Introduction

Among the many proteins and metabolites secreted by adipocytes, adiponectin serves a unique role as a beneficial, anti-diabetic, insulin-sensitizing factor that has anti-atherogenic and anti-inflammatory properties in the vasculature [1]. Chronic inflammation is demonstrated by an increased expression of pro-inflammatory cytokines and elevated infiltration of macrophages into adipose tissue. Among the pro-inflammatory cytokines, tumor necrosis factor-alpha (TNF α) decreases insulin sensitivity and disrupts the homeostasis of lipid and glucose metabolism [2, 3]. A decline in adiponectin (ACRP30) production contributes to pathogenesis of insulin resistance [4]. Oxidative stress has been implicated in pathogenesis, such as in complications of insulin resistance in type 2 diabetes [5, 6]. Excessive generation of reactive oxygen species (ROS) is the underlying mechanism of type 2 diabetes, resulting in an accelerated rate of apoptosis in growth-arrested cells [7]. In addition, reduced plasma adiponectin levels have been described in patients with diabetes, presumably as a result of an increase in ROS [8, 9]. It was demonstrated that the functional consequence of ROS production in adipocytes and the associated insulin resistance and alterations in serum levels of adiponectin, revealed that the increases in ROS are associated with an induced inflammatory response in the adipocytes [10].

Several kinds of seed and their alcoholic or aqueous extracts exerted an anti-hyperglycemic effect on humans, rats and mice [11, 12]. Tanaka et al. [11] proved an anti-hyperglycemic action of the seeds of Eriobotrya japonica, traditionally used as a Kampo medicine called “pipaye” in Chinese and “biwayo-to” in Japanese, and indicated that this effect might be due to the antioxidant effect of flavonoids in the seeds. In addition, Niwano et al. [13] investigated anti-obese characteristics of wide range of traditional Okinawa foodstuffs. Consequently, four foodstuffs showed clear facilitating effect on lipid degradation in 3T3-L1 adipocytes. Two of four active foodstuffs were fruit parts of Citrus species. It has been reported that citrus species are rich sources for various bioactive compounds such as flavonoids, adrenergic amines, limonoids, and coumarines. Thus, phenolic compounds in the seeds may be the cause of the anti-hyperglycemic effects. Since our final goal is to develop dietary supplements with desirable characteristics using novel functional foodstuffs with potent anti-diabetic activity, we conducted a large scale screening for prevention of TNF α level and the promotion of adiponectin in vitro as a first step. Furthermore, this study assessed the DPPH radical scavenging activities in a variety of dried plant seeds.

Materials and Methods

Reagents

We purchased reagents from the following sources: pre-adipocyte medium (PM), adipocyte maintenance medium (AMM) and adipocyte differentiation medium (ADM) from DS Pharma Biomedical Co., Ltd. (Osaka, Japan); (+)-catechin and 2, 2-diphenyl-1-picryl-hydrazyl (DPPH) from Sigma-Aldrich (St. Louis, MO); mouse adiponectin immunoassay kit and TNF α immunoassay kit from R&D Systems, Inc. (Minneapolis, MN, USA). We obtained all other analytical grade chemicals from Wako Pure Chemical Industries, Ltd. (Osaka, Japan).

Preparation of Ethanol Extracts from the Dried Plant Seeds (PSEE)

We collected 20 kinds of dried plant seeds from the market in Japan (Table 1) and pulverized dried materials in a millser. We added 70% ethanol to each pulverized material sample in a proportion 2–5 times its volume, and mixed it for 24 h. We then filtered the liquid layer of the mixture through a filter paper, and stored the collected filtrate below −20 °C until use. When we used the filtrate in assays, each sample was first returned to room temperature, followed by filtration through a membrane filter (pore size 0.2 μm).

Table 1 Specifics on the color and type of seeds used
Full size table

Determination of Total Phenolics

We measured total phenolics with a modified version of the Folin-Ciocalteu method [14] using 0–30 mg/l (+)-catechin as a standard. Briefly, 100 μl of sample or standard was combined with 100 μl of Folin-Ciocalteu reagent and 100 μl of 2% Na2CO3 solution. We allowed the mixture to sit for 60 min before reading absorbance at 750 nm and calculating the concentration of PSEE as (+)-catechin equivalents per gram of dried plant seed [15].

