Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Variants in ABCG8 and TRAF3 genes confer risk for gallstone disease in admixed Latinos with Mapuche Native American ancestry


Latin Americans and Chilean Amerindians have the highest prevalence of gallstone disease (GSD) and gallbladder cancer (GBC) in the world. A handful of loci have been associated with GSD in populations of predominantly European ancestry, however, they only explain a small portion of the genetic component of the disease. Here, we performed a genome-wide association study (GWAS) for GSD in 1,095 admixed Chilean Latinos with Mapuche Native American ancestry. Disease status was assessed by cholecystectomy or abdominal ultrasonography. Top-10 candidate variants surpassing the suggestive cutoff of P < 1 × 10−5 in the discovery cohort were genotyped in an independent replication sample composed of 1,643 individuals. Variants with positive replication were further examined in two European GSD populations and a Chilean GBC cohort. We consistently replicated the association of ABCG8 gene with GSD (rs11887534, P = 3.24 × 10−8, OR = 1.74) and identified TRAF3 (rs12882491, P = 1.11 × 10−7, OR = 1.40) as a novel candidate gene for the disease in admixed Chilean Latinos. ABCG8 and TRAF3 variants also conferred risk to GBC. Gene expression analyses indicated that TRAF3 was significantly decreased in gallbladder (P = 0.015) and duodenal mucosa (P = 0.001) of GSD individuals compared to healthy controls, where according to GTEx data in the small intestine, the presence of the risk allele contributes to the observed effect. We conclude that ABCG8 and TRAF3 genes are associated with GSD and GBC in admixed Latinos and that decreased TRAF3 levels could enhance gallbladder inflammation as is observed in GSD and GSD-associated GBC.


Gallstone disease (GSD) is a complex gastrointestinal disorder defined by the presence of gallstones in the gallbladder, most of the time made of cholesterol compounds1. The presence of gallstones is common in the world population (~10–20% of presence in adults) and although their appearance can remain silent throughout life, more than 20% of the patients develop symptoms that include intense abdominal pain, jaundice, fever, nausea and vomiting, requiring medical intervention2. Main complications of GSD are acute cholecystitis, acute pancreatitis and bile duct obstruction3, which are generally treated by surgical removal of the gallbladder (cholecystectomy). In Chile, these procedures account for more than 40,000 interventions each year4, with a net cost of more than US $25 million to the public healthcare system. Additionally, GSD is the main risk factor for gallbladder cancer (GBC), a disease that presents a high mortality rate in Chile, with 38.2 deaths per 100,000 inhabitants5. In the majority of cases, GBC is an adenocarcinoma characterized by high lethality due to late diagnosis and the ineffectiveness of chemotherapy/radiotherapy. With the removal of the gallbladder and its stones, GBC risk is considerably reduced6.

Epidemiological studies have shown that ethnicity plays a major role in the prevalence of GSD. Asian and African countries have the lowest prevalence of the disease (<10%), followed by European (~20%) and American (>40%) countries2. Notably, the prevalence in adult Chilean Mapuche Amerindians is estimated to be 49.4% and 12.3% for women and men, respectively7. Likewise, adult Chilean women and men from the general population, resulting from the admixture of Europeans (mainly from Spain) and indigenous Amerindians (mainly Mapuches), have an estimated prevalence of 36.7% and 13.1%, respectively7. Studies in families and case-controls populations have shown that GSD has a strong genetic component8,9. The most important genetic risk factor is the ATP-binding cassette, sub-family G (WHITE), member 8, ABCG8 gene10, an hepatic sterolin transporter where the associated polymorphism (p.D19H, rs11887534) is linked to a gain-of-function responsible for the hyper secretion of cholesterol-saturated bile11. Another important risk factor is the UDP glucuronosyl-transferase 1 family, polypeptide A1, UGT1A1 gene, which has been significantly associated to GSD only in men. Both signals have been replicated in different populations across the world, including admixed Chileans10,12,13, yet they only explain a small portion of the population attributable risk of the disease (ABCG8 PAR = 11.2%, UGT1A1 PAR = 9.9%, combined PAR = 21.2%)14. Recently, a meta-analysis involving 8,720 GSD cases and 55,152 controls identified four additional risk factors, including the Transmembrane 4 L six family member 4, TM4SF4 gene (rs9843304), the Sulfotransferase family 2 A member 1 SULT2A1 gene (rs2547231), the glucokinase regulator, GCKR gene (rs1260326) and the cytochrome P450 family 7 subfamily A member 1, CYP7A1 gene (rs6471717)15. However, none of these novel risk factors have been studied in the Chilean population.

Here we report the results of the first large GWAS for a prevalent complex disease in the admixed Chilean population. We hypothesized that a high-density GWAS for GSD in admixed Latinos could identify population specific variants, define the GSD genetic landscape and reveal novel pathological mechanisms.


Discovery GWAS for GSD in an admixed Chilean population

The Discovery GWAS stage involved 1,235 admixed Chileans Latinos with Mapuche Native American Ancestry genotyped using the Affymetrix Axiom® Genome-Wide LAT 1 World Array 4 (see complete pipeline in Supplementary Fig. 1). After genotype calling and quality controls, 1,095 individuals (529 GSD cases and 566 controls) and 677,835 SNPs remained for analysis (Table 1). We imputed genotypes in these individuals using the 1000 Genomes Project Phase 3 reference panel16 and achieved a total number of 9,433,911 SNPs and Indels. We tested for genetic association with GSD by performing a logistic regression analysis adjusted by age (in years; cases = 51.32 ± 10.67, controls = 49.87 ± 9.57, one-way ANOVA P = 0.018), since other known associated covariates such as sex (women%; cases = 92.43%, controls = 92.57%, Z score test P = 0.928), body mass index (BMI; cases = 29.44 ± 4.28, controls = 29.66 ± 3.94, one-way ANOVA P = 0.376) and native American ancestry proportion (cases = 46.3% ± 7.0%, controls = 45.5% ± 7.5%, one-way ANOVA P = 0.069), did not differ significantly between cases and controls.

Table 1 Clinical characteristics of patients in the Discovery GWAS and independent replication populations.

Genome-wide association results are presented in Fig. 1 and consider a genomic inflation factor λ = 1.02 (Supplementary Fig. 2). While we did not detect genome-wide significant associations, several suggestive signals were observed (P < 1 × 10−5). The top-10 most significant loci were: ELMO1 (rs4446645, P = 3.36 × 10−7), 5q34 locus (rs10463138, P = 2.47 × 10−6), 4q12 locus (rs74537816, P = 3.54 × 10−6), TRAF3 (rs368550004, P = 3.92 × 10−6), 9p21.1 locus (rs4879592, P = 4.66 × 10−6), OLFML2B (rs10918361, P = 4.91 × 10−6), 3p22.2 locus (rs73827633, P = 5.06 × 10−6), ABCG8 (rs11887534, P = 5.24 × 10−6), 11p15.3 locus (rs147367002, P = 7.63 × 10−6) and TRPV1 (rs7223530, P = 8.07 × 10−6) (Table 2). Aside from ABCG8, none of the previous GWAS signals associated with GSD reported in a recent large meta-analysis of individuals with European ancestry15 reached suggestive genome-wide significance (Supplementary Table 1).

