Inhibition of lipid digestion by β-glucanase-treated Candida utilis

Abstract

β-Glucanase-treated Candida utilis (GT) has the potential to prevent cardiovascular disease by suppressing elevated postprandial serum triglyceride levels. To clarify the mechanism, in this study, we investigated the effects of GT on lipid digestion. To evaluate the inhibitory activity of pancreatic lipase and the binding ability of bile acid, various yeast samples were prepared with or without β-glucanase treatment, by two-step in vitro digestion, and by water-soluble/insoluble fractionation. The insoluble fraction of GT displayed lipase inhibition and bile acid-binding effects after in vitro digestion, and insoluble fibers and resistant proteins or peptides were thought to be the components involved. The soluble fraction of GT showed slight effects.

Introduction

According to the WHO fact sheet, cardiovascular disease (CVD) is the leading cause of death worldwide, accounting for 31 % of all deaths in 2016. The development of CVD results from lifestyle diseases and fasting and postprandial serum triglyceride levels are known risk factors for CVD (Iso et al., 2001; Iso et al., 2014). Therefore, it is important to suppress elevated postprandial serum triglyceride levels because it helps prevent CVD. Previously, several food materials and ingredients have been reported to inhibit and delay the digestion and absorption of dietary lipids (Kagawa et al., 1998; Kishimoto et al., 2007; Khossousi et al., 2008; Takagaki et al., 2018).

The cell walls of plants, algae, fungi, and yeast contain dietary fiber, which can provide health benefits such as lowering cholesterol levels (Bell et al., 1999; Sima et al., 2018), stimulation of the immune system (Volman et al., 2008), and regulation of blood glucose (Cassidy et al., 2018). In addition, water-soluble dietary fibers derived from cereals and algae are effective in reducing postprandial serum triglyceride levels (Talati et al., 2009; Yoshinaga and Mitamura, 2019). The cell wall of yeast is primarily composed of mannoproteins, β-glucans, and chitin. Yeast-derived β-glucans, composed of β-1,6-branched and β-1,3-linked linear glucose polysaccharides, are usually water-insoluble but can be solubilized using acid/alkali, enzymatic, or radiation degradation (Klis et al., 2006; Khan et al., 2016). Watersoluble dietary fiber derived from yeast may also regulate postprandial serum triglyceride levels.

Candida utilis    is an edible yeast approved by the U.S. Food and Drug Administration as a safe food additive and is used in many processed foods and as a natural seasoning (Bekatorou et al., 2006). Recently, we reported that β-glucanase-treated C. utilis (GT) suppresses elevated postprandial triglyceride levels (Sakurai et al., 2020a, 2020b). However, little is known about the factors involved, and the underlying mechanism is poorly understood. This study investigated the effects of dietary triglycerides on digestion in vitro to elucidate the mechanism underlying the reduction in postprandial serum triglyceride levels using GT.

Materials and Methods

β-Glucanase treatment of C. utilis    GT was prepared according to a previously described method (Sakurai et al., 2020b). Dry powdered C. utilis (ORG whole) was obtained from Mitsubishi Corporation Life Sciences Ltd. (Tokyo, Japan). The suspension of dry yeast powder (10 %, w/w) in distilled water was adjusted to pH 6.5 with HCl. β-Glucanase (Denazyme GEL-L1/R; Nagase ChemteX Corp., Osaka, Japan) was added at 0.3 % (w/w of yeast) and the mixture was incubated at 60 °C for 5 h. After the pH was adjusted to 7 with NaOH, the mixture was heated at 90 °C for 10 min to inactivate the enzyme, which was then freeze-dried. The freeze-dried powder (GT whole) was resuspended (10 %, w/w) in boiling water and centrifuged (HITACHI himac CR22E, R10A3 rotor; Koki Holdings Co., Ltd., Tokyo, Japan) for 10 min at 3 000 rpm. The precipitate and supernatant were freeze-dried separately.

Preparation of in vitro digests    In vitro digests of yeast were prepared according to the method described by Ma and Xiong (2009) with some modifications. The pH of the suspension of the yeast sample (10 %, w/w) in distilled water was adjusted to 2 using HCl. Pepsin (1:10 000; Nacalai Tesque, Inc., Kyoto, Japan) at a concentration of 4 % (w/w of yeast) was added, and the mixture was incubated while stirring at 37 °C for 2 h. After the pH was adjusted to 8 with NaOH, pancreatin (FUJIFILM Wako Pure Chemical Corp., Osaka, Japan) at a concentration of 4 % (w/w of yeast) was added and the mixture was incubated while stirring at 37 °C for 3.5 h. After the pH was adjusted to 7 with HCl or NaOH, the mixture was heated at 98 °C for 10 min to inactivate the enzymes. The digest was centrifuged (HITACHI himac CR22E, R10A3 rotor; Koki Holdings Co., Ltd.) for 10 min at 7 000 rpm. The precipitate and supernatant were freeze-dried separately. The freeze-dried powder of the supernatant was dissolved in distilled water, transferred into a dialysis bag (500–1 000 Da MWCO; Repligen Corp., Waltham, MA, USA), and dialyzed against distilled water. The dialyzed inner liquid was freeze-dried. The solubility of the yeast samples (undigested and in vitro digest) was calculated as a percentage using the following formula: (%) = [1 − (weight of the freeze-dried powder of precipitation)/(initial weight of yeast samples)] × 100.

