Chen et al.

Orphanet Journal of Rare Diseases (2022) 17:445

https://doi.org/10.1186/s13023-022-02597-y

RESEARCH Open Access

Clinical and genetic characterization

of pediatric patients with progressive familial
intrahepatic cholestasis type 3 (PFIC3):
identification of 14 novel ABCB4 variants
and review of the literatures
Rong Chen1†, Feng‑Xia Yang2†, Yan‑Fang Tan3, Mei Deng1, Hua Li1, Yi Xu2, Wen‑Xian Ouyang3* and
Yuan‑Zong Song1*

Abstract
Background: Progressive familial intrahepatic cholestasis type 3 (PFIC3) is an autosomal recessive disease caused by
pathogenic variants of the gene ABCB4. This study aimed to investigate the ABCB4 genotypic and the clinical pheno‑
typic features of PFIC3 patients.
Methods: The clinical and molecular genetic data of 13 new pediatric patients with PFIC3 as well as 82 reported
ones in the PubMed and CNKI databases were collected and analyzed.
Results: The 13 new PFIC3 patients included six females and seven males, and the main presentations were hepato‑
megaly, splenomegaly, jaundice, and pruritus, as well as increased levels of gamma-glutamyl transpeptidase (GGT).
Fourteen new ABCB4 variants were detected, including eight diagnosed to be likely-pathogenic and six, pathogenic.
Among all the 95 PFIC3 cases, hepatomegaly was observed in 85.3% (81/95), pruritus in 67.4% (64/95), splenomegaly
in 52.6% (50/95), jaundice in 48.4% (46/95), portal hypertension in 34.7% (33/95) and GGT elevation in 100% (88/88)
of the patients. Positive responses at varied degrees to oral ursodeoxycholic acid (UDCA) treatment were observed
in 66.1% (39/59) of the patients, among whom 38.5% (15/39) fully recovered in terms of the laboratory changes.
Although the condition remained stable in 53 patients (58.9%, 53/90), the clinical outcomes were not promising in
the rest 37 cases (41.1%, 37/90), including 7 died, 27 having undergone while another 3 waiting for liver transplanta‑
tion. A total of 96 ABCB4 variants were detected in the 95 patients. PFIC3 patients with biallelic null variants exhibited
earlier onset ages [10.5 (2, 18) vs. 19 (8, 60) months, p = 0.007], lower UDCA response rate [18.2% (2/11) vs. 77.1%
(37/48), p = 0.001], and more unpromising clinical outcomes [80% (12/15) vs. 33.3% (25/75), p = 0.001], compared
with those with non-biallelic null variants.

†
Rong Chen and Feng-Xia Yang have contributed equally to this work
*Correspondence: [email protected]; [email protected]
1
Department of Pediatrics, The First Affiliated Hospital, Jinan University,
Guangzhou 510630, China
3
Department of Hepatopathy, Hunan Children’s Hospital,
Changsha 410007, China
Full list of author information is available at the end of the article

© The Author(s) 2022. Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which
permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the
original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or
other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line
to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory
regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this
licence, visit http://​creat​iveco​mmons.​org/​licen​ses/​by/4.​0/. The Creative Commons Public Domain Dedication waiver (http://​creat​iveco​
mmons.​org/​publi​cdoma​in/​zero/1.​0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 2 of 12

Conclusions: PFIC3 presented with hepatomegaly, pruritus, splenomegaly and jaundice with increased serum GGT
level as a biochemistry hallmark. Although varying degrees of improvement in response to UDCA therapy were
observed, 41.1% of PFIC3 patients exhibited unfavorable prognosis. ABCB4 genotypes of biallelic null variants were
associated with severer PFIC3 phenotypes. Moreover, the 14 novel variants in this study expanded the ABCB4 muta‑
tion spectrum, and provided novel molecular biomarkers for diagnosis of PFIC3 patients.
Keywords: Progress familial intrahepatic cholestasis type 3 (PFIC3), ABCB4 gene, Novel variants