Figure 1
figure 1

Genome-wide association results for GSD in admixed Chileans in the discovery stage. Manhattan plot depicting the association P values for all good quality variants. Red line shows genome-wide significance threshold (P < 5 × 10−8). Top-ten candidate variants surpassing the suggestive genome-wide significance threshold (blue line, P < 1 × 10−5) are shown, which were taken for further replication.

Table 2 Top-10 candidate variants associated with GSD in admixed Chileans in the discovery and replication stages.

Association of ABCG8 and TRAF3 variants with GSD in an independent Chilean population

The top-ten candidate variants were tested for replication in an independent admixed Chilean population composed of 1,643 individuals (626 cases and 1,017 controls) from Santiago de Chile (Table 1). Replication was performed by real-time qPCR with TaqMan probes for most leading variants, with the exception of SNP3 (4q12), SNP4 (TRAF3) and SNP9 (11p15.3), which were examined by proxy SNPs in high linkage disequilibrium (LD r2 > 0.88). Nine of the ten candidates were in Hardy-Weinberg equilibrium, where only the SNP3 showed a significant deviation (p = 2 × 10−40, Table 2), therefore association results from this candidate should be taken with caution. After genotyping and logistic regression analyses adjusted for age (in years; cases = 59.57 ± 12.47, controls = 48.26 ± 12.19, one-way ANOVA P < 0.001), sex (women%; cases = 84.82%, controls = 63.12%, Z score test P < 0.001), BMI (cases = 30.74 ± 5.92, controls = 28.61 ± 5.38, one-way ANOVA P < 0.001) and type 2 diabetes (T2D%; cases = 40.73%, controls = 14.36%, Z score test P < 0.001), we successfully replicated GSD association only of selected variants within the ABCG8 (rs11887534, P = 0.001, OR = 1.59, CI 95% = 1.20-2.11) and TRAF3 (rs12882491, P = 0.003, OR = 1.30, CI 95% = 1.09–1.54) genes (Fig. 2, Table 2). Both variants passed Bonferroni correction for 10 tests (P < 0.005) and displayed an effect with the same direction as it was observed in the discovery stage of the GWAS.

Figure 2
figure 2

Regional association plots for the ABCG8 and TRAF3 signals in the GWAS discovery stage. Locus zoom for ABCG8 (a) and TRAF3 (b) signals in chromosome 2 and 14, respectively. To the left is the P value related to GSD in a log10 scale. SNPs with the highest association to GSD are colored purple. The insert panel denotes the imputed quality score (r2) and the appropriate SNPs are colored in the plot.

To assess for the combined effect between ABCG8 and TRAF3 SNPs with GSD, we performed a meta-analysis17 between the Discovery and Replication stages and observed an increased significance in the association of ABCG8 (rs11887534, P = 3.24 × 10−8, OR = 1.74, CI 95% = 1.45–2.02) and TRAF3 (rs12882491, P = 1.11 × 10−7, OR = 1.40, CI 95% = 1.20–1.60) variants (Table 2). Notably, the OR obtained for ABCG8 represents a PAR of 7.1%, while for TRAF3 (rs12882491, risk allele C) the OR of 1.40 (CI 95% = 1.20–1.60) represents a PAR of 20.2%. Taking together, the combined effect of both associations gives an estimated PAR of 25.9%, meaning that if their contribution for the disease is hypothetically eliminated18, there would be 25.9% fewer affected individuals in the population. In addition, SNPxSNP epistasis analysis showed non-significant interaction between ABCG8 and TRAF3 association signals (P = 0.399), suggesting that the genetic contribution of both risk factors is independent from each other. Altogether these results confirm and replicate the previously observed associations of ABCG8 with GSD and identify TRAF3 as a novel candidate gene associated with the disease in the Chilean Admixed population.

ABCG8 and TRAF3 variants are associated with GBC

Although GBC is the most severe complication of GSD5,6,19, few studies have investigated the effect of ABCG8 as a genetic determinant for this cancer20,21. We therefore examined the orientation and magnitude of the effect for ABCG8 and TRAF3 variants in a Chilean population of GBC patients (n = 397 women), using sex-matched controls from the replication population (n = 667 women) (Table 1). After logistic regression analyses adjusted by age (in years; cases = 62.65 ± 13.08, controls = 43.38 ± 12.13, one-way ANOVA P < 0.001), we observed that both SNPs were associated with GBC (ABCG8 rs11887534: P = 6.9 × 10−4, OR = 1.77, CI 95% = 1.27–2.45; TRAF3 rs12882491: P = 0.045, OR = 1.24, CI 95% = 1.004–1.53), with same direction of the effect and similar risk allele frequency (RAF), as it was observed in the discovery and replication populations (RAFABCG8: GBC = 0.14, discovery = 0.14, replication = 0.12; RAFTRAF3: GBC = 0.66, discovery = 0.69, replication = 0.66) (Fig. 3). We also observed that the association of ABCG8 and TRAF3 variants with GBC was significant only when GBC samples were compared to gallstone-free individuals (Supplementary Table 2), indicating that the risk contribution to GBC was explained by the presence of gallstones. These results support the genetic association of ABCG8 with GBC and reveal TRAF3 as a novel marker for the disease.

Figure 3
figure 3

Effect size for ABCG8 and TRAF3 association signals in admixed Chilean populations. Forest plot calculated for GSD and GBC samples. The effect is shown in OR values and their 95% confidence intervals. Sample size for GSD case-control populations and for GBC cases are shown in parenthesis.

