Intestinal FXR inhibition leading to steatohepatitis via gut‑liver axis dysregulation

• Jung-Hwa Oh

Revision dates for related pages

This AOP is primarily applicable to mammalian species in which the bile acid–FXR–FGF15/FGF19 signalling axis and gut–liver anatomical/physiological organisation are conserved.

• Taxa

  • Human (Homo sapiens): Intestinal FXR–FGF19 signalling, enterohepatic bile acid circulation, and gut–liver axis dysfunction have all been documented in human steatotic liver disease and steatohepatitis. Therefore, the AOP is directly relevant to human health risk assessment.

  • Rodents (especially mouse, Mus musculus): Most mechanistic evidence (intestinal and hepatic Fxr knockout models, diet‑induced steatohepatitis, FXR agonist/antagonist studies) comes from mice, where FGF15 is the functional orthologue of human FGF19. The qualitative sequence of key events is considered conserved between mouse and human, although quantitative aspects may differ.

  • Application to other mammals (e.g. rat) is plausible where a comparable FXR–FGF15/FGF19–bile acid axis exists, but specific evidence may be more limited.

• Sex Available data do not indicate a fundamentally different FXR–FGF15/FGF19–gut‑liver mechanism between males and females. Both sexes develop steatosis and steatohepatitis in relevant models, and both exhibit FXR–FGF19 dysregulation in clinical and experimental settings. The AOP is therefore considered applicable to both males and females, while acknowledging that quantitative susceptibility (e.g. threshold, rate of progression) may differ between sexes in specific models.

• Life stage The AOP is best supported for late juvenile and adult stages, in which:

  • hepatic metabolic capacity is fully developed,

  • enterohepatic circulation of bile acids is established, and

  • diet‑ and obesity‑related metabolic stress can realistically contribute to steatosis and steatohepatitis. Evidence in early developmental stages (neonatal/infant) is more limited, and the role of intestinal FXR and bile acid signalling may differ due to developmental changes in bile acid metabolism and microbiota. Thus, applicability to very early life stages should be considered with caution and may require additional data.

• Other biological context

  • The AOP assumes a functional gut–liver axis with intact portal circulation and enterohepatic bile acid cycling; conditions that fundamentally alter this (e.g. major surgical shunts, absence of bile flow) may fall outside the domain of straightforward applicability.

  • The AOP is most relevant under metabolic stress conditions (over‑nutrition, obesity, insulin resistance) that favour hepatic fat accumulation, since these provide the context in which intestinal FXR perturbation and gut–liver axis dysregulation are most likely to drive progression to steatohepatitis.

Overall, the essentiality of the key events (KEs) in this AOP is supported by a combination of direct intervention studies (genetic and pharmacological manipulation of FXR and its core targets) and indirect correlation data (degree of KE modulation vs severity of downstream KEs and steatohepatitis). Essentiality is strongest for KEs directly controlled by FXR (FGF15/FGF19, hepatic SHP/CYP7A1, BSEP) and somewhat weaker, though still supported, for more distal tissue‑level events (intestinal barrier dysfunction, oxidative stress, inflammation).

MIE – Intestinal FXR, inhibition

Essentiality: High (direct evidence)

• Intestinal FXR–deficient mice (e.g. FxrΔIE) show reduced ileal FGF15/FGF19 expression, increased hepatic bile acid synthesis, exacerbated steatosis and steatohepatitis compared with wild‑type animals under the same dietary challenge.

• Conversely, pharmacological activation of intestinal FXR with gut‑restricted agonists (e.g. fexaramine, or orally administered obeticholic acid with strong intestinal action) increases FGF15/FGF19, normalises bile acid synthesis and attenuates hepatic steatosis and inflammation.

• These gain‑ and loss‑of‑function data indicate that preventing intestinal FXR activation (the MIE) reliably worsens, and enhancing it improves, multiple downstream KEs and the AO.

KE1 – Ileal FGF15/FGF19 secretion, decreased

Essentiality: High (direct evidence)

• FGF15/FGF19 is a direct FXR target in the ileum, and its secretion is necessary to repress hepatic CYP7A1 and maintain bile acid homeostasis. Blocking FGF15/FGF19 (e.g. in FGF15‑deficient mice, or through reduced FXR activation) leads to increased CYP7A1 activity, elevated bile acid synthesis, and aggravated hepatic injury under metabolic or cholestatic stress.

• Restoration of FGF19 (by exogenous administration or FXR agonists) reduces bile acid synthesis, improves liver histology and lowers transaminases in animal models and in clinical studies, demonstrating that correcting KE1 can reverse or attenuate several downstream KEs (altered bile acid synthesis, steatosis, inflammation) and improve the AO.

• The tight causal linkage between FGF15/FGF19 levels and hepatic bile acid synthesis provides strong evidence for essentiality.

KE2 – Ileal bile acid transporters (IBABP, OSTα/β), decreased

Essentiality: Moderate (mixed direct/indirect evidence)

• IBABP and OSTα/β are FXR‑regulated transporters that buffer and export bile acids in ileal enterocytes. Genetic or pharmacological reduction of these transporters disrupts enterohepatic bile acid circulation and can lead to local bile acid accumulation and epithelial stress.

• Experimental data show that knockdown/knockout of OSTα/β or IBABP alters bile acid distribution and can contribute to intestinal injury and barrier defects, which in turn favour gut–liver axis inflammation. However, the extent to which modulation of these individual transporters alone is sufficient to drive the full sequence of downstream KEs up to steatohepatitis is less well quantified.

• Overall, there is reasonable evidence that attenuation or restoration of these transporters modulates downstream events, but essentiality is considered moderate rather than high.

KE3 – Intestinal barrier integrity, decreased

Essentiality: Moderate (strong indirect evidence)

• Multiple models show that increased intestinal permeability and loss of barrier integrity are associated with higher portal LPS levels, Kupffer cell activation and worsening hepatic inflammation and fibrosis. Interventions that preserve or restore barrier function (e.g. FXR agonists, probiotics, prebiotics, SCFAs) often attenuate hepatic inflammation and histological liver damage.

• However, steatohepatitis can also develop via mechanisms that do not primarily rely on barrier breakdown (e.g. pure metabolic/lipotoxic pathways), and barrier damage is not unique to FXR‑mediated toxicity.

• Therefore, barrier dysfunction is clearly an important amplifier and modulator of downstream inflammation and the AO, but it may not be strictly essential in all contexts; essentiality is judged as moderate.

KE4 – Hepatic SHP, decreased; CYP7A1/CYP8B1, derepressed

Essentiality: High (direct evidence)

• Hepatic FXR–SHP signalling is a central negative regulator of CYP7A1 and CYP8B1. In hepatic Fxr or Shp knockout mice, CYP7A1 is derepressed, bile acid synthesis is increased and liver injury is exacerbated under cholestatic or metabolic stress; restoring or enhancing this pathway reduces bile acid synthesis and hepatic damage.

• Pharmacological activation of FXR in the liver increases SHP, represses CYP7A1 and improves bile acid homeostasis and liver histology. Blocking this KE (e.g. by SHP loss or sustained CYP7A1 overexpression) prevents FXR‑FGF19‑mediated feedback from fully normalising bile acid synthesis and liver injury.

• These data show that this KE is essential for transmitting signals from intestinal FXR/FGF15/FGF19 to hepatic bile acid synthesis and downstream events.

KE5 – Hepatic BSEP (ABCB11) and MRP2 (ABCC2), decreased

Essentiality: High (direct evidence for bile acid–mediated toxicity)

• BSEP and MRP2 are key canalicular transporters for bile acid and organic anion excretion. Genetic deficiency or functional inhibition of BSEP in humans and animals causes intrahepatic cholestasis, bile acid retention, hepatocellular injury and progression to steatosis, inflammation and fibrosis.

• Experimental models show that FXR activation increases BSEP/MRP2 expression and improves bile acid excretion and liver histology, whereas FXR suppression or transporter inhibition worsens intrahepatic bile acid accumulation and liver injury.

• Since bile acid retention is a major driver of hepatocellular stress and inflammation, maintenance of this KE is critical, and its disruption is strongly linked to downstream KEs and the AO.

KE6 – Hepatic lipid accumulation (steatosis), increased

Essentiality: High for steatohepatitis as defined

• Steatohepatitis, by definition, requires the presence of hepatic steatosis along with inflammation and cell injury. Genetic or dietary interventions that reduce steatosis (e.g. enhancing β‑oxidation, inhibiting de novo lipogenesis, or correcting insulin resistance) generally decrease the severity of inflammation and histological steatohepatitis, even when other insults are present.

• Conversely, models with severe steatosis promote oxidative stress, inflammatory signalling and progression to steatohepatitis, particularly in the presence of additional hits such as bile acid toxicity or endotoxemia.

• While limited inflammatory changes can occasionally occur without marked steatosis, the overall body of evidence supports hepatic steatosis as an essential prerequisite for the AO in the context of this AOP.

KE7 – Hepatic oxidative and endoplasmic reticulum stress, increased

Essentiality: Moderate (good indirect evidence)

• Numerous studies show that reducing oxidative/ER stress (e.g. via antioxidants, ER stress inhibitors, genetic manipulation of stress pathways) mitigates liver injury, inflammation and fibrosis in steatohepatitis models, suggesting a causal contribution of this KE.

• However, oxidative and ER stress are broadly involved in many forms of liver injury and can be both cause and consequence of other KEs (steatosis, inflammation). Complete elimination of oxidative/ER stress is rarely achieved experimentally, and some degree of inflammation can occur even with partially reduced stress.

• Therefore, this KE is a key amplifier and mediator but may not be strictly essential in all contexts; essentiality is rated as moderate.

KE8 – Hepatic inflammation, increased

Essentiality: High (direct evidence relative to AO definition)

• Steatohepatitis is defined histologically by the coexistence of steatosis and lobular inflammation with hepatocyte injury. Interventions that specifically reduce hepatic inflammation (e.g. NF‑κB inhibitors, anti‑TNF/IL‑1β strategies, macrophage‑targeted treatments) consistently attenuate the severity of steatohepatitis and slow or prevent progression to fibrosis, even when some degree of steatosis persists.

• Conversely, genetic or pharmacological manipulations that enhance inflammatory signalling in the liver exacerbate steatohepatitis.

• Given that inflammation is a defining component of the AO itself, this KE is essential for the expression of steatohepatitis as the endpoint of this AOP.

Summary

Taken together, essentiality is strongest for KEs directly under FXR control (intestinal FXR, ileal FGF15/FGF19, hepatic SHP/CYP7A1 and BSEP) and for the defining hepatic manifestations of the AO (steatosis and inflammation). Intestinal barrier dysfunction and oxidative/ER stress are well‑supported contributors and amplifiers, but may not be strictly required in every possible context. Nonetheless, under the conditions for which this AOP is intended (metabolic stress plus gut–liver axis perturbation), modulation of each KE produces qualitatively consistent changes in downstream events and in the incidence and severity of steatohepatitis, providing an overall moderate‑to‑high level of support for essentiality across the pathway.

1) KER1 – Intestinal FXR inhibition → Ileal FGF15/FGF19 secretion, decreased

Biological plausibility Intestinal FXR is highly expressed in ileal enterocytes and directly controls FGF15/FGF19 transcription through FXR response elements in the FGF15/FGF19 promoter. This FXR–FGF15/FGF19 axis is a central component of bile acid endocrine feedback and is well established in both rodents (FGF15) and humans (FGF19). Loss of FXR activation is therefore expected to reduce FGF15/FGF19 expression and secretion.

Empirical support Multiple in vitro, animal and human studies show that bile acid or FXR agonist exposure robustly induces ileal FGF15/FGF19, whereas FXR inhibition, bile acid depletion, or FXR knockout markedly decrease it. Intestinal FXR knockout mice display reduced FGF15 and increased hepatic bile acid synthesis; FXR agonists or FGF19 analogues restore FGF19 levels and suppress bile acid synthesis. Overall, empirical support is strong and consistent across models.

Quantitative understanding Quantitative relationships between FXR activation and FGF19 output have been described in cell models, ileal explants and human studies (e.g. FGF19 versus C4). However, explicit dose–response models linking defined degrees of FXR inhibition to FGF19 reduction across species are still limited; quantitative understanding is moderate.

2) KER2 – Intestinal FXR inhibition → Ileal bile acid transporters (IBABP, OSTα/β), decreased

Biological plausibility FXR directly regulates the expression of ileal BA transporters, including IBABP (FABP6) and OSTα/OSTβ (SLC51A/B), which are critical for intracellular BA buffering and basolateral export. Reduced FXR activity is expected to decrease transcription of these genes and impair ileal handling of bile acids.

Empirical support FXR agonists increase IBABP and OSTα/β expression in vitro and in vivo, whereas FXR deficiency or inhibition reduces their expression. In FXR‑deficient or BA‑deprived models, intestinal BA transport is altered and BA distribution shifts, consistent with changes in transporter expression. Nevertheless, fewer studies directly quantify transporter changes in the specific context of steatohepatitis, so empirical support is moderate.

Quantitative understanding Dose–response relationships between FXR activity and transporter expression have been demonstrated in cell models, but systematic quantitative data across species and time are limited. Quantitative understanding is low‑to‑moderate.

3) KER3 – Ileal bile acid transporters, decreased → Intestinal barrier integrity, decreased

Biological plausibility Loss of BA buffering and export (IBABP, OSTα/β) can increase intracellular and luminal accumulation of cytotoxic bile acids, which damage epithelial cells and tight junctions. FXR also contributes more broadly to barrier maintenance; reduced transporter expression therefore plausibly contributes to barrier dysfunction.

Empirical support Several studies show that altered BA handling and FXR suppression are associated with decreased tight junction protein expression, increased intestinal permeability and enhanced bacterial translocation. Interventions that improve FXR signalling or modulate the BA pool can restore barrier function. However, most studies manipulate FXR or BA pools globally rather than isolating transporter changes, so causal separation of “transporter effect” from other FXR‑mediated effects is incomplete. Empirical support is moderate.

Quantitative understanding Quantitative links between defined levels of transporter reduction and specific changes in permeability (e.g. TEER, FITC‑dextran flux) are sparse. Barrier disruption is typically reported qualitatively or semi‑quantitatively. Quantitative understanding is low.

4) KER4 – Ileal FGF15/FGF19 secretion, decreased → Hepatic SHP decreased; CYP7A1/CYP8B1 derepressed

Biological plausibility FGF15/FGF19 is a key endocrine mediator of negative feedback from intestine to liver. By activating FGFR4/β‑Klotho in hepatocytes, it induces SHP and represses CYP7A1/CYP8B1, thereby reducing bile acid synthesis. Reduced FGF15/FGF19 is therefore expected to diminish SHP signalling and derepress CYP7A1/CYP8B1.

Empirical support FGF15/FGF19 deficiency or impaired FGFR4/β‑Klotho signalling leads to elevated CYP7A1 expression and increased bile acid synthesis in rodents and humans. Exogenous FGF19 or FXR agonists that restore FGF19 decrease CYP7A1 and BA synthesis. Correlations between lower FGF19 and higher C4 (a surrogate of CYP7A1 activity) in humans further support this relationship. Empirical support is strong.

Quantitative understanding There is good quantitative correlation between serum FGF19 and C4 in humans, and dose–response data for FGF19 analogues exist in clinical and preclinical studies. However, complete quantitative models spanning from intestinal FXR perturbation to hepatic CYP7A1 output are not yet fully developed. Quantitative understanding is moderate.

5) KER5 – Hepatic SHP,decreased; CYP7A1/CYP8B1,increased → Hepatic BSEP/MRP2 decreased

Biological plausibility Hepatic FXR–SHP signalling coordinates BA synthesis and export. Reduced FXR/SHP activity that derepresses CYP7A1/CYP8B1 and increases BA synthesis tends to be accompanied by decreased expression of canalicular BA transporters (BSEP, MRP2), both of which are FXR targets. Thus, sustained reduction in FXR signalling is expected to decrease BSEP/MRP2 in parallel with increased BA synthesis, favouring intrahepatic BA retention.

Empirical support FXR or SHP knockout models show both increased CYP7A1 and decreased BSEP expression, associated with BA retention and liver injury. FXR agonists increase BSEP/MRP2 expression and improve cholestatic phenotypes. However, not all studies directly link CYP7A1 increases with BSEP/MRP2 changes in a time‑resolved manner, and some compensatory responses occur. Empirical support is moderate‑to‑strong.

Quantitative understanding Quantitative transcriptional and protein changes in BSEP/MRP2 after FXR modulation are available, but integrated quantitative models linking specific CYP7A1 increases to transporter changes and BA retention are limited. Quantitative understanding is low‑to‑moderate.

6) KER6 – Hepatic BSEP/MRP2 decreased → Hepatic lipid accumulation (steatosis), increased

Biological plausibility Reduced canalicular export of bile acids leads to intrahepatic BA accumulation, which disrupts hepatocyte metabolism, mitochondrial function and lipid handling. BA‑mediated stress can promote steatosis by impairing β‑oxidation and enhancing lipogenic signalling. Thus, loss of BSEP/MRP2 can indirectly favour steatosis in a susceptible metabolic background.

Empirical support Human and animal models with BSEP deficiency or inhibition show BA retention, cholestasis and frequently co‑existing steatosis. FXR agonists that restore BSEP function often reduce both cholestatic markers and hepatic lipid accumulation. However, steatosis can arise through multiple mechanisms, and BSEP/MRP2 reduction is not universally required. Empirical support is moderate.

Quantitative understanding While BA levels, transporter expression and steatosis scores are often reported together, formal dose–response relationships between transporter inhibition and degree of steatosis are not well characterised. Quantitative understanding is low.

7) KER7 – Hepatic lipid accumulation increased → Hepatic oxidative and ER stress increased

Biological plausibility Hepatic steatosis increases substrate flux through mitochondrial β‑oxidation and ER lipid handling, promoting ROS generation, lipid peroxidation and unfolded protein responses. Excess lipid load also leads to lipotoxic species that trigger ER stress and mitochondrial dysfunction. Thus, increased steatosis is expected to increase oxidative and ER stress.

Empirical support Numerous models show that high‑fat diets or genetic manipulations causing steatosis are accompanied by increased markers of oxidative stress (e.g., MDA, 4‑HNE, 8‑OHdG) and ER stress (e.g., GRP78, CHOP). Interventions that reduce steatosis (e.g., improved insulin sensitivity, enhanced β‑oxidation) typically lower these stress markers, while forced lipid accumulation intensifies them. Empirical support is strong.

Quantitative understanding Semi‑quantitative relationships (e.g., correlation between liver triglyceride content and oxidative stress markers) are frequently reported. Fully quantitative, mechanistic models exist in some systems biology studies but are not yet standard for this specific AOP context. Quantitative understanding is moderate.

8) KER8 – Hepatic oxidative/ER stress increased → Hepatic inflammation increased

Biological plausibility Oxidative and ER stress induce DAMP release, activate pattern‑recognition receptors, inflammasomes and stress‑activated kinases and converge on NF‑κB and AP‑1 activation. These pathways drive pro‑inflammatory cytokine and chemokine expression, recruiting and activating Kupffer cells and infiltrating leukocytes. Thus, increased oxidative/ER stress is expected to promote hepatic inflammation.

Empirical support In steatohepatitis models, oxidative/ER stress markers and inflammatory markers increase in parallel. Genetic or pharmacological attenuation of oxidative/ER stress (e.g., antioxidants, chemical chaperones, inhibition of specific stress pathways) reduces hepatic inflammation and histological scores. Conversely, enhancing oxidative stress or ER stress increases inflammatory responses. Empirical support is strong.

Quantitative understanding Correlations between the magnitude of oxidative/ER stress and inflammatory readouts are documented, but formal dose–response models and temporal mapping at the whole‑pathway level remain limited. Quantitative understanding is moderate.

9) KER9 – Intestinal barrier integrity decreased → Hepatic inflammation increased

Biological plausibility Loss of intestinal barrier integrity increases translocation of LPS and other microbial products into portal blood, where they activate TLRs and other receptors on Kupffer cells and liver sinusoidal cells. This drives NF‑κB activation and production of pro‑inflammatory mediators, contributing to hepatic inflammation. This gut–liver axis mechanism is well recognised.

Empirical support Models with increased intestinal permeability (due to diet, FXR deficiency, microbiota imbalance or chemical insult) show elevated portal LPS, increased hepatic TLR signalling and enhanced liver inflammation. Interventions that restore barrier function (e.g., FXR agonists, probiotics, barrier‑strengthening agents) reduce LPS levels and hepatic inflammatory markers. Empirical support is strong, though barrier damage is not the only route to liver inflammation.

Quantitative understanding Some quantitative data relate permeability measures (e.g., FITC‑dextran flux, zonulin levels) to portal LPS concentrations and inflammatory markers, but there is no standardised quantitative model across species. Quantitative understanding is low‑to‑moderate.

10) KER10 – Hepatic inflammation increased → Steatohepatitis

Biological plausibility Steatohepatitis, by definition, requires steatosis plus lobular inflammation and hepatocyte injury (ballooning). Sustained hepatic inflammation drives cell death, ballooning and fibrogenic activation, converting simple steatosis into steatohepatitis and promoting progression to fibrosis.

Empirical support In animal and human steatohepatitis, inflammatory cell infiltration and cytokine production correlate with histological steatohepatitis severity. Anti‑inflammatory interventions (e.g., NF‑κB or cytokine targeting) reduce lobular inflammation and ballooning, attenuating steatohepatitis even when some steatosis persists. Conversely, enhancing inflammatory signalling exacerbates steatohepatitis. Empirical support is strong.

Quantitative understanding Histological grading systems (e.g., NAS) incorporate inflammatory lesions; quantitative relationships exist between inflammatory scores and progression to steatohepatitis and fibrosis. However, fully quantitative mechanistic models linking specific inflammation magnitudes to AO incidence across populations are still emerging. Quantitative understanding is moderate.

For this AOP, the overall quantitative understanding is semi‑quantitative and heterogeneous across KERs. Proximal FXR‑dependent steps are best characterised: dose–response relationships between bile acid/FXR agonist exposure and ileal FGF15/FGF19 expression and secretion are available from in vitro models, ileal explants and human studies, and there is a robust inverse relationship between circulating FGF19 and markers of hepatic bile acid synthesis (e.g., C4), supporting a reasonably well‑defined quantitative link between the MIE, KE1 and KE4. Likewise, multiple studies provide semi‑quantitative associations between the magnitude of hepatic steatosis (KE6) and markers of oxidative/ER stress (KE7) and inflammation (KE8), as well as between histological activity scores and the severity of steatohepatitis (AO). However, for several intermediate KERs—particularly those involving ileal bile acid transporters and intestinal barrier integrity, and the transition from altered bile acid export (KE5) to steatosis (KE6)—data are largely correlative or categorical, and explicit dose–response and time‑course functions are sparse. Species differences in bile acid composition and FXR ligand potency, and the influence of modulating factors such as metabolic status and microbiota, further complicate quantitative extrapolation to humans. Overall, while the direction and relative sensitivity of most KERs are supported by multiple datasets, fully parameterised response–response functions suitable for formal quantitative AOP (qAOP) modelling exist only for selected parts of the pathway (e.g., FXR–FGF19–CYP7A1), and the AOP is currently best used in a qualitative or semi‑quantitative manner rather than for precise numerical prediction of steatohepatitis incidence.