Introduction 12 unrelated families. The clinical data of the 13 patients

Progressive familial intrahepatic cholestasis (PFIC) were collected for analysis, including their ages, genders,
included a group of rare autosomal recessive diseases history, clinical presentations, laboratory changes, treat-
caused by pathogenic variants of the genes encoding ment and outcomes. Most the data were collected from
proteins related to the formation and transfer of bile the medical record databases in the participating hos-
acids in the liver [1]. The PFIC patient’s onset varied pitals, with partial data from other hospitals being pro-
from the neonatal period to early adulthood, which usu- vided by patients’ parents at their referrals to our clinics.
ally developed fibrosis and end-stage liver disease before This study was approved by the Medical Ethics Com-
adulthood [2]. Based on different causative genes, PFIC mittee of the First Affiliated Hospital, Jinan University,
could be divided into types 1–6, with ATP8B1, ABCB11, and written informed consents were signed by the par-
ABCB4, TJP2, NR1H4, and MYO5B being the causative ents of all patients before this study.
gene, respectively [3].
The gene ABCB4 causing PFIC3 (OMIM # 602347) was Genetic analysis
located on chromosome 7q21, which encoded a liver- Genomic DNA was obtained from peripheral blood
specific canalicular transporter, the Multi-Drug Resistant according to standard procedures, and all patients under-
3 (MDR3) protein [4]. MDR3 translocated phosphatidyl- went next-generation sequencing (NGS) of the targeted
choline from the inner to the outer leaflet of the canali- or whole exomes, to explore the underling genetic causes.
cular membrane, resulting phosphatidylcholine efflux ABCB4 variants detected were then verified by Sanger
into the bile [5, 6]. In the aqueous environment of bile, sequencing. The sequencing results were aligned with
phospholipids form mixed micelles with cholesterol and the ABCB4 gene sequence (ENST00000649586.2), which
bile acids, thereby preventing the formation of choles- was available at Ensembl Genome Browser (www.​ensem​
terol gallstones and the detergent action of free bile acids bl.​org). The variant nomenclature was in agreement with
which was injurious to cholangiocyte membrane [7, 8]. current guidelines of the Human Genome Variation Soci-
Typical clinical features of PFIC3 included jaundice, pru- ety (http://​www.​hgvs.​org/​rec.​html).
ritus, hepatomegaly, and splenomegaly, which could pro-
gress to cirrhosis and liver failure before adulthood. Thus Pathogenicity evaluation
far, due to the lack of pathognomonic clinical symptoms All the variants were classified according to the Ameri-
or signs, the definitive diagnosis of PFIC3 relied on the can College of Medical Genetics and Genomics (ACMG)
ABCB4 genetic analysis [9]. standards and guidelines.
In the recent years, the clinical application of molecu- The allele frequencies of the identified variants were
lar genetic techniques facilitated the timely diagnosis collected from the 1000 Genomes Project (https://​www.​
of PFIC patients, and an increasing number of PFIC3 inter​natio​nalge​nome.​org), the Genome Aggregation
patients were reported around the world [10–15]. How- Database (gnomeAD, http://​gnomad-​sg.​org), the Human
ever, the molecular and clinical characteristics of this Gene Mutation database (http://​www.​hgmd.​cf.​ac.​uk/​
condition, generally as a rare liver disease, remained yet ac/​index.​php), and all relevant literatures in database of
far from being completely understood. This study ana- PubMed (https://​pubmed.​ncbi.​nlm.​nih.​gov).
lyzed the phenotypic and genotypic features of PFIC3 Conservation of mutated amino acids was analyzed
patients by reporting 13 new pediatric patients and by comparatively aligning the amino acid sequences of
reviewing the relevant literatures. ABCB4 orthologs collected from the Ensembl Genome
Browse. The 20 primate homologous proteins include
Methods human, angola colobus, tarsier, chimpanzee, bonobo,
Subjects and ethical approval bushbaby, black snub-nosed monkey, gorilla, drill, capu-
The research subjects in this study included 13 pediat- chin, gelada, gibbon, macaque, olive baboon, mouse
ric patients including seven males and six females from lemur, golden snub-nosed monkey, sooty mangabey,
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 3 of 12

coquerel’s sifaka, pig-tailed macaque, and bolivian squir- liver failure. Nine patients were alive with stable condi-
rel monkey. tion, and the rest two lost contact.
The pathogenicity of the novel missense variants was All the 13 patients were either homozygous or com-
predicted by using the three online programs PolyPhen-2 pound heterozygous for ABCB4 variants (Additional
(http://​genet​ics.​bwh.​harva​rd.​edu/pph2), Mutation Taster file 1: Fig. S1). As listed in Fig. 1, a total of 23 ABCB4 vari-
(http://​www.​mutat​ionta​ster.​org) and PROVEAN (http://​ ants were detected in the 13 new PFIC3 patients, and 14
prove​an.​jcvi.​org/seq_submit.php). PolyPhen-2 analysis variants were not detected in any PFIC3 patients previ-
identified variant as probably damaging if the probabil- ously, including eight missense c.2782A > G(p.Arg928Gly),
ity was > 0.85, and possibly damaging if the probability c.1645C > T(p.Arg549Cys), c.1801G > A (p.Ala601Thr),
was > 0.15. Mutation Taster value close to 1 indicated a c.1406G > A(p.Arg469Lys), c.716C > T(p.Ser239Leu),
high security of the prediction. PROVEAN predicted a c.3230C > T(p.Thr1077Met), c.2914G > A(p.Asp972Asn),
variant as “deleterious” if the prediction score was < − 2.5. and c.965 T > C(p.Leu322Pro), three frameshift c.3100_
Moreover, the online bioinformatics tools NNspl (http://​ 3101insA(p.Ile1034A snfs*4), c .3789delA(p.Gly-
www.​fruit​fly.​org/seq_tools/splice.html) and Human 1264Alafs*38) and c.879dupA(p.Ala294Serfs*62), two
Splicing Finder (http://​www.​umd.​be/HSF3/HSF.shtml) splicing-site c.136-2A > G and c.80 + 1G > C, as well as one
were used to assess the potential of splicing-site variants nonsense variant(s) c.2123G > A(p.Trp708*).
to disrupt normal splicing. The variants c.1801G > A(p.Ala601Thr) and c.3230C >
T(p.Thr1077Met) were included in the database
gnomeAD with the allele frequencies of 0.007962‰
Review of the literatures
and 0.1991‰, respectively, while the rest 12 novel vari-
Electronic databases including PubMed and CNKI
ants have neither been reported in any official literatures
(https://​www.​cnki.​net) were retrieved by using the key-
nor included in any variant databases, to the best of our
words “ABCB4” and “PFIC3”. The genotypic and pheno-
knowledge. The amino acid sequences of the homologous
typic data of the pediatric patients with clear molecular
peptides in a total of 20 primates were aligned compara-
genetic diagnosis and detailed clinical information were
tively, and the results showed the eight novel missense
collected and analyzed.
variants were localized in highly conserved regions of all
the 20 primate homologous proteins (Fig. 2).
Statistical analysis In the study, six out of the 14 novel variants were classi-
Data were analyzed with the use of IBM SPSS Statistics fied to be pathogenic, and remaining eight as likely path-
26 software (IBM, Armonk, NY, USA). Normally distrib- ogenic, according to the ACMG standards. The relevant
uted data were expressed as mean ± SD and then com- evidences were listed in detail in Table 2.
pared using Student’s t-test. Data of skewed distribution
were presented as the median values (P25, P75), and
Findings of literature review
comparisons were conducted by means of Mann–Whit-
The clinical and molecular genetic data of 82 pediatric
ney U-test. Categorical variables were expressed as per-
PFIC3 patients, who were definitely diagnosed with bial-
centages, and statistical differences were compared by
lelic ABCB4 variants and had relatively complete clini-
Chi-Square or Fisher’s exact test. Statistical significance
cal information in 20 official literatures [11–13, 15–31],
was set at p < 0.05.
were summarized in Additional file 2: Table S1. Together
with the 13 new PFIC3 patients in this study, a total of
Results 95 patients (38 females, 39 males and 18 gender unde-
Clinical and genetic characteristics of the 13 new PFIC3 scribed) were in-depth analyzed in terms of the pheno-
patients typic and genotypic features.
Table 1 summarized the clinical information of the 13 The ages of symptom onset were 18 (7, 50) months.
new PFIC3 patients from 12 unrelated families. The ages Hepatomegaly was observed in 85.3% (81/95), pruritus in
of symptom onset were 36 (8, 67) months. As the com- 67.4% (64/95), splenomegaly in 52.6% (50/95), jaundice in
monest clinical presentation in this cohort, hepatomegaly 48.4% (46/95), portal hypertension in 34.7% (33/95) and
was observed in 13 patients, followed by splenomegaly failure to thrive (14.7%, 14/95) of all the cases. In addi-
in 11, jaundice in seven, pruritus in four cases. The bio- tion, gastrointestinal bleeding, discolored stools, gall-
chemistry hallmark was markedly increased levels of stone, abdominal distension, ascites, cholecystitis, and
gamma-glutamyl transpeptidase (GGT) in all patients. reduced bone density were also observed some patients.
On the last follow up, two patients demonstrated unfa- Regarding serum biochemistry, 88 patients (100%, 88/88)
vorable outcomes: one was waiting for liver transplanta- exhibited increased levels of GGT. Besides, the elevation
tion due to hepatic decompensation and one had died of of aspartate transaminase (ALT), alanine transaminase
Table 1 Clinical information of the 13 new PFIC3 patients
Patient Sex Family history Age at Clinical Age Anthropometry and laboratory changes at first referral Pathologic Outcomes at the
No onset presentations Features last follow up
Wt Ht ALT AST GGT​ TBIL DBIL TBA 25(OH)D
(kg) (cm)

