**Keywords:** coumarin; pteryxin; HO-1; Nrf2; oxidative stress; *Peucedanum japonicum* Thunb; RAW264.7 cells

#### **1. Introduction**

Some *Peucedanum* species belonging to the *Apiaceae* family contain therapeutic properties and are used in traditional medicine against various conditions, including sore throats, coughs, colds, and headaches [1]. A species of *Peucedanum japonicum* Thunb has been used as a folk medicine in Japan, Taiwan, and China, and the antioxidant and antityrosinase active compounds were found in the leaf extract of the *Peucedanum* species [2,3]. Recent studies have demonstrated that the ethanol (EtOH) extract of *P. japonicum* has an anti-obesity effect and that it contains coumarin-related compounds that the affect diabetes and obesity, both of which are bioaccessible to the systemic tissues [4–9].

Oxidative stress, with the excess production of reactive oxygen species (ROS), is related to an increased risk of developing several diseases, including obesity and diabetes mellitus. The ROS and reactive nitrogen species (RNS), due to the oxidative stress in the cells, induce antioxidant enzymes such as SOD, glutathione peroxidase, and thioredoxin (Trx) as the first line of defense. The Nrf2 (Nuclear factor-erythroid-2-related factor)-ARE (antioxidant response element) signaling responds to the cell damage with the excess production of ROS and RNS or electrophiles. The Nrf2 dissociates from the Kelch-like ECH-associated protein 1 (Keap1) by electrophiles and the oxidative stress, which regulates the expression of the ARE region containing phase II detoxifying/antioxidant enzymes, such as glutathione *S*-transferase, NAD(P)H quinone oxidoreductase-1, Trx, and heme oxygenase-1 (HO-1) [10]. The Nrf2 plays a significant role in the regulation of adipocyte differentiation, obesity, and insulin resistance [11]. Certain dietary chemopreventive agents target the Keap1 by oxidizing or chemically modifying its specific cysteine thiols, which can induce ARE-mediated genes expression [12,13]. In our previous studies, it was demonstrated that marine natural products modulate the HO-1 protein expression through Nrf2 activation in both normal cells and cancer cells [14,15].

This study assessed the Nrf2 activity in various edible medicinal plants, and the highest Nrf2 activity was found in the EtOH extract of *P. japonicum* leaves. In addition, this study shows that pteryxin was the active compound in the extract, which was enhanced by the HO-1 protein expression through the Nrf2-ARE signaling.

#### **2. Materials and Methods**

#### *2.1. Materials*

Coumarin and 3,4-dihydrocoumarin were purchased from the FUJI Firm Wako Pure Chemical Corporation (Osaka, Japan) and pyranocoumarin was obtained from Sigma-Aldrich Co. LLC (St. Louis, USA). The products of the antibodies, such as anti-Nrf2 (Santa Cruz Biotechnology, Inc., TX, USA), anti-HO-1 (StressMarq Biosciences, Inc., Victoria, Canada), and anti-β-actin (FUJI Firm Wako Pure Chemical Corporation) were used for detecting the protein expressions. The cytotoxicity was determined using 3-(4,5-dimethyl-2-thiazlyl)-2,5-diphenyltetrazolium bromide (MTT, FUJI Firm Wako Pure Chemical Corporation).

#### *2.2. Isolation of Pteryxin*

Pteryxin was isolated from the dried-leaf powder of *P. japonicum*. The dried-leaf powder of *P. japonicum* (20 g) was extracted by 50% EtOH (210 mL) using a Dionex ASE 350 accelerated solvent extractor (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The extract was loaded on a Diaion HP20 column (100 × 20 mm I.D., Mitsubishi Rayon Aqua Solutions Co. Ltd., Tokyo, Japan), then the sample was sequentially eluted, each with 100 mL of 50% and absolute EtOH. The EtOH fraction was evaporated in vacuo, and the residue (325 mg) was separated by centrifugal partition chromatography (Easy-PREPccc, 318 mL of coil column, Kutuwa Sangyo, Hiroshima, Japan) in the two-phase solvent system of n-hexane/chloroform/70% methanol (9:1:10 in *v*/*v*/*v*). Its lower layer (mobile phase) was separated at 3.0 mL/min and 1110 rpm. The collected fractions (31.8 mg) were purified on a reversed-phase chromatographic column (XBridge C18 column, 150 × 19 mm, I.D., 5 μm particle size, Waters Corp., MA, USA) at the flow rate of 12.0 mL/min by the elution of formic acid/H2O/acetonitrile (0.1/55/45 in *v*/*v*/*v*) using an HPLC apparatus (PU-980 HPLC Pump, JASCO Corp., Tokyo, Japan), then the yield (0.029%) of 5.8 mg pteryxin was purified.

