Post on 21-Apr-2023
CL IN ICAL STUDIES
InvestigationoftheLith6 candidate genesAPOBEC1 andPPARGin humangallstone diseaseClemens Schafmayer1,2, Henry Volzke3, Stephan Buch4,5, Jan Egberts1, Annika Spille5,Huberta von Eberstein2, Andre Franke4, Markus Seeger5, Sebastian Hinz1, Abdou ElSharawy4,Dieter Rosskopf6, Mario Brosch5, Michael Krawczak7, Ulrich R. Foelsch5, Anton Schafmayer8,Frank Lammert9, Stefan Schreiber2,4, Fred Faendrich1, Jochen Hampe5, JuergenTepel1
1 Department of General Surgery and Thoracic Surgery, Christian-Albrechts-University Kiel, Kiel, Germany
2 POPGEN Biobank, Christian-Albrechts-University Kiel, Kiel, Germany
3 Institute for Community Medicine, Ernst Moritz Arndt University Greifswald, Greifswald, Germany
4 Institute of Clinical Molecular Biology, Christian-Albrechts-University Kiel, Kiel, Germany
5 Department of General Internal Medicine, Christian-Albrechts-University Kiel, Kiel, Germany
6 Institute of Pharmacology, Ernst Moritz Arndt University Greifswald, Greifswald, Germany
7 Institute of Medical Statistics and Biometry, Christian-Albrechts-University Kiel, Kiel, Germany
8 Department of General Surgery, Klinikum Luneburg, Luneburg, Germany
9 Department of Internal Medicine I, University of Bonn, Bonn, Germany
Keywords
APOBEC1 – gallstone disease – Lith genes –
PPARG
Correspondence
PD Dr Juergen Tepel, Department of General
and Thoracic Surgery, Christian-Albrechts-
Universitat Kiel, Arnold-Heller-Strasse 12,
24105 Kiel, Germany
Tel: 149 431 597 4308
Fax:149 431 597 3883
e-mail: jtepel@chirurgie-sh.de
Received 21 August 2006
accepted 13 April 2007
DOI:10.1111/j.1478-3231.2007.01536.x
AbstractBackground: Genetic susceptibility contributes to the aetiology of gallbladder
diseases as shown by multiple epidemiological studies. A major gallstone suscept-
ibility locus (Lith6) was identified in 2003 by quantitative trait locus mapping in
mice. Two attractive positional and functional candidate genes in apolipoprotein B
mRNA-editing protein (APOBEC1) and peroxisome proliferator-activated receptor
g (PPARG) are located in this interval. Aims: To investigate APOBEC1 and PPARG
as candidate genes for common symptomatic gallstone disease in humans. Patients
and methods: Eight hundred and ten patients who underwent cholecystectomy for
symptomatic gallstone disease (median age of onset 50) were compared with 718
sex-matched control individuals. An independent additional sample included 368
gallstone patients and 368 controls. Control individuals were sonographically free of
gallstones. Haplotype tagging and all known coding single nucleotide polymorph-
isms were genotyped for PPARG (N = 32) and APOBEC1 (N = 11). Results: The
investigated high-risk patient sample provides a power of greater than 80% for the
detection of odds ratios down to 1.45. No evidence of association of the two genes in
the single-point tagging markers, coding variants and in the sliding window
haplotype analysis was detected (all nominal single point P-values 4 0.04). A
logistic regression analysis including age, sex and BMI as covariates was also negative
(nominal P-values Z0.08). Conclusions: In the investigated German samples, no
evidence of association of APOBEC1 and PPARG with gallstone susceptibility was
detected. Systematic fine mapping of the complete Lith6 region is required to
identify the causative genetic variants for gallstone in mice and humans.
Gallbladder disease represents a major healthcareproblem both in Europe and the USA. Approximately12% of the US population are affected by gallstonesand more then 750 000 cholecystectomies are per-formed each year (1). Age, gender, ethnicity, obesity,diabetes and parity have all been implicated asrisk factors for the development of gallstones (2, 3).A genetic component in the susceptibility to gallstones
has been recognized as early as 1937 (4). The initialstudies have utilized pathological section statistics (4);more recent studies have turned to ultrasoundsurveys to investigate the familial clustering of chole-lithiasis (5). There is both strong evidence forfamilial clustering and an increased concordance ofthe trait in monocygotic twins as compared withdizygotic twins (6).
Liver International (2007)910 c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard
Liver International ISSN 1478-3223
Systematic, genome-wide scans of gallstone traits inexperimental crosses of inbred mouse strains haveyielded a susceptibility map of ‘Lith’ loci (7–11).Several of these loci have been found in differentmouse crosses, thus strengthening the case for robustsusceptibility loci in the mouse (11). The humanorthologues of these loci are thus natural positionalcandidates for gallstone susceptibility also in humans.Of note, among all mouse crosses specifically aimed atidentifying Lith loci in mice, only one failed to detectlinkage between cholesterol gallstones and Chr 6. Thelocus was named Lith6 when it was detected withsignificant linkage in the D2�CAST intercross (12).Coincident loci with significant or suggestive linkagepeaks were identified in D2�PERA (11) and129S1�CAST intercrosses (12). Positional candidategenes that colocalize with Lith6 and are involved inlipid homeostasis include the peroxisome proliferator-activated receptor gamma (PPARG) and apolipo-protein B mRNA-editing protein (APOBEC1).Although the exact source of the cholesterol destinedfor biliary excretion is not known, increased biliarycholesterol concentrations may be the result of geneticalterations in lipid metabolism in which both candi-date genes are involved (13). The PPARG gene maps tohuman chromosome 3p25.2, which corresponds tomouse chromosome 6E3. The peroxisome prolifera-tor-activated receptors (PPARs) are members of thenuclear hormone receptor subfamily of transcriptionfactors. PPARs form heterodimers with retinoid Xreceptors and these heterodimers regulate the tran-scription of various genes. There are three knownsubtypes of PPARs, PPAR-a, PPAR-d and PPAR-g.PPAR-g has a central role in adipocyte differentiation.Two protein isoforms of PPAR-g have been identifiedthat stem from the same gene because of alternativepromoter usage and alternative mRNA splicing.PPARg2 is specifically expressed in adipose tissue anddiffers from the more ubiquitously expressed PPARg1by the presence of an additional 28 amino acid N-terminal exon (14, 15).