Measurement of molecular weight distribution    The molecular weight distribution and quantity of high-molecular components retained inside the dialysis bag were measured using size-exclusion chromatography (SEC). Yeast samples were suspended in distilled water and centrifuged for 3 min at 15 000 rpm. The supernatant was used for HPLC analysis after filtering through a 0.45 µm membrane filter. HPLC was performed using an Agilent 1200 system (Agilent Technologies, Santa Clara, CA, USA) and an RI detector. A YMC Pack Diol 300 (5 µm, 300 × 8.0 mm I.D.; Nacalai Tesque, Inc.) column was used with ultra-pure water and the flow rate was maintained at 0.7 mL/min. Pullulan was used as the reference standard (molecular weights: 5.8, 12.2, 23.7, 48, 100, 186, and 380 kDa, SHOWA DENKO K.K., Tokyo, Japan). The concentration of dialyzed samples for evaluation was adjusted to the same quantity as before dialysis.

Measurement of viscosity    The viscosity of yeast samples or dietary fiber dissolved in 0.1 M Tris-HCl (pH 7) was measured using a Tuning Fork Vibro Rheometer (RV-10000; A&D Co., Ltd., Tokyo, Japan). Briefly, 10 mL of sample solution was used to measure static viscosity (V_s), which was converted from the vibration resistance, with 1.0 mm of amplitude for 5 min. The viscosity of each sample was calculated from the initial static viscosity values using the following equation: viscosity = V_s/d, where d is the density of the sample. All measurements were performed at 37.0 °C ± 0.5 °C, and each sample was measured three times. Psyllium (Nippon Garlic Corp., Gunma, Japan), pectin (Sansho Co., Ltd., Osaka, Japan), and gum arabic (Nacalai Tesque, Inc.) were used for comparison.

Determination of pancreatic lipase inhibition    The effect of yeast samples on pancreatic lipase activity was evaluated according to the method of Sakae and Sekizaki (2014) with some modifications. A lipid emulsion was prepared as follows: 2 g of olive oil (Ajinomoto Co., Inc., Tokyo, Japan), 250 mg of lecithin from soybean (FUJIFILM Wako Pure Chemical Corp.), and 100 mg of sodium cholate (FUJIFILM Wako Pure Chemical Corp.) were diluted to 50 mL with Tris buffer (0.1 M Tris-HCl, 0.1 M NaCl, pH 7), and sonicated for 5 min using an ultrasonic homogenizer (Sonifier 250; Branson Ultrasonics Corp., Danbury, CT, USA). Porcine pancreatic lipase (Sigma-Aldrich Co. LLC, St. Louis, MO, USA) was diluted to 1 mg/mL with Tris buffer. Yeast samples were suspended in a buffer solution. Thereafter, 100 µL of yeast suspension was mixed with 100 µL of lipid emulsion and 50 µL of enzyme solution and incubated with shaking at 37 °C for 1 h. The mixture was then heated at 98 °C for 2 min and 0.2 M NaHCO₃ solution was added to inactivate the enzymes. The concentration of released free fatty acids was measured using an enzymatic method (NEFA C-test Wako; FUJIFILM Wako Pure Chemical Corp.). The inhibitory activity was calculated as the percentage (%) = [1 − (C_s/C_b)] × 100, where C_b and C_s represent the released free fatty acid concentrations in the blank and samples, respectively. Orlistat (Tokyo Chemical Industry Co., Ltd., Tokyo, Japan) was used as the positive control, and cellulose (FUJIFILM Wako Pure Chemical Corp.) was used as the negative control (Oagawa et al., 2015).

Determination of bile acid-binding ability    The bile acid-binding ability of yeast samples was evaluated according to the method described by Matsumoto et al. (2011) with some modifications. Yeast samples were added to 5 mL of bile acid solution (2 mM sodium bile acid, 0.1 M phosphatase buffer, pH 7). The suspension was incubated at 37 °C for 2 h with shaking at 100 rpm. After centrifugation at 15 000 rpm for 10 min, the supernatant was passed through a 0.45 µm membrane filter. The concentration of bile acid in the supernatant was measured using an enzymatic method (Total Bile Acid-test Wako; FUJIFILM Wako Pure Chemical Corp.). The binding ability was calculated as the percentage (%) = [1 − (C_s/C_b)] × 100, where C_b and C_s represent bile acid concentrations in the blank and samples, respectively. The binding capacity was calculated as the amount of bile acid-binding per sample weight. Four types of bile acids (sodium salts), cholic acid, glycocholic acid (Tokyo Chemical Industry Co., Ltd.), taurocholic acid (Tokyo Chemical Industry Co., Ltd.), and deoxycholic acid (Sigma-Aldrich Co. LLC) were used. Cholestyramine resin (Sigma-Aldrich Co. LLC) was used as the positive control, and cellulose was used as the negative control.

Statistical analysis    Results are expressed as mean ± standard deviation. Statistical differences were evaluated using one-way analysis of variance with Tukey's post-hoc test. Differences were considered statistically significant at p < 0.05.

