Two novel intronic variants in ABCB4: Clinical features, molecular mechanisms, and literature review

Abstract

BACKGROUND

Multidrug resistance protein 3 (MDR3) is expressed in the capillary duct membranes of liver cells and functions in the transport of phospholipids. ABCB4 variants can cause defects in the expression and function of MDR3. Intronic variants of ABCB4 may cause the aberrant deletion or retention of introns in the open reading frame, thereby affecting the normal expression of MDR3. Although numerous studies on the mechanisms of ABCB4 missense variants have been conducted, research on intronic variants remains scarce.

AIM

To explore the pathogenicity of two newly discovered intronic variants of ABCB4.

METHODS

In 2018-2021, two patients with two novel ABCB4 intronic variants were enrolled and their clinical characteristics were analyzed. The Hospital’s Ethical Review Board approved the study and informed consent was obtained. The pathogenicity of the intronic variants was predicted and analyzed in silico. Minigene analysis was performed to investigate their effects on splicing patterns, and their effects on ABCB4 messenger RNA (mRNA) expression and MDR3 expression were explored. Data were evaluated using t-tests. A literature review of ABCB4 intronic variants was conducted.

RESULTS

The main clinical manifestation in both patients was cholestasis. Liver pathology analysis showed bile duct damage, inflammation, and fibrosis that differed between the patients. Three different pathogenicity prediction tools showed similar results, suggesting that these variants interfere with normal splicing. Minigene experiments demonstrated that c.537-32G>T creates a new alternative intron-splicing receptor, resulting in the retention of a 16-bp intronic sequence. The c.833+2T>C mutation created an alternative intron-splicing donor, resulting in the retention of a 67-bp intronic sequence. Neither variant greatly affected mRNA expression (P < 0.05), but MDR3 expression was downregulated by both (P < 0.05). Literature on ABCB4 intronic variants remains limited, and research into the in vitro pathogenic mechanisms of ABCB4 intron variants is especially scarce.

CONCLUSION

The c.537-32G>T and c.833+2T>C variants of ABCB4 are splicing variants, with 3’ and 5’ splicing sites, respectively. Clinical data, bioinformatic predictions, and in vitro experiments indicate that both are pathogenic.

Key Words: ABCB4; Multidrug resistance protein 3; Gene; Mutation; Variant; Sequencing; Intron; Splice; Liver

Core Tip: This study focuses on two rare ABCB4 intronic variants identified in a clinic through gene sequencing. The clinical characteristics of the patients mainly were cholestasis. And the variants were predicted in silico tools which showed pathogenic. Through literature review, we found two intron variants have not been reported, and there is little research on intron variants. In vitro experiments confirmed that both variants lead to intron retention. These two variants have little effect on ABCB4 messenger RNA and both downregulate multidrug resistance protein 3 expression. Which can explain the clinical pathogenicity of the two cases.

INTRODUCTION

ABCB4, a member of the ABC subfamily, is located on chromosome 7 (7q21) and is responsible for encoding multidrug resistance protein 3 (MDR3). MDR3 is composed of 12 transmembrane domains, six intracellular domains, six extracellular loops, and a linker peptide that connects the N- and C-terminal transmembrane domain nucleotide-binding domains. It uses the energy obtained from adenosine triphosphate (ATP) hydrolysis to transport specific substrates. ABCB4 variants can cause defects in the expression and function of MDR3. MDR3 is expressed in the capillary duct membranes of liver cells and functions in the transport of phospholipids, mainly phosphatidylcholine[1]. Under pathological conditions, ABCB4 variants and MDR3 function impairment lead to MDR3 downregulation, dysfunction, and phosphatidylcholine transport disorder, resulting in a series of cholestasis-related diseases. Different ABCB4 variants have different effects on MDR3 expression and function, and different clinical phenotypic spectra. The main clinical disease spectra related to ABCB4 variants include progressive familial intrahepatic cholestasis, low phospholipid associated cholelithiasis syndrome, intrahepatic cholestasis of pregnancy, and drug-induced liver injury. Among these, progressive familial intrahepatic cholestasis-3-associated histopathology is difficult to distinguish from that of other hereditary or acquired chronic cholestasis, such as biliary atresia, Alagille syndrome, or sclerosing cholangitis, and may overlap other phenotypes[2]. In recent years, we found that many patients with ABCB4 intronic variants have clinical manifestations of cholestasis[3]. However, the pathogenicity of these new variants is unclear and further research is needed.

