Abstract
Background
Intrahepatic cholestasis of pregnancy (ICP) is characterised by troublesome maternal pruritus, raised serum bile acid levels and increased fetal risk. Mutations of the ABCB4 gene encoding the hepatobiliary phospholipid transporter have been identified in a small proportion of patients with cholestasis of pregnancy. In a recent prospective study on 693 patients with cholestasis of pregnancy, a cut‐off level for serum bile acid (⩾40 μmol/l) was determined for increased risk of fetal complications.
Objectives
To investigate whether common combinations of polymorphic alleles (haplotypes) of the genes encoding the hepatobiliary ATP‐binding cassette (ABC) transporters for phospholipids (ABCB4) and bile acids (ABCB11) were associated with this severe form of cholestasis of pregnancy.
Methods
For genetic analysis, 52 women with bile acid levels ⩾40 μmol/l (called cases) and 52 unaffected women (called controls) matched for age, parity and geographical residence were studied. Gene variants tagging common ABCB4 and ABCB11 haplotypes were genotyped and haplotype distributions were compared between cases and controls by permutation testing.
Results
In contrast with ABCB11 haplotypes, ABCB4 haplotypes differed between the two groups (p = 0.019), showing that the severe form of cholestasis of pregnancy is associated with the ABCB4 gene variants. Specifically, haplotype ABCB4_5 occurred more often in cases, whereas haplotypes ABCB4_3 and ABCB4_7 were more common in controls. These associations were reflected by different frequencies of at‐risk alleles of the two tagging polymorphisms (c.711A: odds ratio (OR) 2.27, p = 0.04; deletion intron 5: OR 14.68, p = 0.012).
Conclusion
Variants of ABCB4 represent genetic risk factors for the severe form of ICP in Sweden.
Intrahepatic cholestasis of pregnancy (ICP) is a reversible cholestatic liver disease of undefined aetiology and pathogenesis. It is characterised by pruritus and raised serum bile acid levels (⩾10 μmol/l), with onset in the second half of pregnancy and persisting until delivery.1,2,3 Although essentially benign from a maternal viewpoint, ICP is associated with increased fetal risks, such as preterm delivery in 19–60%4,5,6 and fetal distress in 22–41% of affected pregnancies.4,7,8 Intrauterine fetal death has been reported in 0.8–1.6% of ICP cases,4,5,7,8 but was detected in only 0.4% of women in a recent prospective study on ICP.9 Fetal complications were correlated with maternal serum bile acid levels, increasing markedly when levels exceeded 40 μmol/l.9
The aetiology and pathogenesis of ICP is still unknown. Family clustering and varying incidence in different geographical regions suggest that genetic and environmental factors and their interactions probably play a part.1 The hypothesis that genetically determined dysfunction of hepatocanalicular lipid transporters contributes to ICP was first supported by Jacquemin et al,10 who reported the coexistence of ICP and a frameshift mutation in the ABCB4 gene encoding the hepatobiliary transporter for phosphatidylcholine in a large consanguineous family. Since then, additional variants of the ABCB4 gene have been identified in a small number of patients with ICP.11,12,13,14 Also, single British and Finnish patients with ICP carried mutations in the ATP8B1 gene encoding a potential membrane transporter for phosphatidylserine in the liver and intestine.15,16
The biliary secretion of bile acids and other biliary lipid constituents across the canalicular membrane of hepatocytes is mediated by a set of distinct ATP‐binding cassette (ABC) transporters. The bile acid export pump ABCB11 and the phospholipid transporter ABCB4 are key players in this process. ABCB4 translocates the major biliary phospholipid, phosphatidylcholine (lecithin), from the inner to the outer leaflet of the canalicular membrane, and ABCB11 is the predominant bile acid efflux system of hepatocytes. Further, inherited deficiencies of ABCB11 and ABCB4 have been established as causes of progressive familial intrahepatic cholestasis type 2 and type 3, respectively.17,18 These are chronic liver diseases that may rapidly progress to cirrhosis and often necessitate liver transplantation in early childhood. ABCB4 dysfunction is assumed to decrease biliary phospholipid levels, thereby causing toxic membrane damage via an excess of biliary bile acids, whereas ABCB11 deficiency causes intrahepatic retention of bile acids.
From a prospective cohort study on 693 women with ICP, we selected a subgroup with severe ICP (n = 52). In these patients, we studied whether common sets of polymorphisms (haplotypes) covering most of the genetic variation in the hepatobiliary transporter genes ABCB4 and ABCB11 differed from haplotypes observed in healthy controls.