Results

Characteristics of fractionated yeast samples    The flow of the preparation and fractionation of the yeast samples used for evaluation is shown in Fig. 1. Dry yeast powder (ORG whole: A) was treated with β-glucanase to obtain whole GT: B, and then fractionated into insoluble (GT ppt: C) and soluble fractions (GT sup: D) using hot water extraction. ORG whole (A), GT whole (B), and GT sup (D) were treated with pepsin and then pancreatin to mimic the state of digestion in the human body. These were fractionated into insoluble (ORG-digest ppt: E, GT-digest ppt: F) and soluble fractions (GT-digest sup: G, GT sup-digest: H) by centrifugation.

Fig. 1.

Flow of preparation and fractionation of the yeast samples. ORG, origin (starting yeast material); GT, β-glucanase-treated yeast; ppt, precipitate; sup, supernatant. Values in parentheses are yields from ORG whole.

The composition and solubility of the yeast samples (ORG whole, GT whole, and GT sup) are shown in Table 1. To obtain water-soluble dietary fiber, including β-glucan from the cell wall of yeast, endo β-1,3-glucanase containing no other enzymes was used. The solubility of GT whole increased from 9.4 % to 32.1 % after β-glucanase treatment. Based on the solubility of GT whole and the amount of dietary fiber in GT sup, the dietary fiber in GT whole was calculated to be 16.3 % insoluble and 83.7 % soluble.

Table 1. Nutritional composition and solubility of the yeast samples

	ORG whole	GT whole†	GT sup†
Components, %
Moisture	3.4	2.2	3.4
Ash	3.1	3.1	7.0
Fat	6.1	6.0	3.7
Protein	60.5	62.2	10.9
Carbohydrate	26.9	26.5	75.0
Dietary fiber*	24.6	25.5	66.5
Solubility, %
undigested	9.4	32.1
with pepsin	43.5	65.3
with pepsin & pancreatin	56.2	75.9

*  Enzymatic-gravimetric method and liquid chromatography

†  Nutritional composition of GT from Sakurai et al. (2020b)

The molecular weight distribution of the soluble fraction of undigested GT and ORG was measured using SEC-RI (Fig. 2). β-Glucanase treatment produced components estimated to have molecular weights of 3 or more than 500 kDa in the soluble fraction of GT.

Fig. 2.

Molecular weight distribution of the soluble fraction of undigested GT and ORG, and the calibration curve of standard pullulan.

Next, the viscosity of the aqueous solutions of GT sup and several dietary fibers was measured (Fig. 3). Even at concentrations as high as 20 %, the viscosity of the GT sup solution was approximately 17 mPa·s, which was considerably lower than that of psyllium or pectin. Additionally, in vitro digestion did not affect the viscosity.

Fig. 3.

Viscosity of aqueous solutions of GT sup and several dietary fibers.

Pancreatic lipase inhibitory activity    Triglycerides ingested from the diet are hydrolyzed by lipase and then absorbed from the intestinal tract. Lipase in pancreatic juice degrades triglycerides in the small intestine. Inhibition of pancreatic lipase results in the suppression or delay of triglyceride digestion and absorption (Ngoh et al., 2017). Therefore, the pancreatic lipase inhibitory activity of the undigested and in vitro digests of the yeast samples was evaluated (Fig. 4). The undigested whole (A, B) and insoluble fractions of ORG (E-1, E-2) and GT (C, F-1, F-2) both had significant inhibitory activity compared to cellulose (NC). The inhibitory activity of the insoluble fraction of GT increased after pepsin digestion (C<F-1) but decreased following digestion with pancreatin (F-1>F-2). When comparing the inhibitory activity of whole and insoluble fractions of ORG and GT at the same stage of digestion, GT exhibited higher inhibition than ORG on undigested and pepsin digests (A<B≈C, E-1<F-1) and lower inhibition than ORG on pepsin-pancreatin digests (E-2≥F-2). In contrast, the soluble fraction of GT (D, G-1, G-2, H-1, and H-2) did not show significant lipase inhibitory activity. However, the inhibition of pancreatic lipase by the soluble fraction of GT was evaluated again with increased sample concentration (Fig. 5). At high concentrations, the undigested and pepsin digests (D, G-1, H-1) showed significant inhibitory activity compared to cellulose (NC), whereas the pepsin-pancreatin digests (G-2, H-2) did not show any inhibitory activity. These results suggest that the insoluble fraction had a higher lipase inhibitory activity than the soluble fraction.

Fig. 4.

Pancreatic lipase inhibition by yeast samples (ORG, white; GT, black; control, gray). The final concentration of each sample was as follows: whole (A and B), 4 %; other yeast samples, equivalent to its content in whole 4 %; PC, positive control (orlistat), 0.004 %; NC, negative control (cellulose), 4 %. Data are expressed as mean ± standard deviation (n = 3), and different letters (a-g) show significant differences (p < 0.05).

Fig. 5.

Pancreatic lipase inhibition by the soluble fraction of GT (GT, black; control, gray). The final concentration for each sample was as follows: undigested GT sup (D), 5 %; other yeast samples, equivalent to its content in undigested GT sup 5 %; PC, positive control (orlistat), 0.004 %; NC, negative control (cellulose), 4 %. Data are expressed as mean ± standard deviation (n = 3), and different letters (a-d) show significant differences (p < 0.05).

