Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2022 Feb 1.
Published in final edited form as: Hum Genet. 2020 Jul 15;140(2):309–319. doi: 10.1007/s00439-020-02205-8

A Pooled Genome-wide Association Study Identifies Pancreatic Cancer Susceptibility Loci on Chromosome 19p12 and 19p13.3 in the Full-Jewish Population

Samantha A Streicher 1, Alison P Klein 2,3, Sara H Olson 4, Robert C Kurtz 5, Laufey T Amundadottir 6, Andrew T DeWan 1, Hongyu Zhao 7,8, Harvey A Risch 1
PMCID: PMC8395137  NIHMSID: NIHMS1612387  PMID: 32671597

Abstract

Jews are estimated to be at increased risk of pancreatic cancer compared to non-Jews, but their observed 50–80% excess risk is not explained by known non-genetic or genetic risk factors. We conducted a GWAS in a case-control sample of American Jews, largely Ashkenazi, including 406 pancreatic cancer patients and 2,332 controls, identified in the dbGaP, PanScan I/II, PanC4 and GERA data sets. We then examined resulting SNPs with P<10−7 in an expanded sample set, of 539 full- or part-Jewish pancreatic cancer patients and 4,117 full- or part-Jewish controls from the same data sets. Jewish ancestries were genetically determined using seeded FastPCA. Among the full-Jews, a novel genome-wide significant association was detected on chromosome 19p12 (rs66562280, per-allele OR=1.55, 95% CI=1.33–1.81, P=10−7.6). A suggestive relatively independent association was detected on chromosome 19p13.3 (rs2656937, OR=1.53, 95% CI=1.31–1.78, P=10−7.0). Similar associations were seen for these SNPs among the full and part Jews combined. This is the first GWAS conducted for pancreatic cancer in the increased-risk Jewish population. The SNPs rs66562280 and rs2656937 are located in introns of ZNF100-like and ARRDC5, respectively, and are known to alter regulatory motifs of genes that play integral roles in pancreatic carcinogenesis.

Keywords: Pancreatic cancer, Ashkenazi Jews, Genome-wide association study, Principal component analysis

Introduction

For pancreatic cancer, the 1-year relative survival proportion is about 15% and the 5-year proportion less than 3%, on a population basis (Gong et al. 2011). In the United States, for both men and women, pancreatic cancer ranks as the fourth most-frequent cause of cancer death, behind lung, prostate, and colorectal cancer for men, and lung, breast, and colorectal cancer for women (American Cancer Society 2020). However, in contrast to most other cancers, mortality rates for pancreatic cancer are increasing and pancreatic cancer is predicted to be the third leading cause of cancer death by 2022, exceeded only by lung and breast/prostate cancer (Rahib et al. 2014).

Over the past 65 years, pancreatic cancer has consistently been found to be more frequent among Jews, especially Ashkenazi Jews, than non-Jews (MacMahon 1955; Haenszel 1961; Newill 1961; Moldow and Connelly 1968; Seidman 1970; Wynder et al. 1973; Greenwald et al. 1975; Mack et al. 1985; Coogan et al. 2000; Eldridge et al. 2011; Hamada et al. 2019). The two most recent large studies that examined the risk of pancreatic cancer in Jewish compared to non-Jewish subjects, Eldridge and colleagues (Eldridge et al. 2011) and Risch and colleagues (Risch et al. 2015) found a hazard ratio (HR)=1.43, 95%, confidence interval (CI)=1.30–1.57 and an odds ratio (OR)=1.81, 95% CI=1.05–3.10 (Eldridge et al. 2011; Risch et al. 2015), respectively. Increased pancreatic cancer risk among Ashkenazi Jews can be partially explained by the greater prevalence of two BRCA1 mutations (185delAG and 5382insC) and one BRCA2 mutation (6174delT) that together account for about 9.1% of the disease in Jews (Risch et al. 2006) and the greater prevalence of non-O ABO blood groups among Ashkenazi Jews that accounts for about 2% of the excess disease risk in Jews (Streicher et al. 2017). Other known genetic and non-genetic risk factors in sum are not appreciably higher in the Jewish population compared to non-Jewish whites (Eldridge et al. 2011; Hamada et al. 2019).

Both Hamada and colleagues and Eldridge and colleagues examined established non-genetic pancreatic cancer risk factors, including cigarette smoking, obesity, and diabetes, and found that they did not explain the higher risk of incident pancreatic cancer for Ashkenazi Jews in the Health Professionals Follow-up Study (HPFS) or the higher risk of pancreatic cancer mortality for Jewish ethnicity in the Cancer Prevention Study II (CPS II), respectively (Eldridge et al. 2011; Hamada et al. 2019). We recently evaluated associations of the 16 genome-wide significant common low risk pancreatic cancer genetic variants identified from five white European genome-wide association studies (GWASes), in a large sample of Jews (Streicher et al. 2017). We found that for 13 of the 16 genetic variants, among Jews, either a significant association was seen with risk comparable to the published point estimates for white European subjects in general, or the directions and magnitudes of association were consistent with the published estimates. We also found that the average OR between the full-Jewish and non-Jewish white European populations was similar for these single nucleotide polymorphisms (SNPs) (1.25 vs. 1.26) (Amundadottir et al. 2009; Petersen et al. 2010; Wolpin et al. 2014; Childs et al. 2016; Zhang et al. 2016; Streicher et al. 2017).

Since the discrepancy between the known 11% excess risk of pancreatic cancer, due to BRCA1 and BRCA2 mutations and non-O ABO blood groups, and the observed 50–80% excess risk of pancreatic cancer in the Jewish population has yet to be explained by established non-genetic or genetic risk factors, we conducted a GWAS in a sample of Jewish subjects to investigate Jewish genetic contributions to risk of pancreatic cancer.

Methods

Study, reference marker, and imputation data sets

Our study, reference marker, and imputation data sets have previously been described in detail. Our current study used the same definition for full-Jewish, part-Jewish, and non-Jewish white European subjects as our previous study that examined the impact of sixteen established pancreatic cancer susceptibility loci in Jewish subjects (Streicher et al. 2017).

Briefly, our study data set included the Pancreatic Cancer Cohort and Case-Control Consortium I/II (PanScan I/II) (accession phs000206.v4.p3), Pancreatic Cancer Case-Control Consortium (PanC4) (accession phs000648.v1.p1), and the Resource for Genetic Epidemiology Research on Adult Health and Aging (GERA) (accession phs000674.v1.p1) data sets, all of which were obtained from the database of Genotypes and Phenotypes (dbGaP) with genotypes and basic phenotype information. The PanScan I/II and PanC4 data sets contained pancreatic cancer case and control subjects, while the GERA data set was used as a control-only data set. Three data sets included in dbGaP were also obtained from PanScan I/II and PanC4 investigators. These data contained genotypes and additional information on self-reported race and ethnicity from individual study sites. Subjects in these data sets were used as reference marker subjects to help genetically identify full-Jewish and part-Jewish subjects from principal component plots. Both the 1000 Genomes Project Phase 3 (October 2014 release) and The Ashkenazi Genome Consortium (TAGC) data sets were used as reference panels for haplotype estimation (pre-phasing) and genotype imputation (The 1000 Genomes Project Consortium 2012; Carmi et al. 2014). All necessary Institutional Review Board approvals and data use agreements were granted from Yale University and collaborating institutions.

