Applied Microbiology and Biotechnology

, Volume 97, Issue 18, pp 8273–8281

Cholesterol-lowering probiotics: in vitro selection and in vivo testing of bifidobacteria

Authors

  • Alessandra Bordoni
    • Department of Agro-Food Sciences and TechnologiesUniversity of Bologna
  • Alberto Amaretti
    • Department of Life SciencesUniversity of Modena and Reggio Emilia
  • Alan Leonardi
    • Department of Life SciencesUniversity of Modena and Reggio Emilia
  • Elisa Boschetti
    • Department of Biomedical and Neuromotor ScienceUniversity of Bologna
  • Francesca Danesi
    • Department of Agro-Food Sciences and TechnologiesUniversity of Bologna
  • Diego Matteuzzi
    • Department of Pharmaceutical SciencesUniversity of Bologna
  • Lucia Roncaglia
    • Department of Life SciencesUniversity of Modena and Reggio Emilia
  • Stefano Raimondi
    • Department of Life SciencesUniversity of Modena and Reggio Emilia
    • Department of Life SciencesUniversity of Modena and Reggio Emilia
Applied microbial and cell physiology

DOI: 10.1007/s00253-013-5088-2

Cite this article as:
Bordoni, A., Amaretti, A., Leonardi, A. et al. Appl Microbiol Biotechnol (2013) 97: 8273. doi:10.1007/s00253-013-5088-2

Abstract

Thirty-four strains of bifidobacteria belonging to Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium bifidum, Bifidobacterium breve, Bifidobacterium longum, and Bifidobacterium pseu-docatenulatum were assayed in vitro for the ability to assimilate cholesterol and for bile salt hydrolase (BSH) against glycocholic and taurodeoxycholic acids (GCA and TDCA). Cholesterol assimilation was peculiar characteristic of two strains belonging to the species B. bifidum (B. bifidum MB 107 and B. bifidum MB 109), which removed 81 and 50 mg of cholesterol per gram of biomass, being the median of specific cholesterol absorption by bifidobacteria 19 mg/g. Significant differences in BSH activities were not established among bifidobacterial species. However, the screening resulted in the selection of promising strains able to efficiently deconjugate GCA and TDCA. No relationship was recognized between BSH phenotype and the extent of cholesterol assimilation. On the basis of cholesterol assimilation or BSHGCA and BSHTDCA activities, B. bifidum MB 109 (DSMZ 23731), B. breve MB 113 (DSMZ 23732), and B. animalis subsp. lactis MB 2409 (DSMZ 23733) were combined in a probiotic mixture to be fed to hypercholesterolemic rats. The administration of this probiotic formulation resulted in a significant reduction of total cholesterol and low-density cholesterol (LDL-C), whereas it did not affect high-density cholesterol (HDL-C) and HDL-C/LDL-C ratio.

Keywords

BifidobacteriumProbioticCholesterolBile salt hydrolaseIn vivoIn vitro

Introduction

A pioneering approach in the dietary treatment of hypercholesterolemia is represented by the utilization of probiotics, i.e. live microbes which, when administered in adequate amounts, confer a health benefit to the host (FAO/WHO working group 2001). Probiotic bacteria, mostly belonging to lactic acid bacteria (LAB) and bifidobacteria which are capable to colonize the intestinal microbiota, have been demonstrated to exert a number of beneficial health effects through a variety of mechanisms (Rossi and Amaretti 2010; Williams 2010; Nagpal et al. 2012) and have recently attracted considerable attention due to potential anti-cholesterol properties. In fact, several in vivo trials report that consumption of probiotics reduced the systemic cholesterol levels and caused a decrease in blood lipids as well (Usman and Hosono 2000; Pereira and Gibson 2002b; Liong and Shah 2006; Jones et al. 2012a, b, 2013).