Bile acid-binding ability    Bile acids are secreted by the liver and are released into the duodenum. The strong surface activity of bile acids promotes the emulsification of lipids and improves triglyceride hydrolysis by pancreatic lipase. In addition, it increases the solubility of degraded lipids and promotes their absorption. Bile acid-binding agents inhibit these effects by inhibiting the formation of bile acid micelles (Hosomi et al., 2015). Therefore, the bile acid-binding ability of the undigested whole and insoluble fractions of in vitro digests of ORG and GT was first evaluated (Fig. 6). The bile acid-binding rate of all yeast samples was significantly higher than that of cellulose (NC). Fig. 7 shows the amount of bile acid-binding per sample weight. The bile acid-binding capacity of GT increased after pepsin digestion and slightly decreased following pancreatin digestion (B<F-1, F-1≥F-2). In addition, GT had a higher binding capacity than ORG after in vitro digestion (E-1<F-1, E-2<F-2).

Fig. 6.

Binding ability of the yeast samples (ORG, white; GT, black; control, gray) to bile acids (cholic acid, deoxycholic acid, glycocholic acid, and taurocholic acid). The final concentration of each sample was as follows: whole (A and B), 4 %; other yeast samples, equivalent to its content in whole 4 %; PC, positive control (cholestyramine resin), 0.4 %; NC, negative control (cellulose), 4 %. Data are expressed as mean ± standard deviation (n = 3), and different letters (a–g) show significant differences (p < 0.05).

Fig. 7.

Binding capacity of the yeast samples (ORG, white; GT, black) to bile acids (cholic acid, deoxycholic acid, glycocholic acid, and taurocholic acid). Data are expressed as mean ± standard deviation (n = 3), and different letters (a–e) show significant differences (p < 0.05).

Next, the bile acid-binding ability of the soluble fraction of GT was evaluated at the same high sample concentrations as in the pancreatic lipase inhibition assay (Fig. 8). The bile acid-binding rate of in vitro digests of GT whole (G-1, G-2) was significantly higher than that of cellulose (NC). In contrast, the undigested and in vitro digests of GT sup (D, H-1, H-2) did not bind bile acid.

Fig. 8.

Binding ability of the soluble fraction of GT (GT, black; control, gray) to bile acid (cholic acid, deoxycholic acid, glycocholic acid, and taurocholic acid). The final concentration for each sample was as follows: undigested GT sup (D), 5 %; other yeast samples, equivalent to its content in undigested GT sup 5 %; PC, positive control (cholestyramine resin), 0.4 %; NC, negative control (cellulose), 5 %. Data are expressed as mean ± standard deviation (n = 3), and different letters (a–d) show significant differences (p < 0.05).

Discussion

In this investigation, all yeast samples, except for the soluble fraction of pepsin-pancreatin digests of GT (G-2, H-2), displayed a pancreatic lipase inhibition effect (Figs. 4, 5). The mechanisms of pancreatic lipase inhibition involve indirect or direct effects on pancreatic lipase. The former type of mechanism is unique to pancreatic lipase, where the reaction proceeds in a heterogeneous system and is related to the emulsification state of the substrate and the interaction between the enzyme and substrate. Bile acid adsorption is able to lower cholesterol levels and has been considered as one of the former mechanisms (Ogawa et al., 2015; Hosomi et al., 2015; Wang et al., 2015).

Effect of the insoluble components on lipid digestion    Various insoluble dietary fibers, including β-glucan derived from yeast cell wall, have been reported to have bile acid-binding ability (Sima et al., 2018; Zacherl et al., 2011). The dietary fiber in ORG whole was mostly insoluble, and the fiber from GT whole was 16.3 % insoluble. Furthermore, before pepsin digestion and after pancreatin digestion, the bile acidbinding rate appeared to be higher for ORG than for GT (Fig 6; A>B, E-2>F-2), except for glycocholic acid. Thus, the insoluble dietary fiber is presumed to be one of the factors involved in bile acid-binding. Nevertheless, GT displayed a high binding rate, although the insoluble dietary fiber content of GT was only 1/7th that of ORG. In addition, the binding rate of the pepsin digest was not significantly different between ORG and GT (E-1≈F-1), except for deoxycholic acid. This result indicated that components other than insoluble dietary fiber were involved in bile acid-binding.

