One of the biggest mistakes I think we can make in complex gut cases is assuming that a supplement will behave inside the human body exactly the way it behaves in a laboratory experiment. TUDCA is a good example. On paper, TUDCA looks highly attractive. It is a relatively hydrophilic bile acid that has been studied for effects involving bile flow, endoplasmic-reticulum stress, mitochondrial stability, apoptosis, and cellular protection. But there is another organism standing between the capsule and the human cell: the microbiome. Before I assume that somebody taking TUDCA is going to receive the effects attributed to intact TUDCA, I want to know what happens to the molecule as soon as it enters that person’s intestinal ecosystem. Which bacteria encounter it? Which enzymes act on it? Does it remain TUDCA? Is the taurine removed? What happens to that taurine? What happens to the remaining bile-acid molecule? Which secondary bile acids are ultimately produced? Which receptors do those bile acids activate? And can the intestinal epithelium and mitochondria tolerate the resulting metabolic load?
This became particularly important in one case I have been studying. Her stool bile-acid profile showed cholic acid, CA, at only 0.16% and chenodeoxycholic acid, CDCA, at only 1.22%, giving a combined primary bile-acid fraction of just 1.38%. In contrast, deoxycholic acid, DCA, was 52.16% and lithocholic acid, LCA, was 37.34%, giving a combined DCA + LCA fraction of 89.50%. Her LCA ratio was 0.72 and technically fell within the laboratory reference range, but this is exactly why I do not interpret ratios in isolation. A normal relationship between LCA and DCA does not mean the overall bile-acid ecosystem is normal. If nearly 90% of the measured pool consists of DCA and LCA while only 1.38% consists of CA and CDCA, the important question becomes: where did the primary bile acids go?
Most people think of bile as something the liver produces to digest fat. That is true, but incomplete. Bile acids are also signaling molecules. The liver begins by manufacturing primary bile acids, particularly CA and CDCA, and usually conjugates them to glycine or taurine before secretion. These conjugated bile acids enter the small intestine, help emulsify dietary fat, and are then largely reabsorbed and returned to the liver through the enterohepatic circulation. A smaller fraction escapes into the lower intestine, where bacterial metabolism becomes extremely important. One of the first major steps is often bile salt hydrolase, BSH. Certain intestinal bacteria possess BSH enzymes that remove glycine or taurine from conjugated bile acids. For example, conjugated CDCA → BSH activity → free CDCA. BSH does not directly convert CDCA into LCA. It opens the gate for what can happen next. Once CDCA has been deconjugated, another group of bacterial enzymes can perform 7α-dehydroxylation through the bai pathway, producing CDCA → LCA. The same principle applies to cholic acid: CA → DCA. The larger sequence therefore becomes conjugated CDCA → BSH deconjugation → free CDCA → bai-mediated 7α-dehydroxylation → LCA, while conjugated CA → BSH deconjugation → free CA → bai-mediated conversion → DCA.
When I see CA at 0.16%, CDCA at 1.22%, DCA at 52.16%, and LCA at 37.34%, I immediately ask whether an unusually active network of deconjugation followed by secondary bile-acid conversion is operating. In this case, I am particularly interested in the prominent Enterococcus cluster because some Enterococcus species possess BSH enzymes. That does not mean simply seeing Enterococcus proves that it is responsible for all of the bile-acid abnormalities. Taxonomy is not metabolic flux. The important question is whether the ecosystem as a whole is expressing enough BSH activity to create a large pool of deconjugated bile acids that downstream organisms can transform. I think of the pathway as a network: BSH opens the gate, and bai-active bacteria perform much of the secondary transformation.
Now add TUDCA. TUDCA stands for tauroursodeoxycholic acid, essentially UDCA conjugated to taurine. When bacterial BSH encounters TUDCA, it can potentially cleave the molecule into UDCA + taurine. The molecule we swallowed as TUDCA has now become two separate substrates, and the microbiome can potentially use both. The first is UDCA. The second is taurine. In somebody with sulfur sensitivity or abnormal sulfur metabolism, the taurine side deserves attention. Five hundred milligrams of TUDCA represents approximately 1 mmol. If completely deconjugated, it could theoretically release roughly 125 mg of taurine containing approximately 32 mg of sulfur. That does not mean 32 mg of sulfur automatically becomes hydrogen sulfide. Some taurine will be absorbed, some may remain luminal, transit matters, microbial composition matters, competition matters, and host metabolism matters. But it demonstrates an important principle: somebody can follow a very low-sulfur diet while simultaneously delivering a concentrated taurine conjugate directly into the gastrointestinal tract every day.
