Mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic bacteria

Cholesterol homeostasis in our body is regulated primarily by the liver and also by the intestine. The liver regulates the de novo synthesis of cholesterol, the production of bile salts, which are the principal form of cholesterol excretion, the secretion of lipids in the bloodstream in the form of VLDL, the synthesis and storage of cholesterol, and also modulates the expression of the cellular receptor responsible for the uptake of blood lipoproteins by the liver (LDL receptor). Regarding the intestine, it regulates the absorption and excretion of dietary cholesterol, and the excretion of bile salts in the faeces⁽ Reference Zhao and Wright ^{13 )}. Thus, for treatment with probiotics to provide a hypocholesterolaemic effect, these micro-organisms must act mainly on the liver and intestine.

Probiotic bacteria can regulate the host’s cholesterol metabolism although it is important to consider that different strains may have unique clinical effects since each strain may use different mechanisms, alone or in combination, to cause the hypocholesterolaemic effect⁽ Reference Lahti, Salonen and Kekkonen ^{3 ,} Reference Tok and Aslim ^{14 )}. Then, we will describe and discuss the potential mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic bacteria (Fig. 1).

Fig. 1 Main mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic bacteria. Some probiotic bacteria are able to (1) deconjugate bile salts by the action of the bile salt hydrolase enzyme. Also, (2) co-precipitation of cholesterol with deconjugated bile salts, (3) assimilation and incorporation of cholesterol into the cell membranes, (4) conversion of cholesterol to coprostanol by the cholesterol reductase enzyme and (5) inhibition of the gene expression of the intestinal cholesterol transporter Niemann–Pick C1 like 1 (NPC1L) are some of the mechanisms by which probiotic bacteria reduce the absorption of intestinal cholesterol and lipids. The (6) fermentation of indigestible carbohydrates by probiotic bacteria leads to the production of propionate, which inhibits the activity of the liver enzyme 3-hydroxy-methyl-3-glutaryl-CoA (HMG-CoA) reductase. Additionally, these bacteria can modulate lipid metabolism (7), contributing to the hypocholesterolaemic effect.

Co-precipitation of cholesterol with the deconjugated bile salts

Intestinal cholesterol needs to be emulsified by the bile salts in order to be absorbed since it is a hydrophobic substance. Thus, the absorption of intestinal cholesterol is dependent on the action of the bile salts⁽ Reference Gérard ^{17 )}.

However, as previously discussed, some probiotic bacteria are capable of deconjugating bile salts in the intestinal lumen. Furthermore, in the deconjugated form, bile salts are less effective as emulsifiers than conjugated bile salts and do not form stable micelles⁽ Reference Lye, Rahmat-Ali and Liong ^{30 )}. Additionally, deconjugated bile salts at acidic pH are protonated and precipitated, and, consequently, intestinal cholesterol can be co-precipitated with deconjugated bile salts and will not be absorbed by the host⁽ Reference Fava, Lovegrove and Gitau ^{12 ,} Reference Ahn, Kim and Lim ^{31 )}.

Klaver & Meer⁽ Reference Klaver and Meer ^{32 )}, by means of an in vitro study, were one of the first to note that the cholesterol co-precipitation process is dependent on the presence of bile salts in the deconjugated form. Ahn et al. ⁽ Reference Ahn, Kim and Lim ^{31 )} observed a significant positive correlation between the ability of deconjugation of the bile salts and precipitation of cholesterol by L. acidophilus SNUL020, SNUL01 and FM01. Also, it was noticed that L. plantarum Lp91 co-precipitates greater amounts of cholesterol compared with L. plantarum Lp21 and NCDO82. These strains exhibited a lower ability to deconjugate sodium glycocholate compared with Lp91⁽ Reference Kumar, Grover and Batish ^{33 )}. Thus, the presence of bile salts in the deconjugated form is essential for the occurrence of this mechanism.

Nevertheless, the co-precipitation of cholesterol with deconjugated bile salts is highly dependent on the pH of the medium. First, the enzyme BSH needs a slightly acidic pH, usually between 5 and 6, to have optimum performance⁽ Reference Guo, Liu and Zhang ^{2 )}. Second, deconjugated bile salts precipitate at a pH less than 6. Moreover, the precipitation of cholesterol is co-dependent on the precipitation of these bile salts⁽ Reference Klaver and Meer ^{32 )}. Third, when the pH of the medium reaches a value of 7, cholesterol can re-dissolve in the medium⁽ Reference Tahri, Grill and Schneider ^{34 )}.

Thus, it is believed that the co-precipitation of cholesterol with deconjugated bile salts may not significantly contribute to the decrease in the absorption of intestinal cholesterol in vivo since the physiological pH of the intestinal lumen varies over a range of neutrality and alkalinity. However, in in vitro studies, this mechanism is listed as being responsible for the decreased absorption of intestinal cholesterol by the regular consumption of the probiotic bacteria.