Since the binding rate of both ORG and GT decreased after in vitro digestion (A>E-1≥E-2, B≥F-1>F-2), proteins and peptides were seemingly involved in bile acid-binding. In general, resistant proteins are rich in hydrophobic amino acids, and thus bind strongly to bile acids through hydrophobic interactions (Iwami et al., 1985; Higaki et al., 2006). The bile acid-binding capacity is presumed to increase after digestion because the rate of indigestible components that can bind bile acids increases. The bile acid-binding capacity of GT increased after pepsin digestion and slightly decreased following pancreatin digestion (Fig. 7; B<F-1, F-1≥F-2), suggesting that pepsin-resistant and indigestible proteins or peptides are involved in bile acid-binding for GT, and the pepsin digest of GT-containing substances has a particularly high bile acid-binding capacity and is partially degraded by pancreatin. The hydrophobic peptide (VAWWMY), derived from soybean protein, has a strong bile acid-binding ability comparable to that of cholestyramine resin (Nagaoka et al., 2010). Conversely, the binding capacity of ORG was almost unchanged after in vitro digestion (A≈E-1≈E-2), except for deoxycholic acid. In addition, GT had a higher binding capacity than ORG after in vitro digestion (E-1<F-1, E-2<F-2). Since GT has a higher content of protein in its insoluble components as compared to ORG, these results suggest a higher binding capacity for proteins and peptides than for dietary fibers. Moreover, owing to the partial removal of dietary fiber from the cell wall by β-glucanase treatment, GT displayed a larger surface area of exposed protein than ORG. This is also presumed to be related to bile acid-binding.

Since all insoluble fractions had bile acid-binding ability, it was presumed that this also affected pancreatic lipase inhibition. However, the pepsin digests of GT had a bile acid-binding rate equal to or lower than that of ORG (Fig. 6; E-1≥F-1), but the pancreatic lipase inhibitory activity was higher than that of ORG (Fig. 4; E-1<F-1). These results suggest that factors other than bile acid-binding are also involved in pancreatic lipase inhibition. Several peptides have been reported to directly inhibit pancreatic lipase, similar to the antiobesity drug orlistat. The peptide VVYP present in globin digests interacts directly with the enzyme and inhibits lipase action (Kagawa et al., 1996). Hydrophobic peptides derived from quail bean display bile acid-binding ability and lipase inhibitory activity, but there is no correlation between the two. It has been proposed that hydrophobic amino acid residues on peptides directly inhibit lipases (Ngoh et al., 2017). Since GT has a larger surface area of exposed protein than ORG, the proteins in GT are presumed to be easily degraded and able to yield hydrophobic peptides that directly inhibit lipases, especially after pepsin digestion.

Effect of the soluble components on lipid digestion    Some soluble dietary fibers have been reported to inhibit or delay the digestion and absorption of lipids by affecting lipid emulsification and diffusion of the gastrointestinal content, which involves changes in viscosity. Psyllium contains both water-soluble and insoluble dietary fibers and inhibits the digestion and absorption of lipids by increasing the viscosity of the contents in the gastrointestinal tract, leading to the adsorption of bile acid (Khossousi et al., 2007; Zacherl et al., 2011). Water-soluble β-glucans derived from oats and barley are highly viscous and are capable of binding to bile acids and inhibiting pancreatic lipase (Kim and White, 2010; Zhai et al., 2020). The viscosity of the β-glucan solution is affected by its concentration and molecular weight, and the higher these are, the higher the viscosity (Kim and White, 2010; Khan et al., 2016; Sun et al., 2020). Pectin disrupts emulsification and inhibits pancreatic lipase reactions owing to its viscosity and charge (Koseki et al., 1989; Espina-Ruiz et al., 2014). Since the viscosity of GT sup was not as high as that of psyllium or pectin (Fig. 3), factors other than viscosity seemed to affect the lipase inhibition of GT sup.

Some soluble β-glucans, proteins, and peptides have been reported to have bile acid-binding ability. Low-viscosity soluble β-glucans, such as beer yeast obtained after gamma irradiation (27.9–175 kDa) and oat-derived acid-degraded products (1.83–27.5 kDa), display higher bile acid-binding capacity at lower molecular weights (Khan et al., 2016; Sun et al., 2020). However, the effect of these β-glucans on the lipase reaction has not been studied. Although the soluble fraction of GT contained low-molecular-weight components, including β-glucan solubilized by β-glucanase treatment, and had low viscosity (Figs. 2, 3), the pepsin-pancreatin digest of GT sup (H-2) did not bind bile acid or inhibit pancreatic lipase (Figs. 4, 5, 8). β-Glucan is not digested by pepsin and pancreatin. Therefore, components other than β-glucan that were not bound to other molecules seemed to affect lipase inhibition and bile acid adsorption of the soluble fraction of GT before pancreatin digestion.

As shown in Fig. 5, the soluble fraction of the pepsin digest of GT whole had higher lipase inhibitory activity than the undigested or pepsin digest of GT sup (G-1>D≈H-1). These results indicated that the components were insoluble before in vitro digestion and were newly solubilized after pepsin digestion and inhibited lipase. As shown in Fig. 8, the soluble fraction of the pepsin digest of GT whole (G-1) displayed a bile acid-binding effect. ε-Polylysine (Tsujita et al., 2003; Kido et al., 2003) and salmon protamine (Hosomi et al., 2015) inhibit pancreatic lipase by bile acid-binding through hydrophobic and electrostatic interactions. Rice bran protein contains an oligo-glutamic acid peptide that exhibits bile acid-binding ability (Wang et al., 2015). C. utilis is protein-rich and contains lysine, threonine, valine, and glutamic acid (Bekatorou et al., 2006). Significantly, peptides with bile acid-binding ability, such as ε-polylysine, salmon protamine, and oligo-glutamic acid, produced when C. utilis was digested in vitro, which may have inhibited the lipase reaction. The bile acid-binding ability of the soluble fraction of the pepsin digest of GT whole was present even after pancreatin digestion, although it was slightly weakened (G-1≥G-2>NC), suggesting that indigestible substances are involved in this adsorption. However, pancreatin digestion may have produced some components that inhibit bile acid adsorption in the emulsion because the pancreatin digests of GT whole (G-2) did not inhibit pancreatic lipase (Figs. 4, 5).

In addition, water-soluble soybean polysaccharide (SSPS) with a hydrophobic protein-bound structure is considered to form a stable layer on the surface of emulsified micelles, which prevents access of bile salt and lipase, and therefore inhibits lipid digestion (Udomrati et al., 2019). Gum arabic also displays lipase inhibitory activity despite its low viscosity; its arabinogalactan protein is presumed to have a similar effect (Pasquier et al., 1996). Yeast cell walls contain proteins bound to mannan and β-glucan, since the viscosity of GT sup was similar to that of gum arabic (Fig. 3), the undigested and pepsin digests of the soluble fraction of GT may contain glycoproteins that inhibit lipase, similar to the activity displayed by SSPS and gum arabic. In addition, after pancreatin digestion, the loss of lipase inhibition by GT sup (Fig. 5; H-1>H-2≈NC) was presumably the result of the digestion of this glycoprotein.

Although our previous study indicated that GT sup exerts a suppressive effect on postprandial serum triglyceride elevation (Sakurai et al., 2020b), the pepsin-pancreatin digest of GT sup (H-2) did not inhibit pancreatic lipase (Figs. 4, 5), suggesting that GT sup may act not only at the lipid digestion stage, but also at later stages such as the lipid absorption stage. It has been suggested that resistant maltodextrin and isomaltodextrin suppress lipid absorption and promote lipid excretion by inhibiting micelle decomposition and stabilizing micellar structures (Kishimoto et al., 2009; Takagaki et al., 2018). Lipoprotein lipase and hepatic triglyceride lipase, which are lipid-metabolizing enzymes, are responsible for hydrolyzing triglycerides that exist in the blood in a chylomicron state. Globin peptides activate these lipases and lower postprandial blood triglyceride levels (Kagawa et al., 1996). GT sup may also suppress increase in postprandial serum triglyceride levels by mechanisms other than inhibition of lipid digestion, as mentioned earlier. Thus, further research is required to clarify the mechanisms underlying the effects of GT sup.

Conclusions

In conclusion, the present results suggest that GT has an inhibitory effect on the digestion of dietary triglycerides through several mechanisms. Insoluble components, such as dietary fiber and pepsin-resistant and indigestible proteins or peptides, are presumed to bind bile acids and inhibit pancreatic lipase. The soluble dietary fiber derived from the yeast cell wall by β-glucanase treatment had poor viscosity and may display lipase inhibitory activity in complexes with proteins or peptides. However, the detailed mechanism of action, especially for the water-soluble fraction, has not yet been properly elucidated, and further studies are required. In our previous study, we revealed that GT sup has a suppressive effect on postprandial serum triglyceride elevation (Sakurai et al., 2020b). However, in this study, the ingestion of GT whole seems to have a more potent effect than GT sup in controlling serum triglyceride levels.

Conflict of interest    The authors declare no conflict of interest.