Replication Analyses for ABCG8 and TRAF3 variants in Europeans populations

Since we originally identified ABCG8 as a risk factor for GSD in a German population from Kiel10,11, we therefore examined ABCG8 rs11887534 and TRAF3 rs12882491 variants in an extended sample from the original POPGEN-Kiel population consisting in 1,938 individuals (1,027 cases and 911 controls)22, as well as in 4,154 individuals (882 cases and 3,272 controls) from the north-east side of Germany identified as SHIP-Greifswald23 (Table 1). We found that ABCG8 rs11887534 replicated both in POPGEN-Kiel (P = 5.36 × 10−10, OR = 2.00, CI 95% = 1.61–2.50) and SHIP-Greifswald (P = 1.29 × 10−6, OR = 1.87, CI 95% = 1.45–2.41) populations. Interestingly, no replication was observed for TRAF3 rs12882491 either in the POPGEN-Kiel (P = 0.65, OR = 0.97, CI 95% = 0.84–1.11) or the SHIP-Greifswald (P = 0.39, OR = 1.06, CI 95% = 0.93–1.20) samples. Moreover, a difference in orientation and/or magnitude of the effect with the Chilean samples was also reflected in terms of risk allele frequency (RAFTRAF3 Chilean discovery = 0.69, replication = 0.66, POPGEN-Kiel = 0.34, SHIP-Greifswald = 0.35). Altogether, these results demonstrate a consistent effect of ABCG8 association with GSD and indicate an ethnic-specific effect for the TRAF3 variant in admixed Latinos with Native Amerindian ancestry.

TRAF3 expression in human gallbladder and duodenal samples

TRAF3 is a member of the tumor necrosis factor (TNF) receptor-associated factor (TRAF) protein family that mediates signal transduction during the activation of immune and inflammatory responses24,25,26. To date, seven TRAF genes are known in humans (TRAF1 to TRAF7) and the products encoded by these genes are involved in cytokine production and cell survival27. Considering that the acute and chronic inflammation of the gallbladder (cholecystitis) are known hallmarks of GSD pathophysiology28, we explored the expression levels of the TRAF3 protein in 9 gallbladder mucosa samples (4 cases and 5 controls) and mRNA in 25 gallbladder (12 cases and 13 controls) and 41 duodenal (22 cases and 19 controls, independent from the mRNA samples) samples. We used the duodenal mucosa as a comparison for three reasons: First, it is a tissue in contact with bile coming from the liver and gallbladder; second, it is easily accessible by endoscopy in order to obtain biopsies from affected and healthy individuals; and third, evidence show the existence of a similar gene expression pattern compared with both liver and gallbladder tissues29. Immunohistochemical staining in the gallbladder mucosa revealed that TRAF3 protein was observed in different gallbladder cell types such as the epithelium, smooth muscle fibers, arterioles and veins (Fig. 4a). Notably, we found a significant decrease of protein levels in the duodenal mucosa of GSD cases compared to control individuals (P < 0.001, two-tailed t-test; Fig. 4b,c and Supplementary Fig. 3) and also observed a significant decrease of the gene transcripts in the set of mRNA samples from duodenal (P < 0.001, two-tailed t-test) and gallbladder mucosa (P = 0.015, two-tailed t-test) (Fig. 4d). These results show lower levels of TRAF3 in GSD affected individuals, suggesting a functional contribution of this gene during the inflammatory response observed during the onset or development of the disease.

Figure 4
figure 4

TRAF3 expression in human gallbladder and duodenal tissues. (a) Immunohistochemical analysis of TRAF3 expression in normal human gallbladder. Left panel: Negative control without primary antibody; Middle panel: Positive staining was observed for TRAF3 in the cell epithelium (1), muscle fibers (2), arterioles (3), and veins (4); Right panel: Zoom-in (60x) showing TRAF3 localization in the mucosal epithelium. (b) Western blot for TRAF3 in biopsies of duodenal mucosa tissue from GSD cases and control individuals. (c) Determination of TRAF3 protein levels in the duodenal mucosa (as is observed in b). (d) Differential expression analyses for TRAF3 mRNA in duodenal (left panel) and gallbladder mucosa tissues (right panel) in GSD cases and control samples.


Herein we report that common genetic variants within the ABCG8 and TRAF3 genes confer risk to GSD and GBC in the admixed Chilean population. The ABCG8 polymorphism rs11887534 (D19H) has been reported in several world populations including Latinos15, thus the results reported by the present study serves as a novel replication and adds strength to the observed association of the gene. In the discovery stage, the leading ABCG8 SNP showed a lower magnitude of association, compared with the estimated from the German population10, which could be explained by the difference in the risk allele frequency (C allele) between Chilean and German controls (0.079 vs. 0.05, respectively). This could be a reason for not being able to detect in the discovery stage a genome-wide signal (P < 5 × 10−8) in the ABCG8 locus, suggesting that larger sample sizes are required to reach such significance level in future studies on this population.

ABCG8 was consistently associated with GSD in all the populations analyzed in this study, however, TRAF3 association was only observed in Chilean samples. These results suggested that TRAF3 association might have an ethnic-specific effect. Interestingly, data from the 1000 genomes project from Peru (PEL), Mexico (MXL), Colombia (CLM) and Puerto Rico (PUR) show a higher risk allele frequency compared with the one from the overall European population (RAF PEL = 0.84, MXL = 0.55, CLM = 0.53, PUR = 0.49, EUR = 0.34) indicating that would be of interest to validate the TRAF3 association in closely related Latino populations like Peru, Argentina and Bolivia, that also have a high GSD prevalence30,31,32.

GSD has known confounding factors such as age, gender, BMI and T2D, and past studies have considered some of these biological variables into their association analyses10,15,33,34. Furthermore, admixed populations present a special challenge in GWAS analyses due to the presence of 2 or more ancestry components that needs to be considered when performing the association testing35,36. In this regard, we performed 2 additional analyses to examine whether the association of ABCG8 and TRAF3 variants with GSD could be affected by these confounding factors. First, besides age as the main covariate, we included the 5 first eigenvectors from the principal component analysis (PCA, see methods) to take into account any hidden population structure that could still be present. Second, we adjusted the logistic regression by age, the aforementioned 5 eigenvectors and the other known GSD covariates such as gender and BMI to test their effect regardless of their distribution in cases and controls. Remarkably, we found that these additional adjustments did not alter the selection of ABCG8 or TRAF3 as candidate genes for the replication stage since the association P values for their leading variants remained above the significance cutoff of P < 1 × 10−5 (Supplementary Table 3).

Here we also observed that ABCG8 and TRAF3 variants are significantly associated with GBC in admixed Native Americans. For instance, the ABCG8 rs11887534 variant has been previously associated with this malignancy in China21, India20 and European populations37 and most of the studies show that the GBC risk is more pronounced in patients that also had biliary stones in the gallbladder, as it is likely observed in this Chilean cohort (i.e. composed of 90% women GBC patients with cholesterol gallstones). This agrees with our findings where both ABCG8 and TRAF3 variants only contribute to risk when comparing GBC subjects with gallstone free controls38. A recent report from a group in India showed a GWAS scan for GBC in their population and found that the genes ABCB1 and ABCB4 were significantly associated with GBC38. We looked for their association in our GSD GWAS discovery population and found that none of the 3 reported SNPs showed at least nominally significant P values (Supplementary Table 4). This could be indicating that these new GBC risk factors are not be necessarily associated to the GSD phenotype linked to GBC in the Chilean population.