#### *2.3. Analysis of Pteryxin*

The structure of pteryxin was determined by its 1H and 13C-NMR spectra (Avance III HD Ascend 400 MHz spectrometer, Bruker Billerica, MA, USA), and its molecular formula was determined by Q-TOF LC/MS (Agilent 6530 Accurate-Mass Q-TOF LC/MS system Agilent Technologies, CA, USA) on a reversed-phase chromatographic column (ACQUITY UPLC BEH C18, 50 × 2.1 mm I.D., 1.7 μM particle size, Waters Corp., MA, USA) at 40 ◦C. The mobile phase consisting of a 0.1% formic acid aqueous solution / 0.1% formic acid containing acetonitrile (1:1) was carried out at the flow rate of 0.4 mL/min by a linear gradient to 0.1% formic acid aqueous solution/0.1% formic acid containing acetonitrile (1:19) at 3 min. The high-resolution mass spectra (HRMS) was measured under the following conditions: a positive ion mode; a desolvation temperature, 350 ◦C; a desolvation pressure, 40 psig; and a desolvation gas flow, 8 L/min.

#### *2.4. Cell Culture*

RAW264.7 cells (mouse macrophages) were obtained from American Type Culture Collection (VA, USA). Cells were cultured in DMEM medium (including 10 % FBS, 100 U/mL penicillin and 100 μg/mL streptomycin) at 37 ◦C in a 5% CO2 atmosphere.

#### *2.5. Cell Survaival*

The cell viability treatment, with or without a test sample in a well, was examined by an MTT assay, as previously reported [16]. After the culture, MTT (0.05%) was added to each well and incubated for 3 h. The formazan reduced from an MTT was extracted with DMSO (100 μL) and determined as an index of the surviving cells at 570 nm using a microplate reader (BIO-RAD Model 550, BIO-RAD, CA, USA).

#### *2.6. Activity of Nrf2-ARE Signaling*

The activity of Nrf2-ARE signal treatment, with or without a test sample, was examined by a reporter assay, as previously reported (14,15). The RAW264.7 cells, with or without the test sample, were pre-cultured on a 12-well microplate (5 <sup>×</sup> 105 cells/well) for 24 h, and then the pGL4.37 [luc/ARE/Hygro] plasmid and VIOFECTINE (as the transfection reagent) were co-transfected in cells. After 24 h, the cells were washed twice in PBS and lysed in 100 μL lysis buffer. The luciferase activity of the lysate cells (50 μL) was assayed using a luciferase substrate, then the chemiluminescence (CL) in cells was measured by a microplate reader (GLOMAX MULTI Detection system, Promega, WI, USA). The protein concentration of the cells was determined using a BCA protein assay kit (Thermo Fisher Scientific, Inc., MA, USA).

#### *2.7. Nuclear Extraction and Determination of Nrf2*

The Nrf2 translocation in nucleus was examined as previously reported (14). Cells (1.0 <sup>×</sup> 106 cells/mL) with or without the compound (50 and 100 μM) were incubated for 24 h. After 24 h, the cells were treated with trypsin. The nucleus of the cells was extracted using an extraction kit (NE-PER Nuclear and Cytoplasmic Extraction Reagent, Thermo Fisher Scientific K.K, Yokohama, Japan). Nrf2 protein was detected using anti-Nrf2 by a Western blot analysis, and the Nrf2 protein expression was expressed as % of the untreated sample cells (control).

#### *2.8. Protein Expression by Western Blot Analysis*

The assessed cells were treated with the lysis buffer. The cellular lysates were centrifuged at 13,800 *g* for 5 min. The cellular extracts were separated on SDS-polyacrylamide gels (4–12% SDS-polyacrylamide, Invitrogen, CA, USA) and transferred to a nitrocellulose membrane (iBlot Gel Transfer Mini, Invitrogen) using an iBlot Gel Transfer Device (Invitrogen). The protein was detected with the antibodies, such as Nrf2 and HO-1, and the protein expression was determined by densitometry analysis.