Different mechanisms have been proposed for the cholesterol-lowering ability of probiotic bacteria; among them, bile salt hydrolase (BSH) activity (Begley et al. 2006; Swann et al. 2011; Kumar et al. 2012; Jones et al. 2013). Bile acids are efficiently conserved under normal conditions by a process termed enterohepatic recirculation. Conjugated and unconjugated bile acids are absorbed by passive diffusion along the entire gut and by active transport in the terminal ileum (Carey and Duane 1994). Deconjugated bile salts are less efficiently reabsorbed than their conjugated counterparts (De Smet et al. 1998; Jones et al. 2013), which results in the excretion of larger amounts of free bile acids in feces. Thus, microbial BSH activity has the effect to increase de novo synthesis of bile salts in the liver in a homeostatic response to replace those lost with excretion, leading to a reduction in serum cholesterol (Kumar et al. 2012). BSH activity can also indirectly limit cholesterol absorption by reducing its solubility, since bile acids are less efficient in lipids solubilization than bile salts (Reynier et al. 1981).

Some potentially probiotic bacteria, belonging to the species of Lactobacillus and Bifidobacterium indigenous to the gastrointestinal tract (e.g., Lactobacillus acidophilus, Lactobacillus casei, Lactobacillus fermentum, Lactobacillus gasseri, Lactobacillus helveticus, Lactobacillus paracasei, Lactobacillus plantarum, Lactobacillus rhamnosus, Bifidobacterium adolescentis, Bifidobacterium animalis, Bifidobacterium breve, Bifidobacterium infantis, and Bifidobacterium longum), were demonstrated to produce BSH (De Smet et al. 1995; Liong and Shah 2005; Kumar et al. 2012).

Furthermore, probiotic bacteria have a saccharolytic metabolism, yielding acetic acid and lactic acids through carbohydrates fermentation, and acidify the luminal content. Acidifying conditions decrease the solubility of bile salts and particularly of their deconjugated forms (Klaver and Vandermeer 1993), further promoting the excretion of bile acids.

Assimilation is another purported mechanism involved in lowering cholesterol levels by some bifidobacteria and LAB (Dambekodi and Gilliland 1998; Grill et al. 2000; Pereira and Gibson 2002a). Cholesterol can be assimilated mostly during bacterial growth, being bound without transformations onto the cellular surface and incorporated within the membrane phospholipid bilayer (Noh et al. 1997; Kimoto et al. 2002; Lye et al. 2010a; Ooi and Liong 2010). In addition, it was demonstrated that uptake may be accompanied by co-precipitation with bile salts, and/or transformation to coprostanol, similarly resulting in cholesterol removal (Tahri et al. 1997; Lye et al. 2010b).

The identification of probiotics that combine the beneficial human health effects intrinsic to their microbial species to the capability to affect blood cholesterol levels is a major goal in the cholesterol-lowering nutritional strategy. This study aimed to identify probiotic bifidobacteria potentially effective in the lowering blood serum cholesterol levels, through a screening of cholesterol assimilation and BSH activity in 34 strains of human origin. The best performing strains were given to rats receiving a hypercholesterolemic diet, in order to evaluate the effects of the probiotic mixture on lipemic values.

Material and methods

Materials

All chemicals were purchased from Sigma-Aldrich unless otherwise stated. Bifidobacterium strains, belonging to B. adolescentis, B. animalis, B. animalis subsp. lactis, Bifidobacterium bifidum, B. breve, B. longum subsp. infantis, B. longum subsp. longum, and B. pseudocatenulatum (Table 1), were obtained from ATCC (Manassas, VA, USA), DSMZ (Braunschweig, Germany), BUSCoB strain collection (Scardovi Collection of Bifidobacteria, Dept. of Agro-Environmental Science and Technology, University of Bologna, Italy), and our own collection (Dept. of Life Sciences, University of Modena and Reggio Emilia, Modena, Italy). The bacteria were cultured in Lactobacilli MRS broth (BD Difco, Sparks, MD, USA) containing 0.5 g/L l-cysteine·HCl, and were anaerobically incubated at 37 °C for 24 h. Growth was determined by measuring the biomass dry weight (DW).
Table 1

Screening of bifidobacteria for cholesterol absorption in Cholesterol-MRS and for BSH specific activity against GCA and TDCA