Several genetic variants have been identified in thePPAR-g locus. The most common variation is asubstitution of proline with alanine at position 12 inexon 1 of PPARg2, with a frequency of about 12% inCaucasians (16) that has been implicated in theprotection against type II diabetes (17). A rare gain-of-function mutation (Pro115Gln) has been observedin four of 121 unrelated obese German subjects (18).The PPAR-g has also been implicated in cases ofinsulin resistance and partial lipodystrophy (19–21).
Apolipoprotein B, a major protein component ofcirculating plasma lipoproteins, exists in two forms:
apoB-100 and apoB-48. In humans, apoB-100 issynthesized in the liver and is present in very low-density lipoproteins and their metabolic products. Onthe other hand, apoB-48 is synthesized in the smallintestine and is present in chylomicrons and chylomi-cron remnants in the form of apoB. The APOBEC1 isinvolved in the production of apoB-48 from apoB-100and is located on human chromosome 12p13.3 – itsmouse homologue Apobec1 is also located in themouse Lith6 locus on chromosome 6F1. Comparedwith control mice, Apobec1 knockout mice fed alithogenic diet display increased biliary cholesterolsecretion rates, decreased bile salt secretion rates, andaccelerated cholesterol gallstone formation (14). How-ever, mice exhibiting a similar apoB-100-only pheno-type generated by abolition of the APOBEC1 targetnucleotide in Apob showed reduced gallstone suscept-ibility, indicating that APOBEC1 may exert additionalfunctions other than Apob mRNA processing thataffect cholesterol gallstone formation. Interestingly,no large investigation of mutations in the PPARG orAPOBEC1 genes for human gallstone susceptibility hasbeen published.
Despite the advanced stage of genetic investigationin the mouse system, there has been a relative scarcityof genetic studies in the human system. An associationof pigment gallstones with a promoter variation in theUDP-glucuronosyltransferase 1A1 (UGT1A1) gene hasbeen reported in patients with chronic haemolyticdisorders (22, 23) and cystic fibrosis (24). Investiga-tions of the 7a-hydroxylase gene (18) and apolipopro-teins A and B (18–20) in 105–221 and apolipoproteinE in 37–169 patients (25, 26) did not yield replicatedassociation findings. A recent genome-wide linkagescan in Mexican Americans identified significant link-age (LOD 3.7) of gallstone susceptibility on chromo-some 1p (27). This represents the first genome-wideinvestigation of gallstone susceptibility in humans andis an excellent starting point for further systematicmapping efforts.
In this report, we have investigated two functionalcandidate genes within the mouse Lith6 locus forassociation with symptomatic gall stone disease in alarge sample of patients who underwent cholecystect-omy for early onset of symptomatic gallstone disease.
Methods
Patients and phenotypes
Patients who underwent cholecystectomy for cholecys-tolithiasis from 2001 to 2005 in the surgical depart-ments of the hospitals at Kiel, Luneburg, Rendsburg,Schleswig, Flensburg, Husum, Niebull, Heide, and
Liver International (2007)c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard 911
Schafmayer et al. Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease
Eckernforde (all in Northern Germany) were contactedthrough mail by the respective hospitals and invitedto participate in this study. For patients who didnot respond, one written reminder was sent. Indivi-duals who agreed to participate were contacted by thePOPGEN (http://www.popgen.de) recruitment project(28): they were interviewed by mail questionnaire anda venous EDTA blood sample was obtained either atthe POPGEN offices or from the patients’ generalpractitioners. For both cases and controls, this studywas restricted to probands of German ethnicity, i.e.only individuals whose parents were born in Germanywere included. The study protocols were approved bythe institutional ethics committee and the public dataprotection agency. Written, informed consent wasobtained from all study participants. The first conse-cutive 810 participants were included in this study.
Gallstone-free control individuals were recruitedfrom consecutive patients in the years 2003–2004 fromthe routine clinical ultrasound of the Department ofGeneral Internal Medicine, Kiel. Patients with malig-nant disorders were excluded. Index patients wereidentified on the basis of ultrasound reports, andN = 277 probands were recruited using the POPGENinfrastructure using the protocols described above.Additional controls, N = 441, were obtained frompopulation-derived control individuals from thePOPGEN project [patients identified on the basis of theinhabitants register (28)], who underwent an additionalphysical examination at the POPGEN facilities thatincluded an abdominal ultrasound by a trained physi-cian. There was no significant difference in allele fre-quencies between the two control samples (P4 0.1).
DNA was prepared using the FlexiGene chemistry(Qiagen, Hilden, Germany) according to the manu-facturer’s protocols from all samples. This patientsample has been used in a previous study of theauthors (29). An overview of the patient samples isgiven in Table 1.
An additional study population (panel B) fromNorthern Germany utilized 368 cases (median age of
onset 53 years) and 368 sex-matched controls re-cruited from the Study of Health in Pomerania(SHIP), a population-based study representative forthe adult population of West Pomerania (30). Casesreported a history of cholecystectomy or showedevidence of cholelithiasis on ultrasound examination,whereas controls were free of gallstone disease asevidenced by ultrasound. An overview of the studypopulation is given in Table 1.