Quality control (QC) and Fast principal component analysis (FastPCA)

QC was as described in detail previously (Streicher et al. 2017). Briefly, subjects from PanScan I were genotyped on the Illumina HumanHap550 Infinium array, subjects from PanScan II were genotyped on the Illumina Human610-Quad array, subjects from PanC4 were genotyped on the Illumina HumanOmniExpressExome array, and subjects from GERA were genotyped on the Affymetrix Axiom Genome-wideEUR array. QC was performed separately for each genotyping platform. Sample replicates, failed samples, subjects with <98% genotyping completion, subjects with missing sex information, indeterminate X chromosome heterozygosity, or discordance in reported vs. genotyped sex (reported females with >0.25 and reported males with <0.80 X chromosome heterozygosity), and related subjects (π^≥0.20) were excluded from all analyses. Duplicate variants, variants with call rates <98%, extreme Hardy-Weinberg equilibrium departures in controls (P<10−7), monomorphic variants either in cases or controls, copy number variants, and variants in non-autosomal chromosomes were also excluded from all analyses. After QC, 530,771 SNPs were available for analysis on the Illumina HumanHap550 Infinium array, 553,743 SNPs on the IlluminaHuman610-Quad array, 804,262 SNPs on the Illumina HumanOmniExpressExome array, and 596,652 SNPs on the Affymetrix Axiom Genome-wide EUR array. FastPCA was performed as described in detail previously (Streicher et al. 2017). In brief, four FastPCAs were sequentially conducted with the 80,722 post-QC genotyped SNPs common to all four genotyping platforms using EIGENSOFT v6 (Galinsky et al. 2016). First, FastPCA was performed on 55,930 PanScan/PanC4/GERA subjects of various races and ethnicities, including reference marker subjects. The FastPCA PC2 vs. PC1 plot enabled visualization of the PanScan/PanC4/GERA subjects (Supplementary Figure 2). In total, 54,829 white European subjects to the right of the diagonal line in Supplementary Figure 2 were retained, and a second FastPCA was run with these white European subjects, plus reference marker subjects. FastPCA PC2 vs. PC1 was again plotted to visualize the white European subjects (Supplementary Figure 3). From this, 47,545 white European subjects to the left of the diagonal line in Supplementary Figure 3 were retained, and a third FastPCA was run with these white European subjects and reference marker subjects. PC2 vs. PC1 from FastPCA was plotted to visualize the spread between the full-Jewish, part-Jewish, and non-Jewish white European subjects (Supplementary Figure 4). In total, 4,657 full- or part-Jewish subjects to the left of the two intersecting diagonal lines in Supplementary Figure 4 were retained, and a fourth FastPCA was run on these 4,657 full- or part-Jewish subjects plus reference marker subjects. Finally, PC2 vs. PC1 from FastPCA was again plotted with reference marker subjects to visualize the spread among the full-Jewish, 3/4 Jewish, 1/2 Jewish, and 1/4 Jewish subjects (Figure 1). All lines were placed on these four FastPCA plots based on location of the reference marker subjects and where subject clusters thinned.

Figure 1.

Figure 1.

Open in a new tab

FastPCA plot of components 2 v. 1 for 4,657 PanScan/PanC4/GERA Jewish subjects and Jewish or part-Jewish reference marker subjects. The left vertical line demarcates separation between 1/4 Jewish and 1/2 or 3/4 Jewish subjects. The right vertical line demarcates separation between 1/2 or 3/4 Jewish subjects and full-Jewish subjects.

Study population

The full-Jewish subjects and the full- plus part-Jewish subjects were visually identified on the final FastPCA plot (Figure 1). There were no differences in Jewish identification seen in the FastPCA plot according to genotype platform (Supplementary Figure 5). In the individual site studies, reference marker subjects reporting ancestry variously as “Jewish” or “Half-Jewish” or as number of Jewish grandparents facilitated identification of the relevant regions in the PCA plots. In total, 406 full-Jewish pancreatic cancer patients and 2,332 full-Jewish controls, all with complete age, sex, case/control status and genotype information were included in the GWAS. Additionally, 133 part-Jewish pancreatic cancer patients and 1,785 part-Jewish controls, also with complete age, sex, case/control status and genotype information, were added to the full-Jewish pancreatic cancer patients and full-Jewish controls, respectively. This resulted in an extended sample of 539 full- plus part-Jewish pancreatic cancer patients and 4,117 full- plus part-Jewish controls. GWAS SNPs with risk-association P<10−7 in full-Jewish subjects were examined in this extended sample set of full- plus part-Jews. The 539 full- plus part-Jewish cases from the PanScan I/II and PanC4 studies comprise the largest known Jewish pancreatic cancer patient group. To date, an independent validation is not possible with any known data set.

GWAS

Pre-phasing, using SHAPEIT v2, and genotype imputation, using IMPUTE v2, were performed separately for 3,541 white European subjects on the Illumina HumanHap550 array, 3,621 white European subjects on the Illumina Human610-Quad array, 7,485 white European subjects on the Illumina HumanOmniExpressExome array, and 40,182 white European subjects on the Affymetrix Axiom Genome-wide EUR array (O’Connell et al. 2014; Huang et al. 2015). Prior to imputation, variants that could not be mapped from human genome version 18 (hg18) to human genome version 19 (hg19) were removed. The manufacturers’ annotation files and the University of California Santa Cruz (UCSC) genome browser were used to align genotypes for imputation; therefore, variants not in these files were also removed (Illumina HumanHap550 Infinium array N = 1,745; Illumina Human610-Quad array N = 2,071; Illumina HumanOmniExpressExome array N = 21,488; Affymetrix AxiomGenome-wide EUR array N = 6,895) (Speir et al. 2016). Exome-labeled variants were also removed from the Illumina HumanOmniExpressExome array because of their poor representation in the 1000 Genomes Project Phase 3 and TAGC reference panels (N =132,930). Both the 1000 Genomes Project Phase 3 (October 2014 release) and TAGC data sets were used as reference panels for pre-phasing and genotype imputation (The 1000 Genomes Project Consortium 2012; Carmi et al. 2014). After imputation, SNPs with information metric (info) <0.7 were excluded from all subsequent analyses (Huang et al. 2015). Remaining IMPUTE genotype probabilities were converted to genotypes based on a hard call threshold of 0.49999. In order to exclude SNPs with false associations, varying ranges in exclusion criteria were examined to yield an acceptable genomic inflation factor (Price et al. 2010): Ultimately, SNPs with minor allele frequencies (MAFs) >0.45 or <0.05, missing fraction >0.005, extreme Hardy-Weinberg equilibrium departures in controls (P<10−6), case or control MAF relative platform differences exceeding 18%, or heterogeneity comparisons of case or control platform allele proportions χ2 P<0.2 were removed (Supplementary Figure 1) (Fleiss et al. 2003). Case and control MAF relative platform differences were calculated by subtracting the case or control MAF for each platform from the average case or control MAF for the remaining platforms and dividing the difference by the average of the remaining platforms. In total, 1,286,673 genotyped and imputed SNPs remained after SNP filtering as just described. Data from the four genotyping platforms were pooled into one large logistic regression analysis, and association analyses were conducted in PLINK v1.90-beta using unconditional logistic regression, adjusted for age (in 10-year categories), sex, and 6 principal components from a full- plus part-Jewish exact PCA. Per-allele ORs (i.e., OR trends by number of variant alleles), 95% CIs, and P-values were calculated. SNP associations were considered statistically significant at P < 10−7·3(i. e. , 5 × 10−8) and suggestive at P < 10−7 .

Results

In the full-Jewish subjects, a genome-wide significant locus on chromosome 19p12 associated with risk of pancreatic cancer was identified: rs66562280, OR=1.55, 95% CI=1.33–1.81, P=10−7.6. An essentially independent suggestive locus on chromosome 19p13.3 was also associated with risk of pancreatic cancer: rs2656937, OR=1.53, 95% CI=1.31–1.78, P=10−7.0. (Table 1, Figures 2 and 3, Supplementary Tables 2 and 3). When Illumina-only platforms were analyzed the results remained consistent: OR=1.35, 95% CI=1.09–1.68, P=0.0073 for rs66562280 and OR=1.40, 95% CI=1.13–1.74, P=0.0024 for rs2656937. A quantile-quantile plot of GWAS P-values of SNP associations in the full-Jewish subjects indicated appropriate control of type I errors, with a genomic inflation λ value of 1.043 (Price et al. 2010) (Figure 4).