DPPH Radical Scavenging Activity

Employing the procedure described by Negro [16], we measured the free radical scavenging activity of PSEE, dissolving the PSEE in water to obtain a series of sample solutions of different concentrations (0–30 mg dry weight of seed/ml of extract solution). After mixing a sample solution (100 μl) with 100 μl of 800 μM DPPH-ethanol solution and waiting 30 min, we measured the absorbance of the sample solution at 520 nm against a blank. DPPH radical scavenging activity was exhibited as follows. The 50% inhibitory concentration (IC50; concentration of sample required to scavenge 50% of DPPH radicals) values were determined by the method of probit-graphic interpolation for five concentration levels. All tests and analyses were run in triplicate and averaged.

Cell Culture

We obtained mouse 3T3-L1 cells (JCRB9014) from the Health Science Research Resources Bank (Osaka, Japan) and adjusted them to 20,000 cells/ml in PM, planting them at a ratio of 200 μl of the cell suspension into 96-well culture plates and incubating them in an atmosphere of humidified 5% CO2 at 37 °C for 4 days. Thereafter, we induced differentiation by addition of ADM. This mixture was replaced with fresh AMM after 48 h. We then changed the medium every 2 days until the cells were used for experiments and on day 4 after the induction of adipocyte differentiation, added 2 or 4 μl PSEE to the culture medium (the maximum concentration usable without inducing cytotoxicity) for the purpose of preparing the mature adipocytes in vitro. Then, at 48 h after the addition of PSEE, we determined the amount of secreted adiponectin and TNF α in the culture mediums of adipocytes with and without PSEE.

Measurement of TNF α, and Adiponectin Concentrations in Culture Media

We measured the amounts of adiponectin in culture medium of adipocytes using an quantikine mouse adiponectin/Acrp30 immunoassay kit and determine the TNF α levels in culture medium with a commercially available ELISA kit using an Ultrospec Visible Plate Reader II 96 (GE Healthcare Ltd., England). The adiponectin and TNF α levels in the samples were calculated by the following equations.

$$ {\hbox{Adiponectin }}\left( {{\hbox{ng}}/{\hbox{ml}}} \right) = {1}.{7499} \times {\hbox{A45}}0{\hbox{nm}} \times {\hbox{A45}}0{\hbox{nm}} + {1}.{14}0{1} \times {\hbox{A45}}0{\hbox{nm}} $$
$$ {\text{TNF}}\,\alpha \,\left( {{{{\text{pg}}} \mathord{\left/{\vphantom {{{\text{pg}}} {{\text{ml}}}}} \right.\kern-\nulldelimiterspace} {{\text{ml}}}}} \right) = 646.71 \times \left( {{\text{A}}450{\text{nm}} - 0.02} \right) + 0.7924 $$

Statistical Analysis

We present all data as mean ± SD. The statistical comparison between the groups was carried out using ANOVA or Students t-test. P values <0.05 were considered statistically significant.

Results and Discussion

Total Phenolic Content of Ethanol Extracts of Dried Seeds

Previously, we published two papers describing the characteristics of broad beans [17, 18]. It was found that extracts from the broad beans possess antioxidant and cellular growth properties. We furthered this research by examining dried seeds of plants marketed in Japan, to determine whether or not they also exhibit the above-mentioned activities. As is shown in Fig. 1, the total phenolic content in the ethanol extracts of most of the studied plant seeds was high and varied significantly. The total phenolic content varied among species from 0.67 to 4.63 mg (+)-catechin equivalents/g (Fig. 1). The ethanol extracts from pumpkin seeds contained the lowest amounts. Bitter melon seeds of the species Cucurbitaceae contained roughly two times the total phenolic content of pumpkin or cucumber seed. Although the reasons for this are unknown, the pumpkin seeds may be due to contain the lowest amounts in the species Cucurbitaceae for the following reason. It may be the differences on the distribution of fatty acids and sterols in the seed’s content, between pumpkin and bitter melon. Moreover, the pumpkin seeds may be containing similar fatty acid and sterol with bitter melon. However, there are some differences in their phenolic profile, which also reflect their seeds’ differences. Edible burdock seeds contained the highest amount of total phenolics (4.63 mg (+)-catechin equivalents/g) followed by Crown daisy, Japanese radish and komatsuna seeds. Oomah et al. [19] evaluated for variability in phenolic constituents of eight lupin (Lupinus angustifolius L.) genotypes grown at four locations in south central Alberta in 2004. Phenolic compounds in genotypes varied minimally from 11.9 to 14.7 mg catechin equivalent for total phenolic contents. Our results showed total phenolic content for the lupins to be at least two times lower than the values found in their studies.