Genotyping
DNA samples were evaluated by gel electrophoresis andadjusted to 20–30 ng/mL DNA content using the Pico-green fluorescent dye (Molecular Probes – Invitrogen,Carlsbad, CA, USA). One microliter of genomic DNAwas amplified by the GenomiPhi (Amersham, Uppsala,Sweden) whole genome amplification kit and fragmen-ted at 99 1C for 3 min. One hundred nanograms of DNAwere dried overnight in TwinTec hardshell 384-wellplates (Eppendorf, Hamburg, Germany) at room tem-perature. Genotyping was performed on these platesusing the SNPlex chemistry (Applied Biosystems, FosterCity, CA, USA) on an automated platform with TecanFreedom EVO and 384-well TEMO liquid handlingrobots (TECAN, Mannedorf, Switzerland). As imple-mented by the proprietary SNPlex chemistry, the pro-ducts of the whole-genome amplification reaction wereincubated with pooled oligos in order to perform anallele-specific oligonucleotide ligation assay. Ligationproducts were subsequently amplified using universalbiotinylated primers. Polymerase chain reaction pro-ducts were bound to streptavidin-coated plates andhybridized to a mixture of fluorescently labelled oligo-nucleotides, which assume a defined position in thesubsequent capillary electrophoresis owing to coupledelectrophoretic mobility modifiers (‘‘ZipChutes’’). Theresultant ZipChute mixture was then separated on ABI3730� L sequencers. Genotypes were reviewed manuallyusing the GENEMAPPER 4.0 (Applied Biosystems) software.All process data were logged and administered through adatabase-driven LIMS system (31). Genotyping forPPAR-g nonsynonymous coding single nucleotide poly-morphisms (SNPs) Pro12Ala (rs1801282), Pro115Gln(rs1800571) and for mutations Phe388Leu, Arg425Cyswas performed using Taqmans (Foster City, CA, USA)allelic discrimination assays, as described before (12).Minor groove binder (MGB) probe and primer se-quences are available upon request from the authors.
SNP selection and data analysis
Single nucleotide polymorphisms were selected fromHapMap (www.hapmap.org) by the automated
Table 1. Overview of the patient sample
Panel
Patients Controls
Male (%)N Age�Age atDiagnosis� N Age�
Panel A 810 55 50 718 64 50Panel B 368 59 53 368 64 39
�Median values reported.
Both patient samples are independent and have no overlap. Patient
panel A was recruited in Kiel and panel B in Greifswald (both in Northern
Germany).
Liver International (2007)912 c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard
Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease Schafmayer et al.
selection of haplotype tagging SNPs for Caucasiansfrom the CEU data set (settings: Mendelian errors: 0,minor allele frequency: 0.01, HWE cut-off: 0.01). Inaddition, coding SNPs with a frequency 4 0.01 wereselected from dbSNP. Moreover, for PPARG the raremutation Pro115Gln and two mutations located in theligand-binding domain (R425C and F388L) wereselected based on literature reports as described in theintroduction.
This study was performed using a case–controldesign: Sliding window haplotype analysis was per-formed using COCAPHASE through the UNPHASED suiteof programs (http://www.rfcgr.mrc.ac.uk/�fdudbrid/software/unphased/) (32). Single-point genotypic andallelic tests of association on sex-matched genotyperesults were performed using w2 statistics on contin-gency tables. For all tests, nominal P-values are re-ported. Logistic regression analyses were performedusing SPSS version 11.
Results
The expected power of the combined samples A and Bof 1178 cases and 1068 controls was estimated forsingle-point allelic effects over an odds ratio of one totwo at a nominal significance level of 0.05 (33) for(minor) allele frequencies of 0.1, 0.25 and 0.5 of apotential susceptibility marker. As shown in Fig. 1, thepower for the detection of odds ratios greater than1.45 is approximately 80% or greater in all models. Formore frequent susceptibility factors, odds ratios of1.27 would also be detectable at this power level.
For both candidate genes, haplotype-tagging SNPswere generated using the HapMap genotype data(http://www.hapmap.org) for the candidate gene re-gions (34). Twenty-eight SNP markers in the PPARG
gene were selected. These 28 markers provide goodcoverage as determined by the tagging functionality(35) in HAPLOVIEW (36) (settings: see ‘Methods’). Inaddition, all coding SNPs with an allele frequency4 1% in Caucasians as judged by dbSNP or theliterature were included. In addition, the major SNPsthat have been reported to be associated with diabetesor obesity were genotyped (16–18). A selection ofSNPs that have been reported to cause lipodystrophicinsulin resistance was also genotyped (19–21). Figure 2shows the distribution of markers across the gene andthe regional haplotype structure as generated from theHapMap Caucasian genotypes (category CEU) byHAPLOVIEW. As shown in Fig. 2, the markers tag allmajor haplotype blocks of this gene.
The SNP panel was genotyped in both patientpanels as shown in Table 1. No marker showeddeparture from Hardy–Weinberg equilibrium(P4 0.1), indicating robust genotyping performancein this experiment. Single-point allelic and genotypicassociation tests were performed, yielding nominal P-values ranging from 0.046 to 0.98. In order to improvepower – i.e. to detect association with variants on thehaplotypes not directly tagged by one of the SNPs inthe experiments – a sliding window haplotype analysisusing window sizes of two to five markers wasperformed in the high-risk patient sample (panel A).Here, a nominal significance level of 0.04 was obtainedfor the window size 3 for the haplotype spanning fromrs2938395 to rs796290 and from rs4684104 tors709163 for the window size 5 respectively. However,none of the neighbouring haplotypes showed evidenceof association. Coding SNPs P12A showed no signifi-cant difference in allele or genotype frequencies be-tween cases and controls. For variant rs1800571, theannotation of the NCBI-reference sequence P113Qwas used. This SNP is often also referred to as P115Qin the literature. Coding mutations P113Q, P388L andR425C were monomorphic in both cases and controlsin all investigated samples. Table 2 lists all the resultsfor this gene.
Eleven SNPs were selected for APOBEC1 using thesame approach as described above for PPARG. Theknown coding SNP rs2302515, which leads to anamino acid exchange of isoleucine to methionine, wasincluded in the design. This variant showed no evi-dence of association with the gallstone trait (P4 0.1 inboth the genotypic and allelic single-point associationtest). Single-point analyses for genotypic and allelicassociation yielded nominal P-values in the range from0.06 to 0.94. The sliding window-haplotype analyses inthe high-risk patient sample (panel A) resulted innominal significance levels between 0.10 and 0.91.
Fig. 1. The figure shows a graphical representation of a powerestimation in the combined samples A and B at a significancelevel of Po 0.05 for a two-sided test over an odds ratio (OR)range of 1.0–2.2. The graph was generated using PS-power (33)and shows the test power as a function of the OR (x-axis).
Liver International (2007)c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard 913
Schafmayer et al. Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease
The SNP overview and all results of the associationanalysis are presented in Table 3 for APOBEC1 (Fig. 3).
In order to look for a genotype-specific gallstonerisk in the presence of the covariates of age, sex andBMI, a logistic regression analysis over both patientsamples was performed. For each marker, the modelcontained age, sex and BMI and the respective markergenotypes. Age, sex and BMI were consistently signifi-cant in all models (Po 0.01) – the results for theindividual genetic markers are reported in Tables 2 and3 and are consistently negative.
Discussion
The investigation of gallstone susceptibility in mice hasled to a genomic map of gallstone risk loci in thisspecies. The ‘Lith6’ locus was formally reported in 2003as a new quantitative trait locus (QTL) for cholesterolgallstones from an intercross of CAST/Ei and DBA/2J
(37). This region on mouse chromosome 6 had shownsome degrees of linkage evidence in multiple earlierstudies (9, 10, 38) and was also seen in a QTL analysisfrom an intercross of PERA/Ei and I/LnJ inbred mousestrains with an LOD score of 2.0 (11). Thus, Lith6represents one of the best replicated Lith loci in themouse system. The investigation of human homologuesof the functional candidate genes in these intervalsrepresents an intuitive approach to potential risk var-iants for the common gallstone trait in humans (11, 13).
The patient panel A reported here has utilized over800 patients, who have been operated for symptomaticgallstone disease. The population median age of affec-tion in North Germany for gallstones is about 65 years(30), i.e. approximately 50% of the patients developthe disease up to this age. In order to further improvepower, only patients with an age of onset below 65years were included in this study, yielding a medianage of onset of 50 years in cases. In a further attempt to
Fig. 2. Overview of the physical and genetic structure of the peroxisome proliferator-activated receptor gamma (PPARG) gene regionas generated from the Caucasian HapMap genotypes by HAPLOVIEW (36): The upper panel shows the identifiers of the single nucleotidepolymorphisms (SNPs) in relation to their physical position according to the NCBI genome assembly build 34. The coordinates of thegenomic interval are given in this panel. The exon structure of the PPARG transcripts is given below the SNP list also on the physicalsequence scale. Below the exon structure, a translation of the physical positions into the equidistant presentation of linkagedisequilibrium is presented. The lower panel gives an overview of the pairwise linkage disequilibrium structure of the locus (D0),as generated by HAPLOVIEW (36) from the Caucasian HapMap data.
Liver International (2007)914 c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard
Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease Schafmayer et al.
Tab
le2.
Res
ults
of
the
gen
etic
asso
ciat
ion
anal
ysis
of
the
inve
stig
ated
taggin
gan
dco
din
gSN
Ps(b
old
-typ
e)in
the
PPA
RG
gen
e
dbSN
Pid
Posi
tion�
Panel
APa
nel
BA
1B
Panel
A
MA
Fca
seM
AF
contr
ol
OR
Rec
OR
dom
P alle
licP g
eno
MA
Fca
seM
AF
contr
ol
OR
Rec
OR
dom
P alle
licP g
eno
P reg
HA
P2H
AP3
HA
P4H
AP5
rs7632481
50 fl
anki
ng
0.4
50.4
51.0
30.9
70.9
10.8
10.4
50.4
80.8
10.8
50.3
00.5
80.3
80.7
10.2
20.1
40.1
9rs
12635120
50 fl
anki
ng
0.2
00.2
10.7
40.9
40.4
00.5
20.1
90.1
80.8
81.0
80.7
90.7
70.4
00.4
00.2
50.1
40.0
9rs
2960420
50 fl
anki
ng
0.3
50.3
70.8
90.8
60.2
50.3
30.3
80.3
71.0
71.0
10.8
40.9
40.4
40.6
90.5
90.3
00.2
5rs
4279078
50 fl
anki
ng
0.1
20.1
20.6
41.0
40.9
50.4
50.1
10.0
80.7
11.4
0.1
30.1
70.4
80.9
50.4
20.3
30.2
5rs
4684846
50 fl
anki
ng
0.3
0.2
90.9
21.1
40.5
70.1
70.2
90.3
20.7
40.8
80.2
40.4
80.7
60.4
10.4
00.5
20.3
4rs
12493718
50 fl
anki
ng
0.2
90.2
90.9
31.1
0.6
90.3
50.2
90.3
10.8
10.9
20.4
30.7
10.7
00.5
50.6
80.4
10.6
2rs
7638903
50 fl
anki
ng
0.1
40.1
50.8
30.9
40.5
50.8
10.1
60.1
70.9
70.9
10.5
90.8
20.1
90.8
60.3
60.4
80.6
0rs
10510418
50 fl
anki
ng
0.2
90.2
80.8
71.0
90.8
40.2
80.3
00.2
61.1
1.4
0.1
00.0
50.1
10.4
30.5
20.6
60.5
9rs
6802898
50 fl
anki
ng
0.1
40.1
50.9
60.9
40.6
30.8
70.1
70.1
80.8
90.9
10.5
60.8
40.2
30.3
80.5
70.4
60.4
1rs
1801282
ex1
P12A
0.1
40.1
50.9
20.9
20.4
80.7
60.1
60.1
71.0
90.9
10.6
80.8
00.1
70.8
10.5
60.7
00.8
4rs
4135256
intr
on
0.0
20.0
3N
Aw
0.9
10.7
60.5
60.0
30.0
3N
Aw
1.1
0.7
70.7
60.7
80.9
50.9
50.9
40.9
1rs
2120825
intr
on
0.1
30.1
30.8
91.0
30.9
00.8
90.1
40.1
50.9
80.9
30.7
10.9
10.2
40.7
90.9
20.7
10.1
1rs
1800571
ex3
P113Qz
00
NAw
NAw
NAw
NAw
00
NAw
NAw
NAw
NAw
NAw
–‰–‰
–‰–‰
rs10510419
intr
on
0.1
50.1
50.7
51.1
40.5
00.2
20.1
40.1
70.7
20.7
90.1
40.3
40.9
60.7
90.5
50.0
70.1
6rs
2938395
intr
on
0.3
40.3
40.9
90.9
50.7
80.8
50.3
50.3
21.0
41.3
40.2
40.0
60.1
80.3
60.0
40.1
20.1
8rs
4135268
intr
on
0.0
70.0
90.6
10.8
40.1
70.3
90.0
70.0
51.0
31.2
70.3
20.5
80.4
70.5
40.6
00.7
10.3
9rs
796290
intr
on
0.5
00.4
81.0
91.2
20.5
10.1
10.4
90.5
00.9
50.9
30.7
90.9
00.1
60.5
40.6
40.2
70.8
1–
ex6
F388L
00
NAw
NAw
NAw
NAw
00
NAw
NAw
NAw
NAw
NAw
–‰–‰
–‰–‰
rs709157
intr
on
0.2
70.2
70.9
61.0
10.9
80.9
50.2
80.2
70.8
41.1
50.7
20.2
60.5
10.5
90.3
20.8
20.6
2rs
1175540
intr
on
0.3
20.3
11.1
10.9
90.8
20.7
30.3
10.2
91.0
61.1
90.3
90.4
30.1
80.6
60.8
80.5
00.1
0–
ex7
R425C
00
NAw
NAw
NAw
NAw
00
NAw
NAw
NAw
NAw
NAw
–‰–‰
–‰–‰
rs3856806
ex7
H447H
0.1
40.1
41.2
0.9
90.8
80.8
20.1
50.1
50.6
21
0.8
30.6
90.6
00.8
00.4
30.2
30.2
4rs
1152003
30 fl
anki
ng
0.3
40.3
31.1
61.0
60.4
30.7
00.3
70.3
41.0
51.3
0.3
20.1
00.2
50.4
50.1
80.1
80.3
1rs
9833097
30 fl
anki
ng
0.1
20.1
1.7
21.1
70.1
40.3
00.1
30.1
21
1.0
20.9
40.9
80.2
00.1
50.2
20.1
70.1
9rs
1152004
30 fl
anki
ng
0.1
70.1
90.5
90.8
70.1
10.1
80.1
70.1
70.8
31.0
60.8
70.8
10.4
70.2
50.3
00.2
80.3
6rs
4684104
30 fl
anki
ng
0.4
0.4
10.9
30.9
50.6
40.8
90.4
30.4
11.2
21.1
60.3
40.6
10.6
00.8
10.2
90.3
00.0
4rs
1181795
30 fl
anki
ng
0.1
00.1
02.6
60.9
70.9
00.2
70.1
20.1
11.0
31.1
60.4
60.7
20.8
40.5
60.3
20.1
70.3
2rs
3936555
30 fl
anki
ng
0.1
30.1
51.0
30.8
40.1
90.2
80.1
30.1
20.8
31.1
30.5
70.6
90.3
60.3
50.0
90.2
20.2
5rs
1699348
30 fl
anki
ng
0.4
40.4
21.0
81.0
90.5
30.7
30.4
90.4
51.3
11.3
40.1
40.1
90.1
10.0
80.2
60.2
3–z
rs709163
30 fl
anki
ng
0.2
60.2
90.6
90.8
30.0
60.1
30.2
50.2
80.7
10.8
40.1
70.3
90.0
80.1
90.1
3–z
–zrs
709162
30 fl
anki
ng
0.4
40.4
31.0
81.0
30.6
30.8
50.4
80.4
41.2
81.3
30.1
70.2
00.2
20.3
7–z
–z–z
rs1797886
30 fl
anki
ng
0.4
80.5
00.9
00.9
20.4
90.7
60.4
50.4
60.8
31.0
70.5
00.1
50.4
2–z
–z–z
–z
� Rel
ativ
eposi
tions
are
report
edfo
rtr
ansc
ript
NM
_015869
(corr
espondin
gis
ofo
rmPP
ARg2
).
wNo
OR
or
P-va
lue
calc
ula
ted
ow
ing
tolo
wal
lele
freq
uen
cy.
zPro
113G
lnis
equiv
alen
tto
Pro115G
lndes
crib
edby
Ris
tow
etal
.(1
998).
‰Monom
orp
hic
mar
kers
excl
uded
from
hap
loty
pe
anal
ysis
.
zP-v
alue
report
edfo
rth
efirs
tm
arke
rin
the
hap
loty
pe
win
dow
.
The
min
oral
lele
freq
uen
cies
(MA
F)fo
rca
ses
and
contr
ols
are
report
ed.O
Rdom
and
OR
recre
ferto
the
OR
under
the
rece
ssiv
eordom
inan
tm
odes
of
inher
itan
cere
spec
tive
ly.P-
valu
esar
ere
port
edfo
ral
lelic
(Palle
lic)a
nd
gen
oty
pic
asso
ciat
ion
(Pgeno).
The
colu
mns
HA
P2to
HA
P5re
ferto
asl
idin
gw
indow
hap
loty
pe
anal
ysis
usi
ng
CO
CA
PHA
SE.T
he
resu
lts
ofth
ere
gre
ssio
nan
alys
isar
egiv
enunder
the
hea
din
g‘‘p
reg’’.
SNP,
single
nucl
eotide
poly
morp
his
ms;
OD
,odds
ratio.
Liver International (2007)c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard 915
Schafmayer et al. Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease
maximize the power of this investigation, only gall-stone-negative (as determined by abdominal ultra-sound investigation) control individuals with amoderately higher median age (64 years) were used.Many polygenic disorders are characterized by a strongcorrelation between the age of onset of affectedrelatives, as has been documented for instance inbreast cancer (39) and Alzheimer disease (40). It isindeed plausible that the genetic influence upon thedevelopment of such disorders is reflected by the age atwhich individuals develop the disorder (41). Thegallstone hazard ratio may therefore be higher in theseyounger patients than the background familial risk oftwo to four reported for gallstones (4, 5, 42–46). Thus,the investigated sample represents a ‘high-risk’ groupof patients that should provide an increased experi-mental power as compared with a general populationsample. In addition, nominal P-values were reportedfor all tests, including the haplotype associations tofurther improve power.
The international HapMap project (www.hapmap.org) has generated a wealth of genotype and markerinformation that significantly aids in the design ofcandidate gene studies (34). For this candidate genestudy, a primary haplotype-tagging approach waschosen, i.e. the genetic variation at both loci wascaptured with a set of carefully selected variants. Thistagging approach is able to detect signals from un-known regulatory or functional elements (35, 47, 48).It thus potentially even offers advantages over a directmutation detection of the coding region of a gene only,because disease susceptibility may also be conferred byother variants, for instance in splice sites or intronicenhancers (49, 50). Tagging SNPs were selected fromthe public HapMap (www.hapmap.org) SNP resources(34). The genotype and allele frequencies of the SNPsinvestigated here were not significantly different fromthe ones for the Caucasian HapMap individuals(P4 0.11; allele frequencies in Tables 2 and 3), thussupporting the selection of variants from this resource.Coding SNPs as described in ‘Results’ were alsoincluded in the marker panel.
In PPARG, only SNP rs10510418 reached an un-corrected single-point significance value of P = 0.046in the genotypic test in panel B. This signal was neithersupported in the other cohort nor in the combinedregression analysis. None of the other single-pointnominal P-values in any of the tests passed thePo 0.05 threshold. The coding SNP P12A was con-sistently negative in both patient panels and thecombined regression analysis (panel A: Pallelic = 0.48;Pgenotypic = 0.76; panel B: Pallelic = 0.68; Pgenotypic =0.80). The P115Q (P113Q), F388L and R425CTa
ble
3.
Res
ults
of
the
gen
etic
asso
ciat
ion
anal
ysis
of
the
inve
stig
ated
taggin
gan
dco
din
gSN
Ps(b
old
-typ
e)in
the
APO
BEC
1gen
e
dbSN
Pid
Posi
tion�
Panel
APa
nel
BA
1B
Panel
A
MA
Fca
seM
AF
contr
ol
OR
Rec
OR
dom
P alle
licP g
eno
MA
Fca
seM
AF
contr
ol
OR
Rec
OR
dom
P alle
licP g
eno
P reg
HA
P2H
AP3
HA
P4H
AP5
rs10772577
30flan
king
0.4
30.4
20.9
50.9
10.6
00.6
50.3
80.4
10.7
20.8
80.1
70.3
10.3
80.1
60.1
80.1
00.1
6rs
2377084
30flan
king
0.4
00.3
90.9
70.9
0.6
10.5
20.3
40.3
80.6
80.8
60.1
00.2
30.2
20.6
40.7
70.8
90.9
0rs
7968271
30flan
king
0.0
80.0
81.1
80.9
30.6
80.8
50.0
70.0
90.9
50.7
30.1
50.3
00.1
90.5
40.7
20.9
10.8
2rs
10431307
30flan
king
0.2
30.2
10.9
60.8
90.3
50.4
80.1
70.2
10.4
70.8
30.0
60.0
80.1
90.5
00.7
30.7
10.8
6rs
6488547
Intr
on
0.0
30.0
2N
Aw
0.8
0.3
30.3
20.0
20.0
2N
Aw
0.9
40.8
70.8
70.4
40.5
40.6
90.6
00.8
4rs
2302515
ex3
I80M
0.1
30.1
41.6
60.9
90.8
70.5
20.1
20.1
50.9
30.7
40.1
10.1
80.5
20.5
70.5
80.8
50.8
5rs
7137790
Intr
on
0.1
80.1
91.6
1.0
30.4
30.2
80.1
50.1
90.7
20.8
80.0
60.0
90.6
70.4
90.7
00.7
50.8
1rs
1015102
Intr
on
0.2
00.1
90.8
90.9
0.3
40.6
20.1
60.1
60.6
1.0
90.9
80.2
30.8
00.5
30.7
20.8
1–z
rs7973596
Intr
on
0.1
20.1
31.3
1.0
90.4
30.7
10.1
20.1
01.4
20.7
60.2
80.1
50.7
90.7
10.6
7–z
–zrs
6488567
50flan
king
0.1
70.1
71.7
21.0
10.5
70.2
60.1
40.1
41.3
40.9
30.9
40.5
60.5
70.5
3–z
–z–z
rs10845677
50flan
king
0.3
50.3
71.2
81.0
20.3
20.2
10.3
60.3
40.9
10.8
80.5
00.6
80.5
1–z
–z–z
–z� G
ene
istr
ansc
ribed
from
the
gen
om
icm
inus
stra
nd.
wNo
OR
calc
ula
ted
due
tolo
wal
lele
freq
uen
cy.
zP-v
alue
report
edfo
rth
efirs
tm
arke
rin
the
hap
loty
pe
win
dow
.
Colu
mn
hea
din
gs
are
iden
tica
lto
the
PPA
RG
gen
eove
rvie
w.
SNP,
single
nucl
eotide
poly
morp
his
ms;
OD
,odds
ratio.
Liver International (2007)916 c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard
Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease Schafmayer et al.
mutations were not found in any of the 1178 patientsand 1068 controls screened. These SNPs were includedfor the sake of completeness: for the syndromaticmutations identified in lipodystrophic insulin resis-tance, their presence in the investigated patient samplewas unlikely owing to the very low frequency of theunderlying syndrome. Variant P115Q (P113Q) wasnot identified in additional patient samples apart fromthe original study (18), indicating an exceedingly lowfrequency of the variant in German Caucasians (51).
The coding SNP rs2302515, which leads to anamino-acid exchange at amino acid position 60 in theAPOBEC1 gene, also showed no evidence of associa-tion with the trait under study. This variant is alsolocated outside the functionally most relevant regionsof the protein, which were studied in depth in anearlier mutagenesis experiment (52).
In summary, this association study is negative forboth genes in both patient samples. A logistic regres-sion analysis in the presence of covariates sex, age andBMI was negative. Thus, in the investigated Germansample, these two loci do not represent major riskfactors for common symptomatic gallstone disease.The mouse Lith6 locus spans a large amount ofgenomic sequence in the mouse and – consequently –also in the orthologous human region. Thus, furtherfine mapping of the mouse gallstone loci and identifi-cation of the causative variants in the animal system isneeded to transfer these discoveries successfully to thehuman disease. Potentially, association mapping of therespective human regions may offer advantages owingto the lower level of disequilibrium in human popula-tions and the abundant genomic resources in humans.Ultimately, the mapping efforts in mice and humans
Fig. 3. Overview of the physical and genetic structure of the APOBEC1 gene region. This gene is transcribed from the genomic minusstrand. The panels are otherwise organized as in Figure 2.
Liver International (2007)c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard 917
Schafmayer et al. Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease
may have to be combined to untangle the genetic basisof gallstones.
Acknowledgements
This study was supported by the German National Genome
Research Network (NGFN) through the POPGEN Biobank
(BmBF 01GR0468), the National Genotyping Platform and
the NGFN network for environmental disorders. The co-
operation of all patients, their families and physicians who
participated in this study is gratefully acknowledged. Espe-
cially, the help of Christine Hoell, Huberta von Eberstein
(POPGEN Biobank project), the ‘‘Genotyping TA’s’’ and of
the heads of the surgical departments Ilka Vogel (Stadtisches
Krankenhaus Kiel), Marco Sailer (Hamburg Bergedorf),
Hermann Dittrich (Rendsburg), Jurgen Belz (Husum),
Rainer Quaschling (Eckernforde), Hodjat Shekarriz (Schles-
wig), Volker Mendel (Flensburg), Werner Neugebauer
(Flensburg), Jiri Klima (Niebull), Friedrich Kallinowski
(Heide) and Anton Schafmayer (Luneburg) is gratefully
acknowledged. Helpful discussions and scientific guidance
from Thomas Bosch (Kiel) are especially acknowledged.
References
1. Graves EJ, Owings MF. 1996 summary: National Hospital
Discharge Survey. Adv Data 1998; 1–12.
2. Paigen B, Carey MC. Gallstones. In: King RA, ed. The Genetic
Basis of Common Diseases. London: Oxford University Press,
2002; 298–335.
3. Everhart JE, Khare M, Hill M, Maurer KR. Prevalence and
ethnic differences in gallbladder disease in the United States.
Gastroenterology 1999; 117: 632–9.
4. Korner G. Uber die familare Haufung der Gallenblasenkran-
kheiten. Z Menschl Vererbun Konstitutionslehre 1937; 20:
528–82.
5. Nurnberg D, Berndt H, Pannwitz H. Familiare Haufung von
Gallensteinen. Dtsch Med Wschr 1989; 114: 1059–63.
6. Katsika D, Grjibovski A, Einarsson C, Lammert F, Lichten-
stein P, Marschall HU. Genetic and environmental influences
on symptomatic gallstone disease: a Swedish study of 43,141
twin pairs. Hepatology 2005; 41: 1138–43.
7. Lammert F, Carey MC, Paigen B. Chromosomal organization
of candidate genes involved in cholesterol gallstone forma-
tion: a murine gallstone map. Gastroenterology 2001; 120:
221–38.
8. Khanuja B, Cheah YC, Hunt M, et al. Lith1, a major gene
affecting cholesterol gallstone formation among inbred
strains of mice. Proc Natl Acad Sci USA 1995; 92: 7729–33.
9. Paigen B, Schork NJ, Svenson KL, et al. Quantitative trait loci
mapping for cholesterol gallstones in AKR/J and C57L/J
strains of mice. Physiol Genom 2000; 4: 59–65.
10. Lammert F, Wang DQ, Wittenburg H, et al. Lith genes
control mucin accumulation, cholesterol crystallization, and
gallstone formation in A/J and AKR/J inbred mice. Hepatol-
ogy 2002; 36: 1145–54.
11. Wittenburg H, Lyons MA, Li R, Churchill GA, Carey MC,
Paigen B. FXR and ABCG5/ABCG8 as determinants of
cholesterol gallstone formation from quantitative trait locus
mapping in mice. Gastroenterology 2003; 125: 868–81.
12. Hampe J, Franke A, Rosenstiel P, et al. A genome-wide
association scan of nonsynonymous SNPs identifies a sus-
ceptibility variant for Crohn disease in ATG16L1. Nat Genet
2007; 39: 207–11.
13. Wang DQ, Afdhal NH. Genetic analysis of cholesterol
gallstone formation: searching for Lith (gallstone) genes.
Curr Gastroenterol Rep 2004; 6: 140–50.
14. Tontonoz P, Hu E, Graves RA, Budavari AI, Spiegelman BM.
mPPAR gamma 2: tissue-specific regulator of an adipoctye
enhancer. Genes Dev 1994; 8: 1224–34.
15. Fajas L, Auboeuf D, Raspe E, et al. The organization,
promoter analysis, and expression of the human PPARgam-
ma gene. J Biol Chem 1997; 272: 18779–98.
16. Yen CJ, Beamer BA, Negri C, et al. Molecular scanning of the
human peroxisome proliferator activated receptor gamma
(hPPAR gamma) gene in diabetic Caucasians: identification
of a Pro12Ala PPAR gamma 2 missense mutation. Biochem
Biophys Res Commun 1997; 241: 270–4.
17. Altshuler D, Hirschhorn JN, Klannemark M, et al. The
common PPARgamma Pro12Ala polymorphism is associated
with decreased risk of type 2 diabetes. Nat Genet 2000; 26:
76–80.
18. Ristow M, Muller-Wieland D, Pfeiffer A, Krone W, Kahn CR.
Obesity associated with a mutation in a genetic regulator of
adipocyte differentiation. N Engl J Med 1998; 339: 953–9.
19. Barroso I, Gurnell M, Crowley VE, et al. Dominant negative
mutations in human PPARgamma associated with severe
insulin resistance, diabetes mellitus and hypertension. Nature
1999; 402: 880–3.
20. Hegele RA, Cao H, Frankowski C, Mathews ST, Leff T.
PPARG F388L, a transactivation-deficient mutant, in familial
partial lipodystrophy. Diabetes 2002; 51: 3586–90.
21. Agarwal AK, Garg A. A novel heterozygous mutation in
peroxisome proliferator-activated receptor-gamma gene in a
patient with familial partial lipodystrophy. J Clin Endocrinol
Metab 2002; 87: 408–11.
22. Haverfield EV, McKenzie CA, Forrester T, et al. UGT1A1
variation and gallstone formation in sickle cell disease. Blood
2005; 105: 968–72.
23. Chaar V, Keclard L, Diara JP, et al. Association of UGT1A1
polymorphism with prevalence and age at onset of chole-
lithiasis in sickle cell anemia. Haematologica 2005; 90: 188–99.
24. Wasmuth HE, Keppeler H, Herrmann U, Schirin-Sokhan R,
Barker M, Lammert F. Coinheritance of Gilbert syndrome
associated UGT1A1 mutation increases the gallstone risk in
cystic fibrosis. Hepatology 2006; 43: 738–41.