TRAF3 belongs to the TNF receptor associated factor (TRAF) protein. TRAF3 function has been related to the promotion of autoimmunity and predisposition to various cancers39, including multiple myeloma40, Hodgkin lymphoma41 and splenic neoplasms42. The protein physically interacts with the NF-κB-inducing kinase (NIK) promoting NIK degradation by the proteasome, which results in the inactivation of the non-canonical NF-κB pathway43. Indeed, siRNA knockdown of TRAF3 in human cancer cell lines stabilizes NIK and activates NF-κB dependent transcription44. Interestingly, recent evidence indicates that activation of NF-κB results in invasion, lymphangiogenesis and tumor growth in GBC tissue and cell lines45,46, and therefore genetic alterations in one of its modulators could be directly involved in GBC pathogenesis and may serve as a target of future therapies and prevention.

Inflammation is a hallmark of GSD pathogenesis and it has been suggested that the formation of biliary stones is preceded by histopathologic alterations in the gallbladder wall that indicate tissue inflammation47, including edema, increased wall thickness, decreased motility and altered transport function in the epithelium. Therefore, a crosstalk between cholesterol stone formation and inflammatory processes might exist that promote the development or progression of the disease and suggests that ABCG8 and TRAF3 could be both relevant actors in the pathology. First, it is known that the ABCG8 variant promotes gallstone formation by enhancing the efflux of cholesterol-saturated bile11. Second, lower expression levels of TRAF3 in duodenal mucosa and gallbladder epithelium of GSD patients (as observed in the present study) could be responsible for increased production of pro-inflammatory factors, such as NF-κB, IL-6 and IL-12B, as well as for decreased activity of anti-inflammatory factors such as type 1 interferon and IL-1044,48,49,50. Notably, the IL-10 has already been associated with GSD and GBC51. Interestingly, according to the Genotype-Tissue Expression (GTEx) project database, which offers gene expression and quantitative trait loci associations from 53 human tissues52, individuals with the TRAF3 risk allele (rs12882491, C) present decreased TRAF3 expression levels in the small intestine, the closer tissue to our duodenum data (Supplementary Fig. 4). Although the correlation is not statistically significant, the direction of the effect agrees with our findings, indicating that the GWAS risk variant could be functionally linked to the decreased gene levels observed in the study.

In summary, our results confirm the association of ABCG8 and identify TRAF3 as a novel candidate gene for GSD and GBC in the high-risk Chilean admixed population. We hypothesize that both genes are functionally involved in the cascade of events that triggers gallstone formation and the inflammatory response observed in affected patients. Further research is needed to determine the role of the variations within the TRAF3 gene that could explain the observed differential gene expression and its contribution to the pathophysiology of the disease.


Ethical statement

All procedures involving genetic and clinical data usage from Chilean GSD and GBC patients were approved by the Ethical Scientific Committee at the Pontificia Universidad Catolica de Chile and were conducted in accordance with the guidelines of the National Commission on Science and Technology (CONICYT-Chile). All participants provided informed consent. In the German replication samples, Ethical Scientific Committees have previously approved these studies22,53.

Study participants

Discovery stage

The Chilean GWAS discovery population consisted in 1,235 individuals, belonging to the ANCORA family health centers in Santiago de Chile (Puente Alto, La Florida and La Pintana). This sample constitutes an admixed group (European-Amerindian) aged between 20–80 years old, of middle-low income, from an urban area and representative of the general Chilean population. Cases were defined as patients with reported cholecystectomy or by the presence of gallstones in their gallbladder assessed by ultrasound. Control individuals had normal gallbladders, with neither a history of gallbladder surgery nor gallstone findings in ultrasound. Since women have a higher risk to develop gallstones compared to men, the population was mainly composed of women (1,013 individuals: 489 cases and 524 controls), corresponding to the 93.05% of the sample. After genotyping and QCs (see ahead), 1,095 individuals remained for further analysis (529 cases and 566 controls, Table 1). To reduce the contribution of confounding factor in the following genetic association analyses, cases and controls were matched for known GSD covariates and comorbidities, including age at sampling, sex distribution and BMI (Table 1). We excluded T2D individuals from the discovery stage in order to reduce the number of factors to consider in the adjusted analyses.

Chilean GSD replication stage

The independent replication population consisted of 1,643 Chilean individuals (626 cases and 1,017 controls), coming from the ANCORA family health centers and from a population base study in La Florida, which has been previously described7. Information regarding age, sex, BMI and T2D was gathered for all individuals to control for covariates/comorbidities in the genetic association analyses.

Chilean gallbladder cancer (GBC) population

A secondary Chilean replication population was derived from GBC patients, with more than 90% presenting gallstones. The population is composed of a total number of 397 incident and prevalent women cases, recruited from different health centers of the country including: Hospital Sótero del Río, Clinical Hospital of the Pontificia Universidad Católica de Chile, Hospital Regional de Valdivia, Hospital Regional de Temuco and Hospital Regional de Concepción. GBC phenotype was defined according to criteria and steps described by the Classification of Malignant Tumors and the American Joint Committee on Cancer (TNM-AJCC) where only adenocarcinomas were included. Each histological cut was reviewed and selected by two independent expert pathologists. After delimiting the healthy tissue, either adjacent to the tumor or other non-tumoral tissues such as liver or cystic ganglion, blocks were selected for sequential cuts of 10 µm thickness and used for extraction of genomic DNA. DNA concentration and integrity were determined by the NanoDrop ND-1000® Spectrophotometer and by agarose gel electrophoresis, respectively.

European replication populations

We analyzed two European GSD population from Germany. The first population comes from the POPGEN biobank of the Medicine Faculty at the University of Kiel, consisting in 1,938 individuals (1,027 cases and 911 healthy controls) where GSD status was determined by cholecystectomy or ultrasound22. Considering that GSD prevalence increases with age10, the statistical analysis performed on this sample did not include age as a covariate, since stone-free controls were chosen and they had a higher median age (59.43 years old) than affected individuals (45.5 years old). The second population comes from the Study of Health in Pomerania (SHIP) from the Greifswald region in northeast Germany, comprising 4,154 subjects (882 GSD cases and 3,272 healthy controls) where GSD cases were identified by cholecystectomy/ultrasound and control individuals by senlf-reports23.