References

•  Bekatorou,  A.,  Psarianos,  C., and  Koutinas,  A.A. (2006). Production of food grade yeasts. Food Technol. Biotechnol., 44, 407-415.
•  Bell,  S.,  Goldman,  V.M.,  Bistrian,  B.R.,  Arnold,  A.H.,  Ostroff,  G., and  Forse,  R.A. (1999). Effect of β-glucan from oats and yeast on serum lipids. Crit. Rev. Food Sci. Nutr., 39, 189-202.
•  Cassidy,  Y.M.,  McSorley,  E.M., and  Allsopp,  P.J. (2018). Effect of soluble dietary fibre on postprandial blood glucose response and its potential as a functional food ingredient. J. Funct. Foods, 46, 423-439.
•  Espinal-Ruiz,  M.,  Parada-Alfonso,  F.,  Restrepo-Sánchez,  L.P.,  Narváez-Cuenca,  C.E., and  McClements,  D.J. (2014). Impact of dietary fibers [methyl cellulose, chitosan, and pectin] on digestion of lipids under simulated gastrointestinal conditions. Food Funct., 5, 3083-3095.
•  Higaki,  N.,  Sato,  K.,  Suda,  H.,  Suzuka,  T.,  Komori,  T.,  Saeki,  T.,  Nakamura,  Y.,  Ohtsuki,  K.,  Iwami,  K., and  Kanamoto,  R. (2006). Evidence for the existence of a soybean resistant protein that captures bile acid and stimulates its fecal excretion. Biosci. Biotechnol. Biochem., 70, 2844-2852.
•  Hosomi,  R.,  Miyauchi,  K.,  Yamamoto,  D.,  Arai,  H.,  Nishiyama,  T.,  Yoshida,  M., and  Fukunaga,  K. (2015). Salmon protamine decreases serum and liver lipid contents by inhibiting lipid absorption in an in vitro gastrointestinal digestion model and in rats. J. Food Sci., 80, H2346-H2353.
•  Iso,  H.,  Imano,  H.,  Yamagishi,  K.,  Ohira,  T.,  Cui,  R.,  Noda,  H.,  Sato,  S.,  Kiyama,  M.,  Okada,  T.,  Hitsumoto,  S.,  Tanigawa,  T., and  Kitamura,  A. (2014). Fasting and non-fasting triglycerides and risk of ischemic cardiovascular disease in Japanese men and women: The Circulatory Risk in Communities Study (CIRCS). Atherosclerosis, 237, 361-368.
•  Iso,  H.,  Naito,  Y.,  Sato,  S.,  Kitamura,  A.,  Okamura,  T.,  Sankai,  T.,  Shimamoto,  T.,  Iida,  M., and  Komachi,  Y. (2001). Serum triglycerides and risk of coronary heart disease among Japanese men and women. Am. J. Epidemiol., 153, 490-499.
•  Iwami,  K.,  Sakakibara,  K., and  Ibuki,  F. (1986). Involvement of post-digestion ‘hydrophobic’ peptides in plasma cholesterol-lowering effect of dietary plant proteins. Agric. Biol. Chem., 50, 1217-1222.
•  Kagawa,  K.,  Matsutaka,  H.,  Fukuhama,  C.,  Watanabe,  Y., and  Fujino,  H. (1996). Globin digest, acidic protease hydrolysate, inhibits dietary hypertriglyceridemia and Val-Val-Tyr-Pro, one of its constituents, possesses most superior effect. Life Sci., 58, 1745-1755.
•  Kagawa,  K.,  Matsutaka,  H.,  Fukuhama,  C.,  Fujino,  H., and  Okuda,  H. (1998). Suppressive effect of globin digest on postprandial hyperlipidemia in male volunteers. J. Nutr., 128, 56-60.
•  Khan,  A.A.,  Gani,  A.,  Masoodi,  F.A.,  Amin,  F.,  Wani,  I.A.,  Khanday,  F.A., and  Gani,  A. (2016). Structural, thermal, functional, antioxidant & antimicrobial properties of β-D-glucan extracted from baker's yeast (Saccharomyces cereviseae)—Effect of γ-irradiation. Carbohydr. Polym., 140, 442-450.
•  Khossousi,  A.,  Binns,  C.W.,  Dhaliwal,  S.S., and  Pal,  S. (2008). The acute effects of psyllium on postprandial lipaemia and thermogenesis in overweight and obese men. Br. J. Nutr., 99, 1068-1075.
•  Kido,  Y.,  Hiramoto,  S.,  Murao,  M.,  Horio,  Y.,  Miyazaki,  T.,  Kodama,  T., and  Nakabou,  Y. (2003). ε-Polylysine inhibits pancreatic lipase activity and suppresses postprandial hypertriacylglyceridemia in rats. J. Nutr., 133, 1887-1891.
•  Kim,  H.J. and  White,  P.J. (2010). In vitro bile-acid binding and fermentation of high, medium, and low molecular weight β-glucan. J. Agric. Food Chem., 58, 628-634.
•  Kishimoto,  Y.,  Oga,  H.,  Tagami,  H.,  Okuma,  K., and  Gordon,  D.T. (2007). Suppressive effect of resistant maltodextrin on postprandial blood triacylglycerol elevation. Eur. J. Nutr., 46, 133-138.
•  Kishimoto,  Y.,  Yoshikawa,  Y.,  Miyazato,  S.,  Oga,  H.,  Yamada,  T.,  Tagami,  H.,  Hashizume,  C., and  Yamamoto,  K. (2009). Effect of resistant maltodextrin on digestion and absorption of lipids. J. Heal. Sci., 55, 838-844.
•  Klis,  F.M.,  Boorsma,  A., and  De Groot,  P.W.J. (2006). Cell wall construction in Saccharomyces cerevisiae. Yeast, 23, 185-202.
•  Koseki,  M.,  Tsuji,  K.,  Nakagawa,  Y.,  Kawamura,  M.,  Ichikawa,  T.,  Kazama,  M.,  Kitabatake,  N., and  Doi,  E. (1989). Effects of gum arabic and pectin on the emulsification, the lipase reaction, and the plasma cholesterol level in rats. Agric. Biol. Chem., 53, 3127-3132.