Assimilation and incorporation of cholesterol into the cell membrane of the probiotics

The amino acids that comprise the peptidoglycan of the cell walls of the probiotic bacteria and the exopolysaccharides secreted by these micro-organisms are primarily responsible for the strong link between the cholesterol in the medium and the cell surface of the probiotic bacteria. Subsequently, this cholesterol is assimilated into the cell membrane of the bacteria and eliminated with them through the faeces⁽ Reference Lye, Kuan and Ewe ^{35 )}.

Currently, it is known that the ability to assimilate the cholesterol present in the medium is highly strain-specific⁽ Reference Lye, Rahmat-Ali and Liong ^{30 )}. Thus, in a study conducted with thirty-four strains of Bifidobacterium (B. animalis, B. bifidum, B. longum, B. adolescentis and B. pseudocatenulatum), it was observed that these bacteria were capable of assimilating 4 to 81 mg cholesterol/g dry biomass. The strains B. bifidum MB 109 and B. bifidum MB 107 were able to assimilate the highest amounts among the different strains studied⁽ Reference Bordoni, Amaretti and Leonardi ^{10 )}. Also, among the different species of lactic acid bacteria isolated from dairy products (ewe milk, traditional yoghurt, and sour buttermilk) traditionally eaten in Iran, it was observed that the L. brevis strains were able to assimilate more cholesterol (80% of the cholesterol after 9 h of incubation) than the other strains of lactic acid bacteria⁽ Reference Iranmanesh, Ezzatpanah and Mojgani ^{36 )}.

This strain specificity could be the consequence of the differences in the capacity of production of exopolysaccharides since there is a correlation between the amount of exopolysaccharide produced and the quantity of cholesterol assimilated by the strain⁽ Reference Tok and Aslim ^{14 )}. It is likely that the composition and structure of the peptidoglycan of the bacterial cell walls will also influence the ability of each strain to assimilate the cholesterol.

Additionally, differences in the expression of genes involved in intracellular cholesterol transport could be responsible for the differences observed between the strains. Recently, it was observed that the high ability exhibited by B. bifidum PRL2010 to assimilate cholesterol was due to the increased expression of the genes encoding the ABC-type carriers. When maintained in medium containing cholesterol the expression of these genes increased up to two times compared with the bacteria of the same strain maintained in a cholesterol-free medium. It was also observed that none of the genes related to cholesterol catabolism was up-regulated⁽ Reference Zanotti, Turroni and Piemontese ^{37 )}.

As a consequence of the cholesterol assimilation process, the fatty acid composition of the bacteria membrane is modified. Generally, there is an increase in the concentration of SFA and unsaturated fatty acids in the cell membranes⁽ Reference Miremadi, Ayyash and Sherkat ^{25 ,} Reference Zanotti, Turroni and Piemontese ^{37 –} Reference Lye, Rusul and Liong ^{39 )}. This modification in the composition is required since the membrane must maintain its integrity and fluidity after the cholesterol is assimilated. Also, this modification can lead to changes in the tensile strength and the charge of those membranes, which allow these bacteria to be more resistant to adverse conditions present in the intestinal lumen, preventing the occurrence of cell lysis⁽ Reference Kumar, Nagpal and Kumar ^{18 ,} Reference Lye, Rusul and Liong ^{39 )}. Perhaps, that is why Miremadi et al. ⁽ Reference Miremadi, Ayyash and Sherkat ^{25 )} observed that there was a positive correlation between the tolerance to the bile salt and the overall content of the assimilated cholesterol (r 0·808; P<0·05). Then, those strains that exhibited the greatest bile salt tolerance showed higher overall cholesterol assimilation in comparison with those strains that had a low tolerance.

In vivo studies were performed so that the hypocholesterolaemic effect of the regular consumption of probiotic bacteria that are able to assimilate cholesterol in vitro could be evaluated. In this manner, it was observed that Sprague–Dawley rats fed a hypercholesterolaemic diet and treated for 2 weeks with B. longum SPM1207 (10⁸–10⁹ CFU/ml) showed a significant reduction in serum TC and LDL-cholesterol concentration⁽ Reference Lee, Jang and Baek ^{40 )}. Also, Kumar et al. ⁽ Reference Kumar, Grover and Batish ^{33 )} selected the strains of L. plantarum Lp91 and Lp21 after performing an in vitro study, where those strains assimilated larger amounts of cholesterol in de Man, Rogosa and Sharpe (MRS) broth supplemented with Thio-0·3% ox-bile in comparison with L. plantarum NCDO82. Thus, Sprague–Dawley rats fed a hypercholesterolaemic diet supplemented with L. plantarum Lp91 and Lp21 (10⁸ CFU/g, for 21 d) showed a reduction in serum TC, LDL-cholesterol, VLDL and TAG concentrations.

