Although BA is the most common reason for pediatric liver transplantation, little is known about its pathophysiology. Recent years have seen the emergence of new pathomechanisms that have thrown light on environmental toxins as the isoflavonoid biliatresone [64, 78–80]. According to observation on sheep and studies done on zebrafish and BALB/c mice, biliatresone can cause EHBD-BA [80, 81]. In the present study, we document for the first time, that biliatresone has been shown to have an effect on C57BL/6J mice. This demonstrates that the mechanism of action of biliatresone is independent of the immunological background as found in BALB/c compared to C57BL/6J.
It is well known that biliatresone reduces GSH and downregulates the repair protein HSP90 [70, 73–75, 77, 82]. Depletion of GSH triggers an intracellular linear signaling cascade through upregulation of RhoU and Hey2, resulting in the downregulation of the transcription factor SOX17. This cascade is highly conserved and has been documented in zebrafish and BALB/c mice as well as various cholangiocyte cultures [73, 82–87]. Therefore, it is suggestive, that our results were based on the described effect of biliatresone, mirroring the influence on the same intracellular signaling cascades.
In this study, the artificial biliatresone-induced BA model was successfully introduced in C57BL/6J mice and replicated human EHBD-BA. Due to early euthanasia and investigation of jaundiced animals we were able to evaluate very early onset of the BA, which presented similar to a choledochal cyst [4, 5].
When biliatresone was injected into C57BL/6J mice, we noticed a window of susceptibility consisting of time and dose (Fig. 3A). This time dependent outcome has also been reported according to a prior investigation employing BALB/c mice. Those injected with biliatresone 72 hours after birth did not exhibit clinical indications of BA [66]. These results are consistent with the RRV model and the dose- and time dependent morphological disruption of the EHBD in zebrafish [2, 88–90]. Hereby, the virus kind, virus quantity and injection time point have been reported as being critical in the RRV model for the induction of experimental virus-mediated BA similar phenotype [88, 90, 91]. The preferred murine RRV model was not found to accurately simulate BA in all its clinical diversity [89]. It remains questionable whether viral induced BA models are suitable for translational research in BA [92]. In addition immune dependency after KP is questioned by clinical data showing non-significant differences after intravenous immunoglobulin (IVIG) treatment [93].
In the biliatresone-induced BA model, we found that the prognosis of artificially introduced postnatal EHBD-BA was dosage, time and weight dependent (Figs. 1 and 3A). The observed time-dependent efficacy reinforces the reports of neonatal susceptibility in organoids and mice [94]. Since BA only occurs in newborns, it can be assumed that the susceptibility in humans is developmental and exposure to a toxin, such as biliatresone leads to the outbreak.
Jaundice, growth retardation, clay-colored feces and stained urine were all evident in this model (Fig. 2A and B). Additionally, serum studies showed distinct variations in liver- and bile-alteration between the control group and the mice given 70 µg of biliatresone. Cholestatic milieu causes liver damage as seen by significantly higher levels of AP, γ-GT and GLDH (Fig. 4A-C) [95–98]. A higher level of total bilirubin suggests a reduction of liver's metabolic function (Fig. 4D) [99, 100].
Gross morphology of biliatresone treated murine neonates revealed a hydropic gall bladder and an enlarged, twisted EHBD indicating EHBD-BA (Fig. 5A and B). By histological examination of EHBD stained with HE, the extrahepatic obliteration was confirmed. The presence of a liver portal extension strengthened the case for cholestasis.
Since biliatresone has only been evaluated in vivo in BALB/c mouse strains so far, the discovery that we were able to show BA-like characteristics in C57BL/6J strain is quite intriguing. It is well known that BALB/c and C57BL/6J mice differ in their immune characteristics for the defense against infections. Comparing these two mice strains, different results employing experimental BA-models have been observed. In a comparative study obtaining the RRV model only 5 of the 37 (13.5%) C57BL/6J mice were affected by symptoms of BA after receiving RRV injection. According to reports, this strain is almost resistant. On the other hand, 37 out of 55 BALB/c mice (67%) developed experimental BA. The BALB/c mice were said to be extremely vulnerable [90]. In further studies obtaining the RRV model up to 86% of BALB/c mice developed experimental BA after RRV injection [88, 101, 102].
Strain-related different outcomes have been reported likewise in experimental mice models for leishmaniosis, helminths, melanoma metastases, tuberculosis, Pasteurella pneumotropica and Chlamydia [103–110]. BALB/c mice exhibit a stronger Th2- and M2-dominant immune response than C57BL/6 mice displaying a Th1- and M1-dominant immunological response [111, 112]. Varying outcome between different strains strongly indicates important differences in the immune system function linked to variations in M1/M2 [113]. However, these variations are still not fully understood.
BALB/c mice treated with biliatresone developed jaundice in 60% including death with jaundice 20% [66, 75]. The overall jaundice rate in biliatresone-treated C57BL/6J mice in this study was 42.1% (n = 8). Treatment without jaundice led to a mortality rate of 28.9% in BALB/c mice as opposed to 21.0% (n = 4) in C57BL/6J mice. This cross-strain comparison shows higher similarity than the percentual results in RRV studies. Both strains respond likewise to biliatresone treatment, resulting in the same BA-like phenotype, which is linked to the previously discussed molecular pathways of GSH depletion and SOX17-reduction leading to bile duct obstruction [74–77, 83].
C57BL/6 mice are described to be highly resistant towards infectious models compared to BALB/c mice. In the study of Yang et al. the total mortality of neonates treated with 80 µg biliatresone was 48.9%, which is compared to a mortality of 11.0% in C57BL/6J mice treated with 70 µg biliatresone much higher, but similar to the mortality in our pretest 75% (n = 3) in neonates treated with 80 µg biliatresone [66]. The described resistance of C57BL/6J mice as reported in various infection models including RRV, helminths and leishmaniosis could not be observed. This indicates that C57BL/6J mice are not resistant to biliatresone and the onset of the biliatresone-induced BA model is not mediated by genotypical immunological differences.
The comparison between BALB/c and C57BL/6J mice regarding the evaluation of biliatresone treatment as a potential toxin in humans is not completed yet. Study designs should include treatment under equal conditions. Euthanasia should be performed on determined days in both strains to evaluate BA phenotype at the same stages of development. Yang et al. euthanized their animals on day 18 extending the phenotype profile after exhibiting clinical symptoms, which separates the 60% of mice with BA in death 20%, recovery 22.2% and jaundice for more than two weeks 17.8%. Further biliatresone-induced BA-studies aiming for the observation of the impact of the immune system should include immune cell KO-models to evaluate the hypothesis of immune independent BA-induction.
The dose of 60 µg biliatresone, as described in our pretest, did not result in clinical symptoms of BA and 100% survival (n = 3) (Fig. 1). If the amount remains below an effective range, toxic dose or viral load has been demonstrated to be ineffective. Therefore, it is reasonable to expect that rodents and humans will always be exposed to low dosage hazardous precursors. As long as the effect is not increased by the immature organism's susceptibility different doses of such precursors may not necessarily alter development of healthy newborns and they do not go on to acquire BA. The susceptibility is strongly associated to dose, time and weight during exposure as demonstrated in this study.
It is currently unknown, yet which internal or external factor is most likely to cause postnatally induced EHBD-BA in newborn infants [8]. A toxic metabolic end product might arise due to a change in the metabolism depending on dysbiosis combined with disease promoting dietary intake. The human biliary and intestinal microbiota is a very complex network and part of recent discussion to be involved in progression of several diseases [114–124]. Retracing digestive potential of the microbiota it was observed that Rhizoctonia solani inoculated sugar beet roots take on a structure resembling biliatresone and that the bacterium Clostridium can break down a soybean isoflavone into biliatresone [125–127]. On the other hand plants use phytoalexins, such as the isoflavonoids for the defense against microbiota. These are linked to several promoting and affecting disease progressions in humans and might have toxic effects after ingestion, which may be targeted with future therapeutics [128]. A fetal exposure to the toxin may be an explanation and a neonatal exposure may happen while nursing if biliatresone migrates through the placenta. Additionally, the newborn's intestine may already be dysbiotically populated and is not yet colonized by intestinal gut bacteria. When hazardous precursors like isoflavonoids are consumed along with a dysbiosis, the sensitive organism's normal growth may be altered.
The demonstration of similar outcome between BALB/c and C57BL/6J challenges the conventional understanding that the immune system plays a major role in the pathophysiology of BA and supports the theory that toxic alteration influences the pathogenesis. NAC-treatment might be a promising tool to encounter toxic GSH-reduction and mitigate liver damage [71–74, 129].