Fig. 1
figure 1

Total phenolic content of ethanol extracts from the seeds of 20 dried plants marketed in Japan. Each end of the vertical bar represents the standard error of multiple (3) samples

Full size image

DPPH Radical Scavenging Activities of Ethanol Extracts of Dried Seeds

Table 2 shows the DPPH radical-scavenging activities (IC50 values) of PSEE and the catechin. The IC50 values of PSEE were 0.025 ~ 35.6 μg/ml to DPPH radical-scavenging activities. The IC50 value of the scavenging activity was the lowest in the pumpkin seeds (0.025 μg/ml), showing a high scavenging activity. This showed a low value from catechin (0.98 μg/ml), which was used as the positive control. The scavenging activity decreased in the order of the cucumber (0.030 μg/ml), lettuce (0.065 μg/ml), potherb mustard (0.084 μg/ml), and broccoli seeds (0.105μg/ml). The PSEE clearly displayed scavenging activity against the DPPH radical (Table 2). It was also observed that the radical scavenging activities in ethanol extracts of most of the plant seeds in this study were high and varied significantly. The effect of pumpkin seeds on scavenging DPPH radicals is 0.025 μg/ml, which is stronger than that of the Eriobotrya japonica seed extract (0.96 μg/ml) [20]. The results reveal that the PSEE could be used as free radical inhibitors and primary antioxidants. As other, we have found phenolic content consistent with the DPPH radical scavenging activity [21] for seed extracts other than green soybean, pumpkin, cucumber, and lettuce. The correlation coefficient between the extracted phenolic content and scavenging activity was 0.803 (data not shown). In the case of pumpkin, cucumber and lettuce ethanol extracts, although they contained lower phenolic amounts, they showed aggressive antioxidant activity with regard to the DPPH assay (IC50≦0.065 μg/ml). Moreover, in the case of green soybean extracts, although they contained higher phenolic amounts, they showed a low scavenging activity (IC50 = 35.6 μg/ml). As Liolios et al. [22] have said, this can be explained by the fact that phenolic content does not necessarily correspond to antioxidant activity, which is the result of the combined action and the cooperation of a wide range of compounds, including phenolics (free, glycosides, oligomers and polymers), peptides, organic acids and other components. The difference in phenolic content and antioxidant activity could also be due to the biochemical properties of seeds or to minute differences in the biochemical pathways. Although the green soybean, pumpkin, cucumber, and lettuce seeds contain fatty acids, for example, there are some differences in their phenolic profile.

Table 2 DPPH radical-scavenging activity in ethanol extracts from the seeds of 20 dried plants marketed in Japan as determined by the IC50, expressed as μg dry seed weight/ml extract solution. All data were average (±SD) of three replicates
Full size table

Adiponectin Production

Results of adiponectin production of PSEE on mouse adipocytes are summarized in Fig. 2. Adiponectin/Acrp30 immunoassay kit results showed that adiponectin production levels significantly increased in the 48-h PSEE-preloaded adipocytes relative to control cells. Results showed that secreted adiponectin increased 2.5-fold in 2 μl komatsuna compared to ethanol control, and adiponectin increased 4.0-fold in 4 μl green soybean (P < 0.01; Fig. 2 a and b). The results also revealed that the two additional volumes of PSEE, a low volume of 2 μl and a high volume of 4 μl were almost all secreted in similarly increased levels from preloaded vs. control cells (1.2 ~ 2.5- and 1.1 ~ 4.0-fold, respectively). At concentrations above 2 μl, however, adiponectin-production stimulation was reduced at the 4 μl dose of PSEE in comparison with the 2 μl PSEE. Despite the positive outcome of adiponectin-production stimulation with the PSEE dose of 2 μl, the present findings indicate that PSEE was associated with a dose–dependent inhibition of the adiponectin production. Perhaps, there may be the optimum concentration in less than 2 μl for adiponectin production in the cells. On the other hand, no significant effect on adiponectin release from adipocytes was elicited by preloading with Japanese radish, edible burdock, bitter melon or broccoli for 48 h durations.