Table 1.

One significant pancreatic cancer susceptibility locus (P<10−7.3) and one suggestive locus (P<10−7) found in the PanScan/PanC4/GERA full-Jewish subjects’ genome-wide association study and examined in PanScan/PanC4/GERA full- plus part-Jewish subjects.

Chromosomea Gene SNP Positionb Minor allele Major allele Minor allele frequency Number of Subjects
Subjects Allelic OR (95% CI)c Cases Controls Cases Controls P
19p12 ZNF100-like rs66562280 21833014 T C full-Jewish 1.55
(1.33–1.81)		 0.43 0.32 398 2328 10−7.6
full- plus part-Jewish 1.56
(1.36–1.78)		 0.42 0.31 519 4109 10−10
19p13.3 ARRDC5 rs2656937 4892087 C T full-Jewish 1.53
(1.31–1.78)		 0.44 0.34 406 2319 10−7
full- plus part-Jewish 1.50
(1.31–1.57)		 0.42 0.31 539 3989 10−8.4
Open in a new tab
a

Cytogenetic region according to National Center for Biotechnology Information (NCBI) Human Genome Build 37 (hg19).

b

SNP position according to NCBI Human Genome Build 37 (hg19).

c

Allelic odds ratio and 95% confidence interval adjusted for age, sex, and exact PC1-PC6 from the full- plus part-Jewish population principal component analysis plot.

d

For full- plus part-Jewish subjects rs66562280 and rs2656937 did not quite meet our quality control filter inclusion criteria for SNP discovery.

Figure 2.

Figure 2.

Open in a new tab

Manhattan plot of SNP P-values from the full-Jewish GWAS. The Y-axis shows the negative base ten logarithm of the P-values and the X-axis shows the chromosomes. The genome-wide significance threshold, P<10−7.3 (i.e. P<5×10−8) is shown in red.

Figure 3.

Figure 3.

Open in a new tab

Regional plots of SNP P-values from the full-Jewish GWAS in a +/−400 kb window around rs66562280 and rs2656937. The chromosomal band is given above each plot. The X-axis shows the chromosome and physical location (Mb), the left Y-axis shows the negative base ten logarithm of the P-values, and the right Y-axis shows recombination activity (cM/Mb) as a blue line. Positions, recombination rates, and gene annotations are according to NCBI’s build 37 (hg 19) and the 1000 Genomes Project Phase 3 European data set.

Figure 4.

Figure 4.

Open in a new tab

Q-Q plot of SNP P-values from the full-Jewish GWAS. The Y-axis shows the negative base ten logarithm of the observed p-values and the X-axis shows the negative base ten logarithm of the expected p-values. Genomic inflation λ=1.043.

In the full- plus part-Jewish extended subject set, the associations with pancreatic cancer for rs66562280 and rs2656937 remained similar: OR=1.56, 95% CI=1.36–1.78, P=10−10 and OR=1.50, 95% CI=1.31–1.57, P=10−8.4, respectively (Table 1). However, in the full- plus part-Jewish subjects, rs66562280 and rs2656937 did not quite meet our QC filter inclusion criteria for SNP discovery (SNPs were removed with MAFs >0.45 or <0.05, missing fraction >0.005, extreme Hardy-Weinberg equilibrium departures in controls (P<10−6), case or control MAF relative platform differences exceeding 18%, or heterogeneity comparisons of case or control platform allele proportions χ2 P<0.2, see the GWAS section under Materials and Methods for more detail). For rs66562280, the control Affymetrix MAF difference = 23.5% and control χ2 P = 0.002. For rs2656937, the control Illumina HumanOmniExpressExome MAF difference = 24.5% (Supplementary Tables 2 and 3). None of the SNPs meeting the QC filter inclusion criteria were found in the full- plus part-Jewish subjects to be significant at the P=10−7 level. In non-Jewish white European subjects (comprised of Eastern, Northern, Southern, and Western non- Jewish white Europeans), rs66562280 and rs2656937 also did not meet our QC filter inclusion criteria for SNP discovery. Bearing this in mind, in non-Jewish white European cases and controls, OR=2.00, 95% CI=1.92–2.08, P=10−267 for rs66562280 and OR=1.12, 95% CI=1.07–1.16, P=10−6.7 for rs2656937. For rs66562280, the control Illumina HumanOmniExpressExome MAF difference = 21.2%, the control Affymetrix MAF difference = 44.1%, the control χ2 P = 10−217.6, and the case χ2 P = 10−1.1. For rs2656937, the control χ2 P = 10−6.6 (Supplementary Tables 2 and 3). Across the full-Jewish, full- plus part-Jewish, and non-Jewish white Europeans subjects and between cases and controls, the percent distributions were similar for age (<50, 51–60, 61–70, 71–80, >80) and sex (Male, Female) (Supplementary Table 1).

Association results were similar between the genotype threshold method (above) in PLINK v1.09 and the genotype dosage method in SNPTEST v2 (Marchini and Howie 2010) (Supplementary Table 4). For sensitivity purposes, when the line that separates full-Jewish from part-Jewish subjects on the FastPCA plot in Figure 1 was moved slightly in either direction and full-Jewish subjects’ ORs and P-values were recalculated to include or exclude ~30 full-Jews, the results were consistent with the original full-Jewish subjects’ analysis (Supplementary Figure 6 and Supplementary Table 5).

The population attributable fraction (PAF) of pancreatic cancer in the full-Jewish subjects for the minor allele (T) of rs66562280 was 27.7%, and for the minor allele (C) of rs2656937 was 28.2% (Supplementary Table 6). The PAF is calculated under the assumption of Hardy-Weinberg equilibrium. The PAF calculation is based on the fact that for each SNP there are three mutually exclusive types of genotype carriers (e.g., MM, Mm, and mm, where M=major allele and m=minor allele) and for two SNPs there are thus nine possible combinations. When the PAF variables are multiple mutually exclusive states that comprise the whole population, PAF=∑Pe(OR−1)1+∑Pe(OR−1), over all the states except the reference one (say mm1×mm2 ), where for each of the eight non-reference allele pairs, Pe=the genotype frequency in controls and OR=the odds ratio. These two SNP loci are approximately 17 mb apart and according to the linkage-disequilibrium (LD) calculator LDpair (https://ldlink.nci.nih.gov/?tab=ldpair), the two SNPs are not in appreciable LD in European populations (R2=0.02, Dʹ=0.078), thus their individual minor allele frequencies can also be used to calculate their combined population attributable fraction, PAFTotal = 1 − (1 − PAFi)(1 − PAFj) = 48.1% (Witte et al. 2014).