Strain

Adsorption

BSHGCA

BSHTDCA

mg/g

U/mg

U/mg

B. animalis MB 103

16

0.28

0.01

B. animalis MB 104

15

0.36

0.06

B. animalis MB 105

11

0.71

0.13

B. animalis subsp. lactis ATCC 27536

17

0.14

0.06

B. animalis subsp. lactis MB 238

20

0.24

0.03

B. animalis subsp. lactis MB 2409 (DSMZ 23733)

19

0.77

0.10

B. animalis subsp. lactis WC 0459

11

0.69

0.65

B. bifidum MB 107

81

0.15

0.07

B. bifidum MB 109 (DSMZ 23731)

50

0.10

0.07

B. breve MB 113 (DSMZ 23732)

18

1.10

0.18

B. breve MB 233

31

0.09

0.06

B. breve MB 234

24

0.23

0.03

B. breve MB 235

29

0.49

0.04

B. breve MB 252

19

0.09

0.03

B. breve MB 283

12

0.37

0.47

B. longum subsp. infantis ATCC 15697

8

0.67

0.48

B. longum subsp. infantis MB 257

22

0.02

0.02

B. longum subsp. infantis MB 263

27

0.31

0.18

B. longum subsp. longum MB 203

23

0.09

0.10

B. longum subsp. longum MB 214

25

1.33

0.12

B. longum subsp. longum MB 216

16

0.18

0.14

B. longum subsp. longum MB 217

19

0.12

0.05

B. longum subsp. longum MB 218

17

0.23

0.06

B. longum subsp. longum MB 220

17

0.07

0.16

B. longum subsp. longum MB 221

12

0.61

0.16

B. longum subsp. longum MB 222

29

0.55

0.14

B. longum subsp. longum MB 224

20

0.24

0.08

B. longum subsp. longum MB 244

35

0.00

0.13

B. adolescentis MB 239 (DSMZ 18350)

19

0.32

0.06

B. pseudocatenulatum MB 114

8

0.74

0.13

B. pseudocatenulatum MB 115

4

0.10

0.01

B. pseudocatenulatum MB 116 (DSMZ 18353)

11

0.40

0.20

B. pseudocatenulatum MB 237

15

0.63

0.53

B. pseudocatenulatum MB 243

19

0.59

0.10

Values are means, n = 3. Cholesterol absorption, SD always < 7 mg/g; BSHGCA and BSHTDCA, SD always < 0.12 U/mg

The lyophilized probiotic formulation used in the animal trial was supplied by Probiotical (Novara, Italy). It contained equal amounts of B. bifidum MB 109, B. breve MB 113, and B. animalis subsp. lactis MB 2409, and potato maltodextrin as excipient. The strains have been accepted for deposit by DSMZ for patent purposes and named B. bifidum DSMZ 23731, B. breve DSMZ 23732, B. animalis subsp. lactis DSMZ 23733.

Cholesterol assimilation

Cholesterol was dissolved at 50 g/L in Tween 80, then added to MRS at the final concentration of 0.1 g/L. Cholesterol-supplemented MRS was inoculated (10 % v/v) with 16 h MRS cultures and incubated at 37 °C in anaerobiosis. After 24 h, bacterial growth and residual cholesterol were analyzed. MRS cultures and non-inoculated cholesterol-MRS were used as controls.

Cholesterol was analyzed by HPLC, according to the method of Hansbury and Scallen (1978), though slightly modified. The culture was centrifuged (7,000×g for 5 min at 4 °C), the supernatant was filtered (0.22 μm), and 1 μl was injected into a HPLC device (Agilent 1100, Agilent Technologies, USA) equipped with a variable wavelength detector and ZORBAX Eclipse XDB-C18 column (rapid resolution, 1.8 μm particle size, 4.6 × 50 mm, Agilent). Isocratic elution was performed with 2.5 ml/min acetonitrile. Cholesterol was identified at 210 nm by retention time (4 min) and quantified by interpolation of calibration curves. Standard solutions were prepared diluting proper amounts of cholesterol stock solutions in MRS broth. Linearity was demonstrated from 0.01 to 0.1 g/L (r2 = 0.994). The limit of detection (calculated as LOD = 3·(Sy/x)/b, where Sy/x is the residual standard deviation and b is the slope of the linear calibration) was 0.01 g/L.