Nevertheless, some in vivo studies have found conflicting results. Park et al. ⁽ Reference Park, Kim and Shin ^{41 )} did not attribute this mechanism to the hypocholesterolaemic effect observed by treatment with L. acidophilus ATCC 43121 (3×10⁷ CFU/d) since the concentration of cholesterol in the faeces of these animals did not differ from that of the control group. Also, the treatment with L. rhamnosus LC705 (two capsules/d containing 2×10¹⁰ CFU, during 4 weeks), a strain capable of assimilating cholesterol in vitro, did not affect the serum TC, LDL-cholesterol, HDL-cholesterol and TAG concentrations of adult subjects, with moderate elevation in serum cholesterol⁽ Reference Hatakka, Holma and El-Nezami ^{42 )}. However, this disagreement between the results of in vitro and in vivo studies is expected since the in vitro studies do not reproduce the real environmental conditions of the gastrointestinal tract⁽ Reference Madani, Mirlohi and Yahay ^{43 )}.

A limitation observed by us during the revision of this mechanism consists of the lack of quantification of the cholesterol excreted in the faeces by the subjects treated with probiotic bacteria. The result of this quantification could help to confirm the occurrence of the assimilation of cholesterol by the probiotics tested in vivo. Thus, studies that evaluate the occurrence of this mechanism should quantify the cholesterol excreted in the faeces.

According to Miremadi et al. ⁽ Reference Miremadi, Ayyash and Sherkat ^{25 )}, there is a negative correlation between the residual cholesterol in the medium and the growth of probiotic bacteria, which indicates that cholesterol assimilation is growth-dependent. Although probiotic bacteria that are non-growing (stationary phase) or that are dead are also able to assimilate cholesterol, it is only in smaller amounts, when compared with the bacteria that are in the growth phase⁽ Reference Lye, Rahmat-Ali and Liong ^{30 ,} Reference Sirilun, Chaiyasut and Kantachote ^{44 )}. Thus, the reduction of cholesterol concentration in the medium occurs as a consequence of the assimilation of growing cells and the adhesion of cholesterol to the surface of the dead cells⁽ Reference Miremadi, Ayyash and Sherkat ^{25 )}.

Thus, we can conclude that the ability of bacteria to assimilate the cholesterol in the medium appears to be dependent on the growth of the strain and is strain-specific. Then, if the strain is able to assimilate cholesterol, but does not survive the passage through the gastrointestinal tract, the occurrence of the hypocholesterolaemic effect can be compromised.

The presence of bile salts in the deconjugated form plays an important role in the process of assimilation of cholesterol by the probiotic bacteria. The bile salts in the deconjugated form increase the assimilation of cholesterol into the cell membrane of the probiotic bacteria since it increases the permeability, as well as the fluidity and porosity of the membranes, allowing the incorporation of cholesterol⁽ Reference Lye, Rahmat-Ali and Liong ^{30 )}. However, unlike what happens with the mechanism of co-precipitation of cholesterol, the presence of bile salts in the deconjugated form is not essential for the assimilation of cholesterol⁽ Reference Tok and Aslim ^{14 )}.

Despite being little studied, probiotic bacteria are also capable of binding bile salts present in the medium to their cell surface⁽ Reference Lye, Rahmat-Ali and Liong ^{30 )}. Pigeon et al.⁽ Reference Pigeon, Cuesta and Gililliand ^{45 )} observed that the strains of L. delbrueckii subsp. bulgaricus LB-18 and LB-10442 were capable of binding higher amounts of cholic acid than the other strains of Streptococcus thermophiles studied. According to these authors, the strains that produced the largest amounts of exopolysaccharides were capable of binding larger amounts of bile salts.

Similar to what happens with assimilated cholesterol, the bile salts will be eliminated along with the bacteria through the faeces. In this manner, endogenous cholesterol will be redirected to the production of more bile salts, which may improve the cholesterol serum profile of the host. Therefore, more studies to evaluate this mechanism should be conducted. Subsequently, after the identification of the strains capable of binding the bile salts, in vivo studies should be performed for the evaluation of the hypocholesterolaemic effect of this mechanism.

Conversion of cholesterol to coprostanol

In the intestinal lumen dietary cholesterol and endogenous cholesterol excreted via the transintestinal cholesterol efflux can be metabolised by the colonic microbiota. Thus, cholesterol can be reduced to coprostanol (5β-cholestan-3β-ol) and in minor amounts to coprostanone. These metabolites have low intestinal absorption and are eliminated with the faeces, which leads to a decrease in the intestinal absorption of cholesterol⁽ Reference Gérard ^{17 )}.

The efficiency of cholesterol conversion to coprostanol is mainly the result of the activity and abundance of cholesterol-reducing bacteria in the intestinal microbiota⁽ Reference Gérard ^{17 )}. Thus, numerous attempts have been made to identify and isolate the cholesterol-reducing bacteria, so it can be used in the prevention and treatment of hypercholesterolaemia.