Patients and methods
The cohort with ICP and controls
Between 1 February 1999 and 31 January 2002, all pregnant women in the Västra Götaland region of Sweden were prospectively screened for ICP, defined as otherwise unexplained pruritus of pregnancy combined with fasting bile acid levels ⩾10 μmol/l. The epidemiological results of this study have been published recently.9 In Sweden, all pregnancies are monitored by midwives in local antenatal clinics. All 106 antenatal clinics and the six departments of obstetrics in the region participated in the study.
In all, 45 485 pregnancies leading to deliveries were recorded during the study period, and 937 women (2.1%) complained of generalised pruritus during pregnancy, 693 (1.5%) of whom had ICP. On the basis of prospectively recorded fetal complication rates, an increased fetal risk was observed exclusively in patients with bile acid levels ⩾40 μmol/l.9 Therefore, we defined a mild and a severe form of ICP, the women with severe ICP had bile acid levels ⩾40 μmol/l.9 Women with severe ICP (n = 130; 96 in the observational study9 and 34 in the treatment study19) were invited to participate in the genetic study. Forty eight women who accepted and all additional patients (n = 4) who presented with severe ICP shortly after the end of these two studies were enrolled in the genetic study after they gave written informed consent; no more than these four additional patients were available for haplotype analysis. For control purposes, 52 healthy women without ICP were recruited. All controls reported uneventful pregnancies and were matched with respect to age, parity and geographical residence.
The study was approved by the ethics committee of the Faculty of Medicine at the University of Göteborg, Göteborg, Sweden.
Liver chemistry tests
Blood samples were obtained after a 12‐h fast. Total serum bile acid concentrations were analysed with an enzymatic colorimetric method (Enzabile, Biostat Diagnostic Systems, Stockport, UK). Activities of serum transaminases and γ‐glutamyl transpeptidase, and serum bilirubin levels were analysed with standard laboratory methods.9
Genotyping
Genomic DNA was isolated from EDTA‐anticoagulated blood using the QIAamp (Qiagen, Hilden, Germany) protocol. DNA concentrations were determined fluorometrically (Bio‐Rad Laboratories, Hercules, California, USA), using the dye PicoGreen (Molecular Probes, Leiden, The Netherlands).
In general, single‐nucleotide polymorphisms (SNPs) were genotyped using solution‐phase hybridisation reactions with 5′ nuclease and fluorescence detection (TaqMan assays) on the ABI PRISM 7000 Sequence Detection System (Applera, Norwalk, Connecticut, USA). Primer and probe sequences for allele‐specific polymerase chain reactions are available at http://www.appliedbiosystems.com. Polymerase chain reactions contained 5–20 ng genomic DNA, 1× TaqMan Universal Master Mix, 900 nM of each primer and 200 nM of VIC‐labelled and FAM‐labelled probes in 25 μl‐reactions (Applera). Amplification conditions were 95°C for 10 min, 40 cycles at 92°C for 15 s and 60°C for 1 min.
We ascertained the results of TaqMan assays by direct Big Dye Termination cycle sequencing on the ABI PRISM 310 Genetic Analyzer (Applera). The 5‐bp deletion in intron 5 of the ABCB4 gene20 was visualised on silver‐stained polyacrylamide gels.21
Genotype analysis
Consistency in genotype frequencies with Hardy–Weinberg equilibrium was confirmed using an exact test22 and visualised in triangular de Finetti diagrams,23 using software by T Wienker and T Strom (http://ihg.gsf.de/snps.html).
To test for association, we compared the distribution of alleles and genotypes in contingency tables by χ2 test and Armitage's trend test, respectively.24
Haplotype analysis
Haplotypes are defined as particular combinations of SNP alleles present on single chromosomes. The haplotype structure of the human ABCB4 and ABCB11 genes has been defined by Pauli‐Magnus et al,20 on the basis of resequencing of the genes in 56 and 93 healthy Caucasians, respectively. We identified the minimum sets of SNPs tagging the common haplotypes (frequencies >3%) of both genes using the Best Enumeration of SNP Tags algorithm.25
Haplotypes were reconstructed from the genotype data of cases and controls using the PHASE algorithm v.2.0, which is based on bayesian inference.26 To allow referral to specific haplotypes, we have followed the nomenclature used by Pauli‐Magnus et al20 (ie, ABCB11_1, ABCB11_2, ...), irrespective of the actual haplotype frequencies observed in our study.
To test for significant differences in haplotype distributions between patients with ICP (hereafter called cases) and controls, permutation tests were performed, as implemented in the PHASE algorithm.26 The permutation test checks the null hypothesis that case and control haplotypes are a random sample from a single set of haplotype frequencies versus the alternative that cases are more similar to each other than to controls. Haplotypes of cases and controls were permuted 10 000 times, which generates empirical p values at the 0.05 level while controlling for multiple testing.
Additional statistical analysis
Means, standard deviations (SDs), standard errors (SE) medians and ranges were calculated as indicated. Student's t test was used to compare differences in continuous variables between groups. Dichotomous variables were compared using Fisher's exact test. All tests were conducted at a significance level of 0.05.
