|Home | About | Journals | Submit | Contact Us | Français|
Many current experiments investigating the effects of diet, dietary supplements, and pre- and probiotics on the intestinal environments do not take into consideration the potential for using bile salts as markers of environmental change. Intestinal bacteria in vertebrates can metabolize bile acids into a number of different structures, with deamidation, hydroxyl group oxidation, and hydroxyl group elimination. Fecal bile acids are readily available to sample and contain a considerable structural complexity that directly relates to intestinal morphology, bile acid residence time in the intestine, and the species of microbial forms in the intestinal tract. Here we offer a classification scheme that can serve as an initial guide to interpret the different bile acid patterns expressed in vertebrate feces.
Electrospray mass spectrometry is particularly well suited for the analysis of negatively charged molecules, such as bile acids. When applying the instrument to the analysis of fecal matter, it is a natural assumption to anticipate that many of the signals will originate from partly digested food, including mono- and diacyl phospholipids; mono-, di-, and triglycerides; and a variety of fatty acids. Such signals are seldom observed because evolution has increased the efficiency of the pancreatic enzymes and intestinal absorption to a near maximum, and fecal flora is exposed to the leftovers. The great bulk of signals from fecal material consists of a background level of difficult-to-identify peaks originating from a morass of bacterial membranes over which are superimposed comparatively large peaks of bile acids. It is only in situations in which intestinal input and output balance is disturbed (rapid transit, floral disturbance, intestinal disease) that the classic lipids are encountered.
Currently, experiments and advances in nutrition are limited by the difficulty in objectively defining and measuring the effects of nutrients, pre- and probiotics, and microbial flora alterations on intestinal cell health and uptake in a mixed aerobic/anaerobic environment dominated by dynamic micelle formation. For the most part, the large peaks of fecal bile acids are ignored in studies of nutrition, yet these chemical compounds display considerable structural complexity, one that directly reflects their passage through the intestinal environment. These structural alterations have the potential to open a hidden window into the functioning of the normal intestinal state and to address how the physiological environment is changed in the face of either altered nutrition or health.
Bile acids are derived from cholesterol (Fig. 1) and take on the structure of planar amphipathic molecules having a hydrophobic face with methyl groups and a hydrophilic face with hydroxyl groups. Bile acids are produced by every class of vertebrate animals and show substantial diversity across species (1). In bile, bile acids rapidly form mixed micelles with secreted cholesterol and phospholipid (2). Once in the small intestine, these mixed micelles alter their composition with the major solubilized lipids becoming dietary fatty acids, monoglycerides, and fat-soluble vitamins. These mixed micelles function in complex ways to solubilize hydrophobic compounds while concurrently accelerating their diffusion toward the intestinal epithelium (3). Interaction with food particles can also have the opposite effect of slowing digestion, as observed for starch/bile acid complexes (4). Bile acids enter the intestine as di- and trihydroxylated acyl conjugates, most often with the amino acid taurine (see Figure 1 for the structure of taurocholic acid), but in some species, conjugates are formed with glycine.
Within the intestine, bile acids share the environment with 1 × 1013 commensal bacteria, integral components of the human body collectively containing 100 times more genes than the human genome (5). The intestinal microbiota species in both humans and mice consist largely of Bacteroidetes and Firmicutes, with Clostridia representing >95% of the Firmicutes (6, 7). The intestinal appearance of conjugated bile acids directly affect the bacterial flora because they have intrinsic antimicrobial properties and, in addition, stimulate the enterocytes to secrete undefined antimicrobial compounds (8). For example, oral bile acids have been shown to reduce commensal bacterial overgrowth (9). In response to the presence of bile acids, bacteria structurally alter or “damage” bile salts. Essentially, it is a race between the ability of bile acids to retain their original structures and to maintain a high intraluminal intestinal concentration (necessary for hydrophobic lipid digestion) and the ability of bacteria to alter the structures and detergent properties of bile salts. The initial step in “damage” is the removal of the amino acid on the side chain, a process termed deconjugation (see Figure 1 for the structures of taurocholic acid and deconjugated cholic acid). A great number of commensal bacterial species contain bile salt hydrolases responsible for deconjugation, see Ridlon et al. (10) for a partial listing. Once deconjugated, most trihydroxy bile acids are sufficiently polar to remain in the intestinal lumen, but a portion of dihydroxy bile acids flip-flop across intestinal cell membranes and exit the intestinal lumen altogether. There is evidence that the process of deconjugation increases the overall excretion of bile acids in feces (11), and this idea has formed the basis of attempts to decrease serum cholesterol using probiotics possessing bile acid deconjugation activity (12, 13).