25. Fischer S, Dolu MH, Zundt B, Meyer G, Geisler S, Jungst D.
Apolipoprotein E polymorphism and lithogenic factors in
gallbladder bile. Eur J Clin Invest 2001; 31: 789–95.
Liver International (2007)918 c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard
Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease Schafmayer et al.
26. Juvonen T, Kervinen K, Kairaluoma MI, Lajunen LH, Kesa-
niemi YA. Gallstone cholesterol content is related to apolipo-
protein E polymorphism. Gastroenterology 1993; 104:
1806–13.
27. Puppala S, Dodd GD, Fowler S, et al. A Genomewide Search
Finds Major Susceptibility Loci for Gallbladder Disease on
Chromosome 1 in Mexican Americans. Am J Hum Genet
2006; 78: 377–92.
28. Krawczak M, Nikolaus S, von Eberstein H, El Mokhtari NE,
Schreiber S. PopGen: population-based recruitment of
patients and controls for the analysis of complex genotype-
phenotype relationships. Commun Genet 2006; 9: 55–61.
29. Schafmayer C, Tepel J, Franke A, et al. Investigation of the
Lith1 candidate genes ABCB11 and LXRA in human gall-
stone disease. Hepatology 2006; 44: 650–7.
30. Volzke H, Baumeister SE, Alte D, et al. Independent risk
factors for gallstone formation in a region with high chole-
lithiasis prevalence. Digestion 2005; 71: 97–105.
31. Hampe J, Wollstein A, Lu T, et al. An integrated system for
high throughput taqman based SNP genotyping. Bioinfor-
matics 2001; 17: 654–5.
32. Dudbridge F. Pedigree disequilibrium tests for multilocus
haplotypes. Genet Epidemiol 2003; 25: 115–21.
33. Dupont WD, Plummer WD. PS power and sample size
program available for free on the Internet. Controlled Clin
Trials 1997; 18: 274.
34. Altshuler D, Brooks LD, Chakravarti A, Collins FS, Daly MJ,
Donnelly P. A haplotype map of the human genome. Nature
2005; 437: 1299–320.
35. de Bakker PI, Yelensky R, Pe’er I, Gabriel SB, Daly MJ,
Altshuler D. Efficiency and power in genetic association
studies. Nat Genet 2005; 37: 1217–23.
36. Barrett JC, Fry B, Maller J, Daly MJ. Haploview: analysis and
visualization of LD and haplotype maps. Bioinformatics 2005;
21: 263–5.
37. Lyons MA, Wittenburg H, Li R, et al. Lith6: a new QTL
for cholesterol gallstones from an intercross of CAST/Ei
and DBA/2J inbred mouse strains. J Lipid Res 2003; 44:
1763–71.
38. Wittenburg H, Lammert F, Wang DQ, et al. Interacting
QTLs for cholesterol gallstones and gallbladder mucin in
AKR and SWR strains of mice. Physiol Genom 2002; 8:
67–77.
39. Claus EB, Risch NJ, Thompson WD. Age at onset as an
indicator of familial risk of breast cancer. Am J Epidemiol
1990; 131: 961–72.
40. Pankratz N, Byder L, Halter C, et al. Presence of an APOE4
allele results in significantly earlier onset of Parkinson’s
disease and a higher risk with dementia. Mov Disord 2005;
21: 45–9.
41. Pankratz VS, de Andrade M, Therneau TM. Random-effects
Cox proportional hazards model: general variance compo-
nents methods for time-to-event data. Genet Epidemiol 2005;
28: 97–109.
42. Huddy GPB. A study of the family histories of 300 patients
suffering from chronic upper abdominal lesions. Lancet 1925;
2: 276–8.
43. van der Linden W, Simonson N. Familial occurrence of
gallstone disease. Incidence in parents of young patients.
Hum Hered 1973; 23: 123–7.
44. Gilat T, Feldman C, Halpern Z, Dan M, Bar-Meir S. An
increased familial frequency of gallstones. Gastroenterology
1983; 84: 242–6.
45. Sarin SK, Negi VS, Dewan R, Sasan S, Saraya A. High familial
prevalence of gallstones in the first-degree relatives of gall-
stone patients. Hepatology 1995; 22: 138–41.
46. Kratzer W, Kachele V, Mason RA, et al. Gallstone prevalence
in Germany: the Ulm Gallbladder Stone Study. Dig Dis Sci
1998; 43: 1285–91.
47. Ke X, Durrant C, Morris AP, et al. Efficiency and consistency
of haplotype tagging of dense SNP maps in multiple samples.
Hum Mol Genet 2004; 13: 2557–65.
48. Hirschhorn JN, Daly MJ. Genome-wide association studies
for common diseases and complex traits. Nat Rev Genet 2005;
6: 95–108.
49. Helms C, Cao L, Krueger JG, et al. A putative RUNX1
binding site variant between SLC9A3R1 and NAT9 is asso-
ciated with susceptibility to psoriasis. Nat Genet 2003; 35:
349–56.
50. Tokuhiro S, Yamada R, Chang X, et al. An intronic SNP in a
RUNX1 binding site of SLC22A4, encoding an organic cation
transporter, is associated with rheumatoid arthritis. Nat
Genet 2003; 35: 341–8.
51. Schaffler A, Barth N, Schmitz G, Zietz B, Palitzsch KD,
Scholmerich J. Frequency and significance of Pro12Ala and
Pro115Gln polymorphism in gene for peroxisome prolifera-
tion-activated receptor-gamma regarding metabolic para-
meters in a Caucasian cohort. Endocrine 2001; 14: 369–73.
52. Teng BB, Ochsner S, Zhang Q, Soman KV, Lau PP, Chan.
Mutational analysis of apolipoprotein B mRNA editing
enzyme (APOBEC1) structure-function relationships of
RNA editing and dimerization. J Lipid Res 1999; 40: 623–35.
Liver International (2007)c� 2007 The Authors. Journal compilation c� 2007 Blackwell Munksgaard 919
Schafmayer et al. Lith6 candidate genes APOBEC1 and PPARG in human gallstone disease