Once taurine is liberated, some intestinal organisms can metabolize it through pathways such as taurine → isethionate → sulfite → hydrogen sulfide, H₂S. I do not think about this as one organism causing everything. I think in terms of cross-feeding. One bacterium may deconjugate TUDCA, another may process taurine, another may generate hydrogen, another may reduce sulfur compounds, and mucin degraders may liberate additional sulfur-containing substrates. Organisms such as Enterococcus, Bilophila, Desulfovibrio, mucin degraders, and fermentative organisms may therefore function as parts of a metabolic consortium rather than isolated abnormalities. Fermentation also produces hydrogen, but hydrogen does not necessarily accumulate. Sulfur-reducing organisms can consume hydrogen and use it as reducing power. That means somebody can have substantial fermentation without necessarily showing massive hydrogen accumulation. This is why one question I repeatedly ask is: where did the hydrogen go?
Hydrogen sulfide itself is not inherently toxic. The body produces H₂S physiologically, and at appropriate concentrations it participates in normal signaling. Colonocytes also possess mitochondrial machinery designed to oxidize and clear sulfide, including SQOR, ETHE1, and downstream sulfur-oxidation pathways. The problem occurs when microbial production exceeds epithelial clearance capacity. At sufficiently high exposure, H₂S can inhibit cytochrome-c oxidase, Complex IV, suppress mitochondrial respiration, and interfere with butyrate oxidation. This matters because healthy colonocytes oxidize large amounts of butyrate and consume substantial oxygen in the process. That oxygen consumption helps maintain physiological epithelial hypoxia and limits oxygen leakage toward the intestinal lumen. This favors obligate anaerobes and helps prevent facultative respiratory organisms from gaining an energetic advantage. If excess H₂S impairs Complex IV and colonocyte respiration, oxygen consumption falls. More oxygen may remain available near the epithelial surface, while inflammatory signaling can simultaneously increase nitric oxide and nitrate availability. The model therefore becomes excess microbial H₂S → Complex IV stress → impaired butyrate oxidation → reduced colonocyte oxygen consumption → increased epithelial oxygen/nitrate availability → competitive advantage for facultative pathobionts. This is one reason repeatedly killing bacteria may not permanently change an ecosystem. If the metabolic environment still favors the same organisms, they have a reason to return.
The bile-acid side may reinforce the sulfur side. TUDCA itself does not directly become DCA. DCA primarily comes from CA → DCA, while LCA primarily comes from CDCA → LCA. But after TUDCA is deconjugated, it produces UDCA, and UDCA can enter additional microbial transformation pathways. Under appropriate conditions, microbial metabolism can move the UDCA skeleton toward LCA-related products. So I am not arguing that TUDCA directly created her DCA of 52.16%. I am arguing that TUDCA is entering an intestinal ecosystem already showing extraordinary capacity for deconjugation and secondary bile-acid transformation. The fate of TUDCA may therefore be very different from what we would predict by looking only at its direct effects on human cells.
The next question is why the primary bile-acid fraction is so low. Rapid microbial conversion is one possible explanation. If bacteria efficiently convert CA → DCA and CDCA → LCA, very little CA or CDCA may remain in stool. But there is another possibility I have been investigating: what if the liver is also receiving a signal telling it to manufacture less new primary bile acid? This brings us to FXR. FXR is a bile-acid-sensitive nuclear receptor, and one of its important locations is the ileum. When appropriate bile acids activate intestinal FXR, intestinal cells release FGF19. FGF19 travels through the portal circulation back to the liver and signals through the FGFR4/β-Klotho pathway, ultimately suppressing CYP7A1, the major rate-limiting enzyme of the classical bile-acid-synthesis pathway. The normal feedback loop is therefore bile-acid signaling → intestinal FXR activation → FGF19 rises → FGF19 reaches the liver → FGFR4 signaling → CYP7A1 falls → new bile-acid synthesis falls. This is normal physiology. The liver needs a feedback brake so it does not manufacture unlimited amounts of bile acids.
My question is whether, in certain abnormal microbial ecosystems, this normal feedback system could become part of the pathology. Imagine that BSH and bai activity are rapidly converting primary bile acids into secondary bile acids while at the same time the altered enterohepatic bile-acid environment is producing enough receptor signaling to tell the liver to reduce CYP7A1 and manufacture less new primary bile acid. Now the system could become trapped: less new primary bile acid is produced while the remaining primary bile acids are rapidly converted into secondary bile acids. The proposed loop becomes microbial bile-acid transformation → secondary-dominant enterohepatic environment → altered FXR–FGF19 feedback → CYP7A1 suppression → reduced new primary bile-acid synthesis → persistence of an overwhelmingly secondary-dominant pool.