Many vertebrate species, such as humans, have an anaerobic cecum, and in this environment, the process of damage continues because a limited number of commensal bacteria species are capable of removing the 7-hydroxyl group from deconjugated di- and trihydroxy bile acids to form structures termed 7-deoxy bile acids. The most common 7-deoxy bile acids are named lithocholic acid (a monohydroxy bile acid) and deoxycholic acid (a dihydroxy bile acid), as shown in Figure 1. In the domestic mouse, the corresponding dihydroxy bile acid is 3α,7β-dihydroxy-5β-cholan-24-oic acid. Other potential bile salt bacterial metabolites include the keto- or oxobile acids, where ≥1 of the hydroxyl groups are oxidized (resulting in the loss of 2 protons). The redox status of the intestinal environment is a difficult to measure but important feature of diet and bacterial flora, and the proportion of oxo bile acids can range from 0 to nearly 20% of total bile acids (14). After their reuptake in the ileum by the intestinal apical bile salt transporter, both intact and altered bile acids are returned to the liver in portal blood. Damaged bile acids are “repaired” in part or completely by hepatic enzymes (15). The extent of repair (particularly for rehydroxylation at position 7) is species dependent; however, all deconjugated bile acids are reconjugated before being resecreted back into bile. The complete pathway of hepatic secretion, function in the intestine lumen, intestinal cell reuptake, return to the liver (and in some cases hepatic repair of damage), and resecretion is termed enterohepatic cycling. As a result of enterohepatic cycling, biliary bile acids in any 1 species consist of mixtures of di- and trihydroxylated bile acid structures, conjugated with taurine (or, rarely, glycine).
Fecal bile salts represent the remaining small percentage of bile acids that escape intestinal cell reuptake and become excreted. Understanding the pattern of fecal bile acid output begins with the mixture of bile acids secreted by the liver into bile. What is observed in the feces are those bile acids that are a) additional reflections of commensal bacterial modification in the small intestine; b) the fraction of bile acids that are not passively absorbed along the length of the intestine; and c) the remaining fraction of bile acids that are not actively absorbed in the ileum. For example, the absorption of bile acids in the intestinal region from the distal ileum to the rectum is apparently inhibited by the proportion of taurine present (16): d) those bile acids that survive additional modifications (and absorption) in the large bowel and e) those bile acids trapped within or adsorbed to dietary components such as fiber. A large number of fiber types have been administered to experimental animals. For example, in soluble cellulose-fed hamsters, the amounts of deoxycholic and lithocholic acids appearing in feces nearly doubled (17). Overall, the literature on the effect of different fiber types on human fecal bile acid output is somewhat mixed (subject to the type of fiber being administered), ranging from reduced, unchanged, or increased (18). It is unfortunate that the term “fiber” encompasses so many different chemical structures, each with their own metabolic interactions, and all without established mechanisms to provide definitive explanations on the observed outcomes. Together, factors a) through e) combine to allow for a species and diet-dependent escape of bile acids into the feces. For example, in humans ~400 to 800 mg of bile acids escape small intestinal uptake and are exposed to microbial reactions in the large intestine (10). In general, the rate of intestinal transit is a minor factor in altering the amount of bile acids that appear in the feces (19). A number of studies examined the impact of high-saturated fat and high-meat diets on fecal bile acid output, without finding significant results (20, 21). The resulting final pattern as it appears in the feces can be grouped into different classifications, based on the type of structural changes seen in the bile acid molecules.