The cholesterol-reducing bacteria must have the cholesterol reductase enzyme. This enzyme can be found in some strains of probiotic bacteria, like L. acidophilus ATCC 314, L. acidophilus FTCC 0291, L. bulgaricus FTCC 0411, L. casei 1311 FTDC ATCC 393, B. bifidum PRL2010, and Eubacterium coprostanoligenes ATCC 51222⁽ Reference Gérard ^{17 ,} Reference Zanotti, Turroni and Piemontese ^{37 ,} Reference Lye, Rusul and Liong ^{39 )}. There is little information in the scientific literature about the strains positive for this enzyme; however, the activity of this enzyme is strain-specific.

Recently, it has been observed that the cholesterol reductase enzyme can be modulated by the medium in which the bacteria are located. Thus, the cells of B. bifidum PRL2010 cultivating in the medium containing cholesterol have an increased expression of the gene BBPR_0519 in comparison with the cells cultivated in the absence of cholesterol. The BBPR_0519 gene encodes the enzyme aldo/keto reductase, which converts cholesterol in coprostanol⁽ Reference Zanotti, Turroni and Piemontese ^{37 )}.

Some in vivo studies were designed to investigate the hypocholesterolaemic effect of the treatment with cholesterol-reducing bacteria. In this manner, the treatment (2×10⁷ cells/ml) with E. coprostanoligenes ATCC 51222 increased the coprostanol:cholesterol ratio in the contents of the digestive tracts of experimental rabbits, in comparison with the controls. Subsequently, this result was assigned to the hypocholesterolaemic effect observed in the animals treated⁽ Reference Li, Buhman and Hartman ^{46 )}.

However, some studies have failed to prove that the hypocholesterolaemic effect observed was a result of an increase in the conversion of cholesterol to coprostanol. In this manner, Sprague–Dawley rats fed a cholesterol-enriched diet and treated with L. acidophilus ATCC 43121 (2×10⁶ CFU/d), for 3 weeks, showed faecal excretion of coprostanol similar to that of the control group⁽ Reference Park, Kim and Shin ^{26 )}. A similar result was found in the study conducted by Guo & Li⁽ Reference Guo and Li ^{20 )}, where treatment with L. casei F0822 (10⁸ or 10⁹ CFU/ml, for 3 weeks) did not change the faecal concentration of coprostanol. Also, healthy subjects who had consumed 100 ml of milk fermented with L. acidophilus SNUL01 (10⁸ CFU/ml, during 3 weeks) exhibited faecal excretion of coprostanol similar to that of the placebo group⁽ Reference Ahn, Kim and Lim ^{31 )}. However, none of the studies mentioned above reported whether those probiotic bacteria have the cholesterol reductase enzyme.

The activity of the cholesterol reductase enzyme is affected by the presence of bile salts and by the type of bile salts present in the medium. Thus, the conversion of cholesterol in coprostanol is reduced in the medium supplemented with oxgall and taurocholic acid⁽ Reference Lye, Rusul and Liong ^{39 )}. Therefore, the utilisation of a strain positive for the cholesterol reductase enzyme is not sufficient to increase the conversion of cholesterol to coprostanol, as the environmental conditions in the gastrointestinal tract, such as the presence/absence of cholesterol and bile salts, can influence the expression and activity of this enzyme.

Inhibition of the expression of intestinal cholesterol transporter Niemann–Pick C1 like 1 in the enterocytes

Through transporters present in the membrane of enterocytes, the cholesterol present in the intestinal lumen is absorbed. In this way, the transporter Niemann–Pick C1 like 1 (NPC1L1) is indispensable for the absorption of dietary and biliary cholesterol. Thus, NPC1L1 knockout (NPC1L1-/-) mice absorb 69% less cholesterol than the NPC1L1 knockin (+/+) mice. Moreover, they are resistant to the hypercholesterolaemic effect induced by treatment with cholesterol-enriched diets⁽ Reference Altmann, Davis and Zhu ^{47 ,} Reference Davis, Zhu and Hoos ^{48 )}.

Based on the importance of the NPC1L1 cholesterol transporter, an investigation of the effect of the regular consumption of probiotics on the expression of these transporters in the gut of their consumers was initiated.

In an in vitro study with Caco-2 enterocytes, treatment with L. plantarum Lp27 was able to reduce the expression of the cholesterol transporter NPC1L1, as well as reducing the absorption of cholesterol in those cells⁽ Reference Huang, Wu and Wang ^{49 )}. Huang & Zheng⁽ Reference Huang and Zheng ^{7 )} observed that Caco-2 enterocytes treated with L. acidophilus ATCC 4356 expressed lower amounts of mRNA and protein of the cholesterol transporter NPC1L1than the control. Additionally, heat-killed ATCC 4356 was unable to inhibit NPC1L1 expression, although the conditioned medium of this strain was able to reduce NPC1L1 expression even after being inactivated by heat. Thus, the authors concluded that ATCC 4356 secreted some sort of heat-stable molecules, for example, small oligopeptides or lipids, which were responsible for inhibiting the expression of this NPC1L1 transporter.