1 Female No 4.2Y Pruritus, spleno‑ 5Y 17 105 277 232 589 48.8 32.9 88.3 4.87 NA Aged 8Y, awaiting LT
megaly, hepato‑ (− 0.6SD) (− 1.2SD)
megaly, jaundice
2 Female Elder sister of 10.8Y Jaundice, hepato‑ 11.3Y NA NA 63 149 302 115.4 109.4 125.3 NA NA Died of liver failure at
patient 3 megaly, spleno‑ the age 12Y
megaly, gastroin‑
testinal bleeding,
Chen et al. Orphanet Journal of Rare Diseases

portal hypertension
3 Male Younger brother of 7.2Y Hepatomegaly, 8.5Y 27.5 131 109 88 95 3.7 2 7.5 NA NA Loss of follow-up
patient 2 splenomegaly (− 0.3SD) (− 0.3SD)
4 Female No 3Y Hepatomegaly, 6.7Y 17 109 83 56 72 15.5 6.6 93 28 Nodular cirrhosis Aged 9Y, alive
splenomegaly, (− 2.8SD) (− 2.9SD)
failure to thrive
(2022) 17:445

5 Female Elder sister and 1M Jaundice, hepato‑ 2M 5.7 58.8 43 35 637 67.2 20 154.4 20.7 NA Aged 4 M, alive
brother died of liver megaly (+ 0.8SD) (+ 0.6SD)
failure
6 Male Elder brother with 3D Jaundice, hepato‑ 6Y 23.5 116 381 173 307 222 100.8 63.3 5.36 NA Aged 7Y, alive
PFIC3, LT at 9y megaly, spleno‑ (+ 0.8SD) (− 0.4SD)
because of liver megaly, pruritus
failure
7 Female Elder sister died of 6M Hepatomegaly, 8M 7.5 68 120 175 108 67.3 56.4 280.2 NA NA Aged 1.3Y, alive
intracranial hemor‑ splenomegaly, (-1.3SD) (-1.5SD)
rhage jaundice
8 Male No 1Y Hepatomegaly 1.6Y 12.5 80.4 184 130 102 6.1 3.8 98.6 27.6 NA Aged 2Y, alive
(+ 0.9SD) (-0.7SD)
9 Female No 3Y Hepatomegaly, 3Y 12.5 94 186 207 202 17.1 3.6 148.2 NA NA Aged 5.1Y, alive
splenomegaly, (− 1.3SD) (− 0.4SD)
jaundice, discolored
stools
10 Male Elder brother died 6.2Y Hepatomegaly, 6.2Y 21.8 115.6 127 125 354 11.7 6.5 22.2 NA NA Aged 8.3Y, alive
of liver failure splenomegaly (− 0.5SD) (+ 0.2SD)
11 Male Mother with 5Y Pruritus, hepato‑ 6Y 16.1 106.5 228 99 136 21.6 7.4 276 NA Liver fibrosis Aged 9.4Y, alive
intrahepatic megaly, spleno‑ (− 2.8SD) (− 2.5SD)
biliary stones. Elder megaly, failure to
brother and sister thrive
died of liver failure
12 Male No 10 M Jaundice, hepato‑ 2.5Y 14 95 30 85 206 50.1 17.1 114.5 NA NA Loss of follow-up
megaly, spleno‑ (− 0.4SD) (− 1SD)
megaly
Page 4 of 12
 Chen et al. Orphanet Journal of Rare Diseases

Table 1 (continued)
Patient Sex Family history Age at Clinical Age Anthropometry and laboratory changes at first referral Pathologic Outcomes at the
(2022) 17:445

No onset presentations Features last follow up

Wt Ht ALT AST GGT​ TBIL DBIL TBA 25(OH)D
(kg) (cm)