This classification scheme is based on profiles generated by electrospray mass spectrometry. A large literature exists on fecal bile acid analysis using a variety of methods (see Setchell et al. (18) for a review); however, the ease and growing use of electrospray mass spectrometry make this a method of choice. It is convenient to break bile salt structural changes into 6 categories (I–VI), each of which is dependent on intestinal architecture (length, presence or absence of a cecum), transit time, diet, and the environment of the resident bacteria. The scheme also loosely follows the time course change in the different structures.
In category I are found intestinal systems in which the appearance of the fecal bile acids is essentially identical to that of gallbladder bile. In general, this pattern is somewhat exceptional, in that the great majority of fecal bile acids show at least some alterations. Still, there are a number of evolutionary pathways by which bile salt structural stability can be achieved, and these are broken down into 3 subcategories:
Bile acids are unchanged due to the bile acid possessing a chemical structure resistant to bacterial degradation, rapid intestinal transit, and/or the absence of intestinal bacteria. For example, the giant panda is a short-gut carnivore that consumes a fiber-rich bamboo diet. This combination, along with a rapid (5 h) intestinal transit time leaves the intestine nearly bacteria free, with a resulting Ia bile salt pattern (22).
Bile acids are unchanged due to the bile acid possessing a chemical structure resistant to bacterial deconjugation. These bile acids are exposed to bacteria, yet due to unique structural properties, are impervious to bacterial enzymes. See IIc for a discussion on the structures that achieve resistance to bacterial deconjugation.
This pattern is seen in species that use bile alcohols (conjugated with sulfate) that survive intestinal passage unchanged. This pattern is shown in Figure 2 for the Zebra danio (Zebrafish, Danio rerio), which demonstrates a pattern of fecal bile salts that is essentially identical to previous characterizations of bile salts from zebrafish bile (23). In contrast to bile acids conjugated with amino acids, most bacteria do not alter bile salt sulfates. There are a number of vertebrate Paenungulates that also use C27 bile alcohols (1). In these mammals, methanogenesis by commensal bacteria is an important sink for metabolic hydrogen, and the presence of C24 bile acids, physiologically similar in properties to long-chain fatty acids, would strongly inhibit methanogenesis (24).
In category II are found intestinal systems in which bile acids are partly deconjugated. The extent of deconjugation is of interest because the amino acid conjugate taurine contains a sulfonic acid moiety. Bacteria catabolize the released taurine into isethionic acid (25) and toxic hydrogen sulfide. The latter compound both enhances intestinal cell proliferation (lowers cancer cell apoptosis) and inhibits short-chain fatty acid metabolism (10). Diets high in protein enhance the exposure of colonocytes to hydrogen sulfide (26). Category II patterns can be broken down into 3 subcategories.
Bile acids that are partly deconjugated but without dehydroxylation of the bile salt nucleus. In species showing this pattern, the cecum is small or absent, and the intestinal residence time of those bile acids that escape reuptake at the terminal ileum is short. The nucleus and side chains of the bile acids look identical to those of gallbladder bile, with the exception that some of the bile acids have lost their taurine (or glycine) conjugates.
Bile salts are partly deconjugated and additionally show dehydroxylation. An example is shown in Figure 3, for the Snowy owl (Bubo scandiacus). Dehydroxylation (loss of the 7α-hydroxyl group on the bile salt nucleus) is a characteristic marker for the presence of an anaerobic cecum. In many species, dehydroxylation of trihydroxylated bile acids leads to the formation of the dihydroxy bile acid deoxycholic acid. The increase in the proportion of dihydroxy bile acids in the intestine may be important for maintaining the water balance between diarrhea and constipation because it is currently thought that dihydroxy bile acids act as endogenous laxatives (27–29). Because deoxycholic acid is a bacterial metabolite, its proportion in feces is quite low in neonates and slowly increases in juveniles. For example, in a study of 66 species of bovids, the proportion of deoxycholic acid in bile increased from nearly 0 at 20 d to a steady-state plateau at the age of 1 y (30). Adults of any species show steady-state levels of deoxycholic acid until reaching an advanced age, when the proportion of deoxycholic acid again starts to increase (31), likely due to a lower bile acid output and slowing of intestinal transit. External factors that affect the intestinal bacteria also affect the proportion of dehydroxylated bile acids. For example, humans consuming a probiotic drink (Lactobacillus paracasei) and calcium phosphate exhibited increased fecal dehydroxylated bile acids (32). Biliary bile often consists of mixtures of di- and trihydroxy bile acids, and the loss of the 7α-hydroxyl group from a dihydroxy bile acid can lead to the formation of monohydroxy lithocholic acid. Lithocholic acid shows unusual properties for a bile acid in that it is fairly insoluble in water, is readily absorbed by membranes, and forms complexes with lignins and complex polysaccharides. Because the rate of uptake by intestinal membranes is close to the rate of formation of lithocholic acid, what appears in feces is a number that instead reflects the volume and excretion rate of intestinal fiber solids. As a result, any event that affects intestinal mass and excretion rate will also affect the proportion of lithocholic acid appearing in feces. All of the lithocholic acid recovered in feces is formed in the large intestine, a dynamic place and time, where the drying out of the fecal mass “traps” both a proportion of lithocholic acid and the bacteria that form it. Because the proportion of lithocholic acid does not increase in fecal material outside the body, the remaining lithocholic acid is a reflection of a brief snapshot from the last moments of bacteria drying out in the large intestine and being exposed to the rapid temperature change of excreted feces.
Bile acids show partial deconjugation due to a mixture of bile acids that are either susceptible or resistant to bacterial deconjugation. Unlike in category Ib, only a portion of the bile salt structures are difficult for bacteria to enzymatically remove the taurine conjugate. The easy-to-deconjugate portion of the bile salt pool is deconjugated as in IIa; however, a pool of difficult-to-deconjugate bile acids remains in conjugated form. Deconjugation resistance is accomplished in a number of ways, some of which include a) the addition of a hydroxyl group to the α-carbon neighboring the taurine, providing steric hindrance at position 23 for C24 bile acids (33) and position 25 for C27 bile acids; b) the addition of a methyl functional group to the taurine itself to provide steric hindrance at the taurine (34, 35); c) the use of an 8-carbon side chain with a methyl group on the carbon next to the bile acid carboxylic acid providing steric hindrance on position 25 (as occurs on all C27 bile acids); and d) indirect protection of the terminal taurine by nuclear sulfation or glycosylation. Here, “enrichment” implies that conjugated bile acids are actively taken up by intestinal cells, conjugated with sulfate or glucose, and then resecreted into the intestine. Once protected by the additional polar conjugate, these bile acids then pass into the feces unchanged.
Although in category II, some bile acid conjugates remain, in category III, all the bile acids are completely deconjugated. Again, there are 3 subcategories based on the presence and amount of 7-dehydroxylation. For a survey of vertebrate intestinal morphology, including the presence and extent of an anaerobic cecum in different species, see Stevens (36).
Bile acids are completely deconjugated without 7-dehydroxylation. In these species, a cecum is absent, and the transit time in the intestine is sufficiently slow to allow all bile acids present to become completely deconjugated.
Bile acids are completely deconjugated, with partial or complete 7-dehydroxylation. An example is shown in Figure 4, for human feces. Nearly all of the m/z 407 trihydroxy C24 bile acid peak (now very small compared with the proportion found in biliary bile) has been shifted into the m/z 391 dihydroxylated C24 bile acid peak. In humans, this represents a cholic acid to deoxycholic acid shift. Correspondingly, nearly all of the original m/z 391 peak (as it appeared in bile) has been shifted to m/z 375 monohydroxylated C24 bile acids. In humans, this represents a chenodeoxycholic acid to lithocholic acid shift. A comprehensive list of the identities of bile acids identified in human feces is given in Setchell et al. (18).
Bile acids are completely deconjugated, but have 7-dehydroxylation–resistant structures. In a parallel to structurally conferred deconjugation resistance (listed in IIc), selected bile acid structures can interfere with the ability of bacterial enzymes to remove the 7α-hydroxyl group. Resistance can be achieved in a number of ways: a) insertion of a double bond between positions -6 and -7 on the steroid nucleus; b) the addition of a hydroxyl group to nearby position -6; and c) oxidizing the hydroxyl group at position -7 to a keto group.
This category contains species in which fecal bile acids are not observed. Many of these are hindgut fermenters, featuring a long residence time in an intensive bacterial environment. Pathways for the complete aerobic and anaerobic degradation of bile salts by bacteria have been proposed (37), yet it is still unclear exactly what has happened to the bile acids.
This category contains partly or completely deconjugated bile acids enriched in sulfates. An example is shown in Figure 5, for the domestic mouse (Mus musculus). Compared with biliary bile, it appears as though the C24 bile acids have exchanged taurine conjugation for sulfate conjugates. In Figure 5, the large peak at m/z 487 is a trihydroxy C24 bile acid conjugated with sulfate; this conjugate form is found only in trace proportions in biliary bile. The process of sulfate enrichment implies that deconjugated bile acids are passively absorbed by intestinal cells, sulfated (likely as part of a detoxification mechanism), and then resecreted back into the intestine. Intestinal cell metabolism of both bile acids and nutrients is often ignored because the intestine is far more than a simple thoroughfare for nutrients to pass through. Although bile acid sulfates are subject to some hydrolysis by bacteria during colonic transit, a significant portion survives intact in the feces (38, 39). These survivors are likely 7-sulfates because the 3-sulfates are more readily removed (40). Figure 5 also shows a minor peak at m/z 407 for deconjugation-resistant taurine conjugated C24 bile acids, a reflection of its 6-hydroxylated deconjugation-resistant bile acid structures, as discussed in the IIIc section. The dihydroxy bile acids chenodeoxycholic acid and deoxycholic acid act as endogenous osmosensors, inducing concentration-dependent water secretion when perfused into the human colon (28). This water-inducing ability is lost when the bile acids are sulfated (41). Among humans, selected children with constipation were found to have high proportions of 3-sulfated dihydroxy bile acids (pattern V instead of the normal pattern IIIb), and sulfation is known to abolish the secretory activity of these bile acids (42). One the other hand, commensal bacterial metabolism of dietary fiber releases the short-chain fatty acid butyrate, which, in turn, promotes the absorption of water from the intestine (43). Fiber supplementation and butyrate-producing pre- and probiotics in the diet have been shown to increase the amount of both fecal butyrate and butyrate-producing bacteria to promote water absorption and to exhibit antidiarrheal properties (44). In the domestic mouse, changing the diet by the addition of a probiotic (inulin) also appeared to promote intestinal sulfation, at the expense of dehydroxylation, as shown in Figure 6. The amount of deconjugation remained relatively complete in both the control and inulin-fed mice. In contrast, Figure 7 is a profile of feces from an IL-10 knockout mouse, commonly used as a model for intestinal inflammation. When comparing Figure 5 (control mouse) and Figure 7 (IL-10 knockout mouse), the most striking difference is the lack of intestinal sulfation (a missing pattern V) in the knockout, suggesting that this normal detoxification response is blunted in the IL-10 knockout. One possibility is that the intestinal epithelial cells, damaged by immune factors, are unable to metabolize bile salts normally. The inability to sulfate toxic monohydroxy C24 bile acids likely contributes to further damage to the intestinal epithelium.
Fecal bile acids are easy to sample and contain a considerable structural complexity that directly relates to intestinal morphology, their residence time in the intestine, and the species of microbial forms that they encounter. This proposed classification scheme has the potential to serve as an initial guide in interpreting the different bile salt patterns expressed in vertebrate feces and to shed light on the dark world of the intestinal environment.
We thank the Steve Sarro, Director of Animal Programs, National Aviary, Pittsburgh, PA, for collecting the samples of snowy owl feces, and Dr. Alan Hofmann for a critical reading of the manuscript. Both authors have read and approved the final version of this manuscript.