Results
Clinical characteristics of cases
We included 52 cases with serum bile acid concentrations ⩾40 µmol/l in our study. The mean (SD) age of the patients was 30.1 (5.1) years. Heredity for pruritus of pregnancy was reported by 19 (37%) women. The parous women (n = 24) had experienced a total of 36 pregnancies, of which 27 (75%) had been associated with pruritus of pregnancy. Intrauterine fetal deaths had occurred in 2/36 pregnancies (5.6%), both associated with pruritus. Gallstone disease (cholecystectomy or symptomatic, ultrasound‐verified gallstones) was reported by 5 (9.6%) women.
In our study, the 52 women developed symptomatic ICP after a mean of 29.1 (SD 6.4) weeks of gestation. The mean serum bile acid concentration was 111 (SD 64) µmol/l. Bile acid levels were correlated significantly with bilirubin levels (r = 0.38, p = 0.005) and alanine aminotransferase activities (r = 0.51, p<0.001). Serum bilirubin levels and γ‐glutamyl transpeptidase activities were raised in 44% and 20% of the patients, respectively. Spontaneous premature birth in singleton pregnancies occurred in 27% of the pregnancies, whereas the total prematurity rate (including inductions, preterm caesarean sections or multiple pregnancies) was 33%. Asphyxial events (operative delivery due to asphyxia, umbilical arterial pH <7.05 or Apgar Score <7 at five minutes) occurred in 13% of the pregnancies. Meconium staining of amniotic fluid and green staining of placenta and membranes (indicating a longer period of meconium passage) were registered in 60% and 44% of the cases, respectively.
Genetic analysis
The Best Enumeration of SNP Tags algorithm identified a minimum set of five polymorphisms tagging the common haplotypes (frequencies >3%) of the ABCB4 gene and six haplotype‐tagging SNPs covering the ABCB11 gene (tables 1 and 2). These tagging variants define six common ABCB4 and 11 common ABCB11 haploytpes with population frequencies >3% in our study. The tagging polymorphisms were subsequently genotyped in the 52 patients with severe ICP and 52 matched controls. None of the genotype frequencies deviated from the frequencies expected for a large, randomly mating population, as shown by accordance with an exact test for Hardy–Weinberg equilibrium.
Table 1 ABCB11 genotypes and haplotypes.
| Haplotypes | htSNP (codon position) | Haplotype frequencies* | ||||||
|---|---|---|---|---|---|---|---|---|
| Intron −1 (−1952) | Intron −1 (−1155) | Exon 13 (1331) | Intron 18 (−17) | Intron 19 (−17) | Exon 24 (3084) | Cases (%) | Controls (%) | |
| ABCB11_1 | T | T | C | A | C | G | 16.6 (1.6) | 16.9 (1.7) |
| ABCB11_2 | C | C | T | A | C | G | 17.2 (1.8) | 17.2 (2.3) |
| ABCB11_3 | C | C | T | C | T | A | 6.0 (1.6) | 7.4 (1.9) |
| ABCB11_4 | T | T | C | C | T | A | 11.6 (1.7) | 10.9 (1.3) |
| ABCB11_5 | T | T | C | C | C | A | 3.7 (1.2) | 5.8 (1.5) |
| ABCB11_6 | C | C | C | A | C | G | 7.7 (1.6) | 4.1 (1.4) |
| ABCB11_7 | T | C | C | A | C | G | 5.0 (1.4) | 6.5 (1.8) |
| ABCB11_9†‡ | C | C | C | C | T | A | 5.6 (1.3) | 5.4 (2.0) |
| ABCB11_10‡ | T | T | T | A | C | G | 6.1 (0.9) | 4.1 (0.9) |
| ABCB11_11‡ | T | C | C | C | T | A | 3.5 (1.2) | 6.5 (2.0) |
| ABCB11_12‡ | C | C | T | C | C | A | 3.7 (1.1) | 5.2 (1.4) |
Case; woman with intrahepatic cholestasis of pregnancy; control, healthy woman without intrahepatic cholestasis of pregnancy; htSNP, haplotype‐tagging single‐nucleotide polymorphism.
Values are mean (SE).
*Haplotype frequencies do not differ significantly between cases and controls (p = 0.23), as indicated by permutation tests with 10 000 permutations.26
†The haplotype ABCB11_8 is omitted, as it was observed at a frequency of <3% in both cases and controls.
‡As haplotypes ABCB11_9–ABCB11_12 were observed but not designated in the study by Pauli‐Magnus et al,20 numbers are assigned according to their frequencies in this study.