Genomic DNA extraction and genotyping

Genomic DNA was extracted from samples of the Chilean discovery population, replication and GBC incident cases (alive patients) using peripheral blood (Invisorb Blood Universal kit, Invitek, Berlin, Germany). For GBC prevalent cases, DNA was extracted from normal (non-tumoral) tissue samples using the QIAamp DNA-Mini-Kit © (Qiagen, Mainz, Germany). Extracted DNA was quantified with a nanodrop spectrophotometer to a fixed concentration of 50 ng/μl. For the 2 European populations, DNA extraction methods is described in their respective publications. Genome-wide genotyping on the Chilean discovery population was performed with the Affymetrix Axiom® Genome-Wide LAT 1 World Array 4 at the Cologne Center for Genomics (University of Cologne, Germany). The microarray includes probes to identify 818,155 SNPs and it has been optimized to enhance the capture of variants in populations with Amerindian ancestry54. Genotype calling was done with Affymetrix Power Tools software, including the SNPolisher v.1.17.0 R package. For the POPGEN-Kiel and SHIP-Greifswald European populations, the Axiom® Genome-Wide Platform EUR 1 = (674,518 markers) and the Affymetrix Genome-Wide Human SNP Array 6.0 (1,879,489 markers) were used for genome-wide genotyping, respectively. For the Chilean replication and GBC populations, genotyping was carried out by real-time quantitative PCR (qPCR) using TaqMan probes and the TaqMan Universal Master Mix (Thermo Fisher Scientific Inc, Foster City, USA). qPCR reactions and genotype calling were performed in the StepOnePlusTM System (Thermo Fisher Scientific Inc, Foster City, USA).

GWAS quality control procedures

After genotyping of the initial 1,235 samples for the Chilean discovery stage, we used the Affymetrix Power tools protocol and removed 6 individuals and 5,136 variants, leaving 1,229 subjects and 813,019 genotypes, then converted them to PLINK v1.9 format55,56 for further quality controls57. We performed an identity by descent (IBD) analysis to detect cryptic relationships in the GWAS, identifying 115 individuals having an IBD score > 0.185, indicating close relatedness (2nd degree and further). In order to detect population substructure, we performed a principal component analysis (PCA) with the SMARTPCA tool included in the EIGENSOFT 6.0.1 package software58 using the 1000 genome project phase 3 as a reference population, similar to what was described in a recent study that included the present GWAS cohort59. We identified 15 individuals as outliers (>6 standard deviations) and then used the first 5 eigenvectors as covariates in an extended GWAS analysis. One individual had inconsistent sex identification (PLINK sex check), eight had high genotype heterozygosity rates (>5 standard deviations) and no individuals had < 97% of genotype successfully called. Combining all the individuals detected with the methods described above, we obtained a total number of 134 unique samples and excluded them, leaving 1,095 good quality samples (529 cases and 566 controls). At the variant level, we identified 2,940 SNPs deviating from Hardy-Weinberg equilibrium (HWE < 1 × 10-6) in controls, 170 SNPs showing 1% or more genotype missingness (GENO > 0.01) and 122,074 SNPs had a minor allele frequency of less than a 1% in the total population (MAF < 0.01), giving a total number of 125,184 SNPs removed, and leaving 677,835 for further analysis. For the POPGEN-Kiel population, samples were imputed with IMPUTE2 using the 1000 Genomes phase 3 reference panel (build 37). TRAF3 and ABCG8 variants were selected from all variants surpassing MAF > 0.01 and an INFO > 0.4. For the SHIP-Greifswald population, TRAF3 and ABCG8 variants summary statistics were obtained from the study authors53.

Global Ancestry Estimation

Global ancestry proportions for each individual in the Chilean discovery population was calculated with the ADMIXTURE v.1.3.0 software60 using the reference panel of the 1000 Genomes project Phase-3 and 11 individuals from the Mapuche-Huilliche ethnic group as a Chilean Native American reference panel (HUI)59. The mean proportion of the Amerindian ancestry was calculated in cases and controls and tested for significant deviation using a one-way ANOVA test.

Genotype imputation and GWAS

Genotypes were imputed using IMPUTE2 v2.3.1 without the prephasing step in order to gain accuracy61,62. We used the genotype reference panel coming from the 1000 Genomes Project Phase 3 for autosomic variation (October 2014 release) and 1000 Genomes Project Phase 1 for the X chromosome (March 2012 release). After imputation, we selected SNPs and Indels with an INFO score > 0.4, MAF > 0.01 and HWE > 1 × 10−6 leaving 9,433,911 analysis-ready variants. Information on whether a candidate variant comes from imputation or actual chip genotyping is presented in Table 1. Genetic association analyses were performed using the score statistic as an additive logistic regression model implemented in SNPTEST v2.563 and adjusted for age as a covariate. Other GSD known confounding factors such as gender, BMI and TD2 were either matched or non-existent in cases and controls (see results), therefore we did not include them in the main association analysis in order to avoid overfitting. The Population Attributable Risk (PAR) was calculated for ABCG8 and TRAF3 leading SNPs in the combined Chilean discovery and replication population, using the formula: PAR = PRF × (RR − 1)/1 + PRF × (RR − 1) × 100, where PRF is the prevalence of the risk factor in the general population. The allelic odds ratio (OR) was used as an approximation for the estimated relative risk (RR) of disease due to exposure to the risk allele. Combined PAR for the 2 risk variants was calculated as: Combined PAR = 1 − (1 − PAR1) ×(1 − PAR2). SNPxSNP interaction analysis was performed with PLINK v1.9 epistasis function.

Genotyping in replication analyses

TaqMan probes used for the 10 candidate variants coming from the discovery stage are shown in the Supplementary Table 5. The TRAF3 leading indel (rs368550004) was genotyped using a high resolution melting (HRM) procedure, as described elsewhere64. Briefly, HRM was performed using specific DNA primers for TRAF3 (Forward primer = 5′-TCACATTCCATACAATTACC-3′, reverse primer = 5′-ATTAACAAGGAACAACCGAT-3′) and MeltDoctor High-Resolution Melting (HRM) Reagents (Applied Biosystems, cat: 4415440).

Statistical analyses in the replication populations

Replication association analysis was performed using additive logistic regression tests with PLINK v1.9, adjusting by age, sex, BMI and T2D status as they were available for the data. Fixed effect meta-analysis was performed between the discovery stage and variants showing successful replication after Bonferroni correction for 10 tests (P < 0.005), using the inverse-variance method implemented in the –meta-analysis command in PLINK. Forest plots were generated using the rmeta package v.2.16 in R. Regional association plots were generated using LocusZoom (v1.1)65.