I want to be very clear that stool bile-acid percentages alone cannot prove that this is happening. That is why I want the hypothesis to remain falsifiable. If I suspected excessive FXR–FGF19 feedback, I would want to look at FGF19 together with serum C4, 7α-hydroxy-4-cholesten-3-one, which reflects hepatic bile-acid synthesis, ideally alongside quantitative circulating bile-acid measurements. If FGF19 were relatively high while C4 were suppressed, that would support a strong feedback brake on hepatic bile-acid synthesis. If C4 were high, then the liver may actually be trying very hard to manufacture new bile acids, and the low stool CA/CDCA would be better explained by downstream microbial conversion, altered reabsorption, transit, or another mechanism. A useful biological hypothesis should be capable of being proven wrong.
TGR5 adds another layer, and I keep it separate from FXR because the two receptors are doing very different things. A simplified way to think about it is FXR → FGF19 → liver → CYP7A1 feedback, whereas TGR5 → cAMP signaling → enteroendocrine, enteric neural, secretory, vascular, and metabolic effects. DCA and particularly LCA are important endogenous TGR5 ligands. When I see a bile-acid profile containing more than 89% DCA and LCA among these measured fractions, I therefore do not only ask what is happening to hepatic bile-acid synthesis. I also ask what an unusually secondary-dominant environment may be doing to TGR5 signaling in different regions of the gastrointestinal tract.
This regional distinction is particularly important because I think one of the common mistakes in gastrointestinal medicine is treating “motility” as if the entire digestive tract moves at one speed. It does not. The stomach, small intestine, ileum, and colon have different neural circuits, different enteroendocrine cells, different receptor distributions, and different physiological jobs. The same signaling molecule can therefore create very different effects depending on where in the gastrointestinal tract the receptor is being activated.
In the distal intestine and colon, high DCA/LCA exposure can engage TGR5 on enteroendocrine cells, enteric neural pathways, and secretory systems. One plausible sequence is DCA/LCA-rich distal exposure → TGR5 activation → enteric neural and secretory signaling → increased colonic secretion and propulsion → loose stool or urgency. So a patient with excessive secondary bile acids can have very frequent or loose bowel movements. But this does not tell us what is happening upstream.
The stomach and small intestine can simultaneously behave very differently. TGR5 is also expressed within myenteric neural circuits, including inhibitory motor pathways. Experimental DCA/TGR5 signaling has been associated with recruitment of inhibitory enteric mechanisms involving nitric oxide. Nitric oxide is one of the major inhibitory neurotransmitters used by the gut to relax smooth muscle. Appropriate inhibitory signaling is essential for coordinated propulsion, because the intestinal segment ahead of a moving bolus needs to relax. But excessive or mistimed inhibitory signaling could theoretically reduce contractile output or disturb coordination.
The proximal model I am therefore interested in is secondary bile-acid exposure → TGR5 activation on inhibitory enteric pathways → cAMP/neural signaling → nitric-oxide-mediated relaxation → reduced or poorly coordinated proximal contractility → impaired gastric or small-intestinal clearance. Food and secretions can then remain in the upper gastrointestinal tract longer than expected. Microbes receive more time to ferment available substrates, producing bloating, pressure, distension, belching, fermentation, or early satiety.
At the same time, the same individual can have a very different TGR5 output farther downstream: DCA/LCA reaches the colon → TGR5-dependent neural and epithelial signaling → greater secretion and propulsion → loose stool or urgency.
This creates what I think is a very important regional model:
Upper GI/small intestine → impaired or poorly coordinated clearance → prolonged residence time → fermentation and distension
while simultaneously:
Colon → bile-acid/TGR5-driven secretion and propulsion → loose stools or urgency.
These findings are not contradictory. They may represent different regional outputs of the same abnormal bile-acid environment.
This is why loose stools do not prove that the small intestine has good motility.
Someone can have diarrhea several times per day and still have impaired gastric emptying, impaired fasting clearance, abnormal small-intestinal transit, or a weak migrating motor complex. The colon can be moving fluid rapidly while the upper gastrointestinal tract is clearing poorly. I think this distinction is particularly important in patients who present with the seemingly contradictory combination of severe postprandial bloating or distension together with predominantly loose stools.