Table 2 ABCB4 genotypes and haplotypes.
| Haplotypes | htSNPs and deletions (codon position)* | Haplotype frequencies† | |||||
|---|---|---|---|---|---|---|---|
| Intron 5 del AGAAA (−66 to −62) | Exon 6 (504) | Exon 8 (711) | Exon 16 (1954) | Intron 26 (−16) | Cases (%) | Controls (%) | |
| ABCB4_1 | 0 | T | A | A | C | 45.5 (2.2) | 50.5 (1.2) |
| ABCB4_2 | 0 | C | A | A | C | 21.6 (2.0) | 17.5 (1.0) |
| ABCB4_3 | 0 | C | T* | A | C | 5.4 (0.8) | 8.7 (0.8) |
| ABCB4_4 | 0 | T | A | A | T* | 11.3 (2.1) | 8.0 (1.1) |
| ABCB4_5 | 1* | T | A | A | C | 5.8 (0.6) | 0.9 (0) |
| ABCB4_7‡ | 0 | C | T* | G* | C | 1.7 (0.2) | 10.2 (0.4) |
Case; woman with intrahepatic cholestasis of pregnancy; control, healthy woman without intrahepatic cholestasis of pregnancy; del, deletion; htSNP, haplotype‐tagging single‐nucleotide polymorphism.
Values are mean (SE).
*Asterisks indicate haplotype‐tagging alleles.
†Haplotype frequencies differ significantly between cases and controls (p = 0.019), as indicated by permutation tests with 10 000 permutations.26
‡The haplotype ABCB4_6 is omitted, as it was observed at a frequency of <3% in both cases and controls.
Table 1 shows that the distributions of ABCB11 haplotypes do not differ between cases and controls. However, ABCB4 haplotypes differ significantly between both groups (table 2), as confirmed by permutation testing (see Methods, p = 0.019, 10 000 permutations). The haplotype ABCB4_5 occurs more frequently in cases than in controls (5.8% v 0.9%), whereas haplotypes ABCB4_3 and ABCB4_7 are more common in controls (8.7% v 5.4% and 10.2% v 1.7%, respectively).
Table 3 shows that these associations are reflected by significantly different allele and genotype frequencies of the polymorphisms tagging the ABCB4 haplotypes 3, 5 and 7. For the haplotype‐tagging SNP c.711A→T, which tags haplotypes ABCB4_3 and ABCB4_7, allele frequencies differed significantly between cases and controls (c.711A: odds ratio 2.27, 95% confidence interval (CI) 1.04 to 4.96, p = 0.037). The common OR for c.711A→T genotypes was 2.07 (p = 0.047). Allele frequencies of the SNP c.1954A→G (tagging haplotype ABCB4_7 only) also differed between cases and controls (OR 3.98, 95% CI 1.08 to 14.72, p = 0.027). Of note, six cases were heterozygous carriers of the deletion in intron 5; this was not observed in any control in our study (OR 14.68, p = 0.012; table 3).
Table 3 Distributions of alleles and genotypes for selected ABCB4 tagging polymorphisms in 52 patients with intrahepatic cholestasis of pregnancy and 52 matched controls.
| ABCB4 tagging polymorphism | Number (frequency) of alleles or genotypes | Tests for association* | |||
|---|---|---|---|---|---|
| Cases (2n = 104) | Controls (2n = 104) | OR | 95% CI | p value | |
| c.711A | 93 (0.89) | 82 (0.79) | 2.27 | 1.04 to 4.96 | 0.037 |
| c.711T | 11 (0.11) | 22 (0.21) | |||
| c.711A/A | 42 (0.81) | 33 (0.63) | 2.07 | 0.047 | |
| c.711A/T | 9 (0.17) | 16 (0.31) | |||
| c.711T/T | 1 (0.02) | 3 (0.06) | |||
| c.1954A | 101 (0.97) | 93 (0.89) | 3.98 | 1.08 to 14.72 | 0.027 |
| c.1954G | 3 (0.03) | 11 (0.11) | |||
| c.1954A/A | 49 (0.94) | 42 (0.81) | 5.43 | 0.033 | |
| c.1954A/G | 3 (0.06) | 9 (0.17) | |||
| c.1954G/G | 0 (0) | 1 (0.02) | |||
| wt intron 5 | 46 (0.88) | 52 (1.00) | 14.68 | 0.81 to 267.66 | 0.012 |
| del intron 5 | 6 (0.12) | 0 (0) | |||
Case; woman with intrahepatic cholestasis of pregnancy; control, healthy woman without intrahepatic cholestasis of pregnancy; del, deletion; wt, wild type.
*OR, 95% CI and p values were determined using contingency table statistics (see Methods). ORs were calculated relative to low‐risk alleles (c.711T, c.1954G, no deletion in intron 5).
The supplementary figs A and B (provided online at http://www.gutjnl.com/supplemental) show de Finetti diagrams illustrating genotype and allele frequencies of the tagging polymorphisms. In these triangular diagrams, the frequencies of homozygotic genotypes are plotted using the left and right diagonal axes for homozygotes, whereas the frequencies of heterozygotes are shown on the left vertical axis. Consistency with Hardy–Weinberg equilibrium is indicated by the parabola, and allele frequencies are depicted by the intersection with the bottom perpendicular. For both ABCB4 polymorphisms, the positions within the triangular diagrams and the allele frequencies differ significantly between patients with severe ICP and controls.
Multiple regression analysis did not show any significant association between liver chemistry tests and ABCB4 genotypes; however, mean (SD) serum bile acids concentrations tended to be higher in patients carrying at least one at‐risk ABCB4 allele (146 (30) v 107 (9) μmol/l, p = 0.160). γ‐Glutamyl transpeptidase activities, on the other hand, did not differ. In our cohort, only 3 of 13 (23%) patients for whom serum γ‐glutamyl transpeptidase activities were available had raised levels and only one patient carried an at‐risk ABCB4 allele.
DNA sequence analyses showed no known mutations in exon 14 of the ABCB4 gene for any patient tested (not shown). In particular, the C→A transversion in codon 54611 and the single‐nucleotide deletion starting in codon 571 (1712delT) previously described in single cases10 were not detected, consistent with a previous report on Finnish patients with ICP.27
Discussion
ICP is associated with maternal pruritus and increased fetal risk.4,5,6,7,8,28 As the observational part of our ICP study showed that fetal complication rates did not increase until bile acid levels exceeded 40 µmol/l,9 we focused our genetic analysis on these high‐risk patients with severe ICP. This approach differs from previous studies on ICP genetics.10,11,12,13,14,29,30,31 Our strategy, studying well‐defined patients with more extreme phenotypes, is presumably more likely to identify clinically relevant susceptibility genes. Further, we recently showed that the subgroup of patients with severe ICP was more likely to benefit from treatment with ursodeoxycholic acid.19 Ursodeoxycholic acid improved pruritus and all biochemical markers of cholestasis in patients with severe ICP. Mean bile acid levels decreased by 79%, bilirubin concentrations by 50% and alanine aminotransferase activities by 80% (all p⩽0.01).19
It is still unclear whether environmental or genetic impairment of the bile‐secretory apparatus is a prerequisite for liver injury in the case of ICP. Bile secretion normally depends on the integrity of an ensemble of membrane transport systems in hepatocytes and cholangiocytes.1,32 The transport of all three major biliary lipids (ie, bile acids, phosphatidylcholine and cholesterol) across the canalicular membrane of hepatocytes into bile is mediated by ATP‐dependent export pumps, known as ABC transporters.33 Two major transporters are ABCB4 (formerly known as multidrug resistance gene 3, MDR3), which translocates phosphatidylcholine across the hepatocanalicular membrane,34 and the bile acid export pump ABCB11, which is the predominant canalicular efflux system for conjugated bile acids.35 Mutations in the genes encoding these hepatobiliary transporters have been identified as causative in progressive familial intrahepatic cholestasis in children.12,18 In the light of these findings, ABCB4 and ABCB11 represent promising candidate genes, which could also determine the degree of cholestatic injury in patients with ICP. Thus, we investigated common variants of both genes in patients with severe ICP and matched controls. The study design was based on the expectation that there would be higher frequencies of the contributing genetic variants in patients than in a group of matched controls without the disease.
Pauli‐Magnus et al20 have defined the complete haplotype structure of the ABCB4 and ABCB11 genes by sequencing the promoters, the coding regions and the flanking intronic parts of the corresponding genes in a large number of individuals of European descent. Genetic variants have arisen from single historical mutational events and are therefore non‐randomly associated with neighbouring alleles that were present on the particular chromosomal segments on which the mutation occurred (linkage disequilibrium). Haplotypes are defined as particular combinations of alleles that are inherited together on single chromosomes.36 It is noteworthy that only few (<10) common haplotypes are observed in whole chromosomal regions, and these account for most of the genetic variation among individuals. In complex diseases, haplotype analysis is generally expected to yield more information than analysis of individual polymorphisms,37 the rationale behind the design of the International HapMap Project that has released its first dataset recently.36
Owing to the block‐like structure of linkage disequilibrium and the substantial correlations of SNPs with many of their neighbours,36 the number of distinct combinations of SNP alleles encountered in human DNA samples is only a small fraction of the theoretical number of haplotypes if all alleles were distributed randomly. This haplotype structure is the basis for selection of a parsimonious set of SNPs, called haplotype‐tagging SNPs, which represent most of the haplotype variation in a population. In our study, we identified the tagging polymorphisms using the Best Enumeration of SNP Tags algorithm,25 which uses a powerful bayesian approach (PHASE) to reconstruct haplotypes,26 and compared haplotype frequencies between cases and controls by permutation tests, thereby accounting for multiple testing.26 Whereas ABCB11 haplotypes did not differ between cases and controls (table 1), the distribution of ABCB4 haplotypes differed significantly between cases and controls, as determined by permutation testing (p = 0.019; table 2). The at‐risk ABCB4 haplotypes were characterised by tagging gene variants (deletion in intron 5, c.711A→T; table 3). The deletion in intron 5 tagging haplotype ABCB4_5 was detected in 12% of the cases but not in controls, indicating a potentially much larger role for this ABCB4 mutation in the case of severe ICP than previously acknowledged.
Our findings confirm the association between ICP and the SNP c.711A→T detected previously in a large UK cohort of 184 patients with ICP with bile acid levels >14 μmol/l.29 Both studies independently confirm the association between ICP and the common allele of this ABCB4 SNP (designated 743A in the study by Müllenbach et al29). As noted previously,29 it is not unusual that the common allele represents the risk‐associated variant in complex diseases—for example, a common truncating splice‐site mutation has been associated with sarcoidosis recently.38 Similar frequency deviations for the SNP c.711A→T were observed in a recent smaller study comparing the ABCB4 sequence in 21 patients with ICP (bile acid levels >20 μmol/l) and 40 controls,13 but did not reach statistical significance. In the study by Pauli‐Magnus et al,13 the intron 5 deletion was detected in one patient with ICP and two controls, indicating that this polymorphism is not ICP specific, but study participants were not matched for age, parity or geographical residence.13 Although we observed markedly different allele and genotype frequencies for the SNP c.1954A→G in our ICP cases and controls, it has been assumed that the resulting missense mutation (R652G) is a neutral polymorphism that occurs with similar frequencies in cases and controls.12,17,29,39 However, in contrast with a French study,12 in our Swedish patients and in patients from Switzerland,13 ICP seemed to be more prevalent among carriers of arginine at codon 652, and sequence comparisons of ABCB4 genes identifies that the glycine residue is conserved in other species such as rat and hamster.
The increased prevalence of ABCB4 gene variants in ICP is consistent with the very strong association between ICP and gallstone disease in our observational study,9 which detected a 15‐fold increase in gallstone disease among patients with severe ICP. Notably, ABCB4 mutations have been linked to a symptomatic and recurrent form of gallbladder and intrahepatic cholesterol cholelithiasis.40,41
Our study could not exclude that the non‐severe form of ICP, which we defined by bile acid levels <40 μmol/l,9 might also be associated with ABCB4 haplotypes, but this group of patients was not available for genetic analysis. However, only exclusion of the genetic association in a large cohort of patients with mild ICP would unambiguously prove that ICP severity depends on genetic and not environmental factors.
The distributions of common ABCB11 haplotypes did not differ between cases and controls. Our findings and those of Pauli‐Magnus et al13 showed more genetic variability in ABCB4 than in ABCB11 in patients with ICP. Thus, our genetic analysis does not support the idea that common ABCB11 variants have a role in the pathogenesis of ICP. However, it is possible that as yet unidentified rare variants of the ABCB11 gene might play a part in ICP pathogenesis, as a recent study on 57 Finnish patients with ICP30 linked two intragenic ABCB11 SNPs to ICP. Our findings are in line with a recent Finnish family study,31 in which segregation of haplotypes and multipoint linkage analysis excluded ABCB11 as being responsible for the pathogenesis of ICP in two families with dominant inheritance. ABCB11 haplotype structure has also been investigated in patients with primary biliary cirrhosis, primary sclerosing cholangitis20 or gallstone disease, without any specific findings.42
In conclusion, our study identifies a genetic association between common ABCB4 gene variants and severe ICP in Swedish patients. Haplotype analysis was used to guide our search for relevant ICP‐associated polymorphisms. Our findings confirm and extend the results of previous genetic association studies13,29 for the clinically most important ICP subtype. We speculate that genetic risk profiling might help to stratify cases at the time of diagnosis, even before bile acid concentrations exceed levels that indicate an increased fetal risk.
Acknowledgements
We thank all colleagues and midwives from the Västra Götaland region, Sweden, for enrolling and taking care of study patients. Study nurses Ann Christiansson and Gunilla Kaplan (Göteborg, Sweden) and Maren Klöckner and Valerie Pucken (Bonn, Germany) were of invaluable help during clinical and experimental studies.
Abbreviations
ABC - ATP‐binding cassette
ICP - intrahepatic cholestasis of pregnancy
SNP - single‐nucleotide polymorphism
Footnotes
Funding: This study was supported in part by grants from the Department of Research and Development (FoU), Västra Götaland, Sweden, the Ministry of Innovation, Science, Research and Technology of North‐Rhine‐Westphalia (MIWFT‐NRW) and BONFOR, University of Bonn, Germany.
Competing interests: None declared.
References
- 1.Lammert F, Marschall H U, Glantz A.et al Intrahepatic cholestasis of pregnancy: molecular pathogenesis, diagnosis and management. J Hepatol 2000331012–1021. [DOI] [PubMed] [Google Scholar]
- 2.Milkiewicz P, Elias E, Williamson C.et al Obstetric cholestasis. BMJ 2002324123–124. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Poupon R. Intrahepatic cholestasis: from bedside to bench to bedside. Liver Int 200525467–468. [DOI] [PubMed] [Google Scholar]
- 4.Fisk N M, Storey G N. Fetal outcome in obstetric cholestasis. BJOG 1988951137–1143. [DOI] [PubMed] [Google Scholar]
- 5.Rioseco A J, Ivankovic M B, Manzur A.et al Intrahepatic cholestasis of pregnancy: a retrospective case‐control study of perinatal outcome. Am J Obstet Gynecol 1994170890–895. [DOI] [PubMed] [Google Scholar]
- 6.Bacq Y, Sapey T, Brechot M C.et al Intrahepatic cholestasis of pregnancy: a French prospective study. Hepatology 199726358–364. [DOI] [PubMed] [Google Scholar]
- 7.Laatikainen T, Tulenheimo A. Maternal serum bile acid levels and fetal distress in cholestasis of pregnancy. Int J Gynaecol Obstet 19842291–94. [DOI] [PubMed] [Google Scholar]
- 8.Alsulyman O M, Ouzounian J G, Ames‐Castro M.et al Intrahepatic cholestasis of pregnancy: perinatal outcome associated with expectant management. Am J Obstet Gynecol 1996175957–960. [DOI] [PubMed] [Google Scholar]
- 9.Glantz A, Marschall H U, Mattsson L A. Intrahepatic cholestasis of pregnancy: relationships between bile acid levels and fetal complication rates. Hepatology 200440467–474. [DOI] [PubMed] [Google Scholar]
- 10.Jacquemin E, Cresteil D, Manouvrier S.et al Heterozygous non‐sense mutation of the MDR3 gene in familial intrahepatic cholestasis of pregnancy. Lancet 1999353210–211. [DOI] [PubMed] [Google Scholar]
- 11.Dixon P H, Weerasekera N, Linton K J.et al Heterozygous MDR3 missense mutation associated with intrahepatic cholestasis of pregnancy: evidence for a defect in protein trafficking. Hum Mol Genet 200091209–1217. [DOI] [PubMed] [Google Scholar]
- 12.Jacquemin E, De Vree J M, Cresteil D.et al The wide spectrum of multidrug resistance 3 deficiency: from neonatal cholestasis to cirrhosis of adulthood. Gastroenterology 20011201448–1458. [DOI] [PubMed] [Google Scholar]
- 13.Pauli‐Magnus C, Lang T, Meier Y.et al Sequence analysis of bile salt export pump (ABCB11) and multidrug resistance p‐glycoprotein 3 (ABCB4, MDR3) in patients with intrahepatic cholestasis of pregnancy. Pharmacogenetics 20041491–102. [DOI] [PubMed] [Google Scholar]
- 14.Gendrot C, Bacq Y, Brechot M C.et al A second heterozygous MDR3 nonsense mutation associated with intrahepatic cholestasis of pregnancy. J Med Genet 200340e32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Müllenbach R, Bennett A, Tetlow N.et alATP8B1 mutations in British cases with intrahepatic cholestasis of pregnancy. Gut 200554829–834. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Painter J N, Savander M, Ropponen A.et al Sequence variation in the ATP8B1 gene and intrahepatic cholestasis of pregnancy. Eur J Hum Genet 200513435–439. [DOI] [PubMed] [Google Scholar]
- 17.Jacquemin E. Role of multidrug resistance 3 deficiency in pediatric and adult liver disease: one gene for three diseases. Semin Liver Dis 200121551–562. [DOI] [PubMed] [Google Scholar]
- 18.Thompson R, Strautnieks S. BSEP: function and role in progressive familial intrahepatic cholestasis. Semin Liver Dis 200121545–550. [DOI] [PubMed] [Google Scholar]
- 19.Glantz A, Marschall H U, Lammert F.et al Intrahepatic cholestasis of pregnancy: a placebo controlled randomised trial of ursodeoxycholic acid vs. dexamethasone on maternal and fetal outcome and metabolism. Hepatology 2005421399–1405. [DOI] [PubMed] [Google Scholar]
- 20.Pauli‐Magnus C, Kerb R, Fattinger K.et alBSEP and MDR3 haplotype structure in healthy Caucasians, primary biliary cirrhosis and primary sclerosing cholangitis. Hepatology 200439779–791. [DOI] [PubMed] [Google Scholar]