Differential gene expression analysis of TRAF3

For expression and histological analyses (see below) human gallbladder mucosa samples were used as described66. Briefly, duodenal mucosa samples were extracted from gallstone and gallstone-free individuals that met the following criteria: women 18 to 35 years of age, non-obese (BMI 18–30 kg/m2), without significant pathologies and consumption of any drug 2 months prior to recruitment and not pregnant at the time of contact. Duodenal biopsies were taken from the second segment of the duodenum (distal to the ampulla of Vater), during gastrointestinal endoscopy performed at the Endoscopic Unit of the Gastroenterology Department of the Pontificia Univesidad Católica de Chile. Total RNA from was extracted using the PureLinkTM RNA Mini Kit (Ambion Life technologies, Carlsbad, USA) and reverse-transcribed into cDNA with a High-Capacity cDNA Reverse Transcription Kit (Thermo Fisher Scientific Inc., Carlsbad, USA). Quantitative RT-PCR from 25 gallbladders (12 cases and 13 controls) and 41 duodenal (22 cases and 19 controls) samples was performed using specific DNA primers for TRAF3 (Forward primer = 5′- TACAGCGTGTCAAGAGAGCATCGT-3′, reverse primer = 5′- CACAACCTCTGCTTTCATTCCGACAATAG-3′) and SYBR-Green Brilliant III Ultra-Fast SYBR-Green qPCR Master Mix (Agilent Technologies Inc., Santa Clara, USA). Data were expressed in arbitrary units and were normalized to RNA18S levels. Statistical significance was determined using a two-tailed t-test (*P < 0.05).

Western blot

Fifty micrograms of total proteins from duodenal mucosa lysate of 9 samples (5 control and 4 cases, all independent from the mRNA samples), was separated on a 12% SDS-PAGE gel and then transferred onto nitrocellulose membranes (Millipore, Billerica, MA). Using 5% nonfat dry milk for blocking, membranes were incubated overnight at 4 °C with rabbit anti-TRAF3 antibody (1:1000, ABCAM Ltd, Cambridge, UK) and with rabbit polyclonal ε-Cop antibody (1:5000, donated by Dr. Monty Krieger of the Biology Department, Massachusetts Institute of Technology, Cambridge, USA) for 1 h at room temperature. Goat anti-rabbit IgG-HRP were used as secondary antibodies (1:1000, Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA) for 1 h at room temperature. The signals were detected using the chemiluminescence kit ECL Western Blotting (Pierce Biotechnology, Inc., Rockford, IL, USA). Complete Western Blot gels from where Fig. 4b. was composed is presented in the Supplementary Fig. 3.

Immunohistochemistry analysis

Two-micron sections of gallbladder samples were fixed with formalin and embedded in paraffin. For TRAF3 immunohistochemistry analysis, the rabbit anti-TRAF3-ab155298 antibody was used (1:100, ABCAM, Ltd, Cambridge, UK) and was visualized using the Envision Flex/HRP SM802 complex method (Dako, Agilent Technologies Company, Santa Clara, CA 95051, USA). Labeled sections were examined and captured using an Olympus BX51 microscope (Olympus, Tokyo, Japan) with the software Stereo Investigator v.11.03 (MBF bioscience, Williston, VT, USA) and analyzed with the Image-Pro Express program (Media Cybernetics, Bethesda, MD, USA) as described previously66.

Data Availability

All genotype files for the Chilean discovery population are available upon request to the corresponding authors.


  1. Acalovschi, M. Cholesterol gallstones: from epidemiology to prevention. Postgrad Med J 77, 221–229 (2001).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  2. Lammert, F. et al. Gallstones. Nat Rev Dis Primers 2, 16024, (2016).

    Article  PubMed  Google Scholar 

  3. Lammert, F. & Miquel, J. F. Gallstone disease: from genes to evidence-based therapy. J Hepatol 48(Suppl 1), S124–135, (2008).

    CAS  Article  PubMed  Google Scholar 

  4. Bay, C. et al. Access to cholecystectomy among patients attended at primary family health centers. Rev Med Chil 144, 317–324, (2016).

    Article  PubMed  Google Scholar 

  5. Chianale, J., Valdivia, G., del Pino, G. & Nervi, F. Gallbladder cancer mortality in Chile and its relation to cholecystectomy rates. An analysis of the last decade. Rev Med Chil 118, 1284–1288 (1990).

    CAS  PubMed  Google Scholar 

  6. Nervi, F. [Cancer of the gallbladder in Chile]. Rev Med Chil 129, 979–981 (2001).

    CAS  Article  PubMed  Google Scholar 

  7. Miquel, J. F. et al. Genetic epidemiology of cholesterol cholelithiasis among Chilean Hispanics, Amerindians, and Maoris. Gastroenterology 115, 937–946 (1998).

    CAS  Article  PubMed  Google Scholar 

  8. Gilat, T., Feldman, C., Halpern, Z., Dan, M. & Bar-Meir, S. An increased familial frequency of gallstones. Gastroenterology 84, 242–246 (1983).

    CAS  PubMed  Google Scholar 

  9. Sarin, S. K., Negi, V. S., Dewan, R., Sasan, S. & Saraya, A. High familial prevalence of gallstones in the first-degree relatives of gallstone patients. Hepatology 22, 138–141 (1995).

    CAS  PubMed  Google Scholar 

  10. Buch, S. et al. A genome-wide association scan identifies the hepatic cholesterol transporter ABCG8 as a susceptibility factor for human gallstone disease. Nat Genet 39, 995–999, (2007).

    CAS  Article  PubMed  Google Scholar 

  11. von Kampen, O. et al. Genetic and functional identification of the likely causative variant for cholesterol gallstone disease at the ABCG5/8 lithogenic locus. Hepatology 57, 2407–2417, (2013).

    CAS  Article  Google Scholar 

  12. Srivastava, A., Srivastava, A., Srivastava, K., Choudhuri, G. & Mittal, B. Role of ABCG8 D19H (rs11887534) variant in gallstone susceptibility in northern India. J Gastroenterol Hepatol 25, 1758–1762, (2010).

    CAS  Article  PubMed  Google Scholar 

  13. Zhan, L. et al. Prevalence of ABCB4 polymorphisms in gallstone disease in han-Chinese population. Am J Transl Res 8, 1218–1227 (2016).

    CAS  PubMed  PubMed Central  Google Scholar 

  14. Buch, S. et al. Loci from a genome-wide analysis of bilirubin levels are associated with gallstone risk and composition. Gastroenterology 139, 1942–1951 e1942, (2010).

    CAS  Article  PubMed  Google Scholar 

  15. Joshi, A. D. et al. Four Susceptibility Loci for Gallstone Disease Identified in a Meta-analysis of Genome-Wide Association Studies. Gastroenterology 151, 351–363 e328, (2016).

    CAS  Article  PubMed  Google Scholar 

  16. Genomes Project, C. et al. A global reference for human genetic variation. Nature 526, 68–74, (2015).

    ADS  CAS  Article  Google Scholar 

  17. Evangelou, E. & Ioannidis, J. P. Meta-analysis methods for genome-wide association studies and beyond. Nat Rev Genet 14, 379–389, (2013).