TGR5-mediated enteroendocrine signaling creates another possible regional brake. TGR5 activation on intestinal L cells can influence GLP-1 and PYY release. These hormones do more than regulate glucose and appetite. They participate in the normal “ileal brake,” a feedback system through which nutrients and signaling molecules reaching the distal gut tell the upper gastrointestinal tract to slow delivery. Physiologically, this is useful because it prevents overwhelming the distal intestine. But in an abnormal bile-acid environment, I am interested in whether excessive or poorly timed TGR5 signaling could exaggerate this feedback. The proposed sequence would be secondary bile-acid exposure → TGR5 activation → GLP-1/PYY signaling → stronger distal brake → slower gastric or proximal intestinal delivery → greater fullness and residence time → more opportunity for fermentation, while the colon simultaneously remains secretory and loose.
The important point is therefore not that TGR5 universally “slows motility” or universally “causes diarrhea.” Both statements are too simplistic. The better question is: which TGR5-expressing cell is being activated, in which intestinal region, by which bile acid, at what concentration, and for how long? A receptor located on an inhibitory myenteric neuron may produce a different physiological outcome from the same receptor located on an enteroendocrine L cell or within a colonic secretory/propulsive circuit.
This regional framework may also explain why bowel frequency is such an unreliable marker of actual gastrointestinal clearance. A patient may report three or four loose bowel movements per day while still experiencing severe bloating, food sitting for hours, early satiety, fermentation, or recurrent SIBO. The distal colon may be clearing fluid rapidly, while the small intestine remains relatively stagnant. From an ecological perspective, it is the residence time in the small intestine that matters greatly for bacterial expansion, not simply how often stool leaves the colon.
I am therefore interested in a possible sequence such as secondary bile-acid dominance → abnormal regional TGR5 signaling → proximal inhibitory/brake-like physiology + distal secretory/propulsive physiology → impaired small-intestinal clearance despite loose stool → increased fermentation and microbial residence time → further bile-acid transformation. That could create yet another feedback loop within the larger model.
Hydrogen sulfide adds another complication because these signaling systems do not operate independently. H₂S can influence enteroendocrine, neuronal, mitochondrial, and smooth-muscle physiology. In an ecosystem where TUDCA is potentially supplying taurine while secondary bile acids are simultaneously stimulating TGR5, the final phenotype may therefore represent the combination of bile-acid receptor signaling + sulfur metabolism + mitochondrial capacity + regional motility, rather than one receptor being simply “on” or “off.”
Another part of the model I am investigating involves vascular physiology. TGR5 signaling can interact with pathways involving endothelial nitric oxide, NO. Nitric oxide relaxes blood vessels, so one hypothesis I am considering is whether unusually strong postprandial secondary-bile-acid signaling could contribute to excessive splanchnic vasodilation in a susceptible person. The proposed sequence is high secondary-bile-acid exposure → TGR5/cAMP signaling → eNOS/nitric-oxide signaling → mesenteric vasodilation → increased abdominal blood pooling → reduced effective central circulating volume → compensatory sympathetic activation → tachycardia, adrenaline, dizziness, weakness, or POTS-like symptoms. Stool bile-acid percentages cannot prove that this is occurring because fecal bile acids do not tell us the concentration of DCA or LCA reaching the portal or systemic circulation. But the pathway provides a plausible biological bridge between intestinal bile-acid metabolism and apparently neurological or cardiovascular post-meal symptoms and gives us another hypothesis that can be tested rather than assumed.
The mitochondrial side is equally important. DCA and LCA are not inherently pathological molecules; they are normal products of microbial bile-acid metabolism. The problem depends on dose, location, concentration, conjugation, duration of exposure, and host capacity. At sufficiently high local concentrations, hydrophobic unconjugated bile acids can become stressful to cell membranes and mitochondria. Experimental DCA exposure can increase oxidative stress and disturb mitochondrial membrane function. Calcium may amplify that injury because mitochondrial permeability transition is strongly influenced by the interaction between mitochondrial calcium burden, ROS, membrane potential, and redox state. The model I am exploring is hydrophobic bile-acid stress → membrane disturbance + ROS → abnormal intracellular Ca²⁺ handling → increased mitochondrial Ca²⁺ burden → lower threshold for mitochondrial permeability transition → loss of mitochondrial membrane potential → impaired oxidative phosphorylation → reduced ATP → epithelial dysfunction or injury.
This creates an important paradox. TUDCA itself may have mitochondria-protective properties while the microbial ecosystem receiving TUDCA may simultaneously be generating metabolites capable of mitochondrial injury. Both can be true because they involve different molecules, different concentrations, different compartments, and different stages of metabolism. That is precisely why evaluating a supplement only by its direct effect on cultured human cells can be misleading.
Once epithelial energy metabolism begins to fail, barrier integrity can deteriorate. Microbial products such as LPS may gain greater access to host immune signaling, activating TLR4 → NF-κB → inflammatory signaling → iNOS/nitric oxide/nitrate production. That nitrate can serve as a respiratory substrate for certain facultative organisms. Now another self-reinforcing loop appears: dysbiosis → barrier dysfunction → LPS exposure → inflammation → nitrate production → respiratory advantage for facultative organisms → persistent dysbiosis. At this point inflammation is no longer merely the consequence of the microbial ecosystem. Inflammation is helping construct the ecological environment that allows that ecosystem to persist.
Putting the model together, what I am investigating looks like this: BSH-active ecosystem → deconjugation of primary bile acids → free CA and CDCA → bai-mediated conversion → DCA and LCA accumulation. At the same time, TUDCA → BSH cleavage → UDCA + taurine → taurine enters sulfur cross-feeding → increased H₂S pressure → Complex IV and butyrate-oxidation stress → reduced colonocyte oxygen consumption → greater oxygen/nitrate availability → facultative pathobiont persistence. Potentially alongside this, secondary-bile-acid-dominant enterohepatic signaling → altered FXR–FGF19 feedback → CYP7A1 suppression → reduced new primary bile-acid synthesis, while DCA/LCA-rich signaling → region-specific TGR5 activation. Proximally, this could contribute to inhibitory neural signaling, GLP-1/PYY-mediated brake physiology, and impaired clearance; distally, it could increase secretion and propulsion. The patient could therefore experience bloating, fermentation, and impaired small-intestinal clearance at the same time as loose stools or urgency. TGR5-related vascular signaling could theoretically add postprandial splanchnic pooling and sympathetic compensation in susceptible individuals. Finally, high local hydrophobic bile-acid exposure → ROS + Ca²⁺ dysregulation → mitochondrial stress → reduced epithelial ATP production → barrier dysfunction → inflammation → further ecological disruption.
This is why I do not look at CDCA of 1.22% and simply say, “low primary bile acids.” I want to know why they are low. Is the liver failing to manufacture enough? Is CYP7A1 being suppressed by excessive feedback? Is BSH rapidly deconjugating what arrives? Are bai-active organisms converting CDCA into LCA almost immediately? Is CA being converted into DCA? Is ileal reabsorption abnormal? Is transit prolonging microbial exposure? Is regional TGR5 signaling altering proximal versus distal motility? Are loose stools misleading us into assuming small-intestinal clearance is normal? Are several of these mechanisms occurring simultaneously? The same laboratory pattern can potentially be produced by very different physiological failures, and the implications would be completely different.
That is also why I do not think TUDCA is simply “good” or “bad.” I see it as another substrate entering a living metabolic network. Its biological effect depends partly on what the host does with it and partly on what the microbiome does with it before the host ever sees the intact molecule. The questions I keep returning to are: which organism receives the molecule first, which enzyme acts on it, where does the taurine go, where does the sulfur go, where does the hydrogen go, which bile acid is produced next, which receptor sees that bile acid, where in the gastrointestinal tract is that receptor located, what happens to proximal clearance, what happens to distal secretion, what happens to vascular tone, what happens to mitochondrial calcium and respiration, and what signal ultimately reaches the liver?
That is what I mean by the TUDCA feedback trap. It is not one pathway. It is the possibility that several normally separate systems begin reinforcing one another: BSH-driven deconjugation → secondary bile-acid conversion → sulfur cross-feeding → H₂S pressure → mitochondrial dysfunction → loss of epithelial hypoxia → pathobiont persistence → altered FXR signaling → altered hepatic bile-acid synthesis → region-specific TGR5 signaling → proximal slowing or impaired coordination + distal secretion/propulsion → abnormal residence time → further microbial transformation → vascular and neural effects → further distortion of the bile-acid ecosystem. At that point, simply adding more bile support, killing one organism, restricting one nutrient, or looking at one stool ratio may miss the larger system.
The real question becomes: what is the host–microbial network doing with the molecule after we put it into the gut, and how does the answer change depending on which region of the gastrointestinal tract receives the resulting bile acids?