Discussion

Six GWASes have been conducted for pancreatic cancer in white European populations, and together they have identified and replicated 20 genome-wide significant pancreatic cancer genetic susceptibility loci (Amundadottir et al. 2009; Petersen et al. 2010; Wolpin et al. 2014; Childs et al. 2016; Zhang et al. 2016; Klein et al. 2018). Here, we report the first GWAS for pancreatic cancer in the increased-risk Jewish population. We observed a genome-wide significant association with pancreatic cancer risk in a novel region on chromosome 19p12 for rs66562280, which is located in an intron of ZNF100-like (also known as LOC400682). ZNF100-like is a protein-coding gene residing between two ZNF genes, ZNF100 and ZNF429 (~70 kb away from ZNF100 and ~110 kb away from ZNF429), and in the same 400 kb region as several other ZNF protein genes (Figure 3). ZNF genes encode transcription factors, which play integral roles in regulating gene expression, and therefore coordinate biological processes such as differentiation, development, metabolism, apoptosis, autophagy, and stemness maintenance. ZNF100 and ZNF429 genes are both KRAB (kruppel-associated box) genes with C2H2 (CysHis2) zinc finger motifs (Bellefroid et al. 1993). A cluster of highly related KRAB-ZNF genes, including ZNF100 and ZNF429, has been identified in the chromosomal region 19p12-p13.1 (Bellefroid et al. 1993; Rink et al. 2013). ZNF91, the most studied gene in this cluster, has been shown to have upregulated gene expression in acute myelogenous leukemia (AML), prostate cancer, and colon cancer (Unoki et al. 2003; Paschke et al. 2014; Ma et al. 2016). Moreover, in AML, inhibited expression of ZNF91 suppresses cell growth and induces apoptosis, and in colon cancer, overexpressed ZNF91 interacts with NFKB (nuclear factor kappa B), and together they promote colon carcinogenesis by upregulating HIF-1α (hypoxia-inducible factor 1-α) (Unoki et al. 2003; Paschke et al. 2014; Ma et al. 2016). Since HIF-1α expression is upregulated in pancreatic cancer tissue and NFKB activation is increased when HIF-1α is transiently transfected into pancreatic cancer cells, it is reasonable to hypothesize that the increased expression of NFKB and HIF-1α in pancreatic cancer is also due, at least in part, to ZNF91, and possibly other related genes in the same cluster (Sun et al. 2007; Cheng et al. 2011). Additionally, according to haploreg v4, rs66562280 alters regulatory motifs for HNF1 – HNF7 (hepatocyte nuclear factors 1–7), HLTF (helicase like transcription factor), and IRF (interferon regulatory factor) genes. The log-odds (LOD) score for the position weight matrices (PWMs) does not drastically change depending on the expression of the minor allele (T) or the major allele (C) in rs66562280 when bound to regulatory motifs for HNF1 or IRF (Ward and Kellis 2016) . Although this suggests that binding does not occur at important positions and does not severely disrupt the regulatory motifs, it is still possible that the T allele could affect transcription factor binding. The PanC4 GWAS identified the genome-wide significant SNP, rs7214041, whose minor allele (T) alters the same HNF1 regulatory motif (also with a small change in LOD score for the PWMs) as does the minor allele in rs66562280 (Ward and Kellis 2016). Furthermore, the PanC4 GWAS found suggestive evidence of an association with pancreatic cancer risk for rs7310409 located in HNF1A and rs6073450 located 20kb away from HNF4A, and the most recent combined PanScan/PanC4/PANDoRA GWAS revealed significant associations at rs4795218 located in HNF1B and rs2941471 located in HNF4G (Childs et al. 2016; Klein et al. 2018). Remarkably, Bailey and colleagues recently performed an integrated genomic analysis of 456 pancreatic tumors and showed that HNF genes, specifically HNF4G, HNF4A, HNF1N, and HNF1A, primarily drive pancreatic-associated intraductal papillary mucinous neoplasms (Bailey et al. 2016).

When rs66562280 is bound to a regulatory motif for HLTF, the PWM LOD score does dramatically change depending on the expression of the minor allele (C) or the major allele (T) (Ward and Kellis 2016). HLTF expression has been seen to be altered either through promoter hypermethylation and gene silencing, in colon and gastric cancers, or by expression of HLTF truncated proteins that lack functional domains, in cervical, head and neck, and thyroid cancers (Dhont et al. 2016). Additionally, HLTF is a member of the SWI/SNF family of chromatin-remodeling factors, which has recently been implicated as one of ten molecular pathways activated in pancreatic tumors (Bailey et al. 2016). HLTF is expressed in normal pancreas tissue, but the role of HLTF in pancreatic cancer is not yet clear (Gong et al. 1997; Sandhu et al. 2012). We also observed an essentially independent suggestive association with risk of pancreatic cancer in a novel region on chromosome 19p13.3 for rs2656937, located in an intron of ARRDC5 (Arrestin domain-containing 5). ARRDC5 is a gene in the alpha arrestin family of proteins which are responsible for turning on and off the coupling of GPCRs (G protein-coupled receptors) to heterotrimeric G proteins (Kang et al. 2014). The SNP is 11kb away from UHRF1 (ubiquitin like with PHD and ring finger domains 1), a gene known to be involved in pancreatic carcinogenesis (Abu-Alainin et al. 2016). It has been shown that the expression of UHRF1 is increased in pancreatic cancer tissue compared to normal tissue, and that suppression of UHRF1 expression in pancreatic cancer cells decreases growth and enhances apoptosis and cell-cycle arrest (Cui et al. 2015; Abu-Alainin et al. 2016). Furthermore, chromosomal deletions have been detected in pancreatic cancer tissue for the chromosomal band 19p13.3, which includes both ARRDC5 and UHRF1 (Nowak et al. 2005; Liang et al. 2014).

According to haploreg v4.1, rs2656937 also alters 11 regulatory motifs with large PWM LOD score changes depending on the expression of the major allele (T) or the minor allele (C) for 6 of the 11 regulatory motifs (Ward and Kellis 2016): FOXP1 (Forkhead box P1), MYC (v-myc avian myelocytomatosis viral oncogene homolog), NRSF/REST (RE1 silencing transcription factor), NANOG (NANOG homeobox), SETDB1 (SET domain bifurcated 1), and THAP1 (THAP domain containing 1). Moreover, FOXP1, MYC, NRSF/REST, and NANOG have known involvement in pancreatic cancer carcinogenesis (Azevedo-Pouly et al. 2013; Lu et al. 2013; Hessmann et al. 2016; Qiu et al. 2016). Since rs66562280 and rs2656937 are located in gene introns, both SNPs likely affect pancreatic cancer risk through regulatory motifs or other non-direct mechanisms. It is also possible that neither of these SNPs is a causal variant, but rather are haplotype tag SNPs in high linkage disequilibrium (LD) with causal variants.

In order for our study to be powered sufficiently to detect statistically significant associations, we had to pool data across different genotyping platforms within one manufacturer (Illumina) and with a platform of a different manufacturer (Affymetrix), the latter including only control subjects. We chose to use the genotype threshold method rather than the dosage method after imputation because only the threshold method allowed us to calculate SNP MAFs for the control-only Affymetrix platform post-imputation. We used this information, along with SNP MAFs from the three other genotyping platforms to calculate case or control MAF relative platform differences and heterogeneity comparisons of case or control platform allele proportions χ2 P, which enabled us to remove SNPs with inconsistent genotypes.

Because genotype testing differences between manufacturers could have affected control genotyping frequencies but not case frequencies, we very narrowly limited analyzed SNPs to those with close case and control frequencies within the same manufacturer and between the two manufacturers, within 18% and not exceeding a χ2 P=0.2. The ORs and 95% CIs between full-Jewish subjects genotyped on Illumina only platforms and Illumina plus Affymetrix platforms were similar, with substantially smaller P-values resulting from the combined Illumina plus Affymetrix platforms. This supports the ability of our restricted SNP analysis to provide good genotype comparability, enabling the addition of control subjects from the Affymetrix platform for power only without substantially altering detection of the associations.

While the imputation quality for rs66562280 was high overall (info=0.72), when the imputation quality was examined by platform and ethnicity, it was always higher for Jewish subjects compared to non-Jewish white European subjects (for Jewish subjects info=0.80, 0.86, and 0.87 on the Illumina HumanHap550 Infinium, Human 610-Quad, and HumanOmniExpressExome, respectively, and for non-Jewish white European subjects info=0.71, 0.72, and 0.74 on the same platforms, respectively). In addition to the 1000 Genomes reference panel, TAGC data set was used as a reference panel to increase imputation quality for Jewish subjects, which likely explains the higher imputation quality across Illumina platforms for rs66562280 in Jewish subjects. On the Affymetrix platform, rs66562280 was genotyped. For rs66562280 in Jewish subjects there was high agreement between the control MAF from the Affymetrix platform (MAF=0.32) and the control MAFs from the Illumina platforms (MAF=0.34–0.38). According to the Genome Aggregate Database (gnomAD), a large-scale sequencing project, for rs66562280, the MAF for European subjects (MAF=0.30, N=15470) is appreciably lower than the MAF for Ashkenazi Jewish subjects (MAF=0.49, N=190) (Sherry et al. 2001). For rs66562280 in non-Jewish white European subjects, similar to the gnomAD findings, the MAF was lower in non-Jewish white European subjects compared to Ashkenazi Jewish subjects from the Affymetrix platform (MAF=0.24 vs. MAF=0.32, respectively). However, dissimilar to the gnomAD findings, the MAF was higher in non-Jewish white European subjects compared to Ashkenazi Jewish subjects from the Illumina platforms (MAF=0.41–0.43 vs. MAF=0.34–0.38, respectively). Most likely, the MAF in non-Jewish white European subjects from the Illumina platforms was erroneously high due to the lower imputation quality.

The OR in the PanScan/PanC4/GERA full-Jewish subjects was considerably higher for the association between rs2656937 and risk of pancreatic cancer compared to the OR in the PanScan/PanC4/GERA non-Jewish white European subjects. For rs66562280, the OR was larger in the PanScan/PanC4/GERA non-Jewish white European subjects compared to the full-Jewish subjects; however, as discussed above, the discordant imputed allele frequencies on the Illumina platforms and genotyped allele frequencies on the Affymetrix platform, due to lower imputation quality, probably distort the magnitude of the OR in non-Jewish white Europeans. The association between rs66562280 and risk of pancreatic cancer in non-Jewish white European PanScan/PanC4 subjects (Illumina only platforms) is essentially null. While independent replication of these SNPs is not possible with any known data set, as discussed above for rs66562280, data from the gnomAD large sequencing project report a higher MAF for these top two SNPs in Ashkenazi Jews compared to other races and ethnicities. This further supports the possibility of a more important role for these top two SNPs in Ashkenazi Jews: For rs66562280, the MAF=0.30 (N=15470) in Europeans and the MAF=0.49 (N=190) in Ashkenazi Jews, while for rs2656937, the MAF=0.27 (N=18826) in Europeans and the MAF=0.39 (N=290) in Ashkenazi Jews (Sherry et al. 2001).

Why these SNPs could be functionally relevant in full Jews but not in non-Jewish white Europeans is unclear, but may implicate other pathway components not identified to-date. Nevertheless, the allele frequency differences and larger magnitudes of association in the full Jews suggests that these two SNPs could explain much of the increased risk of pancreatic cancer in the Jewish population compared to the non-Jewish white European population. Combined with BRCA1 and BRCA2 mutations and frequency differences in non-O ABO blood groups, these markers in total are estimated to explain about 60% of the disease risk in Jews, or essentially all of the excess of a 1.8-fold risk compared to white European non-Jews. Witte and colleagues argue that the PAF is a problematic measure since it can give estimates that are much larger than other measures and it can exceed 100% if not calculated properly (Witte et al. 2014). The PAF is often incorrectly applied for large collections of SNPs, with high linkage disequilibrium and no per-se unexposed group. However, in our analysis where two SNPs could explain much of the increased risk in Jewish subjects and 20% of Jews are estimated to be homozygous wildtype for each SNP, use of the PAF furthers the analysis.

As discussed in our previous paper, where the same study data set was examined (Streicher et al. 2017), we are confident that most of the full- or part-Jewish subjects in our study are of Ashkenazi decent, since participants from the sub-studies included here were mainly recruited in the United States, where the overwhelming majority of Jews are of Ashkenazi descent (Raphael M.L. 2009). This is supported by the observation that none of the Jewish subjects seeded into our PCA were located in plot regions associated with north African, sub-Saharan African, or eastern Middle-East or Asian ancestry populations, locales that include non-Ashkenazi Jewish populations.

Finally, some limitations of our work should be considered. While we attempted to obtain data sets with as many pancreatic cancer cases as possible, our full-Jewish pancreatic cancer GWAS was somewhat underpowered to detect associations of more modest magnitude (OR~1.15) and MAF (~0.30). A power analysis with 406 full-Jewish pancreatic cancer cases and 2,332 full-Jewish controls would give a GWAS >70% power to detect an association size of 1.45 for a SNP with a MAF >0.37 (at P=10−7·3) (Skol et al. 2006). In addition, it is possible that additional SNPs are located near rs66562280 and rs2656937 that are not seen on the regional plots, but that are associated with pancreatic cancer in the full-Jewish subjects, since only ~10% of genotyped and imputed SNPs met the QC filter inclusion criteria for SNP discovery in the full-Jewish GWAS. Future studies are needed to replicate the associations between rs66562280 and rs2656937 and risk of pancreatic cancer in a full-Jewish study sample, and to increase the sample size of such subjects in order to detect additional SNPs that may be associated with pancreatic cancer in the increased-risk Jewish population. To date, the 539 full- plus part-Jewish cases from the PanScan and PanC4 studies comprise the largest known Jewish pancreatic cancer patient group, which makes replication impossible with any known data set.

If the two SNPs identified here are confirmed in association with increased risks of pancreatic cancer of the magnitudes seen here, they would address a substantial fraction of the excess risk of this disease observed in Ashkenazi Jews, and may be useful for testing of Jewish individuals with family histories of pancreatic cancer not otherwise explained by known pancreatic cancer predisposition mutations.

Supplementary Material

439_2020_2205_MOESM2_ESM
439_2020_2205_MOESM1_ESM

Acknowledgements

High performance computing (HPC) was supported by HPC facilities operated by, and the staffs of, the Yale Center for Research Computing and Yale’s W.M. Keck Biotechnology Laboratory, as well as US National Institutes of Health grants RR019895 and RR029676, which helped fund the HPC cluster.

The cooperation of 30 Connecticut hospitals, including Stamford Hospital, in allowing patient access is gratefully acknowledged.

Funding:

This work was supported by US National Institutes of Health/National Cancer Institute grant F31 CA177153. The Yale Connecticut Pancreas study was supported by US National Institutes of Health/National Cancer Institute grant R01 CA098870. The Memorial Sloan Kettering Cancer Center Pancreatic Tumor Registry is supported by US National Institutes of Health/National Cancer Institute grant P30 CA008748, the Geoffrey Beene Foundation, the Arnold and Arlene Goldstein Family Foundation, and the Society of the Memorial Sloan Kettering Cancer Center. The Johns Hopkins University study is supported by US National Institutes of Health/National Cancer Institute grants P50 CA062924 and R01 CA097075.

Footnotes

Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.

Conflict of interest/competing interests: The authors declare that they have no conflicts of interest.

Ethics approval: All necessary Institutional Review Board approvals and data use agreements were granted from Yale University and collaborating institutions.

Code availability: The code created for the current study is available from the corresponding author on reasonable request.

Availability of data and material:

The data sets analyzed during the current study are available in the database of Genotypes and Phenotypes (dbGaP): accession phs000206.v4.p3, phs000648.v1.p1, and phs000674.v1.p1).

References

  1. Abu-Alainin W, Gana T, Liloglou T, Olayanju A, Barrera LN, Ferguson R, Campbell F, Andrews T, Goldring C, Kitteringham N, Park BK, Nedjadi T, Schmid MC, Slupsky JR, Greenhalf W, Neoptolemos JP, Costello E (2016) UHRF1 regulation of the Keap1-Nrf2 pathway in pancreatic cancer contributes to oncogenesis. J Pathol 238: 423–433 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. American Cancer Society. (2020) Cancer Facts & Figures. American Cancer Society, Atlanta. [Google Scholar]
  3. Amundadottir L, Kraft P, Stolzenberg-Solomon RZ, Fuchs CS, Petersen GM, Arslan AA, Bueno-de-Mesquita HB, Gross M, Helzlsouer K, Jacobs EJ, LaCroix A, Zheng W, Albanes D, Bamlet W, Berg CD, Berrino F, Bingham S, Buring JE, Bracci PM, Canzian F, Clavel-Chapelon F, Clipp S, Cotterchio M, de Andrade M, Duell EJ, Fox JW Jr., Gallinger S, Gaziano JM, Giovannucci EL, Goggins M, Gonzalez CA, Hallmans G, Hankinson SE, Hassan M, Holly EA, Hunter DJ, Hutchinson A, Jackson R, Jacobs KB, Jenab M, Kaaks R, Klein AP, Kooperberg C, Kurtz RC, Li D, Lynch SM, Mandelson M, McWilliams RR, Mendelsohn JB, Michaud DS, Olson SH, Overvad K, Patel AV, Peeters PH, Rajkovic A, Riboli E, Risch HA, Shu XO, Thomas G, Tobias GS, Trichopoulos D, Van Den Eeden SK, Virtamo J, Wactawski-Wende J, Wolpin BM, Yu H, Yu K, Zeleniuch-Jacquotte A, Chanock SJ, Hartge P, Hoover RN (2009) Genome-wide association study identifies variants in the ABO locus associated with susceptibility to pancreatic cancer. Nat Genet 41: 986–990 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Azevedo-Pouly A, Jiang J, Sutaria D, Schmittgen T. (2013) Implications during pancreatic acinar ductal trans-diferentiation [abstract]. In AACR, Vol Proceedings of the 104th Annual Meeting of the American Association for Cancer Research, Washington, DC. [Google Scholar]
  5. Bailey P, Chang DK, Nones K, Johns AL, Patch AM, Gingras MC, Miller DK, Christ AN, Bruxner TJ, Quinn MC, Nourse C, Murtaugh LC, Harliwong I, Idrisoglu S, Manning S, Nourbakhsh E, Wani S, Fink L, Holmes O, Chin V, Anderson MJ, Kazakoff S, Leonard C, Newell F, Waddell N, Wood S, Xu Q, Wilson PJ, Cloonan N, Kassahn KS, Taylor D, Quek K, Robertson A, Pantano L, Mincarelli L, Sanchez LN, Evers L, Wu J, Pinese M, Cowley MJ, Jones MD, Colvin EK, Nagrial AM, Humphrey ES, Chantrill LA, Mawson A, Humphris J, Chou A, Pajic M, Scarlett CJ, Pinho AV, Giry-Laterriere M, Rooman I, Samra JS, Kench JG, Lovell JA, Merrett ND, Toon CW, Epari K, Nguyen NQ, Barbour A, Zeps N, Moran-Jones K, Jamieson NB, Graham JS, Duthie F, Oien K, Hair J, Grutzmann R, Maitra A, Iacobuzio-Donahue CA, Wolfgang CL, Morgan RA, Lawlor RT, Corbo V, Bassi C, Rusev B, Capelli P, Salvia R, Tortora G, Mukhopadhyay D, Petersen GM, Australian Pancreatic Cancer Genome I, Munzy DM, Fisher WE, Karim SA, Eshleman JR, Hruban RH, Pilarsky C, Morton JP, Sansom OJ, Scarpa A, Musgrove EA, Bailey UM, Hofmann O, Sutherland RL, Wheeler DA, Gill AJ, Gibbs RA, Pearson JV, Waddell N, Biankin AV, Grimmond SM (2016) Genomic analyses identify molecular subtypes of pancreatic cancer. Nature 531: 47–52 [DOI] [PubMed] [Google Scholar]
  6. Bellefroid E, Marine J-C, Ried T, Lecocq P, Riviere M, Amemiya C, Poncelet D, Coulie P, de Jong P, Szpirer C, Ward D, Martiall J (1993) Clustered organization of homologous KRAB zinc-finger genes with enhanced expression in human T lymphoid cells. EMBO J 12: 1363–1374 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Carmi S, Hui KY, Kochav E, Liu X, Xue J, Grady F, Guha S, Upadhyay K, Ben-Avraham D, Mukherjee S, Bowen BM, Thomas T, Vijai J, Cruts M, Froyen G, Lambrechts D, Plaisance S, Van Broeckhoven C, Van Damme P, Van Marck H, Barzilai N, Darvasi A, Offit K, Bressman S, Ozelius LJ, Peter I, Cho JH, Ostrer H, Atzmon G, Clark LN, Lencz T, Pe’er I (2014) Sequencing an Ashkenazi reference panel supports population-targeted personal genomics and illuminates Jewish and European origins. Nat Commun 5: 4835–4843 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Cheng ZX, Sun B, Wang SJ, Gao Y, Zhang YM, Zhou HX, Jia G, Wang YW, Kong R, Pan SH, Xue DB, Jiang HC, Bai XW (2011) Nuclear factor-kappaB-dependent epithelial to mesenchymal transition induced by HIF-1alpha activation in pancreatic cancer cells under hypoxic conditions. PLoS One 6: e23752. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Childs EJ, Chaffee KG, Gallinger S, Syngal S, Schwartz AG, Cote ML, Bondy ML, Hruban RH, Chanock SJ, Hoover RN, Fuchs CS, Rider DN, Amundadottir LT, Stolzenberg-Solomon R, Wolpin BM, Risch HA, Goggins MG, Petersen GM, Klein AP (2016) Association of common susceptibility variants of pancreatic cancer in higher risk patients: A PACGENE study. Cancer Epidemiol Biomarkers Prev 25: 1185–1191 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Coogan PF, Rosenberg L, Palmer JR, Strom BL, Zauber AG, Stolley PD, Shapiro S (2000) Nonsteroidal anti-inflammatory drugs and risk of digestive cancers at sites other than the large bowel. Cancer Epidemiol Biomarkers Prev 9: 119–123 [PubMed] [Google Scholar]
  11. Cui L, Chen J, Zhang Q, Wang X, Qu J, Zhang J, Dang S (2015) Up-regulation of UHRF1 by oncogenic Ras promoted the growth, migration, and metastasis of pancreatic cancer cells. Mol Cell Biochem 400: 223–232 [DOI] [PubMed] [Google Scholar]
  12. Dhont L, Mascaux C, Belayew A (2016) The helicase-like transcription factor (HLTF) in cancer: loss of function or oncomorphic conversion of a tumor suppressor? Cell Mol Life Sci 73: 129–147 [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Eldridge RC, Gapstur SM, Newton CC, Goodman M, Patel AV, Jacobs EJ (2011) Jewish ethnicity and pancreatic cancer mortality in a large U.S. cohort. Cancer Epidemiol Biomarkers Prev 20: 691–698 [DOI] [PubMed] [Google Scholar]
  14. Fleiss JL, Levin B, Paik MC. 2003. The comparison of proportions from several independent samples. In Statistical methods for rates and proportions. Wiley-Interscience, Hoboken, New Jersey. [Google Scholar]
  15. Galinsky KJ, Bhatia G, Loh PR, Georgiev S, Mukherjee S, Patterson NJ, Price AL (2016) Fast principal-component analysis reveals convergent evolution of ADH1B in Europe and East Asia. Am J Hum Genet 98: 456–472 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Gong X, Kaushal S, Neville C, Nguyen T, Clark T, Khatib Z, Valentine M, Look A (1997) Developmental regulation of Zbu1, a DNA-binding member of the SWI2/SNF2 family. Dev Biol 183: 166–182 [DOI] [PubMed] [Google Scholar]