    CAS  Article  PubMed  Google Scholar 

  18. Levine, B. What does the population attributable fraction mean? Prev Chronic Dis 4, A14 (2007).

    PubMed  Google Scholar 

  19. Lazcano-Ponce, E. C. et al. Epidemiology and molecular pathology of gallbladder cancer. CA Cancer J Clin 51, 349–364 (2001).

    CAS  Article  PubMed  Google Scholar 

  20. Srivastava, A., Tulsyan, S., Pandey, S. N., Choudhuri, G. & Mittal, B. Single nucleotide polymorphism in the ABCG8 transporter gene is associated with gallbladder cancer susceptibility. Liver Int 29, 831–837, (2009).

    CAS  Article  PubMed  Google Scholar 

  21. Xu, H. L. et al. Cholesterol metabolism gene polymorphisms and the risk of biliary tract cancers and stones: a population-based case-control study in Shanghai, China. Carcinogenesis 32, 58–62, (2011).

    CAS  Article  PubMed  Google Scholar 

  22. Krawczak, M. et al. PopGen: population-based recruitment of patients and controls for the analysis of complex genotype-phenotype relationships. Community Genet 9, 55–61, (2006).

    Article  PubMed  Google Scholar 

  23. Volzke, H. et al. Cohort profile: the study of health in Pomerania. Int J Epidemiol 40, 294–307, (2011).

    Article  PubMed  Google Scholar 

  24. Sun, S. C. The noncanonical NF-kappaB pathway. Immunol Rev 246, 125–140, (2012).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  25. Tseng, P. H. et al. Different modes of ubiquitination of the adaptor TRAF3 selectively activate the expression of type I interferons and proinflammatory cytokines. Nat Immunol 11, 70–75, (2010).

    CAS  Article  PubMed  Google Scholar 

  26. Hacker, H., Tseng, P. H. & Karin, M. Expanding TRAF function: TRAF3 as a tri-faced immune regulator. Nat Rev Immunol 11, 457–468, (2011).

    CAS  Article  PubMed  Google Scholar 

  27. Chung, J. Y., Park, Y. C., Ye, H. & Wu, H. All TRAFs are not created equal: common and distinct molecular mechanisms of TRAF-mediated signal transduction. J Cell Sci 115, 679–688 (2002).

    CAS  PubMed  Google Scholar 

  28. Schirmer, B. D., Winters, K. L. & Edlich, R. F. Cholelithiasis and cholecystitis. J Long Term Eff Med Implants 15, 329–338 (2005).

    Article  PubMed  Google Scholar 

  29. Kampf, C. et al. Defining the human gallbladder proteome by transcriptomics and affinity proteomics. Proteomics 14, 2498–2507, (2014).

    CAS  Article  PubMed  Google Scholar 

  30. Moro, P. L. et al. Gallstone disease in Peruvian coastal natives and highland migrants. Gut 46, 569–573 (2000).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  31. Palermo, M., Berkowski, D. E., Cordoba, J. P., Verde, J. M. & Gimenez, M. E. Prevalence of cholelithiasis in Buenos Aires, Argentina. Acta Gastroenterol Latinoam 43, 98–105 (2013).

    PubMed  Google Scholar 

  32. Claros, N., Laguna, R., Ponce, R. & Feraudy, I. ¿Cuál es la prevalencia de litiasis de la vía biliar principal en pacientes con colecistolitiasis sintomática? Revista chilena de cirugía 59, 127–131 (2007).

    Google Scholar 

  33. Sodhi, J. S. et al. Prevalence of gallstone disease in patients with type 2 diabetes and the risk factors in North Indian population: a case control study. Indian J Gastroenterol 33, 507–511, (2014).

    ADS  Article  PubMed  Google Scholar 

  34. Sun, H. et al. Gender and metabolic differences of gallstone diseases. World J Gastroenterol 15, 1886–1891 (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  35. Medina-Gomez, C. et al. Challenges in conducting genome-wide association studies in highly admixed multi-ethnic populations: the Generation R Study. Eur J Epidemiol 30, 317–330, (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  36. Price, A. L., Zaitlen, N. A., Reich, D. & Patterson, N. New approaches to population stratification in genome-wide association studies. Nat Rev Genet 11, 459–463, (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  37. Stender, S., Frikke-Schmidt, R., Nordestgaard, B. G. & Tybjaerg-Hansen, A. Sterol transporter adenosine triphosphate-binding cassette transporter G8, gallstones, and biliary cancer in 62,000 individuals from the general population. Hepatology 53, 640–648, (2011).

    CAS  Article  PubMed  Google Scholar 

  38. Mhatre, S. et al. Common genetic variation and risk of gallbladder cancer in India: a case-control genome-wide association study. Lancet Oncol 18, 535–544, (2017).

    Article  PubMed  Google Scholar 

  39. Zapata, J. M. et al. Lymphocyte-specific TRAF3 transgenic mice have enhanced humoral responses and develop plasmacytosis, autoimmunity, inflammation, and cancer. Blood 113, 4595–4603, (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  40. Annunziata, C. M. et al. Frequent engagement of the classical and alternative NF-kappaB pathways by diverse genetic abnormalities in multiple myeloma. Cancer Cell 12, 115–130, (2007).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  41. Otto, C. et al. Genetic lesions of the TRAF3 and MAP3K14 genes in classical Hodgkin lymphoma. Br J Haematol 157, 702–708, (2012).

    CAS  Article  PubMed  Google Scholar 

  42. Ruiz-Ballesteros, E. et al. Splenic marginal zone lymphoma: proposal of new diagnostic and prognostic markers identified after tissue and cDNA microarray analysis. Blood 106, 1831–1838, (2005).

    CAS  Article  PubMed  Google Scholar 

  43. Liao, G., Zhang, M., Harhaj, E. W. & Sun, S. C. Regulation of the NF-kappaB-inducing kinase by tumor necrosis factor receptor-associated factor 3-induced degradation. J Biol Chem 279, 26243–26250, (2004).

    CAS  Article  PubMed  Google Scholar 

  44. Bista, P. et al. TRAF3 controls activation of the canonical and alternative NFkappaB by the lymphotoxin beta receptor. J Biol Chem 285, 12971–12978, (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  45. Yao, H. S. et al. Annexin A4-nuclear factor-kappaB feedback circuit regulates cell malignant behavior and tumor growth in gallbladder cancer. Sci Rep 6, 31056, (2016).