Clinical trials that evaluated this mechanism are scarce since it is necessary to perform a biopsy for tissue sample collection. Therefore, an experimental study conducted with hamsters that were fed with a hyperlipidaemic diet and treated with a mix of L. acidophilus La5 and B. animalis subsp. lactis Bb12 (2×10⁹ of each strain, for 5 weeks), it was observed that those expressed lower amounts of the transporter NPC1L1 in the gut when compared with the animals that were fed with a hyperlipidaemic diet (positive control) and those that were not fed with the hyperlipidaemic diet (negative control). The treatment with the probiotic was also able to decrease serum TC, oxidised LDL-cholesterol and TAG concentration, as well as the hepatic TC and TAG concentration⁽ Reference Stancu, Sanda and Deleanu ^{11 )}.

Similar results were observed in rats treated with L. plantarum Lp27 (10⁹ CFU/d), isolated from kefir grains, for 4 weeks. These animals expressed minor amounts of the transporter NPC1L1 in the duodenal and jejunal regions and exhibited a reduction in serum TC, LDL-cholesterol and TAG concentrations in comparison with the control group. The serum HDL-cholesterol concentration was unchanged; also, a reduction in liver TC and TAG concentrations was observed⁽ Reference Huang, Wu and Wang ^{49 )}. According to Altmann et al.⁽ Reference Altmann, Davis and Zhu ^{47 )}, the transporter NPC1L1 is mainly expressed on the surface of the enterocytes located in the jejunal region. In this manner, when it becomes less expressed in this region, cholesterol absorption is compromised.

The molecular mechanisms used by the probiotic bacteria to reduce NPC1L1 transporter expression are not fully understood. Huang & Zheng⁽ Reference Huang and Zheng ^{7 )} observed that when liver X receptors (LXR) are depleted by small interfering RNA in Caco-2 cells, NPC1L1 expression is no longer decreased by treatment with L. acidophilus ATCC 4356, and no reduction in cholesterol uptake is observed. In this manner, it has been proposed that some molecules produced by the probiotic bacteria will be able to connect and stimulate the LXR, down-regulating NPC1L1 protein expression in the small intestine.

Inhibition of hepatic synthesis of cholesterol

Inhibition of the activity of the hepatic enzyme 3-hydroxy-methyl-3-glutaryl-CoA (HMG-CoA) reductase, a limiting enzyme in the process of hepatic cholesterol synthesis, is another mechanism described in the literature, as responsible for the hypocholesterolaemic effect of the regular consumption of probiotic bacteria. The inhibition of the hepatic activity of HMG-CoA reductase occurs mainly as a consequence of the action of SCFA⁽ Reference Pavlović, Stankov and Mikov ^{51 )}.

SCFA are the endproducts of bacterial fermentation of non-digestible carbohydrates by the host. Due to the low pKa (<4·8) of these acids and the pH of about 7 present in the colonic region, SCFA are usually found in the form of anions⁽ Reference Guilloteau, Martin and Eeckhaut ^{52 )}. The major SCFA produced are acetate, propionate and butyrate; other bacterial endproducts produced in minor amounts include lactate, succinate, formate, valerate, caproate, isobutyrate, 2-methyl-butyrate and isovalerate⁽ Reference Vipperla and O’Keefe ^{53 )}.

Each SCFA plays a different role in our body, and some of them play opposite roles. In the inhibition of the activity of HMG-CoA reductase, SCFA propionate and acetate are the main ones involved in the process⁽ Reference Fava, Lovegrove and Gitau ^{12 )}. After being absorbed, acetate reaches the liver, where it is converted to acetyl-CoA by the activity of the enzyme acetyl-CoA synthetase. Acetyl-CoA is then used for the synthesis of cholesterol by the activity of the enzyme HMG-CoA reductase. Thus, the SCFA acetate increases serum cholesterol concentration⁽ Reference Henningsson, Bjiirck and Nyman ^{54 )}. On the other hand, the SCFA propionate acts as an inhibitor of lipogenesis and cholesterogenesis. Propionate is able to inhibit the incorporation of acetate into the molecules of steroids since it inhibits the activity of HMG-CoA reductase⁽ Reference Wolever, Spadafora and Cunnane ^{55 )}. Furthermore, propionate is able to stimulate the hepatic synthesis of bile salts, by increasing the activity of the 7α-hydroxylase enzyme⁽ Reference Arora, Sharma and Frost ^{56 )}. Therefore, propionate has a hypocholesterolaemic effect, while acetate presents a hypercholesterolaemic effect.

Thus, for the probiotic micro-organisms to have a hypocholesterolaemic effect, sufficient amounts of propionate must be produced to offset the effects of acetate in cholesterogenesis. Therefore, several in vivo studies were designed to investigate the hypocholesterolaemic effect of the SCFA produced by the probiotic micro-organisms. In this manner, Sprague–Dawley rats fed a cholesterol-enriched diet and treated with L. plantarum MA2 (10¹¹ cells/d, for 5 weeks) exhibited higher caecal concentrations of propionate than the control group. Additionally, the probiotic treatment lowered the serum TC, LDL-cholesterol and TAG concentrations, as well as the liver TC and TAG concentrations⁽ Reference Wang, Xu and Xi ^{57 )}. Moreover, the hypocholesterolaemic effect of the treatment with different doses (10⁸ or 10⁹ CFU/ml; for 3 weeks) of L. casei F0822 was associated with an increase in caecal concentrations of propionate and total SCFA. Additionally, the propionate:acetate ratio was increased in the group treated with the probiotic in comparison with the control group⁽ Reference Guo and Li ^{20 )}. Thus, part of the hypocholesterolaemic effect observed in these studies may be attributed to the increased production of the SCFA propionate.

However, conflicting results have been found in clinical trials. It was seen that healthy, mildly hypercholesterolaemic male subjects, who consumed the fermented milk kefir (500 ml/d), for 4 weeks, did not exhibit any changes in serum TC, LDL-cholesterol, HDL-cholesterol or TAG concentrations in comparison with baseline. Also, the faecal concentration of total SCFA and propionate was increased in those subjects supplemented with kefir ⁽ Reference St-Onge, Farnworth and Savard ^{58 )}. Nonetheless, male subjects with atherosclerotic plaques in the carotid wall, who had ingested an oat-based drink fermented by L. plantarum DSM 9843 (10¹¹ CFU/d) for 4 weeks, exhibited a similar faecal concentration of total SCFA, acetate and propionate, compared with the placebo group. Additionally, serum TC, LDL-cholesterol, HDL-cholesterol and TAG concentrations were not affected by treatment with the fermented drink⁽ Reference Karlsson, Ahrné and Molin ^{8 )}.

The lack of association between the hypocholesterolaemic effect and the increased faecal concentration of the SCFA propionate in the clinical trials could be a consequence of the nature of the sample utilised for the determination of these acids. It is known that a greater production of SCFA occurs in the caecum; from this region toward the distal colon, the concentration of these acids decreases progressively (up to 95%). Along the route, the SCFA are rapidly absorbed by the colonocytes. Furthermore, in the distal colon, the production of SCFA is small, due to the low availability of substrate⁽ Reference Kumar, Nagpal and Verma ^{59 )}. In this manner, when the concentration of SCFA is evaluated in the faeces, the results will not reflect the actual concentration of the SCFA produced by the intestinal microbiota. Moreover, the experimental studies usually evaluated the concentration of the SCFA in the caecum content, where most of the SCFA is produced.

Another possible limitation of those studies is the fact that the bacteria of the genus Lactobacillus are not capable of producing SCFA with more than two carbons. Therefore, the increase in propionate production observed in all studies that utilised this genus of bacteria would be a consequence of the metabolic activity of the other bacteria present in the intestinal microbiota of the host. Those micro-organisms would convert the lactate produced by Lactobacillus into propionate⁽ Reference Guo and Li ^{20 )}. Therefore, this increase in propionate production is affected by the initial composition of the intestinal microbiota of the host.

The increase in the total amount of SCFA produced may decrease intracolonic pH. As seen previously, at low pH values the bile salts are protonated and precipitated. Thus, the increase in the production of the SCFA contributed to the occurrence of the co-precipitation of cholesterol with the deconjugated bile salts⁽ Reference Klaver and Meer ^{32 )}. Furthermore, the decrease in intracolonic pH could favour an increase in the production of the BSH enzyme by the bacteria, which would consequently lead to an increase in the excretion of cholesterol by the host⁽ Reference Miremadi, Ayyash and Sherkat ^{25 )}.

In conclusion, the hypocholesterolaemic effect of the SCFA depends on the amount of propionate and acetate that is produced by the probiotic bacteria in the intestinal lumen since they have opposite effects on the hepatic synthesis of cholesterol. Thus, to have the hypocholesterolaemic effect, it is necessary to consume probiotic bacteria that produce more propionate than acetate during the fermentation process. Additionally, the production of SCFA can be increased when the host consumes a prebiotic.