BSH assays

BSH activity was determined in the cell-free extract by measuring the amount of amino acids released from conjugated bile salts. Cells of 24 h MRS cultures were harvested by centrifugation (7,000×g for 5 min at 4 °C), washed twice with 0.1 M phosphate buffer (pH 6.0), and suspended in the same buffer to an OD600 of 3.0 AU. A 4-ml cell suspension was subjected to mechanical disruption through one passage at 20 kPsi in the One Shot Cell Disrupter (Constant Systems, UK) and centrifuged (13,000×g for 15 min at 4 °C) to remove whole cells and debris. To measure BSH activity, 20 μl of cell-free extract were mixed with 360 μl of 0.1 M phosphate buffer (pH 6.0) and 20 μl of 200 mM sodium salt of glycocholic acid (GCA) and taurodeoxycholic acid (TDCA), and were incubated at 37 °C. A volume of 100 μl of reaction mixture was sampled after 10 and 30 min and was added to 100 μl of 15 % (w/v) trichloroacetic acid to stop the reaction. The sample was centrifuged (13,000×g for 5 min at 4 °C) and the free amino groups in the supernatant were quantified by ninhydrin reaction (Meyer 1957). A unit of BSH activity was defined as 1 μmol of amino acids released in 1 min from the substrate. The specific activity of cell-free extracts was defined as units per milligram of proteins. Protein concentration was determined with copper/Folin phenol reagent assay kit (Sigma-Aldrich). Four independent biological duplicates with duplicated measurements were performed for each sample.

Animals and diet

The study protocol was approved by the Animal Care Committee of the University of Bologna (Prot. 20600-X/10). Principles of laboratory animal care were followed and the study complied with European Union regulations for the care and use of laboratory animals. Twenty-four male Wistar rats, aged 30 day, were obtained from Harlan (Milano, Italy) and fed diets supplied by Mucedola (Settimo Milanese, Italy). The standard chow diet (ST) was AIN-93M. Hypercholesterolemic diets consisted of AIN-93M supplemented with: 3 % cholesterol + 0.1 % cholic acid (HC1); 1 % cholesterol + 0.1 % cholic acid (HC2). The rats were housed in a controlled environment (temperature 21 °C, humidity 55 ± 5 %) with a 12-h light/dark cycle. Throughout the study, food and water were provided ad libitum, food consumption being recorded daily and rat body weight (bw) weekly.

After a 7-day adaptation to ST diet, rats were turned to hypercholesterolemic diet HC1 for 14 days, then were randomly divided into three dietary groups (each group n = 8) receiving ST, HC2, and probiotic-supplemented HC2 (HC2P) for 30 days. Every morning starting from day 15, the dietary groups received a 6 ml gastric gavage: ST and HC2 groups received water, while HC2P group received a water suspension of probiotic mixture, where the lyophilized probiotic formulation was reconstituted with water (10 % w/v) and diluted as appropriate to achieve the dosage of 0.33 × 109 cfu/day of each strain.

Blood samples were collected after a 12-h fasting at the end of the adaptation period (T0), at the end of the hypercholesterolemic period (T15), and at the end of the dietary treatment (T45). Blood was taken from the tail vein, and centrifuged immediately at 1,500×g for 15 min at 4 °C to obtain serum. Serum samples were stored at −80 °C until analyses.

Blood analyses

The serum concentration of total cholesterol (TC), LDL cholesterol (LDL-C), HDL cholesterol (HDL-C), and triglyceride (TG) was measured at T0, T15, and T45 using cholesterol/cholesteryl ester, HDL and LDL/VLDL cholesterol, and triglyceride quantification kits (BioVision Inc., CA, USA), according to the manufacturer’s instructions.

Results

Cholesterol adsorption by bifidobacteria

In order to develop a screening protocol for cholesterol adsorption by bifidobacteria, the kinetics of growth and cholesterol absorption was determined for two randomly chosen strains, B. longum subsp. infantis ATCC 15697 and B. bifidum MB 109, which were cultured for 40 h in MRS medium supplemented with 0.1 g/L cholesterol (Fig. 1). Both strains entered into the stationary phase within 24 h and adsorbed cholesterol, mostly during the growth phase. After 24 h, cholesterol adsorbed by B. longum subsp. infantis ATCC 15697 and B. bifidum MB 109 was 34 and 52 %, respectively. Due to the wide difference in biomass production, that was 4.6-fold higher in B. longum subsp. infantis ATCC 15697 than in B. bifidum MB 109 (P < 0.05), adsorption results were normalized by the DW, being 7.2 and 48.9 mg/g, respectively.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-013-5088-2/MediaObjects/253_2013_5088_Fig1_HTML.gif
Fig. 1

Time-course cholesterol adsorption by B. longum subsp. infantis ATCC 15697 (a) and B. bifidum MB 109 (b), cultured batch-wise in Cholesterol-MRS. Symbols: DW, filled circle; cholesterol, open circle; cholesterol absorption per g of biomass, open upright triangle

Based on these results, 34 Bifidobacterium strains (B. animalis, n = 6; B. bifidum, n = 2; B. breve, n = 7; B. longum, n = 13; B. adolescentis and B. pseudocatenulatum, n = 6) were screened for cholesterol adsorption, residual cholesterol being evaluated after 24 h of growth in MRS medium supplemented with 0.1 g/L cholesterol. To take into account the differences in growth yields, the extent of cholesterol removal was normalized by the DW.

Bifidobacteria exhibited adsorption levels in the range between 4 and 81 mg per g of dry biomass (median, 19 mg/g; mean ± SD, 21 ± 13 mg/g) (Table 1, Fig. 2a). The strains B. bifidum MB 109 and B. bifidum MB 107 adsorbed the highest amount of cholesterol (50 and 81 mg/g, respectively). The other strains presenting values higher than the 75th percentile (23 mg/g) belonged to B. breve, B. longum subsp. infantis, B. longum subsp. longum (three, one, and three strains, respectively) (Table 1). On average, the strains belonging to the species B. bifidum exhibited the highest adsorption levels (P < 0.05), but the numerosity of the strains belonging to this species was too low to state conclusively that cholesterol adsorption is a specific trait of B. bifidum. With regard to the other species, due to the high strain-specific variability within B. longum, statistically significant differences were not found between B. longum and each B. breve, B. animalis, and B. pseudocatenulatum. Nonetheless, significant differences were found between B. breve and both B. animalis and B. pseudocatenulatum (P < 0.05). In particular, B. animalis and B. pseudocatenulatum exhibited the lowest ability to absorb cholesterol, most of the strains belonging to these species presenting values lower than the general median.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-013-5088-2/MediaObjects/253_2013_5088_Fig2_HTML.gif
Fig. 2

Screening of cholesterol absorption and BSH activity in 34 strains of Bifidobacterium (B. animalis, n = 6; B. bifidum, n = 2; B. breve, n = 7; B. longum, n = 13; B. adolescentis and B. pseudocatenulatum, n = 6). Cholesterol absorption after 24 h of growth in Cholesterol-MRS (a); BSH specific activity against GCA (b); BSH specific activity against TDCA (c). Boxes indicate the median and 25th and 75th percentiles; whiskers indicate 10th and 90th percentiles. Dotted lines indicate means; means with a common symbol are not significantly different (P ≥ 0.05) according to one-way ANOVA with Tukey as post test

According to the results of the screening, B. bifidum MB 109 was selected as a component of the probiotic formulation to be administered to hypercholesterolemic rats.

BSH activity in Bifidobacteria

Bifidobacteria were screened for intracellular BSH activity, which was determined as the capability of cell extracts to hydrolyze the bile salts GCA and TDCA (Table 1, Fig. 2b, c). BSHGCA specific activity was found in the range between 0 and 1.33 U/mg (median, 0.29 U/mg; mean ± SD, 0.38 ± 0.32 U/mg). Significant differences could not be established among the species, due to the wide distribution of BSHGCA and to the low numerosity of some species (P > 0.05) (Table 1, Fig. 2b). Strains belonging to the species B. animalis, B. breve, B. longum, and B. pseudocatenulatum (three, one, three, and two strains, respectively) presented values higher than the 75th percentile (0.60 U/mg). In particular, B. longum subsp. longum MB 214 exhibited the highest activity (P < 0.05), followed by B. breve MB 113 (1.32 and 1.10 U/mg, respectively). BSHTDCA specific activity was found in the range between 0.01 and 0.65 U/mg (median, 0.10 U/mg; mean ± SD, 0.15 ± 0.14 U/mg). Likewise BSHGCA, significant differences could not be established among the species (P > 0.05) (Table 1, Fig. 2c). Strains belonging to the species B. animalis, B. breve, B. longum, and B. pseudocatenulatum (one, two, four, and two strains, respectively) presented values higher than the 75th percentile (0.15 U/mg). In particular, B. animalis subsp. lactis WC 0459 exhibited the highest activity (P < 0.05), followed by B. pseudocatenulatum MB 237 (0.65 and 0.53 U/mg, respectively). No relationship was established between BSHTDCA and BSHGCA specific activities (r2 = 0.14). However, five out of the nine strains with BSHTDCA higher than the 75th percentile (i.e. B. animalis subsp. lactis WC 0459, B. breve MB 113, B. longum subsp. infantis ATCC 15697, B. longum subsp. longum MB 221, and B. pseudocatenulatum MB 237) presented BSHGCA higher than the 75th percentile as well. Besides, four out of the nine strains with BSHTDCA lower than the 25th percentile (i.e., B. breve MB 252, B. longum subsp. infantis MB 257, B. longum subsp. longum MB 217, B. pseudocatenulatum MB 115) exhibited BSHGCA lower than the 25th percentile as well.

According to the results of the screening of BSH activities and to technological properties, B. breve MB 113, and B. animalis subsp. lactis MB 2409 were selected as components of the probiotic formulation to be administered to hypercholesterolemic rats.

In vivo effects of probiotic supplementation

No differences in food consumption (data not shown) and bw (Fig. 3) were detected among the experimental groups during the trial (P ≥ 0.05). In order to induce hypercholesterolemia, during the first period of dietary treatment (T0–T15), all rats received the hypercholesterolemic HC1 diet. Significant differences were detected in the lipemic values between the beginning and the end of the first experimental period (Table 2). In particular, TC and LDL-C levels increased and HDL-C drastically decreased, causing a significant decrease of the HDL-C/LDL-C ratio at T15.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-013-5088-2/MediaObjects/253_2013_5088_Fig3_HTML.gif
Fig. 3

Changes in rat body weight under the different experimental conditions. Rat body weight was measured weekly. Data are means ± SD of 24 rats at T0–T15, and of 8 rats in each group in the following periods. In the period T15–T45, no differences in body weight gain was detected among the different experimental groups

Table 2

Lipemic values at the beginning (T0) and after the dietary treatment with HC1 diet (T15)

 

T0

T15

TC (mg/dl)

99.33 ± 10.03

145.46 ± 6.51*

LDL-C (mg/dl)

18.52 ± 1.07

22.31 ± 3.86°

HDL-C (mg/dl)

65.22 ± 3.10

21.85 ± 4.50*

HDL/LDL

3.54 ± 0.25

1.00 ± 0.22*

TG (μmol/dl)

117.80 ± 18.12

56.64 ± 6.29*

Data are means ± SD, n = 24

°P < 0.01; *P < 0.001, statistical analysis was carried out by the Student’s t test

In the second part of the study (T15–T45), the hypercholesterolemic rats were turned to diverse dietary treatments that differently influenced lipemia. In particular, ST and HC2 groups received water, while HC2P group received a water suspension of probiotic mixture. Serum TC and LDL-C levels (Figs. 4a and 5b, respectively), which increased during the consumption of the hypercholesterolemic diet HC1 for 14 days, decreased in rats fed ST and HC2P diets, whereas no modifications were observed in rats fed the HC2 diet.
https://static-content.springer.com/image/art%3A10.1007%2Fs00253-013-5088-2/MediaObjects/253_2013_5088_Fig4_HTML.gif
Fig. 4

Serum cholesterol level after the second part of the experimental dietary treatment. TC (panel a), LDL-C (panel b), HDL-C (panel c) and HDL/LDL ratio (panel d) were evaluated in the end of the second part of the experimental dietary period (T45) and compared to values obtained at T0 and T15. Data at T0 and T15 are means ± SD of 24 animals, and data at T45 are means ± SD of 8 animals in each dietary group. Statistical analysis was by the one-way ANOVA (TC, P < 0.01; LDL-C, P < 0.001; HDL-C, P < 0.001; HDL/LDL ratio, P < 0.001) with Tukey as post test. Different letters indicate statistical significance (P < 0.05)

https://static-content.springer.com/image/art%3A10.1007%2Fs00253-013-5088-2/MediaObjects/253_2013_5088_Fig5_HTML.gif
Fig. 5

Serum triglyceride level after the second part of the experimental dietary treatment. Serum TG level was evaluated in the end of the second part of the experimental dietary period (T45) and compared to values obtained at T0 and T15. Data at T0 and T15 are means ± SD of 24 animals, and data at T45 are means ± SD of 8 animals in each dietary group. Statistical analysis was by the one-way ANOVA (P < 0.001) with Tukey as post test. Different letters indicate statistical significance (at least P < 0.05)

The HDL-C level, which decreased in the T0–T15 period, significantly increased only in rats fed again the ST diet (Fig. 4c). The HDL/LDL ratio, decreased during the T0–T15 period, raised to a level even higher than T0 in rats turned back to the ST diet, while no significant modifications were observed in the other group (Fig. 4d). Serum TG level, decreased during the T0–T15 period, increased during the T15–T45 treatments, regardless the diet and probiotic administration (Fig. 5).

Discussion

Accumulating evidence suggests that probiotics exert various health biological properties, one of them being the anticholesterolemic activity through bile salts hydrolysis and cholesterol assimilation (Nagpal et al. 2012). In this light, the selection of specific strains and the evidence of their effectiveness, resulting in control of lipemic values, can be exploited to formulate novel probiotic foods or supplements that can exert a role in the CVD prevention. The cholesterol-lowering properties of lactobacilli has been assessed in vivo in several studies in humans and animal models, mostly consisting in the consumption of supplements and fermented foods containing selected strains of Lactobacillus (Kumar et al. 2011; Jones et al. 2013). Solid evidence of hypocholesterolemic effects of bifidobacteria has still to be provided, since they have been administered mostly in association with lactobacilli (Andrade and Borges 2009; Sadrzadeh-Yeganeh et al. 2010; Ooi and Liong 2010; Ejtahed et al. 2011). Furthermore, BSH activity is generally much higher in bifidobacteria than in lactobacilli (Tanaka et al. 1999), thus, the anti-cholesterol effect of bifidobacteria deserves to be investigated more exhaustively.

The present study aimed to develop and validate an anticholesterolemic probiotic formulation, based on a judicious selection of potential cholesterol-lowering bifidobacteria. Thirty-four human Bifidobacterium strains were screened for cholesterol assimilation and BSH activity, which proved to be strain-specific features. Cholesterol assimilation was a peculiar characteristic of two strains belonging to the species B. bifidum (B. bifidum MB 107 and B. bifidum MB 109), which removed 81 and 50 mg of cholesterol/g of biomass, respectively. It is relevant that the median of specific cholesterol absorption by bifidobacteria was 19 mg/g.

Among bifidobacterial species, significant differences in BSH activities could not be established, due to the wide distribution of both BSHGCA and BSHTDCA and to the low numerosity of some species. However, the screening resulted in the selection of promising strains able to efficiently deconjugate GCA and TDCA. No relationship was established between BSH phenotype and the extent of cholesterol assimilation, consistently with a previous study that explored the relationship among bile tolerance, the ability to deconjugate bile salts, and the ability to remove cholesterol (Tahri et al. 1997).

The strains B. bifidum MB 109, B. breve MB 113, and B. animalis subsp. lactis MB 2409 were selected for the probiotic formulation to be administered to rats. B. bifidum MB 109 cholesterol adsorption was among the highest observed values, and both B. breve MB 113 and B. animalis subsp. lactis MB 2409 presented high BSH activity, B. breve MB 113 being among the strains with both BSHTDCA and BSHGCA higher than the 75th percentile. The selection took into account also the technological properties of the strains, so that they could be manufactured in adequate amount and incorporated in the lyophilized formula without losing viability.

Hypercholesterolemia was induced in rats, in order to study the effect of the probiotic supplementation in animals with impaired lipemic values. In rats, hypercholesterolemia can be obtained by a combination of high cholesterol with abundant saturated fat (Hashimoto et al. 1998; Rahman et al. 2007), or by cholic acid supplementation (Ryu et al. 2006). In the present study, it was induced by feeding rats with the hypercholesterolemic diet HC1, containing 3 % cholesterol and 0.1 % cholic acid, for 14 days. Albeit the short period on the HC1 diet, a significant increase of serum TC and LDL-C was achieved, as well as a drastic reduction of HDL-C level (P < 0.001). The reported modification in total and fractionated cholesterol, as well as a reduction of serum TG level, are in agreement with previous works in which hypercholesterolemia was induced by diets enriched in cholesterol and cholic acid (Zulet and Martinez 1995; Hashimoto et al. 1998; Ryu et al. 2006). The reduction in serum TG level could be explained considering that during fasting, VLDLs carry most of the TG in plasma, VLDL TG and serum TG levels being almost the same. VLDL concentration is regulated not only by the rate of synthesis but also by the rate of clearance, i.e., VLDL transformation into LDL (Ginsberg 1998). It is therefore conceivable that fasting serum TG levels is dependent on VLDL/LDL conversion, in turn related to the high cholesterol concentration in the HC1 diet.

In the second part of the study, one group of rats was shifted to the ST diet to evaluate the effect of a reduced cholesterol intake. Two other groups continued on a hypercholesterolemic diet (HC2), one also receiving the probiotic formulation. Although the reduction of dietary cholesterol appeared the most effective treatment, the administration of the probiotic formulation also resulted in a significant reduction of total cholesterol and LDL-C. The probiotic treatment did not affect HDL-C levels and HDL-C/LDL-C ratio, but the relevant decrease of TC and LDL-C suggests that the selected probiotic strains acted in concert to counteract the diet-induced hypercholesterolemia induced in the animal model. The increase of fasting TG level in all dietary groups at T45 could be related to the reduction of cholesterol dietary concentration to 1 % or to normal value, this reducing LDL formation from VLDL.

The data herein reported support the use of properly selected probiotic strains in the control of cholesterol levels. In this study, bifidobacteria have been selected through a screening of two claimed mechanisms involved in cholesterol-lowering activity, BSH and cholesterol assimilation. This screening allowed us to supplement hypercholesterolemic rats with a probiotic mixture which actually improved the diet-induced dyslipidemia. Although a preeminent role of the cholesterol concentration in the diet must be acknowledged and the mechanisms of the observed effect were not conclusively assessed, the effectiveness of the proposed probiotic formulation represents a step ahead in the nutritional treatment of the diet-induced hypercholesterolemia. Further studies are needed for establishing the cholesterol-lowering action of B. bifidum WC 0417, B. breve WC 0420, and B. breve WC 0463 in humans, but results herein reported clearly evidence the advantage of the screening of strains prior to clinical studies.

Acknowledgments

This study was supported by a grant from Probiotical S.p.A. (Novara, Italy).

Conflict of interest

The authors declare that they have no conflict of interest.

Copyright information

© Springer-Verlag Berlin Heidelberg 2013