#### **3. Results**

#### *3.1. Determination of Pteryxin*

The chemical structure of (+)-pteryxin ([α] 24 <sup>D</sup> = 10.9◦ (*c* 0.13, EtOH)) was determined by the following 1H and 13C NMR spectra. The 1H NMR (400 MHz, CDCl3, δ<sup>H</sup> 7.26): 7.59 (1H, d, *J* = 9.5 Hz, H-4), 7.35 (1H, d, *J* = 8.6 Hz, H-5), 6.80 (1H, d, *J* = 8.6 Hz, H-6), 6.63 (1H, d, *J* = 5.0 Hz, H-4- ), 6.22 (1H, d, *J* = 9.5 Hz, H-3), 6.03 (1H, qq, *J* = 7.2, 1.5 Hz, H-3"), 5.35 (1H, d, *J* = 5.0 Hz, H-3- ), 2.09 (3H, s, OCOCH3), 2.00 (3H, dq, *J* = 7.2, 1.5 Hz, H-4"), 1.86 (3H, dq, *J* = 4.5, 1.5 Hz, H-5"), 1.46 (3H, s, 2- -CH3), 1.43 (3H, s, 2- -CH3). 13C NMR (100 MHz, CDCl3, δ<sup>C</sup> 77.0): δ<sup>C</sup> 169.8 (OCOCH3), 166.9 (C-1"), 159.7 (C-2), 156.6 (C-7), 154.0 (C-9), 143.1 (C-4), 137.9 (C-3"), 129.2 (C-5), 127.4 (C-2"), 114.4 (C-6), 113.3 (C-3), 112.5 (C-10), 107.3 (C-8), 77.3 (C-2- ), 70.5 (C-3- ), 60.1 (C-4- ), 25.3 (2- -CH3), 22.2 (2- -CH3), 20.7 (OCOCH3), 20.4 (C-5"), 15.6 (C-4- ).

In addition, the molecular formula was established as C21H22O7 on the basis of HRMS (*m*/*z* 387.1448 [M+H]+, calcd. 387.1438, and *m*/*z* 409.1262 [M+Na]+, calcd. 409.1258). The chemical structure of pteryxin is shown in Figure 1. Pteryxin is an angular-type khellacton coumarin substituted acyl groups. In this study, the Nrf2 activity of the major types of coumarins, such as angular type pyranocoumarin, simple coumarin, and its dihydotype coumarin (3,4-dihydrocoumarin), were assessed together with the pteryxin (Figure 1).

**Figure 1.** Chemical structures of pteryxin and the derivatives used in this study.

#### *3.2. Nrf2-ARE Signaling*

*P. japonicum* leaves were extracted with ethanol, and the cytotoxicity of the extract was evaluated by an MTT assay. As shown in Figure 2a, the cytotoxicity of the extract was not detected in the range of the test concentrations. The various concentrations (100, 200 and 400 μg/mL) of the extract were evaluated for the activity of the Nrf2-ARE signaling in RAW264.7 cells (Figure 2b). The extract was significantly activated in a dose-dependent manner, suggesting that the Nrf2 activator is contained in the extract (Figure 2b).

**Figure 2.** Activation of the Nrf2 (Nuclear factor-erythroid-2-related factor)-ARE (antioxidant response element) signaling in the presence of the ethanol extract of *Peucedanum japonicum* Thunb leaves in RAW264.7 macrophage cells. (**a**) Cell viability with treated samples at the test concentrations was examined by an MTT assay. The cell viability was expressed as % of the control cells without sample. The Nrf2-ARE signaling activity of the EtOH extract of the *P. japonicum* leaves was evaluated by the reporter assay as described in the text. (**b**) The effect of the various concentrations (100–400 μg/mL) of the EtOH extract of *P. japonicum* leaves in RAW264.7 cells. The activity (%) was indicated as % of induction for the control cells without a sample. Data were expressed as mean ± SD, and the significant difference was analyzed by the student's *t*-test. \* *p* < 0.01 indicated as a significant difference from the control.

Pteryxin was isolated from the ethanol extract of *P. japonicum* leaves. As shown in Figure 3a, the cytotoxicity of pteryxin was not detected in the range of the test concentrations. When pteryxin was placed in the Nrf2-ARE signaling cell system, the Nrf2 activity was significantly induced in the concentration range of 25–100 μM, indicating that pteryxin is one of the active compounds in the *P. japonicum* leaves (Figure 3b).

**Figure 3.** Nrf2-ARE signaling activity due to pteryxin in RAW264.7 macrophage cells. The Nrf2-ARE signaling activity in the presence of pteryxin (25–100 μM) was evaluated by the reporter assay, as described in the text. (**a**) Cell viability with treated samples at the test concentrations was examined by an MTT assay. The cell viability was expressed as % of control cells without a sample. (**b**) The effect of the various concentrations (25–100 μM) of pteryxin in RAW264.7 cells. The activity was indicated as induction (%) for control cells without sample. Data were expressed as mean ± SD, and the significant difference was analyzed by the student's *t*-test. \* *p* < 0.01 indicated a significant difference from the control.

#### *3.3. Nrf2 Expression in Cytoplasm and Nuclei*

The expression of cytoplasmic Nrf2 in the presence of a compound was determined by Western blot analysis. The Nrf2 was accumulated in the cytoplasm in a dose-dependent manner, and the accumulation of the transcription factor Nrf2 was also detected in the nucleus, suggesting that the cytoplasmic Nrf2 translocated into the nucleus through the Nrf2-ARE signaling (Figure 4a,b). Consequently, the transcription factor Nrf2 would be activated on the ARE regions, resulting in the expression of the HO-1 protein.

#### *3.4. HO-1 Expression*

The expression of the antioxidant protein HO-1 in the presence of the target compound was determined by Western blot analysis. The protein expression increased in the concentration range of 25–100 μM (Figure 5a,b). Consequently, the Nrf2 in the nucleus was enhanced by the HO-1 protein expression (Figure 4). This result indicates that the pteryxin plays a significant role in delaying the oxidative stress in a biological system.

#### *3.5. Nrf2-ARE Signaling Activity and HO-1 Protein Expression by Coumarin Derivatives*

The Nrf2-ARE signaling activity of pteryxin was compared to the other types of coumarins, such as simple coumarin, 3,4-dihydrocoumarin, and pyranocoumarin. The Nrf2 activity of these compounds was determined by the reporter assay (Figure 6a). As a result, the Nrf2 activity was detected by the α,β−carbonyl coumarins, except for 3,4-dihydrocoumarin, which suggested that the electrophyllicity in the molecule contributes to the cysteine thiol oxidation of Keap1 and leads to the activation of the Nrf2-ARE signaling. Particularly, the activity of pteryxin, which was the highest of the evaluated coumarin derivatives, indicated that the structure of khellacton is suitable for the electrophyllicity.

**Figure 4.** Expression of the transcription factor Nrf2 protein due to pteryxin in RAW 264.7 macrophage cells. The Nrf2 protein expression in the presence of a compound was detected by western blot analysis and determined by densitometry. (**a**) Cytoplasmic Nrf2 protein expression and (**b**) nuclear Nrf2 protein expression in the presence of a compound. Data were expressed as mean ± SD, and the significant difference was analyzed by the student's *t*-test. \* *p* < 0.01 and \*\* *p* < 0.05 indicated a significant difference from the control.

**Figure 5.** HO-1 (heme oxygenase-1) expression due to pteryxin in RAW 264.7 macrophage cells. The HO-1 protein expression due to pteryxin (25–100 μM) on the Nrf2-ARE signaling in the cells was examined. (**a**) Western blot analysis of the HO-1 protein in the presence of a compound. (**b**) Densitometry analysis of the expression of the HO-1 protein. Data were expressed as mean ± SD, and the significant difference was analyzed by the student's *t*-test. \* *p* < 0.01 indicated a significant difference from the control.

**Figure 6.** Activation of Nrf2-ARE signaling and the HO-1 protein expression due to coumarin derivatives in the RAW 264.7 macrophage cells. (**a**) The activation of Nrf2-ARE signaling due to the coumarin derivatives (50 μM) was assessed by the reporter assay. (**b**) The HO-1 protein expression due to the various coumarins (50 μM) was detected by Western blot analysis. The expression of the HO-1 protein was determined by a densitometry analysis. Data were expressed as mean ± SD, and the significant difference was analyzed by the student's *t*-test. \* *p* < 0.01 indicated a significant difference from the control.

The HO-1 protein expression, due to the structurally different coumarins, was assessed together with pteryxin. The pteryxin and pyranocoumarin, in common with an angular-type skeleton, presented a high expression, which had a similar result to that of the Nrf2 activity (Figure 6b). Particularly, the pteryxin, which consisted of khellacton-substituted acyl groups, had the highest Nrf2 activity involving the HO-1 protein expression. These results suggest that the difference in the Nrf2 activity due to the compounds may be dependent on the individual potential electrophyllicity.

#### **4. Discussion**

*Peucedanum* species are used as a traditional medicine for sore throats, coughs, colds, and headaches [1]. Recent studies have demonstrated that the extract of *P. japonicum* Thunb plays a role in suppressing obesity. Particularly, some coumarins in the *Peucedanum* species have been examined based on their anti-diabetes and anti-obesity activities [4,7,8]. In addition, several anti-obesity components, including pteryxin and the other coumarin derivatives, were found in *P. japonicum* Thunb, and they were mainly exerted by inhibition of lipogenesis in the adypocytes.

A more recent study indicated that the oxidative stress in the hypothalamus induces insulin resistance and obesity. The Nrf2 activity then suppressed the hypothalamic oxidative stress, subsequently improving the resistance of insulin and leptin related to obesity [17]. This study demonstrated that the leaf extract of *P. japonicum* Thunb has the physiological function of activating the Nrf2-ARE signaling to avoid cell damage by the excess production of the reactive oxygen species (ROS) under oxidative stress (Figure 2). In addition, the pteryxin, as one of the main Nrf2 activators in the extract, dissociated the Nrf2 from Keap1, and then the Nrf2 translocated into the nucleus, activated the ARE region containing the promoter and enhancer regions-mediated antioxidant enzyme, HO-1 (Figures 3–5). Some α,β carbonyls in the molecule will be potential electrophiles that react with the nucleophile protein, Keap1 [18]. When the cysteine residue in the Keap1 is oxidized by an electrophile, the Nrf2 part from Keap1 binds to the ARE region in the DNA sequences. Our results, together with previous knowledge, suggest that the coumarins that are effective against diabetes mellitus and obesity may also act in conjunction with the Nrf2 activity [8].

Coumarins are a large class of plant secondary compounds with a benzopyrone skeleton; they are distributed across four major sub-types: simple coumarins, furanocoumarins, pyran-substituted coumarins, and pyranocoumarins. The pteryxin is an angular-type pyranocoumarin (khellacton coumarin) substituted acyl group (Figure 1). In this study, the Nrf2 activity of the structurally different coumarins—angular-type pyranocoumarin, simple coumarin, and 3,4-dihydrocoumari were examined together with the pteryxin, and the activation of the Nrf2 activity was detected by the α,β carbonyl coumarins, except for 3,4-dihydrocoumarin (Figure 6a). The HO-1 expression was also similar to the result of the Nrf2 signaling activity (Figure 6b). A previous study demonstrated that the α,β carbonyl in 1,2-naphtoquintione is an electrophile that resulted in the nucleophile protein, Keap1, which is added to the carbon β by Michael addition [18]. A similar result was obtained, except for 3,4-dihydrocoumarin, which suggested that the α,β carbonyl in the molecule plays a significant role as an electrophile for the Nrf2 activator. Also, pteryxin and pyranocoumarin indicated a high Nrf2 activity, which suggested that the angular skeleton will be an effective structure for an Nrf2 activator. In addition, the pteryxin, consisting of khellacton-substituted acyl groups, indicated the highest Nrf2 activity involving HO-1 expression, suggesting that the acyl groups may also contribute to the electrophyllicity in the molecule.

Choi et al. reported the anti-adipogenic and anti-diabetic effects of cis-3,4-diisovalerylkhellactone isolated from *P. japonicum* Thunb [8]. This khellacton coumarin substituted acyl group has a structure similar to that of pteryxin, which may have potential Nrf2 activity. In a previous study, the effect anti-obesity effect due to pteryxin was elucidated through animal testing [14]. The Nrf2 plays a significant role in the regulation of obesity and insulin resistance [11]. The Nrf2 activity suppressed the hypothalamic oxidative stress, resulting in the improvement of the resistance of insulin and leptin related to obesity [17]. Thus, the Nrf2 active function will play a key role in prohibiting obesity and diabetes mellitus. In addition, pteryxin will be a useful agent for functional foods, preventing the metabolic syndrome based on insulin resistance.

#### **5. Conclusions**

In this study, the highest Nrf2 activity was found in EtOH extract of *P. japonicum* leaves, and its Nrf2 active compound was identical to that of pteryxin. The accumulation of the Nrf2 in the nucleus due to pteryxin induced the expression of the antioxidant protein, HO-1. In addition, the Nrf2 active function, due to pteryxin, was suggested to hold electrophillicity due to the α,β-carbonyl and/or substituted acyl groups in the molecule modulating the dissociation of Nrf2 from the Keap 1.

**Author Contributions:** All authors participated in the design of the study. T.O. isolated pteryxin used in this study; J.T. organized this study including experiment, interpreted the data, and wrote the manuscript.

**Funding:** Works described here were supported by Grants from Okinawa Industrial Technology Center (OITC) and Okinawa Science and Technology Promotion Center (OSTC).

**Acknowledgments:** We acknowledge graduate students in our laboratory, Okinawa College for their technical support.

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

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

*Article*