13 Male Mother with ICP 4.3Y Hepatomegaly, 5Y 18 107 156 145 197 12.9 8.3 108.6 NA Ductal proliferation Aged 5.4Y, alive
splenomegaly, (− 0.5SD) (− 1SD) and inflammatory
pruritus infiltration
Reference ranges: ALT: (5–40U/L); AST: (5–40U/L); GGT: (8–50U/L); Tbil: (5.1–23 μmol/L); Dbil: (0.6–6.8 μmol/L); TBA: (0–10 μmol /L); 25(OH)D: (≥ 20 ng/ml). Among the five patients with 25(OH)D analyzed, two exhibited
vitamin D deficiency(< 20 ng/ml)
Y year; M month; D day; ICP intrahepatic cholestasis of pregnancy; Wt weight; Ht height; ALT alanine aminotransferase; AST aspartate aminotransferase; GGT​ γ-glutamyl transpeptidase; Tbil total bilirubin; Dbil direct
bilirubin; TBA total bile acids; 25(OH)D 25-hydroxyvitamin D; NA not available; LT liver transplantation.
Page 5 of 12
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 6 of 12

Fig. 1 ABCB4 genotypes of the 13 new PFIC3 patients from 12 unrelated families. The arrows indicated the probands in each family, and the
deceased individuals were shown with a slash. Circles and squares represented females and males, respectively. The different ABCB4 variants were
illustrated in different colors in this figure

(AST), total bilirubin (TBil), direct bilirubin (DBil), total Genotype–phenotype correlation
bile acids (TBA) was observed in 91.9% (68/74), 98.4% In this study, the frameshift, nonsense, canonical ± 1 or 2
(61/62), 45.5% (30/66), 45.3% (24/53) and 93.8% (45/48) splicing-site variants and exons deletion were defined as
of the patients, respectively. null variants, while missense and non-canonical splicing-
Positive responses at varied degrees to oral ursodeoxy- site, as non-null variants, according to the ACMG stand-
cholic acid (UDCA) treatment were observed in 66.1% ards. The ABCB4 phenotypes of the 95 PFIC3 patients were
(39/59) of the patients, of whom 38.5% (15/39) fully categorized into two groups: biallelic null variants (n = 18)
recovered in terms of the laboratory changes. In clinical and non-biallelic null variants (n = 77). It was found that
outcome analysis on the last follow-up, five out of all 95 PFIC3 patients with biallelic null variants exhibited earlier
patients were excluded due to loss of contact; the condi- onset ages [10.5 (2, 18) vs. 19 (8, 60) months, p = 0.007],
tion remained stable in 53 patients (58.9%, 53/90), while lower UDCA response rate [18.2% (2/11) vs. 77.1% (37/48),
the clinical outcomes were not promising in the rest 37 p = 0.001)], and more unpromising clinical outcomes [80%
cases (41.1%, 37/90), including 7 died, 27 having under- (12/15) vs. 33.3% (25/75), p = 0.001], compared with those
gone while another 3 waiting for liver transplantation. with non-biallelic null variants genotypes.
Among the 95 PFIC3 patients, a total of 96 differ-
ent ABCB4 variants were detected, and missense vari-
ants 60.4% (58/96) was on top of the list, followed by Discussion
frameshift (14.6%, 14/96), nonsense (11.5%, 11/96), splic- This study described 13 new PFIC3 patients from 12
ing-site (12.5%, 12/96), exons deletion (1%, 1/96), as sum- unrelated families, and among the 23 ABCB4 variants
marized in Fig. 3. detected, 14 were not reported previously in any official
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 7 of 12

Fig. 2 Comparative alignment of the homologous peptides affected by the eight novel missense variants in 20 primate species. The eight novel
missense variants all affected a highly conserved amino acid residue of MDR3 protein

literatures, including eight missense, three frameshift, Ser239Leu), and c.3230C > T(p.Thr1077Met) proved to
two splicing-site, and one nonsense variant(s). The two be in trans with a previously-reported pathogenic vari-
frameshift variants c.3100_3101insA(p.Ile1034Asnfs*4) ant by testing parents (PM3) [11, 21]. All the novel mis-
and c.879dupA(p.Ala294Serfs*62) as well as the nonsense sense variants cosegregated with the PFIC3 phenotype in
variant c.2123G > A(p.Trp708*) introduced premature this study (PP1). It was well-known that ABCB4 missense
stop codons in the MDR3 residues 1037, 355 and 708, variants were a common mechanism for PFIC3 develop-
respectively. Another frameshift variant c.3789delA(p. ment (PP2) [21, 24, 27, 32–34]. In silico prediction sug-
Gly1264Alafs*38) led to prolongation of the MDR3 gested them to be disease-causing/ deleterious/possibly
molecule though the loss of the original stop codon. damaging/probably damaging, and with the involved
The canonical splicing-site variants c.136-2A > G and amino acid residues all being highly conserved among 20
c.80 + 1G > C disrupted the normal acceptor or donor primates (PP3). Moreover, the biochemical and clinical
splicing site in the ABCB4 introns 2 and 3, respectively. presentations of the ten patients were quite specific for
According to the ACMG standards and guidelines, the PFIC3 (PP4). Although in vitro or in vivo functional anal-
six novel variants above were all null ABCB4 variants, ysis was not performed due to technical limitation, the
and was diagnosed to be pathogenic, with the relevant evidences above rendered the eight novel missense vari-
evidences listed in Table 2. ants all “likely pathogenic” and supported the diagnosis
The remaining eight novel missense variants were of PFIC3 in the patients, since according to the ACMG
all absent or included rather rarely included in pub- standards, a variant classified as likely pathogenic typi-
lic databases (PM2), and among them, c.2782A > G(p. cally has sufficient evidence that a health-care provider
Arg928GlyA), c.1645C > T(p.Arg549Cys), c.716C > T(p. can use the molecular testing information in clinical
Table 2 Novel ABCB4 variants and the pathogenicity classification
Patient Nucleotide changes Amino acid changes Variant types ACMG evidences ACMG classification Allele Frequency In silico verdict
No
Chen et al. Orphanet Journal of Rare Diseases

1000 gnomAD Mutation PolyPhen-2 PROVEAN HSF NNsplice

Genomes Taster

1 c.2782A > G p.Arg928Gly Missense PM2 + PM3 + PP1-4 Likely pathogenic NI NI D ProbD D – –
1 c.136-2A > G – Splicing PVS1 + PM2 + PP1 + PP3 + PP4 Pathogenic NI NI – – – S S
2/3 c.1645C > T p.Arg549Cys Missense Likely pathogenic NI NI D ProbD D – –
(2022) 17:445

PM2 + PM3 + PP1-4

4 c.1801G > A p.Ala601Thr Missense PM2 + PP1-4 Likely pathogenic NI 0.007962‰ D PossD D – –
4 c.1406G > A p.Arg469Lys Missense PM2 + PP1-4 Likely pathogenic NI NI D ProbD D – –
7 c.3100_3101insA p.Ile1034Asnfs*4 Frameshift PVS1 + PM2 + PP1 + PP4 Pathogenic NI NI – – – –
7 c.80 + 1G > C – Splicing PVS1 + PM2 + PP1 + PP3 + PP4 Pathogenic NI NI – – – S S
8 c.716C > T p.Ser239Leu Missense PM2 + PM3 + PP1-4 Likely pathogenic NI NI D ProbD D – –
9 c.3230C > T p.Thr1077Met Missense PM2 + PM3 + PP1-4 Likely pathogenic NI 0.1991‰ D ProbD D – –
10 c.2914G > A p.Asp972Asn Missense PM2 + PP1-4 Likely pathogenic NI NI D B N – –
10 c.965 T > C p.Leu322Pro Missense PM2 + PP1-4 Likely pathogenic NI NI D ProbD D – –
12 c.2123G > A p.Trp708* Nonsense PVS1 + PM2 + PP1 + PP4 Pathogenic NI NI – – – – –
12 c.3789delA p.Gly1264Alafs*38 Frameshift PVS1 + PM2 + PP1 + PP4 Pathogenic NI NI D – – – –
13 c.879dupA p.Ala294Serfs*62 Frameshift PVS1 + PM2 + PP1 + PP4 Pathogenic NI NI – – – – –

According to the ACMG criteria [35]: PVS1, null variant (nonsense, frameshift, canonical ± 1 or 2 ss, etc.) in a gene where loss of function is a known mechanism of disease; PM2, absent from controls (or at extremely low
frequency if recessive) absent from controls in 1000 Genomes Project, or gnomAD; PM3, for recessive disorders, detected in trans with a pathogenic variant; PP1, co-segregation with disease in multiple affected family
members in a gene definitively known to cause the disease; PP2, missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease; PP3,
multiple lines of computational evidence support a deleterious effect on the gene or gene product; PP4, patient’s phenotype or family history is highly specific for a disease with a single genetic etiology; -, not applicable;
NI not included; D disease causing/deleterious; PossD possibly damaging; ProbD probably damaging; B benign; N neutral/not affecting; S splicing potential alteration of splicing, HSF human splicing finder
Page 8 of 12
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 9 of 12

Fig. 3 The distribution of the 96 ABCB4 variants detected in the 95 PFIC3 patients in this study. The novel ABCB4 variants identified in this study
were marked with asterisks. E2-E28 represented the 27 encoding ABCB4 exons

decision making when combined with other evidence of biliary epithelial cells [40]. Due to the impaired MDR3
the disease in question [35]. function, PFIC3 patients lacked phosphatidylcholine in
The MDR3 protein was primarily expressed in the liver, the bile to form micelles, and thus the very detergent
functioning as a floppase that translocated specifically bile liberated GGT from the canalicular membrane, giv-
phosphatidylcholine from the inner to the outer leaflet ing rise to cholangitis with high serum GGT activity [41,
of the hepatocytes canalicular membranes [36]. Phos- 42].
phatidylcholine was solubilized by canalicular bile salts to At present, medical treatment was the first line of ther-
form mixed micelles, therefore protecting the biliary tree apy offered to PFIC3 patient [9, 43], and the major goal of
from exposure to toxic and detergent effects of bile salts medical treatment was to relieve symptoms, improve the
[37]. The ABCB4 variants in this study impaired the flop- nutritional status, and to treat or prevent complications
pase function of the MDR3 protein, and thus the deple- due to cirrhosis and portal hypertension [38]. UDCA
tion of phosphatidylcholine and elevation of hydrophobic was the most common medicine in patients with PFIC3
bile acids in the biliary tubules damaged the integrity of [44], and the PFIC3 patients with residual phosphatidyl-
the cholangiocyte membrane, leading to the development choline secretion and MDR3 expression, especially those
of intrahepatic cholestasis and presenting as hepatomeg- with missense variants, responded to UDCA in 70% of
aly, pruritus, splenomegaly, jaundice and portal hyper- cases [45], and even in those with cirrhosis, UDCA could
tension. Besides, low phosphatidylcholine levels would be delay PFIC3 progression [12]. In this study, 66.1% (39/59)
expected to destabilize micelles and promote lithogenic- of the patients had positive responses to oral UDCA, of
ity of bile with crystallization of cholesterol, which could whom 38.5% (15/39) fully recovered in terms of the labo-
facilitate liver damage by obstructing small bile ducts ratory changes. This was not surprising since UDCA had
[38]. This could explain the occurrence of gallstones in a multiple mechanisms of action in cholestatic disorder
small number of children with PFIC3 as shown in Addi- including protection of cholangiocytes against cytotoxic-
tional file 2: Table S1. ity of hydrophobic bile acids, stimulation of hepatobiliary
In this study, all PFIC3 patients exhibited increased secretion of hydrophobic bile acids, inhibition of liver
serum GGT levels, constituting a biochemistry hallmark cell apoptosis, as well as anti-inflammation and immu-
of this condition. Serum GGT was deemed to be mainly nomodulation [46, 47].
of hepatobiliary origin and has been used as a “liver test” Although most PFIC3 patients showed varying degrees
for decades [39]. The reasons for elevated GGT values in of improvement in response to UDCA therapy, unfa-
those with hepatobiliary disease included de novo syn- vorable prognosis was observed in some cases. Actually,
thesis, release of membrane-bound GGT (by detergent this condition was progressive in the majority of affected
effects of bile acids), regurgitation of bile into the blood patients, and carried a high risk of developing cirrhosis
stream, and change in permeability or destruction of and liver failure during the first 2 decades of life [48]. In
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 10 of 12

this study, 41.1% (37/90) of PFIC3 patients had a poor Supplementary Information
prognosis, including 7 died, 27 having undergone while The online version contains supplementary material available at https://​doi.​
another 3 waiting for liver transplantation. On the last org/​10.​1186/​s13023-​022-​02597-y.
follow up, the condition remained stable in 53 patients
Additional file 1. Figure S1. ABCB4 genotypes of the 12 unrelated families
(58.9%, 53/90), but their long-term prognoses were on Sangersequencing or next generation sequencing. Arrows indicated
still uncertain, which needed to be followed up. So far, the mutations. Since Sangervalidation through forward sequencing or
like other end-stage liver disease, liver transplantation reverse sequencing, the base of the peak map maybe the reverse comple‑
men tation sequence of the base detected.
remained the last resort in patients unresponsive to med-
Additional file 2. Table S1. Clinical and molecular genetic data of previ‑
ical treatment [49]. Nevertheless, the lack of donor liver ously reported 82 PFIC3 patients.
organ and lifelong burden of immunosuppressive therapy
restricted the treatment option for this devastating con-
Acknowledgements
dition [50]. We sincerely thank the families of the patients for their cooperation and
ABCB4 variants exhibited remarkable heterogene- participation in this study.
ity, and the extent to which they impaired MDR3 flop-
Author contributions
pase activity determined the course and outcome of RC,YZS, YFT, FXY performed data collection of all patients; RC drafted the
the PFIC3 patients [21]. Depending on whether they original manuscript; FXY helped conduct the literature review; YZS conceptu‑
affected the traffic, activity, or stability of the protein, alized and designed the study, critically reviewed and revised the manuscript;
MD and HL conducted variant interpretation and reviewed the manuscript;
ABCB4 variants could be classified as follows: (I) defec- YZS, WXOY, YX managed and followed up the pediatric patients. All authors
tive synthesis, mainly nonsense and frameshift vari- contributed to manuscript revision, read and approved the submitted version.
ants, (II) affect protein maturation, (III) with little or All authors read and approved the final manuscript.
no effect on protein maturation but defective proteins, Funding
(IV) affect the stability and (V) variants without detect- This study got financial support from the National Natural Science Founda‑
able effects, providing the strong basis for the develop- tion of China (No. 81974057), the Clinical Frontier Technology Program of
the First Affiliated Hospital of Jinan University, China (No. JNU1AF-CFTP-
ment of genotype-based therapies for PFIC3 [24]. This 2022-a01228) and Science and Technology Plan Project of Guangzhou city
study found that PFIC3 patients with biallelic null vari- (No. 202201020088).
ants exhibited earlier onset ages, lower UDCA response
Availability of data and materials
rate and more unpromising clinical outcomes, which The datasets generated and analyzed for this study were available from the
clearly indicated that null ABCB4 variants were asso- corresponding author on reasonable request.
ciated with severer PFIC3 phenotypes. Understanding
the genotype–phenotype correlation contributed to the Declarations
prediction of prognosis and provided additional guid-
Ethics approval and consent to participate
ance to physicians and patients about the likely disease This study was approved by the Medical Ethics Committee of the First Affili‑
course [15]. ated Hospital, Jinan University, and written informed consents were signed by
the parents of all patients before this study.

Consent for publication

Conclusions Not applicable.
PFIC3 presented with hepatomegaly, pruritus, spleno-
Competing interests
megaly and jaundice with increased serum GGT level The authors declare that the research was conducted in the absence of any
as a biochemistry hallmark. Although varying degrees commercial or financial relationships that could be construed as a potential
of improvement in response to UDCA therapy were conflict of interest.

observed, 41.1% of PFIC3 patients exhibited unfavora- Author details

ble prognosis. ABCB4 genotypes of biallelic null variants 1
Department of Pediatrics, The First Affiliated Hospital, Jinan University,
were associated with severer PFIC3 phenotypes. Moreo- Guangzhou 510630, China. 2 Department of Infectious Diseases, Guangzhou
Women and Children’s Medical Center, Guangzhou 510120, China. 3 Depart‑
ver, the 14 novel variants in this study expanded the ment of Hepatopathy, Hunan Children’s Hospital, Changsha 410007, China.
ABCB4 variant mutation spectrum, and provided novel
molecular biomarkers for the definite diagnosis of PFIC3 Received: 1 August 2022 Accepted: 20 November 2022

patients.

Abbreviations References
PFIC3: Progressive familial intrahepatic cholestasis type 3; GGT​: Gamma-gluta‑ 1. Davit-Spraul A, Gonzales E, Baussan C, Jacquemin E. Progressive familial
myl transpeptidase; UDCA: Ursodeoxycholic acid; MDR3: Multi-drug resistant intrahepatic cholestasis. Orphanet J Rare Dis. 2009;4:1.
3; NGS: Next-generation sequencing; ACMG: American College of Medical 2. Jacquemin E. Progressive familial intrahepatic cholestasis. Clin Res Hepa‑
Genetics and Genomics; ALT: Aspartate transaminase; AST: Alanine transami‑ tol Gastroenterol. 2012;36(Suppl 1):S26–35.
nase; TBil: Total bilirubin; DBil: Direct bilirubin; TBA: Total bile acids.
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 11 of 12

3. Martínez-García J, Molina A, González-Aseguinolaza G, Weber ND, Smer‑ diagnosis and response to treatment. Pediatr Gastroenterol Hepatol Nutr.
dou C. Gene therapy for acquired and genetic cholestasis. Biomedicines. 2015;18(3):202–8.
2022;10(6):1238. 24. Delaunay JL, Durand-Schneider AM, Dossier C, Falguières T, Gautherot
4. Alam S, Lal BB. Recent updates on progressive familial intrahepatic J, Davit-Spraul A, et al. A functional classification of ABCB4 variations
cholestasis types 1, 2 and 3. Outcome and therapeutic strategies. World J causing progressive familial intrahepatic cholestasis type 3. Hepatology.
Hepatol. 2022;14(1):98–118. 2016;63(5):1620–31.
5. Falguières T, Aït-Slimane T, Housset C, Maurice M. ABCB4. Insights 25. Deng M, Guo L, Song Y. clinical and genetic analysis of a family affected
from pathobiology into therapy. Clin Res Hepatol Gastroenterol. by progressive familial intraphepatic cholestasis type 3. Zhonghua Yi Xue
2014;38(5):557–63. Yi Chuan Xue Za Zhi. 2018;35(5):686–90.
6. Oude Elferink RPJ, Paulusma CC. Function and pathophysiologi‑ 26. Li A, Dong Y, Xu Z, Wang F, Wang L, Yan J, et al. Progressive familial intra‑
cal importance of ABCB4 (MDR3 P-glycoprotein). Pflugers Arch. hepatic cholestasis type 3: a report of two cases in one pedigree. J Clin
2007;453(5):601–10. Hepatol. 2020;36(7):1601–4.
7. Bull LN, Thompson RJ. Progressive familial intrahepatic cholestasis. Clin 27. Saleem K, Cui Q, Zaib T, Zhu S, Qin Q, Wang Y, et al. Evaluation of a novel
Liver Dis. 2018;22(4):657–69. missense mutation in gene causing progressive familial intrahepatic
8. Venot Q, Delaunay JL, Fouassier L, Delautier D, Falguières T, Housset C, cholestasis type 3. Dis Mark. 2020;2020:6292818.
et al. A PDZ-like motif in the biliary transporter ABCB4 interacts with the 28. Zhang W, Lin R, Lu Z, Sheng H, Xu Y, Li X, et al. Phenotypic and molecular
scaffold protein EBP50 and regulates ABCB4 cell surface expression. PLoS characteristics of children with progressive familial intrahepatic cholesta‑
ONE. 2016;11(1): e0146962. sis in South China. Pediatr Gastroenterol Hepatol Nutr. 2020;23(6):558–66.
9. Srivastava A. Progressive familial intrahepatic cholestasis. J Clin Exp Hepa‑ 29. Bai J, Li L, Liu H, Liu S, Bai L, Ning H, et al. A novel compound heterozy‑
tol. 2014;4(1):25–36. gous mutation in ABCB4 gene in a pedigree with progressive familial
10. Degiorgio D, Colombo C, Seia M, Porcaro L, Costantino L, Zazzeron L, et al. intrahepatic cholestasis 3: a case report. Ann Transl Med. 2021;9(5):426.
Molecular characterization and structural implications of 25 new ABCB4 30. Lipiński P, Ciara E, Jurkiewicz D, Płoski R, Wawrzynowicz-Syczewska M,
mutations in progressive familial intrahepatic cholestasis type 3 (PFIC3). Pawłowska J, et al. Progressive familial intrahepatic cholestasis type 3:
Eur J Hum Genet. 2007;15(12):1230–8. report of four clinical cases, novel ABCB4 variants and long-term follow-
11. Jacquemin E, De Vree JM, Cresteil D, Sokal EM, Sturm E, Dumont M, up. Ann Hepatol. 2021;25: 100342.
et al. The wide spectrum of multidrug resistance 3 deficiency: from 31. Küçükçongar Yavaş A, Çavdarlı B, Ünal Uzun Ö, Uncuoğlu A, Gündüz M.
neonatal cholestasis to cirrhosis of adulthood. Gastroenterology. A novel etiologic factor of highly elevated cholestanol levels: pro‑
2001;120(6):1448–58. gressive familial intrahepatic cholestasis. J Pediatr Endocrinol Metab.
12. Fang LJ, Wang XH, Knisely AS, Yu H, Lu Y, Liu LY, et al. Chinese children 2020;33(5):665–9.
with chronic intrahepatic cholestasis and high γ-glutamyl transpeptidase: 32. Degiorgio D, Corsetto PA, Rizzo AM, Colombo C, Seia M, Costantino L,
clinical features and association with ABCB4 mutations. J Pediatr Gastro‑ et al. Two ABCB4 point mutations of strategic NBD-motifs do not prevent
enterol Nutr. 2012;55(2):150–6. protein targeting to the plasma membrane but promote MDR3 dysfunc‑
13. Khabou B, Mahjoub B, Barbu V, Balhoudi N, Wardani A, Sfar MT, et al. tion. Eur J Hum Genet. 2014;22(5):633–9.
Phenotypic variability in Tunisian PFIC3 patients harboring a complex 33. Morita SY, Kobayashi A, Takanezawa Y, Kioka N, Handa T, Arai H, et al. Bile
genotype with a differential clinical outcome of UDCA treatment. Clin salt-dependent efflux of cellular phospholipids mediated by ATP binding
Chim Acta. 2018;486:122–8. cassette protein B4. Hepatology. 2007;46(1):188–99.
14. Schatz SB, Jüngst C, Keitel-Anselmo V, Kubitz R, Becker C, Gerner P, 34. Park HJ, Kim TH, Kim SW, Noh SH, Cho KJ, Choi C, et al. Functional charac‑
et al. Phenotypic spectrum and diagnostic pitfalls of ABCB4 deficiency terization of ABCB4 mutations found in progressive familial intrahepatic
depending on age of onset. Hepatol Commun. 2018;2(5):504–14. cholestasis type 3. Sci Rep. 2016;6:26872.
15. Al-Hussaini A, Lone K, Bashir MS, Alrashidi S, Fagih M, Alanazi A, et al. 35. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards
ATP8B1, ABCB11, and ABCB4 genes defects: novel mutations associated and guidelines for the interpretation of sequence variants: a joint con‑
with cholestasis with different phenotypes and outcomes. J Pediatr. sensus recommendation of the American College of Medical Genetics
2021;236:113-23.e2. and Genomics and the Association for Molecular Pathology. Genet Med.
16. Chen HL, Chang PS, Hsu HC, Lee JH, Ni YH, Hsu HY, et al. Progressive 2015;17(5):405–24.
familial intrahepatic cholestasis with high gamma-glutamyltranspepti‑ 36. Prescher M, Kroll T, Schmitt L. ABCB4/MDR3 in health and dis‑
dase levels in Taiwanese infants: role of MDR3 gene defect? Pediatr Res. ease–at the crossroads of biochemistry and medicine. Biol Chem.
2001;50(1):50–5. 2019;400(10):1245–59.
17. Giovannoni I, Santorelli FM, Candusso M, Di Rocco M, Bellacchio E, 37. Delaunay JL, Bruneau A, Hoffmann B, Durand-Schneider AM, Barbu V,
Callea F, et al. Two novel mutations in African and Asian children with Jacquemin E, et al. Functional defect of variants in the adenosine triphos‑
progressive familial intrahepatic cholestasis type 3. Dig Liver Dis. phate–binding sites of ABCB4 and their rescue by the cystic fibrosis
2011;43(7):567–70. transmembrane conductance regulator potentiator, ivacaftor (VX-770).
18. Kubitz R, Bode J, Erhardt A, Graf D, Kircheis G, Müller-Stöver I, et al. Chole‑ Hepatology. 2017;65(2):560–70.
static liver diseases from child to adult: the diversity of MDR3 disease. Z 38. Stättermayer AF, Halilbasic E, Wrba F, Ferenci P, Trauner M. Variants in
Gastroenterol. 2011;49(6):728–36. ABCB4 (MDR3) across the spectrum of cholestatic liver diseases in adults.
19. Dzagania T, Engelmann G, Häussinger D, Schmitt L, Flechtenmacher C, J Hepatol. 2020;73(3):651–63.
Rtskhiladze I, et al. The histidine-loop is essential for transport activity of 39. Lonardo A, Romagnoli D. Gamma glutamyl transferase: a novel cardiovas‑
human MDR3. A novel mutation of MDR3 in a patient with progressive cular outfit for an old liver test. Indian J Med Res. 2016;143(1):4–7.
familial intrahepatic cholestasis type 3. Gene. 2012;506(1):141–5. 40. Sotil EU, Jensen DM. Serum enzymes associated with cholestasis. Clin
20. Ramraj R, Finegold MJ, Karpen SJ. Progressive familial intrahepatic chol‑ Liver Dis. 2004;8(1):41–54.
estasis type 3: overlapping presentation with Wilson disease. Clin Pediatr 41. van Mil SW, Houwen RH, Klomp LW. Genetics of familial intrahepatic
(Phila). 2012;51(7):689–91. cholestasis syndromes. J Med Genet. 2005;42(6):449–63.
21. Gordo-Gilart R, Andueza S, Hierro L, Martínez-Fernández P, D’Agostino 42. Davit-Spraul A, Gonzales E, Baussan C, Jacquemin E. The spectrum of liver
D, Jara P, et al. Functional analysis of ABCB4 mutations relates clinical diseases related to ABCB4 gene mutations: pathophysiology and clinical
outcomes of progressive familial intrahepatic cholestasis type 3 to the aspects. Semin Liver Dis. 2010;30(2):134–46.
degree of MDR3 floppase activity. Gut. 2015;64(1):147–55. 43. van der Woerd WL, Houwen RH, van de Graaf SF. Current and future
22. Sun HZ, Shi H, Zhang SC, Shen XZ. Novel mutation in a Chinese patient therapies for inherited cholestatic liver diseases. World J Gastroenterol.
with progressive familial intrahepatic cholestasis type 3. World J Gastro‑ 2017;23(5):763–75.
enterol. 2015;21(2):699–703. 44. Chen HL, Wu SH, Hsu SH, Liou BY, Chen HL, Chang MH. Jaundice revisited:
23. Boga S, Jain D, Schilsky ML. Presentation of progressive familial intrahe‑ recent advances in the diagnosis and treatment of inherited cholestatic
patic cholestasis type 3 mimicking Wilson disease: molecular genetic liver diseases. J Biomed Sci. 2018;25(1):75.
Chen et al. Orphanet Journal of Rare Diseases (2022) 17:445 Page 12 of 12

45. Vitale G, Gitto S, Vukotic R, Raimondi F, Andreone P. Familial intrahepatic

cholestasis: new and wide perspectives. Dig Liver Dis. 2019;51(7):922–33.
46. Jhaveri MA, Kowdley KV. New developments in the treatment of pri‑
mary biliary cholangitis–role of obeticholic acid. Ther Clin Risk Manag.
2017;13:1053–60.
47. Manley S, Ding W. Role of farnesoid X receptor and bile acids in alcoholic
liver disease. Acta Pharm Sin B. 2015;5(2):158–67.
48. Colombo C, Vajro P, Degiorgio D, Coviello DA, Costantino L, Tornillo L,
et al. Clinical features and genotype-phenotype correlations in children
with progressive familial intrahepatic cholestasis type 3 related to ABCB4
mutations. J Pediatr Gastroenterol Nutr. 2011;52(1):73–83.
49. Englert C, Grabhorn E, Richter A, Rogiers X, Burdelski M, Ganschow R.
Liver transplantation in children with progressive familial intrahepatic
cholestasis. Transplantation. 2007;84(10):1361–3.
50. Wei G, Cao J, Huang P, An P, Badlani D, Vaid KA, et al. Synthetic human
ABCB4 mRNA therapy rescues severe liver disease phenotype in a BALB/c.
Abcb4 mouse model of PFIC3. J Hepatol. 2021;74(6):1416–28.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in pub‑
lished maps and institutional affiliations.

Ready to submit your research ? Choose BMC and benefit from:

• fast, convenient online submission

• thorough peer review by experienced researchers in your field
• rapid publication on acceptance
• support for research data, including large and complex data types
• gold Open Access which fosters wider collaboration and increased citations
• maximum visibility for your research: over 100M website views per year

At BMC, research is always in progress.

Learn more biomedcentral.com/submissions