- 21.Mergenthaler S, Eggermann K, Tomiuk J.et al Exclusion of a disease relevant role of PAX4 in the aetiology of Silver‐Russell syndrome: screening for mutations and determination of imprinting status. J Med Genet 200037e44. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Elston R C, Forthofer R. Testing for Hardy‐Weinberg equilibrium in small samples. Biometrics 197733536–542. [Google Scholar]
- 23.De Finetti B. Conziderazioni matematiche sul'ereditarieta mendeliana. Meiron 192663–41. [Google Scholar]
- 24.Sasieni P D. From genotypes to genes: doubling the sample size. Biometrics 1997531253–1261. [PubMed] [Google Scholar]
- 25.Sebastiani P, Lazarus R, Weiss S T.et al Minimal haplotype tagging. Proc Natl Acad Sci USA 20031009900–9905. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Stephens M, Scheet P. Accounting for decay of linkage disequilibrium in haplotype inference and missing‐data imputation. Am J Hum Genet 200576449–462. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Eloranta M L, Heiskanen J T, Hiltunen M J.et al Multidrug resistance 3 gene mutation 1712delT and estrogen receptor α gene polymorphisms in Finnish women with obstetric cholestasis. Eur J Obstet Gynecol Reprod Biol 2002104109–112. [DOI] [PubMed] [Google Scholar]
- 28.Berg B, Helm G, Petersohn L.et al Cholestasis of pregnancy. Clinical and laboratory studies. Acta Obstet Gynecol Scand 198665107–113. [DOI] [PubMed] [Google Scholar]
- 29.Müllenbach R, Linton K J, Wiltshire S.et alABCB4 gene sequence variation in women with intrahepatic cholestasis of pregnancy. J Med Genet 200340e70. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Eloranta M L, Hakli T, Hiltunen M.et al Association of single nucleotide polymorphisms of the bile salt export pump gene with intrahepatic cholestasis of pregnancy. Scand J Gastroenterol 200338648–652. [DOI] [PubMed] [Google Scholar]
- 31.Savander M, Ropponen A, Avela K.et al Genetic evidence of heterogeneity in intrahepatic cholestasis of pregnancy. Gut 2003521025–1029. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Trauner M, Meier P J, Boyer J L. Molecular pathogenesis of cholestasis. N Engl J Med 19983391217–1227. [DOI] [PubMed] [Google Scholar]
- 33.Small D M. Role of ABC transporters in secretion of cholesterol from liver into bile. Proc Natl Acad Sci USA 20031004–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Smit J J, Schinkel A H, Oude Elferink R P.et al Homozygous disruption of the murine mdr2 P‐glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell 199375451–462. [DOI] [PubMed] [Google Scholar]
- 35.Gerloff T, Stieger B, Hagenbuch B.et al The sister of p‐glycoprotein represents the canalicular bile salt export pump of mammalian liver. J Biol Chem 199827310046–10050. [DOI] [PubMed] [Google Scholar]
- 36.Altshuler D, Brooks L D, Chakravarti A.et al A haplotype map of the human genome. Nature 20054371299–1320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Wasmuth H E, Matern S, Lammert F. From genotypes to haplotypes in hepatobiliary diseases: one plus one equals (sometimes) more than two. Hepatology 200439604–607. [DOI] [PubMed] [Google Scholar]
- 38.Valentonyte R, Hampe J, Huse K.et al Sarcoidosis is associated with a truncating splice site mutation in BTNL2. Nat Genet 200537357–364. [DOI] [PubMed] [Google Scholar]
- 39.Reichel C, Schneider G, Jacquemin E.et al Prevalence of mutations of the ABCB4‐gene encoding the multidrug resistance 3 glycoprotein (MDR3) in pregnant women with and without intrahepatic cholestasis of pregnancy (ICP). Hepatology 200134212A [Google Scholar]
- 40.Rosmorduc O, Hermelin B, Boelle P Y.et al ABCB4 gene mutation‐associated cholelithiasis in adults. Gastroenterology 2003125452–459. [DOI] [PubMed] [Google Scholar]
- 41.Lucena J F, Herrero J I, Quiroga J.et al A multidrug resistance 3 gene mutation causing cholelithiasis, cholestasis of pregnancy, and adulthood biliary cirrhosis. Gastroenterology 20031241037–1042. [DOI] [PubMed] [Google Scholar]
- 42.Schafmayer C, Tepel J, Franke A.et al Investigation of the Lith1 candidate genes ABCB11 and LXRA in human gallstone disease. Hepatology 200644650–657. [DOI] [PubMed] [Google Scholar]