Diseases are often caused by genetic variations that affect the splicing of precursor messenger RNA (pre-mRNA). Previous studies have shown that gene variants affecting RNA splicing account for 15%-30% of pathogenic variants and exceed 50% for certain genes/exons[4]. A key step in the gene expression of eukaryotes is the precise splicing of pre-mRNA, including recognition of the junction between exons and introns and the excision of introns. Previous studies have elucidated the structure and function of ABCB4, and the pathogenic mechanisms of its mutants in different liver diseases. However, research on missense variants is the main focus. Although numerous in vitro studies on the mechanisms of ABCB4 missense variants have been conducted, including cellular and animal experiments[5,6], research on intronic variants remains relatively scarce[7-9].

In this study, we detected two intronic variants in patients with clinical manifestations of cholestasis and used bioinformatic software to predict their impact on gene splicing. We then elucidated their effects on splicing through minigene splice assays. Our aim was to evaluate the pathogenicity and pathogenic mechanisms of these two newly identified intron variants. This comprehensive analysis of the pathogenicity and mechanisms of ABCB4 intronic variants, integrating clinical data with in vitro experiments, will facilitate early clinical intervention and decision-making for patients with relevant variants, while also contributing to the development of therapeutics.

MATERIALS AND METHODS

Patients

Among the outpatients and hospitalized patients presenting with cholestasis at the Department of Difficult Liver Diseases of The Second Hospital of Nanjing from 2018 to 2021, there were 219 patients whose etiology could not be clarified by routine physical examination, biochemistry, imaging, or liver puncture pathologic examination. Indications include diagnosing the causes of complex liver diseases, distinguishing overlapping diseases, assessing the degree of inflammation and fibrosis, etc.[10].

Next-generation sequencing detected ABCB4 variants in seven of these patients, including two patients with intronic variants. This study was approved by the Medical Ethics Committee of Nanjing Hospital Affiliated to Nanjing University of Traditional Chinese Medicine (No. 2021-LY-kt052). Informed consent was obtained from all patients prior to the study.

Genetic and in silico analysis

Genomic DNA was obtained from peripheral blood according to standard procedures, and next-generation sequencing of the target exome or whole exome was performed for all patients to explore potential genetic causes. The detected ABCB4 variant was verified by Sanger sequencing. Sequencing was conducted by KingMed Diagnostics Co. (Nanjing, Jiangsu Province, China). The sequencing-identified variants were mapped against common genomic databases (the Exome Sequencing Project, Human Gene Mutation Database, and Genome Aggregation Database) to determine whether the mutation sites were novel. Simultaneously, the National Center for Biotechnology Information and PubMed databases were searched to determine whether they were de novo variants. Three splice-variant prediction bioinformatics programs (SpliceAI[11], NNSplice and FSplice[12]) were used to assess the likelihood of the splice-site variants disrupting normal splicing and whether they could be pathogenic. The pathogenicity of the definitive variants was determined in accordance with the guidelines established by the American College of Medical Genetics and Genomics[12].

Minigene splicing assay

The pathogenicity of the two intronic ABCB4 variants was analyzed using a minigene assay. A partial sequence of the intron that contained the possible pathogenic variant, and the full-length sequence of the adjacent ABCB4 exons were cloned into pcDNA3.1. The constructed vectors were verified by Sanger sequencing (Supplementary Figure 1). A wild-type (WT) plasmid was used as a normal control. This analysis was performed by Wuhan BioEagle Biotechnology Co. (Wuhan, Hubei Province, China).

The WT and mutant plasmid expression vectors were transfected into HEK293T cells (Cell Bank of the Academy of Sciences, Beijing, China). The cells were harvested 48 hours after transfection and total cellular RNA was extracted using the FastPure^® Cell/Tissue Total RNA Isolation Kit V2. Total RNA was then reverse transcribed to complementary RNA (cDNA) and amplified using the following primers: MDR3-F: 5’ACATGGTCCTGCTGGAGTTC’, MDR3-R: 5’GCTGATGCCCAGTTCAAAGT3’, glyceraldehyde-3-phosphate dehydrogenase (GAPDH)-F: 5’GAAGGTGAAGGTCGGAGTCA3’, GADPH-R: 5’GACAAGCTTCCCGTTCTCAG3’. The amplified products were analyzed on 2% agarose gel and exposed under an ultraviolet (UV) gel imager. When the samples produced two target bands in the electrophoresis results, the two bands were cut under the UV lamp, and the gel was purified and recovered using the Agarose Gel Purification and Recovery Kit (PROTEINBIO, Nanjing, Jiangsu Province, China). For samples with only one band, the polymerase chain reaction (PCR) products were sent directly to Tianlin Biotechnology Co. (Nanjing, Jiangsu Province, China) for sequencing.