•  Ma,  Y. and  Xiong,  Y.L. (2009). Antioxidant and bile acid binding activity of buckwheat protein in vitro digests. J. Agric. Food Chem., 57, 4372-4380.
•  Matsumoto,  K.,  Kadowaki,  A.,  Ozaki,  N.,  Takenaka,  M.,  Ono,  H.,  Yokoyama,  S.I., and  Gato,  N. (2011). Bile acid-binding ability of kaki-tannin from young fruits of persimmon (Diospyros kaki) in vitro and in vivo. Phytother. Res., 25, 624-628.
•  Nagaoka,  S.,  Nakamura,  A.,  Shibata,  H., and  Kanamaru,  Y. (2010). Soystati (VAWWMY), a novel bile acid-binding peptide, decreased micellar solubility and inhibited cholesterol absorption in rats. Biosci. Biotechnol. Biochem., 74, 1738-1741.
•  Ngoh,  Y.Y.,  Choi,  S.B., and  Gan,  C.Y. (2017). The potential roles of Pinto bean (Phaseolus vulgaris cv. Pinto) bioactive peptides in regulating physiological functions: Protease activating, lipase inhibiting and bile acid binding activities. J. Funct. Foods, 33, 67-75.
•  Ogawa,  A.,  Kobayashi,  T.,  Sakai,  F.,  Kadooka,  Y., and  Kawasaki,  Y. (2015). Lactobacillus gasseri SBT2055 suppresses fatty acid release through enlargement of fat emulsion size in vitro and promotes fecal fat excretion in healthy Japanese subjects. Lipids Health Dis., 14, 1-10.
•  Pasquier,  B.,  Armand,  M.,  Castelain,  C.,  Guillon,  F.,  Borel,  P.,  Lafont,  H., and  Lairon,  D. (1996). Emulsification and lipolysis of triacylglycerols are altered by viscous soluble dietary fibres in acidic gastric medium in vitro. Biochem. J., 314, 269-275.
•  Sakae,  A. and  Sekizaki,  E. (2014). Brief method for lipase activity measurement, and its application for inhibition effect by Pu'er tea. Bulletin of Kiryu University (Kiryu Daigaku Kiyou), 25, 69-74. (in Japanese)
•  Sakurai,  T.,  Kitahara,  C.,  Sato,  T., and  Katsumata,  T. (2020a). Compositions containing yeast-derived protein. Japan Kokai Tokkyo Koho, P2020-183364A, Nov. 12. (in Japanese)
•  Sakurai,  T.,  Sato,  T.,  Kitahara,  C.,  Furuta,  K.,  Fukuda,  Y.,  Kaneda,  T.,  Kosaka,  H.,  Asano,  Y., and  Katsumata,  T. (2020b). Effect of water-soluble dietary fiber from Candida utilis on postprandial serum triglyceride and safety evaluation of long-term and excessive intake of C. utilis extract in healthy subjects —randomized, double-blind, placebo-controlled, crossover and parallel group study—. Jpn. Pharmacol. Ther., 48, 811-826
•  Sima,  P.,  Vannucci,  L., and  Vetvicka,  V. (2018). β-glucans and cholesterol (Review). Int. J. Mol. Med., 41, 1799-1808.
•  Sun,  T.,  Li,  J.,  Qin,  Y.,  Xie,  J.,  Xue,  B.,  Li,  X.,  Gan,  J.,  Bian,  X., and  Shao,  Z. (2020). Rheological and functional properties of oat β-glucan with different molecular weight. J. Mol. Struct., 1209, 127944.
•  Takagaki,  R.,  Ishida,  Y.,  Sadakiyo,  T.,  Taniguchi,  Y.,  Sakurai,  T.,  Mitsuzumi,  H.,  Watanabe,  H.,  Fukuda,  S., and  Ushio,  S. (2018). Effects of isomaltodextrin in postprandial lipid kinetics: Rat study and human randomized crossover study. PLoS One, 13, e0196802.
•  Talati,  R.,  Baker,  W.L.,  Pabilonia,  M.S.,  White,  C.M., and  Coleman,  C.I. (2009). The effects of barley-derived soluble fiber on serum lipids. Ann. Fam. Med., 7, 157-163.
•  Tsujita,  T.,  Sumiyoshi,  M.,  Takaku,  T.,  Momsen,  W.E.,  Lowe,  M.E., and  Brockman,  H.L. (2003). Inhibition of lipases by ε-polylysine. J. Lipid Res., 44, 2278-2286.
•  Udomrati,  S.,  Pantoa,  T.,  Gohtani,  S.,  Nakajima,  M.,  Uemura,  K., and  Kobayashi,  I. (2019). Effects of water-soluble soybean polysaccharide on rheological properties, stability and lipid digestibility of oil-in-water emulsion during in vitro gastrointestinal digestion. Int. J. Food Sci. Technol., 55, 1437-1447.
•  Volman,  J.J.,  Ramakers,  J.D., and  Plat,  J. (2008). Dietary modulation of immune function by β-glucans. Physiol. Behav., 94, 276-284.
•  Wang,  J.,  Shimada,  M.,  Kato,  Y.,  Kusada,  M., and  Nagaoka,  S. (2015). Cholesterol-lowering effect of rice bran protein containing bile acid-binding proteins. Biosci. Biotechnol. Biochem., 79, 456-461.
•  Yoshinaga,  K. and  Mitamura,  R. (2019). Effect of Undaria pinnatifida (Wakame) on postprandial serum lipid response in humans. J. Jpn. Soc. Nutr. Food Sci. (Nippon Eiyo Shokuryo Gakkaishi), 72, 267-273. (in Japanese)
•  Zacherl,  C.,  Eisner,  P., and  Engel,  K.H. (2011). In vitro model to correlate viscosity and bile acid-binding capacity of digested water-soluble and insoluble dietary fibres. Food Chem., 126, 423-428.
•  Zhai,  H.,  Gunness,  P., and  Gidley,  M.J. (2020). Barley β-glucan effects on emulsification and in vitro lipolysis of canola oil are modulated by molecular size, mixing method, and emulsifier type. Food Hydrocoll., 103, 105643.