Reduction of body fatness

Considering that subjects with high body fatness are more likely to develop dyslipidaemia, and the intestinal microbiota is capable of modulating the body fatness of its host⁽ Reference Bäckhed ^{50 )}, changes in the intestinal microbiota composition by the regular consumption of probiotics could be an important mechanism to reduce body fatness⁽ Reference Aronsson, Huang and Parini ^{60 )} and indirectly improve the serum lipid profile.

In this manner, it has been suggested that some strains of probiotic bacteria could increase lipolysis and reduce the lipogenesis process of their host, consequently improving the serum lipid profile⁽ Reference Cheik, Rossi and Guerra ^{61 ,} Reference Moon, Baik and Cha ^{62 )}. Those strains would be capable of down-regulating the expression of genes related to the lipogenesis process such as the enzymes fatty acid synthase and acetyl CoA carboxylase, as well as LXR-α and sterol regulatory element-binding protein 1. Also, it could also up-regulate the expression of genes related to the lipolysis process, such as the enzyme carnitine palmitoyltransferase-1, and PPAR-α⁽ Reference Moon, Baik and Cha ^{62 –} Reference Miyoshi, Ogawa and Higurashi ^{64 )}.

Although the molecular mechanism by which these micro-organisms mediate this modulation is not fully known, it is likely that this effect is a result of the interaction between the probiotic bacteria or their metabolites (SCFA and bioactive peptides) and the host⁽ Reference Bäckhed ^{50 )}. Thus, more studies that evaluate how the regular consumption of probiotics can reduce body fatness should be performed for a better understanding of how this mechanism influences the hypocholesterolaemic effect of probiotic bacteria.

Mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic yeasts

Yeasts have diverse fermentative activities, and therefore have been used for the production of many kinds of fermented foods⁽ Reference Yoshida, Yokoi and Wada ^{65 )}. Although many strains of yeasts are used in the food industry, there are few studies evaluating the hypocholesterolaemic effect of the regular consumption of probiotic yeast. Thus, little is known about the mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic yeast (Fig. 2).

Fig. 2 Main mechanisms responsible for the hypocholesterolaemic effect of regular consumption of probiotic yeasts. Some probiotic yeasts are able to (1) deconjugate bile salts by the action of the bile salt hydrolase enzyme. The (2) co-precipitation of cholesterol with deconjugated bile salts and (3) assimilation and incorporation of cholesterol into the cell membranes are some of the mechanisms by which probiotic yeasts reduce the absorption of intestinal cholesterol and lipids. Additionally, (4) consumption of the polysaccharides extracted from the cell wall of the yeast can be used as substrate by the intestinal microbiota, which may lead to an increase in the production of propionate, which inhibits the activity of the liver enzyme hydroxy-methyl-3-glutaryl-CoA reductase.

The hypocholesterolaemic effect of probiotic yeasts is strain-specific, as reported by Yoshida et al. ⁽ Reference Yoshida, Yokoi and Wada ^{65 )}; among eighty-one strains of yeasts evaluated, only forty-eight were able to decrease the serum TC concentration when administered as a dietary admixture at a concentration of 10% ad libitum, for 7 d. This variability observed between the strains has been attributed to the differences in the composition of the cell wall. The cell wall of yeasts is mainly composed of β-glucan and α-mannan, two types of dietary fibres; however, the structures of those polysaccharides differ according to the strain studied⁽ Reference Yoshida, Yokoi and Ohishi ^{66 )}.

As consequence of the high concentration of nucleic acid in yeast, which could increase consumer’s serum uric acid concentration, as well as the high doses of yeast polysaccharides tested, the hypocholesterolaemic effect of polysaccharides present in the cell wall of yeast has been studied in the isolated form⁽ Reference Waszkiewicz-Robak ^{67 )}. Thus, Sprague–Dawley rats fed a cholesterol-enriched diet, supplemented with β-glucan (30 g/kg of diet; for 8 weeks) isolated from the baker’s yeast Saccharomyces cerevisiae, exhibited a reduction in serum TC and LDL-cholesterol concentrations, while HDL-cholesterol and TAG concentrations were not affected by the treatment⁽ Reference El-Arab, Foheid and El-Said ^{68 )}. Similar results were observed in the study by Wilson et al.⁽ Reference Wilson, Barbato and Nicolosi ^{69 )}, where Syrian golden hamsters fed on a hypercholesterolaemic diet supplemented with 10 g of yeast β-glucan/100 g of diet, for 12 weeks, showed a reduction in serum TC, non-HDL-cholesterol and TAG concentrations. Additionally, treatment with yeast β-glucan reduced the aortic fatty streak area as compared with the control group.

The hypocholesterolaemic effect of these polysaccharides has also been observed in clinical trials. In this manner, the cell wall extract of Kluyveromyces marxuanus YIT 8292 (2 and 4 g/d, for 4 weeks) when consumed by mildly hypercholesterolaemic subjects reduced serum TC and LDL-cholesterol concentrations. A similar effect was observed when normal and hypercholesterolaemic subjects consumed a yogurt supplemented with 3 or 4 g/d of polysaccharides extracted from the cell wall of K. marxuanus YIT 8292, for 8 weeks⁽ Reference Yoshida, Naito and Ohishi ^{70 )}.