Quantitative real-time-PCR assay

To explore the effects of the intronic variants on mRNA and protein levels, intronic variant expression vectors containing the complete cDNA of ABCB4 were constructed. The WT plasmids contained the coding DNA sequence region of the full segment of ABCB4. Total RNA was extracted from HEK 293T cells using quantitative real-time-PCR and reverse-transcribed for mRNA expression analysis. Relative amounts were analyzed using GAPDH (ABclonal, Wuhan, Hubei Province, China) as an internal reference. For further details, refer to the ChamQ SYBR quantitative PCR Master Mix (Vazyme, Nanjing, Jiangsu Province, China) instructions for 40 cycles.

Western blot

Proteins were collected from HEK293T cells after transfection in six-well plates for 48 hours. All procedures were performed on ice. The processed protein samples were separated by 6% (w/v) sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred onto polyvinylidene difluoride membranes, which were then closed with skimmed milk. Anti-MDR3 P3II-26 antibody (Thermo, MA, United States) (1:200) was used as the primary antibody for the target protein and GAPDH (ABclonal) (1:10000) was used as the internal reference. Horseradish peroxidase goat anti-mouse immunoglobulin G (1:5000) was used as the horseradish antibody. Fluorescent color development was performed using an ultrasensitive luminescent liquid kit (Biosharp, Beijing, China) and a Fluorchem M instrument (ProteinSimple, Beijing, China).

Review of the literature

We comprehensively reviewed the ABCB4-related literature on the PubMed and China National Knowledge Infrastructure databases using the keywords “ABCB4” OR “ATP binding cassette subfamily b member 4” OR “MDR3” OR “multidrug resistance protein 3” AND “splice” OR “splicing” OR “intron” OR “intronic”. The literature review was restricted to clinical, in silico, and in vitro studies that employed sequencing and explicitly addressed intronic variants in ABCB4.

Statistical analysis

Categorical data are presented as n, while the continuous data prepared during the data processing for the quantitative PCR and western blot were summarized as mean ± SD. Pairwise comparisons were performed using the Student’s t-test. Descriptive analyses were conducted with SPSS 25.0 software (IBM Corp., Armonk, NY, United States), and statistical graphs were generated using GraphPad Prism version 9.0 (GraphPad Software, San Diego, CA, United States).

RESULTS

Clinical features

The patients were both adults. Their liver function indicated increases in bilirubin and transaminase levels, and differences were especially evident in their degree of cholestasis and elevated alkaline phosphatase and gamma-glutamyl transferase levels. Imaging revealed cirrhosis, splenomegaly, and cholecystitis. Liver biopsy revealed that the two patients showed different degrees of bile duct damage, inflammation, and fibrosis. The bile duct damage, inflammation, and fibrosis of the patient carrying the c.537-32G>T variant (patient 1) were less severe, and their immunohistochemical MDR3 expression was normal. The pathology of the patient with the c.833+2T>C variant (patient 2) included S4 fibrosis, bile duct deficiency, and reduced immunohistochemical MDR3 expression (Table 1).

Table 1 Clinical characteristics, ABCB4 variant information, pathological features, and immunohistochemical results of two patients.

Patient	Sex	Age of disease	Clinical features	TB (max)	ALT (max)	ALP (max)	GGT (max)	Imaging findings	Gene name	Reference sequence information	Allele 1 (amino acid)	Allele 2 (amino acid)	Zygosity	Inflammation grading and fibrosis stage	Bile ducts	MDR3 IHC
1	Male	21	Chronic cholestasis	29	159	346	1016	Liver cirrhosis, calcification of right lobe of liver, cholecystitis, mild splenomegaly	ABCB4	NG 007118.3	c.2362C>T (p.R788W)	c.537-32G>T (p. ?)	Homozygote	G2-3S2	Reduction	Normal
2	Male	37	Chronic cholestasis, cirrhosis, portal hypertension, and upper gastrointestinal bleeding	21	51	386	219	Liver cirrhosis, splenomegaly, portal vein enlargement, portal hypertension, cholecystitis	ABCB4	NG 007118.3	c.833+2T>C (p. ?)	(-)	Heterozygote	G1-3S4	Absence	Reduction

Biochemical indicators represent the highest test values at initial diagnosis of the patient. Reference values: Total bilirubin (3-19 μmol/L); Alanine aminotransferase (5-40 U/L); Alkaline phosphatase (40-129 U/L); Glutamyl transpeptidase (0-50 U/L). G: Inflammation grading; S: Fibrosis stage; MDR3: Multidrug resistance protein 3; IHC: Immunohistochemistry; TB: Total bilirubin; ALT: Alanine aminotransferase; ALP: Alkaline phosphatase; GGT: Glutamyl transpeptidase.

Plasmid expression vector sequencing and silico analysis

Both intronic variants may interfere with the splicing of ABCB4. Sequencing demonstrated that, for c.537-32G>T, an intronic sequence of 16-bp was retained on the right side of exon 6 when compared to the WT sequence (Supplementary Figure 1). For c.833+2T>C, a 67-bp intronic sequence was retained on the left side of exon 8 (Supplementary Figure 1). SpliceAI suggested two possible pathways through which both variants could affect splicing, while NNsplice and FSplice suggested that both variants may affect the location of the splice donor (Table 2). Both variants were classified as having uncertain significance, according to the American College of Medical Genetics and Genomics guidelines.

Table 2 Prediction results of bioinformatics software for two intron variants.

Variants	Gene name	Reference sequence information	SNP	gnomAD	HGMD	SpliceAI			NNSplice	FSplice	ACMG
						Type	Score	Pre-mRNA position
c.537-32G>T	ABCB4	NG 007118.3	(-)	(-)	(-)	Acceptor loss	0.18	-32 bp	Donor loss	Donor loss	Uncertain significance (PS3 + PM2 + PM4 + PP3)
						Acceptor gain	0.09	-16 bp
c.833+2T>C	ABCB4	NG 007118.3	(-)	(-)	(-)	Donor loss	0.99	2 bp	Donor loss	Donor loss	Uncertain significance (PS3 + PM2 + PM4 + PP3)

SNP: Single nucleotide polymorphism; gnomAD: Genome aggregation database; HGMD: Human gene mutation database; Pre-mRNA: Precursor messenger RNA; ACMG: American college of medical genetics and genomics; PS: Pathogenic strong; PM: Pathogenic moderate; PP: Pathogenic supporting.

ABCB4 minigene splicing analysis

The effects of the two intronic variants were analyzed using minigene splice assays. In the DNA electrophoresis results (excluding the mock sample) only mutation-c.833+2T>C had one band, whereas the rest of the samples had two bands (Figure 1A). To determine the desired destination bands, we performed sequence analysis using molecular biology analysis software. The upper band of WT1 showed normal splicing; the lower band showed abnormal splicing with an exon deletion. The lower band of WT2 was a normal splicing reference sequence, and the upper band showed an abnormal splicing sequence. We found that the c.537-32G>T variant resulted in the creation of a new intronic shear acceptor in front of the original intronic shear receptor site, leading to the retention of a 16-bp intronic sequence in the exon (Figure 1B). Similarly, the c.833+2T>C variant caused the original intron-shearing donor site to become inactive, followed by the activation of another intron-shearing donor site, resulting in the retention of a 67-bp intronic sequence (Figure 1C).

Figure 1 Minigene splice assays in variants of c.537-32G>T and c.833+2T>C. A: Agarose gel electrophoresis of real-time polymerase chain reaction products; B: Multiple sequence alignment of the variant c.833+2T>C [mutation, wild-type (WT) 1, and exon 8 + exon 9 of ABCB4] demonstrates the retention of a 67-bp intronic sequence; C: Multiple sequence alignment of the c.537-32G>T variant (mutation, WT2, and exon 6 + exon 7 of ABCB4) demonstrates the retention of a 16-bp intronic sequence. MOCK: PcDNA3.1 empty vector plasmid; WT: Wild-type; Mut: Mutation.

mRNA expression

ABCB4 mRNA expression in the two variants was similar to that of ABCB4-WT (P < 0.05) (Figure 2), suggesting that these variants have little effect on ABCB4 mRNA content.

Figure 2 Messenger RNA expression levels of wild-type and intron mutants in ABCB4. Relative quantity analysis was performed with reference to glyceraldehyde-3-phosphate dehydrogenase. These values represent the average ± SD of the four independent experiments. The data were analyzed in GraphPad Prism 9.5.1. ^aP < 0.05. WT: Wild-type; mRNA: Messenger RNA.

MDR3 expression

In vitro, we identified mature (160 kDa) and immature (140 kDa) forms of MDR3, consistent with previous findings[6,13]. The results showed that the expression levels of MDR3 in the two mutants were lower than those in the WT, with relative expression levels of 0.883 and 0.838, respectively (P < 0.05) (Figure 3).

Figure 3 Multidrug resistance protein 3 expression levels of wild-type and intron mutants in ABCB4. A: Western blot bands of ABCB4-wild-type (WT) and two intron mutants. Among them, WT was transfected with ABCB4 WT plasmid, and c.537-32G>T and c.833+2T>C were transfected with mutant plasmids, respectively; B: Analysis of multidrug resistance protein 3 (MDR3) expression level. The ratio of MDR3 to glyceraldehyde-3-phosphate dehydrogenase was taken as a relative quantity, and the data were standardized against the WT. These data are the average ± SD of three independent experiments. The grayscale value of the western blot strip was measured using Image J software. The data were analyzed in GraphPad Prism 9.5.1. ^bP < 0.05. WT: Wild-type; MDR3: Multidrug resistance protein 3; GAPDH: Glyceraldehyde-3-phosphate dehydrogenase.

Findings of the literature review

We retrieved 102 records from the China National Knowledge Infrastructure and PubMed databases, 14 of which were associated with ABCB4 intron variants. The references comprised 14 publications, including three in Chinese[7,14,15] and 11 in English[8,9,16-24]. These included eight clinical studies, seven bioinformatics analyses, three sequencing studies, three in vitro experimental studies, and one review article (Supplementary Table 1 and Supplementary Figure 2).

DISCUSSION

Among the many previously studied clinical cohorts, missense variants in exons are the most commonly reported ABCB4 defects, whereas intron variants are rare. In this study, we detected two novel intron variants of ABCB4 in patients with unexplained liver disease and evaluated their pathogenicity.

Three different in silico analyses suggested that the two variants may affect the correct splicing of introns. We then used micro-gene splicing analysis in vitro to verify that the two variants were splicing variants that caused abnormal splicing patterns. These variants caused splice-site substitutions and retained a redundant intron sequence, which affected the mRNA and protein expression of ABCB4 to different degrees.

The patient with the c.833+2T>C variant (patient 2) showed a more severe clinical presentation, with the development of cirrhosis, portal hypertension, and upper gastrointestinal bleeding at the first visit. Liver biopsy showed dense and proliferative hepatocytes in the lobules, with a small amount of focal necrosis. The portal area was enlarged, inflammatory cells slightly increased, and fibers significantly increased, resulting in multiple bridging fibrosis. Moreover, the bile duct was missing. Immunohistochemistry revealed decreased MDR3 expression. The in vitro experiments suggest that this variant affects the splicing of ABCB4. This variant is classified as a class I intronic variant that disrupts conserved sequences at both ends of the intron (e.g., canonical GT-AG dinucleotides)[25]. Pathogenesis is caused by the inactivation of the 5’ splice site and the activation of the potential splicing initiation recognition site, resulting in the retention of a 67-bp intronic fragment that would normally be removed. Interestingly, this variant exhibited minimal impact at the mRNA level compared to the WT, suggesting that it does not induce premature mRNA degradation. Western blot analysis revealed a decrease in MDR3 protein levels, which may have resulted from an in-frame frameshift caused by the variant producing an unstable or non-functional protein. This mechanism likely contributed to the phenotype observed in patient 2. In 2020, Sticova et al[26] reported an adjacent variant site, c.833+1G>T, which existed only in the index patient and strongly suggested pathogenicity. Therefore, the destruction of highly conservative GT-AG splicing motifs (± one or two slicing junctions) is strong evidence toward being identified as a pathogenic variant. Only one variant was detected in patient 2. However, the severe clinical manifestation and decreased MDR3 expression in the liver biopsy confirmed the strong pathogenicity of classical splice site variants.

The clinical manifestation in the patient with variant c.537-32G>T (patient 1) was chronic cholestasis. Imaging did not reveal liver cirrhosis. The pathology of the liver puncture included mild-to-moderate interfacial inflammation. Fibrous hyperplasia of the liver was observed, mainly in the portal area. The bile ducts were also reduced. However, MDR3 expression was normal. Considering that this patient carried both the intronic variant c.537-32G>T and the missense variant c.2362C>T, and since our previous research as well as the 2008 study by Schneider et al[18] confirmed that c.2362C>T downregulates phosphatidylcholine translocation, we conducted an in vitro experimental analysis of c.537-32G>T. The in vitro experiments suggested that this variant affects splicing, and that its pathogenesis involves a new intron terminal splice site before the original intron splice site is activated, resulting in the retention of a 16-bp intronic sequence. This variant can be classified as a class II intronic mutation, involving the activation of a cryptic splice site within a deep intron, leading to pseudoexon inclusion. The mRNA expression level of c.537-32G>T was not significantly different from that of the WT, indicating that the variant does not introduce a premature termination codon that would trigger mRNA degradation. Western blot analysis revealed a reduction in MDR3 protein levels relative to the WT. Although the decrease was less pronounced than that caused by the 833+2T>C variant, it partially explained the phenotype observed in patient 1.

A review of the literature revealed that research on the introns of ABCB4 is limited, with most studies focusing on sequencing and clinical investigations. In vitro mechanistic studies are also scarce. Two in vitro studies have reported mutations leading to frameshift deletions. Schneider et al[18] reported that the intronic variant c.3486+5G>A activates a cryptic splice site, resulting in an in-frame deletion of 54 bp (3465-3518). A Chinese study indicated that the c.2065-8T>G variant induces intron retention[14]. The patients described in that report presented with clinical phenotypes consistent with pathogenic effects. These reported cases share similarities with our observations, likely due to frameshift alterations that compromise functional MDR3 expression.

Currently, research on the treatment of patients with ABCB4 intronic variants is limited. Considering that these mutations may cause substantial frameshift alterations, the clinical manifestations and prognosis of carriers are highly heterogeneous and ultimately dependent on the pathogenic mechanisms of the individual mutation sites. Some studies have reported the effectiveness of ursodeoxycholic acid therapy[14], while others have reported poor therapeutic responses[7]. These patients may eventually require artificial liver support or liver transplantation for end-stage disease. Research on molecular therapies targeting ABCB4 also remains limited. In 2019, Song et al[27] found that miRNA-378a-5p negatively regulates MDR3 expression at both the mRNA and protein levels, which may represent a potential therapeutic strategy for a subset of patients. Other therapeutic approaches for intronic mutations exist, including splice-switching antisense oligonucleotides, small-molecule splicing modulators, RNA editing, and gene therapy approaches[28], although they have not yet been applied to ABCB4. These modalities show potential for future research into the treatment of ABCB4 intronic variants.

However, the sample size of the current study was small, limiting our ability to reach statistically significant conclusions to some extent. In addition, owing to economic limitations and subjective patient factors, this study failed to improve on the pedigree investigation of the probands. Functional experiments to further explore the effects of these two variants on the activity and stability of the MDR3 protein are warranted.

CONCLUSION

We identified two new ABCB4 intron variants and analyzed their pathogenicity in vitro. Both variants resulted in the retention of an intronic sequence within the reading frame, thereby impairing MDR3 expression. After excluding other cholestatic liver diseases, gene detection can be used in patients with jaundice accompanied by itching and whose liver function suggests elevated levels of conjugated bilirubin, transaminase, and alkaline phosphatase, especially in those with a family history of cholestasis. This could determine whether the patient has a related disease caused by gene variants. Novel variants that may affect splicing, especially intronic variants, can be analyzed using splice-site prediction tools. Different ABCB4 variants have different pathogenicity. Pathogenicity can be further verified, and the pathogenic mechanism explored, using in vitro assays. In-depth exploration of the pathogenic mechanisms is crucial for establishing a comprehensive understanding of ABCB4 genotype-phenotype correlations, and is also expected to guide clinical treatment.

ACKNOWLEDGEMENTS

We would like to express our gratitude to all the patients and researchers who participated in this study and to the biostatistician for their rigorous review.