  17. Gong Z, Holly EA, Bracci PM (2011) Survival in population-based pancreatic cancer patients: San Francisco Bay area, 1995–1999. Am J Epidemiol 174: 1373–1381 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Greenwald P, Korns RF, Nasca PC, Wolfgang PE (1975) Cancer in United States Jews. Cancer Res 35: 3507–3512 [PubMed] [Google Scholar]
  19. Haenszel WJ (1961) Cancer mortality among the foreign-born in the United States. J Natl Cancer Inst 26: 37–132 [PubMed] [Google Scholar]
  20. Hamada T, Yuan C, Yurgelun M, Perez K, Khalaf N, Morales-Oyarvide V, Babic A, Nowak J, Rubinson D, Giannakis M, Ng K, Kraft P, Stampfer M, Giovannucci E, Fuchs C, Ogino S, Wolpin B (2019) Family history of cancer, Ashkenazi Jewish ancestry, and pancreatic cancer risk. Br J Cancer 120: 848–854 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Hessmann E, Schneider G, Ellenrieder V, Siveke JT (2016) MYC in pancreatic cancer: novel mechanistic insights and their translation into therapeutic strategies. Oncogene 35: 1609–1618 [DOI] [PubMed] [Google Scholar]
  22. Huang J, Howie B, McCarthy S, Memari Y, Walter K, Min JL, Danecek P, Malerba G, Trabetti E, Zheng HF, Consortium UK, Gambaro G, Richards JB, Durbin R, Timpson NJ, Marchini J, Soranzo N (2015) Improved imputation of low-frequency and rare variants using the UK10K haplotype reference panel. Nat Commun 6: 8111–8119 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Kang DS, Tian X, Benovic JL (2014) Role of beta-arrestins and arrestin domain-containing proteins in G protein-coupled receptor trafficking. Curr Opin Cell Biol 27: 63–71 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Klein AP, Wolpin BM, Risch HA, Stolzenberg-Solomon RZ, Mocci E, Zhang M, Canzian F, Childs EJ, Hoskins JW, Jermusyk A, Zhong J, Chen F, Albanes D, Andreotti G, Arslan AA, Babic A, Bamlet WR, Beane-Freeman L, Berndt SI, Blackford A, Borges M, Borgida A, Bracci PM, Brais L, Brennan P, Brenner H, Bueno-de-Mesquita B, Buring J, Campa D, Capurso G, Cavestro GM, Chaffee KG, Chung CC, Cleary S, Cotterchio M, Dijk F, Duell EJ, Foretova L, Fuchs C, Funel N, Gallinger S, JM MG, Gazouli M, Giles GG, Giovannucci E, Goggins M, Goodman GE, Goodman PJ, Hackert T, Haiman C, Hartge P, Hasan M, Hegyi P, Helzlsouer KJ, Herman J, Holcatova I, Holly EA, Hoover R, Hung RJ, Jacobs EJ, Jamroziak K, Janout V, Kaaks R, Khaw KT, Klein EA, Kogevinas M, Kooperberg C, Kulke MH, Kupcinskas J, Kurtz RJ, Laheru D, Landi S, Lawlor RT, Lee IM, LeMarchand L, Lu L, Malats N, Mambrini A, Mannisto S, Milne RL, Mohelnikova-Duchonova B, Neale RE, Neoptolemos JP, Oberg AL, Olson SH, Orlow I, Pasquali C, Patel AV, Peters U, Pezzilli R, Porta M, Real FX, Rothman N, Scelo G, Sesso HD, Severi G, Shu XO, Silverman D, Smith JP, Soucek P, Sund M, Talar-Wojnarowska R, Tavano F, Thornquist MD, Tobias GS, Van Den Eeden SK, Vashist Y, Visvanathan K, Vodicka P, Wactawski-Wende J, Wang Z, Wentzensen N, White E, Yu H, Yu K, Zeleniuch-Jacquotte A, Zheng W, Kraft P, Li D, Chanock S, Obazee O, Petersen GM, Amundadottir LT (2018) Genome-wide meta-analysis identifies five new susceptibility loci for pancreatic cancer. Nat Commun 9: 556. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Liang JW, Shi ZZ, Shen TY, Che X, Wang Z, Shi SS, Xu X, Cai Y, Zhao P, Wang CF, Zhou ZX, Wang MR (2014) Identification of genomic alterations in pancreatic cancer using array-based comparative genomic hybridization. PLoS One 9: e114616. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Lu Y, Zhu H, Shan H, Lu J, Chang X, Li X, Lu J, Fan X, Zhu S, Wang Y, Guo Q, Wang L, Huang Y, Zhu M, Wang Z (2013) Knockdown of Oct4 and Nanog expression inhibits the stemness of pancreatic cancer cells. Cancer Lett 340: 113–123 [DOI] [PubMed] [Google Scholar]
  27. Ma J, Mi C, Wang K, Lee J, Jin X (2016) Zinc finger protein 91 (ZFP91) activates HIF-1α via NF-κB/p65 to promote proliferation and tumorigenesis of colon cancer. Oncotarget 7: 36551–36562 [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Mack TM, Berkel J, Bernstein L, Mack W (1985) Religion and cancer in Los Angeles county. Cancer Ins Monogr 69: 235–245 [PubMed] [Google Scholar]
  29. MacMahon B (1955) The ethnic distribution of cancer mortality in New York City. Acto Unio Int Contra Cancrum 16: 1716–1724 [Google Scholar]
  30. Marchini J, Howie B (2010) Genotype imputation for genome-wide association studies. Nat Rev Genet 11: 499–511 [DOI] [PubMed] [Google Scholar]
  31. Moldow RE, Connelly RR (1968) Epidemiology of pancreatic cancer in Connecticut. Gastroenterology 55: 677–686 [PubMed] [Google Scholar]
  32. Newill VA (1961) Distribution of cancer mortality among ethnic subgroups of the white population of New York City, 1953–58. J Natl Cancer Inst 26: 405–417 [PubMed] [Google Scholar]
  33. Nowak NJ, Gaile D, Conroy JM, McQuaid D, Cowell J, Carter R, Goggins MG, Hruban RH, Maitra A (2005) Genome-wide aberrations in pancreatic adenocarcinoma. Cancer Genet Cytogenet 161: 36–50 [DOI] [PubMed] [Google Scholar]
  34. O’Connell J, Gurdasani D, Delaneau O, Pirastu N, Ulivi S, Cocca M, Traglia M, Huang J, Huffman JE, Rudan I, McQuillan R, Fraser RM, Campbell H, Polasek O, Asiki G, Ekoru K, Hayward C, Wright AF, Vitart V, Navarro P, Zagury JF, Wilson JF, Toniolo D, Gasparini P, Soranzo N, Sandhu MS, Marchini J (2014) A general approach for haplotype phasing across the full spectrum of relatedness. PLoS Genet 10: e1004234. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Paschke L, Rucinski M, Ziolkowska A, Zemleduch T, Malendowicz W, Kwias Z, Malendowicz LK (2014) ZFP91-a newly described gene potentially involved in prostate pathology. Pathol Oncol Res 20: 453–459 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Petersen GM, Amundadottir L, Fuchs CS, Kraft P, Stolzenberg-Solomon RZ, Jacobs KB, Arslan AA, Bueno-de-Mesquita HB, Gallinger S, Gross M, Helzlsouer K, Holly EA, Jacobs EJ, Klein AP, LaCroix A, Li D, Mandelson MT, Olson SH, Risch HA, Zheng W, Albanes D, Bamlet WR, Berg CD, Boutron-Ruault MC, Buring JE, Bracci PM, Canzian F, Clipp S, Cotterchio M, de Andrade M, Duell EJ, Gaziano JM, Giovannucci EL, Goggins M, Hallmans G, Hankinson SE, Hassan M, Howard B, Hunter DJ, Hutchinson A, Jenab M, Kaaks R, Kooperberg C, Krogh V, Kurtz RC, Lynch SM, McWilliams RR, Mendelsohn JB, Michaud DS, Parikh H, Patel AV, Peeters PH, Rajkovic A, Riboli E, Rodriguez L, Seminara D, Shu XO, Thomas G, Tjonneland A, Tobias GS, Trichopoulos D, Van Den Eeden SK, Virtamo J, Wactawski-Wende J, Wang Z, Wolpin BM, Yu H, Yu K, Zeleniuch-Jacquotte A, Fraumeni JF Jr., Hoover RN, Hartge P, Chanock SJ (2010) A genome-wide association study identifies pancreatic cancer susceptibility loci on chromosomes 13q22.1, 1q32.1 and 5p15.33. Nat Genet 42: 224–228 [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Price AL, Zaitlen NA, Reich D, Patterson N (2010) New approaches to population stratification in genome-wide association studies. Nat Rev Genet 11: 459–463 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Qiu D, Han F, Zhuang H, Li Q, Zhang X (2016) Overexpression of FoxP1 is a novel biomarker of malignant human pancreatic cancer. Int J Clin Exp Med 9: 9054–9063 [Google Scholar]
  39. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM (2014) Projecting cancer incidence and deaths to 2030: the unexpected burden of thyroid, liver, and pancreas cancers in the United States. Cancer Res 74: 2913–2921 [DOI] [PubMed] [Google Scholar]
  40. Raphael ML 2009. The Columbia history of Jews and Judaism in America. Columbia University Press, New York. [Google Scholar]
  41. Rink L, Ochs MF, Zhou Y, von Mehren M, Godwin AK (2013) ZNF-mediated resistance to imatinib mesylate in gastrointestinal stromal tumor. PLoS One 8: e54477. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Risch HA, McLaughlin JR, Cole DE, Rosen B, Bradley L, Fan I, Tang J, Li S, Zhang S, Shaw PA, Narod SA (2006) Population BRCA1 and BRCA2 mutation frequencies and cancer penetrances: a kin-cohort study in Ontario, Canada. J Natl Cancer Inst 98: 1694–1706 [DOI] [PubMed] [Google Scholar]
  43. Risch HA, Yu H, Lu L, Kidd MS (2015) Detectable symptomatology preceding the diagnosis of pancreatic cancer and absolute risk of pancreatic cancer diagnosis. Am J Epidemiol 182: 26–34 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Sandhu S, Wu X, Nabi Z, Rastegar M, Kung S, Ma S, Ding H (2012) Loss of HLTF function promotes intestinal carcinogenesis. Mol Cancer 11: 1–12 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Seidman H (1970) Cancer death rates by site and sex for religious and socioeconomic groups in New York City. Environ Res 3: 234–250 [DOI] [PubMed] [Google Scholar]
  46. Sherry S, Ward M-H, Kholodov M, Baker J, Phan L, Smigielski E, Sirotkin K (2001) dbSNP: the NCBI database of genetic variation. Nucleic Acid Research 29: 308–311 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Skol AD, Scott LJ, Abecasis GR, Boehnke M (2006) Joint analysis is more efficient than replication-based analysis for two-stage genome-wide association studies. Nat Genet 38: 209–213 [DOI] [PubMed] [Google Scholar]
  48. Speir ML, Zweig AS, Rosenbloom KR, Raney BJ, Paten B, Nejad P, Lee BT, Learned K, Karolchik D, Hinrichs AS, Heitner S, Harte RA, Haeussler M, Guruvadoo L, Fujita PA, Eisenhart C, Diekhans M, Clawson H, Casper J, Barber GP, Haussler D, Kuhn RM, Kent WJ (2016) The UCSC genome browser database: 2016 update. Nucleic Acids Res 44: D717–725 [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Streicher S, Klein A, Olson S, Amundadottir L, DeWan A, Zhao H, Risch H (2017) Impact of sixteen established pancreatic cancer susceptibility Loci in American Jews. Cancer Epidemiol Biomarkers Prev 26: 1540–1548 [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Sun H-C, Qiu Z-J, LIU J, Sun J, Jiang T, Huang K-J, Yao M, Huang C (2007) Expression of hypoxia-inducible factor-1 alpha and associated proteins in pancreatic ductal adenocarcinoma and their impact on prognosis. Int J Oncol 30: 1359–1367 [PubMed] [Google Scholar]
  51. The 1000 Genomes Project Consortium (2012) An integrated map of genetic variation from 1,092 human genomes. Nature 491: 56–65 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Unoki M, Okutsu J, Nakamura Y (2003) Identification of a novel human gene, ZFP91, involved in acute muelogenous leukemia. Int J Oncol 22: 1217–1223 [PubMed] [Google Scholar]
  53. Ward LD, Kellis M (2016) HaploReg v4: systematic mining of putative causal variants, cell types, regulators and target genes for human complex traits and disease. Nucleic Acids Res 44: D877–881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Witte JS, Visscher PM, Wray NR (2014) The contribution of genetic variants to disease depends on the ruler. Nat Rev Genet 15: 765–776 [DOI] [PMC free article] [PubMed] [Google Scholar]
  55. Wolpin BM, Rizzato C, Kraft P, Kooperberg C, Petersen GM, Wang Z, Arslan AA, Beane-Freeman L, Bracci PM, Buring J, Canzian F, Duell EJ, Gallinger S, Giles GG, Goodman GE, Goodman PJ, Jacobs EJ, Kamineni A, Klein AP, Kolonel LN, Kulke MH, Li D, Malats N, Olson SH, Risch HA, Sesso HD, Visvanathan K, White E, Zheng W, Abnet CC, Albanes D, Andreotti G, Austin MA, Barfield R, Basso D, Berndt SI, Boutron-Ruault MC, Brotzman M, Buchler MW, Bueno-de-Mesquita HB, Bugert P, Burdette L, Campa D, Caporaso NE, Capurso G, Chung C, Cotterchio M, Costello E, Elena J, Funel N, Gaziano JM, Giese NA, Giovannucci EL, Goggins M, Gorman MJ, Gross M, Haiman CA, Hassan M, Helzlsouer KJ, Henderson BE, Holly EA, Hu N, Hunter DJ, Innocenti F, Jenab M, Kaaks R, Key TJ, Khaw KT, Klein EA, Kogevinas M, Krogh V, Kupcinskas J, Kurtz RC, LaCroix A, Landi MT, Landi S, Le Marchand L, Mambrini A, Mannisto S, Milne RL, Nakamura Y, Oberg AL, Owzar K, Patel AV, Peeters PH, Peters U, Pezzilli R, Piepoli A, Porta M, Real FX, Riboli E, Rothman N, Scarpa A, Shu XO, Silverman DT, Soucek P, Sund M, Talar-Wojnarowska R, Taylor PR, Theodoropoulos GE, Thornquist M, Tjonneland A, Tobias GS, Trichopoulos D, Vodicka P, Wactawski-Wende J, Wentzensen N, Wu C, Yu H, Yu K, Zeleniuch-Jacquotte A, Hoover R, Hartge P, Fuchs C, Chanock SJ, Stolzenberg-Solomon RS, Amundadottir LT (2014) Genome-wide association study identifies multiple susceptibility loci for pancreatic cancer. Nat Genet 46: 994–1000 [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Wynder EL, Mabuchi K, Maruchi N, Fortner JG (1973) Epidemiology of cancer of the pancreas. J Natl Cancer Inst 3: 645–647 [DOI] [PubMed] [Google Scholar]
  57. Zhang M, Wang Z, OBazee O, Jia J, Childs EJ, Hoskins J (2016) Three new pancreatic cancer susceptibility signals identified on chromosomes 1q32.1, 5p15.33 and 8q24.21. Oncotarget 7: 66328–66343 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

439_2020_2205_MOESM2_ESM
NIHMS1612387-supplement-439_2020_2205_MOESM2_ESM.docx (1.9MB, docx)
439_2020_2205_MOESM1_ESM
NIHMS1612387-supplement-439_2020_2205_MOESM1_ESM.docx (64.4KB, docx)

Data Availability Statement

The data sets analyzed during the current study are available in the database of Genotypes and Phenotypes (dbGaP): accession phs000206.v4.p3, phs000648.v1.p1, and phs000674.v1.p1).

RESOURCES