The hypocholesterolaemic effect caused by the consumption of the polysaccharides extracted from the cell wall of the yeast may be due to an increase in the faecal excretion of bile salts, which increases the demand for cholesterol for the hepatic synthesis of bile salts. Moreover, these polysaccharides can be used as a substrate by the intestinal microbiota, which may lead to an increase in the production of SCFA, especially propionate, which as previously described can decrease the hepatic synthesis of cholesterol⁽ Reference Yoshida, Yokoi and Ohishi ^{66 )}.

Similar to probiotic bacteria, some strains of yeasts are capable of producing the BSH enzyme, and consequently deconjugating bile salts⁽ Reference Liu, Xie and Xiong ^{71 )}. This would lead to an increase in the excretion of endogenous cholesterol, which would stimulate the hepatic synthesis of bile salts, and consequently reduce the amount of absorbed intestinal cholesterol; hence the formation of micelles will be compromised⁽ Reference Kumar, Nagpal and Kumar ^{18 )}.

Liu et al. ⁽ Reference Liu, Xie and Xiong ^{71 )} observed that the BSH activity of K. marxianus K1 and M3 is proportional to the amount of cholesterol that is removed from the culture medium. Also, partly, the decrease in the concentration of cholesterol is due to its co-precipitation with the deconjugated bile salts. Thus, the mechanism of co-precipitation of cholesterol carried out by yeasts resembles that made by bacteria. In this manner, Wistar rats fed a high-cholesterol diet and treated with different doses (5, 10 and 20 g/kg body weight) of K. marxianus M3 isolated from Tibetan mushrooms, for 7 weeks, showed a reduction in serum TC, LDL-cholesterol and TAG concentrations, and an increase in serum HDL-cholesterol concentration⁽ Reference Xie, Zhang and Liu ^{72 )}.

Yeasts are also able to assimilate the cholesterol present in the medium; however, this mechanism is highly strain-specific and growth-dependent. Thus, from the nine strains of yeasts evaluated, only Saccharomyces cerevisiae KK1, Saccharomyces cerevisiae 832, Saccharomyces boulardii and Isaatchenkia orientalis KK5.Y.1 were able to assimilate significant amounts of cholesterol present in the culture medium, in comparison with the other strains⁽ Reference Psomas, Fletouris and Litopoulou-Tzanetaki ^{73 )}.

Psomas et al. ⁽ Reference Psomas, Fletouris and Litopoulou-Tzanetaki ^{73 )} observed that the yeast cells that were grown in the medium containing both oxgall and cholesterol were more resistant to lysis by sonication. After enzymic lysis, the authors concluded that the assimilated cholesterol is incorporated into the membrane or cell wall and is not metabolised by them. Therefore, the cholesterol assimilation mechanism of probiotic yeasts is similar to that of probiotic bacteria.

In an experimental study, the consumption of Saccharomyces boulardii (12×10¹⁰ CFU/kg of diet; for 14 d) was able to prevent and treat hypercholesterolaemia and hypertriacylglycerolaemia in golden Syrian hamsters fed on a diet enriched with 0·1% cholesterol⁽ Reference Girard, Pansart and Verleye ^{74 )}. The hypocholesterolaemic effect of S. boulardii may be a consequence of the assimilation of cholesterol⁽ Reference Psomas, Fletouris and Litopoulou-Tzanetaki ^{73 )}. Also, Aloğlu et al. ⁽ Reference Aloğlu, Özer and Öner ^{75 )} attributed the hypocholesterolaemic effect observed by them to the high cholesterol assimilation capacity (73·33%) of Cryptococcus humicola M5-2. Thus, Wistar rats fed a cholesterol-enriched diet supplemented with 0·1% of C. humicola M5-2 lyophilised (1×10⁷ cells/ml) had a significant reduction in serum TC and TAG concentrations in comparison with the control.

In a clinical study, Ryan et al. ⁽ Reference Ryan, Hanes and Schafer ^{76 )} attributed the decrease in serum remnant lipoprotein particle concentrations in hypercholesterolaemic adults treated with Saccharomyces boulardii (1·4×10¹⁰ CFU/capsule, eight capsules/d, for 8 weeks) to the ability of this strain to assimilate the cholesterol present in the medium. However, serum TC, LDL-cholesterol and HDL-cholesterol concentrations were not altered by the treatment with the yeast.

Further studies are needed to evaluate the hypocholesterolaemic effect of probiotic yeasts, as the results are promising, although clinical trials are scarce. Through more studies, it will also be possible to identify more mechanisms through which regular consumption of yeast can cause the hypocholesterolaemic effect.