Fig. 2
figure 2

Effects of dried plant seed ethanol extracts on adiponectin production in 3T3-L1 adipocytes. Dried plant seed ethanol extracts at 2 μl (a) and 4 μl (b) regulated adiponectin secretion. The adiponectin production activities of ethanol (control) and dried plant seed ethanol extracts are indicated by unshaded and shaded columns, respectively. Values are the mean ± SD of three measurements. **P < 0.01, *P < 0.05 compared with controls

Full size image

Ethanol extracts at 2 μl from komatsuna and common bean exhibited the highest adiponectin production activity on adipocytes. It increased by 12.7 and 12.1 ng/ml, respectively, and the ethanol extracts from Japanese radish and bitter melon at the same concentration showed this activity lowered by only 0.91 and 1.1 ng/ml, respectively. These results, however, were not consistent with the findings of DPPH radical scavenging activity and/or total phenolic content. Zhang et al. [23] indicated that a stilbene constituent (hydrangeic acid) promoted adipogenesis of 3T3-L1 cells. Hydrangeic acid significantly increased the amount of adiponectin released into the medium, and the translocation of glucose transporter 4 (GLUT4). Hydrangeic acid also increased mRNA levels of adiponectin, peroxisome proliferator-activated receptor gamma2 (PPARgamma2), GLUT4, and fatty acid-binding protein (aP2) while it decreased the expression of TNF α mRNA. However, it did not activate PPAR gamma directly different from troglitazone in a nuclear receptor cofactor assay system. Since some seeds [24, 25] contain the stilbene, it may be that the extracts from the seeds used in our study would have a similar composition to the stilbene used in their studies. However, it would seem that different components of PSEE may affect adiponectin production in various ways.

TNF α Levels

Studies have suggested a correlation between high adiponectin levels, and low TNF α levels [26]. Therefore, we investigated the probability that PSEE induce insulin sensitivity by reducing the secretion of TNF α. Figure 3 shows that PSEE treatment reduced TNF α secretion by about 100% (P < 0.01). Although increases were detected after 2 μl spinach and 4 μl okra PSEE treatments (Fig. 3a and b), all adiponectin-production-enhancing seeds except for 2 μl spinach and 4 μl okra effectively reduced secretion of the TNF α. However, increased levels of TNF α in 2 μl spinach were also reduced by 4 μl spinach extract treatment. The precise mechanism by which okra affects TNF α secretion remains to be determined. These indicate that spinach and okra extracts are not exceptionally responsible for TNF α-induced inhibition of adiponectin production in adipocytes. At concentrations above 2 μl, additionally, TNF α levels were not reduced at the 4 μl dose of PSEE in comparison with the 2 μl PSEE. Despite the positive outcome of TNF α inhibition with the PSEE dose of 2 μl, the present findings indicate that PSEE was not associated with a dose–dependent inhibition of the TNF α secretion. Perhaps, there may be the optimum concentration in less than 2 μl for TNF α inhibition in the cells. Our studies demonstrate that PSEE is capable of increasing secreted adiponectin production in adipocytes, when TNF α production is markedly inhibited in the adipocytes. Importantly, blockade of the TNF α action effectively reversed the decrease in adiponectin secretion. Adiponectin and TNF α mutually regulate expression. Down-regulation of TNF α has been proposed as a mechanism for the protective effects of saturated fatty acids, and the pungent constituent of ginger [27, 28]. Moreover, Yoshida et al. [29] reported that the flavonoids (hesperetin and naringenin) prevent TNF α from down-regulating the transcription of two antilipolytic genes, perilipin and PDE3B. These effects are mediated through the inhibition of the extracellular-signal-regulated kinase (ERK) pathway. Although the precise mechanism by which PSEE affects TNF α inhibition remains to be determined, it may be due to a mechanism occurring in vitro by PSEE, such as fatty acids and/or flavonoids.

Fig. 3
figure 3

Effects of dried plant seed ethanol extracts on TNF α levels in 3T3-L1 adipocytes. Dried plant seed ethanol extracts at 2 μl (a) and 4 μl (b) regulated TNF α levels. The TNF α levels of ethanol (control) and PSEE are indicated by unshaded and shaded columns, respectively. Values are the mean ± SD of three measurements. **P < 0.01, *P < 0.05 compared with controls

Full size image