    ADS  CAS  Article  PubMed  PubMed Central  Google Scholar 

  46. Jiang, X. et al. cIAP2 promotes gallbladder cancer invasion and lymphangiogenesis by activating the NF-kappaB pathway. Cancer Sci 108, 1144–1156, (2017).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  47. Maurer, K. J., Carey, M. C. & Fox, J. G. Roles of infection, inflammation, and the immune system in cholesterol gallstone formation. Gastroenterology 136, 425–440, (2009).

    CAS  Article  PubMed  Google Scholar 

  48. Hacker, H. et al. Specificity in Toll-like receptor signalling through distinct effector functions of TRAF3 and TRAF6. Nature 439, 204–207, (2006).

    ADS  CAS  Article  PubMed  Google Scholar 

  49. Oganesyan, G. et al. Critical role of TRAF3 in the Toll-like receptor-dependent and -independent antiviral response. Nature 439, 208–211, (2006).

    ADS  CAS  Article  PubMed  Google Scholar 

  50. Lalani, A. I. et al. Myeloid cell TRAF3 regulates immune responses and inhibits inflammation and tumor development in mice. Journal of immunology 194, 334–348, (2015).

    CAS  Article  Google Scholar 

  51. Hsing, A. W. et al. Variants in inflammation genes and the risk of biliary tract cancers and stones: a population-based study in China. Cancer Res 68, 6442–6452, (2008).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  52. Carithers, L. J. et al. A Novel Approach to High-Quality Postmortem Tissue Procurement: The GTEx Project. Biopreserv Biobank 13, 311–319, (2015).

    Article  PubMed  PubMed Central  Google Scholar 

  53. Volzke, H. et al. Independent risk factors for gallstone formation in a region with high cholelithiasis prevalence. Digestion 71, 97–105, (2005).

    Article  PubMed  Google Scholar 

  54. Hoffmann, T. J. et al. Design and coverage of high throughput genotyping arrays optimized for individuals of East Asian, African American, and Latino race/ethnicity using imputation and a novel hybrid SNP selection algorithm. Genomics 98, 422–430, (2011).

    CAS  Article  PubMed  Google Scholar 

  55. Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81, 559–575, (2007).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  56. Chang, C. C. et al. Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 4, 7, (2015).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  57. Anderson, C. A. et al. Data quality control in genetic case-control association studies. Nat Protoc 5, 1564–1573, (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  58. Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 38, 904–909, (2006).

    CAS  Article  PubMed  Google Scholar 

  59. Vidal, E. A. et al. Whole genome sequence of Mapuche-Huilliche Native Americans. bioRxiv (2018).

  60. Alexander, D. H., Novembre, J. & Lange, K. Fast model-based estimation of ancestry in unrelated individuals. Genome Res 19, 1655–1664, (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  61. Howie, B. N., Donnelly, P. & Marchini, J. A flexible and accurate genotype imputation method for the next generation of genome-wide association studies. PLoS Genet 5, e1000529, (2009).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  62. Marchini, J. & Howie, B. Genotype imputation for genome-wide association studies. Nat Rev Genet 11, 499–511, (2010).

    CAS  Article  PubMed  Google Scholar 

  63. Marchini, J., Howie, B., Myers, S., McVean, G. & Donnelly, P. A new multipoint method for genome-wide association studies by imputation of genotypes. Nat Genet 39, 906–913, (2007).

    CAS  Article  PubMed  Google Scholar 

  64. Wittwer, C. T., Reed, G. H., Gundry, C. N., Vandersteen, J. G. & Pryor, R. J. High-Resolution Genotyping by Amplicon Melting Analysis Using LCGreen. Clinical Chemistry 49, 853–860, (2003).

    CAS  Article  PubMed  Google Scholar 

  65. Pruim, R. J. et al. LocusZoom: regional visualization of genome-wide association scan results. Bioinformatics 26, 2336–2337, (2010).

    CAS  Article  PubMed  PubMed Central  Google Scholar 

  66. Barrera, F. et al. Effect of cholecystectomy on bile acid synthesis and circulating levels of fibroblast growth factor 19. Ann Hepatol 14, 710–721 (2015).

    CAS  PubMed  Google Scholar 

Download references


We gratefully acknowledge GSD and GBC affected and unaffected individuals who participated in this study. This work was funded by grants from Fondo de Areas Prioritarias (FONDAP) Center for Genome Regulation (1509000) and Fondo Nacional de Desarrollo Científico y Tecnológico (FONDECYT) 1140353 and 1180848 to G.V.D, 1130303 to J.F.M and from the Postdoctoral grant 3160400 to E.R. The Immunohistochemical analysis was supported by the Unidad de Microscopia Avanzada (de Ciencias Biológicas o de Medicina) facility at the Pontificia Universidad Catolica de Chile. The PopGen 2.0 network was supported by a grant from the German Federal Ministry of Education and Research (01EY1103). SHIP-Greifswald is part of the Community Medicine Research net of the University of Greifswald (Germany), which is funded by the Federal Ministry of Education and Research (grants 01ZZ9603, 01ZZ0103, and 01ZZ0403), the Ministry of Cultural Affairs, as well as the Social Ministry of the Federal State of Mecklenburg–West Pomerania, and the network “Greifswald Approach to Individualized Medicine” funded by the Federal Ministry of Education and Research (grant 03IS2061A). SHIP-Greifswald genome-wide data have been supported by the Federal Ministry of Education and Research (grant 03ZIK012) and a joint grant from Siemens Healthcare (Erlangen, Germany) and the Federal State of Mecklenburg–West Pomerania. The University of Greifswald is a member of the Center of Knowledge Interchange program of the Siemens AG and the Caché Campus program of the InterSystems GmbH. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Author information

Authors and Affiliations



B.I.B., G.V.D. and J.F.M. conceived and designed the experiments. J.F.M., W.L., A.F., S.H., G.B., W.S., C.S., H.V., U.V., G.H., M.M.L., K.P., C.B. and J.C.R., collected the data. B.I.B., E.P.-P., L.A., S.B., M.T., E.R., G.D.U. and G.V.D. performed experiments or analyzed the data. P.N., J.L.S., R.A.G. and J.H. helped in drafting the manuscript. B.I.B., G.V.D. and J.F.M. wrote the paper. All authors reviewed the manuscript for medical and scientific accuracy.

Corresponding authors

Correspondence to Giancarlo V. De Ferrari or Juan Francisco Miquel.

Ethics declarations

Competing Interests

The authors declare no competing interests.

Additional information

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Rights and permissions

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 license, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons license 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 license, visit

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Bustos, B.I., Pérez-Palma, E., Buch, S. et al. Variants in ABCG8 and TRAF3 genes confer risk for gallstone disease in admixed Latinos with Mapuche Native American ancestry. Sci Rep 9, 772 (2019).

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI:

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing